 

1.... J.” n 3;. . ‘5 ‘ll‘lt .lliz‘ [1‘ :ll.
, a". .e ,._,z..‘ “1.34,, #33 «é a 5.. gigai

“$431.. 31%.:
M . ‘ .?

 

 

100?-

This is to certify that the
dissertation entitled

POLYELECTROLYTES BASED BIOMIMETIC INTERFACES FOR
BIOELECTRONIC APPLICATIONS

presented by

NEERAJ KOHLI

has been accepted towards fulﬁllment
of the requirements for the

PhD degree in Chemical EngineerinL

 

 

 

 

 

Major Professor’ 5 Signature
l/ I? /o//7

Date

MSU is an Afﬁnnative Action/Equal Opportunity Institution

—- -n- --l-I-l-I-I-O-I-l-I-i-‘-O-I-I-I-D-O-O-O-I-I.

.O-o-v-I-U—l-O-l-O-I-O-I,_O-l-l-O-O-I-I-I-O-U-D-O-l-O-.-I-O-O-I-D-‘-. -

 

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.

 

DAIEDUE

DATEDUE

DATEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRCIDaIeDue.indd-p.1

POLYELECTROLYTES BASED BIOMIMETIC INTERFACES FOR
BIOELECTRONIC APPLICATIONS

By

Neeraj Kohli

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Chemical Engineering and Materials Science

2007

ABSTRACT

POLYELECTROLYTES BASED BIOMIMETIC INTERFACES FOR
BIOELECTRONIC APPLICATIONS

By

Neeraj Kohli

Biomimetic interfaces consisting of polyelectrolyte multilayers, lipid bilayers,
proteins and other nanostructured components have been fabricated and characterized in
this study. While each chapter addresses a unique architecture or issue, the underlying
theme of the study is the development of nanostructured biomimetic architectures that
express protein activities and can be used to produce high-value devices and processes.
Neuropathy target esterase (NT E) is a membrane protein found in human neurons.
Binding of certain organophosphorus (OP) compounds to NTE is believed to cause OP-
induced delayed neuropathy (OPIDN), a type of paralysis for which there is no effective
treatment. Mutations in NTE have also been linked with serious neurological diseases,
such as Lou Gehrig’s disease. In the ﬁrst part of this dissertation, a nanostructured
biosensor to rapidly and sensitively measure the esterase activity of a fragment of NTE
known as NEST is presented. The biosensor was fabricated by co-immobilizing two
enzymes tyrosinase and NEST. The biosensor gives dose-dependent decrease in sensor
output in response to known NEST (or NTE) inhibitors. The second part of this
dissertation presents a theoretical model for bi-enzyme electrode biosensors that use
substrate recycling to increase sensitivity. The model was validated using a rotating disk
electrode on which tyrosinase and NEST were co-immobilized. The model’s predictions

were then used to quantify the effects of mass transport, partition coefﬁcients, and

enzyme kinetics on the biosensor’s metrological properties. This approach can be used to
optimize the performance of bi-enzyme electrodes that involve substrate recycling. The
third part of this study presents novel methods to produce arrays of bilayer lipid
membranes (BLMs) on pattemed polyelectrolyte multilayers. Liposomes composed of 1,
2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 1, 2-dioleoyl-sn-glycero-3-
phosphate (DOPA) were found to adsorb preferentially on poly
(dimethyldiallylammonium chloride) (PDAC) and poly(allylarnine hydrochloride) (PAH)
surfaces, and in much smaller amounts onto sulfonated poly(styrene) (SPS) surfaces.
Poly(ethylene glycol) (m-dPEG acid) coated surfaces were found to resist liposome
adsorption. These results allowed the fabrication of BLM microarrays by exposing
substrates patterned with PDAC, PAH and m-dPEG to DOPA/DOPC vesicles. The fourth
part of this study presents a novel biomimetic interface consisting of an electrolessly
deposited gold ﬁlm overlaid with a tethered BLM (tBLM). Two model membrane
biomolecules, the ionophore valinomycin and NEST, were incorporated into the tBLM,
and the activity of the resulting biomimetic interfaces was measured. Microcontact
printing (uCP) of electrolessly deposited gold patterns on glass slides was also used to
generate arrays of tBLMs. Such arrays may be useful for high-throughput screening of
drugs and chemicals that interact with cell membranes. The ﬁfth part of this study
presents an approach to fabricate high-quality, 3D patterned bionanocomposite layered
ﬁlms on substrates whose surface properties are incompatible with existing self-assembly
methods. The last part of this study presents an approach that can be used to create stable,
nanostructured, amphiphilic and cross-linkable PAMAMOS-DMOMS dendrimer

patterns.

Dedicated to my parents

iv

ACKNOWLEDGEMENTS

I would like to thank my advisors, Dr. Ilsoon Lee and Dr. Mark Worden, for their
invaluable mentorship, constant support and encouragement. It has been a pleasure
working with them. I am also grateful to them for recruiting me and giving me the
opportunity to advance my studies at Michigan State. I would also like to thank them for
being patient with me while reviewing my Ph.D. thesis and encouraging me all along the

way.

I would like to thank my committee members for agreeing to mentor me and for
their advice and support during the course of my dissertation. I would like to thank Dr.
Robert Ofoli for his inputs and for his careful critique of this dissertation, Dr. Merlin
Bruening for his assistance, and for providing me with the generous use of his laboratory

facilities.

During my years at MSU, I developed friendships that I will never forget and it
remains one of the most valuable possessions that I carry forward. Ram, Vaibhav and
Abhyuday are my earliest friends here, right from the Shaw Hall days and although
everyone has gone their separate ways, we share the best of memories. I would like to
thank the A—20 gang, Panneet, Anish, Nimisha, Smita, Rajneesh, Monojit, Gayatri for
being the best friends and roommates. I would also like to thank Devesh, Sachin Vaidya,
Sachin J adhav, Deep, Karuna and Lavanya for being great friends. I would especially like
to thank Devesh and Sachin for all the good times that we had.

I have been most inﬂuenced in my life, by my parents and sister who have

sacriﬁced a lot, and constantly encouraged me to seek higher goals and have never

doubted me. This dissertation would not have been possible if not for their love and
support. Finally I would like to thank my girl friend Vandana for her support especially
during the ﬁnal years of this dissertation. She has always been there for me, especially at
the worst times when the end seemed just as far away as the beginning. Her constant love

and encouragement has been the beacon that has guided me to the end.

vi

TABLE OF CONTENTS

1 INTRODUCTION ...................................................................................................... 1
1.1 Overvrew .......................................... 1
1.2 Nanostructured biosensor for neuropathy target esterase activity ...................... 4
1.3 Highly sensitive phenol sensor based on layer by layer assembly ..................... 5

1.4 Theoretical and experimental study of bi-enzyme electrodes with substrate
recycling .......................................................................................................................... 6

1.5 Arrays of lipid bilayers and liposomes on polyelectrolyte multilayer templates 8

1.6 Tethered lipid bilayers on electrolessly deposited gold for bio-electronic
applications ..................................................................................................................... 9

1.7 Intact transfer of layered bio-nanocomposite ﬁlms on polyelectrolyte templates

......................................................................................... 11

2 NANOSTRUCTURED BIOSENSOR FOR MEASURING NEUROPATHY

TARGET ESTERASE ACTIVITY ............................................................ _ ...................... 14
2.1 ABSTRACT ...................................................................................................... 14
2.2 INTRODUCTION ............................................................................................ 15
2.3 EXPERIMENTAL SECTION .......................................................................... 18

2.3.1 Materials ................................................................................................... 18
2.3.2 NEST expression and puriﬁcation ............................................................ 18
2.3.3 Preparation of phenyl valerate solution .................................................... 19
2.3.4 Preparation of gold electrode for NEST biosensor ................................... 19
2.3.5 Preparation of gold electrode for measuring the activity of AChE and
BChE 21

2.3.6 Ellipsometry .............................................................................................. 21
2.3.7 Potential step amperometry and other measurements ............................... 21

vii

2.4 RESULTS AND DISCUSSION ....................................................................... 22

2.4.1 Ellipsometry .............................................................................................. 22
2.4.2 Amperometric response ............................................................................ 23

2. 4. 2.1 Dependence of current response on working potential and pH ........... 23
2.4.2.2 Measurement of esterase activity using NEST biosensor ..................... 23
2.4.3 Amperometric response to catechol and phenol ....................................... 24
2.4.4 Inhibition of esterase activity .................................................................... 25
2.4.5 Storage stability ........................................................................................ 25
2.4.6 Immobilization of AChE and BChE ......................................................... 26
2.4.7 Signiﬁcance of NEST biosensor ............................................................... 27
2.5 CONCLUSIONS ............................................................................................... 28
2.6 RECOMMENDATIONS FOR FUTURE WORK ........................................... 28

3 HIGHLY SENSITIVE PHENOL SENSOR BASED ON LAYER-BY-LAYER

ASSEMBLY ..................................................................................................................... 43
3.1 ABSTRACT ...................................................................................................... 43
3 .2 INTRODUCTION ............................................................................................ 44
3.3 EXPERIMENTAL SECTION .......................................................................... 46

3.3.1 Preparation of poly-L-lysine (PLL), and tyrosinase (Tyr) multilayer
assemblies ................................................................................................................. 46
3.3.2 Ellipsometry .............................................................................................. 46
3.3.3 Reﬂectance Fourier transform infrared spectroscopy ............................... 47
3.3.4 Potential step amperometry ....................................................................... 47
3.4 RESULTS AND DISCUSSIONS ..................................................................... 47
3.4.1 Ellipsometry and FTIR ............................................................................. 47
3.4.2 Dependence of current response on working potential and pH ................ 48

viii

3.4.3 Amperometric response to catechol and phenol ....................................... 48

3.4.4 Effect of number of PLL-Tyr bilayers ...................................................... 49
3.4.5 Cyclic voltammetry ................................................................................... 49
3.4.6 Discussions ........................................ 50
3 .5 CONCLUSIONS ............................................................................................... 5 1
3.6 RECOMMENDATIONS FOR FUTURE WORK ........................................... 51

4 THEORETICAL AND EXPERIMENTAL STUDY OF BI-ENZYME

ELECTRODES WITH SUBSTRATE RECYCLING ...................................................... 58
4.] ABSTRACT ...................................................................................................... 58
4.2 INTRODUCTION ............................................................................................ 58
4.3 EXPERHVIENTAL SECTION .......................................................................... 61

4.3. 1 Materials ................................................................................................... 61
4.3.2 Preparation of gold electrode .................................................................... 61
4.3.3 Chronoamperometry and other measurements ................... ' ...................... 62
4.4 THEORETICAL MODEL ................................................................................ 62
4.4.1 Enzyme kinetics ........................................................................................ 62
4.4.2 Assumptions .............................................................................................. 64
4.4.3 Model equations ........................................................................................ 66
4.4.4 Boundary conditions ................................................................................. 67
4.4.5 Validation of the model ............................................................................ 73
4.4.6 Simulations and Discussions ..................................................................... 75
4.5 CONCLUSIONS ............................................................................................... 78

5 ARRAYS OF LIPID BILAYERS AND LIPOSOMES ON PATTERNED
POLYELECTROLYTE TEMPLATES ............................................................................ 89

ix

5.1 ABSTRACT ...................................................................................................... 89

5.2 INTRODUCTION ............................................................................................ 90
5.3 EXPERIMENTAL SECTION .......................................................................... 93
5.3.1 Materials ........................................ 93
5.3.2 Preparation of stamps ................................................................................ 94
5.3.3 Preparation of liposomes .......................................................... p ................. 94
5.3.4 Preparation of Polyelectrolyte Multilayers (PEMs) .................................. 95
5.3.5 Preparation of arrays: ................................................................................ 95
5.3.6 Total internal reﬂection microscopy: ........................................................ 96
5.3.7 Fluorescence recovery after pattern photobleaching (FRAPP): ............... 98
5.3.8 Theory and Data Analyses for Fluorescence recovery after pattern
photobleaching .......................................................................................................... 99
5.3.9 Theory ..................................................................................................... 100
5.3.10 Data Analysis and curve ﬁtting ............................................................... 102
5.3.11 Microgravimetric experiments: ........................................... ' .................... 103
5.4 RESULTS AND DISCUSSIONS ................................................................... 103
5.4.1 Characterization of liposome adsorption by TIRFM .............................. 103
5.4.2 Formation of arrays of lipid bilayers ...................................................... 106
5.4.3 Assessment of liposome adsorption and rupture to form bilayers .......... 107
5.4.4 Characterization of liposome adsorption and rupture by QCM .............. 109
5.5 CONCLUSIONS ............................................................................................. 1 1 1
5.6 RECOMMENDATIONS FOR FUTURE WORK ......................................... 112

6 TETHERED LIPID BILAYERS ON ELECTROLESSLY DEPOSITED GOLD
FOR BIOELECTRONIC APPLICATIONS ................................................................... 126

6.1 ABSTRACT .................................................................................................... 126

6.2 INTRODUCTION .......................................................................................... 1 27

6.3 EXPERIMENTAL SECTION ........................................................................ 129
6.3.1 Materials ................................................................................................. 129
6.3.2 Preparation ofstamps ...................................... 130
6.3.3 Synthesis of colloidal gold particles ....................................................... 130
6.3.4 Formation of gold ﬁlm and patterns by electroless deposition ............... 130
6.3.5 Preparation of liposomes for ﬂuorescence and electrochemical
measurements .......................................................................................................... l 3 1
6.3.6 Formation of tBLMs containing valinomycin ........................................ 132
6.3.7 Formation of NEST-containing liposomes (NEST-DOPC liposomes) .. 132
6.3.8 Preparation of phenyl valerate micellar suspension ................................ 133
6.3.9 Formation and characterization of NEST-containing tBLMs ................. 133
6.3.10 Arrays of lipid bilayers ........................................................................... 134
6.3.11 Electrochemical impedance spectroscopy (EIS) measurements ............. 134
6.3.12 Fluorescence recovery after pattern photobleaching (F RAPP) .............. 135
6.3.13 Other measurements ................................................................................ 135

6.4 RESULTS AND DISCUSSIONS ................................................................... 136
6.4.1 Formation of gold ﬁlms .......................................................................... 136
6.4.2 Functional tBLMs containing valinomycin ............................................ 137
6.4.3 Functional BLM containing NEST protein ............................................. 138
6.4.4 BLM ﬂuidity measurement using FRAPP .............................................. 139
6.4.5 Fabrication and characterization of gold patterns ................................... 139
6.4.6 Fabrication of tBLM arrays .................................................................... 140
6.4.7 Signiﬁcance of results and potential applications ................................... 141

6.5 CONCLUSIONS ............................................................................................. 142

6.6 RECOMMENDATIONS FOR FUTURE WORK ......................................... 143

xi

7 INTACT TRANSFER OF LAYERED BIONANOCOMPOSITE ARRAYS ....... 154

7.1 ABSTRACT .................................................................................................... 154
7.2 INTRODUCTION.................................................... ...................................... 154
7.3 EXPERIMENTAL SECTION ........................................................................ 157
7.3.1 Materials ................................................................................................. 157
7.3.2 Preparation of polyelectrolyte multilayers (PEMs) ................................ 158
7.3.3 Fluorescent labeling of PAMAM dendrimers, sADH and sDH ............. 15 8
7.3.4 Fabrication of 3-D structures or architectures ........................................ 159
7.3.5 Other measurements ................................................................................ 160
7.4 RESULTS AND DISCUSSIONS ................................................................... 160
7.5 CONCLUSIONS ............................................................................................. 165

8 NAN OSTRUCTURED CROSS-LINKABLE MICROPATTERNS VIA

AMPHIPHILIC DENDRIMER STAMPING. . .............................................................. 177
8. 1 ABSTRACT .................................................................................................... 177
8.2 INTRODUCTION .......................................................................................... 1 78
8.3 EXPERIMENTAL SECTION ........................................................................ 1 81
8.4 RESULTS AND DISCUSSIONS ................................................................... 183
8.5 CONCLUSIONS ............................................................................................. 188

9 REFERENCES ....................................................................................................... 195

xii

LIST OF FIGURES

Figure 2.1: Molecular architecture of the NEST biosensor. ............................................. 30

Figure 2.2: Ellipsometric thicknesses after the successive addition of following layers:
thioctic acid (point a), PLL-Tyr ﬁrst bilayer (point b), PLL-Tyr second bilayer (point
c), PLL-Tyr third bilayer (point (I), and PLL and NEST ﬁnal bilayer (point e). ...... 31

Figure 2.3: Effect of working potential on the response current of the enzyme electrode in
0.1 M phosphate buffer (pH 7.0) with (i) and without (ii) 12 uM phenyl valerate
solution, in 0.1 M phosphate buffer at an applied potential of -0.1 V (vs Ag/AgCl).
................................................................................................................................... 32

Figure 2.4: Effect of pH on the response current of the electrode, in the presence of 12
uM phenyl valerate solution, in 0.1 M phosphate buffer at an applied potential of —
0.1 V (vs Ag/AgCl). .................................................................................................. 33

Figure 2.5: (a) Current time response of the NEST biosensor to the addition of aliquots of
4 uM phenyl valerate, in 0.1 M phosphate buffer, pH 7.0, at an applied potential of -
0.1 V (vs Ag/AgCl). (b) Calibration plot. ................................................................. 34

Figure 2.6: Control experiment: Current time response on an electrode containing only
tyrosinase. The electrode was assembled in exactly the same way as NEST
biosensor, except that the ﬁnal NEST layer was not deposited. ............................... 35

Figure 2.7: (a) Current time response of the NEST biosensor to the addition of aliquots of
4 uM phenol in 0.1 M phosphate buffer, pH 7.0 at an applied potential of -0.1 V (vs
Ag/AgCl). (b) Calibration plot .................................................................................. 36

Figure 2.8: (a) Current time response of the NEST biosensor to the addition of aliquots of
8 pM catechol in 0.1 M phosphate buffer, pH 7.0 at an applied potential of -O.1 V
(vs Ag/AgCl). (b) Calibration plot. ........................................................................... 37

Figure 2.9: Current time response of NEST biosensor to the addition of 8 uM phenyl
valerate followed by the addition of 10 pM of NEST inhibitor PMSF, in 0.1 M
phosphate buffer, pH 7.0 ........................................................................................... 38

Figure 2.10: Current time response of NEST biosensor to the addition of 8 pM phenyl
valerate followed by the addition of 100 uM of NEST inhibitor PMSF, in 0.1 M
phosphate buffer, pH 7.0 ........................................................................................... 39

Figure 2.11: Current time response of NEST biosensor to the addition of 8 uM phenyl
valerate followed by the addition of 100 uM of NEST inhibitor PMSF, in 0.1 M
phosphate buffer, pH 7.0 ..................................................................... 40

Figure 2.12: Current time response of a bi-enzyme electrode consisting of tyrosinase and
AChE to the addition of aliquots of phenyl acetate to obtain ﬁnal phenyl acetate

xiii

concentrations as increments of 8 uM, in 0.1 M phosphate buffer, pH 7.0, at an
applied potential of -O.1 V (vs Ag/AgCl) ................................................... 41

Figure 3.1: Ellipsometric thickness and FTIR absorbances. at 1665 cm‘1 for PLL-Tyr
multilayered ﬁlms. .................................................................................................... 53

Figure 3.2: FTIR spectra of PLL-Tyr ﬁlms composed of up to 6 bilayers. ...................... 54

Figure 3.3 (a) Current time response of the gold electrode containing 6 PLL-Tyr bilayers
to the addition of aliquots of 8 uM catechol in 0.1 M phosphate buffer, pH 7.0 at an
applied potential of -O.1 V (vs Ag/AgCl). (b) Calibration plot. ............................... 55

Figure 3.4: Effect of number of PLL-Tyr bilayers on the current sensitivity of the enzyme
electrode to catechol. ................................................................................................ 56

Figure 3.5: (a) Cyclic voltammograms of the enzyme electrode containing 2 PLL-Tyr
bilayers, in the presence of different concentrations of catechol: (i) 3.5 uM, (ii) 7
uM, and (iii) 10.5 uM. (b) Representative example of a calibration plot. All the
measurements were performed in 0.1 M phosphate buffer, pH 7.0, and scan rate of
100 mVs“. ................................................................................................................. 57

Figure 4.1: Molecular architecture of bi-enzyme electrode .................................... 79

Figure 4.2: Schematic representation of a rotating disk bi-enzyme electrode and principle
of its functioning in the presence of phenyl valerate substrate. S1, S2, S3 and Q4 denote
the substrate phenyl valerate, phenol, catechol and o-quinone, respectively. E1 denotes
NEST esterase activity. E 2 and E3 denote tyrosinase’s phenolase and catecholase activity,

 

respectively. L denotes the thickness of the enzyme ........................................... 80
Sc:
c

Figure 4.3: Reciprocal plots of S gt and (1 versus the square root of rotation rate for
S ct

electrode A ........................................................................................... 81

Figure 4.4: Cathodic sensitivity, S; h , in the presence of phenol as a function of rotation
rate ..................................................................................................... 82

Figure 4.5: Cathodic sensitivity, ng , in the presence of phenyl valerate as a function of

rotation rate .......................................................................................... 83

xiv

Figure 4.6: Concentration proﬁles of (a) phenyl valerate (b) phenol, catechol and o-
quinone normalized to phenyl valerate bulk concentration (S,(oo)) as a ﬁmction of
relative position (x/L) within the bi-enzyme interface .......................................... 84

Figure 4.7: Concentration proﬁle of o-quinone normalized to phenyl valerate bulk
concentration(Sl(oo))at various rotation rates.................. ............................... 85

Figure 4.8: Current sensitivity, ng, as a function of amount of NEST esterase activity

(61) and tyrosinase’s catecholase activity (63) .................................................. 86

Figure 4.9: Signal ampliﬁcation in bi-enzyme electrode due to the recycling of catechol.
For simulation, the following values of different parameters were used: Pm =

0.009lcm/s, De = 2.2 x 10'5 cmz/s, (0 =500 rpm ......................................... 87
Figure 5.1: Structure of (a) DOPC, (b) DOPA, (c) NBD-PC ......................................... 114
Figure 5.2: Structures of (a) PDAC, (b) SPS, (c) PAH, and (d) m-dPEG acid. ............. 115

Figure 5.3: Schematic representation of the process for the fabrication of arrays of BLMs
on PDAC and PAH patterned substrates. ............................................................... 116

Figure 5.4: Schematic representation of the process for the fabrication of arrays of BLMs
on m-dPEG acid patterned substrates. .................................................................... 117

Figure 5.5: Experimental set-up for ﬂuorescence recovery after pattern photobleaching
using EPI—illumination. ........................................................................................... 118

Figure 5.6: Adsorption curves of (A) liposomes (10% DOPA, 90% DOPC) on PDAC. (B)
liposomes (10% DOPA, 90% DOPC) on SPS. (C) liposomes (20% DOPA, 80%
DOPC) on PDAC. (D) liposomes (20% DOPA, 80% DOPC) on SPS ................... 119

Figure 5.7: a) Adsorption of liposomes (10% DOPA, 90% DOPC) on a glass slide coated
with PEMs with m-dPEG acid being the topmost layer. b) Buffer-wash experiments
to study liposome desorption from PEMs. The top and bottom curves depict
desorption of liposomes from PEMs with PDAC and m-dPEG as the top layer,
respectively. At t=0, adsorption of liposomes (which have adsorbed for at least 45
min) is halted by introducing liposome-free buffer. In each curve, the ﬂuorescence
intensity has been normalized by the corresponding ﬂuorescence value obtained
prior to initiation of the buffer wash. ...................................................................... 120

Figure 5.8: Fluorescence images showing (a) line patterns on a PDAC patterned substrate

(b) circular patterns on a PDAC patterned substrate (c) line patterns on a PAH
patterned substrate. ................................................................................................. 121

XV

Figure 5.9: Fluorescence microscopy images showing (a) line patterns on m-dPEG acid
patterned substrate (b) circular patterns on m-dPEG patterned substrate. .............. 122

Figure 5.10: Fluorescence recovery after pattern photobleaching (FRAPP) proﬁles on
PEMs with (a) PDAC and (b) PAH as topmost layer. Only post bleach ﬂuorescence
intensity normalized against the corresponding pre bleach ﬂuorescence value is
shown. The solid lines in the ﬁgures represent ﬁts to the recovery data set with
models (Wright, Palmer et a1. 1988) describing populations with a single mobile
fraction and an immobile ﬁ'action. Plots of residuals vs. time are also indicated
below each ﬁgure. Average values obtained with these models are summarized in
Table 1 .................................................................................................................... 123

Figure 5.11: Changes in QCM resonant frequency (Curves i and iii) and dissipation
(Curves ii and iv) versus time for the adsorption of liposomes on (a) PEMs having a
top layer of PDAC (b) PEMs having a top layer of PAH. ...................................... 124

Figure 6.1: Structure of various lipid molecules: (a) DGP (reservoir lipid); (b) DPGP
(mobile lipid); (c) NBD-PE. ................................................................................... 144

Figure 6.2: Schematic representation of the process for the deposition of Au ﬁhns: (a)
plasma-treated glass slide; (b) slide after silanization with MPS; (c) modiﬁed slide
after dipping in Au particle solution; ((1) formation of conductive Au ﬁlms after
seeding step ............................................................................................................. 145

Figure 6.3: Schematic representation of the process for the fabrication of Au patterns: (a)
stamping of MPS on a glass slide; (b) MPS patterns on a glass slide; (c) MPS-
pattemed modiﬁed glass slide after heating; ((1) MPS-patterned slide aﬁer dipping in
Au colloid solution; (e) formation of Au patterns after seeding step. .................... 146

Figure 6.4: F E-SEM images showing the grth of gold on glass slides: (a) after dipping
in colloidal solution; (b) seeding once; (c) seeding twice; (d) seeding three times. 147

Figure 6.5: a) Representation of a tethered bilayer membrane containing an ionophore
(valinomycin). (b) Electrochemical spectra in a 50 mM KCl / 50 mM NaCl aqueous
solution of a tethered lipid bilayer (curve i), and a valinomycin-containing lipid
bilayer (curve ii). (c) Equivalent circuit of a bilayer. ............................................. 148

Figure 6.6: Fluorescence recovery after pattern photobleaching (FRAPP) proﬁle of lipid
bilayers on electrolessly deposited gold ﬁlms. The dots represent post-bleach
ﬂuorescence intensity normalized against the corresponding pre-bleach ﬂuorescence
value. The solid line in the ﬁgure represents the ﬁt of the recovery data. .............. 149

Figure 6.7: Optical microscopy images of gold: (a) circular patterns on a glass slide; (b)
line patterns on a glass slide .................................................................................... 150

xvi

Figure 6.8: EDS image showing (a) X-rays collected from Region 1 of a SEM image; (b)
X-rays collected from Region 2 of a SEM image. There is a peak corresponding to
Au in X-rays collected from Region 1 but not in X-rays from Region 2, conﬁrming
the selective deposition of Au on the patterned surface. ........................................ 151

Figure 6.9: (a) Topographical AFM image of gold patterns; (b) 3-D image of the pattems;
(c) pattern height on glass slide. All these images were obtained in air using tapping
mode ........................................................................................................................ 152

Figure 6.10: Fluorescence image showing line patterns of a lipid bilayer consisting of
NBD-PE as the ﬂuorophores. The image was obtained using a ﬁlter cube (Ex: 465-
495/DM: SOS/Em: 515-555). .................................................................................. 153

Figure 7.1: Schematic representation of the procedure used for printing: (a) stamp coated
with sequential layers of proteins (sADH or sDH) and PEMs (PDAC/SPS bilayers)
(Case 1); alternating layers of sADH and G4 PAMAM dendrimers on PEMs (case
2); and PDAC/SPS bilayer sandwiched between sADH layers (case 3). (b) Glass
slide coated with PEMs (10.5 bilayers). (c) Patterned substrate ............................. 167

Figure 7.2: (a) Two-dimensional AFM image of sADH patterns, on a PEM (10.5
PDAC/SPS bilayers) coated glass substrate. (b) Cross-sectional AFM image of
sADH patterns, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate. ...... 168

Figure 7.3: Two-dimensional and cross-sectional AF M images: (a—b) when a
(sADH)1(PDAC/SPS)20 multilayer ﬁlm was transferred to a PEM (10.5 PDAC/SPS
bilayers) coated glass substrate; (c-d) when a (sADH)1(PDAC/SPS)40 multilayer ﬁhn
was transferred to a PEM (10.5 PDAC/SPS bilayer) coated glass substrate. ......... 169

Figure 7.4: Fluorescence images of: (a) the circular patterns obtained on the transfer of a
multilayer ﬁlm consisting of a layer of ﬂuorescently labeled sADH and 20
PDAC/SPS bilayers, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate;
(b) the PDMS stamp surface after printing. ............................................................ 170

Figure 7.5: Schematic representation of the process used for multilevel and
multicomponent stamping ....................................................................................... 171

Figure 7.6: (a) Red ﬂuorescence from vertically printed (Alexa ﬂuor labeled-
sDH)1(PDAC/SPS)20 lines on a PEM (10.5 PDAC/SPS bilayers) coated glass
substrate; (b) Green ﬂuorescence from horizontally printed (F ITC labeled-
sADH)1(PDAC/SPS)20 lines, on a PEM (10.5 PDAC/SPS bilayers) coated glass
substrate;(c) Digitally combined ﬂuorescence image showing both red and green
ﬂuorescence ............................................................................................................ 172

Figure 7.7: (a) Fluorescence image of the patterned ﬁlms of sADH, G4 PAMAM
dendrimers and PEMs (3O PDAC/SPS bilayers) with ﬂuorescently labeled sADH as
the topmost layer, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate. (b)
Fluorescence image of the line patterns of ﬂuorescently labeled dendrimers

xvii

sandwiched between patterned sADH layer and PEMs (30 PDAC/SPS), on a PEM
(10.5 PDAC/SPS bilayers) coated glass substrate. ................................................. 173

Figure 7.8: (a—b) Two-dimensional and cross-sectional AFM images of a patterned ﬁlm
containing sequential layers of sADH, PAMAM dendrimers and PEMs (30
PDAC/SPS bilayers), with sADH as the topmost layer, on a PEM (10.5 PDAC/SPS
bilayers) coated glass substrate. (c-d) Two-dimensional and cross-sectional AFM
images of a patterned ﬁlm containing sequential layers of sADH and PAMAM
dendrimers with sADH as the topmost layer, on a PEM (10.5 PDAC/SPS bilayers)
coated glass substrate. ............................................................................................. 174

Figure 7.9: (a—b) Fluorescence images of the patterned ﬁlms of a PDAC/SPS bilayer
sandwiched between F ITC labeled sADH base layer and Alexa ﬂuor labeled sADH
topmost layer, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate, (a)
Green ﬂuorescence emanating from FITC labeled sADH base layer, obtained using
the following ﬁlter set, Ex: 465-495/DM: SOS/Em: 515-555; (b) Red ﬂuorescence
emanating from Alexa ﬂuor labeled sADH topmost layer, obtained using the
following ﬁlter set, Ex: 510-560/DM: 565/Em: 590-690. (c) Cross-sectional
topographical image of a patterned ﬁlm containing sequential layers of sADH and 1
PDAC/SPS bilayer, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate. (d)
Topographical AFM image of a patterned ﬁlm consisting of a PDAC/SPS bilayer
sandwiched between two sADH layers, on a PEM (10.5 PDAC/SPS bilayers) coated

glass substrate. ........................................................................................................ 175
Figure 7.10: Arrays of amphiphilic proteins obtained on patterned substrate using
directed self-assembly .................... 176
Figure 8.1: Schematic representation of PAMAMOS-DMOMS dendrimer ............... 185

Figure 8.2: (a) An example of an AFM image showing unstable circular patterns on a
glass slide obtained from a 0.5% PAMAMOS dendrimer solution. (b-d)
Visualization of PAMAMOS dendrimer patterns on various substrates. (b) Optical
micrograph of the circular patterns on a glass substrate (c) Fluorescence image of
the circular pattern on a glass substrate; ((1) Fluorescence image of the line pattern
on a SPS surface .............................................................................. 186

Figure 8.3: AFM images of cross-linked PAMAMOS dendrimer patterns. (a) Patterned
glass substrate stamped by 1wt.-% dendrimer solution in methanol for a contact time
of 5 min; (b) Patterned silicon wafer stamped by a 1wt.—% dendrimer solution in
methanol for a contact time of 25 min; (c) Pattern height on glass substrate; ((1)
Pattern height on silicon wafer; (e) 3-D image of the patterned glass substrate; (1) 3-
D image of the patterned silicon wafer ................................................... 187

Figure 8.4: Crosslinking of PAMAMOS-DMOMS dendrimers into a covalently bonded
network structure ............................................................................ 188

xviii

Figure 8.5: Covalent bonding of PAMAMOS-DMOMS dendrimers to glass surfaces
having silanol surface groups ............................................................... 189

xix

LIST OF TABLES

Table 2-1: Performance of bi-enzyme electrodes containing tyrosinase and different
esterases. Phenyl valerate gave the highest current sensitivity for bi-enzyme
electrodes containing tyrosinase and NEST, and also for bi-enzyme electrodes
containing tyrosinase and BChE. Phenyl acetate gave the highest sensitivity for bi-
enzyme electrodes containing tyrosinase and AChE. ............................................... 42

Table 4.1: Kinetic characteristics of tyrosinase immobilized in bi-enzyme electrode.
Values of k and K were obtained from the literature (Coche-Guerente, Labbe et a].
2001). However, enzyme concentration was calculated using Equations 4.40 and
4.41 ............................................................................................... 85

Table 5-1: Average lipid diffusion coefﬁcients (D), and average mobile fractions (m) for
BLMs formed on PEMs with PDAC and PAH as the topmost layer. The estimates
were obtained by averaging parameters ﬁt from all recovery curves. .................... 125

Table 8-1: Effects of concentration and contact time on the stability and uniformity of
patterns on different substrates ............................................................ 194

XX

Cdl

NOMENCLATURE

Double layer capacitance
Diffusion coefﬁcient of catechol, phenol or phenyl valerate in bulk solution

Diffusion coefﬁcient of catechol, phenol or phenyl valerate inside the enzyme

layer

Concentration of active sites of NEST inside the bi-enzyme electrode
Concentration of mono-phenolase active sites inside the bi-enzyme electrode

Concentration of mono-phenolase active sites inside the bi-enzyme electrode

Faraday constant
Number of electrons transferred

Current density

Partition coefﬁcient

Catalytic turnover number
Michaelis-menten constant

Rate of mass transfer in the bulk
Permeability inside the enzyme layer

Dimensionless parameter that compares the enzymatic rate of species i with its

diffusion inside the bi-enzyme layer

xxi

Ra Charge transfer resistance

RS Resistance of solution

S1 Phenyl valerate concentration inside the enzyme layer
S 2 Phenol concentration inside the enzyme layer

S3 Catechol concentration inside the enzyme layer

Q4 Quinone concentration inside the enzyme layer
Kl Reaction length related to phenyl valerate
7i. 2 Reaction length related to phenol

7t 3 Reaction length related to catechol

ZW Warburg-diffusion impedance

ABBREVIATIONS
AF M Atomic force microscopy
AChE Acetylcholinesterase
ALS Amyotrophic lateral sclerosis
BChE Butyrylcholinesterase
BLM Bilayer lipid membrane

CHAPS [(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate

xxii

CVD

DGP

DPGP
DOPA
DOPC
DTT
EIS
EDS
EDTA
FITC
FRAP
FRAPP
FRET
FTIR
HEPES
ITP
LBL
pCP

MPS

Chemical vapor deposition

1 ,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N—[3-(2-pyridyldithio)
propionate

1 ,2-di-0-phytanyl-sn-glycero-3-phosphoethanolamine
1,2-dioleoyl-sn-Glycero-3-Phosphate (Monosodium Salt)
1 ,2—dioleoyl—sn-Glycero—3-Phosphocholine
Dithiothreitol

Electrochemical impedance spectroscopy

Energy dispersive spectroscopy
Ethylenediarninetetraacetic acid

Fluorescein isothiocyanate

Fluorescence recovery after photobleaching
Fluorescence recovery after pattern photobleaching
Fluoresence resonance energy transfer

Fourier transform infrared spectroscopy
4-(2-hydroxyethyl)piperazine-1 -ethanesulfonic acid
Intact transfer printing

Layer by layer assembly

Microcontact printing

3-mercaptopropyltrimethoxy silane

xxiii

NBD PC 1-palmitoyl-2-[6-[(7-nitro-2-l ,3-benzoxadiazol-4-yl)amino]hexanoyl]-sn-
glycero-3-phosphocholine

NTE Neuropathy target esterase

OP Organophosphorus

OPIDN Organophosphorus compounds induced delayed neuropathy

PEG Polyethylene glycol

PEO Polyethylene oxide

PEM Polyelectrolyte multilayer

PMSF Phenylmethylsulfonyl ﬂuoride

PAH Poly(allylamine hydrochloride)

PDAC Poly(diallyldimethyl ammonium chloride)

PDMS Poly(dimethylsiloxane)

PLL Poly-L-lysine

PMT Photomultiplier tube

POPS Polymer-on-polymer stamping

PVD Physical vapor deposition

QCM Quartz crystal microbalance

SAM Self-assembled monolayer

SEM Scanning electron microscopy

SPS Sulfonated polystyrene

xxiv

sADH Secondary alcohol dehydrogenase

sDH Sorbitol dehydrogenase

tBLM Tethered bilayer lipid membrane

TEM Transmission electron microscopy

TIRFM Total internal reﬂection ﬂuorescence microscopy

TYR Tyrosinase

XXV

1 INTRODUCTION
1.1 Overview

Proteins are a class of bio-macromolecules, alongside carbohydrates, lipids, and
nucleic acids, which make up the primary constituents of biological organisms. Proteins
are polymers made up of a speciﬁc sequence of amino acids joined together by peptide
bonds (Glass 1966; Alzagtat and Alli 2002; Vinson 2006). Proteins perform a wide
variety of biological functions inside all the living cells. Some proteins are enzymes,
which catalyze chemical reactions. Other proteins play structural or mechanical roles.
Some other essential functions performed by proteins include immune response, and the
storage and transport of various molecules.

Recent progress in proteomics allow proteins having desired functionalities to be
identiﬁed, produced in recombinant hosts, puriﬁed, and optimized for particular
applications. However, widespread scientiﬁc and commercial exploitation of proteins
requires development of improved biomimetic interfaces with which to measure their
activities. The development of interfaces could be especially challenging for membrane
proteins (Sanders and Oxenoid 2000; Marsh, Horvath et a1. 2002) because they also
require an artiﬁcial cell membrane, called bilayer lipid membrane (BLM), to express their
activities. Ideally, biomimetic interfaces should have the following performance
properties: (1) versatile, for use with multiple classes of proteins, (2) customizable, for
use in diverse applications, (3) robust, for long lifetimes and use in harsh conditions, and
(4) mass-producible, for inexpensive fabrication.

This dissertation describes the fabrication and characterization of functional, three

dimensional biomimetic interfaces consisting of proteins, lipid bilayers, polyelectrolytes

1

and other nanostructured components. These interfaces have some of the above
mentioned desired characteristics, and were assembled using self assembly tools such as
layer by layer assembly (LBL) introduced by Decher, microcontact printing (uCP)
developed by Whitesides and polymer on polymer stamping (POPS) deveIOped by
Hammond and co-workers. LBL assembly can be used to deposit polyelectrolyte
multilayers (PEMs) on most substrates. PEMs are thin ﬁlms (Decher and Hong 1991)
formed by electrostatic interactions between oppositely charged polyelectrolyte species to
create alternating layers of sequentially adsorbed ions. PEMs are economical to produce,
and have been extended to functional polymers (Fou and Rubner 1995), colloids (Lee,
Zheng et al. 2002; Zheng, Lee et al. 2002), biomaterials (Lvov, Ariga et al. 1995),
separations (Bruening, Harris et al. 2001; Dai, Jensen et al. 2001; Balachandra, Dai et al.
2002; Dai, Balachandra et al. 2002; Stanton, Harris et al. 2003; Miller and Bruening
2004), and templates for selective electroless metal deposition (Lee, Hammond et al.
2003). uCP (Kumar and Whitesides 1993) is a soft lithographic technique used in
physics, chemistry, materials science and biology to transfer patterned thin organic ﬁlms
to surfaces, with sub-micron resolution. Unlike other fabrication methods that merely
provide topographic contrast between the feature and the background, uCP also allows
chemical contrast to be achieved via selection of an appropriate ink. Microcontact
printing offers several advantages over conventional photolithographic techniques
because it is simple to perform and is not diffraction-limited. This technique has been
used to make patterns of various small and large molecules on metals and silicon
substrates (St. John and Craighead 1996; Xia, Kim et al. 1996; Kohli, Dvomic et al.

2004); as well as to deposit proteins (Kohli, Worden et al. 2005) and polyelectrolyte

aggregates (Lee, Ahn et al. 2004). POPS is an approach that combines LBL assembly and
ItCP to generate alternating regions of different chemical functionalities on a surface by
using graﬁ, diblock copolymers or polyelectrolytes as ink (Jiang and Hammond 2000;

Jiang, Zheng et al. 2002).

While each chapter in this thesis addresses a unique architecture or issue, the
underlying theme of the study is the development of improved biomimetic architectures
that express protein activities. These interfaces have outstanding potential for conducting
fundamental studies of nanoscale biological processes and for developing new protein-
based technologies, including high-throughput drug screening systems and novel
biosensors.

Chapter 2 of this dissertation presents a nanostructured biosensor interface for
measuring the activity and inhibition of NEST. In Chapter 3, an approach to generate
highly sensitive phenol sensors is presented. In Chapter 4, a theoretical and experimental
study of bi-enzyme electrodes where substrate recycling takes place is presented. In
Chapter 5, an efﬁcient approach for the fabrication of arrays of lipid bilayers on patterned
PEMs is presented. In Chapter 6, a novel biomimetic interface consisting of an
electrolessly deposited gold ﬁlm overlaid with a tethered BLM (tBLM) is presented. In
Chapter 7, an approach is presented that allows high-quality, 3D patterned
bionanocomposite layered ﬁlms to be constructed on substrates whose surface properties
are incompatible with existing self-assembly methods. Chapter 8 presents an approach
that can be used to create stable, nanostructured, amphiphilic and cross-linkable

PAMAMOS—DMOMS dendrimer patterns.

1.2 Nanostructured biosensor for neuropathy target esterase activity

Organophosphorous compounds (OP) can cause at least two kinds of neurotoxic
syndromes in humans and animals (Glynn 1999; Sigolaeva, Makower et al. 2001;
Sokolovskaya, Sigolaeva et al. 2005). The ﬁrst is the acute and sometimes fatal
neurotoxicity that results from covalent inhibition of acetylcholinesterase (AChE) at
cholinergic synapses. The second, OP-induced delayed neuropathy (OPIDN) is
characterized by ﬂaccid paralysis of the lower limbs 1-3 weeks after exposure to certain
neuropathic OPs. These compounds covalently modify the neuronal membrane protein
known as neuropathy target esterase (NT E). Recovery from this disease is usually poor,

and there is no known effective treatment.

In addition to its role in chemically induced disease processes, NTE has recently
been implicated in well-known motor-neuron diseases that occur spontaneously, without
exposure to OP compounds. One of our collaborators, Dr. Rudy J. Richardson
(Toxicology Program, Department of Environmental Health Sciences, University of
Michigan) recently co-authored a groundbreaking conference paper that links a mutation
in the gene encoding NTE to amyotrophic lateral sclerosis (ALS) (VandenDriessche
2005), also known as Lou Gehrig's disease. This paper, entitled, “Neuropathy target
esterase gene mutations cause motor neuron disease,” was presented at the 2005

American Society of Human Genetics meeting, Oct. 25-29, 2005, in Salt Lake City, UT.

Because NTE plays a central role in both chemically induced and spontaneously
occurring neurological diseases, approaches that can help measure its esterase activity

and inhibition are of tremendous scientiﬁc and commercial importance. In this part of the

dissertation, the ﬁrst continuous, electrochemical biosensor for measuring the esterase
activity of a fragment of NTE, known as NEST is presented. The biosensor also gives
dose-dependent drop in sensor output to known NEST inhibitors. The interface was
formed using LBL assembly of a polyelectrolyte poly-L-lysine and two proteins NEST
and tyrosinase. Note that this is the ﬁrst time that NEST has been immobilized in an

active conformation on a surface.

While the development of NEST biosensor was the primary purpose of this study,
we found that the same NEST biosensor approach can also be extended to measure the
activity and inhibition of other medically relevant esterases such as acetylcholinesterase
(AChE) and butyrylcholinesterase (BChE). AChE is found primarily in the blood and
neural synapses (Herz and Kaplan 1973; Millard and Broomﬁeld 1995; Davis, Britten et
al. 1997; Lejus, Blanloeil et al. 1998; Hsieh, Deng et al. 2001). Pseudocholinesterase,
also known as BChE, is found primarily in the liver (Darvesh, Hopkins et al. 2003). Both
of these enzymes are esterases and catalyze the hydrolysis of the neurotransmitter
acetylcholine into choline and acetic acid, a reaction necessary to allow a cholinergic

neuron to return to its resting state after activation.
1.3 Highly sensitive phenol sensor based on layer by layer assembly

Phenols are present in waste-waters and exhaust gases of many industries,
including oil reﬁneries, coke plants, resins and plastics, dyes and chemicals, and textiles
(Stanca, Popescu et al. 2003; Solna, Sapelnikova et al. 2005; Solna and Skladal 2005).
Phenols, even in low concentrations, can cause serious problems because of their toxicity.

The sensing of phenolic compounds is therefore essential to evaluate the risk of

environmental samples on our health. Chapter 3 presents an approach that can be used to
develop highly sensitive phenol sensors by making alternative layers of PLL and
tyrosinase on a gold electrode using LBL assembly. The sensitivities obtained using poly-
L—lysine and tyrosinase multilayer ﬁlms are about 2 orders of magnitude greater than
those reported using other polyelectrolytes and tyrosinase multilayer systems, suggesting
that the composition of polyelectrolytes may inﬂuence important factors such as enzyme
loading and turnover number. The work described in Chapter 2 and Chapter 3 of this
dissertation was done in collaboration with Devesh Srivastava (Advisor: Ilsoon Lee) in
the Department of Chemical Engineering and Materials Science. The work consists of

equal contribution from Devesh Srivastava and the author of this dissertation.

1.4 Theoretical and experimental study of bi-enzyme electrodes with substrate

recycling

During the past decade, biosensors, because of their high selectivity, speciﬁcity
and low cost for mass production, have become very popular in areas such as food
technology, environmental monitoring and biomedical devices (Allen 1997; Vo-Dinh and
Cullum 2000; Prodromidis and Karayannis 2002). Biosensors use a variety of methods
for detecting the analyte of interest, including electrochemical (potentiometric,
amperometric, conductometric) (Dzyadevych, Soldatkin et a1. 2002; Gerard, Chaubey et
al. 2002) optical (absorbance, ﬂuorescence, luminescence, light scattering, refractive
index) (DiazGarcia and ValenciaGonzalez 1995; Barak, Ferguson et al. 1997; Karlsson
and Falt 1997), thermal (Abel, Weller et al. 1996; Ramanathan and Danielsson 2001),

piezoelectric (Muramatsu, Dicks et al. 1987; Saini, Hall et al. 1991; Davis, Britten et al.

1997), and magnetic (Baselt, Lee et al. 1998). The electrochemical mode is generally the

preferred mode of transduction for enzyme based biosensors.

The sensitivity of the biosensors can be increased dramatically (or ampliﬁed) by
incorporating a substrate recycling scheme. In these devices the shuttle analyte is not
only measured once, but is reconverted to be measured again. The shuttle analyte can
either be the analyte of interest (the enzyme substrate) or can be produced by an

additional enzyme reaction.

The range of analytes for which reporting interfaces (biosensors) can be
developed can be increased substantially by coupling multiple protein activities into
reaction pathways. For example, NEST biosensor (described earlier), couples NEST

esterase activity with two tyrosinase oxidation activities.

In view of numerous applications of such multi-enzyme interfaces with ampliﬁed
responses, for low concentrations of substrate, a theoretical model for a bi-enzyme
rotating disk electrode having substrate recycling is presented in Chapter 4 of this
dissertation. The bi-enzyme electrode was fabricated by co-immobilizing two enzymes:
tyrosinase and NEST. The model was validated by studying the response of the bi-
enzyme rotating disk electrode to phenyl valerate under varying rotating speeds. This
model can help quantify the inﬂuence of mass transfer in bulk and enzyme layer, enzyme
kinetics and electron transfer kinetics on the performance of biosensors. This information

can be helpful in optimizing the performance characteristics of biosensors.

1.5 Arrays of lipid bilayers and liposomes on polyelectrolyte multilayer templates

Cell membranes represent one of the major structural components of biological
cells. Cell membranes consist of lipids, membrane proteins and other major
biomacromolecules interposed between them. These membrane proteins perform many
vital sensing, signaling, catalytic, transport, and bioelectronic processes at the molecular
scale. Widespread scientiﬁc and commercial exploitation of membrane proteins requires
development of improved biomimetic interfaces that can express and /or to measure their

activities.

Biosensors or biomimetic interfaces that rely on membrane bound proteins or
enzymes, often incorporate artiﬁcial bilayers (BLMs) in their design. The BLMs are
necessary in order to preserve the functional characteristics of the proteins that are bound
to them. BLMs may be deposited on the biosensor interface using Langmuir-Blodgett
deposition methods or liposome adsorption. The choice of lipids used to construct BLMs
is important, as the characteristics of the resulting BLM have a very strong inﬂuence on
the activity of the membrane protein. Properties such as membrane ﬂuidity, which is an
indicator of the ability of lipid molecules to diffuse freely through the bilayer, play
important roles in maintaining the correct conformation of the embedded biomolecules.
For certain membrane proteins such as ion channels, it is also important to have BLMs
that are highly insulating, since the presence of large defects will render such sensors

ineffective (Raguse, Braach-Maksvytis et al. 1998).

In recent years, considerable developments have been made in the area of

polymeric cushions upon which BLMs can be deposited (Sackmann and Tanaka 2000).

In addition to being able to provide ionic reservoirs, polymeric cushions allow extra-
membrane portions of membrane proteins to attain their necessary conformation and to
retain biological activity. Without the cushions, interactions between the substrate surface
and protein might impede proper function. The use of PEMs has been proposed as one

such viable method for a polymeric cushion.

Chapter 5 presents a novel approach to fabricate arrays of BLMs on patterned
PEMs. The patterned PEMs were deposited using microcontact printing and alternating
deposition of positively charged and negatively charged polyelectrolytes. Arrays of
BLMs were fabricated using exposure of patterned PEMs to liposomes of various
compositions. The deposition of liposomes on the PEMs was monitored using total
internal reﬂection ﬂuorescence microscopy, ﬂuorescence microscopy and gravimetry.
The resulting BLMs were characterized using ﬂuorescence recovery after pattern
photobleaching (FRAPP). Bilayer formation on two different kinds of PEM substrates
was studied. Such arrays can be used in applications such as high throughout screening,
and for conducting fundamental studies of nanoscale biological processes. This work was
done in collaboration with Sachin Vaidya (Advisor: Robert Ofoli) in the Department of
Chemical Engineering and Materials Science. The work consists of equal contributions

from Sachin Vaidya and the author of this dissertation.

1.6 Tethered lipid bilayers on electrolessly deposited gold for bio-electronic

applications

Biomimetic interfaces, consisting of tBLMs with membrane proteins incorporated

into them or bound to them, have primarily been used for sensing applications. New

approaches that can increase the range of tBLMs applications to include areas such as
biocatalysis and biofuel cells are of scientiﬁc and commercial interest. One of the major
challenges in the commercialization of such biomimetic interfaces is the cost of gold
electrodes on which the tBLM are deposited. Although methods such as chemical vapor
deposition (CVD) and physical vapor deposition (PVD) are generally used to deposit
gold, these methods are very expensive and also suffer ﬁ'om some limitations such as: i)
they cannot uniformly coat the pores of porous substrates, ii) they generally require an
adhesion layer. Therefore, there is a need to reduce the fabrication cost and overcome

some of the limitations associated with CVD and PVD.

Electroless deposition is an inexpensive metal-plating technique that, unlike CVD
and PVD, can easily be performed on a laboratory bench top and in aqueous solutions. It
can be used to tune the properties of arrays or ﬁlms of metals with nanometer-scale
precision on virtually any surface, including glass, silicon wafers, and mechanically
ﬂexible plastics. Porous substrates are generally used as electrodes in biocatalytic
reactors and biofuel cells. CVD and PVD are expensive and poorly suited to uniformly
coating the internal pore surfaces of porous substrates. However, electroless deposition
allows the metal plating solution to be drawn in to pores because of capillary action, thus

offering the potential for both coating of internal pores and low processing costs.

Chapter 6 presents an economical and versatile approach that can easily be
extended to fabricate functional and nanostructured biomimetic interfaces on a wide
variety of surfaces including electrodes of biocatalytic reactors, thus increasing the
repertoire of tBLM based applications. The approach entails combining electroless metal

deposition and molecular self assembly to sequentially deposit a conductive gold layer

10

and tBLM on a substrate. To conﬁrm the activity of tBLMs, two biomolecules were also
incorporate into them: the ionophore valinomycin and NEST. Electrochemical impedance
spectroscopy (EIS), UVNisible spectroscopy and ﬂuorescence recovery after pattern
photobleaching (FRAPP) were used to characterize the resulting biomimetic interfaces
and conﬁrm membrane biomolecule activity. Microcontact printing was used to deposit
arrays of tBLMs on electrolessly deposited gold patterns. The resulting arrays were
characterized using atomic force microscopy (AFM), ﬁeld-emission scanning electron

microscopy and energy dispersive spectroscopy (EDS).

1.7 Intact transfer of layered bio-nanocomposite films on polyelectrolyte

templates

Functional, 3-D nano and microarrays composed of molecules such as proteins,
lipids, polyelectrolytes and dendrimers have potential applications in drug screening
devices, novel biosensors, biocatalysis, optoelectronics, and other devices (Blawas and
Reichert 1998; Lee, Zheng et al. 2002; Kidambi, Chan et al. 2004). Chapter 7 of this
dissertation presents an approach that can be used to deposit high-quality microarrays of
layered, bionanocomposite ﬁlms on a wide variety of substrates. As examples of this
approach, the fabrication and characterization of some 3—D nanostructured architectures
(or structures) consisting of dehydrogenase enzymes, polyamidoamine (PAMAM)

dendrimers and PEMs, is presented.

The approach presented in this chapter overcomes a problem encountered when
using microcontact printing to establish a pattern on the target surface and then building

sequential layers on the pattern via LBL self-assembly. Amphiphilic molecules tend to

11

adsorb both to the patterned features as well as the underlying substrate, resulting in low-
quality patterns. By circumventing this problem, this research is expected to signiﬁcantly
extend the range of surfaces and layering constituents that can be used to fabricate 3D,
patterned, bionanocomposite structures. Fluorescence microscopy and AFM were the
primary characterization techniques used to demonstrate the feasibility of this approach

for making these 3D bionanocomposite arrays.

1.8 Nanostructured cross-linkable micropatterns via amphiphilic dendrimer

stamping

Dendrimers are radially symmetrical polymeric molecules that are grown by sequential
addition of branched monomers to the outer shell.7 With each new addition (generation),
the molecular weight of the dendrimer and the density of its terminal functional groups
increase substantially. Radially layered poly-(amidoamine-organosilicon) (PAMAMOS)
dendrimers represent a unique, highly diverse class of amphiphilic and cross-linkable

globular-shaped dendrimers.

In this Chapter, we report the ﬁrst application of microcontact printing of the amphiphilic
and cross-linkable PAMAMOS-DMOMS dendrimers on glass slides, silicon wafers and
polyelectrolyte multilayers (PEMs) in which the pattern average thickness was controlled
by spin-self assembly (i.e., spin-inking). The resulting 3-D micro-pattemed amphiphilic
networks were characterized by optical microscopy and atomic force microscopy (AFM).
Also, the effects of dendrimer ink concentration, inking method and contact time on the
thickness, uniformity and stability of the deposited patterns are presented. The results

provide a framework for controlling the geometry of the deposited patterns. The lateral

12

footprint of the pattern can be controlled by the shape of the elastomeric stamp, and the
thickness of the patterns can be controlled by adjusting the spin coating method, the
surface properties of the stamp, and the substrate used. The results also conﬁrmed the
well-known inﬂuences of spin speed, concentration, and solvent on the thickness of spin
coated ﬁlms. These versatile dendrimer arrays have potential applications in the
development of high throughput drug screening systems and could serve as templates for
nanoparticles growth or electroless metallization in development of nano metal reactors

on micropattemed ﬁlms.

13

2 NANOSTRUCTURED BIOSENSOR FOR MEASURING NEUROPATHY

TARGET ESTERASE ACTIVITY
2.1 ABSTRACT

Neuropathy target esterase (NTE) is a membrane protein found in human neurons, and
other cells, including lymphocytes. Binding of certain organophosphorus (OP)
compounds to NTE is believed to cause OP-induced delayed neuropathy (OPIDN), a type
of paralysis for which there is no effective treatment. Mutations in NTE have also been
linked with serious neurological diseases, such as Lou Gehrig’s disease. In this paper, for
the ﬁrst time, a nanostructured biosensor containing a domain of NTE known as NEST
has been developed. The biosensor was fabricated, using layer by layer assembly
approach, by immobilizing a layer of NEST on top of multilayers consisting of a
polyelectrolyte (poly-L-lysine) and a protein (tyrosinase). The biosensor has a response
time on the order of seconds and gives dose-dependent decrease in sensor output in
response to known NEST (or NTE) inhibitors. Potential applications of the biosensor
include studying the fundamental reaction kinetics of NTE mutants, screening OP
compounds for NTE inhibition, and investigating the effect of NTE mutations on NTE
esterase activity. While the development of a NEST biosensor was the primary purpose
of this study, we found that the approach developed for NEST can also easily be extended
to immobilize and measure the activity of other medically relevant esterases such as
acetylcholinesterase (AChE) and butyrylcholinesterase (BChE). Based on the measured
sensitivities, phenyl valerate was found to be the best substrate for NEST and BChE. On

the other hand, phenyl acetate gave the highest sensitivity with AChE. Potential step

14

amperometry and ellipsometry were the primary characterization techniques used in this

study.
2.2 INTRODUCTION

Neuropathy target esterase (NTE), a membrane-bound esterase found in neurons of
vertebrates (Glynn 1999; Atkins and Glynn 2000; van Tienhoven, Atkins et al. 2002; Li,
Dinsdale et al. 2003; Makhaeva, Sigolaeva et al. 2003; Kropp, Glynn et al. 2004), has
been shown to be necessary for embryonic development in mice, and is believed to be
involved in cell-signaling pathways and lipid trafﬁcking (Glynn 1999). NTE has serine
esterase activity and can hydrolyze ester, peptide, and amide bonds. The nucleophilic
serine residue (active site) of NTE attacks the carbonyl carbon atom of the substrate,
forming a covalent acyl-enzyme intermediate, which is subsequently hydrolyzed. A
consequence of this reaction mechanism is that the esterase activity of NTE is susceptible
to covalent inhibition by organophosphorus esters (OPS) with which it forms an
analogous phosphyl-enzyme intermediate. Irreversible binding of some OP compounds to
the active serine site results in a debilitating neural disease known as (OP)-induced
delayed neuropathy (OPIDN) (Glynn 1999). Symptoms of OIPDN include ﬂaccid
paralysis of the lower limbs, which becomes evident two to three weeks after exposure to
neuropathic OPs. Recovery ﬁom this disease is usually poor, and there is no speciﬁc
treatment. Because NTE is difﬁcult to produce for research purposes, research to study its
esterase activity is typically done using a fragment of the NTE protein that contains the
esterase activity and can be more easily produced One such fragment known as NEST
(NTE gs_t_erase domain) (Atkins and Glynn 2000; Forshaw, Atkins et al. 2001; Kropp,
Glynn et al. 2004) reacts with esters and inhibitors in a manner very similar to NTE.

15

Widespread and long—tenn use of OP compounds in industry and agriculture has
made these hazardous compounds a part of environment, posing a health risk. Also,
neuropathic compounds could potentially be used as chemical weapons. In addition to its
role in chemically induced disease processes, NTE has been implicated in well-known
motor-neuron diseases that don’t require exposure to OP compounds. Recently, a
conference paper linked a mutation in the gene encoding NTE to arnyotrophic lateral

sclerosis (ALS), also known as Lou Gehrig’s disease (Rainier, Bui et al.).

Because NTE plays a central role in both chemically induced and spontaneously
occurring neurological diseases, approaches that can help measure its esterase activity
and inhibition are of tremendous scientiﬁc and commercial importance. Conventionally,
the esterase activity of NTE (or NEST) is measured using two distinct steps. In the ﬁrst
step, a solution containing phenyl valerate is brought into contact with NEST or NTE
protein solution, whose esterase activity reacts with a portion of the artiﬁcial substrate
phenyl valerate to form phenol. In the second step, the concentration of phenol in the
solution is determined either colorimetrically, in the presence of 4-arnino antipyrine
(Kayyali, Moore et al. 1991), or electrochemically, in the presence of tyrosinase enzyme
(Sigolaeva, Makower et al. 2001; Sokolovskaya, Sigolaeva et al. 2005). Tyrosinase
converts phenol ﬁrst to catechol and then to o-quinone, which can be measured
electrochemically at an electrode (Makhaeva, Sigolaeva et al. 2003). The current
generated by the electrode increases with the amount of o-quinone present, thus giving an
indirect measurement of the amount of NTE esterase activity present during the ﬁrst step.
To test for esterase inhibition, this procedure is repeated both in the absence and presence

of a putative inhibitor (e.g. an OP compound). A reduced signal indicates inhibition of

16

the esterase activity. This method has the disadvantages of being slow, and requiring two
steps, making it unsuitable for some important applications, such as high-throughput
screening of compounds for NTE inhibition and continuous, on-line, environmental

monitoring to detect chemical warfare agents that target NTE.

This paper presents the ﬁrst continuous, electrochemical biosensor for real-time,
rapid measurement of NEST (or NTE) esterase activity. The biosensor was fabricated by
co-immobilizing NEST protein and tyrosinase enzyme on an electrode using layer by
layer assembly approach (Decher 1997). To our knowledge, this is the ﬁrst time NEST
has been immobilized in an active conformation on an electrode. Potential applications of
this sensor include detecting the presence of chemical weapons that target NTE,
screening industrial and agricultural OP compounds for NTE inhibition, studying the
fundamental reaction kinetics of NTE, and investigating the effect of NTE mutations

found in ALS patients on NTE’s enzymatic properties.

While the primary purpose of this paper was the development of NEST biosensor,
we found that the approach can also easily be extended to measure the activity of other
medically relevant esterases such as AChE and BChE. AChE, also known as true
cholinesterase is found primarily in neural synapses (Herz and Kaplan 1973; Davis,
Britten et al. 1997; Lejus, Blanloeil et al. 1998; Hsieh, Deng et al. 2001). BChE, also
known as pseudocholinesterase is found primarily in the blood plasma (Lejus, Blanloeil
et al. 1998; Darvesh, Hopkins et al. 2003). Although both of these enzymes are capable
of hydrolyzing the neurotransmitter acetylcholine into choline and acetic acid, a reaction
necessary to allow a cholinergic neuron to return to its resting state after activation, they

differ in their respective reactivities toward substrates. AChE prefers acetylcholine,

l7

whereas BChE prefers butyrylcholine. Inhibition of neural AChE can produce cholinergic
toxicity and death. BChE inhibition is not known to have direct toxic effects, but it can

serve as a biomarker of exposure to OP compounds (Thompson and Richardson 2004).
2.3 EXPERIMENTAL SECTION

2.3.1 Materials

Thioctic acid, poly-L-lysine (PLL) (molecular weight ~ 15,000), tyrosinase (Tyr),
sodium phosphate (monobasic and dibasic), ethylenediaminetetraacetic acid (EDTA),
sodium chloride, 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesu1fonate
(CHAPS) and isopropyl thiogalactoside (IPTG), AChE, and BChE were obtained from
Sigma (St. Louis, MO). All the chemicals were of the highest purity available. Ultrapure
water was supplied by a Nanopure-UV four-stage puriﬁer (Bamstead International,

Dubuque, IA); the puriﬁer was equipped with a UV source and a ﬁnal 0.2 pm ﬁlter.

2.3.2 NEST expression and purification

NEST was expressed and puriﬁed according to published procedures (Atkins and
Glynn 2000). Brieﬂy, DNA fragment encoding NEST was cloned into pET-21b vector,
and the resulting expression vector was transformed into E. coli BL21(DE3). An
overnight culture of transformed E. coli was inoculated with M9 media containing
ampicillin and grown in a ferrnentor. IPTG was added to the resulting cell culture after a
day to induce the expression of NEST. The resulting cells were collected 4 h after
induction by centrifugation and subjected to protein expression techniques. Brieﬂy, 5 g of
cell paste was suspended in 30 ml of PEN buffer (50mM potassium phosphate/0.3 M

NaCl/O.5 mM EDTA, pH 7.8) containing 2% CHAPS and tip sonicated four times. The

18

cell lysate was centrifuged at 2000 g for 30 min at 4°C, the supernatant was collected,
and about 7 mL of supernatant was added to a mini column (volume 10 mL) containing 3
mL of Ni-NTA resin. The mini column was rotated at room temperature for 20 min,
centrifuged at 2000 g for 20 sec and then the top solution was drawn off. The histidine
tagged NEST was eluted from the Ni-NTA resin using 10 mL of PEN buffer containing
0.3% CHAPS and 0.3 M imidazole. The protein purity was determined using SDS PAGE
and protein concentration was determined using BioRad Dc protein assay kit using
albumin as the standard. For long term storage, 25% glycerol was added to the protein
solution, which was then stored at -20 °C. The cloning and fermentation part of this
procedure was done by Dr. Jun Sun (Manager, Protein Expression Laboratory, MSU) and

the puriﬁcation part was done by the author of this dissertation.

2.3.3 Preparation of phenyl valerate solution

To prepare phenyl valerate solution, 15 mg of phenyl valerate was dissolved in 1
mL of dimethylforrnamide (DMF), and 15 mL of water containing 0.03% Triton was
added slowly under vigorous stin'ing. For potential step voltammetry experiments, small
aliquots of the resulting phenyl valerate micellar solution (5.286 mM) were added to the

phosphate buffer to obtain the desired concentrations.

2.3.4 Preparation of gold electrode for NEST biosensor

Tyrosinase is a copper-containing oxidase (Forzani, Solis et al. 2000; Coche-
Guerente, Labbe et al. 2001), which possesses two different activities, as illustrated in

reaction 2.].

l9

OH OH O
OH

0
Phenol Catechol o—quinone

(2.1)

The ﬁrst step is referred to as the enzyme’s hydroxylase activity (also known as
cresolase activity) where phenol is hydroxylated by the aid of molecular oxygen to
produce catechol. In the second step, known as the catecholase activity, the enzyme
oxidizes catechol to o-quinone and is simultaneously oxidized by oxygen to its original
form, with the production of water. The reaction product, o-quinone, is electrochemically
active and can be reduced back to the catechol form at low applied potentials, as

illustrated in reaction 2.2.
o-quinone + 2H+ + 2e' ——-> catechol (2.2)

These characteristics of tyrosinase were exploited by us to fabricate a NEST
biosensor, capable of measuring the NEST’s esterase activity and its inhibition, by co-

immobilizing NEST and tyrosinase on a gold electrode using LBL assembly approach.

The molecular architecture of the biosensor interface is shown schematically in
Figure 2.1. Gold electrodes cleaned in Piranha solution, were 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 PLL solution for 45 min. The PLL solution was prepared by

20

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 PLL-Tyr 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 1 h. The electrodes were
then washed with water, dried under nitrogen and dipped in phosphate buffer (0.1 M, pH

7 .0) for testing.

2.3.5 Preparation of gold electrode for measuring the activity of AChE and BChE

To immobilize AChE or BChE, we used the same procedure as for the NEST
biosensor, except that instead of the NEST solution, the ﬁnal dipping was done in 0.1

mg/mL phosphate buffer solution (pH 7.0) of AChE or BChE.

2.3.6 Ellipsometry

Ellipsometric measurements were obtained with rotating analyzer ellipsometer
(model M-44; J.A. Woollan Co. Inc., Lincoln, NE) using WVASE32 software. The
thickness values for dried ﬁlms were determined using 44 wavelengths between 414.0
and 736.1 nm. The angle of incidence was 75° for all experiments. Refractive indices of
ﬁlms containing PLL and proteins were assumed to be n = 1.5, k = 0. These optical
constants compare well with those determined for 4 bilayer ﬁlms consisting of poly-L-

lysine and tyrosinase using ellipsometry.

2.3.7 Potential step amperometry and other measurements

The electrodes (sensors) were maintained at a potential of -100 mV (vs Ag/AgCl
reference electrode) using a BAS CV-SOW electrochemical analyzer. The esterase

21

activity of NEST biosensor was monitored by measuring the output current for a variety
of phenyl valerate concentrations, under stirred conditions. The NEST protein converts
phenyl valerate to phenol, which gets converted to o-quinone by tyrosinase. The 0-
quinone gets reduced at the electrode’s surface, resulting in the generation of current. The
electroreduction of o—quinone produces catechol which again gets converted to o-quinone

by tyrosinase, thus amplifying the signal.

To measure inhibition of the esterase activity, a known quantity of phenyl valerate
was added to the phosphate buffer (pH 7.0), under stirred conditions. After the
stabilization of current, a known amount of NEST inhibitor was added, and the resulting
drop in current was measured. Unless otherwise speciﬁed, all the errors reported in this

chapter are 0' (where a is the standard deviation obtained using 10 electrodes).
2.4 RESULTS AND DISCUSSION

2.4.1 Ellipsometry

Ellipsometry was used to conﬁrm the deposition of different layers that make up
the NEST biosensor. As shown in Figure 2.2, the thickness increase following the
addition of ﬁrst PLL and Tyr bilayer was approximately 9.3 :I: 0.4 nm. The thickness
increase for the next two PLL-Tyr bilayers was the same and approximately equal to 7.2
d: 0.3 nm. The thickness increase following the addition of ﬁnal PLL-NEST bilayer was

approximately 6.6 :t 0.3 nm.

22

2.4.2 Amperometric response

2.4.2.1 Dependence of current response on working potential and pH

The various experimental parameters (such as pH and applied potential), which
can affect the amperometric determination of phenyl valerate, were optimized. The effect
of applied potential on the amperometric response of the sensor was tested in the range
between 0.05 and -0.20 V. Figure 2.3 shows the steady state response of the biosensor
both in the presence and absence (background current) of phenyl valerate. The
background current is thought to result largely from the direct reduction of dissolved
oxygen. The highest signal-to background ratio, was obtained at -0.1 V. At working
potential more negative then -0.1 V, higher signals were obtained, but the background
current also increased distinctly. Therefore, a working potential of -0.1 V was used for

further studies.

The effect of pH was also studied in the pH range 5.5 to 8.0 in 0.1 M phosphate
buffer at working potential of -0.1 V. As shown in Figure 2.4, the response current

attained a maximum value at pH 7.0. This pH was used for further studies.

2.4.2.2 Measurement of esterase activity using NEST biosensor

Figure 2.5a displays a typical current-time response under the optimal
experimental conditions after the successive addition of aliquots of phenyl valerate in the
buffer to obtain ﬁnal concentrations as increments of 4 uM. A well deﬁned reduction
current, proportional to the amount of phenyl valerate, was observed. The response time

of the electrode was less then 20 5, due to the nano-scale thickness of the interface. The

response to phenyl valerate was linear (R2 = 0.981) in the range 0.5 [AM to 12 uM, and it

23

reached saturation at approximately 30 uM (Figure 2.5b). The limit of detection was 0.5

nM at a signal-to-noise ratio of 3. The reproducibility of the sensor was investigated at a
phenyl valerate concentration of 4 MM; the mean current was approximately 350 nAcm'z,

with a standard deviation of approximately 10%. Figure 2.6 shows a control experiment
which was done on an electrode containing only PLL-Tyr bilayers. As expected, a
relatively very small rise in steady state current was observed on the addition of phenyl
valerate. The small rise can be attributed to the presence of small amount of phenol

produced due to auto hydrolysis of phenyl valerate solution.

It may be noted that the primary purpose of this study was the development of an
approach that can be used to measure NEST esterase activity and inhibition; 3 PLL-tyr
layers and one NEST layer gave us sufﬁcient current sensitivity to check for the
feasibility of the proposed approach. However, if some application requires higher
sensitivities, then our preliminary results (data not shown) suggest that it can be achieved

by ﬁirther increasing the number of tyrosinase or NEST layers.

2.4.3 Amperometric response to catechol and phenol

Since tyrosinase can convert both catechol and phenol to quinone, the
amperometric response of the NEST biosensor to these compounds was also studied.
Figures 2.7a and 2.8a show the current-time response curve after the successive addition
of aliquots of phenol and catechol, respectively. Figure 2.7b and 2.8b show the

corresponding phenol and catechol calibration plots. The response to phenol was linear

(R2 = 0.980) in the range 1 to 25 uM, with an average sensitivity of approximately 410 (:I:

30) nAuM'lcm'z, and it reached saturation at approximately 65 pM. On the other hand,

24

the response to catechol was linear (R2 = 0.982) in the range 1 to 40 uM, with a

sensitivity of 2.5 (i 0.1) uAnM'l cm'z, and it reached saturation at approximately 80 M.

2.4.4 Inhibition of esterase activity

To measure inhibition of the esterase activity, an aliquot of phenyl valerate was
added to the phosphate buffer. After a steady biosensor signal was obtained, a known
quantity of phenylmethylsulfonyl ﬂuoride (PMSF), a non-neuropathic compound
previously shown to inhibit NEST (or NTE) esterase activity, was added to the phosphate
buffer solution, and the resulting drop in current was measured. As shown in Figures 2.9,
2.10 and 2.11 there was no decrease on the addition of 10 [1M PMSF, a 20% (:t 3%)
decrease in response on the addition of 100 uM PMSF and a 70% (d: 4%) decrease on the
addition of 1000 uM PMSF. PMSF inhibition of NEST esterase activity reduces the
amount of phenol and subsequently o-quinone produced. Therefore, less of o-quinone
gets reduced at the electrode surface, leading to a drop in current. A similar dose-
dependent drop in current was observed when a neuropathic OP compound, MIPAFOX
(data not shown), was added to the phosphate buffer. These results suggest that the NEST

biosensor can be used for dose dependent detection of NEST inhibitors.

2.4.5 Storage stability

The sensors storage stability was tested by storing it at 4°C in phosphate buffer,
pH 7.0, over one month, while intermittently measuring the current response to phenyl
valerate standard solution every other day. The results showed that the activity of sensor

remained stable for a week and then reduced gradually, with a half life of 15 days.

25

2.4.6 Immobilization of AChE and BChE

While the development of NEST biosensor was the primary purpose of this study,
we found that the same PLL-Tyr multilayer interface could also be used to immobilize
and measure the activity of other esterases such as AChE and BChE. Figure 2.12
displays a typical current-time response, for a bi-enzyme electrode containing tyrosinase
and AChE, after the successive addition of aliquots of phenyl acetate to the phosphate

buffer. A reduction current proportional to the amount of phenyl acetate was observed.
The response to phenyl acetate was also found to be linear (R2 = 0.989) in the range 0.5

uM to 16 uM and it reached saturation at approximately 40 uM. No signiﬁcant rise in
current was observed, when phenyl valerate instead of phenyl acetate was used as a
substrate on this interface. This result suggests that phenyl acetate is a much better
substrate for acetylcholinesterase than phenyl valerate. Current time response curves
similar to Figures 2.5 and 2.12 were also obtained with bi-enzyme electrodes consisting
of tyrosinase and BChE, with the highest current sensitivity being obtained when phenyl
valerate was used as a substrate. Table 2.1 summarizes the current sensitivities obtained
with three different bi-enzyme interfaces. Control experiments were also done in which
each of the substrate was delivered to a gold electrode containing only PLL-Tyr bilayers.

The current sensitivities obtained in these control experiments were always less than 0.5
nAuM'lcm'z. Collectively, these interfaces can be used for simultaneously detecting the

presence of a variety of compounds which inhibit cholinesterases and also for studying

cholinesterase reaction kinetics.

26

2.4.7 Significance of NEST biosensor

This new biosensor approach to measuring NEST esterase activity and its
inhibition can in principle easily be extended to full length NTE. The approach offers
several advantages over the old two step method. First, it requires only a single step to
measure NEST (or NTE) esterase activity. Because the NEST esterase activity is co-
immobilized with tyrosinase on the sensor interface, the presence of phenyl valerate
triggers sequential reactions that result in an electrical signal. Second, the nanometer-
scale thickness of layers in the sensing interface provides a very short diffusion path
giving a rapid response time (less than 10 seconds). Third, the biosensor is suitable for
continuous, real-time measurements of esterase activity. Fourth, the biosensor is
designed to achieve signal ampliﬁcation via recycle of o-quinone to catechol, thus
increasing the sensitivity of the sensor. Fifth, the biosensor interface is generated by
ﬂexible, layer-by-layer, molecular self-assembly methods that would allow it to be
assembled on electrodes inside microﬂuidic channels, thus enabling the production of
high-density biosensor arrays consisting of various esterases for high-throughput

applications.

This combination of desirable properties makes this interface well suited for
important applications, including studying the kinetic properties of esterases such as
NEST protein, high-throughput screening of compounds for NEST (or NTE) inhibition
and continuous, on-line, environmental monitoring to detect chemical warfare agents that

target NEST (or NTE) and other esterases.

27

2.5 CONCLUSIONS

A biosensor has been developed that allows the activity of NEST to be measured
continuously. The biosensor was fabricated by layer-by-layer assembly approach to co-
immobilize NEST and tyrosinase on a gold electrode. Ellipsometry provided evidence for
the sequential assembly of the multiple layers that make up the interface. Constant
potential amperometry allowed NEST enzyme activity to be measured with a rapid
response time (< 10 s). The biosensor gave dose-dependent response to known non-
neuropathic (PMSF) and neuropathic (Mipafox) NEST inhibitors. The same interface can
also be used to immobilize and measure the activity of other medically relevant esterases

such as AChE and BChE.
2.6 RECOMMENDATIONS FOR FUTURE WORK

It will also be interesting to work on approaches that can help us improve the long-
term stability of the bi-enzyme electrodes. One of the approaches could be dipping the bi-
enzyme electrode in some stabilizing solution such as sucrose or trehalose solution.
Another approach could be using some kind of cross-linking agent such as glutaric
dialdehyde which can cross-link proteins and poly-L-lysine. It will also be equally
interesting to try this approach on other medically relevant esterases such as alkaline

phosphatase. In this case phenyl phosphate may be tried as one of the substrate.

Another recommendation will be to integrate micro-ﬂuidics and semi-conductor
fabrication methods to generate bi-enzyme biosensor electrode arrays. Semi-conductor
fabrication methods can be used to generate gold arrays on a silicon chip. Because all the

interfaces presented in this paper can be fabricated using layer-by—layer assembly from

28

different solutions, microﬂuidics can be used to fabricate different interfaces on different
electrodes. Micro-ﬂuidics in conjunction with potential step voltammetry can then be
used to measure the response of different bi-enzyme electrodes to different analytes.
These electrode arrays may have outstanding potential for food-safety and homeland

security applications.

29

  
 

a, We. a

H3 NH3
Poly-Iysme

Thioctic acid *4»
Gold
electrode

  
 

Tyrosinase —— & é”
N
00C

Figure 2.1: Molecular architecture of the NEST biosensor.

30

 

Poly-lysine!
iTyrosinase (3.5
bilayers)

 

35

 

30 « i “I
25 ‘ § (d)
mi

15 ~ ‘9

Thickness (nm)

 

 

 

O ‘ h F I I T
O 1 2 3 4 5

Number of bilayers

Figure 2.2: Ellipsometric thicknesses after the successive addition of following layers:
thioctic acid (point a), PLL-Tyr ﬁrst bilayer (point b), PLL-Tyr second bilayer (point c),
PLL-Tyr third bilayer (point (1), and PLL and NEST ﬁnal bilayer (point e).

31

 

1.6

1.4 i
mg 1.2 I
Z 1 ‘ (i)
3 0.3-
4.;
C
93 0.6-
L
8 0.4-
0.2 ("I
O F I I I I

 

 

 

-025 -o.2 -o.15 -o.1 -o.05 o 0.05 0.1
Potential (V)

Figure 2.3: Effect of working potential on the response current of the enzyme electrode in
0.1 M phosphate buffer (pH 7.0) with (i) and without (ii) 12 uM phenyl valerate solution,
in 0.1 M phosphate buffer at an applied potential of -0.1 V (vs Ag/AgCl).

32

 

.3
..§

1.2 -
NA

E 1 -
E’

0.8 4
3

E 0.6 4
m
t:

3 0.4 -
O

0.2 —

O l I I

 

 

 

Figure 2.4: Effect of pH on the response current of the electrode, in the presence of 12
uM phenyl valerate solution, in 0.1 M phosphate buffer at an applied potential of -O.1V
(vs Ag/AgCl).

33

1.50

 

_t

N

O
l

.0

co

0
l

.0

0)

o
1

Addition of 4uM
Phenyl valerate

0.30 ~ \

0.00 I r
0 50 100 1 50 200

Time (sec)

(3)

Current (uA/cmz)

 

 

 

 

 

 

 

0 ‘ I j— I l I I

0 5 10 15 20 25 30 35
Concentration (uM)

(b)

Figure 2.5: (a) Current time response of the NEST biosensor to the addition of aliquots of
phenyl valerate to obtain ﬁnal phenyl valerate concentrations as increments of 4 uM, in

0.1 M phosphate buffer, pH 7.0, at an applied potential of -0.1V (vs Ag/AgCl). (b)
Representative example of a calibration plot.

34

 

 

 

 

NA
E 0.30 1
5:
3 0 20 J Addition of 80M
E ' Phenylvalerate
9
5
o 0'10 ‘W
0.00 I I I I I I I I I
O 10 20 30 40 50 50 70 80 90 100

Time (see)

Figure 2.6: Control experiment: Current time response on an electrode containing only
tyrosinase. The electrode was assembled in exactly the same way as NEST biosensor,
except that the ﬁnal NEST layer was not deposited.

35

7.00

 

NA 6.00 r

5.00 -

A/cm

3 4.00 -

3.00 - Addition of4uM
Phenol

Current

2.00 ~

 

 

 

0.00 ‘¥—_ r T r

0 20 40 60 80
Time (sec)

(8)

 

 

 

O I I I I f I

 

 

0 10 20 30 40 50 60
Concentration (uM)

(b)

Figure 2.7: (a) Current time response of the NEST biosensor to the addition of aliquots of
phenol in 0.1 M phosphate buffer (pH 7.0) to obtain ﬁnal phenol concentrations as
increments of 4 uM, at an applied potential of -0.1 V (vs Ag/AgCl). (b) Representative

example of a calibration plot.

36

 

100
90 r
80 -
70 -
60 -
50 -

40 ' Addition of 8uM
30 ‘ Catechol

20 .

10 1* \

O I I 'r
0 40 80 120 160

Time (sec)

(8)

I I I I I l I

Current (pA/cmz)

I

 

 

 

140

120 -
E 100 _
80 -
60 -
40 —
20 —

0 If I I I

0 20 40 60 80 1 00
Concentration (uM)

(b)

 

2)

Current (uA/c

 

 

 

Figure 2.8: (a) Current time response of the NEST biosensor to the addition of aliquots of
catechol in 0.1 M phosphate buffer (pH 7.0) to obtain ﬁnal catechol concentrations as
increments of 8 nM, at an applied potential of -0.1 V (vs Ag/AgCl ). (b) Representative
example of a calibration plot.

37

 

0.90 - Addition of 10'5M
A 0.30 - / PMSF

    

0,20 _ Addition of 8uM
/ Phenyl valerate

 

 

 

 

0.00 a i i
0 50 100 150

Time (sec)
Figure 2.9: Current time response of NEST biosensor to the addition of phenyl valerate in
phosphate buffer (0.1 M, pH 7.0) to obtain a ﬁnal phenyl valerate concentration of 8 uM

followed by the addition of NEST inhibitor PMSF to obtain a ﬁnal PMSF concentration
of 10 pM.

38

 

  
   
 

 

 

 

 

A 3:: 1 - Addition of 104M
N _ PMSF
5 0.70 " 20 % decrease
2 0.60 -
3 0.50 J
g 0.40 4
g 0.30 J Addition of 8uM
O 0.20 L—J / Phenyl valerate
0.10 .
0.00 e a ' ' '
0 20 40 60 80 100

Time (sec)

Figure 2.10: Current time response of NEST biosensor to the addition of phenyl valerate
in phosphate buffer (0.1 M, pH 7.0) to obtain a ﬁnal phenyl valerate concentration of 8
uM followed by the addition of NEST inhibitor PMSF to obtain a ﬁnal PMSF
concentration of 100 uM.

39

 

_ Addition of 10'3 M
090 PMSF

      

7O %
Decrease

Addition of 8uM " L.

Phenyl Valerate

 

 

 

 

0.00 i u i
0 50 100 150

Time (sec)

 

Figure 2.11: Current time response of NEST biosensor to the addition of phenyl valerate
in phosphate buffer (0.1 M , pH 7.0) to obtain a ﬁnal phenyl valerate concentration of 8
uM followed by the addition of NEST inhibitor PMSF to obtain a ﬁnal PMSF
concentration of 1000 uM. ‘

40

 

 

Current (uA)

 

 

 

 

Addition of 8uM
2 Phenyl acetate
1 1 \
O I I I I I j
0 10 20 30 40 50 60 70

Time (sec)

Figure 2.12: Current time response of a bi-enzyme electrode consisting of tyrosinase and
AChE to the addition of aliquots of phenyl acetate to obtain ﬁnal phenyl acetate
concentrations as increments of 8 nM, in 0.1 M phosphate buffer, pH 7.0, at an applied
potential of -0.1 V (vs Ag/AgCl).

41

Table 2-1: Performance of bi-enzyme electrodes containing tyrosinase and different
esterases. Phenyl valerate gave the highest current sensitivity for bi-enzyme electrodes
containing tyrosinase and NEST, and also for bi-enzyme electrodes containing tyrosinase
and Butyrylcholinesterase. Phenyl acetate gave the highest sensitivity for bi-enzyme
electrodes containing tyrosinase and Acetylcholinesterase.

 

 

 

 

 

Enzyme Substrate Average
used sensitivity
(nApM'lcm'z)
NEST Phenyl 87 i 8
valerate
Acetylcholinesterase Phenyl 180 i 25
acetate
Butyrylcholinesterase Phenyl 25 i 10
valerate

 

 

 

 

42

3 HIGHLY SENSITIVE PHENOL SENSOR BASED ON LAYER-BY-LAYER

ASSEMBLY
3.] ABSTRACT

Phenols can be toxic even in low concentrations. The sensing of phenolic compounds is
therefore essential to evaluate the health risk of environmental samples. Interestingly, we
found that nanostructured thin ﬁlms, consisting of bilayers of PLL and tyrosinase on a
gold electrode, can also be used as highly sensitive sensors for detecting phenols. The
ﬁlms were deposited using layer by layer assembly approach and characterized using
ellipsometry, Fourier transform infrared spectroscopy (FTIR), and constant potential
amperometry. The current response of the ﬁlms to catechol and phenol was found to be

linear up to 6 bilayers. The increase in current sensitivity for catechol with each PLL-Tyr
bilayer was approximately (1.18 :t 0.15) pAitM'lcm'2 which is at least 2 orders of

magnitude higher than that reported using other polyelectrolytes and tyrosinase
multilayer systems, suggesting that the composition of polyelectrolytes may inﬂuence

important factors such as enzyme loading and turnover number.

43

3.2 INTRODUCTION

Phenols are present in wastewaters and exhaust gases of many industries, including oil
reﬁneries, coke plants, resins and plastics, dyes and chemicals, and textiles (Kararn and
Nicell 1997). Phenols can be toxic even in low concentrations. They generally get
adsorbed through skin and cause severe systemic reactions. The sensing of phenolic
compounds is therefore essential to evaluate the health risk of environmental samples.
Colorimetry, liquid and gas chromatography, and capillary electrophoresis are commonly
used techniques for phenol determination (T ownshend 1995; Vogel 1998). However,
these techniques are time consuming, require special expertise on part of users, and have
low sensitivity. Phenols can also be detected electrochemically by their direct oxidation
at high overpotentials. However, this approach suffers from poor selectivity, due to

interference by other electroactive compounds such as ascorbic acid.

To overcome these limitations, numerous biosensors based on immobilized
tyrosinase have been proposed (Desprez and Labbe 1996; Ducey and Meyerhoff 1998;
Kulkami, Karve et al. 1998; Dall‘Orto, Danilowicz et al. 1999; Wang, Lu et al. 2000;
Ahn, Beaudette et al. 2001; Freire, Duran et al. 2002; Stanca, Popescu et al. 2003; Rajesh,
Takashima et al. 2004; Solna, Sapelnikova et al. 2005; Solna and Skladal 2005). As
mentioned in Chapter 2, tyrosinase is a copper containing enzyme that catalyzes the
ortho-hydroxylation of monophenols (cresolase or monophenolase activity) to 0-

diphenols by oxygen and also catalyzes the oxidation of o-diphenols to o-quinones.

The underlying principle of most tyrosinase based biosensors is that the reaction

product, o-quinone, is electrochemically active, and gets reduced at the electrode surface

44

at low potential, resulting in the generation of current. Moreover, the electroreduction of
o-quinone produces catechol (o-diphenol) which again gets converted to o-quinone by
tyrosinase. This recycling results in signal ampliﬁcation; as a result higher sensitivity can

be achieved in these sensors in comparison to conventional designs.

The current sensor systems are limited by the amount of protein immobilized on
the electrode surface. The study of such electrodes has shown that the sensitivity of the
sensor is proportional to the surface density of protein on the electrode surface (F orzani,
Solis et al. 2000). Therefore, an increase in protein loading on the surface could promise
a signiﬁcant improvement in sensitivity. LBL assembly is a useﬁrl technique for
increasing the protein content on the electrode surface. In this technique, a substrate is
alternatively dipped in a polycationic and polyanionic solution leading to the formation of
multilayered structures stabilized by electrostatic interactions. This technique has been
used in the past for making alternating layers of tyrosinase, and polyelectrolytes such as
poly(allylarnine hydrochloride) (PAH) (Forzani, Solis et al. 2000; Coche-Guerente,

Labbe et al. 2001; Forzani, Teijelo et al. 2003). However, the highest sensitivity obtained
in these systems was approximately 0.005 itAjuM'lcm'2 for each tyrosinase and PAH

bilayers.

The sensitivity of the sensors can further be increased by varying the composition
of polyelectrolyte solution. As some polyelectrolyte composition may lead to enhanced
protein loading or it may provide protein with a more native environment where it can
exhibit higher turnover number. In this Chapter, the fabrication of multilayer ﬁlms

consisting of a polyelectrolyte poly-L-lysine (PLL) and tyrosinase is presented. The

45

resulting multilayer assemblies were characterized using ellipsometry, FTIR and
ellipsometry. The current response of the ﬁlms to catechol and phenol as a ﬁrnction of
number of PLL-Tyr bilayers was studied up to 8 bilayers. The results of this study are
unique in that the resulting PLL-Tyr multilayer ﬁhns gave current sensitivity which is at
least 2 orders of magnitude greater than those reported using other polyelectrolytes and

tyrosinase multilayer systems (Forzani, Solis et a1. 2000).
3.3 EXPERIMENTAL SECTION

3.3.1 Preparation of poly-L-lysine (PLL), and tyrosinase (Tyr) multilayer

assemblies

Initially, the gold electrode was derivatized by dipping it in 5 mM solution of thioctic
acid in ethanol for 30 min. The electrode was washed with ethanol, dried under nitrogen
and dipped in PLL solution for 45 min. The PLL solution was prepared by dissolving 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 0.2 mg/mL solution of Tyr for l h. The last two steps
were repeated to create multiple PLL-Tyr bilayers. After the formation of PLL—Tyr
bilayers, the electrodes were dipped in phosphate buffer (0.1 M, pH 7.0) for testing. All

the experiments were done at room temperature.

3.3.2 Ellipsometry

Ellipsometric measurements were obtained with rotating analyzer ellipsometer (model
M-44; J.A. Woollan Co. Inc., Lincoln, NE) using WVASE32 software. The thickness
values for ﬁlms were determined using 44 wavelengths between 414.0 and 736.1 nm. The

angle of incidence was 75° for all experiments. Refractive indices of ﬁlms containing

46

PLL and Tyr were assumed to be u = 1.5, k=0. These optical constants compare well with

those determined for 4 PLL-Tyr bilayer ﬁlms using ellipsometry.

3.3.3 Reflectance Fourier transform infrared spectroscopy

Reﬂectance FTIR spectroscopy was performed using a Nicolet Magna-IR 560
spectrometer containing a PIKE grazing angle (80°) attachment. The spectra were

collected with 256 scans using a MCT detector.

3.3.4 Potential step amperometry

The electrodes (sensors) were maintained at a potential of -100 mV (vs Ag/AgCl
reference electrode) using a BAS CV-50W electrochemical analyzer. To measure the
response, a small aliquot of phenol or catechol was added to a beaker containing the

sensor immersed in a stirred buffer solution, and the current was monitored.
3.4 RESULTS AND DISCUSSIONS

3.4.1 Ellipsometry and FTIR

Ellipsometry was used to provide proof for the assembly of the multilayers.
Figure 3.1 shows the ellipsometric thicknesses in air for different number of PLL-Tyr
bilayers. The thickness increase after the addition of each bilayer was 7.2 i 0.4 nm, with
the exception of the initial bilayer, which was 9.3 :t 0.4 nm thick. On the other hand, the
reported thickness increase for PAH-Tyr bilayers was approximately 2.3 d: 0.3 nm

(Forzani, Teijelo et al. 2003).

Figure 3.2 shows the FTIR spectra of PLL-Tyr ﬁlms containing up to 6 bilayers.
The amide absorbances at 3297 cm.1 (amide A, due to N-H asymmetric and symmetric

47

stretching), 1665 cm'1 (amide 1, due to the stretching mode of the carbonyl group coupled

to the amide linkage), and 1540 cm'I (amide II, due to the N-H in-plane bending of the

polypeptide chains), dominates the spectrum of PLL-Tyrﬁhns. These peaks compare

well with the published FTIR spectrum of tyrosinase in solid state (KBr pellet), where

amide A peak was observed at 3295 cm], amide I at 1653 cm'l, and amide II at 1541 cm'

'. Negative shifts of amide band frequencies in relation to the native state are generally

found in denatured proteins, because of mass loss due to decrease in the number of
hydrogen bonds (Forzani, Teijelo et al. 2003). Although most of the tyrosinase amide
bands appeared at the same frequency as that in the solid-state, the positive shifts of
amide I band could be associated with a more structured protein. FT IR also conﬁrmed the
step-wise growth of PLL-Tyr ﬁlms, as the absorbance values approximately increased

linearly (Figure 3.1) with the number of bilayers, conﬁrming the ellipsometric data.

3.4.2 Dependence of current response on working potential and pH

The various experimental parameters that can affect the amperometric
determination of catechol (such as pH and applied potential) were optimized using the
procedure outlined in Chapter 2. The optimum current sensitivity (highest signal to

background ratio) was obtained at -0.1 V and pH = 7.

3.4.3 Amperometric response to catechol and phenol

Figure 3.3a displays a typical current-time response under the optimal
experimental conditions for a gold electrode containing 6 PLL-Tyr bilayers after the
successive addition of aliquots of catechol. A well deﬁned reduction current, proportional

to the amount of catechol was observed. The response time of the electrode was less then

48

20 5, due to the nano-scale thickness of the interface. The response to catechol was linear
(R2 = 0.989) in the range 5 nM to 40 itM and it reached saturation at approximately 85

uM (Figure 3.3b). The limit of detection was 5 nM at a signal-to-noise ratio of 3. The

reproducibility of the sensor was investigated at a catechol concentration of 8 uM; the
mean current was approximately 57 itAcm'z, with a standard deviation of approximately
8%. Similarly, the response of this sensor (containing 6 PLL-Tyr bilayers) to phenol was
also studied. The response to phenol was linear (R2 = 0.990) in the range 1 to 25 uM,
with an average sensitivity of approximately 1.2 (d: 0.15) uAuM'lcm'z, and it reached

saturation at approximately 70 11M.

3.4.4 Effect of number of PLL-Tyr bilayers

Figure 3.4 shows the current sensitivity to catechol as a function of PLL-Tyr
bilayers. The increase in current sensitivity was approximately linear up to 6 bilayers and

it reached saturation at around 8 bilayers. The increase in current sensitivity in the linear
region with each bilayer (with the exception of ﬁrst bilayer where it was 1.4 d: 0.1 uAuM'
lcm'z) was approximately 1.18 :I: 0.15 uAuM’lcm'z. The current sensitivity
corresponding to 6 PLL-Tyr bilayers was 7.1 d: 0.55 uAuM'lcm'z. Similarly, the increase

in sensitivity to phenol was also found to be linear and equal to 0.2 i 0.02 uApM'lcm'2

for each bilayer.

3.4.5 Cyclic voltammetry

While constant potential amperometry was the primary sensing technique used in

this study, other sensing techniques such as cyclic voltammetry were also tried. Figure

49

3.5a shows the cyclic voltammograms of the enzyme electrodes (containing 2 PLL-Tyr
bilayers) at different concentrations of catechol. The electrocatalytic cathodic current
indicates the effective reduction of o-quinone. For all catechol concentrations, the
reduction current begins at 0.18 V and then reaches saturation at higher negative
potentials. As expected, the reductive current was found to increase linearly with catechol
concentration and then it levels off at a catechol concentration of approximately 85 uM

(Figure 3.5b).

3.4.6 Discussions

The ellipsometric thickness increase after the addition of each PLL-Tyr bilayer
was approx. 7.2 nm which is about three times higher than that reported for PAH-Tyr
bilayers (2.3 nm). The current sensitivity obtained in the PLL-Tyr system for each

bilayer was approximately 1.18 :t 0.15 uAuM'lcm'z, which is more than 2 orders of

magnitude higher than that reported for PAH-Tyr bilayers (Forzani, Solis et al. 2000).
Using a theoretical analysis, Forzani et al (Forzani, Solis et al. 2000) have shown that the

current sensitivity in layered tyrosinase-polyelectrolyte system is directly proportional to

the surface coverage (i.e mol/cmz) of the enzyme and its turnover number. If the enzyme

density (i.e mol/cm3) is assumed to be the same in both PAH-Tyr and PLL-Tyr system,

then the ellipsometric data suggests that the enzyme surface coverage in PLL-Tyr
bilayers is approximately three times higher than that in PAH-Tyr bilayers. However, this
three times increase in enzyme surface coverage alone cannot explain more than 2 orders
of magnitude increase in current sensitivity observed with PLL-Tyr bilayers, and it

suggests that other factors such as higher turnover number or different enzyme density

50

may be responsible for increased sensitivity. It is possible that PLL-Tyr bilayers may
provide tyrosinase with a more natural environment where it may be in a more native
conformation and thus exhibits higher turnover number, or PLL-Tyr bilayers may have a

different enzyme density than PAH-Tyr bilayers.
3.5 CONCLUSIONS

An approach to develop highly sensitive phenol sensors is presented. Layer by
layer assembly approach was used to deposit PLL-Tyr bilayers on a thioctic acid
derivatized gold electrode. Ellipsometry and FTIR provided strong evidence of bilayer
assembly. The current sensitivity to catechol and phenol was found to increase linearly up

to 6 PLL-Tyr bilayers. The increase in sensitivity corresponding to catechol and phenol
was 1.18 pApM'lcm'2 and 0.2 pAuM'lcm'z, respectively. This value is at least two orders
of magnitude greater than that reported using other polyelectrolyte-tyrosinase systems,

suggesting that the composition of polyelectrolytes may inﬂuence important factors such

as enzyme loading, density, and turnover number.
3.6 RECOMMENDATIONS FOR FUTURE WORK

Because the entire interface used in this study can be assembled using layer by
layer assembly from different solutions, it may be possible to assemble this interface
inside microﬂuidics channels, and study the effect of different parameters such as shear

rate on the stability on the sensor.

Nanosized functional particles represent an area of great research interest. In the

ﬁeld of biosensors and biocatalysis, nanosized materials offer the potential for extremely

51

high surface area to volume ratio, thus allowing immobilization of large amounts of
biomolecules per unit projected area. Carbon nanoparticles, such as fullerenes, carbon
nanotubes (CNTs), and exfoliated graphite nanoplatelets (xGnPs), provide high
conductivity and surface area without excessive diffusional resistance. These
nanoparticles are widely used to prepare solid electrode systems and supporting
substrates in electrochemical biosensors, due to their high chemical inertness and wide
range of working potentials with low electrical resistance. Therefore, it will be interesting
to study the effect of addition of these nanoparticles on the sensitivity of PLL-Tyr
bilayers. It is possible that addition of these nanoparticles may lead to higher sensitivity
either due to lower diffusional resistance, or increased enzyme loading and higher

turnover number.

52

 

800 0.18

Absorbance at 1665 cm'1

 

 

700 - — 0.16
r 600 — " 0-14
< _
m 500 - 0.12
8 L 0.1
c 400 c
E 300 "i ~ 0.06
200 - r— 0.04
100 — — 0.02
O I T I l I I I I 0

 

0 1 2 3 4 5 6 7 8 9
Number of Bilayers

Figure 3.1: Ellipsometric thickness and FTIR absorbances at 1665 cm'1 for PLL-Tyr
multilayered ﬁlms.

53

 

 

33 03-.

 

 

38 2a-.

 

9.5

i 0.0m

I
.0
o
m

. obb

 

 

wmoo

wNOO

mmoo Nsoo nooo
<<m<o==3¢oqm 33-;

.500

_ 0.00
ANOO

Absorbance

Figure 3.2: FTIR spectra of PLL-Tyr ﬁlms composed of up to 6 bilayers.

54

300

 

A 250‘

N

O

O
1

150 1

100 _ Addition of 8 pM
catechol

\

I I I I I I

10 20 30 40 50 60 70
Time(sec)

(a)

Current (uA/cm2

 

 

01
O O
I

 

O

 

Current (uA/cmz)
N
O
O

 

 

0 l l I l
0 20 40 60 80 100
Concentration (uM)

(b)

 

Figure 3.3 (a) Current time response of the gold electrode containing 6 PLL-Tyr bilayers
to the addition of aliquots of 8 uM catechol in 0.1 M phosphate buffer, pH 7.0 at an
applied potential of -0.1 V (vs Ag/AgCl). (b) Calibration plot.

55

 

 

 

 

at" 9'
E 8'
'70 7‘
21 6*
<5. 5‘
T: 4'
5 2‘
0 1-
0 ﬂ i i i
O 2 4 6 8 10

Number of Bilayers

Figure 3.4: Effect of number of PLL-Tyr bilayers on the current sensitivity of the enzyme
electrode to catechol.

56

 

30

(iii)

~25
L20

 

En’
Current (pA/cmz)

 

 

 

 

 

 

 

- 10
~ 5
- 0
I I I '5
0.25 0.15 0.05 -0.05 -0.15
E (volts)
(3)
120
(c 100 7
E
0 80 A
2
§ 60 -
E
g 40 —
:3
0 20 -
O r l i I
0 20 40 60 80 100
Concentration (,uM)
(b)

Figure 3.5: (a) Cyclic voltammograms of the enzyme electrode containing 2 PLL-Tyr
bilayers, in the presence of different concentrations of catechol: (i) 3.5 uM, (ii) 7 MM,
and (iii) 10.5 nM. (b) Representative example of a calibration plot. All the measurements

were performed in 0.1 M phosphate buffer, pH 7 .0, and scan rate of 100 mVs'l.

57

4 THEORETICAL AND EXPERIMENTAL STUDY OF BI-ENZYME

ELECTRODES WITH SUBSTRATE RECYCLING
4.1 ABSTRACT

The range of analytes for which biosensors can be deve10ped can be increased
substantially by coupling multiple enzyme activities into reaction pathways. The
sensitivity of the biosensors can also be increased dramatically (or ampliﬁed) by
incorporating a substrate recycling scheme. For low concentrations of substrate, this
chapter presents a theoretical model for a bi-enzyme biosensor that achieves signal
ampliﬁcation via substrate recycling. The bi-enzyme electrode was fabricated by co-
immobilizing two enzymes: Tyr and NEST. The model was validated by assembling the
bi-enzyme interface on a rotating disk electrode and measuring the biosensor’s response
to phenyl valerate under varying rotating speeds. The model can help quantify the
inﬂuence of mass transport, partition coefﬁcients, enzyme, and electron-transfer kinetics
on the metrological characteristics of the bi-enzyme electrode. This information can help

optimize the performance of biosensors that use substrate recycling.
4.2 INTRODUCTION

Biosensors based on enzymes are becoming popular in various ﬁelds of analytical
chemistry because of their high selectivity, speciﬁcity and low cost for mass production.
In particular, enzyme electrodes have been developed for many applications such as
electrochemical immunoassays, monitoring of metabolites, and detection of water and air
pollutants (Uchiyama, Hasebe et al. 1993; Rajesh, Takashima et al. 2004; Solna and

Skladal 2005).

58

The range of analytes for which reporting interfaces (biosensors) can be
developed can be increased substantially by coupling multiple enzyme activities into
reaction pathways. For example, NEST biosensor (described in Chapter 2), couples
NEST esterase activity with two tyrosinase oxidation activities: cresolase activity and

catecholase activity, resulting in electrochemical detection of phenyl valerate.

The sensitivity of the reporting interfaces can be increased substantially by
incorporating a substrate recycling scheme. In these devices, an analyte that is oxidized
or reduced at the electrode is electrochemically recycled so that it can interact with the
electrode more than once, thus amplifying the electrode’s signal. The shuttle analyte can
either be the analyte of interest (the enzyme’s substrate) or a compound that is produced
by an additional enzyme reaction. Ampliﬁcation by substrate recycling has been achieved
by several mechanisms. One uses a chemical reaction to regenerate the enzyme’s
substrate (Uchiyama, Hasebe et al. 1993). Another uses a second enzymatic reaction,
which converts the enzyme’s product back to the substrate. In these two mechanisms,
transduction is achieved by monitoring the production of a coproduct or the consumption
of a co-substrate (Bauer, Eremenko et al. 1996). A third mechanism is based on
electrochemically recycling the reaction product back to the substrate (Coche-Guerente,
Labbe et al. 1999). In the latter case, substrate recycling is achieved by the amperometric

transduction step.

Mathematical models can be used to analyze the simultaneous mass transfer and
reaction steps that occur in biosensors that use substrate recycling in order to improve
their metrological characteristics. Such a theoretical analysis can help elucidate the rate-

limiting step(s) and evaluate the performance characteristics of substrate-recycling

59

biosensors. In work relevant to the NEST biosensor, Gorton et al.(Burestedt, Narvaez et
a1. 1996) theoretically considered and experimentally demonstrated the various rate
limiting steps for tyrosinase enzyme modiﬁed electrodes with substrate recycling. This
model assumed that diffusional resistance within the enzyme layer was negligible.
Coche-Guérente et al., in a series of papers, presented a theoretical model for an
immobilized enzyme layer with electrochemical substrate-recycling scheme (Coche-
Guerente, Labbe et al. 1999; Coche-Guerente, Labbe et al. 2001). This model, which was
applied to tyrosinase modiﬁed electrodes, includes the inﬂuence of diffusion, partition
coefﬁcient, enzyme, and electron-transfer kinetics. However, these theoretical analyses
have been limited to single enzyme electrodes and are thus not suitable for bi-enzyme

electrodes such as the NEST biosensor.

In this Chapter, a theoretical model is presented for a bi-enzyme rotating disk
electrode consisting of NEST and tyrosinase. The molecular architecture of the bi-
enzyme electrode is shown in Figure 4.1. The NEST protein converts phenyl valerate to
phenol, which is converted to o-quinone by tyrosinase. The o-quinone is
electrochemically reduced to catechol at the electrode’s surface, resulting in current. A
portion of the catechol produced is then converted to o-quinone by tyrosinase. Catechol
thus serves as shuttle analyte that can undergo successive cycles of enzymatic oxidation-
electrochemical reduction (substrate recycling), resulting in an ampliﬁcation of

biosensor’s response.

The NEST biosensor model includes the inﬂuence of the mass transport,
permeation through the enzyme layers and enzyme kinetics. This model is expressed in

dimensionless form to minimize the number of constants that must be evaluated. The

60

biosensor was assembled on a rotating disk electrode, and the biosensor’s performance
was measured at a variety of rotational velocities and substrate concentrations to evaluate

the constants and validate the model.

4.3 EXPERIMENTAL SECTION

4.3.1 Materials

Thioctic acid, PLL (molecular weight ~ 15,000), tyrosinase, sodium phosphate
(monobasic and dibasic), sodium chloride were obtained from Sigma (St. Louis, MO).
Ultrapure water was supplied by a Nanopure-UV four-stage puriﬁer (Barnstead
International, Dubuque, IA); the puriﬁer was equipped with a UV source and a ﬁnal 0.2

pm ﬁlter.

4.3.2 Preparation of gold electrode

The molecular architecture of the biosensor is shown schematically in Figure 4.1. Gold
rotating disk electrodes (Metrohm Limited. Herisau, Switzerland) were polished with
alumina powder and dipped in 5 mM solution of thioctic acid in ethanol for 30 min. The
electrodes were washed with ethanol, dried under nitrogen and dipped in PLL solution for
45 min. The PLL solution was prepared by adding 12 mg of poly-lysine in 50 ml of 20
mM phosphate buffer (pH 8.5). The electrodes were then rinsed with water and dipped in
equimolar solution of tyrosinase and NEST in 0.1 M phosphate buffer for 45 min. The
last two steps were repeated 3 times to create 3 bilayers of PLL and tyrosinase/NEST.
The electrodes were then washed with water, dried under nitrogen and dipped in
phosphate buffer (0.1 M, pH 7.0) for testing. All the experiments were done at room

temperature.

61

4.3.3 Chronoamperometry and other measurements

The bi—enzyme rotating disk electrodes containing NEST and tyrosinase were maintained
at a potential of -100 mV (vs Ag/AgCl reference electrode) using a CHI 660B
electrochemical analyzer (CH instruments, Austin, TX). To perform electrochemical

measurements, the bi-enzyme rotating disk electrodes (area = 0.07 cmz) were dipped in

0.1 M phosphate buffer (pH 7.0), and small aliquots of phenyl valerate, phenol or
catechol solution were added. The steady state current was then measured at different
rotating speeds. Ellipsometric measurements were done using a WVASE 32 (J .A.
Woollam Co. Inc., Lincoln, NE) ellipsometer. The angle of incidence was 75° for all

experiments. The refractive indices of ﬁlms were assumed to be n =1 .5, k=0.
4.4 THEORETICAL MODEL

4.4.1 Enzyme kinetics

Tyrosinase is a binuclear copper containing enzyme that catalyzes the oxidation of
catechols to o-quinone (catecholase activity). For catechol as the substrate, it has been
reported (Coche-Guerente, Labbe et al. 1999) that the overall rate (v) in solution can be

described by Michaelis-Menten formalism (Equation 4.1) with an apparent Michaelis

constant K app = 220 uM.

v = kcatlEtllSl
K3,” + S

 

(4.1)

62

where E t} S and kcat represent the enzyme concentration, catechol concentration and

turnover number, respectively.

Since, the experiments conducted in the present study involve substrate concentrations

around two orders of magnitude less than K gpp , the kinetics can be assumed to be ﬁrst

order relative to the substrate concentration.

Besides its catecholase activity, tyrosinase is also able to catalyze ortho-
hydroxylation (monophenolase activity) of monophenols to o-diphenols (catechols) that,
in turn, are oxidized to corresponding o-quinones (catecholase activity). Because the
hydroxylation activity of tyrosinase is expressed in conjunction with oxidation of o-
diphenol to its o-quinone, some authors have deﬁned monophenolase activity as the
complete conversion of monophenols to o-quinone. Indeed, the hydroxylation step
proceeds and has been shown to be much slower than the oxidation of o-diphenol to o-
quinone and is therefore considered to be the rate limiting step (Coche-Guerente, Labbe
et a1. 2001). The enzymatic oxidation of phenol to o-quinone has also been shown to

follow Michaelis-Menten formalism (Coche-Guerente, Labbe et al. 2001), with an

apparent K 3117‘” of 250 uM in homogeneous solution.

Similarly, the esterase activity of NTE (or NEST) can convert phenyl valerate to

phenol, and this reaction also follows Michaelis-Menten formalism, with an apparent

K razpp of approximately 1.4 mM in homogeneous solution.

63

4.4.2 Assumptions

Figure 4.2 is a schematic representation of a bi-enzyme rotating disk electrode, modiﬁed
with an enzyme layer that contains NEST and tyrosinase, and has a thickness L. The

sequential steps that lead to an electrochemical signal in the presence of phenyl valerate

substrate S 1 are as follows:

(1) Mass transfer of phenyl valerate (S 1), phenol (S 2), catechol (S3), and quinone (Q4)
through a stagnant ﬁlm between the bulk and the enzyme layer. For rotating disk
electrode, this ﬁlm has a thickness 6 = 1.61Dé/301/6af-1/2 (Coche-Guerente,
Labbe et al. 1999), where De, 0 anda) represent the diffusion coefﬁcient, kinematic
viscosity and rotation speed, respectively. For simplicity, the diffusion coefﬁcients of
S], S 2, S3 and Q4 in the bulk aqueous phase and stagnant ﬁlm were assumed to be
identical (De).

(2) Partitioning of S 1, S 2, S3 and Q4 from the stagnant film into the enzyme layer. The

kinetics of partitioning were assumed to be rapid so that the interfacial concentrations

in the ﬁlm and enzyme layers remained at equilibrium. The partition coefﬁcients (kp)

were assumed to be identical for S1, S 2, S3 and Q4. Thus, yielding the following

equilibrium expressions.

[SIIL- = kaSIIL+ (4.2)

[S2IL- = kaSzlL+ (4-3)

64

[S3IL- = kp[S3]L+ (4.4)
[Q4IL-= kaQ4IL+ (4-5)

(3) Diffusion of S1, 52, S3 and Q4 within the enzymatic layer of thickness L. The

enzymatic layer was assumed to behave like a semi-permeable membrane. The model

was simpliﬁed by assuming an identical diffusion coefﬁcient (Df) for S1, S 2, S3 and

Q4-

(4) At steady state conditions, all the substrates S1, S 2, S3 and Q4 are present in the

enzymatic layer, although only S1 is present in the bulk solution. Inside the enzyme

layer, for low concentrations of phenyl valerate, the conversion of phenyl valerate

(S 1) to phenol can be given by the following ﬁrst order equation:

v1 = k1IE1IISli (4.6)

K1

where E1 denotes the total concentration of active NEST, and K1 is the apparent Km

value. The rate of conversion of phenol to o-quinone and catechol to o-quinone has been
shown to be given by similar ﬁrst order expressions (Coche-Guerente, Labbe et al. 1999)

(Coche-Guerente, Labbe et al. 2001):

_ k2[52][52]

(4.7)
2

65

k3IE3IIS3]

(4.8)
3

Where, E 2 and E3 represent the concentrations of monophenolase and catecholase active
sites, respectively.

(5) Electrode potential was assumed to be sufﬁciently negative so that the
electrochemical reduction step is not rate limiting.

4.4.3 Model equations

Using the same approach as the one already reported in a theoretical treatment of

biosensors (Bartlett and Whitaker 1987; Coche-Guerente, Labbe et al. 1999; Coche-

Guerente, Labbe et al. 2001), the equations describing the concentrations of S1, S 2, S3

and Q4 at steady state are as follows:

 

 

 

2
a s1 _ s1 _ (4.9)
2
a $2 _ $2 + S1 :0 (4.10)
525 s
3 _ 3 :0 (4.11)

66

2
a S S
Q4+2+3

 

6x2 x22 132

(4.12)

where x is the distance from electrode surface. As deﬁned below, 7&1, 7&2 and 7&3

represent the reaction lengths (Coche-Guerente, Labbe et al. 1999) related to phenyl

valerate (S 1), phenol (S 2), and catechol (S3):

1 1
D K — D K —
f 1)2 7"2=( f 2)2
klEl k2Ez

 

 

71.1=(

4.4.4 Boundary conditions

 

)2 (4.13)

To determine concentration proﬁles within the enzyme layer, equations 4.9-4.12 were

solved with the following boundary conditions:

1) Applied potential is sufﬁciently negative that

[Q4]x=0 = 0 (4.14)
2) Since phenyl valerate and phenol are not electro-active
[65‘] -o [532] —0 (415)
6x x=0— ’ 6x x=0" '
3) Only phenyl valerate is present in the bulk
[Sl]x = 00 = Sl(°0)a [S2]x = 00 = 0. [S3Ix = 00: 0’ [Q4]x = 00 = 0 (4-16)

67

4) At steady state, ﬂux of phenyl valerate, phenol, catechol and quinone across the ﬁlm

equals that entering the enzyme layer.

 

D is; =De[kS(oo)—[S] ]
f ax x=L kpé‘ p1 Ix=L
as2 __De
Bria—L ' —k 51.921sz

=L p

04%] Linda
6x sz kpd

 

 

6Q4 __ De
Dfi ax jx=L _ kp6[Q4]x=L

5) From law of conservation of mass, for any x inside the enzyme layer:

[Q4] +[51] + [52] +153] = kpSl (co)

6S (3Q
:- "FDf[-E'x;]x=0= ”FDf[-—i]x:0

6) Current density = J =
6):

L
A

68

(4.17)

(4.18)

(4.19)

(4.20)

(4.21)

(4.22)

Solution of equations 4.9 to 4.12 gave the following analytical expressions for

concentration proﬁles of S i, S 2, S3 and Q4. The following expressions were also

conﬁrmed by Matlab and Mathematica.

 

 

 

 

 

 

 

 

 

k S (00)
S1: 1’ 1 cosh(i) (4.23)
srnh 61 + cosh 61
me
ngl 'nht9 he
51 +005
kpSl(oo) 4912 x me 1 1 x
52 = ( )[cosh(—)—( )cosh —]
Pm‘91 . e —02 41 Pm92 .
srnh 191 +cosh I91 2 l ————srnh 02 +cosh 192
me me
(4.24)
2
S —k S(oo)[1— 1 - 1 ( 61 )
3 p 1 ngl sinh6 + ht9 ngl 11119 no 92-92
1 COS 1 SI 1+COS 1 2 1
m m
e e
P 19
m 3 sinh I93 +cosh I93

 

P (9
m 1 sinh 191 + cosh I91
sinh i] (4.25)

 

 

 

 

 

me x e
x(l— )]x[cosh——
Pm92 - 3 Pm63 - 3
srnh 6 + cosh t9 coshH + srnh t9
2 2 3 3
me me

Q4= k pS160) - S1 - Sz — S3 (4.26)

69

- e _ 0" ,0 ,9
where ”‘18—? Pm ——L—'— l— XI’ 2: KLZ, 3: x3

me is the mass transfer coefﬁcient across the stagnant ﬁlm, and Pm denotes the

P
permeability inside the enzyme layer. The ratio J"— which compares the mass transfer in
m

e
enzyme layer to that in bulk is also known as the Sherwood number. The dimensionless

parameters 61, 92, and 93, also known as Thiele modulus, compare the enzymatic

reaction rates of phenyl valerate, phenol and catechol substrates, respectively, with their

diffusion in the enzymatic layer of thickness L.

The cathodic current sensitivity (SC pv) of the electrode toward phenyl valerate

substrate can then be derived from Equations 4.25 and 4.22:

 

 

 

 

 

 

 

2
Jp 1 1 6
52,, =—S——’(’v=m-)2FP63[1 P_m91 —P_m_61 (212)
—sinh61 +cosh6l —sinh61 +cosh61 62 01
m m
e e
Pm61 . P :63
smh 61 + cosh 61 sinh 63 + cosh 63
m
xa- P g M” Pm; ] (4.27)
m 2 sinh 62 +cosh 62 3 cosh 63 +sinh 63
me me

70

where SI (00) denotes the bulk phenyl valerate concentration. This equation demonstrates
that the bio-electrode response is affected by the rate of mass transport in the stagnant

ﬁlm (me) and the enzyme layer (Pm), as well as by the enzymatic activities toward

phenyl valerate ((91), phenol (6’2) and catechol (03). Equation 4.27 was derived assuming

that only phenyl valerate is present in the bulk. However, if only phenol is present in the
bulk, then it has been shown that the following relation can be derived (Coche-Guerente,

Labbe et a1. 2001):

Pm .
-— 63 s1nh 03 + cosh 63

 

 

J m
SC =_£L=2Fpm63(1— 1 )X( e )
ph S (00) P
2 —m6?2 sinh 62 +cosh 62 —"—’63 cosh 03 +sinh 63
me me

(4.28)

where S; h and 52 (oo) denote the phenol sensitivity and bulk concentration, respectively.

Similarly, if only catechol is present in the bulk, then the following relations can be

derived (Coche-Guerente, Labbe et al. 1999):

 

 

 

 

P 6
C 1— m 3 sinh63 —cosh63
Sc — J“ - —2FP me
C‘ _ — m P sinhB
S3 (00) —”icosh 93 + 3 (4'29)
me 93

71

a

 

 

J 1
S“ = ————‘” = 2FP
C’ m P sinh6
S3 ((1)) _"_1_ cosh 63 + 3 (4.30)
me 3

where S St denotes the sensitivity of the electrode in the presence of catechol at an applied

potential of -O.1 V (where major contribution to current comes from the reduction of o-

quinone), and S gt denotes the sensitivity at an applied potential of 0.5 V (where major
contribution to current comes from oxidation of catechol which couldn’t be converted to

o-quinone by tyrosinase). For rotating disk electrodes we can assume me to be given by

the following Levich equation (Coche—Guerente, Labbe et al. 1999):

2/3
De

 

m = v
e _

1.613V1/6a) “2 (4.31)

On rearranging equations 4.29 and 4.30 and substituting the value of me from equation

4.31, the following relations can be obtained. The following relations were used by us

(data shown in the next section) to validate the model and also estimate some parameters.

 

 

 

C 1
Sct —2/3 1/6 - "5
S“ =(—1+cosh63)+(1.613PmDe v 63s1nh6’3)w (4.32)
ct
1
. —2/31/6 __
s1nh19 v cosh6
_1_ =( 1 3)+(1.613 e 3)w 2 (4.33)
5a: 2FPm 93 2F
C

72

4.4.5 Validation of the model

The model was validated using NEST and tyrosinase containing bi-enzyme electrodes.

Four electrodes (A, B, C and D) were used, and the steady state cathodic and anodic

current sensitivities SE, and SS , were measured at two different catechol concentrations

(3 and 5 11M), under varying electrode rotation speeds. Similarly the cathodic current

sensitivities, S; h and ng, were also measured at different phenol (3 and 5 11M) and

C

phenyl valerate concentrations (3 and 5 11M). While, the cathodic sensitivity, S ct ,

remained practically constant with rotation rate, S; h and ng , decreased with rotation

rate.
C
S t

c
Figure 4.3 shows reciprocal plots of Sec-1t and Tversus the square root of
S

ct

rotation rate. Both plots, as predicted by the model (equations 4.32 and 4.33), showed
linearity with correlation coefﬁcients (R2) greater than 0.98, suggesting that the model

reasonably represent the electro-enzymatic processes occurring in the presence of

catechol. From the slopes and intercepts of the ﬁtted lines in Figures 4.3a and 4.3b,

Equations 4.32 and 4.33 can be used to determine63= 1.81 (:1: 0.2), Pm = 0.0091 (3;

0.0015) cm/s and De = 2.2 (:1: 1.2 x 10'5) cmZ/s. Alternatively, De can also be determined
separately using Levich plot on a bare gold electrode. The obtained value of De compares

well with published values of De for catechol (Coche—Guerente, Labbe et al. 1999).

73

Figure 4.4 shows the current sensitivity to phenol as a function of electrode
rotation speed. The sensitivity of the electrode was found to decrease with increasing

rotation rates. In principle, Equation 4.28 could be ﬁt to the data shown in Figure 4.4 and

determine 62 , Pm and 63. However, the rotating disk electrode system used by us can

give data points at only six rotation speeds. To get more reliable values of the parameters,

the values of Pm and 63 obtained with catechol present in the bulk solution were used,

and Equation 4.28 was ﬁt to the data in Figure 4.4 to give a best ﬁt value of (92 = 0.23 (i

0.05).

Figure 4.5 shows the current sensitivity to phenyl valerate as a function of
electrode rotation speed. Like phenol, the sensitivity of the electrode to phenyl valerate

decreased with increasing rotation rate. Equation 4.27 was ﬁtted to the experimental data

using the previously determined Pm, 62 and 63 parameters (63= 1.81, 192= 0.23, Pm

=0.0091 crn/s). A best ﬁt value of 61: 1.1 (d: 0.25) was obtained.

To summarize the results, the average values of Pm, 61 , 92 and 63 came out to be

0.0091 cm/s, 1.1, 0.23 and 1.81, respectively. The thickness of the interface (L) as

measured using an ellipsometer was 25 nm. If we assume partition coefﬁcient k1, to be 1,

 

. . D f k p -8 2
then the value of Df usmg the relatlon,Pm = L , would be 2.27 x 10 cm /s.

Although these calculations may be approximate, but it suggests that the diffusion in the

enzyme ﬁlm is approximately 3 orders of magnitude slower than in bulk electrolyte.

74

Similar effects have already been reported in layered polyelectrolyte ﬁlms (Coche-

Guerente, Labbe et al. 2001). Assuming that the immobilized tyrosinase has the same

kinetic constants as in the homogeneous solution, the concentrations [E2] and [E3] of

monophenolase and catecholase active sites can be determined using the following

 

 

relations:
D f022

(k2 /1<2 152] = L2 (434)
D f 193?

(k3 /K31E3] = L2 (4.35)

Table 4.1 shows the calculated values of [E2] and [E3] along with the reported values of

kinetic constants used to determine them (Coche-Guerente, Labbe et a1. 2001). Although
these calculations are coarse approximation, they suggest monophenolase active sites

represent only 66% of the catecholase sites and about 40 % of the total sites.

4.4.6 Simulations and Discussions

Figure 4.6 show the simulated concentration proﬁles of phenyl valerate, phenol, catechol

and o-quinone normalized to phenyl valerate bulk concentration (S1(oo)) as a function of

relative position (x/L) within the interface. The concentration proﬁle was simulated using

equations 4.23-4.26, along with the experimentally determined values of different

parameters (Pm=0.0091 cm/s, (91 =1.1,t92 = 0.23 , 93 =1.81, (0:500 rpm and De = 2.2

x 10'5 cmz/s). As expected, the concentration of o-quinone, and the concentration

75

as
gradients of phenyl valerate (-a—1—) and phenol concentration (—a—Z-) , at the electrode
x x

surface (x = 0) are zero.

The model’s predictions help explain the observed (Figure 4.5) decrease in
current with increasing rotation rates in the presence of phenyl valerate. The
electrochemical transduction step regenerates catechol from o-quinone. A portion of the
catechol is lost by diffusion through the stagnant ﬁlm, and the remainder is oxidized to o-
quinone, which increases the sensor’s output. As the electrode rotation rate increases, the
ﬁlm mass transfer coefﬁcient also increases and increasing ﬁaction of the catechol is lost.
As a consequence, less catechol is recycled to o-quinone, and the o-quinone
concentration at the electrode is reduced, leading to lower sensitivity for phenyl valerate
detection. Consistent with this mechanism, Figure 4.7 shows that the predicted

concentration gradient of o-quinone at x=0 decreases as the rotation rate increases.

Figure 4.8 shows the simulated current sensitivity, ng , as a function of amount

of NEST esterase activity (61) and tyrosinase’s catecholase activity ((93 ), assuming Pm =
0.0091 cm/s and a) = 500 rpm. At low 61 values, ng increases as 61 increases,

indicating that the NEST activity is rate limiting. However, as 61 increases, the 52,,
curve approaches an asymptote, indicating that the biosensor response is becoming

limited by the catechol recycling (63 ). For a given, 61, the current sensitivityng

increases with 63 , because at higher values of catecholase activity enzymatic recycling

76

of o-quinone becomes more and more efﬁcient, leading to ampliﬁed responses or higher

sensitivities.

Ampliﬁcation factor (AF) can be deﬁned as the ratio of current sensitivities in the

C . . C
presence (S pv) and absence of catecholase act1v1ty (S pv, 63 = 0 ).

SC
V
AF: 6 p (4.36)

Spv,63 = 0

 

The following relation for the AF can be derived from Equation 4.33 and 4.42. This
equation shows that AF is a function of Thiele modulus 63 and the ratio P—m , which is

me

the reciprocal of Sherwood number.

Pm .
—— 03 811111 63 + 00811 03

 

P m
AF = (1 + —'" )63( P6 ) (4'43)
me in— 19 cosh 0 + sinh (9
m 3 3 3
6

Figure 4.9 shows predicted AF values as a function of 193, assuming experimentally
realistic values of Pm = 0.0091 cm/s and a) = 500 rpm. Ampliﬁcation increases with 93

and, depending upon the rotation rate, ampliﬁcation factors as high as 20 can be
achieved. The mathematical model predicts about 2.5-fold signal ampliﬁcation via

recycling for our bi-enzyme electrode at a rotation speed of 500 rpm.

77

Collectively, the results of this study present an important contribution from the
point of view of optimization of bi-enzyme biosensors that involve substrate recycling.
The results suggest strategies that can be used to increase biosensor sensitivity. For the
NEST biosensor, increasing the tyrosinase concentration and decreasing the mass transfer
coefﬁcient give higher signals for a given NEST loading. This result is interesting since
tyrosinase is relatively inexpensive and commercially available, while NEST is not
commercially available, and its expression and puriﬁcation requires special expertise. In
order to achieve higher signals, currently, we are exploring different ways of increasing
tyrosinase loading on the surface. We are also looking at the possibility of using this
interface as a sensor for detecting NEST inhibitors such as organophosphorous

compounds.
4.5 CONCLUSIONS

This chapter presented a theoretical treatment for bi-enzyme electrodes with substrate
recycling. The model was validated by studying the response of bi-enzyme rotating disk
electrode consisting of two enzymes, NEST and tyrosinase, to phenyl valerate, phenol
and catechol under varying rotating speeds. The validated model helped us determine and
quantify the inﬂuence of important parameters such as mass transport in the bulk and
enzyme layer, partition coefﬁcients, enzyme kinetics and catechol recycling on the
sensitivity of the sensor. This information can be helpful for optimizing the metrological

characteristics of bi-enzyme electrodes.

78

NEST

.
NH3

  
 

 

3 BI-layers

 

Gold
electrode

Figure 4.1: Molecular architecture of bi-enzyme electrode

79

 

I Electrode I

 

Df
(:14 83
containing
layer 2

‘
r

 

 

 

Stagnant ii
film 5 o

 

 

Bulk
solution

 

Figure 4.2: Schematic representation of a rotating disk bi-enzyme electrode and principle
of its functioning in the presence of phenyl valerate substrate. S1, S2, S3 and Q4 denote
the substrate phenyl valerate, phenol, catechol and o—quinone, respectively. E1 denotes

NEST esterase activity. E2 and E3 denote tyrosinase’s phenolase and catecholase activity,
respectively. L denotes the thickness of the enzyme.

80

 

‘7 y = 0.0155x + 0.00093

E R2 = 0.9818
4-0 0.003 -

'1‘ .

0 0.002 —

E,
m “ 0 001

o __

(I) .

\

V-

0 . . .

 

 

 

0.05 0.07 0.09 0.11 0.13 0.15
w-1/2 (rad-1/281/2)

 

    

 

 

 

(a)
10
9 - y=46.641x+2.1358
8 - R2=0.9853
(U 7 _
‘5
Q 6-
013 5 4
(I) 4 -
3 _
2 _
1 _
O l T l l
0.05 0.07 0.09 0.11 0.13 0.15

w—1/2 (rad-1/281/2)

(b)

SC
ct
Figure 4.3: Reciprocal plots of S 3, and a

 

versus the square root of rotation rate for

Ct
electrode A.

81

260;
240:
220:
200-

180- i

Current (Acmlmol)

160a

 

 

ﬁr I 1

do . 100 ' 150 200 250 ' 300 ' 350

w (radian/s)

Figure 4.4: Cathodic sensitivity, SIC? h , in the presence of phenol as a function of rotation

rate.

82

180-
1601
140-
120:
100:

80-

Current (Acmlmol)

 

60 l r l ' l ' l l
50 100 150 200 250 300
w (radians/s)

 

i I T T

350

C

Figure 4.5: Cathodic sensitivity, S pv ,

in the presence of phenyl valerate as a function of

rotation rate.

83

Phenyl Valerate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 0.5

3,

a;

5 0.4 -

E

g 0.3 —

0

c:

O

0 0.2 . . . .

0 0 2 0.4 0 6 0 8 1
x/L
(a)
0.06 0.7 A
’5‘ ...................................... 3,
‘5. 0 05 “t i ——> H 0'6 (0-
Q +catechol\" 0 5 \
c 004 -t/ +quinone ° 8
g +phenol «~ 0.4 “(T-U:
g 0.03 «l 3:
E W” 0.3 0:)
Q) 0 02 0- o
O —- 0.2
8 001 L \ 8
O . 0.1 0
0 3 TL ; : i 1. O
0 0.2 0.4 0.6 0.8 1

X/L
(b)

Figure 4.6: Concentration proﬁles of (a) phenyl valerate (b) phenol, catechol and o-
quinone normalized to phenyl valerate bulk concentration (SI (00)) as a ﬁmction of
relative position (x/L) within the bi-enzyme interface

84

0.035

 

 

 

 

 

E
g 0.03 1
(U
*2 0.025 1 / \. 500 rpm
0)
0 o”...
C 0.02 ‘ ,3.
O .0.
0 .. - 1000 rpm
Io 0.015 _ ...
g I: 1500 rpm
2 0.01 - $233 3:
(U o. “4“" .3000 rpm
8 -
O 0.005 1 ”41“
Z

O I l I l I

O 0.2 0.4 0.8 1

0.6
x/ L

Figure 4.7: Concentration proﬁle of o-quinone normalized to phenyl valerate bulk

concentration (S1 (00)) at various rotation rates

85

 

 

 

 

 

 

 

 

 

 

 

 

‘0‘

E

g 03 =7.24
S;

g 93 =5.43
..>_

'g 0, =3.62
8

._. 03 =1.81
C

93

E, 0 . . . .

0 2 4 6 8 10

B1 (NEST esterase activity)

Figure 4.8: Current sensitivity, ng , as a function of amount of NEST esterase activity

(«91) and tyrosinase’s catecholase activity (63 ).

86

 

  
    

 

 

25
5
.5 20 4
(U
U.
c: 15 -
.9
8 10 _ Ampliﬁcation in bi-
SE enzyme electrode
3
< 5 ‘

0

 

Figure 4.9: Signal ampliﬁcation in bi-enzyme electrode due to the recycling of catechol.
For simulation, the following values of different parameters were used: Pm =
0.009lcm/s, D, = 2.2 x 10'5 cmz/s, a) = 500 rpm.

87

Table 4.1: Kinetic characteristics of tyrosinase immobilized in bi-enzyme electrode.
Values of k and K were obtained from the literature (Coche-Guerente, Labbe et al. 2001).
However, enzyme concentration was calculated using equations 4.40 and 4.41.

Monophenolase - Catecholase
activity activity
K (moI/cm 3) (2.5 1 0.3) x 10'7 (2.2 1 0.2) x 10'7
k(s'1) 20:2 760130
-1 3 -1 7 9
k/K (mol cm s ) 8.4 x 10 3.45 x 10
Enzyme concentration (mol/cm 3 ) 2.29 x 10'6 3.45 x 10'6

88

5 ARRAYS OF LIPID BILAYERS AND LIPOSOMES ON PATTERNED

POLYELECTROLYTE TEMPLATES

5.] ABSTRACT

This paper presents novel methods to produce arrays of lipid bilayers and liposomes on
patterned polyelectrolyte multilayers. We created the arrays by exposing patterns of
poly(dimethyldiallylammonium chloride) (PDAC), polyethylene glycol (m-dPEG acid),
and poly(allylarnine hydrochloride) (PAH) on polyelectrolyte multilayers (PEMs) to
liposomes of various compositions. The resulting interfaces were characterized by total
internal reﬂection ﬂuorescence microscopy (TIRFM), ﬂuorescence recovery after pattern
photobleaching (FRAPP), quartz crystal microbalance (QCM), and ﬂuorescence
microscopy. Liposomes composed of 1, 2-dioleoyl-sn-glycero-3-phosphocholine
(DOPC) and l, 2-dioleoyl-sn-glycero-3-phosphate (monosodium salt) (DOPA) were
found to preferentially adsorb on PDAC and PAH surfaces. On the other) hand, liposome
adsorption on sulfonated poly(styrene) (SPS) surfaces was minimal, due to electrostatic
repulsion between the negatively charged liposomes and the SPS-coated surface.
Surfaces coated with m-dPEG acid were also found to resist liposome adsorption. We
exploited these results to create arrays of lipid bilayers by exposing PDAC, PAH and m-
dPEG patterned substrates to DOPA/DOPC vesicles of various compositions. The
patterned substrates were created by stamping PDAC (or PAH) on SPS-topped
multilayers, and m-dPEG acid on PDAC-topped multilayers, respectively. This technique
can be used to produce functional biomimetic interfaces for potential applications in

biosensors and biocatalysis, for creating arrays that could be used for high-throughput

89

screening of compounds that interact with cell membranes, and for probing, and possibly

controlling, interactions between living cells and synthetic membranes.

5.2 INTRODUCTION

Cell membranes are complex moieties composed primarily of a bilayer lipid membrane
(BLM) and associated membrane proteins. These membranes represent one of the major
structural components of cells and are responsible for many vital cellular functions.
Biomimetic interfaces that can mimic cell membranes and their functionalities have
potential applications as biosensors and can also provide a platform for fundamental
investigations of biomolecular behavior (Stelzle, Weissmuller et al. 1993; Heysel, Vogel
et al. 1995; Sackmann 1996; Raguse, Braach-Maksvytis et al. 1998; Nikolelis, Hianik et
al. 1999). To mimic cell membranes, supported BLMs (sBLMs) have been formed on
glass, silica and unfunctionalized metal surfaces (Ariga and Okahata 1989; Kalb, Frey et
al. 1992; Sackmann 1996; Tien, Barish et al. 1998; Tien and Ottova 1998; Asaka, Ottova
et al. 1999; Plant 1999; Boxer 2000; Ross, Bondurant et al. 2001; F avero, D'Annibale et
al. 2002; Wiegand, Arribas-Layton et al. 2002; Graneli, Rydstrom et al. 2003). Although
sBLMs have enabled researchers to probe properties such as phase transition, lateral
diffusion, permeation, and lipid-protein interactions, they have two serious limitations:
(1) they do not provide space between the bilayer and the underlying substrate to
accommodate hydrophilic moieties of trans-membrane proteins, and to allow lateral
mobility of membrane components, and (2) they lack the ionic reservoirs that most cell

membranes need to ensure their biological function.

90

To overcome these limitations, new approaches that involve the use of
hydrophilic cushions on which BLMs can be deposited are being adopted by researchers.
Such cushions have consisted of hydrogels, polymeric tethers, polymer ﬁlms and
polyelectrolyte multilayers (PEMs) (Raguse, Braach—Maksvytis et al. 1998; Cassier,
Sinner et al. 1999; Naumann, Schmidt et al. 1999; Stora, Lakey et al. 1999; Wagner and
Tamm 2000; Zhang, Longo et al. 2000; Cornell, Krishna et al. 2001; Krishna, Schulte et
a1. 2001; Krysinski, Zebrowska et al. 2001; Sinner and Knoll 2001; Kugler and Knoll
2002; Naumann, Prucker et al. 2002; Proux-Delrouyre, Elie et a1. 2002; Zebrowska,
Krysinski et al. 2002; Zhang, Vidu et al. 2002; Ma, Srinivasan et al. 2003; Moya, Richter
et al. 2003; Naumann, Schiller et a1. 2003; Naumann, Walz et al. 2003; Perez-Salas,
Faucher et al. 2003; Terrettaz, Mayer et al. 2003; Yin, Burns et al. 2003; Moncelli,
Becucci et al. 2004; Munro and Frank 2004). PEMs offer the following advantages
(Cassier, Sinner et al. 1999; Zhang, Longo et al. 2000; Kugler and Knoll 2002; Zhang,
Vidu et al. 2002; Ma, Srinivasan et al. 2003; Moya, Richter et al. 2003; Perez-Salas,
F aucher et al. 2003): (1) they are robust and easy to fabricate, (2) they can be deposited
on virtually any surface, (3) they can provide a reservoir for electron transfer mediators
and cofactors for sensor applications, and (4) their porosity and ﬂexibility may allow the
protein to exist in its natural conformation while bound to the BLM. Lipid bilayers
composed of negatively charged lipids like 1, 2-dioleoy1-sn-glycero-3-phosphate
(monosodium salt) (DOPA), 1-stearoyl-2-oleoyl-phosphatidylserine (SOPS), and 1,2-
dimyristoyl-sn—glycero-3-[phospho-rac-(1-g1ycerol)] (sodium salt) (DMPG) blended with
other zwitterionic lipids like 1, 2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 1-

palmitoyl-2-oleoylphosphatidylcholine (POPC) have already been shown to form on

91

polyelectrolytes such as poly(allylarnine hydrochloride) (PAH), poly(ethyleneimine)
(PEI) and poly(diallyldimethyl ammonium chloride) (PDAC) coated substrates (Sohling
and Schouten 1996; Cassier, Sinner et al. 1999; Zhang, Booth et a1. 2000; Zhang, Longo
et al. 2000; Kugler and Knoll 2002; Ma, Srinivasan et al. 2003). These studies (Kugler
and Knoll 2002) have shown that upon increasing the percentage of charged lipids, lipid
coverages increase, and diffusion coefﬁcients decrease. In another approach (Tiourina,
Radtchenko et al. 2002), PEMs were adsorbed on melamine formaldehyde latex particles
which were soluble at low pH, resulting in the formation of thin polyelectrolyte shells
upon dissolving the core. Lipid bilayers were then formed on the empty shells by
exposing them to charged vesicles. The properties of this system as an artiﬁcial cell were

then evaluated.

Arrays of BLMs have also been fabricated on glass and gold surfaces (Groves,
Ulman et al. 1997; Groves, Boxer et al. 1998; Hovis and Boxer 2000; Kung, Groves et al.
2000; Kung, Kam et a1. 2000; Zhang, Longo et al. 2000; Kam and Boxer 2001; Groves
and Boxer 2002; Proux-Delrouyre, Elie et al. 2002; Saccani, Castano et a1. 2003). In one
approach, a patterned poly(dimethyl siloxane) (PDMS) stamp was brought into contact
with an sBLM formed on a glass slide and then removed (Hovis and Boxer 2000).
Approximately 90% of the lipids in areas in contact with the stamp were transferred to
the stamp’s surface, resulting in arrays of BLM patches separated from one another by
regions of bare glass. The same group showed that the bilayer can be preassembled
directly onto oxidized PDMS surfaces and then transferred intact to the glass slide.
Bilayer patches in the resulting arrays were found to be fully ﬂuid and stable under water.

To date, this methodology has been applied only to a limited number of surfaces.

92

There is general interest in techniques that can help extend these types of
approaches to other substrates, and also address the two limitations of sBLMs described
earlier. The approach in this study is based on the ionic layer-by-layer (LBL) assembly
technique introduced by Decher (Decher and Hong 1991), microcontact printing (pCP)
developed by the Whitesides group (Kumar and Whitesides 1993), and the polymer-on-
polymer stamping process (POPS) developed by Hammond and coworkers (Jiang and

Hammond 2000; J iang, Zheng et al. 2002).

In this paper (Kohli, Vaidya et al. 2006 ), we present methods to fabricate arrays
of BLMs and liposomes on PEMs. Arrays of BLMs were created by exposing PDAC
patterns, polyethylene glycol (m-dPEG acid) patterns, and PAH patterns on PEMs to
liposomes of various compositions. Total intemal reﬂection ﬂuorescence microscopy
(TIRFM) and quartz crystal microbalance (QCM) gravimetry were used to monitor
liposome adsorption to PEMs. Fluorescence recovery after pattern photobleaching
(FRAPP) and ﬂuorescence microscopy were also used to characterize the resulting

interfaces.
5.3 EXPERIMENTAL SECTION

5.3.] Materials

Sulfonated poly(styrene) (SPS) (Mw ~70,000), poly(diallyldimethyl ammonium
chloride) (PDAC) (MW ~100,000), and poly(allylarnine hydrochloride) (PAH) were
obtained from Sigma (St. Louis, MO). 1,2-dioleoyl-sn-glycero-3-phosphate
(monosodium salt) (DOPA, Figure 5.1b), 1,2-dioleoyl-sn-glycero-3-phosphocholine

(DOPC, Figure 5.1a), and 1-palmitoyl-2-[6-[(7-nitro-2-1,3-benzoxadiazol-4-

93

yl)amino]hexanoyl]-sn-glycero-3-phosphocholine(l6:0-06:0 NBD PC, Figure 5.1c) were
purchased from Avanti Polar Lipids (Alabaster, AL). NBD-PC serves as the ﬂuorescence
probe. l-tetradecanethiol, 4-(2-hydroxyethyl)piperazine-1—ethanesulfonic acid sodium
salt (HEPES) was obtained from Fluka (St. Louis, MO). The m-dPEG acid was
purchased from Quanta Biodesign (Powell, OH). Structures of PDAC, PAH, SPS and m-
dPEG acid are given in'Figure 5.2. Sylgard 184 silicone elastomer kit (Dow Corning,
Midland MI) was used to prepare the poly(dimethylsiloxane) (PDMS) stamps for uCP.
The ﬂuorosilanes were purchased from Aldrich Chemical (St. Louis, MO). Ultrapure
water was supplied by a Nanopure-UV four-stage puriﬁer (Barnstead International.

Dubuque, IA); the puriﬁer was equipped with a UV source and a ﬁnal 0.2 um ﬁlter.

5.3.2 Preparation of stamps

The PDMS stamps were made by pouring a 10:1 solution of elastomer and
initiator over a prepared silicon master. The silicon master was pretreated with
ﬂuorosilanes to facilitate the removal of the PDMS stamps from the silicon masters. The
mixture was allowed to cure overnight at 60 °C. The masters were prepared in the
Microsystems Technology Lab at MIT and consisted of features (parallel lines and

circles) ﬁ'om 1 to 20 um.

5.3.3 Preparation of liposomes

Small unilamellar liposomes of two different compositions were prepared: (1)
89% DOPC, 10% DOPA and 1% NBD-PC (referred in the paper as 90% DOPC/10%
DOPA), (2) 79% DOPC, 10% DOPA and 1% NBD-PC (referred in the paper as 80%

DOPC/20% DOPA). These liposomes were prepared by mixing appropriate amounts of

94

lipids in chloroform. This mixture was then dried under nitrogen, making sure lipid
formed a thin, cake-like ﬁlm on the walls of the test tube. The residual chloroform was
removed under high vacuum. The lipids were then reconstituted in HEPES buffer (pH
7.4, 100 mM NaCl), and the resulting liposome solution was sonicated until it became

clear using a Bransonic bath sonicator (Branson Ultrasonics Corporation, Danbury, CT).

5.3.4 Preparation of Polyelectrolyte Multilayers (PEMs)

A Carl Zeiss slide stainer equipped with a custom-designed ultrasonic bath was
connected to a computer to perform LBL assembly (Yoo, Shiratori et al. 1998; Lee,
Zheng et al. 2002; Zheng, Lee et al. 2002; Lee, Hammond et al. 2003). For PDAC/SPS
and PAH/SPS multilayers, the concentration of SPS, PDAC and PAH solution was 0.01
M, 0.02 M and 0.01M respectively as based on the molecular repeat units. All
polyelectrolyte solutions contained 0.1 M NaCl and were at a pH of 7.0. To deposit
PEMs, the glass slides were cleaned with a dilute Lysol water mixture in a sonicator.
These slides were then dried under N2 gas and were further cleaned using Harrick plasma
cleaner (Harrick Scientiﬁc Corporation, Broading Ossining, NY) for 10 min at 20 Pa. To
form the ﬁrst polyelectrolyte bilayer, the slides were immersed for 20 min in a PDAC (or
PAH) solution. Following two sets of 5 min water rinse with agitation, the slides were
subsequently placed in a SPS solution for 20 min. They were rinsed again with water, and

this process was repeated to build multiple layers.

5.3.5 Preparation of arrays:

Two different schemes were used to prepare arrays of BLMs. Scheme 1 (Figure

5.3): A PDMS stamp was dipped in a 250 mM solution of PDAC (or 200 mM solution of

95

PAH) in 75/25 ethanol-water mixture for about 20 min (J iang and Hammond 2000; J iang,
Zheng et al. 2002). The stamp was then” washed with ethanol, dried under nitrogen and
brought in contact with a glass slide that was coated with 5 PDAC/SPS (or PAH/SPS)
bilayers, with SPS forming the uppermost layer. The stamp was removed after 15 min,
and the resulting PDAC (or PAH) patterns were then rinsed with water to remove
unbound or loosely bound PDAC (or PAH). The stamp was then exposed to
DOPC/DOPA liposomes of varying compositions. Scheme 2 (Figure 5.4): A PDMS
stamp was dipped in a 100 11M solution of m-dPEG acid in a 75/25 ethanol-water mixture
for 30 min (Kidambi, Chan et al. 2004). The stamp was then washed, dried under
nitrogen and brought in contact with the glass slide coated with 4.5 PDAC/SPS
multilayers with PDAC as the topmost layer. The stamp was removed after 20 min and
the resulting m-dPEG acid patterns were then rinsed with water to remove excess m-
dPEG acid. These patterns were then exposed to negatively charged DOPC/DOPA
liposomes. All the ﬂuorescence images were obtained using the Nikon Eclipse E 400
microscope (Nikon, Melville, NY) having a ﬁlter cube (Ex: 465-495/DM: SOS/Em: 515-

555)

5.3.6 Total internal reﬂection microscopy:

The experimental setup and the ﬂow cell for TIRFM have been described previously
(Gajraj and Ofoli 2000). Brieﬂy, the apparatus consists of an inverted microsc0pe (Zeiss
Axiovert 135M, Carl Zeiss Inc. Thomwood, NY), the 488 nm line of a 5W continuous
wave argon ion laser (Lexel Lasers Model 95, Fremont, CA), a side—on photomultiplier
tube (Hamarnatsu R4632, Bridgewater, NJ) jacketed in a thermoelectrically cooled

housing (TE l77-TSRF, Products for Research, Danvers, MA), a CCD camera (NT 1,

96

VE1000, Dage-MTI, Michigan City, IN), and a modular automation controller (MAC
2000, Ludl Electronic Products, Hawthorne, NY) that regulates the voltage supply to the
photo-multiplier tube and controls translation of the X—Y stage. A double syringe pump
system (Model 551382, Harvard Apparatus South Natick, MA) was used for infusion and
withdrawal of sample solutions at identical rates from a custom designed ﬂow cell. For
measurements that require photon counting, an SR400 photon counter (Stanford Research
Systems, Sunnyvale, CA) was used after the output signal from the PMT was ampliﬁed
by a fast preampliﬁer (SR445, Stanford Research Systems, Sunnyvale, CA).
Fluorescence intensities were recorded using software written in Labview 6.0 (National
Instruments, Austin, TX). A 500 nm long-pass band ﬁlter was used to separate the
excitation and emission wavelengths. An optical chopper (SR 540, Stanford Research
Systems, Sunnyvale, CA) was used to prevent unintended photobleaching of ﬂuorophores
during experiments. The photon counter was triggered by an output reference voltage
from the chopper, so that data collection only occurred during periods when the ﬂow cell
was illuminated by the monitoring beam. For TIRFM experiments, the ﬂow cell was
initially ﬁlled with buffer. Subsequently, a liposome solution was introduced into the
ﬂow cell at a controlled ﬂow rate. After ﬁlling the ﬂow cell, the infusion was halted, and
adsorption was allowed to continue for 1 h, followed by a buffer wash (for desorption
experiments or FRAPP experiments). Measurements were done at a temperature of 22°C
a: 1°C, which is above the phase transition temperature of both DOPC (-20°C) and DOPA

(-8°C) (Albrecht, Johnston et al. 1982).

97

5.3.7 Fluorescence recovery after pattern photobleaching (FRAPP):

The experimental conﬁguration used for FRAPP has been depicted in Figure 5.5.
Stripe patterns were imposed on the substrate by directing the 488 nm laser beam
(expanded through a 5X beam expander (Edmund optics, Barrington NJ) through a 50 or
100 line per inch Ronchi ruling (Edmund optics, Barrington,NJ) placed in a real image
plane. Placing a ruling in a back image plane, results in projection of a sharply focused
pattern of alternating dark and bright ﬁ'inges on the substrate in the sample plane. This
back real image plane is located near the epi-port of the Axiovert 135M microscope,
coincident with the ﬁeld iris diaphragm of a ﬂuorescence light illuminator (Zeiss)
mounted through the epi-port. The ﬂuorescent light illuminator has a lens that projects
the beam through a Zeiss ﬁlter cube (Ex: 450-490/DM: 510/Em: 515-565), and a 32X
Zeiss objective on to the substrate. For focusing the objective, a ﬂuorescent coating was
brushed onto a glass slide using a yellow ﬂuorescent marker (Sanford). The objective
position was then adjusted till the fringes appeared in sharp focus. The fringe spacing was
estimated by using a reticule containing 10 lines per centimeter in the eyepiece. The glass
slide was then replaced by the sample slide. Prior to initiating photobleaching, minute

adjustments were made to the objective focus knob to bring the ﬁ‘inges into sharp focus.

Beam alignment was checked at the start of each experiment and objective
focusing was performed prior to initiation of every FRAPP experiment. For FRAPP
experiments, it is crucial that the monitoring beams and the photobleaching beams be
precisely coincident. Checking that the beams coincided was accomplished with the
following steps. (1) The monitoring and photobleaching beams were projected on to a

screen some distance away from the vibration isolation table, and beam recombination

98

was checked.(2) Occasionally, as an additional check for recombination, once the laser
beam was directed through the Ronchi ruling so that fringe pattern was formed on the
ﬂuorescently coated glass slide, a CCD camera or the microscope oculars was used to re-
conﬁrm that the beams recombined by observing the overlaying of the faint monitoring
fringes over the photobleaching fringes (Robeson 1995). For all experiments, the
monitoring beam intensity ranged from 1 11W to 5 uW. For photobleaching, the beam
intensity required was approximately 0.5 W. Additional neutral density ﬁlters (NDF)
were required to attenuate the monitoring beam to the desired level. The use of additional
NDFs causes signiﬁcant deviation in the path of the monitoring beam, so checking for
recombination is especially crucial. A high bleaching intensity is required to obtain a
sufﬁcient bleach depth within a reasonable time. Typical bleach times varied from 350

ms to 500 ms. Stripe periodicity in the sample plane was approximately 25 pm.

For FRAPP experiments, liposome adsorption on the polyelectrolyte coated
substrates was initiated by directly introducing the liposome solution at a ﬂow rate of
0.34 mein for approximately 10-12 minutes using the syringe pump. The inﬁrsion was
then halted and adsorption was allowed to continue for approximately 45 minutes. The
flow cell was subsequently ﬂushed with 4-5 ﬂow cell volumes of buffer in order to

remove the ﬂuorescently labeled liposomes in the bulk solution.

5.3.8 Theory and Data Analyses for Fluorescence recovery after pattern

photobleaching

FRAP is a technique that is commonly used to obtain estimates of translational
(lateral) mobilities of proteins or lipids. There are two principal variants of this method.

The ﬁrst involves using a focused laser beam to create a small spot. This is known as spot

99

photobleaching and can be effected using epi- illumination as well as using TIRFM. In
the second method (FRAPP), a laser beam is passed through a Ronchi ruling placed in a
back image plane to create a pattern of alternating dark and bright stripes of well deﬁned
periodicity over a broadly illuminated area. The principle advantage of FRAPP over spot
photobleaching lies in the well deﬁned characteristics of the pattern obtained in the
sample plane. For spot photobleaching, the recovery kinetics and shape depend very
strongly on the shape of the focused spot at the interface. The precise shape of the spot is
difﬁcult to discern as the beam travels through several optics in order to form the spot. As
a result there is considerable uncertainty in measurements of diffusion coefﬁcient
especially for non-ideal samples. In contrast the well deﬁned periodicity of the stripe

pattern produced in the sample plane offers two distinct advantages: (i) It allows us to
measure slow as well as fast diffusion coefﬁcients (10'10 cmZ/s to 10'7 cmz/s), as the

stripe periodicity can easily be varied and (ii) It allows us to examine samples where
multiple populations with different mobilities coexist, using models that describe such
populations. Other variants of FRAPP involve creation of the stripe pattern using the
intersection of two beams either in the sample plane (using TIRF illumination) or in a
back image plane (using EPI illumination). The advantage of using TIRFM—FRAPP over
EPI—FRAPP is that due to the surface selectivity of TIRFM, a buffer wash to eliminate
bulk ﬂuorophores is not necessary. The trade-off is the complexity in the experimental

conﬁguration.

5.3.9 Theory

A complete mathematical analysis of ﬂuorescence recovery after pattern

photobleaching has been presented by Starr and Thompson (Starr and Thompson 2002).

100

A brief description of this analysis is presented below. Consider a sample of mobile
ﬂuorescent lipids with diffusion coefﬁcient D. The ﬂuorescence emission from such a

sample can be described as:

F(t)=Q j [I(x,y)C(x,y,t>dxdy (5,1,
where Q is a proportionality constant that incorporates factors such as ﬂuorophore
quantum yield and the instrument constant , I(x, y) is the monitoring beam intensity and
C(x, y, t) is the concentration of unbleached molecules as a ﬁmction of position and time.
The equation that describes the intensity proﬁle for a Gaussian-shaped laser beam

intersected by a Ronchi ruling placed in a back image plane is given by

 

It add

I(x, y) = [Ziexp [— 2(x S: y )].[1+ 2 c" cos(nkx)] (52)

Here, 10 is the intensity at the origin, 3 is the 1/e2 radius of the expanded beam, k is the

spatial frequency of the pattern, superimposed on the sample plane, deﬁned by:

k =— ,6" =—(-l)[_2_] (5.3)

and a is the spatial period of the stripe pattern. The concentration proﬁle of diffusing

ﬂuorophores is,

 

1 00 00 _ .2 _ 2
0050:2771); j jC(x',y'.0).exp[—(" x) 4“,”)? ”1000' (5.4)

101

 

f0

pu

11

Hi

The initial concentration of unbleached ﬂuorophores in the sample is given by:

C(x,y.0)=Cexp[—K1(x.y)] (5.5)

where C is the total ﬂuorophore concentration and K is a constant that incorporates the

bleach pulse duration, quantum yield and absorptivity of the ﬂuorophores.

After developing Equation (5.1) using Equations (5.2), (5.3), (5.4) and (5.5), for cases
where the stripe periodicity is small compared to the illuminated area, the following
equation can be derived for a single mobile species model, which assumes that there is

one population of mobile ﬂuorophores, in addition to an immobile fraction.

2 2
¢(z) = ¢(0) +iz‘-[1 — ¢(0)].[1-[%).{exp[_ 4:21): J + éexp[— 36:2Dt]}] (5.6)

for t2 0, where ¢(t) is the ratio of the postbleach ﬂuorescence (t>0 , aﬁer the bleach

 

 

pulse) to the prebleach ﬂuorescence ¢(t<0), p represents the fraction 0f the ﬂuorophores
present that are mobile and a and D have already been deﬁned previously. This analysis

neglects rapidly decaying (higher order) terms.

5.3.10 Data Analysis and curve fitting

The model was ﬁt to the data using OriginPro 7.5 (OriginLab Corporation, Northampton
MA) which uses the Levenberg-Marquardt algorithm for non-linear least squares ﬁtting.
The post-bleach ﬂuorescence emission intensity was normalized against the pre-bleach

ﬂuorescence emission intensity (KO).

102

5.3.11 Microgravimetric experiments:

A QCM analyzer (5 MHz crystals, Maxtek Inc., Research Quartz Crystal
Microbalance, Santa Fe Springs, CA) linked to a computer with RQCM data log software

(Maxtek Inc.) was used for microgravimetric measurements. Quartz crystals (Maxtek
Inc.) (geometrical area 1.26 cm2) were used. The QCM crystals were cleaned in piranha

solution (seven parts by volume concentrated sulfuric acid and three parts hydrogen
peroxide) for 30 s and then dipped in a 1 mM solution of 3-mercaptopropanoic acid in
ethanol for 24 h. PAH/SPS and PDAC/SPS multilayers were then deposited ex-situ on
the crystals according to previously described procedures. Frequency changes of the
quartz crystals were measured after obtaining a baseline oscillation frequency change
using a 10 mM HEPES buffer (pH 7.4). The measurements were done at room

temperature.
5.4 RESULTS AND DISCUSSIONS

5.4.1 Characterization of liposome adsorption by TIRFM

Curves A and B in Figure 5.6 depict adsorption of negatively charged liposomes
composed of 90% DOPC/10% DOPA to PEMs, with positively charged PDAC and
negatively charged SPS as the top layers, respectively. The higher intensities obtained in
Curve A indicate that the liposomes adsorbed preferentially onto PDAC, presumably due
to electrostatic interactions. To investigate the role of liposome charge on subsequent
adsorption, we repeated the experiments with liposomes composed of 80% DOPC and
20% DOPA (Curves C and D in Figure 5.6), with PDAC and SPS as the top layer,

respectively. The higher concentration of negatively charged lipids (DOPA) in the

103

liposomes further increased the preference of liposomes for PDAC, suggesting that
adsorption of liposomes on PEMs is signiﬁcantly inﬂuenced by electrostatic interactions

between the charged lipids and polyelectrolytes.

Curves C and D depict experiments conducted with a separate batch of liposomes
from those represented by Curves A and B. Due to variability in liposome characteristics
(due to small changes in labeling ratios, etc.), the magnitude of ﬂuorescence emission
may vary somewhat from batch to batch. However, unlike the case for Curve A, Curve
C for adsorption on PDAC does not appear to approach a saturation value within the
duration of the experiment. This may be the result of multilayer deposition at the higher
concentrations of DOPA. To avoid the possibility of multilayer formation, liposomes

composed of 10% DOPA and 90% DOPC were used for all remaining experiments.

To determine the reversibility of liposome adsorption to PDAC/SPS PEMs, the
ﬂow cell was ﬂushed with 3-4 volumes of liposome-free buffer after liposome
adsorption. The wash signiﬁcantly reduced the ﬂuorescence emission when SPS was the
top layer, but had a much smaller effect when PDAC was the top layer (data not shown).
This result suggests that the lipids are binding more strongly to PDAC than to SPS. It is
possible that the buffer wash experiments reﬂect, in part, depletion of liposomes in the
bulk liquid. However, the intensity of the evanescent wave decays exponentially with
distance from the interface, and has a penetration depth of only about 80 nm in the
TIRFM setup used to conduct these experiments. As a result, illumination of ﬂuorophores
in the bulk solution would only make a small contribution to the total ﬂuorescence

emission measured. In fact, this interfacial sensitivity allowed TIRFM to monitor

104

liposome adsorption on slides coated with lipid monolayers to show that binding kinetics

are strongly inﬂuenced by buffer ionic strength (Kalb, Frey et al. 1992).

Liposome adsorption on surfaces coated with m-dPEG acid (see structure in
Figure 5 .2) was also studied. Figure 5.7a shows liposome adsorption onto PEMs with m-
dPEG as the top layer, followed by a wash with liposome free buffer. While some lipids
did adsorb onto the m-dPEG acid layers, ﬂushing the ﬂow cell with 3-4 ml of buffer after
adsorption resulted in a large decrease in ﬂuorescence for m-dPEG acid coated
substrates. This suggests that the liposomes were only loosely bound. By contrast, a
buffer wash after approximately 45 min of liposome adsorption produced only a small

decrease in ﬂuorescence emission in the case of PEMs topped with PDAC (Figure 5.7b).

In the experiments depicted in Figure 5.7b, liposome adsorption to PEMs with
either PDAC (top curve) or m-dPEG acid (bottom curve) as the topmost layer was halted
by introducing buffer (at t = 0 min on the graph), and then the ﬂuorescence emission
intensities were monitored to track the process of desorption. To enable direct
comparison of the results, the ﬂuorescence emission intensities in each data set were
normalized against the corresponding ﬂuorescence emission intensity recorded prior to
initiation of the buffer wash. There was a 70% decrease in the ﬂuorescence emission
intensity for the m-dPEG case, compared to a 10% decrease for PDAC. The apparent
weakness of liposome adsorption on surfaces coated with both SPS and m-dPEG was
conﬁrmed when ﬂuorescence microscopy was used to characterize arrays of BLMs
deposited on patterns of PEMs with PDAC, SPS, and m-dPEG as the t0pmost layer
(results discussed below). While the mechanism by which PEG resists biomolecular

adsorption is not completely understood, it is believed to stem either from steric

105

exclusions, which is an entropic effect caused by the unfavorable change in free energy
associated with the dehydration and conﬁnement of polymer chain with high
conformational freedom (Harris 1992), or from long range electrostatic repulsions
(Feldman, Hahner et al. 1999; Kreuzer, Wang et al. 2003). Previous studies from our
group have shown that m-dPEG acid can also resists the adsorption of PDAC (Kidambi,
Chan et al. 2004), which is a positively charged polyelectrolyte. Therefore, we believe
the resistance of m-dPEG acid to negatively charged liposomes stems primarily from

steric repulsion.

5.4.2 Formation of arrays of lipid bilayers

Figures 5.8a and 5.8b show ﬂuorescence images of the line and circular patterns,
respectively, of liposomes deposited on PDAC. As clearly shown by these results for
both linear and circular arrays, liposomes bound preferentially to PDAC (brightly
ﬂuorescing features), and negligibly to SPS (dark, featureless background). Similar
results were obtained when liposomes were deposited on a surface stamped with a weak
polyelectrolyte (PAH) instead of a strong polyelectrolyte (PDAC) on SPS (Figure 5.8c).

These results are consistent with those obtained during the adsorption experiments

(Figure 5.6).

In another approach, m—dPEG acid was stamped on a PEM-coated glass slide,
with PDAC being the topmost layer. The m-dPEG acid molecule has a carboxylic acid
group on one end. At a pH above the pKa, the acid group has a negative charge and can
be stamped onto PDAC, resulting in patterns of m-dPEG acid on PDAC(Kidambi, Chan

et al. 2004). Upon subsequent exposure to the liposome solution, liposome adsorption

106

occurred on the exposed PDAC, but not on the m-dPEG acid patterns (Figure 5.9).
Therefore, the ﬂuorescence patterns in Figure 5.9 are the negative replicas of those in.
Figure 5.8. This ability to make either the positive or the negative image of the stamp
adds to the versatility of this technique. Similar ﬂuorescence images were obtained when

PEM coated glass slides, with PAH being the topmost layer were used (data not shown).

5.4.3 Assessment of liposome adsorption and rupture to form bilayers

FRAPP was used to determine if the adsorbed liposomes remained intact on the surface
or ruptured to form a bilayer. Figures 5.10a and 5.10b show the recovery curves for lipids
deposited on PEMs with PDAC and PAH as the topmost layer, respectively.
Fluorescence recovery for lipids on PDAC and PAH was adequately described by the
single mobile species model, which assumes that there is one population of mobile
ﬂuorophores, in addition to an immobile fraction. Unlike spot photobleaching, where
100% recovery of ﬂuorophores is possible for a completely formed BLM, only 50%

recovery is theoretically possible in FRAPP under the same conditions.

The average D values on PAH and PDAC were comparable (0.224 x 10'8 cmz/s
and 0.288 x 10'8 cmz/s, respectively). The average mobile fraction (11) on PAH (0.68) was
signiﬁcantly higher than that on PDAC (0.22). Nollert et al (Nollert, Kiefer et al. 1995)
have described two outcomes when liposomes adsorb to surfaces. In the ﬁrst, liposomes
adsorb but do not rupture. As a result, they produce an immobile fraction because the
ﬂuorescently tagged lipids cannot freely migrate between liposomes. In the other
scenario, liposomes adsorb, rupture and spread to form sBLMs. Therefore, the higher p.

values observed on PAH suggest that adsorbed liposomes more readily ruptured to form a

107

bilayer on PAH than on PDAC. Similar results have been reported in a previously
published AFM study which showed that liposomes composed of DOPA adsorb onto a
PDAC-coated surface, but most remain intact (Luo, Liu et al. 2001). On the other hand,
liposomes form a continuous sBLM on substrates coated with PEI, which is a weak
polyelectrolyte like PAH. However, other feasible mechanisms could also explain the
trends seen in Table 5.1. For instance, lipids adsorbed onto PDAC may form

disconnected bilayer patches.

The diffusion coefﬁcients obtained in this study are comparable to those reported
for BLM containing SOPS and POPC deposited on PDAC/SPS multilayers (0.2 X 10'8
cmZ/s) (Zhang, Longo et al. 2000; Ma, Srinivasan et al. 2003). Our values are, however,
lower than those reported for DMPC:DOPA (10:1) BLM deposited on SPS/PAH
multilayers (Cassier, Sinner et al. 1999). In our system, the dye molecules were present in
both the upper and lower leaﬂet. However, in the reported systems (Cassier, Sinner et al.
1999) the dye molecule is present only in the upper leaﬂet and thus they are measuring
the diffusion coefﬁcient only of the upper leaﬂet. The upper leaﬂet would be less
inﬂuenced by the underlying substrate and thus would be expected to exhibit a higher
diffusion coefﬁcient than the bottom leaﬂet. Moreover, in our study, NBD was attached
to one of the hydrophobic tails of phosphocholine; Cassier et. al. have reported that
diffusion coefﬁcients can increase by a factor of four if the NBD molecule is attached to
the head-group than to the tail of the phosphocholine molecule (Cassier, Sinner et al.

1999).

108

5.4.4 Characterization of liposome adsorption and rupture by QCM

QCM analysis has been used effectively to study both kinetics of liposome
adsorption and subsequent liposome ﬁrsion to form a bilayer (Ohlsson, Tjamhage et al.
1995; Keller and Kasemo 1998; Graneli, Rydstrom et al. 2003). Adsorption of material
onto a QCM chip results in a decrease in the chip’s oscillating frequency. Under ideal
conditions, there is a linear relationship between the change in the resonant frequency

(A f) of the QCM chip, and the change in mass (Am) due to adsorption:

Am 2 C % (4.7)

where C is the mass sensitivity of the crystal, and n is the overtone number.

The A f curve corresponding to sBLM formation on silicon dioxide has been
shown to have two distinct phases (Keller and Kasemo 1998): (1) adsorption of intact
liposomes, and (2) liposome rupture and bilayer formation. When adsorbed liposomes
rupture, water trapped inside and between them is released, resulting in a loss of adsorbed
mass. Thus, Phase 1 is characterized by an increase in mass (decrease inf), and Phase 2 is
characterized by a subsequent decrease in mass (increase inf). There is no net measurable
dissipation shift for BLM formation, as a completely formed sBLM is very compact and
couples strongly to the motion of the QCM crystal. On the other hand, intact liposome
adsorption is characterized by large dissipation shifts, as adsorbed liposomes are
substantially larger and less compact structures. Thus, liposomes can be subjected to

large deformation under shear stress, resulting in increased energy dissipation.

109

Figures 5.11a (Curve i) and 5.11b (Curve iii) show the A f curves for liposome
interactions with a QCM chip coated with PEMs having PDAC and PAH as the topmost
layer, respectively. Neither curve exhibits two distinct phases, suggesting that BLM
formation on PDAC and PAH does not follow the two-phase process reported for silicon
dioxide (Keller and Kasemo 1998). The A f value measured on PDAC (—95 Hz :t 10 Hz)
lies between -26 (:t 4) Hz and -110 (:t 10) Hz, values generally observed for a completely

formed bilayer and for intact liposomes of similar size, respectively. This result, and also
the large dissipation shift (4.5 x 106) corresponding to liposome adsorption (Figure 5.1 la

Curve ii), suggest that most of the liposomes on PDAC remain intact and don’t rupture to
form a bilayer. The previously described FRAPP results on PDAC, which showed about

22% mobile fraction, are also consistent with this hypothesis.

In contrast, liposome adsorption on PAH (Figure 5.11b) resulted in a

comparatively smaller frequency change (Curve iii, - 45 Hz i 10Hz) and dissipation shift
(Curve iv, 1.2 x 10'6 i 0.1 x 106). These results, together with the FRAPP results, which

showed about 68% mobile fraction on PAH, suggest that most of the liposomes on PAH

fused to form a BLM.

The results presented above indicate that this novel method offers several
advantages over previously reported methods used to create sBLM microarrays. The use
of uCP affords precise control over the 2-D geometry of the array at the micron (and
potentially sub-micron) scale. The use of a PEM layer between the underlying surface
and the biomimetic interface adds versatility, because PEM can be adsorbed to virtually

any surface and because the PEM layer could serve as a hydrophilic reservoir needed for

110

some membrane proteins to ﬁrnction properly. This approach also provides greater
control over the physical and chemical properties of the biomimetic interface, because (1)
the thickness of the PEM reservoir can be controlled with nanometer precision via the
number of polyelectrolyte layers deposited, (2) the diverse range of PEMs available
allows the chemical properties of the reservoir to be customized, (3) the strength of
liposome deposition can be controlled, based on charge or the use of m-dPEG acid, and
(4) the degree to which adsorbed liposomes rupture to form BLM can be controlled via

the chemical properties of the top polyelectrolyte layer.
5.5 CONCLUSIONS

We have developed an approach for fabricating arrays of BLMs and liposomes on
PEMs that entails (1) establishing a patterned microarray of PEM and/or m-dPEG acid on
a surface and (2) selectively depositing liposomes on either the features or the
background of the pattern. TIRFM and ﬂuorescence microscopy results indicated that
liposomes composed of DOPA and DOPC adsorb strongly on PDAC and PAH surfaces
but only weakly on SPS and poly(ethylene glycol) (m-dPEG acid) coated surfaces. We
exploited these tendencies to create clean, patterned microarrays of BLMs and liposomes
on PAH and PDAC. The mechanism of liposome rupture and bilayer formation is not yet
fully understood. However, FRAPP and QCM results suggest that a higher fraction of
liposomes rupture to form bilayers on PAH surfaces than on PDAC, where a signiﬁcant

percentage of the adsorbed lipids are essentially immobile.

The versatility of the approach is illustrated by its unique ability to form clean

microarrays in both positive and negative image of the uCP stamp. Potential

lll

applications include novel biosensors and biocatalysts, microarrays for high-throughput
screening of compounds that interact with cell membranes, and micropattemed interfaces
to study, and possibly controlling, interactions between living cells and synthetic

membranes.
5.6 RECOMMENDATIONS FOR FUTURE WORK

One possible study for examining liposome rupture and bilayer formation on
charged, PEM-coated surfaces involves making epi-FRAPP measurements on liposomes
that adsorb to the PEMs in the presence of an electric ﬁeld. The presence of the electric
ﬁeld can alter the conformation of the charged PEMs, possibly altering the surface
characteristics and liposome rupture. Such a study would elucidate the underlying
mechanisms for liposome rupture and bilayer formation, and provide insight to tune the
surface characteristics to optimize it for BLM deposition. Another interesting method to
study vesicular fusion and rupture is using ﬂuorescence resonance energy transfer
(FRET), the radiation-less energy transfer from a donor molecule to an acceptor
molecule. The amount of energy transfer is dependent on the extent of overlap between
the emission spectrum of the donor and the absorption spectrum of the acceptor. This rate
of energy transfer varies inversely with the 6th power of the distance between the donor
and the acceptor (Lakowicz 1983). FRET has been used for studying the mechanism of
initial rupture of liposomes using single vesicle ﬂuorescence assays(Johnson, Ha et al.
2002). It would be useful to probe the mechanism of liposome rupture on the highly

charged polyelectrolyte substrates using such assays.

112

It would also be helpful to characterize liposome adsorption under a multitude of
ﬂow rates and ionic strengths to examine the shape of the adsorption proﬁle, since the
ionic strength of the solution will have a strong inﬂuence on polyelectrolyte structure of
the highly charged polyelectrolytes. In addition, this study would provide insight into

multi-bilayer membrane formation.

113

69 C") H O —\/\/W
\ /\/ RFC ‘ O _
N 0 WW
/ G o
(a)
O
‘1? H

HO, R\O\)\O/o OMW
O

(b)

/O\
N ‘ N
\ /
O JOK/VWT ‘
69 (1-0 H ,.o N
\NA/i) OW/Wv
/ e O

(C)
Figure 5.1: Structure of (a) DOPC, (b) DOPA, (c) NBD-PC

114

H 0

N
/®\ SO3'N3+
(a) (b)
/\/°\/\ /\/° 0“
H300 o
NH3+ NH3+ VT
(0) ((0

Figure 5.2: Structures of (a) PDAC, (b) SPS, (c) PAH, and (d) m-dPEG acid.

115

, PDMS stamp coated with
4"” PDAC/PAH

 

 

Stamping PDAC/PAH on
0"“ PEMs

 
 

. . e. .... , .. SPS
\h . . L,_ ”.K"

 

PDAC/ PAH patterns

 

M” H H Arrays of Lipid bilayers

3 5

 

lmﬂ‘l

Figure 5.3: Schematic representation of the process for the fabrication of arrays of BLMs
on PDAC and PAH patterned substrates.

116

m-dPEG acid patterns

 

Figure 5.4: Schematic representation of the process for the fabrication of arrays of BLMs
on m-dPEG acid patterned substrates.

117

5x beam expander

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ar-ion laser D
Optical flats
Ronchi
rulin
sample <3 lens
/'_1
L]
objective Filter
block detector

Figure 5.5: Experimental set-up for ﬂuorescence recovery after pattern photobleaching

using EPI-illumination.

118

140000
120000 ~
100000 —
80000 -
60000 —
40000 —
20000 d
.- -w-..w~e‘D:SPS
0 “ sz I I l I l I
0 500 1000 1500 2000 2500 3000 3500 4000

Time (seconds)

 

C: PDAC

A: PDAC

    
 

Fluorescence (A.U)

A..- B: SPS

 

 

 

 

Figure 5.6: Adsorption curves of (A) liposomes (10% DOPA, 90% DOPC) on PDAC. (B)
liposomes (10% DOPA, 90% DOPC) on SPS. (C) liposomes (20% DOPA, 80% DOPC)
on PDAC. (D) liposomes (20% DOPA, 80% DOPC) on SPS.

119

 

 

 

 

mama
5 20000 - a
55'
8 15000 -
c
O
U
3 10000 -
g Buffer wash
E: 50W)-

0 I l f l
0 1003 20M) mxn MK” HM”
Time (seconds)
12

 

 

 

 

 

o PDAC
g 1 . b
o . o
E 0.8 . 1K
o
:I
E 0.6 ‘
E
I: 0.4
E
z 0.2 <
o I I t
0 500 1000 1500 42000

Time (seconds)

Figure 5.7: a) Adsorption of liposomes (10% DOPA, 90% DOPC) on a glass slide coated
with PEMs with m-dPEG acid being the topmost layer. b) Buffer-wash experiments to
study liposome desorption from PEMs. The top and bottom curves depict desorption of
liposomes from PEMs with PDAC and m-dPEG as the top layer, respectively. At t=0,
adsorption of liposomes (which have adsorbed for at least 45 min) is halted by
introducing liposome-free buffer. In each curve, the ﬂuorescence intensity has been
normalized by the corresponding ﬂuorescence value obtained prior to initiation of the
buffer wash.

120

 

(C)

Figure 5.8: Fluorescence images showing (a) line patterns on a PDAC patterned substrate
(b) circular patterns on a PDAC patterned substrate (c) line patterns on a PAH patterned
substrate.

121

m-dPEG acid patterned region

./

      

    

50pm

50pm

(a) (b)
Figure 5.9: Fluorescence microscopy images showing (a) line patterns on m-dPEG acid
patterned substrate (b) circular patterns on m-dPEG patterned substrate.

122

0.25-

lntensi

 

0.15-4

 

0.1-

 

0.45:
0.40:
0.35;
0.30;
025:

Fluorescence

0.20—

 

0.15:

 

0.10

 

 

Figure 5.10: Fluorescence recovery after pattern photobleaching (FRAPP) proﬁles on
PEMs with (a) PDAC and (b) PAH as topmost layer. Only post bleach ﬂuorescence
intensity normalized against the corresponding pre bleach ﬂuorescence value is shown.
The solid lines in the ﬁgures represent ﬁts to the recovery data set with models (Wright,
Palmer et al. 1988) describing populations with a single mobile fraction and an immobile
fraction. Plots of residuals vs. time are also indicated below each ﬁgure. Average values
obtained with these models are summarized in Table 1

123

 

 

 

 

 

 

 

 

0 "‘ I I l I 12
1 2 3 4
_20 q 10
32" .3 A
v -40 ~ - “3
a (I) O
‘_
5 'i ’6 V
3 '60 ‘ I g
0' ~ 4 '4:
0) <0
5 o.
Ll. -30 - ._
< A ~ 2 g
D
-100 - _ O
-120 -2
Time(min)
4.5
A - 3.5
N (CA
.I, 'o
a.
5 - 2.5 75
0C) .2
4.:
g, s.
(D L 1.5 5
LI 1 .1)
< O
- 0.5
-O.5

 

 

 

Time(min)

(b)

Figure 5.11: Changes in QCM resonant frequency (Curves i and iii) and dissipation
(Curves ii and iv) versus time for the adsorption of liposomes on (a) PEMs having a top
layer of PDAC (b) PEMs having a top layer of PAH.

124

Table 5-1: Average lipid diffusion coefﬁcients (D), and average mobile fractions (m) for
BLMs formed on PEMs with PDAC and PAH as the topmost layer. The estimates were
obtained by averaging parameters ﬁt from all recovery curves.

 

 

 

 

Average diffusion , ,
Substrate coefficient (x 10.3 cmzls) Average mobile fraction
PAH D 0.224 t 0.01 p 0.68 :l: 0.05
PDAC D 0.288 :h 0.03 [1 0.22 :l: 0.07

 

 

 

 

 

125

 

6 TETHERED LIPID BILAYERS ON ELECTROLESSLY DEPOSITED GOLD

FOR BIOELECTRONIC APPLICATIONS

6.1 ABSTRACT

This paper presents the formation of a novel biomimetic interface consisting of an
electrolessly deposited gold ﬁlm overlaid with a tethered bilayer lipid membrane (tBLM).
Self-assembly of colloidal gold particles was used to create an electrolessly deposited
gold ﬁlm on a glass slide. The properties of the ﬁlm were characterized using ﬁeld-effect
scanning electron microscopy (FE-SEM), energy dispersive spectroscopy (EDS), and
atomic force microscopy (AFM). Bilayer lipid membranes (BLMs) were then tethered to
the gold ﬁlm by ﬁrst depositing an inner molecular leaﬂet using a mixture of 1,2-
dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-[3-(2-pyridyldithio)propionate]

(DGP), 1,2-di-0-phytanyl-sn-glycero-3-phosphoethanolamine (DPGP) and cystamine in
ethanol onto a freshly prepared electrolessly deposited gold surface. The outer leaﬂet was
then formed by the fusion of liposomes made from DPGP or 1,2-dioleoyl-sn-glycero-3-
phosphocholine (DOPC) on the inner leaﬂet. To provide functionality, two membrane
biomolecules were also incorporated into the tBLMs: the ionophore valinomycin and a
segment of neuropathy target esterase (NT E) containing the esterase domain.
Electrochemical impedance spectroscopy (EIS), UV/Visible spectroscopy and
ﬂuorescence recovery after pattern photobleaching (FRAPP) were used to characterize
the resulting biomimetic interfaces and conﬁrm membrane biomolecule activity.
Microcontact printing was used to form arrays of electrolessly deposited gold patterns on
glass slides. Subsequent deposition of lipids yielded arrays of tBLMs. This approach can
be extended to form functional biomimetic interfaces on a wide range of inexpensive

126

materials, including plastics. Potential applications include high-throughput screening of
drugs and chemicals that interact with cell membranes, and for probing, and possibly
controlling, interactions between living cells and synthetic membranes. In addition, the
gold electrode provides the possibility of electrochemical applications, including

biocatalysis, bio-fuel cells, and biosensors.

6.2 INTRODUCTION

Biomimetic interfaces, consisting of tBLMs with membrane proteins incorporated
into them or bound to them, have primarily been used for sensing applications (Raguse,
Braach-Maksvytis et al. 1998; Stora, Lakey et al. 1999; Wagner and Tamm 2000; Yin,
Burns et al. 2003). Novel approaches that can increase the range of tBLMs applications to
include areas such as biocatalysis and biofuel cells are of scientiﬁc and commercial
interest. Another major challenge in the commercialization of such biomimetic interfaces
is the costs involved to produce the gold substrates needed to deposit the tBLM on the
electrode. Although methods such chemical vapor deposition (CVD) and physical vapor
deposition (PVD) have been used to deposit gold, these methods are very expensive and
also suffer from several limitations. Therefore, there is a need for approaches that can
help reduce the cost of fabrication of substrates and overcome some of the limitations

associated with CVD and PVD.

Electroless deposition is an inexpensive metal-plating technique that can easily be
performed on a laboratory bench top and in aqueous solutions (Supriya and Claus 2004;
Guan, Chen et al. 2005). Unlike CVD and PVD, it doesn’t require any adhesion layer and

can be used to tune the properties of arrays or ﬁlms of metals with nanometer-scale

127

precision on virtually any surface, including glass, silicon wafers, and mechanically
ﬂexible plastics. Due to their high surface area to volume ratio, porous substrates are
generally used as electrodes in biocatalytic reactors and biofuel cells. CVD and PVD are
expensive and poorly suited to uniformly coating the internal pore surfaces of porous
substrates. However, electroless deposition, because of capillary action offers the

potential for both coating the internal pores and low processing costs.

Arrays of BLMs on PEMs may be ideal for applications involving membrane
proteins that require either a cofactor or mediator for their activity. The cofactors and
mediators can be attached or entrapped near the membrane protein using the ﬁmctional
groups on the polyelectrolyte. However, BLMs on PEMs have low impedance; as a result
they cannot be used for applications requiring high impedance such as ion channels based
studies. On the other hand, tBLMs formed using tethers like PEG have been shown in
literature to have high impedance (Raguse, Braach-Maksvytis et al. 1998). Therefore,

approaches that can help generate arrays of such tBLMs are of scientiﬁc importance.

This chapter presents (Kohli, Hassler et al. 2006) an economical and versatile
approach that can easily be extended to fabricate functional and nanostructured
biomimetic interfaces on a wide variety of surfaces including electrodes of biocatalytic
reactors, thus increasing the repertoire of tBLM based applications. The approach entails
combining electroless metal deposition, and molecular self assembly, to sequentially
deposit a conductive gold layer and a tBLM on a substrate. The properties of the gold
ﬁlm were characterized using ﬁeld-effect scanning electron microscopy (FE-SEM),
energy dispersive spectroscopy (EDS), and atomic force microscopy (AFM). The
ﬂuidity of the tBLM was measured using ﬂuorescence recovery after pattern

128

photobleaching (FRAPP). The suitability of the interface to maintain membrane proteins
or other biomolecules in a ﬁmctional conformation was shown using the ionophore
valinomycin, and a segment of neuropathy target esterase (NT E) containing its catalytic
domain (NEST). To further demonstrate the versatility'of this approach, microcontact
printing was combined with electroless gold deposition and layer by layer to deposit
arrays of tBLMs. The resulting arrays were characterized using atomic force microscopy
(AFM), ﬁeld-emission scanning electron microscopy, ﬂuoresence microscopy, and

energy dispersive spectroscopy (EDS).

6.3 EXPERIMENTAL SECTION

6.3.1 Materials

1,2-dipalmitoyl-sn-glycero-3-phosphoethanolarnine-N—[3-(2-pyridyldithio) propionate]
(DGP, Figure 6.1a), 1,2-dioleoyl—sn-glycero-3-phosphocholine (DOPC), 1,2-di-0-
phytanyl-sn-glycero-3-phosphoethanolamine (DPGP, Figure 6.1b) and 1,2-dioleoyl-sn-
glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (NBD-PE, Figure
6.10) were purchased from Avanti Polar Lipids (Alabaster, AL). Cystamine, 3-
mercaptopropyltrimethoxy silane (MPS), gold chloride trihydrate (HAuCl4'3H20),
hydroxylamine, dithiothreitol (DTT), ﬂuorosilanes and sodium citrate dihydrate were
obtained from Sigma (St. Louis, MO). Poly (ethylene glycol) PEG-silane 2000 was
obtained from Nektar Therapeutics (San Carlos, CA). Sylgard 184 silicone elastomer kit
(Dow Corning, Midland, MI) was used to prepare the poly(dimethylsiloxane) (PDMS)

stamps for uCP. Deionized (DI) water was supplied by a Nanopure-UV four-stage

129

puriﬁer (Barnstead International, Dubuque, IA); the puriﬁer was equipped with a UV

source and a ﬁnal 0.2 pm ﬁlter. All aqueous solutions were prepared using DI water.

6.3.2 Preparation of stamps

The PDMS stamps were made by pouring a 10:1 (w/w) solution of elastomer and
initiator over a pretreated silicon master, which acts as a mold, to allow the surface
morphology of the stamp to form a negative replica of the master. The silicon master was
pretreated with ﬂuorosilanes to facilitate the removal of the PDMS stamps. The mixture
was allowed to cure overnight at 60 °C and then peeled off. The masters were prepared in
the Keck microfabrication facility at MSU and consisted of features (parallel lines and

circles) from 1 to 20 um.

6.3.3 Synthesis of colloidal gold particles

106 mL of 2.2 mM aqueous sodium citrate was rapidly boiled in a ﬂask, and 1 mL of
24.3 mM HAuC14’3H20 was rapidly added under vigorous stirring. The solution was
boiled for 15 min, cooled to room temperature, and stored at 4 °C.(Supriya and Claus

2004)

6.3.4 Formation of gold film and patterns by electroless deposition

The gold ﬁlms were fabricated as illustrated in Figure 6.2 using a variation of the
method reported by Frens (Frens 1973; Supriya and Claus 2004). Glass slides were
cleaned in piranha solution (7 parts by volume concentrated H2804, and 3 parts 30%
(v/v) H202) then washed with DI water and dried under nitrogen. The slides were dipped

into a 1% (v/v) solution of MPS in methanol for 10 min with stirring, rinsed with

130

methanol, and heated to 110 °C for 2 h to complete the silanization. The surface-modiﬁed
slides were then immersed in colloidal gold solution at room temperature (22 °C i 1°C)
for 2 h to allow thiol bonds to form between gold nanoparticles and sulfur groups. The
slides were then rinsed with DI water and immersed in 200 mL of aqueous solution
containing 2.6 mg of hydroxylamine and 20 mg of HAuCl4°3HzO under constant agitation
for 30 min. This step, called seeding, resulted in chemical reduction of gold ions onto the
immobilized gold nanoparticles. The seeding process was repeated three times to achieve

a continuous, conductive ﬁlm.

To obtain gold patterns (Figure 6.3), a PDMS stamp was dipped into 1 mM
solution of MPS in ethanol for 5 min. The stamp was washed with absolute ethanol, dried
under nitrogen, and brought into contact with a glass slide. The stamp was removed after
2 min, and the resulting MPS patterned slide was washed with ethanol to remove loosely
bound MPS. The slide was heated to 110°C for 2 h and immersed in colloidal gold
solution. The patterned slide was then sonicated in DI water for 10 min to remove loosely
bound gold particles from the background of the glass slides. The gold-particle-patterned
slide was then exposed to the seeding solution for 30 min three times to create gold

patterns.

6.3.5 Preparation of liposomes for ﬂuorescence and electrochemical measurements

For ﬂuorescence experiments, small unilamellar liposomes (~ 20 i 5 nm) were prepared
by adding 20 mg of DPGP and 0.2 mg of NBD-PE in 1 mL of chloroform. This mixture
was then dried under nitrogen, making sure the lipid formed a thin cake-like ﬁlm on the

walls of the test tube. The residual chloroform was removed under high vacuum. The

131

lipids were then reconstituted in 20 mL of 0.1 M aqueous NaCl solution, and the resulting
liposome solution was sonicated with a Bransonic bath sonicator (Branson Ultrasonics
Corporation, Danbury, CT) until it became clear. Liposomes were prepared in the same
way for electrochemical experiments, except no NBD-PE was added. The liposome size
was determined using Coulter N4 MD Particle Size Analyzer (Coulter Electronics Inc.,

Hialeah, FL)

6.3.6 Formation of tBLMs containing valinomycin

The inner leaﬂet of the tBLM was formed by dipping a ﬁeshly prepared electrolessly
deposited gold slide into an ethanolic solution of 1 mM DGP, 1 mM DPGP, and 500 uM
cystamine for 2 h. The slide was washed with absolute ethanol and dried under nitrogen.
The modiﬁed slide was dipped in liposome solution (1 mg/mL) for 45 min and washed

with 0.1 M aqueous NaCl solution to obtain the outer leaﬂet.

An ethanolic stock solution of valinomycin (2 mg/mL) was diluted in the aqueous
solution to obtain a valinomycin concentration of 5 x 10'7 M and allowed to equilibrate

for 1 h at room temperature. No change in impedance was observed on valinomycin
insertion in this aqueous solution (0.1 M NaCl). However, the impedance dropped when
this solution was replaced with another solution containing 0.05 M NaCl and 0.05 M
KCl, providing evidence of the insertion and activity of this potassium-selective

ionophore into the tBLM.

6.3.7 Formation of NEST-containing liposomes (N EST-DOPC liposomes)

NEST was expressed and puriﬁed according to published procedures (Atkins and Glynn
2000). For incorporation into liposomes, NEST solution (0.1 mg/mL) in PEN buffer (50

132

mM phosphate buffer, 0.5 mM EDTA and 300 mM NaCl) containing 0.3% (w/v) CHAPS
was mixed with DOPC (10 mg/mL in 9%(w/v) CHAPS) in 1:4 ratio (protein wtzDOPC
wt) and dialyzed against 500 volumes of PEN buffer containing 1 mM DTT overnight.

(Atkins and Glynn 2000)

6.3.8 Preparation of phenyl valerate micellar suspension

To prepare phenyl valerate solution, 15 mg of phenyl valerate was dissolved in 1 ml of
dimethylforrnamide (DMF), and 15 m1 of water containing 0.03% Triton was added

slowly under stirring.

6.3.9 Formation and characterization of NEST-containing tBLMs

The inner leaﬂet of the lipid bilayer was formed as described above. The NEST-DOPC
liposome solution was pumped across the inner leaﬂet at a rate of 0.34 mL/min for 12
min using a syringe pump. Flow was then halted, and adsorption was allowed to continue
for approximately 1 h. Five to six ﬂow cell volumes of PEN buffer were then passed
through the system to remove the liposomes in the bulk solution. Further details

regarding the setup and design of the ﬂow cell can be found in a previous publication

(Gajraj and Ofoli 2000).

To measure NEST activity in the tBLM, the PEN buffer in the ﬂow cell was
replaced with a micellar suspension of phenyl valerate and allowed to react with the
NEST-containing tBLM for 20 min. The resulting suspension was collected, and the
amount of phenol produced was measured according to published procedures (Kayyali,
Moore et al. 1991; Atkins and Glynn 2000). Brieﬂy, 400/LL of resulting suspension was

mixed with 400 uL of 1.23 mM 4-arninoantipyrine containing 3.8 mg of sodium dodecyl

133

sulfate and allowed to react for 5 min. To develop color, 200 uL of 12.1 mM potassium
ferricyanide solution was then added to this solution, and after 5 min the absorbance (A)
at 486 nm was measured. The spectrophotometer was blanked using suspension obtained
in the same way, except the phenyl valerate suspension was treated with a tBLM
containing no NEST protein. The amount of phenol produced was determined ﬁ'om the
absorbance using a calibration plot depicting A436 as a function of known phenol

concentrations.

6.3.10 Arrays of lipid bilayers

To obtain arrays of tBLMs, a gold-patterned glass slide was dipped in 1 mM solution of
PEG-silane 2000 in absolute ethanol for 20 min, rinsed with absolute ethanol, and heated
at 120 °C for 2 h to bind PEG-silane to the exposed glass background of the array. The
bottom leaﬂet of the tBLM was deposited on the gold features as described above. The
top leaﬂet was then added by dipping the modiﬁed slide in liposome solution for 45 min
and Washing with 0.1 M NaCl solution. Because the PEG-background region resisted

liposome adsorption, liposomes adsorbed only on the modiﬁed gold patterned region.

6.3.11 Electrochemical impedance spectroscopy (EIS) measurements

Impedance measurements were done using an electrochemical analyzer (CH 650A, CH
Instruments, Austin, TX). The impedance spectrum was obtained by sweeping an applied
potential of 50 mV from 10'1 Hz to 104 Hz, superimposed on a DC offset of 0 V. A three-
electrode conﬁguration was used with a Ag/AgCl reference electrode, a platinum counter

electrode, and a modiﬁed gold working electrode having an area of 0.16 cmz.

134

Commercially available software (Z-view, Scribner Associates, Southern Pines, NC) was

used to ﬁt electrical circuit models to the impedance spectra.

6.3.12 Fluorescence recovery after pattern photobleaching (FRAPP)

Liposomes were adsorbed on the electrolessly deposited gold substrates by
causing the liposome solution to ﬂow over them at a rate of 0.34 mL/min for 10-12 min
using a syringe pump. The infusion was then halted, and liposome adsorption was
allowed to continue for approximately 1 h. The ﬂow cell (Gajraj and Ofoli 2000) was
subsequently ﬂushed with four to ﬁve ﬂow cell volumes of 0.1 M NaCl to remove the
remaining liposomes in the bulk solution and FRAPP measurements were performed. All
experiments were done at room temperature, which is above the phase transition

temperature of DPGP and DOPC (Albrecht, Johnston et al. 1982).

6.3.13 Other measurements

All ﬂuorescence images were obtained with a Nikon Eclipse E 400 microscope (Nikon,
Melville, NY) using a ﬁlter cube (Ex: 465-495/DM: SOS/Em: 515-555). Field-emission
scanning electron microscopy (FE-SEM) images and energy dispersive spectroscopy
(EDS) images were obtained using a JEOL 6400 V scanning electron microscope (JEOL
USA, Inc., Peabody, MA) equipped with a Noran EDXS detector. Atomic force
microscopy (AF M) images were obtained with a Nanoscope IV multimode scope (Digital
Instruments, Santa Barbara, CA) using etched silicon probes in tapping mode. The
thickness of the patterned arrays was determined using cross-sectional analysis of the

AFM images.

135

6.4 RESULTS AND DISCUSSIONS

6.4.1 Formation of gold ﬁlms

Figure 6.4a shows a FE-SEM image of a monolayer of gold nanoparticles on a glass
slide. The particles were initially not in contact, so the interface was not conductive.
Grabar et al. studied adsorption of gold particles on MPS-coated glass slides and
concluded that the particle coverage is controlled initially by difﬁision, and at later times

by interparticle repulsion (Grabar, Smith et al. 1996). Colloidal gold particles derived

from [AuCl4]' have a negative charge resulting from strongly adsorbed Cl' and/or

[AuClz]' produced by incomplete reduction of [AuCl4]'. This interparticle repulsion

prevents the close packing of colloidal gold on the surface.

However, the gold coverage was increased by the seeding step, as shown in
Figure 6.2 and Figure 6.4(b-d). The rate of Au3+ reduction on the immobilized gold

particles greatly exceeds that in the seeding solution; thus, little new nucleation occurs in
the solution, and virtually all of the reduced Au is used to grow the gold
particles.(Supriya and Claus 2004) The FE-SEM images in Figure 6.4(b-d) show the
growth of gold ﬁlm on a glass slide following each successive seeding step. The ﬁnal
gold ﬁlm had a root mean square (RMS) roughness of around 10 nm (data not shown).
For comparison, a smooth glass slide, a frosted glass slide, and a gold-sputtered silicon
wafer have RMS roughnesses of about. 0.6, 14, and 4 nm, respectively (Tristram-Nagle,

Petrache et al. 1998; Naumann, Schiller et al. 2003).

136

6.4.2 Functional tBLMs containing valinomycin

Figure 6.5a depicts the spatial organization of the molecules that make up the tBLM.
The mobile lipid, DPGP, (Figure 6.1b), forms the bulk of the BLM. The hydrophilic
portion of the reservoir lipid DGP, Figure 6.1a, establishes the ionic reservoir between
the BLM and the gold surface. The hydrophobic portion of the reservoir lipid embeds in
the BLM, tethering it to the gold surface. The bulky thiobenzyl moiety of the tethering
lipids not only binds the lipid to the gold, but also increases the lateral spacing between
adjacent reservoir lipid molecules. Cystamine competes with the thiobenzyl group for
surface binding sites on the gold, and thus it provides a means to further control the

spacing between the reservoir lipids.

Figure 6.5b (curve i and ii), shows Bode plots of tBLM in the absence and

presence of valinomycin, respectively, in aqueous solution containing 50 mM KCl/SO
mM NaCl. BLMs provide an excellent barrier to ions. Introduction of the K+-selective
ionophore valinomycin into BLM imparts a valinomycin-mediated transmembrane K4r

ﬂux, making it possible to conﬁrm the functionality of valinomycin-containing BLM by

measuring differential impedance to potassium and sodium ions.

Figure 6.50 depicts the electrical analogue model generally used to describe tBLM
systems. In this model, the membrane capacitance (Cm) is bypassed by a variable ohmic
resistance (Rm), which represents the ionophore-mediated ion ﬂux and any background
ion leakage. The membrane capacitance is in series with an effective interfacial

capacitance, Cd], which includes the Helmholtz and diffuse capacitances. The model

137

parameters calculated by Z-view software to ﬁt the data in Figure 6.5b (curve ii) were Cm

= 0.59 uF/cmz, Cd] = 3.9 uF/cmz, and Rm: 46 K0 cmz.

The Cd] value in the assembled bilayer system was found to be smaller than the

published interfacial capacitance of a gold surface (~ 20-30 uF/cmz), possibly because of

the presence of thiobenzyl groups and cystamine molecules partially blocking the
surface, or because the tether organizes the water in the reservoir region such that the

dielectric constant within the reservoir region is lower than that of the bulk aqueous

solution.(Raguse, Braach-Maksvytis et al. 1998) The value of Cm (0.59 uF/cmz)

compared well with the values generally reported for BLM (~O.5 uF/cmz). Additional

evidence for the formation of a functional BLM on electrolessly deposited gold ﬁlms is
provided by the decrease in Rm upon the insertion of valinomycin (Figure 6.5b, curve ii),
as indicated by the change in slope at around 10 Hz instead of at around 5 Hz (Figure
6.5b, curve i) for a lipid bilayer containing no ion channel. The impedance increased
again (data not shown) when the aqueous solution (50 mM KCl/SO mM NaCl) was
replaced with a solution (100 mM NaCl) containing no KCl. These results provide strong

evidence that the valinomycin is embedded in an active conformation in an intact tBLM.

6.4.3 Functional BLM containing NEST protein

The phenyl valerate assay was used to conﬁrm whether tBLMs could immobilize NEST
in a functional conformation. Incubation of phenyl valerate with NEST-containing

BLMs on gold resulted in the production of about 2 :l: 0.19 nmol/min of phenol over an

area of 1 cm2. Incubation of phenyl valerate with NEST-DOPC liposomes in solution

138

resulted in the production of 40 :t 1.2 nmol/min of phenol per pg of NEST protein. This
result suggests the immobilization of approximately 50 ng/cm2 of active NEST in

tBLMs. To our knowledge, this is the ﬁrst time NEST has been immobilized in an active

conformation on a lipid bilayer.

6.4.4 BLM ﬂuidity measurement using FRAPP

Figure 6.6 shows the ﬂuorescence recovery curve for a DGP—DPGP-cystamine modiﬁed
gold slide contacted with ﬂuorescently labeled liposome solution. Fluorescence recovery
was adequately described by a single mobile species model described in Chapter 5, which
assumes only one population of diffusive ﬂuorophores in addition to an immobile

fraction. Equation 5.6 was ﬁtted to the FRAPP data, and average best-ﬁt values of D
(0.48 x 10'8 :1: 0.055 x 10'8 cmz/sec) and m (0.87 i 0.096) were obtained. These values

compare well with published values for tBLM. (Munro and Frank 2004)

These results provide strong evidence that the preponderance of the surface is
covered by a tBLM. If the surface were covered primarily by unruptured liposomes,
extremely low D and m values would be expected, because ﬂuorescently tagged lipids

would not transfer freely between adjacent liposomes.

6.4.5 Fabrication and characterization of gold patterns

Figures 6.7a and 6.7b show circular and line gold patterns, respectively, on a glass slide.
The lighter features show regions that were in contact with the stamp and are thus
covered with gold. The well-deﬁned features of the patterns and the clean background

suggest negligible adsorption of gold particles on the background glass slide.

139

EDS was also used to conﬁrm selective gold deposition on the patterned regions.
In this technique, the area of interest is bombarded with an electron beam of high energy
(~10 KeV), causing release of X-rays, the wavelengths of which are characteristic of the
elements in that region. Figures 6.8a and 6.8b show that Region 1 (MPS patterned region)
has a gold peak, but Region 2 (background glass slide) does not. These results conﬁrm
that gold particles adsorb selectively on the MPS-pattemed region. Because the
penetration depth of the electron beam is around 10 um, X-rays from the underlying glass
substrate are also collected by the detector. As a result, peak related to Si can be seen in
both ﬁgures. Peaks due to other elements generally present in the glass, such as 0, Ca
and A1, are not observed, because X-rays associated with these elements lie outside the.

energy range shown.

Figure 6.9a shows an AFM image of gold patterns. The light areas are deposited
gold regions, while dark areas are the underlying substrate. Figures 6% and 6.9c show
the 3-D image and topographical image, respectively, of the resulting patterns. The gold

regions had an average height of 55 i 7 nm and average RMS roughness of 10 nm.

6.4.6 Fabrication of tBLM arrays

PEG coatings have been shown to resist the adsorption of cells as well as
macromolecules, such as proteins and polyelectrolytes, through a mechanism that is
believed to stem either from steric exclusions between the biomolecule and the PEG
chain (Harris 1992) or from long-range electrostatic repulsions (Feldman, Hahner et al.

1999)

140

Figure 6.10 shows the ﬂuorescence image of tBLM line patterns. The sharp
contrast and clean boundaries between the ﬂuorescent line features and the
nonﬂuorescent background regions, as expected, indicate that ﬂuorescent liposomes
bound strongly to the tethering layer (gold patterned region) but not the PEG-coated

background region.

While the above-mentioned studies were done on electrolessly deposited gold
ﬁlms on a glass slide, we were also successful in forming electrolessly deposited gold
ﬁlms on mechanically ﬂexible substrates, such as plastics, using negatively charged
polyelectrolytes and amine-terminated positively charged polyelectrolytes (data not

shown).

6.4.7 Signiﬁcance of results and potential applications

Collectively, the results of this study represent an important milestone toward
development of products and processes based on the activities of membrane proteins or
other biomolecules. The FE-SEM, EDS and AFM studies conﬁrmed electroless
deposition of gold on an inexpensive substrate (e. g., glass) and showed that the thiclmess
of the gold layer can be controlled. The FRAPP and E18 results showed that silanization,
electroless gold deposition, and molecular self-assembly of lipids can be combined to
form tBLM on the electrolessly deposited gold. The results with valinomycin and NEST
indicated that an ionophore and a fragment of a membrane protein can be bound to the
tBLM in an active conformation, so that the activities of these biomolecules can be
expressed and measured. The array results indicated that discrete units of the tBLM can

be deposited as a high-density microarray that could potentially be used for high-

141

throughput studies of membrane protein or ionophore function. Ongoing research
involves applying the interface described here in devices to characterize the functional
properties of membrane proteins and cellular interactions with membrane proteins, drug

screening, biosensors, biocatalytic reactors, and biological fuel cells.

6.5 CONCLUSIONS

A versatile approach has been developed that can easily be extended to fabricate
functional and nanostructured biomimetic interfaces on a wide variety of surfaces. The
approach entails sequentially depositing onto the surface a silane layer, a conductive gold
layer, a hydrophilic reservoir, and a tBLM. The diffusion coefﬁcient of the tBLMs on
electrolessly deposited gold was comparable to that on vapor-deposited gold. Two
different classes of membrane biomolecules, the ionophore valinomycin, and the esterase
NEST, exhibited their activities when embedded in the tBLM, indicating that the
interface can incorporate membrane biomolecules in an active conformation. Because the
approach deposits both a conductive gold electrode and a tBLM in close proximity, it can
be used for bioelectronic applications, in which the biomolecule’s activity is coupled to
an electrical signal. Such applications include devices to characterize the functional
properties of membrane proteins, biosensors, biocatalytic reactors, and biological fuel
cells. Because the entire interface can be fabricated using layer-by-layer assembly from
different solutions, this approach may allow biomimetic interfaces to be assembled inside
microﬂuidic channels, thus enabling the production of high-density biosensor arrays for

hi gh-throughput applications.

142

6.6 RECOMMENDATIONS FOR FUTURE WORK

It will be interesting to deposit functional tBLMs on the internal pores of a porous
substrate such as alumina. This may be achieved by coating the alumina pores using a
low molecular weight, amine terminated polyelectrolyte. Electroless deposition may then
be used to deposit gold inside these pores by exploiting the afﬁnity of gold nanoparticles
to primary amines. tBLMs may then be formed on the electrolessly deposited gold using
the procedure outlined in this Chapter. These tBLMs may be made functional by
incorporating membrane proteins having biocatalytic potential such as cytochrome c
oxidase. The properties of the modiﬁed alumina substrate, as an electrode in a biofuel cell

or a biocatalytic reactor, may then be evaluated.

143

NO

Figure 6.1: Structure of various lipid molecules: (a) DGP (reservoir lipid); (b) DPGP
(mobile lipid); (c) NBD-PE.

144

SH SH SH SH

 

 

 

 

 

OH OH 0H 0H 0H Silanization g
[l l l l l] , /s|.i\o,sli\o,sli\O/sli\
O O O (r
Plasma treated glass slide l i i l J
(a) (b)
Au colloidal
solution

        
  

 

 

 

 

 

 

. . . . in . .
/Sl'\O/Sl'\O/SI'\O/Sl'\ : seed 9 /Sl'\O/S|'\
O O O O O O
l | l | l l
r l I
(d)

Figure 6.2: Schematic representation of the process for the deposition of Au ﬁhns: (a)
plasma-treated glass slide; (b) slide after silanization with MPS; (0) modiﬁed slide after
dipping in Au particle solution; (d) formation of conductive Au ﬁhns after seeding step

145

 

 

“—2
‘8 g
5... i
/
a *2
:8
—Q’.
/
—-Q’.
/
*2
W i; "—9
—\2 19A
if“ it i
__ _ _/\_
29A. “’ i
s _: «—-E- a é «—
\ 8 \
a?” g- g jib/T3?
_ °- _ _/\_
fa“ f" .10
—s —2
l—s -—s

 

 

 

 

(e)

 

 

 

 

Figure 6.3: Schematic representation of the process for the fabrication of Au patterns: (a)
stamping of MPS on a glass slide; (b) MPS patterns on a glass slide; (c) MPS-patterned
modiﬁed glass slide after heating; (d) MPS-pattemed slide aﬁer dipping in Au colloid

solution; (e) formation of Au patterns after seeding step.

146

 

(c) ((0

Figure 6.4: FE-SEM images showing the growth of gold on glass slides: (a) after dipping
in colloidal solution; (b) seeding once; (c) seeding twice; ((1) seeding three times.

147

Valinomycin

   

Mobile

lipid \

0.0.06

III. IV

   

T

F‘FHMH

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L i“ Reservoir or
- ... . _ _ .l u .. tethering IIpId
Spacer
molecule \ Thiobenzyl group
Gold electrode
(a)
6
5 r
«A C
E m
8 l |
f3; 4 Rs 1 I Cdl
3 34 —l l—
2 . . . .
-1 0 1 2 3 4 Rm
Log(Frequency)

(b) (c)

Figure 6.5: a) Representation of a tethered bilayer membrane containing an ionophore
(valinomycin). (b) Electrochemical spectra in a 50 mM KCl / 50 mM NaCl aqueous
solution of a tethered lipid bilayer (curve i), and a valinomycin-containing lipid bilayer
(curve ii). (c) Equivalent circuit of a bilayer.

148

 

 

 

0.9 —* - , - ~ ., ~ I..- ~ ~ ~ ~~ ~ - ~~~~~~ 9+
3 0.85 — , 3.. . ;, l
5 0 e. o” “z. . :. 0:: : .£;3.<.’.. ’&:’O 0‘ 1‘
a o 8 __ _ 0'0 O‘: ‘ .00 ‘ 1’ . ...~.:' 0.. ‘00 j
2 . ’. e. . “ .fo . 9 ‘

o 0' .0
2 .
"- l
3 I
.5 I
E l
5 0.65 - l
l
0.6 r T I I T I I

0 20 40 60 80 100 120 140
Time (minutes)

Figure 6.6: Fluorescence recovery after pattern photobleaching (FRAPP) proﬁle of lipid
bilayers on electrolessly deposited gold ﬁhns. The dots represent post-bleach
ﬂuorescence intensity normalized against the corresponding pre-bleach ﬂuorescence
value. The solid line in the ﬁgure represents the ﬁt of the recovery data.

149

Gold
patterns

   

(a) (b)

Figure 6.7: Optical microscopy images of gold: (a) circular patterns on a glass slide; (b)
line patterns on a glass slide.

150

 

 

 

30000

27000. 51 30000
27000 51
24000. 24000
c 21000. c 21000
0 18000. c 18000
: 15000. U 15000
n
t 12000‘ t 12000
S 9000. 5 9000
6000. 6000
3000. nu 3000
n g” n
1.700 2.000 2.300 2.600 2.900 .700 2.000 2.300 2.600 2.900
keV keV

 

Figure 6.8: EDS image showing (a) X-rays collected from Region 1 of a SEM image; (b)
X-rays collected from Region 2 of a SEM image. There is a peak corresponding to Au in
X-rays collected from Region 1 but not in X-rays from Region 2, conﬁrming the selective
deposition of Au on the patterned surface.

151

 

25 pm
(a) (b)

55 n m (U, Mlh- “’l-"M {Vh‘flﬂ1I’i-N".“ﬂl
|
I'

kmmrmmi I _

l r

MK}

20 ,um

(C)

Figure 6.9: (a) Topographical AFM image of gold patterns; (b) 3-D image of the patterns;
(0) pattern height on glass slide. All these images were obtained in air using tapping
mode.

152

 

Figure 6.10: Fluorescence image showing line patterns of a lipid bilayer consisting of
NBD-PE as the ﬂuorophores. The image was obtained using a ﬁlter cube (Ex: 465-
495/DM: SOS/Em: 515-555).

153

7 INTACT TRANSFER OF LAYERED BIONANOCOMPOSITE ARRAYS

7.1 ABSTRACT

Microarrays containing multiple, nanostructured layers of biological materials would
enable high-throughput screening of drug candidates, investigation of protein-mediated
cell adhesion, and fabrication of novel biosensors. In this paper, we present an approach
that allows high-quality microarrays of layered, bionanocomposite ﬁlms to be deposited.
The approach uses layer-by—layer self-assembly to preestablish a multilayered structure
on an elastomeric stamp, and then uses microcontact printing to transfer the 3-D structure
intact to the target surface. This approach extends the method previously used for intact
transfer of polyelectrolyte multilayer patterns to include amphiphilic biomolecules, such
as proteins and dendrimers. The approach overcomes a problem encountered when using
microcontact printing to establish a pattern on the target surface and then building
sequential layers on the pattern via layer-by-layer self-assembly. Amphiphilic molecules
tend to adsorb both to the patterned features as well as the underlying substrate, resulting
in low—quality patterns. By circumventing this problem, this research signiﬁcantly
extends the range of surfaces and layering constituents that can be used to fabricate 3D,

patterned, bionanocomposite structures.

7.2 INTRODUCTION

Elastomeric patterning methods, such as microcontact printing (uCP) (Kumar
and Whitesides 1993), microcontact molding (uCM) (Kim and Xia 1995), and micro
ﬂuidics (Kim , Park et al. 2001), allow a two-dimensional pattern to be fabricated on a

surface. In pICP, an elastomeric stamp having the desired topographic pattern is coated

154

with a molecular ink and then brought into contact with the surface. Removal of the
stamp leaves behind a thin layer of ink having the same pattern as the stamp. Based on
the chemical contrast between the inked features and the ink-ﬂee background, additional
layers can then be selectively deposited on either the features or background by directed
self-assembly (Lee, Zheng et al. 2002; Kidambi, Chan et al. 2004).

Biological activity can be imparted to these structures by incorporating protein or
other biomolecules during the LBL assembly process. Protein and DNA patternings using
photolithography or electron beam lithography have been reported (Michel, Lussi et al.
2002; Zhang, Tanii et al. 2004). In these studies, patterned interfaces were created having
discrete regions of reactive terminal groups, onto which proteins or DNA molecules were

subsequently adsorbed.

The biological activity can further be augmented by co-immobilization of other
macromolecules, like dendrimers or polyelectrolyte multilayers (PEMs) both of which
allow the local environment of the protein to be customized. Dendrimers are radially
symmetrical polymeric molecules that are grown by sequential addition of branched
monomers to the outer shell. With each new addition (generation), the molecular weight
of the dendrimer and the density of its terminal functional groups increases substantially
(Tomalia, Naylor et al. 1990; Dvornic and Tomalia 1996; Tomalia and Dvornic 1996;
Kohli, Dvornic et al. 2004). Dendrimers can have a highly ﬁmctionalized outer surface,
making them promising candidates for applications in drug delivery (Boas and Heegaard
2004), imaging (Balzani, Ceroni et al. 2003) and sensing (Ghosh and Banthia 2004).
Dendrimers can also serve as functional frames to encapsulate small molecules needed by

the proteins, including enzyme cofactors and electron mediators. PEMs can be used to

155

immobilize hydrophobic membrane proteins onto hydrophilic substrates, entrap ionic or
polar small molecules needed by the proteins, and act as an ion-selective barrier to screen

out interfering molecules (Lvov, Ariga et al. 1995).

In some cases, layer-by-layer assembly of 3-D structures onto patterns deposited
by p.CP is hindered by the lack of sufﬁcient chemical contrast between the features and
background (i.e., the bare surface), making it difﬁcult to deposit additional layers cleanly
on only the feature or the background. This problem can be particularly challenging
when a layer of amphiphilic molecules (e.g., proteins and dendrimers) is to be deposited,
because amphiphilic molecules can adsorb both to hydrophilic and hydrophobic surfaces.
Polyelectrolyte multilayers also present challenges, because the alternating anionic and
cationic layers could adsorb to either positive or negative charges on the background

surface.

In this chapter, we present an approach that overcomes the abovementioned
difﬁculties in establishing well-deﬁned, 3D bionanocomposite patterns (Kohli, Worden et
al. 2005). The approach (intact transfer printing, ITP) entails combining spin self-
assembly (Kohli, Dvornic et al. 2004) and layer-by-layer self-assembly (Lvov, Ariga et
al. 1995; Kidambi, Chan et al. 2004) to pre-establish a multilayered structure on an
elastomeric stamp, and then using uCP to transfer the 3-D structure intact to the target
surface. While uCP was recently used to transfer preformed PEMs to a substrate (Park
and Hammond 2004), this paper presents for the ﬁrst time conclusive evidence of
formation of bionanocomposite layered structures on a micropattemed stamp and

subsequent transfer of the structures intact to a target substrate.

156

As examples, in this paper we present the fabrication and characterization of
some 3-D nanostructured architectures (or structures) consisting of dehydrogenase
enzymes, polyamidoamine (PAMAM) dendrimers and PEMs, possible using this
approach. For the present study, generation four (G 4) PAMAM dendrimers and two
different dehydrogenase enzymes were used: secondary alcohol dehydrogenase and

sorbitol dehydrogenase (sDH).

This research is expected to signiﬁcantly extend the range of surfaces and
layering constituents that can be used to fabricate 3D, patterned bionanocomposite
structures. Such structures have a broad range of potential applications, including
fabricating protein-containing micro-arrays, studying mechanisms of protein-mediated
cell adhesion (MacBeath and Schreiber 2000; Berg, Yang et al. 2004), diagnosing disease
states (Fodor, Read et al. 1991), constructing biosensors, and investigating interactions

between proteins and other molecules.

7.3 EXPERIMENTAL SECTION

7.3.1 Materials

G4 amine-terminated PAMAM dendrimer (10 wt % solution in methanol, MW, 14,215)
sulfonated poly(styrene) (SPS) (MW ~70,000), sorbitol dehydrogenase (sDH),

poly(diallyldimethyl ammonium chloride) (PDAC) (MW ~100,000) were obtained from

Sigma (St. Louis, MO). Fluorescein isothiocyanate (FITC) and Alexa ﬂuor 568 were
obtained from Invitrogen (Carlsbad, CA). Secondary alcohol dehydrogenase (sADH) was

expressed and puriﬁed according to published procedures (Burdette and Zeikus 1994).

157

Sylgard 184 silicone elastomer kit (Dow Coming, Midland, MI) was used to prepare the
poly(dimethylsiloxane) (PDMS) stamps for uCP. Ultrapure water supplied by a
Nanopure-UV four-stage puriﬁer (Barnstead International, Dubuque, IA); the puriﬁer

was equipped with a UV source and a ﬁnal 0.2 pm ﬁlter.

7.3.2 Preparation of polyelectrolyte multilayers (PEMs)

A Carl Zeiss slide stainer equipped with a custom-designed ultrasonic bath was
connected to a computer to perform LBL assembly (Lee, Zheng et al. 2002; Lee,
Hammond et al. 2003; Kidambi, Chan et al. 2004). For PDAC/SPS multilayers, the
concentration of SPS and PDAC was 0.01 M and 0.02 M, respectively, as based on the
molecular repeat units. All polyelectrolyte solutions contained 0.1 M NaCl and were at a

pH of 7.0. PEMs were deposited on the glass slide using the procedure described earlier.

7.3.3 Fluorescent labeling of PAMAM dendrimers, sADH and sDH

FITC was predissolved in acetone and added to an aqueous solution of PAMAM
dendrimers. The resulting solution was allowed to stand overnight with occasional
stirring. The total amount of dye was adjusted to label just one amino group per PAMAM

molecule on average, assuming perfect reaction efﬁciency. The PAMAM solutions were
then dialyzed against pure water (pH, 7.4) using membrane tubing (Mw cutoff 2000) to

dialyze out unreacted FITC.

sADH was labeled using both FITC (green dye) and alexa ﬂuor (red dye). A 5.5
mg sample of sADH, in 2 mL Tris buffer was dialyzed against a 0.1 M sodium
bicarbonate solution, pH= 8.5, for 24 h. The bicarbonate solution was changed every 6 h

during the dialysis process. FITC (or alexa ﬂuor) was then added to the resulting protein

158

solution, in 1:1 mol ratio and continuously stirred for 10 h. The protein solution was then
dialyzed against phosphate buffer (pH 7.4) for 24 h, changing the buffer every 6 h, to

remove excess F ITC (or alexa ﬂuor).

Alexa ﬂuor was added to the sDH protein solution (2 mg/ml), in 1:1 mol ratio and
continuously stirred for 6 h. The protein solution was then dialyzed against phosphate

buffer (pH 7.4) for 24 h, changing the buffer every 6 h, to remove excess alexa ﬂuor.

7.3.4 Fabrication of 3-D structures or architectures

Figure 7.1 depicts the different 3-D architectures (or structures) that were fabricated and
studied. In Case 1, a PDMS stamp was dipped in a solution of FITC labeled sADH (1
mg/ml) or alexa ﬂuor labeled sDH solution (1 mg/mL) in 20 mM phosphate buffer (pH
7.4) for 2 h. The stamp was washed thoroughly with water and then either 20 or 40
PDAC/SPS multilayers were grown on it according to previously described procedure.
The modiﬁed PDMS stamp was washed with water, dried under nitrogen and brought
into contact with a glass slide coated with 10.5 PDAC/SPS bilayers. The stamp was
removed after 20 min and the resulting patterned substrate was washed thoroughly with
water to remove loosely bound molecules. In case 2, a PDMS stamp was dipped in a
solution of sADH (1 mg/mL) (either FITC labeled or unlabeled) in 20 mM phosphate
buffer (pH 7.4) for 2 h. The stamp was washed with water and dipped in 10 “M aqueous
solution of PAMAM dendrimers (pH 7.4) (FITC labeled or unlabeled) for 30 min. The
stamp was rinsed with water and 30 PDAC/SPS multilayers were grown on to it. The
modiﬁed PDMS stamp was washed with water, dried under nitrogen and brought into

contact with a glass slide coated with 10.5 PDAC/SPS bilayers. The stamp was removed

159

after 20 min and the resulting patterned substrate was washed thoroughly with water. In
case 3, a plasma cleaned PDMS stamp was dipped in FITC labeled sADH solution (1
mg/ml, pH 7.4 in phosphate buffer) for 2h. The stamp was rinsed with water and 1
PDAC/SPS bilayer was grown in it. The stamp was then washed with water and dipped in
alexa ﬂuor labeled sADH solution (1 mg/mL, pH 7.4 in phosphate buffer) for 2h. The
modiﬁed PDMS stamp was then washed with water, dried under nitrogen and brought
into contact with a glass slide coated with 10.5 PDAC/SPS bilayers. The stamp was
removed after 20 min and the resulting patterned substrate was washed thoroughly with

water.

7.3.5 Other measurements

All the ﬂuorescence images were obtained with the Nikon Eclipse E 400 microscope
(Nikon, Melville, NY) using two ﬁlter sets, one for FITC (Ex: 465—495/DM: SOS/Em:
515-555), and the other one for alexa ﬂuor (Ex: 510-560/DM: 565/Em: 590-690). AFM
images were obtained with a Nanoscope IV multimode scope (Digital Instruments, Santa
Barbara, CA). The microscope was equipped with tapping-mode etched silicon probes.
The thickness analysis of the patterns was determined using cross-sectional analysis of

the AFM images.
7.4 RESULTS AND DISCUSSIONS

Figure 7.1 depicts the procedure (ITP) used for fabricating different 3-D architectures or
structures. Brieﬂy, a multilayered ﬁlm consisting of proteins, polyelectrolytes or
dendrimers is assembled on top of an elastometric stamp using layer by layer assembly,

and then transferred intact to a substrate which is oppositely charged from the multilayer

160

ﬁlm’s top surface. The choice of the base layer is a critical element in ITP process. The
base layer should strongly and evenly adsorb on stamp; non uniform pattern transfer can
occur without a smooth base layer. However, the attractive interactions between the base
layer and stamp should be weak enough to allow easy detachment from the stamp during
printing. To achieve this balance, we exploited hydrophobic interactions between the
PDMS stamp and sADH (or sDH) base layer in the fabrication of all 3-D structures. The
adsorption of this base layer using non-electrostatic interactions provides a charged
surface for the assembly of other layers. The ITP process was not reproducible and non-

uniform when other macromolecules such as PDAC or SPS were used as base layers.

Figure 7.2a and 7.2b show the two dimensional and cross-sectional AFM images,
respectively, when a single layer of sADH was transferred. The light regions show the
deposited sADH patterns, while the dark regions are the underlying substrate. The sADH
patterns had an average height of 7 2t 2 nm. Figures 7.3a and 7.3b show the two-
dimensional and cross-sectional images, respectively, when a multilayer ﬁlm (see Figure
7.1, case 1) consisting of sADH base layer and 20 PDAC/SPS bilayers,
(sADH)1(PDAC/SPS)20, was transferred on a PEM coated glass substrate. The average
pattern height was approx. 67 :t 3 nm. On the other hand, the average pattern height (see
Figures 7.3c and 7.3d) was approx. 127 i 4 nm when a (sADH)1(PDAC/SPS)40
multilayer ﬁlm was transferred. On taking into account the contribution from sADH
layer (approx. 7 nm), the pattern height was found to linearly increase with the number of
PDAC/SPS bilayers and determined to be approximately 60 ﬂ: 3 nm for 20 PDAC/SPS

bilayers and 120 i 4 nm for 40 PDAC/SPS bilayers. From these values, the average

161

height of a pair of PDAC/SPS was determined to be 3 :l: 0.15 nm. This agrees well with

literature values of 3.4 nm for similar deposition conditions (Kidambi, Chan et al. 2004).

Fluorescence microscopy was used to check the efﬁciency and feasibility of the
multilayer ﬁlm transfer. Figure 7.4a shows the ﬂuorescence image of the circular patterns
obtained on the transfer of a multilayer ﬁlm consisting of a layer of ﬂuorescently labeled
sADH and 20 PDAC/SPS bilayers. As can be seen by the ﬂuorescent features of the
circular arrays and clean background regions, the multilayer ﬁlm transfer occurred only
in the patterned region and negligibly in the background (unpattemed) region. The
ﬂuorescence image of the stamp (Figure 7.4b) itself revealed little ﬂuorescence in the
circular region where the transferred ﬁlm existed. On the other hand, the background
region retained ﬂuorescence. This suggests that the transferred ﬁlm was complete, and
the multilayer ﬁlm remained completely intact during the transfer process. The ITP
process was reproducible over an area of around 1 cm x 1 cm and is possibly

reproducible over larger areas also.

The ability to make multilevel and multicomponent patterns consisting of
amphiphilic molecules such as proteins and PEMs has potential applications in
biocatalysis and also for the fabrication of practical devices. To demonstrate multilevel
and multicomponent patterning, two different multilayer composite ﬁlms were stamped
on a surface (Figure 7.5). One multilayer ﬁlm consisting of alexa ﬂuor labeled sDH and
20 PDAC/SPS bilayers was stamped ﬁrst on a PEM (10.5 PDAC/SPS bilayers) coated
surface. This was followed by the stamping of a second multilayer ﬁlm consisting of
FITC labeled sADH and 20 PDAC/SPS bilayers at an angle perpendicular to the ﬁrst
printed lines. Figure 7.6a shows the red ﬂuorescence from vertically printed

162

(sDH)1(PDAC/SPS)20 lines and Figure 7.6b shows green ﬂuorescence from horizontally

printed (sADH)1(PDAC/SPS)20 lines. Figure 7.60 shows a digitally combined image

showing both the red and green ﬂuorescence, suggesting the transfer of both multilayer
systems. The pattern height in the crossed region and uncrossed region using the AFM
cross-sectional analyses was found to be approximately 125-130 nm and 65-70 nm,
respectively. As expected, these results suggest the combination of multilayer ﬁlms at the

crossover regions and intact transfer at the uncrossed regions.

The possibility of multicomponent multilayer ﬁlm transfer was further
demonstrated by transferring a multilayer ﬁlm (see Figure 7.1, Case 2) consisting of
sADH, G4 PAMAM dendrimers and PDAC/SPS bilayers,(sADH)1(
PAMAM)1(PDAC/SPS)30, intact to a surface. Fluorescence microscopy was used to
establish the existence of alternating protein and dendrimer layers in the multilayered
ﬁlms after deposition. In one case (Figure 7.7a), only the protein was ﬂuorescently
labeled. In the other case (Figure 7.7b), only the dendrimer was labeled. Fluorescence
observed in both Figures conﬁrms the presence of both the protein and dendrimer layers.
AFM was then used to conﬁrm incorporation of PEM bilayers into the multilayered
ﬁlms. Figures 7.8a and 7.8b show the AFM images of a patterned ﬁlm containing
sequential layers of protein, dendrimers and PEMs (30 PDAC/SPS bilayers) on a PEM
coated substrate. Figures 7.80 and 7.8d show the AFM images of a patterned ﬁlm
containing sequential layers of protein and dendrimers. The height of patterns was
approximately 101 :l: 4 nm with PEMs, and approximately 9 :t 2 nm without PEMs.
Based on these data, the average height of each PDAC/SPS bilayer was estimated to be

about 3.07 i 0.14 nm, which is in agreement with published values.

163

Further proof of the potential advantages of ITP was demonstrated by stamping a
multilayer ﬁlm (Figure 7.1, Case 3) consisting of a PDAC/SPS bilayer sandwiched
between two sADH layers. Fluorescence microscopy was again used to conﬁrm the
presence of different protein layers. The base sADH layer was ﬂuorescently labeled with
a green dye FITC, and the top sADH layer was labeled with a red dye alexa Fluor. The
observance of both the green (Figure 7.9a) and red ﬂuorescence (Figure 7.9b), conﬁrms
the presence of both sADH layers. AF M was then used to conﬁrm the incorporation of
PDAC/SPS bilayer in between these sADH layers. Figure 7.90 shows the cross-sectional
topographical image of a patterned ﬁlm containing sequential layers of sADH and l
PDAC/SPS bilayer. The average height of pattern was 10 d: 1 nm. On taking into account
the contribution from sADH layer (approximately 7 nm), the height corresponding to
each PDAC/SPS bilayer can be estimated to be 3 i 1 nm, which is consistent with our

other results. Figure 7.9d shows the topographical image of a patterned ﬁlm consisting of
(sADH)1(PDAC/SPS)1(SADH)1. The average height of the patterns was 17 i 2 nm.

Based on these data, the height of the topmost sADH layer was estimated to be approx. 7

:thm.

We believe electrostatic interactions between the enzymes, dendrimers and PEMs
are responsible for stabilizing these different multilayered structures. Weak
polyelectrolytes change their conformation or charge density with pH.(Yoo, Shiratori et
al. 1998; Yang and Rubner 2002) Thus, the shape and stability of the resulting 3-D
structures formed with weak polyelectrolytes often vary with pH. To avoid such effects,
we used the strong polyelectrolytes SPS and PDAC, whose charge density is relatively

unaffected by pH.

164

Our novel approach, in which bionanocomposite arrays are preestablished on a
stamp and then transferred intact to the target substrate, is based on topographical
contrast between the feature and background regions of the pattern, rather than chemical
contrast. Thus, the new method offers signiﬁcant advantages over the conventional,
directed-self-assembly approach in cases when the chemical contrast is marginal or when
amphiphilic or zwitterionic molecules (e.g., proteins) are involved. In such cases,
adsorption is likely to occur on both the background and feature regions, leading to poor
resolution. This effect is clearly illustrated by the much cleaner patterns seen in Figure
7.7a, where the SADH-containing pattern was transferred intact, than in Figure 7.10,
where SADH was adsorbed from solution onto both the dendrimer features and the SPS
background.

This research signiﬁcantly extends the range of surfaces and layering
constituents that can be used to fabricate 3D, patterned, bionanocomposite structures.
Such structures have a broad range of potential applications, including fabricating
protein-containing microarrays for screening drug candidates, studying mechanisms of
protein-mediated cell adhesion (Mac Beath and Schreiber 2000; Berg, Yang et al. 2004)
diagnosing disease states (Fodor, Read et al. 1991) constructing biosensors, and

investigating interactions between proteins and other molecules

7.5 CONCLUSIONS

In this paper, different 3-D architectures have been fabricated and characterized
using a unique multilayer approach. This multilayer approach, in which
bionanocomposite arrays are preestablished on a stamp and then transferred intact to the

target substrate, is based on the topographical contrast, rather then chemical contrast.

165

Thus this method offers several advantages over conventional, directed self assembly
approach in cases when the chemical contrast is marginal or when amphiphilic or
zwitterionic (e.g., proteins). In such cases, adsorption is likely to occur on both the
background and feature regions, leading to poor resolution. This research, thus,
signiﬁcantly extends the range of interfaces and layering constituents that can be used to

fabricate 3-D bionanocomposite arrays.

166

PAMAM

Protein (sADH or sDH) d endrimers

—
'l'
l

I

W .. .. —"‘"-

(a) Case 1 Case 2 Case 3

PDAC/SPS

\

 

 

;/ Glass slide coated wrth PEMs

 

 

 

(b)

Bionanocomposite Arrays

/

 

 

 

 

 

(C)

Figure 7.1: Schematic representation of the procedure used for printing: (a) stamp coated
with sequential layers of proteins (sADH or sDH) and PEMs (PDAC/SPS bilayers) (Case
1); alternating layers of sADH and G4 PAMAM dendrimers on PEMs (case 2); and
PDAC/SPS bilayer sandwiched between sADH layers (case 3). (b) Glass slide coated
with PEMs (10.5 bilayers). (c) Patterned substrate

167

50pm

 

50pm

(a)

7nm I'll

50pm
(b)

Figure 7.2: (a) Two-dimensional AFM image of sADH patterns, on a PEM (10.5
PDAC/SPS bilayers) coated glass substrate. (b) Cross-sectional AFM image of sADH
patterns, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate.

168

35pm

 

35pm 35pm
67nm 127nm W
(b) 35m (d) 35pm

Figure 7.3: Two-dimensional and cross-sectional AFM images: (a—b) when a
(sADH)1(PDAC/SPS)20 multilayer ﬁlm was transferred to a PEM (10.5 PDAC/SPS
bilayers) coated glass substrate; (c-d) when a(sADI-I)103DAC/SPS)40 multilayer ﬁlm was
transferred to a PEM (10.5 PDAC/SPS bilayer) coated glass substrate.

169

 

(a) (b)

Figure 7.4: Fluorescence images of: (a) the circular patterns obtained on the transfer of a
multilayer ﬁlm consisting of a layer of ﬂuorescently labeled sADH and 20 PDAC/SPS
bilayers, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate; (b) the PDMS
stamp surface after printing.

170

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PDMS stamp coated with
sorbitol dehydrogenase
(sDH) and 20 PDAC/SPS
bilayer

Vertically printed
sDH patterns on
a PEM coated
glass slide

PDMS stamp coated with
secondary alcohol
dehydrogenase (sADH)
and 20 PDAC/SPS bilayer

Horizontally printed sADH
and vertically printed sDH
patterns

Figure 7.5: Schematic representation of the process used for multilevel and

multicomponent stamping

171

 

(a) (b)

    

20pm

©

Figure 7.6: (a) Red ﬂuorescence from vertically printed (Alexa ﬂuor labeled-
sDH)1(PDAC/SPS)20 lines on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate;
(b) Green ﬂuorescence from horizontally printed (FITC labeled-sADH)1(PDAC/SPS)20
lines, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate;(c) Digitally combined
ﬂuorescence image showing both red and green ﬂuorescence

172

   

(a) (b)

Figure 7.7: (a) Fluorescence image of the patterned ﬁlms of sADH, G4 PAMAM
dendrimers and PEMs (30 PDAC/SPS bilayers) with ﬂuorescently labeled sADH as the
topmost layer, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate. (b)
Fluorescence image of the line patterns of ﬂuorescently labeled dendrimers sandwiched
between patterned sADH layer and PEMs (30 PDAC/SPS), on a PEM (10. 5 PDAC/SPS
bilayers) coated glass substrate.

173

25pm 251"“

   

25pm 25pm
(a) (e)

100nm {1‘ "1

 

25pm 25pm
(b) (d)

Figure 7.8: (a-b) Two-dimensional and cross-sectional AFM images of a patterned ﬁlm
containing sequential layers of sADH, PAMAM dendrimers and PEMs (30 PDAC/SPS
bilayers), with sADH as the topmost layer, on a PEM (10.5 PDAC/SPS bilayers) coated
glass substrate. (c-d) Two-dimensional and cross-sectional AFM images of a patterned
ﬁlm containing sequential layers of sADH and PAMAM dendrimers with sADH as the
topmost layer, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate.

174

 

 

(C) (d)

Figure 7.9: (a—b) Fluorescence images of the patterned ﬁlms of a PDAC/SPS bilayer
sandwiched between FITC labeled sADH base layer and Alexa ﬂuor labeled sADH
topmost layer, on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate, (a) Green
ﬂuorescence emanating from FITC labeled sADH base layer, obtained using the
following ﬁlter set, Ex: 465-495/DM: SOS/Em: 515-555; (b) Red ﬂuorescence emanating
from Alexa ﬂuor labeled sADH topmost layer, obtained using the following ﬁlter set, Ex:
510-560/DM: 565/Em: 590-690. (0) Cross-sectional topographical image of a patterned
ﬁlm containing sequential layers of sADH and 1 PDAC/SPS bilayer, on a PEM (10.5
PDAC/SPS bilayers) coated glass substrate. (d) Topographical AFM image of a
patterned ﬁlm consisting of a PDAC/SPS bilayer sandwiched between two sADH layers,
on a PEM (10.5 PDAC/SPS bilayers) coated glass substrate.

175

 

Figure 7.10: Arrays of amphiphilic proteins obtained on patterned substrate using directed
self-assembly

176

8 NANOSTRUCTURED CROSS-LINKABLE MICROPATTERNS VIA

AMPHIPHILIC DENDRIMER STAMPING

8.1 ABSTRACT

Microcontact printing was used to deposit stable, nanostructured, amphiphilic and cross-
linkable patterns of PAMAMOS-DMOMS dendrimer multilayers onto silicon wafers,
glass, and polyelectrolyte multilayers. The effects of dendrimer ink concentration,
contact time, and inking method, on the thickness, uniformity and stability of the
resulting patterns were studied using optical microscopy, ﬂuorescence microscopy, AF M,
and contact-angle analysis. Dendrimer ﬁlm thickness was found to be controllable via

conditions used during spin-self assembly.

177

8.2 INTRODUCTION

Functional, 3-D nanostructures and microstructures on surfaces can serve
as excellent molecular templates for applications in optoelectronics and biotechnology
(Rudolph 1994; Blawas and Reichert 1998; Kidambi, Chan et al. 2004). Coupling stable,
amphiphilic ﬁlms with hydrophobic biological molecules can yield biomimetic interfaces
able to reproduce biological functions in vitro (Trojanowicz 2001). Recently, thin ﬁlms
of amphiphilic molecules have been used to couple proteins to carbon nanotubes
(Berzina, Troitsky et al. 1996; Feng, Li et al. 2003). 3-D patterns of amphiphilic
molecules can also be used in photon harvesting, organic and polymeric
electroluminescent devices (e.g., light-emitting diodes), organic solid-state lasers, and
photonic band gap materials (Arias-Marin, Amault et al. 2000; Flueraru, Schrader et al.
2000; Schulz, Dietzel et al. 2001; Krebs, Spanggaard et al. 2003). Micro- and nanoscale
partitioning of regions with different chemical composition, charge, or environmental
conditions is a widely used biological motif, as evidenced by the many membrane-
separated organelles and membrane-mediated signaling mechanisms found in cells (Tien

and Ottova 2000; Tavare, Fletcher et al. 2001).

Dendrimers make up a unique, highly diverse class of polymers that have well-
deﬁned macromolecular architectures and are almost perfectly monodisperse. Radially
layered poly-(amidoamine-organosilicon) (PAMAMOS) dendrimers are especially
versatile, amphiphilic and cross-linkable globular-shaped macromolecules (Dvornic, de
Leuze-Jallouli et al. 2000). PAMAMOS dendrimers having dimethoxymethylsilyl
(DMOMS) end-groups can be denoted as PAMAMOS-DMOMS (p, q), where p and q are
integers that deﬁne the generation of the polyamidoamine (PAMAM) interior and the

178

number of exterior layers of organosilicon (OS) branch cells, respectively (see Figure
8.1). PAMAMOS are generally prepared from commercially available PAMAMs with
different relative degrees of amino end-group conversion (i.e., OS substitution). In this

study, PAMAMOS-DMOMS (2,1) dendrimers were used (see Experimental section).

Microcontact printing (uCP) is a soft lithographic technique used in physics,
chemistry, materials science and biology to transfer patterned thin organic ﬁhns to
surfaces with sub-micron resolution (Kumar, Biebuyck et al. 1994; Wilbur, Kumar et al.
1996). Unlike other fabrication methods that merely provide topographic contrast
between the feature and the background, quP also allows chemical contrast to be
achieved via selection of an appropriate ink. uCP offers advantages over conventional
photolithographic techniques because it is simple to perform and is not diffraction
limited. These techniques have been used to make patterns of various small and large
molecules on metals and silicon substrates (St. John and Craighead 1996; Xia, Kim et al.
1996; Yang, Tryk et al. 1996) as well as to deposit proteins, biological cells (Yang,
Mendelsohn et al. 2003; Berg, Yang et al. 2004; Kim, Doh et al. 2004), polymer thin
ﬁlms (POPS) (Jiang, Zheng et al. 2002), controlled particle cluster arrays (Lee, Zheng et
al. 2002) and their selective metal plating (Lee, Hammond et al. 2003), and
polyelectrolyte aggregates (Lee, Ahn et al. 2004). uCP of PAMAM dendrimers has
recently been reported (Ghosh, Lackowski et al. 2001; Arrington, Curry et al. 2002;
Bittner, Wu et al. 2002; Li, Kang et al. 2002; Wu, Bittner et al. 2002; Li, Muir et al.
2003). In these studies the effect of dendrimer concentration on the pattern thickness was

evaluated, and new approaches for electroless metallization of these patterns were

179

suggested. However, no attempt was made to study the effect of other process parameters

controlling the patterns.

In this paper we report the ﬁrst application of microcontact printing of the
amphiphilic and cross-linkable PAMAMOS-DMOMS dendrimers on glass slides, silicon
wafers and polyelectrolyte multilayers (PEMs) in which the pattern average thickness
was controlled by spin-self assembly (i.e., spin-inking). The resulting 3-D micro-
pattemed amphiphilic networks were characterized by optical microscopy and atomic
force microscopy (AFM). Also, the effects of dendrimer ink concentration, inking
method and contact time on the thickness, uniformity and stability of the deposited
patterns are presented. The results provide a framework for controlling the geometry of
the deposited patterns. The lateral footprint of the pattern can be controlled by the shape
of the elastomeric stamp, and the thickness of the patterns can be controlled by adjusting
the spin coating method, the surface properties of the stamp, and the substrate used. The
results also conﬁrmed the well-known inﬂuences of spin speed, concentration, and
solvent on the thickness of spin coated ﬁlms (Pethrick and Rankin 1999; Cho, Char et al.

2001).

180

8.3 EXPERIMENTAL SECTION

Glass slides (Corning Glass Works, Corning, NY.) were cleaned with Alconox precision
cleaner (Alconox Inc., New York, NY) in a Bransonic ultrasonic cleaner (Branson
Ultrasonics Corporation, Danbury, CT) followed by sonication in pure water. They were
then dried under nitrogen ﬂow and subjected to oxygen plasma treatment in Harrick
plasma cleaner (Hanick Scientiﬁc Corporation, Broadway Ossining, NY) for 5 min at 20
Pa vacuum. Polyelectrolyte multilayers (10 bilayers) were deposited on the glass slides
using a standard procedure described elsewhere (Lee, Zheng et al. 2002; Zheng, Lee et al.
2002) with sulfonated poly(styrene) (SPS) as the polyanion and

poly(dimethyldiallylammonium chloride) (PDAC) as the polycation.

Si (100) wafers (Silicon Sense Inc., Nashua, NH) were cleaned by immersion in
Piranha solution (70% sulfuric acid and 30% hydrogen peroxide) at room temperature for

30 min, then rinsed with DI water and dried under nitrogen.

Patterned polydimethylsiloxane (PDMS) stamps were fabricated by pouring a
10:1 mixture of Sylgard 184 elastomer/curing agent (Dow Coming, Midland, MI) over a
patterned silicon master. The mixture was cured for approximately 24 h at 60°C and then

carefully peeled off the master.

Second generation (G2) PAMAMOS-DMOMS (2, 1) dendrimers were obtained
as a 20% w/w methanolic solution from Dendritech, Inc. (Midland, MI). This solution

was further diluted to the desired concentrations (from 0.01% to 1wt.-%) with methanol.

181

Two different methods were utilized to apply ink: spin inking and dip inking. For
spin inking, the stamps were cleaned and treated in the plasma cleaner for 30 s to make
their surfaces hydrophilic. They were then coated with ink using a pipette and
subsequently spun at 3000 rpm for 20 s. Monolayer coating of polymers by this spin self-
assembly has been reported previously (Jang, Kim et al. 2003; Liu, Wang et al. 2003) .
In the dipping method, the stamps were immersed in ink for several minutes and then
dried under nitrogen ﬂow. The ink was then transferred to the silicon substrate by

bringing the stamp into conformal contact with the substrate.

PAMAMOS dendrimers were labeled with ﬂuorescein isothiocyanate (F ITC)
using the standard procedure (Yu and Russo 1996). FITC was predissolved in acetone
and added to methanol solutions of PAMAMOS. The resulting solutions were then
allowed to stand overnight with occasional stirring. The total amount of dye was adjusted
to label just one amino group per PAMAMOS molecule on average, assuming perfect
reaction efﬁciency. The PAMAMOS solutions were then dialyzed against pure water
using sterilized and rinsed membrane tubing. Water for rinsing and dilution was supplied
by a Bamstead Nanopure Diamond-UV puriﬁcation unit (Barnstead International,

Dubuque, Iowa) equipped with a UV source and ﬁnal 0.2 um ﬁlter.

Optical-microscope images were obtained using a Nikon Eclipse ME 600
microscope, and ﬂuorescence images were obtained using the Nikon Eclipse E 400
microscope (Nikon, Melville, NY). Advancing contact angle measurements were
performed with a SEQ Contact Angle Analyzer (Phoenix 450, Surface Electro Optics
Corporation Ltd., Korea). AFM images were obtained with a Nanoscope IV multimode

scope (Digital Instruments, Santa Barbara, CA). The microscope was equipped with

182

tapping-mode etched silicon probes. The thickness analysis of the micr0patterned ﬁlms

was determined using cross-sectional analysis of the AF M images.

8.4 RESULTS AND DISCUSSIONS

The thickness of nanostructured micropattems was found to be controllable, and the
optimum stamping conditions for the uniform nanostructured micropattemed amphiphilic
ﬁlms with high lateral resolution were determined (see Figs. 8.2, 8.3). For all substrates
tested, the pattern average thickness increased with dendrimer concentration over the
concentration range from 0.1% to 1wt.-% when the stamp was spin-coated at 3000 rpm
for 20 s. In this way, the transferred pattern average thickness can be controlled in the
range of 100 to 550 nm by controlling the spin-coating conditions. Thin and unstable
patterns (e.g., see Fig. 8.2a) were formed at concentrations less than 0.5 wt %, while
higher concentrations (>0.5 wt %) resulted in thicker and stable patterns (e.g., see Figs.
8.2 (b-d) and 3). These results conﬁrm the suitability of the spin-inking method as an
alternative to the dip-inking (Wilbur, Kumar et al. 1996) to control the height and
uniformity of the nanostructured and micropattemed ﬁhns and is consistent with the
previously reported dependence of spin coated ﬁlm thickness on concentration, spin
speed, and solvent (Pethrick and Rankin 1999; Cho, Char et al. 2001; Chiarelli, Johal et

al. 2002).

Several parameters were systematically varied to obtain high-quality patterns
having smooth, intact regions of deposited ink, whose shape faithfully reproduced the
topography of the stamp. Under non-optimal conditions, undesirable pattern features

were observed, including ink diffusion at the boundaries, which resulted in ragged edges,

183

and dendrimer aggregation within inked regions, which resulted in irregularly shaped ink
microdomains separated by ﬁssures in case of line patterns and doughnut shaped
structures in case of circular patterns (e.g., Fig. 8.2a). The effects of dendrimer
concentration and contact time on pattern characteristics are presented qualitatively in
Table 8.1 for glass slides, silicon wafers, and polyelectrolyte multilayer-coated substrates
having SPS as the surface layer. As can be seen in Table 8.1, silicon wafers and
polyelectrolyte coated glass slides with sulfonated polystyrene (SPS) as the top-most
layer required longer contact times for pattern transfer than bare glass surfaces. For a
glass substrate, a 1wt.-% solution of PAMAMOS spin-coated at 3000 rpm for 20 s and a
contact time of 5 min resulted in the formation of uniform, stable patterns. For silicon
wafers, a lwt.-% solution of PAMAMOS dendrimers spin-coated twice at 3000 rpm for
20 s and a contact time of 30 min resulted in uniform, stable patterns. For an SPS surface
a 1 wt.-% solution dip-coated for 30 min and a contact time of 1 h resulted in the best
patterns. At lower concentrations and contact times, thinner and unstable patterns were
formed that showed ink aggregation. The reasons for dendrimer aggregation under these
conditions are not clear from these data alone. However, mechanisms involved in
polyelectrolyte aggregation under similar conditions have been explained elsewhere (Lee,
Ahn et al. 2004). While aggregation could be eliminated by increasing the dendrimer
concentration and contact time, at very high concentrations and contact times, diffusion

also became a problem.

The optical microscopy image presented in Figure 8.2b shows excellent contrast
between the PAMAMOS pattern produced and the substrate, indicating successﬁrl

multilayered deposition of PAMAMOS. Fluorescent labeling of PAMAMOS-DMOMS

184

was necessary for lower concentrations of dendrimer inks, because printed features were
difﬁcult to visualize by optical microscopy alone. Previous developmental work on
PAMAMOS-DMOMS has shown that up to 80 or 90% of the amino end-groups of
PAMAM starting material typically reacts with (3-acryloxypropyl)
dimethoxymethylsilane, leaving approximately 10-20% of the end-groups as primary or
secondary amines that are available for reaction with suitable electrophiles (Dvornic, dc
Leuze-Jallouli et al. 2000). Hence, the dendrimers could be covalently bound to
ﬂuorescein isothiocyanate. Figures 8.20, 8.2d show ﬂuorescence images of circular
patterns on a glass substrate and the line patterns on a polyelectrolyte coated glass slides

with SPS as the top surface.

Figure 8.3a shows an AFM image of a patterned glass substrate stamped with a
1wt.-% dendrimer solution, for a contact time of 5 min, and Figure 8.3b shows an AFM
image of patterned silicon wafer stamped with the same dendrimer solution for a contact
time of 30 min. The light areas are deposited dendrimers, while the dark areas are the
underlying substrate. The structure heights for these patterns are depicted in Figures 8.30
and 8.3d, respectively. The results indicate consistent coverage where the stamp met the
substrate and uniform average thickness across the pattern. The pattern shape, as seen in
the cross-sectional views (Figure 8.30 and 8.3d), was found to be round for most of the
patterns transferred. This surface curvature, we believe, arises from a complex interplay
of forces that occur during deposition and drying of the dendrimer ink, as well as
unavoidable edge effects due to the size and shape of the AFM tip. The 3-D images
shown in Figures 8.3e and Figure 8.3f indicate that the upper edges of the line patterns

exhibited a slight height variance, which tended to decrease with the thickness of the

185

patterns for all substrates examined. Although rinsing with water after deposition
smeared the patterns, very stable patterns could be obtained if the substrates were air-
dried for an hour or cured in an oven at 120°C for an hour. This high stability is
attributed to the crosslinking reaction depicted in Figures 8.4 and 8.5. The reaction is
reported (Dvornic, de leuze Jallouli et al. 1999; Dvornic, Li et al. 2002) to proceed
through two steps: (i) water hydrolysis of methoxysilyl (Si-OCH3) end-groups to yield
the corresponding silanols, Si-OH, and (ii) subsequent condensation of these silanols to
form siloxane (Si-O-Si) interdendrimer bridges. The second reaction is self catalyzed by
the basic PAMAM interiors and is easily accomplished either by direct exposure of
methoxysilyl-ﬁrnctionalized PAMAMOS-DMOMS dendrimer precursors to atmospheric
moisture, or by controlled addition of water either in the form of vapor (e.g., in a
humidity chamber), or blended as liquid into a dendrimer solution. Since the process is a
chain reaction in which the water consumed in the hydrolysis step is regenerated in the

condensation step, less than the stoichiometric amount of wateris needed.

Dip-inking produced thinner and relatively non-uniform (i.e., poorly covered and
aggregated) features, while spin-inking produced thicker and more uniform features. This
effect is attributed to spin coating giving a thicker layer of dendrimer ink than dip
coating. An untreated stamp had a water advancing contact angle of 102° before plasma
treatment and 6° after plasma treatment, indicating that the plasma treatment rendered the
stamp surface hydrophilic. Coating the plasma-treated stamps with hydrophobic
PAMAMOS dendrimer ink increased the surface hydrophobicity, and hence the contact
angle. Spin-coated stamps had a higher contact angle (60°) than those dip—coated (around

45°), suggesting that the roughness, thickness and morphology of the thin ﬁlms formed in

186

both cases are different. Because the thickness of the dendrimer ink layer can be
controlled by adjusting the spin rate, spin coating should allow the properties of the

patterned surface to be more ﬁnely controlled than dip coating.

Due to the very regular size and globular shape of the dendrimers and their
controlled amphiphilicity and cross-linkable nature, the micropattemed arrays of
PAMAMOS ﬁlms offer advantages over other macromolecular materials in the
fabrication of the 3-D bioconjugate surfaces and biomimetic interfaces for applications
including producing biochips for proteomics, pharmaceutical screening processes, and
addressing fundamental questions in cell adhesion (Blawas and Reichert 1998; Kane
1999; Mac Beath and Schreiber 2000; Yang, Mendelsohn et al. 2003; Berg, Yang et al.
2004; Kim, Doh et a1. 2004). The cage-like structure and controllable porosity of
dendrimers, coupled with the unique ability of PAMAMOS dendrimers to form highly
cross-linked structures, makes the structures developed in this paper well suited to
encapsulate molecular mediators and cofactors for redox enzymes used in bioelectronic
applications (Willner and Katz 2000). Moreover, these versatile dendrimer arrays could
serve as templates for nanoparticles grth or electroless metallization in deve10pment of
nano metal reactors on micropattemed ﬁlms (Balough, de leuze Jallouli et al. 1999;
Dvornic and Owen 2002). Further characterization of patterned surfaces and in situ
chemical reactions of these ﬁlms is currently underway for electronic devices and sensor

applications.

187

8.5 CONCLUSIONS

This paper reports the ﬁrst micropattemed deposition of stable, nanostructured,
amphiphilic and cross-linkable PAMAMOS-DMOMS dendrimer multilayers onto silicon
wafers, glass surfaces, and polyelectrolyte multilayers. Using optical microscopy,
ﬂuorescence microscopy, AFM, and contact-angle analysis, the effects of dendrimer ink
concentration, contact time, and inking method, on the thickness, uniformity and stability
of the resulting patterns were studied. In the spin—self-assembly process, pattern average
thickness increased with increasing dendrimer ink concentration at constant spin speed.
Interdendrimer crosslinking via siloxane condensation allowed the 3-D amphiphilic and
multilayered structures to remain stable for several days, even after washing if they were
kept in air for an hour or heat treated at 120 °C in an oven. The resulting highly cross-
linked networks have the ability to encapsulate nanoparticles and to serve as molecular
templates for chemical and physical modiﬁcations in opto-electronic and biomimetic

interface applications.

188

Pl1(2-y)/
PAMAM N

Gx

 

—(\CH2)2’E_ O_ (CH2)3— SI— CH3

O’CH3

O_CH3

Figure 8.1: Schematic representation of PAMAMOS-DMOMS dendrimer.

189

 

500.0 nm ’ ‘ ’

250.0 nm 1 i 7 1

0.0 nm 50 ,um

 

 

(C) (d)

Figure 8.2: (a) An example of an AFM image showing unstable circular patterns on a
glass slide obtained from a 0.5% PAMAMOS dendrimer solution. (b-d) Visualization of
PAMAMOS dendrimer patterns on various substrates. (b) Optical micrograph of the
circular patterns on a glass substrate (0) Fluorescence image of the circular pattern on a
glass substrate; ((1) Fluorescence image of the line pattern on a SPS surface.

190

60pm

| 60pm

450 nm _

 

 

  

(0

Figure 8.3: AFM images of cross-linked PAMAMOS dendrimer patterns. (a) Patterned
glass substrate stamped by 1wt.-% dendrimer solution in methanol for a contact time of 5
min; (b) Patterned silicon wafer stamped by a 1wt.—% dendrimer solution in methanol for
a contact time of 25 min; (0) Pattern height on glass substrate; ((1) Pattern height on
silicon wafer; (0) 3-D image of the patterned glass substrate; (f) 3—D image of the
patterned silicon wafer.

191

 

 

Illa-y) o - OCHB
PAg/IfM N (CH2)2—|(|3—O—(CH2)3—Si—CH3
\
_ Z
Hydrolysis
I|1(2-y) O OCH3

l |
"N (CH2)2—C—O—(CH2)3—Si—CH3

OH
Y
L —z

 

 

Cross-linking Reaction

PAMAM
H3co\Si Gx
0/ \CH3
s'i—CH,
00H3

Figure 8.4: Crosslinking of PAMAMOS-DMOMS dendrimers into a covalently bonded
network structure.

192

PAMAMOS-DMOMS

 

OH OH OH OH
Si Si Si Si
/ \O/ \O/ \O/ \

 

I
/ \O/ \O/ \O/

PAMAM
Interior
N

I. I. I,
SI S S

 

 

 

 

 

 

 

Figure 8.5: Covalent bonding of PAMAMOS-DMOMS dendrimers to glass surfaces

having silanol surface groups.

193

Table 8-1: Effects of concentration and contact time on the stability and uniformity of

patterns on different substrates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Glass Slides Silicon SPS
Wafers Surface
Dendrimer Contact Time
Solution (min)
Concentration“
(Wt 70)
1 3 5 10 30 60
0.1 % Poor Fair Good Good Fair Fair coverage
coverage coverage coverage coverage coverage with some
with with some with some with with some aggregation
aggregation aggregation aggregation diffusion aggregation
and
aggregation
0.5 Fair Fair Good Good Good Good
coverage coverage coverage coverage coverage coverage
with some with very with
aggregation little diffusion
aggregation
1.0 F air Good Good Good Good Good
coverage coverage coverage coverage coverage coverage
with very with
little diffusion
aggregation

 

* Concentration less than 0.1 wt % resulted in negligible and unstable pattern transfers

194

 

9 REFERENCES

Abel, A. P., M. G. Weller, et al. (1996). "Fiber-optic evanescent wave biosensor for the
detection of oligonuclcotides." An_alvtical Chemistm 68(17): 2905-2912.

Ahn, Y. B., L. A. Beaudette, et al. (2001). "Survival of a GFP-labeled polychlorinated
biphenyl degrading psychrotolerant Pseudomonas spp. in 4 and 22 degrees C soil
microcosms." Microbial Ecology 42(4): 614-623.

Albrecht, 0., D. S. Johnston, et al. (1982). "Stable Biomembrane Surfaces Formed by
Phospholipid Polymers." Biochimica Et Biophysica Acta 687(2): 165-169.

Allen, T. G. J. (1997). "The 'sniffer-patch' technique for detection of neurotransmitter
release." Trends in Neurosciences 20(5): 192-197.

Alzagtat, A. A. and I. Alli (2002). "Protein-lipid interactions in food systems: a review."
International Journal of Food Sciences and Nutrition 53(3): 249-260.

Ariga, K. and Y. Okahata (1989). "Polymerized monolayers of single—chain , double-
chain, and triple chain silane amphiphiles and permeation control through the
monolayer-immobilized porous glass plate in an aqeuous-solution." Journal of the
American Chemical Societv 1 11: 5618-5622.

Asaka, K., A. Ottova, et al. (1999). "Mediated electron transfer across supported bilayer
lipid membrane (s-BLM)." Thin Solid Films 354(1-2): 201-207.

Atkins, J. and P. Glynn (2000). "Membrane Association of and Critical Residues in the

Catalytic Domain of Human Neuropathy Target Esterase." The Journal of
Biological Chemistry 275(32): 24477-24483.

Balachandra, A. M., J. H. Dai, et a1. (2002). "Enhancing the anion-transport selectivity of
multilayer polyelectrolyte membranes by templating with Cu2+."
_mcromolecules 35(8): 3171-3178.

Balzani, V., P. Ceroni, et al. (2003). Luminescent dendrimers. Recent advances.
Dendrimers V: Functional and vaerbranched Building Blocks. Photophvsical
Properties, Applications in Materials and Life Sciences. 228: 159-191.

195

Barak, L. S., S. S. G. Ferguson, et al. (1997). "A beta-arrestin green ﬂuorescent protein
biosensor for detecting G protein-coupled receptor activation." Journal of
Biological Chemistry 272(44): 27497-27500.

Bartlett, P. N. and R. G. Whitaker (1987). "Electrochemical Immobilization of
enzymes. 1 .Theory." Journal of Electroa_r;alvti.cal Chemistry 224: 27-35.

Baselt, D. R., G. U. Lee, et a1. (1998). "A biosensor based on magnetoresistance
technology." Biosensors & Bioelectronics 13(7-8): 731-739.

Bauer, C. G., A. V. Eremenko, et a1. (1996). "Zeptomole-detecting biosensor for alkaline
phosphatase in an electrochemical immunoassay for 2,4—dichlorophenoxyacetic
acid." Analﬁical Chemistry 68(15): 2453-2458.

Berg, M. C., S. Y. Yang, et al. (2004). "Controlling mammalian cell interactions on
patterned polyelectrolyte multilayer surfaces." Langmuir 20(4): 1362-1368.

Boas, U. and P. M. H. Heegaard (2004). "Dendrimers in drug research." Chemical
Society Reviews 33(1): 43-63.

Boxer, S. G. (2000). "Molecular transport and organization in supported lipid
membranes." Current Opinion in Chemical Biology 4(6): 704-709.

Bruening, M. L., J. J. Harris, et al. (2001). "Ultrathin nanoﬁltration membranes prepared
by layer-by-layer deposition of polyelectrolytes on porous supports." Abstracts of
Papers of the AmericaLChemicLISociegg 222: U388-U388.

Burdette, D. S. and J. G. Zeikus (1994). "Puriﬁcation of acetaldehyde dehydrogenase and
alcohol dehydrogenases from Thermoanaerobacter ethanolicus 39B and
characterization of the secondary-alcohol dehydrogenase (2 degrees Adh) as a
bifunctional alcohol dehydrogenase--acetyl-COA reductive thioesterase."
Biochemical Joumgﬂ 302: 163-170.

Burestedt, E., A. Narvaez, et a1. (1996). "Rate-Limiting Steps of Tyrosinase-Modiﬁed
Electrodes for the Detection of Catechol." _An_alytigal Chemistry 68(9): 1605-
1611.

196

Cassier, T., A. Sinner, et al. (1999). "Homogeneity, electrical resistivity and lateral
diffusion of lipid bilayers coupled to polyelectrolyte multilayers." Colloids and
Biointerfaces 15: 215-225.

Coche-Guerente, L., P. Labbe, et al. (1999). "Ampliﬁcation of arnperiometric biosensor
responses by electrochemical substrate recycling Part 1. Theoretical treatment of
the catechol-polyphenol oxidase system " Journal of Electroagrlvtical Chemigry
470: 53-60.

Coche-Guerente, L., P. Labbe, et al. (2001). "Ampliﬁcation of amperometric biosensor
responses by electrochemical substrate recycling. 3. Theoretical and experimental
study of the phenol-polyphenol oxidase system immobilized in Laponite
hydrogels and layer-by-layer self-assembled structures." _Aﬁlytical Chemistry
73(14): 3206-3218.

Cornell, B. A., G. Krishna, et al. (2001). "Tethered-bilayer lipid membranes as a support
for membrane-active peptides." Biochemical Society Transactions 29: 613-617.

Dai, J. H., A. M. Balachandra, et al. (2002). "Controlling ion transport through multilayer
polyelectrolyte membranes by derivatization with photolabile functional groups."
Macromolecules 35(8): 3164-3170.

Dai, J. H., A. W. Jensen, et al. (2001). "Controlling the permeability of multilayered
polyelectrolyte ﬁlms through derivatization, cross-linking, and hydrolysis."
Langmuir 17(3): 931-937.

Dall'0rto, V. C., C. Danilowicz, et al. (1999). "Comparison between three amperometric
sensors for phenol determination in olive oil samples." Analytical Letters 32(10):
1981-1990.

Darvesh, S., D. A. Hopkins, et al. (2003). "Neurobiology of butyrylcholinesterase."
Nature Reviews Neuroscience 4(2): 131-138.

Davis, L., J. J. Britten, et al. (1997). "Cholinesterase - Its signiﬁcance in anaesthetic
practice." Anaesthesia 52(3): 244-260.

Decher, G. (1997). "Fuzzy nanoassemblies: Toward layered polymeric multicomposites."
Science 5330: 1232-1237.

197

Decher, G. and J. D. Hong (1991). "Buildup of ultrathin multilayer ﬁlms by a self-
assembly process.2. consecutive adsorption of anionic and cationic bipolar
amphiphiles and polyelectrolytes on charged surfaces." Berichte Der Bunﬂ
Gesell Scﬁrﬂ-Physical Chemistry Chemical Physics 95: 1430-1434.

Desprez, V. and P. Labbe (1996). "A kinetic model for the electroenzymatic processes
involved in polyphenol-oxidase-based amperometric catechol sensors." Journal of
Electroanalytical Chemistry 415(1-2): 191-195.

DiazGarcia, M. E. and M. J. ValenciaGonzalez (1995). "Enzyme catalysis in organic
solvents: A promising ﬁeld for optical biosensing." Talanta 42(11): 1763-1773.

Ducey, M. W. and M. E. Meyerhoff (1998). "Microporous Gold Electrodes as Combined
Biosensor/Electrochemical Detectors in Flowing Streams." Electroanalysis 10(3):
157-162.

Dvornic, P. R. and M. J. Owen (2002). ACS symposium series 838.American Chemicaﬂ
Soceity: 236-259.

Dvornic, P. R. and D. A. Tomalia (1996). "Recent advances in dendritic polymers
Current Opinion in Colloid and Interface Science 1: 221-235.

Dzyadevych, S. V., A. P. Soldatkin, et a1. (2002). "Assessment of the toxicity of methyl
parathion and its photodegradation products in water samples using
conductometric enzyme biosensors." Analﬂica Chimica Acta 459(1): 33-41.

Favero, G., A. D'Annibale, et al. (2002). "Membrane supported bilayer lipid membranes
array: preparation, stability and ion-charmel insertion." Analy_tica Chimica Acta
460(1): 23-34.

Feldman, K., G. Hahner, et al. (1999). "Probing resistance to protein adsorption of
oligo(ethylcne glycol)-terminated self-assembled monolayers by scanning force
microscopy." Journal of the American Chemical Society 121(43): 10134-10141.

Flueraru, C., S. Schrader, et al. (2000). MCLC S&T, Section B: Nonlinear thics 25(1-
4): 247-252.

Fodor, S. P. A., J. L. Read, et al. (1991). "Light-Directed, Spatially Addressable Parallel
Chemical Synthesis." Science 251(4995): 767-773.

 

198

Forshaw, P. J ., J. Atkins, et al. (2001). "The catalytic domain of human neuropathy target
esterase mediates an organophosphate-sensitivc ionic conductance across
liposome membranes." Journal of Neurochemistrv 79(2): 400-406.

Forzani, E. S., V. M. Solis, et al. (2000). "Electrochemical behavior of polyphenol
oxidase immobilized in self-assembled structures layer by layer with cationic
polyallylamine." Analytical Chemistry 72(21): 5300-5307.

Forzani, E. S., M. L. Teijelo, et al. (2003). "Effect of the polycation nature on the
structure of layer-by-layer electrostatically self-assembled multilayers of
polyphenol oxidase." Biomacromolecules 4(4): 869-879.

Fou, A. C. and M. F. Rubner (1995). "Molecular-level processing of conjugated
polymers.2. layer-by-layer manipulation of in-situ polymerized p-type doped
conducting polymers " Macromolecules 28: 71 15-7120.

Freire, R. S., N. Duran, et a1. (2002). "Electrochemical biosensor-based devices for
continuous phenols monitoring in environmental matrices." Jourﬂl of the
Brazilian Chemical Society 13(4): 456-462.

Frens, G. (1973). "Controlled nucleation for regulation of particle-size in monodisperse
gold suspensions " Nature-Physical Science 241: 20-22.

Gajraj, A. and R. Y. Ofoli (2000). "Quantitative technique for investigating
macromolecular adsorption and interactions at the liquid-liquid interface."
Langmuir 16(9): 4279-4285.

Gerard, M., A. Chaubey, et al. (2002). "Application of conducting polymers to
biosensors." Biosensors & Bioelectronics 17(5): 345-359.

Ghosh, S. and A. K. Banthia (2004). "Fluorophore-labeled polyamidoamine (PAMAM)
dendritic architectures: Synthesis and ﬂuorescence sensing properties."
Supramolecular Chemistry 16(7): 487-494.

Glass, G. B. J. (1966). "Introduction - Gastric Proteins and Mucoproteins." American
iournal of digestive diseases 11(2): 79-&.

Glynn, P. (1999). "Neuropathy target esterase." Biochemical Joumaﬂ 344: 625-631.

199

Gotschy, W., K. Vonrnetz, et al. (1996). Applied Physics B: Lasers and Optics B63(4):
381-384.

Grabar, K. C., P. C. Smith, et al. (1996). "Kinetic control of interparticle spacing in Au
colloid-based surfaces: Rational nanometer-scale architecture " Journal of the
American Chemical Society 118: 1148-1153.

Graneli, A., J. Rydstrom, et al. (2003). "Formation of supported lipid bilayer membranes
on Si02 ﬁ'om proteoliposomes containing transmembrane proteins." Langmuir
19(3): 842-850.

Groves, J. T. and S. G. Boxer (2002). "Micropattem formation in supported lipid
membranes." Accounts of Chemical ReseaLch 35(3): 149-157.

Groves, J. T., S. G. Boxer, et a1. (1998). "Electric ﬁeld-induced critical demixing in lipid
bilayer membranes." Proceedings of the National Academy of Sciences of the
United States of America 95(3): 935-938.

Groves, J. T., N. Ulman, et al. (1997). "Micropatteming ﬂuid lipid bilayers on solid
supports." Science 275(5300): 651-653.

 

Guan, F., M. A. Chen, et al. (2005). "Fabrication of patterned gold microstructure by
selective electroless plating." Applied Surface Science 240(1-4): 24-27.

Harris, J. M. (1992). Poly(ethylene glycol) chemistry : biotechnical and biomedical
applications. New York, Plenum Press.

Herz, F. and E. Kaplan (1973). "Review - Human Erythrocyte Acetylcholinesterase."
Pediatric research 7(4): 204-214.

Heysel, S., H. Vogel, et al. (1995). "Covalent Attachment of Functionalized Lipid
Bilayers to Planar Wave-Guides for Measuring Protein-Binding to Biomimetic
Membranes." Protein Science 4(12): 2532-2544.

Hovis, J. S. and S. G. Boxer (2000). "Patterning barriers to lateral diffusion in supported
lipid bilayer membranes by blotting and stamping." angrauir 16(3): 894-897.

200

Hsieh, B. H, J. F. Deng, et al. (2001). "Acetylcholinesterase inhibition and the
extrapyramidal syndrome: A review of the neurotoxicity of organophosphate. "
Neurotoxicology 22(4). 423-427.

J iang, X. and P. T. Hammond (2000). "Selective deposition in layer-by-layer assembly:
Functional graft copolymers as molecular templates " Langmuir 16: 8501-8509.

J iang, X., H. Zheng, et al. (2002). "Polymer-on-polymer stamping: Universal approaches
to chemically patterned surfaces " Langmuir 18: 2607-2615.

Johnson, J. M., T. Ha, et al. (2002). "Early steps of supported bilayer formation probed
by single vesicle ﬂuorescence assays." Biophysical Journal 83(6): 3371-3379.

Kalb, E., S. Frey, et a1. (1992). "Formation of Supported Planar Bilayers by Fusion of
Vesicles to Supported Phospholipid Monolayers." Biochimica Et Biophysica Acta
1103(2): 307-316.

Kam, L. and S. G. Boxer (2001). "Cell adhesion to protein-micropattemed-supported
lipid bilayer membranes." Jouﬂal of Biomedical Materials Research 55(4): 487-
495.

Kararn, J. and J. A. Nicell (1997). "Potential applications of enzymes in waste treatment."
Journal of Chemical Technology and Biotechnology 69(2): 141-153.

Karlsson, R. and A. Falt (1997). "Experimental design for kinetic analysis of protein-
protein interactions with surface plasmon resonance biosensors." Journal of
Immunological Methods 200(1-2): 121-133.

Kayyali, U. S., T. B. Moore, et a1. (1991). "Neurotoxic Esterase (Nte) Assay - Optimized
Conditions Based on Detergent-Induced Shifts in the Phenol/4-Aminoantipyrine
Chromophore Spectrum." Journal of Analytical Toxicology 15(2): 86-89.

Keller, C. A. and B. Kasemo (1998). "Surface speciﬁc kinetics of lipid vesicle adsorption
measured with a quartz crystal microbalance." Biophysical Journal 75(3): 1397-
1402.

Kidambi, S., C. Chan, et a1. (2004). "Selective depositions on polyelectrolyte multilayers:
Self-assembled monolayers of m-dPEG acid as molecular template " Journal of
the American Chemical Society 126: 4697-4703.

201

Kim, B. and Y. W. Xia, G. M. (1995). "Polymer microstructures formed by molding in
capillaries." Nature 376: 6541.

Kim , Y. D., C. B. Park, et al. (2001). "Stable sol-gel rnicrostructured and microﬂuidic
networks for protein patterning." Biotechnology and Bioengineering 73(5): 331.

Kohli, N., P. Dvornic, et al. (2004). "Nanostructured Cross-linkable Micropattems via

Amphiphilic Dendrimer Stamping." Macromolecular Rapid Communications
25(9): 935-941.

Kohli, N., P. R. Dvornic, et al. (2004). "Nanostructured crosslinkable micropattems by

amphiphilic dendrimer stamping." Macromolecular Rapid Communications 25(9):
935-941.

Kohli, N., B. L. Hassler, et a1. (2006). "Tethered lipid bilayers on electrolessly deposited
gold for bioelectronic applications." Biomacromolecules 7: 3327-3335.

Kohli, N., S. Vaidya, et al. (2006 ). "Arrays of lipid bilayers and liposomes on patterned
polyelectrolyte templates." Journal of Colloid and Interface Science 301: 461-469.

Kohli, N., R. M. Worden, et al. (2005). "Intact transfer of layered, bionanocomposite
arrays by microcontact printing." Chem. Commun. 3: 316-318.

Kreuzer, H. J., R. L. C. Wang, et al. (2003). "Hydroxide ion adsorption on self-assembled
monolayers " Joumaﬂf the America Chemical Society 125(27): 8384-8389.

Krishna, G., J. Schulte, et al. (2001). "Tethered bilayer membranes containing ionic
reservoirs: The interfacial capacitance." Langmuir 17(16): 4858-4866.

Kropp, T. J ., P. Glynn, et a1. (2004). "The mipafox-inhibited catalytic domain of human
neuropathy target esterase ages by reversible proton loss." Biochemistry 43(12):
3716-3722.

Krysinski, P., A. Zebrowska, et al. (2001). "Tethered mono- and bilayer lipid membranes
on Au and Hg." Langmuir 17(13): 3852-3857.

Kugler, R. and W. Knoll (2002). "Polyelectrolyte-supported lipid membranes."
Bioelectrochemistry 56(1-2): 175-178.

202

Kulkami, P., M. Karve, et al. (1998). "Metalloenzyme biosensors: Part 1 - Amperometric
tyrosinase biosensor for phenols using composite biopolymer matrix." Indian
Jourml of Chemistry I 37(1): 7-10.

Kumar, A., H. A. Biebuyck, et al. (1994). "Patterning self-assembled monolayers-
Applications in materials science." Langmuir 10: 1498-1511.

Kumar, A. and G. M. Whitesides (1993). "Patterning self-assembled monolayers-

Kung, L. A., J. T. Groves, et al. (2000). "Printing via photolithography on
micropartitioned ﬂuid lipid membranes." Advanced Materials 12(10): 731-+.

Kung, L. A., L. Karn, et al. (2000). "Patterning hybrid surfaces of proteins and supported
lipid bilayers." Langr_nuir 16(17): 6773-6776.

Lakowicz, J. R. (1983). Principles of ﬂuorescence spectroscopy. New York, Plenum
Press.

Lee, I., P. T. Hammond, et al. (2003). "Selective electroless nickel plating of particle
arrays on polyelectrolyte multilayers." Chemistry of Materials 15: 4583-45 89.

Lee, 1., H. Zheng, et al. (2002). "Controlled cluster size in patterned particle arrays via
directed adsorption on conﬁned surfaces." Advgced Materials 14: 572-577.

Lejus, C., Y. Blanloeil, et al. (1998). "Cholinesterases." Annales Francaises D Anesthesie
ct de Reanimation 17(9): 1122-1135.

Li, Y., D. Dinsdale, et al. (2003). "Protein domains, catalytic activity, and subcellular
distribution of neuropathy target esterase in mammalian cells." Journal of
Biological Chemistry 278(10): 8820-8825.

Luo, G. B., T. T. Liu, et al. (2001). "Investigation of polymer-cushioned phospholipid
bilayers in the solid phase by atomic force microscopy." Langmuir 17(13): 4074-
4080.

Lvov, Y., K. Ariga, et al. (1995). "Assembly of multicomponent protein ﬁlms by means
of electrostatic layer-by-layer adsorption." Journal of the American Chemical
Society 117: 6117-6123.

203

Ma, C., M. P. Srinivasan, et al. (2003). "Supported lipid bilayers lifted from the substrate

by layer-by-layer polyion cushions on self-assembled monolayers." Colloids and
Surfaces B-Biointerfaces 28(4): 319-329.

MacBeath, G. and S. L. Schreiber (2000). "Printing proteins as microarrays for high-
throughput function determination" Science 289: 1760.

Makhaeva, G. F., L. V. Sigolaeva, et al. (2003). "Biosensor detection of neuropathy
target esterase. in whole blood as a biomarker of exposure to neuropathic
organophosphorus compounds." Journal of Toxicology and Environmental
Health-Part A 66(7): 599-610.

Marsh, D., L. I. Horvath, et al. (2002). "Interaction of membrane-spanning proteins with
peripheral and lipid-anchored membrane proteins: perspectives from protein-lipid
interactions (Review)." Molecular Membrane Biology 19(4): 247-255.

Michel, R., J. W. Lussi, et al. (2002). "Selective molecular assembly patterning: A new
approach to micro- and nanochemical patterning of surfaces for biological
applications " Langrpuir 18(8): 3281-3287.

Millard, C. B. and C. A. Broomﬁeld (1995). "Anticholinestcrascs - Medical Applications
of Neurochemical Principles." Journal of Neurochemisg 64(5): 1909-1918.

Miller, M. D. and M. L. Bruening (2004). "Controlling the nanoﬁltration properties of
multilayer polyelectrolyte membranes through variation of ﬁlm composition."
Langr_nuir 20(26): 11545-11551.

Moncelli, M. R., L. Becucci, et al. (2004). "Tethered bilayer lipid membranes self-
assembled on mercury electrodes." Bioelectrochemistry 63(1-2): 161-167.

Moya, S., W. Richter, et al. (2003). "Freeze-ﬁacture electron microscopy of lipid
membranes on colloidal polyelectrolyte multilayer coated supports."
Biomacromolecules 4: 808-814.

Munro, J. C. and C. W. Frank (2004). "Adsorption of lipid-functionalized poly(ethylene
glycol) to gold surfaces as a cushion for polymer-supported lipid bilayers."
Langmuir 20(8): 3339-3349.

204

Murarnatsu, H., J. M. Dicks, et a1. (1987). "Piezoelectric Crystal Biosensor Modiﬁed with
Protein-a for Determination of Immunoglobulins." Amlytical Chemiﬁy 59(23):
2760-2763.

Naumann, C. A., 0. Prucker, et al. (2002). "The polymer-supported phospholipid bilayer:
Tethering as a new approach to _ substrate-membrane stabilization."
Biomacromolecules 3(1): 27-35.

Naumann, R., S. M. Schiller, et al. (2003). "Tethered lipid Bilayers on ultraﬂat gold
surfaces." Langmuir 19(13): 5435-5443.

Naumann, R., E. K. Schmidt, et al. (1999). "The peptide-tethered lipid membrane as a
biomimetic system to incorporate cytochrome c oxidase in a functionally active
form." Biosensors & Bioelectronics 14(7): 651-662.

Naumann, R., D. Walz, et al. (2003). "Kinetics of valinomycin-mediated K+ ion transport
through tethered bilayer lipid membranes." Journal of Electroanalytical Chemistry
550: 241-252.

Nikolelis, D. P., T. Hianik, et al. (1999). "Biosensors based on thin lipid ﬁlms and
liposomes." Electroanalysis 11(1): 7-15.

Nollert, P., H. Kiefer, et al. (1995). "Lipid Vesicle Adsorption Versus Formation of
Planar Bilayers on Solid-Surfaces." Biophysical Journal 69(4): 1447-1455.

0hlsson, P. A., T. Tjamhage, et al. (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." Bioelectrochemistpya and Bioeneagetics 38(1): 137-
148.

Park, J. and P. T. Hammond (2004). "Multilayer transfer printing for polyelectrolyte
multilayer patterning: Direct transfer of layer-by-layer assembled micropattemed
thin ﬁlms " Advaraced Materials 16: 520-524.

Perez-Salas, U. A., K. M. Faucher, et al. (2003). "Characterization of a biomimetic
polymeric lipid bilayer by phase sensitive neutron reﬂectometry." Langmuir
19(19): 7688-7694.

205

Plant, A. L. (1999). "Supported hybrid bilayer membranes as rugged cell membrane
mimics." Langmuir 15(15): 5128-5135.

Prodromidis, M. I. and M. I. Karayannis (2002). "Enzyme based amperometric
Biosensors for food analysis." Electroanalysis 14(4): 241-261.

Proux-Delrouyre, V., C. Elie, et al. (2002). "Formation of tethered and streptavidin-
supported lipid bilayers on a microporous electrode for the reconstitution of
membranes of large surface area." Langr_nuir 18(8): 3263-3272.

Raguse, B., V. Braach-Maksvytis, et al. (1998). "Tethered lipid bilayer membranes:
Formation and ionic reservoir characterization." Langmuir 14(3): 648-659.

Rainier, S., M. Bui, et al. Neuropathy target esterase gene mutations cause motor neuron
disease, American Society of Human Genetics, Annual Meeting, Salt lake city,
Utah, 2005

Raj esh, W. Takashima, et al. (2004). "Amperometric phenol biosensor based on covalent
immobilization of tyrosinase onto an electrochemically prepared novel copolymer
poly (N -3-aminopropyl pyrrole-co-pyrrole) ﬁlm." Seasors and Actuﬂors B-
Chemical 102(2): 271-277.

Ramanathan, K. and B. Danielsson (2001). "Principles and applications of thermal
biosensors." Biograors & Bioelectronics 16(6): 417-423.

Robeson, J. L. (1995). Application of total internal reﬂection ﬂuorescence to probe
surface diffusion and orientation of adsorbed protiens. Chemical Egineering.
Pittsburgh, Carnegie Mellon University: 183.

Ross, E. E., B. Bondurant, et al. (2001). "Formation of self-assembled, air-stable lipid
bilayer membranes on solid supports." Langmuir 17(8): 2305-2307.

Saccani, J ., S. Castano, et al. (2003). "A phospholipid bilayer supported under a
polymerized Langmuir ﬁlm." Biophysical J oumal 85(6): 3781-3787.

Sackmann, E. (1996). "Supported membranes: Scientiﬁc and practical applications."
Science 271(5245): 43-48.

206

Sackmann, E. and M. Tanaka (2000). "Supported membranes on soft polymer cushions:
fabrication, characterization and applications." Trends in Biotechnology 18(2):
58-64.

Saini, S., G. F. Hall, et al. (1991). "Organic-Phase Enzyme Electrodes." Analyaica
Chimica Acta 249(1). 1-15.

Sanders, C. R. and K. Oxenoid (2000). "Customizing model membranes and samples for
NMR spectroscopic studies of complex membrane proteins." Biochimica Et
Biophysica Acta-Biomembraﬂa 1508(1-2): 129-145.

Sigolaeva, L. V., A. Makower, et al. (2001). "Bioelectrochemical analysis of neuropathy
target esterase activity in blood." Analﬂical Biochemistry 290(1): 1-9.

Sinner, E. K. and W. Knoll (2001). "Functional tethered membranes." Current Opinion in
Chemical Biology 5(6): 705-711.

Sohling, U. and A. J. Schouten (1996). "Investigation of the adsorption of dioleoyl-L-
alpha-phosphatidic acid mono- and bilayers from vesicle solution onto
polyethylenimine-covered substrates." Langmuir 12(16): 3912-3919.

Sokolovskaya, L. G., L. V. Sigolaeva, et al. (2005). "Improved electrochemical analysis
of neuropathy target esterase activity by a tyrosinase carbon paste electrode
modi.ed by l-methoxyphenazine methosulfate." Biotechnology Letters 27: 1211-
1218.

Solna, R., S. Sapelnikova, et al. (2005). "Multienzyme electrochemical array sensor for
determination of phenols and pesticides." Talanta 65(2): 349-357.

Solna, R. and P. Skladal (2005). "Amperometric ﬂow-inj ection determination of phenolic
compounds using a biosensor with immobilized laccase, peroxidase and
tyrosinase." Electroanalysis 17(23): 2137-2146.

St. John, P. M. and H. G. Craighead (1996). "Microcontact printing and pattern transfer
using trichlorosilanes on oxide substrates " Applied Physics Letters 68: 1022-
1024.

Stanca, S. E., I. C. Popescu, et al. (2003). "Biosensors for phenol derivatives using
biochemical signal ampliﬁcation." Talanta 61(4): 501-507.

 

207

Stanton, B. W., J. J. Harris, et al. (2003). "Ultrathin, multilayered polyelectrolyte ﬁlms as
nanoﬁltration membranes." Langmuir 19(17): 7038-7042.

Starr, T. E. and N. L. Thompson (2002). "Fluorescence pattern photobleaching recovery
for samples with multi-component diffusion." Biophysigrl Chemistry 97(1): 29-
44.

Stelzle, M., G. Weissmuller, et al. (1993). "On the Application of Supported Bilayers as
Receptive Layers for Biosensors with Electrical Detection." Journal of Physical
Chemistpy 97(12): 2974-2981.

Stora, T., J. H. Lakey, et a1. (1999). "Ion-channel gating in transmembrane receptor
proteins: Functional activity in tethered lipid membranes." Angewandte Chemic-
International Edition 38(3): 389-392.

Supriya, L. and R. 0. Claus (2004). "Solution-based assembly of conductive gold ﬁlm on
ﬂexible polymer substrates." Langmuir 20(20): 8870-8876.

Tavare, J. M., L. M. Fletcher, et al. (2001). "Review - Using green ﬂuorescent protein to
study intracellular signalling." J ouraal of Endocrinology 170(2): 297-306.

Terrettaz, S., M. Mayer, et al. (2003). "Highly electrically insulating tethered lipid
bilayers for probing the function of ion channel proteins." Langmuir 19(14):
5567-5569.

Thompson, C. M. and R. J. Richardson (2004). Anticholinesterase Insecticides. Pesticide
Toxicology and International Regplation. T. C. Marrs and B. Ballantyne.
Chichester, John Wiley & Sons Limited: 89-127.

Tien, H. T., R. H. Barish, et al. (1998). "Supported bilayer lipid membranes as ion and
molecular probes." _Apalytical Sciences 14(1): 3-18.

Tien, H. T. and A. L. Ottova (1998). "Supported planar lipid bilayers (s-BLMs) as
electrochemical biosensors." Electrochimica Acta 43(23): 3587-3610.

Tien, H. T. and A. L. Ottova (2000). Membrane Biophysics as viewed from experimental
bilayer lipid membrtﬂea. Amsterdam; New York, Elsevier.

208

Tomalia, D. A. and P. R. Dvornic (1996). In Polymeric Materials Encyclopedia,
Salamone, J .C.LEd., CRC Press:Boca Raton 3: 1814-1830.

 

 

Tomalia, D. A., A. M. Naylor, et al. (1990). "Starbust dendrimers-Molecular level control
of size, shape, surface-chemistry, topology, and ﬂexibility from atoms to
macroscopic matter." Angewandte Chemie-Intemational Edition 102: 119-157.

Townshend, A. (1995). Encyclopedia of Analytical Science. London, Academic Press.

Tristrarn-Nagle, S., H. I. Petrache, et al. (1998). "Effect of substrate roughness on D
spacing supports theoretical resolution of vapor pressure paradox." Biophysical
Journal 74(2): A203-A203.

Trojanowicz, M. (2001). "Miniaturized biochemical sensing devices based on planar
bilayer lipid membranes." Fresenius Journal of Aﬂlyticarl Chemistry 371(2): 246-
260.

Uchiyama, S., Y. Hasebe, et al. (1993). "Enzyme-Based Catechol Sensor Based on the
Cyclic Reaction between Catechol and 1,2-Benzoquinone, Using L-Ascorbate and
Tyrosinase." Analytica Chimica Acta 276(2): 341-345.

van Tienhoven, M., J. Atkins, et al. (2002). "Human neuropathy target esterase catalyzes
hydrolysis of membrane lipids." Journal of Biological Chemisg 277(23): 20942-
20948.

VandenDriessche, T. (2005). "Gene therapy ﬂexes muscle - A European Society of Gene
Therapy commentary on progress in gene therapy for Duchenne muscular
dystrophy and amyotrophic lateral sclerosis." Journal of Gene Medicine 7(9):
1255-1256.

Vinson, V. J. (2006). "Proteins at work - Introduction." Science 312(5771): 211-211.

 

Vo-Dinh, T. and B. Cullum (2000). "Biosensors and biochips: advances in biological and
medical diagnostics." Fresenius Journal of Analytical Chemistry 366(6-7): 540-
551.

Vogel, A. I. (1998). Textbook of Quantitative Chemical Analysis. Essex, UK, Longrnan.

209

Wagner, M. L. and L. K. Tamm (2000). "Tethered polymer-supported planar lipid
bilayers for reconstitution of integral membrane proteins: Silane-
polyethyleneglycol-lipid as a cushion and covalent linker." Biophysical Journal
79(3): 1400-1414.

Wang, J ., J. M. Lu, et al. (2000). "Lab-on-a-Cable for electrochemical monitoring of
phenolic contaminants." Analytical Chemsim 72(11): 2659-2663.

Wiegand, G., N. Arribas-Layton, et al. (2002). "Electrical properties of supported lipid
bilayer membranes." Journal of Physical Chemistry B 106(16): 4245-4254.

Willner, I. and E. Katz (2000). "Integration of layered redox proteins and conductive
supports for bioelectronic applications." Angewandte Chemie-Intemational
Edition 39(7): 1180-1218.

Wright, L. L., A. G. Palmer, et al. (1988). "Inhomogeneous Translational Diffusion of
Monoclonal-Antibodies on Phospholipid Langrnuir-Blodgett Films." Biophysical
Journal 54(3): 463-470.

Xia, Y., E. Kim, et al. (1996). "Microcontact printing of alkanethiols on copper and its
application in microfabrication " Chem. Mater. 8: 601-603.

Yang, S. Y. and M. F. Rubner (2002). "Micropatteming of polymer thin ﬁlms with pH-
sensitive and cross-linkable hydrogen-bonded polyelectrolyte multilayers."
Journal of the American Chemical Society 124(10): 2100-2101.

Yang, X. M., D. A. Tryk, et al. (1996). "Surface enhanced Raman imaging of a patterned
self-assembled monolayer formed by microcontact printing on a silver ﬁlm "
Applied Physics Letters 69: 4020-4022.

Yin, P., C. J. Burns, et al. (2003). "A tethered bilayer sensor containing alamethicin
channels and its detection of arniloride based inhibitors." Biosensors &
Bioelectronics 18(4): 389-397.

Yoo, D., S. S. Shiratori, et al. (1998). "Controlling bilayer composition and surface
wettability of sequentially adsorbed multilayers of weak polyelectrolytes."
Macromolecules 31(13): 4309-4318.

210

Zebrowska, A., P. Krysinski, et al. (2002). "Electrochemical studies of blocking
properties of solid supported tethered lipid membranes on gold."
Bioelectrochemistry 56(1-2): 179-184.

Zhang, G. J ., T. Tanii, et al. (2004). "Patterning of DNA nanostructures on silicon surface
by electron beam lithography of self-assembled monolayer " Chemical
Communications 7: 786-787.

Zhang, L., R. Vidu, et al. (2002). "Electrochemical and surface properties of solid-
supported, mobile phospholipid bilayers on a polyion/alkylthiol layer pair used
for detection of antimicrobial peptide insertion." Langmuir 18(4): 1318-1331.

Zhang, L. 0., C. A. Booth, et al. (2000). "Phosphatidylserine/cholesteral bilayers
supported on a polycation/alkylthiol layer pair." Journal of Colloid and Interfacﬁ
Science 228(1): 82-89.

Zhang, L. Q., M. L. Longo, et al. (2000). "Mobile phospholipid bilayers on a
polyion/alkylthiol layer pair." Langmuir 16(11): 5093-5099.

Zheng, H., 1. Lee, et al. (2002). "Two component particle arrays on patterned
polyelectrolyte multilayer templates " ched Materials 14(8): 569-572.

211

   

Iglliljjgijjjjjjjju