mi .....:.,..§_.%.~ . ...1 .1 . , . . . . ., kwﬁdﬂiuﬁ

 

— l

1" LIBRARY
Michigan State a;
University

 

____I

 

This is to certify that the
dissertation entitled

DESIGN AND CHARACTERIZATION OF
NANOSTRUCTURED BIOMIMETIC INTERFACES
CONTAINING BILAYER LIPID MEMBRANES

presented by

SACHIN RAMANLAL JADHAV

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Chemical Engineering

 

 

 

 

WW ZOO/warm

 

Major Professor’s Signature

9/30/1009

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:lProj/Aoc&Pres/ClRC/Dateoue.indd

 

DESIGN AND CHARACTERIZATION OF
NAN OSTRUCTURED BIOMIMETIC INTERFACES
CONTAINING BILAYER LIPID MEMBRANES

By

Sachin Ramanlal J adhav

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Chemical Engineering

2009

ABSTRACT

DESIGN AND CHARACTERIZATION OF NANOSTRUCTURED BIOMIMETIC
INTERFACES CONTAINING BILAYER LIPID MEMBRANES

By
Sachin Rarnanlal J adhav

Cell membranes carry out several molecular recognition, communication,
transport, and catalytic functions to maintain cellular homeostasis. Various membrane
proteins and other biomolecules embedded in the cell membrane contribute speciﬁc
physiological functions and often require a lipid bilayer environment to express their
activities. Biomimetic interfaces containing artiﬁcial bilayer lipid membrane (BLM) with
embedded membrane proteins can effectively reproduce functions of biological cell
membranes in vitro. These interfaces have many potential applications in fundamental
and applied research. BLM-containing interfaces can be used in ﬁmctional proteomics
research to characterize novel membrane proteins, to investigate receptor-drug
interactions for high-throughput drug screening, and in development of biosensor systems

that exploit speciﬁc interactions between membrane proteins and analytes.

The central theme of this work is to develop robust, model membrane interfaces
that are easy to fabricate, directly addressable by various analytical techniques, and
suitable for advanced research and commercial applications. Three different BLM
interfaces, namely planar unsupported BLMs, supported BLMs (sBLM), and tethered
BLMs (tBLM), were fabricated and characterized using electrochemical techniques.
Planar BLM and tBLM interfaces were used for functional characterization of ion
channel membrane proteins. PorB class II (PorBII), a transmembrane pore-forming

protein from Neisseria meningitidis, was incorporated in a tBLM interface and the ion

transport mediated by porin was measured using electrochemical impedance
spectroscopy (E18). E18 study of PorBII porin incorporated in a tBLM interface showed
that the porin’s conductance decreases with increase in applied DC potential. The porin
conductance values were restored after returning to lower potentials. Parallel single
channel studies conducted using planar BLMs revealed that PorBII forms a trimeric pore
in BLM that undergoes a reversible, voltage-dependent closure at high transmembrane
potentials, corroborating tBLM ﬁndings. In addition to measuring proteins’ activities,
novel methods were developed to incorporate membrane proteins into an interface. A
tBLM interface was fabricated using membrane fractions derived from
Schizosaccharomyces pombe yeast cells. The approach will facilitate a direct
incorporation of recombinant membrane proteins overexpressed in S. pombe into BLM
interfaces without the need to separate the proteins from the membrane, purify the

proteins, and then deposit the resulting proteoliposomes.

A sBLM interface was developed to characterize interactions between polymeric
nanoparticles and cell membranes. This interface provided a rapid and convenient way to
screen nanoparticles’ ability to disrupt BLMs, thereby helping assess nanoparticles’
toxicity and suitability for commercial applications (e.g., gene delivery into cells).
Finally, BLM containing interfaces were formed on microelectrode arrays to provide a
cost-effective, hi gh-throughput platform that allows a rapid and simultaneous
electrochemical characterization of proteins for functional proteomics research or

nanoparticle characterization.

Copyright by
SACHIN RAMANLAL JADHAV

2009

DEDICATED

TO

MY FAMILY

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ IX
LIST OF FIGURES ............................................................................................................ X
NOMENCLATURE ........................................................................................................ XX
1. INTRODUCTION ....................................................................................................... 1
1.1. Overview and objectives ..................................................................................... 1
1.2. Dissertation Outline ............................................................................................. 4
1.3. Background .......................................................................................................... 4
1.3.1. Biomimetic interfaces containing bilayer lipid membranes ........................ 4

2.

1.3.2. Biomimetic interface for PorB class II porin ﬁ'om Neisseria meningitidis.6
1.3.3. Investigation of voltage dependent closure of PorBII porin using planar

BLM and tBLM interfaces .......................................................................................... 7
1.3.4. Fabrication of tBLMs using membrane fractions of Schizosaccharomyces
pombe yeast cells ......................................................................................................... 7
1.3.5. Interaction of nanoparticles with supported BLMs ..................................... 8
1.3.6. Integration of biomimetic interfaces with gold electrode arrays ................. 9
BIOMIMETIC INTERFACES CONTAINING BILAYER LIPID MEMBRANES 11
2.1. Abstract .............................................................................................................. 11
2.2. Introduction ....................................................................................................... 12
2.2.1. Patch clamp technique ............................................................................... 13
2.2.2. Planar unsupported BLMs ......................................................................... 14
2.2.3. Solid supported BLMs ............................................................................... 15
2.2.4. Electrochemical characterization of BLM ................................................. 17
2.2.4.1. Electrochemical impedance spectroscopy (EIS) ................................... 17
2.2.4.2. Cyclic voltammetry (CV) 19
2.3. Materials and methods ....................................................................................... 20
2.3.1. Chemicals .................................................................................................. 20
2.3.2. Planar bilayer lipid membrane measurements ........................................... 21
2.3.3. sBLM formation on glassy carbon electrodes ........................................... 21
2.3.4. Preparation of liposomes ........................................................................... 22
2.3.5. Fabrication of tethered bilayer lipid membrane on gold substrate ............ 22
2.3.6. Instrumentation .......................................................................................... 23
2.4. Results and discussion ....................................................................................... 24
2.4.1. Planar BLMs measurements of gramicidin activity .................................. 24
2.4.2. Characterization of sBLM interface .......................................................... 25
2.4.3. Fabrication of tBLM on gold electrode ..................................................... 27
2.4.3.1. Characterization of a tethering lipid monolayer .................................... 28
2.4.3.2. Characterization of a tBLM ................................................................... 29

vi

2.4.3.3. Characterization of valinomycin and gramicidin activity using tBLM
interface ......................................................................................................... 29

3. FUNCTIONAL CHARACTERIZATION OF PORB CLASS II PORIN FROM
NEISSERIA MENINGIHDIS USING A TETHERED BILAYER LIPID MEMBRANE 48

3.1. Abstract .............................................................................................................. 48
3.2. Introduction ....................................................................................................... 49
3.3. Materials and methods ....................................................................................... 51
3.3.1. Expression and puriﬁcation of PorBII porin ............................................. 51
3.3.2. Planar bilayer lipid membrane measurements ........................................... 53
3.3.3. Fabrication and characterization of tBLM ................................................. 53
3.4. Results and discussion ....................................................................................... 54
3.4.1. Measurement of PorBII activity using planar BLM .................................. 54
3.4.2. tBLM characterization and measurement of PorB II activity .................... 55
3.4.2.1 . Electrochemical impedance spectroscopy ............................................. 55
3.4.2.2. Cyclic voltammetry ............................................................................... 58

3.5. Conclusions ....................................................................................................... 59

4. VOLTAGE DEPENDENT CLOSURE OF PORB CLASS 11 PORIN FROM
NEISSERIA MENHVGIYYDIS INVESTIGATED USING ELECTROCHEMICAL

IMPEDANCE SPECTROSCOPY IN A TETHERED BLM INTERFACE ..................... 66
4.1. Abstract .............................................................................................................. 66
4.2. Introduction ....................................................................................................... 67
4.3. Materials and methods ....................................................................................... 69
4.4. Results and discussion ....................................................................................... 69

4.4.1. Voltage dependence of PorBII conductance in planar BLM ..................... 69
4.4.2. EIS study of a voltage dependent channel closure in a tBLM interface....72
4.4.3. Voltage dependent closure of PorB II in symmetric tBLM ....................... 74
4.5. Conclusions ....................................................................................................... 74

5. FABRICATION OF HIGHLY INSULATING TETHERED BILAYER LIPID
MEMBRANES USING YEAST CELL MEMBRANE FRACTIONS FOR

MEASURING ION CHANNEL ACTIVITY ................................................................... 82
5.1. Abstract .......................................................................................................... ‘....82
5.2. Introduction ....................................................................................................... 83
5.3. Materials and methods ....................................................................................... 84

5.3.1. Preparation of membrane fractions ............................................................ 84
5.3.2. Fabrication of tethered bilayer lipid membrane on gold surface ............... 85
5.3.3. Instrumentation .......................................................................................... 85
5.4. Results and discussion ....................................................................................... 88
5.4.1. Characterization of membrane fraction of yeast cells ............................... 88
5.4.2. Characterization of tBLM formation using QCMD .................................. 88
5.4.3. Fluorescence recovery after pattern photobleaching (FRAPP) study of
microsomal sBLM on glass ....................................................................................... 91
5.4.4. Electrochemical characterization of tBLM ................................................ 94
5.5. Conclusions ....................................................................................................... 96

vii

6. ELECTROCHEMICAL CHARACTERIZATION OF INTERACTIONS
BETWEEN NANOPARTICLES AND SUPPORTED BILAYER LIPID MEMBRANES

 

 

... . . . ................................................................................................................................. 1 1 l
6.1. Abstract ............................................................................................................ 111
6.2. Introduction ..................................................................................................... 112
6.3. Methods and instrumentation .......................................................................... 114

6.3.1 . Method ..................................................................................................... 114
6.3.2. Instrumentation ........................................................................................ l 15
6.4. Results and discussion ..................................................................................... 115
6.4.1. Interface formation .................................................................................. 1 15
6.4.2. Interaction of PAMAM dendrimers with sBLM ..................................... 115
6.4.3. Interaction of PEG NPs with sBLM ........................................................ 118
6.5. Conclusions ..................................................................................................... 120

7. INTEGRATION OF BIOMIMIETIC INTERFACES WITH GOLD ELECTRODE

ARRAYS ......................................................................................................................... 134
7.1. Abstract ............................................................................................................ 134
7.2. Introduction ..................................................................................................... 135
7.3. Methods ........................................................................................................... 136

7.3.1 . Microelectrode array fabrication -- ............................... 136
7.3.2. Deposition of insulation layer over microelectrode arrays ...................... 137
7.3.3. Fabrication of tBLM on microelectrode arrays ........ -. .............................. 138
7.3.4. Fabrication of NEST sensor ........... - .................................... 138
7.3.5. NEST solution assay ................................................................................ 139
7.4. Results and discussion ..................................................................................... 139
7.4.1. tBLM on microelectrode arrays on glass ................................................. 139
7.4.2. tBLM on microelectrode arrays on silica substrates ............................... 141
7.4.3. NEST sensor on microelectrode arrays ................................................... 144
7.4.4. Effect of temperature on biomimetic interfaces ...................................... 147
7.4.4.1. Temperature dependence of tBLM impedance ................................... 147
7.4.4.2. Temperature dependence of NEST activity ......................................... 148

7.5. Conclusions ..................................................................................................... 149
8. REFERENCES ....................................................................................................... i .167

viii

 

LIST OF TABLES

Table 6.1: Characteristics of PAMAM dendrimers, along with the parameters showing
their effect on sBLM ............................................................................................ 122

Table 6.2: Characterization of PEG NPs’ interaction with sBLM studied using CV.
[Electrode area of GCE is 7.064 mmz] ................................................................ 123

Table 7.1: The properties of impedance extraction and digitization circuit (IDC) that can
be installed under each element of a sensor array. CMOS: complementary metal—
oxide—semiconductor. .......................................................................................... 150

Table 7.2: The absolute values of electrochemical properties for tBLMs deposited on
Design Si-I microelectrodes on silicon ................................................................ 151

Table 7.3: The area normalized values of electrochemical properties for tBLMs
deposited on Design Si-I microelectrodes on silicon. ......................................... 152

ix

LIST OF FIGURES

Figure 2.1: Planar BLM set-up. Lipids were painted across a small aperture in a partition
that separates two compartments of BLM chamber. The ionic current mediated by
ion transport across BLM through the ion channel proteins was measured by a
pair of Ag/AgCl reference electrodes. ................................................................... 32

Figure 2.2: Supported BLM interfaces. (A) sBLM deposited on a silica or glass surface.
A thin layer of water separates the BLM from the surface of substrate. (B) sBLM
can be deposited on gold surface by fornring a SAM of alkanethiols followed by
depositing the upper leaﬂet of sBLM using liposome vesicles. (C) Charged
polymers and polyelectrolytes have been used as polymeric cushions to separate
sBLM from electrode surface and create ion reservoir. (D) Freely suspended
BLMs are deposited on anodized alumina membranes to create sBLM architecture
that mimics planar BLMs. [(Richter, R. P. et al., 2006)] ...................................... 33

Figure 2.3: A tethered BLM (tBLM) architecture showing the decoupling of BLM from '
electrode surface using hydrophilic spacer molecule creating an ion reservoir
beneath BLM. The ion reservoir also provides space to accommodate the bulky
extramembraneous domains of transmembrane proteins. tBLM offers a stable
BLM interface that is suitable for studying ion channel proteins using
electrochemical impedance spectroscopy. ............................................................. 34

Figure 2.4: A scheme for fabrication of a tBLM. A self assembled monolayer (SAM) of
tethering lipids was ﬁrst deposited on gold electrodes. After SAM formation,
liposomes made of DOPC lipids were used to deposit the upper leaﬂet of tBLM
using vesicle ﬁrsion technique. tBLM formation was characterized using EIS and
CV. Ion channel protein was then be added to the electrolyte solution to allow its
incorporation in tBLM. [Figure not to scale] ........................................................ 35

Figure 2.5: A Randles equivalent circuit model was ﬁtted to tBLM impedance data to
calculate the values of electrochemical properties. R3 represents the solution
resistance, CM represents the capacitance of the membrane, RM represents the

resistance of membrane and CDL represents the double-layer capacitance, which
includes the Hemholtz capacitance ........................................................................ 36

Figure 2.6: The ion transport activity of gramicidin peptides measured using a planar
BLM interface. DPhytPC lipids were painted across 250 um aperture. (A) Single
channel currents corresponding to a conductance value of 25 pS were obtained.
Incorporation of multiple channels proportionally increased the current response.
(B) Single channel study was conducted to obtain the current-potential (I-V)
relationship for grarrricidin peptides. I-V curve was linear in the range of +110
mV to -110 mV. Experimental conditions: Both bilayer compartments had 1 mL

 

of 20 mM HEPES buffer (pH 7.4) containing 1 M KCl; gramicidin concentration
in cis chamber was 1 pg/mL. ................................................................................. 37

Figure 2.7: Cyclic voltammograrn for bare GCE (curve 1) and for sBLM deposited on
GCE (curve 2). Well deﬁned oxidation and reduction peaks were obtained at bare
GCE surface. The peak currents were completely diminished after sBLM
formation suggesting that sBLM did not have large pinhole defects. ................... 38

Figure 2.8: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for bare GCE (triangles) and for sBLM deposited on
GCE (squares). An equivalent circuit shown in Figure 2.9 was ﬁtted to impedance
data. The solid lines passing through the data points represent the model ﬁt. ...... 39

Figure 2.9: An equivalent circuit used to model impedance data obtained for bare GCE
and sBLM on GCE. R3 represents the resistance of solution, CDEFECT is the
double layer capacitance of GCE at the defect sites in sBLM, CM is the

membrane capacitance, RM represents membrane resistance, and Zw is the
Warburg impedance. .............................................................................................. 40

Figure 2.10: (A) Impedance (Log Z) data, and (B) phase angle data were plotted as a
ﬁmction of Log ﬁequency for DPPTE tethering lipid monolayer (triangles) and
tBLM (squares) formed with DOPC liposomes in 100 mM NaCl solution. An
equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data The solid
lines passing through the data points represent the model ﬁt. ............................... 41

Figure 2.11: Cyclic voltarnmogram for bare gold electrode (curve 1) and for DPPTE-
DOPC tBLM deposited on gold surface (curve 2). Well deﬁned oxidation and
reduction peaks were obtained at bare gold surface. The peak currents were
completely diminished after tBLM formation suggesting that sBLM did not have
large pinhole defects. ............................................................................................. 42

Figure 2.12: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-
DOPC tBLM after valinomycin incorporation (triangles). tBLM impedance was
taken in 100 mM NaCl. Valinomycin was added to give a ﬁnal concentration of 5
M in an electrolyte solution. After valinomycin incorporation, 100 mM KCl was
added. An equivalent circuit shown in Figure 2.5 was ﬁtted to impedance data.
The solid lines passing through the data points represent the model ﬁt. Membrane
resistance decreased signiﬁcantly due to the ion transport through the tBLM,
whereas there was no signiﬁcant change in membrane capacitance. The solution
resistance decreased due to increase in the ionic strength of an electrolyte solution
after 100 mM KCl addition. .................................................................................. 43

Figure 2.13: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-
DOPC tBLM after gramicidin incorporation (triangles) in 100 mM NaCl. An

xi

equivalent circuit shown in Figure 2.5 was ﬁtted to impedance data. The solid
lines passing through the data points represent the model ﬁt. Gramicidin was
added to give a ﬁnal concentration of 1 M protein (monomer) in an electrolyte
solution. Membrane resistance decreased signiﬁcantly due to the ion transport
through the tBLM, whereas there was no signiﬁcant change in membrane
capacitance ............................................................................................................. 44

Figure 2.14: Cyclic voltarnmogram for tBLM (curve 1) and for DPPTE-DOPC tBLM
aﬁer gramicidin incorporation (curve 2). The redox peaks were absent in both
voltammograms whereas there was an increase in charging current after the
addition of granricidin suggesting the passage of ions through tBLM by
granricidin while excluding the passage of ferricyanide ions through tBLM. ...... 45

Figure 2.15: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM in 100 mM NaCl (squares),
for DPPTE-DOPC tBLM after gramicidin incorporation (triangles) in 100 mM
NaCl, and for gramicidin-incorporated tBLM after replacing the electrolyte
solution to 100 mM tetrarnethylarnmonium chloride (open diamonds). An
equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data. The solid
lines passing through the data points represent the model ﬁt. Gramicidin was
added to give a ﬁnal concentration of 1 M protein (monomer) in an electrolyte
solution. ................................................................................................................. 46

Figure 2.16: Impedance (Log Z) data were plotted as a function of Log frequency for
DPPTE-DOPC tBLM and for DPPTE-DOPC tBLM after gramicidin
incorporation at different protein concentration and ionic strength of NaCl
solution. Squares: tBLM; triangles: tBLM containing 10 nM gramicidin in 10 mM
NaCl solution; circles: tBLM containing 10 nM gramicidin in 100 mM NaCl
solution; stars: tBLM containing 50 nM gramicidin in 100 mM NaCl solution;
diamonds: tBLM containing 100 nM gramicidin in 100 mM NaCl solution. The
membrane resistance decreased with increase in the ionic strength of solution and
the protein concentration. ...................................................................................... 47

Figure 3.1: pET17bNT-PorBII construct used to transform E. colz'BL21 cells for
expressing PorBII. The PCR product of Neisseria meningitidis PorBII gene delta
1 to 62 was cloned into pGEM-Teasy, digested with NdeI and BamHI, and ligated
into pET17b (Novagen), which was engineered with an insert of an N-terminal
hexa histidine tag and an rTev protease cleavage site, yielding pET17bNT-PorBII.
............................................................................................................................... 61

Figure 3.2: (A) Recordings of DPhytPC/n-decane BLM showing PorBII channel
' insertions. (B) Histogram of channel conductance of PorBII in DPhytPC BLM,
obtained from 1000 channel insertion recordings, showing the probability of
occurrence of particular conductance unit. The conductance values were pooled in

bins of 0.2 nS. Experimental conditions: Both bilayer compartments had 1 mL of

20 mM HEPES buffer (pH 7.4) containing 1 M KCl; PorBII concentration in cis
chamber was 100 pg/mL; Applied potential was +10 mV. ................................... 62

xii

Figure 3.3: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-
DOPC tBLM after PorBII incorporation (triangles) in 100 mM NaCl. An
equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data. The solid
lines passing through the data points represent the model ﬁt. PorBII was added to
give a ﬁnal concentration of 1 M protein (monomer) in an electrolyte solution.
Membrane resistance decreased signiﬁcantly due to the ion transport through the
tBLM, whereas there was no signiﬁcant change in membrane capacitance. ........ 63

Figure 3.4: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log ﬁ'equency for DPPTE—DOPC tBLM (squares) and for DPPTE-
DOPC tBLM after PorBII incorporation (triangles) in 50 mM calcium chloride.
An equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data. The
solid lines passing through the data points represent the model ﬁt. PorBII was
added to give a ﬁnal concentration of 1 M protein (monomer) in an electrolyte
solution. ................................................................................................................. 64

 

Figure 3.5: CV curves for a bare gold electrode coated with a tethering lipid monolayer
(curve 1), a tBLM formed by deposition and rupture of DOPC liposomes on top
of the DPPTE tethering lipid monolayer (curve 2), and the DPPTE-DOPC tBLM
after PorBII incorporation (curve 3). The redox peaks were absent after bilayer
formation but were observed after PorBII incorporation, suggesting PorBII-
mediated transport of ferricyanide ions through tBLM. ........................................ 65

Figure 4.1: Unitary response of recombinant PorBII porin incorporated in DPhytPC lipid
bilayer. The magnitude and length of applied potential is indicated below the
current recording. Single channel insertion of 4.2 ms was observed at +10 mV.
Increasing the potential to +100 mV closes the porin in three discrete steps,
indicative of closure of each conducting monomer. The porin channel was
reopened by reducing the potential to +10 mV, indicating that the porin closure is
completely reversible ............................................................................................. 76

Figure 4.2: Normalized single channel conductance (Ratio of conductance at respective

potential [G] to the conductance at +10 mV [GIOmV] plotted as a function of
applied potential. Five separate experiments were conducted at each potential.
Lipids used- DPhytPC, PorBII concentration- 1 pg/mL, other experimental
conditions were same as described in Figure 3.2. ................................................. 77

Figure 4.3: Open probability calculated from Equation 4.2 using experimentally obtained

It and V0 values. Open probability, shown using solid line, is overlaid on the
voltage dependent porin conductance data. ........................................................... 78

Figure 4.4: Impedance (Log Z) data were plotted as a function of Log frequency for
DPPTE-DOPC tBLM at 200 mV DC potential (squares), for DPPTE-DOPC
tBLM after PorBII incorporation at 200 mV DC potential (triangles), and for
DPPTE-DOPC tBLM containing PorBII after toggling the DC potential to 0 mV
(diamonds) in 100 mM NaCl. An equivalent circuit shown in Figure 2.5 was ﬁtted

xiii

to the impedance data. The solid lines passing through the data points represent
the model ﬁt. PorBII was added to give a ﬁnal concentration of 1 uM protein
(monomer) in an electrolyte solution ..................................................................... 79

Figure 4.5: Fractional NRM plotted as a function of applied DC potential used in EIS
for a PorBII incorporated DPPTE-DOPC tBLM (diamonds) and gramicidin
incorporated DPPTE-DOPC tBLM (squares). ...................................................... 80

Figure 4.6: (A) Impedance (Log Z) data were plotted as a function of Log ﬁequency for
DPPTE-DPPTE tBLM at 0 mV DC potential (squares) and for DPPTE-DPPTE
tBLM after PorBII incorporation at 0 mV DC potential (triangles). (B)
Impedance (Log Z) data were plotted as a function of Log ﬁ'equency for DPPTE-
DPPTE tBLM at 200 mV DC potential (squares) and for DPPTE-DPPTE tBLM
after PorBII incorporation at 200 mV DC potential (triangles) in 100 mM NaCl.
An equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data. The
solid lines passing through the data points represent the model ﬁt. PorBII was
added to give a ﬁnal concentration of 1 uM protein (monomer) in an electrolyte
solution. ................................................................................................................. 81

Figure 5.1: Fabrication of tBLM using liposome or yeast membrane fractions. A self
assembled monolayer of tethering lipid was deposited on the gold substrate, and
the liposomes or microsomes were added to deposit the upper tBLM leaﬂet.
Gramicidin was incorporated into the tBLM, and gramicidin-mediated ion
transport across the tBLM was monitored using EIS. [Figure not to scale] .......... 98

Figure 5.2: Schematic diagram of the FRAPP experimental setup. sBLM: supported
bilayer lipid membrane; PMT: photomultiplier tube. [Figure not to scale] .......... 99

Figure 5.3: Transmission electron microscopy (TEM) image of (A) DOPC liposomes,
and (B) membrane fractions of S. pombe yeast cells. Liposome and membrane
ﬁactions were stained using 1% uranyl acetate solution. The size of liposome and
membrane fractions was well below 200 nm which would be necessary for facile
rupture of these vesicles to yield planar bilayer membranes. .............................. 100

Figure 5.4: Quartz crystal rnicrobalance study for formation of tBLM using liposome
and tBLM using yeast microsomes. (A) Liposome deposition directly on the silica

surface is shown in Curve 1 (A f data) and Curve la (ADissipation data).
Liposome deposition on the tethering lipid SAM is shown in Curve 2 (Af data)

and Curve 2a (ADiSSipation data). (B) Microsome deposition directly on the
silica surface is shown in Curve 1 (A f data) and Curve 1a (ADissipation data).
Microsome deposition on the tethering lipid SAM is shown in Curve 2 (Af data)
and Curve 2a (ADissipation data). ................................................................... 101

Figure 5.5: Fluorescence recovery curves for sBLM on glass formed using POPC
liposomes incorporated with 2% NBD-PC obtained using FRAPP. ................... 102

xiv

 

 

Figure 5.6: Fluorescence recovery curves for sBLM on glass formed using liposomes
obtained using lipids extracted from microsomes and mixed with 2% NBD-PC.
............................................................................................................................. 103

Figure 5.7: Fluorescence recovery curves for sBLM on glass formed using microsomes
incorporated with 2% NBD-PC. .......................................................................... 104

Figure 5.8: Diffusion coefﬁcient and mobile fraction data obtained using FRAPP were
plotted for different sBLMs deposited on glass substrate. .................................. 105

Figure 5.9: Comparison of tBLM formed with DOPC liposome (A), and with membrane
fractions of yeast cells (B) over tethering lipid SAM using EIS. Impedance data
(Log Z) were plotted as a function of Log ﬁ'equency for tethering lipid monolayer
(triangles) and tBLM (squares). An equivalent circuit shown in Figure 5.8 was
used to ﬁt impedance data. The solid lines passing through the data points
represent model ﬁt. The resistance of membrane increases whereas the
capacitance of membrane decreases after the formation of bilayer membrane.

[Electrode area is 0.2 cm2] .................................................................................. 106

Figure 5.10: A Randles equivalent circuit was ﬁtted to the impedance data to calculate
the values of electrochemical characteristics. Rs represent the resistance of
solution, CM represents the capacitance of membrane, RM represents the

resistance of membrane and CDL represents the capacitance of double layer that
includes the Hemholtz capacitance ...................................................................... 107

Figure 5.11: Impedance spectrum for tBLM recorded at different intervals after addition
of nricrosomes. The order of curves from bottom to top is- tethering lipid
monolayer, tBLM at 30 min, tBLM at 1 h, tBLM at 2 h, tBLM at 3 h, tBLM at 5
h, tBLM at 8 h, tBLM at 22 h and tBLM at 24 h. The membrane resistance
increased whereas the membrane capacitance decreased with time of incubation.
The initial rate of bilayer formation was higher, which decreases with time. The
impedance spectrum for tBLM at 22 h, and 24 h were same indicating that the
tBLM formation is complete. The electrolyte solution was replaced after the
tBLM formation. The impedance spectrmn of tBLM did not change after
replacing the electrolyte solution with fresh 100 mM NaCl solution. [Electrode

area is 0.2 cm ] .................................................................................................... 108

Figure 5.12: (A) Cyclic voltarnmograrn of bare gold (curve 1) was plotted along with the
voltarnmograrn obtained after tBLM fabrication (curve 2). The redox peaks were
subdued after the formation of tBLM. (B) Cyclic voltarnmograrn for tBLM
formed using membrane ﬁactions of yeast cells (curve 1), and microsomal tBLM
after gramicidin incorporation (curve 2). The redox peaks were absent in both
tBLM voltarnmograrns whereas there was an increase in charging current after the
addition of gramicidin suggesting the passage of ions through tBLM by

XV

grarrricidin while excluding the passage of ferricyanide ions through tBLM.
[Electrode area is 0.2 cm2] .................................................................................. 109

Figure 5.13: Measurement of gramicidin ion channel activity using EIS. (A) tBLM was
fabricated by fusion of DOPC liposome over tethering lipid monolayer and (B)
tBLM was fabricated by fusion of membrane fractions of yeast cell over tethering
lipid monolayer. Impedance (Log Z) data for tBLM (squares) and tBLM after
gramicidin incorporation (triangles) were plotted as function of Log ﬁ'equency.
Gramicidin was added as an ethanol solution to give a ﬁnal concentration of 1
uM. Membrane resistance drops due to the passage of sodium ions through tBLM.

[Electrode area is 0.2 cm2] .................................................................................. 110

Figure 6.1: sBLM formation on GCE. GCE was ﬁrst oxidized at +1.5 V to impart
negative charge. DOPC sBLM was self assembled on oxidized GCE by
depositing DOPC lipids, solubilized in chloroform, on GCE and immediately
transferring the interface in an aqueous electrolyte solution. sBLM stability was
tested using EIS and CV. Nanoparticles, when added in required concentration,
disrupted sBLM and exposed electrode area which can be calculated from the
redox peaks obtained using CV. In one of the proposed mechanisms, dendrimers
form so called ‘dendrimer ﬁlled vesicles’ after their interaction with sBLM
(Mecke, A. et 01., 2005b). .................................................................................... 124

Figure 6.2: Cyclic voltarnmograrns showing the effect of lower generation PAMAM
dendrimers on sBLM: PAMAM GZ (curve 2) and PAMAM G3 (curve 3).

[Electrode area: 7.064 mmz] ............................................................................... 125

Figure 6.3: Cyclic voltarnmograrns showing the effect of higher generation PAMAM
dendrimers on sBLM: PAMAM G4 (curve 4), PAMAM GS (curve 5), PAMAM
G6 (curve 6), and PAMAM G7 (curve 7). Curve 1 represents the sBLM before

dendrimer addition. [Electrode area: 7 .064 m2] ............................................... 126

Figure 6.4: Impedance (Log Z) data were plotted as a function of Log frequency for a
bare GCE surface (triangles), sBLM (squares), and sBLM after interaction with
PAMAM G5 dendrimers (diamonds). An equivalent circuit shown in Figure 2.9
was used to ﬁt the impedance data. The disruption of sBLM by PAMAM G5
increased the membrane capacitance and decreased the membrane resistance.

[Electrode area: 7 .064 m2] ............................................................................... 127

Figure 6.5: Structures of new PEG NPs synthesized using ‘click’ chemistry by Dr. Baker
group in Department of Chemistry at MSU. The basic backbone of the structure is

shown above the table. The group —R2 is hydrogen (-H) in all NPs. .................. 128

Figure 6.6: Cyclic voltammograms showing the effect of uncharged PEG NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG uncharged NPs (curve 2).

[Electrode area: 7.064 mmz] ............................................................................... 129

xvi

Figure 6.7: Cyclic voltammograrns showing the effect of PEG amine NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG amine NPs (curve 2).

[Electrode area: 7.064 mm2] ............................................................................... 130

Figure 6.8: Cyclic voltarmnograrns showing the effect of PEG acid NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG acid NPs (curve 2).

[Electrode area: 7 .064 m2] ............................................................................... 131

Figure 6.9: Cyclic voltarnmograrns showing the effect of PEG guanidine NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG guanidine NPs (curve 2).

[Electrode area: 7.064 mmz] ............................................................................... 132

Figure 6.10: Impedance (Log Z) data were plotted as a function of Log frequency for a
sBLM on GCE (squares), and sBLM after interaction with PEG guanidine NPs
(triangles). An equivalent circuit shown in Figure 2.9 was used to ﬁt the
impedance data. The disruption of sBLM by PEG guanidine NPs increased the
membrane capacitance and decreased the membrane resistance. [Electrode area:

7.064 mmz] .......................................................................................................... 133

Figure 7.1: (A) First generation Design Glass-I microelectrode array deposited on glass
substrate. SU8 photoresist was used as insulation layer to deﬁne the area of ten
gold electrode elements. (B) Second generation Design Si-I microelectrode array
deposited on silicon substrate. Silicon nitride was used as insulation layer to
deﬁne the electrode size. (C) Design Si-II microelectrode array deposited on
silicon substrate. Silicon nitride was used as insulation layer to deﬁne the
electrode size. ...................................................................................................... 153

Figure 7.2: The experimental set-up for E18 measurements of tBLM using IDC (Yang,
C. et al., 2009). .................................................................................................... 154

Figure 7.3: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log ﬁ'equency for tethering lipid monolayer (triangles) deposited on 3
mm diameter electrodes on Design Glass-I rrricroelectrode array, tBLM formed
using DOPC liposomes (squares), and for tBLM after granricidin incorporation
(diamonds) in 100 mM NaCl. Gramicidin was added to give a ﬁnal concentration
of 1 pM protein (monomer) in an electrolyte solution. ....................................... 155

Figure 7.4: (A) Impedance (Log Z) data were plotted as a function of Log frequency for
tBLM (squares) deposited on 500 um diameter electrodes on Design Si-I
nricroelectrode array, and for tBLM after gramicidin incorporation (triangles) in
100 mM NaCl. (B) Impedance (Log Z) data were plotted as a function of Log
Frequency for tBLM (squares) deposited on 250 um diameter electrodes on
Design Si-I microelectrode array, and for tBLM after gramicidin incorporation
(triangles) in 100 mM NaCl. Gramicidin was added to give a ﬁnal concentration
of 1 uM protein (monomer) in an electrolyte solution. An equivalent circuit

xvii

shown in Figure 2.5 was ﬁtted to the impedance data. The solid lines passing
through the data points represent the model ﬁt. .................................................. 156

Figure 7.5: tBLM’s area normalized membrane resistance before (bar with no vertical
stripes) and after gramicidin incorporation (bar with vertical stripes) for tBLMs
deposited on various size microelectrodes. The sealing nature of tBLMs was
conﬁrmed by the measurement of gramicidin-mediated ion conduction across
tBLM. .................................................................................................................. 157

Figure 7.6: Current density of catechol sensor as a function of catechol concentration
(conc.) obtained using (A) constant potential arnperometry under stirring, and (B)
using cyclic voltammetry ..................................................................................... 158

Figure 7.7: Charge transfer resistance (RCI‘) obtained for catechol sensor as a function
of catechol concentration. .................................................................................... 159

Figure 7.8: Current density of NEST sensor deposited on macroelectrode (5 mm in
diameter) as a function of phenyl valerate concentration (conc.) obtained using
(A) constant potential arnperometry under stirring, and (B) using cyclic
voltammetry. ........................................................................................................ 160

Figure 7.9: An equivalent circuit used to model impedance data from NEST sensor
experiments. R3 represents the resistance of solution, CDL is the double layer

capacitance, RM represents membrane resistance, and Zw is the Warburg
impedance. ........................................................................................................... 161

Figure 7.10: (A) Cyclic voltarnmograrn obtained for NEST sensor on 500 um
nricroelectrode after addition on 10 M phenyl valerate (curve 1), and after
addition of 1 mM PMSF (curve 2). (B) Impedance data (Log Z) were plotted as a
function of Log frequency for NEST sensor on 500 um nricroelectrode after
addition on 10 uM phenyl valerate (squares), and after addition of 1 mM PMSF
(triangles). An equivalent circuit shown in Figure 7.9 was used to ﬁt impedance
data. The solid lines passing through the datapoints represents the ﬁt .............. 1 62

Figure 7.11: (A) Membrane resistance of DPPTE-DPPC tBLM was plotted as a function
of temperature. Membrane resistance values stabilize after 40°C, which
corresponds to the phase transition temperature of DPPC. (B) Admittance data
(Log Y) of DPPTE-DPPC tBLM plotted as a function of l/temperature to obtain
activation energy .................................................................................................. 163

Figure 7.12: Impedance data (Log Z) was plotted as a function of Log frequency for
DPPTE-DPPC tBLM at 50°C (squares), and after addition of 1 uM gramicidin at

50°C (triangles). [Electrode area is 0.2 cm2] ....................................................... 164

xviii

Figure 7.13: (A) NEST speciﬁc activity obtained using solution assay as a function of
temperature. (B) Log(N EST speciﬁc activity) was plotted as a function of
1/temperature to obtain activation energy using solution assay of NEST ........... 165

Figure 7.14: (A) Current density obtained using electrochemical sensor as a function of
temperature. (B) Log(Current density) was plotted as a function of l/temperature
to obtain activation energy using electrochemical sensor of NEST. ................... 166

xix

NOMENCLATURE

Chemicals

Ag/AgCl Silver/silver chloride

ATP Adenosine triphosphate

C8E4 Tetraoxyethylene mono-n—octylether

CaC12 Calcium chloride

DMPC 1,2-dimyristoyl-sn-glycero—3-phosphocholine

DOPC 1,2-dioleoyl-sn-glycero-3-phosphocholine

DPhytPC 1 ,2 -diphytanoyl—sn-glycero-3 -phosphocholine

DPPTE 1 ,2-dipalmitoyl-sn-glycero-3-phosphothioethanol

DT'I‘ Dithiothreitol

EDTA Ehtylenediamine tetraacetic acid

GCE Glassy carbon electrode

HEPES 4-(2-hydroxyethy1)-1-piperazineethanesulfonic acid

IPTG Isopropyl-B-D-thiogalactopyranoside

KCl Potassium chloride

MgC12 Magnesium chloride

NBD-PC l-oleoyl-2-[6—[(7-nitro-2-l ,3-benzoxadiazol-4~yl)amino]hexanoyl]-sn-
glycero-3 -Phosphocholine

NEST A catalytically active ﬁagment of NTE

XX

NHS

NP

NTA

NTE

OP

PAH

PAMAM

PDAC

PEG

PLL

PMSF

POPC

PorBII

PorBIII

SPS

TMAC

Symbols

A

CBULK

N-hydroxysuccinimide

Nanoparticles

Nitrilotriacetic acid

Neuropathy target esterase

Organophosphorus

Poly(allylamine hydrochloride)
Polyarrridoarnine
Poly(diallyldimethylammonium chloride)
Polyethylene glycol

Poly-L-lysine

Phenylmethanesulfonyl ﬂuoride

l-palimotoyl, 2-oleoyl -sn-glycero-3-phosphocholine
PorB class II porin ﬁom Neisseria meningitidis
PorB class III porin

Sulfonated polystyrene

Tetramethylammoniurn chloride

Electrode area

Stripe periodicity

Bulk concentration of redox species

f(0)

f(t)

ip

Double layer capacitance

Capacitance of defects in sBLM

Faradaic capacitance

Membrane capacitance

Constant phase element

Diffusion coefﬁcient

Faradays constant

Pre-bleach ﬂuorescence intensity (t < 0)

Post-bleach ﬂuorescence intensity (t > O)

Conductance

Maximum open conductance

Voltage independent conductance

Charging current

Current amplitude

Peak faradaic current

Mobile fraction of sBLM

xxii

NRM

Number of electrons transferred

Normalized membrane resistance

Charge transferred during oxidation/reduction
Gas constant

Charge transfer resistance

Membrane resistance

Solution resistance

Time

Temperature

Scan rate

Applied potential

Voltage at which the channel has equal probability of being in open or

closed state

Admittance

Impedance

Imaginary part of impedance

Real part of impedance

xxiii

Zw Warburg impedance

9M Phase delay

CO Frequency

Terminologies

BLM Bilayer lipid membrane

LBL Layer-by-layer

SAM Self assembled monolayer

sBLM Supported BLM

tBLM Tethered BLM

Instrumentation

AF M Atomic force microscopy

CV Cyclic voltammetry

CHI CH Instruments

CPA Constant potential arnperometry

DLS Dynamic light scattering

EIS Electrochemical impedance spectroscopy
FRAPP Fluorescence recovery after pattern photobleaching
IDC Impedance extraction and digitization circuit
J EOL Japan electronic optics laboratories

QCM Quartz crystal microbalance

xxiv

QCM-D Quartz crystal microbalance with dissipation monitoring

TEM Transmission electron microscopy

XXV

1. INTRODUCTION

1.1. Overview and objectives

The cell membrane carries out many molecular recognition, communication,
transport, and catalytic functions needed to maintain cellular homeostasis. The primary
structural component of a cell membrane, the bilayer lipid membrane (BLM), is formed
by spontaneous self assembly of arnphiphilic lipid molecules. Various membrane proteins
and other biomolecules embedded in the lipid bilayer contribute speciﬁc physiological
functions and often require a lipid bilayer environment to express their activities.
Artiﬁcial BLMs, formed using synthetic arnphiphilic lipids, can effectively— nrimic the
structure and function of biological cell membranes, and thus allow membrane processes
to be studied in a well-deﬁned system. For example, a membrane protein can be
incorporated into a BLM of known lipid composition and the protein’s functional
properties can be characterized. These interfaces offer a powerful research tool for
carrying out biophysical investigations, as components can be added as desired to better
mimic the natural cellular environment. The ability of different macromolecules to
interact and change the structure of cell membranes can also be studied using BLM
interfaces. BLM-containing interfaces can be used in functional proteomics research to
characterize novel membrane proteins, develop membrane protein interfaces for hi gh-
throughput drug screening studies, to investigate receptor-drug interactions, for
investigating interactions between nanoparticles and cell membranes, and in developing
biosensor systems that exploit speciﬁc interactions between membrane proteins and

analytes.

 

Development of biosensors based on membrane proteins poses several challenges.
Membrane proteins are difﬁcult to purify, often less stable than soluble proteins and may
need a lipid bilayer environment to express their activities. In some cases, special
methods are required for reconstituting membrane proteins in artiﬁcial BLMs. Hence,
there is a need for improved experimental tools that will allow the membrane proteins to
be reconstituted in BLMs and the proteins’ activities to be simultaneously expressed and
measured in vitro. These tools will allow a functional characterization of membrane
proteins and the effect of different environmental variables on protein activity to be

studied.

BLM-containing interfaces can be used to study the interactions of
macromolecules, such as charged nanoparticles (NPs), with cell membranes.
Nanoparticles have proven to be of great value in many commercial applications,
including their use as gene and drug delivery agents. Charged polymeric NPs, in
particular, are highly effective in disrupting lipid bilayers, thereby mediating the
internalization of molecular cargos into the host cell. This property of NPs raises a
concern that NP-mediated defect formation in cell membranes might lead to cytotoxicity.
BLM interfaces provide a rapid and convenient way to screen NPs for their interactions
with lipid bilayers, thereby helping assess nanoparticles’ toxicity and suitability for
commercial applications.

Electrochemical techniques offer a powerful and convenient way of monitoring
BLM-related phenomena. Electrochemical data can be acquired and analyzed online to

provide quantitative and meaningful information in real time about bilayer-associated

physical processes, including molecular function of membrane proteins or BLM-NP

interactions.

The objective of the present research was to develop robust, dynamic model
membrane interfaces that have several desirable properties: (1) can be deposited using
molecular self assembly, (2) are easily analyzed by analytical techniques, and (3) are
suitable for carrying out membrane protein assays and other biophysical studies. Toward
this end, the proposed research focused on the following research tasks: (1) develop a
tethered BLM (tBLM) interface for measuring the ion transport activity of PorB class II
(PorBII) outer membrane protein from Neisseria meningitidis, (2) investigate the voltage
dependence of PorBII activity in tethered and planar BLM interface, (3) develop high
impedance tBLMs using the membrane fractions, (4) study the effect of an NP’s physical
properties on its interaction with supported BLMs (sBLM), and (5) adapt BLM interfaces
to an array platform designed for hi gh-throughput electrochemical characterization of

proteins.

This research focused on development of a tBLM interface for PorBII porin that
would facilitate the design of biosensors for ion transporting membrane proteins, and
would provide a stable BLM system for functional characterization of membrane
proteins. For example, it will allow study of voltage-gated channel proteins using
impedance spectroscopy. Depositing tBLMs using membrane fractions will allow a facile
incorporation of membrane proteins in BLM interfaces. Potential gene delivery agents
can be screened using sBLM interface, while also providing understanding of membrane-
nanoparticle interactions. sBLM interface will also facilitate the design of novel gene or

drug delivery agents. Finally, integrating BLM interfaces with array microsystem will

provide 3 hi gh-throughput platform for studying various membrane related processes.
While each chapter in this dissertation focuses on unique issue, the underlying theme is
designing improved biomimetic architectures for studying cell membrane processes and
developing novel interfaces that express protein activities. The interfaces were designed
such that they are easily adaptable to high-throughput screening platforms and will

facilitate development of other protein biosensors.
1.2. Dissertation Outline

Chapter 2 of this dissertation presents the architectures and fabrication of different
BLM-containing interfaces. The use of planar unsupported BLM and tethered BLM
(tBLM) interfaces to measure activities of small ion transporting molecules is discussed.
In Chapter 3, a tBLM-containing biomimetic interface for a transmembrane pore-forming
membrane protein, PorB class II (PorBII) porin ﬁ'om Neisseria meningitidis, is presented.
Chapter 4 presents the ﬁmctional characterization of voltage dependent conductance
properties of PorBII using planar BLM and tBLM interfaces. In Chapter 5, fabrication of
tBLM interface using the membrane ﬁ'actions from Schizosaccharomyces pombe yeast
cells is presented. Chapter 6 presents the development of a novel supported BLM (sBLM)
interface for screening BLM-nanoparticle interactions. In Chapter 7, adaptation of tBLM

interface and other protein interfaces to gold electrode arrays is presented.

1.3. Background

1.3.1. Biomimetic interfaces containing bilayer lipid membranes

Ion channels, a class of membrane proteins that control ion transport across

membranes, are involved in critical physiological functions such as generation of action

potential in nervous tissue, regulation the heartbeat cadence, and muscle contraction.
Ion-channel malfunction is responsible for many cardiac and neuronal diseases. Hence,
ion channels are important drug targets. Hi gh-throughput methods are needed to screen
compounds for activity against ion channels and thereby identify new lead drug
candidates. However, studying ion channel activity in vitro is a very challenging task

because ion channels express their activities under specialized arnphiphilic environments.

The functions of membrane proteins can be reproduced in laboratory using
biomimetic interfaces that consist of artiﬁcial BLMs embedded with membrane proteins.
The conventional methods to characterize ion channel activity include patch clamp
technique, and planar BLM. Patch clamp technique can be carried out using whole cells
or cell membrane ﬁ'actions. Though sensitive to measure single channel activity, patch
clamp technique is laborious, highly skill-demanding and has a very low throughput.
Unsupported planar BLMs offer an easy and powerful alternative to patch clamp
technique. Planar BLMs can be painted across small apertures. A single channel activity
of ion channel proteins can be measured. However, planar BLMs are fragile and difﬁcult
to adapt to high-throughput platforms. Solid supported BLMs offer a chemically and
mechanically stable BLM system compared to conventional BLMs that can be
investigated using range of analytical techniques. However, supported BLMs provides
less sensitive measurements than conventional BLMs. In this chapter, the design and
characterization of different BLM interfaces is discussed. Three membrane interfaces,
namely planar unsupported BLMs, supported BLMs (sBLM), and a tethered BLM
(tBLM) are presented. Fabrication and characterization of these interfaces using

electrochemical techniques is discussed. Model ion transporting peptide, gramicidin, was

used to demonstrate the measurement of ion channel activity using a planar BLM and a

tBLM interface.
1.3.2. Biomimetic interface for PorB class II porin from Neisseria

meningitidis

tBLM approach has been widely used to study small molecules such as ionophore
valinomycin, and gramicidin peptides. However, incorporation of full-length
transmembrane protein in tBLM interface and measurement of its ion transport activity
still remains a challenging task. In this chapter, we present the incorporation of a
transmembrane porin channel in a tBLM interface and the characterization of protein’s
activity using electrochemical techniques. PorBII is a pore-forming outer membrane
protein from Neisseria sp. The possible role of PorBII in mediating Neisserial infection
makes this protein a potential target for therapeutic drug development. Recombinant
PorBII from N. meningitidis was expressed in E. coli and puriﬁed from inclusion bodies.
PorBII was found to get rapidly incorporated in tBLM and transported ions across tBLM.
PorBII activity in the presence of different ionic species was measured using
electrochemical impedance spectroscopy (EIS). Though the interface was developed for
measuring PorBII activity, the interface can be easily extended to other pore-forming
proteins such as outer membrane protein A (OmpA), a-hemolysin, and others. PorBII
was provided by Yi Zheng and Dr. R. M. Garavito ﬁ'om the Department of Biochemistry

and Molecular Biology at Michigan State University (MSU).

1.3.3. Investigation of voltage dependent closure of PorBII porin using

planar BLM and tBLM interfaces

Porins allow transport of ions and other nutrients across cell membranes of
bacteria. The porin-mediated transport processes can be regulated by binding of a
chemical molecule, by transmembrane potential, or by a mechanical stimulus. The
understanding of ion channel gating mechanisms is important for design of targeted
therapies. Additionally, ion channel safety pharmacology has been an important area in
drug discovery. In this chapter, we studied the voltage dependent conductance properties
of PorBII porin using planar BLM and tBLM interfaces. EIS was used to show a
reversible, voltage dependent closure of PorBII porin incorporated in a tBLM. Single
channel studies were conducted in parallel using planar BLMs to conﬁrm tBLM results.
The approach involving tBLM and EIS can be easily automated and adapted for a rapid
preliminary screening of compounds that modulate activity of voltage-gated proteins.
After identiﬁcation of possible hits, a detail characterization can be followed using planar
BLM technique. This work was conducted in collaboration with Yi Zheng and Dr. R. M.
Garavito from the Department of Biochemistry and Molecular Biology at MSU. Planar
BLM experiments were carried out in collaboration with Dr. Sreenivasarao Kota at

Department of Chemical Engineering and Materials Science (CHEMS) at MSU.
1.3.4. Fabrication of tBLMs using membrane fractions of
Schizosaccharomyces pombe yeast cells

Yeast cells are often used as host cells for expressing recombinant eukaryotic

proteins that require post-translational modiﬁcations. In most membrane biosensor

studies, puriﬁed recombinant membrane proteins are utilized. Membrane proteins are
often difficult to purify and may lose activity during the downstream processing of the
protein. After puriﬁcation, most membrane proteins need to be reconstituted in liposomes
to restore protein’s activity. In this study, we described an approach to reconstitute -
membrane proteins in their native BLM enviromnent on an interface without need of
puriﬁcation and subsequent reconstitution in liposomes. Membrane fractions derived
from S. pombe yeast cells were used to simultaneously deposit the upper leaﬂet of tBLM
and the membrane proteins of S. pombe. Electrochemical and optical techniques were
used to monitor BLM formation. Highly insulating tBLMs were obtained using the
membrane fractions, which were found suitable for measuring the ion transport activity
of gramicidin peptides. This approach could streamline efforts to characterize novel
proteins and screen drug candidates for activity against membrane proteins. This work
was carried out in collaboration with Dr. Dexin Sui and Dr. R. M. Garavito ﬁom the
Department of Biochemistry and Molecular Biology at MSU. Fluorescence recovery
after pattern photobleaching (FRAPP) experiments were performed in collaboration with
Aaron Greiner and Dr. R.Y. Ofoli from Department of Chemical Engineering and

Materials Science at MSU.
1.3.5. Interaction of nanoparticles with supported BLMs

Since last decade, nanoparticles have been introduced in myriad of commercial
applications including cosmetics, paints, textiles and others. Nanoparticles also have
tremendous application in biosensor development. Incorporation of nanoparticles greatly
enhances the surface area, thereby increasing the sensitivity of biosensor by several folds.

The small size of nanoparticles provides them unique properties such as low

irnmunogenicity. This property has been exploited in medical applications wherein
nanoparticles are being used as gene and drug delivery vehicles. Recent research has
suggested that the increasing use of nanoparticles in commercial products poses
signiﬁcant health risks. Therefore, a detail understanding of molecular mechanisms that

causes nanoparticle induced cellular toxicity is required.

It has been hypothesized that nanoparticles disrupt cell membranes to transport
materials inside the cell. Damage to cell membranes could also induce cytotoxic effects
in host cells. In this study, a supported BLM interface was developed for studying the
interaction of nanoparticles with BLMs. Monodisperse positively charged
polyamidoamine (PAMAM) nanoparticles of different size and generation number were
used to study the effect of nanoparticles’ physical properties on their ability to disrupt
BLMs. To understand the effect of surface ﬁmctionality, novel synthetic nanoparticles
were designed and screened for their interaction with sBLMs. This work was carried out

in collaboration with Dr. G. Baker ﬁ'om the Department of Chemistry at MSU.
1.3.6. Integration of biomimetic interfaces with gold electrode arrays

With completion of Human Genome Project, the focus has now shifted towards
structural and functional characterization of proteins encoded by newly discovered gene
sequences. Novel soluble and membrane proteins are being identiﬁed and expressed
using recombinant DNA technology. There is a need for development of tools that will
allow a rapid, cost-effective, multi-parameter, parallel, and high-throughput
characterization of multiple classes of proteins. In this study, gold electrode arrays were
designed and fabricated using photolithography. The microelectrode array platform was

used to deposit different protein interfaces. High impedance tBLMs were deposited on

microelectrodes ranging in size from 3 mm in diameter down to 50 pm in diameter.
Gramicidin peptide was incorporated in tBLMs and the ion transport across BLM was
measured using EIS. Further, tBLM interface was used to study the effect of temperature
on bilayer ﬂuidity and gramicidin activity.

In addition to tBLMs, a biosensor interface for measuring the esterase activity of
NEST, a catalytically active fragment of a membrane protein Neuropathy Target Esterase
(NT E), was fabricated on microelectrode arrays. NTE is found in human neurons, and its
malﬁmction or inhibition by organophosphorus (OP) compounds leads to several
neurological diseases including organophosphate induced delayed neuropatlry and Lou
Gehrig’s disease. A NEST biosensor was used to measure the esterase activity of NEST,
the hydroxylation activity of tyrosinase, and NEST inhibition by phenylmethylsulfonyl
ﬂuoride (PMSF). Further, NEST sensor was utilized to study the temperature dependence
of NEST and tyrosinase activity. This work was carried out in collaboration with Dr.

Andrew Mason from the Department of Electrical and Computer Engineering at MSU.

10

2. BIOMIMETIC INTERFACES CONTAINING BILAYER
LIPID MEMBRANES

2.1. Abstract

The biological functions of cell membranes can be reproduced in the laboratory
using biomimetic interfaces that consist of artiﬁcial bilayer lipid membranes (BLM). In
this chapter, the design and characterization of conventional, and modern biomimetic
interfaces containing BLMs is discussed. Electrophysiological methods such as patch
clamp technique and the planar BLM techniques are the ‘gold standard’ techniques used
for measuring a single channel activity of ion channel proteins. However, these
techniques are labor intensive, less stable and have low throughput. Solid supported
artiﬁcial BLMs (sBLM) offer stable BLM interfaces that can be self assembled on variety
of substrates and are readily addressable by several analytical techniques. Tethered BLMs
(tBLM) are specialized solid supported BLMs that provide necessary molecular
architecture for carrying out ion channel assays. In a tBLM approach, BLM is decoupled
from electrode surface using a hydrophilic chemical spacer thereby creating an ion
reservoir. The electrical properties of the tBLM and ion passage mediated through the
incorporated ion channels can be measured using electrochemical impedance
spectroscopy (EIS). The design and characterization of planar BLM, sBLM, and tBLM
interfaces is described. Planar BLM and tBLM interfaces were used to measure the ion

transport activity of gramicidin peptides.

11

2.2. Introduction

Ion channels are transmembrane proteins that allow conduction of different ions
across the hydrophobic lipid bilayer of a cell membrane (Miller, C., 1986). Ion channels
are involved in various physiological processes such as generation of action potential in
excitable cells like muscle cells, cardiac cells and neurons. Ion channels also control the
physiological processes in non-excitable cells such as release of hormone vasopressin
ﬁ'om posterior pituitary gland. Ion channels can be selective or non-selective towards ion

transport. Depending on the controlling factor, ion channels can be classiﬁed as

1. Ligand-gated channels: A chemical molecule (ligand) binds to the ion channels

and causes a change in the conformation that may open or close the channel.

2. Voltage-gated channels: A change in membrane potential can cause opening or
closing of the channel.

3. Mechano-sensitive channels: A mechanical stimulus such as shear stress acts as
an activator for opening of an ion channel.

Due to their critical physiological role in human body, ion channels are of
immense importance to the pharmaceutical and biotechnology industry. Drugs acting on
the ion channels have an annual sale of more than $6 billion (Wang, X. B. and Li, M.,
2003). The major therapeutic indications include the heart diseases like cardiac
arrhythmias and hypertension, neuronal diseases such as epilepsy and anxiety, pain and
diabetes. Also, some drugs may have undesired interactions with the ion channels leading

t0 fatal side effects. Therefore, ion channel safety pharmacology has been equally
irImportant area in drug discovery. There is a tremendous need for development of high

throughput techniques for measuring ion channel activities in drug discovery process.

12

2.2.1. Patch clamp technique

In 1978, biophysicists Erwin Neher and Bert Sakmann revolutionized the ﬁeld of
electrophysiology when they successfully measured the electric current through ion
channels in cell membrane (Neher, E. et al., 1978; Sigworth, F. J. and Neher, E., 1980).
The technique was called patch clamp method which won them the Nobel Prize for
medicine in 1991. Conventional patch clamping (Wang, X. B. and Li, M., 2003) was
completely carried out under the microscope using a glass micropipette which has a
conducting solution and a rrricroelectrode in it. The cells were taken on a glass slide;
micropipette tip was moved towards a cell surface using a micromanipulator and pressed
gently against cell membrane. A small suction was then applied to pull a patch of
membrane into the opening of pipette tip and create a seal. The seal resistance was
monitored by applying the test voltage pulses. A sea] resistance of order gigaohms, thus
called ‘gigaseal’, was required for accurate recordings with low noise signal and
formation of a stable seal. Formation of gigaseal is very much dependent on the surface
characteristics of pipette tip. Smooth surfaces are required for the formation of gigaseal
which allows the low interferences by avoiding the leakage currents. Once the gigaseal is
formed, the activity of the channel through the patch of cell membrane can be measured.
Whole cell measurements of ion channel activity can be done by applying additional
suction to rupture a membrane patch. The effect of drug on channel activity can be
measured by adding the drug in external solution and measuring the ion currents. Voltage
0811 be manipulated to control the current across the whole cell membrane. Though
Sensitive, patch clamp technique have critical disadvantages. The technique is very labor

intensive and skill demanding. Highly skilled and experienced electro physiologist would

13

be required for successful measurements. It has a very low throughput. Conventional
patch clarrrp can allow only 7-8 cell measurements per day even by a highly skilled

professional.
2.2.2. Planar unsupported BLMs

The planar unsupported BLM is an alternative ‘standard’ technique used to study
ion-transporting membrane proteins. The approach involves forming a BLM across an
aperture in a teﬂon or delrin partition that separates two aqueous solutions [Figure 2.1]
(Miller, C., 1986; Ottovaleitrnannova, A. and Tien, H. T., 1992; Tien, H. T. and Ottova,
A. L., 2000; Tien, H. T. and Ottova, A. L., 2001). BLM is formed by painting a
phospholipid solution across the aperture (100 pm - 2 mm in diameter) and then allowing
the lipid ﬁlm to gradually thin into a bilayer membrane. A surfactant-solubilized
membrane protein can then be added to one of the solutions, resulting in incorporation of
the protein into the BLM. The activity of the protein can then be measured
electrochemically. Planar BLM technique shares the advantages of patch clamp
technique. It is very sensitive and can resolve the ion charmel function at single channel
level. Single channel measurements can be used to characterize the channel’s
conductance states, open probability, and gating properties (Miller, C., 1986). Kinetic
behavior of ion channels can be accurately determined in time resolution of tens of
microseconds. Membrane potential is controlled at speciﬁc voltages which allow ion

currents to be measured under well deﬁned conditions, giving accurate results. The tight
cOntrol over the transmembrane potential allows the investigation of voltage dependence

of the channel conductance, as well as the channel closure.

14

The major limitation of the planar BLMs is that they are very fragile lasting only
a hour before breaking. Also, the technique is not well-suited for high-throughput studies,
due to the relative low stability of the BLM and issues related to the scaling up of the
system. Attempts are being made to develop high-throughput planar bilayer
methodologies using silicon substrates that are etched to create micrometer holes
(Pantoja, R. et al., 2004) and then using polydimethoxylsilane (PDMS) polymer
ﬁarnework to fabricate nricroﬂuidic channels to create two aqueous compartments of
bilayer chamber (Klemic, K. G. et al., 2002a; Klemic, K. G. et al., 2002b; Wang, X. B.

and Li, M., 2003).
‘ 2.2.3. Solid supported BLMs

Supported BLMs (sBLM) provide more stable and mechanically robust system
than planar BLMs (Sackrnann, E., 1996). sBLM can be deposited on various hydrophilic
substrates such as metals (e.g., platinum) (Snejdarkova, M. et al., 1997), glass (Cremer,
P. S. and Boxer, S. G., 1999), silica, mica (Radler, J. et al., 1995; Richter, R. P. and
Brisson, A. R., 2005) and gold (Peng, Z. Q. et al., 2002). Small unilamellar liposome
vesicles made of synthetic phospholipids fuse and rupture (vesicle fusion technique) over
the hydrophilic surfaces (such as silica, glass) due to electrostatic interactions between
the surface groups of substrate and the head group charges of the phospholipids [Figure
2.2A]. On gold substrates, a self assembled monolayer (SAM) of alkanethiols is ﬁrst

deposited to create the lower leaﬂet of sBLM and then the liposome vesicles were used to
deposit the upper leaﬂet of BLM using vesicle ﬁrsion technique [Figure 2.2B]. sBLMs
are suitable to study several classes of membrane processes, including redox proteins and

those that involve simple receptor-analyte binding. sBLMs also offer a suitable platform

15

for electrochemical investigation of BLM disruption by nanoparticles. However, sBLMs
are less suitable for studying protein-mediated ion transport, due to the lack of a
hydrophilic ion reservoir between the sBLM and the underlying substrate. Various
approaches have been used to create an ion reservoir, including depositing the sBLM
onto a polymeric cushion [Figure 2.2C] (Mauro, A. et al., 1988; Naumann, C. A. et al.,
2002; Fortig, A. et al., 2004) or polyelectrolyte multilayer (Kugler, R. and Knoll, W.,
2002), and attaching biotinylated vesicles to the surface using avidin-biotin interaction
followed by fusion of liposomes to form a BLM (Berquand, A. et al., 2003). Formation
of freely suspended sBLMs on anodized alumina membranes is also reported, wherein
the freely suspended BLM essentially acts as planar BLM having inﬁnite ion reservoir on
both sides of membrane [Figure 2.2D]. These approaches have been met with limited
success, in part, because defects in the BLM allow ion leakage and preclude formation of

an insulating membranes (Richter, R. P. et al., 2006).

The tethered BLM (tBLM) architecture [Figure 2.3] (Lang, H. et al., 1992; Lang,
H. et al., 1994; Cornell, B. A. et al., 1997; Raguse, B. et al., 1998; Krishna, G. et al.,
2003; Terrettaz, S. et al., 2003) overcomes the drawbacks associated with sBLM by
simultaneously providing a highly insulating BLM and an ion reservoir between the BLM
and an underlying electrode. The approach uses custom tethering lipids, whose
architecture includes one terminal group (such as thiol-) that binds to the electrode (gold),
another terminal group that anchors into the BLM, and an intermediate hydrophilic
region (e. g., a polyethylene glycol (PEG) moiety). Cornell and coworkers prepared
Synthetic tethered lipids with a phytanoyl lipidic core and a hydrophilic spacer group

Composed of tetraethylene glycol and hemisuccinate ester (Cornell, B. A. et al., 1997;

16

Raguse, B. et al., 1998; Woodhouse, G. et al., 1999; Krishna, G. et al., 2001). Similarly,
Knoll and coworkers (Bunjes, N. et al., 1997; Krishna, G. et al., 2001; Sinner, E. K. and
Knoll, W., 2001; Atanasov, V. et al., 2005; Atanasov, V. et al., 2006; Vockenroth, I. K.
et al., 2007) have synthesized thiolipids which have a thioctyl- group as a surface binding
residue, 3 tetraethylene glycol chain as a hydrophilic spacer group and phytanoyl chains
as hydrocarbon tail. Vo gel and coworkers have also reported synthesis of thiolipids with
similar chemical composition (Lang, H. et al., 1992; Lang, H. et al., 1994; Terrettaz, S. et
al., 2003). Though gold is the most widely used substrate in tBLM studies, Beccuci and
coworkers have used mercury surface to deposit tBLMs (Moncelli, M. R. et al., 2004;
Becucci, L. et al., 2005a; Becucci, L. et al., 2005b; Becucci, L. et al., 2006; Becucci, L.
and Guidelli, R., 2007). First, a self assembled monolayer (SAM) of the tethering lipids is
deposited on gold surface. After SAM formation, remaining lipids needed to establish the
bilayer are deposited, usually by the vesicle fusion technique [Figure 2.4] (Atanasov, V.
et al., 2005). An ideal tBLM possesses a well deﬁned ion reservoir, is mechanically and
biochemically stable, is accessible to electrical measurements, and retains the ﬂuid nature
of BLM. Cornell and coworkers successfully demonstrated the use of tBLM interface as
ion channel immunosensor wherein the change in the ion transport through antibody-
modiﬁed-gramicidin channels was used to identify and detect the antigen concentration

in test solution (Cornell, B. A. et al., 1997).

2.2.4. Electrochemical characterization of BLM

2.2.4.1. Electrochemical impedance spectroscopy (EIS)

The functional characterization of tBLM can be performed using electrochemical

impedance spectroscopy. Absence of redox reactions in normal tBLM measurements

17

prevents the use of amperometric techniques. In impedance spectral analysis, a small
sinusoidal potential is applied in addition to a biased DC potential, and the resulting
current is measured as a function of frequency (Bard, A. and Faulkner, L., 2001). The
frequency sweep is generally carried out between 0.01 to 10000 Hz for lipid bilayer
measurements. Impedance represents the net resistance to current ﬂow offered by the
interface at applied potential. Impedance data can be modeled using a suitable equivalent
circuit which is an arrangement of electrical components such as capacitors, resistors and

inductors. Impedance is deﬁned as the ratio of applied voltage to the resulting current.

1 V . . .
Z (w) = 37(25- = 75(COS(6M) “ 1 311109114 )) = ZR + ’21 ...... (Eq. 2.1)

where, ZR = real part of impedance, Z I = imaginary part of impedance, Y = admittance,

VM = amplitude of voltage, 1M = amplitude of the current, 0M = phase delay and a) =

the radial frequency. The plot of real and imaginary part of impedance is called the
‘Nyquist plot’ whereas the plot of absolute value of impedance as a ﬁrnction of frequency

is called ‘Bode plot’.

tBLM’s impedance can be modeled using a Randles equivalent circuit shown in

Figure 2.5. The electrical elements can be roughly attributed to the speciﬁc components

of the interface. In an equivalent circuit shown in Figure 2.5, resistance of solution (R3)
represents the ohmic drop across the electrolyte solution; capacitance (CM), and
resistance of membrane (RM) represent the electrochemical properties of BLM, and the

capacitance of double layer (CDL) is a characteristic of the electrical double layer at the

18

electrode surface. tBLM formation is characterized from the values obtained for the

electrochemical properties associated with BLM namely the capacitance (0.5 - 0.8 chm-

2) and the resistance (1 - 100 MQcmz) of the membrane (Raguse, B. et al., 1998;
Atanasov, V. et al., 2005).
2.2.4.2. Cyclic voltammetry (CV)

CV is a powerful technique to characterize the electron transfer reactions at
electrode surface (Bard, A. and Faulkner, L., 2001). In a typical CV experiment, the
potential of the electrode is cycled between two values of potential at speciﬁed scan rate
and the current produced due to electron transfer is measured. The redox couple present
in the solution (or immobilized on the surface) get oxidized or reduced at the standard
redox potential for that particular redox reaction. CV measurements can be used to
determine the diffusion coefﬁcient of the redox species in the solution, to determine the
effective electrode area available for the electron transfer, and to provide insights into the
mechanism of the reaction (such as multiple electron transfer reactions). Multiple redox

reactions having different standard redox potentials can also be interrogated.

Before taking BLM’s CV measurement, the ﬁrst step is to record cyclic
voltarnmograrns of a bare electrode of known area at different scan rates in an electrolyte
solution (usually, 100 mM NaCl) containing a redox couple (1 mM of potassium
ferricyanide and 1 mM of potassium ferrocyanide). A duck shaped cyclic voltarnmograrn
with well deﬁned reduction and oxidation current peaks around the standard redox
potential (220 mV vs Ag/AgCl) is obtained. Either reduction or oxidation peak current

Value is plotted against the square root of scan rate and the slope of the curve is used to

19

obtain the diffusion coefficient of redox species according to the following Randles-

Sevcik equation (Bard, A. and Faulkner, L., 2001)

:1. = (2.69 x 105) 123/2A CBULK (1)12)” (Eq 2.2)

where, ip is the peak current (uA), n is the number of electrons transferred (1), A is
electrode area exposed, C BULK is the bulk concentration of redox species (1><10'6

molcm-3), D is the diffusion coefﬁcient (cmZ/s) and V is the scan rate (V/s). In a tBLM

interface, BLM forms an insulating dielectric ﬁlm that hinders the electron transfer at the
electrode surface. For an insulating BLM that is free of major defects, no reduction or
oxidation currents are obtained. Defects in BLM would provide the redox species an
access to electrode surface. The magnitude of the redox currents is directly proportional

to the defect area of BLM.

2.3. Materials and methods

2.3.1. Chemicals

The phospholipids 1,2-dipalmitoy1-sn-glycero-3-phosphothioethanol (DPPTE),
1,2-diphytanoyl-sn-glycero—3-phosphothioethanol (DPhytPC), and 1,2—dioleoyl-sn-
SIYCero-3-phosphocholine (DOPC) were obtained from Avanti Polar Lipids Inc,
Alabaster, AL. All other chemicals including chloroform, n-decane, ethanol, NaCl, KCl,
4“(Z-hydroxyethyl)-l-piperazineethanesulfonic acid (HEPES) were obtained from Sigma

Aldrich, St. Louis, MO. Deionized (DI) water (18.2 MQcmz) was supplied by a

Narropure—UV four-stage puriﬁer (Bamstead International, Dubuque, IA). All aqueous
sollrtions are prepared in DI water.

20

2.3.2. Planar bilayer lipid membrane measurements

Planar BLMs were painted across a 250 um diameter aperture in a delrin cup
(Warner Instruments, Hamden, CT) using a solution of DPhytPC lipids in n-decane (20
mgmL). 10 mM HEPES buffer, pH 7.4 containing 1 M potassium chloride was used as
electrolyte solution in cis and trans sides of BLM chamber (1 mL each). After painting
the BLM, the BLM stability was tested by holding the potential at +100 mV and -100 mV
for 10 min each. Ethanol stock solution of gramicidin was diluted and added to cis side of

BLM chamber to achieve the required concentration.

The cis compartment (voltage command) was connected to the head stage input,
and trans compartment was held at ground using a pair of Ag/AgCl reference electrodes
immersed in both sides of BLM chamber. Unitary currents were recorded using a bilayer-
clamp BC-535 ampliﬁer (Warner Instruments, Hamden, CT) and digitized using an
analog-to»digital converter (Digidata 1440A; Axon Instruments, Sunnyvale, CA). Data
were ﬁltered through eight-pole low pass Bessel ﬁlter (LPF-8; Warner Instruments,
Hamden, CT) and analyzed using pClarnplO software (Axon Instruments, Sunnyvale,

CA).
2.3.3. sBLM formation on glassy carbon electrodes

Glassy carbon electrodes (GCE) (Bioanalytical Systems, West Lafayette, IN)
Were sequentially polished with 1000, 300 and 50 nm alumina slurry, followed by
Wﬁshing with DI water and methanol. The electrode was ultrasonicated for 2 min in DI
Water to remove physically adsorbed alumina, given ﬁnal rinses with methanol and DI
Water, and then dried under a nitrogen stream. The GCE was placed in 100 mM NaCl

solution and oxidized at a potential of +1500 mV for 3 min (Du, L. W. et al., 2006). It

21

Ii
II‘."
c "L

was then washed with DI water and dried under a nitrogen stream. Five uL of DOPC
lipid solution (5 mg/mL in chloroform) was applied to the electrode surface, and the
electrode was immediately immersed in an electrolyte solution (100 mM sodium
phosphate buffer, pH 7.4, containing 1 mM of potassium ferrocyarride and 1 mM of
potassirmr ferricyanide). CV was used to monitor changes in the sBLM’s electrochemical

properties until a steady-state situation was observed.
2.3.4. Preparation of liposomes

A chloroform solution of DOPC phospholipids was dried in a test tube under a
slow stream of nitrogen to yield a thin lipid ﬁlm that was freeze dried for at least 5 h to
evaporate residual organic solvent. The dried lipid was hydrated at 22°C using 0.01 M
HEPES buffer, pH 7.4, containing 150 mM NaCl to obtain multilamellar vesicles, which
were then ultrasonicated using a bath sonicator to yield small unilamellar vesicles. The
liposome solution was subjected to 20 cycles of sonication for l min followed by cooling
on ice for 1 rrrin. The solution was then diluted with HEPES buffer to give 1 mM lipid
concentration, and the liposomes’ size distribution was immediately measured using

dynamic light scattering (Brookhaven Instruments Corporation, Holtsville, NY).
2.3.5. Fabrication of tethered bilayer lipid membrane on gold substrate

The gold substrates were fabricated by Lans Goddard Associates, Foster City,
CA, by deposition of a 15 nm adhesion layer of chromium over a silicon wafer, followed
by deposition of a 100 nm layer of gold using e-bearn evaporation. These substrates were
cleaned in piranha solution (7 parts concentrated sulfuric acid and 3 parts 30% hydrogen
Peroxide) for 30 s and washed with DI water and ethanol [Cautionz Piranha solution is a

vigorous oxidant and should be used with extreme caution]. Cleaned gold electrodes

22

were dried in nitrogen and irmnersed in chloroform solution containing 1 mM DPPTE
tethering lipids and 0.5 mM DOPC mobile lipids for 24 h. DPPTE has a terminal thiol
group that forms a covalent bond gold surface, thus creating a SAM. The SAM
formation is usually rapid for the ﬁrst hour of incubation but it takes several hours to give
a uniform monolayer (Diao et al., 2001). After SAM formation, the electrodes were

washed with chloroform to remove weakly attached lipids and dried in a nitrogen stream.

The electrodes were transferred to an electrochemical cell with 0.2 cm2 of electrode area

exposed to the electrolyte solution. The upper tBLM leaﬂet was deposited using the
vesicle fusion technique. Small, unilamellar DOPC liposomes were added to a ﬁnal
concentration of 0.25 mM lipids. DOPC was chosen because of its low phase transition
temperature, which obviates the need for temperature control during the experiments.
Bilayer formation was allowed to proceed for 24 h, after which the liposome solution was
replaced with fresh 100 mM NaCl solution. EIS and CV data were collected. For
valinomycin experiments, 5 uM of valinomycin was added to electrolyte solution (100
mM NaCl) and impedance was recorded. After 1 h of incubation, the a stock solution of
KCl was added to give a 100 mM KCl concentration in electrolyte solution. For
gramicidin experiments, 1 pM peptide was added to an electrolyte solution (100 mM
NaCl) and the protein incorporation into tBLM was allowed to proceed for an hour, and

then EIS and CV were recorded.
2.3.6. Instrumentation

Dynamic light scattering experiments were performed at 25°C using a 90 Plus
Nanoparticle Size Analyzer (Brookhaven Instruments Inc., Holtsville, NY).

Measurements were obtained at an angle of 90° after appropriate sample dilution. Ten

23

runs were performed for each sample. Ellipsometric measurements were carried out
using rotating analyzer ellipsometer (Model M-44; J .A. Woollan Co. Inc., Lincoln, NE)
rimning WVASE32 software. The SAM’s thickness value was determined using 44

wavelengths between 414.0 nm and 736.1 nm. The angle of incidence was 75°, and the

SAM’s refractive index was assumed to be described by n = 1.5 and k = 0. The

advancing contact angle for the tethering lipid SAM was measured by pendant drop
shape analysis using a software-controlled SEO Contact Angle Analyzer (Phoenix 450,
Surface Electro Optics Corporation Ltd., Korea). Electrochemical measurements were
performed using a CHI66OB electrochemical workstation (CH Instruments Inc., Austin,
TX). The three-electrode electrochemical cell consisted of an Ag/AgCl reference
electrode, platinum wire as auxiliary electrode and a gold or GCE as working electrode.
EIS was carried out in 100 mM NaCl solution using a 0 V DC potential and a 5 mV
sinusoidal potential across the ﬁ'equency range of 0.01 to 10,000 Hz. Electrochemical
parameters were calculated by ﬁtting an appropriate equivalent circuit to the impedance
data, using Z-view software (Scribner Associates, Southern Pines, NC). CV was
performed in 100 mM NaCl solution containing 1 mM potassium ferricyanide as a redox
species. The potential was cycled in a range of 500 mV to -200 mV relative to an

Ag/AgCl reference electrode at a scan rate of 50 mV/s.

2.4. Results and discussion

2.4.1. Planar BLMs measurements of gramicidin activity

Planar BLMs were painted using DPhytPC phospholipids. The stability of the

BLM was tested by applying a potential of +100 mV and -100 mV for 10 min each.

24

Gramicidin was added to the cis compartment of BLM chamber at a concentration of 1
pg/mL. A single channel current of 25 pS was obtained [Figure 2.6A]. Incorporation of
multiple channel units resulted in proportional increase in the channel currents. The
cmrent-voltage relationship was linear in a potential range of ~110 mV to +110 mV
[Figure 2.6B]. The channel current-voltage relationship was syrmnetric for positive and

negative potentials.
2.4.2. Characterization of sBLM interface

sBLMs have been deposited on hydrophilic supports such as mica and silica and
on the conductive materials such as platinum (Saclcmann, E., 1996). sBLM formation on
metal surfaces requires cutting the metal while it is immersed in the lipid solution, to
ensure that the lipids contact a fresh metal surface (Sackrnann, E., 1996; Tien, H. T. and
Ottova, A. L., 2000). GCE provides a convenient and cost-effective alternative to metal
electrodes such as platinum, offering good conductivity, a wide potential window, and
the possibility of chemical ﬁrnctionalization (Niwa, O. and Tabei, H., 1994). To increase
the hydrOphilicity of the surface, we oxidized the GCE at 1.5 V for 3 min in a 100 mM
NaCl solution. The resulting formation of carboxylate ions imparts a net negative charge
to the surface of electrode, which associates with the positively charged choline moiety of

the DOPC lipids.
Upon cooling, lipid bilayers undergo a phase transition from a highly ﬂuid, liquid-

crystalline phase (La) to a more viscous, gel phase (LB). Mecke et a1 (Mecke, A. eta1.,

2005a; Mecke, A. et al., 2005b) reported that polycationic polymers selectively interact

with the lipid bilayers in the liquid-crystalline phase (i.e., above the transition

25

temperature). To ensure that the sBLM was fully ﬂuid during the room-temperature
studies, unsaturated phospholipid DOPC (phase transition temperature = -4°C) was
chosen. In contrast, saturated phospholipid having 14 carbon hydrocarbon chains
(myristyl-) has a phase transition temperature of about 28°C. sBLM formation was
monitored using CV and EIS. Figure 2.7 shows the CV curves for bare GCE and DOPC
sBLM. The bare GCE curve exhibited the classic duck shaped proﬁle, with well-deﬁned

peaks for ferrocyanide oxidation and ferricyanide reduction. The cathodic and anodic

peak currents were 43.5 and 44.2 uA/cmz, respectively [Figure 2.7]. The peak splitting of

70 mV was indicative of a reversible electron transfer. The sBLM curve was essentially
ﬂat, with no redox peaks, indicating that the sBLM forms an effective dielectric barrier
that prevents redox species from reaching and transferring electrons at the electrode

surface [Figure 2.7].

Figure 2.8 shows EIS curves of the bare GCE and the DOPC sBLM. An
equivalent circuit shown in Figure 2.9 was used to ﬁt impedance data and obtain the

values of electrochemical parameters. The bare GCE showed a very small charge transfer

resistance (RM in the equivalent circuit) (6 Qcmz), and most of the impedance spectrum

was representative of Warburg impedance, indicating mass transfer resistance. In the

absence of sBLM, the defect capacitance in equivalent circuit is equal to the bare GCE

capacitance (CDEFECT = CGCE) which was obtained to be 35 uF/cmz. A membrane
resistance of sBLM (RM in the equivalent circuit), which is equal to the charge transfer

resistance, was found to be 85 kﬂcmz. RM values greater than 1 Mﬂcm2 have been
reported for tethered lipid bilayer systems, suggesting that the sBLM formed on the GCE

26

has pin-hole defects. Nevertheless, the sBLM still effectively shields the electrode from

the ferricyanide. An equivalent circuit assumed for the sBLM involves a parallel

arrangement of capacitors: one corresponding to the membrane (CM), and the other

corresponding to membrane defects (CDEFECT)- The overall capacitance is the sum of

both capacitances. At the defects, where the electrolyte can contact the electrode surface,

the capacitance per area of defect should be similar to CGCE- Before nanoparticle

addition, the overall capacitance was determined to be 0.75 uF/cmz, a value typical of

those reported for sBLM-coated electrodes (Gao, H. et al., 2001). Under these conditions,

the fraction of the total area occupied by defects, and thus the CDEFECT term, is
negligibly small.
2.4.3. Fabrication of tBLM on gold electrode

A variety of electrode materials have been used for tBLM studies, including gold
(Sinner, E. K. and Knoll, W., 2001; Schiller, S. M. et al., 2003; Atanasov, V. et al., 2005;
Kohli, N. et al., 2006a; Jadhav, S. R. et al., 2008b) and mercury (Moncelli, M. R. et al.,
2004; Becucci, L. et al., 2005a; Becucci, L. et al., 2005b). The choice of substrate is
governed by factors including surface smoothness, availability of tethering molecules that
bind to the surface, and electrical conductivity. Gold has been the most popular substrate
for tBLM experiments because of its chemical stability, high conductivity, facile thiol
chemistry to couple tethering lipids, and compatibility with microfabrication methods for

eventual application in electrode arrays.

27

2.4.3.1. Characterization of a tethering lipid monolayer

We have deposited a tBLM on gold surface by slight variations of published
procedures (Raguse, B. et al., 1998; Naumann, R. et al., 2003a; Naumann, R. et al.,
2003b; Atanasov, V. et al., 2005). The inner leaﬂet of the BLM was deposited by dipping
a gold substrate in a chloroform solution of DPPTE tethering lipids. A self assembled
monolayer (SAM) of tethering lipids was formed after 24 h, which was characterized
using contact angle measurement, ellipsometry, EIS and CV. Ellipsometry indicated that
the tethering lipid SAM had a thickness of 2.7 d: 0.1 nm, which is comparable to the
length of tethering lipid molecule. The SAM exhibited advancing and receding contact
angles of 109 :1: 3° and 105 :1: 3° respectively, consistent with formation of a uniform,
ordered, hydrophobic SAM ﬁlm on the gold substrate. This ﬁnding is reasonable,
because binding between the thioethanol end group and the gold should orient the
tethering lipid molecules so that the hydrophobic dipalmitoyl tails face outward. A Bode
plot showing EIS results for the SAM is shown in Figure 2.10 (curves marked with

triangles). Fitting a Randles equivalent circuit model [Figure 2.5] to impedance data gave

a monolayer capacitance of 1.12 uF/cm2 and a monolayer resistance of 68 kﬂcmz. In

contrast, the capacitance of the bare gold surface was 25 uF/cmz. The large drop in

capacitance value suggested that most of the gold surface was covered with the SAM
(Diao, P. et al., 1999). The CV data obtained in the presence of potassium ferricyanide

showed that SAM formation decreased the redox peak current by 96%, conﬁrming

formation of uniform insulating barrier on the gold (Diao, P. et al., 1999).

28

2.4.3.2. Characterization of a tBLM

The outer leaﬂet of tBLM was deposited using the vesicle fusion technique in
which the liposomes made of DOPC lipids were allowed to fuse over tethering lipid

SAM. EIS results taken at various times after addition of vesicles to the SAM indicated

membrane resistance value (RM) increased rapidly during the ﬁrst few hours and then

approached an asymptotic value after about 10 h. This trend is consistent with the
mechanism of sBLM formation on planar glass surfaces, in which vesicles initially fuse
onto the surface at random locations, forming bilayer patches, and then gaps are
gradually ﬁlled (Cremer, P. S. and Boxer, S. G., 1999). The electrochemical properties of

BLM were obtained by ﬁtting the Randles equivalent circuit to impedance data using Z-

view software [Figure 2.10]. The ﬁtting gave a capacitance value of 0.709 uF/cm2 for
tBLM, which is in agreement with the reported capacitance value of 0.5 — 0.8 llF/cm2 for

BLM (Schiller, s. M. et al., 2003). The resistance value of 5.98 MQcmz, indicated

formation of an insulating tBLM that is impermeable to sodium and chloride ions. CV of
the tBLM in the presence of ferricyanide lacked redox peaks [Figure 2.11], further
conﬁrming formation of a defect-free tBLM that is impermeable to ferricyanide anions.
2.4.3.3. Characterization of valinomycin and gramicidin activity using tBLM
interface

In an electrolyte solution bathing the preformed tBLM, an aliquot of ethanolic
stock solution containing the ionophore valinomycin (5 uM) or gramicidin peptide (1
uM) was added and the incorporation of the ionophore or protein was allowed for an

hour. Partitioning of valinomycin or gramicidin into the tBLM in a functional

29

conformation was conﬁrmed by EIS. Valinomycin ionophores selectively transported
potassium ions across the BLM. In a 100 mM NaCl electrolyte solution, incorporation of

valinomycin did not change the membrane resistance. Addition of 100 mM KCl to the

solution (bathing valinomycin incorporated tBLM) decreased the RM value from 5.98

Mﬂcm2 to 0.98 Mﬂcrn2 [Figure 2.12], indicating the transport of potassimn ions across

tBLM. The results also conﬁrmed the ion selectivity of valinomycin ionophores.
Gramicidin is a penta decapeptide that forms a dimeric ion channel in BLMs. The
monomers of gramicidin peptide gets partitioned in the lower and the upper leaﬂet of
BLM and are free to diffuse in a ﬂuid bilayer environment. Gramicidin peptides are

selective towards alkali metal ions for transporting ions across BLM. Incorporation of

gramicidin peptides in tBLM decreased the membrane resistance ﬁ'om 3.08 MQcm2 to

0.34 MQcm2 [Figure 2.13]. CV results showed a slight increase in charging current after

gramicidin incorporation in tBLM and no redox peaks were obtained [Figure 2.14]. It
suggested that gramicidin did not allow ferricyanide ion transport across tBLM.
Replacing the NaCl solution to 100 mM tetrarnethylammonium chloride or 50 mM
barium chloride results in the recovery of the tBLM membrane resistance observed
before gramicidin addition, suggesting the larger cations such as tetrarnethylammonium
and barium ions are not transported across BLM by gramicidin [Figure 2.15]. These
results provide evidence for the ion selectivity of gramicidin peptide. Additionally, the
effect of ion concentration and the protein concentration on membrane resistance was
studied. At a ﬁxed concentration of gramicidin peptides, a larger drop in the membrane

resistance was obtained when an ionic strength of the electrolyte solution was increased

30

[Figure 2.16]. Increasing the protein concentration had the similar effect due to

partitioning of higher number of protein molecules inside BLM [Figure 2.16].

31

To
Potentiostat

 

Reference
Electrode

 

Figure 2.1: Planar BLM set-up. Lipids were painted across a small aperture in a partition
that separates two compartments of BLM chamber. The ionic current mediated by ion
transport across BLM through the ion channel proteins was measured by a pair of
Ag/AgCl reference electrodes.

32

IIIIII I
III IIIIIIIIII

A) sBLM on silica

 

 

IIIIIIIII

(33:21" a? (I: 19.?

 

C) Polymer cushioned
BLM

W9

B) BLM on gold using
SAM of alkanethiols

IIIII IIIIII IIII
IIIIIIIIIII
-

D) Freely suspended
BLM

 

 
  

Figure 2.2: Supported BLM interfaces. (A) sBLM deposited on a silica or glass surface.
A thin layer of water separates the BLM from the surface of substrate. (B) sBLM can be
deposited on gold surface by forming a SAM of alkanethiols followed by depositing the
upper leaﬂet of sBLM using liposome vesicles. (C) Charged polymers and
polyelectrolytes have been used as polymeric cushions to separate sBLM from electrode
surface and create ion reservoir. (D) Freely suspended BLMs are deposited on anodized

alumina membranes to create sBLM architecture that mimics planar BLMs. [(Richter, R.

P. et al., 2006)]

33

Protein channel

Lipid bilayer

Aqueous
layer

Spacer
molecules

 

Figure 2.3: A tethered BLM (tBLM) architecture showing the decoupling of BLM ﬁ‘om
electrode surface using hydrophilic spacer molecule creating an ion reservoir beneath
BLM. The ion reservoir also provides space to accommodate the bulky
extramembraneous domains of transmembrane proteins. tBLM offers a stable BLM
interface that is suitable for studying ion channel proteins using electrochemical
impedance spectroscopy.

34

I

 

 

 

. (WWW?

 

 

 

 

 

 

 

 

 

 

Elffimi 1:41»? "” sari) Tethering ‘i'.:fw.,;'_', " " ‘ '3'; .-..,13$.."ﬁ‘L’LLLiul'Iﬂ
Gold electrode lipids SAM on gold electrode
Liposomes
vesicles
‘.

_' Ion I Iéagig

A' channel '””
tBLM containing ion protein tBLM

channel protein

Figure 2.4: A scheme for fabrication of a tBLM. A self assembled monolayer (SAM) of
tethering lipids was ﬁrst deposited on gold electrodes. After SAM formation, liposomes
made of DOPC lipids were used to deposit the upper leaﬂet of tBLM using vesicle fusion
technique. tBLM formation was characterized using EIS and CV. Ion channel protein was
then be added to the electrolyte solution to allow its incorporation in tBLM. [Figure not
to scale]

35

 

 

Rs i i CDL

 

 

 

 

Figure 2.5: A Randles equivalent circuit model was ﬁtted to tBLM impedance data to
calculate the values of electrochemical properties. R3 represents the solution resistance,
CM represents the capacitance of the membrane, RM represents the resistance of

membrane and CDL represents the double-layer capacitance, which includes the
Hemholtz capacitance.

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    

5‘ M F {‘1
E 4“
:: 3 -
§2_ ~ r“ q 1 HI "I H
h
3
o 1-
O-v- Let W Lulu-J Li I”.
‘1 l I I
o 5000 10000 15000
Time(ms)
B 4‘
3- y=0.025x
2. R2=0.9992

Current (pA)
O

 

 

-1 _

-2 s

-3 -

-4 T . I . I I
-1 50 -100 -50 0 50 1 00 1 50

Applied voltage (mV)

Figure 2.6: The ion-transport activity of gramicidin peptides measured using a planar
BLM interface. DPhytPC lipids were painted across 250 um aperture. (A) Single channel
currents corresponding to a conductance value of 25 pS were obtained. Incorporation of
multiple channels proportionally increased the current response. (B) Single channel study
was conducted to obtain the current-potential (I-V) relationship for gramicidin peptides.
I-V curve was linear in the range of +1 10 mV to -110 mV. Experimental conditions: Both
bilayer compartments had 1 mL of 20 mM HEPES buffer (pH 7.4) containing 1 M KCl;
gramicidin concentration in cis chamber was 1 pg/mL.

37

50-
40— . 1
q“ 30-

 

/

\2

-‘ N
O O O
A 1

 

-10 _
-20 _
-30 _
-40 l I I I I I I
500 400 300 200 100 O -100 -200

Potential (mV)

Current (pA cm

 

 

Figure 2.7: Cyclic voltarnmograrn for bare GCE (curve 1) and for sBLM deposited on
GCE (curve 2). Well deﬁned oxidation and reduction peaks were obtained at bare GCE
surface. The peak currents were completely diminished after sBLM formation suggesting
that sBLM did not have large pinhole defects.

38

Log Z (Q cmz) >
OD #5 0| 0)

N
1

 

 

W
m
C

1

Phase angle (°)

l l l l l

O 1 2 3 4

Log Frequency (Hz)

 

 

-¥ N 03 h 01 a)
O O O o C o o
1 1 1 1 1 1

I
—L

l l l l l

0 1 2 3 4

Log Frequency (Hz)

Figure 2.8: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for bare GCE (triangles) and for sBLM deposited on GCE
(squares). An equivalent circuit shown in Figure 2.9 was ﬁtted to impedance data. The
solid lines passing through the data points represent the model ﬁt.

39

CDEFECT

_| I.—
Rs ——I l—
c

 

 

 

 

 

—WW—W—
RII Zw

Figure 2.9: An equivalent circuit used to model impedance data obtained for bare GCE
and sBLM on GCE. R3 represents the resistance of solution, CDEFECT is the double
layer capacitance of GCE at the defect sites in sBLM, CM is the membrane capacitance,
RM represents membrane resistance, and Zw is the Warburg impedance.

40

>

Log Z (:2 cm2)
J:- 01 O)

CO
1

 

 

 

2 I I T l l l
-2 -1 0 1 2 3 4
Log Frequency (Hz)

m
(D
O

1

- Phase angle (°)

N 00 h 01 O) \l

O O O O O O
l - l l l l l -

 

 

0 l 1 If T l l

-2 -1 0 1 2 3 4
Log Frequency (Hz)

Figure 2.10: (A) Impedance (Log Z) data, and (B) phase angle data were plotted as a
function of Log frequency for DPPTE tethering lipid monolayer (triangles) and tBLM
(squares) formed with DOPC liposomes in 100 mM NaCl solution. An equivalent circuit
shown in Figure 2.5 was ﬁtted to the impedance data. The solid lines passing through the
data points represent the model ﬁt.

41

 

 

 

 

 

 

 

A 100 — . 1
E \
0 5o ._
i
:: o f
5 \2
h
5 -50 -
o
-100 —
-150 l l l l l l l
500 400 300 200 100 o -100 -200

Potential (mV)

Figure 2.11: Cyclic voltarnmograrn for bare gold electrode (curve '1) and for DPPTE-
DOPC tBLM deposited on gold surface (curve 2). Well deﬁned oxidation and reduction
peaks were obtained at bare gold surface. The peak currents were completely diminished
aﬁer tBLM formation suggesting that sBLM did not have large pinhole defects.

42

Log Z (:1 cm2) >
.b (ﬂ 0) \l

(a)
1

 

 

2 T l l l
-2 -1 0 1 2
Log Frequency (Hz)

 

 

o l I l I

.b

 

-2 -1 0 1 2
Log Frequency (Hz)

A

Figure 2.12: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-DOPC
tBLM after valinomycin incorporation (triangles). tBLM impedance was taken in 100
mM NaCl. Valinomycin was added to give a ﬁnal concentration of 5 M in an electrolyte
solution. After valinomycin incorporation, 100 mM KCl was added. An equivalent circuit
shown in Figure 2.5 was ﬁtted to impedance data. The solid lines passing through the
data points represent the model ﬁt. Membrane resistance decreased signiﬁcantly due to
the ion transport through the tBLM, whereas there was no signiﬁcant change in
membrane capacitance. The solution resistance decreased due to increase in the ionic

strength of an electrolyte solution after 100 mM KCl addition.

43

>
0" a) ‘1
1 1 - - 1

Log Z (:2 cm2)
A

 

 

3 a
2 _
1 I l l T I l
-2 -1 0 1 2 3 4
Log Frequency (Hz)

60-

 

- Phase angle (°)
2’3

N (A) -5
O O O
1 1 - 1

 

 

.3
O

I r l l

O 1 2 3 4
Log Frequency (Hz)

I

N
I

..x

Figure 2.13: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-DOPC
tBLM after gramicidin incorporation (triangles) in 100 mM NaCl. An equivalent circuit
shown in Figure 2.5 was ﬁtted to impedance data. The solid lines passing through the
data points represent the model ﬁt. Gramicidin was added to give a ﬁnal concentration of
1 M protein (monomer) in an electrolyte solution. Membrane resistance decreased
signiﬁcantly due to the ion transport through the tBLM, whereas there was no signiﬁcant
change in membrane capacitance.

 

 

 

 

 

0.8 —
... 0.6 ~
‘2'
<3 0.2 —
E o -
s 1 _—'
3 -O.2
o I ’—
-0.4 '
'0-6 | l l I I l
300 200 1 00 0 -1 00 -200

Potential (mV)

Figure 2.14: Cyclic voltammogram for tBLM (curve 1) and for DPPTE-DOPC tBLM
after gramicidin incorporation (curve 2). The redox peaks were absent in both
voltarnmograrns whereas there was an increase in charging current after the addition of
gramicidin suggesting the passage of ions through tBLM by gramicidin while excluding
the passage of ferricyanide ions through tBLM.

45

O1

    
 

1

Log Z (0 cm2)
A

 

 

 

 

 

3 _
2 Z
1 I T I I I l
-2 -1 0 1 2 3 4
Log Frequency (Hz)
B 90
80 a
E: 70 j
%, 60 -
C
'5 501
3
m 40 -
.:
“,- 30 —
20 ~
10 I I I I ﬁ I
-2 -1 O 1 2 3 4
Log Frequency (Hz)

Figure 2.15: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM in lOOmM NaCl (squares), for
DPPTE-DOPC tBLM after gramicidin incorporation (triangles) in 100 mM NaCl, and for
gramicidin-incorporated tBLM after replacing the electrolyte solution to 100 mM
tetramethylammonium chloride (open diamonds). An equivalent circuit shown in Figure
2.5 was ﬁtted to the impedance data. The solid lines passing through the data points
represent the model ﬁt. Gramicidin was added to give a ﬁnal concentration of 1 uM
protein (monomer) in an electrolyte solution.

46

 

 

 

4 I I I I l I
-2 -1.5 -1 -0.5 O 0.5 1

Log Frequency (Hz)

Figure 2.16: Impedance (Log Z) data were plotted as a function of Log frequency for
DPPTE-DOPC tBLM and for DPPTE-DOPC tBLM aﬁer gramicidin incorporation at
different protein concentration and ionic strength of NaCl solution. Squares: tBLM;
triangles: tBLM containing 10 nM gramicidin in 10 mM NaCl solution; circles: tBLM
containing 10 nM gramicidin in 100 mM NaCl solution; stars: tBLM containing 50 nM
gramicidin in 100 mM NaCl solution; diamonds: tBLM containing 100 nM gramicidin in
100 mM NaCl solution. The membrane resistance decreased with increase in the ionic
strength of solution and the protein concentration.

47

3. FUNCTIONAL CHARACTERIZATION OF PORB
CLASS 11 PORIN FROM NEISSERIA MENIN GI T IDIS
USING A TETHERED BILAYER LIPID MEMBRANE

3.1. Abstract

PorB class II (PorBII) from Neisseria meningitidis is a pore-forming, outer-
membrane protein that can translocate to the host-cell membrane during Neisserial
infections. This study describes development of tethered bilayer lipid membrane (tBLM)
system to measure PorBII’s conductance properties. tBLM was fabricated by depositing a
self assembled monolayer (SAM) of l,2-dipalmitoyl-sn-glycero-3-phosphothioethanol
(DPPTE) tethering lipid on a gold electrode and then using 1,2-dioleoyl-sn-glycero-3-
phosphocholine (DOPC) liposomes to deposit the upper tBLM leaﬂet. Electrochemical
impedance spectroscopy (EIS) and cyclic voltammetry (CV) were used to monitor tBLM

formation and subsequent PorBII incorporation. The highly insulating tBLM exhibited a

membrane resistance and capacitance of 2.5 MQcm2 and 0.7 uF/cmz, respectively.

PorBII was incorporated into the tBLM in an active conformation, as evidenced by its
mediation of ion passage and the decrease in membrane impedance. After PorBII
incorporation, the membrane resistance decreased to 0.6 MQcmz. As expected, the
PorBII charmel was found to be non-selective, allowing the transport of both cations and
anions. CV results indicated that ferricyanide ions can also pass through the pores. The

PorBII-containing biomimetic interface developed in this study could potentially be used

to screen for compounds that modulate PorBII activity.

48

3.2. Introduction

Porins are pore-forming, transmembrane proteins that allow the passage of water,
ions, and nutrients through cell membranes. PorB class II (PorBII) outer membrane porin
from Neisseria meningitidis ﬁrnctions similarly to other gram-negative bacterial porins
(Murakami, K. et al., 1989; Qi, H. L. et al., 1994; Rudel, T. et al., 1996; Minetti, C. et al.,
1997b). It forms a B-barrel structure in the Neisserial membrane (Weiss, M. S. et al.,
1991; Weiss, M. S. and Schulz, G. E., 1992) and has been implicated in the infection
process (Weel, J. F. L. et al., 1991; Weel, J. F. L. and Vanputten, J. P. M., 1991). PorBII
can also translocate vectorially from the Neisserial membrane to host-cell membranes
(Lynch, E. C. et al., 1984). There, it interferes with the phagosome maturation and causes
a rapid inﬂux of calcium into the cell, which can trigger apoptotic cell death (N aumann,
M. et al., 1999). Thus, chemical agents that interfere with PorBII activity might be

possible therapeutic agents to counteract Neisserial infections.

Artiﬁcial bilayer lipid membranes (BLM) have been extensively used to mimic
biological cell membranes for studying membrane processes, such as signal transduction,
1i gand-receptor interactions and ion transport through cell membranes
(Ottovaleitmannova, A. and Tien, H. T., 1992; Tien, H. T., 1995; Sackmann, 13., 1996;
J anshoff, A. and Steinem, C., 2006). The conventional planar bilayer approach involves
fonning a BLM across a small aperture in a partition between two aqueous solutions
(T ien, H. T. and Ottova, A. L., 2001). Such free-standing BLMs are well suited to study
protein-mediated ion transport, but they are very fragile and are typically stable for only
about an hour. Tethered bilayer lipid membrane (tBLM) architecture (Lang, H. et al.,

1992; Lang, H. et al., 1994; Cornell, B. A. et al., 1997; Raguse, B. et al., 1998; Krishna,

49

G. et al., 2003) provides a stable BLM platform for studying ion channel proteins on an
electrode’s surface. In the tBLM approach, tethering molecules are used to establish a
hydrophilic layer between the surface and the bilayer. These molecules bind to the
electrode on one end, bind to the lipids that compose the lower bilayer leaﬂet on the other
end, and have a hydrophilic central domain (Bunjes, N. et al., 1997; Sinner, E. K. and
Knoll, W., 2001; Schiller, S. M. et al., 2003; Terrettaz, S. et al., 2003; Jadhav, S. R. et
al., 2008a). This domain establishes a water-containing ion reservoir that facilitates ion
transfer across the tBLM and also provides space to accommodate extra-membraneous
protein domains. Electrochemical methods have been used to characterize tBLM
containing a variety of proteins or ionophores, including valinomycin (N aumann, R. et
al., 2003a; Naumamr, R. et al., 2003b), gramicidin (Atanasov, V. et al., 2005), mellitin
(Becucci, L. and Guidelli, R., 2007), alamethicin (Yin, P. et al., 2003), OmpF porin
(Stora, T. et al., 1999), M2 peptide (Vockenroth, I. K. et al., 2007) and cytochrome c
oxidase (Giess, F. et al., 2004; Jeuken, L. J. C. et al., 2006). Several electrode materials
have been used for tBLM studies, such as gold (Schiller, S. M. et al., 2003; Terrettaz, S.
et al., 2003; Kohli, N. et al., 2006a) and mercury (Moncelli, M. R. et al., 2004; Becucci,
L. et al., 2005a; Becucci, L. et al., 2005b). The present study describes the functional
incorporation of PorBII porin into a planar BLM interface, and a tBLM interface on a
gold electrode. A highly insulating tBLM was deposited on a gold electrode in a two-
step process, and solubilized PorBII porin was incorporated. Changes in tBLM
impedance due PorBII incorporation were measured using electrical impedance

spectroscopy (EIS) and cyclic voltammetry (CV).

50

3.3. Materials and methods

3.3.1. Expression and puriﬁcation of PorBII porin

The PCR product of Neisseria meningitidis PorBII gene delta 1 to 62 was cloned
into pGEM-Teasy (Promega), digested with NdeI and BamHI, and ligated into pET17b
(Novagen), which was engineered with an insert of an N-terminal hexa histidine tag and
an rTev protease cleavage site, yielding pET17bNT-PorBII [Figure 3.1]. E. coli BL21-
DE3 cells transformed with in plasmid pET17bNT-PorBII were inoculated into Luria-
Bertani (LB) broth containing ampicillin (100 ug/mL) and grown at 37 °C. When the
optical density at 600 nm (OD 600) of the cell culture reaches 1.0, a ﬁnal concentration
of 0.5 mM isopropyl-ﬂ-D-thiogalactopyranoside (IPTG) was added to induce the protein
expression and the cell culture was incubated for an additional 3 h before harvested by

centrifugation and stored at -70°C.

PorBII was puriﬁed using a modiﬁed procedure (Y. Zheng, D. Sui, and R. M.
Garavito, manuscript in preparation) derived ﬁom protocols described elsewhere (Young,
J. D. E. et al., 1983; Qi, H. L. et al., 1994; Minetti, C. et al., 1997a; Minetti, C. et al.,
1998; Song, J. M. et al., 1999). Briﬂy, for every gram of cell culture paste thawed, it was
resuspended with 3 mL TEN buffer (50 mM Tris-HCl, 1 mM EDTA, 100 mM NaCl pH
= 8.0). The cell suspension was pressurized up to 20,000 psi through Emulsiﬂex-C3

twice to thoroughly disrupt the cell membranes. Subsequently, phenylmethylsulfonyl

ﬂuoride (PMSF), MgClz, and DNase I, were added to the cell lysate at the ﬁnal

concentration of 1 mM, 20 mM, and 0.5 ug/mL, respectively. After 20 min of incubation

on ice, the mixture was centrifuged at 10,000 rpm in an 88-34 rotor for 20 min at 4 °C.

51

The cell pellet was retained and homogenized with 0.5% (w/v) deoxycholic acid in the
TEN buffer to wash off contaminant proteins. The second wash with TEN buffer alone
was carried out to remove the deoxycholic acid. The washed cell pellet was unfolded
with 10 mL TEN-urea buffer (8 M urea, one tablet of protease inhibitor cocktail obtained
ﬁom Roche, Indianapolis, IN), resuspended, and sonicated for 3 min. The denatured

protein solution was centriﬁrged at 12,000 rpm in SS-34 rotor for 20 min to remove any

insoluble debris. The supernatant was mixed with an equal volume of 10% Zwittergent®

3-14 in TEN buffer. The PorBII protein mixture was then refolded by slowly (1 mL/min)
diluting into 2X TEN buffer with 20 mM CaC12 and 0.05% Zwittergent® 3-14 to a ﬁnal
protein concentration of 3 mg/mL. The refolding PorBII solution was also centrifuged to

remove any insoluble matter. The supernatant was loaded onto 20 mL Q-sepharose ion

exchange column equilibtrated with buffer B (20 mM Tris-HCl, 200 mM NaCl, 0.05%
Zwittergent® 3-14, pH 8.0). After 3 column volume wash of buffer B, the refolded
PorBII protein was eluted by buffer B with a salt gradient from 200 mM to 1 M NaCl.

PorBII fractions were pooled and loaded to metal chelating Ni-NTA column. The column

was washed extensively with wash buffer (20 mM Tris-HCl, 200 mM NaCl, pH 8.0, 25
mM imidazole) containing 0.05% Zwittergent® 3-14. The PorBII protein was then
eluted from the column by increasing imidazole concentration to 200 mM. The eluted
protein is concentrated to about 1 mg/mL using ultraﬁltration.

For cleaving hexa histidine tag from protein, the buffer was exchanged into rTev
protease cleavage buffer (50 mM Tris-HCl, 0.1mM EDTA, 1 mM dithiothreitol)

containing 0.2% n-decyl-ﬂ-D-maltoside or 0.02% n-dodecyl-ﬂ-D-maltoside. A higher

52

protein to rTev protease ratio was used to cleave off the hexa histidine tag from PorBII (
for every 5 A.U. PorBII, 1 A.U. of rTev at OD 280 nm). The cleavage reaction was
incubated at room temperature for 16 h. Cleaved PorBII was centrifuged to remove any
insoluble matter. The supernatant was loaded to freshly equilibrated metal chelating Ni-
NTA column to remove the rTev protease and the ﬂow through was collected and
concentrated to 5~10 mg/mL, using arnicon ultra concentrators with molecular mass
cutoff of 50 kDa (Millipore, Billerica, MA). Size exclusion chromatography (Superose 6
10/300 GL) was carried out to access the apparent molecular weight and the oligomeric

state of the PorBII. The ﬁnal puriﬁed protein, at >95% purity, was solubilized in 20 mM

Tris-HCl buffer pH 8.0 containing 200 mM NaCl and 0.05% Zwittergent® 3-14.

3.3.2. Planar bilayer lipid membrane measurements

Planar BLM were painted across a 250 um diameter aperture in a delrin cup
(Warner Instruments, Hamden, CT) using a solution of DPhytPC lipids in n-decane (20
mg/mL). 10 mM HEPES buffer, pH 7.4 containing 1 M potassium chloride was used as
electrolyte solution in both sides of BLM chamber (1 mL each). After painting the BLM,
the BLM stability was tested by holding the potential at +100 mV and -100 mV for 5 min
each. Detergent solubilized stock solution of PorBII was diluted in 20mM aqueous
solution of tetraoxyethylene mono-n-octylether (C8E4). The diluted PorBII solution was

added to cis side to achieve the required concentration of protein.

3.3.3. Fabrication and characterization of tBLM

SAM of DPPTE was assembled on cleaned gold substrates for 24 h. After SAM

formation, the electrode was transferred to an electrochemical cell and small, unilamellar

53

of DOPC were added to deposit the upper tBLM leaﬂet. Bilayer formation was allowed

to proceed for 24 h, after which the liposome solution was replaced with ﬁesh 100 mM

NaCl solution. The same procedure was used to create tBLM in CaClz solution, except a

50 mM CaC12 solution was used instead of 100 mM NaCl. Puriﬁed PorBII solution was

added to the pre-formed tBLM to give a ﬁnal PorBII concentration of 1.0 uM protein
(monomer) in electrolyte solution. Porin incorporation was allowed for an hour, and then
the porin-containing solution was replaced with fresh electrolyte solution.
Electrochemical measurements were performed using a CHI660B electrochemical
workstation (CH Instruments Inc., Austin, TX). BIS was carried out in 100 mM NaCl
solution using a 0 mV DC potential and a 5 mV sinusoidal potential across the ﬁ'equency
range of 0.001 to 10,000 Hz. Electrochemical parameters were calculated by ﬁtting a
Randles equivalent circuit to the impedance data [Figure 2.5], using Z-view software
(Scribner Associates, Southern Pines, NC). CV was performed in 100 mM NaCl solution
containing 1 mM potassium ferricyanide as a redox species. The potential was cycled in a
range of 500 mV to -200 mV relative to an Ag/AgCl reference electrode at a scan rate of

50 mV/s.

3.4. Results and discussion

3.4.1. Measurement of PorBII activity using planar BLM

Planar BLMs were formed by painting either DOPC or DPhytPC lipids across a
250 um aperture in dehin cup. PorBII was added to the cis side of the chamber to give a
ﬁnal concentration of 10 ng/mL in buffer solution. After PorBII addition, step increases

in current were observed indicating incorporation of PorBII channels in BLM. Similar

54

conductance values were obtained using DPhytPC, and DOPC phospholipids. However,
DPhytPC formed more stable BLMs than DOPC and hence, was used for further
experiments. Using single channel experiments, it was observed that the channel remains
fully open at +10 mV and the conductance remains constant for the duration of
experiment (20 min). Around 1000 incorporations were examined to obtain the
conductance distribution of PorBII in DPhytPC BLMs [Figure 3.2A]. It was found that
the channel has varied conductance states (studied at +10 mV) with the most probable
conductance state (60%) of 4.2 nS [Figure 3.2B]. It has been reported that porins from
Neisseria Sp. exhibits multiple conductance states (Mauro, A. et al., 1988; Song, J. M. et
al., 1998). The conductance state value was higher than the one obtained by other groups
(Mauro, A. et al., 1988; Rudel, T. et al., 1996; Song, J. M. et al., 1998) for PorBII ﬁom
N. gonorrhoeae and for PorBIII from N. meningitidis. Factors such as type of lipids used
to form BLM, method of forming BLM, and the presence of solvent in BLM might be

responsible for the higher conductance value obtained with our system.

3.4.2. tBLM characterization and measurement of PorB II activity

3.4.2.1. Electrochemical impedance spectroscopy

tBLM was deposited and characterized as described earlier in Section 2.3.5.

DOPC liposomes were used to deposit the upper leaﬂet of tBLM and E18 was used to

monitor tBLM formation. EIS results indicated that the membrane resistance value (RM)

increased rapidly during the ﬁrst few hours after adding vesicles to the SAM and then
approached an asymptotic value after about 24 h. This trend is consistent with the
mechanism of sBLM formation on planar glass surfaces, in which vesicles initially fuse

onto the surface at random locations, forming bilayer patches, and then gaps are

55

gradually ﬁlled over time (Cremer, P. S. and Boxer, S. G., 1999). A Bode plot showing
BIS results for the tBLM 24 h after vesicle addition is shown in Figure 3.3 (squares).

Fitting the Randles equivalent circuit shown in Figure 2.5 to impedance data gave a total

membrane capacitance (CM) value of 0.709 uF/cmz, which is in agreement with the
reported tBLM capacitance value of 0.5 — 0.8 [IF/61112 (Schiller, S. M. et al., 2003). The

RM value of 2.50 Mﬂcm2 for a tBLM indicated the formation of highly insulating tBLM

that is impermeable to sodium and chloride ions. Addition of 1 uM PorBII (monomer) to

the electrolyte solution after tBLM fabrication decreased RM from 2.50 Mﬂcm2 to 604

kilcm2 for the tBLM produced and characterized in 100 mM NaCl solution [Figure 3.3].

These results are consistent with a conceptual model in which PorBII monomers self-
assemble in the bilayer membrane to form a trimeric pore that allows ion passage across

the BLM (See planar BLM results in Chapter 4) thereby decreasing the membrane

resistance. EIS data for a tBLM formed in 50 mM CaC12 solution gave CM and RM
values of 0.81 rrF/cm2 and 24 MQcmz, respectively [Figure 3.4]. The tBLMs prepared in

CaClz solution consistently provided higher membrane resistance values than those

prepared in NaCl, possibly because of tighter lipid packing in the presence of calcium

salts. PorBII addition resulted in a membrane resistance decrease from 24 MQcm2 to
1.93 MQcmz. The large drops in RM following PorBII addition suggested that PorBII

signiﬁcantly increases ion transport across the tBLM. A large ARM obtained with tBLM

interface is in agreement with the large conductance obtained with the planar BLM

56

studies. However, the PorBII-mediated ion inﬂux inside host cell is reported to be
regulated by the nucleotides (ATP and GTP) that may prevent toxic effects to the target

cell (Rudel, T. et al., 1996).

In contrast to the large change in resistance values, addition of PorBII to the
tBLM had little effect on membrane capacitance, because the area occupied by the porin

inside tBLM is negligibly small. A control experiment was conducted to determine

whether the Zwittergent® 3-14 detergent used to solubilize PorBII affected the tBLM’s

impedance properties. Detergent was added to obtain the same total detergent
concentration as was present after addition of detergent-solubilized protein. The

impedance spectra of the pre-formed tBLM was unaffected by detergent addition.
We used the conductance data obtained from planar BLMs to estimate the number

of channels in tBLM interface. The bulk (multiple) channel conductance in tBLM can be

obtained using following relationship-

1 G 1 + 1
__ = 0 = __
MT TAL RMBLM RPORE

 

R = GMBLM +GP0 139(3-1)
MTOTAL

where, RMTOT AL stands for the RM value measured in the presence of PorBII in tBLM,
RMBLM represents the RM obtained in absence of PorBII and RPORE represents the
pore resistance (GMTOTAL, GMBLM and GPORE represent the corresponding

conductance values). GPORE obtained from Equation 3.1 was divided by the average

single-channel conductance obtained from planar BLM study to calculate a PorBII

channel density of about 354 channels per cm2 of tBLM.

57

3.4.2.2. Cyclic voltammetry

Cyclic voltarnmograrns of the tBLM in the presence of ferricyanide [Figure 3.5]
lacked redox peaks, conﬁrming formation of a high-impedance tBLM that is
impermeable to ferricyanide anions. However, PorBII addition resulted in a well deﬁned
peak for ferricyanide reduction, presumably due to PorBII-mediated ferricyanide-ion
transport across the bilayer. Minetti et a1. studied permeation of oligosaccharides through
proteoliposomes containing PorBII (Minetti, C. et al., 1997b) and estimated a channel

pore diameter of 1.6 nm, which is greater than the ionic radii of most of common ions.

These results indicate that detergent-solubilized PorBII readily incorporates itself
into pre-formed tBLM in an active conformation. This ﬁnding is consistent with PorBII’s
previously reported ability to spontaneously incorporate itself into cell membranes,
thereby facilitating the infection process (Lynch, E. C. et al., 1984; Naumann, M. et al.,
1999). The relative ease with which PorBII can be inserted into a pre-formed tBLM
provides a facile method for electrochemical characterization of the channel’s activity. In
this method, the tBLM’s resistance is measured before and after PorBII addition, and the
resulting change in resistance is taken to be a measure of the protein’s channel-forming
activity. The tBLM interface was easily fabricated using commercially available DPPTE
lipids and provided a large drop in membrane resistance following PorBII addition. This
approach obviates the need for custom-synthesized tethering lipids that are used in tBLM
based interfaces for studying ion transport across BLM (Raguse, B. et al., 1998; Stora, T.
et al., 1999; Schiller, S. M. et al., 2003). In contrast to PorBII, many ion-transporting
membrane proteins cannot be easily transferred from aqueous solution into a preformed

tBLM. An alternative method for such proteins is to ﬁrst embed the proteins into

58

liposomes, and then to use the resulting proteoliposomes to simultaneously deposit both
the top leaﬂet of the tBLM and the protein (V ockenroth, I. K. et al., 2007). However, this
approach precludes direct measurement of the tBLM’s resistance before and after protein
addition.

This study describes for the ﬁrst time the functional characterization of a
Neisserial PorBII porin using EIS. The facile method, which is based on molecular self-
assembly, is suitable for miniaturization and automation, and thus may be adapted to
biosensor arrays. PorBII was shown to be an excellent model ion charmel for tBLM

studies, because it is easily incorporated into pre-formed tBLM in concentrations that

give substantial ARM values. The PorBII-containing biomimetic interfaces were found

to be stable and are suitable for applications including hi gh-throughput screening for lead
compounds that modulate PorBII activity, and Optimization of methods to form

biomimetic interfaces that express membrane-protein activities.
3.5. Conclusions

A novel biomimetic interface was developed by incorporating the Neisserial
PorBII porin into a tBLM formed on gold using DPPTE tethering lipid and DOPC mobile
lipids. Simultaneously, PorBII activity and its conductance distribution were
characterized using conventional planar BLM technique. Detergent-solubilized PorBII
spontaneously incorporated into hi ghly-insulating tBLM in a ﬁrnctional conformation,
allowing its ion-transport properties to be characterized using EIS and CV. EIS studies

showed that contacting a tBLM with 1 uM PorBII solution resulted in roughly one order

of magnitude decrease in RM in both NaCl and CaClz solutions, although both the RM

59

and ARM values were higher in CaClz than in NaCl. CV studies indicated that PorBII

also passes ferricyanide ions. The PorBII-containing biomimetic interface has potential
utility for high-throughput screening of compounds that modulate PorBII activity and as a

model interface for developing tBLM-based technologies.

60

Aval (142)

Pstl (174)
EcoRl (175)

BamHl (206)

 
  
  
 
 

Apa Ll (4006)

Ap

Pstl (3584) PorBII

pET-1 7NT-PorBll

  
 

4333 bp

    

Ndel (1242)

Aval (1396)
Apa Ll (2760)

 

on
Apa LI (2260)

 

Figure 3.1: pET17bNT-PorBII construct used to transform E. coliBL21 cells for
expressing PorBII. The PCR product of Neisseria meningitidis PorBII gene delta 1 to 62
was cloned into pGEM-Teasy, digested with NdeI and BamHI, and ligated into pET17b
(Novagen), which was engineered with an insert of an N-terminal hexa histidine tag and
an rTev protease cleavage site, yielding pET17bNT-PorBII.

61

>
(O
O
O

l

h C) \l

01 O 01

O O O
1 1 1

Current (pA)

300-

150-

 

0 I I I I l

 

0 5 10 15 20 25
Time (ms)

W
(A3
0

1

-‘* N N
01 o 01
1 1 1

Probability (%)
8

 

 

0.2 0.8 1.4 2 2.6 3.2 3.8 4.4 5
Conductance (nS)

Figure 3.2: (A) Recordings of DPhytPC/n—decane BLM showing PorBII channel
insertions. (B) Histogram of channel conductance of PorBII in DPhytPC BLM, obtained
from 1000 channel insertion recordings, showing the probability of occurrence of
particular conductance unit. The conductance values were pooled in bins of 0.2 nS.
Experimental conditions: Both bilayer compartments had 1 mL of 20 mM HEPES buffer
(pH 7.4) containing 1 M KCl; PorBII concentration in cis chamber was 100 pg/mL;

Applied potential was +10 mV.

62

 

 

A 6.5 1
5.5 -
"E
o 4.5 -
9.
N
or 3.5 —
o
.1
2.5 —
1.5 I I I I I I
-2 -1 0 1 2 3 4

Log Frequency (Hz)

 

 

10 T I I I I

 

-2 -1 0 1 2 3 4
Log Frequency (Hz)

Figure 3.3: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-DOPC
tBLM after PorBII incorporation (triangles) in 100 mM NaCl. An equivalent circuit
shown in Figure 2.5 was ﬁtted to the impedance data. The solid lines passing through the
data points represent the model ﬁt. PorBII was added to give a ﬁnal concentration of 1
M protein (monomer) in an electrolyte solution. Membrane resistance decreased
signiﬁcantly due to the ion transport through the tBLM, whereas there was no signiﬁcant
change in membrane capacitance.

63

Log 2 (:2 cm2)
h
01

 

 

 

-3 -2 -1 0 1 2 3 4

390—

604
70-
5,60-
50-A
40~
.30—

20-

1 0 I ﬂ I I I I
-3 -2 -1 0 1 2 3 4
Log Frequency (Hz)

'9 (°)

 

Phase an

 

 

.1

Figure 3.4: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for DPPTE-DOPC tBLM (squares) and for DPPTE-DOPC
tBLM after PorBII incorporation (triangles) in 50 mM calcium chloride. An equivalent
circuit shown in Figure 2.5 was ﬁtted to the impedance data. The solid lines passing
through the data points represent the model ﬁt. PorBII was added to give a ﬁnal
concentration of 1 uM protein (monomer) in an electrolyte solution.

64

 

 

 

 

 

 

 

NA 4‘ 1——-'
IE 3_
< 2‘ 3 =
1
:: ,_ 2 . /
C
s 0.
3
0 -1-
-2 _\/
'3 1 1 I 1 I T 1
500 400 300 200 100 0 -100 -200

Potential (mV)

Figure 3.5: CV curves for a bare gold electrode coated with a tethering lipid monolayer
(curve 1), a tBLM formed by deposition and rupture of DOPC liposomes on top of the
DPPTE tethering lipid monolayer (curve 2), and the DPPTE-DOPC tBLM after PorBII
incorporation (curve 3). The redox peaks were absent after bilayer formation but were
observed after PorBII incorporation, suggesting PorBII-mediated transport of
ferricyanide ions through tBLM.

65

4. VOLTAGE DEPENDENT CLOSURE OF PORB CLASS II
PORIN FROM NEISSERIA MENIN GI T IDIS
INVESTIGATED USING ELECTROCHEMICAL
IMPEDANCE SPECTROSCOPY IN A TETHERED BLM
INTERFACE

4.1. Abstract

Electrochemical impedance spectroscopy (EIS) was used to show a reversible,
voltage dependent closure of PorB class II (PorBII) porin from Neisseria meningitidis
incorporated in a tethered bilayer lipid membrane (tBLM) interface. The voltage
dependent conductance properties of recombinant PorBII porin were characterized using
two bilayer lipid membrane (BLM) interfaces: the conventional planar bilayer membrane
and a tBLM interface. A stable tBLM interface was fabricated by depositing a self
assembled monolayer (SAM) of 1,2-dipalmitoyl-sn-glycero-3-phosphothioethanol
(DPPTE) tethering lipids on gold surface and then liposomes made of 1,2-dioleoyl-sn-
glycero-3-phoshocholine (DOPC) lipids were used to deposit the upper leaﬂet of BLM.

At 0 mV bias DC potential, incorporation of PorBII mediated transport of ions across

tBLM decreasing the membrane resistance (RM) from 2.5 MQcm2 to 0.6 MQcm2
(ARM = 1.9 Mﬂcmz). As DC potential was increased, the values for normalized RM
(ARM normalized to the RM of tBLM at respective potential) decreased, indicating a

lower conductance of PorBII. At 200 mV bias DC potential, the normalized RM value

was 20% of the value observed at 0 mV DC. The decrease in conductance of PorBII
porin at higher DC potentials suggested a voltage dependent closure of PorBII. Further, it

was observed that changing the DC potential from 200 mV back to 0 mV reproduced the

66

ARM value observed previously at 0 mV, indicating that the voltage dependent closure

of PorBII was reversible. Single channel studies, conducted using planar DPhytPC BLM,
conﬁrmed the reversible, voltage dependent closure of PorBII porin at high
transmembrane potentials. The trimeric porin closes in three discrete steps, each step
corresponding to closure of one conducting monomer unit. The approach involving the
use of tBLM interface and impedance spectroscopy provides a simple, high-throughput

method for preliminary characterization of voltage-gated proteins.
4.2. Introduction

The conductance of Neisserial porins is modulated by transmembrane potential,
thereby allowing regulation of ion and nutrient transport across Neisserial cell
membranes (Blake, M. S., Gotschlich, EC, 1987; Mauro, A. et al., 1988; Bainbridge, G.
et al., 1998). Using planar bilayer lipid membrane (BLM) technique, a voltage dependent
closure of porin has been reported for Neisserial porins such as protein I ﬁ'om Neisseria
gonorrhoeae (Young, J. D. E. et al., 1983), PorB class II (PorBII) porin from N.
gonorrhoeae (Rudel, T. et al., 1996), and PorB class III porin from N. meningitidis
(Song, J. M. et al., 1998). Blake and co-workers have done extensive structural and
functional characterization of porins from Neisseria sp. (Blake, M. S. and Gotschlich, E.
C., 1982; Lynch, E. C. et al., 1984; Blake, M. S., Gotschlich, EC, 1987; Mauro, A. et
al., 1988; Qi, H. L. et al., 1994; Song, J. M. et al., 1998; Song, J. M. et al., 1999). They
have reported that recombinant PorBII porin from N. meningitidis exhibits similar
functional properties as wild type PorBII protein. They found that the recombinant
PorBII assembles as a trimer in the membrane, forming pores that are ~1.6 nm in

diameter (liposome swelling assay) (Minetti, C. et al., 1997b).

67

In the present study, we have used two BLM interfaces to characterize the voltage
dependent conductance properties of PorBII porin, namely a planar BLM, and a tethered
bilayer lipid membrane (tBLM) interface. Using planar BLM technique, ion transfer
through a single channel can be measured electrochemically to characterize the channel’s
conductance states, open probability, and gating properties (Miller, C., 1986). The
voltage across the BLM can be tightly controlled to study the voltage dependence of the
channel conductance, as well as the channel closure. However, the major limitation of the
planar BLMs is that its very fragile lasting only a couple of hours before breaking. tBLM
combines the stability of a supported BLM with an ion reservoir between the BLM and
electrode. Electrochemical impedance spectroscopy (EIS) has been used to study ion-
transporting membrane proteins embedded in tBLM interfaces (N aumann, R. et al.,
2003a; Naumann, R. et al., 2003b; Schiller, S. M. et al., 2003; Giess, F. et al., 2004;
Moncelli, M. R. et al., 2004; Atanasov, V. et al., 2005; Kohli, N. et al., 2006a; Becucci,
L. and Guidelli, R., 2007; Becucci, L. et al., 2008; Jadhav, S. R. et al., 2008a; Jadhav, S.
R. et al., 2008b; Knoll, W. et al., 2008). Although tBLMs have been used to study
membrane-insertion rate of voltage-dependent alarnethicin peptides (Yin, P. et al., 2003),
and used with patch-clamp techniques to study ion channel gating by voltage-induced
stress (Andersson, M. et al., 2008), use of IS to study the voltage dependent change in a
transmembrane protein’s conductance properties due to channel closure has not been
explored. In this study, we report, for the ﬁrst time, use of IS to study the voltage-
dependent closure of a transmembrane protein embedded in a tBLM interface. Changes

in PorBII conductance were measured at different DC potentials in tBLM interface using

68

EIS, while single channel measurements were carried out in planar BLMs to conﬁrm

voltage-dependent closure of PorBII.
4.3. Materials and methods

Please refer to the Section 3.3 for the material and methods used. The voltage
depedent properties were studied in a potential range of +110 mV to -110 mV in planar
BLM experiments. In tBLM experiments, DC potential used in EIS studies was changed

from +200 mV to -200 mV while the sinusoidal potential was 5 mV.

4.4. Results and discussion

4.4.1. Voltage dependence of PorBII conductance in planar BLM

At higher transmembrane potentials, PorBII conductance was found to decrease in
three discrete steps suggesting that the pore has a trimeric structure [Figure 4.1]. Each
step decrease in the conductance corresponded to closure of one conducting monomer
unit. The closure pattern was not constant, with one, two or even three monomers closure
events occuring at same time. Figure 4.1 indicates that the fully closed form of the porin
still exhibits a low level of conductance. PorBII’s closed-state conductance at a
membrane potential of 100 mV was 26% that of the fully Open trimeric pore. When the
applied potentials were returned to lower potentials (between +50 mV to -50 mV), the
fully open channel conductance was restored indicating that the voltage dependent
channel closure was reversible. The plot of normalized conductance (ratio of conductance
at respective applied potential to the conductance value at +10 mV) as a function of
applied potential [Figure 4.2] showed that the pore closure starts after :t 50 mV. Rudel et

al. observed similar channel closure behavior for PorBII porin from N. gonorrhoeae

69

(Rudel, T. et al., 1996). The conductance study was limited to potentials between +120
mV to -120 mV range because BLM was found to be unstable at potentials above + 120
mV and below -120 mV. At positive potentials, it appeared that the closed conductance
state of the porin approaches asymptotic steady state value. The following version of
Boltzmann distribution used by Song et al (Song, J. M. et al., 1998) was ﬁtted to the

voltage dependent conductance data to obtain the voltage dependent parameters of porin-

G —G nF
GNZXG = E( _ V0) ................................ Eq. (4.1)
W

 

where, G MAX is the maximum open channel conductance, G is the channel conductance
observed at respective applied potential V, GMN is a voltage independent conductance,
F is Faraday’s constant, K is gas constant, T is temperature, n is the number of charges

transferred across BLM, and V0 is the voltage at which the channel has equal probability

of being in open or closed state. The right hand side of Equation 4.1 was plotted as a

function of applied potential V for positive and negative potentials and the values for n
and V0 were obtained from the slope and intercept of the plot, respectively. The voltage
dependent ﬁ'ee energy (AG = nFVo) can be calculated using experimentally obtained
voltage-dependent parameters. At positive potentials, the AG value of 13.4 kJmol-l was

obtained whereas at negative potentials, AG value of 12.7 kJmol-1 was obtained. AG

represents the voltage dependent energy that must be overcome for a channel to undergo

70

conformational change from open to closed state. Further, experimentally obtained n and

V0 values were used to obtain the open probability (Popen) using following equation

1

1+ exp n+F (:1: V0+) + exp "—F (2; VO—) 59 (4-2)

 

P open =

 

 

where 11+, Vo+ and n_, V0- are the n and V0 values obtained for the positive and

negative potentials, respectively. The ﬁtting of open probability to the voltage dependent

porin conductance data is shown in Figure 4.3.

The planar BLM results indicated that PorBII porin self-assembles as a trimeric
pore in BLM, the channel is ﬁrlly open at low transmembrane potentials, and that the
porin undergoes a reversible voltage dependent closure at higher transmembrane
potentials. These results are in agreement with the reports that Neisserial porins show a
large decrease in the single channel conductance at higher transmembrane potentials
(Young, J. D. E. et al., 1983; Mauro, A. et al., 1988; Rudel, T. et al., 1996). Planar BLM
studies conducted by Rudel et al. on PorBII porin from N. gonorrhoeae showed a
decrease in PorBII conductance at transmembrane potentials above +50 mV and below -
50 mV (Rudel, T. et al., 1996). Studies conducted by Song et a1 using PorB class III
porin from N. meningitidis showed that the wild type and the recombinant porin have
similar voltage dependent conductance properties (Song, J. M. et al., 1998). They

observed that PorBIII porin undergoes voltage dependent closure even at low applied

potentials (V0 = 14.8 mV). However, in our studies, we did not observe porin closure

71

upto +50 mV. The exact mechanism of voltage dependent closure of porins is still not
clear. Nonetheless, the voltage gating property of porins is crucial for regulating the

conduction of ions and molecules across the cell membranes of bacteria.
4.4.2. EIS study of a voltage dependent channel closure in a tBLM

interface

Unlike planar BLM, tBLM provides a mechanically robust BLM system that is
addressable using various analytical techniques and adaptable to the automated high-
tlrroughput platforms. However, measurements using tBLM interface are less sensitive
than planar BLMs. tBLMs are more suitable for measuring bulk channel behavior as
opposed to single channel activity measurable by planar BLMs. tBLM’s impedance was

measured at different DC potentials ranging from +200 mV to -200 mV; a representative

curve obtained at 200 mV is shown in Figure 4.4. The RM and CDL values changed

with applied potential. The electrical double layer is believed to undergo rearrangement

in response to a change in applied DC potential, resulting in altered ion distribution inside
and around tBLM (Cevc, G. et al., 1981), and, consequently, altered RM and CDL

values. As the membrane capacitance is largely dominated by the dielectric constant of

the BLM’s hydrophobic core, there was negligible change observed in the CM value.

The difference in RM value of tBLM before and after PorBII incorporation (ARM) is a

measure of ion conduction by porin across tBLM. From the impedance data at different

DC potentials, the normalized RM value, deﬁned as the ratio of ARM to the tBLM’s

RM at that potential before protein addition, was calculated and was used to plot the

72

fractional NRM (ratio of NRM at respective potential to NRM at 0 mV DC) as a

function of applied DC potential [Figure 4.5]. As shown in Figure 4.5 (diamonds), the

fractional NRM value decreased continuously ﬁ'om 0 mV to 200 mV. These results

indicated that the conductance of PorBII containing tBLM decreased at higher DC
potentials, possibly due to a voltage dependent closure of PorBII. Toggling the DC

potential back and forth between 0 mV and +200 mV gave consistent and reproducible

NRM values, indicating that the pore closure is completely reversible. These

observations were consistent with planar BLM results. Comparison of Figures 4.2 and
Figure 4.5 (diamonds) indicates that the effect of transmembrane potential on PorBII
conductance in tBLMs follow different trend than the one observed in planar BLMs. This
result is not surprising considering the signiﬁcant differences between the two
experimental systems. Unlike the planar bilayer, the tBLM has the working electrode in
close proximity to the bilayer and tethering linkages that bind the lower leaﬂet of the
bilayer to the electrode. This difference would be expected to result in differences in the
charge distribution across the tBLM, the ion concentration in the vicinity of the bilayer,
and the electrical double layer structure, all factors that could inﬂuence the

transmembrane potential.

To establish that the voltage dependence of ARM was a porin-related
phenomenon, the tBLM experiments were repeated using the cation-selective peptide
gramicidin in place of PorBII. There was a <10% change in NRM values when the DC
potential was varied from -200 mV to +200 mV [Figure 4.5 (squares)]. Small changes in

NRM observed at high positive potentials for tBLMs containing gramicidin have been

73

attributed to cation repulsion by the positively charged electrode (Yin, P. et al., 2003).
This study clearly demonstrates the ability of EIS and tBLM interface to differentiate

between the voltage dependent and voltage independent protein activities.
4.4.3. Voltage dependent closure of PorB II in symmetric tBLM

To address the effect of membrane asymmetry on the PorBII conductance
behavior, the upper leaﬂet of tBLM was deposited using liposomes made of DPPT E. The
DPPTE liposomes were prepared by hydrating the dry lipid ﬁlm in 10 mM HEPES
buffer, pH 7.4 at 60°C. Small unilamellar vesicles were obtained by sonicating the lipid
suspension (using bath sonicator) above the phase transition temperature of DPPTE (~
40°C). DPPTE liposomes were incubated with DPPTE SAM for 24 h to complete tBLM
formation. The tBLM experiments were conducted at 22°C, a temperature at which the

tBLM exists in a gel state. As a result, a very high impedance tBLM with the membrane

resistance of 41.2 Mﬂcm2 was obtained. After addition of 1 11M PorBII (monomer), the
RM value decreased to 4.75 Mﬂcrn2 at 0 mV DC potential (Figure 4.6A). However, at
200 mV DC potential, the RM value decreased from 42.5 MQcm2 to 37.3 M9cm2

(Figure 4.6B), corresponding to a normalized RM of 0.12. These results indicated that

PorBII shows a voltage dependent behavior in both symmetric and asymmetric tBLMs.

4.5. Conclusions

The study demonstrates, for the ﬁrst time, the use of tBLM along with EIS to

measure voltage-dependent closure of a channel protein. A strong dependence of ARM

on applied voltage was observed for PorBII but not gramicidin, indicating voltage-

74

induced closure of PorBII. Single channel measurements conducted using the traditional
planar BLM technique clearly demonstrated a three step closure of the porin at higher
transmembrane potentials decreasing porin conductance, supporting the EIS results. The
ﬁnding that voltage-dependent closure of channel proteins can be measured in tBLM
using EIS is signiﬁcant, because this approach can be readily miniaturized and adapted to
robotic systems for hi gh-throughput applications including functional proteomics,

biosensors, and drug screening.

75

 

 

<Q 50pAL
V 53
H

c

o

t

:1

U

O__1"——1_

8,; 100mV

«1
= E

0v
>

 

 

Time (s)

Figure 4.1: Unitary response of recombinant PorBII porin incorporated in DPhytPC lipid
bilayer. The magnitude and length of applied potential is indicated below the current
recording. Single channel insertion of 4.2 nS was observed at +10 mV. Increasing the
potential to +100 mV closes the porin in three discrete steps, indicative of closure of each
conducting monomer. The porin channel was reopened by reducing the potential to +10
mV, indicating that the porin closure is completely reversible.

In}
N
1

A
1

P
m
1

{§§§§§§’1§§§§}

.°
03
1

*1

P
#
1

.°
N
1

(I

'0.2 I I I I l I I I I ﬁ
-150 -120 -90 -60 -30 0 30 60 90 120 150

Applied potential (mV)

Normalized
conductance (GIG 10mV)
O

 

 

Figure 4.2: Normalized single channel conductance (Ratio of conductance at respective

potential [G] to the conductance at +10 mV [GlomV] plotted as a function of applied
potential. Five separate experiments were conducted at each potential. Lipids used-
DPhytPC, PorBII concentration- 1 pg/mL, other experimental conditions were same as
described in Figure 3.2.

77

9.0.09 .-*

NhOQ-‘N

lllllJ
W
:94
‘

Normalized
conductance (GIG 10mV)
O

 

-0.2 I 1 1 1 I 1 I r I 1
-150 -120 -90 -60 -3O 0 30 60 90 120 150

Applied potential (mV)

 

Figure 4.3: Open probability calculated from Equation 4.2 using experimentally obtained
It and V0 values. Open probability, shown using solid line, is overlaid on the voltage
dependent porin conductance data.

78

2
Log Z ((2 cm )

 

 

 

1 I I I I I I
-2 -1 0 1 2 3 4
Log Frequency (Hz)

Figure 4.4: Impedance (Log Z) data were plotted as a ﬁrnction of Log frequency for
DPPTE-DOPC tBLM at 200 mV DC potential (squares), for DPPTE-DOPC tBLM after
PorBII incorporation at 200 mV DC potential (triangles), and for DPPTE-DOPC tBLM
containing PorBII after toggling the DC potential to 0 mV (diamonds) in 100 mM NaCl.
An equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data The solid
lines passing through the data points represent the model ﬁt. PorBII was added to give a
ﬁnal concentration of 1 uM protein (monomer) in an electrolyte solution.

79

 

 

1 7 In.

E i I I Q I i i

i, 08 1 I . 2 T

s

s I 1

E 0.6 -

".5 I 1

z 04 4

.3 I i

4 (

.5 I I

<1

V 0 I I I I I
-250 -1 50 -50 50 1 50 250

DC potential (mV)

Figure 4.5: Fractional NRM plotted as a function of applied DC potential used in EIS
for a PorBII incorporated DPPTE-DOPC tBLM (diamonds) and gramicidin incorporated
DPPTE-DOPC tBLM (squares).

80

 

 

1 I I I I I I

-3 -2 -1 0 1 2 3 4
Log Frequency (Hz)

 

 

 

-3 -2 -1 0 1 2 3 4 .
Log Frequency (Hz)

Figure 4.6: (A) Impedance (Log Z) data were plotted as a function of Log ﬁequency for
DPPTE-DPPTE tBLM at 0 mV DC potential (squares) and for DPPTE-DPPTE tBLM
after PorBII incorporation at 0 mV DC potential (triangles). (B) Impedance (Log Z) data
were plotted as a function of Log frequency for DPPTE-DPPTE tBLM at 200 mV DC
potential (squares) and for DPPTE-DPPTE tBLM after PorBII incorporation at 200 mV
DC potential (triangles) in 100 mM NaCl. An equivalent circuit shown in Figure 2.5 was
ﬁtted to the impedance data. The solid lines passing through the data points represent the
model ﬁt. PorBII was added to give a ﬁnal concentration of 1 M protein (monomer) in
an electrolyte solution.

81

5. FABRICATION OF HIGHLY INSULATING TETHERED
BILAYER LIPID MEMBRANES USING YEAST CELL
MEMBRANE FRACTIONS FOR MEASURING ION
CHANNEL ACTIVITY

5.1. Abstract

A tethered bilayer lipid membrane (tBLM) was fabricated on a gold electrode
using l,2-dipalmitoyl-sn-glycero-3-phosphothioethanol as a tethering lipid and the
membrane fractions of Schizosaccharomyces pombe yeast cells to deposit the upper
leaﬂet. The membrane ﬁactions were characterized using transmission electron
microscopy (TEM) and dynamic light scattering (DLS) and found to be similar in size to
small unilamellar vesicles of synthetic lipids. The dynamics of membrane ﬁaction
deposition and rupture on the tethering lipid layer were measured using quartz crystal
microgravimetry (QCM). Fluorescence recovery after patterned photobleaching (F RAPP)
was used to study the formation of microsomal sBLM on a glass substrate. The
electrochemical properties of the tBLM were characterized using electrical impedance

spectroscopy (EIS) and cyclic voltammetry (CV). The tBLM’s electrical resistance was
greater than 1 MQcmz, suggesting a defect-free membrane. The suitability of tBLM
produced using membrane ﬁ'actions for measuring ion-channel activities was shown by a

decrease in membrane resistance from 1.6 MQcm2 to 0.43 MQcm2 following addition of

granricidin. The use of membrane fractions to form high-quality tBLM on gold
electrodes suggests a new approach to characterize membrane proteins, in which the
upper leaﬂet of the tBLM is deposited, and overexpressed membrane proteins are

incorporated, in a single step. This approach would be especially useful for proteins

82

whose activity is lost or altered during extraction, puriﬁcation, and reconstitution, or

whose activities are strongly inﬂuenced by the lipid composition of the bilayer.
5.2. Introduction

Typically, tBLM are fabricated in a two-step process, in which a self-assembled
monolayer of the tethering lipid is ﬁrst bound to the electrode, and then the remainder of
the BLM is added via liposome adsorption and rupture. Lipid-tethering chemistries have
been established for a variety of electrode materials, including gold (Raguse, B. et al.,
1998), silica (Atanasov, V. et al., 2005) and mercury (Becucci, L. et al., 2005a). A
variety of methods have been developed to deposit the upper BLM leaﬂet on the tethering
layer and incorporate a variety of ionophores, peptides, and transmembrane proteins. In
one commonly used approach, the puriﬁed protein is reconstituted into liposomes, and
the resulting proteoliposomes simultaneously deposit the upper BLM leaﬂet and the
protein (Graneli, A. et al., 2003; Zebrowska, A. and Krysinski, P., 2004; Elie-Caille, C. et
al., 2005). In a second method, the liposomes are used to deposit the upper BLM leaﬂet.
Then, proteins that are either naturally soluble or detergent solubilized are added to the
surrounding aqueous phase and allowed to become embedded into the BLM, in some
cases with the assistance of detergent removal. Both of these approaches require the
target protein to be puriﬁed and then reconstituted into a BLM. These steps can be
problematic for membrane proteins whose activity is lost or altered during extraction,
puriﬁcation, and reconstitution, or whose activities are strongly inﬂuenced by the lipid
composition of the membrane. In such cases, it may be preferable to form the tBLM by
direct deposition of semi-puriﬁed membrane fractions (microsomes) containing the target

protein, thereby circumventing the abovementioned processing steps.

83

The goal of this study was to fabricate a high-quality tBLM using the membrane
ﬁactions of yeast cells that would be suitable for the studying ion transport across BLM.
Previous attempts to form sBLM using rrricrosomes containing membrane proteins did
not yield an insulating bilayer membrane (Elie-Caille, C. et al., 2005; J euken, L. J. C. et
al., 2006) and were not suitable to study the ion-channel proteins using electrochemical
techniques. This study describes for the ﬁrst time a method to produce high-impedance
tBLM on a gold electrode using membrane fractions. Schizosaccharomyces pombe was
selected as the source of microsomes because it has many desirable traits as an expression
host for recombinant membrane proteins. It is robust, easy to grow, capable of expressing
even bulky and high molecular weight eukaryotic proteins, and, unlike E. coli, can carry
out many types of post-translational modiﬁcation. The dynamics of membrane-fraction
deposition and rupture on the tethering-lipid layer were measured using quartz crystal
microbalance (QCM) gravimetry. The electrochemical properties of the resulting tBLM
were characterized using electrical impedance spectroscopy (EIS) and cyclic
voltammetry (CV). The suitability of the interface for ion channel assays was

demonstrated by measuring the impedance drop following gramicidin addition.

5.3. Materials and methods

5.3.1. Preparation of membrane fractions

S. pombe cells were grown in YPD medium (20 g/L yeast peptone, 10 g/L yeast
extract and 2% w/v glucose) at 30°C for 48 h. Cells were harvested by centrifugation at
5,500g for 15 min at 4°C and suspended in lysis buffer (25 mM Tris buffer, pH 8.0,
containing 250 mM sucrose). The cells were then passed through an Emulsiﬂex-C3

homogenizer at 28 kpsi three times. The cell debris was separated using centrifugation at

84

46,300g for 20 rrrin at 4°C. The supernatant was transferred to ultracentrifuge tubes and
centrifuged at 245,000g for 110 min at 4°C. The membrane pellet was frozen at -80°C.
To prepare membrane ﬁactions, a portion of the membrane pellet was suspended in 20
mM Tris-HCl buffer, pH 8.0, containing 200 mM NaCl and 10% glycerol. The
membrane fraction suspension was ultrasonicated for 5 min with 30 s pulses and cooling
for 1 min after each pulse. Finally, the microsome suspension was passed through a 200

nm ﬁlter.
5.3.2. Fabrication of tethered bilayer lipid membrane on gold surface

Cleaned gold electrodes were placed in a chloroform solution containing 1 mM
DPPTE tethering lipid and 0.5 mM DOPC mobile lipids for 24 h. After SAM formation,
the electrodes were washed with chloroform to remove weakly attached lipids and dried

using nitrogen. The electrodes were transferred to an electrochemical cell that maintained

0.2 cm2 of electrode area in contact with electrolyte solution (100 mM sodium chloride).

A liposome (7.86 mg/mL) or membrane-fraction solution (7.6 mg/mL) was added to
electrolyte solution. After allowing 24 h for the bilayer to develop, the solution was
replaced with fresh 100 mM sodium chloride solution. Gramicidin was added as an
ethanolic solution to g’ve a ﬁnal concentration of 1 M protein in the electrolyte solution
[Figure 5.1]. At least one hour was allowed for gramicidin to partition into the bilayer,

and then EIS was performed.
5.3.3. Instrumentation

Transmission electron microscopy (TEM) images were taken using a JEOL

100CX transmission electron microscope (JEOL USA Inc., Cranford, NJ) at an

85

acceleration voltage of 80 kV. A quartz Crystal microbalance (QCM) analyzer (Research
Quartz Crystal Microbalance, Maxtek Inc., Santa Fe Springs, CA) linked to a computer
with Maxtek RQCM Data Log software was used for microgravimetric measurements.

The Maxtek quartz crystals, which oscillated at 5 MHz, were sandwiched between two

gold electrodes having a surface area of 1.26 cmz. The procedure used to clean the QCM

electrodes and establish a tethering-lipid SAM was identical to that previously described
for the gold-coated silicon wafers. After SAM formation, the crystal was mounted on a

crystal holder. Frequency changes were measured at 22°C after obtaining a baseline

oscillation frequency in a 10 mM HEPES buffer (pH 7.4).

DLS, ellipsometry, and electrocherrrical measurements were carried out as

described in Section 2.3.6.

Fluorescence recoveries after pattern photobleaching (F RAPP) measurements
were carried out in collaboration with Aaron Greiner and Dr. R. Y. Ofoli from
Department of Chemical Engineering and Materials Science. The following description
of FRAPP set-up (shown in Figure 5.2) is contributed by Aaron Greiner. FRAPP was
used to measure the translational diffusion and mobile fraction of sBLMs formed on
silica surface (Starr, T. E. and Thompson, N. L., 2002). A double syringe pump system
(Harvard Apparatus, Holliston, MA) was used to infuse and withdraw solutions
simultaneously from a custom-made 1 mL ﬂow cell. The 488 nm laser line from an argon
ion laser (95, Lexel Laser, Fremont, CA) was directed through a 5x beam expander
(Edmund Optics, Inc., Barrington, NJ), 3 ﬁlter cube (Ex: 450-490/DM: 510/Em: 515-
565), and a pair of optical ﬂats that toggle between a low intensity (20uW) monitoring

beam for continuous ﬂuorescence emission measurements, and a high intensity (500

86

mW) beam for photobleaching chromophores. Light passing through the sample was
collected by an inverted microscope (Zeiss Axiovert 135M, Carl Zeiss, Thomwood, NY)
with a Zeiss 32x objective lens. The light intensity was measured by a side-on
photomultiplier tube (PMT, Hamarnatsu R4632, Bridgewater, NJ) connected to a photon
counter (SR400 Stanford Research Systems, Sunnyvale, CA) and a fast preampliﬁer
(SR445 Stanford Research Systems, Sunnyvale, CA). A Ronchi ruling (50 lines per inch,
Edmund Optics, Inc., Barrington, NJ) in the back image plane of the microscope created
a ﬁinge pattern on the sample. An aperture placed in the image plane in front of the PMT
restricted the observation area, resulting in an illuminated area of approximately 200 um
in diameter and an observed area of 75 um. This arrangement prevented unbleached
ﬂuorophores from diffusing from outside the bleached (patterned) area into the
observation zone during ﬂuorescence recovery measurements. A Ronchi ruling was
placed in the rear image plane of the microscope, creating a striped pattern (periodicity of
25 um in the sample plane) of alternating light and dark fringes on the sBLM.
Chromophores inh illuminated zones were photobleached by exposure to the high intensity
laser beam. Unbleached chromophores in the dark zones diffused into the photobleached
areas, leading to recovery of a spatially homogeneous ﬂuorescence signal. The time
constant and functional form of this recovery are given by (Starr, T. E. and Thompson, N.

L.,2002),
m 8 47:21): 1 367:2Dt
f(t)=f(0)+3[l-f(0)] 1—[—,—]exp ——,—— +3exp ——,— 13.1.6.1)
7! a a

where, f(t) is the post-bleach ﬂuorescence intensity (t > 0) normalized with respect to

the constant pre-bleach ﬂuorescence intensity (t < 0), with t = 0 being the time of the

87

bleach pulse, D is the average translational diffusion coefﬁcient of the chromophores

within the bleached area, m is the mobile fraction, and a is the stripe periodicity of the

Ronchi ruling in the sample plane. Equation 5.1 assumes a single diffusing species,
corresponding to the chromophores in the ﬂuid sBLM for our experiments, and an

immobile species of chromophores that reside in intact unruptured vesicles attached to

the surface.

5.4. Results and discussion

5.4.1. Characterization of membrane fraction of yeast cells

Transmission electron microscopy indicated that the membrane fractions had a
spherical shape and size comparable to the small unilamellar vesicles produced using
DOPC [Figure 5.3]. Dynamic light scattering showed a weight-averaged diameter of 20.5
:1: 0.2 nm for the liposomes and 47.4 i 1.4 nm for the membrane ﬁ'action, respectively.
Both of these size distributions lie in the range (less than 200 nm) reported to facilitate

liposome rupture to form sBLM (Cremer, P. S. and Boxer, S. G., 1999).

5.4.2. Characterization of tBLM formation using QCMD

Formation of an upper tBLM leaﬂet by fusion of S. pombe membrane fractions on
the tethering-lipid SAM is analogous to the widely used approach in which liposomes of
synthetic lipids are used to deposit the upper bilayer leaﬂet on an alkanethiol SAM (Peng,
Z. Q. et al., 2002). QCM gravimetry was used to explore the kinetics of microsome

deposition and rupture during tBLM formation. The method involves measuring the

change in oscillating frequency (Af) due to adsorption of material onto the quartz

88

crystal’s surface, and then using the Sauerbrey equation (Sauerbrey, G., 1959) to estimate
the mass adsorbed from the Af data. The tethering lipid SAM and the lipid bilayer are
compact, rigid, structures that couple strongly to the crystal’s motion, resulting in
minimal viscoelastic energy loss and thus low ADisszpation values. In contrast, the
adsorbed vesicles are loosely bound structures that undergo signiﬁcant deformation,

resulting in larger ADiSSipation values. Adsorption of intact vesicles onto an oxidized

gold surface has been shown to cause a large change in ADissipation values (Keller,

(3. A. and Kasemo, B., 1998).
Figure 5.4A compares the Af and ADissipation curves for liposome and

microsome deposition onto silica-coated QCM crystals. The liposome curve exhibits a
rapid decrease to a minimum, followed by an increase to a steady-state value. Keller and
Kasemo (Keller, C. A. and Kasemo, B., 1998) observed such a two-phase trend during
liposome adsorption on silica. In the ﬁrst phase, the intact vesicles adsorb on the surface,
decreasing the oscillation frequency. In the second phase, the adsorbed vesicles rupture

to form a bilayer, releasing entrapped water thereby increasing the frequency. The
presence of some intact adsorbed vesicles on surface contributes to net ADiSSipation.
In our experiments, there was a net change in ADissipation values, suggesting that

not all adsorbed vesicles rupture. The microsome curve showed a continuous decline,

rather than a two-phase trend. The larger magnitude of the Af and ADiSSipdtiOﬂ

values obtained for microsomes than for liposomes suggests that the microsomes’

89

adsorption rate exceeded their rupture rate during the period of the experiment, resulting

in both bilayer patches and intact microsomes on the surface.

Figure 5.4B compares the Af and ADISSipdtiOH curves for liposome and

nricrosome deposition on a tethered monolayer. Both curves exhibit a continuous decline

in Af This trend was previously reported for liposome deposition on a gold surface with

a chemisorbed alkanethiol SAM (Keller, C. A. and Kasemo, B., 1998). In that case, the

steady-state Af value obtained on the alkanethiol SAM was half that on the bare silica,

indicating deposition of only one leaﬂet, rather than two. Also, Af was shown to decrease
with time as a decaying exponential, suggesting a ﬁrst order kinetic model, in which
adsorption and rupture steps occur simultaneously to form the upper sBLM leaﬂet

(Keller, C. A. and Kasemo, B., 1998; Richter, R. et al., 2003). Af values for deposition of

both liposomes and membrane fractions on a tethering lipid monolayer [Figure 5.3A],
decayed continuously but did not follow ﬁrst order kinetics. The change in Af values
obtained for liposome fusion over tethering lipid SAM was greater than the one reported
for liposome rupture over alkanethiol SAM (Keller, C. A. and Kasemo, B., 1998),
presumably because the tethering lipid SAM was not as tightly packed and ordered as the

alkanethiol SAM, allowing mobile lipids from the liposomes to incorporate into the

tethering lipid SAM. Rupture kinetics of liposomes and membrane fractions on the

tethering lipid monolayer was slower than on silica. The higher ADissipation values

indicated that most of the vesicles remained adsorbed on surface during the period of

experiment. In our experience, development of a high-impedance tBLM using liposomes

9O

and microsomes on the tethering lipid monolayer takes several hours, compared to the 25
min duration of the QCM experiments. This observation is consistent with our impedance
results, wherein the membrane resistance required about 10 h to reach the steady-state

value.

The QCM studies also demonstrated that microsome rupture kinetics were slower
than those of DOPC liposomes. Liposome rupture kinetics have been found to be strongly
affected by the phospholipid composition (Richter, R. et al., 2003). Bilayer formation for
liposomes composed of single type of phospholipids occurs faster than liposomes
composed of different phospholipids (Richter, R. P. and Brisson, A. R., 2005). Membrane
fractions are a complex mixture of phospholipids and other membrane constituents that
collectively determine the relative rates of microsome binding and rupture on a surface.
Ohlsson et a1 (Ohlsson, P. A. et al., 1995) found that the rupture of proteoliposomes of

ganglioside GMl receptor protein did not follow the same rupture behavior as liposomes.

The proteoliposomes exhibited a continuous decrease in Af magnitude over an hour

without achieving a steady state value, similar to the trends we observed for microsomes.
However, the adsorbed microsomes did ultimately undergo rupture to form upper leaﬂet

of tBLM as shown by the electrochemical characterization results.
5.4.3. Fluorescence recovery after pattern photobleaching (F RAPP)
study of nricrosomal sBLM on glass

After rupture, BLM should provide a ﬂuid environment for lateral diffusion of
lipid and protein molecules. This study, conducted in collaboration with Dr. R.Y. Ofoli’s

lab, used Fluorescence recovery after photo-bleaching (FRAP) to determine the

91

translational diffusion coefﬁcient of lipids embedded in the BLM, as well as the BLM’s
mobile ﬁaction (Starr, T. E. and Thompson, N. L., 2002; Kohli, N. et al., 2006b;
Lapinski, M. M. et al., 2007). sBLM was deposited on a hydrophilic transparent surface
(mostly, glass) using liposomes or nricrosomes containing around 2% of ﬂuorescently
labeled lipids. Then, a small region of BLM was bleached using intense laser radiation,
and the rate of ﬂuorescence recovery due to a lateral diffusion of ﬂuorescent lipid
molecules in the photo-bleached region was monitored. In FRAPP technique, the photo-
bleaching is performed on a patterned region. A Ronchi ruling (50 lines per inch) was
used to create well deﬁned striped pattern resulting in photobleaching of half of the
illuminated area. Since only 50% area is photo-bleached, a maximum of 50% ﬂuorescent

recovery is possible.

We performed three experiments to study the ﬂuidity of microsomal BLM on
glass surface. In the ﬁrst experiment, modeled after that reported by Eli-calli et al. (Elie-
Caille, C. et al., 2005), liposomes were created containing 1-oleoyl-2-[6-[(7-nitro-2-l,3-
benzoxadiazol-4-yl)amino]hexanoyl]-sn-glycero-3 -phosphocholine (N BD-PC) and 1-
palrrritoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) in a ratio of 2:98. These
liposomes were then mixed with yeast microsomes in 1:1 ratio (dry weight basis). The
mixture was sonicated for 15 min at 40°C using a bath sonicator. The microsome-

liposome mixture was then used to deposit sBLM on glass surface for 3 h. FRAPP

experiments were conducted in several locations across the sBLM, and the D and m

values were calculated by ﬁtting Equation 5.1 to the ﬂuorescence recovery data. In the
second experiment, rrricrosomal suspension was mixed with 2% by weight NBD-PC and

sonicated for 15 min at 40°C. The resulting ﬂuorescently tagged microsomes were used

92

to deposit sBLM on a glass surface for 3 h, and then FRAPP analysis was performed. In
the third experiment, a solvent mixture (chloroformzmethanolzwater in ratio 68:31 :1) was
used to extract lipids ﬁ'om the microsomes using reported procedure (Schneiter, R. and
Daum, G., 2006). The extracted and dried microsomal lipids were then mixed with 2% by
weight NBD-PC and sonicated in buffer at 40°C for 15 min. The resulting liposomes

were used to form sBLMs, and FRAPP analysis was performed.

The model described in Equation 5.1 ﬁtted well to data for POPC sBLM (R2 =

0.98) [Figure 5.5], microsome-POPC sBLM (R2 = 0.93), and microsome lipid sBLM (R2

= 0.96) [Figure 5.6] whereas the model did not ﬁt well to data for pure microsomal

sBLM (R2 = 0.78) [Figure 5.7]. The diffusion coefﬁcient was similar for all four cases

[Table 5.1]. The mobile fractions or the POPC liposome, microsome-POPC liposome
mixture, and liposomes made ﬁ'om extracted rrricrosomal lipids have relatively high
mobile fractions (between 0.67 and 0.81). However, the mobile fraction for the

nricrosomes was substantially lower (0.26). In addition, the standard deviation among

replicate D and m values for the microsomes was considerably larger than for the other

lipid mixtures [Figure 5.7]. It indicated that the model described in Equation 5.1 may not
adequately describe the complexity of nricrosomal BLM. These ﬁndings are consistent
with the hypothesis that the microsomes produce a heterogeneous interface that may

contain a signiﬁcant number of unruptured microsomes. Unbleached, labeled lipids

would not be able to diffuse into or out of unruptured microsomes, resulting in a low m

value. A higher m value obtained for the liposomes produced from extracted microsomal

93

lipids suggests that a non-extracted component of the rrricrosomes, such as membrane
proteins, may slow down the rupture kinetics. From the FRAPP data, it can be inferred
that presence of proteins prolongs the rupture process resulting in lower mobile fraction

that hiders the diffusion of ﬂuorescent species.
5.4.4. Electrochemical characterization of tBLM

The thickness of tBLM deposited using DOPC liposome and yeast membrane
ﬁactions was determined using ellipsometry to be 5.7 :I: 1.0 nm and 6.2 i 0.9 nm,
respectively. The slightly higher thickness for tBLM formed using membrane fractions
may have been due to the presence of yeast membrane proteins in the resulting tBLM.
The EIS data for a tBLM formed using DOPC liposomes [Figure 5.9A] and S. pombe

microsomes [Figure 5.9B], were analyzed using to a Randles equivalent circuit model

[Figure 5.10]. Z-view software was used to obtain membrane resistance (RM) values of
1.59 MQcm2 and 1.62 Mﬂcm2 for the liposomes and microsomes, respectively, and

membrane capacitance (CM) values were 0.78 11F/cm2 and 0.68 p.F/cm2 for the

liposomes and rrricrosomes, respectively. In both cases, these data are consistent with
production of a highly insulating tBLM. Figure 5.11 shows the impedance spectrum for
tBLM taken at different times. The membrane’s resistance increased and its capacitance
decreased consistently, until steady-state was reached somewhere between the 8 h and 22
h time points. These data are consistent with our QCM results, which suggested that
rrricrosome rupture to form a tBLM requires longer than 30 min. The highly insulating

nature of the tBLMs produced using microsomes was further conﬁrmed by CV, wherein

94

the redox peaks for ferro/ferricyanide system were totally subdued, and only a curve

representative of charging current was obtained [Figure 5.12A].

The suitability of the microsome-derived tBLM for characterizing ion charmel
activity was evaluated using gramicidin. Active gramicidin channels are dimers, in which
one monomer exists in each leaﬂet of the bilayer, and aligrrrrrent of the monomers creates
a channel through the entire bilayer. Gramicidin selectively passes alkali metal ions, so
the presence of these ions results in a decrease in membrane resistance (N aurnann, R. et

al., 2003b). Incorporation of gramicidin in tBLM formed with membrane ﬁactions

decreased the membrane resistance from 1.62 MQcm2 to 0.43 Mﬂcm2 [Figure 5.133]. A

similar resistance drop (ﬁ'om 1.61 MQcm2 to 0.10 kﬂcmz) was observed following

incorporation of gramicidin into tBLM formed using DOPC liposomes [Figure 5.13A].
The gramicidin-containing tBLM produced using microsomes was further characterized
using CV. As shown in Figure 5.12B, addition of gramicidin did not yield a signiﬁcant
redox peak characteristic of ferricyanide, despite the observed decrease in resistance
measured using EIS. Because gramicidin selectively passes sodium but not ferricyanide,
these results are also consistent with a defect-free tBLM in which ion channels can
exhibit their characteristic activities. '

The results presented above establish for the ﬁrst time that a high-impedance
tBLM can be formed using S. pombe microsomes. The QCM results showed that both
microsomes and DOPC liposomes adsorb strongly onto both silica and a tethering lipid
SAM, but that the rate of microsome rupture is slower for than for DOPC liposomes. The

EIS and CV results showed that tBLM produced using microsomes have similar

95

membrane resistance and capacitance to those produced using DOPC liposomes, and

allow the activity of ion channels to be measured.

This novel approach extends previously reported methods that used microsomal
preparations to characterize the activities of membrane proteins that do not require an
insulating membrane on interface, such as cytochrome c oxidase, Na-K ATPase and
mitochondrial membrane proteins (Elie-Caille, C. et al., 2005). The ability to fabricate a
high impedance tBLM using membrane fractions suggests the possibility that
recombinant membrane proteins overexpressed in S. pombe could be directly
incorporated into tBLM on gold electrodes without the need to (1) separate the proteins
ﬁ'om the membrane, (2) purify the proteins, and (3) reconstitute the proteins into
liposomes. This approach would be particularly well suited to characterize membrane
proteins that lose activity during one of these processing steps. It could also simplify the
use of biorrrimetic interfaces to characterize membrane proteins advancing efforts to
rapidly characterize novel proteins in a high-throughput mode [for drug screening or

functional proteomics research.

5.5. Conclusions

Membrane fractions recovered from homogenized S. pombe cells were
characterized using transmission electron microscopy and dynamic light scattering and
found to resemble small, unilamellar liposomes. QCM results indicated that the
microsomes and DOPC liposomes exhibited similar adsorption kinetics on the tethering-
lipid SAM, characterized by a continuous increase in adsorbed mass over the 25 minute
experiment. On silica, a similar QCM proﬁle was exhibited for the microsomes, but the

liposomes exhibited distinct adsorption and rupture phases. EIS and CV conﬁrmed that

96

both liposomes and microsomes formed highly insulating tBLM on gold that had high
membrane resistance values. The high impedance tBLM formed using membrane
ﬁactions effectively mimics the properties of the biological cell membrane and is suitable
model to study membrane protein activities. The subsequent addition of granricidin
resulted in a substantial drop in membrane resistance, indicating the ﬂow of ions across
the tBLM. These results demonstrate that tBLM derived ﬁom microsomes are suitable
for ion channel studies. This novel method may provide a means to establish a
biomimetic interface consisting of recombinant membrane proteins in a tBLM without
the need to separate the target protein from the membrane, purify it, and then reconstitute
it into proteoliposomes. This capability could streamline efforts to characterize novel

proteins and screen drug candidates for activity against membrane proteins.

97

I

 

 

. 1111111111.

 

      

   

 

 

 

.. . '. .-1. «I. \, t ,.(,‘I. . “7'
2*?“36332wﬁzwwvm¢~4u3ﬁ4mi - - ._.,_ ., ......... INN“ ..~,
e e Ing I!“ "A ' ..-—".1... "‘ u" . m Lire-:1" ..."..- . ...—'m. ...13- 3 6.
(;Id I Ct d Iﬂﬂds

 

 

Liposomes Microsomes
V

II a, .11 111511-

«f I T ' '
I '- ’7 on
Eﬁwﬁwwﬁgqt,,mg._.ggﬁ channel

Microsome tBLM containing protein tBLM with microsome
ion channel protein proteins

 

 

 

, z _
.......

 

Figure 5.1: Fabrication of tBLM using liposome or yeast membrane fractions. A self
assembled monolayer of tethering lipid was deposited on the gold substrate, and the
liposomes or microsomes were added to deposit the upper tBLM leaﬂet. Gramicidin was
incorporated into the tBLM, and gramicidin-mediated ion transport across the tBLM was

monitored using EIS. [Figure not to scale]

98

 

 

  

Laser light
Diehroic I I mf— (488 nm)
mrrror I

iv Filter Ronchi

Filter - ruling

-3. To PMT

 

Figure 5.2: Schematic diagram of the FRAPP experimental setup. sBLM: supported
bilayer lipid membrane; PMT: photomultiplier tube. [Figure not to scale] -

99

 

Figure 5.3: Transmission electron microscopy (TEM) image of (A) DOPC liposomes,
and (B) membrane fractions of S. pombe yeast cells. Liposome and membrane fractions
were stained using 1% uranyl acetate solution. The size of liposome and membrane
fractions was well below 200 nm which would be necessary for facile rupture of these
vesicles to yield planar bilayer membranes.

100

— 5.00E-06

 

 

 

 

 

-- 4.005-06
—- 3.00E-06 3
Hg
1'
-- 2.005-06 .%
52
-_ Q
1.00an q
-- 0.005+00
'140 I T I r I -1oOOE-06
0 5 10 15 20 25
Time (min)
B 10 - -~ 3.90E-06
0
-10 - \ ., ' ' H ' -~ 2.90E-06
.. .. ' e
A ‘20 ‘ f .. \2a a
g 1 a . cu
v -30 - . ‘5 -- 1.905-06 .%
‘h ...
<1 40 1 g
Q
-50 - \ -- 9.005-07
50 - m1]? 1a \2
-70 I I r I I -1.005-07

 

 

0 5 10 15 20 25
Time (min)

Figure 5.4: Quartz crystal microbalance study for formation of tBLM using liposome
and tBLM using yeast microsomes. (A) Liposome deposition directly on the silica

surface is shown in Curve 1 (A f data) and Curve 1a (ADissipation data). Liposome
deposition on the tethering lipid SAM is shown in Curve 2 (A f data) and Curve 2a
(ADissipation data). (B) Microsome deposition directly on the silica surface is
shown in Curve 1 (Alf data) and Curve 1a (ADissipation data). Microsome
deposition on the tethering lipid SAM is shown in Curve 2 (Zlf data) and Curve 2a
(ADissipation data).

101

 

Normalized Fluorescence Intensity

 

 

Residual fluorescence

 

Time (s)

Figure 5.5: Fluorescence recovery curves for sBLM on glass formed using POPC
liposomes incorporated with 2% NBD-PC obtained using F RAPP.

102

9.0.0
.853

0.25 -

5 .

0.15-

0.10-

.o
f

0.00-

 

.5
8

 

Residual fluorescence

 

Normalized Fluorescence Intensity

l
01

Figure 5.6: Fluorescence recovery curves for sBLM on glass formed using liposomes

Time (6)

obtained using lipids extracted from microsomes and mixed with 2% NBD-PC.

103

0.25—
0.20-
0.15:
0.10—
0.05-

0.00-

 

 

Residual fluorescence

/

 

Normalized Fluorescence Intensity

Figure 5.7: Fluorescence recovery curves for sBLM on glass formed using nricrosomes

incorporated with 2% NBD-PC.

F'I
15

I ' I ' I ' I

20 25 30 35 40

Time (s)

104

 

 

 

 

 

 

 

       

 

 

 

 

 

 

 

 

 

 

 

 

 

2.5 _ IDiffusion coefﬁcient
A [1] Mobile fraction
N” 2 -1
E
5';
E a. 1.5 —
’5 o
o g 1 -
o
g I
'6
g 0.5 —
D
0_ , Illllll
POPC sBLM Microsome- Microsome Extracted
POPC sBLM sBLM microsome
(1 :1 ) lipids sBLM

Figure 5.8: Diffusion coefﬁcient and mobile fraction data obtained using FRAPP were
plotted for different sBLMs deposited on glass substrate.

105

 

2.5 I I I T I I I
-2 -1 0 1 2 3 4
Log Frequency

 

 

 

2.5 I I I I I I
-2 -1 0 1 2 3 4

Log Frequency (Hz)

Figure 5.9: Comparison of tBLM formed with DOPC liposome (A), and with membrane
ﬁactions of yeast cells (B) over tethering lipid SAM using EIS. Impedance data (Log Z)
were plotted as a function of Log frequency for tethering lipid monolayer (triangles) and
tBLM (squares). An equivalent circuit shown in Figure 5.8 was used to ﬁt impedance
data. The solid lines passing through the data points represent model ﬁt. The resistance of
membrane increases whereas the capacitance of membrane decreases after the formation

of bilayer membrane. [Electrode area is 0.2 cm2]

106

 

 

 

 

 

Figure 5.10: A Randles equivalent circuit was ﬁtted to the impedance data to calculate
the values of electrochemical characteristics. Rs represent the resistance of solution, CM
represents the capacitance of membrane, RM represents the resistance of membrane and
CDL represents the capacitance of double layer that includes the Hemholtz capacitance.

107

 

     

 

 

6.5 -
Increasing time of

A incubation
g, 5.5 —
N
3’
.1 4.5 -

3.5 4

-2 -1 0 1 2 3 4
Log Frequency (Hz)

Figure 5.11: Impedance spectrum for tBLM recorded at different intervals after addition
of microsomes. The order of curves from bottom to top is- tethering lipid monolayer,
tBLM at 30 min, tBLM at l h, tBLM at 2 h, tBLM at 3 h, tBLM at 5 h, tBLM at 8 h,
tBLM at 22 h and tBLM at 24 h. The membrane resistance increased whereas the
membrane capacitance decreased with time of incubation. The initial rate of bilayer
formation was higher, which decreases with time. The impedance spectrum for tBLM at
22 h, and 24 h were same indicating that the tBLM formation is complete. The electrolyte
solution was replaced after the tBLM formation. The impedance spectrum of tBLM did
not change after replacing the electrolyte solution with fresh 100 mM NaCl solution.

[Electrode area is 0.2 cm ]

108

00
O
l

 

 

Current (pA)

 

.25 I T I I I I
400 300 200 1 00 O -1 00 -200

 

 

 

 

 

 

.5
.b

 

-006 I I I I I I I
400 300 200 1 00 0 -1 00 -200

Potential (mV)

 

Figure 5.12: (A) Cyclic voltarnmograrn of bare gold (curve 1) was plotted along with the
voltarnmograrn obtained after tBLM fabrication (curve 2). The redox peaks were subdued
after the formation of tBLM. (B) Cyclic voltarnmograrn for tBLM formed using
membrane ﬁactions of yeast cells (curve 1), and microsomal tBLM after gramicidin
incorporation (curve 2). The redox peaks were absent in both tBLM voltammograms
whereas there was an increase in charging current aﬁer the addition of gramicidin
suggesting the passage of ions through tBLM by gramicidin while excluding the passage

of ferricyanide ions through tBLM. [Electrode area is 0.2 cm2]

109

 

 

4.5 I I I I I I

 

-2 -1.5 -1 -o.5 o 0.5
Log Frequency (Hz)

A

 

 

 

4.5 r I I I I ‘1
-2 -1.5 -1 -0.5 0 0.5 1

Log Frequency (Hz)

Figure 5.13: Measurement of gramicidin ion channel activity using EIS. (A) tBLM was
fabricated by fusion of DOPC liposome over tethering lipid monolayer and (B) tBLM
was fabricated by fusion of membrane fractions of yeast cell over tethering lipid
monolayer. Impedance (Log Z) data for tBLM (squares) and tBLM after gramicidin
incorporation (triangles) were plotted as function of Log frequency. Gramicidin was
added as an ethanol solution to give a ﬁnal concentration of 1 uM. Membrane resistance

drops due to the passage of sodium ions through tBLM. [Electrode area is 0.2 cm2]

110

 

6. ELECTROCHEMICAL CHARACTERIZATION OF
INTERACTIONS BETWEEN NANOPARTICLES AND
SUPPORTED BILAYER LIPID MEMBRANES

6.1. Abstract

Electrochemical techniques were used to investigate the interactions of polymeric
nanoparticles (NPs) with 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) supported
bilayer lipid membranes (sBLM) on a glassy carbon electrode. For the ﬁrst part of the
study, PAMAM dendrimers were used as model nanoparticles. PAMAM dendrimers are
hyperbranched monodisperse polymeric nanoparticles whose molecular size, molecular
weight, and number of functional groups can be controlled. Dendrimers of generation 2
through generation 4 (62-64) did not cause signiﬁcant damage to the sBLM, whereas
GS-G7 dendrimers created large defects in sBLM. In cyclic voltammetry (CV) studies,
the peak current obtained after dendrimer addition was used to estimate the defect area.
Electrochemical impedance spectroscopy (EIS) showed an increase in capacitance values
after addition of G5-G7 dendrimers, consistent with a decrease in sBLM surface
coverage. Nanoparticle mass, diameter, and surface charge density all increase
signiﬁcantly with generation number, suggesting that these variables may inﬂuence
nanoparticles’ ability to rupture sBLM. To further understand the effect of functional
group and charge density of the NPs, custom nanoparticles were synthesized having a
polyethylene glycol (PEG) backbone. Both positively and negatively charged NPs were
able to disrupt sBLM having a zwitterionic head group. PEG NPs having a functional
group with multiple charged groups created larger defects in sBLM suggesting that

higher charge density increases the NP-BLM interaction. The method presented in this

111

study provides a convenient and sensitive way to detect interactions between
nanoparticles and lipid bilayers, and may be useful in evaluating nanoparticles’ safety

proﬁle or suitability as drug or gene delivery vehicles.
6.2. Introduction

Recently, nanoparticles (NPs) have been introduced in various commercial
applications including cosmetics, paints, textiles and others. The small size of NPs
provides them unique properties. Incorporation of NPs in biosensor interfaces increases
the surface area for enzyme adsorption by several folds, thereby increasing enzyme
loading and in turn, the sensitivity of biosensor. In medical applications, a low
irnmunogenicity of NPs has been exploited by using NPs as gene and drug delivery
vehicles. Charged polymeric NPs, in particular, are highly effective in disrupting lipid
bilayers, mediating the internalization of molecular cargos into the host cell. This
property raises a concern that NP-mediated defect formation in cell membranes might
lead to cytotoxicity. Therefore, a detail understanding of NP-BLM interactions is
required.

In this study, a supported BLM interface was developed for studying the
interaction of NPs with bilayer lipid membranes. Monodisperse positively charged
polyamidoamine (PAMAM) dendrimers of different size and generation number were
used as model NPs to study the effect of nanoparticles’ physical properties on their ability
to disrupt BLMs. Polyarnidoarnine (PAMAM) dendrimers are hyperbranched polymeric
NPs that are formed in concentric layers [Figure 6.1]. As each subsequent layer
(generation) is added, the molecular size, molecular weight, and number of functional

groups increases by a known amount [Table 6.1]. This property makes dendrimers well

112

 

suited for use as monodisperse NPs having a desired size, shape, and surface charge
(Zeng, F. W. and Zimmerman, S. C., 1997). Recently, PAMAM dendrimers have also
been investigated as non-viral vectors for the gene delivery (Dufes, C. et al., 2005; Zhou,
J. H. et al., 2006) that offer the potential for low immunogenicity (Malik, N. et al., 2000;
Duncan, R. and 1220, L., 2005) and may protect the gene from DNase activity. The
mechanism of gene delivery involves internalization of the charged polymer by host
cells, though a mechanism that may involve disruption of the host cell membrane
(Domanski, D. M. et al., 2004). Dendrimers’ ability to cross cell membranes makes also
makes them an alternative vehicle for drug delivery (Esfand, R. and Tomalia, D. A.,
2001; Crarnpton, H. L. and Simanek, E. E., 2007). Atomic force microscopy studies have
shown that PAMAM dendrimers form 1540 nm holes in 1,2-dimyristoyl-sn-glycero-3-
phosphocholine (DMPC) supported bilayer lipid membrane (sBLM) on mica (Diao, P. et
al., 1999; Mecke, A. et al., 2004; Hong, S. P. et al., 2006). These studies concluded that
the high surface charge density of amino groups found in higher generation dendrimers

was responsible for disruption of bilayer membranes.

sBLMs are more robust and stable than unsupported planar bilayer membranes
and can be deposited on a variety of hydrophilic surfaces, such as oxidized glassy carbon,
silica, and mica (Sackrnann, E., 1996). Even though a sBLM is only about 5 nm thick, it
is an excellent electrical insulator. Thus, disruption of the sBLM by NPs could, in
principle, be measured electrochemically. The approach described in this study provides
a rapid, sensitive, and convenient electrochemical method to detect disruption of sBLM
by NPs. The approach involves fabricating a sBLM on a glassy carbon electrode (GCE),

and then using electrochemical methods to monitor interactions between NPs and the

113

sBLM. Disruption of a l,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) sBLM formed

on a GCE was investigated using cyclic voltammetry (CV) and electrochemical

impedance spectroscopy (EIS).

6.3. Methods and instrumentation

6.3.1. Method

Each GCE (Bioanalytical Systems, West Lafayette, IN) was sequentially polished
with 1000, 300 and 50 nm alumina slurry, followed by washing with de-ionized (DI)
water and methanol. The electrode was ultrasonicated for 2 min in DI water to remove
physically adsorbed alumina, given ﬁnal rinses with methanol and DI water, and then
dried under a nitrogen stream. The GCE was placed in 100 mM NaCl solution and
oxidized at a potential of ISOOmV for 3 min(Du, L. W. et al., 2006). It was then washed
with DI water and dried under a nitrogen stream. Five uL of DOPC lipid solution (5
mg/mL in chloroform) was applied to the electrode surface, and the electrode was
immediately immersed in an electrolyte solution (100 mM sodium phosphate buffer, pH
7.4, containing 1 mM of potassium ferrocyanide and 1 mM of potassium ferricyanide).
After 20 min, PAMAM dendrimers (G2 - G7, provided by Dr. Steve Kaganove,
Michigan Molecular Institute, Midland, MI) were added to obtain a ﬁnal concentration of
20 uM (based on the surface amino groups). PEG NPs were added to a concentration 1
uM (based on the molecular weight). CV was then used to monitor changes in the
sBLM’s electrochemical properties until a steady-state situation was observed [Figure

6.1].

114

 

6.3.2. Instrumentation

Electrochemical measurements were performed using a CHI66OB electrochemical
workstation (CH Instruments Inc., Austin, TX). EIS was performed in 100 mM sodium
phosphate buffer, pH 7.4, containing 1 mM potassium ferrocyanide and 1 mM potassium
fenicyanide. The DC potential was E’ = 220 mV vs a Ag/AgCl reference electrode, and a
5 mV sinusoidal potential was applied across the frequency range of 0.1 to 10,000 Hz. A
modiﬁed Randles equivalent circuit model (Bard, A. and Faulkner, L., 2001) was ﬁtted
to the impedance data using Z-view software (Scribner Associates, Southern Pines, NC).
CV was performed in same electrolyte solution as used for E18 studies. The potential was

cycled in a range of 500 mV to -200 mV at a scan rate of 50 mV/s.

6.4. Results and discussion

6.4.1. Interface formation
The CV and E18 characterization of sBLM have been discussed in Section 2.4.2.
6.4.2. Interaction of PAMAM dendrimers with sBLM

After sBLM formation, PAMAM dendrimers were added to the electrolyte
solution, and their interaction with the sBLM was monitored electrochemically. The
system typically reached steady state 15 min after the addition of dendrimer solution, but
the electrochemical parameters were monitored for 30 min to ensure a constant value.
The dendrimer concentration was calculated based on surface amino groups to ensure that
the same number of charged groups was used in all experiments. The CV curves obtained
after addition of G2 and G3 dendrimers are shown in Figure 6.2. No redox peaks were

obtained for G2, and only a small peak current of 0.12 uA was obtained for G3. These

115

 

results indicate that lower generation dendrimers cause minimal defects in the sBLM, and
are consistent with the AF M study by Mecke et al (Mecke, A. et al., 2005b), which found
that G2 and G3 primarily adsorbed on the sBLM, without creating obvious holes. The CV
curves obtained after addition of G4-G7 dendrimers are shown in Figure 6.3. The peak
current increased roughly linearly with generation mnnber for G5 — G7, suggesting that
larger dendrimers more effectively create defects in the bilayer, through which the redox
species can reach the electrode surface. The peak current values for G2-G7 dendrimers
are summarized in Table 6.1. These values were used with the Randles-Sevick equation
[Equation 2.2] to calculate the approximate sBLM defect area value for each generation

(Bard, A. and Faulkner, L., 2001)

ip = (2.69 X 105) 113/214 CBULK (DV)1/2 (Eq 2.2)

where, ip is the peak current (uA), n is the number of electrons transferred (1), A is
electrode area exposed (for bare GCE = 0.0706 cm2), C BULK is the bulk concentration of

redox species (1><10-6 molcm-3), D is the diffusion coefﬁcient ((8.4><10-6 cmz/s) and V is
the scan rate (0.05 V/s). The diffusion coefficient of the redox species was calculated
ﬁom the slope of the plot of peak current vs scan rate for a bare GCE surface.

EIS results indicated that the overall capacitance increased signiﬁcantly with

generation number, presumably due to a larger fraction of electrode area being exposed to

the electrolyte solution [Figure 6.4], and hence a signiﬁcant CDEFECT value. Large peak

current and higher capacitance values obtained for high generation dendrimers were a

direct measure of the ability of dendrimer to pull out lipids from the sBLM. Based on the

116

 

data provided in Table 6.1, G7 dendrimers have a 36 times greater mass, 2.6 times greater

diameter, and a 4.6 times greater surface charge density than G2 dendrimers.

We noted that the molecular mass, diameter, and surface charge density of
PAMAM dendrimers all increase with generation number. Therefore, the relative role of
nanoparticle properties in determining their ability to disrupt BLMs was not fully
understood ﬁ'om these studies. Mecke et al (Mecke, A. et al., 2004) and Hong et al
(Diao, P. et al., 1999; Hong, S. P. et al., 2006) have studied defect formation by PAMAM
dendrimers in DMPC sBLM on mica using AFM. A mechanism in which the dendrimer
molecules take off the lipids from the sBLM forming so called ‘dendrimer ﬁlled vesicles’
was proposed (Mecke, A. et al., 2005a; Mecke, A. et al., 2005b). They reported that the
size and the surface charge density are both important properties of a dendrimer that
would act together in determining its efﬁcacy to interact with BLM. Our results indicated
that a nanoparticle may have to have sufﬁcient size for enough interactions between the
charged functional groups of NP and the head group of phospholipids. A nanoparticle
with a particle size equal to or greater than thickness of BLM (5 nm) might be required
for enough NP-BLM interaction that results in disruption of BLM. Recently, Ginzburg
and Balijepalli (Ginzburg, V. V. and Balijepailli, S., 2007) developed a thermodynamic
model for interaction of NPs with sBLM. Their simulations show that the NPs need to
have sufﬁcient size to have interaction with large number of phospholipids molecules and
create a defect in BLM. They concluded that the diameter of a nanoparticle should be
comparable or greater than the thickness of bilayer membrane. Our results are in

agreement with their simulation results.

117

 

6.4.3. Interaction of PEG NPs with sBLM

Three main parameters associated with a nanoparticle can be deduced as critical
ones for its interaction with lipid bilayer membranes, namely the size of a nanoparticle,
the surface functional group (and thereby surface charge), and the surface charge density
of a nanoparticle. The issue of nanoparticle size has been well addressed with the study
on PAMAM dendrimers. Different polycationic polymers have been reported in literature
for their interaction with sBLM, such as polyethyleneirnine (PEI), polylysine and
diethylarninoethyl-dextran (Mecke, A. et al., 2005a; Hong, S. P. et al., 2006; Leroueil, P.
R. et al., 2008). All of these polycationic polymers were found to be effective in
disrupting sBLM. A recent report by Leroueil et a1 (Leroueil, P. R. et al., 2008) has
shown that inorganic NPs such as amine coated gold and silica NPs are effective in
disrupting supported BLMs. These reports further support the hypothesis that NPs with a

particle size equal to or greater than thickness of BLM are effective in disrupting BLMs.

Further studies wee required to investigate how a presence of charge on a NP,
type of surface charge, and the surface charge density affects the NP-BLM interaction.
We proposed a hypothesis that a charged surface is essential for BLM-NP interaction.
Further, for a sBLM formed using zwitterionic lipids, the NP-BLM interaction should be
independent of type of charge on NP. Finally, increase in the charge density of
nanoparticle shall increase the BLM disrupting efﬁciency of the NP. To test these
hypotheses, novel polyethylene glycol (PEG) NPs synthesized by Dr. Baker’s group in
Department of Chemistry at MSU were used. These nanoparticles were synthetically
tailored to provide desired size, functionality and charge density to the NP. The structures

of these NPs have been shown in Figure 6.5. The basic backbone was composed of

118

 

hydrophobic as well as hydrophilic groups, providing similar chemical enviromnent as
that of PAMAM backbone (Diao, P. et al., 1999; Baker, G. L. et al., 2008). These novel
NPs were not monodisperse but they have a very narrow particle size distribution. The
particle size for all of these NPs varied from 15-30 nm. The diameter of PEG NPs was
greater than 5 nm and hence, they satisfy the criteria suggested by the thermodynamic
model. PEG NPs have been custom designed to give particular functionality and charge
on the surface. The monomers of these polymeric NPs have been linked via polyethylene
glycol chain. These NPs were functionalized to provide cationic groups such as amine
(PEG amine) and guanidine (PEG guanidine) groups on surface. PEG guanidine NPs
would provide higher cation density per functional group compared to PEG amine NPs.
One of the NPs has anionic carboxylic acid groups (PEG acid) on the side chain which
will be exposed on surface. The basic uncharged PEG NPs did not have any
functionalized side chain and hence did not have a surface charge. These NPs provided
range of functional groups that can be assessed for their role in determining the efﬁciency

of the interaction of NPs with sBLMs.

Similar to the PAMAM study, sBLM interface was assembled on GCE. CV and
EIS were used to characterize the PEG NPs’ interactions with sBLM. The nanoparticle-
mernbrane interaction was quantitated using the peak current data obtained from CV and
the capacitance values obtained ﬁom impedance data. The peak current data is shown in
Table 6.2. Uncharged PEG NPs had negligible effect on sBLM [Figure 6.6] whereas all
other three NPs namely PEG amine [Figure 6.7], PEG acid [Figure 6.8] and PEG
guanidine [Figure 6.9] caused signiﬁcant defect formation in sBLM. PEG amine and

PEG acid were found to be almost equally effective in disrupting sBLM. PEG guanidine

119

 

was 55% more effective in disrupting sBLM than PEG amine. EIS results indicated that

following the addition of PEG guanidine NPs, the membrane capacitance of sBLM

increased from 0.75 uF/cm2 to 15 [.lF/sz [Figure 6.10] corresponding to a surface

coverage of only 0.44 (56% of sBLM was removed). These results suggested that surface
charge is essential for a NP to interact with BLMs. NPs should have suitable functional
group that can interact with the head group of the phospholipids. The charged groups of
polymers interact strongly with head group of phospholipids making membrane lose its
planar morphology and reduce its free energy by creating a hole (Ginzburg, V. V. and
Balijepailli, S., 2007). In our studies, both positively and negatively charged NPs
disrupted sBLM formed using phospholipids with zwitterionic head group. The higher
efﬁciency of PEG guanidine NPs in disrupting sBLM compared to PEG amine suggested
that increasing the charge density of a NP leads to greater NP-BLM interaction. This
study establishes that the molecules used in this study can be potential gene or drug
delivery agents and provides a basis for designing new drug and gene delivery agents.
The approach will also facilitate development of new methods to assess the toxicity of the

NPs to be used in products intended for human or animal use.
6.5. Conclusions

A biomimetic interface consisting of a sBLM deposited on a GCE was used to
study the interaction of various NPs with lipid bilayers. CV and E18 studies showed that
dendrimer addition dramatically decreased impedance and increased ferricyanide peak
currents, presumably by creating defects in the sBLM. Higher generation dendrimers had
a much stronger effect on the sBLM than did lower generation dendrimers. CV data were

used to estimate the defect area as a function of generation. ElS results showed an

120

 

increase in overall capacitance values following dendrimer addition, consistent with a
decrease in sBLM surface coverage. PEG NPs were used to assess the effect of different
surface charge and charge density on the BLM disrupting ability of NPs. The study
showed that suitable surface charge is essential for NPs to interact with BLMs. Higher
charge density provided better efﬁciency. The sBLM interface provides a rapid and
convenient way to screen NPs for strong interactions with lipid bilayers and may have
utility in assessing nanoparticles’ toxicity or suitability for intracellular genes or drug

delivery.

121

 

 

 

 

 

 

 

 

 

 

 

 

Surface _ Area of electrode

Molecular ammo game“ exposed after NP-

Sample weight groups diameter BLM interaction

per (nm) 2
molecule (mm )

Bilayer N/A N/A 2.2 0.044 :t 0.008
PAMAM G2 3256 16 2.9 0.082 i 0.015
PAMAM G3 6909 32 3.6 0.088 d: 0.010
PAMAM G4 14215 64 4.5 0.430 d: 0.023
PAMAM G5 28826 128 5.3 1.761 t 0.164
PAMAM G6 58048 256 6.7 2.902 t 0.190
PAMAM G7 116493 512 7.6 4.077 d: 0.156
Bare GCE * N/A N/A N/A 7.065 t 0.324

 

 

 

 

 

 

 

Table 6.1: Characteristics of PAMAM dendrimers, along with the parameters showing
their effect on sBLM.

122

 

 

 

 

 

 

 

 

 

 

 

Area of electrode
N rti 1 Molecular Peak current exposed after NP'
anopa c e weight (11A) BLM interaction
2
(mm )
Uncharged PEG NP 98000 0.126 :1: 0.021 0.057 :1: 0.012
PEG-amine NP 87000 3.767 :1: 0.227 2.251 :1: 0.131
PEG-guanidine NP 68800 5.30 :1: 0.604 3.503 :1: 0.062
PEG-acid NP 82100 3.433 :1: 0.108 2.119 :1: 0.347

 

 

 

Table 6.2: Characterization of PEG NPs’ interaction with sBLM studied using CV.

[Electrode area of GCE is 7.064 m2]

123

 

* "5" . cccce

 

 

Carbon electrode

DOPC
lipids

 

   

 

1 i *
5mg [5 mgr trauma

{3 : pAMAM 8% l3 803318110888“
sBLM with defects dend'ime' sBLN'I on carbon
e ectro e

Figure 6.1: sBLM formation on GCE. GCE was ﬁrst oxidized at +1.5 V to impart
negative charge. DOPC sBLM was self assembled on oxidized GCE by depositing DOPC
lipids, solubilized in chloroform, on GCE and immediately transferring the interface in an
aqueous electrolyte solution. sBLM stability was tested using EIS and CV. Nanoparticles,
when added in required concentration, disrupted sBLM and exposed electrode area which
can be calculated from the redox peaks obtained using CV. In one of the proposed
mechanisms, dendrimers form so called ‘dendrimer ﬁlled vesicles’ after their interaction
with sBLM (Mecke, A. et al., 2005b).

124

 

0.15 —

 

 

Current (uA)

 

-001 l I I I I 7 I I
500 400 300 200 100 0 -100 -200

Potential (mV)

 

Figure 6.2: Cyclic voltarnmograrns showing the effect of lower generation PAMAM
dendrimers on sBLM: PAMAM G2 (curve 2) and PAMAM G3 (curve 3). [Electrode

area: 7.064 m2]

125

 

 

 

 

 

 

 

 

8_ 7 _
A6.—
< 4- 6 K i
3 5
E 2‘ 4 It
s. i ' ‘\
3-2—IV 1
(J

4‘ /

-5-

-8 I I I I I P I

 

500 400 300 200 100 0 -100 -200
Potential (mV)

Figure 6.3: Cyclic voltarnmograms showing the effect of higher generation PAMAM
dendrimers on sBLM: PAMAM G4 (curve 4), PAMAM GS (curve 5), PAMAM G6
(curve 6), and PAMAM G7 (curve 7). Curve 1 represents the sBLM before dendrimer

addition. [Electrode area: 7.064 m2]

126

 

 

 

 

 

1.5 I I I I I I T I I 1
-1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4
Log Frequency (Hz)

Figure 6.4: Impedance (Log Z) data were plotted as a ﬁmction of Log frequency for a
bare GCE surface (triangles), sBLM (squares), and sBLM aﬁer interaction with PAMAM
G5 dendrimers (diamonds). An equivalent circuit shown in Figure 2.9 was used to ﬁt the
impedance data. The disruption of sBLM by PAMAM G5 increased the membrane

capacitance and decreased the membrane resistance. [Electrode area: 7.064 m2]

127

 

 

 

N

tth~

91

IN:
Rz-N ’N /
O 0
O H
0 ° ° ‘9'
O K 0
N“
N“Ns
Ra

 

 

 

 

 

 

 

 

 

 

 

 

 

 

R R Molecular

NPS 1 3 weight

PEG -H 98000

PEG-amine -(CH2)5-NH2 87000

-(CH20CH2)5-

PEG: -(CH2)5-NHCHNHNH2 68800

guamdme
- CH -oco— CH -
PEG-acid ( 2)6 ( 2);: 82100
coon

 

Figure 6.5: Structures of new PEG NPs synthesized using ‘click’ chemistry by Dr. Baker
group in Department of Chemistry at MSU. The basic backbone of the structure is shown

above the table. The group -R2 is hydrogen (-H) in all NPs.

128

.o .o
N 03
l l

 

 
  

.0
o .3
l l

 

Current (pA)

b
_\
\

 

 

'0.2 I j I I I I I
500 400 300 200 100 0 -100 -200
Potential (mV)

Figure 6.6: Cyclic voltarnmograrns showing the effect of uncharged PEG NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG uncharged NPs (curve 2).

[Electrode area: 7.064 m2]

129

 

 

-h 01
1 l

l

 

I

\1

1

Current (pA)
N —‘ 0 —¥ N 00

l

 

 

l
00

I I l I l I l

500 400 300 200 100 0 -100 -200
Potential (mV)

Figure 6.7: Cyclic voltarnmograrns showing the effect of PEG amine NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG amine NPs (curve 2). [Electrode

area: 7.064 mm ]

130

 

03 4:5
1 l

 

JOAN
ll
' N
I
I
I
I

1

Current (pA)

['6
l

 

 

'3 I I I I I I 7

500 400 300 200 100 0 -100 -200
Potential (mV)

Figure 6.8: Cyclic voltammograrns showing the effect of PEG acid NPs on sBLM:
sBLM (curve 1%, and sBLM after interaction with PEG acid NPs (curve 2). [Electrode

area: 7.064 mm ]

131

 

 

Current (pA)

 

.4 P 7 I I I I I
500 400 300 200 1 00 0 -100 -200

Potential (mV)

 

Figure 6.9: Cyclic voltarnmograrns showing the effect of PEG guanidine NPs on sBLM:
sBLM (curve 1), and sBLM after interaction with PEG guanidine NPs (curve 2).

[Electrode area: 7.064 m2]

132

 

 

1.5 I I I I ‘l
-1 0 1 2 3

Log Frequency (Hz)

.b

Figure 6.10: Impedance (Log Z) data were plotted as a function of Log ﬁ'equency for a
sBLM on GCE (squares), and sBLM after interaction with PEG guanidine NPs
(triangles). An equivalent circuit shown in Figure 2.9 was used to ﬁt the impedance data.
The disruption of sBLM by PEG guanidine NPs increased the membrane capacitance and

decreased the membrane resistance. [Electrode area: 7.064 mm ]

133

 

 

7. INTEGRATION OF BIOMIMIETIC INTERFACES WITH
GOLD ELECTRODE ARRAYS

7 .1. Abstract

Recent advances in the recombinant DNA technology have led to the expression
of novel soluble and membrane proteins in bacterial and mammalian cells. These proteins
can be incorporated into biomimetic interfaces that allow the proteins’ activities to be
expressed and rapidly characterized, thereby advancing functional proteomics research
To facilitate these applications, there is a need for development of tools that will allow a
rapid, cost-effective, multi-parameter, simultaneous hi gh-throughput characterization of
multiple classes of proteins. In this study, development of a microelectrode array system
for the characterization of membrane proteins is reported. Gold microlectrode arrays were
fabricated on glass or silicon substrate using photolithography. The arrays were covered
with insulating layer of silicon nitride to deﬁne the electrode size. Tethered bilayer lipid
membrane (tBLM) interfaces were deposited on microelectrode arrays. Electrochemical
techniques were used to characterize tBLM properties using a commercial potentiostat as
well as novel custom-made circuits developed speciﬁcally for array microsystem. Highly
insulating tBLMs with near-gigaohm membrane resistances were obtained on gold
microelectrodes. Incorporation of synthetic ionophores and a channel peptide decreased
membrane resistance signiﬁcantly, thereby conﬁrming the suitability for carrying out ion
channel assays. The dependence of tBLM ﬂuidity on temperature was also investigated.
The membrane resistance value of a l,2-dipalmitoyl-sn-glycero-3-phosphothioethanol
(DPPTE)-l,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) tBLM stabilized above a

40°C corresponding to a phase transition temperature (PTT) of the lipid. Below PIT,

134

incorporation of gramicidin peptides did not mediate ion transport. Above PT‘T,
gramicidin mediated ion transport decreased the membrane resistance indicating the ﬂuid
nature of a tBLM. In addition to tBLMs, a biosensor interface for measuring the esterase
activity of a membrane protein, NEST, was deposited on microelectrode arrays. NEST
sensor response was decreased by 26% after addition of 1 mM phenylmethylsulfonyl
ﬂuoride (PMSF) indicative of NEST inhibition. The dependence of the functional activity
of NEST on temperature was investigated. NEST activity increased with increase in
temperature in the range of 21°C to 31°C. Above 31°C, the activity decreased amost

linearly with increase in temperature.

7 .2. Introduction

With completion of Human Genome Project, the focus has now shifted towards
structural and functional characterization of new proteins. Novel soluble and membrane
proteins are being identiﬁed and expressed using recombinant DNA technology. There is
a need for development of cost-effective, hi gin-throughput tools for rapid characterization
of the newly identiﬁed proteins. Biosensor arrays, in which multiple biomimetic
interfaces are fabricated on miniaturized electrodes onto a single device, will allow
simultaneous characterization of multiple proteins or enable a single assay to be
performed on many samples in parallel. Biosensor arrays could greatly reduce screening
times and increase throughput in a drug discovery process. These biosensor arrays can be
easily adapted to array microsystem that allows multi-channel measurements. Center for
Nanostructured Biomimetic Interfaces (CNBI) at MSU has developed sensor interfaces
for several soluble proteins such as tyrosinase, glucose oxidase (Lu, J. et al., 2007) and

dehydrogenases (Hassler, B. L. and Worden, R. M., 2006; Kohli, N. et al., 2006a;

135

Hassler, B. L. et al., 2007a; Hassler, B. L. et al., 2007b; Kohli, N. et al., 2007; Lu, J. et
al., 2007; Hassler, B. L. et al., 2008). The progress made with biosensors for soluble
proteins has not been matched with the progress in designing biosensors for membrane
proteins, because of the difﬁculties associated with expression, puriﬁcation and
reconstitution of these proteins. Moreover, membrane proteins need to be embedded in
BLM for expressing their activity. In this study, gold electrode arrays were designed and
fabricated using photolithography. A microelectrode array platform was developed by Dr.
Mason’s group at the Department of Electrical and Computer Engineering at MSU. The
array platform was used to deposit different protein interfaces. High impedance tBLMs
were deposited on microelectrodes ranging in size from 3 mm in diameter down to 50 um
in diameter. Gramicidin peptide was incorporated in tBLMs and the ion transport across
BLM was measured using EIS. Further, tBLM interface was used to study the effect of

temperature on bilayer ﬂuidity and gramicidin activity.
7.3. Methods

7.3.1. Microelectrode array fabrication

Microelectrode arrays were fabricated using a lift-off photolithography process.
The silicon wafers were diced to required dimension and cleaned in acetone. Then, diced
wafers were ultrasonicated in acetone and DI water for 15 min each. Finally, the wafers
were thoroughly rinsed with deionized (DI) water and dried in a nitrogen stream. A
negative photoresist AZ 5214B (Clariant, Sommerville, NJ) was spin coated (3000 rpm,
35 sec) on silicon wafers and pre-baked at 95°C for l min. This step deposits on wafers
approximately 1.5 pm thick photoresist ﬁlm. To pattern microelectrodes, the wafer

covered with photoresist ﬁlm was exposed to UV light through an optical mask. The

136

 

exposed area of the photoresist gets cross linked during a post-baking at 120 °C (for 2
min). The cross linked photoresist loses its photosensitivity. Then, the whole wafer was
exposed to UV light (no mask) for l min to degrade the previously unexposed
photoresist. The degraded photoresist pattern was dissolved and removed by treating the
wafer with AZ MIF 300 developer solution (Microchem Corporation, Newton, MA).

Finally, the wafers were thoroughly rinsed with DI water.

Physical vapor deposition (Edwards Auto3 06 thermal evaporator, Cerarnisis Ltd.,
London) technique was used to deposit titanium and gold layers on the photoresist
patterned silicon wafers. First, a 5 nm thick titanium ﬁhn was deposited at a rate of 0.2 A
per sec as an adhesion layer prior to gold deposition. Then, a gold ﬁlm of 100 nm in
thickness was deposited at a rate of 0.5 A per sec. After gold deposition, the electrodes
were cleaned and ultrasonicated in acetone for 20 min. The electrodes were rinsed using

DI water and dried in nitrogen stream.
7.3.2. Deposition of insulation layer over microelectrode arrays

A 200 nm thick insulation layer of silicon nitride was deposited on microelectrode
arrays using plasma enhanced chemical vapor deposition (Plasma 80plus, Oxford
Instruments, UK). The electrode sizes were deﬁned by using photolithography. Brieﬂy,
1.5 pm thick layer of a positive photoresist S1813 was spin coated on silicon nitride
covered microelectrode array, followed by heating at 95 °C for 8-10 min for pre-baking
of the photoresist. The area of required electrode sizes on an array was exposed to UV
light for 10 sec through a patterned optical mask. The photoresist in the exposed areas
gets degraded with UV light which can be dissolved and subsequently removed after

treatment with 352 development solution. The unexposed area remains unaffected.

137

 

 

Finally, microelectrode array was rinsed with DI water and dried. The S1813 photoresist
covered area (unexposed area in the previous UV light treatment) protects the underlying
silicon nitride ﬁlm during the subsequent reactive ion etching process. The exposed
silicon nitride area was removed using dry reactive ion etching (Plasma Quest, UK)
process. After etching, the S1813 photoresist ﬁlm was dissolved and removed by
ultrasonicating the electrodes in acetone. The ﬁnal rinsing was done using DI water. The
electrodes were cleaned in piranha solution, washed with DI water, dried in nitrogen

stream, and stored at room temperature until used.
7.3.3. Fabrication of tBLM on microelectrode arrays

DPPTE-DOPC tBLM was fabricated on microelectrode arrays using a same
procedure described in Section 2.3.5. Instead of an electrochemical cell, the
measurements were done in a small reservoir (~500 uL) created using a

polydirnethoxysiloxane (PDMS) block.
7.3.4. Fabrication of NEST sensor

NEST was expressed in E. coli and puriﬁed using reported procedures (Atkins, J.
and Glynn, P., 2000; Kohli, N. et al., 2007). NEST sensor was deposited on
macroelectrodes or microelectrode array using the procedure described by Kohli et al.
(Kohli, N. et al., 2007). Brieﬂy, the electrodes were cleaned in piranha solution and
dipped in 5 mM solution of thioctic acid in ethanol for 1 hr. The electrodes were washed
with ethanol, dried under nitrogen and dipped in poly-L-lysine (PLL) solution for 30 min.
PLL solution was prepared by adding 12 mg of PLL in 50 mL of 20 mM phosphate
buffer (pH 8.5). The electrodes were then rinsed with water and dipped in a solution of

tyrosinase (0.2 mg/ml) for l h. The last two steps were repeated 3.5 times to create 3.5

138

 

 

 

PLL-tyrosinase bilayers with PLL being the topmost layer. The electrodes were washed
with water and dipped in a solution of NEST protein (0.1 mg/mL) in 0.1 M phosphate
buffer, pH (7.0) for l h. The electrodes were then washed with water, dried under
nitrogen and dipped in phosphate buffer (0.1 M, pH 7.0) for testing. Phenyl valerate was
used as the substrate for NEST. To prepare phenyl valerate substrate solution, 15 mg of
phenyl valerate was dissolved in 1 mL of dirnethylforrnarnide (DMF), and 15 mL of

water containing 0.03% Triton X—100 was added slowly under vigorous stirring.

7.3.5. NEST solution assay

To 100 uL of 5.263 mM phenyl valerate solution, 50 uL of enzyme solution (with
proper dilution) was added and incubated at the target temperature for 20 min. Then, 100
uL of reaction stopping solution (9.5 mg/mL of sodium dodecyl sulfate and 1.23 mM 4-
aminoantipyrine) was added and incubated at room temperature for 5 min. Finally, 50 [LL
of 12.1 mM potassium ferricyanide solution was added to the the mixture and incubated

at room temperature for 5 min. The absorbance of the solution was measured at 486 nm.

7.4. Results and discussion

7.4.1. tBLM on microelectrode arrays on glass

The ﬁrst generation microelectrode arrays were fabricated on glass. SU8
photoresist was used as insulation layer to deﬁne the area of ten gold electrode elements
with electrode size ranging from 3 mm diameter to 250 pm in diameter (Design Glass-I)
[Figure 7.1A]. Tethered bilayer lipid membranes (tBLM) were deposited on these sensing
elements as described previously (Raguse, B. et al., 1998; Atanasov, V. et al., 2005;

Jadhav, S. R. et al., 2008b). The tBLM acts as an insulating dielectric barrier for ion

139

 

 

It

 

 

transport, thereby increasing the impedance of the bio-interface. Gramicidin peptides
selectively transported alkali metal ions through the tBLM and therefore lowered the
impedance of the modiﬁed tBLM in the presence of alkali metal ions in the electrolyte
solution. Dr. Mason’s lab has developed a compact, low power, impedance extraction and
digitization circuit (IDC) that can be installed under each element of a sensor array
(Yang, C. et al., 2009). The properties of IDC are listed in Table 7.1 IDC circuit was used
to locally extract and digitize sensor impedance information. IDC can measure current
response ranging from 78 fA to 100 nA, suitable for tBLM measurements (Yang, C. et
al., 2009). The experimental set-up for E18 measurements of tBLM using IDC is shown
in Figure 7.2. A data acquisition card (Agilent E3630A) was used to generate the
sinusoidal stimulus and collect output data from the IDC. Impedance measurements were
conducted over the frequency range of 10 mHz to 100 Hz using IDC after formation of
DPPTE monolayer, then tBLM, and ﬁnally after the incorporation of gramicidin peptides
in tBLM bathed in 100 mM sodium (alkali metal) chloride solution. A Randles equivalent

circuit shown in Figure 2.5 was used to model the impedance data obtained using IDC,

with the membrane capacitance (CM) and membrane resistance (RM) representing the
properties of the lipid bilayer membrane. For a DPPTE monolayer, the CM and RM
values were measured to be 0.635 uF/cm2 and 20 kﬂcm2 respectively. For the tBLM, a

CM and RM of 0.505 uF/cmz and 425 chm2 were obtained [Figure 7.3]. The

capacitance value for tBLM was in agreement with the values obtained for quality tBLMs
in our studies with commericial gold. A slightly lower value of membrane resistance

suggested that there might be imperfections in tBLM at the edges of the electrode.

140

 

Incorporation of gramicidin peptides in tBLM decreased the membrane resistance from

425 to 59 kﬂcm2 due to the passage of sodium (alkali metal) ions through the tBLM.

These results demonstrate that the compact IDC circuit can successfully characterize the
impedance spectra of sensor materials. A chip containing an array of IDC units coupled

to on-top gold electrodes would allow simultaneous measurement of all sensor elements.
7.4.2. tBLM on microelectrode arrays on silica substrates

Two designs of second-generation microelectrode arrays were deposited on
silicon substrates to provide ﬂexibility in choosing the desired sizes and patterns of

electrodes. Design Si-I had ten gold electrode elements (deposited in an area of 3 X 3

m2) on a 15 X 30 mm2 size silicon substrate [Figure 7 .18], while Design Si-II had forty

gold electrodes (deposited in a 3 X 3 mm2 area) on a 12.5 X 12.5 mm2 size silicon

substrate. The size of gold electrodes on an array was varied and controlled by patterning
silicon nitride insulation layer using photolithography. In one of the Design Si-I electrode
arrays (Design Si-IA), ﬁve electrode sizes (800 pm, 400 um, 200 um, 100 um and 50 um
in diameter) were patterned with two electrodes of each size. While in Design Si-IB,
microelectrode array having ﬁve electrodes each of sizes 250 pm and 500 pm diameter
were fabricated. The maximum diameter of a gold electrode that can be accommodated
on Design Si-I electrodes was 1 mm. In case of Design Si-II electrodes [Figure 7.1C],
four electrode sizes (200 um, 150 um, 100 um, and 50 um in diameter) were patterned
with ten electrodes of each size. The maximum diameter of a gold electrode that can be

accommodated on Design Si-II electrodes was 250 pm.

141

 

tBLM was deposited on gold microelectrode arrays in a two step process
described in Section 2.3.5. After BLM deposition, the electrochemical properties of
tBLM were measured using electrochemical impedance spectroscopy (EIS). Figure 7.4A
and Figure 7.4B, show the impedance spectrum obtained for tBLM deposited on 250 um
and 500 um diameter microelectrodes, respectively. For a tBLM deposited on a 500 um

diameter electrodes, 3 highly insulting tBLM was obtained with a membrane resistance of
1.29 Mﬂcmz. The suitability of tBLM for ion channel assays was conﬁrmed by
measuring ion transport activity of gramicidin peptides. Incorporation of gramicidin in
tBLM decreased the RM value to 0.029 MQcmZ, conﬁrming ion transport across tBLM

[Figure 7.4A]. Similarly for a 250 um electrode, the tBLM resistance was decreased by

almost three order of magnitude (from 0.9 MQcm2 to 0.0015 MQcmz) following

gramicidin incorporation [Figure 7.5B]. Electrochemical properties of the tBLM

deposited on different size microelectrodes are summarized in Table 7.2 (absolute values

of the constants) and in Table 7.3 (normalized to 1 cm2 electrode area). As the electrode

size got smaller, the absolute value of membrane resistances increased, while the
membrane capacitance values decreased. Compared to membrane capacitance value of
1.2 uF obtained using a gold macroelectrode 5 mm in diameter, the microelectrodes had
a membrane capacitance in nanofarads (nf). Similarly, microelectrodes had higher
absolute membrane resistance values compared to macrolectrodes, while electrodes
below 500 pm in diameter had near-gigaohm resistances. During patch clamp
experiments conducted using liposome vesicles, obtaining a gigaohm resistance is a pre-

requisite for ion channel assays. Gigaohrn resistance membranes are characteristic of

142

 

 

 

good quality BLMs; stray currents do not contribute signiﬁcantly at such high resistance
thereby permitting the measurement of small ion channel currents. Obtaining near-
gigaohm resistances with tBLMs on microelectrodes offers the possibility of using these
membranes for sensitive single channel activity measurements. The sealing property of
tBLMs on microelectrodes was conﬁrmed by measuring the ion conduction activity of
gramicidin peptides. The membrane resistance of tBLMs on all microelectrodes (ﬁom
800 um down to 50 um diameter electrodes) decreased signiﬁcantly following gramicidin
incorporation [Figure 7.5].

The area-normalized electrochemical properties of tBLMs deposited on
microlectrodes deviate somewhat ﬁom those obtained using macroelectrodes [Table 7.3].

For electrodes smaller than 500 um diameter electrode, the area normalized membrane

resistance was below 1 MQcmz, and it decreased as the electrode size got smaller.

Similarly, the area normalized membrane capacitance values increased as the electrode
size decreased. These trends are likely due to electrode-edge effects, arising from
incompletely formed tBLM at the electrode edge. The increase in ratio of circumference
to area as electrodes get smaller is consistent with this hypothesis. In addition to the edge
effects, surface characteristics of electrode, such as surface roughness, also play a major
role. Higher roughness might cause formation of unorganized monolayer of tethering
lipids having defects at the areas with higher roughness. This, in turn, leads to formation
of non-insulating tBLMs that are unsuitable for ion channel assays. A higher surface
roughness can also be imparted during fabrication process of electrodes. Deposition of
microelectrode arrays using photolithography involves multiple steps of photoresist

coating and removal. Incomplete removal of photoresist leaves residues on the electrode

143

 

h

 

 

surface, which can lead to defects in tBLM, and hence non-ideal charge transfer across

the bilayer.
7.4.3. NEST sensor on microelectrode arrays

In addition to tBLMs, a biosensor interface was fabricated on gold macro- and
microelectrodes. The biosensor measures the esterase activity of NEST, a catalytically
active fragment of a membrane protein Neuropathy Target Esterase (NTE) (Atkins, J. and
Glynn, P., 2000; Forshaw, P. J. et al., 2001; Wang, J. and Richardson, R. J ., 2003; Kropp,
T. J. et al., 2004). NTE is an important membrane protein in nerve cells. Inhibition of
NTE by organophosphorus compounds leads to neurological diseases, such as
organophosphate-induced delayed neuropathy and motor neuron disease (Glynn, P.,
2006b; Glynn, P., 2006a). In collaboration with Dr. Rudy J. Richardson at the University
of Michigan, the CNBI has developed a biomimetic interface for NEST that can detect
presence of organophosphorus compounds (Kohli, N. et al., 2007). The biosensor
contains two enzymes, NEST and tyrosinase, immobilized using the positively charged
polyelectrolyte PLL. NEST’s esterase activity hydrolyzes the substrate (phenyl valerate)
to phenol, which is subsequently oxidized to catechol and then to o-quinone by
tyrosinase. o-quinone diffuses through enzymatic layers and gets rapidly reduced to
catechol at the electrode surface. In constant potential arnperometry (CPA) experiments,
the electrode is held at negative potential (-100 mV), which results in o-quinone being
reduced to catechol at the electrode. The cyclic interconversion of catechol and o-quinone

results in signal ampliﬁcation by the interface.

One disadvantage of this design is that conversion of catechol to o-quinone can

cause an irreversible suicide inactivation of tyrosinase (Chang, T. S., 2009; Rarnsden, C.,

144

I

 

 

2009). Higher accumulation of catechol might result in deactivation of tyrosinase over
time. To minimize this effect, we developed an approach that uses cyclic voltammetry
(CV) for measuring sensor response. In CV experiments, the potential is applied for a
brief period of time, thereby limiting the amount of catechol regenerated, compared to
constant potential arnperometry. Secondly, part of catechol liberated due to o-quinone
oxidation during a forward scan is converted back to quinone during a reverse scan.
These two factors may signiﬁcantly improve biosensor lifetime. To test this hypothesis, a
concentration-current relationship was obtained using CPA and CV for freshly tyrosinase
modiﬁed electrode (3 layers of PLL and tyrosinase). After the experiment, the electrode

was rinsed and used again to generate the concentration-current curve. The concentration-

current plot obtained for the ﬁrst use (R2 = 0.999) and second use (R2 = 0.999) of same
catechol sensor aﬁer rinsing using CPA is shown in Figure 7 .6A. The sensitivity (uAuM-

1cm-z) of tyrosinase sensor was found to decrease by 34% after the ﬁrst use, possibly due

to suicide inactivation of enzyme activity. The same experiment carried out using CV
showed a 7% decrease in the sensitivity, suggesting that the extent of enzyme
deactivation almost half in CV which could be due to less regeneration of catechol
[Figure 7.6B]. In addition to CV, we explored the possibility of using EIS to monitor the

catechol sensor response. EIS study showed that increasing the catechol concentration

decreased the charge transfer resistance (RCT) values. RCT is related to concentration of

reduced species as follows (Bard, A. and Faulkner, L., 2001)

145

 

 

RT
”F10 ............................................................ Eq. (7.1)

 

RCT =
. 0 *(l—a) *a
10 — ”FA/C CO CR ................................ Eq.(7.2)

where, i0 is an exchange current (Amperes), k0 is the electron transfer coefﬁcient, C0*
is the bulk concentration of oxidized species, and C 12* is the bulk concentration of

reduced species, and A is area of electrode. In Figure 7.7, RC7 was plotted as a function

of Co'l/z. It follows a linear (R2 = 0.996) relationship, indicating that RC1“ can be used

as a measure of sensor response.
The concentration-current response obtained for NEST activity on a

macroelectrode (~1 cm2 area) using CPA and CV are shown in Figure 7.8A and Figure
7.8B respectively. Using CV, the sensitivity of a NEST sensor was 2.08 i 0.1 uAuM-
1cm'z, which was 8.7 times higher than the sensitivity obtained using CPA (240 :l: 10

nAuM-lcm'z). The sensor response increased linearly (R2 = 0.998) with phenyl valerate

concentration in a range of 0 — 20 uM, reaching saturation at around 30 uM. NEST

sensor was deposited on gold microelectrode arrays on silicon wafer using the same

method. On a 500 mm diameter electrodes, 3 current sensitivity of 1.02 i 0.05 uAuM-
1cm.2 was obtained using CV. EIS was used to monitor the change in RCT after addition

of phenyl valerate. RCT of NEST sensor on 500 mm diameter electrode decreased from

146

 

 

4.58 chm2 to 1.58 chm2 after addition of 10 uM phenyl valerate, suggesting that RCT

can also be used as a measure of NEST activity. An equivalent circuit shown in Figure
7.9 was used to model the impedance data. The NEST sensor on the array was then used
to measure an organophosphate-induced inhibition of NEST esterase activity. Addition of
1 mM of the non-neuropathic NEST inhibitor phenylmethylsulfonyl ﬂuoride (PMSF)
decreased the current response by 28% [Figure 7.10A]. PMSF inhibits NEST esterase
activity, thereby lowering the phenol and subsequent o-quinone production, which leads
to decrease in sensor response. EIS results showed that inhibition of NEST activity by

PMSF leads to increase in the charge transfer resistance [Figure 7.10B], once again

indicating that RCT can be used as measure of NEST inhibition [Figure 7.10B].

7.4.4. Effect of temperature on biomimetic interfaces

7 .4.4.1. Temperature dependence of tBLM impedance

The temperature dependence of a tBLM’s electrochemical properties was
investigated. The tBLM was fabricated on gold macroelectrodes by depositing a lower
leaﬂet of DPPTE tethering lipids, and then the upper leaﬂet was deposited using
liposome vesicles made from 1,2—dipalmitoyl-sn-glycero-3-phosphocholine (DPPC)
phospholipids. Both leaﬂets of the resulting DPPTE-DPPC tBLM had lipids with
saturated hydrophobic chains. The temperature was varied from 22°C to 60°C, with

tBLM incubated at each temperature until a steady state response was achieved. Figure

7.11A shows the RM values for tBLM plotted as a function of temperature. The

membrane resistance decreased to 40°C, after which it remained essentially constant up to

60°C. These results indicate that after 40°C, there was no change in insulating property of

147

 

 

 

 

BLM. Phospholipids with palmitoyl hydrocarbon chains have a phase transition
temperature (PTT), at which the bilayer undergoes a transition from a gel state to a ﬂuid,
liquid-crystalline phase, of 40°C. These results imply that, in the gel phase, as the
temperature was increased from 22°C to PTT, the membrane resistance decreased. Then,

above the FTP, the BLM’s resistance is essentially independent of temperature.

The ﬂuid nature of saturated BLM was further conﬁrmed using gramicidin
channels. For ion conduction, free diffusion of the monomers in ﬂuid BLM is necessary.
It was observed that addition of gramicidin to symmetric tBLM at 22°C did not decrease
membrane resistance, indicating that BLM was in a gel state. However, addition of

gramicidin to symmetric tBLM at 50°C, which is above PTT of palmitoyl chains,
decreased the membrane resistance from 116 kﬂcm2 to 77 chm2, conﬁrming the ﬂuid
nature of tBLM [Figure 7.12].

7.4.4.2. Temperature dependence of NEST activity

The effect of temperature on the NEST activity was studied using a solution assay
and using electrochemical NEST sensor in a temperature range of 22°C to 60°C. Figure
7.13A shows the temperature dependence of NEST activity in a solution assay. NEST
activity increased from 21°C to 31°C, and above 31°C, NEST activity decreased,

suggesting inhibition of NEST at higher temperatures. Fitting the Arrhenius equation to

the NEST activity data between 31°C to 60°C gave activation energy of -l 9.3 kJmol-IK-l

[Figure 7.13B]. Similar results were obtained using the electrochemical NEST sensor.
NEST activity was found to increase with temperature in a range from 21°C to 31°C.

Above 31°C, the activity gradually decreased [Figure 7.14A]. At 50°C, NEST showed

148

 

 

 

 

only 55% of the activity obtained at 21°C. Fitting the Arrhenius equation to the decrease

in NEST activity provided an activation energy of ~14.8 kJmol'lK-1 [Figure 7.14B],

which is similar to the value obtained in solution.. Cooling down the temperature back to
20°C did not recover the NEST activity, suggesting that higher temperatures may cause

irreversible denaturation of NEST.
7.5. Conclusions

This study described the integration of tBLM and NEST biosensor interfaces with
gold microelectrode arrays on glass and silicon substrates. The results indicate that highly
insulating tBLMs can be deposited on microelectrodes ranging in size from 3 mm in
diameter down to 50 pm in diameter. High impedance tBLMs with near-gigaohm
resistances were obtained and were found to be suitable for characterization of ion
conduction activity of channel peptides. The microelectrode array were also suitable for
studying other membrane proteins such as NEST. NEST sensor response decreased in the
presence of PMSF, indicative of NEST inhibition. NEST was found to undergo a
temperature dependent deactivation above 31°C. The resulting biosensor arrays have
potential applications in functional proteomics, biosensor development, and hi gh-

throughput drug screening.

149

 

 

 

 

 

 

 

 

 

 

 

Technology 0.5 pm CMOS
Chip area 0.06 mm2

Power supply 3 Volts

Power 6 uW

Amplitude conversion linearity 50 dB
Phase conversion RMS error 2.7%

Maximum current range 100 nA
Sensitivity 78 fA

 

 

 

 

Table 7 .1: The properties of impedance extraction and digitization circuit (IDC) that can
be installed under each element of a sensor array. CMOS: complementary metal—oxide—
semiconductor.

150

 

 

 

 

 

 

 

 

 

RM of tBLM
Electrode CM of tBLM RM of tBLM with
diameter (um) (nF) (M52) gramicidin

(MS!)
50 2 2390 754
100 l 490 334
200 2 391 166
250 2 1 860 3.10
400 2 398 96.9
500 3 654 14.9
800 6 471 36.3

 

 

 

 

 

 

Table 7.2: The absolute values of electrochemical properties for tBLMs deposited on
Design Si-I microelectrodes on silicon.

151

 

 

 

 

 

 

 

 

 

 

 

RM of tBLM
EICCtl'Ode C M of tBLM RM of tBLM With
manger (chm‘2> (Mnemz) 3mm"?
(Mﬂcm )
50 138 0.047 0.015
100 24.2 0.039 0.026
200 6.38 0.12 0.052
250 3.66 0.91 0.002
400 1.94 0.50 0.122
500 1.55 1.29 0.029
800 1.26 2.36 0.182

 

 

 

 

 

Table 7.3: The area normalized values of electrochemical properties for tBLMs
deposited on Design Si-I microelectrodes on silicon.

152

 

 

 

 

Figure 7.1: (A) First generation Design Glass-I microelectrode array deposited on glass
substrate. SU8 photoresist was used as insulation layer to deﬁne the area of ten gold
electrode elements. (B) Second generation Design Si-I microelectrode array deposited on
silicon substrate. Silicon nitride was used as insulation layer to deﬁne the electrode size.
(C) Design Si-II microelectrode array deposited on silicon substrate. Silicon nitride was
used as insulation layer to deﬁne the electrode size.

153

 

 

Serial data path

 

digital control

W

sinus id
v stimu us

  
         
   
 
  
 

 

 

III IIIIII IIIII III.

 

 

 

 

 

...-IIIIIUIIIIIII...

 

 

lock-in IDC
chip

 

 

tBLM biosensor array on
the pattened gold electrodes
(on glass chip)

 

 

Figure 7.2: The experimental set-up for EIS measurements of tBLM using IDC (Yang,
C. et al., 2009).

 

154

 

 

4.5 I I I I
—2 -1 0 1 2
Log Frequency (Hz)

 

w
8
J

 

- Phase angle (°)
8 8 8 3 8

O.)
O
r

 

N
O

 

I T I I

-1 0 1 2
Log Frequency (Hz)

I
N

Figure 7.3: (A) Impedance (Log Z) data and (B) phase angle data were plotted as a
function of Log frequency for tethering lipid monolayer (triangles) deposited on 3 mm
diameter electrodes on Design Glass-I microelectrode array, tBLM formed using DOPC
liposomes (squares), and for tBLM after gramicidin incorporation (diamonds) in 100 mM
NaCl. Gramicidin was added to give a ﬁnal concentration of 1 uM protein (monomer) in
an electrolyte solution.

155

 

 

'F

 

 

 

3.5 r I I I T 1
-2 -1 0 1 2 3

Log Frequency (Hz)

 

.h

 

4 I I I I I I
-2 -1 0 1 2

Log Frequency (Hz)

 

00
h

Figure 7.4: (A) Impedance (Log Z) data were plotted as a function of Log frequency for
tBLM (squares) deposited on 500 um diameter electrodes on Design Si-I microelectrode
array, and for tBLM after gramicidin incorporation (triangles) in 100 mM NaCl. (B)
Impedance (Log Z) data were plotted as a function of Log Frequency for tBLM (squares)
deposited on 250 um diameter electrodes on Design Si-I microelectrode array, and for
tBLM after gramicidin incorporation (triangles) in 100 mM NaCl. Gramicidin was added
to give a ﬁnal concentration of 1 M protein (monomer) in an electrolyte solution. An
equivalent circuit shown in Figure 2.5 was ﬁtted to the impedance data. The solid lines
passing through the data points represent the model ﬁt.

156

 

-‘ N
-‘ 01 N 01
1 1 1 r

Membrane resistance, RM
2
(Mﬂcm )
.0
U'l

 

 

50 100 200 250 400 500 800
Electrode diameter (um)

Figure 7.5: tBLM’s area normalized membrane resistance before (bar with no vertical
stripes) and after gramicidin incorporation (bar with vertical stripes) for tBLMs deposited
on various size microelectrodes. The sealing nature of tBLMs was conﬁrmed by the
measurement of gramicidin-mediated ion conduction across tBLM.

157

 

 

1 f

2
on

a)
L

N
1

 

Current density (uAcm
.h

III
‘
-1

 

I I I I

30 40 50 60 70
Catechol conc (uM)

O
.1.
C
N
C

B 160—

1

u—L

h

o
1

A

N

C
1

100 i

 

Current density (pAcm

N «b O) m
C O o o C
1 1 1 1

 

I I I I I I I

50 75 100 125 150 175
Catechol conc. (pM)

O
N
0'!

Figure 7.6: Current density of catechol sensor as a function of catechol concentration
(conc.) obtained using (A) constant potential arnperometry under stirring, and (B) using
cyclic voltammetry.

158

I

 

 

12000 -

10000 -

8000 ~

6000 -

Rcr (Q)

4000 -

2000 ‘

 

 

o I I l I l I I
0 50 100 150 200 250 300 350

cod/2 ((mOI cm-3)-1I2)

Figure 7 .7: Charge transfer resistance (RCT) obtained for catechol sensor as a function
of catechol concentration.

159

 

 

O)

-2
)
or

A

l

00

1

1

N
1
O

—b
1
O

 

 

Current density (uAcm

O

I H I I I

10 15 2o 25 30
Phenyl valerate conc. (uM)

O
01

W
\I
o

1

UT 03
O o
1 1

.5
o
1

.

N
O
1

Current density (pAcm )
-x on
O O

 

O

 

I I I I I I I

10 15 20 25 30 35 40
Phenyl valerate conc. (uM)

O
01

Figure 7.8: Current density of NEST sensor deposited on macroelectrode (5 mm in
diameter) as a function of phenyl valerate concentration (conc.) obtained using (A)
constant potential arnperometry under stirring, and (B) using cyclic voltammetry.

160

 

 

 

 

 

 

 

Figure 7.9: An equivalent circuit used to model impedance data ﬁom NEST sensor
experiments. R3 represents the resistance of solution, CDL is the double layer
capacitance, RM represents membrane resistance, and Zw is the Warburg impedance.

161

 

 

 

 

 

 

 

 

 

 

 

A 0.03 -
0.02 —
2 0.02 —
3
E 0.01 -
5 0.01 ~
0
0.00 “Ir-
F—f
'0.01 I I I l I I
400 300 200 100 0 -100 -200
Potential (mV)
B 7.5 —
g 6.5 -
N
or
o
-‘ 5.5 .
4.5 I I I I I I
-1 -O.5 0 0.5 1 1.5 2
Log Frequency (Hz)

Figure 7.10: (A) Cyclic voltarnmograrn obtained for NEST sensor on 500 um
microelectrode after addition on 10 11M phenyl valerate (curve 1), and after addition of 1
mM PMSF (curve 2). (B) Impedance data (Log Z) were plotted as a function of Log
frequency for NEST sensor on 500 um microelectrode after addition on 10 M phenyl
valerate (squares), and after addition of 1 mM PMSF (triangles). An equivalent circuit
shown in Figure 7.9 was used to ﬁt impedance data. The solid lines passing through the
datapoints represents the ﬁt.

162

 

 

 

 

 

 

 

 

 

 

 

A 6 '
O
5.8 -
NA
5 5 6
Q, .
a?
5.4 2
a O
o
.1
5.2 - e
O
O
O . .
5 I T I I l
20 30 40 50 60 70
Temperature (°C)
B -5 -
-5.2 ~
”A
E -54 —
$9..
>.
3,-5.6 -
..1 y = -4450.cx + 9.2398
'5'8 “ R2 = 0.9438 .
"6 I I I I I
0.0029 0.003 0.0031 0.0032 0.0033 0.0034
1IT (1IK)

Figure 7.11: (A) Membrane resistance of DPPTE-DPPC tBLM was plotted as a ﬁrnction
of temperature. Membrane resistance values stabilize after 40°C, which corresponds to
the phase transition temperature of DPPC. (B) Admittance data (Log Y) of DPPTE-
DPPC tBLM plotted as a function of l/temperature to obtain activation energy.

163

 

 

4 I I I I
-2 -1 0 1 2

Log Frequency (Hz)

 

Figure 7.12: Impedance data (Log Z) was plotted as a function of Log frequency for
DPPTE-DPPC tBLM at 50°C (squares), and after addition of 1 uM gramicidin at 50°C

(triangles). [Electrode area is 0.2 cm ]

164

 

 

>

NEST speciﬁc activity
(Units/mg)
a
t
O

 

o I I I I I I

 

1 0 20 30 40 50 60 70
Temperature (°C)

 

 

 

B 1.6 -

g 1.4 .

.2

out

o

«I 1.2 5

.3 a

'5 s 1-

3.2

‘0 c

E a 0.8 ‘7 .

E.

a 0.6 I I I I I
3 0.0029 0.003 0.0031 0.0032 0.0033 0.0034

1rr (1IK)

Figure 7.13: (A) NEST speciﬁc activity obtained using solution assay as a function of

temperature. (B) Log(NEST speciﬁc activity) was plotted as a function of l/temperature
to obtain activation energy using solution assay of NEST.

165

 

 

 

 

 

 

 

 

 

 

 

 

A
NA 12 " O
' e
510-
g c
v 8 _
b e
8 6-
e e
B 4 .
5
'5 2* .
o 0 I I I I I I
10 20 30 40 50 60 70
Temperature (°C)
B 1 2 -
1.1 J
’5 e
'5 1 ‘
C
3 .A 0.9 -
E {5, 0.8—
!; <
a 3; 0.7 .
9.
31 0,6 - y =1787.7x- 4.7719
-’ 0 5 , R2 = 0.9769
0.4 I l ‘I *I I
0.003 0.0031 0.0032 0.0033 0.0034 0.0035
1” [11K]

Figure 7.14: (A) Current density obtained using electrochemical sensor as a function of
temperature. (B) Log(Current density) was plotted as a function of l/temperature to
obtain activation energy using electrochemical sensor of NEST.

166

 

 

 

8. REFERENCES

Andersson, M., Okeyo, G., Wilson, D., Keizer, H., Moe, P., Blount, P., Fine, D.,
Dodabalapur, A., and Duran, RS. 2008. Voltage-induced gating of the
mechanosensitive MscL ion channel reconstituted in a tethered lipid bilayer
membrane. Biosensors & Bioelectronics 23(6):919-923.

Atanasov, V., Atanasova, P.P., Vockenroth, I.K., Knorr, N., and Koper, I. 2006. A
molecular toolkit for highly insulating tethered bilayer lipid membranes on
various substrates. Bioconjugate Chemistry 17(3):631-637.

Atanasov, V., Knorr, N., Duran, R.S., Ingebrandt, S., Offenhausser, A., Knoll, W., and
Koper, I. 2005. Membrane on a chip: A functional tethered lipid bilayer
membrane on silicon oxide surfaces. Biophysical Journal 89(3):]780-1788.

Atkins, J ., and Glynn, P. 2000. Membrane association of and critical residues in the
catalytic domain of human neuropathy target esterase. Journal of Biological
Chemistry 275(32):24477-24483.

Bainbridge, G., Gokce, 1., and Lakey, J .H. 1998. Voltage gating is a fundamental feature
of porin and toxin beta-barrel membrane channels. F ebs Letters 431(3):305-308.

Baker, G.L., Vogel, EB, and Smith, MR. 2008. Glass transitions in polylactides.
Polymer Reviews 48(1):64-84.

Bard, A., and Faulkner, L. 2001. Electrochemical methods: Fundamentals and
Applications. John Wiley and Sons, New York.

Becucci, L., and Guidelli, R. 2007. Kinetics of channel formation in bilayer lipid
membranes (BLMs) and tethered BLMs: Monazomycin and melittin. Langmuir
23(10):5601-5608.

Becucci, L., Guidelli, R., Peggion, C., Toniolo, C., and Moncelli, M.R. 2005a.
Incorporation of channel-forming peptides in a H g-supported lipid bilayer.
Journal of Electroanalytical Chemistry 576(1):121-128.

Becucci, L., Innocenti, M., Salvietti, E., Rindi, A., Pasquini, I., Vassalli, M., Foresti,
M.L., and Guidelli, R. 2008. Potassium ion transport by gramicidin and
valinomycin across a Ag(111)«supported tethered bilayer lipid membrane.
Electrochimica Acta 53(22):6372-6379.

Becucci, L., Moncelli, M.R., and Guidelli, R. 2006. Impedance spectroscopy of OmpF

porin reconstituted into a mercury-supported lipid bilayer. Langmuir 22(3):134l-
1346.

167

Becucci, L., Moncelli, M.R., Naumann, R., and Guidelli, R. 2005b. Potassium ion
transport by valinomycin across a Hg—supported lipid bilayer. Journal of the
American Chemical Society 127(38): l 33 16-13323.

Berquand, A., Mazeran, P.E., Pantigny, J., Proux-Delrouyre, V., Laval, J.M., and
Bourdillon, C. 2003. Two-step formation of streptavidin-supported lipid bilayers
by PEG-triggered vesicle fusion. Fluorescence and atomic force microscopy
characterization. Langmuir 19(5): 1700-1 707.

Blake, MS, and Gotschlich, EC. 1982. Puriﬁcation and Partial Characterization of the
Maj or Outer-Membrane Protein of Neisseria-Gonorrhoeae. Infection and
Immunity 36(1):277-283.

Blake, M.S., Gotschlich, EC. 1987. Functional and immunological properties of
pathogenic Neisseria surface proteins. Inouye M, editor. John Wlley and Sons,
Inc., New York.

Bunjes, N., Schmidt, E.K., Jonczyk, A., Rippmann, F., Beyer, D., Ringsdorf, H., Graber,
P., Knoll, W., and Naumann, R. 1997. Thiopeptide—supported lipid layers on solid
substrates. Langmuir 13(23):6188-6194.

Cevc, G., Svetina, S., and Zeks, B. 1981. Electrostatic Potential of Bilayer Lipid-
Membranes with the Structural Surface-Charge Smeared Perpendicular to the
Membrane-Solution Interface - an Extension of the Gouy-Chapman Diffuse
Double-Layer Theory. Journal of Physical Chemistry 85(12): 1762-1767.

Chang, TS. 2009. An Updated Review of Tyrosinase Inhibitors. International Journal of
Molecular Sciences 10(6):2440-2475.

Cornell, B.A., BraachMaksvytis, V.L.B., King, L.G., Osman, P.D.J., Raguse, B.,
Wieczorek, L., and Pace, RJ. 1997. A biosensor that uses ion-channel switches.
Nature 387(6633):580-583.

Crarnpton, H.L., and Simanek, BE. 2007. Dendrimers as drug delivery vehicles: non-
covalent interactions of bioactive compounds with dendrimers. Polymer
International 56(4):489-496.

Cremer, RS, and Boxer, S.G. 1999. Formation and spreading of lipid bilayers on planar
glass supports. Journal of Physical Chemistry B 103(13):2554-2559.

Diao, P., Jiang, D.L., Cui, X.L., Gu, D.P., Tong, R.T., and Zhong, B. 1999. Studies of

structural disorder of self-assembled thiol monolayers on gold by cyclic
voltammetry and ac impedance. Journal of Electroanalytical Chemistry
464(1):61-67.

Domanski, D.M., Klajnert, B., and Bryszewska, M. 2004. Inﬂuence of PAMAM
dendrimers on human red blood cells. Bioelectrochemistry 630-2): 1 89-191.

168

 

Du, L.W., Liu, X.H., Huang, W.M., and Wang, E.K. 2006. A study on the interaction
between ibuprofen and bilayer lipid membrane. Electrochimica Acta 51(26):5754-
5760.

Dufes, C., Uchegbu, IF, and Schatzlein, A.G. 2005. Dendrimers in gene delivery.
Advanced Drug Delivery Reviews 57(15):217.7-2202.

Duncan, R., and 1220, L. 2005. Dendrimer biocompatibility and toxicity. Advanced Drug
Delivery Reviews 57(15):2215-2237.

Elie-Caille, C., Fliniaux, O., Pantigny, J., Maziere, J.C., and Bourdillon, C. 2005. Self-
assembly of solid-supported membranes using a triggered fusion of phospholipid-

enriched proteoliposomes prepared from the inner mitochondrial membrane.
Langmuir 21(1 0):4661 -4668.

Esfand, R., and Tomalia, DA. 2001. Poly(amidoamine) (PAMAM) dendrimers: from
biomimicry to drug delivery and biomedical applications. Drug Discovery Today
6(8):427—436.

Forshaw, P.J., Atkins, J ., Ray, D.E., and Glynn, P. 2001. The catalytic domain of human
neuropathy target esterase mediates an organophosphate-sensitive ionic

conductance across liposome membranes. Journal of Neurochemistry 79(2):400-
406.

Fortig, A., Jordan, R., Graf, K., Schiavon, G., Purrucker, O., and Tanaka, M. 2004. Solid-
supported biomimetic membranes with tailored lipopolymer tethers.
Macromolecular Symposia 210:329-33 8.

Gao, H., Feng, J., Luo, G.A., Ottova, AL, and Tien, HT 2001. Some electrochemical
features of supported bilayer lipid membranes. Electroanalysis 13(1):49-53.

Giess, F., Friedrich, M.G., Heberle, J ., Naumann, R.L., and Knoll, W. 2004. The protein-
tethered lipid bilayer: A novel mimic of the biological membrane. Biophysical
Journal 87(5):3213-3220.

Ginzburg, V.V., and Balijepailli, S. 2007. Modeling the thermodynamics of the
interaction of nanoparticles with cell membranes. Nano Letters 7:3716-3722.

Glynn, P. 2006a. A mechanism for organophosphate-induced delayed neuropathy.
Toxicology Letters 162(1):94-97.

Glynn, P. 2006b. Neurodegeneration involving neuropathy target esterase (NTE).
Toxicology Letters 164:S9-S9.

Graneli, A., Rydstrom, J ., Kasemo, B., and Hook, F. 2003. Formation of supported lipid
bilayer membranes on SiOZ from proteoliposomes containing transmembrane
proteins. Langmuir 19(3):842-850.

169

 

Hassler, B.L., Amundsen, T.J., Zeikus, J .G., Lee, 1., and Worden, R.M. 2008. Versatile
bioelectronic interfaces on ﬂexible non-conductive substrates. Biosensors &
Bioelectronics 23(10): 1481-1487.

Hassler, B.L., Dennis, M., Laivenieks, M., Zeikus, J.G., and Worden, R.M. 2007a.
Mutation of Tyr-218 to Phe in Thermoanaerobacter ethanolicus secondary alcohol
dehydrogenase: Effects on bioelectronic interface performance. Applied
Biochemistry and Biotechnology 143:1-15.

Hassler, B.L., Kohli, N., Zeikus, J.G., Lee, 1., and Worden, R.M. 2007b. Renewable
dehydrogenase-based interfaces for bioelectronic applications. Langmuir
23(13):7127-7133.

Hassler, B.L., and Worden, R.M. 2006. Versatile bioelectronic interfaces based on
heterotrifunctional linking molecules. Biosensors & Bioelectronics 21(11):2146-
2154.

Hong, S.P., Leroueil, P.R._, Janus, E.K., Peters, J.L., Kober, M.M., Islam, M.T., Orr, B.G.,
Baker, J .R., and H011, M.M.B. 2006. Interaction of polycationic polymers with
supported lipid bilayers and cells: Nanoscale hole formation and enhanced
membrane permeability. Bioconjugate Chemistry 17(3):728-734.

Jadhav, S.R., Sui, D., Garavito, R.M., and Worden, R.M. 2008a. Fabrication of highly
insulating tethered bilayer lipid membrane using yeast cell membrane fractions

for measuring ion channel activity. Journal of Colloid and Interface Science
322(2):465-472.

Jadhav, S.R., Zheng, Y., Garavito, R.M., and Worden, R.M. 2008b. Functional
Characterization of PorB class II Porin ﬁ'om Neisseria meningitidis using a
Tethered Bilayer Lipid Membrane. Biosensors and Bioelectronics 24(4):831-835.

Janshoff, A., and Steinem, C. 2006. Transport across artiﬁcial membranes - an analytical
perspective. Analytical and Bioanalytical Chemistry 385(3):433-451.

Jeuken, L.J.C., Connell, S.D., Henderson, P.J.F., Gennis, R.B., Evans, SD, and Bushby,
R.J. 2006. Redox enzymes in tethered membranes. Journal of the American
Chemical Society 128(5):l7l 1-1716.

Keller, C.A., and Kasemo, B. 1998. Surface speciﬁc kinetics of lipid vesicle adsorption
measured with a quartz crystal microbalance. Biophysical Journal 75(3):l397-
1402.

Klemic, K.G., Klemic, J .F ., Reed, M.A., and Sigworth, F.J. 2002a. Micromolded PDMS
planar electrode allows patch clamp electrical recordings from cells. Biosensors &
Bioelectronics 17(6-7):597-604.

Klemic, K.G., Li, X.H., Klemic, J.F., Reed, M.A., and Sigworth, F.J. 2002b. Improved
planar PDMS patch electrode array. Biophysical Journal 82(1):l6lA-161A.

170

 

Knoll, W., Koper, 1., Naumann, R., and Sinner, E.K. 2008. Tethered bimolecular lipid
membranes - A novel model membrane platform. Electrochimica Acta
53(23):6680-6689.

Kohli, N., Hassler, B.L., Partlrasarathy, L., Richardson, R.J., Ofoli, R.Y., Worden, R.M.,
and Lee, I. 2006a. Tethered lipid bilayers on electrolessly deposited gold for
bioelectronic applications. Biomacromolecules 7 (12):3327-3335.

Kohli, N., Srivastava, D., Sun, J., Richardson, R.J., Lee, 1.8., and Worden, R.M. 2007.
Nanostructured biosensor for measuring neuropathy target esterase activity.
Analytical Chemistry 79(14):5196-5203.

Kohli, N., Vaidya, S., Ofoli, R.Y., Worden, R.M., and Lee, I. 2006b. Arrays of lipid
bilayers and liposomes on patterned polyelectrolyte templates. Journal of Colloid
and Interface Science 301 (2)2461-469.

Krishna, G., Schulte, J., Cornell, B.A., Pace, R., Wieczorek, L., and Osman, PD. 2001.
Tethered bilayer membranes containing ionic reservoirs: The interfacial
capacitance. Langmuir 17(16):4858-4866.

Krishna, G., Schulte, J., Cornell, B.A., Pace, R.J., and Osman, PD. 2003. Tethered
bilayer membranes containing ionic reservoirs: Selectivity and conductance.
Langmuir l9(6):2294-2305.

Kropp, T.J., Glynn, P., and Richardson, R.J. 2004. The mipafox-inhibited catalytic
domain of human neuropathy target esterase ages by reversible proton loss.
Biochemistry 43(12):3 716-3 722.

Kugler, R., and Knoll, W. 2002. Polyelectrolyte-supported lipid membranes.
Bioelectrochemistry 56(1 -2): 1 75-178.

Lang, H., Duschl, C., Gratzel, M., and Vogel, H. 1992. Self-Assembly of Thiolipid
Molecular Layers on Gold Surfaces - Optical and Electrochemical
Characterization. Thin Solid Films 210(1-2):818-821.

Lang, H., Duschl, C., and Vogel, H. 1994. A New Class of Thiolipids for the Attachment
of Lipid Bilayers on Gold Surfaces. Langmuir 10(1):l97-210.

Lapinski, M.M., Castro-Forero, A., Greiner, A.J., Ofoli, R.Y., and Blanchard, G.J. 2007.
Comparison of liposomes formed by sonication and extrusion: Rotational and
translational diffusion of an embedded chromophore. Langmuir 23:11677-11683.

Leroueil, P.R., Berry, S.A., Duthie, K., Han, G., Rotello, V.M., McNemy, D.Q., Baker,
J .R., Orr, B.G., and H011, M.M.B. 2008. Wide varieties of cationic nanoparticles
induce defects in supported lipid bilayers. Nano Letters 8(2):420-424.

171

 

Lu, J., Drzal, L.T., Worden, R.M., and Lee, I. 2007. Simple fabrication of a highly

sensitive glucose biosensor using enzymes immobilized in exfoliated graphite
nanoplatelets Naﬁon membrane. Chemistry of Materials 19:6240-6246.

Lynch, E.C., Blake, M.S., Gotschlich, EC, and Mauro, A. 1984. Studies of Porins -
Spontaneously Transferred ﬁom Whole Cells and Reconstituted ﬁom Puriﬁed

Proteins of Neisseria-Gonorrhoeae and Neisseria-Meningitidis. Biophysical
Journal 45(1):]04-107.

Malik, N., Wiwattanapatapee, R., Klopsch, R., Lorenz, K., Frey, H., Weener, J .W.,
Meijer, E.W., Paulus, W., and Duncan, R. 2000. Dendrimers: Relationship
between structure and biocompatibility in vitro, and preliminary studies on the
biodistribution of I-125-1abelled polyamidoamine dendrimers in vivo. Journal of
Controlled Release 65(1-2):133-148.

Mauro, A., Blake, M., and Labarca, P. 1988. Voltage Gating of Conductance in Lipid
Bilayers Induced by Porin ﬁ'om Outer-Membrane of Neisseria-Gonorrhoeae.

Proceedings of the National Academy of Sciences of the United States of America
85(4):]071-1075.

Mecke, A., Lee, D.K., Rarnamoorthy, A., Orr, B.G., and H011, M.M.B. 2005a. Synthetic
and natural polycationic polymer nanoparticles interact selectively with ﬂuid-
phase domains of DMPC lipid bilayers. Langmuir 21(19):8588-8590.

Mecke, A., Majoros, I.J., Patri, A.K., Baker, J.R., Holl, M.M.B., and Orr, B.G. 2005b.
Lipid bilayer disruption by polycationic polymers: The roles of size and chemical
functional group. Langmuir 21(23):10348-10354.

Mecke, A., Uppuluri, S., Sassanella, T.M., Lee, D.K., Rarnarnoorthy, A., Baker, J .R., Orr,
B.G., and H011, M.M.B. 2004. Direct observation of lipid bilayer disruption by
poly(arnidoamine) dendrimers. Chemistry and Physics of Lipids l32(l):3-14.

Miller, C. 1986. Ion channel reconstitution. Plenum press, New York.

Minetti, C., Blake, MS, and Remeta, DP. 1998. Characterization of the structure,
function, and conformational stability of PorB class 3 protein front Neisseria
meningitidis - A porin with unusual physicochemical properties. Journal of
Biological Chemistry 273(39):25329-2533 8.

Minetti, C., Blake, M.S., Tai, J .Y., and Remeta, D.P. 1997a. Characterization of native
and recombinant forms of Por A and Por B gene products from Neisseria
meningitidis. Biophysical Journal 72(2):THAM4—THAM4.

Minetti, C., Tai, J.Y., Blake, M.S., Pullen, J.K., Liang, S.M., and Remeta, D.P. 1997b.
Structural and ﬁmctional characterization of a recombinant PorB class 2 protein
from Neisseria meningitidis - Conformational stability and porin activity. Journal
of Biological Chemistry 272(16): 10710-10720.

172

 

 

 

Moncelli, M.R., Becucci, L., and Schiller, SM. 2004. Tethered bilayer lipid membranes
self-assembled on mercury electrodes. Bioelectrochemistry 63(1-2): 161-167.

Murakarni, K., Gotschlich, EC, and Seiff, ME. 1989. Cloning and Characterization of
the Structural Gene for the Class 2 Protein of Neisseria-Meningitidis. Infection
and Immunity 57(8):2318-2323.

Naumann, C.A., Prucker, O., Lehrnann, T., Ruhe, J ., Knoll, W., and Frank, C.W. 2002.
The polymer-supported phospholipid bilayer: Tethering as a new approach to
substrate-membrane stabilization. Biomacromolecules 3(1):27-35.

Naumann, M., Rudel, T., and Meyer, T.F. 1999. Host cell interactions and signalling with
Neisseria gonorrhoeae. Current Opinion in Microbiology 2(1):62-70.

Naumann, R., Schiller, S.M., Giess, F., Grohe, B., Hartman, K.B., Karcher, 1., Koper, I.,
Lubben, J ., Vasilev, K., and Knoll, W. 2003a. Tethered lipid Bilayers on ultraﬂat
gold surfaces. Langmuir 19(13):5435-5443.

Naumann, R., Walz, D., Schiller, S.M., and Knoll, W. 2003b. Kinetics of valinomycin-
mediated K+ ion transport through tethered bilayer lipid membranes. Journal of
Electroanalytical Chemistry 550:241-252.

Neher, E., Sakrnann, B., and Steinbach, J .H. 1978. Extracellular Patch Clamp - Method
for Resolving Currents through Individual Open Channels in Biological-
Membranes. Pﬂugers Archiv-European Journal of Physiology 375(2):219-228.

Niwa, O., and Tabei, H. 1994. Voltammetric Measurements of Reversible and Quasi-
Reversible Redox Species Using Carbon-Film Based Interdigitated Array
Microelectrodes. Analytical Chemistry 66(2):285-289.

Ohlsson, P.A., Tjarnhage, T., Herbai, E., Lofas, S., and Puu, G. 1995. Liposome and
Proteoliposome Fusion onto Solid Substrates, Studied Using Atomic-Force
Microscopy, Quartz-Crystal Microbalance and Surface-Plasmon Resonance -
Biological-Activities of Incorporated Components. Bioelectrochemistry and
Bioenergetics 38(1): 1 37-148.

Ottovaleitrnannova, A., and Tien, HT. 1992. Bilayer Lipid-Membranes - an
Experimental System for Biomolecular Electronic Devices Development.
Progress in Surface Science 41 (4):337-445.

Pantoja, R., Nagarah, J.M., Starace, D.M., Melosh, N.A., Blunck, R., Bezanilla, F., and
Heath, J .R. 2004. Silicon chip-based patch-clamp electrodes integrated with
PDMS rrricroﬂuidics. Biosensors & Bioelectronics 20(3):509-517.

Peng, Z.Q., Tang, J.L., Han, X.J., Wang, E.K., and Dong, SJ. 2002. Formation of a
supported hybrid bilayer membrane on gold: A sterically enhanced hydrophobic
effect. Langmuir 18(12):4834-4839.

173

 

Qi, H.L., Tai, J .Y., and Blake, MS. 1994. Expression of Large Amounts of Neisseria]
Porin Proteins in Escherichia-Coli and Refolding of the Proteins into Native
Trimers. Infection and Immunity 62(6):2432-243 9.

Radler, J., Strey, H., and Sackrnann, E. 1995. Phenomenology and Kinetics of Lipid
Bilayer Spreading on Hydrophilic Surfaces. Langmuir 11(11):4539-4548.

Raguse, B., Braach-Maksvytis, V., Cornell, B.A., King, L.G., Osman, P.D.J., Pace, R.J.,
and Wieczorek, L. 1998. Tethered lipid bilayer membranes: Formation and ionic
reservoir characterization. Langmuir 14(3):648-659.

Rarnsden, C. 2009. Suicide-inactivation of tyrosinase by catecholic substrates. F ebs
Journal 276:30-30.

Richter, R., Mukhopadhyay, A., and Brisson, A. 2003. Pathways of lipid vesicle
deposition on solid surfaces: A combined QCM-D and AF M study. Biophysical
Journal 85(5):3035-3047.

Richter, R.P., Berat, R., and Brisson, AR. 2006. Formation of solid-supported lipid
bilayers: An integrated view. Langmuir 22(8):3497-3505.

Richter, RP, and Brisson, AR. 2005. Following the formation of supported lipid
bilayers on mica: A study combining AFM, QCM-D, and ellipsometry.
Biophysical Journal 88(5):3422-3433.

Rudel, T., Schmid, A., Benz, R., Kolb, H.A., Lang, F ., and Meyer, T.F. 1996. Modulation
of Neisseria porin (PorB) by cytosolic ATP/GTP of target cells: Parallels between
pathogen accommodation and mitochondrial endosymbiosis. Cell 85(3):391-402.

Sackrnann, E. 1996. Supported membranes: Scientiﬁc and practical applications. Science
271(5245):43-48.

Sauerbrey, G. 1959. Verwendung Von Schwingquarzen Zur Wagung Dunner Schichten
Und Zur Mikrowagung. Zeitschrzft Fur Physik 155(2):206-222.

Schiller, S.M., Naumann, R., Lovejoy, K., Kunz, H., and Knoll, W. 2003. Archaea
analogue thiolipids for tethered bilayer lipid membranes on ultrasmooth gold
surfaces. Angewandte Chemie-International Edition 42(2):208-211.

Schneiter, R., and Daum, G. 2006. Extraction of yeast lipids. Xiao W, editor. Humana
Press, Totowa, NJ. .

Sigworth, F.J., and Neher, E. 1980. Single Na+ Channel Currents Observed in Cultured
Rat Muscle-Cells. Nature 287(5781):447-449.

Sinner, E.K., and Knoll, W. 2001. Functional tethered membranes. Current Opinion in
Chemical Biology 5(6):705-711.

174

 

Snejdarkova, M., Rehak, M., and Otto, M. 1997. Stability of bilayer lipid membranes on
different metallic supports. Biosensors & Bioelectronics 12(2):]45-153.

Song, J .M., Minetti, C., Blake, MS, and Colombini, M. 1998. Successful recovery of the
normal electrophysiological properties of PorB (class 3) porin ﬁom Neisseria

meningitidis after expression in Escherichia coli and renaturation. Biochimica Et
Biophysica Acta-Biomembranes 1370(2):289-298.

Song, J .M., Minetti, C., Blake, MS, and Colombini, M. 1999. Meningococcal PorA/Cl,
a channel that combines high conductance and high selectivity. Biophysical
Journal 76(2):804-813 .

Starr, TE, and Thompson, N.L. 2002. Fluorescence pattern photobleaching recovery for
samples with multi-component diffusion. Biophysical Chemistry 97(1):29-44.

Stora, T., Lakey, J .H., and Vogel, H. 1999. Ion-channel gating in transmembrane receptor
proteins: Functional activity in tethered lipid membranes. Angewandte Chemie-
International Edition 38(3):3 89-392.

Terrettaz, S., Mayer, M., and Vogel, H. 2003. Highly electrically insulating tethered lipid
bilayers for probing the function of ion channel proteins. Langmuir l9(14):5567-
5569.

Tien, HT. 1995. Self-Assembled Lipid Bilayers as a Smart Material for Nanotechnology.
Materials Science & Engineering C-Biomimetic Materials Sensors and systems
3(1):7-12.

Tien, H.T., and Ottova, AL. 2000. Membrane Biophysics as viewed ﬁom experimental
bilayer lipid membranes. Elsevier, New York.

Tien, H.T., and Ottova, AL. 2001. The lipid bilayer concept and its experimental
realization: from soap bubbles, kitchen sink, to bilayer lipid membranes. Journal
of Membrane Science 189(1):83-117.

Vockenroth, I.K., Atanasova, P.P., Long, J .R., Jenkins, A.T.A., Kno11, W., and Koper, I.
2007. Functional incorporation of the pore forming segment of AChR M2 into

tethered bilayer lipid membranes. Biochimica et Biophysica Acta (BBA) -
Biomembranes 1768(5):l 1 14-1120.

Wang, J ., and Richardson, R.J. 2003. Tentative models for the three-dimensional
structure of the NTE esterase domain (NEST): Predictions ﬁom threading and
docking. Toxicological Sciences 72: 1487.

Wang, X.B., and Li, M. 2003. Automated electrophysiology: High throughput of art.
Assay and Drug Development Technologies 1(5):695-708.

Weel, J.F.L., Hopman, C.T.P., and Vanputten, J.P.M. 1991. Bacterial Entry and
Intracellular Processing of Neisseria-Gonorrhoeae in Epithelial-Cells -

175

 

Irnrnunomorphological Evidence for Alterations in the Maj or Outer-Membrane
Protein Pib. Journal of Experimental Medicine 174(3):705-715.

Weel, J .F .L., and Vanputten, J .P.M. 1991. Fate of the Major Outer-Membrane Protein-
P.Ia in Early and Late Events of Gonococcal-Infection of Epithelial-Cells.
Research in Microbiology l42(9):985-993.

Weiss, M.S., Abele, U., Weckesser, J., Welte, W., Schiltz, E., and Schulz, GE. 1991.
Molecular Architecture and Electrostatic Properties of a Bacterial Porin. Science
254(5038):1627-1630.

Weiss, MS, and Schulz, GE. 1992. Structure of Porin Reﬁned at 1.8 Angstrom
Resolution. Journal of Molecular Biology 227 (2):493-509.

Woodhouse, G., King, L., Wieczorek, L., Osman, P., and Cornell, B. 1999. The ion
channel switch biosensor. Journal of Molecular Recognition 12(5):328-334.

Yang, C., Jadhav, S.R., Worden, R.M., and Mason, A.J. 2009. Compact low-power
impedance extractor and digitizer for sensor array microsystems. IEEE Journal of
Solid-State Circuits 44(10):2844-2855.

Yin, P., Bums, C.J., Osman, P.D.J., and Cornell, BA. 2003. A tethered bilayer sensor
containing alametlricin channels and its detection of arniloride based inhibitors.
Biosensors & Bioelectronics 18(4):389-397.

Young, J.D.E., Blake, M., Mauro, A., and Cohn, Z.A. 1983. Properties of the Major
Outer-Membrane Protein ﬁom Neisseria-Gonorrhoeae Incorporated into Model

Lipid-Membranes. Proceedings of the National Academy of Sciences of the
United States of America-Biological Sciences 80(12):3 83 1-3 835.

Zebrowska, A., and Krysinski, P. 2004. Incorporation of Na+,K+-ATP-ase into the
thiolipid biomimetic assemblies via the fusion of proteoliposomes. Langmuir
20(25):11127-11133.

Zeng, E.W., and Zimmerman, SC. 1997. Dendrimers in supramolecular chemistry: From
molecular recognition to self-assembly. Chemical Reviews 97(5):1681-1712.

Zhou, J.H., Wu, J.Y., Hafdi, N., Behr, J.P., Erbacher, P., and Peng, L. 2006. PAMAM
dendrimers for efﬁcient siRNA delivery and potent gene silencing. Chemical
Communications (22):2362-2364.

176