• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

The interpretation of ionic currents produced by controlled DNA translocation through the alpha-hemolysin nanopore

Dissertation
Author: Robert Frank Purnell
Abstract:
The past decade has seen the emergence of nanopores as highly sensitive single molecule detectors. Recently, there has been interest in using this technique to rapidly and inexpensively sequence single molecules of DNA. In this process, DNA is electrophoretically driven through the nanopore bathed in an electrolyte solution, and the resultant fluctuations in the current carried by the nanopore are used to characterize the DNA. For DNA sequencing, each of the four nucleotides in DNA - adenine, thymine, cytosine, and guanine -- must be detected individually and produce a current signals that are differentiable and identifiable. The primary focus of this thesis is to characterize and interpret blockade currents carried by the biological nanopore alpha-hemolysin (aHL) in the presence of single-stranded DNA (ssDNA). Examination of blockade currents produced by homopolymers of adenine, thymine and cytosine reveals the chemical orientation (3' leading or 5' leading) and identity of nucleotides of the homopolymer are important determinants of blockade current. In a follow up study, we find these current signals are highly sensitive to the identity of substituted nucleotides at multiple locations in an immobilized polythymine. Surprisingly, this sensitivity is neither a function of the geometry of the pore, nor the volume occupied by the substituted base. Blockade currents are in fact governed by base specific interactions between DNA and the aHL protein itself, a finding consistent with recent work on a mutant form of aHL. These results represent a significant contribution towards understanding the origins of blockade currents carried by aHL, and may prove useful in guiding the further development of nanopores for DNA analysis and sequencing. This thesis concludes with a discussion on sources of variability in the experimental system, their effects on currents measured in blocked (i b ) and clear (i o ) pores, and the validity of eliminating variation in ib through normalization by io .

TABLE OF CONTENTS List of Figures Lists of Tables and Equations vi Acknowledgements vii-viii Vita ix-xi Abstract xii-xiii Chapter 1 Introduction 1-9 Section 1.1 Methods for DNA Analysis 1 Section 1.2 Nanopore DNA Sequencing 1 Section 1.3 Experimental Concept 2 Section 1.4 Principle of Operation 3 Section 1.5 Stochastic Sensing 3 Section 1.6 Types of Nanopores 6-7 Section 1.7 DNA Sequencing with alpha hemolysin 7 Section 1.8 Thesis Overview 7-10 Subsection 1.8.a Motivation 7 Subsection 1.8.b Experimental Techniques 8-9 Chapter 2 Experimental Components and Influence on Current Data 10-23 Section 2.1 alpha-hemolsyin 10-12 Subsection 2.1.a Structure and Function 10 Subsection 2.1.b Role of the aHL structure in Blockade Current 11-12 Section 2.2 Lipid Membranes 12-14 Section 2.3 DNA 15-19 Subsection 2.3. a Structure and function 15-17 Subsection 2.3.b DNA Sensing 18 Section 2.4 Interpretation of DNA Translocation Signals 18-23 Subsection 2.4.a Immobilized DNA 18 i l l

Subsection 2.4. b Free translocation of DNA 19-22 Subsection 2.4.c Limitations due to Noise 22-23 Chapter 3 Methods for Controlled Translocation 24-39 Section 3.1 DNA Trapping and Flossing with Rotaxanes 24-29 Section 3.2 Rotaxane formation with a Magnetic Bead 29-38 Subsection 3.2.a Preliminary Results 31-34 Subsection 3.2.b Alternative Explanations 34-38 Chapter 4 Blockade Currents ofStreptavidin ImmobilizedHomopolymers 39-57 Section 4.1 Abstract and Introduction 39-41 Section 4.2 Experimental -Immobilization Protocol 41-42 Section 4.3 Blockade currents for Homopolymers 43-48 Section 4.4 Discuss and Conclusion 48-57 Chapter 5 Examination of PolyT Strands with Single Nucleotide Substitutions of A, G, and C 58-71 Section 5.1 Abstract and Introduction 58-60 Section 5.2 Experimental 60-63 Section 5.3 Results and Discussion 63-71 Chapter 6 Additional Sources of Variation in Currents Carried by Nanopores 72-77 Section 6.1 Normalization of Blockade Current by Open Pore Current 72 Subsection 6.1 .a Rationale 70-72 Section 6.2 Experimental Variations 72-76 Subsection 6.2.a ib/io and Variations in Current Across Experiments 72 Subsection 6.2.b ib/io and Variations in Current During an Experiment ; 72-76 Section 6.3 A True Value for ib? 76-79 Chapter 7 Conclusions and the Future ofNanopore DNA Sequencing 80-81 Appendix A Experimental Conditions from Chapter 5 81-85 Appendix B Determining the Number of Events for Each Strand 86 Appendix C Membrane Resistances from Chapter 6 87 References 88-94 IV

List of Figures 1.1 The Principle ofNanopore Sequencing 2 1.2 Stochastic Sensing 4 1.3 The Ideal DNA sequencing Nanopore 7 1.4 Experimental Approaches 9 2.1 alpha-hemolysin structure 11 2.2 Bilayer Formation Process 14 2.3 ssDNA Structure 16 3.1 Trapping DNA in the aHL pore 25 3.2 Monitoring Formation of a DN A/PEG Rotaxane 26 3.3 Idealized Current Trace for Trapped A/C ssDNA Copolymer 27 3.4 Current Measurements for A/C Copolymer ssDNA Rotaxane 28 3.5 Schematic of Opposing Electrophoretic and Magnetic Force 30 3.6 Conductance States from A/C Copolymer Trapped with Magnetic Bead 32 3.7 Translocation of an A40 Rotaxane 35 3.8 Threading an A 40 ssDNA strand trapped by Stem Side Insertion 36 4.1 Immobilization of a DNA strand in a Nanopore 40 4.2 Procedure and Resulting Currents in the Open and Blocked States 43 4.3 Blockade Currents for Strands Threaded 5' end first 45 4.4 Currents for Strands Threaded 3' end first 47 4.5 5' and3' Threading 54 5.1 Schematic Diagram of single substituted polyT DNA strand 60 5.2 Blockade currents ofpolyT, polyTAt% andpolyl\w 61 5.3 Correction for a Shift in Blockade Current 63 5.4 Graph of Mb for A, C, and G substitutions in a polyT strand at Positions 6-17 64 5.5 Average hfor polyToji and polyT 65 6.1 i0 and it Measurements During an Experiment 77 v

List of Tables 4.1 Blockade Currents from 5' leading strands ofpolyA, polyC, andpolyT 46 4.2 Blockade Currents from 3' leading strands ofpolyA, polyC, andpolyT 48 4.3 Concentrations and Probabilities for 5' and 3' leading orientations 49 4.4 Measured 3' and 5' leading blockage currents for poly A, polyC, andpolyT 55 4.5 Concentrations and probabilities for simultaneous 5' and 3' experiments 55 6.1 The Effect of Normalization on the consistency of Reported Current Data 75 List of Equations 2.1 Inverse Debye Length 17 2.2 Electrophoretic Force on ssDNA 20 3.1 Magnetic Force 31 6.1 Pore Conductivity 73 6.2 Solution Conductivity and Temperature 73 6.3 Solution Conductivity and Concentration 73 6.4 Normalization ofii/i0 73 6.5 Normalization w/compensation for Gmem 78 vi

ACKNOWLEDGEMENTS I thank my advisor Jacob Schmidt for his support and guidance at every step throughout my career as a graduate student at UCLA. His patience makes him an excellent mentor and his hands-off approach encourages his students - both undergraduate and graduate - to explore their topics independently while thinking critically. As one of the first graduate students in Jake's research group, it has been especially rewarding to participlate in building a research lab, provide advice to the new graduate students, and mentor undergraduate researchers. I'd like to thank my undergraduate students who worked for me in the research lab and attended my discussion sections for providing me with the opportunity to mentor them. Training them as naive undergraduate students and watching them develop into independent minded researchers in the laboratory and classroom was particularly rewarding. In particular, I'd like to thank Kunal Mehta, an undergraduate who became a great researcher and colleague, for gathering data, providing intellectual discussion, and writing papers. The process of scientific research can be extremely satisfying one day and utterly frustrating the next. I'd like to thank my fellow colleagues and members of the Schmidt lab during my time at UCLA for many hours of (sometimes) stimulating discussion and moral support. They include Tae-Joon Jeon, Noah Malmstadt, Jason Poulos, Dave Wendell, Hyun-woo Bang, Robert Tan, and Denise Wong. I would also like to thank Dr. Aleksei Aksimentiev for discussions on VMD simulations and Dr. Edward vii

Buck for his hours on the phone assisting me with setting up the bilayer amplifier and recording system. This thesis contains several reprints of figures and text from published articles. Figure 3.2 (p. 26) has been printed with permission from the following source with the required permission of the journal and author: Sanchez-Quesada, J., Saghatelian, A, Cheley, S., Bayley, H., Ghadiri, M.R. Single DNA Rotaxanes of a Transmembrane Pore Protein. Angewandte Chemie, International Addition, 2004, 42, (23), 3063-3067. The text and figures in Chapter 4 are based on "Purnell, R.F., Mehta, K.K., and Schmidt, J. J. Nucleotide Identification and Orientation Discrimination of DNA Homopolymers Immobilized in a Protein Nanopore. Nanoletters 2008, 8, (9), pp. 3029-3034." Chapter 5 is based on "Purnell, R.F. and Schmidt, J.J. Discrimination of Single Base Substitutions in a DNA Strand Immobilized in a Biological Nanopore. ACS Nano, 2009, 3 (9), pp. 2533-2538." For journal articles 4 and 5,1 would like to acknowledge Stephen Cheley for his kind donation of aHL protein that was used the experiments and following funding sources which supported these studies: National Science Foundation Career Award, American Chemical Society Petroleum Research Fund, and the UCLA Faculty Research Grant Program. Vl l l

VITA August 10th, 1981 Born, Plainfield, New Jersey 2003 B.S. Bioengineering University of California, Berkeley Berkeley, California 2003-2005 M.S. Biomedical Engineering University of California, Los Angeles Los Angeles, California 2004-2008 Graduate Student Instructor Department of Bioengineering University of California, Los Angeles PUBLICATIONS AND PRESENTATIONS Malmstadt, N., Nash, M.A., Purnell, R.F., and Schmidt, J.J (2006). Microfluidic lipid bilayer membrane fabrication by solvent extraction: An automated platform for ion channel studies. Presented at 2006 University of California System- wide Bioengineering Symposium, Los Angeles, California. Malmstadt, N., Nash, M.A., Purnell, R.F., and Schmidt, J.J (2006). Automated formation of lipid bilayer membranes in a microfluidic device. Nano Letters, 6 (9): 1961-1965 IX

Mehta, K.K., Purnell, R.F., and Schmidt J.J. (2008) Effect of Strand Orientation on Blockage Currents of Single Stranded DNA Immobilized with a Biological Nanopore. Presented at the Biophysical Society Conference, Long Beach, California. Purnell, R.F. (2008). Identifying DNA homopolymers with a biological nanopore. Presented at the UC Bioengineering Symposium, Riverside, California. Purnell, R.F., Mehta, K.K., and Schmidt, J.J (2008). Nucleotide Identification and Orientation Discrimination of DNA Homopolymers Immobilized in a Protein Nanopore. Nano Letters, 8 (9):3029-3034. Purnell, R.F., and Schmidt, J.J (2009). Discrimination of Single Base Substitutions in a DNA strand Immobilized in a Biological Nanopore. ACS Nano, 3, (9): 2533- 2538. Purnell, R.F., Mehta, K.K., and Schmidt J.J. (2009). Nucleotide Identification and Orientation Discrimination of DNA Homopolymers Immobilized in a Protein Nanopore. Presented at the Biophysical Society Conference, Boston, Massachusetts. Purnell, R.F., Mehta, K.K., and Schmidt J.J. (2009). Nucleotide Identification and Orientation Discrimination of DNA Homopolymers Immobilized in a Protein Nanopore. Presented at the National Human Genome Research Institute Meeting, San Diego, California. Purnell, R.F., Mehta, K.K., and Schmidt J.J. (2008). Single Molecule Measurements of DNA Immobilized in a Biological Nanopore. Presented at the Biophysical Society Conference, Long Beach, California. Purnell, R.F., Mehta, K.K., and Schmidt J.J. (2007) Controlled DNA Translocation through a Biological Nanopore for Rapid DNA Sequencing. Presented at the Frontiers in Nanosystems Conference, Los Angeles, California. Purnell, R.F., Mehta, K.K., and Schmidt J.J. (2007) Controlled DNA Translocation through a Biological Nanopore for Rapid DNA Sequencing. Presented at the Biophysical Society Conference, Baltimore, Maryland. Purnell, R.F and Schmidt J.J. (2006). Controlling DNA Translocation through a Biological Nanopore for Rapid DNA Sequencing. Robert Purnell and Jacob J. Schmidt. Presented at the University of California System-wide Bioengineering Symposium, Los Angeles, California. x

Purnell, R.F. and Schmidt J.J. (2006). Magnetically Controlled DNA Translocation through a Nanopore. Presented at the 2006 UCLA Engineering Research Review, Los Angeles, California. XI

ABSTRACT OF THE DISSERTATION The Interpretation of Ionic Currents Produced by Controlled DNA Translocation through the alpha-Hemolysin Nanopore by Robert Frank Purnell Doctor of Philosophy in Biomedical Engineering University of California, Los Angeles, 2009 Professor Jacob J. Schmidt, Chair The past decade has seen the emergence of nanopores as highly sensitive single molecule detectors. Recently, there has been interest in using this technique to rapidly and inexpensively sequence single molecules of DNA. In this process, DNA is electrophoretically driven through the nanopore bathed in an electrolyte solution, and the xii

resultant fluctuations in the current carried by the nanopore are used to characterize the DNA. For DNA sequencing, each of the four nucleotides in DNA - adenine, thymine, cytosine, and guanine - must be detected individually and produce a current signals that are differentiable and identifiable. The primary focus of this thesis is to characterize and interpret blockade currents carried by the biological nanopore alpha-hemolysin (aHL) in the presence of single-stranded DNA (ssDNA). Examination of blockade currents produced by homopolymers of adenine, thymine and cytosine reveals the chemical orientation (3' leading or 5' leading) and identity of nucleotides of the homopolymer are important determinants of blockade current. In a follow up study, we find these current signals are highly sensitive to the identity of substituted nucleotides at multiple locations in an immobilized polythymine. Surprisingly, this sensitivity is neither a function of the geometry of the pore, nor the volume occupied by the substituted base. Blockade currents are in fact governed by base specific interactions between DNA and the aHL protein itself, a finding consistent with recent work on a mutant form of aHL. These results represent a significant contribution towards understanding the origins of blockade currents carried by aHL, and may prove useful in guiding the further development of nanopores for DNA analysis and sequencing. This thesis concludes with a discussion on sources of variability in the experimental system, their effects on currents measured in blocked (it,) and clear (i0) pores, and the validity of eliminating variation in ib through normalization by i0. xiii

Chapter 1 - Introduction Methods for DNA Analysis 1.1 Methods for DNA analysis Encoded in the sequence of nucleotides of a DNA strand are the instructions for the regulation and maintenance of biological functions. Recent success of the human genome project has created a demand for techniques that extract this information quickly and cheaply. Despite efforts to reduce the cost of traditional Sanger-sequencing based methods with the development of array based methods, the estimated cost to sequence a human genome still ranges anywhere from the optimistic figure of 50,000 to well over 100,000 US dollars '. This is due primarily to the quantity of fluorescently labeled nucleotides and DNA polymerase enzyme required to run the necessary number of primer extension reactions. 1.2 Nanopore DNA Sequencing An alternative approach for sequencing DNA involves direct examination of the DNA structure and sequence by threading through a biological or synthetic nanopore.3 Nanopores provide a suitable environment for direct sequential examination of DNA; their confined geometry requires individual DNA strands, which can form aggregate secondary structures in solution4'5, to unwind and pass through in a single file manner. Thus, it does not require costly chemical reagents associated with Sanger-based sequencing methods. In addition, since this process examines a single strand at a time, the amount of DNA required would be significantly less than used in Sanger Sequencing reactions. 1

Time Figure 1.1 The Principle of Nanopore Sensing a) The application of a voltage across the nanopore generates an open pore current, b) DNA (silver) is driven through the nanopore, resulting in a transient decrease in current for a short time (At) to a blocked state, c) Complete translocation is indicated when open pore current is restored, d) In a DNA sensing experiment translocation events are characterized by the magnitude of the current change Ai and the duration of the event At. 1.3 Experimental Concept To analyze a single DNA strand, an electrical potential is applied across a nanometer scale pore bathed in an electrolyte solution. This voltage potential generates an ion current and drives negatively charged DNA through the pore. In the presence of a DNA strand, the current is temporarily reduced to a blocked state. Strands are characterized by the magnitude of the current during DNA translocation and the duration of the blocked state. A conceptual diagram of this process is shown in figure 1-1. 2

1.4 Principle of Operation This process can be viewed as molecular Coulter-counting. In a Coulter-counter, hydrostatic pressure is used to drive an ionic solution containing biomolecules through an aperture; a voltage, applied across the aperture using two silver chloride reference electrodes, probes the conductance of the fluid in the aperture. The passage of a bio- molecule through the aperture generates a change in resistance to ion flow through the aperture, as predicted by R = p*L/A, where p is the resistivity of the solution, A is the cross-sectional area and L is the length of the channel. Since the ionic current carried by the channel is directly related to cross-sectional area, identification of molecules of similar size from their characteristic resistance pulses requires the aperture have dimensions comparable that of the target molecule(s). Accordingly, the Coulter-counter, used for whole blood counting, has an aperture size on the order of 10s of microns in diameter,6 or about the size of a red or white blood cell.7 Similarly, DNA sequencing, or detection of single molecules, requires a nanometer scale pore to produce currents capable of sensing the differences in the structures of nucleotides on a DNA strand. 7.5 Stochastic Sensing There are several important differences between sensing at the micro and nanometer scale. In nanopores, the process of molecular detection has been termed 'Stochastic Sensing'. In contrast to a traditional Coulter-counter, the current carried by a nanopore is sensitive to the molecular structure of an analyte, the charge distribution in a Q q channel, and the pH of the solution. These properties have been exploited to monitor chemical reactions near the channel10"12 and to detect toxins, ions, and biopolymers.13 A 3

demonstration of the power of stochastic sensing using the alpha-hemolysin (aHL) protein (aHL) was presented in work by Gu et al, where an electrolyte solution containing the divalent cations cadmium, cobalt and nickel produced three identifiable blockade currents when examined by a single mutant aHL pore equipped with a divalent cation binding site.14 A schematic of this experiment is shown in figure 1.2. This binding site (red in figure) consisted of several histidines introduced into pore wall through genetic mutation of the aHL sequence. These results suggest the mutant pore was sensitive to the charge as well as the molecular weight of the analyte, while the wild type pore, which lacked the binding site, was not. Time OOOOOO1 Figure 1.2 Stochastic Sensing with an Engineered aHL nanopore a) Under an applied voltage, charged ions (green spheres) are drawn through the nanopore, creating a current (green trace), b) In the mutant nanopore, binding of charge carriers to the recognition site (red) results in a decrease in the magnitude of the current (red trace). 4

1.6 - Types of Nanopores Nanopores sensors can be categorized into two groups: biological and inorganic. Recent development of nanofabrication techniques employing ion track etching, feedback controlled ion beam sculpting, and electron-beam etching, have enabled the production of nanometer-sized inorganic pores.15"17 Durability of inorganic nanopores enables integration into a variety of experimental conditions. This technology has been used to detect dsDNA,18 the formation of a DNA protein complex on a template strand,19 and to analyze force-induced extension of DNA strands trapped in the nanopore.20'21 By contrast, biological nanopores operate at a limited range of experimental conditions and require a fragile lipid membrane for proper function which typically breaks under applied 99 voltages greater than 200 mV. In addition, different ion channel proteins require special methods for reconstitution into membranes. aHL can be prepared in buffer solution and added straight to experimental solution; however, other proteins must be prepared in a detergent solution (MspA)23, reconstituted in vesicles, or require a concentration gradient of salt across the bilayer for successful insertion (Maxi-K).24 Nevertheless, the majority of nanopore sensing experiments have focused on using biological nanopores, because fabrication of artificial nanopores is expensive and requires further development for reliable mass production, limiting this technology to research laboratories. Biological nanopores, including aHL and the recently developed mutant form of MspA currently being investigated by the Gundlach Lab at the University of Washington25 represent an attractive alternative. In contrast to inorganic pores, they can be produced cheaply through well established recombinant DNA techniques. Mutations 5

in the gene sequence of a nanopore protein can generate reproducible modifications to the pore structure, enabling engineering of biological nanopores as sensors, equipped with binding sites specific to a single molecule. The natural protein folding process produces copies identical in structure; by contrast, synthetic nanopores cannot be engineered at this scale, and their sizes vary due to uncertainty in the manufacturing process. In addition, biological nanopores, specifically aHL, have dimensions which allow passage of single stranded DNA (ssDNA), but not double-stranded DNA (dsDNA),26 making them useful 97 98 for nucleic acid hybridization studies, ' while fabrication of artificial nanopores at this scale not yet a reproducible process. These advantages have enabled more extensive research on biological nanopores, in spite of the limitations on experimental conditions. 1.7 - The Concept of DNA Sequencing with aHL In the process of nanopore DNA sequencing, the bases attached to the DNA backbone are identified as the strand is passed through the pore. The dimensions of aHL are similar ssDNA, making it a suitable candidate for DNA analysis. For DNA sequencing, the pore must be sensitive enough to detect the difference in molecular structure for each of the 4 bases in DNA.29 Given the molecular weight of each nucleotide, with guanine being heaviest, followed by adenine, cytosine and then thymine in descending order, one might expect, based on the principles of coulter counting, the currents would be unique. Further, we might expect the magnitudes of blockade current for the purines (the larger bases) to be smaller than pyrimidines. An additional requirement to produce this idealized current trace is that each nucleotide must be read 6

b) c) A r t A A r t HnffBnnffHi 3SS3b3ooo HnHnHHHHI oodoBoHbob €) J d) open state t Ai blocked state | L. -L T _ - " open state j C ,. , A Time Figure 1.3 The Ideal DNA sequencing Nanopore In an ideal nanopore DNA sensor, a DNA strand is driven through a nanopore, pictured here embedded in an insulating membrane. During the translocation of the DNA strand fluctuations in the current signal correspond to the order of the nucleotides on the strand. In this example thymine bases (blue trace) block the least, followed by cytosine (red), adenine (green) and guanine (purple. individually as it passes through the nanopore. This idealized nanopore sequencing result is shown in figure 1-3. 1.8 - Thesis Overview 1.8.a Motivation While much progress has been made in elucidating the underlying mechanisms that give rise to observed blockade currents in DNA translocation, extraction of the 7

nucleotide sequence from the ionic current has been complicated by the speed of translocation (~ 1 us / base), the minute difference between blockade currents for different nucleotides (typically no greater than 10 pA),2'30 and the sensitivity of blockade currents to characteristics of the polynucleotide other than the base composition, including the direction of threading of the strand31 and the formation of secondary structure.32 To overcome these limitations and further probe the sensitivity of nanopores for DNA sequencing, recent studies have explored several methods of reducing translocation rate. One approach, which involved controlling the temperature and viscosity of the DNA solution and using a lower magnitude voltage to induce translocation, reduced the speed of DNA passage through the pore rate through an artificial nanopore by approximately an one order of magnitude; however, this also reduced the magnitude of the current signal, thus leading to only modest improvements in the SNR33. 1.8.b Experimental This thesis focuses on the interpretation of high precision current measurements obtained from controlling the rate of DNA translocation. This first approach involves DNA trapping and 'flossing'3 ' 5 where an ssDNA strand is trapped in a rotaxane formation with a hairpin on one end, and a magnetic microbead or streptavidin on the other. This experiment and its results are discussed in detail in chapter 3. In the second approach, discussed in chapters 4 and 5, strands were immobilized for a brief time in the pore using a terminal streptavidin, a technique first employed by Kasianowicz36 and later used for sensing of DNA oligomer hybridization. Using streptavidin immobilization, we 8

discovered critical determinants of blockade current include the identity of nucleotides in a homopolymer, the threading orientation (5' leading or 3' leading), and the identity and position of singly substituted nucleotides at multiple positions in an otherwise homopolymer strand. In these experiments, we also observed temperature and conductivity of the DNA solution appeared to have a distinct affect on current in the open and blocked states. Chapter 6 discusses these observations in the context of previous work where blocked current data was reported as a percentage of the open pore current. a) Trapping b) Immobilization Figure 1-4 - Experimental Approaches Two approaches for controlling DNA translocation through a aHL nanopore embedded in a lipid bilayer membrane, a) DNA, represented by silver beads is trapped in the pore by a terminal hairpin and magnetic bead (silver silver sphere), which enables application of a magnetic force to oppose electrophoretic force supplied by the transmembrane voltage, b) DNA terminated with a single streptavidin (diamond) is immobilized in the pore. 9

Chapter 2 Experimental Components The aHL sensing system is composed of three principle components, the aHL protein pore, a lipid bilayer membrane, the electrolyte solution containing the DNA to be examined, and low noise amplifier to measure and record the resultant current signal. This chapter discusses the natural structure and function of each system and their respective influences on current signals observed in the aHL nanopore. 2.1 aHL 2.1.a Structure and Function of aHL The aHL protein (figure 2.1) belongs to a family of toxic transmembrane channel proteins that evolved to serve as a self-defense mechanism for the bacterium Staphylococcus aureus. The protein is secreted as water soluble monomers, which self- assemble on the surface of target cells to form a nonspecific heptameric pore structure composed of two regions: a vestibular cavity and (3-barrel, each with a length of approximately 5 nm, the approximate thickness of a lipid membrane. The outer wall of the p-barrel is composed largely of hydrophobic residues, which stabilize the pore structure when inserted into the membrane; the interior is made up of largely hydrophilic residues. Thus, the insertion of aHL amounts to puncturing a relatively large hole in the cell wall, which disturbs the ionic equilibrium, leading to osmotic swelling and cell lysis. The fact that these pores have been useful for DNA sensing is purely a coincidence. 10

Vestibule Constriction Beta barrel Figure 1 -The Structure of the alpha-Hemolysin Nanopore The alpha hemolysin (aHL) is composed of two regions, a vestibular cavity and a beta barrel or stem. The inner pore diameter ranges from 4.6 nm, in the vestibule, to a mi ni mum of ~ 1.4 nm, in the constriction, located between the stem and vestibule. The interior wall of the pore lumen (in shadow) is made up of predominantly hydrophilic amino acid residues, to allow for high permeability of electrolytes, while the exterior wall is composed of hydrophobic amino acids which facilitate insertion into a eel I membrane. 2. Lb Role of aHL Structure in Blockade Currents The interpretation ionic current carried by aHL in the blocked state amounts to understanding nanometer scale interactions between two single molecules and their effects on blockade current. The complete structure of aHL, determined by x-ray crystallography in 1996 by Song et al,26 has proven invaluable in relating the structure of 11

the pore to its conductance. The first single molecule experiments revealed otHL conductance in a 1M KC1 solution is noticeably larger when positive voltage is applied to the solution bathing the entrance to the p-barrel of the pore, giving a ratio of G+/G- of 1.5 at pH 7.39 This phenomenon is related to the structure of the pore, which ranges from 4.7 nm at its maximum in the vestibule, to a minimum of 1.4 nm at the entrance to the stem region.26 Under applied voltage, this structure produces an asymmetrical distribution of energetic barriers to current flow along the channel axis. There is also a slight anion selectivity of the pore at neutral pH,40 which has been linked to the presence of charged residues Lysine and Glutamate, located in the pore constriction. Open pore currents can also exhibit transient reductions in blockade current, or 'gating', in the presence of divalent cations in acidic conditions (pH 5.8) due to deprotonation of amino acid residues in the constriction, e.g., Glutamate (pkA = 4.4) and stem, e.g. Aspartate (pkA = 3.7).40,41 Since the structure of aHL is highly sensitive to the electrolyte conditions, DNA sensing experiments are conducted in basic solutions that contain chelating agents such as EDTA to bind and sequester divalent cations present in solution. 2.2 Lipid Membranes For proper function as a stochastic sensor, aHL must self-assemble into lipid bilayers. Lipid bilayers are 5 nm thick structures composed of amphiphilic molecules that contain a hydrophobic lipid tail on one end and a hydrophilic phosphate group on the other. 2 This property of drives self-assembly into structures (e.g., vesicles, liposomes, and cell membranes) that minimize the exposure of the lipid tails to the aqueous phase. Natural functions of the lipid bilayer include maintenance of the ionic gradient across a 12

cell, transport of metabolites to and from the extracellular medium, and protection against pathogens. In nanopore sensing experiments, lipid bilayers electrically isolate solutions containing molecules to be examined by the nanopore. They can be formed by contacting phospholipid monolayers on aqueous droplets in a lipid solution,43 through organic solvent extrusion in a microfluidic device,44 or by painting a lipid solution across an aperture drilled in a plastic partition. In each of these scenarios, the process is the essentially the same: two monolayers are brought together to form a single stable bilayer. In traditional DNA sensing experiments, free-standing lipid bilayers are created by painting a lipid solution of 1-3% (w/v) diphytanoylphosphatidylcholine in «-decane on 100 ^m orifices in 10 fjm thick Teflon partitions.45 The Teflon partition is sandwiched between two measurement chambers filled with a buffer solution of 1 M KC1, 1 mM EDTA, and 10 mM Tris-HCl. The self-assembly process of free-standing bilayers, shown in figure 2.2, is driven by the Laplace Pressure difference across the interface separating the aqueous a lipid phases,45 which extrudes the solvent from between the monolayers, forming an annulus between the bilayer and the Teflon film. This process is monitored electrically through resistance and capacitance measurements across the Teflon film. In our system, formation of a stable bilayer is indicated by an increase in the measured capacitance and resistance to 25 pF and 100 G£2, respectively. At 120 mV, this gives a leakage current of approximately 1-2 pA. After a stable membrane is formed, 1 /uL of 1.7 fig/mL otHA in 100 mM Tris-HCl, 200 mM NaCl, at pH 8.2 is added to the cis side of the chamber. Insertion of single aHL channel is accompanied by an abrupt 13

increase in current from 1-2 pA, to 115 pA at 120 mV. The membrane leakage current, which is typically 1-2% of the current observed after aHL incorporation, is typically ignored in most aHL sensing experiments. The effect of membrane leakage current on DNA sensing experiments will be discussed further in chapter 6. a) b) Figure 2.2 - Bilayer Formation Process a) Traditional bilayer experiments are formed by applying a lipid solution of n-decane across a 100-500 mm aperture in a teflon partition that is sandwiched between two teflon reservoirs, b) The resistance and capacitance of the bilayer are monitored by applying a voltage across silver chloride reference electrodes, c) After application of the lipid solution, a pressure difference across the lipid/aqueous interface drives extrusion of the n-decane solution from center of the aperture, forming a bilayer circumscribed by a solvent annulus. 14

Full document contains 109 pages
Abstract: The past decade has seen the emergence of nanopores as highly sensitive single molecule detectors. Recently, there has been interest in using this technique to rapidly and inexpensively sequence single molecules of DNA. In this process, DNA is electrophoretically driven through the nanopore bathed in an electrolyte solution, and the resultant fluctuations in the current carried by the nanopore are used to characterize the DNA. For DNA sequencing, each of the four nucleotides in DNA - adenine, thymine, cytosine, and guanine -- must be detected individually and produce a current signals that are differentiable and identifiable. The primary focus of this thesis is to characterize and interpret blockade currents carried by the biological nanopore alpha-hemolysin (aHL) in the presence of single-stranded DNA (ssDNA). Examination of blockade currents produced by homopolymers of adenine, thymine and cytosine reveals the chemical orientation (3' leading or 5' leading) and identity of nucleotides of the homopolymer are important determinants of blockade current. In a follow up study, we find these current signals are highly sensitive to the identity of substituted nucleotides at multiple locations in an immobilized polythymine. Surprisingly, this sensitivity is neither a function of the geometry of the pore, nor the volume occupied by the substituted base. Blockade currents are in fact governed by base specific interactions between DNA and the aHL protein itself, a finding consistent with recent work on a mutant form of aHL. These results represent a significant contribution towards understanding the origins of blockade currents carried by aHL, and may prove useful in guiding the further development of nanopores for DNA analysis and sequencing. This thesis concludes with a discussion on sources of variability in the experimental system, their effects on currents measured in blocked (i b ) and clear (i o ) pores, and the validity of eliminating variation in ib through normalization by io .