.sfifpwfifi
.Ns 1.. . 4am.

-- . . . . .
awn?” _ 15%.

.53“ .1.

. . V Jihzderfldh. .
b. .avuumhwufiuwm» er, . ‘ .
11..”11fiL-Hflvhw. n V ‘ , . . V mun...» r
‘.......:. .svrfh... 5mm, .. .. . , . .a.
@wnfi?nuf #flfififi . , . . . .. .

2...:
I I. :3
.

‘4. ..
. . .39,
km” ‘ 7

x. ,
43¢... 5.1. . US.

3.3.35.2: ‘ , . .
.. r33}: . ‘ . . , ad...
. ‘ . , . . .3: .6 ,
. a

”(5:35“... 5
u 2...} I h. . (I... u
._.5....... 354:5 . z: .

a
3.. fivbw

«at. h.

3......
3
In: I k
. 3:255! .
A iv}! 10930.1! Hr

. :«mfifikus.
.52....3...‘ .
£‘fiI3’ ‘05-. ‘10

.11.},
. 9.41.4

is. .
:6? gen: $9.: 2. .. ‘
1.3va , .z.v.mm..mwml ‘

 

«Rank...

This is to certify that the
dissertation entitled

DETECTION OF EMERGING PANDEMIC INFLUENZA
STRAINS BY SURFACE PLASMON RESONANCE AND
ELECTRICALLY-ACTIVE MAGNETIC NANOPARTICLE-

BASED BIOSENSOR

presented by

TRACY KAMIKAWA

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Biosystems Engineering;

 

 

MW

 

Major Professor’s Signature

\ZL/f/g ,26, 10/0

Date

MSU is an Affirmative Action/Equal Opportunity Employer

 

LIBRARY
Michigan State
University

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KzlProlechrelelRCIDateDue.indd

 

DETECTION OF EMERGING PANDEMIC INFLUENZA STRAINS BY SURFACE
PLASMON RESONANCE AND ELECTRICALLY-ACTIVE MAGNETIC
NANOPARTICLE-BASED BIOSENSOR
By

Tracy Kamikawa

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Biosystems Engineering

2010

ABSTRACT
DETECTION OF EMERGING PANDEMIC INFLUENZA STRAINS BY SURFACE
PLASMON RESONANCE AND ELECTRICALLY-ACTIVE MAGNETIC
NANOPARTICLE-BASED BIOSENSOR
By

Tracy Kamikawa

Rapid detection technologies including surface plasmon resonance (SPR) and
nanomaterial based biosensors are emerging as sensitive, specific, and rapid diagnostic
tools for the detection of highly pathogenic viruses. This research demonstrates the novel
application of SPR and nano-biosensors for Influenza A virus (FLUAV) detection,
utilizing specificity of binding between FLUAV hemagglutinin (HA) and host sialic acid
(SA) receptors, which determines viral infectivity and transmissibility. In SPR, SA
receptors functionalize a gold sensor surface and a microfluidic system passes Over
recombinant HA, with binding indicated by a measurable increase in mass at the surface.
In the nano-biosensor, nanostructured materials serve as both magnetic concentrator and
biosensor transducer. Aniline monomer is coated around gamma iron oxide cores and
made electrically active by acid doping. The synthesized electrically active polyaniline
coated magnetic (EAM) nanoparticles are adapted in an electrochemical biosensor.
Biologically modified EAMs immunomagnetically concentrate target HA bound to SA
capture probes, and lO-minute electrochemical detection follows application of
glycan/HA/EAM complexes to screen printed carbon electrodes. Experimental results

indicate that the SPR and biosensor systems are able to detect FLUAV HA at 31.4 nM in

2% mouse serum and 1.4 uM in 10% mouse serum, respectively.

C0pyright by
TRACY KAMIKAWA

2010

ACKNOWLEDGEMENTS

I sincerely thank everyone who contributed to this research over the years. I offer
especial appreciation towards Dr. Evangelyn Alocilja for her tireless and unconditional
support, not only during my graduate program but from the beginning of my
undergraduate work, when she offered a home and family to me when mine was many
miles away. Her guidance has shaped not only my scientific career but also my personal
character, and I will carry her passion for improving the lives of others throughout my
life. I am also thankful for the patience and encouragement from Dr. Dorothy Scott, who
tirelessly supported my work at the FDA and strengthened the biologics side of my
research. Both mentors have fostered supportive and encouraging work environments,
and I thank them as well as the members of their labs for the help that they have given me
in my research. They have made this process enjoyable! I also thank the members of my
Ph.D. guidance committee: Dr. Shantanu Chakrabartty, Dr. Daniel Grooms, and Dr.
Bradley Marks for their continuous support.

I would like to also express my unending gratitude to my family and friends who
have offered guidance, patience, and love throughout this long journey. All that I have

and will accomplish is because of and for you all!

Tracy Kamikawa

iv

I would like to thank E. Torres-Chavolla, A. Pastor-Lecha, and the Center for
Advanced Microscopy at Michigan State University for providing TEM assistance and
expertise, and E. B. Setterington, H. Miller, D. Zhang, M. Huarng, and S. Pal for input
and use of cyclic voltammetry reagents. I also thank Dr. N. Razi, A. Tran-Crie, and Dr.
D. Smith of The Consortium for Functional Glycomics for helpful advice and assistance
with biotinylated carbohydrate compounds.

This work was supported by Critical Path Funding from the US Food and Drug
Administration, Center for Biologics Evaluation and Research (FDA CBER). “The
findings and conclusions in this publication have not been formally disseminated by the
Food and Drug Administration and should not be construed to represent any Agency

determination or policy.”

 

 

 

 

TABLE OF CONTENTS
LIST OF TABLES 1
LIST OF FIGURES xi
CHAPTER 1: INTRODUCTION 1
CHAPTER 2: LITERATURE REVIEW 6
2.1 INFLUENZA VIRUS ........................................................................................ 6
2.1.1 Viral Infectivity .................................................................................... 7
2.1.2 Epidemic Spread .................................................................................. 8
2.1.3 Pandemic Spread .................................................................................. 8
2.1.4 Antigenic Shifting ................................................................................ 9
2.1.5 Emergence of Highly Pathogenic Influenza Strains .......................... 10
2.1.5.1 Natural Reservoirs ................................................................. 10
2.1.5.2 Human-to-Human Transmissibility ....................................... 10
2.1.6 The Need for a Novel Biosensor Technology .................................... 12
2.2 [NFL UENZ4 IMMUNE GLOBULIN. ............................................................ 15
2.2.1 Applications ....................................................................................... 15
2.2.2 FLUIGIV as Prophylaxis ................................................................... 16
2.2.3 FLUIGIV as Treatment ...................................................................... 17
2.2.4 Current Applications of Antibody Therapy ....................................... 18
2.2.5 Potential Applications of Antibody Therapy ..................................... 19
2.2.5.1 Antivirals ................................................................................ 19
2.2.5.2 Resistant or Highly Virulent Pathogens ................................. 19
2.2.5.3 Immunocompromised Individuals ......................................... 20
2.2.5.4 Toxin Neutralization .............................................................. 20
2.2.5.5 Antibody Therapy as Combined with Chemotherapy ........... 20
2.2.6 Shortcomings of Immune Sera ........................................................... 22
2. 3. VIROLOGICAL METHODS .......................................................................... 23
2.3.1 Viral Isolation Culture ....................................................................... 23
2.3.2 Complement Fixation ......................................................................... 24
2.3.3 Hemagglutination Inhibition .............................................................. 24
2.3.4 Microneutralization ............................................................................ 25
2.3.5 Immunofluorescent Antibody Staining ............................................. 26
2.3.6 Enzyme Linked Immunosorbent Assay ............................................. 26
2.3.7 RT-PCR .............................................................................................. 27
2.4 COMMERCIAL DIAGNOSTIC TEST KITS .................................................. 28
2. 5 BIOSENSORS ................................................................................................ 31
2.5.1 Surface Plasmon Resonance Sensors ................................................. 31
2.5.2 Quartz Crystal Microbalance ............................................................. 33
2.5.2.1 Applications ........................................................................... 34
2.5.2.2 Rupture Event Scanning ........................................................ 35

vi

2.5.3 Colorimetric Sensors .......................................................................... 36

 

 

2.5.4 Gold Nanoparticles and Quantum Dots ............................................. 36
2.5.4.1 Shortcomings ......................................................................... 37

2.5.5 Microarrays ........................................................................................ 38

2.6 TOWARDS IMPROVED TECHNOLOGY ..................................................... 39
2.6.1 Requirements of Biosensor Technology ............................................ 39

2. 7 CONDUCTING POLYMERS ......................................................................... 41
2.7.1 Electronic Structures of Conducting Polymers .................................. 42
2.7.2 Nanoparticles as Biological Sensory Labels ...................................... 43

2.7.3 Polymers in Biosensors ...................................................................... 45
2.7.4 Conducting Polymers as Transducers ................................................ 46

2.7.5 Polyaniline ......................................................................................... 46
2.7.5.1 Synthesis ................................................................................ 47

2.7.5.2 Forms ..................................................................................... 49

2.7.5.3 Electrical Conduction Properties ........................................... 51

2.7.5.4 Electrochemistry .................................................................... 56

2.7.6 Electrically Active Magnetic Polyaniline .......................................... 5 8
2.7.7 Cyclic Voltammetry ........................................................................... 59
CHAPTER 3: RESEARCH HIGHLIGHTS - 62
3.1 RESEARCH NOVELTY. ................................................................................. 62
3.2 RESEARCH SIGNIFICANCE ........................................................................ 65
3.3 HYPOTHESJS ................................................................................................ 67
3.4 RESEARCH OBJECTIVES ............................................................................ 68
CHAPTER 4: RESEARCH MATERIALS AND METHODS 69
4.1 OBJECTIVE I .' Surface Plasmon Resonance—based binding assay ............ 69
4.1.1 SPR Assay Design ............................................................................. 69
4.1.1.1 Reagents and Chemicals ........................................................... 69

4.1.1.2 Equipment ................................................................................. 70

4.1.1.3 SPR (Biacore) Chip Preparation and Immobilization ............... 70

4.1.1.4 Chip Regeneration .................................................................... 73

4.1.2 SPR Binding between Glycan Receptors and Hemagglutinin ........... 73
4.1.2.1 Binding Assay and Sensitivity Testing ..................................... 73

4.1.2.2 Specificity Testing .................................................................... 74

4.1.3 Characterization Studies .................................................................... 74
4.1.3.1 HA Receptor Binding Domain Binding Assessment ................ 74

4.1.3.2 HA Preparation ......................................................................... 74

4.1.3.3 Serum Experiments ................................................................... 75

4.1.3.4 Statistical Analysis .................................................................... 75

4.2 OBJECTIVE 2: SPR to Detect Ab-Mediated Binding Inhibition ................... 76
4.2.1 SPR Inhibition Assay Design ............................................................. 76
4.2.1.1 Reagents .................................................................................... 76

4.2.1.2 HA/glycan Neutralization by Monoclonal Antibody ................ 76

4.2.1.3 Specificity Testing .................................................................... 77

4.2.2 Characterization Studies .................................................................... 78

vii

4.2.2.1 Antibody Testing ...................................................................... 78

 

4.2.2.4 Serum Experiments ................................................................... 78
4.2.2.5 Statistical Analysis .................................................................... 78

4. 3 OBJECTIVE 3: H5N1—Targeted Biosensor Design ..................................... 80
4.3.1 Biosensor Design ............................................................................... 80
4.3.1.1 Reagents and Chemicals ........................................................... 80
4.3.1.2 Methodology of Supporting SPR Assay ................................... 81
4.3.1.3 Biosensor Architecture .............................................................. 82

' 4.3.1.4 Gold Nanoparticle Synthesis ..................................................... 82
4.3.2 Electrically Active Polyaniline Coated Magnetic Nanoparticles ....... 83
4.3.2.1 EAM Synthesis ......................................................................... 83
4.3.2.2 EAM Nanoparticle Characterization ......................................... 83
4.3.2.3 EAM Immunofunctionalization ................................................ 84
4.3.2.4 EAM Structural Characterization ............................................. 85
4.3.2.5 Spectral Analysis ...................................................................... 85
4.3.3 Biosensor Fabrication ........................................................................ 85
4.3.3.1 SPCE Modification ................................................................... 85
4.3.4 Preconcentration Preparation Technique ........................................... 86
4.3.4.1 Sample Preparation ................................................................... 86
4.3.4.2 Capture Experiments ................................................................. 86
4.3.5 Stepwise Preparation Technique ........................................................ 87
4.3.5.1 Sample Preparation and Capture Experiments ......................... 87
4.3.6 Biosensor Testing ............................................................................... 91
4.3.6.1 Testing Apparatus ..................................................................... 91
4.3.6.2 Detection and Data Analysis ..................................................... 91
4.3.6.3 Sensitivity and Specificity Testing ........................................... 94
4.3.6.4 Complex Matrix Testing ........................................................... 94
4.3.6.5 Statistical Analysis .................................................................... 95
4.4 OBJECTIVE 4: Biosensor to Distinguish a2,3 v. (12,6 Receptor Binding....96
4.4.1 Biosensor Design ............................................................................... 96
4.4.1.1 Reagents and Chemicals ........................................................... 96
4.4.1.2 Biosensor Fabrication ............................................................... 97
4.4.2 Biosensor Testing ............................................................................... 97
4.4.2.1 Sample Preparation ................................................................... 97
4.4.2.2 Capture Experiments ................................................................. 97
4.4.2.3 Detection and Data Analysis ..................................................... 98
4.4.2.4 Sensitivity and Specificity Testing ........................................... 98
4.4.2.5 Complex Matrix Testing ........................................................... 99
4.4.2.6 Statistical Analysis .................................................................... 99
CHAPTER 5: RESULTS AND DISCUSSION ........ - 101
5.1 OBJECTIVE I : Surface Plasmon Resonance—Based Binding Assay ......... 101
5.1.1 SPR Binding Assay Design ............................................................ 101
5.1.1.1 Glycan Immobilization and Chip Stability ............................. 101
5.1.2 SPR Binding Between Glycan Receptors and Hemagglutinin ........ 103
5.1.2.1 Confirmation of H5 Recognition ............................................ 103

viii

5.1.2.2 Chip Regeneration .................................................................. 104

 

 

5.1.2.3 Specificity Testing .................................................................. 104

5.1.2.4 Clade Specificity ..................................................................... 105

5.1.2.5 HA Receptor Binding Domain Binding Assessment .............. 105

5.1.2.6 HA Preparation ....................................................................... 108

5.1.2.7 Serum Experiments ................................................................. 109

5.1.2.8 Structural Morphology Characterization ................................ 110

5. 2 OBJECTIVE 2: SPR to Detect Ab-Mediated Binding Inhibition ................. 111
5.2.1 SPR Neutralization Assay Design ................................................... 1 11
5.2.1.1 Neutralization Ability of Anti-H5 Monoclonal Antibody ...... 111

5.2.1.2 Neutralization Specificity ....................................................... 111

5.2.1.3 Antibody Testing: Anti-HA versus Anti-HA2 ........................ 114

5.2.1.4 Serum Experiments ................................................................. 115

5. 3 OBJECTIVE 3: H5N1—Targeted Biosensor Design ................................... I I 7
5.3.1 Biosensor Design ............................................................................. 117
5.3.1.1 Supporting SPR data ............................................................... 117

5.3.1.2 Electrochemical Detection ...................................................... 119

5.3.1.3 Biosensor Sensitivity .............................................................. 124

5.3.1.4 Magnetic Separation by EAMs ............................................... 124

5.3.1.5 Preparation Effects .................................................................. 126

5.3.1.6 Nonspecific Binding ............................................................... 127

5.3.1.7 Biosensor Specificity .............................................................. 129

5.3.1.8 Structural Morphology Characterization ................................ 130

5.4 OBJECTIVE 4: Biosensor to Distinguish a2,3 v. a2,6 Receptor Binding. 133
5.4.1 Biosensor Design ............................................................................. 133
5.4.1.1 Glycan Sequences .................................................. 133

5.4.1.2 Avian FLUAV-Targeted Biosensor ........................................ 133

5.4.1.3 Human FLUAV-Targeted Biosensor ...................................... 134

5.4.1.4 Specificity ............................................................................... 134
CHAPTER 6: CONCLUSION AND FUTURE RESEARCH - 137
APPENDIX A: STATISTICAL ANALYSIS RESULTS 142
A. 1 ANOVA ANALYSIS OF SPR RESULTS ....................................................... 142
A.2ANOVA ANALYSIS OF BIOSENSOR RESULTS ......................................... 151
REFERENCES - ..... - -- 157

 

ix

LIST OF TABLES

Table 1. Applications for which antibody therapy combined with

chemotherapy has shown preliminary effectiveness ................................... 21
Table 2. SPR assays: Gaps in research ..................................................................... 63
Table 3. Biosensor technology: Gaps in research ..................................................... 64
Table 4. Glycan structure and binding predictions ................................................... 72
Table 5. Biotinylated saccharide sequences and predicted binding to H5N1 ......... 126
Table A-1. SPR Results: Least Square Means ......................................................... 142
Table A-2. SPR Results: Estimates .......................................................................... 148
Table A-3. Biosensor Results: Average AQ and Standard Deviation ..................... 151
Table A-4. Biosensor Results: Least Squares Means .............................................. 153
Table A—5. Biosensor Results: Estimates ................................................................. 154

LIST OF FIGURES

Images in this dissertation are presented in color.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Schematic representation of the structure of the Influenza A virion (adapted
and modified from Zhuang, 2009). ........................................................... 7

General structure of glycan receptor with terminal sialic acids (adapted and
modified from Blixt et al., 2004). ............................................................ 11

SPR theory based on changes in refractive index due to
immobilized target, with sensorgram output (adapted and modified
from GE Healthcare, 2010). ..................................................................... 33

Butterworth van Dyke (BVD) electrical model of a quartz crystal
resonator (adapted and modified from Eun et al., 2002). ........................ 36

Aniline in the presence of hydrochloric acid, oxidized with

ammonium peroxydisulfate to yield polyaniline emeraldine
hydrochloride (adapted and modified from Stej skal and Gilbert,

2002). ....................................................................................................... 48

Forms of polyaniline (adapted and modified from Stej skal et al., 1996) 50

Geometric structure of polyaniline in polyemeraldine state. (a)

Before protonation, (b) formation of bipolarons after 50%

protonation, (c) formation of polarons after 50% protonation, and (d)
polaron lattice formed after polaron separation (adapted and

modified from Stafstrom et al., 1987) ..................................................... 53

Deprotonation of polyaniline in presence of chloride (alkaline
medium). (a) Polyaniline emeraldine salt is converted to (b)
polyaniline emeraldine base (adapted and modified from Stej skal

and Gilbert, 2002) .................................................................................... 55
Schematic cyclic voltammogram for redox couple undergoing single

electron oxidation-reduction process ....................................................... 6O
Biacore SA sensor chip and instrumentation ............................................ 72

Testing schematic. (a) Screen-printed carbon electrode (SPCE)

consisting of two electrodes: carbon working electrode and

silver/silver chloride counter/reference electrode, (b) schematic of

the three electrode voltammetry system (adapted and modified from

Bard and Faulkner, 2000). ....................................................................... 88

xi

Figure 12. Testing schematic. Stepwise preparation method .................................... 89
Figure 13. Testing schematic. Preconcentration preparation method. ....................... 90
Figure 14. SPCE and potentiostat setup ..................................................................... 93

Figure 15. Glycan/HS binding experiments. (a) Triplicates of H5 dilutions
binding to 3’SLN; Regeneration: 60 s of 10mM glycine pH 2.5 and
30 s of 50mM NaOH at 100 til/min, (b) triplicates of H5 dilutions
binding to 3’SLN; Regeneration: 60 s of 10mM glycine pH 2.5 and
18 s of 50mM NaOH at 100 III/min, and (c) triplicates of H5
dilutions binding to 3’ SLex; Regeneration: 60 s of 10mM glycine
pH 2.5 and 30 s of 50mM NaOH at 100 til/min. ................................... 102

Figure 16. H5 Indonesia and H3 Wyoming binding to 3’SLN. Single replicate
shown for clarity. (a) H5 (A/V ietnarn/ 1203/04) 140nM; H5
(A/IndonesiaS/OS) 140nM; H5 Indonesia 140nM + anti-H5
Indonesia 1:250, 1:500, and 1:1000; anti-H5 Indonesia 1:250 and (b)

H3 (A/Wyoming/3/03) at 286nM, 94.3nM, 31.4nM, 10.6nM, and
3.53nM ................................................................................................... 103

Figure 17. (a) H5 at 286nM, 94.3nM, 31.4nM; HA1 H5 at 286nM, 94.3nM,
31.4nM, and (b) H5 at 286nM, 94.3nM, 31.4nM; HA1 H3 at 286nM,
94.3nM, 31.4nM .................................................................................... 107

Figure 18. H5 1.4uM pretreatments. (a) Tween-20 (b) heat treatment at 37
degrees C overnight, 37 degrees C for 4 h, 37 degrees C with
bromelain and 2-ME for 4 h, and 37 degrees C with bromelain for 4
h .............................................................................................................. 108

Figure 19. TEM imaging of (a) synthetic glycans and (b) purified recombinant
H5 HA .................................................................................................... 110

Figure 20. Neutralization experiments. (a) 3’SLN/HS neutralization by anti-H5
monoclonal antibody: H5 140nM; H5 140nM + anti-H5 1:4000,
1:2000, 1:1000, 1:500, 1:250; anti-H5 1:250 only, (b) 3’SLN/HS
binding inhibition by anti-H1 (HlNl/Pan): H5 140nM; H5 140nM +
anti-H1 1:250; anti-H1 1:250 only, and (c) 3’SLN/HS binding
inhibition by anti-H3 (A/Shandong/9/93): H5 140nM; H5 140nM +
anti-H3 1:500; anti-H3 1:250 only ......................................................... 113

xii

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Antibody binding to 3’SLN/HS precomplex. (a) Injection 1: H5

140nM, Injection 2: buffer, anti-HA2 H5 1:500, 1:1000, or 1:2000;
Injection 1: anti-HA2 1:250, Injection 2: buffer, and (b) Injection 1:

H5 140nM, Injection 2: anti-H5 neutralizing monoclonal antibody or
anti-HA2 H5 1:250 ................................................................................ 114

Serum effects. . (a) H5 at 286nM, 94.3nM, 31.4nM, 10.6nM, and

3.53nM prepared in 2% mouse serum binding to 3’SLN and (b) H5
140nM, H5 140nM + anti-H5 1:4000, H5 140nM + anti-H5 1:2000,

H5 140nM in 1% serum, H5 140nM + anti-H5 1:4000 in 1% serum,

H5 140nM + anti-H5 1:2000 in 1% serum ............................................ 116

Supporting SPR results. . (a) H5 140nM binding to H5-specific

glycan 3’SLex, as inhibited by 1% mouse serum and anti-H5

monoclonal antibody 1:500, (b) H5 140nM binding to H5-specific

glycan 3’SLN, as inhibited by 1% mouse serum and anti—H5

monoclonal antibody 1:500, (c) H5 140nM binding to H5-specific

glycan 3’SLex, as inhibited by cross-reactivity of anti-H1 polyclonal
antibody; H5* 140nM binding to 3’SLex, and (d) antibody testing on
HS-specific glycan 3’SLN ..................................................................... 118

Stepwise preparation method. (a) Delta Q values of (A) 3’SLex

100uM + H5 1.4uM, (B) 3’SLex IOOuM + H5 700nM, (C) 3’SLex
IOOuM + H5 360nM, (D) CT/Sda SOOuM + H5 1.4uM, (E) 3’SLN
500uM + no HA, (F) no glycan + H5 1.4IIM, and (G) no glycan + no
HA, and (b) CV of 3’SLex IOOuM + H5 1.4uM ................................... 122

Preconcentration preparation method. (a) Delta Q values of (A)

3’SLex 100uM + H5 1.4uM + 10% mouse serum, (B) 3’SLex

100uM + H5 700nM + 10% mouse serum, (C) 3’SLex lOOuM + H5
360nM + 10% mouse serum, (D) GT3 IOOuM 0+ H5 1.4uM + 10%
mouse serum, (E) 3’ SLex 100uM + no HA, (F) no glycan + H5

1.4uM + 10% mouse serum, and (G) no glycan + no HA, and (b) CV

of 3’SLex 100uM + H5 1.4uM + 10% mouse serum ............................ 123

Cyclic voltammetry results. (a) H5 concentration study as a function

of preparation method and comparison to negative controls and

blanks, as numbered and described in Table A-3. Group (A) 1, (B) 2,

(C) 3, (D) 24, (E) 25, (F) 27, (G) 26, (H) 9, (I) 10, (J) 11, (K) 14, (L)

13, (M) 12, (N) 15, (O) 16, (P) 17, (Q) 18, and (R) 19. (b) Response

for H5 1.4uM using different preparation methods. (A) 1, (B) 8, (C)

9, (D) 20, and (E) 21. For the respective samples, mean AQ i SD, n

= 3 (SD = standard deviation, n = no. of replicates). ............................. 125

xiii

Figure 27. Comparison of different preparation methods. (A) 3’SLex IOOuM +
H5 1.4uM, stepwise, (B) 3’SLex IOOuM + H5 1.4uM,
preconcentration, (C) 3’SLex 100uM + H5 1.4uM + 10% mouse
serum, preconcentration, (D) 3’SLex 100uM + H5* 1.4uM + 10%
mouse serum, preconcentration, and (E) 3’SLN 100uM + H5 1.4p.M,
stepwise .................................................................................................. 127

Figure 28. Specificity investigation using Hl-based negative controls and
preconcentration preparation. (A) 3’SLex lOOpM + H5 1.4uM +
10% mouse serum + anti-HS—EAMS, (B) 3’SLex 100uM + H1
1.4uM + 10% mouse serum + anti-HS—EAMS, (C) 3’SLex lOOuM
+ H1 1.4uM + 10% mouse serum + anti-Hl—EAMS, (D) no glycan
+ H1 1.4uM +10% mouse serum + anti-Hl—EAMS, and (E) no
glycan + no HA + anti-Hl—EAMS ....................................................... 130

Figure 29. TEM imaging. (a) TEM and electron diffraction micrograph (inset)
of EAM polyaniline nanoparticles with gamma iron (III) oxide cores,
(b) TEM of EAMs immunofunctionalized with anti-H5 antibody, (c)
3’SLex/H5/anti-H5—EAM complex, magnetically separated and
washed, with H5 prepared with 10% mouse serum, and (d) 3’SLex
/H5/anti-H5—EAM complex, magnetically separated and washed,
with H5 prepared without serum ............................................................ 132

Figure 30. H5 binding to (12,3 versus a2,6-linked glycan receptors using
preconcentration method. (A) 3’S-Di-LN IOOuM + H5 1.4uM +
10% mouse serum, (B) 3’S-Di-LN IOOuM + H5 700nM + 10%
mouse serum, (C) 6’S-Di-LN IOOpM + H5 1.4uM +10% mouse
serum, and (D) aHS—EAMS only ........................................................ 135

Figure 31. H1 binding to a2,3 versus a2,6-linked glycan receptors using
preconcentration method. (a) 6’S-Di-LN IOOpM + H1 1.4uM + 10%
mouse serum, (b) 6’S-Di-LN IOOpM + H1 700nM + 10% mouse
serum, (c) 3’S-Di-LN 100uM + H1 1.4uM + 10% mouse serum, and
(d) otHl—EAMs only ............................................................................ 136

xiv

CHAPTER 1: INTRODUCTION

On June 11, 2009, the World Health Organization declared a global pandemic of
novel HlNl Influenza A virus (FLUAV). Novel HlNl FLUAV was first Observed in
humans in Mexico and the United States beginning in March, 2009, and since then the
virus has spread rapidly across all 50 states. At the time of the pandemic declaration,
novel HlNl FLUAV had been reported in 70 countries across the globe (CDC, 2009).
The emergence of swine origin HlNl from natural animal reservoirs brought the
devastating capabilities of F LUAV viruses into public consciousness, although these
viruses have been plaguing global economics for decades. Prior to the H1N1 pandemic,
global pandemics had previously occurred throughout history, with varying causes and
consequences. Of particular note was the H1N1 pandemic of 1918, referred to as the
“Spanish flu,” which was extremely virulent and led to 20-40 million deaths worldwide
(Reid et al., 2001).

The HlNl pandemics of 1918 and 2009 differ in their characteristics. Both strains
caused pandemics, but for different reasons. The 1918 HlNl strain was widespread,
infecting one third of the world’s population, and was also the most virulent of all
historical pandemic strains, leading to case fatality rates of >2.5% (Bumet and Clark,
1942; Marks and Beatty, 1976; Reid et al., 2001; Taubenberger and Morens, 2006). The
2009 HlNl strain also experienced rapid spread, but was relatively less severe, with the
cumulative total number of cases worldwide at 380,000 with a <1% death rate (as of 4
October 2009; see WHO, 2009). The ability of a F LUAV strain to transmit from human-
to-human is thus essential for a pandemic to occur, whether or not the strain is highly

lethal. Typically, FLUAV strains must be specific for the human receptors high in the

respiratory tract, so that transmission by aerosol or surface contact is facilitated, as
opposed to those strains specific for the receptors deeper in the respiratory tract. Animal
FLUAV strains are typically specific for the less-accessible receptors.

The highly pathogenic avian influenza virus subtype H5N1 is one such strain.

HSN 1 has caused serious losses in the poultry industry and continues to affect wild fowl
populations across the world, but particularly throughout Asia. The receptor recognition
of H5N1 does not lend itself to human-to-human transmission. However, the ability of

F LUAV viruses to easily mutate could lead a highly pathogenic avian strain to achieve
human infectivity via specificity for the upper respiratory tract receptor, eventually
leading to a pandemic with casualties of 1918 proportions. This could occur naturally via
existing animal reservoirs, or as a result of bioterrorism efforts. Pathogenic H5N1 has
been generated in the laboratory, and by similar methods of recombinant DNA
technology, a highly pathogenic FLUAV could also gain human-to-human
transmissibility, and thus pandemic potential (Hatta et al., 2001a,b).

H5 has been identified as a subtype of FLUAV that could most likely be transmitted
to humans (Webby and Webster, 2003). In fact, in those cases where humans have been
infected by close contact with H5Nl infected birds, >30% of cases were fatal (Webby
and Webster, 2003). If H5N1 were to become more easily transmitted from human-to-
human, like an epidemic of “seasonal flu” which is carried by sneezing and casual
contact, the world population would be at risk for widespread and highly fatal infection.

Vaccine development and production have proven challenging, especially when
demand is high in the face of a worldwide crisis. Vaccine design is based on assumptions

of which strains could gain prevalence in the following year, and as demonstrated by the

 

 

inability of the 2009 vaccine to prevent the H1N1 pandemic, these assumptions are not
always correct.

In association with F LUAV epidemics, the United States experiences annual direct
and indirect costs of up to $12 billion, associated with doctor’s office visits,
hospitalizations, medications, and work productivity losses (Solvay, 2010). Epidemic
FLUAV strains are in general far less virulent and severe than pandemic strains, and the
costs associated with a pandemic could be far greater. As reported by the US.
Department of Health and Human Services, production of over 125 million doses of
pandemic HlNl vaccine has cost approximately $8 billion (N ewborg, 2009). During the
height of the H1N1 pandemic, many public schools were forced to halt all operations to
prevent further spread. The Brookings Institute predicts that a nationwide school closure
for four weeks could lead to $10-47 billion in lost economic activity, which represents
0.1«0.3 percent of the GDP. The Congressional Budget Office predicts that loss of GDP
could reach 4.5 percent in the event of a pandemic on the scale of the 1918 Spanish
Influenza pandemic (Amico, 2009).

Understanding the infectivity of F LUAV is essential in preventing the next pandemic. A
method of rapidly testing emerging pandemic viruses could be the first line of defense
against a historically non-human-transmissible strain which has gained human
transmissibility by natural or unnatural routes. Targeted vaccine development could then
proceed before wide spread, or other preventative measures such as quarantine or
medicinal treatment could be undertaken. Current human and animal diagnostic methods
for virus detection are generally based on internationally recognized methods of isolation

culture with immunocytological confirmation of viral antigen (Alexander et al., 2005;

Charlton et al., 2009). The entire process could take up to 21 days, involving high costs
of reagents and labor. The development of rapid detection devices has thus become
increasingly necessary for environmental and agricultural disease surveillance, and to
provide early detection of potentially human pandemic FLUAV strains for limiting
spread and severity.

Biosensors are attractive alternatives for early identification of infectious pathogens
such as F LUAV, and offer low cost, speed, and ease of operation as compared to their
conventional counterparts. A wide range of detection platforms and targets are under
investigation, spanning all fields of public health, and major advances have been made in
recent years, with evolution still continuing.

Nanotechnology has offered a whole new world of possibilities to biodetection
systems. Nanostructured materials such as gold and magnetic nanoparticles have been
1 demonstrated to be effective biosensor transducer materials. A recent advancement is the
development of nanostructures with both magnetic and conductive properties. Typically,
these are presented with a core/shell (c/s) of magnetic/electrically active materials. One
application exploits the conductive nature of polyaniline and the magnetic nature of iron
(III) oxide to generate electrically active polyaniline coated magnetic nanoparticles
(EAMs) with optimal combined properties of strength, flexibility, and electrochemical
activity. Current literature indicates that these EAMs have not yet been applied in the
detection of highly pathogenic F LUAV.

This dissertation describes the characterization of binding between the FLUAV
surface glycoprotein responsible for host infectivity, hemagglutinin (HA), and the host

cell carbohydrate (glycan) receptor. Surface plasmon resonance (SPR) and an

electrochemical biosensor platform were investigated as compatible assays using the
same glycan/HA pairs. Because a disposable biosensor technology has not yet been
reported, the SPR system was utilized to ploy the interactions between H5N1 HA and
appropriate glycan receptors, as well as to identify HS-targeted antibodies with the ability
to neutralize this binding. Serum matrix effects were also evaluated in the SPR system
performance. Once high avidity binding pairs were identified on the SPR system, an
electrochemical biosensor was designed and fabricated, utilizing immunofunctionalized
EAMs as both the magnetic concentrator of the target glycan/HA complex and the
transducer in the electrochemical detection of EAM nanoparticles on a screen printed
carbon electrode. The binding between Influenza hemagglutinin and glycan receptors was
then characterized on the biosensor platform for correlation to SPR results. Once the
biosensor platform fabrication technique was established, other glycan/HA pairs were

investigated, including human targets.

CHAPTER 2: LITERATURE REVIEW

2.1 INFLUENZA VIRUS

Influenza virus A (FLUAV) is an acute viral disease agent of the respiratory tract
(Stevens et al., 2006a), which is classified as a genus of the Orthomyxoviridae family
(WHO, 2006). Millions of people worldwide are affected annually by FLUAV, either by
epidemics of “seasonal flu,” or, less commonly, by infection with a pandemic strain such
as H5Nl, “bird flu,” or HlNl, “swine flu.” The 2009 swine-origin H1N1 pandemic
exemplifies the speed with which a human-transmissible FLUAV can spread worldwide.
The strain had circulated in pig herds for decades, but once humans became infected, the
virus achieved global spread in a matter of weeks (Michaelis, 2009).

Hemagglutinin (HA) and neuraminidase (NA) are integral membrane proteins.
M2 ion channel protein is inserted through the lipid bilayer. Virion matrix protein Ml
underlies the lipid bilayer. The segmented genome exists as eight RNA single-strands, to

form ribonucleoproteins with transcriptase proteins PBl, PB2, and PA (Figure 1).

_—r"-'

Hemagglutinin (HA)
Neuraminidase (NA)

Lipid bilayer

M2 (Ion channel)

 

Figure 1. Schematic representation of the structure of the Influenza A virion
(adapted and modified from Zhuang, 2009).

2.1.1 Viral Infectivity

The viral surface glycoproteins hemagglutinin (HA) and neuraminidase (NA) are
used to name and characterize each FLUAV strain, as they are largely responsible for
viral infectivity (Stevens et al., 2006a). HA mediates FLUAV host specificity and host
cell entry (Stevens et al., 2006a; Wiley and Skehel, 1987). Of the sixteen known HA and
nine known NA serotypes, only three HA and two NA have adapted sufficiently to
become pandemic in humans. Pandemics recorded in history have included H1N1 in
1918 and 2009, H2N2 in 1957, and H3N2 in 1968. Birds are thought to act as the main
reservoir for FLUAV, because all identified serotypes circulate in the avian population
(Neumann and Kawaoka, 2006; Stevens et al., 2006a; Stevens et al., 2006b;

Taubenberger et al., 2005).

2.1.2 Epidemic Spread

Epidemics of FLUAV are largely facilitated by the ease and speed of human-to-
hmnan transmissibility by aerosol (Wright and Webster, 2001). These “seasonal flu”
epidemics are responsible for 36,000 deaths and 200,000 hospitalizations in the US.
annually, leading to costs for the nation of over $10 billion (HSC, 2005). These
epidemics are caused by antigenic drifi, by which the HA undergoes relatively minor
changes that result from the selection of mutant viruses by antibodies generated against
the prominent HA antigenic type currently circulating in the human population (Wright
and Webster, 2001). Each year, the body produces antibodies against the prominent
F LUAV strain, either naturally or after vaccination. Once the targeted HA antigen
mutates to a form no longer recognizable by the antibodies, the body loses resistance and
an annual epidemic results (NIAID, 2005).

2.1.3 Pandemic Spread

Of even greater concern are FLUAV pandemics, or global disease outbreaks,
which result in high mortality rates, (WHO, 2004; Wright and Webster, 2001). The
FLUAV strains that have adapted to produce human pandemics include HlNl (1918,
2009), H2N2 (1957), and H3N2 (1968) (Stevens et al., 2006a). The most virulent of these
was the 1918-1919 outbreak of H1N1 , known as the “Spanish flu,” which resulted in 20-
40 million deaths worldwide (Reid et al., 2001). It is believed that the extreme virulence
of the virus was the main cause of the high mortality rate, as opposed to other factors
such as lack of antimicrobial agents. RNA sequencing has not revealed the reason for

such high pathogenicity (Neumann and Kawaoka, 2006). Such human pandemics are

caused by antigenic shifts in the HA of FLUAV, which occur less frequently than the
antigenic drifts associated with FLUAV epidemics (Wright and Webster, 2001).
2.1.4 Antigenic Shifting

Antigenic shifts result from a replacement of the genomic RNA segment encoding
HA, and allow F LUAV strains to jump fi'om one animal species to another. An antigenic
shift may oCcur by one of three ways. In the first route, an avian strain and a human strain
are both passed to an intermediate host such as a chicken or pig. When the viruses co-
infect the same cell, genes from the avian and human strains mix to yield a new strain
that can spread to humans (N IAID, 2005). The avian strain provides the new HA
genomic segment via reassortment with the human strain (WHO, 2006). FLUAV lends
itself to such reassortment because the segmented nature of its genome allows for the
exchange of entire genes between different viral strains when they cohabitate the same
cell (Shaw et al., 1992). When the reassortant strain further evolves to gain human-to-
human transmissibility, a pandemic could arise (N IAID, 2005). It is commonly presumed
that the 1918, 1957, 1968, and 2009 pandemic strains were the result of such
reassortrnents (Stevens et al., 2006a). An antigenic shift may also occur when an avian
strain jumps directly from a bird or duck to a human without undergoing a genetic change,
or when an avian strain jumps from a bird to an intermediate host and then to humans
without undergoing a genetic change. When any of these new strains gain human-to-

human transmissibility, a FLUAV pandemic could arise (N IAID, 2005).

2.1.5 Emergence of Highly Pathogenic Influenza Strains

2.1.5.1 Natural Reservoirs

The emergence of highly pathogenic F LUAV from natural animal reservoirs is a
likely threat, as evidenced by the 2009 swine-origin HlNl pandemic. Historically, human
infections by avian FLUAV strains including H5N1, H9N2, H7N7, and H7N3 have been
more commonplace, and could serve as a model for animal-to-human infection
mechanisms with applicability for avian, swine, or other emerging primary hosts
(Matrosovich M.N. et al., 2004). H5N1 in particular has become epizootic in domestic
fowl throughout Asia since 2003, is spreading to European and Afiican bird populations,
and has been confirmed in human cases. Since 2003, 385 confirmed human cases of
H5N1 have been reported to the World Health Organization (WHO), including 243
deaths (as of 19 June 2008, see WHO, 2008). While human-to-human transmission has
not yet shown effectiveness, the high mortality rate is cause for concern (Stevens et al.,
2006c).
2. 1. 5.2 Human-to-Human T ransmissibility

Human-to-human transmission of these infections, and thus the potential for
pandemic, is dependent upon HA receptor specificity (Matrosovich M.N. et al., 2004;
Stevens J. et al., 2006a). HA mediates F LUAV host specificity and host cell entry, and
binds to glycan receptors with terminal sialic acids (Stevens J. et al., 2006a; Wiley and
Skehel., 1987). Avian FLUAV preferentially bind to sialic acids (Figure 2) connected to
galactose by 0:23 linkages on lower respiratory tract ciliated cells, whereas human
FLUAV preferentially bind to (126-linked sialic acids on nonciliated cells found in the

nose and throat (Matrosovich M.N. et al., 2004; Stevens J. et al., 2006a).

10

 

 

 

:Man

O
.
:Sia I :GIcNAc

 

 

 

 

Figure 2. General structure of glycan receptor with terminal sialic acids (adapted
and modified from Blixt et al., 2004).

The specificity for the 0L2.6 receptor of isolates from the 1957 (H2N2) and 1968
(H3N2) pandemics is the primary reason that human-to-human transmission was possible
(Matrosovich M.N. et al., 2004; Tumpey T.M. et al., 2007). It can be assumed that the
1918 (HlNl) pandemic FLUAV strains were also specific for the (12.6 receptor.
Conversely, human infection by avian F LUAV that preferentially bind to the a2.3
receptor will not transmit efficiently from human-to-human and will thus not yield a
pandemic. The H5Nl outbreak in Hong Kong in 1997 was specific for the G23 receptor
(Matrosovich M.N. et al., 2004). In this instance, the F LUAV strain was highly virulent
but transmission between humans did not occur (Class et al., 1998; Suarez et al., 1998;
Subbarao et al., 1998; Yuen et al., 1998).

Humans possess both a2.3 and Q26 receptors, with a greater density of (x23 in the
lower respiratory tract, and a2.6 in the upper respiratory tract. It is believed that in those

cases where avian F LUAV has infected humans, via the handling of dead poultry or other

11

close contact with contaminated materials or infected people, the avian virus has not
switched receptor specificity, but has bound its 0(2.3 receptors deep in the respiratory
tract, where they are able to efficiently replicate. This may explain the so far inefficient
human-to-human transmission of H5N1. If an avian FLUAV mutates to achieve
recognition of the human 0(2.6 receptor in the upper airway, the virus will be easily
transmitted by the sneezing and coughing seen with current epidemics (Shinya et al.,
2006). As shown in the Vietnam and Turkey outbreaks, the virus obtained a greater
affinity to the a2.6 receptors due to mutations or reassortment of genes with circulating
human F LUAV (Krug, 2003; Stevens J. et al., 2006c). Indeed, the 1918, 1957, and 1968
pandemic FLUAV strains are thought to have been alterations of avian FLUAV strains
with a2.3 specificities (Tumpey TM. et al., 2007).
2.1.6 The Need for a Novel Biosensor Technology

FLUAV continues to circulate in Asian poultry markets, and in all plausibility, once
a lethal FLUAV strain is transmitted from poultry to humans it may not be possible to
prevent the virus from acquiring human-to-human transmissibility via antigenic shifts
(Krug, 2003; Webster et al., 2002). A detection method for rapidly identifying 0(2.6
specificity will be essential in this event, so that an avian FLUAV not previously known
to be transmissible among humans can be identified as pandemic immediately after such
transmissibility is acquired. Preventative measures against widespread human infection
can then occur in a timely manner. For example, patients infected with (12.6 strains may
be quickly treated with antiviral drugs or placed under stringent quarantine before the
infection is passed along to others (Krug, 2003). Pandemic viruses thus emerge from

avian progenitors via HA receptor specificity alteration, but because the underlying

12

[I

mechanism of these shifts is unclear, it is difficult to predict such a progression with
certainty. Thus, a method of rapidly determining the binding specificity of an emerging
FLUAV strain will help to predict the potential for human-to-human transmission and the
likely emergence of a pandemic strain.

The development of a biosensor technology that functions on the differential and
specific binding of F LUAV HA to host SA is a significant initiative with regard to
disease monitoring and homeland security. Such a FLUAV biosensor can probe the
infection path of an emerging FLUAV strain, providing a starting point for environmental
virus monitoring useful in tracking the course of virus circulation. Should F LUAV be
engineered for use as a bioterrorism agent, such a biosensor could detect the appearance
of a virus strain lethal to humans and perhaps possessing a novel HA subtype (Amano
and Cheng, 2005).

The high death rates associated with past pandemics and the ease of human-to-human
transmission could make human FLUAV attractive to bioterrorists. Even if a progression
from (12.3 to a2.6 specificity does not occur by the natural mechanisms described above,
lethal human FLUAV may be generated by the reverse genetic system by which
transfection of multiple DNAs occurs without a helper virus (F odor et al., 1999;
Neumann et al., 1999). Recombinant DNA techniques have been used to generate
pathogenic H5N1 virus in the laboratory (Hatta et al., 2001a,b), and it is likely that such
methods could also transfer human-to-human transmissibility or antiviral resistance to a
pandemic avian influenza virus (Hay et al., 1985; Pinto et al., 1992; Air et al., 1999).

There is great interest in maximizing the availability of blood and blood

components during a pandemic. The FDA and the American Association of Blood Banks

13

(AABB) Pandemic Influenza Task Force have been working to anticipate the impact of a
major F LUAV pandemic on the sustained availability of blood supplies in the US.
(Williams, 2007). A F LUAV biosensor capable of identifying the human infectivity and
transmissibility of a novel FLUAV strain would be useful as a screening tool. Quick
identification of a human transmissible F LUAV strain could initiate preventative
measures such as the mobilizing of blood supplies before irreversible spread of the
disease. In the event that a highly pandemic FLUAV strain is identified, the AABB
request for relaxation of certain current regulatory standards related to donor eligibility

and testing may be considered.

14

2.2 INFLUENZA IMMUNE GLOBULIN
2.2.1 Applications

One of the unique abilities of influenza viruses, and the reason there is such great
concern about future pandemics, is their ability for quick mutation, as described in 2.1.3
and 2.1.4. By the same mechanisms of antigenic drift and shift that allow F LUAV strains
to change receptor specificity and host range, a F LUAV strain could also achieve drug-
resistance. Because vaccines may take several weeks to become effective, therapies are
needed that could be used pre- and post-exposure, and as treatment, for pandemic
influenza in settings where vaccines would not have time to take effect. To this end,
passive antibody-based therapies have shown promise with their versatility, specificity,
and low toxicity, with applications for both prophylaxis and treatment of F LUIGIV
infections. In fact, the versatility of antibodies makes antibody-based therapies potentially
useful against any existing pathogen (Casadevall, 1996). Immune sera therapy was first
reported by Behring and Kitasato (1890) for treatment of diphtheria and tetanus, followed
by treatment of bacterial infections such as those caused by Streptococcus pneumoniae
and Haemophilus influenzae, as well as viral infections such as those caused by measles,
Poliomyelitis, and Varicella zoster (Casadevall and Scharff, 1995). Studies indicated that
serum therapy was effective in reducing mortality associated with meningococcal
meningitis, Haemophilus influenzae meningitis, and diphtheria, the last of which is still
treated by antibody therapy (Alexander, 1943a; Alexander, 1943b; Casadevall and
Scharff, 1994; F lexner, 1913; F othergill, 1937; McCloskey, 1985). Serum therapy was
replaced with antimicrobial chemotherapy a century later when toxicity problems

associated with heterologous sera were discovered. Side effects included fever and chills

15

as well as the rash, proteinuria, and arthralgia typical of ‘serum sickness,’ and were likely
due to immune complex formation (F einberg, 1936; Rackemann, 1942). However, in
recent years antimicrobial chemotherapy has also faced lowered applicability as
irnmunocompromised individuals, old and new pathogen emergence, and antimicrobial
drug resistance all showed increases (Casadevall and Scharff, 1995). Antibody-based
therapies are thus regaining applicability. Human ‘gamma globulin’ has replaced
heterologous sera, and typically is formulated as human immunoglobulin for intravenous
administration (IVIG) (Barandun et al., 1962; Prince et al., 1986; Zolla-Pazner and
Gomy, 1992). Recent technological advances in synthesizing of human antibody reagents
have eliminated serum toxicity problems and have propelled antibody-based therapies
forward, and they may fill the gap in infectious disease treatment (Casadevall, 1996;
Casadevall and Scharff, 1995; Kohler and Milstein, 1975; Wright et al., 1992).
2.2.2 FLUIGIV as Prophylaxis

In terms of public health, prevention is especially important, and indeed the use of
antibodies is in general more effective for prophylaxis than for therapy (Casadevall, 1996;
Cross, 1995). High-titer human anti-HIV (human immunodeficiency virus) Ig (I-IIVIG)
was protective against viral challenge in chimpanzees, with amount of antibody
administered and amount of challenge virus playing significant roles in ability to protect
(Prince et al., 1991). Subject animals remained free of infection and no primary immune
response was detected, leading the authors to conclude that HIV vaccines should thus
induce neutralizing antibody and that cell-mediated immunity induction may not be

necessary for HIV protection (Prince et al., 1991).

16

Intranasal antibody prophylaxis has shown great promise against viral respiratory
tract infections in animals and is beginning to show efficacy in human clinical studies.
For example, a single intranasal application of gamma globulin to mice was
prophylactically effective for about 72 hr against influenza A virus (Weltzin and Monath,
1 999).

In the case of antibody in mucosal secretions, protection is achieved by immune
exclusion, in which antibody activity is combined with the physical barrier of the mucus
blanket of the respiratory tract, and by direct neutralization of viral infectivity, in which
binding of antibody to virus particles prevents them from interacting with cell receptors
so they cannot infect target cells (Weltzin and Monath, 1999).

2.2.3 FLUIGIV gum

In terms of treatment, antibody therapy may be useful for those infections by drug-
resistant pathogens or by pathogens for which no antimicrobial drugs are available, or in
the event of antiviral rationing during a severe pandemic. Antibody-based therapies as
treatment are most effective when administered early in the course of disease. For
example, serum administration in the treatment of pneumococcal pneumonia was most
effective if within 3 days of symptom onset (Casadevall and Scharff, 1994).

To address a gap in H5N1 treatment, Luke et al. (2006) concluded that convalescent
human H5N 1 plasma could be usefiil as treatment for H5Nl infection, as evidenced by
studies from the Spanish Influenza pandemic of H1N1 in 1918-1919 which reported that
patients with influenza complicated by pneumonia experienced a reduction in mortality
and symptom improvement when treated with transfusions of influenza-convalescent

human blood products (Luke et al., 2006).

17

A patient who presented flu-like symptoms and whose tracheal aspirates tested
positive for H5N1 in southern China in 2006 was treated with the convalescent plasma
from another patient who had recovered from H5N1 infection 4 months prior (Zhou et al.,
2007). Poultry in the region had suffered from infection by a predominantly clade 2.3
HSN 1 variant. After the first plasma transfusion, the patient’s viral load was diminished
until undetectable after 32 h, and was released 2 months later after complete recovery.
Viral load reduction occurred in conjunction with neutralizing-antibody titer increase,
which Zhou et al. (2007) concluded may have been the result of the convalescent plasma
treatment as well as the normal humoral immune response. Virus isolated from both the
patient and the plasma donor were F ujian-like H5Nl variants which presented close
genetic relation with greater than 99% homology in HA genes (Zhou et al., 2007).
Passive irnmunotherapy is thus an option of interest with regard to FLUAV infection
treatment.

2.2.4 Current Applications of Antibody Therapy

Currently, antibody therapy is used for reducing infections in immunocompromised
patients; for postexposure prophylaxis against measles and hepatitis; for treatment of
botulism, diphtheria, and snake bites; and for prophylaxis and treatment of viral
infections caused by cytomegalovirus (CMV), parvovirus, rotavirus, enterovirus, and
varicella (Barnes et al., 1982; Bodensteiner et al., 1979; Brunell et al.,1972; Bussel and
Cunningham-Rundles, 1985; Conti et al., 1994; Frickhofen et al., 1990; McCloskey,
1985; Pemrington, 1990; Reed et al., 1988; Tacket et al., 1984; Watt, 1978; Wilfert et al.,

1977). High-risk patients, such as those with AIDS or organ transplant recipients, also

18

may be effectively treated with polyclonal antibody to reduce incidence of infection
(Mofenson et al., 1994; Stratta et al., 1992; Yap, 1994).
2.2.5 Potential Applications of Antibody Therapy
2.2.5.1 Antivirals

Antibodies are inherently able to neutralize virus, typically by binding to the virus
within its host receptor binding domain, or by otherwise blocking attachment to this
region. Antibodies as prophylaxis against viruses such as CMV, HIV, respiratory
syncytial virus, and parvovirus are under current study (Aulitzky et al., 1991; Barbas et
al., 1992; Kim, 1987; Zolla-Pazner and Gorny, 1992).
2. 2. 5.2 Resistant or Highly Virulent Pathogens

There are also an increasing number of pathogens which have achieved drug-
resistance, such as Pseudomonas aeruginosa, S. pneumoniae, and E. Faecium, for which
antibiotics have become obsolete (Austrian, 1994; Cameron et al., 1993; Casadevall and
Scharff, 1995; Pier et al., 1989; Schlaes et al., 1993). Various highly virulent pathogens
such as methicillin-resistant Staphylococcus aureus are also lacking in effective
antimicrobial agents. C. parvum and vancomycin-resistant enterococcus have no
available antimicrobial therapy (Casadevall, 1996). Antibody therapy has shown promise
for treatment of pneumococcus and staphylococcus (Casadevall and Scharff, 1994;
Correa et al., 1994; Ramisse et al., 1993). In such cases, conjunctive effects of antibody
therapy and chemotherapy could slow resistance development (Casadevall and Scharff,

1995)

19

2. 2. 5.3 Immunocompromised Individuals

Antibody therapies may be useful to target pathogens such as invasive fungi that
affect primarily immunocompromised patients for whom antimicrobial therapy is
ineffective (Casadevall, 1996). In such individuals, low-virulence organisms could cause
infection that is either difficult or impossible to treat, and antibody therapy is optimal for
the enhancement of immune function (Casadevall and Scharff, 1995).
2. 2. 5.4 Toxin Neutralization

Antibody therapies could be useful for neutralization of toxins, such as those
introduced by snake bites, spider bites, diphtheria, or tetanus. Toxic shock syndrome is
also a potential target, and IVIG administration has been shown to improve toxic shock
syndrome due to Streptococcus pyogenes (Barry et al., 1992; See and Chow, 1989;
Talkington et al., 1993).
2.2. 5.5 Antibody Therapy as Combined with Chemotherapy

Antibody therapy could stand alone for prophylactic purposes, but for treatment of
infection, a combination with chemotherapy is attractive. Antibodies promote microbial
killing, implying that a combination of both treatments would cause an amplified joint
response. Also, a combined response would require less of the toxic antibiotic as well as

less of the costly antibody therapy (Casadevall and Scharff, 1995).

20

Table 1. Applications for which antibody therapy combined with chemotherapy has

shown preliminary effectiveness.

Pathogen Chemotherapy

Antibody

 

Cytomegalovirus Ganciclovir
(Wilson et al., 1987; Reed et al., 1988)

Murine immune sera

 

Haemophilus influenzae Sulfonamide
(Alexander, 1943a,b)

Rabbit, horse immune sera

 

Herpes simplex virus Acycloguanosine
(Cho et al., 1976)

Human immune globulin

 

Lassa virus Ribavirin
(Jahrling et al., 1984)

Monkey immune sera

 

Neisseria meningitidis Sulfanilarnide
(Branham, 1935; Branham, 1937)

Horse immune sera

 

Staphylococcus aureus Penicillin
(Sonea et al., 1958)

Human gamma globulin

 

Staphylococcus aureus Chlorarnphenicol
(Fisher, 1957)

Human gamma globulin

 

Streptococcus pneumoniae Sulfapyridine
(Powell and J amieson, 1939)

Rabbit immune sera

 

21

2.2.6 Shortcomings of Immune Sera

Immune sera in its current state of development presents deficiencies. Monoclonal
antibody preparations, which are homogeneous immunoglobulins generated in vitro by
hybridoma or recombinant DNA technologies, recognize one epitope and could offer 100
to 200 times the activity of polyclonal immune globulin (Lang et al., 1993). In contrast,
immune sera contains antibodies of varying specificities and isotypes. Shortcomings have
included lot-to-lot variation and low specific antibody content (Felton, 1928; Weisman et
al., 1994). Each lot of IVIG could include plasma from more than 2000 donors and is
produced by multiple alcohol precipitations and centrifugations, as well as further
manufacturer-specific processing (Cohn et al., 1944; Kistler and Nitschmann, 1962). An
evaluation of 100 lots of IVIG from several products to determine opsonic activity for
Staphylococcus epidermidis, Haemophilus influenzae, Escherischia coli, and various
serotypes of Streptococcus, revealed lot variability as dependent on organism and
manufacturer. Variation in opsonic activity within one IVIG lot was significantly affected
by donor pool as opposed to manufacturing method (Weisman et al., 1994). Clinical
reports of IVIG inefficacy could be improved if pathogen-specific antibody content of

IVIG products is known (Weisman et al., 1993; Weisman et al., 1994).

22

2.3 VIROLOGICAL METHODS
2.3.1 Viral Isolation Culture

Conventional virological methods for virus analysis are well established (Amano
and Cheng, 2005). The “gold standard” for virus detection is viral isolation culture with
immunocytological confirmation of viral antigen, which follows internationally
recognized methods (Alexander et al., 2005; Charlton et al., 2009). Isolation and
propagation of influenza virus A, B, and C requires a biosafety level 3 (ESL-3)
diagnostic laboratory as well as certain reagents, which often reserves the process for
national reference or research laboratories. Virus is inoculated into the chorioallantoic sac
of embryonated eggs and after 24-48 h incubation, undergoes two to three blind passages,
which may take up to 21 days. The specimen produces a cytopathic effect (CPE), which
is then confirmed as F LUAV by hemagglutination inhibition (HI), immunofluorescent
antibody (IFA) staining, or reverse-transcriptase polymerase chain reaction (RT-PCR)
with the use of reference antisera or monoclonal antibody (Amano and Cheng, 2005;
WHO, 2007b; WHO SEARO, 2007). Pathotyping typically necessitates experimental
inoculation of 4-8-week-old chickens (Alexander et al., 2005; Charlton et al., 2009).
These methods each require bench times of approximately 2-4 hours (Boon et al., 2001).
Serological or molecular biological characterization methods may follow, which would
require approximately 48-72 hours each (Amano and Cheng, 2005). By definition, virus
isolation has a sensitivity and specificity of 100%, but does depend on virus viability.

Time to report including isolation and typing is 2-4 weeks.

23

2.3.2 Complement Fixation

The complement fixation (CF) test mainly detects antibodies to type-specific
nucleoproteins (Ziegler et al., 1997). The test is based on the use of complement, a
biological substance present in normal animal sera (Amano and Cheng, 2005).
Complement reacts with almost any antigen-antibody complex because it lacks
specificity. Sheep red blood cells (sRBC) are the indicators. A positive result occurs
when the complement is bound to an antigen-antibody complex and cannot react with
sRBC, leaving sRBC unlysed. A negative result occurs when there is no antigen-antibody
complex for sRBC to bind to, allowing complement to cause lysis of sRBC. The CF test
requires experienced personnel and laboratory time (Amano and Cheng, 2005). Prince
and Leber (2003) have shown that CF gives false-negative antibody response results
following influenza virus vaccination. CF is less sensitive than ELISA (Masihi and Lange,
1980). Full fixation requires 4 hours at 0 degrees C (Kahn, 1921).

2.3.3 Hemagglutination Inhibition

The hemagglutination inhibition (HI) method is more sensitive than CF for
detecting antibody responses to naturally occurring influenza A and B (Prince and Leber,
2003). H1 is a serological assay, detecting antibodies to strain-specific hemagglutinins.
Hemagglutinin can cause agglutination in the presence of erythrocytes. The HA
agglutination test traditionally identifies influenza, and the HI test is used to determine
hemagglutinin subtype. To perform the HI test, specimen is mixed with antisera to known
HA subtypes. Agglutination is inhibited when the antisera type matches the test sample.
In the presence of an HA binding molecule, the observed H1 is proportional to the

concentration of the inhibitor (Alvarez et al., 2010; Salk et al., 1944). HI assays require

24

laboratory expertise and skill, but are highly reliable, universally recognized, and
preferred for WHO global influenza surveillance (Amano and Cheng, 2005). However,
there are limitations to the process, as HI assays for influenza virus antibodies are not
widely available, are dependent upon the quality of erythrocytes used, and require the use
of replication competent virus (Hassantoufighi et al., 2009; Noah et al., 2009; Prince and
Leber, 2003; Stephenson et al., 2003). Thus BSL-3 facilities are required as is a very
large supply of virus (Allwinn et al., 2010; Hancock et al., 2009; Miller et al., 2010;
Schultsz et al., 2009). Additionally, H1 is limited by low sensitivity, subtype cross-
reactivity, and variability, and because it does not distinguish between infectious and
non-infectious virus particles, may be entirely inappropriate for H5Nl work (J ulkunen et
al., 1985; Massicot and Murphy, 1977; Tsai et al., 2009; WHO, 2007a). Time to report is
several hours.
2.3.4 Microneutralization

Microneutralization is another serological assay, which requires a small amount of
sermn to be tested in a microtitre plate for virus neutralization by FLUAV specific
antibody (WHO, 2007b). Microneutralization confirmed by western blot analysis is
beginning to gain preference over HI and offers consistent results (Hancock et al., 2009;
Kayali et al., 2008; Kitphati et al., 2009; Rowe et al., 1999; Schultsz et al., 2009; Sirskyj
et al., 2010; Tsai et al., 2009). This is a specific and sensitive assay for detecting strain-
specific antibody in serum that can typically detect lower titers than HI.
Microneutralization also presents limitations with complexity of standardization, low-
throughput, and extensive training requirements (Petric et al., 2006; Stelzer-Braid et al.,

2008; Tsai et al., 2009 WHO, 2007b). ESL-3 facilities are required for live virus

25

handling, and results are obtained within 3 days (Hancock et al., 2009; Kitphati et al.,
2009; Schultsz et al., 2009; Sirskyj et al., 2010; WHO, 2007b).
2.3.5 Immunofluorescent Antibody Staining

Immunofluorescent antibody (IF A) staining directly detects influenza antigens in
clinical samples by their interaction with F LUAV strain-specific monoclonal antibodies
that are directly or indirectly fluorescently tagged. A fluorescent microscope is required
for visualization. Compared to cell culture, IF A staining has asensitivity of 70-100% and
specificity of 80-100%, with time to report within 24 hours (WHO, 2007b).
2.3.6 En_zyme Linked Immunosorbent Assay

Enzyme linked irnmunosorbent assay (ELISA) methods do not require fresh
erythrocytes, virus manipulation, or subjective visual results interpretation (Alvarez et al.,
2010). ELISA techniques are under study for detection of anti-influenza antibodies in
animal or human serum samples (Blitvich et al., 2003; De Boer et al., 1990; Hall et al.,
1995; He et al., 2007; Prabakaran et al., 2009; Stelzer-Braid et al., 2008). However,
ELISA methods may be subject to false positive results as infection or seasonal F LUAV
vaccination could generate cross-reactivity of antibodies (Stelzer-Braid et al., 2008). This
was seen in a commercial ELISA for detection of anti-H5 HA antibodies, which was used
to screen vaccinated sera, with accurate identification of high levels of anti-H5 antibodies.
However, antibodies against seasonal H3N2 and HlNl cross-reacted with the H5 antigen
(Stelzer-Braid et al., 2008). Results from ELISA methods have also shown poor
predictive value with H1 or microneutralization assays (Ceyhan et al., 2010; Kayali et al.,

2008; Rowe et al., 1999).

26

2.3.7 RT-PCR

Reverse transcription polymerase chain reaction (RT-PCR) is becoming
increasingly effective for viral detection, typing, and subtyping (Zhang and Evans, 1991).
FLUAV genetic materials may be detected by RT-PCR in samples with very low levels
of viral particles, because of genetic multiplication by polymerase enzyme (WHO,
2007b). Viral nucleic acids are extracted fi'om clinical specimens and cDNA is then
synthesized by in-vitro reverse transcription of viral RNA. The cDNA is amplified with
specific primers and DNA polymerase. Detection of the amplified product can be
achieved by fluorescence and luminescence measurements (Amano and Cheng, 2005).
RT-PCR has higher sensitivity and shorter detection time than conventional methods.
Compared to the “gold standar ” of virus cultivation, Atrnar et al. (1996) reported that
the RT-PCR assay had a sensitivity, specificity, and efficiency of 95, 98, and 97%,
respectively, compared with 75, 100, and 93%, respectively, for the best commercially
available diagnostic kit (Becton Dickinson Directigen). Compared to the 2-10 days
required for culture, PCR only requires 24 hours (Magnard et al., 1999). RT-PCR is thus
an effective alternative to virus isolation for FLUAV detection (Atmar et al., 1996).
Drawbacks of this method include high cost of reagents and thermocyler equipment,
requirement of specific oligonucleotide primers from WHO influenza reference and
collaborating centers, BSL-2 requirement, high rate of false positive results, and
complicated procedure (Ellis and Zambon, 2002; WHO, 2007b). Real-time RT-PCR is
also under study for FLUAV detection, utilizing a one-tube protocol and fluorogenic
hydrolysis type probes, with sensitivity and specificity comparable to virus isolation and

HI (Lee and Suarez, 2004; Spackman et al., 2002).

27

2.4 COMMERCIAL DIAGNOSTIC TEST KITS

Commercial diagnostic test kits directly detect influenza A or B virus-associated
antigens or enzyme in throat swabs, nasal swabs, or nasal washes. These tests generally
have 70% sensitivity and 90% specificity for viral antigens (Montalto, 2003). Time to
report is approximately 30 minutes (Amano and Cheng, 2005).

Directigen (Beckton Dickinson Diagnostic Systems, Sparks, Maryland) is an enzyme
immunoassay (EIA) membrane test for influenza A and B (Reina et al., 2002). Enzyme-
conjugated monoclonal antibodies specific to influenza A or B are used. Visualization of
the captured influenza antigen-antibody couple is achieved by an enzymatic color
development reaction. This is the first commercially available rapid assay kit that
distinguishes between viral antigens from influenza A and B. The test has a sensitivity of
75-87% and specificity of 93-97% (Gavin and Thomson, 2003). Time to report is
approximately 25 minutes (Amano and Cheng, 2005).

QuickVue Influenza A/B (Quidel Corporation, San Diego, California) uses
immunochromatography to detect influenza A and B without differentiation (Gavin and
Thomson, 2003). Extraction is required to allow targeting of nucleoprotein from
influenza. Nucleoproteins in the specimen react with reagents to produce a color change
on the test strip. Manufacturer data show a specificity of 96-99% and sensitivity from 73-
82%. Time to report is approximately 10 minutes (Amano and Cheng, 2005).

ZStatFlu (ZymeTx, Inc., Oklahoma City, Oklahoma) is a neurarninidase assay that
achieves specificity using modified sialic acid (SA). In the presence of neuraminidase,
the bromoindole that is bonded to SA is released, forming insoluble dyes that indicate a

positive response (Shimasaki et al., 2001). Influenza A and B are not discriminated

28

(Gavin and Thomson, 2003). The test has a specificity of 98.7% but a poor sensitivity of
62.2% as reported by the manufacturer. Two minutes of testing and 20 minutes of
incubation are required, resulting in a time to report of approximately 22 minutes (Amano
and Cheng, 2005).

The BioStar OIA Flu A/B (Invemess Medical Innovations, Louisville, Colorado) is a
nucleoprotein antibody assay that detects the change in film thickness due to the binding
of antigen-antibody complex to a silicon wafer surface. Any antigen in the specimen is
captured by the immobilized antibodies, resulting in an increase in film thickness that is
detected as a color change due to a shift in the reflected light path (Amano and Cheng,
2005). Influenza A and B are not discriminated (Gavin and Thomson, 2003). Studies
have shown that the sensitivity of this test may vary from 51.4-71.8% (95% CI),
depending on the source of the specimens (Schultze et al., 2001). Specificity ranges from
69-79% (Gavin and Thomson, 2003). In particular, a negative result on the FLU OIA
should be confirmed by direct fluorescent antibody (DFA) and culture (Hindiyeh et al.,
2000). Time to report is approximately 16-20 minutes (Schultze’et al., 2001).

Binax NOW Flu A and Flu B (Invemess Medical Innovations, Louisville, Colorado)
detects influenza nucleoprotein antigen in specimens using an immunochromatographic
membrane test. Influenza antigen present in the specimen bind to gold-conjugated anti-
influenza antibodies in the test strip. The sample line is then formed when the antigen-
conjugate complexes are captured by the immobilized antibodies (Amano and Cheng,
2005). Sensitivity is reported as 82% and specificity is 94% (Gavin and Thomson, 2003).

Time to report is approximately 1-2 hours.

29

The conclusion to be drawn from these current rapid assay techniques is a lack of
binding partner novelty; most assays depend on detecting Influenza viruses by
interactions with Influenza—specific antibodies. While the sialic acid receptor has been
investigated in neuraminidase binding, there lacks a biosensor technology that exploits

Influenza hemagglutinin specificity for host sialic acid receptors.

30

2.5 BIOSENSORS

A biosensor is the integration of a biological component with an electronic,
electrochemical, optical, or acoustic transducer, with the intention of quantifying a
physiological or biochemical change in terms of an electrical response (Blum and Coulet,
1991; D’Souza, 2001; Ivnitski et al., 1999; Jin et al., 2008; Muhammad-Tahir et al., 2007;
Pal et al., 2007; Pal et al., 2008a; Pal et al., 2008b; Pal and Alocilja, 2009; Turner et al.,
1987; Zhang and Alocilja, 2008). Biosensors for quick and reliable FLUAV detection are
of interest to minimize sample handling and the need for highly skilled laboratory
technicians. An attractive development is single-step direct sensing, in which separation,
incubation, and signal-reporting agents are eliminated. Label-free techniques showing
potential for virus detection include surface plasmon resonance (SPR) biosensors and
acoustic biosensors, both of which have shown subnanogram detection limits.
Alternatively, colorimetric sensors employing functional polymers are promising for
direct viral analysis without the need for instruments (Amano and Cheng, 2005).
2.5.1 Surface Plasmon Resonance Sensors

SPR biosensors are optical sensors that exploit special electromagnetic wave
frequencies to probe interactions between an analyte in solution and a biomolecular
recognition element immobilized on the sensor surface. This direct technique utilizes
these biomolecular recognition elements to recognize and capture analyte in a liquid
sample producing a local increase in the refractive index at a thin metal film surface
(Figure 3). Optical means can then be used to accurately measure the refractive index
increase (Homola, 2003; Meeusen et al., 2005). This biomolecular interaction screening

technique does not require labeling of the ligand or the receptor, allowing virtually any

31

complex to be screened with minimal assay development. The very'high sensitivity of
SPR also lends itself to flexibility in application (Cooper, 2003). Schofield and Dimmock
(1996) first reported the use of SPR for influenza virus detection. The sensor chip was
coated with a polymer matrix coupled with monoclonal antibody for influenza virus. The
influenza virus was injected into the flow system and binding affinity with the surface
antibody was monitored. Dissociation and association rate constants were comparable to
those from an affinity ELISA (Schofield and Dirnmock, 1996). SPR has been used to
study the interaction between influenza hemagglutinin (HA) and its cell surface receptor
sialic acid (SA) using a sensitive microscale binding assay (Takemoto et al., 1996). BHA
is a soluble form of HA that results when the protease bromelain cleaves HA from the
virus near the viral membrane (Brand and Skehel, 1972). Under low-pH-induced
conditions, BHA trimers form soluble aggregates called rosettes which bind specifically
to the fetuin-derivitized sensor surface. The tight binding due to the multivalent
interaction between the BHA rosettes and the fetuin-derivatized sensor surface was
quantitated using measurements of association rate, dissociation rate, and dissociation

constant (Takemoto et al., 1996). Time to report is 10 minutes (Yang et al., 2006).

32

   
 
  

Optical

Polarized detection

light

Intensity —-|
/ WW
II

   

   

 

 

 

 

 

 

 

l
Sensor chip, ll
gold film Resonance
06 A 6 A 6 ' l srgnal
/o o N _ =3]
0 ° / ‘K 0 0 Sensorgram
Flow channel

 

 

Figure 3. SPR theory based on changes in refractive index due to immobilized target,
with sensorgram output (adapted and modified from GE Healthcare, 2010).

2.5.2 Quartz ngstal Microbalance

Acoustic biosensors are based on quartz crystal resonators (Cooper, 2003).
Sauerbrey first demonstrated the sensitivity of the quartz crystal microbalance (QCM)
towards mass changes at the surface of QCM electrodes (Bruckenstein and Shay, 1985;
Henderson, 1991; Plausinaitis et al., 2001; Sauerbrey, 1959). The Sauerbrey equation
presents a linear correlation between the mass change and resonant frequency shift, and is
dependent on the linear sensitivity factor Cf which is a fundamental property of the QCM
crystal:

Afs = - Cf- Am (1)
where, Am is the change in mass per unit area, in g/cm2, Afs is the observed frequency
change, in Hz, and Cf is the sensitivity factor of the crystal.

The electrical behavior of a crystal resonator near series resonance is represented by the

Butterworth van Dyke (BVD) electrical model of a quartz crystal resonator (Figure 4).

33

QCM immunosensors fimction on the principle that adsorption of substances on the
surface of a quartz crystal changes its resonance oscillation frequency (Eun et al., 2002).
QCM is an attractive low-cost technique for monitoring interaction of biomolecules on
functionalized surfaces. As in SPR biosensors, interaction affinity and kinetics analyses
can be performed in real-time and without labeling, by monitoring changes in the crystal
resonant frequency. However, QCM biosensors offer more detailed information than SPR
systems, since the acoustic response also accounts for visco-elastic property and receptor-
ligand complex charge changes (Cooper, 2003).
2. 5. 2.1 Applications

In influenza research, QCM has been applied to the study of virus/receptor
interactions (Amano and Cheng, 2005). Sato et al. (1996) utilized QCM to study binding
of FLUAV to monosialoganglioside in membranes. The receptor functions of
gangliosides GM3 were found to be influenced by surrounding matrix lipids (Sato et al.,
1996). Cooper et al. (2001) developed a sensitive, economical direct method for virus
detection in which type 1 herpes simplex virus interacted with specific antibodies
covalently attached to the oscillating surface of a QCM. As the amplitude of oscillation
of the QCM was increased, the virions were detached and the resulting acoustic noise
was detected. Sensitivity approaches detection of a single virus particle (Cooper et al.,
2001). A QCM immunosensor for the detection of cymbidium mosaic potexvirus
(CymMV) and odontoglossum ringspot tobamovirus (ORSV) utilized QCMs pre-coated
with virus-specific antibodies. Binding of virions to the immobilized antibodies resulted
in a reduction of resonance oscillation frequency dependent on the amount of virus bound

and the resulting increase in mass at the QCM surface. The QCM assay was faster than

34

ELISA with comparable sensitivity (Eun et al., 2002). Irnmunochips of QCM crystal
coated with two monoclonal antibodies against dengue virus envelope protein and non-
structural protein were found to have a 100-fold greater sensitivity than conventional
sandwich ELISA and a shorter approximate detection time of 1 hour. Blood specimens
could be used to detect virus in the viremia phase (Su et al., 2003). Ultrasensitive QCM
for detection of M13-phages in the liquid phase showed an increase in the signal to noise
ratio by a factor of more than 6 when the resonant frequency of the quartz crystal was
increased from the typical range of 5-20 MHz to the ultrasensitive range of 39-110 MHz.
The detection limit was improved by a factor of 200. The ultrasensitive QCM sensors
were chemically milled (Uttenthaler et al., 2001). Time to report is approximately 10
minutes (Leca-Bouvier and Blum, 2005).
2. 5.2.2 Rupture Event Scanning

Rupture event scanning exploits the piezoelectric property of the quartz, by which
the quartz deforms under application of an electric field (Ward and Buttry, 1990). As the
magnitude of the electric field is increased, the oscillation amplitude increases and
acceleration of particles on the surface increases, causing the surface to exert an
increasing force on the particles. Eventually the bonds between particle and surface are
ruptured, and the quartz crystal can detect and convert the acoustic emission of the
rupture into an electrical signal, thus providing information on particle presence, numbers,
and affinity. Sample preparation is minimal, and time to report is a few minutes (Cooper

et al., 2001; Cooper, 2003; Dultsev et al., 2000; Dultsev et al., 2001).

35

 

 

 

Neil—W—

Lm Cm Rm

Figure 4. Butterworth van Dyke (BVD) electrical model of a quartz crystal
resonator (adapted and modified from Eun et al., 2002).
2.5.3. Colorimetric Seps_o_r_s_

Colorimetric sensors using functional polymers enable direct analysis of target
analytes through a color change. “Smart” materials with desirable physical, optical or
electrical properties that respond to an environmental stimulus are synthesized for the
application (Amano and Cheng, 2005). A colorimetric influenza sensor has been
developed that uses a polydiacetylene bilayer assembled on glass slides. The
polydiacetylene layer is functionalized with an analog of SA, the natural receptor for HA
recognition. The SA ligand serves as a molecular recognition element and the conjugated
polymer backbone signals binding at the surface. Binding of viral HA to the SA residues
results in a visible transition of blue to red film color, with color change quantified by
visible absorption spectroscopy. Time to report is several minutes (Charych et al., 1993).
2.5.4 Gold Nanoparticles and Quantum Dots

Nano-size gold particles (AuNPs) as well as semiconductor colloidal quantum dots
(QDs) have attracted interest for sensing applications. A commercial company (Genomic

Profiling Systems, Bedford, MA) has developed a strip test for influenza diagnosis

36

(NIAID, 2006). Influenza virus in the sample binds to antibody-coated gold particles
forming complexes that are drawn up the strip via capillary action. These complexes meet
and bind the antibodies on the strip, turning the line red due to the reflection of light from
the gold when a sufficient number of these complexes are captured. However, the method,
while easy, is not sensitive, requiring a large amount of virus, in the millions, to induce a
color change. They are developing a portable MultiPath technology, which uses digital
imaging to detect individual fluorescent particles instead of the large number of gold
particles. By this method, fluorescent particles bound by virus can be counted
individually, and virus can be detected at levels thousands of times lower than by gold
particles. The company is still performing feasibility studies on this technology (N IAID,
2006).
2.5.4.1 Shortcomings

QD technology is relatively immature and limited by intermittent luminescence,
influence of different dyes bound to protein linkers, influence of the colloidal nature of
the sensing environment, and relatively high expense. An epitaxial quantum dot (eQD)
biosensor has been proposed, in which rows of eQDs emitting at specific wavelengths are
functionalized with biotinylated antibodies. Upon excitation, each eQD will emit
photoltuninescence radiation, which is expected to be modified in the presence of trapped
viruses. eQDs avoid the intermittent luminescence effect. A prototype for influenza A
virus detection utilizes a thiol-biotin-avidin-biotinylated antibody architecture (Dubowski,
2006). Cadmium telluride QDs have been used as a proton sensor to detect proton flux

driven by ATP synthesis in chromatophores. QD-labeled chromatophores were applied as

37

a virus detector to detect the H9 avian influenza virus based on antibody-antigen reaction
(Y un et al., 2007).
2.5.5 Microarrays

A smart CMOS chip for influenza detection has been developed by CombiMatrix
Corp. of Portland, OR. Any known flu strain can be identified in 4 hours, and the
microarray does not require operation by skilled technicians. The CMOS format
electronically identifies binding events. The chip can be optically scanned, but because it
is an active device it can also eliminate the need for fluorescent tags and optical scanners.
The system is not yet portable (Johnson, 2005).

A new ELISA test for influenza virus uses Zanamivir-biotin conjugates.
Biotinylated inhibitors were fixed to an avidin-coated plate and serial dilutions of
influenza virus were added. Unbound virus particles were washed off and anti-HA
serum-horseradish peroxidase (HRP) conjugate and chromogenic substance were added
to detect captured virus (McKimm-Breschkin et al., 2003). The sensitivity was 5
hemagglutinating units (HAU), where 1 HAU is defined as the highest dilution of stock
virus that completely agglutinates a standard erythrocyte suspension (WHO, 1953). Time

to report for ELISA is approximately 2-4 hours (King, 2006).

38

2.6 TOWARDS IMPROVED TECHNOLOGY

Traditional viral assays such as MDCK cell culture, complement fixation, and
hemagglutinin-inhibition are still used widely, but cannot meet the demands for fast and
direct detection at the point of care and for quick screening in case of a bioterrorism event.
Commercially available influenza diagnostic tests provide quick results, but their
sensitivities are lower than real-time RT-PCR and virus culture. In general, rapid test
sensitivity ranged from 10-70% as compared to RT-PCR for the detection of novel H1N1,
and thus a negative result on a rapid test can not rule out novel influenza A infection
(CDC, 2009; Hurt, 2009). A sensitive and specific biosensor technology is thus important
to enable rapid and specific disease diagnosis on-site (Amano and Cheng, 2005;
Muhammad-Tahir et al., 2005b; Pal et al., 2007; Radke and Alocilja, 2005). There is a
need for a biosensor technology with a shorter detection time and comparable or superior
sensitivity as compared to standard ELISA. The technology must offer speed comparable
or superior to commercially available diagnostic test kits with a time to report of less than
30 minutes. Point-of-care applicability is essential. On-site handling and use may be
facilitated by a single use disposable platform, and the need for only an inexpensive
handheld signal reader.
2.6.1 Reguirements of Biosensor Technology
New biosensor technology must improve upon commercially available test kits by
decreasing definitive turnaround time and offering quantitative results as opposed to
subjective color change assessments. A potential aim is the specific identification of HA
receptor preference as an indicator of the pandemic potential of a FLUAV strain. Since

most available test kits function on influenza antigen-antibody binding, an assessment of

39

binding between hemagglutinins of F LUAV to sialic acid receptors would be novel and
informative. As previously described, studies have shown that human FLUAV, including
pandemic and seasonal epidemic viruses, bind the specific a2.6 sialic acid receptor on
host cells, whereas avian FLUAV bind the (12.3 sialic acid receptor (Matrosovich M.N. et
al., 2004; Stevens J. et al., 2006a). This preferential binding could serve as a novel
platform upon which a biosensor could identify FLUAV strains that could achieve
human-to-human transmissibility and pandemic potential. Such a rapid biosensor
technology would have applications for monitoring animal influenza infections and for
screening emerging FLUAV strains for their potential for human infectivity for both

disease monitoring and biosecurity.

40

2.7 CONDUCTING POLYMERS

Polymers were initially thought to be insulators, until the discovery that poly
(sulphur nitride) [(SN)x] becomes superconducting at low temperatures (Suman et al.,
2005). Simple doping with oxidizing agents (p-doping) or reducing agents (n-doping) can
enhance the electrical conductivity of [(SN)x] by several orders. In fact, a field comprised
of many different conductive polymers has emerged. These polymers can be doped and
undoped in reversible reactions via electrochemical techniques including changes in pH
or redox potential (Malhotra et al., 2006; Paul et al., 1985). Yoshino et al. (1983) found
that tetrafluorocyano-quinodimethane doping of poly-p-phenylenesulfide (PPS) increased
the electrical conductivity of non-doped PPS by more than ten orders of magnitude, and
double doping served to increase maximum conductivity as compared to single doping.
McDiarmid and Heeger found that the electrical conductivitiy of polyacetylene could be
enhanced by several orders of magnitude by doping with oxidizing and reducing agents
(Gerard et al., 2002). Hydrogen bromide (HBr) solution was used to dope polyacetylene,
resulting in an increase of electrical conductivity by an order of six. HBr-doped
polyacetylene was found to be slightly more stable than IZ-doped polyacetylene under
conditions of heat and air exposure (Lee et al., 1989). Polymers such as polyaniline,
polypyrrole, polyindole, poly-para-phenylene (PPP), polythiophene, and polyfuran have
been studied to improve biomolecular stability of emerging sensor technologies, with
applications in biosensors, chemical sensors, solar cells, firel cells, diodes, field-effect-
transistors, and rechargeable batteries (Ayesh, 2009; Diaz et al., 1979; Gajendran et al.,
2008; Ivory et al., 1979; Kawai et al., 1991; Kim and Wamser, 2006; Kim et al., 2009;

Malhotra et al., 2006; Rabolt et al., 1980; Syritski et al., 2005; Yoshino et al., 1983).

41

Pyrrole was electrochemically polymerized on a platinum surface to produce a durable
polypyrrole film of 0.8 pm thickness which was used as an electrode in cyclic
voltammetry measurements. Polypyrrole as an organic electrode material offers improved

conductivity, good stability in electrochemical environments, and strong adhesion to the

metal surface (Diaz et al., 1979).

2.7.1 Electronic Structures of Conducting Polymers

In contrast to saturated polymers, conducting polymers have one unpaired electron
via chemical bonding. Each carbon atom in the polymer backbone has a pi electron.
There is a delocalization of electrons along the backbone due to the overlapping of
orbitals of successive carbon atoms, which are in sp2pz configuration in pi bonding.
These partially filled molecular orbitals cause a charge mobility which offers the polymer
electrical conductivity as well as high electron affinity and low ionization potential. The
pi bonds are susceptible to electrochemical and chemical oxidation or reduction (Ivory et
al., 1979; Malhotra et al., 2006). External electric field can easily polarize the electrons-
in these delocalized systems, leading to attractive nonlinear optical properties (Gonnan

and Grubbs, 1991).
Electrical conductivity (0') is due to the existence and ability for movement of
charge carriers, and is expressed as,
o = nep. (2)
where, n is the number of charge carriers per unit volume, e is the electronic charge, and

u is the carrier mobility. Doped conjugated polymers are good electrical conductors due

to the presence of charge carriers in the p-electron polymer system, as introduced by

42

doping, which could potentially occur at every monomer. Charge carrier mobility results
from p-electron delocalization along the polymer backbone (Heeger and Smith, 1991).

Electrical conduction properties of semiconductors can be controlled by addition of
foreign atoms into the semiconductor lattice. The dopant atoms may have an electron
excess or deficit, leading to corresponding n or p type semiconductors. Conducting
polymers may be similarly described. The doping levels of conducting polymers are
comparatively high, in contrast to those of semiconductors. Charge transfer occurs
between the dopant atom and the polymer chain, which is thus partially oxidized (p-
doping) or reduced (n-doping) (Lyons, 1994).
2.7.2 Nanoparticles as Biological Sensogy Labels

Drug delivery, chemical sensors, biosensors, optoelectronics, and electrochemical
devices are all potential applications of nanoparticles of conducting polymers (Berggren
et al., 2007; Burroughes et al., 1988; Diaz et al., 1979; Drummond et al., 2003; Huang et
al., 2002; Jager et al., 2000; MacDiarmid, 2001; Malinauskas et al., 2005; Mannakos et
al., 1998; Potyrailo et al., 2006; Smela 2003). Such nanoparticles have a high surface
area and quantum size effect, offering them extraordinary physicochemical properties.
When their composition, surface structure, and agglomeration are strictly controlled, the
nanoparticles can have optimal electrical, mechanical, and chemical properties.
Nanoparticles with uniform sizes have been achieved by the hard template method which
utilizes anodized aluminum oxide or colloidal particles with empty pores. Dispersion
polymerization has been used to synthesize a polymer shell around monodisperse Si02
particles. The 150-700nm polymeric particles were hollowed by HF etching to remove

the Si02 cores. The size of the hollow core and the shell thickness were able to be varied

43

by size and their monodisperse nature allowed the formation of well-ordered colloidal
crystals (Xu and Asher, 2004). However, hard methods are expensive, require the use of
strong acids or bases, and require various template sizes. Soft template materials such as
functionalized organic acids or polymeric stabilizers are also useful, but present many of
the same limitations of the hard process. Surfactant micelles are self-assembled organic
media that can block aggregation (Mann, 2000). A template-free, one-step chemical
synthesis procedure has been developed to fabricate unagglomerated polypyrrole
nanospheres with controlled size under mild conditions. The sphere sizes are controlled
by manipulation of the volume ratio of two liquids, such as water and octanol, which
form reversed micelle droplets (Kim et al., 2009). Gold nanoparticles (AuNPs) of 1.4mm
diameter linked to an oligonucleotide have been shown to be susceptible to radio
frequency magnetic fields, offering remote electronic control over reversible DNA
hybridization behavior. Monofunctionalization with L-lysine of AuNPs yields 2 nm
nanoparticles that can serve as the building blocks for peptide chains (Hamad-Schifferli
et al., 2002; Sung et al., 2004). Iron (III) oxide (F e304) nanoparticles of 12 nm diameter
have been encapsulated in large unilamellar vesicles of dipalmitoylphosphatidylcholine
(DPPC) via reverse-phase evaporation (Wijaya and Hamad-Schifferli, 2007). Conjugated
polymers have been applied as artificial muscles due to their electroactive nature. Use of
such polymers as actuators in biomedical devices is gaining attention (Smela, 2003).
Polyacetylene, a conjugated polymeric semiconductor, has been used as the basis of a
semiconductor device, which operates by the presence of a surface electric field. The
optical properties of the polymer are changed by the formation of charged solitons, and

optical absorption occurs below the band gap. These optoelectronic effects are useful,

44

especially in combination with the processibility of the polymer (Burroughes et al., 1988).
Gold nanoparticles (AuNPs) conjugated to DNA have been shown to enhance in vitro
protein translation by a combination of nonspecific adsorption of the ribosome to the
AuNP-DNA and specific binding to the mRN A, which is a different perspective to the
common belief that nonspecific adsorption is a barrier for utilizing NPs. In fact, it was
shown that nonspecific adsorption was essential for expression enhancement (Park and
Hamad-Schifferli, 2010).
2.7.3 Poymers in Biosensors

Electrochemical detection of selected DNA sequences or mutated genes can be
achieved by electrochemistry at polymer-modified electrodes, electrochemical
amplifications with nanoparticles, and electrochemistry of DNA-specific redox reporters
(Drummond et al., 2003). Conjugated polymer actuators are of interest for physiological
applications due to their ability to be operated in aqueous media. Polypyrrole is
particularly stable under these conditions. A polypyrrole-gold bilayer microactuator has
been microfabricated with the ability to move other rnicrocomponents (Jager et al., 2000).
Nanostructurized conducting polymers have electrochemical applications such as sensors,
batteries, supercapacitors, and energy converters (Malinauskas et al., 2005). Conducting
polymers are useful for immobilizing and stabilizing biomolecules onto a sensor surface
via physical adsorption, electrochemical entrapment and covalent attachment via
coupling chemistry of ethyl-dimethyl-aminopropylcarbodiimide (EDC) and N-hydroxy-
succinirrride (NHS). The polymer itself can bind protein molecules, and electrochemical
synthesis allows simultaneous direct polymer deposition onto the electrode while also

trapping biomolecule targets (Bartlett and Whitaker, 1988; Gambhir et al., 2001 ).

45

Polymers optimal for this application have functional groups to facilitate covalent
binding. For example, polypyrrole has been galvanostatically electropolymerized to
entrap anti-IgG onto a platinum surface (Andreescu and Sadik, 2004; Sadik et al., 2002).
Enzymes can be covalently linked to the surface of functionalized conducting polypyrrole
films alter carbodiimide activation (Schuhmann et al., 1990). A glucose biosensor was
produced by electropolymerization of pyrrole-modified biotin to enable avidin and biotin-
labeled glucose oxidase (Cosnier and Lepellec, 1999). Immobilization of polyclonal
antibodies onto a conducting polypyrrole membrane has exhibited improved activity
compared to immunosensors made by physical entrapment or adsorption (Bender and
Sadik, 1998). Electrochemical detection has been demonstrated with modified electrodes,
nanoparticle amplification, and DNA-mediated charge transport (Drummond et al., 2003).
2.7.4 Conducting Polymers as Transducers

Conducting polymers can have applications as biosensor transducers, serving to
convert a biochemical signal that results from the interaction of a biological component
into a measurable electronic signal. Physical transducers may be electrochemical, thermal,
piezoelectric, and spectroscopic (Svorc et al., 1997). Amperometric biosensors for
example measure the current produced from the oxidation or reduction of a reactant
under constant applied potential (Malhotra et al., 2006).
2.7.5 Polyaniline

Polyaniline is perhaps the most studied conducting polymer in a family that also
includes polypyrrole, polyacetylene, and polythiophene. As both electrical conductors
and organic compounds, these materials are attractive for their flexibility and robustness.

In particular, polyaniline boasts highly controllable chemical and electrical properties,

46

simple sythesis, low cost, good environmental stability in different solutions, and strong
biomolecular interactions (Ahuja et al., 2007 ; Feast et al., 1996; Ryder et al., 1997; Sarno
et al., 2005). Polyaniline is also unique in the appearance of a single broad polaron band
deep in the gap, with a narrow band near the conduction-band edge, while all other
known conducting polymers reveal two broad polaron bands (Stafstrom et al., 1987).
Polarons cause delocalized unpaired electrons and distortions of the polymer chain,
which are confined to certain phenyl groups and adjacent NH groups (Focke et al., 1987;
Glarum and Marshall, 1986).

2. 7.5.1 Synthesis

Synthesis of polyaniline from its monomer form, aniline, proceeds via chemical or
electrochemical synthesis. Electrochemical polymerization utilizes a standard three-
electrode configuration in an electrochemical bath, consisting of working, counter, and
reference electrodes. Working electrodes may be composed of gold, platinum, nickel, or
chromium. Polymerization may occur by potential scanning, constant potential
(potentiostatic), or constant current (galvanostatic) (Ahuja et al., 2007).

In chemical synthesis, aniline monomer in aqueous solution is polymerized by step-
growth in the presence of an oxidizing agent and a protonating acid. Aniline cation
radicals are produced and then participate in pernigraniline polyaniline chain growth or
recombine into benzidine and N-phenyl-p-phenylenediamine. Oxidative polymerization
converts the perrrigraniline into emeraldine (Stejskal and Gilbert, 2002).

Kim and Wamser (2006) demonstrated that the use of aniline as the active redox
material in a dye-sensitized solar cell using a porphyrin sensitizer leads to the formation

of polyaniline, which acts as the hole transport medirun. At low light intensities, the solar

47

cell offers 0.8% overall energy conversion efficiency (Kim and Wamser, 2006). As
diagrammed in Figure 5, aniline has been polymerized in the presence of equimolar
proportions of hydrochloric acid with oxidation by ammonium peroxydisulfate salt,

yielding polyaniline emeraldine hydrochloride (Stej skal and Gilbert, 2002).

 

 

 

NHzHCI
+ 5 n (NH4)23208 >
4" cr' -
Cl
"l' H
N I.
l" +. +.
\
_ N N
l l .
H H

N
+2nFKH+5nthSO4+5n(NHOfiNLI

Figure 5. Aniline in the presence of hydrochloric acid, oxidized with ammonium
peroxydisulfate to yield polyaniline emeraldine hydrochloride (adapted and
modified from Stejskal and Gilbert, 2002).

48

2. 7.5.2 Forms

Polyaniline may exist in various forms, with corresponding levels of electrical
conductivity (Figure 6). Each structure has a characteristic color, and the forms may
undergo interconversion based on the conditions. Leucoemeraldine is the fully reduced
form of polyaniline, emeraldine is 50% oxidized, and penrigraniline is fully oxidized
(Ray et al., 1989; Stejskal et al., 1996). Each of these oxidation states may exist in base
form or may become protonated to its salt form by acid treatment. Of the characteristic
polyaniline forms, green protonated emeraldine, as produced by oxidative polymerization
of aniline, is of particular interest and importance due to its high electrical conductivity
and stability. As compared to other polymers, the C6 benezenoid rings of emeraldine can
rotate, causing alterations to the electronic structure. Additionally, emeraldine is not
charge conjugation symmetric, and its carbon rings and nitrogen atoms form a
generalized “A-B” polymer (Pouget et al., 1991). Stronger oxidizing conditions generate
blue protonated pernigraniline, which is also expected to be conducting. Reduction by
alkali results in violet perrrigrarriline base or colorless leucoemeraldine, which are not

electrically conducting (Stejskal et al., 1996).

49

H H H
I l

 

 

 

\ +'\
+e +e
15 Li J <—>L J
+0
IN IN
I l |
H H H
Colorless Green protonated Blue protonated
leucoemeraldine emeraldine pernigraniline

 

 

 

 

 

 

 

 

+

'\*e *H / +H+ I +2H+
OE QN‘ E

T Blue emeraldine
H base

 

 

Violet pernigraniline
base

 

 

 

l

 

 

Figure 6. Forms of polyaniline (adapted and modified from Stejskal et al., 1996).

50

2. 7.5.3 Electrical Conduction Properties

The high electrical conductivity of polyaniline is determined by the polaronic state
of the polymer, which in turn is determined by redox state, proton content, and stearic
hindrance (Grzeszczuk and Szostak, 2003). The electronic properties of polyaniline may
be modified by variation of either the number of protons, number of electrons, or both.
Addition of a protic solvent such as hydrochloric acid or sulfuric acid yields a conducting
form of polyaniline, with an increase in conductivity of an up to ten order of magnitude
as compared to its undoped insulating form (Sarno et al., 2005 ; Ahuja et al., 2007). This
is due to protonation (“proton doping”) of formerly unprotonated sites. Grzeszczuk and
Szostak (2003) found that hysteresis of the switching process of electrochemically
produced thin films of polyaniline was highest when hydrochloric acid was used as the
counterion, as compared to trichloroacetic acid or perchloric acid. The anions entered a
polymer phase during the electropolymerization process that was performed in aqueous
acid solution (Grzeszczuk and Poks, 1995; Grzeszczuk and Szostak, 2003; Grzeszczuk
and Zabinska—Olszak, 1993; Poks and Grzeszczuk, 1997).

Green protonated emeraldine has a conductivity of the order of l S/cm, which
places emeraldine in the semiconductor range, between common polymers (0 < 10"-9
S/cm) and metals (0 > 10"4 S/cm) (Stejskal et al., 1996). The Pauli susceptibility is
linearly proportional to the percentage of protonation (Stafstrom et al., 1987). Protonation
of the emeraldine base generates a polysemiquinone radical cation, or polaron, and a
polaron conduction band is formed by coulombic repulsion (MacDiarmid et al., 1987).
Stafstrom et al. (1987) proposed that protonation causes phase segregation of

unprotonated and protonated domains, and suggest a two-step transition of polyaniline in

51

its polyemeraldine form from undoped to proton-doped (Figure 7). Instability of
bipolarons leads to formation of polarons and eventually a polaron lattice, whose single
polaron band structure was shown to be accountable for observed optical transitions
(Stafstrom et al., 1987). Polarons occur at the midgap via removal of an electron from a
neutral nonconducting polymer which has a full valence band and empty conduction band
(band gap). Bipolarons are generated by further oxidation and removal of a second

electron (Stafstrom et al., 1987).

52

(a)

LOO 6pm

protonation n

l
"' 1| (2n)H
r 'I
©N
internal
H H redox reaction
H ‘l' H
(c) N
"I'O polaron©
1": H separation
r f r
N N

\ t©©l '1“ 2.

Figure 7. Geometric structure of polyaniline in polyemeraldine state. (a) Before
protonation, (b) formation of bipolarons after 50% protonation, (c) formation of
polarons after 50% protonation, and (d) polaron lattice formed after polaron
separation (adapted and modified from Stafstrom et al., 1987).

53

Polyaniline in the emeraldine oxidation state has been converted from insulating ((3
~ 10"-10 S/cm) to conducting (0 ~ 5 S/cm) by doping with 1M aqueous HCl, yielding
emeraldine hydrochloride, the corresponding salt (MacDiarmid et al., 1987). This
protonic acid doping process means that the number of electrons associated with the
polymer does not change. The metallic emeraldine hydrochloride was shown by
MacDiarmid et al. (1987) to be a delocalized semiquinone radical cation with a polaron
conduction band, with the nitrogen atoms holding most of the charge (MacDiarmid et al.,
1987)

Protonated polyaniline, in the emeraldine salt (emeraldine hydrochloride) state, has
been converted to nonconducting blue emeraldine base by deprotonation in alkaline
medium (Figure 8). The polyaniline transitions into polyaniline emeraldine base (Stej skal

etaL,1996)

54

(a) CI' III

«I m"

| l J

 

H H n
CI' I deprotonation
(b) H

 

 

 

Figure 8. Deprotonation of polyaniline in presence of chloride (alkaline medium). (a)
Polyaniline emeraldine salt is converted to (b) polyaniline emeraldine base (adapted
and modified from Stejskal and Gilbert, 2002).

55

Hole-doped (p-doped) polyaniline is the more common form of conducting
polyaniline. Chaudhuri and Sarrna (2006) investigated electron-doping (n-doping), in
which synthesis required deprotonation of the amine N atoms (-NH-) in the polymer,
using a very strong base, n-butyl lithium (nBuLi). The resulting lithiated polyaniline was
unstable and reacted exotlrerrnally with moisture, which was likely due to electron-rich N
centers. To address instability, further complexation with electron-deficient boron
trihalides was necessary. While this negated the effect of the previous n-doping, and
generated a nonconducting end product, the reduced polyaniline did exhibit high
efficiency deep blue photoluminescence, with applications in thin, flexible display panels.
The n-doping step was able to deprotonate 75% of N atoms, leading to a strongly
nucleophilic form with potential for attachment of various functional groups (Chaudhuri
and Sarrna, 2006).

Additionally, polyaniline offers efficient electronic charge transfer, making it
attractive for use in biosensors as well as batteries, fuel cells, and electrodes (Liu et al.,
2005 ; Scott et al., 2005; Grennan et al., 2006). Magnetic polyaniline nanoclusters have
been described in the literature as lightweight yet mechanically strong, with various
combinations of magnetic cores and doping agents. Magnetic core materials include iron
(11, III) oxide, hydroxyl iron, and Li Ni Ferrite, with hydrochloric acid, phosphoric acid,
and toluene as doping agents (Poddar et al., 2004; Zhang et al., 2005 ; Dallas et al., 2006;
Jiang and Li, 2006; Xue et al., 2006).

2. 7.5.4 Electrochemistry
The high electrical conductivity of polyaniline is dependent on redox state, proton

content, and stearic hindrance. The characteristic redox switching process of polyaniline

56

is important for understanding its physical and chemical properties. Switching of
polyaniline from its fully reduced leucoemeraldine state to the conducting 50% oxidized
emeraldine state, and then from the emeraldine state to the fully oxidized pernigraniline
state, generates two peaks as observed by cyclic voltammetry. The observed emeraldine
state can occur over a potential range. The redox behavior of polyaniline is fundamentally
asymmetric, with oxidation transition occuning at a slower rate than reduction.
Additionally, the pH of the medium is important, with electrochemical activity lost in the
presence of neutral or alkaline medium (Gospodinova et al., 1996; Grzeszczuk and
Szostak, 2003; Hong and Park, 2005).

The redox transition of polyaniline typically occurs over the potential range from
-200 to 400 mV using a saturated calomel electrode. Proton ejection or injection
accompanies redox transitions. Oxidation of leucoemeraldine to emeraldine and from
emeraldine to pernigraniline was shown to be accompanied by proton ejection. The
proton injection that accompanies polyaniline reduction is incomplete for the transition
from emeraldine to leucoemeraldine. Proton equilibration is thus a slow process. Ybarra
et al. (2000) qualitatively demonstrated these proton ejection and injection processes by
using the amperometric mode of a rotating ring-disk electrode, which exhibited
significantly lower ring current during the reduction response. This indicates that the
polymer is not in protonic equilibrium with the electrolytic phase (Ybarra et al., 2000).

Grzeszczuk and Szostak (2003) utilized various counterions in the reversible
electrochemical doping of polyaniline, and found that the thermodynamic and kinetic CV
characteristics of the reversible state switching of the polymer were largely dependent on

anion nature. Factors included size, geometry, hydrogen-bonding, and basicity of the

57

anions. Formation of transition states between reduced and oxidized states was found to
require less energy when hydrogen-bonding interactions assisted the transition
(Grzeszczuk and Szostak, 2003).

2.7.6 Electrically Active Magnetic Polyaniline

Nanotechnology has progressed to the point where particles can be engineering
consistently at the nanoscale, for application in various biomedical and engineering fields.
Properties of nanoparticles differ significantly from their bulk counterpart. Novel
properties such as superparamagnetism and macroscopic quantum tunnelling emerge
when the size of the particles is reduced below the single domain limit. For iron and iron
oxide, this occurs at 15-20 nm (Poddar et al., 2004).

Magnetic nanoparticles have applications as contrast agents in magnetic resonance
imaging (MRI) and as agents of targeted drug delivery (Babes et al., 1999; Lacava et al.,
2001; Moghinri et al., 2001). A common hurdle encountered in such applications is
opsonization, in which particles injected into the bloodstream become coated by plasma
proteins or other biological circulatory components (Davis, 1997; Portet et al., 2001;
Ramge et al., 2000). Particles that are resistant to such coating will be cleared more
slowly, allowing improved drug performance. Such evasive particles have been
developed with coatings of dextran, polyethylene glycol (PEG), poloxamers and
polyoxamines (Lacava et al., 2001). Small hydrodynamic radius (<20nm) is also
important for the particles to reach the target cells (Gref et al., 1994; Moghimi et al.,
2001). Iron oxide nanoparticles have been synthesized and derivatized with dextran or
albumin, and the influence of their size and surface composition was assessed in vitro

using human dermal fibroblasts to characterize the interaction between cells and particles.

58

Derivatized particles were found to induce cell behavior alterations as compared to the
effects of underivatized particles, indicating that cell response can be specifically directed
by the engineering of particles on the nanoparticle surfaces (Berry et al., 2003).
2.7.7 _Cyclic voltammetgy

The electrochemical properties of a system can be explored using linear sweep
voltammetry techniques such as cyclic voltammetry. In an electrochemical cell with a
conventional three-electrode set-up, a potential is applied which is ramped linearly versus
time at a particular scan rate (V /s or mV/s) from an initial potential, E1, to a final
potential, E2, and back again to E1. In a reversible redox system, a redox couple will
undergo a one electron oxidation-reduction process, described by the equation

Ox + e' H R (3)

The output is presented as a cyclic voltarnmogram (CV), which illustrates the resulting
current (I) measured while scanning the potential range, represented with respect to the
potential (E) (Figure 9). The oxidation process occurs at the cathode and can be
represented by the cathodic peak potential, Epc, which corresponds to the point where the
current reaches the maximum, 1pc. The reduction process occurs at the anode and can be
represented by the anodic peak potential, Epa, as corresponds to the anodic maximum
current, Ipa. These reactions create a concentration gradient at the surface of the
electrode for both species, leading to a diffusion controlled mass transfer process of
species Ox from the electrolyte to the surface of the electrode. This transport is referred
to as ionic charge transfer or mass transfer (Vyas and Wang, 2010). Redox reactions can
be quantitatively assessed by the Ep or Ip values, the ratio of peak currents, [pa/1pc, the

separation of peak potentials, Epa — Epc, or the integral of current, AQ.

59

 

’pc -- /\

§ / \\

.C / ,,/::;"'_
‘ 8 _ / ,
w /

'3 \ /

2 __ /

’pa Epa

 

Figure 9. Schematic cyclic voltammogram for redox couple undergoing single
electron oxidation-reduction process.

The peak current, 1p, in a reversible redox system is characterized by the equation

F 3 l/ 2
I, = 0.4463 — n3/2A133’2C3v‘”
RT (4)
where F = Faraday’s constant (Q/mol), R = turiversal gas constant (J/mol-K), T =
temperature (K), n = number of electrons exchanged in the reaction, A = surface area of
electrode (cm2), D0 = diffusion coefficient of the electroactive species (cm2/s), C*=

concentration of the electroactive species (mol/cm3), and v = scan rate (V/s) (Bard and

Faulkner, 2000).

60

The peak potential, Ep, in a reversible redox system is characterized by the

equation

R T
E = E — 1 . 1 9 —
p 1 / 2 O ”F (5)
Mass transfer for ionic species in electrolytes near electrodes occurs between the
redox couple and the electrode, and can occur by various phenomena, including
diffusional transport under concentration gradients, migration transport of oppositely
charged ions under electrode electric field, and convection transport due to physical
electrolyte stirring. In the case of electrodes modified with electroactive or redox films,
such as conducting polymer films, the redox behavior becomes more complex. Here,
electron transfer from the electrode surface to the film occurs simultaneously with ionic
transfer from electrolyte to film. Electro-neutrality is thereby maintained (Lyons, 1994).
Cyclic voltammetry is an effective method of characterizing electrochemical
systems, and offers valuable information on redox reactions. The research described here
will examine the power of cyclic voltammetry in determining the concentration of an

electrically active species as an indicator of the presence of the target. Integral of current,

AQ, and peak currents, IP, will be used for quantitative analysis.

61

CHAPTER 3: RESEARCH HIGHLIGHTS
3.1 RESEARCH NOVELTY

The development of both an SPR-based assay as well as a nanoparticle-based
biosensor offer innovativeness in structure and application. SPR has been used
extensively in the literature as a sensitive and specific method for characterizing the
avidity, specificity, and kinetics of binding between various partners. Typical reactions
involve protein-protein binding, and while carbohydrate-protein interactions have been
described, to date no literature has been reported that investigates the specific interaction
between host carbohydrate (glycan) receptors and F LUAV hemagglutinin glycoprotein
by SPR (Table 2). Protein microarray technology has been reported as an appropriate
methodology for identifying this interaction; however, the indication of binding by
fluorescence is less quantitative than desirable.

The biosensor architecture is novel in its preparation, with multiple crosslinkers and
signal enhancers applied to achieve repeatable and sensitive binding interactions between
the carbohydrates and proteins. Electrically active polyaniline coated magnetic
nanoparticles (EAMs) are applied dually as magnetic concentrator of the carbohydrate—
protein—antibody complex, as well as the biosensor transducer. While this dual function
has been reported previously, the biosensor is novel in its design, with the molecules of
interest (carbohydrate and protein) applied to the working electrode in a single step,
without the need for preimmobilization of a specific antibody. The biosensor application
also presents innovation in the detection of pandemic-indicative FLUAV hemagglutinin

protein. Table 3 demonstrates the novelty of this research by outlining previous

62

contributions to the fields as well as highlighting gaps in the knowledge base that may be

addressed by the current research.

Table 2. SPR assays: Gaps in research

 

SPR assay: monoclonal antibodies against carbohydrate epitope (Ohlson et al., 2000).

 

SPR assay: H5N1 adjuvanted vaccine preparations against F LUAV (Khurana et al.,
2010)

 

Protein rrricroarrays: spot printing on glass slides for protein-protein interactions
(MacBeath and Schreiber, 2000).

 

Carbohydrate microarrays: spot printing on glass slides for carbohydrate-protein
interactions as observed by fluorescence intensities (Blixt et al., 2004).

 

Needed: SPR assay for quantitative carbohydrate-protein binding characterization

 

Needed: SPR assay for pandemic F LUAV H5N1 identification

 

Needed: SPR assay for measuring antibody-mediated inhibition of carbohydrate-protein
binding

 

63

Table 3. Biosensor technology: Gaps in research

 

Polyaniline synthesis (Li et al., 2007a): 10-50nm diameter polymerized in vanadic acid.

 

Electrically active polyaniline coated magnetic nanoparticle (EAM) synthesis (Li et al.,
2007b): Diameter of 0.5-5 um.

 

Polyaniline based antibody immunochromatographic biosensor: bovine viral diarrhea
virus (Muhammad Tahir et al., 2005a). Polyaniline polymerized in phenylphosphonic
acid (PPA), 4-hydroxybenzenesulphonic acid (HBSA), sulfobenzoic acid (SBA),
hydrochloride acid (HCl), perchloric acid (PA).

 

Polyaniline based antibody immunochromatographic biosensor: human serum albumin
detection; colloidal gold—antibody conjugates (Kim et al., 2000).

 

Polyaniline based antibody immunochromatographic biosensor: polyaniline magnetic
nanoparticles conjugated to antibody; screen printed silver electrodes (Yuk et al., 2009).

 

Polyaniline based antibody immunochromatographic biosensor: Bacillus anthracis (Pal
and Alocilja, 2009).

 

Polyaniline based enzyme amperometric biosensor: glucose oxidase immobilized on a
Prussian Blue—modified platinum electrode (Garj onyte and Malinauskas, 2000).

 

Immunochromatographic biosensor with signal enhancement by colloidal gold
conjugated to progesterone-ovalbumin (Jennes et al., 1986; Laitinen and Vuento, 1996).

 

Immunochromatographic biosensor with signal enhancement by colloidal gold
conjugated to polyclonal antibody: Salmonella typhimurium (Paek et al., 1999).

 

Electrically active polyaniline coated magnetic nanoparticle as immunomagnetic
concentrator of Bacillus anthracis endospores (Pal and Alocilja, 2009).

 

DNA biosensor with signal transduction and amplification by glucose oxidase catalyzed
deposition of cupric hexacyanoferrate (CuHCF) NPs: FLUAV (Chen et al., 2010).

 

Needed: EAM based carbohydrate biosensor

 

Needed: EAM based direct-charge transfer biosensor

 

Needed: EAM and gold nanoparticles as signal enhancers: carbohydrate/protein binding

 

Needed: EAM based biosensor for Influenza A virus detection

 

64

3.2 RESEARCH SIGNIFICANCE

The binding between host glycan receptors and the glycoprotein hemagglutinin on
the F LUAV surface is essential for FLUAV infectivity and transmission, and this
interaction is thus a prime target for study. Understanding the avidity and specificity of
these interactions, as dependent on FLUAV strain and glycan structure, is essential to
understand the mechanism of infection as well as to neutralize this binding by antibody-
based therapies. SPR offers a valuable technique for binding characterization and binding
neutralization studies with repeatable and quantitative results. The SPR assay requires
very low concentrations and volumes of ligand and analyte, and presents the
hemagglutinin analyte in a physiologically relevant aqueous system. Antibody-based
therapies for prophylaxis and treatment of F LUAV infection are gaining interest to
replace or augment chemotherapy techniques, and their current shortcomings, including
lot-to-lot variation, variation due to donor pool, and low specific antibody content, could
be circumvented by an SPR screening assay of donor plasma. Hyperimmune donor
plasma could be rapidly and accurately screened by the proposed SPR assay and ranked
based on glycan/protein neutralization activity to obtain a high potency FLUIGIV
product.

The use of EAMs as the biosensor target concentrator and signal transducer is the
result of the combination of the desirable chemical, electrical, and mechanical properties
of the conducting polymer, polyaniline, and the magnetic properties of the core material,
iron oxide. EAMs are valuable for their nanoscale dimensions, which provide an
increased surface to volume ratio upon which biological events can occur. The magnetic

property of the EAMs in conjunction with their propensity for biological surface

65

modification, allows the target to be quickly and easily identified and separated from
irrelevant background material, reducing matrix interference. This separation technique
will be advantageous for identifying low levels of target in complex samples, in particular
the Serum or respiratory secretions from which hemagglutinin or whole FLUAV virions
are typically isolated. Ideally, the magnetic power of the EAMs will eliminate the need
for time- and reagent-consurrring pre-enrichment steps. The electrochemical and magnetic
properties of the EAMs lend flexibility to the biosensor design, with the ability for any
strain of F LUAV to be specifically identified via a compatible antibody and
corresponding glycan receptor.

The strength of this research lies not only in the value of the. SPR assay and
biosensor individually, but also in the conjunctive applicability of both methodologies.
Parallel testing of carbohydrate-protein interactions on both systems offers a basis for
comparison from which improvements to both assays can be identified. The current
research developed a sensitive and specific SPR assay for characterizing glycan binding
to hemagglutinin from pandenric F LUAV and the neutralization of such binding. A
complementary carbohydrate based biosensor platform was developed to identify H5N1
hemagglutinin based on binding to a corresponding glycan receptor. A similar biosensor
platform was also investigated to differentiate between human-transmissible and non-

human-transmissible FLUAV strains.

66

lHl

 

3.3 HYPOTHESIS
This research is based on the following hypothesis:

Irnmunofunctionalized electrically active polyaniline coated magnetic nanoparticles
(EAMs) will concentrate target hemagglutinin from serum matrix by their magnetic
properties, and will function as the transducer in reporting a FLUAV-specific
biodetection event by their electrical properties, with results correlative to Surface

Plasmon Resonance (SPR) measurements.

67

3.4 RESEARCH OBJECTIVES

This research is based on the following specific objectives:

Objective 1: Design of a Surface Plasmon Resonance (SPR) based binding assay.
Objective 2: Design of an SPR based assay for detection of antibody-mediated binding
inhibition.

Objective 3: Design and fabrication of an EAM based electrochemical biosensor for
detection of H5N1 hemagglutinin.

Objective 4: Design and fabrication of an EAM based electrochemical biosensor for
identification of htunan-transmissible FLUAV strains.

Objective 5: Evaluation of sensitivity and specificity of the EAM based electrochemical

biosensors.

68

CHAPTER 4: RESEARCH MATERIALS AND METHODS
4.1 OBJECTIVE 1
Surface Plasmon Resonance—based binding assay
This objective was aimed at characterizing the ability of Surface Plasmon

Resonance technology to detect carbohydrate/protein binding with repeatability,
sensitivity, and specificity, for application in a glycan/hemagglutinin binding assay.
4.1.1 SPR Assay Desigp
4.1.1.1 Reagents and Chemicals

The biotinylated carbohydrate compounds 3’SLex (B157), 3’SLN (B84), 6’SLN
(B87), CT/Sda (B204), and GD2 (B184) were provided by the Carbohydrate
Synthesis/Protein Expression Core of The Consortium for Functional Glycomics funded
by the National Institute of General Medical Sciences grant GM621 16.. The following
reagent was obtained through the NIH Biodefense and Emerging Infections Research
Resources Repository, NIAID, NIH: H5 Hemagglutinin (HA) Protein from Influenza
Virus, AN ietnam/ 1203/04 (H5N1), Recombinant from baculovirus, NR-10510. 6xHis
tagged H5 hemagglutinin (HA) protein from 293 cell culture, AN ietnam/ 1203/04
(HSNI); C-terrninal 6xHis tagged H1 hemagglutinin (HA) protein from 293 cell culture,
A/South Carolina/l/18 (H1N1); and 6xHis tagged H5 hemagglutinin (HA) protein from
293 cell culture, A/Indonesia/S/OS (H5N1), were purchased from Immune Technology
Corp. (New York, NY). Recombinant full-length H3 protein from baculovirus,
A/Wyoming/3/03 (H3N2) was purchased fiom Prospec Protein Specialists (Rehovot,

Israel).

69

HBS-P buffer; HBS-EP buffer; 10 mM glycine pH 2.0, 2.5, and 3.0; 50 mM NaOH;
10 mM sodium acetate pH 4.0 and Sensor Chip SA (Streptavidin) were purchased from
GE Healthcare (Piscataway, NJ). Avidin/Biotin blocking kit was purchased from Vector
Laboratories, Inc. (Burlingarne, CA). Sodium chloride 5 M, ethylenediaminetetraacetic
acid (EDTA), phosphoric acid, bromelain, 2-mercaptoethanol (2-ME), and Tween-2O
were purchased from Sigma-Aldrich Co. (St. Louis, MO).
4.1.1.2 Equipment

The Biacore 3000 instrument offers automated Surface Plasmon Resonance
detection (Jason-Moller et al., 2006). Interaction analysis between proteins,
carbohydrates, nucleic acids and small molecules is possible on this real-time, label-free,
and contact-free system. Binding information for strong or weak, fast or slow interactions
can be obtained in the form of yes/no binding, binding specificity, binding affinity or
kinetics. As previously described, the Biacore-SPR system measures changes in mass at a
biospecific surface that occur due to the interaction of interest. One interaction partner is
immobilized onto a gold film surface, while the other is passed over in solution. The
Biacore offers integrated microfluidics, with very small dead-volume, low dispersion,
and fast liquid exchange rates (Figure 10). Information is output as a sensorgram of
Resonance Units (RU) versus time (s) (Jason-Moller et al., 2006).
4.1.1.3 SPR (Biacore) Chip Preparation and Immobilization

Previous glycan microarray analysis has been performed to probe HA specificities
for glycan receptors (Blixt et al., 2004; Stevens et al., 2006a). From this work,
corresponding glycan/HA pairs were chosen. The surface chemistry of the Biacore

Sensor Chip SA consists of streptavidin covalently immobilized on a carboxymethylated

70

 

dextran matrix, and has a listed binding capacity of 2 1800 Resonance Units (RU) of a
biotinylated oligonucleotide (Figure 10) (GE, 2004). The sealed sensor chip equilibrated
to room temperature for 60 min and was docked in the Biacore 3000 instrument. The
instrument was primed 3 times with filtered and degassed HBS-P buffer. Sensor Chip SA
was conditioned with three consecutive l-min injections of NaCl 1 M in NaOH 50 mM
(GE, 2004). Biotinylated glycans were diluted to 1 uM in Biacore HBS-P buffer and 8 pl
were injected over a Biacore SA chip at 10 III/min. Glycans were immobilized via
streptavidin/biotin interaction to saturation at approximately 300 resonance units (RU).
Streptavidin has extraordinarily high affinity for biotin (K; 10'14 to 10"6M). The
streptavidin/biotin interaction is also resilient, stable against heat, denaturants, proteolytic
enzyme action, and extreme pH (Laitinen et al., 2007). Flow cells 2-4 were immobilized
separately by manual injection: 3’SLex on flow cell 2, CT/Sda on flow cell 3, and 3’SLN
on flow cell 4. Flow cell 1 was left blank. The immobilized chip was blocked with two 30
s pulses of avidin, one 1 min pulse of HBS-P buffer, and two 30 s pulses of biotin. A
second chip was immobilized with the same glycans as Chip 1 at the lowest
immobilization level possible, around 20 RU. A third chip was immobilized with GD2,

6’SLN, and 3’SLN followed by avidin/biotin blocking.

71

 

Figure 10. Biacore SA sensor chip and instrumentation.

Table 4. Glycan structure and binding predictions

 

Chip Common Saccharide Name and Spacer
1, 2 Name

 

Fcl blank --

Fc2 3 ’ SLex Neu5Aca2-3GalBl-4[Fucor1-3]GlcNAcB-SpNH-LC-LC-Biotin

Fc3 CT/Sda Neu5Aca2-3 [GalNAcB l -4]GalB l -4GlcNAcB-SpNH-LC-LC-Biotin
Fc4 3 ’SLN NeuSAca2-3GalB1-4GlcNAcB-SpNH-LC-LC-Biotin

 

Chip

 

Fcl blank --

Fc2 GD2 Neu5Acor2-8Neu5Aca2-3[GalNAcB1-4]GalBl-4GchSpNH-LC-LC-B
Fc3 6’ SLN Neu5Acor2-6Galj31-4GlcNAc[3-SpNH-LC-LC-Biotin

Fc4 3 ’ SLN Neu5Acc12-3GalB1-4GlcNAcB-SpNH-LC-LC-Biotin

 

72

4.1.1.4 Chin Regeneration

H5 at 280 nM was injected over the immobilized glycans at 5 III/min to establish
activity of the surface. Once binding was observed, regeneration was explored to
completely remove bound H5 while retaining immobilization level and biological activity
of the glycans. Regeneration buffers including glycine 10 mM at pH 3.0, 2.5, and 2.0,
EDTA 50 mM + NaCl 0.5 M, NaCl 1 M, acetate 4.0, phosphoric acid 50 mM, and NaOH
50 mM were injected over the chip at 100 III/min. A two-injection regeneration scheme
consisting of 60 s of glycine 10 mM pH 2.5 and 30 s 50 mM NaOH was compared to a
two-injection regeneration scheme consisting of 60 s of glycine 10 mM pH 2.5 and 18 s
‘ 50 mM NaOH.

4.1.2 SPR Binding between Gflcan Receptors and Hemagglutinin
4.1.2.1 Binding Assay and Sensitivity Testing

Binding between recombinant H5 HA protein and synthetic glycan receptors was
investigated by injecting H5 samples serially diluted at a 1:3 ratio over the immobilized
glycans 3’SLex, CT/Sda, and 3’SLN, for 5 min at 5 ul/min, with regeneration scheme
following 20 min dissociation time. The H5 molarities were 286 nM, 94.3 nM, 31.4 nM,
10.6 nM, and 3.53 nM. Regeneration included 60 s of 10 mM glycine pH 2.5 and 30 s of
50 mM NaOH at 100 III/min. A truncated dilution series was also performed using the
regeneration of 60 s of glycine 10 mM pH 2.5 and 18 s 50 mM NaOH. This binding study
offered results from which the lowest detection limit of the SPR assay for H5 was
obtained. Testing for each H5 molarity was performed in triplicate. Recombinant H3 HA
(A/Wyoming/3/2003) was injected over the same immobilized glycans under the same

conditions and concentrations, and was used as the negative control. The lowest dilution

73

of H5 that produced a signal distinguishable from the control was taken as the sensitivity
of detection. Binding of H5 HA (A/Indonesia/5/2005) was also tested at 140 nM.
4.1.2.2 Specificity Testing

The specificity of the assay was investigated using H1, anti-H1, H3, anti-H3, and
glycans nonspecific for H5, 6’SLN (a2,6 binder), GD2 (a2,8 binder), and CT/Sda (a2,3
binder). H1 HA (A/South Carolina/l/18, HlNl) at 1:3 serial dilution was injected over
the glycans GD2, 3’SLN, and 6’SLN, which were immobilized to saturation. H3 HA
(A/Wyoming/3/03) was injected over glycans 3’SLex, CT/Sda, and 3’SLN, which were
immobilized to saturation. Testing for each dilution was performed in triplicate.
4.1.3 Characterization Studies
4.1.3.1 HA Receptor Binding Domain Binding Assessment

The binding of the HA1 segments of H5 and H3, which contain the receptor
binding domains, to glycans 3’SLex, CT/Sda, and 3’SLN was investigated. HA1 H5 and
HA1 H3 were prepared at 1:3 serial dilutions, with concentrations of 286 nM, 94.3 nM,
and 31.4 nM, and compared to the same concentrations of full length H5. The binding
experiment was repeated with rtmning and sample buffers prepared as I-IBS-P with 0.1%
BSA and 0.5% glycerol. HA1 H5 and HA1 H3 were compared to the positive control, H5
at 140 nM, and 3 glycerol concentration curve from 0-1%.
4.1.3.2 HA Preparation

To obtain a more consistent set of monomers and trimers, H5 was pretreated with

heat, bromelain, 2-mercaptoethanol (2-ME), and Tween-20. H5 at 1.4 uM was heated at

37 °C for 4 h or overnight, with or without bromelain 100 ug/ml and 2-ME 0.1 M. H5 at

74

1.4 IIM was heated at 56 °C for 10 min with or without subsequent cooling at 4 °C. H5 at
1.4 uM was also treated with Tween-20 at 0.1-0.02%.
4.1.3.3 Serum Experimenm

The complexity of biological samples was considered, with the ultimate goal of a
F LUIGIV screening assay in mind. The binding of H5 and H3 to glycans 3’SLex, CT/Sda,
and 3’SLN as described in 4.1.2.1 and 4.1.2.2 was repeated with HA at 1:3 serial dilution
prepared in mouse serum (ICR SCID) at 2% final concentration by volume. Background
binding to glycans 3’SLex, CT/Sda, and 3’SLN and the blank cell was investigated using
mouse serum (ICR SCID) at 0.5-10% in buffer.
4.1.3.4 Statistical Analysis

Each experiment was performed in triplicate to account for equipment or user
variation. The samples were double referenced, by subtracting either the blank fcl or
nonbinders CT/Sda, GD2, or 6’SLN from the binder results, as well as subtracting a
buffer run to compensate for irrelevant machine fluctuations. The SA chips were assumed
to have the same physical properties, and the glycans were assumed to be immobilized to
saturation. The peak RU at the end of the injection cycle was taken as an indicator of
binding. The effects of different HAS, glycans, anti-HA antibodies, and HA concentration
were assessed to calculate the lower detection limit and specificity of the SPR-based
assay. The differences between the means for each sample peak were calculated and
analyzed based on single factor analysis of variance (ANOVA) to a significance of 95%

(or = 0.05) (Tables A-l, A-2), using SAS software (SAS, Cary, NC).

75

4.2 OBJECTIVE 2
SPR to Detect Ab—Mediated Binding Inhibition

Although monoclonal antibodies do not mimic the complexity of immune sera or
donor plasma, the ability of HSNl-specific monoclonal antibodies to neutralize
previously observed binding between recombinant hemagglutinin and synthetic mimics
of host glycan receptors was investigated as proof-of-concept. Also described are
investigations into preparation techniques for the involved reagents for optimal binding
results.
4.2.1 SPR Inhibition Assay Desigp
4. 2. 1.1 Reagents

The following reagent was obtained through the NIH Biodefense and Emerging
Infections Research Resources Repository, NIAID, NIH: Monoclonal Anti-Influenza
Virus H5 Hemagglutinin (HA) Protein (VN04—2), AN ietnam/ 1203/04 (H5N1), (ascites,
Mouse), NR-2728. Polyclonal anti-influenza virus H1 hemagglutinin (HA) protein,
HlNl/Pan, (rabbit); polyclonal anti-influenza virus H5 hemagglutinin (HA) protein,
A/Indonesia/5/05 (H5N1), (rabbit); and polyclonal anti-influenza virus HA2 H5
hemagglutinin (HA) protein, A/Vietnam/ 1203/04 (H5N1), (rabbit), were purchased from
Immune Technology Corp. (New York, NY). Anti—influenza virus H3, A/Shandong/9/93
(H3N2), (mouse IgGl), was purchased from Prospec Protein Specialists (Rehovot, Israel).
Mouse serum (ICR SCID) was purchased from Bioreclamation, Inc. (Liverpool, NY).
4.2.1.2 HA/glycan Neutralization by Monoclonal Antibody

H5 HA (Vietnam) at 140nM was incubated with a 1:2 serial dilution of anti-H5

monoclonal antibody (shown to be neutralizing for H5 HA in standard hemagglutination

76

inhibition assays) for 10 min at 25 °C and injected over the glycan chip surface to
investigate the ability of the antibodies to neutralize the glycan/HS binding.
Concentrations of anti-H5 monoclonal antibody were 1:250, 1:500, 1:1000, 1:2000, and
114000. Binding was assessed by an increase in RU. After 25 min dissociation time, the
glycan surface was regenerated with 60 s of 10 mM glycine pH 2.5 and 30 s of 50 mM
NaOH at 100 111/mm.
4. 2. 1.3 Specificity Testing

The specificity of the assay was investigated using H1, anti-H1, H3, anti-H3, and
glycans nonspecific for H5, 6’SLN (a2,6 binder), GD2 ((12,8 binder), and CT/Sda (a2,3
binder). H1 at 140 nM was preincubated with 1:2 serial dilution of anti-H1 or anti-H5 and
injected over the same glycans. H5 at 140 nM was preincubated with 1:2 serial dilution of
anti-H3 and injected over glycans GD2, 3’SLN, and 6’SLN to observe cross-protection.
Testing for each dilution was performed in triplicate. The anti-H1 and anti-H3 antibodies
were polyclonal preparations, and while this does not offer optimal comparison to the
neutralizing activity of the monoclonal anti-H5, reagent availability for these different
Influenza strains necessitated these comparisons for proof-of-concept. Also, the multiple-
epitope recognition ability of a polyclonal population would better mimic the complexity
of a natural patient plasma sample, and thus probing the cross-reactivity of these anti-H1
and anti-H3 polyclonal antibodies may in fact offer a more application-authentic

evaluation.

77

4.2.2 Characterization Studies
4.2.2.1 Antibody Testing

An antibody that binds outside of the receptor binding domain was of interest for
future application in the biosensor format. The ability of the anti-HA2 H5 polyclonal
antibody to bind to the glycan/HS precomplex was investigated. H5 at 140 nM was
injected over the immobilized glycans for 10 min at 5 III/min. After 1 min dissociation
and no regeneration, a 1:2 serial dilution of anti-HA2 H5 monoclonal antibody was
injected over the glycan/H5 complex for 5 min at 5 ul/min. After regeneration, the
experiment was repeated with anti-H5 monoclonal antibody.
4. 2.2.4 Serum Experiments

The complexity of biological samples was considered, with the ultimate goal of a
FLUIGIV screening assay in mind. The neutralization experiment described in 4.2. 1.2
was repeated with H5 at 140 nM prepared in mouse serum (ICR SCID) at 1% final
concentration by volume. Background binding to glycans 3’SLex, CT/Sda, and 3’SLN
and the blank cell was investigated using mouse serum (ICR SCID) at 0.5-10% in buffer.
4. 2. 2.5 Statistical Analysis
Each experiment was performed in triplicate to nullify the effect of equipment or

user variation. The samples were double referenced, by subtracting either the blank fcl or
nonbinders CT/Sda, GD2, or 6’SLN from the binder results, as well as subtracting a
buffer run to compensate for irrelevant machine fluctuations. The SA chips were assumed
to have the same physical properties, and the glycans were assumed to be immobilized to
saturation. The peak RU at the end of the injection cycle was taken as an indicator of

binding. The effects of different HAS, glycans, anti-HA antibodies, and HA concentration

78

were assessed to calculate the lower detection limit and specificity of the SPR-based
assay. The differences between the means of each sample were calculated and analyzed
based on single factor analysis of variance (ANOVA) to a significance of 95% ((1 = 0.05)

(Tables A-l, A-2), using SAS software.

79

4.3 OBJECTIVE 3
H5Nl—Targeted Biosensor Design
This objective is aimed towards the development of an electrochemical biosensor

for the detection of the same glycan/hemaggglutinin binding described in 4.1.
4.3.1 Biosensor Design

4. 3. 1.1 Reagents and Chemicals

The biotinylated carbohydrate compounds 3’SLex (B157), 3’SLN (B84), GT3

(B108), and 6’SLN (B87) were provided by the Carbohydrate Synthesis/Protein
Expression Core of The Consortium for Functional Glycomics fimded by the National
Institute of General Medical Sciences grant GM62116. The following reagent was
obtained through the NIH Biodefense and Emerging Infections Research Resources
Repository, NIAID, NIH: H5 Hemagglutinin (HA) Protein from Influenza Virus,
AN ietnam/ 1203/04 (H5N1), Recombinant from baculovirus, NR-10510 (Source A H5,
referred to as H5). The following reagent was obtained through the NIH Biodefense and
Emerging Infections Research Resources Repository, NIAID, NIH: Monoclonal Anti-
Influenza Virus H5 Hemagglutinin (HA) Protein (VN04-2), A/Vietnam/1203/04 (HSNl),
(ascites, Mouse), NR-2728. 6xHis tagged H5 hemagglutinin (HA) protein from 293 cell
culture, AN ietnam/ 1203/04 (H5Nl) (Source B H5, referred to as H5*); C—terminal 6xHis
tagged H1 hemagglutinin (HA) protein from 293 cell culture, A/ South Carolina/ N] 8
(HlNl); and polyclonal anti-influenza virus H1 hemagglutinin (HA) protein, H1N1/Pan,
(rabbit), were purchased fi'om Immune Technology Corp. (New York, NY).

All solutions and buffers used in the biosensor study were prepared in de-ionized

(DI) water (from Millipore Direct-Q system). Iron (III) oxide (y-Fe203) nanopowder,

80

aniline monomer, ammonium persulfate, hydrochloric acid (HCl), methanol, diethyl
ether, hydrogen tetrachloroaurate (III) trihydrate, sodium citrate dehydrate,
glutaraldehyde, Polysorbate-20 (Tween-20), phosphate buffered saline (PBS), trizma
base, casein, sodium phosphate (dibasic and monobasic), and streptavidin were purchased
from Sigma-Aldrich (St. Louis, MO). Solutions were prepared as follows: PBS buffer (10
mM PBS, pH 7.4), wash buffer (10 mM PBS, pH 7.4, with 0.05% Tween-20), phosphate
buffer (100 mM phosphate buffer, pH 7.4), casein blocking buffer (100 mM Tris-HCl
buffer, pH 7.6, With 0.1% w/v casein), and glycine blocking buffer (67 uM glycine in 10
mM PBS, pH 7.4). HBS-P buffer, 10 mM glycine pH 2.5, and 50 mM NaOH were
purchased from GE Healthcare (Piscataway, NJ). Avidin/Biotin blocking kit was
purchased fi'om Vector Laboratories, Inc. (Burlingarne, CA). Mouse serum (ICR SCID)

was purchased from Bioreclamation, Inc. (Liverpool, NY).

4.3.1.2 Methodology of Supporting SPR Assay

Glycan partners were chosen for the HAs of interest based on widely accepted HA
specificities, as previously investigated using glycan microarrays (Blixt et al., 2004;
Stevens et al., 2006). Biotinylated glycans were diluted to 1 M in Biacore HBS-P buffer
and 8 III were injected over a Biacore Streptavidin (SA) chip at 10 III/min. Glycans were
immobilized to saturation at approximately 300 resonance units (RU). H5 HA (V ietrrarn)
at 140 nM was incubated with a serial dilution of anti-H5 monoclonal antibody (shown to
be neutralizing for H5 HA in standard hemagglutination inhibition assays) for 10 nrin at
25 °C and injected over the glycan chip surface to investigate the ability of the antibodies
to neutralize the glycan/HS binding. Binding was assessed by an increase in RU. After 25

rrrin dissociation time, the glycan surface was regenerated with 60 s of 10 mM glycine pH

81

2.5 and 18 s of 50 mM NaOH at 100 til/min. Anti-H5 monoclonal antibody binding to the
glycan/H5 complex was also investigated. H5 at 140 nM was injected for 10 min at 5
III/min. After 1 min dissociation and no regeneration, anti-H5 monoclonal antibody was
injected over the glycan/H5 complex for 5 min at 5 ul/min.
4. 3. 1.3 Biosensor Architecture

The platform for electrochemical detection of the HA of interest was a screen
printed carbon electrode (SPCE) comprised of three electrodes, including a working,
common reference, and counter electrode, screen printed on low-cost polyester backing
(Gwent Group, UK). The overall dimension of the sensor chip was 22 x 12 mm, with a
4mm diameter working electrode surrounded by a 1.5 mm wide, partially circular (270°)
common reference and counter electrode. The working electrode was composed of
carbon and the common reference and counter electrode of silver/silver chloride
(Ag/AgCl). Manufacturer’s specifications listed the resistance of the carbon as 50 Ohms

at 12 microns and the resistance of the silver as 320 mOhms at 25 microns.

4.3.1.4 Gold Nanoparticle Synthesis

Application of the carbon-based glycans directly onto the screen-printed carbon
electrode would result in an insulating device. To enhance the electron transducer, thus
amplifying response crurent and improving detection limits, gold nanoparticles (AuNPs)
were applied to the SPCEs (Daniel and Astruc, 2004; Lin et al., 2008; Willner et al.,
2007). AuNPs were synthesized according to a published procedure and their size,
spectroscopic properties, and magnetic profiles have been previously characterized (Hill
and Mirkin, 2006; Zhang et al., 2009). The referenced synthesis procedure required

hydrogen tetrachloroaurate (III) trihydrate aqueous solution (lmM, 50 mL) to be stirred

82

while heated. Vigorous reflux was achieved, followed by titration with 5 mL of 38.8 mM

sodium citrate. The solution shifted from yellow to the deep red characteristic of the

AuNPs.
4.3.2 Electrically Active Polyaniline Coated Magpetic Nanoparticles

4.3.2.1 EAM Synthesis

Aniline monomer was polymerized around gamma iron (III) oxide (y-F e203) cores
to obtain magnetic/polyaniline core/shell (c/s) nanoparticles (Sharma et al., 2005).
Commercially manufactured y-Fe203 nanoparticles were sonicated and dispersed in 50
ml of 1 M HCl, 10 ml deionized (DI) water, and 0.4 ml aniline monomer at 0 °C for 1 h.
The y-F e203 : monomer weight ratio was fixed at 1:0.6. 1 g ammonium persulfate in 20
ml DI water was added as oxidant while the mixture was stirred at 0 °C. As electrically-
active polyaniline, typically green, was formed over the y-Fe203 nanoparticles, typically
brown, the color of the solution visibly transitioned from rust brown to dark green. The
reaction proceeded for 4 h with continuous stirring at 0 °C. The solution was filtered,
washed with 1 M HCl, 10% methanol, and diethyl ether, and dried for 18 h. The resulting
green solid was ground into fine powder and stored in a vacuum desiccator.
4.3.2.2 EAM Nanoparticle Characterization

The electrically-active magnetic/polyaniline c/s NPs have been previously
characterized in terms of structure, size, magnetization, and conductivity (Pal et al.,
2008a; Pal and Alocilja, 2009). Magnetic characterization and room temperature
hysteresis measurements of the EAMs were performed by Pal and Alocilja (2009) using a
superconducting quantum interference device (Quantum Design MPMS SQUID). A

magnetic field cycling range of + 20 kOe to -20 kOe at 300 K constant temperature was

83

used to measure M-H hysteresis loops. The effect of polyaniline on the saturation
magnetization (Ms) values of the EAMs were calculated. Super paramagnetic behavior
was investigated by calculating coercivity (He) and retentivity (MR) of the EAMs.
Blocking temperatures of the EAMs were also investigated using zero field cooled-field
cooled (ZFC-FC) measurements from 5 K to 300 K temperature range and 100 Oe
applied magnetic field.

The solid form of the EAMs was evaluated for electrical conductivity. A hydraulic
press (Fisher Scientific, NJ) applied 10,000 psi to compress approximately 0.25 grams of
sample into 1.5-2 mm thick pellets. A Four Point Probe (Lucas/Signaton Corporation,
Pro4, CA) then measured room temperature electrical conductivity (Pal and Alocilja,

2009).

4.3.2.3 EAM Immunofunctionalization

EAM nanoparticles were immunofunctionalized with either anti-H5 monoclonal
antibody IgG2 or anti-H1 polyclonal antibody. Desiccated EAM polyaniline
nanoparticles were dissolved in 100 mM phosphate buffer (pH 7.4) to obtain a
concentration of 10 mg/ml, and sonicated for 15 min. The EAM polyaniline nanoparticles
were then conjugated with anti-H5 monoclonal antibodies by direct physical adsorption
as previously described and confirmed by Pal and Alocilja (2009). Anti-H5 monoclonal
antibody IgG2 (mouse ascites fluid) or anti-H1 polyclonal antibody (rabbit) was added to
the EAM polyaniline nanoparticles to obtain an antibodyzEAM ratio of 1:10 by volume.
The solution was incubated for 1 h at 25 °C in a rotational hybridization oven (Amerex
Instruments, Inc., Lafayette, CA). Following adsorption of antibody onto the EAM

nanoparticles, the immunofunctionalized nanoparticles were magnetically separated using

84

a F lexiMag Magnetic Separator (Spherotech, Inc., Lake Forest, IL) to remove any
unbound antibody in the supernatant. The anti-HA—EAM complexes were washed twice
with blocking buffer consisting of 100 mM tris—HCl buffer (pH 7.6) with 0.1% (w/v)
casein with magnetically separated supernatant discarded each time. The anti-HA—EAM
complexes were then resuspended in 100 mM phosphate buffer (pH 7.4). The anti-HA—
EAM complexes were prepared on the day of testing and stored at 4 °C until use.
4.3.2.4 EAM Structural Characterization

The structural morphologies of the EAMs and immunofunctionalized EAMs were
analyzed using a transmission electron microscope (TEM, Japan Electron Optics
Laboratories, JEOL 100CX 11). Selected area electron diffraction performed by the
200kV JEOL 2200 field emission TEM was used to study the crystalline nature of the
EAMs.
4.3.2.5 Spectral Analysis

Pa] and Alocilja (2009) previously analyzed the UV-visible spectra of the EAMs
using a UV-VIS-NIR scanning spectrophotometer (UV -3101PC, Shimadzu, Kyoto,
Japan). EAMs at 10 mg/ml were dispersed in de-ionized water by sonication for 10 min.
the nanoparticle suspension was transferred to a quartz cuvette and the sample was
scanned with a 300 to 1000 nm wavelength range using a step size of 1 nm to determine
absorbance.

4.3.3 Biosensor Fabrication

4.3.3.1 SPCE Modification

SPCE chips were prepared by removing the overlaying mesh and foam (Gwent,

Inc., UK). Each chip was washed with 2 ml sterile DI water and air dried for 15 min. As

85

described in Lin et al. (2008), 25 pl of 2.5 mM glutaraldehyde solution as crosslinker
were applied to the working area and incubated at 4 °C for l h. The SPCEs were then
washed with 2 ml DI water and air dried at 25 °C for 15 min. 25 pl of AuNP solution
were applied to the glutaraldehyde-treated working electrode and incubated at 4 °C for 1
h. The SPCEs were then washed with 2 ml DI water and air dried at 25 °C for 15 min. 20
pl of streptavidin at 1 pg/ml were applied to the working area and dried at 4 °C for 2 h or

ovenright.
4.3.4 Preconcentration Preparation Technigue

4.3.4.1 Sample Preparation

Glycans were prepared at 3x desired concentration in 0.01 M PBS. HAS were
prepared at 3x desired concentration in 0.01 M PBS with 10% mouse serum (ICR SCID)
by volume. 30 pl each of glycan and HA were incubated for 15 min at 25 degrees C in a
rotational hybridization oven. 30 pl of the appropriate anti-HA—EAM complex was then
added to the glycan/HA solution and incubated for 20 min at 25 degrees C in a rotational
hybridization oven. The glycan/HA/anti-HA—EAM complexes were magnetically
separated and washed twice with 0.01 M PBS containing 0.05% Tween-20 for 5 minutes

and resuspended in 0.01 M PBS.

4. 3. 4.2 Capture Experiments
The SPCE chips prepared with glutaraldehyde, AuNPs, and streptavidin were then
treated with the biotinylated glycan/HA/anti-HA—EAM complex. 90 pl of the solution

was applied to the treated SPCE and incubated at 25 degrees C for 15 min. The SPCE

was washed with 2 ml DI water and air dried at 25 degrees C for 15 min (Figure 13).

86

4.3.5 Stepwise Preparation Technigue
4.3.5.1 Sample Preparation and Capture Experiments

25 pl of the desired glycan concentration were added to the working area of the
glutaraldehyde, AuNPs, and streptavidin treated electrode and allowed to incubate at 25
degrees C for 30 min. Excess was rinsed with 2 ml DI water and air dried at 25 degrees C
for 15 min. Available sites were blocked with sequential additions of 25 pl Avidin D and
biotin solutions for 30 nrin each, with DI water rinse and air dry after each. 25 pl of the
desired H5 concentration were added, incubated at 25 degrees C for 30 min, rinsed with 2
ml DI water, and air dried at 25 degrees C for 15 min. 100 pl of anti—HA—EAM complex
solution were added to the electrode, incubated at 25 degrees C for 15 min, rinsed with 2

ml DI water, and air dried at 25 degrees C for 15 min (Figure 12).

87

(a) SPCE

Working electrode
(carbon)

Counter/Reference electrode
(silver/silver chloride)

 

‘b’ —0

Current

Potential

 

 

 

 

Reference electrode

Counter electrode Working electrode

Figure 11. Testing schematic. (a) Screen-printed carbon electrode (SPCE) consisting
of two electrodes: carbon working electrode and silver/silver chloride
counter/reference electrode, (b) schematic of the three electrode voltammetry
system (adapted and modified from Bard and Faulkner, 2000).

88

EAM .0 ~Y* anti-HA Antibody
.O

 

Preincubate anti-HA Ab Magnetically separate Ab-modified

+ EAM nanoparticles EAMs, discard supernatant, wash

Glycans HAs Ab-

“ QC) EAMs 6‘;

I. 60 co 8

SPCE I I
working %
electrode

I M m
Apply 25ml glycans to Apply 25 ml HA to Apply 100 ml anti-
SPCE working working electrode. HA Ab - EAM
electrode. Incubate 30 Incubate 30 min, conjugates. Incubate
min, wash, dry. wash, dry. 15 min, wash, dry.

Figure 12. Testing schematic. Stepwise preparation method.

89

Glycans HA in 10%
{- mouse serum

 

 

AA "100
AA Q20
Ab-EAMS
93,0 ‘8: 5

 

J 9’ ,. —> 509

I ~ I
\G'/ \5/
30 pl glycans + Add 30 pl anti-HA Magnetically
30 pl HA in 10% Ab - EAM complex. separate

 

 

 

 

 

mouse serum. Incubate 20 min. glycan/HA/Ab/EAM

Incubate 15 min. complexes from
mouse serum.
Wash.

GchaanA/AbIEAM .53.. 0.0.5
\I

complexes
as
working

electrode Apply 90 “I
glycan/HA/Ab/EAM
complexes.

Incubate 15 min, wash, dry.

Figure 13. Testing schematic. Preconcentration preparation method.

90

4.3.6 Biosensor Testing
4. 3. 6.1 Testing Apparatus

Cyclic voltarnmetric measurements were performed using a 263A
potentiostat/galvanostat (Princeton Applied Research, MA, USA) connected to a personal
computer. Data collection and analysis were controlled through the PowerSuite
electrochemical software operating system (Princeton Applied Research, Wellesley,

MA). SPCE chips purchased from Gwent Inc. (UK) are shown in Figure 11.

4. 3. 6.2 Detection and Data Analysis

100 pl of 0.1 M HCl solution were applied to cover the entire SPCE electrode area
and allowed to incubate for 5 min. The SPCE electrodes were connected to the
potentiostat and cyclic voltammetry was performed at a scan rate of 55 mV/sec and a
cyclic scan range of -0.4 to 1 V, with four consecutive 2 min scans recorded (Figure 14).
Previous experimentation indicated that the third scan produced the most pronounced
current flow differences for different samples and was chosen for analysis. For each
experiment, including positive and negative controls and blanks, three replications were
performed. The samples were calibrated against a negative control, also repeated in
triplicate, which consisted of the anti-HA—EAM application step alone. The total charge
transferred, AQ, was computed from the cyclic voltammograrn as the integral of current,
according to the relationship

I = AQ/At (6)
where, I = current (A), AQ = charge transferred (C), and At = time elapsed (s)
(Kuzrretsov, 1995). The AQ values described in this paper were calculated from the

current and time interval data generated by the potentiostat. Standard deviations and

91

mean AQ values of the third scans for the triplicate data sets were calculated.

The presence of the target is indicated by an increase in total charge transferred into
the SPCE surface. Target HA labeled with the immunofunctionalized EAMs were
captured on the SPCE surface, and the EAMs, consisting of conductive polyaniline
synthesized around a magnetic y-Fe203 core, were made electrically active by acid
doping. An applied external cyclic potential causes polyaniline to switch redox states,
transferring charge into the SPCE surface. Higher current recorded by the potentiostat

indicates more target in the sample (Figure 14).

92

 

Figure 14. SPCE and potentiostat setup.

93

4. 3. 6.3 Sensitivity and Specificity Testing

The lowest detection limit of the biosensor for H5 was investigated. The prepared
biosensors were tested using three samples at 1:2 dilution in 0.01 M PBS to obtain H5 at
1.4 pM, 700 nM, and 360 nM. Testing for each dilution was performed in triplicate. Anti-
HA—EAM complexes without glycan or HA were tested as the control. The lowest
dilution of H5 that produced a signal distinguishable from the control was taken as the
sensitivity of detection, but because our H5 dilution series was limited to three samples
by reagent availability, this sensitivity is not a conclusive analytical sensitivity, but the
detection limit for the experimental concentrations tested here.

The specificity of the biosensor was investigated using H1, anti-H1, and glycans
nonspecific for H5, GT3 (a2,8 binder), 6’SLN (a2,6 binder), and 6’S-Di-LN (a2,6
binder). The H1 was prepared at 1.4 pM in 0.01M PBS, the non-H5 binding glycans were
prepared at 100 pM, and the EAMs were immunofunctionalized with anti-H1 at 1:10
using the method described in 2. 6.

4.3. 6.4 Complex Matrix Testing

The complexity of biological samples was considered, as the ultimate application of
the biosensor as an in-field detection system would require testing of blood or sputum
samples. In the preconcentration method, the HA samples were prepared to consist of
10% mouse serum. After complexing the glycan/HA/anti-HA—EAM, the magnetic
separation and washing technique was investigated for its ability to specifically isolate

the target HA from a complex serum matrix.

94

4. 3. 6.5 Statistical Analysis

Each sample preparation was tested in triplicate with the biosensors to account for
the effect of equipment or user variation. The prepared biosensors were assumed to have
the same physical properties. For each experiment, the cyclic voltammetry (CV) data
were obtained as a curve of current versus potential (I vs. E), including 1020 points for
each scan cycle from -0.4 to 1 V. The mean and standard deviations of the AQ values
were calculated for each sample preparation, including negative controls and blanks. The
differences between the means were calculated and analyzed based on single factor
analysis of variance (ANOVA) to a significance of 95% (CI = 0.05) (Tables A-3-A-5),
using SAS software. The effects of different HAS, glycans, anti-HA antibodies, and HA
concentration were assessed to calculate the lower detection limit of the biosensor as well
as the biosensor specificity. Oxidation (anodic) and reduction (cathodic) peak currents
were also determined from the CV data, which consisted of an oxidation reaction (first
half) and a reduction reaction (second half). The peak currents were determined at the

corresponding peak potentials for each experimental run.

95

4.4 OBJECTIVE 4
Biosensor to Distinguish (12,3 v. (12,6 Receptor Binding
This objective is aimed towards modification of the previously described

electrochemical biosensor (Objective 3) to detect a2,6 receptor specificity as an indicator
of pandemic potential. The ability to distinguish between a2,3 and (12,6 linked receptors
was also important for application in the event that a historically avian (a2,3) F LUAV
strain acquires hrunan (a2,6) transmissibility.
4.4.1 Biosensor Desim
4. 4. 1.1 Reagents and Chemicals

The biotinylated carbohydrate compounds 3’S-Di-LN (B178) and 6’S-Di-LN
(B179) were provided by the Carbohydrate Synthesis/Protein Expression Core of The
Consortium for Functional Glyconrics funded by the National Institute of General
Medical Sciences grant GM62116. The following reagent was obtained through the NIH
Biodefense and Emerging Infections Research Resources Repository, NIAID, NIH: H5
Hemagglutinin (HA) Protein from Influenza Virus, AN ietnam/ 1203/04 (H5N1),
Recombinant from baculovirus, NR-10510 (Source A H5, referred to as H5). The
following reagent was obtained through the NIH Biodefense and Emerging Infections
Research Resources Repository, NIAID, NIH: Monoclonal Anti-Influenza Virus H5
Hemagglutinin (HA) Protein (VN 04-2), AN ietnam/ 1203/04 (HSNl), (ascites, Mouse),
NR-2728. 6xHis tagged H5 hemagglutinin (HA) protein from 293 cell culture,
AN ietnam/ 1203/04 (H5N1) (Source B H5, referred to as H5*); C-terminal 6xHis tagged
H1 hemagglutinin (HA) protein ftom 293 cell culture, A/South Carolina/ 1/18 (HlNl);

and polyclonal anti-influenza virus H1 hemagglutinin (HA) protein, HlNl/Pan, (rabbit),

96

were purchased from Immune Technology Corp. (New York, NY). All solutions and
buffers used in the biosensor study were obtained and prepared as in 4.3.1.1.
4.4.1.2 Biosensor Fabrication
The gold nanoparticles were prepared as in 4. 3. 1.4. The EAMs were synthesized,

immunofunctionalized with anti-H5 or anti-H1 antibodies, and characterized as described
in 4. 3.2. 1-4. 3.2. 5. The SPCES described in 4.3.1.3 were treated as in 4.3.3.1.
4.4.2 Biosensor Testing
4. 4. 2.1 Sample Preparation

The preconcentration preparation method was followed. Glycans were prepared at
3x desired concentration in 0.01 M PBS. HAS were prepared at 3x desired concentration
in 0.01 M PBS with 10% mouse serum (ICR SCID) by volrune. 30 pl each of glycan and
HA were incubated for 15 min at 25 degrees C in a rotational hybridization oven. 30 pl of
the appropriate anti-HA—EAM complex was then added to the glycan/HA solution and
incubated for 20 min at 25 degrees C in a rotational hybridization oven. The
glycan/HA/anti-HA—EAM complexes were magnetically separated and washed twice
with 0.01 M PBS containing 0.05% Tween-20 for 5 minutes and resuspended in 0.01 M

PBS.

4. 4. 2.2 Capture Experiments
The SPCE chips prepared with glutaraldehyde, AuNPs, and streptavidin were then
treated with the biotinylated glycan/HA/anti-HA—EAM complex. 90 pl of the solution

was applied to the treated SPCE and incubated at 25 degrees C for 15 min. The SPCE

was washed with 2 ml DI water and air dried at 25 degrees C for 15 min (Figure 13).

97

4. 4. 2.3 Detection and Data Analysis

Cyclic voltammetry was performed on the treated SPCES using the potentiostat as
described in 4. 3. 6.2. For each experiment, including positive and negative controls and
blanks, three replications were performed. The samples were calibrated against a negative
control, also repeated in triplicate, which consisted of the anti-HA—EAM application
step alone. The AQ values were calculated from the current and time interval data
generated by the potentiostat. Standard deviations and mean AQ values of the third scans
for the triplicate data sets were calculated.

The presence of the target is indicated by an increase in total charge transferred
across the electrodes. Target HA labeled with the immunofunctionalized EAMs were
captured on the SPCE surface, and the EAMs, consisting of conductive polyaniline
synthesized around a magnetic y-Fe203 core, formed an electrical circuit between the

silver electrodes, with current recorded by the potentiostat.

4. 4. 2.4 Sensitivity and Specificity Testing

The lowest detection limit of the biosensor for H5 and H1 was investigated. The
prepared biosensors were tested using two samples at 1:2 dilution in 0.01 M PBS to
obtain H5 at 1.4 pM and 700 nM. Testing for each dilution was performed in triplicate.
Anti-HA—EAM complexes without glycan or HA were tested as the control. The lowest
dilution of H5 or H1 that produced a signal distinguishable from the control was taken as
the sensitivity of detection. The H5-targeted biosensor required H5 incubation with 02,3-
linked glycan 3’S-Di-LN and EAM immunofimctionalization with anti-H5, while the H1-
targeted biosensor required H1 incubation with a2,6-linked glycan 6’S-Di-LN and EAM

immunofunctionalization with anti-H1.

98

The specificity of the biosensor for H5 was investigated using H1, anti-H1, and
a2,6-linked glycan 6’S-Di-LN. The specificity of the biosensor for H1 was investigated
using H5, anti-H5, and (12,3-linked glycan 3’S-Di-LN.

The specificity of the biosensor was investigated using H1 , anti-H1, and glycans
nonspecific for H5, GT3 (a2,8 binder), 6’SLN (a2,6 binder), and 6’S-Di-LN (a2,6
binder). The H1 was prepared at 1.4 pM in 0.01 M PBS, the non-H5 binding glycans
were prepared at 100 pM, and the EAMs were immunofunctionalized with anti-H1 at
1:10 using the method described in 2. 6.

4. 4. 2.5 Complex Matrix Testing

The complexity of biological samples such as blood or respiratory samples was
considered. The HA samples were prepared to consist of 10% mouse serum. After
complexing the glycan/HA/anti-HA—EAM, the magnetic separation and washing
technique was investigated for its ability to specifically isolate the target HA from a

complex serum matrix.

4. 4. 2.6 Statistical Analysis

Each sample preparation was tested in triplicate with the biosensors to nullify the
effect of equipment or user variation. The prepared biosensors were assumed to have the
same physical properties. For each experiment, the cyclic voltammetry (CV) data was
obtained as a curve of current versus potential (I vs. E), including 1020 points for each
scan cycle from -0.4 to 1 V. The mean and standard deviations of the AQ values were
calculated for each sample preparation, including negative controls and blanks. The
differences between the means were calculated and analyzed based on single factor

analysis of variance (AN OVA) to a significance of 95% ((1 = 0.05) (Table A-3-A-5),

99

using SAS software. The effects of different HAS, glycans, anti-HA antibodies, and HA
concentration were assessed to calculate the lower detection limit of the biosensor as well
as the biosensor specificity. Oxidation (anodic) and reduction (cathodic) peak currents
were also determined from the CV data, which consisted of an oxidation reaction (first
half) and a reduction reaction (second half). The peak currents were determined at the

corresponding peak potentials for each experimental run.

100

CHAPTER 5: RESULTS AND DISCUSSION

5.1 OBJECTIVE 1
Surface Plasmon Resonance—based Binding Assay
5.1.1 SPR Binding Assay Desigp
5.1.1.1 Glycan Immobilization and Chip Stability

Glycans were chosen based on their predicted binding to the HAS of interest. On
Chip 1, H5 is predicted to bind 3’SLex and 3’SLN but not CT/Sda. On Chip 3, H5 is
predicted to bind 3’SLN but not GD2 or 6’SLN. 3’SLN and 6’SLN were chosen for
comparison because their sialylated receptors are similar in structure except for the a2,3
versus a2,6 linkage. The samples were double referenced, by subtracting either the blank
fcl or nonbinder CT/Sda from the binder results, as well as subtracting a buffer run to
compensate for irrelevant machine fluctuations. Chips 1 and 3, immobilized with glycans
to saturation, were found to be stable and reusable over a 3 month period, exhibiting
repeatable binding to H5. The regeneration reagents, glycine 10 mM pH 2.5 and 50 mM
NaOH, reliably removed bound HA from the surface, without interfering with the
streptavidin/biotin linkage or damaging the biological activity of the glycan receptors.
The minimrun immobilization level on chip 2 did not provide a reliable RU level, with
fluctuations indistinguishable from background noise. This low glycan immobilization

did not yield repeatable binding with H5.

101

(a) 1600 A / 286 nM
1200
300 94.3 nM
400 /
O / 31.4 nM
. ‘< 10.6 nM
D

0 500 1000 1500 2000

Response UnIts (RU)

 

 

 

 

 

Time (s)
A
(b) S 1600

EC,

é, 1200

5 800

‘1’ 94.3 nM

4

g 00 A— / 31.4 nM

o

a: b

0 500 1000 1500 2000
Time (s)
(c) A

S
96 1600 286 nM
% 1200 94.3 nM
D
g 300 31.4 nM
g 400 10.6 DM
3 3.53 nM
a, 0
o: D

 

 

0 400 800 1200 1600 2000
Time (s)

Figure 15. Glycan/H5 binding experiments. Comparing glycans and regenerations.
(a) Triplicates of H5 dilutions binding to 3’SLN; Regeneration: 60 s of 10mM
glycine pH 2.5 and 30 s of 50mM NaOH at 100 pl/min, (b) triplicates of H5 dilutions
binding to 3’SLN; Regeneration: 60 s of 10mM glycine pH 2.5 and 18 s of 50mM
NaOH at 100 pl/min, and (c) triplicates of H5 dilutions binding to 3’ SLex;
Regeneration: 60 s of 10mM glycine pH 2.5 and 30 s of 50mM NaOH at 100 pI/min.

102

 

 

 

 

 

 

 

5 40° ‘ H5 140 nM
g 300
(I)
'2‘ 200
3 100 H5 140 nM + anti-H5 1:250
g léHs 140 nM + anti-H5 1:500
3 0 H5 140 nM + anti-H5 1:1000
3 Anti-H5 1:250
0: —100 W

0 400 800 1200 1600 2000

(b) Time (s)
5 400 4
E
,2 300 H3:
5 200 286 nM
o 94.3 nM
‘8 100 31.4 nM
g 0 10.6 nM
can, 3.53 nM
-100 r
0 400 800 1200 1600 2000
Time (s)

Figure 16. H5 Indonesia and H3 Wyoming binding to 3’SLN. Single replicate shown
for clarity. (a) H5 (A/Vietnam/1203/04) 140nM; H5 (A/Indonesia5/05) 140nM; H5
Indonesia 140nM + anti-H5 Indonesia 1:250, 1:500, and 1:1000; anti-H5 Indonesia
1:250 and (b) H3 (A/Wyoming/3/03) at 286nM, 94.3nM, 31.4nM, 10.6nM, and
3.53nM.

5.1.2 SPR Binding Between Gucan Receptors and Hemagglutinin
5.1.2.1 Confirmation of H5 Recognition

SPR analysis demonstrated a high avidity, specific binding between HS-specific
012,3-linked glycan receptors and recombinant H5 HA (A/Vietnam/1203/04). The SPR
binding curves were not fit to a model because the binding of the aggregated H5 yielded

an interaction that was not 1:1. The glycan/HS binding also did not reach equilibrium.

103

The aggregated nature of the H5 would result in many available receptor binding
domains on one complex, and could lead to binding and rebinding of the HA to the
immobilized glycans as the HA “walks” along the surface. The kinetics and binding
constants are thus not explored here, in favor of yes/no binding results. The H5 molarity
range from 3.53-286 nM exhibited a dose response in which response level could be
correlated with H5 concentration (Figure 15). H5 at 3.53 nM was indistinguishable from
the negative controls as well as the blanks, and H5 at 10.6 nM was thus taken as the
lower limit of detection and the sensitivity of the system. Although both are H5-specific,
H5 showed higher avidity binding to 3’SLN than 3’SLex (Figure 15). Both of these
glycans offered the same lower limit of detection.
5.1.2.2 Chip Regeneration

The dilution series using the different regeneration schemes were compared, and
triplicates revealed better repeatability when the surface was regenerated with 60 s of 10
mM glycine pH 2.5 and 18 s of 50 mM NaOH at 100 pl/min, as compared to the 30 s
NaOH pulse (Figure 15). The 30 s NaOH pulse may have begun to strip the immobilized
glycan surface, preventing a repeatable level of HA to be bound.
5.1.2.3 Specificity Testing

The specificity of the H5-targeted system was investigated using H1, anti-H1, H3,
anti-H3, and HS-nonspecific glycans. Binding of H1 alone and anti-H1 alone to the
glycans GD2, 6’SLN, and 3’SLN revealed negligible binding that was not statistically
different fi'om buffer alone. Binding of preincubated H1 and anti-H1 also revealed
negligible binding to all glycans, indicating that despite the polyclonal nature of the anti-

Hl antibodies, there is little cross-reactivity between the Hl-based negative controls and

104

the HS-targeted SPR assay. Recombinant H3 HA (A/Wyoming/3/O3) prepared at the
same concentration range showed no cross-reactivity with the H5-specific glycans,
indicating that H3 can be used as an appropriate negative control for the H5 detection
system (Figure 16). The H5-nonspecific glycans, CT/Sda, GD2, and 6’SLN were
assessed for their binding to H5, and this was seen to be within the statistical range of a
buffer injection on the H5-specific glycan, or within the range of H5 binding to the blank
flow cell. H5 thus was shown to bind 012,3 receptors with higher avidity than a2,6 or
012,8 receptors, and we can thus conclude that the SPR assay offers H5-specificity based
on sialic acid receptor preference.
5.1.2.4 Clade Specificity

Recombinant H5 HA (A/Indonesia/S/OS) was found to show negligible binding to
the H5-specific glycans that was not statistically different from the H3 negative controls
or the buffer injections (Figure 16). This may indicate that the immobilized glycans are
specific for the Vietnam H5N1 strain from clade l but not for the Indonesia H5N1 strain
from clade 2.1.3 (WHO, 2009). The H5-specific glycans may be clade-specific.
Alternatively, the predicted composition of the Vietnam H5N1 strain as a large aggregate
may allow for stronger binding to the glycans, whereas an H5 preparation composed
mainly of monomers and trimers may offer lower affinity binding.
5.1.2.5 HA Receptor Binding Domain Binding Assessment

The receptor binding domain on HA is located within the HA1 sequence, and is
responsible for binding to the host glycan receptor. We investigated the binding of the
HA1 segment of H5 and H3 to glycans 3’SLex, CT/Sda, and 3’SLN. As compared to the

binding between glycans and recombinant H5, HA1 H5 showed negligible binding to H5-

105

specific and nonspecific glycans (Figure 17). HA1 H5 binding was not statistically
different fiom the negative controls or blanks. The HA1 H3 showed slight cross-
reactivity with H5-specific 3’SLex and 3’SLN, generating 2-5% of the signal produced
by corresponding concentrations of H5 (Figure 17). This may be attributable to slight
overlap of glycan specificities. In an attempt to improve glycan/HA1 H5 binding, the
binding experiment was repeated with running and sample buffers of HBS-P with 0.1%
BSA and 0.5% glycerol. HA1 H5 and HA1 H3 were compared to H5 140 nM and a
glycerol concentration curve from 0-1%. Under these conditions, HA1 H5 did not bind
any glycans, and the slight binding of HA1 H3 was also reduced to a similar level that is

statistically similar to the negative controls and blanks.

106

600

400

200

Response Units (RU)

—200

(b

V

600

400

200

Response Units (RU)

-200

/ H5 286 nM

/ H5 94.3 nM
k / H5 31.4 nM

l
I ‘— HA1 H5

P
0 500 1000 1500 2000

 

 

 

 

 

Time (s)

/ H5 286 nM

_ /H5 94.3 nM
‘//H5314nM

‘— HA1 H3
F

 

 

 

 

O 500 1000 1500 2000
Time (s)

Figure 17. (a) H5 at 286nM, 94.3nM, 31.4nM; HA1 H5 at 286nM, 94.3nM, 31.4nM,
and (b) H5 at 286nM, 94.3nM, 31.4nM; HA1 H3 at 286nM, 94.3nM, 31.4nM.

107

A
N
V

A 600
D

E

.9 400
'E

3 200
i

8. 0
0)

d)

o:

(b)

5‘ 300
93

g 200
C

D

(D 100
(D

S

g 0
o

CI

0 400 800 1200 1600 2000

Tween treatments:

/H5 1.4 mM
/ + 0.02%

 

 

 

 

 

 

 

 

 

1L \‘t 0.05%
\+ 0.1%
*
0 400 800 1200 1600 2000
Time (3)
Heat treatments:
A /37 deg C overnight
T— 37 deg C for 4 h
/ 37 deg C with
*r bromelain and 2-ME
+ for 4 h

 

 

V 37 deg c with
bromelain for 4 h.

Time (s)

Figure 18. H5 1.4pM pretreatments. (a) Tween-20 (b) heat treatment at 37 degrees
C overnight, 37 degrees C for 4 h, 37 degrees C with bromelain and 2-ME for 4 h,
and 37 degrees C with bromelain for 4 h.

5.1.2.6 HA Preparation

Because we conclude that the H5 is present as large aggregates, the HA was

pretreated in an effort to break down these complexes into more consistent trimer

preparations. Increasing Tween-20 from 0.02 to 0.1% resulted in a 50% drop in SPR

signal for H5 at 1.4 pM (Figure 18). Bromelain and 2-ME treatment served to completely

108

destroy H5 biological activity, and no glycan binding was observed with these
preparations (Figure 18). Heat treatment at 37 degrees C lowered the binding activity of
the H5 at this concentration, although 4 h treatment showed a five-fold reduction of
binding as compared to overnight heat treatment (Figure 18). A longer heat treatment
may have resulted in more consistent trimers while a short heat treatment only served to
shock the sample without generating any viable trimers. Heat treatment at 5 6 degrees C
for 10 min, with or without subsequent 4 degrees C treatment to cease heat effects,
completely destroyed binding ability. All pretreatrnents were thus ineffective in bringing
uniformity to the HA aggregates for a more repeatable signal, and all served to depress or
destroy the glycan/HA binding.
5.1.2. 7 Serum Experiments

The SPR binding assay was also repeated with mouse serum matrix at 1-2% by
volume. H5 at 286 nM, 94.3 nM, 31.4 nM, 10.6 nM, and 3.53 nM, both with and without
2% mouse serum (ICR SCID) by volume, were injected over glycans 3’SLex, CT/Sda,
and 3’SLN. The presence of 2% serum depressed the SPR signal, but the H5
concentrations retained a dose response, if statistically lower than the H5 dilution series
without serum. A comparison of each H5 concentration with 2% serum revealed a 70-
85% drop in SPR signal as compared to the corresponding H5 concentration prepared
without serum (Figure 22). H3, previously shown to offer negligible binding to the H5-
specific glycans, showed slight increase of binding signal with presence of 2% serum,
likely due to nonspecific binding effects.

Background binding of serum to the chip or immobilized glycans was investigated

by injecting varying concentrations of serum in buffer over the glycan surface with no

109

HA or antibody added. Serum at 10% showed approximately 10 RU of nonspecific
binding to 3’SLN, but lower concentrations were not statistically different from the
buffer injections with no serum added. This likely indicates that the addition of 1-2%
serum inhibits glycan/HA binding by nonspecifically binding to the HA, but we cannot
conclude that the serum binds nonspecifically to the glycans or the chip.
5.1.2.8 Structural Morphology Characterization

The synthetic glycans and recombinant HA were analyzed by a JEOL (Peabody,
MA) 100CX 11 Transmission Electron Microscope (TEM) to obtain their structural

morphologies (Figure 19). 1% uranyl acetate was used as stain.

 

Figure 19. TEM imaging of (a) synthetic glycans and (b) purified recombinant H5
HA.

110

5.2 OBJECTIVE 2
SPR to Detect Ab-Mediated Binding Inhibition
5.2.1 SPR Neutralization Assay Desigp
5.2.1.1 Neutralization Ability of Anti-H5 Monoclonal Antibody

Preincubating H5 with neutralizing anti-H5 monoclonal antibody resulted in a
neutralization of glycan/H5 binding on the SPR system. Anti-H5 monoclonal antibody
IgG2 (mouse ascites fluid) at 1:500 neutralized the binding between H5 at 140 nM and
H5-specific glycans 3’SLex and 3’SLN (Figure 20). The glycan/H5 binding was
significantly reduced by anti-H5 monoclonal antibody 1:2000, and the antibody
concentration range from 1:250 to 1:4000 displayed a reproducible dose response.

Convalescent H5N1 plasma has been reported to have a neutralizing antibody titer
when diluted to 1:80, and the range of antibody concentrations tested here can thus be
concluded to be within a physiologically relevant range (Zhou et al., 2007). The anti-H5
monoclonal antibody dilution of 1:4000 offered some glycan/HS binding inhibition, and
the most neutralizing dilution tested, 1:250, offered complete neutralization. Even if a
patient convalescent plasma contains a lower protein content than the anti-H5
monoclonal preparation, the requirement of a 1:80 plasma dilution falls far lower than the
tested monoclonal range of 1:250 to 1:4000, offering evidence that the neutralization
experiments described here have physiological relevance.
5. 2. 1.2 Neutralization Specificity

The glycan/H5 binding showed slight inhibition by anti-H1 polyclonal antibody at
1:250 but the anti-H1 did not cause complete neutralization as observed with anti-H5 at

the same concentration (Figure 20). The binding of H5 to glycans 3’SLex and 3’SLN was

111

also slightly inhibited by anti-H3 polyclonal antibody at 1:500 but this concentration did
not cause the complete neutralization observed with the same concentration of anti-H5
monoclonal antibody (Figure 20). These minor inhibitions of glycan/H5 binding by anti-
HI and anti-H3 polyclonals can be attributed to their composition as compared to the
anti-H5 monoclonal, as the nature of a polyclonal antibody offers a matrix similar in
complexity to a serum matrix, which was also seen to interfere with glycan/HA binding.
We conclude that the anti-H1 and anti-H3 polyclonals at high concentrations do interfere
with binding but not necessarily by binding within the receptor binding domain of H5 in

a CI'OSS-pl'OICCtIVC manner.

112

(a)

400

200

Response Units (RU)

(b)

400

200

Response Units (RU)

(C)

600
400
200

Response Units (RU)

 

H5 140 nM

H5 + anti-H5:

 

 

 

 

 

 

 

 

 

 

A
1:4000
1:2000
121000
1:500
/ 1:250
anti-H5 1:250
b
0 500 1000 1500 2000
Time (s)
A
A <— H5 140 nM
r 5 H5 + anti-H1 1:250
T“ a ‘— anti-H5 1:250
’
0 500 1000 1500 2000
Time (s)
A
‘— H5 140 nM
H5 + anti-H3 1:500
<—— anti-H312250
b
0 500 1000 1500 2000

Time (s)

Figure 20. Neutralization experiments. (a) 3’SLN/HS neutralization by anti-H5
monoclonal antibody: H5 140nM; H5 140nM + anti-H5 1:4000, 1:2000, 1:1000,
1:500, 1:250; anti-H5 1:250 only, (b) 3’SLN/H5 binding inhibition by anti-H1
(HlNl/Pan): H5 140nM; H5 140nM + anti-H1 1:250; anti-H1 1:250 only, and (c)
3’SLN/HS binding inhibition by anti-H3 (AlShandong/9/93): H5 140nM; H5 140nM

+ anti-H3 1:500; anti-H3 1:250 only.

113

(a)

S

E 600
.‘é’

5 400
3:3 200
O

3 o

O

m

(b)

S

E. 600
g

D 400
§ 200
O

B o

G)

D:

 

 

 

0 400 800 1200 1600 2000

Time (s)

 

 

/

‘\

 

b

0 400 800 1200 1600 2000

Time (s)

 

1 Injection 1:

H5 140 nM
Injection 2:
buffer

anti-HA2 1:500
anti-HA2 1:1000

L anti-HA2 122000

Injection 1:
anti-HA2 1:250
Injection 2:
buffer

Injection 1:
H5 140 nM
Injection 2:
anti-H5 neut mAb

Injection 1:
H5 140 nM
Injection 2:
Anti-HA2 1:250

Figure 21. Antibody binding to 3’SLN/HS precomplex. (a) Injection 1: H5 140nM,
Injection 2: buffer, anti-HA2 H5 1:500, 1:1000, or 1:2000; Injection 1: anti-HA2
1:250, Injection 2: buffer, and (b) Injection 1: H5 140nM, Injection 2: anti-H5
neutralizing monoclonal antibody or anti-HA2 H5 1:250.

5.2.1.3 Antibody Testing: Anti-HA versus Anti-1142

The ability of a polyclonal antibody against the HA2 portion of H5 to bind to the

already-formed glycan/HA complex was investigated. This was expected to offer binding

because the HA2 segment does not include the receptor binding domain, the area which

114

is utilized in glycan/HA binding. After an injection of H5 at 140 nM over glycans
3’SLex, CT/Sda, and 3’SLN, the anti-HA2 H5 polyclonal antibody at 1:500, 1:1000, and
1:2000 was injected before regeneration. The anti-HA2 H5 polyclonal antibody did not
show a further increase in RU, indicating that the antibody did not bind the glycan/HA
complex (Figure 21). This could be due to the aggregated nature of the H5, if the HA2
portion of the H5 was hidden within the aggregate. The sequential binding procedure was
repeated with anti-H5 monoclonal antibody, previously shown to be neutralizing due to
binding within the glycan/HA receptor binding domain. However, this monoclonal
exhibited a further increased SPR signal, which indicated that the second injection of
anti-H5 monoclonal antibody also bound to the already formed glycan/HS complex, thus
forming a glycan/HS/ anti-H5 complex (Figure 21). The monoclonal thus did not displace
the glycan but instead bound the H5 outside of the receptor binding domain.
Alternatively, this may also be a result of the aggregated nature of the H5, as the
monoclonal may have bound to an available receptor binding domain exposed on the
large aggregate. The anti-H5 monoclonal antibody may thus neutralize glycan/H5
binding when preincubated with H5, and may also additionally bind an already formed
glycan/H5 complex without displacing the glycan. This would present the anti-H5
monoclonal antibody as an appropriate reagent in a sandwich-type assay, if reaction
sequence is maintained.
5.2.1.4 Serum Experiments

The neutralization experiment described in 2. 6 was repeated with H5 at 140 nM
prepared in 1% mouse serum. The signal was similarly depressed as seen in the

glycan/HS + 2% serum binding experiment. Comparing H5 at 140 nM preincubated with

115

anti-H5 monoclonal antibody from 1:250 to 124000 both with and without 1% serum,
revealed a 20-70% drop in SPR signal for those samples prepared with serum (Figure

22). A depressed dose response was still observed for the serial dilutions of anti-H5

 

 

 

 

monoclonal antibody.

(3) A
E 150
.413 100
C
:3
a) 50
(D
C
a 0 ~
8
a: -50 5

0 400 800 1200 1600 2000
Time (s)
H5 140 nM

00) 5 A
86 H5 140 nM + anti-H5:
a? 400 1:4000
5 /1:2ooo
°’ r
g 200 H5 140 nM + 1% serum
Q.
(a fi
0 H5 140 nM + anti-H5
m 0 "\ I 1:4000 + 1% serum

 

 

0 400 800 1200 1600 2000 H5 140 nM + anti-H5
- 0
Time (s) (12000 +1 /0 serum

 

Figure 22. Serum effects. (a) H5 at 286nM, 94.3nM, 31.4nM, 10.6nM, and 3.53nM
prepared in 2% mouse serum binding to 3’SLN and (b) H5 140nM, H5 140nM +
anti-H5 1:4000, H5 140nM + anti-H5 1:2000, H5 140nM in 1% serum, H5 140nM +
anti-H5 1:4000 in 1% serum, H5 140nM + anti-H5 1:2000 in 1% serum.

116

5.3 OBJECTIVE 3

H5N1—Targeted Biosensor Design

5.3.1 Biosensor Desigp
5. 3. 1.1 Supporting SPR data

Biosensor work proceeded on the basis of previous SPR results. As described in
5.1., SPR analysis demonstrated a high avidity, specific binding between H5-specific
012,3-linked glycan receptors and recombinant H5, with concentration series offering an
observable dose response. Preincubating H5 with neutralizing anti-H5 monoclonal
antibody resulted in neutralization of glycan/HS binding on the SPR system. Anti-H5
monoclonal antibody IgG2 (mouse ascites fluid) at 12500 neutralized the binding between
H5 at 140 nM and H5-specific 012,3 linked glycans 3’SLex and 3’SLN (Figure 23; Table
4). Again, this antibody dilution falls within the range of physiological relevance, where
plasma 1:80 is neutralizing (Zhou et al., 2007). The glycan/H5 binding showed slight
inhibition by anti-H1 polyclonal antibody at 12250 but the anti-H1 did not cause complete

neutralization as observed with anti-H5 at the same concentration (Figure 23).

117

A
O
O

x
3’SLe immobilized:
‘— H5140 nM

/ H5140 nM + 1% serum
:=.- :— H5 140 nM + anti-H5 mAb 1:500

   

O

 

Response
Units (RU)
N
O
O

 

0 1000 2000 3000

 

 

 

 

 

 

 

 

 

 

 

 

Time (s)
a; 5 900 A
E g 600 3’SLN immobilized:
§ ,9 ‘\ H5 140 nM
g :3) 300 / H5140 nM + 1% serum
0 r“ t 4— H5 140 nM + anti-H5 mAb 1:500
a
0 1000 2000 3000
Time (s)
A , X. .. ,
o A 400 ‘_ 3SLe ImmobIIIzed.
g 3 H5 140 nM
3;; 20° ‘ <— H5 140 nM + anti-H1 pAb
£33; 0 j" ‘— H5*140nM
-200 5
0 1000 2000 3000
Time (s)
A 600 A 4 Injection 1:
3:1 a 400 H5 140 nM, 10 min, 5 pllmin
3‘5 200 Injection 2:
8 'E 0 antI-H5 mAb 12500,
a: 3 p 5 min, 5 pl/min

0 1 000 2000 3000

Time (s)

Figure 23. Supporting SPR results. (a) H5 140nM binding to H5-specific glycan
3’SLex, as inhibited by 1% mouse serum and anti-H5 monoclonal antibody 1:500,

(b) H5 140nM binding to H5-specific glycan 3’SLN, as inhibited by 1% mouse
serum and anti-H5 monoclonal antibody 1:500, (c) H5 140nM binding to H5-specific
glycan 3’SLex, as inhibited by cross-reactivity of anti-H1 polyclonal antibody; H5*
140nM binding to 3’SLex, and ((1) antibody testing on H5-specific glycan 3’SLN.

118

The order of interaction was found to be important, as described in 5. 2. 1.3. The
glycan/H5 binding was not neutralized when the same anti-H5 monoclonal antibody was
allowed to react with the already formed glycan/H5 complex. Following typical H5-
binding, a further increased SPR signal indicated that the second injection of anti-H5
monoclonal antibody also bound, forming a glycan/H5/ anti-H5 complex (Figure 23).
The subsequently added anti-H5 monoclonal antibody thus did not displace the glycan
but instead bound the H5 in a region outside of the receptor binding domain or in an
available binding domain if the H5 is present as a trimer or larger aggregate. This is in
contrast to the neutralization experiment, in which the anti-H5 monoclonal antibody
binds within, or otherwise blocks, the glycan receptor binding domain on H5. This anti-
H5 monoclonal antibody is thus appropriate for use in both the SPR neutralization assay
as well as the biosensor sandwich-type assay.

The SPR assay was also repeated with a 1% mouse serum matrix. Although binding
was still observed, the results indicated that the glycan/H5 binding was inhibited by the
addition of serum to the sample buffer (Figure 23).

5. 3. 1.2 Electrochemical Detection

The schematic representation of the detection mechanism of the EAM based
electrochemical biosensor illustrates the electrode architecture and sample preparation
methods. The detection principles is based on an electrochemical sandwich assay in
which a specific glycan functions as a capture probe while an antibody specific for HA
serves as a detector probe. The glycan is labeled with biotin, the HA is labeled with
EAMs, and the SPCE is modified with streptavidin. The glycans are anchored to the

SPCE via high affinity streptavidin-biotin interactions. In the stepwise preparation

119

method, capture probe, target, and detector probe are applied to the streptavidin modified
SPCE sequentially with wash and dry steps between each. In the preconcentration
preparation method, capture probe and target are preincubated, followed by incubation
with detector probe. Magnetic separation then removes unbound material, including
mouse serum, and the glycan/HA/anti-HA—EAM complexes are applied to the SPCE in
one step. Excess is washed and target present on the SPCE biosensor surface is detected
by cyclic voltammetry measurement of the redox activity of the EAMS.

Cyclic voltammetry was used for the electrochemical characterization of the EAM-
modified targets. From the cyclic voltammogram (CV), binding can be quantified by the
intensity of redox peaks or by the AQ, calculated as the integral of current (Kuznetsov,
1995). The cyclic voltammogram of glycan/HA/anti-HA—EAM complexes in 0.1 M HCl
with the scan range of -0.4 to 1.0 V and rate of 20 mV/s exhibited two stable redox peaks
which are characteristic of the polyaniline conducting polymer, confirming successful
capture of the EAM-captured targets. For the preconcentration preparation method shown
in Figure 25, the anodic peak at 0.07 V corresponds to the switching of leucoemeraldine
base to emeraldine salt, and the peak at 0.76 V indicates the switch from emeraldine to
penrigrarriline salt (Arora et al., 2007; Gospodinova et al., 1996). The cathodic peaks
occurred at -0.13 and 0.33 V. For the stepwise preparation method, the anodic peaks
occurred at 0.09 and 0.77 V, while the corresponding cathodic peaks occurred at -0.14
and 0.37 V. For both methods, the anodic and cathodic peak currents are more defined for
the highest H5 concentration, with the lower H5 concentrations and blanks displaying
statistically similar peaks. The decrease in redox peak intensity with decreasing H5

concentration is expected since less target means lower concentration of glycan/HA/anti-

120

HA—EAM complex present on the electrode. The CV of the blanks, which consisted of
the immunofunctionalized EAMs with no glycan or HA, also displayed the characteristic
redox peaks of the polyaniline, indicating that immunofunctionalization of the EAMs did
not alter their native electrochemical behavior (Figures 24-25). Any EAMs present in this
case would be due to low levels of nonspecific binding or insufficient washing, and the
presence of visible though low intensity redox peaks indicates that the EAMs can
generate redox signals even at very low concentrations. Comparison of the blanks to the
glycan/HA/anti-HA—EAM complexes showed that while the anodic and cathodic peak
potentials did show variation based on HA concentration, the peaks were located within
the same voltage range, indicating that complex formation of the immunofunctionalized

EAMs with the glycan/HA also did not affect electrochemistry.

121

 

 

9)
v

(

 

0.5
0.4
0.3

 

 

 

Delta Q (mC)

 

I——I
bfi

 

 

 

 

 

 

 

I-T-Il

 

 

 

 

 

 

 

 

 

 

 

(b)

 

 

 

 

Current (uA)
o

 

 

 

 

-O.5 0 0.5 1.0

 

 

Potential (V)

Figure 24. Stepwise preparation method. (a) Delta Q values of (A) 3’SLex 100pM +
H5 1.4pM, (B) 3’SLex 100pM + H5 700nM, (C) 3’SLex 100pM + H5 360nM, (D)
CT/Sda 500pM + H5 1.4pM, (E) 3’SLN 500pM + no HA, (F) no glycan + H5 1.4pM,
and (G) no glycan + no HA, and (b) CV of 3’SLex 100pM + H5 1.4pM.

122

 

A
N
v

 

 

 

 

 

 

 

 

 

 

 

 

Delta Q (mC)

 

 

 

 

 

 

 

 

 

 

 

 

 

(b) 40

/\ 3°
20

 

 

 

 

 

Current (uA)
o _x
o

 

 

 

 

 

-o.5 o 0.5 1.0
' Potential (V)

 

 

 

Figure 25. Preconcentration preparation method. (a) Delta Q values of (A) 3’SLex
100pM + H5 l.4p.M + 10% mouse serum, (B) 3’SLex l00uM + H5 700nM + 10%
mouse serum, (C) 3’SLex 100pM + H5 360nM + 10% mouse serum, (D) GT3 100uM
0+ H5 1.4uM + 10% mouse serum, (E) 3’ SLex 100p.M + no HA, (F) no glycan + H5
l.4p.M + 10% mouse serum, and (G) no glycan + no HA, and (b) CV of 3’SLex
100pM + H5 1.4uM + 10% mouse serum.

123

5. 3. 1.3 Biosensor Sensitivity

The biosensor platform showed correlation to the SPR assay results. The sensitivity
of the biosensor platform was explored by testing a range of H5 concentrations. The
preconcentration preparation method yielded an average AQ value of 0.474 mC for the
H5 at 1.4 uM binding to 3’SLex. The lower concentrations of 700 nM and 360 nM
displayed significantly decreased AQ values which were not statistically different from
each other or from the blanks (Figure 26a(H-J); Tables A-3, A-5). The stepwise
preparation method yielded an average AQ value of O. 1 88 mC for the H5 at 1.4 pM,
which was within the range of the preconcentration method blanks and was statistically
lower than the preconcentration value for H5 1.4 uM (Tables A-3, A-S). The lower H5
concentrations of the stepwise method were not statistically different from each other but
also showed significantly lower AQ values than the 1.4 uM stepwise (Figure 26a(A-C);
Tables A-3, A-S). The sensitivity of both preconcentration and stepwise preparation
methods to detect H5 using biosensors prepared with 3’SLex were thus taken to be 1.4
uM (Figures 24-25). We conclude that the preconcentration method, which includes two
magnetic separation and wash steps, is better able to isolate the target HA, thus offering a
consistently higher AQ value than the equivalent concentrations prepared using the
stepwise method.
5.3.1.4 Magnetic Separation by EAMs

The preconcentration HA preparations included 10% mouse serum, which the
stepwise HA did not, but the increased signal for the preconcentration method is not
likely attributable to nonspecific binding due to the mouse serum. It can be observed that

the preconcentration method when performed with the same concentrations of glycan and

124

H5 with and without 10% mouse serum yielded similar AQ values, though still
statistically different (P = 0.0324) (Figure 26b(B,C); Tables A-3, A-S). It is a likely
conclusion then that the magnetic separation technique was able to fully extract the target
HA from the 10% mouse serum matrix to yield a similar signal to that obtained when the
sample was prepared with no serum. This is an improvement on the SPR assay, in which

1% mouse serum depressed the signal as described in 5.1.2.4 and 5. 2. 1.4 (Figure 22).

A
N
V

 

 

 

 

 

 

 

 

    

0.6
0.5
60.4
£03
00.2
901
T).
o
ABCDEFGHIJKLMNOPQR
(b) A 0.6.
00-5
£04
00.3
30.2
(D
00.1
0

 

A B C D E
Figure 26. Cyclic voltammetry results. (a) H5 concentration study as a function of
preparation method and comparison to negative controls and blanks, as numbered
and described in Table A-3. Group (A) 1, (B) 2, (C) 3, (D) 24, (E) 25, (F) 27, (G) 26,
(H) 9, (1) 10, (J) 11, (K) 14. (L) 13, (M) 12. (N) 15, (0) 16,(P) 17, (Q) 18, and (R) 19-
(b) Response for H5 l.4p.M using different preparation methods. (A) l, (B) 8, (C) 9,
(D) 20, and (E) 21. For the respective samples, mean AQ :l: SD, n = 3 (SD = standard
deviation, n = no. of replicates).

125

Table 5. Biotinylated saccharide sequences and predicted binding to H5Nl

 

 

 

 

 

 

Common Saccharide Name and Spacer Predicted
Name to bind H5Nl
3 ’ SLex NeuSAcoc2-3Gall3 l -4[Fucal -3]GlcNAcB-SpNH Yes
3’SLN NeuSAcaZ-BGalB1-4GlcNAcB-SpNH Yes
3 ’ S-Di-LN Neu5Aca2-3 [GalB l —4GlcNacB 1 -3]2B-SpNH Yes
6’SLN Neu5Aca2-6GalB l-4GlcNAcB-SpNH No
GT3 Neu5 Aca2-8Neu5Aca2—8Neu5Acon2-3 Gal B l AGch-SpNH No
6’S-Di-LN Neu5Aca2-6[GalBl-4GlcNacB 1-3]2[3-SpNH No

5.3.1.5 Preparation Effects

The signals generated for the same H5 concentration, 1.4 pM, were compared using
different preparation methods. The preconcentration method, with or without 10% mouse
serum added to the HS, yielded statistically higher AQ values than the stepwise method
(Figure 27b). The stepwise method did confirm that H5 binds to 3’SLN with statistically
higher avidity than it binds to 3’SLex, which is confirmatory to SPR results (Figure
27a,e). However, both of these stepwise values fell far lower than the preconcentration
method values. Source A H5 was also shown to be a better binder to 3’SLex than Source
B H5*. H5*, while the same F LUAV strain as Source A H5, yielded a far lower AQ
value when preconcentrated with 3’SLex than for the 3’SLex/HS (Source A)
preconcentration result (Figure 27c,d). However, 3’SLex /HS* preconcentration did yield
a higher AQ value with statistical significance as compared to the 3’SLex/H5 (Source A)

prepared stepwise (Figure 27a,d; Tables A-3, A-S). We can conclude that the

126

preconcentration method offers a more robust response and that H5 from Source A offers
stronger binding to the 3’SLN and 3’SLex than H5*, possibly due to the predicted

aggregated nature of Source A H5.

 

0.6

 

0.5

 

 

 

 

p—u—g

 

 

0.4

 

 

0.3

 

 

 

 

 

0.2

 

t—u—ua

Delta Q (mC)

 

0.1 _—

 

 

 

 

 

 

 

 

 

0
A B c o E _l

Figure 27. Comparison of different preparation methods. (A) 3’SLex 100uM + H5
1.4uM, stepwise, (B) 3’SLex 100pM + H5 1.4pM, preconcentration, (C) 3’SLex
100uM + H5 1.4uM + 10% mouse serum, preconcentration, (D) 3’SLex 100uM +
H5* 1.4uM + 10% mouse serum, preconcentration, and (E) 3’SLN 100uM + H5
1.4uM, stepwise.

5.3.1.6 Nonspecific Binding

In both preparation methods, the blanks yielded AQ values that were statistically
lower than the reading from the HS-specific glycan/H5 interaction, with H5 at 1.4 pM

and glycans 3’SLN or 3’SLex. For the stepwise preparation method, the presence of H5

127

at 1.4 nM, whether incubated after the nonbinder glycan 6’SLN or after no glycan,
resulted in average AQ values lower with statistical significance than the 3’SLN/HS
response, but higher with statistical significance than the blanks with no H5 added to
either the HS-specific glycan or no glycan (Figure 26a(A, D-G); Tables A-3, A-5). The
absence of H5 yielded repeatable blank tests. The presence of H5 in those blanks which
resulted in higher AQ values than in those blanks without H5 indicates that there may be
low levels of nonspecific binding between H5 and the SPCE surface or any of the
immobilized partners previously incubated on the SPCE. Further blocking could prove
useful to eliminate nonspecific binding.

For the preconcentration method, the blanks both with and without H5 were
repeatable and within a statistically similar range (Figure 26a(K-R); Tables A-3, A-S).
The blanks, including the GT3/HS interaction, were also statistically lower than the
3’SLex/HS interaction. The negative control which included no glycan and no HA but
only the anti-HS—EAM antibody complex yielded the highest AQ value of the blanks,
but this remained below the positive control (Figure 26a(N); Tables A-3, A-S).

The preconcentration method did not include a blocking step, while in the stepwise
method the SPCE surface was blocked with avidin and biotin after incubation with the
biotinylated glycans or, when no glycan was included in the sample, before addition of
HA or anti-HA—EAM complexes. The preconcentration method does not lend itself to
blocking with avidin and biotin, since all of the interaction partners, including glycan,
HA, and anti-HA—EAM are added simultaneously as an already formed complex.
However, the lack of a blocking step does not appear to influence the signal with

nonspecific binding effects. The magnetic separation step serves to eliminate irrelevant

128

material which could interfere with target binding.
5. 3. 1. 7 Biosensor Specificity

The specificity of the system was investigated using a series of H1 samples. In the
preconcentration method, the H1, diluted to 1.4 uM with 10% mouse serum, was
preincubated with the HS-specific glycan 3’SLex and subsequently incubated with EAMs
conjugated with either anti-H5 or anti-H1 antibodies. The samples containing both H1
and anti-Hl—EAM complexes showed an increase in AQ as compared to the samples
with no H1 or with anti-HS—EAM complexes (Figure 26a(O-R)). This may indicate that
the H1 and anti-H1 antibodies interact and cause slightly higher levels of nonspecific
binding as compared to H1 alone or anti-H1 alone. However, the levels of all Hl-based
blanks remain within the statistical range of the HS-based blanks (Tables A-3, A-S). This
indicates that despite the polyclonal nature of the anti-H1 antibodies, there is little cross-
reactivity with the HS-targeted biosensor which improves upon the Biacore system
(Figure 230). Both stepwise and preconcentration methods yielded AQ values for the
binding to the HS-nonspecific glycans, GT3 or 6’SLN, which were distinguishably lower
than their corresponding positive binder, 3’SLex or 3’SLN. We conclude that the

biosensor is highly specific for H5.

129

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I i l I
1 l
— A B C D E

 

Figure 28. Specificity investigation using Hl-based negative controls and
preconcentration preparation. (A) 3’SLex 100uM + H5 1.4uM + 10% mouse serum
+ anti-HS—EAMs, (B) 3’SLex 100pM + H1 1.4uM + 10% mouse serum + anti-H5—
EAMs, (C) 3’SLex lOOuM + H1 1.4uM + 10% mouse serum + anti-Hl—EAMs, (D)
no glycan + H1 1.4uM +10% mouse serum + anti-Hl—EAMs, and (E) no glycan +
no HA + anti-Hl—EAMs.

5.3. 1.8 Structural Morphology Characterization

The EAM polyaniline nanoparticles, EAMs immunofunctionalized with anti-H5
antibody, and glycan/HA/anti-HA—EAM complex were analyzed by a JEOL (Peabody,
MA) IOOCX 11 Transmission Electron Microscope (TEM) to obtain their structural
morphologies. 1% uranyl acetate was used to stain anti-H5 antibody, HA, and glycans.
The crystalline nature of the EAM nanoparticles was also studied by selected area
electron diffraction using the JEOL 2200F S field emission TEM. As shown in Figure

29a, the TEM and electron diffraction micrograph revealed EAM polyaniline

130

nanoparticle sizes in the 25-100 nm range. As observed in the TEM image, the darkest
circular areas correspond to the y-Fe203 cores which are surrounded by the lighter
colored polyaniline polymerized around the cores. Immunoftmctionalization of the EAM
nanoparticles yields a cloudier border as compared to the crisp edge of the EAM
nanoparticles alone, indicating that immunofunctionalization was effective (Figure 29b).
TEM imaging of the 3’SLex/H5/anti-H5—EAM antibody complex after two magnetic
separations and washes resulted in a web-like boundary which could be attributed to the
binding of the H5 and glycan, forming a more branched complex than the EAMs or
immunofimctionalized EAMs alone. When comparing the 3’SLex/I-IS/anti-H5—EAM
antibody complex prepared with H5 with and without 10% mouse serum, the TEM
images reveal similarly shaped aggregates, indicating that there is no nonspecific binding
of the serum components to the complex (Figure 29c,d). This is in confirmation of the
cyclic voltammetry results (Figure 26b(B,C)). The backgrounds of the images do reveal
that the sample prepared with mouse serum has a cloudier supernatant, suggesting the

benefit of a more thorough washing, although the AQ values are not affected.

131

 

Figure 29. TEM imaging. (a) TEM and electron diffraction micrograph (inset) of
EAM polyaniline nanoparticles with gamma iron (III) oxide cores, (b) TEM of
EAMs immunofunctionalized with anti-H5 antibody, (c) 3’SLex [HS/anti-HS—EAM
complex, magnetically separated and washed, with H5 prepared with 10% mouse
serum, and (d) 3’SLex /H5/anti-H5—EAM complex, magnetically separated and
washed, with H5 prepared without serum.

132

5.4 OBJECTIVE 4
Biosensor to Distinguish (12,3 v. (12,6 Receptor Binding
5.4.1 Biosensor Desig
5. 4. 1.1 Glycan Sequences

The preconcentration method was also utilized to compare a (12,3 versus 0L2,6
linked glycan receptors. 3’S-Di-LN and 6’S-Di-LN were chosen for comparison purposes
as their saccharide sequences were identical except for the sialic acid linkage, ensuring
that any differences in binding would be the result of this linkage. 3’S-Di-LN was
predicted to bind H5 due to the a2,3 preference of avian F LUAV, and 6’S-Di-LN was
predicted to bind H1 due to the a2,6 preference of human F LUAV.
5.4.1.2 Avian FL UA V- Targeted Biosensor

The H5-specific glycan 3’S-Di-LN at 100 uM bound H5 at 1.4 uM and 700 nM,
prepared to contain 10% mouse serum. The glycan/HS complex was
immunomagnetically separated from the mouse serum and other extraneous unbound
material using EAMs immunofunctionalized with anti-H5 monoclonal antibody. The AQ
values of 3’S-Di-LN binding to H5 at 1.4 uM and 700 nM were not statistically different

from each other, but were statistically higher than the AQ values of H5 at 1.4 uM binding

to the H5-nonspecific glycan 6’S-Di-LN at 100 nM.

133

 

5.4.1.3 Human FL UA V- Targeted Biosensor

The glycan with predicted binding to human HA, 6’S-Di-LN, was shown by the
preconcentration method to specifically bind H1 (HlNl A/ South Carolina/l/18) while
not binding H5. The (12,6 glycan, 6’S-Di-LN, at 100 nM, bound H1 at 1.4 uM and 700
nM, prepared to contain 10% mouse serum. The glycan/H5 complex was
immunomagnetically separated from the mouse serum and other extraneous unbound
material using EAMs immunofunctionalized with anti-H1 monoclonal antibody. The AQ
values of 6’S-Di-LN binding to H1 at 1.4 uM and 700 nM were not statistically different
from each other, but were statistically higher than the AQ values of H1 at 1.4 uM binding
to the HS-specific glycan 3’S-Di-LN at 100 uM.
5. 4. 1.4 Specificity

The AQ value of 6’S-Di-LN/HS was within the statistical range of the negative
control, in which only anti-HS—EAMS with no glycan or HA were incubated on the
SPCE. Similarly, the AQ value of 3’S-Di-LN/Hl was within the statistical range of the
negative control, in which only anti-Hl—EAMS with no glycan or HA were incubated on
the SPCE. From these results, we can conclude that the biosensor is able to distinguish
a2,3 versus a2,6 sialic acid linkages with repeatability. Because human transmissibility
and thus pandemic potential rely on a2,6 specificity of the FLUAV strain, these results
offer proof of concept that the biosensor is able to identify pandemic strains, and to

distinguish them from nonpandemic strains.

134

 

 

 

 

 

 

F Delta 0 (mC) T

 

    

    

 

 

 

A B

g E

c. 3

E E

E E

a J

O O

-0.5 __ 0_ 0.5 _ -8 -0.5 __ 0_ 0.5

l 1 I _

,_..__— Potentia|(V) : Potentia|(V) E
C D

E E

E E

:1 :

O O

-o.5 — o— 0.5 — '8 -o.5 — o——— 0.5 —

Potential (V) Potential (V)

Figure 30. H5 binding to a2,3 versus a2,6-linked glycan receptors using

-8

preconcentration method. (A) 3’S-Di-LN 100pM + H5 1.4uM + 10% mouse serum,

(B) 3’S-Di-LN 100p.M + H5 700nM + 10% mouse serum, (C) 6’S-Di-LN 100uM +
H5 1.4uM +10% mouse serum, and (D) aHS—EAMs only.

135

Delta Q (mC) l

 

 

 

 

 

 
  
 
    

 
      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A . 10 B 1 10
o 0.5 _ 1 a o 0.5 —— 18
| . .
’- 4 ’- 4
<5 <
c. 2 3 2
E o E o
E E
5 —4 5 —4
0 Potential (V) _ -3 0 Potential (V) -3
l |
10 10
C o 05 1 D o 05 1
8 8
6 6
4 kvAqu“ 4
<5 2 a “(We 2
I: o 3: ' \ o
C E
g 4 g 4
3 —— _8 3 — — _8
0 Potential(\/) _ .. 0 _ Potential (V)

 

 

Figure 31. H1 binding to a2,3 versus a2,6—linked glycan receptors using
preconcentration method. (a) 6’S-Di-LN 100uM + H1 1.4uM + 10% mouse serum,
(b) 6’S-Di-LN 100uM + H1 700nM + 10% mouse serum, (c) 3’S-Di-LN 100uM + H1
1.4uM + 10% mouse serum, and (d) aHl—EAMs only.

136

CHAPTER 6: CONCLUSION AND FUTURE RESEARCH

In this dissertation, SPR and biosensor assays were explored as tools for pandemic
FLUAV detection. A novel SPR assay was designed, which utilized H5-specific sialic
acid receptors to specifically and sensitively identify H5 HA protein. The sensitivity of
the HS-targeted assay in the detection of recombinant H5 hemagglutinin (H5Nl
AN ietnam/ 1203/04) was found to be 10.6 nM in buffer and 31.4 nM in 2% mouse
serum. The SPR assay demonstrated high avidity of binding between H5 and a2,3-linked
glycans 3’SLex and 3’SLN with statistically lower binding between H5 and a2,6-linked
6’SLN, a2,8-linked GD2, and 012,3-linked CT/Sda, which is confirmatory to expected
results and demonstrates that the SPR assay can characterize HA by sialic acid receptor
preference. The biosensor showed high specificity for H5 as compared to H1 (I-llNl
A/South Carolina/ 1/ 18) and H3 (H3N2 A/Wyoming/3/03).

Our results indicate that the SPR assay could identify plasma with high neutralizing
activity, as inhibitors of glycan/HA binding, for facilitating high potency FLUIGIV
manufacture. The SPR neutralization assay has shown a range of anti-H5 monoclonal
antibodies from 1:250 to 1:4000 to be neutralizing against glycan/H5 binding. These
highly diluted samples ensure physiologically relevant inclusion, as a 1:80 convalescent
plasma dilution has previously been shown in patient testing to have a neutralizing
antibody titer (Zhou et al., 2007). The assay may facilitate large-scale F LUIGIV
screening to reliably identify plasma that is highly neutralizing against pandemic avian
influenza viruses with 012,3 specificity.

From these results, we expect that the SPR-based assay can be easily modified to

similarly detect HAS with 012,6 specificity, an indicator of human pandemic potential.

137

 

The SPR assay may serve as the first line of identification of a historically avian 012,3-
specific FLUAV that antigenically shifts to become 012,6 specific and thus transmissible
from human-to-human.

Future work will include the development of a pandemic HlNl targeted assay, in
which H1 HA will be identified by Hl-specific sialic acids. Screening for HlNl-specific
FLUIGIV would produce passive therapies that could be useful in the face of a vaccine
shortage as seen with the 2009 novel HlNl pandemic. Clinical samples of plasma or
nasal fluid from animal or human subjects would increase complexity of the system, and

the mouse serum matrix tested here is only intended to offer a first step towards a more

 

complex system. Because clinical experimentation indicates that highly diluted samples
offer neutralizing activity, any limitations of the SPR system as a plasma screening assay
due to matrix interference may be reduced when testing high dilutions. Identification of
multimeric recombinant HA, pseudovirus particles, or whole virus in complex matrices
such as serum or respiratory secretions is another long-term goal. The development of
such an assay which identifies FLUAV HA based on specificity to host sialic acids is a
significant initiative with applications in surveillance, serodiagnosis, and homeland
security.

Further work is required to optimize the SPR assay in terms of sensitivity to detect
HA at concentrations reflecting the viral load in an influenza infected patient.
Preconcentration of target analyte is a viable option. Nonspecific binding can be further
reduced by improving blocking techniques. The sensitivity, specificity, and repeatability
of our novel method are promising. The SPR assay design is easily adaptable to detection

of other FLUAV strains, including the current swine-origin HlNl. The Biacore SPR

138

assay is an appropriate technique for understanding the specificity and avidity of
glycan/HA partners and for probing cross-clade protection of anti-HA antibodies, and
ultimately could find applicability as a screening assay for highly-neutralizing plasma.

An electrochemical biosensor was developed which utilized electrically active
polyaniline coated magnetic nanoparticles (EAMs) both as a magnetic separator and a
biosensor transducer. .

The sensitivity of the biosensor prepared with 3’SLex or 3’SLN in the detection of
recombinant H5 hemagglutinin(H5N1 AN ietnarn/ 1203/04) was found to be 1.4 uM in
10% mouse serum. The sensitivity of the biosensor prepared with 3’S-Di-LN in the
detection of recombinant H5 was found to be 700 nM in 10% mouse serum. The
biosensor sensitivity for H1 hemagglutinin(H1Nl A/ South Carolina! 1/1 8) as prepared
with 6’S-Di-LN was found to be 700 nM. The biosensor demonstrates high avidity of
binding between H5 and 012,3-linked glycans 3’SLex, 3’SLN, and 3’S-Di-LN, with
statistically lower binding between H5 and 012,6-linked 6’SLN and 6’S-Di-LN, and 012,8-
linked GT3, which is confirmatory to expected results and demonstrates that the
biosensor can characterize HA by sialic acid receptor preference. The avian FLUAV
targeted biosensor showed high specificity for H5 as compared to H1 and the human
FLUAV targeted biosensor also offered proof of concept for H1 binding to 012,6 sialic
acids, indicating that the biosensor is easily adaptable to an 012,6 targeted biosensor using
appropriate 012,6 linked sialic acid receptors. The biosensor architecture and fabrication
and testing techniques are easily amenable to the detection of any FLUAV HA subtype.

The biosensor system is rapid to results, with signal detection time at 8 minutes or

less. The five-hour SPCE preparation may be performed offline, with SPCE storage for

139

 

months prior to testing. Using the preconcentration method, the entire sample preparation
time requires 75 minutes, including complex incubation, magnetic separations, washes,
and SPCE application.

The biosensor technology is thus able to repeatably distinguish 012,3-specific
FLUAV strains from 012,6 specific strains, and could thus offer frontline detection of an
emerging human-transmissible strain. Once a highly pathogenic F LUAV strain achieves
human transmissibility via antigenic shifting, the strain could cause a human pandemic,
and a point-of-care biosensor such as that proposed here, could be used at hospitals,
doctors’ offices, or borders as the first line in detection, prophylaxis, and mobilizing of
treatments. The biosensor technology offers quick and reliable identification of the
receptor preference of a F LUAV strain, which is the key characteristic involved in host
range and pandemic potential.

This research shows the ability of the EAMs to immunomagnetically separate target
HA from serum matrix. This capacity will be exploited in future applications in which
whole or pseudotyped virus will be identified in complex matrices such as serum or
respiratory secretions. The large size of a whole or pseudotyped virus in comparison to
the recombinant HA proteins tested here may introduce stearic hindrance effects. The
sandwich biosensor assay is attractive in that while the whole virus may be large, there
will be many HA proteins covering the surface, allowing a similarly high number of
irnmunofuntionalized EAMs to cover the surface of any captured virions, and thus
leading to no signal loss. This is in comparison to the direct label-free SPR assay, in
which stearic hindrance effects could lead to signal loss if fewer virions are captured to

the immobilized glycans.

140

 

The results indicate that the biosensor technology is valuable as a rapid, specific,
and sensitive detection method with applicability at point-of-care for identifying highly
pathogenic avian influenza viruses with 012,3 specificity or for identifying human
influenza viruses with 012,6 specificity, and for differentiating between pandemic and
nonpandemic strains. This is important from an agricultural as well as a biosecurity
standpoint. The development of such a biosensor technology which identifies FLUAV

HA based on specificity to host sialic acids is a significant initiative with applications in

 

disease monitoring and homeland security.
In summary, this research demonstrates the applicability of both SPR and biosensor
platforms for the detection of FLUAV HA using strain-specific glycan receptors, for

purposes of prophylaxis, treatment, and early detection.

141

A.1

APPENDIX A: STATISTICAL ANALYSIS RESULTS

AN OVA ANALYSIS OF SPR RESULTS

Table A-1. SPR Results for Different HA-Glycan Pairs: Least Square Means

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Peak
Group Description Estimate Standard DF t Value Pr > |t|
(RU) Error
1 H5 0.25ug/ml 3'Slex 4.730 1.1847 2 3.99 0.0001
2 H5 0.74ug/ml 3'Slex 19.707 1.8258 2 10.79 <0.0001
3 H5 0.74ug/ml 3'Slex Chip 3 24.410 0.3151 2 77.47 <0.0001
4 H5 2.2ug/ml 3'Slex 59.036 3.3999 2 17.36 <0.0001
5 H5 2.2ug/ml 3'Slex Chip 3 82.855 3.4933 2 23.72 <0.0001
6 H5 6.6ug/ml 3'Slex 145.210 25.1075 2 5.78 <0.0001
7 H5 6.6ug/ml 3'Slex Chip 3 228.900 2.5977 2 88.12 <0.0001
8 H5 V BEI 10ug/ml 3'Slex 397.040 15.9780 2 24.85 <0.0001
9 H5 20ug/ml 3'Slex 745.060 27.1531 2 27.44 <0.0001
10 H5 0.25ug/ml 3'SLN 10.309 1.2086 2 8.53 <0.0001
11 H5 0.74ug/ml 3'SLN 39.811 2.9209 2 13.63 <0.0001
12 H5 2.2ug/ml 3'SLN 131.520 5.3201 2 24.72 <0.0001
13 H5 6.6ug/ml 3'SLN 2 361.500 23.1912 2 15.59 <0.0001
14 H5 10ug only 3'SLN 897.170 13.9810 2 64.17 <0.0001
15 H5 20ug/ml 3'SLN 1493.760 32.0269 2 46.64 <0.0001
16 H1 2.2ug/ml 3'SLN -4.159 0.5703 2 -7.29 <0.0001
17 H1 6.6ug/ml 3'SLN -7.628 1.7263 2 -4.42 <0.0001
18 H16.6+aHl 1:500 3'SLN -3.276 0.7730 2 -4.24 <0.0001
19 H16.6+aH1 1:1000 3'SLN -4.443 0.8485 2 -5.24 <0.0001
20 anti-H1 1:500 3'SLN -2.698 16.6168 2 -0.16 0.8712
21 H16.6+aH5 1:500 3'SLN -0.127 0.3061 2 -0.42 0.6786
22 H16.6+aH5 1:1000 3'SLN 2.050 0.6675 2 3.07 0.0025
23 anti-H5 1:500 3'SLN 1.857 16.6168 2 0.11 0.9112
24 H56.6+antil :250 3'SLN 41.658 4.6252 2 9.01 <0.0001
25 H56.6+antil :500 3'SLN 46.624 5.1730 2 9.01 <0.0001
26 H56.6+anti1:1000 3'SLN 14.581 0.4621 2 31.55 <0.0001
27 HA1 H5 2.2ug/ml 3'SLN -8.397 0.6350 2 -13.22 <0.0001
28 HA] H5 2.2ug/ml 3'Slex 0.156 16.6168 2 0.01 0.9925
29 HA1 H5 6.6ug/ml 3'SLN -9.242 1.7890 2 -5.17 <0.0001
30 HA1 H5 6.6ug/ml 3'Slex 1.286 16.6168 2 0.08 0.9384
31 HA1 H5 20ug/ml 3'SLN -8.921 0.2980 2 -29.94 <0.0001
32 HA1 H5 20ug/ml 3'Slex 1.094 16.6168 2 0.07 0.9476
33 HA1 H3 2.2ug/ml 3'SLN -9.093 0.5135 2 -17.71 <0.0001

 

 

142

 

 

 

 

 

Table A-1. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

34 HA1 H3 2.2ug/ml 3'Slex 5.872 16.6168 2 0.35 0.7243
35 HA1 H3 6.6ug/ml 3'SLN -9.115 0.0545 2 -167.24 <0.0001
36 HA1 H3 6.6ug/ml 3'Slex 15.639 16.6168 2 0.94 0.3482
37 HA1 H3 20ug/ml 3'SLN -8.384 0.7060 2 -11.88 <0.0001
38 HA1 H3 20ug/ml 3'Slex 30.125 16.6168 2 1.81 0.0719
39 D5 0% -4.008 3.9640 2 -l.01 0.3136
40 D4 0.25% -3.870 3.5615 2 -l.09 0.279
41 D3 0.5% -2.911 5.4010 2 -0.54 0.5907
42 D2 0.75% 4.780 4.2085 2 1.14 0.2579
43 D1 1% 8.779 2.0315 2 4.32 <0.0001
44 HA1 H5 2.2ug/ml - D3 -0.722 1.3285 2 -0.54 0.5879
45 HA1 H5 6.6ug/ml - D3 -2.171 1.5290 2 -1.42 0.1577
46 HA1 H5 20ug/ml - D3 -2.980 0.8140 2 -3.66 0.0004
47 HA1 H3 2.2ug/ml - D3 -1.663 0.4065 2 -4.09 <0.0001
48 HA1 H3 6.6ug/ml - D3 -1.914 0.2190 2 -8.74 <0.0001
49 HA1 H3 20ug/ml - D3 -1.285 0.6345 2 -2.02 0.0447
50 BEI H5 10ug/ml - D3 77.953 16.6168 2 4.69 <0.0001
51 H510ugH5mAbl :250 3'Slex 21.825 4.1990 2 5.20 <0.0001
52 H510ugH5mAb1 :500 3'Slex 23.629 3.1844 2 7.42 <0.0001
53 H510ugH5mAb1:1k 3'Slex 26.229 4.6675 ' 2 5.62 <0.0001
54 H510ugH5mAb1 :2k 3'Slex 120.000 15.4232 2 7.78 <0.0001
55 H510ugH5mAbl :4k 3'Slex 352.290 40.1883 2 8.77 <0.0001
56 H510ugH5mAb1 :8k 3'Slex 546.050 16.6620 2 32.77 <0.0001
57 antiHSmAb 1:250 3'Slex 15.695 2.8137 2 5.58 <0.0001
58 H510ug+1%s 3'Slex 77.637 16.6168 2 4.67 <0.0001
59 H510mAb1 :250+1%s 3'Slex 14.951 16.6168 2 0.90 0.3697
60 H510mAb1 :500+1%s 3'Slex 17.416 16.6168 2 1.05 0.2963
61 H510mAbl:1k+l%s 3'Slex 18.973 16.6168 2 1.14 0.2554
62 H510mAbl :2k+1% 3'Slex 37.571 16.6168 2 2.26 0.0252
63 H510mAb1 :4k+1% 3'Slex 54.793 16.6168 2 3.30 0.0012
64 antiH51:250+1% 3'Slex 7.813 16.6168 2 0.47 0.6389
65 H510ugH5mAbl :250 3'SLN 58.851 5.2785 2 11.15 <0.0001
66 H510ugH5mAb1 :500 3’SLN 75.372 6.6805 2 11.28 <0.0001
67 H510ugH5mAblzlk 3'SLN 75.357 23.7535 2 3.17 0.0018
68 H510ugH5mAbl :2k 3'SLN 358.640 9.4680 2 37.88 <0.0001
69 H510ugH5mAbl :4k 3'SLN 824.760 15.7440 2 52.39 <0.0001
70 H510ugH5mAbl :8k 3'SLN 1129.820 10.9750 2 102.94 <0.0001
71 antiHSmAb 1:250 3'SLN 46.071 5.1215 2 9.00 <0.0001
72 H510ug+1%s 3'SLN 334.670 16.6168 2 20.14 <0.0001

143

 

 

Table A-1. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

73 H510mAb1:250+1%s3'SLN 47.372 16.6168 2 2.85 0.005
74 H510mAb1:500+1%s 3'SLN 55.637 16.6168 2 3.35 0.001
75 H510mAbl:1k+l%s3'SLN 50.081 16.6168 2 3.01 0.003
76 H510mAb1:2k+1% 3'SLN 148.240 16.6168 2 8.92 <0.0001
77 H510mAbl:4k+1% 3'SLN 229.790 16.6168 2 13.83 <0.0001
78 antiH51:250+1% 3'SLN 21.135 16.6168 2 1.27 0.2054
79 H5VBEIlO+aIl:250 3'Slex 283.670 16.6168 2 17.07 <0.0001
80 H5VBEIlO+aH1250 3'Slex 213.410 16.6168 2 12.84 <0.0001
81 HSVIT 10ug/ml 3'Slex .3759 16.6168 2 -0.23 0.8213
82 H5VIT10+aV12250 3'Slex 322.860 16.6168 2 19.43 <0.0001
83 H5VIT10+aV1z500 3'Slex 13.606 16.6168 2 0.82 0.4142
84 H5VIT10+aVl:1000 3'Slex 5.772 16.6168 2 0.35 0.7288
85 H51ndoIT10ug/ml3'Slex 3.032 16.6168 2 0.18 0.8555
86 H511T10+al1:2503'S1ex 9.693 16.6168 2 0.58 0.5606
87 H511T10+a11:5003'Slex 9.933 16.6168 2 0.60 0.5509
88 H511T10+a11210003'Slex 6.612 16.6168 2 0.40 0.6913
89 a-Indolz250 3'Slex -0.893 16.6168 2 -0.05 0.9572
90 H511r10+av1:250 3'Slex 107.880 16.6168 2 6.49 <0.0001
91 H511T10+aH11z250 3'Slex 27.783 16.6168 2 1.67 0.0966
92 H1SC10ug/ml3'Slex -2407 16.6168 2 -014 0.885
93 H1SC10+a-H11:250 3'Slex 10.842 16.6168 2 0.65 0.5151
94 HISC10+a-H11:5003'Slex 2.822 16.6168 2 0.17 0.8654
95 HISC10+a-H11000 3'Slex -1.585 16.6168 2 -0.10 0.9241
96 a-HlPanlz250 3'Slex 8.872 16.6168 2 0.53 0.5942
97 HlSClO+a-V1:250 3'Slex 47.264 16.6168 2 2.84 0.0051
98 H1SC10+a-Il:250 3'Slex -0973 16.6168 2 -0.06 0.9534
99 H5VBEIIO+aIlz2503'SLN 282.010 16.6168 2 16.97 <0.0001
100 H5VBEIlO+aH1250 3'SLN 211.990 16.6168 2 12.76 <0.0001
101 H5VIT10ug/ml3'SLN -5.064 16.6168 2 -0.30 0.761
102 H5VIT10+aV1z250 3'SLN 321.350 16.6168 2 19.34 <0.0001
103 H5VIT10+aV1:500 3'SLN 12.589 16.6168 2 0.76 0.4499
104 H5VIT10+aV1z1000 3'SLN 4.572 16.6168 2 0.28 0.7836
105 H51ndoIT10ug/ml3‘SLN 1.951 16.6168 2 0.12 0.9067
106 H511T10+a11z250 3'SLN 8.287 16.6168 2 0.50 0.6187
107 H511T10+a11z500 3'SLN 8.820 16.6168 2 0.53 0.5964
108 H511T10+aIlz1000 3'SLN 5.219 16.6168 2 0.31 0.7539
109 a-Indol:250 3'SLN -2001 16.6168 2 -0.12 0.9043
110 HSIIT10+aVlz250 3'SLN 106.560 16.6168 2 6.41 <0.0001
111 HSIIT10+aHllz250 3'SLN 26.665 16.6168 2 1.60 0.1107
112 H1SC10ug/m13'SLN 3413 16.6168 2 -0.21 0.8375
113 H1SC10+a-H11:2503'SLN 9.177 16.6168 2 0.55 0.5816
114 HISC10+a-H11:5003'SLN 1.816 16.6168 2 0.11 0.9131

 

 

144

 

 

 

 

 

Table A-1. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

115 HlSC10+a-H11000 3'SLN -2.806 16.6168 2 -0.17 0.8661
116 a-Hl Pan 1:250 3'SLN 7.629 16.6168 2 0.46 0.6468
117 H1SC10+a-Vl :250 3'SLN 45.888 16.6168 2 2.76 0.0065
118 HISC10+a-Il :250 3'SLN -2.140 16.6168 2 -0.13 0.8977
119 H5 6.6ug + 1%s 3'Slex 65.617 16.6168 2 3.95 0.0001
120 H5 20ug + 1%s 3'Slex 443.440 16.6168 2 26.69 <0.0001
121 H5 6.6ug + 1%s 3'SLN 238.340 16.6168 2 14.34 <0.0001
122 H5 20ug + 1%s 3'SLN 999.190 16.6168 2 60.13 <0.0001
123 H3 2.2ug 3'Slex -3.137 16.6168 2 -0.19 0.8505
124 H3 6.6ug 3'Slex -2.666 16.6168 2 -0.16 0.8728
125 H3 11.9ug 3'Slex -3.233 16.6168 2 -0.19 0.846
126 H5 0.24ug + 2%s 3'Slex -5.936 16.6168 2 -0.36 0.7214
127 H5 0.73ug + 2%s 3'Slex -6.668 16.6168 2 -0.40 0.6888
128 H5 2.2ug + 2%s 3'Slex -7.298 16.6168 2 -0.44 0.6612
129 H5 6.6ug + 2%s 3'Slex -4.873 16.6168 2 -0.29 0.7697
130 H5 20ug + 2%s 3'Slex 14.021 16.6168 2 0.84 0.4002
131 H3 0.24ug + 2%s 3'Slex 26.359 16.6168 2 1.59 0.1148
132 H3 0.73ug + 2%s 3'Slex -1.701 16.6168 2 -0.10 0.9186
133 H3 2.2ug + 2%s 3'Slex 26.717 16.6168 2 1.61 0.11
134 H3 6.6ug + 2%s 3'Slex -1.709 16.6168 2 -0.10 0.9182
135 H3 11.9ug + 2%s 3'Slex 10.403 16.6168 2 0.63 0.5322
136 H3 2.2ug 3'SLN -5.191 16.6168 2 -0.31 0.7552
137 H3 6.6ug 3'SLN -5.334 16.6168 2 -0.32 0.7487
138 H3 11.9ug 3'SLN -5.593 16.6168 2 -0.34 0.7369
139 H5 0.24ug + 2%s 3'SLN 3.197 16.6168 2 0.19 0.8477
140 H5 0.73ug + 2%s 3'SLN 2.306 16.6168 2 0.14 0.8898
141 H5 2.2ug + 2%s 3'SLN 7.072 16.6168 2 0.43 0.671
142 H5 6.6ug + 2%s 3'SLN 13.290 16.6168 2 0.80 0.4251
143 H5 20ug + 2%s 3'SLN 28.034 16.6168 2 1.69 0.0937
144 H3 0.24ug + 2%s 3'SLN 41.077 16.6168 2 2.47 0.0146
145 H3 0.73ug + 2%s 3'SLN -4.963 16.6168 2 -0.30 0.7656
146 H3 2.2ug + 2%s 3'SLN 41.955 16.6168 2 2.52 0.0126
147 H3 6.6ug + 2%s 3'SLN -4.804 16.6168 2 -0.29 0.7729
148 H3 11.9ug + 2%s 3'SLN 3.519 16.6168 2 0.21 0.8326
149 H5 HBSEP 4hr 37 3'Slex 76.432 16.6168 2 4.60 <0.0001
150 H5 Brom 4hr 37 3'Slex 10.813 16.6168 2 0.65 0.5162
15] H5 Brom2ME 4hr37 3'Slex -6.798 16.6168 2 -0.41 0.6831
152 H5 HBSEP o/n 37 3'Slex 309.130 16.6168 2 18.60 <0.0001
153 H5 Brom o/n 37 3'Slex 4.153 16.6168 2 0.25 0.803
154 H5 Brom2ME o/n37 3’Slex -3.114 16.6168 2 -0.19 0.8516
155 H5 0.02%tween 3'Slex 578.480 16.6168 2 34.81 <0.0001

 

 

145

 

 

 

 

 

 

Table A-1. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

156 H5 0.05%tween 3'Slex 398.980 16.6168 2 24.01 <0.0001

157 H5 0.08%tween 3'Slex 351.030 16.6168 2 21.13 <0.0001

158 H5 0.1%tween 3'Slex 230.320 16.6168 2 13.86 <0.0001

159 H5 10min56 +4deg 3'Slex 0.514 16.6168 2 0.03 0.9754

160 H5 10min 56 only 3'Slex 4.732 16.6168 2 0.28 0.7762

161 H5 HBSEP 4hr 37 3'SLN 121.000 16.6168 2 7.28 <0.0001

162 H5 Brom 4hr 37 3'SLN -21.223 16.6168 2 -1.28 0.2035

163 H5 Brom2ME 4hr37 3'SLN -62.782 16.6168 2 -3.78 0.0002

164 H5 HBSEP o/n 37 3'SLN 548.840 16.6168 2 33.03 <0.0001

165 H5 Brom o/n 37 3'SLN -2.915 16.6168 2 -0.18 0.861

166 H5 Brom2ME o/n37 3'SLN -19.331 16.6168 2 -1.16 0.2466

167 H5 0.02%tween 3'SLN 959.430 16.6168 2 57.74 <0.0001

168 H5 0.05%tween 3'SLN 754.050 16.6168 2 45.38 <0.0001

169 H5 0.08%tween 3'SLN 668.130 16.6168 2 40.21 <0.0001

170 H5 0.1%tween 3'SLN 464.360 16.6168 2 27.95 <0.0001

171 H5 10min56 +4deg 3‘SLN -0.257 16.6168 2 -0.02 0.9877

172 H5 10min 56 only 3'SLN 4.547 16.6168 2 0.27 0.7847

173 HA5 10ug 3'Slex 102.180 16.6168 2 6.15 <0.0001
HA5 10ug + anti-HA2 1:250

174 3'Slex 107.140 16.6168 2 6.45 <0.0001
HA5 10ug + anti-HA2 1:500

175 3'Slex 101.050 16.6168 2 6.08 <0.0001
HA5 10ug + anti-HA2

176 1:1000 3'Slex 98.152 16.6168 2 5.91 <0.0001
HA5 10ug + anti-HA2 '

177 1:2000 3'Slex 98.072 16.6168 2 5.90 <0.0001

178 anti-HA2 1:250 3'Slex 0.878 16.6168 2 0.05 0.9579
HA5 10ug + anti-H5 mAb

179 1:1000 3'Slex 185.930 16.6168 2 11.19 <0.0001

180 HA5 10ug 3'SLN 347.810 16.6168 2 20.93 <0.0001
HA5 10ug + anti-HA2 1:250

181 3'SLN 350.130 16.6168 2 21.07 <0.0001
HA5 10ug + anti-HA2 1:500

182 3'SLN 334.310 16.6168 2 20.12 <0.0001
HA5 10ug + anti-HA2

183 1:1000 3'SLN 331.490 16.6168 2 19.95 <0.0001
HA5 10ug + anti-HA2

184 1:2000 3’SLN 332.750 16.6168 2 20.02 <0.0001

185 anti-HA2 1:250 3'SLN 0.834 16.6168 2 0.05 0.96
HA5 10ug + anti-H5 mAb

186 1:1000 3'SLN 640.140 16.6168 2 38.52 <0.0001

187 buffer -1.827 0.4462 2 -4.10 <0.0001

 

 

146

 

 

 

 

 

 

Table A-1. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

buffer different regen

188 7mMNaOH 2.5MNaC1 -4.892 0.2458 2 -19.90 <0.0001
buffer 1%mouse serum

189 3'Slex 6.327 0.2388 2 26.50 <0.0001
buffer 1%mouse serum

190 3'SLN 0.412 0.2103 2 1.96 0.0523

191 buffer 10%serum 3'Slex 19.465 2.8250 2 6.89 <0.0001

192 buffer 5%serum 3'Slex 29.515 0.4850 2 60.86 <0.0001

193 buffer 2%serum 3'Slex 15.315 0.6550 2 23.38 <.0001

194 buffer 1%serum 3'Slex 5.865 0.4550 2 12.89 <.0001

195 buffer 10%serum 3'SLN 37.400 5.8000 2 6.45 <.0001

196 buffer 5%serum 3'SLN 35.050 0.5500 2 63.73 <.0001

197 buffer 2%serum 3'SLN 13.750 0.3500 2 39.29 <.0001

198 buffer 1%serum 3'SLN 3.130 0.0500 2 62.60 <.0001
buffer 10%serum 3'Slex after

199 3 min dissoc 1.920 0.7400 2 2.59 0.0104
buffer 5%serum 3'Slex after

200 3 min dissoc —1.009 0.0715 2 -14.10 <.0001
buffer 2%serum 3'Slex after

201 3 min dissoc -1.018 0.6445 2 -1.58 0.1165
buffer 1%serum 3'Slex after

202 3 min dissoc -0.993 0.4830 2 -2.06 0.0415
buffer 10%serum 3'SLN

203 after 3 min dissoc 10.095 3.2050 2 3.15 0.002
buffer 5%serum 3'SLN after

204 3 min dissoc 0.706 0.0730 2 9.67 <.0001
buffer 2%serum 3'SLN afier

205 3 min dissoc -0.769 0.4310 2 -1.78 0.0764
buffer 1%serum 3'SLN after

206 3 min dissoc -0.839 0.4110 2 -2.04 0.043

 

 

 

 

 

 

 

147

 

 

 

Table A-2. SPR Group Comparisons: Estimates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Label Peak Estimate Standard DF t Value Pr > |t|
(RU) Error

Group 1 vs. 2 -14.977 2.1764 2 -6.88 <0.0001
Group 1 vs. 4 -54.306 3.6004 2 -15.08 <0.0001
Group 1 vs. 6 -140.480 25.1354 2 -5.59 <0.0001
Group 1 vs. 8 -392.310 16.0218 2 -24.49 <0.0001
Group 1 vs. 9 -740.330 27.1790 2 -27.24 <0.0001
Group 2 vs. 4 -39.330 3.8591 2 ~10.19 <0.0001
Group 2 vs. 6 -125.500 25.1738 2 -4.99 <0.0001
Group 2 vs. 8 -377.330 16.0820 2 -23.46 <0.0001
Group 2 vs. 9 -725.360 27.2144 2 -26.65 <0.0001
Group 4 vs. 6 -86.170 25.3366 2 -3.40 0.0009
Group 4 vs. 8 -338.000 16.3357 2 -20.69 <0.0001
Group 4 vs. 9 -686.030 27.3651 2 -25.07 <0.0001
Group 6 vs. 8 -251.830 29.7604 2 -8.46 <0.0001
Group 6 vs. 9 -599.860 36.9821 2 -16.22 <0.0001
Group 8 vs. 9 -348.020 31.5054 2 -11.05 <0.0001
Group 2 vs. 3 -4.703 1.8528 2 -2.54 0.0122
Group 4 vs. 5 -23.818 4.8746 2 -4.89 <0.0001
Group 6 vs. 7 -83.691 25.2415 2 -3.32 0.0012
Group 10 vs. 11 -29.502 3.1611 2 -9.33 <0.0001
Group 10 vs. 12 -121.210 5.4557 2 -22.22 <0.0001
Group 10 vs. 13 -351.190 23.2227 2 -15.12 <0.0001
Grog) 10 vs. 14 -886.870 14.0332 2 -63.20 <0.0001
Group 10 vs. 15 -1483.450 32.0497 2 -46.29 <0.0001
Group 11 vs. 12 -91.705 6.0692 2 -15.11 <0.0001
Group 11 vs. 13 -321.690 23.3744 2 -13.76 <0.0001
Group 11 vs. 14 -857.360 14.2829 2 -60.03 <0.0001
Group 11 vs. 15 -1453.950 32.1598 2 —45.21 <0.0001
Group 12 vs. 13 -229.980 23.7936 2 -9.67 <0.0001
Group 12 vs. 14 -765.660 14.9590 2 -51.18 <0.0001
Group 12 vs. 15 -1362.250 32.4657 2 41.96 <0.0001
Group 13 vs. 14 -535.670 27.0796 2 -19.78 <0.0001
Group 13 vs. 15 -1132.260 39.5418 2 -28.63 <0.0001
Group 14 vs. 15 -596.590 34.9455 2 -17.07 <0.0001
Group 1 vs. 10 -5.579 1.6924 2 -3.30 0.0012
Group 2 vs. 11 -20.104 3.4446 2 -5.84 <0.0001
Grog) 4 vs. 12 -72.479 6.3137 2 -11.48 <0.0001
Group 6 vs. 13 -216.290 34.1792 2 -6.33 <0.0001
Group 8 vs. 14 -500.140 21.2312 2 -23.56 <0.0001
Group 9 vs. 15 -748.700 41.9882 2 -17.83 <0.0001

148

 

 

Table A-2. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Group 12 vs. 16 135.670 5.3506 2 25.36 <0.0001
Group 13 vs. 17 369.130 23.2554 2 15.87 <0.0001
Group 16 vs. 17 3.469 1.8181 2 1.91 0.0583
Group 18 vs. 19 1.167 1.1478 2 1.02 0.3112
Group 17 vs. 18 -4.352 1.8915 2 -2.30 0.0228
Group 17 vs. 19 -3.186 1.9235 2 -1.66 0.0998
Group 21 vs. 22 -2.177 0.7343 2 -2.96 0.0035
Group 18 vs. 21 -3.149 0.8314 2 —3.79 0.0002
Group 19 vs. 22 -6.492 1.0796 2 -6.01 <0.0001
Group 24 vs. 25 -4.965 6.9392 2 -0.72 0.4754
Group 24 vs. 26 27.078 4.6482 2 5.83 <0.0001
Group 25 vs. 26 32.043 5.1936 2 6.17 <0.0001
Group 12 vs. 27 139.910 5.3579 2 26.11 <0.0001
Group 53 vs. 57 10.534 5.4500 2 1.93 0.0552
Group 54 vs. 55 -232.290 43.0462 2 -5.40 <0.0001
Group 54 vs. 56 -426.050 22.7045 2 -18.77 <0.0001
Group 54 vs. 57 104.310 15.6777 2 6.65 <0.0001
Group 55 vs. 56 -193.760 43.5054 2 -4.45 <0.0001
Group 55 vs. 57 336.590 40.2867 2 8.35 <0.0001
Group 56 vs. 57 530.360 16.8979 2 31.39 <0.0001
Group 8 vs. 51 375.210 16.5205 2 22.71 <0.0001
Group 8 vs. 52 373.410 16.2922 2 22.92 <0.0001
Group 8 vs. 53 370.810 16.6457 2 22.28 <0.0001
Group 8 vs. 54 277.040 22.2074 2 12.48 <0.0001
Group 8 vs. 55 44.751 43.2481 2 1.03 0.3025
Group 8 vs. 56 -149.010 23.0850 2 -6.45 <0.0001
Group 8 vs. 57 381.340 16.2238 2 23.51 <0.0001
Group 65 vs. 66 —16.521 8.5142 2 -1.94 0.0542
Group 65 vs. 67 -16.506 24.3329 2 -0.68 0.4986
Group 65 vs. 68 -299.790 10.8400 2 -27.66 <0.0001
Group 65 vs. 69 -765.910 16.6053 2 -46.12 <0.0001
Group 65 vs. 70 -1070.960 12.1784 2 -87.94 <0.0001
Group 65 vs. 71 12.780 7.3547 2 1.74 0.0844
Group 66 vs. 67 0.015 24.6750 2 0.00 0.9995
Group 66 vs. 68 -283.270 11.5876 2 -24.45 <0.0001
Group 66 vs. 69 -749.390 17.1027 2 -43.82 <0.0001
Group 66 vs. 70 -1054.440 12.8483 2 -82.07 <0.0001
Group 66 vs. 71 29.301 8.4178 2 3.48 0.0007
Group 67 vs. 68 -283.280 25.5709 2 -11.08 <0.0001
Group 67 vs. 69 -749.400 28.4974 2 -26.30 <0.0001

149

 

 

Table A-2. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Group 67 vs. 70 -1054.460 26.1664 2 -40.30 <0.0001
Group 67 vs. 71 29.286 24.2994 2 1.21 0.23
Group 68 vs. 69 —466.120 18.3716 2 -25.37 <0.0001
Group 68 vs. 70 -77l.180 14.4946 2 -53.20 <0.0001
Group 68 vs. 71 312.570 10.7644 2 29.04 <0.0001
Group 69 vs. 70 -305.060 19.1918 2 -15.90 <0.0001
Group 69 vs. 71 778.690 16.5561 2 47.03 <0.0001
Group 70 vs. 7] 1083.740 12.1112 2 89.48 <0.0001
Group 14 vs. 65 838.320 14.9443 2 56.10 <0.0001
Group 14 vs. 66 821.800 15.4951 2 53.04 <0.0001
Group 14 vs. 67 821.820 27.5626 2 29.82 <0.0001
Group 14 vs. 68 538.530 16.8853 2 31.89 <0.0001
Group 14 vs. 69 72.416 21.0557 2 3.44 0.0008
Group 14 vs. 70 -232.640 17.7741 2 -13.09 <0.0001
Group 14 vs. 71 851.100 14.8896 2 57.16 <0.0001
Group 187 vs. 188 3.064 0.5095 2 6.01 <0.0001
Group 187 vs. 190 -2.239 0.4933 2 -4.54 <0.0001
Group 188 vs. 190 -5.303 0.3235 2 -16.39 <0.0001
Group 189 vs. 190 5.915 0.3182 2 18.59 <0.0001
Group 191 vs. 192 -10.050 2.8663 2 -3.51 0.0006
Group 191 vs. 193 4.150 2.8999 2 1.43 0.1545
Group 191 vs. 194 13.600 2.8614 2 4.75 <0.0001
Group 192 vs. 193 14.200 0.8150 2 17.42 <0.0001
Grog 192 vs. 194 23.650 0.6650 2 35.56 <0.0001
Group 193 vs. 194 9.450 0.7975 2 11.85 <0.0001
Group 191-194 vs. 195-198 -4.793 1.6378 2 -2.93 0.004
Group 191 vs. 199 17.545 2.9203 2 6.01 <0.0001
Group 192 vs. 200 30.524 0.4902 2 62.26 <0.0001
Group 193 vs. 201 16.333 0.9189 2 17.77 <0.0001
Group 194 vs. 202 6.858 0.6636 2 10.34 <0.0001
Group 195 vs. 203 27.305 6.6266 2 4.12 <0.0001
Group 196 vs. 204 34.344 0.5548 2 61.90 <0.0001
Group 197 vs. 205 14.519 0.5552 2 26.15 <0.0001
Group 198 vs. 206 3.969 0.4140 2 9.59 <0.0001
Group 187 vs. 199-206 -2.839 0.6197 2 -4.58 <0.0001
Group 187 vs. 16-21 1.894 2.8290 2 0.67 0.5041
Group 187 vs. 51-53 -25.724 2.3893 2 -10.77 <0.0001
Group 187 vs. 65-67 -71.687 8.4243 2 -8.51 <0.0001
Group 187 vs. 71 -47.898 5.1409 2 -9.32 <0.0001

 

 

 

 

 

 

150

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3:3 4 233 22-42-25 - 23: m2 - 218m 722m 58% o28&-m2 8:282 mm
Rood 4 SSS 22-42-28 - 24:: m2 - 212: 228 5824 628224.422 62382 _N
886 4 22.9 25-42-88 + 2828 888 .82 + 23: .2: + 232 38.2% 5824 8&2: 2885 om
Smod 4 Nwfimmd E<m::._-u§ + E on + Swath—w o: 288$ 3
wooed 4 mhmamd Swami—$-55 + 85:8 8308 $2 + 2.3: :1: + zoom—m oE 0:085 3
S23 4 383 22:22-28 + 2828 888 .82 + 213 :E + 212: 08.2% 582» 8&2: 2882 2
m _ 5o 4 game 22-42-28 + 2838 888 22 + 21: 2: + 2.32 082% 9824 8&85 8682 2
$86 4 mmmmmd E<mlmmޤ + <3 0: + 592w on 2585 2
25.0 4 223 25-42-88 + 8:28 888 .82 + 214.2 WE + 232 m5 5824 8288822 2882 2
$88 4 egg 25-42-28 + 92 ea + 232 082% 5824 8&4: 2882 2
3.8.0 4 _mmmmd 2<mlwmé§ + 39:8 0258 $3 + 23: m5 + 583w 05 2585 S
388 4 29.3 254.42-42-28 224:8 888 .82 + 2253 22 + 2.52 382% 862w 8&2: 2882 :
2:3 4 $23 22-42-28 + :8an 886 $2 + 232 2: + 2122 08.2% :82» 8&822 8682 2
288 4 3.42.0 24.42-82-288 + 2828 8:2: 2.2 + 211 a: + 2.32 082% e824 8&822 2582 a
good 4 28% 22:42-88 + 223: mm + 2.32 .82.... 5824.. 8&4: 6882 a
good 4 83 2 .o Emlfiméqm - <3 on - 282% o: 83$on \1
Race 4 2.: 3 2212-28 - 21: E - 5828 ea 8an2 o
223 4 $2 3 22:22.25 - 23: E - 232 08.2% 5824 o28&-m2 64382 m
28¢ 4 2886 241.42-82-88 - 211 E - 232 82.4 82:4 62824-42 8282 4
8:3 4 $83 25:42-24 - 2258 m2 - 232 082% 5824 628.248: 8:2on m
88.0 4 ”2.22 22-82308 - 232 mm - 222 382% 5824 628,444: 62382 N
”23 4 $33 22-42-28 - 213 mm - 2.32 82% :82» 628.243 8282 _
68v SHE 2222.880 moi 90
am 4 3 on

 

 

 

 

 

nouamgn Ewe-3% 2:“ O4 any—93.. NEE.— naobmvéa 39.3.59 .8.“ 3:53— ..8535 .m.< wish.

mhADmHm mOmZHwO—m EC mum>d<z< <>OZ<

N.<

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$88 4 3828 2.214288 + 2888 888 .82 + 282: 422 + 232 22-5-3 88:4 8&42: 2882 mm
3.88 4 883 42.21288 - .2 ea - 8824 ea 2882 3.
wooed 4 Elwood m2<mlm3é§ .. <3 0: - amok—m oz 25er mm
28¢ 4 22.3 24.21.2288 + 2888 8:68 .82 + 2822 E + 232 22.5.3 8824 69288-42: 2882 2
Reed 4 Ease 2.8.2-8288 + 2888 888 .82 + 213 423 + 2452 2.29% 8824 8&8842: 2882 _m
.423 4 Sage 242122-28 + 2888 8.58 .82 + 22: 2: + 232 22.582 8822 8&82: 8682 cm
88.8 4 5.28 2.421288 + 2888 888 :2 + 23... E + 2.52 22.5.3 8824 8288-42: 8682 am
28.8 4 22.3 2.42-4288 + 2888 888 .82 + 24:: 422 + 2452 22.5.2... 8824 82482: 8682 mm
2:: 4 $.28 2<2|m2288 - 21.2 42 - 8822 ea 83.82 R
owood 4 5386 2<mllm3é§ - <30: .. 283w o: 8252on cm
28.8 4 Ease 2.8.2-8288 - <28 .. 2§m 22.4. 8824 62824-42 6832on mm
888 4 2228 2342-42-88 - 21: m2 1 2128 2.2.8 8824 6284286842 62382 E
223 4 2.28.8 2.8.2-8288 - 28% m2 - 245% 22.4.. 8822 62888-42 8an2 mm

 

 

 

 

 

.882an .3. 824...

152

Table A—4. Biosensor Results for Different HA-Glycan Pairs: Least Squares Means

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AQ t
Group Estimate Standard DF Value Pr > |t|
(mC) Error

1 0.1878 0.010840 2 17.33 <.0001
2 0.1219 0.005348 2 22.79 <.0001
3 0.1266 0.009954 2 12.72 <.0001
4 0.0891 0.004502 2 19.78 <.0001
5 0.1 186 0.017980 2 6.6 <.0001
6 0.1175 0.022740 2 5.16 <.0001
7 0.1340 0.007329 2 18.29 <.0001
8 0.4205 0.008652 2 48.6 <.0001
9 0.4744 0.022980 2 20.64 <.0001
10 0.2815 0.018840 2 14.94 <.0001
1 1 0.2432 0.022580 2 10.77 <.0001
12 0.2533 0.037310 2 6.79 <.0001
13 0.2784 0.008902 2 31.27 <.0001
14 0.2881 0.017220 2 16.74 <.0001
15 0.3322 0.028100 2 11.82 <.0001
16 0.2259 0.011470 2 19.7 <.0001
17 0.2974 0.000146 2 2040.9 <.0001
18 0.2957 0.006815 2 43 .4 <.0001
19 0.2518 0.020700 2 12.17 <.0001
20 0.2788 0.022170 2 12.58 <.0001
21 0.2438 0.003730 2 65.37 <.0001
22 0.2852 0.047880 2 5.96 <.0001
23 0.2434 0.080990 2 3.01 0.0037
24 0.1388 0.005600 2 24.78 <.0001
25 0.0895 0.001457 2 61.42 <.0001
26 0.0796 0.008042 2 9.89 <.0001
27 0.1547 0.016660 2 9.29 <.0001
28 0.1401 0.012480 2 11.23 <.0001
29 0.1433 0.003428 2 41.81 <.0001
30 0.0996 0.011210 2 8.88 <.0001
31 0.0835 0.006500 2 12.85 <.0001
32 0.1461 0.016450 2 8.89 <.0001
33 0.0688 0.005662 2 12.15 <.0001
34 0.0670 0.002516 2 26.63 <.0001
35 0.1390 0.004000 2 34.75 <.0001

 

 

 

 

 

 

153

 

 

Table A-5. Biosensor Group Comparisons: Estimates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A t
Comparison Egimate Standard DF Value Pr > |t|
(mC) Error

Group 1 vs. 9 -0.2866 0.02541 2 -11.28 <.0001
Group 1 vs. 21 -0.056 0.01146 2 -4.89 <.0001
Group 1 vs. 28 0.0477 0.01652 2 2.89 0.0053
Group 9 vs. 21 0.2306 0.02329 2 9.9 <.0001
Group 9 vs. 28 0.3343 0.02615 2 12.78 <.0001
Group 21 vs. 28 0.1037 0.01302 2 7.96 <.0001
Group 28 vs. 29 -0.0032 0.01294 2 -0.24 0.8074
Group 28 vs. 30 0.0406 0.01677 2 2.42 0.0183
Group 28 vs. 31 0.0566 0.01407 2 4.03 0.0001
Group 28 vs. 33 0.0714 0.0137 2 5.21 <.0001
Group 28 vs. 35 0.0011 0.0131 2 0.09 0.9309
Group 29 vs. 30 0.0437 0.01172 2 3.73 0.0004
Group 29 vs. 31 0.0598 0.00735 2 8.14 <.0001
Group 29 vs. 32 -0.0028 0.0168 2 -0.17 0.8668
Group 29 vs. 34 0.0763 0.00425 2 17.95 <.0001
Group 10 vs. 35 0.1425 0.01926 2 7.4 <.0001
Group 2 vs. 35 -0.0171 0.00668 2 -2.56 0.0126
Group 4 vs. 30 -0.0105 0.01208 2 -0.87 0.3875
Group 24 vs. 31 0.0553 0.00858 2 6.45 <.0001
Group 26 vs. 33 0.0108 0.00984 2 1.1 0.2773
Group 33 vs. 34 0.0018 0.0062 2 0.29 0.7731
Group 30 vs. 34 0.0326 0.01149 2 2.83 0.0061
Group 31 vs. 33 0.0147 0.00862 2 1.71 0.0926
Group 32 vs. 34 0.0791 0.01664 2 4.76 <.0001
Group 35 vs. 33 -0.0702 0.00693 2 -10.13 <.0001
Group 2 vs. 28 -0.0183 0.01357 2 -1.35 0.1831
Group 3 vs. 28 -0.0136 0.01596 2 -0.85 0.3982
Group 4 vs. 28 -0.0511 0.01326 2 -3.85 0.0003
Group 5 vs. 28 -0.0215 0.02188 2 -0.98 0.3292
Group 28 vs. 10-18 -0.1372 0.01415 2 -9.7 <.0001
Group 29 vs. 10-18 0134 0.0075 2 -17.88 <.0001
Group 15 vs. 33 0.2634 0.02867 2 9.19 <.0001

154

 

 

 

 

Table A-5. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Group7vs. 34 0.067 0.00775 2 8.65 <.0001
Grouplvs. 29 0.0445 0.01136 2 3.92 0.0002
Group 14 vs.3l 0.2046 0.0184 2 11.12 <.0001
Grouplvs2 0.06594 0.01208 2 5.46 <.0001
Grouplvs.3 0.06125 0.01471 2 4.16 0.0001
Group2vs.3 -0.0047 0.0113 2 -0.42 0.6798
Gr01m9vs. 10 0.1929 0.02972 2 6.49 <.0001
Group9vs. 11 0.139 0.02074 2 6.7 <.0001
Group 10 vs.11 0.03831 0.02941 2 1.3 0.1983
Group1,2,3 vs. 4,5,6,7 0.03063 0.00919 2 3.33 0.0016
Group 9-11vs.12-19 0.05519 0.01429 2 3.86 0.0003
Group8vs.9 -0.0539 0.02456 2 -2.2 0.0324
Group1vs.9 -0.2866 0.02541 2 -11.28 <.0001
Group1vs.8 02327 0.01387 2 -16.78 <.0001
Group 1 vs. 20 -0091 0.02468 2 -3.69 0.0005
Grouplvs.21 -0.056 0.01146 2 -4.89 <.0001
Group 20 vs.21 0.03501 0.02248 2 1.56 0.1253
Group 2123 vs. 2427 0.1418 0.03176 2 4.46 <.0001
Group1,2,3 vs. 9,10,11 -0.l876 0.01349 2 -1391 <.0001
Group1,2,3 vs.21,22,23 -0112 0.03182 2 -352 0.0009
Group 9,10,11 vs.

21,22,23 0.0756 0.03377 2 2.24 0.0293
Group9vs. 20 0.1956 0.03194 2 6.13 <.0001
Group 2,3 vs. 4,5,6,7 0.00943 0.00944 2 1 0.3222
Group10,11vs.12-19 -00155 0.0163 2 -095 0.3461
Group 12 vs.13 -0.0251 0.03836 2 -0.65 0.516
Group 12 vs. 14 -0.0348 0.04109 2 -0.85 0.4006
Group 12 vs. 15 -0.0789 0.04671 2 -l.69 0.0969
Grog) 12vs. 16 0.02737 0.03903 2 0.7 0.4862
Group12vs. 17 -0.0441 0.03731 2 -l.18 0.2426
Group 12 vs. 18 -00424 0.03793 2 -1.12 0.2684
Group 12 vs. 19 0.00149 0.04266 2 0.03 0.9723
Group13 vs. 14 -0.0097 0.01938 2 -0.5 0.6174
Group13 vs. 15 -0.0538 0.02948 2 -1.83 0.0733
Grown vs. 16 0.05245 0.01452 2 3.61 0.0007
Group13 vs.17 -0019 0.0089 2 -2.13 0.0374

 

 

 

155

 

 

 

 

Table A-5. Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Group 13 vs. 18 -0.0173 0.01121 2 -1.55 0.1278
Group 13 vs. 19 0.02656 0.02253 2 1.18 0.2435
Group 14 vs. 15 -0.0441 0.03296 2 -1.34 0.1865
Group 14 vs. 16 0.06218 0.02069 2 3.01 0.004
Group 14 vs. 17 -0.0093 0.01722 2 -0.54 0.5926
Group 14 vs. 18 -0.0076 0.01852 2 -0.41 0.683
Group 14 vs. 19 0.0363 0.02692 2 1.35 0.1831
Group 15 vs. 16 0.1063 0.03035 2 3.5 0.0009
Group 15 vs. 17 0.03483 0.0281 2 1.24 0.2205
Group 15 vs. 18 0.0365 0.02892 2 1.26 0.2123
Group 15 vs. 19 0.0804 0.0349 2 2.3 0.0251
Group 16 vs. 17 -0.0715 0.01147 2 -6.23 <.0001
Group 16 vs. 18 -0.0698 0.01334 2 -5.23 <.0001
Group 16 vs. 19 -0.0259 0.02366 2 -1.09 0.2789
Group 17 vs. 18 0.00166 0.00682 2 0.24 0.8081
Group 17 vs. 19 0.04557 0.0207 2 2.2 0.032
Group 18 vs. 19 0.0439 0.02179 2 2.01 0.0489
Group 8 vs. 15 0.08828 0.0294 2 3 0.0041
Group 24,27 vs. 25,26 0.06223 0.00969 2 ‘ 6.42 <.0001
Group 21 vs. 24,27 0.09706 0.00955 2 10.17 <.0001
Group 17,18 vs. 16,19 0.05768 0.01231 2 4.68 <.0001
Group 12-15 vs. 16-19 0.02029 0.01406 2 1.44 0.1548
Group 9 vs. 1215 0.1864 0.02623 2 7.11 <.0001
Group 9 vs. 16-19 0.2067 0.02379 2 8.69 <.0001

 

 

 

 

 

 

156

 

REFERENCES

Ahuja T., Mir I.A., Kumar D., and Rajesh (2007). Biomolecular immobilization on
conducting polymers for biosensing applications. Biomaterials 28(5): 791-805.

Air G.M., Ghate A.A., and Stray SJ. (1999). Influenza neuraminidase as target for
antivirals. Adv. Virus Res. 54: 375-402.

Alexander H.E. (1943a). Treatment of Haemophilus influenzae infections and of
meningococcic and pneumococcic meningitis. Am. J. Dis. Child. 66: 172-187.

Alexander H.E. (1943b). Experimental basis for treatment of Haemophilus influenzae
infections. Am. J. Dis. Child. 66: 160-171.

Amano Y. and Cheng Q. (2005). Detection of influenza virus: traditional approaches and
development of biosensors. Anal. Bioanal. Chem. 381: 156-164.

 

Amico C. (2009). How much will the H1N1 flu cost the US? [WWW article]. URL,
http://www.pbs.org/newshour/updates/health/july-dec09/flu-costs_10-08.html, Accessed
February 16, 2010.

Arora, K., Prabhakar, N., Chand, S., and Malhotra, B.D. (2007). Escherichia coli
genosensor based on polyaniline. Anal. Chem. 79(16): 6152-6158.

Atrnar R.L., Baxter B.D., Dominguez E.A., and Taber L.H. (1996). Comparison of
Reverse Transcription-PCR with Tissue Culture and Other Rapid Diagnostic Assays for
Detection of Type A Influenza Virus. J. Clin. Microbiol. 34: 2604—2606.

Aulitzky W.E., Schulz T.F., Tilg H., Niederwieser D., Larcher K., Otberg L., Scriba M.,
Martindale J ., Stern A.C., Grass P., Mach M., Dierich MP, and Huber C. (1991). Human
monoclonal antibodies neutralizing cytomegalovirus (CMV) for prophylaxis of CMV
disease: report of a phase I trial in bone marrow transplant recipients. J. Infect. Dis. 163:
1344-7.

Austrian R. (1994). Confronting drug-resistant pneurnococci [editorial]. Ann. Intern. Med.
121: 807-9.

Ayesh AS. (2009). Dielectric Properties of Polyethylene Oxide Doped with NHA Salt.
Polym. J. 41: 616-621.

Babes L., Denzoit B., Tanguy G., Le Jeune J .J ., and Jallet P. (1999). Synthesis of iron

oxide nanoparticles used as MRI contrast agents: a parametric study. J. Colloids Interface
Sci. 212: 474—82.

157

Barbas CF. 111, Crowe J .E., Cababa D., Jones T.M., Zebedee S.L., Murphy B.R.,
Chanock R.M., and Burton DR. (1992). Human monoclonal Fab fragments derived fi'om
a combinatorial library bind to respiratory syncytial virus F glycoprotein and neutralize
infectivity. Proc. Natl. Acad. Sci. USA 89:10164-10168.

Barandun S., Kistler P., Jennet F., and Isliker H. (1962). Intravenous administration of
human gamma-globulin. Vox Sang. 7: 157.

Bard A.J. and Faulkner L.R. Electrochemical Methods: Fundamentals and Applications.
New York: John Wiley & Sons, Inc., 2000.

Barnes G.L., Doyle L.W., Hewson P.H., Knoches A.M., McLellan J .A., Kitchen W.H.,
and Bishop RF. (1982). A randomised trial of oral garnmaglobulin in low-birth-weight
infants infected with rotavirus. Lancet. 1(8286): 1371-1373.

Barry W., Hudgins L., Donta ST, and Pesanti EL. (1992). Intravenous immunoglobulin
therapy for toxic shock syndrome. J. Amer. Med. Assoc. 267: 3315-6.

Bartlett RN. and Whitaker R.G. (1988). Strategies for the development of amperometric
enzyme electrodes. Biosensors. 3(6): 359-379.

Behring EA. and Kitasato S. (1890). Uber das Zustandekommen der Diphtherie-
Immunitat und der tetanus-immunitat bei thieren. Deutsch Med. Wochenschr. 49: 1113.

Bender S. and Sadik DA. (1998). Direct electrochemical immunosensor for
polychlorinated biphenyls (PCBs). Environ. Sci. Technol. 32: 788-797.

Berggren M., Nilsson D., and Robinson ND. (2007). Organic materials for printed
electronics. Nat. Mater. 6: 3-5.

Berry C.C., Wells 8., Charles S., and Curtis A.S.G. (2003). Dextran and albumin
derivatised iron oxide nanoparticles: influence on fibroblasts in vitro. Biomaterials. 24:
4551-4557.

Blixt 0., Head S., Mondala T., Scanlan C., Huflejt M.E., Alvarez R., Bryan M.C., Fazio
F., Calarese D., Stevens J ., Razi N., Stevens D.J., Skehel J .J ., van Die 1., Burton D.R.,
Wilson I.A., Cummings R., Bovin N., Wong C.-H., and Paulson J .C. (2004). Printed
covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc. Nat.
Acad. Sci. 101(49): 17033-17038.

Blum L.J. and Coulet P.R. (Eds.), 1991. Biosensor Principles and Applications. Marcel
Dekker, New York.

Bodensteiner J .B., Morris H.H., Howell J .T., and Schochet SS. (1979). Chronic ECHO

type 5 virus meningoencephalitis in X-linked hypogammaglobulinemia: treatment with
immune plasma. Neurology 29: 815-9.

158

Boon A.C., French A.M., Fleming D.M., and Zambon MC. (2001). Detection of
Influenza A Subtypes in Community-Based Surveillance. J. Med. Virol. 65: 163—170.

Brand CM. and Skehel J .J . (1972). Crystalline antigen from the influenza virus enve10pe.
Nature. 238: 467-474.

Branham SE. (1935). Protection of mice against meningococcus infection by polyvalent
antimeningococcic serum. Public Health Rep. 50: 768-778.

Branham SE. (1937). Sulphanilamide, serum, and combined drug and serum therapy in
experimental meningococcus and pneumococcus infections in mice. Public Health Rep. n
52: 685-695. '

Brunell P.A., Gershon A.A., Hughes W.T., Riley H.D. Jr., and Smith J. (1972).
Prevention of varicella in high risk children: a collaborative study. Pediatrics 50: 718-22.

Burnet F. and Clark E. Influenza: a survey of the last 50 years in the light of modern
work on the virus of epidemic influenza. Melbourne: MacMillan, 1942. L

 

Burroughes J .H., Jones C.A., and Friend RH. (1988). New semiconductor device physics
in polymer diodes and transistors. Nature. 335: 137-141.

Bussel J .B. and Cunningham-Rundles C. (1985). Intravenous usage of garnmaglobulin:
humoral immunodeficiency, immune thrombocytopenic purpura, and newer indications.
Cancer Invest. 3: 361-6.

Cameron M.L., Schell W.A., Bruch S., Bartlett J .A., Waskin H.A., and Perfect J .R.
(1993). Correlation of in vitro fluconazole resistance of Candida isolates in relation to
therapy and symptoms of individuals seropositive for human immunodeficiency virus
type I. Antimicrob. Agents. Chemother. 37: 2449-53.

Centers for Disease Control and Prevention (CDC). (Aug. 10, 2009). Interim Guidance
for the Detection of Novel Influenza A Virus Using Rapid Influenza Diagnostic Tests
[WWW document]. URL, http://www.cdc.gov/h1n1flu/guidance/rapid_testing.htm.
Accessed December 10, 2009.

Casadevall A. (1996). Antibody-Based Therapies for Emerging Infectious Diseases.
Emerg. Infect. Dis. 2(3): 200-208.

Casadevall A. and Scharff MD. (1994). “Serum therapy” revisited: animal models of
infection and the development of passive antibody therapy. Antimicrob. Agents
Chemother. 38: 1695-1702.

Casadevall A. and Scharff MD. (1995). Return to the past: the case for antibody-based
therapies in infectious diseases. Clin. Infect. Dis. 21: 150-61.

159

Grennan K., Killard A.J., Hanson C.J., Cafolla A.A., and Smyth MR. (2006).
Optimisation and characterization of biosensors based on polyaniline. Talanta. 68(5):
1 591 -1600.

Centers for Disease Control and Prevention (CDC). (2009). Novel HlNl Flu:
Background on the Situation [WWW document]. URL,
http://www.cdc.gov/hln1flufbackgroundhtm, Accessed June 10, 2010.

Charych D.H., Nagy J .O., Spevak W., and Bednarski MD. (1993). Direct Colorimetric
Detection of a Receptor-Ligand Interaction by a Polymerized Bilayer Assembly. Science.
261: 585-588.

Chaudhuri D. and Sarma DD. (2006). Blue emitting polyaniline. Chem. Commun. 2681-
2683.

Cho C.T., Feng K.K., and Brahmacupta N. (1976). Synergistic antiviral effects of adenine
arabinoside and humoral antibodies in experimental encephalitis due to Herpesvirus
hominis. J. Infect. Dis. 133: 157-167.

Class E.C., Osterhaus A.D., van Beek R., De Jong J .C., Rimmelzwaan G.F., Senne D.A.,
Krauss S., Shortridge K.F., and Webster R.G. (1998). Human influenza A H5N1 virus
related to a highly pathogenic avian influenza virus. Lancet. 351: 472-477.

Cohn E.J., Oncley J .L., Strong L.E., Hughes W.L., and Armstrong SH. (1944). Chemical,
clinical and immunological studies on the products of human plasma fractionation: I. The
characterization of protein fractions of human plasma. J. Clin. Invest. 23: 417-432.

Conti D.J., Freed B.M., Gruber S.A., and Lempert N. (1994). Prophylaxis of primary
cytomegalovirus disease in renal transplant recipients: a trial of ganciclovir vs.
immunoglobulin. Arch. Surg. 129: 443-7.

Cooper MA. (2003). Label-free screening of bio-molecular interactions. Anal. Bioanal.
Chem. 377: 834-842.

Cooper M.A., Dultsev F .N., Minson T., Ostanin V.P., Abell C., and Klenerman K. (2001).
Direct and sensitive detection of a human virus by rupture event scanning. Nat
Biotechnol. 19: 833—837.

Correa A.G., Baker C.J., Schutze GE, and Edwards MS. (1994). Immunoglobulin G
enhances C3 degradation on coagulase-negative staphylococci. Infect. Irnmun. 62: 2362-
6.

Cosnier S. and Lepellec A. (1999). Poly (pyrrole-biotin): a new polymer for biomolecule
grafting on electrode surfaces. Electrochim. Acta. 44: 1833-1836.

160

 

Cross A. (1995). Intravenous immunoglobulins to prevent and treat infectious diseases. In:

Atassi M.Z. and Bixler G.S.J., Eds. Irnmunobiology of Proteins and Peptides VIII. New
York: Plenum Press, 1995.

Dallas P., Moutis N., Devlin E., Niarchos D., and Petridis D. (2006). Characterization,
electrical and magnetic properties of polyaniline/maghemite nanocomposites.
Nanotechnology. 17(19): 5019-5026.

Davis SS. (1997). Biomedical applications of nanotechnology—implications for drug
targeting and gene therapy. Trends Biotechnol. 15(6): 217—224.

Diaz A.F., Kanazawa K.K., and Gardini GP. (1979). Electrochemical polymerization of
pyrrole. J. Chem. Soc., Chem. Commun. 635-636.

Drummond T.G., Hill M.G., and Barton J .K. (2003). Electrochemical DNA sensors. Nat.
Biotechnol. 21 :1 192-1 199.

D’Souza SF. (2001). Review: Microbial Biosensors. Biosens. Bioelectron. 16: 337-353.

Dubowski J .J . (2006, Oct 3). Research opportunity for MSc and PhD students in physics,
material engineering, bio-physics and nanomedicine. Laboratory for Quantum
Semiconductors and Laser-based Nanotechnology [WWW document]. URL,
http://www.gel.usherbrooke.ca/cegi/documents/Epitaxial_QD_Device_v2.pdf. Accessed
June 2, 2010.

Dultsev F .N., Ostanin V.P., and Klenerman D. (2000). “Hearing” Bond Breakage.
Measurement of Bond Rupture Forces Using a Quartz Crystal Microbalance. Langmuir.
16: 5036-5040.

Dultsev F.N., Speight R.E., Fiorini M.T., Blackburn J.M., Abell C., Ostanin V.P., and
Klenerman D. (2001). Direct and Quantitative Detection of Bacteriophage by “Hearing”
Surface Detachment Using a Quartz Crystal Microbalance. Anal Chem. 73: 3935-3939.

Ellis J .S. and Zambon M.C. (2002). Molecular diagnosis of influenza. Rev. Med. Virol.
12: 375—389.

Eun A.J., Huang L., Chew F .T., Li SF, and Wong SM. (2002). Detection of two orchid
viruses using quartz crystal microbalance (QCM) immunosensors. J. Virol. Methods. 99:
71—79.

Expert Committee on Influenza, WHO Tech. Rep. Ser. No. 64. (1953).

Feinberg S.M. The therapy of (horse) serum reactions. General rules in the administration
of therapeutic serums. J. Amer. Med. Assoc. 107: 1717-1719.

161

 

!.""

 

Felton L.D. (1928). The units of protective antibody in antipneumococcus sermn and
antibody solution. J. Infect. Dis. 43: 531-42.

Fisher M.W. (1957). Synergism between human gamma globulin and chloroarnphenicol
in the treatment of experimental bacterial infections. Antibiot. Chemother. 7: 315-321.

Flexner S. (1913). The results of the serum treatment in thirteen hundred cases of
epidemic meningitis. J. Exp. Med. 17: 553-576.

Focke W.W., Wnek GE, and Wei Y. (1987). Influence of oxidation state, pH, and
counterion on the conductivity of polyaniline. J. Phys. Chem. 91(22): 5813-5818.

F odor E., Devenish L., Engelhardt O.G., Palese P., Brownlee G.G., and Garcia-Sastre A.
(1999). Rescue of influenza A virus from recombinant DNA. J. Virol. 73: 9679-9682.

Fothergill L.R.D. (193 7). Haemophilus influenzae (Pfeiffer bacillus) meningitis and its
specific treatment. New Engl. J. Med. 216: 587-590.

Frickhofen N., Abkowitz J .L., Safford M., Berry J .M., Antunez-de-Mayolo J ., Astrow A.,
Cohen R., Halperin1., King L., Mintzer D., Cohen B., and Young NS. (1990). Persistent
B19 parvovirus infection in patients infected with human immunodeficiency virus type 1
(HIV-1): a treatable cause of anemia in AIDS. Ann. Intern. Med. 113: 926-33.

Gajendran P. and Saraswathi R. (2008). Polyaniline—carbon nanotube composites. Pure
Appl. Chem. 80(11): 2377-2395.

Gambhir A., Gerard M., Mulchandani A., and Malhotra B.D. (2001). Coimmobilization
of urease and glutamate dehydrogenase in electrochemically prepared polypyrole-
polyvinyl sulfonate films. Appl. Biochem. Biotechnol. 96: 249-258.

Garjonyte R. and Malinauskas A. (2000). Amperometric glucose biosensors based on
Prussian Blue—and polyaniline—glucose oxidase modified electrodes. Biosens.
Bioelectron. 15(9-10): 445-451.

GE Healthcare. (2010). About Biacore Systems [WWW document]. URL,
http://www.biacore.com/lifesciences/technology/introduction/data_interaction/index.htrn1
Accessed July 10, 2010.

Gerard M., Chaubey A., and Malhotra B.D. (2002). Application of Conducting Polymers
to Biosensors. Biosens. Bioelectron. 17: 345-359.

Glarum SH. and Marshall J .H. (1986). In situ potential dependence of poly(aniline)
paramagnetism. J. Phys. Chem. 90(23): 6076.

162

 

Gorman CB. and Grubbs RH. (1991). Conjugated Polymers: The Interplay Between
Synthesis, Structure, and Properties. In Conjugated Polymers. Bredas, J .L., and Silbey, R.
(eds). Kluwer Academic Publishers, pp. 1-48.

Gospodinova, N., Mokreva, P., and Terlemezyan, L. (1996). Concomitant processes in
the redox switching of polyaniline. Polymer International 41(1): 79-84.

Gref R., Minamitake Y., Peracchia M.T., Trubetskoy V., Torchilin V., and Langer R.
(1994). Biodegradable long-circulating polymeric nanospheres. Science 18(263): 1600—
1603.

Grzeszczuk M. and Poks P. (1995). Analysis of charge transport impedance in the
reduction of thin films of conducting polyaniline in aqueous trichloroacetic acid solutions.
J. Electroanal. Chem. 387(1-2): 79-85.

Grzeszczuk M. and Szostak R. (2003). Electrochemical and Raman studies on the redox
switching hysteresis of polyaniline. Solid State Ionics 157(1-4): 257-262.

Grzeszczuk M. and Zabinska-Olszak G. (1993). Ionic transport in polyaniline film
electrodes: an impedance study. J. Electroanal. Chem. 359(1-2): 161-174.

Harnad-Schifferli K., Schwartz J .J ., Santos A.T., Zhang S., and Jacobson J .M. (2002).
Remote electronic control of DNA hybridization through inductive coupling to an
attached metal nanocrystal antenna. Nature. 415: 152-155.

Hatta M., Gao P., Halfrnann P., and Kawaoka Y. (2001a). Molecular basis for high
virulence of Hong Kong H5N1 influenza A viruses. Science. 293: 1840-1842.

Hatta M., Neumann G., and Kawaoka Y. (2001b). Reverse genetics approach towards
understanding pathogenesis of H5N1 Hong Kong influenza A virus infection. Philos.
Trans. R. Soc. Lond. B. Biol. Sci. 356: 1841-1843.

Hay A.J., Wolstenholme A.J., Skehel J .J ., and Smith M.H. (1985). The molecular basis of
the specific anti-influenza action of arnantadine. EMBO J. 4: 3021-3024.

Heeger A. and Smith P. (1991). Solution Processing of Conducting Polymers:
Opportunities for Science and Technology. In Conjugated Polymers. Bredas, J .L., and
Silbey, R. (eds). Kluwer Academic Publishers, pp. 141-210.

Hindiyeh M., Goulding C., Morgan H., Kenyon B., Langer J ., Fox L., Dean G.,
Woolstenhulme D., Tumbow A., Billetdeaux E., Shakib S., Gordon C., Powers A.,
Vardeny G., Johnson M., Skodack-Jones L., and Carroll K. (2000). Evaluation of BioStar
FLU OIA assay for rapid detection of influenza A and B viruses in respiratory specimens.
J. Clin. Virol. 17: 119—126.

163

 

Homeland Security Council (HSC). (2005, Nov 1). National Strategy for Pandemic
Influenza [WWW document]. URL, http://www.whitehouse.gov/homeland/nspipdf.
Accessed March 24, 2007.

Homola J. (2003). Present and firture of surface plasmon resonance biosensors. Anal.
Bioanal. Chem. 377: 528—539.

Hong S.Y. and Park SM. (2005) Electrochemistry of conductive polymers 36. pH
dependence of polyaniline conductivities studied by current-sensing atomic force
microscopy. Journal of Physical Chemistry B 109(19): 9305-9310.

Huang W., Taylor S., Fu K.F., Zhang D.H., Hanks T.W., Rao A.M., and Sun Y.P. (2002).
Attaching proteins to carbon nanotubes via diimide-activated arnidation. Nano Lett. 2:
3 1 1-3 14.

Hurt A.C., Baas C., Deng Y.M., Roberts S., Kelso A., and Barr LG. (2009). Performance
of influenza rapid point-of-care tests in the detection of swine lineage A(H1N1) influenza
viruses. Influenza Other Respi. Viruses. 3: 171-176.

Ivnitski D., Ihab A.H., Plarnen A., and Ebtisarn W. (1999). Biosensors for detection of
pathogenic bacteria. Biosens. Bioelectron. 14: 599-624.

Ivory D.M., Miller G.G., Sowa J .M., Shacklett L.W., Chance RR, and Baughman RH.
(1979). Highly conducting charge-transfer complexes of poly(p-phenylene). J. Chem.
Phys. 71: 1506.

Jahrling P.B., Peters C.J., and Stephen EL. (1984). Enhanced treatment of Lassa fever by
immune plasma combined with ribavirin in cynomolgus monkeys. J. Infect. Dis. 149:
420-427.

Jager E.W.H., Smela E., and Inganas O. (2000). Microfabricating conjugated polymer
actuators. Science. 290(5496): 1540-1545.

Jason-Moller L., Murphy M., and Bruno J. (2006). Overview of Biacore systems and
their applications. Curr. Protoc. Protein Sci. S45: 19.13.

Jennes L., Conn P.M., and Stumpf W.E. (1986). Synthesis and use of colloidal gold-
coupled receptor ligands. In: Methods in Enzymology, Vol. 124. Conn P.M., Ed.
Academic Press, New York, pp. 36-47.

Jiang J ., Li LC, and Xu F. (2006). Preparation, characterization and magnetic properties

of PAN I/La-substituted LiNi ferrite nanocomposites. Chinese J. Chem. 24(12): 1804-
1809.

164

 

Jin, J .H., Zhang, D., Alocilja, EC, and Grooms, D. (2008). Label-free DNA sensor on
nano-porous silicon-polypyrrole chip for monitoring Salmonella species. IEEE Sensors
Journal. 8(6): 891-895.

Johnson RC. (2005, Nov 28). EE Times: CMOS chip slashes time needed to ID flu
strains [WWW document]. URL,
http://www.eetimes.com/showArticle.jhtml?articleID=174401018. Accessed June 27,
2007.

Kahn KL. (1921). Studies on Complement Fixation: I. The Rate of Fixation of
Complement at Different Temperatures. The Journal of Experimental Medicine. 34: 217-
230.

Kawai T., Kuwabara T., and Yoshino K. (1991). Electrochemical Preparation of an
Insulating Thin Film and Its Characterization. Jpn. J. Appl. Phys. 30: L1192-L1194.

Khurana S., Chearwae W., Castellino F ., Manischewitz J ., King L.R., Honorkiewicz A.,
Rock M.T., Edwards K.M., Del Giudice G., Rappuoli R., and Golding H. (2010).
Vaccines with MF 59 adj uvant expand the antibody repertoire to target protective sites of
pandemic avian H5N1 influenza virus. Sci. Trans]. Med. 2(15): lSraS.

 

Kim H.-S. and Wamser CC. (2006). Photoelectropolymerization of aniline in a dye-
sensitized solar cell. Photochem. Photobio. Sci. 5: 955-960.

Kim J .-H., Cho J .-H., Cha G.-S., Lee C.-W., Kim H.-B., and Pack S.-H. (2000).
Conductimetric membrane strip immunosensor with polyaniline-bound gold colloids as
signal generator. Biosens. Bioelectron. 14(12): 907-915.

Kim S.W., Cho H.G., and Park CR. (2009). Fabrication of Unagglomerated Polypyrrole
Nanospheres with Controlled Sizes from a Surfactant-Free Emulsion System. Langmuir.
25(16): 9030-9036.

King D. (2006). Development of Rapid Field-Based Diagnostics to Improve the Control
of Exotic Livestock Diseases. Speech from Agricultural Biotechnology International
Conference 2006 [WWW document]. URL,
www.abic2006.org/pdf/speakers/DonaldKing.pdf. Accessed April 20, 2007.

Kistler P. and Nitschmann H. (1962). Large scale production of human plasma fractions.
Vox. Sang. 7: 414-424.

Kohler G. and Milstein C. (1975). Continuous cultures of fused cells secreting antibody
of predefined specificity. Nature. 256: 495-497.

Kim KS. (1987). Efficacy of human immunoglobulin and penicillin G in treatment of
experimental group B streptococcal infection. Pediatr. Res. 21: 289-92.

165

 

Krug RM. (2003). The potential use of influenza virus as an agent for bioterrorism.
Antiviral Research. 57: 147-150.

Kuno A., Uchiyarna N., Koseki-Kuno S., Ebe Y., Takashima S., Yarnada M., and
Hirabayashi J. (2005). Evanescent-field fluorescence-assisted lectin microarray: a new

strategy for glycan profiling. Nature Methods. 2: 851-856.

Lacava L.M., Lacava Z.G.M., Da Silva M.F., Silva O., Chaves S.B., Azevedo R.B.,
Pelegrini F., Gansau C., Buske N., Sabolovic D., and Morais RC. (2001). Magnetic
resonance of a dextran-coated magnetic fluid intravenously administered in mice.
Biophys. J. 80: 2483—2486.

Laitinen M.P.A. and Vuento M. (1996). Affinity immunosensor for milk progesterone:
identification of critical parameters. Biosens. Bioelectron. 11(12): 1207-1214.

Laitinen O.H., Nordlund H.R., Hytonen V.P., and Kulomaa MS. (2007). Brave new
(strept)avidins in biotechnology. Trends Biotechnol. 25(6): 269-277.

Lang A.B., Cryz S.J., Schurch U., Ganss M.T., and Bruderer U. (1993). Immunotherapy
with human monoclonal antibodies. J. Irnmunol. 151: 466-72.

Leca-Bouvier B. and Blum L.J. (2005). Biosensors for Protein Detection: A Review.
Analytical Letters. 38:1491-1517.

Lee C.-W. and Suarez D.lL. (2004). Application of real-time RT-PCR for the quantitation
and competitive replication study of H5 and H7 subtype avian influenza virus. J. Virol.
Meth.]19:151-158.

Lee M.S., Tzeng J .S., Chen Y.C., and Shirarnatsu T. (1989). Stability of Electrical and
Photovoltaic Characteristics of HBr Solution-Doped Polyacetylene. Jpn. J. Appl. Phys.
28: 1008-1012.

Li, G., J iang, L., and Peng, 11. (20078) One-dimensional polyaniline nanostructures with
controllable surfaces and diameters using vanadic acid as the oxidant. Macromolecules
40(22): 7890-7894.

Li, X., Shen, J .Y., Wan, M.X., Chen, Z.J., and Wei, Y. (2007b) Core-shell structured and
electro-magnetic functionalized polyaniline composites. Synthetic Metals 157(13-15):
575-579.

Liu Y., Matsumura T., Imanishi N., Hirano A., Ichikawa T., and Takeda Y. (2005).
Electrochem. Solid State Lett. 8(2): A599-A602.

Luke T.C., Kilbane B.M., Jackson J .L., and Hoffman S.L. (2006). Meta-Analysis:

Convalescent Blood Products for Spanish Influenza Pneumonia: A Future H5N1
Pandemic? Ann. Int. Med. 145(8): 599-609.

166

Lyons, M.E.G. (1994) Chapter 1, Electroactive Polymer Electrochemistry, Part 1:
Fundamentals. Plenum Press: New York, NY.

MacBeath G. and Schreiber S.L. (2000). Printing proteins as microarrays for high-
throughput function determination. Science. 8: 1760-1763.

MacDiarmid AG. (2001). Synthetic metals: A novel role for organic polymers (Nobel
Lecture). Angew. Chem. Int. Ed. 40: 2581-2590.

MacDiarmid A.G., Chiang J .C., Richter AF, and Epstein A.J. (1987). Polyaniline: a new
concept in conducting polymers. Synthetic Metals. 18(1-3): 285-290.

Magnard C., Valette M., Aymard M., and Lina B. (1999). Comparison of Two Nested
PCR, Cell Culture, and Antigen Detection for the Diagnosis of Upper Respiratory Tract
Infections due to Influenza Viruses. J. Med. Virol. 59: 215—220.

Malhotra B.D., Chaubey A., and Singh SP. (2006). Prospects of conducting polymers in
biosensors. Ana]. Chirn. Acta 578: 59-74.

Malinauskas A., Malinauskiene J ., and Rarnanavicius A. (2005). Topical Review:
Conducting polymer-based nanostructurized materials: electrochemical aspects.
Nanotechnology. 16: R5 1-62.

Mann S. (2000). The Chemistry of Form. Angew. Chem. Int. Edit. 39(19): 3392-3406.
Marinakos S.M., Brousseau LG 111, Jones A., and F eldheim D1. (1998). Template
synthesis of one-dimensional Au, Au-poly(pyrrole), and poly(pyrrole) nanoparticle arrays.
Chem. Mater. 10: 1214-1219.

Marks G and Beatty W.K. Epidemics. New York: Scribners, 1976.

Masihi K.N. and Lange W. (1980). J. Immunol. Methods 36: 173—179.

Matrosovich M.N., Matrosovich T.Y., Gray T., Roberts N.A., and Klenk H.-D. (2004).
Human and avian influenza viruses target different cell types in cultures of human airway
epithelium. Proc. Natl. Acad. Sci. USA. 101: 4620-4624.

McCloskey R.V. (1985). Corynebacterium diphtheriae (diphtheria). In: Mandel] G.L.,
Douglas R.G. Jr., Bennett J .E., Eds. Principles and practice of infectious diseases. 2nd ed.
New York: John Wiley and Sons, 1171-1174.

Michaelis M., Doerr H.W., and Cinatl J. Jr. (2009). Novel swine-origin influenza A virus

in humans: another pandemic knocking at the door. Med. Microbiol. Immunol. 198: 175-
183.

167

McKimm-Breschkin J .L., Colman P.M., Jin B., Krippner G.Y., McDonald M., Reece
P.A., Tucker S.P., Waddington L., Watson K.G., and Wu W.Y. (2003). Angew. Chem.
Int. Ed. Eng]. 42: 3118—3121.

Meeusen, C., Alocilja, EC, and Osbum, W. (2005). Detection of E. coli OlS7:H7 Using
a Miniaturized Surface Plasmon Resonance Biosensor. Transactions of the ASAE.
48(6):2409-2416.

Mofenson L.M., Moye J. Jr., Korelitz J ., Bethe] J ., Hirschhom R., and Nugent R. (1994).

Crossover of placebo patients to intravenous immunoglobulin confirms efficacy for

prophylaxis of bacterial infections and reduction of hospitalizations in human

immunodeficiency virus-infected children. Pediatr. Infect. Dis. J. 13: 477-84. .211

Moghimi S.M., Hunter A.C., and Murray J .C. (2001). Long-circulating and target-
specific nanoparticles: theory to practice. Pharm. Rev. 53: 283—318.

Montalto NJ. (2003). An Office-Based Approach to Influenza: Clinical Diagnosis and
Laboratory Testing. Am. Fam. Physician 67: 111—118. L

 

Muhammad-Tahir, Z., Alocilja, EC, and Grooms, D.L. (2005a). Polyaniline synthesis
and its biosensor application. Biosens. Bioelectron. 20(8): 1690-1695.

Muhammad-Tahir, Z., Alocilja, EC, and Grooms, D.L. (2005b). Rapid detection of
bovine viral diarrhea virus as surrogate of bioterrorism agents. IEEE Sensors Journal.
5(4):757-762.

Muhammad-Tahir, Z., Alocilja, EC, and Grooms, D.L. (2007). Indium Tin Oxide-
Polyaniline Biosensor: Fabrication and Characterization. Sensors Journal. 7:1123-1140.

National Institute of Allergy and Infectious Disease (N IAID). (2005). Focus on the Flu:
Antigenic Drift vs. Antigenic Shift [WWW document]. URL,
http://www3.niaid.nih.gov/news/focuson/flu/illustrations/antigenie/antigenicdrifthtm.
Accessed April 18, 2007.

National Institute of Allergy and Infectious Disease (N IAID). (2006). Flu (Influenza):
Improving the Gold (Bead) Standard for Flu Tests [WWW document]. URL,
http://www3.niaid.nih.gov/healthscience/healthtopics/Flu/Research/ongoingResearch/Arc
hive/GoldBeadStandardforFluTests.htrn. Accessed June 27, 2007 .

Neumann G. and Kawaoka Y. (2006). Host Range Restriction and Pathogenicity in the
Context of Influenza Pandemic. Emerg. Infect. Dis. 12(6): 881-886.

Neumann G., Watanabe T., Ito H., Watanabe S., Goto H., Gao P., Hughes M., Perez D.R.,
Donis R., Hoffmann E., Hobom G., and Kawaoka Y. (1999). Generation of influenza A
viruses entirely from cloned cDNAs. Proc. Natl. Acad. Sci. USA. 96:9345-9350.

 

168

Newborg, H. (2009). Swine Flu Scare: It’s All About the Adjuvant! [WWW document].
URL, http://www.infowars.com/swine-flu-scare-its-all-about-the-adjuvant]. Accessed
May 6, 2010.

Ohlson S., J ungar C., Strandh M., and Mandenius C.-F. (2000). Continuous weak-affinity
immunosensing. Trends Biotechnol. 18(2): 49-52.

Pack S.-H., Lee C.-W., Yook S.—H., Kwon O.-H., and Park Y.-N. (1999). Performance
control strategies of one-step irnmuno-chromatographic assay system for Salmonella
typhimurium. Anal. Lett. 32(2): 335—360.

Pal, S. and Alocilja, EC. (2009). Electrically active polyaniline coated Magnetic (EAPM)
Nanoparticle as Novel Transducer in Biosensor for Detection of Bacillus anthracis
Spores in Food Samples. Biosens. Bioelectron. 24(5): 1437-1444.

Pal, S., Alocilja,E.C., and Downes, RP. (2007). Nanowire Labeled Direct-Charge
Transfer Biosensor for Detecting Bacillus species. Biosensors & Bioelectronics Journal.
22:2329-2336.

 

Pal, S., Setterington, E., and Alocilja, EC. (2008). Electrically-Active Magnetic
Nanoparticles for Concentrating and Detecting Bacillus anthracis Spores in a Direct-
Charge Transfer Biosensor. IEEE Sensors Journal. 8(6): 647-654.

Pa], S., Ying, W., Alocilja,E.C., and Downes, RP. (2008). Sensitivity and Specificity
Performance of a Direct-Charge Transfer Biosensor for Detecting Bacillus cereus in
Selected Food Matrices. Biosystems Engineering Journal. 99(4): 461-468.

Park S. and Harnad-Schifferli K. (2010). Enhancement of In Vitro Translation by Gold
Nanoparticle—DNA Conjugates. ACS Nano. [Epub online].

Paul H.W., Ricco A.J., and Wrighton MS. (1985). Resistance of Polyaniline Films as a
Function of Electrochemical Potential and the Fabrication of Polyaniline-Based
Microelectronic Devices. J. Phys. Chem. 89: 1441.

Pennington J .E. (1990). Newer uses of intravenous immunoglobulins as anti- infective
agents. Antimicrob. Agents Chemother. 34: 1463-6.

Pier G.B., Thomas D., Small G., Siadak A., and Zweerink H. In vitro and in vivo activity
of polyclonal and monoclonal human immunoglobulins G, M, and A against
Pseudomonas aeroginosa lipopolysaccharide. Infect. Immun. 57: 174-9.

Pinto L.H., Holsinger L.J., and Lamb RA. (1992). Influenza virus M2 protein has ion
channel activity. Cell. 69: 517-528.

169

Poddar P., Wilson J .L., Srikanth H., Morrison S.A., and Carpenter BE. (2004). Magnetic
properties of conducting polymer doped with manganese-zinc ferrite nanoparticles.
Nanotechnology. 15(10): S570-574.

Poks P. and Grzeszczuk M. (1997). Temperature effects in thin film polyaniline
electrodes. Pol. J. Chem. 71(8): 1 140-1 150.

Portet D., Denoit B., Rump E., Lej eunne J .J ., and Jallet P. (2001). Nonpolymeric coatings
of iron oxide colloids for biological use as magnetic resonance imaging contrast agents. J.
Colloids Interface Sci. 238: 37—42.

Potyrailo RA. (2006). Polymeric sensor materials: Toward an alliance of combinatorial
and rational design tools? Angew. Chem. Int. Ed. 45: 702-723.

Pouget J .P., Jozefowicz M.E., Epstein A.J., Tang X., and Macdiarmid AG. (1991). XRay
Structure of Polyaniline. Macromolecules 24(3): 779-789.

Powell HM. and Jarnieson WA. (1939). Combined therapy of pneumococcic rat
infections with rabbit antipneumococcic serum and sulfapyridine (2-sulfanily1
aminopyridine). J. Immunol. 36:459-465.

Prince A.M., Reesink H., Pascual D., Horowitz B., Hewlett I., Murthy K.K., Cobb K.E.,
and Eichberg J .W. (1991). Prevention of HIV Infection by Passive Immunization with
HIV Immunoglobulin. AIDS Res. Hum. Retroviruses 7( 12): 971-973.

Prince 0., Hemming V., and Chanock R. (1986). The use of purified immunoglobulin in
the therapy of respiratory syncytial virus infection. Pediatr. Infect. Dis. 5: $201-$203.

Prince HE. and Leber AL (2003). Comparison of Complement Fixation and
Hemagglutinin Inhibition Assays for Detecting Antibody Responses following Influenza
Virus Vaccination. Clin. Diagn. Lab. Irmnunol. 10: 481—482.

Rabolt J .F., Clarke T.C., Kanazawa K.K., Reynolds J .R., and Street GB. (1980). Organic
Metals: Polyparaphenylene Sulfide Hexafluoroarsenate. J. Chem. Soc., Chem. Commun.
347-348.

Rackemann RM. (1942). Allergy: serum reactions, with particular reference to the
prevention and treatment of tetanus. J. Amer. Med. Assoc. 226: 726-733.

Radke, S. and Alocilja, EC. (2005). A Microfabricated Biosensor for Detecting
Foodbome Bioterrorism Agents. IEEE Sensors Journal. 5(4): 744-750.

Ramge P., Unger R.E., Oltrogge J .B., Zenker D., Begley D., Kreuer J ., von Briesen H.
(2000). Polysorbate-80 coating enhances uptake of polybutylcyanoacrylate (PBCA)-

nanoparticles by human and bovine primary brain capillary endothelial cells. Eur. J.
Neurol. 12: 1931—2934.

170

Ramisse F., Szatanik M., Binder P., and Alonso J .-M. (1993). Passive local
irnmunotherapy of experimental staphylococcal pneumonia with human intrave- nous
immunoglobulin. J. Infect. Dis. 168: 1030-3.

Ray A., Asturias G.E., Kershner D.L., Richter A.F., Macdiarmid A.G., and Epstein
A.J. (1989). Polyaniline - Doping, Structure and Derivatives. Synthetic Metals 29(1):
E141-E150.

Reed E.C., Bowden R.A., Dandliker P.S., Lilleby K.E., and Meyers J .D. (1988).
Treatment of cytomegalovirus pneumonia with ganciclovir and intravenous
cytomegalovirus immunoglobulin in patients with bone marrow transplants. Ann. Intern.

Med. 109: 783-8.

Reid A.H., Taubenberger J.K., and Fanning T.G. (2001). The 1918 Spanish influenza:
integrating history and biology. Microbes Infect. 3:81-87.

Reina J ., Padilla E., Alonso F., Ruiz De Gopegui E., Munar M., and Mari M. (2002).
Evaluation of a New Dot Blot Enzyme Immunoassay (Directigen Flu A + B) for
Simultaneous and Differential Detection of Influenza A and B Virus Antigens from
Respiratory Samples. J. Clin. Microbiol. 40: 3515—3517.

Sadik O.A., Gheorghiu E., Xu H., Andreescu D., Balut C., Gheorghiu M., and Bratu D.
(2002). Differential Impedance Spectroscopy for Monitoring Protein Immobilization and
Antibody-Antigen Reactions. Anal. Chem. 74: 3 142-3 150.

Sato T., Serizawa T., and Okahata Y. (1996). Binding of influenza A virus to
monosialoganglioside (GM3) reconstituted in glucosylceramide and sphingomyelin
membranes. Biochim. Biophys. Acta. 1285: 14—20.

Schlaes D.M., Binczewski B., and Rice LB. (1993). Emerging antimicrobial resistance
and the immunocompromised host. Clin. Infect. Dis. 17(suppl 2): S527-36.

Schofield DJ. and Dimmock NJ. (1996). Determination of affinities of a panel of IgGs
and F abs for whole enveloped (influenza A) virions using surface Plasmon resonance. J.
Virol. Methods. 62: 33—42.

Scott M.T., Spinks G.M., and Wallace G. (2005). Micro-humidity sensors based on a
processable polyaniline blend. Sens. Actuators B: Chem. 107(2): 657-665.

Schuhmann W., Larnmert R., Uhe B., and Schemidt H.L. (1990). Polypyrrole, a new
possibility for covalent binding of oxido-reductases to electrode surfaces as a base for
stable biosensors. Sensor Actuator B. 1: 537-541.

Schultze D., Thomas Y., and Wunderli W. (2001). Evaluation of an Optical

Immunoassay for the Rapid Detection of Influenza A and B Viral Antigens. Eur. J. Clin.
Microbiol. Infect. Dis. 20: 280—283.

171

 

See RH. and Chow A.W. (1989). Microbiology of toxic shock syndrome: overview. Rev.
Infect. Dis. 11(suppl 1): SSS-60.

Shaw M.W., Arden NH, and Maassab HF. (1992). New Aspects of Influenza Viruses.
Clinical Microbiology Reviews. 5(1): 74-92.

Shimasaki C.D., Achyuthan K.E., Hansjergen J .A., and Appleman J .R. (2001). Rapid
diagnostics: the detection of neuraminidase activity as a technology for high-specificity
targets. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 356: 1925—193 1.

Shinya K., Ebina M., Yamada S., Ono M., Kasai N., and Kawaoka Y. (2006). Avian flu:
influenza virus receptors in the human airway. Nature. 440: 43 5-6.

Smela E. (2003). Conjugated polymer actuators for biomedical applications. Adv. Mater.
15: 481-494.

Solvay (2010). Cost of Influenza [WWW document]. URL, http://www.solvay-
influenza.com/aboutinfluenza/costof'mfluenza/0,,2655-2-0,00.htm, Accessed May 4, 2010.

 

Sonea S., Barduas A., and F rapier A. (1958). Combined protective action of human
gamma globulin and antibiotics when administered simultaneously in experimental
staphylococcal infections. Rev. Can. Biol. 17: 1 10-1 15.

Spackman E., Senne D.A., Myers T.J., Bulaga L.L., Garber L.P., Perdue M.L., Lohman
K., Daum LT, and Suarez D.L. (2002). Development of a Real-Time Reverse
Transcriptase PCR Assay for Type A Influenza Virus and the Avian H5 and H7
Hemagglutinin Subtypes. J. Clin. Microbiol. 40(9): 3256-3260.

Stafstrom S., Bredas J .L., Epstein A.J., Woo H.S., Tanner D.B., Huang W.S., and
MacDiarmid AG. (1987). Polaron lattice in highly conducting polyaniline: Theoretical
and optical studies. Phys. Rev. Lett. 59(13): 1464-1467.

Stej skal J. and Gilbert R.G. (2002). Polyaniline. Preparation of a conducting polymer
(IUPAC Technical Report). Pur Appl. Chem. 74(5): 857-867.

Stejskal J ., Kratochvil P., and Jenkins AD. (1996). The formation of polyaniline and the
nature of its structures. Polymer. 37(2): 367-369.

Stevens J ., Blixt 0., Glaser L., Taubenberger J .K., Palese P., Paulson J .C., and Wilson
I.A. (2006a). Glycan Microarray Analysis of the Hemagglutinins from Modern and
Pandemic Influenza Viruses Reveals Different Receptor Specificities. J. Mol. Biol. 355:
1 143-1 1 5 5.

Stevens J ., Blixt 0., Paulson J .C., and Wilson l.A. (2006b). Glycan microarray

technologies: tools to survey host specificity of influenza viruses. Nature Reviews:
Microbiol. 4: 857-864.

172

 

 

Stevens J ., Blixt 0., Tumpey T.M., Taubenberger J .K., Paulson J .C., and Wilson I.A.
(2006c). Structure and Receptor Specificity of the Hemagglutinin from an H5Nl
Influenza Virus. Science. 312: 404-410.

Stratta R.J., Shaefer M.S., Cushing K.A., Markin R.S., Reed E.C., Langnas A.N., Pillen
T.J., and Shaw B.W. Jr. (1992). A randomized prospective trial of acyclovir and immune
globulin prophylaxis in liver transplant recipients receiving OKT3 therapy. Arch. Surg.
127: 55-64.

Su C.C., Wu T.Z., Chenb L.K., Yang H.H., and Tai DP. (2003). Development of
irnmunochips for the detection of dengue viral antigens. Anal. Chim. Acta. 479: 117—123. L

 

Suarez D.L., Perdue M.L., Cox N., Rowe T., Bender C., Huang J., and Swayne DE.
(1998). Comparisons of highly virulent H5N1 influenza A viruses isolated from humans
and chickens from Hong Kong. J. Virol. 72: 6678-6688.

 

Subbarao K., Klimov A., Katz J. Regnery H., Lim W., Hall H., Perdue M., Swayne D., D
Bender C., Huang J ., Hemphill M., Rowe T., Shaw M., Xu X., Fukuda K., and Cox N. '
(1998). Characterization of an avian influenza A (H5N1) virus isolated from a child with

a fatal respiratory illness. Science. 279: 393-396.

Suman S., Singhal R., Sharma A.L., Malhotra B.D., and Pundir CS. (2005).
Development of a lactate biosensor based on conducting copolymer bound lactate
oxidase. Sensor Actuator B. 107: 768-772.

Sung K.-M., Mosley D.W., Peelle B.R., Zhang S., and Jacobson J .M. (2004). Synthesis
of Monofunctionalized Gold Nanoparticles by Fmoc Solid-Phase Reactions. J. Am.
Chem. Soc. 126: 5064-5065.

 

Svorc J ., Miertus S., Katrlik J ., and Stred’ansky M. (1997). Composite Transducers for
Amperometric Biosensors. The Glucose Sensor. Anal. Chem. 69: 2086-2090.

Syritski V., Gyurcsanyi R.E., Opik A., and Toth K. (2005). Synthesis and
characterization of inherently conducting polymers by using Scanning Electrochemical
Microscopy and Electrochemical Quartz Crystal Microbalance. Synth. Met. 152 (1-3):
133-136.

Tacket C.O., Shandera W.X., Mann J .M., Hargrett N.T., and Blake RA. (1984). Equine
antitoxin use and other factors that predict outcome in type A food- borne botulism. Am.
J. Med. 76: 794-8.

Takemoto D.K., Skehel J .J ., and Wiley DC. (1996). A Surface Plasmon Resonance

Assay for the Binding of Influenza Virus Hemagglutinin to Its Sialic Acid Receptor.
Virology. 217: 452—458.

173

Talkington D.F., Schwartz 3., Black C.M., Todd J .K., Elliott J ., Breiman RR, and

F acklam RR. (1993). Association of phenotypic and genotypic characteristics of invasive
Streptococcus pyogenes isolates with clinical components of streptococcal toxic shock
syndrome. Infect. Immun. 61: 3369-74.

Taubenberger J .K., Reid A.H., Lourens R.M., Wang R., Jin G., and Farming T.G. (2005).
Characterization of the 1918 influenza virus polymerase genes. Nature. 437: 889-893.

Tumpey T.M., Maines T.R., Van Hoeven N., Glaser L., Solorzano A., Pappas C., Cox
N.J., Swayne D.E., Palese P., Katz J .M., and Garcia-Sastre A. (2007). Science. 315: 655-
659.

Tumer A.P.F., Karube 1., and Wilson G.S. (Eds), 1987. Biosensors Fundamentals and
Applications. Oxford University Press, Oxford, UK.

Uttenthaler E., Schraml M., and Mandel J ., Drost S. (2001). Ultrasensitive quartz crystal
microbalance sensors for detection of M13-Phages in liquids. Biosens. Bioelectron. 16:
73 5—743.

Vyas RN. and Wang B. (2010). Electrochemical analysis of conducting polymer thin
films. Int. J. Mol. Sci. 11: 1956-1972.

Ward MD. and Buttry DA. (1990). In Situ Interfacia] Mass Detection with Piezoelectric
Transducers. Science. 249: 1000-1007.

Watt C.H. Jr. (1978). Poisonous snakebite treatment in the United States. J. Amer. Med.
Assoc. 240: 654-6.

Webby R.J. and Webster R.G. (2003). Are We Ready for Pandemic Influenza? Science.
302(5650): 1519-1522.

Webster R.G., Guan Y., Peiris M., Walker D., Krauss S., Zhou N.N., Govorkova E.A.,
Ellis T.M., Dyrting K.C., Sit T., Perez DR, and Shortridge KR (2002). Characterization
of H5N1 influenza viruses that continue to circulate in geese in southeaster China. J.
Virol. 76: 118-126.

Weisman L.E., Cruess DR, and Fischer G.W. (1993). Standard versus hyper immune
intravenous immunoglobulin in preventing or treating neonatal bacterial infections. Clin.
Perinatol. 20: 21 1-224.

Weisman L.E., Cruess DR, and Fischer G.W. (1994). Opsonic activity of commercially
available standard intravenous immunoglobulin preparations. Pediatr. Infect. Dis. J. 13:
1122-5.

Weltzin R. and Monath T.P. (1999). Intranasal antibody prophylaxis for protection
against viral disease. Clin. Micr. Rev. 12(3): 383-393.

174

 

Wijaya A. and Hamad-Schifferli K. (2007). High-Density Encapsulation of F 6304
Nanoparticles in Lipid Vesicles. Langmuir. 23: 9546-9550.

Wiley DC. and Skehel J .J . (1987). The Structure and Function of the Hemagglutinin
Membrane Glycoprotein of Influenza Virus. Ann. Rev. Biochem. 56: 365-94.

Wilfert C.M., Buckley R.H., Mohanakumar T., Griffith J .F., Katz S.L., Whisnant J .K.,
Eggleston P.A., Moore M., Treadwell E., Oxman M.N., and Rosen F .S. (1977). Persistent

and fatal central-nervous system echovirus infections in patients with agarnma-
globulinemia. New Engl. J. Med. 296: 1485-9.

Williams A.E., PhD. (2007). Measures To Maximize Blood and Blood Component
Availability during a Pandemic. Office of Blood Research and Review (OBRR) Guidance
Concept Paper.

Wilson E.J., Medearis D.N. Jr., Hansen LA, and Rubin R.H. 9-(1-3-Dihydroxy-2-
Propoxymethyl) guanine prevents death but not immunity in murine cytomegalovirus-
infected normal and immunosuppressed BALB/c mice. Antimicrob. Agents Chemother.
31: 1017-1020.

World Health Organization (WHO). (2004, March 2). Avian influenza A(H5N1)- update
31: Situation (poultry) in Asia: need for a long-term response, comparison with previous
outbreaks [WWW document]. URL, http://www.who.int/csr/don/2004_03_02/en/.
Accessed March 24, 2007.

World Health Organization (WHO). (2006, Feb). Avian influenza (“bird flu”) — Fact
sheet [WWW document]. URL,
http://www.who.int/mediacentre/factsheets/avian_influenza/en/. Accessed March 23,
2006.

World Health Organization (WHO) Regional Office for South-East Asia. (2007).
Guidelines on Laboratory Diagnosis of Avian Influenza [WWW document]. URL,
http://www.searo.who.int/LinkFiles/CDS_CDS-Guide]ines-Laboratory.pdf. Accessed
December 10, 2009.

World Health Organization (WHO). (2008, June 19). Cumulative Number of Confumed
Human Cases of Avian Influenza A/(HSNl) Reported to WHO [WWW document]. URL,
http://www.who.int/csr/disease/avian_influenza/country/cases_table_2008_06_1 9/en/inde
x.htm]. Accessed June 26, 2008.

World Health Organization (WHO). (2009). Pandemic (HlNl) 2009 — update 69 [WW
document]. URL, http://www.who.int/csr/don/2009_10_09/en/index.htrnl, Accessed June
3, 2010.

Wright A., Shin S.-U., and Morrison S.L. (1992). Genetically engineering antibodies:
progress and prospects. Crit. Rev. Immunol. 12: 125-168.

175

 

Wright RE and Webster R.G. 2001). Orthomyxoviruses. In: Knipe D.M. and Howley
P.M. (Eds). Fields Virology, 4 ed. Lippincott Williams & Wilkins, Philadelphia: pp.
1533-1579.

Xu X. and Asher SA. (2004). Synthesis and Utilization of Monodisperse Hollow
Polymeric Particles in Photonic Crystals. J. Am. Chem. Soc. 126: 7940-7945.

Xue W.Y., Qiu H., Fang K., Li J ., Zhao J .W., and Li M. (2006). Electrical and magnetic
properties of the composite pellets containing DBSA-doped polyaniline and Fe
nanoparticles. Synth. Met. 156(11-13): 833-837.

Yang G., Cho N.-H., and Kim G.-Y. (2006). Sensing of the Insecticide Carbamate
Pesticides by Surface Plasmon Resonance. Published by the American Society of
Agricultural and Biological Engineers. 2006 ASAE Annual Meeting. Paper number
067125.

Yap PL. (1994). Does intravenous immune globulin have a role in HIV-infected patients?
Clin. Exp. Immunol. 97(suppl 1): 59-67.

Ybarra G., Moina C., Florit M.I., and Posadas D. (2000). Proton exchange during the
redox switching of polyaniline film electrodes. Electrochem. Solid-State Lett. 3(7): 330-
332.

Yoshino K., Kyokane J ., Ozaki M., Yun MS, and Inuishi Y. (1983). Effect of Double
Doping on Electrical Conductivity of Poly-p-Phenylenesulfide. Jpn. J. Appl. Phys. 22:
L289-L290.

Yuen K.Y., Chan P.K., Peiris M., Tsang D.N., Que T.L., Shortridge K.F., Cheung F.T.,
To W.K., Ho E.T., Sung R., and Cheng AF. (1998). Clinical features and rapid viral
diagnosis of human disease associated with avian influenza A H5N1 virus. Lancet. 351:
467-471.

Yuk J .S., Jin J .-H, Alocilja EC, and Rose J .B. (2009). Performance enhancement of
polyaniline-based polymeric wire biosensor. Biosens. Bioelectron. 24(5): 1348-1352.

Yun Z., Zhengtao D., J iachang Y., Fangqiong T., and Qun W. (2007). Using cadmium
telluride quantum dots as a proton flux sensor and applying to detect H9 avian influenza
virus. Analytical Biochem. 364(2): 122-7.

Zhang, D. and Alocilja, EC. (2008). Characterization of nano-porous silicon-based DNA
biosensor for the detection of Salmonella Enteritidis. IEEE Sensors Journal. 8(6): 775-
780.

Zhang W.D. and Evans DH. (1991). J. Virol. Methods. 33:165-189.

176

 

Zhang Z., Wan M., and Wei Y. (2005). Electromagnetic functionalized polyaniline
nanostructures. Nanotechnology. 16(12): 2827-2832.

Zhou B., Zhong N., and Guan Y. (2007). Treatment with Convalescent Plasma for
Influenza A (HSNl) Infection. N. Engl. J. Med. 357(14): 1450-1451.

Zhuang X. (2009). Visualizing Individual Influenza Particles in Living Cells [WWW
document]. URL, http://zhuang.harvard.edu/cellentry.html#Figure1. Accessed February
20, 2010.

Ziegler T., Katz J .M., Cox N.J., and Regnery H.L. (1997). Influenza viruses, p. 673-678.
In N.R. Rose, E.C. de Macario, J .D. Folds, H.C. Lane, and RM. Nakarnura (ed.), Manual
of clinical laboratory immunology. 5th ed. American Society for Microbiology,
Washington, DC.

Zolla-Pazner S. and Gomy MK. (1992). Passive immunization for the prevention and
treatment of HIV infection. AIDS 6: 1235-47.

177