a“
unhwa-tmrsh

«5.1 ‘30»

.. win.
I. .
ymfikfim:
hvdwvfla. .
of:§.,
..~..!a.ruh. t:

hvmfifiwmuh.

A. 4.. 10, u. L
35,. afifiia. an»...
tn§.ealh.flru

(3.

if... {Lia
5.7.3.1.?
34%..“ Am.

. - r
5n! .1 in:
sunniéflmm

\.

. {3513. {4...
1...! .411; 1'!)
if. .f

 

200%

LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

INFLAMMATION AND IDIOSYNCRATIC DRUG REACTIONS:
INFLAMMATORY MECHANISMS AND INTERACTIONS IN A
MURINE MODEL OF TROVAFLOXACIN HEPATOTOXICITY

presented by

Patrick Joseph Shaw

has been accepted towards fulfillment
of the requirements for the

 

 

Ph.D. degree in Pharmacology & Toxicology

”lbw A3231")

Major Professor’s Sipnature

?.ob~©?

Date

 

 

MSU is an affirmative-action, equal-opportunity employer

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 cJClR-C-lfiateouejndd-p. 15

H——_—.—

INFLAMMATION AND IDIOSYNCRATIC DRUG REACTIONS:
INFLAMMATORY MECHANISMS AND INTERACTIONS IN A MURINE
MODEL OF TROVAFLOXACIN HEPATOTOXICITY
BY

Patrick Joseph Shaw

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Pharmacology and Toxicology

2008

ABSTRACT
INFLAMMATION AND IDIOSYNCRATIC DRUG REACTIONS:
INFLAMMATORY MECHANISMS AND INTERACTIONS IN A MURINE MODEL
OF TROVAFLOXACIN HEPATOTOXICITY
BY

Patrick Joseph Shaw

Drug-induced liver injury is the leading cause of acute liver failure in the
United States and is a major concern for both public health and the
pharmaceutical industry. Idiosyncratic adverse drug reactions (lADRs), a rare
form of drug-induced liver injury, have been the reason for the majority of
postrnarket regulatory actions on drugs. The liver is often a target of lADRs.
lADRs are characterized by the toxicity being unrelated to the pharmacology of
the drug and do not demonstrate obvious dose or time dependence. The erratic
occurrence and lack of mechanistic evidence makes lADRs very difficult to
predict. Hepatotoxicity induced by the fluoroquinolone antibiotic trovafloxacin
(TVX) exhibited these characteristics. The mechanism underlying TVX-induced
idiosyncratic hepatotoxicity is unknown. We and others have hypothesized that
an inflammatory stress, commonplace and erratic in people, could alter the
threshold for toxicity of certain drugs precipitating an lADR.

This dissertation tested the hypothesis that an inflammatory stress could
precipitate idiosyncrasy—like TVX hepatotoxicity in mice. Administration of a
nonhepatotoxic dose of TVX 3 h before a nonhepatotoxic dose of either

lipopolysaccharide (LPS) or peptidoglycan-lipoteichoic acid mixture caused

significant heIDatocelIular necrosis and apoptosis. Levofloxacin (LVX), a
fluoroquinolone antibiotic without lADR liability in humans, did not interact with
LPS to cause hepatotoxicity. The remaining studies focused on understanding
the mechanisms underlying TVX/LPS-induced liver injury.

Gene expression analysis at a time before the onset of liver injury
segregated mice to their respective treatment groups. Therefore, gene
expression analysis was able to distinguish TVX/LPS-treated mice from all other
treatment groups.

Furthermore, LPS-induced increases in TNFa, lFNy, thrombin activation,
PAH and VEGF were enhanced by TVX. The progression of TVX/LPS-induced
liver injury was dependent on PMN activation, TNFa, IFNy, thrombin activation,
PAH and VEGF. Based on this finding, mice were killed at a time near the onset
of liver injury to explore how these mediators of inflammation interact with one
another and the cascade of events which leads to TVX/LPS-induced
hepatotoxicity. TNFa, lFNy, PAH and VEGF potentially interacted to form
several cycles of dysregulated inflammation. These potential vicious cycles of
inflammation might be involved in TVX/LPS-induced liver injury.

In summary, novel proinflammatory properties and potential cycles of
inflammation were identified which might be involved in various models of
inflammatory tissue injury. Additionally, these studies support the possibility of
predicting and identifying mechanisms underlying lADRs by utilization of a

drug/LPS coexposure model.

Copyright by
PATRICK JOSEPH SHAW
2008

 

 

ACKNOWLEDGMENTS

There are a lot of people who have helped me along the way and I would
like to thank them for their contributions. No matter how many times I edit or
rewrite this section I will not be able to express how grateful I truly am to

everyone who has assisted me.

First, and foremost, I would like to thank my mentor, Dr. Robert Roth. The
decision to come to MSU and in turn, to join Bob's laboratory has been one of the
best decisions I have ever made. As a result of working and interacting daily with
Bob, I have developed into not only a better scientist, but also a better person. I
have truly been lucky to have an advisor who has been an amazing mentor and a
great friend. There are few things I will cherish more from MSU than Friday night
happy hours for nachos and beer. I truly admire Bob’s passion for science; and
only hope that I can carry this eagerness for science with me throughout my

career.

Dr. Patricia Ganey also deserves my gratitude. Patti played a significant
role in my development as a scientist and my education in the laboratory. She
has been a mentor to me throughout my graduate career. I have been extremely
fortunate to have two truly amazing mentors to be there for me. There is really
nothing I can say or do to truly express how much I appreciate what Bob and

Patti have done for me as a scientist and a person.

 

 

In addition to Bob and Patti, there have been several people at MSU who
have helped me along the way. I owe a great deal of appreciation to my other
committee members, Dr. Lapres and Dr. Kaminski, for their continued guidance
and advice regarding my dissertation. Within the laboratory, Dr. Jane Maddox
and former research assistant Sandra Newport helped me throughout my
research. Jane and Sandy trained me and always were available for questions,
even the really dumb ones. Dr. James Luyendyk, a former graduate student in
the lab, has also played a significant role in my research. He helped train me in
the laboratory when I first joined, and put up with me as I stumbled through the
first couple months in the laboratory. Jim taught me a lot in the short time I
worked with him. His approach to research shaped my graduate career, and I am

lucky to have had the chance to work with him.

I would also like to thank the people who have either rotated or worked
with me to contribute to the research presented in this dissertation. Marie
Hopfensperger, Susan Achberger and Aaron Fullerton have all contributed to the
studies presented in this dissertation, and I appreciate their contributions.
Additionally, a dissertation is a team effort and I want to thank all of the current
and former members of the laboratory. Dr. Bryan Copple, Dr. Xiaomin Deng,
Erica Sparkenbaugh, Dr. Francis Tukov, Wei Zou, Dr. Steve Bezdecny, Christine
Dugan, Jingtao Lu, Theresa Eagle, Emilie Evenson, Dr. Sachin Devi, Nicole
Crisp and Allen Macdonald have all been of assistance to me and I am grateful

for their contributions.

The Department of Pharmacology and Toxicology also deserves
tremendous thanks. Dr. Stephanie Watts and Dr. Keith Lookingland have been
great leaders of the graduate program at MSU. Additionally, the administrative
staff within the department has been a great help and always available to assist
me. The Center for Integrative Toxicology also deserves my gratitude. Amy
Swagart and Carol Chvojka have always been helpful to me with financial and

. travel problems.

Beyond MSU, I have had the opportunity to collaborate with a number of
exceptional scientists. l have worked closely with colleagues at Abbott
Laboratories. I owe thanks to Dr. Eric Blomme, Dr. Jeff Waring, Michael Liguori
and Amy Ditewig for their assistance and collaboration. Additionally, I owe a big
thank you to Dr. Robert Stachlewitz who had me in his laboratory at Worcester,
MA for a summer. Working with Rob at Abbott Laboratories was an amazing

experience for which I am grateful.

I am extremely grateful to the Society of Toxicology for everything it has
offered me. Additionally, I owe a big thank you to the Michigan regional chapter
of SOT. My experience within SOT has been a big part of my growth as a
scientist. I have had the opportunity to work alongside exceptional scientists and
people thanks to the SOT. I especially want to thank Betty Eidemiller, who

assisted me in more ways than I can list during my time with SOT.

Additionally, I would like to thank those who provided funding for my

research. First off, thank you to Bob, who funded my studies through NIH grants.

vi

 

"“““l

In addition, I would like to thank the Center for Integrative Toxicology, MSU, SOT
and ASPET for the continued support in my graduate research and the

attendance of scientific meetings.

Finally, the only reason I was able to make this far was by the unwavering
support and love from my family and friends. I owe a big thank you to some close
friends: Brandon and Deirdre. Additionally, thank you to Sarah for being there for
me during the writing of this dissertation, and understanding when I wasn’t in the

best mood.

I have an amazing family who deserve a lot of my appreciation. My
parents have been there for me throughout, and I owe everything to them. I am
eternally grateful. They have always dropped everything to help me, and I truly
appreciate the sacrifices they made for me to get where I am today. Thank you
Mom and Dad. Additionally, my step—parents, Greg and Carol, have always been
supportive and willing to help [me out, and I am grateful. Furthermore, I would like
to thank Eric, my brother and best friend. He has always been there to hear
about the frustrations and encourage me. I wish I could put into words how much
he has helped me. I don’t want to get all emotional, but I would not be anywhere
near where I am today without my brother. I especially want to thank Eric, Mom

and Dad, without whom none of this would have been possible.

vii

TABLE OF CONTENTS

LIST OF TABLES .......................................................................... xiii
LIST OF FIGURES ......................................................................... xiv
LIST OF ABBREVIATIONS .............................................................. xviii
CHAPTER 1
General Introduction and Specific Aims ....................................................... 1
1.1 Idiosyncratic adverse drug reactions ...................................... 2
1.1.1 Overview of idiosyncratic adverse drug reactions ....... 2
1.1.2 Hypothetical mechanisms of idiosyncratic adverse
drug reactions .............................................................. 4
1.2 Inflammation ........................................................................... 20
1.2.1 Overview of inflammatory stress .................................. 20
1.2.2 Tumor necrosis factor a (TNFa) .................................. 27
1.2.3 Neutrophils ................................................................... 30
1.2.4 Interferon y (IFNy) ........................................................ 31
1.2.5 The hemostatic system ................................................ 34
1.2.6 Vascular endothelial growth factor (VEGF) ................. 38
1.3 Trovafloxacin-induced idiosyncratic liver injury ...................... 43
1.3.1 Overview of trovafloxacin ..................................... 43

1.3.2 Trovafloxacin-induced hepatotoxicity in people ..........44
1.3.3 Interaction between trovafloxacin and

inflammation ...................................................... 46

1.4 Hypothesis and specific aims ......................................... 48

1.5 Overview and significance of dissertation .......................... 50

CHAPTER 2

Lipopolysaccharide and trovafloxacin coexposure in mice causes

idiosyncrasy-Iike liver injury dependent on tumor necrosis factor-alpha ..... 52

2.1 Abstract ..................................................................... 53

2.2 Introduction ................................................................ 54

2.3 Materials and methods .................................................. 55

2.3.1 Materials ........................................................... 55

2.3.2 Animals ............................................................. 55

2.3.3 Experimental protocol .......................................... 56

2.3.4 Histopathology ................................................... 57

2.3.5 TNFa analysis ................................................... 57

2.3.6 Statistical Analyses ............................................. 57

2.4 Results ...................................................................... 58

2.4.1 Dose-response and timecourse of liver injury ............ 58

2.4.2 Comparison of trovafloxacin and levofloxacin ............ 58

2.4.3 Timecourse of TNFa concentration in plasma ........... 63

2.4.4 Pentoxifylline study .............................................. 68

viii

2.4.5 Etanercept inhibition of TNFa activity ...................... 73
2.5 Discussion .................................................................. 80

CHAPTER 3
Trovafloxacin enhances the inflammatory response to a gram- -negative or
a gram- positive bacterial stimulus, resulting in CD18-dependent liver

injury in mice ................................................................................. 86
3.1 Abstract ..................................................................... 87
3.2 Introduction ................................................................ 89
3.3 Materials and methods .................................................. 90

3.3.1 Materials ........................................................... 90
3.3.2 Animals ............................................................. 90
3.3.3 Experimental protocols ........................................ 90
3.3.4 Histopathology ................................................... 91
3.3.5 Cytokine measurements ....................................... 91
3.3.6 Neutrophil staining .............................................. 91
3.3.7 Statistical Analyses ............................................. 92
3.4 Results ...................................................................... 93

3.4.1 TVX coexposure with either LPS or PGN-LTA

causes hepatotoxicity .......................................... 93
3.4.2 TVX enhances cytokine induction by either

inflammatory stimulus .......................................... 93
3.4.3 Effect of TVX on microbial stimuli-induced hepatic

neutrophil accumulation ........................................ 103

3.4.4 Effect of CD18 neutralization on TVX/LPS- and
TVX/PGN-LTA-induced liver injury and inflammation ..103

3.5 Discussion .................................................................. 1 13
CHAPTER 4
TNFa acting at both p55 and p75 receptors is essential for synergistic
hepatotoxicity from TVX/LPS coexposure ........................................... 119
4.1 Abstract ..................................................................... 120
4.2 Introduction ................................................................ 121
4.3 Materials and methods .................................................. 123
4.3.1 Materials ........................................................... 123
4.3.2 Animals ............................................................. 123
4.3.3 Experimental protocols ....................................... 123
4.3.4 Histopathology ................................................... 124
4.3.5 Cytokine measurements ....................................... 124
4.3.6 Neutrophil staining .............................................. 124
4.3.7 Hemostatic system measurements ......................... 124
4.3.8 Statistical analyses ............................................. 124
4.4 Results ...................................................................... 126
4.4.1 - p55"' and p75"' mice are protected from TVX/LPS-
induced liver injury .............................................. 126

4.4.2 TNFa neutralization attenuates TVX/LPS-induced

ix

inflammatory cytokines and chemokines .................. 126
4. 4. 3 TVX/LPS-induced hepatic neutrophil accumulation

is independent of TNFa ........................................ 131
4. 4. 4 TNFa neutralization attenuates TVX/LPS-induced
hemostatic system activation ................................. 131
4.4.5 TVX and TNFa coexposure causes hepatotoxicity ..... 138
4.5 Discussion .................................................................. 143

CHAPTER 5
Coexposure of mice to trovafloxacin and lipopolysaccharide, a model of
idiosyncratic hepatotoxicity, results in a unique gene expression profile

and interferon gamma-dependent liver injury ....................................... 148
5.1 Abstract ..................................................................... 149
5.2 Introduction ................................................................ 150
5.3 Materials and methods .................................................. 152
5.3.1 Materials ........................................................... 152
5.3.2 Animals ............................................................. 152
5.3.3 Experimental protocols ......................................... 152
5.3.4 ALT activity and histopathology .............................. 153
5.3.5 RNA isolation ..................................................... 153
5.3.6 Gene array analysis ............................................ 153
5.3.7 Hepatic neutrophil accumulation ............................ 154
5.3.8 Plasma cytokine measurements ............................ 154
5.3.9 Statisitical methods ............................................. 155
5.4 Results ...................................................................... 156
5.4.1 Development of hepatocellular injury after TVX/LPS,
but not LVX/LPS coexposure ................................. 156
5.4.2 Hepatic global gene expression changes before the
onset of hepatotoxicity .......................................... 156
5.4.3 Gene expression changes within TVX/LPS, TVXNeh
and Veh/LPS treatment groups ............................... 161
5.4.4 Several gene expression changes in TVX/LPS-
treated mice are involved in interferon signaling ........ 161
5.4.5 Timecourse of plasma concentrations of IL-18
and IFNy ........................................................... 171
5.4.6 TVX/I7PS- induced hepatocellular Injury in
lL-18"' mice ....................................................... 171
5.4.7 TVX/LPS- induced liver Injury in lFNy"' mice .............. 171
5.4.8 TVX/LPS-induced hepatic neutrophil accumulation
and proinflammatory cytokines in IF Ny”' mice ........... 176
5.5 Discussion .................................................................. 183
CHAPTER 6
The role of the hemostatic system in a murine model of idiosyncratic liver
injury induced by trovafloxacin and lipopolysaccharide coexposure ........... 200
6.1 Abstract ..................................................................... 201

6.2
6.3

6.4

6.5

CHAPTER 7

Introduction ................................................................ 203
Materials and methods .................................................. 205
6.3.1 Materials ........................................................... 205
6.3.2 Animals ............................................................. 205
6.3.3 Experimental protocols ......................................... 205
6.3.4 Heparin treatment ................................................ 205
6.3.5 Histopathology .. ............................................ 206
6.3.6 Hemostatic system measurements ......................... 206
6.3.7 Cytokine measurements ....................................... 206
6.3.8 Statisitical analyses ............................................. 206
Results ...................................................................... 207
6.4.1 Changes in the hemostatic system in TVX/LPS-

induced liver injury ............................................... 207
6.4.2 Anticoagulant heparin protects from TVX/LPS-

induced liver injury ............................................... 212
6.4.3 PAl-1"' mice are protected from TVX/LPS-

induced liver injury ............................................... 219
6.4.4 Role of coagulation system activation and PAl-1 in

TVX/LPS-induction of cytokines .............................. 219
Discussion .................................................................. 228

Vascular endothelial growth factor has a proinflammatory role which
is critical to trovafloxacin and lipopolysaccharide coexposure-

induced liver injury ......................................................................... 233
7.1 Abstract ..................................................................... 234
7.2 Introduction ................................................................ 235
7.3 Materials and methods .................................................. 236

7.3.1 Materials ........................................................... 236
7.3.2 Animals ............................................................. 236
7.3.3 Experimental protocols ......................................... 236
7.3.4 Histopathology .................................................... 237
7.3.5 Cytokine measurements ....................................... 237
7.3.6 Neutrophil staining ............................................... 237
7.3.7 Hemostatic system measurements ......................... 237
7.3.8 Statisitical analyses ............................................. 237
7.4 Results ...................................................................... 238

7.4.1 Timecourse of plasma concentration of VEGF ........238
7.4.2 Effect of VEGF neutralization on TVX/LPS-

induced hepatotoxicity .......................................... 238
7.4.3 Effect of VEGF neutralization on plasma
concentrations of cytokines ................................... 245
7.4.4 Effect of VEGF neutralization on TVX/LPS-induced
. hepatic neutrophil accumulation .............................. 245
7.4.5 Effect of VEGF neutralization on the hemostatic
system .............................................................. 245

xi

7.5 Discussion .................................................................. 255

CHAPTER 8

Summary and conclusions ............................................................... 259
8.1 Summary of research .................................................... 260
8.2 Proposed pathways of TVX/LPS-induced liver injury ........... 266
8.3 Major findings and implications ....................................... 273
8.4 Knowledge gaps and future studies ................................. 276

REFERENCE LIST ......................................................................... 278

xii

Table 1.1

Table 1.2

Table 1.3

Table 1.4

Table 1.5

Table 5.1

Table 5.2

Table 5.3

Table 6.1

LIST OF TABLES

Concordance of LPS/drug exposure in rats for lADR-
causing drugs in humans ............................................... 19

The effects of TNFa on other mediators of liver injury .......... 29

The effects of IF Ny on other mediators of inflammatory
liver injury .............................................................................. 33

The effects of the hemostatic system and hypoxia on
mediators of inflammatory liver injury ............................... 37

The effects of VEGF on mediators of inflammatory
liver injury ................................................................... 41

Pathways highly affected by the 142 functional genes
selectively changed by TVX/LPS-treatment compared to
all other treatment groups .............................................. 166

Genes selectively altered in expression by TVX/LPS
that are regulated by interferons ..................................... 169

Supplemental table: Functional genes selectively
changed by TVX/LPS-treatment compared to all
other treatment groups .................................................. 192

Scorin of histopathology of livers from wild-type and
PAl-1' ‘ mice cotreated with TVX/LPS ............................... 222

xiii

LIST OF FIGURES

Images in this dissertation are presented in color

Fig. 1.1 Hypothetical relationship between inflammation and

drug idiosyncrasy ......................................................... 1 7
Fig. 1.2 Toll-like receptor 4 signaling pathways ............................. 23
Fig. 1.3 A summary of the inflammatory process ........................... 26
Fig. 2.1 Dose response and development of liver injury from

TVX/LPS cotreatment in mice ......................................... 60

Fig. 2.2 Comparison of TVX/LPS with LVX/LPS coadministration ...... 62

Fig. 2.3 Liver histopathology in mice cotreated with LPS and

either TVX or LVX ........................................................ 65
Fig. 2.4 Timecourse of TNFa concentration in the plasma of

treated mice ................................................................ 67
Fig. 2.5 The effect of pentoxifylline (PTX) on LPS-induced TNFa

expression and TVX/LPS-induced liver injury ..................... 70
Fig. 2.6 Effect of PTX on TVX/LPS-induced liver pathology ............. 72

Fig. 2.7 The effect of etanercept on TVX/LPS-induced TNFa
expression and liver injury ............................................. 75

Fig. 2.8 Effect of etanercept on TVX/LPS-induced liver pathology ..... 77

Fig. 2.9 Effect of etanercept treatment given at the peak of
plasma TNFa on TVX/LPS-induced liver injury .................. 79

Fig. 3.1 Development of liver injury after TVX/LPS or
TVX/PGN-LTA coexposure ............................................ 95

Fig. 3.2 Histopathology of livers from mice treated with TVX/LPS
or TVX/PGN-LTA ......................................................... 97

Fig. 3.3 Effect of TVX pretreatment on LPS- and PGN-LTA-
induced increases in cytokines ....................................... 99

Fig. 3.4 Effect of TVX pretreatment on LPS- and PGN-LTA-
induced increases in chemokines .................................... 102

xiv

 

 

Fig. 3.6

Fig. 3.?

its.
”is.

Fig.5.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.
Fig.
Fig.
Fig.
Fig.

Fig.
Fig.

Fig.

3.5

3.6

3.7

3.8

4.1

4.2

4.3

4.4

4.5

4.6
4.7
5.1
5.2
5.3

5.4
5.5

Effect of TVX pretreatment on LPS- and PGN-LTA-
induced hepatic neutrophil accumulation ........................... 105

Effect of CD18 neutralization on TVX/LPS- and
TVX/PGN-LTA-ind uced liver injury .................................. 107

Effect of CD18 neutralization on TVX/LPS— and
TVX/PGN-LTA-induced hepatic neutrophil accumulation ...... 109

Effect of CD18 neutralization on TVX/LPS- and
TVX/PGN-LTA-induced cytokine increases ....................... 111

The role of TNF receptors in TVX/LPS-induced
liver injury ................................................................... 128

Effect of TNFa inhibition on TVX/LPS-induced
increases in plasma cytokines ........................................ 130

Effect of TNFa inhibition on TVX/LPS-induced
increases in plasma chemokines ..................................... 133

Effect of TNFor inhibition on TVX/LPS-induced hepatic
PMN accumulation ....................................................... 135

Effect of TNFa inhibition on hemostatic system

dysregulation mediated by TVX/LPS coexposure ............... 137
TVX/TNFa coexposure-induced liver injury ........................ 140
Histopathology of TVX/TNFa-induced liver injury ................ 142
Development of TVX/LPS-induced liver injury .................... 158

Hierarchical clustering of hepatic gene expression profiles 160
Venn diagram depiction of probeset regulation of

TVXNeh-, Veh/LPS- and TVX/LPS-treated mice relative

to VehNeh controls ...................................................... 163
Timecourse of lL-18 and IF Ny plasma concentrations .......... 173

TVX/LPS-induced liver injury is attenuated in lL-18"' mice 175

lFNy"‘ mice are resistant to TVX/LPS-induced liver injury ..... 178

Fig.

Fig.

Fig.
Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.
Fig.

Fig.

Fig.

Fig.

Fig.

5.7

5.8

5.9

6.1

6.2

6.3

6.4

6.5

6.7

7.1
7.2

7.3

7.4

7.5

7.6

TVX/LPS-induced hepatic PMN accumulation is
unchanged in IFNy"‘ mice ............................................... 180

Plasma concentrations of proinflammatory cytokines
are reduced in lFNy"' mice ............................................. 182

Hypothesized role of lFNy in TVX/LPS-induced liver injury 188

Effect of fluoroquinolones and LPS on coagulation
system activation ......................................................... 209

Effect of fluoroquinolones and LPS on plasma
concentration of active PAl-1 .......................................... 211

Effect of fluoroquinolones and LPS on hepatic sinusoidal
fibrin deposition ............................................................ 214

Effect of heparin treatment on TVX/LPS—induced hepatic
fibrin deposition and liver injury ....................................... 216

Histopathology of heparin-treated and PAI-1"' mice
treated with TVX/LPS ................................................... 218

Effect of PAH deficiency on TVX/LPS-induced hepatic
fibrin deposition and liver injury ....................................... 221

Effect of heparin treatment or PAl-1 deficiency on
TVX/LPS-induced increases in plasma cytokines ................ 225

Timecourse of VEGF plasma concentration ....................... 240

Effect of VEGF neutralization on TVX/LPS-induced
plasma VEGF induction and liver injury ............................ 242

Protection from TVX/LPS-induced liver pathology by
VEGF neutralization ..................................................... 244

Effect of VEGF neutralization on TVX/LPS-induced
increase of cytokines .................................................... 247

VEGF neutralization did not affect TVX/LPS-induced
increases in chemokines ............................................... 249

Effect of VEGF neutralization on TVX/LPS-induced
hepatic neutrophil accumulation ...................................... 251

xvi

Fig. 7.7

Fig. 8.1

Fig. 8.2

The role of VEGF in TVX/LPS-induced hemostatic
system activation ......................................................... 254

Proposed pathways to TVX/LPS-induced liver injury ............ 268

Possible proinflammatory cycles induced by
TVX/LPS coexposure ....... * ............................................ 272

xvii

ALT
ANOVA
ATP
cAMP
CYP
DAG
DlLI
DNA
egr—1
Erk
FAK
FDA
Flk-1
Flt-1
HIF-1or
HPC
lADR
IFNy
IFNGR1
IFNGR2
IKK-i
IL-1
IL-4
IL—6
lL-1O
lL-12
lL-15
IL-18
lP-1 0
IP3
IRAK
IRF3
Jak1
JNK
KC
LBP
LPS
LTA
LVX
MAPK
MOP-1
MHC
MlP-1or
MlP-2

LIST OF ABBREVIATIONS

alanine aminotransferase

analysis of variance

adenine triphosphate

cyclic adenosine monophosphate
cytochrome P450
sn-1,2-diacylglycerol

drug-induced liver injury
deoxyribonucleic acid

early growth response-1
extracellular regulated kinase
focal adhesion kinase

US. Food and Drug Administration
VEGF receptor 2

VEGF receptor 1
hypoxia-inducible factor a
hepatocyte

idiosyncratic adverse drug reaction
interferon y

interferon 7 receptor 1

interferon y receptor 2

le kinase-i

interleukin 1

interleukin 4

interleukin 6

interleukin 10

interleukin 12

interleukin 15

interleukin 18

lFNy-inducible protein 10

inositol (1 ,4,5)-triphosphate
lL-1-receptor—associated kinase
interferon regulatory factor 3

janus activated kinase 1
c-jun-N-terminal kinase
keratinocyte chemoattractant
LPS-binding protein
lipopolysaccharide

lipoteichoic acid

levofloxacin

mitogen-activated protein kinase
monocyte chemoattractant protein-1
majOr histocompatibility complex
macrophage inflammatory protein 1a
macrophage inflammatory protein 2

xviii

MPO
MyDBB
NFKB
NK cells
NKT cells
p55
p75
PAl-1
PAR-1
PBMC
PGN

Pl
PI3K
PIP2
PKC
PLCy
PMN
PTX
RIP
ROS
Sck
SODZ
STAT1
TAB-2
TACE
TAK1
TAT
TBK1
TCR
TIRAP
TLR
TNFa
t-PA
TRAF6
TRAM
TRIF
TVX
u-PA
VEGF
VEGFR
VRAP

myeloperoxidase

myeloid differentiation factor 88
nuclear factor-KB

natural killer cells

natural killer T cells

TNF receptor 1

TNF receptor 2

plasminogen activator inhibitor 1
protease activated receptor-1
peripheral blood mononuclear cell
peptidoglycan

pharmacologic interaction
phosphoinositide 3-kinase
phosphatidylinositol (4,5)-biphosphate
protein kinase C

phospholipase C y

neutrophil

pentoxifylline

receptor-interacting protein

reactive oxygen species

Shc-Iike protein

superoxide dismutase 2

signal transducer and activator of transcription 1
TAK1 binding protein
TNFoI-converting enzyme
transforming growth factor-b-associated kinase
thrombinzantithrombin

TANK binding kinase 1

T cell receptor

Toll/lL-1 R domain-containing adapters
toll-like receptor

tumor necrosis factor a

tissue plasminogen activator

tumor necrosis factor-associated factor 6
TRlF-related adapter molecule
TlR-containing adapter molecule
trovafloxacin

urokinase plasminogen activator
vascular endothelial growth factor
VEGF receptor

VEGF receptor-associated protein

xix

CHAPTER 1

General Introduction and Specific Aims

1.1 Idiosyncratic adverse drug reactions

1.1.1 Overview of idiosyncratic adverse drug reactions

Adverse drug reactions are a serious problem for not only the public
health, but also for pharmaceutical companies and drug-regulatory agencies. In a
study in the United Kingdom, adverse drug reactions accounted for more than
6% of hospital admissions. Of these admissions due to adverse drug reactions,
the mortality rate was 2% (1). In addition to the risk to public health, adverse drug
reactions are a major issue for drug development. A significant amount of time
and money is expended in the effort to predict the risk of adverse reactions from
drug candidates. Despite comprehensive preclinical drug testing and clinical trials,
over 10% of drugs approved during 1975-2000 were either withdrawn from the
market or have been highly restricted in use (2). In 1998, the pharmaceutical
industry spent over 20 billion dollars on drug discovery and development, with
screening assays and toxicity testing accounting for about 20% of the total
amount spent (3). Despite such extensive efforts, in 1999 over 258,000 post—
marketing adverse events were reported in the United States, suggesting that
this is a persistent major issue (4).

Adverse drug reactions can occur in any number of tissues, but the liver is
often the target organ. Of the 28 drugs removed from the US market between
1976 and 2005, 6 were withdrawn due to hepatotoxicity (5). Drug-induced liver
injury (DILI) accounts for more than 50% of acute liver failure cases (6). It is

associated with significant mortality; therefore, a number of drugs which have

been associated with DILI have been removed from the market. For example,
bromfenac (7), troglitazone (8, 9) and tienilic acid (10) have been completely
removed from the market due to hepatotoxicity. In addition, hepatotoxicity of
other drugs such as trovafloxacin (I'VX) (11, 12), nefazodone (13), and
nevirapine (14, 15) has led to “black box” warnings limiting their use. DILI is the
leading cause for the withdrawal of drugs from the market by either the US.
Food and Drug Administration or pharmaceutical companies (16).

An important subset of adverse drug reactions which cause DILI are
idiosyncratic adverse drug reactions (lADRs), which account for 13-17% of all
cases of acute liver failure (6, 17). lADRs typically occur in a small fraction of
people (generally < 1%) within the range of doses used clinically. The exact
mechanisms underlying lADRs are unknown but typically do not involve the
pharmacological properties of the drug. In addition, lADRs lack an obvious dose-
dependence, meaning that a dose which causes toxicity in some patients does
not in others. Another characteristic of lADRs is that the onset of toxicity relative
to the duration of drug therapy is variable. Finally, there is a wide range in the
severity of the reactions depending on the drug and individuals.

Despite extensive research, animal models do not exist which reproduce
the hepatotoxicity caused by lADRs. The development of animal models is
necessary to predict those drugs which cause lADRs and to decrease human
suffering. A predictive animal model would be beneficial for several reasons.
Prediction of drug candidates that could cause lADRs would prevent their

development into marketed pharmaceuticals and thereby reduce risk to public

health. In addition, it would prevent pharmaceutical companies from sending
such candidates to clinical trials or to market and would thereby save money
spent on clinical trials, marketing and potential lawsuits from patients affected by
lADRs.

Drugs which lead to lADRs are usually not identified in preclinical testing
due to their typically rare occurrence and the use of relatively small numbers of
animals in toxicity testing. The inability of animal tests to predict lADRs may be
due, in part, to the reaction being idiosyncratic in animals as well as humans, and
thus an extremely large number of animals would be needed to detect toxicity. It
has been estimated that to predict an lADR confidently, toxicity testing would
require 30,000 animals to be treated (18). In addition, the current animal testing
paradigms might not include sufficient biological diversity to elucidate lADR
toxicities. Since such large studies are not possible for drug candidates, it is
critical that the modes of action of lADRs are better understood to develop

predictive models.

1.1.2 Hypothesized mechanisms of idiosyncratic adverse drug reactions
Despite extensive research, the mechanisms underlying lADRs remain
poorly understood and incompletely characterized. There exist several obstacles
to understanding lADRs. A substantial challenge is that an animal model for the
early detection of hepatic lADRs is currently unavailable. In addition, the tissue
from afflicted individuals is often difficult to obtain for research purposes,

although the DILI network is trying to address this obstacle. Even when tissue

from affected individuals is available, the tissue would have been harvested long
after injury developed and is likely, therefore, to be of limited value for
mechanistic studies. However, despite such limitations and difficulties, progress
has been made in understanding ‘ lADRs. Such progress has led to the
development of several diverse theories about lADR pathogenesis. To this point,
none of the hypotheses to explain lADR pathogenesis have been proved or
disproved. The prevalent hypotheses to explain lADR toxicity and supporting

experimental evidence are described in more detail below.

Reactive Intermediate Hypothesis

One theory for the mechanism of lADRs is that a drug is metabolized into
a reactive metabolite, which might bind with important cellular proteins, damage
membrane integrity, alter calcium homeostasis or other intracellular signaling in
ways which could lead to toxicity and that susceptible individuals have
polymorphisms in the bioactivating enzyme(s) (19). Indeed, there are several
cases in which a drug linked with lADRs has the ability to form a metabolite
which is reactive (20). The reactive intermediate hypothesis can be closely
associated with all of the hypotheses to be described, especially if a reactive
metabolite and not the parent drug is the agent involved in the toxicity.

Troglitazone is an antidiabetic drug which was linked with serious
idiosyncratic hepatotoxicity (9). Research conducted after troglitazone was
removed from the market showed that it is metabolized in the rat to five

intermediates with the ability to form glutathione conjugates that appear in bile

(21). In addition, metabolic activation by cytochrome P450 3A4 (CYP3A4) forms
reactive metabolites which bind to proteins and nucleophiles (22). Whether these
form protein adducts that play a role in toxicity is unknown. In addition, if the
protein adducts are formed and involved in toxicity, the degree of protein adducts
that constitute a threshold for troglitazone lADRs is unknown. Furthermore,
several drugs which form reactive metabolites are not associated with an
increased risk of lADRs (23). Moreover, one would expect an “intrinsic” (dose-
related) toxicity picture in the absence of some metabolism-related sensitivity
factor that renders a small fraction of patients susceptible to lADRs. Thus,
although the reactive intermediate hypothesis is a reasonable one, a causal link
between reactive metabolite generation and hepatotoxicity has not been

established conclusively for drugs that cause lADRs.

Genetic Polymorphism Hypothesis

A related theory is that genetic polymorphisms among individuals can
cause differences in the toxic responses of individuals to drugs. Many
polymorphisms can lead to drug metabolism differences among individuals,
leading to differences in pharmacokinetics and reactive intermediate formation
(24). Human polymorphisms in genes encoding cytochrome P450 drug
metabolizing enzymes have been identified and could lead to differences in drug
metabolism and clearance that could render some individuals more susceptible

to toxicity. In addition, it is possible that a polymorphism in drug metabolizing

enzymes might lead to the formation of a reactive intermediate not seen in the
majority genotype.

Alternatively, a genetic polymorphism in a protective gene, such as an
anti-inflammatory cytokine, might render individuals more susceptible to normally
nontoxic doses of drugs, resulting in an lADR. For this hypothesis to explain
lADRs, the genetic polymorphism of people on drug therapy would have to be as
rare as the lADR itself or the lADR would have to be a result of a rare
combination of several more common polymorphisms. Even if this explains the
rarity of lADRs, the genetic polymorphism hypothesis does not explain other
characteristics of lADRs such as the variability in the onset of toxicity.

An example often referenced by supporters of the importance of
polymorphisms is toxicity caused by isoniazid, a first-line drug used in the
prevention and treatment of tuberculosis. Isoniazid has been linked to several
cases of liver injury (25). The susceptibility of individuals to isoniazid-induced
liver injury has been linked to a polymorphism resulting in a rapid acetylator
phenotype (25, 26). It was hypothesized that the rapid acetylators produce more
of a reactive metabolite which causes hepatocellular necrosis. However, several
epidemiological studies failed to find an association between the rapid acetylation
polymorphism and liver injury (27). Another example of this hypothesis is evident
from a study in which individuals were treated with the idiosyncratic drug,
diclofenac. It was found that individuals who developed a toxic response had a
greater rate of polymorphisms in the interleukin 10 (IL-10) and interleukin 4 (IL-4)

genes than the group of individuals who did not develop a toxic response to

diclofenac (28). An association between lADRs and genetic polymorphisms does
exist with some drugs; however, their roles remain uncertain, and it remains likely

that other factors play a role in precipitating lADRs.

Hapten Hypothesis

A widely accepted theory to explain lADRs is that they result from an
adaptive immune response. Some clinical characteristics of lADRs such as the
delayed onset of toxicity, the lack of a simple dose-response relationship and
eosinophilia have led some to postulate that lADRs are mediated by adaptive
immunity (29). This has led to the formation of two related hypotheses. The
hapten hypothesis states that a chemically reactive drug or a reactive metabolite
binds to an endogenous protein. This protein adduct is then seen as a foreign
antigen capable of initiating immunological recognition (30). According to this
hypothesis, the drug-modified protein must be processed by antigen-presenting
cells and presented to T cells. This results in sensitization of the T cells to the
foreign antigen. The immune system develops memory to the foreign antigen,
and upon subsequent exposure to the drug, robust immune system activation
occurs, resulting in the formation of autoantibodies and/or the activation of
cytotoxic T cells targeting self proteins (31). It is important to understand that
both sensitizing and challenging exposures are required in this hypothetical
mechanism.

In support of the hapten hypothesis, the presence of autoantibodies has

been detected in patients with hepatic lADRs after exposure to several drugs,

including diclofenac, troglitazone, halothane and tienilic acid (28, 32). The study
which found autoantibodies in the sera of patients who experienced diclofenac
hepatotoxicity also reported the presence of autoantibodies in some patients
treated with diclofenac who did not develop hepatotoxicity (28). Such a finding
was also found in halothane-treated patients, in whom autoantibodies were found
whether they developed toxicity or not (33). Thus, from these reports, a clear
cause and effect relationship between autoantibodies and idiosyncratic
hepatotoxicity is lacking. The clinical evidence supporting the role of the adaptive
immune system may in some cases be explained by immune system activation
occurring secondary to tissue damage. Efforts have been undertaken to show the
involvement of the specific immune system in hepatotoxic lADRs; however, in all
of the current animal models of drug immunogenicity, an adaptive immune
response was detected in the absence of liver damage (34). Accordingly,
experimental support for this hypothesis is incomplete, and an animal model of

drug hepatotoxicity with an adaptive immune mechanism has not emerged so far.

The danger hypothesis

A theory closely related to the hapten hypothesis described above is the
danger hypothesis, which proposes that a damaging immune system activation
occurs only if the drug binds to a protein which causes some type a of a stress
response, such as inflammation or cell death, resulting in a ‘danger’ signal (35).
Thus, according to the danger hypothesis, the formation of a drug-protein adduct

is insufficient to cause injury, a secondary signal during sensitization such as

mild cell death or cytokine release then results in adaptive immune system
activation and pathogenesis (36). It has been postulated that reactive drug
metabolites themselves could cause this danger signal, and this is what
determines which reactive metabolites lead to lADRs (37, 38). However, the
‘danger’ signal could be from a number of independent factors including an

infection causing an innate immune response, resulting in an inflammatory stress.

The pharmacological interaction (PI) hypothesis

The PI hypothesis is closely related to the hapten and danger hypotheses,
in that it suggests an active role for the adaptive immune system in the
development of lADRs. The PI hypothesis proposes that drugs bind reversibly to
the major histocompatibility complex (MHC) and T cell receptor (T CR) complex. It
is hypothesized that the drug then acts like a superantigen to elicit an adaptive
immune system response, precipitating an lADR (39). Much of the early work
leading to the development of the Pl hypothesis was done with
sulfamethoxazole; which caused proliferation of T cells isolated from
sulfamethoxazole lADR patients (40). However, there is no evidence that an
lADR drug binding to the MHC-TCR complex is capable of eliciting an immune
response. The role as a possible superantigen to the MHC:TCR complex has not
not been shown with any other drugs linked with hepatotoxic lADRs. In addition,
evidence is also lacking in support of a causal link between an adaptive immune

response and the precipitation of a hepatotoxic lADR.

1O

Mitochondrial dysfunction hypothesis

Another hypothesis for lADRs is that mitochondrial dysfunction and
disturbances in mitochondrial integrity by oxidative stress are an underlying
cause. Mitochondria play a critical role in providing the cell with energy,
controlling the process of apoptosis and regulating intracellular oxidative stress.
Mitochondrial dysfunction can encompass several changes such as decreased
adenosine triphosphate (ATP) production, mitochondrial reactive oxygen species
(ROS) production or depolarization of the mitochondrial membrane potential.

One way in which mitochondrial dysfunction can occur is through DNA
alteration. Mitochondrial DNA alterations which could result in dysfunction are
rare but are seen in humans. It was found in a epidemiological study that >12 in
100,00 people either had mitochondrial DNA disease or were at risk to develop it;
these results reflect the minimum prevalence of mtDNA disease and pathogenic
mtDNA mutations (41). It is hypothesized that either a mitochondrial disease or
polymorphism could alter mitochondrial function and render cells sensitive to a
drug, resulting in idiosyncratic toxicity (42). It is also postulated that genetic or
acquired mitochondrial abnormalities can lead to silent and gradually
accumulating mitochondrial injury which reaches a threshold and abruptly
triggers liver injury (43).

There is extensive evidence linking lADR drugs with mitochondrial
alterations. Troglitazone, tolcapone, diclofenac, valproic acid, and isoniazid are
some of the drugs which cause lADRs and which have mitochondrial liability in

hepatocytes (43-48). In addition, diclofenac and troglitazone are cytotoxic to

11

HepGZ cells through a mitochondrial mechanism (49, 50). In one study,
superoxide dismutase 2 ($002) heterozygote mice, a model of silent
mitochondrial abnormality, were chronically treated with troglitazone. This
treatment had no effect on wild-type mice but resulted in hepatocellular necrosis
in 8002*" mice (51). However, the hypothesis fails to explain the apparent lack
of dose dependence that characterizes lADRs. In addition, there are several
drugs that cause mitochondrial alterations in vitro but have not resulted in
adverse drug reactions in people.

It is of importance to note that mitochondrial dysfunctions can be induced
by a number of independent factors such as xenobiotics which might be taken
concurrently, hypoxia or inflammation. Therefore, it is possible that alterations in

mitochondrial function play a role in other hypothesized mechanisms of lADRs.

Failure to adapt hypothesis

Another hypothesis of lADRs is that a small fraction of people develop
minor liver toxicity in response to a drug. Most of these individuals “adapt” and
experience a resolution of liver injury even in the continued presence of the drug.
However, it is proposed that a small fraction of these people fail to “adapt”, and
the injury progresses to overt toxicity (52). Reports of isoniazid hepatotoxicity
seem to support this theory, inasmuch as 15% of patients taking isoniazid
experience minor alanine aminotransferase (ALT) elevations, but less than 1%

develop symptomatic hepatitis with continued treatment (53).

12

The mechanisms underlying the “adaptation” phenomenon are unknown.
Adaptation may not be recognized in clinical trials because drug treatment is
stopped when the serum ALT activity rises to greater than 3 times the upper limit
of normal, making it impossible to diStinguish between patients who would and
would not adapt. In addition, there are currently few animal models in which
adaptation can be studied. However, future studies made possible by the DILI
Network will attempt to address these issues and determine possible reasons for
increased susceptibility of certain individuals to lADRs. It is also of importance to
note that the ‘failure to adapt’ hypothesis does not discount other hypotheses of
lADRs, as toxicity may be due to any number of mechanisms to which certain

individuals cannot adapt and therefore experience an lADR.

Multiple determinant hypothesis

The multiple determinant hypothesis proposes that idiosyncratic reactions
are the result of multiple, discrete but necessary factors or processes all
occurring simultaneously (54). Each factor has an independent probability of
occurring, but all of them are required to precipitate an lADR, thus accounting for
the rare occurrence rate. According to the hypothesis, an idiosyncratic reaction
would only occur in an individual if all the critical steps occur within an
appropriate time. An equation for the probability of an idiosyncratic reaction is
proposed below:

PIADR = Pchem X Pexp X IDenv X Pgenei Where,

13

Pma is the probability of an lADR, Pchem is the probability contributed by
chemical properties, Pexp is the probability determined by the drug exposure to
the critical organ(s), Pam, represents probabilities determined by environmental
factors (drug coexposure, inflammation, etc.) and Pgene is the probability related
to genetic factors (54).

The multiple determinant hypothesis is a rather general and
encompassing hypothesis which takes into account the other hypotheses
mentioned above. However, it is important to understand in more detail the
mechanistic aspects of lADRs to develop predictive animal models. Inasmuch as
environmental and genetic factors might play a role in the probability of a specific
drug causing an lADR, it is important to determine which factors are important to
toxicity and why.

The hypothesis implies that an underlying factor has the potential to lower
the toxicity threshold of a drug, rendering a normally therapeutic dose toxic.
Several factors have the potential to affect the susceptibility of an individual to
drug toxicity including age, gender, coexposure to other pharmacological agents,

drug metabolism differences, and state of health.

Inflammatory stress hypothesis
In the multiple determinant hypothesis, one environmental factor that
might render an individual sensitive to a normally nontoxic drug dose is

inflammatory stress. This idea has led to the inflammatory stress hypothesis,

14

which states that an episode of inflammation has the potential to interact with
concurrent drug therapy to precipitate an lADR.

Inflammatory episodes are commonplace in people and occur erratically
throughout life. Many are modest enough that they go unnoticed. A hypothetical
relationship between inflammation and lADRs is illustrated in Fig. 1.1. For
therapeutically useful drugs, the pharmacologic effect is seen at much smaller
doses than signs of toxicity. Most drugs are developed so that the range between
a therapeutic dose and the smallest toxic dose (ie., the therapeutic window) is as
large as possible. As dose is increased, toxicity is seen (such as kidney toxicity in
Fig. 1.1) and death ensues at large doses. Liver toxicity in this example is not
observed because the toxicity threshold lies at doses higher than those that are
lethal. The hypothesis is that a modest inflammatory stress can decrease the
threshold for hepatic toxicity, thereby shrinking the therapeutic window and
resulting in a toxic response at a normally safe and pharmacologically effective
dose of the drug. In this case, an lADR would occur at a dose which is nontoxic
to individuals not experiencing a concurrent inflammatory episode. The erratic
nature of inflammatory episodes can explain the unpredictable nature of lADRs.

Lipopolysaccharide (LPS), a component of gram-negative bacterial cell
walls, is one agent that can induce an inflammatory stress as described in more
detail below. Experimental models have been developed in which nontoxic doses
of lADR-causing drugs are rendered hepatotoxic upon coexposure to a nontoxic
dose of LPS. For example, rats became susceptible to hepatotoxicity from

several drugs known to cause lADRs when they were concurrently exposed to a

15

 

Fig. 1.1. Hypothetical relationship between inflammation and drug
idiosyncrasy. Drug A is a relatively safe and efficacious drug. The asterisk
indicates the usual therapeutic dose. The safety margin between
pharmacological effect and kidney toxicity is quite large. A modest inflammatory
response shifts the threshold for liver toxicity and precipitates and idiosyncratic

response (55).

16

< 9.5 no econ

 

\

.xmhdozu

  
 

. .6...
58m 355.

 

 

esuodsau

17

nontoxic dose of LPS. Drugs known to cause lADRs in humans such as
trovafloxacin, ranitidine, sulindac, chlorpromazine and diclofenac were all
rendered hepatotoxic to rats when coupled with a nontoxic dose of LPS (56-60)
(Table 1.1). Drugs in the same pharmacologic class which were not associated
with lADRs in humans were used when available. These drugs not associated
with human lADRs did not interact with inflammatory stress to cause
hepatotoxicity in animal models (58, 59).

Of the drugs tested, only the ones linked with lADRs in humans interacted
with a concurrent inflammatory stress to cause hepatotoxicity in rats. This
concordance suggested a potential role for inflammation in the mechanism of
human lADRs. The results in animal models suggest that an inflammatory
episode caused by LPS or other factors could render an individual susceptible to
hepatotoxicity at normally nontoxic drug doses, thus causing an idiosyncratic
reaction. A challenge still lies in understanding mechanisms of the hepatotoxicity
observed with coexposure to LPS and an lADR-causing drug. The remainder of
the Introduction and subsequent chapters of the thesis will explore inflammatory
stress in greater detail and present work to develop and explore an

inflammation/drug interaction model of TVX toxicity in mice.

18

Table 1.1. Concordance of LPS/drug coexposure model in rats for lADR-

causing drugs in humans

 

 

Linked to
. . LPS/drug coexposure
Drug hepatotoxrcIty in _
hepatotoxm to rats?

humans?
Trovafloxacin Yes Yes
Levofloxacin No No
Chlorpromazine Yes Yes
Ranitidine Yes Yes
Famotidine No No
Sulindac Yes Yes
Diclofenac Yes Yes

 

19

1.2 Inflammation

1 .2.1 Overview of inflammatory stress

Inflammation is an innate immune system process critical for the host’s
defense against infection and foreign substances. The inflammatory response is
a complex process encompassing the recruitment of cells, release of cytokines
and other biologically active mediators, vasodilation, hemostatic system
activation and complement activation. The magnitude of an inflammatory
response depends on the cause and varies from one individual to the next.
Modest inflammatory episodes occur sporadically and are commonplace in
people. Inflammation occurs in response to a number of stimuli including tissue
injury, microbial pathogens and other foreign substances.

As mentioned above, recognition of microorganisms by various cell types
within the body induces an inflammatory response. Components of gram-
negative bacteria have been measured In the plasma of individuals and are
increased by conditions such as gastrointestinal disturbances, alcohol
consumption, surgery, alterations in diet, etc. (55, 61). In turn, a great deal of
inflammation research has focused on host responses to gram-negative bacterial
cell wall constituents. Endotoxin is a component of gram-negative bacterial cell
walls and is released when bacteria undergo cell division or are damaged by
antibiotics (62). A major, biologically active component of endotoxin is LPS.
Chapter 3 presents some studies exploring the interaction between TVX and

gram-positive bacterial cell wall components peptidoglycan and lipoteichoic acid,

20

which can also induce inflammation. However, the majority of the work will
explore in detail the interaction between TVX and LPS. The mechanism by which
LPS induces an inflammatory response will described below.

Toll-like receptors (TLRs) are conserved pattern recognition receptors that
recognize bacterial components (63). The effects of LPS are elicited primarily
through the activation of TLR4. LPS-binding protein and CD14 are required for
presentation of LPS to TLR4; and the interaction of the co-receptor MD-2 with
dimerized TLR4 is required to elicit activation of TLR4 by LPS (64, 65). After LPS
activates TLR4, the resulting responses can be divided into those dependent on
myeloid differentiation factor 88 (MyDBB) and those independent of MyD88. The
signaling pathways activated by TLR4 activation by LPS are described in more
detail below and summarized in Fig. 1.2.

The activated TLR4 dimer recruits Toll/lL-1R domain-containing adapters
(T IRAP), TlR-containing adapter molecule (TRlF) and TRIP-related adapter
molecule (T RAM) (66). TIRAP recruitment and activation results in My088
recruitment. My088 is an adapter protein which activates inflammatory signaling
pathways. The activation of My088 leads to the recruitment and phosphorylation
of members of the lL-1-receptor-associated kinase (IRAK) family (67).
Phosporylated IRAK then dissociates from My088 and interacts with tumor
necrosis factor receptor-associated factor 6 (TRAF6) (68). Activated TRAF6
associates with TAK-1 binding protein-2 (TAB-2), causing activation of
transforming growth factor-B-associated kinase 1 (TAK1). TAK1 is a mitogen-

activated protein kinase (MAPK). At this point, TAK1 activates the p38 MAPK

21

Fig. 1.2. Toll-like receptor 4 signaling pathways. Schematic summary of TLR4
signaling following activation by LPS. LPS-binding protein (LBP) and soluble
CD14 assist in the presentation of LPS to the TLR4 dimer. The TLR4 dimer is
associated with MD-2 in the cell membrane. Upon LPS binding, Toll/lL-1R
domain-containing adapter (T IRAP) and TRIF-related adapter molecule (T RAM)
are recruited to the intracellular domain of TLR4. TIRAP recruitment and
activation allows for myeloid differentiation factor 88 (MyD88) recruitment and
activation. IL-1-receptor-associated kinase (IRAK) binds to the activated MyD88
and is in turn activated. IRAK activation allows for the binding and subsequent
activation of tumor necrosis factor receptor-associated factor 6 (TRAF6). TRAF6
binds to a complex formed by TAK-1 binding protein-2 (T AB-2) and transforming
growth factor-B-associated kinase 1 (T AK1). The binding of TRAF6 to the TAB-
2fl'AK1 complex results in TAK1 phosphorylation. TAK1 is a mitogen-activated
protein kinase (MAPK) which in turn activates p38 and JNK MAPK pathways and
NF-KB. The recruitment of TRAM also results in MyDBB-independent signaling.
Toll/IL-1 R domain-containing-containing adapter molecule (TRIF) binds to TRAM
and becomes activated. TRIF then serves as an adapter molecule for the
activation of TRAF6, again leading to MAPK and NF-KB activation. TRIF binds to
the TANK binding kinase1 (T BK1)/IKB kinase-l (lKK-i) complex which is believed
to phosphorylate and activate TBK1. TBK1 phosphorylates and activates the
transcription factor interferon regulatory factor 3 (IRF3), the activation of which

results in interferon a and [3 (IFNa/B) expression.

22

 

 

 

 

W088
IRAK:
TRAF6
TRAF6
2 IRF3
TAK1
NF-KB
p38 MAPK JNK MAPK NF {B MAPKs IFNalp

23

pathways, c-jun-N-terminal kinase (JNK) MAPK pathways, and the inflammatory
transcription factor, nuclear factor-KB (NF-KB) (69). NF-KB, p38 and JNK
activation leads to gene/protein expression of several mediators of inflammation
and cell death (70).

In addition, LPS activation of TLR4 results in a MyD88-independent
response via the recruitment of TRAM and TRIF (also known as TICAM1) (66).
TRIF then can directly activate TRAF6 with the help of receptor-interacting
protein (RIP), resulting in MAPK and NF-KB activation independent of My088. In
addition, TRIF can bind to the TANK binding kinase 1 (T BK1)/II<B kinase-i (lKK-i)
complex (71). TRIF binding of the TBK1/lKK-i complex is believed to
phosphorylate and activate TBK1. Activated TBK1 directly activates interferon
regulatory factor 3 (IRF3), a transcription factor the activation of which results in
interferon or and [3 expression (72).

TLR4 is a conserved receptor expressed on a number of cell types
including macrophages, endothelial cells, neutrophils, dendritic cells and
platelets; therefore, these cell types are activated following LPS activation of the
signaling pathways described above. TLR4 activation causes NFKB and MAPK
pathway activation, which in turn activates other transcription factors such as
early growth response-1 (egr-1). The cellular activation and transcription factor
activation by LPS leads to the release of cytokines, neutrophil proteases,
vascular endothelial growth factor, coagulation system activation and
complementactivation. This collection of factors induced by LPS can have

several profound effects such as hemostasis, edema, altered blood flow, microbe

24

Fig, 1.3. A summary of the inflammatory process. LPS activates TLR4,
leading to transcription factor activation as described in Fig. 1.2. The signal
transduction following TLR4 activation causes the activation of several cell types,
resulting in several inflammatory processes including the production and release
of cytokines: interleukin 1 (IL-1), interleukin 6 (IL-6), interferon y (lFNy), lL-10 and
tumor necrosis factor a (T NFa). In addition to the production of cytokines, cellular
activation will result in neutrophil protease release, vascular endothelial growth
factor (VEGF) release, ROS production, coagulation system activation and
complement activation. These inflammatory processes can induce several
several physiological changes such as hemostasis, edema, altered blood flow,
microbe destruction or host cell injury. The degree of inflammatory episode then
dictates the result. If modest, the inflammatory episode could be inconsequential,
beneficial or could increase the sensitivity of the host to another insult. However,
a severe inflammatory episode has the potential to cause direct tissue injury to

host

25

TLR4

Transcription factor activation
(NFch, egr-1, etc.) and
genelprotein expression

 

6

21mm:
macrophages. endothelial cells,
neutrophils, dendritic cells, platelets

ll

cytokines (IL-1, IL-6, IFN7, IL-10, TNFa, etc.)
proteases (elastase and cathepsln G)
vascular endothelial growth factor (VEGF)
reactive oxygen species (ROS)
activated coagulation system
activated complement

l

hemostasis, edema,
altered blood flow, microbe
destruction, host cell Injury, etc.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

modest severe
lnconsequentlal, beneficlal
or Increased sensitivity “35‘” injury

 

 

 

 

26

destruction or host cell injury; but can also feedback to cause more cellular
activation of macrophages, endothelial cells, neutrophils, dendritic cells, platelets
and other cell types. Fig. 1.3 illustrates a simplified version of the inflammatory
stress induced by LPS exposure. '

The magnitude and location of the inflammatory episode dictate its
ultimate effects. If the inflammatory episode is severe and uncontrolled, it can
lead to host tissue injury. However, inflammation is essential for the defense of
an organism against foreign pathogens. If the inflammatory episode is modest it
can lead to one of several consequences: it might not have any effect, it might be
beneficial by killing a foreign organism, or the inflammatory stress could sensitize
the tissue to another insult potentially resulting in injury (55, 73). Based on this
latter condition, we have hypothesized that an inflammatory stress induced by
LPS could precipitate a toxic response to a therapeutic and normally nontoxic
dose of a drug. The following sections will discuss the role of selected
inflammatory factors in liver injury and their interactions with other inflammatory

mediators.

1.2.2 Tumor necrosis factor a (TNFa)

TNFa is a pleiotropic cytokine that induces a number of cellular responses
that include cell proliferation, production of inflammatory mediators, upregulation
of adhesion molecules and programmed cell death. It is a key mediator of
inflammatory responses, including both tissue damage and host defense

mechanisms (74, 75). TNFor plays a critical role in several models of

27

hepatotoxicity including endotoxemia, viral hepatitis, acetaminophen
hepatotoxicity and ischemia/reperfusion (76-79).

The main cellular sources of TNFa production are macrophages, but
several other cell types produce ‘TNFa including mast cells, endothelial cells,
stellate cells, fibroblasts and neuronal cells (80, 81). TNFa is produced as a
transmembrane protein, which is biologically active, but it may also be released
as a soluble form via proteolytic cleavage by TNFoI-converting enzyme (TACE)
(80). Large amounts of TNFa are produced in response to TLR activation by
microbial products, including LPS. Resident liver macrophages, ie, Kupffer cells,
are a major source of TNFa production and release following TLR4 activation by
LPS in liver (82).

The biological effects of TNFa are elicited via two high affinity cell surface
receptors, p55 (TNF-R1) and p75 (T NF-R2) (83). The two TNF receptors are
structurally similar but functionally very different. Expression of p55 is found in a
wide variety of mammalian cell types, whereas the expression of p75 is typically
found only on immune cells (80). Signaling through the p55 receptor is the key
mode of TNFa signaling in most cell types. The cells of the lymphoid system are
the exception, in which signaling through the p75 receptor plays a major role.

The intracellular domains of p55 and p75 are the main difference between
the two receptors. The intracellular domain of the p55 receptor contains a death
domain, which couples the receptor’s activation to caspase activation and cell
death (84). The p75 receptor lacks the death domain. However, both receptors

recruit members of the TRAF family when activated. Activation of TRAF proteins

28

Table 1.2. The effects of TNFa on other mediators of liver injury

 

Mediator of liver injury

TNFa effects

 

Neutrophils (PMNs)

 

 

Induces PMN proliferation or apoptosis
depending on environment and concentration of
TNFq (85)

Induces activation of respiratory burst by PMNs
via both the p55 and p75 receptor or
potentiates effects of PMN stimuli (86)
Increases PMN adhesion molecules on
endothelial cells leading to increased “rolling”

(37)

 

IFNy

Induces Th1 response leading to lFNy

production (88)

 

Hemostatic system

Increases tissue factor expression on
endothelial cells, promoting coagulation system
activation (89)

Stimulates lL-6 production, which causes new
platelet formation with increased prothrombotic

activity (90)

 

 

VEGF

 

Increases VEGF expression (91)
Synergistically enhances VEGF activation of

egr-1 (92)

 

29

 

leads to MAPK activation and NF-KB activation, as described earlier. Ligand
activation of the receptors is another functional difference; membrane bound
TNFa has the ability to activate both p55 and p75 receptors (80, 93), whereas
soluble TNFa only activates the p55 receptor and is the dominant signal for p55
activation (94).

The role of each receptor has been studied in several models of liver
injury. The p55 receptor has been studied more extensively in hepatotoxicity, and
evidence has emerged for critical roles in endotoxemia, acetaminophen and
carbon tetrachloride (77, 95, 96). In contrast, roles for both receptors have been
demonstrated only in a few models of hepatotoxicity, such as from concanavilin A,
Pseudomonas aeruginosa exotoxin A and adenovirus (97-99).

As summarized in Table 1.2, TNFor can modulate several inflammatory
factors that contribute to various models of liver injury. The role of TNFa and the
effects of TNFa on these inflammatory factors will be explored in an
inflammation/drug interaction model of TVX toxicity in mice in subsequent

chapters.

1.2.3 Neutrophils

Neutrophils (PMNs) are a major component of the innate immune system.
They are recruited to inflammatory sites and are capable of phagocytosing and
enzymatically digesting microbes. They circulate in the blood and are recruited to
the site of invading microorganisms or dead/dying cells. Proinflammatory signals

of microorganisms or dead/dying cells cause the upregulation of selectin

30

adhesion molecules on both PMNs and endothelial cells. The selectins cause
PMNs to bind to and “roll” along endothelial cells. In response to inflammatory
cytokines and chemokines, PMNs then increase expression of integrins, i.e.
CD11b/CD18, on their cell surface. The expression of integrins causes PMNs to
extravasate from the blood and into the tissue. Chemokines released from
homeostatically altered cells can contribute to this process. The extravasation of
PMNs into the tissue is critical for their antimicrobial function and for cytotoxicity
to host cells (100). lnfiltrated PMNs can become activated by a number of
mediators that cause release of their granules, which contain serine proteases
(such as elastase and cathepsin G), myeloperoxidase (MPO), defensins and
vascular endothelial growth factor (VEGF). Although PMNs are critical to host-
defense against pathogens, the proteases and reactive oxygen species released
from activated PMNs can cause tissue injury. Indeed, PMNs play a major role in
several models of liver injury including endotoxemia, ischemia/reperfusion and

LPS/ranitidine interaction (101-103).

1.2.4 Interferony(IFNy)

IFNy is a type II interferon and is integral to both the innate and adaptive
immune responses. It plays a crucial role in innate immune host defense
mechanisms by inducing neutrophil activation and activating macrophage
functions such as phagocytosis, respiratory burst and cytokine secretion (104). In
addition, lFNy plays a critical role in the adaptive immune system host defense by

promoting secretion of lgGZA antibodies by B cells, antigen presentation,

31

induction of Th1 cell differentiation and maturation of T cells (104, 105). The
importance of lFNy in host defense is demonstrated by the increased
susceptibility of IFNy"' mice to a variety of infections (106). However, IFNy is also
an important mediator of inflammatory injury, such as septic shock (107).

The cellular sources of lFNy are T cells, dendritic cells, natural killer (NK)
cells, and natural killer T (NKT) cells. The control of lFNy production in these cell
types is mediated by a variety of factors. NK, NKT and dendritic cells can be
stimulated to produce IFNy by cytokines such as interleukins 12,15 and 18 (IL-12,
lL-15, and lL-18) by activated macrophages and dendritic cells (108, 109). In
addition, the production of IFNy by T cells is induced by activation of the T cell
receptor complex and interaction of C028 with B7 proteins (109).

The biological effects of lFNy are mediated through the lFNy receptor,
which is composed of two integral membrane proteins, IFNGR1 and IFNGR2
(110). The lFNy receptor is expressed on nearly all cell types. IFNGR1 plays an
important role in ligand binding, ligand trafficking through the cell and signal
transduction; whereas IFNGR2 plays only a minor role in ligand binding but is
required for signaling (110-112). Both subunits are required for eliciting the
effects of IFNy. The intracellular domains of IFNGR1 and IFNGR2 have binding
sites for janus activated kinase 1 (Jak1) and Jak2, respectively. Jak1 and Jak2
bind to the intracellular domains in unstimulated cells. IFNy binds to IFNGR1
thereby generating binding sites for IFNGR2 (113). This oligomerization of the

subunits leads to transphosphorylation and activation of Jak1 and Jak2 (114).

32

Table 1.3. The effects of IFNy on other mediators of inflammatory liver

injury

 

Mediator of liver
injury

IF Ny effects

 

Neutrophils (PMNs)

Induces the expression of a wide variety of
chemokines and adhesion molecules (115, 116)
Enhances the oxidative burst of activated

neutrophils (117-119)

 

TNFa

 

Leads to TNFa production and release from
PMNs (120)
Induces TNFa production itself and enhances

the LPS-induction of TNFa in Kupffer cells (121)

 

Hemostatic system

 

Enhances LPS-induction of tissue factor (122)
Dampens the fibrinolytic response of endothelial

cells to TNFa (123)

 

 

VEGF

 

Induces VEGF release in monocytes and
macrophages (124)
Induces initial VEGF production, but dampens

VEGF production at later times (125)

 

33

 

The activated Jaks then phosphorylate an important tyrosine on IFNGR1,
creating a docking site for signal transducer and activator of transcription 1
(STAT1) (126). STAT1 is then phosphorylated and in turn activated by the JAKs.
Once phosphorylated, STAT1 dislocates from the receptor and translocates to
the nucleus. STAT1 then binds specific promoter sites in lFNy-inducible genes,
resulting in increased expression (127). Thus, IFNy stimulation leads to the
expression of a number of antitumor, proapoptotic, and proinflammatory genes.
lFNy has a number of immunoregulatory functions, generally as a
proinflammatory cytokine. IFNy plays a special role in the liver, where there is an
abundance of NK and NKT cells in addition to the presence of conventional T
cells. NK and NKT cells represent about 25-40% of isolated liver leukocytes
(128-130). It is thus not surprising that lFNy plays a role in several models of liver
injury, including endotoxemia-, acetaminophen- and concavalin A-induced
hepatitis (115, 131, 132). In several models, lFNy has profound effects on other
mediators of liver injury such as PMNs, TNFor, the hemostatic system, and VEGF.
Some of these proinflammatory effects of lFNy are summarized in Table 1.3. A
comprehensive analysis of the effects of lFNy in a single model of inflammatory
injury has not been done, and this would provide a better understanding of the

role of lFNy in liver injury.

1.2.5 The hemostatic system
The hemostatic system encompasses a complex interaction between

platelets, blood vessels, procoagulant factors, coagulation inhibitors and

34

fibrinolytic factors. The two functional arms, coagulation and fibrinolysis, exist in a
delicate balance. The maintenance of such balance is critical to prevent blood
loss from tissue and control fibrin deposition to maintain appropriate blood flow to
tissue. If the hemostatic system is not tightly regulated, blood flow to tissues is
interrupted resulting in tissue ischemia. The significance of the balance of the two
arms is illustrated by the high sensitivity of organs to ischemia/reperfusion insult
(76,133,134)

Coagulation system activation can be initiated by both the extrinsic and
intrinsic pathways. The activation of blood coagulation occurs predominantly
through the extrinsic pathway, beginning with tissue factor expression. However,
the result of both pathways is the cleavage of prothrombin to active thrombin,
also called factor II. Thrombin activation is regulated by the endogenous inhibitor,
antithrombin III, which binds to and inactivates it. Antithrombin III binding to
thrombin forms thrombinzantithrombin (TAT) dimers which can be measured in
the plasma as a marker of coagulation system activation.

Thrombin is a protease with a major role in the vascular homeostasis. It
plays an important role in fibrin deposition in several ways. Most importantly, it
cleaves fibrinogen to fibrin monomers, which polymerize to form fibrin clots (135).
In addition, thrombin positively feeds back on the coagulation pathway by
activating coagulation factors XI, VIII, and V, which in turn can generate more
thrombin (136-138). Also, thrombin elicits diverse biological effects through
cleavage and activation of protease activated receptor-1 (PAR-1) (139, 140).

PAR-1 is highly expressed on platelets which become activated and aggregate

35

following thrombin cleavage of PAR-1 (141). Some of the effects of PAR-1
activation are independent of the hemostatic system; for example, PAR-1
activation can stimulate mast cells and monocytes to release proinflammatory
cytokines such as lL-1, lL-6 and TNFa (142, 143).

The active dissolution of fibrin clots is critical to maintaining normal blood
perfusion of organs; this is controlled by the fibrinolysis arm of the hemostatic
system. Fibrin clots are cleaved and dissolved by the serine protease plasmin
(135). Plasmin is synthesized in an inactive form, plasminogen, by the liver and
secreted into the plasma. Tissue plasminogen activator (t-PA) and urokinase
plasminogen activator (u-PA) convert plasminogen to plasmin. The enzymatic
activity of t-PA and u-PA is inhibited by plasminogen activator inhibitors 1 and 2
(PAH and PAl-2) (144).

PAl-1 is the major endogenous downregulator of fibrinolysis. PAl-1 is
mainly produced by endothelium, but can also be released by other cell types
such as adipose tissue. In addition to its role as a plasminogen activator inhibitor,
PAl-1 has several proinflammatory properties such as enhancing PMN activation,
inducing TNFa production and increasing VEGF expression (145-147). Thus, an
increase in active PAI-1 can increase inflammation and decrease the formation of
plasmin, thereby impairing the lysis of fibrin clots.

If the balance of the two functional components of hemostasis is altered, a
possible outcome is unregulated activation of the hemostatic system, which could
lead to fibrin deposition and occlusive fibrin clots. These clots have the potential

to impair local blood flow and result in tissue hypoxia (103). By these and

36

Table 1.4. The effects of the hemostatic system and hypoxia on mediators

of inflammatory liver injury

 

Mediator of liver injury Hemostatic system and hypoxia effects

 

0 Primary hepatocytes are sensitized to PMN

elastase-mediated cell death in hypoxic

 

conditions (103)

l . . .
Neutrophils (PMNs) . o Hypoxra causes PMN chemokine production by
hepatocytes (148)

o PAl-1 potentiates LPS-induced neutrophil

 

activation (146)

 

- Hypoxemia causes a significant increase in

 

 

 

 

TNF“ macrophage release of TNFa (149)
. Hypoxia enhances mature dendritic cell and
IFNy macrophage lFNy production in response to
LPS or lL-18 (150)
o Hypoxia induces VEGF expression through
hypoxia inducible factor 0t (HlF-1a) activation
VEGF (151)
o PAI-1 increases VEGF expression (145)

 

 

37

 

perhaps other mechanisms, the hemostatic system contributes to liver injury in
endotoxemia, acetaminophen hepatotoxicity and lschemia/reperfusion as
examples (103, 152-154). The hemostatic system and hypoxia could play a role
in liver injury via interactions with various inflammatory mediators, as
summarized in Table 1.4. However, interactions between hemostasis and
inflammation are still not fully understood. In addition to its interactions with
inflammation, hypoxia itself has the ability to cause cell injury directly (155). The
role of the hemostatic system and its effects on these inflammatory factors will be
explored in an inflammation/drug interaction model of TVX toxicity in mice in
subsequent chapters.

One consequence of tissue hypoxia is the stabilization of hypoxia-
inducible transcription factors, which translocate to the nucleus where they
associate with other factors, bind to transcriptional regulatory elements of DNA
and initiate the transcription of genes encoding proteins involved in adaptation to
hypoxia and cell death. One of these proteins is vascular endothelial growth

factor.

1.2.6 Vascular endothelial growth factor (VEGF)

The VEGF family describes splice variants VEGF-A, VEGF-B, VEGF-C,
VEGF-D and placental growth factor. The best studied VEGF variant is VEGF-A,
therefore, VEGF will refer to VEGF-A. It is a cytokine best known for its function
as a stimulator of developmental, adaptive and pathological angiogenesis (156-

159). However, beyond its role in vessel development, VEGF also stimulates

38

differentiation, survival, migration, proliferation, tubulogenesis and vascular
permeability in endothelial cells (157, 160, 161). The importance of VEGFs and
VEGF receptors (VEGF Rs) has been shown in gene targeting studies, where the
removal of a single VEGF allele results in embryonic lethality (156, 162). In
addition to its role in angiogenesis and development, VEGF is involved in the
inflammatory response, which will be discussed in more detail below.

VEGF is produced by endothelial cells, macrophages, activated T cells,
and variety of other cell types as a result of multiple stimuli. As mentioned above,
one potent stimulus of VEGF production is a low oxygen environment, which
results in hypoxia inducible factor a (HIF-1or) stabilization and expression of
VEGF (163). The production of VEGF can also be stimulated by a number of
inflammatory cytokines including IL-1B, IL-10L, IL—6, oncostatin M, TNFa and lL-8
(164-169).

Members of the VEGF family act on tyrosine kinase receptors. The
members of the VEGF family show different affinities for each of these receptors.
VEGF acts through specific binding to two receptors, VEGFR-1 (Flt-1) and
VEGFR-2 (Flk-1). Flt-1 is the only VEGF receptor expressed on monocytes and
macrophages, whereas endothelial cells and hematopoietic stem cells express
both Flt-1 and Flk-1 (170). Flk-1 is considered to be the major mediator of the
effects of VEGF on endothelial cells. Ligand binding of VEGF to the extracellular
domain of the receptors causes dimerization and autophosphorylation of
intracellular tyrosine residues. Several proteins bind to the phosphorylated

tyrosine residues and are activated, including VEGFR-associated protein (VRAP),

39

Shc-like protein (Sck) and phospholipase y (PLCy) (171-173). PLCy activation
leads to the hydrolysis of phosphatidylinositol (4,5)-biphosphate (PIP2) creating
the second messengers sn-1,2-diacylglycerol (DAG) and inositol (1,4,5)-
triphosphate (IP3). IP3 binds to a specific receptor on the endoplasmic reticulum
to release stores of intracellular Ca2+, whereas DAG is an activator of protein
kinase C (PKC). Active PKC initiates a signaling cascade resulting in
extracellular regulated kinase (Erk) activation and transcription of several genes.
In addition, VEGFR activation results in the activation of several other signaling
proteins including phosphoinositide 3-kinase (PI3K), focal adhesion kinase (FAK)
and p38 MAPK (174, 175). The mechanisms by which VEGFR activates these
proteins are still unknown.

Proinflammatory effects of VEGF have been the focus of recent research.
A number of VEGF’s effects are mediated through endothelial cells. Indeed, the
discovery of VEGF arose from its ability to increase microvascular permeability of
endothelial cells (176). In addition, VEGF activates endothelial cells to induce the
production of several chemokines and adhesion molecules (177, 178). It has
profound effects on several mediators of inflammation as summarized in Table
1.5. In accordance its proinflammatory effects, VEGF is involved in the
development of liver injury in animal models of endotoxemia and
ischemia/reperfusion (179, 180). The role of VEGF in rat models of drug/LPS-
induced liver injury has not been examined, but VEGF and Flt-1 mRNA are
selectively upregulated in LPS/ranitidine-treated rats (181). The selective

increase in VEGF and Flt-1 was only seen in the group which developed liver

40

Table 1.5. The effects of VEGF on mediators of inflammatory liver injury

 

Mediator of liver injury VEGF effects

 

o Induces adhesion molecules on endothelial

 

Neutroph'“ (PMNSI cells and modulates PMN trafficking (180)

 

o VEGF blockade in a murine model of sepsis

TNF“ reduces plasma TNFa concentration (179)

 

- Enhances the induction of chemoattractant

IF"? lFNy-inducible protein 10 (lP-10) by IFNY (182)

 

 

- Induces tissue factor expression which can

Hemostatic system activate the coagulation system (183)

o Induces PAl-1 in endothelial cells (184)

 

 

41

 

injury, suggesting that VEGF signaling might play a role in LPS/ranitidine-induced
hepatotoxicity. The mechanisms by which VEGF might participate in the
development of hepatotoxicity remain unclear.

As summarized in Table 1.5, VEGF affects several inflammatory factors
that play critical roles in various models of liver injury. The role of VEGF in the
development of liver injury and the effects of VEGF on these inflammatory factors
will be explored in an inflammation/drug interaction model of TVX toxicity in mice

in subsequent chapters.

42

1.3 Trovafloxacin-induced idiosyncratic liver injury

1.3.1 Overview of trovafloxacin

TVX is an example of a drug linked to human idiosyncratic hepatotoxicity
which was not predicted by preclinical testing or evident in clinical trials. The
antibacterial mechanism of action of quinolones is through inhibition of type II
topoisomerase, bacterial gyrase. This enzyme is critical for bacterial DNA
replication, recombination and repair. Interference with bacterial gyrase results in
the arrest of bacterial cell growth. TVX is a broad-spectrum fluoroquinolone
antibiotic with excellent gram-negative activity and a greater spectrum of
bactericidal activity against gram-positive pathogens compared to other
fluoroquinolones. In addition to its increased bactericidal spectrum, TVX has high
bioavailability after oral dosing and a relatively long elimination half-life, allowing
for once-daily dosing. TVX has excellent penetration into various tissues, with
highest concentration in the liver, spleen, kidney and lung and the lowest
concentration in the brain (185). These qualities made it an extremely attractive
drug.

The bactericidal activity of TVX is exerted by the parent compound, not by
a metabolite. Approximately 50% of a dose of TVX in humans is recovered
unchanged in the feces (43%) and urine (6%). For the portion of TVX that is
metabolized, phase II conjugation, specifically glucuronidation, plays a major role
(186). The metabolism and clearance of TVX is unique compared to other

fluoroquinolones. Conjugative metabolism is common for fluoroquinolones, but

43

oxidative metabolism is the predominant biotransforrnation, which is not seen
with TVX. Urinary excretion plays a major role in clearance of most
fluoroquinolones (187), but not for TVX. The main route of TVX excretion is
biliary; only a small amount undergoes renal excretion or oxidative metabolism
(188).

Recent metabolic studies using a model cyclopropylamine-containing
surrogate molecule for TVX suggested that TVX might be metabolized to a
reactive otB-unstaurated aldehyde (189). The reactive intermediate was only
found in the presence of Cyp1A2 or MP0. MP0 is released exclusively by
activated neutrophils, suggesting that TVX might be metabolized differently in a
pro-inflammatory environment to a reactive species that might contribute to TVX

toxicity.

1.3.2 Trovafloxacin-induced hepatotoxicity in people

Trovafloxacin was released in February, 1998, as a potent new
fluoroquinolone antibiotic. In 1999, reports surfaced linking TVX with
hepatotoxicity, which led to severe restriction of TVX uses and prescription.
Before restrictions were placed on TVX usage, approximately 2.5 million
prescriptions were filled. A total of 140 severe hepatic reactions were reported,
making the incidence of TVX-induced severe hepatic reactions approximately 1
in 18,000 prescriptions. Of these 140 severe hepatic reactions, 14 cases led to
liver failure, ie, approximately 1 in 178,000 prescriptions (190). Examination of

the case reports revealed that duration of TVX therapy in patients did not

44

correlate with a toxic response. The hepatotoxic events linked with M were
equal in males and females and were seen in individuals ranging in age from 21-
91 years old. This clinical profile of TVX hepatotoxicity classified the toxicity as
an lADR.

TVX was the first fluoroquinolone to have its use severely restricted
because of liver toxicity (191). In retrospect, the hepatotoxic potential of TVX was
seen in dogs at doses exceeding therapeutic doses (190). Some dogs developed
elevated liver enzyme signals and centrilobular necrosis. However, because the
toxicity was species-specific, reversible and no serious hepatotoxicity was seen
in clinical trials, the liver injury observed in dogs was not considered relevant for
patients treated with therapeutic doses. Since several other fluoroquinolones
have not been associated with human hepatotoxicity, TVX lADRs do not seem to
be related to the desired pharmacologic properties of the drug (191).

The pathology of TVX hepatotoxicity was similar in several cases reported.
Viral, metabolic and autoimmune causes of hepatitis were excluded in the cases
reported (11, 192, 193). A liver biopsy revealed centrilobular hepatocellular
necrosis with collapsed sinusoids around the central vein, whereas the portal
tracts appeared normal. In addition, parenchyma-based lymphocytes, plasma
cells and eosinophils were present, and these inflammatory cells were highly
concentrated at the peripheral edges of the centrilobular necrosis (11, 192, 193).

Some signs of hepatic dysfunction have also been linked to other
fluoroquinolones such as temafloxacin and ciprofloxacin (194, 195).

Temafloxacin was removed from the market due to hemolytic anemia, but it was

45

also linked to hepatic dysfunction. Both temafloxacin and trovafloxacin share a
unique 2,4-difluorophenyl moiety at position 1 of the molecule which is unseen in
other fluoroquinolones (193). This moiety is not included in the surrogate
molecule mentioned above that formed a reactive metabolite in the presence of
Cyp1A2 or MP0 (189). It is possible that this structural component is a

requirement to idiosyncratic hepatotoxicity of fluoroquinolones.

1.3.3 Interaction between trovafloxacin and inflammation

Clinical reports of TVX hepatotoxicity revealed the presence of
inflammatory cells in liver biopsies (11, 192, 193). Based on these findings, it was
hypothesized that an inflammatory stress might interact with TVX to induce
hepatotoxicity. Indeed, a modest inflammatory stress induced by LPS interacted
with a nontoxic dose of TVX to precipitate liver injury in rats (58). This
hepatotoxic interaction was not seen with coexposure to LPS and levofloxacin
(LVX), a fluoroquinolone lacking the propensity to cause lADRs in humans. TVX
given to rats 2 h after a nonhepatotoxic dose of LPS resulted in midzonal lesions
of coagulative necrosis (58). The finding that TVX, but not LVX, caused liver
injury when administered with LPS, suggested that the interaction between TVX
and an inflammatory stress is independent of the desired pharmacology of the
drug.

Hepatic gene expression analysis identified a unique profile induced by
LPS/TVX-coexposure, including that TVX enhanced the LPS-induced expression

of a number of chemokines. The increased chemokine expression suggested a

46

role for PMNs, which was confirmed by the result that PMN depletion attenuated
LPS/TVX-induced hepatotoxicity (58). The selectivity of LPS coexposure to only
interact with TVX, and not LVX, to cause liver injury, suggests that inflammatory
stress might play a role in TVX hepatotoxicity in humans. However, further study
is needed to determine if this interaction between inflammation and TVX is
species-specific. In addition, understanding the mechanisms of TVX-
inflammation interaction might provide a better understanding of TVX

hepatotoxicity in humans.

47

1.4 Hypothesis and specific aims

The overall hypothesis is that a coexposure of mice to TVX and an
inflammatory stress results in idiosyncrasy-like liver injury that is dependent on
the following factors: TNFa, lFNy, hemostatic system activation and VEGF.
These factors were chosen based on their involvement in other models of
inflammatory liver injury. In addition, these factors were chosen based on the
interactions of these inflammatory mediators with one another outlined in Tables
2-5. I hypothesize that these factors create vicious proinflammatory cycles
possibly involved in the pathogenesis of liver injury. The goal of the specific aims
proposed is to determine which factors are critical for the development of
TVX/LPS-induced liver injury and how each mediator fits into the cascade of
events leading to hepatotoxicity. These general hypotheses will be addressed by

specific hypotheses represented in five specific aims:

Aim 1 Hypothesis: TVX interacts with an inflammatory stress to cause

idiosyncrasy-like liver injury in mice. (Chapters 2 and 3)

Aim 2 Hypothesis: TNFa is critical to the development of hepatotoxicity caused
by TVX/LPS treatment via interactions with PMNs, lFNy, VEGF and/or

hemostasis. (Chapters 2 and 4)

48

Aim 3 Hypothesis: TVX/LPS-induced liver injury is dependent on IFNy, which
positively regulates PMNs, TNFa, hemostatic system activation and/or VEGF.

(Chapter 5)

Aim 4 Hypothesis: Hemostatic system imbalance is involved in TVX/LPS-induced

hepatotoxicity, and this affects PMNs, TNFa, lFNy and/or VEGF. (Chapter 6)

Aim 5 Hypothesis: VEGF is important in the development of liver injury caused by

TVX/LPS treatment via interactions with PMNs, TNFa, IFNy and/or hemostatic

system activation. (Chapter 7)

49

1.5 Overview and significance of dissertation

The studies proposed outline the development of a murine model of TVX
idiosyncratic liver injury and the identification of several factors involved in the
pathogenesis. A murine model of idiosyncratic liver injury is potentially of great
importance. First, it proves that the LPS/IADR-drug interaction seen in rats is not
species-specific. In addition, commonalities across species within the TVX/LPS
coexposure model of liver injury might extrapolate to human TVX hepatotoxicity.
The development of a murine model of idiosyncratic hepatotoxicity will also prove
useful for mechanistic studies through the use of genetically modified mice.
Finally, a predictive animal model of lADRs could be added to preclinical testing
paradigms to eliminate drug candidates with the potential to cause lADRs in
humans. The use of a predictive model might prevent the release of drugs which
could be harmful to patients. In addition, the elimination of drug candidates with
the propensity to cause lADRs could save pharmaceutical companies millions of
dollars in the financial investment in developing drugs that ultimately must be
withdrawn from the market.

Identification of the role of TNFa, lFNy, the hemostatic system and VEGF
in TVX/LPS-induced liver injury is important to understanding the mechanism of
TVX hepatotoxicity in humans. The roles of IFNy, the hemostatic system and
VEGF in hepatotoxicity have not been extensively studied. In addition, a
comprehensive examination as to how these factors affect one another and how
they fit into the cascade of events resulting in liver injury has not been reported.

Such interactions among inflammatory mediators are not only of importance to

50

this model of liver injury, but have the potential to be extrapolated to other

models of inflammatory tissue injury.

51

CHAPTER 2

Shaw, P.J., Hopfensperger, M.J., Ganey, P.E., and Roth, RA. (2007).
Lipopolysaccharide and trovafloxacin coexposure in mice causes
idiosyncrasy-like liver injury dependent on tumor necrosis factor-alpha.

Toxicol Sci. 100(1): 259-266.

52

2.1 Abstract

lADRs occur in a small subset of patients, are unrelated to the
pharmacological action of the drug, and occur without an obvious relationship to
dose or duration of drug exposure. The liver is often the target of these reactions.
Why they occur is unknown. One possibility is that episodic inflammatory stress
interacts with the drug to precipitate a toxic response. We set out to determine if
LPS renders mice sensitive to TVX, a fluoroquinolone antibiotic linked to
idiosyncratic hepatotoxicity in humans, and if the cytokine TNFor is involved in the
development of liver injury. Male mice were treated with a nontoxic dose of TVX
followed 3 h later by a nonhepatotoxic dose of LPS. Coexposure to TVX and LPS
led to a significant increase in liver injury as determined by plasma alanine
aminotransferase activity and histopathological examination. In contrast,
coexposure of mice to LPS and LVX, a fluoroquinolone without liability for
causing lADRs in humans, was not hepatotoxic. Measurements of TNFa
concentration in the plasma revealed a significant, selective increase in
TVX/LPS-treated mice at times prior to and at the onset of liver injury. Treatment
with either pentoxifylline to inhibit TNFor transcription or etanercept to inhibit
TNFa activity significantly reduced TVX/LPS-induced liver injury. The results
suggest that the model in mice is able to distinguish between drugs with and
without the propensity to cause idiosyncratic liver injury and that the

hepatotoxicity is dependent on TNFa.

53

2.2 Introduction

As described in Section 1.1, the ability of an inflammatory stress to
potentiate the hepatotoxicity of numerous xenobiotic agents has led us to
hypothesize that inflammatory stress could alter the toxicity threshold of certain
drugs precipitating idiosyncratic toxicity (55, 73). In rats, a modest inflammatory
episode induced by LPS administration potentiates the hepatotoxicity of several
idiosyncratic drugs including ranitidine, chlorpromazine, diclofenac, sulindac and
trovafloxacin (56-60). Such inflammation-drug interaction leading to
hepatotoxicity has not been demonstrated in mice. The purpose of this study was
to test the hypothesis that an inflammatory stress induced by LPS would
potentiate TVX hepatotoxicity in mice. Furthermore, the study tested the
hypothesis that LVX, a fluoroquinolone without idiosyncratic liability, would not
interact with LPS to cause liver injury. TNFa is a mediator of inflammation
critically involved in several models of liver injury, as described in more detail in
Section 1.2.3. Finally, we tested the hypothesis that TNFor is critically involved in

the development of TVX/LPS-induced liver injury in mice.

54

2.3 Materials and Methods

2.3.1 Materials

Unless othenlvise noted, all chemicals were purchased from Sigma-Aldrich
(St. Louis, MO). Lipopolysaccharide derived from Eschericia coli serotype
O55:B5 was used for these studies. Lot 024K4067 with activity of 9.2 x 106
EU/mg was used for the experiments represented in Fig. 2.1-2.3. Lot 075K4038
with an activity of 3.3 x 106 EU/mg was used for the experiments represented in
Fig. 2.4-2.9. The activity was determined using a colorimetric, kinetic Limulus
amebocyte lysate assay purchased from Cambrex Corp. (Kit 50-650U; East
Rutherford, NJ). TVX and LVX were kind gifts from Abbott Laboratories (Abbott
Park, IL). Infinity ALT reagent was purchased from Therrno Electron Corp.

(Louisville, CO).

2.3.2 Animals

Male, CS7BL/6J mice (Jackson Laboratory, Bay Harbor, ME), 9-11 weeks
old and weighing 21-26 g were used for the studies. Animals were given
continual access to bottled spring water and were fed a standard chow (Rodent
Chow/1' ek8640, Harlan Teklad, Madison, WI) ad libitum. Mice were allowed to
acclimate for 1 week in a 12 h light/dark cycle. They received humane care
according to the criteria outlined in the Guide for the Care and Use of Laboratory
Animals, and procedures were approved by the MSU Committee on Animal Use

and Care.

55

2.3.3 Experimental protocol

Mice fasted for 12 h were given various doses of TVX, LVX or their saline
vehicle by oral gavage. They were then given LPS at 67 x 10‘5 EU/kg or 2.0 x 106
EUlkg (lots 024K4067 or 075K4038, respectively) by intraperitoneal injection 3 h
after drug dosing. During the course of these studies we were forced to change
lots of LPS. The dose of LPS of the initial lot was chosen based on preliminary
dose-response studies for which the objective was to identify a nonhepatotoxic
dose of LPS. For the lot of LPS that was used to complete these studies, a dose
was chosen that was nonhepatotoxic when given alone and produced liver injury
in TVX-cotreated mice that was similar in magnitude and timing to that produced
by the first lot.

Food was returned immediately after LPS administration. Mice were
anesthetized with sodium pentobarbital (50 mg/kg, i.p.) at various times, and
blood was drawn from the vena cava into a syringe containing sodium citrate
(final concentration, 0.76%) and transferred to an Eppendorf tube for preparation
of plasma. The left lateral liver lobe was fixed in 10% neutral buffered formalin
and blocked in paraffin within 72 h. For some studies, mice were treated with
pentoxifylline (200 mg/kg) or sterile saline by intraperitoneal injection 1 h before
LPS injection. In other studies, mice were treated with etanercept (8 mg/kg) or
sterile water by intraperitoneal injection either 1 h before LPS injection (Fig. 2.7
and 2.8) or 1.5 h after LPS dosing (Fig. 2.9). Etanercept (Enbrel, Amgen
Pharmaceuticals) was purchased from the Michigan State University Pharmacy

(East Lansing, MI).

56

2.3.4 Histopathology

Fonnalin-fixed left lateral liver lobes were embedded in paraffin, sectioned
at 5 pm, stained with hematoxylin & eosin and examined by light microscopy.
The tissue sections presented in the figures were from mice with plasma ALT

activity close to the average of the respective treatment group.

2.3.5 TNFa analysis

The plasma concentrations of TNFa were measured using a mouse
inflammation kit (Cat. No. 552364) purchased from BD Biosciences (San Diego,
CA). The BD cytometric bead array analysis was performed on a B0

FACSCaIibur flow cytometer (BD Biosciences).

2.3.6 Statistical analyses

Results are presented as mean j; S.E.M. A 1-,2-, or 3-way analysis of
variance (ANOVA) was used as appropriate after data normalization. For the
TVX/LPS timecourse (Fig. 2.13), an ANOVA on Ranks was used. All pairwise
comparisons were made using Dunn’s method. The criterion for significance was

p < 0.05 for all studies.

57

2.4 Results

2.4.1 Dose-response and timecourse of liver injury

Administration of LPS after TVX caused a significant increase in plasma
ALT activity in a TVX dose-dependent manner (Fig. 2.1A): TVX doses of 80
mg/kg or greater caused hepatotoxicity in LPS-treated mice. TVX alone did not
cause a significant increase in ALT activity up to 1000 mg/kg (data not shown).
Administration of TVX doses greater than 200 mg/kg followed by LPS led to
death within 15 h. A TVX dose of 150 mg/kg and LPS given 3 h later provided a
maximal response with approximately 90% survival of mice; this protocol was
chosen for all additional studies.

To evaluate the time-dependence of liver injury, TVX was administered 3
h before LPS dosing, and plasma ALT activity was measured at various times.
TVX or LPS given alone did not significantly affect ALT activity compared to
control mice at any time evaluated. Plasma ALT activity was significantly
elevated by 9 h after TVX/LPS coexposure and peaked at 15-21 h after LPS (Fig.
2.13).

2.4.2 Comparison of trovafloxacin and Ievofloxacin

Unlike TVX, LVX is not associated with human lADRs. We compared the
hepatotoxic response to each of these in animals cotreated with LPS. The
pharmacologically efficacious dose of TVX is similar in mice and humans (196,

197), and the same is true for LVX (198, 199). We chose a dose of

58

Fig. 2.1. Dose response and development of liver injury from TVX/LPS
cotreatment in mice. A, Mice were given TVX at various doses (8, 25.3, 45, 80,
100, 150, and 200 mglkg; p.o) and then 3 h later LPS (67 x 106 EU/kg; i.p.) or
Veh (sterile saline). Hepatic parenchymal cell injury was estimated 15 h after
LPS administration from increases in plasma ALT activity. n = 4-9 animals/group.
* significantly different from Veh-treated group. B, Mice were treated with TVX
(150 mglkg; p.o.) or Veh and then 3 h later with LPS (67 x 106 EU/kg; i.p.) or Veh.
Plasma ALT activity at various times after LPS dosing is depicted. n = 4-6

animals/group. * significantly different from 0 h group.

59

8
8

E

L

ALT (UIL)

i

ALT(UIL)
- i i .3 .5 $5

. +LPS

 

 

+ Veh

     

 

50 100 150 200
TVX(mglkg)

 

10 20 30 40

Hours after LPS

60

Fig. 2.2. Comparison of TVX/LPS with LVX/LPS coadministration. TVX (150
mglkg), LVX (375 mglkg) or Veh (saline) was administered by oral gavage, and 3
h later LPS (67 x 106 EU/kg; i.p.) or Veh was administered. Mice were
anesthetized and sacrificed 15 h after LPS administration. Hepatic parenchymal
cell injury was estimated as increases in plasma ALT activity. n = 4-6
animals/group. * significantly different from VehNeh. # significantly different from

Veh/LPS.

61

ALT (UIL)

7
6

0|

AGO-h

 

 

 

 

 

000 -
000« r==l Veh
000 TVX
000 . — LVX
000 /
000 1/
500 .

0 r—um-

Veh

62

 

LVX (375 mglkg) to keep the dose ratio of TVX/LVX similar to the ratio of doses
used clinically in humans (200). LVX, TVX or Veh was given 3 h prior to LPS or
Veh, and then mice were sacrificed 15 h later to measure plasma ALT activity
and for histologic examination of the livers. TVX, LVX or LPS were all nontoxic
when administered alone (Fig. 2.2). TVX/LPS coexposure increased ALT activity
in the plasma, suggesting hepatic parenchymal cell injury. ALT activity was not
increased in LVX/LPS-treated mice. A

There were no significant hepatocellular lesions in mice treated with
VehNeh, TVXNeh or LVXNeh (Fig. 2.3A, 2.3B and 2.3C, respectively).
Histopathological examination of livers from TVX/LPS-cotreated mice (Fig. 3E)
revealed hepatocellular necrosis, which was not seen in Veh/LPS- (Fig. 2.30) or
LVX/LPS-treated mice (Fig. 2.3F). Inflammatory cell infiltration was seen in all
LPS-treated groups. The coagulative necrosis seen in the TVX/LPS-treated
group was located predominantly midzonally but could also be found in
centrilobular regions. The appearance of these lesions in TVX/LPS-treated mice
followed the same timecourse as was seen for ALT activity in the plasma (data

not shown).

2.4.3 Timecourse of TNFot concentration in plasma

Mice were treated according to the protocol described above and were
sacrificed at various times (0, 1.5, 3, 4.5 and 6 h) after LPS. These and
subsequent studies were performed with a different lot of LPS than was used to

generate data in Figures 2.1-2.3. The dose used for these studies was 2 x 106

63

Fig. 2.3. Liver histopathology in mice cotreated with LPS and either TVX or
LVX. Mice were treated with TVX (150 mglkg), LVX (375 mglkg) or Veh (p.o) and
then dosed with LPS (67 x 106 EU/kg; i.p.) or Veh. Liver sections are from mice
treated with VehNeh (A), TVXNeh (B), LVXNeh (C), Veh/LPS (D), TVX/LPS (E),
LVX/LPS (F) and killed at 15 h. The arrows indicate randomly distributed,
variably sized foci of coagulative necrosis seen only in TVX/LPS-treated mice
and which were observed predominantly in midzonal regions but also in some

centrilobular regions.

64

. .- 4'-
- V'u'.._v
I."

.‘l -'

 

D -A
.
l'

r“: ~ ”f“ v
1 ' c.” . ‘ ' .

131

65

Fig. 2.4. Timecourse of TNFa concentration in the plasma of treated mice.
Mice were treated with TVX (150 mglkg), LVX (375 mglkg) or Veh (p.o) and then
dosed with LPS (2 x 106 EU/kg; i.p.) or Veh 3 h later. n = 3-10 animals/group. *
significantly different from the same treatment group at 0 h. # significantly

different from Veh/LPS-treated mice at the same time.

66

    
  
 
 

4000. —-o-- VehNeh

—-o— TVXNeh

: --O- LVXNeh
—-e— Veh/LPS

E 3000- —-e-— TVX/LPS

a *# ------e- LVX/LPS

D.

V 2m ‘

Z5

2

I- ‘°°° .......

 

 

Hours after LPS or Veh

67

EU/kg, and despite the large difference in dose based on activity in the Limulus
lysate assay, the results obtained with both lots were similar in terms of the
magnitude and timing of liver injury. Plasma ALT activity was significantly and
selectively increased in TVX/LPS-treated mice starting at 4.5 hrs after LPS (data
not shown). LPS-treated groups showed a significant increase in plasma TNFa
concentration at all times measured (Fig. 2.4). TVX administered prior to LPS
caused greater elevation of TNFa concentration in the plasma compared to
Veh/LPS-treated mice at 3 and 4.5 h after LPS. By contrast, LVX cotreatment

had no effect on the LPS-induced change in plasma TNFa concentration.

2.4.4 Pentoxifylline study

As mentioned above, ALT activity was increased in TVX/LPS-cotreated
mice 4.5 h after LPS administration, and plasma TNFa was selectively increased
at this time. This result raised the possibility of a role for this cytokine in the
development of hepatotoxicity in TVX/LPS-treated mice. Pentoxifylline (PTX) is a
nonspecific phosphodiesterase inhibitor that inhibits LPS-induced TNFa
production by increasing cAMP in monocytes/macrophages. The increase in
cAMP inhibits the translocation and activation of NFKB, which controls TNFor
expression (201). A dose of PTX 200 mg/kg (i.p.) given 1 h before LPS
administration significantly decreased plasma TNFa concentration 1.5 h after
LPS treatment (Fig. 2.5A). This dose of PTX significantly reduced TVX/LPS-
induced liver injury, as estimated by plasma ALT activity 15 h after LPS dosing

(Fig. 2.58). TVXNeh/LPS-treated livers had much less glycogen deposition and

68

Fig. 2.5. The effect of pentoxifylline (PTX) on LPS-induced TNFa expression
and TVXILPS-induced liver injury. A, Mice were treated with PTX (200 mglkg;
i.p.) or Veh (saline) 1 h prior to LPS. Mice were sacrificed 1.5 h after LPS
treatment and plasma TNFa concentration was measured. B, TVX (150 mglkg)
or Veh was administered by oral gavage, followed by PTX or Veh 2 h later. LPS
(2 x 106 EU/kg; i.p.) was given 1 h after PTX dosing. Mice were sacrificed 15 h
after LPS treatment, and ALT activity was measured in the plasma. n = 5-10
animals/group. * significantly different from VehNehNeh control. * significantly

different from TVXNeh/LPS treatment group.

69

§

TNFor. (ngm L)

O

g

g
l

 

 

 

ALT (UIL)

 

VehNehNeh TVXNeh/LPS TVXIPTXILPS

 

VehNehNeh TVXNeh/LPS TVXIPTXILPS

7O

Fig. 2.6. Effect of PTX on TVXILPS-induced liver pathology. Mice were
treated as described in Fig. 5 and sacrificed 15 h after LPS. Liver sections from
mice treated with VehNehNeh (A), TVXNeh/LPS (B), and TVX/PTX/LPS (C)
were examined. The arrows indicate randomly distributed, variably sized foci of
coagulative necrosis and hemorrhage seen only in TVXNeh/LPS-treated mice.
Lesions were not obvious in the TVX/PTX/LPS-treated group despite the slightly

increased ALT activity in the plasma in this group (Fig. SB).

71

.C‘
I

-.: .I .
1" P’

gig}: ,

D
‘0
4. -

‘ .
h. ..
. l‘

1‘.-

‘T a? )' ‘kblfi

 

72

had foci of midzonal hepatocellular necrosis compared to vehicle-treated control
mice (Fig. 2.6B and 2.6A, respectively). PTX administration to TVX/LPS-treated
mice reduced the midzonal hepatocellular necrosis and also reduced the

glycogen depletion compared to the TVXNeh/LPS group (Fig. 2.6C).

2.4.5 Etanercept inhibition of TNFa activity

Etanercept is a recombinant, human soluble TNFa receptor that inhibits
TNFa activity. An etanercept dose of 8 mglkg (i.p.) caused a significant decrease
in plasma TNFa concentration In TVX/LPS-treated mice at 4.5 h after LPS
administration (Fig. 2.7A). This dose of etanercept administered 1 h before LPS
completely protected mice from the TVX/LPS-induced increase in plasma ALT
activity (Fig. 2.73) and from hepatocellular necrosis (Fig. 2.8). The TVX/LPS-
treated mice consistently had midzonal and centrilobular foci of coagulative
necrosis which were not observed when etanercept was administered (Fig. 2.88
and 2.80, respectively).

In an attempt to determine if the prolongation of the LPS-induced plasma
TNFa peak by TVX pretreatment (Fig. 2.4) was critical to TVX/LPS-induced liver
injury, etanercept was administered at 1.5 h after LPS dosing (ie, at the time
plasma TNFa concentration had peaked). Etanercept administration at this time

provided significant reduction in TVX/LPS-induced liver injury (Fig. 2.9).

73

Fig. 2.7. The effect of etanercept on TVXILPS-induced TNFa expression and
liver injury. A, TVX (150 mglkg) or Veh was administered by oral gavage
followed by etanercept (8 mglkg; i.p.) or Veh 2 h later. LPS (2 x 106 EU/kg; i.p.)
was given 1 h after etanercept dosing. Mice were sacrificed 4.5 h after LPS
treatment, and plasma TNFor concentration was measured. B, Mice were treated
as described above and sacrificed 15 h after LPS; ALT activity was measured in
the plasma. n = 4-8 animals/group. * significantly different from VehNehNeh

#

control group. significantly different from TVXNeh/LPS treatment group.

74

  

 

VehNeh/Veh TVXNeh/LPS TVXIEtan/LPS

 

m a. W. m ..
353355

 

 

Mafia

:5. H2

VehNehNeh TVXNeI‘IIU’S TVXlEtanILPS

75

Fig. 2.8. Effect of etanercept on TVXILPS-induced liver pathology. Mice were
treated as described in Fig. 7 and sacrificed 15 h after LPS. Liver sections from
mice treated with VehNehNeh (A), TVXNeh/LPS (B), and TVX/etanercept/LPS
(C) were examined. The arrows indicate randomly distributed, variably sized foci

of coagulative necrosis and hemorrhage seen only in TVXNeh/LPS-treated mice.

76

.t

i

we.

fim'fi

T
revels

 

77

Fig. 2.9. Effect of etanercept treatment given at the peak of plasma TNFa on
TVXILPS-induced liver injury. Mice were treated with TVX (150 mglkg; p.o.) 3 h
before LPS (2 x 10‘5 EUlkg; i.p.). Mice were then treated with etanercept (8
mglkg; i.p.) 1.5 h after LPS dosing and sacrificed 15 h after LPS; ALT activity
was measured in the plasma. n = 5 animals/group. * significantly different from

TVX/LPSNeh-treated group.

78

ALT (UIL)

3

"S

3

§

0

 

 

 

79

 

 

2.5 Discussion

The underlying mechanisms behind hepatic lADRs in humans are
unknown. One of the most widely accepted hypotheses is that they involve
immune-mediated hypersensitivity reactions (202). However, for the majority of
drugs there is limited evidence to support this theory. Another widely accepted
hypothesis is that mitochondrial dysfunction from oxidative stress leads to
hepatotoxic lADRs. There is evidence from human hepatocytes that TVX can
cause oxidative stress and mitochondrial damage (203). In contrast, changes in
markers of oxidative stress were not observed in rats treated with TVX,
suggesting either that the effect may be species-specific or that the observation
in vitro does not pertain in vivo (58). Thus, mechanisms of TVX-mediated liver
injury remain unknown.

In rats, the TVX/LPS interaction was found to precipitate a hepatotoxic
response (58). The timing of dosing in the rat model was different from the
mouse model presented here in that the TVX was given 2 h after LPS
administration. The rats were treated with TVX by i.v. rather than oral
administration, which might explain the differences in protocols that caused
maximally toxic responses. That is, greater time might be needed after oral
closing to reach effective plasma TVX concentration compared to iv. injection.
Additionally, the half-life of TVX in mice is much longer than in rats, and this
might contribute to the differences in the dosing protocol needed to induce
maximal liver injury (204, 205). The development of hepatotoxic TVX-

inflammation interaction in both mice and rats demonstrates that the

80

phenomenon is not species-specific and might have common mechanisms which
could be extrapolated to TVX lADRs in humans.

The degree of TVX/LPS-induced liver injury was much greater in mice (Fig.
2.1) compared to rats (58). This assessment is based on histopathology and on
the fold increase in plasma ALT activity. In mice, the peak plasma ALT activity
was about 30 fold greater than in the rat model, and the liver lesions were more
pronounced. Both moderate and severe hepatotoxic responses have been
reported in people who took TVX (206). The robustness of the murine model of
liver injury resembles the severe hepatotoxicity caused by TVX in humans more
so than the rat model. This might be due to the greater similarity in TVX
pharmacokinetics in mice and humans (204, 205, 207). TVX binding to serum
proteins is greater in rats compared to humans, 92 vs. 70%, respectively (205,
207). The degree of serum protein binding in mice is unavailable, but the more
extensive serum protein binding in rats might contribute to the less robust liver
injury.

Coexposure to LVX and LPS did not produce hepatotoxicity in mice, as
indicated by both plasma ALT activity (Fig. 2.2) and histopathological
examination (Fig. 2.3). Thus, for this class of drugs, the animal model is selective
for a drug that produced lADRs in humans. The difference in response was
probably not due to pharmacokinetic differences, as LVX and TVX have very
similar elimination half-lives (208). Another possible explanation for the selective
hepatotoxicity with TVX/LPS coexposure is that TVX is more potent against GI

bacteria, causing release of LPS into the bloodstream, which, when paired with

81

LPS administration, precipitates a toxic response. This possibility, however, can
be ruled out. If TVX or LVX caused LPS release from the GI tract, then it should
have been reflected in increased plasma TNFor; however, neither of the drugs
alone caused such an increase. Moreover, LVX failed to enhance the increase in
plasma concentration of TNFa or to cause liver injury when coadministered with
LPS. Thus, this model is selective for the lADR-causing TVX in both the
development of liver injury and in the enhancement of LPS-induced increase in
plasma TNFa concentration.

TVX pretreatment selectively prolonged the LPS-induced plasma TNFa
peak before and during the onset of liver injury (Fig. 2.4). TNFa is critically
involved in several models of liver injury including ischemialreperfusion, and
endotoxemia (209, 210). To explore the role of TNFa, both PTX and etanercept
were used to inhibit TNFa activity. PTX pretreatment provided significant
protection from TVXILPS-induced liver injury (Fig. 2.5). In addition to inhibiting
TNFa, PTX has several other effects including preventing platelet aggregation,
decreasing other proinflammatory cytokines, and inhibiting hepatic fibrogenesis
(211). Accordingly, a more selective inhibitor was also used.

Etanercept is a recombinant human soluble TNFa receptor which
specifically neutralizes the activity of TNFot. Pretreatment with etanercept
completely protected mice from TVX/LPS-induced liver injury (Fig. 2.7).
Additionally, etanercept administration at a later time to eliminate the
prolongation by TVX of the LPS-induced plasma TNFa peak also provided

protection (Fig. 2.9). Thus, the prolonged TNFa presence caused by TVX

82

pretreatment seems to be critically involved in the TVX/LPS-induced liver injury.
However, this finding does not rule out a critical role for the initial peak of TNFa
(0-1.5 h after LPS). Whether TNFa directly causes hepatotoxicity or acts
indirectly through other mediatorswill require further investigation.

The cellular source(s) of TNFa in this model have not been explored. In
the liver, Kupffer cells can be stimulated by LPS to release TNFa and other
cytokines. These cells are an important source of TNFa in several models of liver
injury (212, 213) and seem likely to be involved in TVX/LPS-induced liver injury
as well. Similarly, neutrophils were found to be critically involved in the TVX/LPS
model in rats. It seems likely that they play a similar role in the mouse model, but
additional studies are required to confirm this.

The hepatotoxic interaction between TVX and LPS presented here
contrasts to a previous report showing that TVX reduces LPS-induced death in
mice (214). The difference in the effect of TVX could be due to different timing of
TVX administration. The protective effect of TVX was seen when it was
administered at 47, 17 and 1 h before LPS. In our hands, the timing of TVX
administration in relation to LPS was critical. For example, administration of TVX
after LPS dosing did not lead to significant liver injury in mice (data not shown),
suggesting that TVX had to be present in the body during LPS administration to
precipitate liver injury. An alternative explanation for the contrasting response is
that after a lethal dose of LPS (214), TVX might play a different role to reduce
mortality, for example by killing bacteria translocated from the GI tract into the

circulation. In addition, the previous study referenced used Swiss Webster mice

83

whereas C57/BL6 mice were used for these studies, so that strain differences
might contribute to the disparate results.

It has been reported that TVX significantly reduces TNFa concentrations
induced by LPS in mice (214, 215). These results contrast with data presented in
Fig. 2.4. The difference in results could be due to different strains of mice, doses
of LPS (lethal vs. nonhepatotoxic) or different treatment protocols, in which TVX
was given 1 h (214) or 3 h (data presented here) before LPS. In that study, TVX
increased plasma TNFa concentration, an effect not observed in our study. The
plasma concentration of TNFa in control mice was reported to be 1.4 1; 0.5
ng/mL (214), a value that is extremely high for normal mice and might reflect an
ongoing inflammatory response in their controls.

It has also been reported that alatrofloxacin, a prodrug of TVX, decreased
LPS-stimulated expression of TNFa mRNA and protein in vitro in human
peripheral blood mononuclear cells (PBMCs) (215), a result that contrasts with
our findings. Interestingly, in rats cotreated with TVX and LPS, mRNA for TNFa
in liver was not increased, but mRNA for TNF-induced protein was elevated (58),
suggesting a post-transcriptional mechanism for increasing TNFor protein. It is
also possible that Kupffer cells, a major source of TNFa in the liver, respond
differently to the interaction of TVX with LPS than human PBMCs with respect to
TNFa production. Another possibility is that a hepatic metabolite is involved in
the TVX effect on LPS stimulation, and this metabolite might not be produced by
isolated PBMCs. Additionally, our studies would have allowed more time for such

a metabolite to form. Thus, although a previous study provided evidence for an

anti-inflammatory property of TVX, the treatment protocols, mouse strains, and
doses contrast with those employed in this study.

In summary, a modest inflammatory stress induced by LPS rendered TVX,
but not LVX, hepatotoxic in mice. TVX pretreatment prolonged the LPS-induced
increase in TNFor in the plasma. The increase in TNFa plays a critical role in the
development of TVX/LPS-induced liver injury. The demonstration of TVX/LPS
toxicity in both mice and rats indicates that the interaction is not species-specific.
The results suggest the possibility that inflammatory stress underlies the
development of TVX-induced idiosyncratic liver injury and support the potential of
animal models of drug-inflammation interaction as preclinical predictors of lADRs

in humans.

85

CHAPTER 3

Trovafloxacin enhances the inflammatory response to a gram-negative or a
gram-postitive bacterial stimulus, resulting in CD18-dependent liver injury

in mice.

86

3.1 Abstract

Trovafloxacin, a fluoroquinolone antibiotic, was strongly linked with several
cases of idiosyncratic hepatotoxicity leading to the severe restriction of TVX
usage. Previous studies have shown that a modest inflammatory stress induced
by LPS renders nontoxic doses of TVX hepatotoxic in mice. The liver injury is
dependent on TNFa, suggesting TVX might enhance the response to an
inflammatory stress. The purpose of this study was to examine the interaction of
TVX with a subsequent inflammatory stress induced by either a gram-negative or
a gram-positive bacterial stimulus. Mice were given TVX 3 h before LPS (gram-
negative) or a peptidoglycan/lipoteichoic acid (PGN-LTA) mixture isolated from S.
aureus (gram-positive). Administration of TVX, LPS or PGN-LTA was
nonhepatotoxic. In contrast, TVX administration prior to LPS or PGN-LTA
resulted in liver injury starting by 4.5 h which was maximal at 15 h.
Histopathology revealed that TVX/LPS-coexposure resulted in primarily midzonal
hepatocellular cell death, whereas TVX/PGN-LTA-induced necrosis was primarily
centrilobular. LPS or PGN-LTA alone increased plasma concentrations of lL-18,
lL-1B, lL-6, lL-10, keratinocyte chemoattractant (KC), monocyte chemoattractant
protein 1 (MCP-1), macrophage inflammatory protein 1a (MlP-1a), VEGF, IFNy,
TNFa and macrophage inflammatory protein 2 (MIP-2) at 4.5 h. TVX
administration enhanced the LPS induction of all of these cytokines/chemokines.
In contrast, TVX enhanced the PGN-LTA-induced increase of all except TNFa
and IFNy. ,TVX/LPS- and TVX/PGN-LTA-induced liver injury was significantly

attenuated by CD18 antiserum treatment. In summary, TVX significantly

87

enhanced the inflammatory response of mice to either a gram-negative or gram-

positive stimulus and caused hepatotoxicity which was dependent on CD18.

88

3.2 Introduction

A nontoxic inflammatory stress induced by LPS, a cell wall component of
Gram-negative bacteria, can interact with a nontoxic dose of TVX to cause
TNFor-dependent hepatotoxicity in both rats and mice (58, 216). Toll-like
receptors recognize pathogens and bacteria as foreign as described in Section
1.2.1. Whereas LPS activates cells through binding to TLR4, TLR2 serves as the
primary receptor for Gram-positive bacteria and their cell wall components,
peptidoglycan (PGN) and lipoteichoic acid (LTA) (217, 218). Activation of TLR2
by PGN and LTA leads to the activation of the transcription factor NF-KB, which
increases the expression of inflammatory cytokines and chemokines (217). This
study determined if an inflammatory stress induced by gram-positive microbial
products interacts with TVX to cause liver injury.

Additionally, TVX enhanced the LPS-induced increase of TNFa (216).
One purpose of this study was to determine if TVX pretreatment enhanced the
LPS- or PGN-LTA-induced increase of cytokines. The profile of cytokines altered
by TVX coexposure with either inflammatory stimuli were compared. Finally, we
tested that hypothesis that CD18, involved in PMN activation, plays a critical role
in the development of hepatotoxicity resulting from TVX/inflammatory stress

coexposure.

89

3.3 Materials and Methods

3.3.1 Materials

Peptidoglycan and lipoteichoic acid isolated from S. aureus were
purchased from Sigma (St. Louis, MO). Rabbit anti-murine CD18 antiserum was
designed against amino acids 89-100 and purchased from New England Peptide
(Gardner, MA). Please refer for Section 2.3.1 for additional information on this

topic.

3.3.2 Animals

Please refer to Section 2.3.2 for information on this topic.

3.3.3 Experimental protocols

Mice fasted for 12 h were given TVX (150 mglkg) or Veh (saline) by oral
gavage. They were then given LPS (2.0 x 106 EU/kg), PGN-LTA (30 mglkg each)
or Veh (saline) by intraperitoneal injection 3 h later. Food was returned
immediately after this dosing. Mice were anesthetized with sodium pentobarbital
(50 mglkg; i.p.) and killed at either 4.5 or 15 h after the administration of LPS,
PGN-LTA or Veh for various measurements. Blood was drawn from the vena
cava into a syringe containing sodium citrate, resulting in a final concentration of
0.76%. The left lateral lobe was fixed in 10% neutral buffered formalin and

paraffin blocked.

90

For some studies, mice were treated with CD18 antiserum. CD18
antiserum or rabbit control serum (0.25 mL; i.p.) was administered when food

was removed and then again 2 h after LPS, PGN-LTA or Veh administration.

3.3.4 Histopathology

Formalin-fixed, left lateral liver lobes were embedded in paraffin, sectioned
at 5 pm, stained with hematoxylin & eosin and examined by light microscopy.
The tissue sections presented in the figures were from mice with plasma ALT

activity close to the average of the respective treatment group.

3.3.5 Cytokine measurements

The plasma concentrations of lL-1B, TNFa, lL-10, lL-6, lL-18, IFN 7, VEGF,
MCP-1, KC, MlP-2 and MlP-1a were measured using custom Bio-plex cytokine
assays purchased from Bio-Rad Laboratories (Hercules, CA) using the Bio-Plex

200 System (Bio-Rad Laboratories).

3.3.6 Neutrophil staining

Paraffin-embedded, left lateral liver lobes were stained for neutrophils
(PMNs) using a rabbit anti-PMN lg isolated from the serum of rabbits immunized
with rat PMNs (101). lmmunohistochemical staining of PMNs was done using the
protocol as described previously (219). Neutrophil accumulation was then
quantified by counting PMNs per high power field. The slides were coded,

randomized and then visualized using a light microscope.

91

3.3.7 Statistical analyses

Results are presented as mean 1 S.E.M. A student’s t-test or a 2-way
analysis of variance (ANOVA) was used as appropriate after data normalization.
All pairwise comparisons were made using a Tukey test with the criterion for

significance at p < 0.05.

92

3.4 Results

3.4.1 TVX coexposure with either LPS or PGN-LTA causes hepatotoxicity

TVX (150 mglkg) administration 3 h before either LPS or PGN-LTA
mixture caused a significant increase in plasma ALT activity as early as 4.5 h
after the inflammatory stimulus (Fig. 3.1). It increased to a maximum at 15 h after
either inflammatory stimulus. Administration of either LPS or PGN-LTA without
TVX did not cause an increase in plasma ALT activity (data not shown).

The histopathology of the respective treatment groups corroborated the
plasma ALT activity results. TVX/PGN-LTA coexposure resulted in hepatocellular
oncotic necrosis and apoptosis primarily in the centrilobular regions (Fig. 3.2). In
contrast, TVX/LPS coexposure caused hepatocellular death primarily in midzonal

regions (Fig. 3.2).

3.4.2 TVX enhances cytokine induction by either inflammatory stimulus
Administration of either LPS or PGN-LTA caused a significant increase at
4.5 h of plasma concentrations of the following cytokines: lL-1B, TNFa, lL-10, IL-
6, lL-18, lFNy, vascular endothelial growth factor (VEGF), and MCP-1 (Fig. 3.3).
TVX pre-treatment enhanced the LPS-induction of all cytokines listed above (Fig.
3.3). In contrast, TVX treatment enhanced the PGN-LTA-induced increase in all
of the cytokines listed except for TNFa or IFNy at 4.5 h (Fig. 3.3). In addition,

TVX enhanced the induction of chemokines by both LPS and PGN-LTA (Fig. 3.4).

93

Fig. 3.1. Development of liver injury after TVX/LPS or TVXIPGN-LTA
coexposure. Mice were treated with TVX (150 mglkg; p.o.) 3 h before LPS (2 x
106 EU/kg; i.p.) or PGN-LTA (30 mglkg each, i.p.). Mice were killed at various
times and plasma ALT activity was measured. n = 5 animals/group. * significantly

different from same treatment group at 0 h.

94

ALT activity (UIL)

9000 r
7500 ‘
6000 -

4500 , I
3000 /

1500 -

 

- TVX/LPS
= TVXIPGN-LTA

 

0 1 .5 3 4.5 6 15

Hours after inflammatory stimulus

95

Fig. 3.2. Histopathology of livers from mice treated with TVX/LPS or
TVXIPGN-LTA. Mice were treated with TVX (150 mglkg; p.o.) 3 h before either
LPS (2 x 106 EU/kg; i.p.) or PGN-LTA (30 mglkg each, i.p.). Mice were killed at

15h and representative photomicrographs were taken.

96

 

...c. Ind/rune. . levfimfiwa. .. t ...q.
Eucwamwcflwwws} “teammate. .
“flarxnwmmaawmflz. 3...»... L wwwwrauvrifi
2......” {H3473}? .w... 3.51.03... .
A i... . .. sci... . Texcumao. «an... LE
nova! nut—wet“ Dru. [finer/Mm: (Inn-r1 ..
. flusAflrfivfisJ.’ mmwhflfihkvil. .

. 119%.”...fl” reanw. lawman} .. .. Qt.

   
   

H rr a h- .u...

.. «Writ». \yw‘ 1.1/WM... I

«News! as» .
..

JG . #5. u . .....3

 

     

r... a
. J

   

   

 

 

 

97

Fig. 3.3. Effect of TVX pretreatment on LPS- and PGN-LTA-induced
increases in cytokines. Mice were treated with TVX or Veh and then LPS,
PGN-LTA or Veh as described in Section 3.3.3. Mice were killed at 4.5 h, and
plasma concentrations of lL-1B, TNFa, lL-6, IL-18, VEGF, MCP-1, lL-10 and lFNy
were measured. n = 4-6 animals/group. * significantly different from VehNeh-
treated group. * significantly different from TVXNeh-treated mice. “’significantly

different from Veh-treated mice within the same treatment group.

98

-Veh

 

 

 

 

 

 

 

 

 

 

 

 

 

.5
..l
2
it
..I
.r:
0
>
3 § § § E °
('luliedpogm
75 .
-r . 4::
i “i
‘0'. L 4
a) at: m r .
5 . I
f.
>
-. , , 1 ' . . a , f . e
t i t .2— ° ti: 3 2 i.
(10”ij g P." ('I lllle) 9""

99

LPS PGN-LTA

Veh

LPS PGN-LTA

Veh

Fig. 3.3 (cont’d).

- Veh
F.“

 

 

 

 

 

 

 

 

.- IE 0- IS
ash—P ...-I at .‘l
2 5
2 0.
. _

ae+ 2 2
_I _l
8
.g s

sass-38° §§§§§§°

('l ullfiu) I-dow (1w16d) 0l-'Il

 

#0
l_.

LPS PGN-LTA
*
LPS PGN-LTA

Veh
Veh

 

 

s: g s s ° recess"
(1 wlfidl :IOBA l1 wrfid) 1- UOJOIJGIUI

100

Fig. 3.4. Effect of TVX pretreatment on LPS- and PGN-LTA- induced
increases in chemokines. Mice were treated with TVX or Veh and then LPS,
PGN-LTA or Veh as described in Section 3.3.3. Mice were killed at 4.5 h, and
plasma concentrations of KC, MlP-2 and MlP-1or were measured. n = 4-6
animals/group. * significantly different from VehNeh-treated group. " significantly
different from TVXNeh-treated mice. ‘D significantly different from Veh-treated

mice within the same treatment group.

101

 

 

 

 

 

m
M
_m
Mil—ll.
....
... a .. . a .
SE‘OEOV.

"l?

 

#¢

 

 

 

 

 

 

”warm

3835 urn—=2

 

o

Veh

 

 

mmmmo

32:03 8.3::

LPS PGN-LTA

Veh

102

3.4.3 Effect of TVX on microbial stimuli-induced hepatic neutrophil
accumulation

Hepatic neutrophil accumulation was evaluated at 4.5 h after LPS or PGN-
LTA. Both LPS and PGN-LTA alone caused a significant increase in neutrophils
present in the liver (Fig. 3.5). TVX alone did not cause hepatic neutrophil
accumulation. In addition, TVX pre-treatment did not affect the number of

neutrophils present in the liver after either inflammatory stimulus (Fig. 3.5).

3.4.4 Effect of CD18 neutralization on TVXILPS- and TVXIPGN-LTA-induced
liver injury and inflammation

CD18 antiserum was administered as described in Section 3.3.3. CD18
neutralization protected mice from TVXILPS- and TVX/PGN-LTA-induced liver
injury as measured by plasma ALT activity (Fig. 3.6).

0018 antiserum administration significantly attenuated TVX/PGN-LTA-
induced hepatic neutrophil accumulation (Fig. 3.7). In addition, CD18
neutralization reduced the TVX/PGN-LTA-induced increases in TNFa and MOP-1
(Fig. 3.8). In contrast, CD18 neutralization did not affect TVXILPS-induced
hepatic neutrophil accumulation (Fig. 3.7). Similarly, CD18 neutralization did not

affect any cytokines induced by TVX/LPS coexposure (Fig. 3.8).

103

Fig. 3.5. Effect of TVX pretreatment on LPS- and PGN-LTA- induced hepatic
neutrophil accumulation. Mice were treated with TVX or Veh and then LPS,
PGN-LTA or Veh as described in Section 3.3.3. Mice were killed at 4.5 h.
Paraffin-embedded liver lobes were cut and stained for PMNs. n = 4-6
animals/group. * significantly different from VehNeh-treated group. ‘ significantly

different from TVXNeh-treated mice.

104

 

 

 

 

“E: .8. czsd

LPS PGN-LTA

Veh

105

Fig. 3.6. Effect of CD18 neutralization on TVXILPS- and TVXIPGN-LTA-
induced liver injury. Mice were treated with TVX/LPS or TVX/PGN-LTA and
CD18 antiserum or control serum as described in Section 3.3.3. Mice were killed
at 15 h and plasma ALT activity was measured. n = 6-10 animals/group. *

significantly different from control serum treated mice within the same treatment

group.

106

8000 .
- Control
Em CD18 antiserum

5

 

 

\\

88

1000 -

ALT activity (UIL)
88 8

  

 

 

 

TVX/LPS TVXIPGN-LTA

107

Fig. 3.7. Effect of CD18 neutralization on TVXILPS- and TVXIPGN-LTA-
induced hepatic neutrophil accumulation. Mice were treated with TVX/LPS or
TVX/PGN-LTA in addition to CD18 antiserum or control serum as described in
Section 3.3.3. Mice were killed at 4.5 h. Paraffin-embedded liver lobes were
stained for PMNs. n = 6-10 animals/group. 1‘ significantly different from control

serum treated mice within the same treatment group.

108

Hepatic PMNs per HPF
.5
O

— Control
— CD18 antiserum

   
   

. .
" J
. ,
'Ds
5“
- L'." .
35'; l
.
-I~ ., r.
I I J ..

 

 

 

 

TVX/LPS TVXIPGN-LTA

109

Fig. 3.8. Effect of CD18 neutralization on TVXILPS- and TVXIPGN-LTA-
induced cytokine increases. Mice were treated with TVX/LPS or TVX/PGN-
LTA in addition to CD18 antiserum or control serum as described in Section 3.3.3.
Mice were killed at 4.5 h and plasma cytokine concentrations were measured.
Control and CD18 antiserum control mice had equivalent baseline concentrations
of IL-6 (0.05 1 0.005 ng/mL), lL-10 (36 1 6 pg/mL), TNFor (411 1 120 pg/mL), IL-
18 (75 1 15 pg/mL), KC (0.04 1 0.006 ng/mL), MIP-1a (541 1 56 pg/mL), lFNy (9
1 2 pg/mL) and MOP-1 (0.14 1 0.03 ng/mL). n = 6-10 animals/group. 1‘
significantly different from control serum treated mice within the same treatment

group.

110

TVXILPS TVXIPGN-LTA

  

 

 

gasses:

('I wrfld) Ol-"II (‘I wrfid) 9 VII

TVXILPS TVXIPGN-LTA

 

 

 

 

U r U U U V V c
G O G O 0 O
O O G 1' N
1- e- 1-

(1 IIJIDU) 9-'u ('l wlfidl 1°le

140 a
120

111

Fig. 3.8 (cont’d).

{:1 CD18 antiserum

-Conlrol

TVX/LPS TVXIPGN-LTA
TVX/LPS TVXPGN-LTA

 

 

 

 

ééé" itéééé

(1wl5di 1ol-cllllil (1wr6u) l-dow

TVXILPS TVXIPGN -LTA
TVXILPS TVXIPGN-LTA

 

 

 

geese:- estates
(1wrlu)o)l (‘lullfidHNdl

112

3.5 Discussion

Inflammatory episodes are commonplace and occur sporadically. We
hypothesized that an inflammatory stress can decrease the toxicity threshold to
certain drugs, precipitating an idiosyncratic toxicity. Previous studies showed that
TVX enhanced the LPS-induced plasma TNFa increase and that TVX/LPS
coexposure resulted in hepatotoxicity (216). The studies presented here
examined whether TVX enhanced the LPS-induced increases in other cytokines.
In addition, they explored whether TVX enhanced the cytokine response to gram-
positive microbial stimuli, PGN and LTA, and if TVX/PGN-LTA coexposure was
hepatotoxic to mice.

Inflammation is a complex process which can be initiated by the host’s
recognition of microbial products by toll-like receptors. Bacterial components can
be measured in the plasma and are increased by a number of stressors including
alcohol consumption, surgery and gastrointestinal disturbances (61). Bacterial
components of both gram-positive bacteria (PGN and LTA) and gram-negative
bacteria (LPS) have been measured in the plasma and activate toll-like receptors
to induce inflammation. PGN and LTA activate TLR2 and induce NFch activation
through My088-dependent mechanisms (217, 220-222). As described
extensively in Section 1.2.1, LPS activates TLR4 and induces NFch activation
through both My088-dependent and —independent mechanisms.

A nontoxic dose of TVX was rendered hepatotoxic upon coexposure to a
nontoxic dose of either LPS or PGN-LTA (Fig. 3.1). The timecourse of

hepatotoxicity for TVX/LPS- and TVX/PGN-LTA-induced liver injury was similar,

113

inasmuch as plasma ALT activity increased as early as 4.5 h and continued to
progress until 15 h. That TVX interacted with either TLR2- or TLR4-activating
ligands to cause liver injury proves that the TVX/inflammation-induced liver injury
shown previously (216) is not specific to TLR4 activation. Indeed, it suggests that
TVX interacts with an inflammatory stress, irrespective of its source, to
precipitate liver injury. The result suggests that inflammatory stress induced by
either gram-positive or gram-negative bacteria might play a role in TVX
hepatotoxicity.

Despite a similar timecourse of liver injury, the histopathology differed
between TVX/LPS- and TVX/PGN-LTA-induced liver injury (Fig. 3.2). TVX/LPS-
treated mice developed lesions of hepatocellular necrosis and apoptosis primarily
localized to midzonal regions, whereas TVXIPGN-LTA lesions of hepatocellular
necrosis and apoptosis were primarily centrilobular regions. Such a difference in
localization might be due to a difference in TLR2 and TLR4 expression in regions
of the mouse liver, of which little is known. Another possibility is that the
difference in the location of the lesions suggests a difference in the mechanism
of pathology. Therefore, to examine possible mechanisms of pathogenesis,
inflammatory cytokines were measured at 4.5 h, the onset of liver injury, to
determine if TVX enhanced cytokine release in response to bacterial stimuli.

TVX enhanced the LPS- and PGN-LTA-induced increases of several
cytokines: IL-1B, lL-6, lL-18, VEGF, MCP-1 and lL-10. In contrast, only the LPS-
induced increase of TNFa and IFNy was enhanced by TVX, and it did not affect

the PGN-LTA induction of these cytokines (Fig. 3.3). These cytokines were only

114

measured at 4.5 h, therefore it is possible that TVX enhanced PGN-LTA-induced
increases in TNFa and IFNy at other times. Indeed, TVX pretreatment caused a
trend towards an increase in TNFa in PGN-LTA-treated mice, but the difference
was not statistically significant at this time. lFNy induction by LPS, but not PGN-
LTA, was enhanced by TVX, a result possibly related to the differences in TLR2-
and TLR4-activation. TLR4, but not TLR2 activation induces dendritic cells to
produce lL-12 (223, 224), which directly stimulates natural killer and T cells to
produce IFNy (225). It is thus possible that TVX acts to enhance a number of the
steps in this pathway of IFNy production induced by LPS which would not be
activated by PGN-LTA.

Similar to a number of cytokines, TVX enhanced the LPS- and PGN-LTA-
induced increase in KC, MIP-2 and MlP-1a (Fig. 3.4), all of which can be
upregulated as a result of NFIcB activation (226). It is thus likely that the common
upregulation of cytokines by both stimuli was mediated through NFxB activation.
KC, MIP-2 and MlP-10t all have chemotactic activity for neutrophils; therefore,
hepatic PMN accumulation was quantified to determine if the TVX enhancement
of chemokines was associated with an increase in LPS- and/or PGN-LTA-
induced PMN accumulation. Despite an increase in several chemokines, TVX
pretreatment did not affect LPS- or PGN-LTA-induced neutrophil accumulation in
the liver (Fig. 3.5). After LPS/galactosamine treatment, neutrophils roll and
adhere in hepatic postsinusoidal venules independently of KC or MlP-2, but
extravasation of neutrophils into the parenchyma was significantly reduced by

MlP-2 or KC neutralization (227). It is thus possible that TVX enhances LPS- and

115

PGN-LTA-induced PMN extravasation into the parenchyma and in turn PMN
activation. This would not be detected by PMN staining.

To determine if PMN activation is involved in TVX/LPS- or TVXIPGN-LTA-
induced liver injury, mice were pretreated with a neutralizing antibody to CD18, a
82-integrin critical for PMN activation (102, 228). CD18 neutralization attenuated
hepatotoxicity induced by either TVX/LPS or TVX/PGN-LTA coexposure (Fig.
3.6); therefore, PMN activation appears to be a common pathway required for the
progression of liver injury. Neutrophil activation is required for TVX/Inflammation-
induced liver injury in both rats and mice (58), which suggests that this
requirement is not species-specific. It therefore might be an important pathway in
TVX-induced hepatotoxicity in people.

The mechanism of hepatic neutrophil accumulation after TVX/LPS or
TVX/PGN-LTA coexposure is different, inasmuch as CD18 neutralization
reduced TVX/PGN-LTA- but not TVX/LPS-induced PMN accumulation (Fig. 3.7).
This finding that LPS-induced hepatic neutrophil accumulation is CD18-
independent is consistent withprevious reports (229). However, that PGN-LTA-
induced hepatic neutrophil accumulation is CD18-dependent has not been
reported to our knowledge. Further studies are required to understand the
difference in mechanisms of PMN accumulation between these two stimuli.

TVX enhanced the induction of cytokines by either LPS or PGN-LTA,
therefore cytokines were measured to determine if CD18 neutralization affected
the induction of proinflammatory cytokines. CD18 neutralization did not

significantly affect the TVX/LPS induction of lL-6, lL-10, TNFa, IL-1B, KC, MlP-1a,

116

lFNy or MCP-1 (Fig. 3.8). Previously, TNFa was found to be critical for TVXILPS-
induced liver injury (216). The finding that CD18 neutralization did not affect
TNFot induction suggests that PMN activation and TNFor activity either represent
separate hepatotoxic pathways or that PMN activation is downstream of TNFa in
a pathway involved in the pathogenesis in TVXILPS-cotreated mice.

Similar to TVX/LPS, CD18 neutralization did not affect the TVX/PGN-LTA
induction of lL-6, lL-10, lL-1B, KC, MlP-1a or IFNy. However, it did attenuate the
TVX/PGN-LTA-induced increase in TNFa and MOP-1 (Fig. 3.8). Since lL-6, KC
and lL-10 were unchanged by CD18 neutralization, Kupffer cell activation was
probably not affected, and the attenuation of TNFor and MCP-1 may be
neutrophil-dependent. Neutrophils can produce and release MCP-1, and the
depletion of neutrophils significantly reduced MOP-1 production induced by the
injection of apoptotic cells into the peritoneal cavity of mice (230, 231 ). Therefore,
it is likely that the attenuation of MCP-1 induction by CD18 neutralization is due
to the reduction in hepatic neutrophil accumulation. Similarly, the reduction in
TNFa concentration by CD18 neutralization might also be due to the attenuation
of hepatic neutrophil accumulation, inasmuch as neutrophils express TACE on
their extracellular membrane. This enzyme is critical for TNFa cleavage and
release (232). Whether TVX/PGN-LTA-induced liver injury requires TNFa is
unknown, therefore protection by CD18 neutralization could be due to PMN
inactivation or to reduced concentrations of TNFa.

In summary, TVX synergized with a modest inflammatory stress induced

by either a gram-negative or a gram-positive stimulus to cause liver injury in mice.

117

TVX enhanced the LPS- and PGN-LTA-induced increases in proinflammatory
cytokines. However, TVX did not enhance the hepatic neutrophil accumulation
driven by either of these stimuli. CD18 neutralization attenuated TVX/LPS- and
TVX/PGN-LTA-induced liver injury, suggesting that PMN activation plays a
critical role in injury progression in both models. CD18 neutralization attenuated
TVX/PGN-LTA induced increases in hepatic neutrophil accumulation and TNFa
and MOP-1 plasma concentrations. In contrast, it did not affect TVX/LPS-induced
hepatic neutrophil accumulation or proinflammatory cytokine increases. The
results suggest that inflammatory stress induced by either a gram-positive or
gram-negative bacterial products could play a role in the development of TVX-

induced hepatotoxicity and that the pathogenesis is CD18-dependent.

118

CHAPTER 4

Shaw, P.J., Ganey, PE. and Roth, R.A. (2008). TNFa acting at both p55 and

p75 receptors is essential for synergistic hepatotoxicity from TVX/LPS

coexposure. Submitted to JPE T

119

4.1 Abstract

The use of trovafloxacin (TVX), a fluoroquinolone antibiotic, was severely
restricted clue to an association of TVX therapy with idiosyncratic hepatotoxicity
in patients. The mechanisms underlying idiosyncratic toxicity are unknown;
however, one hypothesis is that an inflammatory stress can render an individual
sensitive to the drug. Previously, we reported that treatment of mice with TVX
and lipopolysaccharide (LPS) induced tumor necrosis factor a (TNFa)-dependent
liver injury, whereas TVX or LPS treatment alone was nontoxic. The goal of this
study was to elucidate the role of TNFa in TVX/LPS-induced liver injury. p55"'
(TNFR1) and p754' (T NFR2) mice were protected from hepatotoxicity caused by
TVX/LPS coexposure, suggesting that TVX/LPS-induced liver injury requires
both TNF receptors. TNFor inhibition using etanercept significantly reduced the
TVX/LPS-induced increases in the plasma concentrations of several cytokines
around the time of onset of liver injury. However, despite the reduction in
chemokines, etanercept treatment did not affect the TVX/LPS-induced hepatic
accumulation of neutrophils. In addition, etanercept treatment attenuated
TVX/LPS-induction of plasminogen activator inhibitor-1 (PAl-1), and this was
associated with a reduction in hepatic fibrin deposition. Mice treated with TVX
and a nontoxic dose of TNFa also developed liver injury. In summary, TNFa acts
through p55 and p75 receptors to precipitate an innocuous inflammatory cascade.

TVX enhances this cascade, converting it into one that results in hepatocellular

injury.

120

4.2 Introduction

Trovafloxacin (TVX), a fluoroquinolone antibiotic, is one example of a drug
for which use was restricted severely due to lADRs. TVX was approved for use in
the US. in 1997, and by 1999 its use was associated with 152 cases of serious
hepatic events. Of these, 14 resulted in acute liver failure, 5 patients required
liver transplants and 4 died (233).

One hypothesis regarding the cause of lADRs is that inflammatory stress
alters the toxicity threshold of an individual, rendering a normally therapeutic
dose of a drug toxic (55, 73). In accordance with this hypothesis, nontoxic doses
of TVX and bacterial lipopolysaccharide (LPS) synergized to cause acute liver
injury in both rats and mice (58, 216). In this animal model, TVX pretreatment
enhanced the LPS-induced peak in plasma TNFa concentration. In addition,
TNFa neutralization completely protected mice from TVX/LPS-induced liver injury
(216).

TNFot is a pleiotropic cytokine that stimulates a number of cellular
responses, including proliferation, production of inflammatory mediators,
upregulation of adhesion molecules and programmed cell death. Large amounts
of TNFa are produced in response to several microbial products, including LPS.
TNFa is a key mediator of inflammatory responses, which can result in both
tissue damage and host defense (74, 75). The main cellular source of TNFa is
macrophages, but several other cell types produce TNFa including mast cells,

hepatic stellate cells, endothelial cells, fibroblasts and neuronal cells (80, 81).

121

TNFa plays a critical role in several models of liver injury caused by viral hepatitis,
ischemialreperfusion or hepatotoxic doses of LPS (76, 78, 79, 209).

The biological effects of TNFa are elicited via two high affinity cell surface
receptors, p55 (T NF-R1) andp75 (T NF-R2) (83). The role of each receptor has
been evaluated in several models of liver injury. The p55 receptor has been
studied more extensively and is important in hepatotoxicity caused by LPS,
acetaminophen or carbon tetrachloride (77, 95, 96). In contrast, critical roles for
both receptors have been shown only in a few models of hepatotoxicity, such as
that induced by concanavalin A, Pseudomonas aeruginosa exotoxin A or
adenovirus (234-236). The study presented here was designed to determine the
importance of each receptor in TVXILPS-induced liver injury and to evaluate the

influence of TNFor on other proinflammatory factors in this lADR model.

122

4.3 Materials and Methods

4.3.1 Materials
Recombinant murine TNFa was purchased from R&D Systems

(Minneapolis, MN). Please refer for Section 2.3.1 for additional information on

this topic.

4.3.2 Animals
p55"', p75"', and CS7/Bl6 wild-type controls were purchased from Jackson
Laboratory (Bay Harbor, ME). Please refer to Section 2.3.2 for additional

information on this topic.

4.3.3 Experimental protocols

Mice fasted for 12 h were given TVX (150 mglkg) or Veh (saline) by oral
gavage. They were then given LPS (2.0 x 106 EU/kg), TNFa (50 pg/kg)) or Veh
(saline) by intraperitoneal injection 3 h later. Food was returned immediately after
this closing. Mice were anesthetized with sodium pentobarbital (50 mglkg; i.p.)
and killed at designated times after LPS, TNFa or Veh for various measurements.
Blood was drawn from the vena cava into a syringe containing sodium citrate,
resulting in a final concentration of 0.76%. The left lateral lobe was fixed in 10%

neutral buffered formalin and paraffin blocked.

123

4.3.4 Histopathology

Please refer to Section 2.3.4 for information on this topic.

4.3.5 Cytokine measurements

Please refer to Section 3.3.5 for information on this topic.

4.3.6 Neutrophil staining

Please refer to Section 3.3.6 for information on this topic.

4.3.7 Hemostatic system measurements

Plasma thrombinzantithrombin lll (TAT) dimers were measured using the
Enzygnost TAT ELISA kit purchased from Dade Behring (Marburg, Germany).
Active PAl-1 plasma concentration was measured using an ELISA kit purchased
from Molecular Innovations, Inc. (Novi, MI). Hepatic fibrin immunohistochemistry
and estimation of deposition was done following the protocol described
previously with a slight modification (237), ie, artifactual fibrin staining seen
within vessel lumens in all treatment groups was removed from quantification

calculations.

4.3.8 Statistical analyses

Results are presented as mean 1 S.E.M. A student’s t-test or a 2-way

analysis of variance (ANOVA) was used as appropriate after data normalization.

124

All pairwise comparisons were made using a Tukey test with the criterion for

significance at p < 0.05.

125

4.4 Results

4.4.1 p55"' and p75"' mice are protected from TVXILPS-induced liver injury
To determine the contribution of each TNF receptor to TVX/LPS-induced
liver injury, p55"' and p75" mice were treated with TVX/LPS as described in
Materials and Methods. TVX/LPS coexposure caused significant liver injury at 15
h in control (wild-type) mice. Both p55"' and p75"' mice were resistant to
TVX/LPS-induced liver injury (Fig. 4.1). p75"' mice were completely protected
from TVXILPS-induced liver injury and had significantly reduced plasma ALT
activity compared to control and p55”' mice (Fig. 4.1). Histopathologic
examination of livers corroborated this result, inasmuch as lesions of
hepatocellular necrosis were decreased in p55"' compared to control mice, and

were completely absent in p75"' mice (Fig. 4.1).

4.4.2 TNFa neutralization attenuates TVXILPS-induced inflammatory
cytokines and chemokines

In a previous study, treatment with etanercept, which is a mimic of the
soluble p75 receptor, reduced TVX/LPS-induced increase in plasma TNFa
concentration and protected mice from TVX/LPS-induced liver injury (216).
TVX/LPS-treated mice were dosed with etanercept to determine the effects of
TNFa on the induction of proinflammatory cytokines and chemokines at 4.5 h, a
time near the onset of liver injury (see Fig. 3.1). TNFa inhibition attenuated the

TVX/LPS—mediated induction of IFNy, IL-6, IL-10, MCP-1 and VEGF (Fig. 4.2).

126

Fig. 4.1. The role of TNF receptors in TVXILPS-induced liver injury. Wild-
type, p55"' and p75"' mice were treated with TVX 3 h before LPS as described in
Materials and Methods. Mice were killed 15 h after LPS, and plasma ALT activity
was measured. Photomicrographs were taken of livers from representative mice
from each group. n = 5-8 animals/group. *significantly different from wild-type

group; # significantly different from p55"' group.

127

 

 

 

as I
.8. I

0.5.

 

\\

 

§
('III'I) 1.1V

 

128

Fig. 4.2. Effect of TNFa inhibition on TVXILPS-induced increases in plasma
cytokines. Mice were treated with vehicles or with TVX/LPS in addition to
etanercept or its vehicle as described in Materials and Methods. Mice were
sacrificed 4.5 h after LPS administration. Plasma concentrations of lFNy, lL-6, IL-
10, IL-1B, MCP-1 and VEGF were measured as described in Materials and
Methods. n = 4-6 animals/group. *significantly different from respective VehNeh;

# significantly different from TVXILPSNeh group.

129

an
...
...

 

 

m w w o

2 1

sees s...

 

 

m m m m ..

353a. » caste...

TVX/LPS

VehNeh

TVX/LPS

VehNeh

 

 

 

 

 

 

 

.Tm
m m m m o
seesaw...
..iT|.ll
.i
Tm
.
W m m m ..
38332....

TVX/LPS

VehNeh

TVX/LPS

VehNeh

*#

TVX/LPS

VehNeh

-r‘1

 

 

m m m o

3.52.. mom;

a»

.TT
N m w w o
..

3 £35 7&0!

 

 

 

TVX/LPS

VehNeh

j Etanercept

- Veh

“......

130

The increase in lL-1B plasma concentration following TVX/LPS treatment was not
changed by etanercept treatment (Fig. 4.2). TNFa neutralization significantly

reduced the TVX/LPS induction of chemokines MlP-2, KC and MlP-1a (Fig. 4.3).

4.4.3 TVXILPS-induced hepatic neutrophil accumulation is independent of
TNFa

Previous results pointed to a role for PMNs in the pathogenesis of
TVX/LPS-induced liver injury (Chapter 3). Despite causing a reduction in
chemokines (Fig. 4.3), etanercept treatment did not reduce hepatic neutrophil
accumulation induced by TVX/LPS coexposure. In fact, etanercept treatment

slightly increased neutrophil accumulation (Fig. 4.4).

4.4.4 TNFa neutralization attenuates TVXILPS-induced hemostatic system
acfivafion

The coagulation system plays an important role in TVX/LPS-induced
pathogenesis (presented in Chapter 6). To determine if TNFa plays a role in
TVX/LPS-induced coagulation system activation, TVX/LPS-treated mice were
treated with etanercept and killed at 4.5 h. The dose of etanercept markedly
reduced the TVX/LPS-induced release of TNFa in this model (216). Plasma
thrombin-antithrombin (TAT) dimers, measured as a biomarker of coagulation
system activation, were significantly increased in TVX/LPS-treated mice (Fig.
4.5A). Etanercept treatment caused a trend toward reduction in plasma TAT

dimers, but this difference was not statistically significant. The plasma

131

Fig. 4.3. Effect of TNFa inhibition on TVXILPS-induced increases in plasma
chemokines. Mice were treated with vehicles or with TVX/LPS in addition to
etanercept or its vehicle as described in Materials and Methods. Mice were
sacrificed 4.5 h after LPS administration. Plasma concentrations of MlP-2, KC
and MIP-1or were measured as described in Materials and Methods. n = 4-6
#

animals/group. *significantly different from respective VehNeh group;

significantly different from TVXILPSNeh group.

132

KC (nglm L) MlP-2 (nglmL)

MIP-1a (pglm L)

..L
N
O

 

 

 

 

90 ‘ l=l Etanercept
00 .
* #
3o .
0 _— __
VehNeh TVX/LPS
400 .
— Veh
300 l Etanercept |
200 - I a #
100 ‘ l a
0 __ —_
VehNeh TVX/LPS
600
500 l — Veh
I=I Etanercept
400
*#
300
200

..L
O
O

 

O

 

VehNeh TVX/LPS

133

Fig. 4.4. Effect of TNFa inhibition on TVXILPS-induced hepatic PMN
accumulation. Mice were treated with vehicles or with TVX/LPS in addition to
etanercept or its vehicle as described in Materials and Methods. Mice were
sacrificed 4.5 h after LPS administration. Paraffin-embedded livers were stained

for neutrophils, and the number of neutrophils was quantified as described in

Materials and Methods. n 4-6 animals/group. *significantly different from

respective VehNeh group; * significantly different from TVXILPSNeh group.

134

PMNs per HPF
d N 03 h 0| 0’ N on
O O O O O O O O O

A; L

- Veh

 

a Etanercept

 

 

 

 

 

VehNeh

135

 

TVX/LPS

Fig. 4.5. Effect of TNFa inhibition on hemostatic system dysregulation
mediated by TVX/LPS coexposure. Mice were treated with vehicles or with
TVX/LPS in addition to etanercept or its vehicle as described in Materials and
Methods. They were sacrificed 4.5 h after LPS administration. Plasma
concentrations of (A) TAT dimers and (B) active PAl-1 were measured as
described in Materials and Methods. (C) Hepatic fibrin deposition was stained
immunohistochemically and quanitified as described in Materials and Methods. n
= 4-6 animals/group. *significantly different from respective VehNeh group;

1"significantly different from TVXILPSNeh group.

136

300

 

 

    

 

 

 

250 ‘
3
E 200 -
l— 150-
<
l-

100 ‘

50 ‘ ~
VehNeh TVXILPS

1200 - *
3
E 1000 ‘ _ Veh
g 800 = Etanercept
‘7 600
E
n. *#
g 400
‘6' 200i
<

o ‘*
VehNeh TVXILPS

C A
o g 0014
IE- 0 *
u .5 0012 « _ Veh
a a a: Etanercept
o g 0010- l
1: ._

o.-
.E “a 0.008-
3 o
t; :- 0.006 . #
7, o
In c 0.0“ ‘
'5 .9
g '6 0002-
55, g 0.000 J

VehNeh TVXILPS

137

concentration of active PAl-1, an inhibitor of the fibrinolytic system, was
increased by TVX/LPS coexposure (Fig. 4.58). Etanercept significantly reduced
the TVX/LPS induction of plasma active PAl-1 (Fig. 4.58). Fibrin deposition in
tissue occurs if the rate of coagulation system activation exceeds the rate of
fibrinolysis. TVX/LPS coexposure caused a significant increase in sinusoidal
fibrin deposition in the liver at 4.5 h, which was significantly reduced by

etanercept treatment (Fig. 4.5C).

4.4.5 TVX and TNFct coexposure causes hepatotoxicity

TVXILPS-induced liver injury is dependent on TNFa (216), but to
determine if TNFa alone could interact with TVX, mice were treated with TVX
and recombinant murine TNFa as described in Materials and Methods. They
were killed 15 h after TNFa treatment, the time of maximal plasma ALT activity in
TVXILPS-treated mice. TVX or TNFot treatment alone did not cause an increase
in plasma ALT activity (Fig. 4.6). However, TVXfTNFor coexposure increased
plasma ALT activity. Histopathological evaluation of liver sections corroborated
the lack of injury from TVX or TNFa alone (Fig. 4.7). In contrast, TVX/TNFa
coexposure caused hepatocellular necrotic and apoptotic lesions primarily in

centrilobular and midzonal regions of liver lobules, and these lesions extended to

periportal regions in some severely affected mice.

138

Fig. 4.6. TVXfI’NFa coexposure-induced liver injury. Mice were treated with
TVX 3 h before recombinant murine TNFa as described in Materials and

Methods. Mice were killed 15 h after TNFa administration, and plasma ALT

activity was measured. n - 4-5 animals/group. *significantly different from

TVXNeh group; " significantly different from Veh/TNFa group.

139

ALT activity (UIL)

8
O
O

2000 -

1000 .

 

— Veh
I: TVX

 

 

 

 

Veh TNFa

140

Fig. 4.7. Histopathology of TVXITNFa-induced liver injury. Mice were treated
with TVX 3 h before recombinant murine TNFa as described in Materials and
Methods. Mice were killed 15 h after TNFor administration, and photomicrographs

were taken of representative livers.

141

 

 

       

 

    

 

 

142

4.5 Discussion

Previously, we reported that a nontoxic dose of TVX interacts with a
nontoxic dose of LPS to cause TNFa-dependent liver injury in mice (216). The
critical role of TNFa in TVXILPS-induced liver injury was based upon TNFa
neutralization, however the role of each TNF receptor was not studied. Activation
of the p55 receptor results in two main signals: NFch activation and activation of
caspases leading to apoptosis (238, 239). Inasmuch as ligation of the p55
receptor can result in NFKB activation and cell death, it is not surprising that the
p55 receptor is critical to several models of liver injury (77, 95, 96, 234-236).
Similar to other models dependent on TNFor and p55, TVXILPS-induced liver
injury was significantly attenuated in p554' mice. It is possible that p554' mice
have reduced liver injury due to decreased plasma TNFoc concentrations, since
TNFor induction by LPS was found to be attenuated in p55"‘ mice (96). However,
p55 was not involved in the production of TNFa by galactosamine/LPS
coexposure (240).

The function of the p75 receptor is less understood compared to the p55
receptor. Similar to p55, activation of the p75 receptor causes NFch activation,
but it does not result in caspase activation (241). The critical role of the p75
receptor in hepatotoxicity is unclear: it is not involved in some models of liver
injury that are dependent on TNFa (96, 240) but is involved in others (234-236).
Indeed, p75"' mice were completely protected from TVXILPS-induced liver injury

and had significantly reduced hepatocellular injury compared to p554‘ mice.

143

Why p75"' mice are completely protected from TVXILPS-induced
hepatotoxicity is unclear. It is possible that the p75 receptor is playing two roles
in the progression of TVXILPS-induced hepatotoxicity. It has been suggested
that the p75 receptor acts to bind TNFa and transfer it to the p55 receptor,
resulting in p55 activation at lower concentrations of TNFa (242). Therefore,
without the p75 receptor present, the threshold for TNFa—dependent liver injury
might be higher than the concentration that is achieved. Additionally, the p75
receptor can cooperate with the p55 receptor to enhance necrotic cell death in
response to TNFa (243). To account for the minor hepatotoxicity seen in p55”'
mice, p75 activation must also cause minor hepatocellular injury independent of
p55 following TVX/LPS coexposure. Therefore, it is possible that combined
activation of p55 and p75 synergize to cause extensive hepatocellular necrosis
which is not seen when either receptor is absent. In addition, the p75 receptor is
involved in the LPS-induced production of TNFa (96). It is possible that the
complete protection in p75”' mice resulted from a combination of these
mechanisms, inculding a reduction in LPS-induced TNFa production.

TNFa was involved in the TVXILPS-mediated increases in several
inflammatory cytokines: lFNy, lL-6, MCP-l, VEGF, MlP-2, KC and MIP-10t. The
attenuation of lL-6 and MlP-2 after TNFot neutralization was also seen in another
model of drug/LPS coexposure-induced hepatotoxicity (244). The reduction of
such a large number of cytokines might be due to a decrease in TNFor-driven

NFKB activation mediated through p55 and p75 receptor activation (238, 241).

144

Several of the cytokines that required TNFa for their release have
chemotactic properties. Therefore, we measured PMN accumulation in livers of
these mice. The hepatic accumulation of PMNs induced by TVX/LPS was not
decreased by TNFa inhibition. It is thus likely that the TVXILPS-induced hepatic
PMN accumulation is mediated by selectins and other adhesion molecules as
seen in endotoxemia or by TNFa-independent sinusoidal contraction (227, 245).
It is possible that TNFa is not involved in PMN accumulation but is needed for
PMN activation, since TNFa can promote neutrophil activation in vitro (246). If
TNFor enhances PMN activation and degranulation, it might explain the slight
Increase in hepatic PMN accumulation in TVXILPS/etanercept-treated mice,
since when TNFa is present the accumulated PMNs might be activated,
degranulate and undergo clearance from the tissue.

In addition to being critical to TVX/LPS upregulation of cytokines, it is
possible that TNFa is involved in the progression of liver injury by enhancing
hemostasis. TVX/LPS coexposure caused fibrin deposition in liver sinusoids, and
treatment with anticoagulant heparin significantly reduced TVXILPS-induced liver
injury (presented in Chapter 6). TNFa has the potential to interact with the
hemostatic system in several ways. It can induce tissue factor which activates
the coagulation system, but it also increases PAI-1 expression which could
depress fibrinolysis (244, 247, 248). Indeed, TVXILPS-induced increases in
active PAl-1 and hepatic fibrin deposition were TNFa-dependent, whereas
coagulation system activation showed a trend but was not significantly reduced

following TNFa inhibition. The results suggest that if tissue factor induction

145

occurs during TVX/LPS coexposure, it is TNFa-independent, but that a slight
reduction in coagulation system activation by etanercept along with a more
pronounced reduction in active PAl-1 was able to prevent hepatic fibrin
deposition.

Based on the importance of TNFa in TVXILPS-induced liver injury, we
examined whether TVX can interact with a dose of TNFot to induce similar
hepatocellular damage. Indeed, TVX treatment prior to a nonhepatotoxic dose of
recombinant murine TNFa resulted in significant liver injury. Other studies have
shown that TNFa by itself does not cause liver injury in mice but can when
administered with galactosamine or a DNA synthesis inhibitor (249, 250). It is
unclear from these results whether TVX sensitized mice to TNFa-induced liver
injury or vice versa. However, the hepatocellular lesions in TVX/TNFor-treated
mice appear similar to those seen after galactosamine/1' NFa coexposure,
suggesting commonalities in mechanisms (251). Recently, cytochrome P450 2E1
(CYP2E1) induction by pyrazole was shown to sensitize mice to TNFa-induced
liver injury (252). It is unlikely that TVX/TNFa-induced liver injury is related to an
effect on CYP2E1 activity by TVX, as TVX treatment alone did not have any
effect on CYP2E1 expression (unpublished results). However, to exclude this
possibility CYP2E1 activity would need to be measured following TVX exposure.
It is also possible that TVX treatment reduced TNFa clearance, and that the
prolonged presence of TNFa resulted in cell death. This is consistent with the
prolonged presence of TNFor in the plasma of TVXILPS-treated mice compared

to LPS treatment alone (216). TNFa inactivation and clearance is mediated by

146

soluble forms of the two receptors (253). It is possible that TVX reduces the
cleavage or expression of these receptors, in turn reducing TNFa clearance.
Further studies are required to better understand the mechanism by which TVX
and TNFa interact to cause hepatocellular damage.

In summary, TVXILPS-induced liver injury depended on the presence of
both TNF receptors, p55 and p75. The p75 receptor may play an even more
important role than the p55 receptor in the progression of TVXILPS-induced
hepatotoxicity. At the onset of liver injury, the TVX/LPS coexposure-related
increase in several cytokines, active PAH and hepatic fibrin was TNFa-
dependent. However, despite the observation that the induction of chemokines
was TNFa-dependent, the hepatic PMN accumulation was independent of TNFa.
The critical role of TNFa in TVXILPS-induced liver injury was likely through
upregulation of cytokines and activation of the hemostatic system. The
observation that the liver injury could be reproduced by substituting TNFa
administration for LPS supports the critical importance of this cytokine in the

pathogenesis.

147

CHAPTER 5

Shaw, P.J., Ditewig, A.C., Waring, J.F., Liguori, M.J., Blomme, E.A., Ganey,
PE. and Roth, R.A. (2008). Coexposure of mice to trovafloxacin and
lipopolysaccharide, a model of idiosyncratic hepatotoxicity, results in a
unique gene expression profile and interferon gamma-dependent liver

injury. Submitted to Toxicological Sciences.

148

5.1 Abstract

The antibiotic trovafloxacin (TVX) has caused severe idiosyncratic
hepatotoxicity in people, whereas levofloxacin (LVX) has not. Mice cotreated with
TVX and lipopolysaccharide (LPS), but not with LVX and LPS, develop severe
hepatocellular necrosis. Mice were treated with TVX and/or LPS, and hepatic
gene expression changes were measured before liver injury using gene array.
Hepatic gene expression profiles from mice treated with TVX/LPS clustered
differently from those treated with LPS or TVX alone. Several of the probesets
expressed differently in TVXILPS-treated mice were involved in interferon
signaling and the JAK/STAT pathway. A timecourse of plasma concentrations of
interferon gamma (IFNy) and interleukin-18 (IL-18), which directly induced IFNy
production, revealed that both cytokines were selectively increased in TVX/LPS-
treated mice. Both IL-18"' and lFNy"‘ mice were significantly protected from
TVXILPS-induced liver injury. In addition, lFNy”' mice had decreased plasma
concentrations of TNFa, lL-18 and lL-1B when compared to wild-type mice. In
conclusion, the altered expression of genes involved in type II interferon signaling
in TVXILPS-treated mice led to the finding that lL-18 and IFNy play a critical role

in TVXILPS-induced liver injury.

149

5.2 Introduction

Despite the rare occurrence of lADRs, they represent a serious hazard to
public health and are an important issue for pharmaceutical companies and drug-
regulatory agencies. Currentpreclinical testing protocols fail to identify drugs that
cause lADRs because predictive animal or in vitro models are lacking. Despite
the prevalence and threat of lADRs, the mechanisms underlying them are still
unknown. As described in Section 1.1.2, we and others have hypothesized that
an episode of inflammatory stress can render an individual susceptible to a
normally nontoxic drug dose, thereby precipitating an lADR (55, 73, 254). In
concordance with this hypothesis, an inflammatory stress renders mice sensitive
to TVX-induced hepatotoxicity (56-59, 216).

TVX is a fluoroquinolone antibiotic which has seen limited use due to its
association with idiosyncratic hepatotoxicity in people. LVX is a fluroroquinolone
antibiotic not associated with hepatotoxicity. In accordance with the lack of
hepatotoxicity seen in people, coexposure to TVX/LPS, but not to LVX/LPS, was
hepatotoxic to mice and rats (58, 216).

Global gene expression analysis of livers at a time of maximal
hepatotoxicity revealed distinct clustering of LPS/TVX-treated rats. We showed
recently that mice cotreated with TVX and LPS also develop hepatocellular
necrosis. Similar to the rat model, LVX did not interact with LPS to cause liver
injury (216). Here we test the hypothesis that at a time before the onset of liver

injury, TVX/LPS treatment of mice leads to a distinct pattern of gene expression.

150

In testing this hypothesis, we identified several genes involved in type II
interferon signaling that were changed by TVX/LPS coexposure. IFNy is a
proinflammatory cytokine that plays a key role in both the innate immune system
and modulation of the adaptive immune system (255). It is a critical mediator of
liver injury from several xenobiotic agents (132, 256, 257). Accordingly, we
hypothesized that lFNy plays a critical role in TVXILPS-induced liver injury and

investigated its influence on the induction of other proinflammatory cytokines.

151

5.3 Materials and Methods

5.3.1 Materials

Please refer to Section 2.3.1 for information on this topic.

5.3.2 Animals

Please refer to Section 2.3.2 for information on this topic. In addition, IL-
18"' mice and C57/Bl6 wild-type controls were purchased from Jackson
Laboratory (Bay Harbor, ME). lFNy”' female Balb/C mice were a kind gift from Dr.
Alison Bauer (MSU, East Lansing, MI), and corresponding female Balb/C control

mice were purchased from Jackson Laboratory (Bay Harbor, ME).

5.3.3 Experimental protocol

In a previous study, mice cotreated with nonhepatotoxic doses of TVX and
LPS developed liver injury (216). The dose of LVX was chosen to approximate
the ratio of TVX/LVX doses used clinically in humans. Mice were fasted for 12 h
before each experiment. TVX (150 mglkg), LVX (375 mglkg) or their saline
vehicle was administered to mice by oral gavage and then given LPS (2.0 X 106
EU/kg, i.p.) 3 h later. Food was returned immediately after LPS administration.
Mice were anesthetized with sodium pentobarbital (50 mglkg, i.p.) at various
times, and sacrificed by exsanguination. The left lateral liver lobe was fixed in
10% neutral buffered formalin and blocked in paraffin within 72 h. The right

medial liver lobe was flash frozen in liquid nitrogen for total RNA isolation.

152

5.3.4 ALT activity and histopathology

Plasma ALT activity was measured spectrophotometrically using Infinity
ALT reagent purchased from Therrno Electron Corp. (Louisville, CO). Forrnalin-
fixed liver Iobes were processed and embedded in paraffin. Paraffin sections

were cut at 5 pm and stained with hematoxylin and eosin.

5.3.5 RNA isolation

Frozen liver samples (50 mg of tissue per sample) were immediately
added to 2 mL of TRlzol reagent (lnvitrogen Life Technologies, Carlsbad,
California). One mL of the tissue homogenate was transferred to a microfuge
tube, and total RNA was extracted with chloroform followed by nucleic acid
precipitation with isopropanol. The pellet was washed with 80% ethanol and
resuspended in molecular biology grade water. Nucleic acid concentration was
determined spectrophotometrically at 260 nm (Smart-Spec, Bio-Rad Laboratories,
Hercules, CA), and RNA integrity was evaluated using an Agilent bioanalyzer

(Agilent Technologies, Model 2100, Foster City, CA).

5.3.6 Gene array analysis

Microarray analysis was performed using the standard protocol provided
by Affymetrix, Inc. (Santa Clara, CA). Briefly, approximately 5 ug of total RNA
was reversed transcribed into cDNA using a Superscript II Double-Strand cDNA
synthesis kit (lnvitrogen Life Technologies, Carlsbad, California). The primer

used for the reverse transcription reaction was a modified T7 primer with 24

153

thymidines at the 5’ end (Affymetrix). The sequence was
5’GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-(dT)24-3’. cDNA
was purified via phenol/chloroform/isoamylalcohol (lnvitrogen Life Technologies,
Carlsbad, California) extraction and ethanol precipitation. Biotin-labeled cRNA
was synthesized according to the manufacturer’s instructions from the cDNA
using the Enzo RNA Transcript Labeling Kit (Affymetrix). The labeled cRNA was
then purified using RNeasy kits (Qiagen, Valencia, CA). cRNA concentration and
integrity were evaluated. Approximately 20 pg of cRNA was then fragmented in
a solution of 40 mM Tris-acetate, pH 8.1, 100 mM potassium acetate, and 30 mM
magnesium acetate at 94°C for 35 minutes. Fragmented, labeled cRNA was
hybridized to an Affymetrix mouse genome array, 430A 2.0 which contains
sequences corresponding to roughly 22,600 transcripts, at 45°C overnight using
an Affymetrix Hybridization Oven 640. The array was subsequently washed and
stained twice with strepavidin-phycoerythrln (Molecular Probes, Eugen, OR)
using a Gene-Chip Fluidics Workstation 400 (Affymetrix). The array was then

scanned using the Affymetrix GeneChip® Scanner 3000.

5.3.7 Hepatic neutrophil accumulation

Please refer to Section 3.3.6 for information on this topic.

5.3.8 Plasma cytokine measurements
The plasma concentrations of IFNy, lL-18, IL-6, lL-10, MIP-1or, KC, MlP-2,

MCP-1, VEGF, TNFor and IL-10 were measured using bead-plex kits purchased

154

from Bio-Rad Laboratories and measured using a Bio-Plex 200 system (Hercules,

CA).

5.3.9 Statistical methods

ALT activity and plasma protein concentration results are presented as
mean 1 S.E.M. A 1-,2-, or 3-way analysis of variance (ANOVA) was used as
appropriate after data normalization. All pairwise comparisons were made using
Dunn’s method. The criterion for significance was p < 0.05.

The results of the microarrays were analyzed using Rosetta Resolver error
models, and ratios were built for each treatment array compared to the VehNeh
control using the Resolver System. This analysis calculated a p-value for every
gene’s fold change relative to VehNeh using the Rosetta Resolver error model
(258). The gene expression change was considered significant if it had a p-value
<0.001. Genes were considered regulated if the p-value was <0.001 for 2 of 3
VehNeh-, 2 of 3 TVXNeh-, 2 of 3 LVXNeh-, 4 of 5 Veh/LPS-, 3 of 4 LVX/LPS-, 5
of 7 TVXILPS-treated mice. Average link heuristic criteria and the Euclidean
distance metric for similarity measure were used to perform the agglomerative
cluster analysis of the treatment arrays using Rosetta Resolver software. Gene
expression profile analysis was done using Ingenuity Pathway Analysis

purchased from Ingenuity Systems (Redwood, CA).

155

5.4 Results

5.4.1 Development of hepatocellular injury after TVXILPS-, but not
LVXILPS-coexposure

The timecourse of hepatocellular injury was examined to determine the
time of the onset of liver injury. Treatment with TVX, LVX, LPS, or LVX/LPS did
not increase plasma ALT activity (Fig. 5.1). TVX/LPS coexposure, however,
caused a significant increase in plasma ALT activity as early as 4.5 h which

continued to increase through 15 h (Fig. 5.1).

5.4.2 Hepatic global gene expression changes before the onset of
hepatotoxicity

Hepatic gene expression was evaluated at 3 h after LPS, a time prior to
the onset of liver injury. Analysis of Affymetrix Genechip 430 2.0 Array data
identified probesets defined as regulated relative to VehNeh-treated mice, and
these were subjected to hierarchical clustering analysis (Fig. 5.2). Within the
cluster of LPS-treated mice, LVXILPS- and Veh/LPS-treated mice were
distinguished from TVXILPS-treated mice (Fig. 5.2). Thus, hierarchical clustering
applied to gene expression analysis was able to distinguish TVXNeh from
LVXNeh-treated mice. In addition, at a time before liver injury, a separate cluster
occurred for TVXILPS-treated mice, the only treatment that produced

hepatotoxicity.

156

Fig.5.1. Development of TVXILPS-induced liver injury. Mice were treated with
TVX (150 mglkg), LVX (375 mglkg) or Veh (saline) orally and then 3 h later with
either LPS (2 x 106 EU/kg; i.p.) or Veh (saline). Mice were sacrificed at various
times after LPS dosing, and plasma ALT activity was measured. n= 45
animals/group. * significantly different from 0 h respective control group. #

significantly different from all other treatment groups at the same time.

157

 

*#
3 6000 —-e— VehNeh
-. —-t-1— TVXNeh
D --e— LVXNeh
V —-e— Veh/LPS
3 —er— TVXILPS
.— 4000‘ .......... LVX/LPS
.2
H
0
(U
I- 2000-
:r‘ *t
*#
Om
0.0 1.5 3.0 4.5 0.0 150

 

Hours after LPS or Veh

158

Fig. 5.2. Hierarchical clustering of hepatic gene expression profiles. Mice
were treated with TVX (150 mglkg), LVX (375 mglkg) or Veh (saline) orally and
then 3 h later with either LPS (2 x 106 EU/kg; i.p.) or Veh (saline). Mice were
sacrificed 3 h after LPS or Veh dosing, and total RNA was isolated from the liver.
Gene expression was evaluated using Affymetrix Genechip 430 2.0 Arrays. RNA
from each mouse was analyzed using a separate array. Gene expression profiles
are analyzed relative to VehNeh-treated mice. Green represents probesets
downregulated, whereas red represents probesets upregulated with respect to

VehNeh-treated mice.

159

 

 

 

 

ps0.001; n= 0345

160

5.4.3 Gene expression changes within TVXILPS, TVXNeh and Veh/LPS
treatment groups

The gene expression changes in the TVXILPS-treated group are important
because they can be anchored to hepatotoxicity. TVX/LPS treatment resulted in
2156 gene expression changes compared to VehNeh-treated mice. The gene
expression changes induced by TVX/LPS treatment were compared to those
induced by TVX or LPS alone revealing a large number of genes changed
selectively in TVXILPS-treated mice (Fig. 5.3). The majority (580/773) of these
TVX/LPS treatment-selective genes were downregulated. In addition, a large
number of genes (983) were commonly regulated by TVX/LPS and Veh/LPS
treatment. TVXILPS and TVXNeh treatments resulted in 619 similarly regulated
genes. As expected, there were relatively few genes regulated by both TVX/Veh

and Veh/LPS treatments which were not also affected by TVXILPS-treatment.

5.4.4 Several gene expression changes in TVXILPS-treated mice are
involved in interferon signaling

Gene expression changes in the TVXILPS-treated mice were of interest
due to the development of hepatotoxicity and were examined further. To do this,
a list of genes expressed differently (p<0.001) in the TVXILPS-treated mice
compared to all other treatment groups (VehNeh, TVXNeh, LVXNeh, Veh/LPS,
and LVX/LPS) was created using TVXILPS-treatment as the baseline for ratio
building. 254 probesets were selectively altered by TVX/LPS treatment; of these

probesets, 142 represented functional genes as determined by Ingenuity

161

Fig. 5.3. Venn diagram depiction of probeset regulation of TVXNeh-,
Veh/LPS- and TVXILPS-treated mice relative to VehNeh-treated controls.
Mice were treated with TVX and/or LPS as described in Methods. Three hours
after LPS administration, total RNA was isolated from the liver, and gene
expression was evaluated using Affymetrix Genechip 430 2.0 Arrays. The
number of probesets increased or decreased relative to VehNeh-treated mice is

shown. Probesets were defined as regulated if p <0.001.

162

1364

i173

VehILPS

 

163

Systems analysis. A full list of these 142 TVXILPS-selective genes can be found
in Supplemental Table 5.3.

The 142 genes selectively changed in expression by TVX/LPS were
analyzed using Ingenuity Pathway Analysis to determine highly affected
pathways. Table 5.1 Is a list of pathways impacted by the genes selectively
altered by TVX/LPS treatment in order of increasing p—values. Pathways with p-
values > 0.1 were excluded from the list. In addition to the p-values, the fraction
(ratio) of genes affected within each pathway is reported.

Of the pathways affected, two which had the highest ratios, low p-values
and are involved in several models of liver injury were JAK/Stat signaling and
Interferon signaling (Table 5.1). Type I and II interferons signal via specific
receptors through JAK/Stat signaling within cells (259). To examine further the
potential role of interferons in TVXILPS-induced gene expression changes,
genes in the set of 142 TVXILPS-selective genes were examined for their
relationship to interferons. Of the 142 genes, 26 (18.3%) are regulated by
interferons (Table 5.2). Of these IFN-regulated genes, all but 3 were increased in
expression. A number of the genes have potential importance in the development
of hepatotoxicity by being pro-apoptotic, chemokines, pro-inflammatory or
involved in immune responses (Table 5.2). The majority of these genes were
regulated by IFNy. Accordingly, we examined the role of lFNy in TVX/LPS-

induced liver injury.

164

Table 5.1. Pathways highly affected by the 142 functional genes selectively
changed by TVXILPS-treatment compared to all other treatment groups.
Mice were treated as described in methods. Three hours after LPS administration,
RNA was isolated from the liver, and gene expression was evaluated using
Affymetrix Genechip 430 2.0 Array. Probesets were defined as regulated if
p<0.001. The 254 gene-associated probesets changed by TVXILPS-coexposure
when compared to all other treatment groups were analyzed by Ingenuity
Pathway Analysis. Of these 254 probesets, 142 were identified as functional
genes. The pathways highly affected (p <0.1) are listed along with the ratio of the
number of genes changed selectively by TVX/LPS to the total number of genes
in the pathway as identified by the Ingenuity database. The specific genes
selectively changed by TVX/LPS that are in each pathway are listed. Some of
these genes are involved in more than one pathway and are, therefore, listed

more than once.

165

 

5000 .55 5005 .255 .5592 2.50:2 50.5 8.585 0525.065 cacao:
8.22.2 .5225. 505.5 205.5 2.50:2 59o 8055.5 55.295 08382
0.2.2.2 .225 .500 2.me2 $55 8.5555 55.205 5-...
ER .2555 .5005 .050 58.5 8.525 55.205 2552ch
cozmzwom
meow .5 5.2.2 .205 .2220. .2.._0m_> 58.0 8-52.5 :8 20.95 0:80:55ch cease...
50... .5595. 2.50:2 58.0 8-52.5 . 55.205 .28ch 93.2
5x22: .555 5.22.2 .2522 .5595. 2.50:2 ~30 3.58.. 55.855 50.0.2
252“. .59 .235 .5202 53.5 8.58.? 8558 5x552
mmw<zo 055050.
20.2: 582.2 .000 .805 .50“. .522. 53o 8.5%.» $95 9588 8585-252
«5005 .5250 .225 .5005 58.0 8.is 55.295 5522..
0.225. 505.5 .e0m_m 2.50:2 5505 8.555 55.205 .28ch £80
Z52“. .55 .2550 2.205. .5512 58.0 8-58.5 9:205 man.
252“. .55 5005 .255 0.50 Jess: «mod 8-552 853% mxmaxu
E05 5.82.2 .225 .50”. 5.50.0 8.55.5. 55.855 “.05
255 .555: .0292 .2“_05> 2.50:2 Sod 8.553 5.859228 as s 55.20.... 06%:
2x02: .225 .50“. .55005 2.50:2 Ego 3-52.5 55.29... o;
_ .555: 5.20.2 .2
.5: .5225d 020052 .5005 .255 .e0m_m .8052 Sod 8-523 855555: 59ch
52.2.5 20.2: .5005 .5322 .225
22.2.55 .805 .50“. .55005 .5822 2.50:2 59o 8-55.5 550:0... 8:8.ch cases $82
52.2.55
.55 .222 .305 .50... 2.50:2 5.322
220.00 ..mmm: .225 2.550 .220. .5252 53.5 8-52.5 55.205 5580.. 28588.0
50:00 030m 03.2.0 5355a.

 

...m 030...

166

 

 

w¥mm<2 .m._.<._.m .mfiwom .<.mv.u_z .Nmmo<
mom—m .vOmzm .<n_0m_>
Zw._.n_ .mumum. .<_.Zv_n_0

<FZ¥DO .9205.
$.02 ...szmm—w 40,0...
<_‘Zv_oo .<n.mm0 .Emum:
MES—.2 .xom ...—2.40. .xmo .EZDJO
m._.<._.m do”. .xmo
_.._.<ZQZ® .0400
w¥m¢<§ .Zm._.n_ <sz00 NED—2 .<_mvEZ
<.<._I .Ewm
«Zn—......— Jim.
2w<n_ .<_.Zv_oo .NEGH .NEQS. .mO...

mNod
mmod
mmod

nvod
omod
wmod
nwod
Sued
owed
wNod
..mod
mood
mmod

mom—55¢
NOflth
Noam—mod

No-m_mm.m
No-m_Nm.w
mom—mad
Nonnumok
NOMK _‘ .m
mom—no.0
Nommmh
mom—Nod
Nonmhnd
No-wmm.m

959.95 .20000. 003000 50.0.0-0
95055 5_50.0_05 5.03 0_;00..0>E<
55.2.0.5 22.”.

00.00300. 02.09.0050
085% 220 2.8 0.26 ..00
52.9.5.5 ..-.02
5.52000 mmeo>
05.2.95 5.550558 0300030..
55.9.5.5 .00..
65.600002: 0.050020
05.2.0.5 0.20.5.0
2.0.5.5505 c09E<
05.2.90 .3523“...
05.2.95 .2080. 009000.02. _>.<

20.28. 9.5 032

167

Table 5.2. Genes selectively altered in expression by TVXILPS that are
regulated by interferons Mice were treated as described in methods. Three
hours after LPS administration, RNA was isolated from the liver, and gene
expression was evaluated using Affymetrix Genechip 430 2.0 Array. Probesets
were defined as regulated if p<0.001. Of the 142 genes changed selectively by
TVXILPS-coexposure, the 26 listed have been reported to be regulated by

interferons as determined by Ingenuity Pathway analysis.

168

 

5.500.025 .50... 0. 00290. 0.5 4 0.00.00 02.100.00.00 00.52020 .(0010 00000.2.
0500050. . 0000 0500050. I
00.00.8000. 02.0025 0.... 0 m 4 000.00 000200.050 0.2002 000.52 .0 N.N0.2.
020.00.002-20 . w 0500... 0500... I
.553500 22.. .0 0.02.5. 50.0>..00 2 mm 4 05000. 0.0.0.0 00.02.00-509... 9.00.2.2 .0 080.2.
. 0000 I I
h .. 4 050200090 00.0.:0...5 00.0.0.0. 0N0w_ .0 0 0000.2.
2.2000000 20. 4 . .200. 00.0500. 00.0.0.0. Em. .0I0lom2022.
090550.222. 0... 4 2 .2080. 2 0.0.8.0.... 20:... .mmmoaomv.
0.0 4 05000.0 0002.0. 0Ew000 00.2.0.0. 00.0. .0 2.2.2.
. 2000 I
02.02.00 000 0.0.0.500. 00.09.00. 0 .2 4 . 2:00.00. 00.50.00 22.00.00. ...—<0. .0 .002N2.
0500050. . < ._ 5520 - I
00:00.. .00..0.0050.0 000000 . 01.2 2 N. 4 0.0.0.000 0:5..0008005... .00.). < (..I ..0 2NmmN2.
0500050. . . - I I
0.555. 02.02050... $0.00 _ 0:2 2. 4 8.00. 0 N ......n_.358o.m.: 0.. u: .0 x .852.
- . N 0200.0... 00000000 I
0.6.080 o... 5 2 4 50.”. was 5.8.8.0250» 5.0., ~55 .0 5505.2.
8.00.9.0 0.09.39. .00.0.2 8. 4 5 0:05.. 0.60. 0-x-0. 00059.0 30x0 02.83.
00.20.0000 2.00.500 .0N-n__s_ 2.00 4 N 0000.. 2.00. 0-x-0. 00.200000 N._oxo .0 00N0.2.
0:505 :00 0255.0 ..00 0.0 D 50.00.00 .0000 «.010 .0IxI0Nm2m2.
.. . . . . 0.... 9.0 085.6. 5.; 2.8.09.0
. N .m 550.0 00.20.08 I
. 0 4 0.00.00 x..00-002-x..00 0.50m NmI._Im .0 080.2.
0.2.80-0... 5.2 4 0.38.00.25.00 2mm 0:. .202.
.0000. 0. 0000.308. 3.00.00 000000 02030.0”. 0.00.. 0000 30:50 .0000
0.0". 0000 00000000

 

«.0 0.00...

169

 

0000.30.
0.000230. 00.050.000.00

02.03080 00.20E000
.0 00.2008 00.000.000.00

55005 0>..00.x0 0. 0500050.

00..000.5 2.5.2....
0. 0500050. 2000000. 02.0002

00..000.5 2.0.2....
0. 0500050. 2000000. 02.0002

000500.. 00. .0.00. .9000

00.9.0... 2.5.22.2
0.0500050. 2000000. 02.0000

5000000 . 015:0 00.550020

0.0
0..
0.2
0..
..om
0.0.
0..
0..
0.0

N.

44444444>>

<
.200. 0.2.0.0 .0..0...0000 .0.3050>
N 020.00..
0 2.... ..0. .0 05500035 0 ....20

0 0000005000

.0 0.02.00 000 0030500.. .0020
.00000000..E

.N 050.3E5.0 00.08.0030

0 00..000.5 00.2008 .0 35500030

N 00..000.5 00.2030 .0 55.50.0030

20000500..
00. .0.00. .02020 00.038-00.20
0500.0.0.0..00 .0024.
N 00...
8.0000.
-000 .050.0..050..0 00.50.... 0.0.0.0
0 .000 000.0 ...03035 0E0500.0.0
.<

<00w>
NOSSw
0._.<._.w
NDOw
MwOOw
NwOOw

N<. .O._m
0<Z_0mww

NZnEh.

m<§w0

.01000002.
06:50.2.
06:38.02.
0:00.822.
02:88.0:

0:00.022.

.0I.220~2.
0:00.22.

.mlwm .mmv.

002000202.

.0008. 2.... 030.

170

5.4.5 Timecourse of plasma concentrations of IL-18 and lFNy

IL-18 induces the production of lFNy in several cell types. Neither TVX nor
LVX increased plasma concentrations of IL-18 or IFNy when given alone. LPS
treatment caused a significant increase in both lL-18 and lFNy plasma
concentrations at all times measured (Fig. 5.4A, B). TVX treatment prior to LPS
caused a further increase in plasma concentrations of lL-18 and lFNy at both 4.5
and 6 h (Fig. 5.4A, B). In contrast, LVX did not affect either plasma IL-18 or IFNy

induction by LPS (Fig. 5.4A, B).

5.4.6 TVXILPS-induced hepatocellular injury in lL-18"' mice

To explore the role of lL-18 in TVXILPS-induced liver injury in mice, wild-
type and lL-18"' mice were treated with TVX/LPS. lL-18"‘ mice had significantly
reduced plasma ALT activity compared to the wild-type controls (Fig. 5.5A). In
corroboration with the ALT values, TVX/LPS treatment resulted in midzonal
hepatocellular necrosis in wild-type control mice (Fig. 5.58), which was less

extensive in lL—18"' mice (Fig 5.50).

5.4.7 TVXILPS-induced liver injury in IFNy 4' mice

To examine the role of IFNy in TVXILPS-induced liver injury, wild-type and
IFNy"' female Balb/C mice were treated with TVX/LPS. TVX or LPS treatment
alone did not cause hepatocellular injury in female Balb/C mice (data not shown).
TVXILPS-treatment resulted in significant liver injury in the female Balb/C mice,

as seen in male CS7/BI6 mice. In lFNy"' mice, the TVXILPS-induced increase in

171

Fig. 5.4. Timecourse of IL-18 and IFNy plasma concentrations. Mice were
treated with TVX (150 mglkg), LVX (375 mglkg) or Veh (saline) orally and then 3
h later with either LPS (2 x 106 EU/kg; i.p.) or Veh (saline). Mice were killed at
various times, and plasma concentrations of lL-18 (A) or IFNy (B) were
measured. n = 4-6 animals/group. * significantly different from the respective
treatment group at 0 h. * significantly different from VehILPS-treated mice at the

same time.

172

lL-18 (pglmL)

IF Ny (pglmL)

— <9 — VehNeh
- a?» — TVXNeh
- -0- LVXNeh

4000 . - -O - VehILPS

 

 

 

 

—O— TVXILPS
.......... vaijs *#
3000 « *
*
/ ‘~
/ ......... s. *#
2000 - / °°°°°°°° :‘Kx *
/ \\£
I
I \\
1000 . / "mi \\ a:
#
0 0 o o
0 1 2 3 4 5 6
Hours after LPS or Veh
300 -
—-o— VehNeh *#
$01 _‘§— TVXNeh
* #
200 e
150-
100 -
50 .
o .............
0 1 2 3 4 5 6
Hours after LPS or Veh

173

Fig. 5.5. TVXILPS-induced liver injury is attenuated in lL-18"' mice. Male,
C57Bl/6J wild-type or lL-18 "' mice were treated with TVX/LPS as described in
Methods. A: Mice were sacrificed 15 h after LPS treatment, and plasma ALT
activity was measured. Wild-type controls and lL-18”' mice had equivalent
baseline values of 55 i 15 U/L. n = 6-9 animals/group. * significantly different
from wild-type control. 8: Representative photomicrograph of liver from a wild-
type mouse sacrificed 15 h after TVXILPS-treatment. Midzonal lesions of
coagulative necrosis were observed. Arrows highlight necrotic lesions. C:
Representative photomicrograph of liver from an IL-18"' mouse sacrificed 15 h
after TVXILPS-treatment. Few lesions of necrosis were observed. Arrows

highlight necrotic lesions.

174

 

 

A 35. .5<

lL-18-I-

wild-typo

 

175

plasma ALT activity was markedly attenuated compared to wild-type controls (Fig.
5.6A). Midzonal coagulative necrotic lesions obvious in wild-type mice were not

seen in lFNy"' mice (Fig. 5.68 and 5.60).

5.4.8 TVXILPS-induced hepatic neutrophil accumulation and pro-
inflammatory cytokines in lFNy"' mice

Hepatic PMN accumulation and plasma concentrations of several
cytokines and chemokines were measured at a time before the onset of liver
injury in wild-type controls and in IFNy”‘ mice. Plasma ALT activity values were
equivalent to baseline for both wild-type and lFNy"’ mice (data not shown).
Hepatic PMN accumulation was also similar in IFNy”' mice compared to wild-type
mice 6 h after TVX/LPS treatment (Fig. 5.7). The plasma concentrations of lL-6,
MCP-1, lL-10, KC, MlP-2, MlP-1a, and VEGF were similar in wild-type and IFNy4'
mice given TVX/LPS (data not shown). In contrast, TVXILPS-treated lFNy"' mice
had significantly reduced plasma concentrations of TNFa, lL-18, and lL-1B

compared to TVXILPS-treated wild-type mice (Fig. 5.8A, B, C).

176

Fig. 5.6. lFNy"' mice are resistant to TVXILPS-induced liver injury. Female,
BaIb/CJ wild-type or lFNy"' mice were treated as decribed in Methods. A: Mice
were sacrificed 15 h after LPS treatment, and plasma ALT activity was measured.
VWld—type controls and IF Ny"' mice had equivalent baseline values of 55 j; 15 U/L.
n = 6-10 animals/group. * significantly different from wild-type control. B:
Representative photomicrograph of liver from a wild-type mouse sacrificed 15 h
after TVXILPS-treatment. Midzonal lesions of coagulative necrosis were
observed. Arrows highlight necrotic lesions. C: Representative photomicrograph
of liver from an IFNy"' mouse sacrificed 15 h after TVXILPS-treatment. Few

lesions of necrosis were observed.

177

 

 

Ime m m
33...?
A

IFN-I-

wild-typo

 

178

Fig. 5.7. TVXILPS-induced hepatic PMN accumulation is unchanged in
IFNy"' mice. Female Balb/C wild-type and IFNy "' mice were treated with 'NX
and LPS as described in Methods. Mice were sacrificed 6 h after LPS
administration, and hepatic PMN accumulation was quantified as described in
methods. Wild-type controls and IF Ny"' mice had equivalent baseline values of 1

i 0.5 PMNs per HPF. n = 5 animals/group.

179

 

 

 

Wild-type

 

 

 

wwwmm

"E: ..on mz_>_n_

IFN-I-

180

Fig. 5.8. Plasma concentrations of proinflammatory cytokines are reduced
in lFNy"' mice. Female, Balb/C wild-type and lFNy"' mice were treated with TVX
and LPS as described in Methods. Mice were sacrificed 6 h after LPS
administration, and plasma concentrations of TNFa (A), lL-18 (B) and lL-1B (C)
were measured by Bio-Rad bead-plex. Wild-type controls and lFNy"' mice had
equivalent baseline concentrations of TNFa (210 1 41 pg/mL), lL-18 (25 1 10
ngmL) and lL-1B (75 1 16 pg/mL). n = 5 animals/group. * significantly different

from wild-type controls.

181

A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wild-ty pe

IFN-I-

 

 

... ...
* N * N *
.ir .ir
w w
.III. ...-M .II. M .l].
w w
mmmommmmmommmmmm.
3 .52: cuzp 3 see c w... 3 .553 n 3.

B C

182

5.5 Discussion

This study confirmed previous results that TVX-pretreatment, but not LVX-
pretreatment, interacted with an inflammatory stress induced by LPS to cause
liver injury in mice. We extended this to define the early timecourse of
hepatocellular injury as measured by plasma ALT activity. TVXILPS-treated mice
had a small but significant increase in plasma ALT activity as early as 4.5 h after
LPS, whereas there was not a significant increase in plasma ALT activity in any
other treatment group at any time measured (Fig. 5.1). Based on these results,
hepatic gene expression changes were examined at 3 h, a time just before the
onset of liver injury.

Similar to results obtained in LPS/TVX-treated rats after the onset of liver
injury (58), we observed that TVXILPS-coexposure resulted in unique gene
expression changes in mice prior to the onset of liver injury (Fig. 5.2). The distinct
gene expression profiles were observed despite differences in species, timing,
and routes of LPS and TVX administrations in the two studies (58). These results
in mice suggest that global gene expression change is an earlier marker of liver
toxicity than plasma ALT activity in this model.

Although, numerous groups of genes might be involved in injury, the set of
genes selectively affected by TVX/LPS was chosen as a starting point for
mechanistic analysis. There were several pathways affected by TVX/LPS-
treatment that suggest that TVX magnifies the inflammatory response induced by
LPS treatment (Table 5.1). This finding is consistent with the finding that TVX

pretreatment enhances the plasma concentrations of several pro-inflammatory

183

cytokines induced by LPS (See Fig. 3.2). Of the pathways impacted by TVX/LPS-
treatment, JAK/STAT and interferon signaling pathways had among the highest
ratios and low p—values.

The actions of lFNy are exerted through the activation of the JAK/STAT
pathway (259). Of the 142 TVXILPS-selectively expressed genes, 26, or 18%,
can be regulated by interferons (Table 5.2). Furthermore, a significant number of
these genes have roles in functions related to hepatotoxicity, such as apoptosis,
leukocyte migration, pro-inflammatory cytokine production and immune response.
Most (23/26) of these IFN-regulated genes were enhanced in expression in
TVXILPS-treated mice, suggesting an increase in interferon signaling. The
majority of these are known to be regulated by lFNy, rather than by lFNoc or IFNB.
lRF-1, one of the genes selectively upregulated by TVX/LPS, encodes for a
transcription factor activated as a result of JAK/STAT activation by IFNy (259).
This pathway is activated in several other models of liver injury, including
concavalin A, acetaminophen, ischemialreperfusion and LPS/galactosamine (115,
116, 131, 132). However, the role of IFNy or IL-18 has not been examined in a
model of liver injury from drug/inflammation interaction; therefore, the role of IL-
18 and IFNy was explored in TVXILPS-induced injury.

The first step was to measure the plasma concentrations of lL-18 and IF Ny.
IL-18 stimulates IF Ny production (260). The plasma concentrations of both IL-18
and IF Ny were selectively increased by TVX/LPS treatment (Fig. 5.4A, B). This
enhancement and prolongation of the lFNy induction by LPS has not been shown

previously in any drug/LPS-coexposure model of liver injury. The mechanism by

184

which this interaction occurs is unknown. However, we reported previously that
TVX pretreatment prolongs the LPS-induced plasma TNFa peak (216). It is thus
possible that this prolongation of the TNFa peak drives the prolongation of the
plasma IFNy peak.

Based on the selective increases in lL-18 and lFNy in TVXILPS-treated
mice, we examined the role each cytokine plays in the development of TVXILPS-
induced hepatotoxicity using transgenic mice. IL-18 and lFNy are components of
the same signaling pathway. Both lL-18"' mice and IFNy”' mice had significantly
reduced hepatocellular injury following TVXILPS-treatment compared to their
respective wild-type controls (Fig. 5.5 and 5.6), implicating pathophysiological
roles for lL-18 and IF Ny. Female Balb/C mice were used to generate the data for
Fig. 56-8. The reason for the gender and strain change related to the availability
of lFNy"' mice of this background. Despite the wild-type controls having lower
plasma ALT activity in response to TVXILPS-treatment, the extent of
histopathologically evident hepatocellular injury was similar to that seen in male
C57Bll6 mice. It is unknown whether the difference seen in the magnitude of ALT
activity increase in response to TVXILPS is due to a strain or gender difference.
Regardless, the pathway by which lL-18 increases lFNy is common to both male
CS7BI/6 and female Balb/C mice. Thus, the observation that both male 057BI/6
lL-18"' mice and female Balb/C |FNy"' mice were protected from TVX/LPS-
induced liver injury suggests that the importance of this pathway in TVXILPS-

induced liver injury is not gender— or strain-specific.

185

Several inflammatory cytokines and chemokines were measured in wild-
type and |FNy"' mice treated with TVX/LPS at a time before the onset of liver
injury. This time was selected to exclude the possibility that a difference seen
was a result of injury. The plasma concentrations of several cytokines were
unchanged, but TNFa, lL-1B, and IL-18 were significantly reduced in lFNy”' mice.
Previously, TNFa inhibition attenuated TVXILPS-induced liver injury (216). IFNy
enhances the production and release of TNFa, as shown in vitro (121). In vivo,
lFNy induction of TNFa production is mediated by lRF-1 (261), a transcription
factor selectively upregulated in TVXILPS-treated mice (Table 5.2). Interestingly,
TNFa inhibition reduced the TVXILPS-induced increase in IFNy (see Fig. 4.2),
suggesting that each of these cytokines regulates the expression of the other.
Additionally, lFNy"' mice had reduced levels of lL-1B and lL-18. These cytokines
exist as pro-forms of the proteins in cells and are cleaved by caspase 1 to their
active forms. The mRNA levels of IL-16 and lL-18 were unchanged in gene
expression analysis. Therefore, we hypothesize that IFNy positively feeds back to
increase caspase 1 activity either directly or indirectly. The consequent increase
in lL-18 could then cause more production of lFNy, potentially resulting in a
proinflammatory vicious cycle.

lFNy causes apoptosis in primary hepatocytes (262), and it is possible that
TVX directly sensitizes hepatocytes to IFNy-induced cell death. Additionally, TVX
enhanced the LPS-induced increases in both TNFa (216) and lFNy, which

synergize to cause primary hepatotocyte cell death (263). It is also possible that

186

Fig. 5.9. Hypothesized role of IFNy in TVXILPS-induced liver injury. Dashed
arrows represent untested hypotheses. LPS causes hepatic neutrophil
accumulation and TNFa release. TVX enhances the LPS-induced increase in
TNFa (216). TNFa can increase lL-18, which in turn induces lFNy production. In
addition, lFNy can feedback to increase both lL-18 and TNFa (Fig. 8). This has
the potential to create a vicious, proinflammatory cycle of lFNy and TNFa
production. In addition, PMNs migrate into tissues and become activated,
causing death of hepatic parenchymal cells. It is possible that lFNy enhances
PMN activation. In addition, TVX might act directly on hepatocytes to sensitize
them to cell death induced directly by lFNy, lFNy/TNFa coexposure or toxic

factors released from PMNs.

187

TVX acts directly on hepatocytes to enhance TNFor/lFNy-induced cell death. Fig.
9 illustrates the proposed pathway to TVXILPS-induced liver injury and the
interaction between TNFa and IFNy.

In addition to effects on other proinflammatory cytokines, IFNy can play a
direct role in the regulation and activation of many cell types that might be
involved in liver injury. For example, PMN activation is critical for TVX/LPS-
dependent liver injury (see Fig. 3.6) and IFNy can increase PMN activation in
vitro as measured by oxidative burst and differential gene expression (118, 264,
265). Although PMN accumulation after TVXILPS-cotreatment was unchanged in
IFNy”' mice, it is possible that activation of accumulated PMNs was decreased in
these mice, and this could be a reason for the reduced hepatotoxicity in response
to TVXILPS (as depicted in Fig. 5.9). In addition, IFNy plays a role in the
activation of CD8+ T cells and macrophages (104, 266). Accordingly, it is
possible that IFNy is involved in hepatotoxicity by activating one or more of these
cell types.

In summary, TVX interacts with LPS, but not LVX, to cause liver injury in
mice. That hepatotoxicity only occurred for the drug with lADR-potential in
humans raises the possibility that inflammatory stress might play a role in the
pathogenesis of idiosyncratic liver injury caused by TVX in people. Gene
expression analysis revealed distinct clustering by treatment at a time before the
onset of liver injury. A small group of genes was identified which changed in
expression only after TVXILPS-treatment compared to all other treatment groups.

Of these genes, a large number were related to interferon signaling. This

189

observation led to experiments that demonstrated the involvement of lL-18 and

IF Ny in the pathogenesis of TVXILPS-induced liver injury in mice.

190

Supplemental Table 5.3. Functional genes selectively changed by TVXILPS-
treatment compared to all other treatment groups. Mice were treated as
described in methods. Three hours after LPS administration, RNA was isolated
from the liver, and gene expression was evaluated using Affymetrix Genechip
430 2.0 Array. Probesets were defined as regulated if p<0.001. The 254 gene-
associated probesets changed by TVXILPS-coexposure when compared to all
other treatment groups were analyzed by Ingenuity Pathway Analysis. Of these
254 probesets, 142 were identified as functional genes. The specific genes

selectively changed by TVX/LPS are listed.

191

mEmEou coma: ac:

 

£55; :8 .8536 :8 mm. > use 5855.. 55, 269.555 510 scrmrmmmvmi
mm b 2% cmma 5me 5 ~82:
ommEmc S .9525
<20 2 $839 3258995 m. 3 4 $2: 252855.565 32x8 5:83va
cozmzwom
=8 c. 265% .8582 955:: we b 2822: £8 38 6:883?
951 cemmoao.
o.» 4 2:898 8&8 «moo eezmoo 6:858va
we 4 o 8.59:8 555% __8-uc__co ooooo 5U 58:
3 b 925:8 22.8 55 598 5 82m:
3 > F 5255 85583 296 =8 FzEnZo eminence:
v 9:59:00
5 b -582 as 6.323% coma. Humans:
m 9.39:8
3 4 -589 as 3.3233 8mm 56881
N .m mmmfi .mEEmEoo
to 4 5558 5.2.5825: 25cm 851m swimwear:
2558953 92 4 5.3259229 .28 51:2va
omcoamo. 0.52:5 BoEB macaw
52625ch 59.5 _ 0:5. mm b 255 2 8555 5:55:52 5.2%. “mime EN:
3 4 mm 52.8 525995 5: E m8E< 5:258:
mm b >.< 59989.85. 20% afloat:
Z 4 38.8: xmc< can? a 35::
9x82. 2 528.8. 3” 4 .5 557555852 vfieoz< “micrometer
oomommo xn_<.>_ m9m>=om mdm 4 6382 2995.? Na «mm—9‘ “argue”:
lounges". .025» £30
.235 E soc—wrong 33:39". Eon. 532.63. 2:2. 2.00 980 cocoa—cow

 

Qm aim... 39.028396

192

m._.<._.w

w9m>=om .:o=m>=om xmwimSZ
oomommo

¥a<_2 8 xomnuooc o>=mmoc
8380

x532 so 8883 5268:

£288. :8 8 :28.qu magma

8:20:88 .2858: .nNdzz

820.88 .2888: .5N-n:.2

c8888.: oEEEocoofio

mé

o.m

QN
m.N_.
m.N_.

06
ms
Wmm
N.m
N.N
m.m
cor
wdw
o.N
fin
o.N
9v
0N

44>>><<>>>> >44‘4><

o 88$ 888cm: 2889.3. o..=w
< gcznam .m 889.

582 8:955: 25:93 <95
Eon N :csnam N 883
8825 88.25: 25283 $35
8889
Loss 53ch 55528 meow
m $92888 3850on .25 woman
S 322585 858% .98 2.88
mm 8652633128 $2.28
N 8980.:
283 52.8 3..-:8 Emma
2 $8885 x8. 98 2x8
:R 83 559528 258
9 0:8:
@5ch .88 x oEomoEoEo m _. ”.moxo
m
.88: £55 0-x-ov 558558 2on
N
959. 955 0-x-ov 55.5258 «.88

88:8; 08.808
3.5 we; «88me x8-> xmo

owmaoemc«53506-0-.05880 H.200
F 88895 528.0 E.z_._o
3 Enzo-0 $2040

988 mmmsx 950:0 (x10

66:339.
science:
scrmmoevi
5:828:
5:153:

scmcueovmeee
cc 5 sconce,

smhwoemmee
.5 m mmsmmee
en ooemmv.

51583.3
5:853:
5:88,:
scmmvmmvi
55:85:?
on «memmes

.mrmvmomi
Cc c.eo~omvs

8.2.8. a.» cane 5:55.285

193

0000.0 02.0.0.8 03.000900.
00.8000. 088:...
52.0.8008 009.00 . 01.2
008000. 00:08..
50.0.8008 009.00 . 01.2

832.00
mumz .0 02.0.3000 0>...000

80.020.008.08 0.8800008

0.8800008

de
V.N_.
V...
50
N0
mé

©.N_.
0.0N

NV
...N

0.0
m0
fin. ....
MN
...m
mN

QN

44>>4>>>>>4>>><44

F 000000.08 080...
< ._ 000.0 90.0800
9.300058%... 5.02

8.00. o .N 28.008020...
0 .0 2.0.0.030
0.90... 509800.00 .900
N .0 2:00.030
0.900 .c09.c00.:0 .900
F 000.080.053.000
.2 0.0500029008000030

.. 0.8.0 8.0.08 8.8.8.. 000
P 000200020
.2300 0.50.00
.0009. 050.080.080.20
8.0000. 80.0020
003:0 0.50.00
00088008000820
0 00.0.8: 00.3.0“.
@0808... 020808 .0..>
088800000 00.85 am“. 00..>
0. 0.0.08 .80.”.
00059.0 0.00 3.0“.
N 00.008; 0000008
on 0E.> 0.00.00.85.80 0.0->
P 8.3.2:.
0.000000. 8.0000. mmmm.

SAC—2....

<.<._I

0..-NI
$00400
N<0..OO
_.._.<Zn_ZO

..<...0
..000

0400
@000

0.30
NON“.

won.
oOxmm
20.4“.
Nwhm

FEW—mm

.0r000000.
.0r000000.
.0rxr.00.00.
.0i010000.0.
.0i000000.
.0r00.0~0.

.0nm1000000.
.0 000000.

.0HOH000000.
.0 0 00.000.

.0H0000F0.
.0 000000.

.0Hoo.0~0.
.010000.0.
.0 000000.
.0i0000F0.

.0I0~.0.0.

.0008. 0.0 0.0.00 300520000

194

>8.0EE0..5-08
0.03.0000 0.2.. .0 0.052 00.02.00

>0>>:.00
00.0.2 .0 :0..0.:00. 0>...000

02“.. 0:0 02.... 00.0.0000

0..0.0000-0.0
8905800508

:0..0>..00
0:0 808.880. 0.89.00.

0.0
0.0
0.00
0.0
0.0.
N...
0.0
oNN
m. ....
n. 2.
N00
0.0
0.0
méu

0N
0.0

0.0.0.

0...

>4><<444444>>4>44>

00 000000 000.000 5.0.00.2 0000..

50.08 05050 000 .0802 0.205.
0 0005.. 000:0. 000:...
0.0.90 00.02.00-609... 00.0032
0 0005.. 000:0.
0.0.90 00.02.00-509... 00.0032
5.00. 02. 08:05-00. .5:

8.0000. 508.000.. >..0:00 2.0.. «in...
8.0000 05.0:90 000005-025 EN ..<<_v.
080008088 0 :2. mznw
50.08 .
05050 <52 02$ 00:02.5 0m<_.mz>.
0:00 0000.0:c0x0

00.0.0.5.0 888.5 000.0.
.. 8.00. 0.8.0.000. 8.0.00.5 Em.
4.8.0000. n. 5.00.8.5 <05...
000.00
000005 05500 8.0.8.5 n.._.0_
. 50.08

05050 8.00. £2,000 00...-5305 .0000.
. 8.0.000. .0.:0E00_0>00

00.08.090.008. Em...
0 00.00... <20 .0 5.5.00. 00_
.0000. .
0500.08 8.00200 8.2.00.0.5 .520.
n. 508.000.0380..
80.00: 000080088... .0002...

.0H0H0000.0.
.0 0 000000.

.0:0000.0.
.0:0|000.00.
.0H0000.0.
.m_000.00.
.0 m 00.000.
.0 0000.0.
.0|0|.00000.
.0H0H0000.0.
.0m800.
.0 000000.
.0|.0.0.0.
.010.00.0.

.mlm000.0.
.0 0 000000.

.0Ihmovwv _.

000000000.

.0008. 0.0 0.00. 300520000

195

0:00 .0..0:0..00..E .0. 0.508000.

0.00550 .00”. 5 00295
:0..0>..00 9.02
00000.05 .:0..000.000 0.240

82,500
00.0.2 .0 :0..0.000. 02.000:

00:0000.
>8.0:.E0..5 05.080 «...-0
00:0000. 000..0 0>..00.x0
:0..0.0.E 00.09.00.
000000.000 x...0:. .0.0..000..x0

N0
NNN
0.0
...m
0.0
FNV
0.0.
0.0
odh
0.0

0.00

0.0..
0.0
0.0

0.0
Nov
1N.-

0.0

4 <>>>4 4 4 b 4}

>4 44 444

000.0E>.00 .0..0:0..00...2
m 000:0. 0.5.0.00
000.080.0000
058030008020
0 . 00052-0 0.0.8098
-100005300000800
to.
50:80 02.70.00.980. 500.00...
050000
-0520 .00 000.000.0208...
. 00000. 0.0.0.0 00.500

... .00... 00.0.0002
00.0 8.0000. 80802
000.000.8008 00.000.002
0.00 .8..0.:5 0:00-0
5 80880 0:00 00.800200
E9. 0000.. .0 8.00. 80.00:
0.80 00.020. 0.00-0
5 808.80 0:00 00.800200
E9. 0000.. .0 8.00. .0080:
00000800 ..00 80.0008 .0802
00000.90 ..00 80.0008 .982
0:00 008000.
80.58 :0..0..:0.0...0 0.0.022

m: 5000.52.90...

0.. 0000..0000_.0.0:. x...0.2
0....-5088 05808.5 03.).

him-.0...
00.00

00.0
<..v.mn:n.
..<D..In.
m¢m_n.n.
5....0n.
...—.032

mummz
n.Z

N.mv_...z

(.mvEZ
0DDwZ
dun—DMZ

000:).
m ......2

0.6.5.5.
..n..xu..2

film-.0000.
.0 000000.

.0IX|000000.
.0u0u00.0.0.
.0:0000.0.
.mwm00000.
.0u01000000.
.0 0 000.00.

.mnm00000.
.0 0 000000.

.0n0u000000.

.0H000000.
.m 00.000.
.0 0 000000.

00000.0.
.0 0.0.00.

0.0.0020.
.0 0 00.0.0.

......08. 0.0 0.00. 300520000

196

:00000>0..x0
0:0 800.80 00.09.00.

>03:.00
00.02 .0 800.000. 02.000

05000.0 2.020...
0. 00:0000. 0.000000. 02.000:

000800008
0000000 . 01.2 .0 05000008
000E00

<20 0. 00:0000. .000.0000-0.0
80008.8

0...
YN

0.0..
0..

0.0
0.0..

0.0
N0
0...
N...
00

0.0

b 4>4 4V4 4 >4

4>4>44

000508.008 .-0..0.<
N 0000.000 00.0000 0:20

5.00.000.
.. .<
0.0.0.0 00.00... 20.0.00 8.0
0 50.0.0 .005. 05m.
0 0000.0 0.80. 9.0
00.0.8: 0:0000:0
.0._> 0.00._0:.00:00.0000. .0.->
5x.00.
. 58050.00 .0 8.0.0000
0
50.0.0 0:08:88 0:80.880
0 00...
8.0000.-:0:
.000.0:000...0 0500.... 50.0.0
. .<>. 00>.
000.0...0008 0500.... 50.0.0
00.9.8:
50:0. 0:0 000.0:000:0
0
.00... 0.8.0 ...:0000 080000.20
0002.0
80:88 ..8.0.000. 5.00.0.0
<0. ._:0000 0.8.0.000.
.. 0008:0008 50.0.0

0.02.0me
mmzmw

w..mw
...<00 .0
002.".
002m.
..mm
xom
.2060

00.2.9...

NZO...0
..<.00._.n.
Zm...n.
0<_>.wn.

mum—n.

<03. ..000

.0H00.0.0.
.0 000000.

.0:~...~0.
.0H.00000.
.0I000000.
.0 .0.0.0.
.mmm..o~0.
.010:0..0.0.
.0 0 .000.0.

.0I0I. .800.

0:00.000.

0:02.000.

.0|-.000.
0101000000.
0.0.1.8000.

0000000.

......000. 0.0 0.00. 300.00.00.00

197

mwozm m>=m2xo 2 $582

3.2% 252$.
2 $539 xomnumfi m>zmmmz

3.29m 252$
2 $532 xomnnmmfi m>=mmmz

toamcmz c2 .39: >50:

23055: B 8:525 .25

w._.
fem
adv
EN
...m
mé
EN
9m
mum
NN

n...
m6

QN
0N
QN
m...

Nwm

<><<>44>4<4>>4444

.mtucocoozE
.N $33.56 233.3%
m 9:206
3336 E .ommmaqsw
N 9.53%
95.23 “.0 83235
P mcEfiEoo
EmEou c059. BEE _m>_>5w
ND 55895
“.o 55:59 Emacmamu £8<
.3322“ :23 2390
89.298 ammofi>£Eavom
.mtoamcmz Bow oEEm 2:260
.mtoamcg EN
553%: 92393 Eauow
29.035: :0.
.90.: E296 83:09:22;
9.385 3.75%

m no:
-mctmmEEEEm .863 mc_o__aw
v so:
-mctmmbcfifia .889. mc_o__am
or no:
-mctmmEcEEE 35$ mc_o__aw
P so:
-mEEwBEEQm .883 mc_o__aw
v
.2525 .2938 ammosEmmE

NDOw
mwoow
NwOOw

Fonz—2w
Naomgzw
mm wOOJw
mm<m04m

250.5
3.5 ..O._m
9150.5

2.. Kim
.=v_w

mwmuw

Vwmmw
Swan—w

Pwmmw

szimww

“mumaopomwvr
flmnxummmmmv.
“mimopavvr
“mnmvommcr
“mnmuoovmvv.
“mnvammvvr
fim mmoomwv
“a m wooomvp
“mumvomuer
“a mummvvr

.mmpe¢o~vr
“a crummvr

“mumuomrmmvr
“mnmnnmwew
“unnumvmmFVF
umuxuuumvmqw

“muovrmvvv

.528. 3 2.3 35529.5

198

 

0000.000
00.09.00. 520.060.00.06

000000.000
5.0.0.0 .30.0EE0=0.-0.0
00.00000 00.09.00.

0000:0000 000.00.000
.0 000.030 00.08.500.000

0000000 v.03). 00.02.00

...N
9N
md
FR
Nd
...v
odm

QN

4)} 4 PD <<>4 <4 4 4}} {D 44

0 0.0.000 00342... 00000... 00.N
00000090300 00.0.00x
< .900.

0.30.0 .0..0....0000 .0.:000>
. 000000000000 00.0.5
.Nm.

000.000 00..00:_000-0...:0.0:
UN 0.00 0.30....

0 0.0.0.0 000000.802...

0 0.0.0.0 00.0.00000
00.0000. 000000.. 0.03....

. 500000000800...
N 0000.80....0000....

_.

0800050.. 3.00. 000005-000»
0000 0003.21...
00000000000550 00.00;...
00.003 0.E00.0.000>w
N 00.00.00
0 03. ..E .0 0000000000 0....2w
. 0.0500 00.050 6.0.0
0 0000000000. .0
5.02.00 000 0000000.. .0005
m 00.0.0.000 0.0.000
5.000.. 0.0.. 8.0.2-030
N 000000 .0885
N 00.0.0.000 0.0E00 ....om.

moz<u.N
on

<u.0m>
Pan—D

.NmmD
ONmmD...

mun—20.2...
Mun—(KI...

5th
N20...

0.0.0..
6......
...<._.

n..moz>w

NOEDw
2.me

m._.<...w
mom—(Pm

N0<...w
Nommaw

.0wmn0000.0.
.0 000.00.

.0H000000.
.0 000000.

.0H0.0000.
.0n.00000.
.0 000000.

.0ImN ..ng

.mmm00000.
.0 0 0000.0.

.0 mw000000.
.0:0000.0.
.0u000.00.
.0 000000.

.0wmw.000.0.
.0 000000.

.0I0I000000.
.0I0I000000.

.0H000.00.
.0 000000.

.0008. 0.0 0.000 0.0052005

199

CHAPTER 6

Shaw, P.J., Fullerton, A.F., Ganey, PE. and Roth, R.A. (2008). The role of

the hemostatic system in a murine model of idiosyncratic liver injury

induced by trovafloxacin and lipopolysaccharide coexposure.

200

6.1 Abstract

The use of the fluoroquinolone antibiotic TVX was severely restricted in
1999 due to its association with idiosyncratic hepatotoxicity. Previously, we
reported that a nontoxic dose of TVX interacts with a nontoxic dose of LPS to
cause robust hepatocellular injury in mice. This interaction with LPS was not
seen in mice treated with LVX, a fluoroquinolone not associated with
hepatotoxicity in people. TVXILPS-coexposure caused an increase in plasma
ALT activity as early as 4.5 h after LPS administration and which progressed
through 15 h. We examined the role of the hemostatic system in TVX/LPS-
induced liver injury. At the onset of liver injury, coexposure to TVXILPS, but not
exposure to TVX, LVX, LPS or LVX/LPS, caused increased plasma
concentration of thrombin-antithrombin dimers and decreased plasma circulating
fibrinogen. LPS treatment induced a small increase in plasma plasminogen
activator inhibitor-1 (PAl-1) concentration, and TVX pretreatment enhanced this
effect. TVX/LPS coexposure also resulted in hepatic fibrin deposition.
Anticoagulant heparin administration reduced TVXILPS-induced hepatic fibrin
deposition and liver injury. PAl-1"’ mice treated with TVX/LPS exhibited similar
fibrin deposition to wild-type mice but had significantly reduced hepatocellular
injury. PAl-1"' mice, but not heparin-treated mice, had reduced plasma
concentrations of several cytokines compared to TVXILPS-treated controls. In
summary, TVXILPS-coexposure caused an imbalance in the hemostatic system,
resulting in increased thrombin activation, plasma concentrations of PAH and

hepatic fibrin deposition. Both thrombin activation and PAl-1 play a critical role in

201

the progression of TVXILPS-induced liver injury, but through different modes of

action.

202

6.2 Introduction

The hemostatic system encompasses a number of factors involved in the
complex interactions of platelets, von Willebrand factor, the coagulation system,
anticoagulants and the fibrinolytic system. The activation of blood coagulation
occurs predominantly through the tissue factor pathway, which leads to the
formation of thrombin from prothrombin. Thrombin is a protease with a number of
biological activities including the cleavage of circulating fibrinogen to fibrin, which
polymerizes to form insoluble fibrin. Dissolution of fibrin is mediated by plasmin.
The fibrinolytic system is controlled primarily by plasminogen activator inhibitor-1
(PAl-1), which inhibits the production of active plasmin. In addition to its role in
fibrinolysis, PAl-1 has several other proinflammatory properties (146, 147).

The coagulation and fibrinolytic systems exist in a delicate balance to
prevent widespread blood loss while controlling fibrin deposition. If the balance of
these two components is altered, a possible outcome is unregulated activation of
the hemostatic system, which could lead to fibrin deposition and production of
occlusive thrombi. These in turn have the potential to alter blood flow and result
in local tissue hypoxia, thereby contributing to tissue injury (103, 152, 154).

Whether an alteration in the hemostatic system occurs in the TVX/LPS-
coexposure model of liver injury in mice has not been determined. The studies
presented here were designed to test the hypothesis that TVXILPS-coexposure
in mice results in hemostatic system dysregulation and that this plays a role in
the development of TVXILPS—induced liver injury. To examine this hypothesis,

biomarkers of thrombin activation, active PAl-1 concentrations and hepatic fibrin

203

deposition were evaluated at the time of onset of liver injury. To determine the
importance of the hemostatic system in the development of TVXILPS-induced
liver injury, mice were treated with anticoagulant heparin. The importance of PAI-
1 in TVXILPS-induced hepatotoxicity was determined using PAl-1"’ mice.
Heparin and PAl-1 can have profound effects on inflammation. Heparin
attenuates ischemialreperfusion-induced inflammatory responses (153), whereas
PAl-1 induces proinflammatory cytokine expression (147). Therefore, we looked
to determine what role thrombin activation and PAl-1 play in the TVX/LPS-
induction of proinflammatory cytokines, some of which contribute to the

pathogenesis in this model (216).

204

6.3 Materials and Methods

6.3.1 Materials

Please refer for Section 2.3.1 for information on this topic.

6.3.2 Animals

Please refer to Section 2.3.2 for information on this topic. In addition, PAI-
1"' and C57Bl/6J mice were purchased from Jackson Laboratory (Bay Harbor,
ME).

6.3.3 Experimental protocols

Mice fasted for 12 h were given various doses of TVX or Veh (saline) by
oral gavage. They were then given LPS (2.0 x 106 EUIkg) or Veh (saline) by
intraperitoneal injection 3 h later. Food was returned immediately after this
dosing. Mice were anesthetized with sodium pentobarbital (50 mglkg; i.p.) and
killed at designated times after the dose of LPS or Veh for various measurements.
Blood was drawn from the vena cava into a syringe containing sodium citrate
resulting in a final concentration of 0.76%. The left lateral lobe was fixed in 10%

neutral buffered formalin and paraffin blocked.

6.3.4 Heparin treatment
Heparin (3000 U/kg, so.) was administered at 2, 6 and 10 h after LPS

administration. Mice were killed at 15 h for experiments used in the generation of

205

Figs. 6.5A, 6.7A, and 6.70. Mice sacrificed at 4.5 h (Figs. 6.58, 6.50, 6.8) were

treated with heparin only at 2 h after LPS dosing.

6.3.5 Histopathology

Please refer to section 2.3.4 for information on this topic.

6.3.6 Hemostatic system measurements

Please refer to Section 4.3.7 for information on this topic.

6.3.7 Cytokine measurements

Please refer to Section 3.3.5 for information on this topic.

6.3.8 Statistical analyses

All bar graph results are presented as mean 1 S.E.M. A 1- or 2-way
analysis of variance (ANOVA) was used as appropriate after data normalization.
All multiple pairwise comparisons were done using Tukey’s Test. The criterion for

significance was p < 0.05.

206

6.4 Results

6.4.1 Changes in the hemostatic system in TVXILPS-induced liver injury

TVXILPS-coexposure causes hepatotoxicity in mice, whereas treatment
with TVX, LVX, LPS, or LVX/LPS is nontoxic (216). In a preliminary study in
TVXILPS-treated mice, the plasma concentration of TAT dimers, a biomarker of
coagulation system activation, peaked at 4.5 h after LPS (data not shown). This
corresponds to the onset of liver injury (see Fig. 3.1). In further studies at 4.5 h,
treatment with TVX, LVX, LPS or LVX/LPS did not significantly affect the plasma
concentration of TAT dimers, whereas TVXILPS-treated mice had a significant
elevation (Fig. 6.1A).

Circulating fibrinogen is consumed during coagulation system activation.
Treatment with either TVX or LVX alone did not affect the plasma fibrinogen
concentration (Fig. 6.1 B). Administration of LPS alone caused a small increase in
plasma fibrinogen concentration (acute phase response) which was unaffected
by pretreatment with LVX. In contrast, TVXILPS-coexposure caused a marked
decrease in plasma fibrinogen concentration (Fig. 6.18), suggesting coagulation
system activation.

The plasma concentration of active PAl-1 was measured as a marker of
fibrinolytic system downregulation. No effect on active PAl-1 concentration was
seen when mice were treated with TVX or LVX alone (Fig. 6.2). LPS treatment

caused a small increase in plasma active PAI-1 concentration, which was

207

Fig. 6.1. Effect of fluoroquinolones and LPS on coagulation system
activation. Mice were treated as described in Materials and Methods with TVX,
LVX or Veh and then 3 h later with LPS or Veh. They were sacrificed 4.5 h after
LPS administration, and plasma concentrations of TAT (A) and fibrinogen (B)
were measured. n = 4-6 animals/group.*significantly different from respective

control group without LPS treatment. ” significantly different from Veh/LPS group.

208

 

 

 

       

  

#
T
LPS

 

“{‘{“{“““““
000.00.x...“.....03000
v.0.0>0»0.Q.0.0.0.0.0.Qv0.0.0. O O

TTI

    

 

* #
LPS
*

 

’ D ’

 
 

 

 

_b

 

 

 

 

 

 

Veh

 

Veh

 

 

 

 

mTVX
-LVX
mTVX
—va

1:: Veh
1:: Veh

 

 

 

 

100

0 0 m 0 0 0
30.0.... 00000.5...

3.0... :0.

209

Fig. 6.2. Effect of fluoroquinolones and LPS on plasma concentration of
active PAl-1. Mice were treated as described in Materials and Methods with TVX,
LVX or Veh and then 3 h later with LPS or Veh. They were sacrificed 4.5 h after
LPS administration, and plasma concentration of PAH was measured. n = 4-6
animals/group.*significantly different from respective control group without LPS

treatment. " significantly different from Veh/LPS group.

210

active PAl-1 (nglmL)

§

55
O

S

8
o

 

=Veh
mTVX
—LVX

    

       
      
       
       
     
     
 

   

9:

9"...
V
A

.9
A

6",- .
0 go

9
O
A

-.-
o o
9’ :0

A

O
o
’o

A. . A

.0
A

     

‘9

  

 

 

 

Veh

211

 

markedly enhanced by TVX pretreatment (Fig. 6.2). In contrast, LVX
pretreatment did not affect the LPS-induced increase in active PAI-1 (Fig. 6.2).
When the rate of fibrin polymerization exceeds the rate of fibrinolysis,
fibrin deposition occurs. Administration of TVX or LVX alone did not result in
hepatic fibrin deposition (Fig. 6.3). Treatment with LPS alone caused a slight, but
significant increase in sinusoidal fibrin when compared to VehNeh-treated mice
(Fig. 6.3). LVX pretreatment did not affect the LPS-induced fibrin deposition, but
TVXILPS-cotreated mice had significantly increased hepatic sinusoidal fibrin

compared to both VehNeh- and LPSNeh-treated mice (Fig. 6.3).

6.4.2 Anticoagulant heparin protects from TVXILPS-induced liver injury

To determine if the activation of the coagulation system seen in TVX/LPS-
treated mice plays a critical role in the progression of liver injury, mice were
treated with anticoagulant heparin. Heparin inhibits coagulation system activation
by increasing the affinity of endogenous antithrombin for thrombin and other
serine proteases of the coagulation pathway. Extensive fibrin deposition
observed in the livers of TVXILPSNeh-treated mice at 4.5 h was significantly
reduced by heparin treatment (Fig. 6.4A). Treatment with heparin also reduced
the increase in plasma ALT activity in mice given TVX/LPS (Fig. 6.43). This
protection was confirmed by histopathological examination. TVX/LPSNeh-
treated mice developed large lesions of necrosis with evidence of apoptosis in
some, whereas TVX/LPS/heparin-treated mice had less frequent and smaller

necrotic lesions (Fig. 6.5, top).

212

Fig. 6.3. Effect of fluoroquinolones and LPS on hepatic sinusoidal fibrin
deposition. Mice were treated as described in Materials and Methods with TVX,
LVX or Veh and then 3 h later with LPS or Veh. They were sacrificed 4.5 h after

LPS administration, and hepatic fibrin deposition was quantified

immunohistochemically as described in Materials and Methods. n 4-6
animals/group.*significantly different from respective control group without LPS

treatment. # significantly different from Veh/LPS group.

213

.°
.5
O

i=1Veh
mrvx
-|_vx

9.0.0
288

Sinusoidal Fibrin Deposition
(Fraction Total Area)
9
S

 

 

 

 

 

*#

 
  

     
   
    
     
   
   
   
         
       
 

':3°8:$’3’3’3‘.
v v v v ‘

$333..

’9

'Q
A

   
   
        
    
  

:0
.0

O
O

6:
0.

.
A

 

'0

   

 

 

 

Veh

 

 

-,_¢,,;

 

VehNeh ..

 

 

 

 

 

o ', '
‘ h - .
4
2?.» . -
. ‘ ’, °
51-. - ..
. I . t
. /~ ,fi,
‘ .
.. 7
. .-
M I ' .
I.
I y“ o
- 3.' s
I
. ~ . .
. . . . ....
.’ ‘ b.
... ,
. .‘
p

 

 

 

 

 

 

 

 

 

 

These data were generated by Aaron Fullerton.

214

 

Fig. 6.4. Effect of heparin treatment on TVXILPS-induced hepatic fibrin
deposition and liver injury. Mice were treated as described in Materials and
Methods with TVX/LPS and heparin or saline vehicle. (A) They were sacrificed at
4.5 h, and fibrin deposition was quantified as described in Materials and Methods.

n - 4-6 animals/group.*significantly different from respective control group

without TVX/LPS treatment. " significantly different from TVX/LPS group not
treated with heparin. (B) Mice were sacrificed 15 h after LPS dosing, and plasma
ALT activity was measured. n = 8-12 animals/group.*significantly different from

TVXILPSNeh-treated mice.

215

Slnusoldal Flbrln Deposltlon
(Fractlon Total Area)

ALT (UIL)

 

 

 

0.25 l *
— Control
020 _ L:I Heparin
0.15 -
0.10 -
0.05 -
0.00 -
VehNeh TVXILPS
8000 _ — Control
:2: Heparin
6000 —
4000 - *
T
2000 « .L
o _

 

 

 

 

 

 

TVX/LPS

These data were generated by Aaron Fullerton.

216

Fig. 6.5. Histopathology of heparin-treated and PAl-1"' mice treated with
TVXILPS. Mice were treated as described in Materials and Methods with
TVXILPSNeh or TVXILPS/heparin (top). In addition, wild-type or PAl-1"' mice
were treated with TVX/LPS as described in Materials and Methods (bottom).
Photomicrographs were taken of representative livers from mice sacrificed at 15

h after LPS administration.

217

 

 

 

.../Hie... u»

. ...?

...

r . ...siy .
'

 

 

   

. t .I

czaozamuii

..,\.-.»

 

218

6.4.3 PAl-1"' mice are protected from TVXILPS-induced liver injury

To determine if PAl-1 plays a role in hepatic fibrin deposition at the onset
of liver injury, wild-type and PAl-1"' mice were treated with TVX/LPS and killed at
4.5 h. TVX/LPS-coexposurecaused significant fibrin deposition in wild-type mice
as well as in PAl-1"' mice (Fig. 6.6A). To confirm that PAl-1"' mice do not have
reduced fibrin deposition at a different time, mice were killed at 15 h and hepatic
fibrin deposition was quantified. At 15 h, PAl-1"' mice had the same amount of
fibrin deposition as wild-type mice (Fig. 6.68) but had significantly reduced
hepatocellular injury compared to wild-type mice after TVXILPS-coexposure (Fig.
6.6C). Histopathologic evaluation of liver sections confirmed this protective effect.
Midzonal lesions of hepatocellular necrosis and apoptosis observed in TVX/LPS-
treated wild-type mice (Fig. 6.5, bottom) were less frequent and smaller in PAl-1"‘
mice (Table 6.1). TVX/LPS coexposure caused grade 4 or 5 (most severe)
necrosis in 8/14 wild-type mice, whereas none of the 9 PAl-1"' mice were grade 4
or 5. PAl-1"' mice showed a trend toward decreased inflammation and

hemorrhage that did not reach statistical significance.

6.4.4 Role of coagulation system activation and PAl-1 in TVXILPS-induction
of cytokines

Heparin-treated, wild-type and PAI-1"' mice were cotreated with TVXILPS
and killed at 4.5 h to determine what roles heparin and PAl-1 have in the
TVXILPS-induction of inflammatory cytokines. Heparin treatment did not change

the TVXILPS-induced increases in lL-18, IL-6, KC, lL-10, MCP-1, VEGF, or

219

Fig. 6.6. Effect of PAH deficiency on TVXILPS-induced hepatic fibrin
deposition and liver injury. Wild-type and PAl-1"' mice were treated as
described in Materials and Methods with TVX and then 3 h later with LPS. They
were sacrificed at 4.5 h (A) or 15 (B), and fibrin deposition was quantified as
described in Materials and Methods. n = 4-6 animals/group.*significantly different
from VehNeh. (C) Mice were sacrificed 15 h after LPS dosing, and plasma ALT

activity was measured. n = 13-22 animals/group.*significantly different from wild-

type.

220

A

 

*Tll.

 

 

 

TVXILPS

    

VehNeh

- Wild-type

C: PAI-1 "'

 

a

o

mwmmm

nm 0. 0 0 0
.82 32 5.88":
528%.. $5... 328:...»

 

 

 

- Wild-type

=1 PAl-l '4'

 

 

 

m m m. m.
0 o o o
:25 .23 5:08“:
8:83.. :3: 3282.5

w.
o

TVXILPS

VehNeh

— Wild-type
i=i PAl-1"'

 

 

 

 

 

 

4000 l

C

3000 «
00

3:: S<

TVXILPS

221

Table 6.1. Scoring of histopathology of livers from wild-type and PAI-1"'

 

 

 

mice cotreated with TVXILPS.
Mouse strain Necrosis Inflammation Hemorrhage ,
Wild-type 3.2 i 0.4 1.4 d: 0.1 0.6 :l: 0.2
PAl-1"' 2.0 :l: 0.2 * 1.0 1: o o 1 o

 

 

Wild-type and PAI-1"' mice were treated with TVX and LPS as described in
Materials and Methods. They were sacrificed 15 h after LPS. Liver sections were
cut 5 mm thick and stained with hematoxylin and eosin, and resultant slides were
scored by a pathologist for necrosis, inflammation and hemorrhage. The scoring
scale was set from 0-5 with the following criteria: no observation (0), mild (1),
mild to moderate (2), moderate (3), moderate to severe (4), and severe (5).
Scores are reported as average :I: S.E.M. n = 9-15 animals/group.*significantly

different from wild-type group.

222

TNFa. However, it enhanced the TVXILPS-induction of lL-1B and lFNy (Fig. 6.7).
In contrast, PAl-1"' mice had significantly reduced levels of several cytokines
after TVX/LPS treatment. These included lL-1B, lL-6, KC, lL-10, MCP-1, and
IFNy (Fig. 6.7). There wasa trend toward decreased plasma concentration of
TNFa in PAl-1"' mice, but this was not statistically significant. Of the cytokines

measured, only lL-18 and VEGF were increased to the same degree in wild-type

and PAl-14' mice by TVXILPS-coexposure (Fig. 6.7).

223

Fig. 6.7. Effect of heparin treatment or PAl-1 deficiency on TVXILPS-
induced increases in plasma cytokines. Mice were treated with vehicles or
with TVXILPS and sacrificed 4.5 h after LPS administration. Plasma
concentrations of lL-1B, lL-18, lL-6, KC, lL-10, MCP-1, VEGF, lFNy, and TNFa
were measured as described in Materials and Methods. n = 4-6
animals/group.*significantly different from VehNeh. # significantly different from

TVX/LPS control group.

224

uW////////////%

 

 

 

1111111

.. Theme 3....

TVX/LPS

Veh

         

 

 

 

 

 

.W//////////%% *M/%///////%

 

TVX/LPS

Veh

Eff“: PAi-1+
225

’////A Heparin

- Control

Fig. 6.7 (cont’d).

lL-18 (pglm L)

 

 

 

 

Veh TVX/LPS
‘0 .
3 3° ‘
E
2’ 2.
‘9
i 10 <
o .
TVX/LPS
4000 1
3 3000 .
E
U)
3. 20004
2
=i' 1000
o .
TVX/LPS
- Control
‘ ff PAM"
m Heparin

226

Fig. 6.7 (6°"t'd)'

TVX/LPS

Veh

*
*
.I
o o o m
m m m

TVX/LPS

eh

#
*
an
*
*
v
mumnmuo

4|

...

   

 

...
i .3

 

... W .
AJctmavdn—ZF

_ TVX/LPS

Veh

- C°“"°'

f PAI-l"

W/IA Heparin

227

6.5 Discussion

Previous studies showed that TVX, but not LVX, interacts with a nontoxic
dose of LPS to cause hepatotoxicity in mice (216). The studies presented here
examined whether TVXILPS-coexposure altered the hemostatic system and, if so,
what role the alteration played in the pathogenesis. A detailed timecourse of
TVXILPS-induced liver injury showed that liver injury begins at ~4.5 h and peaks
at 15 h after LPS administration (see Fig. 3.1). To understand the importance of
the hemostatic system in the progression of liver injury, hemostatic system
biomarkers were measured at the onset of liver injury.

The hemostatic system normally functions as a balance between
prothrombotic and antithrombotic factors to maintain homeostasis. Based on an
increase in thrombin activation and a decrease in plasma fibrinogen
concentration, TVXILPS-coexposure activated the coagulation system, but TVX,
LVX, LPS or LVX/LPS treatment did not (Fig. 6.1). LPS treatment alone induced
a small, but statistically significant increase in plasma fibrinogen. This response
has been seen previouSly and likely reflects an acute phase response to LPS
(267). Coagulation system activation accompanying liver injury occurs after
cotreatment with LPS and other agents (268-270). PAl-1 inhibits fibrinolysis via
inhibition of plasminogen activators (271). Like its effect on coagulation system
activation, TVX, but not LVX, enhanced the LPS-induced increase in plasma
concentration of active PAl-1 (Fig. 6.2). Whether TVX acts directly to enhance
the LPS induction of coagulation factors and PAl-1 is unknown. It is possible that

TVX directly upregulates tissue factor or acts indirectly by enhancing the LPS

228

induction of cytokines such as TNFa that can induce tissue factor production
(272). TVX enhances the LPS-induced production of TNFa, and it is likely that
this plays a role in the TVXILPS-induced coagulation system activation (216).
TNFa is required for TVXILPS-induced liver injury, and the increased TNFa in
TVXILPS-treated mice affected coagulation system activation and PAI-1
production (Chapter 4).

The TVXILPS-mediated coagulation system activation and impairment of
the fibrinolytic system suggests an imbalance in the hemostatic system that
favors fibrin formation. Indeed, TVXILPS-coexposure did result in increased in
fibrin deposition compared to all other treatment groups (Fig. 6.3). Sinusoidal
fibrin can impair hepatic blood flow, and thus oxygen delivery to hepatocytes,
causing tissue hypoxia (103). Hypoxia could play a role in the pathogenesis of
TVXILPS-induced liver injury in a number of ways. For example, it sensitizes
hepatocytes to cytotoxicity by proteases and hypochlorous acid (HOCl) released
by activated neutrophils (103). In addition, hypoxia can alter cellular homeostasis
by inducing oxidative stress and can directly cause cell death (273-275).

To address whether thrombin activation was involved in the pathogenesis
of liver injury, heparin was used to inhibit coagulation. Heparin inhibited thrombin
activation and decreased the TVXILPS-induced fibrin deposition which in turn
protected mice from TVXILPS-induced liver injury (Fig. 6.4). This suggests that
coagulation system activation plays a critical role in the fibrin deposition and the
pathogenesis. However, whether hepatic fibrin deposition, thrombin activation of

receptors or both are critical for TVXILPS-induced liver injury is unknown.

229

Heparin can attenuate inflammation (153); accordingly, to address if this
might be involved in its protective effect, plasma cytokine concentrations were
measured. Heparin treatment failed to reduce plasma concentrations of any of
the cytokines induced by TVXILPS-treatment (Fig. 6.7), suggesting that heparin
did not provide protection against liver injury by reducing inflammatory cytokines.
This result also suggests that coagulation system activation acts downstream of
any effects initiated by cytokines. Heparin treatment significantly increased
TVXILPS-induction of IFNy. This enhancement might have been a result of the
inhibition of thrombin, which can drive a Th2 response that downregulates lFNy
expression (142). In addition, after heparin treatment there was a slight, but not
statistically significant decrease in the plasma concentration of VEGF, the
expression of which can be driven by hypoxia (276).

In other models, there is evidence that PAl-1 can play a role in both fibrin
deposition and the progression of liver injury (277-279). However, hepatic fibrin
deposition at both 4.5 and 15 h was similar in wild-type and PAl-1"' mice treated
with TVX/LPS, yet both plasma ALT activity and histopathologic evaluation
indicated lesser hepatocellular injury in the PAl-1"' mice (Fig. 6.5 and 6.6).
Therefore, the results suggest that protection from TVXILPS-induced liver injury
in PAl-1"‘ mice is not related to fibrin deposition.

In addition to its role in fibrinolysis, PAl-1 has proinflammatory properties
including induction of lL-6 (147). Moreover, PAl-1 can enhance LPS-induced
neutrophil activation (146). Relative to wild-type mice, PAI-1"' mice treated with

TVX/LPS had reduced concentrations of lL-1B, IL-6, lL-10, lFNy, MCP-1 and KC

230

at 4.5 h (Fig. 6.7). This result suggests a proinflammatory role for PAI-1 in the
regulation of several cytokines. Our result for IFNy is in contrast to a report that
PAl-1"' mice have an enhanced plasma lFNy response to a toxic dose of LPS
(280). It may be that the induction of IFNy occurs by a different mechanism after
TVXILPS-coexposure compared to administration of a toxic dose of LPS by itself.
It is also possible that the attenuated cytokine response in PAl-1"' mice is
secondary to a reduction in injury; however plasma concentrations were
measured at 4.5 h, when there is only a slight increase in plasma ALT activity
which is not attenuated in PAI-1"' mice (data not shown). Therefore, it seems
likely that the reduced cytokine response underlies the protection seen in PAl-1"’
mice. However, it has been reported that PAl-1"' endothelial cells are resistant to
apoptosis (281). If hepatocytes are also resistant to apoptosis in these mice, this
might play a role that attenuated liver injury in PAl-1"' mice.

In summary, TVX interacted with LPS to enhance coagulation system
activation, increase PAl-1 production and cause liver injury. In contrast, LVX did
not alter the hemostatic system or cause liver injury when coadministered with
LPS. The TVXILPS-induced imbalance in the hemostatic system resulted in
hepatic fibrin deposition. Anticoagulant heparin inhibited thrombin activation
which significantly reduced fibrin deposition and protected mice from TVX/LPS-
induced hepatocellular necrosis without decreasing plasma concentrations of
proinflammatory cytokines. This suggests that active thrombin, perhaps through
enhanced fibrin deposition, contributes to the liver injury. PAl-1"‘ mice did not

have attenuated hepatic fibrin deposition but did have smaller plasma

231

concentrations of several inflammatory cytokines and less TVXILPS-induced liver
injury. Accordingly, coagulation system activation was critical for TVXILPS-
induced fibrin deposition, whereas PAl-1 was not, and the important role of PAl-1
in promoting the liver injury'appears to be unrelated to its ability to downregulate
fibrinolysis. It is likely that thrombin activation and proinflammatory cytokines
comprise two independent pathways that contribute to TVXILPS-induced liver
injury, inasmuch as the removal of either one attenuated, but did not eliminate
liver injury; in order to achieve complete protection from TVXILPS-induced liver

injury the inhibition of both was required.

232

CHAPTER 7

Vascular endothelial growth factor has a proinflammatory role which is

critical to trovafloxacin and lipopolysaccharide coexposure-induced liver

injury.

233

7.1 Abstract

TVX is a fluoroquinolone antibiotic which has caused idiosyncratic
hepatotoxicity in a small fraction of people. Animal models predictive of drugs
that cause lADRs are lacking in preclinical safety testing. Previously, we showed
that a modest inflammatory stress induced by LPS renders mice sensitive to
nonhepatotoxic doses of TVX. In contrast, LVX, a fluoroquinolone antibiotic that
has not been associated with lADRs in humans, does not cause hepatotoxicity in
mice cotreated with LPS. TVXILPS-induced liver injury is dependent on both
coagulation and TNFa pathways. VEGF is a cytokine with proinflammatory
properties which has the potential to be a part of either of the pathways involved
in the pathogenesis. The purpose of this study was to explore interactions
between VEGF and other proinflammatory mediators of liver injury and, in turn,
the role of VEGF in TVXILPS-induced liver injury. TVX prolonged the LPS-
induced increase in plasma VEGF. VEGF neutralization attenuated the
TVXILPS-induced increases in both hepatic fibrin deposition and plasma TNFa
concentration. Additionally, VEGF neutralization protected mice from TVX/LPS-
induced liver injury, as reflected by both plasma ALT activity and liver
histopathology. In summary, VEGF has proinflammatory properties and plays a

critical role in the progression of TVXILPS-induced liver injury.

234

7.2 Introduction

VEGF has been studied extensively for its angiogenic properties. However,
recent work has focused on its proinflammatory properties. Several cytokines can
affect VEGF expression; conversely, VEGF can induce TNFa, neutrophil
chemokines and tissue factor (92, 179, 180). Inasmuch as VEGF has the
potential to interact with several pathways involved in TVXILPS-induced liver
injury, its interactions with various mediators of liver injury were examined. In
accordance with its proinflammatory effects, VEGF is involved in the
development of liver injury in animal models of endotoxemia and
ischemialreperfusion (179, 180).

The purpose of this study was to test the hypothesis that VEGF is involved
in the pathogenesis of TVXILPS-induced liver injury. Furthermore, the potential
interaction of VEGF with neutrophil accumulation, TNFa, IFNy and hepatic fibrin
deposition, all of which play a role in TVXILPS-induced liver injury, was explored
in this study. These pathways involved in the pathogenesis were explored at a
time near the onset of liver injury. Additionally, the study explored novel
proinflammatory properties of VEGF that might be of importance for inflammatory

tissue injury.

235

7.3 Materials and Methods

7.3.1 Materials
Please refer for Section 2.3.1 for information on this topic. VEGF
antiserum and control rabbit serum were a kind gift from Dr. David Briscoe

(Harvard Medical School, Boston, MA).

7.3.2 Animals

Please refer to Section 2.3.2 for information on this topic.

7.3.3 Experimental protocols

Mice fasted for 12 h were given TVX (150 mglkg) or Veh (saline) by oral
gavage. They were then given LPS (2.0 x 106 EU/kg) or Veh (saline) by
intraperitoneal injection 3 h later. Food was returned immediately thereafter. Mice
were anesthetized with sodium pentobarbital (50 mglkg; i.p.) and killed at
designated times after the administration of LPS or Veh for various ,
measurements. Blood was drawn from the vena cava into a syringe containing
sodium citrate resulting in a final concentration of 0.76%. The left lateral lobe was
fixed in 10% neutral buffered formalin and paraffin blocked.

For VEGF neutralization studies, VEGF antiserum or control serum was

administered (0.5 mL, i.p.) 15 h before and (0.8 mL, i.p.) 2 h after LPS.

236

7.3.4 Histopathology

Please refer to section 2.3.4 for information on this topic.

7.3.5 Cytokine measurements

Please refer to Section 3.3.5 for information on this topic.

7.3.6 Neutrophil staining

Please refer to section 3.3.6 for information on this topic.

7.3.7 Hemostatic system measurements

Please refer to Section 4.3.7 for information on this topic.

7.3.8 Statistical analyses

Results are presented as mean i S.E.M. A Student’s t-test or a 2-way
analysis of variance (ANOVA) was used as appropriate after data normalization.
All pairwise comparisons were made using Tukey’s test with the criterion for

significance at p < 0.05.

237

7.4 Results

7.4.1 Timecourse of plasma concentration of VEGF

TVX alone did not have any effect on plasma VEGF concentration (Fig.
7.1). LPS treatment caused a significant increase in plasma VEGF at all times
measured (Fig. 7.1). TVX treatment prior to LPS enhanced the LPS-induced

increase in plasma concentration of VEGF at 4.5 (Fig. 7.1).

7.4.2 Effect of VEGF neutralization on TVXILPS-induced hepatotoxicity

To explore the role of VEGF in TVXILPS-induced liver injury, VEGF was
neutralized by VEGF antiserum, which reduced the TVXILPS-induced increase in
plasma VEGF at 4.5 h (Fig. 7.2A). TVXILPS-treated mice were treated with
control serum or VEGF antiserum and then killed at 15 h. TVXILPS-induced liver
injury was attenuated by VEGF neutralization (Fig. 7.23). Histopathologic
evaluation confirmed that mice were protected from TVXILPS-induced
hepatotoxicity by VEGF neutralization. TVX/LPS coexposure caused
hepatocellular necrosis and apoptosis primarily localized to the midzonal regions
(Fig. 7.3). VEGF neutralization reduced the size and frequency of lesions

induced by TVX/LPS coexposure (Fig. 7.3).

238

Fig. 7.1 Timecourse of VEGF plasma concentration. Mice were treated with
TVX or Veh 3h before LPS or Veh as described in Materials and Methods. They
were killed at various times and plasma concentration of VEGF was measured. n
= 5—7 animals/group. *signiflcantly different from 0 h within same treatment group.

1"significantly different from Veh/LPS treatment group at same time.

239

VEGF (pglmL)

1200 '

 

- -0 - VehNeh
- «3 - TVXNeh

 

 

 

 

Hours after LPS or Veh

240

Fig. 7.2. Effect of VEGF neutralization on TVXILPS-induced plasma VEGF
induction and liver injury. Mice were treated with TVXILPS or VehNeh in
addition to VEGF antiserum or control serum as described in Materials and
Methods. (A) They were killed 4.5 h after LPS administration. Plasma
concentration of VEGF was measured as described in Materials and Methods. n
= 4-6 animals/group. (B) Mice were killed 15 h after LPS administration and
plasma ALT activity was measured. n = 6-10 animals/group. *significantly
different from respective VehNeh group. 1”significantly different from

TVXILPS/control serum-treated group.

241

1000 1

800 4

VEGF (ngmL)
8

§

_ Control
12:: VEGF antiserum

 

 

 

8
8

ALT activity (UIL)

O

 

”5

s
3

.L
O
8

 

 

 

 

VehNeh TVXILPS

- CODUOI *
D VEGF antiserum

 

 

 

VehNeh TVXILPS

242

Fig. 7.3. Protection from TVXILPS-induced liver pathology by VEGF
neutralization. Mice were treated with TVX/LPS or VehNeh in addition to VEGF
antiserum or control serum as described in Materials and Methods. They were
killed 15 h after LPS administration, and photomicrographs were taken of

representative livers.

243

    

   

\ U 5027'5.‘

..m‘z; - 1.;- ”.49.

 
  
 

4 .
5" .- ' “a?“ 7*? . ...-
' bkiyfvfffizxfi 3’37"

  

4!

      
 

   
  
   

    
   

~ . 1.9. . .

'\ ‘ '. 5 Q l _ ‘
Pf); ‘5 ‘33?“ ‘ “C‘<J§.f ‘
’1’“ 3 :5 tsfii‘fxrt‘ll‘lat- .

    

-‘. I -: .

. "l I;"'\\"};‘ ‘13 '5‘ 9“
.'..-v‘.-;. J. '

- V 3“ ‘ ..is

.; n 5 ’ :
TVXILPSIanti-VEGF

244

7.4.3 Effect of VEGF neutralization on plasma concentrations of cytokines
Mice were treated as described in Materials and Methods and killed at 4.5
h, a time near the onset of liver injury. VEGF antiserum by itself did not affect the
plasma concentrations of any cytokines or chemokines measured (Fig. 7.4 and
7.5). TVX/LPS coexposure caused an increase in the plasma concentrations of
TNFa, lL—10, lL-1l3, lL-6, lL-18 and IFNy (Fig. 7.4). VEGF neutralization reduced
the TVXILPS-induced increase in TNFa and lL-10, but did not affect the induction
of other cytokines (Fig. 7.4). In addition to inducing a number of cytokines,
TVX/LPS coexposure caused an increase in the chemokines MCP-1, KC and

MlP-1a, and this increase was unaffected by VEGF neutralization (Fig. 7.5).

7.4.4 Effect of VEGF neutralization on TVXILPS-induced hepatic neutrophil
accumulation

Mice were treated with TVX/LPS and killed at 4.5 to measure hepatic
neutrophil accumulation. TVX/LPS coexposure caused neutrophils to accumulate

in the liver, and this was unaffected by VEGF neutralization (Fig. 7.6).

7.4.5 Effect of VEGF neutralization on the hemostatic system

TVX/LPS coexposure altered the balance of the hemostatic system to
cause hepatic fibrin deposition (See Chapter 6). Hemostatic system biomarkers
were measured at 4.5 h, a time at the onset of liver injury. VEGF antiserum by

itself did not cause any changes in TAT dimers, active PAl-1 or hepatic fibrin

245

Fig. 7.4. Effect of VEGF neutralization on TVXILPS-induced increase of
cytokines. Mice were treated with TVX/LPS or VehNeh in addition to VEGF
antiserum or control serum as described in Materials and Methods. They were
killed 4.5 h after LPS administration. Plasma concentrations of TNFa, lL-10, IL-
10, lL-6, lL-18 and lFNy were measured as described in Materials and Methods.
n = 4-6 animals/group. *significantly different from respective VehNeh group.

“significantly different from TVXILPS/control serum-treated group.

246

 

 

 

 

 

 

3000
TE" 6000 "El soon
3 E * t
V 4000
l; 3000 3.
l- :' 2000
o o __ __
VehNeh TVX/LPS
000 400 - *
3 “W A 300 -
g 400 g zoo
SE- “? .
:' 200 i 1°01
0 ° ‘ __
VehNeh TVX/LPS
35m 120 i *
sooo * 100
.i 2500 * 3 so.
E L3
E 2000 5 so.
3' 1500 r-
‘7 z 40-
: 1“” E
500 20‘
° _"‘ __ 0‘ __
VehNeh TVX/LPS VehNeh TVX/LPS

- Control
VEGF antiserum

247

Fig. 7.5. VEGF neutralization did not affect TVXILPS-induced increases in
chemokines. Mice were treated with nTVX/LPS or VehNeh in addition to VEGF
antiserum or control serum as described in Materials and Methods. They were
killed 4.5 h after LPS administration. Plasma concentrations of MOP-1, KC and
MlP-1a were measured as described in Materials and Methods. n = 4-6
animals/group. *significantly different from respective VehNeh group.

”significantly different from TVXILPS/control serum-treated group.

248

 

 

*
TVXILPS

 

 

- Contra
175 ‘ =- VEGF aritlssrum
VohNd‘l

memes...

35.9.. Eu...

200 «

 

—COntrol

 

TVXILPS

VehNeh

 

 

150 - — VEGantlserum

175 1

J

1

35.9.. 9.

one

 

 

  

—VEGFantlsenln

m
_

TVXILPS

249

 

mwmm

35.2. eves.

Fig. 7.6. Effect of VEGF neutralization on TVXILPS-induced hepatic
neutrophil accumulation. Mice were treated with TVX/LPS or VehNeh in
addition to VEGF antiserum or control serum as described in Materials and
Methods. They were killed 4.5 h after LPS administration. Liver sections were

stained immunohistochemically for neutrophils and quantified as described in

Materials and Methods. n - 4-6 animals/group. *significantly different from

respective VehNeh group.

250

PMNs per HPF

801
60'

40‘

 

 

- Control
:23 VEGF antiserum

 

 

 

 

 

 

VehNeh TVXILPS

251

deposition (Fig. 7.7). VEGF neutralization caused a trend towards a decrease in
TVXILPS-induced increase in TAT dimers (Fig. 7.7A). The plasma concentration
of active PAl-1 was increased by TVX/LPS coexposure, and this was unaffected
by treatment with VEGF antiserum (Fig. 7.73). Hepatic fibrin deposition was
increased by TVX/LPS coexposure (Fig. 7.7C). VEGF neutralization significantly
attenuated the TVXILPS-induced increase in fibrin deposition in the liver (Fig.

7.70).

252

Fig. 7.7. The role of VEGF in TVXILPS-induced hemostatic system
activation. Mice were treated with TVX/LPS or VehNeh in addition to VEGF
antiserum or control serum as described in Materials and Methods. They were
killed 4.5 h after LPS administration. Plasma concentrations of (A) TAT dimers
and (B) active PAl-1 were measured as described in Materials and Methods. (C)
Frozen liver sections were stained immunohistochemically for fibrin and
quantified as described in Materials and Methods. n = 4-6 animals/group.
*significantly different from respective VehNeh group. #significantly different from

TVXILPS/control serum-treated group.

253

A

TVXILPS

     

m VEGF antiserum
VehNeh

— Control

 

 

m m m m m m o
3.9.. was... e5

 

 

 

 

*TT‘

m

M

In

mm .
0

... _
1m m m m m

1

3 5.9.. Es". 258

TVXILPS

VehN eh

_ Control

 

 

m m m M
n. n. 0 0.
32a 33 .3392...

Sauce-o 5...... less...»

TVXILPS

VehNeh

254

7.5 Discussion

VEGF is produced by endothelial cells, macrophages, activated T cells,
and a variety of other cell types as a result of several stimuli, including hypoxia,
lL-1B, lL-1a, lL-6, oncostatin M, TNFoc and lL-8 (163-169). LPS caused an
increase in plasma VEGF concentration at 1.5, 3, 4.5 and 6 hours. TVX
treatment enhanced the LPS-induced increase in VEGF (Fig. 7.2). It is possible
that this is a direct effect of TVX to enhance the, stimulation of one or more cell
types to produce and release VEGF in response to LPS. However, it is likely that
the prolonged VEGF plasma concentration is a secondary effect. TVX enhanced
the LPS-induced increase in TNFa (Fig. 2.4), which can induce VEGF expression
in macrophages (91). Additionally, TVX/LPS coexposure causes hemostatic
system dysregulation which leads hepatic fibrin deposition (Chapter 6). Fibrin
deposition can alter blood flow and cause local tissue hypoxia, which induces
VEGF expression (151). Therefore, it is possible that the prolongation of VEGF is
a downstream effect in a cascade of inflammatory events.

To address whether VEGF is involved in the pathogenesis of liver injury,
VEGF antiserum was used to neutralize it. VEGF neutralization decreased the
TVXILPS-induced increase in plasma concentration of VEGF. VEGF activity was
not measured, and it is likely that VEGF antiserum reduced activity to a greater
degree than the plasma concentration of VEGF. However VEGF neutralization
attenuated TVXILPS-induced liver injury (Fig. 7.2 and 7.3), suggesting that VEGF

plays a critical role in the pathogenesis.

255

To determine possible mechanism(s) by which VEGF contributes to
TVXILPS-induced liver injury, plasma cytokine concentrations and hemostatic
system parameters were measured. TVX enhanced the LPS induction of several
proinflammatory cytokines, of which TNFa and lFNy are known to be involved in
the pathogenesis (Chapters 2, 3 and 5). To determine if VEGF plays a role in the
induction of cytokines, mice were killed at a time near the onset of liver injury.
VEGF neutralization did not alter the TVXILPS-induced increase in lL-1B, lL-6,
IL-18 or lFNy. However, it attenuated the TVXILPS-induced increase of TNFa
and lL-10 (Fig. 7.4). The reduced lL-10 plasma concentration is likely due to the
reduced plasma concentration of TNFa, inasmuch as lL-10 induction is primarily
TNFa-dependent (282). The ability of VEGF to enhance TNFa concentration is
consistent with a report in which VEGF neutralization reduced the plasma
concentration of TNFa in a model of sepsis (179). It is possible that this effect is
mediated by activation of early growth response-1, a transcription factor which
can be activated by VEGF and which increases TNFa expression (92). TNFa
neutralization attenuated the TVXILPS-induced increase of VEGF (Fig. 4.2). It is
therefore possible that VEGF and TNFa upregulate one another’s production,
creating a vicious proinflammatory cycle which could contribute to TVXILPS-
induced liver injury.

TVX/LPS coexposure induced increases in the plasma concentrations of
several chemokines, KC, MCP-1 and MlP-1a, and these were unchanged by
VEGF neutralization (Fig. 7.5). This result was surprising, inasmuch as VEGF

neutralization reduced TNFa concentration and TNFa neutralization attenuated

256

the TVXILPS-induced increase in these chemokines (Fig. 4.2 and 4.3). However,
it is possible that VEGF neutralization did not reduce the concentration of TNFa
to a great enough extent to see a difference in chemokine induction.

The TVXILPS-induced increase in KC, MCP-1 and MlP-1a appears to be
VEGF-independent, but it was possible that VEGF played a role in neutrophil
accumulation independent of these chemokines. Accordingly, livers were stained
immunohistochemically for PMNs. VEGF neutralization did not affect TVX/LPS-
induced neutrophil accumulation in the livers of treated mice (Fig. 7.6). This
result is in contrast to a study which found that VEGF neutralization significantly
attenuated MCP-1 production and hepatic neutrophil accumulation induced by
ischemialreperfusion (180). Thus, it is likely that the mechanism of MCP-1
induction and hepatic PMN accumulation between TVX/LPS coexposure and
ischemialreperfusion is different. It is possible that VEGF interacted with
neutrophils to enhance activation, which is critical for TVXILPS-induced
hepatotoxicity.

TVX/LPS coexposure caused hemostatic system dysregulation, which is
involved in the pathogenesis (see Chapter 6). VEGF can potentially interact with
the hemostatic system in several ways. For example, VEGF induces tissue factor,
which causes coagulation system activation (183), and PAl-1, which
downregulates fibrinolysis (184). VEGF neutralization caused trends toward
attenuated plasma concentrations of TAT dimers and PAl-1. The TVX/LPS-
induced increase in PAl-1 and hepatic fibrin deposition was TNFaodependent

(Fig. 4.5); therefore, it is possible that the reduction by VEGF neutralization is

257

secondary to its effect on TNFa. The trends toward a decrease in concentrations
of TAT dimers and PAI-1 may have resulted in reduced hepatic fibrin deposition
after TVX/LPS coexposure (Fig. 7.7). The ability of VEGF neutralization to
reduce hepatic fibrin deposition is a novel finding and might be involved in other
models of inflammatory tissue injury. TVXILPS-induced hepatic fibrin deposition
might be involved in the pathogenesies. Additionally, fibrin deposition could alter
blood flow and cause local tissue hypoxia, which can drive VEGF expression
(151). Inasmuch as VEGF neutralization reduced fibrin deposition, it is possible
that fibrin deposition and VEGF enhance one another to create a feedforward
cycle involved in pathogenesis.

In summary, TVX prolonged the LPS-induced increase of VEGF and
caused hepatotoxicity. VEGF neutralization attenuated liver injury caused by
TVX/LPS coexposure. Additionally, VEGF neutralization reduced the TVX/LPS-
induced increase in TNFot and hepatic fibrin deposition. It is possible that VEGF
is involved in dysregulated feedforward cycles with both TNFa and the

hemostatic system, which could drive the development of liver injury.

258

CHAPTER 8

Summary and conclusions

259

8.1 Summary of research

At the outset, the hypothesis was tested that TVX interacts with an
inflammatory stress to cause idiosyncrasy-like liver injury in mice. Initial dose
response studies were conducted to find nontoxic doses of TVX (Fig. 2.1A), LPS
and PGN-LTA. TVX, up to 1000 mglkg, was administered orally to mice without
finding a hepatotoxic dose. TVX administered 3 h before a nonhepatotoxic dose
of LPS, which activates TLR4, caused hepatocellular injury. Preliminary dose-
response studies revealed that mice treated with 150 mglkg TVX before LPS
developed significant liver injury with a mortality rate less than 10% (Fig. 2.1A);
therefore, 150 mglkg TVX was used for all subsequent studies. Additionally, TVX
interacted with an inflammatory stress induced by PGN-LTA coexposure, which
activates TLR2, to cause liver injury (Fig. 3.1). A timecourse study of TVX/LPS-
and TVX/PGN-LTA-induced liver injury found that plasma ALT activity was
increased by 4.5 h and progressed to a maximum by 15h (Fig. 2.13, 3.1). The
primary histopathologic finding in TVXILPS-treated mice was lesions of
hepatocellular necrosis and apoptosis primarily in midzonal regions (Fig. 2.3).
TVX-PGN-LTA-treated mice had similar lesions, but these were primarily located
in centrilobular regions of the liver (Fig. 3.2). Overall, these results showed that
TVX interacts with an inflammatory stress, induced by LPS or PGN-LTA, to
cause hepatotoxicity in mice.

In a subsequent study, an equiefficacious dose of LVX did not interact with

LPS coexposure to cause liver injury, whereas TVX/LPS coexposure did (Fig.

260

2.2). Accordingly, the propensity of the fluoroquinolones to cause idiosyncratic
hepatotoxicity in people tracked with the potential of each to interact with LPS
coexposure to cause hepatotoxicity in mice. Inasmuch as LVX/LPS did not cause
liver injury, type II topoisiomerase inhibition alone by fluoroquinolones is not
sufficient to interact with LPS to cause liver injury.

Neutrophils are critical to the pathogenesis of several models of liver injury.
Both LPS and PGN-LTA induced accumulation of neutrophils within the liver
which was unaffected by TVX pretreatment (Fig. 3.5). Prevention of neutrophil
activation by CD18 antiserum attenuated both TVX/LPS- and TVX/PGN-LTA-
induced liver injury (Fig. 3.6). However, CD18 antiserum did not affect TVX/LPS-
induced increases in plasma concentrations of cytokines (Fig. 3.8). These results
suggest that neutrophil activation plays a role which is independent of the
induction of proinflammatory cytokines in TVXILPS-induced liver injury.

TNFa is critical to several models of liver injury, therefore we hypothesized
that this cytokine plays a role in TVXILPS-induced hepatotoxicity via interactions
with other factors known to be important in inflammatory liver injury. These
included PMNs, IFNy, VEGF and the hemostatic system. TVX enhanced the
LPS-induced increase in plasma TNFa (Fig. 2.4). In contrast, LVX did not affect
this. TNFa neutralization, via etanercept administration, attenuated TVX/LPS-
induced liver injury (Fig. 2.7 and 2.8). Additionally, p55" and p75"' mice were
protected from TVXILPS-induced liver injury (Fig. 4.1), suggesting that both TNF

receptors, play a role. To explore further the role of TNFa in the pathogenesis, a

nontoxic dose of TNFa was administered in place of LPS. TVX/TNFa-

261

coexposure caused hepatotoxicity (Fig. 4.6), suggesting that TVX acts, at least in
part, at a point downstream of TNFa production.

TNFa has the potential to be involved in TVXILPS-induced hepatotoxicity
in several ways. To explore possible interactions with PMNs, IFNy, VEGF and/or
the hemostatic system, mice were killed at a time near the onset of liver injury.
TNFa neutralization attenuated TVXILPS-induced increases in the plasma
concentrations of lFNy, IL-6, MCP-1, VEGF, MIP-2, KC and MlP-1a (Fig. 4.2 and
4.3). However, despite a reduction in plasma concentrations of chemokines,
hepatic PMN accumulation induced by TVX/LPS coexposure was unaffected by
TNFot neutralization (Fig. 4.4). Dysregulation of the hemostatic system by
TVX/LPS coexposure was also lessened by etanercept treatment. The TVX/LPS-
induced increase in the plasma concentration of PAl-1 was attenuated by TNFa
neutralization (Fig. 4.5A), whereas the plasma concentration of TAT dimers was
unaffected (Fig. 4.53). However, TNFa neutralization reduced TVXILPS-induced
hepatic fibrin deposition (Fig. 4.50). Overall, the TVXILPS-induced increases in
plasma concentrations of lFNy and VEGF and dysregulation of the hemostatic
system were TNFa-dependent.

Hepatic gene expression analysis at a time before the onset of liver injury
distinguished TVXILPS-treated mice from those treated with TVX or LPS alone
(Fig. 5.2). Further analysis of genes selectively changed by TVX/LPS compared
to all other treatment groups suggested a role for interferon signaling, inasmuch
as 18% of the genes were increased and regulated by interferons (Table 5.2).

Therefore, the hypothesis was tested that IFNy plays a role in the progression of

262

liver injury induced by TVX/LPS coexposure. TVX enhanced the LPS-induced
increase in the plasma concentrations of lFNy and lL-18, which stimulates cells to
release lFNy (Fig. 5.4). Both |L-18"' and |FNy"' mice were protected from
TVXILPS-induced hepatotoxicity (Fig. 5.5 and 5.6). These results suggest that
IFNy is a critical mediator involved in the pathogenesis.

The role which IFNy plays in the TVXILPS-induced increase of plasma
concentrations of TNFa, VEGF and hepatic neutrophil accumulation were
explored. After TVXILPS coexposure, IFNy"' mice had reduced plasma
concentrations of TNFa, lL-18 and lL-1B, whereas VEGF was unchanged
compared to wild-type mice (Fig. 5.9). TVXILPS-induced hepatic neutrophil
accumulation was unchanged in IFNy"' mice (Fig. 5.7). These results suggest
that TVXILPS-induced increases in plasma VEGF and hepatic neutrophil
accumulation are independent of lFNy, but IFNy plays a role in the increased
plasma concentrations of TNFa and lL-18.

Hemostatic system dysregulation can result in hepatic fibrin deposition
and has the potential to be involved in the progression of liver injury through
several mechanisms. Therefore, the hypothesis tested was that an imbalance in
the hemostatic system plays a role in the pathogenesis of TVXILPS-induced liver
injury. At the onset of liver injury, coexposure to TVX/LPS, but not exposure to
TVX, LVX, LPS or LVX/LPS caused coagulation system activation as measured
by increased plasma concentration of thrombin-antithrombin dimers and
decreased plasma circulating fibrinogen (Fig. 6.1). LPS treatment induced a

small increase in plasma PAl-1 concentration, and TVX pretreatment enhanced

263

this effect (Fig. 6.2). This imbalance in coagulation and fibrinolysis induced by
TVX/LPS coexposure was associated with hepatic fibrin deposition (Fig. 6.3).
Anticoagulant heparin administration reduced TVXILPS-induced hepatic fibrin
deposition and liver injury (Fig. 6.4). PAl—1"' mice treated with TVX/LPS exhibited
similar fibrin deposition to wild-type mice yet had less hepatocellular injury (Fig.
6.6). These results suggest that TVX/LPS coexposure caused an imbalance in
the hemostatic system, resulting in increased thrombin activation, plasma
concentrations of PAH and hepatic fibrin deposition. Furthermore, both thrombin
activation and PAI-1 play a critical role in the progression of TVXILPS-induced
liver injury, but through different modes of action.

To examine the mechanisms by which thrombin inhibition and PAl-1"'
mice are protected from TVXILPS-induced liver injury, plasma concentrations of
several cytokines were measured at a time near the onset of liver injury. After
TVX/LPS coexposure, PAl-1”' mice had reduced plasma concentrations of
several cytokines including IFNy, but not TNFa or VEGF compared to wild-type
controls (Fig. 6.7). In contrast, heparin treatment did not attenuate the TVX/LPS-
induced increase in the plasma concentrations of any cytokines measured (Fig.
6.7). These results suggest that PAl-1, but not thrombin, has a proinflammatory
role during TVX/LPS coexposure.

VEGF is a cytokine with proinflammatory properties and has the potential
to participate in several pathways involved in the pathogenesis. TVX prolonged
the LPS-induced increase in plasma concentration of VEGF (Fig. 7.1). VEGF

neutralization protected mice from TVXILPS-induced liver injury, as reflected by

264

both plasma ALT activity and liver histopathology (Fig. 7.2 and 7.3). VEGF
neutralization did not affect the TVXILPS-induced increases in plasma
concentrations of lL-1p, lL-6, IL-18, lFNy, MCP-1, KC or MIP-1a (Fig. 7.4 and
7.5). However, it reduced the increase in plasma TNFa induced by TVX/LPS
coexposure (Fig. 7.4). Hepatic neutrophil accumulation was unaffected by VEGF
neutralization (Fig. 7.6), but it caused a trend toward a reduction in TVX/LPS-
induced plasma concentrations of TAT dimers and PAl-1 (Fig. 7.7).This was
associated with significantly decreased hepatic fibrin deposition (Fig. 7.7). These
results suggest that VEGF plays a critical role in the progression of TVX/LPS-
induced liver injury, perhaps by promoting increased TNFa and hepatic fibrin
deposition.

In summary, TVX interacts with an inflammatory stress to cause liver
injury in mice which is dependent on PMN activation, TNFa, lFNy, thrombin, PAI-
1 and VEGF. Furthermore, TNFa, lFNy and VEGF interact with one another to
create possible cycles of inflammation. These proinflammatory cycles and the

proposed pathways of TVXILPS-induced liver injury are outlined in Section 8.2.

265

8.2 Proposed pathways of TVXILPS-induced liver injury

Figure 8.2 illustrates proposed pathways to TVXILPS-induced liver injury.
These pathways are based on the results presented in this dissertation and are
likely to be modified as we learn more about this model. LPS caused coagulation
system activation, hepatic PMN accumulation and increased the plasma
concentration of TNFa. TVX treatment enhanced the LPS-induced increase in
coagulation system activation (Fig. 6.1) and TNFa (Fig. 2.4).

LPS-induced hepatic neutrophil accumulation was unaffected by TVX.
Additionally, PMN accumulation was not reduced by neutralizing or removing
TNFa (Fig. 4.4), IFNy (Fig. 5.7) or VEGF (Fig. 7.6). Neutrophil activation is critical
to TVXILPS-induced liver injury (Fig. 3.6), but is not required for the induction of
cytokines by TVXILPS coexposure (Fig. 3.9). Therefore, hepatic PMN
accumulation and activation appear to be a mechanism required for TVX/LPS-
induced hepatotoxicity which does not interact with any other pathways explored.
However, this interpretation is subject to change, as some of the cytokines
involved in TVXILPS-induced liver injury might affect PMN activation.

Coagulation system activation induced by LPS was enhanced by TVX (Fig.
6.1); and along with TNFa (Fig. 4.5) and VEGF (Fig. 7.7) played a critical role in
TVXILPS-induced hepatic fibrin deposition (Fig. 6.3). However, TNFa and VEGF
contribute to hepatic fibrin deposition by a mechanism independent of
coagulation system activation, since neutralization of either did not affect plasma

concentrations of TAT dimers (Fig. 4.5 and 7.7). However, coagulation system

266

Fig. 8.1. Proposed pathways to TVXILPS-induced liver injury. See Section

8.2 for a detailed explanation.

267

\. %./a.

.....muflmee E

......w ... i/.././rv....\\

Ems-“fines
2.-.: I 6.5: I u0w> :35... :93

 

 

 

_mn_.__/ /xH\

activation did not increase plasma concentrations of any of the cytokines
examined (Fig. 6.7). Thrombin activation was critical to the progression of
TVXILPS-induced liver injury (Fig. 6.3), likely at least in part, via hepatic fibrin
deposition.

TVX enhanced the LPS-induced increase in the plasma concentration of
TNFa, which appears to be a central player in TVXILPS-induced liver injury.
TVXILPS-induced increases in the plasma concentrations of VEGF (Fig. 4.2),
chemokines (Fig. 4.3), PAl—1 (Fig. 4.5), lL-18 (data not shown) and lFNy (Fig.
4.2) were TNFa-dependent. Due to their TNFa-dependence and because TNFa
was enhanced before these factors, they are likely downstream of TNFa.
Furthermore, TNFa and lFNy could have directly caused hepatocellular death
which might be enhanced by TVX.

Likely proinflammatory cycles were identified, inasmuch as several factors
were found which upregulate one another. The removal or inhibition of VEGF
(Fig. 7.4), PAI-1 (Fig. 6.7) or IFNy (Fig. 5.8) attenuated the TVXILPS-induced
increase in plasma concentration of TNFa. Conversely, TNFor neutralization
reduced the appearance of each of these. Similarly, lFNy"' mice had a reduced
plasma concentration of lL-18 (Fig. 5.8) following TVXILPS coexposure, and IL-
18 is a known inducer of IFNy expression. PAl-1 also had other proinflammatory
properties beyond upregulating TNFa, such as increasing plasma concentrations
of chemokines and lFNy (Fig. 6.7) after TVX/LPS administration. These vicious

proinflammatory cycles are summarized in Fig. 8.2. Several of these properties

269

found were novel and the inflammatory cycles of upregulation have the potential

to be involved in the pathogenesis.

270

Fig. 8.2. Possible proinflammatory cycles induced by TVXILPS coexposure.
Based on the results obtained, there exist several possible loops of uncontrolled
upregulation. These vicious cycles of inflammation could be involved in

TVXILPS-induced liver injury. See Section 8.2 for additional information.

271

 

 

\

 

> lL-18
'IFNy

 

 

 

 

TNFa
11

r
PAl-1 '—

 

 

J

VEGF

272

Section 8.3 Major findings and implications

1. TVX, but not LVX interacts with an inflammatory stress to cause liver
injury in mice. This observation had been reported in rats (58). However,
the development of hepatotoxic TVX-inflammation interaction in mice
demonstrates that the phenomenon is not species-specific and might be
extrapolated to TVX lADRs in humans. Furthermore, the degree of
TVXILPS-induced liver injury was much greater in mice (Fig. 2.1)
compared to rats (58). Both moderate and severe hepatotoxic responses
have been reported in people who took TVX (206). The robustness of the
murine model of liver injury resembles the severe hepatotoxicity caused
by TVX in humans more so than the rat model. Additionally, that TVX, but
not LVX interacts with an inflammatory stress to cause hepatotoxicity
suggests that this model could distinguish between fluroquinolones based
on their propensity to cause lADRs. This suggests that the
drug/inflammation animal model could potentially be used as a preclinical

tool to help select a drug candidate based on its potential lADR liability.

2. TVX interacted with an inflammatory stress induced by either gram-
negative or gram-positive stimuli. That TVX interacted with either TLR2- or
TLR4-activating ligands to cause liver injury proves that
TVX/inflammation-induced liver injury is not specific to TLR4 activation.

Indeed, it suggests that TVX interacts with an inflammatory stress,

273

regardless of its source, to precipitate liver injury. The result suggests that
inflammatory stress induced by either gram-positive or gram-negative
bacteria might play a role in TVX hepatotoxicity. This is of particular
importance inasmuch as TVX was approved for treating both gram—

positive and gram-negative bacterial infections by the FDA.

. TVXILPS-coexposure resulted in unique gene expression changes in mice
prior to the onset of liver injury. These results suggest that global gene
expression change is an earlier marker of liver toxicity than plasma ALT
activity in this model. It is possible that gene expression analysis could be
used to identify drugs with lADR liability when coadministered with LPS,
even if the coexposure does not result in hepatotoxicity. It is possible that
in the drugflnflammation model gene expression analysis offers a more
sensitive endpoint to filter out than liver injury to identify drug candidates

with lADR liability.

. TVXILPS-induced liver injury was dependent on PMN activation, TNFa,
IFNy, thrombin activation, PAH and VEGF. It is possible that these

mediators of inflammatory liver injury are involved in the pathogenesis of

TVX-induced hepatotoxicity in people.

274

5. TVXILPS-induced liver injury is dependent on both TNF receptors. The
p55 receptor has been well studied, but the role of the p75 receptor in
inflammatory liver injury is unclear. The p75 receptor is not involved in
some models of liver injury that are dependent on TNFa (96, 240) but is
involved in others (234-236). Indeed, p754' mice were protected to a
greater degree than p55”' mice from TVXILPS-induced liver injury.
Therefore, TVX/LPS coexposure would be an ideal model to use for

further studies examining the role of the p75 receptor in drug-induced liver

injury.

6. The comprehensive studies exploring the roles of TNFa, IFNy, PAH and
VEGF following TVX/LPS coexposure led to the finding that several
proinflammatory cycles appear to be involved in the pathogenesis. It is
possible that they result in an unregulated inflammatory response resulting
in host tissue injury. These proinflammatory cycles could be involved in
numerous models of inflammatory tissue injury and might not be specific

to the liver or to TVX/LPS coexposure.

275

Section 8.4 Knowledge gaps and future studies

Several important findings were identified by this research including the
development of an animal model of TVX-induced hepatotoxicity. It is possible
that the drug/inflammation model could be used preclinically to identify drugs with
the propensity to cause lADRs. Further studies with additional fluoroquinolones
would provide more positive and negative comparators and would add merit to
the model. Additionally, extensive studies with other drugs linked with lADRs in
people would be needed to validate the model. It is essential to determine the
rate of false negatives and false positives identified by the model before any
widescale use in drug development.

The findings presented elucidated some of the mechanisms involved in
TVXILPS-induced liver injury. However, the mechanism by which TVX interacts
with TLR2 and TLR4 activation to cause liver injury is still unknown. Future
studies will examine signaling pathways initiated by TLR activation to determine if
TVX pretreatment alters the initial signaling pathways. The identification of
specific pathways might lead to the finding of a specific enzyme or adapter
protein which TVX interacts with.

Whether TVX affects hepatocytes directly was not examined. It is possible
that TVX directly sensitizes to insults such as PMN proteases, hypoxia, TNFa or
IFNy. A direct sensitization of hepatocytes by TVX might account for the

extensive liver injury caused by TVX/LPS coexposure.

276

TVX treatment prior to a nonhepatotoxic dose of recombinant murine
TNFa resulted in significant liver injury. Other studies have shown that TNFa by
itself does not cause liver injury in mice but can when administered with
galactosamine or a DNA synthesis inhibitor (249, 250). Indeed, the hepatocellular
lesions in TVXfI'NFa-treated mice appear similar to those seen after
galactosamine/TNFa coexposure, suggesting commonalities in mechanisms
(251). This raises the possibility that TVX affects DNA synthesis of eukaryotic
cells, rendering them sensitive to inflammatory stress. Additional studies need to
be done to determine if TVX interacts with eukaryotic cells to affect DNA
synthesis to a greater degree than other fluoroquinolones. This could provide
understanding of the mechanism by which TVX causes liver injury in people,
whereas other fluoroquinolones do not.

Another focus which requires further research is examination of specific
cell types. Neutrophil activation was not measured. Furthermore, it is unknown
whether any of the cytokines involved in the pathogenesis affect PMN activation.
It is possible that they cause liver injury, at least in part, by enhancing PMN
activation. Additionally, it is unknown whether cell types other than PMNs are
involved in TVXILPS-induced hepatotoxicity. It is possible that T cells, NK cells,
platelets and Kupffer cells are involved in the pathogenesis, and requires further

examination.

277

10.

11.

12.

13.

Reference List

. Pirrnohamed M, James S, Meakin S, Green C, Scott AK, Walley TJ, et al.

Adverse drug reactions as cause of admission to hospital: prospective
analysis of 18 820 patients. BMJ 2004 Jul 3;329(7456):15-19.

Lasser KE, Allen PD, Woolhandler SJ, Himmelstein DU, Wolfe SM, Bor
DH. Timing of new black box warnings and withdrawals for prescription
medications. JAMA 2002 May 1;287(17):2215-2220.

Michelson S, Joho K. Drug discovery, drug development and the emerging
world of pharmacogenomics: prospecting for information in a data-rich
landscape. Curr Opin Mol Ther 2000 Dec;2(6):651-654.

Sibbald B. Cisapride, before and after: still waiting for ADE-reporting
reform. CMAJ 2001 Nov 13;165(10):1370.

Kaitin Kl. Drug Safety Withdrawals in the US. not linked to speed of FDA
approval. Tufts Center for the Study of Drug Development Impact Report
2005 2005 Sep;7(5).

Ostapowicz G, Fontana RJ, Schiodt FV, Larson A, Davern TJ, Han SH, et
al. Results of a prospective study of acute liver failure at 17 tertiary care
centers in the United States. Ann Intern Med 2002 Dec 17;137(12):947-
954.

Hunter EB, Johnston PE, Tanner G, Pinson CW, Awad JA. Bromfenac
(Duract)-associated hepatic failure requiring liver transplantation. Am J
Gastroenterol 1999 Aug;94(8):2299-2301.

Graham DJ, Green L, Senior JR, Nourjah P. Troglitazone-induced liver
failure: a case study. Am J Med 2003 Mar;114(4):299-306.

Murphy EJ, Davern TJ, Shakil AO, Shick L, Masharani U, Chow H, et al.
Troglitazone-induced fulminant hepatic failure. Acute Liver Failure Study
Group. Dig Dis Sci 2000 Mar;45(3):549-553.

Bemuau J, Mallet L, Benhamou JP. [Hepatotoxicity due to tienilic acid].
Gastroenterol Clin Biol 1981 Jun;5(6-7):692-693.

Lazarczyk DA, Goldstein NS, Gordon SC. Trovafloxacin hepatotoxicity.
Dig Dis Sci 2001 Apr;46(4):925-926.

Nicholson SC, Webb CD, Moellering RC, Jr. Antimicrobial-associated
acute hepatitis. Pharrnacotherapy 2002 Jun;22(6):794-796.

Choi S. Nefazodone (Serzone) withdrawn because of hepatotoxicity.
CMAJ 2003 Nov 25;169(11):1187.

278

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Maniar JK, Shah SR, Verma R, Kamath R, Gupte P, Maniar A.
Nevirapine—induced fulminant hepatitis. J Assoc Physicians India 2006
Dec;54:957-958.

Sanne I, Mommeja-Marin H, Hinkle J, Bartlett JA, Lederman MM,
Maartens G, et al. Severe hepatotoxicity associated with nevirapine use in
HIV-infected subjects. Jlnfect Dis 2005 Mar 15;191(6):825-829.

Temple RJ, Himmel MH. Safety of newly approved drugs: implications for
prescribing. JAMA 2002 May 1;287(17):2273-2275.

Hussaini SH, Farrington EA. Idiosyncratic drug-induced liver injury: an
overview. Expert Opin Drug Saf 2007 Nov;6(6):673-684.

Lee WM. Drug-induced hepatotoxicity. N Engl J Med 2003 Jul
31 ;349(5):474-485.

Kaplowitz N, Aw TY, Simon FR, Stolz A. Drug-induced hepatotoxicity. Ann
Intern Med 1986 Jun;104(6):826-839.

Walgren JL, Mitchell MD, Thompson DC. Role of metabolism in drug-
induced idiosyncratic hepatotoxicity. Crit Rev Toxicol 2005 Apr;35(4):325-
361.

Kassahun K, Pearson PG, Tang W, McIntosh I, Leung K, Elmore C, et al.
Studies on the metabolism of troglitazone to reactive intermediates in vitro
and in vivo. Evidence for novel biotransforrnation pathways involving
quinone methide formation and thiazolidinedione ring scission. Chem Res
Toxicol 2001 Jan;14(1):62-70.

He K, Talaat RE, Pool WF, Reily MD, Reed JE, Bridges AJ, et al.
Metabolic activation of troglitazone: identification of a reactive metabolite
and mechanisms involved. Drug Metab Dispos 2004 Jun;32(6):639—646.

Uetrecht J. Evaluation of which reactive metabolite, if any, is responsible
for a specific idiosyncratic reaction. Drug Metab Rev 2006;38(4):745-753.

Williams DP, Park BK. Idiosyncratic toxicity: the role of toxicophores and
bioactivation. Drug Discov Today 2003 Nov 15;8(22):1044-1050.

Mitchell JR, Zimmerman HJ, lshak KG, Thorgeirsson UP, Timbrell JA,
Snodgrass WR, et al. Isoniazid liver injury: clinical spectrum, pathology,
and probable pathogenesis. Ann Intern Med 1976 Feb;84(2):181-192.

Huang YS. Genetic polymorphisms of drug-metabolizing enzymes and the

susceptibility to antituberculosis drug-induced liver injury. Expert Opin
Drug Metab Toxicol 2007 Feb;3(1):1-8.

279

27.

28.

29.

30.

31.

32.

33.

35.

36.

37.

38.

Gurumurthy P, Krishnamurthy MS, Nazareth O, Parthasarathy R, Sarrna
GR, Somasundaram PR, et al. Lack of relationship between hepatic
toxicity and acetylator phenotype in three thousand South Indian patients
during treatment with isoniazid for tuberculosis. Am Rev Respir Dis 1984
Jan;129(1):58-61.

Aithal GP, Ramsay L, Daly AK, Sonchit N, Leathart JB, Alexander G, et al.
Hepatic adducts, circulating antibodies, and cytokine polymorphisms in
patients with diclofenac hepatotoxicity. Hepatology 2004 May;39(5):1430-
1440.

Uetrecht J. Screening for the potential of a drug candidate to cause
idiosyncratic drug reactions. Drug Discov Today 2003 Sep 15;8(18):832-
837.

Park BK, Coleman JW, Kitteringham NR. Drug disposition and drug
hypersensitivity. Biochem Pharrnacol 1987 Mar 1;36(5):581-590.

Robin MA, Le RM, Descatoire V, Pessayre D. Plasma membrane
cytochromes P450 as neoantigens and autoimmune targets in drug-
induced hepatitis. J Hepatol 1997;26 Suppl 1:23-30.

Obermayer-Straub P, Strassburg CP, Manns MP. Autoimmune hepatitis. J
Hepatol 2000;32(1 Suppl):181-197.

Kitteringham NR, Kenna JG, Park BK. Detection of autoantibodies
directed against human hepatic endoplasmic reticulum in sera from
patients with halothane-associated hepatitis. Br J Clin Phamnacol 1995
Oct;40(4):379-386.

Park BK, Pirmohamed M, Kitteringham NR. Role of drug disposition in
drug hypersensitivity: a chemical, molecular, and clinical perspective.
Chem Res Toxicol 1998 Sep;11(9):969—988.

Liu ZX, Kaplowitz N. Immune-mediated drug-induced liver disease. Clin
Liver Dis 2002 Aug;6(3):755-774.

Kaplowitz N. Idiosyncratic drug hepatotoxicity. Nat Rev Drug Discov 2005
Jun;4(6):489-499.

Uetrecht JP. New concepts in immunology relevant to idiosyncratic drug
reactions: the "danger hypothesis" and innate immune system. Chem Res
Toxicol 1999 May; 1 2(5):387-395.

Seguin B, Uetrecht J. The danger hypothesis applied to idiosyncratic drug
reactions. Curr Opin Allergy Clin Immunol 2003 Aug;3(4):235-242.

280

39.

40.

4

A

42.

43.

44.

45.

46.

47.

48.

49.

50.

Pichler WJ. Pharmacological interaction of drugs with antigen-specific
immune receptors: the p-i concept. Curr Opin Allergy Clin Immunol 2002
Aug;2(4):301-305.

Engler OB, Strasser l, Naisbitt DJ, Cerny A, Pichler WJ. A chemically inert
drug can stimulate T cells in vitro by their T cell receptor in non-sensitised
individuals. Toxicology 2004 Apr 1;197(1):47-56.

. Chinnery PF, Johnson MA, Wardell TM, Singh-Kler R, Hayes C, Brown DT,

et al. The epidemiology of pathogenic mitochondrial DNA mutations. Ann
Neurol 2000 Aug;48(2):188—193.

Waring JF, Anderson MG. Idiosyncratic toxicity: mechanistic insights
gained from analysis of prior compounds. Curr Opin Drug Discov Devel
2005 Jan;8(1):59-65.

Boelsterli UA, Lim PL. Mitochondrial abnormalities-a link to idiosyncratic
drug hepatotoxicity? Toxicol Appl Pharmacol 2007 Apr 1;220(1):92-107.

Haasio K, Koponen A, Penttila KE, Nissinen E. Effects of entacapone and
tolcapone on mitochondrial membrane potential. Eur J Pharmacol 2002
Oct 18;453(1):21-26.

Keller BJ, Yamanaka H, Thurman RG. Inhibition of mitochondrial
respiration and oxygen-dependent hepatotoxicity by six structurally
dissimilar peroxisomal proliferating agents. Toxicology 1992;71(1-2):49—61.

Petrescu l, Tarba C. Uncoupling effects of diclofenac and aspirin in the
perfused liver and isolated hepatic mitochondria of rat. Biochim Biophys
Acta 1997 Feb 15;1318(3):385-394.

Schwab CE, Tuschl H. In vitro studies on the toxicity of isoniazid in
different cell lines. Hum Exp Toxicol 2003 Nov;22(11):607-615.

Shishido S, Koga H, Harada M, Kumemura H, Hanada S, Taniguchi E, et
al. Hydrogen peroxide overproduction in megamitochondria of
troglitazone-treated human hepatocytes. Hepatology 2003 Jan;37(1):136-
147.

Boelsterli UA. Diclofenac-induced liver injury: a paradigm of idiosyncratic
drug toxicity. Toxicol Appl Pharmacol 2003 Nov 1;192(3):307-322.

Tirmenstein MA, Hu CX, Gales TL, Maleeff BE, Narayanan PK, Kurali E,
et al. Effects of troglitazone on HepGZ viability and mitochondrial function.
Toxicol Sci 2002 Sep;69(1):131-138.

281

51.

52.

53.

54.

55.

56.

57

58.

59.

60.

61.

Ong MM, Latchoumycandane C, Boelsterli UA. Troglitazone-induced
hepatic necrosis in an animal model of silent genetic mitochondrial
abnormalities. Toxicol Sci 2007 May;97(1):205-213.

Watkins PB. Idiosyncratic liver injury: challenges and approaches. Toxicol
Pathol 2005;33(1):1-5.

Black M, Mitchell JR, Zimmerman HJ, lshak KG, Epler GR. Isoniazid-
associated hepatitis in 114 patients. Gastroenterology 1975
Aug;69(2):289-302.

Li AP. A review of the common properties of drugs with idiosyncratic
hepatotoxicity and the "multiple determinant hypothesis" for the
manifestation of idiosyncratic drug toxicity. Chem Biol Interact 2002 Nov
10;142(1-2):7-23.

Roth RA, Luyendyk JP, Maddox JF, Ganey PE. Inflammation and drug
idiosyncrasyuis there a connection? J Pharmacol Exp Ther 2003
Oct;307(1):1-8.

Buchweitz JP, Ganey PE, Bursian SJ, Roth RA. Underlying endotoxemia
augments toxic responses to chlorpromazine: is there a relationship to
drug idiosyncrasy? J Pharmacol Exp Ther 2002 Feb;300(2):460-467.

. Deng X, Stachlewitz RF, Liguori MJ, Blomme EA, Waring JF, Luyendyk JP,

et al. Modest inflammation enhances diclofenac hepatotoxicity in rats: role
of neutrophils and bacterial translocation. J Pharmacol Exp Ther 2006
Dec;319(3):1191-1199.

Waring JF, Liguori MJ, Luyendyk JP, Maddox JF, Ganey PE, Stachlewitz
RF, et al. Microarray analysis of lipopolysaccharide potentiation of
trovafloxacin-induced liver injury in rats suggests a role for
proinflammatory chemokines and neutrophils. J Pharmacol Exp Ther 2006
Mar;316(3):1080-1087.

Luyendyk JP, Maddox JF, Cosma GN, Ganey PE, Cockerell GL, Roth RA.
Ranitidine treatment during a modest inflammatory response precipitates
idiosyncrasy-like liver injury in rats. J Pharmacol Exp Ther 2003
Oct;307(1):9-16.

Zou W, Devi SS, Younis HS, Roth RA, Ganey PE. Inflammation-mediated
Liver Injury in Rats Cotreated with Sulindac and Lipopolysaccharide
[Abstract]. The Toxicologist 2008;102.

Hellman J, Roberts JD, Jr., Tehan MM, Allaire JE, Warren HS. Bacterial
peptidoglycan-associated lipoprotein is released into the bloodstream in

gram-negative sepsis and causes inflammation and death in mice. J Biol
Chem 2002 Apr 19;277(16):14274-14280.

282

62.

63

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

Zahringer U, Lindner B, Rietschel ET. Molecular structure of lipid A, the
endotoxic center of bacterial lipopolysaccharides. Adv Carbohydr Chem
Biochem 1994;50:21 1-276.

. Janssens S, Beyaert R. Role of Toll-like receptors in pathogen recognition.

Clin Microbiol Rev 2003 Oct;16(4):637-646.

Jerala R. Structural biology of the LPS recognition. Int J Med Microbiol
2007 Sep;297(5):353-363.

Kitchens RL, Thompson PA. Modulatory effects of sCD14 and LBP on
LPS-host cell interactions. J Endotoxin Res 2005;11(4):225-229.

Akira S, Takeda K. Toll-like receptor signalling. Nat Rev Immunol 2004
Ju|;4(7):499-511.

Cao Z, Henzel WJ, Gao X. IRAK: a kinase associated with the interleukin-
1 receptor. Science 1996 Feb 23;271(5252):1 128-1131.

Cao Z, Xiong J, Takeuchi M, Kurama T, Goeddel DV. TRAF6 is a signal
transducer for interleukin-1. Nature 1996 Oct 3;383(6599):443-446.

Ninomiya-Tsuji J, Kishimoto K, Hiyama A, lnoue J, Cao Z, Matsumoto K.
The kinase TAK1 can activate the NIK—I kappaB as well as the MAP
kinase cascade in the lL-1 signalling pathway. Nature 1999 Mar

1 8;398(6724):252-256.

Heumann D, Roger T. Initial responses to endotoxins and Gram-negative
bacteria. Clin Chim Acta 2002 Sep;323(1-2):59-72.

Yamamoto M, Sato S, Hemmi H, Hoshino K, Kaisho T, Sanjo H, et al.
Role of adaptor TRIF in the MyD88-independent toll-like receptor signaling
pathway. Science 2003 Aug 1;301(5633):640-643.

Schafer SL, Lin R, Moore PA, Hiscott J, Pitha PM. Regulation of type |
interferon gene expression by interferon regulatory factor-3. J Biol Chem
1998 Jan 30;273(5):2714-2720.

Ganey PE, Luyendyk JP, Maddox JF, Roth RA. Adverse hepatic drug
reactions: inflammatory episodes as consequence and contributor. Chem
Biol Interact 2004 Nov 1;150(1):35-51.

Kaufrnann SH. Bacterial and protozoal infections in genetically disrupted
mice. Curr Opin Immunol 1994 Aug;6(4):518-525.

Tracey KJ, Cerami A. Tumor necrosis factor, other cytokines and disease.
Annu Rev Cell Biol 1993;9:317-343.

283

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

Colletti LM, Kunkel SL, Walz A, Burdick MD, Kunkel RG, Wilke CA, et al.
The role of cytokine networks in the local liver injury following hepatic
ischemialreperfusion in the rat. Hepatology 1996 Mar;23(3):506-514.

Ishida Y, Kondo T, Tsuneyama K, Lu P, Takayasu T, Mukaida N. The
pathogenic roles of tumor necrosis factor receptor p55 in acetaminophen-
induced liver injury in mice. J Leukoc Biol 2004 Jan;75(1):59-67.

Raftery G, Griffiths B, Kay L, Kane D. Chronic viral hepatitis and TNF-
alpha blockade. Rheumatology (Oxford) 2007 Aug;46(8):1381-1382.

Shimizu S, Yamada Y, Okuno M, Ohnishi H, Osawa Y, Seishima M, et al.
Liver injury induced by lipopolysaccharide is mediated by TNFR-1 but not
by TNFR-2 or Fas in mice. Hepatol Res 2005 Mar;31(3):136-142.

Wajant H, Pfizenmaier K, Scheurich P. Tumor necrosis factor signaling.
Cell Death Differ 2003 Jan;10(1):45-65.

Thirunavukkarasu C, Watkins SC, Gandhi CR. Mechanisms of endotoxin-
induced NO, lL-6, and TNF-alpha production in activated rat hepatic
stellate cells: role of p38 MAPK. Hepatology 2006 Aug;44(2):389—398.

Tracey KJ, Cerami A. Tumor necrosis factor: an updated review of its
biology. Crit Care Med 1993 Oct;21(10 Suppl):S415-S422.

Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor
superfamilies: integrating mammalian biology. Cell 2001 Feb
23; 1 04(4):487-501.

Tartaglia LA, Ayres TM, Wong GH, Goeddel DV. A novel domain within
the 55 kd TNF receptor signals cell death. Cell 1993 Sep 10;74(5):845-
853.

Murray J, Barbara JA, Dunkley SA, Lopez AF, Van O, X, Condliffe AM, et
al. Regulation of neutrophil apoptosis by tumor necrosis factor-alpha:
requirement for TNF R55 and TNFR75 for induction of apoptosis in vitro.
Blood 1997 Oct 1;90(7):2772-2783.

Dri P, Haas E, Cramer R, Menegazzi R, Gasparini C, Martinelli R, et al.
Role of the 75-kDa TNF receptor in TNF-induced activation of neutrophil
respiratory burst. J Immunol 1999 Jan 1;162(1):460-466.

Ohira H, Abe K, Yokokawa J, Takiguchi J, Rai T, Shishido S, et al.
Adhesion molecules and CXC chemokines in endotoxin-induced liver
injury. Fukushima J Med Sci 2003 Jun;49(1):1-13.

Hanlon AM, Jang S, Salgame P. Signaling from cytokine receptors that
affect Th1 responses. Front Biosci 2002 May 1;7:d1247-d1254.

284

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

Friedl J, Puhlmann M, Bartlett DL, Libutti SK, Turner EN, Gnant MF, et al.
Induction of permeability across endothelial cell monolayers by tumor
necrosis factor (TN F) occurs via a tissue factor-dependent mechanism:
relationship between the procoagulant and permeability effects of TNF.
Blood 2002 Aug 15;100(4):1334-1339.

Esmon CT. Possible involvement of cytokines in diffuse intravascular
coagulation and thrombosis. Baillieres Best Pract Res Clin Haematol 1999
Sep;12(3):343-359.

Perez-Ruiz M, Ros J, Morales-Ruiz M, Navasa M, Colmenero J, Ruiz-del-
Arbol L, et al. Vascular endothelial growth factor production in peritoneal
macrophages of cirrhotic patients: regulation by cytokines and bacterial
lipopolysaccharide. Hepatology 1999 Apr;29(4):1057-1063.

Mechtcheriakova D, Schabbauer G, Lucerna M, Clauss M, De MR, Binder
BR, et al. Specificity, diversity, and convergence in VEGF and TNF-alpha
signaling events leading to tissue factor up-regulation via EGR-1 in
endothelial cells. FASEB J 2001 Jan;15(1):230-242.

Grell M, Douni E, Wajant H, Lohden M, Clauss M, Maxeiner B, et al. The
transmembrane form of tumor necrosis factor is the prime activating ligand
of the 80 kDa tumor necrosis factor receptor. Cell 1995 Dec 1;83(5):793-
802.

Grell M, Wajant H, Zimmermann G, Scheurich P. The type 1 receptor
(CD120a) is the high-affinity receptor for soluble tumor necrosis factor.
Proc Natl Acad Sci U S A 1998 Jan 20;95(2):570-575.

Morio LA, Chiu H, Sprowles KA, Zhou P, Heck DE, Gordon MK, et al.
Distinct roles of tumor necrosis factor-alpha and nitric oxide in acute liver

injury induced by carbon tetrachloride in mice. Toxicol Appl Pharmacol
2001 Apr 1;172(1):44-51.

Peschon JJ, Torrance DS, Stocking KL, Glaccum MB, Otten C, Willis CR,
et al. TNF receptor-deficient mice reveal divergent roles for p55 and p75 in
several models of inflammation. J Immunol 1998 Jan 15;160(2):943-952.

Kusters S, Tiegs G, Alexopoulou L, Pasparakis M, Douni E, Kunstle G, et
al. In vivo evidence for a functional role of both tumor necrosis factor
(TNF) receptors and transmembrane TNF in experimental hepatitis. Eur J
Immunol 1997 Nov;27(11):2870-2875.

Schumann J, Bluethmann H, Tiegs G. Synergism of Pseudomonas

aeruginosa exotoxin A with endotoxin, superantigen, or TNF results in T.
Immunol Lett 2000 Oct 3;74(2):165-172.

285

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

Hayder H, Blanden RV, Korner H, Riminton DS, Sedgwick JD, Mullbacher
A. Adenovirus-induced liver pathology is mediated through TNF receptors
1 and II but is independent of TNF or Iymphotoxin. J Immunol 1999 Aug
1;163(3):1516-1520.

Jaeschke H, Smith CW. Mechanisms of neutrophil-induced parenchymal
cell injury. J Leukoc Biol 1997 Jun;61(6):647-653.

Hewett JA, Schultze AE, VanCise S, Roth RA. Neutrophil depletion
protects against liver injury from bacterial endotoxin. Lab Invest 1992
Mar;66(3):347-361 .

Jaeschke H, Farhood A, Bautista AP, Spolarics Z, Spitzer JJ, Smith CW.
Functional inactivation of neutrophils with a Mac-1 (CD11b/CD18)
monoclonal antibody protects against ischemia-reperfusion injury in rat
liver. Hepatology 1993 May;17(5):915-923.

Luyendyk JP, Shaw PJ, Green CD, Maddox JF, Ganey PE, Roth RA.
Coagulation-mediated hypoxia and neutrophil-dependent hepatic injury in
rats given lipopolysaccharide and ranitidine. J Pharmacol Exp Ther 2005
Sep;314(3):1023-1031.

Boehm U, Klamp T, Groot M, Howard JC. Cellular responses to interferon-
gamma. Annu Rev Immunol 1997;15:749-795.

Hasbold J, Hong JS, Kehry MR, Hodgkin PD. Integrating signals from IF N-
gamma and lL-4 by B cells: positive and negative effects on CD40 ligand-
induced proliferation, survival, and division-linked isotype switching to
lgG1, lgE, and IgG2a. J Immunol 1999 Oct 15;163(8):4175-4181.

Shtrichman R, Samuel CE. The role of gamma interferon in antimicrobial
immunity. Curr Opin Microbiol 2001 Jun;4(3):251-259.

Nakamura S, Otani T, ljiri Y, Motoda R, Kurimoto M, Orita K. IFN-gamma-
dependent and -independent mechanisms in adverse effects caused by
concomitant administration of lL-18 and lL-12. J Immunol 2000 Mar

15; 1 64(6):3330-3336.

Salkowski CA, Thomas KE, Cody MJ, Vogel SN. Impaired IFN-gamma
production in IFN regulatory factor-1 knockout mice during endotoxemia is
secondary to a loss of both lL-12 and lL-12 receptor expression. J
Immunol 2000 Oct 1;165(7):3970-3977.

Walker W, ste-Amezaga M, Kastelein RA, Trinchieri G, Hunter CA. lL-18
and C028 use distinct molecular mechanisms to enhance NK cell
production of lL-12-induced IFN-gamma. J Immunol 1999 May
15;162(10):5894-5901.

286

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

Bach EA, Aguet M, Schreiber RD. The IFN gamma receptor: a paradigm
for cytokine receptor signaling. Annu Rev Immunol 1997;15:563-591.

Farrar MA, Schreiber RD. The molecular cell biology of interferon-gamma
and its receptor. Annu Rev Immunol 1993;11:571-611.

Soh J, Donnelly RJ, Kotenko S, Mariano TM, Cook JR, Wang N, et al.
Identification and sequence of an accessory factor required for activation
of the human interferon gamma receptor. Cell 1994 Mar 11;76(5):793-802.

Bach EA, Tanner JW, Marsters S, Ashkenazi A, Aguet M, Shaw AS, et al.
Ligand-induced assembly and activation of the gamma interferon receptor
in intact cells. Mol Cell Biol 1996 Jun;16(6):3214-3221.

Briscoe J, Guschin D, Rogers NC, Watling D, Muller M, Horn F, et al.
JAKs, STATS and signal transduction in response to the interferons and
other cytokines. Philos Trans R Soc Lond B Biol Sci 1996 Feb
29;351(1336):167-171.

Jaruga B, Hong F, Kim WH, Gao B. lFN-gammalSTAT1 acts as a
proinflammatory signal in T cell-mediated hepatitis via induction of multiple
chemokines and adhesion molecules: a critical role of lRF-1. Am J Physiol
Gastrointest Liver Physiol 2004 Nov;287(5):G1044-G1052.

Tsung A, Stang MT, lkeda A, Critchlow ND, lzuishi K, Nakao A, et al. The
transcription factor interferon regulatory factor-1 mediates liver damage
during ischemia-reperfusion injury. Am J Physiol Gastrointest Liver Physiol
2006 Jun;290(6):G1261-G1268.

Cassatella MA, Cappelli R, Della B, V, Grzeskowiak M, Dusi S, Berton G.
Interferon-gamma activates human neutrophil oxygen metabolism and
exocytosis. Immunology 1988 Mar;63(3):499-506.

Ellis TN, Beaman BL. Interferon-gamma activation of polymorphonuclear
neutrophil function. Immunology 2004 May;112(1):2-12.

Tennenberg SD, Fey DE, Lieser MJ. Oxidative priming of neutrophils by
interferon-gamma. J Leukoc Biol 1993 Mar;53(3):301-308.

Meda L, Gasperini S, Ceska M, Cassatella MA. Modulation of
proinflammatory cytokine release from human polymorphonuclear
leukocytes by gamma interferon. Cell Immunol 1994 Sep;157(2):448-461.

Kawada N, Mizoguchi Y, Kobayashi K, Morisawa S, Monna T, Yamamoto

S. Interferon gamma modulates production of interleukin 1 and tumor
necrosis factor by murine Kupffer cells. Liver 1991 Feb;11(1):42-47.

287

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

Collins PW, Noble KE, Reittie JR, Hoffbrand AV, Pasi KJ, Yong KL.
Induction of tissue factor expression in human monocyte/endothelium
cocultures. Br J Haematol 1995 Dec;91(4):963-970.

Amman V, Stemme S, Rymo L, Risberg B. Interferon-gamma modulates
the fibrinolytic response in cultured human endothelial cells. Thromb Res
1995 Mar 1;77(5):431-440.

Malaguarnera L, lmbesi R, Di RM, Scuto A, Castrogiovanni P, Messina A,
et al. Action of prolactin, IFN-gamma, TNF-alpha and LPS on heme
oxygenase-1 expression and VEGF release in human
monocytes/macrophages. Int lmmunopharrnacol 2005 Aug;5(9):1458—
1469.

Ray PS, Fox PL. A post-transcriptional pathway represses monocyte
VEGF-A expression and angiogenic activity. EMBO J 2007 Jul
25;26(14):3360-3372.

Greenlund AC, Farrar MA, Viviano BL, Schreiber RD. Ligand-induced IFN
gamma receptor tyrosine phosphorylation couples the receptor to its
signal transduction system (p91). EMBO J 1994 Apr 1;13(7):1591-1600.

Darnell JE, Jr., Kerr IM, Stark GR. Jak-STAT pathways and transcriptional
activation in response to IFNs and other extracellular signaling proteins.
Science 1994 Jun 3;264(5164):1415-1421.

Hashimoto W, Takeda K, Anzai R, Ogasawara K, Sakihara H, Sugiura K,
et al. Cytotoxic NK1.1 Ag+ alpha beta T cells with intermediate TCR
induced in the liver of mice by lL-12. J Immunol 1995 May 1;154(9):4333-
4340.

Liu ZX, Govindarajan S, Okamoto S, Dennert G. NK cells cause liver
injury and facilitate the induction of T cell-mediated immunity to a viral liver
infection. J Immunol 2000 Jun 15;164(12):6480-6486.

Wiltrout RH. Regulation and antimetastatic functions of liver-associated
natural killer cells. Immunol Rev 2000 Apr;174:63-76.

Lee HJ, Oh YK, Rhee M, Lim JY, Hwang JY, Park YS, et al. The role of
STAT1/lRF-1 on synergistic ROS production and loss of mitochondrial

transmembrane potential during hepatic cell death induced by LPS/d-GalN.
J Mol Biol 2007 Jun 15;369(4):967-984.

Liu ZX, Govindarajan S, Kaplowitz N. Innate immune system plays a

critical role in determining the progression and severity of acetaminophen
hepatotoxicity. Gastroenterology 2004 Dec; 127(6):1 760-1774.

288

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

Gwag BJ, Lee YA, Ko SY, Lee MJ, Im DS, Yun BS, et al. Marked
prevention of ischemic brain injury by Neu2000, an NMDA antagonist and
antioxidant derived from aspirin and sulfasalazine. J Cereb Blood Flow
Metab 2007 Jun;27(6):1142-1151.

Terada Y, Inoshita S, Kuwana H, Kobayashi T, Okado T, Ichijo H, et al.
Important role of apoptosis signal-regulating kinase 1 in ischemic acute
kidney injury. Biochem Biophys Res Commun 2007 Dec 28;364(4):1043-
1049.

Sidelmann JJ, Gram J, Jespersen J, Kluft C. Fibrin clot formation and
lysis: basic mechanisms. Semin Thromb Hemost 2000;26(6):605-618.

Baglia FA, Walsh PN. Thrombin-mediated feedback activation of factor XI
on the activated platelet surface is preferred over contact activation by
factor Xlla or factor Xla. J Biol Chem 2000 Jul 7;275(27):20514-20519.

Nesheim ME, Mann KG. Thrombin-catalyzed activation of single chain
bovine factor V. J Biol Chem 1979 Feb 25;254(4):1326-1334.

Pieters J, Lindhout T, Hemker HC. In situ-generated thrombin is the only
enzyme that effectively activates factor VIII and factor V in thromboplastin-
activated plasma. Blood 1989 Aug 15;74(3):1021-1024.

Coughlin SR. Sol Sherry lecture in thrombosis: how thrombin 'talks' to
cells: molecular mechanisms and roles in vivo. Arterioscler Thromb Vasc
Biol 1998 Apr;18(4):514-518.

Kahn ML, Nakanishi-Matsui M, Shapiro MJ, lshihara H, Coughlin SR.
Protease-activated receptors 1 and 4 mediate activation of human
platelets by thrombin. J Clin Invest 1999 Mar;103(6):879-887.

Kinlough-Rathbone RL, Perry DW, Guccione MA, Rand ML, Packham MA.
Degranulation of human platelets by the thrombin receptor peptide
SFLLRN: comparison with degranulation by thrombin. Thromb Haemost
1993 Dec 20;70(6):1019-1023.

Naldini A, Sower L, Bocci V, Meyers B, Carney DH. Thrombin receptor
expression and responsiveness of human monocytic cells to thrombin is
linked to interferon-induced cellular differentiation. J Cell Physiol 1998
Oct;177(1):76-84.

Umarova BA, Dugina TN, Shestakova EV, Gluza E, Strukova SM.
Activation of rat mast cells upon stimulation of protease-activated receptor
(PAR-1). Bull Exp Biol Med 2000 Apr;129(4):314-317.

Hayashi T. [Tissue-type plasminogen activator (t-PA) and plasminogen
activator inhibitor (PAI)]. Rinsho Byori 1994 Apr;42(4):346-351.

289

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

Hjortland GO, Lillehammer T, Somme S, Wang J, Halvorsen T, Juell S, et
al. Plasminogen activator inhibitor-1 increases the expression of VEGF in
human glioma cells. Exp Cell Res 2004 Mar 10;294(1):130-139.

Kwak SH, Wang XQ, He Q, Fang WF, Mitra S, Bdeir K, et al. Plasminogen
activator inhibitor-1 potentiates LPS-induced neutrophil activation through
a JNK-mediated pathway. Thromb Haemost 2006 May;95(5):829-835.

Renckens R, Roelofs JJ, de W, V, Florquin S, Lijnen HR, Carmeliet P, et
al. The role of plasminogen activator inhibitor type 1 in the inflammatory

response to local tissue injury. J Thromb Haemost 2005 May;3(5):1018-

1025.

Laurens M, Defamie V, Scozzari G, Schmid-Alliana A, Gugenheim J,
Crenesse D. Hypoxia-reoxygenation-induced chemokine transcription is
not prevented by preconditioning or intermittent hypoxia, in mice
hepatocytes. Transpl Int 2005 Apr;18(4):444-452.

Ertel W, Morrison MH, Ayala A, Chaudry IH. Hypoxemia in the absence of
blood loss or significant hypotension causes inflammatory cytokine
release. Am J Physiol 1995 Jul;269(1 Pt 2):R160-R166.

Murata Y, Ohteki T, Koyasu S, Hamuro J. IFN—gamma and pro-
inflammatory cytokine production by antigen-presenting cells is dictated by
intracellular thiol redox status regulated by oxygen tension. Eur J Immunol
2002 Oct;32(10):2866-2873.

Forsythe JA, Jiang BH, lyer NV, Agani F, Leung SW, Koos RD, et al.
Activation of vascular endothelial growth factor gene transcription by
hypoxia-inducible factor 1. Mol Cell Biol 1996 Sep;16(9):4604-4613.

Ganey PE, Luyendyk JP, Newport SW, Eagle TM, Maddox JF, Mackman
N, et al. Role of the coagulation system in acetaminophen-induced
hepatotoxicity in mice. Hepatology 2007 Oct;46(4):1177-1186.

Harada N, Okajima K, Uchiba M. Dalteparin, a low molecular weight
heparin, attenuates inflammatory responses and reduces ischemia-
reperfusion-induced liver injury in rats. Crit Care Med 2006 Jul;34(7):1883-
1891.

Hewett JA, Roth RA. The coagulation system, but not circulating
fibrinogen, contributes to liver injury in rats exposed to lipopolysaccharide
from gram-negative bacteria. J Pharmacol Exp Ther 1995 Jan;272(1):53-
62.

Smith MK, Mooney DJ. Hypoxia leads to necrotic hepatocyte death. J
Biomed Mater Res A 2007 Mar 1;80(3):520-529.

290

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

Ferrara N, Carver-Moore K, Chen H, Dowd M, Lu L, O'Shea KS, et al.
Heterozygous embryonic lethality induced by targeted inactivation of the
VEGF gene. Nature 1996 Apr 4;380(6573):439-442.

Ferrara N. Vascular endothelial growth factor and the regulation of
angiogenesis. Recent Prog Horrn Res 2000;55:15-35.

Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors.
Nat Med 2003 Jun;9(6):669-676.

Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular
endothelial growth factor is a secreted angiogenic mitogen. Science 1989
Dec 8;246(4935):1 306-1309.

Carmeliet P. VEGF gene therapy: stimulating angiogenesis or angioma-
genesis? Nat Med 2000 Oct;6(10):1102-1 103.

Zachary I. Signaling mechanisms mediating vascular protective actions of
vascular endothelial growth factor. Am J Physiol Cell Physiol 2001
Jun;280(6):C1375-C1386.

Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein M,
et al. Abnormal blood vessel development and lethality in embryos lacking
a single VEGF allele. Nature 1996 Apr 4;380(6573):435-439.

Vink A, Schoneveld AH, Lamers D, Houben AJ, van der GP, van Diest PJ,
et al. HIF-1 alpha expression is associated with an atheromatous
inflammatory plaque phenotype and upregulated in activated
macrophages. Atherosclerosis 2007 Dec;195(2):e69-e75.

Charalambous C, Pen LB, Su YS, Milan J, Chen TC, Hofman FM.
Interleukin-8 differentially regulates migration of tumor-associated and
normal human brain endothelial cells. Cancer Res 2005 Nov
15;65(22):10347-10354.

Chu SC, Tsai CH, Yang SF, Huang FM, Su YF, Hsieh YS, et al. Induction
of vascular endothelial growth factor gene expression by proinflammatory
cytokines in human pulp and gingival fibroblasts. J Endod 2004

Oct; 30(1 0):704-707.

Kawaguchi M, Akagi M, Gray MJ, Liu W, Fan F, Ellis LM. Regulation of
vascular endothelial growth factor expression in human gastric cancer
cells by interleukin-1beta. Surgery 2004 Sep;136(3):686-692.

Konishi N, Miki C, Yoshida T, Tanaka K, Toiyama Y, Kusunoki M.
Interleukin-1 receptor antagonist inhibits the expression of vascular
endothelial growth factor in colorectal carcinoma. Oncology 2005;68(2-
3):138-145.

291

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

Repovic P, Fears CY, Gladson CL, Benveniste EN. Oncostatin-M
induction of vascular endothelial growth factor expression in astroglioma
cells. Oncogene 2003 Nov 6;22(50):8117-8124.

Yeh HH, Lai W, Chen HH, Liu HS, Su WC. Autocrine IL-6-induced Stat3
activation contributes to the pathogenesis of lung adenocarcinoma and
malignant pleural effusion. Oncogene 2006 Jul 20;25(31):4300-4309.

Cross MJ, Dixelius J, Matsumoto T, CIaesson-Welsh L. VEGF-receptor
signal transduction. Trends Biochem Sci 2003 Sep;28(9):488-494.

Takahashi T, Yamaguchi S, Chida K, Shibuya M. A single
autophosphorylation site on KDR/Flk-1 is essential for VEGF-A-dependent
activation of PLC-gamma and DNA synthesis in vascular endothelial cells.
EMBO J 2001 Jun 1;20(11):2768—2778.

Warner AJ, Lopez-Dee J, Knight EL, Feramisco JR, Prigent SA. The Shc-
related adaptor protein, Sck, forms a complex with the vascular-
endotheliaI-grth-factor receptor KDR in transfected cells. Biochem J
2000 Apr 15;347(Pt 2):501-509.

Wu LW, Mayo LD, Dunbar JD, Kessler KM, Ozes ON, Warren RS, et al.
VRAP is an adaptor protein that binds KDR, a receptor for vascular
endothelial cell growth factor. J Biol Chem 2000 Mar 3;275(9):6059-6062.

Qi JH, Claesson-Welsh L. VEGF-induced activation of phosphoinositide 3-
kinase is dependent on focal adhesion kinase. Exp Cell Res 2001 Feb
1;263(1):173—182.

Rousseau S, Houle F, Landry J, Huot J. p38 MAP kinase activation by
vascular endothelial growth factor mediates actin reorganization and cell
migration in human endothelial cells. Oncogene 1997 Oct;15(18):2169-
2177.

Bates DO, Hillman NJ, Williams B, Neal CR, Pocock TM. Regulation of
microvascular permeability by vascular endothelial growth factors. J Anat
2002 Jun;200(6):581-597.

Reinders ME, Sho M, lzawa A, Wang P, Mukhopadhyay D, Koss KE, et al.
Proinflammatory functions of vascular endothelial growth factor in
alloimmunlty. J Clin Invest 2003 Dec;112(11):1655-1665.

Kayisli UA, Demir R, Erguler G, Arici A. Vasodilator-stimulated
phosphoprotein expression and its cytokine-mediated regulation in

vasculogenesis during human placental development. Mol Hum Reprod
2002 Nov;8(11):1023-1030.

292

179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

Tsao PN, Chan FT, Wei SC, Hsieh WS, Chou HC, Su YN, et al. Soluble
vascular endothelial growth factor receptor-1 protects mice in sepsis. Crit
Care Med 2007 Aug;35(8):1955-1960.

Tsuchihashi S, Ke B, Kaldas F, Flynn E, Busuttil RW, Briscoe DM, et al.
Vascular endothelial growth factor antagonist modulates leukocyte
trafficking and protects mouse livers against ischemialreperfusion injury.
Am J Pathol 2006 Feb;168(2):695—705.

Luyendyk JP, Lehman-McKeeman LD, Nelson DM, Bhaskaran VM, Reilly
TP, Car BD, et al. Unique gene expression and hepatocellular injury in the
lipopolysaccharide-ranitidine drug idiosyncrasy rat model: comparison with
famotidine. Toxicol Sci 2006 Apr;90(2):569-585.

Boulday G, Haskova Z, Reinders ME, Pal S, Briscoe DM. Vascular
endothelial growth factor-induced signaling pathways in endothelial cells
that mediate overexpression of the chemokine IFN-gamma-inducible
protein of 10 kDa in vitro and in vivo. J Immunol 2006 Mar 1;176(5):3098-
3107.

Schabbauer G, Schweighofer B, Mechtcheriakova D, Lucerna M, Binder
BR, Hofer E. Nuclear factor of activated T cells and early growth
response-1 cooperate to mediate tissue factor gene induction by vascular
endothelial growth factor in endothelial cells. Thromb Haemost 2007
Jun;97(6):988-997.

Pepper MS, Ferrara N, Orci L, Montesano R. Vascular endothelial growth
factor (VEGF) induces plasminogen activators and plasminogen activator
inhibitor-1 in microvascular endothelial cells. Biochem Biophys Res
Commun 1991 Dec 16;181(2):902-906.

Fischman AJ, Babich JW, Alpert NM, Vincent J, Wilkinson RA, Callahan
RJ, et al. Pharmacokinetics of 18F-Iabeled trovafloxacin in normal and
Escherichia coli-infected rats and rabbits studied with positron emission
tomography. Clin Microbiol Infect 1997 Jun;3(3):379.

Dalvie DK, Khosla N, Vincent J. Excretion and metabolism of trovafloxacin
in humans. Drug Metab Dispos 1997 Apr;25(4):423-427.

Outman WR, Nightingale CH. Metabolism and the fluoroquinolones. Am J
Med 1989 Dec 29;87(6C):37S-42$.

Melnik G, Schwesinger WH, Teng R, Dogolo LC, Vincent J. Hepatobiliary

elimination of trovafloxacin and metabolites following single oral doses in
healthy volunteers. Eur J Clin Microbiol Infect Dis 1998 Jun;17(6):424-426.

293

189.

190.

191.

192.

193.

194.

195.

196.

197.

198.

199.

200.

Sun Q, Zhu R, Foss FW, Jr., Macdonald TL. Mechanisms of trovafloxacin
hepatotoxicity: studies of a model cyclopropylamine-containing system.
Bioorg Med Chem Lett 2007 Dec 15;17(24):6682-6686.

Stahlmann R. Clinical toxicological aspects of fluoroquinolones. Toxicol
Lett 2002 Feb 28;127(1-3):269-277.

De SA, De SG. Adverse reactions to fluoroquinolones. an overview on
mechanistic aspects. Curr Med Chem 2001 Mar;8(4):371-384.

Chen HJ, Bloch KJ, Maclean JA. Acute eosinophilic hepatitis from
trovafloxacin. N Engl J Med 2000 Feb 3;342(5):359-360.

Lucena MI, Andrade RJ, Rodrigo L, Salmeron J, Alvarez A, Lopez-Garrido
MJ, et al. Trovafloxacin-induced acute hepatitis. Clin Infect Dis 2000
Feb;30(2):400—401 .

Blum MD, Graham DJ, McCloskey CA. Temafloxacin syndrome: review of
95 cases. Clin Infect Dis 1994 Jun;18(6):946-950.

Zimpfer A, Propst A, Mikuz G, Vogel W, Terracciano L, Stadlmann S.
Ciprofloxacin-induced acute liver injury: case report and review of
literature. Virchows Arch 2004 Jan;444(1):87-89.

Girard AE, Girard D, Gootz TD, Faiella JA, Cimochowski CR. In vivo
efficacy of trovafloxacin (CP-99,219), a new quinolone with extended
activities against gram-positive pathogens, Streptococcus pneumoniae,
and Bacteroides fragilis. Antimicrob Agents Chemother 1995
Oct;39(10):2210-2216.

Sokol WN, Jr., Sullivan JG, Acampora MD, Busman TA, Notario GF. A
prospective, double-blind, multicenter study comparing clarithromycin
extended-release with trovafloxacin in patients with community-acquired
pneumonia. Clin Ther 2002 Apr;24(4):605-615.

Croom KF, Goa KL. Levofloxacin: a review of its use in the treatment of
bacterial infections in the United States. Drugs 2003;63(24):2769-2802.

Onyeji CO, Bui KQ, Owens RC, Jr., Nicolau DP, Quintiliani R, Nightingale
CH. Comparative efficacies of levofloxacin and ciprofloxacin against
Streptococcus pneumoniae in a mouse model of experimental
septicaemia. Int J Antimicrob Agents 1999 Jul;12(2):107-114.

Lubasch A, Keller I, Borner K, Koeppe P, Lode H. Comparative
pharmacokinetics of ciprofloxacin, gatifloxacin, grepafloxacin, levofloxacin,
trovafloxacin, and moxifloxacin after single oral administration in healthy
volunteers. Antimicrob Agents Chemother 2000 Oct;44(10):2600-2603.

294

201.

202.

203.

204.

205.

206.

207.

208.

209

210.

211.

Witkamp R, Monshouwer M. Signal transduction in inflammatory
processes, current and future therapeutic targets: a mini review. Vet Q
2000 Jan;22(1):11-16.

Uetrecht J. Screening for the potential of a drug candidate to cause
idiosyncratic drug reactions. Drug Discov Today 2003 Sep 15;8(18):832-
837. .

Liguori MJ, Anderson MG, Bukofzer S, McKim J, Pregenzer JF, Retief J,
et al. Microarray analysis in human hepatocytes suggests a mechanism
for hepatotoxicity induced by trovafloxacin. Hepatology 2005
Jan;41(1):177-186.

Ng W, Lutsar I, Wubbel L, Ghaffar F, Jafri H, McCracken GH, et al.
Pharmacodynamics of trovafloxacin in a mouse model of cephalosporin-

resistant Streptococcus pneumoniae pneumonia. J Antimicrob Chemother
1999 Jun;43(6):811-816.

Teng R, Girard D, Gootz TD, Foulds G, Liston TE. Pharmacokinetics of
trovafloxacin (OP-99,219), a new quinolone, in rats, dogs, and monkeys.
Antimicrob Agents Chemother 1996 Mar;40(3):561-566.

Nightingale SL. From the Food and Drug Administration. JAMA 1999 Jul
7;282(1):19.

Teng R, Harris SC, Nix DE, Schentag JJ, Foulds G, Liston TE.
Pharmacokinetics and safety of trovafloxacin (OP-99,219), a new
quinolone antibiotic, following administration of single oral doses to
healthy male volunteers. J Antimicrob Chemother 1995 Aug;36(2):385-394.

Ernst ME, Ernst EJ, Klepser ME. Levofloxacin and trovafloxacin: the next
generation of fluoroquinolones? Am J Health Syst Pharm 1997 Nov
15;54(22):2569-2584.

. Colletti LM, Remick DG, Burtch GD, Kunkel SL, Strieter RM, Campbell DA,

Jr. Role of tumor necrosis factor-alpha in the pathophysiologic alterations
after hepatic ischemialreperfusion injury in the rat. J Clin Invest 1990
Jun;85(6):1936-1943.

Tiegs G, Walter M, Wendel A. Tumor necrosis factor is a terminal
mediator in galactosamine/endotoxin-induced hepatitis in mice. Biochem
Pharmacol 1989 Feb 15;38(4):627-631.

Windmeier C, Gressner AM. Pharmacological aspects of pentoxifylline

with emphasis on its inhibitory actions on hepatic fibrogenesis. Gen
Pharmacol 1997 Aug;29(2):181-196.

295

212.

213.

214.

215.

216.

217.

218.

219.

220.

221.

Kiemer AK, Muller C, Vollmar AM. Inhibition of LPS-induced nitric oxide
and TNF-alpha production by alpha-lipoic acid in rat Kupffer cells and in
RAW 264.7 murine macrophages. Immunol Cell Biol 2002 Dec;80(6):550-
557.

Tsukada S, Enomoto N, Takei Y, Hirose M, lkejima K, Kitamura T, et al.
Dalteparin sodium prevents liver injury due to lipopolysaccharide in rat
through suppression of tumor necrosis factor-alpha production by Kupffer
cells. Alcohol Clin Exp Res 2003 Aug;27(8 Suppl):7S-11S.

Khan AA, Slifer TR, Araujo FG, Suzuki Y, Remington JS. Protection
against lipopolysaccharide-induced death by fluoroquinolones. Antimicrob
Agents Chemother 2000 Nov;44(11):3169-3173.

Purswani M, Eckert S, Arora H, Johann-Liang R, Noel GJ. The effect of
three broad-spectrum antimicrobials on mononuclear cell responses to
encapsulated bacteria: evidence for down-regulation of cytokine mRNA
transcription by trovafloxacin. J Antimicrob Chemother 2000
Dec;46(6):921-929.

Shaw PJ, Hopfensperger MJ, Ganey PE, Roth RA. Lipopolysaccharide
and trovafloxacin coexposure in mice causes idiosyncrasy-like liver injury
dependent on tumor necrosis factor-alpha. Toxicol Sci 2007
Nov;100(1):259-266.

Schwandner R, Dziarski R, Wesche H, Rothe M, Kirschning CJ.
Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by
toll-like receptor 2. J Biol Chem 1999 Jun 18;274(25):17406-17409.

Yoshimura A, Lien E, lngalls RR, Tuomanen E, Dziarski R, Golenbock D.
Cutting edge: recognition of Gram-positive bacterial cell wall components
by the innate immune system occurs via Toll-like receptor 2. J Immunol
1999 Jul 1;163(1):1-5.

Yee SB, Hanumegowda UM, Hotchkiss JA, Ganey PE, Roth RA. Role of
neutrophils in the synergistic liver injury from monocrotaline and bacterial
lipopolysaccharide exposure. Toxicol Sci 2003 Mar;72(1):43-56.

Majcherczyk PA, Rubli E, Heumann D, Glauser MP, Moreillon P. Teichoic
acids are not required for Streptococcus pneumoniae and Staphylococcus
aureus cell walls to trigger the release of tumor necrosis factor by
peripheral blood monocytes. Infect Immun 2003 Jul;71(7):3707-3713.

Moreillon P, Majcherczyk PA. Proinflammatory activity of cell-wall

constituents from gram-positive bacteria. Scand J Infect Dis
2003;35(9):632-641.

296

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

Weber JR, Moreillon P, Tuomanen El. Innate sensors for Gram-positive
bacteria. Curr Opin Immunol 2003 Aug; 1 5(4):408-41 5.

Gerosa F, Baldani-Guerra B, Lyakh LA, Batoni G, Esin S, Winkler-Pickett
RT, et al. Differential regulation of interleukin 12 and interleukin 23
production in human dendritic cells. J Exp Med 2008 Jun 9;205(6):1447-
1461. -

Perrin-Cocon L, Agaugue S, Diaz 0, Vanbervliet B, Dollet S, Guironnet-
Paquet A, et al. Th1 disabled function in response to TLR4 stimulation of
monocyte-derived DC from patients chronically-infected by hepatitis C
virus. PLoS ONE 2008;3(5):e2260.

Del VM, Bajetta E, Canova S, Lotze MT, Wesa A, Parmiani G, et al.
Interleukin-12: biological properties and clinical application. Clin Cancer
Res 2007 Aug 15;13(16):4677-4685.

Ouaaz F, Li M, Beg AA. A critical role for the RelA subunit of nuclear
factor kappaB in regulation of multiple immune-response genes and in
Fas-induced cell death. J Exp Med 1999 Mar 15;189(6):999-1004.

Li X, Klintman D, Liu Q, Sato T, Jeppsson B, Thorlacius H. Critical role of
CXC chemokines in endotoxemic liver injury in mice. J Leukoc Biol 2004
Mar;75(3):443-452.

Deng X, Luyendyk JP, Zou W, Lu J, Malle E, Ganey PE, et al. Neutrophil
interaction with the hemostatic system contributes to liver injury in rats

cotreated with lipopolysaccharide and ranitidine. J Pharmacol Exp Ther
2007 Aug;322(2):852-861.

Jaeschke H, Farhood A, Fisher MA, Smith CW. Sequestration of
neutrophils in the hepatic vasculature during endotoxemia is independent

of beta 2 integrins and intercellular adhesion molecule-1. Shock 1996
Nov;6(5):351-356.

Yoshimura T, Takahashi M. IFN-gamma-mediated survival enables
human neutrophils to produce MCP-1/CCL2 in response to activation by
TLR ligands. J Immunol 2007 Aug 1;179(3):1942-1949.

Shiratsuchi Y, lyoda T, Tanimoto N, Kegai D, Nagata K, Kobayashi Y.
Infiltrating neutrophils induce allospecific CTL in response to immunization
with apoptotic cells via MCP-1 production. J Leukoc Biol 2007
Feb;81(2):412-420.

Walcheck B, Herrera AH, St HC, Mattila PE, Whitney AR, Deleo FR.
ADAM17 activity during human neutrophil activation and apoptosis. Eur J
Immunol 2006 Apr;36(4):968-976.

297

233.

234.

235.

236.

237.

238.

239.

240.

241.

242.

Bertino J, Jr., Fish D. The safety profile of the fluoroquinolones. Clin Ther
2000 Jul;22(7):798-817.

Hayder H, Blanden RV, Korner H, Riminton DS, Sedgwick JD, Mullbacher
A. Adenovirus-induced liver pathology is mediated through TNF receptors
l and II but is independent of TNF or Iymphotoxin. J Immunol 1999 Aug
1;163(3):1516—1520.

Kusters S, Tiegs G, Alexopoulou L, Pasparakis M, Douni E, Kunstle G, et
al. In vivo evidence for a functional role of both tumor necrosis factor

(T NF) receptors and transmembrane TNF in experimental hepatitis. Eur J
Immunol 1997 Nov;27(11):2870-2875.

Schumann J, Bluethmann H, Tiegs G. Synergism of Pseudomonas
aeruginosa exotoxin A with endotoxin, superantigen, or TNF results in T.
Immunol Lett 2000 Oct 3;74(2):165-172.

Copple BL, Banes A, Ganey PE, Roth RA. Endothelial cell injury and fibrin
deposition in rat liver after monocrotaline exposure. Toxicol Sci 2002
Feb;65(2):309-318.

Natoli G, Costanzo A, Moretti F, Fulco M, Balsano C, Levrero M. Tumor
necrosis factor (T NF) receptor 1 signaling downstream of TNF receptor-
associated factor 2. Nuclear factor kappaB (NFkappaB)-inducing kinase
requirement for activation of activating protein 1 and NFkappaB but not of
c-Jun N-terminal kinase/stress-activated protein kinase. J Biol Chem 1997
Oct 17;272(42):26079-26082.

Varghese J, Chattopadhaya S, Sarin A. Inhibition of p38 kinase reveals a
TNF-alpha-mediated, caspase-dependent, apoptotic death pathway in a
human myelomonocyte cell line. J Immunol 2001 Jun 1;166(11):6570-
6577.

Nowak M, Gaines GC, Rosenberg J, Minter R, Bahjat FR, Rectenwald J,
et al. LPS-induced liver injury in D-galactosamine—sensitized mice requires
secreted TNF-alpha and the TNF-p55 receptor. Am J Physiol Regul Integr
Comp Physiol 2000 May;278(5):R1202-R1209.

Dopp JM, Sarafian TA, Spinella FM, Kahn MA, Shau H, de VJ. Expression
of the p75 TNF receptor is linked to TNF-induced NFkappaB translocation
and oxyradical neutralization in glial cells. Neurochem Res 2002
Nov;27(11):1535-1542.

Tartaglia LA, Pennica D, Goeddel DV. Ligand passing: the 75-kDa tumor
necrosis factor (T NF) receptor recruits TNF for signaling by the 55-kDa
TNF receptor. J Biol Chem 1993 Sep 5;268(25):18542-18548.

298

243.

244.

245

246.

247.

248.

249.

250.

251.

252.

253.

254.

Pelagi M, Cumis F, Colombo B, Rovere P, Sacchi A, Manfredi AA, et al.
Caspase inhibition reveals functional cooperation between p55— and p75—
TNF receptors in cell necrosis. Eur Cytokine Netw 2000 Dec;11(4):580-
588.

Tukov FF, Luyendyk JP, Ganey PE, Roth RA. The role of tumor necrosis
factor alpha in Iipopolysaccharide/ranitidine-induced inflammatory liver
injury. Toxicol Sci 2007 Nov;100(1):267-280.

. Kamochi M, Kamochi F, Kim YB, Sawh S, Sanders JM, Sarembock I, et al.

P-selectin and lCAM-1 mediate endotoxin-induced neutrophil recruitment
and injury to the lung and liver. Am J Physiol 1999 Aug;277(2 Pt 1):L310-
L319.

Dri P, Haas E, Cramer R, Menegazzi R, Gasparini C, Martinelli R, et al.
Role of the 75-kDa TNF receptor in TNF-induced activation of neutrophil
respiratory burst. J Immunol 1999 Jan 1;162(1):460-466.

Nawroth P, Handley D, Matsueda G, De WR, Gerlach H, Blohm D, et al.
Tumor necrosis factor/cachectin-induced intravascular fibrin formation in
meth A fibrosarcomas. J Exp Med 1988 Aug 1;168(2):637-647.

Takeshita Y, Takamura T, Hamaguchi E, Shimizu A, Ota T, Sakurai M, et
al. Tumor necrosis factor-alpha-induced production of plasminogen
activator inhibitor 1 and its regulation by pioglitazone and cerivastatin in a
nonmalignant human hepatocyte cell line. Metabolism 2006
Nov;55(11):1464-1472.

Schwabe RF, Brenner DA. Mechanisms of Liver Injury. I. TNF-alpha-
induced liver injury: role of IKK, JNK, and ROS pathways. Am J Physiol
Gastrointest Liver Physiol 2006 Apr;290(4):G583-6589.

Sheri HM, Pervaiz S. TNF receptor superfamin-induced cell death: redox-
dependent execution. FASEB J 2006 Aug;20(10):1589-1598.

Gezginci S, Bolkent S. The effect of Z-FA.FMK on D-galactosamine/TNF-
alpha-induced liver injury in mice. Cell Biochem Funct 2007
May;25(3):277-286.

Wu D, Cederbaum A. Cytochrome P4502E1 sensitizes to tumor necrosis
factor alpha-induced liver injury through activation of mitogen-activated
protein kinases in mice. Hepatology 2008 Mar;47(3):1005-1017.

Bemelmans MH, van Tits LJ, Buun'nan WA. Tumor necrosis factor:
function, release and clearance. Crit Rev Immunol 1996;16(1):1-11.

Tafazoli S, Spehar DD, O'Brien PJ. Oxidative stress mediated
idiosyncratic drug toxicity. Drug Metab Rev 2005;37(2):311-325.

299

 

 

255.

256.

257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

Varrna TK, Lin CY, Toliver-Kinsky TE, Shem/cod ER. Endotoxin-induced
gamma interferon production: contributing cell types and key regulatory
factors. Clin Diagn Lab Immunol 2002 May;9(3):530-543.

Kim WH, Hong F, Radaeva S, Jaruga B, Fan S, Gao B. STAT1 plays an
essential role in LPS/D-galactosamine-induced liver apoptosis and injury.
Am J Physiol Gastrointest Liver Physiol 2003 Oct;285(4):G761-G768.

Kusters S, Gantner F, Kunstle G, Tiegs G. Interferon gamma plays a
critical role in T cell-dependent liver injury in mice initiated by concanavalin
A. Gastroenterology 1996 Aug;1 11(2):462-471.

Rajagopalan D. A comparison of statistical methods for analysis of high
density oligonucleotide array data. Bioinformatics 2003 Aug
12;19(12):1469-1476.

Stark GR, Kerr IM, Williams BR, Silverman RH, Schreiber RD. How cells
respond to interferons. Annu Rev Biochem 1998;67:227-264.

Tsutsui H, Matsui K, Okamura H, Nakanishi K. Pathophysiological roles of
interleukin-18 in inflammatory liver diseases. Immunol Rev 2000
Apr; 1 74: 1 92-209.

Senaldi G, Shaklee CL, Guo J, Martin L, Boone T, Mak TW, et al.
Protection against the mortality associated with disease models mediated
by TNF and IFN-gamma in mice lacking IFN regulatory factor-1. J
Immunol 1999 Dec 15;163(12):6820-6826.

Kano A, Watanabe Y, Takeda N, Aizawa S, Akaike T. Analysis of IFN-
gamma-induced cell cycle arrest and cell death in hepatocytes. J Biochem
1 997 Apr; 1 21 (4):677-683.

Adamson GM, Billings RE. Cytokine toxicity and induction of NO synthase
activity in cultured mouse hepatocytes. Toxicol Appl Pharmacol 1993
Mar;119(1):100-107.

Edwards SW, Say JE, Hughes V. Gamma interferon enhances the killing
of Staphylococcus aureus by human neutrophils. J Gen Microbiol 1988
Jan;134(1):37-42.

Suzuki K, F urui H, Kaneko M, Takagi K, Satake T. Priming effect of
recombinant human interleukin-2 and recombinant human interferon-

gamma on human neutrophil superoxide production.
Arzneimittelforschung 1990 Oct;40(10):1176-1 179.

Adams DO, Hamilton TA. The cell biology of macrophage activation. Annu
Rev Immunol 1984;2:283-318.

300

267.

268.

269.

270.

271.

272.

273.

274.

275.

276.

277.

Koj A, Gauldie J, Regoeczi E, Sauder DN, Sweeney GD. The acute-phase
response of cultured rat hepatocytes. System characterization and the
effect of human cytokines. Biochem J 1984 Dec 1;224(2):505-514.

Luyendyk JP, Copple BL, Barton CC, Ganey PE, Roth RA. Augmentation
of aflatoxin B1 hepatotoxicity by endotoxin: involvement of endothelium
and the coagulation system. Toxicol Sci 2003 Mar;72(1):171-181.

Luyendyk JP, Lehman-McKeeman LD, Nelson DM, Bhaskaran VM, Reilly
TP, Car 80, et al. Coagulation-dependent gene expression and liver injury
in rats given lipopolysaccharide with ranitidine but not with famotidine. J
Pharmacol Exp Ther 2006 May;317(2):635-643.

Yee SB, Hanumegowda UM, Copple BL, Shibuya M, Ganey PE, Roth RA.
Endothelial cell injury and coagulation system activation during synergistic
hepatotoxicity from monocrotaline and bacterial lipopolysaccharide
coexposure. Toxicol Sci 2003 Jul;74(1):203-214.

Ranby M, Brandstrom A. Biological control of tissue plasminogen
activator-mediated fibrinolysis. Enzyme 1988;40(2-3):130-143.

Gnant MF, Turner EM, Alexander HR, Jr. Effects of hyperthermia and
tumour necrosis factor on inflammatory cytokine secretion and
procoagulant activity in endothelial cells. Cytokine 2000 Apr;12(4):339-347.

Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC,
Schumacker PT. Mitochondrial reactive oxygen species trigger hypoxia-
induced transcription. Proc Natl Acad Sci U S A 1998 Sep
29;95(20):11715-11720.

Chandel NS, McClintock DS, Feliciano CE, Wood TM, Melendez JA,
Rodriguez AM, et al. Reactive oxygen species generated at mitochondrial
complex Ill stabilize hypoxia-inducible factor-1alpha during hypoxia: a
mechanism of O2 sensing. J Biol Chem 2000 Aug 18;275(33):25130—
25138.

Smith MK, Mooney DJ. Hypoxia leads to necrotic hepatocyte death. J
Biomed Mater Res A 2007 Mar 1;80(3):520-529.

Forsythe JA, Jiang BH, lyer NV, Agani F, Leung SW, Koos RD, et al.
Activation of vascular endothelial growth factor gene transcription by
hypoxia-inducible factor 1. Mol Cell Biol 1996 Sep;16(9):4604-4613.

Arteel GE. New role of plasminogen activator inhibitor-1 in alcohol-induced
liver injury. J Gastroenterol Hepatol 2008 Mar;23 Suppl 12854-859.

301

278.

279.

280.

281.

282.

Bergheim I, Guo L, Davis MA, Duveau I, Arteel GE. Critical role of
plasminogen activator inhibitor-1 in cholestatic liver injury and fibrosis. J
Pharmacol Exp Ther 2006 Feb;316(2):592-600.

Deng X, Lu J, Lehman-McKeeman LD, Malle E, Crandall D, Ganey PE, et
al. p38 MAPK-dependent TNF-{alpha} converting enzyme is important for
liver injury in hepatotoxic interaction between lipopolysaccharide and
ranitidine. J Pharmacol Exp Ther 2008 Apr 4.

Renckens R, Pater JM, van der PT. Plasminogen activator inhibitor type-
1-deficient mice have an enhanced IFN-gamma response to
lipopolysaccharide and staphylococcal enterotoxin B. J Immunol 2006 Dec
1;177(11):8171-8176.

Balsara RD, Castellino FJ, Ploplis VA. A novel function of plasminogen
activator inhibitor-1 in modulation of the AKT pathway in wild-type and
plasminogen activator inhibitor-1-deficient endothelial cells. J Biol Chem
2006 Aug 11;281(32):22527-22536.

Wanidworanun C, Strober W. Predominant role of tumor necrosis factor-

alpha in human monocyte lL-10 synthesis. J Immunol 1993 Dec
15;151(12):6853-6861.

302

 

Illllllllllljllrilli