Calcium in Living Cells

Series Editors

Leslie WilsonDepartment of Molecular, Cellular and Developmental Biology

University of California

Santa Barbara, California

Paul MatsudairaDepartment of Biological Sciences

National University of Singapore

Singapore

Academic Press is an imprint of Elsevier30 Corporate Drive, Suite 400, Burlington, MA 01803, USA
525 B Street, Suite 1900, San Diego, CA 92101-4495, USA32, Jamestown Road, London NW1 7BY, UKLinacre House, Jordan Hill, Oxford OX2 8DP, UK
First edition 2010

Copyright � 2010 Elsevier Inc. All rights reserved

No part of this publication may be reproduced, stored in a retrieval systemor transmitted in any form or by any means electronic, mechanical, photocopying,recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology RightsDepartment in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333;email: [email protected]. Alternatively you can submit your request online byvisiting the Elsevier web site at http://elsevier.com/locate/permissions, and selectingObtaining permission to use Elsevier material

NoticeNo responsibility is assumed by the publisher for any injury and/or damage to personsor property as a matter of products liability, negligence or otherwise, or from any useor operation of any methods, products, instructions or ideas contained in the materialherein. Because of rapid advances in the medical sciences, in particular, independentverification of diagnoses and drug dosages should be made

ISBN: 978-0-12-374841-6ISSN: 0091-679X

For information on all Academic Press publicationsvisit our website at elsevierdirect.com

Printed and bound in USA

10 11 12 10 9 8 7 6 5 4 3 2 1

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

David L. Armstrong (183), Membrane Signaling Group, Laboratory of Neurobiology,National Institute of Environmental Health Sciences, NIH, Durham, North Carolina,USA

Darryl A. Auston (113), Center for Biomedical Engineering and Technology, andDepartment of Physiology, University of Maryland School of Medicine, Baltimore,Maryland, USA

S. Baudet (67), Ricerca Biosciences SAS, Saint Germain sur l’Arbresle, FranceD.M. Bers (67), Department of Pharmacology, University of California, Davis, Davis,California, USA

Donald M. Bers (1), Department of Pharmacology, University of California, DavisSchool of Medicine, Davis, California, USA

Francis Burton (225), School of Life Sciences, University of Glasgow, United KingdomChristian Erxleben (183), Membrane Signaling Group, Laboratory of Neurobiology,National Institute of Environmental Health Sciences, NIH, Durham, North Carolina,USA

L. Hove-Madsen (67), Cardiovascular Research Centre CSIC-ICCC, Hospital de laSanta Creu i Sant Pau, Barcelona, Spain

Joseph P.Y. Kao (113), Center for Biomedical Engineering and Technology, andDepartment of Physiology, University of Maryland School of Medicine, Baltimore,Maryland, USA

Eric Karplus (263), Science Wares Inc., Falmouth, Massachusetts, USAOle Johan Kemi (225), School of Life Sciences, University of Glasgow, UnitedKingdom

Gong Li (113), Center for Biomedical Engineering and Technology, and Department ofPhysiology, University of Maryland School of Medicine, Baltimore, Maryland, USA

Mark A. Messerli (91), BioCurrents Research Center, Cellular Dynamics Program,Marine Biological Laboratory, Woods Hole, Massachusetts, USA

Andrew L. Miller (263), Biochemistry and Cell Biology Section and State Key Labora-tory of Molecular Neuroscience, Division of Life Science, HKUST, Clear Water Bay,Kowloon, Hong Kong, PR China

Richard Nuccitelli (1), BioElectroMed Corp., Burlingame, California, USAChris W. Patton (1), Hopkins Marine Station, Stanford University, Pacific GroveCalifornia, USA

Taufiq Rahman (199), Department of Pharmacology, Tennis Court Road, Universityof Cambridge, Cambridge, United Kingdom

ix

x Contributors

Martyn Reynolds (225), Cairn Research Limited, Faversham, Kent, United KingdomKelly L. Rogers (263), The Walter and Eliza Hall Institute of Medical Research,

Parkville, AustraliaGodfrey Smith (225), School of Life Sciences, University of Glasgow, United KingdomPeter J. S. Smith (91), BioCurrents Research Center, Cellular Dynamics Program,

Marine Biological Laboratory, Woods Hole, Massachusetts, USAColin W. Taylor (199), Department of Pharmacology, Tennis Court Road, University

of Cambridge, Cambridge, United KingdomSarah E. Webb (263), Biochemistry and Cell Biology Section and State Key Laboratory

of Molecular Neuroscience, Division of Life Science, HKUST, Clear Water Bay,Kowloon, Hong Kong, PR China

Michael Whitaker (153), Institute of Cell and Molecular Biosciences, Medical School,Newcastle University, Framlington Place, Newcastle upon Tyne, United Kingdom

Jody A. White (183), Membrane Signaling Group, Laboratory of Neurobiology,National Institute of Environmental Health Sciences, NIH, Durham, North Carolina,USA

Robert Zucker (27), Molecular and Cell Biology Department, University of Californiaat Berkeley, Berkeley, California, USA

PREFACE

This volume of Methods in Cell Biology is a sequel to the oft-consulted Volume

40 of the series edited by Richard Nuccitelli that oVered a practical guide to the

study of calcium in living cells. Much in that volume remains relevant and this

volume oVers updates of chapters contributed to the original volume. But in the

decade and a half that have elapsed since the publication of Volume 40, as calcium

signaling has continued to find itself a ubiquitous element of cell regulation, new

technical advances have oVered themselves to the field and existing methods have

been refined.

This volume retains the bedrock of an understanding of calcium buVering and

the manipulation of intracellular free calcium concentration in cells: the subtleties

and peculiarities of an ion that acts at submicromolar concentrations and that is

very actively regulated by cellular buVers and pumps are covered extensively by the

early chapters on calcium buVers; a detailed treatment of dynamic changes in free

calcium achieved by the photosensitive release of calcium from buVers that under-go light-induced changes in calcium aYnity follows on.

Calcium-sensitive electrodes oVer the most quantitative approach to measuring

calcium concentrations within cells and in solutions. Two chapters in this volume

provide a deep understanding of both spatially homogeneous calcium sensing and

of the use of calcium-sensitive electrodes to measure standing fluxes and gradients

of calcium.

Calcium-sensitive fluorescent dyes present some advantages in measuring intra-

cellular free calcium over electrodes—what is lost in precision can be gained in

convenience and time resolution. The two chapters on calcium-sensitive fluores-

cent approaches cover low molecular mass indicators and the newer recombinant

techniques based on green fluorescent protein.

In many circumstances, particularly in studying neuronal calcium signaling in

individual neurones, patch clamp methods are king. Two chapters are devoted to

patch clamp analysis of calcium signaling. One of these concentrates on calcium

channels at the plasma membrane—an approach that remains key to understand-

ing neuronal signaling mechanisms; the other highlights the remarkable achieve-

ment of using patch clamp techniques to study both the aggregate and single

channel properties of calcium release channels in the membranes of intracellular

calcium stores.

The final two chapters of the volume explain the state-of-the-art in imaging

calcium signals. Confocal and multiphoton microscopy have much improved the

spatial and temporal resolution of the measurement of calcium signals, revealing,

among other things, how very localized calcium signals play a part in the versatile

xi

xii Preface

repertoire of this key second messenger. Low-intensity photon imaging using

aequorin has provided an approach best suited to the long-term recording of

calcium signals associated with cell division and pattern formation, situations in

which photobleaching and light-induced damage preclude the use of fluorescent

probes.

I thank all the authors of this volume for having made possible, as I see it, such a

valuable and detailed contribution to the methodological state-of-the-art in the

field. I also thank Zoe Kruze and Narmada Thangavelu of Elsevier for their help

and patience in bringing this volume into being.

Michael Whitaker

CHAPTER 1

METHODS IN CELL BIOLCopyright 2010, Elsevier Inc.

A Practical Guide to the Preparation ofCa2þ BuVers

Donald M. Bers,* Chris W. Patton,† and Richard Nuccitelli‡

*Department of PharmacologyUniversity of California, DavisSchool of Medicine Davis, California, USA

†Hopkins Marine StationStanford University, Pacific GroveCalifornia, USA

‡BioElectroMed Corp.Burlingame, California, USA

A

OGY,All rig

bstract

VOL. 99 0091hts reserved. 1 DOI: 10.1016/S0091

-679X-679X

I. In
troduction II. R ationale
A.
Which Ca2þ BuVer Should You Use? B. EGTA: The Workhorse of Biological Ca2þ Chelators C. BAPTA Family of Ca2þ BuVers
III. M
ethods A. Basic Mathematical Relationships B. Temperature, Ionic Strength, and pH Corrections
IV. M
aterials A. [Ca2þ] Measurement and Calibration Solutions B. Preparing BuVer Solution C. Software Programs
V. D
iscussion and Summary R eferences
Abstract

Calcium (Ca2þ) is a critical regulator of an immense array of biological

processes, and the intracellular [Ca2þ] that regulates these processes is �10,000

lower than the extracellular [Ca2þ]. To study and understand these myriad Ca2þ-dependent functions requires control and measurement of [Ca2þ] in the nano- to

micromolar range (where contaminating Ca2þ is a significant problem). As with

/10 $35.00(10)99001-8

2 Donald M. Bers et al.

pH, it is often essential to use Ca2þ buVers to control free [Ca2þ] at the desired

biologically relevant concentrations. Fortunately, there are numerous available

Ca2þ buVers with diVerent aYnities that make this practical. However, there are

numerous caveats with respect to making these solutions appropriately with

known Ca2þ buVers. These include pH dependence, selectivity for Ca2þ (e.g., vs.

Mg2þ), ionic strength and temperature dependence, and complex multiple equili-

bria that occur in physiologically relevant solutions. Here we discuss some basic

principles of Ca2þ buVering with respect to some of these caveats and provide

practical tools (including freely downloadable computer programs) to help in the

making and calibration of Ca2þ-buVered solutions for a wide array of biological

applications.

I. Introduction

Cell biologists quickly learn how important it is to control the ionic composition

of the solutions used when studying cellular biochemistry, physiology, and phar-

macology. BuVering the pH of the solutions we use is so routine that one can

hardly imagine making a biological solution without the careful selection of the

appropriate pH buVer and measurement of pH in the resulting solution. Indeed,

there are an array of popular zwitterionic amino acid pH buVers introduced by

Good et al. (1966) that are in widespread use (e.g., HEPES) and which complement

the natural physiological pH buVers for these purposes. In contrast, there has been

less attention to buVering and measuring [Ca2þ] because extracellular [Ca2þ] levelsare typically in the millimolar range and such concentrations are easily measured

and prepared. However, intracellular [Ca2þ] ([Ca2þ]i) is quite another matter

because these levels are more typically in the 100 nM–10 mM range which is not

as easily prepared or measured. For example, your source of distilled water could

easily have trace Ca2þ contamination in the range of 1–10 mM. This range of

contaminant Ca2þ can also come from chemicals and biochemicals commonly

used to make solutions. Additionally, there is often a considerable amount of

endogenous Ca2þ in biological tissue or cell samples which is not easily removed

or controlled. Therefore, when we are interested in studying intracellular reactions,

Ca2þ buVering is extremely important.

In this chapter, we will present a practical guide to the preparation of Ca2þ

buVer solutions. Our goal is to emphasize the methods and important variables to

consider while making the procedure as simple as possible. We will also introduce

computer programs which may be of practical use to many workers in this field.

One is a spreadsheet useful in making and validating simple Ca2þ calibration

solutions. The others are more powerful and extensive programs for the calcula-

tion of [Ca2þ] (and other metals and chelators) in complex solutions with multiple

equilibria. These programs have been developed and described with maximum ease

of use in mind.

Table IMixed stability co

Ca2þ buVera

CDTA

EGTA

Quin 2

BAPTA

Fura-2

Dibromo-BAPTA

4,40-Difluoro-BAPTA

Nitr-5 photolyzed

5-Methyl-50-nitro-BA5-Mononitro-BAPTA

NTA

ADA

Citrate

5,50-Dinitro-BAPTA

aAbbreviations: C

N-N-N0-N0-tetraacetiADA, acetamidomin

bMeasured at pH

1. A Practical Guide to the Preparation of Ca2þ BuVers 3

II. Rationale

A. Which Ca2þ BuVer Should You Use?

When selecting the appropriate Ca2þ buVer for your application, the main

consideration is to choose one with a dissociation constant (Kd) close to the desired

free [Ca2þ]. The ability of a buVer to absorb or release ions and thus to hold the

solution at a given concentration of that ion is greatest at its Kd. Just as you should

not choose PIPES (pKa¼6.8) to buVer a solution at pH 7.8, choosing a Ca2þ buVerwith a Kd far from the desired [Ca2þ] set point is a mistake. As a rule of thumb, the

buVer’s Kd should not lie more than a factor of 10 from your desired [Ca2þ].In addition, the buVer should exhibit a much greater aYnity for Ca2þ than Mg2þ

since intracellular [Mg2þ] is typically 10,000-fold higher than [Ca2þ]i. Fortunately,about a dozen suitable buVers are available spanning the range from 10 nM to

100 mM (Table I). There are also a large number of fluorescent Ca2þ indicators

(see Chapter 5) that can also serve as Ca2þ buVers, giving one the opportunity to

both buVer and measure free [Ca2þ] with the same reagent. We will not focus on

nstants for useful Ca2þ buVers at 0.15 M ionic strength in order of Ca2þ aYnity

log K0Ca Kd

K0Ca (pH 7.4)/

K0Ca (pH 7.0) K0

Ca /K0Mg (pH 7.4) References(pH 7.4)

7.90 13 nM 2.7 120 Martell and Smith (1974, 1977),

Bers and MacLeod (1988)

7.18 67 nM 6.2 72,202 Martell and Smith (1974, 1977),


6.84 144 nM 1.15 25,114 Tsien (1980)

6.71 192 nM 1.14 158,244 Tsien (1980)

6.61 242 nM 1.14 72,373 Grynkiewicz et al. (1985)

5.74 1.83 mM 1.02 63,000 Tsien (1980)

5.77 1.7 mMb – – Pethig et al. (1989)

5.2 6.3 mMb – – Tsien and Zucker (1986)

PTA 4.66 22 mMb – – Pethig et al. (1989)

4.4 40 mMb – – Pethig et al. (1989)

3.87 134 mM 2.5 8 Martell and Smith (1974, 1977),


3.71 191 mM 1.24 32 Nakon (1979)

3.32 471 mM 1.03 1.3 Martell and Smith (1974, 1977),


2.15 7 mMb – – Pethig et al. (1989)

DTA, cyclohexilinedinitrilo-N-N-N0-N0-tetraacetic acid; EGTA, Ethylene glycol bis (b-aminoethylester)

c acid; BAPTA, 1,2-bis(o-aminophenoxy)ethane-N-N-N0-N0-tetraacetic acid; NTA, nitriloacetic acid;

odiacetic acid.

7 and 0.1 M ionic strength.


these fluorescent indicators here, but they can be substituted for the buVers de-

scribed (especially when the fundamental binding properties have been measured).

Ethylene glycol bis(b-aminoethylether)-N,N,N0,N0-tetraacetic acid (EGTA) is one

of the best-known Ca2þ buVers, and it can be a reliable buVer in the range of

10 nM–1 mM [Ca2þ] at the typical intracellular pH of 7.2. However, if your goal is

tomake buVers in the 1–10 mMrange, BAPTA (1,2-bis(o-aminophenoxy)ethane-N,

N,N0,N0-tetraacetic acid) or dibromo-BAPTA (Br2-BAPTA) would be better

choices.

B. EGTA: The Workhorse of Biological Ca2þ Chelators

By far the most popular Ca2þ buVer has been EGTA. This molecule has been

used extensively because its apparent dissociation constant (Kd) at pH 7 (0.4 mM) is

close to intracellular Ca2þ levels and it has a much higher aYnity for Ca2þ than for

Mg2þ (�100,000 times higher around neutral pH). However, the preparation of

Ca2þ buVers using EGTA is complicated by the strong pH dependence of its Ca2þ

aYnity (see Fig. 1 and Table I). Thus, while the free [Ca2þ] would be about 400 nM

when EGTA is half saturated with Ca2þ at pH 7, the free [Ca2þ] in this same

solution would decrease by nearly 10-fold to 60 nM by simply raising the pH to

7.4! Therefore, the pH of Ca2þ buVers made with EGTA must be very carefully

controlled, and the calculation of the appropriate amounts of EGTA and Ca2þ to

use must be made at the desired pH. The purity of the EGTA is also a variable that

can cause substantial errors, as large as 0.2 pCa units in the free [Ca2þ] (Bers, 1982;Miller and Smith, 1984).

6 7 84

5

6

7

8

9 EGTA

BAPTA

Br2-BAPTA

0.001

0.01

0.1

1

10

pH

Log

K� C

a (a

ppar

ent C

a2+ a

ffini

ty)

Fre

e [C

a2+] (mM

) fo

r1

Ca

: 2 li

gand

rat

io

Fig. 1 The pH dependence of apparent aYnities (K0Ca) for EGTA, BAPTA, and Br2-BAPTA at 20 �C

and 150 mM ionic strength.


There are many papers explaining how to calculate the proper amounts of

EGTA and Ca2þ that must be combined to obtain a given free [Ca2þ] (some are

listed below). Due to the steep pH dependence and slightMg2þ sensitivity, both pH

and Mg2þ must be considered in the calculation and it is best accomplished by

computer. We provide a program for such calculations and describe it below.

Systematic errors in EGTA purity and pH can be a real practical problem (Bers,

1982), even with the best calculations for solution preparation. Thus, we also

recommendmeasuring the free [Ca2þ] whenever possible (see below andChapter 3).

C. BAPTA Family of Ca2þ BuVers

Roger Tsien developed an analogue of EGTA in which the methylene links

between oxygen and nitrogen atoms were replaced with benzene rings to yield a

compound called BAPTA (Tsien, 1980; Fig. 2). This compound exhibits a much

lower pH sensitivity and much higher rates of calcium association and dissocia-

tion. These characteristics are mainly due to the fact that BAPTA is almost

completely deprotonated at neutral pH. Moreover, modifications of BAPTA

have been made to provide Ca2þ buVers with a range of Kd values covering the

biologically significant range of 0.1 mM–10 mM (see Table I; Pethig et al., 1989).

However, one disadvantage compared with EGTA is that the BAPTA family of

buVers exhibits a greater ionic strength dependence (see Figs. 3–5). In particular,

increasing ionic strength from 100 to 300 mM decreases the apparent aYnity

constant, K0Ca for BAPTA or Br2-BAPTA by almost threefold, whereas the

COO–

COO–

COO–

COO– –OOC

–OOC

–OOC

–OOC

N

NN

X XX = H; BAPTA

X = Br; Br2-BAPTA

N

OO

O O

EGTA

Fig. 2 Structural formulas for the Ca2þ chelators EGTA (top) and BAPTA and Br2-BAPTA

(bottom).

1 8 15 22 29 361

1.5

2

2.5

3

Temperature (�C)

Data

Prediction

K� C

a (in

106

M−1

)K

� Ca

(in 1

06 M

−1)

0 0.05 0.1 0.15 0.2 0.25 0.30.7

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

Ionic strength (M)

Data

Prediction

A

B

Fig. 3 EGTA apparent Ca2þ aYnity (K0Ca) is influenced by temperature (A) and ionic strength (B).

The experimental data in A is from Harrison and Bers (1987) at pH 7.00 and 0.19 M ionic strength and

in B from Harafuji and Ogawa (1980) at pH 6.8 and 22 �C. Predicted values are based on the

temperature and ionic strength corrections described in the text.


EGTA aYnity is only reduced by about 30%. In contrast, raising temperature from

1 to 36 �C approximately doubles the apparent aYnity of all three of the Ca2þ

buVers shown in Figs. 3–5 (i.e., EGTA, BAPTA, and Br2-BAPTA).

III. Methods

A. Basic Mathematical Relationships

From the forgoing and the data shown in Figs. 1 and 3–5, it is clear that one

needs to know quantitatively how the buVers being used are altered by the typical

range of experimental conditions (e.g., pH, temperature, and ionic strength). While

1 8 15 22 29 362

2.5

3

3.5

4

4.5

5

Temperature (�C)

ΔH = 0

Data

Prediction

A

B

0.1 0.15 0.2 0.25 0.30

1

2

3

4

5

6

7

Ionic strength (M)

Data

Prediction

K� C

a (in

106

M−1

)K

� Ca

(in 1

06 M

−1)

Fig. 4 BAPTA apparent Ca2þ aYnity (K0Ca) is influenced by temperature (A) and ionic strength (B).

The experimental data is fromHarrison and Bers (1987) at pH 7.00 and 0.19 M ionic strength (A) and at

pH 7.00 and 22 �C (B). Predicted values are based on the temperature and ionic strength corrections

described in the text.


we do not want to belabor the equations, it may be useful for some readers if we lay

out some of the basics. If you are not interested in the equations, you can ignore

this section and the next (and still use the programs as more of a black box). We

hope we have accounted for things as well as possible.

In the sections above, we used Kd to talk about Ca2þ aYnity. That Kd was the

apparent overall dissociation constant, which we will get back to below (see

Eq. (5)). It is more traditional to set out the mathematical expressions starting

with the simple definition of the Ca2þ association constant KCa

KCa ¼ ½CaR�Ca�½R�½ ð1Þ

1 8 15 22 29 362

3

4

5

6

Temperature (�C)

A

B

Data

Prediction

0.1 0.15 0.2 0.25 0.30

1

2

3

4

5

6

7

8

Ionic strength (M)

Data

Prediction

K� C

a (in

105

M−1

)K

� Ca

(in 1

05 M

−1)

Fig. 5 The eVect of temperature (A) and ionic strength (B) on the apparent Ca2þ aYnity (K0Ca) of Br2-

BAPTA. The experimental data is from Harrison and Bers (1987) at pH 7.00 and 0.19 M ionic strength

(A) and at pH 7.00 and 22 �C (B). Predicted values are based on the temperature and ionic strength

corrections described in the text.


where R is the Ca2þ buVer. This expression is not too useful directly, because we do

not know any of the variables on the right side. It is generally more useful to have

[Ca2þ] or bound Ca2þ ([CaR]) in terms of known quantities, like total Ca2þ ([Cat])

or total chelator ([Rt]). One of the complicating factors is also that Ca2þ buVerslike EGTA or BAPTA exist in multiple unbound forms in diVerent states of

protonation. Then for a tetravalent Ca2þ buVer like EGTA, the total of the non-

Ca2þ bound forms of the buVer is

Rt½ � � CaR� ¼ ½R� þ ½HR� þ ½H2R� þ ½H3R� þ ½H4R�� ð2Þ


where we have omitted valency for simplicity. R (or R4�) is the form which binds

Ca2þ most avidly and it is convenient to transform Eq. (1) to one with an apparent

aYnity constant for Ca2þ (K0Ca) for a given pH

K0Ca ¼

½CaR�½Ca�½R� �

½R�½Rt� � ½CaR� ð3Þ

or using Eqs. (1) and (2)

K0Ca ¼ KCa

R½ �R½ � þ HR½ � þ H2R½ � þ H3R½ � þ H4R½ � ð4Þ

Then it is a simple matter to show that

K0Ca ¼

KCa

1þ Hþ½ � KH1½ � þ Hþ½ �2KH1KH2 þ Hþ½ �3KH1KH2KH3 þ Hþ½ �4KH1KH2KH3KH4

ð5Þwhere KH1–KH4 are the four acid association constants for the buVer. Now if we

know KCa, the pH, and KH1–KH4, we can calculate K0Ca. This K0

Ca is thus the

apparent aYnity for a given [H] where pH¼� log10 ([H]/gH), and gH is the activity

coeYcient for protons under the experimental conditions (see below). This K0Ca is

the reciprocal of the dissociation constant, Kd discussed in the previous section.

Eq. (3) can also be manipulated to yield

CaR½ �= Ca½ � ¼ K0Ca Rt½ � �K

0Ca CaR½ � ð6Þ

which is the linearization for Scatchard plots of Bound/Free ([CaR]/[Ca]) versus

Bound ([CaR], where slope¼–K0Ca and x-intercept¼ [Rt]). One can also solve for

[CaR] obtaining the familiar Michaelis–Menten form.

½CaR� ¼ ½Rt�1þ 1= K

0Ca Ca�Þ½� ð7Þ

Solving for free [Ca] is more complicated because we do not know [CaR] a priori,

but substituting [CaR]¼ [Cat]� [Ca] we can get a quadratic solution

Ca½ �2 þ ð½Rt� � ½Cat� þ 1=K0CaÞ½Ca� � Cat½ �=K 0

Ca ¼ 0 ð8ÞSimilar equations can be developed for Ca2þ binding to the protonated form (e.g.,

H-EGTA) which also binds Ca2þ with a lower aYnity (e.g., see Harrison and Bers,

1987). For example, when we include Ca2þ binding to the singly protonated form

of EGTA (or HR3�) the following term must be added to the apparent aYnity

expression on the right-hand side of Eq. (5)

KCa2

1=ð Hþ½ �KH1Þ þ 1þ Hþ½ �KH2 þ Hþ½ �2KH2KH3 þ ½Hþ�3KH2KH3KH4

ð9Þ


where KCa2 is the Ca2þ association constant for the chelator in the singly proto-

nated form, HR. This provides some basics of the relationships for a single

chelator. However, more complicated solutions have multiple equilibria (e.g.,

other cations that bind EGTA and other Ca binding moieties) which cannot be

readily solved simultaneously in an analytical manner.

It should, however, be noted that it is simpler to go from free [Ca2þ] to [Cat],

especially with no Ca2þ competitors. This is because all of the chelators which

might bind Ca2þ will be in equilibrium with the same free [Ca2þ]. Thus, one couldsimply use a series of equations like Eq. (7) for diVerent chelators if you know the

values on the right-hand side. Then you can simply add up free [Ca2þ] plus the[CaR] values from the chelators to obtain the [Cat]. If free [Ca2þ] is not known(or chosen) it requires multiple versions of equations like Eq. (8) to be solved

simultaneously. Thus, iterative computer programs are useful (see below).

B. Temperature, Ionic Strength, and pH Corrections

While the above explains the theoretical basis for calculating the pH eVect onK0

Ca, we should clarify how we normally correct for temperature, ionic strength,

and pH for the experimental conditions used. Again, those not interested in the

details can skip this section. Thus, the final apparent aYnity (or K0Ca) should

include correction for temperature and ionic strength as well as pH. Indeed, both

proton aYnity (KH1–KH4) and metal aYnity constants (e.g., KCa) should be ad-

justed for the experimental temperature and ionic strength before adjusting for pH

as above.

1. Temperature Corrections

The standard way to correct equilibrium constants for changes in temperature

depends on knowledge of the enthalpy (DH) of the reaction.

log10K0 ¼ log10K þ DH 1=T � 1=T

0� �

= 2:303� Rð Þ ð10Þ

where temperature, T is in �K, DH is in kcal/mol and R is 1.9872�10�3 kcal/

(mol�K). Unfortunately, the DH values are not known for all the constants we

might like. For example, for EGTA they are known for the first two acid associa-

tion constants (KH1 and KH2) and the higher aYnity Ca2þ constant (KCa1). This is

generally suYcient for calculations with EGTA (see Fig. 3A). However, no DHvalues have been reported for individual constants for BAPTA and Br2-BAPTA.

Harrison and Bers (1987) measured the temperature dependence of the apparent

K0Ca for BAPTA and Br2-BAPTA. We have fit that data, varying the value of the

DH for KCa. This is somewhat empirical because there is likely to also be tempera-

ture dependence of KH1–KH4. However, the data was well described using DHvalues (for KCa) of 4.7 and 5.53 kcal/mol for BAPTA and Br2-BAPTA, respectively

(see Figs. 4 and 5). Also, since BAPTA and Br2-BAPTA are almost completely


unprotonated already at neutral pH (see Fig. 1), the adjustments toKH1 and KH2 are

less important than with EGTA.

However, it should be noted that one cannot simply use the DH values reported

by Harrison and Bers (1987) for the overall K0Ca of BAPTA and Br2-BAPTA (3.32

and 4.04 kcal/mol) as suggested byMarks andMaxfield (1991). That is because the

intrinsic eVect of increasing temperature on the K0Ca (with DH¼0) is to reduce the

K0Ca (due to the intrinsic temperature dependence of the ionic strength adjustment,

see Fig. 4A and below). Consequently, the apparent overall DH for K0Ca (3.32 for

BAPTA) is smaller than the actual DH for KCa required (our estimate is 4.7 kcal/

mol).

Additionally, Harrison and Bers (1987) found the K0Ca for Br2-BAPTA to be

somewhat higher than the value predicted by the initial values reported by Tsien

(1980). We find that using a slightly higher KCa (log KCa¼6.96 rather than 6.8)

allowed a considerably better fit to the array of experimental data shown in Fig. 5.

2. Ionic Strength Corrections

Ionic strength can also dramatically alter the K0Ca (see Figs. 3–5). We use the

procedure described by Smith and Miller (1985) with ionic equivalents (Ie) rather

than formal ionic strength (Ie¼0.5SCi|zi|, where Ci and zi are the concentration

and valence of the ith ion). We will use the terms equivalently here. Then the

expression used to adjust for ionic strength is

log10K0 ¼ log10K þ 2xyðlog10fj � log10f

0jÞ ð11Þ

where K0 is the constant after conversion, K is the constant before, x and y are the

valences of cation and anion involved in the reaction. The terms log10fj and log10fj0are adjustment terms related to the activity coeYcients for zero ionic strength and

desired ionic strength, respectively. To adjust for ionic strength:

log10fi ¼AIe

1=2

1þ Ie1=2

� bIe ð12Þ

where b is a constant (0.25). A is a constant which depends on temperature and the

dielectric constant of the medium (e)

A ¼ 1:8246� 106

eTð Þ3=2ð13Þ

whereT is the absolute temperature (�K) and e is the dielectric constant for water. Thedielectric constant is temperature dependent and can be found from tables, but the

following equation provides an excellent empirical description over the range 0–50 �C.

e ¼ 87:7251þ 0:3974762T þ 0:0008253T 2 ð14Þ


where T is in �C. Thus, there is some intrinsic temperature dependence in the ionic

strength adjustment itself (see Fig. 4A, broken line). These corrections provide a

reasonably good description of the influence of ionic strength on theK0Ca inFigs. 3–5.

a. Activity CoeYcient for ProtonsThe association constants as usually reported (e.g., in Martell and Smith, 1974,

1977) are often called stoichiometric (or concentration) constants. These terms are

sensible because they imply (correctly) that they are to be used with concentrations

or stoichiometric amounts in chemical equilibria (e.g., as in Eq. (1)). While we

routinely talk about ion concentrations in ‘‘concentration’’ or ‘‘stoichiometric’’

terms, the usual exception is pH (where pH¼� log Hydrogen ion activity or

10�pH¼aH¼gH[Hþ]).

Thus, one can simply convert pH to [Hþ] and go ahead using the ‘‘stoichiomet-

ric’’ constants at face value. That is, then everything is in concentration terms and

not activity. This is the way we have done it in our programs.

The alternative is to change the stoichiometric constants to ‘‘mixed’’ constants

(for proton interactions, or KH1–KH4 only). Then you can still use pH (or 10�pH

rather than 10�pH/gH) in your calculations. Thus, acid association constants (KH1–

KH4) should be divided by the value of gH. Then you can multiply the constant by

the proton activity (since they are always of the same order in the equations (see

Eq. (5))). That is to say that [Hþ]KH1¼ ([Hþ]gH)(KH1/gH), where [Hþ]gH¼10�pH.

This method seems a bit more awkward, but the result is the same.

The proton activity coeYcient, gH varies with both temperature and ionic

strength. The empirical relationship we devised to describe this relationship is the

following

gH ¼ 0:145045� exp �B� Ieð Þ þ 0:063546� exp �43:97704� Ieð Þ þ 0:695634

ð15Þwhere B¼0.522932�exp(0.0327016�T)þ4.015942 and Ie is ionic strength and T

is temperature (in �C). This gives very good estimates of gH from 0 to 40 �C and

from 0 to 0.5 M ionic strength. This expression was sent to Alex Fabiato for use in

his computer program (Fabiato, 1991). While there is a typographical error in text

(the first coeYcient was erroneously 1.45045), the correct expression is in the

program as it was distributed.

IV. Materials

A. [Ca2þ] Measurement and Calibration Solutions

1. Measuring [Ca2þ]

While we can calculate the free [Ca2þ] or [Cat] for our solutions with the

computer programs to be described below, there are still many potential

sources of error (e.g., contaminant Ca2þ, systematic errors in pH, impurities in


chemicals, etc.). Thus, it is valuable to measure the free [Ca2þ] to check that the

solutions are as you expected (especially for complex solutions). Ca2þ sensitive

electrodes are a convenient way to do this (see Chapter 3). We normally use

Ca2þ minielectrodes (as described in Chapter 3) or commercial macroelectrodes.

Both can be connected to a standard pH meter, but it is best to have a meter

which can read in increments of 0.1 mV. We have had good luck with Orion brand

Ca2þ-electrodes and they can be stable for 6 months or so. However, they are

rarely as good as the home-made minielectrodes. These minielectrodes are very

easy to make and are sensitive to changes in free [Ca2þ] down to 1 nM or beyond.

They do not last as long as commercial macroelectrodes, but they are extremely

cheap to make (per electrode) and can be discarded if they get contaminated with

protein or are exposed to radioactive molecules. One can also use fluorescent

indicators, once suitably calibrated, in an analogous way. The only disadvantage

there is the more limited dynamic range of these Ca2þ indicators (10-fold

above and below the Kd) versus electrodes which can give linear responses over

the 10 nM–1 M range.

2. Spreadsheet for Calibration Calculations

Making up calibration solutions for Ca2þ-electrodes (or fluorescent indicators)is really a simpler version of the multiple equilibria problem which will be discussed

below (with respect to MaxChelator), because we really only need to consider the

Ca2þ-EGTA buVer system. This approach is based on the paper by Bers (1982).

This method has the following general steps:

1. Calculate how much total Ca2þ (or free [Ca2þ]) is required for the desired

solutions (using known constants, corrected as above). All solutions should have

the same dominant ionic constituents as the solutions to be measured (e.g.,

140 mM KCl, 10 mM HEPES).

2. Measure the free [Ca2þ] with a good quality Ca2þ electrode compared to free

[Ca2þ] standards without EGTA (at higher [Ca2þ] where [Ca2þ] is more easily

controlled).

3. Accepting (for the moment) that the values from the electrode are all correct,

allows the calculation of bound Ca2þ ([CaR]) from free [Ca2þ] and total [Ca2þ].4. Scatchard plot analysis allows the independent measurement of the apparent

K0Ca and total [EGTA] in your solutions and experimental conditions (even with

systematic errors). Note that the Scatchard plot is very sensitive and deviates from

linearity at very low [Ca2þ] where Ca2þ-electrodes can become sub-Nernstian in

response (see Figs. 6 and 7).

5. Using these ‘‘updated’’ values of total [EGTA] and K0Ca you can recalculate

the free [Ca2þ] in the solutions. Then you can either use the free [Ca2þ] predictedfrom the electrode directly or you can recalculate from the total [Ca2þ] and

Ca calibration For entry of pCaSolution conditions K �Ca calculation (see A32..G47) Regression analysis (see I29-K35)

7.2 pH B= 5.12535712 Intermed 4.954 [EGTA]tot (mM)0.15 M ionic equiv (0.5*sum |zi|Ci) Gamma H= 0.76295887 H activity coefficient 0.9980 r^2

23 � C Log [H] = −7.0825011 Range for linear regression for scatchard5 mM EGTA [H] (M) = 8.27E − 08 should be linear electrode/scatchard slope

500 ml bottle K �Ca

Assn= 6.363E + 06 Log K �Ca

= 6.80365 6.86839 = log K �Ca

from scatchardK

d= 1.57E-07 M or 0.1572mM K �

Ca Discn M= 1.572E − 07 1.35E−07 = K �

Ca dissociation from scatchard

" (nM)= 157.2 135.4 nMInitial Ca-free Ca-total ml 100 mM V-Ca Ca-free Ca-free Ca-bound B/F Regresn InterpCa (nM) (mM) CaCl

2 (mV) (M) (nM) (mM) line B/F mediate (nM) pCa1 8.5000 3.162 0.099 0.493 −152 5.38E − 09 5.383 0.099 18319.405 35857.379 −4.86E − 03 2.75 8.5612 8.0000 10 0.299 1.496 −142.6 1.14E − 08 11.355 0.299 26341.654 34376.672 −4.65E − 03 8.70 8.0603 7.5000 31.623 0.838 4.188 −131.5 2.74E − 08 27.411 0.838 30553.384 30400.143 −4.12E − 03 27.55 7.5604 7.0000 100 1.944 9.722 −117.1 8.60E − 08 85.997 1.944 22608.359 22226.140 −3.01E − 03 87.48 7.0585 6.5000 316.228 3.340 16.701 −101.4 2.99E − 07 299.131 3.340 11165.474 11917.990 −1.61E − 03 280.25 6.5526 6.0000 1000 4.322 21.609 −85 1.10E − 06 1099.966 4.321 3928.010 4674.610 −6.32E − 04 924.58 6.0347 5.5000 3162.278 4.766 23.831 −70.4 3.51E − 06 3506.127 4.763 1358.415 1409.423 −1.87E − 04 3381.53 5.4718 5.0000 10000 4.932 24.662 −51.9 1.52E − 05 15232.059 4.917 322.822 268.545 −2.13E − 05 17,312.44 4.7629 4.5000 31622.777 5.007 25.034 −38.1 4.56E − 05 45563.879 4.961 108.884 −55.893 5.30E − 05 63,647.15 4.196

10 3.0000 1,000,000 5.999 29.995 0 9.38E − 04 938455.736 5.061 5.392 −790.218 1.05E − 03 1,046,090.19 2.98011 3.0000 1,000,000 5.999 29.995 0 9.38E − 04 938455.736 5.061 5.392 −790.218 1.05E − 03 1,046,090.19 2.98012 3.0000 1,000,000 5.999 29.995 0 9.38E − 04 938455.736 5.061 5.392 −790.218 1.05E − 03 1,046,090.19 2.980

2.0 10 mM 10 mM 28.2 From regression3.0 1mM 1mM 0.8 [EGTA]tot K-Ca-EGTA4.0 100mM 100mM −29.6 4.953608052 mM 7.38567143 × 10^6/M

Avg slope= 28.9 99.07% % pure 6.868389983 = log KSlope (mV)= 29 Regression I/J14-: 21 Regression I/J13-: 21

mV offset at 1 mM Ca= 0.8 Slope B/F intercept Slope B/F intercept-7385.67143 36585.72 −6516.46 32752.75

Temperature and ionic strength correction SE of coeff 135.41 548.75 435.32 1663.88Std cond Ionic str Final Final R^2 0.9980 571.40 0.9697 2324.14

Temp 20 incl T eff 23� C F stat 2,975 6 224 7I-Eq 0.100 0.150 0.150 M Delta H Valence Reg sum Sq 9.71E + 08 1.96E + 06 1.21E + 09 3.78E + 07Stoich Log K Log K � Log K � K � (M) kcal/mol 2*x*yconstK1 9.47 9.3576 9.3138 2.060E + 09 −5.8 8K2 8.85 8.7657 8.7219 5.271E + 08 −5.8 6K3 2.66 2.6038 2.6038 4.016E + 02 0 4K4 2 1.9719 1.9719 9.374E + 01 0 2KCa 10.97 10.7453 10.6840 4.831E + 10 −8.1 16KCa2 5.3 5.1315 5.1315 1.353E + 05 0 12Log f 0.109225 0.12327007Temp 293 296A 0.507424 0.51006648Epsil 80.1057 79.0197311

Recalculated

−200

−160

−120

−80

−40

0

40

23456789

Ele

ctro

de r

esp

(mV

)

pCa

Electrode calibration

Initial pCa

Recalculated

No EGTA

−1.E + 04

0.E + 00

1.E + 04

2.E + 04

3.E + 04

4.E + 04

0.0 2.0 4.0 6.0

Bou

nd/fr

ee

Bound (mM)

Scatchard plot

Data

Regression

Fig. 6 Excel spreadsheet used to prepare Ca2þ calibration buVers using a Ca2þ electrode. This version is used when you want to start

with the pCa of the calibration solutions as input and determine howmuch total Ca2þ is needed to achieve the desired free [Ca2þ]. It alsoallows updating of the apparentK0

Ca and free [Ca2þ] in the calibration solutions. This and related spreadsheets can be freely downloaded

(see text for details).

0 1 2 3 4 50

40,000

30,000

20,000

10,000

Regression line

Bou

nd/F

ree

(mM

/mM

)

Bound Ca2+-EGTA (mM)

3 4 5 6 7 8 9

−150

−100

−50

0

pCa

Ele

ctro

de r

espo

nse

(mV

)

Original pCaElectrode pCaRecalc. pCa

A B

Fig. 7 Scatchard plot (A) and electrode calibration curves (B) for the spreadsheet shown in Fig. 6. The

Scatchard plot allows estimation of the total [EGTA] (x-intercept) and the apparent association

constant, K0Ca (-slope). The Scatchard plot is very sensitive to the detection limit of the Ca2þ electrode.

The leftmost two points in A are the lowest free [Ca2þ] in the calibration curve in B (and are not included

in the regression). The three calibration curves shown are for the original (or planned pCa), the pCa

predicted solely by the electrode and the pCa after recalculation, using the values determined in the

Scatchard plot along with the total Ca2þ added to the buVers. In this instance, there was good agreement

between the three curves, but this is not always the case (see Bers, 1982).


updated constants. The latter is necessary for the lowest free [Ca2þ] where the

electrode response is becoming nonlinear (�pCa 9).

We use a spreadsheet (Excel) to greatly simplify all of these steps (see Fig. 6).

There are three basic versions of this spreadsheet: one for starting with free

[Ca2þ] as the input (DMB-CAF-2010.xls), one for pCa as input (DMB-PCA-

2010.xls), and one for total Ca2þ as input (DMB-CAT-2010.xls). These can be

freely downloaded from the MaxChelator site as described below. We will walk

you through the use of this spreadsheet in making a series of free [Ca2þ]standards here.

The fields for the input of data are shaded dark gray. For the pCa version of the

spreadsheet in Fig. 6 you proceed as follows (the others versions are completely

analogous):

1. Enter your solution conditions (upper left, pH, ionic strength, temperature,

total [EGTA], and bottle size you will use). The K0Ca values are then automatically

adjusted for the selected temperature, pH, and ionic strength (lower left box).

2. Enter the desired pCa values. The free [Ca2þ], total [Ca2þ], and ml of 100 mM

Ca2þ stock are automatically calculated (next three columns) using the adjusted

K0Ca.


3. Enter the mV readings from a Ca2þ electrode (including values for Ca2þ

standards lacking EGTA at 100 mM, 1 mM, and 10 mM [Ca2þ] and the electrode

reading at 1 mM free [Ca2þ] as the ‘‘oVset’’). This is the fourth column (V-Ca) and

you can choose the electrode slope (rather than assume the average). The free

[Ca2þ], Ca2þ-bound to EGTA, and the bound/free (B/F) are then automatically

calculated (based on the electrode response and total [Ca2þ]).4. Those calculated values (light shaded box, yellow in downloaded file) will be

subject to linear regression Scatchard analysis (automatically). The Scatchard plot

and Electrode calibration curves (Fig. 6, bottom) can be inspected to check

linearity. If values within the regression window are not on the linear range, they

can throw oV the analysis. The top and bottom two [Ca2þ] are excluded from the

regression to allow calculations of [Ca2þ] for solutions outside that range.5. Finally, the free [Ca2þ] and pCa are automatically recalculated using the

measured K0Ca and total [EGTA] (from the auto-analysis) as well as the total

Ca2þ values (last two columns). The Electrode calibration curve and Scatchard

plot allow you to get an overview of the results. (Fig. 6).

We routinely use this for calibration solutions for both Ca2þ-electrodes and

fluorescent indicators. In addition to improving the reliability of Ca2þ calibration

solutions, one of the convenient aspects of this spreadsheet is that you can see all

the details of what is going on. For example, you can see that the EGTA is almost

completely saturated as you get up to 10 mM free [Ca2þ]. In this range we usually

believe the electrode, rather than our ability to pipette within 1% of the required

volume. On the other hand, as you approach the detection limit of the electrode

(e.g., �pCa 9), we use the recalculated pCa values. The measured versus predicted

K0Ca, EGTA purity and [Ca2þ] can also be useful in identifying potential systematic

errors or changes in your procedures.

B. Preparing BuVer Solution

1. Basic Steps in Solution Preparation

There are no hard and fast rules or special tricks to make these buVers, butspecial care in weighing and pipetting, and common sense can help avoid some

potential problems. The water should be well purified to minimize contamination

with Ca2þ and other metals. We usually use water that is first distilled and then run

through a water purification system containing at least one ion exchange column

(e.g., Nanopure, from Barnstead). This provides water with resistivity of

>15 MOhm-cm. Starting with good water like this is important for removal of

other metal contaminants as well as Ca2þ. There can also be contaminating Ca2þ

and metals in the salts and chemicals used to make solutions. In the end, it is

typical to find 1–3 mM free [Ca2þ] in nominally Ca2þ-free solutions. This can be

checked with a Ca2þ-electrode.


Some people include 1–2 mM TPEN, a heavy metal chelator in Ca2þ-buVersolutions. This can chelate submicromolar amounts of heavy metals, which may

or may not be chelated by the dominant Ca2þ buVer. This may not be important in

routine applications, but may ensure that the Ca2þ-sensitive process under study

will not be altered by trace amounts of other metals. All solutions should be made

and stored in clean plastic ware (careful washing and extensive rinsing in deionized

water is required). Glass containers should be avoided. EGTA can leach Ca2þ out

of glass leading to gradual increase in free [Ca2þ] in the solutions. We have often

been able to store Ca2þ calibration solutions for more than 6 months in polypro-

pylene bottles (provided that there is no organic substrate to foster bacterial

growth).

An accurate [Ca2þ] standard is important for making Ca2þ buVers. It is diYcult

to make accurate [Ca2þ] using CaCl2 � 2H2O typically used to make physiological

solutions. This is because the hydration state varies making stoichiometric weigh-

ing imprecise. CaCO3 can be more accurately weighed, but has the disadvantage

that you must then drive oV the CO2 with prolonged heating and HCl, unless

HCO3 is desired in the solutions (which is a weak Ca2þ buVer itself). A convenient

alternative is to buy a CaCl2 standard solution and we use a 100 mM CaCl2solution from Orion (BDH also sells an excellent 1 M CaCl2 standard). To save

money, one can titrate a larger volume of CaCl2 to the same free [Ca2þ] as the

Orion standard using a Ca2þ-electrode.It is also important to prepare accurate stock solutions of Ca2þ chelators. EGTA

from diVerent commercial sources diVer somewhat in purity (Bers, 1982; Miller

and Smith, 1984), but manufacturers provide purity estimates that help (we find

that purity typically ranges from 95 to 100% of the stated purity). BAPTA has also

been reported to contain 20% water by weight (Harrison and Bers, 1987), but can

be dried at 150 �C until the weight is constant to assure removal of water. If one

measures the total buVer concentration (as described in section above) this prob-

lem can be largely obviated. We typically measure the purity of each lot of EGTA

or BAPTA that we use, taking this approach. Then we often keep track on the

bottle itself, so that we can confirm the value upon subsequent tests with the same

batch. EGTA (in the free acid form) is also not very soluble because of the acid pH.

For neutral pH solutions, it is practical to dissolve EGTA with KOH in a 1:2

stoichiometry, since at neutral pH two of the four protons on EGTA are disso-

ciated (vs. all four for BAPTA).

When Ca2þ is added to EGTA solutions, 2 mol of Hþ are released for each mole

of Ca2þ bound. Thus, the pH should always be adjusted as the Ca2þ is being added

or afterward. The strong pH dependence of the K0Ca of EGTA (Fig. 1) emphasizes

the importance of this point. We typically measure [Ca2þ] and pH simultaneously

just before the solutions are brought up to final volume (for approximate pH

adjustment) and after, for final pH adjustment (as close to the third decimal

place as possible) and [Ca2þ] measurement. The solutions are also checked again

later to assure consistency. The rigorous attention to pH adjustment will obviously

be less crucial for the BAPTA buVers.


It may well be asked, why not just use BAPTA rather than EGTA? The main

reason is expense, BAPTA is about 30 times more expensive. The other reason is

that EGTA is the ‘‘Devil we know’’ and indeed we do know much about its

chemistry (e.g., metal binding constants, DH values). For applications with small

volumes of solution though, it may be quite reasonable to replace EGTA with

BAPTA.

The ionic strength contribution of the pH buVer should also be included in the

ionic strength calculation (Ie¼0.5SCi|zi|). This requires calculation of the fraction

of buVer in ionized form (i.e., not protonated).

2. Potential Complications

Not all of the desired constants have been determined for the metals and

chelators of interest. This places some limitations on how accurately one can

predict the free [Ca2þ] of a given complex solution or determine how much total

Ca2þ is required to achieve a desired free [Ca2þ]. The same is true for other species

of interest (e.g., Mg2þ, Mg2þ-ATP). Some Ca2þ buVers also can interact with Ca2þ

in multiple stoichiometries (e.g., the low aYnity Ca2þ buVer, NTA (nitrilotriacetic

acid) can form Ca2þ-NTA2 complexes). There can also be systematic errors in pH

measurements (Illingworth, 1981) or purity of reagents. Purity can be estimated as

described above.

The pH problem is actually quite common, especially with combination pH

electrodes. To put it simply, the reference junction of some electrodes (particularly

with ceramic junctions) can develop junction potentials which are sensitive to ionic

strength. This problem can be exacerbated when the ionic strength of the experi-

mental solutions diVers greatly from the pH standards (typically low ionic strength

phosphate pH standard buVers). A systematic error in solution pH of about 0.2 pH

units is not at all uncommon. As is clear from Fig. 1, this could translate into a 0.4

error in log K0Ca and produce a two- to threefold diVerence in free [Ca2þ] even

where EGTA is at its best in terms of buVer capacity.While measuring the free [Ca2þ] with an electrode can be extremely valuable, it is

not foolproof either. Ca2þ electrodes are not perfectly selective for Ca2þ (see

Chapter 3). For example, the selectivity of these electrodes for Ca2þ over Mg2þ

is about 30,000–100,000 (Schefer et al., 1986). This roughly corresponds to the

diVerence in intracellular concentrations. Thus, a 100 nM Ca2þ solution with

1 mM Mg2þ would look to the electrode like a 110–130 nM Ca2þ solution. For

the Ca2þ electrodes described in Chapter 3 (using the ETH 129 chelator), the

interference by Na or K is less. For 140 mM Na or K in a 100 nM Ca2þ solution,

the apparent [Ca2þ] would be only about 101 nM.

Some Ca2þ buVers can also interfere with Ca2þ electrodes. Citrate, DPA (dipi-

colinic acid), and ADA (acetamidoiminodiacetic acid), three low aYnity Ca2þ

buVers were found to interfere with Ca2þ electrode measurements, while NTA

did not (Bers et al., 1991). Interestingly, citrate, DPA, and ADA (which modified

electrode behavior) also modified Ca2þ channel characteristics, but NTA did not.


When Ca2þ electrodes cannot be practically used, one may still be able to use

optical indicators such as the fluorescent indicators fura-2, indo-1, Fluo-4, Fluo-

5N for selected [Ca2þ] ranges, or the metallochromic dyes antipyralazo III, mur-

exide, or tetramethylmurexide for higher free [Ca2þ] (Kd�200 mM, 3.6, and

2.8 mM, respectively, Ohnishi, 1978, 1979; Scarpa et al., 1978). Of course, these

indicators require calibration too.

A general potential complication with Ca2þ buVers is that they may alter the very

processes one is interested in studying with Ca2þ buVers. For example, EGTA and

other Ca2þ-chelators have been documented to increase the Ca2þ sensitivity of the

plasmalemmal and SR Ca2þ-ATPase pumps and also of Naþ/Ca2þ exchange

(Berman, 1982; Sarkadi et al., 1979; Schatzmann, 1973; Trosper and Philipson,

1984). For example, 48 mMEGTAdecreased the apparentKCa ofNaþ/Ca2þ exchange

in cardiac sarcolemmal vesicles from 20 to 5 mMCa2þ (Trosper and Philipson, 1984).

These points above are not meant to discourage one from using Ca2þ buVers,but simply to point out some of the potential problems that one might encounter.

Being aware of what might occur can help troubleshoot, when things do not make

sense. Clearly, the use of Ca2þ buVer solutions is essential for the understanding ofCa2þ-dependent phenomena. Our aim here is to provide helpful information.

C. Software Programs

While the above spreadsheet is useful for very simple Ca-EGTA or Ca-BAPTA

solutions used for calibrations, it is not suYcient for more complex buVers that onetypically uses experimentally (which include Mg2þ in addition to Ca2þ and multiple

anionic species like ATP that bind Ca2þ and Mg2þ). Several computer programs

have been described (Bers et al., 1994; Brooks and Storey, 1992; Fabiato, 1988;

McGuigan et al., 1991; Schoenmakers et al.., 1992; Taylor et al., 1992), but we will

focus on, MaxChelator developed by one of the authors (CWP Bers et al., 1994). We

have seen above that care is needed in using Ca2þ electrodes. This is equally true for

any software used to determine free metal concentrations in the presence of chela-

tors. In both cases, careful measurement of environmental conditions is needed:

temperature, pH, and ionic strength, as well as attention to the quality and accuracy

of measurement of all reagents. In addition, software is aVected by the choice of

stability constants, quality of the code, and the particular algorithms used, and of

course, the understanding of those using the software (being dependent on personal

knowledge and the ease of use of the software).

Fabiato and Fabiato (1979) broke ground for average users by publishing their

paper on using a hand held programmable calculator to determine free [Ca2þ] or[Mg2þ] in the presence of EGTA. Before then complicated and user unfriendly

software running on main frames and mini computers was all that was available.

Use of Ca2þ electrodes was also just starting and not easy for most labs to

implement. The Fabiato code opened this door, but was somewhat limited.

Richard Steinhardt’s lab used the Fabiato paper to write a version for the Apple

2e, and one of us (CWP) further developed this to a program known as the


MaxChelator series of programs, first introduced in the 1994 version of this

chapter.

There was no internet 16 years ago when the first edition of this chapter was

presented. The compilation of useful stability and thermodynamic constants (e.

g., Martell and Smith, 1974, 1977) has not grown with the explosion of biological

use of Ca2þ buVers and novel Ca2þ indicators (although resources are available

at the National Institute of Standards and Technology (NIST) web site http://

www.nist.gov/srd/nist46.htm). For most of these new compounds, accurate sta-

bility constants have not been determined. Further, there is some disagreement

over which constants and algorithms are best. However, as implied above, it is

valuable to be able to calculate appropriate stoichiometric concentrations of, for

example, Ca2þ, Mg2þ, EGTA, and ATP to use in your solutions to obtain the

desired free [Ca2þ] and [Mg2þ] and [Mg2þ-ATP]. On the other hand, there is no

substitute for actually measuring the concentration when possible to

avoid imperfections in the calculations and also systematic errors (McGuigan

et al., 2007).

Two commonly used programs are Chelator by Theo Schoenmakers

(Schoenmakers et al., 1992) and the MaxChelator series by one of the authors

(CWP). Chelator is written for DOS and has not been updated since 1992 (making

it less broadly useful in 2010) as fewer computers and users run DOS programs and

the user interface is dated. The MaxChelator series expanded into Windows (both

16 and 32 bit), andmore recently into the web via Javascript to be more OS neutral.

This website has downloadable versions of the MaxChelator suite, Chelator, and

several other related tools (including the Bers’ Spreadsheets as in Fig. 6): http://

maxchelator.stanford.edu/downloads.htm

1. Ideal Software Criteria

1. First and foremost is the software has to give the correct answer or at least

close enough that it does not aVect the experimental conclusions (and allows

measurement verification).

2. Must be easy to use. Users should not be confused as to where to enter

information or what information to enter.

3. Adaptable. There should be an easy way to enter diVerent constants and

possibly even allow for diVerent methods of doing some of the calculations.

4. Source code available so the knowledge is not lost with the programmer/

researchers.

No current software handles all these requirements well.

http://www.nist.gov/srd/nist46.htm

http://www.nist.gov/srd/nist46.htm

http://maxchelator.stanford.edu/downloads.htm

http://maxchelator.stanford.edu/downloads.htm


2. Accuracy

We think that the constants and algorithms for the calculations in these pro-

grams are appropriate, but the ambiguities in available fundamental constants,

some nuances in their application and the systematic experimental errors discussed

above conspire such that solution making by recipe is imperfect. Experience and

direct [Ca2þ] measurement are the best ways to limit inaccuracies in the long run.

Indeed, blind acceptance of the calculations, and a presumption that there are no

systematic errors (in either the solution making, pH, temperature, or in the

calculations) enhances the likelihood for inaccuracies.

3. Ease of Use and Adaptability

Early versions of these programs were DOS based and developed within various

labs, with user interfaces not consistent with present day expectations. Patton’s

MaxChelator has attempted to maintain a user friendly interface that has evolved

during the past 15 years. One can input either desired free concentrations of Ca2þ,Mg2þ, Mg-ATP to obtain the total concentrations required or vice versa, and there

are simple intuitive screens for these inputs. We are not aware of a commercial

program that does these calculations. To ensure adaptability in this future, code

should be available as open source to maximize access for future improvements

(including by others).

Both Chelator and MaxChelator allow for additional chelators and sets of

constants to be created or changed (a useful feature), but do not allow for their

inner workings or equations to be changed. If programs were open sourced then

the inner algorithms could be changed to try out diVerent ideas. Software could be

‘‘tweaked’’ and refined to hopefully overcome its limitations. Another issue for the

future is whether there will be suYcient interest in the continual evolution of these

software suites.

4. Other Things to be Aware of When Doing This Work

In line with the aforementioned concerns, Patton et al. (2004) mentioned several

precautions. First, pH control is critical (within 0.01 pH unit) especially for EGTA.

Moreover, when metals bind to chelators, Hþ is released aVecting pH, and that

increases the importance of appropriate pH buVer choice. Second, chelators

cannot reduce free metal concentration to zero. An equilibrium is set up, and

[Ca2þ] and [Mg2þ] (like [Hþ]) are always finite. If proteins or other moieties in your

system have higher aYnity for the metal than your chelator, it can complicate

chelator eVectiveness. Contamination with Ca2þ is almost always present. Third,

select the right chelator for metal concentrations of interest. Just like pH buVerswhich work in a range of �0.5 pH units, chelators work in a range of �0.5pKd

(�0.3–3Kd). Using too high Kd allows contaminant Ca2þ to strongly influence

[Ca2þ] at the low levels, while too low Kd will result in saturation and loss of


buVering near the higher end. Note that the lower aYnity buVers typically have

higher oV-rates and thus equilibrate faster and damp rapid [Ca2þ] spikes more

eVectively.

5. Why Use Software and Where to Get MaxChelator?

Despite all the caveats, using software to calculate free [Ca2þ], [Mg2þ], and [Mg-

ATP] is necessary to have a reasonable chance of getting the solutions right. And

we think that MaxChelator is a useful tool in this regard. More information and

downloads (free) are available at http://maxchelator.stanford.edu/. Whenever

practical, it is also highly desirable to measure the [Ca2þ] using either electrodes

or fluorescent Ca2þ indicators to confirm the predictions and check for reproduc-

ibility. These measurements are less practical for other metals (even Mg2þ) or

anions, for which electrodes and fluorescent indicators are less available.

6. MaxChelator for Windows

The earliest MaxChelator eVort was a DOS program which was then moved to

Windows (Winmaxc), and the latest version is posted at the above website. The

current Windows version allows visualization in two or three dimensions, some of

the key factors that aVect the result. The source code is hundreds of pages long andis complied under the Delphi (Visual Pascal) environment (not posted). The files of

constants are editable using a text editor and any number of files of constants can

be maintained. However, the algorithms used to calculate the eVects of tempera-

ture and ionic concentration are hidden, limiting the flexibility of this version. On

the other hand, it is straightforward to use and multiple metals and chelators can

be easily used together (e.g., Ca2þ, Mg2þ, Ba2þ, BAPTA, Br2BAPTA, and ATP).

7. Javascript Web Versions

Not everyone wants to use windows software, so the algorithms have been

ported to Javascript which runs on all platforms that have a browser with Java-

script enabled (with syntax similar to C programming language). One limitation is

that the math libraries for interpreted Javascript are not as accurate as those for

compiled programs (and rounding errors can create limitations, especially with

simultaneous use of multiple metals and chelators). Some people also disable

Javascript because of the fear of malware.

An advantage of Javascript, besides running on most computers, is that the

source code is readily accessible and can be saved, edited, and then run on any

machine. If the result is an improvement, it can be shared. Another advantage is

the simple user interface. Everything is in front of you all the time, and it is very

easy and intuitive to change pH, temperature, ionic concentration, or metal/

chelator concentrations. Several variants are available for either online calcula-

tions or download. Some are simple binary Ca-EGTA or Mg-ATP calculators like

http://maxchelator.stanford.edu/


the one in the screenshot below. One simply chooses the calculation type at the top,

enter the temperature, pH, and ionic strength (line 2) and the two known

Ca-EGTA concentrations. Not only are the traditional find free Ca2þ and find

total Ca2þ calculations performed, but also the occasionally useful find total

EGTA given the free and total (or free and bound) Ca2þ levels can be performed.

There are also the slightly more complex versions for Ca–Mg-ATP-EGTA equili-

bria. Finally, there is the more comprehensive version (Web MaxC) that allows any

combination of cations (Al3þ, Ba2þ, Ca2þ, Cd2þ, Cu2þ, Fe2þ, Mg2þ, Sr2þ, Zn2þ)and 12 diVerent chelators (including EGTA, BAPTA, Br2BAPTA, EDTA, ATP,

ADP, and citrate), but the simplicity and functionality are the same as the simple Ca-

EGTA version above. There are also versions posted that use the Schoenmakers

constants and conditions and other versions will be posted as they are written. These

programs can thus be helpful in designing solutions with particular free ion con-

centrations, but should be used with understanding of the limitations.

V. Discussion and Summary

It is important to be able to prepare solutions with buVered [Ca2þ], and often

these solutions are complicated by multiple equilibria, and theoretical and practi-

cal limitations. Here we have discussed some of the basic principles that are

involved, several key factors that complicate the process and provide some practi-

cal tools and advice to increase the probability that one can make the desired

solution. However, neither the calculations nor the solution preparation nor

measurement are foolproof. One must be alert to some of the potential caveats,

and make independent measurements when possible.

Often it is useful to make a very careful set of calibration standards at a selected

ionic strength, temperature, and pH, using simpler solutions (e.g., containing

simple Ca-EGTA buVers) for standardization of either a Ca2þ electrode or fluo-

rescent Ca2þ indicator (as in Fig. 6). Note also that there are [Ca2þ] solution sets


sold commercially for this purpose, but they may not mimic your preferred con-

ditions (and we have not used them). Once your electrode or fluorescent indicator

is calibrated, you can use it to measure [Ca2þ] in more complex solutions, where

solution predictions are less reliable. These more complex solutions could be a

series of solutions of diVerent [Ca2þ] or [Mg-ATP], for example, to activate

skinned muscle fiber contraction, expose to permeabilized cells, dialyze into cells

via patch pipettes or use directly in biochemical assays in vitro. This is certainly a

rational and practical approach. One practical caveat is that the aYnity of most

fluorescent Ca2þ indicators changes (usually decreases two- to fourfold) in the

cellular environment versus in protein-free solutions (Harkins et al. 1993; Hove-

Madsen and Bers, 1991; Konishi et al., 1988; Uto et al., 1991) and this seems to be

due to the interaction of the indicators with cellular proteins (which can be

mimicked in vitro). So precise control and measurement of [Ca2þ]i in cells are

both very diYcult to fully achieve. On the other hand, the importance of [Ca2þ]makes it important to measure and try to control [Ca2þ] as best one can. Aware-

ness of the limitations may seem daunting, but should not dissuade one from these

valuable experiments. Even relative [Ca2þ] changes and imperfect control or

measurement of [Ca2þ] are of value in understanding these processes.

Acknowledgements

This work was supported by a grant from the National Institutes of Health (HL30077).

References

Berman, M. C. (1982). Stimulation of calcium transport of sarcoplasmic reticulum vesicles by the

calcium complex of ethylene glycol bis(b-aminoethyl ether)-N,N0,-tetraacetic acid. J. Biol. Chem.

257, 1953–1957.

Bers, D. M. (1982). A simple method for the accurate determination of free [Ca] in Ca-EGTA solutions.

Am. J. Physiol. 242, C404–C408.

Bers, D. M., and MacLeod, K. T. (1988). Calcium chelators and calcium ionophores. In ‘‘Handbook of

Experimental Pharmacology,’’ (P. F. Baker, ed.), Vol. 83, pp. 491–507. Springer-Verlag, Berlin.

Bers, D.M., Hryshko, L. V., Harrison, S.M., andDawson, D. (1991). Citrate decreases contraction and

Ca current in cardiac muscle independent of its buVering action. Am. J. Physiol. 260, C900–C909.

Bers, D. M., Patton, C. W., and Nuccitelli, R. (1994). A practical guide to the preparation of Ca buVers.

Methods Cell Biol. 40, 3–29.

Brooks, S. P. J., and Storey, K. B. (1992). Bound and determined: A computer program for making

buVers of defined ion concentrations. Anal. Biochem. 201, 119–126.

Fabiato, A. (1988). Computer programs for calculating total from specified free or free from specified

total ionic concentrations in aqueous solutions containing multiple metals and ligands. Methods

Enzymol. 157, 378–417.

Fabiato, A. (1991). Ca2þ buVering: computer programs and simulations. In ‘‘Cellular Calcium:

A Practical Approach,’’ (J. G. McCormack, and P. H. Cobbold, eds.), pp. 159–176. New York,

Oxford University Press.

Fabiato, A., and Fabiato, F. (1979). Calculator programs for computing the composition of the

solutions containing multiple metals and ligands used for experiments in skinned muscle cells.

J. Physiol. Paris 75, 463–505.


Good, N. E., Winget, G. D., Winter, W., Connolly, T. N., Izawa, S., and Singh, R. M. M. (1966).

Hydrogen ion buVers for biological research. Biochemistry 5, 467–477.

Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985). A new generation of Ca indicators with greatly

improved fluorescence properties. J. Biol. Chem. 260, 3440–3450.

Harafuji, H., and Ogawa, Y. (1980). Re-examination of the apparent binding constant of ethlene glycol

bis(b-aminoethyl ethylene)-N,N,N0,N0-tetraacetic acid with calcium around neutral pH. J. Biochem.

87, 1305–1312.

Harkins, A. B., Kurebayashi, N., and Baylor, S. M. (1993). Resting myoplasmic free calcium in frog

skeletal muscle fibers estimated with fluo-3. Biophys. J. 65, 865–881.

Harrison, S. M., and Bers, D. M. (1987). The eVect of temperature and ionic strength on the apparent

Ca-aYnity of EGTA and the analogous Ca-chelators BAPTA and dibromo-BAPTA. Biochim.

Biophys. Acta 925, 133–143.

Hove-Madsen, L., and Bers, D. M. (1991). Indo-1 binding in permeabilized myocytes alters its spectral

and Ca binding properties. Biophys. J. 63, 89–97.

Illingworth, J. A. (1981). A common source of error in pH measurements. Biochem. J. 195, 259–262.

Konishi, M., Olson, A., Hollingworth, S., and Baylor, S. M. (1988). Myoplasmic binding of fura-

2 investigated by steady-state fluorescence and absorbance measurements. Biophys. J. 54, 1089–1104.

Marks, P. W., and Maxfield, F. R. (1991). Preparation of solutions with free calcium concentration in

the nanomolar range using 1, 2-bis(o-aminophenoxy)ethane-N,N,N0,N0-tetraacetic acid. Anal. Bio-

chem. 193, 61–71.

Martell, A. E., and Smith, R. M. (1974). Critical Stability Constants, Vol. 1. Plenum, New York.

Martell, A. E., and Smith, R. M. (1977). Critical Stability Constants, Vol. 3. Plenum, New York.

McGuigan, J. A. S., Luthi, D., and Buri, A. (1991). Calcium buVer solutions and how to make them:

A do it yourself guide. Can. J. Physiol. Pharmacol. 69, 1733–1749.

McGuigan, J. A., Kay, J. W., Elder, H. Y., and Luthi, D. (2007). Comparison between measured and

calculated ionised concentrations in Mg2þ/ATP, Mg2þ/EDTA and Ca2þ/EGTA buVers; influence of

changes in temperature, pH and pipetting errors on the ionised concentrations.Magnes. Res. 20, 72–81.

Miller, D. J., and Smith, G. L. (1984). EGTA purity and the buVering of calcium ions in physiological

solutions. Am. J. Physiol. 246, C160–C166.

Nakon, R. (1979). Free metal ion depletion by Good’s buVers. Anal. Biochem. 95, 527–532.

Ohnishi, S. T. (1978). Characterization of the murexide method: Dual wavelength spectrophotometry of

cations under physiological conditions. Anal. Biochem. 85, 165–179.

Ohnishi, S. T. (1979). A method of estimating the amount of calcium bound to the metallochromic

indicator arsenazo III. Biochim. Biophys. Acta 586, 217–230.

Patton, C., Thompson, S., and Epel, D. (2004). Some precautions in using chelators to buVer metals in

biological solutions. Cell Calcium 35, 427–431.

Pethig, R., Kuhn, M., Payne, R., Adler, E., Chen, T.-H., and JaVe, L. F. (1989). On the dissociation

constants of BAPTA-type calcium buVers. Cell Calcium 10, 491–498.

Sarkadi, B., Shubert, A., and Gardos, G. (1979). EVect of Ca-EGTA buVers on active calcium transport

in inside-out red cell membrane vesicles. Experientia 35, 1045–1047.

Scarpa, A., Brinley, F. J., and Dubyak, G. (1978). Antipyralazo III, a middle range Ca2þ metallochro-

mic indicator. Biochemistry 17, 1378–1386.

Schatzmann, H. J. (1973). Dependence on calcium concentrations and stoichiometry of the calcium

pump in human red cells. J. Physiol. 235, 551–569.

Schefer, U., Ammann, D., Pretsch, E., Oesch, U., and Simon, W. (1986). Neutral carrier based Ca2þ-selective electrode with detection limit in the subnanomolar range. Anal. Chem. 58, 2282–2285.

Schoenmakers, T. J., Visser, G. J., Flik, G., and Theuvenet, A. P. (1992). CHELATOR: an improved

method for computing metal ion concentrations in physiological solutions. Biotechniques 870–874,

876–879.

Smith, G. L., and Miller, D. J. (1985). Potentiometric measurements of stoichiometric and apparent

aYnity constants of EGTA for protons and divalent ions including calcium. Biochim. Biophys. Acta

839, 287–299.


Taylor, R. B., Trimble, C., Valdes, J. J., Wayner, M. J., and Chambers, J. P. (1992). Determination of

free calcium. Brain Res. Bull. 29, 499–501.

Trosper, T. L., and Philipson, K. D. (1984). Stimulatory eVect of calcium chelators on Naþ–Ca2þ

exchange in cardiac sarcolemmal vesicles. Cell Calcium 5, 211–222.

Tsien, R. Y. (1980). New calcium indicators and buVers with high selectivity against magnesium and

protons: design, synthesis and properties of prototype structures. Biochemistry 19, 2396–2404.

Tsien, R. Y., and Zucker, R. S. (1986). Control of cytoplasmic calcium with photolabile tetracarbox-

ylate 2-nitrobenzhydrol chelators. Biophys. J. 50, 843–853.

Uto, A., Arai, H., and Ogawa, Y. (1991). Reassessment of fura-2 and the ratio method for determina-

tion ofintracellular Ca2þ concentrations. Cell Calcium 12, 29–37.

CHAPTER 2


Photorelease Techniques for Raising orLowering Intracellular Ca2þ

Robert ZuckerMolecular and Cell Biology DepartmentUniversity of California at BerkeleyBerkeley, California, USA

OGY,All rig

A

Vhts

bstract

OL. 99 0091reserved. 27 DOI: 10.1016/S0091-

-679X679X

I.
I ntroduction II. N itr Compounds
A.
Chemical Properties B. Calculating [Ca2þ]i Changes in Cells
III.
D M-Nitrophen A. Chemical Properties B. Calculating Changes in [Ca2þ]i
IV.
D iazo Compounds A. Chemical Properties B. Calculating EVects of Photolysis
V.
I ntroduction into Cells VI. L ight Sources V II. C alibration V III. P urity and Toxicity IX. B iological Applications
A.
Ion Channel Modulation B. Muscle Contraction C. Synaptic Function D. Other Applications
X.
C onclusions R eferences
/10 $35.00(10)99002-X

28 Robert Zucker

Abstract

The quantitative manipulation of intracellular calcium concentration ([Ca2þ]i) is avaluable instrument in the modern cell biologists’ toolbox for unraveling the many

cell processes controlled by calcium. I summarize here the major classes of photo-

sensitive calcium chelators used to elevate or reduce [Ca2þ]i, with an emphasis on

their physicochemical properties and methods of calculating magnitudes and kinet-

ics of eVects on [Ca2þ]i of flashes and steady light, in order to encourage the choice of

the best substance for particular applications. The selection and calibration of

appropriate light sources, and procedures for introducing the chelators into cells,

spatially restricting [Ca2þ]i changes, and measuring the profiles of [Ca2þ]i changesimposed by photolysis, are also described. The final section describes a selection of

biological applications.

I. Introduction

Photolabile Ca2þ chelators, sometimes called caged Ca2þ chelators, are used to

control [Ca2þ]i in cells rapidly and quantitatively. A beam of light is aimed at cells

filled with a photosensitive substance that changes its aYnity for binding Ca2þ.Several such compounds have been invented that allow the eVective manipulation

of [Ca2þ]i in cells. These compounds oVer tremendous advantages over the alter-

native methods of microinjecting Ca2þ salts, pharmacologically releasing Ca2þ

from intracellular stores, or increasing cell membrane permeability to Ca2þ using

ionophores, detergents, electroporation, fusion with micelles, or activation of

voltage-dependent channels, in terms of specificity of action, repeatability and

reliability of eVect, maintenance of cellular integrity, definition of spatial extent,

and rapidity of eVect, all combined with the ability to maintain the [Ca2þ]i changefor suYcient time to measure its biochemical or physiological consequences. Only

photosensitive chelators allow the concentration of Ca2þ in the cytoplasm of intact

cells to be changed rapidly by a predefined amount over a selected region or over

the whole cell. Since loading can precede photolysis by a substantial amount of

time, cells can recover from the adverse eVects of the loading procedure before the

experiments begin. The ideal photosensitive Ca2þ chelator does not exist, but

would have the following properties.

1. The compound could be introduced easily into cell, by microinjection or by

loading a membrane-permeating derivative that would be altered enzymatically to

an impermeant version trapped in cells.

2. The compound could be loadedwithCa2þ to such a level that the unphotolyzed

formwould buVer the [Ca2þ]i to near the normal resting level, so its introduction into

cells would not perturb the resting Ca2þ level. Additionally, by adjusting the Ca2þ

loading or selecting chelator variants, the initial resting Ca2þ level could be set to

somewhat higher or lower than the normal resting concentration.

2. Photorelease Techniques for Raising or Lowering Intracellular Ca2þ 29

3. The chelator should be chemically and photolytically stable.

4. Photolysis by a bright flash of light should allow rapid changes in the free

Ca2þ level; this characteristic requires rapid photochemical and subsequent dark

reactions of the chelator.

5. Photolysis should be achievable with biologically appropriate wavelengths,

which requires a high quantum eYciency and absorbance at wavelengths that

readily penetrate cytoplasm but cause little biological damage, that is, that are

not highly ionizing. For the chelator to be protected from photolysis by light

needed to view the preparation would also be useful.

6. The photoproducts, or postphotolysis buVer mixture, should continue to

buVer Ca2þ, and so hold it at the new level in the face of homeostatic pressure

from membrane pumps and transport processes.

7. Neither the unphotolyzed chelator nor its photoproducts should be toxic, but

rather should be inert with respect to all ongoing cellular molecular and physio-

logical processes. Three classes of compounds, the nitr series, DM-nitrophen, and

the diazo series, share enough of these properties to have generated intense interest

and widespread popularity, and form the subjects of this review.

Numerous more general reviews of photolabile or caged compounds, which

contain some information on photolabile Ca2þ chelators, have appeared (Adams

and Tsien, 1993; Gurney, 1993; Kao and Adams, 1993; Kaplan and Somlyo, 1989;

McCray and Trentham, 1989; Ogden, 1988; Parker, 1992; Walker, 1991). Reviews

focused more on photosensitive Ca2þ chelators may also be consulted (Ashley

et al., 1991a; Ellis-Davies, 2003; Gurney, 1991; Kaplan, 1990).

II. Nitr Compounds

A. Chemical Properties

The first useful class of photosensitive Ca2þ chelators was developed by Roger

Tsien. This nitr class of compounds relies on the substitution of a photosensitive

nitrobenzyl group on one or both of the aromatic rings of the Ca2þ chelator 1,2-bis

(o-aminophenoxy)ethane-N,N,N0,N0-tetracetic acid (BAPTA) (Adams and Tsien,

1993; Adams et al., 1988; Kao and Adams, 1993; Tsien and Zucker, 1986). Light

absorption results in the abstraction of the benzylic hydrogen atom by the excited

nitro group and oxidation of the alcohol group to a ketone. The resulting nitro-

sobenzoyl group is strongly electron withdrawing, reducing the electron density

around the metal-coordinating nitrogens and reducing the aYnity of the tetracar-

boxylate chelator for Ca2þ. In the first member of this series, nitr-2, methanol is

formed as a by-product of photolysis, but in subsequent members (nitr-5, nitr-7,

and nitr-8) only water is produced. Photolysis of nitr-2 is also slow (200 ms time

constant). For the other nitr chelators, the dominant photolysis pathway is much

faster (nitr-7, 1.8 ms; nitr-5, 0.27 ms; and nitr-8, not reported). For these reasons,

30 Robert Zucker

nitr-2 is no longer used. For the three remaining nitr compounds, photolysis is

most eYcient at the absorbance maximum for the nitrobenzhydrol group, about

360 nm, although light between 330 and 380 nm is nearly as eVective. The quantumeYciency of the Ca2þ-bound form is about 1/25 (nitr-5, 0.035 and nitr-7, 0.042) and

is somewhat less in the Ca2þ-free form (0.012 and 0.011). The absorbance at this

wavelength is 5500 M�1 cm�1 (decadic molar extinction coeYcient) for nitr-5 and

nitr-7, and 11,000 M�1 cm�1 for nitr-8. The structures of the nitr series of com-

pounds are given in Fig. 1; and the photochemical reaction of the most popular

member of this group, nitr-5, is shown in Fig. 2. The physico-chemical properties

of these and other photosensitive chelators are summarized in Table I.

These chelators share the advantages of the parent BAPTA chelator: high

specificity for Ca2þ over Hþ and Mg2þ (Mg2þ aYnities, 5–8 mM), lack of

dependence of Ca2þ aYnity on pH near pH 7, and fast buVering kinetics. One

limitation is that the drop in aYnity in the nitr compounds after photolysis is

relatively modest, about 40-fold for nitr-5 and nitr-7. The Ca2þ aYnity of nitr-5

drops from 0.15 to 6 mM at 120-mM ionic strength after complete photolysis.

These aYnities must be reduced at higher ionic strength, roughly in proportion to

the tonicity (Tsien and Zucker, 1986). By incorporating a cis-cyclopentane ring

into the bridge between the chelating ether oxygens of BAPTA, nitr-7 was created

COO−

COO− COO− COO− COO− COO− COO−

O

O O

O O

OO

O

O

O

O

O

O

OHOH

OH OH

O2N

HO

OMe

MeHH

H H H

N N

NO2NO2

NO2 NO2

nitr-5 nitr-7

nitr-9nitr-8

O O O

N

N

N

COO− COO−COO− COO−

N

COO−

Me

N

COO−COO−

Fig. 1 Structures of the nitr series of photolabile chelators, which release calcium on exposure to light.

NO2 NO+

Photolyzed nitr-5(low Ca2+ affinity)

H2O

CH3

N N

CH3H

nitr-5(high Ca2+ affinity)

HO

OO O

O

O

O OO

Ca2+

Ca2+

−O2C

−O2C −O2C

−O2C

O

N

CO2− CO2

−

CO2− CO2

−

N

hn

Fig. 2 Reaction scheme for the photorelease of nitr-5.


with significantly higher Ca2þ aYnities (54 nM, decreasing to 3 mM after photol-

ysis at 120 mM ionic strength). To increase the change in Ca2þ-binding aYnity on

photolysis, nitr-8 was created with a 2-nitrobenzyl group on each aromatic ring of

BAPTA. Photolysis of each group reduces aYnity only about 40-fold, as for nitr-

5 and nitr-7, but photolysis of both nitrobenzyl groups reduces aYnity nearly

3000-fold, to 1.37 mM, with a quantum eYciency of 0.026. Finally, nitr-9 is a

dicarboxylate 2-nitrobenzhydrol with a low Ca2þ aYnity that is unaVected by

photolysis; this compound can be used to control for nonspecific eVects of the

photoproducts.

Initially, nitr-5 was the substance most often applied in biological experiments,

largely because it was the first photolabile chelator to have most of the qualities of

the ideal substance. The limited aYnity for Ca2þ of this substance in the unpho-

tolyzed form requires that it be lightly loaded with Ca2þ when introduced into

cells; otherwise, the resting [Ca2þ]i will be too high. However, the compound in a

lightly loaded state contains little Ca2þ to be released on photolysis. Nitr-7 alle-

viates this problem with an aYnity closer to that of normal resting [Ca2þ]i, but itssynthesis is more diYcult and its photochemical kinetics are significantly slower.

Both compounds permit less than two orders of magnitude increase in [Ca2þ]i,generally to only the low micromolar range, and then only with very bright flashes

or prolonged exposures to steady light to achieve complete photolysis. Nitr-8 per-

mits a much larger change in [Ca2þ]i. Photolysis kinetics for this compound have

not yet been reported. Neither nitr-8 nor the control compound nitr-9 is presently

commercially available; nitr-5 and nitr-7 are supplied by CalBiochem (La Jolla,

California).

Table IProperties of photosensitive Ca chelators

Compound

(availability)

tPhot(ms)

lmax

(nm) Q.E.Ca Q.E.free

Before

photolysis

After

photolysis

Before

photolysis

After

photolysisCa-bind-

ing on-

rate

KD-Ca

(mM)

KD-Mg

(mM)

KD-Ca

(mM)

KD-Mg

(mM)

e10-Ca(M�1 cm�1)

e10-free(M�1 cm�1)

e10-Ca(M�1 cm�1)

e10-free(M�1 cm�1)

Nitr-5 0.27 365 0.035 0.012 0.145 8.5 6.3 8 5450 5750 13,800 27,300 0.5

Nitr-7 1.8 365 0.042 0.011 0.054 5.4 3.0 5 5780 5540 24,700 10,000 0.2

Nitr-8 – 365 0.026b – 0.5 – 1370 – 11,000 1100 �50,000 �20,000 –

Nitr-9 – 365 �0.02 �0.02 �1000 �10 �1000 �10 �5500 �5500 �15,000 �25,000 –

Azid-1 <2.0 342 1.0 0.9 0.23 8 120 8 33,000 27,000 11,500 5550 0.8

DM-

nitrophen

0.015;1.9 370 0.18 0.18 0.007c 0.0017c 4200;89 2.5 4330 4020 3150 3150 0.02

NP-EGTA 0.002 345 0.23 0.23 0.08c 9 1000 9 975 975 1900 1900 0.017

DMNPE-4 – 347 0.09 0.09 0.048d 7 1000 7 5140 5140? 5140? 5140? 0.01?

NDBF-

EGTA

0.01;0.52 330 0.7 0.7 0.1 15 1000 15 18,400 18,400? 18,400? 18,400? –

Diazo-2 0.134 370 0.057a 0.030a 2.2 5.5 0.073 3.4 2080 22,200 700 2080 0.8

Diazo-4 – 370 0.030a,b 0.030a,b 89 – 0.055 2.6 4600 46,000 <500 <500 0.8

Diazo-3 0.24 375 0.048 0.048 >1000 20 >1000 20 2100 22,800 700 2100 0.8

tPhot, photolysis time constant; lmax, absorbance maximum of Ca-loaded compound; most eVective photolysis wavelength; Q.E.Ca,free, quantum eYciency,

Ca-bound (free); KD-Ca,Mg, Ca (Mg) dissociation constant (1/aYnity) at 0.1–0.15 M ionic strength; and e10-Ca,free, decadic absorbance extinction coeYcient of

Ca-bound (free) compound.a10% of absorbed photons produce a nonphotolyzable photoproduct similar in absorbance and aYnities to unphotolyzed diazo.bFor photolysis of each site.cAt pH 7.2; doubles for each 0.3 pH unit reduction.dAt pH 7.2; increases 2.5� for each 0.2 pH unit reduction.


The latest addition to the nitr-like class of compounds based on BAPTA is

Azid-1 (Adams et al., 1997). This compound was derived from the high-aYnity

fluorescent indicator derivative of BAPTA, fura-2, by addition of an azido substit-

uent to fura’s benzofuran-3 position. Unlike fura-2, neither this compound nor its

photoproducts are fluorescent; and unlike the other nitr compounds and the

dimethoxynitrophenyl class of Ca2þ chelators (see below), it relies on the photo-

sensitivity of an aromatic azide rather than a nitrobenzyl group. UV absorption

peaking at 372 nm (342 nm for the Ca2þ-bound form) probably leads to formation

of a nitrene which steals hydrogen from water to produce an amidine, which with

another hydrogen converts to a nitrenium that rapidly combines with water to

form an amidinium that reacts with OH� to produce the final low-aYnity electron-

withdrawing benzofurane-3-one photoproduct plus ammonia. Thus, photolysis

absorbs one net proton and produces one molecule of ammonia for each molecule

of azid-1 photolyzed, which can lead to an elevation of pHi in weakly buVered cells.

This disadvantage is counterbalanced by substantial advantages. Photolysis of

both Ca2þ-bound and Ca2þ-free forms of zaid-1 is phenomenally eYcient

(Q.E.�1), and azid-1 is very UV-dark, absorbing at 33,000 M�1 cm�1 when

Ca2þ-bound (or 27,000 M�1 cm�1 when free); these factors combine to make it

250–300 times more sensitive to light than nitr-5! Moreover, its Ca2þ-aYnity drops

from 230 nM to 120 mM, on photolysis, a change that is 12 times the change in nitr-

5 aYnity on photolysis. Like the nitr compounds, it hardly binds Mg2þ at all

(KD¼8 mM), and its Ca2þ-binding (�109 M�1 s�1) and photolysis rates (t<2 ms)are equally rapid. In most respects, azid-1 comes closest to the ideal-caged Ca2þ

compound. Unfortunately, its synthesis is quite diYcult, and it has never been

commercially available; at present, apparently none exists at all.

B. Calculating [Ca2þ]i Changes in Cells

If nitr-5 or azid-1 is photolyzed partially by a flash of light, the reduction in Ca2þ

aYnity of a portion of the chelator occurs within �0.3 ms. During this period of

photolysis, low-aYnity buVer is being formed and high-aYnity buVer is vanishingwhile the total amount of Ca2þ remains unchanged. As the buVer concentrationschange, Ca2þ ions reequilibrate among the new buVer concentrations by shifting

from the newly formed low-aYnity photoproduct to the remaining unphotolyzed

high-aYnity caging chelator. Since the on-rate of binding is close to the diVusionlimit (as calculated from Adams et al., 1988; see also Ashley et al., 1991b), this

equilibration occurs much faster than photolysis, and Ca2þ remains in quasi-

equilibrium throughout the photolysis period. The [Ca2þ]i in a cell rises smoothly

in a step-like fashion over a period of 0.3 ms from the low level determined by the

initial concentrations of total Ca2þ, and unphotolyzed chelator to a higher level

determined by the final concentrations of all the chelator species after partial

photolysis. At least in the case of nitr-5, [Ca2þ]i remains under the control of the

low- and high-aYnity species, so the elevated Ca2þ is removed only gradually by

extrusion and uptake into organelles. Thus, nitr-5 and azid-1 are well suited to

34 Robert Zucker

producing a modest but quantifiable step-like rise in response to a partially

photolyzing light flash, or a gradually increasing [Ca2þ]i during exposure to steady

light. Subsequent flashes cause further increments in [Ca2þ]i. These increments

actually increase because, with each successive flash, the remaining unphotolyzed

chelator is loaded more heavily with Ca2þ. Eventually, unphotolyzed nitr-5 or

azid-1 is fully Ca2þ-bound, and subsequent flashes elevate Ca2þ by smaller incre-

ments as the amount of unphotolyzed chelator drops.

If a calibrated light source is used that photolyzes a known fraction of nitr in the

light path, or in cells filled with chelator and exposed either fully or partially to

light, then the mixture of unphotolyzed nitr and photoproducts may be calculated

with each flash (Lando and Zucker, 1994; Lea and Ashley, 1990). The diVerentquantum eYciencies of free and Ca2þ-bound chelators must be taken into account.

Simultaneous solution of the buVer equations for photolyzed and unphotolyzed

chelators and native Ca2þ buVers predicts the [Ca2þ]i. For suYciently high nitr-5

concentration (above 5 mM), the native buVers have little eVect and usually may

be ignored in the calculation. Further, since [Ca2þ]i depends on the proportion of

chelator loaded with Ca2þ, the exact chelator concentration in the cell makes little

diVerence, at least in small cells or cell processes.

If the cell is large, the light intensity will drop as it passes from the front to the

rear of the cell. Knowing the absorbance of cytoplasm and chelator species at

360 nm, and the chelator concentration before a flash, the light intensity and

photolysis rate at any point in the cytoplasm may be calculated. A complication

in this calculation is that nitr-5 photoproducts have very high absorbance (Ca2þ-free photoproduct, 24,000 M�1 cm�1 and Ca2þ-bound photoproduct,

10,000 M�1 cm� l) (Adams et al., 1988). As photolysis proceeds, the cell darkens

and photolysis eYciency is reduced by self-screening. For azid-1 the situation is

reversed: its photoproducts have much lower absorbance (11,500 and

5000 M�1 cm�1) for Ca2þ-bound and free species or 1/3 and 1/4 of the respective

unphotolyzed forms. Thus light penetrates more deeply as photolysis proceeds.

Regardless, with estimation of the spatial distribution of light intensity from Beer’s

Law, the spatial concentrations of photolyzed and unphotolyzed chelator can be

computed; from this calculation follows the distribution of the rise in [Ca2þ]i. Thesubsequent spatial equilibration of [Ca2þ]i can be calculated by solving diVusionequations, often in only one dimension, using the initial [Ca2þ]i and chelator

distributions as the boundary conditions. EVects of endogenous buVers, uptake,and extrusion mechanisms on the rise in [Ca2þ]i can be incorporated into the

calculations. Simulations of the temporal and spatial distribution of [Ca2þ]i havebeen devised (Zucker, 1989) and applied to experimental data on physiological

eVects of [Ca2þ]i; the predicted changes in [Ca2þ]i have been confirmed with Ca2þ-sensitive dyes (Lando and Zucker, 1989). Simplified and approximate models using

the volume-average light intensity to calculate volume-average photolysis rate and

average [Ca2þ]i changes often suYce when the spatial distribution of [Ca2þ]i is notimportant, for example, in small cells or processes or when estimating the change in

[Ca2þ]i in a cell after diVusional equilibration has occurred.


III. DM-Nitrophen


Graham Ellis-Davies followed a diVerent strategy for releasing Ca2þ—by

attaching a 2-nitrobenzyl group to one of the chelating amines of ethylenediami-

netetraacetic acid (EDTA) to form the photosensitive chelator dimethoxynitro-

phenyl-EDTA or DM-nitrophen (Ellis-Davies and Kaplan, 1988; Kaplan and

Ellis-Davies, 1988). Photolysis by UV light in the wavelength range 330–380 nm

cleaves the DM-nitrophen with a quantum eYciency of 0.18 through multiple

intermediate pathways (McCray et al., 1992) to form iminodiacetic acid and a

Hþ-absorbing 2-nitrosoacetophenone derivative, with 65% of the photoproducts

formed with a time constant of 15 ms and the rest with t¼1.9 ms (Faas et al., 2005,

2007). A simplified reaction is shown in Fig. 3. Although DM-nitrophen binds

Ca2þ with an aYnity of 7 nM at pH 7.2, Ca2þ-bound chelator forms a photoprod-

uct-binding Ca2þ with a 4-mM aYnity, while the free form (and Mg2þ-boundforms, see below) photolyze to a 90 mM-KD photoproduct at an ionic strength of

150 mM. These values are from the most recent of the continually evolving models

of DM-nitropen photolysis (Ayer and Zucker, 1999; Bollmann and Sakmann,

2005; Faas et al., 2005, 2007; Kaplan and Ellis-Davies, 1988; Neher and Zucker,

1993). Thus, complete photolysis of Ca2þ-DM-nitrophen can elevate Ca2þ over

50,000-fold, much more than photolysis of the nitr compounds or azid-1. This

significant advantage is counterbalanced to some extent by the facts that the

photoproducts buVer Ca2þ so weakly that the final [Ca2þ]i will be determined

largely by native cytoplasmic buVers, and that the Ca2þ liberated by photolysis of

DM-nitrophen will be removed more readily by extrusion and uptake pumps.

CO2−

CO2−

CO2−

CO2− CO2

−

CO2−

−O2C

−O2CCa2+

Ca2+

NO2 NO

ON

DM-nitrophen

(high Ca2+ affinity)

DM-nitrophen photolysis products

(low Ca2+ affinity)

OCH3

OCH3 OCH3OCH3

N N

NH

+hn

Fig. 3 Structure of and reaction scheme for DM-nitrophen, which releases calcium on exposure to

light.

36 Robert Zucker

The absorbance of Ca2þ-saturated and free DM-nitrophen is 4330 and

4020 M�1 cm�1, respectively, and 0.18.

A serious complication of DM-nitrophen is that it shares the cation-binding

properties of its parent molecule EDTA. In particular, Hþ and Mg2þ compete for

Ca2þ at the hexacoordinate-binding site. The aYnity of DM-nitrophen for Mg2þ

at pH 7.2 is 1.7 mM, whereas the photoproducts bind Mg2þ with aYnities of about

2 mM. Further, both the Ca2þ- and Mg2þ-aYnities of DM-nitrophen are highly

pH-dependent (Grell et al., 1989), doubling for each 0.3 units of pH increase. Thus,

in the presence of typical [Mg2þ]i levels of 1–3 mM, DM-nitrophen that is not

already bound to Ca2þ will be largely in the Mg2þ-bound form. Further, excess

DM-nitrophen will suck Mg2þ oV ATP, which binds it substantially more weakly,

compromising the ability of ATP to serve as an energy source or as a substrate for

ATPases. Finally, photolysis of DM-nitrophen will lead to a jump in [Mg2þ]i aswell as [Ca2þ]i, and to a rise in pH. Unless controlled by native or exogenous pH

buVers, this pH change can alter the Ca2þ and Mg2þ aYnities of the remaining

DM-nitrophen. In the absence of Ca2þ-loading, DM-nitrophen even may be used

as a caged Mg2þ chelator (Ellis-Davies, 2006). Attributing physiological responses

to a [Ca2þ]i jump, therefore, requires control experiments in which DM-nitrophen

is not charged with Ca2þ. DM-nitrophen currently is sold by CalBiochem.

To circumvent the problems arising fromMg2þ competing for the Ca2þ-bindingsite of DM-nitrophen, a second generation derivative of ethylene glycol bis(b-aminoethylether)-N,N,N0,N0-tetraacetic acid (EGTA, which binds Mg2þ only

very weakly) coupled to a light-sensitive ortho-nitrophenyl group was developed

(Ellis-Davies and Kaplan, 1994). This compound, nitrophenyl-EGTA or NP-

EGTA, is very rapidly cleaved (t¼2 ms) (Ellis-Davies, 2003) to Hþ-absorbingimidodiacetic acid photoproducts with eVective Ca2þ-KD of 1 mM, 12,500-fold

higher (lower aYnity) than that of the unphotolyzed cage (80 nM) at pH 7.2, with

pH-dependence similar to that of EGTA, EDTA, and DM-nitrophen. Unlike DM-

nitrophen, Mg2þ binding to NP-EGTA is negligible (9 mM before and after

photolysis). Quantum eYciency (0.23) is similar to that of DM-nitrophen, and

higher than for the nitr compounds, but less than that of azid-1. However,

photolysis eYciency is seriously limited by its low absorbance (975 M�1 cm�1),

only 1/6—1/4 those of the nitr compounds and DM-nitrophen, and less than 3% of

azid-1’s absorbance.

More recently, a dimethoxy-ortho-nitrophenyl derivative of EGTA (DMNPE-4)

was introduced (Ellis-Davies and Barsotti, 2006), with somewhat higher Ca2þ

aYnity (48 nM), dropping with time constants of 10 and 17 ms to 1 mM on

photolysis, low Mg2þ-aYnity (7 mM), and under half the quantum eYciency

(0.09) but over five times the absorbance (5140 M�1cm�1), thus twice the photoly-

sis eYciency of NP-EGTA. An additional very slow phase releasing 30% of caged

Ca2þ with t�667 ms was observed.

Ellis-Davies’ lab has also produced a new generation of EGTA-based chelators

using the novel photosensitive chromophore nitrodibenzofuran or NDBF-EGTA

(Momotake et al., 2006). This compound binds Ca2þ with KD¼100 nM at pH 7.2,


presumably with an on-rate similar to that of the other EGTA derivatives. Ca2þ

aYnity drops sharply to �1 mM on photolysis with time constants of 14 and

520 ms. Quantum eYciency (0.7) and absorbance (18,400 M�1 cm�1) are extremely

high, as is the change in Ca2þ aYnity (10,000-fold at pH 7.2), making this a very

attractive candidate for future-caged Ca2þ research.

B. Calculating Changes in [Ca2þ]i
Calculating [Ca2þ]i changes on photolysis of NP-EGTA and its congeners is
similar to that for the nitr compounds (if the pH dependence of binding constants

is ignored), sinceMg2þ binding is not an issue. Since the chelators’ Ca2þ aYnities is

similar to resting cytoplasmic [Ca2þ]i levels, filling cells with a half-Ca2þ-loadedchelator will not disturb [Ca2þ]i but can release substantial amounts of Ca2þ

(�1 mM) which will be reduced about 100-fold by the cell’s endogenous buVers.However, except for NDBF-EGTA, the low absorbance usually limits flash pho-

tolysis to at most about 20%.

Quantifying changes in [Ca2þ]i caused by photolysis is much more diYcult for

DM-nitrophen. The initial level of [Ca2þ]i before photolysis depends upon the total

concentrations of Mg2þ, Ca2þ, DM-nitrophen, ATP, and native Ca2þ buVers,because at least two buVers (DM-nitrophen and endogenous buVers) compete

for Ca2þ, two buVers (ATP and DM-nitrophen) compete for Mg2þ, and, afterpartial photolysis, both cations also bind to the two photoproducts. Calculating

equilibrium Ca2þ levels involves simultaneous solution of at least six nonlinear

buVer equations (Delaney and Zucker, 1990), which is a tedious chore at best.

Also, the various dissociation constants depend on ionic strength, and have been

measured only at 150 mM. The high aYnity of DM-nitrophen for Ca2þ might

appear to dominate the buVering of Ca2þ in cytoplasm, but this idea is misleading.

A solution of DM-nitrophen that is 50% saturated with Ca2þ will hold the free

[Ca2þ]i at 7 nM at pH 7.2; this action will be independent of the total DM-

nitrophen concentration. However, 5 mM DM-nitrophen with 2.5 mM Ca2þ and

5 mM Mg2þ will buVer free [Ca2þ]i to about 2 mM; now doubling all concentra-

tions results in a final [Ca2þ]i of around 5 mM. Since the total [Mg2þ]i available, asfree or weakly bound to ATP, is several millimolar, partially Ca2þ-loaded DM-

nitrophen may bring the resting Ca2þ level to a surprisingly high level. Because the

solution is still buVered, this [Ca2þ] may be reduced only gradually by pumps and

uptake, but eventually Ca2þwill be pumped oV the DM-nitrophen until the [Ca2þ]iis restored to its normal level. Then photolysis may lead to only tiny jumps in

[Ca2þ]i. However, if a large amount of Ca2þ-loaded-DM-nitrophen is introduced

into a cell relative to the total [Mg2þ]i, Ca2þ can be buVered to low levels while

photolysis can release a large amount. In fact, if enough DM-nitrophen is intro-

duced into cells with no added Ca2þ, it may gradually absorb Ca2þ from cytoplasm

and intracellular stores and photolysis can produce a substantial jump in [Ca2þ]i.Therefore, both resting and the postphotolysis levels of Ca2þ may vary over very

wide ranges, depending on [DM-nitrophen]i, [Mg2þ]i, and cellular [Ca2þ]i control

38 Robert Zucker

processes, all of which are diYcult to estimate or control. Thus, quantification of

changes in [Ca2þ]i is not easy to achieve.

The situation may be simplified by perfusing cells with Ca2þ-DM-nitrophen

solutions while dialyzing out Mg2þ and mobiles endogenous buVers (Neher and

Zucker, 1993; Thomas et al., 1993). Of course, this procedure will not work in

studies of cell processes requiring Mg2þ-ATP or if perfusion through whole-cell

patch pipettes is not possible.

Another consequence of Mg2þ binding by DM-nitrophen is that cytoplasmic

Mg2þ may displace Ca2þ from DM-nitrophen early in the injection or perfusion

procedure, leading to a transient rise in [Ca2þ]i before suYcient DM-nitrophen is

introduced into the cell (Neher and Zucker, 1993; Parsons et al., 1996; Thomas

et al., 1993). Such a ‘‘loading transient’’ was accurately predicted from models of

changes of the concentrations of total [Ca2þ]i, [Mg2þ]i, ATP, native buVer, andDM-nitrophen during filling from a whole-cell patch electrode (R. S. Zucker,

unpublished). Since this process may have important physiological consequences,

controlling it is important. The process may be eliminated largely by separating the

Ca2þ-DM-nitrophen-filling solution in the pipette from the cytoplasm by an

intermediate column of neutral solution [such as dilute EGTA or BAPTA] in the

tip of the pipette, which allows most of the Mg2þ to escape from the cell before the

DM-nitrophen begins to enter. Then most of the loading transient occurs within

the tip of the pipette.

One method of better controlling the change in [Ca2þ]i in DM-nitrophen experi-

ments is to fill cells with a mixture of Ca2þ-DM-nitrophen and another weak Ca2þ

buVer such as N-hydroxyethylethylenediaminetriacetic acid (HEEDTA) or l,3-

diaminopropan-2-ol-tetraacetic acid (DPTA) (Neher and Zucker, 1993). These

tetracarboxylate Ca2þ chelators have Ca2þ aYnities in the micromolar or tens of

micromolar range. If cells are filled with such a mixture without Mg2þ, the initialCa2þ level can be set by saturating the DM-nitrophen and adding appropriate

Ca2þ to the other buVer. Then photolysis of DM-nitrophen releases its Ca2þ onto

the other buVer; the final Ca2þ can be calculated from the final buVer mixture in

the same fashion as for the nitr compounds. Since all the constituent aYnities are

highly pH dependent, a large amount of pH buVer (e.g., 100 mM) should be

included in the perfusion solution, and the pH of the final solution adjusted

carefully.

The kinetic behavior of DM-nitrophen and the NP-EGTAs is much more

complex than their equilibrium reactions. Photolysis proceeds rapidly (ts¼0.2

and 2 ms for DM-nitrophen, 2 ms for NP-EGTA), but the on-rate of Ca2þ binding

is much slower, about 20 mM�1 ms�1 (Ellis-Davies, 2003; Faas et al., 2005, 2007).

This characteristic has particularly interesting consequences for partial photolysis

of partially Ca2þ-loaded chelator. A flash of light will release some Ca2þ, whichinitially will be totally free. If the remaining unphotolyzed and unbound chelator

concentration exceeds that of the released Ca2þ, this Ca2þ will rebind, displacing

Hþ within milliseconds and producing a brief [Ca2þ]i ‘‘spike’’ (Ellis-Davies et al.,

1996; Grell et al., 1989; Kaplan, 1990;McCray et al., 1992), followed by a near-step


fall in pH. IfMg2þ is also present, a secondary relaxation of [Ca2þ]i follows becauseof the slower displacement of Mg2þ from DM-nitrophen (Ayer and Zucker, 1999;

Delaney and Zucker, 1990; Escobar et al., 1995, 1997). Moreover, if a steady UV

source is used to photolyze DM-nitrophen, rebinding continually lags release,

leading to a low (micromolar range) free [Ca2þ]i while the illumination persists.

When the light is extinguished, the [Ca2þ]i drops rapidly to a low level under control

of the remaining chelator (Zucker, 1993). In the case of DM-nitrophen bound to

Mg2þ, achievement of equilibrium is somewhat slower (tens of milliseconds). Thus,

a reversible ‘‘pulse’’ of [Ca2þ]i is generated, the amplitude of which depends on light

intensity and the duration of which is controlled by the length of the illumination.

This situation remains so until the remaining unphotolyzed cage becomes fully

saturated with Ca2þ, whereupon [Ca2þ]i escapes from the control of the chelator,

imposing a practical limit on the product of [Ca2þ]i and duration of about

0.75 mM s for DM-nitrophen. Similar kinetic considerations apply when Ca2þ is

passed by photolysis from a Ca2þ cage to another buVer such as BAPTA, EGTA,

HEEDTA, orDPTA. Judicious selection of buVers and buVer ratiosmay be used to

shape this Ca2þ ‘‘spike’’ to match a hypothetical naturally occurring [Ca2þ]i(t)waveform and test its physiological consequences (Bollmann and Sakmann,

2005). If this behavior is considered undesirable, it may be avoided by using only

fully Ca2þ-saturated DM-nitrophen, due to its extremely high Ca2þ-aYnity, for

which rebinding to unphotolyzed chelator is impossible. Thus, the kinetic complex-

ity of the nitrophen class of chelators can be turned to experimental advantage,

greatly magnifying the flexibility of experimental [Ca2þ]i control.

IV. Diazo Compounds


In some experiments, being able to lower the [Ca2þ]i rapidly, rather than raise it,

is desirable. For this purpose, caged chelators were developed. Initial attempts

involved attachment of a variety of photosensitive protecting groups to mask one

of the carboxyl groups of BAPTA, thus reducing its Ca2þ aYnity until restored by

photolysis. Such compounds displayed low quantum eYciency (Adams et al., 1989;

Ferenczi et al., 1989) and their development has not been pursued. A more suc-

cessful approach (Adams et al., 1989) involved substituting one (diazo-2) or both

(diazo-4) of the aromatic rings of BAPTA with an electron-withdrawing diazoke-

tone that reduces Ca2þ aYnity, much like the photoproducts of the nitr com-

pounds. Figure 4 shows the structures of the diazo series of chelators. Photolysis

converts the substituent to an electron-donating carboxymethyl group while

releasing a proton; the Ca2þ aYnity of the photoproduct is thereby increased.

The reaction is illustrated in Fig. 5.

Diazo-2 absorbs one photon with quantum eYciency 0.03 to increase aYnity, in

433 ms, from 2.2 mM to 73 nM at 120-mM ionic strength (or to 150 nM at 250 mM

COO−

N N N N N

Mediazo-2 diazo-3 diazo-4

O O OMe

O

O O

OOO

N+ N+ N+ +N

N−N− N− −N

COO− COO− COO− COO− COO− COO− COO− COO− COO−

Fig. 4 Structures of the diazo series of photolabile chelators, which take up calcium on exposure to

light.

O

OCH

CH3

N2

−O2C

−O2C −O2C−O2C

−O2C−O2C

O O O O

H+

O

O

C+HC N2

N

CO2−

CO2−

N N N N N

hn

CO2−

CO2−

CO2−

CO2−

CO2−

CH3 CH3

H2O

H2C +

Ca2+

Diazo-2(low Ca2+ affinity)

Photolyzed diazo-2(high Ca2+ affinity)

Fig. 5 Reaction scheme for the photolysis of diazo-2.

40 Robert Zucker

ionic strength). The absorbance maximum of the photosensitive group is

22,200 M� l cm�1 at 370 nm, and drops to negligible levels at this wavelength

after photolysis. A small remaining absorbance reflects formation of a side product

of unenhanced aYnity and unchanged molar extinction coeYcient in 10% of the

instances of eVective photon absorption. This ‘‘inactivated’’ diazo still binds Ca2þ

(with some reduction in absorbance), but is incapable of further photolysis. The

Ca2þ-bound form of diazo-2 has about one-tenth the absorbance of the free form,

dropping to negligible levels after photolysis, with quantum eYciency of 0.057 and

a time constant of 134 ms. Binding of Ca2þ to photolyzed diazo-2 is fast, with an

on-rate of 8�108 M�1 s�1. Mg2þ binding is weak, dropping from 5.5 to 3.4 mM

after photolysis, and pH interference is small with this class of compound.

One limitation of diazo-2 is that the unphotolyzed chelator has suYcient Ca2þ

aYnity that its incorporation into cytoplasm is likely to reduce resting levels to

some degree, and certainly will have some eVect on [Ca2þ]i rises that occur

physiologically. To obviate this problem, diazo-4 was developed with two


photolyzable diazoketones. Absorption of one photon increases the Ca2þ aYnity

from 89 to 2.2 mM (with a 10% probability of producing a side-product with one

inactivated group). Absorption of two photons (with a probability assumed to

equal the square of the probability of one group absorbing one photon, and with a

measured quantum eYciency of 0.015) results in further increase of the aYnity to

55 nM, a total increase of 1600-fold. This large increase in aYnity is, to some

extent, oVset by the small fraction of diazo-4 that can be doubly photolyzed

readily. Thus, a flash of light produces a variety of species: unphotolyzed, singly

photolyzed, doubly photolyzed, singly inactivated, doubly inactivated, and singly

photolyzed-singly inactivated, with a variety of transition probabilities among

species (Fryer and Zucker, 1993). Unphotolyzed diazo-4 is highly absorbent

(46,000 M�1 cm�1 at 371 nm for the free form; about 4600 M�1 cm�1 for the

Ca2þ-bound form). The singly photolyzed species have absorbances of half these

values and doubly photolyzed diazo-4 has negligible absorbance at this wave-

length. Inactivation causes little change in absorbance.

A third member of this series, diazo-3, has a diazoketone attached to half the

cation-coordinating structure of BAPTA, and has negligible Ca2þ aYnity.

On photolysis, diazo-3 produces the photochemical intermediates of diazo-2 plus

a proton, and may be used to control for these eVects of photolysis of the diazo

series. At one time, diazo-2 and diazo-3 (but not diazo-4) were commercially

available (Molecular Probes, Eugene, and Oregon), but these stocks appear to

have been exhausted.

B. Calculating EVects of Photolysis

As for the nitr compounds, equilibration is faster than photolysis, so a flash of

light leads to a smooth step transition in the concentration of Ca2þ chelator

species. If the percentage of photolysis caused by a light flash is known, the

proportions of photolyzed and inactivated diazo-2, or of the six species of diazo-

4, can be calculated. Usually, diazo is injected without any added Ca2þ, so the

eVect of photoreleased buVers is to reduce the [Ca2þ]i from its resting value. This

change can be calculated only if the total Ca2þ bound to the native buVer in

cytoplasm as well as the characteristics of that buVer are known. These character-istics often can be inferred from available measurements on cytoplasmic Ca2þ

buVer power and the normal resting [Ca2þ]i level. The more usual application of

these substances is to reduce the eVect of a physiologically imposed rise in [Ca2þ]i.In many cases, the magnitude of the source of this Ca2þ is known, as in the case of

a Ca2þ influx measured as a Ca2þ current under voltage clamp or the influx

through single channels estimated from single channel conductances. Also, the

magnitude of the total Ca2þ increase in a response can be estimated frommeasured

increases in [Ca2þ]; and estimates of cytoplasmic buVering. With this information,

the expected eVect of newly formed diazo photoproducts on a physiological rise in

[Ca2þ]i can be calculated by solving diVusion equations that are appropriate for the

distribution of Ca2þ sources before and after changing the composition of the

42 Robert Zucker

mixture of buVers in the cytoplasm. Examples of such solutions of the diVusionequation exist for spherical diVusion inward from the cell surface (Nowycky and

Pinter, 1993; Sala and Hernandez-Cruz, 1990), cylindrical diVusion inward from

membranes of nerve processes (Stockbridge and Moore, 1984; Zucker and

Stockbridge, 1983), diVusion from a point source (Fryer and Zucker, 1993;

Stern, 1992), and diVusion from arrays of point sources (Fogelson and Zucker,

1985; Matveev et al., 2002, 2004, 2006, 2009; Pan and Zucker, 2009; Simon and

Llinas, 1985; Tang et al., 2000; Yamada and Zucker, 1992). For large cells, the

spatial nonuniformity of light intensity and photolysis rate also must be consid-

ered, taking into account the absorbances of all the species of diazo and the

changes in their concentration with photolysis. Like azid-1, the self-screening

imposed by diazo chelators is reduced with photolysis, so successive flashes (or

prolonged illumination) are progressively more eVective.

V. Introduction into Cells

Photolabile chelators are introduced into cells by pressure injection from micro-

pipettes, perfusion from whole-cell patch pipettes, or permeabilization of the cell

membrane. Iontophoresis is also suitable for diazo compounds, since this proce-

dure inserts only the Ca2þ-free form. For the caged Ca2þ substances, this method

of introduction requires that the chelator load itself with Ca2þ by absorbing it from

cytoplasm or intracellular stores. Filling cells from a patch pipette has the special

property that, if the photolysis light is confined to the cell and excludes all but the

tip of the pipette inside the cell, the pipette barrel acts as an infinite reservoir of

unphotolyzed chelator. Then the initial conditions of solutions in the pipette can be

restored within minutes after photolysis of the chelator in the cell. The nitr and

diazo compounds are soluble at concentrations over 100 mM and DM-nitrophen

is soluble at 75 mM, so levels in cytoplasm exceeding 10 mM can be achieved

relatively easily, even by microinjection, making the exogenous chelator com-

pound the dominant buVer.Nitr and diazo chelators also have been produced as membrane-permeant acet-

oxymethyl (AM) esters (Kao et al., 1989). Exposure of intact cells to medium

containing these esters (available from CalBiochem and Molecular Probes, respec-

tively) might result in the loading of cells with nearly millimolar concentrations, if

suYcient activity of intracellular esterase is present to liberate the membrane-

impermeant chelator. However, nitr-5 or nitr-7 introduced in this manner is not

bound to Ca2þ, so it must sequester Ca2þ from cytoplasm, from intracellular stores,

or after Ca2þ influx is enhanced, for example, by depolarizing excitable cells. The

final concentration, level of Ca2þ loading, and localization of the chelator are

uncertain, so this method of incorporation does not lend itself to quantification

of eVects of photolysis unless cells are coloaded with a Ca2þ indicator.

During loading and other preparatory procedures, the photolabile chelators may

be protected fromphotolysis with low passUV-blocking filters in the light path of the


tungsten or quartz halide beams used for viewing. For more detail on these filling

procedures, see Gurney (1991). Other methods of loading cells, used primarily with

other sorts of caged compounds, are discussed by Adams and Tsien (1993).

VI. Light Sources

Photolysis of caged Ca2þ chelators requires a bright source of near UV light.

If speed is unimportant, an ordinary mercury or xenon arc lamp may be used.

Mercury lamps have a convenient emission line at 366 nm. Exposure can be

controlled with a shutter, using MgF-coated Teflon blades for particularly bright

sources. Lamps of 100–150 W power with collimating quartz lenses provide suY-cient energy to photolyze�25% of caged Ca2þ compounds in�2 s. Bulbs of larger

power only generate bigger arcs, with more energy in a larger spot of similar

intensity. With additional focusing, photolysis can be achieved in one-tenth the

time or even less. These light sources are the appropriate choice in applications

using reversible [Ca2þ]i elevation with DM-nitrophen.

Fast events require the use of a laser or xenon arc flashlamp. The xenon lamps

are less expensive and cumbersome; convenient commercial systems are available

from Chadwick-Helmuth (El Monte, California), Rapp Optoelektronik (Hamburg,

Germany), TILL Photonics (Grafelting, Germany), and Cairn Research (Faver-

sham, UK). These flashlamps discharge up to 200–300 J electrical energy across

the bulb to provide a pulse of �1 ms duration with up to 300 mJ energy in the

330�380-nm band. The Chadwick-Helmuth unit includes only a power supply and

lamp socket, so a housing with focusing optics must be constructed (see Rapp and

Guth, 1988). Focusing can be accomplished with a UV-optimized elliptical reflec-

tor or with quartz refractive optics. The reflector can be designed to capture more

light (i.e., have a larger eVective numerical aperture), but reflectors have greater

physical distortion than well-made lenses. In practice, the reflector generates a

larger spot with more total energy, but somewhat less intensity, than refractive

methods. One advantage of reflectors is that they are not subject to chromatic

aberration—focusing is independent of wavelength—so the UV will be focused in

the same spot as visual light. This is not true of refractive lenses. To focus and aim

them accurately at the sample, a UV filter must be used to block visual light and the

beam must be focused on a fluorescent surface for visualization. Both types of

housing are available from Rapp Optoelektronik. Using either system, photolysis

rates approaching 80–90% in one flash are achievable. This rate may be reduced by

imposing neutral density filters or reducing discharge energy, but the relationship

between electrical and light energy is not linear and should be measured with a

photometer. Flashlamps can be reactivated only after their storage capacitors have

recharged, setting the minimal interval between successive maximal flashes at

several seconds or more.

F1ashlamps are prone to generating a number of artifacts. The discharge causes

electrical artifacts that can burn out semiconductors and op amps, and reset or

44 Robert Zucker

clear digital memory in other nearby equipment. Careful electrostatic shielding,

wrapping inductors with paramagnetic metal, power source isolation, and using

isolation circuits in trigger pulse connections to other equipment prevent most

problems, which are also diminished in pulsed mercury lamps (Denk, 1997). The

discharge generates a mechanical thump at the coil used to shape the current pulse

through the bulb; this thump can dislodge electrodes from cells or otherwise

damage the sample. Mechanical isolation of the oVending coil solves the problem.

Lamp discharge also produces an air pressure pulse that can cause movement

artifacts at electrodes, which can be seen to oscillate violently for a fraction of a

second when videotaped during a flash. This movement can damage cells severely,

especially those impaled with multiple electrodes. Small cells sealed to the end of a

patch pipette often fare better against such mistreatment. To reduce this source of

injury, the light can be filtered to eliminate all but the near UV. Commercial Schott

filters (UG-I, UG-Il), coated to reflect infrared (IR) light, serve well for this

purpose, but can cut the 330�380-nm energy to 30% or less. Liquid filters to

remove IR and far UV also have been described (Tsien and Zucker, 1986).

Removing IR reduces temperature changes, which otherwise can exceed 1 �C per

flash, whereas removing far UV prevents the damaging eVects of ionizing radia-

tion. Chlorided silver pellets and wires often used in electrophysiological recording

constitute a final source of artifact. These components must be shielded from the

light source or they will generate large photochemical signals.

To simply aim and focus the light beam directly onto the preparation is easiest.

If isolating the lamp from the preparation is necessary, the light beam may be

transmitted by a fiber optic or liquid light guide, with some loss of intensity. If a

microscope is being used already, the photolysis beam may be directed through the

epifluorescence port of the microscope. The lamp itself, or a light guide, may be

mounted onto this port. Microscope objectives having high numerical aperture

and good UV transmission will focus the light quite eVectively onto a small area,

which can be delimited further by a field stop aperture. With the right choice of

objectives and proper optical coupling of the lamp to the light guide and the guide

to the microscope port, light intensities 25 times greater than those obtained by

simply aiming the focused steady lamp or flashlamp can be achieved—suYcient to

half-photolyze DM-nitrophen in 25 ms of steady bright light. TILL Photonics

make a xenon arc spectrophotometer (the Polychrome) with eYcient optical

coupling to several commercial epifluorescence microscopes. Half reflective mir-

rors can be used to combine the photolysis beam with other light sources, such as

those used for [Ca2þ]i measurement. However, as the optical arrangement becomes

more complex, photolysis intensity inevitably decreases.

The newest development in light sources is the high intensity light-emitting diode

(Bernardinelli et al., 2005). This rapidly evolving and inexpensive technology can

already produce 365-nm UV light at 50 mW/cm2 (with LEDs made by Prizmatix,

Modi’in Ilite, Israel, e.g.), or about20%of the intensityof a collimatedxenonarc lamp.

It is often important to restrict photolysis to one region of a cell (Wang and

Augustine, 1995). With epi-illumination, this may be done with a field stop


diaphragm, or by conveying the photolysis beam through a tapered quarty fiber

optic filter to the cell surface (Eberius and Schild, 2001; Godwin et al., 1997).

Lasers provide an alternative source of light with the advantages of a coherent

collimated beam that is focused much more easily to a very small spot. Pulsed

lasers such as the frequency-doubled ruby laser or the XeF excimer laser provide at

least 200 mJ energy at 347 or 351 nm in 50 and 10 ns, with possible repetition rates

of 1 and 80 Hz, respectively. Liquid coumarin-dye lasers, with up to 100 mJ

tunable energy in the UV and pulse duration, are also available. Inexpensive

nitrogen lasers providing lower pulse energies (0.25 mJ) in 5-ns pulses at 337 nm

also have been developed (Engert et al., 1996) and, with appropriate focusing,

might be useful. To date, lasers have found their widest application in studies of

muscle contraction. More information on these laser options is contained in

discussions by Goldman et al. (1984) and McCray and Trentham (1989).

An adaptation of laser photolysis is the two-photon absorption technique (Denk

et al., 1990). A colliding-pulse mode-locked Ti:Sapphire laser generating 100-fs

pulses of 630-nm light at 80 MHz is focused through a confocal scanning micro-

scope. Photolysis of UV-sensitive caged compounds requires simultaneous absorp-

tion of two red photons, so photolysis occurs only in the focal plane of the scanning

beam. This behavior restricts photolysis to about 1 mm3 in three dimensions, but for

most compounds the photolysis rate is so slow, due to their extremely limited two-

photon cross sections, that several minutes of exposure are required with currently

available equipment. The best results were achieved with azid-1 and NDBF-EGTA

(Brown et al., 1999; DelPrincipe et al., 1999; Momotake et al., 2006), as expected

from their high single photon absorbances. Azid-1 could be fully photolyzed in the

two-photon focal volume with a 10-ms pulse train of 7 mW average power, with a

retention time of the released Ca2þ in this volume of about 150 ms. This technique isexpensive and specialized, and is still under development, but may have practical

applications in revealing the precise localization within cells or subcellular orga-

nelles of fixed targets of Ca2þ action or of highly localized Ca2þ buVers.Near-UV light alone seems to have little eVect on most biological tissues, with

the obvious exception of photoreceptors and the less obvious case of smooth

muscle (Gurney et al., 1987). Control experiments on the eVects of light on

unloaded cells, and on the normal physiological response under study, can be

used to ascertain the absence of photic eVects.

VII. Calibration

When designing a new optical system or trying a new caged compound, being

able to estimate the rate of photolysis of the apparatus used is important. This

information is necessary to adjust the light intensity or duration for the desired

degree of photolysis, and to insure that photolysis is occurring at all.

In principle, the fraction (F) of a substance photolyzed by a light exposure of

energy J can be computed from the formula e–(J–J0)¼ (I�F)/0.1, where J0 is the

46 Robert Zucker

energy needed to photolyze 90% of the substance and is given by J0 ¼hcA/Qel,where h is Planck’s constant, c is the speed of light, A is Avogadro’s number, Q is

the quantum eYciency, e is the decadic molar extinction coeYcient, and l is the

wavelength of the light. In practice, however, this equation is rarely useful for the

following reasons.

1. Measuring the energy of the incident light on a cell accurately is diYcult,

especially for light of broad bandwidth with varying intensity at diVerentwavelengths.

2. The quantum eYciency, although provided for all the photolabile chelators, is

not such awell-defined quantity. The value depends critically on how it is measured,

which is not always reported. In particular, the eVective quantum eYciency for a

pulse of light ofmoderate duration (e.g., from a flashlamp) is often greater than that

of either weak steady illumination or a very brief pulse (e.g., from a laser), because

of the possibility of multiple photon absorptions of higher eYciency by photochem-

ical intermediates. This phenomenon has been noted to play a particularly strong

role in nitr-5 photolysis (McCray and Trentham, 1989). Thus, apparent diVerencesin quantum eYciencies between diVerent classes of chelators may be mainly the

results of diVerent measurement procedures.

3. Finally, the quantum eYciency is a function of wavelength, which is rarely

given.

A more practical and commonly adopted approach is mixing a partially Ca2þ-loaded photolabile chelator with a Ca2þ indicator in a solution with appropriate

ionic strength and pH buVering, and measuring the [Ca2þ] change in a small

volume of this solution, the net absorbance of which is suYciently small to

minimize inner filtering of the photolyzing radiation. Suitable indicators include

fura-2, indo-1 (Grynkiewicz et al., 1985), furaptra (Konishi et al., 1991), f1uo-3,

rhod-2 (Minta and Tsien, 1989), Calcium GreenTM, OrangeTM, and CrimsonTM

(Eberhard and Erne, 1991), arsenazo III (Scarpa et al., 1978), and fura-red

(Kurebayashi et al., 1993). The choice depends largely on available equipment.

Fura-2, indo-1, and furaptra are dual-excitation or-emission wavelength fluores-

cent dyes, allowing more accurate ratiometric measurement of [Ca2þ], but theyrequire excitation at wavelengths that photolyze the photolabile Ca2þ chelators

and are subject to bleaching by the photolysis light. The former problem may be

minimized by using low intensity measuring light with a high sensitivity detection

system. Furaptra is especially useful for DM-nitrophen, because of its lower Ca2þ

aYnity. Fluo-3 and rhod-2 were designed specifically for use with photolabile

chelators (Kao et al., 1989), being excited at wavelengths diVerent from those

used to photolyze the chelators, but they are not ratiometric dyes and are diYcult

to calibrate accurately. Calcium Green, Orange, and Crimson suVer the same

limitation, but they are often used because of their fast kinetics and bright intensity,

allowing the accurate tracking of fast changes in [Ca2þ]i. Arsenazo and antipyr-

alazo are metallochromic dyes that change absorbance on binding Ca2þ,


fortunately at wavelengths diVerent from those at which the photolabile chelators

show any significant absorbance. However, these dyes are also diYcult to calibrate

for absolute levels of [Ca2þ], although changes in [Ca2þ] may be determined fairly

accurately. Fura-red is a ratiometric dye excited by visible light, so it might have

some application in calibrating photolysis. A problem common to all the fluores-

cent indicators is that their fluorescent properties may be altered by the presence of

photolabile chelators, which generally are used at millimolar levels whereas the

indicators are present at 100 mM or less. The photolabile chelators often produce

contaminating fluorescence, which also may be Ca2þ-dependent and may partially

quench the fluorescence of the indicators (Hadley et al., 1993; Zucker, 1992). Thus,

the indicators must be calibrated in the presence of photolabile chelator at three

well-controlled [Ca2þ] levels, preferably before and after exposure to the photolysis

flash, before they can be used to measure the eVects of photolysis on [Ca2þ] (Neher

and Zucker, 1993). The low and high [Ca2þ] calibrating solutions may be made

with excess Ca2þ or another buVer such as EGTA or BAPTA, but the intermediate

[Ca2þ] solution is more diYcult to generate, since photolysis of the chelator will

release some Ca2þ and change the [Ca2þ]i and pH in this solution unless it contains

a very high concentration of controlling chelator and pH buVer.The calibration procedure is generally the same for any combination of chelator

and indicator. A small sample of the mixture is placed in a 1-mm length of micro-

cuvette with a 20-mm pathlength (Vitro Dynamics, Rockaway, New Jersey) under

mineral oil to prevent evaporation. This cuvette is exposed repeatedly to the

photolysis beam or to flashes, which should illuminate the whole cuvette uniformly,

and the [Ca2þ] after each flash or exposure is measured using a microscope-based

fluorescence or absorbance photometer. A small droplet of solution under mineral

oil alone would work, and may be necessary if the photolysis beam is directed

through the microscope and illuminates a very small area, but sometimes the

fluorescent properties of the indicators are aVected by the mineral oil. This eVectwould be detected in the procedure for calibrating the chelator-indicator mixture,

but is best avoided using the microcuvettes, in which contact with oil is only at the

edges, the fluorescence or absorbance change of which need not be measured.

In some applications, such as whole-cell patch clamping of cultured cells, using

the cell as a calibration chamber can be easier than any other procedure.

The expected changes in [Ca2þ] depend on the chelator used. The nitr and diazo

chelators should lead to a stepwise rise or fall in [Ca2þ] after each exposure; the

results can be fitted to models of the chelators and their photoproducts, using their

aYnities and the relative quantum eYciencies of free and bound chelators (Fryer

and Zucker, 1993; Lando and Zucker, 1989). The percentage photolysis of the

chelator in response to each light exposure is the only free parameter, and is varied

until the model fits the results. In the case of the high-aYnity DM-nitrophen, little

rise in [Ca2þ] will occur until the total amount of remaining unphotolyzed chelator

equals the total amount of Ca2þ in the solution, whereupon the [Ca2þ] will increasesuddenly. Equations relating initial and final concentrations of DM-nitrophen,

48 Robert Zucker

total [Ca2þ], and photolysis rate (Zucker, 1993) then may be used to calculate

photolysis rate per flash or per second of steady light exposure.

Most photosensitive compounds also undergo substantial absorbance changes

after photolysis. These changes can be monitored during repeated exposure to the

light source without a Ca2þ indicator; the number of flashes or the duration of light

exposure required to reach a given percentage photolysis then can be determined.

Realizing that photolysis proceeds exponentially to completion (Zucker, 1993),

these data can be used to determine the photolysis rate directly. Ideally, both

methods should be used to check for consistent results. A final method for

determining photolysis rate is using high pressure liquid chromatography

(HPLC) to separate and quantify parent chelators and photoproducts in the

reaction solution after partial photolysis (Walker, 1991).

VIII. Purity and Toxicity

When experiments do not work as planned, the first suspected source of error is

the integrity of the photolabile chelator. DiVerent procedures have proved most

useful for testing the diVerent classes of compounds. The nitr and diazo com-

pounds undergo large absorbance changes on binding calcium and photolysis.

A 100 mM solution (nominally) of the chelator is mixed with 50 mM Ca2þ in

100 mM chelexed HEPES solution (pH 7.2), and 0.3 ml is scanned in a 1-mm

pathlength spectrometer. Then 1 ml 1 MK2EGTA is added to bring the [Ca2þ] to 0,

and the sample is scanned again. Finally, 1 ml 5 M CaCl2 is added to provide excess

Ca2þ, and a third scan is recorded. The first scan should be midway between the

other two. If the first scan is closer to the excess Ca2þ scan, it is indicative of a lower

than expected concentration of the chelator, probably because of an impurity.

Alternatively, Ca2þ may have been present with the chelator, which may be

checked by running a scan on the chelator with no added Ca2þ and comparing

the result with a scan with added EGTA; they should be identical. Ca2þ free and

Ca2þ-saturated chelator solutions also are scanned before and after exposure to

UV light suYcient to cause complete photolysis; the spectra are compared with

published figures (Adams et al., 1988, 1989; Kaplan and Ellis-Davies, 1988) to

determine whether the sample was partially photolyzed at the outset. The Ca2þ

aYnities of unphotolyzed and photolyzed chelators can be checked by measuring

the [Ca2þ] of 50%-loaded chelators with a Ca2þ-selective electrode.The absorbance of DM-nitrophen and some related chelators is almost Ca2þ

independent, so these procedures are not eVective. A solution of DM-nitrophen

nominally of 2 mM concentration is titrated with concentrated CaCl2 until the

[Ca2þ] measured with an ion-selective electrode suddenly increases; this change

indicates the actual concentration of the chelator and gives an estimate of purity.

The aYnity of the photolysis products can be measured as for the other chelators;

spectra before and after photolysis indicate whether the sample was already

partially photolyzed.


Purities of 80–90% are typical for commercial samples of all the chelators, but

occasional batches of 60% purity or less have been seen; these also sometimes show

high degrees of toxicity. Whether such low purity is the result of poor synthesis or

storage is unclear. Nitr compounds decompose detectably after only 1 day at room

temperature, and exposure to ambient fluorescent lighting for 1 day causes detect-

able photolysis. Chelators should be shipped on dry ice and stored at�80 �C in the

dark; even under these conditions they do not last forever. Repeated thawing and

freezing also degrades the compounds.

Some of the photolabile Ca2þ chelators display a degree of biological toxicity in

some preparations. Commercial samples of nitr-5 have been seen to lyse sea urchin

eggs (R. S. Zucker and L. F. JaVe, unpublished results) and leech blastomeres

(K. R. Delaney and B. Nelson, unpublished results) within minutes. Zucker and

Haydon (1988) found that nitr-5 blocked transmitter release within 10 min of

perfusion in snail neurons, whereas DM-nitrophen has no similar eVect(P. Haydon, unpublished results). These eVects are not caused by the photoproducts,since photolysis is not necessary for the problems to occur. DM-nitrophen has been

observed to reduce secretion in chromaYn cells; higher chelator concentrations,

photolyzed to give the same final [Ca2þ]i level, caused less secretion (C. Heinemann

and E.Neher, unpublished results). The eVect was overcome partially by inclusion of

glutathione in the perfusion solution, as reported for the photoproducts of other

2-nitrobenzhydrol-based caged compounds (Kaplan et al., 1978). These signs

of toxicity have been observed sporadically; whether they are properties of the

chelators themselves or of impurities in the samples used is unclear. The chelators

have been applied successfully to a wide range of preparations without obvious

deleterious results, although subtle eVects may have been missed.

IX. Biological Applications

A brief synopsis of the earliest biological applications of the caged Ca2þ chela-

tors follows along with a much more selective sampling of the more recent and

extensive literature. This is included in this chapter because many of the original

papers include a wealth of detail about methodology and interpretation of Ca2þ

photorelease technology.

A. Ion Channel Modulation

1. Potassium and Nonspecific Cation Channels

The first and still one of the major applications of photosensitive Ca2þ chelators

is analysis of Ca2þ-dependent ion channels in excitable cells. In 1987, Gurney et al.

first used nitr-2,-5, and-7 to activate Ca2þ-dependent Kþ current in rat sympathetic

neurons. These researchers found that a single Ca2þ ion binds to the channel with

rapid kinetics and 350 nM aYnity.

50 Robert Zucker

The next application of the nitr chelators was in an analysis of Ca2þ-activatedcurrents in Aplysia neurons (Lando and Zucker, 1989). We found that Ca2þ-activated Kþ and nonspecific cation currents in bursting neurons were linearly

dependent on [Ca2þ]i jumps in the micromolar range, as measured by arsenazo

spectrophotometry and modeling studies. Both currents relaxed at similar rates

after photolysis of nitr-5 or nitr-7, reflecting diVusional equilibration of [Ca2þ]inear the front membrane surface facing the light source. Potassium current relaxed

more quickly than nonspecific cation current, after activation by Ca2þ entry during

a depolarizing pulse, because of the additional voltage sensitivity of the Kþ

channels. This diVerence was responsible for the more rapid decay of hyperpolar-

izing afterpotentials than of depolarizing afterpotentials.

The role of Ca2þ-activated Kþ current in shaping plateau potentials in gastric

smooth muscle was explored by Carl et al. (1990). In fibers loaded with nitr-5/AM,

Ca2þ photorelease accelerated repolarization during plateau potentials and

delayed the time to subsequent plateau potentials, suggesting a role for changes

in [Ca2þ]i and Ca2þ-activated Kþ current in slow wave generation.

Another current modulated by [Ca2þ]i is the so-called M current, a muscarine-

blocked Kþ current in frog sympathetic neurons. Although inhibition is mediated

by G-protein coupling of the receptor to phospholipase C, resting M current is

enhanced by modest elevation of [Ca2þ]i (some tens of nanomolar) and reduced by

greater elevation of [Ca2þ]i, which also suppresses the response to muscarine

(Marrion et al., 1991). As for ventricular ICa (see below), several sites of modula-

tion of M current by [Ca2þ]i apparently exist. In these experiments, [Ca2þ]i waselevated by photorelease from nitr-5 and simultaneously measured with fura-2.

Step changes in [Ca2þ]i imposed by diazo-2 photolysis and monitored with bis-

fura-2 fluorescence changes have also been used to characterize the modulation of

cGMP-gated ion channels by [Ca2þ]i (Rebrik et al., 2000).

The after-hyperpolarization that follows spikes in rat hippocampal pyramidal

neurons is caused by a class of Ca2þ-dependent Kþ channels called IAHP channels.

This after-hyperpolarization and the current underlying it rise slowly to a peak 0.5 s

after the endof abriefburstof spikes.Ca2þphotorelease fromnitr-5 orDM-nitrophen

activates this currentwithout delay (Lancaster andZucker, 1994), and the currentmay

be terminated rapidly by photolysis of diazo-4 (but see conflicting results of Sah and

Clements, 1999), suggesting that the delay in its activation following action potentials

is caused by a diVusion delay between points of Ca2þ entry and the IAHP channels.

The Ca2þ sensitivity of the mechanoelectrical transduction current in chick

cochlear hair cells was studied using nitr-5 introduced by hydrolysis of the AM

form (Kimitsuki and Ohmori, 1992). Elevation of [Ca2þ]i to 0.5 mM (measured

with fluo-3) diminished responses to displacement of the hair bundle, and acceler-

ated adaptation during displacement when Ca2þ entry occurred. Preventing Ca2þ

influx blocked adaptation. Evidently, adaptation of this current was the result of

an action of Ca2þ ions entering through the transduction channels.

In guinea pig hepatocytes, noradrenaline evokes a rise in Kþ conductance after a

seconds-long delay. Photorelease of Ca2þ from nitr-5 and use of caged inositol


1,4,5-trisphosphate (caged-IP3) show that this delay arises from steps prior to or

during generation of IP3 (Ogden et al., 1990), which releases Ca2þ from intracellu-

lar stores to activate Kþ current.

2. Calcium Channels

The first application of DM-nitrophen was in a study of Ca2þ channels in chick

dorsal root ganglion neurons (Morad et al., 1988). With divalent charge carriers,

inactivation by photorelease of intracellular Ca2þ occurred within 7 ms, whereas

with monovalent charge carriers a nearly instantaneous block occurred, especially

when Ca2þ was released extracellularly. A similar rapid block of monovalent

current through Ca2þ channels was observed in response to photorelease of

extracellular Ca2þ in frog ventricular cells (Nabauer et al., 1989). DiVerent Ca2þ-binding sites may be exposed if altered conformational states are induced in the

channels by the presence of diVerent permeant ions.

The regulation of Ca2þ current (ICa) in frog atrial cells by [Ca2þ]i also has been

studied with nitr-5 (Charnet et al., 1991; Gurney et al., 1989). Rapid elevation of

[Ca2þ]i potentiated high-voltage-activated or L-type ICa and slowed its deactiva-

tion rate when Ba2þ was the charge carrier, after a delay of several seconds.

Inclusion of BAPTA in the patch pipette solution blocked the eVect of nitr-5

photolysis. The similarity of eVect of Ca2þ and cAMP and their mutual occlusion

suggest a common phosphorylation mechanism.

Regulation of ICa in guinea pig ventricular cells appears to be more complex

(Bates and Gurney, 1993; Hadley and Lederer, 1991). A fast phase of inactivation

reflects a direct action on Ca2þ channel permeation, since ICa inactivation caused

by photorelease of Ca2þ from nitr-5 is independent of the phosphorylation state of

the channels and does not alter gating currents. A late potentiation is also present,

the magnitude of which depends on the flash intensity delivered during a depolar-

izing pulse, but not on the initial [Ca2þ]i level, the degree of loading of nitr-5, or thepresence of BAPTA in the patch pipette. This result suggests that, during a

depolarization, nitr-5 becomes locally loaded by Ca2þ entering through Ca2þ

channels, and that the Ca2þ-binding site regulating potentiation is near the channel

mouth. Larger [Ca2þ]i jumps elicited by photolysis of DM-nitrophen evoke greater

ICa inactivation, but no potentiation, perhaps because of the more transient rise in

[Ca2þ]i when DM-nitrophen is photolyzed.

DM-nitrophen loaded with magnesium in the absence of Ca2þwas used to study

the Mg2þ-nucleotide regulation of L-type ICa in guinea pig cardiac cells (Backx

et al., 1991; O’Rourke et al., 1992). In the presence of ATP, a rise in [Mg2þ]i to50–200 mM led to a near doubling of the magnitude of ICa. Release of caged ATP

also increased ICa. Therefore, the eVect on Ca2þ channels was caused by a rise in

Mg2þ-ATP. Nonhydrolyzable ATP analogs worked as well as ATP, soMg2þ-ATP

seems to modulate Ca2þ channels directly.

We microinjected Aplysia neurons with nitr-5, DM-nitrophen, or diazo-4 to

characterize Ca2þ-dependent inactivation of Ca2þ current (Fryer and Zucker,

52 Robert Zucker

1993). Elevation of [Ca2þ]i to a few micromolar with nitr-5 caused little inactiva-

tion, but photolysis of DM-nitrophen rapidly inactivated half the ICa, presumably

that in the half of the cell facing the light source. Thus, inactivation requires high

[Ca2þ]i levels and occurs rapidly in all channels, even if they are closed. Experi-

ments with diazo-4 showed that an increase in buVering power reduced the rate of

inactivation of ICa modestly. DiVusion-buVer reaction simulations suggest that

Ca2þ acts at a site within 25 nm of the channel mouth (see also Johnson and

Byerly, 1993).

B. Muscle Contraction

One of the earliest applications of photolabile Ca2þ chelators was initiating

muscle contraction in frog cardiac ventricular cells by photorelease of extracellular

Ca2þ from DM-nitrophen (Nabauer et al., 1989). The strength of contraction

elicited by a stepwise rise in [Ca2þ]e showed a membrane potential dependence

that was indicative of entry through voltage-dependent Ca2þ channels rather than

of transport by Naþ–Ca2þ exchange.

Several laboratories have used caged Ca2þ chelators to study Ca2þdependentCa2þ release from the sarcoplasmic reticulum in rat ventricular myocytes.

Valdeolmillos et al. (1989) loaded cells with the AM form of nitr-5, Kentish et al.

(1990) subjected saponin-skinned fibers to solutions containing Ca2þ-loaded nitr-5,and Nabauer and Morad (1990) perfused single myocytes with DM-nitrophen

loaded with Ca2þ. Photolysis elicited a contraction blocked by ryanodine or

caVeine, procedures that prevent release of Ca2þ from the sarcoplasmic reticulum,

implicating Ca2þ-induced Ca2þ release, which could be confined to a portion of a

fiber by localized photolysis (O’Neill et al., 1990).

Gyorke and Fill (1993) used Ca2þ-DM-nitrophen to show that the cardiac

ryanodine receptors adapt to maintained [Ca2þ]i elevation, remaining sensitive to

larger [Ca2þ]i changes and responding by releasing still more Ca2þ. In smooth

muscle from guinea pig portal vein, the IP3-dependent release of Ca2þ was itself

dependent upon [Ca2þ]i (Iino and Endo, 1992). Ca2þ photoreleased from DM-

nitrophen and measured with fluo-3 accelerated Ca2þ release from a ryanodine-

insensitive, IP3-activated store. The possibility that adaptation reflected slow

unbinding of Ca2þ from the channels following a flash-induced Ca2þ ‘‘spike’’

was refuted by demonstrating a rapid deactivation of channel function to a sudden

drop in [Ca2þ]i imposed by diazo-2 (Velez et al., 1997).

Ca2þ-loaded nitr-5 was used in skinned frog and scallop muscle fibers to show

that the rate-limiting step in contraction is not the time-course of the rise in [Ca2þ]ibut rather the response time of the contractile machinery (Ashley et al., 1991b; Lea

and Ashley, 1990). Using isolated myofibrillar bundles from barnacle muscle, Lea

and Ashley (1990) showed that nitr-5 photolysis elevating [Ca2þ]i by 0.2–1.0 mMCa2þ not only activated contraction directly and rapidly but also evoked a slower

phase of contraction that was dependent on Ca2þ-induced Ca2þ release from the

sarcoplasmic reticulum. Analysis of [Ca2þ]i steps imposed by DM-nitrophen or


diazo-2 revealed kinetics of muscle contraction and relaxation steps following the

binding and unbinding of Ca2þ to troponin C (Ashley et al., 1993).

The first biological application of the caged chelator diazo-2 was in the study of

muscle relaxation. Mulligan and Ashley (1989) showed that rapid reduction in

[Ca2þ]i in skinned frog semitendinosus muscle resulted in a relaxation similar to

that occurring normally in intact muscle, indicating that mechanochemical events

subsequent to the fall in [Ca2þ] were rate limiting. However, Lannergren and Arner

(1992) reported some speeding of isometric relaxation after photolysis of diazo-2,

loaded in the AM form into frog lumbrical fibers. Lowered pH slowed relaxation

to a step reduction in [Ca2þ]i (Palmer et al., 1991), perhaps accounting for a

contribution of low pH to the sluggish relaxation of fatigued muscle. In contrast

to frog muscle, photorelease of Ca2þ chelator caused a much faster relaxation in

skinned scallop muscle than in intact fibers (Palmer et al., 1990), suggesting that, in

these cells, relaxation is rate limited primarily by [Ca2þ]i homeostatic processes.

C. Synaptic Function

Action potentials evoke transmitter release in neurons by admitting Ca2þ

through Ca2þ channels. Because of the usual coupling between depolarization

and Ca2þ entry, assessing the possibility of an additional direct action of mem-

brane potential on the secretory apparatus has been diYcult. Photolytic release of

presynaptic Ca2þ by nitr-5 perfused into presynaptic snail neurons cultured in

Ca2þ-free media was combined with voltage clamp of the presynaptic membrane

potential to distinguish the roles of [Ca2þ]i and potential in neurosecretion (Zucker

and Haydon, 1988), revealing no direct eVect of membrane potential on transmit-

ter release.

Hochner et al. (1989) injected Ca2þ-loaded nitr-5 into crayfish motor neuron

preterminal axons, and used a low-[Ca2þ] medium to block normal synaptic

transmission. They found that action potentials transiently accelerated transmitter

release evoked by modest photolysis of nitr-5. However, Mulkey and Zucker

(1991) used fura-2 to show that the extracellular solutions used by Hochner et al.

(1989) failed to block Ca2þ influx through voltage-dependent Ca2þ channels.

When external Ca2þ chelators or channel blockers eliminated influx completely,

spikes failed to have any influence on transmitter release, even when it was

activated strongly by photolysis of intracellularly injected Ca2þ-loaded DM-

nitrophen.

Delaney and Zucker (1990) confirmed at the squid giant synapse that in a Ca2þ-free medium, action potentials have no eVect on transmitter release triggered by a

rise in [Ca2þ]i upon photolysis of presynaptically injected DM-nitrophen. Flash

photolysis of DM-nitrophen produced a transient postsynaptic response resem-

bling normal excitatory postsynaptic potentials. The intense phase of transmitter

release was probably caused by the brief spike in [Ca2þ]i following partial photol-

ysis of partially Ca2þ-loaded DM-nitrophen. This response began a fraction of a

millisecond after the rise in [Ca2þ]i, a delay similar to the usual synaptic delay

54 Robert Zucker

following Ca2þ influx during an action potential; both delays had the same

temperature dependence. Thus photolysis of DM-nitrophen caused a [Ca2þ]i tran-sient resembling that occurring normally at transmitter release sites in the vicinity

of Ca2þ channels that open briefly during an action potential. After the secretory

burst, a moderate phase of transmitter release persisted for 15 ms, corresponding

to a relaxation in [Ca2þ]i measured with fura-2 that probably reflected slow Ca2þ

displacement of Mg2þ bound to unphotolyzed DM-nitrophen.

Similar responses to partial flash photolysis of lightly Ca2þ-loaded DM-nitro-

phen were observed at crayfish neuromuscular junctions (Lando and Zucker,

1994). Transmitter release evoked by slow photolysis of Ca2þ-DM-nitrophen

using steady illumination also has been studied at this junction (Mulkey and

Zucker, 1993). The rate of quantal transmitter release, measured as the frequency

of miniature excitatory junctional potentials (MEJPs), was increased �1000-fold

during the illumination. Brief illumination (0.3–2 s) evoked a rise in MEJP fre-

quency that dropped abruptly back to normal when the light was extinguished, as

would be expected from the reversible rise in [Ca2þ]i that should be evoked by such

illumination, which leaves most of the DM-nitrophen unphotolyzed (Zucker,

1993). Longer light exposures caused an increase inMEJP frequency that outlasted

the light signal, as would be expected from the rise in resting [Ca2þ]i after photoly-sis of most DM-nitrophen. These experiments illustrate the utility of steady pho-

tolysis of partially Ca2þ-loaded DM-nitrophen in generating reversible changes in

[Ca2þ]i in cells.

At cultured snail synapses, FMRFamide inhibits asynchronous transmitter

release elicited by [Ca2þ]i elevated by photolysis of presynaptic nitr-5 (Man-Son-

Hing et al., 1989), and blocks synchronous release to partial flash photolysis of

partially Ca2þ-loaded DM-nitrophen (Haydon et al., 1991). As at crayfish and

squid synapses, these flash-evoked postsynaptic responses resembled the spike-

evoked responses and were triggered by the spike in [Ca2þ]i that results when DM-

nitrophen is used in this fashion.

At leech serotonergic synapses, a presynaptic Ca2þ uptake process may be

activated by photolysis of presynaptic DM-nitrophen; blocking it with zimelidine

or by external Naþ removal eliminated the presynaptic transport current and

prolonged the postsynaptic response, uncovering a contribution of this process

to the termination of transmitter release (Bruns et al., 1993).

DM-nitrophen has been used extensively to probe the steps involved in exocyto-

sis in endocrine cells. Measuring [Ca2þ]i changes with furaptra, we and others

(Heinemann et al., 1994; Neher and Zucker, 1993; Thomas et al., 1993) observed—

in bovine chromaYn cells and rat melanotrophs—three kinetic secretory phases in

response to [Ca2þ]i steps to �100 mM, reflecting release from diVerent vesicle

pools. Prior exposure to a modest [Ca2þ]i rise primed phasic responses to a

subsequent step in [Ca2þ]i, indicating that [Ca2þ]i not only triggers exocytosis but

also mobilizes vesicles into a docked or releasable status. After exocytosis, another

[Ca2þ]i stimulus often evoked a rapid reduction in membrane capacitance signaling

a [Ca2þ]i-dependent compensatory endocytosis.


The release of synaptic transmitter by action potentials is often enhanced for about

a second after short bouts of presynaptic activity (synaptic facilitation), and formuch

longer after sustained activity lasting minutes (posttetanic potentiation, PTP). We

have used photolytic release of presynaptic Ca2þ from DM-nitrophen to induce

facilitation without electrical activity (Kamiya and Zucker, 1994). Photolysis of

diazo-2 or diazo-4 to terminate the [Ca2þ]i increase lingering briefly after a short

spike train abolished facilitation immediately at crayfish neuromuscular junctions.

PTP induced by longer stimulation, both in this preparation and inAplysia neuronal

synapses (Fischer et al., 1997), was abolished more slowly by rapid reduction of the

prolonged residual [Ca2þ]i resulting frommitochondrial overload (Tang and Zucker,

1997). Thus, facilitation and PTP arise from residual [Ca2þ]i acting on distinct

molecular targets diVerent from the secretory trigger, which is also activated by

Ca2þ. We compared responses to [Ca2þ]i steps on flash photolyzing DMNPE-4 at

weakly transmitting but strongly facilitating neuromuscular junctions to responses at

strongly transmitting but depressible junctions (Millar et al., 2005), and concluded

that the diVerence in response kinetics was best explained by a diVerence in the state

of Ca2þ-dependent priming, such that strongly transmitting synapses were already

preprimed at rest by a priming target tuned to have a higher Ca2þ-sensitivity. Thisled, in turn, to the development of a comprehensive model of synaptic transmission,

facilitation, and depression that comprised three Ca2þ-dependent processes—vesicle

mobilization to docking sites, priming of docked vesicles, and activation of mem-

brane fusion (Pan and Zucker, 2009).

We (Lando and Zucker, 1989) and Heidelberger et al. (1994) were the first to use

DM-nitrophen photolysis to characterize the Ca2þ-cooperativity of secretion at

neuromuscular junctions and retinal bipolar neurons; subsequently, we (Ohnuma

et al., 2001) used NP-EGTA and Kasai et al. (1999) used DM-nitrophen to show

diVerences in the Ca2þ-dependence and sensitivity of peptidergic or aminergic

large dense core vesicle fusion and cholinergic small clear vesicle fusion at central

molluscan synapses and in PC12 cells. Hsu et al. (1996) reported that transmitter

release at squid giant synapses decayed to step [Ca2þ]i increases produced by

NP-EGTA photolysis; subsequent higher steps evoked more release, indicating

transmitter stores had not been depleted, suggesting either an adaptation of the

release process, as the authors proposed, or possibly vesicle heterogeneity in

sensitivity to release, or the operation of mobilization or priming processes

enabling release of previously undocked or unprimed vesicles at higher [Ca2þ]i.Caged Ca2þ photolysis has been used extensively in the last decade in many

elegant experiments, especially from the laboratories of Erwin Neher and Bert

Sakmann, to kinetically characterize in detail the secretory trigger for neurosecre-

tion, primarily at the giant synapse of the calyx of Held (Bollmann and Sakmann,

2005; Bollmann et al., 2000; Felmy et al., 2003a,b; Hosoi et al., 2007; Sakaba et al.,

2005; Schneggenburger and Neher, 2000; Wadel et al., 2007; Wang et al., 2004;

Young and Neher, 2009). Ca2þ uncaging from DM-nitrophen has been used to

probe the kinetics and cooperativity of Ca2þ binding to the secretory trigger,

kinetic consequences of SNARE protein and synaptotagmin mutation, eVects of

56 Robert Zucker

Ca2þ on synaptic facilitation, the dependence of release kinetics on the time-course

of local [Ca2þ]i changes, heterogeneity of vesicle Ca2þ-sensitivity and release

kinetics, the role of Ca2þ in mobilizing vesicles to replenished pools depleted in

synaptic depression, and to compare secretion evoked by global [Ca2þ]i manipula-

tion in uncaging to local [Ca2þ]i influx though voltage-dependent channels to

address the question of distance of the secretory target from Ca2þ channels and

its changes in development. Kinetic studies of the Ca2þ-dependence of secretion

using DM-nitrophen have also been conduced on cochlear hair cells (Beutner et al.,

2001) and photoreceptors (Duncan et al., 2010).

Zoran et al. (1991) used Ca2þ photorelease to study synapse maturation. Spike-

evoked transmitter release begins only several hours after cultured snail neurons

contact a postsynaptic target. DM-nitrophen photolysis showed this developmen-

tal change to result from the delayed increase in Ca2þ-sensitivity of the secretory

machinery.

Long-term potentiation and depression (LTP and LTD) in mammalian cortical

synapses are involved in cognitive processes such as memory consolidation and

spatial learning. We found that a brief but strong postsynaptic [Ca2þ]i elevationwas suYcient to induce LTP in rat hippocampal CA1 synapses onto the injected

neuron, while a more prolonged but modest [Ca2þ]i elevation specifically induced

LTD, and a brief but modest Ca2þ rise could elicit either (Malenka et al., 1988;

Neveu and Zucker, 1996a,b; Yang et al., 1999). By terminating the [Ca2þ]i risefollowing a brief aVerent tetanus by photoactivating the Ca2þ chelator diazo-4, we

showed that postsynaptic [Ca2þ]i must remain elevated for several seconds before it

can induce (Malenka et al., 1992). We also found that long-lasting changes in

synaptic transmission at CA3 hippocampal pyramidal cells can be produced by

postsynaptic [Ca2þ]i elevations induced by DM-nitrophen, NP-EGTA, or

DMNPE-4 photolysis (Wang et al., 2004).

A diVerent form of LTD in cerebellar Purkinje neurons that plays a role in

motor skill learning, parallel fiber synapses are depressed when their activity

coincides with postsynaptic firing, especially when the latter is triggered by climb-

ing fiber input. Lev-Ram et al. (1997) showed that photolytic release of caged Ca2þ

from nitr-7 could replace postsynaptic spiking and that photolytic release of either

caged NO or caged cGMP could replace parallel fiber activity; simultaneous

uncaging of Ca2þ and either NO or cGMP could induce LTD without any

electrical stimulation at all. Kasono and Hirano (1994) showed that a modest

release of Ca2þ from nitr-5 depressed responses to glutamate application to a

dendrite only when the stimuli were temporally paired. Using DMNPE-4,

Tanaka et al. (2007) found that a suYciently high and prolonged Ca2þ elevation

alone could induce LTD, and that the threshold for LTD induction was history-

dependent.

Depolarization-induced suppression of inhibition (DSI) is another form of

cortical synaptic plasticity, which is mediated by activation of postsynaptic endo-

cannabinoid synthesis by activity-induced [Ca2þ]i elevation and subsequent retro-

grade regulation of inhibitory transmitter release. We found identical DSI


sensitivities to uniform postsynaptic [Ca2þ]i elevation by NP-EGTA photolysis vs.

the volume-average of the highly nonuniform [Ca2þ]i elevation on opening voltage-

sensitive Ca2þ channels by depolarizations (Wang and Zucker, 2001), implying

that the enzymatic targets of postsynaptic Ca2þ entering through Ca2þ channels in

activating DSI are not tightly colocalized with the channels—a situation exactly

opposite of the case for Ca2þ activation of classical transmitter release.

Long-lasting synaptic regulation also occurs at developing neuromuscular junc-

tions, where repeated activation of one of two motor neuron inputs results in a

postsynaptic Ca2þ-dependent compensatory or homeostatic reduction in presyn-

aptic transmitter release to action potentials at terminals facing the activated

receptors. Using focal DM-nitrophen or nitr-5 photolysis to mimic the localized

postsynaptic [Ca2þ]i elevation seen to accompany the stimulus normally used to

induce this selective persistent depression, we were able to induce a similar syn-

apse-specific modification (Cash et al., 1996a). Subsequently, synapses made by the

modified motor neuron onto other muscle fibers also became depressed by the

spread of an unidentified presynaptic intracellular signal (Cash et al., 1996b).

D. Other Applications

The tight regulation of cytoplasmic [Ca2þ]i is essential for ensuring that Ca2þ can

act reliably and eYciently as a localized second messenger of a huge variety of

cellular processes. Endogenous buVers play a defining role in this process, and an

appreciation of the functions of these buVers and their characteristics (aYnities,

binding kinetics, mobility, and localization) is crucial to our understanding how

Ca2þ performs its central cellular functions. Use of photosensitive Ca2þ chelators

has become an important tool in the estimation of cytoplasmic buVer characteristics,and much eVort has gone into developing procedures and protocols for defining

them with some precision. Some of the best examples of this sort of analysis come

from the laboratories of Stephen Bolsover, Istvan Mody and Julio Vergara, and

ErwinNeher, whose papers should be consulted for the analytical details (Faas et al.,

2007; Fleet et al., 1998; Nagerl et al., 2000; Naraghi et al., 1998; Xu et al., 1997).

In addition to these major areas of application of caged Ca2þ chelators, this

method of [Ca2þ]i manipulation has been used to address an increasingly diverse

range of biological problems. Nitr and diazo compounds were inserted by AM

loading into fibroblasts that were activated by mitogenic stimulation to produce

[Ca2þ]i oscillations monitored using f1uo-3 (Harootunian et al., 1988). Photore-

lease of Ca2þ from nitr-5 enhanced and accelerated the oscillations, whereas

release of caged chelator by photolysis of diazo-2 inhibited them. Nitr-7 photolysis

caused not only an immediate rise in [Ca2þ]i liberated from the photolyzed chela-

tor, but also elicited a later rise in [Ca2þ]i (Harootunian et al., 1991). This eVect wasshown, pharmacologically, to be caused by IP3-sensitive stores, suggesting that an

interaction between [Ca2þ]i and these stores underlies the [Ca2þ]i oscillations.Photorelease of Ca2þ from DM-nitrophen has been used to study the binding

kinetics of Ca2þ to the Ca2þ-ATPase of sarcoplasmic reticulum vesicles (DeLong

58 Robert Zucker

et al., 1990). The relaxation of the [Ca2þ] step, measured by arsenazo spectropho-

tometry after photolysis, revealed the kinetics of binding to the ATPase. Changes

in the Fourier transform infrared spectrum consequent to photorelease of Ca2þ

from nitr-5 provided information on structural changes in the ATPase after

binding Ca2þ (Buchet et al., 1991, 1992). In a final application to the study of

enzyme conformational changes, photolysis of Mg2þ-loaded DM-nitrophen was

used to form Mg2þ-ATP rapidly to activate Naþ/Kþ exchange, the state of which

was monitored by fluorescence of aminostyrylpyridinium dyes (Forbush and

Klodos, 1991). Rate-limiting steps were measured at 45 s�1 by this method.

Ca2þ has been implicated in the control of filopodial activity in the responses of

growth cones of developing neurons to environmental cues. Pioneer neurons lay

out peripheral aVerent pathways in developing grasshoppers. We loaded pioneer

neurons by de-esterification of the AM esters of DM-nitrophen and calcium green

(Lau et al., 1999) and showed that elevation of local [Ca2þ]i in a growth cone to

� 1 mM for just 10 s was suYcient to activate subsequent filopodial prolongation

and induce the formation of new filopodia at spots with high actin concentration

(labeled with rhodamine-phalloidin).

In other applications, Gilroy et al. (1991) and Fricker et al. (1991) microinjected

Ca2þ-loaded nitr-5 into guard cells of lily leaves and showed that photorelease of

about 600 nM intracellular Ca2þ (measured with fluo-3) initiated stomatal pore

closure. Kao et al. (1990) loaded Swiss 3T3 fibroblasts with nitr-5/AM and showed

that photolysis that elevated [Ca2þ]i by hundreds of nanomolar (measured by

fluo-3) triggered nuclear envelope breakdown, an early step in mitosis, while

having little eVect on the metaphase to anaphase transition. Control experiments

using nitr-9 showed no eVect of reactive photochemical intermediates or products.

Groigno and Whitaker (1998) initiated chromosome disjunction and segregation

in embryonic sea urchin cells by Ca2þ photorelease from NP-EGTA or by photol-

ysis of caged IP3. Ca2þ buffers prevented chromatid separation but not the later

stages of anaphase, indicating a specific role for Ca2þ in early anaphase chromo-

some disjunction. Tisa and Adler (1992) used electroporation to introduce Ca2þ-loaded nitr-5 or DM-nitrophen into Escherichia coli bacteria, and showed that

elevation of [Ca2þ]i enhanced tumbling behavior characteristic of chemotaxis

whereas photorelease of caged chelator from diazo-2 decreased tumbling. Photo-

lysis of diazo-3, which reduces pH without aVecting [Ca2þ]i, caused only a small

increase in tumbling. Mutants with methyl-accepting chemotaxis receptor proteins

still responded to Ca2þ, whereas mutants of specific Che proteins did not, indicat-

ing that the action of these proteins lay downstream of the Ca2þ signal.

X. Conclusions

Interest in photolabile Ca2þ chelators has been intense. Their range of applica-

tion has broadened well beyond the original nerve, muscle, and fibroblast prepara-

tions. They remain one of the most valuable tools for the precise definition of


calcium’s roles and mechanisms of action in cell biology. They have attracted and

challenged some of the best minds in physiology, resulting in great conceptual and

skillful sophistication in the rapid evolution of this technology, which shows little

sign of abating.

Acknowledgments

I thank Steve Adams for valuable discussion and Joseph Kao for drawings of chelator structures.

The research done in my laboratory in this area was supported primarily by National Institutes of

Health Grant NS 15114.

References

Adams, S. R., and Tsien, R. Y. (1993). Controlling cell chemistry with caged compounds. Annu. Rev.

Physiol. 55, 755–784.

Adams, S. R., Kao, J. P. Y., Grynkiewicz, G., Minta, A., and Tsien, R. Y. (1988). Biologically useful

chelators that release Ca2þ upon illumination. J. Am. Chem. Soc. 110, 3212–3220.

Adams, S. R., Kao, J. P. Y., and Tsien, R. Y. (1989). Biologically useful chelators that take up Ca2þ

upon illumination. J. Am. Chem. Soc. 111, 7957–7968.

Adams, S. R., Lev-Ram, V., and Tsien, R. Y. (1997). A new caged Ca2þ, azid-1, is far more photosensi-

tive than nitrobenzyl-based chelators. Chem. Biol. 4, 867–878.

Ashley, C. C., GriYths, P. J., Lea, T. J., Mulligan, I. P., Palmer, R. E., and Simnett, S. J. (1991). Use of

fluorescent TnC derivatives and ‘‘caged’’ compounds to study cellular Ca2þ phenomena. In ‘‘Cellular

Calcium: A Practical Approach,’’ (J. G. McCormack, and P. H. Cobbold, eds.), pp. 177–203. Oxford

University Press, New York.

Ashley, C. C., Mulligan, I. P., and Lea, T. J. (1991). Ca2þ and activation mechanisms in skeletal muscle.

Q. Rev. Biophys. 24, 1–73.

Ashley, C. C., Lea, T. J., Mulligan, I. P., Palmer, R. E., and Simnett, S. J. (1993). Activation and

relaxation mechanisms in single muscle fibres. Adv. Exp. Med. Biol. 332, 97–114.

Ayer, R. K., Jr., and Zucker, R. S. (1999). Magnesium binding to DM-nitrophen and its eVect on the

photorelease of calcium. Biophys. J. 77, 3384–3393.

Backx, P. H., O’Rourke, B., and Marban, E. (1991). Flash photolysis of magnesium-DM-nitrophen in

heart cells. A novel approach to probe magnesium- and ATP-dependent regulation of calcium

channels. Am. J. Hypertens. 4, 416S–421S.

Bates, S. E., and Gurney, A. M. (1993). Ca2þ-dependent block and potentiation of L-type calcium

current in guinea-pig ventricular myocytes. J. Physiol. (Lond.) 466, 345–365.

Bernardinelli, Y., Haeberli, C., and Chatton, J. Y. (2005). Flash photolysis using a light emitting diode:

An eYcient, compact, and aVordable solution. Cell Calcium 37, 565–572.

Beutner, D., Voets, T., Neher, E., and Moser, T. (2001). Calcium dependence of exocytosis and

endocytosis at the cochlear inner hair cell aVerent synapse. Neuron 29, 681–690.

Bollmann, J. H., and Sakmann, B. (2005). Control of synaptic strength and timing by the release-site

Ca2þ signal. Nat. Neurosci. 8, 426–434.

Bollmann, J. H., Sakmann, B., and Borst, J. G. (2000). Calcium sensitivity of glutamate release in a

calyx-type terminal. Science 289, 953–957.

Brown, E. B., Shear, J. B., Adams, S. R., Tsien, R. Y., and Webb, W. W. (1999). Photolysis of caged

calcium in femtoliter volumes using two-photon excitation. Biophys. J. 76, 489–499.

Bruns, D., Engert, F., and Lux, H. D. (1993). A fast activating presynaptic reuptake current during

serotonergic transmission in identified neurons of Hirudo. Neuron 10, 559–572.

Buchet, R., Jona, I., andMartonosi, A. (1991). Ca2þ release from caged-Ca2þ alters the FTIR spectrum

of sarcoplasmic reticulum. Biochim. Biophys. Acta 1069, 209–217.

60 Robert Zucker

Buchet, R., Jona, I., and Martonosi, A. (1992). The eVect of dicyclohexylcarbodiimide and cyclopia-

zonic acid on the diVerence FTIR spectra of sarcoplasmic reticulum induced by photolysis of caged-

ATP and caged-Ca2þ. Biochim. Biophys. Acta 1104, 207–214.

Carl, A., McHale, N. G., Publicover, N. G., and Sanders, K. M. (1990). Participation of Ca2þ-activatedKþ channels in electrical activity of canine gastric smooth muscle. J. Physiol. (Lond.) 429, 205–221.

Cash, S., Dan, Y., Poo, M.-M., and Zucker, R. (1996). Postsynaptic elevation of calcium induces

persistent depression of developing neuromuscular synapses. Neuron 16, 745–754.

Cash, S., Zucker, R. S., and Poo, M.-M. (1996). Spread of synaptic depression mediated by presynaptic

cytoplasmic signaling. Science 272, 998–1001.

Charnet, P., Richard, S., Gurney, A. M., Ouadid, H., Tiaho, F., and Nargeot, J. (1991). Modulation of

Ca currents in isolated frog atrial cells studied with photosensitive probes. Regulation by cAMP and

Ca2þ: A common pathway? J. Mol. Cell. Cardiol. 23, 343–356.

Delaney, K. R., and Zucker, R. S. (1990). Calcium released by photolysis of DM-nitrophen stimulates

transmitter release at squid giant synapse. J. Physiol. (Lond.) 426, 473–498.

DeLong, L. J., Phillips, C. M., Kaplan, J. H., Scarpa, A., and Blasie, J. K. (1990). A new method for

monitoring the kinetics of calcium binding to the sarcoplasmic reticulumCa2þ-ATPase employing the

flash-photolysis of caged-calcium. J. Biochem. Biophys. Methods 21, 333–339.

DelPrincipe, F., Egger, M., Ellis-Davies, G. C., and Niggli, E. (1999). Two-photon and UV-laser flash

photolysis of the Ca2þ cage, dimethoxynitrophenyl-EGTA-4. Cell Calcium 25, 85–91.

Denk, W. (1997). Pulsing mercury arc lamps for uncaging and fast imaging. J. Neurosci. Methods 72,

39–42.

Denk, W., Strickler, J. H., and Webb, W. W. (1990). Two-photon laser scanning fluorescence micros-

copy. Science 248, 73–76.

Duncan, G., Rabl, K., Gemp, I., Heidelberger, R., and Thoreson, W. B. (2010). Quantitative analysis of

synaptic release at the photoreceptor synapse. Biophys. J. 98, 2102–2110.

Eberhard, M., and Erne, P. (1991). Calcium binding to fluorescent calcium indicators: Calcium green,

calcium orange and calcium crimson. Biochem. Biophys. Res. Commun. 180, 209–215.

Eberius, C., and Schild, D. (2001). Local photolysis using tapered quartz fibres. Pflugers Arch. 443,

323–330.

Ellis-Davies, G. C. (2003). Development and application of caged calcium. Methods Enzymol. 360,

226–238.

Ellis-Davies, G. C. (2006). DM-nitrophen AM is caged magnesium. Cell Calcium 39, 471–473.

Ellis-Davies, G. C., and Barsotti, R. J. (2006). Tuning caged calcium: Photolabile analogues of EGTA

with improved optical and chelation properties. Cell Calcium 39, 75–83.

Ellis-Davies, G. C. R., and Kaplan, J. H. (1988). A new class of photolabile chelators for the rapid

release of divalent cations: Generation of caged Ca and caged Mg. J. Org. Chem. 53, 1966–1969.

Ellis-Davies, G. C., and Kaplan, J. H. (1994). Nitrophenyl-EGTA, a photolabile chelator that selective-

ly binds Ca2þwith high aYnity and releases it rapidly upon photolysis. Proc. Natl. Acad. Sci. USA 91,

187–191.

Ellis-Davies, G. C., Kaplan, J. H., and Barsotti, R. J. (1996). Laser photolysis of caged calcium: Rates of

calcium release by nitrophenyl-EGTA and DM-nitrophen. Biophys. J. 70, 1006–1016.

Engert, F., Paulus, G. G., and BonhoeVer, T. (1996). A low-cost UV laser for flash photolysis of caged

compounds. J. Neurosci. Methods 66, 47–54.

Escobar, A. L., Cifuentes, F., and Vergara, J. L. (1995). Detection of Ca2þ-transients elicited by flash

photolysis of DM-nitrophen with a fast calcium indicator. FEBS Lett. 364, 335–338.

Escobar, A. L., Velez, P., Kim, A. M., Cifuentes, F., Fill, M., and Vergara, J. L. (1997). Kinetic

properties of DM-nitrophen and calcium indicators: Rapid transient response to flash photolysis.

Pflugers Arch. 434, 615–631.

Faas, G. C., Karacs, K., Vergara, J. L., and Mody, I. (2005). Kinetic properties of DM-nitrophen

binding to calcium and magnesium. Biophys. J. 88, 4421–4433.

Faas, G. C., Schwaller, B., Vergara, J. L., andMody, I. (2007). Resolving the fast kinetics of cooperative

binding: Ca2þ buVering by calretinin. PLoS Biol. 5, e311.


Felmy, F., Neher, E., and Schneggenburger, R. (2003). Probing the intracellular calcium sensitivity of

transmitter release during synaptic facilitation. Neuron 37, 801–811.

Felmy, F., Neher, E., and Schneggenburger, R. (2003). The timing of phasic transmitter release is Ca2þ-dependent and lacks a direct influence of presynaptic membrane potential. Proc. Natl. Acad. Sci. USA

100, 15200–15205.

Ferenczi, M. A., Goldman, Y. E., and Trentham, D. R. (1989). Relaxation of permeabilized, isolated

muscle-fibers of the rabbit by rapid chelation of Ca-2þ-ions through laser-pulse photolysis of caged-

Bapta. J. Physiol. (Lond.) 418, P155.

Fischer, T. M., Zucker, R. S., and Carew, T. J. (1997). Activity-dependent potentiation of synaptic

transmission from L30 inhibitory interneurons of aplysia depends on residual presynaptic Ca2þ but

not on postsynaptic Ca2þ. J. Neurophysiol. 78, 2061–2071.

Fleet, A., Ellis-Davies, G., and Bolsover, S. (1998). Calcium buVering capacity of neuronal cell cytosol

measured by flash photolysis of calcium buVer NP-EGTA. Biochem. Biophys. Res. Commun. 250,

786–790.

Fogelson, A. L., and Zucker, R. S. (1985). Presynaptic calcium diVusion from various arrays of single

channels. Implications for transmitter release and synaptic facilitation. Biophys. J. 48, 1003–1017.

Forbush, B., 3rd, and Klodos, I. (1991). Rate-limiting steps in Na translocation by the Na/K pump. Soc.

Gen. Physiol. Ser. 46, 210–225.

Fricker, M. D., Gilroy, S., Read, N. D., and Trewavas, A. J. (1991). Visualisation and measurement of

the calcium message in guard cells. Symp. Soc. Exp. Biol. 45, 177–190.

Fryer, M. W., and Zucker, R. S. (1993). Ca2þ-dependent inactivation of Ca2þ current in Aplysia

neurons: Kinetic studies using photolabile Ca2þ chelators. J. Physiol. (Lond.) 464, 501–528.

Gilroy, S., Fricker, M. D., Read, N. D., and Trewavas, A. J. (1991). Role of calcium in signal

transduction of Commelina guard cells. Plant Cell 3, 333–344.

Godwin, D. W., Che, D., O’Malley, D. M., and Zhou, Q. (1997). Photostimulation with caged

neurotransmitters using fiber optic lightguides. J. Neurosci. Methods 73, 91–106.

Goldman, Y. E., Hibberd, M. G., and Trentham, D. R. (1984). Relaxation of rabbit psoas muscle fibres

from rigor by photochemical generation of adenosine-50-triphosphate. J. Physiol. (Lond.) 354,

577–604.

Grell, E., Lewitzki, E., Ruf, H., Bamberg, E., Ellis-Davies, G. C., Kaplan, J. H., and deWeer, P. (1989).

Caged-Ca2þ: A new agent allowing liberation of free Ca2þ in biological systems by photolysis. Cell.

Mol. Biol. 35, 515–522.

Groigno, L., andWhitaker, M. (1998). An anaphase calcium signal controls chromosome disjunction in

early sea urchin embryos. Cell 92, 193–204.

Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985). A new generation of Ca2þ indicators with greatly


Gurney, A. (1991). Photolabile calcium buVers to selectively activate calcium-dependent processes.

In ‘‘Cellular Neurobiology: A Practical Approach,’’ (J. Chad, and H. Wheal, eds.), pp. 153–177. IRL

Press, New York.

Gurney, A. (1993). Photolabile caged compounds. In ‘‘Fluorescent and Luminescent Probes for

Biological Activity—A Practical Guide to Technology for Quantitative Real-Time Analysis,’’

(W. T. Mason, ed.), pp. 335–348. Academic Press, San Diego.

Gurney, A. M., Tsien, R. Y., and Lester, H. A. (1987). Activation of a potassium current by rapid

photochemically generated step increases of intracellular calcium in rat sympathetic neurons. Proc.

Natl. Acad. Sci. USA 84, 3496–3500.

Gurney, A. M., Charnet, P., Pye, J. M., and Nargeot, J. (1989). Augmentation of cardiac calcium

current by flash photolysis of intracellular caged-Ca2þ molecules. Nature 341, 65–68.

Gyorke, S., and Fill, M. (1993). Ryanodine receptor adaptation: Control mechanism of Ca2þ-inducedCa2þ release in heart. Science 260, 807–809.

Hadley, R. W., and Lederer, W. J. (1991). Ca2þ and voltage inactivate Ca2þ channels in guinea-pig

ventricular myocytes through independent mechanisms. J. Physiol. (Lond.) 444, 257–268.

62 Robert Zucker

Hadley, R. W., Kirby, M. S., Lederer, W. J., and Kao, J. P. (1993). Does the use of DM-nitrophen,

nitr-5, or diazo-2 interfere with the measurement of indo-1 fluorescence? Biophys. J. 65, 2537–2546.

Harootunian, A. T., Kao, J. P., and Tsien, R. Y. (1988). Agonist-induced calcium oscillations in

depolarized fibroblasts and their manipulation by photoreleased Ins(1, 4, 5)P3, Caþþ, and CaþþbuVer. Cold Spring Harb. Symp. Quant. Biol. 53(Pt. 2), 935–943.

Harootunian, A. T., Kao, J. P., Paranjape, S., Adams, S. R., Potter, B. V., and Tsien, R. Y. (1991).

Cytosolic Ca2þ oscillations in REF52 fibroblasts: Ca2þ-stimulated IP3 production or voltage-depen-

dent Ca2þ channels as key positive feedback elements. Cell Calcium 12, 153–164.

Haydon, P. G., Man-Son-Hing, H., Doyle, R. T., and Zoran, M. (1991). FMRFamide modulation of

secretory machinery underlying presynaptic inhibition of synaptic transmission requires a pertussis

toxin-sensitive G-protein. J. Neurosci. 11, 3851–3860.

Heidelberger, R., Heinemann, C., Neher, E., and Matthews, G. (1994). Calcium dependence of the rate

of exocytosis in a synaptic terminal. Nature 371, 513–515.

Heinemann, C., Chow, R. H., Neher, E., and Zucker, R. S. (1994). Kinetics of the secretory response in

bovine chromaYn cells following flash photolysis of caged Ca2þ. Biophys. J. 67, 2546–2557.Hochner, B., Parnas, H., and Parnas, I. (1989). Membrane depolarization evokes neurotransmitter

release in the absence of calcium entry. Nature 342, 433–435.

Hosoi, N., Sakaba, T., and Neher, E. (2007). Quantitative analysis of calcium-dependent vesicle

recruitment and its functional role at the calyx of Held synapse. J. Neurosci. 27, 14286–14298.

Hsu, S. F., Augustine, G. J., and Jackson, M. B. (1996). Adaptation of Ca2þ-triggered exocytosis in

presynaptic terminals. Neuron 17, 501–512.

Iino, M., and Endo, M. (1992). Calcium-dependent immediate feedback control of inositol 1,4,5-

triphosphate-induced Ca2þ release. Nature 360, 76–78.

Johnson, B. D., and Byerly, L. (1993). Photo-released intracellular Ca2þ rapidly blocks Ba2þ current in

Lymnaea neurons. J. Physiol. (Lond.) 462, 321–347.

Kamiya, H., and Zucker, R. S. (1994). Residual Ca2þ and short-term synaptic plasticity. Nature 371,

603–606.

Kao, J. P., and Adams, S. R. (1993). Photosensitive caged compounds: Design, properties, and

biological applications. In ‘‘Optical Microscopy: New Technologies and Applications,’’

(B. Herman, and J. J. Lemasters, eds.), pp. 27–85. Academic Press, New York.

Kao, J. P., Harootunian, A. T., and Tsien, R. Y. (1989). Photochemically generated cytosolic calcium

pulses and their detection by fluo-3. J. Biol. Chem. 264, 8179–8184.

Kao, J. P., Alderton, J. M., Tsien, R. Y., and Steinhardt, R. A. (1990). Active involvement of Ca2þ in

mitotic progression of Swiss 3T3 fibroblasts. J. Cell Biol. 111, 183–196.

Kaplan, J. H. (1990). Photochemical manipulation of divalent cation levels. Annu. Rev. Physiol. 52,

897–914.

Kaplan, J. H., and Ellis-Davies, G. C. (1988). Photolabile chelators for the rapid photorelease of

divalent cations. Proc. Natl. Acad. Sci. USA 85, 6571–6575.

Kaplan, J. H., and Somlyo, A. P. (1989). Flash photolysis of caged compounds: New tools for cellular

physiology. Trends Neurosci. 12, 54–59.

Kaplan, J. H., Forbush, B., 3rd, and HoVman, J. F. (1978). Rapid photolytic release of adenosine 50-triphosphate from a protected analogue: Utilization by the Na:K pump of human red blood cell

ghosts. Biochemistry 17, 1929–1935.

Kasai, H., Kishimoto, T., Liu, T. T., Miyashita, Y., Podini, P., Grohovaz, F., and Meldolesi, J. (1999).

Multiple and diverse forms of regulated exocytosis in wild-type and defective PC12 cells. Proc. Natl.

Acad. Sci. USA 96, 945–949.

Kasono, K., and Hirano, T. (1994). Critical role of postsynaptic calcium in cerebellar long-term

depression. Neuroreport 6, 17–20.

Kentish, J. C., Barsotti, R. J., Lea, T. J., Mulligan, I. P., Patel, J. R., and Ferenczi, M. A. (1990).

Calcium release from cardiac sarcoplasmic reticulum induced by photorelease of calcium or Ins(1, 4,

5)P3. Am. J. Physiol. 258, H610–H615.


Kimitsuki, T., and Ohmori, H. (1992). The eVect of caged calcium release on the adaptation of the

transduction current in chick hair cells. J. Physiol. (Lond.) 458, 27–40.

Konishi, M., Hollingworth, S., Harkins, A. B., and Baylor, S. M. (1991). Myoplasmic calcium tran-

sients in intact frog skeletal muscle fibers monitored with the fluorescent indicator furaptra. J. Gen.

Physiol. 97, 271–301.

Kurebayashi, N., Harkins, A. B., and Baylor, S. M. (1993). Use of fura red as an intracellular calcium

indicator in frog skeletal muscle fibers. Biophys. J. 64, 1934–1960.

Lancaster, B., and Zucker, R. S. (1994). Photolytic manipulation of Ca2þ and the time course of slow,

Ca2þ-activated Kþ current in rat hippocampal neurones. J. Physiol. (Lond.) 475, 229–239.

Lando, L., and Zucker, R. S. (1989). ‘‘Caged calcium’’ in Aplysia pacemaker neurons. Characterization

of calcium-activated potassium and nonspecific cation currents. J. Gen. Physiol. 93, 1017–1060.

Lando, L., and Zucker, R. S. (1994). Ca2þ cooperativity in neurosecretion measured using photolabile

Ca2þ chelators. J. Neurophysiol. 72, 825–830.

Lannergren, J., and Arner, A. (1992). Relaxation rate of intact striated muscle fibres after flash

photolysis of a caged calcium chelator (diazo-2). J. Muscle Res. Cell Motil. 13, 630–634.

Lau, P. M., Zucker, R. S., and Bentley, D. (1999). Induction of filopodia by direct local elevation of

intracellular calcium ion concentration. J. Cell Biol. 145, 1265–1275.

Lea, T. J., and Ashley, C. C. (1990). Ca2þ release from the sarcoplasmic reticulum of barnacle

myofibrillar bundles initiated by photolysis of caged Ca2þ. J. Physiol. (Lond.) 427, 435–453.

Lev-Ram, V., Jiang, T., Wood, J., Lawrence, D. S., and Tsien, R. Y. (1997). Synergies and coincidence

requirements between NO, cGMP, and Ca2þ in the induction of cerebellar long-term depression.

Neuron 18, 1025–1038.

Malenka, R. C., Kauer, J. A., Zucker, R. S., and Nicoll, R. A. (1988). Postsynaptic calcium is suYcient

for potentiation of hippocampal synaptic transmission. Science 242, 81–84.

Malenka, R. C., Lancaster, B., and Zucker, R. S. (1992). Temporal limits on the rise in postsynaptic

calcium required for the induction of long-term potentiation. Neuron 9, 121–128.

Man-Son-Hing, H., Zoran, M. J., Lukowiak, K., and Haydon, P. G. (1989). A neuromodulator of

synaptic transmission acts on the secretory apparatus as well as on ion channels.Nature 341, 237–239.

Marrion, N. V., Zucker, R. S., Marsh, S. J., and Adams, P. R. (1991). Modulation of M-current by

intracellular Ca2þ. Neuron 6, 533–545.

Matveev, V., Sherman, A., and Zucker, R. S. (2002). New and corrected simulations of synaptic

facilitation. Biophys. J. 83, 1368–1373.

Matveev, V., Zucker, R. S., and Sherman, A. (2004). Facilitation through buVer saturation: Constraints

on endogenous buVering properties. Biophys. J. 86, 2691–2709.

Matveev, V., Bertram, R., and Sherman, A. (2006). Residual bound Ca2þ can account for the eVects of

Ca2þ buVers on synaptic facilitation. J. Neurophysiol. 96, 3389–3397.

Matveev, V., Bertram, R., and Sherman, A. (2009). Ca2þ current versus Ca2þ channel cooperativity of

exocytosis. J. Neurosci. 29, 12196–12209.

McCray, J. A., and Trentham, D. R. (1989). Properties and uses of photoreactive caged compounds.

Annu. Rev. Biophys. Biophys. Chem. 18, 239–270.

McCray, J. A., Fidler-Lim, N., Ellis-Davies, G. C., and Kaplan, J. H. (1992). Rate of release of Ca2þ

following laser photolysis of the DM-nitrophen-Ca2þ complex. Biochemistry 31, 8856–8861.

Millar, A. G., Zucker, R. S., Ellis-Davies, G. C., Charlton, M. P., and Atwood, H. L. (2005). Calcium

sensitivity of neurotransmitter release diVers at phasic and tonic synapses. J. Neurosci. 25, 3113–3125.

Minta, A., and Tsien, R. Y. (1989). Fluorescent indicators for cytosolic sodium. J. Biol. Chem. 264,

19449–19457.

Momotake, A., Lindegger, N., Niggli, E., Barsotti, R. J., and Ellis-Davies, G. C. (2006). The nitrodi-

benzofuran chromophore: A new caging group for ultra-eYcient photolysis in living cells. Nat.

Methods 3, 35–40.

Morad, M., Davies, N. W., Kaplan, J. H., and Lux, H. D. (1988). Inactivation and block of calcium

channels by photo-released Ca2þ in dorsal root ganglion neurons. Science 241, 842–844.

64 Robert Zucker

Mulkey, R. M., and Zucker, R. S. (1991). Action potentials must admit calcium to evoke transmitter

release. Nature 350, 153–155.

Mulkey, R. M., and Zucker, R. S. (1993). Calcium released by photolysis of DM-nitrophen triggers

transmitter release at the crayfish neuromuscular junction. J. Physiol. (Lond.) 462, 243–260.

Mulligan, I. P., and Ashley, C. C. (1989). Rapid relaxation of single frog skeletal muscle fibres following

laser flash photolysis of the caged calcium chelator, diazo-2. FEBS Lett. 255, 196–200.

Nabauer, M., and Morad, M. (1990). Ca2þ-induced Ca2þ release as examined by photolysis of caged

Ca2þ in single ventricular myocytes. Am. J. Physiol. 258, C189–C193.

Nabauer, M., Ellis-Davies, G. C., Kaplan, J. H., and Morad, M. (1989). Modulation of Ca2þ channel

selectivity and cardiac contraction by photorelease of Ca2þ. Am. J. Physiol. 256, H916–H920.

Nagerl, U. V., Novo, D., Mody, I., and Vergara, J. L. (2000). Binding kinetics of calbindin-D(28k)

determined by flash photolysis of caged Ca2þ. Biophys. J. 79, 3009–3018.Naraghi, M., Muller, T. H., and Neher, E. (1998). Two-dimensional determination of the cellular Ca2þ

binding in bovine chromaYn cells. Biophys. J. 75, 1635–1647.

Neher, E., and Zucker, R. S. (1993). Multiple calcium-dependent processes related to secretion in bovine

chromaYn cells. Neuron 10, 21–30.

Neveu, D., and Zucker, R. S. (1996). Long-lasting potentiation and depression without presynaptic

activity. J. Neurophysiol. 75, 2157–2160.

Neveu, D., and Zucker, R. S. (1996). Postsynaptic levels of [Ca2þ]i needed to trigger LTD and LTP.

Neuron 16, 619–629.

Nowycky, M. C., and Pinter, M. J. (1993). Time courses of calcium and calcium-bound buVers

following calcium influx in a model cell. Biophys. J. 64, 77–91.

O’Neill, S. C., Mill, J. G., and Eisner, D. A. (1990). Local activation of contraction in isolated rat

ventricular myocytes. Am. J. Physiol. 258, C1165–C1168.

O’Rourke, B., Backx, P. H., and Marban, E. (1992). Phosphorylation-independent modulation of

L-type calcium channels by magnesium-nucleotide complexes. Science 257, 245–248.

Ogden, D. (1988). Cell physiology. Answer in a flash. Nature 336, 16–17.

Ogden, D. C., Capiod, T., Walker, J. W., and Trentham, D. R. (1990). Kinetics of the conductance

evoked by noradrenaline, inositol trisphosphate or Ca2þ in guinea-pig isolated hepatocytes.

J. Physiol. (Lond.) 422, 585–602.

Ohnuma, K., Whim, M. D., Fetter, R. D., Kaczmarek, L. K., and Zucker, R. S. (2001). Presynaptic

target of Ca2þ action on neuropeptide and acetylcholine release in Aplysia californica. J. Physiol.

(Lond.) 535, 647–662.

Palmer, R. E., Mulligan, I. P., Nunn, C., and Ashley, C. C. (1990). Striated scallop muscle relaxation:

Fast force transients produced by photolysis of Diazo-2. Biochem. Biophys. Res. Commun. 168,

295–300.

Palmer, R. E., Simnett, S. J., Mulligan, I. P., and Ashley, C. C. (1991). Skeletal muscle relaxation with

diazo-2: The eVect of altered pH. Biochem. Biophys. Res. Commun. 181, 1337–1342.

Pan, B., and Zucker, R. S. (2009). A general model of synaptic transmission and short-term plasticity.

Neuron 62, 539–554.

Parker, I. (1992). Caged intracellular messengers and the inositol phosphate signaling pathway.

In ‘‘Intracellular Messengers. Neuromethods Vol. 20,’’ (A. A. Boulton, G. B. Baker, and

C. W. Taylor, eds.), pp. 369–396. Humana Press, Totowa, NJ.

Parsons, T. D., Ellis-Davies, G. C., and Almers, W. (1996). Millisecond studies of calcium-dependent

exocytosis in pituitary melanotrophs: Comparison of the photolabile calcium chelators nitrophenyl-

EGTA and DM-nitrophen. Cell Calcium 19, 185–192.

Rapp, G., and Guth, K. (1988). A low cost high intensity flash device for photolysis experiments.


Rebrik, T. I., Kotelnikova, E. A., and Korenbrot, J. I. (2000). Time course and Ca2þ dependence of

sensitivity modulation in cyclic GMP-gated currents of intact cone photoreceptors. J. Gen. Physiol.

116, 521–534.


Sah, P., and Clements, J. D. (1999). Photolytic manipulation of [Ca2þ]i reveals slow kinetics of

potassium channels underlying the after hyperpolarization in hippocampal pyramidal neurons.

J. Neurosci. 19, 3657–3664.

Sakaba, T., Stein, A., Jahn, R., and Neher, E. (2005). Distinct kinetic changes in neurotransmitter

release after SNARE protein cleavage. Science 309, 491–494.

Sala, F., and Hernandez-Cruz, A. (1990). Calcium diVusion modeling in a spherical neuron. Relevance

of buVering properties. Biophys. J. 57, 313–324.

Scarpa, A., Brinley, F. J., Jr., TiVert, T., and Dubyak, G. R. (1978). Metallochromic indicators of

ionized calcium. Ann. NY Acad. Sci. 307, 86–112.

Schneggenburger, R., and Neher, E. (2000). Intracellular calcium dependence of transmitter release

rates at a fast central synapse. Nature 406, 889–893.

Simon, S. M., and Llinas, R. R. (1985). Compartmentalization of the submembrane calcium activity

during calcium influx and its significance in transmitter release. Biophys. J. 48, 485–498.

Stern, M. D. (1992). BuVering of calcium in the vicinity of a channel pore. Cell Calcium 13, 183–192.

Stockbridge, N., andMoore, J. W. (1984). Dynamics of intracellular calcium and its possible relationship

to phasic transmitter release and facilitation at the frog neuromuscular junction. J. Neurosci. 4,

803–811.

Tanaka, K., Khiroug, L., Santamaria, F., Doi, T., Ogasawara, H., Ellis-Davies, G. C., Kawato,M., and

Augustine, G. J. (2007). Ca2þ requirements for cerebellar long-term synaptic depression: role for a

postsynaptic leaky integrator. Neuron 54, 787–800.

Tang, Y., and Zucker, R. S. (1997). Mitochondrial involvement in post-tetanic potentiation of synaptic

transmission. Neuron 18, 483–491.

Tang, Y., Schlumpberger, T., Kim, T., Lueker, M., and Zucker, R. S. (2000). EVects of mobile buVers

on facilitation: Experimental and computational studies. Biophys. J. 78, 2735–2751.

Thomas, P., Wong, J. G., and Almers, W. (1993). Millisecond studies of secretion in single rat pituitary

cells stimulated by flash photolysis of caged Ca2þ. EMBO J. 12, 303–306.

Tisa, L. S., and Adler, J. (1992). Calcium ions are involved in Escherichia coli chemotaxis. Proc. Natl.

Acad. Sci. USA 89, 11804–11808.

Tsien, R. Y., and Zucker, R. S. (1986). Control of cytoplasmic calcium with photolabile tetracarbox-

ylate 2-nitrobenzhydrol chelators. Biophys. J. 50, 843–853.

Valdeolmillos, M., O’Neill, S. C., Smith, G. L., and Eisner, D. A. (1989). Calcium-induced calcium

release activates contraction in intact cardiac cells. Pflugers Arch. 413, 676–678.

Velez, P., Gyorke, S., Escobar, A. L., Vergara, J., and Fill, M. (1997). Adaptation of single cardiac

ryanodine receptor channels. Biophys. J. 72, 691–697.

Wadel, K., Neher, E., and Sakaba, T. (2007). The coupling between synaptic vesicles and Ca2þ channels

determines fast neurotransmitter release. Neuron 53, 563–575.

Walker, J. W. (1991). Caged molecules activated by light. In ‘‘Cellular Neurobiology: A Practical

Approach,’’ (J. Chad, and H. Wheal, eds.), pp. 179–203. IRL Press, New York.

Wang, S. S., and Augustine, G. J. (1995). Confocal imaging and local photolysis of caged compounds:

Dual probes of synaptic function. Neuron 15, 755–760.

Wang, J., and Zucker, R. S. (2001). Photolysis-induced suppression of inhibition in rat hippocampal

CA1 pyramidal neurons. J. Physiol. (Lond.) 533, 757–763.

Wang, J., Yeckel, M. F., Johnston, D., and Zucker, R. S. (2004). Photolysis of postsynaptic caged Ca2þ

can potentiate and depress mossy fiber synaptic responses in rat hippocampal CA3 pyramidal

neurons. J. Neurophysiol. 91, 1596–1607.

Xu, T., Naraghi, M., Kang, H., and Neher, E. (1997). Kinetic studies of Ca2þ binding and Ca2þ

clearance in the cytosol of adrenal chromaYn cells. Biophys. J. 73, 532–545.

Yamada, W. M., and Zucker, R. S. (1992). Time course of transmitter release calculated from simula-

tions of a calcium diVusion model. Biophys. J. 61, 671–682.

Yang, S. N., Tang, Y. G., and Zucker, R. S. (1999). Selective induction of LTP and LTD by

postsynaptic [Ca2þ]i elevation. J. Neurophysiol. 81, 781–787.

66 Robert Zucker

Young, S. M., Jr., and Neher, E. (2009). Synaptotagmin has an essential function in synaptic vesicle

positioning for synchronous release in addition to its role as a calcium sensor. Neuron 63, 482–496.

Zoran, M. J., Doyle, R. T., and Haydon, P. G. (1991). Target contact regulates the calcium respon-

siveness of the secretory machinery during synaptogenesis. Neuron 6, 145–151.

Zucker, R. S. (1989). Models of calcium regulation in neurons. In ‘‘Neural Models of Plasticity:

Experimental and Theoretical Approaches,’’ (J. H. Byrne, and W. O. Berry, eds.), pp. 403–422.

Academic Press, Orlando, FL.

Zucker, R. S. (1992). EVects of photolabile calcium chelators on fluorescent calcium indicators. Cell

Calcium 13, 29–40.

Zucker, R. S. (1993). The calcium concentration clamp: Spikes and reversible pulses using the photo-

labile chelator DM-nitrophen. Cell Calcium 14, 87–100.

Zucker, R. S., and Haydon, P. G. (1988). Membrane potential has no direct role in evoking neurotrans-

mitter release. Nature 335, 360–362.

Zucker, R. S., and Stockbridge, N. (1983). Presynaptic calcium diVusion and the time courses of

transmitter release and synaptic facilitation at the squid giant synapse. J. Neurosci. 3, 1263–1269.

CHAPTER 3


Making and Using Calcium-SelectiveMini- and Microelectrodes

L. Hove-Madsen,* S. Baudet,† and D. M. Bers‡

*Cardiovascular Research Centre CSIC-ICCCHospital de la Santa Creu i Sant PauBarcelona, Spain

†Ricerca Biosciences SASSaint Germain sur l’Arbresle, France

‡Department of PharmacologyUniversity of California, Davis,Davis, California, USA

A

OGY,All rig

bstract


-679X-679X

I. In
troduction A. Main Characteristics of Ca2þ-Selective Electrodes
II. R
ationale III. M ethods
A.
Preparation of Minielectrodes B. Application of Minielectrodes C. Preparation of Ca2þ-Selective MEs D. Application of Ca2þ-Selective MEs
IV. D
iscussion R eferences
Abstract

Detection and measurement of intracellular calcium concentration ([Ca2þ]i)have relied on various methods, the popularity of which depends on their ease of

use and applicability to diVerent cell types. Historically, Ca2þ-selective electrodeshave been used concomitantly with absorption indicators such as arsenazo-III,

but their interest has been eclipsed by the introduction of a large number of

fluorescent calcium probes with calcium sensitivities varying from the nanomolar

/10 $35.00(10)99003-1

68 L. Hove-Madsen et al.

to the micromolar range such as fura-2, indo-1, fluo-4, and many others. In this

chapter, we emphasize the utility of Ca2þ-selective electrodes and show that their

use is complementary to use of fluorescent indicators; indeed, each method has

advantages and disadvantages. We first describe the preparation and application

of Ca2þ-selective minielectrodes based on the Ca2þ ligand ETH 129 (Schefer

et al., 1986) that have a larger dynamic range and faster response time than most

commercially available calcium electrodes. The second part of the chapter is

dedicated to ETH 129-based Ca2þ-selective microelectrodes (MEs), and their

application in the determination of [Ca2þ]i in cardiac cells. Since numerous

reviews and books have been dedicated to the theoretical aspects of ion-selective

ME principles and technology, this chapter is not intended for investigators who

have no experience with MEs.

I. Introduction

A. Main Characteristics of Ca2þ-Selective Electrodes

The key advantage of the Ca2þ-selective electrodes is the wide dynamic range

of their response (e.g., from pCa 9 to 1), as compared, for example, to

fluorescent and metallochromic Ca2þ indicators that typically have a dynamic

range of four or less pCa units (Fig. 1). There has been developed a plethora of

useful fluorescent calcium probes with calcium sensitivities varying from the

nanomolar to the micromolar range such as fura-2, indo-1, fura red, fluo-4,

furaptra, fluo-5N, and others (Grynkiewicz et al., 1985; Harkins et al., 1993;

Lipp et al., 1996; Picht et al., 2006; Shannon and Bers, 1997). These are widely

used and are extremely important tools for study of Ca2þ, but Ca2þ-selectiveelectrodes are a valuable complementary tool. For more basic reference to

electrode technology and electrophysiology, we suggest monographs by

Ammann (1986), Purves (1981), and Thomas (1982). An electronic introduc-

tion to ion-selective electrodes can be found at www.nico2000.net/Book/

Guide1.html.

Their response is based on a semiempirical equation (Nicolski–Eisenman equa-

tion) derived from the Nernst equation:

Ex ¼ Eo þ RT=ZxF lnðax þKpotxy azx=zyy Þ ð1Þ

where Ex is the ion-selective electrode potential, Eo is a constant, R, T, Z, and F

have their usual meaning, ax is the activity of the ion that is measured (activity

(a) is related to concentration (C) by the relation: a¼gC where g is the activity

coeYcient) and Kxypot is the selectivity coeYcient. This expression is strictly

valid for activities only, but if the activity coeYcients do not change, they can

be used with free concentrations too. This is often for convenience, since

solutions and chemical equilibria are more often described in these concentra-

tions terms. So, if x is Ca2þ and y is Naþ (the most common interfering cation

http://www.nico2000.net/Book/Guide1.html

http://www.nico2000.net/Book/Guide1.html

100

Indo-1

Minielectrode

Microelectrode

pCa

8

Pot

entia

l (m

V)

75

50

25

0

−25

−50

−757 6 5 4 3

Flu

ores

cenc

e ra

tio

3

2

1

0

Fig. 1 Dynamic range of Ca2þ-selective electrodes and indo-1. The electrode potential of Ca2þ-selective mini and MEs is shown together with the fluorescence ratio (400/470) for indo-1. Measure-

ments were performed in a KCl buVer containing 140 mM KCl, 10 mM HEPES, 10 mM NaCl, and

1 mM EGTA. Notice that indo-1 is suitable for measurements between pCa 7.5 and 5, while the

dynamic range for the Ca2þ electrode is wider, ranging from pCa 9 to 1 for minielectrodes and 7.5 to

1 for MEs.

3. Calcium Selective Mini- and Microelectrodes 69

for the Ca2þ-selective ligand), the relationship becomes, for 30 �C and changing

to log 10:

ECa ¼ Eo þ 30 log Ca2þ� �þKpot

NaCa Naþ½ �2� �

ð2Þ

Thus for the case of no interfering ion and when extracellular Ca2þ¼ [Ca2þ]ref,then the potential diVerence (DE) between two solutions of diVerent [Ca2þ] reducesto the Nernst equation:

DE ¼ 30 logð½Ca2þ�=½Ca2þ�refÞ ð3ÞThe response of an ion-selective electrode to changes in free [Ca2þ] is much slower

than the fluorescent, bioluminescent, and metallochromic Ca2þ indicators. Thus,

Ca2þ electrodes are ideal for measurements of slow changes over wide ranges of

[Ca2þ], but not appropriate for very rapid changes of free [Ca2þ] although the Ca2þ

electrodes can respond in the millisecond range at higher [Ca2þ] (Bers, 1983). Theresponse time can also be improved by using Ca2þ selective electrodes with a

concentric inner micropipette that reduces the longitudinal resistance of the

Ca2þ-selective resin and thereby decreases the electrical time constant (Fedirko

et al., 2006; Ujec et al., 1979). These limitations, together with the physical size of

Ca2þ-selective electrodes put them at disadvantage with fluorescent and lumines-

cent calcium indicators for measurements of dynamic changes in calcium levels in


cellular environments. Moreover, about the time that indo-1 and fura-2 were

synthesized (Grynkiewicz et al., 1985), a Ca2þ-ligand of improved selectivity

(ETH 129) was introduced (Schefer et al., 1986), which made it more realistic to

measure low Ca2þ levels typical of intracellular environments, since the electrode

response was Nernstian until pCa 8–9 (depending on the ionic background).

However, the availability and popularity of the fluorescent indicators did to a

large extent eclipse and limit further use and characterization of ETH 129-based

electrodes.

II. Rationale

We will not present an extensive review of the advantages and disadvantages of

Ca2þ-selective electrodes compared to fluorescent indicators, but the purpose of

this chapter is rather to show how the advantages of both approaches can be

combined for specific purposes. For example, a classical problem with Ca2þ

indicators is to convert the fluorescent signal to actual [Ca2þ]. In fact, it is

commonly acknowledged that in vitro calibration curves of fluorescent calcium

indicators are not applicable to in vivo or in situ conditions, because the indicators

bind to intracellular constituents, which modifies their excitation/emission spec-

trum and decreases the apparent aYnity of Ca2þ for the probe (Blatter and Wier,

1992; Harkins et al., 1993; Hove-Madsen and Bers, 1992; Konishi et al., 1988).

In vivo calibration curves are even more diYcult to obtain and are highly depen-

dent on the cell type under study. Ca2þ-selective electrodes are still one of the most

straightforward ways to measure and quantify Ca2þ because (1) their behavior is

not appreciably altered by the intracellular milieu (ions and proteins); (2) they can

be easily included in an electrophysiological setup and do not require extensive

and expensive apparatus; (3) their linear (Nernstian) response simplifies the con-

version of the voltage signal to [Ca2þ]i; (4) their behavior allows determination

of wide ranges of pCa (see above); and (5) the Ca2þ ligand itself does not change

(or buVer) [Ca2þ]i.Thus, Ca2þ-selective electrodes continue to be a good choice for the preparation

of calibration solutions for Ca2þ determinations and for measurements of dissoci-

ation constants for Ca2þ-binding compounds such as ethylene glycol bis(b-amino

ethyl ether)-N,N,N0,N0-tetraacetic acid (EGTA), indo-1, 1,2-bis(o-aminophenoxy)

ethane-N,N,N0,N0-tetraacetic acid (BAPTA), and oxalate under diVerent experi-mental conditions (Bers, 1982; Harrison and Bers, 1987, 1989; Hove-Madsen and

Bers, 1992, 1993a; Hove-Madsen et al., 1998) Indeed, commercially available Ca2þ

electrodes are largely directed towards determination of the calcium concentration

in solutions or biological fluids, but Ca2þ electrodes can also be used to measure

cellular Ca2þ buVering and changes in the free [Ca2þ] in cell suspensions (Hove-

Madsen and Bers, 1993a, 1993b; Hove-Madsen et al., 1998) or combined with

other electrophysiological techniques (Kang and Hilgemann, 2004; Kang et al.,

2003). Moreover, Ca2þ electrodes are economical, easy to prepare, and they can


easily be used for measurements of free [Ca2þ] in experimental solutions or

biological fluids. Indeed, for those studying Ca2þ-dependent processes, there is

no practical or economical reason why Ca2þ electrodes should not be used as

routinely as pH electrodes.

III. Methods

A. Preparation of Minielectrodes

Ca2þ-selective minielectrodes can be prepared by dipping polyethylene (PE)

tubes (typically �5 cm) in a membrane solution (see composition below). We

have tried other types of tubing but polyvinyl chloride (PVC) tubing appears to

absorb ETH 129 from the Ca2þ-selective membrane, resulting in a faster loss of

sensitivity as compared to the PE tubing. On the other hand, materials such as

Teflon tubing absorb little ETH 129 but the PVC membrane does not adhere well

to the tubing. As a result, the electrodes are more easily damaged, although they

may have a longer lifetime if handled with care.

The dimensions of the tubing vary from a diameter less than 1 to �3 mm.

Electrodes prepared with the membrane solution described below result in Ca2þ-selective membranes that are a few hundred micrometers thick. With diameters

larger than 5 mm, the Ca2þ-selective membrane bursts more easily during

handling, but this problem may be overcome by inserting a ceramic plug into the

tubing, before dipping it in the membrane solution as described by Orchard et al.

(1991). For general purposes we have used inner electrode diameters of 1.67 mm

(PE 240, Clay Adams).

After dipping the PE tubing in the membrane solution, the Ca2þ-selectivemembrane is allowed to dry overnight. Then the electrode is filled with an appro-

priate filling solution, which should correspond to experimental conditions (see

below). After filling the electrode, it is allowed to equilibrate for at least 3 days in a

glass vial containing the filling solution (but see below).

1. Preparation and Use of the Ca2þ-Selective Ligand

The Ca2þ-selective membrane can be prepared as described by Schefer et al.

(1986) (all from Fluka/Sigma-Aldrich; Table I).

ETH 129 is dissolved inN-phenyl-octyl-ether (NPOE) under vigorous stirring in

a small glass vial (Solution 1). At the same time, PVC is dissolved in the THF;

when completely dissolved, potassium TCPB is added (Solution 2). When the

components of Solutions 1 and 2 are completely dissolved, the two solutions are

mixed, and the membrane solution is ready to use, or can be stored in a glass vial,

closed with a Teflon screwcap and protected from light. If THF evaporates from

the membrane solution during storage, a small amount of THF can be added to

achieve the desired viscosity of the membrane solution.

Table IPreparation of the PVC-based Ca2þ-selective ligandminielectrodes

Component* Amount

Solution 1

ETH 129 25 mg

N-phenyl-octyl-ether (NPOE) 451.5 ml

Solution 2

Polyvinyl chloride (PVC) 250 mg

Kþ-tetrakis chlorophenyl borate (TCPB) 12.9 mg

Tetrahydrofuran (THF) �5 ml

*Components can be obtained from Sigma-Aldrich (St. Louis, MO)


The filling solution used for the minielectrode depends on the experimental

solutions. Generally, the ionic composition should mimic the environment in

which measurements of Ca2þ are planned, and the [Ca2þ] of the filling solution

should also be in the range of the measured values. With measurements of low

[Ca2þ], Ca2þ in the filling solution can be buVered with EGTA to the desired free

[Ca2þ]. However, we have not obtained good results with filling solutions with a

pCa higher than 7.5. We typically use a Ca2þ-EGTA buVer with 1 mM free Ca2þ as

a filling solution.

2. Electrode Characteristics

The resistance of the minielectrodes is 1–2 MO, which normally makes it possi-

ble to use a standard pH/ion meter to monitor the electrode potential. We have

used either a commercial pH/ion meter (Orion pH/ion analyzer) or preamplifier

(A311J, Analog Devices). When using commercial pH/ion meters, it is normally

necessary to make an adapter cable that connects the minielectrode to the meter

input. We have used a chart recorder (Soltec) or an electronic data acquisition

device (Linseis) for continuous monitoring of the electrode potential.

The lifetime of the minielectrode depends on the [Ca2þ] the electrode is used to

measure, and on the composition of the experimental solutions. For measurements

in solutions without protein or interfering ions, the detection limit for the electro-

des increases slowly with time and the electrodes will have a detection limit in the

subnanomolar range and a Nernstian response down to 10 nM for at least 1 month

after preparation. Figure 2 shows the change in the response of a Ca2þ electrode

with time.

Two days after filling, the electrode response was still ‘‘super-Nernstian’’ at low

[Ca2þ] (i.e., >29 mV per 10-fold change in [Ca2þ]), but normal within 7 days after

filling.Notice thatwedid still obtainaNernstian responsedowntopCa8 for2months

after filling the electrode. The response time at low [Ca2þ], however, got slower with

Pot

entia

l (m

V)

2 Months

2 Days

50

25

0

−25

−50

−75

−100

pCa

8 7 6 5 4

Fig. 2 Electrode potential of a Ca2þ-selective minielectrode 2, 7 and 60 days after filling of the

minielectrode.Measurements were performed in a KCl buVer as in Fig. 1. Notice the ‘‘super-Nernstian’’

response of electrodes 2 days after filling (circles). Seven days after the filling (squares), electrode

response was linear down to a free [Ca2þ] of less than 10 nM. The slope of the regression line was

�28.4 mV/pCa. Two months after filling (diamonds), the electrode response was linear to pCa 7.5, but

response time was slowed at high pCa.


time and measurements below 30 nM are only practical with fairly fresh electrodes.

We generally fill electrodes once a week to obtain the best results. However, if the

Ca2þ electrodes are used to measure micromolar or higher [Ca2þ] in protein-free

solutions, the same electrode can be used for longer periods (up to several months).

In the presence of cellular proteins, a small oVset in the electrode response is seen

at the first exposure to protein. Then, no further alteration of the electrode

response occurs, but the response time of the Ca2þ electrode is increased after

exposure to protein, and measurements of free [Ca2þ] below 10 nM are more

diYcult in the presence of protein concentrations higher than 10 mg/ml.

The response time of the Ca2þ electrodes can be of critical importance in some

applications. Figure 3 compares the response times of a Ca2þ electrode and indo-1

fluorescence to a decrease in the [Ca2þ] in a suspension of permeabilized myocytes

(3 mg/ml) where cellular Ca2þ uptake processes have been blocked with thapsigar-

gin and ruthenium red.

We examined the response to a decrease in [Ca2þ], as this may be a more

stringent test than an increase in [Ca2þ]. Notice that when 2 mM EGTA was

added to lower the free [Ca2þ], the electrode response was 94% complete in 1 s

while the indo-1 signal was 97% complete in 1 s. A slow final phase, lasting several

seconds, is apparent in the electrode signal only (see amplified inset).

In Fig. 3 the response time was examined under experimental conditions where

spatial inhomogeneities in the myocyte suspension are minimized by buVeringCa2þ with indo-1 and oxalate. However, under some experimental conditions, an

50Indo

Elec

ElecEGTA

40

30

20

Free

[Ca]

(mM

)

Free

[Ca]

(mM

)

10

−10 10 15−5 0

Time (s)

5

1.5

1

0.5

0

0 10Time (s)

0

Fig. 3 Comparison of response time of a Ca2þ-selective minielectrode and indo-1. Free [Ca2þ] wasmeasured in a suspension of permeabilized rabbit ventricular myocytes. Ca2þ uptake into sarcoplasmic

reticulum and mitochondria was inhibited with thapsigargin and ruthenium red, respectively. The initial

free [Ca2þ] was 32 mMwhich is near saturation for indo-1, resulting in a very noisy trace. That is because

a small change in fluorescence ratio corresponds to a large in [Ca2þ] at this level (see Fig. 1). At the

arrow, 2 mMEGTAwas added to the cell suspension to lower free [Ca2þ]. Both the electrode and indo-1signal were more than 90% complete in 1 s. The inset shows the response of indo-1 and Ca2þ electrode at

low [Ca2þ]. Notice that the indo-1 signal was 100% complete in 2 s (and actually undershoot slightly)

while the final completion of the electrode response was slower.


apparently slower electrode response may result from inhomogeneities rather than

a slower electrode response per se. This is illustrated in Fig. 4, where free [Ca2þ]was monitored with Ca2þ electrode and indo-1 simultaneously in a myocyte

suspension in the absence and presence of oxalate.

In Fig. 4A, Ca2þ addition to the cells causes a rapid increase in [Ca2þ], which is

subsequently sequestered by the sarcoplasmic reticulum (SR). In the absence of

10 mM oxalate (Fig. 4A), the electrode response appears to be slower than the

corresponding indo-1 signal. However, when oxalate is subsequently added to the

cell suspension (Fig. 4B), the measured change in free [Ca2þ] after a Ca2þ addition

is similar for indo-1 and the Ca2þ electrode. It should be noted that oxalate not

only buVers the free [Ca2þ], but also increases the Ca2þ uptake rate in the SR, and

thereby the removal of Ca2þ from the cell suspension. Thus, despite inducing a

faster rate of change in free [Ca2þ], oxalate eliminates the diVerence between Ca2þ

electrode and indo-1 signal by eliminating spatial inhomogeneities in free [Ca2þ] inthe myocyte suspension. Indeed, indo-1 is expected to be less sensitive to spatial

inhomogeneities as it diVuses into the permeabilized cells and binds to cellular

proteins (Hove-Madsen and Bers, 1992, 1993b). In contrast, the Ca2þ electrode

can only measure the Ca2þ outside the permeabilized cells, and inhomogeneities

during uptake or release of Ca2þ from the cells are therefore likely to occur,

resulting in erroneous measurements with the Ca2þ electrode.

2.5

0

2

1.5

1

0.5

0

15

10

5

0Fr

ee [C

a] (mM

)

Free

[Ca]

(mM

)

250 500

Time (s) Time (s)

750

Ca electrode

Ca electrodeIndo-1 Indo-1

1000 0 50 100 150

Fig. 4 EVect of inhomogeneities in [Ca2þ] on a Ca2þ-selective minielectrode and indo-1 response in

permeabilized rabbit ventricular myocytes. Panel A shows simultaneous measurements of Ca2þ uptake

in digitonin-permeabilized myocytes with both a Ca2þ-selective minielectrode and indo-1. Ca2þ uptake

in mitochondria was inhibited with ruthenium red. Ca2þ was added at time¼0 and was largely

accumulated by the SR. Under these conditions, the response of the Ca2þ electrode was slower than

indo-1. Panel B shows Ca2þ uptake by precipitating intra-SR Ca2þ and thereby preventing buildup of a

[Ca2þ] gradient. Notice that Ca2þ uptake is much faster in the presence of oxalate with no apparent

diVerence between electrode and indo-1 response (from Hove-Madsen and Bers (1993a) with

permission).


3. Storage of Minielectrodes

After PE tubes have been dipped in an ETH 129 membrane solution and allowed

to dry overnight, the dry electrodes can be stored in a closed glass vial for long

periods. We have filled minielectrodes that had been stored for 3 years and the

electrodes made with PE tubing still had a resistance of 1–2 MO with a linear

response down to less than 10 nM Ca2þ after filling. Electrodes made with PVC

tubing had higher resistance (�50 MO) but were also functional, although slower

and less sensitive. Storage of the electrodes in plastic vials results in ‘‘Ca2þ-selectiveplastic containers,’’ as the ETH 129 slowly diVuses into the container. Once the

Ca2þ electrodes are filled with the filling solution, however, the response time

increases and the electrodes gradually lose sensitivity.

B. Application of Minielectrodes

Minielectrodes can be used for a number of purposes. The most straightforward

application is the preparation of solutions where Ca2þ is buVered with chelators

such as EGTA, EDTA, or BAPTA as described by Bers (1982). We have developed

a spreadsheet that allows calculation of the actual pCa of these solutions, based on

the Nernstian response of the minielectrodes (see Chapter 1). Furthermore, the

spreadsheet allows determinations of the Kd and the purity of the Ca2þ chelator

used to prepare the solution. Thus we have used the minielectrodes to determine


the Kd for EGTA, BAPTA, and oxalate in buVer solutions (Bers, 1982; Harrison

and Bers, 1987, 1989; Hove-Madsen and Bers, 1993a). More comprehensive pro-

grams to calculate the amount of Ca2þ and Ca2þ buVer needed to prepare solu-

tions have been developed (e.g., MaxChelator by C. Patton, Marine Biology

Institute, Monterey, CA; http://maxchelator.stanford.edu/downloads.htm; cf.

Chapter 1).

We have also used the minielectrodes to characterize the binding of Ca2þ to

indo-1 in vitro and in cell suspensions, in order to calibrate the indo-1 and furaptra

signals when used in cell suspensions (Hove-Madsen and Bers, 1992; Hove-

Madsen et al., 1998; Shannon and Bers, 1997). In agreement with previous studies

of fura-2 (Konishi et al., 1988) we found that indo-1 binds extensively to cellular

proteins and causes a � fourfold increase in the Kd for Ca2þ-indo-1 in permeabi-

lized myocytes.

The minielectrodes can also be used to titrate the passive Ca2þ binding sites in

permeabilized myocytes where the cellular Ca2þ uptake and release process are

inhibited. Using the same titration method, we have measured total Ca2þ uptake in

the SR in permeabilized myocytes, by inhibiting Ca2þ uptake in the mitochondria

and release of Ca2þ through the SR Ca2þ release channels (Hove-Madsen and

Bers, 1993a).

Finally, we have used the minielectrodes together with indo-1 for online mea-

surements of the Ca2þ uptake rate in the SR in permeabilized ventricular myocytes

(Hove-Madsen and Bers, 1993b) and to examine the eVects of phospholamban

phosphorylation and temperature on the uptake rate (Hove-Madsen et al., 1998;

Mattiazzi et al., 1994), and we have determined the inhibition of Ca2þ uptake by

thapsigargin in order to measure the number of SR Ca2þ pump sites (Hove-

Madsen and Bers, 1993b). In experiments measuring Ca2þ uptake rates caution

should, however, be taken when using minielectrode because of the above men-

tioned possibilities of inhomogeneities in Ca2þ in cell suspensions and slowing of

the electrode response.

C. Preparation of Ca2þ-Selective MEs

1. Glass Tubing Preparation

We have used nonfilamented capillaries (150 mm outer diameter, 15 cm long,

from Clark Electromedical Instruments, UK or World Precision Instruments,

USA). The glass is cut in the middle with a diamond pen or glass scorer (with

care to avoid deposition of dirt). Both ends of the capillaries (now 7.5 cm long) are

lightly fire-polished; a whole batch can be prepared and kept in a small glass

beaker, preferentially in a dust-proof container. Cleaning of the micropipettes

prior to pulling has been a matter of debate, but as 99.8% of the glass after pulling

is newly exposed (Deyhimi and Coles, 1982), we do not find such procedures to be

necessary.

http://maxchelator.stanford.edu/downloadshtm; cf.


2. Microelectrode Pulling and Silanization

MEs can be prepared ‘‘on-demand’’ and in ‘‘batch’’ methods. The first method

consists in preparing (i.e., pulling and silanizing) MEs on the experimental day,

which has the obvious advantage of having ‘‘freshly’’ prepared electrodes and of

being able to manufacture as many as desired a day. Protocols have been described

(reviewed in Ammann, 1986). The batch method that we use consists of preparing

a batch of MEs that can be kept for several days in a dry, dust-proof container.

MEs are pulled on a programmable horizontal stage puller (Model P80 PC or

comparable, Sutter Instruments, USA). By trial-and-error, and according to the type

of biological preparations studied, a satisfactory shape ofMEcanbe found.However,

several aspects ofMEpulling have to be taken into account when designing its shape.

As ion-selective MEs have an intrinsically high resistance (i.e., >50–100 GO), thesignal-to-noise ratio has to be minimized. This can be achieved in two ways: the first,

and most obvious, is to increase the tip diameter, within certain limits, which are

dictatedby the sizeof the cell type. Inour case,we impale cardiac cells inwholemuscle.

Cell length is typically between 50 and 150 mm and tip diameters less than 0.5 mm are

needed. However, there is a trade-oV between tip diameter and detection limit of the

electrode (see below). That is, increasing tip diameter improves electrode response (in

terms of detection limit and speed of response), but is more likely to damage the cell

during impalement. Another simple way to decrease the resistance is to decrease the

length of the ME shank. This further helps to reduce capacitative artifacts, that are

encountered when the level of physiological solution fluctuates in the experimental

bath (Vaughan-Jones and Kaila, 1986), and also helps electrolyte filling (see below).

In contracting muscular preparations, the shank should also possess some flexibility

to avoid dislodging of the ME during a contraction. Again, by trial-and-error, an

adequate shape that fulfills all these requirements can be found. Their shape was

designed to reduce their resistance by making them steeply tapered (having a shank

length of approximately 150 mm and diameter of 20 mm at 10 mm above the tip).

Under lightmicroscopical observation, the tip diameterwas estimated tobe�0.5 mm.

MEs are dehydrated, tip up in an aluminium block, at 200 �C for 12 h. Our

experience, and of others (Vaughan-Jones and Wu, 1990), has been that a better

silanization and longer lifetime of the MEs are achieved this way.

The silanization protocol consists of spritzing 300 ml of N,N-dimethyltrimethyl-

silylamine (Fluka) onto the aluminum block and rapidly placing a glass lid on top

of the dish. Care must be taken not to inhale vapor from the silane vial or during its

introduction in the dish. In our hands, once opened, the silane vial can be kept for

at least 2 months without losing its properties. Silanization procedure lasts for

90 min. The lid is then removed and the MEs are baked for another 60 min, which

drives oV the excess silane vapor. The aluminum block and the MEs are then

placed into an air-tight plastic container, also containing desiccant. We advise not

to keep the MEs more than a week, because repetitive openings of the container

and insuYcient seal quality will progressively make the electrodes lose their

hydrophobicity.


3. Electrolyte Filling of the ME

The filling or conducting electrolyte is introduced in the ME from the back

(back-filling), but because of the hydrophobicity of the glass wall (due to the

silanization), the tip does not get filled immediately. The strategy to fill the ME

completely depends on the absence or presence of the ligand at its very tip.

In the case where the ME is first filled with the electrolyte, the back of the ME is

connected, via a flexible tubing, to a 50-ml syringe, whose plunger has been

previously pulled to 5 ml. Positive pressure is then applied and, as assessed by

microscopic observation, the electrolyte creeps along the ME wall and fills the tip.

The biggest air bubbles are then removed by gently flicking the ME, held tip down.

Although we have initially applied this procedure to filamented MEs, a similar

success rate (more than 90%) has been obtained with nonfilamented MEs

(surprising, because silanization is expected to limit aqueous filling ease).

Whatever the method used, if the electrolyte column was interrupted by air

bubbles then, gentle heating of the tip of theME, under microscopic control, with a

tungsten–platinum wire, according to the device described by Thomas (1982), can

remedy this problem. Once filled with electrolyte to the tip, the hydrophobic ligand

can be introduced (see below).

In the case where the electrolyte is added after the Ca2þ ligand, additional

problems exist. In fact, caution has to be taken to avoid the presence of air at the

electrolyte–ligand interface. If a traditional whisker is used, great care has to be

taken not to accidentally disrupt the column of ligand, which can lead to mixing of

oil and water, making unstable ion-selectiveMEs. If a heating filament is used, care

has to be taken not to heat the ligand because it is likely that the local high

temperature may damage the ligand properties.

We have used a filling solution that has an ionic composition mimicking the

intracellular medium (in mM): Naþ: 10; Kþ: 140; HEPES: 10; EGTA: 1; pH 7.1

(at 30 �C) and pCa 7. This solution is in fact identical to the calibrating solution

having the same pCa (see below and also Orchard et al. (1991)) for additional

comments).

Our experience has been that it is best to minimize the time between electrolyte

and ligand filling. We prefer to fill the ME with the ligand as soon as the ion-

selective ME is filled with the electrolyte (although we managed to draw the ligand

in the tip 2–3 h after electrolyte filling). If MEs are left overnight with the filling

solution, it can be very diYcult to draw ligand into the tip; this observation could

be explained by a progressive glass hydration, causing it to lose its capability to

retain the ligand.

4. Preparation and Use of the Ca2þ-Selective Ligand

For repetitive and long-lasting Ca2þ measurements with Ca2þ-selective MEs

dissolving the Ca2þ ligand in a ‘‘cocktail’’ containing PVC is useful (Tsien and

Rink, 1981). Although the cocktail available from Fluka (Cocktail II containing

Table IIPreparation of the PVC-based Ca2þ-selective ligand MEs

Componenta Amount

Solution 1

ETH 129 (or ETH 1001) 27.5 mg

NPOE 500 mlNaþ-tetraphenyl borate 5.53 mg

Solution 2

Solution 1 200 mlPVC 36 mg

THF 400 ml

aComponents can be obtained from Sigma-Aldrich (St. Louis,

MO) and see Table I for abbreviations.


94% (w/w) NPOE, 5% (w/w) ETH 129, and 1% (w/w) sodium TPB) is satisfactory

to start with (Ammann et al., 1987), small volumes sometimes provided (0.1 ml) do

not facilitate handling. We prefer to make up our own cocktail, in a larger NPOE

volume, with the same proportions. The composition, for 500 ml of NPOE is given

in Table II.

Because of the small volumes and required stirring, it is preferable to work with

flat-bottomed, small volume glass vials, and miniature stirring bars. ETH 129 and

TPB are dissolved with vigorous stirring in NPOE, in a 2-ml glass vial (Solution 1).

Solution 1 can be kept at room temperature for several months, with a Teflon

screw cap, protected from light. When the final cocktail (Solution 2) is prepared,

PVC is dissolved in THF with stirring. 0.2 ml of Solution 1 is then added, stirred,

and finally sonicated. THF is allowed to partially evaporate to approximately half

of the initial volume and the final cocktail is finally poured in a 0.5-ml conical vial

(Clark Electromedical Instruments).

Before dipping the ME tip, THF is allowed to evaporate until the mixture has

the consistency of a thick syrup. Experience helps to determine the adequate

consistency of the ligand. In fact, if not enough THF has evaporated, the ETH

129 sensor is too diluted and the evaporation in theMEmay cause retraction of the

gel, yielding poor responses. On the other hand, in some instances, we have

managed to fill the electrodes even if the ligand appeared to be solidified (as a

rule, the thickest mixture which will fill the tip is best).

Because of the small diameter of the tip and the viscosity of the ligand, negative

pressure is required at the back of the electrode. This is achieved by a >10-ml

syringe connected, via a 3-way stopcock, to a flexible Teflon tubing connected to

the back of the electrode with soft tubing. Vacuum is then applied by pulling the

plunger out and by blocking it with a rod/block or collar placed along the plunger

(care should be taken to regularly check the vacuum of the system). Observation of

the ME and measurements of the ligand column height are performed under


microscopic control, with a microforge (e.g., as described by Thomas (1982). In

brief, the microscope body is laid on its back so that the stage (removed) would be

vertical and the eyepieces are oriented upwards. A long-working distance objective

(40�) is used and the ME and the ligand vial are held independently by two

micromanipulators.

Depending on glass, shape, and tip diameter 10–30 min of negative pressure is

typically required to fill the tip. Vacuum is slowly released before lifting the

electrode from the ligand. A column height of less than 300 mm is preferable,

because it decreases the electrode sensitivity to changes in temperature and level

of the bath in the experimental chamber (Vaughan-Jones and Kaila, 1986). Our

experience is that, depending on the tip diameter, column heights between 50 and

250 mm yielded acceptable electrode responses.

Once the electrode is filled with both the ligand and the electrolyte, we prefer not

to let the ME equilibrate in Tyrode or high pCa solutions, because this favors the

deposition of dirt on the tip of the ME and might contribute to clogging. Rather,

we place them, tip up, in a drilled plastic plate, protected by an upside down glass

beaker. Before calibration, the column height is rechecked because THF in the

column continues to evaporate, leading to shrinkage of the PVC gel.

As THF evaporates, the stock solution becomes even thicker. Periodically

enough THF must be added to decrease the viscosity of the mixture. This process

is hastened by mixing it with a glass rod and then on a Vortex mixer (maximal

setting). Sonication can also be used, but does not give better results.

5. Double-Barreled Ca2þ-Selective MEs

For Ca2þ-selective MEs measurements, one must measure both the potential of

the Ca2þ electrode and the local voltage (typically with a KCl-filled ME; see

below). Both electrodes can be built into a single double-barreled electrode,

where one barrel is the Ca2þ electrode and the other is the voltage electrode. We

have not had much luck using these for intracellular recording, but they can be

extremely useful for measurement of local extracellular or interstitial [Ca2þ] inmulticellular preparations (Bers, 1983, 1985, 1987; Bers and MacLeod, 1986;

Shattock and Bers, 1989).

Double-barreled electrodes can be pulled from 2 to 2.5 mm diameter theta-style

tubing (R and D Optical Systems, MD) on a Brown–Flaming P-77 micropipette

puller (Sutter Instruments, CA, USA). For extracellular recording, the tips of the

electrodes are carefully broken under microscopic observation to have 4–12 mmoverall tip diameters. Two methods of silanization of the Ca2þ barrel are practical.

(1) Distilled water is injected into the reference barrel. A hypodermic needle

containing tri-n-butylchlorosilane is introduced into the Ca2þ barrel, �1 ml ofsilane is displaced into the shank of each electrode, and electrodes are placed

in a 200 �C oven, tips up, for 5 min and then cooled. (2) a stream of silanizing N,

N-dimethyltrimethylsilylamine (TMSDMA vapor is passed through the Ca2þ

barrel (with or without warming), while a stream of nitrogen gas is passed through


the reference barrel (under pressure) to prevent silanization of the reference barrel

(which would result in both barrels being Ca2þ-sensitive).The larger tips of these electrodes make the filling easy. Both barrels can be

easily backfilled. The silanized barrel is backfilled with a reference solution con-

taining 10 mMCaCl2 and 100 mMKCl and the nonsilanized barrel with a solution

containing 140 mM NaCl. A column of the neutral Ca2þ ion-exchange cocktail

ETH 1001 or 129 (Fluka Chemical, Ronkonkoma, NY) 50–250 mm long is easily

drawn into the silanized barrel. Ca2þ electrodes with these tip diameters exhibit

Nernstian behavior over the range 10 mM–10 mM Ca2þ (Bers and Ellis, 1982).

The resistance of these MEs was typically 1–5 GO for the Ca2þ-sensitive barrel and1–4 MO for the reference barrel.

The impedance of the two barrels is very diVerent, but their fast response allowsrelatively rapid interstitial [Ca2þ] monitoring. To match signal response kinetics to

voltage steps a variable-passive R-C filter can be added to the reference barrel

signal after the signal has come from operational and oVset amplifiers. This filter is

adjusted while a square voltage pulse is fed into the bath until the best matching

with the Ca2þ barrel response is obtained (Bers, 1983). These Ca2þ electrodes

typically exhibit Nernstian behavior at least over the range 10 mM–10 mM Ca2þ.This is satisfactory for typical extracellular [Ca2þ] measurements and the double-

barreled electrodes are easier to calibrate and use than intracellular impalements

(described in more detail below).

6. Calibrating Bath and Solution Perfusion

It is preferable to calibrate ion-selective MEs in the experimental chamber in

which measurements are made or to have the calibrating bath as close as possible in

design and proximity. Our calibration chamber is a ‘‘flow-through’’ type (volume:

0.1 ml), immediately adjacent to the experimental chamber. Note that it is conve-

nient that the experimental chamber is viewed from the front, and not from above.

7. Calibration Procedure

The bath electrode is either an Ag wire (chlorided by dipping it in bleach for

15–20 min) or an agar bridge. Ideally, a conventional electrode (3 M KCl filled)

should also be immersed in the bath, and the diVerential voltage (ion-selective ME

minus conventional) should be read. We use a commercial amplifier (FD-223 from

WPI) or a home built amplifier using varactor bridge preamplifiers (AD311J,

Analog Devices) as described by Thomas (1982).

We have adopted the following method to quickly select suitable MEs. The ME

is mounted in its holder and advanced into the calibrating bath, allowing the trace

to stabilize. If the device used to measure the signal has a resistance measurement

feature, it is worth measuring this parameter. In fact, our experience has been that

for the sharp Ca2þ MEs having resistances ranging between 100 and 250 GO were

suitable for our experiments, in terms of linearity and detection limit of the


calibration curve. Within this range, the higher the resistance, the lower the

detection limit for a given batch of MEs. At resistances higher than 300 GO, thedetection limit decreased sharply, probably because of the small tip diameter

(Ammann, 1986). Finally, although low resistance MEs tended to also have low

detection limits, they were not suitable for our experiments, because of the large tip

diameter that could seriously damage the cell membrane during impalement. By

contrast, we have found the height of the ligand column not to be a valuable

predictor of ME performance (although this height was always kept 50–250 mm).

In our hands, an ME can be used a few minutes after ligand filling.

After equilibration in the control physiological solution (in our case, an HEPES-

based Tyrode, containing 2 mM Ca2þ), ion-selective ME potential is adjusted to

0 mV. Our 2 mM Ca2þ Tyrode gives a voltage reading corresponding to an

intracellular calibrating solution of pCa 2.6. Then, flow is switched to a solution

of high pCa (between 7.5 and 9). At 30 �C (our experimental temperature), the

theoretical slope of the relationship between voltage and pCa is �30 mV/pCa, so

that, between pCa 2.6 and pCa 8, the theoretical voltage should be �162 mV.

However, as the electrode detection limit decreases at high pCa, a practical

compromise is often necessary for acceptability (e.g., readings more negative

than �150 mV). A Ca2þ-selective ME meeting this criteria may then be calibrated

over a wider range of [Ca2þ]. After the calibration is completed, the ME is moved

into the experimental chamber and equilibrated until stable. Conventional,

3 M KCl-filled MEs are pulled from the same glass and with the same character-

istics as the Ca2þ electrodes, but are not silanized.

D. Application of Ca2þ-Selective MEs

1. Sharp Ca2þMEs for [Ca2þ]i Measurement

We and others have had some, but limited success in making reliable [Ca2þ]imeasurements in cardiac myocytes with these electrodes (Bers and Ellis, 1982;

Marban et al., 1980), and a few other groups have had some luck with other cell

types such as skeletal muscle fibers (Allen et al., 1992; Blatter and Blinks, 1991;

Lopez et al., 2000), Aplysia neurones (Gorman et al., 1984), and photoreceptors

(Levy and Fein, 1985) using an earlier developed Ca2þ ionophore (ETH 1001;

Sigma-Aldrich 21192) instead of ETH 129 in the cocktail. ETH 1001-based MEs

generally do not make electrodes that have quite as low a detection limit as ETH

129 (Schefer et al., 1986), but for some reason ETH1001 seems to be of greater

practical utility for intracellular Ca2þ MEs. We have done some preliminary

cardiac muscle experiments that are consistent with this notion (resting

pCa¼6.34�0.15; mean�SD; n¼10 determinations). However, with the excellent

fluorescent Ca2þ indicators available now that are easy to use, one would need a

compelling reason to tackle this challenging electrophysiological approach for

[Ca2þ]i and the references above should then help.

Control

200 ms 100 ms

Rabbit ventricle Rat ventricle

[Ca]

o mM

Control

Control550

Control

+Citrate

+Citrate

+Citrate

+Citrate

500480

5202.5 m

N/m

m2

5 mN

/mm

2

500

490

A B

Fig. 5 Measurements of [Ca2þ]o with double-barreled Ca2þ-selective MEs during single steady state

contractions in (A) rabbit and (B) rat ventricular muscle (0.5 Hz, 30 �C). The [Ca2þ]o and tension are

shown in the absence and presence of 10 mM citrate (which limits [Ca2þ]o depletion by buVering

[Ca2þ]o. Bath [Ca2þ]o¼0.5 mM (dotted line). Data was from Shattock & Bers, (1989), as presented in

Bers (2001) (with permission).


2. Measuring Extracellular [Ca2þ] with Double-Barreled MEs

Double-barreled Ca2þ MEs can record rapid changes in extracellular [Ca2þ]([Ca2þ]o) between cells in multicellular preparations such as isolated cardiac tra-

beculae (Bers, 1983, 1985, 1987; Bers and MacLeod, 1986; Shattock and Bers,

1989). Figure 5A shows that one can detect small [Ca2þ]o depletions during

individual steady state rabbit cardiac action potentials and contractions. More-

over, when [Ca2þ]o is buVered by the low aYnity fast buVer citrate these depletionscan be suppressed. Note that these [Ca2þ]o depletions reflect net cellular Ca2þ

influx (in excess of eZux) early in the contraction and net Ca2þ eZux later in the

contraction, such that [Ca2þ]o returns to the bath level. In cardiac myocytes the

depletion is driven mainly by Ca2þ influx via Ca2þ channel current and to some

extent by Naþ/Ca2þ exchange (which can mediate Ca2þ influx at positive Em when

[Ca2þ]i is low). As [Ca2þ]i rises in the cell during the heartbeat because of Ca2þ

entry and SR Ca2þ release, it causes enhanced Ca2þ eZux (mainly via Naþ/Ca2þ

exchange in cardiac myocytes), and this allows [Ca2þ]o to recover. Note that action

potential repolarization greatly enhances the driving force for Ca2þ eZux via Naþ/Ca2þ exchange, further enhancing the recovery of [Ca2þ]o to the bath level.

In rat ventricular muscle the [Ca2þ]o signals are remarkably diVerent (Fig. 5B).In the rat there is only a very brief phase of [Ca2þ]o depletion (for �20 ms), which

gives way to a large rise in [Ca2þ]o during the contraction. At first this result

seemed surprising in light of the rabbit results in Fig. 5A. However, when we

consider the diVerences in action potential shape and that [Naþ]i is higher in rat

ventricular myocytes (Shattock and Bers, 1989), the explanation became clear. The

rat (and mouse) ventricle exhibit very short action potential duration compared to

30

60

−30

0

30

−90

−60

0

30

0

−60

−30

A

Em

or

EN

a/C

a (m

V)

Na/

Ca

exch

ange

driv

ing

forc

e(E

Na/

Ca−

Em

) m

V


rabbit (or human ventricle), and this drives rapid Ca2þ extrusion via Naþ/Ca2þ

exchange at a time when [Ca2þ]i is very high (Fig. 6B). In the rabbit ventricle, the

longer action potential plateau keeps Naþ/Ca2þ exchange in check, delaying

extrusion until a later time where [Ca2þ]i is lower. Another implication of

Fig. 5B is that there is net Ca2þ eZux during the contraction in rat (vs. net influx

in rabbit). This means that there must be net Ca2þ influx between contractions in

rat ventricle, and the [Ca2þ]o trace in Fig. 5B is actually going below the bath by the

end of the trace to restore the steady state balance before the next beat (�1.5 s

later). Note that during a steady state heartbeat, total Ca2þ influx must equal total

Ca2þ eZux (i.e., there is no net gain or loss of Ca2þ at the steady state).

Extracellular Ca2þ-MEs are also useful for assessing nonsteady state Ca2þ fluxes

on a longer time scale (Bers and MacLeod, 1986; MacLeod and Bers, 1987).

Figure 7A shows that when 0.5 Hz stimulation is stopped there is a very slow

small rise in [Ca2þ]o over many seconds (net Ca2þ eZux), and upon resumption of

stimulation (now at 1 Hz) that there is a net [Ca2þ]o depletion which develops over

Rabbit ventricle Rat ventricleB

Ca efflux

250 500

Ca influx

Time (ms)

0 250 500

Time (ms)

aNai= 7.2 mM aNai= 7.2 mM

ENa/CaENa/Ca

Em

Em

Fig. 6 Changes in the reversal potential of the Naþ/Ca2þ exchange (ENa/Ca) during the action

potential (Em) and Ca2þ transient in rabbit and rat ventricle. Changes in electrochemical driving force

for Naþ/Ca2þ exchange (ENa/Ca�Em) are shown in the bottom panels, assuming a 3:1 stoichiometry of

Naþ/Ca2þ exchanger and aNai are measured Naþ activity values (Shattock & Bers, 1989). Ca2þ

transients driving the contraction are assumed to be the same for both species (resting [Ca2þ]i¼150 nM,

peak [Ca2þ]i¼1 mM, 40 ms after the AP initiation). Note that Ca2þ eZux is low during rest in

rabbit myocytes because of the low [Ca2þ]i (despite a significant driving force). Based on data in

Shattock & Bers, (1989), as modified in Bers (2001), with permission.

Tension200mN

A Control

B Ryanodine

0.2 mM[Ca]o

[Ca]o1 mV

[Ca]o1 mV

Tension80mN

0.2 mM [Ca]o

30 s

Fig. 7 Measurements of [Ca2þ]o using double-barreled extracellular Ca2þ MEs in rabbit ventricular

muscle. Local [Ca2þ]o and tension are shown starting during steady state stimulation at 0.5 Hz with a

couple of pauses (0.5 and 1 min) and a period of 1 Hz stimulation. Spikes on [Ca2þ]o trace are stimulus

artifacts, and bath [Ca2þ]o was 0.2 mM. The same protocol was used before (A) and after equilibration

with 1 mM ryanodine (B; from Bers & MacLeod, (1986), with permission).


several beats. What we showed with other experiments was that this depletion

reflects the filling of SR Ca2þ stores (i.e., net transfer of Ca2þ from the extracellular

space to the SR). Moreover, the rise in [Ca2þ]o during rest reflects the gradual loss

of SR Ca2þ that depends on diastolic leak of Ca2þ from the SR and net Ca2þ

extrusion by Na/Ca exchange (thermodynamically favored in resting rabbit ven-

tricular muscle; Fig. 6A). This also allowed new insight at the time as to exactly

how ryanodine works in cardiac myocytes. Figure 7B shows that loss of cellular

Ca2þ during rest is much faster after ryanodine exposure, but that during 1 Hz

pacing the cell (and SR) can take up Ca2þ. This showed that ryanodine makes the

SR leaky, but not so much as to abolish SR Ca2þ uptake (i.e., Ca2þ could still be

driven into the SR). Once the repeated high [Ca2þ]i signals stop driving SR Ca2þ

uptake, the Ca2þwithin the SR is lost very quickly. Thus dynamics of cellular Ca2þ

flux balance can be readily assessed by double-barreled Ca2þ-MEs, both with

relatively high time resolution during steady state conditions and for longer

changes that occur in nonsteady state conditions. This makes them a nice comple-

ment to fluorescent indicators and voltage clamp studies.

3. Troubleshooting Ca2þ-Selective MEs

The ME cannot be filled with the ligand:

1. The ligand may be too thick because of the THF evaporation. Redilute PVC

by adding small amounts of THF and stirring the mixture to homogeneity.


2. The tip diameter is too small. This may entail preparing MEs with diVerentshapes and testing them (ligand filling and calibration) on the same day. Relying on

ME resistance may give misleading results because it is also aVected by the

geometry of the shank.

3. ME silanization may be insuYcient, because of insuYcient time of silaniza-

tion (or insuYcient beaker seal during exposure to silane vapors), old silane, glass

rehydration (storage problem), and insuYcient dehydration.

The ME gives bad calibration curves (subNernstian slopes, low detection limit).

1. Make sure that the calibrating solutions are adequate. Check them with

commercial macroelectrodes or the above described ETH 129-based minielec-

trodes, as described in previously published methods (e.g., Bers, 1982).

2. The ligand may be too old. This is a common occurrence. Ligand lifetime is

sensitive to exposure to light and air.

3. The ETH 129 is too diluted. This can occur by not letting THF evaporate

enough before ligand filling. In this case, it is better to let all the THF evaporate

and then, to re-add small amounts (few tens of microliter) of THF and stir the

mixture until a syrup-like solution is obtained. It can also happen if after adding

THF to a gelled cocktail, the cocktail is not mixed to homogeneity (by vortexing or

sonication).

IV. Discussion

We here describe the design and preparation of Ca2þ-selective mini- and micro-

electrodes, based on the ligand ETH 129 in a PVC matrix. Both mini- and micro-

electrodes have excellent responses during in vitro calibration with a response time

and detection limit superior to that of most commercially available minielectrodes.

These electrodes can have multiple applications. We do note, however, that for

measuring [Ca2þ]i, the ligand ETH 1001 may be preferable.

Other Ca2þ-selective ionophores with a higher selectivity for Ca2þ against other

ions, such as K23E1 (Suzuki et al., 1995), have been developed after ETH 129 but

their detection limit for calcium is considerably inferior to that of ETH 129-based

electrodes (Suzuki et al., 1995). Thus, K23E1 may be useful for clinical analysis of

calcium in plasma or other biological fluids, but ETH 129-based electrodes remain

superior for calcium measurements in the submicromolar concentration range.

ETH 129-based minielectrodes are economical, easy to prepare, and have suc-

cessfully been used for purposes where the response time of the electrode is appro-

priate. This includes preparation of calibration solutions, determination of the Kd

for EGTA, BAPTA, and oxalate in buVer solutions (Bers, 1982; Harrison and Bers,

1987; Harrison and Bers, 1989; Hove-Madsen and Bers, 1993a), and calibration of

indo-1 and furaptra signals cell suspensions (Hove-Madsen and Bers, 1992; Hove-

Madsen et al., 1998; Shannon and Bers, 1997). The minielectrodes have also been


used to titrate the passive Ca2þ binding sites in permeabilized myocytes where the

cellular Ca2þ uptake and release process are inhibited. The same titration method

has also been used to measure total Ca2þ uptake in the SR of permeabilized

cardiomyocytes (Hove-Madsen and Bers, 1993a).

Ca2þ minielectrodes have also been used together with indo-1 for online mea-

surements of the Ca2þ uptake rate in the SR in permeabilized ventricular myocytes

(Hove-Madsen and Bers, 1993b), to examine the eVects of phospholamban phos-

phorylation and temperature (Hove-Madsen et al., 1998; Mattiazzi et al., 1994)

and to measure the number of SR Ca2þ pump sites by titration with the selective

pump inhibitor thapsigargin (Hove-Madsen and Bers, 1993b; Hove-Madsen et al.,

1998). In these applications, the electrode response time can become a limiting

factor and it is important to use a fresh Ca2þ electrode for each experiment, and to

minimize inhomogeneities in cell suspensions that the calcium electrodes cannot

detect.

Although the use of Ca2þ selective electrodes for measurements of dynamic

changes in some biological systems is limited by their response time (particularly

at submicromolar concentration), the potential benefits of combining Ca2þ selec-

tive electrodes with other experimental techniques are underexplored. Indeed,

calcium selective electrodes have successfully been used to monitor [Ca2þ] insidepatch pipettes (Kang and Hilgemann, 2004; Kang et al., 2003). New Ca2þ sensors

based on coating of microcantilevers with ion-selective self-assembled monolayers

have also been developed (Ji and Thundat, 2002) and may be useful in the mapping

of Ca2þ channels and transporters on the cell surface. Indeed, measurement of the

change in extracellular ion concentration with ion-selective MEs has also been

shown to provide a noninvasive means for functional mapping of the location and

density of potassium channels (Korchev et al., 2000; Messerli et al., 2007) and for

the quantification of transmembrane Ca2þ flux (Bers, 1983, 1985, 1987; Bers and

MacLeod, 1986; Shattock and Bers, 1989; Smith et al., 1999).

Thus, in spite of the overwhelming predominance of fluorescent Ca2þ indicators,

Ca2þ-selective electrodes and biosensors still remain a valuable supplement to

many imaging and electrophysiological techniques in molecular and cellular

physiology.

Acknowledgments

This work was supported by a grant from the National Institute of Health (HL30077) to DMB and a

grant from the Spanish Ministry of Science and Technology (SAF2007-60174) to LHM.

References

Allen, P. D., Lopez, J. R., Sanchez, V., Ryan, J. F., and Sreter, F. A. (1992). EU 4093 decreases

intracellular [Ca2þ] in skeletal muscle fibers from control and malignant hyperthermia-susceptible

swine. Anesthesiology 76, 132–138.

Ammann, D. (1986). Ion-Selective Microelectrodes. Principles, Design and Application Springer,

Berlin.


Ammann, D., Buhrer, T., Schefer, U., Muller, M., and Simon, W. (1987). Intracellular neutral carrier-

based Ca2þ microelectrode with subnanomolar detection limit. Pflugers Arch. 409, 223–228.

Bers, D. M. (1982). A simple method for the accurate determination of free [Ca] in Ca-EGTA solutions.

Am. J. Physiol. 242, C404–C408.

Bers, D. M. (1983). Early transient depletion of extracellular Ca during individual cardiac muscle

contractions. Am. J. Physiol. 244, H462–H468.

Bers, D. M. (1985). Ca influx and sarcoplasmic reticulum Ca release in cardiac muscle activation during

postrest recovery. Am. J. Physiol. 248, H366–H381.

Bers, D. M. (1987). Mechanisms contributing to the cardiac inotropic eVect of Na pump inhibition and

reduction of extracellular Na. J. Gen. Physiol. 90, 479–504.

Bers, D. M. (2001) Excitation–Contraction Coupling and Cardiac Contractile Force, 2nd ed., Kluwer

Academic Press, Dordrecht, NL.

Bers, D. M., and Ellis, D. (1982). Intracellular calcium and sodium activity in sheep heart Purkinje

fibres. EVect of changes of external sodium and intracellular pH. Pflugers Arch. 393, 171–178.

Bers, D. M., and MacLeod, K. T. (1986). Cumulative depletions of extracellular calcium in rabbit

ventricular muscle monitored with calcium-selective microelectrodes. Circ. Res. 58, 769–782.

Blatter, L. A., and Blinks, J. R. (1991). Simultaneous measurement of Ca2þ in muscle with Ca electrodes

and aequorin. DiVusible cytoplasmic constituent reduces Ca2þ-independent luminescence of

aequorin. J. Gen. Physiol. 98, 1141–1160.

Blatter, L. A., andWier, W. G. (1992). Agonist-induced [Ca2þ]i waves and Ca2þ-induced Ca2þ release in

mammalian vascular smooth muscle cells. Am. J. Physiol. 263, H576–H586.

Deyhimi, F., and Coles, J. A. (1982). Rapid silylation of a glass surface: Choice of reagent and eVect of

experimental parameters on hydrophobicity. Helv. Chim. Acta 65, 1752–1759.

Fedirko, N., Svichar, N., and Chesler, M. (2006). Fabrication and use of high-speed, concentric Hþ- andCa2þ-selective microelectrodes suitable for in vitro extracellular recording. J. Neurophysiol. 96,

919–924.

Gorman, A. L., Levy, S., Nasi, E., and Tillotson, D. (1984). Intracellular calcium measured with

calcium-sensitive micro-electrodes and Arsenazo III in voltage-clamped Aplysia neurones. J. Physiol.

353, 127–142.

Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985). A new generation of Ca2þ indicators with greatly


Harkins, A. B., Kurebayashi, N., and Baylor, S. M. (1993). Resting myoplasmic free calcium in frog

skeletal muscle fibers estimated with fluo-3. Biophys. J. 65, 865–881.

Harrison, S. M., and Bers, D. M. (1987). The eVect of temperature and ionic strength on the apparent

Ca-aYnity of EGTA and the analogous Ca-chelators BAPTA and dibromo-BAPTA. Biochim.

Biophys. Acta 925, 133–143.

Harrison, S. M., and Bers, D. M. (1989). Correction of proton and Ca association constants of EGTA

for temperature and ionic strength. Am. J. Physiol. 256, C1250–C1256.

Hove-Madsen, L., and Bers, D. M. (1992). Indo-1 binding to protein in permeabilized ventricular

myocytes alters its spectral and Ca binding properties. Biophys. J. 63, 89–97.

Hove-Madsen, L., and Bers, D. M. (1993). Passive Ca buVering and SR Ca uptake in permeabilized

rabbit ventricular myocytes. Am. J. Physiol. 264, C677–C686.

Hove-Madsen, L., and Bers, D. M. (1993). Sarcoplasmic reticulum Ca2þ uptake and thapsigargin

sensitivity in permeabilized rabbit and rat ventricular myocytes. Circ. Res. 73, 820–828.

Hove-Madsen, L., Llach, A., and Tort, L. (1998). Quantification of Ca2þ uptake in the sarcoplasmic

reticulum of trout ventricular myocytes. Am. J. Physiol. 275, R2070–R2080.

Ji, H. F., and Thundat, T. (2002). In situ detection of calcium ions with chemically modified micro-

cantilevers. Biosens. Bioelectron. 17, 337–343.

Kang, T. M., and Hilgemann, D. W. (2004). Multiple transport modes of the cardiac Naþ/Ca2þ

exchanger. Nature 427, 544–548.

Kang, T.M., Markin, V. S., and Hilgemann, D.W. (2003). Ion fluxes in giant excised cardiac membrane

patches detected and quantified with ion-selective microelectrodes. J. Gen. Physiol. 121, 325–347.


Konishi, M., Olson, A., Hollingworth, S., and Baylor, S. M. (1988). Myoplasmic binding of fura-

2 investigated by steady-state fluorescence and absorbance measurements. Biophys. J. 54, 1089–1104.

Korchev, Y. E., Negulyaev, Y. A., Edwards, C. R., Vodyanoy, I., and Lab, M. J. (2000). Functional

localization of single active ion channels on the surface of a living cell. Nat. Cell. Biol. 2, 616–619.

Levy, S., and Fein, A. (1985). Relationship between light sensitivity and intracellular free Ca concen-

tration in Limulus ventral photoreceptors. A quantitative study using Ca-selective microelectrodes.

J. Gen. Physiol. 85, 805–841.

Lipp, P., Luscher, C., and Niggli, E. (1996). Photolysis of caged compounds characterized by ratio-

metric confocal microscopy: A new approach to homogeneously control and measure the calcium

concentration in cardiac myocytes. Cell Calcium 19, 255–266.

Lopez, J. R., Contreras, J., Linares, N., and Allen, P. D. (2000). Hypersensitivity of malignant

hyperthermia-susceptible swine skeletal muscle to caVeine is mediated by high resting myoplasmic

[Ca2þ]. Anesthesiology 92, 1799–1806.

MacLeod, K. T., and Bers, D. M. (1987). EVects of rest duration and ryanodine on changes of

extracellular [Ca] in cardiac muscle from rabbits. Am. J. Physiol. 253, C398–C407.

Marban, E., Rink, T. J., Tsien, R. W., and Tsien, R. Y. (1980). Free calcium in heart muscle at rest and

during contraction measured with Ca2þ-sensitive microelectrodes. Nature 286, 845–850.

Mattiazzi, A., Hove-Madsen, L., and Bers, D. M. (1994). Protein kinase inhibitors reduce SR Ca

transport in permeabilized cardiac myocytes. Am. J. Physiol. 267, H812–H820.

Messerli, M. A., Corson, E. D., and Smith, P. J. (2007). Measuring extracellular ion gradients from

single channels with ion-selective microelectrodes. Biophys. J. 92, L52–L54.

Orchard, C. H., Boyett, M. R., Fry, C. H., and Hunter,M. (1991). The use of electrodes to study cellular

Ca2þ metabolism. In ‘‘Cellular Calcium. A Practical Approach,’’ (J. G. McCormack, and

P. H. Cobbold, eds.), Oxford University Press, New York.

Picht, E., DeSantiago, J., Blatter, L. A., and Bers, D. M. (2006). Cardiac alternans do not rely on

diastolic sarcoplasmic reticulum calcium content fluctuations. Circ. Res. 99, 740–748.

Purves, R. D. (1981). Microelectrode Methods for Intracellular Recording and Ionophoresis. Academic

Press, New York.

Schefer, U., Ammann, D., Pretsch, E., Oesch, U., and Simon, W. (1986). Neutral carrier based Ca2þ-selective electrode with detection limit in the subnanomolar range. Anal. Chem. 58, 2282–2285.

Shannon, T. R., and Bers, D. M. (1997). Assessment of intra-SR free [Ca] and buVering in rat heart.

Biophys. J. 73, 1524–1531.

Shattock, M. J., and Bers, D. M. (1989). Rat vs. rabbit ventricle: Ca flux and intracellular Na assessed

by ion-selective microelectrodes. Am. J. Physiol. 256, C813–C822.

Smith, P. J. S., Hammar, K., Porterfield, D. M., Sanger, R. H., and Trimarchi, J. R. (1999). Self-

referencing, non-invasive, ion selective electrode for single cell detection of trans-plasma membrane

calcium flux. Microsc. Res. Tech. 46, 389–417.

Suzuki, K., Watanabe, K., Matsumoto, Y., Kobayashi, M., Sato, S., Siswanta, D., and Hisamoto, H.

(1995). Design and synthesis of calcium and magnesium ionophores based on double-armed diaza-

crown ether compounds and their application to an ion sensing component for an ion-selective

electrode. Anal. Chem. 67, 324–334.

Thomas, R. C. (1982). Ion-Selective Microelectrodes: How to Make and Use Them. Academic Press,

New York.

Tsien, R. Y., and Rink, T. J. (1981). Ca2þ-selective electrodes: A novel PVC-gelled neutral carrier

mixture compared with other currently available sensors. J. Neurosci. Methods 4, 73–86.

Ujec, E., Keller, O., Machek, J., and Pavlik, V. (1979). Low impedance coaxial Kþ selective microelec-

trodes. Pflugers Arch. 382, 189–192.

Vaughan-Jones, R. D., and Kaila, K. (1986). The sensitivity of liquid sensor, ion-selective microelec-

trodes to changes in temperature and solution level. Pflugers Arch. 406, 641–644.

Vaughan-Jones, R. D., and Wu, M. L. (1990). pH dependence of intrinsic Hþ buVering power in the

sheep cardiac Purkinje fibre. J. Physiol. 425, 429–448.

CHAPTER 4


Construction, Theory, and PracticalConsiderations for using Self-referencingof Ca2þ-Selective Microelectrodes forMonitoring Extracellular Ca2þ Gradients

Mark A. Messerli and Peter J. S. SmithBioCurrents Research CenterCellular Dynamics ProgramMarine Biological LaboratoryWoods Hole, Massachusetts, USA

A

OGY,All rig

bstract


-679X-679X

I. In
troduction II. C aSM Construction
A.
Micropipette Fabrication B. Silanization C. Microelectrode Construction
III. P
roperties of CaSMs A. Response to Ion Activity B. Selectivity C. Spatial Resolution D. Response Time
IV. S
elf-referencing of CaSMs A. DiVerential Concentration Measurement B. DiVerential Concentration Determination C. Calculation of Flux D. Correction for Ca2þ BuVering E. Measurement of Voltage Gradients F. Positional Artifacts
V. C
a2þ Flux Measurements R eferences
/10 $35.00(10)99004-3

92 Mark A. Messerli and Peter J. S. Smith

Abstract

Ca2þ signaling in the extra- and intracellular domains is linked to Ca2þ

transport across the plasma membrane. Noninvasive monitoring of these result-

ing extracellular Ca2þ gradients with self-referencing of Ca2þ-selective micro-

electrodes is used for studying Ca2þ signaling across Kingdoms. The quantitated

Ca2þ flux enables comparison with changes to intracellular [Ca2þ] measured with

other methods and determination of Ca2þ transport stoichiometry. Here, we

review the construction of Ca2þ-selective microelectrodes, their physical charac-

teristics, and their use in self-referencing mode to calculate Ca2þ flux. We also

discuss potential complications when using them to measure Ca2þ gradients near

the boundary layers of single cells and tissues.

I. Introduction

Regulation of resting [Ca2þ]i and the control of spatial and temporal dynam-

ics during Ca2þ signaling require coordinated transport between membrane-

separated compartments, giving rise to Ca2þ fluxes across organelles and the

plasma membrane. Movement of Ca2þ across the plasma membrane via trans-

porters, exchangers, or channels gives rise to minute gradients of [Ca2þ] in the

extracellular boundary layer that reflect changes in [Ca2þ]i. The near real-time

extraction of these gradients requires a detection method that is not disturbing

to the local chemical environment, functions over a wide dynamic range, and

possesses high sensitivity, selectivity, and spatial resolution. For these reasons

extracellular Ca2þ gradients have been monitored with self-referencing of Ca2þ-selective microelectrodes (CaSMs), enabling noninvasive characterization of

Ca2þ transport and signaling events. Unlike most fluorescent or luminescent

indicators, CaSMs were originally developed for measuring both intracellular

and extracellular [Ca2þ] (listed in Lanter et al., 1982). Measurement of minute

Ca2þ gradients on the outside of cells was limited by electrical drift in the

system. For this reason, a modulation technique was introduced (Kuhtreiber

and JaVe, 1990) that enabled reduction of drift and provided a simple means

for calculating Ca2þ flux. The method was later coined ‘‘self-referencing’’ and

has been extended to other ion-selective microelectrodes and amperometric

microelectrodes enabling characterization of fluxes of many diVerent analytes

(Messerli et al., 2006; Smith et al., 2007). Measurement of Ca2þ fluxes with self-

referencing has enabled direct comparison of Ca2þ fluxes measured with other

techniques including radioactive tracers, fluorescent and luminescent ion indi-

cators, and voltage clamp. We will first discuss the construction and general

properties of CaSMs before discussing their use with the self-referencing

approach.

4. Electrochemical Measurement of Ca2+ Flux 93

II. CaSM Construction

A. Micropipette Fabrication

Ion-selective microelectrodes are based on an ion-selective solvent or liquid

membrane, immobilized in the tip of a glass micropipette with a backfilling

electrolyte. The glass micropipette housings are pulled from 1.5 mm outer diameter

borosilicate (TW150-4 World Precisions Instruments, Sarasota, FL), aluminosili-

cate (A150-100-10 Sutter Instruments, Novato, CA), or quartz glass (Q150-110-10,

Sutter Instruments). Inner filaments, commonly used to load electrolyte solutions

to the tips of micropipettes, are avoided. Although the glass body is fragile, it

provides distinct advantages over other materials including low cost, excellent

resistive properties necessary for use with the high-resistance liquid membranes,

and easy fabrication of small tips. Micropipettes are pulled, silanized, and stored in

bulk, �50 per wire rack. The glass is pulled down to a final edge slope of 0.15–0.17

and an inner tip diameter of 2–3 mm. Borosilicate and aluminosilicate micropip-

ettes are pulled on a horizontal heated filament puller (P-97, Sutter Instruments)

while quartz pipettes are pulled on a horizontal laser puller (P-2000, Sutter Instru-

ments). Latex gloves are worn during handling of the glass before silanization.

B. Silanization

The hydrophilic glass surfaces are coated with a hydrophobic silane to enable

adhesion and high electrical resistance between the glass and the hydrophobic

liquid membrane. While many forms of silanization exist, we prefer vapor deposi-

tion of N,N-dimethyltrimethylsilylamine (cat# 41716 Sigma-Aldrich, St. Louis,

MO) as it enables rapid and uniform coating of numerous micropipettes, simulta-

neously. A wire rack of micropipettes is placed in a small solid wall metal box

(8 cm�8 cm�10 cm) with a swinging door within the oven so that the silane vapor

can be trapped in a small region around the pipettes. Prior to coating, the glass is

dried for 20 min at 240 �C under vacuum (28 in Hg). This shortens drying time and

decreases loss of hydroxyl groups (Deyhimi and Coles, 1982; Munoz et al., 1983).

Higher temperatures may dry glass more quickly as well; however, this silane has

ignited two out of four times at �250 �C. After drying, atmospheric pressure is

recovered by purging the oven with Argon. A small volume of silane (20 mL) isdropped into a tiny glass beaker in the metal enclosure and the door to the

enclosure is closed before the oven door is closed. The glass is exposed to the silane

vapor for 20 min before removing and placing the micropipettes in a sealed bell jar

with desiccant in the bottom. Functional CaSMs have been produced from micro-

pipettes that have been stored in this manner for up to a month. This method has

reduced variation in the quality of silanization.


C. Microelectrode Construction

Standard, electrolyte-based CaSMs consist of a short column of liquid membrane

(�30 mm) with a longer column of Ca2þ containing electrolyte (5 mm) used to make

electrical contact with the voltage recording headstage via a Ag/AgCl wire. Commer-

cially available vented pipette holders (WPI, Sarasota, FL, Warner Instruments,

Hamden, CT) are used to immobilize the CaSMs while loading and recording from

the high input impedance electrometers �1015 Ω (BioCurrents Research Center,

Woods Hole, MA; Molecular Devices, Sunnyvale, CA; Warner Instruments, Ham-

den,CT). CaSMs are constructed by first backfilling a fewmillimeters of the electrolyte

into a silanized micropipette with a long blunt needle and syringe, before tip loading

the liquid membrane. The backfilling electrolyte has varied from 100 nM Ca2þ buV-ered with 5 mM EGTA, 10 mMHEPES with 90 mMKCl (Tsien and Rink, 1981) to

simply 100 mM CaCl2 (Kuhtreiber and JaVe, 1990). However, based on further

discussion below it will be shown that the backfilling solution should be based on the

bath [Ca2þ] with additional electrolyte, 100 mMKCl, to make electrical contact with

the Ag/AgCl wire. Ca2þ-selective liquid membranes can be mixed in the lab or

purchased premixed (cat# 21048 (ETH1001), cat# 21196 (ETH129); Sigma-Aldrich,

St. Louis, MO).

Tip loading of the liquid membrane is performed under microscopic control,

displayed in Fig. 1A. The electrolyte filled micropipette on the right is positioned

near a loading pipette on the left, a tip brokenmicropipette that has been dip-loaded

with liquidmembrane. Both the loading pipette and the CaSM are connected to air-

filled syringes with plastic tubing so that pressure can be applied. The threaded

plunger (TP) syringe in Fig. 1A allows a small controlled pressure to create a small

bulge of liquid membrane away from the loading pipette which aids loading of the

CaSM.A plastic syringe (PS) with a three-way valve for the CaSM enables applying

and venting pressure before loading and before removing the CaSM from the

electrode holder. After positioning both the loading pipette and the CaSM within

the field of view under the microscope objective, Fig. 1B, pressure is applied to the

back of the CaSM to push the electrolyte to the tip. Pressure is vented and the tip of

the CaSM is immediately positioned within the liquid membrane bulge held in the

loading pipette. Surface tension will immediately draw the liquid membrane into

the silanizedmicropipette. A combination of pressure and suction is used to achieve

a liquid membrane column of the desired length (�30 mm). After the desired length

is achieved, the CaSM tip is removed from the liquid membrane, the back of the

CaSM is vented to atmospheric pressure, and the CaSM is removed from its holder.

III. Properties of CaSMs

A. Response to Ion Activity

The potential across the Ca2þ-selective liquid membrane in the tip of the CaSM

is comprised of two phase boundary potentials, between the interfaces of the liquid

membrane with (1) the backfilling solution and (2) the extracellular medium, and

Loading pipette

A B

Loading pipetteCaSM CaSM

TPPS

Fig. 1 Ca2þ-selective microelectrode tip filling station. (A) Micropositioners on each side of an

upright microscope are used to position the tips of a loading pipette and a CaSM in the field of view.

The stage has been removed. The threaded plunger (TP) syringe and the plastic syringe (PS) are

connected to the loading pipette and CaSM via plastic tubing enabling application of pressure and

suction to control the length of liquid membrane loaded into the CaSM from the loading pipette.

(B) Higher magnification of A showing the close positioning of the glass loading pipette and glass

CaSM. The system is mounted on a large metal plate to reduce vibration during loading.


also the diVusion potential between the two ends of the column of liquid mem-

brane (Bakker et al., 1997). The diVusion potential is considered negligible as bulk

movement of Ca2þ across the liquid membrane does not occur during common

usage with high-impedance electronics and no current flow. The inner phase

boundary potential is considered constant due to the rigorous clamping of Ca2þ

with buVers or with high concentrations of Ca2þ in the backfilling electrolyte. The

external phase boundary potential, for an ideal ion-selective microelectrode, is

related to the extracellular ion activity by the Nernst equation,

E ¼ EO þ S logai ð1Þwhere ‘‘Eo’’ is the sumof constant potential contributions, ‘S’is theNernstian slope¼(2.3RT) / (ziF) (R, T, and F hold their standardized meanings) and ‘‘ai’’ is the activity

of the primary ion. Constant potential contributions are comprised of the boundary

potentials and liquid junction potentials that exist across the circuit comprising the

reference and measuring electrodes. The valence ‘‘zi’’ of Ca2þ produces a slope only

half as steep (�29 mV/order magnitude change in Ca2þ) compared to monovalent

ions. The high selectivity of the twoCa2þ liquidmembranes discussed here alongwith

the generally standard physiological media that are used enables us to perform flux


calculations according to the Nernst equation listed above. However, in complex

media with significantly interfering ions, the slope of response can be reduced. The

decrease in slope can be predicted using the Nicolsky–Eisenman equation for ions of

similar valence and the extended Nicolsky–Eisenman equations for ions of diVerentvalence (Bakker et al., 1994). In some cases it may be more practical to perform an

empirical determination of the slope of the line describing the relationship between

measured voltage and the change in ionic activity. This determination is performed

by making up the working medium with slightly higher and lower concentrations of

Ca2þ and determining the slope of response. A sub-Nernstian response may reflect

the presence of an interfering ion or of a substance that is fouling the microelectrode.

According to the Nernst equation, the voltage output is dependent on ionic

activity. However, as ion activity is directly proportional to ion concentration, via

the activity coeYcient, and the changes that occur to the activity coeYcient due to

changes in ionic strength are negligible during self-referencing in physiological

saline, we will use concentration in place of activity for further discussion.

B. Selectivity

A primary motivation for early development of CaSMs was to monitor intracel-

lular [Ca2þ] (Lanter et al., 1982; Tsien and Rink, 1981). This required a liquid

membrane with high selectivity for monitoring the low resting [Ca2þ]i (�100 nM)

in the presence of higher concentrations of potentially interfering ions including

Kþ (�120 mM), Naþ (�10 mM) and Mg2þ (�1 mM). Accordingly, two diVerentCa2þ ionophores with very high selectivity were reported (Ammann et al., 1975,

1987; Lanter et al., 1982). Some of their selectivity coeYcients for Ca2þ over other

common cations are listed in Table I. Selectivity for Ca2þ over these cations is

relatively good compared to liquid membranes for other ions. However, not all

inorganic or organic ions have been tested and may therefore act as interferents.

Not only do interfering ions reduce the electrode’s voltage response to the primary

ion but they also slow the response time of the electrode (Bakker et al., 1997). This

point is particularly important when using the electrodes in self-referencing mode

where a temporal component is part of the modulation approach. In biological

applications, it is critical for the investigator to empirically test the voltage

response of a CaSM in the medium in which the experiments are to be performed.

Simple solutions of the primary ion are not suYcient. Additionally, the CaSM

should be tested for interference or fouling due to the addition of transport

blockers or cellular poisons.

C. Spatial Resolution

Small, micron-sized sensors give rise to high spatial resolution. The spatial

resolution is defined first by the external surface area of the Ca2þ-selective liquid

membrane, but also by the sampling time and the distance between the source of

the Ca2þ transport and the CaSM. This holds true for the high-impedance

Table ISelectivity coeYcients of two diVerent Ca2þ-selective liquidmembranes

Interfering ion (M)

Selectivity coeYcients (log CaMPot)

ETH1001 ETH129a

Kþ �5.4b �7.2

Naþ �5.5b �5.8

Mg2þ �4.9b �6.7

NH4þ �5.0c �3.6

Hþ �4.4c �2.5

aAmmann et al. (1987).bLanter et al. (1982).cAmmann et al. (1975).


headstages �1015 Ω that are typically used, which help to decrease the bulk

movement of Ca2þ between the medium and liquid membrane. Spatial resolution

is decreased due to diVusion of Ca2þ in the bulk medium from nearby transport

events. DiVusion of Ca2þ from 10 to 20 mm away will reach the CaSM in only �20

and �80 ms, indicating that the sampled volume is much larger than the immedi-

ate dimensions of the CaSM tip. As these events are diVusing from regions further

away, the local concentration change that they produce near the tip of the CaSM

will be much smaller (proportional to 1/r2) than the signals from events immedi-

ately in front of the CaSM. The decay in signal with distance is evident from

measurements of extracellular Kþ gradients due to eZux through single Kþ

channels (Messerli et al., 2009). The sampling domain of the CaSM is therefore

slightly larger than the surface area of the liquid membrane and decays rapidly

with increasing distance from the surface.

D. Response Time

Self-referencing of CaSMs requires the use of CaSMs with relatively short

response times so that the CaSM can reach equilibrium in a short period of time

at its new position. The response time of CaSMs is governed by the ability to

provide charge to the sensing node. In an ideal measuring system, diVusionthrough the unstirred layer at the surface of the electrode defines the response

time of the sensors when the liquid membrane is equilibrated with the salt of an ion

to which the electrode responds (Bakker et al., 1997). For ion-selective microelec-

trodes, this process may occur so quickly that the electronics of the system slow the

measured response (Ammann, 1986). Low input impedance of the amplifier and

parasitic capacitances in the circuit will draw more charge than an ideal system

therefore slowing the response time of the system. Amplifier input impedances of

Table IICaSMs based on ionophore ETH1001 possess short response times inphysiological saline over a range of [Ca2þ]

Response times (t95%ms)

0.1–1 mM 1–10 mM 10–1 mM 1–0.1 mM

48�7 53�10 58�9 81�10

Physiological saline consists of (in micromolar) 120 NaCl, 5 KCl, 2 MgCl2,

10 HEPES with the CaCl2 concentration listed above. CaSMs remained stationary

during the experiment, while three adjacent streams of media (1 mL/min) were rapidly

positioned (<8 ms) in front of the measuring electrode. These measurements describe

the response time of the entire measuring system for 4 CaSMs.


�106 GO are typically used to accommodate ion-selective microelectrodes that

have high resistances, 1–20 GO (Ammann, 1986). Even with the best electronics,

the time constant (RC, resistance times capacitance) of the CaSM itself imposes a

low pass filter. Resistance is primarily dependent on the tip diameter and length of

the column of liquid membrane and capacitance is primarily dependent on the

thickness and dielectric constant of the wall of the glass micropipette. To reduce

the resistance of the CaSMs, they are fabricated with relatively large tips of 2–3 mminner diameter and with short columns of liquid membrane �30 mm. The short

columns are achieved by tip loading the liquid membrane as discussed above.

Capacitance can be lowered by using thicker walled borosilicate glass (1.5 mm O.

D. 0.84 mm I.D. cat# 1B150-6, WPI Sarasota, FL). The construction design listed

above has produced CaSMs with response times shorter than 100 ms, Table II.

In practice, a slight deviation from the expected length does not change the

response time very much, at least when considering the use of these electrodes in

the self-referencing application.

IV. Self-referencing of CaSMs

A. DiVerential Concentration Measurement

Self-referencing of CaSMs was developed to measure extracellular Ca2þ gradi-

ents/currents that may have existed near previously characterized extracellular

voltage gradients (Kuhtreiber and JaVe, 1990). For example, relatively steady

eZux of Ca2þ across the plasma membrane gives rise to a gradient of Ca2þ with

a higher concentration near the cell. Self-referencing of CaSMs is implemented by

measuring the [Ca2þ] at two points in that Ca2þ gradient. The electrical variation

of a single CaSM due to thermal noise, �100–200 mV, of the high-impedance

sensors and chemical drift is too large to enable measurement of such small

diVerences in extracellular Ca2þ. As a result, a frequency sensitive method of


detection was explored based on response times of about 1 s reported for CaSMs

commonly used at that time (Ammann, 1986). The general measuring protocol

includes intermittent collection of ion concentration by a single CaSM at a position

near the biological preparation and then at a position some distance away orthog-

onal to the source. A CaSM is shown in Fig. 2A and B at the two positions, next to

a mouse pancreatic islet. The excursion distance in this case is 20 mm but can vary

between 5 and 50 mm depending on the size of the cell or cellular preparation.

At each pole of excursion, the CaSM is allowed to reach equilibrium (�0.25 s)

before recording the average local ion concentration for about 1 s. Considering

that the CaSMs possess response times of less than 0.1 s, 0.25 s is plenty of time to

reach equilibrium. The CaSM is then immediately positioned to the opposite pole,

and allowed to reach equilibrium before recording the local ion concentration. The

movement of the CaSM between the two positions is controlled by stepper motors

set to move the sensor at a rate of 40 mm/s such that it takes 0.25 s to reach its new

Near pole (E1)

A

Near pole

Far poleE2 E1

ΔE = E1− E2

ΔE

E2 E1d d d d

C

B

Far pole (E2)

50mm

Fig. 2 Ca2þ flux measurements performed with self-referencing of a Ca2þ-selective microelectrode

near a mouse pancreatic islet. (A) In the near pole the CaSM collects the average [Ca2þ]-dependentpotential for 1 s, E1. (B) After movement to the far pole and equilibration, the average [Ca2þ]-dependentpotential is collected again for 1 s, E2. (C) Data collection scheme portraying Ca2þ eZux. The auto-

mated determination of the diVerential [Ca2þ]-dependent potential, DE, is used to determine Ca2þ flux.

This measuring scheme continues, as defined by the user. The [Ca2þ]-dependent potential is discarded,(d), during periods of movement (�0.25 s) and during equilibration in the new position (�0.25 s).


position. A diVerential concentration recording between the two poles of excursion

is collected, about every 3 s. The measurement scheme continues until stopped by

the user. The diVerential recording possesses peak to peak noise of �10 mV while

longer periods of signal collection and averaging can enable extraction of concen-

tration diVerences that give rise to 1 mV diVerences or a 0.008% diVerence from the

background [Ca2þ]. Figure 2C illustrates this collection scheme during Ca2þ eZux

where E1 and E2 are the recorded ion concentration–dependent potentials at the

two poles. The recording collected during movement and equilibration at the new

pole is discarded, labeled ‘‘d’’ in Fig. 2C. A diVerential Ca2þ measurement is

collected over a period of about 3 s, which is faster than the low frequency drift,

thus reducing its influence on the measurement. Signal averaging at each pole over

a period of 1 s reduces the influence of the high frequency noise. Measurement of

the diVerential ion concentration–dependent voltage, DE, between the two posi-

tions over time enables further enhancement of the signal-to-noise ratio.

This modulation approach was termed self-referencing, referring to the fact that

the measurements, collected by a single CaSM, are compared to each other in

order to determine the concentration diVerence between the two points. The signal

collected by a single CaSM at one point in space and time is referenced to the

signal collected by that same CaSM at a diVerent point in space and time in order

to reduce electrical drift of the measuring system. The CaSM has its own

bath reference electrode. While this diVerential measurement could be achieved

with two similar, CaSMs, positioned at known distances from the source, the

sensitivity would suVer from the signal drift and noise inherent to two separate

measuring systems.

Measured diVerential voltages of �10s mV are extracted from relatively large

oVset potentials�100s mV by using a combination of amplification methods. Low

gain must be used with the large oVset potentials in order to keep the signal within

the dynamic range of the amplifier. As a result, low resolution digital systems will

not be able to register small changes in the diVerential voltage. A 12-bit system

with a dynamic range of �10 V provides only 4.9 mV/bit resolution while a 16-bit

system provides only 0.3 mV/bit. Additional amplification prior to digitization is

necessary to resolve signals at or below 1 mV. Two separate methods for amplifica-

tion are used; (1) a nearly equal and opposite electrical oVset is supplied before

amplification (sample hold mode) and (2) a running average of the low gain

measurement is subtracted from the real-time input before amplification (RC

subtract mode). Sample hold mode applies a known voltage that is selected either

manually or automatically from the signal after a set duration of time to null the

oVset potential before applying 103 times gain. The primary disadvantage for this

mode is that drift can take the system back out of the dynamic range of the

amplifier so that a new potential must be applied regularly. The advantage is

that it does not need an additional correction factor to compensate for the signal

lost due to the filtering that occurs in RC subtract mode. In RC subtract mode, a

high-pass filter is used to collect a running average potential that is subtracted

from the potentials collected in the near and far pole. The signals are then amplified


103 times before digitizing. Typically, this mode employs a high-pass filter with a

time constant of 10 s. RC subtract allows amplification for systems with large drift

but involves a correction factor to oVset the high-pass filter. The correction factor

will be dependent on the time constant of the high-pass filter and the period of data

acquisition. For standard conditions, a period of 3.3 s (0.3 Hz translation frequen-

cy), 40 mm/s translation speed, 10 s time constant of the high-pass filter along with

data selection of the last 70% of the half cycle, we calculate that the signal is 7%

smaller than a square wave with similar rise time.

Automated, repetitive positioning of the CaSMs is controlled by three stepper

motors arranged in an X, Y, Z configuration with the Z plane parallel to the plane

of the stage of the microscope. Smooth linear motion is obtained by coupling each

of the stepper motors to a lead screw controlling the position of three small

translational plates connected together to form a three-dimensional positioner

(BioCurrents Research Center, Woods Hole, MA). Low voltage control of the

stepper motors prevents electrical feedback to the high-impedance headstage of the

CaSM. Positioning can be achieved over a working distance of 3–4 cm with

submicron resolution and repeatability (Danuser, 1999). A computer interface

enables repetitive motion and positional control with the Faraday cage closed.

B. DiVerential Concentration Determination

The relationship between the measured diVerential voltage and the diVerentialion concentration between the two poles of excursion for an ideal CaSM can be

determined using the Nernst equation.

E1 � E2 ¼ EO þ S logCið Þ1 � EO þ S logCið Þ2DE ¼ S logCið Þ1 � S logCið Þ2DE ¼ logCS1 � logCS2

i 1ð Þ i 2ð ÞCS1

i 1ð Þ !

ð2Þ
DE ¼ log
CS2

i 2ð Þ

‘‘E1’’, ‘‘Ci(1)’’, and ‘‘S1’’ are the measured voltage, [Ca2þ] and slope of the voltage–

log(Ci) graph for the near pole of excursion. The subscript 2 labels the same

parameters for the far pole of excursion. The slow changing constant potential

contributions ‘‘Eo’’ are reduced if not eliminated by calculating the diVerencebetween potentials over short periods of time.

Equation (2) enables a clear picture of the relationship of the sensitivity of

detection to the background ion concentration during measurements. For a

given [Ca2þ] change due to cellular flux, the concentration in the position next to

the cell, Ci(1), is the sum of the background ion concentration and the concentra-

tion change generated by the source while Ci(2), in most cases, is close to the

background ion concentration. It is easier to generate a larger DE when the

background concentration of the measured ion is lower as the ratio of Ci(1)/Ci(2)


will be much larger/smaller for the same Ca2þ eZux/influx on lowered background

[Ca2þ]. This has led to the lowering of the background [Ca2þ] in order to generate

DE with a greater signal-to-noise ratio, see Table IV. Care must be taken to ensure

that changing the background concentration does not interfere with normal cellu-

lar activity.

Rearrangement of Eq. (2) relates the [Ca2þ] in the near pole of excursion to the

[Ca2þ] at the far pole of excursion.

Cið1Þ ¼ CS2S1

ið2Þ � 10DES1 ð3Þ

For an ideal electrode, the voltage output is close to the Nernstian slope over a

wide range of [Ca2þ] so S1 ¼ S2 ¼ S ¼ 2:3RTziF

. Therefore, Eq. (3) simplifies to

Ci 1ð Þ ¼ Ci 2ð Þ � 10DES ð4Þ

For minute fluxes that are typically measured with self-referencing, the average

concentration of Ca2þ at the far pole, position 2, is not too diVerent from the

average concentration of Ca2þ in the bulk solution. Therefore the diVerence in

[Ca2þ] between the two points of excursion can be described as follows:

DC ¼ Ci 1ð Þ � Ci 2ð Þ ¼ Cbath � 10DES � Cbath ð5Þ

A primary assumption here is that the excursion distance is small compared to the

extent that the gradient extends out into the bulk solution so that the concentration

diVerence between the two excursion points is linear. For minute gradients

measured from small cells (�10 mm diameter), an excursion of 10 mm will most

likely sample over a distance in which the concentration diVerence is not linear andtherefore will lead the investigator to underestimate the true flux. Incorrect estima-

tion of the true flux could also occur during a two-point measurement in a more

intense, extended gradient, where the concentration of the ion in the far pole is

substantially diVerent from the background concentration of the ion. In both of

these cases, a three-point measurement should be performed in order to (1) more

carefully map the concentration gradient with a third point to ensure a linear

relationship or determine a more accurate nonlinear relationship and (2) to deter-

mine the concentrations in the gradient relative to the background concentration

of the ion in the bath.

The selectivity of Ca2þ liquid membranes is relatively good compared to liquid

membranes for other ions. Therefore, measurement of Ca2þ gradients in the

presence of a constant concentration of an interfering ion or in the presence of

a gradient of an interfering ion is not a major concern. However, specific

circumstances may require the use of higher concentration of an interfering

ion and the two cases need to be addressed. Details necessary to account for

these situations have been addressed previously (Messerli et al., 2006; Smith

et al., 2007).


C. Calculation of Flux

The diVerential concentration measurement is converted to flux to provide a

direct representation of the number of ions passing through a unit area per unit

time. Calculation of flux enables comparison of Ca2þ transport between diVerentsystems as it takes into account the diVusion coeYcient of Ca2þ, the distance overwhich the diVerential concentration measurement was acquired, the surface geom-

etry of the source, and the distance of measurement from the source. It also

provides a value for comparison of Ca2þ flux measured with self-referencing of

CaSMs to other methods for monitoring Ca2þ including intracellular fluorescent

and luminescent ion indicators and radioactive tracer flux studies. For planar

sources where the measuring electrode is relatively close to a large source, such

as a tissue, sheet of cells or large diameter cell, and the diVerential concentrationis measured over a small distance ‘‘Dx’’ within the gradient next to the source,

flux (J) is

J ¼ �DDCDx

ð6Þ

where ’’D’’ is the diVusion coeYcient of Ca2þ. By this model, at equilibrium the

flux measured at some distance from the source is the same as the flux at the surface

of the source. According to this equation, eZux, a higher concentration of Ca2þ

near the source, is identified by a negative flux.

In order to determine flux at the cell surface for known surface geometries, it is

useful to calculate analyte flow, that is, the quantity of substance (Q) moving per

unit time (Henriksen et al., 1992). Flow is the same for all concentric surfaces

surrounding the source surface. Flux at the source surface is the flow divided by the

surface area of the source. Therefore, radially emanating flow from a cylindrical

surface is

Flow ¼ Q

t¼ � 2pD

ln b=að Þ DCð Þ ð7Þ

where ‘‘D’’ is the diVusion coeYcient of the analyte and ‘‘a’’ and ‘‘b’’ are the radial

distances between the center of the cylinder and the electrode tip at the near and far

poles, respectively. These equations have been adapted from Crank (1967). Ana-

lyte flux at the surface of the cylinder is then determined by dividing by its surface

area 2prl. A caveat of this approach is the assumption that the flow is equal at all

points around the cylinder and along the shaft of the cylinder. An alternative is to

calculate flux per unit length by dividing by 2pr (Henriksen et al., 1992).

The flow from a spherical source is

Flow ¼ Q

t¼ �4pD

ab

b� aDCð Þ ð8Þ

Flux at the cell surface can then be determined by dividing by the spherical surface

area 4pr2.


D. Correction for Ca2þ BuVering

The presence of Ca2þ buVers or binding agents with the appropriate aYnity can

lead to collapse of Ca2þ gradients by shuttle buVering (Speksnijder et al., 1989).

Ca2þ can diVuse from the surface of the cell in either its free state or bound to the

buVer. CaSMs only measure the free concentration of Ca2þ. The actual Ca2þ

flux at a source is the sum of the measured free Ca2þ flux and the unmeasured

Ca2þ flux diVusing as Ca2þ bound to buVer.

JCa total ¼ JCa measured þ JCa Buffer ð9ÞKnowing the conditions under which the Ca2þ flux was measured including the

[Ca2þ] of medium, dissociation constant, Kd, of the buVers and concentration of

the buVers present, a simple relationship can be derived to determine the ratio of

Ca2þ diVusing bound to buVer compared to the freely diVusing Ca2þ. ShuttlebuVering of Hþ is a bigger concern than for Ca2þ due to the larger number of

Hþ buVers that are used in physiological media. The equations that exist to correct

for shuttle buVering of Hþ (Messerli et al., 2006; Smith et al., 2007) can be adapted

for use with Ca2þ flux correction and may be necessary under specific

circumstances.

xi ¼ DB

DCa2þB½ � Kd

Kd þ Ca2þ� �� 2 ð10Þ

The correction factor, ‘‘xi’’, is the ratio of the Ca2þ bound buVer flux to the free

Ca2þ flux. Therefore

JCa total ¼ JCa measured 1þ xi þ þ xnð Þ ð11Þwhere a number of diVerent Ca2þ buVers (xi þ þ xn) may be collapsing the Ca2þ

gradient. The correction factor is based on three criteria, the ratio of the diVusioncoeYcients of the Ca2þ-buVer complex to free Ca2þ, the Ca2þ buVer concentration,and the dissociation constant, Kd, of the Ca

2þ buVer compared to the [Ca2þ] of themedium. The Kd is the inverse of the more commonly used ‘‘stability constant’’ also

known as the association constant. Only Ca2þ buVers/binding agents that have Kd

values near the range of the extracellular [Ca2þ] will act as eVective shuttle buVersduring self-referencing of CaSMs. Table III lists a few of these compounds. Note

that two of the compounds, ADA and Bicine, are commonly used as Hþ buVers.Generally, the Ca2þ Kd values of other Good buVers are not in the range of normal

extracellular [Ca2þ] or are very poor Ca2þ chelators (Dawson et al., 1986).

E. Measurement of Voltage Gradients

The use of CaSMs with self-referencing is subject to a similar problem that

occurs with the use of intracellular CaSMs; specifically they detect not only

changes in ion concentration but also voltage. Extracellular voltage gradients

Table IIIList of Ca2þ binding compounds that can act as shuttle buVers for extra-cellular Ca2þ in the 0.1–1.0 mM range

Ligand log(Kd)

Pyrophosphate �5.0

N-(2-acetamido)iminodiacetic acid (ADA) �4.01

ATP �3.8

Citric acid �3.5

Oxalic acid �3.0

Polyphosphate �3.0

N,N-bis(hydroxyethyl)-glycine (Bicine) �2.8

Values were obtained fromDawson et al. (1986) except as noted 1(Lance et al., 1983).


have been mapped near many diVerent systems (Borgens et al., 1989; Nuccitelli,

1986) Extracellular electric fields generated by cells are generally very small espe-

cially in high conductivity media such as animal saline. However, in lower conduc-

tivity saline ion transport can give rise to relatively large electric fields >1 mV/10 mm which can be detected with self-referencing microelectrodes. Plants for

example, drive transcellular currents through them, as a result of ion transport

across single cells or tissues. In low conductivity medium, these currents generate

substantial voltage gradients next to the cells, coexisting with the concentration

gradients of the transported ions. The diVerential voltage measured by the CaSM

will be the sum of the voltage diVerences due to the [Ca2þ] diVerence and the

voltage diVerence. For example, a peak voltage diVerence during oscillating cur-

rent influx of about 9 mV would occur over a 10-mm distance immediately in front

of a lily pollen tube. Peak current density around 0.4 mA/cm2 was measured at a

distance of about 20 mm from the cell surface with a medium resistivity of about

5000 O cm (Messerli and Robinson, 1998). This voltage diVerence is just above thebackground noise of the system used at that time,�5 mV for Ca2þ, (Messerli et al.,

1999). The voltage signals detected by the self-referencing CaSM peaked about six

times larger than the diVerential voltage due to current flux indicating that the

extracellular electric field could have contributed to the calculated Ca2þ flux by up

to 15%.

F. Positional Artifacts

Self-referencing of CaSMs near solid objects can generate position dependent

artifacts. Movement of Ca2þ across the external interface between the liquid

membrane and bathing medium may occur through current driven and zero net

current mechanisms (Bakker and MeyerhoV, 2000). Release of Ca2þ by the CaSM

restricts its sensitivity in bulk medium by leading to a modification of the local ion


concentration at the tip of the microelectrode. During self-referencing, when the

CaSM is positioned near a solid object, released Ca2þ can accumulate between the

CaSM and the object in a short period of time leading to an artificially higher

concentration of Ca2þ in the constrained space. Likewise, uptake of Ca2þ by the

CaSM can lead to a depletion of Ca2þ in the constrained space. These artifacts are

most apparent in solutions of low background [Ca2þ]. Figure 3 shows examples of

both extremes where CaSMs are self-referenced near a 100-mm diameter glass

bead. The CaSM is moving in a path so that the plane of its tip is always parallel

to the near surface of the bead to enable the ISM to get closer to the surface. The

ion trapping eVect is reduced when the path of excursion orients the plane of the tip

of the CaSM perpendicular to the near surface of the solid object, as shown in

Fig. 2A and B, because the liquid membrane surface cannot get as close to the solid

object. EZux of Ca2þ from the microelectrode tip occurs when constructed with

100 mM CaCl2 backfilling solution, originally performed by Kuhtreiber and JaVe(1990). Accumulation of the released Ca2þ in less than 1 s can be detected when the

bath [Ca2þ] is 50 mM but not when it is 2 mM, giving rise to an artificial eZux of

Ca2þ from the solid glass bead. Reducing the concentration of the primary ion in

the backfilling solution is one method of reducing the ion leak (Bakker and

MeyerhoV, 2000). However, when used with self-referencing this can lead to an

artifact of the opposite polarity shown in Fig. 3. The CaSM constructed with

−40

−30

−20

−10

0

10

20

30

40

50

0 5 10 15 20 25 30

Distance from bead (mm)

Diff

eren

tial v

olta

ge (mV

)

100 mM Ca2+ backfill, 50mM Ca2+ bath

100 mM Ca2+ backfill, 2 mM Ca2+ bath

100 nM Ca2+ backfill, 50mM Ca2+ bath

100 nM Ca2+ backfill, 2 mM Ca2+ bath

50mM Ca2+ backfill, 50mM Ca2+ bath

Fig. 3 Ca2þ movement across the tip of a CaSM can be detected in low background [Ca2þ] near asolid object. Electroneutral exchange of Ca2þ out of the tip of a CaSM (filled box) or into the tip of the

CaSM (filled circle) can give rise to accumulation or depletion of the local [Ca2þ] between a solid object

and the tip of the CaSM. In higher bath [Ca2þ] (empty box, empty circle) the accumulation or depletion

is insignificant compared to the background [Ca2þ] and is therefore not detected. In lowered bath [Ca2þ]careful balancing of the backfilling [Ca2þ] with the bath [Ca2þ] can reduce (filled triangle) if not

eliminate net movement of Ca2þ across the liquid membrane.


100 nM Ca2þ in the backfilling solution generates an influx of Ca2þ into the CaSM

tip, depleting the local [Ca2þ] in the bath and giving rise to an artificial Ca2þ influx

into the glass bead. Again this can be detected in the 50 mM Ca2þ solution but not

the 2 mM Ca2þ containing bath solution. The artifact can be reduced by matching

the backfilling [Ca2þ] with the [Ca2þ] in the bath. Other methods for eliminating

Ca2þ flux across the tip of the CaSM include current clamping (Lindner et al.,

1999; Pergel et al., 2001), or using the solid contact ion-selective electrode design

(Lindner and Gyurcsanyi, 2009).

V. Ca2þ Flux Measurements

Extracellular Ca2þ flux measurements have been performed on a number of

diVerent systems some of which are listed in Table IV, ranging from animal

neurons and muscle to tip growing root hairs, pollen tubes, and fungi. Measured

Ca2þ fluxes are relatively small ranging between 0.1 and 10 pmol cm�2 s�1 encour-

aging measurements from cells in reduced background [Ca2þ] 0.1 mM. The limit

of flux sensitivity for a typical self-referencing CaSM with �10 mV near real-time

variation performed in 1 mM bath [Ca2þ] is about �6.3 pmol cm�2 s�1, an order

of magnitude higher than in 0.1 mM bath [Ca2þ]. Considering the large trans-

plasma membrane electrochemical driving force on Ca2þ, reduction of extra-

cellular [Ca2þ] by an order of magnitude did not cause noticeable problems

for the diVerent preparations, at least over the few hour period during

which measurements were acquired as noted by multiple authors listed in

Table IV.

While eZux of Ca2þ in cells at rest is expected to be relatively small, the

measured influx of Ca2þ, presumably through channels, is also relatively small.

Active single 0.5 pS Ca2þ channels at a density of 1 mm�2 should give rise to a Ca2þ

influx of about 47 pmol cm�2 s�1. Although as noted by Hille (2001) voltage-gated

Ca2þ channels exist at low density and low open probabilities (<0.1) even with

strong depolarizing potentials indicating that low channel density and activity is

suYcient to account for measured changes in [Ca2þ]i. The channel density and

activity used above may be overestimates of actual Ca2þ channel density. Also,

weak influx may also be a result of the Ca2þ amplification cascades that exist to

release Ca2þ from intracellular stores after influx through the plasma membrane.

Additional directions for the use of self-referencing with CaSMs include the

study of electroneutral Ca2þ transporters/exchangers and extracellular Ca2þ sig-

naling (Breitwieser, 2008). Ca2þ selective microelectrodes have been instrumental

in providing the sensitivity for defining the complex transport of the Naþ/Ca2þ

exchanger (Kang and Hilgemann, 2004) and the P-type plasma membrane Ca2þ

pump (PMCA) in neurons (Thomas, 2009). With self-referencing of ion-selective

microelectrodes, transport stoichiometries could be determined noninvasively

from the outside of intact cells.

Table IVCalcium flux measurements acquired from preparations representing multiple Kingdoms

Preparation Ca2þ flux (pmol cm�2 s�1) Conditions

Bath [Ca2þ](mM) Reference

Aplysia californica bag cell �1 to �5 Rest, H2O2 0.1 Duthie et al. (1994)

�1 to �5 Rest, thapsigargin None added

(0.5 mM

EGTA)

Knox et al. (1996)

Rana catesbeiana hair cell �0.5 Rest 0.05 Yamoah et al. (1998)

þ5.0 Stimulated

Callinectes sapidus olfaction �2.5 15% ASWa 0.1 Gleeson

et al. (2000b)

�4.0 AFWb 0.1 Gleeson

et al. (2000a)

Sclerodactyla

briareus smooth muscle

�1.0 to �4.0 Rest, Ach.c 0.1 Devlin and Smith

(1996)

�1 to �7.5 Muscarinic

agonists

0.1 Devlin et al. (2003)

Busycon canaliculatum

cardiac muscle

�1 to �4 Rest, FMRFamide 0.1 Devlin (1996)

Mouse ova �0.02 Rest 0.05 Pepperell

et al. (1999)þ0.6 Bepridil additiond

�0.2 Replenished Naþ

þ0.08 to þ0.35 Addition of EGFe None added Hill et al. (1999)

Lilium longiflorum

pollen tubes

þ2 to þ20 Germination 0.1 Pierson et al. (1994)

þ5 to þ38f Oscillating

tip growth

0.13 Messerli et al. (1999)

Root hairs þ4.3 to þ7.2 (alfalfa) Tip growth,

nod factor

Not listed Herrmann and Felle

(1995)

þ2.5 (S. alba) Tip growth 0.1

þ0.07 to þ1.2 (A. thaliana) Osmotic regulation 0.1 Lew (1998)

Neurospora crassa hyphae �0.1 Voltage

dependence

0.05 Lew (2007)

þ0.1 to þ1.5 Osmotic regulation 0.05 Lew and Levina

(2007)

Ceratopteris richardii spores �3.5 top Gravity sensing 0.02–0.05 Chatterjee

et al. (2000)þ0.5 bottom

Physcomitrella

patens filaments

þ1–3 Gravity sensing 0.1 Allen et al. (2003)

aArtificial seawater.bArtificial freshwater.cAcetylcholine.dBepridil was added to block the plasma membrane Naþ/Ca2þ exchanger.eEpidermal growth factor.fCalculated flux at cell surface.


Acknowledgments

The BioCurrents Research Center is funded by NIH:NCRR grant P41 RR001395.


References

Allen, N. S., Chattaraj, P., Collings, D., and Johannes, E. (2003). Gravisensing: Ionic responses,

cytoskeleton and amyloplast behavior. Adv. Space Res. 32(8), 1631–1637.

Ammann, D. (1986). Ion-Selective Microelectrodes. Springer-Verlag, Berlin.

Ammann, D., Guggi, M., Pretsch, E., and Simon, W. (1975). Improved calcium ion-selective electrode

based on a neutral carrier. Anal. Lett. 8(10), 709–720.

Ammann, D., Buhrer, T., Schefer, U., Muller, M., and Simon, W. (1987). Intracellular neutral carrier-

based Ca2þ microelectrode with subnanomolar detection limit. Pflugers Arch. Eur. J. Physiol. 409,

223–228.

Bakker, E., and MeyerhoV, M. E. (2000). Ionophore-based membrane electrodes: New analytical

concepts and non-classical response mechanisms. Anal. Chim. Acta 416, 121–137.

Bakker, E., Meruva, R. K., Pretsch, E., andMeyerhoV, M. E. (1994). Selectivity of polymer membrane-

based ion-selective electrodes: Self-consistent model describing the potentiometric response in mixed

ion solutions of diVerent charge. Anal. Chem. 66, 3021–3030.

Bakker, E., Buhlmann, P., and Pretsch, E. (1997). Carrier-based ion-selective electrodes and bulk

optodes. 1. General characteristics. Chem. Rev. 97, 3083–3132.

Borgens, R. B., Robinson, K. R., Vanable, J. W. J., and McGinnis, M. E. (1989). Electric Fields in

Vertebrate Repair. Alan R. Liss, Inc., New York.

Breitwieser, G. E. (2008). Extracellular calcium as an integrator of tissue function. Int. J. Biochem. Cell

Biol. 40, 1467–1480.

Chatterjee, A., Porterfield, D. M., Smith, P. J. S., and Roux, S. J. (2000). Gravity-directed calcium

current in germinating spores of Ceratopteris richardii. Planta 210, 607–610.

Crank, J. (1967). The Mathematics of DiVusion. Oxford University Press, London.

Danuser, G. (1999). Photogrammetric calibration of a stereo light microscope. J. Microsc. 193, 62–83.

Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M. (1986). Data for Biochemical

Research. Oxford University Press, Oxford.

Devlin, C. L. (1996). A vibrating Ca2þ-selective electrode measures Ca2þ flux induced by the neuropep-

tide FMRFamide in a gastropod ventricle. Comp. Biochem. Physiol. 116A(2), 93–100.

Devlin, C. L., and Smith, P. J. S. (1996). A non-invasive vibrating calcium-selective electrode measures

acetycholine-induced calcium flux across the sarcolemma of a smooth muscle. J. Comp. Physiol. 166,

270–277.

Devlin, C. L., Amole, W., Anderson, S., and Shea, K. (2003). Muscarinic acetylcholine receptor

compounds alter net Ca2þ flux and contractility in an invertebrate smooth muscle. Invert. Neurosci.

5, 9–17.

Deyhimi, F., and Coles, J. A. (1982). Rapid silyation of a glass surface: Choice of reagent and eVect of

experimental parameters on hydrophobicity. Helv. Chim. Acta 65, 1752–1759.

Duthie, G. G., Shipley, A., and Smith, P. J. S. (1994). Use of a vibrating electrode to measure changes in

calcium fluxes across the cell membranes of oxidatively challenged Aplysia nerve cells. Free Rad. Res.

20(5), 307–313.

Gleeson, R. A., Hammar, K., and Smith, P. J. S. (2000a). Sustaining olfaction at low salinities: Mapping

ion flux associated with the olfactory sensilla of the blue crab Callinectes sapidus. J. Exp. Biol. 203,

3145–3152.

Gleeson, R. A., McDowell, L. M., Aldrich, H. C., Hammar, K., and Smith, P. J. S. (2000b). Sustaining

olfaction at low salinities: Evidence for a paracellular route of ion movement from the hemolymph to

the sensillar lymph in the olfactory sensilla of the blue crab Callinectes sapidus. Cell Tissue Res. 301,

423–431.

Henriksen, G. H., Raman, D. R., Walker, L. P., and Spanswick, R. M. (1992). Measurement of net

fluxes of ammonium and nitrate at the surface of barley roots using ion-selective microelectrodes.

Plant Physiol. 99, 734–747.

Herrmann, A., and Felle, H. (1995). Tip growth in rooth hair cells of Sinapis alba L.: Significance of

internal and external Ca2þ and pH. New Phytol. 129, 523–533.


Hill, J. L., Hammar, K., Smith, P. J. S., and Gross, D. J. (1999). Stage-dependent eVects of epidermal

growth factor on Ca2þ eZux in mouse oocytes. Mol. Reprod. Dev. 53, 244–253.

Hille, B. (2001). Ion Channels of Excitable Membranes. Sinauer Associates, Inc., Sunderland.

Kang, T. M., and Hilgemann, D. W. (2004). Multiple transport modes of the cardiac Naþ/Ca2þ

exchanger. Nature 427, 544–548.

Knox, R. J., Jonas, E. A., Kao, L.-S., Smith, P. J. S., Connor, J. A., and Kaczmarek, L. K. (1996). Ca2þ

influx and activation of a cation current are coupled to intracellular Ca2þ release in peptidergic

neurons of Aplysia californica. J. Physiol. 494(3), 627–639.

Kuhtreiber, W. M., and JaVe, L. F. (1990). Detection of extracellular calcium gradients with a calcium-

specific vibrating electrode. J. Cell Biol. 110, 1565–1573.

Lance, E. A., Rhodes, C. W., III, and Nakon, R. (1983). Free metal ion depletion by ‘‘Good’s’’ buVers.

Anal. Biochem. 133, 492–501.

Lanter, F., Steiner, R. A., Ammann, D., and Simon, W. (1982). Critical evaluation of the applicability

of neutral carrier-based calcium selective microelectrodes. Anal. Chim. Acta 135, 51–59.

Lew, R. R. (1998). Immediate and steady state extracellular ionic fluxes of growing Arabidopsis thaliana

root hairs under hyperosmotic and hypoosmotic conditions. Physiol. Plant. 104, 397–404.

Lew, R. R. (2007). Ionic currents and ion fluxes in Neurospora crassa hyphae. J. Exp. Bot. 58(12),

3475–3481.

Lew, R. R., and Levina, N. N. (2007). Turgor regulation in the osmosensitive cut mutant of Neurospora

crassa. Microbiology 153, 1530–1537.

Lindner, E., and Gyurcsanyi, R. E. (2009). Quality control criteriz for solid-contact, solvent polymeric

membrane ion-selective electrodes. J. Solid State Electrochem. 13, 51–68.

Lindner, E., Gyurcsanyi, R. E., and Buck, R. P. (1999). Tailored transport through ion-selective

membranes for improved detection limits and selectivity coeYcients. Electroanalysis 11, 695–702.

Messerli, M. A., and Robinson, K. R. (1998). Cytoplasmic acidification and current influx follow

growth pulses of Lilium longiflorum pollen tubes. Plant J. 16, 87–91.

Messerli, M. A., Danuser, G., and Robinson, K. R. (1999). Pulsatile influxes of Hþ, Kþ and Ca2þ lag

growth pulses of Lilium longiflorum pollen tubes. J. Cell Sci. 112, 1497–1509.

Messerli, M. A., Robinson, K. R., and Smith, P. J. S. (2006). Electrochemical sensor applications to the

study of molecular physiology and analyte flux in plants. In ‘‘Plant Electrophysiology,’’

(A. G. Volkov, ed.), pp. 73–107. Springer, Berlin, Heidelberg, New York.

Messerli, M. A., Collis, L. P., and Smith, P. J. S. (2009). Ion trapping with fast response, ion-selective

microelectrodes enhances detection of extracellular ion channel gradients. Biophys. J. 96, 1597–1605.

Munoz, J.-L., Deyhimi, F., and Coles, J. A. (1983). Silanization of glass in the making of ion-sensitive

microelectrodes. J. Neurosci. Methods 8, 231–247.

Nuccitelli, R. (ed.) (1986). Ionic currents in development. Progress in Clinical and Biological Research.

Alan R. Liss, Inc, New York.

Pepperell, J. R., Kommineni, K., Buradagunta, S., Smith, P. J. S., and Keefe, D. L. (1999). Transmem-

brane regulation of intracellular calcium by a plasma membrane sodium/calcium exchanger in mouse

ova. Biol. Reprod. 60, 1137–1143.

Pergel, E., Gyurcsanyi, R. E., Toth, K., and Lindner, E. (2001). Picomolar detection limits with current-

polarized Pb2þ ion-selective membranes. Anal. Chem. 22, 509–514.

Pierson, E. S., Miller, D. D., Callaham, D. A., Shipley, A., Rivers, B. A., Cresti, M., and Hepler, P. K.

(1994). Pollen tube growth is coupled to the extracellular calcium ion flux and the intracellular

calcium gradient: EVect of BAPTA-type buVers and hypertonic media. Plant Cell 6, 1815–1828.

Smith, P. J. S., Sanger, R. H., and Messerli, M. A. (2007). Principles, development and applications of

self-referencing electrochemical microelectrodes to the determination of fluxes at cell membranes.

In ‘‘Electrochemical Methods for Neuroscience,’’ (A. C. Michael, and L. M. Borland, eds.),

pp. 373–405. CRC Press, Boca Raton.

Speksnijder, J. E., Miller, A. L., Weisenseel, M. H., Chen, T. H., and JaVe, L. F. (1989). Calcium buVer

injections block fucoid egg development by facilitating calcium diVusion. Proc. Natl. Acad. Sci. USA

86, 6607–6611.


Thomas, R. C. (2009). The plasma membrane calcium ATPase (PMCA) of neurones is electroneutral

and exchanges 2 Hþ for each Ca2þ or Ba2þ ion extruded. J. Physiol. 587, 315–327.

Tsien, R. Y., and Rink, T. J. (1981). Ca2þ-selective electrodes: A novel PVC-gelled neutral carrier

mixture compared with other currently available sensors. J. Neurosci. Methods 4, 73–86.

Yamoah, E. N., Lumpkin, E. A., Dumont, R. A., Smith, P. J. S., Hudspeth, A. J., and Gillespie, P. G.

(1998). Plasma membrane Ca2þ-ATPase extrudes Ca2þ from hair cell stereocilia. J. Neurosci. 18(2),

610–624.

CHAPTER 5


Practical Aspects of Measuring IntracellularCalcium Signals with Fluorescent Indicators

Joseph P. Y. Kao, Gong Li, and Darryl A. AustonCenter for Biomedical Engineering and Technology, andDepartment of PhysiologyUniversity of Maryland School of MedicineBaltimore, Maryland, USA

OGY,All rig

A

Vhts

bstract

OL. 99 0091reserved. 113 DOI: 10.1016/S0091

-679X-679X

I.
I ntroduction II. F luorescent Ca2þ Indicators III. L oading Indicators into Cells
A.
Limited Aqueous Solubility of AM Esters B. Dye Compartmentalization: Loading of Indicator into Subcellular
Compartments Other than the Cytosol

(

C.
Dye Leakage or Extrusion from Cells D. Procedure for Loading
IV.
M anipulation of [Ca2þ] A. Using EGTA and BAPTA as Extracellular Ca2þ BuVers B. Lowering Extracellular [Ca2þ] C. Divalent Cation Ionophores D. BuVering Changes in Intracellular [Ca2þ]
V.
C onversion of Indicator Fluorescence Signal into Values of [Ca2þ] A. Calibrating a Nonratiometric Fluorescent Indicator B. Calibrating a Ratiometric Fluorescent Indicator
VI.
R eporting Indicator Fluorescence Intensity Changes without Calibration A. Reporting Relative Changes in Fluorescence: F/F0 and DF/F0 B. Caveat in Interpreting Relative Fluorescence Changes: Indicator
Fluorescence is Not a Linear Function of [Ca2þ]
V II. M easuring [Ca2þ] in Mitochondria
A.
Estimating the Fraction of Intracellular Rhod-2 Indicator that Resides inMitochondria
B.
Minimizing Rhod-2 Loading in the Cytosol C. Monitoring Cytosolic and Mitochondrial [Ca2þ] Simultaneously
/10 $35.0010)99005-5

114 Joseph P. Y. Kao et al.

V
III.
1

conc2

while

emis

excit

over

C

Sy

en

A

e

sio

at

a

oncluding Remarks

mbols used: Ca2þ, free calcium ion; [Ca2þ], concentration of free calcium ions; [Ca

tration of free calcium ions.

n excitation spectrum is taken by monitoring fluorescence emission intensity at a fixe

xcitation light is scanned through a wavelength range over which the sample can abso

n intensity is plotted as a function of the excitation wavelength. To collect an emissi

ion light at a fixed wavelength is delivered to the sample while the emission intensity

wavelength range. Here, the emission intensity is plotted as a function of emission w

2þ]i,

d wa

rb l

on s

is m

avel

R
eferences
Abstract

The use of fluorescent indicators for monitoring calcium (Ca2þ) signals and for

measuring Ca2þ concentration ([Ca2þ]) in living cells is described. The following

topics are covered in detail: (1) ratiometric and nonratiometric fluorescent indicators

and the principles underlying their use, (2) techniques for loadingCa2þ indicators and

Ca2þ buffers into living cells, (3) calibration of indicator fluorescence intensity

measurements to yield values of intracellular [Ca2þ], (4) analysis of nonratiometric

fluorescence intensity data and caveats relating to their interpretation, (5) techniques

for manipulating intracellular and extracellular [Ca2þ], and (6) the use of fluorescent

indicators tomonitorCa2þ signals inmitochondria. The chapter aims topresent these

fundamental topics in a manner that is practically useful and intuitively accessible.

The origins of key mathematical equations used in the article are outlined in two

appendices.

I. Introduction

In the application of anymeasurement technique, a body of practical knowledge is

shared by experienced practitioners. Although important for making successful

measurements, such lore, which sometimes seems arcane, often is not described

explicitly or explained in journal publications. In this respect, measuring [Ca2þ]1

with fluorescent indicators is no exception. The purpose of this chapter is to gather in

one place some of the most common and useful practical information relevant to the

use of fluorescent Ca2þ indicators. Such a collection of information is hoped to

alleviate the frustration of those who are novices at using fluorescent indicators.

II. Fluorescent Ca2þ Indicators

The commonly available fluorescent indicators for Ca2þ fall into two operation-

al classes: dual-wavelength ratiometric dyes and single-wavelength nonratiometric

dyes (Table I). Chemical structures of some of the indicators listed in Table I are

shown in Fig. 1. For nonratiometric indicators, a change in [Ca2þ] brings about acorresponding change in the intensity of the indicator’s fluorescence excitation and

emission spectra,2 whereas the wavelengths of the excitation and emission spectral

cytosolic

velength

ight. The

pectrum,

onitored

ength.

Table IProperties of common fluorescent Ca2þ indicatorsa

Indicator type Kd (nM)

Absorption maxima

(nm) Emission maxima (nm)

Ca2þ-free Ca2þ-bound Ca2þ-free Ca2þ-bound

Nonratiometric

Monomeric

Quin2 115b 352 332 492 498

Fluo-2/Fluo-8c 380 – 492 – 514

Fluo-3 400 503 506 526 526

Fluo-4 345 491 494 �516 516

Calcium Green-1TM 190 506 506 531 531

Calcium Green-2TM 550 506 503 536 536

Calcium Green-5NTM 14d 506 506 532 532

Oregon Green 488 BAPTA-1TM 170 494 494 523 523

Oregon Green 488 BAPTA-2TM 580 494 494 523 523

Rhod-2 1.0d 556 553 576 576

Calcium OrangeTM 328 549 549 575 576

Calcium CrimsonTM 185 589 589 615 615

Dextran-conjugatede

Fluo-4 dextran (MW 10,000) �600 �494 �494 �518 �518

Calcium Green-1 dextran (MW 3000–70,000)f �240–540f �509 �509 �534 �534

Oregon Green 488 BAPTA-1 dextran (MW 10,000) �265 �496 �496 �524 �524

Ratiometric

Monomeric

Fura-2 224b 362 335 512 505

Fura RedTM 140 473 436 670 655

Indo-1 250b 349 331 485 410

Dextran-conjugatede

Fura(-2) dextran (MW 10,000) �240 364 338 501 494

aData from Tsien (1980), Grynkiewicz et al. (1985), Minta et al. (1989), Haugland (1992), and Molecular Probes, The

Handbook (web publication, Invitrogen Corporation).bEVective Kd in the presence of 1 mMMg2þ. (Generally, competition byMg2þ slightly reduces the aYnity of any indicator

for Ca2þ.)cDiVerent names for the same molecule.dmM.eThe Kd and absorption and emission maxima of dextran-conjugated indicators can vary from lot-to-lot and is dependent

on the molecular weight of the dextran used as well.fKd is reported to be diVerent between low- and high-MW versions: �540 nM for MW 3000, �240–250 nM for MW

10,000 and 70,000.

5. Measuring [Ca2þ] with Fluorescent Indicators 115

peaks remain essentially unchanged. Excitation spectra of Fluo-3 (Minta et al.,

1989), a nonratiometric indicator, at saturating and ‘‘zero’’ [Ca2þ] are shown in

Fig. 2. It can be seen that peak excitation occurs at �505 nm irrespective of

whether the indicator is Ca2þ-free or Ca2þ-bound—the defining characteristic of

a nonratiometric indicator. In contrast, ratiometric indicators exhibit not only

intensity changes with changing [Ca2þ] but the Ca2þ-free and Ca2þ-bound forms

CO2−

CO2− CO2

−CO2

−

CO2−

CO2−

−O2C

−O2C

CO2−

CO2− −O2C

−O2C

CO2−

CO2− −O2C

−O2C

−O2C −O2C−O2C

CO2−

CO2−

−O2C−O2C

CO2−

CO2−

−O2C−O2C

−O2C

−O2C−O2C

N N N N

N

N

O O O O O

CO2−

CO2−

CO2−

CO2−

N

N

N

OOO

O

O

N

N N

OO

N N

OO

−O2C

N N

OO

O

O

CO2−

N

NH

N

N

N+

H

N

O

O Rhod-2

Indo-1Fura-2

BAPTAEGTA Quin2

NH

O

O

O

Calcium green-2

Oregon green 488 BAPTA-2ClF

O

O

O

O

X

XFluo-

R

X

2 or −8*

3

4

H H or CH3

CH3

CH3

Cl

F

R

X

X X

X

X

O−

O−

−O

Fig. 1 Structures of selected fluorescent Ca2þ indicators and the Ca2þ chelators, EGTA, and BAPTA.

All molecules are represented in their polycarboxylate, Ca2þ-sensitive forms. The following conventions

have been used in these structural drawings: (1) Implicit carbon: Every unlabeled vertex, whether

internal or terminal, represents a carbon atom. (2) Implicit hydrogen: Every carbon has a suYcient

number of (undrawn) hydrogens to make the total number of bonds to that carbon equal to 4. (3)

Explicit heteroatoms: non-carbon, non-hydrogen atoms (e.g., O, N) are labeled explicitly; hydrogens

attached to the heteroatom are also explicitly drawn. For example, OH is equivalent to

CH3–CH¼CH–CH2–OH. *DiVerent names for the same molecule.


350

Ca2+-bound

Ca2+-free

400

Wavelength (nm)

Flu

ores

cenc

e in

tens

ity

450300

Fig. 3 Excitation spectra of Ca2þ-bound and Ca2þ-free forms of Fura-2 (lemission¼505 nm).

440420 460

Wavelength (nm)

Ca2+-free

Ca2+-bound

Flu

ores

cenc

e in

tens

ity

480 500 520

Fig. 2 Excitation spectra for Fluo-3 (lemission¼525 nm). The Ca2þ-free form of Fluo-3 is �100 times

less bright than the Ca2þ-bound form.


of the indicator actually have distinct spectra, the maxima in which occur at

diVerent wavelengths (the spectra show wavelength shifts). The two ratiometric

indicators most commonly used are Fura-2 and Indo-1 (Grynkiewicz et al., 1985).

For Fura-2, significant shifts are observed in the excitation spectra (Fig. 3) but not

in the emission spectra. Indo-1 shows a significant shift primarily in its emission

spectra. For nonratiometric indicators, because intensity monitored at a single-


wavelength is the only experimental measurement that is related to [Ca2þ], intensitychanges arising from factors unrelated to changes in [Ca2þ] (e.g., changes in cell

thickness, leakage of indicator from the cell) can confound interpretation of the

intensity data. In contrast, because the Ca2þ-free and Ca2þ-bound forms of ratio-

metric indicators are characterized by spectral peaks at diVerent wavelengths,

intensity measurements can be made at two diVerent wavelengths, and the ratio

between these intensities is quantitatively related to [Ca2þ] (Grynkiewicz et al.,

1985). Obtaining a ratio minimizes the eVect of many artifacts that are unrelated

to changes in [Ca2þ]—for example, a change in cell thickness or indicator loss from

the cell would aVect intensities at the two wavelengths equally, so the eVect wouldcancel when the two intensities are ratioed. The two commonly used ratiometric

indicators, Fura-2 and Indo-1, require excitation in the ultraviolet (UV) range,

whereas most of the common nonratiometric dyes use visible excitation light.

Although the ratiometric dyes can be calibrated more reliably (Section V), some-

times avoiding using UV light for excitation may be necessary (e.g., UV can excite

significant autofluorescence in some biological preparations and can photolyze

photosensitive ‘‘caged’’ compounds). Clearly, in practice, instrumentation for

using ratiometric indicators is more complex than that for nonratiometric

indicators.

Quin2 (Tsien, 1980; Tsien et al., 1982) is the archetypal tetracarboxylate indicator

listed in Table I (structure in Fig. 1). Its properties and applications as a nonratio-

metric indicator have been reviewed in detail (Tsien and Pozzan, 1989). However,

Quin2 has been superseded by new generations of nonratiometric and ratiometric

indicators. Of the nonratiometric indicators listed in Table I, the Fluo and Calcium

Green series as well as Oregon Green 488 BAPTA-2 incorporate fluorescein chro-

mophores and are, therefore, excited at wavelengths typical of fluoresceins. The

Fluo dyes, Calcium Green-2 and Oregon Green 488 BAPTA-2 exhibit the largest

intensity changes in their transition from Ca2þ-free to Ca2þ-bound forms (�100-

fold; Haugland, 1992;Minta et al., 1989). This change can be an advantage because,

for a given rise in [Ca2þ], these indicators give a larger increase in brightness

compared to other nonratiometric indicators. Because fluorescence quantum

eYciency3 can range only from 0 to 1, the large intensity diVerence between Ca2þ-bound and Ca2þ-free forms implies that the Ca2þ-free forms of the two indicators

must be only weakly fluorescent. Some researchers find this fact annoying because

cells with relatively low resting [Ca2þ]i (cytosolic free Ca2þ concentration) would

have most of the indicator in the Ca2þ-free form and therefore would be quite dim.

Rhod-2, Calcium Orange, and Calcium Crimson are indicators that incorporate

rhodamine-type chromophores and therefore are excited at much longer wave-

lengths than are the Fluo and Calcium Green dyes. When the acetoxymethyl (AM)

3 Fluorescence quantum eYciency, symbolized as FF or QF, is the fraction of total light absorbed

that is emitted as fluorescence. Fluorescence quantum eYciency may also be thought of as the proba-

bility that a molecule will emit fluorescence after absorbing a photon. Being a probability, the quantum

eYciency can have a value between 0 and 1.


ester is used to load cells, Rhod-2 loads well into mitochondria; up to 80% of the

intracellular dye is located in these organelles. The use of Rhod-2 to monitor Ca2þ

signals in mitochondria is outlined in Section VII.

A choice of ratiometric indicator can be made on practical grounds. Typically,

Fura-2 is excited alternately at two diVerent wavelengths, whereas the emission is

collected at a single fixed wavelength. Therefore, the pair of intensity measure-

ments, whether in imaging or in single-cell microfluorometry, must be collected

sequentially. Indo-1, on the other hand, usually is excited at a fixed wavelength

whereas emission is monitored simultaneously at two diVerent wavelengths, that is,emission from the Ca2þ-bound and Ca2þ-free forms of the indicator can be

collected simultaneously. Therefore, Indo-1 potentially can give better temporal

resolution. However, in conventional imaging, Indo-1 can be more diYcult to use

because the two emission images, usually collected through slightly diVerentoptical paths, can be diYcult to keep in spatial registration. Fura-2 has been the

most widely used ratiometric Ca2þ indicator, both in conventional imaging and in

single-cell measurements. Indo-1 has, however, been used successfully in UV laser-

scanning confocal imaging applications (Motoyama et al., 1999; Niggli et al., 1994;

Sako et al., 1997). Fura RedTM is touted as a ratiometric indicator whose excita-

tion and emission wavelengths are both in the visible range. This indicator suVersfrom having very low fluorescence quantum eYciency (�0.013 in the Ca2þ-freeform; J.P.Y. Kao, unpublished results4). Fura Red diVers from the other ratio-

metric indicators because its fluorescence intensity decreases upon binding Ca2þ.The relatively low quantum eYciency implies that higher indicator concentrations

and/or higher excitation light intensities are required.

The dextran-conjugated dyes are biopolymers with pendant indicator molecules.

The dextran-conjugated indicators listed in Table I are available with dextran

molecular weights of 3000, 10,000, or 70,000 (Invitrogen Corporation, Molecular

Probes Brand). Being membrane-impermeant, dextran conjugates must be loaded

into cells by an invasive technique such as microinjection. Whereas the monomeric

indicators can leak out of cells at a steady rate (Section III.C), dextran-conjugated

indicators tend to have long residence times in cells. Therefore, dextran-conjugated

dyes can be useful in applications in which long-term monitoring of [Ca2þ]i isrequired. Instances also occur in which cells rapidly transport monomeric dyes

into internal organelles (Hepler andCallaham, 1987) but do not do so when dextran

conjugates are used. Because the conjugates are made by covalent attachment of

monomeric indicators to dextran polymers, individual indicator monomers can

reside in slightly diVerent local microenvironments on the polymer. Therefore, the

conjugates, rather than having a unique Kd and identical spectral properties, are

characterized by a range of microscopic Kds and a distribution of spectral proper-

ties. These characteristics provide a likely explanation for lot-to-lot variations inKd

and spectral characteristics.

4 The quantum eYciency of the Ca2þ-free form of Fura Red was determined relative to carboxy-

SNARF-1.

Esterases

IntracellularExtracellular

Indicator

Indicator

Membrane-permeant

Ca2+

AMO2C

AMO2CAMO2C

AMO2CCO2AM

CO2AMCO2AM

CO2AM

Trapped inside cell

Indicator

Ca2+-sensitive-insensitive

−O2C−O2C

CO2−

CO2−

Fig. 4 Schematic representation of how incubation with the acetoxymethyl (AM) ester results in

intracellular accumulation of a polycarboxylate indicator. The hydrophobic (lipophilic) AM ester

readily diVuses into the cell through the cell membrane. Abundant cellular esterases cleave the AM

ester groups to generate the Ca2þ-sensitive form of the indicator which, being a polyanion, cannot

escape through the cell membrane and is, therefore, trapped inside the cell.


III. Loading Indicators into Cells

The common fluorescent indicators for Ca2þ are polycarboxylate anions that

cannot cross lipid bilayer membranes and therefore are not cell-permeant. In the

negatively charged form, the indicators can be introduced into cells only by

microinjection or through transient cell permeabilization, procedures that require

some special equipment and skill.5 By far the most convenient way of loading an

indicator into cells is incubating the cells in a dilute solution or dispersion of the

AM ester of the indicator. This process is represented schematically in Fig. 4. The

AM group is used to mask the negative charges on the carboxyl groups present in

the indicator molecule. The AM ester form of the indicator is uncharged and

hydrophobic. Consequently, it can pass through the cell membrane and enter the

cell interior. The carboxyl groups in the indicator, however, are essential to the

ability of the indicator molecule to sense Ca2þ; therefore, the AM groups must be

removed once the AM ester has entered the cell. Because the AM group is labile to

enzymatic hydrolysis by esterases present in the cell, the AM esters are processed

intracellularly to liberate the Ca2þ-sensitive polycarboxylate form which, being

multiply charged, becomes trapped inside the cell. Trapping of the polyanionic

form of the indicator allows cells to accumulate up to hundreds of micromolar of

5 A variety of techniques for loading membrane-impermeant species into cells is discussed byMcNeil

(1989, 2001).


the Ca2þ-sensitive form of the indicator when incubated with micromolar concen-

trations of the AM ester in the extracellular medium. Several factors influence the

eYciency and quality of indicator loading via the AM ester, and will be discussed

subsequently.

A. Limited Aqueous Solubility of AM Esters

AM esters of the common Ca2þ indicators have molecular weights in excess of

1000. Being large uncharged organic molecules, these esters have very low solubili-

ty in aqueous media. For example, at 25 �C, the solubility of the AM ester of Fura-

2 in pure water is only 0.11 mM (Kao et al., 1990). In biological media, in which the

ionic strength is typically �0.15 M, the solubility of Fura-2 AM would be even

lower. Addition of AM ester in excess of the solubility limit simply would result in

precipitation of solid AM ester, which is eVectively unavailable for loading cells. Inaddition, fine particles of solid AM ester often adhere well to the outer surfaces of

cells or to the extracellular matrix and can contribute large Ca2þ-insensitivefluorescence signals to the measurement.6 A convenient solution to the solubility

problem is the use of Pluronic F-127, a mild nonionic surfactant,7 as a dispersing

agent for AM esters. Typically, aliquots of Pluronic and AM ester stock solutions

in dimethylsulfoxide (DMSO) are mixed intimately before dispersal into an aque-

ous medium.8 The Pluronic is presumed to sequester the AM ester in micellar form,

thus preventing precipitation, and the micelles serve as a steady source to replenish

AM esters taken up by cells. The net result is significantly improved loading of

indicators into cells. Details of the loading procedure are described in Section III.D

below.

B. Dye Compartmentalization: Loading of Indicator into Subcellular CompartmentsOther than the Cytosol

1. Minimizing Compartmentalization

In typical experiments, one usually wishes to monitor changes in the concen-

tration of Ca2þ in the cytosol; therefore, ideally, one would like the indicator to

be loaded exclusively into the cytosol. This ideal situation is almost never

realized for two reasons. First, because the AM ester form of the indicator is

membrane-permeant, it can enter not only the cytosol but all subcellular

6 This is a problemwith AM ester that are fluorescent (e.g., Fura-2 AM and Indo-1 AM) but not with

nonfluorescent AM esters (e.g., AM ester of the Fluo series).7 Pluronic F-127 is manufactured by BASF Wyandotte Corporation, Wyandotte, Wisconsin. Being

a surfactant used on the industrial scale, it is inexpensive.8 With slight warming, Pluronic F-127 can be dissolved in DMSO at almost 25% (w/v). Such a highly

concentrated stock solution is inconvenient to use, however: as the solution absorbs moisture from the

air, the Pluronic will precipitate, and the resulting particulate suspension is diYcult to pipette. A 15–20%

solution in DMSO is a convenient formulation.


membrane-enclosed compartments as well. Although this process occurs to a

large extent in the cytosol, enzymatic hydrolysis of AM esters also can take

place within subcellular organelles. Therefore, some fraction of the indicator

molecules tend to be trapped in noncytosolic compartments. Second, some cell

types actively endocytose material from the incubation medium (Malgaroli

et al., 1987), including dispersed AM esters, which are then hydrolyzed to

release fluorescent indicator molecules within organelles of the endocytotic

pathway. Presumably, indicator molecules liberated in this way can end up in

a variety of organelles that are connected to the endocytotic pathway by

vesicular traYc. Because most subcellular organelles [e.g., endoplasmic reticu-

lum (ER), lysosomes] tend to have high intraorganellar [Ca2þ] (>mM), indica-

tors confined to these organelles would be saturated with Ca2þ and would

contribute a high-[Ca2þ] fluorescence signal that would not vary with changes

in cytosolic [Ca2þ]. Therefore, the net eVect of compartmentalized dye is

biasing measured cytosolic free Ca2þ concentration toward higher values.

The first cause of dye compartmentalization just stated is a reflection of an

inherent imperfection of the AM ester loading technique and cannot be remedied

easily. The second cause is a cell biological process and can be attenuated.

Because endocytosis is a temperature-dependent process, cells loaded at lower

temperatures with AM ester tend to show less compartmentalization of indicator

(Malgaroli et al., 1987). The following results illustrate this point. In REF52

fibroblast cells, roughly 30% of total intracellular Fura-2 was in noncytosolic

compartments when loading via the AM ester was carried out at 37 �C, whereasonly 10% was compartmentalized when loading was performed at 23 �C. Although

endocytosis is known to be blocked at 10 �C in mammalian cells and at 4 �C in

amphibian cells, loading at the lowest biologically permissible temperature does

not necessarily yield the best results, because processing of AM esters by esterases

in the cytosol is also temperature-dependent. At very low temperatures, the con-

centration of indicator accumulated in the cytosol can be quite low. Optimal

loading temperature is determined empirically to be the temperature at which

dye compartmentalization is minimized while good cytosolic loading is main-

tained. In practice, convenience often dictates loading at room temperature as a

reasonable compromise.9

2. Assessing Extent of Compartmentalization

The extent of indicator compartmentalization can be estimated through a simple

semiquantitative procedure that is based on the observation that micromolar

concentrations of digitonin primarily permeabilizes the plasma membrane

9 When loading is done in air, the incubation medium should be bicarbonate-free (i.e., some other

buVer such as HEPES should be used to maintain the pH of the medium). Otherwise, steady loss of CO2

will shift the CO2–HCO3�–CO3

2� equilibrium and rapidly alkalinize the medium.

0

Flu

ores

cenc

e in

tens

ity (

a.u.

)

100 200

Digitonin

Triton X-100

Fi

Fd

Fb

Time (s)

300 400

Fig. 5 Procedure for assessing dye compartmentalization. REF52 fibroblast incubated with 1 mMFura-2 AM dispersed with Pluronic F-127 in Minimal Essential Medium (MEM) for 60 min in an air

incubator at 37 �C.Measurementwas done inHanks’ Balanced Salt Solution (HBSS) containing 2.6 mM

EGTA (suYcient to reduce extracellular [Ca2þ] to <1 mM), and buVered at pH 7.4 with HEPES.

The concentration of digitonin was 20 mM; that of Triton X-100, 1%, w/v. (a.u. = arbitrary units).


and release cytosolic dye, whereas 1% Triton X-100 can permeabilize and release

dye from subcellular organelles (e.g., see Kao et al., 1989). The procedure consists

of monitoring total fluorescence from a cell or a field of cells, preferably bathed

in low-Ca2þ medium,10 and treating the cells first with digitonin and then

with Triton X-100. Figure 5 shows the procedure being applied to a cell loaded

with Fura-2 AM. The fluorescence measured before treatment (Fi) represents

contributions from cytosolic and compartmentalized dye plus background.

After digitonin release of cytosolic dye, the fluorescence measured (Fd) represents

compartmentalized dye plus background. The fluorescence level attained after

Triton X-100 treatment is considered the background (Fb). The fraction of total

intracellular dye that is compartmentalized in organelles is simply (Fd�Fb)/

(Fi�Fb).

10 Nominally Ca2þ-free medium (i.e., medium from which Ca2þ salts have been omitted) or, better

still, medium containing suYcient EGTA to make [Ca2þ]< 1 mM (see Section IV.B.1). It is important to

keep [Ca2þ] low in this experiment because when [Ca2þ] is elevated, permeabilized cells lose their

integrity rapidly, and dye molecules may leak from organelles even before Triton is applied.

B

Frac

tion

rem

aini

ng

0 5 10 15

Time (min)

20

Proben.

Sulfin.

Noinhibitor

35�C

25

1.0

0.9

0.8

0.7

0.6

1.0A

0.9

0.8

Frac

tion

rem

aini

ng

0.7

0.6

0.50 10 20

35 �C

30 �C

Time (min)

30

Fig. 6 (A) Time course of indicator loss at 30 and 35 �C from REF52 cells loaded with Fura-2. (B)

EVect of probenecid (1 mM) and sulfinpyrazone (75 mM) on indicator extrusion from REF52 cells at

35 �C; arrowmarks the time of drug application. For all measurements, the cells were loaded with Fura-

2 by incubation with AM ester/Pluronic dispersion in HEPES-buVered HBSS at 25 �C. Experiments

were done in HBSS at the specified temperatures. Total Fura-2 fluorescence (FT¼F340þF380) is

monitored over time. Each trace was normalized by dividing by the value of FT at time 0.


C. Dye Leakage or Extrusion from Cells

Once an indicator is loaded into cells, it leaks out at a rate that is strongly tempera-

ture-dependent. For mammalian cells, the loss rate is maximal at 37 �C and drops oVsharply as temperature is lowered,11 as shown in Fig. 6A. Loss of indicator occurs by

an extrusion mechanism for organic anions (Di Virgilio et al., 1988) and can be

blocked eVectively by inhibitors of uric acid transport such as probenecid and

sulfinpyrazone (DiVirgilio et al., 1988, 1990), as illustrated inFig. 6B. Sulfinpyrazone

at tens to hundreds ofmicromolar has been found to be eVective, whereas probenecidworks at millimolar dosage. High concentrations of inhibitor virtually can stop

indicator extrusion but also can induce signs of cellular stress such as blebbing.

Therefore, using the minimum concentration of inhibitor necessary to reduce the

rate of indicator loss to a tolerable level without attempting to block the process

altogether is advisable. In general, slowing indicator loss by lowering the temperature

at which experiments are conducted is preferable to using transport inhibitors.

11 The sharp increase in rate of indicator leakage with rising temperature is another reason why

compartmentalization is a more severe problem when cells are incubated with AM esters at higher

temperatures. Higher temperature accelerates loss of dye from the cytosol but not from organelles, so

compartmentalized dye becomes a larger proportion of the total intracellular dye content.


Probenecid and sulfinpyrazone are both hydrophobic organic acids and, as such, are

practically insoluble in water. These molecules must be neutralized with a stoichio-

metric amount of base (e.g., NaOH) before an aqueous stock solution can be

prepared.

D. Procedure for Loading

Stock solutions of the AM ester of the indicator can be made with dry DMSO as

solvent. Typical concentrations can be in the range of 0.1–10 mM. Such DMSO

solutions may be stored safely in screw-capped polypropylene microcentrifuge

tubes in the freezer for many months without apparent degradation. Dry DMSO

is used for making a solution of Pluronic F-127 at a concentration of 15–20% (w/w

or w/v is not crucial). This stock solution can be stored at room temperature. When

exposed to air, this concentrated stock solution slowly absorbs moisture until

Pluronic F-127 begins to precipitate. Because a heterogeneous Pluronic stock is

diYcult to transfer, preparing a fresh stock is preferable at this stage. Usually one

can load cells by incubation in medium containing (nominally) 0.1 to a few tens of

micromolar AM ester in a Pluronic dispersion. Typically 0.5–1 mL of Pluronic

stock is suYcient for dispersing 1–10 nmol AM ester. Thus, 1 mL of 20% Pluronic

stock is adequate for dispersing 1 mL of 10 mM AM ester stock into 1 mL of

medium to yield a 10 mM AM ester solution (1000-fold dilution). Using a dilution

of �1:1000 minimizes the DMSO concentration in the final loading medium (to a

few parts per thousand). Premixing the requisite volumes of AM ester stock and

Pluronic stock is advisable since this minimizes the chances of AM ester precipita-

tion during aqueous dispersal. The Pluronic-AM mixture is then dispersed into

aqueous loading medium.

Serum proteins [e.g., bovine serum albumin (BSA)] often have a salutary eVecton cells and also can improve loading eYciency, possibly by acting as hydrophobic

carriers for the AM ester and preventing its precipitation. Thus, including BSA

(0.5–1%) or serum (a small percentage) in the incubation medium may be

advantageous.

Enzymatic processing of an AM species to yield the Ca2þ-sensitive indicator

typically consists of sequential hydrolysis of four to eight AM ester groups,

depending on the particular indicator chosen. Only when all the AM groups on a

molecule are hydrolyzed does the molecule become properly Ca2þ-sensitive. Forthis reason, after removal of AM-containing loading medium, allowing the cells

some extra time (e.g., 20 min at room temperature) to complete intracellular

processing of the most recently trapped, partially hydrolyzed AM species can be

useful in some cases.

It is reassuring to convince oneself that suYcient AM ester is present in the

loading medium, so the amount of indicator taken up by cells is not limited by the

amount available. For a typical example, assume that 1 million cells are loaded in

2 mL medium containing 2 mMAM ester. The total amount of AM ester available


is 4 nmol. If the cells are 20 mm in diameter, then the 1 million cells have a total

intracellular volume of 4.2 mL. Further, if cells are loaded to a final intracellular

indicator concentration of 150 mM (a generous estimate), then the total amount of

AM uptake by cells is 0.63 nmol, which is still much less than the 4 nmol available

in the loading medium.

IV. Manipulation of [Ca2þ]

In studying Ca2þ-dependent cellular processes, raising or lowering intracellular

or extracellular [Ca2þ] is frequently desirable. Conventional techniques for achiev-ing these ends require the use of Ca2þ buVers or ionophores and will be discussed in

this section.

A. Using EGTA and BAPTA as Extracellular Ca2þ BuVers

Because it is highly selective for binding Ca2þ over Mg2þ,12 EGTA is the most

commonly used Ca2þ buVer. However, because two of the ligand atoms in EGTA

are tertiary alkylamino nitrogens, the two highest pKas of EGTA are 8.90 and

9.52,13 implying that at physiological pH EGTA will exist primarily as protonated

species—a fact that is illustrated more quantitatively in Fig. 7. For example, Fig. 7

shows that, at pH 7.2, �98% of EGTA in solution exists as H2EGTA2�, �2% as

HEGTA3�, and only a negligible fraction is in the EGTA4� form. Therefore, the

Ca2þ-binding reaction near physiological pH is fairly represented as

H2EGTA2� þ Ca2þ > CaEGTA2� þ 2Hþ

That two Hþ ions are liberated in the binding reaction means that the binding of

Ca2þ by EGTA should have very steep pH dependence, as a plot of pK0d(Ca)

14

versus pH indeed shows (Fig. 8). For a concrete example, a drop in pH from 7.2 to

7.1 changes the K0d(Ca) of EGTA by a factor of �1.6, that is, small errors in pH

can lead to significant uncertainties in the dissociation constant. In contrast,

12 ForEGTA,DpKd¼ pKd(Ca2þ)� pKd(Mg2þ)¼ 5.58; therefore, EGTAbindsCa2þmore tightly than

Mg2þ by a factor of 380,000 (i.e., 105.58). For comparison, in the case of EDTA, DpKd ¼ 1.78, which

represents only a 60-fold diVerence in EDTA’s aYnity for Ca2þ and Mg2þ. BAPTA [1,2-bis(o-aminophe-

noxy)ethane-N,N,N0,N0-tetraacetic acid] has a selectivity similar to that of EGTA: DpKd ¼ 5.20.13 At 25�C and 0.10M ionic strength. Data pertaining to EGTA that are used in this section are from

Martell and Smith (1974).14 In the metal chelator literature, Kd is used for the ‘‘absolute’’ (or intrinsic) dissociation constant

and represents the dissociation constant characterizing the fully deprotonated form of the chelator. K0d

represents Kd that has been corrected for the weakening eVect of acidic pH (thus K0d is the working

dissociation constant at a specific pH). This convention (Kd vs. K0d) is not followed consistently in the

applications literature. Details of how pH correction is applied to convert Kd into K0d are described in

Appendix 1.

100

H2EGTA2−

Free

EG

TA e

xist

ing

asH

2EG

TA2−

, HE

GTA

3−, o

r E

GTA

4− (%

)

EGTA4−

HEGTA3−

50

60

7 8 9pH

10 11 12

Fig. 7 Percentage of free EGTA existing as H2EGTA2�, HEGTA3�, and EGTA4� as a function of

solution pH. Calculations performed with data for EGTA (at 0.1 M ionic strength, 25 �C) tabulated by

Martell and Smith (1974); see Appendix 1 for algebraic details.


knowing that the two highest pKas of BAPTA are 5.47 and 6.36 (Tsien, 1980), one

infers that the ability of BAPTA to bind Ca2þ should be only very weakly

dependent on pH, as shown in Fig. 8. Comparison of the two traces in Fig. 8

shows that BAPTA has the advantage of being only weakly pH dependent in the

physiological pH range. The fact that the two traces cross between pH 7.2 and 7.3

implies that EGTA has the potential advantage of being a progressively stronger

binder of Ca2þ above the crossover point (e.g., about two- and ninefold stronger

than BAPTA at pH 7.5 and 7.8, respectively). The pH insensitivity of BAPTA

makes it a less troublesome Ca2þ buVer to use, although it is more costly than

EGTA.

B. Lowering Extracellular [Ca2þ]

In an experiment, lowering extracellular [Ca2þ] is often desirable. Depending on

how low one wishes to clamp the extracellular [Ca2þ], one of the approaches

described in the following sections may be adopted. The procedures require either

a stock solution of 1 M Na2H2EGTA at a pH near neutral or a stock solution of

1 M Na4BAPTA.

12

12

11

11

10

10

9

9

pH

BAPTA

EGTA

8pK� d

8

7

7

6

6

5

4

Fig. 8 Plot of pK0d(Ca) versus pH for EGTA and BAPTA. Calculations performed with data for

EGTA from Martell and Smith (1974) and for BAPTA from Tsien (1980).


1. Lowering [Ca2þ] to <1 mM but not Approaching 0

a. Using EGTANa2H2EGTA can be added directly to calcium-containing medium at a concen-

tration that is 3–4 times15 the concentration of Ca2þ in the medium. Because binding

of Ca2þ to H2EGTA2� releases protons and acidifies the medium, simultaneously

adding TRIS base [tris(hydroxylmethyl)aminomethane] at a concentration equal to

�2.2 times the Ca2þ concentration in the medium is also necessary. This amount of

TRIS scavenges protons released from H2EGTA2� and maintains the pH of the

medium at nearly the level before EGTA addition. This procedure can be applied

confidently down to pH �6.8. At pH levels that are much lower, the strong pH

sensitivity weakens EGTA and makes it inconvenient to use as a Ca2þ scavenger.

b. Using BAPTAAdd Na4BAPTA at a concentration equal to 3 times the Ca2þ concentration in

the medium. Essentially no pH adjustments are necessary. Because of its relative

insensitivity to decreasing pH (Fig. 8), BAPTA can be used more conveniently for

scavenging Ca2þ at lower pH values than EGTA.

15 3 times if pH � 7.0, 4 times if pH < 7.0.


2. Lowering [Ca2þ] to a Level Approaching 0

If one desires a medium in which the free Ca2þ concentration approaches

‘‘zero,’’ one can prepare nominally calcium-free medium16 that also contains

EGTA or BAPTA at millimolar concentrations. Again, to avoid problems of

rapid weakening of the Ca2þ aYnity of EGTA with decreasing pH, using

BAPTA at pH<7 is more convenient.

3. Setting Extracellular [Ca2þ] to a Precisely Known Value

When the extracellular [Ca2þ] must be known precisely, medium containing a

well-defined Ca2þ buVer system at a fixed pH must be made. The preparation of

such solutions is detailed in Chapter 9.

C. Divalent Cation Ionophores

1. Properties of Br-A23187 and Ionomycin

Br-A2318717 and ionomycin are ionophores that form complexes with divalent

metal cations; the complexes are lipid-soluble and thus can cross cellular mem-

branes. These two ionophores are commonly used to increase the permeability of

biological membranes to Ca2þ. Understanding the diVerences between the two

makes it possible to make a judicious choice during an experiment.

Ionomycin can lose two acidic hydrogens and, as a dianion, can form an

uncharged 1:1 complex with a divalent metal ion such as Ca2þ or Mg2þ. Br-A23187 can lose a single acidic hydrogen and form an uncharged 2:1 complex

with a divalent metal ion. This diVerence makes ionomycin potentially more

eVective in binding and transporting divalent cations (e.g., two molecules of Br-

A23187 are needed to bind and carry a single Ca2þ whereas only one molecule of

ionomycin is suYcient).

Compared with Br-A23187, ionomycin has somewhat better selectivity for Ca2þ

over Mg2þ; ionomycin prefers Ca2þ by a factor of �2, whereas Br-A23187 shows

essentially no preference for one cation over the other (Liu and Hermann, 1978). In

addition, these ionophores actually do not bind Ca2þ very tightly [e.g., Kd(Ca2þ)�

100 mM for ionomycin; J.P.Y. Kao, unpublished results.18] These factors suggest

that the two ionophores would be ineYcient in mediating Ca2þ transport when

relatively low Ca2þ concentrations are involved (i.e., at [Ca2þ]�Kd, e.g., <1 mM),

because at such low Ca2þ concentrations, only a minute fraction of total

16 Because calcium is a ubiquitous ‘‘contaminant’’ in the environment, nominally calcium-free

solutions still can contain micromolar levels of Ca2þ.17 The parent compound, A23187, is fluorescent and should be avoided in fluorescence work. The

presence of the bromine atom in 4-bromo-A23187 (Br-A23187) eVectively quenches the intrinsic

fluorescence of the ionophore and makes the molecule useful in fluorescence microscopy.18 Determined by absorption spectroscopy at pH 11, at which all ionomycin in solution would be

present as the fully deprotonated dianion.


ionophore is actually engaged in Ca2þ binding and transport. This problem

becomes evident when either ionophore is used in calibrating Ca2þ indicators in

cells (see Section V.B).

The most significant diVerence between the two ionophores lies in the pH

dependence of their ability to transport Ca2þ (Liu and Hermann, 1978). Transport

of Ca2þ by Br-A23187 approaches a maximum at pH 7.5, whereas Ca2þ transport

by ionomycin does not reach a maximum until pH 9.5. The pH at which half-

maximal transport is achieved is �6.4 for Br-A23187 and �8.2 for ionomycin.

Therefore, if one desires to increase transport of extracellular Ca2þ into cells in

acidic media (pH<7.0), Br-A23187 is a much better choice than ionomycin.

2. Using Br-A23187 and Ionomycin

Ionomycin can be obtained as either the free acid or the Ca2þ salt. Br-A23187 is

available as the free acid. All forms are soluble in dry DMSO, which can be used to

prepare stock solutions. Because these ionophores are very hydrophobic, they are

bound avidly by serum proteins. Serum proteins such as BSA, when present in the

medium, greatly reduce the eVectiveness of ionophores and, if possible, should be

left out of the experimental medium when ionophores are to be used. Otherwise,

much higher concentrations of ionophore must be used. Br-A23187 and ionomycin

have been used at concentrations ranging from 10�7 to 10�5 M. In addition to

increasing Ca2þ flux across the plasma membrane, Br-A23187 and ionomycin also

transport Ca2þ out of intracellular calcium stores into the cytosol. Therefore, in

the presence of these ionophores, intracellular calcium stores are rapidly depleted

(Kao et al., 1990).

D. BuVering Changes in Intracellular [Ca2þ]

1. Increasing Intracellular Ca2þ BuVering Capacity by BAPTA Loading

When a change in [Ca2þ]i (a Ca2þ signal) is correlated with a biological process,

one can ascertain whether the Ca2þ signal is essential in the process by blocking the

change in [Ca2þ]i with a calcium chelator. By far the easiest way to introduce extra

Ca2þ buVering capacity into cells is by incubation with BAPTA AM in Pluronic

dispersion. Compared with the AM esters of common Ca2þ indicators, BAPTA

AM has much higher aqueous solubility—15 mM at 25 �C (Kao et al., 1990).

Therefore, BAPTA can be loaded eYciently into cells via the AM ester. BAPTA

AM is loaded into cells in precisely the same way that AM esters of indicators are

loaded. Cells can be loaded with AM esters of BAPTA and an indicator simulta-

neously. Figure 9 illustrates the eVect of intracellular BAPTA loading on normal

changes in [Ca2þ]i. Figure 9A shows the changes in [Ca2þ]i in a REF52 cell loaded

with Fura-2 in response to sequential application of 1 mM bradykinin and 1 mMBr-A23187. Figure 9B shows the responses in a similar cell loaded with Fura-2 and

BAPTA. These results clearly demonstrate that the presence of suYcient BAPTA

practically eliminates the rapid and transient rises in [Ca2þ]i elicited by an agonist.

1.2

1.0

0.8

0.6

[Ca2+

] i (m

M)

0.4

0.2

0.00 10

A B

Bra

dyki

nin

Bra

dyki

nin

Br–

A23

187

Br–

A23

187

20 30

−BAPTA +BAPTA

0 10 20

Time (min)Time (min)

30

Fig. 9 BuVering action of intracellular BAPTA on changes in [Ca2þ]i. (A) Changes in [Ca2þ]i of aREF52 cell treated with 1 mMbradykinin and then 1 mMBr-A23187. (B) Changes in [Ca2þ]i of a REF52

cell, preloaded with BAPTA, in response to the same treatments as in (A). Cells were loaded with 1 mMFura-2 AM in Pluronic dispersion for 85 min at 25 �C. For (B), 20 mMBAPTA AMwas also present in

the incubation medium. Experiments were done in HBSS.


Indeed, even the massive rise resulting from a combination of Ca2þ influx and

discharging of intracellular calcium stores mediated by Br-A23187 is suppressed

substantially by the buVering action of BAPTA.

2. Possible Controls for the Use of BAPTA

Similar to EGTA, BAPTA is a chelator not only for Ca2þ but also for other

multivalent metal cations. Thus, one may wish to ensure that any inhibitory eVectobserved when using BAPTA is caused strictly by the ability of BAPTA to buVerCa2þ, and not because it is scavenging other biochemically important metal ions

such as Zn2þ. The reagent used to control for heavy metal scavenging by BAPTA is

TPEN (N,N,N0,N0-tetrakis(2-pyridylmethyl)ethylenediamine) (Fig. 10), a mem-

brane-permeant metal ion chelator that shows a marked preference for binding

heavy metal cations over Ca2þ (Anderegg et al., 1977). Whereas the Kd(Ca2þ) of

TPEN is 40 mM (Arslan et al., 1985), Kd(Zn2þ) is 2.6�10�16 M (Anderegg and

Wenk, 1967). This enormous selectivity for binding heavy metal ions over Ca2þ

enables TPEN to scavenge heavy metal ions very eVectively, even in the presence of

N N

NN

N

N N

O O

BAPTATPEN Half-BAPTA

OH3C

CO2−

CO2− CO2

−−O2C−O2C −O2C

N N

Fig. 10 Structures of TPEN, BAPTA, and half-BAPTA.


millimolar levels of Ca2þ. That TPEN is membrane-permeant means it can be

applied without using any special procedures. Dry DMSO can be used to prepare

stock solutions of TPEN. Typically, TPEN is used in aqueous medium at a

concentration of 10�6–10�5 M.

BAPTA loading through the AM ester is very eYcient; high concentrations of

BAPTA may be accumulated intracellularly. Thus, ascertaining that observed

inhibitory eVects are not the result of cytotoxicity arising from the presence of

high concentrations of a foreign organic anion may be important. In this case, the

control reagent is N-(o-methoxyphenyl)iminodiacetic acid,19 sometimes referred

to as ‘‘half-BAPTA.’’ As can be seen from Fig. 10, half-BAPTA is essentially

chemically identical to BAPTA except that the molecule is only half of BAPTA.

Because the full tetracarboxylate structure of BAPTA is crucial for Ca2þ binding,

half-BAPTA, lacking such a structure, shows only very weak aYnity for Ca2þ

(Kd�3 mM; J.P.Y. Kao, unpublished results). Half-BAPTA is thus expected to

mimic BAPTA in all chemical respects except for the ability to buVer Ca2þ at

physiological concentrations. The AM ester of half-BAPTA is sporadically

available from commercial vendors. Cell loading via the AM ester can be done

as described previously for other AM esters.20

V. Conversion of Indicator Fluorescence Signal intoValues of [Ca2þ]

Although raw fluorescence signals from intracellularly trapped Ca2þ indicators

can be informative in a qualitative way, one still must perform some calibration

before even semiquantitative estimates of [Ca2þ]i can be made. Basic principles of,

19 A trivial name is anisidine-N,N-diacetic acid.20 Because it is processed by esterases to generate only the Ca2þ-insensitive half-BAPTA and yet it is

processed intracellular in the same way that all AM esters are, half-BAPTA AM can also be used as a

control for possible artifacts from AM ester hydrolysis.


as well as experimental procedures for, calibration for ratiometric and nonratio-

metric indicators are discussed in this section.

A. Calibrating a Nonratiometric Fluorescent Indicator

For a nonratiometric indicator that increases fluorescence emission on binding

Ca2þ, the free Ca2þ concentration is given by

Ca2þ� � ¼ Kd

F � Fmin

Fmax � F

� �ð1Þ

where Fmin is the indicator fluorescence intensity at zero [Ca2þ] (when all

indicator molecules in the sample are Ca2þ-free), Fmax is the indicator fluores-

cence at saturatingly high [Ca2þ] (when all indicator molecules are present as

the Ca2þ-bound form), and F is the measured fluorescence intensity for which

we wish to find a corresponding value of [Ca2þ]. To arrive at a correspondence

between measured F and [Ca2þ]i, Kd, Fmin, and Fmax all must be known.

Whereas Kd usually is predetermined in vitro, Fmin and Fmax must be obtained

in situ. The most straightforward approach would be to try to equilibrate the

indicator-loaded cell with solutions that contain ‘‘zero’’ [Ca2þ] and then high

[Ca2þ]. In practice, however, deficiencies inherent in a nonratiometric indicator

make this approach unattractive. Interpretation of intensity changes is con-

founded by dye leakage, which causes the total indicator fluorescence from the

cell (and therefore Fmin and Fmax) to decrease with time. This basic flaw of

nonratiometric indicators means that obtaining good quantitative estimates of

[Ca2þ]i in cells in which dye leakage or extrusion occurs at significant rates

would be diYcult. In such cases, a laborious calibration would not be justified.

An alternative semiquantitative calibration procedure developed for Fluo-3 is

discussed next.

The calibration procedure for Fluo-3 depends on the fact that, in vitro,

FMn¼0.2Fmax, where FMn is the fluorescence intensity when Fluo-3 is saturated

completely with Mn2þ (Kao et al., 1989; Minta et al., 1989). That Fmax¼100Fmin is

also known from in vitro measurements. Because both Fmax and Fmin can be

expressed in terms of FMn, the only parameter that must be determined experimen-

tally is FMn. In situ calibration then consists of:

1. applying micromolar levels of ionomycin or Br-A23187 to increase perme-

ability of the cell to divalent metal ions;

2. adding suYcient MnCl2 (typically twice the concentration of Ca2þ in solu-

tion)21 to ensure saturation of intracellular Fluo-3; and

21 In Step 2, it is best if the medium contains no carbonate, bicarbonate, or phosphates, which can

form insoluble precipitates with Mn2þ and thus reduce the concentration of free Mn2þ. Moreover, the

light scattering by the particulate precipitates can add considerable noise to the fluorescence signal.


3. permeabilizing the cell with digitonin to release Fluo-3 to permit estimation

of fluorescence background.

Because the fluorescence intensity measured after Step 3 is just the background

signal (including cellular autofluorescence), whereas the intensity after Step 2 is

(FMnþbackground), one can obtain FMn by subtraction. The calibration proce-

dure described here is based on the following assumptions: (1) indicator fluores-

cence intensity is not diminishing rapidly as a result of leakage; (2) the fluorescence

properties (Fmin, Fmax, and FMn) of the indicator are known from in vitromeasure-

ments and are the same in cells as in vitro; and (3) the Kd of the indicator is also the

same in cells as in vitro.

Quin2 is an example of an indicator the fluorescence of which is quenched

completely by heavy metal ions. For calibration of such an indicator, see the

review on Quin2 by Tsien and Pozzan (1989).

B. Calibrating a Ratiometric Fluorescent Indicator

A dual-wavelength ratiometric indicator allows excitation spectral intensity or

emission spectral intensity of the indicator to be monitored at two diVerentwavelengths. If F1 is the fluorescence intensity at wavelength l1, F2 is the fluores-

cence intensity at wavelength l2, and R¼F1/F2, then the free Ca2þ concentration

can be shown to be (Grynkiewicz et al., 1985)

½Ca2þi ¼ Kd

R� Rmin

Rmax � R

� �sf ;2sb;2

� �ð2Þ

where Rmin is the limiting value of the ratio R when all the indicator is in the Ca2þ-free form and Rmax is the limiting value of R when the indicator is saturated with

Ca2þ.22 Experimentally, the factor sf,2/sb,2 is simply the ratio of the measured

fluorescence intensity when all the indicator is Ca2þ-free to the intensity measured

when all the indicator is Ca2þ-bound, with both intensity measurements taken at

l2. On the right side of Eq. (2), with the exception of Kd, which is an intrinsic

property of the indicator, all other terms are ratios of intensities; in forming these

ratios, problems associated with cell shape and dye concentration changes cancel.

Using Eq. (2) to calculate [Ca2þ] requires that Kd be known and that Rmin, Rmax,

and sf,2/sb,2 be determined experimentally. A typical calibration entails

1. increasing Ca2þ permeability of the cell with ionomycin or Br-A23187 in the

presence of ‘‘zero’’ extracellular Ca2þ (EGTA or BAPTA in nominally Ca2þ-free medium; Section IV.B.2), so all intracellular Ca2þ could be depleted;

22 Usually, l1 and l2 are chosen so that intensity measured at l1 consists mostly of fluorescence

emitted by the Ca2þ-bound form of the indicator, whereas intensity at l2 consists mostly of fluorescence

from the Ca2þ-free form. Choosing the wavelength pair in this way increases the diVerence betweenRmin

and Rmax, making it possible to map [Ca2þ] onto a wider range of R values.


2. increasing extracellular [Ca2þ] in the presence of Ca2þ ionophore so that

Ca2þ could enter the cell to saturate intracellular indicator;

3. permeabilizing the cell with digitonin (at concentrations prescribed in

Section III.B.2) to release cytosolic dye so the background signal may be

measured.23

Although the procedure seems straightforward, a few empirical findings are

helpful in performing a successful calibration:

1. Many cell types do not tolerate severe calcium deprivation well. During Step

1, these cells often become fragile or leaky or, in the case of adherent cells, detach

from the substrate. In many cases, however, one can compensate for the total

absence of Ca2þ by supplementation with elevated concentrations of Mg2þ. Thus,raising the extracellular [Mg2þ] to 5–20 mM can help maintain cell integrity during

a long calibration. Although Mg2þ should bind Fura-2 to a limited extent and

slightly alter its fluorescence spectrum (Grynkiewicz et al., 1985), in practiceRmin is

not aVected significantly by Mg2þ supplementation.

2. Because ionomycin and Br-A23187 become very ineYcient at Ca2þ transport

when intra- and extracellular free Ca2þ concentrations are below micromolar

levels, depleting the cell of Ca2þ entirely is quite diYcult. Therefore, one often

must wait a long time for trueRmin to be reached in Step 1. In the example shown in

Fig. 11, the interval between ionomycin addition and attainment of Rmin was in

excess of 90 min.

3. Step 2 could be performed in two ways. One could add, in combination with

fresh ionophore if desired, suYcient Ca2þ to bind to all the EGTA or BAPTA that

was introduced in Step 1 and still have a large excess of free extracellular Ca2þ, aswell as suYcient TRIS base to counteract any acidification arising from the Ca2þ-EGTA binding reaction. Alternatively, one could replace the medium from Step 1

with nominally Ca2þ-free medium and then add a large excess of Ca2þ in combi-

nation with a fresh dose of ionophore. The aim is to initiate massive Ca2þ influx

into the cell at a rate that overcomes any Ca2þ extrusion mechanism that the cell

may mobilize. In practice, concentrations of ionophore ranging from 10�6 to

10�5 M and external [Ca2þ] in the range of one to several tens of millimolar can

be used.

23 At controlled concentrations, digitonin releases primarily cytosolic dye whereas compartmenta-

lized dye remains with the permeabilized cell and would be subtracted out as background. The

assumption is that intraorganellar [Ca2þ] is significantly higher (> mM) than cytosolic [Ca2þ], socompartmentalized indicator would be essentially completely Ca2þ-bound and thus contribute a con-

stant background to the measured 340- and 380-nm fluorescence signals from the cell. This assumption

would fail if significant amounts of dye are compartmentalized into organelles that do not have high

luminal [Ca2þ]. An alternative approach to obtaining a background reading is to add MnCl2 (at a

concentration equal to or greater than the Ca2þ concentration in the medium) at the same time as the

digitonin so that compartmentalized dye can also be quenched as ionophores transport Mn2þ into the

organelles. Using such an approach assumes that cellular autofluorescence is the true background and

ignores the contribution of compartmentalized dye to the background.


4. A stable baseline is obtained quickly only if the dye released from cells by

digitonin permeabilization is swept rapidly away from the region directly above the

microscope objective. Otherwise, fluorescence from the released dye still will be

captured by the objective and, thus, contribute to the measured background. Once

swept away and diluted into the bulk medium, the released dye contributes

negligibly to the background.

0

A45 6

F�340

F�380

F�f,380

BG340

BG380

20

Flu

ores

cenc

e in

tens

ity (

a.u.

)

40 60

Time (min)

80 100 120

B

1 3

R=

F34

0/F

380

20

10

0

20 40 60 80 100

6

Rmax

Rmin

0 20 40

0.8

60

Time (min)

80 100 120

0.7

0.6

0.5

1 2 3

5

2

F�b,380

Fig.11 (Continued)

2.0C

1 2 3

1.0

0.00 10 20

Time (min)

30 40

[Ca2+

] i (m

M)

Fig. 11 Procedure for in situ calibration of intracellular Fura-2. (A) Fluorescence intensity traces

acquired at 340- and 380-nm excitation. Time marker arrow correspond to (1) addition of 50 nM

vasopressin; (2) exchange into Ca2þ-free phosphate-buVered saline (PBS) containing 10 mM MgCl2,

2 mM EGTA, pH 7.4; (3) addition of 10 mM ionomycin; (4) exchange into nominally Ca2þ-free saline,pH 7.4; (5) addition of 10 mM ionomycinþ20 mM CaCl2; and (6) 20 mM digitonin. Dotted lines mark

fluorescence levels corresponding to various parameters discussed in Section V.B. (a.u. = arbitrary

units). (B) F340/F380 ratio trace derived from the data in (A). Dotted lines markRmin andRmax (0.566 and

16.6, respectively, in this experiment). The parameter sf,2/sb,2 is 10.7. Inset. The portion of the trace from

20 to 118 min at higher resolution on the vertical scale to reveal the gradualness with which Rmin is

approached. (C) [Ca2þ]i trace derived from the ratio trace by using Eq. (2) in Section V.B. Only the first

40 min of the experiment are shown. This REF52 cell was incubated with 1 mMFura-2 AM in Pluronic

dispersion in HBSS for 90 min at 25 �C before being transferred to fresh HBSS for measurement.


Elevation of [Ca2þ]i by ionophore can lead to rapid cell lysis and loss of indica-

tor, sometimes before Rmax can be determined confidently. Almost paradoxically,

raising the extracellular [Ca2þ] to 10–30 mM (rather than just a few mM) in this

procedure appears, in some cases, to have a protective eVect on cell structure so

lysis is deferred and Rmax can be reached. If high extracellular [Ca2þ] is used, themedium should be free of phosphate salts, bicarbonate/carbonate, and even sul-

fate, since these ions can form precipitates with Ca2þ.Typical data from an experiment performed on a REF52 cell loaded with Fura-

2 are shown in Fig. 11. Shown in Fig. 11A are the two raw data traces, F 0340 and

F 0380, collected when the cell is excited alternately with 340-nm and 380-nm light.

The fluorescence signals measured after digitonin permeabilization are the back-

ground intensities, BG340 and BG380, that must be subtracted from the respective

traces to yield the true F340 and F380 (i.e., F340¼F 0340�BG340 and

F380¼F 0380�BG380). The ratio trace is simply a point-by-point division,


R¼F340/F380 and is shown in Fig. 11B. Rmin is the limiting value of R that is

reached during Ca2þ deprivation, whereas Rmax is the limiting value of R reached

after treatment with ionophore at high [Ca2þ].24 The factor sf,2/sb,2 is essentially

(F 0f,380�BG380)/(F 0

b,380�BG380). Using these experimentally derived parameters

and a predetermined Kd (224 nM; Grynkiewicz et al., 1985) in Eq. (2), one can

convert the F340/F380 ratio trace into a plot of [Ca2þ]i as a function of time

(Fig. 11C).

This procedure has the advantage that all spectroscopically derived parameters,

namely Rmin, Rmax, and sf,2/sb,2, that are especially sensitive to environmental

changes are determined in situ with the indicator residing in the intracellular

environment. Only the equilibrium dissociation constant is determined in vitro.

Rmin determined by Ca2þ deprivation is assumed to be the true value. In view of the

ineVectiveness of currently available ionophores at low [Ca2þ], one would be

justified in concluding that true Rmin would be diYcult to reach25 and that Rmin

is easy to overestimate. An overestimate of Rmin results in underestimation of

[Ca2þ].Finally, it is worthwhile to examine the eVects of errors in Rmin, Rmax, and

sf,2/sb,2 on the derived value of [Ca2þ]. For simplicity, one assumes that errors in

the three parameters are independent. Because sf,2/sb,2 is related linearly to [Ca2þ](see Eq. (2)), a percentage error in sf,2/sb,2 translates into the same percentage error

in [Ca2þ]. Inspection of Eq. (2) reveals that errors in Rmin should aVect primarily

low values of [Ca2þ] (corresponding to R values near Rmin). Error in Rmax, on the

other hand, aVects the way in which all the R values are scaled and, therefore,

should influence all derived values of [Ca2þ]. These expectations are borne out bycalculation.26

24 From Fig. 11B, the ratio values near Rmax are seen to oscillate significantly because, at saturating

[Ca2þ], the fluorescence of the indicator excited at 380 nm (Fb,380 ¼F 0b,380 � BG380) is very weak and

cannot be determined with high precision. In forming the ratio, because Fb,380 is a small number and

occurs in the denominator, noise fluctuations in Fb,380 become magnified into large-amplitude fluctua-

tions in Rmax. Therefore, one must average a large number of points to obtain a reliable estimate of

Rmax. Alternatively, the fluorescence intensity data (both F340 and F380) can be smoothed first before a

ratio is formed.25 Rather than estimating Rmin directly from the lowest values attained in the ratio trace, curve-

fitting the portion of the ratio trace that represents the slow descent towards Rmin is also a reasonable

approach. As expected, Rmin obtained by exponential curve-fitting is somewhat lower than that

estimated directly from the ratio trace.26 When one uses parameters similar to those for Fura-2 inREF52 cells as determined onour instrument

(Rmin¼ 0.5,Rmax¼ 15, and sf,2/sb,2¼ 12), a 10% overestimation ofRmin leads to�19% underestimation of

[Ca2þ] at 50 nM, �10% at 100 nM, and �2% at 500 nM. A 10% overestimation of Rmax leads to

underestimation of [Ca2þ] by �9.5% at 50 nM, �10.9% at 500 nM, and �12.5% at 1 mM. A 10% under-

estimation ofRmax results in overestimation of [Ca2þ] by�11.8% at 50 nM,�14% at 500 nM, and�16.5%

at 1 mM.

2.5

2.0

1.5

1.0

0.5

0 10

Time (s)

ΔF

F

F/F

0 o

r Δ

F/F

0

F0

F0

20 30

0.0

Fig. 12 Two conventions, F/F0 and DF/F0, for reporting fluorescence changes relative to baseline

fluorescence intensity. Note: DF/F0¼F/F0�1.


VI. Reporting Indicator Fluorescence Intensity Changeswithout Calibration

A. Reporting Relative Changes in Fluorescence: F/F0 and DF/F0

With the widespread use of nonratiometric indicators, which are diYcult to

calibrate, it has become common to report not [Ca2þ], but rather indicator fluores-cence changes. The convention is to report either the fluorescence intensity relative

to baseline intensity (F/F0), or the change in fluorescence intensity relative to

baseline intensity (DF/F0¼ (F�F0)/F0). Figure 12 illustrates these two conventions.

From the above definitions and from the graphs in Fig. 12, it is apparent that the

two reporting conventions are simply related: DF/F0¼F/F0�1. It is important to

stress that in order for these relative measurements to be meaningful, F and F0

should be intensities that have been background-subtracted.

B. Caveat in Interpreting Relative Fluorescence Changes: Indicator Fluorescence is Not aLinear Function of [Ca2þ]

Because a nonratiometric indicator becomes brighter when it binds Ca2þ, anincrease in indicator fluorescence implies an increase in [Ca2þ]. Once fluorescence

intensity data have been converted into relative changes, however, there is perhaps


a natural tendency to regard the relative change in intensity as reflecting an

equivalent relative change in [Ca2þ]. For example, a doubling of intensity relative

to baseline (F/F0¼2 or DF/F0¼1) is often used to infer a doubling of [Ca2þ]. Suchan inference should never be made because it is always incorrect. A quantitative

analysis is presented below.

As shown in Appendix 2, the total fluorescence, FT, emitted by a solution of

Ca2þ indicator is governed by the expression

FT / QCaIneCaInfCaIn þQIneIn 1� fCaInð Þ ð3Þwhere QCaIn and QIn are the fluorescence quantum eYciencies of the Ca2þ-boundand Ca2þ-free forms of the indicator, respectively, eCaIn and eIn are the extinctioncoeYcients of the two forms of the indicator at the excitation wavelength, and fCaInis the fraction of the indicator that is in the Ca2þ-bound form. Knowing that

fCaIn¼ [Ca2þ]/([Ca2þ]þKd), we can rewrite the expression to show its dependence

on [Ca2þ] more explicitly:

FT / QIneIn þ QCaIneCaIn �QIneInð Þ Ca2þ� �

Ca2þ� �þKd

ð4Þ

The only variable in the expression is [Ca2þ]; all other parameters, being intrinsic

characteristics of a particular indicator, are constants. The above expression shows

that whereas [Ca2þ] can range from 0 to any arbitrary positive value, the total

fluorescence, FT is bounded. This behavior is shown in Fig. 13.When [Ca2þ]¼0, all

of the indicator is Ca2þ-free, and the fluorescence has a minimum value that

depends on the intrinsic brightness (QIneIn) of the Ca2þ-free form of the indicator.

At saturating [Ca2þ] ([Ca2þ]Kd), all of the indicator is Ca2þ-bound, and the

fluorescence has a maximum value that depends on the intrinsic brightness

(QCaIneCaIn) of the Ca2þ-bound form of the indicator. Once the indicator molecules

are saturated, further increasing [Ca2þ] brings no increase in fluorescence. There-

fore, as can be seen from Eq. (4) and Fig. 13, fluorescence intensity is a nonlinear

function of [Ca2þ]. This nonlinearity is the reason that a relative change in indica-

tor fluorescence does not imply an equal relative change in [Ca2þ]. Figure 13 showsthat the discrepancy depends on the extent to which the indicator is already bound

to Ca2þ: Starting from a relatively low [Ca2þ], increasing [Ca2þ] by an increment,

DCa1, results in a fluorescence increase, DF1. From the now-higher [Ca2þ], a

further identical increment of DCa2 (¼DCa1) brings a much smaller fluorescence

increase, DF2.

The error in using relative fluorescence changes to infer relative [Ca2þ] changescan be analyzed quantitatively for a specific example. Fluo-4 is a nonratiometric

indicator that is commonly used with 488-nm excitation. The extinction coeYcient

of Fluo-4 changes only by a few percent upon binding Ca2þ

(eIn� eCaIn¼77,000 M�1 cm�1 at 488 nm); the Ca2þ-bound form is at least 100

times more fluorescent than the Ca2þ-free form (QCaIn¼0.14, QIn�0.0014); and

Kd¼345 nM. The quantitative relationship between [Ca2þ] and fluorescence can

Upper bound ∝ QCaIn εCaIn

ΔF2

ΔCa 1

ΔCa 2

ΔF1

[Ca2+]

00

FT

Lower bound ∝ QIn εIn

Fig. 13 Indicator fluorescence intensity is a nonlinear function of [Ca2þ]. At [Ca2þ]¼0, all indica-

tor molecules in solution are in the Ca2þ-free form, and indicator fluorescence is at the lower

bound. Whereas [Ca2þ] can range from 0 to any arbitrarily large value, indicator fluorescence

cannot exceed an upper bound, which is reached when all indicator molecules in solution are in

the Ca2þ-bound form. The relationship between fluorescence intensity and [Ca2þ] is hyperbolic.

The consequence is that successive equal increments in [Ca2þ] do not result in equal increments of

fluorescence intensity (compare the fluorescence increments DF1 and DF2 resulting from two equal

increments in [Ca2þ]).


be calculated by using these parameters in Eq. (4). Figure 14 shows the relative

change in Fluo-4 fluorescence for diVerent increments in [Ca2þ], up to a 10-fold

change ([Ca2þ]/[Ca2þ]0¼10). Because resting [Ca2þ]i is typically in the range 50–

100 nM, the starting [Ca2þ] was assumed to be [Ca2þ]0¼75 nM for the calculation.

Figure 14 shows clearly that the relative change in fluorescence is never a good

measure of the true relative change in [Ca2þ]. F/F0 significantly underestimates

[Ca2þ]/[Ca2þ]0, and the error increases severely as the change in [Ca2þ] becomes

larger.

VII. Measuring [Ca2þ] in Mitochondria

As mentioned in Section II, when cells are incubated with the AM ester of

Rhod-2, the indicator preferentially loads into mitochondria. The structures of

two fluorescent dyes, TMRM and TMRE, which also accumulate preferentially

into mitochondria, and Rhod-2 AM are shown in Fig. 15A. The positively charged

structure in these molecules that enables preferential loading into mitochondria

is highlighted with thick lines in Fig. 15A. Figure 15B shows that, rather than

A

B

Fig. 15 (A) Structures of Rhod-2 AM (bromide salt), as well as TMRM and TMRE (percholorate

salts), two dyes that accumulate preferentially into mitochondria. Note that in each case, the

dye molecule bears a permanent positive charge. (B) A series of related resonance structures

showing that the positive charge can be located on diVerent atoms in the molecule; that is, the charge

is delocalized.

2

2

1

1

03

3

4

4

5

5

6

6

7

7

8

8

9

9

10

10

2

1

0

3

4

5

6

7

8

9

10

[Ca2+]/[Ca2+]0

FF0

[Ca2+]0= 75 nM

[Ca2+]

[Ca2+]0

Fig. 14 Specific example illustrating that a relative change in fluorescence (F/F0) of the indicator,

Fluo-4, does not accurately reflect the true relative change in [Ca2þ] ([Ca2þ]/[Ca2þ]0).


being isolated on a single atom, the positive charge can reside on many

diVerent atoms in the structure—that is, the positive charge is delocalized over

the entire highlighted structure. A hydrophobic organic ion whose charge

is delocalized can pass through lipid membranes. Mitochondria maintain a re-

markably negative membrane potential—the mitochondrial lumen is typically at

�150 to �200 mV relative to the cytosol. Therefore, Rhod-2 AM, which is a

hydrophobic organic cation whose positive charge is delocalized, can permeate

through the plasma membrane and the mitochondrial membranes and preferen-

tially partition into the negative lumen of mitochondria. In mitochondria, cleavage

of AM ester groups by esterases liberates the Ca2þ-sensitive form of Rhod-2 -

(bearing multiple nondelocalized negative charges), which is not membrane-per-

meant and thus trapped in the mitochondrial lumen. Therefore, Rhod-2 can be

used to monitor intramitochondrial Ca2þ signals (Babcock et al., 1997; Tsien and

Bacskai, 1995).

A. Estimating the Fraction of Intracellular Rhod-2 Indicator that Resides in Mitochondria

While Rhod-2 can be preferentially loaded into mitochondria, the discrimina-

tion against loading into other subcellular compartments is imperfect. Some

intracellular Rhod-2 is expected to reside in the cytosol and nonmitochondrial

organelles. A simple procedure based on diVerential permeabilization of cellular

membranes can be used to estimate the fraction of intracellular Rhod-2 that

actually resides in mitochondria. The procedure is a modification of the one

described in Section III.B.2. The procedure consists of monitoring total Rhod-

2 fluorescence from a cell or a group of cells bathed in low-Ca2þ medium and

treating the cells sequentially with (1) a Ca2þ ionophore (ionomycin or Br-

A23187), (2) the mild detergent digitonin to permeabilize the plasma membrane,

and (3) a strong detergent, for example, Triton X-100 or sodium dodecyl sulfate

(SDS), to permeabilize all membranes. Figure 16 shows the procedure being

applied to a vagal sensory neuron that had been incubated with 1 mM Rhod-

2 AM for 1 h at room temperature. Application of ionomycin abolishes significant

diVerences in [Ca2þ] between diVerent subcellular compartments. This ensures that

Rhod-2 in all compartments is at comparable levels of Ca2þ-binding, and thus

would contribute fluorescence intensity in proportion to their actual content in

each compartment. Once the fluorescence reaches a steady baseline after ionomy-

cin treatment, digitonin permeabilization of the plasma membrane allows cytosolic

Rhod-2 to escape,27 giving a decrement in total fluorescence (labeled ‘‘C’’ in

Fig. 16). Subsequent permeabilization of all cellular membranes by SDS allows

27 Digitonin treatment leads to release of Rhod-2 from the nucleus as well. The nuclear pores have a

size exclusion limit of 35–40 kDa; molecules with molecular mass less than the exclusion limit can freely

exchange between the nucleoplasm and cytosol. Therefore, with respect to low-molecular-mass solutes

such as simple ions (e.g., Ca2þ, Naþ, Cl�) and small organic molecules (e.g., glucose, ATP, fluorescent

indicators), the nucleo-cytoplasm functions as a single ‘‘cytosolic’’ compartment.

Iono

60 s

C

M + NM

Rho

d-2

fluor

esce

nce

(arb

itrar

y un

its)

DigitSDS

0-Ca/BAPTA

Fig. 16 A diVerential permeabilization experiment for estimating subcellular fractions of Rhod-2

indicator. A rabbit vagal sensory neuron was incubated with 1 mM Rhod-2 AM for 1 h at 23 �C and

then bathed in nominally Ca2þ-free physiological saline to which 2 mM Na4BAPTA was added (0-Ca/

BAPTA). The total Rhod-2 fluorescence from the cell was monitored. Ionomycin (Iono, 2 mM) was

applied to dissipate Ca2þ gradients between subcellular compartments. Digitonin (Digit, 20 mM)

permeabilized the plasma membrane selectively to release cytosolic Rhod-2 (intensity decrement

marked ‘‘C’’). Sodium dodecyl sulfate (SDS; 0.25%, w/v) permeabilized all cellular membranes to

release Rhod-2 from organellar compartments (decrement marked ‘‘MþNM’’). The durations of

reagent applications are indicated by the bars at the bottom.


Rhod-2 to escape from mitochondria as well as nonmitochondrial organelles; this

causes a further decrement in fluorescence (labeled ‘‘MþNM’’ in Fig. 16). For

four neurons tested, the ratio of the noncytosolic fraction, MþNM, to the cyto-

solic fraction, C, was

MþNM

C¼ 5:36

This provides one required algebraic condition; a second condition is

MþNM þ C ¼ 1

that is, intracellular Rhod-2 must be in the mitochondria, in nonmitochondrial

organelles, or in the cytosol. Since there are three variables, a third algebraic

condition is required, and this can be obtained by performing the permeabilization

experiment on cells whose incubation with Rhod-2 AM had been done in the

presence of a protonophore (e.g., CCCP, FCCP, 2,4-dinitrophenol),28 which

28 It is advisable to use oligomycin (e.g., 10 mM), a blocker of the mitochondrial F1F0-ATP synthase,

in conjunction with the protonophore. In discharging the mitochondrial membrane potential, the

protonophore eliminates the Hþ electrochemical gradient that is used by the ATP synthase to generate

ATP. This causes the ATP synthase to run in reverse—as an ATPase—and rapidly deplete cellular ATP.

Oligomycin, by blocking ATPase action, helps to preserve the cellular ATP pool.


abolishes the mitochondrial membrane potential and thus eliminates the driving

force for preferential partition of Rhod-2 AM into mitochondria. In these cells

with depolarized mitochondria, the fluorescence decrement caused by digitonin

still represents loss of cytosolic Rhod-2 (C), but the decrement caused by SDS may

be attributed to Rhod-2 loss from nonmitochondrial organelles (NM). In four

neurons loaded in the presence of 5 mM CCCP,

NM

C¼ 0:623

which provides the last algebraic condition required to solve for the three

unknowns, C, NM, and M. Using all the three conditions together yields

C¼0.156, NM¼0.097, and M¼0.747—about 75% of intracellular Rhod-2 reside

in mitochondria, with �15% in the cytosol and �10% in other organelles.29

B. Minimizing Rhod-2 Loading in the Cytosol

Having a few percent of Rhod-2 residing in nonmitochondrial organelles is likely

to be unimportant. The luminal [Ca2þ] in these organelles (e.g., lysosomes, ER, etc.)

is not expected to changemarkedly. Therefore, the Rhod-2 fluorescence signal from

these organelles should not change significantly during an experiment and thus can

be considered operationally to be part of the background fluorescence. In contrast,

Rhod-2 in the cytosol, although much less than that in the mitochondria, may

contribute a contaminating cytosolic Ca2þ signal when one attempts to measure

mitochondrial [Ca2þ]. The cytosolic fraction can beminimized by taking advantage

of the ubiquitous cellular transporters that extrude organic anions from the cytosol

into the extracellular space (this is the process discussed in Section III.C). After they

have been incubatedwithRhod-2AMat room temperature, cells can be transferred

into medium containing no AM ester and incubated for 20–30 min at 37 �C to

accelerate extrusion of Rhod-2 from the cytosol. Thereafter, the Ca2þ-responsiveRhod-2 signal should be predominantly mitochondrial.

29 This likely underestimates mitochondrial loading, because we assumed that in the presence of

protonophore, Rhod-2 loading into mitochondria was negligible. Without the benefit of the mitochon-

drial membrane potential, Rhod-2 AM can still diVuse into the depolarized mitochondria and be

processed by esterases therein. Therefore, the decrement caused by SDS should contain both contribu-

tions from nonmitochondrial organelles and from passively loaded, depolarized mitochondria. If we

assume that on a per-volume basis, the cytosol and mitochondria have comparable capacity to process

AM esters, then the Rhod-2 content in depolarizedmitochondria relative to that in the cytosol should reflect

the ratio of the mitochondrial and cytosolic volumes. Depending on cell type, mitochondria-to-cytosol

volume ratio can range from �1/5 to �1/3. Using the lower figure changes the third algebraic condition to

(C/5þNM)/C¼ 0.623, leading to themodified estimates:C¼ 0.156,NM¼ 0.066, and M¼ 0.778.Using the

higher figure of 1/3 leads to the condition, (C/3þNM)/C¼ 0.623, which yields C¼ 0.156, NM¼ 0.045, and

M ¼ 0.798. Comparing these numbers with those obtained in the main text shows that this additional

correction only changed the estimate by a few percent.

Caf

60 s

0-Ca

MitoCyto

ΔF/F

0=

0.5

Fig. 17 Simultaneous measurement of Ca2þ signals in the cytosol and in mitochondria. A rabbit vagal

sensory neuron was incubated with 1 mM each Rhod-2 AM and Fluo-3 AM for 1 h at 23 �C and then

superfused with nominally Ca2þ-free physiological saline (0-Ca). Rhod-2 and Fluo-3 fluorescence,

excited at 543 and 488 nm, respectively, were imaged simultaneously by laser-scanning confocal micros-

copy. Data are represented as fluorescence change relative to baseline (DF/F0). A 5-s pulse of caVeine

(Caf, 10 mM) was delivered by superfusion. The durations of reagent applications are indicated by the

bars at the bottom.


C. Monitoring Cytosolic and Mitochondrial [Ca2þ] Simultaneously

Because indicators whose AM esters are uncharged load primarily into the

cytosol, while Rhod-2 preferentially loads into mitochondria, one can monitor

Ca2þ signals in the two compartments simultaneously. For measuring cytosolic

[Ca2þ], one should select an indicator whose excitation and emission wavelengths

do not interfere with Rhod-2 measurement. Since Rhod-2 is a rhodamine-based

indicator, a fluorescein-based indicator would be suitable for the cytosolic mea-

surement (e.g., members of the Fluo family of indicators). Figure 17 shows an

experiment where the cytosolic and mitochondrial Ca2þ signals are monitored

simultaneously in a vagal sensory neuron being stimulated with a brief pulse of

caVeine (an agonist that activates ryanodine receptor Ca2þ channels to release

Ca2þ from Ca2þ stores in the ER). The cytosolic and mitochondrial Ca2þ

transients have very diVerent decay kinetics: the time for the Ca2þ signal to

decay by 80% was t80%¼7.7 s in the cytosol and t80%¼65.1 s in mitochondria.

VIII. Concluding Remarks

Fluorescent Ca2þ indicators have contributed enormously to our understanding

of intracellular calcium regulation. For those who are beginning to use these

indicators, the technical details can seem bewildering. This compendium of


common techniques has aimed to set in order the body of practical empirical

knowledge that underlies successful measurements of [Ca2þ] through the use of

fluorescent indicators.

Appendix 1. The fraction of a polybasic acid that exists in aparticular state of protonation

The native forms of the Ca2þ buVers and indicators discussed in this chapter are

polybasic acids; that is, they are species with multiple dissociable protons (e.g.,

H4EGTA). Almost invariably, the fully deprotonated species is the one that

actually binds Ca2þ with high aYnity; therefore it is useful to be able to estimate

the fraction of the indicator or buVer in solution that actually exists in the fully

deprotonated form. In general, deprotonation of a polybasic acid is characterized

by a sequence of dissociation equilibria. For the specific example of a tetrabasic

acid, H4A, the sequence of stepwise dissociation reactions and the corresponding

equilibrium constants are:

H4A > Hþ þH3A� K1 ¼ Hþ½ H3A

�½ H4A½

H3A� > Hþ þH2A

2� K2 ¼Hþ½ H2A

2�� H3A

�½

H2A2� > Hþ þHA3� K3 ¼

Hþ½ HA3��

H2A2��

HA3� > Hþ þA4� K4 ¼Hþ½ A4��

HA3��

If we define C0 to be the total concentration of the acid, irrespective of the state

of protonation, then the fraction that is present as the fully deprotonated,

tetraanionic form, A4�, is

a4 ¼A4�� C0

ðA1:1Þ

To derive a useful expression for a4, we first recognize that the total concentration,C0, encompasses the concentrations of all the possible protonated forms:

C0 ¼ H4A½ þ H3A�½ þ H2A

2�� þ HA3�� þ A4�� ðA1:2ÞBecause all the concentrations are related to each other through the dissociation

equilibrium constants,K1 throughK4, the concentration of any particular protonated

species can be written in terms of the concentration of any other species. In the

present case, it is convenient to express all the concentrations in terms of [A4�]:


HA3�� ¼ Hþ½ K4

A4�� ðA1:3Þ

H2A2�� ¼ Hþ½

K3

HA3�� ¼ Hþ½ 2K3K4

A4�� ðA1:4Þ

H3A�½ ¼ Hþ½

K2

H2A2�� ¼ Hþ½ 3

K2K3K4

A4�� ðA1:5Þ

H4A½ ¼ Hþ½ K1

H3A�½ ¼ Hþ½ 4

K1K2K3K4

A4�� ðA1:6Þ

By using these four relations, the expression for the total concentration can be

written in terms of [A4�]:

C0 ¼ Hþ½ 4K1K2K3K4

A4�� þ Hþ½ 3K2K3K4

A4�� þ Hþ½ 2K3K4

A4�� þ Hþ½ K4

A4�� þ A4��

ðA1:7ÞDividing through by [A4�] leads to

C0

A4�� ¼ 1

a4¼ Hþ½ 4

K1K2K3K4

þ Hþ½ 3K2K3K4

þ Hþ½ 2K3K4

þ Hþ½ K4

þ 1 ðA1:8Þ

Writing the right side of Eq. (A1.8) as a fraction with a common denominator and

then inverting the fraction gives the desired final expression

a4 ¼ K1K2K3K4

Hþ½ 4 þ Hþ½ 3K1 þ Hþ½ 2K1K2 þ Hþ½ K1K2K3 þK1K2K3K4

: ðA1:9Þ

An important feature of Eq. (A1.9) to notice is that each term in the expression

actually represents the contribution of a particular protonated form, thus:

K1K2K3K4 , A4�

Hþ½ K1K2K3 , HA3�

Hþ½ 2K1K2 , H2A2�

Hþ½ 3K1 , H3A�

Hþ½ 4 , H4A

This insight makes it easy to write the fraction of the polybasic acid that is in a

particular form: the term representing the particular protonated form appears in

the numerator, while the denominator is simply the sum of all the possible terms.

For example, the fraction existing as HA3� is

a3 ¼ Hþ½ K1K2K3


; ðA1:10Þ

and the fraction existing in the doubly deprotonated H2A2� form is


a2 ¼ Hþ½ 2K1K2


ðA1:11Þ

The plots shown in Fig. 7 were generated using the above expressions for a2, a3,and a4 in conjunction with the four stepwise dissociation constants for EGTA.

In footnote 14, it was stated that Kd represents the ‘‘absolute’’ or intrinsic

dissociation constant characterizing the fully deprotonated form of the chelator

(e.g., A4� in the case above). At any pH where not all of the chelator is in the fully

deprotonated form, Kd must be corrected for the weakening eVect of acidic pH; the

corrected, or ‘‘conditional,’’ dissociation constant is K0d. As one would expect, the

correction factor is the fraction of chelator that exists in the fully deprotonated

form at the desired pH (e.g., a4 in the case of a tetrabasic acid like EGTA). Thus,

for a tetrabasic chelator,

K0d ¼

Kd

a4ðA1:12Þ

The plots shown in Fig. 8 were generated using Eq. (A1.12).

Appendix 2. Deriving an expression for the amount offluorescence emitted by a solution of fluorescent indicator

Light absorption by a solution containing a light-absorbing molecule, such as a

colorimetric or fluorescent indicator, is described by the Beer–Lambert Law:

A ¼ � logI

I0¼ elc ðA2:1Þ

whereA is the absorbance (or ‘‘optical density’’) of the solution, I0 is the intensity of

a light beam impinging on the solution, I is the intensity after the beam has passed

through the solution (I0� I¼ Iabs is the amount of light absorbed), e is the molar

extinction coeYcient (also known as themolar absorptivity), l is the thickness of the

solution throughwhich the light beam passes, and c is the concentration of the light-

absorbing molecule. The equation can be rearranged to the exponential form:

I ¼ I0e�2:303elc ðA2:2Þ

By convention, e has units of M�1cm�1 (i.e., l mol�1cm�1), l is measured in cm,

and c is measured in units of molarity (M, or mol l�1). For typical imaging

experiments where fluorescent indicators are loaded into cells, numerical limits

may be defined for the three parameters of interest:

e < 50; 000M�1 cm�1

l < 50� 10�4 cm

c < 100� 10�6M


The basis for these limits is as follows. First, indicators are small molecules whose

extinction coeYcients almost never surpass 50,000. Second, with extremely rare

exceptions, cells do not exceed 50 mm in diameter. Third, incubation with AM

esters usually achieves intracellular indicator concentration of several tens of

micromolar, and when indicators are introduced through a whole-cell patch elec-

trode, the concentration is usually kept below 100 mM to ensure that excessive

Ca2þ buVering capacity is not introduced into the cell. Applying these limits gives

2:303elc < 0:0575 � 1

Knowing that for x�1, e�x�1�x, we can recast Eq. (A2.2) as

I ¼ I0 1� 2:303elcð Þ ¼ I0 � 2:303elcI0 ðA2:3Þwhich rearranges to a simple expression of the amount of light absorbed by the

sample:

I0 � I ¼ Iabs ¼ 2:303elcI0 ðA2:4ÞThis approximate expression is accurate to within �1% for absorbances less than

0.06 (i.e., elc<0.06).

After absorbing a photon, a fluorescent molecule may reemit the absorbed

quantum of energy as fluorescence. The probability that after absorbing a photon,

a molecule will emit a photon of fluorescence is known as the quantum eYciency of

fluorescence. Quantum eYciency is usually symbolized as Q or f (Greek phi). The

amount of fluorescence emission should be the amount of light absorbed multi-

plied by the quantum eYciency:

F ¼ 2:303QelcI0 ðA2:5ÞIn this expression, only Q and e are intrinsic molecular properties, and these can

diVer substantially between the Ca2þ-free and Ca2þ-bound forms of the indicator

(symbolized as In and CaIn, respectively). For typical nonratiometric indicators,

QCaInQIn, while eCaIn� eIn. The total concentration of indicator, CT, is the sum

of the concentrations of the Ca2þ-bound and Ca2þ-free forms:

CT ¼ CaIn½ þ In½ ðA2:6ÞMaking use of the dissociation equilibrium constant:

CaIn > Ca2þ þ In Kd ¼ Ca2þ� �

In½ CaIn½

we can deduce the fraction of indicator that is Ca2þ-bound, fCaIn, and the fraction

that is Ca2þfree, fIn:

fCaIn ¼ Ca2þ� �

Ca2þ� �þKd

and fIn ¼ Kd

Ca2þ� �þKd

ðA2:7Þ


and, by necessity, fCaInþ fIn¼1. Therefore,

CaIn½ ¼ fCaInCT and In½ ¼ fInCT ¼ 1� fCaInð ÞCT ðA2:8ÞThe total fluorescence, FT, from a solution of indicator contains contributions

from both CaIn and In forms:

FT ¼ FCaIn þ FIn ¼ 2:303QCaIneCaInl CaIn½ I0 þ 2:303QIneInl In½ I0¼ 2:303QCaIneCaInlfCaInCTI0 þ 2:303QIneInlfInCTI0

¼ 2:303lCTI0 QCaIneCaInfCaIn þQIneInfIn½ ðA2:9Þ

This shows that the total fluorescence depends on the intrinsic molecular

properties of In and CaIn and the relative abundance of the two forms:

FT / QCaIneCaInfCaIn þQIneInfIn ðA2:10ÞMoreover, because the product, Qe, is a composite measure of a molecule’s ability

to absorb light and then emit fluorescence, we may think of Qe as the ‘‘intrinsic

brightness’’ of a fluorescent molecule. The brightness contribution of each indica-

tor form to the total fluorescence is weighted by the relative abundance of each

form. Finally, because fCaInþ fIn¼1, we can write

FT / QCaIneCaInfCaIn þQIneIn 1� fCaInð Þ ðA2:11Þ

References

Anderegg, G., and Wenk, F. (1967). Pyridinderivate als Komplexbildner. VIII. Die Herstellung je eines

neuen vier- und sechszahnigen Liganden. Helv. Chim. Acta 50, 2330–2332.

Anderegg, G., Hubmann, E., Podder, N. G., andWenk, F. (1977). Pyridinderivate als Komplexbildner.

XI. Die Thermodynamik der Metallkomplexbildung mit Bis-, Tris- und Tetrakis[(2-pyridyl)methyl]-

aminen. Helv. Chim. Acta 601, 123–140.

Arslan, P., Di Virgilio, F., Beltrame, M., Tsien, R. Y., and Pozzan, T. (1985). Cytosolic Ca2+ homeo-

stasis in Ehrlich and Yoshida carcinomas. A new, membrane-permeant chelator of heavy metals

reveals that these ascites tumor cell lines have normal cytosolic free Ca2+. J. Biol. Chem. 260,

2719–2727.

Babcock, D. F., Herrington, J., Goodwin, P. C., Park, Y. B., and Hille, B. (1997). Mitochondrial

participation in the intracellular Ca2+ network. J. Cell Biol. 136, 833–844.

Di Virgilio, F., Steinberg, T. H., Swanson, J. A., and Silverstein, S. C. (1988). Fura-2 secretion and

sequestration in macrophages. A blocker of organic anion transport reveals that these processes occur

via a membrane transport system for organic anions. J. Immunol. 140, 915–920.

Di Virgilio, F., Steinberg, T. H., and Silverstein, S. C. (1990). Inhibition of Fura-2 sequestration and

secretion with organic anion transport blockers. Cell Calcium 11, 57–62.

Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985). A new generation of Ca2+ indicators with greatly


Haugland, R. P. (1992). ‘‘Handbook of Fluorescent Probes and Research Chemicals.’’ Molecular

Probes, Eugene, Oregon.


Hepler, P. K., and Callaham, D. A. (1987). Free calcium increases during anaphase in stamen hair cells

of Tradescantia. J. Cell Biol. 105, 2137–2143.

Kao, J. P. Y., Harootunian, A. T., and Tsien, R. Y. (1989). Photochemically generated cytosolic calcium


Kao, J. P. Y., Alderton, J. M., Tsien, R. Y., and Steinhardt, R. A. (1990). Active involvement of Ca2+ in

mitotic progression of Swiss 3T3 fibroblasts. J. Cell Biol. 111, 183–196.

Liu, C.-M., and Hermann, T. E. (1978). Characterization of ionomycin as a calcium ionophore. J. Biol.

Chem. 253, 5892–5894.

Malgaroli, A., Milani, D., Meldolesi, J., and Pozzan, T. (1987). Fura-2 measurement of cytosolic free

Ca2+ in monolayers and suspensions of various types of animal cells. J. Cell Biol. 105, 2145–2155.

Martell, A. E., and Smith, R. M. (1974). ‘‘Critical Stability Constants,’’ Vol. I. Plenum, New York.

McNeil, P. L. (1989). Incorporation of macromolecules into living cells. Meth. Cell Biol. 29, 153–173.

McNeil, P. L. (2001). Direct introduction of molecules into cells. Curr. Protoc. Cell Biol. 20.1.1–20.1.7.

Minta, A., Kao, J. P. Y., and Tsien, R. Y. (1989). Fluorescent indicators for cytosolic calcium based on

rhodamine and fluorescein chromophores. J. Biol. Chem. 264, 8171–8178.

Motoyama, K., Karl, I. E., Flye, M. W., Osborne, D. F., and Hotchkiss, R. S. (1999). Effect of Ca2+

agonists in the perfused liver: Determination via laser scanning confocal microscopy. Am. J. Physiol.

276, R575–R585.

Niggli, E., Piston, D. W., Kirby, M. S., Cheng, H., Sandison, D. R., Webb, W. W., and Lederer, W. J.

(1994). A confocal laser scanning microscope designed for indicators with ultraviolet excitation

wavelengths. Am. J. Physiol. 266, C303–C310.

Sako, Y., Sekihata, A., Yanagisawa, Y., Yamamoto, M., Shimada, Y., Ozaki, K., and Kusumi, A.

(1997). Comparison of two-photon excitation laser scanning microscopy with UV-confocal laser

scanning microscopy in three-dimensional calcium imaging using the fluorescence indicator Indo-1.

J. Microsc. 185, 9–20.

Tsien, R. Y. (1980). New calcium indicators and buffers with high selectivity against magnesium and

protons: Design, synthesis, and properties of prototype structures. Biochemistry 19, 2396–2404.

Tsien, R. Y., and Bacskai, B. J. (1995). Video rate confocal microscopy. In ‘‘Handbook of Biological

Confocal Microscopy,’’ (J. B. Pawley, ed.), pp. 459–478. Plenum Press, New York.

Tsien, R. Y., and Pozzan, T. (1989). Measurement of cytosolic free Ca2+ with quin2. Meth. Enzymol.

172, 230–262.

Tsien, R. Y., Pozzan, T., and Rink, T. J. (1982). Calcium homeostasis in intact lymphocytes: Cytoplas-

mic free calcium monitored with a new, intracellularly trapped fluorescent indicator. J. Cell Biol. 94,

325–334.

CHAPTER 6


Genetically Encoded Probes forMeasurement of Intracellular Calcium

Michael WhitakerInstitute of Cell and Molecular BiosciencesMedical School, Newcastle University, Framlington PlaceNewcastle upon Tyne, United Kingdom

A

OGY,All rig

bstract


-679X-679X

I. In
troduction II. G enetically Encoded Sensors
A.
The Cameleon Family B. Camgaroos C. Pericam G-CaMP Family
III. A
pplications of Genetically Encoded Sensors A. Targeting to Subcellular Locations B. Tissue-Specific Expression
IV. U
se of Genetically Encoded Calcium Sensors V. C onclusions
R
eferences
Abstract

Small, fluorescent, calcium-sensing molecules have been enormously useful in

mapping intracellular calcium signals in time and space, as chapters in this volume

attest. Despite their widespread adoption and utility, they suVer some disadvan-

tages. Genetically encoded calcium sensors that can be expressed inside cells by

transfection or transgenesis are desirable. The last 10 years have been marked by a

rapid evolution in the laboratory of genetically encoded calcium sensors both

figuratively and literally, resulting in 11 distinct configurations of fluorescent

proteins and their attendant calcium sensor modules. Here, the design logic and

performance of this abundant collection of sensors and their in vitro and in vivo

use and performance are described. Genetically encoded calcium sensors have

proved valuable in the measurement of calcium concentration in cellular

/10 $35.00(10)99006-7

154 Michael Whitaker

organelles, for the most part in single cells in vitro. Their success as quantitative

calcium sensors in tissues in vitro and in vivo is qualified, but they have proved

valuable in imaging the pattern of calcium signals within tissues in whole animals.

Some branches of the calcium sensor evolutionary tree continue to evolve rapidly

and the steady progress in optimizing sensor parameters leads to the certain hope

that these drawbacks will eventually be overcome by further genetic engineering.

I. Introduction

Small, fluorescent, calcium-sensing molecules have been enormously useful

in mapping intracellular calcium signals in time and space, as chapters in this

volume attest. Despite their widespread adoption and utility, they suVer some

disadvantages.

All low molecular mass fluorescent cytoplasmic calcium sensors are highly

charged molecules, so cross the cell’s plasma membrane very poorly. They are

placed into the cytoplasm by microinjection using fine-tipped micropipette or a

patch clamp pipette in whole cell mode. This limits their utility. Cell-permeant

fluorescent calcium sensors can be made by masking the charged carboxylic groups

by forming acetoxymethyl (AM) esters. Once inside the cell, the ester bonds are

cleaved, trapping the sensor in the cell. It is straightforward to bathe cells in culture

with the aposensor at low concentration and these AM esters have been very

widely used. One major drawback of the method is that the calcium sensor finds

itself not only in the cytoplasm, but also in intracellular compartments such as the

endoplasmic reticulum (ER) (Silver et al., 1992). Calcium concentrations are

higher in the ER than in the cytoplasm, so this leads to a significant unwanted

fluorescence signal from sensor in the ER that makes interpretation of the true

cytoplasmic concentration changes diYcult. It is also very challenging to use low-

molecular-mass fluorescent calcium sensors in whole animals.

For these reasons, genetically encoded calcium sensors that can be expressed

inside cells by transfection or transgenesis are desirable. One such sensor is

aequorin, a calcium-sensing protein found in the jellyfish Aequoria victoria. Origi-

nally, aequorin was isolated as a protein from jellyfish and placed inside cells by

microinjection (Baker, 1978; Gilkey et al., 1978). More recently, a construct

encoding recombinant aequorin has been used to express the aequorin apoprotein

in cells directly (see Chapter 10). Aequorin is a luminescent molecule and at the

concentrations used inside cells emits relatively few photons compared to fluores-

cent molecules at appropriate excitation intensities (Varadi and Rutter, 2002b).

However, proteins that are fluorescent at the visible wavelengths best suited to

fluorescence imaging are relatively rare. As it happens, A. victoria also expresses

a fluorescent protein, green fluorescent protein (GFP), and it is the work that

has produced the variously colored versions of GFP that has improved our

knowledge of this fluorophore and led to recombinant fluorescent calcium sensors

(Tsien, 2010).

6. Genetically Encoded Probes for Measurement of Intracellular Calcium 155

The first recombinant fluorescent calcium sensors were described by Tsien and

Persecchini in 1997 (Miyawaki et al., 1997; Persechini et al., 1997; Romoser et al.,

1997). They were based on a concatenation of a recombinant calcium-binding

domain with GFP-derived fluorescent protein pairs. This approach has bred a

family of these cameleon indicators, so called because they are based on a long

tongue-like interaction between calmodulin (CaM) and a binding peptide and

change color (Miyawaki et al., 1997). Later, when it was realized that the GFP

beta-can structure lent itself to circular permutation without loss of function

(Baird et al., 1999), insertion of a calcium-binding domain within the GFP

(Baird et al., 1999) or concatenated to new N- or C-terminals (Nakai et al., 2001)

led to a second family of calcium sensors based on the fluorescence of a single

GFP-derived molecule, the camgaroos, pericams and their relatives.

The last 10 years have been marked by a rapid evolution in the laboratory of

these two families and their relatives, both figuratively and literally, as random

mutagenesis and clonal selection in bacteria has on occasion been used to opti-

mize the properties of the sensors (Griesbeck et al., 2001). This rapid diversifica-

tion has generated not only continuing improvements in the performance of the

sensors, but also a plethora of choice. Reviews have been written to track progress

in the field (Barth, 2007; Demaurex and Frieden, 2003; Garaschuk et al., 2006;

Griesbeck, 2004; Hires et al., 2008; KotlikoV, 2007; Mank and Griesbeck, 2008;

Miyawaki, 2003a,b, 2005; Pozzan and Rudolf, 2009; Solovyova and Verkhratsky,

2002; Zacharias et al., 2000). Most of the new variants have first been tested by

their makers in living cells as proof of principle rather than to answer substantial

questions in biology. I shall first set out the evolution of this growing tribe of

genetically encoded calcium sensing probes, dealing with the two broad families

in turn and then describe their application and utility in various biological

settings.

II. Genetically Encoded Sensors

A. The Cameleon Family

1. Origins

The family founders were described in three papers that followed rapidly in

succession in 1997. Their conception was aided by previous work in which GFP

had been altered by directed mutagenesis to produce diVerent colored variants

with altered excitation and emission spectra (Heim et al., 1995). As an aside, these

diVerently colored variants are sometimes referred to collectively as GFPs, though

they are not green. Persechini’s group described a construct (FIP-CBsm) in which a

red-shifted excitation variant of GFP (RSGFP; Delagrave et al., 1995, hereafter

GFP) and blue fluorescent protein (BFP) are linked by a sequence that includes 17

amino acids from the calmodulin-binding domain of avian myosin light chain

kinase (MLCK). This novel protein indirectly senses calcium concentrations inside


cells, as when calcium increases, endogenous calmodulin becomes activated and

binds to the MLCK calcium-binding domain. This in turn alters the disposition of

the attached GFPs and leads to changes in Forster resonance energy transfer

(FRET) between the blue and green proteins (Romoser et al., 1997).

FRET is the phenomenon on which the cameleon sensor family relies. It occurs

between closely apposed fluorophores that have overlapping emission and excita-

tion spectra (Jares-Erijman Elizabeth and Jovin Thomas, 2003). In this example,

the emission spectrum of BFP overlaps with the excitation spectrum of GFP. The

extent of FRET depends on the degree of overlap between the two spectra, the

orientation of the fluorescence dipoles and crucially, the distance between them.

There is a very steep sixth power relationship with distance, so the energy transfer

is very sensitive to distance between fluorophores over the range 1–10 nm (Jares-

Erijman and Jovin, 2003). Calmodulin binds to the helical MLCK sequence by

wrapping its two lobes around it (Ikura et al., 1992). In FIP-CBsm, the steric bulk

of the calmodulin molecule when it binds to the MLCK peptide linker forces the

BFP and RFP further apart and reduces FRET (Romoser et al., 1997). FRET can

be measured in a variety of ways (Jares-Erijman Elizabeth and Jovin Thomas,

2003; Visser et al., 2010), but conceptually the simplest method is to excite the

donor fluorophore, here BFP, and measure the emission of both the donor and the

acceptor, here GFP. FRET takes place by nonradiative energy transfer, so high

levels of FRET transfer energy from BFP to GFP, reducing BFP emission at

around 440 nm and increasing GFP emission at 510 nm. Calmodulin binding

reduces FRET, increasing emission at 440 nm and reducing emission at 510 nm.

These changes can be expressed as a ratio of emission at the two wavelengths, a

value independent of the concentration of the protein. In HEK-239 cells expressing

FIP-CBsm, ratio changes (F510/F440) of around three- to fourfold could be observed

after raising free intracellular calcium concentration with the calcium ionophore

ionomycin (Romoser et al., 1997).

FIP-CBsm relied on endogenous calmodulin to generate a calcium-sensitive

FRET signal between GFPs. Tsien’s construct concatenated Xenopus laevis cal-

modulin and an MLCK calmodulin-binding peptide, M13 (Ikura et al., 1992),

together between BFP and GFP and also in an analogous construct between two

other GFP variants, enhanced cyan fluorescent protein (ECFP) and enhanced

yellow fluorescent protein (EYFP). In this concatenated configuration, binding

of calcium to calmodulin causes it to loop back toward the M13 peptide (the

cameleon’s tongue) as it binds, reducing the distance between the two GFP

variants and enhancing FRET (Miyawaki et al., 1997). This study beautifully

exemplifies the power of the cameleon concept linked to selective mutagenesis:

the original BFP/GFP construct (cameleon-1) worked well in vitro, but did not

express suYciently in mammalian cells; the enhanced variant with mammalian

codon usage (EBFP/EGFP—cameleon-2) showed much improved expression, but

the best expression, brightness, and signal-to-noise data were seen with enhanced

cyan and yellow variants of GFP (ECFP/EYFP—yellow cameleon-2). These ben-

efits came, however, at the expense of a lower FRET change between calcium


bound and unbound forms (1.5 vs. 1.8 when expressed as a ratio of emission

wavelengths) and a greater pH sensitivity. Mutagenesis can also be applied to the

calcium-binding aYnity of the calmodulin moiety: calmodulin has two classes of

calcium-binding sites and site-directed mutations in either high-(K0d 70 nM) or low

(K0d 11 mM) aYnity sites give rise to constructs in which high-aYnity sites are

suppressed to give a monotonic binding curve (K0d 4.4 mM: cameleon-3) or low-

aYnity sites are altered to give an enhanced range over four orders of magnitude of

calcium concentration (K0ds of 83 nM and 700 mM: cameleon-4). The third dimen-

sion of modification adds signal tags to the constructs. Nuclear localization tags

gave cameleon-2nu and ER localization tags produced yellow cameleon-3er (K0d

4.4 mM) and cameleon-4er (K0ds of 83 nM and 700 mM).

Tsien’s seminal paper also exemplifies some challenges in the approach: on the

one hand, the complexities of permutation and combination of mutant variants

and their concomitant properties and on the other hand, the relatively low magni-

tude of FRET modulation by calcium over a very wide range of concentrations.

The subsequent proliferation of family members results from attempts to improve

brightness and dynamic range, but at the expense of adding to the combinatorial

complexity.

Persechini’s second sensor design also concatenated GFPs, MLCK peptide, and

calmodulin, though in diVerent order. A calmodulin whose EF hand calcium-

binding sites had been reversed in order (CN-CaM) was added to the FIP-CBsm

C-terminal to BFP to make FIP-CA (Persechini et al., 1997). This produced a

sensor with a monotonic FRET response and a K0d of 100 nM. Variants with lower

aYnities for calcium were obtained by mutating the MLCK calmodulin-binding

peptide sequence, rather than the calmodulin calcium-binding sites. As with FIP-

CBsm, calmodulin binding reduced FRET, the ratio (now expressed as F440/F510)

increasing approximately 1.7-fold over the calcium dynamic range. The interaction

was markedly pH sensitive in the range 6.5–7.5. This configuration of calmodulin

and calmodulin-binding peptide did not lead to later variants and appears to have

been an evolutionary dead end.

The cameleon family of calcium sensors is shown in Fig. 1.

2. Evolution

The EYFP in yellow cameleon-2 and-3 shows an apparent pKa of 6.9, so

contains a significant proportion of the protonated species at physiological pH

(Miyawaki et al., 1999). The protonated species does not participate in FRET

(Habuchi et al., 2002). As pH can vary by several tenths of a pH unit when cells are

stimulated; changes in pH would be read as changes in calcium ion concentration.

Two adjacent point mutations in EYFP (V68L and Q69k) lower the pKa to 6.1,

markedly reducing the pH sensitivity in the physiological range (Miyawaki et al.,

1999). Replacing EYFP with EYFP-V68L/Q69K abolished pH sensitivity above

pH 6.9 (Miyawaki et al., 1999). This substitution produces yellow cameleon-2.1

(YC2.1; K0ds for calcium: 100 nM and 4.3 mM) and yellow cameleon-3.1 (YC3.1;

Fig. 1 Schematic depiction of the diVerent classes of genetically encoded calcium sensors. EYFP and EGFP variants for individual

sensors are shown to the right, as are the identities of the red-emitting sensors.


K0d for calcium: 1.5 mM) with around a twofold diVerence in 528/476 nm emission

ratios in calcium-free and calcium-saturating conditions. Recalling that the calci-

um-dependent signal from FIP-CBsm relied on binding of endogenous calmodulin,

an obvious concern would be that YCs would be perturbed by such interactions

and also perhaps themselves perturb downstream calcium-signaling pathways.

In fact, EC50s for YC2.1 and YC3.1 stimulation of calmodulin-dependent phos-

phodiesterase were two to three orders of magnitude greater than for calmodulin

and the sensors were unperturbed by addition of 3 mM calmodulin. Of course, the

YC constructs will buVer calcium inside cells. This was tested by studying the

calcium oscillations induced in HeLa cells induced by addition of histamine. At a

YC3.1 concentration of 150 mM, calcium oscillations were evident whereas at

concentrations greater than 300 mM, oscillations were not seen, though the overall

magnitude of the response was little altered. The loss of oscillations suggests

calcium buVering. Below around 20 mM, the fluorescent signal was too faint to

give acceptable signal-to-noise ratios (Miyawaki et al., 1999). Thus, working YC

concentrations in the range 40–150 mM do not substantially perturb calcium-

dependent signaling mechanisms.

Yellow fluorescent proteins, besides being sensitive to pH, are more prone than

GFP to photobleaching and to quenching by biological anions such as chloride.

Because YFPs show such utility as one of the partners in the CFP/YFP FRET

couple, this defect is worth fixing. Mutagenesis by error-prone PCR and expression

in Escherichia coli uncovered a mutation to methionine in residue 69 that was much

more resistant to chloride quenching than EYFP-V68L/Q69K, twice as resistant to

photobleaching, with a pKa of 5.7 rather than 6.1 and of comparable spectral

properties including brightness (Griesbeck et al., 2001). This YFP is known as

citrine, and substituted for EYFP-V68L/Q69K as the FRET acceptor produced

the cameleons YC2.3 and YC3.3. These two cameleons express well at 37 �C, showa ratio change of around 1.5 to calcium over their dynamic range and are pH

insensitive down to around pH 6.5. To demonstrate the utility of YC3.3 in an

acidic compartment, it was targeted to the Golgi using an 81 residue N-terminal

construct from human galactosyl transferase type II. The sensor was saturated

when expressed in the Golgi, suggesting high resting levels of free calcium concen-

tration in this cellular compartment (Griesbeck et al., 2001).

The CFP/citrine couple was also used in an ER-targeted sensor, Cameleon

D1ER. Here, the rationale was to design a sensor based on the M13/CaM-biding

pair that would be insensitive to interaction with endogenous calmodulin (Palmer

et al., 2004), as had been reported (Hasan et al., 2004; Heim and Griesbeck, 2004).

The M13 and CaM were co-mutated to provide a binding pair that would not

interact strongly with endogenous calmodulin. Cameleon D1ER has a very wide

range of calcium sensitivity with K0ds of 0.81 and 60 mM, appropriate for ER

calcium sensing, and was successfully used in HeLa cells to monitor cytoplasmic

and ER calcium simultaneously in conjuction with Fura2 (Palmer et al., 2004).

The GFP family of proteins is remarkable in possessing a visible wavelength

fluorophore that is formed through an oxidation reaction involving adjacent


amino acids (Tsien, 1998). Fluorescence develops relatively slowly when the pro-

tein is expressed in cells, the process of what is known as maturation taking tens of

minutes to hours; maturation is also temperature dependent, oxidation to form the

fluorophore being the rate-limiting step. Another potential diYculty with FRET-

based probes using the CFP/YFP partners is that maturation of YFP is substan-

tially slower than that of CFP, particularly at mammalian body temperatures

(Miyawaki et al., 1999), a very important consideration especially for expression

in transgenic mammals. If the YFP partner of the FRET couple matures more

slowly than the CFP partner, then the sensors dynamic range is compromised, as

mature CFP in a sensor that contains immature YFP will contribute to the 476 nm

emission in the absence of 528 nm emission from the same construct, so that the

overall population 528/476 emission ratio will be depressed as a function of the

proportion of disparately matured sensor constructs (as illustrated by the behavior

of YC6.1 discussed below; Evanko and Haydon, 2005). The F46L mutation in

YFP greatly accelerates oxidation to the mature fluorophore and four additional

point mutations contributed to create a construct that matured two orders of

magnitude faster than EYFP from a urea-denatured state (Nagai et al., 2002;

Rekas et al., 2002); because of its resulting brightness, this YFP construct was

given the name Venus. Venus also has a low pKa (6.0) and low sensitivity to

chloride, comparable to citrine in these respects (Griesbeck et al., 2001), though

it lacks citrine’s improved resistance to photobleaching. Substitution of Venus for

EYFP-V68L/Q69K resulted in a new rapidly maturing yellow cameleon (YC2.12).

Bright YC2.12 fluorescence was seen to develop rapidly after ballistic transfection

of Purkinje cells in cerebellar slices, though the fold ratio change after depolariza-

tion suggests that its dynamic range was not much altered from earlier family

members (Nagai et al., 2002).

The challenge of improving dynamic range was addressed systematically by

altering the orientation of the YFP fluorescence dipole relative to the CFP dipole

(Jares-Erijman and Jovin, 2003) to maximize FRET (Nagai et al., 2004). Changes

in orientation were achieved by circular permutation (see below, Section II.B.1) of

the Venus construct. The YC3.12-based construct with EYFP-V68L/Q69K sub-

stituted by circularly permutated Venus with a new N-terminal at Asp-173 (termed

YC3.60) showed the largest increase in fluorescence emission ratio dynamic range

between calcium free and calcium-bound forms in vitro: around 6.6-fold compared

to 2.1-fold for YC3.12. This large improvement in dynamic range was verified by

expression of each the two sensors in HeLa cells and challenge with ATP to raise

cytoplasmic free calcium levels (Nagai et al., 2004). This study also illustrates the

important point that altering the properties and conformation of the FRET

partners at the N- and C-terminals of the sensor can also alter the apparent calcium

activation characteristics of the calmodulin-M13 inner pair as measured by FRET.

YC3.60 showed a monotonic increase with calcium concentration, as would be

expected from a construct based on the monotonically increasing cameleon-3

(Miyawaki et al., 1997), but the apparent dissociation constant for YC3.60 is

0.25 mM, compared to 4.4 mM for cameleon-3. YC2.60, based on cameleon-2,


has a single high-aYnity K0d of 40 nM, compared to the two K0

ds of 70 nM and

11 mM of cameleon-2 (Miyawaki et al., 1997). YC4.6 has K0ds of 58 nM and

14.4 mM, compared to K0ds of 83 nM and 700 mM in cameleon-4. The YCX.60

series of cameleons show the rapid maturation and low pH and chloride sensitiv-

ities of their Venus forbears, YCX.12, and are the best performing native M13-

based cameleons to date.

Both citrine and the circularly permutated Venus (cpv) of the YCX.60 series were

used as alternative acceptors in a further series of cameleons based on Cameleon

D1ER (Palmer et al., 2006). Computational design of novel M13 and CaM-based

binding pairs led to Cameleons D2, D3, and D4 and D2cpv, D3cpv, and D4cpv,

oVering a wide range of calcium aYnities, good sensor dynamic range (the cpv series

comparable to YC3.60), and insensitivity to endogenous calmodulin. This cameleon

set showed good performance in reporting cytoplasmic and mitochondrial calcium

concentrations in HeLa cells and peri-plasmalemmal calcium concentrations in

hippocampal neurones when localized with the appropriate targeting sequences.

ECFP/EYFP-based cameleons require excitation at near-UV wavelengths.

It would be convenient to have FRET-based calcium sensors that can be excited

at visible wavelengths. One possibly solution is to use a FRET couple in which

GFP is paired with a red fluorescent protein. GFP-like red fluorescent proteins are

found in corals (Baird et al., 2000; Miyawaki et al., 2003b). However, they are less

tractable than GFP and its variants as they oligomerize, mature very slowly via a

green-emitting intermediate and in general, show low extinction coeYcients and

quantum yield (Miyawaki et al., 2003b). A GFP/RFP cameleon has been devel-

oped using a DsRed variant—a tandem dimer mutant (Yang et al., 2005). The

maturation rate is tens of hours and the emission ration change is less than 1.2-fold

when cells expressing the sensor are challenged with ionomycin (Yang et al., 2005).

3. Changing the Sensor Mechanism 1

Solution NMR showed that the calmodulin-binding peptide of calmodulin-

dependent kinase kinase (CKKp) has a diVerent relation to the two lobes of

calmodulin than M13 peptide (Truong et al., 2001). The structural modeling

suggested that the peptide might be concatenated in a recombinant construct

between the N- and C-terminal lobes of calmodulin. Calculations suggested that

if ECFP and EYFP-V68L/Q69K were attached to the N- and C-terminals of the

split calmodulin, then the distance between the fluorophores when calcium was

bound and the calmodulin interacting with its binding peptide might be less than

40 A, rather than the 50–60 A in M13-based YC2.1. Given the sixth power depen-

dency of FRET on distance between fluorescent dipoles (Jares-Erijman and Jovin,

2003), this approach promised an improvement of the dynamic range of the ratio

of fluorescence emission. The splitting of the N- and C-domains of calmodulin in

this construct (termed YC6.1) led to a monotonic calcium-binding curve with a K0d

of 110 nM, in some respects more suited to measurement of smaller changes in

intracellular free calcium concentration. While in the event, YC6.1 showed a more


modest fold emission ratio change than predicted (2.1 vs. 1.4 for YC2.1 in parallel

experiments), the twofold change was expressed over a narrower range of calcium

concentrations (0.05–1 mM) in the physiologically relevant cytoplasmic range.

YC6.1 of course suVers from the pH and chloride sensitivity and the slow

maturation of its EYFP-V68L/Q69K fluorophore that we discussed above. Repla-

cing EYFP-V68L/Q69K with Venus (Evanko and Haydon, 2005) gives the sensor

VC6.1 (Venus cameleon 6.1: the nomenclature is confusing and unhelpful, given

that the Venus CaM–M13 cameleons are known as YC2.12 and YC3.12). VC6.1

shows a emission ratio change of around 2.1-fold between zero and saturating

calcium concentrations. Thus, as with substitution with Venus for EYFP-V68L/

Q69K to produce YC2.12 from YC2.1, dynamic range is not much altered, while

improvements in maturation and pH and chloride sensitivity are obtained. It

would be logical to develop a YC6 sensor that contains the circularly permutated

Venus used in YC2.6 and YC3.6 (Nagai et al., 2004); this would be predicted to

much improve the ratio dynamic range.

Small improvements in dynamic range for YC6.1 and VC6.1 can be obtained by

excluding from analysis cells that express a low resting YFP/CFP ratio (Evanko

and Haydon, 2005): the authors very reasonably suggest that this screens out cells

in which the YFP partner is less-mature relative to its CFP pair.

4. Changing the Sensor Mechanism 2

One potential disadvantage of calmodulin-based sensors is that calmodulin is a

near-ubiquitous protein with many binding partners. It is possible that calmodulin-

based sensors may suVer interference from binding partners when expressed in the

cytoplasm or other cellular compartments. While there is no direct evidence to

support this conjecture, it is nonetheless true that performance in vivo does not

always mirror the sensor properties demonstrated in vitro (Hasan et al., 2004; Heim

and Griesbeck, 2004). With this potential pitfall in mind, a sensor has been devel-

oped based on troponin C, a calcium-binding protein and close homologue of

calmodulin that is, however, expressed only in muscle. The approach was to

concatenate TnC with CFP and citrine (Heim and Griesbeck, 2004). While devel-

oping these CFP–TnC–citrine sensors, a variant strategy was pursued to concate-

nate TNI, a TnC-binding partner, alongside TnC by analogy with theM13 binding

partner of calmodulin in the classical cameleons; this was unsuccessful. The con-

structs showing the greatest change in FRET between calcium-free and calcium-

bound forms contained a chicken skeletal muscle TnC with an N-terminal 14

residue truncation, TN-L15, and a human cardiac TnC, TN-humTnC. TN-L15

showed a 140% change and TN-humTnC a 120% change, measured in the absence

of magnesium ion. At physiological (1 mM) magnesium concentrations, the

dynamic ranges were 100% and 70%, respectively. Apparent dissociation constants

were 470 nM for TN-L15 and 1.2 mMfor TN-humTnC. The TnCEF hand calcium-

binding sites in TN-L15 were mutated to give K0ds of 300 nM and 29 mM. The pH

sensitivities were similar to the other CFP/citrine-based sensors, with a reduction in


dynamic range below pH 6.8 and little eVect in the physiological range pH 6.8–7.3.

Calcium oV rates were similar to or slightly faster than that of YC2.3 (Heim and

Griesbeck, 2004). The TN-L15 sensor was targeted to the plasma membrane using

GAP43, Ras, or Synaptobrevin. In direct comparison with YC2.1 and YC 3.3, it

showed markedly greater sensitivity and no diminution of dynamic range.

Mutations to EF hands III and IV and substitution of citrine with a circularly

permutated variant, Citrine-cp174 produced a TnC-based sensor that showed no

magnesium dependence, a fourfold dynamic range and a K0d of 2.5 mM —TN-XL.

TN-XL has a very fast oV rate with a dominant component with a time constant of

142 ms (Mank et al., 2006). Expressed in Drosophila under a UAS/Gal4 neuronal

promoter, it showed response times to calcium signals at the neuromuscular junction

significantly faster thanother sensors—YC2.0,YC3.3, Inverse Pericam,G-CaMP1.3,

andG-CaMP1.6. Furthermutagenesis and rearrangement of the TnC domain gave a

higher aYnity variant, modestly named TN-XXL, that was capable of long-term

monitoring of individual neuronal responses in flies and mice (Mank et al., 2008).

B. Camgaroos

1. Circular Permutation of EYFP

Remarkably perhaps, the beta-can that surrounds the cyclized and oxidized

fluorophore is amenable to circular permutation, by which is meant the insertion

of a peptide linker between N- and C-terminals of the protein and the creation of a

newN- and C-terminal pair elsewhere in the sequence, in the loops that connect the

component beta-sheets and in the beta sheets themselves (Baird et al., 1999). As we

have seen, circular permutation of Venus led to YC2.60 and YC3.60, the two

cameleons with the largest emission ratio dynamic range (Nagai et al., 2004). The

discovery that N- and C-terminals of EYFP could be rearranged prompted the

discovery that a calcium sensor could be fashioned by insertion of calmodulin

within EYFP itself.Xenopus calmodulin was inserted between residues 144 and 146

of each of ECFP, EYFP, and EGFP. Each of these constructs was a calcium

sensor, with the EYFP insertion giving the largest calcium response. In calcium-

free conditions, the construct absorbs predominantly at 400 nm, while in calcium-

saturating conditions, the dominant absorption peak is at 490 nm. The 400-nm

absorption is due to the protonated form of EYFP and the 490-nm absorption to

the unprotonated form. As discussed in Section II.A.2, in EYFP, the protonated

species is not fluorescent (Habuchi et al., 2002), so the excitation spectrum shows a

single peak at 490 nm and both the excitation and emission spectra are strongly

dependent on calcium concentration, with around an eightfold increase in emission

intensity at saturating calcium concentrations. Calcium binding was monotonic

with an apparent dissociation constant of 7 mM. Calcium binding clearly shifts the

proportion of protonated and unprotonated forms at constant pH, so the pKa’s for

the two forms are diVerent: 10.1 and 8.9, respectively. Continuing the whimsical

tradition, this calcium sensor is termed Camgaroo-1, because it is yellowish, carries


a smaller companion (the calmodulin) in a pouch, can bounce high in signal and

may spawn improved progeny (Baird et al., 1999) The increase in fluorescence

intensity after addition of histamine to Camgaroo-1 expressing HeLa cells was a

modest 40% and the characteristic calcium spiking activity was almost invisible, so

the sensor is not quite as bouncy as its name implies when sensing cytoplasmic free

calcium; however, addition of ionomycin caused an overall sevenfold increase in

fluorescence. The modest increase observed in response to histamine is almost

certainly due to the 7 mM K0d, high relative to the calcium increase from around

100 nM to 1 mM expected when histamine is added to HeLa cells.

Camgaroo-1 does not fold well at 37 �C and could not be targeted to intracellu-

lar organelles, for example, mitochondria (Baird et al., 1999). In an attempt to live

up to another of its attributes, the possibility that it may spawn improved progeny,

Camgaroo-1 was subjected to error-prone PCR mutagenesis (Griesbeck et al.,

2001); selection of the brightest clone after expression in E. coli revealed a point

mutation of residue 69 to methionine. This new sensor, Camgaroo-2, had

very similar calcium-binding properties and fluorescence dynamic range as

Camgaroo-1, but expressed far more brightly in HeLa cells grown at 37 �C. Theresponse to histamine a (5% fluorescence increase) was lower even than for Cam-

garoo-1, but targeting to mitochondria using the targeting sequence of subunit

VIII of cytochrome c oxidase was demonstrated. Mitochondrial calcium increases

that raised the resting fluorescence signal by about 70% were demonstrated in

response to histamine and subsequent addition of ionopmycin gave an overall 1.5-

fold increase in fluorescence signal (Griesbeck et al., 2001), lower than that ob-

served with cytoplasmic Camgaroo-2, perhaps because the resting mitochondrial

calcium concentration is higher than that of the cytoplasm.

Using a similar camgaroo-like strategy, the EF hand calcium-binding site was

introduced into EGFP between residues 144–145, 157–158, or 172–173 (Zou et al.,

2007). These Ca-G family sensors had extinction coeYcients and quantum yields

comparable to EGFP. They operate in the ratiometric mode and with excitation at

398 and 490 nm showed a sensor dynamic range of 1.8 at a 510-nm emission

wavelength. Comprising a single EF hand-binding site, the apparent dissociation

constants are in the millimolar range (0.4–2 mM) and are, therefore, suitable only

for monitoring high calcium environments such as the ER. They are markedly pH

sensitive, with a pKa of around 7.5. Expressed in the ER of HeLa and BHK-21cells,

they showed modest ratio changes in response to agonists (Zou et al., 2007).

C. Pericam G-CaMP Family

1. Pericams

In pursuit of the idea that the clefts introduced into the beta can structure by

circular permutation might make the fluorophore more accessible to solution

protons and so susceptible to structural changes brought about by reorientation

of concatenated peptides, Miyawaki’s group developed the pericam series of


sensors (Nagai et al., 2001). Circular permutation of EYFP-V68L/Q69K to give an

EYFP with Y145 as the N-terminal and N144 as the C-terminal (cpEYFP)

produced an EYFP variant that could be concatenated with M13 and calmodulin

(bearing the E104Q mutation that conferred a monophasic calcium-binding curve;

Miyawaki et al., 1997). The construct with calmodulin at the N-terminal (CaM–

cpEYFP–M13) showed no calcium-dependent properties, confirming the finding

reported for a cpGFP variant (Nakai et al., 2001), but the opposite concatenation

(M13–cpEYFP–CaM) gave a construct that showed threefold brighter 520 nm

fluorescence in high calcium media compared to calcium-free media when excited

at 485 nm. This construct was given the name pericam (from a circularly permuted

YFP and CaM—calmodulin). Pericam was the prototype from which three peri-

cams with enhanced features were developed. Flash pericam has three additional

point mutations that confer an eightfold increase in 520-nm fluorescence on

calcium binding. Flash Pericam is a single wavelength, nonratiometric indicator

with a K0d of 0.7 mM .Knowing that substitution of phenylalanine at residue 203 in

YFP conferred fluorescence on the protonated form, this mutation was introduced

into Flash Pericam. The result, Ratiometric Pericam, was a sensor whose emission

ratio at 520 nm when excited at 494 nm or 415 nm changes 10-fold between

calcium-free and calcium-saturating conditions with a K0d of 1.7 mM; this excita-

tion ratio sensor is functionally analogous to fura-2 (Grinkiewicz et al., 1985).

Further semirandommutagenesis of Ratiometric Pericam gave a single wavelength

construct whose fluorescence intensity at 513–515 nm decreased on calcium bind-

ing—Inverse Pericam (K0d; 0.2 mM). Two advantages of Inverse Pericam are that it

is bright and has excitation/emission characteristics similar to fluorescein; the latter

advantage it shares with Flash Pericam: these two YFP-based indicators are

functionally equivalent to the Fluo-3 and Fluo-4 single wavelength calcium sensors

(Gee et al., 2000; Kao et al., 1989; Minta et al., 1989). Expression in HeLa cells

showed that Ratiometric Pericam and Inverse Pericam expressed significantly

better at 37 �C than did Flash Pericam. Ratiometric Pericam gave a 2.5-fold

increase in excitation ratio emission after addition of histamine, while Flash and

Inverse Pericams oVer a �100% increase and decrease in signal, respectively, with

the same agonist. As might be expected from our earlier discussion of the camgar-

oos, the calcium-free and calcium-bound forms of all three pericams showed

diVerent pKa’s and all three have pH sensitive emissions in the physiological pH

range. Miyawaki showed a proof of principle that the excitation ratio-based

Ratiometric Pericam can be used in the context of confocal imaging (Shimozono

et al., 2002); recent confocal microscopes based on acousto-optical filters oVerturnkey solutions to excitation ratiometric imaging.

2. GCaMPs

Single wavelength nonratiometric sensors that use the same sensor strategy

as pericams but are based on circularly permutated GFP rather than EYFP

were developed at almost the same time as the pericams, their development


preceding the pericams’ by a matter of months (Nakai et al., 2001). Both the CaM–

cpGFP–M13 and M13–GFP–CaM concatenations were tested: only the latter

showed significant calcium-sensing properties. Twenty-six variants of the M13–

N149cpGFPC144–CaM concatenate were tested and the variant that showed the

greatest fluorescence increase in HEK-239 cells after ATP addition was termed

G-CaMP (presumably for green fluorescent calmodulin protein). In HEK-239

cells, G-CaMP gave a 1.5-fold increase in fluorescence in response to ATP and a

fourfold increase in response to ionomycin. G-CaMP has very similar fluorescence

parameters to Flash Pericam, with an excitation maximum at 489 nm, an emission

maximum at 509 nm and a 4.5-fold increase in fluorescence on calcium binding

(cf. eightfold for Flash Pericam). The apparent dissociation curve was monotonic,

with a K0d of 0.24 mM. As with the camgaroos and pericams and for the same

reasons, the sensor signal is strongly pH dependent in the physiological range. The

association time constant for calcium binding was strongly calcium dependent and

varied from 250 ms at low calcium concentration to 2.5 ms at higher concentra-

tions; the dissociation time constant was 200 ms. G-CaMP expresses poorly at

37 �C. G-CaMP-expressing smooth muscle showed a response to rapid depolari-

zation of around 50%, with a time course comparable to that previously measured

with Fluo-3. Carbachol addition gave a 2.5-fold increase. pH was monitored in

these experiments and did not change (Nakai et al., 2001).

This first GCaMP family member, later designated GCaMP1, had very weak

fluorescence when expressed at physiological temperatures compared to GFP

itself. This was addressed by introducing two mutations V163A and S175G that

were known to improve the temperature-dependent maturation of GFP to give a

variant known as G-CaMP1.6 (Ohkura et al., 2005); this increased brightness

about 40-fold. However, these modifications did not lead to adequate maturation

above 30 �C. The G-CaMP construct was subjected to error-prone PCR mutagen-

esis and the clones fluorescing most brightly at 37 �C were selected (Tallini et al.,

2006). The two new mutations in the brightest clone were identified (D180Y and

V93I), but it also turned out that the RSET leader sequence that had been added to

facilitate purification of the expressed protein was essential for thermal stability at

37 �C. This construct, GCaMP2, is around 200 times brighter than G-CaMP1 at

37 �C (with an extinction coeYcient at 487 nm of 19,000 and a quantum yield of

0.93 with emission at 508 nm) and shows the same four- to fivefold increase in

fluorescence a saturating calcium concentrations when compared to calcium-free

conditions. Though not reported, it should be assumed that this sensor remains

pH-sensitive. GCaMP2 was expressed using tissue-specific promoters in transgenic

animals and calcium transients were detected in granule cells in cerebellar slices

(Diez-Garcia et al., 2005) and in isolated heart in vitro and in adult and embryonic

heart in vivo (Tallini et al., 2006). Some insight into the sensor mechanism of

GCaMP2 is aVorded by its crystal structure (Akerboom et al., 2009; Wang

et al., 2008).

Even so, in HEK293 cells, GCaMP2 fluorescence is still 100-fold lower than

GFP itself (Tian et al., 2009). HEK293 cell medium-throughput screening assays


were used to identify brighter GCaMP2 mutants; attention was also paid to

improving the sensitivity to small calcium changes through mutations of the

CaM EF hands and of the M13/CaM interaction domains. The upshot was

GCaMP3, with a dynamic range of 12, due to a twofold decrease in calcium-free

fluorescence and a 1.5-fold increase in calcium-saturated fluorescence relative to

GCaMP2, and a K0d of 0.66 mM (Tian et al., 2009).

3. Cases 12 and 16

The Case (presumably Calcium sensor) constructs were developed by analyzing

the linker sequences between M13 and cpEYFP/GFP and cpEYFP/GFP and

calmodulin and the three key residues 148, 145, and 203 in the pericams and

G-CaMPs (Souslova et al., 2007). Based on this analysis, constructs were made

containing the G-CaMP linker sequences and the cpEYFP derived from Ratio-

metric Pericam. Nine point mutants were made with alterations in both the linker

sequences and in the three key residues within cpEYFP. As expected, combinations

of Asp148 and Phe203 produced ratiometric indicators akin to Ratiometric Peri-

cam, while Asn or Glu at residue 148 combined with Phe203 had a single excitation

peak at 490 nm. The Glu148/Thr145 and Glu148/S145 variants showed a 14.5-fold

increase in 490-nm fluorescence between calcium-free and calcium-bound forms.

The E148/S145 variant of these pericam-G-CaMP hybrids was optimized for

folding at 37 �C using error-prone PCR, resulting in a variant with a 12-fold

dynamic range named Case12. Substituting Thr for Ser at the 145 position of

Case12 gave Case16, with a 16.5-fold dynamic range. The apparent dissociation

constant for both Cases 12 and 16 was 1 mM. Like the pericams and G-CaMP

sensors, the calcium-bound forms of Cases 12 and 16 (pKa 7.2)—and thus their

fluorescence—are aVected by any changes in pH within the physiological range.

III. Applications of Genetically Encoded Sensors

A. Targeting to Subcellular Locations

Low molecular mass fluorescent calcium sensors do make their way to intracel-

lular compartments (Silver et al., 1992) and can be used to measure calcium there,

but they are diYcult to target precisely (Varadi and Rutter, 2002b). One of the two

major advantages of genetically encoded calcium sensors is that chimeric con-

structs and signaling tags can target them specifically to subcellular locations.

Methods to achieve some of these specific localizations had already been developed

for GFP itself and for the calcium sensor aequorin (De Giorgi et al., 1996). The

ability to target cameleons YC-3er and YC-4er was demonstrated in the study in

which cameleons were first described (Miyawaki et al., 1997).


1. Endoplasmic Reticulum

ER calcium concentrations have been measured using low molecular mass

calcium sensors and with aequorin (Solovyova and Verkhratsky, 2002), but it

seems fair to say that the cameleon-based sensors (YC-3er and YC-4er) have

given the best estimates of ER calcium concentration and turnover (Foyouzi-

Youssefi et al., 2000; Graves and Hinkle, 2003a,b; Varadi and Rutter, 2002a; Yu

and Hinkle, 2000). In summary, cameleon-based indicators have presented a

picture of the ER as an organelle with resting calcium concentrations in the

range 250–600 mM and a very active calcium turnover that depends very heavily

on the activity of the SERCA ATPase (Demaurex and Frieden, 2003). Transgenic

YC3.3er has been engineered to give tissue-specific expression in mouse pancreatic

beta cells (Hara et al., 2004). The interpretation of calcium changes in the ER

measured by cameleon indicators is tempered by the finding that pH changes

within the organelle may interfere with estimates of dynamic calcium concentra-

tion (Varadi and Rutter, 2004). Improved sensors for ER calcium are now avail-

able (Palmer Amy et al., 2004; Zou et al., 2007).

2. Mitochondria

Mitochondrial targeting of recombinant aequorin was achieved using the

N-terminal presequence of subunit VIII of cytochrome oxidase (Rizzuto et al.,

1992). The same targeting strategy was used to locate ratiometric pericam within

mitochondria (Robert et al., 2001) and to show that the pericam tracked beat to

beat calcium changes in cardiomyocytes, just as did aequorin. Cameleon probes

targeted to mitochondria were eVective only at very low expression levels

(Arnaudeau et al., 2001). In a comparison of mitochondrially targeted cameleon

(mtYC2), camgaroo-2, and Ratiometric Pericam (Nagai et al., 2001) in HeLa

cells, it was found that Ratiometric Pericam was the most reliable and faithful

of the sensors (Filippin et al., 2003). Mislocalization and poor expression of the

mitochondrially targeted YC2 sensor could be improved by inserting a tandem

repeat of the subunit VIII presequence as the targeting sequence (2mtYC2)

(Filippin et al., 2005). 2mtYC2 was used successfully to demonstrate calcium

handling by skeletal muscle mitochondria during contraction (Rudolf et al.,

2004). Insertion of a tandem targeting repeat was an ineVective strategy

for the preferred citrine or Venus variants (Filippin et al., 2005), but in contrast,

the D2cpv, D3cpv, and D4cpv cameleons (Palmer et al., 2006) functioned

well as mitochondrial calcium sensors when targeted with the cytochrome

oxidase tandem repeat (Palmer et al., 2006). These constructs are now the

recommended genetically encoded mitochondrial calcium sensors. An recent

overview of calcium sensor approaches in mitochondria is available (Pozzan

and Rudolf, 2009).


3. Peroxisome

Cameleon D3cpv was furnished with a modified peroxisome localization

sequence (D3cpv-KVK-SKL) to monitor calcium concentrations in this organelle

in HeLa cells in response to agonists or depolarization (Drago et al., 2008).

4. Golgi

The Citrine cameleon YC3.3 has been expressed in the Golgi using an 81 residue

N-terminal sequence from human galactosyl transferase type II (Griesbeck et al.,

2001); it was saturated, oVering no useful information but that the Golgi has a very

high resting calcium concentration.

5. Plasma Membrane

Sub-plasmalemmal calcium concentrations may diVer from those in bulk cyto-

plasm. Localized calcium concentrations around secretory vesicles were shown to

be higher than those in cytoplasm by using a phogrin chimera to target YC2 to

secretory vesicle membrane (Emmanouilidou et al., 1999). A number of targeting

strategies have proved successful in localizing sensors to the plasma membrane.

The cpVenus cameleon YC3.60 has been targeted using a Ki-Ras chimera (Nagai

et al., 2004). The TN-L15 sensor localized to the plasma membrane as GAP43,

Ras, or synaptobrevin chimeras (Heim and Griesbeck, 2004). Localization

can also be achieved with a myristoyl/palmitoyl N-terminal tag (Zacharias

et al., 2002), an approach that was used with the cameleon D series (Palmer

et al., 2006). A chimera of GCaMP2 and synaptotagmin (SyGGCamp2) has been

used to monitor synaptic calcium signals, in this case in vivo in zebrafish (Dreosti

et al., 2009).

B. Tissue-Specific Expression

The other major advantage of genetically encoded calcium sensors is tissue-

specific expression in intact organisms.

1. YC2.1

The first transgenic tissue-specific expression of genetically encoded calcium

sensors was demonstrated in plants. YC2.1 was expressed in Arabidopsis guard

cells of the leaf, first using a CaMV promoter (Allen et al., 1999) and then a guard

cell-specific det promoter (Allen et al., 2000), demonstrating that aspects of the

calcium-signaling response in guard cells were under diVerential genetic control.

YC3.1 was used in transgenic Aradidopsis plants to visualize calcium signals in the

pollen grain (Iwano et al., 2004).


YC2.1 was expressed transgenically in Caenorhabditis elegans pharyngeal

muscle under the control of the pharyngeal-specific myo-2 promoter (Kerr

et al., 2000) and tracked calcium changes during pharyngeal pumping;

YC3.1 tracked temporal changes more faithfully than YC2.1, being the

faster sensor, but YC2.1 tracked calcium changes to basal level more faith-

fully than YC3.1, as might be expected from its lower K0d. Expression of

YC2.12 in C. elegans touch neurons under the control of the mec-4 promot-

er identified a role for specific ion channels in the touch response (Suzuki

et al., 2003).

The UAS/Gal4 tissue-specific expression system was used to express YC2.1 in a

subset of the antennal lobe projection neurones of Drosophila in order to study

odorant responses in the antennal lobe and mushroom body calyx in vivo

(Diegelmann et al., 2002; Fiala et al., 2002). Odorant-specific patterns of neuronal

excitation were seen in both the antennal lobe and the calyx. In the former, the

EYFP/ECFP emission ratio changes were 1.23�0.23% (mean and sem) and in the

latter 0.6�0.06%. In the antennal lobe, the changes in sensor signal were observed

in spatially restricted regions of around 10–30 mm diameter, the size of individual

glomeruli. These very small changes were nonetheless reproducible, with distinct

and reproducible patterns of activity from fly to fly associated with diVerentodorants.

The same UAS/Gal 4 technology was used to express YC2 in neurones of

larval Drosophila (ReiV et al., 2002) to the evolution of calcium signaling in

presynaptic terminals innervating larval muscle. A 28% emission ratio change

was measured in vivo during spike train stimulation of the neuromuscular junc-

tion and signals of this magnitude could be resolved in single synaptic boutons;

there were no detectable diVerences in neuromuscular junction physiology

between wild-type and transgenic larvae. This study illustrates the point that

targeted expression of genetically encoded sensors in individual neurones is for

some applications superior to the use of low molecular mass synthetic calcium

indicators, as the specificity of expression more than compensates for the loss of

brightness.

In a similarly mature use of YC2.1 sensor technology coupled to UAS/Gal4

transgenic expression, neuronal calcium measurement coupled with electrophysi-

ology was used to identify thermosensory neurones in the larval nervous system

in vivo (Liu et al., 2003). Changes in emission ratio of 10–50% were associated with

heating and cooling. A functional map of thermosensory neurones was generated

and it was found that neurones with diVerent temperature responses were anato-

mically segregated.

YC2.1 was also used in zebrafish to record the behavior of Rohon-Beard (RB)

neurones during the fish’s escape response (Higashijima et al., 2003). This careful

study started with transient expression of the YC2.1 transgene in the RB neurones to

show proof of principle before generating transgenic lines in which the calcium signals

in the RB neurons could be correlated with the escape response in conscious fish.


2. YC3.3er

YC3.3er (the citrine-based sensor) was expressed in the beta cells of transgenic

mice under the control of the mouse Insulin 1 promoter (Hara et al., 2004). The

sensor signal could be detected in isolated pancreatic islets and addition of thapsi-

gargin or carbachol gave the expected decrease in the 535/485 emission ratio.

3. Camgaroos and Inverse Pericam

UAS/Gal 4 expression was used to create transgenic Drosophila that expressed

camgaroos-1 and-2 in the mushroom bodies of adult brain (Yu et al., 2003).

Dissected fly brains were used. Camgaroo-2 fluorescence in the mushroom bodies

was much more intense than that of camgaroo-1, but the camgaroo-1 emission

ratio signal on potassium depolarization was more than double that of camgaroo-2

(38% vs. 14% in the mushroom body lobe and 83% vs. 28% in the mushroom body

itself ). It was shown that these increases were not due to changes in pH. Applica-

tion of the putative mushroom body transmitter, acetylcholine, causes ratio

changes of a few percent. In this setting, camgaroo-2, although brighter, showed

substantially lower ratio changes than camgaroo-1; it also underwent significantly

faster photobleaching.

Inverse pericam is an intensity-coded sensor that decreases its fluorescence as

calcium increases. Addition of DsRed2 to the C-terminal of inverse pericam

produces a ratiometric indicator whose 615/510 nm emission ratio increases as

calcium increases. This indicator (DsRed2-referenced inverse pericam (DRIP))

requires dual excitation and dual emission optics (Shimozono et al., 2004). The

DsRed2 fluorescence is a passive, calcium-independent signal that is proportional

to the concentration of the sensor and helps control for alterations in overall

fluorescence intensity due for example to movement artifacts. DRIP was expressed

transgenically in worms under the control of the myo 2 promoter that is specific for

pharyngeal muscle. Ratio changes of 30–40% were measured in worms undergoing

fast pharyngeal pumping.

After screening six sensors (flash pericam, inverse pericam,G-CaMP, camgaroo-2,

YC2.12, andYC3.12) for calcium sensitivity in stably transfected fibroblast cell lines,

the two with the greatest dynamic range (inverse pericam: �40% and camgaroo-2:

þ170%), together withYC3.12 that gave inconclusive results in the fibroblast expres-

sion screen but is optimized for expression at 37 �C, were used to generate transgenic

mice under the control of the TET expression system (Hasan et al., 2004); the TET

system allows tissue-specific expression by crossing the TET mice with mice expres-

sing the TET transactivator under tissue-specific control. TET sensor mice were

crossed with a line expressing the transactivator under the control of the alpha-

calmodulin/calcium dependent kinase II (aCamKII) promoter. All mice developed

normally. Five highly expressing lines were obtained out of 36 transgenic lines: two

YC3.12, two camgaroo-2, and one inverse pericam. Expression patterns in brain

slices and excised retina were analyzed by two-photonmicroscopy. They appeared to


be mosaic, not mapping directly to the known patterns of aCamKII expression.

Neocortical expression could also be imaged through the thinned skull in anaesthe-

tized mice. Two photon fluoresence recovery after photobleaching suggested that as

much as half the fluorescence signal was immobile and this together with punctuate

staining patterns suggested that this immobile sensor fraction might be due to

interaction between the M13 and CaM moieties of the sensors and their normal

cellular targets. Cellular and synaptic stimulation of pyramidal neurones in

cortical slices using sharp and patch microelectroded gave 5–10% increases

in 535 nm fluorescence using wide field imaging and around 20–100% for cam-

garoo-2 and �30% for inverse pericam using two photon imaging. In the retina, a

ganglion cell subset was strongly labeled in YC3.1-expressing mice, but no light-

evoked responses were detected. In camgaroo-2 expressing lines, bleaching occurred

in the retina too quickly for measurements to be made. In one inverse pericam-

expressing mouse, 7 of 12 ganglion cells tested showed a transient decrease in

fluorescence attributable to a calcium increase in response to light. Sensors were

imaged in the olfactory bulb in vivo using wide fieldmicroscopy. Camgaroo-2 expres-

sing mice showed a 1–3% increase in response to odors, while inverse pericam gave

�8% decrease. Each distinct odor evoked a unique pattern of activity, similar odors

evoking similar patterns.

This thoughtful study established four main facts: around half of the transgeni-

cally expressed cameleon family sensor was immobile; this reduced sensitivity and

made quantitation of the calcium signals impossible; nonetheless, it was possible to

observe patterns of neuronal activity; YC3.12 was not an eVective transgenic

sensor. The study also reports unpublished experiments in which transgenic mice

expressing YC3.0 under the control of a b-actin promoter gave only 1–2% ratio

changes during wide filed imaging in cerebellar slices. The high proportion of

immobile sensor in transgenic animals remains for the moment inexplicable—it

was not seen in the stably transfected fibroblast lines.

4. GCaMP

G-CaMP (Nakai et al., 2001) was expressed in mice under the control of a

smooth muscle myosin heavy chain promoter and was expressed in vascular and

nonvascular smooth muscle (Ji et al., 2004). The signatures of inotropic (ion

channel) and metabotropic (InsP3-mediated) postsynaptic signaling could be dis-

tinguished in single excised smooth muscle cells.

In a set of experiments strikingly parallel to those with YC2.1 (Diegelmann

et al., 2002; Fiala et al., 2002), but using two photon imaging, G-CaMP was

expressed in a subset of projection neurones in Drosophila antennal lobe to

demonstrate that diVerent odorants activated specific patterns of glomeruli

(Wang et al., 2003). Individual glomeruli are diVerentially sensitive to a given

odorant and more are recruited as the odorant concentration is increased.

Increases of fluorescence of up to 50% (at 525 nm) were measured in responsive

glomeruli.


Transgenic expression of GCaMP2 has been achieved in mouse heart (Tallini

et al., 2006). The TET system was used: the GCAMP2 sequence was placed

downstream of a weakened a myosin heavy chain promoter (aMHC) and seven

tetO enhancer sequences to permit suppression of gene expression using doxycy-

cline. These mice were crossed with others with a hemizygous aMHC-tetracycline

transactivator allele. The doubly transgenic mice expressed GCaMP2 only in the

heart. Doxycycline suppression of the transgene was essential, as mice constitu-

tively expressing GCaMP2 from birth showed markedly enlarged hearts, a pheno-

type comparable to that seen in mice overexpressing calmodulin. This phenotype

was avoided entirely by administering doxycycline in utero and until 13–15 weeks

postpartum. Subsequent removal of doxycycline for up to 6 weeks led to no

detectable cardiomegaly. Robust GCaMP signals were present 4 weeks after

doxycycline removal.

Striking wide field fluorescence images of cardiac calcium transients in whole

mouse heart beating at up to 300 beats/min were obtained with anaesthetized,

ventilated open-chested mice, the first to be recorded under wholly physiological

conditions with the heart under normal load. As expected sympathetic stimulation

with isoproterenol markedly increased the calcium signal and also increased end

diastolic calcium concentration. Signal-to-noise ratios were good and it was possi-

ble to record very clean signals from a single pixel of the 100�100 pixel imaging

array (tens of microns). Using a photodiode array in isolated perfused heart,

signals from a membrane potential sensitive dye and from GCaMP2 were acquired

simultaneously. Association and dissociation kinetics of calcium were rapid (t¼14

and 75 ms, respectively) and unaltered in vivo. Comparison with a fast calcium dye

Rhod2 nonetheless showed that the rise and decay times of the GCaMP2 signal in

beating heart was around 45% slower, but with a three times greater dynamic

range. Calcium sparks could not be observed in isolated ventricular myocytes

expressing GCaMP2. GCaMP2 imaging in open-chested embros from embryonic

day 10 allowed the analysis of the development of the atrio-ventricular node

conduction pathway.

GCaMP2 fused to synaptotagmin localizes to synaptic boutons. It reports the

location of synapses in zebrafish in vivo and shows a linear response over a wide

range of action potential frequencies (Dreosti et al., 2009). It can report spiking

frequencies in optic tectum; it also reports activity in the graded synapses of retinal

bipolar cells. GCaMP2 has also been used to map functional connections in the C.

elegans nervous system (Guo et al., 2009). Connections can be mapped grossly, but

the sensor’s signals are too weak to distinguish direct from indirect connections.

5. TN-L15, TN-XL, and TN-XXL

Acerulean version of TN-L15, cerTN-L15, was used to create a transgenicmouse

line that expressed the sensor widely in brain, especially in the neocortex and

hippocampus (Heim et al., 2007). Calcium changes resulting from two to three

action potentials could be resolved and calcium responses in spiny dendrites of


pyramidal cells could be detected after puYng on glutamate, an excitatory neuro-

transmitter (Garaschuk et al., 2007; Heim et al., 2007). TN-XLwas expressed using

the UAS/Gal4 tissue-specific expression system in Drosophila neuromuscular

junction (Mank et al., 2006). Its rapid oV rate for calciummade it significantly better

at tracking calcium changes than its counterparts. TN-XXL showed improved

sensitivity and long term-stability in sensing calcium signals from fly neurones; in

mice, tuning curves for orientation-specific neurones in visual cortex could be

monitored repeatedly over timescales of days or weeks (Mank et al., 2008).

6. Comparison of the Performance of Genetically Encoded Calcium Sensors

Though progress in the field has been periodically reviewed (Barth, 2007;

Garaschuk et al., 2007; Griesbeck, 2004; Mank and Griesbeck, 2008), few studies

have systematically compared the performance of diVerent genetically encoded

calcium sensors, except to demonstrate the superiority of a novel sensor over its

predecessors. I have discussed above (Section III.B.3) the systematic comparisons

of camgaroo-1 and-2 when expressed in Drosophila mushroom bodies (Yu et al.,

2003) and of inverse pericam, camgaroo-2 and YC3.1 when expressed in mouse

brain (Hasan et al., 2004).

The performance of GCaMP, inverse pericam, and camgaroo-2 was compared

with that of the low molecular mass synthetic indicators X-Rhod-5F and Fluo4-

FF in apical dendrites of pyramidal cells in hippocampal brain slices from 6- to 7-

day-old rats transfected using a biolistic approach and maintained at room

temperature (Pologruto et al., 2004). Images were obtained using two-photon

microscopy. Action potentials were triggered using current injection into the cell

body. Under these conditions, X-Rhod-5F and Fluo4-FF could detect calcium

changes (signal twice that of noise) in the dendrite due to voltage-dependent

calcium channel activation after single action potentials while with the same

criterion GCaMP required five action potentials, camgaroo-2, 33 action potentials,

and inverse pericam over 20. For comparison, the dynamic ranges (DF/F) for thethree sensors under these conditions in vitro was1.8, �2, and �0.25, so the

sensitivity of camgaroo-2 was poor despite its larger dynamic range. Power spec-

trum analysis was used to analyze the fluorescence response during action potential

trains at 20 Hz. Most of the power in the frequency analysis of X-Rhod-5F and

Fluo4-FF fluorescence was at the fundamental frequency, 20 Hz, indicating that

the fluorescence signal tracked each action potential. For the genetically encoded

sensors, no clear peak was observed at 20 Hz, indicating that the sensors were too

slow to track individual action potentials at this stimulation frequency.

It was possible to measure calcium activation curves in situ for the three sensors

and thus their apparent dissociation constants by simultaneously measuring

calcium concentration using a calibrated X-Rhod-5F signal and the fluorescence

signal from the sensor at various levels of stimulation. For inverse pericam (K0d

0.9 mM) and camgaroo-2 (K0d 8 mM), these were comparable to those previously

reported in vitro; however, GCaMP showed a K0d (1.7 mM) almost an order of


magnitude greater than that previously reported in vitro (Nakai et al., 2001).

Because the calcium concentration profile of dendritic action potentials is well

characterized (Pologruto et al., 2004), there seems little doubt that the calcium

dissociation characteristics of GCaMP vary markedly in vitro and in vivo. FRAP

studies in dendrites showed that the all three sensors were mobile, with mobilities

comparable to GFP itself. This result is quite firmly at odds with that reported in

mouse brain (Hasan et al., 2004) and discussed above (Section III.B.3). It may be

that dendrites, being relatively free of organelles, mirror better the behavior of the

sensors in cytoplasm than cell bodies; it should be noted that punctuate staining

was reported in mouse brain (Hasan et al., 2004). It should also be borne in mind

that though the observations on mouse brain slices were carried out at room

temperature, as were these experiments in rat brain slices, in the mouse study,

the sensors had been expressed at body temperature, whereas the biolistically

transfected rat brain slices were maintained throughout at room temperature.

These data, as the authors point out (Pologruto et al., 2004), demonstrate that

the genetically encoded sensors are better-suited to measuring summated neuronal

responses after multiple stimuli, not single action potentials, consistent with their

reported use to monitor patterns of neuronal activity (Fiala et al., 2002; Hasan

et al., 2004; Wang et al., 2003); as it happens, these three studies all described

odorant-specific patterns of neuronal signaling.

As an addendum to the study, Svoboda’s group also provided in vitro solution

X-ray scattering evidence that showed that the calcium-dependent fluorescent

signal of GCaMP, as theorized, depends on a coupled structural change in which

calcium binding to CaM is closely linked to binding of CaM to M13; in contrast,

the calcium-dependent fluorescence signal in camgaroo-2 is solely due to binding

to CaM, the M13 peptide paradoxically playing no part in the sensor response

(Pologruto et al., 2004).

A second comparative study was undertaken at the Drosophila larva neuromus-

cular junction (ReiV et al., 2005), using an approach previously reported (ReiVet al., 2002). The responses of 10 sensors from the three families to 40 and 80 Hz

stimulation of the synaptic bouton were compared. Camgaroos-1 and-2 and flash

pericam did not sense calcium changes in the bouton. YC2.0, 2.3, 3.3, TN-L15,

inverse pericam, and GCaMP1.3 and 1.6 all showed adequate responses (around

5% on average at 40 Hz and 10–15% at 80 Hz) to pulse train stimuli, but none

exhibited dynamic ranges anywhere near comparable to those measured in vitro

(ReiV et al., 2005). None was comparable in performance in this system when

compared to the later developed TN-XL (Mank et al., 2006). In an echo of the

work in rat brain slices, the performance of YC3.3, TN-L15, GCaMP1.6,

GCaMP2, YC2.60, YC3.60, cameleon D3, and TN-XL were compared one with

another and calibrated against a low molecular mass indicator, Oregon-Green-

BAPTA-1 (Hendel et al., 2008). The latter four sensors were around twofold more

responsive than their earlier counterparts. None of the sensors were seen to detect

single action potentials, though YC3.60 and cameleon D3 could detect two action

potentials in succession. None showed the fast temporal response of the low


molecular mass indicator. A theoretical framework in which to consider the pros

and cons of calcium sensors in recording neuronal activity has been adumbrated

(Hires et al., 2008).

GCaMP1.6 and GCaMP2 were compared in pyramidal cells dendrites in

mammalian brain slices transfected ballistically or by electroporation (Mao

et al., 2008) under conditions that allowed comparison with first generation

sensors (Pologruto et al., 2004). Their performance was not significantly better

than GCaMP, even when localized using membrane and cytoskeletal targeting

chimeras (Mao et al., 2008). GCaMP3, however, showed substantial gains in

sensitivity and discrimination (Tian et al., 2009): overall, the signal-to-noise

ratio was much improved and responses in dendrites to single action potentials

could be reliably detected. Direct comparison with TN-XXL and cameleon D3

showed that, although brighter, the two FRET sensors gave smaller fluorescence

changes and less favorable signal-to-noise ratios. GCaMP3 was also more photo-

stable. After either adenoviral transfection or in utero electroporation, calcium

responses in pyramidal neurones could be observed in awake, behaving mice

(Tian et al., 2009). Parallel electrical recordings showed that detectable calcium

responses were associated with three or more action potentials. Calcium responses

were also readily observed in the glomeruli of Drosophila antennal lobe and in

sensory neurones of C. elegans, altogether a methodological tour de force (Tian

et al., 2009).

IV. Use of Genetically Encoded Calcium Sensors

For single cell applications, wide-field fluorescence imaging, spinning disk, or

confocal microscopy are appropriate methods. Dual excitation laser scanning

confocal imging is achievable (Shimozono et al., 2002). For whole animal applica-

tions, particularly in intact brain or brain slices two photon microscopy is recom-

mended, as it reduces tissue damage and oVers improved imaging within tissue (see

Chapter 9; Fan et al., 1999).

Expression of sensors in cells and tissues, as we have seen, can be achieved by

transfection and transgenesis. One advantage of transgenic approaches is that

expression can be confined to a specific tissue or cell type, an advantage even if it

is excised for imaging. Random expression in a subset of cells can more simply be

achieved by using biolistic transfection of excised tissue.

Ratiometric sensors (in this context the FRET-based sensors, ratiometric peri-

cam and DRIP) oVer the advantage that the quantitative signal is in theory

independent of variations in sensor distribution and concentration within cells

(Silver et al., 1992). This allows reliable calibration of the signal in terms of calcium

concentration (see chapter 1). Nonratiometric sensors (e.g., GCaMP3) are ade-

quate for determining changes in calcium concentration, for example, when mea-

suring overall spatial and temporal patterns of calcium signaling. Even in these

circumstances, caution should be exercised in case the responses are nonlinear,


especially at low calcium concentration, so that a subset of smaller signals is

overlooked (ReiV et al., 2005).

In general, genetically encoded calcium sensors are not available commercially,

though Invitrogen oVers YC3.60 (http://probes.invitrogen.com/media/pis/

mp36207.pdf). Some can be obtained for noncommercial use from their creators

(http://www.tsienlab.ucsd.edu/ and http://cfds.brain.riken.jp/). Or you can make

your own using the handbook (Miyawaki et al., 2003a, 2005).

V. Conclusions

Genetically encoded calcium sensors have proved valuable in the measurement

of calcium concentration in cellular organelles, for the most part in single cells

in vitro. Their success as sensors in tissues in vitro and in vivo is qualified. They have

also proved valuable in imaging the pattern of calcium signals within tissues,

particularly in the poikilotherms, C. elegans, Drosophila, and zebrafish. In home-

otherms, the record is largely disappointing, even when tissue is excised and

monitored at room temperature (Pologruto et al., 2004). Striking exceptions are

the use of GCaMP2 to image calcium-signaling patterns in mouse heart (Tallini

et al., 2006) and pyramidal neurones (Tian et al., 2009). For the most part, sensors

are still not capable of sensing individual calcium events in single cells when these

cells are part of tissue, though single cell responses can be monitored in disaggre-

gated cells (KotlikoV, 2007). Some branches of the calcium sensor evolutionary

tree continue to evolve rapidly and the steady progress in optimizing sensor

parameters leads to the certain hope that these drawbacks will eventually be

overcome by further genetic engineering.

Acknowledgments

I thank Jill McKenna for helping with this chapter. Our work is supported by grants from the

Wellcome Trust.

References

Akerboom, J., Rivera Jonathan, D. V., Guilbe Maria, M. R., Malave Elisa, C. A., Hernandez

Hector, H., Tian, L., Hires, S. A., Marvin Jonathan, S., Looger Loren, L., and Schreiter Eric, R.

(2009). Crystal structures of the GCaMP calcium sensor reveal the mechanism of fluorescence signal

change and aid rational design. J. Biol. Chem. 284, 6455–6464.

Allen, G. J., Kwak, J. M., Chu, S. P., Llopis, J., Tsien, R. Y., Harper, J. F., and Schroeder, J. I. (1999).

Cameleon calcium indicator reports cytoplasmic calcium dynamics in Arabidopsis guard cells. Plant

J. 19, 735–747.

Allen, G. J., Chu, S. P., Schumacher, K., Shimazaki, C. T., Vafeados, D., Kemper, A., Hawke, S. D.,

Tallman, G., Tsien, R. Y., Harper, J. F., Chory, J., and Schroeder, J. I. (2000). Alteration of stimulus-

specific guard cell calcium oscillations and stomatal closing in Arabidopsis det3 mutant. Science 289,

2338–2342.

http://probes.invitrogen.com/media/pis/mp36207.pdf

http://probes.invitrogen.com/media/pis/mp36207.pdf

http://www.tsienlab.ucsd.edu/

http://cfds.brain.riken.jp/


Arnaudeau, S., Kelley, W. L., Walsh, J. V., and Demaurex, N. (2001). Mitochondria recycle Ca(2þ) to

the endoplasmic reticulum and prevent the depletion of neighboring endoplasmic reticulum regions.

J. Biol. Chem. 276, 29430–29439.

Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (1999). Circular permutation and receptor insertion

within green fluorescent proteins. Proc. Natl. Acad. Sci. USA 96, 11241–11246.

Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2000). Biochemistry, mutagenesis, and oligomerization

of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97, 11984–11989.

Baker, P. F. (1978). The regulation of intracellular calcium in giant axons of Loligo andMyxicola. Ann.

NY Acad. Sci. 307, 250–268.

Barth, A. L. (2007). Visualizing circuits and systems using transgenic reporters of neural activity. Curr.

Opin. Neurobiol. 17, 567–571.

De Giorgi, F., Brini, M., Bastianutto, C., Marsault, R., Montero, M., Pizzo, P., Rossi, R., and

Rizzuto, R. (1996). Targeting aequorin and green fluorescent protein to intracellular organelles.

Gene 173, 113–117.

Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted

excitation mutants of the green fluorescent protein. Nat. Biotechnol. 13, 151–154.

Demaurex, N., and Frieden, M. (2003). Measurements of the free luminal ER Ca(2þ) concentration

with targeted ‘‘cameleon’’ fluorescent proteins. Cell Calcium 34, 109–119.

Diegelmann, S., Fiala, A., Leibold, C., Spall, T., and Buchner, E. (2002). Transgenic flies expressing the

fluorescence calcium sensor Cameleon 2.1 under UAS control. Genesis 34, 95–98.

Diez-Garcia, J., Matsushita, S., Mutoh, H., Nakai, J., Ohkura, M., Yokoyama, J., Dimitrov, D., and

Knopfel, T. (2005). Activation of cerebellar parallel fibers monitored in transgenic mice expressing a

fluorescent Ca2þ indicator protein. Eur. J. Neurosci. 22, 627–635.

Drago, I., Giacomello, M., Pizzo, P., and Pozzan, T. (2008). Calcium dynamics in the peroxisomal

lumen of living cells. J. Biol. Chem. 283, 14384–14390.

Dreosti, E., Odermatt, B., Dorostkar Mario, M., and Lagnado, L. (2009). A genetically encoded

reporter of synaptic activity in vivo. Nat. Methods 6, 883–889.

Emmanouilidou, E., Teschemacher, A. G., Pouli, A. E., Nicholls, L. I., Seward, E. P., and Rutter, G. A.

(1999). Imaging Ca2þ concentration changes at the secretory vesicle surface with a recombinant

targeted cameleon. Curr. Biol. 9, 915–918.

Evanko, D. S., and Haydon, P. G. (2005). Elimination of environmental sensitivity in a cameleon

FRET-based calcium sensor via replacement of the acceptor with Venus. Cell Calcium 37, 341–348.

Fan, G. Y., Fujisaki, H., Miyawaki, A., Tsay, R. K., Tsien, R. Y., and Ellisman, M. H. (1999). Video-

rate scanning two-photon excitation fluorescence microscopy and ratio imaging with cameleons.

Biophys. J. 76, 2412–2420.

Fiala, A., Spall, T., Diegelmann, S., Eisermann, B., Sachse, S., Devaud, J.-M., Buchner, E., and

Galizia, C. G. (2002). Genetically expressed cameleon in Drosophila melanogaster is used to visualize

olfactory information in projection neurons. Curr. Biol. 12, 1877–1884.

Filippin, L., Magalhaes Paulo, J., Di Benedetto, G., Colella, M., and Pozzan, T. (2003). Stable

interactions between mitochondria and endoplasmic reticulum allow rapid accumulation of calcium

in a subpopulation of mitochondria. J. Biol. Chem. 278, 39224–39234.

Filippin, L., AbadMaria, C., Gastaldello, S., Magalhaes Paulo, J., Sandona, D., and Pozzan, T. (2005).

Improved strategies for the delivery of GFP-based Ca2þ sensors into the mitochondrial matrix. Cell

Calcium 37, 129–136.

Foyouzi-Youssefi, R., Arnaudeau, S., Borner, C., Kelley,W. L., Tschopp, J., Lew, D. P., Demaurex, N.,

and Krause, K. H. (2000). Bcl-2 decreases the free Ca2þ concentration within the endoplasmic

reticulum. Proc. Natl. Acad. Sci. USA 97, 5723–5728.

Garaschuk, O., Milos, R.-I., Grienberger, C., Marandi, N., Adelsberger, H., and Konnerth, A. (2006).

Optical monitoring of brain function in vivo: From neurons to networks. Pflugers Arch. 453,

385–396.

Garaschuk, O., Griesbeck, O., and Konnerth, A. (2007). Troponin C-based biosensors: A new family of

genetically encoded indicators for in vivo calcium imaging. Cell Calcium 42, 351–361.


Gee, K. R., Brown, K. A. Chen, W.-N. U., Bishop-Stewart, J., Gray, D., and Johnson, I. (2000).

Chemical and physiological characterization of fluo-4 Ca2+-indicator dyes. Cell Calcium 27, 97–106.

Gilkey, J. C., JaVe, L. F., Ridgway, E. B., and Reynolds, G. T. (1978). A free calcium wave traverses the

activating egg of the medaka, Oryzias latipes. J. Cell Biol. 76, 448–466.

Graves, T. K., and Hinkle, P. M. (2003a). Ca(2þ)-induced Ca(2þ) release in the pancreatic beta-cell:

Direct evidence of endoplasmic reticulum Ca(2þ) release. Endocrinology 144, 3565–3574.

Graves, T. K., and Hinkle, P. M. (2003b). Endoplasmic reticulum calcium storage and release in cells

expressing misfolded growth hormone. Growth Horm. IGF Res. 13, 36–43.

Griesbeck, O. (2004). Fluorescent proteins as sensors for cellular functions. Curr. Opin. Neurobiol. 14,

636–641.

Griesbeck, O., Baird, G. S., Campbell, R. E., Zacharias, D. A., and Tsien, R. Y. (2001). Reducing the

environmental sensitivity of yellow fluorescent protein. Mechanism and applications. J. Biol. Chem.

276, 29188–29194.

Grinkiewicz, G., Poenie, M., and Tsien, R.Y. (1985). A new generation of Ca2þ indicator with improved

fluorescence properties. J. Biol. Chem. 260, 3440–3450.

Guo, Z. V., Hart, A. C., and Ramanathan, S. (2009). Optical interogation of neural circuits in

Caenorhabditis elegans. Nat. Methods 6, 891–896.

Habuchi, S., Cotlet, M., Hofkens, J., Dirix, G., Michiels, J., Vanderleyden, J., Subramaniam, V., and

De Schryver, F. C. (2002). Resonance energy transfer in a calcium concentration-dependent cameleon

protein. Biophys. J. 83, 3499–3506.

Hara, M., Bindokas, V., Lopez James, P., Kaihara, K., Landa Luis, R., Harbeck, M., and Roe, M. W.

(2004). Imaging endoplasmic reticulum calcium with a fluorescent biosensor in transgenic mice. Am.

J. Physiol. Cell Physiol. 287, C932–C938.

Hasan, M. T., Friedrich, R. W., Euler, T., Larkum, M. E., Giese, G., Both, M., Duebel, J., Waters, J.,

Bujard, H., Griesbeck, O., Tsien, R. Y., Nagai, T., et al. (2004). Functional fluorescent Ca2þindicator proteins in transgenic mice under TET control. PLoS Biol. 2, e163.

Heim, N., and Griesbeck, O. (2004). Genetically encoded indicators of cellular calcium dynamics based

on troponin C and green fluorescent protein. J. Biol. Chem. 279, 14280–14286.

Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature 373, 663–664.

Heim, N., Garaschuk, O., Friedrich, M.W., Mank,M., Milos, R.-I., Kovalchuk, Y., Konnerth, A., and

Griesbeck, O. (2007). Improved calcium imaging in transgenic mice expressing a troponin-C based

biosensor. Nat. Methods 4, 127–129.

Hendel, T., Mank, M., Schnell, B., Griesbeck, O., Borst, A., and ReiV, D. F. (2008). Fluorescence

changes of genetic calcium indicators and OGB-1 correlated with neural activity and calcium in vivo

and in vitro. J. Neurosci. 28, 7399–7411.

Higashijima, S.-i., Masino Mark, A., Mandel, G., and Fetcho, J. R. (2003). Imaging neuronal activity

during zebrafish behavior with a genetically encoded calcium indicator. J. Neurophysiol. 90,

3986–3997.

Hires, S. A., Tian, L., and Looger, L. L. (2008). Reporting neural activity with genetically encoded

calcium indicators. Brain Cell Biol. 36, 69–86.

Ikura, M., Clore, G. M., Gronenborn, A. M., Zhu, G., Klee, C. B., and Bax, A. (1992). Solution

structure of a calmodulin-target peptide complex by multidimensional NMR. Science 256,

632–638.

Iwano, M., Shiba, H., Miwa, T., Che, F.-S., Takayama, S., Nagai, T., Miyawaki, A., and Isogai, A.

(2004). Ca2þ dynamics in a pollen grain and papilla cell during pollination of Arabidopsis. Plant

Physiol. 136, 3562–3571.

Jares-Erijman, E. A., and Jovin, T. M. (2003). FRET imaging. Nat. Biotechnol. 21, 1387–1395.

Ji, G., Feldman,M. E., Deng, K.-Y., Greene, K. S., Wilson, J., Lee, J. C., Johnston, R. C., Rishniw,M.,

Tallini, Y., Zhang, J., Wier, W. G., Blaustein, M. P., et al. (2004). Ca2þ-sensing transgenic mice:

Postsynaptic signaling in smooth muscle. J. Biol. Chem. 279, 21461–21468.

Kao, J. P., Harootunian, A. T., and Tsien, R. Y. (1989). Photochemically generated cytosolic calcium



Kerr, R., Lev-Ram, V., Baird, G., Vincent, P., Tsien, R. Y., and Schafer, W. R. (2000). Optical imaging

of calcium transients in neurons and pharyngeal muscle of C. elegans. Neuron 26, 583–594.

KotlikoV, M. I. (2007). Genetically encoded Ca2þ indicators: Using genetics and molecular design to

understand complex physiology. J. Physiol. 578, 55–67.

Liu, L., Yermolaieva, O., Johnson, W. A., Abboud, F. M., and Welsh, M. J. (2003). Identification and

function of thermosensory neurons in Drosophila larvae. Nat. Neurosci. 6, 267–273.

Mank, M., and Griesbeck, O. (2008). Genetically encoded calcium indicators. Chem. Rev. 108,

1550–1564.

Mank,M., ReiV, D. F., Heim,N., Friedrich,M.W., Borst, A., andGriesbeck, O. (2006). A FRET-based

calcium biosensor with fast signal kinetics and high fluorescence change. Biophys. J. 90, 1790–1796.

Mank, M., Santos, A. F., Direnberger, S., Mrsic-Flogel, T. D., Hofer, S. B., Stein, V., Hendel, T., ReiV

Dierk, F., Levelt, C., Borst, A., BohoeVer, T., Hubener, M., et al. (2008). A genetically encoded

calcium indicator for chronic in vivo two-photon imaging. Nat. Methods 5, 805–811.

Mao, T., O’Connor, D. H., Scheuss, V., Nakai, J., and Svoboda, K. (2008). Characterization and

subcellular targeting of GCaMP-type genetically-encoded calcium indicators. PloS ONE 3, e1796.

Minta, A., Kao, J. P., and Tsien, R. Y.(1989). Fluorescent indicators for cytosolic calcium based on

rhodamine and fluorescein chromophores. J. Biol. Chem. 264, 8171–8178.

Miyawaki, A. (2003). Fluorescence imaging of physiological activity in complex systems using

GFP-based probes. Curr. Opin. Neurobiol. 13, 591–596.

Miyawaki, A. (2003). Visualization of the spatial and temporal dynamics of intracellular signaling.Dev.

Cell 4, 295–305.

Miyawaki, A. (2005). Innovations in the imaging of brain functions using fluorescent proteins. Neuron

48, 189–199.

Miyawaki, A., Llopis, J., Heim, R., McCaVery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997).

Fluorescent indicators for Ca2þ based on green fluorescent proteins and calmodulin. Nature 388,

882–887.

Miyawaki, A., Griesbeck, O., Heim, R., and Tsien, R. Y. (1999). Dynamic and quantitative Ca2þmeasurements using improved cameleons. Proc. Natl. Acad. Sci. USA 96, 2135–2140.

Miyawaki, A., Nagai, T., and Mizuno, H. (2003). Mechanisms of protein fluorophore formation and

engineering. Curr. Opin. Chem. Biol. 7, 557–562.

Miyawaki, A., Sawano, A., and Kogure, T. (2003). Lighting up cells: Labelling proteins with fluor-

ophores. Nat. Cell Biol. (Sept. Suppl.), S1–S7.

Miyawaki, A., Nagai, T., and Mizuno, H. (2005). Engineering fluorescent proteins. Adv. Biochem. Eng.

Biotechnol. 95, 1–15.

Nagai, T., Sawano, A., Park, E. S., and Miyawaki, A. (2001). Circularly permuted green fluorescent

proteins engineered to sense Ca2þ. Proc. Natl. Acad. Sci. USA 98, 3197–3202.

Nagai, T., Ibata, K., Park Eun, S., Kubota, M., Mikoshiba, K., and Miyawaki, A. (2002). A variant of

yellow fluorescent protein with fast and eYcient maturation for cell-biological applications. Nat.

Biotechnol. 20, 87–90.

Nagai, T., Yamada, S., Tominaga, T., Ichikawa, M., and Miyawaki, A. (2004). Expanded dynamic

range of fluorescent indicators for Ca(2þ) by circularly permuted yellow fluorescent proteins. Proc.

Natl. Acad. Sci. USA 101, 10554–10559.

Nakai, J., Ohkura,M., and Imoto, K. (2001). A high signal-to-noise Ca(2þ) probe composed of a single

green fluorescent protein. Nat. Biotechnol. 19, 137–141.

Ohkura, M., Matsuzaki, M., Kasai, H., Imoto, K., and Nakai, J. (2005). Genetically encoded bright

Ca2þ probe applicable for dynamic Ca2þ imaging of dendritic spines. Anal. Chem. 77, 5861–5869.

Palmer, A. E., Jin, C., Reed, J. C., and Tsien, R. Y. (2004). Bcl-2-mediated alterations in endoplasmic

reticulum Ca2þ analyzed with an improved genetically encoded fluorescent sensor. Proc. Natl. Acad.

Sci. USA 101, 17404–17409.

Palmer, A. E., Giacomello, M., Kortemme, T., Hires, S. A., Lev-Ram, V., Baker, D., and Tsien, R. Y.

(2006). Ca2þ indicators based on computationally redesigned calmodulin-peptide pairs. Chem. Biol.

13, 521–530.


Persechini, A., Lynch, J. A., and Romoser, V. A. (1997). Novel fluorescent indicator proteins for

monitoring free intracellular Ca2þ. Cell Calcium 22, 209–216.

Pologruto, T. A., Yasuda, R., and Svoboda, K. (2004). Monitoring neural activity and [Ca2þ] withgenetically encoded Ca2þ indicators. J. Neurosci. 24, 9572–9579.

Pozzan, T., and Rudolf, R. (2009). Measurements of mitochondrial calcium in vivo. Biochim. Biophys.

Acta 1787, 1317–1323.

ReiV, R. F., Thiel, P. R., and Schuster, C. M. (2002). DiVerential regulation of active zone density

during long-term strengthening of Drosophila neuromuscular junctions. J. Neurosci. 22, 9399–9409.

ReiV, D. F., Ihring, A., Guerrero, G., IsacoV Ehud, Y., Joesch, M., Nakai, J., and Borst, A. (2005). In

vivo performance of genetically encoded indicators of neural activity in flies. J. Neurosci. 25,

4766–4778.

Rekas, A., Alattia, J.-R., Nagai, T., Miyawaki, A., and Ikura, M. (2002). Crystal structure of venus, a

yellow fluorescent protein with improved maturation and reduced environmental sensitivity. J. Biol.

Chem. 277, 50573–50578.

Rizzuto, R., Simpson, A. W. M., Brini, M., and Pozzan, T. (1992). Rapid changes of mitochondrial

Ca2þ revealed by specifically targeted recombinant aequorin. Nature 358, 325–327.

Robert, V., Gurlini, P., Tosello, V., Nagai, T., Miyawaki, A., Di Lisa, F., and Pozzan, T. (2001). Beat-

to-beat oscillations of mitochondrial [Ca2þ] in cardiac cells. EMBO J. 20, 4998–5007.

Romoser, V. A., Hinkle, P. M., and Persechini, A. (1997). Detection in living cells of Ca2þ-dependent

changes in the fluorescence emission of an indicator composed of two green fluorescent protein

variants linked by a calmodulin-binding sequence. A new class of fluorescent indicators. J. Biol.

Chem. 272, 13270–13274.

Rudolf, R., Mongillo, M., Magalhaes Paulo, J., and Pozzan, T. (2004). In vivo monitoring of Ca(2þ)

uptake into mitochondria of mouse skeletal muscle during contraction. J. Cell Biol. 166, 527–536.

Shimozono, S., Fukano, T., Nagai, T., Kirino, Y., Mizuno, H., and Miyawaki, A. (2002). Confocal

imaging of subcellular Ca2þ concentrations using a dual-excitation ratiometric indicator based on

green fluorescent protein. Sci. STKE 2002, pl4.

Shimozono, S., Fukano, T., Kimura, K. D., Mori, I., Kirino, Y., and Miyawaki, A. (2004). Slow Ca2þ

dynamics in pharyngealmuscles inCaenorhabditis elegansduring fast pumping.EMBORep. 5, 521–526.

Silver, R. A., Whitaker, M., and Bolsover, S. R. (1992). Intracellular ion imaging using fluorescent dyes:

Artefacts and limits to resolution. Pflugers Arch. 420, 595–602.

Solovyova, N., and Verkhratsky, A. (2002). Monitoring of free calcium in the neuronal endoplasmic

reticulum: An overview of modern approaches. J. Neurosci. Methods 122, 1–12.

Souslova, E. A., Belousov, V. V., Lock, J. G., Stromblad, S., Kasparov, S., Bolshakov, A. P.,

Pinelis, V. G., Labas, Y. A., Lukyanov, S., Mayr, L. M., and Chudakov, D. M. (2007). Single

fluorescent protein-based Ca2þ sensors with increased dynamic range. BMC Biotechnol. 7, 37.

Suzuki, H., Kerr, R., Bianchi, L., Frokjaer-Jensen, C., Slone, D., Xue, J., Gerstbrein, B., Driscoll, M.,

and Schafer, W. R. (2003). In vivo imaging of C. elegans mechanosensory neurons demonstrates a

specific role for the MEC-4 channel in the process of gentle touch sensation. Neuron 39, 1005–1017.

Tallini, Y. N., Ohkura, M., Choi, B.-R., Ji, G., Imoto, K., Doran, R., Lee, J., Plan, P., Wilson, J.,

Xin, H.-B., Sanbe, A., Gulick, J., et al. (2006). Imaging cellular signals in the heart in vivo: Cardiac

expression of the high-signal Ca2þ indicator GCaMP2. Proc. Natl. Acad. Sci. USA 103, 4753–4758.

Tian, L., Hires, S. A., Mao, T., Huber, D., Chiappe, M. E., Chalasani Sreekanth, H., Petreanu, L.,

Akerboom, J., McKinney, S. A., Schreiter, E. R., Bargmann, C. I., Jayaraman, V.,, et al. (2009).

Imaging neural activity in worms, flies and mice with improved GCaMP calcium indicators. Nat.

Methods 6, 875–881.

Truong, K., Sawano, A., Mizuno, H., Hama, H., Tong, K. I., Mal, T. K., Miyawaki, A., and Ikura, M.

(2001). FRET-based in vivo Ca2þ imaging by a new calmodulin-GFP fusion molecule. Nat. Struct.

Biol. 8, 1069–1073.

Tsien, R. Y. (1998). The green fluorescent protein. Ann. Rev. Biochem. 67, 509–544.

Tsien, R. Y. (2010). The 2009 Lindau Nobel Laureate Meeting: Roger Y. Tsien, Chemistry 2008. JoVE

35. http://www.jove.com/index/Details.stp?ID=1575, doi: 10.3791/1575.

http://www.jove.com/index/Details.stp?ID=1575


Varadi, A., and Rutter, G. A. (2002a). Dynamic imaging of endoplasmic reticulum Ca2þ concentration

in insulin-secreting MIN6 cells using recombinant targeted cameleons: Roles of sarco(endo)plasmic

reticulum Ca2þ-ATPase (SERCA)-2 and ryanodine receptors. Diabetes 51(Suppl. 1), S190–S201.

Varadi, A., and Rutter, G. A. (2002b). Green fluorescent protein calcium biosensors. Calcium imaging

with GFP cameleons. Methods Mol. Biol. 183, 255–264.

Varadi, A., and Rutter, G. A. (2004). Ca2þ-induced Ca2þ release in pancreatic islet beta-cells: Critical

evaluation of the use of endoplasmic reticulum-targeted ‘‘cameleons’’. Endocrinology 145, 4540–4549.

Visser, A. J. W. G., Laptenok, S. P., Visser, N. V., van Hoek, A., Birch, D. J. S., Brochon, J. C., and

Borst, J. W. (2010). Time-resolved FRET fluorescence spectroscopy of visible fluorescent protein

pairs. Eur. Biophys. J. 39, 241–253.

Wang, Q., Shui, B., KotlikoV, M. I., and Sondermann, H. (2008). Structural basis for calcium sensing by

GCaMP2. Structure (Lond., Engl.—1993). 16, 1817–1827.

Wang, J. W., Wong, A. M., Flores, J., Vosshall, L. B., and Axel, R. (2003). Two-photon calcium

imaging reveals an odor-evoked map of activity in the fly brain. Cell 112, 271–282.

Yang, X., Xu, P., and Xu, T. (2005). A new pair for inter- and intra-molecular FRET measurement.

Biochem. Biophy. Res. Commun. 330, 914–920.

Yu, R., and Hinkle, P. M. (2000). Rapid turnover of calcium in the endoplasmic reticulum during

signaling. Studies with cameleon calcium indicators. J. Biol. Chem. 275, 23648–23653.

Yu, D., Baird, G. S., Tsien, R. Y., and Davis, R. L. (2003). Detection of calcium transients in

Drosophila mushroom body neurons with camgaroo reporters. J. Neurosci. 23, 64–72.

Zacharias, D. A., Baird, G. S., and Tsien, R. Y. (2000). Recent advances in technology for measuring

and manipulating cell signals. Curr. Opin. Neurobiol. 10, 416–421.

Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2002). Partitioning of lipid-modified

monomeric GFPs into membrane microdomains of live cells. Science 296, 913–916.

Zou, J., Hofer, A. M., Lurtz, M. M., Gadda, G., Ellis, A. L., Chen, N., Huang, Y., Holder, A., Ye, Y.,

Louis, C. F., Welshhans, K., Rehder, V., et al. (2007). Developing sensors for real-time measurement

of high Ca2þ concentrations. Biochemistry 46, 12275–12288.

CHAPTER 7

METHODS IN CELL BIOL

Patch Clamp Methods for StudyingCalcium Channels

David L. Armstrong, Christian Erxleben, and Jody A. WhiteMembrane Signaling GroupLaboratory of NeurobiologyNational Institute of Environmental Health SciencesNIH, Durham, North Carolina, USA

A

OGY,

bstract

VOL. 99 0091183 DOI: 10.1016/S0091

-679X-679X

I. I
ntroduction II. R ationale III. M ethods
A.
Assembling the Patch Clamp Rig B. Making Pipettes C. Making Seals D. Making Recordings
IV. M
aterials V. D iscussion VI. S ummary
R
eferences
Abstract

The patch clamp technique, which was introduced by Neher and Sakmann and

their colleagues in 1981, has allowed electrophysiologists to record ion channel

activity from most mammalian cell types. When well-established precautions are

taken to minimize electrical and mechanical fluctuations, current transients as small

as 0.5 pA and as brief as 0.5 ms can be measured reliably in cell-attached patches of

plasmamembranewith apolishedglass pipettewhen it forms a giga-ohmsealwith the

membrane. Inmany cases, this is suYcient to watch individual channel proteins open

and close repeatedly in real time on metabolically intact cells. No other technique

currently provides a more precise or detailed view of the function and regulation of

calcium channel gating. If antibiotics are added to the pipette to permeabilize the

/10 $35.00(10)99007-9

184 David L. Armstrong et al.

membrane underneath to small monovalent cations, thereby allowing the entire cell

to be voltage-clampedwithout disrupting its contents, the integrated activity of all the

calcium channels in the surface membrane can be measured.

I. Introduction

Calcium ions trigger many fundamental cellular processes by binding to pro-

teins, usually with dissociation constants around 1 mM calcium. However, unlike

other second messengers such as cyclic nucleotides, calcium is neither created nor

destroyed by biological processes. Therefore, regulating calcium-dependent pro-

cesses requires moving calcium ions into and out of cellular compartments. Trans-

port proteins that hydrolyze ATP move calcium ions from lower to higher

concentrations. They maintain resting cytoplasmic calcium at 100–200 nM by

pumping calcium out of the cytosol into the endoplasmic reticulum, the mitochon-

dria, or out across the plasma membrane. By contrast, ion channel proteins, when

the channels are open, allow calcium ions to diVuse rapidly down their electro-

chemical gradient back into the cytosol. Many channels are permeable to calcium

because they pass all cations, including calcium, up to a certain size, but most of

the current is carried by more prevalent ions such as sodium or potassium. In

contrast, some ‘‘calcium-selective’’ channels pass calcium almost exclusively, even

in the presence of other cations.

There are currently three classes of calcium-selective channel proteins (Table I):

three families of voltage-gated calcium channels in the plasma membrane (CaV1-3);

two families of calcium-selective channels in the endoplasmic reticulum membrane;

Table ICalcium-selective channels

Voltage-activated calcium channels in the surface membrane

L type CaV1.1 CACNA1S Skeletal muscle

CaV1.2 CACNA1C Cardiovascular muscle

CaV1.3 CACNA1D Endocrine cells, neurons

CaV1.4 CACNA1F Retina

P/Q type CaV2.1 CACNA1A Nervous system

N type CaV2.2 CACNA1B Nervous system

R type CaV2.3 CACNA1E Nervous system

T type CaV3.1 CACNA1G Brain, heart

CaV3.2 CACNA1H Brain, endocrine cells, heart

CaV3.3 CACNA1I Brain

Calcium release-activated channel (CRAC) in the surface membrane

STIM-gated ORAI1-3?

Ligand-gated calcium channel in the endoplasmic reticulum

IP3-gated ITPR1-3

Ryanodine-gated RYR1-3

7. Patch Clamp Methods 185

and calcium release-activated channels (CRAC) that mediates store-operated calci-

um entry (SOCE) across the plasma membrane (Hille, 2001; Hogan et al., 2010). All

of these proteins provide channels that allow calcium to diVuse into the cytosol whenthe channel opens. Each open channel protein has a unique unitary conductance for

calcium, ranging from approximately 0.1 to several hundred picosiemens (pS), but

the proteins spontaneously cycle between open and closed conformations on a time

scale of milliseconds. It is the rates of these transitions rather than the conductances

which are regulated by physiological events to control the amplitude of calcium

fluxes. Such unitary currents are often diYcult tomeasure because they are small and

individual openings last less than a millisecond. However, the total amount of

current crossing the entire surface membrane of a cell at any time is the product of

the number of channels (N), the fraction of time they spend in the open state (Po),

their unitary conductance (g), and the electrochemical driving force (DV) measured

as the diVerence between the voltage across the membrane (Vm) and the ion’s Nernst

potential.

Even small currents produce physiologically significant increases in intracellular

calcium. For example, a current of only 0.1 pA (pA¼10�12 A), which corresponds

to 0.1 pC of charge per second, or �300,000 divalent ions per second, will transfer

300 calcium ions each millisecond the channel is open. Such a current, 0.1 pA

lasting 1 ms, is just below the current technology of detection with the patch clamp

technique. Nevertheless, it would produce a physiologically significant change in

intracellular calcium. The calcium ions cannot diVuse on average much farther

than a micrometer in a millisecond, so the concentration under the membrane will

rise transiently to 0.5 mM, more than double the resting level of calcium. If there

was only one such channel that opened for 1 ms in every square micrometer of

membrane of a spherical cell with a diameter of 10 mm (volume�0.5 pL; area -

�300 mm2), then the resulting 30 pA current would almost double intracellular

calcium concentration throughout the cell. Action potentials that depolarize cells

for tens of milliseconds will have correspondingly larger eVects. Thus, millisecond

diVerences in calcium channel kinetics have profound consequences for cell physi-

ology and human health (Erxleben et al., 2006).

This calculation also illustrates the danger of expressing recombinant channels

in mammalian fibroblasts. Investigators routinely report currents of a few

nanoamperes, which even inexperienced investigators can measure with the

patch clamp technique. However, in the scenario outlined above, a 3-nA current

would represent a 100� higher density of channels and produce 100� larger

increases in calcium, which might lead to cytotoxic reactions. In most cases, such

recordings are made with exogenous calcium buVers in the cytosol, which not only

prevent cytotoxicity but also preclude analysis of physiological regulation of

calcium channels by calcium-dependent signaling. In addition, because most calci-

um channels have a low probability of opening (Po) less than 0.1, the larger current

density reflects at least 1000 channel proteins per square micrometer, or more than

10% of the space available with close packing. At this density, there might not be

room for each channel protein to be associated with its normal penumbra of


regulatory proteins, even if their expression was upregulated by the cell to com-

pensate for the increased expression of the channels.

Recording the activity of recombinant channels from cells that are dialyzed with

simple salt solutions using the conventional ‘‘whole-cell’’ configuration of the

patch clamp is also dangerous because in the absence of normal metabolic activity,

many calcium channels have conformations from which it is diYcult to elicit

robust activity with physiological stimuli. Nevertheless, if the plasmid drives the

expression of 100–10,000 times more channels than are normally present, even

channels with very little activity might generate quite large currents. We would

argue, however, that identifying the mechanisms that halve or double the activity

of recombinant channels, which only open for 1 or 2 ms every second, might not be

relevant to their physiological functions. Thus, this chapter focuses on the more

diYcult and consequently less frequently used techniques of single-channel record-

ing from cell-attached patches and the perforated patch technique for voltage

clamping metabolically intact cells.

II. Rationale

Calcium signaling can be investigated at many levels. Although the patch clamp

method is a very quantitative technique for measuring ion fluxes across the plasma

membrane, internal stores of calcium are important sources that the patch clamp

technique cannot access. To study calcium release from internal stores through

calcium-selective channels, one must use fluorescent calcium indicators or recon-

stitute the channels from organelles in bilayers. This chapter provides an overview

of the patch clamp method for measuring voltage-activated calcium currents

across the plasma membrane. We have applied this technique primarily to dis-

sociated mammalian cells in vitro culture, and it has also been adapted to studying

neurons in brain slices (Sakmann and Neher, 1995). Very basically, the technique

involves the use of an operational amplifier circuit to clamp voltage changes

between a wire in the patch pipette and a wire in the bath and the measurement

of how much current it takes to hold the voltage constant. When a giga-ohm

(GO¼109 ohms) seal is formed between the glass patch pipette and the cell

membrane, background current fluctuations can be reduced suYciently to detect

picoampere currents. Gigaohm seals are most eVective for stable, low-noise re-

cording when they are in the 10–100 GO range. However, in our experience, most

investigators routinely settle for seals in the 1–10 GO range. The methods de-

scribed below allow us to routinely obtain seals around 50 GO. Unfortunately,

they are all necessary for success.

When the patch clamp technique was introduced almost 30 years ago (Hamill

et al., 1981), it was the only quantitative method available to obtain reliable

physiological information about calcium signals in mammalian cells and the

proteins that mediate them. Now, however, there are increasingly sophisticated

calcium indicators that can be targeted genetically to specific compartments in


specific cells. In addition, genome cloning, fluorescence confocal microscopy, and

structural studies of membrane proteins have advanced to the point where they are

better suited than the patch clamp technique for identifying which channel proteins

are expressed by specific cell types, how their distribution on and traYcking to and

from the cell surface is regulated, and their three-dimensional molecular structures.

Nevertheless, the patch clamp remains the technique of choice for investigating the

channels’ physiological function and regulation at the molecular level in real time.

In the future, genetic manipulation of channel sequence and expression in model

organisms and mass spectrometry of posttranslational modifications might begin

to supplant even these applications of the patch clamp technique.

Three volumes of Methods in Enzymology have already been devoted to the

patch clamp method, so it would be impossible to provide a suYciently detailed

introduction here that would allow a novice to make recordings without consulting

other sources. Scientists with little training in physics and/or membrane physiology

should start with the ion channel primer by Aidley and Stanfield (1996) and the

technical primers by Molleman (2003) and Ogden (1987). More experienced phy-

siologists could move directly to the compendia by Sakmann and Neher (1995) and

by Rudy and Iverson (1992). More specialized topics are covered in subsequent

volumes edited by Conn (1998, 1999). If you are creating your own rig from

scratch, chapter by Jim Rae and Rick Levis in volume 207 is essential reading.

III. Methods

A. Assembling the Patch Clamp Rig

To localize contributions to noise, it is important to assemble the rig slowly,

piece by piece. There are two general classes of noise to eliminate initially. Sinu-

soidal fluctuations arise from electronic interference by power sources and from

mechanical vibrations. In practice, they are often diYcult to distinguish because

they produce oscillations of similar frequency, but both should be eliminated

completely. Later, after giga-ohm seal formation, any remaining noise should

have much higher frequencies. Some of this noise arises from capacitative coupling

between the glass wall of the pipette and the salt solution surrounding it in the

bath, which can be minimized experimentally. Remaining contributions to high

frequency noise can only be reduced by filtering the signal, which limits how brief

an event of a given amplitude one can detect.

To begin, run the calibration tests that are specified by the manufacturer on the

amplifier while it is sitting on a desk. Then assemble the air table, the faraday cage,

the microscope, and the manipulator. They should share a common ground. The

lamp of the microscope will require a DC power supply and each of the manip-

ulator’s motors must be grounded separately. Surprisingly, despite all their metal

parts, many microscopes are not isopotential and need to be grounded at several

points. With the input of the amplifier’s headstage still open and the output filtered

Table IIChecklist for reducing noise

Electrical sources

Faraday cage grounded

Microscope and (all three) manipulator motors grounded

DC supply for microscope lamp or LED

Check for ground loops

Pipette holder clean and dry

Mechanical sources

Table floating freely

Microscope and manipulator tightly secured to table

Electrical lines and perfusion tubing lashed tightly to microscope base or table

No air drafts blowing on rig

Rubber gasket that holds glass capillary in pipette holder in new condition


to 100 Hz, confirm that there are no sinusoidal fluctuations at 10� the highest gain

you plan to use. Table II contains a checklist for major sources of noise.

Finally, test the pipette holder, metal electrodes, and perfusion system by

inserting into the recording chamber a cover slip that has a hemisphere of solidified

sylgard elastomer (184, Dow Corning). Pressing a patch pipette into the sylgard

allows you to obtain seals of up to 100 GO. Silver wires can be chlorided by

dipping them in chlorine bleach for several minutes, then rinsed thoroughly and

air dried. Because silver is toxic to cells, the ground wire is connected to the bath

through an agar salt bridge.

B. Making Pipettes

Commercial programmable pipette pullers are now available, but some experi-

mentation is required inevitably to find the settings that produce pipettes with the

desired overall shape: long and narrow tapers for cell-attached patch recordings to

minimize capacitance with the bath; and short stubby tapers for perforated patch

whole-cell recordings to speed antibiotic diVusion to the tip and minimize access

resistance. Handling the capillaries with wet or greasy hands on the portion that is

heated contributes to variability, but before pulling the pipettes, both ends of the

capillaries should be fire polished gently to increase the longevity of the silver

chloride coating on the wire in the pipette holder. Otherwise, the sharp glass edge

at the back of the pipette scrapes oV a little silver chloride every time you insert a

new pipette. Also make sure the pipette holders and their internal rubber gaskets

are designed to hold capillaries of the same outer diameter as your pipettes, or they

will not fit snugly.

For most low-noise applications, the pipette must be coated with ‘‘Sylgard’’

(Corning 184), an elastomer that is cured with heat, which insulates the pipette

walls from the bath solution. The reported curing time for Sylgard is 10 min at


150 �C; however, a brief exposure (<1 min) to a heating element followed by 10–

20 min at room temperature should be adequate for curing before moving on to

shaping of the tip. The Sylgard is most eVective when it is applied thickly, but

prevents giga-ohm seal formation if it gets on the tip. Our method of applying

Sylgard is illustrated in Fig. 1. We use gravity by mounting the pipette vertically

A

C E

FD

B

Fig. 1 Application of Sylgard. Sylgard (Corning 184), a heat-cured polymer, is applied to patch

pipettes after pulling but before polishing to reduce the capacitance between the glass walls and the bath.

(A) Sylgard is stored in small 1-ml plastic tubes in the freezer until it is ready for use. Room temperature

Sylgard is applied using a 25-gauge hypodermic needle that is bent for ease of application. The needle

hub is aYxed to a plastic tube for ease of handling. (B) The pipettes are placed into a hand-made holder

that holds the pipette upright and allows it to be positioned easily so that the heating element made of

tungsten wire surrounds the tip. (C) The pipette is positioned so that the area from the first narrowing of

the glass to the tip is placed above the heating element. (D) Sylgard is applied in consecutive rings or

‘‘donuts’’ starting at the area where the glass first narrows. This lower ring helps protect against noise if

your bath height changes between perfusions. (E) Between applications of consecutive rings, apply heat

to cure each ring. Remember that heat goes up, so the hottest area near the heating wire is above it. (F)

The final ring should also coat the taper of the pipette and can be placed as close as 50–100 nm from the

tip. Be careful not to get Sylgard on the tip or on the heating element!


inside a loop of tungsten wire across the prongs of an outlet plug. We control the

heat transiently with a rheostat and observe the pipette through an inexpensive,

low power, dissecting scope mounted horizontally on a boom arm. The Sylgard is

precured to a viscous consistency at room temperature when it is first made and

then stored in small 1 ml aliquots in the freezer. We apply it with a bent hypoder-

mic needle by taking up a dollop onto the needle, touching the dollop to the

pipette, and slowly winding it on to the glass by turning the pipette from below.

Care must be taken to avoid touching Sylgard to the heating/curing wire, or the

Sylgard vapors will quickly coat the tip. You will know when you get Sylgard too

close to the tip because when you try to polish it, the Sylgard contracting around

the thin glass walls at the tip literally wrinkles the glass.

Everybody has their own favorite glass that they believe forms the tightest seals,

but their relative intrinsic noise can be tested empirically using the Sylgard hemi-

spheres. The lowest noise is produced by quartz glass capillaries (Levis and Rae,

1993), but they require a special laser puller to create the pipettes. On real cells, the

size and shape of the tip also influence success. It is extremely diYcult to make

pipettes that will routinely make >10 GO seals when you cannot clearly see the tip

while you polish it. An extra long working distance (ELWD) objective with at least

40� magnification is essential but they are expensive, so they must be mounted

below the polishing element that is heated to prevent cracking the lens from

repeated heating. Our polishing strategy is illustrated in Fig. 2. We find that it is

also critical to melt a small bead of the pipette glass onto the apex of the heating

element, usually a small loop of platinum wire, presumably because it prevents the

tip from being coated with metal. We find that pipettes with bullet-shaped tips and

initial resistances between 3 and 5 MO make the best seals.

Filling the pipettes also requires some attention to detail. All pipette solutions

should be filtered through 0.2 or 0.45-mm filter disks but do not use filter disks

prepared with wetting agents. To avoid bubbles and washing the dirt inside the

capillaries down into the tip, both of which reduce seal success, one must first fill

the tip separately by immersing it in a small vial of pipette solution and allowing

the first 20–50 mm to fill by capillary action. Then the rest can be backfilled.

Usually bubbles are visible to the naked eye, and they can be removed by gently

flicking the pipette while it is held between the thumb and the forefinger.

C. Making Seals

To reliably get seals over 10 GO, all the precautions in Table III are important. In

addition you have to be fast. It should take only 3–5 min from filling the pipette to

touching down onto the cell, including mounting the pipette in the holder, manip-

ulating it into the chamber, zeroing out the oVset, measuring the resistance, finding

it in the microscope, and manipulating it just above the cell without touching the

bottom or another cell. This takes practice. It also helps to run a little solution

through the chamber before lowering the pipette into the bath to clear any debris

that accumulates on the surface. To minimize debris from collecting on the patch

A

B C

D E

Fig. 2 Fire-polishing pipettes. Fire polishing the tip allows the user to narrow the tip opening to gain

the desired shape and resistance. (A) The pulled pipettes are placed horizontally in a 2D holder mounted

on the stage of an inverted microscope with a long working distance objective �40�. On the opposite

side of the stage, a 3D manipulator is placed with a small platinum wire loop on which a bead of pipette

glass has been melted. (B) The unpolished pipette tip is brought into the same plane of view as the glass

bead, but at opposite sides of the periphery. (C) When current is passed through the wire, the bead gets

hot and expands into the center of the field. The tip of the pipette is transiently manipulated closer to the

bead until the opening at the tip starts to close. (D)When the tip narrows and attains rounded edges, the

pipette is withdrawn and the heat is turned oV. (E) A finished, polished pipette tip.

Table IIIChecklist for obtaining giga-ohm seals

Cells are healthy

CO2 regulation is accurate

No trypsin for 24 h and no transfection reagents in past 12 h

Maintained in salt solution at room temperature for less than 15 min

Pipettes are functioning properly

Pipettes are stored in dust-free box for �4 h after polishing

Pipette solution filtered through �0.45 mm mesh

Tip filled separately by capillary action before back filling; no bubbles

Pipette holder gasket sized to glass and fits snugly (but do not over tighten)

Suction line intact and dry

No mechanical vibrations transmitted to tip (see Table II)

Solutions

Bath solution has more solutes than pipette solution

Neither solution has proteins or ATP

At least one of the solutions has �0.1 mM Ca2þ



pipette after it is lowered in the bath, gentle positive pressure is applied on the

suction line which is then pinched off until contact with the cell. In addition, bath

solutions are designed to be 10% hyperosmotic to the pipette solution (usually by

adding glucose), so water will stream out of the pipette tip. Axial approaches are

more eVective than lowering the angled pipette straight down on to the cell. Some

manipulators program a fourth axis or you can mount the manipulator at an angle

so one axis is parallel to the pipette. We prefer to monitor the approach of the

pipette to the cell surface by looking at the pipette current trace in response to a

small voltage step. As the pipette touches the cell, the amplitude of the current

diminishes. When it is reduced to 33–50%, we release the positive pressure on the

pipette, and then, if necessary, we apply additional suction. To make these manip-

ulations without disturbing the tip, the small tube connected to the pipette holder

must be firmly anchored to the headstage and the microscope stage to prevent

vibrations. Most of us prefer to apply the suction directly by mouth because the

suction is controlled more easily and changes more evenly. Small flat cells are

obviously harder to patch than large round ones. Visualization of the cell surface

is improved by interference contrast optics, but this can be implemented on very

simple microscopes for the price of two small pieces of black tape (Axelrod, 1981).

Most people we have trained find the rat pituitaryGH cell lines to be the easiest cells

with which to learn patching. They can be obtained from the ATCC.

D. Making Recordings

1. Cell-Attached Patch Recordings

The two most important things to remember about the cell-attached configura-

tion are that the voltage polarities are reversed relative to traditional whole-cell

recording and that there are two membranes in the current path between your

electrodes. The amplifier sets the voltage between the pipette wire and the ground

electrode in the bath, but, by convention, the cell interior is negative to the outside,

so depolarizing the membrane of a cell-attached patch means making the pipette

more negative. However, the cell membrane also contributes a voltage which is not

clamped in the cell-attached configuration, so to accurately determine the voltage

across the patch, the cell’s membrane potential must be set to zero by bathing the

cell in an equimolar potassium solution of 140–150 mM. Unfortunately, that

means all the voltage-gated channels, including the calcium channels, in the

membrane outside the patch will be activated. Therefore, it is essential to reduce

extracellular calcium to avoid flooding the cell with calcium. Some investigators

use calcium buVers, but the bath is infinite relative to the cell volume and most cells

are rapidly depleted of calcium. A more physiological solution is to reduce extra-

cellular calcium to 0.1 mM, which is still 100�more than resting calcium inside the

cell, but produces negligible currents in the presence of 2–5 mM magnesium.

For cell-attached recordings, the primary goal is increasing the signal and

reducing noise. Increasing the signal is usually achieved by increasing the


concentration of the divalent ion in the pipette solution on the outside of the patch.

Practically, 90 mM calcium is as high as one can achieve without making the

pipette solution hyperosmotic to the bath. However, such large currents also invite

unphysiological, or even toxic, calcium responses in the cytoplasm, so most people

use 90 mM barium. The conductance of voltage-gated channels to barium is often

higher than to calcium, and barium blocks many potassium channels, but high

concentrations of divalent ions alter the surface potential of the membrane and

shift voltage-activation and inactivation curves. If all the precautions that are

described here are taken to reduce noise, it is possible to record unitary currents

with physiological concentrations of calcium (Josephson et al., 2010).

Patch pipettes with initial resistances greater than 10 MO rarely make high

resistance seals, so several square micrometers of membrane are usually drawn

into wider, lower resistance pipettes by the suction. Such large areas of membrane

usually contain many diVerent types of channels, many of which have much larger

unitary conductance than calcium channels. Therefore, sodium, potassium, and

chloride must be replaced with impermeant ions and sometimes, ion channel

blockers must be added too. Cesium does not permeate potassium selective chan-

nels, but it does go through many nonselective cation channels, so N-methyl-d-glucamine (NMDG) is a safer substitute for sodium. Chloride can be replaced with

methanesulfonic acid, although a few millimolar chloride must be left in to allow

reversible current movement between the wire and the solution.

To reduce background noise further, coating the pipette with Sylgard and

obtaining higher resistance seals are the two most practical steps. Lowering the

bath also helps although no perfusion system is perfect, so the bath cannot be too

low. Stable recordings also require a drift-free manipulator or lifting the cell oV the

bottom of the chamber, which is easier when calcium is removed from the bath

solution. If you can lift the cell oV the substrate, the lowest noise recordings can be

obtained by putting a layer of inert oil over the surface (Rae and Levis, 1992), but

then the chamber must be designed with the perfusion inlet and outlet at the

bottom.

2. Perforated Patch Recordings

Many people report that they have tried perforated patch recording but could

not get it to work. In our experience, there are two critical steps that befuddle most

beginners until we show them the ‘‘secrets.’’ The two essential steps to successful

perforated patch recording are eYcient solubilization of the antibiotic and optimal

loading of the pipette (Horn and Marty, 1988). We learned how to sonicate from

Robert Rosenberg, a calcium channel researcher at UNC Chapel Hill for many

years, who also used bilayers. He taught us that cylindrical devices are most

eVective, and they must be filled to the height where the water surface is most

agitated. Then placing a covered, round bottom, cylindrical glass tube with less

than 1 ml of solution into the vortex for less than a minute is suYcient to disperse

the nystatin or amphotericin or gramicidin, but this only lasts for an hour or two.


The second secret is empirically determining exactly how far to fill the tip with

antibiotic-free solution. If the antibiotic is too close to the tip, it can permeabilize

the entire cell membrane and prevents giga-ohm seal formation. If it is too far

back, it takes forever for the perforation to proceed. Therefore, each person must

take the time to figure out exactly how long to dip their pipettes in the antibiotic-

free solution by painstakingly examining and measuring how far up the solution

goes for a given count; ‘‘one one thousand, two one thousand, etc.’’ With those

preliminaries, we routinely get access resistances below 20 MO in 5 min and often it

goes down closer to 10 MO.Finally, there is one mistake that even experienced electrophysiologists make

that is much easier to avoid. Most patch clampers almost exclusively use the

‘‘whole cell’’ configuration, in which the membrane underneath the pipette is

disrupted by suction, and the cell is dialyzed with the pipette solution. To achieve

this configuration, they routinely add calcium chelators, such as EGTA, to the

pipette solution. Chelating calcium destabilizes the patch membrane and prevents

cytoplasmic calcium from rising to toxic levels. However, if EGTA is not removed

for perforated patch recordings, suction often leads to cell dialysis with the antibi-

otic, and the current through its channels dwarfs the calcium current.

3. Recordings of Calcium Release-Activated Currents, Icrac

The conductance of individual Icrac channels is too low to measure with the

patch clamp technique although it has been estimated by fluctuation analysis

(Prakriya and Lewis, 2006). While Icrac resulting from overexpression of STIM/

Orai in recombinant systems can be as large as 100 pA/pF at �100 mV, which

translates into 1 nA whole-cell current in an average HEK293 cell, endogenous

Icrac currents are only a few pA/pF or about 10 pA for the average cell. In order to

measure whole-cell currents in the 1–10 pA/pF range, low-noise techniques that

are usually only used for high-resolution single-channel current measurements

need to be employed. Specifically, the whole-cell recording electrodes should be

coated with an elastomer-like Sylgard 184 and, of course, the initial seal resistance

prior to establishing whole-cell configuration should be as high as possible. With

Sylgard-coated and subsequently fire polished electrodes, one should routinely

obtain seal resistances of �50 GO in the cell-attached mode on HEK cells. At

least for HEK293 cells, if calcium is buVered to 100 nM in the pipette, the whole-

cell configuration can then be obtained easily by a single, brief, and gentle suction

pulse.

In classical voltage jump protocols that are used to elicit whole-cell calcium

currents, the eVect of carefully coating the pipette with Sylgard can be readily

observed as a reduction of the fast capacitive transients at the beginning and end

of the step. For Icrac measurements, however, investigators routinely use fast

voltage-ramp protocols (typically �100 mV/s) to measure the quasi steady-state

IV relationship of Icrac. Under those conditions, the contribution of capacitative

current to the total current is less visible since it remains constant during the ramp,


and small changes in the bath level that are almost inevitable during perfusion

changes will change the amplitude of the capacitative current. This can produce a

shift of the instantaneous current–voltage relationship along the current axis,

which can easily be misinterpreted as a shift in the reversal potential of Icrac.

IV. Materials

Although all patch clampers have their favorite vendors, as government-

employed scientists, we cannot reveal our specific preferences here (see

Table IV). If there is someone else at your institution who is patch clamping

successfully, you should consult them first anyway and consider buying what

they have after you try it because they will already be familiar with its operation.

V. Discussion

The patch clamp methods described here are not easy. They have daunting

initial costs in instrumentation, almost $100,000 at current list prices, and a steep

learning curve for both the technical aspects and conceptual foundations of

electrophysiology. Even for experienced investigators, the patch clamp technique

is a laborious method that requires constant trouble shooting. Nevertheless, like

any physical activity, patch clamp performance is always enhanced mysteriously

by a belief in success. Thus, a common refrain around our laboratory is ‘‘If Neher

or Sakmann could do it, then you can do it too.’’

Table IVComponents of a patch clamp rig

Tungsten wire for sylgarding and polishing

Silver wire and pellets

Pipette glass

Pipette puller

Dissecting microscope on swing arm for sylgard coating

Pipette polisher with

Inverted microscope with �40� LWD objective

Coarse manipulator for adjusting heating element

Fluorescent microscope with diVerential interference contrast optics

LED illumination is cheaper, cooler, and electrically quieter

3D micromanipulator with submicron resolution and no drift

Patch clamp amplifier, computer interface, and software

Computer according to the software manufacturer’s specifications

Physiological chamber and perfusion apparatus

Pipette holder and agar bridge ground electrode

Small cylindrical sonicator


VI. Summary

The patch clamp technique provides quantitative records of calcium channel

activity in the plasma membrane at the molecular level in real time in situ.

However, studying calcium channel function and regulation under physiologi-

cal conditions in metabolically intact cells requires more demanding approaches

than the conventional whole-cell recording through ruptured patches on dia-

lyzed cells. Whole-cell recordings through antibiotic-perforated patches and

single-channel recordings from cell-attached patches require additional eVortand attention to detail, but are currently unrivaled in their precision and

detailed view of calcium channel function and physiological regulation of

gating.

Acknowledgments

Our studies have been supported by the NIH intramural program at NIEHS through Grant Z01-

ES080043. We also thank Sue Edelstein for drawing Figures 1 & 2.

References

Aidley, D. J., and Stanfield, P. R. (1996). ‘‘Ion Channels: Molecules in Action.’’ 1st edn. Cambridge

University Press, Cambridge, UK.

Axelrod, D. (1981). Zero-cost modification of bright field microscopes for imaging phase gradient on

cells: Schlieren optics. Cell Biophys. 3, 167–173.

Conn, P. M. (1998). ‘‘Methods in Enzymology: Ion Channels Part B.’’ Vol. 293. Academic Press,

New York, NY.

Conn, P. M. (1999). ‘‘Methods in Enzymology: Ion Channels Part C.’’ Vol. 294. Academic Press,

New York, NY.

Erxleben, C., Liao, Y., Gentile, S., Chin, D., Gomez-Alegria, C., Mori, Y., Birnbaumer, L., and

Armstrong, D. L. (2006). Cyclosporin and Timothy syndrome increase mode 2 gating of CaV1.2

calcium channels through aberrant phosphorylation of S6 helices. Proc. Natl. Acad. Sci. USA 103,

3932–3937.

Hamill, O. P., Marty, A., Neher, E., Sakmann, B., and Sigworth, F. J. (1981). Improved patch-clamp

techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers

Arch. 391, 85–100.

Hille,B. (2001). ‘‘IonChannels ofExcitableMembranes.’’ 3rd edn. SinauerAssociates, Inc., Sunderland,MA.

Hogan, P. G., Lewis, R. S., and Rao, A. (2010). Molecular basis of calcium signaling in lymphocytes:

STIM and ORAI. Annu. Rev. Immunol. 28, 491–533.

Horn, R., and Marty, A. (1988). Muscarinic activation of ionic currents measured by a new whole-cell

recording method. J. Gen. Physiol. 92, 145–159.

Josephson, I. R., Guia, A., Lakatta, E. G., Lederer, W. J., and Stern, M. D. (2010). Ca(2þ)-dependent

components of inactivation of unitary cardiac L-type Ca(2þ) channels. J. Physiol. 588, 213–223.

Levis, R. A., and Rae, J. L. (1993). The use of quartz patch pipettes for low noise single channel

recording. Biophys. J. 65, 1666–1677.

Molleman,A. (2003). ‘‘PatchClamping,An IntroductoryGuide toPatchClampElectrophysiology.’’ John

Wiley & Sons, Chichester, England.

Ogden, D. (1987). ‘‘Microelectrode Techniques: The Plymouth Workshop.’’ 2nd edn. The Company of

Biologists Limited, Cambridge, UK.


Prakriya, M., and Lewis, R. S. (2006). Regulation of CRAC channel activity by recruitment of silent

channels to a high open-probability gating mode. J. Gen. Physiol. 128, 373–386.

Rae, J. L., and Levis, R. A. (1992). A method for exceptionally low noise single channel recordings.


Rudy, R., and Iverson, L. E. (1992). ‘‘Methods in Enzymology: Ion Channels.’’ Vol. 207. Academic

Press, New York, NY.

Sakmann, B., andNeher, E. (1995). ‘‘Single-Channel Recording.’’ 2nd edn. PlenumPress, NewYork,NY.

CHAPTER 8


Nuclear Patch-Clamp Recording fromInositol 1,4,5-Trisphosphate Receptors

Taufiq Rahman and Colin W. TaylorDepartment of PharmacologyTennis Court Road, University of CambridgeCambridge, United Kingdom

A

OGY,All rig

bstract


-679X-679X

I. In
troduction II. N uclear Patch-Clamp Recording III. C hoice of Cells for Analyses of IP3R IV. M ethods
A.
Culture of DT40 Cells B. Isolation of Nuclei C. Solutions for Patch-Clamp Recording D. Patch-Clamp Recording E. Analysis of Single-Channel Records
V. C
oncluding Remarks R eferences
Abstract

Inositol 1,4,5-trisphosphate receptors (IP3R) are ubiquitous intracellular Ca2þ

channels. They are regulated by IP3 and Ca2þ and can thereby both initiate local

Ca2þ release events and regeneratively propagate Ca2þ signals evoked by receptors

that stimulate IP3 formation. Local signaling by small numbers of IP3R underpins

the utility of IP3-evoked Ca2þ signals as a ubiquitous signaling pathway. The

physiological impact of Ca2þ release by very small numbers of IP3R underscores

the necessity to understand the behavior of IP3R at the single-channel level.

In addition, and in common with analyses of every other ion channel, single-

channel analyses have the potential to define the steps linking IP3 binding to

channel opening. Patch-clamp recording, by resolving the openings and closings

of single channels with exquisite temporal resolution, is the most powerful

/10 $35.00(10)99008-0

200 Taufiq Rahman and Colin W. Taylor

technique for analysis of single-channel events. It has contributed enormously to

the understanding of gating and desensitization/inactivation of numerous ion

channels. However, most IP3R reside within intracellular membranes, where they

are inaccessible to conventional patch-clamp recording methods. Here, we describe

the application of nuclear patch-clamp methods to single-channel analyses of

native and recombinant IP3R.

I. Introduction

Inositol 1,4,5-trisphosphate receptors (IP3R) comprise a family of tetrameric

intracellular channels that mediate the release of Ca2þ from the intracellular stores

of almost all animal cells (Foskett et al., 2007; Taylor et al., 1999). Three genes

encode homologous subunits of vertebrate IP3R, and a single gene encodes inver-

tebrate IP3R. The key structural determinants of IP3R activation, although pres-

ently poorly understood, are likely to be similar for all IP3R. Activation is initiated

by binding of IP3 to a conserved IP3-binding core toward the N-terminal of each

subunit (Bosanac et al., 2002), conformational changes then pass via the

N-terminal suppressor domain (Bosanac et al., 2005) to the pore, which is formed

by transmembrane regions lying toward the C-terminus (Boehning and Joseph,

2000; Foskett et al., 2007; Rossi et al., 2009; Taylor et al., 2004). Most IP3R in most

cells are expressed within membranes of the endoplasmic reticulum (ER). DiVerentIP3R subtypes may, however, diVer in their subcellular distributions (Taylor et al.,

1999) and in their modulation by various additional signals and associated pro-

teins (Betzenhauser et al., 2008b; Choe and Ehrlich, 2006; Mackrill et al., 1997;

Patterson et al., 2004; Wojcikiewicz and Luo, 1998). Resolving the roles of diVer-ent IP3R subtypes in the genesis of the complex Ca2þ signals that regulate cellular

activity is an important issue (Futatsugi et al., 2005; Miyakawa et al., 1999;

Sugawara et al., 1997; Wang et al., 2001).

Opening of the intrinsic pore of all IP3R requires binding of IP3 and Ca2þ

(Adkins and Taylor, 1999; Marchant and Taylor, 1997). IP3R can, therefore,

both initiate the Ca2þ signals evoked by receptors that stimulate IP3 formation

and then regeneratively propagate them by Ca2þ-induced Ca2þ release. This dual

regulation of IP3R allows a hierarchical recruitment of Ca2þ release events as the

stimulus intensity increases (Bootman et al., 1997; Marchant and Parker, 2001).

Single IP3R respond first, then several IP3R within a cluster open together to give

larger local events (‘‘puVs’’), and as puVs become more frequent, they ignite

regenerative Ca2þ waves (Bootman and Berridge, 1995; Marchant et al., 1999).

This hierarchy of events allows Ca2þ to function as a local or global messenger, a

feature that underlies its versatility (Berridge et al., 2000). A key point for the

present discussion is that local events involving very few IP3R underlie the Ca2þ

signals that regulate cellular activity. By contrast, for most ion channels, it is the

collective behavior of large numbers of channels, the macroscopic current, that

determines the physiological response, a change in membrane potential, or

8. Patch-Clamp Recording of IP3 Receptors 201

transcellular ion flux, for example. The distinction highlights the particular impor-

tance of single-channel recording in the analysis of IP3R. As with all such analyses,

they provide the highest resolution insight into the opening and closing of single

channels and can thereby reveal details of gating mechanisms (Colquhoun, 2007;

Sivilotti, 2010). But for IP3R and other Ca2þ channels too, openings of individual

channels are the physiologically significant behavior.

The patch-clamp technique, developed originally by Neher and Sakmann (1976)

with subsequent improvements (Hamill et al., 1981), is the most powerful means of

studying the behavior of ion channels in their native environment. It involves

recording currents passing through an electrically isolated, small area (‘‘patch’’)

of biological membrane in response to an applied voltage or ionic gradient

(Fig. 1A and B). Isolation of the patch is achieved by pressing a polished glass

pipette tip of �1 mm diameter (containing electrolyte solution) against the cell-

surface and applying gentle suction to form a very high-resistance ‘‘giga-Ohm’’

(GO) seal (Hamill et al., 1981). The tight seal is crucial because it isolates the patch

both electrically and physically, so reducing background noise and allowing single-

channel events to be resolved (Hamill et al., 1981). Because of the electrical

isolation and low resistance of the patch-pipette relative to the membrane, a

patch can be voltage-clamped by simply applying a potential to the pipette.

These patch-clamp recordings allow the openings and closings of individual chan-

nels to be resolved with submillisecond temporal resolution under optimal condi-

tions (Fig. 1C). The amplitudes of these tiny currents and their dependence on

applied potential and ion concentrations allow the ion selectivity and rates of ion

permeation to be determined. Lurking within the pattern of stochastic openings

and closings is the information from which the sequence of events that leads to

channel gating and desensitization/inactivation can be reconstructed. Comprehen-

sive descriptions of the patch-clamp technique are available from the original

articles (Hamill et al., 1981; Neher and Sakmann, 1976) and subsequent reviews

(Ogden, 1994; Sakmann and Neher, 1995). However, most IP3R are expressed in

membranes of the ER, where they are inaccessible to conventional patch-clamp

techniques. Alternative approaches are therefore needed.

II. Nuclear Patch-Clamp Recording

It is impracticable, despite one heroic success recording from IP3R within the

ER of an intact cell (Jonas et al., 1997), to use patch-clamp techniques routinely to

record the behavior of single channels within the membranes of intracellular

organelles in situ. A more promising approach for single-channel recordings

in situ is the ‘‘optical patch-clamp,’’ where high-resolution optical microscopy in

combination with fluorescent Ca2þ indicators is used to measure the Ca2þ signals

evoked by opening of single or small clusters of IP3R (Demuro and Parker, 2007;

Smith and Parker, 2009). Presently, however, these optical methods can be used

only to measure fluxes through Ca2þ channels, and they lack the temporal

C

O1

O2

20pA

500 msK+

Cell-attached

Inside-out Outside-out

Whole-cell

A

Referenceelectrode

Patch pipette

Ionchannel

Bath

Patch-clampamplifier

B

CIP3

Fig. 1 Conventional patch-clamp recording. (A) A polished glass pipette forms a tight seal against a

biological membrane isolating an area across which the tiny currents passing through small numbers of

open channels can be recorded. (B) DiVerent configurations of conventional patch-clamp recording.

Beginning with a cell-attached patch, an inside-out excised patch can be produced by pulling the patch-

pipette away from the cell, while suction or a strong brief voltage-pulse ruptures the underlying

membrane to give the whole-cell configuration. An outside-out patch can then be produced by pulling

the patch-pipette away from the whole-cell configuration. (C) Typical whole-cell recordings of IP3R3

expressed in the plasma membrane of DT40-KO cells expressing rat IP3R3. PS included IP3 (10 mM),

ATP (5 mM), and a free [Ca2þ] of �200 nM; Kþ was the charge carrier and the holding potential was

�100 mV. C, O1, and O2 show the closed state and the openings of 1 and 2 IP3R, respectively.



resolution of conventional patch-clamp recording. Measuring the electrical activity

of intracellular channels, therefore, presently relies upon redirecting channels to the

plasma membrane, where they then become accessible to conventional patch-clamp

techniques (Xu et al., 2007); reconstituting the channel into an artificial membrane;

or isolating organelles that express the channel and adapting the patch-clamp

technique to record from these membranes. The latter has been used, for example,

to resolve the behavior of the mitochondrial Ca2þ uniporter from mitochondria

stripped of their outer membrane (‘‘mitoplasts’’) (Kirichok et al., 2004) and for

single-channel recordings of the endolysosomal protein, TRPML1, from artificially

enlarged lysosomes (Dong et al., 2008). All three methods have been used to record

single-channel behavior of IP3R.

We observed that DT40 cells express very small numbers of functional

IP3R within the plasma membrane (Dellis et al., 2006) and because DT40 cells

lacking native IP3R are available (Section III), conventional whole-cell patch-

clamp recording has been used by us (Dellis et al., 2008) and others

(Betzenhauser et al., 2008b, 2009a) to examine the behavior of recombinant and

mutant IP3R. Typical recordings from IP3R in the plasma membrane of DT40 cells

are shown in Fig. 1C. A limitation of this approach is that excised patch-clamp

recording, where the ‘‘intracellular’’ composition can be precisely controlled, is

impracticable (because plasma membrane IP3R are too scarce), and with whole-

cell recording, it is diYcult to define reliably the exact concentration of IP3 bathing

the IP3R (Dellis et al., 2006). Detailed descriptions of the methods used for whole-

cell recording of IP3R expressed in the plasma membrane of DT40 cells have been

published (Dellis et al., 2006; Taylor et al., 2009b). We prefer nuclear patch-

clamping (see below) to whole-cell recording because the nuclear envelope is

continuous with the ER (Fig. 2A), wherein reside most IP3R, and it is practicable

to work with excised patches that provide better control of media bathing both

sides of the membrane.

The first electrical recordings from IP3R were made by incorporating native or

purified IP3R into artificial lipid bilayers (Bezprozvanny et al., 1991; Ehrlich and

Watras, 1988; Maeda et al., 1991; Mayrleitner et al., 1991). As with all reconsti-

tuted systems, the lipid composition of the bilayer and the steps involved in

isolating, purifying, and reconstituting IP3R into the bilayer may aVect normal

function of the channel, not least its regulation by accessory proteins (Boehning

et al., 2001a; Foskett et al., 2007; Patterson et al., 2004). Anecdotally, and it

equates with our experience, it seems to be more diYcult to obtain bilayer record-

ings from IP3R than from its close relatives, the ryanodine receptors (Williams,

1995).

Many of the problems with bilayer recording are resolved by using nuclear patch-

clamp recording (Fig. 2). This technique was first introduced in the early 1990s

(Matzke et al., 1990; Mazzanti et al., 1990, 2001; Tabares et al., 1991) and subse-

quently, applied by the laboratories of Clapham (Stehno-Bittel et al., 1995) and

Foskett (Mak and Foskett, 1994) to record single-channel events from native

IP3R in nuclei from Xenopus oocytes. It has, subsequently, been successfully applied

Cytoplasm-out

A B

C

On-nucleus

ONMINM

Lumen-out

Cell

Nucleus

ONMINM

Nucleoplasm

10mm

C

10 pA

500 ms K+

IP3E

Nucleus

Debris

D

K+

Cytosol

Lumen

PS

+40 mV

BS

Fig. 2 Nuclear patch-clamp recording of IP3R. (A) The nuclear envelope comprises an inner (INM)

and outer (ONM) membrane surrounding a luminal space that is continuous with the lumen of the ER.

The continuity of the ONMwith the ERmembrane allows some ER proteins to invade the ONM, where

they become accessible to nuclear patch-clamp recording. (B) Phase-contrast image of a DT40 nuclear

preparation showing nuclei, one of which has debris attached, and an intact cell. (C) Three recording

configurations are used for nuclear patch-clamp recording. The on-nucleus, lumen-out, and cytosol-out

excised patch configurations are analogous to the cell-attached, inside-out, and outside-out configura-

tions of conventional patch-clamp recording (Fig. 1B). (D) An excised lumen-out nuclear patch

illustrating the convention used to report membrane potential. (E) Typical recording from a single

IP3R3 recorded in the lumen-out configuration from the nucleus of a DT40-KO cell stably expressing

rat IP3R3. PS included IP3 (10 mM), ATP (5 mM), and a free [Ca2þ] of �200 nM; Kþ was the charge

carrier and the holding potential was þ40 mV. C denotes the closed state.


to analyses of diVerent IP3R subtypes expressed in diVerent cells, including COS-7

cells (Boehning et al., 2001a), insect Sf9 cells (Ionescu et al., 2006), smooth muscle

(Kusnier et al., 2006), DT40 cells (Betzenhauser et al., 2009b; Dellis et al., 2006;


Rahman et al., 2009), human B lymphoblasts (Cheung et al., 2008), cerebellar

Purkinje cells (Marchenko et al., 2005), and embryonic cortical neurons and fibro-

blasts (Cheung et al., 2008).

The utility of nuclear patch-clamp recording derives from the fact that the outer

nuclear envelope is continuous with the ER membrane (Dingwall and Laskey,

1992) (Fig. 2A). Channels that are normally expressed within ER membranes can,

therefore, pass into the outer nuclear envelope, allowing their activity to be

recorded from patches of nuclear membrane in a near-physiological setting

(Fig. 2). Abundant nuclear pore complexes, each with a large central conduit

linking cytoplasm and nucleoplasm (Mazzanti et al., 2001), might have been

expected to compromise formation of the tight seals required for patch-clamp

recording or at least pollute recordings from conventional channels with lesser

conductances. In practice nuclear pores appear not to cause problems. It seems

unlikely, though it remains possible, that this results from patching onto ER

overlying the nuclear envelope, rather than the envelope itself. It is perhaps more

likely that patches that include nuclear pores are rejected because they fail to form

giga-Ohm seals, or the high-Kþ medium used for nuclear patch-clamp recording

favors closure of nuclear pores (Bustamante and Varanda, 1998).

III. Choice of Cells for Analyses of IP3R

Almost all animal cells express IP3R, most express more than one of the three

vertebrate gene products, and substantial alternative splicing and posttranslational

modifications add further to the diversity of subunits from which IP3R are assem-

bled (Foskett et al., 2007). Assembly of these subunits into homo- and heterote-

trameric structures increases the diversity of functional IP3R enormously (Joseph

et al., 1995, 2000; Wojcikiewicz and He, 1995). Despite this complexity, there have

been many valuable studies of the single-channel behavior of native IP3R in, for

example, Xenopus oocytes, nuclei from oocytes, insect Sf9 cells, and cerebellar

Purkinje neurons, and of native IP3R reconstituted into lipid bilayers (Section II).

But the limitations of such studies are obvious when it to comes to exploring the

structural basis of IP3R activation. This demands a more homogenous population

of IP3R with a defined structure and ideally expressed in a native membrane.

At present, only one expression system provides the ‘‘null background’’ that allows

these demanding criteria to be satisfied: DT40 cells (Fig. 3).

DT40 cells originate from an avian leukosis virus-transformed bursal B cell

(Baba et al., 1985). The uniquely valuable feature of these cells is the unusually

high frequency with which they integrate targeted DNA constructs into their

genome (Buerstedde and Takeda, 1991). This feature, together with the shorter

introns of avian genes, allows targeted disruption of specific genes and has ensured

widespread use of DT40 cells for ‘‘gene-knockouts.’’ In a monumental eVort,Kurosaki and his colleagues used targeted gene disruption to inactivate both

copies of all three IP3R genes in DT40 cells and thereby to generate the first cell

DT40-KO

A B

−10 −8 −6 −4

0

50

100DT40-R3

Log {[IP3], M}

Ca2+

rel

ease

(%

)

- CDT40-KO

- C

DT40-R3

10 pA

1 sK+

IP3

Fig. 3 DT40-KO cells provide a null background for expression of functional IP3R. (A) IP3-evoked

Ca2þ release from permeabilized DT40 cells assessed using a luminal Ca2þ indicator (Tovey et al., 2006).

Permeabilized DT40-KO cells stably expressing rat IP3R3 (DT40-R3) release Ca2þ when stimulated

with IP3, whereas DT40-KO cells are unresponsive. (B) Currents recorded from lumen-out patches from

DT40-KO and DT40-R3 cells. PS included IP3 (10 mM), ATP (5 mM), and a free [Ca2þ] of �200 nM;

Kþ was the charge carrier and the holding potential was þ40 mV. C denotes the closed state.


line lacking functional IP3R (Sugawara et al., 1997). These DT40-KO cells, which

Kurosaki (RIKEN, Japan) has made widely available, provide the only null

background for functional expression of IP3R (Fig. 3). They have been extensively

used by many groups to express each of the three IP3R subtypes and define their

functional properties, and to explore the role of IP3R in many higher order

processes (e.g., Joseph and Hajnoczky, 2007; Miyakawa et al., 1999). A recent

review provides further details of the use of DT40 cells for analyses of Ca2þ

signaling pathways (Taylor et al., 2009b). Here, we describe our use of DT40

cells expressing mammalian IP3R for nuclear patch-clamp recording.

IV. Methods

A. Culture of DT40 Cells

Wild-type DT40 cells (Cell Bank number RCB1464) and DT40-KO cells

(RCB1467) are available from Riken Bioresource Center Cell Bank, Japan

(http://www.brc.riken.jp/lab/cell/). Cells are grown in RPMI 1640 medium (Invi-

trogen) supplemented with 10% fetal bovine serum (FBS, Sigma), 2 mM l-gluta-mine, 1% chicken serum (Sigma), and 50 mM b-mercaptoethanol (Invitrogen) in a

humidified atmosphere containing 5% CO2, ideally at 39–41 �C (matching the

increased body temperature of birds). It is, however, acceptable and more conve-

nient when incubators are shared with mammalian cells to culture DT40 cells at

37 �C without loss of viability. The only obvious eVect is a slowing of growth rate;

the doubling time has been reported to increase from about 10 h at 39–41 �C to

18 h at 37 �C (Mak et al., 2006). The chicken serum must be heat-inactivated

http://www.brc.riken.jp/lab/cell/


(56 �C for 30 min). We avoid antibiotics, but addition of penicillin (10,000 units/

ml) and streptomycin (10 mg/ml) to culture media is optional (Winding and

Berchtold, 2001). The cells grow in suspension in flasks, Petri dishes, or multiwell

plates (Greiner Bio-one). Cells are passaged by 20-fold dilution every 2–3 days

when they reach a density of about 2�106 cells/ml. Avoid growing cells beyond a

density of 2.5�106 cells/ml. Cells (2�106ml�1) in either culture medium or FBS

supplemented with 10% dimethylsulfoxide (DMSO, Sigma) can be frozen and then

stored in liquid nitrogen following standard procedures. We routinely culture cells

for 30–35 passages, before thawing a new frozen stock. For the latter, 1 ml of cells

is dispensed into 20 ml of medium. We find it unnecessary to remove residual

DMSO at this stage. Cells are then passaged after 24 h.

DT40 cells are not easy to transfect with IP3R expression constructs, we there-

fore use cell lines stably expressing rat IP3R1 (GenBank accession number of

GQ233032.1), mouse IP3R2 (AB182290), and rat IP3R3 (GQ233031.1). Details

of the methods and sources of the original clones are provided in previous pub-

lications (Dellis et al., 2006; Rahman et al., 2009; Rossi et al., 2009; Tovey et al.,

2010). Briefly, DT40-KO cells are transfected by nucleofection with linearized

constructs of pcDNA3.2-IP3R using solution T and program B23 (Amaxa) using

5 mg DNA/106 cells. G418 (Invitrogen, 2 mg/ml) is used for selection. Expression

of IP3R in each cell line is quantified by immunoblotting using custom-made anti-

peptide antisera (Cardy et al., 1997; Dellis et al., 2006; Rossi et al., 2009; Tovey

et al., 2010) and, where needed, by3

H-IP3 binding. Functional expression of

IP3R in each DT40 cell line is verified by comparison with DT40-KO cells using

a luminal Ca2þ indicator and a high-throughput assay for IP3-evoked Ca2þ release

(Laude et al., 2005; Tovey et al., 2006) (Fig. 3). Only cell lines shown to express

functional IP3R are used for nuclear patch-clamp recording.

B. Isolation of Nuclei

Several methods have been described for isolation of nuclei; most rely on a

combination of osmotic and mechanical lysis of cells (Boehning et al., 2001a;

Bustamante, 1994; Franco-Obregon et al., 2000; Marchenko et al., 2005). Our

protocol is adapted from that of Boehning et al. (2001a). DT40 cells expressing a

recombinant IP3R (DT40-IP3R cells, 1.5–2�106 cells/ml) are centrifuged (500�g

for 2 min at 4 �C), washed once with ice-cold phosphate-buVered saline (PBS), and

then once with cold nuclear isolation medium (NIM). PBS has the following

composition: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2PO4, KH2PO4, pH 7.4,

with NaOH. NIM comprises: 250 mM sucrose, 150 mM KCl, 3 mM b-mercap-

toethanol, 10 mM Tris–HCl, 1 mM phenylmethanesulfonyl fluoride (PMSF,

Sigma), pH 7.5. Cell pellets are resuspended in NIM supplemented with complete

protease inhibitor cocktail (Roche, 1 mini-tablet/20 ml) and stored on ice for up to

4–5 h. For isolation of nuclei, 1 ml of the cell suspension is homogenized with 3–4

strokes of a Dounce homogenizer (Wheaton Industries, Inc.), which lyses about

5–10% of cells, assessed by staining with Trypan Blue (0.001%). This crude lysate


containing some isolated nuclei is stored in NIM on ice and used within 1 h for

patch-clamp experiments. Because the activity of nuclear IP3R has been reported

to decrease after �40 min at 20 �C (Boehning et al., 2001a), we routinely prepare

fresh nuclei at hourly intervals. DT40 cells are not much larger than their nuclei

(Fig. 2B). The inexperienced eye may therefore find it diYcult to distinguish nuclei

from cells. But nuclei rarely have the smooth surface of intact cells and they often

only partially protrude from broken cells, where the relatively clean exposed

surface allows formation of a giga-Ohm seal. The yield of nuclei can be substan-

tially increased to 50–60% using methods that require incubation in hypo-osmolar

media (Franco-Obregon et al., 2000), but we rarely detect active IP3R after such

isolation procedures. A nuclear isolation kit (Sigma Nuclei EZ Prep) also provides

nuclei in high yield (�85%), but we rarely succeed in forming giga-Ohm seals with

these nuclei. In practice, the low yield of nuclei with our protocol is not a limitation

for nuclear patch-clamp recording.

IP3R have also been reported to be expressed within the inner nuclear membrane

(Humbert et al., 1996; Marchenko et al., 2005). The citrate treatment used to

remove the outer nuclear membrane and so allow patch-clamp recording of

IP3R within the inner membrane (Marchenko et al., 2005) appears not, at least

in our experience, to be readily applicable to DT40 cell nuclei.

C. Solutions for Patch-Clamp Recording

For most recordings, we use Kþ as the charge-carrier. This eliminates the

complexity of having Ca2þ passing through the IP3R regulate its activity, and it

provides larger single-channel currents than with bivalent cations (Rahman et al.,

2009). The bath solution (BS), which bathes the luminal surface of the nuclear

envelope, typically contains 140 mM KCl, 10 mM HEPES, 100 mM 1,2-bis(2-

aminophenoxy)ethane-N,N,N0,N0-tetraacetic acid (BAPTA, tetra potassium salt,

Calbiochem), and a free [Ca2þ] of �200 nM (total CaCl2, 51 mM) adjusted to pH

7.1 with KOH. The usual pipette solution (PS), which bathes the cytosolic surface

of the membrane, contains 140 mM KCl, 10 mM HEPES, 500 mM BAPTA,

Na2ATP (0.5 mM), IP3 (American Radiolabeled Chemicals, Inc.), and a free

[Ca2þ] of �200 nM (total CaCl2, �254 mM) adjusted to pH 7.1 with KOH.

IP3, Ca2þ, and ATP are the three ligands of IP3R whose concentrations must be

adjusted to obtain optimal IP3R activity in patch-clamp recording (Foskett et al.,

2007). The concentration of IP3 in PS can be varied between experiments, depend-

ing on the aim of the analysis; 10 mMwill usually be suYcient to saturate responses

to IP3 (Foskett et al., 2007; Rahman et al., 2009). The potentiating eVects of ATP

diVer between IP3R subtypes with higher concentrations required optimally to

activate IP3R3 (Betzenhauser et al., 2008a; Miyakawa et al., 1999). The pH of PS

must be readjusted after addition of ATP, and its eVects on free [Ca2þ] also need to

be considered. Finally, because Mg2þ aVects the conductance of IP3R (Mak and

Foskett, 1998; Rahman and Taylor, 2009), it is advisable to use ATP of the highest


purity. Freshly prepared dilutions of ATP and IP3 (from frozen stocks) are added

to PS as required.

EVective buVering of the free [Ca2þ], which might reasonably be varied between

nanomolar and several micromolar, requires buVers with appropriate aYnities for

Ca2þ (Patton et al., 2004). For free [Ca2þ] less than �1 mM, BAPTA

(KDCa�192 nM at pH 7.4) is preferable to EGTA because it has faster Ca2þ-

binding kinetics and lesser pH-dependence. Where the free [Ca2þ] of PS is

1–100 mM, we use 5,50-dibromo BAPTA (KDCa�1.83 mM, Fluka), EGTA

(KDCa�67 nM), and/or N-(2-hydroxyethyl)ethylenediamine-N,N0,N0-triacetic

acid (HEDTA, KDCa�2.2 mM, Sigma), alone or in appropriate combinations

(Bers et al., 1994). We initially estimate the amount of CaCl2 required to achieve

the desired free [Ca2þ] using WinMaxC software (http://www.stanford.edu/�-

cpatton/maxc.html) and then measure the free [Ca2þ] of the final media (supple-

mented with ATP, IP3, etc.) directly using either a fluorescent Ca2þ indicator

(Fluo-3, KDCa¼325 nM, Invitrogen) or a Ca2þ-sensitive electrode (Mettler

Toledo Ingold, Fisher Scientific) for higher free [Ca2þ] (Dellis et al., 2006;

Rahman et al., 2009).

The osmolarities of all solutions are adjusted to �290–310 mOsm kg�1 using

glucose and mannitol, and verified using a vapor pressure osmometer (Wescor,

Inc.). This is more important for recordings in the on-nucleus configuration than

for recordings from excised patches (Fig. 2C). PS is prepared to be slightly (�10%)

hypo-osmolar to BS to aid formation of giga-Ohm seals (Hamill et al., 1981). All

recording solutions are filtered using detergent-free 0.2-mm filters (AcrodiscÒ

syringe filters, Pall Corporation) (Ogden, 1994). Fresh recording solutions (with-

out added IP3 or ATP) are prepared monthly and stored at 4 �C.The presence within the nuclear envelope of other large-conductance cation and

Cl� channels (Franco-Obregon et al., 2000; Marchenko et al., 2005; Mazzanti

et al., 2001; Tabares et al., 1991) might potentially contaminate recordings of

nuclear IP3R. In practice, this appears not to be a significant problem. If such

problems should arise, they can be mitigated by replacing KCl in BS and PS with

cesium methanesulfonate (CsCH3SO3): Csþ permeates IP3R but not Kþ channels

(Tovey et al., 2010), while most anion channels are impermeable to CH3SO3�.

D. Patch-Clamp Recording

The equipment required for nuclear patch-clamp recording is the same as that

used for conventional patch-clamp recording (Fig. 1A). The basic rig includes an

amplifier, headstage, electrode holder, micromanipulator, AgCl bath electrode,

data acquisition system (i.e., analog-to-digital converter, computer, and software),

inverted microscope, air table, and a Faraday cage. In addition, a pipette puller

and fire-polisher or microforge are required to fabricate electrodes. Optional

extras include systems for exchange of solutions, an oscilloscope, and a low-pass

8-pole Bessel filter; the latter extends the filtering range down to 0.1 Hz from the

1 to 100 kHz provided by the inbuilt filter. Comprehensive descriptions of the

http://www.stanford.edu/~cpatton/maxc.html




equipment used for patch-clamp recording are presented in relevant chapters of

Sakmann and Neher (1995). Among the many steps taken to minimize electrical

noise, the following are particularly important: appropriate grounding of equip-

ment, use of thick-walled glass capillaries, filling pipettes with PS to the minimal

level required to contact the recording electrode, and minimizing immersion of the

pipette in BS (Rae and Levis, 1992).

Pipettes are pulled from filamented, thick-walled borosilicate glass capillaries

(GC150-10F, Clark Electromedical Instruments) using a Flaming/Brown P-87

horizontal micropipette puller (Sutter Instruments), and then fire-polished to a

tip diameter of �1 mm using a microforge (MF-830, Narishige). It is advisable to

melt a small bead of glass onto the wire of the forge to prevent platinum vapor

from reaching the pipette tip. With monovalent cations as charge carriers, the

single-channel conductance (g) of IP3R is large enough (�360 pS, Section IV.E) to

achieve good signal-to-noise ratios without hydrophobic coating of the patch-

pipette (Penner, 1995). But when g of IP3R is reduced, with pore mutants or with

Ca2þ or Ba2þ as charge carriers, for example, it may be necessary to coat pipette

tips with SylgardTM

(Dellis et al., 2006, 2008). When filled with PS, the pipette

resistance typically remains within the range of 15–20 MO. Pipettes are best

prepared a few hours before experiments. Unused pipettes can, however, be stored

in an air-tight container and used later, but it is advisable to repolish them lightly

before use to remove any impurities accumulated during storage.

Petri dishes are precoated with poly-l-ornithine or poly-l-lysine (0.01%, Sigma)

for 1–2 h, then rinsed twice with deionized water and air-dried. The nuclear

preparation (15 ml) is added to a Petri dish containing BS (1.5 ml) and the cells/

nuclei are allowed to adhere. The dish is then mounted on the stage of an inverted

microscope (Zeiss Axiovert 100) coupled to an assembly of headstage (CV 203 BU,

Molecular Devices) and micromanipulator (PCS-1000, Burleigh Instruments).

Recordings are made at room temperature (�20 �C) in the on-nucleus or excised

configuration (Fig. 2C). The latter is preferable because it allows control of the

medium on both sides of the membrane and eVective control of the voltage acrossthe patch.

A nucleus largely free from debris is first identified (Fig. 2B) and the patch-

pipette is positioned, using the micromanipulator, with its tip just above the

nucleus. A slight positive pressure is applied to the inside of the patch-pipette

before dipping it into the BS to avoid dirt accumulating at the pipette tip and to

prevent backflow of BS into the PS (Hamill et al., 1981). After dipping the pipette

into BS, the pipette capacitance is compensated using the specific oVset on the

amplifier and the pipette resistance (typically �10–15 MO) is noted. As the pipette

tip approaches the nucleus, the positive pressure is relieved. Taking care not to

puncture the nuclear membrane, the pipette is lowered until it contacts the mem-

brane, which should increase the pipette resistance by at least 2 MO. A giga-Ohm

seal (�5 GO) usually forms within a few seconds of applying slight negative

pressure, by suction, to the inside of the pipette; this is usually controlled by an

attached 50-ml syringe or by mouth. Seal formation can sometimes be facilitated


by applying a holding potential of about �40 mV once a high-resistance contact is

established (Ogden, 1994), and occasionally giga-Ohm seals form spontaneously.

Seals >5 GO can be routinely obtained with mild suction providing the nucleus is

immobile and free of debris.

To formanexcised patch, our preferred recording configuration, the patch is pulled

from the nucleus after forming the giga-Ohm seal (Fig. 2C). To prevent formation of

closedvesicles at the tipof thepatch-pipette, excisedpatchesarebriefly (1–2 s) exposed

to air and then reimmersed in BS (Hamill et al., 1981). Prewritten protocols are then

used to record currents through the excised patch at diVerent holding potentials. Thebath electrode is grounded (i.e., 0 mV) and for convenience, the potential across a

nuclear patch (whether attached or excised) is defined as the pipette potential minus

thebathpotential. That is,with symmetricalmedia, a positiveholdingpotentialwould

favor movement of cations from PS (the cytosolic surface) into BS (the luminal

surface) producing an outward current and an upward deflection on the channel

record (Fig. 2D and E) (Franco-Obregon et al., 2000; Mak and Foskett, 1994;

Rahman et al., 2009). For determination of current–voltage (I–V) relationships, and

thereby the single-channel conductance (g) of the channel (Section IV.E), the voltage

across the excised patch can be stepped from�60 toþ60 mV in increments of 20 mV

fromaholding potential of 0 mV.Applyingmore extreme voltages aVects the stabilityof thenuclear patch.For all other experiments, includingkinetic analyses, currents are

typically recorded atþ40 mV for between 1 and 10 min.

Currents are amplified with an Axopatch 200B amplifier in its voltage-clamp

mode, filtered at 1 kHz with a low-pass 4-pole Bessel filter (built into the amplifier),

and digitized at 10 kHz with a Digidata 1322A interface using the PC-based

acquisition software package pClamp 9.2 (Molecular Devices) (Colquhoun,

1994). This filtering, while it inevitably causes some loss of information, has the

eVect of rejecting signals (background noise) that are too brief to reflect the gating

of IP3R. If the filtering frequency is set too low, it will reject events that do reflect

gating of channels, and if set too high, background noise will obscure the openings.

The optimal filtering frequency is, therefore, a compromise that depends upon the

noise and time-course of the channel events; it needs to be optimized empirically.

The sampling rate must, of course, exceed the filter frequency if further valuable

information is not to be lost as the signals are digitized. In practice, digitization

should be 10–20 times faster than the cutoV or ‘‘corner’’ frequency of the filter

(Colquhoun, 1994). Most nuclear patch-clamp studies of IP3R have used 1-kHz

filtering (Dellis et al., 2006; Ionescu et al., 2006; Mak and Foskett, 1997; Rahman

et al., 2009). For presentation, traces can be further filtered oZine using a Gaussian

filter (built within ClampFit).

For the determination of relative permeabilities to cations, asymmetric record-

ing solutions are used. For example, the normal BS can be replaced by a Ba2þ-richBS (50 mM BaCl2, 30 mM KCl, 10 mM HEPES, adjusted to pH 7.1 with KOH),

while the PS remains unchanged (Boehning et al., 2001a; Dellis et al., 2006). The

liquid junction potential (LJP) under this asymmetric condition can be predicted

(5.2 mV at 20 �C), using the ‘‘junction potential calculator’’ (JPCalc, within


pClamp 9.2), which uses the generalized Henderson equation (Barry, 1994). The

calculated LJP is then subtracted from the observed reversal potential (Erev)

obtained from the current–voltage (I–V) plot. We return in Section IV.E, after

considering analysis of the raw traces, to describe how Erev allows the relative

permeability of IP3R to diVerent cations to be calculated.

In both the on-nucleus and excised patch configurations described above, the

cytoplasmic surface of the IP3R lies within the patch-pipette (Fig. 2C); it is, therefore,

diYcult to change the IP3 concentration once the giga-Ohm seal has formed. It is

possible, though diYcult, to perfuse a patch-pipette and thereby to vary the composi-

tion of the ‘‘cytosolic’’ medium while recording channel activity (Hering et al., 1987;

Maathuis et al., 1997), but this technique has not yet been applied to IP3R. Other

options include the cytoplasm-out configuration of nuclear patch-clamp recording

(Fig. 2C), which has been successfully applied to analyses of IP3R in Sf9 cells (Mak

et al., 2007). Alternatively, flash-photolysis of caged-IP3 within the patch-pipette in

either the on-nucleus or excised nuclear patch configuration can be used rapidly to

increase the IP3 concentration bathing the cytosolic surface of the IP3R once the

recording is underway (Rahman et al., 2009). For these flash-photolysis experiments,

pipettes are prepared from thin-walled, nonfilamented borosilicate glass capillaries

(Harvard Instruments) and PS includes d-myo-inositol 1,4,5-trisphosphate, (4,5)-1-

(2-nitrophenyl) ethyl ester (caged-IP3, �100 mM, Calbiochem). After recording for

30–60 s, IP3 can thenbe released intoPSbyphotolysis of caged-IP3usinga singlehigh-

intensity flash (1 ms) from a Xe-flash lamp (XF-10, Hi-Tech Scientific; 240 J with the

capacitor charged to 385 V) passed through a filter (300–350 nM) (Walker et al.,

1987). A problemwith this approach is the diYculty in assessing the concentration of

IP3 to which the IP3R are exposed after flash-photolysis of caged-IP3.

E. Analysis of Single-Channel Records

Two sorts of information can be extracted from single-channel records: the

properties of the open channel (its ability to conduct diVerent ions); and the

sequence of stable states through which the channel passes as it moves between

closed, open, and desensitized conditions. Here, we provide only a brief introduc-

tory summary of the methods used to extract this information from the openings

and closings of channels resolved by patch-clamp recording. The reader interested

in more rigorous and detailed descriptions is advised to begin with two excellent

books (Ogden, 1994; Sakmann and Neher, 1995).

Several software packages are available for analysis of electrophysiological

records. These include ClampFit, which includes the pClamp suite (Molecular

Devices), and DC-soft, which includes SCAN, EKDIST, and HJCFIT (http://

www.ucl.ac.uk/Pharmacology/dcpr95.html); QuB (www.qub.buValo.edu), Pulse/

Patchmaster (HEKA Elecktronik), Tac (Bruxton Inc.), and the Strathclyde Electro-

physiology Software (http://spider.science.strath.ac.uk/sipbs/software_ses.htm).

We use ClampFit and QuB for analyzing records (Dellis et al., 2006; Rahman

et al., 2009).

http://www.ucl.ac.uk/Pharmacology/dcpr95.html

http://www.ucl.ac.uk/Pharmacology/dcpr95.html

http://www.qub.bu&ff;alo.edu

http://spider.science.strath.ac.uk/sipbs/software_ses.htm

0.0

0.0

Cou

nt/to

tal

D

A

B


Drifting baseline in current traces is first checked and corrected manually using

ClampFit. Current-amplitude histograms of recordings with no obvious sub-con-

ductance states are measured with a half-amplitude threshold-crossing criterion

using ClampFit (Sachs et al., 1982; Colquhoun, 1994). Only events lasting longer

than twice the filter rise time (tr¼0.3321/fc, where fc is the cutoV frequency of the

filter) can reach their full amplitude and so be reliably measured in a threshold-

crossing-based idealization procedure (Colquhoun, 1994) (Fig. 4A and B).

In recordings with 1-kHz filtering, the predicted filter rise time (assuming the filter

behaves as a Gaussian filter) is 332 ms. Events lasting<1 ms are, therefore, omitted

from the amplitude histograms. The peaks of the binned current amplitude histo-

grams are fitted in ClampFit by sums of the appropriate number of Gaussian

Log (duration, ms)

0

3

6to= 10 ms

Cou

nt/to

tal

−1 0 1 2

Log (duration, ms)

1 0 1 20

0.02

0.04

0.06 tc1= 1.14 ms(88%)

tc2= 92 ms(12%)

C

0 5 100

500

1000

Num

ber

of e

vent

s

Amplitude (pA)

C

O

C

O

10 pA

100 ms

Fig. 4 Analysis of the behavior of single IP3R from nuclear patch-clamping recording. (A) Fragment

of a raw record from an excised lumen-out nuclear patch with a single functional IP3R3. The recording

conditions were identical to those shown in Fig. 3B. (B) The idealized record of the same trace produced

as described in the text (Section IV.E). This idealized record is used for all subsequent analyses. (C) All-

points current amplitude histogram, showing two peaks, one at 0 pA (the closed state) and a second at

�10 pA (the single open state of one IP3R). (D) Distribution of the open (top) and closed (bottom) life-

times shown as Sigworth–Sine plots (Sigworth and Sine, 1987). The plots suggest a single open state with

to of 10 ms, and two closed states with tc of 1 and 92 ms.


probability density functions (pdfs) (Fig. 4C). Mean current amplitudes are plotted

against the corresponding applied potentials to create I–V curves (see Fig. 5).

The unitary conductance (g) and reversal potentials (Erev) are derived by

linear least-square regression analysis using statistical package such as Prism 5

(GraphPad Software, Inc.) or Origin Pro 7.5 (OriginLab Corporation).

Because g is a fundamental property of any ion channel, reflecting interactions

between permeating ions and the residues that form the pore and lead to it, I–V

relationships are often used as ‘‘fingerprints’’ to help identify channels. With Kþ as

charge carrier, the I–V relationships of mammalian IP3R in excised nuclear patches

from DT40-KO cells are linear across a range of applied potentials (�60 to

þ60 mV) (Fig. 5). From the slopes of these I–V plots, we have consistently

observed two populations of IP3R with unitary Kþ conductances (gK) of either�120 or �200 pS (Dellis et al., 2006; Rahman and Taylor, 2009; Rahman et al.,

2009) (Fig. 5B). Neither current was detected in the nuclear envelope of DT40-KO

cells or from DT40-KO cells stably expressing IP3R in the absence of IP3, or with

IP3 in the presence of a competitive antagonist. These values of gK are lower than

reported (�320–360 pS) for IP3R in the nuclear envelope of mammalian cells

(Foskett et al., 2007), but there is wide variation in published values (from �9 to

�480 pS) (Cheung et al., 2010; Rahman and Taylor, 2009). The reason for these

disparities is unresolved, but it may reflect variable amounts of free Mg2þ in PS

causing a reduction in gK (Mak and Foskett, 1998; Rahman and Taylor, 2009).

It is, however, clear from analyses of I–V relationships that all IP3R have a large gfor monovalent cations and lesser g for bivalent cations (Dellis et al., 2006; Foskett

et al., 2007) (Fig. 5B).

C

500 ms

20 pA

C

C

C

C

−80 −40 40 80

−15

−10

−5

5

10

15

V (mV)

I (pA)

BA

+40 mV

+60 mV

0 mV

−40 mV

−60 mV

Fig. 5 Current–voltage relationship for nuclear IP3R. (A) Currents were recorded from lumen-out

patches excised from the nucleus of DT40-KO cells stably expressing IP3R3. PS included IP3 (10 mM),

ATP (5 mM), and a free [Ca2þ] of �200 nM. Kþ was the charge-carrier and the holding potential was

varied between þ60 and �60 mV as shown. C denotes the closed state. (B) From the slope of the

current–voltage (I–V) relationship, g was 208 pS. Results are means�SEM, n¼4.


Mutations within the putative pore region of IP3R that change g provide direct

evidence that the residues within the P-loop linking the last pair of transmembrane

domains are likely to contribute to the ion-permeation pathway (Boehning et al.,

2001b; Schug et al., 2008). Future work along similar lines is likely to define more

precisely the structural determinants of ion permeation. In addition, our demon-

stration that similar point mutations aVected g of the IP3-activated currents

detected in the plasma membrane of DT40 cells expressing mutant

IP3R provided definitive evidence that the currents were carried directly by IP3R,

rather than by another plasma membrane channel with which IP3R in the ER

might have associated (Dellis et al., 2006).

Resolving the unitary current events associated with opening of individual

IP3R also allows functional IP3R to be counted. The number of active IP3R in a

patch can be estimated from the maximal number of simultaneous openings to the

unitary current level (Horn, 1991) (Figs. 1C and 4C). The likelihood of several

channels opening simultaneously depends upon their Po and the number of chan-

nels (N). We would, for example, need to wait much longer, on average, for six

IP3R with low Po to open simultaneously than for the simultaneous opening of two

IP3R with high Po. We can be confident (p<0.01) that we have detected the entire

complement of active IP3R within a patch, when the recording period is longer

than 5(sNþ1) (Ionescu et al., 2006), where

sN ¼ toN Poð ÞN

" #exp

NtDto

� �ð1Þ

and sN is the mean interval between successive simultaneous openings of all N

IP3R; tD the minimum duration of an open event detectable after filtering

(200 ms in our experiments); and to is the mean channel open time. Confidently,

estimating the number of active IP3R within a patch is important, not the least

because there has been a suggestion that increasing concentrations of IP3 cause

increases in both Po (making it easier to detect simultaneous openings) and the

number of active IP3R (Ionescu et al., 2006). This interesting and unprecedented

behavior, which we fail to see (Rahman et al., 2009), has been invoked to explain

the unusual pattern of quantal Ca2þ release observed for IP3R (Taylor, 1992).

By varying the concentrations of cations on either side of the membrane

(Section IV.D and Fig. 6A), the relative permeability (PBa/PK) can be calculated

using a modified version of the Goldman–Hodgkin–Katz (GHK) equation

(Bezprozvanny and Ehrlich, 1994; Fatt and Ginsborg, 1958):

Erev ¼ RT

2Fln4PBa Ba2þ

� �o

PK Kþ½ ið2Þ

where PBa/PK is the relative permeability to Ba2þ and Kþ, [Kþ]i the [Kþ] in PS,

[Ba2þ]o the [Ba2þ] in BS, Erev the reversal potential (corrected for the LJP, see

Section IV.D), R the universal gas constant, F the Faraday constant, and T is the

C

C

5 pA

500 ms+40 mV

+60 mV

C−60 mV

A B

−80 −40 40 80

−5

5

V (mV)

I (pA)

Fig. 6 Determining the cation-selectivityof IP3R fromnuclear patch-clamprecording. (A)Currentswere

recorded in the same way as described in Fig. 5A, but with the usual PS changed to include Ba2þ (50 mM)

rather thanKþ. The currents recorded at diVerent holding potentials are shown. C denotes the closed state.

(B) I–V relationship showing a reversal potential (Erev) of �23.8�1.4 mV after correction for the liquid

junction potential (Section IV.D). From themodifiedGHKequation, this suggests that the permeability to

Ba2þ relative to Kþ (PBa/PK) is 4.7. The unitary conductance (g) from the slope of the plot is 45�4 pS.


absolute temperature (K). These analyses have established that IP3R (Dellis et al.,

2006; Foskett et al., 2007), like ryanodine receptors (Williams, 2002), are far less

selective (PBa/PK�7) than Ca2þ channels in the plasma membrane (Fig. 6). The

distinction is important because channels within the plasma membrane must be

able to discriminate between the many ions with an electrochemical gradient across

the membrane, whereas Ca2þ is probably the only cation with an appreciable

gradient across the ER membrane (Somlyo et al., 1977).

In addition to revealing the properties of the open pore, single-channel analyses

can also shed light on the steps that lead to its opening. The kinetic analyses of single-

channel records described here require that channel behavior has attained a steady-

state. This is most easily assessed from a stability plot of single-channel open

probability (Po) versus time (Colquhoun, 1994; Weiss and Magleby, 1989). Only

records or parts thereof with an overall steady-state Po should be used for kinetic

analysis. Files with stable baselines are exported asQuB-supported file formats (.ldt).

In QuB, the files are further examined and sections of data with spurious noise are

excluded using the preprocessing module (‘‘Pre’’). Current traces are then idealized

into noise-free, open, and closed transitions using the segmental k-means (SKM)

algorithm in the QuB suite. This uses a hidden Markov model (HMM) to decide

whether each excursion in the record should be classified as an open or closed state

based upon its amplitude (Qin, 2004) (Fig. 4B). The output at this stage is a categori-

zation of every transition into a switch between current amplitudes: a single closed

current amplitude (baseline noise) and one or several amplitudes of the open channel

(s).Where several evenly spaced current amplitudes are detected, it can be diYcult to

resolve whether they arise from openings of several channels or switches between


equally spaced sub-conductance states of a single channel (Rahman and Taylor,

2009). For IP3R, sub-conductance states are rare (Rahman and Taylor, 2009), allow-

ing the simplest possible scheme, a switch between a single closed (C) and open (O)

state (C$O) with arbitrarily chosen rate constants (e.g., 100 s�1), to be used for the

initial idealization (Qin, 2004). Beginning with this simple scheme does not compro-

mise lateranalyses thatmight revealmorecomplex relationshipsbetween several open

and closed states. Direct comparison of raw traces with their idealized versions is

essential at this stage to confirm the fidelity of the idealization procedure.

Hitherto, the analysis, has considered only the amplitudes of the currents, the

next step considers the durations of these events in records from single channels.

This provides the opportunity to resolve diVerent open and closed states and

possible relationships between them, leading to plausible gating schemes. The

distribution of lifetimes of a single state of a channel is described by a single

exponential (Colquhoun, 1994). The analysis attempts iteratively to establish, for

each potential gating scheme (beginning with the simplest, C$O), the number of

exponential functions required to describe the closed and open lifetimes derived

from the idealization procedure. A maximum interval likelihood method (MIL) is

used to fit the lifetimes with pdfs (Qin et al., 1996, 1997, 2000). During this fitting

process, a dead-time of 200 ms (twice the sampling interval) is retrospectively

imposed for the correction of missed events (Sivilotti, 2010).

Dwell-time histograms are generated and displayed with logarithmic abscissa

and square root ordinate (Fig. 4D) (Sigworth and Sine, 1987) and fitted by a

mixture of exponential pdfs, defined in the function f(t) as

fðtÞ ¼Xni¼1

aiti

exp �t=tið Þ ð3Þ

where ai is the fractional area occupied by the ith component in the distribution, such

that the areas corresponding to all components sum to unity, and ti is the time

constant for the ith component. The mean life-time (t) is given by the following

equation:

t ¼Xni¼1

aitið Þ ð4Þ

The Sigworth–Sine transformation (Fig. 4D) allows a single plot clearly to display

dwell-times spanning several orders of magnitude. Individual exponential compo-

nents of the distribution can be directly identified from the peaks of the distribution.

After iterative exploration of alternative gating schemes, the log likelihood ratio

(Colquhoun, 1994) is used to identify the scheme that best fits the data. The chosen

scheme is then used to reidealize the raw data to provide the final gating para-

meters (mean life-times and Po). Although these are the methods we have used to

address the gating of IP3R (Rahman et al., 2009), more sophisticated approaches

exploit the additional information that lurks in the correlations that exist between

transitions (McManus et al., 1985).


Analyses like these identify the numbers of stable open and closed states and

plausible relationships between them. They lead thereby to models of the steps

through which the IP3R passes between its inactive and open states. Such analyses

have so far been rather limited for IP3R, but they clearly suggest the existence of a

single open state and several closed states (Ionescu et al., 2007; Rahman et al., 2009)

(Fig. 4D).

Extending the analysis to patches, in which we detected several IP3R, allowed us

to demonstrate that IP3 causes IP3R to form small clusters of �4–5 channels

within which to is reduced from �10 to �5 ms (Rahman et al., 2009). These

observations lead us to suggest that IP3 contribute to the evolution of elementary

Ca2þ signals by both regulating IP3R activity and by assembling IP3R into clusters,

within which regulation of IP3R by Ca2þ and IP3 is retuned (Rahman and Taylor,

2009; Rahman et al., 2009; Taylor et al., 2009a).

For most channels, including IP3R, single-channel open probability (Po) (rather

than g or the number of active channels) is the behavior that changes as the

stimulus intensity varies. Increasing IP3 or Ca2þ increases Po of IP3R because

both ligands shorten the duration of the closed times, without aVecting to; hence,the probability of finding the channel open (Po) is increased (Foskett et al., 2007;

Rahman et al., 2009). Po is calculated from the fitted amplitude histograms of the

current traces (typically lasting �1 min for IP3R) (Ding and Sachs, 1999):

Po ¼ Ao

Ao þ Ac

ð5Þ

where Ao and Ac are the areas under the curves corresponding to the open and

closed states in the current amplitude histogram, respectively.

When IP3R activity is low, it becomes very diYcult to know how many channels

are contributing because it is unlikely that all will open simultaneously. Under

these conditions, the overall activity is better expressed as NPo which is defined as

(Ching et al., 1999; Rahman et al., 2009):

NPo ¼PNn¼1

ntnð ÞT

ð6Þ

where tn is the total time for which n IP3R are simultaneously open and T is the

duration of the recording.

V. Concluding Remarks

Patch-clamp recording of IP3R expressed within the nuclear envelope allows

single-channel analyses of these otherwise inaccessible intracellular Ca2þ channels

(Figs. 1 and 2). DT40-KO cells provide a null background (Fig. 3) for expression of

recombinant and mutant IP3R allowing functional analysis of IP3R with defined


composition (Taylor et al., 2009b; Tovey et al., 2006). Nuclear patch-clamp

recording of DT40 cells heterologously expressing mammalian IP3R, therefore,

allows single-channel recording with its exquisite temporal resolution to be com-

bined with opportunities to manipulate systematically the structure of the

expressed IP3R. The stability of these patch-clamp recordings in a native mem-

brane and the opportunity to apply them in various configurations (Fig. 2) aVordvaluable opportunities to examine the behavior of small numbers of IP3R directly

(Rahman et al., 2009) and as a means to address the mechanisms underlying

IP3R activation (Rossi et al., 2009).

In the short period during which nuclear patch-clamp analyses have been

applied to IP3R, they have succeeded in confirming that IP3R are large conduc-

tance, relatively nonselective cation channels, and revealed the durations of the

channel openings and closing (Dellis et al., 2006; Foskett et al., 2007; Rahman

et al., 2009) (Figs. 4–6). Together, these insights allow estimates of the likely Ca2þ

fluxes through individual IP3R for comparison with optical measurements of the

elementary Ca2þ signals evoked by IP3 in situ (Shuai et al., 2007, 2008). Combining

site-directed mutagenesis with nuclear patch-clamp recording has provided direct

evidence that the pore of IP3R is formed by residues within the ‘‘P-loop’’ linking

the final pair of transmembrane domains of each IP3R subunit (Boehning et al.,

2001b; Dellis et al., 2006, 2008; Schug et al., 2008). The eVects of a novel family of

synthetic partial agonists on normal and mutant IP3R analyzed by nuclear patch-

clamp recording have shed light on the first stages of IP3R activation by showing

that the initial conformation changes evoked by IP3 binding to the IP3-binding

core pass onward toward the pore entirely via the N-terminal suppressor domain

(Rossi et al., 2009). Similar analyses have revealed the means, whereby ATP

(Betzenhauser et al., 2008b, 2009b), cyclic AMP-dependent protein kinase

(Betzenhauser et al., 2009a), cyclic AMP (Tovey et al., 2010), and various accesso-

ry proteins (Cheung et al., 2008, 2010; Li et al., 2007) modulate IP3R behavior.

Future application of the nuclear patch-clamp technique to IP3R is certain to add

further to our understanding of the stochastic behavior of single and clustered

IP3R and to resolving the structural basis of IP3R activation.

Acknowledgments

This work was supported by grants from the Wellcome Trust, and the Biotechnology and Biological

Sciences Research Council (UK). T. R. is a Drapers’ Company Research Fellow at Pembroke College,

Cambridge.

References

Adkins, C. E., and Taylor, C. W. (1999). Lateral inhibition of inositol 1, 4, 5-trisphosphate receptors by

cytosolic Ca2þ. Curr. Biol. 9, 1115–1118.Baba, T. W., Giroir, B. P., and Humphries, E. H. (1985). Cell lines derived from avian lymphomas

exhibit two distinct phenotypes. Virology 144, 139–151.


Barry, P. H. (1994). JPCalc, a software package for calculating liquid junction potential corrections in

patch-clamp, intracellular, epithelial and bilayer measurements and for correcting junction potential

measurements. J. Neurosci. Methods 51, 107–116.

Berridge, M. J., Lipp, P., and Bootman, M. D. (2000). The versatility and universality of calcium

signalling. Nat. Rev. Mol. Cell Biol. 1, 11–21.

Bers, D. M., Patton, C. W., and Nuccitelli, R. (1994). A practical guide to the preparation of Ca2þ

buVers. In ‘‘Methods in Cell Biology,’’ (R. Nuccitelli, ed.), Vol. 40, pp. 4–29. Academic Press,

San Diego, CA.

Betzenhauser, M. J., Wagner, L. E., II, Iwai, M., Michikawa, T., Mikoshiba, K., and Yule, D. I. (2008).

ATP modulation of Ca2þ release by type-2 and type-3 InsP3R: DiVering ATP sensitivities and

molecular determinants of action. J. Biol. Chem. 283, 21579–21587.

Betzenhauser, M. J., Wagner, L. E., II, Won, J. H., and Yule, D. I. (2008). Studying isoform-specific

inositol 1, 4, 5-trisphosphate receptor function and regulation. Methods 46, 177–182.

Betzenhauser, M. J., Fike, J. L., Wagner, L. E., II, and Yule, D. I. (2009). Protein kinase A increases

type-2 inositol 1, 4, 5-trisphosphate receptor activity by phosphorylation of serine 937. J. Biol. Chem.

284, 25116–25125.

Betzenhauser, M. J., Wagner, L. E., II, Park, H. S., and Yule, D. I. (2009). ATP regulation of type-1

inositol 1, 4, 5-trisphosphate receptor activity does not require walker A-type ATP-binding motifs.

J. Biol. Chem. 284, 16156–16163.

Bezprozvanny, I., and Ehrlich, B. E. (1994). Inositol (1, 4, 5)-trisphosphate (InsP3)-gated Ca channels

from cerebellum: Conduction properties for divalent cations and regulation by intraluminal calcium.

J. Gen. Physiol. 104, 821–856.

Bezprozvanny, I., Watras, J., and Ehrlich, B. E. (1991). Bell-shaped calcium-response curves for Ins(1,

4, 5)P3- and calcium-gated channels from endoplasmic reticulum of cerebellum.Nature 351, 751–754.

Boehning, D., and Joseph, S. K. (2000). Direct association of ligand-binding and pore domains in

homo- and heterotetrameric inositol 1, 4, 5-trisphosphate receptors. EMBO J. 19, 5450–5459.

Boehning, D., Joseph, S. K., Mak, D.-O. D., and Foskett, J. K. (2001). Single-channel recordings of

recombinant inositol trisphosphate receptors in mammalian nuclear envelope. Biophys. J. 81,

117–124.

Boehning, D., Mak, D.-O. D., Foskett, J. K., and Joseph, S. K. (2001). Molecular determinants of ion

permeation and selectivity in inositol 1,4,5-trisphosphate receptor Ca2þ channels. J. Biol. Chem. 276,

13509–13512.

Bootman, M. D., and Berridge, M. J. (1995). The elemental principles of calcium signaling. Cell 83,

675–678.

Bootman,M. D., Berridge,M. J., and Lipp, P. (1997). Cooking with calcium: The recipes for composing

global signals from elementary events. Cell 91, 367–373.

Bosanac, I., Alattia, J.-R., Mal, T. K., Chan, J., Talarico, S., Tong, F. K., Tong, K. I., Yoshikawa, F.,

Furuichi, T., Iwai, M., Michikawa, T., Mikoshiba, K., et al. (2002). Structure of the inositol 1, 4, 5-

trisphosphate receptor binding core in complex with its ligand. Nature 420, 696–700.

Bosanac, I., Yamazaki, H., Matsu-ura, T., Michikawa, M., Mikoshiba, K., and Ikura, M. (2005).

Crystal structure of the ligand binding suppressor domain of type 1 inositol 1, 4, 5-trisphosphate

receptor. Mol. Cell 17, 193–203.

Buerstedde, J. M., and Takeda, S. (1991). Increased ratio of targeted to random integration after

transfection of chicken B cell lines. Cell 67, 179–188.

Bustamante, J. O. (1994). Nuclear electrophysiology. J. Membr. Biol. 138, 105–112.

Bustamante, J. O., and Varanda, W. A. (1998). Patch-clamp detection of macromolecular translocation

along nuclear pores. Braz. J. Med. Biol. Res. 31, 333–354.

Cardy, T. J. A., Traynor, D., and Taylor, C. W. (1997). DiVerential regulation of types 1 and 3 inositol

trisphosphate receptors by cytosolic Ca2þ. Biochem. J. 328, 785–793.

Cheung, K. H., Shineman, D., Muller, M., Cardenas, C., Mei, L., Yang, J., Tomita, T., Iwatsubo, T.,

Lee, V. M., and Foskett, J. K. (2008). Mechanism of Ca2þ disruption in Alzheimer’s disease by

presenilin regulation of InsP3 receptor channel gating. Neuron 58, 871–883.


Cheung, K. H., Mei, L., Mak, D. O., Hayashi, I., Iwatsubo, T., Kang, D. E., and Foskett, J. K. (2010).

Gain-of-function enhancement of IP3 receptor modal gating by familial Alzheimer’s disease-linked

presenilin mutants in human cells and mouse neurons. Sci. Signal. 3, ra22.

Ching, L. L., Williams, A. J., and Sitsapesan, R. (1999). AMP is a partial agonist at the sheep cardiac

ryanodine receptor. Br. J. Pharmacol. 127, 161–171.

Choe, C., and Ehrlich, B. E. (2006). The inositol 1,4,5-trisphosphate receptor (IP3R) and its regulators:

Sometimes good and sometimes bad teamwork. Sci. STKE 363, re15.

Colquhoun, D. (2007). What have we learned from single ion channels? J. Physiol. 581, 425–427.

Colquhoun, D. (1994). Practical analysis of single channel records. In ‘‘Microelectrode Techniques: The

Plymouth Workshop Handbook’’ (D. Ogden, ed.), pp. 101–139. The Company of Biologist Ltd.,

Cambridge.

Dellis, O., Dedos, S., Tovey, S. C., Rahman, T.-U., Dubel, S. J., and Taylor, C. W. (2006). Ca2þ entry

through plasma membrane IP3 receptors. Science 313, 229–233.

Dellis, O., Rossi, A. M., Dedos, S. G., and Taylor, C. W. (2008). Counting functional IP3 receptors into

the plasma membrane. J. Biol. Chem. 283, 751–755.

Demuro, A., and Parker, I. (2007). Multi-dimensional resolution of elementary Ca2þ signals by

simultaneous multi-focal imaging. Cell Calcium 43, 367–374.

Ding, S., and Sachs, F. (1999). Single channel properties of P2X2 purinoceptors. J. Gen. Physiol. 113,

695–720.

Dingwall, C., and Laskey, R. (1992). The nuclear membrane. Science 258, 942–947.

Dong, X. P., Cheng, X., Mills, E., Delling, M., Wang, F., Kurz, T., and Xu, H. (2008). The type IV

mucolipidosis-associated protein TRPML1 is an endolysosomal iron release channel. Nature 455,

992–996.

Ehrlich, B. E., and Watras, J. (1988). Inositol 1, 4, 5-trisphosphate activates a channel from smooth

muscle sarcoplasmic reticulum. Nature 336, 583–586.

Fatt, P., and Ginsborg, B. L. (1958). The ionic requirements for the production of action potentials in

crustacean muscle fibres. J. Physiol. 142, 516–543.

Foskett, J. K., White, C., Cheung, K. H., and Mak, D. O. (2007). Inositol trisphosphate receptor Ca2þ

release channels. Physiol. Rev. 87, 593–658.

Franco-Obregon, A., Wang, H. W., and Clapham, D. E. (2000). Distinct ion channel classes are

expressed on the outer nuclear envelope of T- and B-lymphocyte cell lines. Biophys. J. 79, 202–214.

Futatsugi, A., Nakamura, T., Yamada, M. K., Ebisui, E., Nakamura, K., Uchida, K., Kitaguchi, T.,

Takahashi-Iwanaga, H., Noda, T., Aruga, J., and Mikoshiba, K. (2005). IP3 receptor types 2 and 3

mediate exocrine secretion underlying energy metabolism. Science 309, 2232–2234.

Hamill, O. P., Marty, A., Neher, E., Sakmann, B., and Sigworth, F. J. (1981). Improved patch-clamp

techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers

Arch. 391, 663–678.

Hering, S., Beech, D. J., and Bolton, T. B. (1987). A simple method of fast extracellular solution

exchange for the study of whole-cell or single channel currents using patch-clamp. technique. Pflugers

Arch. 410, 335–337.

Horn, R. (1991). Estimating the number of channels in patch recordings. Biophys. J. 60, 433–439.

Humbert, J.-P., Matter, N., Artault, J.-C., Koppler, P., and Malviya, A. N. (1996). Inositol 1, 4, 5-

trisphosphate receptor is located to the inner nuclear membrane vindicating regulation of nuclear

calcium signaling by inostiol 1, 4, 5-trisphosphate. J. Biol. Chem. 271, 478–485.

Ionescu, L., Cheung, K. H., Vais, H., Mak, D. O., White, C., and Foskett, J. K. (2006). Graded

recruitment and inactivation of single InsP3 receptor Ca2þ-release channels: implications for quantal

Ca2þ release. J. Physiol. 573, 645–662.

Ionescu, L., White, C., Cheung, K. H., Shuai, J., Parker, I., Pearson, J. E., Foskett, J. K., and

Mak, D. O. (2007). Mode switching is the major mechanism of ligand regulation of InsP3 receptor

calcium release channels. J. Gen. Physiol. 130, 631–635.

Jonas, E. A., Knox, R. J., and Kaczmarek, L. (1997). Giga-ohm seals on intracellular membranes:

A technique for studying intracellular ion channels in intact cells. Neuron 19, 7–13.


Joseph, S. K., and Hajnoczky, G. (2007). IP3 receptors in cell survival and apoptosis: Ca2þ release and

beyond. Apoptosis 12, 951–968.

Joseph, S. K., Lin, C., Pierson, S., Thomas, A. P., and Maranto, A. R. (1995). Heteroligomers of type-I

and type-III inositol trisphosphate receptors in WB rat liver epithelial cells. J. Biol. Chem. 270,

23310–23315.

Joseph, S. K., Bokkala, S., Boehning, D., and Zeigler, S. (2000). Factors determining the composition of

inositol trisphosphate receptor hetero-oligomers expressed in COS cells. J. Biol. Chem. 275,

16084–16090.

Kirichok, Y., Krapavinsky, G., and Clapham, D. E. (2004). The mitochondrial calcium uniporter is a

highly selective ion channel. Nature 427, 360–364.

Kusnier, C., Cardenas, C., Hidalgo, J., and Jaimovich, E. (2006). Single-channel recording of inositol

trisphosphate receptor in the isolated nucleus of a muscle cell line. Biol. Res. 39, 541–553.

Laude, A. J., Tovey, S. C., Dedos, S., Potter, B. V. L., Lummis, S. C. R., and Taylor, C. W. (2005).

Rapid functional assays of recombinant IP3 receptors. Cell Calcium 38, 45–51.

Li, C., Wang, X., Vais, H., Thompson, C. B., Foskett, J. K., andWhite, C. (2007). Apoptosis regulation

by Bcl-xL modulation of mammalian inositol 1, 4, 5-trisphosphate receptor channel isoform gating.

Proc. Natl. Acad. Sci. USA 104, 12565–12570.

Maathuis, J. M., Taylor, A. R., Assmann, S. M., and Sanders, D. (1997). Seal-promoting solutions and

pipette perfusion for patch clamping plant cells. Plant J. 11, 891–896.

Mackrill, J. J., Challiss, R. A. J., O’Connell, D. A., Lai, F. A., and Nahorski, S. R. (1997). DiVerential

expression and regulation of ryanodine receptor and inositol 1, 4, 5-trisphosphate receptor Ca2þ

release channels in mammalian tissues and cell lines. Biochem. J. 327, 251–258.

Maeda, N., Kawasaki, T., Nakade, S., Yokota, N., Taguchi, T., Kasai, M., and Mikoshiba, K. (1991).

Structural and functional characterization of inositol 1, 4, 5-trisphosphate receptor channel from

mouse cerebellum. J. Biol. Chem. 266, 1109–1116.

Mak, D.-O. D., and Foskett, J. K. (1994). Single channel inositol 1,4,5-trisphosphate receptor currents

revealed by patch clamp of isolated Xenopus oocyte nuclei. J. Biol. Chem. 269, 29375–29378.

Mak, D.-O. D., and Foskett, J. K. (1997). Single-channel kinetics, inactivation, and spatial distribution

of inositol trisphosphate (IP3) receptors in Xenopus oocyte nucleus. J. Gen. Physiol. 109, 571–587.

Mak, D. O., and Foskett, J. K. (1998). EVects of divalent cations on single-channel conduction

properties of Xenopus IP3 receptor. Am. J. Physiol. 275, C179–C188.

Mak, D.-O. D., White, C., Ionescu, L., and Foskett, J. K. (2006). Nuclear patch clamp electrophysiolo-

gy of inositol trisphosphate receptor Ca2þ release channels. In ‘‘Calcium Signaling,’’ (J. W. Putney,

Jr., ed.), pp. 203–229. Taylor and Francis, Boca Raton.

Mak, D. O., Pearson, J. E., Loong, K. P., Datta, S., Fernandez-Mongil, M., and Foskett, J. K. (2007).

Rapid ligand-regulated gating kinetics of single inositol 1, 4, 5-trisphosphate receptor Ca2þ release

channels. EMBO Rep. 8, 1044–1051.

Marchant, J. S., and Parker, I. (2001). Role of elementary Ca2þ puVs in generating repetitive Ca2þ

oscillations. EMBO J. 20, 65–76.

Marchant, J. S., and Taylor, C.W. (1997). Cooperative activation of IP3 receptors by sequential binding

of IP3 and Ca2þ safeguards against spontaneous activity. Curr. Biol. 7, 510–518.

Marchant, J., Callamaras, N., and Parker, I. (1999). Initiation of IP3-mediated Ca2þ waves in Xenopus

oocytes. EMBO J. 18, 5285–5299.

Marchenko, S. M., Yarotskyy, V. V., Kovalenko, T. N., Kostyuk, P. G., and Thomas, R. C. (2005).

Spontaneously active and InsP3-activated ion channels in cell nuclei from rat cerebellar Purkinje and

granule neurones. J. Physiol. 565, 897–910.

Matzke, A. J., Weiger, T. M., andMatzke, M. A. (1990). Detection of a large cation-selective channel in

nuclear envelopes of avian erythrocytes. FEBS Lett. 271, 161–164.

Mayrleitner, M., Chadwick, C. C., Timerman, A. P., Fleischer, S., and Schinder, H. (1991). Purified IP3

receptor from smooth muscle forms an IP3-gated and heparin-sensitive Ca2þ channel in planar

bilayers. Cell Calcium 12, 505–514.


Mazzanti, M., DeFelice, L. J., Cohn, J., and Malter, H. (1990). Ion channels in the nuclear envelope.

Nature 343, 764–767.

Mazzanti, M., Bustamante, J. O., and Oberleithner, H. (2001). Electrical dimension of the nuclear

envelope. Physiol. Rev. 81, 1–19.

McManus, O. B., Blatz, A. L., and Magleby, K. L. (1985). Inverse relationship of the durations of

adjacent open and shut intervals for C1 and K channels. Nature 317, 625–627.

Miyakawa, T., Maeda, A., Yamazawa, T., Hirose, K., Kurosaki, T., and Iino, M. (1999). Encoding of

Ca2þ signals by diVerential expression of IP3 receptor subtypes. EMBO J. 18, 1303–1308.

Neher, E., and Sakmann, B. (1976). Single-channel currents recorded from membrane of denervated

frog muscle fibres. Nature 260, 799–802.

Ogden, D. (1994). Microelectrode techniques. ‘‘The Plymouth Workshop Handbook.’’ Company of

Biologists, Cambridge.

Patterson, R. L., Boehning, D., and Snyder, S. H. (2004). Inositol 1, 4, 5-trisphosphate receptors as

signal integrators. Annu. Rev. Biochem. 73, 437–465.

Patton, C., Thompson, S., and Epel, D. (2004). Some precautions in using chelators to buVer metals in

biological solutions. Cell Calcium 35, 427–431.

Penner, R. (1995). A practical guide to patch clamping. In ‘‘Single-Channel Recording,’’ (B. Sakmann,

and E. Neher, eds.), pp. 3–30. Plenum Press, New York.

Qin, F. (2004). Restoration of single-channel currents using the segmental k-means method based on

hidden Markov modeling. Biophys. J. 86, 1488–1501.

Qin, F., Auerbach, A., and Sachs, F. (1996). Estimating single-channel kinetic parameters from

idealized patch-clamp data containing missed events. Biophys. J. 70, 264–280.

Qin, F., Auerbach, A., and Sachs, F. (1997). Maximum likelihood estimation of aggregated Markov

processes. Proc. R. Soc. Lond. B 264, 375–383.

Qin, F., Auerbach, A., and Sachs, F. (2000). Hidden Markov modeling for single channel kinetics with

filtering and correlated noise. Biophys. J. 79, 1928–1944.

Rae, J. L., and Levis, R. A. (1992). Glass technology for patch clamp electrodes.Methods Enzymol. 207,

66–92.

Rahman, T., and Taylor, C. W. (2009). Dynamic regulation of IP3 receptor clustering and activity by

IP3. Channels 3, 226–232.

Rahman, T. U., Skupin, A., Falcke, M., and Taylor, C. W. (2009). Clustering of IP3 receptors by IP3

retunes their regulation by IP3 and Ca2þ. Nature 458, 655–659.

Rossi, A. M., Riley, A. M., Tovey, S. C., Rahman, T., Dellis, O., Taylor, E. J. A., Veresov, V. G.,

Potter, B. V. L., and Taylor, C. W. (2009). Synthetic partial agonists reveal key steps in IP3 receptor

activation. Nat. Chem. Biol. 5, 631–639.

Sachs, F., Neil, J., and Barkakati, N. (1982). The automated analysis of data from single ionic channels.


Sakmann, B., and Neher, E. (1995). ‘‘Single-Channel Recording.’’ Plenum Press, New York.

Schug, Z. T., da Fonseca, P. C., Bhanumathy, C. D., Wagner, L., II, Zhang, X., Bailey, B.,

Morris, E. P., Yule, D. I., and Joseph, S. K. (2008). Molecular characterization of the inositol 1, 4,

5-trisphosphate receptor pore-forming segment. J. Biol. Chem. 283, 2939–2948.

Shuai, J., Pearson, J., Foskett, J. K., Mak, D. O., and Parker, I. (2007). A kinetic model of single and

clustered IP3 receptors in the absence of Ca2þ feedback. Biophys. J. 93, 1151–1162.

Shuai, J., Pearson, J. E., and Parker, I. (2008). Modeling Ca2þ feedback on a single inositol 1, 4, 5-

trisphosphate receptor and its modulation by Ca2þ buVers. Biophys. J. 95, 3738–3752.

Sigworth, F. J., and Sine, S. M. (1987). Data transformations for improved display and fitting of single-

channel dwell time histograms. Biophys. J. 52, 1047–1054.

Sivilotti, L. G. (2010). What single-channel analysis tells us of the activation mechanism of ligand-gated

channels: The case of the glycine receptor. J. Physiol. 588, 45–58.

Smith, I. F., and Parker, I. (2009). Imaging the quantal substructure of single IP3R channel activity

during Ca2þ puVs in intact mammalian cells. Proc. Natl. Acad. Sci. USA 106, 6404–6409.


Somlyo, A. V., Shuman, H., and Somlyo, A. P. (1977). Composition of sarcoplasmic reticulum in situ by

electron probe X-ray microanalysis. Nature 268, 556–558.

Stehno-Bittel, L., LuckhoV, A., and Clapham, D. E. (1995). Calcium release from the nucleus by InsP3

receptor channels. Neuron 14, 163–167.

Sugawara, H., Kurosaki, M., Takata, M., and Kurosaki, T. (1997). Genetic evidence for involvement of

type 1, type 2 and type 3 inositol 1, 4, 5-trisphosphate receptors in signal transduction through the

B-cell antigen receptor. EMBO J. 16, 3078–3088.

Tabares, L., Mazzanti, M., and Clapham, D. E. (1991). Chloride channels in the nuclear membrane.

J. Membr. Biol. 123, 49–54.

Taylor, C. W. (1992). Kinetics of inositol 1, 4, 5-trisphosphate-stimulated Ca2þ mobilization. Adv.

Second Mess. Phos. Res. 26, 109–142.

Taylor, C. W., Genazzani, A. A., and Morris, S. A. (1999). Expression of inositol trisphosphate

receptors. Cell Calcium 26, 237–251.

Taylor, C. W., da Fonseca, P. C. A., and Morris, E. P. (2004). IP3 receptors: The search for structure.

Trends Biochem. Sci. 29, 210–219.

Taylor, C. W., Rahman, T., and Pantazaka, E. (2009). Targeting and clustering of IP3 receptors: Key

determinants of spatially organized Ca2þ signals. Chaos 19037102-1-037102-10.

Taylor, C. W., Rahman, T., Tovey, S. C., Dedos, S. G., Taylor, E. J. A., and Velamakanni, S. (2009).

IP3 receptors: Some lessons from DT40 cells. Immunol. Rev. 231, 23–44.

Tovey, S. C., Sun, Y., and Taylor, C. W. (2006). Rapid functional assays of intracellular Ca2þ channels.

Nat. Protoc. 1, 259–263.

Tovey, S. C., Dedos, S. G., Rahman, T., Taylor, E. J. A., Pantazaka, E., and Taylor, C. W. (2010).

Regulation of inositol 1, 4, 5-trisphosphate receptors by cAMP independent of cAMP-dependent

protein kinase. J. Biol. Chem. 285, 12979–12989.

Walker, J. W., Somlyo, A. V., Goldman, Y. E., Somlyo, A. P., and Trentham, D. R. (1987). Kinetics of

smooth and skeletal muscle activation by laser pulse photolysis of caged inositol 1, 4, 5-trisphosphate.

Nature 327, 249–252.

Wang, Y., Chen, J., Wang, Y., Taylor, C. W., Hirata, Y., Hagiwara, H., Mikoshiba, K., Toyo-oka, T.,

Omata, M., and Sakaki, Y. (2001). Crucial role of type 1, but not type 3, inositol 1, 4, 5-trisphosphate

(IP3) receptors in IP3-induced Ca2þ release, capacitative Ca2þ entry, and proliferation of A7r5

vascular smooth muscle cells. Circ. Res. 88, 202–209.

Weiss, D. S., and Magleby, K. L. (1989). Gating scheme for single GABA-activated Cl� channels

determined from stability plots, dwell-time distributions, and adjacent-interval durations. J. Neu-

rosci. 9, 1314–1324.

Williams, A. J. (1995). The measurement of the function of ion channels reconstituted into artificial

membranes. In ‘‘Ion Channels. A Practical Approach,’’ (R. H. Ashley, ed.), pp. 43–67. IRL Press,

Oxford.

Williams, A. J. (2002). Ion conduction and selectivity in the ryanodine receptor channel. Front. Biosci. 7,

1–8.

Winding, P., and Berchtold, M. W. (2001). The chicken B cell line DT40: A novel tool for gene

disruption experiments. J. Immunol. Methods 249, 1–16.

Wojcikiewicz, R. J. H., and He, Y. (1995). Type I, II and III inositol 1, 4, 5-trisphosphate receptor co-

immunoprecipitation as evidence for the existence heterotetrameric receptor complexes. Biochem.

Biophys. Res. Commun. 213, 334–341.

Wojcikiewicz, R. J. H., and Luo, S. G. (1998). DiVerences among type I, II, and III inositol-1, 4, 5-

trisphosphate receptors in ligand-binding aYnity influence the sensitivity of calcium stores to inositol-

1, 4, 5-trisphosphate. Mol. Pharmacol. 53, 656–662.

Xu, H., Delling, M., Li, L., Dong, X., and Clapham, D. E. (2007). Activating mutation in a mucolipin

transient receptor potential channel leads to melanocyte loss in varitint-waddler mice. Proc. Natl.

Acad. Sci. USA 104, 18321–18326.

CHAPTER 9


Confocal and Multiphoton Imaging ofIntracellular Ca2þ

Godfrey Smith,* Martyn Reynolds,† Francis Burton,* andOle Johan Kemi**School of Life SciencesUniversity of GlasgowUnited Kingdom

†Cairn Research LimitedFaversham, KentUnited Kingdom

A
OGY, VOll rights r
A

Les

bstract

. 99 0091-67erved. 225 DOI: 10.1016/S0091-67

I.
W hy Study Ca2þ Signaling with Confocal and Multiphoton Microscopy II. C onfocal Microscopy III. L imitations in Speed of Confocal Imaging IV. L aser Scanning Confocal Microscopy V. T otal Internal Reflection Fluorescence Microscopy VI. F orster Resonance Energy Transfer Microscopy VII. P arallel Scanning Confocal Systems VIII. S pinning Disk Confocal Microscopy IX. P rogrammable Matrix Microscopy X. A dvantages and Disadvantages of Confocal Microscopy XI. M ultiphoton Excitation Laser Scanning Microscopy XII. C a2þ Indicators for Use in Confocal and Multiphoton Microscopy XIII. U se of Dyes for Single-Photon Confocal Microscopy XIV. U se of Dyes for 2P Excitation Microscopy XV. Is It Worth Converting the Intracellular Fluorescence Signal to [Ca2þ]? XVI. C alibration of Single Wavelength Dyes XVII. E stimation of Fmax Values X VIII. E stimation of Fmin or the Dynamic Range of the Dye XIX. C onsequence of Errors in Estimation of Intrinsic and Dye Fluorescence XX. M ultimodal and Multiple Fluorophore Confocal and Multiphoton Microscopy
R
eferences
9X/10 $35.009X(10)99009-2

226 Godfrey Smith et al.

Abstract

This chapter compares the imaging capabilities of a range of systems including

multiphoton microscopy in regard to measurements of intracellular Ca2þ within

living cells. In particular, the excitation spectra of popular fluorescent Ca2þ

indicators are shown during 1P and 2P excitation. The strengths and limitations

of the current indicators are discussed along with error analysis which highlights

the value of matching the Ca2þ aYnity of the dye to a particular aspect of Ca2þ

signaling. Finally, the combined emission spectra of Ca2þ and voltage sensitive

dyes are compared to allow the choice of the optimum combination to allow

simultaneous intracellular Ca2þ and membrane voltage measurement.

I. Why Study Ca2þ Signaling with Confocal and MultiphotonMicroscopy

Ca2þ is a ubiquitous intracellular messenger that controls a large number of

cellular processes, such as gene transcription, excitation, contraction, apoptosis,

cellular respiration, and the activity levels of many cell-signaling messenger cas-

cades. Inside the cell, Ca2þmay, under various conditions, sequester into the sarco/

endoplasmic reticulum, mitochondria, and the nucleus, or exist in the cytosol

either in its free form or as bound to buVers. Typically, a large Ca2þ concentration

gradient is maintained across the plasmamembrane of the cell. Because of diVerentCa2þ channels, pumps, and exchangers on the membranes of the cell or organelles,

Ca2þ fluxes may be created at multiple locations in the cell. Therefore, Ca2þ

concentration and signal may be specific with respect to both location and time.

Moreover, Ca2þ may concentrate in distinct cytoplasmic regions because of tight

physical loci not enclosed by membranes, for example, the dyadic area between the

transverse tubule and the sarco/endoplasmic reticulum of muscle cells. Given the

large number of Ca2þ channels feeding it with Ca2þ from both the extracellular

space (transverse tubule) and the sarco/endoplasmic reticulum, such that the dyad

may transiently have very diVerent localized Ca2þ concentrations compared to the

rest of the cytosol which may be only nanometers away. Thus, Ca2þ localizes in the

cytosol as well as within organellar compartments, and these Ca2þ signals may last

for very short timeframes (ns) or for substantially longer periods of time (min).

Ca2þ signaling per se is not within the remit of this chapter and will not be covered

in any detail, but the interested reader is referred to other sources, for example,

Bootman et al. (2001).

Nonetheless, for the purpose of this chapter, it is important to acknowledge that

the average Ca2þ concentration in any given cell usually ranges 0.01–1 mM, but

that the Ca2þ almost never exists uniformly across the cell, and that local Ca2þ

events may occur with very fast time courses. This therefore requires Ca2þ imaging

of live specimens with high spatial and temporal resolution. Thus, one would

9. Confocal and Multiphoton Imaging of Intracellular Ca2þ 227

ideally want to distinguish Ca2þ in distinct areas that may be within a nanometer

distance from each other, and to record localized Ca2þ events that may last only a

millisecond. Although such requirements tax any given microscopy system, confo-

cal and multiphoton microscopy systems oVer a range of imaging capabilities that

fulfill these criteria.

II. Confocal Microscopy

Optical sectioning by confocal microscopy adds several benefits to Ca2þ imaging.

Since its early development, confocal microscopy has fundamentally transformed

optical imaging to now provide a valuable addition that allows unprecedented

imaging ofminute optical sections within live specimens in close to real-time speeds.

In terms of Ca2þ imaging, this has opened up new fields of study, given that Ca2þ

signaling in many, if not all, biological systems is compartmentalized within small

sections of the cell and occurs often at very high velocities. Confocal microscopy

allows the study of these events within discrete depths of cell or tissue by blocking

light originating outside the plane of focus. This is achieved by the addition of

confocal apertures in front of the illumination source and in the image plane

directly in front of the signal detection system (see Fig. 1); usually, a photomulti-

plier tube (PMT) that rejects out-of-focus light originating from fluorescence

outwith the area of interest (the focal plane), and only allowing in-focus light

through to the PMT (Webb, 1999) (Fig. 1). This is in contrast to regular epifluor-

escence microscopy, in which the majority of the fluorescence is out-of-focus light

that generally reduces the contrast of the in-focus light, and also dramatically

compromises in-focus detail. This occurs since the emitted fluorescence cannot

be discriminated along the Z-axis (top to bottom), and also less along the X- and

Y-axes (although this has also to do with excitation light sources; see later) in

conventional epifluorescence microscopy (Lichtman and Conchello, 2005)

(See also later). Thus, although confocal microscopy also excites the specimen

along the entire Z-axis in line with conventional epifluorescence microscopy, only

in-focus light is allowed to pass the pinhole of the confocal aperture to enter the

signal detector.

Importantly, confocal imaging may be performed on live specimens residing

under physiologic conditions and that are electrically, chemically, mechanically,

and otherwise active and healthy. Specimens may also be electrically and mechani-

cally stimulated and superfused by any given solutions that would not interfere

with the confocal imaging.

The diVerence between regular epifluorescence and confocal microscopy light

capture abilities can be illustrated by the following examples. Considering that the

depth of focus of a high numerical aperture (NA>1.3) objective is restricted to

�0.3 mm,whereas the depth of a fluorescent cell may be�5–25 mm, it becomes clear

that the depth of focus will only constitute�1–5% of the full depth of the cell. Since

epifluorescence microscopy captures light along the entire Z-axis, 95–99% of the

Lightdetector

Dichroicbeamsplitter

Focal plane

Objective

Confocal aperturewith pinhole

Laser

Fig. 1 Overview of the optical pathway of a confocal microscope. The blue path illustrates the

excitation light, whereas the green path illustrates the emitted fluorescence light. Note that the confocal

aperture with the pinhole in front of the light detector (usually, a photomultiplier tube (PMT)) blocks

out-of-focus light.


cell volume will contribute to unwanted out-of-focus background signal, or noise

(Lichtman and Conchello, 2005), and this cannot be distinguished from in-focus

fluorescence. Thus,most of the cell will be out of focus, andwith it, the vastmajority

of the signal will come from out-of-focus areas. In contrast, setting the pinhole of

the confocal aperture to 1 airy unit to achieve true confocality will provide an

optical section or Z-resolution of 0.5–1 mm with the same high NA objective as

described above. This will only allow a minimum of out-of-focus light to reach the

signal detector, without any other interference to the optical pathway or any

secondary digital processing of the signal at the time of recording apart from the

scanning and building of the image itself, which otherwise would have further

compromised the scanning speed.

The full 3D XYZ-resolution, or the ability to discern two points from each

other, will, however, be diVraction-limited as determined by the point spread

function (PSF) set by the optical performance of the microscope. With high NA

objectives (>1.2), this is typically �0.3�0.3�0.6 mm (Cox and Sheppard, 2004).

Although Z-resolution is dramatically diVerent between conventional epifluores-

cence and confocal microscopes, the 2D spatial resolution in the XY-field is

not, though factors, such as excitation wavelength, objectives and the optical


pathway, and diVerent media and surfaces, will also aVect this. However, several

factors associated with the confocal principle allow for improving also XY-

resolution, as compared to widefield epifluorescence microscopy. First, spatial

resolution may be improved by further reducing the pinhole diameter in the

confocal aperture to a size smaller than the width of the central disk of the airy

unit pattern, though this also dramatically reduces light transmission. This

principle works because the pinhole is aligned with the center of the airy unit

pattern of the illuminating beam, which means that any emission originating

from any fluorescent molecules excited by the outer airy rings of the illuminating

beam will be blocked by the confocal aperture; like all light, the illumination

beam also presents with a airy wave pattern consisting of a central bright spot

and outer ring waves that in comparison are more faint. In other words, the

resultant fluorescence emission may be experimentally manipulated to originate

from an area smaller than the airy unit, which cannot be achieved by conven-

tional widefield epifluorescence microscopy. Moreover, because the PSF of the

confocal microscope is narrower at normalized light intensities relative to that of

the conventional widefield microscope, it means the XY spatial resolution will be

�1.4�greater with a confocal microscope than a widefield microscope

(Conchello and Lichtman, 2005). Finally, some of the in-focus light will scatter

on its way through the specimen, due to diVraction, reflection, and refraction as

cell structures interfere with the light path. This also compromises fluorescence,

but not confocal microscopy, as the confocal aperture also blocks scattered light

from reaching the signal detector.

In addition, combining confocal microscopy with total internal reflection fluo-

rescence (TIRF) or Forster resonance energy transfer (FRET) microscopy techni-

ques has the capacity to increase resolution to only a few tens of nanometers

(see below for more detailed information). Other techniques such as narrowing

the boundaries of the PSF by suppressing (de-exciting) the fluorescence from the

edge of the center spot of the airy pattern by stimulated emission depletion

(STED), and other nonlinear optical masking techniques, have further enhanced

optical resolution of confocal microscopes (Bullen, 2008; Willig et al., 2006),

though these techniques are not yet compatible with fast scanning of Ca2þ events

that take place over fast timescales, and will therefore not be discussed here.

Finally, secondary signal processing or deconvolution (computationally reverse

optical distortion to enhance resolution) of the recorded images also serve to

enhance spatial resolution of both confocal and epifluorescence microscopy by a

factor of 2–3.

Because light scattering increases proportionally to increasing thickness of the

specimen, this becomes more of an issue with deeper imaging of thicker speci-

mens. Therefore, appropriately setting the pinhole not only allows for imaging of

thin optical sections, but also aVects the signal-to-noise (SNR) ratio. Whereas a

case made be made that reducing the pinhole diameter increases XYZ-resolution

(especially Z-, but also XY-resolution; see above), opening the pinhole to ap-

proximately match the projected image of the diVraction-limited spot (1.22l/NA,


where l is the illumination wavelength) will substantially increase the SNR with

only minimal reduction in the Z-resolution (Conchello et al., 1994). Thus, this

would increase the quality of the signal with little degradation of the depth

discrimination.

III. Limitations in Speed of Confocal Imaging

The removal of out-of-focus light allows for relatively fast imaging of �0.6 mmthin sections either in spot, line, or frame modes that either may be repeated

sequentially, or combined with stepwise up- or down-focusing through the speci-

men in order to generate 3D reconstructions. 3D sectioning does not allow for

recording of cellular events in real time, but repeated 1D line imaging or even

repeated 2D imaging with reduced frame sizes or restricted pixel numbers in

contrast do allow for relatively fast recording with a temporal resolution of

approaching a microsecond scale. Although not as fast as regular widefield imag-

ing, even 2D frame imaging may still be acquired on fast time scales, usually within

hundreds of millisecond, though there will be a trade-oV between temporal and

spatial resolution. The reason for the lower temporal resolution compared to

widefield imaging is that conventional confocal imaging requires some form of

scanning, that is, sequential pixel sampling, in order to ‘‘build’’ an image, which

thus happens pixel-by-pixel. In contrast, the whole field during widefield imaging is

captured simultaneously either by PMTs or charge-coupled device (CCD) cameras

(Ogden, 1994). For some Ca2þ events, imaging with a temporal resolution in the

order of milliseconds may be satisfactory, but other events may occur considerably

faster than this. Likewise, some events allow the microscopist to sample images

with a low spatial resolution, whereas others require the opposite. Thus, the speed

of confocal image acquisition depends on the mode and the settings of the scanning

and how many pixels are scanned before returning to the same pixel again.

This will be discussed later.

IV. Laser Scanning Confocal Microscopy

Out-of-focus light rejection and image acquisition through a confocal aperture

with a pinhole is the common principle that constitutes confocal microscopes, but

the illumination and excitation principles may diVer between various systems.

First, confocal microscopy by laser scanning the specimen (laser scanning confocal

microscopy, often abbreviated to LSM or LSCM) is the most widely used illumi-

nation and excitation method today.

During LSCM, a laser beam is directed on to the specimen, whereupon it scans

the designated field, which may be a single spot (in reality rarely used for biological

imaging apart from fluorescence recovery after photobleaching (FRAP) applica-

tions), a 1D line, or a 2D frame. The laser is controlled by the use of two oscillating


mirrors in the scanhead that deflect the beam along a fast and slow axes perpen-

dicular to one another. Thus, during a 2D frame scan, the beam is first directed

along the horizontal axis, after which it ‘‘jumps’’ down one pixel and scans the next

line, and this process continues until a full frame has been scanned. This may then

be repeated for serial frame scanning, or the plane of focus may be moved along

the Z-axis for 3D imaging. The fluorescence emission returns along the same light

path (descanning), but is deflected by a dichroic mirror splitting the excitation and

emission lights, such that the emission light only is directed onto the confocal

aperture, where the in-focus light penetrates though the pinhole to reach the signal

detector.

Unfortunately, scanning is a rate-limiting step for gaining fast images, especially

in a 2D frame scan mode, because of the mechanical characteristics of the mirrors.

A typical 512�512 pixel 2D frame may be scanned in �1 s, depending on the

settings under which the scan is performed. When scanspeed is of concern, several

approaches may be taken to increase this, for example, by linescanning instead of

framescanning. In this configuration, the same line is scanned sequentially for a

given period of time to allow detection of one or more events occurring within the

time frame of the scan. Although this provides a high temporal resolution, the

acquired information is limited to a one-pixel wide area of the cell, such that

information about events occurring elsewhere in the cell is missed. Furthermore,

reducing the pixel dwell time (the period over which each pixel is scanned) also

increases scanspeed, but this also reduces the SNR. Reduced SNRmay partially be

compensated for by increasing the laser power, but this may present other problems

such as photodamage to the specimen; especially in live cells, and photobleaching.

Finally, the length of the line or the size of the frame to be scanned may also be

reduced, or fewer pixelsmay be scanned during 2D frame scanning to yield the same

eVect, and the scan may also be run in a bidirectional mode instead of unidirection-

al, though using two lines running in opposite directions may cause a slight oVsetbetween them, which tends to blur the signal. More recently, several approaches

have been taken to increase scanspeeds, in particular for 2D frame imaging, such as

utilizing resonant oscillating mirrors in the scanhead instead of the more conven-

tional galvanometer-driven mechanical mirrors. Other options include arranging

prisms and acousto-optical deflectors into the excitation light path to illuminate the

entire line simultaneously, instead of a pixel-by-pixel illumination applied by the

conventional laser scanning microscopes described above. This means that the

scanning in a 2D frame mode would only involve movement along one dimension

(X), since the other dimension (Y) would all be scanned at once (simultaneously),

and therefore, 2D frame scanning may be performed at linescan speeds or at speeds

approaching video rates, if the emitted fluorescence is deflected onto a linear CCD

camera, though PMTarraysmay also be used. Single PMTswould, however, not be

able to construct the images if lines instead of single pixels are scanned. The caveat

with these approaches is that true confocality will be lost because the pinhole of the

confocal aperture must be replaced by one or more longitudinal slit openings to

accommodate the simultaneous scanning of lines (hence, this is also called slit


scanning). Also, series of holographic or curved mirrors in the scanhead have also

been utilized to scan more than one pixel at a time, but this has remained more of a

rarity compared to the abovementionedmicroscopymodifications (Callamaras and

Parker, 1999; Tsien and Bacskai, 1995).

The aforementioned confocal scanning approaches depend on the use of single-

photon lasers as the source of illumination and excitation. These are based on the

principle that a single photon provides enough energy to excite a single fluorescent

molecule, that is, to ‘‘lift’’ it from a ground state to the ‘‘excited’’ state. The phase

where the fluorophore is lifted to the excited state lasts for femtoseconds (10�15 s),

whereas the fluorophore remains in the higher-energy excited state for picoseconds

(10�12 s) where it undergoes internal conversion and starts to vibrate, which eVec-tively leads to dissipation of energy, such that it drops back to the ground state;

measurable to a time scale of nanoseconds (10�9 s). When this happens, the

fluorophore releases a photon that due of the loss of energy has a longer wavelength

(less energy), and this is what creates the fluorescence emission that may be

measured by signal detectors such as PMTs or CCD cameras. The diVerencebetween the excitation and emission spectra (emission wavelengths being longer

than excitation wavelengths) is called the Stokes shift. The process of excitation and

subsequent relaxation with photon release and fluorescence emission can be illu-

strated by a Jablonski diagram (Fig. 2), and is not restricted to confocal microsco-

py, but is in fact the basis for all fluorescence techniques including epifluorescence

microscopy and spectroscopy. As detailed above, it is the volume of the recorded

fluorescent emission that diVers between confocal and epifluorescence microscopy

modalities, though the volume of excitation may also diVer, but this has to do with

how much of the specimen is subjected to the illumination light. However,

although the laser excites fluorophores along the entire Z-axis of the specimen

(see also above and Fig. 3), peak excitation and as such peak brightness occurs at

the focal plane, whereas out-of-focus excitation decreases with the square of the

distance from the focal plane. This is because the laser excitation beam presents

with an hourglass shape, with the ‘‘waist’’ of the hourglass coinciding exactly with

the focal plane.

Several laser lines have been developed that allow single-photon excitation

of fluorescent Ca2þ indicators (fluorophores), in particular, the multiline

argon ion (Ar-ion) laser that provides high-intensity light from the ultraviolet

(UV) to the green spectrum (�250–514 nm wavelengths), the single-line helium–

neon (He–Ne) lasers that extend the covered spectrum to �633 nm, and argon–

krypton (Ar–Kr) lasers that providehigh-intensity light fromblue to redwavelengths.

Thus, these lasers are well suited for exciting the common Ca2þ indicators dyes and

are also as such much used in Ca2þ signaling research. Recent developments in solid

state and diode lasers have also added more choices for the microscopist. However,

lasers will not be covered in detail here, but the interested reader will find a wealth of

literature on this topic by searching the appropriate literature databases ormicrosco-

py textbooks, literature that covers the topic from both physics and biology

perspectives.

Internal conversion, vibration state(loss of energy) (ps)

High energyexcited states

Absorption(excitation)(fs)

Absorption(excitation)

Wavelength

Rel

ativ

e in

tens

ity

Excitation light

B

A

Low energyground state

Stokes shift

Emission

Emission(ns)

Emission light(longer wavelength)

Fig. 2 (A) A Jablonski diagram showing the energy states of a given fluorophore, including the

process of absorption and emission of longer wavelength light upon excitation. The duration of each

state is also indicated. (B) The absorption and emission spectra of a given fluorophore including the

Stoke’s shift due to emission of a longer wavelength photon upon excitation of the fluorophore.


V. Total Internal Reflection Fluorescence Microscopy

Confocal microscopy coupled to TIRF provides a very thin optical section

of fluorescence excitation that allows imaging with low background noise and

minimal out-of-focus fluorescence. This is because total internal reflection can

only occur when the excitation light beam in a medium of high refractive index

reaches an interface of a medium with a lower refractive index at an angle of

incidence that is greater than the specific critical angle y. When the light is

totally internally reflected, none of it penetrates the medium with the lower

refractive index, and ideally, there is no net energy flux escaping the glass.

Incident light

Incident light

Range of incidenceangles greater thanthe critical angle q

Specimen, low refractive index

Glass, high refractive index

Reflected light

Evanescent field<100 nm

Reflected light

Objective

SpecimenGlass

A

B

Fig. 3 Total internal reflection fluorescence (TIRF) microscopy. (A) Overview including the inci-

dent and reflected laser light paths within the objective. (B) Once the incident light reaches a medium

with a lower refractive index at an angle greater than the critical angle y, the incident light does not

penetrate the specimen, but an electromagnetic field is created that penetrates up to ~100 nm above

the surface, called an evanescent wave. This may excite fluorophores within the range of the evanes-

cent wave.


However, the reflected light generates an electromagnetic field that penetrates

beyond the interface and into the lower refractive index medium as an evanes-

cent wave. This wave, with a wavelength similar to the excitation light beam,

decreases exponentially with the distance into the medium. The penetration

depth of the evanescent wave may be manipulated by changing the angle of

incidence beyond the critical angle, which may be calculated accurately by

knowing the angle of incidence, but will typically be limited to �100 nm or

less (Cleemann et al., 1997; Mashanov et al., 2003). This allows for imaging of

Ca2þ events occurring in close proximity to the plasma membrane of live cells.

However, this does require that the cell is positioned within the evanescent wave

and not above it on the glass surface, which may well be the case for the

majority of the cell if the cell is of a certain size. In our experience, it requires

an experimental eVort to physically position large cells within the evanescent

wave and yet still maintain physiological conditions for the cell. This is required

because the evanescent wave is generated from the glass surface and not from

the boundary of the cell.


VI. Forster Resonance Energy Transfer Microscopy

Confocal microscopy coupled with FRET imaging has experienced something

of a surge of interest recently fromCa2þ researchers, because of the development of

Ca2þ-sensitive cameleons. The FRET process is based upon nonradiative energy

transfer from a fluorophore in an excited state (‘‘donor’’) to another chromophore

(‘‘acceptor’’) that usually is, but does not have to be, a diVerent fluorescent

molecule within a range of 10–100 Angstroms (1–10 nm) (fluorophore) (Jares-

Erijman and Jovin, 2003). Thus, when FRET occurs, it is not only the acceptor

emission that will be recordable, but the measurable emission from the donor will

also be greatly reduced because of the energy transfer. Also, the intensity of the two

emission bands will depend on the distance between the two donor and acceptor

fluorescent proteins or molecules even within the FRET distance. The closer the

distance between the donor and acceptor, the greater the longer-wavelength emis-

sion (from the acceptor) and the less the shorter-wavelength emission (from the

donor). The fluorescence emission from the acceptor is discernible from the donor

emission due to the spectral redshift, such that the instrumental requirement is that

the two emission bands must be separately recordable. The presence of the longer-

wavelength acceptor emission will confirm that the acceptor molecule is within

the FRET distance of the donor, that is, within a distance of �10 nm. Likewise,

the absence of it suggests that the physical distance between the donor and

acceptor is larger than that. Traditionally, FRET imaging has been used to study

protein–protein interactions by tagging diVerent proteins with fluorescent probes

(e.g., yellow or green fluorescent proteins (YFP or GFP)) and thereby study

whether they exist within or outwith the FRET distance.

However, it is the recent development of cameleons that has moved FRET

imaging to also become a sought-after technique with respect to studies of Ca2þ.Cameleons are genetically encoded fluorescent proteins that are sensitive to Ca2þ,in which a blue- or cyan-mutant of GFP (serving as the donor), calmodulin that

can bind to Ca2þ, a calmodulin-binding peptide such as the calmodulin-binding

domain of skeletal muscle myosin light chain kinase (the M13 peptide), and a

long-wavelength yellow mutant or normal GFP (serving as the acceptor) are

tandemly fused (Miyawaki et al., 1997). This configuration forms a stable and

compact complex that remains intact once transfected into the target cell, and

subsequent development has also improved the spectral properties and rendered

the cameleons less susceptible to changes in pH (enhanced GFPs). In the absence

of Ca2þ, the cameleon remains in its linear tandem configuration, whereby the

two GFP mutants (donor and acceptor) at the two flanks of the tandem are too

far apart to create FRET. However, when the concentration of free Ca2þ

increases, calmodulin binds to Ca2þ and undergoes a conformational change

that also leads it to bind and wrap around the M13 peptide, which creates a

compact configuration of the cameleon and therefore brings the donor and

acceptor GFP mutants to within a distance where energy transfer may occur


with much greater eYciency. Thus, the presence of Ca2þ creates FRET by the

cameleon, which may be readily detected. Moreover, because Ca2þ aYnity can be

tuned by incorporating mutations into the calmodulin protein, cameleons may

detect free Ca2þ concentrations in the range 10 nM–10 mM, and this has been

done to visualize local Ca2þ signals in the nucleus, sarco/endoplasmic reticulum,

and the cytosol, by transfecting chimeras of the cameleon that also have the

appropriate localization signals encoded in the complementary DNA (Miyawaki

et al., 1997). A typical example of a cameleon used for FRET imaging of Ca2þ has

an excitation spectrum peak at 442 nm for the donor (blue mutant of GFP), with

the associated emission peaking at 486 nm. This wavelength FRET-excites the

GFP on the acceptor side, which then emits longer wavelength fluorescence with a

peak at 530 nm (Fig. 4).

Several protocols for detecting and measuring FRET eYciency have been

developed, of which some are more applicable to imaging of Ca2þ signals than

others. These include measuring donor quenching, that is, measuring the decrease

in the emission from the donor fluorophore, which appears because some of the

energy emitted by the donor fluorophore is used to excite the acceptor chromo-

phore/fluorophore. This is done by taking the ratio between donor and acceptor

fluorescence during FRET as the numerator, and the same ratio in the absence of

FRET (i.e., by removing either of the donor or acceptor fluorophores) as the

denominator. Like all ratiometric quantifications, this has the advantage that the

measure becomes independent of local variations in fluorescence. The disadvan-

tage is that both the donor and acceptor fluorophores may be quenched by other

factors that would misrepresent the results. Another method for measuring FRET

eYciency is by measuring donor quenching and acceptor photobleaching, that is,

the intensity of the donor emission in the presence of an acceptor relative to the

intensity of the donor emission in the absence of an acceptor. In practice, the

latter is done by first photobleaching the acceptor by illuminating it with light at

the peak excitation spectrum of the acceptor fluorophore before measuring the

donor emission. This protocol relies on the fact that fluorescing itself causes the

fluorophores to lose their ability to fluoresce, a process called photobleaching.

However, because photobleaching may take many minutes to achieve (up to

�20 min), this may become less available in live specimens. Finally, FRET

combined with fluorescence lifetime imaging (FLIM) has introduced a robust

option for Ca2þ measurements, because it largely is unaVected by experimental

conditions such as fluorophore concentrations and excitation intensities. In this

combined FRET–FLIM approach, the change in donor lifetime is measured in

the presence and absence of an acceptor. The principle behind this is that the

period of time the donor will fluoresce (i.e., the lifetime of the donor) depends on

the presence or absence of an acceptor (Levitt et al., 2009). As described above,

the measurement of donor emission in the absence of an acceptor may be done by

first photobleaching the acceptor. However, once the photobleached control

images have been captured, this protocol allows for detailed imaging of Ca2þ

signals over a short time period.

Donor:Blue orcyan FP Cam

Excitation:

A

B

Emission:

Excitation:

486 nm 530 nm

Wavelength

Rel

ativ

e em

issi

on in

tens

ity

Emission:

FRET<100 nm

0 Ca2+ +4 Ca2+

4 Ca2+

+Ca2+

−Ca2+

442 nm

442 nm

486 nm

530 nm

M13

Acceptor:Green oryellow FP

Fig. 4 Cameleon-based Forster resonance energy transfer (FRET). (A) Schematic of the cameleon in

the absence and presence of Ca2þ; note the conformational change in the calmodulin (Cam) and the

Cam-binding domain of myosin light chain kinase (M13) upon binding to Ca2þ that allows for FRET

between the donor and acceptor green fluorescent protein (FP) mutants. (B) The relative emission

intensities at different wavelengths indicate whether or not Ca2þ is present, and hence, FRET occurs.


Since the introduction of cameleons, derivatives have been developed that fuse

Cam with the Cam-binding peptide M13 in a similar fashion as cameleons, but are

based on single GFPs instead of two GFP mutants with diVerent spectral proper-ties that necessitate Ca2þ imaging by FRET. These derivatives, called pericams,

can therefore be imaged by standard epifluorescence or confocal microscopes, and

present with broad Ca2þ sensitivities and diVerent localization sequences that

allow for measurements of a range of Ca2þ concentrations within specific orga-

nelles and intracellular targets (Kettlewell et al., 2009; Nagai et al., 2001).


VII. Parallel Scanning Confocal Systems

The main limitation of the traditional laser scanning confocal microscope sys-

tems discussed so far is their low full frame temporal resolution. At typical rates of

just a few Hertz, an increase of 1–2 orders of magnitude in the speed of image

acquisition is necessary to decipher the fast Ca2þ dynamics of systems such as

neuronal networks or cardiac and muscle tissues.

Recent developments in parallel multispot confocal illumination devices have

gone some way to addressing these concerns, providing high contrast optical

sectioning with typical rates in the 10–100 Hz domain. The multispot system uses

an aperture mask at an illumination plane conjugate with the sample, where

multiple illumination points with nonoverlapping airy disk profiles are projected

to the sample simultaneously. This illumination pattern is then changed in a

sequential fashion such that every image point is uniformly illuminated in a

given time interval. Two approaches have been used in recent years to successfully

achieve this fast pattern change: the Nipkow spinning disk where a patterned disk is

spun at high speeds, and programmable matrix systems (digital mirror and liquid

crystal arrays) where individual pixels can be switched ‘‘on’’ and ‘‘oV’’ at very highspeeds.

The parallel illumination approach has the obvious advantage that all image points

can be illuminated much faster than in a conventional single scan system, but the

drawback is that a 2D imaging detector is required to record the image, which

then becomes the limiting factor for capturing image frames. Limitations of the

detectors (typically CCD cameras) are further exacerbated by the requirement for

fast detection, as these devices have a noise floor below which the fluorescence signal

cannot be resolved. The fundamental problem is that in order to increase frame rates,

the exposure time of a frame has to be reduced, thus limiting the number of detectable

photons in an integration interval. Simultaneously, the transfer rate of data from

the camera has to be increased, a process by which the read noise of CCD cameras

increases. Many of these concerns are resolved by modern high-end electron multi-

plication (EM) CCD cameras such as the iXon 897 (Andor), Evolve (Photometrics),

or ImagEM (Hamamatsu). These EM-CCD cameras can amplify small photoelec-

tron signals above the read noise of the camera; providing the sensitivity and speed

required making these theoretical parallel approaches a practical reality.

VIII. Spinning Disk Confocal Microscopy

These systems illuminate the specimen simultaneously with a large number of

non-overlapping points of light by using multiple (hundreds to thousands) pinholes

arranged in a geometrically precise spiral pattern on a spinning disk (Nipkow disk).

The disk is placed at an image plane conjugate with the sample, and the illumination

is filtered by this disk. Thus, the specimen is raster scanned rather than single-spot


scanned, such that the entire 2D frame may be illuminated semi-simultaneously.

Fluorescence emitted from each illumination point then returns through the same

aperture of the mask to give a set of confocal points in the image. When the disk is

spun at high speed (typically several thousand revolutions per minute), an appar-

ently continuous image is obtained when viewed through an eyepiece. The resultant

image is traversed with scan lines, but precise synchronization of the illumination

and detection systems virtually eliminates this artifact in well configured systems

(Wilson et al., 1996). Thus, this approach allows for very fast scan rates without

significantly compromising the SNR (Wang et al., 2005).

As the input illumination power is spread over a much larger area than with

conventional scanning confocal systems, light throughput can be a serious impedi-

ment to successful implementation of spinning disk systems. This was partially

resolved by enhancing the original design by means of a second disk spinning in

sequence with the Nipkow pinhole array disk. This second disk is sited between the

dichroic and the light source and contains a microlens array that maps a miniature

lens to each pinhole (Tanaami et al., 2002), thus improving the illumination

eYciency by focusing the lightbeam onto the pinhole (Fig. 5A). This also reduces

the backscattering of light at the surface of the Nipkow disk, which substantially

increases SNR. The emission detection pathway is not aVected by this modifica-

tion. Use of specialist cameras and fast versions of the spinning disk head can now

enable imaging rates of up to 2 kHz. However, the drawbacks are that the pinholes

on the spinning disk are inflexible, and the dwell time per pixel is usually very short

(�100 ns), which may severely reduce the SNR, although the frequent illumination

of the same pixel as the disk rotates may compensate for this eVect.

IX. Programmable Matrix Microscopy

These instruments are based on the principle of spatially filtering full field

illumination in a defined pattern at high speed, so as to give it a prescribed,

dynamic structure. By using appropriate filtering patterns, the device can simu-

late the optical behavior of confocal scanning microscopes. These systems are

directly comparable to the spinning disk approach in that the illumination device

consists of an array of small apertures that act as both the illumination and

detection pinholes. The principal diVerence, and advantage, is that the elements

of the array are individually addressable, allowing far greater flexibility in exper-

imental design compared to the spinning disk (Hanley et al., 1999). As a practical

example, selectively and sequentially illuminating individual cells within the field

is possible with the array system, whereas the Nipkow disk can only operate at

full frame.

Two technologies have been used in implementing practical programmable

matrix systems. The first makes use of a digital micromirror device (DMD), an

array of micrometer-sized mirrors whose angle can be independently controlled

to direct illumination to an ‘‘on’’ (confocal) or ‘‘oV’’ (non-confocal) pathway.

A

Nipkow pinhole array disk

Objective

Lightdetector 1(conjugate)

B

Lightdetector 2(nonconjugate)

Light source





Objective

ReflectiveLCOS array

Light detector

Emitted light

Excitationlight beam

Microlens array disk

Microlenspinhole

Fig. 5 (A) Nipkow spinning disk. The figure shows two light beams reaching the specimen and the

light detector, although, in reality, multiple beams reach the specimen and the light detector simulta-

neously. (B) The optical pathway of a programmable matrix microscope. The liquid crystal on silicon

(LCOS) array consists of numerous mm-sized liquid crystal ‘‘pixels’’ that can be switched independently

to reflect the lightbeam onto multiple spots simultaneously.


The alternative uses a reflective liquid crystal display (LCD), an array of microm-

eter-sized liquid crystal ‘‘pixels,’’ the base of which is coated with a reflective layer

(Fig. 5B). Each element is individually addressable, creating a controlled polariza-

tion pattern that is optically transformed into a confocal intensity pattern. So far,

the liquid crystal approach forms the basis for the commercial utilization of this

technique. This allows for using both the confocal image (conjugate with the array


mask) and nonconjugate image to provide further potential enhancements in

dynamic image quality (Heintzmann et al., 2003).

The real power of the programmable matrix approach lies in the ability to

redefine the array composition, without any additional hardware changes. This

enables complete flexibility for optimization of the illumination pattern for the

sample and experiments in hand—from low speed high resolution to high speed

lower resolution studies.

X. Advantages and Disadvantages of Confocal Microscopy

From an experimental point of view, confocal Ca2þ imaging oVers many benefits

over conventional fluorescence microscopy, especially with respect to spatial and

temporal resolution and the ability to optically section the specimen along the

Z-axis down to a resolution of 0.5–1 mm, enabled by the presence of a confocal

aperture with a pinhole in the light path that eVectively gives this imaging modality

its uniqueness, as well as its name. Importantly, confocal microscopy may be

performed in live cells that are electrically, mechanically, and by all other mea-

sures, viable. Moreover, confocal microscopy does not impair the ability to

manipulate the surrounding environment in terms of solutions in which the cells

or specimen are bathed, solution pens, temperature devices, application of phar-

macological drugs and inhibitors, patch pipettes, electrical and mechanical manip-

ulators, etc., as long as these do not interfere with the actual light pathway.

However, the significant drawbacks with confocal microscopy relate firstly to the

unavoidable scattering as well as chromatic aberration with subsequent focal plane

oVsetting that reduces confocal signal collection (Bliton et al., 1993) of the illumi-

nation light as it penetrates the specimen, especially since wavelengths of

�300–600 nm are commonly used for excitation, as those are very prone to

scattering; eVectively limiting the penetration depth to �40–50 mm (or less,

depending on the laser output and the spectral properties of the specimen), and

secondly to the fact that although only the emitted fluorescence produced at or

very near to the focal point will be recorded, excitation will still occur along the

whole Z-axis of the specimen (given molecules able to be excited are present along

the Z-axis). The latter point presents often considerable problems because the

photodamage generated along the entire Z-axis may kill live specimens. Though

this may be mitigated by lowering the laser power, the restricted light capturing

from the small volume accommodated by the pinhole tends to drive toward

microscopists increasing laser power. This point is further accentuated by excita-

tion at short wavelengths (300–500 nm) including UV light, as those are severely

damaging to live specimens due to the associated high energy. Finally, higher

purchase and maintenance costs relative to conventional epifluorescence and wide-

field microscopy may also represent a disadvantage to the researcher.

A diVerent excitation approach with many of the same advantages as confocal

laser scanning microscopes as described above, while also substantially limiting the


major drawbacks of confocal microscopy such as the restricted penetration depth

and photodamage, is oVered by multiphoton microscopy systems.

XI. Multiphoton Excitation Laser Scanning Microscopy

A conceptually diVerent microscope that also allows optical sectioning and high

spatial and temporal resolution live specimen imaging, and that also comes with

other added benefits, is the multiphoton microscope. From a biophysical perspec-

tive, multiphoton imaging rests on an excitation principle that relies on excitation

of the fluorescent molecule by photons that alone do not have enough energy, but

need to combine by simultaneously coming in to a very close proximity of the

fluorophore.

Although the theoretical concept of multiphoton excitation is relatively simple and

has been known and also utilized by physicists for many decades, biological applica-

tions ofmultiphoton excitation are of amore recent date (Denk et al., 1990).Whereas

in confocal single-photon laser illumination, the fluorophore is excited by the absorp-

tion of a single photon, as it provides suYcient energy for the fluorophore to reach an

excited state. Upon return to the ground state, a photon of longer wavelength than

the excitation photon is emitted, and it is this process that creates the fluorescence.

With multiphoton excitation, a fluorophore is excited by the near-simultaneous

(within 10�18 s) absorption of two or more photons that combined provide enough

energy to promote the fluorescent molecule from a ground state to an excited state;

two photons for two-photon (2P) excitation, three photons for three-photon (3P)

excitation, etc., with 2P excitation being by far the most common multiphoton

excitation modality. Thus, with 2P excitation, the fluorophore absorbs two photons

simultaneously, each twice the wavelength and half the energy required formolecular

excitation, and likewise, in 3P excitation, each of the three excitation photons have

three times the wavelength, but only one-third of the energy compared to single-

photon excitation (Helmchen and Denk, 2005).

Once excited, the emitted fluorescence is then proportional to the square of the

excitation intensity in 2P absorption (third power in 3P excitation), and this 2P

excitation (measured in Goeppert-Meyer Units) occurs only at the focal point, as

it is only here that the density of the excitation photons is high enough to ensure

a simultaneous photon arrival to the fluorophore that is suYcient to excite it.

Though this may not occur with all photons in the volume, the probability of it

occurring for the vast majority of the photons is very high. This nonlinear

excitation constitutes the most important physical diVerence from confocal sin-

gle-photon excitation, in that excitation will be confined to a small ellipsoid

volume around the focal point, whereas above and below this point, the density

of photons suYciently close to one another, or in other words the light intensity,

will not be high enough to generate any excitation (Fig. 6). This eVectively means

that out-of-focus fluorescence will not be generated, which again removes the

need for a confocal aperture with a pinhole in the emission light pathway.

Single-photon

BContinuous single-photon laser

Femtosecond pulsed 2P laser

Time

Time

100 fs

10 ns

Excitation

Excitation and emission intensitiesPhotonsFocal point

Emission2P

Single-photon 2P Single-photon 2P

A

Fig. 6 Two-photon (2P) excitation microscopy, including a comparison to single-photon confocal

microscopy. (A) Schematic of illumination lightbeams, in which excitation occurs along the whole Z-

axis with single-photon confocal microscopy, though with highest intensity at the focal point, whereas

excitation is confined to a narrow area around the focal point with 2P microscopy. Correspondingly,

emission occurs along the entire z-axis with confocal microscopy; though out-of-focus light is blocked

by the confocal aperture, whereas emission is confined to a narrow area around the focal point with 2P

excitation microscopy. (B) Schematic of the continuous laser used for single-photon excitation and a

pulsed titanium:sapphire (Ti:Sapphire) laser used for 2P excitation.


Therefore, the optical sectioning and spatial and temporal resolution one may

achieve will be close to that achieved by confocal microscopy or a little poorer,

but not better (Cox and Sheppard, 2004). A clear benefit of this is that there need

not be a confocal aperture. Although crucial for confocal single-photon micros-

copy, confocal apertures themselves reduce light transmission, but also require

that the emission light path is returned through the scanhead including the

mirrors, which further causes a loss of light. Instead, the emitted fluorescence

after 2P excitation may be focused directly onto a PMT without having to be

descanned, as in confocal microscopy, such that 2P microscope systems are

generally more light sensitive. However, the introduction of a confocal aperture

with a pinhole into the 2P emission pathway may improve the spatial resolution

to a confocal standard.


Thus, although the capture of the emitted fluorescence may be set up with

diVerent light paths in 2P microsscopes compared to confocal single-photon

microscopes, it is conceptually similar in that emission light needs to be split by

the excitation light with the insertion of an appropriate dichroic mirror. However,

in the case of 2P microscopy, the emitted fluorescence will in most cases have a very

much shorter wavelength than the excitation wavelength, which is in contrast to

confocal single-photon and conventional epifluorescence microscopes.

Because of the abovementioned properties of 2P excitation and diVerences to

single-photon confocal excitation, 2P excitation oVers several important advan-

tages over single-photon excitation. First, because light scattering declines steeply

with increasing wavelengths, red and IR (670–1100 nm) light can penetrate and

hence excite fluorescent molecules much deeper into the specimen than single-

photon lasers providing excitation wavelengths in the range from �300 to

�600 nm. With suYcient power outputs and optimized optics including corrective

optics for pulse dispersion, penetration depths approaching 1000 mm (1 mm) may

be achieved, which is considerably deeper than the �40–50 mm achievable with

single-photon excitation. Thus, 2P excitation secures an up to �20-fold deeper

penetration depths by using excitation light of twice the wavelength. To achieve

comparable deep imaging with single-photon confocal microscopy, tissue or layer

removal with histologic techniques or penetration by the objective would have been

required. Such mechanical approaches would for obvious reasons compromise the

‘‘intactness’’ and viability of the tissue. For this reason, confocal microscopy is

mainly restricted to studies of single, isolated cells or tissue surfaces, whereas 2P

excitation laser scanning microscopy allows deep tissue imaging of intact organs.

Similarly, 2P excitation microscopes are comparably more often upright rather

than inverted, to allow for tissue preparation to be set up on the stage, whereupon

the objective lens is lowered down to within the optical range for imaging. The

opposite is true for confocal microscopes, which more often than not come

inverted. Furthermore, even with 2P excitation deep tissue imaging, one can be

assured that the vast majority of the fluorescence comes from the focal point and

not from residual scattering (scattering is greatly reduced, but not obliterated, by

long-wavelength excitation). This is because, even in strongly scattering tissue such

as the heart, the density of scattering exciting photons is too low to generate

significant fluorescence. Therefore, this together with the long excitation wave-

lengths also contributes to 2P excitation microscopes being much less sensitive to

light scattering than regular widefield epifluorescence or confocal microscopes.

However, it should be noted that penetration depths also heavily depend on the

specimen, as diVerent tissue properties may degrade the illumination as well as the

emitted fluorescent light. In particular, collagen and myelin are known to scatter

light and thereby restrict penetration (Helmchen and Denk, 2005).

Additionally, IR light is much less phototoxic than shorter wavelengths, such

that it correspondingly may cause only negligible photodamage. Also, because of

the nonlinear excitation, photodamage and bleaching are restricted to the focal

point. Importantly, the reduced photodamage and fluorophore bleaching allows


for longer imaging periods. 2P excitation spectra of most fluorophores are wider

than the equivalent single-photon excitation spectra, whereas the associated emis-

sion spectra after 2P and single-photon excitation are not diVerent, such that the

same excitation wavelength may potentially excite several fluorophores with dis-

tinct emission spectra (Bestvater et al., 2002; Xu et al., 1996). This property will be

explored in more detail later in the chapter. Moreover, whereas the positioning of

the confocal aperture relative to the light detector as well as the focal plane is

critical to ensure that the images of the source and detector apertures are cofo-

cused, this is not the case for 2P excitation microscopes, in which the position of

the light detector is not critical, because fluorescence emission will only be gener-

ated at and tightly around the focal volume.

However, several problems persist with 2P excitation. First, 2P excitation requires

very high light intensities; intensities that would instantly vaporize the specimen if

the light was delivered continuously. However, this is overcome by using lasers that

provide ultrabrief (10�13 s) pulses at very high frequencies (�80–100 MHz), which

thereby generate very high instantaneous energy, but suYciently low average energy

to avoid any substantial damage to the specimen. Thus, the distance between each

pulse is typically �10 ns, whereas the width of each pulse itself is typically 100 fs

(Fig. 6). The high repetition rates match fluorescence lifetimes closely (see above)

such that a good balance between excitation eYciency and onset of saturation is

achieved (Helmchen and Denk, 2005). The most widely used lasers that fulfill this

criterion and provide the necessary wavelengths, are the solid state titanium:sap-

phire (Ti:Sapphire) oscillating (pulsed) lasers. These lasers are tunable, such that the

latest versions are capable of delivering wavelengths within the range �670–

1100 nm, thus, from visible red to infrared (IR), though this range is continuously

expanding while also coming with higher power outputs, as manufacturers keep

developing their product lines. Though most 2P excitation applications may not be

power-limited, more available power does benefit applications requiring deep

tissue penetration. Typical power outputs in the latest Ti:Sapphire lasersmay exceed

4 W at 800 nm, though the optics cause a major loss in power from the outlet of the

laser to the focal point of the objective lens. Glass components of the optical

pathway also cause a dispersion of the �100 fs laser pulses, but this may be

compensated for by corrective optics in order to maximize 2P excitation (Diels

et al., 1985). Other limitations of 2P excitation microscopy are that reflected light

imaging is not possible; only fluorescence imaging, and that it is not suitable for

imaging highly pigmented specimens, as these absorb IR and near-IR light.

XII. Ca2þ Indicators for Use in Confocal andMultiphoton Microscopy

For many years, biology has benefited from the fluorescent dyes or constructs

that bind and therefore can measure the free concentration of Ca2þ [Ca2þ]. Theseindicators have been used to examine the levels and time course of changes


in [Ca2þ] within the cytosol and the various organelles within the cell (e.g.,

nucleus, mitochondria, sarco/endoplasmic reticulum). The two main categories

of molecules used for this purpose are small synthetic organic molecules based

on the fast Ca2þ buVer BAPTA (Tsien, 1980) or modified versions of natural

Ca2þ binding proteins (Miyawaki et al., 2003). Both categories ‘‘sense’’ Ca2þ by

chelating the ion which changes the structure/chemical properties of the ligands.

Several modes of fluorescence have been utilized to report Ca2þ (summarized in

Fig. 7).

Absorbance/quantum yield: In this case, Ca2þ binding causes a change in the

intensity of the fluorescence from the dye; typically, Ca2þ binding causes an

increase in fluorescence (Fig. 7A). This mode is the most common employed by

the fluorescent Ca2þ indicators used in confocal or 2P microscopy. In particular,

+Ca

+Ca +Ca

+Ca

−Ca

−Ca

−Ca

−Ca

Quantum yieldA B

DC

Spectral shift

Wavelength

Fluorescence life time Förster resonance energy transfer

Time (ns)

Wavelength

Wavelength

Ligh

t int

ensi

tyLi

ght i

nten

sity

Ligh

t int

ensi

tyLi

ght i

nten

sity

Fig. 7 Major categories of fluorescence properties of Ca2þ indicators. (A) Change in dye absor-

bance and quantum yield generates a Ca2þ-sensitive change in the fluorescence intensity (e.g., Fluo-3/

4 Rhod-2, Oregon Green, and Fura-Red (inverse relationship)). (B) Spectral shift in the excitation

spectrum as a result of Ca2þ binding to an indicator allows ratiometric measurements (e.g., Fura-2/3/

4/6/FF). (C) Changes in fluorescence life time as a result of Ca2þ binding to an indicator; the

fluorescence decays exponentially within ns of the end of excitation. The rate of decay is Ca2þ

dependant; for example, the decay of Ca2þ-bound Fluo-3 fluorescence is faster than the decay of

unbound Fluo-3. (D) Change in Forster resonance energy transfer (FRET) efficiency as a result of

Ca2þ binding to either the acceptor or donor proteins; Ca2þ binding changes the distance between the

two linked fluorescent proteins. The distance between the donor and acceptor proteins determines the

degree of FRET; an increase in FRET efficiency causes a decrease in donor fluorescence and an

increase in acceptor fluorescence. Black lines indicate excitation spectra and gray lines indicate

emission spectra. Dotted lines represent the Ca2þ-free form of the dye, whereas solid lines represent

the Ca2þ-bound form of the dye.


Fluo-3, Fluo-4, and Oregon Green all show dramatic increase in fluorescence

upon binding to Ca2þ, while Fura-Red fluorescence decreases upon Ca2þ binding.

The optical arrangement required for these dyes is the simplest, that is, single

excitation and emission wavelengths.

Spectral shift: In this case, Ca2þ binding causes a shift in either the excitation

or emission wavelengths. Fura-2 and related dyes show a shift in the excitation

spectrum of the dye with minimal changes in the emission (Fig. 7B). In contrast,

Indo-1 and related dyes show large shifts in the emission spectrum of the dye

with minimal changes in the excitation spectrum. This spectral shift is used to

allow ratiometric measurements. In the case of Fura-based dyes, while exciting

at �340 nm, a rise of Ca2þ will cause an increase in fluorescence, whereas while

exciting at 380 nm causes a decrease. The ratio of fluorescence at these two

wavelengths is a unique function of the [Ca2þ] and is independent of the

concentration of the dye. This is particularly useful when measuring cells that

move and thereby alter the amount of dye within the light path, or comparing

cell compartments with diVerent dye concentrations (e.g., nucleus versus the

cytosol). However, despite all these advantages, the Fura- and Indo-based dyes

are rarely used in confocal microscopy because excitation wavelengths are

short (<400 nm) and therefore not convenient for commonly available lasers.

The shortest wavelength routinely available in commercial systems is 405 nm,

which can be used to excite Fura-based dyes close to the 380 nm excitation

maximal. In this mode, Fura acts as an inverse indicator, a property that has

some value in single-photon confocal and 2P imaging (Ogden et al., 1995;

Wokosin et al., 2004).

Fluorescence lifetime: In this case, the decay of fluorescence at the end of a pulse

of excitation light will take a finite time (ns) to decay. The time course of decay is

aVected by the binding of Ca2þ (Fig. 7C). In the case of Fluo-3, the decay of

fluorescence or lifetime is shorter for the Ca2þ bound form (Sanders et al., 1995).

Since the lifetime is independent of the concentration of the dye, the relative

population of bound and free dye can easily be calculated from this technique.

However, the nontrivial analysis of emission decay required to obtain lifetime data

has minimized the use of this technique.

FRET eYciency: In this case, a donor molecule absorbs photons and can transfer

the associated energy to a close by acceptormolecule via a nonradiative process that

operates at distances less than the wavelength of light (up to 10 nm). The closely

adjacent molecule accepts the energy transfer (thus excites) and then emits light at a

distinct wavelength (Fig. 7D). The most popular pairs of donor and acceptor

molecules are cyan fluorescent proteins (donor) and yellow fluorescent proteins

(acceptor). The eYciency of the energy transfer depends on the proximity of the

two molecules; within 10 nm, the eYciency is high but this drops dramatically as

the distance between the two molecules increases (proportional to 1/(distance)6).

This mechanism can be used to detect Ca2þ binding to either construct since the

change in tertiary structure will alter FRET eYciency and therefore altered

FRET signal.


XIII. Use of Dyes for Single-Photon Confocal Microscopy

In theory, all the fluorescent Ca2þ indicators created for cell biology can be used

in confocal or 2P excitation modes, although in practice, only a limited number of

dyes are routinely used for a series of practical considerations. By far, the majority

of confocal applications use single-wavelength excitation and emission dyes. These

dyes have several advantages: (i) their excitation wavelength is �500 nm or longer

and therefore can use the emission from readily available Argon or Krypton-

Argon lasers, and (ii) the majority of dyes have a good dynamic range (see

below). But they also suVer from a series of disadvantages: (i) calibration is diYcult

because the signal will be a function of both the concentration of the dye (unknown

and variable) and the concentration of Ca2þ and (ii) dye signal may vary with time

due to loss of the dye from the volume or due to photobleaching. All dyes

photobleach, but some dyes are more susceptible than others; in particular,

Fluo-3 and Fluo-4 are amongst those that most rapidly bleach (Thomas et al.,

2000). Dyes that show significant spectral shifts can be used in a single wavelength

mode, for example, Fura-based dyes can be imaged in confocal microscopy by

using a 405 nm laser. This produces an inverse reporter since fluorescence

decreases as Ca2þ increases (Wokosin et al., 2004). Similar measurements using

Indo-based dyes are less easy because the narrow excitation spectrum means that

405 nm laser light only poorly excites the dye and Indo-based dyes appear to be

inherently more prone to photobleaching.

XIV. Use of Dyes for 2P Excitation Microscopy

The utility of a dye for imaging Ca2þ using 2P excitation does not follow directly

from the behavior of the dye in single-photon excitation. There is no guarantee

that 2P excitation will successfully excite the dye since the eYciency of two-photon

excitation appears to be relative to the fluorophore structure (Kim et al., 2008).

A number of Ca-sensitive indicators have been studied over a limited range of 2P

wavelengths (Xu et al., 1996), some dyes (e.g., Oregon Green) have a better 2P

cross-section than the more common single wavelength dyes (Fluo-3); none

showed the expected increase in cross-section as the excitation wavelength

approached 900 nm. In contrast, the 2P cross-section of Fura-based dyes appeared

readily excitable by wavelengths approximately to double the appropriate single-

photon wavelength (Wokosin et al., 2004). In particular, �800 nm light provides

an inverse Ca2þ-sensitive signal that corresponds to the excitation of this dye at

�400 nm, as advocated previously (Ogden et al., 1995). However, few studies to

date have explored the more commonly used dyes (Fluo-3/4 and Rhod-2) at the

longer wavelengths achievable currently with tunable Ti:Sapphire lasers (up to

1100 nm). As shown in Fig. 8, the Fluo-based dyes do not show the peak of

fluorescence anticipated from approximately double single-photon wavelengths

(�1000 nm). The small peaks in the excitation spectrum observed at 400–470 nm

1000

100

10

1000

100

10

1000 1P × 22P

1P × 2

2P

900

Wavelength (nm)

1000 1200

Excitation wavelength (nm)

800

800

1000900800700

700

Fluo-3

Rhod-2B

A

Flu

ores

cenc

e (a

.u.)

600

400

200

0

1000 1100 1200900800700

Flu

ores

cenc

e (a

.u.)

600

400

200

0

Fig. 8 Excitation spectra of Ca2þ bound forms of (A) Fluo-3 and (B) Rhod-2. Emitted fluorescence

was collected at 500–650 nm (Fluo-3) and 550–650 nm (Rhod-2). Black line represents the single-photon

(1P) excitation spectra measured on a Perkin–Elmer spectrophotometer (2 nm slit width); the 1P

wavelength has been scaled by a factor of 2 (�2) to generate comparable wavelengths to two-photon

(2P) spectra. Gray line represents the 2P excitation spectrum of an equivalent concentration of dye (10

mM); excitation light was provided by a Coherent Chameleon XP Ti:Sapphire laser attached to a Zeiss

510 upright laser scanning microscope. Laser power was altered at each wavelength to ensure equivalent

excitation power across the excitation wavelengths. Insets show spectra plotted on log scale.



during single-photon excitation were also paralleled by similar spectra at

800–950 nm during 2P excitation, but longer wavelengths failed to produce the

increased fluorescence anticipated from the single-photon spectrum. The behavior

of the Rhod-dyes (related structurally to Fluo-dyes) has yet to be examined in

detail mainly because accessing wavelengths >1100 nm is technically diYcult.

The practical application of 2P excitation to excite Ca2þ-sensitive dyes is limited;

the majority of applications used �800 nm to excite either Fluo-3 or Rhod-2.

In doing so, these studies are using the higher powers available from lasers at

these wavelengths, rather than attempting to access the higher quantum yields that

may be present at longer wavelengths. Further work is clearly required to deter-

mine the optimal 2P conditions to excite these readily available dyes at 1000–

1100 nm, or alternatively design new dyes with fluorophore structures that are

more easily excited by the 2P approach (Kim et al., 2008).

XV. Is It Worth Converting the Intracellular FluorescenceSignal to [Ca2þ]?

Ca2þ-sensitive dyes are frequently used in conjunction with confocal microscopy

to simply indicate the timing or frequency of transient Ca2þ events. Under these

circumstances, it is not considered necessary to convert the fluorescence signal into

estimates of intracellular [Ca2þ]. This is avoided partly for experimental ease, since

conversion requires knowledge of the concentration of the dye and its aYnity for

Ca2þ, and partly because absolute Ca2þ concentrations are not considered vital to

the interpretation of the experiments. In a number of cases this may be justified,

but whenever amplitude or time course of a Ca2þ transient is considered an impor-

tant variable, then calibration becomes essential for the following several reasons:

1. It is important to distinguish changes in background Ca2þ from one experi-

mental scenario to the next since this determines the subsequent intracellular Ca2þ

buVer power. Any event that generates a Ca2þ transient, for example, Ca2þ influx

via plasmalemmal Ca2þ channels or Ca2þ release from an internal store, will

increase total intracellular Ca2þ by a specific amount, but the extent to which

this increases the free cytosolic Ca2þ concentration will depend on the cellular

buVer power for Ca2þ. This is illustrated in Fig. 9, where total cellular Ca2þ is

increased by a standard amount; the subsequent free Ca2þ transient amplitude can

be �40% larger simply because of small (�10%) changes in resting [Ca2þ]. There-fore, if the basal Ca2þ concentration changes, the interpretation of changes in

transient amplitude has to be made with caution.

2. Intracellular [Ca2þ] levels that almost saturate the indicator cannot be used to

examine moderate changes in the amplitude of the Ca2þ transient. In the absence

of information concerning the maximal Ca2þ signal, it is diYcult to know how close

the dye is to saturation and therefore how sensitive the signal is to changes in peak

Ca2þ level.

Time (s)

Tim

e (s

)

0

3.0

1.5

300(mM)

(mM)

150

T1

T2

Buffercurve

0.5

0

0.5

1.0

1.0

F2

F1

Free[Ca2+]

[Ca2+]free

[Ca2+]total

Total [Ca2+]

Fig. 9 Illustration of the conversion of increments in total cellular Ca2þ to free Ca2þ signal using the

cellular buffer power. Increase of the total cellular Ca2þ by equivalent amounts (T1 & T2) causes a rise

of free [Ca2þ] of F1 and F2 due to differences in the background [Ca2þ]. Cellular Ca2þ buffer is

illustrated by the relationship between total cellular Ca2þ and free Ca2þ. Note that while the amplitude

of the transient increase in free [Ca2þ] depends on the increase in total Ca2þ and the cellular buffer

power, the time course of the decrease will depend on the extent of activation of cellular Ca2þ pumps

and exchangers. Generally, the rate of these processes depends on the free [Ca2þ], therefore the decay ofthe Ca2þ transients of different amplitudes may differ substantially.


XVI. Calibration of Single Wavelength Dyes

Calibration of single wavelength dyes is based on two main assumptions: firstly,

that the dye is at equilibrium with intracellular Ca2þ, that is, the kinetics of the

change of intracellular Ca2þ are slow compared to the rate constants for associa-

tion and dissociation. For the commonly used Fluo-type dyes at an intracellular

concentration of<100 mM, the half time for association and dissociation are of the

order of�1 ms and therefore for most circumstances the equilibration assumption

holds. The second assumption is that all the dye molecules sense the same amount

of light, that is, there is no filtering intrinsic to the biological preparation or by the

dye itself. Typically, the absorption coeYcient of Ca2þ indicators is of the order of

50,000 M�1 cm�1, therefore significant absorbance (>0.05) would occur in a cell

10 mm thick containing >100 mM dye. This consideration is important for single-

photon excitation and is one of the constraints that limit single-photon imaging to

thin (<50 mm) specimens.


With these two criteria satisfied, the Ca2þ concentration can be calculated from

the fluorescence from single wavelength dyes using the following equation:

Ca2þ� � ¼ Kd F � Fminð Þ= Fmax � Fð Þ

where F is the fluorescence signal from the cell/tissue; Kd the dissociation constant

of Ca2þ for the indicator (units M); Fmin the minimum fluorescence achieved when

the dye is essentially Ca2þ free, which practically can be approximated by exposing

the dye to a [Ca2þ], that is, 0.01Kd of the dye; and Fmax is the fluorescence achieved

when the dye is completely Ca2þ bound, which practically can be achieved with

[Ca2þ] of 100Kd of the dye. Measurements of these constants with some degree of

precision inside a cell are, however, diYcult. The dissociation constant can be

measured outside the cell in solutions approximating the intracellular mileau, but it

has been a frequent observation that the value of the Kd is altered by the intracel-

lular environment in a way that is diYcult to mimic by solution chemistry, for

example, mimicking intracellular viscosity and the range of negatively charged

intracellular proteins (Poenie, 1990). Therefore, the best practice is to measure the

dissociation constant within the cell type of interest. The easiest way to achieve this

is by using a glass microelectrode to gain access to the intracellular space. The use

of a series of solutions with a high concentration of Ca2þ buVer (EGTA or

BAPTA) with specific [Ca2þ] can be used to make a series of single cell measure-

ments to allow estimation of Kd. However, it is important to note that this

technique cannot be applied to multicellular preparations where multiple cells in

a tissue are diVerentially loaded with the dye.

XVII. Estimation of Fmax Values

This should be measured on a cell-to-cell basis even within multicellular prepara-

tions and involves exposing the inside of the cell to�50 mMor higher Ca2þ, depend-ing on the aYnity of the dye for Ca2þ. These levels are generally toxic to cells, but if

tolerated for a short time (1–2 s), this may be suYcient to estimate Fmax. These

intracellular [Ca2þ] levels can be achieved rapidly within single cells by perfusion

with a Ca2þ ionophore and raised extracellular Ca2þ (Loughrey et al., 2003).

A second ingenious method used in single voltage clamp experiments is to use an

amphotericin-containing patch pipette that facilitates monovalent cation exchange

across the membrane within the patch and therefore allows low resistance access to

the cell. At the endof the experiment, themembrane is ruptured under the patchusing

a rapid pressure step and the resultant influx of Ca2þ from the patch pipette generates

a rapid rise of intracellular [Ca2þ] that can be used to assess Fmax (Diaz et al., 2001).

A simpler but less reliablemethod is to simply use themicroelectrode to penetrate the

cell and allow extracellular Ca2þ influx in order to record Fmax, but generally Ca2þ

influx occurs in parallel with a rapid loss of intracellular dye so the signals would have

to be interpreted with caution.


Depending on the tissue, there are alternative approaches to estimate Fmax; in

some nerve cells, rapid frequent stimulation of cells can generate intracellular Ca2þ

levels that approach saturation of the dye (Maravall et al., 2000), thus allowing

Fmax values to be estimated.

XVIII. Estimation of Fmin or the Dynamic Range of the Dye

Estimation of Fmin or the dynamic range of the dye is more diYcult. The ratio of

Fmax/Fmin measured outside the cell cannot be assumed to apply inside; estimates

suggesting that values of �70–80% of those seen in free solutions are common

(Poenie, 1990). Again, an averaged value can be obtained using patch pipettes as a

means of establishing buVered [Ca2þ] inside cells. Alternatively, the Fmin in each

experiment can be estimated if the intracellular Ca2þ can be lowered to 1–10 nM

(for Fluo-3) by decreasing extracellular Ca2þ, but this assumes that intracellular

Ca2þ can be readily manipulated by changes in the extracellular environment

which is not always the case for every cell type. The eVect of under- or overestima-

tion of the dynamic range of the indicator is shown in Fig. 10C. Based on the

reported dynamic range of Fluo-based dyes (�100�), large under/overestimates of

the dynamic range (by up to 30%) cause only small errors in Ca2þ estimation and

only in the lower range of [Ca2þ] values relative to the Kd of the dye.

XIX. Consequence of Errors in Estimation of Intrinsic andDye Fluorescence

Prior to conversion of the indicator-based fluorescence to [Ca2þ], the back-

ground or intrinsic fluorescence of the cell/tissue has to be subtracted from the

signal. All cells have an intrinsic fluorescence mainly due to the metabolites beta

nicotinamide adenine dinucleotide (NADH) and flavin adenine dinucleotide

(FAD); their excitation wavelengths are 350–500 nm and emission wavelengths

�450–600 nm. The relative fluorescence of these two metabolites depends on the

metabolic state of the cell/tissue and degree of photobleaching. Thus, intrinsic

cellular fluorescence is significant and variable. The most advisable approach is to

use a dye with a significant basal fluorescence that is many times (>10�) that of the

intrinsic value. This cannot always be achieved; the Fluo-based and Rhodamine-

based dyes are by far the most popular dye groups used in confocal and 2P excitation

microscopy. Their main attraction is a large dynamic range as a result of a low

fluorescence signal from the Ca2þ free form. In this situation, Fmin values are fre-

quently comparable to that of the intrinsic fluorescence of the cell and therefore it is

important to quantify either by parallel measurements on nonloaded tissue or from a

single cell prior to the introduction of the dye. Error in estimation of background

fluorescence (which can be up to 100%) has dramatic eVects on the calculation of

Error in background estimationA B

Error in absolute fluorescence

Kd= 0.4mM Kd= 0.4mM

%E

rror

in [C

a2+] m

easu

rem

ent

%E

rror

in [C

a2+] m

easu

rem

ent

Kd= 0.8mM Kd= 0.8mM

100

−100%

−30%

−10%

+10%

+30%+100%

−50%

+50%

80

Rel

ativ

e flu

ores

cenc

e

60

40

20

0.1

[Ca2+]/Kd

1 10

40

0.1

[Ca2+]/Kd

1 100

100

80

Rel

ativ

e flu

ores

cenc

e

60

40

20

0

20

0

−20

40

20

0

−20

−40

C Error in dynamic range

Kd= 0.4mM

%E

rror

in [C

a2+] m

easu

rem

ent

Kd= 0.8mM

−30%

−10%

+10%

+30%

0.1 1 10

[Ca2+]/Kd

100 4

80

Rel

ativ

e flu

ores

cenc

e

60

40

20

0

2

−2

−4

0

Fig. 10 [Ca2þ] calibration errors. (A) Error due to changes in background fluorescence plotted

against varying [Ca2þ] values normalized by the indicator Kd (0.4 mM, Fluo-3 and 0.8 mM, Fluo5F).

Dynamic range of the dye was set at 100 (maximum attributable to Fluo-3). A typical cellular Ca2þ

concentration range is highlighted by gray boxes (100 nM to 1 mM) for each of the twoKd values. The left

axis (thick line) highlights the relative fluorescence versus [Ca2þ]. The right axis (thin lines) represents the%error in [Ca2þ] due tovariations in the intracellular background.Two levels of backgroundfluorescencewere considered: (i) background fluorescence is equal in magnitude to Fmin; errors of �50% and�100%

background were considered, and (ii) background fluorescence is equal to 5� Fmin; errors of �10% and

�20% of background fluorescence. These two combinations of background and errors superimpose

exactly (i.e., �50% superimposes on �10% and �100% superimposes on �20%) to produce the 4 error

lines shown. (B) Error in absolute fluorescence levels; errors of�10% and�30% are shown. (C) Error in

dynamic range of Fluo- and Rhod-based dyes; errors of�10% and �30% are shown.


intracellular [Ca2þ], particularly at either end of the sensitive range of the indicator, asshown in Fig. 10A. If the dye has a Ca2þ aYnity midway between the extremes of

intracellular Ca2þ, then the error can be small and approximately constant. But if the

dye has a lowerCa2þ aYnity, whichmay be desirable to resolve changes in peakCa2þ,then errors associated with the minimum cellular Ca2þ levels can be large.


These errors can be compounded by errors in estimates of fluorescence or the

variability of signals from one cell to the next. Fig. 10B shows the errors in Ca2þ

based on simple errors in fluorescence changes. The graph illustrates the risk

inherent in using dyes with a relatively high aYnity relative to the physiological

signal. Small errors in the range of fluorescence signals translate to large errors of

intracellular Ca2þ such that the ability to discriminate changes in maximum

physiological response is severely impaired. This can be significantly improved

by using lower aYnity dyes, but at the cost of poor resolution of minimum or

background intracellular [Ca2þ].

XX. Multimodal and Multiple Fluorophore Confocal andMultiphoton Microscopy

Although Ca2þ is an important signaling molecule in a variety of cell types, it by

no means operates alone. Rather, Ca2þ both temporally and spatially interacts

with many other properties and processes in the cell that only in concert orches-

trate cellular function. Thus, some of these processes are dictated by Ca2þ, butsome or not. A good example of this interplay is excitation–contraction coupling

in muscle cells, in which the action potential depolarizes the plasma membrane of

the cell, which causes a small influx of Ca2þ through the membrane. This inward

Ca2þ current stimulates the ryanodine receptor to release bulk Ca2þ from the

sarcoplasmic reticulum, which upon binding to the myofilaments induces the

actin–myosin interaction and the subsequent cellular contraction (Bers, 2002).

The cellular contraction may be imaged by simple black-and-white contrast

edge-detection microscopy, but this is not the case for the intracellular Ca2þ and

membrane potential characteristics that both require more sophisticated methods

such as fluorescence microscopy. Thus, simultaneous imaging with the use of

multiple fluorophores present at the same time in the specimen or combinations

of diVerent imaging modalities in some sense is required for capturing complex

information.

Thus, loading or injecting the specimen with multiple fluorophores allows for

simultaneous recording of diVerent signals, or if simultaneous recordings are not

technically possible, diVerent signals may be recorded sequentially without having

tomanipulate, move, or in any other way perturb the specimen between recordings.

In the latter case, only the optical pathways of the microscope would be altered

between recordings, whereas the specimen would not, since it would already be

loaded with diVerent fluorophores. The use of multiple fluorophores require either

the ability to direct separate emission wavelength bands onto diVerent light detec-tors, or to spectrally separate diVerent fluorophores by diVerent excitation wave-

lengths. Depending on the hardware, both confocal and multiphoton microscopes

can fulfill these requirements and therefore allow for measurements with multiple

fluorophores. Such experiments can be done by simultaneously loading the


specimen with a Ca2þ-sensitive fluorophore and a fluorophore that is sensitive for

another characteristic of the cell, which may also be Ca2þ in a diVerent compart-

ment of the cell with diVerent dynamics or a diVerent concentration range, or a

second fluorophore that may be sensitive to the plasma membrane voltage in an

excitable cell (also called a potentiometric dye). Measurements of Ca2þ and mem-

brane voltage (resting membrane potentials and action potentials) may then be

conducted either simultaneously by exciting both fluorophores at the same time and

capture spectrally diVerent fluorescence emission signals, or sequentially by exciting

each fluorophore separately, that is, one after the other, under otherwise similar

experimental conditions. The latter approach would assume that the experimental

conditions remain the same. Several factors may necessitate this, such as an inabili-

ty to diVerentiate between diVerent emission signals, or an inability to excite more

than one fluorophore at any given time, for example, if the excitation spectra do not

overlap and only one excitation wavelength may be delivered at one time.

The advantage of using multiple fluorophores either simultaneously or sequen-

tially is to increase the information content of the imaging, especially how diVerentprocesses relate to each other spatially and temporally. However, several issues

may limit the applicability of such measurements. Introducing a fluorophore to the

specimen may also change the dynamics of the cellular parameter of interest,

especially in live specimens that rely on stable and constant intra- and extracellular

environments. For instance, most Ca2þ indicators are also Ca2þ chelators that

buVer free Ca2þ, and most fluorophores or the medium they are delivered in may

change biochemical and biophysical properties of the intracellular environment.

This may be accentuated by simultaneous loading with several dyes. DiVerent dyesmay also quench, sequester, or in other ways inhibit each other. Finally, excitation

in itself may cause changes or damage to the specimen, and although this to some

degree is unavoidable, the degree of change or damage may be diVerent or evenaccentuated during sequential recordings.

The single-photon excitation and emission spectra of multiple Ca2þ- and

voltage-sensitive fluorescent dyes are well known. Clearly, some dyes have over-

lapping excitation or emission peaks, or present with broad excitation or emission

spectra such that even if the peaks are separated from one another, the tails of the

spectra still overlap considerably. Overlapping excitation spectra means that

diVerent dyes may be excited simultaneously, but overlapping emission spectra

may result in severely reduced signal specificity, and therefore, certain combina-

tions of fluorescent dyes may be less applicable, such as the potentiometric Di-4-

ANEPPS and Di-8-ANEPPS dyes, and the Ca2þ-sensitive Fluo-3 dye, all exten-

sively used by numerous laboratories for single-fluorophore purposes. All of these

dyes have single-photon excitation peaks at �480–500 nm and emission peaks

at 520–610 nm, respectively, with especially Di-4-ANEPPS and Di-8-ANEPPS

having very broad emission spectra that peak at �610 nm, but that considerably

overlap with the Fluo-3 emission spectrum, even though the latter has its peak

at �525 nm and therefore numerically diVers from Di-4-ANEPPS and

Di-8-ANEPPS and is more narrow (Fig. 11A). This problem may to some degree

800700

Wavelength (nm)

Rhod-2 and RH-237

600500

600

B

Em

issi

on in

tens

ity (

a.u.

)

500

400

300

200

100

0

800

A

800

700

700

Wavelength (nm)

600

600

500

500

Fluo-3 and Di-4-ANEPPS

400

300

200

100Em

issi

on in

tens

ity (

a.u.

)

0

C

800700Wavelength (nm)

600500

Fluo-3 and RH-237

Em

issi

on in

tens

ity (

a.u.

)

700

600

500

400

300

200

100

0

Fig. 11 Single-photon emission spectra of voltage-sensitive Di-4-ANEPPS and RH-237, and Ca2þ-sensitive Fluo-3 and Rhod-2 dyes. The following emission spectra were obtained by spectrophotometry

after simultaneously loading cardiac muscle cells in a high Ca2þ solution (60 mM) with Ca2þ- andvoltage-sensitive dyes and exciting at 488 nm. The presence of intact excitable cells in a Ca2þ-richenvironment provides substrate for both Ca2þ- and voltage-sensitive dyes. (A) Fluo-3 (Ca2þ; first peak)and Di-4-ANEPPS (voltage; second peak). (B) Rhod-2 (Ca2þ; first peak) and RH-237 (voltage; second

peak). (C) Fluo-3 (Ca2þ; first peak) and RH-237 (voltage; second peak).


be avoided as the Ca2þ and voltage signals are spatially separated between intra-

cellular and membrane compartments of the cell, though in muscle cells this may

turn out to be diYcult because of the dense network of plasma membrane trans-

verse tubules that penetrate into the interior of the cell. The same overlap problem

exists with the voltage-sensitive RH-237 and the Ca2þ-sensitive Rhod-2 dyes, with

emission peaks occurring at �580 and �660 nm, respectively, but with especially

the RH-237 emission spectrum being very broad (Fig. 11B). In contrast, Fluo-3

and RH-237 are more distinctly separated from one another. Both Fluo-3 and

RH-237 dyes may be excited by the same single-photon excitation wavelength at


�500 nm (though RH-237 would not be optimally excited by this wavelength), but

have emission spectra that may be spectrally diVerentiated, as Fluo-3 has a narrowemission spectrum that peaks at �525 nm, whereas RH-237 peaks at �660 nm

with a broad spectrum (Fig. 11C). Thus, this combination appears attractive as it

diVerentiates the Ca2þ and membrane potential signals, for which it has also been

used successfully (Fast and Ideker, 2000).

However, several problems arise when transferring from single-photon to 2P

excitation, although several of the voltage-sensitive dyes present with consistent

and reproducible 2P excitation spectra that very much resemble the double single-

photon excitation spectra and may thus be confidently used for meaningful multi-

photon imaging. In contrast, the 2P behavior of many of the Ca2þ-sensitive dyes issomewhat diYcult to interpret (see previous discussion and Fig. 8), although the

ratiometric Fura dyes may be 2P excited in order to provide a meaningful Ca2þ

signal that also captures transient changes over a millisecond scale with high

fidelity (Wokosin et al., 2004). This may be because the Fura dyes have a single-

photon excitation spectrum in the UV range (340–380 nm), and therefore the 2P

excitation spectrum, which approximately is double the single-photon spectrum,

occurs at �800 nm wavelengths, in which the 2P laser power outputs are not

limited. Moreover, this study indicated that several of the Fura dyes, in particular

Fura-4F, may work well when excited with a single IR wavelength, despite their

use as ratiometric single-photon dyes, as judged by the dynamic ranges and SNR

obtained during 2P excitation microscopy in single cardiac muscle cells during

diVerent Ca2þ conditions. In contrast, the Fluo- and Rhod-based Ca2þ-sensitivedyes, all with single-photon excitation peaks at �500–550 nm, present with 2P

excitation spectra that are not immediately predicted by the doubled single-photon

spectra (Fig. 8). In these cases, the 2P excitation spectra are at least partly broken

up and appear blueshifted compared to the doubled single-photon spectra.

Although doubling the single-photon excitation spectrum is often a good predictor

for the 2P excitation spectrum, deviations from this do occur, although these

deviations may neither be systematic nor well understood (Xu et al., 1996; Zipfel

et al., 2003). Alongside this, a reoccurring problem is that the available Ti:Sapphire

pulsed 2P lasers are power-limited at the long wavelengths of 1000–1100 nm that

would correspond to the doubled single-photon excitation spectra of Fluo-3 and

Rhod-2. Because not all fluorophores are easily transferable from single-photon

excitation, for which they were developed, to 2P excitation, this therefore has made

it problematic to use multiple fluorophores simultaneously during 2P excitation

microscopy, and the issue has not yet been fully resolved.

A diVerent approach to capture more complex information has been to combine

several multimodal microscopy techniques inways that also encompass confocal and

multiphoton systems, but also this comes with both advantages and disadvantages.

For instance, diVerentmodes of contrast used on the same specimenmay increase the

information extracted from the images and reduce artifacts. Multimodal microscopy

may also allow for a wider repertoire of fluorophores. However, if confocal and

multiphoton imaging are combined, it requires descanning and insertion of a


confocal aperture with a pinhole into the light pathway, which may reduce the

fluorescence capture after 2P excitation and thus lead to a loss of signal, though a

confocal aperturemay also be set up to increase the spatial resolution ofmultiphoton

images, by restricting the PSF tails. Because of these limitations, the reality is often

that it is diYcult, though not impossible, to achieve optimal performance from each

individual mode when several modes are combined. Nonetheless, the advantages of

simultaneous or near-simultaneous light capture by diVerent modes of microscopy

may, under the right circumstances, far outweigh the disadvantages.

Examples include combinations of confocal or multiphoton with epifluorescence

or diVerential interference contrast microscopes to capture light emission restricted

to the focal plane as well as capturing a widefield view, either simultaneously or

sequentially without having to reorient or replace the specimen. Other options also

include setting up a microscope system that combines confocal and 2P excitation

imaging modes, or 2P excitation and second-harmonic generation (SHG) imaging.

Although these applications tend to serve narrow and specific purposes, they may

allow for imaging of local versus global Ca2þ signaling, or Ca2þ signaling in

combination with for example, metabolic parameters by using 2P excitation to

excite metabolites such as NADH and FAD, or collagen that in particular con-

tributes to the SHG signal (Masters, 2006). A final example of multimodal micros-

copy techniques that may successfully be combined includes the combination of

FRET and FLIM imaging to quantify FRET between two fluorophores, as in the

case of the Ca2þ-sensitive cameleon described above. These examples are not

exhaustive, but serve to illustrate the potential of combining diVerent fluorophoresor microscopy modalities in order to gain information of a more detailed nature.

References

Bers, D. M. (2002). Cardiac excitation–contraction coupling. Nature 90, 182–189.

Bestvater, F., Spiess, E., Stobrawa, G., Hacker, M., Feurer, T., Porwol, T., Berchner-Pfannschmidt, U.,

Wotzlaw, C., and Acker, H. (2002). Two-photon fluorescence absorption and emission spectra of

dyes relevant for cell imaging. J. Microsc. 208, 108–115.

Bliton, C., Lechleiter, J., and Clapham, D. E. (1993). Optical modifications enabling simultaneous

confocal imaging with dyes excited by ultra-violet and visible wavelength light. J. Microsc. 169,

15–26.

Bootman, M. D., Collins, T. J., Peppiatt, C. M., Prothero, L. S., MacKenzie, L., De Smet, P.,

Travers, M., Tovey, S. C., Seo, J. T., Berridge, M. J., Ciccolini, F., and Lipp, P. (2001). Calcium

signalling—An overview. Semin. Cell Dev. Biol. 12, 3–10.

Bullen, A. (2008). Microscopic imaging techniques for drug discovery. Nat. Rev. Drug Discov. 7, 54–67.

Callamaras, N., and Parker, I. (1999). Construction of a confocal microscope for real-time x–y and x–z

imaging. Cell Calcium 26, 271–279.

Cleemann, L., Di Massa, G., and Morad, M. (1997). Ca2þ sparks within 200 nm of the sarcolemma of

rat ventricular cells: Evidence from total internal reflection fluorescence microscopy. Adv. Exp. Med.

Biol. 430, 57–65.

Conchello, J. A., and Lichtman, J. W. (2005). Optical sectioning microscopy. Nat. Methods 2, 920–931.

Conchello, J. A., Kim, J. J., and Hansen, E. W. (1994). Enhanced 3-D reconstruction from confocal

scanning microscope images. II: Depth discrimination vs. signal-to-noise ratio in partially confocal

images. Appl. Opt. 33, 3740–3750.


Cox, G., and Sheppard, C. J. R. (2004). Practical limits of resolution in confocal and non-linear

microscopy. Microsc. Res. Tech. 63, 18–22.

Denk, W., Strickler, J., and Webb, W. (1990). Two-photon laser scanning fluorescence microscopy.

Science 248, 73–76.

Diaz, M. E., TraVord, A. W., and Eisner, D. A. (2001). The eVects of exogenous calcium buVers on the

systolic calcium transient in rat ventricular myocytes. Biophys. J. 80, 1915–1925.

Diels, J. C., Fontaine, J. J., McMichael, I. C., and Simoni, F. (1985). Control and measurement of

ultrashort pulse shapes (in amplitude and phase) with femtosecond accuracy. Appl. Opt. 24,

1270–1282.

Fast, V., and Ideker, R. (2000). Simultaneous optical mapping of transmembrane potential and

intracellular calcium in myocyte cultures. J. Cardiovasc. Electrophysiol. 11, 547–556.

Hanley, Q. S., Verveer, P. J., Gemkow, M. J., Arndt-Jovin, D., and Jovin, T. M. (1999). An optical

sectioning programmable array microscope implemented with a digital micromirror device.

J. Microsc. 196, 317–331.

Heintzmann, R., Sarafis, V., Munroe, P., Nailon, J., Hanley, Q. S., and Jovin, T. M. (2003). Resolution

enhancement by subtraction of confocal signals taken at diVerent pinhole sizes. Micron 34, 293–300.

Helmchen, F., and Denk, W. (2005). Deep tissue two-photon microscopy. Nat. Methods 2, 932–940.

Jares-Erijman, E. A., and Jovin, T. M. (2003). FRET imaging. Nat. Biotechnol. 21, 1387–1395.

Kettlewell, S., Cabrero, P., Nicklin, S. A., Dow, J. A. T., Davies, S., and Smith, G. L. (2009). Changes of

intra-mitochondrial Ca2þ in adult ventricular cardiomyocytes examined using a novel fluorescent

Ca2þ indicator targeted to mitochondria. J. Mol. Cell. Cardiol. 46, 891–901.

Kim, H. M., Kim, B. R., An, M. J., Hong, J. H., Lee, K. J., and Cho, B. R. (2008). Two-photon

fluorescent probes for long-term imaging of calcium waves in live tissue. Chemistry 14, 2075–2083.

Levitt, J. A., Matthews, D. R., Ameer-Beg, S. M., and Suhling, K. (2009). Fluorescence lifetime and

polarization-resolved imaging in cell biology. Curr. Opin. Biotechnol. 20, 28–36.

Lichtman, J. W., and Conchello, J. A. (2005). Fluorescence microscopy. Nat. Methods 2, 910–919.

Loughrey, C. M., MacEachern, K. E., Cooper, J., and Smith, G. L. (2003). Measurement of the

dissociation constant of Fluo-3 for Ca2þ in isolated rabbit cardiomyocytes using Ca2þ wave char-

acteristics. Cell Calcium 34, 1–9.

Maravall, M., Mainen, Z. F., Sabatini, B. L., and Svoboda, K. (2000). Estimating intracellular calcium

concentrations and buVering without wavelength ratioing. Biophys. J. 78, 2655–2667.

Mashanov, G. I., Tacon, D., Knight, A. E., Peckham, M., and Molloy, J. E. (2003). Visualizing single

molecules inside living cells using total internal reflection fluorescencemicroscopy.Methods 29, 142–152.

Masters, B. R. (2006). ‘‘Confocal Microscopy and Multiphoton Excitation Microscopy. The Genesis of

Live Cell Imaging.’’ The International Society for Optical Engineering Press, Bellingham, WA.



882–887.

Miyawaki, A., Mizuno, H., Nagai, T., and Sawano, A. (2003). Development of genetically encoded

fluorescent indicators for calcium. Methods Enzymol. 360, 202–225.



Ogden, D. (ed.) (1994). Microelectrode Techniques. ‘‘The Plymouth Workshop Handbook.’’ 2nd edn.

The Company of Biologists Ltd, Cambridge, UK.

Ogden, D., Khodakhah, K., Carter, T., Thomas, M., and Capiod, T. (1995). Analogue computation of

transient changes of intracellular free Ca2þ concentration with the low aYnity Ca2þ indicator

furaptra during whole-cell patch-clamp recording. Pflugers Arch. 429, 587–591.

Poenie, M. (1990). Alteration of intracellular Fura-2 fluorescence by viscosity: A simple correction. Cell

Calcium 11, 85–91.

Sanders, R., Draaijer, A., Gerritsen, H. C., Houpt, P. M., and Levine, Y. K. (1995). Quantitative pH

imaging in cells using confocal fluorescence lifetime imaging microscopy. Anal. Biochem. 227,

302–308.


Tanaami, T., Otsuki, S., Tomosada, N., Kosugi, Y., Shimizu, M., and Ishida, H. (2002). High-speed

1-frame/ms scanning confocal microscope with a microlens and Nipkow disk. Appl. Opt. 41,

4704–4708.

Thomas, D., Tovey, S., Collins, T. J., Bootman, M. D., Berridge, M. J., and Lipp, P. (2000).

A comparison of fluorescent Ca2þ indicator properties and their use in measuring elementary and

global Ca2þ signals. Cell Calcium 28, 213–223.

Tsien, R. Y. (1980). New calcium indicators and buVers with high selectivity against magnesium and

protons: Design, synthesis, and properties of prototype structures. Biochemistry 19, 2396–2404.

Tsien, R. Y., and Bacskai, B. J. (1995). Video-rate confocal microscopy. In ‘‘Handbook of Biological

Confocal Microscopy.’’ (J. B. Pawley, ed.), Plenum Press, New York.

Wang, E., Babbey, C. M., and Dunn, K. W. (2005). Performance comparison between the high-speed

Yokogawa spinning disc confocal system and single-point scanning confocal systems. J. Microsc. 218,

148–159.

Webb, R. H. (1999). Theoretical basis of confocal microscopy. Methods Enzymol. 307, 3–20.

Willig, K. I., Rizzoli, S. O., Westphal, V., Jahn, R., and Hell, S. W. (2006). STED microscopy reveals

that synaptotagmin remains clustered after synaptic vesicle exocytosis. Nature 440, 935–939.

Wilson, T., Juskaitis, R., Neil, M. A., and Kozubek, M. (1996). Confocal microscopy by aperture

correlation. Opt. Lett. 21, 1879–1881.

Wokosin, D. L., Loughrey, C. M., and Smith, G. L. (2004). Characterization of a range of Fura dyes

with two-photon excitation. Biophys. J. 86, 1726–1738.

Xu, C., Zipfel, W. R., Shear, J. B., Williams, R. M., andWebb,W.W. (1996). Multiphoton fluorescence

excitation: New spectral windows for biological nonlinear microscopy. Proc. Natl. Acad. Sci. USA 93,

10763–10768.

Zipfel, W. R., Williams, R. M., andWebb, W. W. (2003). Nonlinear magic: Multiphoton microscopy in

the biosciences. Nat. Biotechnol. 21, 1369–1377.

CHAPTER 10


The Use of Aequorins to Record andVisualize Ca2þ Dynamics: From SubcellularMicrodomains to Whole Organisms

Sarah E. Webb,* Kelly L. Rogers,† Eric Karplus,‡ andAndrew L. Miller**Biochemistry and Cell Biology Section and State Key Laboratory of Molecular NeuroscienceDivision of Life ScienceHKUST, Clear Water BayKowloon, Hong Kong, PR China

†The Walter and Eliza Hall Institute of Medical ResearchParkville, Australia

‡Science Wares Inc.FalmouthMassachusetts, USA

A

OGY,All rig

bstract


-679X-679X

I. In
troduction II. E xpression of Apoaequorin, GFP-Apoaequorin, and Other Apoaequorin-Based
Spectral Variants in Cells, Tissues, and Whole Organisms
A. Expression of Apoaequorin B. Expression of GFP-Apoaequorin C. Summary of Section II
III. In
troducing Coelenterazines into Cells, Tissues and Embryos IV. T echniques for Detecting Aequorin Luminescence V. C onclusions
R
eferences
Abstract

In this chapter, we describe the practical aspects of measuring [Ca2þ]transients that are generated in a particular cytoplasmic domain, or within a

specific organelle or its periorganellar environment, using bioluminescent,

/10 $35.00(10)99010-9

264 Sarah E. Webb et al.

genetically encoded and targeted Ca2þ reporters, especially those based on

apoaequorin. We also list examples of the organisms, tissues, and cells that

have been transfected with apoaequorin or an apoaequorin-BRET complex, as

well as of the organelles and subcellular domains that have been specifically

targeted with these bioluminescent Ca2þ reporters. In addition, we summarize

the various techniques used to load the apoaequorin cofactor, coelenterazine,

and its analogs into cells, tissues, and intact organisms, and we describe recent

advances in the detection and imaging technologies that are currently being

used to measure and visualize the luminescence generated by the aequorin-

Ca2þ reaction within these various cytoplasmic domains and subcellular

compartments.

I. Introduction

One of the most significant recent developments in the Ca2þ signaling field has

been the general acceptance of the wide-spread heterogeneity of Ca2þ activity

within individual cells; not only at rest, but also most importantly, during stimula-

tion (Berridge, 2009; Rizzuto and Pozzan, 2006; Rutter et al., 2006; Whitaker,

2008). This has led to the concept of dynamic subcellular ‘‘Ca2þ microdomains.’’

As suggested by Rizzuto and Pozzan (2006), this term (especially with regard to its

spatial dimensions) has several diVerent meanings depending on one’s area of

interest. In this chapter, however, like Rizzuto and Pozzan, we use the term in a

general way to describe Ca2þ dynamics that do not involve the entire cell cyto-

plasm, but that remain localized to a specific cytoplasmic domain, or occur within

a particular organelle or its periorganellar environment. Thus, one of the current

challenges researchers are facing in the field of Ca2þ imaging is that of resolving

changes in [Ca2þ] within, and between, various subcellular microdomains.

An eVective strategy to address this challenge that is common to both fluorescence-

and luminescence-based imaging techniques is to exclusively visualize Ca2þ

dynamics in specific microdomains using genetically encoded and targeted Ca2þ

reporters (GET-CRs). These come in two general forms, fluorescent GET-CRs

and bioluminescent GET-CRs, respectively. At the other end of the size spectrum

is the exciting prospect of imaging Ca2þ signals derived from GET-CRs within

freely moving, large, organisms, for example, adult mice (Rogers et al., 2007).

This presents a diVerent set of technical challenges to researchers in the Ca2þ

imaging field.

Fluorescent GET-CRs include the camgaroos (Baird et al., 1999; Griesbeck

et al., 2001), G-CaMPs (Nakai et al., 2001; Ohkura et al., 2005), pericams (Nagai

et al., 2001), case-sensors (Souslova et al., 2007), grafted EF-hands (Zou et al.,

2007), and cameleon-types (Miyawaki et al., 1997; Ishii et al., 2006; Tsuruwaka

et al., 2007; and reviewed by Zorov et al., 2004; McCombs and Palmer, 2008).

Bioluminescent GET-CRs include single protein entities such as aequorin

10. The Use of Aequorins to Record and Visualize Ca2þ Dynamics 265

(Cheung et al., 2006; Torrecilla et al., 2000), obelin (Stepanyuk et al., 2005),

mitrocomin (Inouye and Sahara, 2009), clytin (Inouye, 2008), and photina

(Bovolenta et al., 2007). In addition, there is a growing number of aequorin-

derived bioluminescence resonance energy transfer (BRET)-based complexes

such as the GFP-aequorins (Ashworth and Brennan, 2005; Baubet et al., 2000;

Martin et al., 2007 ; Rogers et al., 2005, 2007), as well as other wavelength-shifted

variants (Gorokhovatsky et al., 2004). To date, apoaequorin alone, or apoae-

quorin in tandem with another BRET protein, has been genetically expressed in a

diverse range of diVerent species; either in the whole organism or in specific

tissues within an intact organism (See Table I and Fig. 1). For example, apoae-

quorin has been ubiquitously expressed in whole zebrafish embryos (Cheung

et al., 2006) or specifically targeted to the Malpighian tubules in Drosophila

(Rosay et al., 1997), while the BRET complexes GFP-apoaequorin and YFP-

apoaequorin have been specifically targeted to neuronal cell subsets of Drosophi-

la (Martin et al., 2007), and the endodermis and pericycle of Arabidopsis roots

(Kiegle et al., 2000), respectively. Furthermore, apoaequorin or apoaequorin-

BRET complexes have been expressed either ubiquitously in the cytosol of cells in

culture or, using specific targeting sequences, in distinct organelles of cells in

culture (see Table II, and a recent review by Gerasimenko and Tepikin, 2005).

Specific organelles targeted include: the ER (Montero et al., 1997), mitochondria

(Rizzuto et al., 1992), the Golgi apparatus (Pinton et al., 1998), the nucleus (Brini

et al., 1993, 1994), gap junctions (George et al., 1998), subplasma membrane

domains (Marsault et al., 1997; Nakahashi et al., 1997), secretory vesicles

(Mitchell et al., 2001), and the outer mantle of secretory granules (Pouli et al.,

1998), to name but a few examples.

In this chapter, we focus on describing the practical uses of bioluminescent GET-

CRs, especially those based on apoaequorin. In addition, we provide the reader (in

table form) with a review of the literature to date listing representative examples of

whole organisms, tissues, and cells that have been transfected with apoaequorin or

an apoaequorin-BRET complex, as well as a list of organelles and subcellular

domains that have been specifically targeted (see Tables I and II). We also summa-

rize diVerent strategies used for loading various derivatives of the apoaequorin

cofactor, coelenterazine (Shimomura et al., 1989) into cells, tissues, and intact

organisms (summarized in Table III). Furthermore, we describe recent advances in

detection and imaging technologies used to measure and visualize light generated by

the aequorin-Ca2þ luminescent reaction within cells, tissues, and intact organisms

(summarized in Table IV and illustrated in Figs. 2 and 3). Our hope is that this

chapter will provide a starting point for researchers wishing to use GET-CRs to

measure or visualize Ca2þ dynamics from cells, tissues, or intact organisms. Fur-

thermore, the references provided in Tables I–IV should lead them to more detailed

information regarding a biological system and/or experimental setup that will com-

plement their own research interest. For loading holoaequorin into cells and embry-

os, we refer readers to the practical methodologies described inMiller et al. (1994), as

Table IExamples where apoaequorin or apoaequorin-BRET complexes have been targeted to a diverse range of diVerent species

Kingdom Class Species Apoaequorin or apoaequorin-BRET targeted to: References

Animalia Mammals Mus musculus (mouse) Whole organism Yamano et al. (2007)

Mitochondrial matrixa Rogers et al. (2007)

Fish Danio rerio (zebrafish) Whole organismb Cheung et al. (2006)

Trunk musculature Cheung (2009)

Amphibians Xenopus laevis (African clawed frog) Plasma membrane of oocytes Daguzan et al. (1995)

Insects Drosophila melanogaster (Fruit fly) Malpighian tubules—diVerent

cellular components

Rosay et al. (1997)

Mushroom bodies and antennal lobesa Martin et al. (2007)

Plantae Dicots Nicotiana plumbaginifolia (Tobacco) Whole organism Knight et al. (1991a)

Arabidopsis thaliana Whole organism Knight et al. (1995, 1996),

Sedbrook et al. (1996)

Guard cells Dodd et al. (2006)

Endodermis and pericycle of rootsc Kiegle et al. (2000)

Solanum tuberosum (Potato) Whole organism Fisahn et al. (2004)

Monocots Triticum aestivum (Winter wheat) Whole organism Nagel-Volkmann et al. (2009)

Moss Physcomitrella patens Whole organism Russell et al. (1996)

Fungi Funguses Phyllosticta ampelicida Whole organism Shaw et al. (2001)

Neurospora cassa Whole organism Nelson et al. (2004)

Aspergillus awamori Whole organism Nelson et al. (2004)

Aspergillus niger Whole organism Nelson et al. (2004),

Bencina et al. (2005)

Yeast Saccharomyces cerevisiae Whole organism Batiza et al. (1996)

Schizosaccharomyces pombe Whole organism Deng et al. (2006)

Protista Amoebozoa Dictyostelium discoideum (Slime mold) Whole organism Cubitt et al. (1995)

Diatoms Phaeodactylum tricornutum Whole organism Falciatore et al. (2000)

Monera Bacteria Escherichia coli Whole organism Knight et al. (1991b)

Bacillus subtilis Whole organism Herbaud et al. (1998)

Streptococcus pneumoniae Whole organism Chapuy-Regaud et al. (2001)

Blue-green algae Anabaena strain sp. PCC7120 Whole organism Torrecilla et al. (2000)

aGFP-aequorin constructs were used.bTransient transfection of apoaequorin mRNA was used.cYFP-aequorin constructs were used.

0 s 2 s 4 s 6 s

8 s 10 s 12 s 14 s

16 sPho

tons

/pix

el

18 s 20 s 22 s

0.20

0.16

0.12

0.08

0.04

0

A

1–2 3–4 5–6 7–8 >8

0 min

Photons/pixel

3 min 6 min 9 minB

0 s 15 s 12

1

30 sOK107

Pho

tons

/pix

el

C

0 16 32 64 128 225Photons

100 ppb 300 ppb 500 ppb 750 ppbAir

D

Fig. 1 Examples of the spatial patterns of Ca2þ signals generated by whole organisms, or by specific

tissues or subcellular organelles within intact organisms, where aequorin or GFP-aequorin were

genetically expressed. (A) Newborn mice stably expressing GFP-aequorin targeted to the mitochondrial

matrix were injected intraperitoneally with native coelenterazine and bioluminescence activity was

recorded with the animals un-restrained and freely moving. These are consecutive video images on to

which have been superimposed the corresponding bioluminescence images. Each panel represents 2 s of

accumulated light. Scale bar is 5 mm. Reproduced with permission, fromRogers et al. (2007). (B) An 18-

somite stage (i.e., �18 h postfertilization (hpf)) zebrafish embryo that was injected with apoaequorin

mRNA at the one-cell stage to transiently express apoaequorin throughout the whole embryo, and then



not much has changed with respect to these particular techniques since Volume 40 of

the Methods in Cell Biology series was published.

II. Expression of Apoaequorin, GFP-Apoaequorin, and OtherApoaequorin-Based Spectral Variants in Cells, Tissues, andWhole Organisms

Microinjected holoaequorin has been used since the late 1960s for monitoring

changes in [Ca2þ]i in diVerent cells and tissues. The earliest reports describe the use

of holoaequorin to detect Ca2þ transients in muscle and nerve cells (Baker et al.,

1971; Ridgway and Ashley, 1967) as well as during activation in medaka eggs

(Ridgway et al., 1977). This approach is only practical, however, for introducing

aequorin into giant cells and large embryos, which are easy to microinject. The

more recent development, from the mid 1980s to early 1990s, of genetic engineering

techniques to introduce and express apoaequorin (the protein moiety of aequorin)

cDNA in cells, tissues, and whole organisms (Inouye et al., 1989; Knight et al.,

1991a,b; Nakajima-Shimada et al., 1991; Prasher et al., 1985; Saran et al., 1994), as

well as to target apoaequorin to specific organelles within cells (Brini et al., 1993;

Rizzuto et al., 1992), has paved the way for aequorin to be used as the Ca2þ

reporter of choice in many more biological systems today, from cells in culture to

complex multicellular organisms.

GFP-aequorin was developed approximately 10 years ago in order to improve

the stability and light emission properties of aequorin for single-cell imaging

(Baubet et al., 2000). Based on the naturally occurring phenomenon of BRET,

GFP-aequorin emits a red-shifted light emission (l¼509 nm) relative to that of

aequorin alone (l¼470 nm) in the presence of elevated free Ca2þ ion concentra-

tions. GFP-aequorin has a number of advantages over aequorin for monitoring

changes in cellular Ca2þ concentrations, including increased stability and total

light output. Furthermore, the expression level and distribution of the GFP reflects

the expression level and distribution of apoaequorin; thus, the expression of

apoaequorin can be directly visualized in living cells or tissues. Although the

incubated with f-coelenterazine starting at the 64-cell stage to reconstitute aequorin. Each panel

represents 120 s of accumulated light and consecutive panels are stepped at 60-s intervals. Scale bar is

200 mm. (C) D. melanogaster (P[GAL4] OK107 line) stably expressing GFP-aequorin in the mushroom

bodies. Exposed fly brains were incubated for >1 h at room temperature with native coelenterazine,

prior to imaging. The first panel shows the whole brain and the localization of GFP in the mushroom

bodies. The following panels show consecutive bioluminescent images, each panel representing 15 s of

accumulated light, following treatment with 70 mM KCl to induce Kþ-depolarization. Scale bar is

100 mm. Reproduced with permission, from Martin et al. (2007). (D) Nine-day old seedlings of

Arabidopsis thaliana (ecotype RLD1) that constitutively express apoaequorin were incubated in the

dark overnight in coelenterazine solution. These images show the total Ca2þ-dependent biolumines-

cence recorded from seedlings exposed to air or to diVerent concentrations of ozone for 1 h. Scale bar is

5 mm. #John Wiley and Sons Ltd. Reproduced with permission, from Evans et al. (2005).

Table IIExamples where apoaequorin or apoaequorin-BRET complexes have been targeted to the cytosol of cells in culture or todistinct intracellular organelles either in cells in culture or intact organisms

Species (common

name

and/or scientific

name) Cell type

Cytosolic expression of

apoaequorin or

apoaequorin-BRET Organelle(s) targeted Comments References

Kingdom: Animalia

Human (Homo

sapiens)

ECV304 (umbilical

vein endothelial cells)

✓ Mitochondria Transient expression Lawrie et al. (1996)

Diploid fibroblasts � (used indo 1-AM

and fura 2-AM

Mitochondria 1� cell cultures(Transient

expression)

Padua et al. (1998)

HEK-293 cells (embry-

onic kidney cell line)

✓ – Some transient

and some stable

expression

Sheu et al. (1993)

✓ – Stable expression Button and

Brownstein (1993)

✓ Nucleus HSV gene

transfer vector

Chamero et al. (2002)

✓ ER lumen,

mitochondrial

matrix

Transient expression Brini et al. (2005)

HEK 293T cells ✓ Nucleus,

mitochondria, ER

Expression of GA

and RA via HSV

gene transfer

vector

Manjarres et al. (2008)

HeLa (immortalized

epithelial cell line)

✓ Nucleus Stable expression Brini et al. (1993)

� (used Fura-2) ER lumen Transient and stable

expression

Montero et al. (1995)

� (used fura 2) ER 1� cell cultures(HSV gene

transfer vector)

Alonso et al. (1998)

✓ Mitochondria,

Golgi, ER

Transient expression Pinton et al. (2000)

� (used indo-1) Golgi, ER Stable expression Missiaen et al. (2004)

Normal adult and

Hailey-Hailey disease

keratinocytes

� (used proton-induced

X-ray emission)

Golgi 1� cell cultures(Transient

expression)

Behne et al. (2003)

Jurkat cells (immorta-

lized T-lymphocytes)

� (used indo PE3 (AM)) ER Transient expression Narayanan et al. (2003)

(continues)

Table II (continued )

Species (common

name

and/or scientific

name) Cell type


apoaequorin or


African green

monkey

(Simia

aethiops)

Kidney COS cell

cultures

✓ – Transient expression Button and Brownstein

(1993)

COS7 cells ✓ ER Transient expression Kendall et al. (1994)

✓ Mitochondriaþouter

mitochondrial

membrane

Transient expression Brandenburger et al.

(1999)

Cow (Bos taurus) Bovine adrenal

medulla

chromaYn cells

� (used fura 2) ER 1� cell cultures (HSV

gene transfer

vector)


� (used fura 2) Mitochondria 1� cell cultures (HSV

gene transfer

vector)

Montero et al. (2000)

✓ (also Fura 4F) ER, nucleus

mitochondria,

1� cell cultures(HSV gene trans-

fer vector)

Villalobos et al. (2002)

✓ Nucleus 1� cell cultures (HSV

gene transfer

vector)


� Secretory granules 1� cell cultures (AdV

gene transfer

vector)

SantoDomingo et al.

(2008)

Bovine adrenal zone

glomerulosa cells

✓ Mitochondrial matrix 1� cell cultures(Transient

transfection)

Brandenburger et al.

(1996)

✓ Mitochondriaþouter mi-

tochondrial membrane

1� cell cultures(Transient

transfection)

Brandenburger et al.

(1999)

Mouse

(Mus

musculus)

Muller glial cells of adult

retinal explants

✓ – Explant culture. Ex-

pression of GA via

AdV gene transfer

vector

Agulhon et al. (2007)

A-11 (nonmetastatic)

and 3LL (metastatic)

Lewis lung cancer cell

lines

✓ – Stable expression Yoshida et al. (1998)

Soma and neurites from

adult superior cervical

ganglion neurons

� (used fura 2) Mitochondria 1� cell cultures (HSV

gene transfer

vector)

Nunez et al. (2007)

Pancreatic b-cells fromintact islets

� (used fura 2) Mitochondria Explant culture. Ex-

pression of GA via

HSV gene transfer

vector

Quesada et al. (2008)

Myoblasts isolated from

the extensor digi-

torum longus muscle

in mdx and normal

C57B1/10 mice

✓ Sub-sarcolemma 1� cell cultures(Transient

expression)

Basset et al. (2004)

Myotubes isolated from

hind leg muscles of

mdx and C57BL101

mice

✓ SR, mitochondriaþ -

plasma membrane


expression)

Robert et al. (2001a)

C2C12 skeletal muscle

cell line

� (used fura 2 AM) Mitochondria Stable expression Challet et al. (2001)

NIH 3T3 fibroblasts � (used Fura-2) ER HSV gene transfer

vector


✓ Outer surface of intracel-

lular membranes

Transient expression Biagioli et al. (2005)

MIN6 (pancreatic b-cellline)

� (used fura 2 AM) Mitochondria Stable expression Nakazaki et al. (1998)

� (used fura 2 AM) ER, secretary vesicle Transient expression

and AdV gene

transfer vector

Mitchell et al. (2001)

Intact pancreatic islets of

Langerhans from

Balb/c mice

✓ Nuclear Explant culture

(HSV gene trans-

fer vector)

Villalobos et al. (2005)

Neuro2A (neuroblasto-

ma cells)

✓ – Transient expression

of GA

Baubet et al. (2000)

Rat

(Rattus

norvegicus)

Cerebellar granule cells � (used fura 2) ER 1� cell cultures(HSV gene trans-

fer vector)


Skeletal muscle

myotubes

✓ (also used fura-2) Mitochondria, nucleus,

SR

1�cell cultures

(Transient

expression)

Brini et al. (1997)

� SR and ER 1� cell cultures(Transient

expression)

Robert et al. (1998)

✓ ER lumen and lumen of

the terminal cisternae of

the SR


expression)

Brini et al. (2005)

(continues)

Table II (continued )

Species (common

name

and/or scientific

name) Cell type


apoaequorin or


L6 myogenic cell line � (used indo PE3 (AM)) ER Transient expression Narayanan et al. (2003)

Ventricular myocytes

from neaonatal

Wistar rats

✓ Mitochondria 1� cell cultures(Transient

expression)

Robert et al. (2001b)

A7r5 cells (aortic

smooth muscle cell

line)

✓ Plasma membrane Transient expression Marsault et al. (1997)

Aortic

smooth muscle cells

� (used fura 2 AM) Mitochondria 1� cell cultures(Transient

expression)

Szado et al. (2003)

Anterior pituitary cells � (used fura 2) ER 1� cell cultures(HSV gene trans-

fer vector)


GH3 cells (pituitary cell

line)

� (used fura 2) ER HSV gene transfer

vector


✓ Nucleus HSV gene transfer

vector


✓ Nucleus, ER Expression of GA

via HSV gene

transfer vector


PC12 cells (Adrenal me-

dulla pheachromo-

cytoma cell line)

� (used fura 2) ER HSV gene transfer

vector


✓ Nucleus HSV gene transfer

vector


✓ Secretory granule

membrane

Transient expression Moreno et al. (2005)

✓ Mitochondria Transient expression Dıaz-Prieto et al. (2008)

✓ Nucleus, mitochondria,

ER

Expression of GA

and RA via HSV

gene transfer

vector

Manjarres et al. (2008)

INS-1 cells

(derived from insulin-

secreting pancreatic

b-cell tumor)

✓ Mitochondria Stable expression Kennedy et al. (1996)

� ER Stable expression Maechler et al. (1999)

H4-IIE cells (hepatoma

cell line)

✓ ER Stable expression Chan et al. (2004)

Tail artery (from male

Wistar rats)

� SR Explant culture

(AdV gene trans-

fer vector)

Rembold et al. (1997)

Hamster

(Cricetulus

griseus)

CHO-K1 cells ✓ – Stable expression Button and Brownstein

(1993), Sanchez-Bueno

et al. (1996)

CHO.T cells ✓ Mitochondria Transient expression Rutter et al. (1996)

CHO cells ✓ Peroxisomes

Mitochondria, ER

Transient expression Lasorsa et al. (2008)

Fruit fly

(Drosophila

melanogaster)

Schneider 2 (S2) cells ✓ – Stable expression Torfs et al. (2002)

Kingdom: Plantae

Soybean (Glycine

max)

Cells in suspension of

[L]., cell line 6.6.12

✓ – Stable expression Mithofer et al. (1999)

Tobacco (Nicoti-

ana tabacum)

Leaf discs ✓ – Stable expression Cessna et al. (2000)

Parsley (Petrose-

linum crispum)

Cells in suspension ✓ – Stable expression Blume et al. (2000)

Kingdom: Fungi

Aspergillus

nidulans

Whole organism ✓ Mitochondria Stable expression Greene et al. (2002)

Saccharomyces

cerevisiae

Whole organism ✓ – Stable expression Nakajima-Shimada et al.

(1991)

� ER lumen Stable expression Strayle et al. (1999)

� Mitochondria Stable expression Jung et al. (2004)

Kingdom: Protista

Trypanosoma

brucei brucei

Procyclic cells ✓ Nucleus Stable expression Xiong and Ruben (1996)

✓ Mitochondria Stable expression Xiong et al. (1997)

AdV, Adenovirus; HSV, Herpes Simplex Virus; CHO.T cells, CHO cells that over-express human insulin receptors; HEK 293T, 293 cells transformed with large

T-antigen from SV40.

Table IIIExamples of the some of the reported coelenterazine loading protocols

Species Coelenterazine loading protocol reported References

Kingdom: Animalia

Mus musculis (Mouse) Native coelenterazine was introduced into adult mice (at 4 mg/kg) by tail-vein

injection and into new-born mice (at 2–4 mg/g) by intraperitoneal injection.

Light emission was recorded immediately

Rogers et al. (2007)

Minced tissues or cells of tissues were incubated with 0.2 ml RPMI 1640

containing 10 mM coelenterazine at 37 �C for 5 h

Yamano et al. (2007)

Danio rerio (Zebrafish) Embryos that had been dechorionated at the 64-cell stage were incubated with

50 mM f-coelenterazine, prepared in 30% Danieau’s solution

Cheung et al. (2006)

Xenopus laevis (African clawed frog) Oocytes were incubated in 2.5 mM coelenterazine in a medium containing

5 mM b-mercaptoethanol

Daguzan et al. (1995)

Drosophila melanogaster (Fruit fly) Malpighian tubules from 4 to 14-day old adults were incubated in Schneider’s

medium containing 2.5 mM coelenterazine for 4–6 h in the dark

Rosay et al. (1997)

Exposed fly brains were incubated in fly ringers solution containing 5 mMnative coelenterazine for >1 h at r.t.

Martin et al. (2007)

Kingdom: Plantae

Nicotiana plumbaginifolia (Tobacco) Seedlings were floated on water containing 2.5 mM coelenterazine o/n at r.t. in

the dark

Knight et al. (1991a, 1996)

Arabidopsis thaliana Seedlings were submerged in 10 mM coelenterazine (which had been dissolved

in ethanol) in dH2O for 7.5 h at r.t. in the dark

Sedbrook et al. (1996)

Seedlings were floated on water containing 2.5 mM coelenterazine o/n at r.t. in

the dark

Knight et al. (1996)

Solanum tuberosum (Potato) Plants were incubated in 5 mM hcp coelenterazine for 8 h Fisahn et al. (2004)

Physcomitrella patens (Moss) Ground up moss tissue was incubated in 0.5 MNaCl, 5 mMmercaptoethanol,

5 mM EDTA, 0.1% gelatin, 10 mM Tris–HCl pH 7.4 containing 2.5 mMcoelenterazine for 6 h or o/n

Russell et al. (1996)

Kingdom: Fungi

Neurospora crassa / Aspergillus niger /

Aspergillus awamori

Vogel’s medium containing 2.5 mM coelenterazine (prepared in methanol) was

inoculated with 1�105 spores/ml. Inoculated medium was incubated for

24 h at 30 �C in the dark

Nelson et al. (2004)

Phyllosticta ampelicida 2.5 mM coelenterazine was pipetted over colonies growing in 1/2� potato

dextrose agar and incubated for 4 h

Shaw et al. (2001)

Schizosaccharomyces pombe Cells were incubated in EMM medium containing 20 mM coelenterazine for

4 h at 30 �CDeng et al. (2006)

Saccharomyces cerevisiae 0.1 volume of 590 mM coelenterazine (prepared in methanol) was added to

25–30 ml yeast culture and incubated for 20 min at r.t.

Batiza et al. (1996)

Kingdom: Protista

Dictyostelium discoidium (Slime mold) Cells were incubated with coelenterazine solution to a final concentration of

50 mM for 24 h at r.t. in the dark

Cubitt et al. (1995)

Kingdom: Monera

Escherichia coli Cells, diluted in 100 mMKCl, 1 mMMgCl2, Tris–HCl, pH 7.5, were incubated

with coelenterazine (final concentration of 2.5 mM) o/n at r.t. in the dark

Knight et al. (1991b)

Bacillus subtilis Bacteria were incubated in TS medium containing 20 mg/ml kanamycin and

2.5 mM coelenterazine h for 1 h in the dark at r.t.

Herbaud et al. (1998)

Anabaena sp. Cells were incubated with coelenterazine (to a final concentration of 2.5 mM)

for 4 h in the dark

Torrecilla et al. (2000)

Tissue Culture Cells

H4-IIE cells Cells were incubated with 5 mM coelenterazine in Krebs–Ringer modified

buVer containing 1 mM EGTA for 1 h at r.t.

Chan et al. (2004)

Neuro2A cells Cells were incubated with 5 mM coelenterazine, 10 mM b-mercaptoethanol

and 4 mM EDTA in PBS for 2–4 h at 4 �CBaubet et al. (2000)

HeLa cells Cells were incubated with 2.5 mM coelenterazine for 6–8 h at 37 �C Brini et al. (1993)

Jurkat cells Cells were incubated with 5 mM coelenterazine-n for 2 h at 37 �C Narayanan et al. (2003)

COS7 cells Cells in suspension were incubated with 2.5 mM coelenterazine in DMEM o/n

at 4 �CKendall et al. (1994)

r.t. is room temperature; o/n is overnight.

Table IVExamples of the detectors used to image aequorin-generated luminescence

Company Type of detector reported Examples of specimens visualized References

Berthold Technologies (U.K.) Ltd,

Hertfordshire, UK

LB980 intensified tube camera Nicotiana plumbaginifolia (tobacco)

seedlings

Wood et al. (2001)

Biospace Lab., Paris, France Cooled GaAs ICCD (3rd generation) Mus musculus (mouse)—intact

animals

Rogers et al. (2007)

Dittie Thermografie, Bonn, Germany Giotto 1.12 microchannel plate-linked

image intensifier tube (1st generation)

Triticum aestivum (winter wheat)

seedlings

Nagel-Volkmann et al. (2009)

EG&G, Berthold Technologies (U.K.)

Ltd, Hertfordshire, UK

Luminograph LB980 low-light camera

system

Phyllosticta ampelicida Shaw et al. (2001)

Hamamatsu Photonics GmbH Deutsch-

land, Herrsching, Germany

C2400-40H ICCD Petroselinum crispum (parsley)

suspension cultures

Blume et al. (2000)

Hamamatsu Photonics Co., Hamamatsu,

Japan

Ultrasensitive VIM camera system (a CCD

camera equipped with an intensifier,

Model C-1400-47)

Arabidopsis thaliana intact plants Furuichi et al. (2001)

VIM photon counting camera Bovine adrenal chromaYn cells Villalobos et al. (2002)

Mouse superior cervical ganglion

neurons

Nunez et al. (2007)

Mouse intact pancreatic islets Villalobos et al. (2005),

Quesada et al. (2008)

C2400-20M ICCD CHO.T cells Rutter et al. (1996)

Photek Ltd., East Sussex, UK RA-IPD H4-IIE cells Chan et al. (2004)

Danio rerio (zebrafish) embryos Cheung et al. (2006)

IPD 3 Drosophila melanogaster (fruit fly)

mushroom bodies

Martin et al. (2007)

Adult mouse retinal explants Agulhon et al. (2007)

Photek 216 ICCD Arabidopsis thaliana seedlings Evans et al. (2005)

EDC-02 ICCD Arabidopsis thaliana seedlings/

leaves

Knight and Knight (2000),

Grant et al. (2000)

Photometrics, Tucson, AZ CH220 CCD imager Arabidopsis thaliana seedlings Sedbrook et al. (1996)

Photonic Science, Robertsbridge, UK Cooled extended ISIS video camera Neuro2A (mouse neuroblastoma

cells)

Baubet et al. (2000)

Princeton Instruments, Trenton, NJ TE/CCD512BKS CCD Nicotiana plumbaginifolia (tobacco)

seedlings

Sai and Johnson (2002)


improved light emission properties of GFP-aequorin are still not completely

understood, it is thought that they likely relate to an improved stability of the

apoaequorin protein or to an increased quantum yield of the Ca2þ-activatedphotoprotein complex, or perhaps both.

The cloned GFP-apoaequorin gene has been engineered to target diVerentsubcellular compartments using similar strategies as those developed and described

already for aequorin (Rizzuto et al., 1992). This approach has also been extended

to include other spectral variants of these aequorin-based reporters, by replace-

ment of the GFP gene with the sequences encoding Venus (YFP), mRFP, and

more recently, mOrange (Bakayan et al., 2009; Curie et al., 2007; Manjarres et al.,

2008). The development of these bifunctional reporters together with new imaging

technologies (see Section IV) has considerably extended the number of applica-

tions possible with aequorin and in particular, has facilitated important advances

in multicompartment measurements of Ca2þ concentrations and in noninvasive

whole animal Ca2þ-imaging studies of the mammalian system.

A. Expression of Apoaequorin

1. Protocol 1: Preparation of Transgenic Zebrafish that Express Apoaequorin in a Tissue-SpecificManner (e.g., in the skeletal musculature)

a. MaterialspiP-HE (apoaequorin plasmid; Inouye et al., 1989)

ap-SK plasmid (a-actin promoter; Higashijima et al., 1997)

pIRES2-EGFP plasmid (Clontech Laboratories, Inc., Mountain View, CA,

USA)

pCMVTNT vector (Promega Corp., Madison, WI, USA)

f-coelenterazine (C-6779; Molecular Probes, Invitrogen Corp., Eugene, OR,

USA)

Methyl cellulose (M0387; Sigma–Aldrich Corp., MO, USA)

b. Methodsi. Preparation of the pa-KS-aeq-IRES-EGFP plasmid. To prepare the pa-KS-

aeq-IRES-EGFP plasmid, use PCR to amplify the apoaequorin gene from the

piP-HE plasmid, with the following oligonucleotide primers: 50-accagaattcatgacaag-caaacaatactcagtcaagcttacatcagac-30 and 50-accagtcgacttaggggacagctccaccgtagag-30,such that EcoR1 and Sal1, are added to the 50 and 30 ends of the apoaequorin

gene, respectively. The apoaequorin gene can then be cloned into the pIRES2-EGFP

plasmid using these restriction sites. Excise the aeq-IRES-EGFP fragment with

EcoR1 and Not1, and then clone it into the ap-SK plasmid to obtain an aeq-

IRES-EGFP fragment with an a-actin promoter (i.e., a-aeq-IRES-EGFP). In paral-

lel, amplify the SV40 late polyadenylation signal (pA) from the pCMVTNT vector

using the following oligonucleotide primers: 50-accagcggccgccagacatgataagatacattg-30 and 50-accagagctctctagaaccggttaccacatttgtagaggtttt-30, adding Not1 to the 50 end,


and Age1,Xba1, and Sac1 to the 30 end of the SV40 late polyadenylation signal. The

SV40 late polyadenylation signal can then be cloned into the pBluescriptII-KSþ

plasmid with Not1 and Sac1, after which the a-aeq-IRES-EGFP fragment can also

be cloned into this plasmid using Xho1 and Not1. The enhanced green fluorescent

protein (EGFP) marker gene is regulated by the IRES-sequence for the subsequent

identification of transgenic fish. Use of the IRES-sequence enables the translation of

both apoaequorin and the EGFP marker from a single mRNA; thus, the expression

level and distribution of EGFP reflects the expression level and distribution of

apoaequorin (Fahrenkrug et al., 1999; Wang et al., 2000).

ii. Generation of transgenic zebrafish that express apoaequorin in the skeletal

muscles. Linearize the pa-KS-aeq-IRES-EGFP plasmid with Xba1 and then

microinject �1 nl (i.e., �100–200 pg) into the center of the blastodisc of the

zebrafish embryos at the one-cell stage. The microinjection pipettes and pressure

injection system used are described in detail in Webb et al. (1997). The injected

embryos should then be maintained at�28.5 �C and screened for the expression of

EGFP after 24 hpf. Embryos (F0) that express EGFP should be raised to adult-

hood for further transgenic germ line screening. This involves: (1) The F0 fish being

crossed with the wild-type fish to get the F1 generation. (2) If some of the F1

embryos express EGFP, then this indicates that one of their parents was transgenic

(i.e., it was heterozygous). The F1 embryos that express EGFP can then be raised

to adulthood and intercrossed with one another to produce the F2 generation.

(3) In the F2 generation, 50% of the oVspring should be heterozygous, 25% should

be homozygous, and 25% should be wild-type. (4) The homozygous F2 transgenic

fish may then be identified by crossing the EGFP-expressing fish with wild-type

fish; if all of the F3 oVspring express EGFP, then their transgenic parent was a

homozygote; if 50% of the F3 embryos express EGFP, then their transgenic parent

was heterozygote. The homozygous F2 fish can then be intercrossed with one

another to obtain stable transgenic lines.

iii. In vivo reconstitution of aequorin. Dechorionate the a-actin-apoaequorintransgenic embryos when they are at the eight-cell stage (we dechorionate embryos

manually with watchmaker’s forceps) and incubate them in a custom-designed

injection/imaging chamber (described in Webb et al., 1997) with 20 mg/ml f-coe-

lenterazine in 30% Danieau’s solution to reconstitute the active aequorin. Prepare

the f-coelenterazine as a stock solution of 2 mg/ml in methanol and dilute it in 30%

Danieau’s solution just prior to use. In this transgenic F2 fish line, the EGFP and

thus the apoaequorin, are expressed in the musculature at low levels at �12 hpf

(i.e., the�6-somite stage) and the level of expression increases in an approximately

linear manner up to�24 hpf. In addition, at�24 hpf EGFP and thus apoaequorin

were expressed throughout the entire musculature, that is, in both the slow and fast

muscles. Thus, this line of muscle-specific apoaequorin-expressing transgenic

zebrafish can be used to visualize and characterize the Ca2þ signals generated in

the trunk musculature during its formation and function. For imaging, these later-

stage embryos may be immobilized with 3% methyl cellulose. We have collected


both temporal and spatial information regarding the Ca2þ signals generated by the

musculature from these embryos up to 96 hpf. For imaging, we use two custom-

built Photon ImagingMicroscope Systems (PIMS; Science Wares, Falmouth, MA,

USA), one based on an IPD-425 (Photek Ltd., Sussex, UK) and the other based on

a back-illuminated EMCCD (DU-897 iXonEMþ camera) that was purchased from

Andor Technology (Belfast, Northern Ireland, UK) and then optimized by Science

Wares Inc. for detecting single photon events at the emission rates typical for

aequorin-based imaging (see Section IV for further details).

B. Expression of GFP-Apoaequorin

2. Protocol 2: Transient and Stable Transfection of Mammalian Cells with GFP-ApoaequorinUsing Plasmid DNA

Many of the expression vectors designed for gene delivery are commercially

available. They contain a number of units, including an immediate-early enhancer/

promoter sequence such as human cytomegalovirus (HCMV), a multiple cloning

region for insertion of the reporter gene, an antibiotic resistance gene (e.g., Ampi-

cillin) for selection of the vector in Escherichia coli, and an additional antibiotic

resistance gene for selection in mammalian cells (e.g., Neomycin G148). For

mammalian expression, a good starting point is to clone the reporter gene into a

vector containing an HCMV promoter (Williams et al., 2005). The constitutive

immediate-early HCMV promoter drives high levels of GFP-aequorin expression

in many mammalian cell lines following transient transfection.

Transfection reagents (e.g., cationic liposomes) enable recombinant DNA deliv-

ery into the nucleus of many immortalized cell types such as HEK293, HeLa,

COS7, CHO and NIH/3T3 cells, with high eYciency. Following transient trans-

fection, stable clones can then be isolated using a combination of drug selection

(e.g., Neomycin (G418) resistance) and cell sorting using flow cytometry. On the

other hand, primary cells (e.g., cortical neurons) are usually transfected with very

low eYciency (i.e., less than 1–5%) using this method and better results can be

obtained by using recombinant viral vectors for gene delivery (Rogers et al., 2005).

a. MaterialsTransfection reagent (e.g., FugeneÒ6 reagent, Roche Applied Science; Lipofec-

tamine, Invitrogen; PolyFect, Qiagen).

Ultrapure plasmid DNA (1 mg/ml)—Plasmid DNA can be purified using a

plasmid purification kit (e.g., QiagenÒpurification kits).

Optimem media (Invitrogen)

35-mm Petri dishes or 8-chamber slides (e.g., those available fromMatTek corp.

or Ibidi Gmbh). Any culture dish will do providing the bottom of the dish is

optimized for high-resolution microscopy on an inverted setup (e.g., dishes

prepared with a glass coverslip mounted underneath a hole cut in the bottom).


This allows the use of objective lenses with high numerical apertures for maximum

light collection.

Native or h-coelenterazine (supplied by Molecular Probes, Invitrogen, US or

Interchim, France; 1-5 mM stock solution prepared in 100% ethanol).

b. Methodsi. Preparation of cells that are transiently transfected with GFP-

apoaequorin. Healthy cell monolayers can be transfected when they are approxi-

mately 50–75% confluent and imaged within 24–48 h following transfection. Wash

cells 1� and resuspend with serum-free medium (with no antibiotic/antimycotic)

and then place cells back into the 37 �C/5% CO2 incubator. For transfecting a 35-

mm dish, prepare a microcentrifuge tube containing 150 ml of serum-free media.

Add 4.5 ml of transfection reagent (this amount may be increased according to the

cell type or reagent used). Vortex the tube and then add 1.5 ml of plasmid DNA

(1 mg/ml). Mix by flicking the tube and leave the tube at room temperature for

approximately 30–45 min to allow the formation of a complex. Following incuba-

tion, remove the cells from the incubator and add 100 ml of the mix drop-wise and

gently swirl the dish before placing back into the incubator. Although optional,

healthy cells can be more easily maintained if the medium is changed after 6 h with

fresh medium containing fetal calf serum (FCS). After 24–48 h, replace the normal

growth medium with a serum-/phenol red-free medium (used for imaging), and

incubate the cells for 1 h with 5mM coelenterazine (native or h-according to the

type of study). Single cells expressing high levels of GFP-aequorin can be selected

by their GFP fluorescence, and used for bioluminescence Ca2þ imaging studies.

ii. Preparation of stable cell lines expressing GFP-apoaequorin. Follow the

protocol for preparing cells for transient transfection (see previous section). At

48 h after transient transfection, start the selection process by adding a selective

medium containing the appropriate antibiotic (e.g., Geneticin, or Puromycin). The

antibiotic concentration used (starting at an upper concentration of 1 mg/ml)

needs to be optimized for diVerent cell types. The medium should be changed

with fresh selective medium every 2–3 days over a period of several weeks. During

this time, the concentration of antibiotic may be gradually decreased. Isolation of

GFP-positive clones into 96-well plates can then be facilitated using flow cytome-

try. In our experience, approximately 10% of the clones survive and proliferate

after a first round of FACS sorting. Since stable transfection is a random integra-

tion event and a large amount of variability is expected, the clones should be

selected based on the level of GFP intensity and homogeneity as well as cellular

morphology, and where instrumentation is available (e.g., Nikon’s Biostation or

the Incucyte from Essen Instruments), clones can also be selected based on growth

curves.


3. Protocol 3: Preparation and Transfection of Organotypic Brain Slices with RecombinantAdenovirus-5 Vector Containing the GFP-Apoaequorin Gene

Recombinant viral vectors, such as human adenovirus serotype 5 (Ad5), adeno-

associated virus (AAV), Sindbis viruses, or retroviruses (e.g., Lentiviruses), are

highly eYcient expression vectors for gene delivery in mammalian cells or tissues

(Tenenbaum et al., 2004). Both Ad5 and the Sindbis virus have been used to

mediate high levels of expression of GFP-apoaequorin in primary neuronal cul-

tures, brain slices, and retinal explant cultures (Rogers et al., 2005). Ad5 was found

to preferentially infect glial cells in cortical- or hippocampal-derived tissue as well

as Muller cells (a type of glial cell) in retinal explant cultures (Rogers et al., 2005).

a. MaterialsVibratome (e.g., Model VT-1000, Leica)

Gas tank (95% O2 / 5 % CO2)

Glass petri dish (3–5 cm diameter)

Membrane filter inserts (e.g., 12 mm TranswellÒ Permeable Supports, Corning)

12-well tissue culture plates

Artificial cerebrospinal fluid (ACSF), pH 7.4 (124 mMNaCl, 3 mMKCl, 2 mM

CaCl2, 1.3 mMMgCl2, 25 mMNaHCO3, 1.25 mMNaHPO4, and 10 mM glucose)

Superglue, Blades, Low temperature melting Agar, Culture medium (50 %

MEM, 25 % HBSS, 25 % Horse Serum, 6.5 mg/ml glucose, 2 mM L-glutamine,

100 U/ml penicillin, 100 mg/ml streptomycin, pH 7.2).

Ad5-GA Viral particles (1�5 x 108 particles).

Specimen chamber for live imaging

b. MethodsOrganotypic slices can be prepared similarly to previously reported methods

(Stoppini et al., 1991). Briefly, prepare 400-600 mm slices as described in section 3b.

Once the slices are cut, gently transfer them using a Pasteur pipette onto a culture

membrane insert and into a 12-well culture plate containing prewarmed culture

medium. Slices can be kept in culture for 4–5 days and viral particles added directly

to the medium approximately 48 h prior to imaging. After viral transfection and

verification of GFP expression, the membrane culture insert with attached slice can

be moved to a larger Petri dish containing growth media (e.g., 35 mm) and the

membrane excised with a scalpel blade. Care should be taken to avoid folding of

the membrane insert, which will cause injury to the tissue. The membrane with

attached slice should then be carefully inserted into an imaging chamber and a

tissue anchor placed on top to secure the tissue (e.g., Series RC-20, Harvard

Apparatus). Once the slice is mounted onto an inverted microscope, a simple

gravity flow system for delivering buVer together with a small peristaltic pump

connected to the output can be used for perfusion (Mohammed et al., 2007).


i. Transgenic mice expressing GFP-apoaequorin reporters. GFP-aequorin has

also been expressed in a number of mammalian cell lines, as well as in organisms

such as Plasmodium bergei (Billker et al., 2004), Drosophila melanogaster (Martin

et al., 2007), and in mice (Rogers et al., 2007). Transgenic mice can provide a

source of cells, tissues, or organs for studies ex vivo or for studies in vivo (Rogers

et al., 2007). In the case of targeted reporters, these are especially useful because

they provide information regarding the localization of any probe-derived signals,

which would otherwise be diYcult because of the low to moderate resolution

aVorded by bioluminescence imaging. In addition, an inducible or conditional

expression system could be introduced into the vector (e.g., Cre/Lox or Tet

inducible elements), to ensure the absence of a phenotype in the event that expres-

sing the reporter ubiquitously from the early stages of development is found to be

detrimental. A conditional expression system also enables expression to be acti-

vated in a specific cell population or at diVerent stages of development. Indeed, the

UAS-Gal4 system enables the specific expression of GFP-aequorin in neuronal

subsets of the fly brain, allowing specific neuroanatomical mapping of Ca2þ

signaling pathways (Martin et al., 2007).

Transgenic mice can be generated via one of two methods: (1) using a ‘classical’

transgenesis approach, where the transgene is randomly integrated into the

genome (Constantini and Lacy, 1981), or (2) by homologous recombination,

which enables directed integration of the transgene (e.g., knock-in of the HPRT

locus; Rogers et al., 2007). Similarly to pronuclear injection, lentiviral vectors can

be used to deliver the transgene into the fertilized mouse egg (Ikawa et al., 2003).

However, these methods can result in random and multiple integrations of the

transgene. In contrast, homologous recombination targets transgene insertion in a

single-copy to a known site in the genome ensuring a more predictable expression

pattern and phenotype based on the known integration site (Bronson et al., 1996).

Transgenic mice conditionally expressing mitochondrially targeted GFP-apoae-

quorin have already been generated using this method (Rogers et al., 2007).

Targeted insertion of the transgene was made 5’ to the X-linked hypoxanthine

phosphoribosyltransferase (HPRT) locus (X-chromosome).

4. Protocol 4: Preparation of Acute Brain Slices from Transgenic Mice ExpressingMitochondrially Targeted GFP-Aequorin

a. MaterialsVibratome (e.g., Model VT-1000, Leica)

Oxygen (95% O2/5% CO2)

Glass petri dish (3–5 cm diameter)

ACSF, pH 7.4 (124 mM NaCl, 3 mM KCl, 1.3 mM MgCl2, 25 mM NaHCO3,

1.25 mM NaHPO4 and 10 mM glucose)

Superglue, Blades, Low temperature melting Agar

Specimen chamber for live imaging (e.g., RC-20 chamber from Harvard

apparatus)


b. MethodsACSF should be prepared fresh on the day of experiments. Two 50 ml falcon tubes

filledwithACSF (Ca2þ free) canbeplacedat�20 �Cfor approximately 1–2hprior to

the preparationof brain slices. This partly frozenmedium is used tofill the bath on the

Vibratome where the brain will be sliced. Once the Vibratome is ready, rapidly

remove the brains from neonates, dissect the cerebellum away, and place the brain

ventrally against the agar block. Horizontal or coronal slices (400 mm) can be cut and

transferred immediately to a small Petri dish containing oxygenated ACSF and

coelenterazine (10 mM) and incubated at room temp, in the dark for 45–60 mins.

Once the slice has been inserted into the imaging chamber, a slice anchor (Harvard

Apparatus) can be used to secure the tissue. Slices shouldbe continuously perfused (at

a flow rate of 1 ml/min) with oxygenated ACSF containing 2 mM CaCl2. Biolumi-

nescence signals can be monitored as previously described (Rogers et al., 2007).

C. Summary of Section II

The commercial development of products, kits, or services enabling genetic

engineering of cells or animals means that these technologies are no longer out

of reach to biologists who have experience in imaging but who have little molecular

biology experience. BRET-based imaging depends on the degree of spectral over-

lap, relative orientation, and the distance between donor and acceptor dipoles.

Fluorescent proteins derived from coelenterates or their variants are therefore

likely to be the most suitable acceptors owing to their structural similarity with

GFP (Tsien, 1998). The recent development of RFP-aequorin (Manjarres et al.,

2008) has provided the means to simultaneously monitor Ca2þ signals from two

diVerent microdomains within a single cell.

III. Introducing Coelenterazines into Cells, Tissues and Embryos

Coelenterazine is the small (�400 Da) prosthetic group that binds with apoae-

quorin to form the active aequorin complex. As coelenterazine is subject to oxida-

tion, it is normally supplied in a sealed vial free of O2. Prior to reconstitution, the

coelenterazine should be stored at �20 �C. In addition, coelenterazine is poorly

soluble in water and so stock solutions are normally prepared in methanol. In this

form, coelenterazine is stable for �3 months at �20 �C. Coelenterazine was

originally isolated from Aequorea victoria; however, in the 1970s a procedure for

chemically synthesizing coelenterazine was developed (Inoue et al., 1975; Kishi

et al., 1972). This procedure has since been used for preparing coelenterazine and

its analogs (Jones et al., 1999). Many of the coelenterazine analogs possess proper-

ties diVerent from those of native coelenterazine. These include half-life, aequorin

regeneration rate, luminescence capacity, emission maximum and membrane

permeability, the latter being due to the lipophilic nature of coelenterazine.

For example, f-coelenterazine has the same half-life as native coelenterazine


(Shimomura, 1991) but when it is reconstituted with apoaequorin to form an

aequorin complex (f-aequorin) the level of luminescence produced on reaction

with Ca2þ is almost 20-fold higher than that produced when native coelenterazine

is used. In addition, f-coelenterazine has the highest permeability through cell

membranes (Shimomura, 1997).

As coelenterazine is lipophilic, apoaequorin-expressing cells, tissues, and whole

organisms can simply be incubated in coelenterazine solution. However, this

method is successful only in tissue culture cells and in simple organisms that

have a large surface area-to-volume ratio where eYcient diVusion occurs. In

more complex, multicellular organisms such as developing vertebrate embryos,

reconstituting aequorin is more of a challenge. In the case of our a-actin-apoae-quorin transgenic zebrafish, we started our f-coelenterazine incubation as early as

the eight-cell stage (i.e., 1.25 hpf ) when the embryonic cells had a large surface

area-to-volume ratio, and embryos were incubated continually in this 20 mg/ml

coelenterazine both up until, and during data collection, which took place from 16

to 48 hpf. In the case of the apoaequorin-expressing transgenic mice, Rogers et al.

(2007) introduced native coelenterazine into adult mice (at 4 mg/kg) by injection

into the tail vein and into new-born mice (at 2–4 mg/g) by intraperitoneal injection.These, and other protocols used for introducing coelenterazine into various intact

organisms and tissue culture cells are summarized in Table III.

IV. Techniques for Detecting Aequorin Luminescence

Currently, several diVerent types of equipment are commercially available that

can be used to detect or visualize aequorin-generated luminescence. These range in

capability, price, design, and commercial availability. At the lower cost end, there

is the simple test tube/culture dish luminometer, which provides only temporal

Ca2þ signaling information, costs just a few hundred US dollars, and is supplied by

several diVerent companies. At the higher cost end are several custom-designed

imaging systems, which provide both temporal and spatial luminescent informa-

tion, as well as bright-field and fluorescence images (if required), to enable the

correlation of Ca2þ signaling events with morphological features and other cellular

changes. These systems are obviously a lot more costly and are built to order by a

small number of specialist companies. Some examples of the types of detectors that

have been used to image aequorin-generated luminescence are shown in Table IV.

It may be diYcult to justify the purchase of expensive single photon imaging

equipment at early stages in a project. Often a photon counting photomultiplier

tube (PMT) can be used in place of an imaging photon detector (IPD) to determine

the timing and amplitude of bioluminescence signals in living systems. By adding a

near-IR light source and an appropriate blocking filter in front of the PMT, a

relatively inexpensive near-IR sensitive camera can be used to continuously moni-

tor morphological development while the PMT reports total bioluminescence

activity. An example of this type of system is shown in Fig. 2A. Accurate correlation

Dark box

IRcamera

Zoomlens

Feed-through

Shutter

Light-proofchamber

FilterLens

IRlighting

Beam-splitter Lens

IR-sensitivecamera

Sample stage(large field of view)

Doorswitch

HV/amplifier/discriminator

PMT

PMT1

ICCD

Sample

ManualPMTshutterIR

blockingfilter

IR ring light

Temperature-controlledstage

Monitor

ComputerPMT system

controller

PerfusionPerfusion

B

A

C

Mirror

CellsGlass

window

Dichroicmirror

(585 DCXR)

D630-60

PMT2

D535-50

Perfusionchamber

1252

IRimage

Count plot

Fig. 2 Schematic representations of three recently developed luminescence detection systems. (A)

A system that combines a PMT with an IR light source and an IR-sensitive camera. This enables both

single photon detection in the visible light spectrum and monitoring the sample continuously in real-

time using IR light. The system was designed and built by Science Wares, Inc., (Falmouth, MA). (B)

A two-channel luminometer, which was designed to simultaneously collect temporal Ca2þ signaling



of bioluminescence signals with morphological events facilitates planning, optimi-

zation, and standardization of protocols for subsequent imaging experiments.

Over the past few years, several novel aequorin-based detection system designs

have been reported. One is the two-channel luminometer, which was made for

dual-wavelength aequorin measurements (see Fig. 2B). This system was designed

by Manjarres et al., 2008 and built by Cairn Research (Faversham, UK). Here, the

luminescence emission from two spectrally distinct aequorins (GFP-aequorin and

RFP-aequorin) that are coexpressed in diVerent subcellular locations within the

same cells is divided by a dichroic mirror and the resulting beams of light are

filtered at 535 and 630 nm and then collected by two separate PMTs (Manjarres

et al., 2008).

Another new Ca2þ detection device is an imaging system capable of acquiring

real-time bioluminescence data from living (and unrestrained) small animals such

as mice (see Fig. 2C). This system was designed by Roncali et al. (2008). It is based

around a Photon ImagerTM intensified CCD camera (Biospace Lab., France)

operating in a photon counting mode. The ICCD camera is set on top of a light-

tight chamber and records optical signals at a video rate of 25 Hz. Motion can also

be monitored by using two cameras, one that records the signal of interest and the

other that is used to video-track the animal. The latter can be achieved by

illuminating the field of view with infrared light. The signals from both cameras

are recorded simultaneously and electronically synchronized. A detailed descrip-

tion of this equipment is given by Roncali et al. (2008). A similar approach has

been used with microscope-based imaging systems to continuously acquire biolu-

minescence image data emitted at short wavelengths while using longer wavelength

illumination to simultaneously record transmitted light images that show mor-

phology (Speksnijder et al., 1990).

One of the most recent developments in bioluminescence detection involves

using an EMCCD detector for single photon imaging (Martin et al., 2007;

Rogers et al., 2008). The best bioluminescence imaging detectors are capable of

single photon detection, and this requires that the detector somehow amplify

the detected signal above background noise. All electronic imaging detectors

ultimately convert incident photons from the sample into detected electrons,

information from cells expressing GFP- and RFP-aequorin in diVerent organelles. The system was

designed by I. M. Manjarres, P. Chamero, M. T. Alonso, and J. Garcıa-Sancho (Universidad de

Valladolid and Consejo Superior de Investigaciones Cientıficas, Valladolid, Spain), and B. Domingo,

F. Molina and J. Llopis (Universidad de Castilla-LaMancha, Albacete, Spain), and was built by Cairne

Research (Faversham, UK). (C) A photon counting-based system, with a video monitoring function

(via an IR-sensitive camera), for whole-body optical imaging of un-restrained, freely moving small

animals, such as mice. The system was designed by E. Roncali, K. L. Rogers and B. Tavitian (Labor-

atiore d’Imagerie Moleculaire Experimentale, INSERM U803, Orsay, France) and M. Savinaud,

O. Levrey and S. Maitrejean (Biospace Lab, Paris). Panels (B) and (C) are modified from Fig. 1 in

Manjarres et al. (2008) and Roncali et al. (2008), respectively.

Sample Photo

Photons

Photo

ns


which get amplified by an electronic circuit. The output of the circuit is monitored

by a computer for recording purposes, and the collected data is then postprocessed

to generate meaningful images (see Fig. 3).

Bioluminescence can be diYcult to image because there is often very little signal,

and because it is diYcult to predict exactly when and where the signal will be

produced. Because the signals are often quite small, it is important to use the most

eYcient optical system possible to collect the light. An overview of the main types

of optical systems used for collecting light from bioluminescent samples is given by

Karplus (2006) and see Fig. 4. In microscopy, it is important to select an objective

with the highest ratio of numerical aperture to magnification (NA/mag). In macro-

imaging, it is important to select a lens with smallest possible working distance and

the lowest possible f-stop, typically known as a ‘‘fast’’ lens. It can be helpful to have

an electronic source on hand to test the eYciency of an optical system in biolumi-

nescence imaging mode and ensure that the various components of the system are

performing as expected (Creton and JaVe, 2001).Another important consideration for bioluminescence imaging is that the sam-

ple being imaged must be kept in a light-tight enclosure to eliminate any direct or

reflected ambient light from reaching the imaging detector, as it could easily

overwhelm the bioluminescence signal. Because the sample is normally kept in

complete darkness during bioluminescence imaging, it can be beneficial to periodi-

cally obtain bright-field and fluorescence images to determine the stage of devel-

opment or morphological condition of the sample.

Such bright-field and fluorescence images are typically orders of magnitude

brighter than bioluminescent images, so they can be obtained in a very short

period of time compared with that required to accumulate a meaningful biolumi-

nescence image. Until the recent development of deep cooled, back-thinned, elec-

tron multiplying charge-coupled devices (EMCCDs), detectors capable of single

photon imaging for bioluminescence were not well suited to acquiring bright-field

and fluorescence images because they used microchannel plates to amplify and

transmit images inside the detector. Microchannel plates blur and distort the

OpticsImaging

detector andelectronics

ns Photons Electrons Computerrecording

Photon dataprocessing

Photon images

Fig. 3 In bioluminescent imaging, some of the photons emitted from a sample are collected by an

optical system and directed onto a detector that converts the incident photons into electrons. These

electrons are processed by electronic circuits that provide data to a computer indicating the time and

position of the detected photons. The computer program then postprocesses the photon data to generate

images.

Imaging mode

Imagedata

Photonimagingsoftware

Shutter driver

Multicore CPU

Dark box

Dark box

BioluminescenceA

B

C

Bright-field Fluorescence

Fluorescencecondenser

Fluorescencecondenser

Epifluorescentfilter set

Epifluorescentfilter set

Fiber optic light guide

Light frommercuryarc lamp

Fiber opticlight guide

Bright-fieldcondenser

Electronic shutter

Electronic shutter

Bioluminescent/fluorescent sample

MotorizedXY stage

MotorizedObjective turret

Motorizedfilter turret

Tubelens

Reducinglens

Electronicshutter

Light fromhalogen lamp

FL slider

Motorizedmirror

Tubelens

Reducinglens

Electronicshutter

*Waterchiller

*Only required for RA-IPD

RA

-IP

D s

yste

mE

MC

CD

sys

tem

alo

neC

omm

on fo

r E

MC

CD

,R

A-I

PD

and

PM

T

EMCCD EMCCD

RA-IPDRA-IPD RA-IPD

CCDCCDCCD

EMCCD

Waterchiller

To microscopeinterface (in CPU)

To microscopeinterface (in CPU)

To EMCCD interface (in CPU)

To RA-IPD interface (in CPU) andRA-IPD controller

To framegrabber(in CPU)

EMCCD interface

Microscopeinterface

Imagedata

Photonimagingsoftware

Shutter driver

Multicore CPU

RA-IPDcontroller

Microscope interface

RA-IPD interface

Frame grabber

Motorizedfocus

Motorizedfocus

High NAobjective

Halogenlamp

Mercuryarc lamp

Fig. 4 Schematic representation of luminescence imaging systems based around a modified EMCCD

and an RA-IPD, which can be used to acquire bright-field and fluorescence imaging information as wel

as collect bioluminescence data. (A) Components that are common to both the EMCCD- andRA-IPD

based systems. (B and C) Components that are specific for the (B) EMCCD- and (C) RA-IPD-based


l

-


images they transmit, have a limited dynamic range, and can also be permanently

damaged if they are exposed to too much signal. While a nonintensified detector

could be used to acquire bright-field and fluorescence images, the problem of blur

and distortion in the bioluminescence image would still remain. In addition, it was

a challenge to adjust the size and registration of images acquired from two diVerentdetectors.

An EMCCD sensor can be fabricated on a substrate alongside one or more

amplifiers with programmable gain settings, so a computer can rapidly change

from a low-gain to a high-gain setting. Thus, a computer-controlled EMCCD is

capable of acquiring bright-field and fluorescence images at low-gain settings, as

well as bioluminescence images at high-gain settings (see Fig. 4). This makes it

possible to eliminate blurring and distortion in all three types of images to be

acquired, improving the spatial resolution in the bioluminescence images by 2–3

times over what can be achieved with a microchannel plate based detector. Because

all three types of images can be obtained with the same sensor, the scale and

registration are identical as well. Furthermore, back-thinned EMCCDs have a

significantly higher quantum eYciency in the visible light spectrum, compared with

photocathode materials used in intensifier-based detectors, so they are able to

respond more eYciently to weak signals.

While EMCCDs are capable of detecting single photon events when the electron

multiplying gain is high enough to overcome the read noise of the output amplifier,

the electron multiplying gain mechanism is subject to substantial statistical varia-

tion. For example, when an output signal of 1000 electrons occurs, it is easily

detected as a meaningful event, but it is not possible to be certain how many input

electrons generated this signal—it may have been just 1, or 2, or 5 input electrons.

As a result, EMCCDs operating in photon counting mode have limited ability to

track large intensity changes. The maximum signal intensity that can be recorded

reliably is essentially determined by the frame rate at which the sensor is read out.

The range can be extended at the expense of the field of view by selecting a small

region of interest, and/or at the expense of spatial resolution by binning together

adjacent pixels on the sensor.

Two additional limitations of EMCCDs arise from the circuitry used to read out

the image data. First, in order to record the signal detected by the CCD sensor,

systems, respectively. The EMCCD can be modified to acquire bioluminescence information as well as

bright-field and fluorescence images. On the other hand, a resistive-anode Imaging Photon Detector

(RA-IPD) can be used in conjunction with a CCD camera, the latter to acquire bright-field and

fluorescent images, when a higher dynamic range and temporal resolution are needed for the biolumi-

nescence signal. For both the EMCCD and RA-IPD-based imaging systems, a high level of automation

for the microscope makes it possible to rapidly switch between the various imaging modes, and also

makes it possible to have the computer run automated acquisition sequences over extended periods,

typically overnight. The motorized focus allows the computer to acquire image stacks in any imaging

mode for three dimensional reconstructions. Both systems were designed and built by Science Wares

Inc., Falmouth, MA, USA.


there must be a period of dead time while the accumulated image data is shifted out

of the active area of the sensor. Second, the signals used to shift the image data

from the active area into the readout amplifier can also impose noise on the output

that looks identical to single photon events in the photon counting mode. Signifi-

cant advances have been made to minimize the noise (called clock-induced charge)

that is generated. However, these two eVects are still significant such that increas-

ing the readout rate beyond a certain level actually degrades the net single photon

imaging performance of the EMCCD.

In situations where significant bioluminescence intensity changes are taking

place over periods less than a second, an intensifier-based detector is likely to be

a better choice than an EMCCD (see Fig. 4). The spatial resolution of commer-

cially available intensifier-based detectors is usually in the order of tens of microns

at the detector input window, which is adequate for many applications, but not as

good as what can be achieved with commercially available EMCCD detectors,

which typically have a pixel size of 8 or 16 mm. There are two main types of

intensifier-based detectors, those with a phosphor image output that is optically

coupled to a visible light CCD, and those with an electrically encoded anode that

produces position sensitive pulses for each detected event. The temporal resolution

of detectors with an optically coupled phosphor output is typically in the range of

tens of milliseconds due to the persistence time of the phosphor and the frame rate

of the CCD.

The best temporal resolution for single photon imaging can be in the order of

tens of nanoseconds, and is achieved only by intensifier-based detectors with an

electrically encoded anode. The dynamic range of such detectors is constrained by

the dead time of the pulse processing electronics, which is typically in the order of a

few microseconds per detected event. Improvements continue to be made in the

spatial resolution of microchannel plates and throughput speed of encoded anode

detectors (Lapington, 2004; Siegmund et al., 2005), but the cost and complexity of

operating such detectors has prevented them from being used widely for biolumi-

nescence imaging so far.

Another factor that should be considered when selecting a detector for biolumi-

nescence imaging is the expected signal-to-noise ratio of the recorded image data

(Karplus, 2006). Frequently, accepting lower spatial resolution can result in a

better signal-to-noise ratio. In situations where a fast or brief signal needs to be

identified with high temporal accuracy, a photocathode detector is often capable of

a better signal-to-noise ratio than an EMCCD. Even though an EMCCD detector

can have 2�–20� higher quantum eYciency than a detector with a photocathode,

at the high readout rates needed for good temporal resolution, incident photons

can still be lost during the dead time needed to transfer image data into the readout

frame of the EMCCD, and the photons that are detected can be obscured by clock-

induced readout noise. Figure 5 shows a comparison of the bioluminescence

images acquired by an EMCCD and an RA-IPD photon imaging system at two

diVerent stages of zebrafish development.

Cleavage period

i iAPview

Sideview

APview

Sideview

4-cell stage1 hpf

ii EM-CCD

1 2 3 4 >5 1–2 3–4 5–6 7–8 >9 1 2 3 4 >5 1–2 3–4 5–6 7–8 >9

Pos

ition

ing

Pro

paga

tion

Dee

peni

ngA

ppos

ition

iii IPD 425 ii EM-CCD iii IPD 425

Sphere stage4 h

Blastula periodBA

Fig. 5 Comparison of the bioluminescence images acquired by the EMCCD and RA-IPD photon

imaging systems during (A) the Cleavage Period and (B) Blastula Period of zebrafish development. (Ai

andBi) Schematics of an embryo froma side (animal pole—AP) viewand top viewat the (Ai) Four-cell stage

(i.e., �1 hpf) and (Bi) sphere stage (i.e., �4 hpf) to show the morphology of the embryo and the typical

patterns of Ca2þ signals (in red) observed at these two stages of development. (Aii, Aiii and Bii, Biii)

Representative AP views of f-aequorin loaded embryos to show the changes in intracellular free Ca2þ that

occur (Aii andAiii) at diVerent timesduring the secondcell division cycle (i.e., two- to four-cell stage) and (Bii

andBiii) at sphere stage. The imageswere acquired using anAndorEMCCD-based imaging system (Aii and

Bii) and a Photek IPD 425-based imaging system (Aiii and Biii). In both cases, luminescence was accumu-

lated for 30 s. Color scales indicate luminescent flux in photons/sec. Scale bars are 200 mm.



V. Conclusions

Living cells, tissues and whole organisms are essentially defined by their complex

spatial structures. Underlying such broad morphological characteristics are much

finer molecular assemblies, such as microdomains within the cytoskeleton, plasma

membrane, nucleus and cytoplasm. It is becoming clear that such microdomains are

the loci of many key signaling events, including those that involve Ca2þ and as such,

they are becoming amajor area of interest for investigators in theCa2þ signaling field.

In recent years, researchers have made spectacular advances on the live-cell imaging

front, due in part to the development of techniques that combine breaking Ernst

Abbe’s diVraction barrier (Abbe, 1873) being compatible with examining living

systems. Such techniques include stimulated emission depletion (STED) microscopy

(Hell, 2007), photoactivated localization microscopy (PALM) (Betzig et al., 2006),

and stochastic optical reconstruction microscopy (STORM) (Rust et al., 2006). It is

imperative, therefore, that Ca2þ imaging also joins the super-resolution revolution,

and indeed significant progress has been made on this front with the development of

fluorescence-based techniques such as single channel Ca2þ nanoscale resolution

(SCCaNR) microscopy (Wiltgen et al., 2009). The continued development of both

fluorescent and luminescent GET-CRs will undoubtedly alsomake a contribution to

this advancement, especially if, in the case of the latter, the intensity of the Ca2þ-mediated luminescent emission can be increased. We find ourselves, therefore, at a

most exciting and opportune time to extend our understanding of Ca2þ signaling in

living cells from the microscopic to the nanoscopic level.

Acknowledgments

We thank Philippe Brulet, Marc Knight, and Jean-Rene Martin, who kindly gave us permission to

use their previously published work. Special thanks also to Osamu Shimomura for his generous support

of aequorin-based imaging over the years. We acknowledge financial support from Hong Kong RGC

GRF grants: HKUST-6241/04M,-6416/06M,-661707 and-662109.

References

Abbe, E. (1873). Beitrage zur Theorie des Mikroskops und der Mikroskopischen Wahrnehmung. Arch.

f. Mikr. Anat. 9, 413–420.

Agulhon, C., Platel, J. C., Kolomiets, B., Forster, V., Picaud, S., Brocard, J., Faure, P., and Brulet, P.

(2007). Bioluminescent imaging of Ca2þ activity reveals spatiotemporal dynamics in glial networks of

dark-adapted mouse retina. J. Physiol. 583, 945–958.

Alonso, M. T., Barrero, M. J., Carnicero, E., Montero, M., Garcia-Sancho, J., and Alvarez, J. (1998).

Functional measurements of [Ca2þ] in the endoplasmic reticulum using a herpes virus to deliver

targeted aequorin. Cell Calcium 24, 87–96.

Ashworth, R., and Brennan, C. (2005). Use of transgenic zebrafish reporter lines to study calcium

signalling in development. Brief Funct. Genomic Proteomic 4, 186–193.

Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (1999). Circular permutation and receptor insertion

within green fluorescent proteins. Proc. Natl. Acad. Sci. USA 96, 11241–11246.


Bakayan, A., Picazo, F., Domingo, B., and Llopis, J. (2009). Fluorescent protein-aequorin fusions as

improved bioluminescent Ca2+ reporters. Acta Physiologica 195 (Suppl. 667), P121.

Baker, P. F., Hodgkin, A. L., and Ridgway, E. B. (1971). Depolarization and calcium entry in squid

giant axons. J. Physiol. 218, 709–755.

Basset, O., Boittin, F. X., Dorchies, O. M., Chatton, J. Y., van Breemen, C., and Ruegg, U. T. (2004).

Involvement of inositol 1, 4, 5-trisphosphate in nicotinic calcium responses in dystrophic myotubes

assessed by near-plasma membrane calcium measurement. J. Biol. Chem. 279, 47092–47100.

Batiza, A. F., Schulz, T., and Masson, P. H. (1996). Yeast respond to hypotonic shock with a calcium

pulse. J. Biol. Chem. 271, 23357–23362.

Baubet, V., Le Mouellic, H., Campbell, A. K., Lucas-Meunier, E., Fossier, P., and Brulet, P. (2000).

Chimeric green fluorescent protein-aequorin as bioluminescent Ca2þ reporters at the single-cell level.

Proc. Natl. Acad. Sci. USA 97, 7260–7265.

Behne, M. J., Tu, C. L., Aronchik, I., Epstein, E., Bench, G., Bikle, D. D., Pozzan, T., and

Mauro, T. M. (2003). Human keratinocyte ATP2C1 localizes to the Golgi and controls Golgi.

Ca2þ stores. J. Invest. Dermatol. 121, 688–694.

Bencina, M., Legisa, M., and Read, N. D. (2005). Cross-talk between cAMP and calcium signalling in

Aspergillus niger. Mol. Microbiol. 56, 268–281.

Berridge, M. J. (2009). Inositol trisphosphate and calcium signalling mechanisms. Biochim. Biophys.

Acta 1793, 933–940.

Betzig, E., Patterson, G. H., Sougrat, R., Lindwasser, O. W., Olenych, S., Bonifacino, J. S.,

Davidson, M. W., Lippencott-Schwartz, J., and Hess, H. F. (2006). Imaging intracellular fluorescent

proteins at nanometer resolution. Science 313, 1642–1645.

Biagioli, M., Pinton, P., Scudiero, R., Ragghianti, M., Bucci, S., and Rizzuto, R. (2005). Aequorin

chimeras as valuable tool in the measurement of Ca2þ concentration during cadmium injury.

Toxicology 208, 389–398.

Billker, O., Dechamps, S., Tewari, R., Wenig, G., Franke-Fayard, B., and Brinkmann, V. (2004).

Calcium and calcium-dependent protein kinase regulate gamete formation and mosquito transmis-

sion in a malaria parasite. Cell 117, 503–514.

Blume, B., Nurnberger, T., Nass, N., and Scheel, D. (2000). Receptor-mediated increase in cytoplasmic

free calcium required for activation of pathogen defense in parsley. Plant Cell 12, 1425–1440.

Bovolenta, S., Foti, M., Lohmer, S., and Corazza, S. (2007). Development of a Ca2þ-activated photo-

protein, PhotinaÒ, and its application to high-throughput screening. J. Biomol. Screen. 12, 694–704.

Brandenburger, Y., Kennedy, E. D., Python, C. P., Rossier, M. F., Vallotton, M. B., Wollheim, C. B.,

and Capponi, A.M. (1996). Possible role for mitochondrial calcium in angiotensin II- and potassium-

stimulated steroidogenesis in bovine adrenal glomerulosa cells. Endocrinology 137, 5544–5551.

Brandenburger, Y., Arrighi, J. F., Rossier, M. F., Maturana, A., Vallotton, M. B., and Capponi, A. M.

(1999). Measurement of perimitochondrial Ca2þ concentration in bovine adrenal glomerulosa cells

with aequorin targeted to the outer mitochondrial membrane. Biochem. J. 341, 745–753.

Brini, M., Murgia, M., Pasti, L., Picard, D., Pozzan, T., and Rizzuto, R. (1993). Nuclear Ca2þ

concentration measured with specifically targeted recombinant aequorin. EMBO J. 12, 4813–4819.

Brini, M., Marsault, R., Bastianutto, C., Pozzan, T., and Rizzuto, R. (1994). Nuclear targeting of

aequorin: A new approach for measuring nuclear Ca2þ concentration in intact cells. Cell Calcium 16,

259–268.

Brini, M., De Giorgi, F., Murgia, M., Marsault, R., Massimino, M. L., Cantini, M., Rizzuto, R., and

Pozzan, T. (1997). Subcellular analysis of Ca2þ homeostasis in primary cultures of skeletal muscle

myotubes. Mol. Biol. Cell 8, 129–143.

Brini, M., Manni, S., Pierobon, N., Du, G. G., Sharma, P., MacLennan, D. H., and Carafoli, E. (2005).

Ca2þ signaling in HEK-293 and skeletal muscle cells expressing recombinant ryanodine receptors

harboring malignant hyperthermia and central core disease mutations. J. Biol. Chem. 280,

15380–15389.

Bronson, S. K., Plaehn, E. G., Kluckman, K. D., Hagaman, J. R., Maeda, N., and Smithies, O. (1996).

Single-copy transgenic mice with chosen-site integration. Proc. Natl. Acad. Sci. USA 93, 9067–9072.


Button, D., and Brownstein, M. (1993). Aequorin-expressing mammalian cell lines used to report Ca2þ

mobilization. Cell Calcium 14, 663–671.

Cessna, S. G., Sears, V. E., Dickman, M. B., and Low, P. S. (2000). Oxalic acid, a pathogenicity factor

for Sclerotinia sclerotiorum, suppresses the oxidative burst of the host plant. Plant Cell 12,

2191–2199.

Challet, C., Maechler, P., Wollheim, C. B., and Ruegg, U. T. (2001). Mitochondrial calcium oscillations

in C2C12 myotubes. J. Biol. Chem. 276, 3791–3797.

Chamero, P., Villalobos, C., Alonso,M. T., andGarcıa-Sancho, J. (2002). Dampening of cytosolic Ca2þ

oscillations on propagation to nucleus. J. Biol. Chem. 277, 50226–50229.

Chamero, P., Manjarres, I. M., Garcıa-Verdugo, J. M., Villalobos, C., Alonso, M. T., and Garcıa-

Sancho, J. (2008). Nuclear calcium signaling by inositol trisphosphate in GH3 pituitary cells. Cell

Calcium 43, 205–214.

Chan, C., Harland, M. L., Webb, S. E., Chen, J., Miller, A. L., and Barritt, G. J. (2004). Evaluation,

using targeted aequorins, of the roles of the endoplasmic reticulum and its (Ca2þþMg2þ)ATP-ases in

the activation of store-operated Ca2þ channels in liver cells. Cell Calcium 35, 317–331.

Chapuy-Regaud, S., Jones, H. E., Campbell, A. K., and Trombe, M. C. (2001). Validation of the use of

aequorin for cytoplasmic free calcium determination by chemiluminescence in Streptococcus pneu-

moniae. J. Soc. Biol. 195, 271–276.

Cheung, C. Y. (2009). Exploring the role of Ca2þ signaling in slow muscle development using intact

zebrafish embryos as an animal model (PhD thesis, HKUST, 2009).

Cheung, C. Y., Webb, S. E., Meng, A., andMiller, A. L. (2006). Transient expression of apoaequorin in

zebrafish embryos: Extending the ability to image calcium transients during later stages of develop-

ment. Int. J. Dev. Biol. 50, 561–569.

Constantini, F., and Lacy, E. (1981). Introduction of a rabbit b-globin gene into the mouse germ line.

Nature 294, 92–94.

Creton, R., and JaVe, L. F. (2001). Chemiluminescence microscopy as a tool in biomedical research.

Biotechniques 31, 1098–1105.

Cubitt, A. B., Firtel, R. A., Fischer, G., JaVe, L. F., and Miller, A. L. (1995). Patterns of free calcium in

multicellular stages of Dictyostelium expressing jellyfish apoaequorin. Development 121, 2291–2301.

Curie, T., Rogers, K. L., Colasante, C., and Brulet, P. (2007). Red-shifted aequorin-based biolumins-

cent reporters for in vivo imaging of Ca2+ signaling. Mol. Imag. 6, 30–42.

Daguzan, C., Nicolas, M. T., Mazars, C., Leclerc, C., andMoreau, M. (1995). Expression of membrane

targeted aequorin in Xenopus laevis oocytes. Int. J. Dev. Biol. 39, 653–657.

Deng, L., Sugiura, R., Takeuchi, M., Suzuki, M., Ebina, H., Takami, T., Koike, A., Iba, S., and

Kuno, T. (2006). Real-time monitoring of calcineurin activity in living cells: Evidence for two distinct

Ca2þ-dependent pathways in fission yeast. Mol. Biol. Cell 17, 4790–4800.

Dıaz-Prieto, N., Herrera-Peco, I., de Diego, A. M. G., Ruiz-Nuno, A., Gallego-Sandin, S.,

Lopez, M. G., Garcıa, A. G., and Cano-Abad, M. F. (2008). Bcl2 mitigates Ca2þ entry and

mitochondrial Ca2þ overload through downregulation of L-type Ca2þ channels in PC12 cells. Cell

Calcium 44, 339–352.

Dodd, A. N., Jakobsen, M. K., Baker, A. J., Telzerow, A., Hou, S. W., Laplaze, L., Barrot, L.,

Poethig, R. S., HaseloV, J., and Webb, A. A. R. (2006). Time of day modulates low-temperature

Ca2þ signals in Arabidopsis. Plant J. 48, 962–973.

Evans, N. H., McAinsh, M. R., Hetherington, A. M., and Knight, M. R. (2005). ROS perception in

Arabidopsis thaliana: The ozone-induced calcium response. Plant J. 41, 615–626.

Fahrenkrug, S. C., Clark, K. J., Dahlquist, M. O., and Hackett, P. B., Jr. (1999). Dicistronic gene

expression in developing zebrafish. Mar. Biotechnol. 1, 552–561.

Falciatore, A., d’Alcala, M. R., Croot, P., and Bowler, C. (2000). Perception of environmental signals

by a marine diatom. Science 288, 2363–2366.

Fisahn, J., Herde, O., Willmitzer, L., and Pena-Cortes, H. (2004). Analysis of the transient increase in

cytosolic Ca2þ during the action potential of higher plants with high temporal resolution:


Requirement of Ca2þ transients for induction of jasmonic acid biosynthesis and PINII gene expres-

sion. Plant Cell Physiol. 45, 456–459.

Furuichi, T., Mori, I. C., Takahashi, K., and Muto, S. (2001). Sugar-induced increase in cytosolic Ca2þ

in Arabidopsis thaliana whole plants. Plant Cell Physiol. 42, 1149–1155.

Gerasimenko, O., and Tepikin, A. (2005). How to measure Ca2þ in cellular organelles? Cell Calcium 38,

201–211.

George, C. H., Kendall, J. M., Campbell, A. K., and Evans, W. H. (1998). Connexin-aequorin chimerae

report cytoplasmic calcium environments along trafficking pathways leading to gap junction biogen-

esis in living COS-7 cells. J. Biol. Chem. 273, 29822–29829.

Gorokhovatsky, A. Y., Marchenkov, V. V., Rudenko, N. V., Ivashina, T. V., Ksenzenko, V. N.,

Burkhardt, N., Semisotnov, G. V., Vinokurov, L. M., and Alakhov, Y. B. (2004). Fusion of Aequoria

victoria GFP and aequorin provides their Ca2þ-induced interaction that results in red shift GFP

absorptionand eYcient bioluminescence energy transfer.Biochem.Biophys.Res.Commun.320,703–711.

Grant, M., Brown, I., Adams, S., Knight, M., Ainslie, A., and Mansfield, J. (2000). The RPM1 plant

disease resistant gene facilitates a rapid and sustained increase in cytosolic calcium that is necessary

for the oxidative burst and hypersensitive cell death. Plant J. 23, 441–450.

Greene, V., Cao, H., Schanne, F. A. X., and Bartelt, D. C. (2002). Oxidative stress-induced calcium

signalling in Aspergillus nidulans. Cell. Signal. 14, 437–443.

Griesbeck, O., Baird, G. S., Campbell, R. E., Zacharias, D. A., and Tsien, R. Y. (2001). Reducing the

environmental sensitivity of yellow fluorescent protein mechanism and applications. J. Biol. Chem.

276, 29188–29194.

Hell, S. W. (2007). Far-field optical nanoscopy. Science 316, 1153–1158.

Herbaud, M. L., Guiseppi, A., Denizot, F., Haiech, J., and KilhoVer, M. C. (1998). Calcium signalling

in Bacillus subtilis. Biochim. Biophys. Acta 1448, 212–226.

Higashijima, S., Okamoto, H., Ueno, N., Hotta, Y., and Eguchi, G. (1997). High-frequency generation

of transgenic zebrafish which reliably express GFP in whole muscles or the whole body by using

promoters of zebrafish origin. Dev. Biol. 192, 289–299.

Ikawa, M., Tanaka, N., Kao, W. W., and Verma, I. M. (2003). Generation of transgenic mice using

lentiviral vectors: a novel preclinical assessment of lentiviral vectors for gene therapy. Mol. Ther. 8,

666–673.

Inoue, S., Sugiura, S., Kakoi, H., and Hasizume, K. (1975). Squid bioluminescence. II. Isolation from

Watasenia scintillans and synthesis of 2-(p-hydroxybenzyl)-6-(p-hydroxyphenyl)-3, 7-dihydroimidazo

[1, 2-a]pyrazin-3-one. Chem. Lett. 4, 141–144.

Inouye, S. (2008). Cloning, expression, purification and characterization of an isotype of clytin, a

calcium-binding photoprotein from the luminous hydromedusa Clytia gregarium. J. Biochem. 143,

711–717.

Inouye, S., and Sahara, Y. (2009). Expression and purification of the calcium binding photoprotein

mitrocomin using ZZ-domain as a soluble partner in E. coli cells. Protein Expr. Purif. 66, 52–57.

Inouye, S., Aoyama, S., Miyata, T., Tsuji, F. I., and Sakaki, Y. (1989). Overexpression and purification

of the recombinant Ca2þ-binding protein, apoaequorin. J. Biochem. 105, 473–477.

Ishii, K., Hirose, K., Iino, M. (2006). Ca2+ shuttling between endoplasmic reticulum and mitochondria

underlying Ca2+ oscillations. EMBO Rep. 7, 390–396.

Jones, K., Hibbert, F., and Keenan, M. (1999). Glowing jellyfish, luminescence and a molecule called

coelenterazine. Trends Biotechnol. 17, 477–481.

Jung, D. W., Bradshaw, P. C., Litsky, M., and PfeiVer, D. R. (2004). Ca2þ transport in mitochondria

from yeast expressing recombinant aequorin. Anal. Biochem. 324, 258–268.

Karplus, E. (2006). Advances in instrumentation for detecting low-level bioluminescence and fluores-

cence. In ‘‘Photoproteins in Bioanalysis,’’ (S. Daunert, and S. Deo, eds.), pp. 199–223. Wiley-VCH,

Weinheim, Germany.

Kendall, J. M., Badminton, M. N., Dormer, R. L., and Campbell, A. K. (1994). Changes in free calcium

in the endoplasmic reticulum of living cells detected using targeted aequorin. Anal. Biochem. 221,

173–181.


Kennedy, E. D., Rizzuto, R., Theler, J. M., Pralong, W. F., Bastianutto, C., Pozzan, T., and

Wollheim, C. B. (1996). Glucose-stimulated insulin secretion correlates with changes in mitochondri-

al and cytosolic Ca2þ in aequorin-expressing INS-1 cells. J. Clin. Invest. 98, 2524–2538.

Kiegle, E., Moore, C. A., HasseloV, J., Tester, M. A., and Knight, M. R. (2000). Cell-type-specific

calcium responses to drought, salt and cold in the Arabidopsis root. Plant J. 23, 267–278.

Kishi, Y., Tanino, H., and Goto, T. (1972). The structure confirmation of the light-emitting moiety of

bioluminescent jellyfish Aequoria. Tetrahedron Lett. 27, 2747–2748.

Knight, H., and Knight, M. R. (2000). Imaging spatial and cellular characteristics of low temperature

calcium signature after cold acclimation in Arabidopsis. J. Exp. Bot. 51, 1679–1686.

Knight, M. R., Campbell, A. K., Smith, S. M., and Trewavas, A. J. (1991a). Transgenic plant aequorin

reports the eVects of touch and cold-shock and elicitors on cytoplasmic calcium. Nature 352,

524–526.

Knight, M. R., Campbell, A. K., Smith, S. M., and Trewavas, A. J. (1991b). Recombinant aequorin as a

probe for cytosolic free Ca2þ in Escherichia coli. FEBS Lett. 282, 405–408.

Knight, M. R., Knight, H., and Watkins, N. J. (1995). Calcium and the generation of plant form. Phil.

Trans. R. Soc. Lond. B 350, 83–86.

Knight, H., Trewavas, A. J., and Knight, M. R. (1996). Cold calcium signaling in Arabidopsis involves

two cellular pools and a change in calcium signature after acclimation. Plant Cell 8, 489–503.

Lapington, J. S. (2004). A comparison of readout techniques for high-resolution imaging with micro-

channel plate detectors. Nucl. Instrum. Methods Phys. Res. A 525, 361–365.

Lasorsa, F. M., Pinton, P., Palmieri, L., Scarcia, P., Rottensteiner, H., Rizzuto, R., and Palmieri, F.

(2008). Peroxisomes as novel players in cell calcium homeostasis. J. Biol. Chem. 283, 15300–15308.

Lawrie, A. M., Rizzuto, R., Pozzan, T., and Simpson, A. W. M. (1996). A role for calcium influx in the

regulation of mitochondrial calcium in endothelial cells. J. Biol. Chem. 271, 10753–10759.

Maechler, P., Kennedy, E. D., Sebo, E., Valeva, A., Pozzan, T., and Wollheim, C. B. (1999). Secreta-

gogues modulate the calcium concentration in the endoplasmic reticulum of insulin-secreting cells.

Studies in aequorin-expressing intact and permeabilized ins-1 cells. J. Biol. Chem. 274, 12583–12592.

Manjarres, I. M., Chamero, P., Domingo, B., Molina, F., Llopis, J., Alonso, M. T., and Garcıa-

Sancho, J. (2008). Red and green aequorins for simultaneous monitoring of Ca2þ signals from two

diVerent organelles. Pflugers Arch. Eur. J. Physiol. 455, 961–970.

Marsault, R., Murgia, M., Pozzan, T., and Rizzuto, R. (1997). Domains of high Ca2þ beneath the

plasma membrane of living A7r5 cells. EMBO J. 16, 1575–1581.

Martin, J. R., Rogers, K. L., Chagneau, C., and Brulet, P. (2007). In vivo bioluminescence imaging of

Ca2þ signalling in the brain of Drosophila. PLoS ONE 2(3), e275.

McCombs, J. E., and Palmer, A. E. (2008). Measuring calcium dynamics in living cells with genetically

encodable calcium indicators. Methods 46, 152–159.

Miller, A. L., Karplus, E., and Jaffe, L. F. (1994). Use of aequorin for [Ca2+]i imaging. Chapter 13.

In: Methods in Cell Biology Volume 40: A Practical Guide to the Study of Ca2+ in Living Cells

(R. Nuccitelli, ed.), pp. 305–338. Academic Press.

Missiaen, L., Van Acker, K., Van Baelen, K., Raeymaekers, L., Wuytack, F., Parys, J. B., De

Smedt, H., Vanoevelen, J., Dode, L., Rizzuto, R., and Callewaert, G. (2004). Calcium release from

the Golgi apparatus and the endoplasmic reticulum in HeLa cells stably expressing targeted aequorin

to these compartments. Cell Calcium 36, 479–487.

Mitchell, K. J., Pinton, P., Varadi, A., Tacchetti, C., Ainscow, E. K., Pozzan, T., Rizzuto, R., and

Rutter, G. A. (2001). Dense core secretory vesicles revealed as a dynamic Ca2þ store in neuroendo-

crine cells with a vesicle-associated membrane protein aequorin chimaera. J. Cell Biol. 155, 41–51.

Mithofer, A., Ebel, J., Bhagwat, A. A., Boller, T., and Neuhaus-Url, G. (1999). Transgenic aequorin

monitors cytosolic calcium transients in soybean cells challenged with b-glucan or chitin elicitors.

Planta 207, 566–574.



882–887.


Mohammed, J. S., Caicedo, H., Fall, C. P., and Eddington, D. T. (2007). Brain slice stimulation using a

microfluidic network and standard perfusion chamber. J. Vis. Exp. 8, 302.

Montero, M., Brini, M., Marsault, R., Alvarez, J., Sitia, R., Pozzan, T., and Rizzuto, R. (1995).

Monitoring dynamic changes in free Ca2þ concentration in the endoplasmic reticulum of intact

cells. EMBO J. 14, 5467–5475.

Montero, M., Avarez, J., Scheenen, W. J. J., Rizzuto, R., Meldolesi, J., and Pozzan, T. (1997). Ca2þ

homeostasis in the endoplasmic reticulum: Coexistence of high and low [Ca2þ] subcompartments in

intact HeLa cells. J. Cell Biol. 139, 601–611.

Montero, M., Alonso, M. T., Carnicero, E., Cuchillo-Ibanez, I., Albillos, A., Garcıa, A. G., Garcıa-

Sancho, J., and Alvarez, J. (2000). ChromaYn-cell stimulation triggers fast millimolar mitochondrial

Ca2þ transients that modulate secretion. Nat. Cell Biol. 2, 57–61.

Moreno, A., Lobaton, C. D., SantoDomingo, J., Vay, L., Hernandez-SanMiguel, E., Rizzuto, R.,

Montero, M., and Alvarez, J. (2005). Calcium dynamics in catecholamine-containing secretory

vesicles. Cell Calcium 37, 555–564.



Nagel-Volkmann, J., Plieth, C., Becker, D., Luthen, H., and DorZing, K. (2009). Cold-induced

cytosolic free calcium ion concentration changes in wheat. J. Plant Physiol. doi:10.1016/j.

jplph.2009.05.002.

Nakai, J., Ohkura, M., and Imoto, K. (2001). A high signal-to-noise Ca2þ probe composed of a single

green fluorescent protein. Nat. Biotechnol. 19, 137–141.

Nakahashi, Y., Nelson, E., Fagan, K., Gonzales, E., Guillou, J.-L., and Cooper, D. M. F. (1997).

Construction of a full-length Ca2+-sensitive adenylyl cyclase/aequorin chimera. J. Biol. Chem. 272,

18093–18097.

Nakajima-Shimada, J., Iida, H., Tsuji, F. I., and Anraku, Y. (1991). Monitoring of intracellular calcium

in Saccharomyces cerevisiae with an apoaequorin cDNA expression system. Proc. Natl. Acad. Sci.

USA 88, 6878–6882.

Nakazaki, M., Ishihara, H., Kakei, M., Inukai, K., Asano, T., Miyazaki, J. I., Tanaka, H., Kikuchi, M.,

Yada, T., and Oka, Y. (1998). Repetitive mitochondrial Ca2þ signals synchronize with cytosolic Ca2þ

oscillations in the pancreatic beta-cell line, MIN6. Diabetologia 41, 279–286.

Narayanan, B., Islam, M. N., Bartelt, D., and Ochs, R. S. (2003). A direct mass-action mechanism

explains capacitative calcium entry in Jurkat and skeletal L6 muscle cells. J. Biol. Chem. 278,

44188–44196.

Nelson, G., Kozlova-Zwinderman, O., Collis, A. J., Knight, M. R., Fincham, J. R. S., Stanger, C. P.,

Renwick, A., Hessing, J. G. M., Punt, P. J., van den Hondel, C. A. M. J. J., and Read, N. D. (2004).

Calcium measurement in living filamentous fungi expressing codon-optimized aequorin.Mol. Micro-

biol. 52, 1437–1450.

Nunez, L., Senovilla, L., Sanz-Blasco, S., Chamero, P., Alonso, M. T., Villalobos, C., and Garcıa-

Sancho, J. (2007). Bioluminescence imaging of mitochondrial Ca2þ dynamics in soma and neurites of

individual adult mouse sympathetic neurons. J. Physiol. 580, 385–395.

Ohkura, M., Matsuzaki, M., Kasai, H., Imoto, K., and Nakai, J. (2005). Genetically encoded bright

Ca2þ probe applicable for dynamic Ca2þ imaging of dendritic spines. Anal. Chem. 77, 5861–5869.

Padua, R. A., Baron, K. T., Thyagarajan, B., Campbell, C., and Thayer, S. A. (1998). Reduced Ca2þ

uptake by mitochondria in pyruvate dehydrogenase-deficient human diploid fibroblasts. Am.

J. Physiol. 274, C615–C622.

Pinton, P., Pozzan, T., and Rizzuto, R. (1998). The Golgi apparatus is an inositol 1, 4, 5-trisphosphate-

sensitive Ca2þ store, with functional properties distinct from those of the endoplasmic reticulum.

EMBO J. 17, 5298–5308.

Pinton, P., Ferrari, D., Magalhaes, P., Schulze-OsthoV, K., Di Virgilio, F., Pozzan, T., and Rizzuto, R.

(2000). Reduced loading of intracellular Ca2þ stores and downregulation of capacitative Ca2þ influx

in Bcl-2-overexpressing cells. J. Cell Biol. 148, 857–862.


Pouli, A. E., Karagenc, N., Wasmeier, C., Hutton, J. C., Bright, N., Arden, S., Schofield, J. G., and

Rutter, G. A. (1998). A phogrin-aequorin chimaera to image free Ca2þ in the vicinity of secretory

granules. Biochem. J. 330, 1399–1404.

Prasher, D., McCann, R. O., and Cormier, M. J. (1985). Cloning and expression of the cDNA coding for

aequorin, a bioluminescent calcium-binding protein.Biochem. Biophys. Res. Commun. 126, 1259–1268.

Quesada, I., Villalobos, C., Nunez, L., Chamero, P., Alonso, M. T., Nadal, A., and Garcıa-Sancho, J.

(2008). Glucose induces synchronous mitochondrial calcium oscillations in intact pancreatic islets.

Cell Calcium 43, 39–47.

Rembold, C. M., Kendall, J. M., and Campbell, A. K. (1997). Measurement of changes in sarcoplasmic

reticulum [Ca2þ] in rat tail artery with targeted apoaequorin delivered by an adenoviral vector. Cell

Calcium 21, 69–79.

Ridgway, E. B., and Ashley, C. C. (1967). Calcium transients in single muscle fibres. Biochem. Biophys.

Res. Commun. 29, 229–234.

Ridgway, E. B., Gilkey, J. C., and JaVe, L. F. (1977). Free calcium increases explosively in activating

medaka eggs. Proc. Natl. Acad. Sci. USA 74, 623–627.

Rizzuto, R., and Pozzan, T. (2006). Microdomains of intracellular Ca2þ: Molecular determinants and

functional consequences. Physiol. Rev. 86, 369–408.

Rizzuto, R., Simpson, A. W. M., Brini, M., and Pozzan, T. (1992). Rapid changes of mitochondrial

Ca2þ revealed by specifically targeted recombinant aequorin. Nature 358, 325–327.

Robert, V., De Giorgi, F., Massimino, M. L., Cantini, M., and Pozzan, T. (1998). Direct monitoring of

the calcium concentration in the sarcoplasmic and endoplasmic reticulum of skeletal muscle myo-

tubes. J. Biol. Chem. 273, 30372–30378.

Robert, V., Gurlini, P., Tosello, V., Nagai, T., Miyawaki, A., Di Lisa, F., and Pozzan, T. (2001b). Beat-

to-beat oscillations of mitochondrial [Ca2þ] in cardiac cells. EMBO J. 20, 4998–5007.

Robert, V., Massimino, M. L., Tosello, V., Marsault, R., Cantini, M., Sorrentino, V., and Pozzan, T.

(2001a). Alteration in calcium handling at the subcellular level in mdx myotubes. J. Biol. Chem. 276,

4647–4651.

Rogers, K. L., Stinnakre, J., Agulhon, C., Jublot, D., Shorte, S. L., Kremer, E. J., and Brulet, P. (2005).

Visualization of local Ca2þ dynamics with genetically encoded bioluminescent reporters. Eur.

J. Neurosci. 21, 597–610.

Rogers, K. L., Picaud, S., Roncali, E., Boisgard, R., Colasante, C., Stinnakre, J., Tavitian, B., and

Brulet, P. (2007). Non-invasive in vivo imaging of calcium signaling in mice. PLoS ONE 2(10), e974.

Rogers, K. L., Martin, J. R., Renaud, O., Karplus, E., Nicola, M.-A., Nguyen, M., Picaud, S.,

Shorte, S. L., and Brulet, P. (2008). Electron-multiplying charge-coupled detector-based biolumines-

cence recording of single-cell Ca2þ. J. Biomed. Opt. 13, 031211, doi:10.1117/1.2937236.

Roncali, E., Savinaud, M., Levrey, O., Rogers, K. L., Maitrejean, S., and Tavitian, B. (2008).

New device for real-time bioluminescent imaging in moving rodents. J. Biomed. Opt. 13, 054035-1-

054035-10.

Rosay, P., Davies, S. A., Yu, Y., Sozen, M. A., Kaiser, K., and Dow, J. A. (1997). Cell-type specific

calcium signalling in a Drosophila epithelium. J. Cell Sci. 110, 1683–1692.

Russell, A. J., Knight, M. R., Cove, D. J., Knight, C. D., Trewavas, A. J., and Wang, T. L. (1996).

The moss, Physcomitrella patens, transformed with apoaequorin cDNA responds to cold shock,

mechanical perturbation and pH with transient increases in cytoplasmic calcium. Transgenic Res. 5,

167–170.

Rust, M. J., Bates, M., and Zhuang, X. (2006). Sub-diVraction-limit imaging by stochastic optical

reconstruction microscopy (STORM). Nat. Methods 3, 793–795.

Rutter, G. A., Burnett, P., Rizzuto, R., Brini, M., Murgia, M., Pozzan, T., Tavare, J. M., and

Denton, R. M. (1996). Subcellular imaging of intramitochondrial Ca2þ with recombinant targeted

aequorin: Significance for the regulation of pyruvate dehydrogenase activity. Proc. Natl. Acad. Sci.

USA 93, 5489–5494.

Rutter, G. A., Tsuboi, T., and Ravier, M. A. (2006). Ca2þ microdomains and the control of insulin

secretion. Cell Calcium 40, 539–551.


Sai, J., and Johnson, C. H. (2002). Dark-stimulated calcium ion fluxes in the chloroplast stroma and

cytosol. Plant Cell 14, 1279–1291.

Sanchez-Bueno, A., Yoshida, R., and Tsuji, F. I. (1996). Regeneration and luminescence of aequorin in

Chinese hamster ovary cells transformed with cDNA for apoaequorin. Int. J. Biochem. Cell Biol. 28,

1045–1049.

SantoDomingo, J., Vay, L., Camacho, M., Hernandez-SanMiguel, E., Fonteriz, R. I., Lobaton, C. D.,

Montero, M., Moreno, A., and Alvarez, J. (2008). Calcium dynamics in bovine adrenal medulla

chromaYn cell secretory granules. Eur. J. Neurosci. 28, 1265–1274.

Saran, S., Nakao, H., Tasaka, M., Iida, H., Tsuji, F. I., Nanjundiah, V., and Takeuchi, I. (1994).

Intracellular free calcium level and its response to cAMP stimulation in developingDictyostelium cells

transformed with jellyfish apoaequorin cDNA. FEBS Lett. 337, 43–47.

Sedbrook, J. C., Kronebusch, P. J., Borisy, G. G., Trewavas, A. J., and Masson, P. H. (1996).

Transgenic AEQUORIN reveals organ-specific cytosolic Ca2þ responses to anoxia in Arabidopsis

thaliana seedlings. Plant Physiol. 111, 243–257.

Shaw, B. D., Kozlova, O., Read, N. D., Turgeon, B. G., and Hoch, H. C. (2001). Expression of

recombinant aequorin as an intracellular calcium reporter in the phytopathogenic fungus Phyllosticta

ampelicida. Fungal Genet. Biol. 34, 207–215.

Sheu, Y. A., Kricka, L. J., and Pritchett, D. B. (1993). Measurement of intracellular calcium using

bioluminescent aequorin expressed in human cells. Anal. Biochem. 209, 343–347.

Shimomura, O. (1991). Preparation and handling of aequorin solutions for the measurement of cellular

Ca2þ. Cell Calcium 12, 635–643.

Shimomura, O. (1997). Membrane permeability of coelenterazine analogues measured with fish eggs.

Biochem. J. 326, 297–298.

Shimomura, O., Musicki, B., and Kishi, Y. (1989). Semi-synthetic aequorins with improved sensitivity

to Ca2þ ions. Biochem. J. 261, 913–920.

Siegmund, O., Vallerga, J., Jelinsky, P., Michalet, X., andWeiss, S. (2005). Cross delay line detectors for

high time resolution astronomical polarimetry and biological fluorescence imaging. IEEE Nucl. Sci.

Symp. Conf. Rec. 1, 448–452.

Souslova, E. A., Belousov, V. V., Lock, J. G., Stromblad, S., Kasparov, S., Bolshakov, A. P.,

Pinelis, V. G., Labas, Y. A., Lukyanov, S., Mayr, L. M., and Chudakov, D. M. (2007). Single

fluorescent protein-based Ca2þ sensors with increased dynamic range. BMC Biotechnol. 7, 37.

Speksnijder, J. E., Sardet, C., and JaVe, L. F. (1990). The activation wave of calcium in the ascidian egg

and its role in ooplasmic segregation. J. Cell Biol. 110, 1589–1598.

Stoppini, L., Buchs, P. A., and Muller, D. (1991). A simple method for organotypic cultures of nervous

tissue. J. Neurosci. Methods 37, 173–182.

Stepanyuk, G. A., Golz, S., Markova, S. V., Frank, L. A., Lee, J., and Vysotski, E. S. (2005).

Interchange of aequorin and obelin bioluminescence color is determined by substitution of one active

site residue of each photoprotein. FEBS Lett. 579, 1008–1014.

Strayle, J., Pozzan, T., and Rudolph, H. K. (1999). Steady-state free Ca2þ in the yeast endoplasmic

reticulum reaches only 10 mM and is mainly controlled by the secretory pathway pump pmr1. EMBO

J. 18, 4733–4743.

Szado, T., Kuo, K. H., Bernard-Helary, K., Poburko, D., Lee, C. H., Seow, C., Ruegg, U. T., and van

Breemen, C. (2003). Agonist-induced mitochondrial Ca2þ transients in smooth muscle. FASEB J. 17,

28–37.

Tenenbaum, L., Chtarto, A., Lehtonen, E., Velu, T., Brotchi, J., and Levivier, M. (2004). Recombinant

AAV-mediated gene delivery to the central nervous system. J. Gene Med. 6, S212–S222.

Torfs, H., Poels, J., Detheux, M., Dupriez, V., Van Loy, T., Vercammen, L., Vassart, G.,

Parmentier, M., and Vanden Broeck, J. (2002). Recombinant aequorin as a reporter for receptor-

mediated changes of intracellular Ca2þ-levels in Drosophila S2 cells. Invert. Neurosci. 4, 119–124.

Torrecilla, I., Leganes, F., Bonilla, I., and Fernandez-Pinas, F. (2000). Use of recombinant aequorin to

study calcium homeostasis and monitor calcium transients in response to heat and cold shock in

cyanobacteria. Plant Physiol. 123, 161–176.


Tsien, R. Y. (1998). The green fluorescent protein. Annu. Rev. Biochem. 67, 509–544.

Tsuruwaka, Y., Konishi, T., Miyawaki, A., and Takagi, M. (2007). Real-time monitoring of dynamic

intracellular Ca2þ movement during early embryogenesis through expression of yellow cameleon.

Zebrafish 4, 253–260.

Villalobos, C., Nunez, L., Montero, M., Garcıa, A. G., Alonso, M. T., Chamero, P., Alvarez, J., and

Garcıa-Sancho, J. (2002). Redistribution of Ca2þ among cytosol and organella during stimulation of

bovine chromaYn cells. FASEB J. 16, 343–353.

Villalobos, C., Nadal, A., Nunez, L., Quesada, I., Chamero, P., Alonso, M. T., and Garcıa-Sancho, J.

(2005). Bioluminescence imaging of nuclear calcium oscillations in intact pancreatic islets of Langer-

hans from the mouse. Cell Calcium 38, 131–139.

Wang, X., Wan, H., Korzh, V., and Gong, Z. (2000). Use of an IRES bicistronic construct to trace

expression of exogenously introduced mRNA in zebrafish embryos. Biotechniques 29, 814–816.

Webb, S. E., Lee, K. W., Karplus, E., andMiller, A. L. (1997). Localized calcium transients accompany

furrow positioning, propagation and deepening during the early cleavage period of zebrafish em-

bryos. Dev. Biol. 192, 78–92.

Whitaker, M. (2008). Calcium signalling in early embryos. Philos. Trans. R. Soc. Lond. B 363,

1401–1418.

Williams, S., Mustoe, T., Mulcahy, T., GriYths, M., Simpson, D., Antoniou, M., Irvine, A.,

Mountain, A., and Crombie, R. (2005). CpG-island fragments from theHNRPA2B1/CBX3 genomic

locus reduce silencing and enhance transgene expression from the hCMV promoter/enhancer in

mammalian cells. BMC Biotechnol. 5, 17.

Wiltgen, S. W., Smith, I. F., Beri, N., and Parker, I. (2009). Super-resolution imaging of Ca2þ flux

through IP3Rs with millisecond temporal resolution and nanometer spatial resolution. Biophys. J. 96,

637a.

Wood, N. T., Haley, A., Viry-Moussaıd, M., Johnson, C. H., van der Luit, A. H., and Trewavas, A. J.

(2001). The calcium rhythms of diVerent cell types oscillate with diVerent circadian phases. Plant

Physiol. 125, 787–796.

Xiong, Z. H., and Ruben, L. (1996). Nuclear calcium flux in Trypanosoma brucei can be quantified with

targeted aequorin. Mol. Biochem. Parasitol. 83, 57–67.

Xiong, Z. H., Ridgley, E. L., Enis, D., Olness, F., and Ruben, L. (1997). Selective transfer of calcium

from an acidic compartment to the mitochondrion of Trypanosoma brucei. Measurements with

targeted aequorins. J. Biol. Chem. 272, 31022–31028.

Yamano, K., Mori, K., Nakano, R., Kusunoki, M., Inoue, M., and Satoh, M. (2007). Identification of

the functional expression of adenosine A3 receptor in pancreas using transgenic mice expressing

jellyfish apoaequorin. Transgenic Res. 16, 429–435.

Yoshida, R., Sanchez-Bueno, A., and Tsuji, F. I. (1998). Tracing the in vivo distribution and dynamics

of cancer cells in mice by luminescence of aequorin-expressing transformants. Biochem. Biophys. Res.

Commun. 242, 659–662.

Zorov, D. B., Kobrinsky, E., Juhaszova, M., and Sollott, S. J. (2004). Examining intracellular organelle

function using fluorescent probes. From animacules to quantum dots. Circ. Res. 95, 239–252.

Zou, J., Hofer, A. M., Lurtz, M. M., Gadda, G., Ellis, A. L., Chen, N., Huang, Y., Holder, A., Ye, Y.,

Louis, C. F., Welshhans, K., Rehder, V., and Yang, J. J. (2007). Developing sensors for real-time

measurement of high Ca2þ concentrations. Biochemistry 46, 12275–12288.

INDEX

1,2-Bis(o-aminophenoxy)ethane-N,N,N’,

N’-tetraacetic acid (BAPTA), 17–18

ionic strength dependent, 5, 7

pH dependent, 4

structural formulas for, 5

temperature, apparent aYnity, 6, 7

2P excitation microscopy

Fluo-based dyes, 248, 249

Rhod-dyes, 249, 250

A

Acetoxymethyl (AM) ester, 120

Aequoria victoria, 154

Aequorins

apoaequorin expression, 268, 277–279

bioluminescent GETCRs, 265

BRET protein complex, 265

coelenterazine, 283–284

expression of GFP-Apoaequorin, 279–283

luminescence

bioluminescence imaging, 287

ca 2þdetection device, 286

detectors types, 267

EMCCD detector, 286, 288

equipment, 284

factors, 290

intensifier-based detector, 290

photon counting photomultiplier, 284

schematic representations, 285

single photon imaging, 290

two-channel luminometer, 286

mitochondrial target, 160

strategies, 265

AM. See Acetoxymethyl (AM) ester

Apoaequorin

BRET complexes

cytosol cell culture, 269–273

diverse range species, 266

materials, 277

methods

in vivo reconstitution, 278

p�-KS-aeq-IRES-EGFP plasmid

preparation, 277–278

transgenic zebrafish generation, 278–279

Azid-1, 33–35

B

Bioluminescence

EMCCD detector, 286

fluorescent GET-CRs, 264

GETCRs, 265

optical systems, 287

single photon detection, 286

vs. EMCCD and an RA-IPD photon imaging

system, 291

Bioluminescence resonance energy transfer

(BRET), 255

aequorins, 265

cytosol cell culture, 269–273

diverse range species, 266

Blue fluorescent protein (BFP), 155

BRET. See Bioluminescence resonance

energy transfer

C

Ca2þ binding compounds, 104, 105

Ca2þ buVers

association constant (KCa), 7, 10

BAPTA family, 5–7

basic steps, solution preparation, 16–18

dissociation constant (Kd), 3

EGTA, 4–6

fluorescent indicators, 3–4

ionic strength corrections, 11–12

measurement, free [Ca2þ], 12–13Michaelis-Menten form, 9

potential complications, 18–19

proton activity coeYcient, 12

software programs, 19–23

spreadsheet for calibration calculations, 13–16

stability constants, 3

temperature corrections, 10–11

Ca2þ flux measurements, 99, 100, 108

301

302 Index

Ca2þ in mitochondria, measurement

cytosolic and mitochondrial Ca2þ, 146Rhod-2 indicator, 143–145

Rhod-2, cytosol, 145–146

Ca2þ, manipulation

buVering changes, 130–132

divalent Cation Ionophores, 129–130

extracellular buVers

BAPTA, 126–127

EGTA, 126–127

lowering Extracellular [Ca2þ], 127–129Ca2þ-selective electrodes

dissociation constant, measurements of, 70

dynamic range of, 69

ETH 129, 86

key advantage, 68

microelectrodes (MEs)

bath calibration, 81

Ca2þ-selective ligand, preparation and use

of, 78–80

calibration procedure, 81–82

double-barreled, 80–81

electrolyte filling of, 78

ETH 1001, ionophore, 82

extracellular [Ca2þ], measurement, 83–85

glass tubing preparation, 76

microelectrode pulling and silanization, 77

solution perfusion, 81

troubleshooting, 85–86

minielectrode

application of, 75–76

electrode potential of, 73

inhomogeneities, eVect of, 75

lifetime of, 72

resistance of, 72

response times of, 73–74

selective ligand, preparation and use, 71–72

storage of, 75

Nicolski-Eisenman equation, 68

Ca2þ-selective liquid membranes, coeVcients, 96,

97

Ca2þselective microelectrodes (CaSMs)

construction

microelectrode, 94

micropipette fabrication, 93

silanization, 93

properties

response time, 97–98

response to Ion Activity, 94–96

selectivity, 96

spatial resolution, 96–97

self referencing

calculation of flux, 103

correction for Ca2þ buVering, 104

diVerential concentration determination,

101–102

diVerential concentration measurement,

98–101

measurement of voltage gradients, 104–105

positional artifacts, 105–107

Caged Ca2þ chelators. See Photolabile Ca2þ

chelators

Calcium release-activated channels (CRAC),

185, 194–195

Cameleon, genetically encoded calcium sensors

CFP/citrine couple, 159

ECFP/EYFP-based cameleons, 161

F46L mutation, 160

FIP-CBsm, 156

FRET, 156

mechanism, 161–163

myosin light chain kinase, 155

YC concentrations, 159

YC2.12 fluorescence, 160

YCX.60, 161

CaSM. See Ca2þselective microelectrodes

CaSMs, construction

microelectrode construction, 94

micropipette fabrication, 93

silanization, 93

CaSMs, properties

Response Time, 97–98

Response to Ion Activity, 94–96

Selectivity, 96

Spatial Resolution, 96–97

CaSMs, self referencing

calculation of flux, 103

correction for Ca2þ buVering, 104

diVerential concentration determination,

101–102

diVerential concentration measurement, 98–101

measurement of voltage gradients, 104–105

positional artifacts, 105–107

Cell-attached patch recordings, 192–193

Coelenterazine, 283–284

Confocal and multiphoton imaging

absorbance/quantum yield, 246–247

advantages and disadvantages, 241–242

calibration of single wavelength dyes, 251–252

conversion of increments, 251

fluorescence lifetime, 247

Fmax values estimation, 252–253

Fmin estimation, 253

Forster resonance energy transfer microscopy

acceptor photobleaching, 236

cameleon, 235, 237

donor quenching, 236

eYciency, 246, 247

Index 303

FRET-FLIM approach, 236

GFP, 235–236

pericams, 237

intrinsic and dye fluorescence, 253–255

laser scanning confocal microscopy

2D frame scan, 231

Jablonski diagram, 232, 233

signal-to-noise, 231

Stokes shift, 232

limitations in speed, 230

multimodal and multiple fluorophore

Di-8-ANEPPS, 256, 257

excitation-contraction coupling, 255

Fura dyes, 258

membrane voltage, 256

overlapping excitation spectra, 256

RH-237 emission spectrum, 257

multiphoton excitation laser scanning

microscopy

2P excitation microscopy, 242, 243

biophysical perspectives, 242

high repetition rates, 245

IR light, 244

NA objectives, 228

parallel scanning confocal systems, 238

programmable matrix microscopy

digital micromirror device, 239

filtering patterns, 239

liquid crystal display, 240

PSF, 229

signal detection system, 227, 228

spatial resolution, 229

spectral shift, 247

spinning disk confocal microscopy, 238–239

total internal reflection fluorescence

microscopy

evanescent wave, 234

refractive index, 233, 234

use of dyes

2P excitation microscopy, 248–250

single-photon confocal microscopy, 248

CRAC. See Calcium release-activated channels

Cytoplasmic [Ca2þ]i, regulation of, 57

D

Diazo compounds, Ca2þ chelators

chemical properties, 39–41

photolysis, eVects of, 41–42

Dibromo-BAPTA (Br2-BAPTA), 4, 11

pH dependent, apparent aYnities, 4


temperature and ionic strength, apparent Ca2þ

aYnity, 5

Digital micromirror device (DMD), 239

DM-nitrophen, Ca2þ chelators

[Ca2þ]i changes, 37–39absorbance of, 36

Ca2þ-and Mg2þ-aYnities of, 36

caged Mg2þ chelator, 36

kinetic behavior of, 38

quantum eYciency, 35

structure of and reaction scheme, 35

Dual-wavelength ratiometric dyes, 114, 115,

118–120

E

Electrode calibration curves, 15, 16

Electron multiplying charge-coupled devices

(EMCCD)

bioluminescence detection, 286

computer-controlled EMCCD, 289

limitations, 289

single photon events, 289

vs. RA-IPD photon imaging system, 290, 291

EMCCD. See Electron multiplying charge-

coupled devices

Enhanced yellow fluorescent protein (EYFP)

circular permutation, 163–164

Ethylene glycol bis(�-aminoethylether)-N,N,N’,

N’-tetraacetic acid (EGTA), 17–18

ionic strength, apparent aYnity, 6

pH dependent, 4–5


temperature, apparent aYnity, 6

Excel spreadsheet, Ca2þ calibration buVers, 14, 15

Excitation spectra

Fluo-3, 117

Fura-2, 117

F

Fluorescence lifetime imaging (FLIM), 236

Fluorescent Ca2þ Indicators

dual-wavelength ratiometric dyes, 114, 115,

118–120

single-wavelength nonratiometric dyes, 114,

115, 118–120

Fluorescent Ca2þ Indicators, properties, 115

Forster resonance energy transfer (FRET)

microscopy

acceptor photobleaching, 236

advantage and disadvantage, 236

cameleon, 235, 237

donor quenching, 236

eYciency, 246, 247

304 Index

Forster resonance energy transfer (FRET)

microscopy (cont.)

FRET-FLIM approach, 236

GFP, 235–236

pericams, 237

Fura-2, 33

G

Genetically encoded calcium sensors

Aequoria victoria, 154

Cameleon family

CFP/citrine couple, 159

ECFP/EYFP-based cameleons, 161

F46L mutation, 160

origin, 155, 157

sensor mechanism, 161–163

YC concentrations, 159

YC2.12 fluorescence, 160

YCX.60, 161

yellow fluorescent protein, 159

Camgaroos, 163–164

cases 12 and 16, 167

GCaMPs, 165–167

green fluorescent protein, 154

pericam, 164–165

subcellular locations

endoplasmic reticulum, 168

golgi, 169

mitochondria, 168

peroxisome, 169

plasma membrane, 169

tissue-specific expression

comparative studies, 174–176

GCaMP, 172–173

inverse pericam, 171–172

TN-L15, TN-XL, 173, 174

TN-XXL, 174

YC2.1, 169–170

YC3.3er (citrine-based sensor), 171

uses, 176–177

GFP-apoaequorin

recombinant viral vectors

materials, 281

methods, 281–282

transgenic mice expressing mitochondrially

materials, 282

methods, 283

transient and stable transfection

materials, 279–280

methods, 280

Green fluorescent protein, 154

H

Human cytomregalovirus (HCMV), 279

I

Icrac. See Calcium release-activated currents

Imaging photon detector (IPD), 284

Indicator fluorescence signal, conversion

nonratiometric fluorescent indicator,

calibration, 133–134

ratiometric fluorescent indicator, calibration,

134–138

Indicators, loading in the cells, 121–122

aqueous Solubility of AM Esters, 121

dye compartmentalization

assessing extent of compartmentalization,

122–124

minimizing compartmentalization,

121–122

dye leakage, 124–125

procedure, 124–125

Inositol 1,4,5-trisphosphate receptors (IP3R)

ATP ligands, 198

behavior analysis, 203

Ca 2þ signals, 190

Ca2þ ligands, 198

cell analysis, 195–196

current–voltage relationship, 204

cytopasm-out configuration, 202

DT40 cell expression, 193

electrical recording, 193

endoplasmic reticulum, 190

inner nuclear membrane expression, 198

intrinsic pore open, 190

IP3, 198

nuclear path-clamp recording, 194

single-channel recording, 191, 193

whole-cell recordings, 192

Intracellular Ca2þ

confocal and multiphoton imaging

2P excitation microscopy, 248–250

advantages and disadvantages, 241–242

Fmax values estimation, 252–253

Fmin estimation, 253

FRET, 235–237

indicators, 245–247

intrinsic and dye fluorescence, 253–255

limitations in speed, 230

LSCM, 230–233

multimodal and multiple fluorophore,

255–259

Index 305

multiphoton excitation laser scanning

microscopy, 242–245

parallel scanning confocal systems, 238

programmable matrix microscopy, 239–241

single wavelength dyes, 251–252

single-photon confocal microscopy, 248

spinning disk confocal microscopy, 238–239

TIRF, 233–234

patch clamp methods, 185

Intracellular calcium signals

fluorescent Ca2þ Indicators

Ion channel modulation, Ca2þ chelators, 49–52

calcium channels, 51–52

potassium and nonspecific cation channels,

49–51

J


L

Laser scanning confocal microscopy (LSCM)

2D frame scan, 231


signal-to-noise, 231

Stokes shift, 232

Liquid crystal display (LCD), 240

M

MaxChelator, 22

Microelectrodes (MEs), Ca2þ-selectivebath calibration, 81

Ca2þ-selective ligand, preparation and use of,

78–80

calibration procedure, 81–82

double-barreled, 80–81

electrolyte filling of, 78

ETH 1001, ionophore, 82

extracellular [Ca2þ], measurement, 83–85

glass tubing preparation, 76

microelectrode pulling and silanization, 77

solution perfusion, 81

troubleshooting, 85–86

Minielectrode, Ca2þ-selectiveapplication of, 75–76

electrode potential of, 73

inhomogeneities, eVect of, 75

lifetime of, 72

resistance of, 72

response times of, 73–74

selective ligand, preparation and use, 71–72

storage of, 75

Multiphoton excitation laser scanning

microscopy

2P excitation microscopy, 242, 243

biophysical perspectives, 242

high repetition rates, 245

IR light, 244

Muscle contraction, Ca2þ chelators, 52–53

Myosin light chain kinase (MLCK), 155

N

Nipkow spinning disk, 239, 240

Nitr compounds, Ca2þ chelators

[Ca2þ]i changes, cells, 33–34azid-1, 33

BAPTA, 29, 30

fura-2, 33

nitr-5, 30–31

nitr-7, 30–31

nitr-8, 30–31

Nitr-5, 30–31

Nitr-7, 30–31

Nitr-8, 30–31

Nuclear patch-clamp recording

conventional techniques, 193

ER membrane, 195

IP3R, 195–196

methods

asymmetric recording solutions, 201

cytoplasm-out configuration, 202

DT40 Cell culture, 196–197

equipments, 199

nuclei isolation, 197–198

optimal filtering frequency, 201

pipette tip approaches, 201

recording configuration, 201

single channel record analysis, 202–208

solutions, 198–199

P

Patch clamp methods

calcium-selective channels, 184

intracellular calcium, 185

materials, 195

methods

calcium release-activated currents, 194–195

cell-attached patch recordings, 192–193

fire-polishing pipettes, 191

giga-ohm seals, 190–192

306 Index

Patch clamp methods (cont.)

perforated patch recordings, 193–194

rig assemble, 187–188

Sylgard application, 188, 189

principles, 186–187

recombinant channels, 185, 186

Perforated patch recordings, 193–194

Photolabile Ca2þ chelators, 33

biological applications

cytoplasmic [Ca2þ]i, regulation of, 57

filopodial activity, control of, 58

ion channel modulation, 49–52

muscle contraction, 52–53

rate-limiting steps, 58

synaptic function, 53–57

calibration, 45–48

diazo compounds

chemical properties, 39–41

photolysis, eVects of, 41–42

DM-nitrophen

[Ca2þ]i changes, 37–39absorbance of, 36

Ca2þ-and Mg2þ-aYnities of, 36

caged Mg2þ chelator, 36

kinetic behavior of, 38

quantum eYciency, 35

structure of and reaction scheme, 35

introduction into cells, 42–43

light sources, 43–45

nitr compounds

[Ca2þ]i changes, cells, 33–34azid-1, 33

BAPTA, 29, 30

fura-2, 33

nitr-5, 30–31

nitr-7, 30–31

nitr-8, 30–31

properties of, 32

purity and toxicity, 48–49

Photomultiplier tube (PMT), 227, 231

Photon counting photomultiplier (PMT), 284

PMT. See Photon counting photomultiplier

Point spread function (PSF), 229

Programmable matrix microscopy

digital micromirror device, 239

filtering patterns, 239

liquid crystal display, 240

Proton activity coeYcient, 12

R

Ratiometric fluorescent indicators

Fura-2, 117

Indo-1, 117

principle, 114, 115, 118–120

Recombinant viral vectors

materials, 281

methods, 281–282

S

Scatchard plot analysis, 13, 15, 16

Shuttle buVers, 104, 105

Single channel recording

current-amplitude histograms, 203

dwell-time histogram, 207

electrophysiological records, 202

IP3R cation-selectivity, 206

kinetic analyses, 206

maximum interval likelihood method, 207

Mg 2þ, 198mutations, 205

open probability, 208

Sigworth–Sine transformation, 207

stability plot, 206

Single-photon confocal microscopy, 248

Single-wavelength nonratiometric dyes, 114, 115,

118–120

SOCE. See Store-operated calcium entry

Software programs, Ca2þ buVers, 19–20

accuracy, 21

ideal software criteria, 20

javascript web versions, 22–23

MaxChelator, 22

use and adaptability, ease of, 21

Store-operated calcium entry (SOCE), 185

Synaptic function, Ca2þ chelators, 53–57

T

Tissue-specific expression

comparative studies, 174–176

GCaMP, 172–173

inverse pericam, 171–172

TN-L15, TN-XL, 173, 174

TN-XXL, 174

YC2.1, 169–170

YC3.3er (citrine-based sensor), 171

Total internal reflection fluorescence (TIRF)

microscopy

evanescent wave, 234

refractive index, 233, 234

Y

Yellow fluorescent protein, 159

VOLUMES INSERIES

Founding Series EditorDAVIDM. PRESCOTT

Volume1 (1964)Methods in Cell Physiology

Edited by David M. Prescott

Volume 2 (1966)Methods in Cell Physiology










Volume 7 (1973)Methods in Cell Biology






307

308 Volumes in Series



Volume 11 (1975)Yeast Cells


Volume 12 (1975)Yeast Cells








Volume 16 (1977)Chromatin and Chromosomal Protein Research I

Edited by Gary Stein, Janet Stein, and Lewis J. Kleinsmith

Volume 17 (1978)Chromatin and Chromosomal Protein Research II


Volume 18 (1978)Chromatin and Chromosomal Protein Research III


Volume 19 (1978)Chromatin and Chromosomal Protein Research IV




Volumes in Series 309

Advisory Board ChairmanKEITHR. PORTER

Volume 21A (1980)Normal Human Tissue and Cell Culture, Part A: Respiratory, Cardiovascular, and

Integumentary Systems

Edited by Curtis C. Harris, Benjamin F. Trump, and Gary D. Stoner

Volume 21B (1980)Normal Human Tissue and Cell Culture, Part B: Endocrine, Urogenital, and

Gastrointestinal Systems

Edited by Curtis C. Harris, Benjamin F. Trump, and Gray D. Stoner

Volume 22 (1981)Three-Dimensional Ultrastructure in Biology

Edited by James N. Turner

Volume 23 (1981)Basic Mechanisms of Cellular Secretion

Edited by Arthur R. Hand and Constance Oliver

Volume 24 (1982)The Cytoskeleton, Part A: Cytoskeletal Proteins, Isolation and Characterization

Edited by Leslie Wilson

Volume 25 (1982)The Cytoskeleton, Part B: Biological Systems and In Vitro Models

Edited by Leslie Wilson

Volume 26 (1982)Prenatal Diagnosis: Cell Biological Approaches

Edited by Samuel A. Latt and Gretchen J. Darlington

Series EditorLESLIEWILSON

Volume 27 (1986)Echinoderm Gametes and Embryos

Edited by Thomas E. Schroeder

Volume 28 (1987)Dictyostelium discoideum: Molecular Approaches to Cell Biology

Edited by James A. Spudich


Volume 29 (1989)Fluorescence Microscopy of Living Cells in Culture, Part A: Fluorescent Analogs,

Labeling Cells, and Basic Microscopy

Edited by Yu-Li Wang and D. Lansing Taylor

Volume 30 (1989)Fluorescence Microscopy of Living Cells in Culture, Part B: Quantitative

Fluorescence Microscopy—Imaging and Spectroscopy

Edited by D. Lansing Taylor and Yu-Li Wang

Volume 31 (1989)Vesicular Transport, Part A

Edited by Alan M. Tartakoff

Volume 32 (1989)Vesicular Transport, Part B


Volume 33 (1990)Flow Cytometry

Edited by Zbigniew Darzynkiewicz and Harry A. Crissman

Volume 34 (1991)Vectorial Transport of Proteins into and across Membranes


Selected fromVolumes 31, 32, and 34 (1991)Laboratory Methods for Vesicular and Vectorial Transport


Volume 35 (1991)Functional Organization of the Nucleus: A Laboratory Guide

Edited by Barbara A. Hamkalo and Sarah C. R. Elgin

Volume 36 (1991)Xenopus laevis: Practical Uses in Cell and Molecular Biology

Edited by Brian K. Kay and H. Benjamin Peng

Series EditorsLESLIEWILSONANDPAULMATSUDAIRA

Volume 37 (1993)Antibodies in Cell Biology

Edited by David J. Asai


Volume 38 (1993)Cell Biological Applications of Confocal Microscopy

Edited by Brian Matsumoto

Volume 39 (1993)Motility Assays for Motor Proteins

Edited by Jonathan M. Scholey

Volume 40 (1994)A Practical Guide to the Study of Calcium in Living Cells

Edited by Richard Nuccitelli

Volume 41 (1994)Flow Cytometry, Second Edition, Part A

Edited by Zbigniew Darzynkiewicz, J. Paul Robinson, and Harry A. Crissman

Volume 42 (1994)Flow Cytometry, Second Edition, Part B

Edited by Zbigniew Darzynkiewicz, J. Paul Robinson, and Harry A. Crissman

Volume 43 (1994)Protein Expression in Animal Cells

Edited by Michael G. Roth

Volume 44 (1994)Drosophila melanogaster: Practical Uses in Cell and Molecular Biology

Edited by Lawrence S. B. Goldstein and Eric A. Fyrberg

Volume 45 (1994)Microbes as Tools for Cell Biology

Edited by David G. Russell

Volume 46 (1995)Cell Death

Edited by Lawrence M. Schwartz and Barbara A. Osborne

Volume 47 (1995)Cilia and Flagella

Edited by William Dentler and George Witman

Volume 48 (1995)Caenorhabditis elegans: Modern Biological Analysis of an Organism

Edited by Henry F. Epstein and Diane C. Shakes

Volume 49 (1995)Methods in Plant Cell Biology, Part A

Edited by David W. Galbraith, Hans J. Bohnert, and Don P. Bourque


Volume 50 (1995)Methods in Plant Cell Biology, Part B

Edited by David W. Galbraith, Don P. Bourque, and Hans J. Bohnert

Volume 51 (1996)Methods in Avian Embryology

Edited by Marianne Bronner-Fraser

Volume 52 (1997)Methods in Muscle Biology

Edited by Charles P. Emerson, Jr. and H. Lee Sweeney

Volume 53 (1997)Nuclear Structure and Function

Edited by Miguel Berrios

Volume 54 (1997)Cumulative Index

Volume 55 (1997)Laser Tweezers in Cell Biology

Edited by Michael P. Sheetz

Volume 56 (1998)Video Microscopy

Edited by Greenfield Sluder and David E. Wolf

Volume 57 (1998)Animal Cell Culture Methods

Edited by Jennie P. Mather and David Barnes

Volume 58 (1998)Green Fluorescent Protein

Edited by Kevin F. Sullivan and Steve A. Kay

Volume 59 (1998)The Zebrafish: Biology

Edited by H. William Detrich III, Monte Westerfield, and Leonard I. Zon

Volume 60 (1998)The Zebrafish: Genetics and Genomics

Edited by H. William Detrich III, Monte Westerfield, and Leonard I. Zon

Volume 61 (1998)Mitosis and Meiosis

Edited by Conly L. Rieder


Volume 62 (1999)Tetrahymena thermophila

Edited by David J. Asai and James D. Forney

Volume 63 (2000)Cytometry, Third Edition, Part A

Edited by Zbigniew Darzynkiewicz, J. Paul Robinson, and Harry Crissman

Volume 64 (2000)Cytometry, Third Edition, Part B

Edited by Zbigniew Darzynkiewicz, J. Paul Robinson, and Harry Crissman

Volume 65 (2001)Mitochondria

Edited by Liza A. Pon and Eric A. Schon

Volume 66 (2001)Apoptosis

Edited by Lawrence M. Schwartz and Jonathan D. Ashwell

Volume 67 (2001)Centrosomes and Spindle Pole Bodies

Edited by Robert E. Palazzo and Trisha N. Davis

Volume 68 (2002)Atomic Force Microscopy in Cell Biology

Edited by Bhanu P. Jena and J. K. Heinrich Horber

Volume 69 (2002)Methods in Cell–Matrix Adhesion

Edited by Josephine C. Adams

Volume 70 (2002)Cell Biological Applications of Confocal Microscopy

Edited by Brian Matsumoto

Volume 71 (2003)Neurons: Methods and Applications for Cell Biologist

Edited by Peter J. Hollenbeck and James R. Bamburg

Volume 72 (2003)Digital Microscopy: A Second Edition of Video Microscopy



Volume 73 (2003)Cumulative Index

Volume 74 (2004)Development of Sea Urchins, Ascidians, and Other Invertebrate

Deuterostomes: Experimental Approaches

Edited by Charles A. Ettensohn, Gary M. Wessel, and Gregory A. Wray

Volume 75 (2004)Cytometry, 4th Edition: New Developments

Edited by Zbigniew Darzynkiewicz, Mario Roederer, and Hans Tanke

Volume 76 (2004)The Zebrafish: Cellular and Developmental Biology

Edited by H. William Detrich, III, Monte Westerfield, and Leonard I. Zon

Volume 77 (2004)The Zebrafish: Genetics, Genomics, and Informatics

Edited by William H. Detrich, III, Monte Westerfield, and Leonard I. Zon

Volume 78 (2004)Intermediate Filament Cytoskeleton

Edited by M. Bishr Omary and Pierre A. Coulombe

Volume 79 (2007)Cellular Electron Microscopy

Edited by J. Richard McIntosh

Volume 80 (2007)Mitochondria, 2nd Edition

Edited by Liza A. Pon and Eric A. Schon

Volume 81 (2007)Digital Microscopy, 3rd Edition


Volume 82 (2007)Laser Manipulation of Cells and Tissues

Edited by Michael W. Berns and Karl Otto Greulich

Volume 83 (2007)Cell Mechanics

Edited by Yu-Li Wang and Dennis E. Discher

Volume 84 (2007)Biophysical Tools for Biologists, Volume One: In Vitro Techniques

Edited by John J. Correia and H. William Detrich, III


Volume 85 (2008)Fluorescent Proteins

Edited by Kevin F. Sullivan

Volume 86 (2008)Stem Cell Culture

Edited by Dr. Jennie P. Mather

Volume 87 (2008)Avian Embryology, 2nd Edition

Edited by Dr. Marianne Bronner-Fraser

Volume 88 (2008)Introduction to Electron Microscopy for Biologists

Edited by Prof. Terence D. Allen

Volume 89 (2008)Biophysical Tools for Biologists, Volume Two: In Vivo Techniques

Edited by Dr. John J. Correia and Dr. H. William Detrich, III

Volume 90 (2008)Methods in Nano Cell BiologyEdited by Bhanu P. Jena

Volume 91 (2009)Cilia: Structure andMotilityEdited by Stephen M. King and Gregory J. Pazour

Volume 92 (2009)Cilia: Motors and RegulationEdited by Stephen M. King and Gregory J. Pazour

Volume 93 (2009)Cilia: Model Organisms and IntraflagellarTransportEdited by Stephen M. King and Gregory J. Pazour

Volume 94 (2009)Primary CiliaEdited by Roger D. Sloboda

Volume 95 (2010)Microtubules, in vitroEdited by Leslie Wilson and John J. Correia

Volume 96 (2010)ElectronMicroscopyof Model SystemsEdited by Thomas Mueller-Reichert


Volume 97 (2010)Microtubules: InVivoEdited by Lynne Cassimeris and Phong Tran

Volume 98 (2010)Nuclear Mechanics & GenomeRegulationEdited by G.V. Shivashankar

Calcium in Living Cells

Documents

Transcript of Calcium in Living Cells