Toward the Development of Nucleic Acid Assays Using Fluorescence Resonance Energy Transfer (FRET) and a Novel

Label Free Molecular Switching Construct

by

Melissa Massey

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Department of Chemistry University of Toronto

© Copyright by Melissa Massey 2011

ii

Toward the Development of Nucleic Acid Assays Using

Fluorescence Resonance Energy Transfer (FRET) and a Novel

Label Free Molecular Switching Construct

Melissa Massey

Doctor of Philosophy

Department of Chemistry University of Toronto

2011

Abstract

The research presented in this thesis introduces design criteria for development of a new type

of self-contained optical biosensor. The study begins with evaluation of a dual label,

fluorescence resonance energy transfer (FRET) bioassay format, and then goes on to

demonstrate a signalling platform that uses an immobilized fluorescent intercalating dye so as

to avoid labelling of both the target and probe strands.

An extensive survey of FRET pairs that can be used to monitor hybridization events in

solution and at solid interfaces was conducted in solution to provide a set of calculated Förster

distances for the extrinsic labels Cyanine 3 (Cy3), Cyanine 5 (Cy5),

Carboxytetramethylrhodamine (TAMRA), Iowa Black Fluorescence Quencher (IabFQ) and

Iowa Black RQ (IabRQ). FRET parameters using thiazole orange (TO) intercalating dye as a

FRET donor for various acceptor dye-labelled DNA conjugates in solution were determined.

Limitations associated with quenching mechanisms other than those mediated by FRET

motivated the development of a molecular switch that contained intercalating dye.

iii

The four binding sites associated with Neutravidin served for assembly of the switch using

biotin interactions. One binding site was used to immobilize an unlabelled oligonucleotide

probe. The adjacent site was used to immobilize a novel biotinylated TO derivative that could

physically reach the probe. On hybridization of the probe with target, the intercalating dye was

captured by the hybrid, leading to a change of fluorescence. This reversible signalling

mechanism offers a method without nucleic acid labelling to detect nucleic acid association at

an interface. A SNP discrimination strategy involving TO and formamide was investigated,

and SNP discrimination without the requirement of thermal denaturation was achieved for

multiple target lengths, including a 141-base pair PCR amplicon in solution. It was determined

that formamide could also provide improvements of signal-to-noise when using thiazole

orange to detect hybridization.

iv

Acknowledgements

I would like to extend my sincere gratitude and thanks to my supervisor, Professor Ulrich J.

Krull. Thank you for your support, guidance, patience and inspiration throughout my

undergraduate and graduate career. Thank you once again for inspiring me to become active

within the university community. The UTM community would not be the same without you. I

have grown as an individual and the campus has grown as a whole thanks to your efforts.

I would also like to thank my committee members, Dr. R. Scott Prosser and Dr. Aaron R.

Wheeler, for their useful suggestions concerning my research and for taking the time to be a

part of my PhD Advisory Committee.

There are several members of the UTM community that I would like thank as well. Thank

you to Dr. Paul Piunno for being a great boss and colleague so many years ago at FONA

Technologies. I am glad that UTM students now have the opportunity to be taught by an

excellent mentor (probably with a few laughs along the way). Thanks to Dr. Peter Mitrakos

for your instrument expertise for both research and teaching purposes. Thank you to Samia

Awadallah and Lin Milne for being friendly faces to chat with and for making the

undergraduate labs run smoothly. Thank you to Carmen Bryson for your administrative

support; it is greatly appreciated that the lab could count on you every time we needed it!

Thank you to the Ontario Ministry of Training, Colleges and Universities, and the University

of Toronto for the provision of scholarships and fellowships, and special thanks to Dina

Gordon Malkin for your generous donations towards the pursuit of education at the University

of Toronto.

My sincere thanks to members of the Chemical Sensors Group, both past and present. In

particular, thank you to Dr. Russ Algar. Thank you for being a friend, colleague, co-author,

and for editing my often ‘long-winded’ work. Thanks also to Andrew Chan, Dr. April Wong,

Dr. Ying Lim and Sameer Al-Abdul-Wahid for many laughs, dinners, and memories. Thanks

to Dr. Taufik Al-Sarraj, Jonathan Cauchi, Dr. Lu Chen, Lori Chong, Yevgenia Kravtsova,

v

Tasmea Mahmud, Eleonora Petryayeva, Anthony Tavares, Uvaraj Uddayasankar, and Dr.

Charles Vannoy for making the lab a great place to work.

Thanks to all my friends outside the academic environment. Thank you to my

swimming/cycling/running friends. We’ve covered a LOT of miles together and I’ve enjoyed

so much of it. I can truly say from experience that graduate school is like a marathon. Thank

you to Lisa Bradley. Your support, kind words, and encouragement over the last 20+ years

have meant so much to me. Thanks to the “TEAM” and to Feo.

Last, but certainly not least, I would like to thank my family. Thank you to my parents and to

my brother for being there with me every step of the way and always believing in me. I love

you.

vi

Overview of Contributing Authors

This thesis presents an original contribution towards the field of fluorescence-based nucleic

acid diagnostics. However, it is important to note that multi-authored publications have been

incorporated into the work presented here, and the contributions put forth by the colleagues I

have had the fortunate opportunity to work with as well as my contributions are outlined

below.

Chapter 1 provides the reader with an introductory review of background material which

places the research presented into context. A portion of the material from this section can be

found within two published book chapters and one review article. These are multi-author

publications, and the material extracted from these works that is presented here is content from

sections that I was responsible for writing. However, I would like to acknowledge Dr. Paul

Piunno and Dr. W. Russ Algar for their contributions in terms of fine-tuning the content and

editing.

Chapters 2 and 3 are based on FRET methods of analysis for nucleic acid hybridization

detection. Both chapters are joint first authored publications by my self and Dr. Algar (see

List of Publications). Chapter 2 involves the analysis of several FRET pairs in solution and a

focus on one particular FRET pair was used to study selectivity both in solution and at an

interface. The solution phase selectivity data was completed by my self, as well as the surface

chemistry required for analysis of selectivity at an interface. Together, Dr. Algar and I

collected and analyzed the FRET experimental data to generate the list of FRET parameters

for several dye/quencher FRET pairs. Chapter 3 involves FRET using the intercalating dye

thiazole orange. I was responsible for the synthesis of the thiazole intercalating dye moieties,

and together Dr. Algar and I designed, completed and analyzed the experiments and

experimental data collected.

Chapter 4 is based on the development of a fluorescent molecular switch for monitoring DNA

hybridization. The majority of the work presented in this chapter has been published in an

article for which I am listed as first author. Additional unpublished data was also collected and

is presented. I was solely responsible for the design and completion of all experiments and the

vii

analysis of the experimental data. I was also responsible for the dye syntheses required,

specifically the synthesis of the biotin-tethered fluorescent intercalating dyes.

Chapter 5 is an extension of work presented in Chapter 4 and addresses the issues of

selectivity. I was solely responsible for all experimental design, completion, and subsequent

data analysis. I also performed all PCR amplifications and purifications required to generate

the PCR amplicons required for the experiments performed.

viii

Table of Contents

Acknowledgements ........................................................................................................................ iv 

Overview of Contributing Authors ................................................................................................ vi 

Table of Contents ......................................................................................................................... viii 

List of Publications ...................................................................................................................... xiii 

List of Tables ............................................................................................................................... xiv 

List of Figures .............................................................................................................................. xvi 

List of Appendices ........................................................................................................................ xx 

1  Introduction ................................................................................................................................ 1 

1.1  Fluorescence as an Analytical Tool .................................................................................... 1 

1.1.1  Theoretical Description of Fluorescence ................................................................ 2 

1.1.2  Excited State Processes and Fluorescence Emission .............................................. 4 

1.1.3  Measurement of Fluorescence ................................................................................ 5 

1.1.4  Fluorescence Resonance Energy Transfer (FRET) ................................................. 6 

1.1.5  Measurement of FRET ............................................................................................ 7 

1.1.6  Time Resolved Fluorescence Measurements (fluorescence lifetime) .................... 9 

1.2  Nucleic Acids as an Analytical Target .............................................................................. 10 

1.2.1  Nucleic Acid Properties and Structure .................................................................. 10 

1.3  Immobilization of Nucleic Acids Using Biotin/Avidin .................................................... 14 

1.3.1  General Characteristics of Avidin, Avidin Derivatives, and Biotin ..................... 15 

1.3.2  Surface Immobilization of (Strept-, Neutr-) Avidin ............................................. 19 

1.4  Fluorescent Dyes and Molecules that Associate with DNA ............................................. 21 

1.4.1  Fluorescent Labelling/Staining of Nucleic Acids ................................................. 22 

1.4.2  General Effects of Molecular Structure on Fluorescence ..................................... 22 

ix

1.4.2.1  Heterocyclic compounds ........................................................................ 23 

1.4.3  Dyes that Associate with Double-stranded DNA (dsDNA) ................................. 24 

1.4.3.1  Intercalation ............................................................................................ 24 

1.4.3.2  Representative Dyes that Bind Through Intercalation ........................... 25 

1.4.3.2.1  Thiazole Orange ................................................................................. 27 

1.5  DNA and Molecular Switching Techniques ..................................................................... 30 

1.5.1  Molecular Beacons and Hairpin Probes ................................................................ 31 

1.5.2  Other Molecular Switching Constructs ................................................................. 34 

1.6  Fiber Optic Sensing Modalities ........................................................................................ 38 

1.7  Oligonucleotides as Targets for Interfacial Hybridization ................................................ 40 

1.7.1  The Influence of Fluorescent Labels on Duplex Stability .................................... 43 

1.7.2  Structural Factors Affecting Hybridization at an Interface ................................... 45 

1.8  Contributions of This Thesis ............................................................................................. 46 

1.8.1  Scope ..................................................................................................................... 46 

1.8.2  Overview of Contributions ................................................................................... 47 

2  Fluorescence Resonance Energy Transfer (FRET) for DNA Biosensors: FRET Pairs and Förster Distances for Various Dye–DNA Conjugates ............................................................. 51 

2.1  Introduction ....................................................................................................................... 52 

2.2  Experimental Methods ...................................................................................................... 54 

2.2.1  Reagents ................................................................................................................ 54 

2.2.2  Instruments ............................................................................................................ 55 

2.2.3  Dye labeled oligonucleotides ................................................................................ 55 

2.2.4  Oligonucleotide immobilization ........................................................................... 55 

2.2.5  Spectral data and Förster distance calculations .................................................... 56 

2.2.6  Selectivity using the TAMRA–IabRQ FRET pair in solution .............................. 58 

2.2.7  Selectivity using the TAMRA–IabRQ FRET pair on glass substrates ................. 58 

x

2.3  Results and Discussion ..................................................................................................... 58 

2.3.1  Förster distances in single and double stranded DNA .......................................... 58 

2.3.2  TAMRA/IabRQ selectivity studies in solution ..................................................... 62 

2.3.3  TAMRA/IABRQ selectivity studies on glass slide substrates .............................. 65 

2.4  Conclusions ....................................................................................................................... 69 

3  Fluorescence Resonance Energy Transfer and Complex Formation Between Thiazole Orange and Various Dye-DNA Conjugates: Implications in Signaling Nucleic Acid Hybridization ............................................................................................................................ 71 

3.1  Introduction ....................................................................................................................... 72 

3.2  Experimental Methods ...................................................................................................... 74 

3.2.1  Reagents ................................................................................................................ 74 

3.2.2  Dye labelled oligonucleotides ............................................................................... 76 

3.2.3  Association studies ................................................................................................ 77 

3.2.4  Instrumentation ..................................................................................................... 78 

3.2.5  Spectral data, lifetime, and Förster distance calculations. .................................... 78 

3.2.6  Octanol–water partition experiments .................................................................... 78 

3.3  Results and Discussion ..................................................................................................... 79 

3.3.1  Energy transfer between thiazole orange and an extrinsic dye label .................... 79 

3.3.2  Potential FRET-based transduction strategies for DNA hybridization ................. 82 

3.3.3  Non-fluorescent complex formation between modified TO and TAMRA ........... 86 

3.4  Conclusion ........................................................................................................................ 91 

4  Towards a Fluorescent Molecular Switch for Nucleic Acid Biosensing ................................. 92 

4.1  Introduction ....................................................................................................................... 93 

4.2  Experimental ..................................................................................................................... 95 

4.2.1  Synthesis of two biotinylated thiazole orange derivatives .................................... 96 

4.2.1.1  Synthesis of TO-bio2 .............................................................................. 98 

4.2.2  Preparation of optical fibers ................................................................................ 100 

xi

4.2.3  Switch assembly in solution ................................................................................ 100 

4.2.4  Switch immobilization and activation ................................................................ 101 

4.2.5  Interfacial hybridization experiments ................................................................. 101 

4.3  Results and Discussion ................................................................................................... 103 

4.4  Conclusions ..................................................................................................................... 119 

5  Isothermal Hybridization Selectivity Enhancement Using Formamide as a Chemical Agent to Influence Duplex Stability ................................................................................................. 121 

5.1  Introduction ..................................................................................................................... 122 

5.2  Experimental Methods .................................................................................................... 125 

5.2.1  Reagents .............................................................................................................. 125 

5.2.2  Instrumentation ................................................................................................... 126 

5.2.3  Use of formamide to control selectivity .............................................................. 126 

5.2.4  PCR amplification of β-actin gene fragments ..................................................... 128 

5.2.5  Selectivity experiments using PCR amplicons ................................................... 128 

5.3  Results and Discussion ................................................................................................... 129 

5.4  Conclusions ..................................................................................................................... 146 

6  Conclusions and Future Work ................................................................................................ 147 

6.1  Future Outlook ................................................................................................................ 147 

6.1.1  Modification of Thiazole Orange ........................................................................ 147 

6.1.2  Investigation of Synthetic Oligonucleotide Analogues and Intracellular Diagnostics .......................................................................................................... 148 

6.1.3  SNP Selectivity Capability at an Interface .......................................................... 148 

6.2  Summary and Conclusions ............................................................................................. 149 

6.2.1  Analysis of FRET Pairs for Monitoring DNA Hybridization ............................. 149 

6.2.2  Development of a Molecular Switch Construct for Monitoring Nucleic Acid Hybridizaton ....................................................................................................... 150 

6.2.3  Isothermal Selectivity and Signal Enhancement Studies using Thiazole Orange and Formamide ................................................................................................... 152 

xii

Appendix I .................................................................................................................................. 153 

Appendix II ................................................................................................................................. 163 

References ................................................................................................................................... 165

Copyright Acknowledgements ..................................................................................................180

xiii

List of Publications

Refereed

1. M. Massey, U.J. Krull. “Towards a Fluorescent Molecular Switch for Nucleic Acid Biosensing” Analytical and Bioanalytical Chemistry 98 (4) (2010) 1605-1614.

2. W.R. Algar, M. Massey, U.J. Krull. “The Application of Quantum Dots, Gold Nanoparticles and Molecular Switches to Nucleic Acid Diagnostics” Trends in Analytical Chemistry 28 (3) (2009) 292-306.

3. W.R. Algar, M. Massey, U.J. Krull. “LabVIEW in an Advanced Undergraduate Analytical Chemistry Laboratory: Assembly of a Modular Fluorimeter and Associated Software” Journal of Chemical Education 86 (1) (2009) 68-71.

4. W.R. Algar, M. Massey, U.J. Krull. “Fluorescence Resonance Energy Transfer and Complex Formation Between Thiazole Orange and Various Dye-DNA Conjugates: Implications in Signaling Nucleic Acid Hybridization” Journal of Fluorescence 16 (4) (2006) 555-567.

5. M. Massey, W.R. Algar, U.J. Krull. “Fluorescence resonance energy transfer (FRET) for DNA Biosensors: FRET pairs and Förster distances for various dye-DNA conjugates” Analytica Chimica Acta 568 (2006) 181-189.

Book Chapters

6. W.R. Algar, Y. Lim, M. Massey, A.K.Y. Wong, Y. Ye, U.J. Krull. “Assembly of Oligonucleotide Probes on Surfaces for Development of Biosensors and Biochips” in Soft Nanomaterials,Volume 2, H.S. Nalwa, Ed., American Scientific Publishers, California (2009).

7. X. Wang, M. Massey, P.A.E. Piunno, U.J. Krull. “Optical Methods of Single Molecule Detection and Applications in Biosensors” in Smart Biosensor Technology, G. Knopf and A. Bassi, Eds., Taylor and Francis Group LLC, New York (2007).

8. M. Massey, P.A.E. Piunno, U.J. Krull. “Challenges in the Design of Optical DNA Biosensors” in Frontiers in Chemical Sensors: Novel Principles and Techniques, G. Orellana, M.C. Moreno-Bondi, Eds., Springer Series on Chemical Sensors and Biosensors, v.3, O.S. Wolfbeis, Series Ed., Springer-Verlag, Berlin Heidelberg (2005).

xiv

List of Tables Chapter 1

Table 1. Avidin and Avidin analogues that can be used as building blocks for

engineering functional surface structures.

18

Table 2. Selected applications of TO for bioanalysis. 29

Table 3. Recent developments in molecular beacon and hairpin probe technology. 32

Chapter 2

Table 4. Oligonucleotide sequences and labels used in FRET experiments. 56

Table 5. Fluorescence and energy transfer data calculated from experimental data

for various donor-acceptor pairs.

59

Chapter 3

Table 6. Oligonucleotide sequences and labels used in FRET experiments. 77

Table 7. Fluorescence and energy transfer data calculated from experimental data

for various TO-acceptor pairs.

80

Table 8. Fluorescence lifetimes measured for TO intercalated in unlabelled, IabFQ,

IabRQ, and Cy5-labelled dsDNA conjugates.

80

Table 9. Energy transfer efficiencies and TO-acceptor distances deduced from TO

fluorescence quenching efficiency changes in TO fluorescence lifetime.

81

xv

Chapter 4

Table 10. Comparison of fluorescence parameters associated with literature values

for Neutravidin in the presence and absence of biotin.

108

Table 11.

Chapter 5

Time-resolved fluorescence data for TO-bio1 and TO-bio2.

113

Table 12. Oligonucleotide sequences used in selectivity experiments. 127

Table 13. Fluorescence lifetime changes of TO-bio2 associated with dsDNA in the

presence and absence of formamide.

137

xvi

List of Figures Chapter 1

Figure 1. Perrin-Jablonski diagram illustrating the molecular and photophysical

processes that occur between absorption and emission of a photon.

3

Figure 2. Structure of DNA. 13

Figure 3. X-ray crystallography image of Avidin illustrating the tetrameric subunit

conformation.

16

Figure 4. Structure of ethidium bromide. 25

Figure 5. Representative structures of the intercalative class and external label

class of cyanine dyes.

26

Figure 6. Structures of thiazole orange and oxazole yellow. 27

Figure 7. Working principles of selected molecular switching applications for

label-free nucleic acid diagnostics.

35

Figure 8. Design of the label-free fluorescent molecular switch hybridization

monitoring construct and schematic representation of the proposed

hybridization strategy.

48

Chapter 2

Figure 9. Fluorescence emission and absorption profiles for ssDNA and dsDNA

conjugates of Cy3 and TAMRA.

60

Figure 10. Melt curve for a Cy3 dsDNA conjugate. 61

Figure 11. TAMRA fluorescence intensity increase as a function of the degree of

base pair mismatch at 24.5 °C and 60 °C in oligonucleotide duplexes.

64

xvii

Figure 12. Fluorescence scans of surface immobilized TAMRA labelled ssDNA

and dsDNA.

66

Figure 13. Surface TAMRA/IabFQ FRET analysis as a function of the number of

base-pair mismatches.

67

Figure 14. Fluorescent signal regeneration as a function of the degree of base pair

mismatching for TAMRA/IabRQ FRET pair at an interface.

69

Chapter 3

Figure 15. Synthetic scheme for the preparation of thiazole orange dye and the

attachment of various side-chains.

75

Figure 16. Fluorescence spectra showing quenching of TO fluorescence by Cy3 and

TAMRA acceptors.

79

Figure 17. Peak fluorescence intensities of TO as a function of environment. 83

Figure 18. Anticipated signal-to-noise ratios for various proposed probe/target

hybrid detection schemes based on experimental data.

84

Figure 19. Absorption and fluorescence spectra for TAMRA labelled ssDNA in the

absence and presence of TO as a function of side-chain identity showing

complex formation as a function of hydrophobicity of TO side-chain.

87

Figure 20. Quenching of TAMRA-ssDNA fluorescence as a function of the relative

concentration of TO-alkyl.

88

Figure 21. Quenching of TAMRA-dsDNA fluorescence as a function of the relative

concentration of TO-alkyl.

89

xviii

Chapter 4

Figure 22. Design of the label-free fluorescent molecular switch hybridization

monitoring construct and schematic representation of the proposed

hybridization strategy.

94

Figure 23. Synthesis of TO-bio1 biotinylated thiazole orange intercalating dye

containing an 18-atom tether.

97

Figure 24. Synthesis of TO-bio2 biotinylated thiazole orange intercalating dye

containing a 26-atom tether.

99

Figure 25. Schematic of site-directed templating method for assembly and

immobilization of molecular switches on optical fibers and

instrumentation used for interrogation of optical fibers via total internal

reflection.

102

Figure 26. Fluorescence emission spectrum of (a) biotinylated thiazole orange and

thiazole orange containing a three PEG-unit chain in the presence of

dsDNA and (b) biotinylated thiazole orange with DNA and Neutravidin.

105

Figure 27. Fluorescence emission spectra of thiazole orange (free dye) and

biotinylated thiazole orange in the presence of Avidin and Neutravidin.

106

Figure 28. Fluorescence emission spectra of tryptophan residues associated with

Neutravidin in the absence and presence of biotinylated thiazole orange.

107

Figure 29. Fluorescence emission spectra of TO-bio2 as a free dye and in ssDNA

and dsDNA environments.

110

Figure 30. Fluorescence spectra of TO-bio2 as a function of NaCl concentration. 111

Figure 31. Surface functionalization of fused silica optical fibers. 114

Figure 32. Selective, pH-dependent capture of Neutravidin and on biotin and

iminobiotin-functionalized fused silica optical fibers.

115

xix

Figure 33. Fluorescence data from optical fibers coated with molecular switch. 117

Figure 34. Plot showing fluorescence intensity of TO-bio2 at the interface

following regeneration and subsequent hybridization.

119

Chapter 5

Figure 35. Schematic of proposed representation of TO-bio2 binding motifs for

both ssDNA and dsDNA and associated changes upon addition of

formamide.

124

Figure 36. TO-bio2 in the presence of non-labelled, fully complementary and 2bpm

dsDNA as a function of % formamide added to solution.

130

Figure 37. TO-bio2 fluorescence as a function of target oligonucleotide

complementarity.

132

Figure 38. Fluorescence emission intensity of TO-bio2 with ssDNA, ssDNA and

formamide, and free dye (no DNA present).

133

Figure 39. Graph showing room temperature fully complementary, 1bpm and 3bpm

discrimination of a 19-mer target using TO-bio2 in combination with

formamide.

134

Figure 40. Fluorescence intensity of TO-bio2 as a function of time after addition of

formamide.

135

Figure 41. Graph showing variable length, room temperature SNP discrimination

using TO-bio2 and formamide.

143

Figure 42. Graphs showing SNP discrimination of 141-base pair PCR amplicon

targets and affects of probe/target ratio on signal-to-noise.

145

xx

List of Appendices Appendix I

Appendix II

Synthetic Procedures and Characterization

Fluorescence Lifetime Supplementary Information

153

163

1

1 Introduction

The work presented in this thesis is an original contribution towards the field of fluorescence-

based nucleic acid diagnostics. It represents a combination of several projects that were explored

as methods for detection of nucleic acid hybridization at an interface and methods for

investigation of fluorescence properties of labeling dyes in the presence of nucleic acids in

solution for applications in nucleic acid analysis. The outcome is a collection of FRET-based

methodology and direct excitation fluorophore signaling constructs which used both steady-state

and time-resolved fluorescence modalities for interrogation. The emphasis of work has been

directed to investigation of the properties of signaling dyes both in solution and at a solid

interface. The development of a molecular switching construct which utilizes a novel thiazole

orange derivative containing a biotin-terminated tether is perhaps one of the more significant

contributions of the work. This represents, to the best of the authors’ knowledge, the first account

of synthesis of a biotinylated intercalating dye with subsequent application as a means to detect

hybridization. Solution-based selectivity analysis using formamide with this dye derivative has

also given new insight into a useful approach for lowering background signals when using this

detection strategy. The outcome is significant given that the intercalating dye can be used to

monitor nucleic acid binding, and the implementation of formamide concurrently provides a

means to reduce adsorption of dye and enhance selectivity of hybridization. The overall end goal

of the work was to explore new nucleic acid binding constructs that exhibit simplicity in both

design and functionality. Chapter 1 provides a background review to place the work done into

context within the field of fluorescence-based nucleic acid diagnostic tools.

1.1 Fluorescence as an Analytical Tool Bioanalytical sensing platforms using fluorescence transduction mechanisms can provide simple,

rapid, sensitive and selective analysis of biorecognition events.1,2 One of the fundamental

requirements of optical transduction techniques is the generation of a photoluminescent signal

indicative of a binding event. Luminescence is defined as emission of ultraviolet, visible or

infrared photons from an electronically excited species.2 Analytical techniques based on

2

transduction of a fluorescence emission signal can be designed to detect selective reactions or

association of molecular components. Fluorescence methods of analysis are particularly

advantageous when one considers a few key design criteria. The variety of fluorescent labels

available and the wide wavelength range of the electromagnetic spectrum that they cover imparts

selectivity based on the judicious choice of one label or a suite of labels that can work together to

create either a multiplexed system or a system based on FRET. These fluorescent labels typically

involve organic dyes, fluorescent nucleobase analogues, fluorescent polymers, and nanomaterials

such as quantum dots or gold nanoparticles.1 Fluorescence emission is a signal that is capable of

providing for immediate response (real-time measurement), low detection limits, and the

potential for portability of instrumentation. Perhaps one of the most significant advantages of

fluorescence methods of analysis is the variety of detection modalities. These include steady

state, time resolved, fluorescence imaging, single molecule spectroscopy, anisotropy

(polarization), microscopy, and the ability to measure in the solution phase and solid phase. A

brief overview of theoretical concepts and measurement techniques for both steady-state and

time-resolved fluorescence is presented in the following section.

1.1.1 Theoretical Description of Fluorescence

Fluorescence and phosphorescence are luminescence phenomena that arise from the initial

absorption of a photon of specific energy (hυ) by an absorbing atom, molecule or lattice to create

an electronically excited state. Luminescence that is a result of initial absorption of energy in the

form of photons is referred to as photoluminescence. Upon relaxation from an electronically

excited state to the ground state, a photon is emitted. In the specific case of fluorescence, the

photon emission occurs from an excited singlet state where the electron in the excited state and

its associated ground state electron are spin paired (spins are of opposite sign). This spin state

allows rapid relaxation of the excited state electron to a ground state molecular orbital by

emission of a photon, which can be detected as a fluorescence signal. There are many other

pathways for an excited state species to return to the ground state. These include internal

conversion and intersystem crossing (leading to phosphorescence or delayed fluorescence),

intramolecular charge transfer and conformational changes as well as a variety of processes

where excited state interactions with other molecules can lead to electron transfer, proton

transfer, energy transfer, excimer or exciplex formation or photochemical transformation.2 The

3

mechanism of fluorescence including the initial absorption of a photon and the processes

occurring in the excited state and emission are shown using a Jablonski diagram as in Figure 1.

Figure 1. Perrin-Jablonski diagram illustrating the molecular photophysical processes that occur between absorption and emission of a photon. The ground and excited singlet electronic states are denoted by S0,S1 and S2, and the first excited triplet state is shown as T1. The Morse potential energy wells for the ground and excited singlet states are shown and contain the vibronic transitions (νn) for excitation and emission. (a) The blue solid and dashed lines represent absorption from (i) S0ν0 and (ii) S0ν1 to different vibrational states of S1 based on the probability of the transition (b) Representation of a nonradiative relaxation pathway: (i) excitation (violet line) to S2ν3 followed by vibrational relaxation (vertical zigzagged line) to S2ν0 and (ii) internal conversion (horizontal line) to S1ν5 followed by vibrational relaxation to S1ν0 and (iii) internal conversion to S0v6 followed by non-radiative vibrational relaxation to the ground state. (c) Radiative relaxation pathway: (i) excitation to S1ν2, followed vibrational relaxation to S1ν0, and (ii) emission of a photon in the form of fluorescence to different vibrational states of S0 based on relative probabilities of transition (d) Another possible nonradiative relaxation pathway: (i) excitation to S1ν2 followed vibrational relaxation to S1ν0, and intersystem crossing (horizontal line) to T1ν1. Following vibrational relaxation to T1ν0, to molecule generally relaxes via non-radiative processes due to the inefficiency of phosphorescence and delayed fluorescence. (Permission to reproduce figure and figure text kindly provided by Dr. Russ Algar).3

Each electronic state can be represented by a potential energy well onto which several vibrational

energy states are superimposed. Within each vibrational energy state a probability distribution is

shown to illustrate the overlapping of states required for absorption. The ground, first, and

second electronic states are denoted as S0, S1 and S2, respectively. Within S0, S1, and S2, the

4

vibrational states are denoted as ν0, ν1, ν2....νn, respectively. Rotational states exist as well, but

are not considered here to maintain overall simplicity.

Absorption of light by a molecule leads to an electronic transition where an electron is promoted

typically from the highest occupied molecular orbital (HOMO) to an unoccupied molecular

orbital of higher energy. In order for this to occur, two criteria must be met: the energy of the

incident photon must equal the energy difference between the ground and excited electronic state

(resonance condition), and the electric field vector of the incident photon must be aligned with

the electronic transition dipole of the molecule (photoselection principle). In polyatomic

molecules, these transitions occur from a bonding to an antibonding orbital on the time-scale of

~10-15 s. In a qualitative sense, the transition occurs from the lowest vibrational state of the

ground state to a vibrational state in the excited state whose vibrational wave function it most

closely resembles.2 The efficiency of absorption at a particular wavelength can be defined using

the Beer-Lambert Law:

log   (1)

where A(λ) is absorbance, and are the incident and transmitted light intensities, respectively,

b is the path length (in cm), c is concentration of absorption species (in mol/L) and ε(λ) is the

molar absorption coefficient (in L mol-1 cm-1). The molar absorption coefficient represents the

probability of photon absorption at a given wavelength. Once a photon is absorbed and an

excited electronic state is achieved, several processes occur which allow the molecule to return

to the ground state. These transitions can be both radiative and non-radiative and the relaxation

process is usually some combination of the two.

1.1.2 Excited State Processes and Fluorescence Emission

The relaxation of a molecule from an excited vibrational state (S1νn) to the lowest energy excited

vibrational state (S1ν0) occurs on a much faster time scale than the radiative relaxation of the

molecule from the electronically excited state to the ground state; hence, fluorescence occurs

from a thermally equilibrated state.2 Vibrational relaxation leads to a transfer of energy as heat to

the surrounding solvent via molecular collisions (sometimes referred to as collisional

deactivation). This non-radiative relaxation process has an approximate time-scale of ~10-12 s.

Internal conversion is another non-radiative mode of relaxation that occurs between two

5

electronic states of the same spin multiplicity. A combination of internal conversion (if a

molecule is promoted to an electronic state higher than S1) and vibrational relaxation results in

the molecule relaxing to the lowest vibrational level of S1. The efficiency of these non-radiative

pathways is the basis for Kasha’s rule which states that fluorescence is almost always observed

from the ground vibrational state of the first excited singlet state (S1ν0). Internal conversion from

S1 to S0 is possible, but is less efficient than internal conversions between higher energy states.

As a result, fluorescence can then compete as a relaxation pathway to the ground state.

Fluorescence occurs if the rate of fluorescence is comparable to the rate of internal conversion

between the S1 and S0 states, and as a consequence, only a relatively small number of molecules

exhibit significant fluorescence in condensed phases. Because energy is lost through these

efficient non-radiative processes, there is a shift to longer wavelengths for fluorescence emission

compared to the wavelength required for absorption. This is referred to as the Stokes shift. Even

though the time scale for emission is approximately 10-15 s, excited molecules can occupy the S1

state on a time-scale that can range from picoseconds to hundreds of nanoseconds before photon

emission.2 This lifetime can be measured and provide important information about the local

environment of the fluorophore.

1.1.3 Measurement of Fluorescence

Fluorescence is usually a result of relaxation from the first excited singlet state to the ground

electronic state via emission of a photon. Fluorescence intensity is defined by an extension of

the Beer-Lambert law:

Φ (1-10εbc) (2)

where F is fluorescence, Φ is quantum yield, K is the instrument response coefficient, I0 is the

intensity of the incident radiation, ε is the molar absorption coefficient, b is the path length, and c

is concentration of the fluorescent species. If the concentration is sufficiently low (<0.01M), then

the equation can be approximated as:

2.303 Φ (3)

When considering the use of fluorophores as an analytical tool, the quantum yield of

fluorescence must be taken into account as it directly relates to the intensity of the fluorescence

6

signal generated from the population of fluorophores under interrogation. The quantum yield of

fluorescence is defined by:

             Φ       Σ      …  (4)

where kr represents the decay lifetime (rate) of the radiative processes and knr represents the

decay lifetime of the non-radiative processes including internal conversion ( ), intersystem

crossing ( ), quenching ( ), energy transfer ( ) and any other process that may lead to non-

radiative mechanisms of relaxation, τs represents the excited state lifetime, or fluorescence decay

time of the singlet state, and is the rate constant for radiative deactivation (S1 → S0) with

emission of a photon.

Molecular fluorescence can be used as an analytical tool due to its high sensitivity to local

environment. Factors including, but not limited to, solvent polarity, hydrogen bonding, pH,

pressure, viscosity, temperature, fluorescence quenching, electric potential, and ions in solution

can all affect fluorescence. Structural components of fluorescent molecules also affect

fluorescence, including: degree of conjugation, substituent effects (both electron donating and

electron withdrawing), and certain types of heterocyclic compounds and their characteristics.

These factors can affect molar absorptivity, quantum yield, and shift excitation and emission

spectra to different wavelengths. As a result, fluorescence can be used to monitor changes in

these parameters or can be used to probe biochemical/bioanalytical processes which render

changes in these parameters within the local environment of a fluorophore.

1.1.4 Fluorescence Resonance Energy Transfer (FRET)

FRET is a nonradiative transfer of energy from an excited donor (D) to an acceptor (A) through

long-range resonance coupling between the relaxation transition dipole of the donor and

excitation transition dipole of the acceptor.4 Energy transfer between two different molecules

can be represented by:

D* + A → D + A*

Since FRET is a resonant interaction, fluorescence intensity of the donor is reduced and the

acceptor (if it is a fluorophore) undergoes FRET sensitized emission. If the acceptor is a dark-

7

absorbing species, the donor species is effectively quenched without the observation of FRET-

sensitized emission from the acceptor. Several combinations of donor-acceptor pairs have been

utilized with DNA, including: two fluorophores5,6, a fluorophore with a dark absorber6-8, a

fluorophore with an intercalator9,10, a fluorophore with a gold nanoparticle11, quantum dot donors

with fluorophore acceptors12-14, and intercalating donors with single-walled carbon nanotube

(SWCT) acceptors15. FRET is strongly distance dependent on a length scale comparable to most

biological macromolecules. As a result, FRET is widely used to determine distances in

biomolecular assemblies and associations and is commonly given the term ‘molecular ruler’. The

efficiency of energy transfer allows one to measure molecular distances between 2-10 nm. In

many analytical applications, the enhancement or decrease in fluorescence intensity of

donor/acceptor moieties can also provide the basis for rapid analytical analyses.

1.1.5 Measurement of FRET

The extent of energy transfer is dependent on several factors. Most notably, FRET is dependent

upon the distance between the donor and acceptor as well as the orientation of the donor

emission transition dipole with respect to the acceptor absorption transition dipole. The emission

spectrum of the donor must also overlap with the absorption spectrum of the acceptor such that

the energy differences between the ground and excited electronic states of the acceptor are equal

to or less than that of the donor. This allows coupling between transitions of equal energy to

occur. This overlap ensures that there are a sufficient number of vibronic transitions which are

equal in energy in both the donor and the acceptor species such that the transitions are in

resonance.2 The energy transfer between the donor and the acceptor is due to several interactions

between the two species. The interactions can be either due to short-range inter- or

intramolecular overlap or long-range Coulombic interactions. Long-range Coulombic

interactions are long-range dipole-dipole interactions, and are also referred to as the Förster

mechanism of interaction.2 This Coulombic term refers to the energy transfer process where the

excited electron occupying the LUMO of the donor relaxes and an electron in the HOMO of the

acceptor is excited to the LUMO of the acceptor simultaneously.2 In this case, long range

interaction occurs over distances up to 8-10 nm (in comparison to short-range interactions which

are a few tens of Angstroms). Förster developed a series of expressions which define this long-

range dipole-dipole transfer of energy, and as a consequence, FRET is sometimes referred to as

8

Förster Resonance Energy Transfer. The dipole-dipole energy transfer rate constant can be

defined as:

                                                                          (5)

where τD is the excited state lifetime of the donor in the absence of energy transfer (absence of

the acceptor), r is the separation distance between the donor and the acceptor, Ro is the Förster

distance, where the transfer efficiency is 50% and energy transfer to the acceptor and decay of

the excited donor are both equally probable.2 The transfer efficiency (E) can be written in terms

of the Förster distance and the separation distance:

  

1   1   1   (6)

Equation 6 also shows how E can be calculated using the fluorescence (F), quantum yield (Φ) or

lifetime (τ) ratios of donor/acceptor (subscripts DA): donor (subscripts D). This expression also

shows that there is a 1/r6 distance dependence for FRET. The Förster distance can be defined

using the following equation:

8.79   10        Φ   (7)

where n is the refractive index of the medium in the wavelength region of spectral overlap, κ2 is

the orientation factor of the transition dipole moments of the donor and the acceptor, ΦD is the

fluorescence quantum yield of the donor in the absence of the acceptor, and J represents the

spectral overlap integral which is defined below:

    (8)

where FD (λ) is the donor fluorescence in the absence of acceptor as a function of wavelength,

and εA is the acceptor molar absorption coefficient at wavelength λ.

The orientation factor (κ2) in Equation 7 can be approximated to a value of 2/3 when the donor

and acceptor are free to sample all orientations. However, the values for κ2 can range from 0 to 4

9

where zero represents a perpendicular transition moment orientation between donor and acceptor

and four represents parallel transition moments between the two species.

1.1.6 Time-Resolved Fluorescence Measurements (fluorescence lifetime)

Fluorescence measurements can provide information about structural changes of fluorescent

biomolecular complexes. Time-resolved methods are used extensively in studies of biological

systems and cellular imaging for bioanalytical analyses16-18, and can arguably provide more

information than steady-state measurements. Fluorescence lifetime refers to an ensemble average

time a fluorophore spends in the excited state following excitation, and corresponds to the time at

which 1/e or 37% of the population of excited fluorophores remains. Time-resolved fluorescence

measurements can give insight into the underlying fluorescence dynamics through resolution of

the lifetimes of the fluorescence intensity decay profiles.19 This can provide information about

populations of the same fluorophore that are experiencing different local environments that a

steady-state emission spectrum would not be able to differentiate. Fluorescence lifetime is

independent of fluorophore concentration, and this is particularly useful when an accurate

measure of the concentration of a species is not known.

Fluorescence lifetime can be measured in both the time domain and frequency domain, and

experiments are done using pulse and phase-modulation fluorimetry, respectively. Fluorescence

lifetime follows an exponential model, and the time-dependent intensity I(t) can be described

using the following equation:

      (9)

where Io is the fluorescence intensity at time zero, t is time and τ is the fluorescence lifetime. If a

population of fluorophores is excited by a short pulse of light, and the intensity is measured as a

function of time, it then follows that a plot of ln I(t) vs t yields a slope of -1/τ for a

monoexponential decay process. The decay in fluorescence intensity is recorded at a time after

the initial pulse is delivered, and deconvolution and curve fitting analyses are used to extract out

the fluorescence lifetime parameters of a sample. For the simple case of a mono-exponential

decay process, Equation 9 can also be written as:

10

  (10)

where the variable α is defined as a pre-exponential factor. In samples which contain

fluorophores that are capable of experiencing different environments, a distribution of decay

times corresponding to different populations of fluorophores in the same sample can be present.

In these cases, a multi-exponential decay model can be used to describe the system. The

contribution of each component defined through the weighting associated with its pre-

exponential factor then becomes an important analytical parameter in defining an overall model

of the system. The multi-exponential decay can be represented as:

 ∑ (11)

where αi is the pre-exponential factor associated with each component i of the model. In this

way, the relative contributions of each component to an amplitude averaged lifetime can be

determined. This can be particularly useful when looking at fluorescent dyes that associate with

DNA as a method for determination of the presence of different binding modes. This can also be

useful for systems that exhibit a significant degree of undesired background fluorescence as

time-gated acquisition of fluorescence signals can be used to suppress short-lived background

fluorescence.

As previously mentioned, there is a wide scope of fluorophores that have been used for the

transduction of nucleic acid hybridization. The following section provides a brief overview of

nucleic acid properties and immobilization strategies relevant to detection of DNA hybridization

and is followed by an overview of fluorescent dyes and their properties relevant to fluorescent

detection methods and association with DNA.

1.2 Nucleic Acids as an Analytical Target

1.2.1 Nucleic Acid Properties and Structure

Nucleic acids are linear biopolymers of nucleotides that contain the key components required for

the mechanism of storage and transfer of genetic information. The genome of an organism is

comprised of deoxyribonucleic acid (DNA). Information encoded in the nucleotide sequence of

DNA is transferred either through replication (creating a duplicate of the DNA molecule) or

11

transcription (DNA sequence is translated to create a complementary messenger ribonucleic acid

(mRNA) molecule where 3-base codons are then read in the mechanism of protein synthesis).

In DNA, each nucleotide monomer is composed of a deoxyribose sugar, a heterocyclic

nitrogenous base, and a phosphate group. Single-stranded nucleic acids are polynucleotides

linked through the phosphate group of each monomer via phosphodiester bridges as shown in

Figure 2. These linkages are formed as 5’-nucleoside monophosphates are added at the 3’-

hydroxyl group of a preceding nucleotide and this gives the sequence a 3’-5’ linkage

directionality. The bases of the nucleotide monomers are composed of either a derivative of a

six-membered heterocyclic aromatic ring containing two nitrogen atoms (pyrimidine) or a fused

ring heterocycle containing a pyrimidine and an imidazole ring (purine). There are two common

naturally occurring pyrimidines that are encountered in nucleic acid structure: (cytosine (C) and

thymine (T)), and two purines: (adenine (A) and guanine (G)). When considering using DNA as

an analytical target or as a building block to create nanoscale constructs, evaluation of relevant

size parameters is important. These include: pitch length, base-pair spacing, number of hydrogen

bonding sites between bases and their associated bond lengths, and the width of double-stranded

DNA (Figure 2).

DNA in its native form exists as a double helix where two complementary polynucleotide chains

associate in an anti-parallel fashion (one strand runs in the 3’-5’ direction while the other runs in

the 5’-3’direction) such that their genetic information is conserved upon replication and thus

creates a repository for genetic information. The ability of DNA to maintain a double helical

structure is a result of several bonding and energetic processes. The structure is dependent on the

strength of interchain hydrogen bonds that exist between the base pairs. Complementary base

pairing occurs between a purine and a pyrimidine such that A pairs with T via two hydrogen

bonding sites and G pairs with C via three hydrogen bonding sites (see Figure 2). The strength

of the hydrogen bonds between the base pairs is approximately 20 kJ/mol which allows the

internuclear distance between non-bonded atoms to be less than their Van der Waals contact

distance providing a strong attractive interaction.20 Hydrogen bonding between water molecules

and the sugar-phosphate backbone also act to stabilize the helical structure. The nucleobases are

also involved in π-stacking and hydrophobic interactions which orient non-polar moieties to the

interior of the double-helix, allowing further stabilization of the helical structure. Charge

screening through electrostatic interaction between counter-ions in the surrounding medium and

12

the polyanionic backbone of the DNA strands minimizes charge repulsion of the two strands

which carry significant charge density and allows the strands to associate. Double-stranded DNA

(dsDNA) can form three major secondary structures denoted as A, B and Z, with B-DNA being

the major conformation DNA adopts in solution. This conformation leads to base-pair spacing of

0.34 nm and a pitch length of 3.4 nm comprised of 10 base pairs. Because the two strands of

dsDNA run anti-parallel to one another, a major and a minor groove exist in the helical structure.

Groove binding molecules can recognize and bind to these grooves and can be used to develop

signalling mechanisms for detection of dsDNA. In nucleic acid sensing platforms, immobilized

oligonucleotides are used as probes to detect complementary targets from a sample solution that

is exposed to the sensing element, and a DNA hybridization event occurs resulting from the

aforementioned association processes. DNA duplexes form, and these duplexes adopt a less

flexible, ‘rigid-rod’ conformation as a result of the hydrogen bonding configuration, in contrast

to the conformational flexibility associated with a single stranded probe oligonucleotide.

Modeling simulations have shown that short oligonucleotides immobilized onto an amine-

functionalized silica substrate can be conformationally described as a “stable hybrid duplex with

B-like character”.21 Monti et al.21 have also shown through detailed modeling simulations of

hybridization (rather than a pre-hybridized duplex) that the hydrogen bonding distances of the

base pairs formed by oligonucleotides immobilized at an interface were similar (average value of

0.29 nm) to those of DNA in solution (see Figure 2).

13

Figure 2. Structure of DNA, including (a) helical dimensions (b) structure and bonding of bases, sugars, and phosphate backbone including base-pairing, hydrogen bond number and length (c) individual nitrogenous bases. The image in (b) is reproduced with permission from reference22 Copyright © 2008 Sinauer Associates, Inc.

14

The publication of the human genome sequence in 2001 and the completion of the human

genome project in 2003 have allowed extensive expansion of the field of research associated

with genomics and subsequent emerging fields of proteomics and metabolomics. The single

nucleotide polymorphism (SNP) working group has determined that 1.42 million SNPs are

distributed throughout the human genome23. Many human diseases result from a single gene

mutation, and the ability to recognize or detect these mutations is central to providing rapid

diagnosis. The inherent ability of single-stranded nucleic acid oligomers to selectively recognize

and bind their complementary targets therefore becomes of great utility for applications

involving DNA detection and SNP analysis. This is the basis for several DNA-based engineering

scaffolds including DNA-based sensing, high-throughput screening, and DNA computing,

electronics and molecular machines.1 In light of these advancements, several synthetic analogues

that overcome some of the limitations or improve binding capabilities of nucleic acid

probe/target hybrids for these particular applications have been created. These include: peptide

nucleic acids (PNA)24-26, locked nucleic acids (LNA)27-30, nucleic acids containing metal ions to

generate metallic-like conduction (M-DNA)31-34, and functional nucleic acids (FNA)35-37 which

include aptamers and nucleic acid enzymes.

The basis of most structural matrices involving nucleic acids immobilized at an interface

involves synthetic, short-chain oligonucleotide sequences. Oligonucleotides are usually defined

as being 2-20 nucleotides in length, although some definitions incorporate strands of nucleotides

up to 100 bases in length.38 The immobilization techniques which utilize the biotin/Avidin

system to place oligonucleotide probes at an interface are outlined in the next section.

1.3 Immobilization of Nucleic Acids Using Biotin/Avidin Immobilized oligonucleotides are the central component to many sensing modalities and have

become increasingly prevalent as structural scaffolds for nanoscale constructs. Many

immobilization strategies have been developed and depend on the type of nucleic acid used (e.g.

DNA, PNA, LNA, etc.), and the solid substrate employed. Immobilization methods normally

fall into five main categories: adsorption, matrix entrapment or cross-linking, covalent

attachment, self-assembled monolayers (SAMS), and affinity binding.39 The focus of this thesis

15

will be affinity binding, and the most common form of affinity binding of oligonucleotides to

solid surfaces is through the use of biotin-avidin linkages.

When considering the construction of functional nanomaterials, a robust chemistry with high

binding affinity is essential to maintain structural integrity. Stable, surface bound components

capable of providing the correct orientation of appropriate biorecognition elements is a key

design consideration.40 The high-affinity interaction of the biotin/avidin system has widespread

bioanalytical applications including biosensors, affinity chromatography, immunoassays, drug

delivery, and affinity-based imaging agents.41,42 Much like the derivitization of nucleic acids to

provide custom-built building blocks that overcome specific analytical challenges, derivatives of

native biotin and avidin have been created to overcome limitations such as charge effects leading

to non-specific adsorption, to provide reversible binding properties where desired, and

fluorescence labelling to provide optical signals for bioanalytical imaging and sensing

applications. The ability to conjugate an oligonucleotide sequence or signalling agent of interest

to a biotin moiety allows both synthetic flexibility and provides a facile method to create

nanostructures of interest.

1.3.1 General Characteristics of Avidin, Avidin Derivatives, and Biotin

Avidin is a tetrameric glycoprotein with a molecular weight of approximately 66,000-68,000 Da.

In nature, it is found in egg whites where it serves as an antibiotic to prevent bacterial

growth.41,43 Avidin and its analogues are capable of binding four biotin molecules for every

avidin molecule (one per protein subunit), which makes it very useful from a nanostructure

templating standpoint (see Figure 3).

16

(a)

(b)

Figure 3. (a) X-ray crystallography ribbon diagram of Avidin illustrating the tetrameric subunit conformation. Binding of a single biotin molecule to each subunit is represented using a ball-and-stick model. (b) Distances from biotin-binding pocket to surface of protein (Å). The image in (a) was accessed from the RCSB Protein Data Bank.44 The distances and image in (b) were created using JMol Viewer45 (available through the RCSB Protein Data Bank) Biotin is rendered in beige, while amino acids associated with the outer surface of the protein are rendered in purple, aqua and red.

Biotin is a small molecule vitamin (vitamin B7 or vitamin H) comprised of a ureido ring fused to

a tetrahydrothiophene ring containing a valeric acid side chain. In nature, it functions as a

coenzyme of carboxylases and is present in all living cells where it acts as a carrier of CO2 and

carboxyl group donor in fatty acid synthesis.46 From an analytical and synthetic standpoint, the

carboxylic acid group present in the biotin structure provides a useful linkage site for attachment

of amine-terminated molecules of interest. In particular, it provides a site for covalent

attachment of amine-terminated oligonucleotides to biotin moieties or biotin moieties to amine-

terminated solid substrates. Biotin derivatives can also be used to fine-tune analytical

applications, and several biotin analogues are commercially available to suit specific synthetic

criteria. These include biotin derivatives that are reactive towards primary amines,

carbohydrates, carboxyl groups, sulfhydryl groups, and photoreactive biotinylation reagents.47

The association between biotin and avidin is one of the strongest known non-covalent

17

ligand/protein interactions with a dissociation constant (Kd) of 1.3 x 10-15 M at pH = 5.0.40 This

renders the bound complex resistant to many denaturation methods including buffer,

temperature, salt and pH extremes. It also shows resistance to certain chaotropic agents,

including guanidine hydrochloride (up to a limit of ~3 M) and urea (~4 M).43,48 The strong

biotin/avidin binding affinity is quite advantageous when considering design criteria for stable,

regenerable surfaces for bioanalytical applications.

Avidin has an isoelectric point of 10.5, which renders the protein positively charged at neutral

pH due to positively charged amino acid residues present on the protein. This can lead to issues

associated with non-specific adsorption, and is particularly problematic with immobilization of

polyanionic oligonucleotide sequences. Furthermore, oligosaccharide (mainly mannose and

glucosamine) components of the Avidin structure can bind to carbohydrate-binding proteins on

cell surfaces which leads to further complications with adsorption.41 To overcome this,

Streptavidin or Neutravidin can be used in template docking strategies for biotinylated molecules

containing significant negative charge density. Table 1 summarizes the important considerations

of Avidin and its analogues for bioanalytical applications.

18

Table 1. Avidin and Avidin analogues that can be used as building blocks for engineering functional surface structures.

Protein M.W. (KDa)

Kd (M)* Dimensions (nm)

pI charge**

advantages limitations

Avidin

isolated from vertebrate egg white

66-68 1.3 x 10-15 at pH 5.0

5.6x5x4

10.5 + hydrophilic, lysine residues provide primary amine sites for bioconjugation

significant charge related non-specific binding

Streptavidin

isolated from bacterial culture of Streptomyces avidinii

53-60 10-13-10-15 4.5x4.5x5.8 5-7.5 n non-glycosylated form of avidin, leads to less non-specific binding

less hydrophilic than avidin, decreased number of lysine residues available for bioconjugation

Neutravidin

commercial protein engineered analog of avidin

60 10-15 approx. proportional to those of native avidin

6.3***

- non-glycosylated form of avidin, reduced number of charge sites on the surface of the protein while leaving amino acid residues in biotin binding pocket the same

CaptavidinTM

avidin derivative with a nitrated tyrosine mutant present in the biotin binding pocket

10-9

at pH 4.0

reversible, pH dependent binding of biotin; dissociation of biotin at pH 10

*Kd with biotin. **Charge at neutral pH. **Charge at neutral pH. ***Can be engineered with pI ranging from 4.7-9.4. 42.

Streptavidin is a biotin-binding protein isolated from the culture of Streptomyces avidinii and is a

deglycosylated form of avidin with reported molecular weights ranging from 53,000-60,000 Da,

reported pI ranging from ~5-7.5, and reported Kd values ranging from 10-13- 10-15 M.41,47,49

Despite the reduced hydrophilicity, decreased number of lysine residues, lower abundance and

higher cost of Streptavidin when compared to avidin, the advantage of decreased non-specific

19

adsorptive effects as a result of reduced charge-sites has led to increased use of Streptavidin in

bioanalytical applications.42 However, the limitations previously mentioned with Streptavidin

led to the synthesis of Neutravidin, a protein-engineered analogue of avidin. Marttila et al.42,

through site-directed mutagenesis and subsequent expression and purification, produced charge

reduced mutants of Avidin with a range of pI values from 4.7-9.4 that exhibit less non-specific

binding and a similar affinity for biotin. By reducing the number of charged residues on the

outer surface of the protein, while leaving the amino acid residues in the biotin-binding pocket

unchanged, the affinity for biotin is retained while reducing the overall charge density of the

protein. Neutravidin, a commercially available form of deglycosylated avidin, has a molecular

weight of approximately 60,000 Da, a Kd value of approximately 10-15 M, and a pI of 6.3 which

makes it particularly useful for bioanalytical immobilization applications.

1.3.2 Surface Immobilization of (Strept-, Neutr-) Avidin

Methods that incorporate immobilization of Neutravidin on a solid substrate to provide a

templating mechanism for multifunctional surface-tethered components are useful because they

provide a means of assembly of multi-component structures onto a surface.50 Biotin and various

forms of avidin can be immobilized onto many types of surfaces including for example: glass,

fused silica optical fibers, silicon wafers, gold, filter membranes and cell membranes.43

Combinations of the many different biotin and Avidin derivatives available allow one to custom-

build a surface chemistry tuned to the specific needs of a bioanalysis. Biotin-avidin methods of

nucleic acid immobilization have been used in fluorescence43,51-53, surface plasmon resonance

(SPR)54,55, piezoelectric56,57 and electrochemical58,59 sensor platforms.

In addition to the ‘tool kit’ of biotin and Avidin derivatives available for construction of

nanoscale scaffolds, the method used to immobilize tethered components is also a key factor to

the overall design of multi-component molecular constructs. The immobilization process can be

broken down into two main considerations; the chemistry associated with attachment of a

biotinylated oligonucleotide to an avidin/streptavidin/neutravidin (ASN) molecule, and the

method of immobilization of ASN to the solid substrate. The immobilization method of the

nucleic acid oligomer portion in these cases is through a covalently bonded biotin moiety to the

oligonucleotide that is then immobilized through affinity binding to a surface functionalized with

ASN. Synthetic methods can be employed to covalently attach one or more biotin moieties to

20

either the 3’ or 5’-terminus of an oligonucleotide as well as internal positions along an

oligonucleotide chain via a biotinylated phosphoramidite.60 A variety of linkers including alkyl-

chain and polyethylene glycol (PEG) chains can be used to vary the distance between the biotin

moiety and the terminal nucleobase. The choice of linker can also impart rigidity or flexibility

depending on the surface design requirements.

There are several methods for immobilization of ASN to solid substrates. These include

physisorption, adsorption, formation of self-assembled monolayers (SAMs) using thiol/gold

interactions, covalent bonding of the protein (through the amino, thiol or carboxylic acid groups

present on the protein) to functionalized surfaces, and affinity binding of ASN to a biotinylated

surface.43 If the application is to immobilize biotinylated oligonucleotides to a surface, these

immobilization methods can be broadly divided into two categories: direct immobilization of

ASN or implementation of an ASN film sandwiched between a biotinylated substrate and a layer

of biotinylated oligonucleotides. Biotinylation of a solid surface followed by attachment of

avidin species through the avidin/biotin affinity interaction creates a surface capable of binding

other biotinylated molecules, and has been reported to provide increased stability and

organization of a streptavidin coated surface film.43 By using a biotinylated surface, the protein

is protected from potential denaturing effects of contact with a hydrophobic surface and ideally

provides a docking mechanism where the protein is strongly bound through two biotinylated

anchors to the surface while providing two distal binding sites facing outwards into

solution.43,50,61-63

The physical phenomenon of binding at an interface inherently invokes a change in the values of

the association constants with respect to solution-phase affinity interactions due to the loss of

degrees of freedom associated with pinning molecular components to a surface. For the specific

cases of biotin/avidin binding, issues of ligand accessibility need to also be addressed.64 One of

the more popular immobilization methods utilized for covalent biotinylation of surfaces for

fluorescence methods of analysis is through aminosilanization of a glass surface (through the use

of aminopropyltrimethoxysilane (APTMS) or aminopropyltriethoxysilane (APTES)) followed by

subsequent covalent linkage of a biotin moiety.53,64 Wayment and Harris64 have specifically

studied the biotin-Neutravidin binding kinetics of an aminosilanized glass surface to which biotin

moieties have been covalently immobilized followed by binding of Neutravidin. Using mixed

silane monolayers comprised of APTES and cyanoethylsilane on glass, a low density of biotin

21

binding sites was created on a glass substrate by reacting the amine functional groups of APTES

with a succinimidyl ester derivative of biotin. Tetramethylrhodamine-labelled Neutravidin

moieties were then bound to the biotin sites to provide a luminescent tag for single-molecule

imaging via total internal reflection fluorescence (TIRF). By observing both the rates of binding

and unbinding, an association constant of Ka = 5.5(+/- 0.2) x 1011 M-1 was calculated for the

biotin/Neutravidin binding at a glass surface.64 However, it should be noted that a range of

values for Ka of biotin/avidin species binding at an interface have been determined, and as

expected the values are highly dependent on factors such as immobilization chemistry, linker

lengths, density of available surface binding sites and kinetic parameters.64,65

Biotin/avidin affinity chemistry to tether biological molecules into an organized structure

capable of monitoring nucleic acid hybridization both in solution and on a surface has been

employed here. Neutravidin was used to provide minimal non-specific adsorption of both the

oligonucleotide probe and to block non-specific adsorption when immobilized at the surface of

an optical fiber. Although the binding affinity of a surface-bound biotin/Neutravidin system is

lower than that of solution phase binding, the affinity is still sufficient to provide anchoring of a

switching construct comprised of multiple surface-tethered components.

1.4 Fluorescent Dyes and Molecules that Associate with DNA Analytical techniques based on fluorescence detection are ubiquitous due to their inherent

sensitivity, selectivity and ability to provide spatial and temporal resolution. However, many

analytes of interest are not fluorescent, and so indirect methods are used to monitor them. These

indirect methods involve the use of fluorophores either through derivitization or covalent

attachment of the fluorophore to the analyte of interest, formation of a fluorescent complex, or

fluorescence quenching processes. Organic dyes are very commonly used as fluorophores for

the monitoring of a biomolecular interaction. Within the umbrella term of ‘organic dyes’, several

subsets of dyes that are utilized to monitor recognition events can be categorized. Some of the

representative dyes and their properties are discussed here as well as effects of molecular

structure on dye fluorescence properties.

22

1.4.1 Fluorescent Labelling/Staining of Nucleic Acids

Fluorescent labelling or association of fluorescent species with nucleic acids can be achieved

through both covalent bonding and non-covalent interactions using a wide variety of labels. A

label can be judiciously chosen based on several criteria including quantum yield, Stokes shift,

spectral emission properties, time-resolved fluorescence properties, multiplexing capability,

susceptibility to photobleaching, ability for functional attachment to other molecules, ability to

discriminate between ssDNA or dsDNA, and the minimization of background fluorescence

signals. Nucleic acids have been labelled with dyes that emit wavelengths ranging from the x-ray

to the IR regions of the electromagnetic spectrum.66-70 However, labels that exhibit fluorescence

in the visible range are the most commonly used fluorophores for analytical applications

involving bioanalysis. In particular, dyes that have the ability to associate with nucleic acids are

advantageous over external labels that show no selectivity towards DNA structures. These dyes

can then be used to signal hybridization events or association with nucleic acid structures and

provide additional information in both the steady state and time-resolved fluorescence regimes.

Because external labels often do not give specific information relating to binding events, dyes

that associate with DNA through intercalation and groove binding can provide an additional

benefit to the overall design of engineered switching constructs. These classes of dyes can also

serve to reduce the overall complexity of design by reducing the number of labels required for

analysis. For example, while two labels would be required for a FRET method of analysis,

methods of detection that require only one label can be designed using intercalating dyes that are

structurally sensitive to differentiate between cases such as the absence of nucleic acid, and

ssDNA and dsDNA motifs.

1.4.2 General Effects of Molecular Structure on Fluorescence

Most fluorophores are aromatic (with many being heterocyclic) and therefore contain delocalized

pi electron systems. The degree of conjugation alters the extent of the pi system and leads to

shifts in excitation and emission maxima and variations in quantum yield. The transitions

associated with these molecules are most commonly the π→π* transitions, and n→π* transitions

when heteroatoms are present within cyclic structures. Many fluorophores contain substituents,

and their effects alter the π→π* and n→π* transition. The degree and nature of the effects

depend on both the nature and position of the substituent, solvent polarity, and pH of the

surrounding medium.

23

1.4.2.1 Heterocyclic compounds

Fluorescence of aromatics which contain one or more heterocyclic nitrogens (azarenes) depend

heavily on solvent polarity. Related heterocycles containing oxygen and sulfur can be similarly

compared, and in the case where heterocycles contain N, O, or S atoms with single bonds to

carbon, the quantum yield is relatively high owing to the orientation of the π electron system of

the rings and the non-bonding orbital of the N, O or S atom.2 Several classes of fluorophores

exist, including coumarins, rhodamines, pyronines, fluoresceins, cyanine and oxazine classes

which represent examples of some of the more common classes of heterocyclic fluorophores.2

Bridging rotational bonds between two ring structures with an oxygen atom can also restrict

bond rotation. This is the main mode of reducing bond rotation between bridging bonds of

substituted phenyl rings in rhodamine dyes, and this allows rhodamine dyes to be fluorescent in

polar media and exhibit high quantum yields in polar solvents.2 The fluorescence emission from

fluorophores is very sensitive to chemical structure as well as the immediate environment, and

therefore these can be be excellent candidates for use as probes of environmental structure and

polarity.

Some of the most common dyes used for bioanalytical analysis are the cyanine dyes. The most

popular dyes within this class are Cy3, Cy5 and Cy7. The cyanine dyes represent a mode of

relaxation/emission that is affected by solvent polarity, viscosity and twist angle. The structural

components of these molecules allow them to possess two resonance structures where there is a

charged and an uncharged nitrogen atom in each resonance structure. This results in symmetry

of charge distribution when both isomers are considered.2 The fluorescence quantum yield and

decay are are dependent upon the twist angle and solvent viscosity. The non-radiative decay rate

depends on twist angle, and consequently quantum yield depends on solvent viscosity/rigidity of

the surrounding media.2 Cyanine dye derivatives can be readily conjugated to nucleic acids, and

many exhibit charged side chains which can be used for improved water solubility or to prevent

self-association, which is a common cause of self-quenching through aggregation for these types

of dyes. Cyanine dyes are extensively used as both FRET donors and acceptors, and are used as

such for the work in this thesis. Bioconjugates involving cyanine dyes with emission

wavelengths beyond 550 nm are valuable tools for analysis due to their multi-wavelength

detection capabilities (multiplexing), compatibility with excitation sources commonly found in

24

fluorescence microscopy and flow cytometry instrumentation, and less background interferences

caused by tissue autofluorescence occurring at lower wavelengths.71

1.4.3 Dyes that Associate with Double-stranded DNA (dsDNA)

Three major modes of binding can be defined when considering dyes that associate with dsDNA,

including: intercalation, groove-binding and association through attractive electrostatic

interaction.72 The focus of this thesis will be mainly on these modes of binding. Several organic

dyes have been shown to intercalate into dsDNA, and relevant details of this binding mode are

outlined below.

1.4.3.1 Intercalation

Intercalators associate with dsDNA by insertion between the stacked base pairs of DNA.

Intercalative modes of binding are governed by a molecular assembly formed as a result of π-

stacking, hydrogen bonding, van der Waals interactions, and hydrophobic effects.72 These

interactions are important for the transfer of the intercalator from a usually polar aqueous

solution to the less polar internal environment of dsDNA, and it has been shown that

hydrophobic effects and van der Waals interactions at the intercalation site are the primary

factors responsible for efficient intercalation.73 As a result, the hydration layer around the DNA

also changes. An overall positive charge on the dye species is also advantageous for intercalation

of small aromatic compounds, and as such, many intercalating dyes are cationic in nature.74

Intercalation leads to an overall lengthening of the double helix, as the DNA must ‘unwind’ in

order to accommodate an external dye moiety. This leads to an increase in the spacing of the

phosphate groups and a release of counter ions from the grooves of the DNA which results in an

energetically favourable contribution to the free energy of binding by increasing entropy.75,76

Cationic intercalators act to further enhance this contribution by providing additional release of

counter cations from the DNA grooves.72

Intercalators usually bind to dsDNA with binding constants in the 104-106 M-1 range, and these

binding constants can be further increased by structural considerations such as bisintercalation

(two intercalating groups on the same molecule).72 The free energy change of intercalation

(ΔGobs) can be described using a sum of the contributions associated with each energetic

intermolecular interaction:

25

Δ  Δ  Δ  Δ  Δ    Δ (12)

where ΔGconf is the free energy associated with the conformational change of the DNA helix,

ΔGt+r represents the loss of translational and rotational degrees of freedom upon association of

the intercalator and dsDNA (bimolecular complex), ΔGhyd is hydrophobic transfer of a

hydrophobic moiety from a hydrophilic solution to the hydrophobic interior of dsDNA, ΔGpe is

the polyelectrolyte effect of released counter ions (usually sodium ions) from the DNA

backbone, and ΔGmol corresponds to the energy of non-covalent interactions.72 There are several

dyes which bind through intercalative methods and the representative classes with a few relevant

examples are discussed below.

1.4.3.2 Representative Dyes that Bind Through Intercalation

Perhaps the most well-known intercalating dye used for fluorescence detection of DNA is

ethidium bromide (EtBr). EtBr is a cationic phenanthridinium compound that binds to dsDNA

with little- to no sequence specificity, with one dye molecule inserting for every 4-5 base pairs .77

It also binds weakly via a non-intercalative binding mechanism only after the intercalative sites

have been saturated.78 Of particular interest is that despite only a 20-30 fold increase in

fluorescence intensity of EtBr upon binding to DNA, there is a significant increase in

fluorescence lifetime from ~1.7 ns in water to approximately 20 ns upon binding to dsDNA.19

This is quite useful from an analytical standpoint considering the dye lifetime shifts out of the

tissue autofluorescence range. Propidium iodide is another commonly used phenanthridine dye,

and other representative intercalating dye classes include the acridine and anthraquinone dyes.72

Figure 4 shows the structure of EtBr.

Figure 4. Structure of ethidium bromide (EtBr).

26

Another major class of intercalating dyes are the cyanine dyes. Non-intercalating (or labeling)

cyanine dyes also exist. The general structures of intercalating vs. labeling cyanine dyes are

shown in Figure 5 below.

(a)

(b)

(c)

Figure 5. Representative structure of the (a) intercalative class of cyanine dyes (X = O or S) and the (b) external label class of cyanine dyes (Cy3 and Cy5 shown).

Intercalating cyanine dyes show high affinity for nucleic acids and exhibit substantial changes in

their photophysical properties upon binding.79 The commonly encountered dyes in this class are

thiazole orange (TO), oxazole yellow (YO), picogreen, SYBR green, and Cyan 2. Homodimeric

analogues of the thiazole and oxazole dyes are also used extensively in DNA analysis, as well as

the thiazole orange derivative TO-PRO-1 which adds a second cationic centre to the dye. This

thesis focuses on the use of thiazole orange as an intercalating dye for probing nucleic acid

structural organization and hybridization.

27

1.4.3.2.1 Thiazole Orange

Thiazole orange (TO) and its derivative oxazole yellow (YO) have been used as intercalative

transduction agents in nucleic acid hybridization assay and sensor platforms. TO is an

asymmetric cyanine dye. It is a non-planar chromophore comprised of a benzothiazole derivative

and a quinolinium ring linked via a monomethine bridge that allows intramolecular rotation to

occur. It also contains two nitrogen atoms which create charge symmetry due to resonance.

Thus, thiazole orange is weakly fluorescent in solution. However, it is highly fluorescent in

viscous or rigid media, which makes it an excellent choice as an intercalator indicative of DNA

hybridization. The structures of thiazole orange and oxazole yellow are shown in Figure 6.

(a)

(b)

Figure 6. Structure of (a) thiazole orange and (b) oxazole yellow.

TO has been reported to provide ~3000-fold fluorescence intensity enhancement upon DNA

binding.80 It has also been reported that the fluorescence ratio between bound and unbound

states of TO and YO is 18,900 and 700, respectively.81 Despite the variation in reported values,

this is a significant increase over the fluorescence intensity enhancement provided by ethidium

bromide. The increase in quantum yield of TO upon binding to dsDNA is due to the restriction

of rotation around the monomethine bridge upon intercalation of the dye into the double helical

structure as the benzothiazole and quinolinium rings adapt to the propeller twist of the base

pairs.82 The flexibility of the monomethine bridge allows the benzothiazole ring to twist relative

to the quinolinium ring.83 The monomethine bridge has a low energy barrier to rotation and

hence is free to rotate in solution, allowing for the electronically excited dye to relax by non-

radiative decay.82 The quantum yield of free TO in solution has been reported to be 2 x 10-4 at

25ºC, and the binding constant for TO is ~106 M-1.84,85 In viscous media or restricted local

28

environments such as intercalation sites in dsDNA, restricted motion leads to enhancements of

quantum yield typically from ~10-1000 fold, with some reports of up to 3200-fold and 18,900-

fold enhancement upon binding of dyes with DNA.80,81,86,87 This response to change in local

environment can be used to signal changes in local DNA structure.87-90 TO is also able to

associate with DNA through a groove-binding mode of association, and this mode of binding

seems to be dependent on structural permutations to the TO parent dye structure.91 Associated

lifetime changes upon intercalation have been reported as well, and significant increases in

lifetimes from 70 ps for the free dye in solution to 3 ns upon intercalation have been reported.92,93

Although the binding affinity of monomeric TO is lower than that of its its dimeric form,

monomeric TO has a higher dsDNA/ssDNA fluorescence ratio.10,94 As a result, TO has found

extensive use in bioanalytical applications and these are summarized in Table 2 below.

29

Table 2. Selected applications of TO for bioanalysis.

Application TO derivative Highlights Ref.

FRET for signalling nucleic acid hybridization

free dye FRET between free TO donor and an acceptor dye (Cy3, Cy 5, TAMRA, Iowa Black RQ, Iowa Black FQ) covalently linked to DNA probe terminus

10

Light-up probe for nucleic acid hybridization and SNP detection

TO-containing acyclic nucleotide phosphoramidite synthesized containing a serinol or glycerol linker

decrease in background fluorescence that is typically observed for TO in a ssDNA environment through design of linker that prevents dye/nucleobase stacking; SNP discrimination when mismatch is located adjacent to the dye

87

“Hybridization-sensitive On-Off DNA probe”

two TO dye moieties covalently linked to one nucleotide

formation of an H-aggregate between two dye moieties that quenches fluorescence in the absence of target and dissociates upon probe/target hybridization to restore signal use of artificial bases to further improve S/N

95,96

Sequence-specific Thiazole Orange-Peptide Bioconjugate

TO covalently linked to the N-terminus of a peptide sequence capable of sequence specific binding of a DNA target

study sequence-selective DNA-binding peptide probes, and the associated DNA-protein interactions that occur upon binding

92,97

TO as a universal base in PNA for hybridization analysis

TO linked to terminal or mid-sequence to a PNA backbone

TO acts as a universal base in a PNA/DNA duplex while maintaining duplex stability and providing SNP discrimination

98,99

Oligonucleotide-conjugated TO probes for hybridization studies; both in solution and at an interface

TO linked to terminal or mid-sequence positions using alkyl or PEG linkers

detection of non-labelled targets and ability to create self-contained biosensors

94,100-103

TO for labelling cancer cells TO synthesized with various substituents on the benzothiazole and quinolinium rings

TO with an amino-ethyl linker positioned at the quinolinium nitrogen was found to be cell membrane penetrable and could be used to label breast cancer cells

104

Bisquinoliniun/TO G-quadruplex –selective fluorescent probe

TO linked via an amide bond to pyridodicarboxamide (PDC) bisquinolinium

TO and PDC both show selectivity towards G quadruplex structures and a structure combining the two resulted in a highly-specific G-quadruplex ligand

105

cationic conjugated polymer (CCP) amplification of TO fluorescence for DNA detection

free TO as a FRET acceptor and fluorescent probe for hybridization and fluorescent cationic polymer as a FRET donor

Visual, sequence specific, instrument-free detection of PNA/DNA duplexes with SNP discrimination where addition of CCP improves sensitivity and selectivity by FRET sensitized enhancement of TO emission

106

TO/Gd chelate contrast imaging agent for MRI detection of cell death

TO-Pro-1 covalently linked to a Gd chelate

GadoTO was preferentially located in condensed nuclei of necrotic cells and enables detection/imaging of cell death by MRI or fluorescence

107

30

TO is an excellent candidate for the detection of label-free targets due to the photophysical

changes that occur upon binding to dsDNA that are not readily observed with external labels.

Many synthetic analogues of TO have been reported in the literature, and the covalent

attachment of tethers for conjugation to biomolecules (mainly DNA and PNA) is commonplace.

Tethers can be attached to either the quinolinium nitrogen or the benzothiazole nitrogen atoms.

However, it has been shown that conjugates with tethers attached at the quinolinium nitrogen

display higher quantum yield values, suggesting that conjugation at the benzothiazole nitrogen

may inhibit or impede intercalation due to the positioning of the dye within the minor groove.97

1.5 DNA and Molecular Switching Techniques Molecular switching technologies play a significant role in the development of bioanalytical

technologies. Binding-induced biomolecular switches used for the detection of nucleic acids are

becoming increasingly prevalent. Strategies that can exhibit ‘‘on/off’’ optical signaling changes

as a function of target binding offer the advantages of eliminating target labeling and are

potentially suitable for intracellular monitoring of DNA or RNA. In this thesis, the use of

binding-induced switching with a single tethered intercalant label and an oligonucleotide probe

provides an elegant, yet simple approach to create a nanoscale construct that can be used in both

solution and at an interface to monitor biomolecular interactions. The necessary molecular

components are organized at an interface through site-directed templating. The structure of

Neutravidin offers binding sites, and biotinylated oligonucleotide and a biotinylated tether

capped with intercalating dye are allowed to self-assemble to provide appropriate molecular level

spacing. It is evident that there are several advantages to this construct, including: easily

assembled modular components; a novel biotinylated intercalating dye which provides stable

binding of the signalling moiety with high binding affinity; the ability to monitor binding events

both in solution and at an interface; analysis through two channels of fluorescence (steady-state

and time-resolved); and label-free target analysis (in this case “label-free” indicates that neither

the probe or target oligonucleotide requires conjugation with a fluorescent probe). In general, the

main design consideration of such approaches is maximization of the signal-to-noise ratio

between bound and unbound states, whether in the steady state or time-resolved fluorescence

analysis modalities. Nucleic acids, due to their selective self-recognition, conformational

freedom, and clinical and biological importance, are ideal building blocks for developing

31

molecular switches for diagnostics. The following section outlines various molecular switching

techniques.

1.5.1 Molecular Beacons and Hairpin Probes

The best-known example of a nucleic-acid molecular switch is the molecular beacon (MB). MBs

have been widely used in nucleic acid diagnostic technology since their introduction by Tyagi

and Kramer.8 The dual-labeled, stem-loop structures unfold in the presence of target and lead to

a change in fluorescence signal mediated by FRET. Conventional MBs use fluorophore-quencher

pairs or fluorophore-fluorophore pairs to generate FRET mediated signals. Under optimized

conditions, MBs have been shown to give 200-fold changes in signal in solution.108 These

relatively simple, early examples of nano-scale engineering overcame the issue of target labeling

and are capable of single nucleotide polymorphism (SNP) discrimination.108 More recently, the

MB field has made use of advances in inorganic nanotechnology to try to improve existing

technologies. Table 3 summarizes some of these developments. 11,109-118

32

Table 3. Recent developments in molecular beacon and hairpin probe technology. The * represents novel/unique design criteria that was explored in the development of the technology.

Donor/Quencher Probe Mode of Detection

Highlights Ref.

Cy3/BHQ LNA*+

DNA

Fluorescence Nuclease-resistant; low cellular protein binding probe; long-term intracellular monitoring (5-24 hrs) of mRNA expression; rapid hybridization kinetics and selectivity

119

Pyrene excimer*/DABCYL

DNA Time-resolved Fluorescence*

40 ns lifetime and 130 nm Stokes’ shift of pyrene excimer; time-resolved methods to enhance sensitivity; ‘tunable intensity’ based on addition of multiple pyrene monomers

109

QDs*/BHQ DNA Fluorescence Three colour multiplex detection; SNP discrimination; increased S/N ratio c.f. SYBR Gold

111

QD*/ Cy5* DNA Fluorescence Ratiometric sensing using two fluorophores; bifunctional peptide linker for better control over number of probes per QD and easier purification

112

Ru(bpy)2(dcbpy)*/ gold electrode

DNA ECL No excitation source; 90 pM LOD; SNP discrimination; reusable sensor configuration

113

TAMRA/gold film* DNA Fluorescence 100-fold increase in fluorescence upon hybridization; SNP discrimination; elimination of dual probe labeling; 10 pM LOD; reusable sensor configuration

114,115

Fluorescein; Rhodamine 6G; Texas Red; Cy 5/Au NP*

DNA Fluorescence >97.5% quenching efficiency by Au NP; better quenching of NIR dyes; >100-fold fluorescence increase on hybridization; SNP discrimination

11

Fluorescein/SWNT* DNA Fluorescence No dual probe labeling; ssDNA/SWNT complexes via pi-stacking; increased S/N ratios c.f. DABCYL-MB; 4 nM LOD; SNP discrimination

120

Oxazine dye (MR121)/no quencher*

DNA Fluorescence No dual probe labeling; oxazine dye quenched by G bases in hairpin configuration; sub-pM LOD; reusable sensor configuration; rapid

117

Py-1 (ChromeoTM P-502)* dye/no quencher

DNA Fluorescence lifetime*

No dual probe labeling; single probe labeling indicated by colour change and >10-fold increase in QY; increased sensitivity via fluorescence lifetime measurements; pM LOD

118

33

As shown in Table 3, quantum dots (QD) have been incorporated into MBs as luminescent

donors.112,113 Different linkage strategies have been investigated, and highlight the importance of

establishing the optimal linkage chemistry between the QD surface and the probe nucleic acid in

order to optimize response.112,121 Given the brightness of QDs, it was natural to introduce more

efficient quenchers into MBs. Both QDs and gold nanoparticles (NP) have been combined to

create MBs, where the gold NP is a very effective fluorescence quencher.121 Gold NPs have also

been used as quenchers in MB configurations using organic fluorophores. Such experiments

have demonstrated >99% quenching efficiency and improvement in overall sensitivity.11

Similarly, single-wall carbon nanotubes (SWCNTs) have recently been used as quenchers for

fluorophore labelled hairpin probes.117 The probe ssDNA and SWCNTs form complexes through

non-covalent pi stacking interactions that result in fluorescence quenching of the dye-labeled

probe. Upon hybridization of the target, decreased affinity between dsDNA and the SWCNTs

allows restoration of the fluorescent signal as the dye lifts off the SWCNT surface. These

complexes were shown to provide better signal-to-noise ratios compared to conventional MBs

using DABCYL as the quencher. The SWCNT-MB complex also exhibited higher thermal

stability than the DABCYL-MB counterpart. In addition to QDs, other less conventional donor

components of MBs have been pyrene excimers. Time-resolved measurements provided better

sensitivity compared to steady-state fluorescence measurements due to the long 40-ns lifetime

and large 130-nm Stokes shift of the pyrene excimer.109 This approach is well suited to

measurements in complicated matrices with high autofluorescence backgrounds. Another less

conventional approach that has been developed is electrochemiluminescence (ECL)-based MBs

using hairpins immobilized on gold electrodes.113 While a number of strategies have been

developed using a metal surface to quench immobilized hairpins that are labeled with

fluorophores,114,115,122,123 the ECL approach is quencher free. Upon target binding, the ECL

reporter is displaced from the electrode surface and the ECL signal is decreased, generating an

‘‘off’’ signal for hybridization. Another quencher-free approach was developed using oxazine

derivatives covalently attached to oligonucleotide hairpin probes.117 These probes have the

typical hairpin structure with the addition of a poly-guanosine tail at one terminus and the

oxazine dye at the other. The oxazine is quenched via photo-induced electron transfer through

stacking with the adjacent guanosine residues in the closed conformation.124 Upon target binding,

the stacking is removed and the fluorescence emission of the dye is recovered.

34

Dyes that exhibit changes in fluorescence intensity upon hybridization without the presence of an

additional quenching species represent a significant simplification of the overall design of

switching technologies. Apart from label selection, another design criterion of MBs is selection

of the hairpin material. The different structures and the energetics of hybridization between

different types of nucleic acid can affect the performance of MBs. In contrast to their DNA

counterparts, MBs based on PNA are insensitive to salt concentration and resist DNA binding

proteins.125 PNA-MBs can also exhibit efficient fluorophore quenching without the typical stem

structure due to the greater hydrophobicity of PNA.125 A particularly important development has

been MBs based on locked nucleic acids (LNAs).119 These MBs are far more resistant to

nuclease degradation and protein adsorption than DNA-based MBs. As such, they have potential

for intracellular monitoring. These MBs can overcome the limitation of false-positive signals that

arise from MBs that have undergone conformational changes not associated with target binding

or enzymatic degradation caused by the cellular machinery. Some limitations remain in the area

of MB technology. Many of the aforementioned examples are restricted to solution-phase assays.

In addition, MBs are also synthetically complex when one considers the dual labelling

component. Although several strategies involving MBs immobilized at an interface have been

explored51,126-128, these suffer from high background signals that are associated with false-

positive signals due to poor stability of the stem-loop structure once it has been immobilized.129

The fluorescence enhancement from a MB probe/target binding event drops from approximately

25-fold in solution to approximately 2-5-fold when the MB is immobilized.8,130 Biosensors

developed using immobilized MBs containing LNA bases have led to improvements in terms of

signal-to-noise.129 However, LNAs are currently a very expensive (~$790.00 CDN for a

minimum guaranteed yield of 10 nmoles of product for a 25-mer dual-labeled probe131 with up to

six LNA insertions and ~$465.00 USD for a minimum guaranteed yield of 7.7 nmoles for a 19-

mer biotinylated, linear probe132 with up to nine LNA insertions) and synthetically complex route

for probe design. LNA has also been shown to have a slower hybridization rate when compared

to a DNA counterpart.110,116,133 It is apparent that there are trade-offs in terms of increasing

signal-to-noise and speed of signal acquisition.

1.5.2 Other Molecular Switching Constructs

Beyond MB technology, the creation of DNA-based nanostructures and optical materials capable

of undergoing a signal change upon probe/target hybridization is of substantial interest in the

35

field of nucleic acid diagnostics. Currently, much of the literature about DNA ‘‘switching’’

devices is associated with the construction of DNA-based nanomachines that, in some cases, use

hybridization as a method of creating a nanomechanical switch.134-137 Functional nanostructures

derived from nucleic acid building blocks are also used for applications in bioelectronics and

molecular computing.138-141 However, more emphasis is now being placed on creating switch-

type technologies to monitor DNA-hybridization events for diagnostic applications that eliminate

the requirement for target labeling. Figure 7 highlights some of these applications.

Figure 7. Working principles of selected molecular switching applications for label-free nucleic acid diagnostics. (a) molecular beacon8; (b) ‘hybrid molecular probe technology143; (c) ‘snap-to-it’ probe142; (d) surface immobilized tethered thiazole orange oligonucleotide probe102; (e) ‘switchable DNA interface’145.

36

One switching technology developed with DNA utilizes an open-closed loop transition without

the traditional stem segment. Morgan et al.142 have created a ‘‘snap to-it’’ probe technology that

uses a PNA probe containing two metal-chelating ligands at each probe terminus that co-ordinate

one metal centre (Figure 7(c)). In the presence of either Ni2+ or Cu2+ (depending on the ligand

used), these coordinating ligands constrain the probe to a loop structure. The termini are also

labeled with a fluorophore (EDANS) and quencher (DABCYL), respectively, creating a

modified MB-type function. The energetics of hybridization are sufficient to overcome the

energy associated with the metal/ligand complexation. The ‘‘snap-to-it’’ probes were shown to

have enhanced specificity for fully complementary targets over single base mismatches

compared to probes lacking metal chelating agents. These ‘‘snap-to-it’’ probes showed a 21 ºC

difference in Tm between the fully complementary and single-base mismatch condition, and

showed a 40-fold increase in fluorescence upon complementary probe/target binding. Similarly,

Martinez et al.143 have developed a ‘‘hybrid molecular probe’’ (HMP) to overcome the

limitations of false-positive signals associated with enzymatic degradation or cellular protein

adsorption with MBs in intracellular environments (Figure 7(b)). The HMP comprises two

strands of ssDNA separated by a covalently-linked PEG spacer. At the distal ends of the DNA

sequences are attached two different fluorophores capable of undergoing FRET. Upon

hybridization, the two strands of DNA hybridize to the target in an orientation that places the two

fluorophores adjacent to one another. The resulting FRET provides the analytical signal, which is

calibrated against variations in intensity due to photobleaching or changes in environment by

making ratiometric measurements. The PEG tether was shown to facilitate hybridization and the

HMP exhibited faster hybridization kinetics compared to a MB that recognized the same target

sequence. The HMP has also been shown to be more effective when compared to a MB in cell

lysate, showing less degradation, and was also capable of monitoring hybridization both in

solution and in an intracellular environment. Buck et al.144 have developed a ‘‘DNA

nanoswitch’’ for the detection of nucleic-acid hybridization based on the conformation of a

Holliday junction and FRET. This switching technology was able to achieve a 30-fold difference

in signal between fully complementary and single-base mismatched targets. It was also shown

that the nanoswitch could detect complementary RNA targets in nM concentrations from a

heterogeneous mixture of RNA. Rant et al.145 have used the fluorescence quenching properties of

gold to create ‘‘switchable DNA interfaces’’ for the detection of non-labeled targets (Figure

7(e)). A 3´-thiol terminated oligonucleotide probe with a Cy3 fluorophore at the 5´-end was

37

grafted in low densities to a gold electrode. By application of an AC potential, charge driven

oscillations of the DNA were monitored by measuring fluorescence intensity. Label-free

transduction of hybridization was based on the difference in switching kinetics between ssDNA

and dsDNA.

When designing fluorescence-based hybridization assays, the transduction and the specific

binding components must be integrated. Single materials that concurrently bind nucleic acids and

generate a change in optical signal are of significant interest. A substantial body of work exists

about the use of fluorescent cationic polythiophenes for signal amplification for detecting nucleic

acid hybridization.146-150 The fluorescence of the cationic polymer derivative is quenched when

it adopts a conjugated planar conformation with ssDNA.149 Upon hybridization, the polymer

adopts a helical, non-planar configuration that leads to a large increase in fluorescence signal.149

Upon target hybridization, the aggregation of Alexa546-labeled probes allowed a single

polythiophene donor to transfer energy to several Alexa546 acceptors. Similarly, the use of

fluorescent dyes which undergo a substantial fluorescence enhancement in the presence of

dsDNA is a very attractive method for the detection of nucleic acid hybridization. These

materials can operate in a switch configuration by being ‘‘switched off’’ when only ssDNA

probes are present and ‘‘switched on’’ when there is a hybridization event. Probe

oligonucleotides with tethered TO dyes have been used to monitor hybridization both in solution

and at an interface as outlined in Table 2 and shown in Figure 7 (d). In addition to intercalating

dyes, fluorescent labels have been incorporated within nucleic-acid sequences. Menacher et al.151

incorporated 5-nitroindole as a surrogate DNA base in a probe sequence and situated a Cy3-

modified nucleobase two positions away. The Cy3 underwent a fluorescence enhancement upon

probe/target hybridization. This approach is similar to the strategy of base-discriminating

fluorescent (BDF) probes.152-155 To create a BDF probe, a nucleoside derivative is synthesized,

usually through either attachment of an extrinsic fluorophore label to the nucleobase or

derivitization of the nucleobase to create an intrinsically fluorescent moeity.152 The fluorescent

dyes are sensitive to their local environment, showing changes in spectral intensity or profile (i.e.

color) as, for example, local hydrophobicity changes between ssDNA and dsDNA. However,

because a mismatch introduces a local change in the conformation of dsDNA, BDF probes are

also sensitive to the difference in environment between the modified deoxyuridine pairing with

adenine or forming a mismatch with another nucleobase. SNP discrimination has been

38

demonstrated using BDF probes.154 Much like the aforementioned MB technologies, the

majority of these molecular switching constructs are used for solution-phase assays, have

complicated synthetic components, or rely on an immobilization platform to assist in the

transduction process (as in the case for the gold electrode, where it is also required to monitor

AC potential as well as optical transduction). These factors add to the overall complexity of

analysis. However, recently there has been a molecular switch created using an intercalating dye

and SWCT system for analysis.15 This solution phase assay relies on the affinity interaction of

ssDNA probes and SWCNT. Ethidium bromide (EtBr) associated with the ssDNA (the dye is

not covalently bound to the ssDNA probe) is quenched by the SWCNT. Upon introduction of

target, the dsDNA is released from the SWCNT, and the fluorescence signal from the EtBr is

restored to provide an analytical signal. Again, this technology is currently restricted to solution

phase assay format. The work presented in this thesis overcomes some of the limitations of these

competing technologies, and these design elements are highlighted in section 1.8.

1.6 Fiber Optic Sensing Modalities Sensing elements based on optical modes of detection represent a common approach for

generating analytical signals based on nucleic acid hybridization or association of molecular

components at an interface. Optical transduction mechanisms based on fiber optics are

commonly used for biorecognition platforms since they can provide sensitive measurements and

are amenable to numerous surface modification options.156 Optical fibers are a commonly used

solid substrate for immobilization of biomolecules, including nucleic acids. The use of optical

fibers provides both an immobilization platform and a method for delivery of light required for

excitation of fluorescent materials used to signal a hybridization event. Further, sensitivity is

derived from the limited penetration depth associated with the evanescent field intensity which is

generated at the interface. These methods of analysis operate based on the principle that as light

strikes an interface between two materials of differing refractive indices (n1 and n2), total internal

reflection results at the critical angle θ, where:

sin θ       (12)

Optical fibers are used routinely for the transmission of electromagnetic radiation over long

distances by confinement of the photons within guided modes of a cylindrical waveguide.

39

Optical radiation propagates through modes of optical fibers by total internal reflection (TIR)

and at the points of reflection, at the interface between the waveguide core and cladding,

localized zones of constructive interference are created. In sensing applications, the cladding

material associated with the optical fiber is removed, and the analyte/aqueous medium becomes

the cladding. This serves to create a standing electric field intensity, known as an evanescent

wave, that extends beyond the waveguide (into the medium of lower refractive index) and decays

exponentially with increasing distance from the interface. The intensity of the evanescent wave

can be determined using the following relationship:

  exp (13)

where ET represents the magnitude of the electric field vector, EoT is the magnitude of the electric

field vector at the interface (area of highest intensity), z is the distance along the normal to the

plane defined by the interface and δ is the characteristic decay length. The characteristic decay

length is dependent on the wavelength (λ) of the light under consideration, the refractive index of

the waveguide core (n1) and the outer medium (n2) and the propagation angle (θ) of the guided

photons relative to the normal to the interface (TIR incidence angle) as given by:

    sin   (14)

The effective sampling depth or penetration depth (dp) is representative of the characteristic

decay length and represents the distance at which the intensity of the evanescent field has

decayed to 1/e with respect to the maximum (at the interface), as defined by:

    sin   (15)

Excitation of a fluorophore that is close to the surface of the waveguide can be achieved via the

evanescent field, and the sampling depth that is characteristic of the wavelength range associated

with visible radiation provides a sensing window for interrogation of short-strand

oligonucleotide hybridization events at an interface. The resulting fluorescence emission is

40

isotropically distributed, and if an oligonucleotide probe is immobilized on the surface on an

optical fiber and participates in the formation of a hybridized complex with a fluorophore that is

also immobilized at the interface, or a fluorophore-labelled target, then a fluorescent signal can

be generated and detected to provide an interfacially-sensitive method for detection of nucleic

acid hybridization or molecular association of modular components.14,157,158 Optical fibers

provide the advantages of small size, flexible geometry and length, and the potential for remote

sensing applications.156

1.7 Oligonucleotides as Targets for Interfacial Hybridization Nucleic acid biosensors are built on the premise that single stranded DNA (ssDNA) is able to

recognize and bind its complementary target. If the surface chemistry for immobilizing ssDNA

on a transduction platform is carefully chosen and the experimental conditions finely tuned, the

immobilized ssDNA film can bind its complement, creating a device with truly useful diagnostic

potential. This methodology can also be extended to consider the building of molecular

constructs and monitoring their interactions at an interface. Control over the immobilization

chemistry leads to the creation of an organized film of probe oligonucleotide that promotes

selective hybridization with target material while concomitantly reducing non-selective

adsorptive effects.159 The physical phenomenon of hybridization at an interface inherently

invokes a change in the values of the association constants with respect to solution-phase

hybridization. There is loss of degrees of freedom associated with pinning the probe species at

the surface, and the introduction of surface free energy and a different electrostatic and solvation

environment. As a consequence, configurational freedom of the probe becomes an important

design element in creating a sensor for detection of a hybridization event.160 Much focus is

directed at control of the local physicochemical environment at the surface due to its great

importance in sensor design. In this work, probe oligonucleotides were immobilized at an

interface using a biotin functionalized tether to allow strong affinity binding, where the tether

allows extension of the probe oligonucleotide into solution. It must also be noted that the

charged interface of a nucleic acid biosensor will affect the local conditions of pH, ionic

strength, and dielectric constant leading to a hybridization environment for the probe-target

duplex which is very different from that experienced in bulk solution. Thus the thermodynamics

of hybridization differ between bulk solution and at an interface.161

41

The most common target materials used for nucleic acid biosensors are synthetic short-chain

oligonucleotides. Most commonly short oligonucleotides (~20 bases in length) are immobilized

onto solid supports as the biorecognition elements (i.e. probes) in a nucleic acid construct. In

this work, a 19-mer probe/target system was used. A short-chain oligonucleotide probe-target

duplex offers the advantage of rapid kinetics of hybrid formation.162-164 The kinetics of probe-

target binding are affected by surface interactions and the local electrostatic environment when

the probe-target hybridization event is brought to a surface, and the kinetics (approach to

equilibrium) and thermodynamics (processes at equilibrium) of hybridization at an interface

become crucial in guiding device design.165 The energetics of binding determine the overall

selectivity of the hybridization event at the sensor surface, and the enthalpy, entropy and

transitional free energy changes between dsDNA and ssDNA can be calculated to optimize

selective target-probe hybridization.162 The thermal denaturation temperature (Tm) of the probe-

target hybrids is generally taken as the indicator of these thermodynamic parameters. The Tm

represents the temperature at which ~50% of the duplexes have denatured and can be used as an

indicator for the stability of a given duplex.166 If the stability of a given nucleic acid duplex is

compromised by the presence of a single base-pair mismatch, or the addition of a denaturant

such as formamide or urea, the Tm for that particular system will be lower than that of its fully

complementary counterpart.167 In this manner, the selectivity of the system can be tuned such

that a sensor containing an ensemble of immobilized oligonucleotides will on average selectively

bind only the fully complementary target at a particular temperature. This temperature must be

above the melt temperature of the mismatched duplex and below the melt temperature of the

fully complementary duplex. Either a large difference in Tm or a sharp melting transition is

required to ensure that the fraction of fully complementary of dsDNA is near unity, while the

fraction of mismatched dsDNA is near zero. In the work considered in this thesis, knowledge of

the Tm of the probe/target duplex is required in order to employ the site-directed templating

strategy of the overall device design. In this case, the temperature required to melt off the target

strand and leave the biotinylated probe and intercalating dye intact as part of the switching

construct is part of the overall design strategy. Factors affecting the Tm of a given duplex

include temperature, pH, ionic strength, presence of a denaturant such as formamide or urea in

solution, and base sequence.168,169

42

Levicky and Horgen165 estimate that a surface coverage of 1012-1013 probes/cm2 on a solid

substrate could be comparable to a solution concentration of ~0.1-1 M solution of probes and

targets. This is a much higher concentration than typically experienced in a bulk solution

experiment, clearly demonstrating the difference between solution and interfacial hybridization

environments. Watterson and coworkers170 have extensively studied the thermal denaturation of

oligonucleotide duplexes at an interface and discovered that factors such as degree of sequence

complementarity, G•C content, pH, ionic strength, local dielectric strength, immobilized strand

density, and the nature of the substrate all contribute to the observed Tm at an interface.170

Watterson et al. 170 have also shown that there is a tendency for greater selectivity at the interface

relative to bulk solution and that the difference in melt temperature between single base-pair

mismatched and fully complementary can be altered as a function of the density of the

immobilized probes. This is extremely advantageous from a sensor perspective since the

thermodynamics of selectivity are enhanced at an interface as opposed to that in bulk solution.

Combinations of probe density, ionic strength, and temperature can be used to optimize the

interfacial hybridization events when compared to the same system in a bulk solution

environment.170 Factors such as the density of ssDNA immobilized at the sensor surface directly

influence the charge density and hence ionic strength at the interface and dramatically alter the

thermodynamics of hybridization with respect to a similar nucleic acid system present in bulk

solution.161 Piunno et al.161 found that the thermodynamic stability of dsDNA at an interface is

very different from that in solution, and may manifest as changes in enthalpy as a function of

ionic strength, asymmetry in melt curves, and reduced dielectric constants. In dilute solution,

individual DNA molecules do not experience one another to a great degree. However, when

densely packed on a surface, neighbouring DNA molecules experience one another due to their

polyanionic nature and electrostatic interactions. Thus it is not surprising that probe density can

influence the Tm of the hybridized duplex, nor that probe density can directly alter the orientation

of the ssDNA probes at a surface.161 At low densities there is a tendency for DNA to collapse

onto the surface, whereas at high density the strands tend to extend into the surrounding solution.

It follows that, if probe orientation is such that the availability of the nucleation site on the

ssDNA where hybridization initiates is compromised, the kinetics of hybridization will be altered

as a function of density.161 Computational predictions were tested experimentally and it was

determined that duplexes that formed at an interface were destabilized with respect to their

counterparts in solution. This was a result of interactions between the immobilized probes and

43

neighbouring strands, and the solid substrate. Ionic strength was also found to be a factor in

duplex stability. The results further showed that selectivity was a function of ionic strength,

immobilized strand density, and the sequence of the nucleic acids of interest.

While it has been stated that the direct comparison of thermodynamics and binding parameters

between surface and solution phase hybridization events can be erroneous due to substantially

different environments161, trends within data sets can be compared and appropriate arguments

made in terms of stability of duplex formation. The work presented in Chapter 5 represents a

solution phase study of selectivity that can be used as a platform for comparison of future work

that would include investigating the same selectivity parameters at an interface using the

molecular switching hybridization platform introduced in Chapter 4.

1.7.1 The Influence of Fluorescent Labels on Duplex Stability

In a typical fluorescence detection scheme for nucleic acid diagnostics, a fluorophore label is

covalently attached to the oligonucleotide target sequences and upon probe-target hybridization,

a fluorescent signal is detected.

The target solution has an impact on the fluorescence properties of the dye since its emissive

properties are inherently sensitive to local environmental factors including pH, ionic strength,

and the presence of species capable of fluorescence quenching or coupling through FRET. The

concentration and the nucleobase composition of the target DNA in solution can also influence

the quantum yield of the fluorophore as well as the degree of structural organization of the

targets to which the fluorophores are tethered.6,10 Properties of the target solution could also

influence the degree of structural organization of targets containing longer sequences of nucleic

acids such as RNA and PCR fragment targets.

Usually any effects the label has on the duplex stability are ignored, or it is assumed that the Tm

is not modified by the presence of an extrinsic label, which may not be true. In studies of G-

quadruplex171 and i-motif DNA structures172, the presence of covalently attached dye labels

destabilized the structures formed and showed a Tm shift by up to 11ºC. On the other hand,

intercalating dyes and molecular beacon-type structures with a fluorophore-quencher pair in

close proximity can potentially form complexes that act to stabilize DNA structures.173-175 A

broadly applicable accurate model may be very difficult to develop as the thermodynamic effects

44

are likely dye specific owing to the vast array of molecular structures and different methods of

dye conjugation. Many dyes are also differently sensitive to each of the four DNA nucleotides

and this can further influence the stability of hybrids that are formed. Moreira et al. 176 have

measured the effects of a large number of commonly used fluorophores and quenchers on the

stability of DNA duplexes using fluorescently labeled oligonucleotides and fluorophore-

quencher labeled molecular beacon probes. The fluorophores investigated were: 6-

Carboxyfluorescein (FAM), hexachlorofluorescein (HEX), Cyanine 3 (Cy3), Cyanine 5 (Cy5),

tetrachlorofluorescein (TET), carboxytetramethylrhodamine (TAMRA), and Texas Red-X.

Quenchers studied were: Black Hole Quenchers 1 and 2 (BHQ1 and BHQ2), and Iowa Black RQ

and FQ (IabRQ, IabFQ). The base sequences and buffer compositions were kept the same for all

experiments so the data sets of all the dye-quencher species studied could be compared. Melt

curves of the fully matched duplexes to which the labels were covalently attached were collected

by measuring the absorption at 260 nm as a function of temperature. It was found that a 3′-

BHQ2 label stabilized the duplex, increasing the Tm by 2.3 ºC. A 5′-Cy5 labeled probe

hybridized with a 3′-IabRQ labeled target (putting the fluorophore-quencher in close proximity)

showed a 4.3 ºC increase in Tm. The Cy3 and Cy5 dyes placed at the 5′-end and hybridized with

a non-labelled sequence also stabilized the duplex, increasing Tm by 1.6 ºC. TAMRA and Texas

Red-X had a marginal stabilizing effect on the duplex and the observed Tm increases were within

the experimental error. FAM and HEX had no apparent stabilizing effect, and only one dye

(TET) destabilized the duplex, and resulted in a Tm decrease of 0.5 ºC. For the quencher labels,

BHQ2 and IabRQ increased the duplex Tm by up to 2.6 ºC and provided for an overall decrease

in ΔG˚60ºC. BHQ1 and IabFQ increased the Tm by 1.2 ºC while DABCYL-dC and TAMRA had

no effect on the duplex stability. For dual-labeled probes (fluorophore and quencher at opposite

termini), such as those used in molecular beacon technology, the effects of stabilization were

additive for FAM-oligo-Dabcyl-dC, HEX-oligo-BHQ2, Cy5-oligo-IabRQ probes. The FAM-

oligo-IabFQ showed a stabilizing effect higher than summative in terms of the ΔG060ºC values,

whereas for Cy3-oligo-BHQ1 the stabilizing effect was slightly less than summative. The

interactions between certain fluorophore-quencher pairs must be taken into consideration when

analyzing these values as some pairs can form complexes which act to stabilize or destabilize the

duplex structure. Dangling ends, which occur when one of the sequences making up the duplex

is longer than the other, were also studied and their presence acted to further increase the duplex

stability imparted by the dye. The thermodynamic stabilization of the duplex by the dyes is

45

mostly entropic in nature. These entropic gains are considerable and are thought to be because

of changes in hydration and organization of counter ions when strands hybridize. For example,

the increase in entropy associated with Cy3 and Cy5 is ΔΔS0vh = +21 cal/(mol K). Although the

actual physical mechanism of the dye/quencher-DNA interactions that cause the stabilizing

effects is unknown, the intermolecular forces of hydrogen bonding and base stacking may

contribute to the process. The increased stabilization of the Cy3 and Cy5 dyes over the other

labels could be due in part to pseudo-intercalating or minor groove binding interactions with the

duplex, as these processes are known to stabilize double-stranded DNA hybrids. Intercalation

can provide significant favourable enthalpic modes of stabilization, and singly charged dye

species capable of intercalation can increase favourable enthalpic contributions by as much as -8

kcal/mol. It should be noted that the thermodynamic measurements for these systems are based

on a particular nucleic acid sequence (20-mer with base composition 5’-

ACCCGTTCACCCTCCCCCAG-3’) and that the sequence length and base composition will

have effects on these values. The stabilizing effects are expected to increase as duplex length

increases, and base sequence effects would be dye-specific as it is known that some dyes are

nucleobase-sensitive. For example, TAMRA fluorescence can be quenched by nearby G

residues, and the mechanism of this phenomenon may have to do with structural properties

which could also affect duplex stability. Therefore, it can be seen that the development of a

thermodynamic model incorporating extrinsic dye labels could be quite a daunting task when one

considers the significant role that environment plays in controlling the sensitivity of the

dye/quencher species.

1.7.2 Structural Factors Affecting Hybridization at an Interface

Early work by Southern and coworkers established fundamental aspects of analyzing

hybridization behaviour of nucleic acids at an interface.177-181 Several conditions for target

oligonucleotide hybridization were investigated using template-mediated synthetic

oligonucleotide arrays on glass substrates. Oligonucleotide duplex formation is dependent on

several structural factors including: oligonucleotide length, base sequence, number of

mismatches, and both probe and target concentration.161,177 Significant information about

hybridization behaviour and oligonucleotide interaction can then be gathered from the

thermodynamic and kinetic data of such experiments including the association and dissociation

constants of nucleic acid binding. Maskos and Southern177 showed that duplex formation

46

between 10-mer oligopyrimidine targets and immobilized octapurine probes is dependent on

nucleobase composition, and that this dependence can be decreased by the use of chaotropic

reagents. Short-chain oligonucleotides were used in order to eliminate secondary structure

formation through intra-strand base-pairing as this would add another level of complexity to the

analysis. Strands with a guanosine (G) at the end of a duplex gave up to 50% higher duplex

yields over those with adenosine (A) in the same position. In terms of mismatches, it was also

shown that higher duplex yields were obtained when a terminal mismatch contained a G residue.

This further confirms reports that G•T and A•G mismatches are less destabilizing than A•C and

A•A mismatches.182 Although the effects of nucleobase composition on stability are usually

‘averaged-out’ in longer double-stranded duplexes, the oligonucleotides immobilized at an

interface are usually between 15 and 25 base-pairs in length. The duplex stability dependence on

base sequence then becomes a parameter to consider in sensor design.

1.8 Contributions of This Thesis

1.8.1 Scope

Nucleic acid biosensing platforms are under continuous development as a method to provide

rapid, reliable detection of pathogens and genetic diseases. Constructs that are simple in design,

assembly, and functionality are an obvious advantage for the detection of nucleic acid targets.

Label-free techniques or techniques that incorporate an intrinsic signaling mechanism eliminate

the requirement for target labelling, which decreases the overall complexity and time required for

analysis. While the concept of label free analysis is the central dogma for sensor designs utilizing

surface plasmon resonance and piezoelectric devices, the attractiveness of a fluorescence method

that eliminates target labelling while providing an overall simplicity in terms of synthesis,

construction and data processing is also of interest from the perspective of providing a definitive

signal for selective target binding and potential for multiplexed analyses. While several methods

report sandwich assays or addition of multiple reagents subsequent to probe/target binding as

‘label free’, this adds to the overall analysis time and fundamentally requires the use of a labelled

reagent which is not part of the sensor construct. Nanoscale assemblies have been reported which

contain all components of the “sensor cargo” in one unit, such as those found in molecular

beacon systems.51 There has also been some work done with immobilization of labelled

oligonucleotide probes at an interface.102,103 The sensor community has witnessed an introduction

47

and subsequent increased motivation towards the use of nanoparticles for use as both engineering

scaffolds and provision of optical signalling mechanisms via FRET or surface enhanced raman

scattering (SERS) due to the unique properties these materials exhibit as a function of moving to

the nanoscale regime.12,183 However, fluorescent dyes still hold a significant role in the

fluorescence based assay and sensor community. Several FRET based methods of analysis still

rely on a fluorescent dye to provide an acceptor for FRET-sensitized emission.12-14,184

Fluorescent dyes, and in particular intercalating and groove binding dyes, can provide one

fundamental advantage over nanoparticle assemblies, which is that the mechanism of

fluorescence change can provide information about the structure of a hybrid rather than acting

only as an external label.

1.8.2 Overview of Contributions

The work presented in this thesis introduces a sequential development and transition of nucleic

acid detection from a FRET-based target-labelled analysis to a label free analysis involving a

singular signalling fluorophore (thiazole orange). The latter is enhanced by a preliminary

investigation of a simple, yet effective method for improving both signal-to-noise and selectivity

in solution based assays. This thesis is divided into six main chapters, including the introduction

that provides the necessary background elements required to put the work that is presented into

context.

Chapters 2 and 3 are based on FRET methods of analysis for nucleic acid hybridization

detection. Chapter 2 is a survey of FRET-based parameters for several external label FRET pairs

applied to nucleic hybridization detection. From this survey, the TAMRA/Iowa BlackRQ FRET

pair was further studied in-depth and showed the ability to provide selectivity for base pair

mismatches in nucleic hybridization experiments both in solution and at an interface. Chapter 3

is dedicated to the concept of FRET based methods of analysis involving the intercalating dye

thiazole orange. Several FRET acceptors using thiazole orange as a donor were studied, and it

was concluded that TO was a viable FRET donor for each FRET pair. Factors such as the

importance of label choice and concentration were evaluated and showed that these factors are

critical in designing FRET pairs for analysis to avoid concentration-mediated dimerization and

aggregation which can lead to undesirable contact-mediated quenching effects.

48

Chapter 4 details the investigation of a novel molecular switching technique for nucleic acid

hybridization at an interface involving a nanoscale construct centred around a novel biotin-

tethered thiazole orange dye derivative. The interest was in extending the conceptual elegance of

a sensor design that integrates the selectivity of hybridization with an intrinsic signalling

mechanism for applications in solid-phase analysis, and as such, a molecular assembly was

developed for nucleic acid detection that is schematically shown in Figure 8.

Figure 8. Design of the fluorescent molecular switch hybridization monitoring construct, and schematic representation of the proposed molecular hybridization switch strategy. Site-directed templating of the biotinylated thiazole orange intercalating dye derivative places the dye in close proximity to a probe oligonucleotide using biotin/Neutravidin affinity chemistry. Upon the introduction of a nonlabeled target oligonucleotide, an enhancement of fluorescence intensity due to the intercalation of the thiazole orange moiety provides a signal that is indicative of a

ybridization event. h

 

The construct makes use of Neutravidin as a templating platform and offers the opportunity to

combine a biotinylated probe oligonucleotide with a nearby thiazole orange (TO) intercalating

dye that is also physically immobilized by means of a biotinylated linker. With the introduction

49

of Neutravidin as a “central hub”, it is possible to have this docking molecule function

simultaneously as an interface for the adjacent placement of the biotinylated probe and signaling

dye on a biotinylated linker as well as for surface immobilization through the use of the distal

biotin binding sites. Immobilization of the sensor constructs onto fused silica optical fibers

provides a substrate which can be used for evanescent wave excitation of fluorophores via total

internal reflection. The use of intercalating dye which has different binding affinities to ssDNA

and dsDNA provides fluorescence intensity and lifetime changes in different environments that

can be used to distinguish intercalation from other binding events. The work herein offers the

first report of the synthesis of a biotinylated intercalating dye species which incorporates a tether

of sufficient length (18-atom or 26-atom tether lengths were examined) to provide flexibility,

mobility and availability of the dye species when it is immobilized for solid-phase assays. By

first anchoring the biotinylated probe to Neutravidin, the immobilized oligonucleotide strand can

site-direct the assembly of the biotinylated linker that is capped with TO. The use of a single

signaling dye simplifies the design when compared to the dye/quencher pairs often encountered

in FRET technologies. The new sensor construct also provides a platform where potentially

hazardous intercalating dyes are permanently placed at an interface, and the handling of solutions

of such dyes is ameliorated. In addition, the synthetic method employed allows one to vary

tether lengths to provide custom-built intercalating dyes with a biotin functionality. The work

presented in this thesis has explored improvements to the overall design and surface assembly of

signaling dyes in proximity to immobilized probe oligonucleotides and provides some significant

advantages over previously existing work done with immobilization of tethered thiazole orange

oligonucleotide probes at an interface 102,103 The switching construct provides a facile method of

immobilization of probe and signaling dye at an interface that can provide a coating similar to

molecular self assembly for better control of surface immobilized components. The switch is

also modular and can be studied both in solution and at an interface. The dye is located adjacent

to the probe oligonucleotide sequence (as opposed to a tether at the probe terminus) and can

therefore provide flexibility in terms of binding location in a hybridized duplex as well as the

potential to decrease background signals that are associated with pi-stacking of tethered dyes

with the single-stranded probe nucleobases. The switch also incorporates Neutravidin which

provides a self-contained way to block non-specific adsorption at the interface. The use of

Neutravidin also provides a means of purification by binding only those components which are

biotinylated. The ability to “plug in” different components using biotin/avidin affinity binding

50

provides a simple method for potential multiplexing capabilities through binding of multiple

dyes with characteristic wavelengths.

Finally, Chapter 5 gives a detailed steady state and time-resolved fluorescence analysis of the use

of formamide as a means of concomitantly enhancing selectivity and signal generation upon

hybridization using biotinylated thiazole orange. This work reports a method to isothermally

provide single nucleotide polymorphism (SNP) detection using the denaturant properties of

formamide. It was also determined that formamide could provide enhanced signal-to-noise when

hybrids formed due to the perturbation of external binding modes of thiazole orange with the

polyanionic backbone of ssDNA and dsDNA. Although thiazole orange is capable of

distinguishing between ssDNA and dsDNA, this method provides improvement in the signal

magnitude of dsDNA over ssDNA, which is important in using thiazole orange as a fluorescence

signalling agent which can act as both an optical label and a probe for hybridization. Time-

resolved fluorescence methods were used to investigate the different fluorescence processes

contributing to the steady-state signals observed, and it was found that formamide alters the

relative contributions of differing binding modes associated with the dye. SNP discrimination

was also achieved for a variety of target lengths, including a 141 base-pair PCR amplicon at

room temperature.

51

2 Fluorescence Resonance Energy Transfer (FRET) for DNA Biosensors: FRET Pairs and Förster Distances for Various Dye–DNA Conjugates Analytica Chimica Acta 568 (2006) 181–189 ©2006 Elsevier

Abstract

Fluorescence resonance energy transfer (FRET) between the extrinsic dye labels Cyanine 3 (Cy3), Cyanine 5 (Cy5), Carboxytetramethylrhodamine (TAMRA), Iowa Black Fluorescence Quencher (IabFQ), and Iowa Black RQ (IabRQ) has been studied. The Förster distances for these FRET-pairs in single- and double-stranded DNA conjugates have been determined. In particular, it should be noted that the quantum yield of the donors Cy3 and TAMRA varies between single- and double-stranded DNA. While this alters the Förster distance for a donor–acceptor pair, this also allows for detection of thermal denaturation events with a single fluorophore. The utility of FRET in the development of nucleic acid biosensor technology is illustrated by using TAMRA and IabRQ as a FRET pair in selectivity experiments. The differential quenching of TAMRA fluorescence by IabRQ in solution has been used to discriminate between 0 and 3 base pair mismatches at 60 °C for a 19 base sequence. At room temperature, the quenching of TAMRA fluorescence was not an effective indicator of the degree of base pair mismatch. There appears to be a threshold of duplex stability at room temperature which occurs beyond two base pair mismatches and reverses the observed trend in TAMRA fluorescence prior to that degree of mismatch. When this experimental system is transferred to a glass surface through covalent coupling and organosilane chemistry, the observed trend in TAMRA fluorescence at room temperature is similar to that obtained in bulk solution, but without a threshold of duplex stability. In addition to quenching of fluorescence by FRET, it is believed that several other quenching mechanisms are occurring at the surface.

52

2.1 Introduction The field of biosensors and biochips for nucleic acid analysis has developed significantly over

the last decade. Sensing platforms are continuously being developed to provide rapid, reliable

detection of pathogens and genetic diseases. 185-189 The array of technology includes piezoelectric 56,187, electrochemical190,191, and optical methodologies. 189,192-194 In many cases the ultimate goal

is to create a sensitive, selective, and reusable field-deployable device for rapid diagnostics.

Current technology is largely limited by the compromise between achieving high sensitivity and

selectivity while maintaining the rapidity and reusability of the devices. Many device designs

have been investigated with the goal of creating simple, practical biosensor technologies. One

promising group of approaches is based on fluorescent signal evolution on optical transduction

elements using fluorescent labels. Though sensitive detection often results, such an approach also

tends to necessitate pre-labeling of the target sequence rather than the probe sequence. This

imparts considerable preparative work to the measurement protocol and takes away from the

concept of a true self-contained biosensor. Fluorescence strategies have been developed to

provide intrinsic signaling of nucleic acid hybrid formation. Fluorescence based biosensor

strategies are often very sensitive in that they can be designed to concurrently offer significant

signal while also minimizing background (noise). In particular, fluorescence resonance energy

transfer (FRET) has been used to develop analytical methods that are suitable for investigation of

the extent of selective binding by observation of associated proximity effects on fluorescence

intensity and lifetimes.

Molecular beacons, which rely on fluorescence resonance energy transfer (FRET) to provide

intrinsic signaling, may suffer from poor selectivity due to their short stem sequences.

Alternatively, unlabeled targets can be detected through the use of an intercalant dye tethered to

immobilized nucleic acid probes102, but at the cost of higher background fluorescence intensities.

FRET is a mechanism by which energy is transferred from an electronically excited donor

molecule to a ground-state acceptor via through-space resonant dipolar coupling. The efficiency

of this process is strongly distance dependent. Popular FRET-based strategies for nucleic acid

detection include: molecular beacons 5, duplex probes195, scorpion primers196, and fluorescence

polarization assays197. Several combinations of donor–acceptor pairs have been utilized,

53

including: two fluorophores5, a fluorophore with a dark absorber7,8, a fluorophore with an

intercalator9, and a fluorophore with a gold particle.82 One approach to development of optical

DNA biosensors that can achieve low fluorescence background is based on the use of FRET

pairs. If the acceptor and donor are in close proximity to one another, the acceptor can absorb the

excitation energy of the donor via FRET, so long as the donor–acceptor separation is within the

Förster distance.2 According to Förster theory, at donor–acceptor distances up to 70% of the

Förster distance, energy transfer efficiency remains above 90%. Similarly, beyond 140% of the

Förster distance, transfer efficiencies drop below 10%. If a biosensor is optimized to operate

between these ‘on/off’ distance extremes, large changes in acceptor or donor fluorescence should

allow comparable sensitivities to direct excitation of fluorophores. The enhanced red-shift

resulting from indirect excitation of a fluorescent acceptor via FRET may allow greater rejection

of background donor fluorescence and further increase sensitivity. Alternatively, the acceptor

can be a ‘dark absorber’ (such as the Iowa Black quenchers, IabFQ and IabRQ), in which case

the extinguished fluorescence of the donor is of analytical interest. In this context, dark absorbers

may find use in sensing applications for eliminating background signals in the absence of probe-

target hybrid formation. As part of this work, the FRET-related properties of Cy3 and TAMRA,

two commonly used dyes in DNA diagnostics, and Iowa Black dark quenchers were examined.

The large family of cyanine dyes have absorption and fluorescence profiles from 400 to 1000 nm

with solubility in a wide range of solvents.198,199 Reactive groups can easily be coupled to these

dyes such that covalent linkage to nucleic acids is feasible. Two of the more common cyanine

dyes used for nucleic acid studies are Cy3 and Cy5, which serve as high efficiency labels to

indicate the presence of a particular nucleic acid sequence. An extension of the polymethine

chain of Cy3 by two carbons gives the structure of Cy5 and shifts the fluorescence emission of

the dye from the green to the red. In particular, Cy3 covalently attached to the 5’ end of a nucleic

acid sequence will form a pi-stacked complex with the terminal base pair of the DNA duplex.200

The open chain of the polymethine bridge, however, allows the dye to exist in many

conformations which can result in a lowering of the quantum yield of fluorescence via non-

radiative energy loss.201

Rhodamine dyes are another well-established class of fluorophores. These dyes, which include

Rhodamine B, Texas Red, and Tetramethyl Rhodamine (TAMRA) amongst others, are highly

fluorescent in polar solvents. TAMRA and other rhodamine dyes are recognized to be

54

advantageous for their greater photostability relative to fluorescein, as well as their lack of a

strong pH-dependence.202 Nonetheless, these dyes remain highly sensitive to environment. For

example, TAMRA is quenched by charge transfer with guanosine in DNA203, while the lifetime

and quantum yield of TAMRA–DNA conjugates are highly sensitive to the local molecular

environment, as well as temperature and ionic strength.202,204

While the importance of selecting appropriate fluorescent labels for DNA in solution should be

apparent, the proper selection of dyes for labeling of surface-immobilized DNA raises new

constraints and concerns. Any fundamentally reusable nucleic acid biosensor uses surface bound

nucleic acid probes. Ultimately, if an immobilized duplex probe was to be used in a FRET sensor

configuration, the system could be custom-designed with a differing number and position of base

mismatches, as well as differing lengths of donor and acceptor labeled strand so that the

energetics could be optimized for development of fluorescence signal upon probe/target

recognition. Fluorescence and quenching events can then allow the simultaneous monitoring of

multiple probes using different fluorophores, and provide the capability to monitor

thermodynamic, kinetic and conformational parameters.195,205-210 However, working at an

interface invariably introduces complexities associated with surface free energy. To reduce

complexity, it is first necessary to understand how a system behaves in solution. It is then

possible to transfer the experiment to an interface and to begin to correlate any changes in

behaviour with interfacial properties.

2.2 Experimental Methods

2.2.1 Reagents

All chemicals were reagent grade or better and used without further purification. Reagent grade

toluene, ethanol, acetone, methanol, dichloromethane, and anhydrous diethyl ether were from

EM Science (Toronto, ON, Canada) and were dried by standard distillation methods where

necessary. Moisture content measurements of dry solvents were verified using an AquaStar®

Karl Fischer Titrator (EMD Chemicals Inc., Gibbstown, NJ, USA). Glass substrates were

25mm×375mm×31mm microscope slides from Fisher Scientific (Pittsburgh, PA, USA).

Hellmanex II glass cleaner was obtained from Hellma GmbH & Co. KG (Müllheim, Germany);

30% hydrogen peroxide, 28.0–30.0% ammonium hydroxide, and 35.6–38.0% hydrochloric acid

were obtained from EMD Chemicals (Gibbstown, NJ). Glass surface modification reagents,

55

including 98% 3-glycidoxypropyltrimethoxysilane and 99.5% N,N-diisopropylethylamine

(Hünig’s base), were also obtained from Sigma-Aldrich (Oakville, ON). Water was purified by

the Milli-Q cartridge system (Millipore Corporation, Mississauga, ON, Canada). Water that was

used in the preparation of buffered oligonucleotide was double-distilled and autoclaved.

2.2.2 Instruments

Ultraviolet–visible absorption spectra were measured using a Bichrom Ltd. (Cambridge, UK)

Libra S22 spectrometer and a Hewlett Packard 8452A Diode-Array Spectrometer (Hewlett

Packard Corporation, Palo Alto, CA, USA). Solution phase fluorescence spectra were measured

using a QuantaMaster PTI Spectrofluorimeter and Felix Software (Photon Technology

International, Lawrenceville, NJ, USA). Fluorescence characterization of experiments on glass

surfaces was done using a Versarray Chipreader 5 μm System (Bio-Rad, Hercules, California,

USA).

2.2.3 Dye labeled oligonucleotides

The oligonucleotide sequences in Table 4 were obtained from Integrated DNA Technologies

(Coralville, IA, USA) and dissolved in 1× PBS buffer at pH 7.0. The base sequence corresponds

to the SMN1 gene fragment used in oligonucleotide diagnostics for spinal muscular atrophy.211

All subsequent dilutions were prepared with 1× PBS buffer. Solutions containing a 1:1 ratio of

probeA/B and targetn oligonucleotides were heated at 95 °C for 5 min and let cool to room

temperature to generate dsDNA.

2.2.4 Oligonucleotide immobilization

Glass substrates were sonicated in glass cleaner for 30 min then treated, successively, with 5:1:1

v/v water: ammonium hydroxide: hydrogen peroxide and 5:1:1 v/v water: hydrochloric acid:

hydrogen peroxide, for ten minutes each, at 80 °C. The substrates were then sonicated twice in

methanol for 15 min, rinsed successively with dichloromethane and diethyl ether, and stored in

an oven at 125 °C until further use. This protocol has been described previously212, and creates

reactive silanol moieties while removing organic and inorganic contaminants. Cleaned and dried

substrates were immersed in a solution of 250 mL refluxing dry toluene (≤25 ppm water), 80 mL

GOPS, and 2.4 mL Hünig’s base, under argon for 24 h. The substrates were recovered and

washed with organic solvent as described for the previous step. An extensive characterization of

56

the GOPS modified surface has recently been published.213 GOPS modified glass substrates

were spotted with oligonucleotide solution (ProbeB in Table 4) using an automated pinspotter.

Spotted substrates were stored in a humid environment then rinsed twice each with 1× TRIS–

SDS buffer and once with distilled water. Table 4. Oligonucleotide sequences and labels used in FRET experiments.

List of Sequences ProbeA 5’- ATT TTG TCT GAA ACC CTG T-Cy3/TAMRA-3’

ProbeB 5’-NH2C12H24-ATT TTG TCT GAA ACC CTG T-TAMRA-3’

Target0 5’-Cy5/IabRQ/IabFQ/no label-ACA GGG TTT CAG ACA AAA T-3’

Target1 5’-IabRQ-- ACA GGG TTA CAG ACA AAA T-3’

Target2 5’-IabRQ-- ACA GGG TTA GAG ACA AAA T-3’

Target3 5’-IabRQ-- ACA GGG TTA GCG ACA AAA T-3’

Abbreviations

Cy3 N,N’-(diisopropyl)-tetramethylindocarbocyanine

Cy5 N,N’-(diisopropyl)-tetramethylindodicarbocyanine

TAMRA Carboxytetramethylrhodamine

IabRQ Iowa Black RQ

IabFQ Iowa Black FQ

ProbeA was used in solution-phase experiments; probeB was immobilized on a glass substrate for experiments at an interface. Targetn sequences were hybridized to the probes, where n represents the number of base pair mismatches.

2.2.5 Spectral data and Förster distance calculations

The donor–acceptor distance–efficiency relationship is given by Eq. (20), where Ro is the Förster

distance, r is the donor–acceptor separation, and E is the energy transfer efficiency. The Förster

57

distance (Eq. (21)) is a characteristic of a donor–acceptor pair, and depends on factors including

refractive index of the surrounding medium, n, the donor quantum yield, ΦD, the relative

orientation between donor emission and acceptor absorption dipoles, and the degree of spectral

resonance between the two species.214 These latter two parameters are described by the

orientation factor, κ2, and spectral overlap integral, J, respectively. The spectral overlap integral

(Eq. (22)) is a function of the fluorescence intensity of the donor, FD, and molar absorptivity of

acceptor, εA, as a function of wavelength, λ, normalized against the total donor emission.214 In

the case of donor and acceptor molecules that are free to sample all orientations, the orientation

factor takes on a value of κ2 = 2/3.

  

 

8.79 10 Φ  

(20)

(21)

 (22)

ΦΦ  

 

(23)

Ultraviolet–visible absorption and fluorescence emission spectra were obtained with 3 and 0.5

μM solutions of oligonucleotide, respectively. The measured sequences used were ProbeA for

ssDNA, and ProbeA/Target0 (no label) for dsDNA. It is important to note that the donor and

acceptor properties should be measured separately since any FRET between them will modulate

these properties. From the spectra obtained, the integrands in Eq. (22) were calculated at 0.5 nm

increments and integrated numerically to determine the spectral overlap. The refractive index

and orientation factor terms in Eq. (21) were taken as n = 1.43 and κ2 = 2/3, corresponding to the

58

refractive index of buffer and non-restricted dye motion, respectively. Quantum yield values for

fluorescent dyes were determined relative to fluorescein dye in sodium borate buffer fixed at pH

9.5. The quantum yield of fluorescein under these conditions is known to be 0.93215 and the

quantum yield, Φ, of other dyes were determined as a ratio (Eq. (23)) of their integrated

emission, Fdλ, corrected for different molar extinction coefficients, ε, at the wavelength of

excitation. Fluorescein was excited at 490 nm.

2.2.6 Selectivity using the TAMRA–IabRQ FRET pair in solution

Fluorescence intensity measurements were recorded at both 24.5 °C and 60 °C (five degrees

below the calculated thermal denaturation temperature, Tm, for the perfectly matched SMN1

probe/target duplex). Sequences studied were TAMRA ssDNA, TAMRA dsDNA (no quencher),

TAMRA/IabRQ dsDNA (fully complementary), and TAMRA/IabRQ dsDNA containing 1, 2,

and 3 centrally-located base pair mismatches corresponding to Probe0, Probe0/Target0,

Probe0/Target0, Probe0/Target1, Probe0/Target2, Probe0/Target3, respectively according to the

sequences outlined in Table 4.

2.2.7 Selectivity using the TAMRA–IabRQ FRET pair on glass substrates

Double-stranded DNA was hybridized in solution according to the protocol outlined in Section

2.2.3 prior to immobilization or alternatively was hybridized on slide. All surface studies

employed pre-hybridization in solution except for the study involving partial regeneration, which

was hybridized on-slide. For on-slide hybridization, eight 1 μL spots of 1 μM target solutions

were spotted on the slide surface. The slide surface was subsequently covered with microscope

cover slips and placed in a humid hybridization chamber for 24 h. The slides were subsequently

rinsed twice with 1× TRIS/SDS buffer and once with sterile water. For regeneration studies on

probe-functionalized substrates, the slides were sonicated in sterile water for 15 min, dried, and

subsequently scanned to record fluorescence intensity.

2.3 Results and Discussion

2.3.1 Förster distances in single and double stranded DNA

The Förster distance is a critical parameter in experiments which utilize FRET to determine

donor–acceptor distances, particularly with respect to transduction of conformational changes or

59

Table 5. Fluorescence and energy transfer data calculated from experimental data for various donor-acceptor pairs.

Donor Excitation Acceptor ssDNA dsDNA wavelength

(nm)

Donor quantum yield, Φ

Spectral overlap, J

(10−10 cm6)

Förster distance, Ro (nm)

Donor quantum yield, Φ

Spectral overlap, J

(10−10 cm6)

Förster distance, Ro (nm)

Cy3 520 Cy3 0.070±0.003 4.2± 0.2 4.2 ± 0.3 0.042 ± 0.002 4.1 ± 0.3 3.8 ± 0.4 TAMRA 4.9 ± 0.2 4.3 ± 0.3 4.9 ± 0.4 4.0 ± 0.4 IabFQ 2.7± 0.1 3.9 ± 0.3 2.5 ± 0.2 3.5 ± 0.4 Cy5 3.0± 0.2 4.0 ± 0.3 3.3 ± 0.4 3.7 ± 0.5 IabRQ 0.95 ± 0.06 3.3 ± 0.2 0.96 ± 0.20 3.0 ± 0.6

TAMRA 556 TAMRA 0.076±0.014 2.7 ± 0.7 3.9 ± 1.3 0.098 ± 0.006 2.6 ± 0.2 4.1 ± 0.4 IabFQ 2.6 ± 0.7 3.9 ± 1.3 2.3 ± 0.2 4.0 ± 0.4 Cy5 4.9 ± 1.3 4.4 ± 1.4 5.5 ± 0.7 4.6 ± 0.7 IabRQ 1.2 ± 0.3 3.5 ± 1.1 1.3 ± 0.2 3.6 ± 0.7

The donor quantum yields and spectral overlap integral for each pair are given separately to allow calculation of the Förster distance for any index of refraction and orientation data. The listed Förster distance employs values of 1.43 and 2/3 for the refractive index and orientation factor respectively. Sequences used were ProbeA for ssDNA, and ProbeA/Target0 (no label) for dsDNA (0.5μM). All values were calculated using the equations outlined in Section 2.2.5. Spectral properties of dyes; Cy3:λabs max = 520 nm (shoulder), 550 nm (peak); λem max = 564 nm; TAMRA: λabs max = 556 nm; λem max = 577 nm; IabFQ: λabs max = 420–620 nm (range); 531 nm (peak); λem max = none; Cy5: λabs max = 605 nm (shoulder), 648 nm (peak); λem max = 668 nm; IabRQ: λabs max = 500–700 nm (range); 656 (peak); λem max = none.

dynamics. In these instances, the interest is in sensitizing acceptor fluorescence (e.g. Cy3–Cy5

pair) or decreasing donor fluorescence (e.g. via a quencher such as IabRQ). Therefore, in

addition to dye properties such as brightness, peak emission wavelength, lifetime, or

photostability, it is also necessary to choose a donor–acceptor pair with an appropriate Förster

distance to maximize the energy transfer behaviour of interest. In nucleic acid diagnostics, the

detection of hybridization events is paramount and, as discussed previously, FRET is a popular

strategy employed in this regard. A list of experimentally determined Förster distances for

various donor–acceptor pairs as DNA conjugates is presented in Table 5.

It is important to note that the Förster distance is dependent on whether the DNA is single

stranded or double stranded. This is thought to be primarily due to changes in the quantum yield

of the donor dyes Cy3 and TAMRA used in these experiments. Variations in Förster distance

could also be due to changes in dye orientation (κ2) since relative orientations of dyes tethered to

biomolecules such as proteins and DNA are not completely randomized.216 However, the widely

accepted approximation of κ2 = 2/3 was used for the FRET values calculated in this work, and

literature reports state that rough approximations by using κ2 = 2/3 over multiple measurements

still yields useful FRET information.216

60

The quantum yield of Cy3 decreases 40% going from ssDNA to dsDNA, and the quantum yield

of TAMRA increases 28% going from ssDNA to dsDNA. The associated emission and

absorption profiles are shown in Figure 9.

Figure 9. Fluorescence emission (0.5 μM) and absorption profiles (3.0 μM) for ssDNA and dsDNA conjugates of Cy3 (a and c) and TAMRA (b and d). In each case the dye is conjugated to the ProbeA sequence. Note that the quantum of yield of Cy3 conjugated to dsDNA is 40% lower, while the quantum yield of TAMRA is 28% higher. The absorption profiles show no significant difference between ssDNA and dsDNA conjugates. Note that the dsDNA samples have twice as much nucleic acid material as the ssDNA samples of equal concentration.

The absorption profiles, and thus the molar absorptivity coefficients, are essentially consistent

between ssDNA and dsDNA. This is also true for the other dyes used in this work. As a

consequence of this observed change in quantum yield, the Förster distance for Cy3–acceptor

61

pairs is larger in ssDNA than dsDNA, and vice versa for TAMRA–acceptor pairs. This is

relevant in FRET studies of hybridization dynamics, wherein it may be necessary to compensate

for a shift in the donor–acceptor Förster distance as the duplex forms. With respect to

diagnostics, it is possible to observe hybridization solely on the basis of the change in quantum

yield. As an example, Figure 10 shows a melt curve in which the relative fluorescence observed

from Cy3 increases as the DNA duplex dissociates.

Figure 10. Melt curve for the Cy3–dsDNA conjugate (ProbeA–Target0). As the temperature increases, the dsDNA melts to ssDNA and Cy3 undergoes a 40% increase in quantum yield. The melt curve is corrected for the change in quantum yield with temperature.

Literature reports about the quantum yield of Cy3–DNA conjugates vary. For example,

Sabanayagam et al.217 report a value of 0.04 for Cy3–dsDNA conjugate—in agreement with

these results—while others report values ranging up to 0.24.218-220 Perhaps more importantly, the

quantum yield has been demonstrated to be a function of base sequence221, stressing the

importance of experimental determination. Data about the quantum yield of TAMRA–DNA

conjugates are not widely available in the literature. A value of 0.10–0.11 has been reported for a

62

TAMRA labeled aptamer.202 This value is very close to the value determined for TAMRA–

dsDNA conjugate. The large uncertainty associated with the TAMRA–acceptor Förster distance

is due to the corresponding uncertainty in TAMRA quantum yield. It is possible that the differing

magnitudes of uncertainty between dyes arise from the different environmental sensitivities of

TAMRA ssDNA conjugates, TAMRA dsDNA conjugates, and Cy3 DNA conjugates.

2.3.2 TAMRA/IabRQ selectivity studies in solution

The selectivity of oligonucleotide hybridization can be significantly different at an interface than

in bulk solution, and factors such as temperature, ionic strength, and pH can be tuned to enhance

the selectivity at an interface.211 The enhancement of selectivity at an interface can sometimes be

more effective than under similar conditions in bulk solution.211 Features such as thermal

denaturation temperature tend to show similar trends when comparing bulk solution and

interfacial environments, and exhibit similar sensitivities to ionic strength.211 Nonetheless, in

most cases, overall selectivity is best for hybridization that takes place in bulk solution. It is

therefore essential to study any new system in solution in order to investigate the factors which

affect the selectivity and fluorescence signal. The fluorescence intensity changes of several

nucleic acid sequences were investigated in this work. The intention was to examine whether

fluorescence intensity changes of the TAMRA/IabRQ FRET pair system could be seen upon

hybridization, and to determine if selectivity differences between perfectly matched duplexes and

those duplexes which contained one, two, and three centrally-located mismatches could be

established. The fluorescence intensity of the TAMRA/IabRQ duplexes as a function of base pair

mismatch (from 0 to 3 centrally located base-pair mismatches) were measured at both 24.5 °C

(room temperature) and 60 °C (5 °C lower than the Tm of the perfectly matched duplex) in order

to establish whether base mismatch discrimination could be observed based on fluorescence

intensity changes imparted by the number of base-pair mismatches present in the duplex. The

data is shown in Figure 11 at these two temperatures for 0.5 μM solutions. Analogous trends

were observed with other concentrations ranging from 0.1 to 1.0 μM (data not shown). The peak

emission intensity for TAMRA in the perfectly matched duplex at 60 °C is 20% larger than at

24.5 °C. Since quantum yield decreases with both increasing temperature and denaturation of

TAMRA labeled oligonucleotides, this demonstrates that a significant proportion of duplex has

denatured. As a consequence, there is a large population of fluorophores and quenchers that are

no longer in close proximity, precluding the possibility of FRET. Therefore, the quencher is

63

unable to quench the fluorescence emission of the TAMRA, and the resulting increase in

TAMRA emission exceeds the decrease in quantum yield. In oligomers consisting of 14–20

nucleotides, each base pair mismatch can reduce the observed Tm by as much as 5 °C.211 This is

intuitive since introducing a base pair mismatch to a nucleic acid hybrid system decreases the

overall stability of the duplex. It is expected that an increasing number of mismatches would lead

to decreasing duplex stability and an increase in TAMRA emission as the number of base pair

mismatches increases. Therefore, near the Tm for the fully matched duplex, the majority of the

population of the original duplexes containing three base pair mismatches will be denatured.

While the relative fluorescence intensities of perfectly matched, one, two, and three base pair

mismatches at 60 °C follows the expected trend of decreased quenching with increasing

mismatch, the data obtained at 24.5 °C is more interesting. The following trend in fluorescence

intensity of the TAMRA/IabRQ system was observed: perfectly matched > 1 bpm > 2 bpm. It is

proposed that this decrease in fluorescence intensity does not lie in more efficient quenching of

the fluorophore via FRET, but quenching by other processes and environmental factors. When

TAMRA is associated with ssDNA, it has a decreased quantum yield relative to dsDNA. If the

rigid environment of the dsDNA protects the TAMRA species from bulk solution when the dye

is attached to the probe oligonucleotide via a short tether, this may lead to a reduction of

collisional deactivation and protection from quenching effects that are sometimes associated with

the ions present in the buffer. In the fully matched duplex, TAMRA is well protected from the

bulk solution due to the charge-screened environment associated with the dsDNA. With respect

to duplexes with one and two base pair mismatches, the rigid rod conformation of the fully

matched duplex is compromised by the presence of the mismatches, yielding an increased

tendency for collisional deactivation. Nonetheless, at this level of instability, it is thought that the

TAMRA and IabRQ moieties remain within the Förster distance, thus resulting in the above

trend. In contrast, the fluorescence intensity associated with the three base pair mismatch breaks

the trend. Breathing effects are known to occur in oligonucleotide duplexes, which would result

in a time-dependent distance changes between the fluorophore and quencher, resulting in an

ensemble average equivalent to an increased steady-state fluorescence intensity for a system

which ventures outside the Förster distance. For the one and two base pair mismatch systems, the

complementarity of the bases near the termini are identical, and minimal breathing occurs at the

termini. However, when three base-pair mismatches are centrally located within the

TAMRA/IabRQ duplex, there appears to be enough destabilization to produce a population of

64

DNA strands wherein the dye and quencher do not lie within the Förster distance. The hypothesis

is that this generates a fluorescence enhancement that more than compensates for the increased

rate of collisional deactivion due to duplex instability, thus leading to a break in the trend. This

suggests a limitation to the analytical usage of the dye-quencher FRET strategy at room

temperature. A stability threshold may exist where the trend between different degrees of

mismatch reverses. As observed in Figure 11 (left), the signal obtained for fully complementary

and three base pair mismatch systems is indistinguishable.

Figure 11. TAMRA fluorescence intensity increase as a function of the degree of base pair mismatch at 24.5°C and 60 °C in ProbeB–Target0–3 duplexes. Please note the difference in maximum fluorescence on the y-axis (2.5×105 for TAMRA/IabRQ system at 24.5 °C vs. 2.5×106 for TAMRA/IabRQ system at 60°C). Measurements were performed in triplicate and error bars represent 1 σ.

In addition, the changes in signal up to two base pair mismatches is minimal and does not

suggest a high degree of discrimination. In contrast to the data obtained at room temperature, the

65

data obtained at 60 °C indicates that discrimination between fully complementary, one, two, or

three base pair mismatches may be possible. This is particularly true between one and two base

pair mismatches where there is four-fold increase in signal. Furthermore, there is no apparent

reversal of the observed trend as is found at room temperature. This suggests that a FRET

strategy for the detection of single nucleotide polymorphisms still requires optimization of

temperature and other experimental conditions affecting stringency.

2.3.3 TAMRA/IABRQ selectivity studies on glass slide substrates

TAMRA/IabRQ investigations on surfaces were undertaken in an effort to determine if FRET

between these two moieties at an interface behaves similarly to that in solution. An interface is a

necessary component of a reusable biosensor, and FRET at an interface is important in the future

development of a low-background biosensor in which labeled probe material eliminates the

requirement for labeled target material. Ferguson et al.222 and Perez-Luna et al.223 have shown

that one major limitation in the currently available strategies using optically-based biosensors is

that non-specific binding/adsorptive effects can lead to false positive signal generation and high

background noise. In contrast to singly-labeled interfacial probes which produce a signal in

response to both the presence of non-complementary sequences and absence of target material 102, a sensor configuration which employs a donor–‘dark’ acceptor FRET pair would ideally

allow the quenching of the background fluorescence associated with the labeled probe sequence.

The TAMRA donor was placed at the 3’-end of the probe sequence, which was tethered to

surface via its 5’-end. The IabRQ acceptor was placed at the 5’-end of the target so that donor

and acceptor were adjacent and well within the Förster distance. Figure 12 shows fluorescence

intensity data for various surface immobilized probe-target hybrids.

66

Figure 12. Surface immobilized TAMRA labeled probeB sequences, including: 1.0 μM TAMRA ssDNA, 1.0 μM TAMRA dsDNA (no quencher; target0) 1.0 μM TAMRA/IabRQ dsDNA (target0); and 0.5 μM samples of same series. Bars on graphs correspond to row of spots in the adjacent fluorescence image. Measurements were performed in triplicate and error bars represent 1 σ.

Comparison of the 0.5μM TAMRA and TAMRA/IabRQ dsDNA hybrids shows that TAMRA

fluorescence was quenched in the presence of IabRQ. However, this alone does not establish a

FRET mechanism of energy transfer between the TAMRA and the IabRQ FRET-pair. TAMRA

is known to have self-quenching properties both as a function of being at a solid interface as well

as properties inherent to the molecule itself.224 TAMRA and other rhodamine derivatives possess

self-quenching properties resulting from homo resonance energy transfer (homoRET). TAMRA

molecules can thus act as non-fluorescent traps or ‘sinks’ for the excited state energy of other

TAMRA molecules.225 This phenomenon arises from the relatively small Stokes shift associated

with TAMRA (λex 556 nm, λem 577 nm found in solution), and thus offers highly favourable

spectral overlap.226 It has also been determined that fluorophores can undergo aggregation self-

quenching at surfaces. Aggregation can cause the formation of a ground-state non-fluorescent

complex227, leading to decreased fluorescence signals associated with surface bound nucleic acid

probes. TAMRA can also undergo fluorescence quenching in the presence of dsDNA containing

G residues in close proximity to the TAMRA. This occurs via a photo induced electron transfer

between the dye and G residues where the nucleobase is oxidized and the dye is reduced.203 The

decreased fluorescence intensity of the 1.0 μM spots with respect to the 0.5 μM is likely due to

67

one or more of the quenching effects described above. The greater intra-strand interactions and

conformational changes at higher immobilization densities result in different degrees of

quenching. Results have shown that decreasing or non-linear increases in fluorescence intensity

with increasing spotting concentrations of TAMRA above 1.0 μM is commonplace. Furthermore,

the greater fluorescence observed with TAMRA–ssDNA conjugates relative to TAMRA–dsDNA

conjugates on a surface—which is opposite to what is observed in solution—also appears to be a

result of surface specific interactions and quenching mechanisms.

One of the fundamental measures of nucleic acid biosensor performance is the ability to detect

single nucleotide polymorphisms. For example, base pair mismatch discrimination is of the

utmost importance when dealing with diseases such as Spinal Muscular Atrophy (SMA),

wherein the difference between healthy and unhealthy individuals rests in a signal base

mutation.211 Figure 13 shows preliminary work towards determining mismatch discrimination

with respect to surface-bound probe sequences labeled with TAMRA.

Figure 13. Surface TAMRA/IabRQ FRET analysis as a function of the number of base-pair mismatches present in the target (1 μM ProbeB TAMRA ssDNA and 1μM ProbeB/Target0–3 dsDNA hybrids). Measurements were performed in triplicate and error bars represent 1 σ.

68

Double-stranded DNA hybrids were pre-hybridized in solution and then immobilized as ‘rigid

rods’ of dsDNA through the linker modification of the probe sequence. Again, the fluorescence

intensity data shows that there is surface quenching associated with the formation of dsDNA at

the surface. Additional quenching is observed with the introduction of IabRQ on the target

sequence. It is interesting to note that the TAMRA/IabRQ dsDNA mismatch data is very similar

to the trend observed for the analogous solution-phase experiment. While the three base pair

mismatch hybrid in solution did not previously follow the trend of decreasing intensity with

increasing mismatch, it can be seen in Figure 13 that the three base pair mismatch shows no

statistically significant deviation from the trend at an interface. This suggests that solution phase

experiments are a reasonable predictor of interfacial behaviour, but that the surface energetics

remain different from those in solution. On-slide hybridization with subsequent regeneration of

single-stranded probe as a function of base pair mismatch was also investigated using the

TAMRA/IabRQ FRET pair. Such an experiment has an obvious interpretation in terms of the

reusability of a biosensor. Sonication of slides with immobilized duplexes in sterile water of zero

ionic strength was employed to generate partial denaturation. This mild method of regeneration

was chosen to ensure that a significant portion of duplex remained. A more vigorous

regeneration process would be expected to yield nearly complete recovery and differentiating

between the number of mismatches would be impossible. As shown in Figure 14, the increasing

fluorescence intensity with increasing number of mismatches confirms that the trend in duplex

stability as a function of mismatch can be reproduced at an interface via the TAMRA/IabRQ

FRET-pair.

69

Figure 14. Fluorescent signal regeneration as a function of the degree of base pair mismatching for the TAMRA/IabRQ FRET pair at the glass surface (1 μM ProbeB TAMRA ssDNA and 1 μM ProbeB/Target1–3 dsDNA hybrids). Measurements were performed in triplicate and error bars represent 1 σ.

2.4 Conclusions The Förster distances, Ro, of several donor–acceptor permutations for the dyes Cy3, Cy5,

TAMRA, IabFQ, and IabRQ have been experimentally determined for both ssDNA and dsDNA

conjugates. With Cy3 and TAMRA donors, the Förster distance varies between single and

double stranded DNA, largely due to changes in donor quantum yield. In the case of Cy3, there

is an increase in quantum yield in going from dsDNA to ssDNA, and with TAMRA there is a

corresponding decrease. It has been demonstrated that this change in quantum yield can be used

to detect the thermal denaturation of dsDNA. With respect to selectivity, the TAMRA/IabRQ

FRET pair was found to effectively discriminate between perfectly matched, one, two, and three

base pair mismatches at 60 °C in solution. In contrast, at room temperature, it is believed that the

fluorescence as a function of base pair mismatch indicates a threshold of duplex stability

between two and three base pair mismatches which renders FRET discrimination of base pair

mismatches unreliable. Surface studies incorporating the TAMRA/IabRQ FRET pair suggest that

70

there are several modes of quenching when TAMRA is placed at an interface. This phenomenon

adds a level of complexity to transduction strategies in biosensor applications since quenching of

TAMRA emission cannot be solely attributed to FRET. Nonetheless, it has been shown that

TAMRA is still quenched by the presence of IabRQ, yielding trends (at room temperature) with

respect to the degree of base pair mismatching present that are essentially consistent with those

observed in solution. Regeneration of the TAMRA labeled probes is possible with sonication of

the solid substrate in sterile water at room temperature. The findings presented in this work can

all be applied to the design and engineering of nucleic acid biosensor platforms based on

fluorescent modes of detection. A significant number of biosensor designs do not consider the

sensitivity of a fluorophore to its immediate environment, or its physical interactions in that

environment. Fluorophore choice is therefore of utmost importance in the design of a sensor

platform.

This work has investigated the potential for FRET based discrimination of 0–3 base pair

mismatches at room temperature—both in solution and at an interface relevant to fiber-optic

biosensor technology. The interest here is in the selectivity of the system rather than the

sensitivity. The data obtained is a first step towards developing a surface-based FRET strategy

for signaling of nucleic acid hybridization events. At this stage, the emphasis is on understanding

how environmental changes associated with immobilization and hybridization of DNA affects

the properties and behaviour of fluorescent dyes.

71

3 Fluorescence Resonance Energy Transfer and Complex Formation Between Thiazole Orange and Various Dye-DNA Conjugates: Implications in Signaling Nucleic Acid Hybridization Journal of Fluorescence 16 (2006) 555-567 ©Springer Science and Business Media, Inc. 2006

Abstract

Fluorescence resonance energy transfer (FRET) was investigated between the intercalating dye thiazole orange (TO), and the dyes Cyanine 3 (Cy3), Cyanine 5 (Cy5), Carboxytetramethyl Rhodamine (TAMRA), Iowa Black FQ (IabFQ), and Iowa BlackRQ (IabRQ), which were covalently immobilized at the end of dsDNA oligonucleotides. In addition to determining that TO was an effective energy donor, FRET efficiency data obtained from fluorescence lifetime measurements indicated that TO intercalated near the middle of the 19mer oligonucleotide sequence that was used in this study. Discrepancies in FRET efficiencies obtained from intensity and lifetime measurements led to the investigation of non-fluorescent complex formation between TAMRA and modified TO. The hydrophobicity of TO was modified by the addition of either an alkyl or polyethylene glycol (PEG) side-chain to study effects of dimer and aggregate formation. It was found that at stoichiometric excesses of modified TO, fluorescence quenching of TAMRA was observed, and that this could be correlated to the hydrophobicity of a TO chain species. The TAMRA:TO-chain association constant for the TO-alkyl system was 0.043±0.002 M−1, while that obtained for the TO-PEG was 0.037±0.002 M−1. From the perspective of method development for the transduction of hybridization events, the work presented evaluates a variety of schemes based on energy transfer between TO and an acceptor dye, and discusses the implications of complex formation in such schemes.

72

3.1 Introduction The work in Chapter 2 introduced FRET-based mechanisms of transduction for extrinsic labels

both in solution and at an interface. This Chapter represents the investigation of FRET-based

transduction strategies in which an intercalating dye that can interact with fluorophore or

quencher end-labelled DNA could signal hybridization events. In some of the strategies

presented, labeling of target is also avoided. While ethidium bromide (EB) is one of the best

known intercalators used in the detection of dsDNA,228 it is just one of many known fluorescent

intercalating dyes, and does not offer optimal analytical performance as a marker for detection of

dsDNA structure due to its relatively low fluorescence enhancement upon binding and relatively

low molar absorptivity. Thiazole orange (TO) is another dye that has been used as an

intercalative transduction agent in nucleic acid hybridization assays. TO is a non-planar

chromophore composed of a benzothiazole derivative and a quinolinium ring linked via a

monomethine bridge. TO has been reported to provide anywhere from a 50- to a 18 900-fold

fluorescence enhancement upon dsDNA binding.81,84 Regardless of whether the enhancement is

in the upper or lower limit of this range, it is a substantial improvement in comparison to the

emission from EB.79,80 It is worth noting that differing experimental conditions including

nucleotide sequence84,229 can influence the degree of fluorescence enhancement that is observed.

The increase in quantum efficiency of TO upon intercalation results from the restriction of

rotation around the monomethine bridge connecting the benzothiazole and quinolinium

heterocycles of the dye. TO is weakly fluorescent in solution and highly fluorescent in viscous

or rigid media, making it an excellent choice as an intercalator that can indicate DNA

hybridization. Selective transduction of dsDNA structure by use of a fluorescent intercalating

dye is highly advantageous in nucleic acid biosensor design in comparison to an indirect method

that only indicates the presence of a fluorescently labelled target. An intercalating dye such as

TO is necessary in this respect due to its strong dependence on stable hybrid formation and low

fluorescence in the presence of single-stranded DNA (ssDNA). However, intercalating dye not

associated with dsDNA still has a finite quantum yield. It may be possible to further decrease

background signal and thereby enhance signal-to-noise in the presence of ssDNA by relying on a

FRET strategy. The presence of a fluorophore or quencher labelled sequence offers the

possibility of two sensitivity enhancement strategies: (1) incorporation of a quencher labelled

sequence to effectively eliminate background fluorescence; or (2) incorporation of a fluorescent

73

acceptor to allow for ratiometric quantification and greater confidence in observed signals. In

this work, the preliminary stages of development of an intercalator-based FRET strategy for

signalling nucleic acid hybridization events is presented. While the work in this chapter does not

report a method in itself, it establishes the first step in developing a methodology ultimately

intended for use in a nucleic acid biosensor. The interest lies in characterizing the energy

transfer, physical behaviour, and interactions between the dye species involved in the FRET-

based transduction strategies proposed. In light of this objective, a minimalist approach is taken

to the systems studied. For example, monomeric TO dye modified with an alkyl or a

polyethylene glycol (PEG) side-chain as a model system has been investigated. In this work, the

side-chain modification allows control over hydrophobicity, but more generally allows the

monomeric TO to be covalently tethered to a probe oligonucleotide, as previously described and

characterized by our group102,103, or to another linking moiety (such as biotin) to create a

switching-type construct. For the work in this particular Chapter, this latter feature has not yet

been exploited so as to isolate the physical behaviour of the dye molecules, without restrictions

and perturbations of the tethering process superimposed. The study remains of practical

importance because the use of a tethered intercalator, though advantageous, is neither

commonplace nor required for effective biosensor design. Indeed, this work determines that the

spectroscopy of the TO-FRET systems that are studied are complicated by the potential for

hydrophobicity and stoichiometrically driven non-fluorescent complex formation with an

acceptor dye. In addition to characterizing complex formation, this work demonstrates energy

transfer from TO to various dye-dsDNA conjugates and the corresponding Förster distances have

been determined. A second interesting aspect of this work is the use of monomeric TO which,

when compared to the array of work using dimeric TO derivatives, [for example83,230-232], is

much less studied. Monomeric TO is advantageous from a biosensor perspective because it

shows much less sequence selectivity for intercalation than dimeric TOTO, and also because the

TO-TO bridge in the dimer occupies the site most effectively used for attachment of a tether.100

In addition, TOTO binds less reversibly to dsDNA which may hinder the use of TOTO in a

reusable biosensor construct, while also exhibiting a smaller dsDNA/ssDNA fluorescence

ratio.94,228 Although FRET between dimeric TOTO species and many other heterodimeric

intercalators has been reported 85,231,233-239, it is believed that this report is the first instance of

FRET between an intercalator and a distinct, extrinsic fluorophore (in this work extrinsic refers

to any non-intercalating dye conjugated to either terminus of an oligonucleotide).

74

3.2 Experimental Methods

3.2.1 Reagents

All chemicals were reagent grade or better and used without further purification. Reagents used

for thiazole orange synthesis (methyl iodide, 2-(methylthio)benzothiazole, 4-methylquinoline,

triethylamine, triethylene glycol monochlorohydrin, sodium iodide) were from the Aldrich

Chemical Company (Milwaukee, WI, USA). Oligonucleotide solutions were prepared in 1× PBS

buffer; buffers were prepared with double-distilled water and subsequently autoclaved. Buffer

salts (sodium chloride, dibasic sodium phosphate) were obtained from Aldrich. Water was

deionized and purified using a Milli-Q cartridge purification system (Millipore Corporation,

Mississauga, ON, Canada) and used in octanol–water partitioning experiments, and 99% octanol

was obtained from Sigma (Oakville, ON, Canada). The synthesis of thiazole orange with various

side-chains was based upon the methods of Brooker et al.240,241 and Carreon et al.97. The

quinolinium compounds were synthesized by treating 4-methylquinoline (lepidine) with three

equivalents of the appropriate iodo-functionalized side chain in refluxing toluene, and the

resulting precipitate was filtered and washed with anhydrous ether. The benzothiazole derivative

was synthesized by treating 2-(methylthio)benzothiazole with three equivalents of iodomethane

in refluxing ethanol. The product was precipitated from solution with anhydrous ether, filtered,

and washed once again with ether. This was followed by solvent extraction in CH2Cl2/H2O, and

the aqueous layer was removed under reduced pressure to yield the desired product. The pale

yellow precipitate was subsequently recrystallized from acetone. The side-chain functionalized

quinolinium compounds were then condensed with the benzothiazole derivative in the presence

of triethylamine in ethanol to give the dye species. The dye derivatives were recrystallized from

MeOH/H2O. This species is referred to as TO-alkyl. For the triethylene-glycol side-chain

quinolinium derivative, triethylene glycol monochlorohydrin was converted to triethylene glycol

monoiodohydrin according to the method of Koizumi et al.242 (not shown). This species is

referred to as TO-PEG. The synthetic schemes for both the TO-PEG and TO-alkyl dyes are

shown in Figure 15 (see Appendix I for synthetic and characterization details).

75

Figure 15. Synthetic scheme for the preparation of thiazole orange dye and the attachment of different side-chains.

76

3.2.2 Dye labelled oligonucleotides

The oligonucleotide sequences in Table 6 were obtained from Integrated DNA Technologies

(Coralville, IA, USA) and dissolved in 1×PBS buffer at pH 7.0. The base sequence corresponds

to the SMN1 gene fragment used in oligonucleotide diagnostics for spinal muscular atrophy.189

All subsequent dilutions were also prepared with 1× PBS buffer. Solutions containing a 1:1 ratio

of probe and target oligonucleotides were heated at 95 °C for 5 min and allowed to slowly cool

to room temperature to generate dsDNA. Oligonucleotide solutions used in the measurement of

spectral properties and energy transfer were 0.5 μM in dsDNA (i.e. 1.0 μM in total

oligonucleotide). Oligonucleotide solutions used to evaluate the potential for background

suppression and ratiometric measurements via FRET were prepared as follows: ssDNA was

mixed with TO-PEG stock solution and PBS buffer to produce a solution 1.0 μM in each, and let

stand for 30 min prior to analysis; 1.0 μM ds-DNA solutions (2.0 μM total oligonucleotide) were

prepared as described above, mixed with an equivalent of TO-PEG, and let stand 30 min.

77

Table 6. Oligonucleotide sequences and labels used in FRET experiments.

List of Sequences

Sequences

Probe 5’-ATT TTG TCT GAA ACC CTG T-Cy3/TAMRA-3’

Target 5’-Cy5/IabRQ/IabFQ-A CAG GGT TTC AGA CAA AAT-3’

Abbreviations

Cy3 N,N’-(diisopropyl)-tetramethylindocarbocyanine

Cy5 N,N’-(diisopropyl)-tetramethylindodicarbocyanine

TAMRA Carboxytetramethylrhodamine

IabRQ Iowa Black RQ

IabFQ Iowa Black FQ

TOa Thiazole orange

Labelled target sequences were hybridized with unlabelled probe sequences to generate dsDNA, and vice versa. TOa is not covalently tethered to either terminus of the oligonucleotide.

3.2.3 Association studies

Solutions used in the study of the association of TO-chain and TAMRA labelled oligonucleotide

were prepared by mixing buffered solutions of free side-chain modified TO and either ssDNA or

dsDNA. The solutions were then mixed for approximately 1 hr. Stock solutions of side-chain

modified TO were 100 and 200 μM for the polyethylene glycol and alkyl modified dyes,

respectively. The former was entirely aqueous, and the latter was dissolved in a 1:1

water:methanol solution. Oligonucleotide solutions for analysis were in the concentration range

of 0.25–0.50 μM.

78

3.2.4 Instrumentation

Ultraviolet-visible absorption spectra were collected using a Bichrom Ltd. (Cambridge, UK)

Libra S22 spectrometer and a Hewlett Packard 8452A Diode-Array Spectrometer (Hewlett

Packard Corporation, Palo Alto, CA). Solution phase fluorescence spectra were collected using a

Quanta-Master PTI Spectrofluorimeter and Felix Software (Photon Technology International,

Lawrenceville, NJ, USA). Moisture content measurements of dry solvents were determined using

an AquaStar® Karl Fischer Titrator (EMD Chemicals Inc., Gibbstown, New Jersey, USA) which

provided a readout of H2O concentration in parts-per-million. Fluorescence lifetime data was

obtained with a time correlated single photon counter (constructed in-house243), driven by a 520

nm femtosecond laser (pulse duration: 200 fs, repetition rate: 15 MHz, bandwidth: 3 nm, mean

power: 30 mW).

3.2.5 Spectral data, lifetime, and Förster distance calculations.

The donor–acceptor distance-efficiency relationships for the donor-acceptor pairs were

calculated (see equations 20-23 Chapter 2). Ultraviolet-visible absorption and fluorescence

emission spectra were obtained with 3 and 0.5 μM solutions of oligonucleotide, respectively.

From the spectra obtained, the overlap integrals, (J) integrands were calculated. The refractive

index and orientation factor terms were taken as n=1.43 and κ2 =2/3, corresponding to the

refractive index of buffer and non-restricted dye motion. The excitation wavelength for TO was

set at 507 nm. Quantum yield values for fluorescent dyes were determined relative to fluorescein

dye in sodium borate buffer fixed at pH 9.5. The quantum yield of fluorescein under these

conditions is known to be 0.93215, and the quantum yield, Φ, of other dyes were determined as a

ratio (Eq. (23) Chapter 2) of their integrated emission, ∫Fdλ, corrected for different molar

absorptivity coefficients, ε, at the wavelength of excitation. Fluorescein was excited at 490 nm.

Lifetime measurements were made with 3 μM solutions and were calculated from fluorescence

decay curves using SPCImage software (Version 2,7,27,38,0) from Becker & Hickl GmbH

(Berlin, Germany). The best fit was obtained by minimizing the chi-squared value.

3.2.6 Octanol–water partition experiments

Standard aqueous solutions of PEG (3.6 μM) and alkyl (3.4 μM) modified TO were prepared. An

equal volume of octanol was added to an aliquot of aqueous standards. The mixtures were slowly

stirred for approximately 18 hrs. The peak absorbance for the aqueous phases were compared to

79

that of the standard solutions and used to determine the concentration of the TO species in each

phase.

3.3 Results and Discussion

3.3.1 Energy transfer between thiazole orange and an extrinsic dye label

Thiazole orange and other intercalating dyes are particularly useful in that their fluorescence

intensity is dependent on selective partitioning of the dye into dsDNA. As a first step towards a

FRET-based strategy employing TO, it has been demonstrated that TO can participate in energy

transfer with an extrinsic dye (i.e. not an intercalator) that is covalently attached to the terminus

of a DNA oligonucleotide. In the presence of an extrinsic label, the fluorescence of intercalated

TO-PEG was found to be significantly quenched. The fluorescence spectra in Figure 16 show

quenching efficiencies of 97%, 88%, 88%, 86%, and 80% for TO-PEG with ds-DNA conjugates

of Cy3, TAMRA, IabFQ, Cy5, and IabRQ as energy acceptors.

Figure 16. Fluorescence spectra (a) showing quenching of TO fluorescence by Cy3 and TAMRA. Note the changes in the emission profiles of Cy3 and TAMRA. Cy3 shows the largest quenching of TO as well as increased emission and a slight red-shift; TAMRA shows similar effects. Quenching of TO fluorescence was also observed with IabFQ, Cy5, and IabRQ in (b). Note that there was no appreciable change in Cy5 emission. All solutions are 0.5 μM in both labelled dsDNA and TO-PEG (where applicable).

80

The Förster distances for these TO acceptor pairs were experimentally determined from their

respective emission and absorption profiles and are tabulated in Table 7.

Table 7. Fluorescence and energy transfer data calculated from experimental data for various TO-acceptor pairs.

Acceptor TO quantum

yield, Φ Spectral overlap, J (10−10 cm6)

Förster distance, Ro (nm)

TO 0.011 ± 0.002 0.76 ± 0.20 2.3 ± 0.7 Cy3 5.2 ± 1.1 3.2 ± 0.9 TAMRA 3.7 ± 0.8 3.0 ± 0.8 IabFQ 2.4 ± 0.5 2.8 ± 0.8 Cy5 1.6 ± 0.4 2.6 ± 0.8 IabRQ 0.6 ± 0.2 2.2 ± 0.8

Note. The spectral overlap integral for each pair is given separately to allow calculation of the Förster distance for any index of refraction and orientation. The listed Förster distance employs values of 1.43 and 2/3 for the refractive index and orientation factor, respectively.

These quenching efficiencies roughly mirror the trend observed in the spectral overlap between

TO and the various acceptors. In the case of Cy3 and TAMRA acceptors, an increase in emission

was also observed. This is thought to be emission sensitized by energy transfer from TO to Cy3

or TAMRA. A slight red-shift of the emission max was also observed with Cy3. Fluorescence

lifetime data suggests that the mechanism of TO quenching includes an energy transfer process

for the IabFQ, IabRQ, and Cy5 acceptor species. The measured fluorescence lifetimes for

intercalated TO in the absence of an acceptor, and in the presence of IabFQ, IabRQ, or Cy5 are

tabulated in Table 8.

Table 8. Fluorescence lifetimes measured for TO intercalated in unlabelled, IabFQ-, IabRQ-, and Cy5-labelled dsDNA conjugates.

Acceptor dye Lifetime (ns) (none) 2.6 IabRQ 1.8 IabFQ 2.2 Cy5 2.0

81

As with steady-state quenching efficiency, the decrease in lifetime mirrors the decrease in

spectral overlap between the dye-acceptor pairs. Unfortunately, the lifetime changes for TO in

the presence of Cy3 and TAMRA could not be determined due to the limited spectral resolution

of the available instrumentation. The intense fluorescence of these two dyes at the excitation

wavelength of TO prevented resolution of TO fluorescence. It was found that the fluorescence

lifetimes of Cy3 and TAMRA in dsDNA were 2.0 and 4.2 ns, respectively, and this prevented

time-based resolution with respect to TO fluorescence. The fluorescence of Cy5-dsDNA was

sufficiently low as not to interfere with the measurement of TO fluorescence at the emission

wavelength (520 nm). Assuming that, on average, TO intercalates at the same position in the

dsDNA regardless of extrinsic label, the correspondence between spectral overlap and both

quenching efficiency and TO lifetime suggests a Förster type energy transfer process for the TO

donor with the IabFQ, IabRQ, and Cy5 acceptors. Calculating the donor–acceptor distance, R,

based on the quenching efficiencies and TO fluorescence lifetimes yields the data summarized in

Table 9.

Table 9. Energy transfer efficiencies and TO-acceptor distances deduced from TO fluorescence quenching efficiency and changes in TO fluorescence lifetime.

Quenching Lifetime Acceptor Efficiency, E Distance, r (nm) Efficiency, E Distance, r (nm)

Cy3 0.97 1.8 - - TAMRA 0.88 2.1 - - IabFQ 0.88 2.0 0.31 3.2 Cy5 0.86 1.9 0.23 3.2 IabRQ 0.80 1.7 0.15 2.9 Average (± standard deviation) - 1.9 (± 0.2) - 3.1 (± 0.2) Approximate number of bases from extrinsic dye labelled terminus

- 5-6

- 9-10

In Equation (24), the subscripts DA and D refer to donor (TO) emission in the presence of

acceptor and donor emission in the absence of acceptor, respectively. FDA and FA are similarly

related to fluorescence intensity; the lifetime of TO is denoted by τ.

E = 1 − (FDA − FA)/FD = 1 –τDA/τD = R06/R0

6 + r6 (24)

82

Examining the data in Table 9, it is immediately obvious that the quenching and lifetime data do

not yield the same energy transfer efficiencies. The hypothesis is that this results from the

formation of a non-fluorescent complex between TO and the various acceptor dyes. A second

quenching pathway resulting from formation of such a complex, superimposed on the FRET-

based quenching pathway is consistent with the discrepancy in Table 9. The fluorescence

lifetime is insensitive to non-fluorescent complex formation and thus represents the true FRET

efficiency. The quenching efficiency observed in steady state measurements is the net effect of

these two pathways. The lifetime calculated donor–acceptor distances correspond to a position

approximately in the center of the 19-base pair oligonucleotide sequence, which does reflect the

most stable position for TO to intercalate (considering the ensemble average). This lends

additional confidence to the lifetime data and suggests that complex formation is able to compete

with intercalation to some extent. In terms of method development for biosensor design, this

highlights a potential source of spurious signals which could plague an assay. Thus an

understanding of the physical interaction between the donor and the acceptor is critical, and the

subject of the discussion in a latter section. The orientation factor, κ2, is often a point of difficulty

in the study of systems with limited mobility.216 Although the acceptor molecules in this work

have a large degree of rotational freedom, it is not necessarily the case that κ2 =2/3 since TO is

essentially immobilized within the double helix (for example, Shins et al.244 have studied dipole

orientation in the TOTO-DNA system). Furthermore, if these acceptor dyes have differing

tendencies to associate or aggregate with DNA, then the orientation factor becomes both a

function of the degrees of freedom associated with the linkage, and the dye-DNA interactions.

An estimation of these effects is beyond the scope of this work, and thus it is unclear if the

observed standard deviation of 0.2 nm for the calculated r values represents true variability in the

TO-acceptor distance, variability in the orientation factor, or possibly contributions from both of

these two effects.

3.3.2 Potential FRET-based transduction strategies for DNA hybridization

Initially, two possible strategies were suggested for improving signaling of hybridization events

through the use of TO and FRET: (1) background suppression through the use of a dark

quencher; and (2) ratiometric measurements with the use of a second fluorophore. With respect

83

to (1), Figure 17 shows the peak emission intensities for various permutations of TO with ssDNA

and dsDNA, as both IabFQ-labelled or unlabelled sequences.

Figure 17. Peak fluorescence intensities (measured at 520 nm) for various solutions containing 1 μM of TO and 1μM ssDNA or dsDNA. The fluorescence intensity is expressed as a logarithmic scale. Note the comparatively low fluorescence of the TO in the absence of DNA or in the presence of ssDNA labelled with IabFQ. Measurements were performed in triplicate and error bars represent 1 σ.

It is immediately obvious that TO has significant background fluorescence in the presence of ss-

DNA, albeit considerably less than that of intercalated TO. However, note that TO in the

presence of IabFQ-labelled ssDNA shows extremely low fluorescence which is comparable to

isolated TO dye in solution. The quenching of non-intercalated TO via FRET is a result of an

association of the positively charged TO with the polyanionic probe sequence. This association is

also largely responsible for the fluorescence observed from non-intercalated TO. Note that the

emission from TO in the absence of DNA is only 4% of that observed in the presence of non-

labelled ssDNA, suggesting that the association partially restricts rotation about the

84

monomethine bridge in TO. Figure 18 shows a comparison of the anticipated signal-to-noise

(S/N) ratios for various hypothetical assay schemes based on the data in Figure 17.

Figure 18. Anticipated signal-to-noise ratios for various proposed probe/target hybrid detection schemes. Refer to the main text for a description of schemes. The values shown at right are calculated based on the data presented in Figs. 16 and 17. Legend: (P) probe; (T) target; (D) displaced sequence; (•) IabFQ; (◦) TO; (✩) TAMRA. The shading of each symbol qualitatively represents its emission intensity.

The ‘noise’ is defined as the TO fluorescence in the absence of probe (P)/target (T) duplex; the

‘signal’ is defined as that in the presence of P/T duplex. These schemes are as follows:

1. The typical assay in which the enhancement of TO fluorescence upon intercalation is used to

detect P/T hybridization events. The anticipated S/N ratio is 21.

2. A modification of the typical assay which uses IabFQ labelled P to suppress background

fluorescence from nonintercalated TO prior to P/T hybridization. Despite the reduction in

intercalated TO fluorescence due to the IabFQ, the much reduced fluorescence from non-

intercalated TO leads to an anticipated S/N ratio increase to a value of 76.

85

3. This is a competitive hybridization scheme in which a IabFQ labelled sequence (D) is

displaced from the probe sequence by the target sequence. Due to the relatively large TO

fluorescence expected from the IabFQ-D/P hybrid with respect to the T/P hybrid, the anticipated

S/N ratio is poor, having a value of 6.

4. This strategy requires that TO is initially associated with IabFQ-labelled ssDNA and

subsequently intercalates in P/T duplex, ideally yielding a S/N ratio of 443. This could be

referred to as a competitive association scheme since it would not be possible to prevent TO

from associating with single-stranded P instead of D. If there was an equal tendency for TO to

associate with P and D, the expected S/N ratio decreases to 40.

Although it may seem counterintuitive, scheme II appears to be a simple means by which to

enhance the sensitivity of a hybridization assay with TO by almost a factor of 4. While scheme

IV offers a much larger potential increase in sensitivity (ca. 22-fold), it is not readily

implemented. Solutions of TO were prepared with different ratios of unlabelled and IabFQ-

labelled sequences, and the TO fluorescence measured. The results indicated that there is no

preference for TO association with the IabFQ-labelled sequence (data not shown). Clearly a

biosensor strategy employing scheme IV will not be easily achieved, although it may be possible

with elaborate molecular ‘switch’ chemistry. Considering ratiometric measurements as described

in section 3.1, the use of a competitive hybridization scheme employing TAMRA labelled target

could potentially increase sensitivity by a factor of five. As shown in scheme V in Figure 18, an

initial P/D hybrid could be displaced in favour of a TAMRA T/P hybrid. Comparing the ratio of

fluorescence intensities at the peak emission wavelengths for each dye (roughly 520 and 570 nm

for TO and TAMRA respectively) before and after the displacement events would ideally yield a

S/N ratio of 110. This estimate is based on the change in emission from TAMRA as dsDNA

conjugate with and without intercalated TO. In practice, the TAMRA would be weakly

associated with ssDNA prior to hybridization. Data previously reported in Chapter 2 shows that

the quantum yield of TAMRA increases by 28% upon hybridization6, and considering this

additional increase in the fluorescence at 570 nm, a S/N ratio of 132 could be approached.

Although one could imagine a similar scheme using Cy3 as an acceptor, a 40% reduction in the

quantum yield of Cy3 accompanies hybridization6, thus counteracting the TO-FRET sensitized

increase in fluorescence. Nonetheless, TAMRA appears to be an excellent candidate for such a

transduction strategy. Although labeling of target is still required in scheme V, the use of TO and

86

FRET in conjunction with an extrinsic dye is still advantageous in practical application. Many

optical biosensor devices use surface sensitive evanescent wave excitation. When only labelled

target is used, non-specific adsorption will yield a fluorescent signal which is indistinguishable

from hybrid formation. With scheme V, nonspecific adsorption could be differentiated since it is

less likely to decrease the TO emission signal via FRET. The effectiveness of this scheme would

be maximized by using extrinsic and intercalator dyes with absorption coefficients and quantum

yields of similar magnitudes.

3.3.3 Non-fluorescent complex formation between modified TO and TAMRA

The formation of non-fluorescent ground-state complexes between dyes, including rhodamines

and cyanines, has been previously reported.245-250 In each case, it is believed that hydrophobic

interactions (in an aqueous environment) are the driving force behind complex formation. TO

was modified in these experiments with an alkyl chain or PEG chain, providing different

hydrophobicities. Fluorescence and absorption spectra for TAMRA labelled ssDNA with a 10-

fold excess of modified TO are shown in Figure 19.

87

Figure 19. Absorption and fluorescence spectra for TAMRA labelled ssDNA in the absence and presence of TO. As the TO is made more hydrophobic by changing the nature of the side chain, complex formation appeared to occur to a greater extent. This is suggested by changes in the absorption spectrum and an increase in fluorescence quenching with increasing hydrophobicity. (a) Absorption profiles for 3.0 μM TAMRA-labelled ssDNA in the absence and presence of ten equivalents of TO-PEG or TO-alkyl. The absorption profile of a stoichiometric equivalent of TO is shown for reference. (b) The fluorescence of 0.4 μM TAMRA is quenched by the formation of a ground-state complex with ten equivalents of TO-PEG and TO-alkyl. Fluorescence is excited at 556 nm, where TAMRA shows strong absorption but TO is only very weakly absorbing.

The spectra show increased fluorescence quenching of TAMRA and a change in absorption

profile in the presence of a 10-fold excess of TO. The effect was also observed to increase as the

TO was made more hydrophobic by replacing the PEG side-chain with an alkyl side-chain, thus

supporting the hypothesis of the role of complex formation. The octanol–water partition

coefficients, Kow, were determined to be 0.10±0.02 and 0.87±0.04 for the TO-PEG and TO-alkyl

systems respectively, confirming the greater hydrophobicity of the derivative with the alkyl side-

chain. Note that the absorption profiles for TO-alkyl and TO-PEG show no significant

differences. The distinctly different absorption profiles in Figure 19 are very likely due to

88

different extents of complex formation. Static quenching by the formation of a non-fluorescent

complex (also known as quenching by pre-association) is known to follow the relationship

described by Eq. (25), where Fo and F are the fluorescence intensities of the unquenched and

quenched states, K is the association constant, and [Q] is the concentration of quencher. This

equation is generally applied to quenching by pre-association251 and should not be confused with

the isomorphous Stern-Volmer equation for dynamic quenching.

F0/F = 1 + K[Q] (25)

As shown in Figure 20, the TAMRA fluorescence of different TO:TAMRA-ssDNA conjugate

ratios were measured and yields the linear relationship defined by Eq. (25) at stoichiometric

excesses of TO.

Figure 20. Quenching of TAMRA-ssDNA fluorescence as a function of the relative concentration of TO-alkyl. The linear fit is applied to ratios beyond 2:1 with the slope representing the TO-TAMRA association constant, K. The linear correlation coefficient is denoted by R2. (b) Analogous data for the association and quenching of TO-PEG with TAMRA-ssDNA. The linear fit is applied beyond ratios of 1:1. The concentration of TAMRA-ssDNA in each case is 0.4 μM. Measurements were performed in triplicate and error bars represent 1 σ.

89

A greater extent of quenching was observed with the alkyl side-chain versus the PEG side chain.

The association constants were 0.043±0.002 M−1 and 0.037±0.002 M−1 respectively. Once again,

this suggests a hydrophobic aspect to the TO-TAMRA interactions. Deviations from linearity

existed at TO-TAMRA-ssDNA ratios that were lower than 2:1 indicating that the TO species

was not quenching the TAMRA fluorescence as effectively as at higher TO-TAMRA ratios. This

may have been due to a weak association with the TAMRA and/or the polyanionic backbone of

the DNA. Recall that both TO species are positively charged and that there may also be

hydrophobic interactions between the TO and the nucleobases. It is possible that the weak

association may have served to decrease the collisional deactivation rates in the local

environment of the TAMRA, thus increasing the observed TAMRA fluorescence. Similar

changes in local environment are thought to lead to an increase in TAMRA fluorescence when

going from ssDNA to dsDNA.6 The interaction of TO with TAMRA-dsDNA conjugates was

also investigated, and Figure 21 shows similar quenching to that observed for TAMRA-ssDNA

conjugates at stoichiometric excesses of TO-PEG. The concentration dependent fluorescence of

TO with respect to unlabelled dsDNA is also shown in Figure 21.

Figure 21. Quenching of TAMRA-dsDNA fluorescence as a function of the relative concentration of TO-alkyl. A linear fit has not been applied due to the competition between intercalation and complex formation. (b) TO fluorescence as a function of TO:dsDNA (unlabelled) ratio. The linear fit is applied to the ratios below 2:1. The concentration of dsDNA in each case is 0.4 μM. Measurements were performed in triplicate and error bars represent 1 σ.

90

In the absence of TAMRA, TO fluorescence increased linearly with TO concentration until

stoichiometric excess was reached. Even so, a significant deviation from linearity only occurred

in the range of a 5-to-10-fold excess, indicating that intercalation at multiple sites in the 19-mer

was supported at room temperature. Binding of TO to the external portions of DNA is also

known to occur at sufficiently high concentrations.252,253 The data between Figures 20 and 21

also show that the quenching efficiency is greater with TAMRA ssDNA conjugates than

TAMRA-dsDNA conjugates, suggesting that intercalation is favoured in the competition

between intercalation and complex formation. Note also that the increase in TAMRA

fluorescence at less than two equivalents of TO is also absent. This is consistent with the

aforementioned hypothesis since the nature of the dsDNA is expected to dominate the local

environment around TAMRA. In this sense, weak association of TO with the TAMRA or

polyanionic backbone of the DNA would not significantly alter the environment of the TAMRA.

There was another point of interest in this analysis in that the deviation from linearity of TO-

dsDNA fluorescence that is seen in Figure 21 at 5- and 10-fold excesses of TO represent changes

of 10 and 19%, respectively. The quenching efficiencies of TO with TAMRA-dsDNA at these

concentrations are 13 and 17%, respectively. While it could be suggested that this similarity

indicates that complex formation occurred once the intercalative capacity of the 19-mer had been

satisfied, this is not the case. Quenching was observed at TO concentrations in which

intercalation retained its linearity, thus indicating that there was likely a competition between the

two processes. In light of this data, a FRET-based methodology for biosensor development

which employs intercalating dyes must be carefully designed. The interaction of TO with

TAMRA—particularly at stoichiometric excesses of the former—has the potential to be

detrimental to the sensitivity of a sensor. It is clear that assays employing TO in the presence of a

second dye should avoid large excesses of the former. In addition, it would be very worthwhile

to tether a single TO molecule to the probe sequence at the terminus opposite the TAMRA label.

The data suggests that a tether which allows intercalation near the middle of the sequence (for

oligonucleotides similar in size to those used in this work) would be effective in allowing FRET

and preventing complex formation. In addition to TO-TAMRA complex formation, it should be

noted that TO-TO dimerization can also occur at high concentrations of the dye. The

dimerization of TO is well characterized in the literature 84,254, and here it is noted that the onset

of dimerization is a function of the side chain modification on the TO, i.e. hydrophobicity. With

the more soluble (in aqueous buffer) TO-PEG, it was found that dimer formation became

91

appreciable in the range of 10–15 μM. Using the less soluble TO-alkyl, it was found that dimer

formation was appreciable even as low as 8 μM. The dimerization process is marked by the

growth of a second absorption peak on the blue side of the monomeric TO absorption peak. The

spectra that were obtained are in good agreement with those reported elsewhere.84,252

3.4 Conclusion It has been demonstrated that TO is capable of acting as a FRET donor for the acceptor dyes

Cy3, TAMRA, IabFQ, Cy5, and IabRQ in dsDNA conjugates. Both fluorescence intensity and

fluorescence lifetime data suggest a FRET mechanism, and the Förster distances for these pairs

have been derived from experimental data. Furthermore, fluorescence lifetime data indicates that

TO intercalates near the center of the double-stranded 19-mer oligonucleotide that was used in

this study. Discrepancies between the FRET efficiency derived from intensity and lifetime data

are attributed to the ability of TAMRA-labelled oligonucleotide to form a non-fluorescent

complex with TO. Complex formation has been investigated using polyethylene glycol and alkyl

side chains attached to TO, and the results indicate greater fluorescence quenching with the more

hydrophobic alkyl side chain. In addition, TO dimerization was also found to be more extensive

when using the more hydrophobic alkyl side-chain modified dye. These results indicate that

combinations of dyes, or use of large quantities of a single dye with hydrophobic tendencies,

must be carefully evaluated to avoid induction of associations between dyes. Physical

associations may affect photophysical properties and lead to discrepancies in assay results that

are anticipated from FRET. With respect to this last point, this work has proposed and evaluated

a variety of FRET schemes using TO for the transduction of DNA hybridization events. It is

found that a roughly four-fold increase in signal-to-noise ratio over a conventional TO-based

assay can be obtained by using IabFQ to suppress background fluorescence from nonintercalated

TO. Similarly, a proposed ratiometric scheme using TAMRA as a fluorescent acceptor dye is

expected to yield a greater than five-fold increase in signal-to-noise ratio.

92

4 Towards a Fluorescent Molecular Switch for Nucleic Acid Biosensing Analytical and Bioanalytical Chemistry 98 (4) (2010) 1605-1614 ©2010 Springer-Verlag

Abstract

A novel fluorescent molecular switch for the detection of nucleic acid hybridization has been explored in relation to the development of a structure that would be amenable for operation when immobilized for solid-phase analyses. The structure was prepared by self-assembly, and used Neutravidin as the central multivalent docking molecule, a newly synthesized biotinylated long-chain linker for intercalating dye that was modified with thiazole orange (TO) at one end, and a biotinylated probe oligonucleotide. Self-assembly of the biotinylated components on adjacent Neutravidin binding sites allowed for physical placement of an oligonucleotide probe molecule next to tethered TO. The TO located at the end of the flexible linker chain was available to intercalate, and could report if a duplex structure was formed by a probe–target interaction by means of fluorescence intensity. Subsequently, regeneration of the single-stranded probe was possible without loss of the intercalator to solution. The switch constructs were assembled in solution and subsequently immobilized onto biotin functionalized optical fibers to complete the sensor design. Solution-phase fluorescence lifetime data showed a biexponential behavior for switch constructs, suggesting intercalation as well as a significant secondary binding mode for the immobilized TO. It was found that the secondary binding mechanism for the dye to DNA could be decreased, thus shifting the dye to intercalative binding modes, by adjusting the solution conditions to a pH below the pI of Neutravidin, and by increasing the ionic strength of the buffer. Preliminary work demonstrated that it was possible to achieve up to a five-fold increase in fluorescence intensity on hybridization to the target.

93

4.1 Introduction Nucleic acid biosensing platforms are being developed as a method to provide rapid, reliable

detection of pathogens and genetic diseases.10 Constructs that are simple in design, assembly,

and functionality are an obvious advantage for the detection of nucleic acid targets. The work

presented in Chapter 3 found that the spectroscopy of the TO-FRET-systems studied were

complicated by the potential for hydrophobicity and stoichiometrically driven non-fluorescent

complex formation with an acceptor dye. Therefore, the motivation for the work presented in

this chapter was to simplify the overall design platform by reducing the signaling process to one

that involved a single intercalating dye strategy. Label-free techniques or techniques that

incorporate an intrinsic signaling mechanism eliminate the requirement for target labelling,

which decreases the overall complexity and time required for analysis. Nanoscale assemblies

have been reported which contain all components of the “sensor cargo” in one unit, such as those

found in molecular beacon systems. The interest is in extension of the conceptual elegance of a

sensor design that integrates the selectivity of hybridization with an intrinsic signalling

mechanism for applications in solid-phase analysis. A molecular assembly for nucleic acid

detection has been developed and is schematically shown in Figure 22.

94

Figure 22. Design of the label-free fluorescent molecular switch hybridization monitoring construct, and schematic representation of the proposed molecular hybridization switch strategy. Site-directed templating of the biotinylated thiazole orange intercalating dye derivative places the dye in close proximity to a probe oligonucleotide using biotin/Neutravidin affinity chemistry. Upon the introduction of a nonlabeled target oligonucleotide, an enhancement of fluorescence intensity due to the intercalation of the thiazole orange moiety provides a signal that is indicative of a hybridization event.

The construct makes use of Neutravidin as a templating platform and offers the opportunity to

combine a biotinylated probe oligonucleotide with a nearby thiazole orange (TO) intercalating

dye that is also physically immobilized by means of a biotinylated linker. Neutravidin has been

reported to have a high binding affinity for biotin, and it offers the advantage of reduced non-

specific adsorption to surfaces while still retaining significant binding affinity: (Ka=5.5 x 1011 M-

1).64,255,256 The biotin binding sites on Neutravidin are arranged with a distance of approximately

1 nm between adjacent sites, and 5 nm for distal sites.257,258 With the introduction of Neutravidin

as a “central hub”, it is possible to have this docking molecule function simultaneously as an

interface for the adjacent placement of the biotinylated probe (width of ssDNA 1 nm)259 and

signaling dye on a biotinylated linker as well as for surface immobilization through the use of the

distal biotin binding sites. Immobilization of sensor constructs onto fused silica optical fibers

provides for evanescent wave excitation of fluorophores via total internal reflection, and is a

well-established method of developing nucleic acid biosensors.184,189 Some intercalating dyes can

exhibit different binding affinities to single-stranded DNA (ssDNA) and double stranded DNA

95

(dsDNA). Such dyes can provide fluorescence intensity and lifetime changes in different

environments that can be used to distinguish intercalation from other binding events. Thiazole

orange (TO) is an asymmetric cyanine dye which is capable of undergoing significant

fluorescence enhancement upon intercalation into dsDNA.80 The work herein appears to offer

the first report of the synthesis of a biotinylated intercalating dye species which incorporates a

tether of sufficient length (18-atom or 26-atom tether lengths were examined) to provide

flexibility, mobility and availability of the dye species when it is immobilized for solid-phase

assays. By first anchoring the biotinylated probe to Neutravidin, the immobilized oligonucleotide

strand can site-direct the assembly of the biotinylated linker that is capped with TO. The use of a

single signaling dye simplifies the design when compared to the dye/quencher pairs often

encountered in fluorescence resonance energy transfer technologies. The new sensor construct

also provides a platform where potentially hazardous intercalating dyes are permanently placed

at an interface, and the handling of solutions of such dyes is ameliorated. In addition, the

conformational changes upon binding that are associated with a construct comprising linear

components are anticipated to be less sterically hindered when compared to hairpin/loop

switching structures.

4.2 Experimental All chemicals were reagent grade or better and used without further purification. Reagents used

for thiazole orange synthesis, including methyl iodide, 2-(methylthio)benzothiazole, 4-

methylquinoline, triethylamine, N-hydroxysuccinimide (NHS), N,N′-diisopropylcarbodiimide

and anhydrous N,N-dimethylformamide (99%) were from Sigma‐Aldrich (Oakville, ON,

Canada). (+)-Biotinyl-3,6-dioxaoctanediamine (EZ-Link® amine-PEG2-biotin) and Neutravidin

were from Thermo Scientific (Rockford, IL, USA) through Fisher Canada (Nepean, ON,

Canada). Reagent-grade toluene, ethanol, acetone, isopropanol, methanol, dichloromethane, and

anhydrous diethyl ether were from EM Science (Toronto, ON, Canada). Glass-cleaning reagents,

including Hellmanex II glass cleaner (Hellma GmbH & Co. KG, Müllheim, Germany), and 30%

hydrogen peroxide, 28.0–30.0% ammonium hydroxide, 35.6–38.0% hydrochloric acid, and 18 M

sulfuric acid were from EMD Chemicals (Gibbstown, NJ, USA). Glass surface modification

reagents, including aminopropyltrimethoxysilane, 99.5% N,N-diisopropylethylamine (Hünig’s

base), N-ethyl-N’-(3-dimethylaminopropyl) carbodiimide (EDC), and biotin were from Sigma-

Aldrich. HPLC-purified modified oligonucleotide sequences were from Integrated DNA

96

Technologies (Coralville, IA, USA). Phosphate-buffered saline (PBS) buffer (pH 7.4) and

acetate buffer (pH 4.67) were prepared using double-distilled water in glass, which was

subsequently autoclaved. Buffer salts (sodium chloride, dibasic sodium phosphate, monobasic

sodium phosphate, sodium acetate) were from Sigma-Aldrich. Water for rinsing and washing

was purified using a Milli-Q cartridge system (Millipore Corporation, Mississauga, ON,

Canada). Fused silica optical fiber was from Polymicro Technologies, LLC (Phoenix, AZ, USA).

4.2.1 Synthesis of two biotinylated thiazole orange derivatives

The synthesis of TO-bio1 began with an iodopropyl side chain that was attached to thiazole

orange based on the methods of Brooker et al.240,241 and Carreon et al.97 and is described in

section 3.2.1 of Chapter 3 as well as Appendix I. The biotinylated thiazole orange moiety was

then synthesized by reacting the previously synthesized thiazole orange dye (0.03 g, 0.051

mmol) in equimolar quantity with EZ-Link® amine-PEG2-biotin (0.02 g, 0.051 mmol) in the

presence of triethylamine in dry DMF at 60 °C for 48 hr. DMF was evaporated under reduced

pressure to yield a red solid. The mass calculated for [C37H49N6O4S2+] was m/z 705.95, and the

value found was 705.11 amu (ESI+). The biotin-functionalized thiazole orange conjugate is

shown in Figure 23.

97

Figure 23. Synthesis of TO-bio1 biotinylated thiazole orange intercalating dye containing a 18-atom tether (including biotin valeryl side-chain).

98

4.2.1.1 Synthesis of TO-bio2

The synthesis of TO-bio2 began with the attachment of an undecanoic acid side chain to thiazole

orange based upon the methods of Svanvik et al.260 and Carreon et al.97, and subsequent

conversion to an NHS ester was based on the methods of Pei et al.261 The carboxylic acid

functionality at the chain terminus was then converted to an NHS ester by reacting the dye

species (30 mg, 0.05 mmol, 1 equiv.) with N-hydroxysuccinimide (NHS) (6.32 mg, 0.055 mmol,

1.1 equiv.) in the presence of DIC (6.93 mg, 0.055 mmol, 1.1 equiv.) in dry dichloromethane.

The reaction was allowed to stir under an inert atmosphere at room temperature for 5 h. The

biotinylated thiazole species was then synthesized by reacting the NHS ester dye derivative (21

mg, 0.035 mmol, 1 equiv.) with EZ-Link®amine-PEG2-biotin (15 mg, 0.039 mmol, 1.1 equiv) in

the presence of DIPEA (50 μL, 0.28 mmol, 8 equiv.) in anhydrous DMF (3 mL). The reaction

was allowed to stir under an inert atmosphere of argon for 20 h at room temperature. The DMF

was evaporated under reduced pressure and the product was rinsed with MilliQ water. The

product was then dissolved in methanol and precipitated with diethyl ether to yield a red solid

(1.27 mg, 4%). The mass calculated for [C45H63N6O5S2+] was m/z 832.15, and the mass found

was 832.30 amu (ESI+). The biotin functionalized thiazole orange conjugate is shown in Figure

24.

99

Figure 24. Synthesis of TO-bio2 biotinylated thiazole orange intercalating dye containing a 26-atom tether (including biotin valeryl side-chain).

100

4.2.2 Preparation of optical fibers

Fused silica fibers (approximately 4 cm in length) were treated with piranha solution (50% v/v

18 M H2SO4 and 50% v/v 30% H2O2) for 24 h to remove the outer cladding. The fibers were

rinsed several times with deionized water and methanol. They were then sonicated in a dilute

solution of Hellmanex in deionized water (2×10 min) and rinsed with copious amounts of

deionized water. The fibers were then subjected to a rigorous cleaning protocol according to

previously established methods.212 A solution of 1:1:5 v/v/v concentrated NH4OH:30% H2O2:

deionized water was prepared, and the fibers were immersed in this solution, with heating to 80

°C for five minutes. The solution was then decanted and the fibers were rinsed several times with

deionized water. The fibers were then immersed in a solution of 1:1:5 v/v/v concentrated

HCl:30%:H2O2:deionized water, heated to 80 °C for five minutes, and the solution was decanted.

The fibers were rinsed successively with methanol, dichloromethane, and diethyl ether, and

stored in an oven at 125 °C until required. Clean and dry fiber substrates were immersed in a

solution of 75 mL refluxing anhydrous toluene, APTMS (1.7 mL, 9.94 mmol), and N,N-

diisopropylethylamine (0.2 mL, 1.15 mmol) under an inert argon atmosphere for 18 h. The

substrates were subsequently washed sequentially with methanol, dichloromethane, and diethyl

ether, and were then dried in a dessicator for a minimum of 12 h. The APTMS functionalized

substrates were immersed in a solution of 75 mL N,N-dimethylformamide (DMF), biotin (0.035

g, 0.14 mmol), and DCC (28 μL, 0.13 mmol), and this was placed on an orbital shaker for 12 h.

The fibers were recovered and successively rinsed with DMF, dichloromethane, and diethyl

ether, and dried in a desiccator.

4.2.3 Switch assembly in solution

Biotin-modified probe oligonucleotide sequences (biotin-C6H12-5′-ATT TTG TCT GAA ACC

CTG T-3′) and target oligonucleotide sequences (5′-A CAG GGT TTC AGA CAA AAT-3′)

were reconstituted through dilution of initial mass in 0.1 M PBS buffer at pH 7.4. The probe base

sequence corresponds to the clinically relevant SMN1 gene fragment used in oligonucleotide

diagnostics for spinal muscular atrophy (SMA).189 Solutions containing a 1:1 ratio of probe and

target oligonucleotides at a concentration of 1 μM were heated at 95 °C for 5 min and allowed to

slowly cool to room temperature to generate dsDNA. Two methods of switch assembly were

investigated. For the first method, after cooling, biotinylated thiazole orange was added in a 1:1

ratio of dye:dsDNA, and the solution was allowed to stand for 30 min prior to the addition of

101

Neutravidin. Neutravidin was added such that all switch components (biotinylated dsDNA,

biotinylated dye, and Neutravidin) were present in a 1:1:1 ratio, with each component at a

concentration of 1 μM. For the second method, biotinylated thiazole orange was added in a 1:1

ratio of dye:Neutravidin, and the solution was allowed to stand 30 min prior to addition of

dsDNA. Fluorescence spectra were recorded after each structural component was added to

monitor the fluorescence signal change as a function of the presence of each component.

4.2.4 Switch immobilization and activation

Molecular switch constructs assembled in solution were immobilized onto biotinylated fused

silica substrates by immersing the fibers in a 1 mL solution containing the switch constructs for

24 h. A schematic of the switch assembly and immobilization process is shown in Figure 25. The

coated fibers were then sonicated in 0.1 M PBS buffer for five minutes, rinsed in deionized

water, and fluorescence spectra were recorded. At this stage, the oligonucleotides immobilized at

the interface were still in the double-stranded form. Probe/target hybrid denaturation was done

using two methods: (1) a chemical denaturation method, and (2) a thermal denaturation/cycling

method. Method (1) was accomplished by sonicating the fibers in an aqueous 4 M urea solution

for five minutes, which generated the ssDNA probe at the surface and eliminated non-

specifically adsorbed material. The fibers were then rinsed using 0.1 M PBS buffer and

deionized water. The second method involved heating the switch coated fibers in 0.1 M PBS

buffer at 65 °C for 20 min to generate the ssDNA probe at the surface. The fibers were allowed

to cool to room temperature. This was followed by a thermal cycle in 0.1 M PBS buffer at 45 °C

for 15 min, after which the fibers were allowed to cool to room temperature.

4.2.5 Interfacial hybridization experiments

Hybridization experiments were done using the switch coated fibers. For fibers prepared using

the chemical denaturation method, the fibers were immersed in a 5 μM solution of fully

complementary target oligonucleotide for 5 h. The fibers were then removed from the solution,

rinsed with both 0.1 M PBS buffer and deionized water, and fluorescence was then determined.

For the fibers prepared using the thermal denaturation method, the fibers were immersed in a 1

μM solution of target oligonucleotide for various times from 40 min to 4 h. All fluorescence

determinations were done at 25 °C. Solution-phase steady-state fluorescence spectra were

acquired using a Quanta-Master PTI spectrofluorimeter and Felix software (Photon Technology

102

International, Lawrenceville, NJ, USA). Fluorescence spectra were collected from the fused

silica optical fibers by modifying the same instrument. Coupling of excitation light via total

internal reflection into the fibers was achieved using laser radiation at 488 nm from a Coherent

Innova 70 CW argon ion laser (Coherent Laser Products, Palo Alto, CA, USA), and an optical

fiber housing that was custom built to fit into the sample cuvette compartment as shown in

Figure 25.184

Figure 25. Schematic of the self-directed templating process used for switch coated optical fiber assembly and instrument assembly used for interrogation of optical fibers via total internal reflection. The modified sample compartment of a PTI Spectrofluorimeter allows laser radiation from a 488 nm Argon ion laser to be coupled into the optical fibers with immobilized switch chemistry present on the surface.

103

Time-resolved fluorescence decay profiles were collected with a PTI Laser-Strobe™ system

using a dye (Coumarin 503, 6 mM, in ethanol) laser (GL-302, Photon Technology International)

tuned to 511 nm, which was pumped by a pulsed nitrogen laser (GL-3300, Photon Technology

International). Coumarin 503 was from Exciton (Dayton, OH, USA). TimeMaster (Photon

Technology International) decay fits were generated using Felix software. Steady state

fluorescence emission spectra were collected from 521 to 650 nm. The solution-phase spectra

were blank subtracted. The emission spectra collected from fiber surfaces using the chemical

activation protocol were normalized using the second-order diffraction peak associated with the

488 nm laser line of the Ar laser in order to compensate for heterogeneity in the coating of fiber

surfaces, and for variations in the laser intensity and angles that coupled into fibers.184 These

data were computed according to Eq. 26 below, where F(λ) represents the fluorescence intensity

recorded at wavelength λ.184

                                     

∑990

)(λF=970λ

                     (26)

This allowed for data at the peak emission wavelength of the biotinylated thiazole orange (530

nm) to be compared at different stages of analysis from the same fiber. In this case, F(λ) in the

numerator corresponds to the peak emission intensity at 530 nm, which represents the maximum

fluorescence emission wavelength of the biotinylated fluorophore. The emission spectra

collected from fiber surfaces using the thermal activation protocol were blank subtracted using a

fiber coated with the biotin immobilization chemistry. Spectra were normalized to the signal at

500 nm prior to blank subtraction.

4.3 Results and Discussion This work represents a preliminary investigation of a novel concept for developing a molecular

switch for solid-phase assays that is capable of detecting label-free target oligonucleotide.

Biotinylated flexible long-chain linkers that were terminated with TO have been synthesized and

used as fluorescent signaling probes both in solution and at an interface to monitor probe/target

oligonucleotide binding. The switching effect arises from both the ability of the construct to be

regenerated, as well as the “on/off” capability of thiazole orange to undergo a fluorescence

quantum yield change upon DNA hybridization. The synthesis of TO-bio1 began with the

104

production of an iodopropyl group attached to the quinolinium nitrogen, which allowed for facile

introduction of an 18-atom linker arm between the parent thiazole orange compound and the

native biotin moiety (Figure 23). Lartia and Asseline have reported that cyanine derivatives that

are covalently linked to internucleotide phosphate groups of oligonucleotide sequences require a

linker with a minimum of eleven atoms.262-264 These results, taken together with the spacing of

adjacent biotin binding sites, suggests that an 18-atom linker for the new construct used herein

would provide sufficient length for the intercalation of TO into double stranded hybrids at an

interface, even though the modified dye would not be covalently attached to the oligonucleotide

probe sequence. The modified TO dye was examined for its ability to associate with dsDNA in

bulk solution to provide a measurable signal corresponding to probe-target binding. The

biotinylated linker was attached at the quinolinium nitrogen on the parent dye moiety. It has been

shown that the attachment of linkers at this dye position facilitates orientation into the minor

groove and hence improves quantum yields of thiazole orange–peptide conjugates, as conjugates

with linkers attached to the benzothiazole nitrogen can adversely affect intercalation.97 The TO–

biotin conjugate was characterized using ESI (positive ion mode) mass spectrometry, UV-visible

spectroscopy and fluorescence spectroscopy. The solution-phase absorption profile of the

biotinylated dye (no DNA present) further confirmed the presence of the biotin moiety (314 nm)

and showed a 2 nm red shift (504 nm to 506 nm) in the absorption peak associated with the

conjugated dye with respect to the unmodified dye. The TO–biotin conjugate showed similar

fluorescence properties to the parent dye compound as shown in Figure 26.

105

Figure 26. (a) Fluorescence emission spectrum of biotinylated thiazole orange in the presence of dsDNA (black) and a thiazole orange moiety functionalized with a -(-CH2)2-O-(CH2)2-O-(CH2)2-OH side chain in the presence of dsDNA (red). Dye and DNA were each 1 uM in concentration (1:1 ratio). (b) Fluorescence emission spectrum of biotinylated thiazole orange in the presence of dsDNA (black) and subsequent binding to Neutravidin (red).

Comparison of the parent with the biotinylated dye showed that the biotinylation did not

significantly alter the photophysical properties (see Figure 26). The spectra are consistent with

literature reports that indicate biotinylation of a dye results in a red shift in the emission

spectrum.265 It is known that some biotinylated dyes can experience a decrease in steady state

fluorescence intensity upon biotinylation and subsequent association with avidin.266,267 As

shown in Figure 26, this was observed with the biotinylated thiazole orange derivative.

Literature reports of this phenomenon have been associated with dyes that are used to track

dye/protein interactions (i.e., no nucleic acids present). In the work presented here, the

biotinylated thiazole orange dye would not be expected to fluoresce in the absence of nucleic

acids, but could perhaps be indicative of competitive binding modes based on differences in

energetics associated with affinity constants. It is plausible that the reduction in fluorescence

intensity could arise from the dissociation of externally bound or intercalated dye molecules

associated with dsDNA upon biotin binding to Neutravidin.

Biotinylated dye/protein binding was further investigated in the work of this thesis by comparing

free dye (no biotin tether) and biotinylated dye, and also the physical form of the protein (Avidin

106

vs. Neutravidin). Since measurements were performed in buffer at physiological pH (0.1 M PBS

at pH =7.4), the pI of the protein becomes a factor in the analysis. The pI of avidin is ~10, while

the pI of Neutravidin is ~6.3.268 Therefore, the charge on Avidin at neutral pH is positive, while

that of Neutravidin is negative. Since the dye species carries a positive charge, it was anticipated

that both the free dye and biotinylated dye would show a higher fluorescence signal upon

interaction with Neutravidin as a result of electrostatic interactions. As shown in Figure 27, this

was the case, but it was also shown that the biotinylated dye exhibited a higher quantum yield in

the presence of both proteins when compared to the free dye. It was hypothesized that this is due

to the biotinlyated dye binding to the two types of protein via the biotin binding pocket, thus

placing the dye in close proximity to the protein surface. Adsorption of dye to the protein

surface would be expected, although with different affinities depending on charge, and is shown

by a broad band emission at a wavelength that is approximately 20 nm red-shifted from the

emission peak of thiazole orange in the presence of dsDNA.

Figure 27. Fluorescence emission spectra of thiazole orange (free dye) and biotinylated thiazole orange in the presence of Avidin and Neutravidin at pH = 7.4.

107

Figure 28. Fluorescence emission spectra of Neutravidin: (a) Fluorescence emission of tryptophan residues residing in biotin binding pocket of Neutravidin (1 μM) (b) Fluorescence emission of tryptophan residues associated with Neutravidin after the addition of TObio2 (5 μM). Inset. Fluorescence difference spectrum of Neutravidin due to biotin, as calculated by subtracting biotin bound fluorescence emission spectrum from biotin-free fluorescence emission spectrum.

Further definitive confirmation of biotin binding was sought because one of the most important

factors required for the proposed molecular switch is to ensure that the components are binding

as anticipated. Perhaps one of the most important interactions required for this construct is the

binding of the biotinylated intercalating dye to the biotin binding pockets on Neutravidin to

create the signalling component of the switch construct. It was therefore important to establish

that the biotinylated dye was not simply adsorbed to the surface of the protein. This was a

concern since Neutravidin, under the conditions used, has an overall net negative charge. While

this is beneficial for creating charge repulsion to decrease adsorption of nucleic acid components

of the switching construct, this advantage could be lost because thiazole orange has a positive

charge as a result of quaternary nitrogen atoms present in the dye structure. A widely used

method for monitoring biotin binding to avidin-based proteins is a HABA (hydroxyaminobutyric

acid) displacement assay, where the decrease in absorbance at ~500 nm is monitored as a

108

function of biotin binding. Due to the high binding affinity of biotin to avidin, biotin displaces

weakly bound HABA molecules, leading to a decrease in absorbance at 500 nm as a function of

the number of biotins which bind to the protein. However, this method is problematic for this

work since thiazole orange has a λabs maximum of ~501 nm. An alternate method was used

based on the work of Kurzban et al. where monitoring of the fluorescence emission of

tryptophan residues present in the biotin-binding pocket can be used to monitor biotin binding.269

This ameliorates the complications associated with overlapping absorption profiles associated

with HABA and TO, and provides a convenient fluorescence-based method of confirmation of

dye binding using a wavelength that lies outside the spectral window associated with thiazole

orange. Literature results have shown that biotin binding reduces tryptophan fluorescence and

results in a blue shift of the emission peak due to a conformational change induced upon binding

in both streptavidin/biotin systems and Neutravidin/biotin systems.269,270 The work presented

here was done using a biotinylated dye and Neutravidin, and showed similar data trends, thus

confirming that the biotinylated dye was binding to the biotin-binding pocket of Neutravidin.

Biotinylated dye was added in 5-fold excess to the protein to ensure that all four binding sites

were occupied. The biotin-induced changes led to a reduction in fluorescence intensity, a blue-

shift of the emission peak, and a reduction in spectral bandwidth as shown in Figure 28. The

plotted spectrum of the difference between the unbound and bound states showed a wavelength

maximum of 348 nm and a spectral bandwidth of 54 nm (Figure 28 inset). The comparison of

data between this work and the work by Kurzban et al.269 and Bottini et al.270 is shown in Table

10.

Table 10. Comparison of fluorescence parameters associated with literature values reported by Kurzban et al. and Bottini et al. with the Neutravidin/biotinylated dye system presented in this work. Streptavidin

/biotin system269 Neutravidin /biotinylated dye

Neutravidin/biotinylated NP270

Concentration ratio

5:1 biotin: Streptavidin 5:1 dye: Neutravidin 4:1 biotin: Neutravidin

Peak fluorescence emission decrease

39% 61% -

Shift in peak emission upon biotin binding

blue shift 4nm blue shift 19nm blue shift 14nm

Spectral bandwidth reduction

7 nm 5 nm 8nm

λmax of difference spectrum

344 nm 348 nm -

Difference spectrum bandwidth 58 nm 54 nm -

109

Further calibration data showed a linear trend for difference spectra obtained using 1-4

equivalents of TO-bio2: Neutravidin (see Appendix I, Figure A.1.1) and demonstrates the

sequential occupation of the four biotin binding sites present on the protein. A deviation from

the linear trend was observed for five and six equivalents of biotinylated dye and reflects the

excess number of biotinylated moieties with respect to the number of available binding sites.

Switch assembly was also confirmed using these biotin-induced changes (see Appendix I, Figure

A.1.1). The calculated number of binding sites for the switch construct was 2.0±0.2, which is

consistent with a site-directed templating strategy in which a biotinylated probe oligonucleotide

and a biotinylated dye moiety bind to Neutravidin through biotin/avidin affinity interactions.

Streptavidin and Neutravidin are both tetrameric proteins which contain three tryptophan

residues per monomer, so similar trends would be expected between the two proteins.270

However, it was observed in this work that there was a much larger blue shift upon biotin

binding than that reported by Kurzban for the streptavidin/biotin system. However, Bottini et al.

have also used the method described by Kurzban et al. for analysis of binding of Neutravidin to

silica quantum dots that were biotin-functionalized. The results of that work showed a similar

blue shift, which lends confidence that a larger blue shift is associated with the use of

Neutravidin than for Streptavidin upon biotin binding. The mechanism of fluorescence

quenching is thought to be due to local structural changes within biotin binding sites where

tryptophan residues that occupy moderately polar and buried local environments are collisionally

quenched due to a biotin-induced conformational change.269

The biotinylated dye showed a very low fluorescence signal in the absence of dsDNA in

solution, as was expected. Upon the introduction of dsDNA, a significant fluorescence intensity

increase was observed. For an equimolar dye/dsDNA ratio, an almost 100-fold increase in

fluorescence intensity was observed for the dye bound to dsDNA over free dye in solution (no

DNA present), but only a 2.5-fold increase was observed when comparing the signal to that

observed in the presence of ssDNA. Figure 29 shows the fluorescence spectra of TO-bio2 as free

dye in solution and in the presence of ssDNA and dsDNA.

110

Figure 29. Fluorescence emission spectra of TO-bio2: (a) free dye in solution; (b) TO-bio2 in the presence of ssDNA; (c) TO-bio2 in the presence of dsDNA. The concentration of each component was 1 μM in 0.1 M PBS buffer, 20 mM NaCl, pH 7.4.

The next step in the construction of the switch was to add Neutravidin to capture the complex

that was formed by the interaction of the biotinylated dye with dsDNA in solution. Neutravidin is

a nonglycosylated form of the avidin class of biotin-binding molecules, and was selected for this

work because it has been reported to have lower nonspecific binding properties than avidin.271

The strong binding affinity between Neutravidin and biotin is still retained when the protein is

deglycosylated.256 Neutravidin provided affinity binding of the biotin labels of the DNA/dye

complex in close proximity to form the mechanical elements of the switch, as well as the

opportunity to subsequently immobilize the switch at a biotinylated interface.

111

Figure 30. a–g Fluorescence spectra of TO-bio2 in the presence of dsDNA in 0.1 M PBS buffer containing: (a) 20 mM NaCl; (b) 100 mM NaCl; (d) 500 mM NaCl; (f) 1 M NaCl; followed by subsequent binding to Neutravidin: (e) 20 mM NaCl; (c) 100 mM NaCl; (h) 500 mM NaCl; (g) 1 M NaCl. The concentration of each switch component was 1 μM.

Figure 30 (and Figure 26(b)) shows that, while a signal decrease occurred for the thiazole orange

moiety upon binding to Neutravidin, the fluorescence intensity was clearly sufficient to identify

that the conjugated dye was associated with dsDNA, and that the magnitude of signal decrease

could be controlled through the manipulation of the ionic strength of the buffer solution. The

ratio of dye/DNA/Neutravidin was set at 1:1:1 to prepare the switch constructs in solution.

Control experiments involving the binding of biotinylated TO and Neutravidin in the absence of

dsDNA showed that the fluorescence signal of the biotinylated dye in the presence of dsDNA

and Neutravidin was approximately 26-fold higher than that of TO-bio1 in the presence of

Neutravidin alone. An approximate nine-fold increase in fluorescence signal of the dye in the

presence of Neutravidin when compared to TO-bio1 alone in solution suggests some non-

specific adsorptive effects that contribute to background fluorescence when the protein is

112

present. This is not surprising considering that the dye is charged and could interact with the

protein.

Steady-state fluorescence may not provide binding details that are integral to understanding the

mechanism of interaction of various structural components in the switch construct. The

interaction of TO-bio1 and TO-bio2 under a variety of conditions was studied using time-

resolved fluorescence spectroscopy. The results summarized in Table 11 indicate that

fluorescence emission from the switch constructs cannot be explained with a single mechanism

such as intercalation. The general hypothesis is that the rate of rotation or torsion around the

methine bridge which links the two heterocycles of cyanine dyes controls the non-radiative

decay of the dyes.93 Netzel et al.93 have shown that the monomeric cyanine dyes BO-PRO-1,

PO-PRO-1, YO-PRO-1 and TO-PRO-1 exhibit either bi- or triexponential emission decay

kinetics, which reflects the presence of different dye/dsDNA modes of binding. Similar bi-

exponential decay curves were observed in this work. However, in work reported by Netzel et al.

for TOPRO-1 and Schweitzer et al. for PicoGreen intercalation into dsDNA, the contribution

attributed to the short lifetime component was much smaller than that of the longer lifetime

component associated with intercalation.93,272 In the absence of DNA, the fluorescence decay

profile of TO-bio1 in 100 mM PBS, 100 mM NaCl pH 7.4 buffer was 0.16 ± 0.03 ns. This short

lifetime is reflected in the low quantum yield seen in the absence of DNA.84 This lifetime

component increased to 0.21 ± 0.02 ns upon the addition of Neutravidin. Pre-exponential factors

and fluorescence lifetime values recorded from solution-phase time resolved fluorescence data

showed that the dye/Neutravidin complexes that were bound with biotinylated ssDNA and

dsDNA were not statistically distinguishable at pH 7.4 using 100 mM NaCl in PBS buffer. It has

been previously reported that the fluorescence lifetime of TO associated with dsDNA in solution

is 2.6 ns.10 The fluorescence lifetime data presented here suggests that either a non-specifically

adsorbed dye species or a different binding mechanism to the DNA contributes to the

fluorescence signals observed. The similarity in fluorescence lifetimes as well as the associated

pre-exponential factors for both the ssDNA case and dsDNA (when bound to Neutravidin) may

indicate that the dye is in sufficient proximity to the probe strand and/or the Neutravidin that

similar interactions that are not associated with intercalation may be taking place. The significant

contribution of the short lifetime component gives some insight into the low signal-to-noise that

has been observed in steady-state fluorescence intensity changes upon the introduction of a target

113

oligonucleotide for hybridization. The pI of Neutravidin is 6.3 ± 0.2273, so at pH 7 the protein

would be negatively charged, while the TO would be positively charged. The fluorescence

lifetimes for TO-bio2 in the “switch” configuration at pH 7.4, with 0.1 M PBS buffer containing

20 mM NaCl and 100 mM NaCl, were recorded and showed a decrease in the pre-exponential

factor for τ1 (the shorter lifetime component) when in the solution of higher ionic strength (9.4 ±

0.9 for 20 mM NaCl vs. 5 ± 1 for 100 mM NaCl). The longer-lifetime component and its

associated pre-exponential remained unchanged within experimental error. Switch constructs

were examined at pH 4.67 in acetate buffer at both 100 mM NaCl and 500 mM NaCl to create a

condition in which the protein carried a net positive charge since pH<pI. Since the dye species is

cationic, there may be less nonspecific binding associated when the protein is positively charged.

It has also been reported that increasing the ionic strength of solutions containing cyanine dyes

can reduce the aggregation of dye species in solution if the dye interaction is via groove binding

with dsDNA.91 The fluorescence lifetime data at pH 4.67 showed a change in the ratio of the

contributions of the two lifetime components. At a lower pH and higher ionic strength, the

contribution of the longer lifetime component associated with intercalation became much more

significant.

Table 11. Time-resolved fluorescence data for TO-bio1 and TO-bio2.

Probe nucleic acid immobilized at an interface in the proximity of an immobilized intercalating

dye that exhibits significant fluorescence enhancement upon dsDNA hybridization is an

attractive method for development of a biosensor platform where target oligonucleotide labelling

is eliminated. Fluorescence studies of the molecular switches immobilized at the surface of

114

fused silica optical fibers were accomplished by modifying the sample compartment of a PTI

Spectrofluorimeter such that 488 nm argon ion laser radiation (biotinylated thiazole orange λexc =

511nm) could be coupled into an optical fiber. The fiber was held in place using a Teflon sleeve

which suspended the fiber within the cuvette housing (see Figure 25).

Neutravidin can be self-assembled onto a biotinylated surface by affinity interaction using the

remaining biotin binding sites present on one face of the Neutravidin after the molecular switch

was constructed in solution. Fused silica optical fibers were used as the immobilization platform

in this work, and Figure 31 shows the solid-phase synthetic route to create biotin-functionalized

surfaces.

(a)

(b)

Figure 31. Surface functionalization of fused silica optical fibers (a) Reaction of surface hydroxyl groups with APTMS to provide amino groups at the surface (b) Surface functionalization with biotin or iminobiotin to provide surface binding sites for Neutravidin immobilization (X= O biotin, X = N iminobiotin).

115

(a) (b)

Figure 32. Selective, pH-dependent capture of Neutravidin using iminobiotin-functionalized fused silica optical fibers. (a) Capture of TAMRA-labelled Neutravidin on biotin-functionalized fibers as a function of pH. The data demonstrates that capture in acetate buffer (pH 4) and ammonium carbonate buffer (pH 11) is not distinguishable within experimental error (expected). (b) Capture of TAMRA-labelled Neutravidin on iminobiotin-functionalized fibers as a function of pH, demonstrating that iminobiotinylated surfaces are capable of pH dependent capture. Fibers were immersed in 1 μM TAMRA-labelled Neutravidin solutions in pH 4 and pH 11 buffers. Data was corrected for change in fluorescence as a function of pH.

The functionality of the surface immobilized chemistry was verified by utilizing the pH

dependent binding properties of iminobiotin as a comparison (see Figure 32). Iminobiotin allows

pH dependent binding and release of Neutravidin. At high pH values (pH >11), iminobiotin

occupies the biotin-binding sites of Neutravidin. At low pH (~4), this binding is disrupted and

results in release or dissociation of the protein from iminobiotin. The pH-dependent capture of

Neutravidin using iminobiotin-functionalized optical fiber substrates was studied using a

carboxytetramethylrhodamine (TAMRA)-labeled Neutravidin species and monitored using

confocal fluorescence microscopy. The results showed that the surface chemistry was functional

116

as selective immobilization in pH 11 buffer was observed as a higher fluorescence intensity

profile due to greater binding of fluorescently labeled Neutravidin. Release of Neutravidin from

the fiber surfaces was also achieved through manipulation of pH (data not shown). A comparison

to biotinylated fibers which were not expected to show a pH dependent binding trend showed

that the biotinylated fibers did not exhibit selective binding.

Fluorescence studies of the immobilized molecular switches on fused silica optical fibers were

accomplished in an intrinsic excitation and fluorescence recovery mode. Non-specifically

adsorbed material was removed by low-power sonication in 0.1×PBS buffer for 5 min. A

significant fluorescence signal from intercalated thiazole orange contained in the switch

construct was evident. A red shift in the fluorescence emission profile occurred, which is

consistent with the work published by Wang and Krull for thiazole orange immobilized at an

interface.102

In order for the switch to be activated for the detection of target oligonucleotide, the dsDNA

used initially for site-directed templating of the dye was denatured using two different methods:

a chemical denaturation and a thermal denaturation method. Chemical denaturation was

accomplished by sonicating the switch-coated fibers in a 4 M urea solution for 5 min. This

denatured the DNA hybrids at the interface, leaving the robust biotin–Neutravidin chemistry

intact, and the biotinylated probe oligonucleotide in single stranded form adjacent to a free TO

molecule on its linker. It has already been shown by Lu et al. that chemical denaturation using 4

M urea of dsDNA hybrids containing a biotinylated probe oligonucleotide bound to streptavidin,

which was in turn immobilized onto a biotin-functionalized optical fiber, still allowed six cycles

of use.48,274-276 This suggests that the integrity of the surface-immobilized

streptavidin/biotinylated probe oligonucleotide was retained despite the seemingly strong

denaturing conditions. Both chemical and thermal treatments were investigated by Lu et al., and

the fluorescence intensity of probe/target hybridization (as signaled by fluorescently-labeled

target oligonucleotide) was found to be only marginally higher upon probe/target hybridization

over the six cycles of use for the thermal regeneration. In the work presented here, both chemical

and thermal treatments were used as methods to denature the templating target strand and to

generate the immobilized molecular switches.

117

Fluorescence intensity data from optical fibers coated with activated molecular switches in

response to the presence of target after using the chemical activation treatment were collected.

The data were calculated using Equation 26, where F(λ) represents the fluorescence intensity

recorded at 530 nm (the fluorescence emission maximum of TO-bio1 and TO-bio2 at the

interface). Subsequent denaturation of the dsDNA template led to a decrease in the fluorescence

signal intensity, as the dye was no longer associated with a DNA hybrid and therefore underwent

a decrease in quantum yield. Removal of additional non-specifically adsorbed material at the

surface was also a probable contribution to the overall decrease in signal intensity observed, and

would be superimposed on the effect of denaturation of the double-stranded hybrids (data not

shown). Immersion of the fibers prepared using the chemical activation method in a solution of

target oligonucleotide (ssDNA) complementary to the probe sequence subsequently lead to a

two-fold increase in the fluorescence signal collected as shown in Figure 33.

(a) (b)

Figure 33. Fluorescence data from optical fibers coated with molecular switch using TO-bio1. Fibers were immersed in a 1 uM solution of switches for 12 hrs and subsequently rinsed in PBS buffer. (a) Signal from switch-coated fibers generated by sonicating fibers in 4 M urea solution for five minutes to generate ssDNA probes at the surface. (b) Signal after fibers immersed in 5 uM target ssDNA solution (5 hrs). Data was averaged over multiple fibers, and error bars represent the standard deviation (1σ) associated with the data set.

However, in such approaches, arguably the most important consideration is the maximization of

the signal-to-noise ratio between bound and unbound target states.277 A thermal

118

denaturation/cycling method was used to see if an improvement in the signal-to-noise for

probe/target hybridization could be achieved. Two different methods of constructing the

molecular switch were also tested. In the first method, dsDNA with intercalated TO-bio2 was

prepared first, and was then added dropwise to a solution of Neutravidin (all components were at

concentrations of 1 μM). The switch constructs were then immobilized onto the surfaces of the

biotinylated optical fibers. In a second approach, TO-bio2 was incubated with Neutravidin (both

1 μM in concentration), followed by the subsequent addition of 1 μM dsDNA (the probe strand

of the duplex contained a biotin functionality at the 5′-terminus). The complexes were then

immobilized on the surface of the biotinylated optical fibers. For thermal denaturation of switch

scaffolding duplexes, the optical fiber surface was immersed in buffer (0.1 M PBS, 20 mM

NaCl, pH 7.4) at 65 °C for 20 min, followed by a 45 °C thermal cycle for 15 min (0.1 M PBS, 1

M NaCl, pH 7.4). These two different types of switch-activated optical fibers were then

immersed in a solution containing 1 μM target oligonucleotide. Using the chemical denaturation

protocol, an approximate two-fold signal increase upon target binding was observed using a 5

μM unlabeled target solution in 0.1 M PBS buffer containing 0.1 M NaCl (see Figure 33). The

results shown in Figure 34 demonstrate significant improvement in signal enhancement upon

target binding in comparison to the chemical denaturation step using 4 M urea, with a five-fold

enhancement of signal seen for 1 μM target oligonucleotide. The time course of signal

development suggests that there is initial nonspecific adsorption of the dye species on the surface

of Neutravidin, and that, over time, this adsorptive interaction is replaced with the interaction of

the dye species with the nucleic acid present. Comparison of the results obtained using TO-bio1

and TO-bio2 (which had the longer tether length: 26-atom tether vs. 18-atom tether) indicated

that an increase in linker length further contributed to the increased fluorescence signal in the

presence of complementary target, which may be due to the improved availability of TO dye by

virtue of the reach of the flexible linker.

119

Figure 34. a–b Graphs showing fluorescence intensity at 530 nm (fluorescence maximum for TO-bio2 at the interface). (a) Switch assembled using site-directed templating method: i, switch-coated fibers after thermal denaturation of templating strand; ii, switch-coated fibers immersed in 1 μM target solution for 40 min; iii, switch-coated fibers immersed in 1 μM target solution for 4.5 h. (b) Switch assembled by association of dye/ Neutravidin, followed by titration of dsDNA into solution: i, switch coated fibers after thermal denaturation/cycling; ii, switch-coated fibers immersed in 1 μM target solution for 40 min; iii, switch-coated fibers immersed in 1 μM target solution for 4.5 h. Data was averaged over multiple fibers, and error bars represent the standard deviation (1σ) associated with the data set.

4.4 Conclusions Preliminary work is reported about the construction of a novel molecular switch based on

biomolecular self-assembly to create a functional sensing platform for oligonucleotides. The

focus has been on examining the functionality of the switch in a solid-phase assay environment,

with the intention being that optimizing the selectivity, sensitivity and speed are aspects that will

follow. The switch assembly was constructed using Neutravidin as a docking molecule. Two

biotinylated components consisting of hybridized probe oligonucleotide and TO on a

biotinylated linker were first allowed to associate. The complex of dsDNA and intercalant was

then exposed to Neutravidin so that the two biotin functionalities of the complex would assemble

side-by-side on one face of the protein. Subsequently, the biotin-binding sites on the other face

of the Neutravidin were used to attach the switches to a solid interface. The success of the work

120

presented here suggests a design platform that can be used for the further development of simple

nanoscale constructs that can be used for bioanalytical applications.

121

5 Isothermal Hybridization Selectivity Enhancement Using Formamide as a Chemical Agent to Influence Duplex Stability

Abstract

The molecular switching strategy outlined in Chapter 4 makes use of a novel, biotinylated intercalating dye moiety. The introduction of a biotinylated tether to an intercalating dye has the potential to provide a facile method of templating DNA-sensitive dyes to an interface that possesses conformational mobility and a means of detecting hybridization at an interface. To this end, the selectivity of hybridization using the biotinylated dye molecule has been investigated and is the theme for this final chapter. An investigation of SNP discrimination has been evaluated in solution as a platform for which future work at an interface can be developed. Selectivity for single nucleotide polymorphism (SNP) detection using a variety of target lengths (19 base and 34 base oligonucleotide sequences and 141 base-pair PCR amplicons) has been achieved using a biotinylated thiazole orange (TO-bio2) derivative as a signalling agent and formamide as a denaturant to influence duplex stability. The ability of thiazole orange to distinguish between single-stranded and double-stranded DNA is enhanced in the presence of formamide due to changes in binding modes associated with the dye/DNA structure. In the absence of formamide, a signal increase of approximately 2-fold for dsDNA over ssDNA was observed, and upon addition of formamide, this signal increase was increased to approximately 8-fold. This performance improvement allowed the system to achieve SNP discrimination for both oligonucleotide targets and PCR amplicons. Analysis was done in as little as ten minutes for oligonucleotide targets. Short-probe oligonucleotides used to detect hybridization in PCR samples were found to be both an effective method for blocking strand re-annealing of PCR products, and provided SNP discrimination when combined with TO-bio2 and formamide. An increase in the ratio of probe:PCR amplicon from 4:1 to 8:1 increased the difference in signal magnitude between fully complementary and 1 base pair mismatch (1 bpm) samples. Fluorescence lifetime measurements have shown that the presence of formamide alters both the fluorescence lifetimes of TO-bio2 and the relative ratios of the contributions of bi-exponential lifetimes associated with TO-bio2 binding to dsDNA for both complementary and three base pair mismatch (3bpm) samples. The addition of formamide to samples containing ssDNA and TO-bio2 resulted in a mono-exponential fluorescence lifetime comparable to free TO-bio2 dye in solution, showing that secondary binding modes may be controlled through the addition of formamide.

122

5.1 Introduction Nucleic acid diagnostic technology is of great clinical and environmental interest. It has

numerous applications in areas such as genetic disease screening, detection of pathogens in water

supplies and foodstuffs, and as a potential detection method for bio-threat agents. According to a

SNP map of the human genome constructed by the International SNP Map Working Group, there

are a reported 1.42 million single nucleotide polymorphisms (SNPs) that are distributed

throughout the human genome. These have specific implications in inherited differences and can

contribute to anthropological characteristics, disease risk and environmental response.23 The

successful implementation of most nucleic acid hybridization platforms (sensors, arrays and

assays) relies on the ability of the system to discriminate base-pair mismatches between probe

and target oligonucleotide sequences in a simple, rapid, and inexpensive way.278 DNA duplex

stability can be readily used to identify mutations present in a sample, and hybridization stability

can be altered through either thermal or chemical means. Many assay methods are predicated on

the use of thermal denaturation profiles to differentiate between fully complementary and

mismatched target oligonucleotides.189,279 Fluorescence intensity can be used for monitoring the

presence of double-stranded DNA. While fluorescence is certainly effective as a method for SNP

discrimination, denaturation must then be corrected for differences in fluorescence intensity as a

function of temperature. The sample analysis process also requires slow temperature changes

(typically 0.1-0.5 °C/min) in order to maintain equilibration of the system during

measurement.280 To overcome this limitation of speed, kinetically-driven methods are being

developed to achieve more rapid analysis times. More recently, chemical methods to increase

selectivity have begun to grow in popularity.280,281 Formamide and urea are able to destabilize

DNA duplexes by competing with Watson-Crick base pairing.282 In particular, formamide can

lower the thermal melt temperature (Tm) of a duplex through disruption of hydrogen bonding

motifs present or by alterations of hydration patterns of double-stranded DNA.282 Because

formamide does not chemically degrade nucleic acids at room temperature, it is frequently used

to lower Tm values for DNA in PCR reactions to avoid DNA degradation. The lowering of Tm of

DNA duplexes using formamide is reported to range from 0.60 to 0.72 °C/% formamide in

solution283-285, or 2.4-2.9 °C/mole formamide282 for DNA duplexes of significant lengths (>800

base pairs). Fuchs et al.281 have recently reported an average lowering of Tm for immobilized

oligonucleotides (16-mer) of 0.58 ± 0.05 °C/% formamide.

123

Ideally, implementation of detection strategies based on hybridization would rely on methods of

analysis that offer rapid analysis, and that minimize reagent consumption and steps in the overall

analysis. A simple, functional fluorescence method to rapidly monitor selectivity in solution at

room temperature has been implemented here using biotinylated thiazole orange. The concept is

based on the ability of formamide to lower the melting temperature of DNA double strands,

coupled with selectivity of thiazole orange for intercalation with dsDNA. In order for a

fluorescent dye to provide sensitive and specific DNA detection, the dye should exhibit low

fluorescence intensity associated with DNA in the single-stranded state, a large intensity change

upon hybridization with target material, and an ability to show a low fluorescence intensity for

partially mismatched targets, ideally with SNP discrimination capability. Thiazole orange is an

asymmetric cyanine dye which binds to DNA through intercalation and electrostatic interaction

of the cationic dye centre with the negatively charged, polyanionic backbone of DNA. The

monomeric form of TO shows a distinct advantage over the more commonly used intercalating

dye ethidium bromide since TO shows a much larger fluorescence enhancement in the presence

of dsDNA (ethidium bromide shows ~twenty fold increase in fluorescence intensity whereas TO

has been reported to show ~3000 fold increase) and also shows preferred selectivity between

ssDNA and dsDNA.234,286,287 Monomeric TO is also advantageous over its homodimeric bis-

intercalative derivative TOTO, since TOTO has similar affinity for single- and double-stranded

DNA. Despite the fluorescence intensity differences associated with TO for ssDNA and dsDNA,

there is still significant interaction associated with ssDNA as shown in Chapter 4. In this work,

it was proposed that the contributions to the overall fluorescence intensity signals arising from

non-intercalative modes of TO association with DNA could be reduced or eliminated by the

presence of formamide in solution.

The simplicity of SNP analysis when using formamide arises from several key features: i) the

analysis can be done at room temperature and therefore eliminates the requirement for additional

heating apparatus as well as time consuming temperature ramping to establish equilibrium

conditions; ii) fluorescence monitoring can be achieved without the requirement for temperature-

corrected data sets based on decreases in fluorescence as a function of increased temperature; iii)

a complicated FRET system is not required for analysis as the system is based on one

intercalating dye capable of discrimination between single and double-stranded DNA sequences

for which formamide further enhances the signal differences between ssDNA and dsDNA; iv)

124

this can be implemented to avoid need for labeling of probe and target strands; v) the system

responds quickly, with analysis completed in minutes.

To demonstrate the capability of the method, changes of fluorescence intensity upon addition of

formamide were monitored for fully complementary hybrids, 1 bpm, 2 bpm, 3 bpm and non-

complementary 19-mer double-stranded DNA constructs in the presence of biotin-tethered TO

intercalating dye. Figure 35 shows a representation of the binding mode changes associated with

the addition of formamide for TO-bound ssDNA and dsDNA.

Figure 35. Schematic of proposed representation of TO-bio2 binding motifs for both ssDNA and dsDNA and associated changes upon addition of formamide.

125

SNP discrimination using 19-mer and 34-mer oligonucleotide targets, as well as for a 141 base β-

actin PCR amplicon generated using symmetric PCR was achieved. Because formamide disrupts

duplex formation, the hybridization efficiency for both fully matched and mismatched duplexes

decreased as the percentage of formamide increased. The addition of formamide led to

fluorescence quenching of the dye for both fully matched and mismatched systems. The

fluorescence intensity changes could be due to the disruptions of bonding interactions between

the dye and backbone of the DNA duplex. This would lead to restoration of rotational freedom of

the dye, resulting in a decrease in fluorescence intensity due to non-radiative modes of relaxation

achieved through collisional deactivation. However, the relative fluorescence intensity between

the matched and mismatched system also changed as a function of percent formamide added.

The relative change was maximized at 18.5% formamide, where the complementary duplex

showed approximately a 4.1 fold difference in signal intensity when compared to the mismatched

system. These results demonstrated advantages over traditional methods of mismatch detection,

allowing operation at room temperature without incorporation of fluorescent signalling dyes into

oligonucleotide probe or target strands. Furthermore, base-pair discrimination could be easily

optimized by altering the quantity of formamide added to sample solutions as a function of

factors that affect duplex stability such as GC content. Time-resolved fluorescence decay

measurements of fully complementary and mismatched duplexes provided further insight into

the changes in microenvironment that were associated with the addition of formamide. Of

particular significance is the ability of the combination of formamide and thiazole orange to

effectively discriminate SNPs in symmetric PCR samples. Re-annealing of PCR strands after

application of heat to separate the double-stranded symmetric PCR products into single-stranded

target would compete with probe binding. A rapid detection of SNPs was demonstrated in such

a competitive binding environment.

5.2 Experimental Methods

5.2.1 Reagents

All chemicals were reagent grade or better and used without further purification. Reagents used

for thiazole orange synthesis, including methyl iodide, 2-(methylthio)benzothiazole, 4-

methylquinoline, triethylamine, N-hydroxysuccinimide (NHS), N,N′-diisopropylcarbodiimide

99.5%, N,N-diisopropylethylamine (Hünig’s base), and anhydrous N,N-dimethylformamide

126

(99%) were from Sigma-Aldrich (Oakville, ON, Canada). The synthetic details for dye synthesis

are outlined in section 4.2.1 of Chapter 4. Formamide (99.5%) was also from Sigma-Aldrich.

(+)-Biotinyl-3,6-dioxaoctanediamine (EZ-Link® amine-PEG2-biotin) and Neutravidin were from

Thermo Scientific (Rockford, IL, USA) through Fisher Canada (Nepean, ON, Canada). Reagent-

grade toluene, ethanol, acetone, isopropanol, methanol, dichloromethane, and anhydrous diethyl

ether were from EM Science (Toronto, ON, Canada).

5.2.2 Instrumentation

Ultraviolet-visible absorption spectra were measured using a Hewlett Packard 8452 Diode-Array

Spectrometer (Hewlett Packard Corporation, Palo Alto, CA, USA). Solution phase steady-state

fluorescence spectra were measured using a QuantaMaster PTI Spectrofluorimeter equipped with

Felix Software. Time resolved fluorescence decay profiles were collected with a PTI Laser-

Strobe™ system using a dye (Coumarin 503, 6 mM, in ethanol) laser (GL-302, Photon

Technology International) tuned to 511 nm, which was pumped by a pulsed nitrogen laser (GL-

3300, Photon Technology International, Lawrenceville, NJ, USA). Coumarin 503 laser dye was

from Exciton (Dayton, OH, USA). TimeMaster (Photon Technology International) decay fits

were generated using Felix software. Steady state fluorescence emission spectra were collected

from 521 to 650 nm. PCR was performed using a Bio-Rad iCycler Thermal Cycler (Bio-Rad

Laboratories Headquarters, Hercules, CA, USA).

5.2.3 Use of formamide to control selectivity

The oligonucleotide sequences in Table 12 were obtained from Integrated DNA Technologies

(Coralville, IA, USA) and dissolved in 0.1 M PBS buffer at pH 7.4 (100 mM PBS, 20 mM

NaCl), with the exception of the β-actin PCR amplicon which was generated in-house. Several

clinically relevant sequences were used for the work, and the probe and target oligonucleotides

are listed in Table 12. Probe A was complementary to a portion of the H. sapiens survival motor

neuron protein coding gene (SMN 1).189 Probe B was complementary to a portion of the H.

sapiens β-actin gene and Probe C was a probe containing a single base pair mismatch. The target

sequences for Probe 1 were fully complementary (both a 19 base pair target equal in length to the

probe and a 34-mer ‘long’ target) and those containing one, two and three base pair mismatches.

The two sequences that were non-complementary to the Probe A sequence were a T20

homopolymer and a mixed base sequence which corresponded to a portion of the S. Enterica

127

invasion protein A coding gene (inv A).14,288 The target sequence for Probe B was a 141 base

pair PCR amplicon associated with the H. sapiens β-actin gene. All subsequent dilutions were

prepared with 0.1 M PBS buffer. Solutions containing a 1:1 ratio of Probe A and the different

target oligonucleotides were heated at 95 °C for 5 min and cooled to room temperature to

generate dsDNA. Formamide was added to buffered solutions of dye and DNA according to

percentages and times outlined in the results and discussion section. Sample solutions involving

the PCR amplicons are described in the following section for clarification.

Table 12. Oligonucleotide sequences used in selectivity experiments.

ProbeA 5’-biotin-ATT TTG TCT GAA ACC CTG T -3’

Targets

fully complementary (equal length target 19-mer)

5’-ACA GGG TTT CAG ACA AAA T-3’

fully complementary (long target 34-mer)

1bpm (long target 34-mer)

5’-TCC TTT ATT TTC CCT T ACA GGG TTT CAG ACA AAA T-3’

5’-TCC TTT ATT TTC CTT T ACA GGG TTT CAC ACA AAA T-3’

1 bpm 5’-ACA GGG TTT CAC ACA AAA T-3’

2 bpm 5’-ACA GAG TTT CAG ACG AAA T -3’

3 bpm 5’-ACA GGG AAA CAG ACA AAA T-3’

T20 (NC) 5’-TTT TTT TTT TTT TTT TTT TT-3’

SAL target (NC) 5’-ATC CAC AGA AGA ATC CAG A-3’

ProbeB

fully complementary for β-actin PCR amplicon

5’-CCC TCC CCC ATG CCA TCC T-3’

ProbeC

1bpm β-actin PCR amplicon

5’-CC TCC CCC ATG CCA CCC T-3’

Target 141-base PCR amplicon containing 3’-GGG AGG GGG TAC GGT AGG A-5’ segment

128

5.2.4 PCR amplification of β-actin gene fragments

A 141 base-pair PCR product was amplified from β-actin template DNA using the following

primer pair; 5’-TCA CCC ACA CTG TGC CCA TC-3’ (forward primer) and 5’-GTG GTG

GTG AAG CTG TAG CC-3’ (reverse primer). Symmetric PCR amplification was performed

using a pre-programmed thermal cycle as follows: a preheating (95 °C for 5 min) step followed

by 45 cycles (at 95 °C for 60 s, 61 °C for 30 s, and 72 °C for 30 s), and a final extension at 70 °C

for 10 min in 100 μL of reaction mixture containing 0.02 μg/mL of template DNA, 1x buffer (10

mM tris-HCl (pH 8.0), 50 mM KCl, 0.08% Nonidet P40), 1.5 mM MgCl2, 0.5 μM of each

primer, 25 nM dNTPs, and 2.5 units of Taq polymerase (Fermentas Life Sciences Canada Inc,

Burlington, ON). PCR products of 141 bp were run on agarose gels (1% w/v) pre-stained with

SYBR Gold (0.8x). Gels were run at 100 V in 1x TE buffer (100 mM Tris, 2 mM EDTA) for 60

min and examined under ultraviolet illumination. PCR products were purified using a Qiagen

PCR purification kit (Qiagen Inc., Mississauga, ON, Canada) and samples were quantified using

UV-visible absorption spectroscopy (λ = 260 nm) using molar absorptivity (ε) values calculated

using an in-house software program developed by Dr. Paul Piunno based on the methods of

Puglisi and Tinoco.168

5.2.5 Selectivity experiments using PCR amplicons

PCR amplified samples were exposed to thermal denaturation (98 °C for 10 min) to obtain

single-stranded forms. Generation of ssDNA using a modified method based on those developed

by Wang et al.289 was found to be an effective method to obtain ssDNA for hybridization

experiments from a symmetric PCR amplification process.289,290 The principle of the method

relies on the use of short oligonucleotide sequences which are complementary to some sequences

on the amplified strand. When added during the denaturating process, short strands hybridize

with the single-stranded PCR strand and suppress PCR strand reannealing. The denaturation

procedure used here involved 10 min incubation at 98 °C for the PCR products, followed by

removal of the samples from heat and addition of excess probe oligonucleotide (see Table 12).

The ‘blocking strands’ used were the probe oligonucleotides of interest. The mixture was then

incubated at 95 °C for three minutes, allowed to cool, and dye was added. The concentration

ratios used were 4:1 and 8:1 for probe oligonucleotide:PCR amplicon and 10:1 dye:PCR

amplicon. Sample concentrations of  PCR amplicons were 0.24 μM in a 50 μL volume (12

pmol).

129

5.3 Results and Discussion Initial studies of the changes of fluorescence intensity as a function of percent formamide in

solution were conducted for fully complementary and 2 bpm double-stranded DNA constructs in

the prescence of TO intercalating dye (see Figure 36).

130

(a)

(b)

Figure 36. (a) TO-bio2 (1μM) in the presence of unlabelled, fully complementary dsDNA (1μM) (black) and unlabelled dsDNA containing 2 bpm (1μM) (grey) as a function of % formamide in 0.1 M PBS buffer (20 mM NaCl) solution. Inset: Results from formamide concentrations ranging from 6.6 % to 22 % to show the magnitude of signal change as a function of increasing formamide concentration. Fluorescence measured at room temperature. (b) Data set plotted to illustrate shape of transition as a function of % formamide in solution (precision ±5%). Dye and DNA concentrations were set at 1μM to provide a 1:1 dye:DNA ratio.

131

The data shown in Figure 36 (b) shows the change in relative fluorescence as a function of added

formamide. The initial steep decrease in fluorescence intensity shows that the initial addition of

as little as 2% formamide in solution affects the TO-bio2 fluorescence for both fully

complementary as well as 2 bpm duplexes. The initial sharp decrease in fluorescence intensity

may be explained by the mechanism of formamide denaturation of DNA duplexes proposed by

Blake and Delcourt282, which suggests that the denaturing effects of formamide are a result of

changes in the ligand association constants associated with water and counter ion binding to both

helical (dsDNA) and coil (ssDNA) states of nucleic acids. The initial sharp decrease can be

attributed to the exchange of water molecules and counter ions associated with the double helix

with formamide.282 The work of Blake and Delcourt showed that formamide acts to destabilize

the double helix by displacement of weakly bound water (hydrate) molecules, and water

molecules occupying groove sites are most readily exchangeable which leads to a change in ionic

potential and consequently changes the sodium ion (counter ion) associated with both helical and

coil states. This may explain the initial drop in fluorescence intensity associated with TO-bio2.

Since the parent dye structure is cationic in nature dye, TO-bio2 would act as a counter ion to the

negatively charged polyanionic backbone of DNA (both in the helical double-stranded state and

coiled single-stranded state). Thus, both electrostatic interactions as well as intercalative modes

of binding would contribute to the overall fluorescence signal intensity associated with dye/DNA

interactions. If formamide is able to displace counter ions from the DNA backbone, this would

essentially release the dye species into solution and restore rotation about the monomethine

bridge, thus reducing the radiative modes of decay by increasing non-radiative collisional

deactivation modes of decay.

The selectivity of hybridization was further examined using different probe/target strand

combinations. Experiments were conducted using an initial concentration of 18.5% formamide

in solution (rather than a titration method as used in the previous experiment) in order to

investigate both signal-to-noise enhancements and improvements in speed of analysis. Figure 37

shows the ability of the system to discriminate fully complementary targets from 2 bpm,

mismatched homopolymer (T20) and mixed base non-complementary targets.

132

Figure 37. TO-bio2 in the presence of (a) unlabelled, fully complementary dsDNA, (b) unlabelled dsDNA containing 2 bpm (c) unlabelled probe with non complementary T20 homopolymer and (d) unlabelled probe with a mixed-base non-complementary sequence. Sample concentrations were 1 μM dsDNA (ProbeA with targets as indicated), 1 μM TO-bio2 in 0.1 M PBS buffer (20 mM NaCl) containing 18.5 % formamide. Steady state fluorescence signals were measured at room temperature after an incubation time of 10 min with formamide.

The data represented here shows that fully complementary dsDNA can be readily distinguished

from a variety of different mismatched targets. The results were consistent with the previous

data set, showing an approximate four-fold change in signal magnitude of the fully

complementary sequence over mismatched targets at 18.5 % formamide. The magnitude of the

change in fluorescence intensity for TO-bio2 with mismatched targets in the presence of

formamide is higher than that observed for TO-bio2 in the absence of formamide. This indicates

that introduction of formamide is able to increase signal-to-noise. This is likely due to

suppression of background fluorescence associated with external binding modes of the dye to

ssDNA (associations with the nucleobases or phosphate backbone). While the addition of

formamide does lead to a decrease in fluorescence emission intensity for fully complementary

dsDNA as a result of the loss of dyes bound via external binding modes, the percent decrease in

fluorescence intensity for dsDNA is less than that for ssDNA, confirming that the introduction of

formamide can improve signal-to-noise as well as provide SNP discrimination. Figure 38

illustrates the ability of formamide to suppress background fluorescence associated with TO-bio2

binding to ssDNA.

133

Figure 38. Fluorescence intensity of TO-bio2 in the presence of a) MCR2 probe ssDNA; b) MCR probe ssDNA + 25 % formamide; c) free TO-bio2. MCR2 probe sequence 5’-TTC AGC CTG TCT GTG ATT GC-3’.

The incubation time was then reduced to five minutes to examine the consequence of speeding

the analysis. It was shown that discrimination between fully complementary targets and those

containing two basepair mismatches as well as non complementary sequences (both mixed base

and homopolymer T20) could still be achieved. Increasing stringency by an increase in the %

formamide added (from 18.5% to 37%) allowed for SNP analysis (see Figure 39).

134

Figure 39. TO-bio2 in the presence of unlabelled, fully complementary dsDNA (FC), unlabelled dsDNA containing one base pair mismatch (1bpm) and unlabelled dsDNA containing three base pair mismatches (3bpm). dsDNA corresponds to ProbeA and corresponding target as indicated in Table 12. Sample concentrations were 1 μM dsDNA, 1 μM TO-bio2 in 0.1 M PBS buffer (20 mM NaCl) containing 37 % formamide. Steady state fluorescence signals were measured at room temperature after an incubation time of ten min with formamide. Measurements were performed in triplicate and error bars represent 1 σ.

Steady state fluorescence intensity at λmax of 526 nm (maximum emission wavelength of TO-

bio2) was monitored over a 10 min period (fluorescence intensity measured at 30 s intervals for

10 min) for both fully complementary duplexes and those containing 2 bpm after addition of

formamide. These results showed no significant change in fluorescence intensity for both

samples over the time frame of measurement. This may suggest that an even shorter analysis

time (<5 min) could be achieved under the proper conditions given that the fluorescence signal

change here occurs within the time frame required to prepare the sample and record a t=0

fluorescence measurement (see Figure 40).

135

Figure 40. Fluorescence intensity as a function of time for TO-bio2 in the presence of (a) unlabelled, fully complementary dsDNA; (b) unlabelled dsDNA containing two base pair mismatches (2bpm). Both samples contain 18.5 % formamide which was added immediately prior to measurement. dsDNA corresponds to ProbeA and corresponding target as indicated in Table 12.

The mechanism of signal transduction for fluorescent probes can rely on several different

processes. The use of time-resolved fluorescence spectroscopy can provide further insight into

complex systems and provides information about populations of the fluorophore that are

experiencing different local environments.291 Fluorescence lifetime analysis is a tool that can

provide information about molecular features of the sample, such as the existence of two or more

conformations or binding states.292

Time-resolved fluorescence data of monomeric cyanine dyes such as TO, BO, YO and PO in the

absence of formamide have shown that multiple binding modes exist between these dyes and

nucleic acid sequences. Netzel et al.93 have shown that the monomeric cyanine dyes BO-PRO-

1, PO-PRO-1, YO-PRO-1 and TO-PRO-1 exhibit either bi- or tri-exponential emission decay

kinetics, which reflects the presence of different dye/dsDNA modes of binding. Bi-exponential

decay curves have also been observed for TO-bio2 in the presence of dsDNA.53 It is postulated

136

that the use of formamide in solution at low percentages may be able to shift the population of

dye molecules in the various binding modes of TO-bio2, with the intercalative mode being

dominant.

Fluorescence lifetime data was collected in order to further understand the fluorescence

processes associated with the selectivity achieved when using formamide as a denaturant.

Thiazole orange is a monomeric cyanine dye and several cyanine dye analogues have been

reported in the literature to show both monoexponential and bi-exponential decay kinetics in the

presence of short oligonucleotide sequences.93 Such fluorescence intensity decay profiles extend

to a tri-exponential fit in the presence of CT-DNA as shown by Netzel et al.93 The

multiexponential decay models show that there are multiple binding/association modes of this

particular class of DNA-binding dyes with DNA. This would be expected due to the combination

of intercalative binding and the cationic nature of the dye, which would lead to external modes of

binding associated with the polyanionic backbone of nucleic acids.

Time-resolved fluorescence decay profiles of fully complementary probe/target hybrids in the

presence and absence of formamide were analyzed to determine if a change in the distribution of

binding modes associated with the dye occurred upon the addition of formamide. The salt

concentration was kept at a relatively low value of 20 mM in order to maintain duplex stability

and maximize the observed equilibrium binding constant of the dye through increased

electrostatic affinity of the dye to dsDNA.236 It was found that there was a shift from a multi-

exponential decay associated with the duplex in the absence of formamide to a mono-exponential

decay profile associated with the duplex in the presence of formamide. Table 13 shows the

change in fluorescence lifetime and exponential prefactor(s) for TO-bio2 associated with

complementary and 3bpm dsDNA duplexes (see Appendix II for fluorescence intensity decay

profiles, curve fitting and residuals). The fluorescence lifetime associated with the mono-

exponential fit suggests that intercalative binding modes remain after introduction of formamide,

while external binding modes are reduced. This result confirms the complexities that exist in data

interpretation for the general use of TO as a fluorescence probe for nucleic acids and that

significant background signals are present when using TO.

137

Table 13. Fluorescence Lifetime Changes of TO-bio2 Associated with dsDNA in the Presence and Absence of Formamide. τ1 (ns) τ2 (ns) α1/α2 χ2 (range for data set) TO-bio2 + dsDNA 3.5±0.2 1.3±0.6 2.2 0.920-1.168 (1:1 dye:duplex ratio) TO-bio2 + dsDNA 2.9±0.1 - - 1.096 18.5% formamide (1:1 dye:duplex ratio) TO-bio2 + dsDNA 3.6±0.1 1.4±0.4 1.9 1.238-1.910 (2:1 dye:duplex ratio) TO-bio2 + dsDNA 2.8±0.1 - - 1.106-1.321 18.5% formamide (2:1 dye:duplex ratio) TO-bio2 + dsDNA 3bpm 3.4±0.1 - - 1.988-3.453 (1:1 dye:duplex ratio) TO-bio2 +dsDNA 3bpm 4.2±0.6 1.8±0.5 0.4 1.039-1.267 18.5% formamide (1:1 dye:duplex ratio) TO-bio2 + dsDNA 3bpm 4.3±0.5 0.7±0.1 0.1 0.808-1.138 37% formamide (1:1 dye: duplex ratio) TO-bio2 dye 0.5±0.1 - - 0.8811-2.219 TO-bio2 + ssDNA (1:1 dye: oligonucleotide ratio) 1.1±0.1 - - 1.554-1.749 TO-bio2 + ssDNA 0.5±0.1 - - 0.8764-.9661 18.5% formamide (1:1 dye: oligonucleotide ratio)

As shown in Table 13, the fluorescence lifetime decay profiles are sensitive to changes in the

local environments of TO-bio2. For TO-bio2 in the presence of fully complementary dsDNA,

the fluorescence lifetime decay profiles are affected by both the dye:DNA ratio as well as the

138

addition of formamide to a buffered solution of DNA and dye. The fully complementary

sequence shows a bi-exponential decay profile in the absence of formamide that reflects binding

of dye through both external binding modes and intercalation. Molecular modeling calculations

for TO-PRO-1 and TO-MET (TO derivative containing an iodohexyl side chain) done by

Prodhomme et al.286 have shown that the thiazole orange parent dye ring system intercalates

completely between the base pairs of DNA helix and the side chain occupies the minor groove

with the benzothiazole and quinolinium rings stacking with the bases. The stacking contribution

predominates over the electrostatic contribution, and intercalation is not significantly altered by

changes in the side chain identity.286 The longer lifetime (τ1) component that is typically

associated with the intercalative binding modes for TO-bio2 provides a larger contribution to the

overall fluorescence lifetime as shown by the ratio of pre-exponential factors, α1/α2, showing that

a tether of significant length containing a biotin docking moiety attached to the dye can provide

intercalative binding to dsDNA. Netzel et al.93 have reported bi-exponential decay kinetics

associated with short dsDNA duplexes for TO-PRO-1, which reflect multiple modes of

dye/DNA binding. The results presented as part of this thesis show trends similar to those

reported by Netzel et al. where the longer lifetime component shows a larger contribution as

reflected in the pre-exponential factors associated with each lifetime value. Upon increasing the

ratio of dye to DNA to 2:1, a shift in the contributions of each fluorescence lifetime component

was observed. These results agree with previously published results (see Chapter 3) indicating

that an average of one TO dye molecule intercalates into each 19-mer duplex associated with the

SMN1 probe/target oligonucleotide sequence, with an overall lifetime value of 2.6 ns.10 As the

ratio of dye to DNA was increased, the contribution of the shorter lifetime component associated

with external modes of binding increased, and a shift in α1/α2 from 2.2 associated with a 1:1

dye:DNA ratio to 1.9 for α1/α2 with a 2:1 ratio of dye:DNA was observed. This would be

expected if the dye:DNA ratio was increased to greater than a 1:1 ratio. If one dye molecule

intercalates into the duplex, it would be expected that if the ratio was increased, an increase in

the external modes of binding would be observed and would show a higher contribution to the

overall lifetime signature. While fluorescence lifetime is not sensitive to concentration, if the

binding modes are shifted as a result of an increase in one binding mode over another as

influenced by a concentration increase, the ratio of the contributions of the lifetimes would be

affected. This is reflected in the change in the ratio of pre-exponential factors (α1/α2 ) while the

lifetime values associated with each of the two lifetime components associated with α1 and α2

139

remain similar within experimental error. Upon addition of 18.5 % formamide to fully

complementary duplexes stained with TO-bio2, the fluorescence lifetime reduces to a mono-

exponential lifetime. It is postulated that the external modes of binding are reduced upon

introduction of formamide, while the intercalative mode of binding is retained. There is also a

decrease in the emission lifetime associated with the intercalative mode of binding upon addition

of formamide as a result of disruption of duplex stability. This decrease in duplex stability could

provide an environment where dye molecules may experience increased torsional rotation about

the monomethine bridge of TO-bio2 which would not be present in a fully matched duplex in the

absence of formamide. A formamide concentration of 18.5 % does not correspond to a

suppression in melt temperature to that below the Tm for the duplex. Therefore, duplexes may

exhibit some destabilization, and would show a corresponding reduction of the bi-exponential

lifetime associated with TO-bio2 in the absence of formamide to a mono-exponential lifetime in

the presence of formamide. This mono-exponential emission lifetime is observed for both a 1:1

dye:DNA ratio and a 2:1 dye:DNA ratio in the presence of formamide, further supporting the

notion that external modes of binding are being affected. Linear dichroism experiments by

Larsson et al.253 showed that externally bound YO dyes are oriented with angles between their

long axes (the directions of their transition moments) and the DNA helical axis which are less

than the corresponding angle for the intercalated molecules, which would be expected to be

approximately 90° (dyes oriented parallel to nucleobases and perpendicular to the helical

axis).253,293 Computational models of TO generated by Ediz et al.294 have shown that TO has a

critical interplanar twisting angle of 60° where the dye decays non-radiatively to the ground

state. The results of the work presented in this thesis are consistent with the proposal that a

planar structural geometry is associated with emissive states while a twisted structure is related

to non-emissive states. It therefore is plausible that modes of external binding likely do not

maintain planar geometries and lead to lower fluorescence intensities and shorter fluorescence

lifetimes as a result of twisted geometries and solvent exposure for the fully complementary

dsDNA sequences. If dye molecules were bound and aligned along a groove site in the DNA,

then the pitch associated with the groove site would give a dye angle orientation of

approximately 40-50 °.295 This is also in agreement with the computational studies by Silva et

al.296 who report that as the dye rotates through interplanar angles higher than 40°, a decrease in

quantum yield is observed due to a decreased absorption intensity associated with the S0→S1

transition. A further hypothesis is that the external binding of positively charged dyes leads to a

140

decreased surface charge on the DNA and a resultant increase in flexibility. Both of these

arguments support a net decrease of fluorescence lifetime associated with an external binding

mode when compared to an intercalative binding mode. The dye:DNA ratios used by Larsson et

al. were different from those used here. They showed that dye:DNA base pair ratios up to 0.2

exhibited intercalative modes of binding for dsDNA that was 164 kbp and that ratios greater than

0.2 showed contribution from external binding modes. In my work, 19 base oligonucleotides

were used. Since it is known that DNA undergoes breathing effects at the ends of duplexes, the

ratio of 1 dye to 19 bases (1:1 dye:DNA duplex) of ~0.05 could actually be much higher or have

a range of values depending on breathing effects. In addition, the nature of the linker attached to

a dye moiety has effects on the binding of the dye, and therefore could also lend preference to

external modes of binding even at low dye:DNA base ratios.

Fluorescence lifetime values were recorded for SMN1 duplexes containing three centrally

located base pair mismatches. The mismatched duplex in the absence of formamide showed a

mono-exponential lifetime. If the dye preferentially binds via intercalation with a lesser

contribution attributed to an external binding conformation as shown by the fully matched

duplexes, the destabilization of three mismatches centrally located in a duplex containing only 19

bases will disrupt the sites associated with external binding modes. The increase in fluorescence

lifetime associated with the intercalative component could arise from the fact that the quantum

yield enhancement of thiazole orange is higher for those sequences containing high purine

content.286 Since the mismatches chosen were three centrally located purine:purine mismatches

(A:A mismatches) which were located adjacent to three guanosine residues on the target strand,

this could lead to an overall fluorescence increase, despite the presence of mismatches, and is

also consistent with the higher fluorescence intensity observed in the steady state for the three

base pair mismatched duplex. This sequence-dependent increase in fluorescence lifetime could

potentially be confirmed through measuring the fluorescence lifetime of a polyA target (future

work). Upon introduction of formamide, the disruption starts to affect the intercalative modes of

binding and as the duplex is denatured, a bi-exponential lifetime signature emerges. At 18.5%

formamide, enough destabilization exists to provide multiple dye states, and a shorter lifetime

component with larger contribution to the overall lifetime is present. This is reflected in a value

for α1/α2 of 0.4. As the percentage of formamide was increased to 37%, the ratio of α1/α2

reduced to 0.1 as binding modes shifted further towards a shorter lifetime component. The

141

shorter lifetime component also showed a decrease in lifetime (1.8 ± 0.5 ns vs. 0.7 ± 0.1 ns),

which reflects further destabilization of the duplex and reduction of the barriers that restrict

rotation about the monomethine bridge in the TO-bio2 dye structure. The bi-exponential decay

fit is in agreement with data obtained by Carlsson et al.297 for oxazole yellow probes. The bi-

exponential decay kinetics were attributed to internal rotation, which resulted in a distribution of

dye species with different conformations and therefore different lifetimes.297 If rotation of the

bridge between the heterocycles is on the same time scale as the fluorescence lifetimes, the

measured lifetime then becomes an average distributed over the different conformations

present.297 If there is significant destabilization to create a wide distribution of conformational

species due to creation of a heterogeneous environment where rotational movement is permitted,

this would lead to a distribution of lifetimes reflecting multiple modes of dye/DNA association.

The 3 bpm system creates a heterogeneous environment for the dye species, and this is further

reflected in the range of associated χ2 values. For the 3 bpm system, enough instability exists to

introduce interplanar twisting. This would increase as a function of the increase in formamide

concentration, and would lead to a decrease in lifetime for the short lifetime component and a

shift of the overall contribution to favour the short lifetime component.

Fluorescence lifetime values were also measured for TO-bio2 bound to ssDNA in both the

presence and absence of formamide. The fluorescence lifetime for free dye in solution and

samples containing TO-bio2 with ssDNA in the presence of formamide showed similar lifetimes,

indicating that formamide disrupts binding of TO-bio2 to ssDNA. The lifetime of the dye

associated with ssDNA in the absence of formamide shows a similar lifetime component to that

of the shorter lifetime component of TO-bio2 associated with dsDNA. This is consistent with the

short lifetime component being indicative of an external binding mode that would be present for

both ssDNA and dsDNA, while intercalative binding modes would not be created in ssDNA

environments. Formamide essentially decreases the background fluorescence associated with

TO-bio2 in the presence of ssDNA, which has significant implications in the use of this dye as a

signalling agent for nucleic acid hybridization events.

In this work, the fluorescence lifetime data complemented the steady state fluorescence intensity

results and therefore provided further understanding of binding processes. The large decrease in

fluorescence intensity associated with both ssDNA and dsDNA can be attributed to structural

changes due to shedding of externally bound dye molecules as a result of formamide

142

interactions, which can readily exchange cations from ion interaction sites associated with the

nucleic acid structures. For duplexes containing mismatches, the formamide acted to destabilize

and denature duplexes and concurrently affected dye interactions. Therefore it is of great

advantage to use formamide as this can provide for improved selectivity of hybridization as well

as improvements in background suppression associated with TO dye interactions. This is a

significant finding that provides for an ability to tune the selectivity and improve signal-to-noise

by the simple addition of one reagent.

Chemical denaturation exploits the fact that even a single mismatch between probe and target

DNA strands will significantly alter duplex stability. The rapid response of formamide-mediated

denaturation observed in this work has been observed by Russom et al.283 and Leidl et al.280

However, in both cases, the target oligonucleotides were labeled and therefore required constant

removal of fluorescently labeled target material to reduce the background fluorescence

associated with denatured probe/target duplexes. While removal of material using solution flow

generated from a microfluidics device is a valid method of background supression, the work

presented in this thesis demonstrates a method that can be used in bulk solution assay.

Furthermore, this also leads to a method where the percentage of formamide required is reduced

(50% formamide required for SNP discrimination by Russom and co-workers 285 as compared to

37% used for this work) due to the fact that strands do not need to be fully dissociated since

destabilization of the duplex will already affect the torsional movement of the dye. In addition,

this work shows that SNP selectivity can be achieved for a variety of target lengths (as shown in

Figure 41), including PCR products.

143

Figure 41. TO-bio2 in the presence of unlabelled, fully complementary dsDNA (FC), and unlabelled dsDNA containing one base pair mismatch (1 bpm) with two different sequences containing 19 bases and 34 bases, respectively. Sample concentrations were 1 μM dsDNA, 1 μM TO-bio2 in 0.1x PBS buffer (20 mM NaCl) containing 37 % formamide. dsDNA corresponds to ProbeA and corresponding target as indicated in Table 12. Steady state fluorescence signals were measured at room temperature after incubation with formamide. Measurements were performed in triplicate and error bars represent 1 σ.

SNP analysis was also done using a 19-base probe/141 base PCR amplicon target. In an

integrated nucleic acid hybridization analysis system, PCR amplification of a relevent segment

of genomic DNA would provide the sample required for analysis. Therefore, the ability to

provide rapid, isothermal SNP discrimination of PCR amplicons would be a meaningful

contribution. The method used here to hybridize short oligonucleotide probes in the presence of

symmetric PCR products was based on methods developed by Wang et al.289 and Hill et al.298 ,

where short oligonucleotide ‘blocking strands’ were used to prevent strand-reannealing in

symmetric PCR products. The method was capable of detecting SNPs in full length PCR

amplicons without the requirement of a fluorescent label on either the probe or target nucleic

acid. Higher formamide percentages were required for PCR products due to a higher GC content

associated with the probe/target recognition sequences as well as to reduce background from an

excess addition of dye that would bind to multiple sites along the backbone of a longer PCR

target. The use of a symmetric PCR product mix, which may have some strand reannealing,

144

despite the excess of short ‘blocking’ strands present in the sample, can be thought of as a

somewhat heterogeneous sample. It was found that the signal-to-noise could be improved upon

increasing the ratio of probe strands from 4:1 to 8:1, as shown in Figure 42. Control experiments

showed that the blocking probe strands were successful and critical in reducing strand re-

annealing of PCR strands (see Figure 42 (iii)). The fluorescence signal intensity was highest for

samples containing PCR products only (no blocking strands) as strand re-annealing of a 141 bp

PCR target would provide multiple binding sites for dye intercalation. A trend was observed for

the data set containing blocking strands in the PCR amplicon mixture. The trend of fluorescence

signal intensity showed partially-complementary blocking strands (11 bpm in a 19-mer strand) >

fully complementary blocking strands > 1bpm blocking strands. Partially-complementary

blocking strands would not compete in the process of re-annealing as well as would fully

complementary or 1bpm samples, and were therefore expected to show a higher fluorescence

signal owing to some PCR amplicons that underwent strand re-annealing. However, the presence

of some complementary bases provided some blocking capability against the re-annealing of

PCR amplicons. The fully complementary samples showed blocking of strand re-annealing, and

provided a fluorescence signal indicative of hybridization of complementary probe/target

material. The 1bpm blocking strands provided the best blocking capability. This is attributed to a

combination of blocking re-annealing, and a concurrent decrease in stability imparted by a single

base mismatch so as to give TO-bio2 enough torsional movement to provide an overall decrease

in quantum yield when compared to the fully complementary sample.

145

(i) (ii)

(iii)

Figure 42. (i) (a) TO-bio2 in the presence of non-labelled, fully complementary dsDNA (FC), and (b) non-labelled dsDNA containing one base pair mismatch (1bpm). The target sequence was a 141 bp β-actin PCR amplicon. Sample concentrations were 0.48 μM PCR amplicon, 1.92 μM probe, and 4.8 μM TO-bio2 in 0.1 M PBS buffer (20 mM NaCl) containing 46 % formamide, to generate a 4:1 probe:PCR target ratio. (ii) Sample concentrations were 0.24 μM PCR amplicon, 1.92 μM probe, and 4.8 μM TO-bio2 in 0.1 M PBS buffer (20 mM NaCl) containing 46 % formamide, to generate 8:1 probe:PCR target ratio. Steady state fluorescence signals were measured at room temperature after incubation with formamide. Error bars represent 1 σ. (iii) Control experiment showing the ability of probe blocking strands in an 8:1 (probe:PCR amplicon) ratio to suppress strand re-annealing of PCR products and provide SNP discrimination. (a) absence of blocking strands (b) partially-complementary blocking strand (11bpm) (c) FC blocking strand (d) 1bpm blocking strand. Measurements were performed in triplicate and error bars represent 1σ.

146

5.4 Conclusions Preliminary work is reported using formamide and a biotin-tethered thiazole orange derivative to

provide rapid, isothermal SNP discrimination in solution for nonlabelled probe/target

recognition. The denaturing properties of formamide have been used to achieve both SNP

discrimination and a method for improving signal-to-noise through counter ion exchange of

cationic dye molecules. Although thiazole orange is able to discriminate between ssDNA and

dsDNA in the absence of formamide, the addition of formamide provides a larger change in

signal magnitude and reduction of fluorescence background. SNP discrimination was achieved

using two different target oligonucleotide lengths and a 141 base pair PCR amplicon.

147

6 Conclusions and Future Work

6.1 Future Outlook The results and discussion chapters of this thesis have outlined and investigated several key

concepts towards development of nucleic acid detection strategies based on optical spectroscopy

methods. The foundational research presented herein serves as a platform from which to launch

future work. The molecular switch strategy presented in this thesis can be thought of as proof-

of-concept of a nanoscale engineered construct that could be used in a variety of platforms in

both solution and at an interface. With this in mind, there are several ideas and applications that

could be explored and are outlined briefly in the following sections.

6.1.1 Modification of Thiazole Orange

The synthetic method presented in this thesis represents a facile strategy by which custom-built

tethers can be introduced between biotin and thiazole orange. Analysis of several tether lengths

should be conducted to optimize the tether length which allows rapid and selective detection of

hybridization. The synthesis of thiazole orange derivatives containing an extended polymethine

bridge between heterocycles of the dye structure should also be investigated. By extending this

bridge, the wavelengths of absorption and emission can be shifted, and this could provide

potential multiplexing capabilities. Further increase in signal-to-noise could also be achieved

using a pair of covalently linked thiazole orange dyes which can exploit exciton coupling effects

through H aggregate formation to quench dye species prior to a hybridization event. Ikeda et

al.95 have developed this strategy by synthesizing a doubly labelled nucleoside containing two

thiazole orange dye molecules. The synthetic strategy could be extended to this work and be

used to create biotinylated two-dye constructs. A mixture of switch constructs could be

effectively immobilized at an interface by immobilizing various dye/ssDNA pairs to central

Neutravidin hubs. This could further be extended to immobilizing this ‘sensing cargo’ to the

surface of a nanoparticle that could provide the necessary excitation source through FRET for

biomedical analyses or other assays.

148

6.1.2 Investigation of Synthetic Oligonucleotide Analogues and Intracellular Diagnostics

Synthetic analogues of DNA and RNA have emerged as potential biorecognition elements for

monitoring nucleic acid hybridization. While natural nucleic acids (DNA, RNA) are typically

used as recognition elements, synthetic nucleobase analogues, including peptide nucleic acids

(PNA) and locked nucleic acids (LNA) have gained recent popularity.98,99,129 Thiazole orange has

been reported to undergo 1000x signal increase upon binding to DNA, and 3000x upon binding

to RNA.80 It would therefore be worthwhile to investigate the fluorescence enhancement upon

binding to LNA structures since these bridged RNA analogues are known to resist enzymatic

degradation; thus, the potential for higher signal enhancement upon binding as well as a method

to move towards complex matrices including the cellular environment may be possible.133

Thiazole orange has been reported to have affinity for tumor cells and has also been reported to

exhibit cell membrane permeability.81,104 Therefore, the molecular switching strategy presented

here has the potential to be used in cancer diagnostics and intracellular work. The construction of

a molecular switch in solution followed by capture at an interface functionalized using

iminobiotin allows the denaturation of the dsDNA probe that is required to template the

dye/probe oligonucleotide and subsequent release of the molecular switch back into solution.

Depending on methodology of cellular work desired, it is convenient that the switch construct

could be used both in solution and at an interface. Ratiometric analyses could also be

investigated using the FRET pairs that were the subject of work in Chapters 2 and 3. An

investigation into PNA analogues may also be useful in reducing charge interactions associated

with the polyanionic backbone of DNA, and the increased affinity of PNA:DNA duplexes for

stability of hybrids.

6.1.3 SNP Selectivity Capability at an Interface

Perhaps the most significant focus of future work would be the application of the isothermal SNP

and signal enhancement capabilities shown in Chapter 5 of this thesis to SNP detection at an

interface. Preliminary selectivity work using the molecular switches immobilized at an interface

showed poor selectivity (data not shown). However, the use of formamide to impart selectivity

and increase signal magnitude for DNA hybrids over ssDNA represents a significant

methodology that can be used across a variety of applications. Arguably, label free methods are

limited due to the existence of a high background that would accompany immobilization of

149

fluorescent dyes at an interface as well as the relatively small number of fluorescent dyes that are

able to distinguish between ssDNA and dsDNA. If use of formamide to decrease external

binding modes decreases this background fluorescence while providing an isothermal method for

SNP analysis, then it would be an invaluable tool for both the advancement of label-free analysis

as well as device design that does not require external heating. One could envision this to have

important applications for sensing elements in microfluidics devices or chip-based detection

strategies.

6.2 Summary and Conclusions Successful design of fluorescent probes to provide changes in fluorescence properties

(fluorescence quantum yield, lifetime, anisotropy, microscopy, single molecule detection,

resonance energy transfer) as a result of target analyte binding is the basis of fluorescence based

optical detection of nucleic acids. Methods that provide overall simplicity while maintaining

functionality is primary to the design of nucleic acid detection platforms. The work presented in

this dissertation represents original research contributions to analytical chemistry, with a

particular emphasis on nucleic acid diagnostics using fluorescence spectroscopy. Fluorescence

spectroscopy in both the steady state and time-resolved regimes has been used to investigate

several aspects towards practical development of nucleic acid diagnostics. The main conclusions

derived from this work are summarized in the sections below.

6.2.1 Analysis of FRET Pairs for Monitoring DNA Hybridization

An extensive survey of the Förster distances of several donor–acceptor permutations involving

Cy3, Cy5, TAMRA, IabFQ, and IabRQ have been experimentally determined for both ssDNA

and dsDNA conjugates. The Cy3 and TAMRA donors showed variation in Förster distance due

to donor quantum yield changes between ssDNA and dsDNA. Selectivity studies focused

specifically on the TAMRA/IabRQ FRET pair, both in solution and at an interface.

Discrimination between fully matched, one, two and three base mismatches at 60°C was

achieved in solution, and interfacial experiments involving the TAMRA/IabRQ FRET pair

suggested that there are several modes of quenching when TAMRA is placed at an interface

which adds a level of complexity to transduction strategies in biosensor applications since

quenching of TAMRA emission could not be solely attributed to FRET. Surfaces of immobilized

TAMRA labelled probe could be regenerated upon sonication of substrates in sterile water at

150

room temperature. A significant number of biosensor designs do not consider the sensitivity of a

fluorophore to its immediate environment, or its physical interactions in that environment, and

the work in this thesis highlights the significance of fluorophore choice in the design of a sensor

platform. FRET analyses by incorporating an intercalating FRET donor were subsequently

investigated, and represent one of the first accounts of FRET using thiazole orange intercalating

dye as a fluorescence donor. TO was capable of acting as a FRET donor for Cy3, TAMRA,

IabFQ, Cy5 and IabRQ acceptor in dsDNA duplexes, and Förster distances for the

donor/acceptor pairs were calculated. Through the use of time-resolved fluorescence and FRET

data, it was shown that thiazole orange intercalates at the center of 19-mer dsDNA duplexes.

FRET efficiencies were also found to vary between calculations derived from steady state vs.

time-resolved data. From this data, it was discovered that TAMRA-labeled oligonucleotides

formed a ground-state non-fluorescent complex with TO and that this quenching phenomenon

was dependent on the hydrophobicity changes imparted by the nature of the side chain present on

the TO dye. The results provided further insight about the importance of fluorophore/FRET pair

choice since physical associations through dimerization or aggregation are known to affect

photophysical properties and would subsequently lead to discrepancies in assay results.

Experimental results aimed at improving signal to noise showed that an approximately four-fold

increase in signal-to-noise ratio over a single fluorophore TO-based assay can be obtained by

using a quencher such as IabFQ to suppress background fluorescence from non-intercalated dye,

and a proposed ratiometric scheme using TAMRA as a fluorescent acceptor dye would be

expected to yield a greater than five-fold increase in signal-to-noise ratio.

6.2.2 Development of a Molecular Switch Construct for Monitoring Nucleic Acid Hybridizaton

Molecular switch technologies have played, and will continue to play, a significant role in the

development of bioanalytical technologies, and represent an example of the impact of nano-scale

engineering on bioanalytical chemistry. Strategies that can exhibit ‘‘on/off’’ optical signaling

changes as a function of target binding offer the advantages of elimination of target labeling and

potential suitability for intracellular monitoring of DNA or RNA. In this thesis, the use of

binding-induced switching with a single tethered intercalant label and an oligonucleotide probe

provides an elegant approach to assemble a nanoscale construct that can be used in both solution

and at an interface to monitor biomolecular interactions. The experimental work presented

151

herein describes the construction of a novel molecular switch technology based on biomolecular

self-assembly to create a functional sensing platform for oligonucleotides. The focus has been on

examination of the functionality of the switch in a solid-phase assay environment, with the

intention being that optimizing the selectivity, sensitivity and speed are aspects that will follow

in the future. A novel thiazole orange derivative containing a biotinylated tether for intercalating

dye has been synthesized, and the success of the work presented here suggests a design platform

that can be used for the further development of simple nanoscale constructs that can be used for

bioanalytical applications. Step-wise analysis of structural components of the overall sensing

platform were investigated. Analysis of iminobiotinylated surfaces prepared in the same manner

as biotinylated surfaces showed that the surface chemistry used was robust and functional.

Confocal fluorescence microscopy data showed that pH dependent binding and release of

fluorescent Neutravidin from fused silica optical fibers was achieved. Steady-state fluorescence

data derived from monitoring fluorescence emission changes from tryptophan residues

associated with the peptide structure of Neutravidin showed that the biotinylated dye was

captured by biotin-binding sites. Regeneration of the single-stranded probe required for target

binding was possible following site-directed templating of probe and dye to adjacent binding

sites using dsDNA. Solution-phase fluorescence lifetime data of molecular switch constructs

showed a bi-exponential fluorescence decay profile, suggesting that intercalation as well as a

significant secondary binding mode for the immobilized TO was present. It was found that the

secondary binding mode could be decreased by adjusting the solution conditions to a pH below

the pI of Neutravidin, and by increasing the ionic strength of the buffer. The result was

enhancement of the role of the intercalative binding modes to the total fluorescence intensity.

Preliminary work demonstrated that it was possible to achieve up to a five-fold increase in

fluorescence intensity on hybridization to the target at an interface. This work represents a

significant advance in terms of providing a strategy that does not require external labelling of

oligonucleotides, and given that FRET pairs in solution gave similar signal magnitudes, this

method moves to a much simpler overall strategy. The use of binding-induced switching with a

single tethered intercalant label and an oligonucleotide probe provides an elegant approach to a

nanoscale engineered construct that can be used in both solution and at an interface to monitor

biomolecular interactions. The concept is to organize molecular components at an interface

through site-directed templating controlled by nucleic acids to provide molecular level

proximity. It is evident that there are several advantages to this construct, including: easily

152

assembled modular components; a novel biotinylated intercalating dye that provides stable

binding of the signalling moiety with high binding affinity; the ability to monitor binding events

both in solution and at an interface; analysis through two channels of fluorescence (steady-state

and time-resolved); and label-free target analysis (in this case “label-free” indicates that neither

the probe or target oligonucleotide requires conjugation with a fluorescent probe).

6.2.3 Isothermal Selectivity and Signal Enhancement Studies using Thiazole Orange and Formamide

One of the initial limitations associated with the molecular switch technology was the issue of

selectivity. Selectivity studies showed that although a signal difference between complementary

and non-complementary nucleic acid targets could be detected, the change was still statistically

small. To this end, the use of formamide as a means of achieving selectivity at room temperature

was investigated in solution. Formamide was used as a denaturant to influence duplex stability.

Isothermal SNP selectivity was achieved for the SMN1 probe/target hybrid for two different

target lengths. Formamide was also shown to enhance the ability of thiazole orange to

distinguish between single-stranded and double-stranded DNA due to changes in binding modes

associated with the dye/DNA structure as shown from time-resolved fluorescence decay data.

Fluorescence lifetime measurements have shown that the presence of formamide alters the

relative ratios of the contributions of bi-exponential lifetimes associated with TO-bio2 binding to

dsDNA and showed that formamide was capable of altering the external binding modes

associated with electrostatic interactions between the cationic dye species and polyanionic

backbone of ssDNA and dsDNA. This increased the magnitude of signal change for thiazole

orange between single-stranded DNA and double-stranded DNA and thus represents a

favourable increase in signal over backround. The system was also capable of SNP

discrimination for a 141 base pair PCR amplicon generated using symmetric PCR. Analysis

could be performed in as little as ten minutes for oligonucleotide targets.

153

Appendix I Synthesis of Thiazole Orange Derivatives and Starting Materials

Confirmation of dye starting materials and final products was provided by 1H-NMR (200 or 500

MHz) and ESI+ Mass Spectrometry. In addition to the distinct colour change that occurred when

thiazole orange was formed (deep red solution), the 1H-NMR spectra showed the disappearance

of the singlet associated with the methyl group on the quinoline ring at 3ppm and the appearance

of a singlet at ~7ppm indicative of the formation of the methine bridge linking the quinoline and

benzothiazole moieties. Tether attachment at the quinolinium nitrogen was also confirmed by a

triplet at ~5ppm.

Synthesis of TO derivatives

1

Preparation of compound 1.

In an oven-dried 250mL round bottom flask containing a stir bar, 2-(methylthio)benzothizole

(5.0 g, 27.6 mmol, 1.0 eq) was dissolved in 40 mL anhydrous EtOH under an argon atmosphere.

Iodomethane (11.7 g, 82.7 mmol, 3.0 eq) was added drop wise via syringe to this solution. The

round bottom was fitted with a condenser and the pale yellow solution was refluxed for 4 hours.

The reaction mixture yielded a pale yellow precipitate. The reaction was cooled to room

temperature, and the crude product was precipitated using diethyl ether. The reaction mixture

was filtered using a scintered glass frit, and washed with ether (3x15 mL). The crude product

was then re-dissolved in water (125 mL) and was extracted wtih dichloromethane (4x75 mL).

The product was then concentrated in vacuo to yield pale yellow needles (2.13 g, 24% yield); 1H-

NMR (200 MHz, d6-DMSO) δ 3.13 (s, 3H, SCH3), 4.11 (s, 3H, NCH3), 7.72-7.76 (m, 1H, Ar),

154

7.80-7.85 (m, 1H, Ar), 8.20 (d, J = 7.70 Hz, 1H, Ar), 8.52 (d, J = 8.10 Hz, 1H, Ar); Exact mass

calculated for [C9H10NS2+] requires m/z 196.28. Found 195.73 (ESI+).

2

Preparation of compound 2

In an oven-dried 250 mL round bottom flask containing a stir-bar, 1,3-diiodopropane (31.0 g,

104.8 mmol, 3.0 eq) was dissolved in 100mL dry toluene under an argon atmosphere. To this

solution, 4-methylquinoline was added (5g, 34.9 mmol, 1.0 eq) dropwise. The solution was then

refluxed for 24 hrs. The reaction was then allowed to cool to room temperature and the resulting

yellow precipitate was collected via vacuum filtration, washed with ether and subsequently

recrystallized from acetone (quantitative yield); 1H-NMR (500MHz, d6-DMSO) δ 2.50 (m, 2H,

CH2), 3.09 (s, 3H, CH3), 3.39 (t, 2H, CH2I), 5.07 (t, 2H, NCH2), 8.01-8.13 (m, 2H, Ar), 8.20-

8.30 (m, 1H, Ar), 8.58 (d, 1H, Ar), 8.72 (d, 1H, Ar), 9.44 (d, 1H, Ar).

155

3

Preparation of compound 3

In an oven-dried 100mL round bottom flask containing a stir-bar, 11-bromoundecanoic acid (4.1

g, 15 mmol, 1.1 eq) was added. 4-methylquinoline (1.85 mL, 14.0 mmol, 1.0 eq) was added.

The reaction was kept under argon and heated to 110°C for 3 hours. The resulting dark purple

residue was dissolved in MeOH (20 mL) and precipitated with ether (80 mL). Pale pink needles

were collected via vacuum filtration (1.35 g, 24% yield); 1H-NMR (200MHz, CDCl3) δ 1.19-

1.34 (m, 12H, CH2), 1.44 (m, 2H, CH2), 1.91-1.92 (m, 2H, CH2), 2.13 (t, 2H, CH2), 2.98 (s, 3H,

CH3), 4.99 (t, 2H, NCH2), 8.02-8.10 (m, 2H, Ar), 8.20-8.30 (m, 1H, Ar), 8.51-8.56 (m, 2H, Ar) ,

9.43 (d, 1H, Ar); Exact mass calculated for [C21H30NO2+] requires m/z 328.23. Found 328.21

(ESI+).

4

Preparation of compound 4

Sodium iodide (3.2 g, 16 mmol, 1.6 eq) was dissolved in 50mL dry acetone under an argon

atmosphere in an oven-dried 100 mL round bottom flask. To this solution, 2-[2-

(chloroethoxy)ethoxy]ethanol (1.68 g, 10 mmol, 1eq) was added drop wise via syringe. A

156

transparent yellow solution formed, and the reaction mixture was refluxed for six hours. The

solution was pale yellow with a white precipitate. The acetone was then evaporated in vacuo and

30mL of diethyl ether was added. The residue was then re-dissolved in acetone and precipitated

with ether to yield a white precipitate composed of NaCl and NaI. The mixture was vacuum

filtered and the filtrate was collected and concentrated in vacuo to yield a bright yellow oil (1.5

g, 58% yield); 1H-NMR (200MHz, CDCl3) δ 3.33 (t, 3J = 6.9 Hz, 3H, CH2I), 3.69 (m, 2H,

CH2OH), 3.73-3.84 (m, 3H, CH2); Exact mass calculated for [C6H23IO3-Na+] requires m/z

282.05. Found 282.88 (ESI+).

5

Preparation of compound 5

In an oven dried 50mL round bottom flask, 2-[2-(iodoethoxy)ethoxy]ethanol (1.5g,

5.8mmol, 1.7eq) was dissolved in 10mL dry toluene under an argon atmosphere. To this

solution, lepidine (0.5g, 3.5mmol, 1eq) was added drop wise via syringe. The reaction was

heated to 110°C and stirred for six hours and then allowed to cool to room temperature. The

residue was dissolved in MeOH and extracted into ether several times. The residue was

concentrated in vacuo and then dissolved in 30mL H2O and extracted with diethyl ether

(3x25mL). The aqueous layer was concentrated in vacuo to yield an orange oil (0.57 g, 40%

157

yield); 1H-NMR (200MHz, d6-DMSO) δ 3.07 (s, 3H, CH3), 3.26-3.57 (m, 8H, CH2O), 4.03 (t, 3J

= 4.1 Hz, 2H, CH2OH), 5.32 (t, 3J = 3.9 Hz, 2H, NCH2), 8.09-8.17 (m, 2H, Ar), 8.20-8.30 (m,

1H, Ar), 8.58 (d, J= 8.2 Hz, 1H, Ar), 8.72 (d, J= 9.0 Hz, 1H, Ar), 9.40 (d, J = 5.9 Hz, 1H, Ar);

Exact mass calculated for [C15H23NO3+] requires m/z 298.34. Found 298.94 (ESI+).

6

Preparation of compound 6 (TO-peg)

In an oven dried 50mL round bottom flask, lepidine-PEG (0.40 g, 0.90 mmol, 1 eq) and

S, 3-Dimethylbenzothiazole-2-thiol (0.32 g, 0.90 mmol, 1 eq) were suspended in 16mL

anhydrous EtOH under an argon atmosphere. Triethylamine was added (0.10 g, 0.97 mmol, 1eq)

drop wise to this suspension by syringe. The yellow reaction mixture immediately turned deep

red in colour upon addition of triethylamine. The reaction was stirred at 45°C for two hours and

then allowed to stir at room temperature for one hour. The reaction mixture was then

concentrated in vacuo and was washed with ether (4x10 mL). The dark red residue was then re-

dissolved in 10mL EtOH by warming the mixture of residue and EtOH. The mixture was

precipitated using diethyl ether and the precipitate was collected via vacuum filtration and

washed with diethyl ether (3x10 mL) to yield a red powder (0.25 g, 47% yield); 1H-NMR

(200MHz, d6-DMSO) δ 3.28-3.57 (m, 8H, CH2O), 3.89 (t, 3J = 3.8 Hz, 2H, CH2OH), 4.04 (s, 3H,

NCH3), 4.81 (t, 3J = 4.0 Hz, 2H, NCH2), 6.94 (s, 1H, CH), 7.37-7.47 (m, 2H, Ar), 7.59-7.67 (m,

1H, Ar), 7.73-7.83 (m, 2H, Ar), 7.95-7.99 (m, 1H, Ar), 8.04-8.08 (m, 1H, Ar), 8.19 (d, J= 8.3Hz,

1H, Ar), 8.55 (d, J=7.1Hz, 1H, Ar), 8.81 (d, J=8.82 Hz, 1H, Ar); Exact mass calculated for

[C24H30N2O3S+] requires m/z 423.56. Found 423.21 (ESI+).

158

7

Preparation of compound 7 (thiazole orange containing undecanoic acid side chain)

In an oven dried 50mL round bottom flask, the quinolinium undecanoic acid derivative (0.18 g,

0.39 mmol, 1eq) and benzothiazolium derivative (0.12 g, 0.39 mmol, 1 eq) were suspended in

10 mL anhydrous EtOH under an argon atmosphere. Triethylamine was added (0.08 g, 0.39

mmol, 1eq) drop wise to this suspension by syringe. The yellow reaction mixture immediately

turned deep red in colour upon addition of triethylamine. The reaction was stirred at 50 °C for

two hours and then allowed to stir at room temperature for one hour. The reaction mixture was

then concentrated in vacuo and was washed with ether (4x10 mL) and the precipitate was

collected via vacuum filtration. The product was re-suspended in a solution containing 7 mL

acetone and 10 mL ether for 45 minutes and the precipitate was collected via vacuum filtration

and washed with diethyl ether (3x10mL) to yield a red powder (0.92 g, 40%yield) Exact mass

calculated for [C29H35N2O2S+] requires m/z 475.24. Found 475.11(ESI+).

159

8

Preparation of compound 8 (thiazole orange containing iodopropyl side chain)

In an oven dried 100mL round bottom flask, the quinolinium iodopropyl derivative (0.68 g, 1.55

mmol, 1eq) and benzothiazolium derivative (0.5 g, 1.55 mmol, 1eq) were suspended in 40 mL

anhydrous EtOH under an argon atmosphere. Triethylamine was added (0.157 g, 1.55 mmol,

1eq) drop wise to this suspension by syringe. The yellow reaction mixture immediately turned

deep red in colour upon addition of triethylamine. The reaction was stirred at room temperature

overnight (~18 hrs). The mixture was concentrated in vacuo and precipitated with ether. The

product was recrystallized from acetone/ether to yield red crystals (0.91 g, 78% yield); 1H-NMR

(500MHz, d6-DMSO) δ 2.40 (m, 2H, CH2), 3.45 (t, 2H, CH2I), 4.04 (s, 3H, NCH3), 4.67 (t, 2H,

NCH2), 6.94 (s, 1H, CH), 7.39 (d, 1H, Ar), 7.41-7.50 (m, 1H, Ar), 7.60-7.65 (m, 1H, Ar), 7.70-

7.85 (m, 2H, Ar), 7.97-8.05 (m, 1H, Ar), 8.10 (d, 1H, Ar) 8.20 (d, 1H, Ar), 8.60 (d, 1H, Ar), 8.85

(d, 1H, Ar); Exact mass calculated for [C21H20IN2S+] requires m/z 459.38. Found 458.98 (ESI+).

160

9

Preparation of compound 9 (TO-bio1: thiazole orange containing biotinylated tether)

The synthesis of TO-bio1 began with an iodopropyl side chain that was attached to thiazole

orange based on the methods of Brooker et al.239, 240 and Carreon et al.97 The quinolinium

compound was synthesized by treating 4-methylquinoline (lepidine) (5 g, 34.9 mmol) with three

equivalents of diiodopropane (30.99 g , 104.8 mmol) in refluxing toluene followed by vacuum

filtration and subsequent recrystallization from acetone (see compound 2). The side-chain

functionalized quinolinium compound (1.36 g, 3.1 mmol) was then condensed with the

benzothiazole derivative (1.0 g, 3.1 mmol) (see compound 1) in the presence of triethylamine

(0.313 g, 3.1 mmol) in ethanol at 55 °C to give the dye species. The dye species was

recrystallized from MeOH/H2O. The biotinylated thiazole orange moiety was then synthesized

by reacting the previously synthesized thiazole orange dye (0.03 g, 0.051 mmol) in equimolar

quantity with EZ-Link® amine-PEG2-biotin (0.02 g, 0.051 mmol) in the presence of

triethylamine in dry DMF at 60 °C for 48 hr. DMF was evaporated under reduced pressure to

yield a red solid. The mass calculated for [C37H49N6O4S2+] was m/z 705.95, and the value found

was 705.11 amu (ESI+).

161

10

Preparation of compound 10 (TO-bio2: thiazole orange containing biotinylated tether)

The synthesis of TO-bio2 began with the attachment of an undecanoic acid side chain to thiazole

orange (see compounds 1, 3 and 7) based upon the methods of Svanvik et al.260 and Carreon et

al.97, and subsequent conversion to an NHS ester was based on the methods of Pei et al.261 The

carboxylic acid functionality at the chain terminus was then converted to an NHS ester by

reacting the dye species (30 mg, 0.05 mmol, 1 equiv.) with N-hydroxysuccinimide (NHS) (6.32

mg, 0.055 mmol, 1.1 equiv.) in the presence of DIC (6.93 mg, 0.055 mmol, 1.1 equiv.) in dry

dichloromethane. The reaction was allowed to stir under an inert atmosphere at room

temperature for 5 h. The biotinylated thiazole species was then synthesized by reacting the NHS

ester dye derivative (21 mg, 0.035 mmol, 1 equiv.) with EZ-Link® amine-PEG2-biotin (15 mg,

0.039 mmol, 1.1 equiv) in the presence of N,N-DIPEA (50 μL, 0.28 mmol, 8 equiv.) in

anhydrous DMF (3 mL). The reaction was allowed to stir under an inert atmosphere of argon for

20 h at room temperature. The DMF was evaporated under reduced pressure and the product was

rinsed with MilliQ water. The product was then dissolved in methanol and precipitated with

diethyl ether to yield a red solid (1.27 mg, 4%). The mass calculated for [C45H63N6O5S2+] was

m/z 832.15, and the mass found was 832.30 (ESI+).

162

Characterization of Switch Assembly Using TO-bio2 Biotin-Binding Assay

(a) (b)

Figure A.1.1. Fluorescence emission spectra of Neutravidin used to determine the number of occupied biotin binding sites. (a) Fluorescence emission of tryptophan residues residing in biotin binding pocket of Neutravidin (1 μM) after the addition of TO-bio2 (5 μM). Inset. Fluorescence difference spectrum of Neutravidin due to biotin, as calculated by subtracting biotin bound fluorescence emission spectrum from biotin-free fluorescence emission spectrum. (b) Calibration data obtained from difference spectra measured for 1-6 equivalents of TO-bio2 per Neutravidin.

163

Appendix II Fluorescence Lifetime Supporting Information (data associated with Table 13-p 137 )

Fluorescence Lifetime Curve Fitting and Residuals

Figures A.II.1. and A.II.2. show the fluorescence intensity decay profiles and fit residuals for the

fluorescence lifetime data presented in Table 13 (see p.137) to demonstrate the quality of fit.

The fluorescence lifetimes (τ), pre-exponential factors (α) and χ2 values are presented in Table

13.

Figure A.II.1. Fluorescence intensity decays (counts) of TO-bio2 intercalating dye in the presence of dsDNA. Curve fits are superimposed on raw data (black) and residuals are reported below the decay profiles. (a) Fluorescence lifetime analysis of TO-bio2 in the presence of dsDNA (1:1 dye duplex ratio) in the absence and presence of formamide. (b) Fluorescence lifetime analysis of TO-bio2 in the presence of dsDNA (2:1 dye duplex ratio) in the absence and presence of formamide.

164

Figure A.II.2. Fluorescence intensity decays (counts) of TO-bio2 intercalating dye in the presence of dsDNA. Curve fits are superimposed on raw data (black) and residuals are reported below the decay profiles. (a) Fluorescence lifetime analysis of TO-bio2 in the presence of dsDNA (1:1 dye duplex ratio) containing 3bpm in the absence and presence of formamide. (b) Fluorescence lifetime analysis of free TO-bio2 and TO-bio2 in the presence of ssDNA (1:1 dye duplex ratio) in the absence and presence of formamide

165

References (1) Functional Nucleic Acids for Analytical Applications; Li, Y.; Lu, Y., Eds.; Springer New

York, 2009.

(2) Valeur, B. Molecular Fluorescence: principles and applications; Wiley-VCH: Weinheim, 2002.

(3) Algar, W. R., University of Toronto, 2010.

(4) DNA Technology; Lakowicz, J. R., Ed.; Kluwer Academic/Plenum Publishers: New York, 2003; Vol. 7.

(5) Zhang, P.; Beck, T.; Tan, W. H. Angew. Chem. Int. Ed. 2001, 40, 402.

(6) Massey, M.; Algar, W. R.; Krull, U. J. Anal. Chim. Acta. 2006, 568, 181.

(7) Li, J. W. J.; Fang, X. H.; Schuster, S. M.; Tan, W. H. Angew. Chem. Int. Ed. 2000, 39, 1049.

(8) Tyagi, S.; Kramer, F. R. Nat. Biotechnol. 1996, 14, 303.

(9) Nakayama, H.; Arakaki, A.; K., M.; Takeyama, H.; Matsunaga, T. Biotechnol. Bioeng. 2003, 84, 96.

(10) Algar, W. R.; Massey, M.; Krull, U. J. J. Fluoresc. 2006, 16, 555.

(11) Dubertret, B.; Calame, M.; Libchaber, A. J. Nat. Biotechnol. 2001, 19, 365.

(12) Algar, W. R.; Krull, U. J. Anal. Chem. 2009, 81, 6562.

(13) Algar, W. R.; Krull, U. J. Anal. Chem. 2010, 82, 400.

(14) Algar, W. R.; Krull, U. J. Langmuir 2010, 26, 6041.

(15) Liu, Y.; Wang, Y.; Jianyu, J.; Wang, H.; Yang, R.; Tan, W. H. Chem. Commun. 2009, 6, 665.

(16) Marti, A. A.; Li, X.; Jockusch, S.; Li, Z.; Raveendra, B.; Kalachikov, S.; Russo, J. J.; Morozova, I.; Puthanveettil, S. V.; Ju, J.; Turro, N. J. Nucleic Acids Res. 2006, 34, 3161.

(17) Han, C. L.; Chen, T. T.; Zu, L. L. Chem. Phys. Lett. 2010, 500, 323.

(18) Ohta, N.; Takakazu, N.; Nagao, I.; Kinjo, M.; Aoki, Y.; Tanaka, M. In Proceedings of SPIE Achilefu, S., Raghavachari, R., Eds.; SPIE: San Jose, CA USA, 2009; Vol. 7190.

(19) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; 3rd ed.; Springer: New York, 2006.

166

(20) Atkins, P.; de Paula, J. Physical Chemistry; 7 ed.; Freeman: New York, 2002.

(21) Monti, S.; Cacelli, I.; Ferretti, A.; Prampolini, G.; Barone, V. J. Phys. Chem. B. 2010, 114, 8341.

(22) Sadava, D.; Heller, H. C.; Orians, G. H.; Purves, W. K.; Hillis, D. Life: The Science of Biology, 8th Ed.; Sinauer Associates and WH Freeman and Company: New York, 2008.

(23) Sachidanandam, S.; Weissman, D.; Schmidt, S. C.; Kakok, J. M.; Stein, L. D.; Marth, G.; Sherry, S.; Mullikin, J. C.; Mortimore, B. J.; Willey, D. L.; Hunt, S. E.; Col, C. G.; Coggill, P. C.; Ric, C. M.; Ning, Z.; Rogers, J.; Bentley, D. R.; Kwok, P.-Y.; Mardis, E. R.; Yeh, R. T.; Schultz, B.; Cook, L.; Davenport, R.; Dante, M.; Fulton, L.; Hillier, L.; Waterston, R. H.; McPherson, J. D.; Gilman, B.; Schaffner, S.; Van Etten, W. J.; Reich, D.; Higgins, J.; Daly, M. J.; Blumenstiel, B.; Baldwin, J.; Strange-Thomann, N.; Zody, M. C.; Linton, L.; Lander, E. S.; Altshuler, D. Nature 2001, 409, 928.

(24) Aoki, H.; Tao, H. Analyst 2005, 130, 1478.

(25) Ananthanawat, C.; Vilaivan, T.; Hoven, V. P.; Su, X. D. Biosens. Bioelectron. 2010, 25, 1064.

(26) Al Attar, H. A.; Monkman, A. P. Ad.v Funct. Mater. 2008, 18, 2498.

(27) Mohrle, B. P.; Kumpf, M.; Gauglitz, G. N. Analyst 2005, 130, 1634.

(28) Martinez, K.; Estevez, M. C.; Wu, Y. R.; Phillips, J. A.; Medley, C. D.; Tan, W. H. Anal. Chem. 2009, 81, 3448.

(29) Veedu, R. N.; Wengel, J. Chem. Biodivers. 2010, 7, 536.

(30) Koshkin, A. A.; Nielsen, P.; Meldgaard, M.; Rajwanshi, V. K.; Singh, S. K.; Wengel, J. J. Am. Chem. Soc. 1998, 120, 13252.

(31) Nokhrin, S.; Baru, M.; Lee, J. S. Nanotechnology 2007, 18.

(32) Spring, B. Q.; Clegg, R. M. J. Phys. Chem. B. 2007, 111, 10040.

(33) Dinsmore, M. J.; Lee, J. S. J. Inorg. Biochem. 2008, 102, 1599.

(34) Lee, J. S.; Latimer, L. J. P.; Reid, R. S. Biochem. Cell Biol. 1993, 71, 162.

(35) Rupcich, N.; Nutiu, R.; Shen, Y.; Li, Y.; Brennan, J. D. The Use of Functional Nucleic Acids in Solid-Phase Fluorometric Assays; Springer Science+Business Media: New York, 2009.

(36) Song, S. P.; Wang, L. H.; Li, J.; Zhao, J. L.; Fan, C. H. Trac-Trend. Ana.l Chem. 2008, 27, 108.

(37) Sefah, K.; Phillips, J. A.; Xiong, X. L.; Meng, L.; Van Simaeys, D.; Chen, H.; Martin, J.; Tan, W. H. Analyst 2009, 134, 1765.

167

(38) http://ghr.nlm.nih.gov/ghr/glossary/oligonucleotide (accessed May 4, 2011)

(39) Di Giusto, D. A.; King, G. C. Top. Curr. Chem. 2006, 261, 131.

(40) Garcia-Aljaro, C.; Munoz, F. X.; Baldrich, E. Analyst 2009, 134, 2338.

(41) Avidin-Biotin Interactions: Methods and Applications; Walker, J. M.; McMahon, R. J., Eds.; Humana Press, 2008.

(42) Marttila, A. T.; Airenne, K. J.; Laitinen, O. H.; Kulik, T.; Bayer, E. A.; Wilchek, M.; Kulomaa, M. S. Febs. Lett. 1998, 441, 313.

(43) Smith, C. L.; Milea, J. S.; Nyugen, G. H. Top. Curr. Chem. 2006, 261, 63.

(44) http://www.pdb.org/pdb/explore/explore.do?structureId=1ldq. (accessed June 26, 2011)

(45) http://www.jmol.org/. (accessed June 26, 2011)

(46) Alberts, B.; Bray, D.; Lewis, J.; Ratt, M.; Roberts, K.; Watson, J. D. Molecular Biology of the Cell; Garland Publishing: New York, 1989.

(47) www.piercenet.com. (accessed May 5, 2011)

(48) Lu, H.; Zhao, Y. J.; Ma, J. M.; Li, W. Y.; Lu, Z. H. Colloid Surface A. 2000, 175, 147.

(49) Chaiet, L.; Wolf, F. J. Arch. Biochem. Biophys. 1964, 106, 1.

(50) Medintz, I. L.; Anderson, G. P.; Lassman, M. E.; Goldman, E. R.; Bettencourt, L. A.; Mauro, J. M. Anal. Chem. 2004, 76, 5620.

(51) Liu, X. J.; Tan, W. H. Anal. Chem. 1999, 71, 5054.

(52) Long, F.; Wu, S.; He, M.; Tong, T.; Shi, H. Biosens. Bioelectron. 2011, 26, 2390.

(53) Massey, M.; Krull, U. J. Anal. Bioanal. Chem. 2010, 398, 1605.

(54) Jin, W.; Lin, X. C.; Lv, S. W.; Zhang, Y.; Jin, Q. H.; Mu, Y. Biosens. Bioelectron. 2009, 24, 1266.

(55) Bassil, N.; Maillart, E.; Canva, M.; Levy, Y.; Millot, M. C.; Pissard, S.; Narwa, W.; Goossens, M. Sensor. Actuat. B-Chem. 2003, 94, 313.

(56) Storri, S.; Santoni, T.; Mascini, M. Anal. Lett. 1998, 31, 1795.

(57) Zhou, X. C.; Huang, L. Q.; Li, S. F. Y. Biosens. Bioelectron. 2001, 16, 85.

(58) Singh, R.; Verma, R.; Kaushik, A.; Sumana, G.; Sood, S.; Gupta, R. K.; Malhotra, B. D. Biosens. Bioelectron. 2011, 26, 2967.

(59) Li, G. J.; Li, X. L.; Wan, J.; Zhang, S. S. Biosens. Bioelectron. 2009, 24, 3281.

168

(60) http://www.idtdna.com/catalog/Modifications/Modifications.aspx?catid=2. (accessed May 5, 2011)

(61) Wu, F. B.; He, Y. F.; Han, S. Q. Clin. Chim. Acta 2001, 308, 117.

(62) Biebricher, A.; Paul, A.; Tinnefeld, P.; Golzhauser, A.; Sauer, M. J. Biotechnol. 2004, 112, 97.

(63) Peluso, P.; Wilson, D. S.; Do, D.; Tran, H.; Benkatasubbaiah, M.; Quincy, D.; Heidecker, B.; Poindexter, K.; Tolani, N.; Phelan, M.; Witte, K.; Jung, L. S.; Wagner, P.; Nock, S. Anal. Biochem. 2003, 312, 113.

(64) Wayment, J. R.; Harris, J. M. Anal. Chem. 2009, 81, 336.

(65) Zhao, S.; Reichert, W. M. Langmuir 1992, 8, 2785.

(66) Kricka, L. J.; Fortina, P. Clin. Chem. 2009, 55, 670.

(67) Zeitz, L.; Lee, R. Science 1963, 142, 1670.

(68) Feulgen, R.; Rossenbeck, H. H-S Z Physiol. Chem. 1924, 135, 203.

(69) Kissane, J. M.; Robins, E. J. Biol. Chem. 1958, 233, 184.

(70) Middendorf LR, Patonay G, inventors; Li-Cor, Inc. (Lincoln, NE, assignee.) Sequencing near infrared and infrared fluorescence labeled DNA for detecting using laser diodes. US Patent 5,230,781. 1993 Jul 27.

(71) Berlier, J. E.; Rothe, A.; Buller, G.; Bradford, J.; Gray, D. R.; Filanoski, B. J.; Telford, W. G.; Yue, S.; Liu, J. X.; Cheung, C. Y.; Chang, W.; Hirsch, J. D.; Beechem, J. M.; Haugland, R. P.; Haugland, R. P. J Histochem. Cytochem. 2003, 51, 1699.

(72) Ihmels, H.; Otto, D. Top. Curr. Chem. 2005, 258, 161.

(73) Ren, J. S.; Jenkins, T. C.; Chaires, J. B. Biochemistry-Us 2000, 39, 8439.

(74) Sartorius, J.; Schneider, H. J. J. Chem. Soc. Perk. T 2. 1997, 2319.

(75) Herzyk, P.; Neidle, S.; Goodfellow, J. M. J. Biomol. Struct. Dyn. 1992, 10, 97.

(76) Friedman, R. A. G.; Manning, G. S. Biopolymers 1984, 23, 2671.

(77) Waring, M. J. J. Mol. Biol. 1965, 13, 269.

(78) Bloomfield, V. A.; Crothers, D. M. C.; Tinoco, I. Physical chemistry of nucleic acids; Harper and Row: New York, 1974.

(79) Kricka, L. J. Ann. Clin. Biochem. 2002, 39, 114.

(80) Lee, L. G.; Chen, C. H.; Chiu, L. A. Cytometry 1986, 7, 508.

169

(81) Rye, H. S.; Yue, S.; Wemmer, D. E.; Quesada, M. A.; Haugland, R. P.; Mathies, R. A.; Glazer, A. N. Nucleic Acids Res. 1992, 20, 2803.

(82) Bordelon, J. A.; Feierabend, K. J.; Siddiqui, S. A.; Wright, L. L.; Petty, J. T. J. Phys. Chem. B. 2002, 106, 4838.

(83) Staerk, D.; Hamed, A. A.; Pedersen, E. B.; Jacobsen, J. P. Bioconjugate Chem. 1997, 8, 869.

(84) Nygren, J.; Svanvik, N.; Kubista, M. Biopolymers 1998, 46, 39.

(85) Rye, H. S.; Quesada, M. A.; Peck, K.; Mathies, R. A.; Glazer, A. N. Nucleic Acids Res. 1991, 19, 327.

(86) Constantin, T. P.; Silva, G. L.; Robertson, K. L.; Hamilton, T. P.; Fague, K.; Waggoner, A. S.; Armitage, B. A. Org. Lett. 2008, 10, 1561.

(87) Bethge, L.; Singh, I.; Seitz, O. Org. Biomol. Chem. 2010, 8, 2439.

(88) Tamulaitis, G.; Zaremba, M.; Szczepanowski, R. H.; Bochtler, M.; Siksnys, V. Nucleic Acids Res. 2007, 35, 4792.

(89) Piatek, A. S.; Tyagi, S.; Pol, A. C.; Telenti, A.; Miller, L. P.; Kramer, F. R.; Alland, D. Nat. Biotechnol. 1998, 16, 359.

(90) Jarikote, D. V.; Krebs, N.; Tannert, S.; Roder, B.; Seitz, O. Chem-Eur J. 2007, 13, 300.

(91) Ghasemi, J.; Ahmadi, S.; Ahmad, A. I.; Ghobadi, S. Appl. Biochem. Biotech. 2008, 149, 9.

(92) Thompson, M. Biomacromolecules 2007, 8, 3628.

(93) Netzel, T. L.; Nafisi, K.; Zhao, M.; Lenhard, J. R.; Johnson, I. J Phys Chem-Us 1995, 99, 17936.

(94) Privat, E.; Melvin, T.; Merola, F.; Schweizer, G.; Prodhomme, S.; Asseline, U.; Vigny, P. Photochem. Photobiol. 2002, 75, 201.

(95) Ikeda, S.; Okamoto, A. Chem.-Asian J. 2008, 3, 958.

(96) Ikeda, S.; Kubota, T.; Yuki, M.; Yanagisawa, H.; Tsuruma, S.; Okamoto, A. Org. Biomol. Chem. 2010, 8, 546.

(97) Carreon, J. R.; Mahon, K. P.; Kelley, S. O. Org. Lett. 2004, 6, 517.

(98) Kohler, O.; Seitz, O. Chem. Commun. 2003, 2938.

(99) Kohler, O.; Venkatrao, D.; Jarikote, D. V.; Seitz, O. Chembiochem 2005, 6, 69.

(100) Privat, E.; Asseline, U. Bioconjugate Chem. 2001, 12, 757.

170

(101) Privat, E.; Melvin, T.; Asseline, U.; Vigny, P. Photochem. Photobiol. 2001, 74, 532.

(102) Wang, X. F.; Krull, U. J. Anal. Chim. Acta 2002, 470, 57.

(103) Wang, X. F.; Krull, U. J. Bioorg. Med. Chem. Lett. 2005, 15, 1725.

(104) Fei, X. N.; Gu, Y. C.; Ban, Y.; Liu, Z. J.; Zhang, B. L. Bioorgan. Med. Chem. 2009, 17, 585.

(105) Yang, P.; De Cian, A.; Teulade-Fichou, M. P.; Mergny, J. L.; Monchaud, D. Angew. Chem. Int. Edit. 2009, 48, 2188.

(106) Li, K.; Liu, B. Anal. Chem. 2009, 81, 4099.

(107) Garanger, E.; Hilderbrand, S. A.; Blois, J. T.; Sosnovik, D. E.; Weissleder, R.; Josephson, L. Chem. Commun. 2009, 4444.

(108) Tyagi, S.; Bratu, D. P.; Kramer, F. R. Nat. Biotechnol. 1998, 16, 49.

(109) Conlon, P.; Yang, C. J.; Wu, Y.; Chen, Y.; Martinez, K.; Kim, Y.; Stevens, N.; Marti, A. A.; Jockusch, S.; Turro, N. J.; Tan, W. H. J. Am. Chem. Soc. 2008, 138, 336.

(110) Wu, Y.; Yang, C. J.; Moroz, L. L.; Tan, W. Anal. Chem. 2008, 80, 3025.

(111) Kim, J. H.; Chaudhary, S.; Ozkan, M. Nanotechnology 2007, 18, 195105.

(112) Medintz, I. L.; Berti, L.; Pons, T.; Grimes, J. F.; English, D. S.; Alessandrini, A.; Facci, P.; Mattoussi, H. Nano. Lett. 2007, 7, 1741.

(113) Zhang, J.; Qi, H.; Li, Y.; Yang, J.; Gao, Q.; Zhang, C. Anal. Chem. 2008, 80, 2888.

(114) Du, H.; Strohsahl, C. M.; Camera, J.; Miller, B. L.; Krauss, T. D. J. Am. Chem. Soc. 2005, 127, 7932.

(115) Strohsahl, C. M.; Miller, B. L.; Krauss, T. D. Nat. Protoc. 2007, 2, 2105.

(116) Yang, C. J.; Wang, L.; Wu, Y.; Kim, Y.; Medley, C. D.; Lin, H.; Tan, W. Nucleic Acids Res. 2007, 35, 4030.

(117) Piestart, O.; Barsch, H.; Buschmann, V.; Heinlein, T.; Knemeyer, J. P.; Weston, K. D.; Sauer, M. Nano Lett. 2003, 3, 979.

(118) Hoefelschweiger, B. K.; Wolfbeis, O. S. J. Fluoresc. 2008, 18, 413.

(119) Wu, Y.; Yang, C. J.; Tan, W. H. Anal. Chem. 2008, 80, 3025.

(120) Yang, R.; Jin, J.; Chen, Y.; Shao, N.; Kang, H.; Ziao, Z.; Tang, Z.; Wu, Y.; Zhu, Z.; Tan, W. H. J. Am. Chem. Soc. 2008, 130, 8351.

(121) Cady, N. C.; Strickland, A. D.; Batt, C. A. Mol. Cell Probe. 2007, 21, 116

171

(122) Stoermer, R. L.; Keating, C. D. J. Am. Chem. Soc. 2006, 128, 13243.

(123) Du, H.; Disney, M. D.; Miller, B. L.; Krauss, T. D. J. Am. Chem. Soc. 2003, 125, 401.

(124) Knemeyer, J. P.; Marme, N.; Sauer, M. Anal. Chem. 2000, 72, 3717.

(125) Kuhn, H.; Demidov, V. V.; Coull, J. M.; Fiandaca, M. J.; Gildea, B. D.; Frank-Kamenetskii, M. D. J. Am. Chem. Soc. 2002, 124, 1097.

(126) Zuo, X.; Yang, X.; Wang, K.; Tan, W.; Li, H.; Zhou, L.; Wen, J.; Zhang, H. Anal. Chim. Acta 2006, 567, 173.

(127) Fang, X.; Liu, X.; Schuster, S.; Tan, W. J. Am. Chem. Soc. 1999, 121, 2921.

(128) Liu, X.; Farmerie, W.; Schuster, S.; Tan, W. Anal. Biochem. 2000, 283, 56.

(129) Martinez, K.; Estevez, M.-C.; Wu, Y.; Phillips, J. A.; Medley, C. D.; Tan, W. Anal. Chem. 2009, 81, 3448.

(130) Bonnet, G.; Tyagi, S.; Libchaber, A. J.; Kramer, F. R. P. Natl. Acad. Sci. USA. 1999, 96, 6171.

(131) http://www.idtdna.com/catalog/DualLabeledLNAProbes. (accessed June 24, 2011)

(132) http://www.exiqon.com/order-lna-oligos. (accessed June 24, 2011)

(133) Wang, L.; Yang, C. J.; Medley, C. D.; Benner, S. A.; Tan, W. J. Am. Chem. Soc. 2005, 127, 15664.

(134) Viasnoff, V.; Meller, A.; Isambert, H. Nano. Lett. 2006, 6, 101.

(135) Kolaric, B.; Sliwa, M.; Brucale, M.; Vallee, R. A. L.; Zuccheri, G.; Samori, B.; Hofkens, J.; De Schryver, F. C. Photochem. Photobiol. Sci. 2007, 6, 614.

(136) Liu, D.; Bruckbauer, A.; Abell, C.; Balasubramanian, S.; Kang, D. J.; Klenerman, D.; Zhou, D. J. Am. Chem. Soc. 2006, 128, 2067.

(137) Liu, H.; Zhou, Y.; Yang, Y.; Wang, W.; Qu, L.; Chen, C.; Liu, D.; Zhang, D.; Zhu, D. J. Phys. Chem. B 2008, 112, 6893.

(138) Yan, H. Science 2004, 306, 2048.

(139) Fortina, P.; Kricka, L. J.; Surrey, S.; Grodzinski, P. Trends. Biotechnol. 2005, 23, 168.

(140) Carbone, A.; Seeman, N. C. P. Natl. Acad. Sci. USA. 2002, 99, 12577.

(141) Dupraz, C. J. F.; Nickels, P.; Beierlein, U.; Huynh, W. U.; Simmel, F. C. Superlattice. Microst. 2003, 33, 369.

172

(142) Morgan, J. R.; Lyon, R. P.; Maeda, D. Y.; Zebala, J. A. Nucleic Acids Res. 2008, 36, 3522.

(143) Martinez, K.; Medley, C. D.; Yang, C. J.; Tan, W. Anal.Bioanal. Chem. 2008, 391, 983.

(144) Buck, A. H.; Campbell, C. J.; Dickinson, P.; Mountford, C. P.; Stoquert, H. C.; Terry, J. G.; Evans, S. A. G.; Keane, L. M.; Su, T. J.; Mount, A. R.; Walton, A. J.; Beattie, J. S.; Crain, J.; Ghazal, P. Anal. Chem. 2007, 79, 4724.

(145) Rant, U.; Arinaga, K.; Scherer, S.; Pringsheim, E.; Fujita, S.; Yokoyama, N.; Tornow, M.; Abstreiter, G. P. Natl. Acad. Sci. USA. 2007, 104, 17364.

(146) Dore, K.; Dubus, S.; Ho, H. A.; Levesque, I.; Brunette, M.; Corbeil, G.; Boissinot, M.; Boivin, G.; Bergeron, M. G.; Boudreau, D.; Leclerc, M. J. Am. Chem. Soc. 2004, 126, 4240.

(147) Raymond, F. R.; Ho, H. A.; Peytavi, R.; Bissonnette, L.; Boissinot, M.; Picard, F. J.; Leclerc, M.; Bergeron, M. G. BMC Biotechnol.2005, 5, 10.

(148) Dore, K.; Leclerc, M.; Boudreau, D. J. Fluoresc. 2006, 16, 259.

(149) Ho, H. A.; Boissinot, M.; Bergeron, M. G.; Corbeil, G.; Dore, K.; Boudreau, D.; Leclerc, M. Angew. Chem. Int. Ed. 2002, 41, 1548.

(150) Ho, H. A.; Dore, K.; Boissinot, M.; Bergeron, M. G.; Tanguay, R. M.; Boudreau, D.; Leclerc, M. J. Am. Chem. Soc. 2005, 127, 12673.

(151) Menacher, F.; Rubner, M.; Berndl, S.; Wagenknect, H. A. J. Org. Chem. 2008, 73, 4263.

(152) Hudson, R. H. E.; Ghorbani-Choghamarani, A. Org. Biomol. Chem. 2007, 5, 1845.

(153) Okamoto, A.; Tainaka, K.; Nishiza, K.; Saito, I. J. Am. Chem. Soc. 2005, 127, 13128.

(154) Okamoto, A.; Tainaka, K.; Ochi, Y.; Kanatani, K.; Saito, I. Mol. Biosyst. 2006, 2, 122.

(155) Okamoto, A.; Saito, Y.; Saito, I. Photochem. Photobiol. 2005, 6, 108.

(156) Leung, A.; Shankar, P. M.; Mutharasan, R. Sensor. Actuator. B 2007, 125, 688.

(157) Monk, D. J.; Walt, D. R. Anal. Bioanal. Chem. 2004, 379, 931.

(158) Piunno, P. A. E.; Krull, U. J.; Hudson, R. H. E.; Damha, M. J.; Cohen, H. Anal. Chem. 1995, 67, 2635.

(159) Massey, M.; Piunno, P. A. E.; Krull, U. J. Frontiers in Optical Sensing: Novel Principles and Techniques; Springer-Verlag: Berlin, 2005; Vol. 3.

(160) Wong, E. L. S.; Chow, E.; Gooding, J. J. Langmuir 2005, 21, 6957.

(161) Piunno, P. A. E.; Watterson, J.; Wust, C. C.; Krull, U. J. Anal. Chim. Acta 1999, 400, 73.

173

(162) Kleinjung, F.; Bier, F. F.; Warsinke, A.; Scheller, F. W. Anal. Chim. Acta 1997, 350, 51.

(163) Beaucage, S. L.; Caruthers, M. H. Tetrahedron Lett. 1981, 22, 1859.

(164) Wolf, S. F.; Haines, L.; Fisch, J.; Kremsky, J. N.; Dougherty, J. P.; Jacobs, K. Nucleic Acids Res. 1987, 15, 2911.

(165) Levicky, R.; Horgan, A. Trends Biotechnol. 2005, 23, 143.

(166) Breslauer, K. J. Methods in Molecular Biology: Protocols for Oligonucleotide Conjugates; Humana Press: Totowa, 1994.

(167) Nelson, J. W.; Martin, F. H.; Tinoco, I. Biopolymers 1981, 20, 2509.

(168) Puglisi, J. D.; Tinoco, I. J. Methods Enzymol. 1989, 180, 304.

(169) Schildkr.C; Lifson, S. Biopolymers 1965, 3, 195.

(170) Watterson, J. H.; Piunno, P. A. E.; Wust, C. C.; Krull, U. J. Langmuir 2000, 16, 4984.

(171) Mergny, J. L.; Maurizot, J. C. Chembiochem 2001, 2, 124.

(172) Mergny, J. L. Biochemistry-Us 1999, 38, 1573.

(173) Marras, S. A. E.; Kramer, F. R.; Tyagi, S. Nucleic Acids Res. 2002, 30.

(174) Tsourkas, A.; Behlke, M. A.; Rose, S. D.; Bao, G. Nucleic Acids Res. 2003, 31, 1319.

(175) Christensen, U. B.; Wamberg, M.; El-Essawy, F. A. G.; Ismail, A. E. H.; Nielsen, C. B.; Filichev, V. V.; Jessen, C. H.; Petersen, M.; Pedersen, E. B. Nucleos. Nucleot. Nucl. 2004, 23, 207.

(176) Moreira, B. G.; You, Y.; Behlke, M. A.; Owczarzy, R. Biochem. Biophys. Res. Commun. 2005, 327, 473.

(177) Maskos, U.; Southern, E. M. Nucleic Acids Res. 1993, 21, 4663.

(178) Maskos, U.; Southern, E. M. Nucleic Acids Res. 1992, 20, 1675.

(179) Southern, E. M.; Casegreen, S. C.; Elder, J. K.; Johnson, M.; Mir, K. U.; Wang, L.; Williams, J. C. Nucleic Acids Res. 1994, 22, 1368.

(180) Williams, J. C.; Casegreen, S. C.; Mir, K. U.; Southern, E. M. Nucleic Acids Res. 1994, 22, 1365.

(181) Tinoco, I.; Uhlenbec.Oc; Levine, M. D. Nature 1971, 230, 362.

(182) Aboulela, F.; Koh, D.; Tinoco, I.; Martin, F. H. Nucleic Acids Res. 1985, 13, 4811.

(183) Cao, Y. W. C.; Jin, R. C.; Mirkin, C. A. Science 2002, 297, 1536.

174

(184) Algar, W. R.; Krull, U. J. Langmuir 2009, 25, 633.

(185) Almadidy, A.; Watterson, J.; Piunno, P. A. E.; Foulds, I. V.; Horgen, P. A.; Krull, U. J. Can. J. Chem. 2003, 81, 339.

(186) Hartley, H. A.; Baeumner, A. J. Anal. Bioanal. Chem. 2003, 376, 319.

(187) Deisingh, A. K.; Thompson, M. Analyst 2001, 126, 2153.

(188) Zhong, Z. B.; Reynolds, R.; Kidd, J. R.; Kidd, K. K.; Jenison, R.; Marlar, R. A.; Ward, D. C. P. Natl. Acad. Sci. USA. 2003, 100, 11559.

(189) Watterson, J. H.; Raha, S.; Kotoris, C. C.; Wust, C. C.; Gharabaghi, F.; Jantzi, S. C.; Haynes, N. K.; Gendron, N. H.; Krull, U. J.; Mackenzie, A. E.; Piunno, P. A. E. Nucleic Acids Res. 2004, 32.

(190) Xie, H.; Zhang, C. Y.; Gao, Z. Q. Anal. Chem. 2004, 76, 1611.

(191) Fojta, M. Electroanal. 2002, 14, 1449.

(192) Mariotti, E.; Minunni, M.; Mascini, M. Anal. Chim. Acta 2002, 453, 165.

(193) Rich, R. L.; Myszka, D. G. J. Mol. Recognit. 2000, 13, 388.

(194) Lee, M.; Walt, D. R. Anal. Biochem. 2000, 282, 142.

(195) Kong, D. M.; Huang, Y. P.; Zhang, X. B.; Yang, W. H.; Shen, H. X.; Mi, H. F. Anal. Chim. Acta 2003, 491, 135.

(196) McKeen, C. M.; Brown, L. J.; Nicol, J. T. G.; Mellor, J. M.; Brown, T. Org. Biomol. Chem. 2003, 1, 2267.

(197) Kwok, P. Y. Hum. Mutat. 2002, 19, 315.

(198) Sturner, D. M. The Chemistry of Heteroaromatic Compounds; Wiley: New York, 1977; Vol. 30.

(199) Mujumdar, R. B.; Ernst, L. A.; Mujumdar, S. R.; Lewis, C. J.; Waggoner, A. S. Bioconjugate Chem .1993, 4, 105.

(200) Norman, D. G.; Grainger, R. J.; Uhrin, D.; Lilley, D. M. J. Biochemistry-Us 2000, 39, 6317.

(201) Cooper, M.; Ebner, A.; Briggs, M.; Burrows, M.; Gardner, N.; Richardson, R.; West, R. J. Fluoresc. 2004, 14, 145.

(202) Unruh, J. R.; Gokulrangan, G.; Wilson, G. S.; Johnson, C. K. Photoch. Photobio. B 2005, 81, 682.

175

(203) Torimura, M.; Kurata, S.; Yamada, K.; Yokomaku, T.; Kamagata, Y.; Kanagawa, T.; Kurane, R. Anal. Sci. 2001, 17, 155.

(204) Vamosi, G.; Gohlke, C.; Clegg, R. M. Biophys J. 1996, 71, 972.

(205) Mergny, J. L.; Lacroix, L.; Teulade-Fichou, M. P.; Houndson, C.; Guittat, L.; Hoarau, M.; Arimondo, B. P.; Vigneron, J. P.; Legn, J. M.; Riou, J. F.; Garestier, T.; Helene, C. P. Natl. Acad. Sci. USA. 2001, 98, 3062.

(206) Mergny, J. L.; Garestier, T.; Rougee, M.; Lebedev, A. V.; Chassignol, M.; Thuong, N. T.; Helene, C. Biochemistry-Us 1994, 33, 15321.

(207) Mergny, J. L.; Boutorine, S. A.; Garestier, T.; Belloc, F.; Rougee, M.; Bulychev, N. V.; Koshkin, A. A.; Bourson, J.; Lebedev, A. V.; Valeur, B.; Thuong, N. T.; Helene, C. Nucleic Acids Res. 1994, 22, 920.

(208) Ellouse, C.; Piot, F.; Takahashi, M. J. Biochem-Tokyo 1997, 121, 151.

(209) Yang, M.; Ghosh, S. S.; Millar, D. P. Biochemistry-Us 1994, 33, 15329.

(210) Yang, M.; Millar, D. P. Methods Enzymol. 1997, 278, 417.

(211) Watterson, J.; Piunno, P. A. E.; Krull, U. J. Anal. Chim. Acta 2002, 469, 115.

(212) Henke, L.; Nagy, N.; Krull, U. J. Biosens. Bioelectron. 2002, 17, 547.

(213) Wong, A. K. Y.; Krull, U. J. Anal. Bioanal. Chem. 2005, 383, 187.

(214) Cheung, H. C. Topics in Fluorescence Spectroscopy, Principles; Plenum Press: New York, 1991; Vol. 2.

(215) Sjoback, R.; Nygren, J.; Kubista, M. Spectrochim. Acta A 1995, 51, L7.

(216) Rasnik, I.; Mckinney, S. A.; Ha, T. Accounts Chem. Res. 2005, 38, 542.

(217) Sabanayagam, C. R.; Eid, J. S.; Meller, A. J. Chem. Phys. 2005, 122, 061103.

(218) Malicka, J.; Gryczynski, I.; Fang, J.; Kusba, J.; Lakowicz, J. R. Journal of Fluorescence 2002, 12, 439.

(219) http://www.glenres.com/ProductFiles/Technical/Extinctioris.html. (accessed September 6, 2005)

(220) Amersham Pharmacia Biotech, Life Science News 4 (2000) 1.

(221) Williamson, J. R.; Raghuraman, M. K.; Cech, T. R. Cell 1989, 59, 871.

(222) Ferguson, G. S.; Chaudury, M. K.; Sigal, G. B.; Whitesides, G. M. Science 1991, 253, 776.

176

(223) Perez-Luna, V. H.; O'Brien, M. J.; Opperman, K. A.; Hampton, P. D.; Lopez, G. P.; Klumb, L. A.; Stayton, P. S. J. Am. Chem. Soc. 1991, 121, 6469.

(224) http://researchchem.psu.edu/mallouk/articles/am_dna2.pdf. (accessed September 6, 2005)

(225) Jablonski, J. Acta. Phys. Pol. A. 1955, 295.

(226) Malicka, J.; Gryczynski, I.; Lakowicz, J. R. Anal. Chem. 2003, 75, 4408.

(227) http://www.biosearchtech.com/download/research/tigr2000.pdf. (accessed September 6, 2005)

(228) Haugland, R. Handbook of molecular probes and research products; Molecular Probes, Inc.: Eugene, 2002; Vol. 9.

(229) Prodhomme, S.; Demaret, J. P.; Vinogradov, S.; Asseline, U.; Morin-Allory, L.; Vigny, P. J. Photoch. Photobio. B. 1999, 53, 60.

(230) Nissum, M.; Jacobsen, J. P.; Faurskov, N. O.; Waage, J. P. Biospectroscopy 1997, 3, 207.

(231) Laib, S.; Seeger, S. J. Fluoresc. 2004, 14, 187.

(232) Bordelon, J. A.; Feierabend, K. J.; Siddiqui, S. A.; Wright, L. L.; Petty, J. T. J. Phys. Chem. B. 2006, 106, 4838.

(233) Benson, S. C.; Singh, P.; Glazer, A. N. Nucleic Acids Res. 1993, 21, 5727.

(234) Le Pecq, J. B.; Paoletti, C. J. Mol. Biol. 1967, 27, 87.

(235) Hyun, K.-M.; Choi, S.-D.; Lee, S.; Kim, S. K. Biochim. Biophys. Acta 1997, 1334, 312.

(236) Petty, J. T.; Bordelon, J. A.; Robertson, M. E. J. Phys. Chem. B. 2000, 104, 7221.

(237) Bunkenborg, J.; Gadjev, N. I.; Deligeorgiev, T.; Jacobsen, J. P. Bioconjugate Chem. 2000, 11, 861.

(238) Bunkenborg, J.; Stidsen, M. M.; Jacobsen, J. P. Bioconjugate Chem. 1999, 10, 824.

(239) Rye, H. S.; Glazer, A. N. Nucleic Acids Res. 1995, 23, 1215.

(240) Brooker, L. G. S.; Keyes, G. H.; Williams, W. W. J. Am. Chem. Soc. 1942, 64, 199.

(241) Brooker, L. G. S.; White, F. L.; Keyes, G. H.; Smyth, C. P.; Oesper, P. F. J. Am. Chem. Soc. 1941, 63, 3192.

(242) Koizumi, M. D.-B., C.; Sauvage, J. P. Eur. J. Org. Chem. 2004, 4, 770.

(243) Major, A.; Barzda, V.; Piunno, P. A. E.; Musikhin, S.; Krull, U. J. Opt. Express 2006, 14, 5285.

177

(244) Shins, J. M.; Agronskaia, A.; deGrooth, B. G.; Greve, J. Cytometry 1999, 37, 230.

(245) Mizukami, S.; Kikuchi, K.; Higuchi, T.; Urano, Y.; Mashima, T.; Tsuruo, T.; Nagano, T. Febs. Lett. 1999, 453, 356.

(246) Daugherty, D. L.; Gellman, S. H. J. Am. Chem. Soc. 1999, 121, 4325.

(247) West, W.; Pearce, S. J. Phys. Chem.-Us. 1965, 69, 1894.

(248) Geoghegan, K. F.; Rosner, P. J.; Hoth, L. R. Bioconjugate Chem. 2000, 11, 71.

(249) Valdes-Aguilera, O. N., D.C. Accounts Chem. Res. 1989, 22, 171.

(250) Kikuchi, K.; Takakusa, H.; Nagano, T. Trac-Trend. Anal. Chem. 2004, 23, 407.

(251) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Kluwer Academic/Plenum Publishers: New York, 1999.

(252) Larsson, A.; Carlsson, C.; Johnsson, M. Biopolymers 1995, 36, 153.

(253) Larsson, A.; Carlsson, C.; Jonsson, M.; Albinsson, B. J. Am. Chem. Soc. 1994, 116, 8459.

(254) Nygren, J.; Andrade, J. M.; Kubista, M. Anal. Chem. 1996, 68, 1706.

(255) Hiller, Y.; Bayer, E. A.; Wilchek, M.; Edward, A. B. Methods Enzymol. 1990, 184, 68.

(256) Hiller, Y.; Gershoni, J. M.; Bayer, E. A.; Wilchek, M. Biochem J. 1987, 248, 167.

(257) Sano, T.; Cantor, C. R. P. Natl. Acad. Sci. USA. 1995, 92, 3180.

(258) Smith, C. L.; Milea, J. S.; Nguyen, G. H. Top. Curr. Chem. 2005, 261, 63.

(259) Soper, S. A.; Owens, C.; Lassiter, S.; Xu, Y.; Waddell, E. In DNA Technology; Lakowicz, J. R., Ed.; Kluwer/Plenum: New York, 2003; Vol. 7.

(260) Svanvik, N.; Westman, G.; Wang, D. Y.; Kubista, M. Anal. Biotchem. 2000, 281, 26.

(261) Pei, R.; Rothman, J.; Xie, Y. L.; Stojanovic, M. N. Nucleic Acids Res. 2009, 37.

(262) Lartia, R.; Asseline, U. Chem. Eur. J. 2006, 12, 2270.

(263) Asseline, U.; Toulme, F.; Thuong, N. T.; Delarue, M.; Montenaygarestier, T.; Helene, C. Embo J. 1984, 3, 795.

(264) Asseline, U.; Bonfils, E.; Dupret, D.; Thuong, N. T. Bioconjugate Chem. 1996, 7, 369.

(265) Furstenberg, A.; Kel, O.; Gradinaru, J.; Ward, T. R.; Emery, D.; Bollot, G.; Mareda, J.; Vauthey, E. Chemphyschem 2009, 10, 1517.

(266) Golosov, A. A.; Karplus, M. J. Phys. Chem. B. 2007, 111, 1482.

178

(267) Abbyad, P.; Shi, X. H.; Childs, W.; McAnaney, T. B.; Cohen, B. E.; Box.er, S. G. J. Phys. Chem. B. 2007, 111, 8269.

(268) http://www.piercenet.com/browse.cfm?fldID=01030702. (accessed May 31, 2011)

(269) Kurzban, G. P.; Gitlin, G.; Bayer, E. A.; Wilchek, M.; Horowitz, P. M. J Protein Chem. 1990, 9, 673.

(270) Bottini, M.; D'Annibale, F.; Magrini, A.; Cerignoli, F.; Arimura, Y.; Dawson, M. I.; Bergamaschi, E.; Rosato, N.; Bergamaschi, A.; Mustelin, T. Int. J. Nanomed. 2007, 2, 227.

(271) Marttila, A. T.; Laitinen, O. H.; Airenne, K. J.; Kulik, T.; Bayer, E. A.; Wilchek, M.; Kulomaa, M. S. Febs. Lett. 2000, 467, 31.

(272) Schweitzer, C.; Scaiano, J. C. Phys. Chem. Chem. Phys. 2003, 5, 4911.

(273) Ramirez, I. B. R.; Ekblad, L.; Jergil, B. J. Chromatogr. B. 2000, 743, 389.

(274) Bier, F. F.; Kleinjung, F.; Scheller, F. W. Sensor. Actuat. B.-Chem. 1997, 38, 78.

(275) Healey, B. G.; Matson, R. S.; Walt, D. R. Anal. Biochem. 1997, 251, 270.

(276) Watts, H. J.; Yeung, D.; Parkes, H. Anal. Chem. 1995, 67, 4283.

(277) Algar, W. R.; Massey, M.; Krull, U. J. Trac-Trend. Anal. Chem. 2009, 28, 292.

(278) Nasef, H.; Beni, V.; O'Sullivan, C. K. Anal. Methods-Uk. 2010, 2, 1461.

(279) Wong, A. K. Y.; Marushchak, D. O.; Gradinaru, C. C.; Krull, U. J. Anal. Chim. Acta 2010, 661, 103.

(280) Liedl, T.; Simmel, F. C. Anal. Chem. 2007, 79, 5212.

(281) Fuchs, J.; Dell'Atti, D.; Buhot, A.; Calemczuk, R.; Mascini, M.; Livache, T. Anal. Biochem. 2010, 397, 132.

(282) Blake, R. D.; Delcourt, S. G. Nucleic Acids Res. 1996, 24, 2095.

(283) Russom, A.; Irimia, D.; Toner, M. Electrophoresis 2009, 30, 2536.

(284) McConaughy, B.L.; Laird, C. D.; McCarthy, B. J. Biochemistry-Us 1969, 8, 3289.

(285) Hutton, J. R. Nucleic Acids Res. 1977, 4, 3537.

(286) Prodhomme, S.; Demaret, J.-P.; Vinogradov, S.; Asseline, U.; Morin-Allory, L.; Vigny, P. J. Photochem. Photobio. B. 1999, 53, 60.

(287) Deniss, I. S.; Morgan, A. R. Nucleic Acids Res. 1976, 3, 315.

179

(288) Ikebukuro, K.; Kohiki, Y.; Sode, K. Biosens. Bioelectron. 2002, 17, 1075.

(289) Wang, R. H.; Tombelli, S.; Minunni, M.; Spiriti, M. M.; Mascini, M. Biosens. Bioelectron. 2004, 20, 967.

(290) Kerman, K.; Vestergaard, M.; Nagatani, N.; Takamura, Y.; Tamiya, E. Anal. Chem. 2006, 78, 2182.

(291) Juskowiak, B. Anal. Bioanal. Chem. 2011, 399, 3157.

(292) Lakowicz, J. R.; Cherek, H.; Gryczynski, I.; Joshi, N.; Johnson, M. L. Biophys. Chem. 1987, 28, 35.

(293) Norden, B. Appl. Spectrosc. Rev. 1978, 14, 157.

(294) Ediz, V.; Lee, J. L.; Armitage, B. A.; Yaron, D. J. Phys. Chem. A. 2008, 112, 9692.

(295) Matsuoka, Y.; Norden, B. Biopolymers 1982, 21, 2433.

(296) Silva, G. L.; Ediz, V.; Yaron, D.; Armitage, B. A. J. Am. Chem. Soc. 2007, 129, 5710.

(297) Carlsson, C.; Larsson, A.; Jonsson, M.; Albinsson, B.; Norden, B. J. Phys. Chem.-Us 1994, 98, 10313.

(298) Hill, H. D.; Vega, R. A.; Mirkin, C. A. Anal. Chem. 2007, 79, 9218.

180

Copyright Acknowledgements Exerpts of this thesis have been republished in part from several references that appear in the List

of Publications. Figure 2 (b) from Chapter one was adapted and reproduced in part, and the

copyright is listed below.*

Chapter 1

Copyright 2009: American Scientific Publishers

Copyright 2007: Taylor and Francis Group LLC

Smart biosensor technology by KNOPF, GEORGE K. Copyright 2007 Reproduced with permission of TAYLOR & FRANCIS GROUP LLC - BOOKS in the format Dissertation via Copyright Clearance Center.

Copyright 2005: Springer Science + Business Media

With kind permission from Springer Science + Business Media: Frontiers in Chemical Sensors, Challenges in the Design of Optical DNA Biosensors, Volume 3, 2005, 227-260, M. Massey, P.A.E. Piunno, U.J. Krull.

Copyright 2009: Elsevier Limited

Reprinted from Trends in Analytical Chemistry, Volume 28, W.R. Algar, M. Massey, U.J. Krull, The Application of Quantum Dots, Gold Nanoparticles and Molecular Switches to Nucleic Acid Diagnostics, 292-306, Copyright 2009, with permission from Elsevier.

Copyright 2008: Sinauer Associates, Inc.

* Figure 11.9 from Sadava, D.; Heller, H. C.; Orians, G. H.; Purves, W. K.; Hillis, D. Life: The Science of Biology, 8th Ed.; Sinauer Associates., p. 240. Copyright © 2008 Sinauer Associates, Inc. Reproduced with permission.

Chapter 2

Copyright 2006: Elsevier Limited

Reprinted from Analytica Chimica Acta, Volume 568, M. Massey, W.R. Algar, U.J. Krull, Fluorescence Resonance Energy Transfer (FRET) for DNA Biosensors: FRET pairs and Förster distances for various dye-DNA conjugates, 181-189, Copyright 2006, with permission from Elsevier.

181

Chapter 3

Copyright 2006: Springer Science + Business Media

With kind permission from Springer Science + Business Media: Journal of Fluorescence, Fluorescence Resonance Energy Transfer and Complex Formation Between Thiazole Orange and Various Dye-DNA Conjugates: Implications in Signaling Nucleic Acid Hybridization, Volume 16, 2006, 555-567, W.R. Algar, M. Massey, U.J. Krull.

Chapter 4

Copyright 2010: Springer Science + Business Media

With kind permission from Springer Science + Business Media: Analytical and Bioanalytical Chemistry, Towards a Fluorescent Molecular Switch for Nucleic Acid Biosensing, Volume 98, 2010, 1605-1614, M. Massey, U.J. Krull.