Role of CrmA in regulating inflammation and apoptosis during cowpox virus infection

MISSING IMAGE

Material Information

Title:
Role of CrmA in regulating inflammation and apoptosis during cowpox virus infection
Physical Description:
xvi, 239 leaves : ill. ; 29 cm.
Language:
English
Creator:
Nathaniel, Rajkumar, 1971-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Cowpox   ( mesh )
Poxviridae Infections   ( mesh )
Serpins   ( mesh )
Inflammation   ( mesh )
Apoptosis   ( mesh )
Capases   ( mesh )
Interleukins   ( mesh )
Immunity   ( mesh )
Chick Embryo   ( mesh )
Department of Molecular Genetics and Microbiology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Molecular Genetics and Microbiology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 2003.
Bibliography:
Includes bibliographical references (leaves 196-238).
Additional Physical Form:
Also available online.
Statement of Responsibility:
by Nathaniel Rajkuma.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029530550
oclc - 70171164
System ID:
AA00025810:00001

Full Text










ROLE OF CrmA IN REGULATING INFLAMMATION AND APOPTOSIS DURING
COWPOX VIRUS INFECTION














By

RAJKUMAR NATHANIEL


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

UNIVERSITY OF FLORIDA


2003















ACKNOWLEDGMENTS

A number of people have contributed to the completion of this work. I would like

to thank my mentor Richard Moyer for allowing me to start this project and also for

helping me complete it. Dick has been extremely patient and forgiving with me these

past 6 and some years. He continuously made efforts to train me as a scientist. Dick

helped me appreciate poxviruses. The members of my committee Lung-Ji Chang, Paul

Gulig, and Michael Clare-Salzler have also been instrumental in guiding me through

these years. They have sat patiently through many more committee meetings than they

would have wanted to. I would also like to thank Dr. Richard Condit and Dr. Sue Moyer

for valuable input and suggestions during joint lab-meetings.

The members of the Moyer Lab need special mention. They have been good

friends and work colleagues. Mike Duke taught me the difference between Republicans

and Democrats. Mike and I have had many philosophical discussions about the world we

live in and I will miss those interactions. Alison Bawden, Traci Ness and Kristin Moon of

the so-called Estrogen Sanctuary made sure I wasn't naive by the time they left the lab.

Ben Luttge and Lauren Brum have been wonderful work mates and strongly encouraged

my graduation process. Peter Turner has been generous with his time and was always

there when I needed technical advice. I am grateful to Marie Becker, for reading my

manuscripts and for her mothering us in the lab. Qianjun Li has been a good friend, work

neighbor and advisor.









I am grateful to Joyce Conners for taking care of all the necessary paperwork. I

would also like to thank the administrative staff of the Department of Molecular Genetics

and Microbiology for their timely help.

I owe a lot to friends at the Graduate Evangelical Fellowship, who have seen me

through my graduate career. I would specially like to mention Chip Appel, whose

constant encouragement and friendship was invaluable at times. Brian and Shigeko

Jackson, Diomy Zamora, Kwabena Amphosa-Managera, Shelly Merves and Ricardo

Gomez have specially seen me through the final stages of completing this project. Their

friendship has been invaluable. My family, have patiently waited for me to finish this

degree and have been a positive influence through all these years. Finally, I would like to

express my heartfelt gratitude to my wife, Leena, for her love, patience and endurance

despite all the strain we have gone through my graduate career. Leena has truly been my

helpmate. I firmly believe in the notion that it takes a village to raise an individual.

There are many others, who have, in little ways influenced my life and shaped the way I

think. Like little drops of water that make a flowing river, these individuals have touched

my life in numerous ways.














TABLE OF CONTENTS
page

ACKNOW LEDGM ENTS .................................................................................................. ii

LIST OF TABLES........................................................................................................... viii

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

ABBREVIATION S ........................................................................................................... xi

ABSTRACT...................................................................................................................... xv

CHAPTER

1 LITERATURE REVIEW ............................................................................................... 1

Introduction..................................................................................................................... 1
Imm une System s............................................................................................................. 1
Innate Imm unity and Inflamm ation................................................................................ 3
Innate Imm une Recognition....................................................................................... 5
Toll and Interleukin-1 Receptors .................................................................................... 6
Toll and Toll Like Receptors (TLR).......................................................................... 6
Interleukin-1 (IL-1).................................................................................................... 8
Inflamm atory M editors ............................................................................................... 17
Tum or Necrosis Factor (TNF) ................................................................................. 17
Interleukin- 18 (IL- 18).............................................................................................. 21
Interleukin-8 (IL-8).................................................................................................. 25
Avian Cytokines............................................................................................................ 29
TNF-ax ...................................................................................................................... 29
IL-1 .......................................................................................................................... 30
IL-18 ........................................................................................................................ 32
CXC Chem okines .................................................................................................... 32
Viral Inhibitors of Inflamm ation................................................................................... 33
TNF Inhibitors ......................................................................................................... 33
IL-1 Inhibitors.......................................................................................................... 34
IL-18 Inhibitors........................................................................................................ 35
Chem okine Inhibitors............................................................................................... 36
Inhibitors of Complement........................................................................................ 38
Inhibitors of Interferons ........................................................................................... 39
Apoptosis ...................................................................................................................... 40









Caspases................................................................................................................... 44
Inflam m atory Caspases-1, -4, -5, -11 and -12.......................................................... 46
Apoptotic Initiator Caspases-2, -8, -9, and -10........................................................ 48
Apoptotic Effector Caspases-3, -6 and -7................................................................ 49
Intrinsic Activation of Caspases: Mitochondrial Permeability Transition Pore...... 51
Extrinsic Activation of Caspases: Death Receptor Mediated Signaling.................. 53
Granzym e B ............................................................................................................. 56
Apoptosis in Avian Species .......................................................................................... 57
Viral Inhibitors of Apoptosis ........................................................................................ 58
Viral Inhibitors of Mitochondria: PT Pore Modulators and Bcl-2 Homologs......... 59
vLIPs: Inhibitors of DISC ........................................................................................ 60
Baculovirus Caspase Inhibitor: P35......................................................................... 61
Baculovirus Inhibitor of Apoptosis Protein (IA P)................................................... 62
Caspase Independent Induction of Apoptosis............................................................... 64
Poxviruses..................................................................................................................... 65
Poxvirus Lifecycle ................................................................................................... 67
Im portance of Poxviruses......................................................................................... 69
Serine Protease Inhibitors (Serpins).............................................................................. 70
Poxvirus Serpins ........................................................................................................... 75
SERP2...................................................................................................................... 76
Cytokine Response M odifier A (Crm A).................................................................. 77
SPI-2 is a Crm A Hom ologue................................................................................... 80
RPV SPI-2 and CPV Crm A Equivalency..................................................................... 81
In Vivo Role of CrmA during CPV Infections of CAMs.............................................. 84

2 M ATERIA LS AN D M ETH OD S.................................................................................. 87

Virology ........................................................................................................................ 87
Cells ......................................................................................................................... 87
Viruses ..................................................................................................................... 87
Viral Stock Preparation and Quantification............................................................. 88
M olecular Techniques................................................................................................... 89
Polym erase Chain Reaction (PCR).......................................................................... 89
DNA Manipulation, Ligation and Transformation.................................................. 91
V iral DN A Preparation ............................................................................................ 92
Sodium Dodecyl Sulphate-Polyacrylamide Gel Electrophoresis (SDS-PAGE)...... 92
Im m unoblot.............................................................................................................. 93
Pulse Labeling.......................................................................................................... 94
DN A Sequencing ..................................................................................................... 94
Sequence Analysis and D database Search................................................................. 95
Recom binant V irus Construction.................................................................................. 95
Transfection of Infected Cells U sing Liposom es..................................................... 95
Selection of Recombinant Virus in the Presence of Mycophenolic Acid ................ 96
Reverse Selection of Recombinant Virus using 6-Thioguanine.............................. 99
Construction of Recombinant RPV Expressing CrmA............................................ 99
Construction of Recombinant RPV Expressing P35 ............................................. 100



v









Construction of Recombinant CPV Expressing SPI-2........................................... 102
Construction of Recombinant CPV Expressing P35 ............................................. 103
Construction of Recombinant CPV Expressing SERP2........................................ 104
Site Directed Mutagenesis of CrmA Gene............................................................. 104
Construction of CPVCrmA D303A....................................................................... 105
Construction of CPVAIL- 1 PR Virus ..................................................................... 105
A poptosis A ssays........................................................................................................ 106
DAPI Staining of Infected Cells........................................................................... 106
Preparation of Infected Cell Extracts for Caspase Activity Assay........................ 109
Ac-DEVD-AMC Cleavage Assay for Caspase Activity ....................................... 109
In Vivo Infection Assays using CAMs........................................................................ 110
Preparation of Chicken Chorioallantoic Membrane for Infections........................ 110
Infecting and Harvesting CAMs............................................................................ 111
Measurement of Reactive Oxygen Intermediates.................................................. I11
M TT R education A ssay........................................................................................... 112
V irus Infectivity A ssay .......................................................................................... 112
Terminal Caspase Activity Assay on Extracts from Infected CAMs.................... 113
Construction of Plasmids Containing Chicken ProlL- 11P or ProlL- 18.................. 114
Plasmid Containing Mouse ProlL- 1 3 .................................................................... 115
Quick Coupled Transcription/Translation System................................................. 115
In Vitro Cleavage Assay for Processing ProlL- 1 3 and ProlL- 18.......................... 115

3 R E SU LT S ................................................................................................................... 117

Equivalency of SPI-2 and CrmA................................................................................ 117
RPVASPI-2::crmA and CPVAcrmA::SPI-2 Recombinant Virus Construction..... 118
RPVASPI-2::crmA and CPVAcrmA::SPI-2 Prevent Apoptosis in Pig Cells........ 119
RPVASPI-2::crmA and CPVAcrmA::SPI-2 Inhibit Caspase Activation............... 124
Recombinant RPV and CPV Viruses Expressing P35........................................... 130
RPVASPI-2::P35 Induces Apoptosis in LLC-PK1 Cells....................................... 131
CPVAcrmA::P35 Blocks Apoptotic Induction in LLC-PK1 Cells........................ 132
Viral Protein Expression: Effect of AraC.............................................................. 134
Recombinant CPV Viruses Expressing SERP2 or CrmA D303A.............................. 137
Infections of LLC-PK1 Cells with CPV Recombinants: DAPI Stained Cells....... 138
Infections of LLC-PK1 Cells with CPV Recombinants: Ac-DEVD-AMC
C leavage A activity .............................................................................................. 141
Recombinant CPV Infections of CAMs ..................................................................... 144
Pock Morphology of Recombinant CPV Infected CAMs..................................... 145
All White Pocks Contain Heterophils.................................................................... 145
Inflammatory Pocks Show High Levels ofNADPH Oxidase............................... 148
P35 and Serp2 Restore Viral Yields from Infected CAMs.................................... 152
P35, Serp2 and CrmA Block Induction of Terminal Caspase Activity within
Infected Pocks................................................................................................... 153
Effect of Caspase Inhibitors on Protease Activity Present in
Infected C A M s................................................................................................. 155
Chicken ProIL-13 Processing Activity in CAMs.................................................. 159









Processing ofProIL-1 p can be Blocked by either SERP2 or P35 within
Inflam m atory Pocks.......................................................................................... 162
Processing ofProlL-18 can be Blocked by either SERP2 or P35 within
Inflam m atory Pocks.......................................................................................... 164
Chicken ProlL- 18 Processing Activity in CPVAcrmA Lysates is Blocked by
Caspase Specific Peptide Inhibitors.................................................................. 166
Caspase-3 Activity in the Presence of Chicken CAM Extracts........................ 171
CPVAIL- 1 PR Fails to Induce Inflammation on CAMs.............................................. 171

4 D ISC U SSIO N ............................................................................................................. 178

Equivalency of SPI-2 and CrmA ................................................................................ 178
Potential A pplications............................................................................................ 181
P35 Replacements of SPI-2 and CrmA....................................................................... 183
Function of CrmA during CPV Infections of CAMs.................................................. 184
Requirement of Intact RCL for CrmA Function In Vivo....................................... 185
Regulation of Inflammation and Apoptosis by CrmA are Distinct Functions....... 185
Inflammation during CPV Infection may not be due to IL- 1 P or IL-18................ 186
Models of Inflammatory Response on CAMs....................................................... 190
Future Studies .............................................................................................. ............... 193

LIST OF REFERENCES................................................................................................ 196

BIOGRAPHICAL SKETCH.......................................................................................... 239


LLI vii














LIST OF TABLES
Table page

1 Characterization of mice deficient for different components of the IL-1 system......... 15
2 List of mice deficient in the IL-1 system ...................................................................... 16
3 Taxonomy of poxviruses............................................................................................... 66
4 List of Primers................................................................................... ............................ 90
5 List of plasmids constructed....................................................................................... 107
6 List of constructed recombinant viruses..................................................................... 108
7 Summary of results using recombinant CPVs............................................................ 177














LIST OF FIGURES
Figure page

1 Signaling by TNF, IL-1R and TLR........................................................................... 11
2 A poptosis .................................................................................................................. 43
3 Poxvirus replication cycle......................................................................................... 68
4 Serpins as physiological regulators........................................................................... 72
5 Serpin R C L ............................................................................................................... 73
6 Inhibition of trypsin by alpha-1 antitrypsin.............................................................. 74
7 Comparison of SPI-2 and CrmA peptide sequences................................................. 83
8 Construction of recombinant RPV............................................................................ 120
9 Construction of recom binant CPV ............................................................................ 121
10 Expression of SPI-2 and CrmA during RPV infections........................................... 122
11 Expression of CrmA and SPI-2 during CPV infections........................................... 123
12 Morphological characteristics of LLC-PK1 cells infected with RPV
derivatives ...................................................................................................... 125
13 Morphological characteristics of LLC-PKI cells infected with CPV
derivatives ...................................................................................................... 126
14 Biochemical changes in LLC-PK1 cells infected by RPV derivatives.................... 128
15 Biochemical changes in LLC-PK1 cells infected by CPV derivatives.................... 129
16 Comparison of P35 expressed in RPV and CPV..................................................... 133
17 Effect of AraC on P35 expression in recombinant RPV and CPV.......................... 136
18 Expression of SERP2 and CrmA D303A recombinant proteins ............................. 139
19 Morphological changes in LLC-PK1 cells infected with CPVAcrmA::SERP2
or CPV Crm A D 303A ..................................................................................... 142
20 Biochemical changes in LLC-PK1 cells infected with CPVAcrmA::SERP2
or CPV Crm A D 303A ..................................................................................... 143
21 Pock morphology on CAMs infected with recombinant viruses............................. 146
22 Histological examination of pocks from infected CAMs........................................ 149
23 Inflammatory pocks show similar levels of oxidative burst.................................... 151
24 P35 and SERP2 restore viral yields from infected CAMs....................................... 154
25 P35, SERP2 and CrmA block terminal caspase activity within CEF cells.............. 156
26 P35, SERP2 and CrmA block terminal caspase activity within infected pocks...... 157
27 Chicken IL-1 3 processing........................................................................................ 161
28 P35 and SERP2 function like CrmA to prevent Caspase-1 mediated
processing of prolL- I P3.................................................................................. 163
29 P35 and SERP2 function like CrmA to prevent processing of chicken
prolL-18 in C A M s ....................................................................................... 165
30 Caspase-1 mediated processing of chicken prolL-18 is blocked by peptides.......... 168
31 Chicken prolL-18 processing activity in CPVAcrmA::lacZ extracts is
inhibited by caspase specific peptides.......................................................... 169









32 Caspase-3 activity in the presence of chicken extracts............................................ 172
33 Construction of CPVAIL-13R................................................................................. 174
34 CPVAIL-1OR fails to induce inflammation during CAM infections ....................... 175














ABBREVIATIONS


AcNPV Autographa californica nucleopolyhedrovirus

AP-1 activator protein- 1

APC antigen presenting cell

BIR baculovirus IAP repeat domain

BV baculovirus

CAM chicken chorioallantoic membrane

CEF chicken embryo fibroblast

CpGV Cydiapomonella granulosisvirus

CPV cowpoxvirus

CrmA cytokine response modifier A

CTL cytotoxic T lymphocytes

CV-1 monkey kidney cell line

DD death domain

DFF DNA fragmentation factor

DNA-PK DNA protein kinase

DISC death receptor induced signaling complex

E.coli gpt E.coli guanine phosphoribosyltransferase

FADD Fas associated death domain

G-CSF granulocyte colony stimulating factor









GM-CSF granulocyte-monocyte colony stimulating factor

HPRT hypoxanthine guanine phosphoribosyltransferase

lAP inhibitors of apoptosis proteins from baculovirus

ICE interleukin 1 1 converting enzyme

IFN interferon

IL interleukin

IRAK IL- 1 R associated kinase

ITA inhibitor of T-cell apoptosis

ITR inverted terminal repeat

JNK Jun N-terminal kinase

kbp kilobase pairs

LLC-PK1 pig kidney cell line

MAPK mitogen activated protein kinase

M-CSF monocyte colony stimulating factor

MMP matrix metalloproteinase

MOI multiplicity of infection

MPA mycophenolic acid

NADP nicotinamide adenine dinucleotide phosphate

NF-KB nuclear factor kappa B

NGF nerve growth factor

NIK NF-KB inducing kinase

NK natural killer cell

NO nitric oxide









OpNPV Orgyia pseudotsugata nucleopolyhedrovirus

P35 anti-apoptotic protein from baculovirus

PARP poly(ADP) ribose polymerase

PBS phosphate buffered saline

PCD programmed cell death

PR-3 proteinase-3

PT permeability transition (mitochondrial)

RPV rabbitpoxvirus

RCL reactive center loop

SAPK stress activated protein kinase

SERP2 serine proteinase inhibitor-2 from myxomavirus

SDS sodium dodecyl sulfate

SPI-1 serine proteinase inhibitor-1 from poxviruses

SPI-2 serine proteinase inhibitor-2 from poxviruses

SPI-3 serine proteinase inhibitor-3 from poxviruses

SPI-7 serine proteinase inhibitor-7 from swinepoxvirus

STO mouse fibroblast cell line

TACE TNF-a converting enzyme

TDS transient dominant selection

TIR Toll/IL- IR domain

TLR Toll like receptor

TNF tumor necrosis factor

TRAF TNF receptor associated kinase









variola virus

vaccinia virus

tumor necrosis factor receptor

TNFR associated death domain


VAR

VV

TNFR

TRADD















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

ROLE OF CrmA IN REGULATING INFLAMMATION AND APOPTOSIS DURING
COWPOXVIRUS INFECTION
By

Rajkumar Nathaniel

May 2003

Chair: Richard W. Moyer
Department: Molecular Genetics and Microbiology


Poxviruses are large DNA viruses that replicate exclusively in the cytoplasm of

the infected host cell. The importance of the host inflammatory response in controlling a

poxvirus infection is supported by the fact that approximately one third of the poxvirus

genome is directed at deflecting the early host response to virus infection. Viral serine

protease inhibitors (serpins) are proteins encoded by poxviruses that interfere with the

host immune response. Cowpox virus (CPV) encodes the serpin CrmA which has been

shown to inhibit inflammation and apoptosis during virus infection. Wild type cowpox

virus forms red, hemorrhagic, non-inflammatory pocks on the chorioallantoic membranes

(CAMs) of embryonated chicken eggs whereas CPV deleted for CrmA forms white,

inflammatory pocks accompanied by lowered yield of virus. In vitro studies have shown

CrmA to be a potent inhibitor ofcaspase-1, as well as caspases-8 and -10. Caspase-1 is

an enzyme responsible for the activation of the pro-inflammatory cytokines IL- 1 P and IL-

18 from their inactive precursors. A CrmA-mediated block in the activation of these pro-









inflammatory cytokines would be consistent with the role of CrmA in controlling

inflammation. CrmA also prevents apoptosis, consistent with the demonstrated inhibition

ofcaspases-8 and -10. The myxoma virus serpin, SERP2, has been shown to inhibit

caspases-1, -8 and -10 as does the non-serpin caspase inhibitor, P35, from baculovirus.

However, CPV recombinants where CrmA is replaced with SERP2, P35 or CrmA D303A

(mutation of the reactive center loop (RCL) which alters substrate specificity) all produce

white, inflammatory pocks on CAMs. Virus yields are restored from the P35 and SERP2

but not CrmA D303A recombinants. Pro-apoptotic caspase-3-like activity is detected

only from CPVAcrmA::lacZ and CPVCrmA D303A infections suggesting that loss of

functional CrmA leads to apoptosis and it is apoptosis, not inflammation which is

responsible for the reduction in virus yield. We did not detect caspase-1 within infected

CAMs through processing of chicken proIL-P or proIL-18, suggesting that caspase-1

mediated control of inflammation is unlikely. We show for the first time that the

regulation of apoptosis and inflammation are two distinct functions of CrmA that may not

depend solely on caspase inhibition.














CHAPTER 1
LITERATURE REVIEW

Introduction

Poxviruses are unique DNA viruses in that they replicate exclusively in the

cytoplasm of infected cells. Smallpox, caused by variola virus, is by far the most

infamous poxvirus and has been associated with human infections for about 2000 years

(102). The ability of poxviruses to evade the effectors of antiviral immunity is evidenced

by numerous poxvirus proteins that interact with the host immune system (271).

Studying the interactions of viral proteins would be useful in understanding the immune

response to a poxvirus infection and would also help in elucidating mechanisms of

immune evasion used by poxviruses. Before discussing the role of poxvirus proteins in

modulating the host immune response, it is necessary to review basic aspects of innate

immunity, mediators of inflammation, apoptosis and poxvirus biology.



Immune Systems

The mammalian immune system is divided into the innate and adaptive

components. Innate immunity provides defense against a wide variety of pathogens

without requiring prior exposure, involves an inflammatory response and takes place at

the local site of immune challenge (reviewed in 172, 293). Although innate immunity

was thought to be nonspecific, considerable advances in our understanding of the innate

immune system have made it clear that this response can also recognize self from non-









self. Innate immunity involves a pre-programmed system that includes receptors and the

signaling mechanisms that induce a full inflammatory response that results in up

regulation of proinflammatory cytokines such as tumor necrosis factor (TNF-ca),

interleukin-1 3 (IL-1 P), IL-6 and interferon y (IFNy). As a consequence of

proinflammatory mediator induction, cells such as macrophages, natural killer cells and

neutrophils are activated and migrate to the site of injury. In addition, increased

generation of reactive oxygen species and degranulation of cytoplasmic granules

containing lytic enzymes occur, that aid in the toxic effects associated with these cells.

The adaptive response, on the other hand, requires prior exposure and takes place

in the draining lymph nodes. In the lymph nodes professional antigen presenting cells

(APC) such as macrophages, B-cells and dendritic cells that arrive from sites of

inflammation activate naYve T and B-lymphocytes which then migrate back to the site of

inflammation. Antibodies are produced by activated B-lymphocytes and function to

neutralize the pathogen itself or offer protection against toxins made by pathogens.

Additionally antibodies bound to antigen enable phagocytic cells to ingest and destroy the

antibody coated or opsonized foreign particle. Antibodies bound to antigen can also

activate plasma proteins called complement, which through a cascade of events can lyse

and destroy certain bacteria and viruses. Activating complement also aids in

phagocytosis of pathogens. While extracellular pathogens can be taken care of by the

actions of antibody and complement, the intracellular pathogens are largely destroyed by

functions of the T-lymphocyte cell mediated response.

Interactions between T-lymphocytes and cells presenting antigenic markers drive

the cell-mediated response. Cytotoxic T cells (CTL) recognize virally infected cells and









cause the infected cell to undergo apoptosis or cell death thus preventing viral replication

and release of progeny. Intracellular pathogens taken up by macrophages can evade lysis

by residing in vesicles that have not fused with the lysosome. A subset to T cells termed

T-Helper 1 (TH 1) cells; promote activation of macrophages which allows fusion of their

lysosomes with vesicles thus destroying the invading pathogen. Protection against

extracellular pathogens mediated by antibodies also involves the function of a third

subset of T-lymphocytes termed T-helper 2 cells (TH2). TH2 cells stimulate B-cells to

proliferate and produce antibody to particular antigens. In most cases the adaptive

response confers life long immunity against re-infection by the same pathogen.



Innate Immunity and Inflammation

Inflammation is the reaction of living tissue to any form of injury. Traditionally

inflammation is defined by the Latin words calor, dolor, rubor and tumor meaning heat,

pain, redness and swelling. These physical observations of inflammation are the effects

of cytokines on local blood flow, which allows the accumulation of fluid and cells in the

inflamed area. Primary mediators of inflammation such as interleukin-1 (IL-1) and tumor

necrosis factor (TNF), which are released from injured tissue, cause dilation of blood

vessels around the area of injury. Endothelial cells lining the blood vessels swell and

partially retract allowing the leakage of fluid into the injured area. The increase in body

temperature is a host response to proinflammatory mediators that decreases viral/bacterial

replication and increases antigen presentation by APCs. Migration and extravasation of

leukocytes depend on interactions with the endothelium. Expression of adhesion

molecules such as selections and integrins and their interaction with corresponding ligands









on both leukocytes and endothelial cells control the emigration of leukocytes through

blood vessels. In addition secondary inflammatory mediators such as chemotactic

factors, chemokine IL-8 for neutrophils, are induced during inflammation and serve to

attract leukocytes to areas of tissue injury.

The purpose of an inflammatory reaction is to clear infection, remove dead tissue

and allow access of the immune system to the damaged area. The main cells initially

seen at sites of inflammation are neutrophils followed by the appearance of macrophages.

Activated neutrophils and macrophages undergo the process of respiratory burst whereby

products such as hydrogen peroxide (H202), superoxide radical (02") and nitric oxide

(NO) are produced. The produced reactive oxygen and nitrogen species are released into

the extracellular milieu and act as microbicidal agents. The release of oxygen radicals

and nitric oxide, as well as proteases present in neutrophilic granules such as lysozymes

and acid hydrolases, also leads to localized tissue destruction. In addition, the activation

of clotting pathways as a result of protease action from inflammatory exudates helps in

sequestering the invading pathogen from spread. Lipid mediators such as leukotrienes

(LTB4), prostaglandins and platelet activating factor (PAF) that are released by

phagocytic cells amplify the inflammatory response. PAF is chemotactic to leukocytes

and activates PMNs. Typically the injured area is replaced by tissue identical to that

which was lost. If restitution of the original tissue is not possible due to the severity of

injury, the area is replaced by fibrous, non-specific scar tissue. An acute inflammatory

process can become chronic if the injured area is continuously damaged while being

repaired.






5

Antigen present in the draining lymph nodes activates the adaptive arm of

immunity, and thus during the later stages of an inflammatory response lymphocytes are

involved. There is a delay of 4 to 7 days from the time of injury before the effects of an

adaptive response are seen. This period is sufficient for pathogens such as poxviruses to

replicate and cause systemic spread. Therefore, in such cases the innate immune

response is crucial in controlling infections during the period before an adaptive response

is seen.



Innate Immune Recognition

The mammalian innate immune response has two steps in recognizing invading

pathogens. The first involves recognition of pathogen associated molecular patterns

(PAMPs), and the second step is based on recognition of molecular markers specific for

self (172). PAMPs are conserved products of microbial metabolism that are distributed

and common among a number of pathogens. One example of a PAMP is

lipopolysaccharide (LPS) common to all gram-negative bacteria. The molecules that

recognize PAMPs are termed pattern recognition receptors (PRRs). PRRs include a

number of molecules that can be cytosolic, membrane bound and secreted in eukaryotic

cells. PRRs function to opsonize antigens, activate complement, initiate proinflammatory

and apoptotic cascades. A common PAMP associated with virus replication is the

formation of double stranded RNA that can bind and thereby activate protein kinase R

(PKR) (69). PKR phosphorylates eukaryotic initiation factor eIF-2a which serves to

block viral and cellular protein synthesis, inducing interferon responses and causes

apoptosis of infected cells (450). Double stranded RNA also activates 2'-5'-









oligoadenylate synthase (OAS) which synthesizes 2'-5'-oligoadenylate. The

oligonucleotide activates RNAaseL that cleaves viral and cellular RNA (210) thus

blocking viral replication and inducing apoptosis. Other intracellular PAMPs include

Nod 1, a cellular protein that binds to and regulates procaspase-9, an enzyme involved in

apoptotic induction (167). Nod2, a protein expressed in monocytes, activates nuclear

factor kappaB (NF-KB), a transcription factor associated with inflammation (294). The

complete range of ligands that interact with Nod proteins is currently unknown.

Extracellular PRRs include mannan-binding lectin (MBL) that binds mannose

residues that are abundant on microbial surfaces (111). Serine proteases MASP 1 and

MASP2 are activated by MBL and cleave C2 and C4 proteins, thereby initiating the

complement cascade. C-reactive protein (CRP) and serum amyloid protein (SAP)

function as opsonins by binding to bacterial phosphorylcholine (361). The macrophage

mannose receptor (MMR) facilitates the phagocytosis of bacterial and fungal pathogens

(111). Macrophage scavenger receptor (MSR) binds LPS and protects against endotoxic

shock (409). Other examples of extracellular PRRs are the Toll/Interleukin-1 family of

receptors.



Toll and Interleukin-1 Receptors

Toll and Toll-Like Receptors (TLR)

The Toll/IL- R family of receptors are conserved from flies to mammals and play

a central role in the induction of cellular innate immune responses. Drosophila Toll

ligand has been identified as Spatzle (221). The Drosophila Toll gene product was

originally found to be required for dorso-ventral polarity (139). Toll protein was later








found to control anti-fungal response in flies (221). The intracellular domain of Toll was

found to be homologous to a similar region in human interleukin-1 receptor (IL-1 R)

(293). Homologs to IL-1R have since been discovered, that now belong to the Toll-like

receptor (TLR) superfamily.

TLRs are transmembrane proteins that contain leucine rich repeat (LRR)

extracellular domains and Toll/IL-iR (TIR) cytoplasmic domains. The mammalian IL-1

system will be discussed in further detail later. Similarities exist between the fly Toll and

mammalian IL-1 system. Engagement of both receptors by the respective ligands results

in the activation of nuclear transcription factors. In addition to Spatzle, there also exist

adapter proteins such as Tube and a protein kinase Pelle. To date there are at least nine

Drosophila and ten mammalian TLRs known (172). In addition to the Toll ligand,

Spatzle, there are six other Spatzle-like proteins (281). TLR4, the first mammalian TLR

identified was shown to activate NF-KB and induce IL-1, IL-6 and IL-8 genes (254).

Studies with the mouse homolog of TLR4 demonstrated its role in the recognition of LPS

and gram-negative bacteria (172). TLR2 and TLR5 also recognize bacterial components

such as peptidoglycan, LPS, lipids from trypanosomes and flagellin (150, 362), (53),

(141). TLR2 binding has been shown to cause internalization of the receptor bound

ligand to the macrophage phagosome (141,426). Whether this is an exception or a rule

for most TLRs is to be seen. Experimental evidence also suggests that TLR1, 2 and 6

may function in a cooperative manner (138, 301). This suggests that heterodimerization

of receptors could in theory increase the repertoire of ligand specificities. TLR9 is

thought to reside intracellularly on lysosomes where it recognizes unmethylated CpG

motifs in bacterial DNA and induces THI response (145).









The initiation of signaling by TLRs could occur directly following receptor

engagement as in the case of TLR4 (254) (Figure 1). Alternatively the ligand may need

to be processed prior to receptor interaction as seen with Drosophila Spatzle, which

requires cleavage by serine protease Easter in order to activate Toll (221). The signaling

events following TLR engagement seem to be conserved from flies to mammals.

Signaling by TLR is depicted in Figure 1. Similar to IL-1 binding IL-1R, MyD88, an

adapter molecule that contains a TIR domain, is recruited to the receptor complex where

it causes autophosphorylation of serine threonine kinase IRAK (255). Complex protein

kinase interactions result in the activation and translocation of NF-KB to the nucleus

where it induces expression of inflammatory cytokine genes. TLR2, 5, 6 and 9 activate

NF-KB, while TLR2, 4, 6 and 9 can induce the transcription factor activator protein-1

(AP-1) activity (433). TLR2 has been shown to mediate NF-KB activity via PI3 kinase,

suggesting the involvement of GTP binding proteins (18). TLR2 involvement in

apoptosis has also been demonstrated (13, 14). The fact that mutant caspase-8 could

inhibit TLR2 activity suggests the possibility of direct activation of caspases by TLR2.

The inflammatory cytokines TNF-a, IL- 1 3, IL-18 and IL-8 that are upregulated by TLRs

(254) will be discussed in detail later. The identification of more signaling events

associated with TLRs will bring to light their different roles in the immune response. The

TLR system is similar to mammalian IL-1R system (293).



Interleukin-1 (IL-1)

IL-1 is one of the earliest cytokines produced in response to cellular damage or

infection. IL-1 P is the prototypical proinflammatory cytokine that belongs to the IL-1









family, which includes in addition IL- la and IL-1 receptor antagonist (IL-IRa) (98).

Both IL- I a and IL- I3 are synthesized as 31 kDa precursors that are processed to 17 kDa

mature cytokines by cysteine proteases calpain and caspase-1 respectively (55, 58, 189,

238, 413). ProlL-laot is processed between residues 112 tol 13 by calpain whereas prolL-

1lP is cleaved by caspase-1 first at Asp27 followed by cleavage at Asp 16 to yield mature

cytokine (198, 395). ProlL-laot also contains a nuclear localization signal at residues 79-

86 (448). Interestingly while prolL- 13 has an absolute requirement for processing to be

functional, prolL-la is biologically active as a precursor (58, 253,270). Neither IL-la

nor -P3 contains signal sequence regions that would target these proteins to the secretary

pathway (99). The predominant soluble form seems to be IL-1 3 P, whereas IL-la is

thought to be cell associated (21). IL-Ra is a soluble protein that is able to bind IL-1

receptors but is unable to transmit a signal and is thought to control IL-1 mediated

inflammation (86).

There are two receptors identified for IL-1. The 80 kDa IL-1 receptor type I (IL-

I RI) is responsible for signal transduction (Figure 1), whereas the 60 kDa type II receptor

(IL-1RII) does not deliver any signal and therefore functions to regulate IL-1 activity by

acting as a sink (370). The IL-1RI shares 45% homology with the cytoplasmic region of

Drosophila Toll protein (121). In addition to the TIR domain, IL-1RI and II contain

three extracellular immunoglobulin domains. Initial binding of IL-1 3 to the receptor is

with low affinity. Crystal structures of IL-1RI have revealed that binding of IL-1 p to the

receptor induces a conformational change in the receptor that allows the third Ig domain

on the receptor to wrap around the ligand (357, 435). This conformational change in the








receptor presumably facilitates the docking of IL- R accessory protein (IL- RAcP)

resulting in high affinity binding of IL-1 and the transduction of a signal.

The key signaling molecules activated are NF-KB, AP-1 and members of the

mitogen activated protein kinase (MAPK) family (166, 276,372, 386). The initial

signaling step post receptor engagement involves the association of adapter molecule

MyD88, a protein containing both a TIR domain and a death domain (DD) (50). The

MyD88 TIR is thought to trimerize with TIR domains of IL-IRI and IL-1RAcP in the

signaling complex, similarly to trimerization seen with the TNF-receptor (293). MyD88

recruits IL-1R associated kinases (IRAKs) -1 and -2 (282, 447). Like MyD88, IRAKs

also contain a DD, and it is conceivable that the recruitment of IRAKs to the signaling

complex is via the DDs. Although IRAK is phosphorylated on activation, a catalytically

inactive molecule is still active in signal transduction, suggesting that kinase activity is

not an absolute requirement (197, 245). IRAK functions to recruit TNF receptor

associated factor (TRAF-6), which in turn is responsible for the activation of NF-KB

inducing kinase (NIK) via kinases TAB-1 and TAK-1 (291). Consequently NIK

activates a complex called the signalsome containing IKB kinases (IKK) (42). TRAF-6

can also activate the signalsome via kinases ECSIT and MEKK1 (202, 220). This

complex activation of a number of protein kinases culminates in the activation and

translocation to the nucleus of transcription factors NF-KB and AP-1 (276, 372).

Interestingly NF-KB activation has been associated with both cell death and cell

proliferation (51). In general the state of the target cell determines the cellular pathway

involved. Post receptor signaling involves the activation of kinases of the MAPK family

that results in the activation and translocation of transcription factors. Unlike IL- I3, IL-

















f? .is]MgJ ^MyDW da D








asoiain ofl TRADD, whc recruts RI n TA2 to 2 th[rIetrAK-2lex

Aciaino KKJC ocur thrug th atvatio ofIK inss rbyRPeiae
oli rizo oIKK c s.eeRADD r
W(AK-2


TAI



NF-%:B I -KB-



Figure 1. Signaling by TNF, IL-1R and TLR. (A) TNF binds to TNFRI and causes the
association of TRADD, which recruits RIP and TRAF2 to the receptor complex.
Activation of IKK occurs through the activation of IKK kinases or by RIP mediated
oligomerization of IKK complexes. TNFRI induces cell death when TRADD recruits
FADD and activates caspase-8. (B) IL-1 and TLR receptor complexes which include
MyD88 and IRAKs induce the ubiquitin dependent association of TRAF6. As a result
IKK kinase Taki is activated which phosphorylates and activates the IKK complex.
Abbreviations: TNF, tumor necrosis factor; TNFR, TNF receptor; TRADD, TNF receptor
associated death domain; FADD, Fas ligand associated death domain; IKK, I kappa B
kinase; NF-KB, nuclear factor kappa B; IL, interleukin; LPS, lipopolysaccharide; IKK, I
kappa B kinase; IRAK, IL-1 receptor associated kinase, Mal, MyD88 adapter-like,
TRAF, TNF receptor associated factor. From reference 42.








IRa does not bind the third Ig domain on the receptor and cannot induce conformational

change in the receptor that would allow IL-1RAcP docking (357). IL-1 mediated signal

transduction was absent in both IL- 1RI-/ and IL-1RAcP-/- mice, suggesting that both the

receptor and accessory proteins were essential components for signaling (77, 126, 212).

The IL-1RII also has three Ig domains but lacks the TIR domain (252). IL-1RII binds all

three forms of IL-1 (252). The lack of the transducing cytoplasmic TIR domain indicates

that it functions as a negative regulator of IL- 1 activity. Consistent with this role, IL-

IRII has very low affinity for IL-1iRa (398). IL-1RII is expressed highly on lymphoid

and myeloid cells where it can be shed and act as a soluble decoy receptor (73). It was

presumed that proteolytic cleavage of the extracellular domains of membrane receptors

results in the soluble forms of these receptors (83). Indeed TNF-a converting enzyme

(TACE) has been shown to shed IL-1RII from cell membranes (334). Soluble forms of

the type I receptor have also been found in circulation. Patients with rheumatoid arthritis

and sepsis have elevated levels of soluble IL- IRs indicating that these receptors function

to regulate IL-1 activity in vivo (19, 125). In areas of local inflammation, tissue necrosis

can release precursor IL- 1p into the extracellular environment where it potentially can be

processed by proteases other than caspase-1. The maturation of IL-1p3 mediated by

enzymes other than caspase-1 is a consideration that may be important in the results

presented in this dissertation. Both IL-1 RI and II can bind precursor IL-1p3 and prevent

its processing, indicating the importance of soluble receptors in inflammatory fluids

(398).

The fact that biologically active IL- 3P was seen in caspase- 1'/ mice suggested the

existence of other proteases that could process prolL- I P3 (100). This observation may be








important in interpreting the results of experiments presented in this thesis. ProlL-1 3 can

be released from cells by an unknown mechanism despite the absence of any secretary

signal sequences in the molecule (33). In addition, cell lysis can also cause the release of

cytoplasmic proIL-1 p during tissue damage (142). A number ofproteases can be found

at sites of inflammation such as in synovial fluid, sites of neutrophil and macrophage

infiltration and at sites of tissue damage. Chymotrypsin and Stapylococcus aureus

protease are able to process the 31 kDa proIL- 1 3 to biologically active fragments ranging

from 17 to 19 kDa whereas the Trypsin cleavage product is inactive (38). Cathepsin G,

collagenase and elastase can also process proIL-1 P to bioactive cytokine (143). In

addition, the importance of such mechanisms in vivo was evidenced by similar processing

capabilities seen in synovial and bronchoalveolar lavage fluids from patients with

inflammatory polyarthritis and sarcoidosis (143). The CTL molecule granzyme A, a

serine protease, can also generate biologically active IL- I3, although this cleavage occurs

at Argl20, four residues downstream of the Asp 116 site used by caspase-1 (169). The

serine protease chymase, derived from dermal mast cells, also converts proIL-1 p to an

active species, suggesting a role for the initiation of inflammation in the skin (261).

Matrix metalloproteinases (MMPs) have also been implicated in the processing ofproIL-

1 P3 (355). Stromelysin (MMP-3) and gelatinase-A (MMP-2 and -B (MMP-9) can

generate biologically active forms of IL- IP. Extended incubation of the mature product

with these enzymes caused degradation. In addition, none of these MMPs were able to

process proIL- la. Proteinase-3, a proteolytic enzyme from activated neutrophils, has

also been shown to produce IL-1i P (71). The balance of accessible proteases at sites of

local inflammation may regulate the availability of active IL- 1 3 and thus modulate acute








and chronic states of inflammation. In summary, a number of proteases other than

caspase-1 can cleave pro-IL- P13 to generate the active cytokine.

A number of knockout mice models in the IL-1 system has been studied and are

listed in Tables 1 & 2. IL-I a is the least well studied, and observations indicate that it is

not involved in the inflammatory responses induced by turpentine or infection with

Listeria (98, 155). IL-1 P13 is crucial for the development of an acute phase response and

in the induction of febrile response to inflammation (472). Quite unexpectedly IL- 1 P1'

mice were hypersensitive to both IL-a and IL- I3, an observation that so far has not been

explained (12). Mice deleted for IL-Ra were seen to spontaneously develop autoimmune

disorders that resembled rheumatoid arthritis, arterial inflammation and were shown to be

highly susceptible to LPS induced lethality (149, 156, 290). IL-1RI knockout mice failed

to induce a biological response to either IL-1 a or 13, indicating that this receptor is

essential for all IL-1 activity (126, 212). In IL-IRAcP"'/ fibroblasts, binding of IL-l a or

IL-IRa was only moderately reduced, whereas IL- 13 binding affinity was reduced 70-

fold indicating the requirement of accessory protein for efficient receptor binding (77).

Similarly these mice failed to induce a biological response to IL-1. Disruption of the

signaling component IRAK caused reduced activation of NF-KB in fibroblasts and

attenuated responses to IL-1 in mice (410). Caspase- 1' /" mice are deficient in processing

IL-13 (209). Other studies have shown mature IL-1 13 to be released despite caspase-1

deficiency (100). Although caspase-1 I is the most important protease required for IL- 13

maturation other enzymes have been implicated to perform this function. The importance

of enzymes other than caspase- I in IL- 113 maturation may be important in certain cases

and will be discussed in relation to the results presented in this dissertation. These








00I





I &
(0





I roz




o4- t. 0Z z 7 7 y



& .0
'a 9 9 9





8 u1tg 0^I t
iiIN
VA0



8


.00
1 2
IN






16



Table 2. List of mice deficient in the IL-1 system.


IL-1 system component Mouse strain (reference)
IL-IRI C57BL/6 x 129J (126); C57BL/6 x 129/SV (212)
IL-1RAcP C57BL/6 (77)
IL-la C57BL/6 x DBA/2 (155)
IL-IP C57BL/6 x DBA/2 (155); C57BL/6 x 129EV (100)
IL-lreceptor antagonist C57BL/6 (149,290); MF-1 (290)
IRAK C57BL/6 (410)
Caspase-1 C57BL/6 x 129 (100)









studies underscore the importance of IL- 1 as a key proinflammatory cytokine that

mediates systemic effects and rapidly activates the innate immune response.

The end result of IL-1 signaling is the expression of genes, most of which are

involved in immunity and inflammation (81). IL-1 induces expression of a number of

cell proliferation factors such as granulocyte colony stimulating factor (G-CSF),

granulocyte-monocyte colony stimulating factor (GM-CSF) and monocyte colony

stimulating factor (M-CSF). In addition IL-1 also stimulates platelet production and

increases prostaglandin E synthesis. The production of a number of cytokines involved

in generating an inflammatory response is also stimulated by IL-1 including IL-1, -2, -3, -

6, -7, -8 and TNF. IL-1 can also activate natural killer (NK) cells, T- and B-cells. IL-1 is

a potent fever-inducing agent (81). The activities of IL- 1 demonstrate its role in

inducing an inflammatory response. The crucial role of IL-1 in forming part of the innate

immune response is borne out by the fact that pathogens such as poxviruses encode

molecules that interfere with IL-1 P function. The CrmA protein encoded by

cowpoxvirus is a potent inhibitor of caspase- I thus affecting the IL- I3 maturation

process (118). In addition cowpoxvirus also encodes a (decoy) receptor that binds IL- 13

preventing interaction with cellular IL-1R (381). Both CrmA and the virus-encoded IL-

I OR molecules are the subjects of this dissertation.



Inflammatory Mediators

Tumor Necrosis Factor (TNF)

Along with IL-1, TNF is considered a primary proinflammatory cytokine.

Although mainly produced by mononuclear cells, TNF can also be produced by a variety









of cells including monocytes, macrophages, mast cells, epithelial cells, neutrophils, as

well as T- and B- lymphocytes (277). It exists in two forms, TNF-a and TNF-P3. TNF-a

is synthesized as a membrane-bound 26 kDa protein that is proteolytically cleaved to a

soluble 17 kDa fragment at Val77 by TNF-a converting enzyme (TACE) (39, 273).

TACE is a membrane-anchored enzyme that belongs to the disintegrin-metalloprotease

family and contains an extracellular zinc-dependent protease domain. TNF-P is a 25 kDa

glycoprotein produced by lymphocytes and is also called lymphotoxin-a (130). TNF-a

and TNF-P share 50% homology at the amino acid level (443). Both of the molecules are

active only as homotrimers (24). Although TNF-a and P have similar affinities to the

TNF receptors, they differ in their ability to cause mitogenic and cytotoxic effects (319,

360).

TNF-P is more mitogenic, as its trimers are seen to be stable for several days and

were effective in inducing fibroblast growth that requires extended presence of the

cytokine (358,360). TNF-a trimers are less stable but have a 200-fold increase in

cytotoxicity over TNF-P3 (360). Gene disruptions in both TNF-ac and -P3 protected mice

from zymosan-induced organ dysfunction, indicating the involvement of these cytokines

in inflammation (436). Both TNF-ca and 3 are able to interact with two receptors that

mediate their biological effects (152). Binding to the receptors results in a wide range of

biological effects including cell proliferation, tumor necrosis, inflammation and

apoptosis, which are dependent on, cell condition and type of associated molecules

recruited to the signaling complex. TNF-ca is very effective in controlling local infection.

However, systemic release of TNF-a is detrimental to the host leading to septic shock








and ultimately death by multiple organ failure due to increased vascular permeability

resulting in loss of blood volume and plasma clotting proteins.

The TNF receptor, TNF-RI, has a molecular weight of 55 kDa and acts as the

main receptor type expressed by most cells (24). TNF-RI is thought to mediate most of

the proinflammatory effects of TNF-a (315). TNF receptor mediated signaling is

depicted in Figure 1. The intracellular domain of TNF-RI contains a death domain (DD)

that is required for apoptotic signaling of cell death (402). TNF-RI is always associated

with silencer of death domains (SODD), a cellular protein that functions to prevent the

inadvertent self-association of TNF-RI death domains (173). Following TNF binding to

TNF-RI, the receptor trimerizes leading to the disassociation of SODD. The adapter

protein TNF-R associated death domain (TRADD) interacts with TNF-RI via the DDs

(159). The Fas associated death domain (FADD) adapter protein that interacts with Fas

during apoptotic signaling does not directly bind TNF-RI, but can be recruited by

TRADD to the TNF-R signaling complex. (160). The association of FADD leads to the

recruitment of procaspase-8 followed by its activation and the initiation of the caspase

cascade.

The N-terminal domain of TRADD has been found to interact with the C-terminal

domain of TNF associated factor (TRAF) 2 (160). TRAF-2 plays a role in the activation

of NF-KB and Jun N-terminal kinase (JNK) by both receptors but is not involved in

apoptosis (161,464). RIP, a cytoplasmic serine/threonine protein kinase, strongly

interacts with TRADD in the receptor-signaling complex and activates NF-KcB and JNK

pathways (159, 232). However RIP disruption (RIP/-) affects only NF-KcB but not JNK








activation (190). Thus is seems that TNF-RI signaling of NF-KB is mediated via RIP

while TRAF-2 enables JNK activation.

TNF-RII is a 75 kDa receptor that is co-expressed with TNF-RI predominantly in

endothelial and blood cells and is involved in passing bound TNF to TNF-RI for

signaling (151, 403). In addition, TNF-RII can directly signal NF-KB activation (152).

TNF-RII has been found to associate with a number of proteins including TRAF-1,

TRAF-2, cellular inhibitor ofapoptosis protein (clAP) -1 and cIAP-2 (343). While

TRAF-1 and -2 can form heterodimers, only TRAF-2 contacts the receptor directly (345).

Similarly clAP-1 and -2 interact indirectly with the receptor by associating with TRAF-1 I

and -2 via their N-terminal BIR (baculovirus IAP repeat) domain (343). Cellular IAPs

have been shown to inhibit the induction of apoptosis by interacting with caspases-3 and -

7 directly as well as target these proteases to degradation by ubiquitination (163, 346).

Cellular lAPs have also been reported to direct the ubiquitination of TRAF-2 (227). It is

unclear at present if clAPs have any role in the activation of NF-kB directly. Expression

of TRAF-2 is sufficient to induce NF-rB and JNK activation under certain conditions

(232,344).

A small subset of TNF binding proteins are the soluble TNF receptors that

represent the extracellular cytokine binding domains that have been shed from the cell

surface by TACE mediated proteolytic cleavage (76, 379). The soluble receptors function

to neutralize the biological activity of TNF by antagonizing the binding of cellular

receptors (94, 179, 364, 379, 406). The outcome of TNF activity depends on the

composition of the signaling complex and can either lead to the induction of apoptosis or

activation of JNK and NF-kB pathways that function to mediate cell survival (232).









In summary both TNF-a and -P are able to bind TNFR and signal via DD

containing proteins initiating a cascade of events leading to the induction of apoptotic cell

death whereas events leading to IKB phosphorylation results in the translocation of NF-

KB to the nucleus where it can induce transcription of genes involved in cell survival and

the proinflammatory response. In addition TNF-oa also causes neutrophil and

macrophage activation and induction to synthesize nitric oxide (NO). TNF-a also

induces the synthesis of acute phase proteins such as C-reactive protein that opsonize

microbes and activate the complement pathway. Mannan-binding lectin (MBL), a PRR

that opsonizes microbes and activates complement is up regulated by TNF-ot. Like IL-

13, TNF-a is also a fever-inducing agent. The important role of TNF, in the innate

immune system and specifically in controlling viral infections, is evidenced by pathogens

such as poxviruses encoding molecules that interfere with TNF function. Cowpoxvirus

encodes five different TNF receptor family homologs that will be discussed in more

detail later.



lnterleukin-18 (IL-18)

IL- 18 was originally discovered as the factor that induced interferon gamma

(IFNy) thus being critical for the control of intracellular pathogens (296). IL-18 is a

potent inducer of CTL and NK cell activity and promotes the differentiation ofT cells to

the THI phenotype (396). While the induction of IL-18 is not entirely clear, it is seen to

be involved in an immune prophylactic response to a number of bacterial and viral

infections (35, 318). Like prolL-13, prolL-18 does not contain a signal peptide sequence,

is intracellular and is processed by caspase-1 (122, 133). The precursor is synthesized as








a 24 kDa peptide that is cleaved at Asp36 to generate the mature 18 kDa cytokine (7).

There exist alternative processing mechanisms that will be discussed later. Unlike IL-1 P,

prolL-18 is constitutively expressed by a variety of cells including Kupffer cells,

macrophages, osteoblasts, keratinocytes, dendritic cells, astrocytes, microglial cells, T

and B cells (99, 285). A molecule purified from urine was shown to bind IL-18 (IL-

18BP) and has been the focus for potential immunotherapy (6,292). IL-18BP was shown

to inhibit the IL-18 induction of IFN y. Interestingly a number ofpoxviruses also encode

a soluble vIL- 18BP underlining the importance of IL- 18 mediated antiviral response (52).

IL-18BP prevents IL-18 from binding its receptor and thus functions as a soluble decoy

receptor to regulate IL- 18 activity.

The 60 to 100 kDa receptor for human IL-18 was originally identified as the IL-1

receptor related protein (IL- 1 Rrp) and is now termed IL-1 8R/IL-18Ra has low binding

affinity for IL-18 (308, 416). Other studies have shown that T cells exhibit high affinity

binding for IL-18 suggesting the presence of a second molecule involved in the binding

complex (466). Subsequently an accessory protein-like (AcPL) receptor that is required

for high affinity binding of IL-18 and transduction of signal was identified (44).

Following binding of IL-18 to IL-1 8R, AcPL also termed IL-1 8R accessory protein

(AP)/IL- 18RP3 is recruited to the receptor complex where it forms a high affinity

heterodimer (336, 369). IL-18R3 does not bind IL-18 by itself and does so only after IL-

18 is bound by low affinity to IL-18Ra (44). Although proteolytic processing of the

extracellular receptor domains can result in soluble forms of both molecules, only soluble

IL-18Ra (s IL-18Roa) binds IL-18 it does not neutralize the effect of IL-18. The

enhancement of affinity for IL- 18 binding to IL-I 8Ra is supported by recent evidence









showing that soluble IL-1 8Rca only in combination with sIL-18Rp3 is able to block IL-18

induction of IFNy (336). Thus the regulation of IL-18 by soluble forms of the receptors

is similar to that seen in IL-1.

The effects of receptor binding and signal transduction are similar for both IL-18

and IL-1. IL-18 signaling complex (comprising of IL-18 bound to IL-18Ra and IL-

18RP3) recruits IRAK via the adapter protein MyD88 (340, 411). MyD88"/- cells do not

respond to IL-18 (1). Upon activation, IRAK dissociates from the receptor complex and

is responsible for the activation of NIK mediated by TRAF-6 (200, 411). The signalsome

containing IKB kinases (IKK) is activated by NIK directly or alternatively by TRAF-6

activation of MEKK1 and ECSIT (42,202, 220). The complex interaction between a

number of protein kinases culminates in the activation and translocation of transcription

factors NF-KB and AP-1 (25, 248, 340). The steps leading to the activation of NF-KB

and AP- 1 seem to be identical both in IL- I3 and IL-18. IL-18 is also involved in

activating the MAPK pathway. IL-18 induced the kinase activities of MAPK and LCK in

THI cells (419). MAPK is involvement in cell growth could possibly explain the effect

of IL-18 in activating NK and T cells (415). Apart from inducing IFNy production from

T, B and NK cells, IL-18 acts in synergy with IL-12 to induce NK cell cytotoxicity and

THl cell differentiation (35). Signaling pathways used by IL-18 and IL-12 are quite

different. While IL-18 activates NF-KB, IL-12 activates the signal transducer and

activator of transcription (STAT) 4 (25). The lack of IFNy production by IL-18 observed

in TH2 cells may be due to the fact that TH2 cells do not express IL-1 8R (466). In

addition to IFNy, IL-18 is a potent inducer of the chemokines MIP- lac, MCP-1, IL-8, as

well as IL-lP3 and TNF-a production in different cell types (321). The ability of IL-18 to









induce an innate response as well as stimulate the production of a number of

inflammatory cytokines underscores its importance as a proinflammatory mediator.

Both IL- 18 and IL- 18R have been targeted for gene disruptions. IL- 18 deficient

mice have impaired IFNy production, defective NK cell activity and reduced THI

response (399). The importance of IL-18 in controlling intracellular pathogens was

demonstrated in IL-18-/' mice infected with the protozoan parasite Leishmania that were

highly susceptible to infection compared to wild type controls (445). Infections with the

extracellular bacteria Stapylococcus aureus caused reduced septicemia whereas these

mice developed more severe septic arthritis, a reaction associated with impaired THI

response (445). Receptor knockout mice have also been characterized. IL-18"-' cells

from these mice had defects in NK cytolytic activity, IFNy production and were impaired

in NF-KB or JNK activity (157). The IL-1 8R-/ mice also failed to respond to IL-18

indicating the essential role of this molecule in mediating signaling.

The availability of mature IL-18 under diverse conditions can determine the

outcome of the immune response. While IL-18 is processed to maturity by caspase-1 at

Asp36, there also exist alternative processing mechanisms. In addition to caspase-1,

caspases-4 and -5 have also been shown to process prolL-18 correctly (122). Caspase-3

on the other hand processed IL- 18 at Asp71 and Asp76 resulting in polypeptides with no

significant biological activity (7, 133). Proteinase-3 (PR-3), a serine protease present in

neutrophil granules has also been shown to process prolL-18 (99, 389). Both membrane

bound and soluble forms of PR-3 enzyme have been identified. PR-3 has been associated

with a number of inflammatory diseases and has recently been implicated as one the

major autoantigens in Hepatitis C virus infection (47, 99, 348, 453). Interestingly PR-3









can also mediate the induction of apoptosis in HL-60 cells that is independent of caspase

activity and loss of mitochondrial membrane potential in spite of being implicated in the

cleavage and activation of caspase-3 (453). Of interest will be future studies on

understanding the biological activities of PR-3 since it seems to play a role in both

inflammation and apoptosis.

In summary, IL- 18 induces the production of inflammatory cytokines (TNF-oc,

IL-113 and IFNy), growth factors (GM-CSF) and chemokines (IL-8, MCP-1, MIP-la and

MIP-1 3). The biological activities associated with IL-18 drive a THI response, enhance

NK cell cytotoxicity and induce macrophage activation. Thus IL-18 plays a crucial role

in the innate immune response against pathogens such as viruses. The fact that

poxviruses such as cowpoxvirus encode molecules to bind IL-18; thereby disrupting

cytokine function, indicate the importance of IL- 18 as an antiviral agent and mediator of

inflammation.



Interleukin-8 (IL-8)

IL-8 was originally identified as a neutrophil chemo attractant from supernatants

of activated monocytes (359, 437,467). IL-8 is a 7 tolO kDa protein that belongs to the

CXC family of chemotactic cytokines termed chemokines and is now called CXCL8.

Chemokines are classified based on the position of the first two N-terminal cysteines that

can be separated by one amino acid (CXC) or can be adjacent (CC) (22). Disulphide

bonds between the first and third Cys as well as the second and fourth Cys are important

for three dimensional folding, receptor recognition and biological activity. IL-8 is

synthesized as a 99 amino acid precursor that is proteolytically processed by the removal








of a 20 amino acid leader sequence (67). Maturation of IL-8 occurs extracellularly with a

number ofproteases being implicated in the cleavage process. Two forms of mature IL-8

have been observed to be secreted from activated neutrophils IL-8 [77] and IL-8 [72]

containing 77 or 72 residues respectively which can be further processed by neutrophil

elastase, cathepsin G and proteinase-3 to give N-terminal truncated forms that are more

active (302). Neutrophil gelatinase B, a secreted matrix metalloproteinase (MMP-9) has

also been shown to cleave mature IL-8 [77] to more active truncated forms (431). Thus

inflammatory exudates containing proteases can potentiate more active forms of IL-8.

Near sites of production, chemokines can form oligomers on endothelial or

extracellular matrix, thereby forming a gradient with higher concentrations of chemokine

near the inflammatory stimulus (298). In addition to attracting neutrophils, IL-8 has also

been shown to have chemotatic effects on a number of other cells including basophils,

eosinophils, T lymphocytes and macrophages (366). The ability of IL-8 to attract

neutrophils has been disputed by both in vitro and in vivo experiments involving

intradermal injections in humans (218, 222, 405). In addition to being secreted by

neutrophils, IL-8 is also secreted by keratinocytes, fibroblasts, endothelial cells and T

cells (310). IL-8 has been demonstrated to have mitogenic effects (353). Subsequently

constitutive IL-8 expression has been observed in many types of human cancers (366).

IL-8 has also been shown to induce angiogenesis (199).

Although monomeric IL-8 is biologically active, IL-8 can also exist as a dimmer

with the disulphide bonds playing important roles in anchoring the cores of the dimmer

together (70). N-terminal residues Glu4, Leu5 and Arg6 (ELR) have been shown to be

absolutely essential for receptor binding (67, 144). The ELR motif is found to be present








only in chemokines that attract neutrophils (22). Synthetic IL-8 containing modifications

of the ELR sequence can function as receptor antagonists that block IL-8 action

suggesting that N-terminal truncations that occur naturally during inflamed states could

contribute to regulation (268). Two IL-8 receptors have been identified, the CXCR1 and

CXCR2 (154, 280). The receptor more selective for IL-8 is CXCR1 (269). Chemokines

exert their function via heterotrimeric seven transmembrane G-protein coupled receptors

(GPCRs) (22). The functions of the two receptors are also different in certain aspects.

While both receptors induce cytosolic free Ca2 and degranulation, CXCR1 mediates

respiratory burst and the activation of phospholipase D (181).

Initial studies on chemokine signaling were performed in human neutrophils

stimulated with IL-8 (22, 298, 310). The functional responses were found to be similar

for all chemokine receptors. Following chemokine binding the receptor, there is a

conformational change in the chemokine at the N-terminus with disulphide bonding

between Cys7 and Cys34 (1st and 3rd Cys bond) playing a major role in receptor binding

and activation (325). The conformational change facilitates further interaction with the

receptor that leads to activation. Activation leads to the exchange of bound GDP for GTP

in the G-protein. GTP bound ca subunit and P3y dissociate. The signal transduction

enzymes phospholipase C (PLC) and phosphatidylinositol-3-OH kinase (PI3K) are

activated by P3y subunit. PLC mediates cleavage of phosphatidylinositol (4,5)-

bisphosphate yielding inositol triphosphate (IP3) and diacylglycerol (DAG). IP3

induction results in release of intracellular stores of calcium giving rise to a transient

increase in free calcium concentration. DAG activates protein kinase C (PKC). PI3K

also generates phosphatidylinositol triphosphate and activates protein kinase B (PKB).









The Py subunit also brings about activation of MAP kinase and then re-associates with

GDP bound a subunits. This complex activation of kinase pathways results in the

activation of transcription factors that up regulate the expression of cell adhesion

molecules, growth factors and inflammatory cytokines that facilitate the motogenic and

mitogenic effects of IL-8.

In neutrophils, the transient increase in cytosolic free calcium and activation of

PKC is responsible for granule release and superoxide production (23). The activation of

GTP'ases results in cytoskeletal restructuring (40, 219). CXCR1 mediated activation of

phospholypase D also leads to superoxide formation in human neutrophils (181, 211).

Activation of map kinases leads to the up regulation of transcription factors (180, 414).

Binding of ligand to CXCR2 can also lead to receptor desensitization mediated by PKC

followed by receptor endocytosis and degradation, a mechanism thought to regulate IL-8

(278,279, 461). While the activation of transcription factors AP-1 and NF-KCB by TNF-a

and IL-I 3 up regulate IL-8, expression of IL-8 is inhibited by interferons a and 13 (262,

371). Since IL-1 3 and TNF-a are elevated during inflammatory states and the fact that

IL-8 is synthesized by a number of cell types, highlights its important in causing local

accumulation of neutrophils in any tissue.

In summary, IL-8 functions as a chemotactic factor for leukocytes increasing

access of immune cells such as neutrophils to sites of inflammation. IL-8 enhances

adhesion molecule binding (132 integrin) to circulating leukocytes thereby facilitating

migration and extravasation of the immune cells across blood vessel walls into inflamed

tissue. IL-8 activates neutrophils to undergo respiratory burst and degrannulation. In








short, IL-8 is particularly important in the innate immune response by directing

neutrophil migration to sites of infection.



Avian Cytokines

A number of homologs for mammalian proinflammatory cytokines have been

identified in birds. A review of avian and mammalian cytokines reveals that even though

they function similarly, some of their primary structures seem to differ significantly.

Avian homologs of TNF-a, IL- 3P, IL-18 and a number of CXC chemokines have been

identified so far. A majority of these molecules were discovered in chickens, in addition

to duck, turkey, pheasant, quail and guinea fowl (384).



TNF-a

TNF-a-like activity can be detected in chick embryos as early as day 1 (452).

The pattern of expression throughout the growing embryo suggested a role for TNF in

apoptosis during differentiation of the neuronal system as well as in tissue remodeling.

Indeed cells undergoing DNA fragmentation associated with apoptosis were

demonstrated to be closely associated with TNF-a immunoreactivity (451). Chicken

macrophages stimulated with bacterial LPS or the intracellular protozoan Eimeria,

secreted a TNF-like activity that was preferentially cytotoxic to chicken fibroblasts

indicating its role in bacterial and coccidial infections (471). Partial purification of

stimulated macrophage supernatant fractions containing TNF-like activity revealed a 17

kDa protein that cross reacted with antibodies to human TNF-a (329). Although the

partially purified protein induced biological activities related to TNF-ca in chicken









macrophages, it was not blocked by antibodies to human TNF-a suggesting structural

differences and/or the presence of other factors in the preparation. Chicken TNF-RI has

recently been cloned and was found to be 47% identical to its human counterpart in the

cytoplasmic death domain (46). In addition a number of death receptors (DRs) related to

TNF-RI have been identified in chickens that are yet to be fully characterized (46).



IL-1

Chicken IL-1 activity was first seen in LPS stimulated spleenocyte supematants

(140). Chicken macrophages treated with LPS also induced secretion of IL-1 activity

(41). The induction of IL-1 was seen to be dependent on protein kinase C and a

calmodulin dependent kinase. Other studies have also shown that IL-1 secretion by

macrophages is dependent on intracellular calcium levels (65). IL-1 also lowered growth

rates in chicken similar to the effects seen with inflammatory agent Sephadex (195). The

chicken IL- IP gene was cloned rather recently and a peptide containing residues 106 to

267, the putative mature IL-P sequence, was shown to be biologically active both in vitro

and in vivo (446).

Unlike the mammalian system where maturation of IL-1 P requires caspase-1 to

process the inactive precursor initially at Asp27 (site 1) followed by cleavage at Aspl 16

(site 2), the chicken proIL- 113P sequence does not contain either of the conserved Asp

residues required for cleavage that would release the mature cytokine (395, 446). The

mechanism of IL-1 P13 maturation in chickens is currently unknown. It remains a

possibility that in chickens, IL-1 13 does not require processing and maybe active as the

precursor itself. The chicken proIL- 113 sequence is similar to proIL- 113 sequences from








frog and fish, which also do not contain the conserved Asp sites found in mammals (475).

Matrix metalloproteinases (MMPs) have also been implicated in the processing of

mammalian proIL-1P (355). Gelatinase A (MMP-2) and B (MMP-9) are among the

MMPs shown to cleave prolL-13. An avian gelatinase that shares homology with MMP-

2 is observed to be over expressed in chicken embryo fibroblasts transformed with Rous

sarcoma virus (64). In addition, a chicken enzyme similar to MMP-9 has been cloned

from macrophages that differ from its mammalian counterpart in sequence and

biochemical function (137). In light of alternative mechanisms of processing observed

for mammalian IL- 1p, it is worth noting that chicken heterophils treated with LPS,

zymosan or phorbol myristate acetate (PMA) show increased MMP production (328).

Although there seem to avian homologs to mammalian enzymes that have been

implicated in IL- l3 maturation, there is no experimental data to prove that any of the

avian enzymes actually function to process avian IL- 1P3.

The chicken IL-1RI was cloned from chicken embryo fibroblasts (136).

Interestingly the cytoplasmic region of chicken IL-1RI is nearly 80 % homologous to

both its mammalian counterpart as well as the Drosophila Toll receptor. Like other TLRs

the chicken receptor also has an extracellular Ig-like region although sequence

divergence is higher in this region. Soluble chicken IL-1RI was shown to inhibit IL-1-

like activity released from LPS induced macrophages confirming an antagonistic role for

this molecule (196). The chicken homologs of IL- la, IL-1 RII or IL- IReceptor

antagonists are yet to be identified.









IL-18
The chicken IL-18 gene was discovered as a result of BLAST search analysis of a

bursal expression sequence tag (EST) database (354). Comparison of the primary

sequence between the chicken and mammalian homologs revealed that the chicken

precursor contains the conserved Asp36 that is required for processing by caspase-1 (7).

A putative peptide representing residues 30 to 198 expressed with an N-terminal

histadine tag was found to induce IFNy synthesis in primary chicken spleen cells. The

IL-18 receptor and accessory signaling protein as well as the antagonist IL-18 binding

protein are yet to be identified in birds.



CXC chemokines

The chicken chemotactic and angiogenic factor (cCAF) is a 9 kDa CXC

chemokine that was originally cloned from Rous sarcoma virus transformed fibroblasts

(29, 388). Recent studies have suggested it to be a homolog of IL-8 (183). Similar to IL-

8, the chicken cytokine can also be post translationally cleaved (by plasmin) to smaller

fragments (243). Experimental evidence points to the C-terminus as being angiogenic

(243). K60 is another chicken CXC chemokine that shares 50% homology to mammalian

IL-8 (367). K60 is secreted from macrophages stimulated with LPS, IL-1l and IFNy

(367). Interestingly both cCAF and K60 are the only chicken chemokines to date that

contain the conserved ELR motif present in all mammalian chemokines that attract

neutrophils (22). A homolog of mammalian CXCRI was cloned from a chicken genomic

library (226). The chicken receptor was found to be 67% homologous to human CXCRI.

The identification of mammalian homologs of IL-8 and related CXC chemokines in birds

demonstrate the similarities between the chemotatic systems.









Viral Inhibitors of Inflammation

The outcome of viral infections depends on both host and viral factors. The early

resolution of infection is the goal of the mounted innate immune reaction. Not

surprisingly, viruses have been shown to encode a number of molecules that interfere

with the early host innate response. Viral products that specifically function to abrogate

proinflammatory cytokines and their effects will be discussed here. Cytokine synthesis is

a typical early host response to viral infection. Different viruses target the same cytokine

pathways underlining their importance as potent anti-viral agents and these include TNF-

a, IL- 1 P, IL- 18, the interferons, complement as well as a number of chemokines. The

strategies employed by viruses function to disrupt cytokine transcription/up regulation,

interfere with receptor binding and modulate cytokine signaling.



TNF Inhibitors

Encoding decoy receptors that bind TNF is a mechanism employed by a number

of poxviruses and herpesviruses. Shope fibroma virus encodes the T2 protein that has

been shown to be a TNF receptor homolog that binds both TNF-a and P3 (249). A

homolog to the T2 protein was also found in Myxoma virus (427). Deletion of Myxoma

virus T2 had no effect on growth of the virus in tissue culture but significantly attenuated

viral lethality in infected rabbits. Cowpoxvirus encodes five different TNF receptor

family homologs CrmB, CrmC, CrmD, CrmE and a soluble CD30 homologue. The

CrmB protein shares homology to TNFRII in the ligand binding region and can bind both

TNF-a and 3 (162). CrmC also shares homology to TNFRII but only binds TNF-a

(375). CrmD found only in some strains of cowpoxvirus and ectromelia virus was able to









bind both TNF-a and P and block their cytolytic activity in vitro (233). In addition to

being present in cowpoxvirus, CrmE is also present in other poxviruses such as vaccinia

virus (352). The cowpox CrmE was also found to only inhibit TNF-ca activity. More

recent work has shown the vaccinia CrmE protein to also bind TNF-a but not P3 (333). In

this study CrmE was found to inhibit the cytotoxic activities of human TNF-a but not

those of mouse or rat. Vaccinia strain USSR deleted for CrmE was attenuated in an

intranasal mouse model of infection. But vaccinia WR strain, engineered to either

express the USSR CrmE, or CPV CrmB or CPV CrmC or CPV CrmE genes displayed

enhanced virulence indicating the influence of viral TNFRs on disease outcome. The

CD30 molecule is a member of the TNF receptor family and is associated with TH2

responses although its exact function role is unclear at present (31). The important role

of CD30 in viral infection is evidenced by the fact that cowpoxvirus encodes a secreted

CD30 homologue. Another virus that interferes with TNF is the herpes human

cytomegalovirus that encodes UL144, which also belongs to the TNFR family (30).

While the ligand that binds UL144 is yet to be identified, its importance in viral infection

is demonstrated by the presence of antibodies to UL144 in patient sera.



IL-1 Inhibitors

Members of poxviruses also encode molecules that interfere with IL- 113 either in

its production or signaling. The subject of this dissertation work is the cowpoxvirus

CrmA protein that will be discussed in more detail later. Briefly, CrmA is believed to

inhibit the processing of proIL-113 by blocking the activity of the processing enzyme,

caspase-1 (331). The importance of IL-1 3 involvement in viral infection is further









evidenced by the expression of soluble IL- I3 receptors encoded by poxviruses. The

vaccinia virus B1 5R protein based on sequence was proposed to function as an IL-1

binding protein (376). Indeed B15R was found to bind IL-113 but not IL-1-at or IL-IRa

(8,381). Viruses lacking B15R were found to be less virulent than wild type when

injected into weanling mice via the intracranial route. Similar experiments in mice

infected via the intranasal route with virus disrupted for B15R accelerated the illness and

mortality indicating a role for IL-1 in modulating virulence. Furthermore B15R was also

found to prevent fever caused by virus infection in mice. When the non-functional B15R

gene in Copenhagen, a virulent strain of vaccinia, was repaired, the resulting virus

infection suppressed fever and attenuated viral virulence (10). Recent evidence shows

that poxviruses can also interfere with signaling events downstream of TNFR and IL-1 RI

such as NF-KB activation although observed responses seem to be specifically different

for each virus tested (295).



IL-18 Inhibitors

The central role of IL- 18 in inducing IFNy that protects against bacterial and viral

infection is well documented (318). Interestingly a soluble cellular IL-18 binding protein

(IL-1 8BP) was identified and found to be more homologous to putative poxviral proteins

than to mammalian IL-18 receptors (292). In the cellular context IL-18BP functions as a

decoy receptor to bind IL-18 preventing its activity. Viral IL-18BPs (vIL-18BP) have

been identified in vaccinia, cowpox, ectromelia, variola and swinepox viruses (292). In

addition Molluscum contagiosum virus was found to contain three open reading frames

with homology to IL-18BP (457). Two of the Molluscum gene products MC53L and









MC54L bound both human and murine IL-18 with high affinity and prevented IL-18

mediated IFNy induction in KG-1 monocytic cells. IL-18 binding activity was found in

culture supematants of a number of vaccinia strains as well as cowpox and ectromelia

viruses (377). Furthermore vIL-18BP of ectromelia virus was found to inhibit NF-KB

activation and IFNy induction in KG-1 monocytic cells in response to IL-18. Cowpox

and ectromelia viruses are rodent pathogens and not surprisingly their vIL- 18BPs

preferentially bound mouse IL-18 over human IL-18 (52). Mice infected peritonealy

with ectromelia deleted for vIL-18BP had increased NK cell cytotoxic activity while the

mutant virus was found to be attenuated (43). IL-18 activity clearly is an important host

response mechanism seeing that a number of viruses block its activity.



Chemokine Inhibitors

The number of mechanisms employed by different viruses to subvert chemokine

activity indicates the importance of this response to control viral infection. Similar to

counter strategies used against cytokines, different herpes and poxviruses encode

molecules that regulate chemokines. Typically these fall into three categories. First are

molecules that are ligands or viral homologs of cellular chemokines. These include

human herpes virus proteins such as vMIP-I, vMIP-II, vMIP-III, that are homologs to

MIP-I a or P3 and function as agonists or antagonists (213,264, 385). Cytomegalovirus

proteins vCXC-1 and -2 as well as Marek's disease virus vIL-8 are homologs of IL-8

(307, 312). Interestingly vIL-8 was found to chemoattract blood mononuclear cells but

not heterophils. Molluscum encodes two proteins that act as antagonists to MCP-I. The









MC148R molecule is able to bind receptors but does not transmit signal as it lacks a

region critical for receptor activation (78, 206).

The second group of viral proteins is chemokine receptor homologs. Virus

encoded chemokine receptor homologs are typically cell membrane bound and function

to bind chemokines. These include the ORF74 protein found in a number of herpes

viruses that binds CXC chemokines (342). The US28 gene ofcytomegalovirus is a

functional CC chemokine receptor (36). In addition this virus also encodes US27, UL33

and UL78 proteins with homology to CC chemokine receptors some of which have been

found to localize on viral membranes (109, 240, 330). Swinepox virus encodes the K2R

molecule that is potential receptor of CXC chemokines (246) (Traci Ness and Richard

Moyer unpublished). Capripoxvirus also contains a putative CC chemokine receptor the

Q2/3L ORF (54). Molluscum encodes MC148, a MIP-1 P3 homolog that was shown to

antagonize the chemotactic activity of a number of CC and CXC chemokines on different

cell types (78).

The third mechanism employed by viruses is to secrete binding proteins that

sequester chemokines and prevent chemotaxis of inflammatory cells to the site of

infection. Examples of these would include the myxomavirus M-T7 protein that was

originally discovered as an IFNy receptor but has subsequently been found to also bind C,

CC and CXC chemokines (214,429). The M-T1 gene product of myxomavirus and

rabbitpoxvirus secreted 35 kDa proteins function similarly in binding CC chemokines

with high affinity and IL-8 (CXC) with a lower affinity but share limited homology (129,

216). In vivo studies indicate that M-T1 deletions in myxoma do not affect virulence but

the absence of M-T1 increased infiltrating monocytes and macrophages at the site of









primary infection (215). More recently M-T1 was found to differ from rabbitpoxvirus 35

kDa protein in that M-T1 I can localize to cell surfaces through interactions at the C-

terminus of the molecule and heparin (363). The breadth of viral modulators of

chemokines suggests that chemotaxis is an important host response to virus infection.



Inhibitors of Complement

Activation of complement pathways can itself lead to the destruction of virus-

infected cells as well as mediate the release potent inflammatory cytokines and is another

mechanism targeted by virus infections. Mammalian regulators of complement

activation function to protect bystander cells from the effects of activated complement.

These regulators are characterized by homologous motifs termed short consensus repeats

(SCRs) (335). The vaccinia complement control protein (VCP) contains four SCRs and

regulates complement by binding to C4b and C3b (204, 251). VCP acts as a cofactor to

the cellular complement regulator serine protease factor I and aids in the cleavage of C4b

and C3b to inactive components (349). The cowpox VCP homolog called

immunomodulatory protein (IMP) was also shown to be functional in inhibiting tissue

damage due to complement at the site of viral infection both in the footpad and

subcutaneous models of mice injections (158, 258). More recently the variola homolog

of VCP termed smallpox inhibitor of complement systems (SPICE) was found to be

nearly 100 fold and 6 fold more potent at inactivating human C3b and C4b respectively

than its vaccinia counterpart (341). Glycoproteins C (gC) of human herpes virus 1 and 2

have also been found to bind C3 components of complement (203). In addition the

herpesvirus saimiri CCPH gene product prevents complement mediated cell damage and









has been found to bind complement component C3d (108). Clearly inhibiting

complement activation and lysis of infected cells allows viruses more time to replicate

and spread.



Inhibitors of Interferons

In addition to the blocking the mentioned immune mechanisms, viral products

also mediate interferon activity and will be briefly discussed here. Poxviruses encode a

number of proteins to bind both interferon type I and II. A soluble type I IFN receptor

has been identified as the vaccinia B18R open reading frame (72, 397). B18R has high

affinity for IFNca of different species, can also be found membrane bound and its absence

from vaccinia attenuated infections in the murine intranasal model. The myxoma M-T7

products are soluble type II IFN receptors that sequester IFNy (274). A unique feature of

M-T7 is that in addition to binding IFNy it also binds IL-8 and other members of the C,

CC and CXC chemokines (214). Soluble IFNy receptor homologs appear to be present in

ectromelia, cowpox and rabbitpox viruses as well (275). Indeed the vaccinia B8R protein

binds IFNy from a different species (9). Another unique molecule is the 38 kDa

glycopeptide secreted by tanapox virus (95). Unlike any other protein known to date, the

38 kDa protein binds IFNy, IL-2 and IL-5 demonstrating yet another way viral proteins

disarm multiple immune pathways with the same molecule. There also exist numerous

protein molecules encoded by different viruses that regulate the intracellular activities of

IFN that will not be discussed here. Blocking IFN and its activity is vital for virus

survival as evidenced by numerous viral strategies discovered to date.









Apoptosis

Apoptosis is a crucial step necessary for the clearance of infected cells and

activated neutrophils, to protect areas adjacent to the inflammatory tissue from injury.

Apoptosis or programmed cell death is a distinct mechanism by which cells die and are

dismantled in an orderly fashion (191). This process is a distinct morphological and

biochemical function that involves cell shrinkage, membrane blebbing, degradation of

proteins, chromatin condensation, DNA fragmentation followed by disintegration of the

cell into multiple membrane enclosed vesicles termed apoptotic bodies (454). The

disassembled cell is finally phagocytosis by its neighbors. This form of cell death differs

from necrosis where cells typically lyse by loss of membrane (191). Apoptosis is part of

normal homeostatic mechanism whereby cells die during embryonic development, tissue

turnover and as a means of defense against pathogens. Metazoan development is

characterized by the over production of cells followed by apoptotic induction to cull the

excess cells at later stages of development such that the relative number of cells required

to achieve proper function is maintained. This phenomenon is first characterized in the

nematode Caenorhabditis elegans by Horvitz and others (468). Processes of assembly

and disassembly constantly maintain the normal human body's approximately 1014 cells.

It is therefore not surprising that inappropriate cell death (either too much or too little)

can give rise to disease conditions. Excessive apoptosis has been implicated in

neurodegenerative disease such as Alzheimer's (242, 387) and in tissue destruction that

occurs after vascular occlusions in the brain and heart (34, 62, 286). Similarly the lack of

apoptosis contributes to the development of tumors (2) and autoimmunity such as

autoimmune lymphoproliferative syndrome in humans (107, 339,439).









Apoptosis can be induced by a number of stimuli such as radiation, growth factor

depravation, receptor-mediated signaling and viral infections. There are basic pathways

by which a cell can be signaled to initiate apoptosis as shown in Figure 2. One pathway

is extrinsic, mediated via receptor-ligand interactions specifically the receptors related to

the tumor necrosis factor (TNF) and nerve growth factor (NGF) families including

TNFR1, TNFR2 and CD95/APO-1/FasR (147). These specialized cell surface receptors

are termed 'death receptors' (DR) (147). Ligand-receptor interaction results in receptor

oligomerization that causes specific cytoplasmic proteins to be recruited to the receptor

complexes. These cytoplasmic "adapter" proteins such as TRADD and FADD, then

interact with a group of cellular proteases called caspases. The activation of caspases

leads to the breakdown of cellular homeostasis that culminates in cell death.

Alternatively the second or intrinsic signaling pathway involves mitochondrial

dysfunction (Fig. 2). External cues such as loss of growth factor signals converge with

internal cues such as DNA damage and trigger the permeablization of the outer

mitochondrial membrane and leakage of mitochondrial pro-apoptotic effectors into the

cytoplasm (147). In addition the cytotoxic granule, granzyme B, a serine protease

produced by CTLs and natural killer (NK) cells has been shown to directly activate

caspases as well as the cytoplasmic protein Bid which activates the mitochondrial

pathway (26, 184). The pro-apoptotic stimulus results in the release of an array of

molecules from mitochondrial membranes, which ultimately cause cellular demise that

can be dependent or independent of caspase activation. The activation of caspases is by

far the main biochemical incident responsible for programmed cell death. Thus signaling

through DRs or the disruption of mitochondrial integrity results in the induction of


























Figure 2. Apoptosis. The induction of apoptosis can be signaled extrinsically involving
TNF receptors or Fas. Receptor complexes aid in the activation of initiator caspases such
as caspase-8 and -10 which in a cascade fashion ultimately in results in the activation of
terminal caspases such as caspase-3, -6 and -7. Cellular substrates that are involved in
DNA repair and homeostasis form targets for activated caspases. Apoptotic signals can
also originate from within the cell when mitochondrial integrity is compromised. Factors
normally residing within mitochondrial membranes such as cytochrome c then leak into
the cytosol and can cause activation of caspases. Mitochondrial proteins like AIF and
EndoG can induce apoptosis independent of caspase activation. Granzyme B can directly
activate caspases or can cleave cytosolic protein Bid. Cleaved Bid translocates to the
mitochondria and associates with Bax. Transcription factor p53 regulates bcl and bax
levels. Upregulation of bcl favors cell survival, whereas upregulation of bax favors cell
death. Abbreviations: Apaf-1, apoptosis associated factor-1; AIF, apoptosis inducing
factor; CASP, caspase; Cyt c, cytochrome c; DFF, DNA fragmentation factor; DNA-PK,
DNA protein kinase; FADD, Fas associated death domain; EndoG, endonuclease G;
TNFr, tumor necrosis factor receptor; TRADD, TNFR associated death domain; UV,
ultra violet












"
A^ S
^ |
/ |

00


0


Cu
2
Cu






0
Cu
~


ON


IM-
sn


0
. T-0
1u0


f-f
CM
eg
a


*w
i.^
U


U!









apoptosis. Finally, neighboring cells and/or macrophages circumvent an inflammatory

response by phagocytosing the vesicles or apoptotic bodies formed as a result of

apoptotic induction.



Caspases

Caspases are defined as cysteine proteases that cleave protein substrates after

specific aspartic acid or the P1 residue (15). There are more than 42 mammalian

caspases described to date (217). Of these 12 are found in humans, caspases-1 to -10

(119), caspase-12 (105) and caspase-14 (91). There are also caspases cloned from

diverse organisms such as worms (459, 460, 468), insects (110, 168,365, 380) and hydra

(66).

Caspases are thiol proteases in that they make use of a cysteine side chain

nucleophile to hydrolyze cleavage of the peptide bond. Like other proteases they are

produced as zymogens and maturation involves cleavage of the propeptide after certain

aspartate residues. Caspase precursors contain an N-terminal prodomain that can vary in

length from 23 to 219 amino acids followed by regions encoding a large (p20) and a

small (pl 0) subunit. Sometimes spacer regions can also be present between subunits that

are excised during the maturation process (90).

The association of one large and one small subunit (p20/pl0 heterodimer) gives

rise to the catalytic domain. Each active enzyme is a tetramer composed of two such

heterodimers. The quaternary structure of caspases is unique among proteases and is

described as a homodimer of heterodimers containing two active sites at opposite ends.

While the requirement for an aspartic acid residue at the P1 position of the substrate is an









absolute, the two residues (P2 and P3) towards the amino side of P1 are not absolute but

are important in having an impact on substrate cleavage. The P4 residue on the other

hand appears to be important in determining substrate specificity. Caspases are activated

in a sequential manner beginning with those that initiate the cascade and ending in the

activation of effector caspases that result in the cleavage of the so-called death substrates.

By nature of their cleavage specificity, most caspases are capable of auto processing.

The length and presence of certain motifs on caspase prodomains are related to

their function. Caspases with large prodomains are typically involved in the initiation of

the apoptotic cascade. The death effector domain (DED) found in caspases -8 and -10

interacts with the DED of signaling adapter proteins that are associated with death

receptors involved in extrinsic signaling. Another motif, the caspase recruitment domain

(CARD) found in caspases -1, -2, -4 and -9 functions in promoting interactions between

one another and with adapter molecules. Short prodomain containing caspases are

typically effector caspases and are usually activated by initiator caspases.

Based on substrate specificities, caspases have been divided into three major

groups (119). Group I caspases are called inflammatory caspases since they function to

activate proinflammatory cytokines and include caspases-1, -4, -5, -11 and -12.

Caspases belonging to Group III function as initiator caspases and include caspases -2, -

8, -9, and -10. The initiator caspases cleave and activate Group II effector caspases that

include caspases-3, -6 and -7. Based on optimal substrate sequence specificity, caspase-2

has been placed in group II and similarly caspase-6 has been placed in group III (119).

The dual role of caspases in mediating apoptosis and inflammation indicates that these









two mechanisms are linked and underlines the importance of caspases in the regulation of

these two cellular processes.



Inflammatory Caspases-1, -4, -5, -11 and 12

Caspase-1 is the prototypical member of this group and was the first caspase to be

characterized. It was originally described as the interleukin-1 3 converting enzyme or

ICE (58, 413). In addition to processing proIL-1 to maturation, caspase-1 has also been

shown to process prolL-18 to the mature cytokine (122, 133). The human caspase-1

precursor is a 45 kDa polypeptide that is processed at 4 distinct sites. Initial cleavage

generates the 11.5 kDa prodomain while successive cleavages generate the large and

small subunits (413). The prodomain appears to be required for dimerization,

autoactivation (430) and activation of caspases-3 and -8 (404). Over expression of

caspase-1 has been linked to apoptotic induction in rat fibroblasts, mammalian COS cells

and neuronal cell lines (117, 134, 260). The first caspase knockout mice created was

caspase-1 deficient mice (209, 225). The importance of caspase-1 in the maturation of

IL- 1 3 and IL-18 is seen in the ability of caspase inhibitors to block the formation of

bioactive cytokines (194). Although mice deficient for caspase-1 are defective in

generating mature cytokines, the animals developed normally (209). There are also

reports that caspase- 1' -- mice were able to release biologically active IL- 13 in response to

subcutaneous injection of turpentine indicating the presence of alternate pathways for

prolL- 13 processing (100). Nevertheless the importance of caspase- 1 in inflammation is

supported by evidence showing that blocking caspase-1 activity or deleting the gene

rescues mice from lethal endotoxic shock (132). Caspase-1 knockout mice were also less








susceptible to hypoxic-ischemic brain injury suggesting the involvement of caspase-1 in

neuronal apoptosis (113,230). While the involvement of caspase-1 in processing

proinflammatory substrates is established, the activation of caspase-1 is less well

understood.

There is evidence to show that caspase-11 is an upstream regulator and is

responsible for the activation ofcaspase-1 (441). Although caspase-11 is a murine

caspase, it is thought to be a homologue of human caspase-4 (60% identical) (187). Like

caspase-1, caspase-11 can be induced by lipopolysaccharide (LPS) (440). Caspase-11

knockout mice are defective in processing prolL-103 (441) and apoptotic induction (187).

There is some evidence showing the activation of caspse- 11 by cathepsin B (356). There

is very little information known about the functions of caspase-4 and -5 (101, 186). Like

caspase- 1, they have been shown to process prolL- 1 3 (186) and prolL- 18 (122).

Caspase-5 can be induced by IFNy and LPS (228) and is found to be involved in receptor

mediated apoptosis (207). Of interest is the observation that caspase-5 can cleave the cell

cycle regulator Max at the unusual cleavage site of glutamic acid (207).

Caspase-12 was first identified in rodents and found to be localized in the

endoplasmic reticulum (ER) where it is activated by excessive stress to the ER (465).

Caspase-12 has been linked to neurodegenerative disorders (205) and was shown to

activate caspase-9 (266). Experimental evidence suggests that ischemic injury may

induce the activation of caspase-12 mediated by calpain (a cysteine endopeptidase

activated by intracellular calcium) resulting in apoptosis (284). It has been demonstrated

that procaspase-12 interacts with TNF receptor associated factor 2 (TRAF2) and is

activated in response to ER stress (465). A recent report suggests that catalytic









functionality of human caspase-12 has been lost since it is reported to contain a number

of mutations that prevent it from being expressed as a full length protein (105).



Apoptotic Initiator Caspases-2, -8, -9, and -10

The first caspase related disease condition described in humans is autoimmune

lymphoproliferative syndrome (ALPS) type II, an autoimmune disorder linked to

mutations in the initiator caspase-10 (439). Initiator caspases represent members

involved in both pathways leading to caspase activation namely the receptor mediated

extrinsic pathway involving caspase-8 and the intrinsic mitochondrial pathway involving

caspase-9. DR induced apoptosis involves receptor ligand interactions that aggregate and

act as scaffolds for caspase activation often termed the death inducing signaling complex

(DISC) (90). The recruitment of initiator caspases to the DISC involves the DEDs in the

prodomains of caspase-8 and -10 that are involved in binding DEDs on adaptor

molecules such as FADD in the DISC (438). Similarly the CARD found in caspase-2

and -9 prodomains has been shown to associate with CARDs on adaptor molecules such

as RAIDD facilitating the recruitment of caspase-2 to the DISC (5, 88). The induced

proximity model proposes that presence of high local concentrations of procaspase-8 in

the DISC is sufficient to stimulate the autoactivation of caspase-8 (283). This

phenomenon of autoactivation has been noted for caspase-9 and -10 (382, 438). An

alternative hypothesis is that conformational changes induced by binding adaptor

molecules can cause activation of proinitiator caspases (147). Activated caspase-8 can

either cleave the terminal effector caspases-3 and -7 directly or can cleave the









cytoplasmic protein Bid, which in turn activates the mitochondrial pathway (223).

Similarly activated caspase-10 has been shown to cleave procaspases-3, -7 and -8 (103).

Apoptosis can also be triggered without engaging the DRs, but instead causing

mitochondria to release proapoptotic molecules into the cytoplasm. One of these is

cytochrome c, a protein involved in electron transport, that now accumulates in the

cytoplasm and binds the apoptotic protease activating factor -1 (Apaf-1), a cytoplasmic

protein that shares homology with the C. elegans ced-4 apoptosis regulator. Cytochrome

c stimulates the formation of the mitochondrial apoptosome consisting of cytochrome c,

Apaf-1 and caspase-9. The binding of procaspase-9 to oligomers of Apaf-1 occurs

through the CARD regions of these two proteins. Like caspase-8, caspase-9 also

undergoes autoactivation induced by proximity of high concentrations of the proenzymes.

Caspase-9 processes the effector caspases-3 and -7 directly and has been shown to be the

apical caspase activated in the mitochondrial pathway responsible for the activation of

caspases-2, -6, -8 and -10 (374). Convergence of the DR and mitochondrial pathways to

cell death occurs as a result of caspase-8 mediated cleavage of cytoplasmic protein Bid,

which facilitates the release of cytochrome c from the mitochondria. This is probably as

a result of a feedback amplification loop since caspase-3 mediates activation of caspases-

2, -6, -8 and -10 following activation by caspase-9 (374).



Apoptotic Effector Caspases-3, -6 and -7

Caspase-3 which was first discovered as a gene product with high homology to

the C. elegans CED-3 protein and which caused apoptosis when overexpressed in Sf-9

insect cells (104). Subsequent work showed caspase-3 to be necessary for the cleavage








of poly (ADP-ribose) polymerase (PARP) an enzyme involved in DNA repair that is

proteolytically cleaved at the onset of apoptosis (289, 407). Caspase-3 is considered to

be the prototypic member of the effector caspases that function as the final executioners

of the apoptotic process (90). While the effector caspases can autoactivate, it has been

suggested that upstream caspases serve as the main activators. The initiator caspases-8, -

9 and -10 have all been found to cleave and activate procaspases-3 and -7 (90), while

caspase-3 has been shown to activate caspase-6 suggesting that this protease acts

downstream of caspase-3 (148). Caspase-3 can also be activated by caspase-1 (407),

caspase-4 (185), and caspase- 11 (187). The involvement of cytotoxic T cell mediated

apoptotic induction is demonstrated in the ability of the serine protease granzyme B to

directly activate caspses-3, -6 and -7 (103, 299).

Effector caspases target substrates that are essential in DNA repair pathways,

maintaining homeostasis and structural elements of cells. In the DNA repair pathway

targeted proteins include PARP, cleaved by all effector caspases, the catalytic subunit of

DNA dependent protein kinase (DNA-PKes) and the 70 kDa protein component of Ul

ribonucleoprotein (U 1-70 kDa) that are cleaved by caspse-3 (56). The characteristic

DNA ladder formation seen in isolated apoptotic nuclei can be induced by caspase-3,

which is known to directly activate caspase activated deoxyribonucease (CAD)/ DNA

fragmentation factor (DFF) (231) and also by caspse-7 (300). In addition, other

substrates include DNA topoisomerase I & II, RNA polymerase I upstream binding factor

(UBF) (57)

The induction of apoptosis gives rise to the cleavage of molecules that affect at

least three protein kinase pathways. Caspase-3 can activate the stress activated protein








kinase/Jun N-terminal kinase (SAPK/JNK) pathway which has been shown to up regulate

proapoptotic genes (90) (455) as well as activate protein kinase C 6 (PKC8) during

apoptosis (123). Activated cell cycle regulators such as cyclin dependent kinases (cdk)

have also been observed to be in apoptotic cells (473). The arrays of effector caspase

substrates include those that facilitate the breakdown of cellular structure. The nuclear

matrix proteins lamin A, B & C are cleaved by caspase-6 (347). All three effectors

cleave the structural nuclear mitotic apparatus protein (NuMA) (148). Cytoskeleton

proteins that are cleaved include actin, gelsoin, fodrin and keratin (373).



Intrinsic Activation of Caspases: Mitochondrial Permeability Transition Pore

The role of mitochondria in apoptosis was established following the discovery of

several mitochondrial proteins, which normally reside in the intermembrane space, but

are released into the cytoplasm following appropriate stimuli. Mitochondria have been

known for some time to be the energy centers of eukaryotic cells. Mitochondrial

transmembrane potential (Aym) represents the potential between ions distributed on either

side of the inner mitochondrial membrane that results in a chemical (pH) and electric

gradient needed for mitochondrial function (20). During the induction of apoptosis, Ay.m

collapse has been found to precede the changes associated with nuclear breakdown in a

number of cell types (208). Although the events leading to Arn disruption and

mitochondrial damage have not yet been elucidated, it is clear that pro- and anti-apoptotic

members of the B-cell lymphoma-2 (Bcl-2) family regulate the intrinsic mitochondrial

pathway.









The Bcl-2 family consists of more than 12 proteins that have been divided into

three groups based on structural and functional similarities (2, 17). Members of the anti-

apoptotic first group including Bcl-2 and Bcl-XL, contain four conserved Bcl-2 homology

(BH) domains (BH1-BH4). The C-terminal hydrophobic tails localize the proteins to the

outer mitochondrial surface while the major portion of these proteins face the cytoplasm.

The second group consists ofpro-apoptotic members including Bax and Bak,

characterized by three BH domains and the C-terminal hydrophobic tail. Members of the

third group are diverse and pro-apoptotic including Bid and Bik, containing only the BH3

domain. Bcl-2 family members can form homodimers, but more importantly the pro- and

anti-apoptotic members interact to form heterodimers. Large numbers of such

heterodimers can be found in the cell. Ultimately the levels of pro- versus anti-apoptotic

family members determine the fate of the cell. An excess ofpro-apoptotic Bcl-2 family

members causes cell death conversely the presence of larger numbers of anti-apoptotic

members is protective.

There are a number of models to explain the mechanism of AYm disruption

mediated by the formation of permeability transition pores (PT) that allow the dissipation

of inner membrane ion gradients (147). Since Bcl-2 members oligomerize, they could

potentially form channels to facilitate protein transport. Alternatively Bcl-2 family

members could interact with other proteins to form channels. The voltage dependent

anion channel (VDAC) can bind several Bcl-2 members that regulate its activity (147).

Similarly the adenine nucleotide translocator protein (ANT) has also been implicated to

play a role in Bax induced apoptosis (244). And finally it has been proposed that Bcl-2

members themselves induce the rupture of mitochondrial membranes thereby disrupting









Aym and facilitating the release of mitochondrial pro-apoptotic proteins into the

cytoplasm.

The release of cytochrome c leads to formation of the mitochondrial apoptosome

in conjunction with cytoplasmic Apaf-1 and procaspase-9, thus activating caspase-9.

Second mitochondrial activator of caspases (SMAC) or direct IAP binding protein with

low pI (Diablo), is a 25 kDa pro-apoptotic polypeptide released from mitochondria (87,

434). The first four amino acids of mature SMAC/Diablo, Ala-Val-Pro-Ile (AVPI) binds

baculovirus inhibitor of apoptosis repeat (BIR) 3 domain of X chromosome encoded

inhibitor of apoptosis protein (XIAP) (59). SMAC/Diablo can also bind the BIR2

domain of XIAP, a region important for XIAP mediated inactivation of caspases-3 and -

7. Thus SMAC binding to XIAP releases active caspases-3, -7 and -9 thereby promoting

cell death. The serine protease Htra2/Omi is also released from mitochondria and has

been shown to interact with XIAP and relieve inactivation of caspases (394). Apoptosis

inducing factor (AIF) and endonuclease G (EndoG) have recently been implicated in cell

death independent of caspase activation (224, 309, 393). Thus as a result of

mitochondrial dysfunction cell death occurs that is both dependent and independent of

caspase activity.



Extrinsic Activation of Caspases: Death Receptor Mediated Signaling

Apoptotic death, an essential feature of the mammalian immune system, regulates

lymphocyte maturation, receptor repertoire selection, homeostasis and cell-mediated

immune response. This pathway of programmed cell death makes use of specialized cell

surface receptors termed the death receptors (DRs) that belong to the tumor necrosis









factor receptor (TNF-R) superfamily. Members of the DR family include Fas/CD95,

TNF-R1, DR3, DR4 and DR5 (188). The common feature of these molecules is the

presence of the cytoplasmic death domain (DD) that is essential for apoptotic signal

transduction. Receptor binding leads to oligomerization and formation of the death

inducing signaling complex (DISC) that allows the recruitment of intracellular adapter

proteins to the DISC.

Interaction between the death domains of the receptors and adapter proteins such

as Fas associated death domain protein (FADD) and TNF-R associated death domain

protein (TRADD) are responsible for the binding. The adaptor molecules also contain

the death effector domains (DED) that allow recruitment of procaspases-8 and -10 to the

DISC. The initiator caspases are autocatalyzed and serve as activators of the terminal

effector caspases that cleave the death substrates. The importance of adapter molecules

like FADD in apoptosis is evidenced in cell lines derived from FADD knockout mice that

fail to undergo apoptosis induced by CD95 ligation. However FADD"-' fibroblasts were

still sensitive to cell death induced by drugs that utilize the mitochondrial pathway (470).

The discovery of viral and cellular proteins that function to block the signal

transduction events downstream of the DISC gave rise to a new family of anti-apoptotic

proteins termed the FLICE inhibitory proteins (FLIPs). The viral FLIP found in certain

herpesviruses and poxviruses inhibit the recruitment of procaspase-8 to the DISC

underlining the importance of this pathway in pathogen clearance. Casper/cFLIP is the

cellular homolog of vFLIP and contains two DEDs and a defective caspase-like domain.

Casper can associate with the DEDs of FADD and caspase-8 thereby interfering with the

formation of a functional DISC. As expected cells from Casper knockout mice are highly









sensitive to receptor mediated apoptotic induction (462). Interestingly gene disruptions

in FADD"', Casper'-, and caspase-8"' cause embryonic lethality at similar stages and

display abnormalities in heart development (432, 462, 463).

Diseases evidence the importance of DR mediated cell death in the regulation of

the immune system where either too little or too much apoptosis occurs. The

lymphoproliferation mutation (lprc) in mice is a point mutation in the CD95 DD that

abolishes apoptotic signal transduction causing symptoms similar to those seen in

systemic lupus erythematosus. Very similar symptoms are seen in generalized

lymphoproliferative disease (gid) mice where a point mutation at the carboxy terminus of

the CD95 ligand blocks ability to ligate the CD95 receptor. In both cases there is an

accumulation of aberrant T cells. In humans, autoimmune lymphoproliferative syndrome

(ALPS) type I is a dysfunction of the CD95-CD95L system showing severe

autoimmunity with non-malignant lymphadenopathy and an increase in T cells. Thus the

failure to purge self-reactive lymphocytes can cause autoimmunity. AIDS is

characterized by a depletion of CD4+ T helper cells in the peripheral blood of patients

with human immune deficiency virus (HIV). Viral regulatory proteins such as HIV-1 Tat

enter non-infected cells and causes them to be hypersensitive to apoptosis by CD95

ligation. The binding of HIV gp 120 to CD4 also increases CD4 T cells to undergo

apoptosis. Thus receptor-mediated apoptosis seems to be a mode of CD4+ T cell death

that plays a role in the immune regulation of HIV infection.









Granzyme B

Cytotoxic T cells CTLs and NK cells can selectively induce apoptosis in by

binding the death receptor CD95 on target cells or alternatively can cause apoptosis via

the cytotoxic granule granzyme B (GrB). GrB, a 35 kDa protein, is a member of the

granzyme family of lymphocyte serine proteases and is the only enzyme known to date

that shares aspartate cleaving specificity with caspases. Granzymes in general are

involved in cytolysis brought about by CTLs and NK cells. There are 5 granzymes

known to date in humans -A, -B, -H, -K and -M (184) the most abundant being GrA and

GrB (320). Granzymes are able to enter the target cells via the pore forming protein

performin. Performin shares structural and functional homology with the complement

membrane attack complex and gets inserted into target cell membranes where it

polymerizes to form perforin pores. Granzymes are synthesized as pre-zymogen

peptides. The pre or leader sequence targets the molecule to the secretary pathway.

Cleavage of the propeptide activates the enzymes. While purified granzymes alone have

no cytotoxic effects, their enzymatic substrates are potent cytolytic agents. GrB has been

shown to cleave procaspases-3, -6, -7, -8, -9, and -10 thus activating both initiator and

effector caspases. Cleavage of the Bcl-2 family member Bid directly by GrB activates

the mitochondrial pathway in apoptosis. In addition GrB has been shown to cleave death

substrates including PARP, DNA-PKcs, and NuMA (16, 114). Thus GrB plays an

important role in the induction of apoptosis through both the intrinsic and extrinsic

pathways.









Apoptosis in Avian Species

Apoptosis in the avian system is less well understood compared to the information

available from the mammalian system. Both mitochondrial mediated and DR mediated

pathways of apoptotic induction have been observed in birds. Most of these studies have

been done in chicken, quail and duck. A number of avian death receptors have been

identified to date. The first avian death receptor protein TVB was identified as the

receptor for avian leucosis sarcoma viruses (49). TVB is 37% identical to human DR-5

death receptor and has been shown to function via its death domain (48, 49). The chicken

B6 (chB6) molecule, widely used as a B lymphocyte marker, does not contain any

apparent DDs, but has been shown to induce apoptosis when cross-linked with antibodies

or overexpressed in murine cells (115, 317). The chicken homolog to human DR-6 is

76% identical to its human counterpart and contains a death domain and two TRAF

binding sites (45). Elevated levels of DR-6 protein correlated with apoptotic induction in

the hen ovary (45). Characterization of partial cDNA sequences of chicken Fas and TNF-

RI have shown 37% and 42% identity to the human homologs respectively (46).

Members of the bcl2 family have also been identified in chickens. The chicken bcl-xl

homolog functions to protect hen granulosa cells from apotosis in the phosphorylated

state (177). The bcl-2 homolog Nr-13 was originally identified as an antiapoptotic

protein activated upon infection by Rous sarcoma virus (124). Purified Nr-13 is able to

bind cytochrome c with high affinity suggesting that it blocks the formation of the

mitochondrial apoptosome involving cytochrome c and Apaf-1 thereby regulating the

induction of apoptosis (265). An IAP protein has also been identified in chickens.









Inhibitor of T-cell apoptosis (ITA) protein contains BIR domains as well as the zinc

finger motif characteristic of IAPs (176).

To date four avian caspases have been cloned and characterized. Chicken

caspases-1, -2, -3 and -6 (174, 175, 178). There are no reports to date of purified chicken

caspases showing biochemical activity. Chicken caspase-1 while being 44% identical to

human caspase-1 has two cysteine residues (QCCRG) in the active site pentapeptide

compared to its human homolog (QACRG) that has only one. The chicken homolog of

caspase-2 is 70% identical to its human counterpart and contains the conserved QACRG

catalytic site sequence. Chicken caspase-3 and -6-like activity was noted in granulose

cells treated with the apoptotic inducer okadaic acid. Both chicken caspase-3 and -6 are

>65% identical to the human enzymes with complete conservation in the active site

region. Caspase-6 gene disruptions in a chicken cell line had no effect on the induction

of apoptosis (347). Chicken caspase-6-like its mammalian counterpart was seen to

cleave lamins A and C. Available reports of apoptosis in the avian system suggest

conservation of both DR and mitochondrial pathways. The identification of novel

molecules in the future will highlight differences and help gain a better understanding of

this process in birds.



Viral Inhibitors of Apoptosis

Viruses are obligate intracellular parasites and therefore induction of an apoptotic

reaction to infection is a highly effective host innate response that controls and limits the

spread of the invading pathogen. Viruses, in order to survive have developed multiple

ways to either neutralize or delay the cell death machinery thereby maximizing the








production of progeny particles. Virus infection itself generates proapoptotic signals.

Adsorption of non-infectious virus particles to the cell surface can trigger apoptotic

signaling. Avian leukosis virus enters cells via the TVB receptor, (a member of the DR

family with significant homology to human DR-5) which has been shown to transduce

apoptotic signal (48). Viral mechanisms that block both the extrinsic and intrinsic

pathways of apoptosis have been identified. The mitochondrial pathway can also be

triggered following virus infection. Viral replication cycle activates cell cycle regulators

such as p53 is thought to involve the Bcl-2 family of proteins to induce mitochondrial

dysfunction (131). Respiratory syncitial virus can activate the ER resident caspase 12

(37). The identification of viral inhibitors in both pathways of programmed cell death

indicates the importance of apoptosis in controlling a viral infection. Inhibitors of both

the intrinsic and extrinsic pathways are discussed.



Viral inhibitors of Mitochondria: PT pore modulators and Bcl-2 homologs

Recently vaccinia virus was shown to inhibit apoptotic induction mediated by

mitochondria (444). Infections of Jurkat cells with vaccinia virus deleted for caspase

inhibitor SPI-2 were found to be resistant to Fas-induced cell death, prevent the loss of

mitochondrial membrane potential and decrease cytochrome c levels as well as caspse-3

activity. The mechanism of vaccinia virus inhibition of mitochondrion mediated

apoptosis is not clearly understood but is believed to involve the ability of the virus to

modulate the permeability transition pore. The Hepatitis B virus HBx protein binds

human VDAC and inhibits apoptotic induction by altering transmembrane potential

(324). Viral mitochondria localized inhibitor of apoptosis (vMIA) encoded by









cytomegalovirus forms a complex with ANT and blocks permeablization of the

mitochondrial outer membrane (128).

Modulations of mitochondrial apoptotic pathways have been observed in a

number of other viruses. Epstein-Barr virus, a y herpesvirus encodes two bcl-2

homologs. The BALF-1 protein contains two BH domains and protects transfected HeLa

cells from undergoing apoptosis when induced with anti-Fas antibody (241). The

mechanism of BALF- 1 function seems to be via association with proapoptotic bcl-2

members, Bax and Bak. The BHRF-1 protein is less similar to bcl-2 and bcl-xl than

BALF-1 but has shown to protect human B cell lines from apoptotic induction with

ionomycin (146, 241). The adenovirus E1B19K protein is also a bcl-2 homolog that

blocks apoptosis during virus infection (327). E 1B19K does not block the activation of

caspase-8 but rather interacts with Bax causing inhibition of cytochrome c release from

mitochondria and subsequent activation of caspase-9 (313). Further characterization has

shown it to specifically block oligomerization of Bax by preventing exposure of the Bax

BH2 epitope (391,392).



vFLIPs: Inhibitors of DISC

Collectively, proteins that inhibit the activation of initiator caspases in the DISC

belong to the family of FLICE (caspase-8) inhibitory proteins (FLIPs). A common

characteristic of FLIPs is the presence of DEDs. DED containing proteins have been

located by database searches of herpes and poxvirus genomes (32). The equine herpes

virus E8 protein interacts with the prodomain of caspase-8 preventing its activation and

possibly its recruitment to the DISC. Ectopic expression of E8 protects HeLa cells from









TNFRI and Fas induced apoptosis (32). MC159 ofMolluscum contagiosm virus contains

two DEDs, binds both caspase-8 and FADD and blocks apoptosis induced by Fas, TNF

and TRAIL (120). Similar vFLIPs have been found in a number of herpes viruses some

of which associate with kinases of the I,:B kinase complex (127, 229).



Baculovirus Caspase Inhibitor: P35

The baculovirus anti-apoptotic P35 proteins are unique in that they do not share

homologies with any other known proteins. Interestingly these proteins share no

homology to poxvirus serpins, yet they inhibit similar target proteinases. P35 was first

identified in the insect virus Autographa californica nuclear polyhedrosis virus (AcNPV)

where P35 mutants induce apoptosis in the Sf-21 insect cell line (isolated from

Spodopterafrugipedra) (68). P35 proteins from a number ofbaculoviruses show distinct

abilities to inhibit apoptosis (267). Most of the P35 functional data generated to date

have used AcNPV P35. In vitro studies have implicated P35 to be a suicide substrate for

caspases (4). P35 has been shown to inhibit all three groups of mammalian caspases and

specifically caspases-1, -3, -6, -7, -8 and -10 (474). In addition P35 has also been shown

to inhibit the C.elgans CED-3, Sfcaspase-1 as well as gingipain (4, 378, 459). Recently,

the crystal structures of P35 with caspase-8 an apical initiator caspase and also with

effector Sf caspase-1 I have revealed clues to its function mechanisms (92, 106, 458). The

reactive site loop (RCL) of P35 associates with the molecule's main P3 sheet core such

that the Asp residue required for protease recognition is exposed at the apex of the loop.

Following cleavage of the RCL, there forms a covalent thio-ester linkage between the

target proteinase and P35. Cleavage also induces a dramatic conformational change in









P35 where the N-terminal of P35 gets repositioned into the active site of the caspase

protecting the thio-ester linkage from hydrolysis. Interestingly the presence of certain

motifs in effector caspases but absent in initiator caspases explains the preference of P35

to complex with terminal rather than apical caspases. Ectopic expression of P35 has

shown it inhibits apoptosis in a number of systems including insects, nematodes and

mammalian cells where the apoptotic signals tested included virus infection, induction by

actinomycin D, induction by Fas and TNF, NGF depravation, developmentally

programmed cell death and induction by X-radiation (259). In each case, P35 was able to

either block completely or significantly delay apoptosis. In the light of these studies, the

broad range of activity ascribed to P35 is suggestive that the molecule is a pan-caspase

inhibitor interrupting the activity of a highly conserved and ubiquitous component of the

death machinery.



Baculovirus Inhibitor of Apoptosis Proteins (lAP)

The family of inhibitor of apoptosis (lAP) was first described in baculoviruses by

Lois Miller and others (75). lAPs have also been identified in yeast, worms, insects and

mammals (350). Two zinc-binding motifs are characteristic of lAPs, the first is the

baculovirus IAP repeat (BIR) and the second being a RING domain. The BIR motif can

be found in multiple copies in a given lAP, while the RING domain is typically a single

copy located at the carboxy terminal end of the molecule (350). The RING domains

function to catalyze the degradation of IAPs and target protein through ubiquitylation

processes. The BIR domains seem to be essential for lAPs to function as anti-apoptotic

agents (89). Baculovirus lAPs were discovered by virtue of their ability to block









apoptosis in Sfcells infected with a AcNPV P35 mutant (80, 337). The CplAP and

OpIAP proteins (isolated from baculoviruses Cydia pomonella granulosisvirus and

Orgyia pseudotsugata nucleopolyhedrovirus respectively) can also inhibit apoptosis

induced by a variety of signals including induction by, actinomycin D, expression of the

reaper and Doom genes of Drosophila melanogaster and expression of ICE in a

mammalian system (153). The baculovirus IAPs mode of inhibiting the Sfcaspase-l

activity is thought to affect the activation of this insect caspase, implying that the

mechanism by which IAPs function is different from the P35 mode of action (365). A

number of lAP homologues have been found in insects and mammals including the

human neuronal apoptosis inhibitory protein (NAIP), cIAP-l, cIAP-2 and XIAP (350).

The mechanism by which viral lAPs act in inhibiting apoptosis is not fully understood.

Studies with the crystal structures of human XIAP has elucidated a possible mechanism

of action that could be similar for all lAPs. The BIR3 domain of XIAP is thought to aid

the cleavage of caspase-9 and directly bind the small subunit at the C-terminus of

caspase-9 inactivating it (383, 390). With respect to caspase-3 and -7, XIAP seems to

have a completely different mode of inhibition than that seen with caspae-9 (60, 164,

338). In this case BIR2 domain is involved and complexes with caspase-3 and -7. A

linker domain N-terminus to BIR2 functions to block the substrate groove in the caspase,

in a backward or opposite orientation that is different from substrate recognition

sequences making it a novel mechanism for caspase inhibition. XIAP is not a substrate

for caspase-3 or -7 but rather functions to mask the active site. It will be of interest to

know whether the viral IAPs also function in a similar manner.









Caspase Independent Induction of Apoptosis

Apoptosis in the absence of active caspases seems to represent a pathway

conserved in slime mold (Dictyostelium), worms (C.elegans) and mammals (234, 297).

Two proteins have been identified to date that are released from mitochondria and are

postulated to play a role in apoptosis independent of caspase activation. The first is a

novel mitochondrial cell death inducer called the apoptosis inducing factor (AIF) (393).

Mammalian AIF is a 57 kDa flavoprotein that shares homology to oxidoreductases from

vertebrates (Xenopus laevis), non-vertebrates (C.elegans), plants, fungi and bacteria

(234). It was originally discovered as the factor able to induce caspase independent

chromatin condensation and DNA fragmentation in isolated nuclei, but itself has not been

seen to exhibit DNAse activity (393). AIF can also cause permeablization of

mitochondrial membranes and induce release of cytochrome c suggesting a role in a

positive feedback loop with the apoptosome pathway (393). Deletion of AIF in mice

causes embryonic lethality (182). While cells from knockout mice remain sensitive to

apoptosis in response to serum withdrawal, they are resistant to apoptosis induced by the

oxidative stress inducer menadione (182). Since AIF is also an oxidoreductase, it is

unclear whether observations seen in KO mice were due to loss of apoptotic activity

alone or due to the elimination of oxidoreductase activity as well.

The second toxic molecule implicated in caspase independent apoptosis is

endonuclease G (EndoG), which, like AIF, is also released from mitochondria (309).

EndoG is a 30 kDa nuclease, which was originally linked to mitochondrial DNA

replication based on its location and substrate specificity (74). Unlike CAD/DFF, EndoG

is able to induce nucleosomal DNA fragmentation independent of caspase activation









(224). It has also been implicated in DNA fragmentation seen in DFF deficient mouse

embryonic fibroblasts following induction of apoptosis by UV-radiation and TNF

treatment (469). A homolog to EndoG in C.elegans has been identified as CPS-6, which

shows similarity in sequence, localization and biochemical properties to mammalian

EndoG (309). Mutant C.elegans deficient in CPS-6 can be rescued by mouse EndoG

(309). The involvement of EndoG in cell death indicates the presence of an apoptotic

pathway, independent of caspase activation that is conserved from worms to mammals.

Although the caspase independent pathway is not completely understood, the

identification of AIF and EndoG clearly supports the existence of an apoptotic program

parallel to caspase activation.



Poxviruses

The Poxviridae are a family of large double stranded DNA viruses that replicate in the

cytoplasm of infected cells. The taxonomy of poxviruses is based on hosts they infect and

include the two subfamilies Chordopoxvirinae containing vertebrate viruses and

Entomopoxvirinae containing the insect poxviruses (Table 2) (271). The

Chorodopoxvirinae comprises eight genera: Orthopoxvirus, Parapoxvirus, Avipoxvirus,

Capripoxvirus, Leporipoxvirus, Suipoxvirus, Molluscipoxvirus and Yatapoxvirus. While

there are three genera within Entomopoxvirinae, these viruses may need to be reclassified

based on recent sequence information that suggests that they may not be as closely

related as previously thought (3, 27). Both variola virus (Orthopoxvirus genus) and

molluscum contagiosum virus (Molluscipoxvirus genus) are the only poxviruses to

exclusively infect humans (271).










Table 3. Taxonomy of poxviruses


Genera Members

Orthopoxvirus Cowpox, monkeypox, buffalopox,
vaccinia, Variola, rabbitpox,
camelpox, ectromelia, raccoonpox,
taterapox, volepox
Parapoxvirus Pseudocowpox, parapoxvirus of red
deer, bovine popular stomatitis virus,
~________ orf, paravaccinia virus
Vertebrate Avipoxvirus Canarypox, fowlpox, juncopox,
Vererat mynahpox, pigeonpox, quailpox,
poxviruses sparrowpox, starlingpox, turkeypox
Capripoxvirus Goatpox, sheeppox, lumpy skin
disease virus
Leporipoxvirus Myxoma virus, rabbit fibroma virus,
hare fibroma virus, squirrel fibroma
virus
Suipoxvirus Swinepox virus
Molluscipoxvirus Molluscum contagiosum virus
Yatapoxvirus tanapox virus, Yaba monkey tumor
virus
Entomopoxvirus A Anomala cuprea, Aphodius
tasmaniae, Demodema boranensis,
Dermolepida albohirtum, Figulus
____ subleavis, Melolontha melolontha
Entomopoxvirus B Amsacta moorei, Acrobasis zelleri,
Insect Arphia conspersa, Choristoneura
t biennis, Choristoneura conflict,
Poxviruses Melanoplus sanguinipes, Oedaleus
~____~___senigalensis, Schistocera gregaria
Entomopoxvirus C Aedes aegypti, Camptochironomus
tentans, Chironimus luridus,
Chironomus plumosus, Chironomus
attenuatus, Goeldichironomus
______________haloprasimus
California harbor sealpox, cotia
Unclassified virus, dolphinpox, embu virus, grey
Uncassified kangaroopox, marmosetpox,
poxviruses Molluscum-like poxvirus, Nile
~~~____~~______crocodilepox, mule deerpox virus


http://www.ncbi.nlm.nih.gov/ICTVdb/Ictv/index.htm









By far the most infamous poxvirus is variola, the causative agent of smallpox

infections in humans. Smallpox produced a systemic disease with a generalized rash in

humans and has been documented for about 2000 years (102). Of importance is that it is

the first and only human disease to date that has been purposefully eradicated. The

concept of variolation i.e. inoculation of dried smallpox scab material into the skin to

induce mild disease and subsequent immunity was practiced since the early 10th century

(102). It was not until 1798 when Edward Jenner demonstrated that inoculation with

cowpox could protect against smallpox. This new method, vaccination, quickly gained

widespread use and eventually led the eradication of smallpox in 1977 as a result of mass

vaccination programs carried out by the World Health Organization.



Poxvirus Lifecycle

Poxvirus genomes are linear double stranded DNA molecules that are covalently

joined at the termini. The internal core region of the genome is highly conserved among

poxviruses whereas variation occurs at the ends of the genome within the inverted

terminal repeat (ITR) regions. Genes located in this region are typically not essential for

viral growth in tissue culture, but contribute to viral pathogenesis in vivo. Poxviruses are

unique in that the entire life cycle is accomplished in the cytoplasm of the infected cell

(Figure 3). The receptor that mediates viral entry is unknown at present. However viral

entry is mediated by membrane fusion that is dependent on pH. Viral gene expression

occurs in three stages. The early genes are transcribed within the viral core that contains

all the transcription machinery. These include growth factors, immune defense

molecules such as CrmA and factors required for DNA replication. Intermediate genes










DNA


rEntry/uncoating


1
Early gene
expression


Intermediate /
transcription factors,
Immunomodulatory 0
proteins


f replication



Intermediate gene
expression

L |
gP Late gene expression
U ,


_j H f/1I _1E\V 1-1 barlyt
IEV_ @Str


EEV Golgi _____
-- Ggwrapping I j
IMV Immature
Particle Assembly


transcription factors
uctural proteins
Late enzymes




j


Figure 3. Poxvirus replication cycle. The cytoplasmic infection of a cell with an
infectious poxvirus particle. Abbreviations: EEV, extracellular enveloped virus; IMV,
intracellular mature virus; IEV, intracellular enveloped virus.


EEV


CEV


4" .









are transcribed concomitantly with DNA replication. Intermediate gene products are

factors for late gene transcription. Late gene products include structural proteins,

enzymes and early transcription factors. The viral DNA is then packaged within the

cytoplasmic viral factories and maturation proceeds to form infectious intracellular

mature virus (IMV) particles. IMV can be enveloped by membranes derived from the

golgi apparatus to form intracellular enveloped virus (IEV) which upon exiting the cell

looses a membrane layer and is termed extracellular enveloped virus (EEV).



Importance of Poxviruses

Poxviruses are extremely efficient in evading the effectors of antiviral immunity. Many

of these mechanisms target conserved cellular pathways. The knowledge gained from the

interactions of these viral proteins would be useful in understanding the immune response

to a poxvirus infection. Such studies also help in elucidating mechanisms of immune

evasion used by the virus. Therapeutic strategies based on models generated from viral

studies could be used to design molecules to modulate immune responses that occur in

other pathological conditions. Vaccinia virus is also an excellent vector for the

expression of foreign proteins, but the virus also elicits a tremendous immune response.

Therefore while the virus has potential as a vector in gene therapy, it might be

advantageous to minimize the side effects of treatment (immuneregulation) without

compromising gene delivery. Identifying and possibly modifying the genes responsible

for the vigorous immune response can improve its use in gene therapy. Finally, though

smallpox has officially been eradicated, interest in smallpox is sustained due to its

potential use as a weapon of mass destruction in biological warfare. Smallpox is also









similar to another related poxvirus namely monkeypox that causes zoonotic infections of

humans producing a clinically identical disease.

Poxviruses encode a number of molecules that aid in subverting and blocking the

host immune response at various levels. As described earlier, poxviral immune evasion

proteins include those that function to block complement activation, TNF, IL-103, IL-18,

IFNa/P signaling via either binding proteins such as the virally encoded IL- 18 binding

protein or receptor homologs such as CrmB-E which bind TNF. In order to modulate

immune cell trafficking poxviruses encode proteins that bind chemokines or are

chemokine homologs. Poxviruses encode intracellular molecules to block activation of

cellular responses to viral replication, which include the double stranded RNA binding

protein E3L and K3L a homolog to cellular initiation factor eIF-2cc. Also included

among the intracellular proteins believed to interact with the host immune system are

poxvirus serpins, which are members of the serine proteinase inhibitor family. The

subject of this dissertation work involves the serpin CrmA first identified in cowpox virus

and subsequently found to have homologs in all orthopoxviruses sequenced to date

including variola and vaccinia viruses. Orthopoxviruses encode a total of three serpins

SPI-1, SPI-2/CrmA and SPI-3, which will be discussed later.



Serine Protease Inhibitors (Serpins)

The serpin protein superfamily is characterized by a conserved structure and

mode of action that can be described as suicide substrate-like inhibition (368). Serpins

are ubiquitous in nature, found in animals, plants, certain viruses and more recently in

bacteria (368, 449). Based on phylogenetic analysis serpins are grouped into 16 clades









and 10 divergent orphans (368). Although most serpins function to inhibit serine

proteases, there exist examples of cross class inhibition such as that exhibited by the

poxviral serpin CrmA and the cellular PI-9 which can inhibit both serine and cysteine

proteases (e.g. caspases). In addition some serpins such as ovalbumin are non-inhibitory

while corticosteroid binding globulin (CBG) and thyroxine binding globulin (TBG)

function as transport proteins (165, 311). In mammals serpins are involved in regulating

a number of processes including coagulation, fibrinolysis, inflammation, complement

activation, organ development and extracellular homeostasis as shown in Figure 4.

A metastable structural conformation is extremely important for the function of inhibitory

serpins. The conserved secondary structure is composed of three P-sheets and at least 7

(typically 9) a-helicies. The exposed reactive center loop (RCL) is a region spanning

about 20 amino acids located at the C-terminus of the molecule which serves as the

proteinase recognition site as shown in Figure 5. Within the RCL lies the P 1 residue that

determines target proteinase specificity. The P1-PI' peptide bond is cleaved during

scissile bond formation with the proteinase. Initially the proteinase forms a non-covalent

Michaelis-like complex by interactions involving residues flanking the scissile bond.

Attack by the active site serine (in the case of serine proteases) on the P1 residue

leads to the formation of a covalent bond between the carbonyl of the P 1 residue and the

protease. At this point, the RCL begins to insert itself into 13-sheet A of the serpin while

still attached to the protease. The consequence of complete loop insertion is a 70A

translocation of the target proteinase that effectively distorts its active site resulting in

inactivation of the enzyme. The example of trypsin inactivation by alpha-1 antitrypsin is

shown in Figure 6. The formed acyl intermediate is stable from hours to weeks and is



















tumor cell invasion


prohormone conversion


SPCI (SERPINAS)
A Till (SERPINCI)
HCFII (SERPINDIJ
PAIl (SERPINEI)
ECM maintenance
and remodelling ---
I' PI(SERPINAI) PNI (SERPINE2)
alACT(SERPINA3) HSP47(SERPINHI)
PAll (SERPINEI) CBP2 (SERPINH2)


hormone transport

/CBG(SERPINA6) I
TBG (SERPINA7) PCI (SERPINAS)




inflammation &
"'- T complement activation
1 PI (SERPINAI)
adACr (SERPINA3)
SKAL(SERPINA4)
MNEI (SERPINBI)
oip-serins (SERPINB*)
blood pressure regulation
\' AGT(SERPINA8)
7- .....angiogenesis
PEDF (SSERPINFI)
maspiR (SERPINBS)?
A Till (SERPINCI)?
P411 (SERPIEI)?
T fibrinolysis


apoptosis
1 OlPI (SERPINAI)
-LIACT(SERPINA3)
av-.sepiins (SERPINB*)?


Figure 4. Serpins as physiological regulators. Serpins are involved in numerous
functions both as positive and negative regulators. From reference 368.
























P14....P4 P3 P2P1-P1' P2' P3' P4'......P14'
t
Scissile Bond




Figure 5. Serpin RCL. Position of the C-terminal reactive center loop (RCL) in the
serpin molecule. P 1 residue within RCL determines specificity for proteinase.











Trypsin (active)


Serpin-proteinase complex


Native alpha-1 antitrypsin


Trypsin (disrupted)


Figure 6. Inhibition of trypsin by alpha-1 antitrypsin. Following formation of the
serpin-proteinase complex, the scissile bond is cleaved (P1 '-P1). RCL insertion into the
backbone of alpha-1 antitrypsin (serpin) causes Trypsin (target proteinase) to be
translocated such that its active site is effectively distorted resulting in inactivation of the
enzyme. From Huntington, et al. Nature, 2000


MM+









eventually hydrolyzed releasing the trapped previously active enzyme and cleaved serpin.

Serpin-protease complexes involving serine proteases are SDS stable and can be readily

detected on reducing gels. Whereas complexes with cysteine proteases, such as with

caspases, are intrinsically unstable due the nature ofthiol ester linkages that are easily

hydrolyzed and can only be detected on native, non denaturing gels.



Poxvirus Serpins

The first poxvirus serpin discovered was CPV CrmA (cytokine response modifier A). A

total of three serpins SPI-1, SPI-2/CrmA and SPI-3 are present in orthopoxviruses. SPI-1

mutants of RPV display host range restriction. SPI-1 can form complexes with cathepsin

G (263) and more recent work suggests that SPI-1 may interact with the viral DNA

replication machinery (235). Disruption of the SPI-3 gene in VV or CPV results in cell

fusion and the formation of giant syncytium (422). SPI-3 expressed in vitro can form

stable complexes with some serine proteinases including trypsin, urokinase and tissue

plasminogen activator (420). Mutational analysis reveals that serpin function is not

required for SPI-3 to inhibit cell fusion during virus infection, suggesting SPI-3 to be a

bi-functional protein, with two separate functional motifs (423).

The SERP1, SERP2 and SERP3 proteins are serpins encoded by Myxoma virus

(135, 314, 428). SERP1 is the only secreted viral serpin and in vitro experiments have

shown SERP 1 to inhibit serine proteinases including human plasmin, urokinase and

tissue plasminogen activator, as well as C Is of the complement system suggesting an

anti-inflammatory activity for this serpin (442). The biochemical properties of SERP1

are similar to SPI-3 of orthopoxviruses. Indeed in vivo studies have proven the potent









action of SERP1 against inflammation (237). Although SERP1 and SPI-3 share similar

inhibitory profiles in vitro, SPI-3 cannot substitute for SERP 1 in myxoma virus infections

of rabbits (442). The SERP2 protein shares a 45% similarity with CrmA/SPI-2 from

Orthopoxvirus and is also functionally similar to CrmA/SPI-2. SERP2 will be discussed

in more detail later. SERP3 was recently identified from myxoma virus and despite

having significant sequence deletions, has been proposed to retain overall serpin fold

(135). Deletion of SERP3 from myxoma virus causes attenuation of disease and prevents

viral spread to secondary sites in rabbits, suggesting a role in virulence. It is not yet

known if this serpin can inhibit proteases.

Little is known about the swine poxvirus (SPV) serpin SPI-7 that contains an Asp

at P1 like CrmA/SPI-2. SPI-7 does not reveal any activity against caspases tested to date,

although SPV mutants deleted for SPI-7 have shown some attenuation (Pierre Mussy &

Richard Moyer unpublished). There are 5 serpins found in the fowlpoxvirus genome, in

addition serpins have been identified in Yaba-like disease virus (YLDV) and lumpy skin

disease virus (LSDV) (250). To date none of the poxvirus serpins are essential for virus

growth in tissue culture.



SERP2

Like CrmA, the myxoma virus serpin SERP2 also contains an Asp residue at the

P1 position in the RCL. SERP2 is 35% identical to CrmA, can block caspases-1 and -8

as well as granzyme B activity (424). With respect to caspase-1, SERP2 is a less

effective inhibitor (K, = 80 nM) than CrmA (K, = 4 pM). Similarly for granzyme B,

SERP2 is four times weaker (K, = 420 nM) compared to CrmA (K; = 100 nM). Deletion









of the SERP2 gene in myxoma virus leads to attenuation of disease as evidenced by

survival of infected rabbits (257). SERP2 has been reported to inhibit the influx of

inflammatory cells into the site of primary inoculation as well as prevent apoptosis of

infected lymphocytes in the rabbit model. While recombinant cowpox virus expressing

SERP2 instead of CrmA failed to block the induction of apoptosis in pig kidney cells,

SERP2 in the context of myxoma virus does inhibit apoptotic induction during infection

of rabbit kidney cells (424). The importance of the P1 residue in SERP2 is evidenced by

the observation that mutant viruses, in which the P I Asp is changed to Ala, cause an

attenuated disease (Amy MacNeill and Richard Moyer unpublished). If SERP2 were to

function simply as a caspase inhibitor in vivo, then a recombinant myxoma virus

expressing CrmA instead of SERP2 should produce similar effects as the wt myxoma

infection. On the contrary CrmA replacements of SERP2 in myxoma virus causes only

~60% mortality with a marked difference in primary lesion pathology (Amy MacNeill

and Richard Moyer unpublished). Thus SERP2 is an important virulence factor for

myxoma and deducing its natural target would provide clues to its role during myxoma

infection as well as its function as a serpin.



Cytokine Response Modifier A (CrmA)

CrmA was the first viral serpin molecule identified (316). It is an example of a

cross class inhibitor in that it has been shown to inhibit both the serine protease granzyme

B as well as members of the caspase family of cysteine proteases (201). CrmA was first

identified based on its ability to alter lesion morphology on chicken chorioallantoic

membranes (CAM). CPV produces characteristic red, hemorrhagic, non-inflammatory








pocks or lesions on 11 day old CAMs. Deletion of the CrmA gene resulted in a mutant

virus that produces white, non-hemorrhagic pocks that contained infiltrating

inflammatory cells consisting mainly of heterophils (analogous to neutrophils in

mammals) (306). Thus the gene responsible for this prevention of heterophil influx was

termed cytokine response modifier-A (CrmA). White pocks also produce much less virus

compared to wt CPV and have the ability to reduce nitroblue tetrazolium (NBT)

consistent with the presence of activated chicken heterophils. Histological analysis of the

white pocks also shows a moderate heterophilic inflammation absent in wt CPV

infections. Another study has shown that CrmA prevents the formation of (14R, 15S)

dihydroxyeicosatetraenoic acid (diHETE), an arachidonic acid metabolite during CPV

infection (305). It was further suggested that these diHETE intermediates could directly

or indirectly lead to the generation of inflammatory mediators. Whether or not this is the

case during CPV infections of CAMs has not yet been established.

Around the same time it was shown that CrmA could inhibit caspase-1 (IL-

lp convertase) in vitro (331). Caspase-1 plays a major role in the maturation ofproIL-l1

and prolL-18, both of which are proinflammatory cytokines (122, 133,413). Therefore

the role played by CrmA in blocking inflammation during CPV infections of CAMs was

thought to be through inhibition of a caspase-1 mediated release of proinflammatory IL-

1p3 and/or IL-18. IL-18 was previously described as the factor that induces synthesis of

interferon y (IFNy). Viral proteins that function to block the activity of both IL-1 3 and

IFNy provide evidence for the importance of these host responses to poxvirus infections.

In addition to controlling IL-1p3 maturation through CrmA, orthopoxviruses also encode a

soluble IL-1p receptor that binds IL-13 (381). Furthermore it is this soluble secreted









receptor, not CrmA that serves to control fever mediated by IL-1 3 in infected mice (192).

Like IL- I3, the significance of IFN in poxvirus infections can be inferred because

poxviruses control IFN both extracellularly by encoding an IFN a/P and y binding

proteins and intracellularly through proteins (vaccinia E3L and K3L genes) that interfere

with signaling events which occur after receptor binding (9, 11, 79). Deletion of either

the soluble receptors or the intracellular viral proteins E3L and K3L also leads to

attenuation of the virus (28, 456).

In addition to caspase-1, CrmA inhibits caspases-4, -5, -8, -9 and -10 at

physiological levels (118). Caspases play a central role in mediating apoptosis.

Consistent with its ability to directly inhibit a number of apical caspases, CrmA has been

shown to block the induction of apoptosis by various stimuli in different systems (117,

171,287, 332, 407,408). Caspases can also be activated by granzyme B, a serine

protease that is part of granules synthesized by CTLs and NK cells. CrmA blocks

granzyme B mediated apoptosis by directly inhibiting granzyme B (322). Although

CrmA is shown to block the activation of caspases, its role in preventing the induction of

apoptosis during CPV infection has only been demonstrated in pig kidney cells (236,

332). Furthermore, no relationship has been found between CrmA expression and virus

yield in tissue culture as both wild type CPV and CrmA deletion mutants produce similar

amounts of infectious virus (332) and CrmA is not required for viral growth in cell

culture.









SPI-2 is a CrmA Homologue

The 38 kDa CrmA homolog found in other orthopoxviruses is designated as

serine protease inhibitor-2 (SPI-2). SPI-2 from rabbitpox virus (RPV) is 92% identical to

CrmA and its expression profile is similar to CrmA in virus-infected cells (236). In vitro

studies show both proteins inhibit caspase-1 with similar efficiencies. In pig kidney cells

CrmA functions to prevent the induction of apoptosis during CPV infections. But wild

type RPV despite the presence of SPI-2 cannot block apoptosis in the swine cells. In

mixed infection experiments, CrmA was able to substitute for SPI-2 in RPVASPI-2

infected cells, but the converse was not true. SPI-2 in RPV could not prevent apoptotic

induction in CPVAcrmA infections. Thus, in spite of their high degree of homology and

their identical properties in vitro, SPI-2 and CrmA do not seem to be functionally

equivalent. Ectromelia virus (EV) SPI-2 shares 94% identity with other poxviral

homologs (425). Similar to CrmA expression of EV SPI-2 also protected cells from TNF

mediated apoptosis and inhibited the activity of caspases-1 and -8.

While most of the information about CrmA has been obtained by ectopic

expression or virus infection of tissue culture cells, its role in vivo is less well understood.

Intranasal infections of mice with CPV deleted for CrmA were attenuated and showed a

higher LD50 by at least one order of magnitude (412). Mice infected with a CrmA

deletion mutant also had decreased pulmonary pathology and reduced inflammation

compared to wild type CPV infections. Conversely, deletion of the CrmA homolog SPI-

2, in vaccinia virus has no effect on virulence in mice (192) despite the fact that vaccinia

SPI-2 functioned to inhibit the induction of apoptosis during virus infection and virus

infected cell extracts were able to block the processing of proIL-1 P. Mice infected via








the intra dermal route with vaccinia virus deleted for SPI-2, showed enhanced pathology

(417). Surprisingly the lesions produced by the SPI-2 deleted virus were significantly

larger than those produced by wild type vaccinia. Thus although SPI-2 and CrmA are

homologous proteins, some differences do exist.



RPV SPI-2 and CPV CrmA Equivalency

Previous studies have shown that LLC-PK1 pig kidney cells infected with wt

CPV maintain a typical morphological structure of the nuclei throughout the cycle of

viral infection until cell lysis occurred (332). However pig kidney cells infected with a

CPVAcrmA mutant virus showed cytopathic effects consistent with apoptotic death.

Studies with RPV infected cells showed that RPVASPI-2 mutant virus infections also

induced typical morphological characteristics associated with apoptosis (236). But more

interestingly infections with wt RPV induced apoptosis. This result was surprising due to

the fact that RPV encodes a functional SPI-2 (CrmA) protein.

Apoptotic cells manifest a number of biochemical changes that include activation

of caspase enzymes, which lead to the cleavage of nuclear proteins such as PARP and

lamin. CPVAcrmA, RPV and RPVASPI-2 infected extracts contain an activity that

cleaved PARP from its native 116 kDa form to a 85 kDa fragment characteristically seen

in cells undergoing apoptosis whereas wt CPV infected extracts failed to do so (236).

The presence of lamin A cleaving activity in the extracts mirrored the results obtained

with PARP cleavage. These biochemical assay results correlated exactly with our

observations of infected cells stained with DAPI, which stains DNA of the cell nucleus.









SPI-2 of RPV shares 92% amino acid homology with CrmA, as shown in Figure

7. However these two proteins differ in the reactive site loop at positions P5 and P6,

thereby not excluding the possibility that the two proteins may have different inhibitory

specificities with respect to activated caspases in apoptotic cells. It has been reported that

there is at least one example of a caspase that recognizes residues away from the P1

cleavage site as far as the P5 residue. Therefore it is possible that the differences

between SPI-2 and CrmA at P5 and P6 may in fact play a role during virus infection.

Published reports from our laboratory have shown that caspases are activated during RPV

but not CPV infections (236). In vitro experiments with purified SPI-2 and CrmA have

indicated that both the proteins could inhibit the activity of caspase-1 with similar

efficiencies. These results suggest that although the two proteins SPI-2 and CrmA

function similarly in vitro, they may not be functionally equivalent in vivo.

Swapping the two genes between the viruses allows us to examine any differences that

might be present between SPI-2 and CrmA. If SPI-2 and CrmA have distinct properties,

then one would expect an infection of LLC-PK 1 cells where CrmA is expressed in place

of SPI-2 in RPV to inhibit the induction of apoptosis (unlike wt RPV infected cells). And

similarly SPI-2 expressed in place of CrmA in CPV should induce apoptosis in LLC-PK1

cells (unlike wt CPV infections). However if there were an additional factor unique to

CPV that is required in addition to CrmA to inhibit the induction of apoptosis, then

blockage of apoptotic induction would be observed only during CPV infection regardless

of SPI-2 or CrmA expression. The results of this project are presented in this thesis and

are discussed.







83





Alignment of SPI-2 and CrmA

1 mdifreiassmkgenvfispasissvltilyygangstaeqlskyvekee 50
III III I IIII1111111I111111111111 IlIi II
1 mdifreiassmkgenvfisppsissvltilyygangstaeqlskyvekea 50

51 nmdkvsaqnisfksinkvygrysavfkdsflrkigdkfqtvdftdcrtid 100

51 dknk...ddisfksmnkvygrysavfkdsflrkigdnfqtvdftdcrtvd 97

101 ainkcvdiftegkinplldeqlspdtcllaisavyfkakwltpfekefts 150
III1 IIIIII l I 111 IlIllIlII II I I111 I111 II
98 ainkcvdiftegkinplldeplspdtcllaisavyfkakwlmpfekefts 147

151 dypfyvsptemvdvsmmsmygkafnhasvkesfgnfsiielpyvgdtsmm 200
III IIIIII I II111 .1111I11II 1I I I II11 II .
148 dypfyvsptemvdvsmmsmygeafnhasvkesfgnfsiielpyvgdtsmv 197

201 vilpdkidglesieqnltdtnfkkwcnsleatfidvhipkfkvtgsynlv 250
111I 111IIII 11I11III111.:: I I11111 I I 1II
198 vilpdnidglesieqnltdtnfkkwcdsmdamfidvhipkfkvtgsynlv 247

251 dtlvksgltevfgstgdysnmcnldvsvdamihktyidvneeyteaaaat 300

248 dalvklgltevfgstgdysnmcnsdvsvdamihktyidvneeyteaaaat 297

301 svlvadcastvtnefcadhpfiyvirhvdgkilfvgrycspttnc* 346
lIi I II IIII IIIIIIII11111II11I11111111I
298 calvadcastvtnefcadhpfiyvirhvdgkilfvgrycspttn*. 342
4
PI


Figure 7. Comparison of SPI-2 and CrmA peptide sequences. P 1 position is indicated
by arrowhead. Sequence alignment was performed using GCG Gap program.









In Vivo Role of CrmA During CPV Infections of CAMs

Apoptosis and inflammation are both important mechanisms by which the host

limits and clears viral infections. The results from this project will help gain a better

understanding of viral pathogenesis in terms of both virus-host interactions and the ability

of virus to modulate the host immune system during an infection.

We use the infection of CAMs of embryonated eggs as a model to study the role

of CrmA on the innate immune response including inflammation and apoptosis. The 10

day old CAM is an excellent model to study an acute inflammatory response to infection.

At this stage of embryonic development there is an absence of circulating mature B or T

cells (63,247), minimal complement protein (116) and the inflammatory response due to

CPV infection is mainly mediated by heterophils (analogous to mammalian neutrophils)

and monocytes (112). In order to understand the role played by CrmA during a CPV

infection of the CAM, we addressed the following questions (i) does CrmA prevent white

pock formation by protease inhibition, (ii) if so, is the function of CrmA solely dependent

on inhibition ofcaspases and (iii) can serpin homologs found in different poxviruses

function similarly?

The first question was addressed by mutating the P 1 Asp residue in the CrmA

RCL to alanine (D303A), which should abolish serpin function with the predicted

consequence of inducing white pocks. Such a result would also mean that an intact

Pl Asp is critical for CrmA's function as an inhibitor of inflammation. The second

question was addressed by replacing CrmA in CPV with other caspase inhibitors such as

P35 ofbaculovirus or SERP2 from myxoma virus. If the in vivo role of CrmA was to

only inhibit caspases, then P35 and SERP2 should completely replace the function of