Development of biomimetic sensing elements based on synthetic nanopores

MISSING IMAGE

Material Information

Title:
Development of biomimetic sensing elements based on synthetic nanopores
Physical Description:
xiii, 123 leaves : ill. ; 29 cm.
Language:
English
Creator:
Heins, Elizabeth A
Publication Date:

Subjects

Subjects / Keywords:
Chemistry thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2005.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Elizabeth A. Heins.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 003320035
System ID:
AA00008990:00001

Full Text










DEVELOPMENT OF BIOMIMETIC SENSING ELEMENTS BASED ON
SYNTHETIC NANOPORES












By

ELIZABETH A. HEINS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005





























Copyright 2005

by

Elizabeth A. Heins






























This dissertation is dedicated to my parents,
Geoffrey A. Heins and Lorraine F. Kreiner.














ACKNOWLEDGMENTS

There are always far more people that help us achieve our goals than we

are ever able to acknowledge. I humbly apologize if I have forgotten anyone

here-please know that I have appreciated all you have done.

I would like to thank my research advisor, Dr. Charles R. Martin, for the

opportunities and experiences that I have had. I would also like to thank the

other members of my research committee, Drs. Richard Yost, Anna Brajter-Toth,

Sean Sullivan, and Barry Ache, for the interest they have taken in my research

and for the thoughtful suggestions made along the way.

I have been privileged to work with a most interesting group of people over

the years. Some have provided great insight into my research and some have

provided the friendship that makes everything worth it. Former Martin Group

members Drs. David Mitchell, Scott Miller, Shufang Yu and Marc Wirtz started

me on the right path. Current members Drs. Zuzanna Siwy and Lane Baker

helped guide me though to the end and have taught me more than they know.

Damian Odom, John Wharton, Lacramiora Trofin, Robbie Sides, Chad Harrell,

Fatih Buyukserin, Heather Hillebrenner, and Miguel Mota shared the journey.

I am also very fortunate to have spent time working with Dr. Paul Cremer

and his group as a Visiting Scholar of Texas A&M University. Along with Dr.

Cremer, Dr. Tinglu Yang and Fernando Albertorio helped fuel my passion for lipid

bilayers. As a part of my collaboration with TAMU, I have also been privileged to








work with Dr. Hagan Bayley (now at Oxford University) and Dr. Stephen Cheley,

using their beautiful aHL in many of my experiments over the years.

This journey is not possible without the love and support of one's friends

and family. Those that have made sure I remembered to laugh include Arthur

Bides, Dave Mitchell, Damian Odom, Zuza Siwy, Fernando Albertorio, Ken

Healy, Liisa Kauri, Lindsey Hebert, Bruce and Charlotte Dare, and John Janzer

(whose sage words of irreverent wisdom kept me sane). My mother and step-

father, Lorraine and Hans Kreiner; my father and step-mother, Geoffrey and

Nancye Heins; and my brothers Wills (and his wife and daughter, Lisa and Leigh-

Anne) and Jonathan Heins, have always been there, even when they had no

idea what I was doing. James Cline, his daughter Kadi, and Richard Cline, and

Jeanette and Bill, Bill, and Michael Jones have also been wonderfully supportive.

I must not forget to acknowledge those who have unknowingly been a

constant source of inspiration: Drs. Kathryn Williams and Samuel Colgate in the

Department of Chemistry at the University of Florida as well as Dr. A. Samuel

Kimball in the Department of English and Drs. Stuart Chalk, Robert Vergenz,

Edward Healy, and Kunisi Venkatasubban in the Department of Chemistry at the

University of North Florida. Dr. Healy was right; in time I did indeed learn some

chemistry.



If we knew what it was we were doing, it would not be called research, would it?

Albert Einstein














TABLE OF CONTENTS

paae

ACKNOW LEDGMENTS ................................................................................... iv

LIST OF FIGURES ........................................................................................... ix

A B S T R A C T ............................................................................................... ...... xii

CHAPTER

1 INTRODUCTION AND BACKGROUND...................................................... 1

Biological Nanopores................................................................................... 1
Synthetic Nanopores as Biomimetic Sensing Elements............................ 11
Irradiation of Polymer Membranes............................................................. 15
Etching of Single Conical Nanopores in Polymer Membranes .................. 16
Sizing of a Conical Nanopore.................................................................... 17
Calculating the Effective Pore Length of a Conical Nanopore................... 18
Rectifying Properties of a Conical Nanopore............................................. 19
The Ratchet Model .................................................................................... 21
Dissertation Overview................................................................................ 21

2 CHEMICAL CONTROL OF CURRENT RECTIFICATION IN A
SYNTHETIC NANOPORE W ITH CROWN ETHER................................... 25

Introduction ........................................................................................ ....... 25
Experimental ............................................................................................. 26
M ate rials ..................................................................................... ....... 26
Preparation of Conical Nanopores ...................................................... 27
Estimation of the Tip Diameter, dt....................................................... 27
Electrochemical Measurements.......................................................... 28
Results and Discussion............................................................................. 29
Chemistry and Charge of the Polyimide Nanopore............................. 29
Effect of 18C6 on Current-Voltage Curves with KCI as the Electrolyte.. 30
Effect of 18C6 on Current-Voltage Curves with LiCI as the Electrolyte.. 35
Effect of PEG" on Current-Voltage Curves....................................... 37
The Effect of 18C6 on the Extent of Rectification................................ 39
Comparison of Synthetic Conical Nanopores vs. a Model Biological
Nanopore, aHL ................................................................................ 42
C onclusion........................................................................................ ........ 45








3 STOCHASTIC SENSING DETECTION OF AN ANIONIC PORPHYRIN
USING A SYNTHETIC CONICAL NANOPORE........................................ 47

Introd auction ............................................................... ..... ........... ..............47
E xpe rim e nta l ...................................... .................. ..................................4 9
M ate ria ls ......................................... ....................... ........................... 49
Preparation of the Conical Nanopores and Estimation of Tip Diameter. 50
Current-Time Recordings.................................................................... 50
Results and Discussion ............................................................................. 51
Current-Time Response of Nanopore in 1 M KCI................................ 52
Voltage and Concentration Dependence of the Current-Time
Response of Nanopore .................................................................... 55
Blockage of Current by H2-TPPS ........................................................ 61
C conclusion ........................................................................................ ........ 69

4 POLYMER FILMS AS SUBSTRATES FOR MOBILE SUPPORTED LIPID
B ILA Y E R S .............................................................................................. ... 7 1

Introduction ........................................................... ................. ........ ............7 1
Experim ental .................................................................................... ......... 76
M ate ria ls ............................................................................ ................. 76
Preparation of Substrates for Surface Characterization and
Fluorescence Experiments................................... ...........................78
Preparation of Vesicles........................................ ...............................78
Dynamic Light Scattering to Determine Vesicle Size.......................... 79
Bilayer Formation................................................................................ 79
Fluorescence Recovery After Photobleaching Experiments .............. 80
Atomic Force Microscopy (AFM) Surface Characterization
Experiments........................................... ...................................... 81
Preparation of the Conical Nanopores and Estimation of Tip Diameter. 82
Electrochemical Measurements .......................................................... 83
Results and Discussion ............................................................................. 84
Sizing Small Unilamellar Vesicles Formed by Extrusion........... .......84
Surface Characterization.......................................................... ........84
Fluorescent Recovery After Photobleaching (FRAP).... ................... 86
Electrochemistry ............................................... ................... ............92
C onclusion..................................................................... ................... ........ 95

5 USE OF THE POLYMER MEMBRANE-SUPPORTED LIPID BILAYER
SYSTEM AS A BIOSENSING PLATFORM............................................... ...96

Introd uctio n ....................................................................... .........................96
E xpe rim enta l ............................................................. ................................... 98
M ate rials ............................................................... ..... ..................... 98
Preparation of the Conical Nanopores and Estimation of Tip Diameter. 99
Preparation of Vesicles for Lipid Bilayer ............................................. 99
Formation of Lipid Bilayer on Polymer Support................................. 100








Electrochemical M easurements ........................................................ 101
Results and Discussion ........................................................................... 102
Conclusion............................................................................................... 109

6 CONCLUSION ......................................................................................... 110

LIST OF REFERENCES ................................................................................ 114

BIOGRAPHICAL SKETCH ............................................................................. 123














LIST OF FIGURES


Figure page

Figure 1-1. Lipid bilayer surrounding a cell showing transmembrane proteins..... 2

Figure 1-2. Examples of biological nanopores shown in ribbon or ball-and-stick
representations................................................................................. ........2...

Figure 1-3. Example of a voltage-gated potassium channe..............................4...

Figure 1-4. Schematic showing the theory of the Coulter counter technique....... 6

Figure 1-5. Dimensions of a/pha-hemolysin (aHL) channel ..............................7...

Figure 1-6. Current-time recordings demonstrating effect on conductance of aHL
nanopore in absence and presence of poly(ethylene) glycol (PEG)
molecules3 and crown ethers. ................................................................... 8

Figure 1-7. Examples of engineered aHL nanopores detecting different types of
a na lyte s .......................................................................................... ........ 10

Figure 1-8. Examples of synthetic nanopores with dimensions analogous to
biological nanopores ............................................................................... 12

Figure 1-9. Examples of sensing-capabilities of single conical nanopore
m em branes ..................................................................................... ....... 14

Figure 1-10. Principle behind the creation of latent tracks in polymeric
membranes via irradiation with swift heavy ions...................................... 15

Figure 1-11. Electrochemical set-up for chemically etching conical nanopores in
polymeric membranes. ............................................................................ 17

Figure 1-12. Electrostatic potential inside conical nanopore........................... 20

Figure 2-1. Conductivity cell apparatus used for obtaining current-voltage
cu rve s............................................................................................. ........ 2 8

Figure 2-2. Chemical structure of polyimide.................................................... 29

Figure 2-3. Current-voltage response through a 1.5 nm conical nanopore pore in
1 M KC I at pH 8.0. ............................................................................ ....... 30








Figure 2-4. Structures of polyethylene glycol) with Mn -200 (PEG200) and
1,4,7,10,13,16-hexaoxacyclooctadecane (18C6).................................... 31

Figure 2-5. Current-voltage response of a 1.5 nm conical nanopore when 18C6
is added in varying concentrations .......................................................... 32

Figure 2-6. Current-voltage response of a 3 nm conical nanopore when 18C6 is
added in varying concentrations............................................................. 34

Figure 2-7. Current-voltage response of a 1.5 nm conical nanopore in 1 M LiCI
e lectro lyte ....................................................................... ........................36

Figure 2-8. Current-voltage response of a 1.5 nm conical nanopore when
PEG200 is added in varying concentrations.............................................. 38

Figure 2-9. Degree of rectification as measured at 500 mV as a function of 18C6
concentration for a 1.5 nm pore............................................................... 40

Figure 3-1. Electrochemical cell set-up for measuring current-time transients... 50

Figure 3-2. 4,4',4",4"'-(porphine-5,10,15,20-tetrayl)tetrakis(benzenesulfonic acid)
(H 2-T P P S ). ...................................................................................... ....... 52

Figure 3-3. Current-time recordings of 1 M KCI pH 8.0................................... 53

Figure 3-4. Current-time recordings showing voltage-dependence of H2-TPPS
translocation events in 1 M KCI pH 8.0 ................................................... 54

Figure 3-5. Histograms showing increase in magnitude of blockade events of H2-
TPPS as the applied potential increases................................................. 56

Figure 3-6. Histograms of dwell time of H2-TPPS in nanopore at (a) 400 mV, (b)
500 mV, and (c) 600 mV. ......................................................................... 58

Figure 3-7. Inverse relationship of translocation time to applied potential.......... 59

Figure 3-8. Current-time recordings showing concentration dependence of H2-
TPPS translocation events in 1 M KCI pH 8.0......................................... 60

Figure 3-9. Schematic definition of terms needed to calculate cross-sectional
area of nanopore in absence of particle. ................................................. 62

Figure 3-10. Schematic definition of terms needed to calculate cross-sectional
area of particle................................................................................. ....... 63

Figure 3-11. Schematic definition of terms needed to calculate cross-sectional
area of nanopore in the presence of particle........................................... 64

Figure 4-1. Structures of lipids used in preparation of vesicles....................... 77








Figure 4-2. Vesicle fusion to substrate............................................................ 79

Figure 4-3. Principle behind FRAP.................................................................. 81

Figure 4-4. Electrochemical cell set-up for bilayer experiments...................... 83

Figure 4-5. AFM images and corresponding line scans of (a) glass under water,
(b) PET under water, and (c) Kapton under water...................................85

Figure 4-6. FRAP images and fluorescence recovery curve of a bilayer
containing 1 mol% PEG2" in egg-PC/Texas Red-PE on glass..............87

Figure 4-7. FRAP images and fluorescence recovery curve of a bilayer
containing egg-PC/Texas Red-PE on glass............................................88

Figure 4-8. FRAP images and fluorescence recovery curve of a bilayer
containing 5 mol% PEG5-PE in egg-PC/Texas Red-PE on PET ........... 89

Figure 4-9. FRAP images and fluorescence recovery curve of a bilayer
containing 5 mol% PEGss0-PE in egg-PC/Texas Red-PE on Kapton.........90

Figure 4-10 Current-voltage curves of (a) single conical nanopore membrane
and (b) 10-conical nanopore membrane in the absence and presence of a
egg-PC/Texas Red-PE bilayer containing 5 mol% PEG5-PE................93

Figure 4-11. One-week stability study of egg PC-PEGSSo-Texas Red-PE bilayer
on K a pton ....................................................................................... ......... 94

Figure 5-1. Representations of (a) alpha-hemolysin with p-cyclodextrin (p-CD)
lodged in lumen and (b) 1-CD ................................ : ................................. 97

Figure 5-2. Vesicle fusion onto conical nanopore membrane ....................... 100

Figure 5-3. Electrochemical cell set-up......................................................... 101

Figure 5-4. Incorporation of a-HL in polymer supported bilayer and sensing B-
C D ........................................................................................................ 102

Figure 5-5. Representative current-time recordings of the (a) supported bilayer,
(b) after the addition of aHL, (c) and after the addition of 3-CD............. 103

Figure 5-6. Current-voltage curves of a single conical nanopore membrane
before and after formation of lipid bilayer .............................................. 104

Figure 5-7. Analysis of 1-CD events in aHL.................................................. 108














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DEVELOPMENT OF BIOMIMETIC SENSING ELEMENTS BASED ON
SYNTHETIC NANOPORES


By

Elizabeth A. Heins

May 2005

Chair: Charles R. Martin
Major Department: Chemistry

The goal of this research has been to develop biomimetic sensing

elements based on single conical nanopores in polymer membranes and to

investigate the properties of such elements as compared to biological nanopores,

particularly the transmembrane protein, alpha-hemolysin (aHL).

In the first part of this work, chemical control of current rectification in

single conical nanopores is explored via the addition of 18-crown-6 ether (18C6)

to the supporting electrolyte. The degree and direction of rectification of the

conical pores are tuned by adding the crown at various concentrations on the tip

or base side of the membrane. The results are interpreted on the basis of steric

and electrostatic interactions of the iK-crown complex with the nanometric

opening of the conical pore.

Stochastic sensing of small molecules in the size regime of medical and

biological interest is studied in the second part of this work. The voltage and








concentration dependencies of a negatively charged porphyrin moving through

single conical nanopores prepared in polymer membranes are demonstrated.

Additionally, a formula has been derived to calculate the amount of expected

current block of a particle in a conical nanopore. This has then been compared

to the experimental data.

The third part of this work is concerned with the development of a stable

supported lipid bilayer on a nanoporous polymer support. The ability to form a

bilayer using vesicles containing lipids with polyethylene glycol moieties was

demonstrated. Characterization of this system by fluorescence and

electrochemical techniques verified the mobility and defect-free nature of the

bilayer, as well as the maintenance of giga-ohm resistance of the system for long

periods of time, suggesting that the system provides a suitable environment for

incorporation of a transmembrane protein.

The final part of this work examines the suitability of the polymer-

supported bilayer system as a sensing platform, through the addition of a pore-

forming protein, alpha-hemolysin, to the system and monitoring the resulting

current-time characteristics. Through the merging of natural and synthetic

technologies, our system may be applied to the development of a real-world

sensing platform for the study and use of transmembrane proteins.












CHAPTER 1
INTRODUCTION AND BACKGROUND

Biological Nanopores

As we strive to develop more technologically advanced devices for

sensing and separation, we find we are inspired by the wonderful designs of

Mother Nature. For example, the lipid bilayer that surrounds cells has many

remarkable systems for selective transport (Figure 1-1). Based on a complex

recognition scheme, molecules are transported across the bilayer by the

appropriate mechanism or excluded from the cell completely. This process is

achieved by a widely studied group of transmembrane proteins that controls the

exchange of ions and small molecules through membranes of biological cells,

contributing to cellular processes such as signaling and transport.1

A key characteristic of biological nanopores is that they have pore

diameters in the size range of biologically relevant molecules and ions, such as

cations (<1 nm diameter), sugars (-1 nm diameter), DNA (~ 2 nm diameter), and

proteins (3-6 nm diameter) to allow the selective passage of appropriate solutes.

Examples of biological nanopores include the KcsA potassium channel,2'3 the

acetylcholine receptor,4 gramicidin,5 the porins of bacteria and mitochondria

(such as maltoporin6'7 and the voltage-dependent ion selective channel (VDAC)8'

9 respectively),10 as well as the bacterial toxin alpha-hemolysin (aHL),11 shown in

Figure 1-2.










0
bkidg aft
UxaphoilMIb"y


..-._ jtTW


oo.Mbon pGinis


Figure 1-1. Lipid bilayer surrounding a cell showing transmembrane proteins.12


Figure 1-2. Examples of biological nanopores shown in ribbon or ball-and-stick
representations: (a) KcsA potassium channel,3 (b) acetylcholine
receptor channel,4 (c) gramicidin channel,5 (d) maltoporin channel,10
(e) CIC channel,8 (f) alpha-hemolysin11 (not comparable scales).








These biological nanopores can be divided into two groups: the small,

highly ion-selective channels of excitable membranes (such as

neurotransmitters) and the large channels which function as conductors of

macromolecules. Many of these channels are known to undergo a change in

conductive properties in response to mechanical stress, a chemical agent, or

change in electrical potential.13

For example, that family of ion channels which responds to a change in

transmembrane potential is said to voltage-gate, allowing ions to pass through

the pore, and hence across the cell membrane, when open. One example of

such a channel is the potassium channel, whose gating properties are shown in

Figure 1-3a and 1-3b. In the open state, voltage-gated channels control ion flow

because they rectify ion currents. This rectification effect for ion channels implies

that ions are conducted in one direction more efficiently than in the other, as

shown in Figure 1-3c. This is observed as a non-ohmic asymmetric current-

voltage characteristic. Ion channels that rectify play a very important role in

setting the resting membrane potential and controlling the excitation threshold for

a cell.13

Biological nanopores can respond in a selective way to a given stimulus.

This sets the stage conceptually for the use of such pores as sensing elements.

Experimental evidence has shown two possible ways to develop sensors from

biological nanopores. The first is through changes in the current-voltage

characteristic resulting from overall changes in the conductance of the biological

nanopore due to the presence of a molecule in the channel.14 The second is









through examination of the current-time recordings, in which the interaction of a

molecule with a nanopore results in either discrete, single events or an increase

in the overall noise of the nanopore. In small, ion-selective channels, it is the

passage of ions that is observed and measured, while in the case of large

channels, it is the passage or blockage of an ionic current that permits the

detection of other molecules in the pore.14

(a) H 1 0 [ Il















(c) 50 pA
40-
30-
20-
10-
,1_______-_____________" mV
-200 -150 -100 50 100 150 200
-10
-20-
-30
-40
--50
Figure 1-3. Example of a voltage-gated potassium channel. (a) Cartoon
showing selectivity of channel. Potassium may pass, but sodium
cannot. (b) Representative current-time recording of potassium
passing through pore.15 (c) Current-voltage curve showing non-ohmic
behavior of the channel.





5


The technology allowing advancements such as the observation of ions

and molecules with biological nanopore was developed in two ways. First, in

1969, Bean et al. demonstrated the ability to reconstitute biological channels in a

planar bilayer system.16 This allowed for the study of one to many biological

nanopores in a controllable, abiotic environment. The second advance was the

development of the patch-clamp technique by Neher and Sakmann in 1976,17

giving researchers the ability to study biological channels in both a whole cell and

a single-channel set-up. These two methods have proven indispensable in the

study of biological nanopores and their development as sensing elements.

In 1970, Hladky and Haydon showed the first example of ion current flow

through a functional biological nanopore,18 using gramicidin incorporated in an

artificial phospholipid bilayer. The fact that biological nanopores would not only

insert into such a bilayer, painted across an orifice in a Teflon sheet, but that they

would also conduct ions across the membrane suggested that larger biological

nanopores might also be used to detect larger molecules in solution.

The literature is rife with examples of biological nanopores through which

the stochastic events of single molecules passing have been recorded. The

interpretation of these events has been based on the Coulter counter

technique.19 This technique, as depicted in Figure 1-4, is based on detection of a

single particle within a nanopore with diameter comparable to the particle being

detected.20 When a particle is passing through a nanopore, the particle will

partially block it. This can be observed as a decrease in ion current. The

duration and amplitude of the blockage are related to the geometry of the








particle. Another consequence of the small pore diameter is that the

translocating particles interact with the pore walls. These interactions, together

with variations in structure of the molecule cause variations in the ion current

signal.




















Time Time
Figure 1-4. Schematic showing the theory of the Coulter counter technique.
When a particle enters the pore, current is temporarily decreased,
resulting in an observable and measurable event.

In 1978, Neher and Steinbach showed the first biological nanopore in

which stochastic events were observed.21 This was the acetylcholine receptor, in

which current-blockages were due to the interaction of a lidocaine derivative with

a single acetylcholine channel. In 2001, Winterhalter and Bezrukov used the

idea of a molecular Coulter counter to study the transport of ATP molecules

through the VDAC nanopore.22 In 2002, they demonstrated these principles in

the transport of maltodextrins through maltoporin.7










vestibule/cavity 3oA



cap

constriction -Ao

transmembrane
Barrel


Figure 1-5. Dimensions of alpha-hemolysin (aHL) channel, a model biological
nanopore.23

Research on the model biological nanopore, aHL, has been key in the

development of biological channels for use as sensing elements.24 Indeed, aHL

is held to be the "gold standard" in biological nanopore-based sensing.

Interestingly, this biological nanopore is not known in nature for a selective ability

to respond to a specific analyte.25 Rather, aHL is known for its ability to allow the

non-selective transport of ions and small molecules out of a cell, leading to cell

death.

aHL is produced by the bacteria Staphlococcus aureus.26 It is a relatively

stable polymeric protein, with monomeric units comprised of 293 amino acid

building blocks each. Seven 33.2 kDa monomeric units assemble into a prepare

complex, which then undergoes heptameric oligomerization. Each of the

monomers contributes a P-stranded hairpin to the assembly of a 14-stranded 03-

barrel pore when inserted in a cell membrane. This pore extends through to the

trans (or intercellular) side of the lipid bilayer. The "cap" of this mushroom-

shaped protein remains on the cis (or extracellular) side of the bilayer.26 As








confirmed by the crystal structure, following oligomerization, the overall

dimensions of the protein, shown in Figure 1-5, are approximately 10 nm by 10

nm with a nanopore inner diameter of -1.4 nm.11

Bayley and co-workers have performed extensive work on the structural

modification of the wild-type aHL.2729 Modification has been achieved through

site-directed mutagenesis with recombinant DNA technology or targeted

chemical modification. By replacing amino acids in the protein with other

naturally occurring or synthesized amino acids, Bayley has shown that it is

possible not only to control what passes through the pore, but also that it is

possible to open and close the pore on command.

(a) (b)
Time. s
0 1 2 3 4 5
J, 0


-20



NOPEG PEG200 PEG2000 PEG8000 -1000 mM

Figure 1-6. Current-time recordings demonstrating effect on conductance of aHL
nanopore in absence and presence of (a) poly(ethylene) glycol (PEG)
molecules3 and (b) crown ethers.31 At increasing concentrations, an
increase in particle-induced noise in the signal is observed.

The studies with aHL have shown it is possible to detect the presence of

an ion or small molecule either by looking at the overall conductance changes of

the pore or through examination of the time series. For example, in 1996,

Bezrukov and coworkers showed measurable effects on the ionic current passing








through the aHL pore in the presence of poly(ethylene glycol) (PEG) molecules.30

The changes in the overall conductance, seen in Figure 1-6a, have been used in

the analysis of the size of the channel as well as the size of the PEG able to

translocate the pore. Recently (2004), they have also shown conductance

changes based on the presence of a cyclic polyether, 18-crown-6 (Figure 1-6b),

in the aHL pore.31 These changes have been attributed to both the sterically

hindered crown in the nanopore as well as the effect of the electric field on the

crown when it is completed with K*.

If a translocating molecule interacts very strongly with the nanopore or

takes a relatively long time to pass through the pore, the current interruption is

strong enough to produce a discrete blockage of ionic current called a blockade.

By examining the time series of such blockades, it is possible to determine size

and concentration information about the molecule. In 1996, Kasianowicz et al.

showed it possible to drive single-stranded RNA or DNA through an aHL

nanopore electrophoretically, with the length of block being proportional to the

polymer length.32 Following this, a series of reports showed that it was possible

not only to determine the length of a DNA or RNA, but that it was also possible to

distinguish between the signals produced by different backbones.33 Additionally,

it was shown possible to resolve, at single base resolution, DNA hairpins of less

than 10 bases3 and also to detect the passage of a complementary strand of

DNA to one covalently bound to the aHL.3








(a) (b) (c)





A. .. -









1 5 lst 10 S

Figure 1-7. Examples of engineered aHL nanopores detecting different types of
analytes. (a) Addition of histadine in the barrel allows for detection of
divalent metal ions. (b) Modification of the restriction in the pore allows
for binding of non-covalent adapters which bind adamantine
derivatives. (c) Covalent attaent attachment of a ligated-polymer in the pore
allows for "capture" of targeted proteins.24

Bayley and coworkers have also shown that it is possible to detect single

small organic molecules through the use of non-covalent modification by taking

advantage of host-guest chemistry.24 For example, P3-cyclodextrin, known to bind

small organic molecules in its hydrophobic interior, is used as a molecular

adapter in the detection of a range of adamantane derivatives. 39 Indeed, it has

been demonstrated that different adamantanes produced characteristic

signatures with distinctive current block and dwell times. Additionally, they have

shown that it is possible to engineer these nanopores in various ways, allowing

for detection of a wide range of molecules and ions (Figure 1-7).4'








Synthetic Nanopores as Biomimetic Sensing Elements

In spite of the significant advances made in developing sensing elements

based on biological nanopores, the practical application of these elements is

tenuous at best, largely because of the fragility of the planar bilayer and the

mobility of the protein within the bilayer.23 In order to fully develop a sensing

element based on the properties of biological nanopores, a synthetic nanopore

system that allows for flexibility in the dimension, composition, and stability of the

pore is necessary. A major hurdle has been matching the extremely small

diameter (1-2 nm) of biological nanopores with such a support.

Advances in nanotechnology have allowed for the development of

synthetic membranes containing single nanopores of dimensions comparable to

biological channels. Multiple approaches in various materials have been utilized

to prepare such nanopore systems and are both described briefly below and

shown in Figure 1-8. These synthetic membranes have shown stochastic

sensing capabilities similar to those demonstrated by biological systems with

larger molecules such as DNA.

One of the synthetic nanopore approaches is the method developed by

Branton and Golovchenko, in which single pores have been prepared in silicon

nitride using focused ion beam etching.45 These nanopores, with lengths and

diameters of <10 nm have been used to measured the length and folding

characteristics of double stranded DNA.4 A second method is based on

lithography. Saleh and Sohn have demonstrated that nanopores prepared in

poly(dimethyl siloxane) (PDMS), using standard lithographic techniques, can be

integrated with microfluidic systems.47 Using such nanopores, they have shown









(a) (b)














(c) (d)
5 pm Si6 N4 membrane




epoxy section MWNT


q..







Figure 1-8. Examples of synthetic nanopores with dimensions analogous to
biological nanopores (scales bars are shown for each image): (a)
nanopore prepared by focused ion beam-etching in silicon nitride', (b)
lithographically-prepared nanochannel in PDMS,48 (c) single multi-
walled carbon nanotube,49 (d) gold replica of single conical nanotube
chemically etched in a polymer membrane.50

that it is possible to detect single A-phage DNA molecules.47 Crooks and

coworkers have shown a strategy for preparing single carbon nanotube








membranes with diameters that can reach nanometer scale.51 They have been

able to demonstrate the Coulter counter principles using DNA.i Interestingly,

the uncharged walls of these nanopores allow determination of the

electrophoretic mobility of molecules, something not achievable with biological

nanopores. A fourth approach to the development of biomimetic synthetic

nanopores involves the use of a chemical etching process to create single

conical nanopores in polymer membranes.52' 53

The use of polymeric membranes in the detection of single particles is not

new. Indeed, the original Coulter counter, patented in 1957, was based on a

pore in a polymer membrane.19 It was many years however, before it was

possible to etch a pore small enough to detect particles <100 nm in size. The

polymeric nanopores prepared by Siwy, Martin, and others have a small-opening

diameter that can be as small as 2 nm, and can be used to sense a variety of

molecules.53'56

Using single conical nanopore membranes, the Martin group has explored

the stochastic sensing properties of these synthetic nanopores, as shown in

Figure 1-9. For example, Martin and coworkers have created a nanodevice to

stochastically sense the electrophoretic movement of DNA molecules and to

discriminate between two different types of DNA.57 Additionally, the Martin group

has shown that it possible to control the voltage-gating properties of such

nanopores through the surface modification of the pore walls using Au-thiol

chemistry.58






















Plasmid DNA
"bumping"


ssDNA translocation


0.1 M KF, pH 6.1
0.1 M KF, DH2.8


SVoltage (mV)
500 1000


-2ur (nA)
Current (nA)


A Modified Membrane in
AA 0.1 M KF, pH 6.1


* ~ I.,~1A


500 1000





+ -


Figure 1-9. Examples of sensing-capabilities of single conical nanopore
membranes. (a) Representative current-time recording showing
different events for ssDNA and plasmid DNA translocations through a
conical polymeric nanopore.57 (b) Representative current-voltage
curves demonstrating control over rectification by chemically modifying
a conical polymeric nanopore.'


-1 A A g4 A A
-0.5.


gg Current (nA)


4 & 'G00'


+






Mercaptoothanalam~ine
I .


+QW
IM


60 pA [
I400 ms








Irradiation of Polymer Membranes

In order to prepare single conical nanopores in polymer membranes for

both the applications described above and the ones in this work, the so-called

track-etch technique was used. It is based on the irradiation of a dielectric film

with swift heavy ions59 and the subsequent chemical development (etching) of

the damaged ion tracks.60 A unique feature of heavy ion irradiation is the

formation of controllable number of damage tracks based on a controllable

number of heavy ions passing through the dielectric film.61 That is to say, each

swift heavy ion which penetrates the foil produces one damage track (Figure 1-

10). Therefore, controlling the number of ions used for irradiation enables one to

prepare membranes with a tailored number of pores ranging from 1 up to 1010

pores/cm2.


Reactive
Polymer chains chain end

Swift heavy ion -",. Chain scission
e.g. Au, Pb, U
Latent track

Figure 1-10. Principle behind the creation of latent tracks in polymeric
membranes via irradiation with swift heavy ions.

Polymer foils, used for these experiments, were irradiated with single swift

heavy ions (e.g. Au, Xe, U ) with 2.2 GeV kinetic energy at the heavy ion

accelerator UNILAC at the Institute for Heavy Ions Research (GSI) in Darmstadt,

Germany. Films irradiated with a single ion were used, leading to a single latent

track in a polymer film. The ions penetrate the membranes at normal incidence,

creating damaged tracks. The penetration depth, in all cases, is greater than the








thickness of the polymer membranes. This was followed by chemical etching to

form pores.

Etching of Single Conical Nanopores in Polymer Membranes

Membrane samples that contained a single conically shaped nanopore

were prepared by anisotropic chemical etching of the heavy-ion irradiated

polymer films, as shown in Figure 1-11. Per the procedure developed by Spohr

et al.,62 the irradiated film was placed between the two halves of a conductivity

cell, with an etch solution added to one half-cell and a stop solution added to the

other half-cell. Each half-cell contained a Pt wire, and a transmembrane

potential of 1 V was applied during etching. The progress of the etching process

was monitored by measuring the transmembrane current.53 Etching was halted

when the desired transmembrane current was measured. The membrane was

then removed from the conductivity cell and placed in water overnight.

The large-diameter opening of the conical nanopore will be referred to as

the "base" of the nanopore and the small-opening as the "tip" of the nanopore.

The nanopores studied here had a base diameter, db, of -2 pm, as measured

using scanning electron microscopy. The tip diameter, d,, was determined

electrochemically videe infra). The etch process also thinned the membrane from

12 to 10 pm.

It has been shown previously that it is possible to etch pores of different

geometries by controlling the etching conditions.63' 4 In addition to cones, such

geometries include cylinders, double cones, and elliptically shaped pores. These

various geometries show potential in the use of the further tailoring of polymer

nanopore membranes in transport studies and particle sizing. The conically-








shaped nanopore was chosen for the work of this dissertation for two reasons.

The first is to mimic the asymmetry in shape that is prevalent in biological

nanopores. 13' The second is that the conical shape leads to a lower nanopore

resistance (relative to the cylindrical geometry), which leads to an increase of

current signal.65


Pt Pt
Reference Working
Electrode Electrode




Stop Etchant


Figure 1-11. Electrochemical set-up for chemically etching conical nanopores in
polymeric membranes.

Sizing of a Conical Nanopore

To determine the diameter of the tip-opening of the conical nanopore, a

polymer membrane containing a single conical nanopore membrane was

mounted in a conductivity cell, and both half-cells were filled with the desired

electrolyte solution. A Ag/AgCI electrode was inserted into each half-cell and the

Keithley picoammeter/voltage source was used to obtain current-voltage curves

associated with ionic migration through the nanopore. As shown in Figure 1-12,

the working Ag/AgCI electrode was in the half-cell solution facing the base of the

nanopore and the reference Ag/AgCI electrode was in the half-cell solution facing

the nanopore tip. The current-voltage curve was obtained by stepping the

voltage in from -1000 mV to +1000 mV at a rate of 5 seconds per step. The sign

convention is as follows: a positive transmembrane potential means that the








working electrode is the anode and a negative transmembrane potential means

that the working electrode is the cathode.

The ionic resistance of an electrolyte-filled conically shaped pore is given

bye

R -4pL Equation (1.1)


where p is the specific resistance of the electrolyte, and L is the length of the

pore, in this case 10 pm. R was determined from the experimental current-

voltage curves and used to calculate dt. However, as will be discussed below,

the current-voltage curves for these conical nanopores are nonlinear. For the

reasons discussed by Apel et al. we use the currents obtained between 200 mV

to obtain R.52 The electrolyte specific resistance was measured using a YSI

conductivity meter (Yellow Springs, OH).

Calculating the Effective Pore Length of a Conical Nanopore

The effective length of a conical nanopore is defined as the length over

which 95% of the resistance is focused.66 It is over this length that the voltage

drop is the greatest, resulting in the strongest electric field. To determine the

effective length, distribution of resistance along the pore axis, from the tip of the

nanopore towards the large pore opening is analyzed. This analysis is

performed by calculating the resistance of cones, R,, of length x e <0,L>, big

diameter dx e <0,dr>, and small opening d.66 dx can be expressed as the

following function of x:

dx = (db-d)x+dL Equation (1.2)








making the resistance Rx:

R = 4p2 + (d-d+Ld Equation (1.3)
x n(d d) dL (d -:d,)x +Ld,

dx and R, are calculated for various values of x. By dividing Rx by the total

resistance of the cone of length L, and opening diameters db and dt (Equation

1.3), one can follow how resistance of a conical pore changes along its length.

Rectifying Properties of a Conical Nanopore

It has been shown that the negatively charged surface of the polymer

nanopore results in the formation of an electric field inside the pore. 5354 The

electrostatic potential inside the conical nanopore with excess surface charge

has been shown to be asymmetric.53'58, 65 At the tip of the conical nanopore, the

electrostatic potential within the nanopore is at its most negative. The

electrostatic potential gradually becomes more positive towards the large-

diameter opening. When no transmembrane potential is being applied, no

transmembrane current flows through the conical nanopore, and the electrostatic

potential (0) felt by a cation along the nanopore axis is schematically shown in

Figure 1-12a.

The left side of the figure represents the electrolyte solution in contact with

the tip of the nanopore. The right side of the figure represents the region of the

pore where the nanopore radius is much greater than the double-layer thickness

and the cation does not "feel" the effects of the internal field. Between these two

areas is the region where the radius of the nanotube is comparable to the

double-layer thickness. In this region, the potential of the cation is lowered due

to electrostatic interactions with the fixed anionic surface charge. As the








nanotube diameter increases, moving in the x-direction, the electrostatic

interactions fall off. This accounts for the upward slope of the electrostatic

potential profile.



(a) Bulk
Solution


No effect ore reg where No effect
felt by I ionic current is I felt by
cation affected by internal cation
0 electrostatic
potential



Distance, I
(b)
At negative applied voltage:




LNcaNo Cation Trap

At positive applied voltage:



L^i + j.ation Trap

External Electric Field Electrostatic Potential Supposition of two
Inside Nanopore Fields
Figure 1-12. Electrostatic potential inside conical nanopore when there is no
applied potential. (a) At negative applied potentials, the trap tilts in
such a manner that there is no electrostatic trap. (b) At positive
applied potentials, the trap tilts in such a manner that a trap for cations
exists.

We assume the superposition of the internal electric field with the external

applied voltage. When a negative potential is applied across the membrane, the

voltage drop caused by the resulting negative current effectively rotates

approximately 1I8t of a turn in the clockwise direction, resulting in no








electrostatic trap, as shown in Figure 1-12b. However, when a positive potential

is applied across the membrane, the voltage drop caused by the resulting

positive current effectively rotates approximately 1/8th of a turn in the

counterclockwise direction (Figure 1-12b). This effectively creates an

electrostatic trap, which in turn causes the reduction, or rectification of the

transmembrane current.

The Ratchet Model

A ratchet is a type of Brownian motor, which has the ability to produce

directed motion in the absence of gravitational force, macroscopic electric fields,

or long-range spatial gradients of chemicals if the system is driven out of

equilibrium.67 This results from Brownian movement of particles in synergy with

an external field, such as a thermal gradient, a non-equilibrium chemical reaction,

or a fluctuating external field.6 In devices based on biased Brownian motion, net

transport occurs by a combination of diffusion and directed motion resulting from

an externally applied time-dependent field. The potential map of such a device is

typically visualized as a series of asymmetric teeth, or a ratchet.

A single conical nanopore, with an asymmetric electrostatic potential map,

can be treated as a single ratchet tooth.5 When subjected to an external electric

field, non-equilibrium perturbations are introduced to the system. Depending on

the magnitude and direction of the external field, the ratchet can be said to rock

in a clockwise or counterclockwise direction resulting in a "tilting" ratchet.

Dissertation Overview

The goal of this research has been to develop biomimetic sensing

elements based on single conical nanopores in polymer membranes and to








investigate the properties of such elements as compared to biological nanopores,

particularly aHL. The previous part of chapter 1 has reviewed the background

information for this dissertation, including biological and synthetic nanopore

sensing strategies, the track-etch process, and the sizing of conical nanopores.

In chapter 2, chemical control of current rectification in single conical

nanopores is explored. Via the addition of 18-crown-6 ether (18C6) to the

supporting electrolyte, tuning the degree of rectification of the nanopore has been

demonstrated. Conical pores prepared in a polymer membrane are shown to

respond to the presence of the crown in a similar way as has been shown for the

aHL pore. The degree and direction of rectification of the conical pores are tuned

by adding the crown at various concentrations on the tip or base side of the

membrane. Further, it is shown that the crown has an effect on the transport

properties of the pore only if it can complex with cations of the primary

electrolyte. The results are interpreted on the basis of steric interactions of the

KI-crown complex with the nanometric opening of the conical pore.

Stochastic sensing provides a means to study transport properties at the

single molecule level, an important issue for many applications, such as drug

delivery. In chapter 3, stochastic sensing of small molecules is demonstrated.

The voltage- and concentration-dependent movements of the negatively charged

porphyrin, 4,4',4",4'"-(porphine-5,10,15,20-tetrayl)tetrakis(benzenesulfonic acid)

(H2-TPPS), through single conical nanopores prepared in polymer membranes

are studied. Additionally, a formula for calculating the expected current block of

a particle in a conical nanopore has been developed and compared it to the








experimental data. From this, it is shown that the observed current response can

be interpreted using a modified Coulter counter model.

Chapter 4 is concerned with the development of a stable supported lipid

bilayer on a nanoporous polymer support as the study and use of

transmembrane proteins in biosensing applications requires the development of

a rugged system that preserves the biological activity of the protein. The ability

to tailor the size and surface chemistry of polymer nanopores makes this support

an attractive option as a sensing platform. The polymer support was prepared by

the track-etch method described in chapter 1. Phosphatidylcholine vesicles,

containing poly(ethylene glycol) terminated phosphoethanolamine lipids, were

prepared by the extrusion method, and the bilayer formed via vesicle fusion.

Characterization of this system by fluorescence and electrochemical techniques

verified the mobility and defect-free nature of the bilayer, as well as the

maintenance of giga-ohm resistance of the system for long periods of time. This

suggests that the system provides a suitable environment for incorporation of a

transmembrane protein.

In chapter 5, the suitability of the polymer-supported bilayer system was

then examined by the addition of a pore-forming protein, alpha-hemolysin, to the

system and monitoring the resultant current-time characteristics. The described

system has a stability not previously demonstrated in biological nanopore-based

sensing systems. The platform has been developed by taking advantage of the

sensing capabilities of aHL and the stability and ruggedness of a polymeric

supported bilayer system. Through the merging of natural and synthetic





24


technologies, this system may be applied to the development of a real-world

sensing platform for the study and use of transmembrane proteins.

Chapter 6 provides some conclusions and further directions for the work

presented in this dissertation.












CHAPTER 2
CHEMICAL CONTROL OF CURRENT RECTIFICATION IN A SYNTHETIC
NANOPORE WITH CROWN ETHER

Introduction

Biological nanopores and channels consist of proteins that span the

phospholipid membranes surrounding cells. Nature has evolved a diverse

collection of such nanopores that perform critical tasks in cellular processes, like

signaling and transport.1 These biological nanopores respond to a variety of

stimuli, including mechanical stress, chemical interaction, and electrical

potential.13 Mimicking features of biological systems with a synthetic nanopore

could have far-reaching implications in biochemical sensing and separation. In

this chapter, it is shown that it is possible to control of the electrophoretic

migration of ions through membranes containing a single conical nanopore via

the addition of an ion-complexing cyclic polyether, 18-crown-6 (18C6).

Recently, the Martin group and others have reported conical nanopores

prepared in polymer membranes that display electrical signaling and selective

transport effects similar to those observed in biological nanopores. 5358 69 In

asymmetric conical nanopores that carry an excess negative surface charge,

transmembrane currents are rectified with a preferential direction of cation flow

from the tip of the conical nanopore to the base of the conical nanopore.52, 53

Further control of the rectification of the synthetic conical nanopores has been

demonstrated by introducing a variable surface charge to the pore walls.5 DNA-








modified conical nanopores that rectify current based on an electromechanical

gate incorporated at the pore surface have also been demonstrated.6

Here it is shown that the addition of 18C6 to an electrolyte solution affords

control over the direction and magnitude of ionic currents through a conical

nanopore, provided the small-diameter of the pore is of molecular dimensions.

The extent of rectification of the conical nanopores is governed by the

concentration of 18C6, and the direction of rectification is governed by the

placement of 18C6 (with respect to the conical pore geometry). It is further

shown that 18C6 has an effect on the transport properties of the pore only if it

can complex with cations of the primary electrolyte. The results are interpreted

on the basis of the occlusion of the conical nanopore and electrostatic

interactions of the KI-18C6 complex with the nanometric opening of the pore.

Experimental

Materials

Polyimide films ((Kapton-50HN, DuPont, 3 cm diameter, 12 pm thick), that

had been irradiated with a single swift heavy ion of 2.2 GeV kinetic energy to

create a single damage track through the film were obtained from GSI

Darmstadt, Germany. Potassium chloride (certified A.C.S., Fisher Scientific),

1,4,7,10,13,16-hexaoxacyclooctadecane (18C6) (99%, Sigma); poly(ethylene

glycol) (PEG") (Mn = 200); sodium hypochlorite (13% active chloride ion,

Sigma), potassium iodide (certified A.C.S., Fisher Scientific), and lithium chloride

(99%, Acros) were used as received. All solutions were prepared in 18 mega-

ohm water (Barnstead, E-pure).








Preparation of Conical Nanopores

Membrane samples that contained a single conically shaped nanopore

were prepared by the anisotropic chemical etching of the heavy-ion irradiated

Kapton films. As per the procedure developed by Spohr et al.,62 the irradiated

film was placed between the two halves of a conductivity cell, with an etch

solution of sodium hypochlorite added to one half-cell and a stop solution of 1 M

potassium iodide added to the other half-cell. Each half-cell contained a Pt wire,

and a transmembrane potential difference of 1 V was applied during etching

using a Keithley 6487 picoammeter/voltage source (Keithley Instruments,

Cleveland, OH). During etching, the solutions in the conductivity cell were kept

at 50 C. The progress of the etching process can be monitored by measuring

the transmembrane current during etching. Etching was halted when a

transmembrane current of 50 pA was measured. The membrane was then

removed from the conductivity cell and placed in water overnight. We will refer

here to the large-diameter opening of the conical nanopore as the "base" of the

nanopore and the small-opening as the "tip" of the nanotubes. All of the

nanopores studied here had a base diameter, db, of 2 pm, as measured using

scanning electron microscopy.70 The tip diameter, dr, was determined

electrochemically videe infra). The etch process also thinned the membrane from

12 pm to 10 pm.

Estimation of the Tip Diameter, dt

The ionic resistance of an electrolyte-filled conically shaped pore is given


by5








R= 4pL Equation (2.1)


where p is the specific resistance of the electrolyte, and L is the length of the

pore, in this case 10 pm. R was determined from the experimental current-

voltage curves and used to calculate d,. For the reasons discussed by Apel et al.

we use the currents obtained between 200 mV to obtain R.52 The electrolyte

specific resistance was measured using a YSI conductivity meter (Yellow

Springs, OH).

Electrochemical Measurements


Ag/AgCI V Ag/AgCI
Reference Working
Electrode Electrode



Electrolyte. \Electrolyte



Small Opening (dt) Large Opening (db)
Figure 2-1. Conductivity cell apparatus used for obtaining current-voltage
curves.

The single conical nanopore membrane was mounted in a conductivity

cell, and both half-cells were filled with the desired electrolyte solution. A

Ag/AgCI electrode was inserted into each half-cell and the Keithley

picoammeter/voltage source was used to obtain current-voltage curves

associated with ionic migration through the nanopore. As shown in Figure 2-1,

the working Ag/AgCI electrode was in the half-cell solution facing the base of the








nanopore and the reference Ag/AgCl electrode was in the half-cell solution facing

the nanopore tip. The current-voltage curve was obtained by stepping the

voltage in 50 mV increments from -500 mV to +500 mV at a rate of 10 s per step.

The sign convention is as follows: a positive transmembrane potential means that

the working electrode is the anode, and a negative transmembrane potential

means that the working electrode is the cathode. The current-voltage curves

reported here are averages of three sequential measurements made on the

same nanopore.

Results and Discussion

Chemistry and Charge of the Polyimide Nanopore

The chemical structure of the polyimide used here is shown in Figure 2-2.

When etched with sodium hypochlorite and then exposed to solutions with a pH

of 8, as used here, the surface of this initially electrical-neutral polymer becomes

negatively charged.5,'70 This negative surface charge can be removed by

exposure to solutions by lowering the pH to 3,6 suggesting that, in accord with

the chemical structure, the negative surface charge is due to deprotonated

carboxylate groups. The maximum surface charge, has been determined to be

approximately 1.5 negative charges per nm2.71






0 0
Figure 2-2. Chemical structure of polyimide.
Figure 2-2. Chemical structure of polyimide.








As a result of the surface charge, the nanopores are cation

permselective,53,72 provided the tip diameter is small relative to the Debye

screening length of the electrolyte solution used. This means that current is

carried through the nanopore predominately by cations, which, as we will see,

simplifies the interpretation of the obtained current-voltage curves. Furthermore,

such cation-permselective conical nanopores show non-linear (rectifying) current-

voltage curves; i.e., for any absolute value of applied transmembrane potential,

the current is higher at negative potentials than at the equivalent positive

potentials (Figure 2-3). As discussed in detail previously, this is because cations

migrating from base to tip experience an electrostatic trap not observed when

cations are migrating from tip to base.58

3














1 M KCI at pH 8.0.




to characterize the pores.30, 73, 74 One of the most successful approaches is
based on the use of polyethylene glyols) (PEGs) (Figure 2-4a). The
-600 -250 250 00
.0 -1N



-3
Potential (mV)
Figure 2-3. Current-voltage response through a 1.5 nm conical nanopore pore in
1 M KCI at pH 8.0.

Effect of 18C6 on Current-Voltage Curves with KCI as the Electrolyte

In the case of biological nanopores, it is common to use soluble polymers

to characterize the pores. 30,7374 One of the most successful approaches is

based on the use of poly(ethylene glycols) (PEGs) (Figure 2-4a). The








interactions of PEGs with biological nanopores such as aHL have been studied

using current-noise analysis.41'74 PEG molecules of size comparable to the pore

diameter produced increased noise due to increased chemical interaction with

the channel walls of aHL.

(a) (b)




HO


Figure 2-4. Structures of molecules investigated: (a) poly(ethylene glycol) with
Mn -200 (PEG2) and (b) 1,4,7,10,13,16-hexaoxacyclooctadecane
(18C6).

Another polyether that has been used to study aHL is the cyclic

polyether,31 1,4,7,10,13,16-hexaoxacyclooctadecane, commonly called 18-

crown-6 (18C6),75 shown in Figure 2-4b. 18C6 is an uncharged molecule with an

outer diameter of approximately 1.15 nm.76 Alone, it might not be expected to

have an effect on the current-voltage properties of the nanopore. However,

because of the available lone pairs on the ether oxygens, crown ethers complex

cations. Because the crown is non-charged, the complex retains the positive

charge of the completed cation. The potassium cation is completed by 18C6

with a formation constant of approximately 100 I/mol in aqueous solution.77 In

contrast, the formation constant between Li+ and 18C6 is effectively zero.78

The effect of 18C6 on the current-voltage response of a conical nanopore

with a tip diameter of 1.5 nm is shown in Figure 2-5. The electrolyte placed on

both sides of the membrane is 1 M KCI, pH 8.0. Figure 2-5a shows the effect of








(a) 4
(+) (-)






0
250 500





-4"
Potential (mV)
(b) 4


I 2




-500 -250 250 500

-2 ,P .



-4
Potential (mV)
Figure 2-5. Current-voltage response of a 1.5 nm conical nanopore when 18C6
is added in varying concentrations to the (a) tip or (b) base side of the
membrane. Each concentration is represented by the following
symbols: -- 1 M KCI, -0- 5 mM 18C6, -6- 10 mM 18C6, -0- 25 mM
18C6, -01- 50 mM 18C6. Each data point represents the average of
three replicate measurements with a standard deviations of <2%.








adding the 18C6 at concentrations varying from 5 mM to 50 mM to the electrolyte

solution facing the tip of the nanopore. At negative applied transmembrane

potentials both the excess uncomplexed K and the charged KW-18C6 complex

are driven electrophoretically from tip to base. While the diameter of the

hydrated potassium (-0.3 nm) is much smaller than the tip diameter, the

diameter of the complex (1.15 nm) is comparable to d,. For this reason, when

electrophoretically driven into the nanopore, the complex occludes the pore tip,

and this is responsible for the decrease in current observed at negative

potentials. Analogous results were obtained with aHL31 because, like the

nanopore used here, the lumen of aHL has a constriction with diameter of ~1.4

nm.11

At positive applied transmembrane potentials the K-1 8C6 complex is

electrophoretically driven away from the nanopore, and charge is carried through

the nanopore by free K* from the solution facing the base of the nanopore. For

this reason, 18C6 has no effect on the transmembrane current at positive

potentials (Figure 2-5a). The opposite trends are observed when 18C6 is added

to the electrolyte solution facing the base of the nanopore (Figure 2-5b). At

positive potentials, the complex is electrophoretically driven into the nanopore

resulting in occlusion and decreased currents. At negative potentials the

complex is driven away from the nanopore, and thus has no effect on the

measured current.

To confirm that this size-based occlusion model is correct, analogous

experiments were conducted on nanopores with larger tip diameters. Figure 2-6





34


(a) 4

2F(+) (-)







S-500 -250 0 250 500






-44
Potential (mV)
(b) 4-
S(+) (-)




I pal 0 2-


S-500 -250 250 500
0 -21




-4
Potential (mV)

Figure 2-6. Current-voltage response of a 3 nm conical nanopore when 18C6 is
added in varying concentrations to the (a) tip or (b) base side of the
membrane. Each concentration is represented by the following
symbols: -*- 1 M KCI, -0-5 mM 18C6, -6- 10 mM 18C6, -0- 25 mM
18C6, -0- 50 mM 18C6.








shows results for a nanopore with dt = 3 nm. Because dt is now significantly

larger than the diameter of the complex, addition of 18C6 does not have a

detectable effect on the current-voltage curve regardless of the polarity of the

transmembrane potential, the electrolyte solution to which the 18C6 is added,

and the 18C6 concentration. Similar results were obtained for nanopores with dt

of 9 and 15 nm. These studies also discount an alternative possibility-that the

small amount of completed K* produces a large decrease in the conductivity of

the 1 M KCI solution. If this were the case, 18C6 would decrease the observed

current regardless of the tip diameter of the nanopore, and this is not what is

observed experimentally.

Effect of 18C6 on Current-Voltage Curves with LiCI as the Electrolyte

The electrophoretic transport of the large KI-18C6 complex into the

nanopore is most likely responsible for the observed effects of added 18C6 on

the current-voltage curves (Figure 2-5). It seemed possible; however, that simple

diffusion-based transport of the 18C6 into the nanopore might be sufficient to

produce the effects observed here. To discount this possibility, analogous

experiments were performed using LiCI as the electrolyte. Because Li is not

completed by 18C6,78 the crown is not electrophoretically driven into the

nanopore; it is however, free to diffuse into the nanopore. The results of these

experiments with LiCI using a nanopore with d, = 1.5 nm and the highest (50 mM)

18C6 concentration are shown in Figure 2-7. Addition of 18C6 has no effect on

the current-voltage curves regardless of which side of the membrane it is added.








(a) 1
(+) ()

0.5



0


-0.5


-1
Potential (mrnV)
(b)
S^ r).1 (++)

0.5




-500 -250 0 250 500

IN
-0.5

-1 "

Potential (mV)

Figure 2-7. Current-voltage response of a 1.5 nm conical nanopore in I M LiCI
electrolyte. The Li* does not form a complex when (a) 50 mM 18C6 or
(b) 50 mM PEG" is added to the tip or base side of the membrane.
Each current-voltage curve is represented by the following symbols:
1 M LiCI, -6- 50 mM 18C6 added to the cis side or -- 50 mM
analyte added to the trans side of the membrane.








This experiment also eliminated the possibility that 18C6 has some special

chemical affinity with the nanopore, and that it is this chemical effect that is

responsible for the results in Figure 2-5. Additionally, the idea was considered

that the effects in Figure 2-5 were due to transport of the K*-18C6 complex into

the nanopore via electroosmotic flow (EOF) as opposed to the electrophoretic

mechanism proposed here. Since the nanopore has negative surface charge,

EOF-based transport is in the same direction as electrophoretic transport.

However, EOF does not require that the species being transported is charged.

Hence, if the dominant transport mechanism driving the 18C6 into the nanopore

was EOF, the results in KCI and LiCI would be the same, and this is not what is

observed experimentally.

Effect of PEG20m on Current-Voltage Curves

To further confirm that electrophoretic transport of the iK-18C6 complex

into the nanopore is responsible for the changes in the current-voltage curves

observed, analogous experiments were conducted with a linear poly(ethylene

glycol), PEG". The molecular weight of PEG2 (ca. 200 Da) is similar to that of

18C6 (264 Da). PEG" is also similar to 18C6 in that each both contain ether

functionalities. There are approximately 5 ether functionalities per PEG2

molecule. However, PEG2 and 18C6 are different in that PEG2 is a linear

molecule and 18C6 is cyclic. Since PEG" is linear, it does not form stable

complexes with metal ions and remains uncharged in a solution of KCI.79

Therefore, as per the LiCI experiments with 18C6, PEG2 is not

electrophoretically driven into the nanopore. In analogy to the LiCI experiments,








(a)






C
'S r
| -500






(b)







C-


-4P '
Potential (mV)


500 -250





-4-
Potential (mV)


250



P~J
b7V


Figure 2-8. Current-voltage response of a 1.5 nm conical nanopore when
PEG2 is added in varying concentrations to the (a) tip or (b) base side
of the membrane. Each concentration is represented by the following
symbols: -0- 1 M KCI, -0- 5 mM PEG200, -6- 10 mM PEG200, -0- 25
mM PEG20, -0- 50 mM PEG200.


250


500


0
8,


500


-250 0

2-2


. -92








addition of PEG200 has essentially no effect on the current-voltage curves (Figure

2-8).

The Effect of 18C6 on the Extent of Rectification

As noted earlier, due to the excess negative charge on the pore wall,

these conical nanopores are cation permselective provided the tip diameter is

small relative to the Debye screening length of the electrolyte solution

employed.5 This cation permselectivity results from the formation of an

electrostatic trap, as reported previously.5 The effects of the electrostatic trap

can be observed in the non-linear (rectifying) current-voltage curve for 1 M KCI in

the absence of added 18C6 (Figure 2-5a). At positive applied transmembrane

potentials K* is electrophoretically transported through the conical nanopore from

the large-diameter opening towards the small-diameter opening. As KW

translocates the conical nanopore, it experiences an electrostatic trap that

hinders translocation, as described previously. Conversely, K*

electrophoretically transported through the conical nanopore from the small-

diameter opening to the large-diameter opening is not hindered by the

electrostatic trap.

A rectifying current-voltage response for the conical nanopore is

measured as a result of this asymmetric hindrance. Thus, a higher magnitude

current is observed at a given negative transmembrane potential relative to the

corresponding positive transmembrane potential. The extent of rectification

observed can be quantified via the ratio, r:


r = Equation (2.2)
'500








where isoo is the measured current at an applied transmembrane potential of -

500 mV and i+soo is the current at +500 mV.

20







o 15



05




10 20 30 40 50
Concentration of 18C6 added to I M KCI (mM)
Figure 2-9. Degree of rectification as measured at 500 mV as a function of 18C6
concentration for a 1.5 nm pore. 18C6 is present at the indicated
concentrations on either the tip (-6-) or base (,--) side of the
membrane. 1 M KCI is represented by the straight line (-).

The steric hindrance of the K-1 8C6 complex affords a method to tune the

overall rectifying properties of the pore. When added to only one side of the

membrane, as described here, 18C6 affects only one branch (positive currents or

negative currents) of the current-voltage curve, producing rectification, albeit

through a different mechanism. Rectification due to 18C6 is steric in origin and

can add to or subtract from the inherent electrostatic rectification of the conical

nanopore, depending on which side of the conical nanopore 18C6 is added. This

is shown in Figure 2-9 as plots of degree of rectification vs. concentration of

added 18C6 from the current-voltage curves in figure 2-5. In the absence of








18C6, the extent of rectification (due the intrinsic electrostatic rectification of the

conical nanopore) is 3.1.

When 18C6 is added to the side of the membrane facing the large-

diameter opening, steric rectification induced by 18C6 adds to the electrostatic

rectification of the conical nanopore. Steric rectification increases when the

concentration of 18C6 is increased, as observed in Figure 2-5b. The extent of

rectification for 18C6 added to the side of the membrane facing the large-

diameter opening is plotted as a function of 18C6 concentration in Figure 2-9. At

the maximum concentration of 18C6 (50 mM), the extent of rectification is 18.8,

or 6 times larger than the electrostatic rectification (3.1) of the conical nanopore

alone.

When 18C6 is added to the side of the membrane facing the small-

diameter opening, steric rectification induced by 18C6 subtracts from the

electrostatic rectification of the conical nanopore. Steric rectification increases

when the concentration of 18C6 is increased, as observed in Figure 2-9. The

extent of rectification for 18C6 added to the side of the membrane facing the

small-diameter opening as a function of 18C6 concentration is also plotted in

Figure 2-9. At the maximum concentration of 18C6 (50 mM), the extent of

rectification is 0.6, or 5 times smaller than the electrostatic rectification (3.1) of

the conical nanopore alone.

If the extent of rectification is 1, then the positive and negative branches of

the current-voltage curve are balanced and there is no rectification. Therefore,

an extent of rectification of 0.6 (50 mM 18C6) indicates that the inherent








electrostatic rectification of the pore is now dominated by steric rectification due

to added 18C6. The extent of rectification measured in control experiments

described previously using LiCI and PEG2 showed little to no change in the

observed rectification.

A higher concentration of 18C6 could also reside in the conical nanopore

due to an enhanced dissociation of K1 8C6 caused by the negatively charged

pore walls. This effect (higher Kd) has been observed for crown ether-Ki

complexes in cation-exchange resins. In this case, the K*-18C6 complex would

be driven into the pore by electrophoresis, the complex would dissociate, and Ki

would enter the double layer to compensate for the negative charge of the

nanopore surface. The uncharged 18C6 present in the pore could only leave via

diffusion or by completing another iC present in solution. Both of these actions

would result in a lowered transmembrane current. Uncharged 18C6 in the pore

would occupy more volume and exclude current-carrying iC. Complexation of

free WK in solution would also lower the current through the nanopore due to the

lowered iC concentration.

Comparison of Synthetic Conical Nanopores vs. a Model Biological
Nanopore, aHL

Chemical sensing and transport of using wild-type a-hemolysin (aHL) and

engineered aHL mutants have been used to demonstrate many of the key

concepts and properties related to biological nanopores.24'80 Results of the

effects of 18C6 analogous to those reported here for synthetic conical nanopores

have also been reported for aHL by Bezrukov et al.31 The results presented

here, in the context of Bezrukov's work, provide a root for comparing the








properties of synthetic conical nanopores with respect to aHL. The current-

voltage response of the nanopore deviates due to the presence of added 18C6

for both biological and synthetic pores. Because the origins and deviations for the

two systems differ, a consideration of the properties of each nanopore system is

necessary.

Since many measurements through nanopores are electrochemical in

nature, similarities in the characteristics of the pores, such as the electric field

across the membrane must be established. A typical voltage used in aHL

experiments is 100 mV, which gives an electric field of 1.0 x 10 V/m over the

pore's length of 10 nm.31 In our measurements with conical nanopores we use

higher voltages, because the effective length of the pore-determined as the

length over which 95% of the resistance is focused-is 140 nm.66 The electric

field across the synthetic nanopores is therefore determined to be 3.6 x 107 V/m

at 500 mV. A greater electric field over the pore's length means that aHL is more

sensitive to small current fluctuations than conical nanopores at a given applied

voltage. However, the synthetic system is robust enough to withstand much

higher voltages than the aHL system, which is limited by the mechanical and

electrical stability of the bilayer.

There are several important geometric considerations that must be

compared between the aHL nanopore and synthetic conical nanopores. The

length of the aHL channel is 10 nm, with a restriction of approximately 1.4 nm

about halfway down the channel.11 On one side of the restriction (the p-barrel

side), the channel has an i.d of 2 nm. On the other side of the restriction (the cap








side) the channel is larger, with an i.d. of -4 nm. The channel of the conical

nanopore is a much longer, 10 pm, although as mentioned, the effective pore

length for electrochemical consideration is only about 140 nm. Instead of having

a restriction halfway down this length, the small opening of the pore is essentially

the restriction.

Both aHL and synthetic conical nanopores are inherently asymmetric in

shape. This asymmetry has been found to be responsible for ion current

rectification. 5881 Dependence of the rectification properties of the two pores on

electrolyte pH indicates that it is the pore surface charge that controls

rectification. In the case of aHL, the presence of 18C6 results in a current-

voltage response that is consistent with the response observed in the case of

synthetic conical nanopores. In the case of aHL, when 18C6 translocates the

pore, current is limited at intermediate values (expressed as conductance by

Bezrukov et al.) due the presence of a physical restriction in the aHL nanopore.

Eventually, when high enough potentials are applied, 18C6 is forced through this

restriction and current again increases.

In the case of the synthetic conical nanopore, the small diameter opening

of the pore is essentially the restriction. Since negatively charged conical

nanopores are inherently cation selective, the electrostatics of the Ki-18C6

complex must be considered. These effects have been described in the context

of previous results obtained. In each case, the effect of 18C6 on both aHL and

synthetic conical nanopores can be easily observed through the current-voltage

response of the pore. Hindered translocation due to steric congestion at the








limiting nanopore radius results in modified current response as a function of

voltage for both natural and synthetic pores.

Each nanopore platform has advantages and disadvantages. For

example, while the structure of aHL is known at the atomic level,11 the synthetic

conical nanopores described here are more robust both chemically and

physically. Additionally, as we have demonstrated previously, the nanopore may

be easily chemically functionalized without the need for genetic engineering.

Conclusion

This work has shown that the addition of 18C6 to an electrolyte solution

affords control over the direction and magnitude of ionic currents through a

conical nanopore, provided the small-diameter of the pore is of molecular

dimensions. The extent of rectification of the conical nanopores is governed by

the concentration of 18C6, and the direction of rectification is governed by the

placement of 18C6 (with respect to the conical pore geometry). It has been

further shown that 18C6 has an effect on the transport properties of the pore only

if it can complex with cations of the primary electrolyte. The results have

potential ramifications in controlling ion transport in nanofluidic channels and in

the development of nanofluidic circuits. The results reported here also have

implications with respect to the design and development of synthetic nanopore-

based sensors.

The observed effects on the current-voltage response of conical

nanopores are explained based on the steric effects of the crown complex

interacting with the pore. The comparable size of the K+-18C6 complex with

respect to the small opening of the conical nanopore is a mitigating factor in the








observed steric effect, which can affect the inherent electrostatic rectification of

the pore. Single conical nanopores and the aHL pore respond in a similar

manner to the presence of crown ethers. This provides us with a synthetic

platform from which to further develop our biomimetic nanopore membranes,

including use of conical nanopores for the study of cellular processes and the

construction of biosensing devices. Ultimately, the hope is to match or exceed

the exquisite sensitivity and selectivity demonstrated with the biological nanopore

aHL in a robust synthetic platform.













CHAPTER 3
STOCHASTIC SENSING DETECTION OF AN ANIONIC PORPHYRIN USING A
SYNTHETIC CONICAL NANOPORE

Introduction

Beginning with the development of the Coulter counter over 50 years

ago,19 there has been a great deal of interest in measuring the movement of

single particles through a channel by monitoring the change in electrochemical

signal.82 Early measurements were limited by the technology available for

creating channels allowing detection of sub-micron sized particles. The advent of

both electrophysiological methods and the production of advanced materials

have extended the Coulter counter technique to channels of nanoscopic

dimensions, allowing the detection of single molecules. 2382

There are two types of nanopore systems commonly used to

stochastically sense single molecules: those based on biological membrane

proteins 14,80 and those based on synthetic membranes. 2382 One of the first

examples of stochastic sensing using a biological channel was reported using the

acetylcholine receptor by Neher and Steinbach.21 Since then, significant work

involving biological channels, including with maltoporin6' 22 and voltage-

dependent anion channel (VDAC) channels,8 have been reported. Additionally, a

large body of work on single molecule detection using the alpha-hemolysin (aHL)

pore has been reported by several groups, including those of Bayley,25, 36 38 80, 83

Deamer,33, 84 Kasianowicz,85 86 and Akeson.34'87 For example, both wild-type








and engineered aHL have been used to stochastically sense a wide range of

analytes including DNA,88 polymers,30 and small molecules in the presence of

molecular adapters38.

Recently, techniques to prepare synthetic membranes containing single

nanopores with dimensions analogous to biological channels have been

reported. A variety of approaches in various materials have been utilized to

prepare such nanopores, including focused ion beam etching of silicon nitride,4'

46 soft lithographic techniques,47'48 embedded carbon nanotubes,' 51' and track-

etched conical nanopores produced in polymeric membranes, such as those

used in this study.52, 53' In the case of large DNA molecules, these single-pore

synthetic membranes have shown stochastic sensing capabilities similar to those

demonstrated by biological systems.58, 65 69,72,89

Of particular interest are the stochastic-sensing properties of single

conical nanopores prepared in polyimide membranes. The ability to prepare

small pores (<10 nm), in addition to excellent pore stability in aqueous solutions

make conical nanopores in polyimide ideal candidates for sensor platforms. In

addition to showing an ability to translocate DNA in a manner analogous to

results obtained with biological nanopores,89 it has been shown that conical

nanopores prepared in polyimide membranes exhibit current stability similar to

that of aHL, which does not fluctuate and maintains a stable baseline in KCI

solutionsS3, o70

Stochastic sensing provides a means to study transport properties at the

single molecule level, an important issue for many applications, such as drug








delivery. It is therefore of interest to study small molecules of biological and

medical interest. Particularly interesting are molecules in the size regime of

therapeutic molecules that are used to treat diseased cells. To this end, the

voltage- and concentration-dependent movements of the negatively charged

porphyrin, 4,4',4",4'"-(porphine-5,10,15,20-tetrayl)tetrakis(benzenesulfonic acid)

(H2-TPPS), through single conical nanopores prepared in polymer membranes

were studied. With a diameter of -2 nm,90 this porphyrin represents a model

molecule, as derivatives of this and other porphyrins are used in the targeted

photodynamic treatment of cancer tumors.91-94

Specifically, the use of a conical nanopore as a molecular Coulter counter-

like device is demonstrated. It is shown that negatively-charged small particles

can be electrophoretically driven through a single nanopore prepared in a

polymeric membrane. It is further shown that conductance through conical pores

prepared in a polymer membrane changes in the presence of a small molecule in

a manner similar to that of biological nanopores by demonstrating both voltage

and concentration dependence. Additionally, it is shown that the observed

current response can be interpreted using a modified Coulter counter model.

Experimental

Materials

Polyimide films (Kapton-50HN, DuPont, 3 cm diameter, 12 pm thick), that

had been irradiated with a single swift heavy ion of 2.2 GeV kinetic energy to

create a single damage track through the film were obtained from GSI

Darmstadt, Germany. Potassium chloride (certified A.C.S., Fisher Scientific),

4,4',4",4"'-(porphine-5,10,15,20-tetrayl)tetrakis(benzenesulfonic acid) (H2-TPPS)








(~98.0%, Sigma), sodium hypochlorite (13% active chloride ion, Sigma) and

potassium iodide (certified A.C.S., Fisher Scientific) were used as received. All

solutions were prepared in 18 mega-ohm water (Bamstead, E-pure).

Preparation of the Conical Nanopores and Estimation of Tip Diameter

Membrane samples that contained a single conically shaped nanopore

were prepared by anisotropic chemical etching of the heavy-ion irradiated Kapton

films, as per the procedure described in chapter 2. The diameter of the tip of the

conical nanopores was subsequently calculated by solving Equation 2.1 for d,


A C, Axopatch Digidata
Reference 200B Working
Electrode Electrode
Computer



1 M KCI 1 M KCI


Tip-opening Base-opening

Figure 3-1. Electrochemical cell set-up for measuring current-time transients.

Current-Time Recordings

The single conical nanopore membrane was mounted in a conductivity

cell, and both half-cells were filled with 1 M KCI, pH 8.0 solution. A Ag/AgCI

electrode was inserted into each half-cell, with the working Ag/AgCI electrode in

the half-cell solution facing the base of the nanopore and the reference Ag/AgCI

electrode in the half-cell solution facing the nanopore tip as shown in Figure 3-1.

A transmembrane potential from -600 mV to +600 mV was applied. The resultant








current was recorded using an Axopatch 200B amplifier (Molecular Devices

Corporation, Union City, CA) in the voltage-clamp mode with a lowpass Bessel

filter at 2 kHz bandwidth. The signal was further digitized by a Digidata 1322x

(Molecular Devices Corporation) at a sampling frequency of 10 kHz. The data

were recorded using pClamp 9.0 (Molecular Devices Corporation). For statistical

analysis, the signal was recorded in three or more 30-second segments at each

applied transmembrane potential. The current-time data reported here are

representative of recordings made on a nanopore with a small diameter opening

of -4.5 nm. Individual events were detected and analyzed with QuB software

available at www.qub.buffalo.edu/.

Results and Discussion

The method of detecting a single particle with a nanopore is based on the

comparable dimensions of the pore diameter and the size of the particle to be

detected.14 When a particle is passing through a nanopore, the particle will

partially block it. This can be observed as decrease in ion current. The duration

and amplitude of the blockage are related to the geometry of the particle.20

Another consequence of the small pore diameter is the interaction of the

translocating molecules with the pore walls.14 These interactions, together with

variations in structure along the molecule, cause variations in the ion current

signal.

The detection of a single analyte molecule with a conical nanopore relies

on similar diameters of the nanopore tip (d,) and the analyte of interest. When an

analyte passes through the nanopore, the flow of ions through the pore is

partially blocked.53'89 This is observed as a transient decrease in the








transmembrane ion current. The duration and amplitude of the blockage are

indicative of both the properties of the analyte (e.g., charge, hydrodynamic

radius) and pore-particle interactions. It has been shown using biological

nanopores, that the identity of the analyte can be determined from the magnitude

and length of the current blocks.14 Additionally, the concentration of the analyte

can be determined by the frequency of the observed current blocks.



HOS SO3H





NH HN





H03S SO3H

Figure 3-2. 4,4',4",4'"-(porphine-5,10,15,20-tetrayl)tetrakis(benzenesulfonic acid)
(H2-TPPS).

Current-Time Response of Nanopore in 1 M KCI

To interpret the effect of 4,4',4",4"'-(porphine-5,10,15,20-tetrayl)tetrakis-

(benzenesulfonic acid) (H2-TPPS) (Figure 3-2) in a single conical nanopore

membrane, it is necessary to first consider the case of supporting electrolyte

alone. In Figure 3-3, the current response as a function of time behavior of a

conical nanopore is demonstrated at various applied transmembrane potentials

in 1 M KCI at pH 8.0. As can be observed, in the absence of the H2-TPPS, a













100 mV


1000 ms


400 mV



jjj_6 lI Al jj ".. N.... lw .i. ".1.1 A .. t m Ij -IIJO r,1 ,


1000 ms


-400 mV


-r ir i M- Wf V" iifii- 7i;7ir
1- ~~ ~ 1 -, -l 6 -


v-i r" r


* -T' I T ."" *


1000ms


Time (ms)


Figure 3-3. Current-time recordings of 1 M KCI pH 8.0 at (a) 100 mV, (b) 400
mV, and (c) -400 mV.


700


650


600


2550


2500


2450


-3700


0.
C -3750



-3800


-- -,W










700
<-

| 650
o -
600-


(b)

2550


S2500


2450


- 3050

S3000


2950


(d)
3500


3400


3300


100 mV


1000 ms
I I


400 mV


1000 ms
i I


500 mV


600 mV


1000 ms
I--------I


Time (ms)


Figure 3-4. Current-time recordings showing voltage-dependence of H2-TPPS
translocation events in 1 M KCI pH 8.0 at (a) 100 mV, (b) 400 mV, (c)
500 mV, and (d) 600 mV.


w








steady-state membrane current is obtained. The exceptional stability of the ion

current signal through Kapton nanopores was demonstrated and studied

previously in detail.3,65

Voltage and Concentration Dependence of the Current-Time Response of
Nanopore

By examining the voltage dependence of a molecule in an electric field, it

is possible to determine at what applied potential a molecule is driven through

the pore electrophoretically. In the case of the negatively charged H2-TPPS,

translocation events are observed when a positive transmembrane potential is

applied, but not when a negative transmembrane potential is applied. This is

because, at pH 8.0, the sulfonate groups of H2-TPPS are deprotonated,90 giving

it a net negative charge of 4.

Given the arrangement of the electrodes, as shown in Figure 3-1, when a

positive transmembrane potential is applied, anions such as the H2-TPPS, are

driven electrophoretically through the conical pore towards the anode, as shown

in Figure 3-4. When a negative potential is applied, the anionic H2-TPPS is

driven away from the pore. Additionally, events are not observed below an

applied potential of +300 mV. This is likely due to a combination of the

electrostatic repulsion of the negative charges on the nanopore surface, as well

as the entropic penalty paid by a molecule entering into a very narrow pore. This

is expressed by the existence of a threshold voltage (+300 mV) for the H2-TPPS

to enter the pore.

The amount of reduction in current, or the mean amplitude of blockade

event, increases at increased applied potentials. The amount of blockage






56



(a) 2000 400 mV (2) 2513.530.02 pA
17650

-1600

1250

o 1000
760.
750

600. (1) 2493.530.05 pA

2650

0
2476 2486 2495 2505 2616 2626
Current (pA)
(b) 0so 500 mV
(2) 3024.79*0.06 pA

600
(1) 2990.060.07 pA

0
4.,

200


0
2930 2960 2970 2990 3010 3030 3060
Current (pA)
(C) 700 600 mV
(2) 3448.180.14 pA
600

(. 600&


0:30 (lb) 3419.950.1 pA
0 300\

2W
(la) 3367.56t0.12 pA


0
3360 3380 3400 3420 3440 3460 3480 3500 3520 3640 3560
Current (pA)

Figure 3-5. Histograms showing increase in magnitude of blockade events of H2-
TPPS as the applied potential increases: (a) 400 mV, (b) 500 mV, and
(c) 600 mV.








caused by the porphyrin can be determined by analyzing histograms of ion

current values that have been fitted with a Gaussian function, as shown in

Figure3-5 for 400 mV, 500 mV, and 600 mV.

The ratio of the current in the presence of the molecule (1) increases

relative to the current when no molecule is present (2). At 600 mV, the

translocation events are represented by two distributions (1a and 1 b) due to two

levels of current block. The distribution represented by 1 b would appear to be

due to a "flicker" in the current due to the very rapid translocation of the

molecules. This may be due events occurring so rapidly that they can not be

resolved. This may also be due to the orientation of the molecule as it passes

through the effective length of the pore.

The duration of the translocation time, or dwell time, is the amount of time

it takes for the H2-TPPS to pass through the effective length of the pore

(calculated as 100 nm from Equations 1-2 and 1-3). The dwell times were found

using the software QuB and analyzed with Excel and Origin. Distributions of

dwell times for the applied voltages 400 mV, 500 mV, and 600 mV are shown in

Figure 3-6. The translocation time, calculated as an arithmetic average, is

observed to decrease with increased applied potential: for 400 mV, the average

dwell time is 30 ms, while for 600 mV, it is only 4 ms. Such a strong voltage

dependence of the translocation time is expected due to the presence of a strong

negative charge of the H2-TPPS molecule.. This is understood by the

relationship between the electrophoretic driving force and voltage-at higher








applied transmembrane potentials, the electric force is greater on the H2-TPPS

molecule in the electric field, causing it to move faster through the nanopore.


400 mV




~30 ms


4 12 20 28 36 44 52 60
Time (ms)

1 500 mV


< t > 18 ms


. *..... I......


2 10 18 26 34 42 50 58 66
Time (ms)


600 mV



~4 ms


2 10 18 26 34 42 50 58
Time (ms)


Figure 3-6. Histograms of dwell time of H2-TPPS in nanopore at (a) 400 mV, (b)
500 mV, and (c) 600 mV.


(b) 160

' 120

0 80
04
z40

0


a300

0200

Z 100


iii .11Ihh....


>








When the average dwell time for each applied potential is plotted against

the inverse of that potential, a proportional relationship is observed, as shown in

Figure 3-7. Extrapolation of the trendline to the axes reveals that the y-intercept

is not at (0, 0), but rather where the value of V is 300, due to the existence of a

threshold potential at which the porphyrin started to translocate the nanopore.

40-

3.0
E
S20-




1/V

Figure 3-7. Inverse relationship of translocation time to applied potential.

In addition to the magnitude and duration of events, the frequency of events

also increases as the applied transmembrane potential increases. Histograms

corresponding to recordings made at 400 mV with 20 nM and 60 nM H2-TPPS

Figure 3-8) were fit using an exponential function of tf where a is a fitting

parameter between 1 and 1.3. An observable increase in events at higher

concentrations is observed; for example, there were 505 events at 60 nM vs 103

events at 20 nM for the times analyzed. Similar trends for DNA translocation

through both biological and other types of synthetic nanopores have been

reported.43, 46, 87,89

The voltage and concentration dependencies support our supposition that

the events correspond to the molecules translocating the pore. The bumping

events are independent of voltage and concentration.32











0 nM H2-TPPS


1 M KCI


1000 ms


20 nM H2-TPPS


1000 ms


60 nM H2-TPPS


1000 ms


Time (ms)


Figure 3-8. Current-time recordings showing concentration dependence of H2-
TPPS translocation events in 1 M KCI pH 8.0 of (a) 0 nM, (b) 20 nM,
and (c) 60 nM at +400 mV.


2600-


2500-


2400


2600-


2500-


2400


2550


2500


2450


LLLAZ:


I I


.1 I


I -.- -








Blockage of Current by H2-TPPS

The last point to consider is the percentage of blockage that is observed

with the H2-TPPS molecules. In the case where the diameter of a nanopore is of

similar size as the diameter of the molecule, it is thought that the current will be

blocked significantly due to sterics.14 The diameter of the H2-TPPS is 2 nm,90

and the nanopore tip diameter is -4.5 nm, based on the calculation from

Equation 2.1. However, a significant current block is not observed relative to the

open-state ionic current. To estimate whether the observed current drop

corresponds to the size of the molecule, theoretical modeling applied for

detecting particles in the Coulter counter technique was used.

This theory, based on Maxwell's expression for the effective resistivity of a

dilute concentration of insulating spheres in solution9 says that the resistive

pulse is proportional to the cube of the diameter of the particle and is inversely

proportional to the 4m power of the diameter of the pore. Proposed originally by

Kubitschek,9 the formula for sizing sub-micron particles was put forth by DeBlois

and Bean in 1970,20 and relates the sensitivity of a Coulter counter to the

relationship between the size of the pore and the size of the particle to be

counted. As a particle enters a pore, the resistance of that pore is increased due

to the temporary displacement of the fluid in the pore. Experimentally, the

expected change in resistance or amount of current block can be found by

calculating the resistance of the pore containing a particle versus the resistance

of the pore in the absence of a particle.

To obtain a numerical answer for the resistance of a conical nanopore

containing a particle, the difference in resistance between the cross-section of








the pore at a given point minus the cross-section of the particle must be

calculated. This required the development of a set of equations applicable to the

conical geometry of the polymer nanopores and was done with Dr. Zuzanna

Siwy. The following two assumptions were made, as per the Coulter counter

theory.20 First, it is assumed that the particle is spherical in shape. Second, it is

assumed that the only influence of the particle as it translocates through the pore

is that the particle expels that volume of solution equal to the volume of the

particle.







y Y RC
------------ ----R--r*

rc











L
Figure 3-9. Schematic definition of terms needed to calculate cross-sectional
area of nanopore in absence of particle.

To determine the radius of a cone at specific location along its axis, it was

first necessary to consider the similarity between two triangular regions both

bounded on one side by the pore wall, one with the other sides represented by z








and y, and one with the other two sides represented by L and (R-rr), as shown in

Figure 3-9, where


y_ (Rc-ro)
z L


Equation (3.1)


where re is the radius of the tip-opening of the conical nanopore, R, is the radius

of the base opening, and L is the effective length of the pore. This equation re-

arranges as follows:


Sz(R r)
y- L
L


Equation (3.2)


Because the radius Y, defined as the radius of the cone at given position of z, is

equal toy + re, the following is obtained:


Yz(R. -r.)
L


Equation (3.3)


The cross-section of a cone at point z is therefore rY2.


Figure 3-10. Schematic definition of terms needed to calculate cross-sectional
area of particle.

To calculate the cross-sectional area of the particle (Figure 3-10), we must

first determine r2 using the Pythagorean Theorem:

r2 = r +( -x) Equation (3.4)








where r, is the radius of the particle, r is the radius of the cross-sectional disc of

the particle, and x is the position of this disc relative to the pore axis. This

equation rearranges to:

r2 = 2rx-x2 Equation (3.5)


The area of this cross-section is therefore

A = r(2rx- x2)


Equation (3.6)


Figure 3-11. Schematic definition of terms needed to calculate cross-sectional
area of nanopore in the presence of particle.

The cross-sectional area at point x of a cone with the particle in it is

determined by


Equation (3.7)


A = (Y2 r2)








where Y is given by Equation 3.3 and r is (2rjx-x2) from Equation 3.5. To

calculate the change in resistance of a cross-section of a cone with a particle in

it, we start with the general formula for resistance:

R = p- Equation (3.8)
A

where p is the specific resistivity of the supporting electrolyte, L is the length and

A is the cross-sectional area of the nanopore. Only the region of the cone over

which 80% of total resistance is focused was considered, which in this case, is

100 nm in this case (from Equations 1.2 and 1.3). The resistance of a cone with

a length of 100 mn, without a particle in it, is defined as follows:


RI = p- L )- Equation (3.9)
T 00 .nm

To calculate the resistance, R, of a conical nanopore with a particle in it, it is

necessary to divide the cone into infinitesimally small slices of thickness, dx.

These slices can be treated as discs. Resistance of each of the slices of cone

with a particle in it is expressed as:


dR = p [ -x- Equation (3.10)
Z y2 -2rx + x2

The particle of radius rj is placed between x=0 and xr=2r,, which are limits of the

following integral:

2r,
R= d[ dxr- 2r,x+x2 Equation (3.11)


which, in turn, becomes:









R=P 2 [ (+ o)(R,-r)+Lrc ]2r2x+X2
10 L Equation (3.12)
p 2r, dC
SF(x)

where xo determines the position of the particle in the cone and x+xo=z. When

the denominator of the integral, F(x), is expanded, the following is obtained:


L(x)=x2 L2 +

2x(R-r2 2(R, r)r 2r
L2 L

x0 2 (Rc- -r.)2 2x, (R, r )rc 2
EL L I


Equation (3.13)


This expansion looks like "Ax+Bx+C." In order to simplify the integration, we

apply this form so that for


P 2r
R = Ax2 +Bx+C

the following terms are obtained:

A =(RC r) +1J

B =2xo (Rc r) 2(R -r)r_
L2 L j
x (R -r) 2x(Rc-re)rc 2
C = 2( -r-)Z--+rc
L L

The integral to solve is of the type

f dr
S-=arctan x + C


Equation (3.14)




Equation (3.14a)

Equation (3.14b)

Equation (3.14c)


Equation (3.15)





67

because Ax2+Bx+C in this case cannot be expressed as A(x-x1)(x-x2). In order to

fit Ax+Bx+C to this form, we use the Binomial Theorem, where


(a+b) +c=a2 + 2ab+b' +c

By applying this to the denominator of Equation 3-14, a and

a = x
B
b =-
2


Equation (3.16)

b can be defined as:

Equation (3.16a)
Equation (3.16b)


which results in:


Equation (3.16c)


This in turn gives the following:


2 2
(a+b)2 +c = x+ +C_ K


B 22

.%Ax + B )2= x 2
2,F


C-_ =1
[2^A-


Equation (3.16d)




Equation (3.17)


Equation (3.18)


the following is obtained:


R f 2

4A) 2,A +
f--- -- +1
I 4 2
lV 4A )


Equation (3.19)


If we let


and


(a+b)2 = ( + 2( )(2+





68

Integration by parts is used, withy, d', and dx defined as follows:


Equation (3.20a)


C^-
4A




&dr
C- d~t


dx-c -4A


Equation (3.20b)





Equation (3.20c)


This in turn leads to the following:


B2
T4A dx

B 2B2


4A
PC- -- (arctan X, arctan Xo)









C- -4A


I(2r,) + B
2,14
4A
r4A


Equation (3.21)









Equation (3.21 a)


Equation (3.21b)


where:


and








To find a numerical solution for the change in resistance of a conical

nanopore with a tip-opening diameter of 4.5 nm and a particle with a diameter of

2 nm, the above set of equations is solved using r, = 1 nm, rc = 2.25 nm, R, = 1

pm, and L = 10 pm;. The position of the particle in the pore, xo, is changed

between 0 and 30 nm, as the maximum increase in resistance caused by the

presence of a single particle is obtained when the particle is placed within the

first 30 nm inside the tip-opening of the cone.

The calculated increase of resistance R caused by a single particle is

compared with the resistance of the cone without a particle in it by dividing the

change in R by Rtot,,:


R = Equation 3.22
R
btaj

The calculated average resistance increase is 2%. Experimentally, the

observed increase of resistance caused by H2-TPPS equal to approximately

2.5%. Comparing the calculated change in current block to the experimental

data, this experimental data is in good agreement with the theory.

Conclusion

The effects on conductance in a conical nanopore resistance in the

presence of H2-TPPS have been explored. Under an electric field, these

negatively charged molecules were shown to move and, at the appropriate

transmembrane potentials, translocate the nanopore. It was also shown that the

length of time it takes for the molecule to translocate the nanopore has an

inversely proportional relationship to the applied potential. It was further shown








that the number of events observed scales with both applied potential and

analyte concentration.

The observed events of H2-TPPS translocating the conical nanopore are

explained using the Coulter counter principles. The comparable size of the H2-

TPPS to the nanopore causes a decrease in the conductance of the pore,

resulting in lower ionic currents. The experimentally observed current decreases

have been shown to be in good agreement with the theoretical value. These

single conical nanopores respond in a manner similar to biological nanopores

when used as molecular Coulter counters. This provides us with a synthetic

platform from which to further develop biomimetic nanopore membrane sensing

platforms.












CHAPTER 4
POLYMER FILMS AS SUBSTRATES FOR MOBILE SUPPORTED LIPID
BILAYERS

Introduction

There has been a great deal of interest in the use of supported lipid

bilayers as components of biosensing platforms.9799 This interest has risen from

the need for a mechanically rugged support in which to incorporate biological

nanopores. These nanopores, such as alpha-hemolysin (aHL), have exhibited

stochastic sensing properties with potential use in real-world sensing.24 Until

recently, it has only been possible to incorporate biological nanopores into planar

lipid bilayers, which span an orifice of 50-150 pm in a Teflon sheet. Supported

only by the buffer solutions on either side of the membrane, the planar bilayer

system has been shown to be mechanically and temporally unstable for a

practical sensing device.23

One attempt to overcome the stability issues inherent in planar bilayer

systems has been to immobilize biological nanopores in or on a substrate in the

absence of a bilayer.23 This strategy, which has involved the incorporation of

aHL into the mouth of a conical nanopore, has proven difficult, both in the

positioning and orientation of the protein and in the ability to sense analytes with

such a system. One reason for this difficulty is the activity of many biological

nanopores occurs only in the biological matrix of a lipid bilayer. Indeed,








monomeric aHL forms a channel through a cell membrane only if the membrane

is of the appropriate thickness and composition.26

In order to develop a bilayer platform for biosensing, it has become critical

to find both a suitable support for lipid bilayers that will provide mechanical

stability and a bilayer of appropriate composition that allows for biological

function of incorporated biological nanopores.1"0

Interest in supported bilayer systems is not new. In the 1980s, Tamm and

McConnell developed the first generation of the so-called solid-supported

membranes. In one of these studies, a lipid bilayer composed of L-a-

dipalmitoylphosphatidylcholine was formed on a oxidized silicon substrate.100 In

order to optimize the substrate surface coverage and the mobility of the bilayer, a

variety of other supports for lipid bilayers have been studied. In addition to

silicon,101'102 the most common substrates, requiring little to no surface

preparation, are glass,13 "104 mica,'o0 and quartz.106 Additionally, there has been

limited success with other substrates such as poly(dimethyl siloxane) (PDMS),107'
108 xerogels/aerogels,109 polymer cushions,97' 110 and metal substrates.99" 111,112

In these cases, surface preparation has proven to be more difficult, often relying

on complicated substrate modification schemes and bilayer formation

techniques. These include formation of bilayers using Langmuir-Blodgett1"' or

Langmuir-Schaefer techniques,113 or the "painting" of a lipid solution in an

organic solvent onto the substrate.114

Much of this previous research has demonstrated that it is difficult to put a

defect-free, mobile bilayer on a support.115 This in turn, has resulted in a variety








of strategies to encourage this bilayer formation. Some such strategies include

the use of hybrid bilayers,116 tethers,115 substrate surface modification including

self-assembled monolayers (SAMs) of thiols,117 grafting of oligonucleotides,118

incorporation of the S-layer crystalline bacterial cell surface layer,119,120 and

probably most successfully, vesicle fusion.106

The formation of supported lipid bilayers via vesicle fusion involves the

preparation of lipid vesicles by either the extrusion121 or sonication method.22

The extrusion method for preparation of small unilamellar vesicles (SUVs) is

preferred as it results in the formation of vesicles with diameters of less than 100

nm.123 Such vesicles have been shown to form relatively defect-free mobile

bilayers on a variety of surfaces. The successful fusion and spreading of lipid

bilayers is based primarily on interactions between the substrate surface and the

vesicles. In particular, the interaction of vesicles and resulting bilayers with

surfaces involves a balance among van der Waals, electrostatic, hydration, and

steric forces.124' 125 Generally speaking, hydrophilic surfaces that have strong

attractive interactions with lipids promote vesicle fusion and subsequent bilayer

formation.

Considerations when choosing a substrate to use as a bilayer support

include the substrate effect on the mobility of the bilayer (the rate of lateral

diffusion), the effect of the substrate on defects in the bilayer (e.g, surface

roughness can inhibit formation of a uniform bilayer), and the complexity of

substrate preparation.97 In order to overcome these challenges, we have elected

to use a polymer membrane containing a single conical nanopore as a substrate








for the fusion of vesicles. The advantages of such a support include good

mechanical properties, such as surface smoothness and flexible substrate (unlike

glass, mica, or quartz), a surface chemistry that easily controlled (e.g., by pH,

chemical modifications), and, with an eye to developing a sensing platform, the

ability to control the number and size of pores through the membrane for use in

sensing applications.63'65

The choice of support is not the only consideration in developing a

supported bilayer system. Indeed, the components of the bilayer play a key role

in the both stability and mobility of the bilayer on a specific support. Additionally,

it is the bilayer components that provide the appropriate environment for the

activity of a biological nanopore.106' 12 An exciting development in preparing

supported lipid bilayer systems has been the incorporation of PEGs into the

system.127 PEG is an inert, water soluble polymer consisting of ethylene oxide

monomer units. Its physiochemical properties have been well established.12

The size and degree of polymerization can be tailored easily and the polymer can

be attached to other moieties via a linker. A property of many polymers,

including PEGs, is that their conformation is dependent on both polymer size and

number of polymer molecules per unit area (grafting density).128'129 At low

grafting density, PEG exists as a random coil or mushroom conformation. As the

grafting density increases, the polymer undergoes a transition from a globular

structure to an extended or brush like configuration. This transition from low to

high density depends on the polymer size and is denoted as the mushroom-to-

brush transition.130








One of the more successful strategies for supporting a bilayer uses PEG

polymers either tethered to a substrate115,131 or incorporated into the lower leaflet

of the bilayer,132 forming a cushion for the lipid bilayer. Additionally, PEGs have

also been conjugated to phospholipids such as phosphatidylethanolamine

(PE).130, 133 Inclusion of PEG-PE in the formation of liposomes has been used as

a means to impart increased steric and temporal stability to the liposome.

Utilizing PEG-PE provides a controlled means of varying the polymer density in

lipid membrane surfaces by controlling the mole fraction of the lipopolymer.

Taking advantage of these two conformation regimes may impart new properties

to supported lipid bilayers.130, 134

By incorporating lipids containing the PEG moiety into vesicles, the need

to chemically modify the substrate surface is circumvented but the advantage of

using PEGs as a part of the bilayer support is retained. Additionally, by using

PEG-PE, the PEG density in the bilayer is controllable, allowing for the

determination of the optimal PEG size and conformation in the bilayer.130 An

added bonus of the use of PEG-conjugated lipids in the formation of vesicles is

the increased fusion of the vesicles to the substrate.135 Finally, we know that a

true biological membrane is comprised of many more components than simply

lipids, and it is the presence of components such as the glycocalyx that provide

the structural stability necessary for membrane integrity.1 It has been

hypothesized that the addition of PEG-PE in a bilayer membrane imparts to the

bilayer structural stabilization in a manner similar to that of the glycocalyx and

other membrane components in a cell membrane.110' 134








The development of a supported lipid bilayer system on a nanopore

polymer membrane is demonstrated here. By incorporating PEG-lipids into the

bilayer membrane, a mechanical and temporal stability, not available in planar

bilayer systems, is shown. Additionally, using fluorescence microscopy, it is

shown that the supported lipid bilayers on polymeric supports are mobile. The

properties of this system are investigated by surface characterization techniques

and electrochemistry.

Experimental

Materials

Polyimide films (Kapton-50HN, DuPont, 3 cm diameter, 12 pm thick) and

poly(ethylene terphthalate) (PET) films, that had been irradiated with a single

swift heavy ion of 2.2 GeV kinetic energy to create a single damage track through

the film were obtained from GSI Darmstadt, Germany. The following were used

as received: potassium chloride (certified A.C.S.), potassium iodide (certified

A.C.S.), formic acid (88% in water), and sodium hydroxide (certified A.C.S.) from

Fisher Scientific; sodium chloride (certified A.C.S.), sodium phosphate

monobasic (A.C.S. reagent > 99%) sodium phosphate dibasic (A.C.S. reagent >

99%). sodium hypochlorite (13% active chloride ion), and ,3-cyclodextrin (purum >

99%(HPLC)) from Sigma-Aldrich. Small unilamellar vesicles were prepared from

1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol]

(PEG-PE) with a molecular weight of 550 Da or 200 Da, L-a-phosphotidylcholine

from egg (egg-PC) (Avanti Polar Lipids, Alabaster, AL) and N-(Texas Red

sulfonyl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (Texas Red-








PE) (Molecular Probes, Eugene OR). Micro-wells were prepared from

poly(dimethylsiloxane) (PDMS) (Dow Corning Sylgard 184, Fisher Scientific) and

borosilicate microscope coverslips (VWR International). All solutions were

prepared in 18 mega-ohm water (Barnstead, E-pure).



(a)

/ 0
0



(b) Ho
CHOC(HzCH20)n-C-N" O* OI- 'O /
N*
NH4


(CH3CH2)3NH


Figure 4-1. Structures of lipids used in preparation of vesicles: (a), L-a-
phosphotidylcholine from egg (egg-PC), (b) 1,2-dioleoyl-sn-glycero-3-
phosphoethanolamine-N-[methoxy(polyethylene glycol] (PEG-PE) with
a molecular weight of 550 Da or 200 Da (PEG-PE), and (c) N-(Texas
Red sulfonyl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine
(Texas Red-PE).








Preparation of Substrates for Surface Characterization and Fluorescence
Experiments

Upon receipt, the borosilicate glass cover slips were cleaned and

annealed in a kiln at 400 C. PET and Kapton membranes were chemically

etched then rinsed overnight in water and dried under nitrogen. Additional PET

and Kapton membranes were used as received. Prior to surface

characterization, polymer and glass samples were exposed to air-plasma for

twenty seconds under full power (Harrick Plasma Cleaner/Sterilizer, Ossining,

NY) on the side of the sample that faces the bilayer. For fluorescence recovery

after photobleaching (FRAP) experiments, a PDMS micro-well was pressure-

sealed to the glass or polymer surface prior to plasma cleaning.

Preparation of Vesicles

Small unilamellar vesicles (SUVs) were prepared by first mixing the

appropriate mole fractions of PEG-PE and Texas Red PE with egg-PC (Figure 4-

1) in chloroform as follows: 99.5 mol% egg-PC/0.5 mol% Texas Red-PE, 94.5

mol% egg-PC/5 mol% PEG-PE/0.5 mol% Texas Red-PE, 98.5 mol% egg-PC/1

mol% PEG2-PE/0.5 mol% Texas Red-PE, 94.5 mol% egg-PC/5 mol%

PEG2-PE/0.5 mol% Texas Red-PE. Each mixture was then evaporated under

a stream of nitrogen, and desiccated in vacuum for 4 hours. Rehydration of the

lipids to 1 mg/mL is done in 150 mM NaCI solution at pH 7.4. The solution is

then frozen in liquid nitrogen followed by thawing in water at 30 C. After 10 such

free-thaw cycles, the resultant large vesicles are extruded through a

polycarbonate ultra-filtration membrane with an average pore size of 50 nm using

a extruder (Northern Lipids, Vancouver, British Columbia, Canada).








Dynamic Light Scattering to Determine Vesicle Size

In order to verify the final diameter of the extruded vesicles, the vesicles

are sized using dynamic light scattering (DLS). This is determined using a

90Plus Particle Size Analyzer by Dynamic Light Scattering from Brookhaven

Instruments (Holtzville, NY). 3 pL of vesicle solution are mixed with 97 pL of 18

mega-ohm water. After allowing the solution to equilibrate to the temperature of

the instrument (25 C), 5 measurements are taken, three minutes apart.

(a)


Figure 4-2. Vesicle fusion to substrate.

Bilayer Formation

Formation of supported lipid bilayers was achieved via vesicle fusion.121

In this method, a bilayer is formed due to the adsorption and subsequent rupture








of vesicles on the substrate. This results in the formation of single bilayer disks.

These disks then coalesce to form a continuous lipid bilayer membrane. SUVs

containing the appropriate mol fractions of egg-PC, PEG-PE, and Texas Red-PE

were fused to either planar borosilicate slides or polymer membranes as shown

in Figure 4-2. Approximately 100 pL of vesicle solution is placed on the substrate

and allowed to incubate. After incubation, the bilayer-covered substrates are

thoroughly rinsed with buffer or water.

Fluorescence Recovery After Photobleaching Experiments

Fluorescence recovery after photobleaching (FRAP) experiments are

based on the principal of observing the recovery of fluorescence in a bleached

area.136 In such an experiment, a pulse of high intensity light is used to

deliberately and irreversibly photobleach the fluorescent molecules in a defined

area. Recovery of fluorescence in this bleached region represents the diffusive

movement of unbleached fluorophores into that region and bleached

fluorophores out of it. Quantitation of fluorescence intensity over time is used to

determine the percent of mobile lipids in the bilayer and diffusion coefficient of

the fluorescent molecule.

For the FRAP experiments in this work, samples were imaged using an

inverted epifluorescence Nikon Eclipse TE2000-U microscope (Nikon

Instruments Inc., Melville, NY) under 10X magnification equipped with a Spectra-

Physics argon-krypton laser (534 nm argon line, 100 mW) (Mountain View, CA).

An area of 17.7 pm2 was bleached under full laser power. The fluorescence

recovery of the bleached spot was regularly imaged using a Roper Scientific

CCD camera (Trenton, NJ), and imaging analysis was done using MetaMorph








software (Universal Imaging, Downingstown, PA). The lateral diffusion

coefficient of 0.5 mol% Texas Red-PE in the lipid bilayer was measured via this

series of scans, by first integrating the fluorescence intensity for the entire

bleached area and plotting this intensity versus elapsed time, as shown

schematically in Figure 4-3. The diffusion coefficient of the fluorescent probe

was calculated from the recovery vs. time curve, fitted with a first-order

exponential-rise function to obtain the ti/2. Five measurements were obtained for

each bilayer on each type of sample.

Fluorescence before Fluorescence after
Fluorescence after
photobleaching photobleaching






ce
O W Fluorescence
0 10 recovery
IL 0

Tim


Bleaching of spot
with laser

Figure 4-3. Principle behind FRAP.

Atomic Force Microscopy (AFM) Surface Characterization Experiments

A Nanoscope-III Atomic Force Microscope (Nanoscope III, Digital

Instruments, Santa Barbara, CA) equipped with a 12 pm scanner (J-scanner,

Digital Instruments) was used. All images were recorded in tapping mode with

the cantilever immersed in a liquid droplet of 18 mega-ohm water placed on the








substrate in the fluid cell. Standard silicon nitride tips (NanoProbes, Digital

Instruments) were used. Images (512 x 512 pixels) were flattened line-by-line

during recording. The Nanoscope v4.4 software (Digital Instruments, Santa

Barbara, CA) controlling the microscope was also used to determine the mean

roughness of the surface topography of the solid supports. The fast scan

direction was set perpendicular to the long axis of the cantilever. The scan

speed was optimized for each image individually (typically 1-2 pm/s).

Preparation of the Conical Nanopores and Estimation of Tip Diameter

Kapton membrane samples containing a single conically shaped nanopore

were prepared by anisotropic chemical etching of the heavy-ion irradiated films,

as per the procedure described in chapter 2. The diameter of the tip of the

conical nanopores was subsequently calculated by solving Equation 2.1 for d.

PET membrane samples containing a single conically-shaped nanopore

were also prepared by anisotropically etching polymer membranes which

possess a single ion track.52 The procedure is similar to that described in

chapter 2, however in this case, the etch solution is 9 M NaOH and the stop

solution is 1 M formic acid in 1 M KCI. The etching process was done at room

temperature and the process is stopped when a current of 0.2 nA observed. The

diameter of the tip of the nanopore is again calculated using Equation 2.1 for d4.

In this case, db is determined by multiplying the etch time by the bulk etch rate

(2.12 nm/min). (Further details regarding the etching conditions are reported

elsewhere.)52,65







Electrochemical Measurements
Measurements to explore the electrochemical properties of the polymer-

bilayer system were made at room temperature (23 0C). Electrolyte solution for

these recordings was 1 M KCI at pH 8.0. The single conical nanopore

membrane was mounted in a conductivity cell. A Ag/AgCI electrode was inserted

into each half-cell solution, as depicted in Figure 4-4. The membrane was

mounted such that the membrane face with the conical nanopore base was in the

half-cell solution containing the working electrode (hereafter designated the trans

side of the membrane). The working electrode potential was controlled with

respect to the electrode in solution on the opposite side of the membrane. The

conical nanopore tip faced this second electrode (hereafter designated the cis

side of the membrane).

1322x AgAgCI 200B AglAgCI
Reference ||Working
Electrode Electrode


Supported
Lipid I MI KCI 1 M KCI
Bilayer
cis-compartment trans-compartment


Figure 4-4. Electrochemical cell set-up for bilayer experiments.
Vesicle solution was added to the cis-side of the polymer membrane and

incubated for 6-12 hours, to allow ample time for vesicle rupture and fusion on

the polymer surface. This was followed by careful rinsing with fresh buffer








solution for 15 minutes. Data were acquired using an Axopatch 200B (Molecular

Devices Corporation, Union City, CA) in the voltage-clamp mode with a lowpass

Bessel filter at 2 kHz bandwidth. The signal was further digitized by a Digidata

1322x (Molecular Devices Corporation) at a sampling frequency of 10 kHz. The

data were recorded using pClamp 9.0 (Molecular Devices Corporation).

Results and Discussion

Sizing Small Unilamellar Vesicles Formed by Extrusion

The size of small unilamellar vesicles prepared from egg-PC, Texas Red-

PE, and either PEG5-PE or PEG20-PE by the extrusion procedure was

determined using the techniques of dynamic light scattering (DLS). These

vesicles were shown to have an average diameter of 80 10 nm. Vesicles

prepared without the addition of the lipopolymer were shown to have a slightly

larger diameter of 85 10 nm. When checked after 7 and 14 days, DLS of the

vesicle solutions showed that the diameter of vesicles containing PEG had not

changed, while the PEG-free vesicles had increased in diameter to

approximately 100 nm. This supports the observations of Marsh et al.,13 that the

presence of PEG both reduces and stabilizes the size of the vesicles.

Surface Characterization

It had long been thought that mobile, defect-free bilayers could only be

formed on extremely smooth surfaces. It has been well-documented that such

bilayers could easily and reproducibly be formed on surfaces such as borosilicate

glass,103,104 with an average roughness of < 0.2 nm. Interested in the use of a

polymer membrane as the support for this system is due to the known

mechanical properties of the polymer membranes, such as flexibility and an








easily tailored surface chemistry. It was therefore necessary to be assured that

the surface roughness was on the same scale as borosilicate glass.


Figure 4-5. AFM images and corresponding line scans of (a) glass under water,
(b) PET under water, and (c) Kapton under water.








AFM characterization of the plasma-cleaned glass, PET, and Kapton

surfaces was carried out by acquiring height images under identical

measurement conditions. The resultant two-dimensional images and

corresponding line scans from the surfaces of the samples are shown in Figure

4-5. These images were acquired in water to determine the roughness of a

hydrated polymer surface, the same conditions used for bilayer formation. The

measured roughness of glass was found to be 0.128 nm, while the measured

roughness of PET was found to be 1.652 nm and Kapton, 0.743 nm. If surface

roughness is truly a key consideration in the formation of mobile supported

bilayers, the results of these measurements would suggest that a bilayer formed

on Kapton would be mobile, while one on PET would not.

Fluorescent Recovery After Photobleaching (FRAP)

Usefulness of a supported bilayer system in a sensing platform depends, in

part, on the mobility of the bilayer.106 This is important if a membrane-bound

biological channel is used as the sensing element, as bilayer mobility is often an

integral part of channel activity.'15 In order to test the mobility on the substrate

surfaces, FRAP experiments were performed using each of the vesicle solutions

described above on borosilicate glass, PET, and Kapton membrane samples.

By choosing two lipopolymers that differ only in molecular weight, the effect

of the conformation of the PEG on bilayer mobility can be addressed.130 Since

the conformation, either mushroom or brush, is dependent on lipopolymer

density,133 it was necessary to be sure that the lipopolymer density was such that

the bilayer was at its maximum mobility. In the case of PEG550, the onset of the

mushroom-to-brush transition commences at a mole fraction of 0.070.130, 134








Therefore, in a 5 mol% PEG5-PE bilayer, the PEG polymer exists in the

mushroom conformation. In the case of PEG2" however, the transition starts at

a mole fraction of 0.014.130' 134 As the polymer size increases, the mole fraction

of the mushroom-to-brush transition decreases. This means that, in bilayers

containing 5 mol% PEG2-PE, the PEG moiety exists in the brush configuration,

while the in 1 mol% PEG2-PE, the PEG remains in the mushroom

conformation.

Before bleach Immediately after bleach <5 min after bleach






0 0 0



1.0-
8

C 0.86-



S0.2 t=21.61.5s
0 D = 4.0 0.2 pm2Zs
z

0.0 100 200 300
Time (sec)

Figure 4-6. FRAP images and fluorescence recovery curve of a bilayer
containing 1 mol% PEG2 in egg-PC/Texas Red-PE on glass.

In experiments conducted on borosilicate glass, by Albertorio et al.,134 it has

been shown that, in the case of PEG550, increasing the mole fraction of the