Hybrid materials

Temperature Compensation for Hybrid Devices:Kinesin’s Km is Temperature Independent**

Robert Tucker, Ajoy K. Saha, Parag Katira, Marlene Bachand,

George D. Bachand, and Henry Hess*

A wide variety of nanodevices integrate biological compo-

nents to provide unique functions. One of the challenges

arising from this hybrid approach is the stabilization of device

operation against temperature changes. The activity of

biological nanomachines, such as enzymes, is strongly

temperature-dependent, often increasing from 50 to 300%

for every 10 8C increase in temperature at saturating substrate

concentrations.[1] Traditionally, temperature is closely con-

trolled in biotechnological processes and microfluidic

devices,[2–4] either to maintain a stable activity or to switch

between active and inactive states of the system. However, in

field-deployable devices it would be desirable to stabilize

internal processes against temperature fluctuations. Inspira-

tion for stabilization strategies is provided by poikilotherm

organisms, which compensate for temperature changes over a

wide range of timescales.[5,6]

Two biological strategies providing instantaneous com-

pensation are of particular interest. On one hand, metabolic

networks can display temperature compensation due to the

existence of feedback loops.[7] The engineering equivalent is

found in the canonical layout for temperature-compensated

electronic circuits.[8] On the other hand, enzymes themselves

can display near constant activity for subsaturating substrate

concentrations ([S]<Km) in a limited temperature range. An

example is lactate dehydrogenase from Alaskan king crab,

whoseMichaelis constant,Km, increases with temperature and

compensates for the increasing vmax value.[9] From an

[�] R. Tucker, A. K. Saha, P. Katira, Prof. H. Hess

Department of Materials Science and Engineering

University of Florida

160 Rhines Hall, Gainesville, FL 32611 (USA)

E-mail: hhess@mse.ufl.edu

M. Bachand, G. D. Bachand

Biomolecular Interfaces & Systems Department

Sandia National Laboratories

PO Box 5800, MS-1413, Albuquerque, NM 87185 (USA)

[��] Financial support was provided by the DARPA Biomolecular MotorsProgram (AFOSR FA 9550-05-1-0274 & FA 9550-05-1-0366) andthe UF Center for Sensor Materials and Technologies (ONRN00014-07-1-0982). This work was performed, in part, at theCenter for Integrated Nanotechnologies, a US Department ofEnergy, Office of Basic Energy Sciences nanoscale scienceresearch center operated jointly by Los Alamos and SandiaNational Laboratories. Sandia is a multiprogram laboratory oper-ated by Sandia Corporation, a Lockheed Martin Company, for theUnited States Deparatment of Energy’s National Nuclear SecurityAdministration under Contract DE-AC04-94AL85000.

DOI: 10.1002/smll.200801510

small 2009, 5, No. 11, 1279–1282 � 2009 Wiley-VCH Verlag Gmb

engineering point of view, this approach is preferable because

of its simplicity and is similar to employing a material with

intrinsic temperature compensation such as INVAR steel.[10]

An interesting target for an exploration of this tempera-

ture-compensationmechanism in the context of hybrid devices

is the ATPase kinesin. Kinesin and other motor proteins have

evolved in nature to provide actuation and transport within

cells.[11,12] Recently, motor proteins and their associated

filaments (microtubules in the case of kinesin) have been

employed for the transport of nanoscale cargo in micro-

fabricated structures,[13–21] and it has been shown that these

‘‘molecular shuttle’’ systems[22,23] can be applied to tasks such

as force measurements,[24] surface imaging,[25] single molecule

manipulation,[26] computing,[27] and biosensing.[28–30] Motor-

driven active transport is thus an attractive alternative to

pressure-driven fluid flow and electro-osmotic flow.[31]

Bohm et al.[32] previously reported that the Km values for

porcine kinesin-1 at temperatures of 25 and 35 8C are 66� 10

and 79� 8mM, respectively. Kawaguchi and Ishiwata deter-

mined that the activation energy of bovine kinesin-1 is

50 kJ mol�1, implying a Q10 at saturating ATP concentrations

equal to two.[33] This suggests that the activity change per

10 8C change in temperature for small ATP concentrations is

significantly smaller than two, implying that the increasingKm

partially compensates for the increasing vmax. Similarly, ATP

consumption assays for Thermomyces lanuginosis kinesin-3

determined an activation energy of 94� 9 kJ mol�1, and an

increase in the Km from 42� 7mM at 25 8C to 1.6� 0.3mM at

50 8C.[34] At low ATP concentrations the Q10 is 0.7 and thus

the Thermomyces kinesin activity falls with increasing

temperature despite an increasing activity at saturating

substrate conditions.

Here, we present for the first time detailed measurements

of the temperature dependence of kinesin motor protein

activity at subsaturating substrate concentrations from 19 to

34 8C for both Drosophila kinesin-1 and Thermomyces

kinesin-3. These measurements were undertaken in the

expectation that – similar to other enzymes with applications

in biotechnology[35] – the Km of kinesin would increase with

increasing temperature with beneficial implications for the

design of kinesin motor powered hybrid devices.

Our velocity measurements for microtubules gliding on

Drosophila kinesin-1 (Figure 1) show the expected Arrhenius-

type increase with increase in temperature for a saturating

ATP concentration (1mM), which replicates the data of

Kawaguchi and Ishiwata[33] for saturating ATP concentrations

(1mM) obtained with bovine kinesin (also see [36]).

H & Co. KGaA, Weinheim 1279

communications

Figure 1. A) Microtubule gliding velocity on Drosophila kinesin-1 as a

function of temperature for various ATP concentrations. Open squares

are the data published by Kawaguchi and Ishiwata [33] for 1mM ATP.

B) Microtubule gliding velocity on Thermomyces kinesin-3 as a function

of temperature for various ATP concentrations.

Figure 3. A) Km and B)maximal velocity vmax as function of temperature.

While Km does not change significantly as a function of temperature, the

dependence of vmax on temperature fits an Arrhenius equation well.

Drosophila kinesin-1 – full squares, black; Thermomyces kinesin-3 –

open triangles, gray.

1280

A plot of velocity as function of ATP concentration for a

series of temperatures (Figure 2) was obtained by interpolat-

ing the data points in Figure 1 for five temperatures for

Drosophila and four for Thermomyces. Fitting these curves

with a Michaelis-Menten equation v([ATP], T)¼ vmax(T)�[ATP]/(Km(T)þ [ATP]) revealed the temperature depen-

dence of the velocity at saturating substrate concentrations

vmax and the Michaelis constant Km (Figure 3). The 34 8C data

point of the Thermomyces data has not been included in the fit

of vmax and Km since the reduced vmax indicates partial

deactivation.

The temperature-dependent vmax parameters were

graphed in Arrhenius plots ln(vmax)¼ vmax, 0� exp(�Ea/RT),

and the activation energies were determined by linear error-

weighted least-square fits to be 53� 5.1 kJ mol�1 (Q10¼ 2.04)

Figure 2. Microtubule gliding velocity as function of ATP concentration

for a series of temperatures obtained by interpolating the data pre-

sented in Figure 1. Lines are fits to Michaelis–Menten functions with

vmax and Km as parameters. A) Drosophila kinesin-1, B) Thermomyces

kinesin-3.

www.small-journal.com � 2009 Wiley-VCH Verlag Gm

and 38� 3.3 kJ mol�1 (Q10¼ 1.67) for Drosophila kinesin-1

and Thermomyces kinesin-3, respectively. Independent of

temperature, the Km values were found to be 70� 10mM and

220� 90mM for Drosophila kinesin-1 and Thermomyces

kinesin-3, respectively. Since the Michaelis constant has been

found to be independent of temperature in the case of the two

tested kinesins, temperature compensation of enzyme activity

cannot be achieved at the expense of enzyme turnover by

reducing the substrate concentration.

It is not surprising that our measurements in the

temperature interval from 19 to 34 8C do not replicate the

dramatic increase of the Km of Thermomyces kinesin-3

observed by Rivera et al. in hydrolysis measurements at

55 8C. The tested temperature range is well below the thermal

optimum for this organism. In addition, several changes in the

properties of Thermomyces kinesin-3 were observed at

temperatures above 45 8C.[34] A second preparation of

Thermomyces kinesin-3 showed significantly altered motility

(more interruptions of gliding motion and more stuck

microtubules) and a Km of 175mM at 23 8C, which indicates

that variations between preparations exist.

It is more challenging to reconcile the increase in Km for

porcine kinesin-1 observed by Bohm et al.[32] with our

observations. A possible explanation for Bohm et al. observa-

tion is the depletion of ATP from the solution over time in the

absence of an ATP replenishing system, which caused an

apparent increase of Km of Drosophila kinesin as the cell was

heated over the course of an hour in our initial experiments.

Only the utilization of an ATP regenerating system[37]

prevented this effect in the presented experiments for both

Thermomyces and Drosophila kinesin.

It could be argued that even if kinesin’sKm would increase

in proportion to vmax, the compensation strategy would

require a reduction of substrate concentration to a fraction

of Km, and consequently an undesirable loss in device

performance. However, organisms routinely operate enzymes

at subsaturating substrate conditions,[6] and even for hybrid

bH & Co. KGaA, Weinheim small 2009, 5, No. 11, 1279–1282

devices it is far from self-evident that the optimum activation

corresponds to themaximum activation. For example, we have

found that a velocity of 200 nm s�1 represents an optimum for

the loading of cargo onto kinesin-driven molecular shuttles

using biotin–streptavidin linkages.[38] Similarly, INVAR steel

sacrifices mechanical properties for a low thermal expansion

coefficient.

For the case of kinesin, we conclude that stabilization

against temperature changes will require the design of a

suitable enzymatic network, adding to the complexity of the

device similar to the recently demonstrated enhanced

robustness of a DNA nanomotor.[39]

Experimental Section

All chemicals are from Sigma–Aldrich (St. Louis, MO) unless

otherwise specified.

Kinesin and microtubule preparation: A kinesin construct

consisting of the wild-type, full-length D. melanogaster kinesin

heavy chain and a C-terminal His-tag was expressed in Escherichia

coli, and purified using a Ni-NTA column as in Reference [40].

T. lanuginosus kinesin was prepared as in Reference [34].

Rhodamine-labeled tubulin (3.2mg mLS1, Cytoskeleton, Denver,

CO) was polymerized at 37 -C for 30min in BRB80 buffer (80mM

PIPES, 1mM EGTA, 2mM MgCl2, pH 6.9) with 1mM GTP (Roche

Diagnostics, Indianapolis IN), 4mM MgCl2, and 5% DMSO, and

subsequently diluted 100-fold into BRB80 with 10mM paclitaxel.

Variable temperature motility assay: A new flow cell was

assembled for each ATP concentration tested, which consisted of

a square coverslip (22T22mm2, FishersFinest no. 1 Fisher

Scientific), double-sided tape as spacer, and a circular coverslip

(15mm diameter, no. 1 Warner Instruments). In the Drosophila

kinesin assays, the inner surfaces of the flow cell were incubated

with a casein solution (0.5mg mLS1 in BRB80) for 5min, and then

a Drosophila kinesin solution (10 nM in BRB80, 0.1mg mLS1

casein, and the desired ATP concentration) for 5min. A motility

solution containing 3.2mg mLS1 microtubules, casein (0.2mg

mLS1), an antifade solution (20mM D-glucose, 20mg mLS1

glucose oxidase, 8mg mLS1 catalase, 10mM dithiothreitol (BioRad

Laboratories, Hercules, CA), an ATP-regeneration system (2000

Units LS1 creatine phosphokinase, 2mM creatine phosphate –

ref.[37]) and ATP (10, 25, 50, 100, 200, or 1000mM) was

introduced, and the edges of the flow cell were sealed with

Apiezon grease to minimize evaporation. In the Thermomyces

kinesin assays, the procedure was identical except the flow cell

was not incubated with casein, the motility solution did not

contain casein, and the ATP concentrations tested were 10, 100,

200, 1000, and 5000mM. The flow cell was placed on the heat

stage (Model RC-20, Warner Instruments), the top coverslip coated

with thermally conductive silver paste and covered with an

aluminum plate, and the ensemble was fastened together with

screws. The temperature of the flow cell was set to the desired

temperature (19, 23, 26, 31, or 34 -C) and automatically regulated

by a temperature controller (Warner Instruments, TC-324B), which

used an electric current and a pair of 20V resistors to maintain a

constant temperature. The microscope objective (Nikon 100T oil

immersion) was heated by a resistive wire (Nichrome 60, Pelican

Wire) powered with a DC regulated power supply (EXTECH1

small 2009, 5, No. 11, 1279–1282 � 2009 Wiley-VCH Verlag Gmb

instruments). The power supply was manually regulated to heat

the objective to the same temperature as the flow cell.

Measurement of temperature: The thermistor from the auto-

matic temperature controller was inserted into the heat stage near

the aluminum plate on top of the flow cell. A second thermocouple

was attached to the objective and read using a multimeter.

Thermocouple readings in the temperature range of interest were

calibrated against an alcohol thermometer.

Measurement of velocity: An Eclipse TE2000-U fluorescence

microscope (Nikon, Melville NY) with a 100T oil objective (NA

1.4), an X-cite 120 lamp (EXFO, Ontario, Canada), a rhodamine

filter cube (no. 48002, Chroma Technologies, Rockingham, VT),

and an iXon EMCCD camera (ANDOR, South Windsor, CT) were

used to image microtubules on the bottom surface of the flow

cells. A series of ten images were taken for each ATP concentration

and temperature, and the gliding velocities of approximately ten

microtubules were measured using image analysis software

(ImageJ v1.37c, National Institutes of Health, USA). The experi-

ment was begun at room temperature (T¼19 -C) and the flow cell

was heated to each new temperature until motility ceased. The

error bars shown in the figures are the standard error of the mean

of the velocity measurements.

Keywords:

drosophila . hybrid materials . Michaelis–Menten kinetics .molecular motors . smart dust

[1] A. Bezkorovainy, M. E. Rafelson, Concise Biochemistry, Marcel

Dekker, New York 1996.

[2] D. L. Huber, R. P. Manginell, M. A. Samara, B. I. Kim, B. C. Bunker,

Science 2003, 301, 352.[3] R. M. Guijt, A. Dodge, G. W. K. van Dedem, N. F. de Rooij, E.

Verpoorte, Lab. Chip 2003, 3, 1.[4] G. Mihajlovic, N. M. Brunet, J. Trbovic, P. Xiong, S. von Molnar, P. B.

Chase, Appl. Phys. Lett. 2004, 85, 1060.[5] J. R. Hazel, C. L. Prosser, Physiol. Rev. 1974, 54, 620.[6] G. N. Somero, Annu. Rev. Ecol. Syst. 1978, 9, 1.[7] P. Ruoff, M. Zakhartsev, H. V. Westerhoff, FEBS J. 2007, 274, 940.[8] P. Horowitz, W. Hill, The Art of Electronics, Cambridge University

Press, Cambridge 1989.

[9] G. N. Somero, Biochem. J. 1969, 114, 237.[10] M. van Schilfgaarde, I. A. Abrikosov, B. Johansson, Nature 1999,

400, 46.

[11] M. Schliwa, G. Woehlke, Nature 2003, 422, 759.[12] J. Howard, Mechanics of Motor Proteins and the Cytoskeleton,

Sinauer, Sunderland, MA 2001.

[13] D. V. Nicolau, H. Suzuki, S. Mashiko, T. Taguchi, S. Yoshikawa,

Biophys. J. 1999, 77, 1126.[14] Y. Hiratsuka, T. Tada, K. Oiwa, T. Kanayama, T. Q. Uyeda, Biophys. J.

2001, 81, 1555.[15] P. Stracke, K. J. Bohm, J. Burgold, H. J. Schacht, E. Unger, Nano-

technology 2000, 11, 52.[16] M. G. L. van den Heuvel, C. T. Butcher, S. G. Lemay, S. Diez, C.

Dekker, Nano Lett. 2005, 5, 235.[17] R. Yokokawa, S. Takeuchi, T. Kon, M. Nishiura, R. Ohkura, K. Sutoh,

H. Fujita, J. Microelectromech. Syst. 2004, 13, 612.[18] Y. Z. Du, Y. Hiratsuka, S. Taira, M. Eguchi, T. Q. P. Uyeda, N.

Yumoto, M. Kodaka, Chem. Commun. 2005, 40, 2080.[19] Y. M. Huang, M. Uppalapati, W. O. Hancock, T. N. Jackson, IEEE

Trans. Adv. Pack. 2005, 28, 564.[20] S. G. Moorjani, L. Jia, T. N. Jackson, W. O. Hancock, Nano Lett.

2003, 3, 633.

H & Co. KGaA, Weinheim www.small-journal.com 1281

communications

1282

[21] C.-T. Lin, M.-T. Kao, K. Kurabayashi, E. Meyhofer, Small 2006, 2,281.

[22] J. R. Dennis, J. Howard, V. Vogel, Nanotechnol. 1999, 10, 232.[23] H. Hess, J. Clemmens, D. Qin, J. Howard, V. Vogel, Nano Lett. 2001,

1, 235.

[24] H. Hess, J. Howard, V. Vogel, Nano Lett. 2002, 2, 1113.[25] H. Hess, J. Clemmens, J. Howard, V. Vogel, Nano Lett. 2002, 2, 113.[26] C. Z. Dinu, J. Opitz, W. Pompe, J. Howard, M. Mertig, S. Diez, Small

2006, 2, 1090.[27] D. V. Nicolau, D. V. Nicolau, G. Solana, K. L. Hanson, L. Filipponi, L.

S. Wang, A. P. Lee, Microelectron. Eng. 2006, 83, 1582.[28] G. D. Bachand, S. B. Rivera, A. Carroll-Portillo, H. Hess, M. Bac-

hand, Small 2006, 2, 381.[29] S. Ramachandran, K.-H. Ernst, G. D. Bachand, V. Vogel, H. Hess,

Small 2006, 2, 330.[30] C. T. Lin, M. T. Kao, K. Kurabayashi, E. Meyhofer,Nano Lett. 2008, 8,

1041.

[31] T. Nitta, H. Hess, Nano Lett. 2005, 5, 1337.

www.small-journal.com � 2009 Wiley-VCH Verlag Gm

[32] K. J. Bohm, R. Stracke, M. Baum, M. Zieren, E. Unger, FEBS Lett.

2000, 466, 59.[33] K. Kawaguchi, S. Ishiwata, Cell Motil. Cytoskeleton 2001, 49, 41.[34] S. B. Rivera, S. J. Koch, J. M. Bauer, J. M. Edwards, G. D. Bachand,

Fungal Genet. Biol. 2007, 44, 1170.[35] R. K. Scopes, Clin. Chim. Acta 1995, 237, 17.[36] K. Kawaguchi, S. Ishiwata, Biochem. Biophys. Res. Commun. 2000,

272, 895.

[37] C. Leduc, F. Ruhnow, J. Howard, S. Diez, Proc. Natl. Acad. Sci. USA

2007, 104, 10847.[38] A. Agarwal, P. Katira, H. Hess, 2008, unpublished results.

[39] J. D. Bishop, E. Klavins, Nano Lett. 2007, 7, 2574.[40] D. L. Coy, M. Wagenbach, J. Howard, J. Biol. Chem. 1999, 274,

3667.

bH & Co. KGaA, Weinheim

Received: October 10, 2008Published online: March 19, 2009

small 2009, 5, No. 11, 1279–1282