Nanoparticles for Cancer Detection and Therapy: …...Abstract Nanoparticles for Cancer Detection...

Nanoparticles for Cancer Detection and Therapy: TowardsDiagnostic Applications of Quantum Dots and Rational

Design of Drug Delivery Vehicles

by

Sawitri Mardyani

A thesis submitted in conformity with the requirementsfor the degree of Doctor of Philosophy

Graduate Department of Institute of Biomaterials and BiomedicalEngineering

University of Toronto

Copyright c© 2011 by Sawitri Mardyani

Abstract

Nanoparticles for Cancer Detection and Therapy: Towards Diagnostic Applications of

Quantum Dots and Rational Design of Drug Delivery Vehicles

Sawitri Mardyani

Doctor of Philosophy

Graduate Department of Institute of Biomaterials and Biomedical Engineering

University of Toronto

2011

This thesis describes observations, techniques and strategies, which contribute to-

wards the development of nanoparticle based detection and treatment of cancer. Quan-

tum dots and biorecognition molecules were studied towards applications in detection

and microgels were used in the rational design of a targeted drug delivery vehicle. The

fluorescence intensity of quantum dots was examined in buffers commonly used in molec-

ular biology. The fluorescence intensity of ZnS-capped CdSe quantum dots (QDs) was

found to vary significantly, depending on the amount of ZnS capping on the QDs or

the concentration, pH and type of buffer the QDs were in. Since fluorescence cannot

reliably be used to quantify QDs, an alternative quantification method was developed,

which does not rely on their fluorescence. This method employs phage display to identify

nanoparticle-specific bacteriophage which were then applied in an assay to quantify QDs

in environments where absorbance or fluorescence spectroscopy are ineffective. Biorecog-

nition molecules, which can direct nanoparticles to a molecular target, were also identi-

fied through phage display. Phage display on whole cells was used to identify a peptide,

which was conjugated with QDs to stain HeLa (cervical cancer) cells. A high-throughput

phage display screening strategy was also developed, which could enable the simultane-

ous identification of multiple biorecognition molecules from a single library. QD-encoded

microbead barcodes were conjugated to protein targets and then used to screen a phage

ii

display library. The beads and the binding phage were then separated using flow cy-

tometry and fluorescence assisted cell sorting. Finally, biorecognition molecules were

combined with nanoparticles to create drug delivery vehicles, which were designed to

protect, deliver and then release chemotherapeutic drugs through an intracellular pH

trigger. PolyNIPAAm and chitosan hydrogels, under 200 nm in diameter, were loaded

with chemotherapeutic drugs, conjugated to transferrin and tested in vitro on HeLa cells.

These projects demonstrate the great potential in this growing field as well as some of

the many challenges that have yet to be overcome.

iii

Dedication

To Ibu,

for never letting me give up.

iv

Acknowledgements

I want to thank everyone. Everyone. More than anything, this was a team effort. If I

learned one thing from this whole experience, it was to ask for help. I want to thank Uki

for teaching me that lesson when I needed it the most. And I want to thank everyone

else for stepping up to help – especially those who stepped up even before I learned how

to ask. In particular, I would like to thank Sana Siddiqui for reading and making sense

of this document long before it made much sense, Dr. Setiadi Yazid for showing me how

to strengthen it and Dr. Tanya Hauck for helping me make the final revisions.

I would like to thank my lab mates for their support, both moral and technical. It

has been a privilege to work with such brilliant and dedicated people. I would also like to

thank the students, whom I’ve had the pleasure to mentor: Wilson Chan, Daniel Pfeffer,

Christopher Park and Carly Petes. I thank them for the technical support they have

given me and for everything they have taught me.

Outside the Chan lab, I have had the opportunity of working with several exceptional

scientists. I thank Dr. Mallika Das and Dr. Hong Zheng from Dr. Eugenia Kumacheva’s

lab for our productive collaboration in developing rationally designed drug delivery ve-

hicles from the hydrogels they synthesized. I also thank Tammy Reid and the technical

support of New England Biolabs for teaching this electrical engineer phage display.

I would like to thank my supervisor, Dr. Warren Chan, for the opportunity to explore

this fascinating field – and for helping me find my way when I got lost. I also thank my

committee members and examiners Dr. Julie Audet, Dr. Chen Wang, Dr. Kevin Truong

and Dr. Heather Sheardown for their helpful discussions and comments. In particular, I

would like to thank my committee chair, Dr. Chris Yip for the diligence, meticulousness

and indefatigability he brings to everything I’ve seen him do.

Finally, I would like to thank my friends and family–and friends who have become

family, for their encouragement, support and prayers. Without them, and without the

One who answers prayers, none of this would have been possible.

v

Contents

1 Introduction 1

1.1 Nanotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Nanoparticles in detection and diagnosis . . . . . . . . . . . . . . . . . . 5

1.2.1 Quantum dots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.3 Biorecognition molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

1.3.1 Antibodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

1.3.2 Aptamers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.3.3 Phage display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.4 Nanoparticles in therapy: developing drug delivery vehicles . . . . . . . . 13

1.4.1 Nanoparticle properties . . . . . . . . . . . . . . . . . . . . . . . . 14

1.4.2 Active Targeting . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1.5 Thesis outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2 Variability of Quantum Dot Fluorescence 22

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.2.1 Synthesis of ZnS-capped CdSe Nanocrystals . . . . . . . . . . . . 25

2.2.2 Composition analysis through inductively coupled plasma spec-

troscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.2.3 Immobilization of QDs in polystyrene microbeads . . . . . . . . . 25

vi

2.2.4 Flow cytometry and statistical analysis . . . . . . . . . . . . . . . 26

2.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3 Quantification of QDs using biorecognition molecules 37

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37


3.2.1 Synthesis and characterization of nanoparticles . . . . . . . . . . . 40

3.2.2 Creation of the affinity matrix . . . . . . . . . . . . . . . . . . . . 42

3.2.3 Phage panning . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.2.4 Screening of clones . . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.2.5 Determination of binding characteristics . . . . . . . . . . . . . . 44

3.2.6 Phage assay for QD quantification . . . . . . . . . . . . . . . . . . 45

3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.3.1 Nanoparticle characterization . . . . . . . . . . . . . . . . . . . . 46

3.3.2 Optimization of affinity matrix . . . . . . . . . . . . . . . . . . . 46

3.3.3 Panning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

3.3.4 Sequences of binding clones . . . . . . . . . . . . . . . . . . . . . 51

3.3.5 Comparison of binding characteristics . . . . . . . . . . . . . . . . 52

3.3.6 Determination of dissociation constant . . . . . . . . . . . . . . . 52

3.3.7 Measuring concentration using phage assay . . . . . . . . . . . . . 57

3.3.8 Robustness against optically interfering agents . . . . . . . . . . . 57

3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4 Phage Display on Whole Cells 61

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61


4.2.1 Phage Panning . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

vii

4.2.2 Sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.2.3 Analysis of binding peptide . . . . . . . . . . . . . . . . . . . . . 65

4.2.4 Peptide-quantum dot conjugation . . . . . . . . . . . . . . . . . . 65

4.2.5 Cell staining experiment . . . . . . . . . . . . . . . . . . . . . . . 66

4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

4.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

5 High Throughput Phage Display 73

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73


5.2.1 Polystyrene bead barcordes . . . . . . . . . . . . . . . . . . . . . 78

5.2.2 Conjugation of barcodes to proteins . . . . . . . . . . . . . . . . . 78

5.2.3 Phage display panning . . . . . . . . . . . . . . . . . . . . . . . . 78

5.2.4 Analysis of bound peptides . . . . . . . . . . . . . . . . . . . . . . 79


5.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

6 Rational Design of Nanoscale Drug Delivery Vehicles 83

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84


6.2.1 Microgel synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . 87

6.2.2 Loading and release of microgel payload . . . . . . . . . . . . . . 88

6.2.3 Conjugation of targeting moiety . . . . . . . . . . . . . . . . . . . 88

6.2.4 In vitro studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89


6.3.1 pH-triggered release of microgel payload . . . . . . . . . . . . . . 90

6.3.2 In vitro release of R6G from polyNIPAM-AAc microgels . . . . . 93

viii

6.3.3 In vitro viability studies of HeLa cells treated with targeted micro-

gel drug delivery vehicles with pH-triggered release . . . . . . . . 93

6.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

7 Conclusions and Future Work 99

7.1 Conclusions and contributions . . . . . . . . . . . . . . . . . . . . . . . . 99

7.2 Future work: Development of quantum dots for diagnostic applications . 101

7.3 Future work: Identification of targeting molecules . . . . . . . . . . . . . 103

7.4 Future work: Rational design of drug delivery vehicles . . . . . . . . . . . 104

Bibliography 105

ix

List of Tables

2.1 Properties of ZnS-capped CdSe QDs used in this study . . . . . . . . . . 28

3.1 Number of phage eluted after each round of panning . . . . . . . . . . . 50

3.2 Washing conditions in each round of panning . . . . . . . . . . . . . . . . 51

3.3 Amino acid sequences of QD-binding peptides . . . . . . . . . . . . . . . 51

6.1 Specifications of Rationally Designed Hydrogel Drug Delivery Vehicles . . 87

x

List of Figures

1.1 The scale of nanoparticles compared to biological entities . . . . . . . . . 3

1.2 Nanoparticles for cancer diagnostics and therapy . . . . . . . . . . . . . . 9

1.3 Phage display screening . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.4 Nanoparticle parameters and their implications for rational design of drug

delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1.5 Summary of projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.1 TEM of quantum dot bead barcode . . . . . . . . . . . . . . . . . . . . . 27

2.2 Variability of QD fluorescence in buffers . . . . . . . . . . . . . . . . . . 29

2.3 Influence of QD capping on variability of QD fluorescence in buffers . . . 31

2.4 Influence of buffer concentration on QD fluorescence . . . . . . . . . . . . 33

2.5 Influence of buffer pH on QD fluorescence . . . . . . . . . . . . . . . . . 34

3.1 Phage assay schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

3.2 Phage display panning . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3.3 Spectra and gel electrophoresis of QDs of different surface chemistries . . 47

3.4 QD Surface Chemistries . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

3.5 Optimization of the affinity matrix . . . . . . . . . . . . . . . . . . . . . 49

3.6 Binding of MAA7 phage on QDs of different surface chemistries . . . . . 53

3.7 Standard curve - Absorbance vs. Phage concentration . . . . . . . . . . . 54

3.8 Competitive assay using MAA-coated QD-specific phage . . . . . . . . . 55

xi

3.9 Scatchard plot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3.10 Measurement of QD concentration by phage assay . . . . . . . . . . . . . 57

3.11 Quantification of QDs by phage assay in the presence of optically interfer-

ing agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

4.1 Evolution of phage display library after subtraction and selection . . . . 63

4.2 Progress of panning experiment on whole cells . . . . . . . . . . . . . . . 67

4.3 Structure of HeLa-binding peptide . . . . . . . . . . . . . . . . . . . . . . 68

4.4 Cell staining with peptide-quantum dot conjugates . . . . . . . . . . . . 69

5.1 Process of screening a molecular library to identify a targeting molecules. 74

5.2 Proposed method of high throughput screening. . . . . . . . . . . . . . . 75

5.3 Barcoding with quantum dots. . . . . . . . . . . . . . . . . . . . . . . . . 76

5.4 High Throughput Screening of Biorecognition Molecules . . . . . . . . . . 77

5.5 Separation of optically coded polystyrene beads into respective populations. 80

5.6 Binding of peptides to protein A. . . . . . . . . . . . . . . . . . . . . . . 81

6.1 pH sensitive targeted drug delivery vehicle. . . . . . . . . . . . . . . . . . 86

6.2 Characterization of pH-sensitive nanogels . . . . . . . . . . . . . . . . . . 91

6.3 In vitro cumulative release of methotrexate from the microgels . . . . . . 92

6.4 Release of R6G in HeLa cells . . . . . . . . . . . . . . . . . . . . . . . . . 94

6.5 Cell viability after 36 hours of incubation with doxorubicin-loaded polyNIPAM-

AAc hydrogels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

6.6 Cell viability under methotrexate treatment in targeted chitosan-based

drug delivery vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

xii

Abbreviations

BRM biorecognition molecule

BSA bovine serum albumin

CS chitosan

CTAB cetyl trimethylammonium bromide

DDV drug delivery vehicle

DIC differential interference contrast

DMEM Dulbecco’s Modified Eagle Medium

Dox Doxorubicin

EDC 1-ethyl 2-(2-dimethyl propyl) carbodiimide hydrochloride

EDTA ethylenediaminetetraacetic acid

ELISA enzyme-linked immunosorbant assay

EPR enhanced permeability and retention

FACS fluorescence activated cell sorting

FWHM full width half maximum

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HRP horseradish peroxidase

HTTC N-[(2hydroxy-3-trimethylammonium)propyl] chitosan chloride

ICP inductively coupled plasma

IPTG isopropyl β-D-1-thiogalactopyranoside

Kd dissociation constant

LP low pass

MAA mercaptoacetic acid

MTX Methotrexate

MUA mercaptoundecanoic acid

NME new molecular entity

NP nanoparticle

xiii

PBS phosphate buffered saline

PEG polyethylene glycol

pfu particle forming units

pKa acid dissociation constant

polyNIPAM-Aac poly(N-isoproplyacrylamide-acrylic acid)

QD quantum dot

R6G rhodamine 6G

TBE tris buffered EDTA

TBS tris buffered saline

TBST tris buffered saline - Tween

TEM transmission electron microscopy

TMB tetramethyl benzene

TOPO tri-noctylphosphine oxide

TPP sodium trypolyphosphate

Tris tris(hydroxymethyl)aminomethane

X-Gal bromo-chloro-indolyl-galactopyranoside

xiv

Academic Contributions

Articles published in refereed journals

1. Mardyani S., Chan W.C.W., Quantification of quantum dots using phage display

screening and assay. Journal of Materials Chemistry, 2009 19 (35): 6321-6323.

2. Lee A.J., Hung A., Mardyani S., Rhee A., Klostranec J.M., Mu Y., Li D., and

Chan W.C.W., Toward the accurate read-out of quantum dot barcodes: Design of

deconvolution algorithms and assessment of uorescence signals in buffer. Advanced

Materials, 2007 19(20):31133118.

3. Zheng J., Ghazani A. A., Song Q., Mardyani S., Chan W.C.W., Wang C., Cellular

Imaging and Surface Marker Labeling of Hematopoietic Cells Using Quantum Dot

Bioconjugates, Laboratory Hematology, 2006 12 (2): 94-98.

4. Jiang W., Mardyani S., Fischer H., Chan W.C.W., Design and characterization

of lysine cross-linked mereapto-acid biocompatible quantum dots. Chemistry of

Materials, 2006, 18 (4): 872-878.

5. Zhang H., Mardyani S., Chan W.C.W., Kumacheva E, Design of biocompatible

chitosan microgels for targeted pH-mediated intracellular release of cancer thera-

peutics. Biomacromolecules, 2006, 7 (5): 1568-1572.

6. Das M., Mardyani S., Chan W.C.W., Kumacheva E., Biofunctionalized pH-

Responsive Microgels for Cancer Cell Targeting. Advanced Materials, 2006, 18

(1): 80-83.

7. Jiang W., Fischer H., Papa E., Mardyani S., Chan W. C. W., Semiconductor

Quantum Dots as Contrast Probes for Whole Animal Imaging. Trends in Biotech-

nology, 2004, 22 (12), 607-609.

xv

Other refereed contributions

1. Mardyani S., Mu Y., Li D., Chan W.C.W., Effects of Biological Buffers on

the Photoluminescence of ZnS-Capped CdSe Quantum Dots (10/2006). BMES,

Chicago, IL. Presented Orally.

2. Mardyani S., Mu Y., Chan W.C.W., High-Throughput Identification of Targeting

Molecules using Quantum Dots, Phage Display, and Flow Cytometry (10/2005).

Canadian Chemical Engineering Conference, Toronto, Ontario. Presented Orally.

3. Zheng J., Song Q., Mardyani S., Chan W.C.W., Wang C., (05/2004) Long-term

and Live Cell Labeling of Hematopoietic Cells Using Quantum Dot Bioconjugates.

International Symposium on Technological Innovations in Laboratory Haematology,

Barcelona, Spain. Presented Orally. Won best paper and best presentation award.

4. Mardyani, S., Singhal, A., Jiang, W., Chan, W.C.W., Interfacing peptides, identi-

fied using phage-display screening, with quantum dots for the design of nanoprobes,

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 5705, pp. 217-

224, 2005.

5. Wang C., Zheng J., Mardyani S., Chan W.C.W., Intracellular and surface marker

labeling of hematopoietic cells using quantum dot bioconjugates, Experimental

Hematology, 33 (7): 121-121 317 Suppl. 1, JUL 2005 (meeting abstract.)

6. Jiang W., Mardyani S., Fischer H., Singhal A., Zheng J., Song Q., Wang C.,

Chan W. C.W., (02/2004) Biological Applications of Semiconductor Quantum

Dots. CIHR/NSERC 2nd Annual Nanomedicine Meeting, Toronto, ON.

xvi

Non-refereed contributions

1. Mardyani S., Jiang W., Lai J., Zhang J., Chan W. C. W., (2004) Bioinspired Ap-

proaches to Building Nanoscale Devices. ”Biological Nanostructures and Applica-

tions of Nanostructures in Biology,” M. Stoscio, M. Dutta eds., Kluwer Publishing.

pp.149-160.

2. Jiang W., Singhal A., Fischer H., Mardyani S., Chan W. C. W., (2004) Engineer-

ing Biocompatible Quantum Dots for Real-time, Ultrasensitive Biological Imaging

and Detection. S. Bhatia, T. Desai, eds. Kluwer Publishing. pp. 137-156.

Patents

1. Chan W. C. W., Fischer H., Mardyani S., Jiang W., Large-Scale Processing of

Biocompatible, Facile-reactive Powdered Semiconductor Quantum Dots. Patent

number: 20060014315.

xvii

Chapter 1

Introduction

Applications of nanotechnology in medicine aim to diagnose and treat diseases by working

at the nanometer scale where biological macromolecules exist and operate [1, 2]. In the

past decade, nanotechnology has emerged as a new and powerful weapon for the detection

and treatment of cancer [1, 3, 4, 5]. Nanoparticles have unique and tunable properties

[6, 7, 8]. When these particles are conjugated to biorecognition molecules, they can home

in to a molecular target (such as a cancer marker) and highlight the presence of cancer

for use of cancer detection [9, 10, 11], or deliver a cytotoxic agent for use in targeted

therapy [12, 13, 14].

This thesis will describe observations, techniques and strategies that can be used

towards the development of nanoparticle based detection and treatment of cancer. In

the realm of detection, two projects will focus on quantum dots, the variability of their

fluorescence in commonly used buffers, and methods to quantify them based on their

surface chemistry. Two projects will also describe the use of phage display screening

to identify biorecognition molecules, which are necessary to confer nanoparticles with

targeting capabilities. The final project, which is in the realm of treatment, describes the

rational design of targeted drug delivery vehicles using two types of polymer nanogels.

This chapter will provide background on nanotechnology, nanoparticles in diagnosis,

1

Chapter 1. Introduction 2

biorecognition molecules and the design of targeted drug delivery vehicles. The first

section gives a brief definition and introduction to nanotechnology. This is followed by

background on the use of nano particles in detection, with a focus on quantum dots. The

subsequent section discusses biorecognition molecules and methods for discovering novel

targeting molecules. This is followed by a section that gives background on the design of

nanoparticle-based drug delivery vehicles. The final section provides an outline of how

the projects in this thesis are laid out and connected.

1.1 Nanotechnology

Nanotechnology is the design of objects with dimensions conveniently described in units

of nanometers (10-9m) [15]. The dimensions of nanoparticles lie between the width of a

DNA double helix (∼2 nm) [16] and the size of viruses (tens to hundreds of nanometers).

Figure 1.1 illustrates the scale of several biologically relevant entities in comparison to

some nanoparticles.

Objects in the nanometer size regime often exhibit properties that are not found in

bulk materials of the same composition. In many cases, these emergent properties are

tunable, meaning that they can be adjusted by changing the shape, size or composition

of the nanoparticle [18, 19]. For example, quantum dots are semiconductor nanocrystals

whose fluorescence emission changes according to size [6]. Smaller quantum dots emit

light at shorter wavelengths and larger quantum dots emit light at longer wavelengths.

Under the same fluorescence excitation, zinc sulfide-capped cadmium selenide quantum

dots with a 2.3 nm core diameters emit blue light (λ ∼ 490 nm), 4.2 nm core diameters

emit yellow light (λ ∼ 560 nm) and 5.5 nm core diameters emit red light (λ ∼ 630nm)

[20].

Another example of size-dependent optical properties is gold nanoshells. Halas and

co-workers have shown that nanoparticles composed of a dielectric core and gold shell


Figure 1.1: The scale of nanoparticles compared to biological entities. Above: the size

of representative naturally occurring molecules [16, 17], organisms and objects between

1 nm to 1,000,000 nm (1 mm). Below: size of representative synthetic nanoparticles

between 1 nm - 1,000 nm (1 µm).


(called gold nanoshells) have absorbance spectra that can be tuned by varying the core

diameter to shell thickness ratio [21]. The higher this ratio (i.e. the thinner the shell,

relatively) the higher the optical resonance wavelength of the nanoshell [21]. Hirsch et.

al. tuned the nanoshells to absorb in the near infra-red window [7]. This is the region

of the electromagnetic spectrum where light absorption by hemoglobin and water are at

a minimum and photons are able to penetrate deeper into tissue (500 µm to cm) [22].

These nanoshells were injected into the tumours of mice, after which an external NIR

laser was used to excite the nanoshells, causing them to release heat and kill surrounding

tumour tissue [7].

Other nanoparticles, such as gold nanorods have absorbance spectra that are depen-

dent on the shape of the particle. In the case of gold nanorods, the peak absorbance

wavelength can be adjusted by changing the rod’s aspect ratio [23]. Like nanoshells, gold

nanorods also convert absorbed light into heat. They have been used for thermal therapy

[24] and thermally enhanced chemotherapy [25] where nanorods in neoplastic tissue are

excited by a near infrared laser outside the tissue to emit heat in the tumour region.

The size and unique properties of nanoparticles make them well suited for many bi-

ological applications [26, 27]. For cancer, which in 2007 surpassed major cardiovascular

disease as the leading cause of death in Canada [28], nanotechnology has the potential

to improve patient prognosis through earlier and more accurate diagnosis as well as con-

trolled and targeted delivery of cancer therapeutics. Already, ultra-small paramagnetic

iron oxide particles (USPIO) 50 nm in diameter have been used clinically as contrast

agents in magnetic resonance imaging for the detection of tumors [29]. In cancer thera-

peutics, several nanoparticle formulations of chemotherapeutic drugs, such as Abraxane

and Doxil have been approved for clinical use [4]. In these novel therapeutics, known

chemotherapeutic drugs (paclitaxel in Abraxane and doxorubicin Doxil) are delivered

in the form of a nanoparticle for improved pharmacokinetic distribution and reduced

toxicity [1, 4].


1.2 Nanoparticles in detection and diagnosis

Treatment for cancer is most effective if the disease is detected early, before it has metas-

tasized or spread beyond the site of the initial tumour. To illustrate: in the United

States, the 5-year relative survival rate for localized breast cancer is 97% [30]. If the

cancer has spread regionally, the survival rate dips to 78% [30]. Once metastasis has

occurred, this survival rate plummets to 23% [30].

The early stages where treatment is most effective, however, are often asymptomatic.

Early detection strategies use imaging systems to detect physiological and anatomical

changes concomitant with oncogenesis, or molecular probes to detect cancer biomarkers

such as proteases, antigens, antibodies, proteins and nucleic acid based markers [31].

Nanoparticles have been used as contrast agents in both strategies.

As MRI contrast agents, superparamagenetic iron oxide nanocrystals are used to

detect liver metastases and metastatic lymph nodes. Upon intravenous injection, the

nanocrystals are phagocytosed by macrophages and monocytes, which take them to the

liver and lymph nodes. Diagnosis is made by observing the differential accumulation of

the nanoparticles in metastatic and inflamed tissue through MRI. [29]

The size and unique properties of nanoparticles make them particularly suited for

use as molecular probes. In fact, one of the earliest demonstrations of nanoparticles in a

biomedical application used 13 nm gold colloids as colorimetric DNA sensors. Mirkin and

colleagues conjugated the gold colloids to oligonucleotides. In the absence of complemen-

tary DNA, the solution of colloids remains monodisperse, forming a deep red suspension.

When complementary single-stranded DNA sequences are introduced to the solution,

they form bridges between the gold colloids, causing them to aggregate. The solution of

aggregated gold colloids appears blue. This forms the basis of a simple genomic detection

system. [32]

Similar in size to macromolecules and viruses, nanoparticles can also access molecular

markers in situ [9]. Many nanoparticles interact strongly with electromagnetic radiation


from within and beyond the visible light spectrum and are thereby able to generate a

strong signal. Tunable properties mean that probes can be made to order (e.g. to have

an absorbance or fluorescence peak at a specific wavelength) and a series of probes can

be made and combined in multiplexed labeling. Because they exemplify all of these

useful qualities, quantum dots in particular have drawn a great deal of interest for their

potential applications in cancer diagnosis [9, 33, 34, 35, 36, 33].

1.2.1 Quantum dots

Quantum dots are semiconductor nanocrystals whose optical and electronic properties

are strongly size-dependent [6]. Larger quantum dots have a smaller bandgap and emit

lower energy photons (towards the red side of the spectrum) and smaller quantum dots

have a larger bandgap and emit higher energy photons (towards the blue side of the

spectrum). Quantum dots of various semiconductor compositions, such as ZnS, CdS,

ZnSe, CdTe and PbS, have size-tunable fluorescence emission that range between the

UV and the infrared [37].

Quantum dots can replace organic dyes in many applications and offer several advan-

tages over organic dyes, such as photostability, long fluorescence lifetime, broad absorp-

tion spectra and narrow, tunable emission spectra [38]. These properties make quantum

dots well suited to applications such as single molecule tracking [39], time-gated imaging

[40] and multiplexed imaging where organic dyes would be of limited use due to their

photobleaching, short fluorescence lifetime and narrow absorption spectra [34]. Their

photostability and tunable emission spectra also enables long-term multicolour imaging,

which can offer real-time insights into how cells and proteins interact [41]. They have

been used to track metastasis [33]. The ability to tune their fluorescence into the near

infrared window has also enabled their use in optically guided surgery [36] and in vivo

cancer imaging [9].

One of the most useful advantages that quantum dots have over organic dyes is their


broad absorption spectra, coupled with their narrow, tunable emission spectra. Tradi-

tional organic dyes have narrow absorption spectra, and a small Stokes shift (difference

between the absorption peak and emission peak) necessitating multiple lasers and filters

for their use in multicolour imaging. Quantum dots, on the other hand, all have high

absorption in the UV region so different colours of quantum dots may be excited by

a single light source and imaged simultaneously. This opens the door for multicolour

imaging [41] and labeling of multiple intracellular targets simultaneously [35].

The labeling of multiple disease markers with quantum dot-based barcodes [42] could

pave the way towards more sensitive and accurate disease detection systems. Quantum-

dot based probes may be used to study the heterogeneity of disease markers and link

them to prognosis for improved diagnostic strategies.

Chapter 2 describes the challenge of working with nanoparticles with properties that

change in response to their environment. This project describes the fluctuations in flu-

orescence intensity of zinc sulfide-capped cadmium selenide quantum dots in biological

buffers. Previous studies have shown that fluorescence fluctuations exist in ZnS-capped

CdSe quantum dots in non-polar solvents [43]. In addition, fluorescence fluctuations

of less commonly used CdTe quantum dots have been studied in aqueous solvents [44].

Studying this phenomenon in commonly used ZnS-capped CdSe quantum dots in biologi-

cally relevant buffers highlighted potential challenges in future applications and prompted

the development of strategies to mitigate it.

Chapter 3 addresses the challenge of detecting and quantifying nanoparticles based

on their surface chemistry. Surface chemistry plays a vital role in the interaction between

a nanoparticle and the biological environment, since it is the surface of the nanoparticle

that is ’seen’ by the biological system rather than the core. Bacteriophage expressing

peptide fusions were identified that bound specifically to quantum dots with different

surface chemistries. These phage were then used in an assay to quantify quantum dots

analogous to the way antibodies are used to detect and quantify antigens.


1.3 Biorecognition molecules

Biorecognition molecules are necessary to direct nanoparticles, such as quantum dots, to a

molecular target. Certain pairs of molecules, such as enzymes and substrates, antibodies

and antigens or complementary chains of DNA, bind strongly and specifically to each

other and not to other molecules in the environment. Biorecognition uses this molecular

complementarity, which is found ubiquitously in biological systems.

To give a nanoparticle homing or targeting capabilities, the nanoparticle is conjugated

to one member of a molecular complement pair, also know as a targeting or biorecognition

molecule. This enables the nanoparticle to specifically bind to the other member of the

complementary pair, also known as the target. This specificity to a molecular target can

be used in diagnosis, where the nanoparticle labels a molecular marker. It can also be

exploited in targeted therapy, where the nanoparticle carries therapeutic elements and

releases them at the target site. The use of biorecognition molecules with nanoparticles

for diagnostic and therapeutic applications is illustrated in Figure 1.2.

There are several strategies for identifying new targeting molecules. This section de-

scribes some of the most common methods of identifying targeting molecules: polyclonal

antibodies, monoclonal antibodies, aptamers, and phage display.

1.3.1 Antibodies

Antibodies are molecules produced by an organism’s immune system in response to an

immunogenic invader or antigen. Antibodies bind specifically to their target antigen,

making them a natural targeting molecule.

The population of antibodies produced by an immunized animal is polyclonal in

that it recognizes collectively all the antigens an animal has been exposed to in the

past. The magnitude of the immune response produced by the antigen determines the

proportion of the polyclonal antibody population that is directed against that antigen.


Figure 1.2: Nanoparticles for cancer diagnostics and therapy. Biorecognition molecules

confer targeting capabilities to nanoparticles such as quantum dots, which can then be

used for detection. Nanoparticles such as polymer microgels can be used to deliver drugs

to cancer cells.


The diversity of the polyclonal antibody population can lead to a lack of specificity

towards the target antigen due to antibodies cross-reacting with non-target antigens.

Also, because production of polyclonal antibodies occurs in living animals, this technology

is limited by the lifespan of the animal and the amount of serum that can be extracted.

[45]

The major drawbacks associated with polyclonal antibodies were overcome in 1975

when Georges Kohler and Cesar Milstein reported the use of hybridomas for producing

monoclonal antibodies [46]. Antibodies are produced in B lymphocytes with each B lym-

phocyte producing one specific antibody. Because of their short lifespan, primary cultures

of B lymphocytes have limited usefulness for producing monoclonal antibodies. Kohler

and Milstein overcame this barrier by fusing B lymphocytes with mutant myeloma cells

from an immortal cell line. After immunizing a mouse with an antigen, B lymphocytes

are harvested from the spleen of the mouse. These B lymphocytes are then fused with

the cells from the immortal cell line. The viable hybrid cells produced by the fusion

are selected and grown in separate wells, where they produced hybridoma colonies, each

producing a single type (monoclonal) of antibody. The wells are then tested to determine

which colonies produce antibodies for the target antigen. These resulting colonies can be

grown in a large culture from which substantial quantities of pure monoclonal antibodies

can be harvested. [47]

Monoclonal antibodies are commonly used in immunofluorescence microscopy, to label

components of a cell or tissue; affinity chromatography, to separate proteins in complex

mixtures; and in diagnostic and therapeutic applications [48]. Monoclonal antibodies are

used to identify tumor markers in cancer detection, to bind and inactivate toxic proteins

and to interrupt signaling processes in the treatment of cancer [49, 45, 50].


1.3.2 Aptamers

Aptamers are functional, binding oligonucleotides (nucleic acid chains) that have been

identified through a process of in vitro screening [51, 52]. Here, a library of randomized

RNA is screened for binding affinity to a protein target. Oligonucleotides that bind to

the target are amplified using the polymerase chain reaction (PCR) and the screening

is repeated. Multiple rounds of screening and amplification will yield an aptamer that

binds with greatest affinity to target protein.

1.3.3 Phage display

Phage display is a powerful tool to find targeting molecules based on their binding affinity.

Phage display screens a library of genetically modified bacteriophage to find a peptide

that binds to a target with the greatest affinity [53].

Bacteriophages are viruses that infect bacteria. They are composed of DNA inside a

protein coat. The DNA inside the phage encodes the amino acid sequences of the proteins

displayed on the protein coat of the phage. Libraries of several million different phages

can be created through random mutations of an insert into the phage DNA. Phage display

subjects these libraries through a process of panning and amplification that selects for

clones that bind most effectively to the target protein.

Figure 1.3 illustrates the phage display panning process on one target protein. First,

the target protein is adsorbed onto the surface of the dish to create an affinity matrix.

Then, the phage display library is incubated with the target protein. After incubation,

the unbound phage are washed away and the bound phage is left behind. These phage

are eluted from the proteins through acid denaturation or competition binding. The

eluted phage are then amplified by infecting and multiplying in bacteria [54]. The new,

amplified phage population is once again panned against the target protein to further

enrich the phage population for phage with high binding affinity to the target. After


Figure 1.3: Phage display screening process. A library of phage expressing random

peptides is incubated with a protein target, which is adsorbed onto a solid substrate.

Unbound phage are then washed away, leaving bound phage, which are eluted by an

acidic buffer or by an excess of a known ligand to the target. The eluted phage are

amplified in bacterial host and harvested or used for another round of panning.


three or more rounds of panning, the DNA of the remaining phage are sequenced to

determine the amino acid sequence of the binding peptide.

Through this process of selection and amplification, the phage expressing a peptide

that binds to the affinity matrix comes to dominate the total phage population and

can thereby be selected against a high background of non-specific phage. By expressing

random peptides on a library of phage, and using affinity purification and amplification, it

is possible to identify peptides that bind to a protein target without any prior knowledge

of the nature of the target [55, 56].

Chapter 4 addresses the challenge of identifying new targeting molecules needed for

the growing library of nanoparticle-based labels designed for multiplexed detection and

diagnosis. In this project a biorecognition molecule was isolated using phage display

screening. A peptide was identified, synthesized and coupled to quantum dots for cell

labeling. This project demonstrated the feasibility of identifying new targeting molecules

to complex targets such as whole cells using phage display. The project also highlighted

the challenges with this technique.

1.4 Nanoparticles in therapy: developing drug de-

livery vehicles

Once cancer is detected, the most commonly used therapies to treat it are surgery, ra-

diation and chemotherapy. The challenge in each of these therapies lies in striking the

balance between being aggressive enough to eliminate all the malignant cells but cautious

enough to avoid damaging surrounding healthy tissue. If a treatment is not aggressive

enough, some malignant cells may survive and grow into a new tumour. If it is too ag-

gressive, it may destroy so much healthy tissue that it compromises normal physiological

functions. Treatments such as surgery, radiation and chemotherapy suffer from a lack of

specificity and selectivity.


Specificity is the ability of a therapy to cause the desired effect [57]–in this case,

remove, kill or stop the growth of cancerous cells. If a therapy is not able to effectively

eliminate malignant cells, then it is lacking in specificity. Selectivity is the ability of a

therapy to cause the desired effect only where it is needed [57]. When cancer treatments

harm healthy cells alongside malignant cells, they lack selectivity. Targeted drug delivery

aims to increase both the specificity and selectivity of treatment.

Targeting the delivery of a therapeutic to the malignant cells while avoiding the

healthy cells promises to increase the efficacy of the therapeutic and reduce its side

effects. Chemotherapeutic drugs conjugated to targeting moieties have been used to

successfully treat tumour bearing mice using doses at which the same free, non-targeted

drug did not cause any cancer regression [58, 59]. These targeted drugs were also safe at

higher doses where free drug was lethal [58, 59]. By lowering the effective dose, ED, i.e.

the dose at which a drug produces an effect and by raising the toxic dose, TD, i.e. the

dose at which it becomes toxic, targeting expands the drug’s therapeutic window.

The spatial location of drug delivery is not the only parameter that needs to be

controlled. The timing of the delivery is also an important consideration. Periodic

dosing of drugs results in peaks and valleys in drug concentration in vivo. The peaks may

represent dangerously high levels of cytotoxic chemicals that are prone to compromise

the drug’s selectivity while the troughs could represent levels so low that the specificity

declines.

1.4.1 Nanoparticle properties

There are several parameters in a nanoparticle’s structure and composition that can be

programmed to create a targeted drug delivery vehicle. By modifying the physical and

chemical properties of nanoparticles, we can control their biological properties such as

circulation half-life, immunogenicity, biodistribution, toxicity, biodegradation, method of

excretion, etc. Figure 1.4 summarizes how the nanoparticle’s core composition, shape,


Figure 1.4: Nanoparticle parameters and their implications for rational design of drug

delivery. Nanoparticle parameters that may be modified for therapeutic applications

include size, shape, surface chemistry, core and targeting molecule. These parameters

have implications on the nature of therapy, efficacy of targeting, mechanism of release

and possible side effects.


surface chemistry, size and attached targeting molecules can be manipulated in the design

of a drug delivery vehicle.

Core material

The nanoparticle core may be synthesized from a wide variety of materials in different

geometries and sizes. Materials such as carbon in carbon nanotubes [60] or gold in gold

nanorods [24] or silica and gold in gold nanoshells [7, 61] have been used to deliver

thermal therapy to cancer cells either through cytotoxic application of heat or in synergy

with chemotherapeutics [25]. The poor vascularization of the tumour also makes it more

susceptible to damage by heat than normal tissue, a factor that aids in selectivity.

Chemotherapeutic agents may be carried in the cores of hollow or porous nanoparti-

cles, also called nanocarriers or nanovectors. Liposomes, dendrimers, fullerenes, polymer

nanospheres and mesoporous silica are examples of nanoparticles that have been used to

carry drugs. These materials may also be programmed with mechanisms by which the

drug may be released. For example, drug release may be triggered by heat, light, ion

concentration, ultrasound, radiation or internal triggers like the reduced pH of tumour

interstitia [62] or of an endocytosed vesicle.

Shape

The geometry of the nanoparticle can be altered to affect its overall properties. For

example, the aspect ratio (i.e. the ratio of the length to the width) of a gold nanorod

affects its absorbance spectra, which can be tuned to convert different wavelengths of

light to heat to apply thermal therapy [24] or trigger drug release.

The geometry of the nanoparticle has also been shown to affect its intake into cells

[63, 64]. One method of cellular uptake of nanoparticles is receptor mediated endocytosis,

whereby the nanoparticle first attaches to a specific receptor on the cell surface that

triggers its uptake. 50 by 14 nm gold nanorods had much less cellular uptake by receptor-

mediated endocytosis than spherical colloids with diameters of 50 nm or 14 nm [63]. This

may suggest that spherical objects have a better chance of getting inside a cell than one


with a high aspect ratio. However, carbon nanotubes have been taken up by cells through

non-receptor-mediated endocytosis regardless of the nanotube’s surface functionalization

or cell type [65]. It has been suggested that this is due to the very high aspect ratio of the

nanotubes. However, further research is needed to verify the mechanism of this as other

studies have shown that carbon nanotubes are internalized by cells through endocytosis

[60].

Nanoparticle geometry can also affect its half-life in the blood. The half-life is the

amount of time it takes for the concentration of nanoparticles in the blood to be re-

duced by 50%. In rats, filamentous micelles stayed in circulation for over a week, while

PEG-coated spherical vesicles were removed from the circulation within days [66]. The

local particle shape from the perspective of the phagocyte, at which phagocytes initially

encounter a particle has also been shown to affect whether it phagocytoses that particle.

For example, if a phagocyte makes first contact with a disk-shaped particle by its narrow

edge, it engulfs it within minutes. If, however, it makes first contact along one of the

flat sides of an identical disk-shaped particle, it will spread along that surface and not

internalize the particle, even after two hours. As such, cylindrical, disc-like and hemi-

spherical nanoparticles outperform their spherical counterparts when it comes to evading

phagocytosis [64].

Surface chemistry

Moving out from the core, the nanoparticle’s surface chemistry can also be mediated

for avoidance of the reticuloendothelial system. This would increase its half-life and

thereby increase its availability to the target cells. The surface chemistry can also effect

the amount of cell uptake by target and non-target cells. In addition, as the interface

between the nanomaterial and the biological environment, the surface chemistry largely

determines whether a nanoparticle suppresses or stimulates immune responses [67].

Adsorption of serum proteins onto the surface of the nanoparticle can mark it for

phagocytosis by macrophages, a process called opsonization. The surface chemistry of


the nanoparticle affects the extent and nature of this adsorption [68] and can be modified

to minimize this opsonization [69, 70]. Without surface modification, nanoparticles have

been reported to be removed from the bloodstream within seconds by macrophages–cells

which remove foreign objects and stimulate other immune cells [71].

Nanoparticle surface charge has been shown to influence the composition of the “hard

protein corona” that forms around the particle incubated in blood plasma. The corona

is composed of proteins adsorbed onto the surface of the nanoparticle. The hard corona

refers to the proteins which adsorb tightly to the nanoparticle as opposed to the soft

corona, which refers to proteins weakly adsorbed and therefore only transiently associated

with the nanoparticle [68]. Positively charged, neutral and negatively charged polystyrene

nanoparticles of different sizes incubated with human plasma, were found to have different

hard protein corona, which included different biologically significant proteins [68].

Addition of highly hydrophilic polyethylene glycol to the surface of the nanoparticle is

an often used strategy to evade the reticuloendothelial system (macrophages) and increase

the blood half-life of a nanoparticle to increase its chances of reaching its intended target

[72, 9]. In one of many examples, CTAB coated gold nanorods are internalized by KB

cells. When the nanorods are coated with polyethylene glycol, this uptake is reduced

nearly 20-fold 3[73]. Alternatively, nanoparticles may be designed for macrophage uptake,

which can be used as a carrier to take the nanoparticles to the target site [29].

Surface chemistry has been shown to modulate cellular uptake of gold nanorods and

colloids [74, 69]. Nanoparticles coated with anionic and hydrophobic groups organized

in sub-nanometer striations could pass through the plasma membrane without disrupt-

ing the phospholipid bilayer [69]. In vitro comparisons between uptake of nanorods by

HeLa cells in media with serum and serum-free media show differences that indicate that

serum proteins adsorbed onto the nanoparticle also affect uptake [74]. In vivo, surface

chemistry has been shown to affect the biodistribution of quantum dots injected intra-

venously into rats. After 90 minutes, the liver was shown to take up 40% of the dose of


mercaptoundecanoic acid-coated lysine crosslinked quantum dots and 99% of the dose of

bovine serum albumin coated quantum dots [75].

Size

Finally, the overall size of the nanoparticle drug delivery system may be tuned for

delivery to the tumour site and then to an intracellular target. Most currently avail-

able nanoparticle formulations of chemotherapeutic drugs target tumours through the

enhanced permeability and retention (EPR) effect. This phenomenon, which allows

nanoparticles to accumulate in a tumour site, occurs because of the unique architecture

of tumour vasculature.

Growing tumours have poorly-made vasculature, characterized by a tortuous archi-

tecture with gaps in the vessel walls, making the vasculature “leaky” [76]. Particles under

200nm in diameter can traverse through these gaps in the vasculature and accumulate

in the tumour interstitia [77]. Poor lymphanic drainage in the tumour further enhances

the selective accumulation of nanoparticles in the tumour interstitia [78].

The size of the drug delivery vehicle may also affect the type of proteins that are

adsorbed onto it [68] with implications on how the immune system responds to the

particle. In addition, nanoparticle size can mediate the efficiency of its uptake into cells.

It has been shown that maximum cellular uptake is achieved with 50nm particles [79, 63].

The attachment of several targeting moieties side-by-side on a nanoparticle increases the

avidity (i.e. the combined strength of multiple bond interactions) compared to free

targeting moieties.

Finally, nanoparticle size has been shown to affect renal clearance. This is important

because after the targeted site is saturated with targeted nanoparticle therapeutics, it is

desirable to clear excess nanoparticles from the bloodstream to avoid any future compli-

cations they may cause. Particles under 5.5 nm in diameter were found to undergo renal

clearance [70].


1.4.2 Active Targeting

In contrast to passive targeting, where nanoparticles accumulate in the tumor due to

EPR, active targeting utilizes biorecognition molecules conjugated to nanoparticles to

confer them with the ability to bind to molecular markers. Antibodies [80], antibody

fragments [81], proteins [82], aptamers [83], peptides [84] and small molecules [85] have

been used against integrins and receptors that are overexpressed in transformed cells or

tumours undergoing angiogenesis. In a strategy akin to the Trojan horse, nanoparticles

may also be coated with nutrients, such as folic acid or glucose, which rapidly growing

cancer cells need to support their heightened metabolism. Chapter 6 describes the de-

velopment and testing of two polymer based drug delivery vehicles which utilize the iron

carrying protein, transferrin, for active targeting.

Chapter 6 addresses the challenge of designing a nanoscale drug delivery vehicle within

the multiple design constraints and opportunities of the biological environment. Design

constraints include the need for a size small enough to fit in through the pores of porous

tumour blood vessels, the requirement to direct the drug carrier to its target, and the

need to release the drug from the carrier once it reaches the target. Receptor-mediated

endocytosis was exploited to transport the drug delivery vehicle into the cell and trigger

the release of the drug. PolyNIPAM acrylic acid and chitosan nanogels (nanometer sized

polymer hydrogels) were tested.

1.5 Thesis outline

The projects described in this document fall under three general categories: 1. Character-

ization/Quantification of quantum dots–towards diagnostic applications 2. Identification

of biorecognition molecules 3. Design of targeted drug delivery vehicles

Figure 1.5 uses the diagram from Figure 1.2 to illustrate how the projects in each

chapter fit into these categories.


Figure 1.5: Schematic summary of projects pursued towards diagnostic applications of

quantum dots and the design of targeted drug delivery vehicles. Chapter 2 examines

the effect of the biological environment on the properties of a nanoparticle. Chapter 3

describes screening for and employing biorecognition molecules to quantify nanoparticles.

Chapters 4 and 5 focus on identification of biorecognition molecules, with whole cell phage

display in chapter 4 and high throughput phage display in chapter 5. Chapter 6 describes

the rational design of targeted drug delivery vehicles.

Chapter 2

Variability of Quantum Dot

Fluorescence

Biological environments can have unexpected effects on nanoparticle properties. This

chapter examines the fluorescence of ZnS-capped CdSe quantum dots of different sizes

and capping thickness in several commonly used biological buffers. The fluorescence of

these nanocrystals changed significantly under the effects of different buffers, different

pHs and different buffer concentrations. The fluctuations in fluorescence of quantum

dots will have important implications in their use as fluorescent probes in quantitative

biological studies.

Some of the findings from this study were published in A.J. Lee, A. Hung, S. Mardyani,

A. Rhee, J.M. Klostranec, Y. Mu, D. Li, and W.C.W. Chan: Toward the accurate read-

out of quantum dot barcodes: Design of deconvolution algorithms and assessment of flu-

orescence signals in buffer. Advanced Materials. 2007. 19(20):3113 - 3118. Copyright

Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

22

Chapter 2. Variability of Quantum Dot Fluorescence 23

2.1 Introduction

Fluorescence-based techniques such as immunohistochemistry, flow cytometry, fluores-

cent microscopy, and DNA microarrays have become cornerstones of biological research.

Recently, quantum dots (QDs) have emerged as a new class of fluorescent probes, which

overcome many of the shortcomings of organic fluorophores. In particular, the wide ab-

sorbance spectra of QDs, their brightness, narrow and tunable emission spectra [86, 20],

and resistance to photobleaching [82] make them very attractive for multiplexed, ultra-

sensitive or long-term studies where organic fluorophores cannot be used. QDs have

been used for long-term labeling and imaging [87, 39, 41], optically guided surgery [36],

deep tissue cancer detection [9], tracking cancer metastasis [33], and optical barcodes

[42, 88]. For QDs to be applicable beyond the proof of concept, it is important to under-

stand the effects of their external environment on their fluorescence. This is especially

pertinent when QDs are used in biological applications, where they will be exposed to

various buffers or different microenviroments and chemical gradients in the living cell.

All of these could potentially affect the fluorescence of QD probes and hence could lead

a researcher to make an incorrect conclusion if these conditions are not considered.

Since the late 1980s, extensive studies by the Ellis group have shown that the lumi-

nescence of CdSe bulk single crystals can be quenched or enhanced by adsorbed ligands

on the surface [89, 90, 91, 92, 93]. Several groups have also shown that adsorbed ligands

affect the fluorescence of CdSe nanocrystals [94, 95, 96, 97]. However, these studies have

been mostly restricted to semiconductors exposed to organic solvents instead of aqueous

solvents, which are more complicated. Only recently has the effect of biologically relevant

molecules on the photoluminescence of QDs been studied.

Nadeau and colleagues showed that bare CdSe QDs were more sensitive to changes

in non-polar solvents than ZnS-capped CdSe QDs [43]. They also showed that water

solubilization of both CdSe and ZnS-capped CdSe QDs caused quenching, especially in

the bare QDs [43]. In a study on the effect of biochemical buffers on CdTe nanocrystals,


Boldt et al. found that highly acidic buffers extinguish photoluminescence through de-

struction of the nanoparticles and high dilution extinguishes photoluminescence due to

desorption of surface ligands that passivate the nanocrystals [44]. Their work, however

was done on CdTe nanocrystals, which are not the predominant QDs used in biological

applications. The present study reports the changes of photoluminescence emission of

bare CdSe and ZnS-capped CdSe QDs, in a number of commonly used aqueous buffers

and environments. These findings have implications in applications of QDs as biological

probes where their photoluminescence intensity may be used to indicate the presence or

measure the quantity of a target molecule.

The most commonly used QD probes in biological applications are CdSe nanocrys-

tals with a ZnS capping, synthesized through either an organometallic approach using

dimethyl cadmium as a precursor, or a greener approach, where dimethyl cadmium is

replaced with cadmium oxide [37, 98]. The ZnS capping stabilizes and increases the flu-

orescence of the QDs by passivating surface traps, [86, 20] maintains the integrity of the

QD core in an oxidative environment [20] and minimizes cytotoxicity. Only a few studies

have been conducted on the effect of aqueous solvents and environments on their photolu-

minescence emission [43, 99, 100]. In this study, the optical and physical properties of the

CdSe and ZnS-capped CdSe QDs are described in Table 1. Rather than modifying the

surface chemistry of the QDs to render them water-soluble, we elected to infuse the QDs

into polystyrene beads for interfacing with aqueous buffers and characterization. This

prevents non-uniform aggregation of QDs in solution and its associated problems with

quantitative measurements. This also makes the fluorescence of QDs easy to characterize

using the technique of flow cytometry, and polystyrene beads are also easy to manipulate

and can be washed and resuspended in different buffers. Furthermore, these QDs are

coated by the surfactant tri-noctylphosphine oxide (TOPO), which is weakly bound to

the QDs [97]. This allows us to gauge the interactions of ions with the ZnS-capped CdSe

surface.


2.2 Materials and Methods

2.2.1 Synthesis of ZnS-capped CdSe Nanocrystals

ZnS-capped, CdSe QDs were synthesized using a previously described organometallic

procedure [101]. For the study of the variability of quantum dot fluorescence in biological

buffers, the green (528 nm emission), yellow (565 nm emission) and orange (585 nm

emission) CdSe QD cores were made in three separate reactions. The green, yellow,

and orange description refers to the fluorescence colour emission of the QDs. In each

reaction, once the core was made, the vessel was cooled to 270 ◦C and the capping solution

(consisting of diethylzinc and hexamethyldisilathiane in tri-n-octylphosphine) was added

to produce the ZnS capping layer on the CdSe core. To make QDs with the same core

but different amounts of the capping solution, aliquots were taken after the addition of

0.0 ml, 1.0 ml, 2.0 ml, 3.0 ml of capping solution.

2.2.2 Composition analysis through inductively coupled plasma

spectroscopy

The composition of the QDs was analyzed by inductively coupled plasma atomic emission

spectroscopy. QDs were initially dried in air and then dissolved in 0.5ml of 70% nitric

acid. The solution was then diluted to 5.0 ml. Calibration standards containing Cd, Se,

Zn, and S were prepared concurrently with the samples.

2.2.3 Immobilization of QDs in polystyrene microbeads

QDs dissolved in chloroform were mixed in propanol at a 15% V/V ratio to make a

solution with a final QD concentration of at least 3µM. QD concentrations were mea-

sured using Beers law and spectrophotometry absorbance measurements using standard

procedures [102]. 5µm polystyrene microbeads from Bangs Laboratories were swelled in


the solution of solvent and QDs and the solution was shaken overnight. The microbeads

were mixed with an excess of 108 QDs per bead. QDs enter the pores of the polystyrene

microbeads and remain embedded in the polystyrene through hydrophobic-hydrophobic

interactions. The QD-stained microbeads were then washed three times in propanol,

dried and resuspended in aqueous buffers of various concentrations and pH. They were

incubated overnight at room temperature and then the fluorescence intensity was mea-

sured through flow cytometry. QD-encoded beads were embedded in resin, sliced by a

microtome and imaged by TEM.

2.2.4 Flow cytometry and statistical analysis

A Coulter Epics XL flow cytometer was used to measure the fluorescence of the QD-

stained beads in various buffers. The flow cytometer measured the fluorescence emision,

side scatter and forward scatter of 10,000 particles in each sample. As some of these

particles may be aggregates of microbeads, broken microbeads, or aggregates of quantum

dots, forward scatter and side scatter, which can be correlated to the size and granularity

of particles, were used to single out the population of monodisperse microbeads from

larger aggregates and smaller particles. Only the fluorescence of these single microbeads

was used in further analysis.

2.3 Results and Discussion

Table 2.1 summarizes the properties of the QDs used in this study including the emission

peak wavelength, full width half maximum (FWHM), and ratio of zinc to cadmium,

which is indicative of the amount of capping in each QD sample.

Figure 2.1 shows a transmission electron micrograph of a sliced quantum dot-encoded

bead. Elemental analysis using the transmission electron microscope showed high con-

centrations of cadmium co-located with the bright spots on the periphery of the bead


Figure 2.1: TEM of cross section quantum dot-encoded bead shows that QDs, indicated

by bright spots, are mainly on the outer edge of the bead.


Table 2.1: Properties of ZnS-capped CdSe QDs used in this study

QD Emission Peak (nm) FWHM (nm) Zn:Cd

G0 524 31 0 : 1

G1 534 31 0.19 : 1

G2 536 29 0.36 : 1

G3 537 29 0.45 : 1

Y0 564 27 0 : 1

Y1 567 32 0.08 : 1

Y2 565 32 0.17 : 1

Y3 567 32 0.24 : 1

O0 587 30 0 : 1

O1 593 34 0.04 : 1

O2 587 33 0.15 : 1

O3 589 33 0.28 : 1

suggesting that these are quantum dots. Quantum dots at the periphery of the bead are

more likely to be exposed to elements in the local environment.

The following commonly used buffers were chosen for this study: HEPES, PBS, Tris

and carbonate buffer. The molecular structures of these buffers are shown in Figure

2.2A. HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) and PBS (phosphate

buffered saline) are widely used in cell culture to maintain physiological pH. Tris (2-

amino-2-(hydroxymethyl)-1,3-propanediol) is used for many applications including gel

electrophoresis, western blots, phage display screening, and immunoassay because it has

a large buffering range. Carbonate buffer is also widely used in biomolecule detection

schemes.

Figure 2.2B shows the relative fluorescence of G1, Y2 and O2 – QDs with similar


Figure 2.2: Variability of QD fluorescence in buffers. A) Chemical structures of the

main molecules in the buffers used in this study. B) Variation in fluorescence intensity

of different colored (green, yellow and orange) QDs, with similar zinc to cadmium ratios

(between 0.15:1 and 0.19:1). Error bars indicate the 95% confidence interval on the mean

normalized fluorescence.


levels of zinc sulfide capping (Zinc-to-Cadmium ratio = 0.15:1 to 0.19:1). The three

different colours of QDs show the same progression of fluorescence in the four different

buffers, where fluorescence intensity is as follows: PBS < carbonate < Tris < HEPES.

There is a large difference, however, in the magnitude of variation. Fluorescence varies

around 10%, in G1, Y2 and about 300% in O2. This higher fluctuation in fluorescence

in QDs with higher emission has also been observed by Boldt et al [44]. In their study

of CdTe quantum dots, they found that the variability of fluorescence intensity dropped

markedly in QDs with diameters under 3.06 nm [44]. This was attributed to the more

defined structure and fewer surface traps of smaller nanocrystals [44]. Using the equations

developed by Peng et al., which relate the first absorption peak position (determined from

spectrophotometry absorbance measurements) to the diameter of the quantum dot, the

core diameters of G1 and Y2 are calculated to be under 3.17 nm [102]. Since this current

study uses quantum dots with a CdSe core, it was not expected that this diameter be

under the 3.06 nm cut-off determined by Eychmuller et al.. using CdTe quantum dots.

Increased variability of fluorescence in larger QDs may also be a result of greater

surface area. The surface area of the CdSe core of O1 is 141% greater than G1 and 55%

greater than Y2. This greater surface area allows for more interaction between the QD

and the environment, thereby increasing the effect of the QDs chemical environment on

the intensity of its fluorescence.

The differences in the effects of each buffer on the QD fluorescence may be attributed

to the chemical structure of the buffer. HEPES and Tris contain tertiary and primary

amines, respectively, which have both been shown to enhance QD fluorescence [97, 103].

These amines bind to low energy trap sites on the QD surface, which are linked to non-

radiative recombination. Binding increases the energy of the trap, removing the site as

an efficient trap, thereby increasing fluorescence emission [103]. Conversely, carbonate

and phosphate act as Lewis acids, which have been shown by the Ellis group to decrease

fluorescence emission from CdSe substrates [89].


Figure 2.3: Influence of QD capping on variability of QD fluorescence in buffers. Vari-

ation in fluorescence intensity of green QDs with different amounts of ZnS capping, in

phosphate, carbonate, Tris and HEPES buffers. Error bars indicate the 95% confidence

interval on the mean normalized fluorescence.


Figure 2.3 shows the effect of different buffers on the fluorescence of G1, G2 and G3,

which are green QDs, all synthesized with the same CdSe core but different amount of

ZnS capping. Inductively coupled plasma spectroscopy showed that the capped QDs had

Zn:Cd ratios of 0.19:1, 0.36:1 and 0.45:1. Data from G0, which had no ZnS capping is

not shown because its fluorescence, as well as the fluorescence of other uncapped QDs

(Y0 and O0) was almost completely quenched in all the buffers studied. As seen in

Figure 1 with QDs of different colors and the same amounts of capping, the magnitude

of variation of fluorescence intensity also differs amongst QDs of the same color with

different amounts of capping. Fluorescence varied 102 ± 13.3%, 110 ± 1.6% and 191

± 4.4% in G1, G2 and G3 respectively. In addition, in G1, fluorescence in HEPES was

greater than fluorescence in Tris. This was reversed in G2 and G3, which had more

ZnS capping. In all three samples, the fluorescence in HEPES or Tris was greater than

fluorescence in carbonate and PBS.

The effect of buffer concentration on the fluorescence of QDs was also studied. Figure

2.4 shows the effect of the concentration of Tris buffer on the fluorescence of G1, G2 and

G3, with different amounts of ZnS capping. The magnitude of fluorescence increases as

the concentration of Tris is increased from 0.1 mM to 10.0 mM. The fluorescence then

decreases in the 100.0 mM Tris buffer. This same trend was observed by McLendon

and colleagues in their study of CdS QDs in simple amines [103]. They found that high

concentrations of triethyl amine (>0.1M) quench the fluorescence of QDs [103]. Gas

chromatography showed that the triethyl amine was oxidized [103]. At lower amine

concentrations, however, they found that fluorescence was enhanced significantly [103].

In our study, we found that in different concentrations of Tris buffer, the fluorescence

of the green QDs with different amounts of capping varied between 42% to 131%. In

these different concentrations of Tris buffer, fluorescence intensity of yellow (Y1, Y2, Y3)

and orange QDs (O1, O2, O3) varied 18%-133% and 264%-514% respectively. As seen

previously, the largest, orange QDs display greater variability in their fluorescence than


Figure 2.4: Influence of buffer concentration on QD fluorescence. Variation in fluores-

cence intensity of green QDs with different amounts of ZnS capping, in different con-

centrations of Tris buffer. Error bars indicate the 95% confidence interval on the mean

normalized fluorescence.


Figure 2.5: Variation in fluorescence intensity of green QDs with different amounts of

ZnS capping in citric acid buffer, pH 5 to 7. Error bars indicate the 95% confidence

interval on the mean normalized fluorescence.

their smaller counterparts.

Figure 2.5 shows the fluorescence intensity of G1, G2 and G3 in citrate buffer of pH 5

to pH 7. This range is significant in the biological context as pH levels between 5 and 7

are found in blood, endosomes and tumour interstitia [62, 104]. The fluorescence of the

QDs was found to increase with increasing pH. This trend of increasing fluorescence of

QDs with increasing pH is also seen with other buffers including Tris, HEPES, PBS and

borate. In the 10mM citric acid, these green QDs showed between 40 - 43% change in

fluorescence between pH 5 to 7. Yellow and orange QDs varied in fluorescence between

22% - 345% in citrate buffer with a pH of 5 to 7. Variation in fluorescence intensity due

to pH is seen in many organic fluorophores. For example, the fluorescence intensity of


FITC at 520nm is about 10 times greater at pH 7 than it is at pH 5 [105].

In all the studies described above, spectrophotometry absorbance measurements did

not show any QDs in the supernatant, indicating that the variation in fluorescence in-

tensity was not due to leakage of QDs from the microbeads.

Although this study was done on TOPO-coated QDs, environmentally triggered vari-

ations in QD fluorescence has also been observed in QDs coated with other ligands. QDs

coated with mercapto acids have shown pH dependent fluorescence changes [99, 100]

CdTe QDs coated with thioglycolic acid and mercapto propanoic acid have also shown

variations in fluorescence intensity in different biological buffers [44].

To mitigate the effect of ligand adsorption on changes in photoluminescence, coating

strategies may be developed to insulate the QD surface from small molecules in the envi-

ronment. This approach has been pursued in the design of new QD barcodes, where the

QDs are at the core of a polymer bead and are therefore isolated from their environment

[106]. The fluorescence signals of these barcodes are far more robust against changes in

the buffer environment.

Keeping in mind the possible fluctuations in QD fluorescence, alternative methods of

reading barcodes that are not so dependent on absolute fluorescence intensity may be

developed. Such a deconvolution strategy, which depends on the ratio of colour intensities

in a barcode rather than the absolute values, have now been demonstrated [107].

On a different note, these variations in QD fluorescence may also be seen as a potential

opportunity rather than an obstacle. If they can be harnessed and made consistent,

the QDs could then be used as probes, detecting their environmental conditions and

reporting them with variations on fluorescence. For example, the pH dependence of

FITC fluorescence has been used to measure the pH of organelles [105].


2.4 Conclusion

QDs offer many advantages over traditional organic fluorophores in biological applica-

tions. Their resistance to photobleaching, brightness and unique optical properties open

the doors to studies that could not be done by organic fluorophores. However, we have

shown that the fluorescence of QDs varies in different buffers, different pHs and at differ-

ent buffer concentrations. This has implications in any application of QDs in quantitative

biological assays. Before such studies can be done, the effect of various small molecules

on the QDs’ fluorescence must be known.

The variability of QD fluorescence depending on environmental conditions is an im-

portant consideration for the future development of QD probes and QD applications.

Simple awareness of this phenomenon would help researchers “de-bug” when they ob-

serve unexpected results in their experiments. In addition, strategies can be developed

to overcome or minimize this behaviour, if it is undesirable in a particular application or

to capitalize on it to provide valuable information or insights.

Chapter 3

Quantification of QDs using

biorecognition molecules

This chapter describes a proof-of-concept study, which demonstrated the quantification

of mercaptoacetic acid-coated ZnS-capped CdSe quantum dots (10 nM detection sensi-

tivity and three order linear dynamic range) in biologically relevant solutions using phage

screening and assay.

This work was published in S. Mardyani and W.C.W. Chan. Quantication of quan-

tum dots using phage display screening and assay. Journal of Materials Chemistry,

19(35):6321 - 6323, 2009. Reproduced by permission of The Royal Society of Chemistry

(http://www.rsc.org/).

3.1 Introduction

Nanostructures have been demonstrated in a wide range of applications in medicine and

biology [108, 10, 109]. As the field advances, the ability to accurately quantify nanos-

tructure concentrations is vital. Current methods to quantify nanoparticles are based

on absorbance and emission spectroscopy [102, 43, 110, 107]. However, this technique is

limited by the optical and physical properties of the core material, which could be influ-

37

Chapter 3. Quantification of QDs using biorecognition molecules 38

enced by the molecular environment (e.g., solvent, interfering species). Techniques such

as atomic spectroscopy have good sensitivities but do not differentiate between metals

in the elemental form (e.g, free metal ion) or in the nanoparticle (e.g., components in

the nanoparticle) [75, 111]. Due to these limitations, there is a need to develop novel

analytical methods to quantify whole nanoparticles (including the surface chemistry).

In this study, we demonstrate a proof-of-concept study of using a phage-based assay to

measure semiconductor nanostructures, also known as quantum dots.

In biochemistry, the presence and quantity of a molecule of interest are often deter-

mined using a reporter or probe linked to a targeting molecule, such as an oligonucleotide,

antibody, protein or peptide, that binds tightly and selectively to the molecule of inter-

est. However, the availability of recognition molecules for nanostructures is limited. As

a first step to developing biological assays for measuring nanostructures, one must cre-

ate a library of bio-recognition molecules (e.g., peptides, aptamers, or antibodies) for

nanostructures.

Here, we demonstrate as a proof-of-concept the use of phage-display screening to iden-

tify peptide molecules that can target and detect mercaptoacetic acid coated zinc sulfide-

capped cadmium selenide (ZnS-capped CdSe) quantum dots in various biological envi-

ronment. The technique of phage display for organizing quantum dots in sub-structures

was pioneered by Belcher and co-workers [112, 113, 114]. However, this technique has

not been utilized for developing immunoassays for quantifying nanostructures.

In this application, whole M13 phage particles displaying peptides serve the role of

the primary antibody in an immunoassay. These phage particles were derived from phage

display screening to identify phage which bind to QDs. In the assay, equal aliquots of

each phage clone were incubated in gelatin-coated 96-well plates, which had been coated

with QDs of various surface chemistries. After washing excess phage off the plate the

amount of phage binding to each QD-coated plate was detected using a horse radish

peroxidase(HRP)conjugated anti-M13 antibody. This is similar to an enzyme-linked


Figure 3.1: Phage assay schematic. In an ELISA, a primary antibody binds to the antigen

of interest and is detected by an enzyme-linked secondary antibody. In the phage assay,

the phage particle binds to the quantum dot of interest and is detected by an anti-phage

horse radish peroxidase conjugate.


immunosorbant assay (ELISA), as shown schematically in Figure 3.1.

Figure 3.2 shows a schematic of the procedure used to select and identify peptides

displaying phage that bind to the QDs. First, the QDs were affixed onto a solid substrate.

The substrate surface coating was optimized for maximum QD adsorption. Phage pan-

ning was then performed according to the methods described in New England Biolabs’

Ph.D. 12 Phage display kit manual [115]. After each round of panning, the eluted phage

were titered to monitor the progress of the panning experiment.

After three rounds of panning, individual phage clones were isolated and their binding

analyzed by an enzyme-linked immunosorbant assay (ELISA). Clones which showed selec-

tivity in binding to their intended target compared to QDs with other surface chemistries

were amplified for further characterization, DNA extraction and sequencing.

One phage that binds specifically to MAA-coated QDs was used to quantify QDs

of different surface chemistries by a phage assay. The effectiveness of this phage assay

was then compared with absorbance spectroscopy in quantifying QDs in solution with

optically interfering agents.


3.2.1 Synthesis and characterization of nanoparticles

Zinc sulfide-capped cadmium selenide quantum dots (ZnS(CdSe) QDs) were synthesized

through an organo-metallic procedure using a hexadecylamin-trioctylphosphine oxide-

trioctylphosphine mixture [86, 116].

Water-soluble mercaptoacetic acid (MAA)-coated QDs were made by the procedure

described by Chan and Nie [82].

The organo-metallic QD synthesis produces hydrophobic QDs coated in tri-n-octyl

phosphine oxide (TOPO). To make water-soluble, mercaptoacetic acid-coated QDs, a

1µM solution TOPO-coated QDs in chloroform is mixed with glacial mercaptoacetic


Figure 3.2: Phage display panning. To identify biorecognition molecules for quantum

dots, quantum dots are adsorbed onto a solid substrate and then incubated with a phage

display library expressing random heptapeptides on the pIII protein coat. Unbound

phage are washed away and bound phage are amplified. Through this process, the library

evolves and becomes dominated with clones that bind to the target QDs.


acid in a 3:1 volume ratio and stirred at room temperature for at least two hours. The

solution was then centrifuged at 3000 rpm for 3 minutes. The supernatant was removed

and the pellet resuspended in pH 11 water. To remove excess mercaptoacetic acid, the

QDs were washed with acetone and again resuspended in pH 11 water.

Mercapto-undecanoic acid-coated QDs (MUA QDs) were synthesized according to

the methods described by Jiang et al [100].

Adapting the procedure of Parak’s group, in a total volume of 100 µl, 6 µM QDs

were mixed with BSA at a 1:20 molar ratio [117]. To these solutions, 10 µl of an EDC

solution of appropriate concentration was added to result in a molar ratio of EDC:QD of

128000:1, 64000:1, 32000:1, ... 4:1, 2:1. The solution was incubated at room temperature

overnight with gentle shaking.

Nanoparticles were characterized by UV spectrophotometry and fluorescence spec-

troscopy. In addition, the QDs of various surface coatings were subjected to gel elec-

trophoresis through a 2% agarose gel in TBE buffer. Gels were imaged by GelDoc XR

under epifluorescence with UV excitation.

3.2.2 Creation of the affinity matrix

The affinity matrix is a solid substrate onto which the target used for phage selection is

affixed. In this case, the affinity matrix was composed of QDs adsorbed onto a polystyrene

plate.

Several coatings were tested to enhance the adsorption of QDs onto NUNC microtitre,

polystyrene 96-well plate. For gelatin coating, 0.1% [w/v] gelatin solution in water was

incubated in each well for 10-15 minutes and removed. For poly-lysine coating, polylysine

(Sigma P4707) was incubated in each well for 5 minutes and then aspirated. The plate

was then rinsed thoroughly with water and allowed to dry for at least two hours. For

BSA-coating, 0.5% BSA in 0.1 sodium bicarbonate buffer, pH 8.6 was incubated in each

well for 1 hour at 4 ◦C and then removed.


After coating, the wells were incubated with 1µM of QDs in bicarbonate buffer

(0.1M sodium bicarbonate, pH 8.6) at 4 ◦C. The plates were then washed 6 times with

TBST (0.1% TWEEN) and the fluorescence of each well was measured in a fluorescence

plate reader. The adsorption of QDs onto the plate was also verified by epifluorescence

microscopy, using 350nm excitation and 520LP emission filters.

3.2.3 Phage panning

Phage panning was performed according to the methods described in New England Bio-

labs’ Ph.D. 12 Phage display kit manual and illustrated in Figure 3.2. Briefly, 1 x 1011

pfu (particle forming units) of phage were incubated on the affinity matrix for an hour at

room temperature. Unbound phage were then washed away with 10 washes using TBST

(with either 0.1% or 0.5% TWEEN). The bound phage were then eluted with glycine-

HCl, pH 2.2. These phage were amplified in E. coli and the amplified phage population

was used in the following round of phage panning. After each round of panning the

eluted phage were titered to monitor the progress of the panning experiment.

3.2.4 Screening of clones

Not all the phage obtained from phage display panning have the desired binding charac-

teristics. Thus, it is necessary to screen the clones obtained from phage display panning.

After three rounds of panning, phage were plated on IPTG/X-Gal plates. Individual

plaques were selected and amplified in 1ml of bacterial culture overnight. The bacte-

ria was pelleted and the binding of the phage was tested using ELISA (enzyme-linked

immunosorbant assay).

A gelatin-coated 96-well plate was incubated with 1µM solutions of QDs of various

surface chemistries and incubated overnight. The binding of each phage clone was tested

on each of the different surface chemistries. Phage were incubated in the wells for one

hour. The plate was washed six times with TBST. Anti-M13 horseradish peroxidase


(HRP) antibody was added to the wells and then washed again six times with TBST.

3,3,5,5-tetramethylbenzidine (TMB) substrate was added until a blue colour developed.

Then the reaction was stopped with 2 M sulfuric acid. The optical density of each well

at 450nm-570nm was read by a plate reader. Clones which showed higher binding to

their intended target compared to QDs with other surface chemistries were amplified

and characterized further.

Phage DNA was extracted using QIAGEN’s M13 DNA Extraction kit and sequenced

at The Centre for Applied Genomics at SickKids Hospital in Toronto, Ontario.

3.2.5 Determination of binding characteristics

The binding characteristics of the phage were determined using the ELISA method de-

scribed in section 3.2.4. The concentration of phage incubated in each QD-coated well

was varied, while the concentration of QDs used to coat the wells was kept constant.

The Scatchard plot is used to determine the dissociation constant (KD) between a

receptor and a ligand. In this case, because each bacteriophage has up to five copies of

the binding peptide expressed at one tip and the quantum dots have multiple sites at

which the peptides might bind, we determined the relative dissociation constant (KDRel)

of the phage-QD system. This procedure is modified from the methods described by

Dyson et al [118].

The overall strategy is to mix varying amounts of MAA-QDs and MAA-QD-specific

phage and let these mixtures come to equilibrium. This is done in a BSA-coated plate

to minimize the binding of free phage to the polystyrene plate. An ELISA is then used

to determine the amount of free phage in these equilibrium mixtures. If we know the

amount of free phage in each mixture and we know the total amount of phage and QDs

that we started with, we can make a Scatchard plot to estimate the relative dissociation

constant of the system.

First, a 96-well plate was incubated with a BSA blocking solution (0.5% [w/v] BSA in


0.1M bicarbonate buffer, pH 8.6) at 4 ◦C for at least 1 hour. This coated plate was then

washed 6 times with tris buffered saline, 0.1% TWEEN-20 (TBST). In the coated plate,

QDs of various concentrations (1nM 50nM) were incubated with a constant amount

of phage (2.5 x 1010 pfu/ml) until equilibrium was reached (1 hour at 37 ◦C). These

solutions were transferred to plates coated with MAA-coated QDs. Free phage from these

equilibrium solutions would bind to the QD-coated plate and be detected by ELISA. To

correlate the ELISA signal with an amount of free phage in solution, a calibration curve

was made on the same plate.

Varying concentrations of phage were incubated on an MAA-QD coated plate for

an hour. The plate washed 6 times with TBST. The plate was then incubated for 1

hour at room temperature with a HRP-anti-M13 MAb conjugate, diluted 1:5000 in BSA

blocking solution. The plate was again washed and then developed with the chromogenic

substrate, 3,3,5,5-tetramethylbenzidine. The reaction was stopped with 2 M sulfuric acid.

A plate reader was used to measure the absorbance at 450nm minus the absorbance at

570nm.

3.2.6 Phage assay for QD quantification

The ELISA set-up described in section 3.2.4 was also used to measure concentrations

of QDs using the phage obtained from phage display panning. Here, however, the con-

centration of QDs used to coat the well was varied while the amount of phage used in

detection was kept constant. QDs were diluted in bicarbonate buffer and the phage

concentration was fixed at 1011pfu/ml.


3.3 Results

3.3.1 Nanoparticle characterization

QDs were characterized through UV-Vis spectrophotometry, fluorescence spectroscopy

and gel electrophoresis. Figure 3.3 A, B, C show the absorbance and emission spectra

of MAA, MUA and BSA-coated QDs respectively. These surface chemistries are shown

schematically in Figure 3.4. Since these QDs have identical core ZnS capped-CdSe quan-

tum dots made from a single batch, they cannot be differentiated by their spectra as

shown in Figure 3.3 A, B and C.

Gel electrophoresis was done with a 2% agarose gel in 0.5x Tris buffered EDTA (TBE).

The gel was subjected to a 100V field in TBE buffer for 30 minutes. The gel was imaged

under UV excitation. Figure 3.3D shows the inverted image

From the image, we can see that MAA-coated QDs traveled furthest, followed by

MUA-coated QDs and then BSA-coated QDs. This indicates that MAA-coated QDs

have the smallest size to charge ratio, followed by MUA-coated QDs and then BSA-

coated QDs. That the QDs traveled the same direction in the gel indicates that their

surface charges are of the same polarity (in this case, negative).

3.3.2 Optimization of affinity matrix

Affinity-based selection of binding clones from a phage display library necessitates an

affinity matrix. Generally, this is a solid substrate onto which the target antigen is

affixed. The phage library is incubated with the affinity matrix, allowing clones with

desired binding properties to bind. Non-binding clones are then washed away.

Figure 3.5 shows the fluorescent signals given by QDs that were incubated in wells of

a polystyrene plate coated with gelatin, poly-lysine, BSA and nothing. Since the gelatin-

coated plate resulted in the highest fluorescence signals from bound QDs, gelatin was used

to coat the plate for all subsequent phage panning and characterization experiments.


Figure 3.3: Spectra and gel electrophoresis of QDs of different surface chemistries. (A)

Absorbance (dashed lines) and fluorescence (solid lines) spectra of (A) MAA-coated QDs,

(B) MUA-coated QDs, (C) BSA-coated QDs are very similar and cannot be used to

distinguish the different surface chemistries. (D) Gel electrophoresis of MAA, MUA and

BSA-coated QDs indicate that these QDs have different charge to size ratios. (MAA =

mercaptoacetic acid; MUA = mercaptoundecanoic acid; BSA = bovine serum albumin)


Figure 3.4: Schematic of the quantum dot surface chemistries tested in this study: MAA

= mercaptoacetic acid; MUA = lysine-crosslinked mercaptoundecanoic acid; BSA =

bovine serum albumin.


Figure 3.5: Optimization of the affinity matrix. Various coatings were tested for their

ability enhance the adsorption of quantum dots of different surface chemistries onto the

plastic substrate (NUNC Maxisorp plates). The fluorescence of the QDs adsorbed onto

the plate after overnight incubation was measured using a fluorescence plate reader.

Gelatin-coated plates resulted in the highest fluorescent signals from the quantum dots.

(Mean ± SD, n = 3)


3.3.3 Panning

The effectiveness of the phage selection or panning process may be monitored by mea-

suring the number of phage eluted after each round of panning. If the selection process is

successful, this number can be expected to increase significantly after each round, as the

phage that bind to the target become an increasingly larger portion of the total phage

population.

Table 3.1 shows the number of eluates from each round. From round 1 to round 2

there is an increase by about four or five orders of magnitude. From round 2 to round

3, there is actually a slight decrease, although there were more phage eluted in round 3

than in round 1. This decrease in the number of binding phage is likely due to the more

stringent washing conditions applied in round 3 than in round 2.

Table 3.1: Number of phage eluted after each round of panning

Target Round 1 Round 2 Round 3

MAA QDs 300, 000 2, 500, 000, 000 18, 000, 000

MUA-Lysine BSA QDs 40, 000 1, 030, 000, 000 95, 000, 000

polymer QDs 106, 000 2, 200, 000, 000 500, 000, 000

Table 3.2 shows the washing conditions used in each round of selection. From round

1 to round 2, the amount of detergent in the wash buffer was increased from 0.1% [v/v]

to 0.5% [v/v]. From round 2 to round 3, the amount of time for each wash was increased

from 10-15 seconds to 1 minute. Washing conditions were changed from round to round

in order to increase the stringency of selection. The significant increase in the time for

each wash is likely the cause for the number of eluted phage in round 3 being lower than

that in round 2. Each round of selection involved 10 washes.


Table 3.2: Washing conditions in each round of panning

Conditions Round 1 Round 2 Round 3

Washing buffer TBS, TBS, TBS,

0.1% TWEEN-20 0.5% TWEEN-20 0.5% TWEEN-20

Time for each wash 10-15 seconds 10-15 seconds 1 minute

# washes 10 10 10

TBS = Tris buffered saline (50mM Tris, 150mM NaCl, pH 7.5)

3.3.4 Sequences of binding clones

Of the 120 clones tested, those in table 3.3 showed promising binding characteristics in

that they bound specifically to their intended target QDs and showed little binding to

QDs of other surface chemistries. The majority of the clones bound either to MAA-

coated QDs, regardless of the surface chemistry of the QDs they were selected on, or

they bound promiscuously to more than one type of QD surface chemistry.

Table 3.3: Amino acid sequences of QD-binding peptides

Code Sequence Surface chemistry of QD target

MAA7 SAMHSKHRQAVP MAA

MAA9 H (Termination -TAG) MAA

BSA5-7 SHKHNTHPRFPL BSA

BSA5-8 VEGSHYFWHREN BSA

BSA21 SHNHQYLYSPEV BSA

Table 3.3 shows the sequences of the peptides displayed on the phage that bound

to various QD targets. The high presence of histidine is to be expected, as histidine is

known for its metal coordinating abilities, and thus may be binding to the ZnS surface


of the capped QDs. These sequences are also populated by many aromatic residues (H,

P, W, Y) and amine-containing residues (K, Q, R).

In phage display panning on carbon nanotubes, Wang et al. found that aromatic

residues such as histidine and tryptophan were highly represented in the binding peptides

[119]. In a study testing the binding of amino acids to inorganic surfaces Willet et al.

concluded that the charge of an amino acid was the primary determinant in its binding

to an inorganic surface [120]. Since all the QDs tested were negatively charged (verified

by the direction they travelled in gel electrophoresis), it is not surprising that the binding

peptide have many positively charged amine-containing residues.

3.3.5 Comparison of binding characteristics

Figure 3.6 shows that MAA7 phage bind more strongly to MAA-coated QDs than to

MUA or BSA-coated QDs. At an input phage concentration of 1011pfu/ml, MAA-coated

QDs produce a signal four times higher than BSA-coated QDs and 10 times higher than

polymer-coated QDs.

3.3.6 Determination of dissociation constant

The relationship between the optical density obtained from the ELISA (OD 450-570nm),

and the amount of phage in solution is shown in Figure 3.7. This plot was used in the

second part of the experiment to determine an unknown free phage concentration based

on an optical density measurement made through ELISA.

Figure 3.8 shows the optical density from ELISA obtained from the competitive assay

QD-phage solutions. The optical density is related to the amount of free phage in solution,

which is inversely related to the QD concentration in the equilibrium phage-QD solutions.

At higher QD concentrations, more phage will be bound to the QDs in solution and less

phage will be free to bind to the plate.

Using the standard curve on Figure 3.7, we determine the concentrations of free phage,


Figure 3.6: Binding of MAA7 phage (expressing the peptide: SAMHSKHRQAVP) on

QDs of different surface chemistries. The optical density (OD 450-570nm) measures the

amount of phage binding to the quantum dots at each concentration of input phage. The

curves indicate that MAA7 phage bind more strongly to MAA-coated QDs than to MUA

or BSA-coated QDs. (mean ± SD, n = 3)


Figure 3.7: Correlation of absorbance from ELISA and concentration of free phage in

solution. This curve is later used to determine the amount of free phage in various phage-

QD mixtures. Error bars indicate the standard deviation (n = 3) (pfu = particle forming

units).

p, in each solution. The concentration of bound phage, x, is calculated from the mass

conservation equation below where pt is the total phage concentration.

x = pt − p (1)

Assuming that each phage binds to one QD, the number of free QDs, q, is again

determined from the mass conservation equation, where qo is the total QD concentration.

q = qo − x (2)

The dissociation constant is the ratio of free quantum dots and phage to bound

quantum dot-phage complexes at equilibrium.

KD =pq

x(4)

substituting for p, based on equation 1

KD =(pt − x)q

x(5)

which can be rearranged as follows

KD =ptq

x− q (6)


Figure 3.8: Competitive assay using MAA-coated QD-specific phage. A constant concen-

tration of phage (5 x 109 pfu/ml) is incubated in solution with QD solutions of various

concentrations (depicted on x-axis). The free phage from these solutions are captured

onto QDs adsorbed to a substrate and measured in an immunoassay (reflected in the ab-

sorbance measurement on the y-axis). As expected, as the QD concentration increases,

more phage are bound to QDs, and less phage are free to bind to the QD-coated plate,

leading to a lower OD signal. Error bars indicate the standard deviation (n = 3)


Figure 3.9: Scatchard plot. The slope of the Scatchard plot is the negative inverse of the

relative dissociation constant. From this plot, KDRel = 5 x 10-9. (x = concentration of

bound phage, pt = total phage concentration, q = concentration of free QDs)

x

ptq=

1

KD

− x

KDpt

(7)

where the slope of the plot ofx

ptqagainst

x

pt

is equal to−1

Kd

.

Equation 7 follows the form of the Scatchard equation

rc

= 1KD

− rKD

(8)

where r is the ratio of the concentration of bound ligand to total available binding

sites ( xpt

in equation 7) and c is the concentration of free ligand (q in equation 7).

Figure 3.9 shows the plot of x/ptq against x/pt. Many points from figure 3.8 were

not transformed and plotted in 3.9 because they were too low (in the flat part of the

competitive binding curve) or at the y-intercept, where [QD] = 0. The apparent or

relative dissociation constant derived from this graph is 5nM (-1/KDRel = - 2 x 108,

therefore KDRel = 5 x 10-9).


Figure 3.10: Measurement of the concentration of quantum dots of different surface

chemistries using the phage assay. The phage assay specifically measures mercaptoacetic

acid-coated quantum dots down to a concentration of 10nM. MAA = mercaptoacetic

acid; MUA = lysine-crosslinked mercaptoundecanoic acid; BSA = bovine serum albumin.

(mean ± SD, n = 3)

3.3.7 Measuring concentration using phage assay

The peptide-displaying phage was then used to measure concentrations of QDs of different

surface chemistries. Figure 3.10a depicts the detection scheme, which resembles the

ELISA described previously. Figure 3.10 shows that the selected phage can be used to

detect concentrations of MAA-coated QDs from 10nM to 1µM, while giving no signal from

BSA-coated or polymer-coated QDs, thus demonstrating specificity to surface chemistry.

3.3.8 Robustness against optically interfering agents

In order to determine whether this method can quantify QDs in the presence of optically

interfering species, we purposely spiked a 500nM solution of MAA-coated QDs with

trypan blue (maximum absorbance = 588nm). Figure 3.11 shows a comparison of the


Figure 3.11: Quantification of QDs by phage assay in the presence of optically interfering

agents. (A) Phage assay can be used to detect QDs in water and trypan blue. (B) A

linear relationship exists between the absorbance signal and QD concentration in distilled

water but not in trypan blue, because the high absorbance of the trypan blue blocks the

absorbance signal of the QDs (3B). (mean ±SD, n=3)


signal obtained from phage-based detection of MAA-coated QDs in water and trypan

blue. The phage assay gives a linear QD concentration curve in both distilled water and

trypan blue. In absorbance spectroscopy, the trypan blue obscures the absorbance of the

QDs.

3.4 Conclusion

Phage display was used to identify peptide-expressing bacteriophage which bind specif-

ically to MAA-coated QDs. An assay for the quantification of QDs in solution was

demonstrated with these bacteriophage and demonstrated sensitivity down to 10nM as

well as minimal binding to QDs with other surface chemistries (MUA and BSA). This

phage assay was also used to quantify nanoparticles in optically interfering media, where

measurements by absorbance spectroscopy are not possible. This phage-based assay has

potential application in detecting, quantifying and discriminating between nanoparti-

cles based on their surface chemistry, the importance of which is becoming increasingly

apparent in biological applications.

Jagota and colleagues, who have identified peptides that bind to carbon nanotubes,

have suggested that these peptides can be used as a means of providing “chemical han-

dles” by which to manipulate the carbon nanotubes [119]. A similar goal is pursued

here. Like monoclonal antibodies, which have been used in innumerable applications

in biochemistry to detect, quantitate, purify, select and manipulate protein targets,

peptide-displaying phage that bind specifically to nanoparticles may do the same for

their nanoparticle targets. The idea of using phage to play the role of antibodies has

been suggested and demonstrated by the founder of phage display screening, George

Smith. He and Petrenko displayed peptides on the major coat protein on the filamentous

phage and demonstrated their use as substitute antibodies—with high specificity and

nanomolar affinities [121].


Here, instead of proteins, the target is nanoparticles but the principles remain the

same. The antibody, or in this case, the phage particle, is used as a link between the target

an the outside world. When bound to a chromophore, or an anti-phage antibody that is

bound to a chromophore, the phage can be used to detect or quantify the nanoparticle.

Chapter 4

Phage Display on Whole Cells

The development of novel nanoparticle probes opens the doors to multiplexed labeling

that could lead to a greater understanding of biological processes and more accurate

diagnosis of pathological states. For this to happen, however, novel nanoparticle probes

must be coupled to targeting molecules which bind specifically to molecular targets. This

chapter describes the identification of targeting peptides through phage display screening

on whole cells.

Some of the findings from this study were published in Mardyani, S., Singhal, A.,

Jiang, W., Chan, W.C.W., Interfacing peptides, identified using phage-display screen-

ing, with quantum dots for the design of nanoprobes, Progress in Biomedical Optics and

Imaging - Proceedings of SPIE 5705, pp. 217-224, 2005. Reproduced with permission.

4.1 Introduction

Isolation of peptides that target the molecular markers of cancer has been pursued

through phage display. This method uses a library of bacteriophages displaying differ-

ent peptides on their protein coat in a multi-round selection process to find an effective

targeting peptide [54].

In 1985, Smith demonstrated phage display as a method to find targeting peptides

61

Chapter 4. Phage Display on Whole Cells 62

for a single protein target [53]. Since then, the technique has since been used on a greater

variety of targets including inorganic substrates [122], nanostructures [112, 123, 119, 124],

whole cells [125] and tumour vasculature [126, 59]. Peptides identified through phage

display have been used for in vivo targeting of quantum dots [72].

Phage display panning on complex targets such as whole cells presents new challenges.

Primarily, because a cell has hundreds or thousands of proteins on its surface, phage

display panning on such a target will likely yield hundreds of binding peptides without

a clear consensus as to which amino acid sequence binds with the greatest affinity [127].

Secondly, cells of different types will express many of the same proteins. If a phage

display panning experiment yields a peptide that binds to a protein that is common to

many cell types (e.g. cancerous and non-cancerous cells), this peptide will not be useful

for detection or targeting since it will not be selective.

The solution to this problem is subtraction panning [125]. Before selecting phage from

the library that bind to the target cell, the library is first depleted of phage that would

bind to a non-target cell, as shown in a Venn diagram in figure 4.1, which demonstrates

this idea for selecting phage that bind to malignant as opposed to healthy cells. In this

way, phage that bind promiscuously to both target and non-target cells are removed from

the library.

One of the major strengths of phage display is that it allows for the identification of

a targeting molecule without having prior knowledge of the target [127]. Because of its

simplicity, potency and wide application, phage display was the strategy chosen for the

identification of targeting molecules in this project.

As an initial demonstration of whole cell phage display was performed to find peptides

that target HeLa (human cervical cancer) cells. STO (mouse fibroblast) cells were used

for the subtraction rounds. These two cell lines were chosen for ease of maintenance.

The subtraction/selection procedure was adapted from Rasmussen et al [125].

After four rounds of panning, the phage clones will be amplified and their DNA


Figure 4.1: Venn diagram representation of the evolution of phage display library after

subtraction and selection. Subtraction removes phage that bind to healthy cells. The

remaining phage are then panned on malignant cells to select for phage that bind to

cancer cells. The resulting phage will bind to cancer cells and not to healthy cells.


extracted and sequenced. This DNA sequencing will reveal the amino acid sequence

of the peptides that bind to the target cells. These peptides will then be synthesized,

conjugated to quantum dots and used for cell staining.


4.2.1 Phage Panning

The Ph.D.-7 phage display library was purchased from New England Biolabs. This

library contains 2.8 x 109 clones of M13 phage with random heptapeptide inserts.

HeLa and STO cells were cultured using Dulbeccos Modified Eagle Brand media sup-

plemented with 10% fetal bovine serum, 5% amphotericin, and 5% penicillin/streptomycin.

The cells were then trypsinized using Trypsin EDTA and washed three times using PBS.

The subtraction (STO) cells were counted and resuspended at 107 cells/ml in PBS, 1%

bovine serum albumin. These cells were incubated at room temperature with 1.5 x 1011

peptide displaying phages with slow shaking for 1 hour. The cells were then centrifuged

at 1500 rpm for 2 minutes. The supernatant, containing phage that did not bind to these

subtraction cells, was used to resuspend another 107 cells. This subtraction process was

repeated three times. After subtraction, the remaining phage were incubated with 5 x

106 selection (HeLa) cells for 4 hours at 4 ◦C with slow shaking. After incubation, the

cells were washed by centrifuging at 1500 rpm for 2 minutes, discarding the supernatant,

and resuspending the remaining cells in PBS, 1% BSA, 1% Tween-20. The wash was

repeated five times, changing centrifuge tube between each wash. The binding phage

were eluted by incubating with 100µl Glycine-HCl (pH 2.2) for 10 minutes on ice. The

cells were centrifuged at 1500 rpm for 2 minutes and the pellet was discarded. The super-

natant was neutralized by adding 15µl Tris-HCl (pH 9). 1µl of this phage solution was

titered and the remaining phage were amplified in 20ml of LB-Tetracycline in a shaking

incubator at 37 ◦C for 4.5 to 6 hours. The amplified phage were then harvested and


titered. The subtraction/selection panning procedure was repeated using these newly

amplified phage. This panning procedure was repeated four times.

4.2.2 Sequencing

After four rounds of panning, DNA from the binding phage was extracted by ethanol

precipitation. This DNA was sequenced by The Centre for Applied Genomics in the

Hospital for Sick Children in Toronto, Ontario. The DNA sequence was then interpreted

to find the amino acid sequence of the displayed heptapeptide.

4.2.3 Analysis of binding peptide

Before the peptide was synthesized, the binding efficiency of the phage was determined.

1.5 x 1011 phage were incubated with 5 x 106 HeLa or STO cells for 4 hours at 4 ◦C with

slow shaking. Unbound phage were then washed off using the procedure described above.

The bound phage were eluted with 100µl Glycine-HCl (pH 2.2) for 10 minutes on ice.

The cells were centrifuged at 1500 rpm for 2 minutes and the pellet was discarded. The

supernatant was neutralized by adding 15µl Tris-HCl (pH 9). These phage were then

titered to determine the number of phage that bound to each cell type.

Once the binding efficiency was determined, the peptide was synthesized at the syn-

thesis facility at the Hospital for Sick Children in Toronto, Ontario.

4.2.4 Peptide-quantum dot conjugation

The synthesized peptide was conjugated to quantum dots using a carbodiimide-mediated

reaction. Peptides and quantum dots were mixed in a 1:25 molar ratio. The carbodiimide,

1-Ethyl-3-(3-dimethylaminopropyl)-carbodiimide, was then added at a 1000 times molar

excess to the peptide, to mediate the formation of a peptide bond from a carboxylic acid

group on the quantum dot and an amine group on the peptide [128].


4.2.5 Cell staining experiment

The peptide-conjugated quantum dots were then incubated with HeLa and STO cells to

demonstrate labeling capabilities. HeLa and STO cells were trypsinized, counted and

plated in 60mm tissue culture dishes with cover slips inside at 106 cells per dish. The

cells were incubated at 37 ◦C, 5% CO2 overnight. After feeding the cells, 50µl of quantum

dot-peptide conjugate was added to one dish of HeLa cells and one dish of STO cells.

Controls with equal amounts of bare, unconjugated quantum dots in HeLa and STO cells

were also studied.

4.3 Results

To find peptides that target HeLa cells, subtraction/selection phage display was per-

formed using STO cells for subtraction and HeLa cells for selection. The panning began

with 1011 plaque forming units of phage expressing approximately 2.8 x 109 different

peptide inserts. After four rounds of panning, the remaining phage pool was found to

be dominated by phage displaying a single peptide sequence. Sequencing phage from

previous rounds revealed that by round three, there was only one peptide sequence in

the phage pool. Thus, in this case, the fourth round of panning was unnecessary.

Figure 4.2 shows the binding efficiency of the phage pools from rounds 1, 2, and 4.

As expected, binding efficiency increased with each additional round of panning. After

round 4, the binding efficiency on HeLa cells was about double that on STO cells.

The peptide displayed on the phage from round 4 was synthesized and used for

quantum-dot based labeling. Figure 4.3 shows the structure of the synthesized pep-

tide. In addition to the seven amino acids of the heptapeptide displayed on the phage,

the peptide was synthesized with a GGGS spacer. On the M13 phage in this library,

this spacer exists between the displayed peptide and the protein coat of the phage. This

4-amino acid chain increases the display of the peptide when it is conjugated to the


Figure 4.2: Progress of panning experiment on whole cells. Each round of panning

begins with 1.5 x 1011 pfu of the phage library. In each round of panning, the phage

undergo three one-hour subtractions on non-target (STO) cells and one, four-hour round

of selection on HeLa cells. These cells are washed five times and the bound phage are

eluted and titered. With each subsequent round of panning, the number of phage that

bind to the target cell increases.


Figure 4.3: The molecular structure of the HeLa-binding peptide identified through phage

display panning. The HFYVSPW hepta-peptide is flanked by a gly-gly-gly-ser spacer on

the C-terminus.


Figure 4.4: Cell staining with peptide-quantum dot conjugates. A) HeLa cells show

significant staining with quantum dot-peptide conjugates. B) Brightfield image of HeLa

cells shows that round cells are stained more than the flat cell at the left of the picture. C)

STO cells show very little staining with quantum dot-peptide conjugates. D) Brightfield

image of STO cells.

quantum dot. The C-terminus of the peptide is amidated to allow the conjugation of

this end of the peptide to the quantum dot.

Figure 4.4 shows STO and HeLa cells stained with the peptide-quantum dot conju-

gate. The left column shows the fluorescence of the 600nm emission quantum dots under

UV excitation and a 520nm long pass emission filter. The HeLa cells show much more

labeling than the STO cells. Examination of a larger population of cells showed that only

about 15% of the HeLa cells showed significant staining while the STOs showed minimal

staining. The HeLa cells that were stained were mainly those that were rounded rather

than flat in morphology. The increased affinity to cells with rounded morphology may

be a result of the phage display panning being performed on trypsinized cells.

Phage display panning on trypsinized cells also limits the potential markers that the


phage might be binding to. As trypsin is a serine protease, it will digest surface-bound

proteins. The phage may have been binding to glycans such as glycosphingolipids ex-

pressed on the cell surface. Since glycans have been implicated in tumour proliferation,

invasion and angiogenesis [129], targeting molecules for these markers could have appli-

cations in cancer research and diagnosis.

A BLAST search found over 100 matches of the peptide sequence to sequences in

the Protein Data Bank. Since the peptide is so short, only seven amino acids long, it is

not surprising that there the number of matches is so large. Further research would be

necessary to narrow in on the possible cell markers that the selected peptide binds to.

The surprising result of this experiment was that only one peptide was identified

through the phage display panning. HeLa cells (human cervical cancer) and STO cells

(mouse fibroblasts) are two very different cell lines. One would expect that using these

two cell types for selection and subtraction of a phage library would yield several targeting

peptides that select for HeLa cells over STOs. However, in this study, after three rounds

of panning, the phage population was reduced to a single clone.

There are many factors that play into this. The directed evolution of phage display

panning is not solely dependent on the selection strategy. The peptides displayed on the

coat of the M13 bacteriophage used in this library are expressed on the PIII protein.

This protein plays a role in the life cycle of the bacteriophage, specifically in infecting

its host for replication. Displayed peptides can decrease the infectivity of the phage [53]

and some peptides might have a greater effect than others on this decreased infectivity.

Phage that display such peptides will have an evolutionary disadvantage to other phage

although they may have higher affinity to the target.

The stringency of the washes, especially in the early rounds of phage display panning

is also an important consideration. In these early rounds, the diverse phage display

library will have relatively few copies of each random peptide sequence. An excessively

stringent wash in these early stages could prematurely remove a potentially good binder.


On the other hand, washes that are not stringent enough will leave poorly binding phage

in the population. In either case, when each round of panning begins with a population of

1011 plaque forming units, and ends with thousands of phage being eluted after selection,

it is highly probable that some potential binders are lost and some poor binders are kept.

This loss of binding phage may be prevented by examining the phage pool for binding

peptides at an earlier stage in the panning, where the phage pool still contains a diverse

population of clones. This would drastically increase the complexity of this phage display

selection process. Clones would need to be isolated, amplified and individually tested.

Many of these clones would likely have poor binding characteristics—either not binding

at all to the target or binding indiscriminately and non-specifically.

Performing subtraction and selection panning on trypsinized cells may have also

played a role in limiting the number of binding phage found in this procedure. Trypsiniz-

ing cells removes many of the cell surface markers that may have differentiated the two

cell lines. To preserve surface bound proteins, adherent cells can be released using EDTA

instead of trypsin. If this was done in a whole cell phage display experiment where STO

cells are used for subtraction and HeLa cells for selection, it is likely that the panning

process would yield a large number of clones – not all of which would have had a useful

application. A more meaningful result could be obtained if the subtraction and selection

cells are more similar to each other and relevant to a specific application. For example,

subtraction on healthy cervical cancer cells and selection on malignant cervical cancer

cells could isolate markers that differentiate the two.

4.4 Conclusion

Phage display screening using STO cells for subtraction and HeLa cells for selection has

yielded a peptide that, when conjugated to quantum dots, can specifically label HeLa

cells. In fact, the peptide labels a subpopulation of HeLa cells with a rounded morphology.


This was accomplished with no prior knowledge of the nature of the target.

We have demonstrated identification of a cell-targeting peptide by phage display

screening and its use in quantum-dot cell staining. This demonstration highlighted some

weaknesses in whole cell phage display panning, which may be overcome with different

screening strategies. The potential for combining quantum dots with phage display for

improved diagnostics remains. These results may be used as guidance for the development

of future strategies that aim to capitalize on these powerful technologies.

Chapter 5

High Throughput Phage Display

This chapter describes a phage screening strategy for whole cells, which was designed

to enable simultaneous identification of multiple biorecognition molecules from a single

library. QD-encoded microbead barcodes were conjugated to protein targets and then

used to screen a phage display library. The beads and the binding phage were then

separated using flow cytometry and fluorescence assisted cell sorting.

5.1 Introduction

Conventionally, the process of identifying biorecognition molecules involves screening a

library of potential candidate targeting molecules on one target at a time. The library

is incubated with the target, allowing some members to bind. Non-binders are washed

off. Then, the remaining binding population is amplified and the screening is repeated

with the binding population. This is done for several rounds. The resulting targeting

molecules are then characterized and tested for their binding to the target molecule.

If the binding has high affinity and specificity, the targeting molecule has been found

and the experiment is complete, otherwise, the process is repeated until the appropriate

molecule is found. Figure 5.1 shows a general flowchart depicting the process of screening

for a targeting molecule, such as a phage displayed peptide. We propose a method for

73

Chapter 5. High Throughput Phage Display 74

Figure 5.1: Process of screening a molecular library to identify a targeting molecules.

The library is incubated with the target, non-binding members are washed away and

the binding population is amplified. The screening is repeated and then the resulting

potential target molecules are characterized and tested for their binding specificity and

affinity. If sufficient specificity and affinity is demonstrated, the experiment is complete

and the targeting molecule has been found. Otherwise, the process is repeated.

high throughput screening of multiple targets in one screening experiment.

In this system, the different targets or antigens are labeled with unique labels and

later separated according to those labels. This is shown in Figure 5.2 as a labeling step

before screening, and a separation step after screening.

The use of quantum dot barcodes [42] would allow for the labeling of hundreds of

potential targets. Figure 5.3 is an illustration of the barcoding concept, showing four of

the eight possible barcodes that can be created with three different colors of beads at two

intensity levels. The intensity levels correspond to the number of quantum dots of each

color in the bead. The barcodes can be read using a fluorimeter, which will measure the

intensity and wavelength of emission. Quantum dot barcodes can also be rapidly read


Figure 5.2: Proposed method of high throughput screening. In high throughput screen-

ing, the different targets are first labeled with unique markers, rounds of screening and

amplification are performed, and then the different targets are separated based on their

unique signals. This would be followed by iterations of verification and testing (not shown

in the diagram).


Figure 5.3: Barcoding with quantum dots. By combining different colours and different

intensity levels in mesoporous beads, a large number of barcodes can be made from a few

distinct colours of quantum dots.


Figure 5.4: High throughput screening of biorecognition molecules. In high throughput

phage screening, target proteins are conjugated to QD barcoded microspheres and incu-

bated with a molecular library (e.g. phage display library, yeast surface display, aptamer,

etc.) Unbound members of the library are washed away and the QD barcodes, with their

attached targets and bound library members are separated by fluorescence activated cell

sorting.

using a flow cytometer [130], which will be used for this particular application.

Figure 5.4 shows three different quantum dot bead barcodes conjugated to molecular

targets used for screening a library of molecules. These barcodes are incubated with

the library, allowing complementary molecules from the library to bind to the molecu-

lar targets. Unbound members of the library are washed away and quantum dot bead

barcodes (along with the molecules bound to them) are separated through fluorescence

activated cell sorting. Biorecognition molecules for each target can then be eluted from

the quantum dot bead barcode.



An attempt was made at implementing this high throughput phage screening strategy

with limited success. The following sections describe the methods used and the results

obtained so they may be built upon for future work.

5.2.1 Polystyrene bead barcordes

10µm, carboxylic acid functionalized polystyrene beads were encoded with green and red

quantum dots by soaking in a solution of quantum dots dissolved in propanol. Quantum

dots diffuse and remain in the beads due to hydrophobic-hydrophobic interactions. QD-

encoded beads were embedded in resin, sliced by a microtome and imaged by TEM.

5.2.2 Conjugation of barcodes to proteins

These beads can then be covalently conjugated, through a carbodiimide-mediated peptide

bond, to their respective target molecules. As a proof of concept, beads encoded with

three different quantum dot codes (green, green-red, red) were respectively conjugated

to three different proteins (protein A, transferrin, streptavidin).

5.2.3 Phage display panning

After conjugation the three separate populations were mixed together and three rounds of

phage display screening were performed on the mixed population. After the third round,

the beads were separated by fluorescence activated cell sorting (FACS). The phage were

then eluted from the sorted beads and then amplified and sequenced. Three different

sequences were found and the peptides encoded by these sequences were synthesized.


5.2.4 Analysis of bound peptides

The binding of the different peptides to each target protein were then tested. The pep-

tides were conjugated to 10µm carboxylic acid functionalized polystyrene beads. Beads

were then incubated for one hour with varying concentrations (10nM - 10µM) of fluo-

rescently labeled protein, dissolved in 10mM tris, pH 8, 3% milk. The beads were then

resuspended in 10mM PBS, 1% BSA and examined by flow cytometry.


Figure 5.5 shows the mixed, and separated bead populations, infused with 536 nm, 610

nm and a mixture of 536 nm and 610 nm emitting quantum dots. FACS was used for

separation.

After three rounds of panning, 15 phage plaques were sequenced and the following

three unique peptides were identified: His-Leu-Tyr-Val-Ser-Pro-Trp (anti-Streptavidin)

His-Arg-Val-Pro-Thr-His-Pro (anti-Protein A) His-Lys-Arg-Pro-Arg-Asn-Asn (anti-Transferrin)

Figure 5.6 shows the binding of the three peptides to protein A. The fluorescence

measures the amount of fluorescently labeled protein A bound to the peptide-coated bead

after incubation and washing. As expected, anti-protein A has much higher binding to

protein A than the other peptides.

Additional experiments were unable to show similar specific binding for the anti-

streptavidin and anti-transferrin peptides. It was expected that the peptide expressed

on phage that was binding to streptavidin-coated beads (i.e. anti-streptavidin peptide)

would bind specifically to streptavidin but not to transferrin or protein A. This was

not confirmed by the binding experiments. Similarly, the peptide that was binding to

transferrin-coated beads (i.e. anti-transferrin peptide) did not show specific binding to

transferrin in binding experiments.

It is possible that phage displaying these peptides survived the selection process


Figure 5.5: Separation of optically coded polystyrene beads into respective populations.

A mixture of green (536nm), red (610nm), and green-red mix (536nm+610nm) quantum

dot-encoded beads (left) is separated by FACS into three unique populations (green-

536nm, green-red-mix-536nm+610nm, and red-610nm).


Figure 5.6: Binding of protein A to peptides. Fluorescence is directly proportional

to the number of protein A molecules bound to the peptide-coated polystyrene bead.

Polystyrene beads were coated with three different peptides: anti-transferrin, anti-

streptavidin and anti-protein A. Here, protein A shows more binding to anti-protein

A peptide (i.e. the peptide obtained through phage display screening on protein A) than

anti-transferrin or anti-streptavidin peptides.


through non-specific binding. In the phage display panning process, it is very possi-

ble that some phage that do not bind specifically to a target are not removed through

the washing steps. These phage could potentially come to be overrepresented in the

phage population due to higher infectivity and heightened rates of multiplication during

the phage amplification stages. Examining the phage pool at an earlier round of panning,

before so many phage amplification stages may yield better results.

5.4 Conclusion

A high-throughput phage display panning procedure, which used quantum dot-encoded

polystyrene beads and FACS was developed. Three targets were tested: streptavidin,

transferrin and protein A. A peptide was identified which bound specifically to protein A.

Peptides that were isolated through this process because of their binding to streptavidin

and transferrin, however, were later found to not bind specifically to their respective

targets.

This project faced obstacles in the development of quantum dot barcodes. Initially,

it was found that many of the quantum dots used for barcoding lost much of their

fluorescence once put in buffer. This lead to studies on the effect of biological buffers on

quantum dot fluorescence, which are described in Chapter 2. Eventually, some quantum

dots were found with high enough fluorescence and quantum yield that even when they

lost some fluorescence in the buffer, they retained enough for detection with FACS.

Since this study was done, improved quantum dot barcodes have been developed that

insulate the quantum dots from their external environment [106]. These barcodes are

much more robust against external variations in pH or ion concentration. The use of these

barcodes coupled with adjustments in the stringency of washes and numbers of rounds

of panning could make this strategy a successful means of performing phage display on

multiple targets simultaneously.

Chapter 6

Rational Design of Nanoscale Drug

Delivery Vehicles

The design of targeted, environmentally responsive drug delivery vehicles is a very attrac-

tive biomedical application of nanoparticles. Targeting can be achieved by conjugating

targeting molecules to the particle surface. The chemistry of the particle may also be

tailored so that the particle releases its payload in response to a biological trigger. This

chapter describes two studies done with the Kumacheva group from the Chemistry De-

partment at the University of Toronto, in designing targeted drug delivery vehicles for

the delivery of chemotherapeutics [131, 132]. Hydrogels were designed for intracellu-

lar delivery and pH-triggered drug release. In vitro studies show that conjugation of

biorecognition molecules to the hydrogel drug delivery vehicle increased the effectiveness

of the drugs in causing cell death.

This work has been published in M. Das, S. Mardyani, W.C.W. Chan, and E. Ku-

macheva: Biofunctionalized pH-responsive microgels for cancer cell targeting. Advanced

Materials. 2006. 18(1):80 - 83. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Re-

produced with permission. Portions of this chapter are also reproduced with permission

from H. Zhang, S. Mardyani, W.C.W. Chan, and E. Kumacheva. Design of biocompatible

83

Chapter 6. Rational Design of Nanoscale Drug Delivery Vehicles 84

chitosan microgels for targeted pH-mediated intracellular release of cancer therapeutics.

Biomacromolecules, 7(5):1568 - 1572, 2006. Copyright 2006 American Chemical Society.

6.1 Introduction

Controlled release of macromolecules from a polymer-based system was first described in

1976 with 1.5 x 1.5 x 0.5 mm3 pellets of protein-mixed polymers implanted in the eyes

of rabbits [133]. Over the years improvements through smaller sizes, different structures,

pH responsiveness, antibody targeting and long circulating half-lives have been demon-

strated. Today, several polymer-based nanosized drug delivery vehicles are in clinical

use to treat several different cancers including Kaposi’s sarcoma (Doxil), non-Hodgkin’s

lymphoma (Vincristine) and recurrent breast cancer (DaunoXome). These are mainly

liposome-based systems and all function through passive targeting [4].

Microgels are colloidal particles made of a crosslinked polymer network, which can be

swollen in a suitable solvent [134]. Several properties of microgels make them good can-

didates for drug delivery vehicles. Their particle size can be controlled, they are stable,

and they can be easily functionalized to respond to changes in their environment such

as pH, ionic strength or temperature [131]. Frechet and colleagues have reported pH-

triggered release of drugs from microgels into macrophages [135]. These microgels were

non-targeted and had diameters of 200 nm or greater. Lyon and colleagues described

temperature-triggered release of drugs from 270 nm microgels with folate-mediated tar-

geting. The drug release, however, was at 37 ◦C, offering little protection for healthy

tissues from the drug [136].

To design particles that would selectively deliver drugs into neoplastic cells, we have

identified the following design constraints and objectives:

1. size small enough to potentially take advantage of the EPR effect.

2. protection of the chemotherapeutic drug and maintenance of its efficacy


3. intracellular release of the chemotherapeutic

4. targeted delivery to cancer cells

To meet these objectives, hydrogels were synthesized with diameters under 200 nm

(to meet objective 1). These microgels were chosen as the core component of this drug

delivery system because of their stability, their ease of synthesis, control over particle size

and their open network structure, which allows for the physical loading of drugs within

their core (objective 2). Microgels can be functionalized to swell and shrink in response

to external stimuli such as temperature and pH. In this case, the two microgels tested

were designed to respond to pH so that they could release their payload in the low pH

environment of an endosomal vesicle (objective 3).

The final objective, targeted delivery to cancer cells, will be met in conjunction with

objective 3 by conjugating the surface of the microgel to a targeting moiety, in this case

transferrin. Transferrin is an iron-carrying protein overexpressed on many cancer cells

[137, 138, 139], that undergoes receptor-mediated endocytosis [140]. In this process,

binding of transferrin to a receptor on the cell surface will cause the cell to internalize

the receptor-ligand complex in a vesicle. As it enters the cytosol, the pH in this vesicle

decreases. The hydrogel which carries the cytotoxic payload has been designed to be pH

sensitive and release this payload when it reaches a microenvironment of pH 5. Figure

6.1 shows a schematic of the drug delivery vehicle carrying and releasing methotrexate

intracellularly.

The immortalized cervical cancer cell line, HeLa, was used for in vitro testing. Table

6.1 summarizes the properties of the two drug delivery vehicles designed for these studies.

The main difference between the two DDVs used in these studies is their composition.

Although an effective demonstration of rational design of targeted drug delivery, the first

DDV was limited in its future applications because of the inherent toxicity of the polymer,

poly(N-isopropylacrylamide-acrylic acid) (polyNIPAM-AAc). A similar drug delivery

vehicle was later designed using the biocompatible, biodegradable polymer, chitosan


Figure 6.1: Transport and intracellular release of chemotherapeutic drug via pH sensitive

targeted drug delivery vehicle. The drug, methotrexate is loaded in a pH sensitive hy-

drogel, conjugated to a targeting molecule, transferrin. The transferrin-conjugated drug

delivery vehicle enters the cell through receptor-mediated endocytosis. As the pH in the

vesicle decreases, the hydrogel releases the drug from its matrix.


Table 6.1: Specifications of Rationally Designed Hydrogel Drug Delivery Vehicles

polyNIPAM-AAc Chitosan

Payload Doxorubicin Methotrexate

Targeting Moiety Transferrin Transferrin

Diameter 150nm 180nm

Release trigger pH pH

Release mechanism shrinking swelling

(CS), a naturally existing polysaccharide.


This section describes the materials and methods used to synthesize and characterize the

microgels; load them with dyes or chemotherapeutic drugs and observe the pH-triggered

release; and then test the microgels in vitro.

6.2.1 Microgel synthesis

Poly(N-isopropylacrylamide-acrylic acid) (polyNIPAM-AAc) microgel particles with mo-

lar ratio NIPAM-AAc 9:1 were synthesized by free radical precipitation polymerization

[141]. At pH∼7.0 the carboxylic groups of AAc are strongly ionized (ζ-potential = -38

mV) while at pH∼4 they are largely protonated and carry only a weak charge (ζ-potential

= -1.2 mV). Particle size was determined by photon correlation spectroscopy. [131]

CS-based microgels were synthesized by grafting (2-hydroxyl) propyl-3-trimethyl am-

monium to the CS backbone and then conducting ionic association between the quater-

nary ammonia on the side chain of the resulting polymer of N-[(2-hydroxy-3-trimethylammonium)

propyl] chitosan chloride (HTCC) and sodium tripolyphosphate (TPP) counterions. By


changing the wt% of HTCC and TTP and changing the ratio of HTCC to TPP, the

mean hydrodynamic diameter of the hydrogel could be tuned between 180nm to 800nm.

Particle size and morphology was determined using electron microscopy. [132]

6.2.2 Loading and release of microgel payload

To demonstrate the pH-triggered release of small molecules from the polyNIPAM-AAc

microgels, the dye R6G was loaded into the gel through electrostatic-incorporation. 10

nmol of R6G was mixed with a 1 wt% microgel solution at pH 7.0. After incorporation

in the microgel, excess R6G in the dispersion was removed by centrifugation. Optical

microscopy images were taken after the dye-loaded gels had been in solution at pH 7.4

for at least 7 days.

For the cell studies on polyNIPAM-AAc microgels, 50 µL of the original microgel

sample and 250 µL of 10-4M R6G was mixed together in 5 mL of 0.01 M phosphate-

buffered saline (PBS) at pH 7.4 and allowed to equilibrate overnight. This mixture was

centrifuged at 12 000 rpm and at 4 ◦C for 1 hour. For Doxorubicin (Dox) infusions, 500

µL of 3.68 x 10-5M Dox solution and 50 µL of original microgel sample were incorporated

in 5 mL of PBS (pH 7.4) before centrifugation. The Dox loading level of the microgels

used in this study was 31.5% (mg Dox/mg polymer).

The CS-microgels were loaded with the chemotherapeutic drug, methotrexate dis-

odium (MTX) through electrostatic forces by mixing them with MTX at a microgel/MTX

ratio of 0.3-0.8 in PBS, pH 7.4. The release of the MTX from the microgel was measured

in PBS buffer at 37 ◦C over a 5-day period. MTX concentration was measured using

UV-vis spectroscopy at λ = 372 nm.

6.2.3 Conjugation of targeting moiety

The conjugation of transferrin and albumin to the microgels was accomplished by carbodiimide-

mediated coupling. A 10 mg/ml stock solution of the proteins was made in 0.01 M PBS at


pH 7.4. 240 µL of this solution was then mixed with 300 µL of loaded polyNIPAM-AAc

gels in 0.01 M PBS at pH 7.4. Then, at least a 10-fold molar excess of 1-ethyl-3-(2-

dimethylaminopropyl) carbodiimide hydrochloride was added to mediate the formation

of an amide bond between carboxylic groups on the gel and amino groups on the protein

[128].

For the chitosan microgels, 80 µL of apo-transferrin solution with the concentration of

10 mg/mL was added to each 100 µL of 0.16 wt % HTCC microgel (180 nm in diameter)

dispersion in PBS at pH 7.4. At least 10-fold molar excess of EDC was mixed for coupling.

The reaction was maintained for at least 2 h. Then excess free EDC was removed by

ultracentrifugation and replaced with the same amount of fresh PBS buffer.

6.2.4 In vitro studies

The demonstration of the targeted small molecule delivery using polyNIPAM-AAc mi-

crogels was done using the R6G-loaded microgels in HeLa cells. HeLa cells were grown

on cover slips in 100 mm tissue culture dishes in 37 ◦C , 5% CO2 until 50% confluency

was reached. R6G-loaded hydrogel particles, conjugated to transferrin, were dispersed in

high glucose Dulbeccos Modified Eagle Medium (DMEM) supplemented with 10% fetal

bovine serum, 1% penicillin and 1% amphotericin B. For the controls, non-conjugated

R6G-loaded microgel particles, particles conjugated to bovine serum albumin and parti-

cles in solution with, but not conjugated to transferrin, were used. Cells were incubated

overnight in this system. After washing the coverslips with 10 mM PBS, the cells were

examined under 20x magnification through differential interference contrast and epifluo-

rescence. Quantitative analysis was based on the measurement of luminescence intensity

per cell using ImagePro software. To keep for future reference, the cells were fixed with

34% paraformaldehyde, followed by three washes with PBS.

For viability testing of Dox-loaded polyNIPAM-AAc microgels, 106 HeLa cells were

placed in high glucose DMEM, supplemented with 10% fetal bovine serum, 1% penicillin,


and 1% amphotericin B in each 60 mm tissue-culture dish and allowed to grow for two

days at 37 ◦C , 5% CO2. Dox-loaded hydrogel particles, conjugated to transferrin, as well

as control systems were dispersed in separate dishes of HeLa cells. The HeLa cells were

then incubated for 36 hours with Dox-loaded hydrogel particles conjugated to transferrin

and control systems. The cells were then trypsinized, washed with PBS, stained with

Trypan Blue, and counted under the microscope using a hemacytometer.

The effectiveness of the CS-microgels as drug delivery vehicles was also tested in

vitro on HeLa cells. 106 HeLa cells were plated on 60mm tissue culture plates and grown

overnight in DMEM supplemented with 10% FBS, 1% penicillin and 1% amphotericin

B in 37 ◦C , 5% CO2. Microgels in various forms (with/without MTX, with/without

transferrin or albumin) were then added to the cells and incubated with the cells for 24

hours. The cells were then washed with PBS, pH 7.4, trypsinized and washed again.

Live and dead cells were stained with 4% Trypan Blue solution and counted using a

hemacytometer.


pH-sensitive microgels with diameters less than 200 nm were synthesized and charac-

terized. These gels were loaded with two different small molecules, conjugated with

targeting proteins and tested in vitro on the immortalized cervical cancer HeLa cell line

to demonstrate their utility in drug delivery.

6.3.1 pH-triggered release of microgel payload

Figure 6.2a shows the variation in polyNIPAM-AAc microgel size in the range 4.0<pH<7.0,

caused by the change in ionization of the COOH-groups and the change in polymer hy-

drophilicity. At pH∼7.0 the microgels are 50% larger in size than at pH∼4.0. The

swelling occurs due to osmotically-driven forces and electrostatic repulsion between the


Figure 6.2: Characterization of pH-sensitive nanogels. (a) Normalized radius of

polyNIPAM-AAc microgel with respect to pH (b) epifluorescent microscopy image of

R6G-loaded gels at pH 7 (c) epifluorescent microscopy image of R6G-loaded gels at pH

4.

deprotonated COOH moieties at pH∼7.0 [134, 142].

Figure 6.2b shows the R6G-loaded gels as discrete bright spots at pH 7. At pH 4,

the fluorescence is diffuse, as the dye has been released from the inside of the gel (Figure

6.2c). An estimate, based on fluorescence intensity measurements, shows that microgels,

loaded with 35% R6G (mg R6G/mg polymer) at pH 4 exhibit 88% cumulative release

after 24 hours, versus 12.5% shown by microgels in media of pH 7.4. This pH-dependent

release is the mechanism for delivering the payload of this polyNIPAM-AAc drug delivery

vehicle to the interior of the cell.

Figure 6.3 shows the in vitro cumulative release of MTX from the CS microgels in PBS

at pH 7.4 and pH 5. The increased release at pH 5, compared to pH 7 is attributed to two

factors. First, MTX was partly protonated at pH = 5 (the values of pKa1 and pKa2 of

methotrexate are 4.84 and 5.51, respectively [143]; thus ionic interactions between MTX

and the microgel became weaker. Second, at reduced pH, enhanced repulsion between

the protonated of the amino groups of polymer molecules led to an increased pore size


Figure 6.3: In vitro cumulative release of methotrexate from the microgels in buffers of

different pH. Higher cumulative release of methotrexate from the chitosan microgel is

seen at pH 5 (filled triangle) than pH 7.4 (filled square). (Mean ± SD, n=3)


in the swollen microgel network, favoring diffusion of MTX.

6.3.2 In vitro release of R6G from polyNIPAM-AAc microgels

Figure 6.4 shows the differential interference contrast (DIC) and fluorescent images

of HeLa cells after incubation with R6G-loaded hydrogels. Cells incubated with the

transferrin-conjugated gels show clear fluorescence, while the controls (albumin-conjugated

gels and unconjugated gels) did not, indicating the specific, endocytosis-mediated release

of the R6G into the cells cytosol. This demonstrates that the targeting molecule enables

the drug to enter the cell through a specific pathway and be released in the target cell.

6.3.3 In vitro viability studies of HeLa cells treated with tar-

geted microgel drug delivery vehicles with pH-triggered

release

Figure 6.5 shows the viability of HeLa cells after incubation with dox-loaded hydrogels.

Cells incubated with gels conjugated with transferrin showed the lowest viability, 28.4

± 5% compared to 65-75 ± 5% in the controls (p < 0.01). Transferrin-conjugated,

Dox-loaded microgels caused higher cell mortality than Dox-loaded microgels in solution

with transferrin, albumin-conjugated Dox-loaded microgels, unconjugated Dox-loaded

microgels and transferrin-conjugated plain microgels. This indicates that the transferrin-

conjugated microgel effectively delivered Dox into the cell.

Figure 6.6 shows the results of a study of cell viability and mortality. HeLa cells

were incubated for 24 hours with (a) regular media with nothing added, (b) free MTX

in the same amount present in MTX-loaded CS-based microgels, (c) CS-based microgels,

(d) transferrin-conjugated HTCC CS-based microgels, (e) non-conjugated MTX-loaded

CS-based microgels, and (f) transferrin-conjugated MTX-microgel complexes.

The mortality of the cells in these systems were 0; 9.4; 11.5; 2.8; 10.5; and 45.2


Figure 6.4: Release of R6G in HeLa cells. Differential interference contrast and epifluores-

cent microscopy images of R6G release in HeLa cells from (a) R6G-loaded polyNIPAM-

AAc gels with no attached targeting molecule, which produced no apparent staining,

(b) R6G-loaded gels conjugated to albumin, which shows some fluorescence, likely due

to non-specific binding and (c) R6G-loaded gels conjugated to transferrin, which gener-

ated significant staining in the HeLa cells. Release of R6G inside cells is inferred by the

apparent diffuse staining of cells as opposed to punctated bright spots, which would be

expected if R6G-loaded gels were adhering to the cell surface. 20x objective, numerical

aperture NA= 0.4, λex = 480 ± 40 nm (100 W Hg lamp), λem = 535 nm.


Figure 6.5: Cell viability after 36 hours of incubation with doxorubicin-loaded

polyNIPAM-AAc hydrogels. The drug delivery conditions tested in vitro were: (a)

transferrin-conjugated Dox-loaded microgels; (b) Dox-loaded microgels in solution with

free transferrin (no conjugation); (c) albumin-conjugated Dox-loaded microgels; (d) plain

Dox-loaded microgels (no conjugation); and (e) transferrin-conjugated plain microgels (in

the absence of Dox). (Mean ± SD, n=3)


Figure 6.6: Cell viability under methotrexate treatment in targeted chitosan-based

drug delivery vehicles. Viability (non-filled)/mortality (striped) of HeLa cells after 24

hour incubation with (a) no microgels or MTX; (b) MTX; (c) CS-based microgels; (d)

transferrin-conjugated CS-based microgels; (e) MTX-loaded CS-based microgels; and (f)

transferrin-conjugated MTX-loaded CS-based microgels. (Mean ± SD, n=3)


%, respectively. The transferrin-conjugated MTX-loaded CS-based microgels demon-

strated a significant (4.8- and 4.5-fold, p < 0.01) increase in cell mortality of HeLa cells

compared to the pure drug and non-bioconjugated MTX-loaded microgels. Two factors

enhanced the efficiency of the biofunctionalized pH-responsive microgel drug carriers in

suppressing HeLa cells. First, bioconjugated microgels were efficiently internalized into

the tumour cells through transferrin receptor-mediated endocytosis. Second, due to pH-

induced swelling of the microgels a faster diffusion-driven release of MTX was achieved

in late endosomes, which has a reported pH of 5.0 [104].

6.4 Conclusion

Two targeted drug delivery vehicles were designed to meet the following objectives:

1. size small enough to potentially take advantage of the EPR effect

2. protection of the chemotherapeutic drug and maintenance of its efficacy

3. intracellular release of the chemotherapeutic

4. targeted delivery to cancer cells

According to size measurements by photon correlation spectroscopy and electron mi-

croscopy, both microgels have diameters less than 200nm, which would meet objective 1

[77]. Whether or not the particles actually do take advantage of the EPR effect can only

be seen through in vivo testing.

Objective 2, protection of the chemotherapeutic drug and maintenance of its ef-

ficacy was demonstrated in Figures 6.5 and 6.6, which show that the drugs loaded

in transferrin-conjugated microgels are able to effect toxicity on their target. These

transferrin-conjugated microgels were designed to undergo receptor-mediated endocyto-

sis and would be exposed to a drop in pH that would cause them to release the drug.

Had the drug delivery vehicle not protected the drug, we would expect the drug to be

leaked into the environment, thereby resulting in equal amounts of cell death in the con-


trols with drug-loaded microgels not conjugated to transferrin, or a lack of toxicity in

the transferrin-conjugated drug-loaded microgel due to degradation of the drug. Neither

of these were observed.

Objective 3, intracellular release, can be deduced through in vitro testing, which

showed increased mortality only when the drug-loaded hydrogel was conjugated to trans-

ferrin, which undergoes receptor-mediated endocytosis. This same evidence could also be

used to say that objective 4, targeted delivery, was met. However, one of the objectives

of targeted delivery is increased selectivity in order to protect non-target cells from the

drug. The monoculture of HeLa cells does not serve as a platform in which to test this

particular objective.

In the future, additional experiments could be performed to test these drug delivery

vehicles in a more realistic environment to see if they still meet the objectives. For

example, a co-culture with a non-malignant cell line or one that does not overexpress

transferrin receptors could show if the transferrin confers the drug delivery vehicle with

selective targeting capabilities or if it simply increases their cellular uptake.

Testing in the in vivo environment would also be of interest because this is where

adsorption of serum proteins, opsonization, phagocytosis, capture into the reticuloen-

dothelial system, non-specific adsorption or uptake into non-target cells or organs and

the associated side effects would come into play. The systemic toxicity of the polymer

hydrogel and how it is broken down or excreted could also be elucidated.

Overall, within the limits of in vitro testing, two drug delivery vehicles which meet

the design objectives were demonstrated. These drug delivery vehicles illustrate some of

the potential that nanoparticles have in biomedical applications. Advances in chemistry,

physics and biology are utilized to design particles with chemical, physical properties

tailored for a specific biological environment. The nano-bio interface is exploited for

targeting and triggered release of the drug. These developments have the potential to

increase the efficacy and reduce the side effects of chemotherapy for improved prognosis.

Chapter 7

Conclusions and Future Work

This chapter will summarize the conclusions and contributions of the projects described

in this thesis. Future directions, which may be pursued as a continuation of this work,

will also be described.

7.1 Conclusions and contributions

This work is expected to have an impact in a number of fields, including the develop-

ment of quantum dot-based strategies for diagnostic applications, the identification of

biorecognition molecules and the design of targeted drug delivery vehicles. The study of

the variation in intensity of fluorescence emission of ZnS-capped CdSe quantum dots in

biologically relevant buffers motivated the development of strategies to design, synthesize

and read quantum dot barcodes. Section 7.2 describes some of these strategies, which

mitigate the effects of environmentally induced fluorescence variations on quantum dot

barcodes.

The phage-based assay offers a way of quantifying quantum dots, that does not de-

pend on their spectroscopic measurements. Using phage display to identify nanoparticle-

specific bacteriophage, we quantified MAA-coated ZnS-capped CdSe quantum dots in

environments where absorbance or fluorescence spectroscopy are ineffective. Methods

99

Chapter 7. Conclusions and Future Work 100

such as this, which make it possible to study and track these nanoparticle probes in

various environments, would facilitate the verification and testing that will necessarily

precede widespread practical applications.

The work on phage display on whole cells and high throughput phage display demon-

strate some of the strengths and weaknesses of phage display as a method to identify

novel targeting molecules. A peptide,was identified using whole cell phage display and

then conjugated to quantum dots for fluorescent labeling of HeLa cells. However, the

fact that only a single clone was left in the library after just three rounds of panning

was a matter of concern. One would expect that panning on such different cells as HeLa

(human cervical carcinoma) and STO (mouse fibroblast) would result in many peptides

that would selectively bind to one target and not the other. Multiple wash steps and

rounds of amplification may have lead to the loss of binding clones and favour the growth

of some clones over others, leading them to dominate the library. In addition, although

trypsinizing the cells prior to phage display panning may have increased the possibility

of finding a biorecognition molecule for a carbohydate-based marker, it also reduced the

number of potential targets on the cell surface.

A high throughput phage display strategy was developed using quantum dot-encoded

microbead barcodes and fluorescence activated cell sorting. This technique has the po-

tential to enable the selection of targeting molecules for multiple targets simultaneously.

In an experiment with three different targets, a biorecognition molecule that recognized

one target was found and its specificity to its target was verified. Phage that bound to

the other two targets were also identified, but specific binding to their respective targets

was not verified. This highlights another challenge in phage display panning, in that the

panning process will sometimes yield weak or non-target binders [144].

In the realm of therapeutics, the rational design of targeted drug delivery vehicles

guided by knowledge of tumour biology was demonstrated. Microgels were formulated

to be small enough to take advantage of the enhanced permeability and retention effect.


The drug payloads were incorporated into the microgels through physical, rather than

chemical means to prevent any loss of activity that chemical conjugation might cause.

These microgels were then conjugated to transferrin, whose receptor is overexpressed in

many cancers. This allowed the drug delivery vehicles to enter the cells through receptor-

mediated endocytosis where the pH of the late endosome would trigger the release of their

drug payloads.

7.2 Future work: Development of quantum dots for

diagnostic applications

Some work has already been done as a result of the study of the variability of quantum

dot fluorescence in chapter 2. Highlighting the sensitivity of quantum dot fluorescence to

their external environment motivated the development of a barcode design and detection

strategy that would be robust against environmentally induced variations in fluorescence.

Individual codes were designed based on unique ratios of quantum dot colours rather

than intensities [107]. These ratio-based barcodes are more stable than intensity-based

barcodes in the face of variations of fluorescence intensity due to environmental factors.

As long as the environment causes the fluorescence intensity of the colours in one barcode

to change the same way (e.g. all decrease in intensity by 50%), the reading of a barcode

designed based on the ratio of quantum dot colours will not change. To further mediate

the effects of the external environment on the barcode reading, the Chan lab developed a

method of synthesizing quantum dot barcodes, which protects and insulates the quantum

dots from their external environment [106].

On the other hand, changes in nanoparticle fluorescence intensity in response to

changes in their external environment can be exploited in the development of probes.

For example, Chen and Rosenzweig developed selective ion probes from cadmium sulfide

quantum dots. When capped with polyphosphate, the luminescence of these QDs were


sensitive to nearly all mono- and divalent cations (i.e. the luminescence changed in re-

sponse to changes in concentration of these ions in the environment). When capped with

thioglycerol, however, the QDs were only sensitive to copper and iron ions. QDs capped

with L-cysteine were sensitive only to zinc ions, with photoluminescence increasing in

increasing concentrations of zinc ions. As zinc ion probes, L-cysteine and thioglycerol-

capped QD probes have a dynamic range up to 20 µM, and a detection limit of ∼0.8

µM. The emission enhancement caused by zinc (II) ions is attributed to the activation

of surface states in the quantum dots. [145]

The assay used to quantify QDs using bacteriophage isolated through phage display

panning shown in chapter 3 furthers a new paradigm for combining nanoparticles and

biorecognition molecules. The classical paradigm is one where the biorecognition molecule

is conjugated to the nanoparticle to give it targeting capabilities in active targeting or

labeling/detection strategies, as illustrated in Figure 1.2. Another paradigm is where

biorecognition molecules are used to enable the organized self-assembly of nanoparticles

via biological scaffolding. Bacteriophage obtained from phage display banning have been

used to template the assembly of nanoparticles into films [113], wires [114], fibers [146],

and nanowires for lithium ion battery electrodes [147].

Here, recognizing the many similarities between nanoparticles and biological molecules

(i.e. in size and surface chemistry), we use a strategy from molecular biology, namely an

enzyme-linked immunosorbant assay, and apply it to nanoparticles. In molecular biology,

antibodies are an invaluable tool for detecting, quantifying, purifying and manipulating

biological molecules. These molecules have similar weights and molecular compositions

but vastly different biological functions and activities. Through molecular complementar-

ity, antibodies will bind to their antigens with exquisite specificity, enabling the detection

or manipulation of a particular type of molecule (i.e. the antigen to this antibody) amidst

a background of a plethora of other molecules. The technique described in chapter 3

could potentially be expanded so that nanoparticle-specific biorecognition molecules can


serve as the new antibodies for nanoparticles. Then, nanoparticle-specific biorecognition

molecules could purify, detect, quantify and manipulate nanoparticles analogous to the

ways that antibodies are used to purify, detect, quantify and manipulate proteins.

7.3 Future work: Identification of targeting molecules

The whole cell phage display panning process may be modified to identify more mean-

ingful biorecognition molecules. To preserve membrane-bound proteins, EDTA instead

of Trypsin should be used to release adherent cells from their substrate. The cell lines

used for subtraction and selection should be more similar and relevant to a specific ap-

plication. Clones may also be obtained and tested from earlier rounds of panning, before

they are lost through multiple rounds of washing and amplification.

For the immediate future, the high throughput phage display technique introduced in

chapter 5 should be refined to demonstrate its feasibility in identifying multiple targets

simultaneously. Refinement may involve reducing the number of wash steps and harvest-

ing a larger number of clones and then testing for their functionality (rather than doing

multiple rounds of panning until the library is reduced to very few clones).

This technique, which utilizes fluorescence assisted cell sorting, may also be used in

whole cell phage display by combining the subtraction and selection steps. The cells used

for subtraction and selection could be stained different colours, incubated with the cell

library and then separated by FACS. The FACS process could also simultaneously serve

as the subtraction, selection and washing steps, as the sheath fluid in the flow cytometer

can wash away unbound phage. This would incorporate a higher level of automation and

consistency into the panning process.


7.4 Future work: Rational design of drug delivery

vehicles

The rational design of drug delivery vehicles is one of the most promising applications

of nanoparticles in medicine. Future work can go in many directions. The drug delivery

vehicles demonstrated in vitro in a monoculture can be tested in a co-culture to assess

the specificity of the targeting. Towards clinical applications, these drug delivery vehicles

may be tested in vivo. In addition, the targeting molecule and the chemotherapeutic used

in the drug delivery vehicle may be tailored for a specific tumour.

Moreover, although nanoparticle-based chemotherapeutics are using known drugs that

have been in use for decades, one of the most promising areas of research is in the use

of drug delivery vehicles to bring new molecular entities (NME) to the clinic. In the

process of drug discovery, many promising NMEs do not make it to clinical applications

because of poor solubility, low circulating half-lifes or toxic side effects. By overcoming

their drawbacks, nanoparticle drug delivery vehicles can offer new life to these NMEs.

Another promising area of research is multifunctional nanoparticles that simultane-

ously deliver therapeutics and diagnostics, or “theragnostics”. Drug delivery vehicles

may be conjugated to quantum dots or other nanoparticle probes to treat and monitor

the tumour.

Nanoparticles may also be used to trigger the release of the drug once it reaches

its target site. For example, gold nanorods and nanoshells can be designed to efficiently

convert near-infrared light into heat. These nanoparticles can trigger the release of a drug

from temperature sensitive microgels. This external trigger would provide an additional

level of control over the timing and location of drug release.

Bibliography

[1] M. Wang and M. Thanou. Targeting nanoparticles to cancer. Pharmacological

Research, 62:90–99, 2010. 1, 1.1

[2] B. Y. S. Kim, J. T. Rutka, and W.C.W. Chan. Nanomedicine. The New England

Journal of Medicine, 365(25):2434–2443, 2010. 1

[3] M. Ferrari. Cancer nanotechnology: Opportunities and challenges. Nature Reviews

Cancer, 5(3):161–171, 2005. 1

[4] D. Peer, J.M. Karp, S. Hong, O.C. Farokhzad, R. Margalit, and R. Langer.

Nanocarriers as an emerging platform for cancer therapy. Nature Nanotechnology,

2:751–760, 2007. 1, 1.1, 6.1

[5] O.C. Farokhzad and R. Langer. Impact of nanotechnology on drug delivery. ACS

Nano, 3(1):16–20, 2009. 1

[6] A.P. Alivisastos. Semiconductor clusters, nanocrystals, and quantum dots. Science,

271:933–937, February 1996. 1, 1.1, 1.2.1

[7] L.R. Hirsch, R.J. Stafford, J.A. Bankson, S.R. Sershen, B. Rivera, R.E. Price,

J.D. Hazle, N.J. Halas, and J.L. West. Nanoshell-mediated near-infrared thermal

therapy of tumors under magnetic resonance guidance. Proceedings of the Na-

tional Academy of Sciences of the United States of America, 100(23):13549–13554,

November 2003. 1, 1.1, 1.4.1

105

Bibliography 106

[8] Y. Sun and Y. Xia. Gold and silver nanoparticles: A class of chromophores with

colors tunable in the range from 400 to 750 nm. The Analyst, 128:686–691, 2003. 1

[9] X. Gao, Y. Cui, R.M. Levenson, L.W.K. Chung, and S. Nie. In vivo cancer targeting

and imaging with semiconductor quantum dots. Nature Biotechnology, 22(8):969–

976, August 2004. 1, 1.2, 1.2.1, 1.4.1, 2.1

[10] Mrinmoy De, P.S. Ghosh, and V.M. Rotello. Applications of nanoparticles in

biology. Advanced Materials, 20(22):4225–4241, 2008. 1, 3.1

[11] N.L. Rosi and C.A. Mirkin. Nanostructures in biodiagnostics. Chemical Reviews,

105:1547–1562, 2005. 1

[12] C. Loo, A. Lowery, N. Halas, J. West, and R. Drezek. Immunotargeted nanoshells

for integrated cancer imaging and therapy. Nano Letters, 5(4):709–711, 2005. 1

[13] L. Brannon-Peppas and J.O. Blanchette. Nanoparticle and targeted systems for

cancer therapy. Advanced Drug Delivery Reviews, 56(1649-1659), 2004. 1

[14] J. Cheng, B.A. Teply, I. Sherifi, J. Sung, G. Luther, F.X. Gu, E. Levy-Nissenbaum,

A.F. Radovic-Moreno, R. Langer, and O.C. Farokhzad. Formulation of function-

alized PLGA-PEG nanoparticles for in vivo targeted drug delivery. Biomaterials,

28:869–876, 2007. 1

[15] Nanoscience and nanotechnologies: opportunities and uncertainties. Technical re-

port, The Royal Society & The Royal Academy of Engineering, July 2004. 1.1

[16] M. Mandelkern, J.G. Elias, D. Eden, and D.M. Crothers. The dimensions of DNA

in solution. Journal of Molecular Biology, 152(1):153–161, 1981. 1.1, 1.1

[17] V.R. Sarma, E.W. Silverton, D.R. Davies, and W.D. Terry. The three-dimensional

structure at 6 A resolution of a human gamma GI immunoglobulin molecule. Jour-

nal of Biological Chemistry, 246(11):3753–3759, 1971. 1.1

Bibliography 107

[18] E. Roduner. Size matters: why nanomaterials are different. Chemical Society

Reviews, 35:583–592, 2006. 1.1

[19] C. Burda, X. Chen, R. Narayanan, and M.A. El-Sayed. Chemistry and properties

of nanocrystals of different shapes. Chemical Reviews, 105(4):1025–1102, 2005. 1.1

[20] B.O. Dabbousi, J. Rodriguez-Viejo, F.V. Mikulec, J.R. Heine, H. Mattoussi,

R. Ober, K.F. Jensen, and M.G. Bawendi. (CdSe)ZnS core-shell quantum dots:

Synthesis and characterization of a size series of highly luminescent nanocrystal-

lites. Journal of Physical Chemistry B, 101:9463–9475, 1997. 1.1, 2.1

[21] S.J. Oldenburg, R.D. Averitt, S.L. Westcott, and N.J. Halas. Nanoengineering of

optical resonances. Chemical Physics Letters, 288:243–247, 1998. 1.1

[22] R. Weissleder. A clearer vision for in vivo imaging. Nature Biotechnology, 19:316–

317, April 2001. 1.1

[23] C.J. Murphy, T.K. Sau, A.M. Gole, C.J. Orendorff, J. Gao, L. Gou, S.E. Hun-

yadi, and T. Li. Anisotropic metal nanoparticles: Synthesis, assembly, and optical

applications. Journal of Physical Chemistry B, 109:13857–13870, 2005. 1.1

[24] X. Huang, I.H. El-Sayed, W. Qian, and M.A. El-Sayed. Cancer cell imaging and

photothermal therapy in the near-infrared region by using gold nanorods. Journal

of the American Chemical Society, 128(6):2115–2120, 2006. 1.1, 1.4.1

[25] T.S. Hauck, T.L. Jennings, T. Yatsenko, J.C. Kumaradas, and W.C.W. Chan. En-

hancing the toxicity of cancer chemotherapeutics with gold nanorod hyperthermia.

Advanced Materials, 20(20):3832–3838, 2008. 1.1, 1.4.1

[26] P.K. Jain, X. Huang, I.H. El-Sayed, and M.A. El-Sayed. Noble metals on the

nanoscale: Optical and photothermal properties and some applications in imaging,

Bibliography 108

sensing, biology and medicine. Accounts of Chemical Research, 41(12):1578–1586,

2008. 1.1

[27] G.M. Whitesides. The ’right’ size in nanobiotechnology. Nature Biotechnology,

21(10):1161–1165, October 2003. 1.1

[28] Statistics Canada. Mortality, summary list of causes 2007. 2010. 1.1

[29] C. Corot, P. Robert, J.-M. Idee, and M. Port. Recent advances in iron oxide

nanocrystal technology for medical imaging. Advanced Drug Delivery Reviews,

58:1471–1504, 2006. 1.1, 1.2, 1.4.1

[30] Cancer facts & figures 2003. Technical report, American Cancer Society, 2003. 1.2

[31] M.D. Ullah and M. Aatif. The footprints of cancer development. Cancer Treatment

Reviews, 35:193–200, 2009. 1.2

[32] R. Elghanian, J.J. Storhoff, R.C. Mucic, R.L. Letsinger, and C.A. Mirkin. Selective

colorimetric detection of polynucleotides based on the distance-dependent optical

properties of gold nanoparticles. Science, 277:1078–1081, August 1997. 1.2

[33] E.B. Voura, J.K. Jaiswal, H. Mattoussi, and S.M. Simon. Tracking metastatic

tumor cell extravasation with quantum dot nanocrystals and fluorescence emission-

scanning microscopy. Nature Medicine, 10(9):993–998, September 2004. 1.2, 1.2.1,

2.1

[34] A.A. Ghazani, J.A. Lee, J.M. Klostranec, Q. Xiang, R.S. Dacosta, B.C. Wilson,

M.S. Tsao, and W.C.W. Chan. High throughput quantification of protein expres-

sion of cancer antigens in tissue microarray using quantum dot nanocrystals. Nano

Letters, 6(2):2881–2886, 2006. 1.2, 1.2.1

[35] X. Wu, H. Liu, J. Liu, K.N. Haley, J.A. Treadway, J.P. Larson, N. Ge, F. Peale,

and M.P. Bruchez. Immunofluorescent labeling of cancer marker Her2 and other

Bibliography 109

cellular targets with semiconductor quantum dots. Nature Biotechnology, 21(1):41–

46, January 2003. 1.2, 1.2.1

[36] S. Kim, Y.T. Lim, E.G. Soltesz, A.M. De Grand, J. Lee, A. Nakayama, J.A. Parker,

T. Mihaljevic, R.G. Laurence, D.M. Dor, L.H. Cohn, M.G. Bawendi, and J.V.

Frangioni. Near-infrared fluorescent type ii quantum dots for sentinel lymph node

mapping. Nature Biotechnology, 22(1):93–97, January 2004. 1.2, 1.2.1, 2.1

[37] I.L. Medintz, H.T. Uyeda, E.R. Goldman, and H. Mattoussi. Quantum dot biocon-

jugates for imaging, labelling and sensing. Nature Materials, 4(6):435–446, 2005.

1.2.1, 2.1

[38] W.C.W. Chan, D.J. Maxwell, X. Gao, R.E. Bailey, M. Han, and S. Nie. Lumi-

nescent quantum dots for multiplexed biological detection and imaging. Current

Opinion in Biotechnology, 13(1):40–46, 2002. 1.2.1

[39] M. Dahan, S. Levi, C. Luccardini, P. Rostaing, B. Riveau, and A. Triller. Diffusion

dynamics of glycine receptors revealed by single-quantum dot tracking. Science,

302:442–445, October 2003. 1.2.1, 2.1

[40] M. Dahan, T. Laurence, F. Pinaud, D.S. Chemla, A.P. Alivisastos, M. Sauer, and

S. Weiss. Time-gated biological imaging by use of colloidal quantum dots. Optics

Letters, 26(11):825–827, June 2001. 1.2.1

[41] J.K. Jaiswal, H. Mattoussi, J.M. Mauro, and S.M. Simon. Long-term multiple

color imaging of live cells using quantum dot bioconjugates. Nature Biotechnology,

21:47–51, January 2003. 1.2.1, 2.1

[42] M. Han, X. Gao, J.Z. Su, and S. Nie. Quantum-dot-tagged microbeads for mul-

tiplexed optical coding of biomolecules. Nature Biotechnology, 19:631–635, July

2001. 1.2.1, 2.1, 5.1

Bibliography 110

[43] J.A. Kloepfer, S.E. Bradforth, and J.L. Nadeau. Photophysical properties of bio-

logically compatible CdSe quantum dot structures. Journal of Physical Chemistry

B, 109(20):996–10003, 2005. 1.2.1, 2.1, 3.1

[44] K. Boldt, O.T. Bruns, N. Gaponik, and A. Eychmuller. Comparative examination

of the stability of semiconductor quantum dots in various biochemical buffers. The

Journal of Physical Chemistry B, 110:1959–1963, 2006. 1.2.1, 2.1, 2.3, 2.3

[45] S.S. Iqbal, M.W. Mayo, J.G. Bruno, B.V. Bronk, C.A. Batt, and J.P. Chambers.

A review of molecular recognition technologies for detection of biological threat

agents. Biosensors & Bioelectronics, 15:549–578, 2000. 1.3.1

[46] A.K. Abbas and C.A. Janeway Jr. Immunology: Improving on nature in the twenty-

first century. Cell, 100:129–138, 2000. 1.3.1

[47] G. Kohler and C. Milstein. Continuous cultures of fused cells secreting antibody

of predefined specificity. Nature, 256(5517):495–497, August 1975. 1.3.1

[48] H. Lodish, A. Berk, P. Matsudaira, C.A. Kaiser, M. Krieger, M.P. Scott,

L. Zipursky, and J. Darnell. Molecular Cell Biology. W. H. Freeman and Company,

5th edition edition, 2002. 1.3.1

[49] J.S. Ross, D.P. Schenkein, R. Pietrusko, M. Rolfe, G.P. Linette, J. Stec, N.E.

Stagliano, G.S. Ginsburg, W.F. Symmans, L. Pusztai, and G.N. Hortobagyi.

Targeted therapies for cancer 2004. American Journal of Clinical Pathology,

122(4):598–609, 2004. 1.3.1

[50] R.M. Nakamura, W.W. Grody, J.T. Wu, and R.B. Nagle, editors. Cancer Diagnos-

tics: Current and Future Trends, chapter 4, pages 35–52. Humana Press: Totowa,

New Jersey, 2004. 1.3.1

Bibliography 111

[51] A.D. Ellington and J.W. Szostak. In vitro selection of rna molecules that bind

specific ligands. Nature, 346(6287):818–822, August 1990. 1.3.2

[52] C. Tuerk and L. Gold. Systemic evolution of ligands by exponential enrichment:

RNA ligands to bacteriophage T4 DNA. Science, 249(4968):505–510, 1990. 1.3.2

[53] G.P. Smith. Filamentous fusion phage: Novel expression vectors that display cloned

antigens on the virion surface. Science, 228(4705):1315–1317, June 1985. 1.3.3, 4.1,

4.3

[54] G.P. Smith and V.A. Petrenko. Phage display. Chemical Reviews, 97(2):3910410,

1997. 1.3.3, 4.1

[55] S.E. Cwirla, E.A. Peters, R.W. Barrett, and W.J. Dower. Peptides on phage: A

vast library of peptides for identifying ligands. Proceedings of the National Academy

of Sciences of the United States of America, 87:6378–6382, August 1990. 1.3.3

[56] J.J. Devlin, L.C. Panganiban, and P.E. Devlin. Random peptide libraries: A source

of specific protein binding molecules. Science, 249:404–406, July 1990. 1.3.3

[57] C.T. Walsh, R.D. Schwartz-Bloom, and R.R. Levine. Levine’s Pharmacology - Drug

Actions and Reactions. Taylor & Francis, seventh edition, 2005. 1.4

[58] J.F. Kukowska-Latallo, K.A. Candido, Z. Cao, S.S. Nigavekar, I.J. Majoros, T.P.

Thomas, L.P. Balogh, M.M. Khan, and J.R. Baker Jr. Nanoparticle targeting of

anticancer drug improves therapeutic response in animal model of human epithelial

cancer. Cancer Research, 65(12):5317–5324, 2005. 1.4

[59] W. Arap, R. Pasqualini, and E. Ruoslahti. Cancer treatment by targeted drug de-

livery to tumor vasculature in a mouse model. Science, 279(5349):377–380, January

1998. 1.4, 4.1

Bibliography 112

[60] N.W.S. Kam, M. O’Connell, J.A. Wisdom, and H. Dai. Carbon nanotubes as

multifunctional biological transporters and near-infrared agents for selective cancer

cell destruction. Proceedings of the National Academy of Sciences of the United

States of America, 102(33):11600–11605, August 2005. 1.4.1

[61] A.M. Gobin, M.H. Lee, N.J. Halas, W.D. James, R.A. Drezek, and J.L. West.

Near-infrared resonant nanoshells for combined optical imaging and photothermal

cancer therapy. Nano Letters, 2(7):1929–1934, 2007. 1.4.1

[62] G.R. Martin and R. K. Jain. Noninvasive measurement of interstitial pH profiles in

normal and neoplastic tissue using fluorescence ratio imaging microscopy. Cancer

Research, 54(21):5670–5674, 1994. 1.4.1, 2.3

[63] B.D. Chithrani and W.C.W. Chan. Elucidating the mechanism of cellular uptake

and removal of protein-coated gold nanoparticles of different sizes and shapes. Nano

Letters, 7(6):1542–1550, 2007. 1.4.1

[64] J.A. Champion and S. Mitragotri. Role of target geometry in phagocytosis. Pro-

ceedings of the National Academy of Sciences of the United States of America,

103(13):4930–4934, March 2006. 1.4.1

[65] K. Kostarelos, L. Lacerda, G. Pastorin, W. Wu, S. Wieckowski, J. Luansivilay,

S. Godefroy, D. Pantarotto, J.-P. Briand, S. Muller, M. Prato, and A. Bianco.

Cellular uptake of functionalized carbon nanotubes is independent of functional

group and cell type. Nature Nanotechnology, 2:108–113, February 2008. 1.4.1

[66] Y. Geng, P. Dalhaimer, S. Cai, R. Tsai, M. Tewari, T. Minko, and D.E. Discher.

Shape effects of filaments versus spherical particles in flow and drug delivery. Nature

Nanotechnology, 2:249–255, April 2007. 1.4.1

[67] M.A. Dobrovolskaia and S.E. McNeil. Immunological properties of engineered nano-

materials. Nature Nanotechnology, 2:469–478, August 2007. 1.4.1

Bibliography 113

[68] M. Lundqvist, J. Stigler, G. Elia, I. Lynch, T. Cedervall, and K. A. Dawson.

Nanoparticle size and surface properties determine the protein corona with possible

implications for biological impacts. Proceedings of the National Academy of Sciences

of the United States of America, 105(38):14265–14270, September 2008. 1.4.1

[69] A. Verma, O. Uzun, Y. Hu, H-S. Han, N. Watson, S. Chen, D.J. Irvine, and

F. Stellacci. Surface-structure-regulated cell-membrane penetration by monolayer-

protected nanoparticles. Nature Materials, 7:588–595, July 2008. 1.4.1

[70] H. Soo Choi, W. Liu, P. Misra, E. Tanaka, J.P. Zimmer, B. Itty Ipe, M.G. Bawendi,

and J.V. Frangioni. Renal clearance of quantum dots. Nature Biotechnology,

25(10):1156–1170, October 2007. 1.4.1

[71] R. Gref, Y. Minamitake, M.T. Peracchia, V. Trubetskoy, V. Torchilin, and

R. Langer. Biodegradable long-circulating polymeric nanospheres. Science,

263(5153):1600–1603, 1994. 1.4.1

[72] M.E. Akerman, W.C.W. Chan, P. Laakkonen, S.N. Bhatia, and E. Ruoslahti.

Nanocrystal targeting in vivo. Proceedings of the National Academy of Sciences

of the United States of America, 99(20):12617–12621, October 2002. 1.4.1, 4.1

[73] T.B. Huff, M.N. Hansen, Y. Zhao, J-X Cheng, and A. Wei. Controlling the cellular

uptake of gold nanorods. Langmuir, 23(4):1596–1599, 2007. 1.4.1

[74] T.S. Hauck, A.A. Ghazani, and W.C.W. Chan. Assessing the effect of surface

chemistry on gold nanorod uptake, toxicity, and gene expression in mammalian

cells. Small, 4(1):153–159, January 2008. 1.4.1

[75] H.C. Fischer, L. Liu, K.S. Pang, and W.C.W. Chan. Pharmacokinetics of nanoscale

quantum dots: In vivo distribution, sequestration, and clearance in the rat. Ad-

vanced Functional Materials, 16(10):1299–1305, 2006. 1.4.1, 3.1

Bibliography 114

[76] E. Ruoslahti. Specialization of tumor vasculature. Nature Reviews Cancer, 2:83–90,

2002. 1.4.1

[77] S.K. Hobbs, W. L. Monsky, F. Yuan, W. G. Roberts, L. Griffith, V. P. Torchilin,

and R. K. Jain. Regulation of transport pathways in tumor vessels: role of tumor

type and microenvironment. Proceedings of the National Academy of Sciences of

the United States of America, 95(8):4607–4612, 1998. 1.4.1, 6.4

[78] Y. Matsumura and H. Maeda. A new concept for macromolecular therapeutics in

cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the

antitumor agent smancs. Cancer Research, 46:6387–6392, December 1986. 1.4.1

[79] B.D. Chithrani, A.A. Ghazani, and W.C.W. Chan. Determining the size and

shape dependence of gold nanoparticle uptake into mammalian cells. Nano Letters,

6(4):662–668, 2006. 1.4.1

[80] W. Jiang, B. Y. S. Kim, J. T. Rutka, and W.C.W. Chan. Nanoparticle-mediated

cellular response is size-dependent. Nature Nanotechnology, 3:145–150, 2008. 1.4.2

[81] X Qian, X.-H. Peng, D.O. Ansari, Q. Yin-Goen, G.Z. Chen, D.M. Shin, L. Yang,

A.N. Young, M.D. Wang, and S. Nie. In vivo tumor targeting and spectroscopic

detection with surface-enhanced Raman nanoparticle tags. Nature Biotechnology,

26(1):83–90, 2008. 1.4.2

[82] W.C.W. Chan and S. Nie. Quantum dot bioconjugates for ultrasensitive nonisotopic

detection. Science, 281(5385):2016–2018, 1998. 1.4.2, 2.1, 3.2.1

[83] O.C. Farokhzad, S. Jon, A. Khademhosseini, T.-N.T. Tran, D.A. LaVan, and

R. Langer. Nanoparticle-aptamer bioconjugates: A new approach for targeting

prostate cancer cells. Cancer Research, 64(21):7668–7672, November 2004. 1.4.2

Bibliography 115

[84] W. Cai, D.-W. Shin, K. Chen, O. Gheysens, Q. Cao, S.X. Wang, S.S. Gambhir,

and X. Chen. Peptide-labeled near-infrared quantum dots for imaging tumor vas-

culature in living subjects. Nano Letters, 6(4):669–676, 2006. 1.4.2

[85] R. Weissleder, K. Kelly, E.Y. Sun, T. Shtatland, and L. Josephson. Cell-specific

targeting of nanoparticles by multivalent attachment of small molecules. Nature

Biotechnology, 23(11):1418–1423, 2005. 1.4.2

[86] M.A. Hines and P. Guyot-Sionnest. Synthesis and characterization of strongly lu-

minescing ZnS-capped CdSe nanocrystals. Journal of Physical Chemistry, 100:468–

471, 1996. 2.1, 3.2.1

[87] B. Dubertret, P. Skourides, D.J. Norris, V. Noireaux, A.H. Brivanlou, and A. Libch-

aber. In vivo imaging of quantum dots encapsulated in phospholipid micelles. Sci-

ence, 298:1759–1762, November 2002. 2.1

[88] X. Gao and S. Nie. Quantum dot-encoded beads. Methods in Molecular Biology,

303:61–71, 2005. 2.1

[89] K.D. Kepler, G.C. Lisensky, M. Patel, L.A. Sigworth, and A.B. Ellis. Surface-bound

carbonyl compounds as lewis acids. photoluminescence as a probe for the binding of

ketones and aldehydes to cadmium sulfide and cadmium selenide surfaces. Journal

of Physical Chemistry, 99(43):16011–16017, 1995. 2.1, 2.3

[90] J.K. Lorenz, T.F. Kuech, and A.B. Ellis. Cadmium selenide photoluminescence as

a probe for the surface adsorption of dialkyl chalcogenides. Langmuir, 14(7):1680–

1683, 1998. 2.1

[91] A. Ivanisevic and A.B. Ellis. Photoluminescent properties of cadmium selenide in

contact with solutions and films of metalloporphyrins. evidence for semiconductor-

mediated adduct formation of oxygen with metalloporphyrins at room temperature.

Journal of Physical Chemistry B, 103(11), 1914-1919 1999. 2.1

Bibliography 116

[92] R.J. Brainard, C.A. Paulson, D. Saulys, D.F. Gaines, T.F. Kuech, and A.B.

Ellis. Modulation of cadmium selenide photoluminescence intensity by adsorp-

tion of silapentanes and chlorinated silanes. Journal of Physical Chemistry B,

101(51):11180–11184, 1997. 2.1

[93] R.J. Brainard and A.B. Ellis. Photoluminescent properties of cadmium selenide in

contact with solutions and films of vaska’s complex: Effects of oxygen and carbon

monoxide. Journal of Physical Chemistry B, 101(14):2533–2539, 1997. 2.1

[94] G. Kalyuzhny and R.W. Murray. Ligand effects on optical properties of CdSe

nanocrystals. Journal of Physical Chemistry B, 109(15):7012–7021, 2005. 2.1

[95] C.F. Landes, M. Braun, and M.A. El-Sayed. On the nanoparticle to molecular

size transition: Fluorescence quenching studies. Journal of Physical Chemistry B,

105(43):10554–10558, 2001. 2.1

[96] S.A. Majetich and A.C. Carter. Surface effects on the optical properties of cadmium

selenide quantum dots. Journal of Physical Chemistry, 97(34):8727–8731, 1993. 2.1

[97] C. Bullen and P. Mulvaney. The effects of chemisorption on the luminescence of

CdSe quantum dots. Langmuir, 22(7):3007–3013, 2006. 2.1, 2.3

[98] Z.A. Peng and X. Peng. Formation of high-quality CdTe, CdSe, and CdS

nanocrystals using cdo as a precursor. Journal of the American Chemical Soci-

ety, 123(1):183–184, 2001. 2.1

[99] X. Gao, W.C.W. Chan, and S. Nie. Quantum-dot nanocrystals for ultrasensitive

biological labeling and multicolor optical encoding. Journal of Biomedical Optics,

7(4):532–537, 2002. 2.1, 2.3

Bibliography 117

[100] W. Jiang, S. Mardyani, H.C. Fischer, and W.C.W. Chan. Design and characteriza-

tion of lysine cross-linked mercapto-acid biocompatible quantum dots. Chemistry

of Materials, 18(4):872–878, 2006. 2.1, 2.3, 3.2.1

[101] C.B. Murray, D.J. Norris, and M.G. Bawendi. Synthesis and characterization of

nearly monodisperse CdE (E= S, Se, Te) semiconductor nanocrystallites. Journal

of the American Chemical Society, 115:8706–8715, 1993. 2.2.1

[102] W.W. Yu, L. Qu, W. Guo, and X. Peng. Experimental determination of the ex-

tinction coefficient of CdTe, CdSe, and CdS nanocrystals. Chemistry of Materials,

15:2854–2860, 2003. 2.2.3, 2.3, 3.1

[103] T. Dannhauser, M. O’Neil, K. Johansson, D. Whitten, and G. McLendon. Photo-

physics of quantized colloidal semiconductors dramatic luminescence enhancement

by binding of simple amines. Journal of Physical Chemistry, 90:6074–6076, 1986.

2.3, 2.3

[104] S. Mukherjee, R.N. Ghosh, and F.R. Maxfield. Endocytosis. Physiological Reviews,

77(3):759–803, 1997. 2.3, 6.3.3

[105] M.J. Geisow. Fluorescein conjugates as indicators of subcellular pH. Experimental

Cell Research, 150:29–35, 1984. 2.3

[106] S. Fournier-Bidoz, T.L. Jennings, J.M. Klostranec, W. Fung, A. Rhee, D. Li, and

W.C.W. Chan. Facile and rapid one-step mass preparation of quantum-dot bar-

codes. Angewandte Chemie (International Edition), 44:5577–5581, 2008. 2.3, 5.4,

7.2

[107] A.J. Lee, A. Hung, S. Mardyani, A. Rhee, J.M. Klostranec, Y. Mu, D. Li, and

W.C.W. Chan. Toward the accurate read-out of quantum dot barcodes: Design of

deconvolution algorithms and assessment of fluorescence signals in buffer. Advanced

Materials, 19(20):3113–3118, 2007. 2.3, 3.1, 7.2

Bibliography 118

[108] R.A. Sperling, P. Rivera Gil, F. Zhang, M Zanella, and W.J. Parak. Biological

applications of gold nanoparticles. Chemical Society Reviews, 37(9):1896–1908,

2008. 3.1

[109] J.M. Klostranec, Q. Xiang, G.A. Farcas, J.A. Lee, A. Rhee, E.I. Lafferty, S.D.

Perrault, K.C. Kain, and W.C.W. Chan. Convergence of quantum dot barcodes

with microfluidics and signal processing for multiplexed high-throughput infectious

disease diagnostics. Nano Letters, 7(9):2812–2818, 2007. 3.1

[110] W. Jiang, A. Singhal, J. Zheng, C. Wang, and W.C.W. Chan. Optimizing the

synthesis of red- to near-IR-emitting CdS-capped CdTexSe1−x alloyed quantum

dots for biomedical imaging. Chemistry of Materials, 18(20):4845–4854, 2006. 3.1

[111] H.C. Fischer, S. Fournier-Bidoz, K.S. Pang, and W.C.W. Chan. Quantitative detec-

tion of engineered nanoparticles in tissues and organs: An investigation of efficacy

and linear dynamic ranges using ICP-AES. Nanobiotechnology, 3(1):46–54, 2007.

3.1

[112] S.R. Whaley, D.S. English, E.L. Hu, P.F. Barbara, and A.M. Belcher. Selection of

peptides with semiconductor binding specificity for directed nanocrystal assembly.

Nature, 405:665–668, June 2000. 3.1, 4.1

[113] S.-W Lee, C. Mao, C.E. Flynn, and A.M. Belcher. Ordering of quantum dots, using

genetically engineered viruses. Science, 296(5569):892–895, 2002. 3.1, 7.2

[114] C. Mao, C.E. Flynn, A. Hayhurst, R. Sweeney, J. Qi, G. Georgiou, B. Iverson, and

A.M. Belcher. Viral assembly of oriented quantum dot nanowires. Proceedings of

the National Academy of Sciences of the United States of America, 100(12):6946–

6951, 2003. 3.1, 7.2

Bibliography 119

[115] New England BioLabs, 240 County Road, Ipswich, MA, 01938. Ph.D.-12 Phage

Display Peptide Library Kit - Rapid Screening of Peptide Ligands with a Phage

Display Peptide Library, 2.7 edition, November 2007. 3.1

[116] D.V. Talapin, A.L. Rogach, A. Kornowski, M. Haase, and H. Weller. Highly

luminescent monodisperse CdSe and CdSe/ZnS nanocrystals synthesized in a

hexadecylamine-trioctylphosphine oxide-trioctylphosphine mixture. Nano Letters,

4:207–211, 1. 3.2.1

[117] R.A. Sperling, T. Pellegrino, J.K. Li, W.H. Chang, and W.J. Parak. Electrophoretic

separation of nanoparticles with a discrete number of functional groups. Advanced

Functional Materials, 16(7):943–948, 2006. 3.2.1

[118] M.R. Dyson, V. Germaschewski, and K. Murray. Direct measurement via phage

titre of the dissociation constants in solution of fusion phage-substrate complexes.

Nucleic Acids Research, 23(9):1531–1535, 1995. 3.2.5

[119] S. Wang, E.S. Humphreys, S.-Y. Chung, D.F. Delduco, S.R. Lustig, H. Wang, K.N.

Parker, N.W. Rizzo, S. Subramoney, Y.-M. Chiang, and A. Jagota. Peptides with

selective affinity for carbon nanotubes. Nature Materials, 2:196–200, 2003. 3.3.4,

3.4, 4.1

[120] R.L. Willet, K.W. Baldwin, K. W. West, and L.N. Pfeiffer. Differential adhesion of

amino acids to inorganic surfaces. Proceedings of the National Academy of Sciences

of the United States of America, 102(22):7817–7822, 2005. 3.3.4

[121] V.A. Petrenko and G.P. Smith. Phages from landscape libraries as substitute

antibodies. Protein Engineering, 13(8):589–592, 2000. 3.4

[122] E. Eteshola, L.J. Brillson, and S.C. Lee. Selection and characteristics of peptides

that bind thermally grown silicon dioxide films. Biomolecular Engineering, 22:201–

204, 2005. 4.1

Bibliography 120

[123] P. Chan, T. Phan, M.C. Kao, C. Dolan, and J.B.-H. Tok. Generating short peptidic

ligands for silver nanowires from phage display random libraries. Bioorganic &

Medicinal Chemistry Letters, 16:5261–5264, 2006. 4.1

[124] R.R. Naik, S. J. Stringer, G. Agarwal, S.E. Jones, and M.O. Stone. Biomimetic syn-

thesis and patterning of silver nanoparticles. Nature Materials, 1:169–172, Novem-

ber 2002. 4.1

[125] U.B. Rasmussen, V. Schreiber, H. Schultz, F. Mischler, and K. Schugart. Tumor

cell-targeting by phage-displayed peptides. Cancer Gene Therapy, 9:606–612, 2002.

4.1, 4.1

[126] R. Pasqualini and E. Ruoslahti. Organ targeting in vivo using phage display peptide

libraries. Nature, 380:364–366, March 1996. 4.1

[127] K.C. Brown. New approaches for cell-specific targeting: identification of cell-

selective peptides from combinatorial libraries. Current Opinion in Chemical Biol-

ogy, 4(1):16–21, 2000. 4.1, 4.1

[128] G.T. Hermanson. Bioconjugate Techniques. Academic Press, 1996. 4.2.4, 6.2.3

[129] M.M. Fuster and J.D. Esko. The sweet and sour of cancer: glycans as novel

therapeutic targets. Nature Reviews Cancer, 5:526–542, July 2005. 4.3

[130] X. Gao and S. Nie. Quantum dot-encoded mesoporous beads with high bright-

ness and uniformity: Rapid readout using flow cytometry. Analytical Chemistry,

76(8):2406–2410, April 2004. 5.1

[131] M. Das, S. Mardyani, W.C.W. Chan, and E. Kumacheva. Biofunctionalized pH-

responsive microgels for cancer cell targeting. Advanced Materials, 18(1):80–83,

2006. 6, 6.1, 6.2.1

Bibliography 121

[132] H. Zhang, S. Mardyani, W.C.W. Chan, and E. Kumacheva. Design of biocom-

patible chitosan microgels for targeted pH-mediated intracellular release of cancer

therapeutics. Biomacromolecules, 7(5):1568–1572, 2006. 6, 6.2.1

[133] R. Langer and J. Folkman. Polymers for the sustained release of proteins and other

macromolecules. Nature, 263(5580):797–800, 1976. 6.1

[134] B.R. Saunders and B. Vincent. Microgel particles as model colloids: theory, prop-

erties and applications. Advances in Colloid and Interface Science, 80:1–25, 1999.

6.1, 6.3.1

[135] N. Murthy, M. Xu, S. Schuck, J. Kunisawa, N. Shastri, and M.J. Frechet. A macro-

molecular delivery vehicle for protein-based vaccines: Acid-degradable protein-

loaded microgels. Proceedings of the National Academy of Sciences of the United

States of America, 100(9):4995–5000, April 2003. 6.1

[136] S. Nayak, H. Lee, J. Chmielewski, and L.A. Lyon. Folate-mediated cell target-

ing and cytotoxicity using thermoresponsive microgels. Journal of the American

Chemical Society, 126:10258–10259, 2004. 6.1

[137] W.P. Faulk, B.-L. Hsi, and P.J. Stevens. Transferrin and transferrin receptors in

carcinoma of the breast. The Lancet, 2(8191):390–392, May 1980. 6.1

[138] I.S. Trowbridge and M.B. Omary. Human cell surface glycoprotein related to cell

proliferation is the receptor for transferrin. Proceedings of the National Academy

of Sciences of the United States of America, 78(5):3039–3043, May 1981. 6.1

[139] R. Sutherland, D. Delia, C. Schneider, R. Newman, J. Kemshead, and M. Greaves.

Ubiquitous cell-surface glycoprotein on tumor cells is proliferation-associated recep-

tor for transferrin. Proceedings of the National Academy of Sciences of the United

States of America, 78(7):4515–4519, July 1981. 6.1

Bibliography 122

[140] R. Newman, C. Schneider, R. Sutherland, L. Vodinelich, and M. Greaves. The

transferrin receptor. Trends in Biochemical Sciences, 7(11):397–400, November

1982. 6.1

[141] H. Xiao, R. Pelton, and A. Hamielec. Preparation and kinetic characteri-

zation of copolymers of acrylamide and poly(ethylene glycol) (meth)acrylate

macromonomers. Polymer, 37(7):1201–1209, 1996. 6.2.1

[142] A. Fernandez-Nieves, A. Fernandez-Barbero, B. Vincent, and F.J. de las Nieves.

Osmotic de-swelling of ionic microgel particles. Journal of Chemical Physics,

119(19):10383–10388, 2003. 6.3.1

[143] J. Grim, J. Chladek, and Martınkova. Pharmacodynamics of methotrexate in non-

neoplastic diseases. Clinical Pharmacokinetics, 42(2):139–151, 2003. 6.3.1

[144] L. Yang and J.P. Nolan. High-throughput screening and characterization of clones

selected from phage display libraries. Cytometry Part A, 71A:625–631, 2007. 7.1

[145] Y. Chen and Z. Rosenzweig. Luminescent CdS quantum dots as selective ion probes.

Analytical Chemistry, 74(19):5132–5138, October 2002. 7.2

[146] C.-Y. Chiang, C.M. Mello, J. Gu, E.C.C.M. Silva, K.J. Van Vliet, and A.M.

Belcher. Weaving genetically engineered functionality into mechanically robust

fibers. Advanced Materials, 19:826–832, 2007. 7.2

[147] K.T. Nam, D.-W. Kim, P.J. Joo, C.-Y. Chiang, N. Meethong, P.T. Hammond, Y.-

M. Chiang, and A.M. Belcher. Virus-enabled synthesis and assembly of nanowires

for lithium ion battery electrodes. Science, 312:885–888, May 2006. 7.2

Nanoparticles for Cancer Detection and Therapy: …...Abstract Nanoparticles for Cancer Detection...

Documents

Transcript of Nanoparticles for Cancer Detection and Therapy: …...Abstract Nanoparticles for Cancer Detection...