Functional analysis of the rabbitpox virus serpin SPI-1

Material Information

Functional analysis of the rabbitpox virus serpin SPI-1 characterization in vitro and in vivo
Alternate title:
Characterization in vitro and in vivo
Moon, Kristin Laurie, 1972-
Publication Date:
Physical Description:
xvi, 256 leaves : ill. ; 29 cm.


Subjects / Keywords:
Apoptosis ( jstor )
Cells ( jstor )
Diseases ( jstor )
Infections ( jstor )
Lesions ( jstor )
Rabbits ( jstor )
Serpins ( jstor )
Vaccinia ( jstor )
Vaccinia virus ( jstor )
Viruses ( jstor )
Cathepsins ( mesh )
Chymotrypsin ( mesh )
Department of Molecular Genetics and Microbiology thesis Ph.D ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Molecular Genetics and Microbiology -- UF ( mesh )
Protease Inhibitors ( mesh )
Rabbits ( mesh )
Research ( mesh )
Serpins ( mesh )
Vaccinia virus -- pathogenicity ( mesh )
Virulence ( mesh )
bibliography ( marcgt )
non-fiction ( marcgt )


Thesis (Ph.D.)--University of Florida, 2000.
Bibliography: leaves 228-255.
General Note:
General Note:
Statement of Responsibility:
by Kristin Laurie Moon.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025671815 ( ALEPH )
52006966 ( OCLC )

Full Text







Copyright 2000


Kristin Laurie Moon

Dedicated to my husband Joel for his love, support and his constant faith in me.


I would like to take this opportunity to thank the many people that have

contributed to this study. First, I would like to thank my mentor Dr. Richard Moyer for

his supervision, support, and guidance through the years. He has provided a unique

environment in which to learn and work that has made the daily process of doctoral

research easier to bear. I wish to thank him for allowing me the chance to think on my

own and the opportunity to try things in the lab that may not have seemed logical at the

time. I am completely satisfied with my experience in his lab and consider him an

excellent teacher. In addition, I wish to thank the members of my supervisory committee.

Along with Dr. Moyer, Dr. Bill Dunn, Dr. John Aris, Dr. Tom Rowe, and Dr. Al Lewin

have each contributed valuable insight to my research project. I came away from every

committee meeting with optimism and great ideas for future experiments. Each member

of my committee took the time to individually meet with me to give me advice and help

me to interpret my results and I thank them for their support.

The members of the departmental administrative, editorial and fiscal staff have

been essential to my success as a graduate student. Their hard work and attention to

detail allows every student in the program more time to focus on classwork and research

and no one would be able to make it through the program without them. In particular, I

wish to acknowledge Joyce Conners for her special contribution. My occasional visits to

her office have provided only a glimpse into her stressful daily routine and I am very glad

that I don't have her job! I wish to thank her for putting up with me for six years and for

being such a delightful person.

The members of the Moyer lab have been a constant source of enjoyment for me

in my years as a doctoral student. Though some of the faces have changed over time, the

lighthearted atmosphere present in the lab has remained the same. In particular, I wish to

acknowledge the following people for their special contributions, although my words

won't do justice to my feelings for each of them. Dr. Pete Turner provided excellent

guidance and tutelage over the years, and I thank him for his assistance. Mike Duke did

an excellent job running the lab, making us smile, exposing us to alternative forms of

music and food, and keeping us informed with (useless) trivia. In addition, he has been a

truly wonderful friend whose daily companionship I miss. I wish to thank Mike Duke,

Lauren Morges, Traci Ness, Dr. Allison Bawden and Dr. Amy MacNeill for their

excellent rabbit-wrangling techniques and I hope they each realize that I would still be

stuck down in Animal Resources without their help. Finally, I want to thank Traci Ness

for many things. She has been my partner in misery and in crime, my second set of hands

in the lab, and a constant source of encouragement and support to me. She remains a

very special friend and adopted family member.

Saving the best for last, I wish to acknowledge my family. They have supported

me through six agonizing years of stress, anxiety, disappointment, excitement, and all of

the ups and downs of scientific research at the graduate level. They have been patient and

understanding, not really knowing what I was doing but all the while having faith that

one day I would finish. Without their love and encouragement I could not have made it

through the rough times, and the good times would not have been nearly as good. Most

of all, I want to thank my husband Joel for everything that he has endured and for the

many sacrifices that he has made so that I could finish my degree. I thank him for his

patience, encouragement, and his prayers. I thank him for the many times that he

challenged me to stick it out and follow through with the goals that I had set. I thank him

for having more faith in me than I ever had in myself and for his unconditional love.



ACKNOW LEDGEM ENTS .......................................................................................... iv

LIST OF TABLES ........................................................................................................ ix

LIST OF FIGURES ....................................................................................................... x

KEY TO ABBREVIATIONS ........................................................................................... xii

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


I INTRODUCTION AND BACKGROUN D ................................................................. 1

Poxvirus Genomic Structure ................................................................................... 3
M olecular Biology of Poxviruses ............................................................................. 4
Poxvirus Infections of Social and Economic Importance to Man .............................. 11
Poxvirus Replication and Spread in the Host ........................................................ 16
Rabbitpox Virus ...................................................................................................... 18
Poxvirus Immune Modulators that Counteract the Host Early Response to
Virus Infection .................................................................................................... 20
Apoptosis .................................................................................................................... 39
Poxvirus Inhibitors of Apoptosis ........................................................................... 48
Serpins ........................................................................................................................ 56
M echanism of Serpin Action .................................................................................. 57
Poxvirus Serpins .................................................................................................... 63
The Orthopoxvirus Serpin SPI-I ............................................................................ 71

2 M ATERIALS AND M ETHODS .......................................................................... 76

Virological Techniques .......................................................................................... 76
Recombinant DNA Techniques ............................................................................ 84
Protein Analysis ..................................................................................................... 91

3 ANALYSIS OF SPI-1 SERPIN ACTIVITY IN VITRO ........................................ 99

4 DISCUSSION OF SPI-1 SERPIN ACTIVITY IN VITRO ...................................... 122

C E L L S .................................................................................................................. 133

C E L L S .................................................................................................................. 157



9 CONCLUDING REMARKS ................................................................................... 226

REFERENCE LIST ........................................................................................................ 228

BIOGRAPHICAL SKETCH .......................................................................................... 256


Table page

1. Taxonomy of the Poxviruses ................................................................................. 2

2. Properties of the Poxvirus Serpins ...................................................................... 64

3. Proposed Poxvirus Serpin Groups ......................................................................... 123

4. Virulence of RPV SPI- I and SPI-2 Mutants in Rabbits ........................................ 183

5. Summary of Rabbit Primary Lesion Histology Following Infection with WtRPV
and Serpin M utants ............................................................................................. 193

6. Virulence of WtRPV, RPVASPI-1 and RPV SPI-1 (-) in Rabbits ........................ 200


Figure paae

1. Poxvirus replication cycle ...................................................................................... 6

2. Poxvirus proteins that interfere with the host early immune response ................. 26

3. Important features of the apoptotic cascade ........................................................ 41

4. Poxvirus inhibitors of apoptosis .......................................................................... 51

5. Serpin structure and mechanism of action ............................................................ 59

6. Construction of the shuttle vector pKMSPI-1 ..................................................... 86

7. Comparison of the SPI-1 reactive site loop (RSL) with representative
inhibitory and noninhibitory serpins ................................................................... 101

8. In vitro SDS-PAGE analysis of SPI-1 activity ....................................................... 104

9. SDS-PAGE analysis of SPI- 1 activity .................................................................... 107

10. SPI-I is cleaved at or near the RSL by the proteinases chymotrypsin and
cathepsin G .......................................................................................................... 110

11. Reaction of SPI-I with cathepsin G in the presence of the competing serpin
ao-antichymotrypsin ........................................................................................... 112

12. Stability of the SPI- 1-cathepsin G complex ............................................................ 115

13. Diagram of SPI- I site-directed mutants .................................................................. 118

14. Effects of RSL mutations on the ability of SPI-1 to form a complex with
cathespin G in vitro ............................................................................................. 120

15. Scheme for creating RPV SPI-1 recombinants ....................................................... 135

16. PCR analysis of RPVASPI-1 DNA and immunoblot screening of infected
cell extracts ........................................................................................................ 138

17. SPI-1 protein expression by wtRPV and RPV SPI-1 site-directed mutant
viruses ................................................................................................................. 14 1

18. Cellular modification of SPI-1 protein is not cell line-specific .............................. 144

19. Host range of wtRPV and RPV SPI-l mutants ....................................................... 147

20. DAPI staining of A549 cells infected with wtRPV and RPV SPI-1 RSL mutants. 150

21. Cleavage of death substrates PARP and lamin A by infected cell extracts ............ 153

22. Induction of Caspase 3 activity in infected cell extracts ......................................... 155

23. PCR analysis of virus constructs used in animal experiments ................................ 174

24. Immunoblot analysis of virus constructs used in animal experiments ................... 177

25. Host range of viruses used in animal experiments .................................................. 180

26. Rabbit rectal temperature changes following infection with wtRPV or
RPV constructs containing serpin mutations ...................................................... 185

27. Rabbit body weight changes following infection with wtRPV or RPV
constructs containing serpin mutations ............................................................... 187

28. Histology of rabbit primary lesions following infection with wtRPV or
RPV serpin m utant viruses .................................................................................. 192

29. Schematic representation of the genomic SPI-1 locus in the viruses
wtRPV, RPVASPI-1 or RPV SPI-I(-) ................................................................ 198

30. Rabbit rectal temperature changes following infection with wtRPV,
RPVASPI- I or RPV SPI- I (-) ............................................................................. 202

31. Rabbit body weight changes following infection with wtRPV, RPVASPI-l
or R PV SPI-1 (-) ................................................................................................. 204

32. Comparision of the primary lesions formed following infection of animals
with wtRPV, RPVASPI-1 or RPV SPI-1 (-) ....................................................... 208


a adenosine
aa amino acid
A alanine
A Angstrom
ACT a 1 -antichymotrypsin
AIDS acquired immunodeficiency syndrome
AIF apoptosis inducing factor
APS ammonium persulfate
AT a 1 -antitrypsin
ATIII antithrombin III
bp base pair
BSA bovine serum albumin
c cytosine
C cysteine
CAM chorioallantoic membrane
CARD caspase recruitment domain
CBP chemokine binding protein
cDNA complementary DNA
cm centimeter
CPE cytopathic effect
CPV cowpox virus
CTL cytotoxic T-lymphocyte
DAPI 4' 6-diamidino-2-phenylindole
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
dsRNA double stranded RNA
DTT dithiothreitol
EDTA ethylenediaminetetraacetic acid
EEV extracellular enveloped virus
EGF epidermal growth factor
EM electron microscopy
ER endoplasmic reticulum
EV ectromelia virus
F phenylalanine
FADD Fas-associating protein with death domain
FBS fetal bovine serum
FPV fowlpox virus

g gravity
g guanosine
HA hemagglutinin
HBSS Hank's buffered standard saline
His histadine
HIV human immunodeficiency virus
HNE human neutrophil elastase
hpi hours post infection
hr hour
ICE interleukin- I 3 converting enzyme
IFN interferon
IL interleukin
IMV intracellular mature virus
ITR inverted terminal repetition
kbp kilobase pair
kDa kilodalton
LB Luria broth
LD50 lethal dose 50
M molar
MCS multiple cloning site
MCV molluscum contagiosum virus
MEM minimal essential media
mg milligram
MHC major histocompatability complex
ptg microgram
ptl microliter
jtM micromolar
min minute
ml milliliter
mm millimeter
mM millimolar
MOI multiplicity of infection
MPV monkeypox virus
mRNA messenger RNA
MV myxoma virus
N asparagine
ng nanogram
NK natural killer
nM nanomolar
nm nanometer
NOS nitric oxide synthase
nt nucleotide
ORF open reading frame
PAGE polyacrylamide gel electrophoresis
PAl plasminogen activator inhibitor

PARP poly(ADP-ribose) polymerase
PBS phosphate-buffered saline
PCR polymerase chain reaction
pfu plaque forming units
PMN polymorphonucleocyte
PMSF phenylmethylsulfonyl fluoride
PT permeability transition
R arginine
RCA regulators of complement activation
RNA ribonucleic acid
RNI reactive nitrogen intermediate
RPV rabbitpox virus
RSL reactive site loop
RT room temperature
S serine
SCCA squamous cell carcinoma antigen
SDS sodium dodecyl sulfate
SPI serine proteinase inhibitor
SPV swinepox virus
t thymidine
T threonine
TBS Tris-buffered saline
TIR terminal inverted repeat
TK thymidine kinase
TNF tumor necrosis factor
tPA tissue plasminogen activator
U unit
UV ultraviolet
V volt
VCP vaccinia virus complement control protein
VGF vaccinia growth factor
VV vaccinia virus
wt wild type

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



Kristin Laurie Moon

December, 2000

Chairman: Richard W. Moyer
Major Department: Molecular Genetics and Microbiology

Poxviruses are the only virus family known to encode functional members of the

serpin family of serine proteinase inhibitors, and the vast majority have been

demonstrated to serve as important virulence factors during infection of animal hosts.

The natural or engineered mutation of several poxvirus serpins has been shown to result

in virus attenuation associated with an impairment in the ability to cause disease. Many

orthopoxvirus members, including vaccinia virus, rabbitpox virus (RPV) and variola

virus, the causative agent of smallpox, have been shown to encode the serine proteinase

inhibitor (SPI)-I gene, implying that SPI-1 expression endows the viruses with a

selective advantage during natural infection. Previous studies have shown that SPI-1

expression is necessary for the complete host range of RPV in tissue culture, which has

been proposed to stem from the ability of SPI- I to inhibit apoptosis. Still, other studies

have questioned the importance of SPI-1 as a virulence factor during animal infections.

Because SPI-1 shares a high degree of homology with members of the serpin

superfamily, it has been widely assumed that any role that SPI-1 serves during infection

is related to its ability to function as a proteinase inhibitor. However this model has

remained untested until now. This study is the first to demonstrate that SPI-1 protein

expressed in vitro has the properties of a functional serpin and appears to preferentially

inhibit target proteinases of the chymotrypsin family, including cathepsin G.

Furthermore, experiments using RPV recombinants expressing SPI-1 proteins with

engineered mutations at amino acids critical for serpin activity are the first to suggest that

it is the ability of SPI-1 to function as an inhibitory serpin which is essential for the

unrestricted host range of RPV. While infection of restrictive cell lines with RPV

mutants lacking SPI-l serpin activity triggered several morphological features of

apoptosis, caspase activation did not occur, implying that SPI-1 may be involved in

preventing a caspase-independent form of cell death during infection. Preliminary

experiments in rabbits suggests that SPI-1 is an important virulence factor in vivo which

appears to inhibit inflammation at the primary site of infection.


Poxviruses are large, double-stranded DNA viruses that replicate exclusively in

the cytoplasm of infected cells. The virus family is divided into two subfamilies: the

Chordopoxviruses and the Entomopoxviruses, which infect vertebrates and insects,

respectively. The Chordopoxvirinae consist of eight genera: Orthopoxvirus,

Parapoxvirus, Avipoxvirus, Capripoxvirus, Leporipoxvirus, Suipoxvirus,

Molluscipoxvirus, and Yatapoxvirus (Table 1). Members of a Chordopoxvirinae genus

are both genetically and antigenically related and have a similar morphology and host

range. In contrast, the Entomopoxvirinae have been divided into three genera,

Entomopoxvirus A, B and C, based on the insect host of isolation (Table 1)(12). Until

recently, little was known regarding the insect poxviruses. However, the recent

sequencing of the genomes of two Entomopoxvirus B viruses has provided insight into

the evolutionary relationship between the two poxvirus subfamilies. Divergence between

the insect and vertebrate poxviruses is indicated by the presence of only 49 identifiable

Chordopoxvirus counterparts in Melanoplus sanquinipes with only 20-48% amino acid

identity among the shared gene products, in addition to the presence of 43 novel ORFs in

five gene families (2). Similarly, only 30% of the Amsacta moorei (AmEPV) genome

was shown to encode genes with vertebrate poxvirus homologues (22).

Table 1. Taxonomy of the Poxviruses

Subfamily Genus Representative Species

Chordopoxvirinae Orthopoxvirus Vaccinia, Variola, Cowpox, Rabbitpox,
Ectromelia, Monkeypox, Skunkpox, Racoonpox
Suipoxvirus Swinepox
Leporipoxvirus Myxoma, Shope Fibroma Virus
Capripoxvirus Sheep-pox, Goatpox. Lumpy Skin Disease Virus
Molluscipoxvirus Molluscum contagiosum virus
Parapoxvirus Orf, Psuedocowpox, Bovine papular stomatitis virus
Avipoxvirus Canarypox, Fowlpox, Quailpox, Turkeypox
Yatapoxvirus Tanapox, Yabapox
Unclassified Macropad poxvirus, Crocodilian poxvirus
Entomopoxvirinae Entomopoxvirus A Melolontha melolontha
Entomopoxvirus B Amsacta moorei. Melanoplus sanguinipes
Entomopoxvirus C Chironimus luridus

Remarkably, 35.1% of the AmEPV genome was shown to encode open reading frames

(ORFs) with no known function and no Entomopoxvirus or Chordopoxvirus homologues

(22). Still, a common Poxviridae core of genes was elucidated following the

identification of genes common to both poxvirus subfamilies, including those encoding

enzymes involved in RNA transcription and modification, DNA replication, protein

processing, virion assembly, and virion structural proteins (2,22).

Poxvirus Genomic Structure

Poxviruses have linear double stranded DNA genomes that range in size from 130

kbp (parapoxviruses) (165,221) to 300 kbp (avipoxviruses) (110). Vaccinia virus, the

prototypic orthopoxvirus, has a genome of nearly 190 kbp and encodes approximately

185 unique, nonoverlapping open reading frames (ORFs) of more than 65 amino acids

which are expressed from both strands of DNA (89). The two strands of all poxvirus

genomic DNAs are joined by hairpin loops, forming a covalently continuous nucleotide

chain (17,83). Inverted terminal repetitions (inverted terminal repeats, ITRs) which are

identical but oppositely oriented are present at both ends of the genome (80,316). Some

poxvirus ITRs include coding regions, often resulting in the presence of diploid copies of

individual genes at both ends of the genome.

Conservation of the central portion of the poxvirus genome exists among viruses

across poxvirus genera but not across subfamilies. Genes located within this central core

region of the genome generally encode components essential for virus growth and include

the DNA polymerase, enzymes responsible for nucleotide metabolism, transcription

factors, and structural proteins. Because these genes are absolutely required for

productive infection under all conditions, genes located within the conserved, central

region of the genome are often referred to as "housekeeping genes." In contrast, much

genetic variability exists within the ITRs at the termini of the various poxviruses. Genes

located within these regions of the poxvirus genome are typically not essential for virus

growth in tissue culture, but the vast majority have been shown to play a role in

promoting virus spread during infection of animals. These include many proteins which

interfere with the host innate response to infection such as inflammation mediated by

complement, interferon, cytokines and chemokines (8,259).

Molecular Biology of Poxviruses

Poxvirus morphology. Poxviruses are the largest animal viruses and are visible

using light microscopy (38). Vaccinia virus virions appear by cryoelectron microscopy

as smooth, brick shaped structures of approximately 350 nm by 270 nm (66). Fixed and

stained thin sections of virions reveal a membrane-bound, dumbbell-shaped

nucleoprotein core which contains the viral supercoiled DNA and associated proteins

together with trypsin-sensitive structures termed lateral bodies, the function of which is

not known (56). The core and lateral bodies are surrounded by glycoprotein-studded lipid


Virus attachment and entry. Study of the method of entry of poxviruses into a

host cell is complicated by the fact that the viruses exist in two major forms, both of

which are infectious. Intracellular mature virus (IMV) represents the majority of

infectious virus and is thought to be contained within a single lipid bilayer (105). IMV

remains within the cytoplasm throughout infection and is mechanically released upon cell

lysis. Extracellular enveloped virus (EEV) consists of an IMV particle with an additional

outer membrane containing at least 10 proteins which are absent from IMV, including the

viral hemagglutinin (HA) (202). The percentage of IMV versus EEV produced during

infection depends on both the virus strain and the type of cell infected. While both IMV

and EEV are infectious, it is thought that EEV is more important for spread of infection

in animals and tissue culture cells (203). Primarily, the outer membrane of EEV is

thought to protect the virus from immune attack, as EEV, but not IMV, is resistant to

neutralization by antibodies (111,295).

A round of poxvirus replication begins with binding of the virus to the host cell

by a poorly understood mechanism that likely involves association with one or more

cellular chemokine receptors (130) (Figure 1). The existence of a unique viral attachment

protein has not been established, but a recent study indicates that IMV and EEV attach to

different cellular receptors (298). Entry of IMV into the host cell is thought to occur in a

pH-independent fusion process which takes place at the plasma membrane (45,64,296).

Entry of IMV into the cell releases the DNase-resistant, membrane-bound nucleoprotein

core into the cytoplasm. In contrast, EEV enters the cell by an endocytosis event that

requires low pH to disrupt the EEV outer membrane, allowing IMV to fuse with the

endosome membrane and release the nucleoprotein core into the cytoplasm (296).

Virus gene expression. Viral early mRNA expression occurs immediately upon

release of the virus core into the cytoplasm and is directed by a DNA-dependent RNA

polymerase complex present within the infecting virion (176). This 500 kDa RNA core

polymerase complex is comprised of two large and six small virally encoded subunits and

contains all of the activities necessary to initiate nonspecific RNA transcription from any

0 Attachment
Entry/Primary D N
c uncotng Secondary replcainA
Early gene expression

DNA polymerase
Intermediate gene
trancription factors
Immune modulators


Late gene

Late gene

Structural proteins
(Nuclear factors?) Early gene transcriptio



4- (!r) 4-
Immature particle

Figure 1. Poxvirus replication cycle. The major events that take place during a single
round of poxvirus replication are depicted. Abbreviations: IMV, intracellular mature
virus; IEV, intracellular enveloped virus; CEV, cell-associated enveloped virus; EEV,
extracellular enveloped virus.

single stranded DNA template (176). Specificity for targeted transcription from poxvirus

early gene promoters requires the early transcription factors RAP94 (94 kDa RNA

polymerase associated protein) and VETF (vaccinia early transcription factor) in addition

to the core RNA polymerase (5) (140). Early gene transcription is initiated at a purine

nucleotide generally 12-17 nucleotides downstream of a conserved core early promoter

sequence (59). Termination of early gene transcripts occurs 50-70 bases downstream of

the termination sequence UUUUUNU and is dependent on the viral polymerase, VETF,

and the viral capping enzyme (222,254). Following termination, the 3' ends of early

mRNA are polyadenylated and the 5' ends are capped, activities which are again

performed by viral enzymes (82,153,176,255). These capped and polyadenylated

transcripts, which are ultimately modified in a fashion to be indistinguishable from

eukaryotic transcripts, are released from the viral core into the infected cell cytoplasm.

Splicing does not take place in any class of poxvirus mRNA. However, the presence of

cryptic splice sites in several poxvirus genes has been detected (R. Moyer, unpublished

results). Transcripts made at early times post infection are translated into a variety of

proteins, including proteins necessary for further uncoating of the virus core, growth

factors and immune modulators, and the enzymes and factors necessary for viral DNA

replication and intermediate gene transcription.

Following early gene expression, a more complete second phase of uncoating

releases a DNase-sensitive nucleoprotein core into the cytoplasm. Intermediate gene

transcription, as well as viral DNA replication (discussed below), ensues after release of

the core into the cytoplasm. Synthesis of the intermediate class of genes depends on the

presence of trans-acting factors produced at early times post-infection and on DNA

replication, but is independent of de novo protein synthesis following DNA replication.

Currently, five intermediate genes (Al L, A2L, GK1, 18 and 13) have been identified,

three of which (AlL, A2L, GK1) are known to encode late gene transcription factors

(122,304). No consensus intermediate gene promoter sequence has been demonstrated to

date, however the factors necessary for intermediate transcription are known (302,303).

These include the viral RNA polymerase, the virus capping enzyme mentioned above,

and virus intermediate transcription factors (VITF)-I and 2 (223). Of these factors

necessary for intermediate gene transcription, one, VITF-2, is known to be a cellular

protein as it can be purified from uninfected cells (224). To date, little is known

regarding the mechanism of intermediate mRNA transcript termination. The capped

intermediate transcripts are heterogeneous with typical 3' poly A ends as well as 5' poly

(A) heads of up to 30 bases in length (16). The 5' poly (A) heads are thought to result

from an RNA polymerase slippage mechanism (175), but their function, if any, is


Transcription from late viral promoters occurs following DNA replication and

the accumulation of intermediate gene products, and has been detected for up to 48 hours

following the synchronous infection of tissue culture cells (190). Many late proteins,

including the major virion components, accumulate in large amounts during this long

period of late gene expression. Other late proteins include factors necessary for early

gene transcription which become packaged in the maturing progeny virions.

Transcription of late genes requires a complex of the viral RNA polymerase and several

late transcription factors produced from intermediate genes. Late gene promoters are

comprised of a core sequence of approximately 20 bp containing a series of consecutive

A or T residues, separated by a region of about 6 bp from a highly conserved TAAAT

element within which transcription initiates (27,60,225). Similar to intermediate viral

mRNA, late mRNAs have capped, heterogeneous 5' ends containing 35-50 nontemplated

polyadenylate residues that are thought to result from RNA polymerase slippage

occurring at the AAA sequence within the late gene promoter (6,28,200,238). In

addition, late mRNAs have heterogeneous 3' ends indicating that late gene transcription

does not terminate uniformly. Since both strands of DNA are transcribed late during

infection, significant levels of double stranded RNA (ds RNA) accumulate during this


Virus DNA replication. In cells synchronously infected with vaccinia virus,

DNA replication begins 1-2 hours after infection and proceeds for at least 10 hours (282).

Replication results in the production of nearly 10,000 genome copies per cell (118).

Viral DNA replication occurs within electron-dense regions of the cell cytoplasm termed

viral factories or virosomes. The cytoplasmic site of replication and the ability of the

viral DNA to replicate within enucleated cells indicates that the virus provides most if not

all of the machinery necessary for replication (205). The viral DNA polymerase is found

as a monomer of 110 kDa and possesses both polymerase and 3'-5' exonuclease activity

(76,283). Replication is thought to initiate by the introduction of single stranded nicks

near the telomeres within the ITRs. Therefore, nicking can occur at either or both ends

of the genome. Nicking results in the formation of an exposed 3' end which is used as a

primer for strand extension by the viral polymerase. Elongation of this strand displaces

the complementary strand which is used as a template for nascent strand synthesis. The

replicated DNA strand then folds back on itself and copies the remainder of the genome.

Replication results in the production of large concatemeric intermediates with multiple

genomes organized in tail-to-tail and head-to-head arrangements (181). Concatameric

molecules are resolved into unit length genomes after the onset of late gene transcription

through a process known as telomere resolution which is mediated by the viral resolvase


Formation of virions and release from the cell. The initial stages of virion

morphogenesis occur in the electron dense poxviral factories. Staining of thin sections of

infected cells with DNA specific indicators reveals large quantities of DNA present in

what has been termed the "viroplasm" within these factories. The first evidence of virus

assembly is the presence of crescent-shaped membranous structures which will

eventually form the envelope of the IMV particle (55). The origin of these crescent

membranes is still under debate. Historically, it had been thought that these viral

membranes are formed de novo. However, later studies using labeled antibodies specific

for several subcellular compartments indicated that these membranes are derived from

cistemae of the intermediate compartment between the endoplasmic reticulum and the

Golgi stacks (261). More recent work demonstrates that crescents within poxvirus

factories contain a single membrane that can form in the absence of host organelle

membranes (105). Regardless of their origin, the crescent membranes assemble around

viroplasm to form spherical, immature particles containing a granular center. Following

production of the immature particles, the viroplasm within the particle condenses to form

a dense nucleoprotein core embedded in a granular matrix. Some studies suggest that the

nucleoprotein enters the immature envelopes just before they are completely sealed (173).

At this stage of virion morphogenesis, the progeny IMV particles are fully

infectious. These virions move out of the assembly areas to the cell periphery where a

fraction will become wrapped by additional membranes derived from the trans-Golgi or

early endosomal network that contain viral proteins destined to be in EEV. These

wrapped particles, known as intracellular enveloped virus (1EV), are transported along

actin-containing microfilaments to the plasma membrane (53). Fusion of IEV with the

plasma membrane results in the externalization of the particle, with loss of the outermost

Golgi-derived membrane. Only a small percentage of the virus which is secreted from

the cell is released as EEV. The majority of the virus adheres to the external cell surface

and is called cell-associated enveloped virions (CEV).

Poxvirus Infections of Social and Economic Importance to Man

Variola. The orthopoxvirus variola is the causative agent of smallpox, which at

its zenith, was endemic throughout the inhabited world with the exception of Australia

and certain islands (38). Two strains of smallpox virus have been distinguished based on

the severity of the disease and the mortality rate: variola major and variola minor. The

case fatality rate of variola major was usually 20% to 30% while variola minor was rarely

associated with a case fatality rate of more than 1%. People of all ages were susceptible

to infection with variola, and one attack of smallpox gave almost complete immunity to

re-infection. Because there is no efficient animal reservoir for variola, a successful

campaign by the World Health Organization was able to eradicate smallpox and the last

reported case occurred in Somalia in 1977.

There are many excellent reviews describing smallpox infection in man (72-74).

The following is a brief description of the typical course of smallpox infection. The

incubation period following infection with either strain of variola was between 10 and 14

days. The onset of sickness was acute, with fever, malaise, headache and backache. On

the third or fourth day following the onset of symptoms, the characteristic smallpox rash

appeared, first on the buccal and pharyngeal mucosa, the face, the forearms and hands

which later spread to the trunk and lower limbs. The lesions of the rash began as

macules, which soon became firm papules and then vesicles that quickly became opaque

and pustular following the influx of inflammatory cells. At about 8 or 9 days after the

onset of the rash, the pustules became umbilicated and dried up and eventually formed

scabs. In addition to the ordinary type smallpox described above, two other forms of

smallpox were recognized. The rare hemorrhagic-type smallpox, most common in

pregnant women, was associated with bleeding from the conjunctiva and mucous

membranes, severe toxemia, and early death, often before development of the rash had

occurred. Flat-type smallpox was also characterized by severe toxemia, but was

associated with a delay in the appearance of the rash and the slow development of the

skin lesions, which were usually flat and soft. This form of smallpox was often seen in

individuals with defects in cell-mediated immunity and was associated with a high

mortality rate.

Monkeypox virus. Monkeypox virus, also a member of the orthopoxvirus genus,

was first discovered in 1958 when it was isolated from smallpox-like lesions among

captive cynomolgus monkeys at the State Serum Institute, Copenhagen (301). Between

1958 and 1968, a similar disease was described in several other captive monkey colonies

in Europe and the United States, but the disease has not been reported in captive monkeys

since that time. In 1970, scientists learned that a smallpox-like disease affecting humans

living in tropical rain forest areas of the Democratic Republic of the Congo was caused

by monkeypox virus. Several outbreaks of monkeypox virus have occurred since 1970,

with all of the cases being localized to small villages within the rain forest regions of

western and central Africa. Clinically, monkeypox infections in humans resembles

smallpox, with affected individuals displaying a pustular rash, fever, respiratory

symptoms and marked lymphadenopathy. In contrast, monkeypox in humans who were

vaccinated against smallpox typically results in a milder form of the disease.

Investigation of monkeypox outbreaks in humans revealed that thirty percent of the cases

occurred in June, July or August coinciding with the period of greatest outdoor activity

such as fanning and hunting (184). Monkeypox is known to naturally infect at least eight

species of primates and four species of squirrel, and it is thought that these animals serve

as a reservoir for the virus in the wild (73).

Molluscum contagiosum virus. Molluscum contagiosum virus (MCV) is the

only member of the molluscipoxvirus genus. Although worldwide in its distribution,

molluscum contagiosum (MC) has been frequently encountered as an easily treatable

disease and has rarely been a cause of serious morbidity. The disorder mainly affects

children, sexually active adults and immunocomprimised individuals. Epidemiologic

studies suggest that transmission may be related to factors such as warmth and humidity

of the climate as well as poor hygiene (68). Transmission of MCV in children is thought

to occur by intimate skin-to-skin contact whereas in adults, MCV is most often

transmitted by sexual contact. MCV causes a benign tumor on the skin which averages

between 3.0-5.0 mm in diameter, although giant lesions of up to 1.5 cm have been

reported (91). The number of lesions in a given individual is usually less than 30, but up

to 100 lesions may coalesce to form a plaque (91). Untreated lesions in

immunocompetent individuals usually resolve within several months, although they may

persist as long as five years. Since 1980, MC has been recognized as a common disorder

among the HIV infected population. The prevalence of MC in HIV infected individuals is

estimated to be as high as 5-18% (157). The severity of MC in HIV-infected patients has

been shown to be inversely related to the CD4+ lymphocyte count (127). MCV has

proven to be difficult to study in the laboratory, as no cell culture or animal model

investigated to date can support MCV replication (248). In fact, MCV has the narrowest

tissue tropism of any poxvirus, as it is only able to replicate in the human epidermis (38).

Recently, the complete genome sequence of the type I subtype of MCV was reported

(245). Analysis of the predicted MCV ORFs revealed that the virus encodes homologues

of all orthopoxvirus genes known to be essential for virus replication (246). However,

the virus was found to lack many of the genes that allow orthopoxvirus replication in

resting cells, perhaps explaining the restriction of MCV productive infection to the

metabolically active keritinocytes of the human epidermis (246). In addition, the MCV

genome was shown to lack genes encoding homologues of many of the orthopox- and

leporipoxvirus proteins necessary for modulating the host immune response to infection.

However, the virus appears to encode several unique proteins unique which are predicted

to allow the persistent infection of MCV in the absence of inflammation (246).

Myxoma virus. Myxoma virus is the most well known member of the genus

leporipoxvirus. Two strains of myxoma virus are endemic in certain rabbit populations.

One strain is maintained in Sylvilagus brasiliensis (tapeti) in South America, while the

second strain is maintained in the bush rabbit Sylvilagus bachmani in the western regions

of North America. In its native host, myxoma virus is transmitted mechanically by

arthropod vectors and causes a localized infection characterized by benign fibromas of

the skin. In contrast, myxoma virus causes a rapid and fatal disease known as

myxomatosis in the European rabbit (Orycolagus cuniculus) which has been the topic of

many excellent reviews (73,74,77). The disease in the European rabbit is characterized

by extensive fulminating internal and external lesions and severe immunodysfunction

accompanied by secondary Gram negative bacterial infections of the respiratory tract

(77). Because of the high mortality associated with myxoma virus infection of European

rabbits, the Australian government released the South American strain of myxoma virus

into the wild in 1950 in an attempt to control the feral rabbit population in that country.

The virus was able to suppress the numbers of European rabbits in Australia for nearly

thirty years. However, over time, the rabbit population once again began to increase as

rabbits became resistant to the disease and attenuated virus strains arose from within wild

populations of 0. cuniculus. Today, myxoma-resistant rabbit populations in Australia

approach the levels found before the original virus release. Recently, a study was

published which explored the basis of myxoma virus attenuation and rabbit resistance to

infection and provides insight into the co-evolution of the virus and its host (30).

Vaccinia. The prototypical member of the orthopoxvirus genus, vaccinia virus is

the most widely characterized member of the poxvirus family and the agent used in the

vaccination effort responsible for the eradication of smallpox. Vaccination involved the

introduction of approximately 108 pfu of virus into the deltoid. A papule appeared at the

vaccination site 4 to 5 days after vaccination, which became vesicular 2 to 3 days later.

The lesion contents rapidly became turbid due to the infiltration of inflammatory cells,

and the central lesion became red and enlarged, reaching its maximum diameter by the 9th

or 1Oth day. Draining lymph nodes became swollen and patients usually developed a

minor fever. The pustule dried from the center outward, and the scab which then formed

sloughed away after approximately 3 weeks, leaving behind a scar which to this day is

the hallmark of vaccination against smallpox. Complications to routine vaccination were

rare, but included progressive vaccinia (vaccinia necrosum), eczema vaccinatum,

generalized vaccinia and postvaccinal encephalitis (74). Routine vaccination of the

general population was discontinued in 1971 and currently only military personnel and

laboratory workers are required to be vaccinated against smallpox.

Much of what is known about poxvirus replication was discovered by studying

vaccinia, which has a wide host range and can infect many cell types. In addition,

because of its ability to elicit a strong humoral and cell-mediated immune response, as

well as the large size of its genome which can accommodate many foreign genes,

researchers are studying the ability of vaccinia virus to be used as a recombinant vaccine

vector (276).

Poxvirus Replication and Spread in the Host

Poxviruses can cause either a localized, self-limiting infection associated with

little dissemination from the initial site of inoculation as is the case with Molluscum

contagiosum, or they may cause a generalized, systemic disease associated with a high

mortality rate, as is seen following infection of humans, rabbits and mice respectively

with variola, rabbitpox virus or ectromelia virus. In some cases, the same virus can be

responsible for causing both types of disease patterns but in different host species. For

example, myxoma virus causes only localized benign fibromas in its native host, S.

brasiliensis (tapeti), whereas infection of the European rabbit (0. cuniculus) causes the

rapid and fatal systemic disease myxomatosis. Because I have restricted my work to

studying rabbitpox virus (RPV) which is associated with systemic infection of the rabbit,

only the pattern of poxvirus disease caused by that virus will be discussed in detail.

Collectively, poxviruses have been shown to utilize all possible routes for

infecting the host (38). Infection via the skin likely occurs through microscopic abrasions

which allow the virus access to the epidermal or dermal layer. Some poxviruses,

including myxoma virus, depend on arthropod vectors for mechanical transmission of the

virus through the skin. Virus is carried in the biting mouthparts of the insect and

deposited in the epidermis or dermis of the host during a bloodmeal. Infection via the

respiratory tract has been demonstrated for variola, rabbitpox, vaccinia and ectromelia

viruses. In fact, epidemiological evidence suggests that smallpox was transmitted in

excretions from the mouth or nose by face-to-face contact. RPV is highly infectious by

the respiratory route, with 1 PFU sufficient for causing disease (26). The oral route of

infection has been demonstrated for ectromelia virus infection of mice and is probably

the result of cannibalism of infected mouse carcasses. However, this route is not thought

to be of major importance in the natural infection of mice by the virus.

Probably the best studied animal models for systemic orthopoxvirus

dissemination and disease in the host are myxoma and rabbitpox in rabbits and mousepox

(ectromelia) in mice. Studies in these animals produced nearly identical results

suggesting common themes in the ability of poxviruses to cause disease. Because the

pattern of spread of rabbitpox virus in the rabbit is thought to be similar to that which

occurs for other poxviruses in their hosts and since I have restricted my work to the use

of RPV, a brief description of the pathogenesis of rabbitpox virus will be given here. For

more thorough reviews, see (38) and (26).

Rabbitpox Virus

Rabbitpox virus was first reported in 1932 when it was shown to be responsible

for a spontaneous epidemic of a smallpox-like disease within a breeding colony of rabbits

housed at the Rockefeller Institute for Medical Research (95). Within two weeks, one

hundred percent of the colony had been infected and forty six percent had died. While

the exact origin of the virus is not known, a neurovirulent strain of vaccinia was being

used to infect rabbits in common areas of the breeding facility. Epidemiological studies

suggest that this virus was the source of RPV, as do sequence comparisons which show

99% nucleotide homology between RPV and vaccinia virus (31).

The outbreak of rabbitpox virus within rabbit colonies resulted in excellent

documentation of the progression of the disease (96-98). Clinically, rabbitpox infection

of rabbits was shown to resemble smallpox infection in humans. Onset of disease was

associated with malaise and fever in conjunction with enlargement of the popliteal and

inguinal lymph nodes. At two to three days post infection, papules or lesions formed on

all parts of the animal which eventually became umbilicated and necrotic. Infection of

the upper respiratory tract often occurred which was accompanied by edema in the lungs

resulting in dyspnea and obligate mouth breathing. Many of the animals developed

conjunctivitis associated with nasal mucopurulent discharges, sometimes stained with

blood. Death occurred within 8-10 days in nearly fifty percent of the cases.

In laboratory studies, rabbitpox has been shown to be an acute severe generalized

infection, the principle features of which are independent of the route of administration of

the virus (26). Following intradermal inoculation, there is a lag or incubation period of

approximately two to three days during which time the virus replicates at the local site of

infection. Initially, virus is found only at the inoculation site. However, virus soon

enters the lymphatics and reaches the draining lymph node as early as 36 hours post

infection (26). Multiplication of virus in the lymph node and the constant generation of

virus from the primary inoculation site results in necrosis of the cells of the lymph node

and the passage of virus through the efferent lymphatics to the blood stream in what is

termed a primary viremia. Virus within the blood is associated with the lymphocyte

fraction suggesting that during viremia, virus is not free but is cell-associated (26,313).

Spread of the virus through the bloodstream results in the invasion of internal organs,

including the bone marrow, spleen and liver, within 2 to 3 days after infection. In the

spleen, virus has been shown to infect lymphocytes, while in the liver, Kupffer and

paranchymal cells become infected (171). Necrosis of these infected cells releases virus

directly into the bloodstream. By day 4 or 5 post infection, this release of virus exceeds

the capacity of the reticuloendothelial system to dispose of it, and virus is once again

detected in the blood creating a "second viremia". Concurrent with this is the onset of

fever which likely arises due to the inability of the spleen to clear the virus and the

resulting systemic infection. Characteristic signs of infection, including nasal and

conjunctival discharges, weakness and lack of interest in food and water usually appear

within 2-3 days of the onset of fever (26). During this period of secondary viremia,

infectious virus can be detected in most tissues examined including the bone marrow,

intestine, nasal mucosa, ovary, vagina, uterus and skin (71). In the skin, virus multiplies

rapidly in the cells of the epidermis, hair follicles, and sweat glands, and skin lesions

appear which become ulcerated within a few days (38). It is at this stage of illness when

large amounts of virus are released into the environment that the animal is most

infectious. Death can occur as early as day 6 post infection in rabbits showing

pronounced respiratory distress. More commonly it occurs between the 8th and 12'h days

when it is frequently preceded by a sharp decrease in body temperature associated with a

severe drop in blood pressure and a rise in serum potassium concentrations to levels

shown to be lethal in experimental models (26). Animal survival is associated with but

not dependent on the presence of circulating antibody which prevents the infection of

new cells and virus content in all tissues and organs falls rapidly (71). Scabs form on the

lesions which resolve between days 10 and 12 leaving behind scarring.

Poxvirus Immune Modulators that Counteract the Host Early Response to
Virus Infection

Members of the Poxvirus family possess large DNA genomes which range in size

from 130 to 300 kbp. Because the viruses replicate exclusively in the cytoplasm of

infected cells, much of the virus genome is devoted to encoding the factors necessary for

autonomous DNA synthesis and gene expression. Still, the study of natural occurring

deletion mutants of vaccinia virus in addition to analysis of engineered virus mutations

has revealed that at least 15% to 28% of the virus genome is not required for virus

replication in tissue culture (206). The vast majority of these genes which are

"nonessential" for virus replication in tissue culture encode factors that enhance virus

infection and spread within the host. While some poxviruses are capable of establishing

persistent infections in susceptible hosts, members of the virus family generally cause

acute primary infections that often result in lifelong immunity to re-infection. The ability

of the host to mount an effective response to infection by poxviruses relies heavily on the

nonspecific arm of the immune response which is present at the time of the host's

primary exposure to the virus. In turn, poxviruses have evolved to encode a large number

of proteins that interfere with the early host mechanisms of anti-viral immunity thus

enabling efficient progeny virus replication and spread throughout the infected organism.

Current models propose that the genes encoding these poxvirus immunomodulatory

proteins were initially acquired from host cells, a hypothesis which is supported by the

high degree of homology between many of the virus proteins and their cellular

counterparts. In this section, the key components of the host innate immunity will be

addressed along with the poxvirus gene products which serve to counteract each of the

nonspecific elements. For a more thorough review, see reference (38).

Following infection with a virus, the host's first response involves the innate, or

nonspecific components of immunity, including the induction of interferons, the

alternative pathway of complement, inflammation, natural killer cells, and apoptosis.

This is followed at later times by the action of the learned responses which include

delayed type hypersensitivity, anti-viral cytotoxic T lymphocytes (CTLs) and virus-

specific antibodies. Although no specific immune response to poxvirus infection can be

detected until approximately 8 days after infection, studies have shown that the

nonspecific components of the immune system are activated almost immediately

following infection.

Interferons. One of the earliest host responses to viral infection is the production

of interferons. Type I interferons (IFNs), which include IFNat and IFNPI, can be induced

by viral infection of virtually all cell types. In contrast, IFNy, the only Type II IFN, is

solely produced by activated T-lymphocytes and natural killer (NK) cells (for a more

thorough review, see reference (117)). In tissue culture, local interferon generation has

been shown to protect neighboring cells from productive virus infection resulting in the

establishment of an antiviral state in surrounding cells. Studies in cell culture have

shown that pre-treatment of cells with interferon prior to infection leads to an abortive

infection (87). The secreted proteins act in an autocrine and paracrine fashion by binding

to specific cell receptors triggering signal transduction and transcription of IFN-

responsive genes. Two genes expressed in response to interferon encode enzymes which

are in part responsible for the antiviral effects of the cytokine. The dsRNA-dependent

protein kinase PKR, and the 2',5'-oligoisoadenylate synthetase are both produced as

latent forms in response to interferon stimulation, but are activated following exposure to

double stranded RNA produced during virus infection. Upon activation, PKR is

responsible for phosphorylating the small subunit of eukaryotic initiation factor 2

(elF2a), inactivating the factor and leading to inhibition of translation (231). The 2',5'-

oligo A synthetase catalyzes the synthesis of a family of oligonucleotides derived from

ATP which collectively activate latent RNase L. RNase L functions to cleave cellular

and viral single-stranded RNA. The end result is the inhibition of mRNA translation in

infected cells (32,119,168). In addition to establishing an antiviral state, interferon is able

to upregulate expression of MHC molecules on the surface of infected cells, which allows

enhanced detection of viral peptides by effector cells of the immune system. While IFNy

is able to upregulate expression of both MHC class I and II molecules, IFNa and 3 can

only exert an effect on MHC class I regulation (117).

In addition to the general effects of interferons listed above, IFNy also serves as a

potent immunoregulatory cytokine (70). IFNy is a powerful activator of macrophages,

which in turn produce the important pro-inflammatory cytokines TNFa and IL- 1 P, as

well as reactive nitrogen intermediates (RNIs) which function to allow more effective

intracellular pathogen killing (62). IFNy has also been shown to induce nitric oxide

synthase (NOS) production in activated macrophages, leading to enhanced production of

nitric oxide and inhibition of viral replication (121).

Studies have shown that poxvirus infection efficiently induces interferon

production both in infected animals and in tissue culture (86-88). This likely occurs

following the production of double stranded RNA, a potent interferon inducer, during the

late period of gene expression in poxvirus infected cells. During poxvirus infection,

increased levels of all types of interferons can be detected in the serum as well as in cells

taken from the site of infection, indicating that interferons are an important host response

in combating the infection (270). Experiments have shown that treatment of mice with

anti-interferon antisera resulted in a significant increase in the amount of virus recovered

from infected animals (114). In addition, administration of alpha or beta interferon to

animals prior to infection has been shown to decrease the severity of the disease as well

as the mortality rate (38). Treatment of mice with monoclonal IFNy antibodies resulted

in lethal infection by vaccinia, a virus which is normally effectively cleared by the

murine immune system (230). The specific importance of IFNy in the resolution of

poxvirus infection is exemplified by the finding that athymic nude mice, which are

normally incapable of clearing vaccinia virus infection, mount an effective immune

response to infection of a VV recombinant expressing IFNy (126,215).

Several poxvirus genes are responsible for specifically counteracting the effects of

interferon during infection (Fig. 2). Two poxvirus proteins inhibit the interferon-induced

block in mRNA translation. By binding and sequestering double stranded RNA produced

during infection, the protein product of the E3L gene prevents activation of the 2',5'-A

synthetase. The mechanism of E3L action and its effect on productive infection will be

described in a later section. The product of the vaccinia virus K3L gene shares homology

with eIF2a, the target of the dsRNA-induced PKR. The vaccinia protein acts as an

alternative phosphorylation substrate for the protein kinase, thus preventing inactivation

of elF2ca and relieving the block in RNA translation. Studies have shown that deletion of

the vaccinia K3L gene results in the increased sensitivity of virus-induced protein

synthesis and virus replication to interferon (25).

In addition to blocking the downstream effects of interferon signalling in infected

cells, many poxviruses encode soluble interferon receptors which are able to bind the

cytokine prior to engagement of the cell receptor. Genes encoding secreted IFNy

receptors (IFNyRs) have been discovered in the leporipoxviruses myxoma and shope

fibroma viruses, as well as the orthopoxvirus members vaccinia, variola, CPV, RPV, and

ectromelia (89,155,177,250,293,294). The myxoma M-T7 protein is the most studied

poxvirus IFNyR, and has been shown to share 30% amino acid identity with the

extracellular, ligand-binding domain of the cellular IFNy receptor a chain (177). The

Figure 2. Poxvirus proteins that interfere with the host early immune response. See
text for specific details. Abbreviations: CC, CC chemokine; R, receptor; NK, natural
killer; IFN, interferon; TNF, tumor necrosis factor; IL, interleukin; V, viral; CCBP,
CC chemokine binding protein; exp'n, expression; PKR, protein kinase R; ICE,
interleukin-1 converting enzyme; RCA, regulator of complement activation; MHC,
major histocompatability complex.

IFNy NK cell C4b C3b
F-- F1 TNF /
+TNF -- 'gF (TX F F
1+ Classical Alternativ
IL-1PZ vlL-1P R pathway pathway

protein is expressed early during infection and following secretion, accumulates as a

stable, soluble glycoprotein capable of binding gamma interferon with high affinity in a

species-specific manner (180). Deletion of M-T7 from myxoma resulted in severe

attenuation in animals which was associated with decreased rabbit mortality and reduced

virus dissemination from the primary site of inoculation (179).

Several poxviruses, including vaccinia (strain WR), CPV and ectromelia, have

been demonstrated to produce a functional interferon a/P receptor during infection

(52,273). Studies have shown that the vaccinia receptor, known as B I 8R, is able to bind

and functionally inhibit IFNa (52,142,273). In contrast to the poxvirus IFNyRs, BI8R is

found both in secreted and cell-associated forms, suggesting that the protein may act not

only to sequester IFNx thus preventing cell binding, but also may interfere with signal

transduction mediated by IFNa receptor binding by functioning as a cytokine sink (289).

Natural and engineered vaccinia virus B 1 8R mutants are attenuated in mice, suggesting a

role for the protein in mediating virulence in vivo (273).

Complement. The complement system is composed of a group of proteins that

are normally present in the serum in an inactive state (1,104). When activated, these

proteins participate in a coordinated series of reactions which lead to the formation of a

membrane attack complex which is capable of cell membrane lysis, as well as the

production of C5a, a factor which is chemotactic for neutrophils. The complement

cascade can be activated by different mechanisms involving two separate pathways: the

classical pathway and the alternative pathway. The classical pathway depends on the

formation of antibody-antigen complexes for initiation, while the alternative pathway is

initiated by the presence of complement component C3b, generated spontaneously at low

levels or as a by-product of the classical pathway. Execution of both pathways result in

the formation of C3 convertase which promotes activation of C5 convertase which is

responsible for catalyzing reactions leading to the assembly of the membrane attack


In addition to formation of the membrane attack complex and lysis of infected

cells, complement can lead to the efficient opsonization of foreign particles (1).

Opsonization can take place following covalent attachment of C3 degradation products to

the foreign particle. Bound C3 is then recognized by C3 receptors on leukocytes

enhancing phagocytosis of the foreign particle by these cells.

Three complement proteins (C3a, C4a, and C5a) are anaphylatoxins, and are able

to induce the release of mediators from mast cells which cause rapid increases in vascular

permeability (1). C3a and C4a receptors are expressed on mast cells, basophils, smooth

muscle cells and lymphocytes while C5a receptors are found on mast cells, basophils,

neutrophils, monocytes/macrophages and endothelium. The major effects of

anaphylatoxin binding to mast cells and basophils are granule exocytosis and release of

vasoactive mediators, such as histamine, which increase vascular permeability and

stimulation of smooth muscles. In addition, C5a is a potent chemokine and is able to

stimulate neutrophil migration and at high doses is able to cause respiratory burst and the

production of reactive oxygen intermediates. The combined action of these factors

contributes to inflammation at the site of complement activation.

Because of the many effects of complement activation, regulation of the

complement cascades are tightly regulated in vivo. Most of the regulation centers on the

formation of C3b and C4b, components crucial for activation of the downstream effectors

of both complement pathways. Complement activation on host cells is inhibited by

several membrane regulators of complement activation, or RCA. The activity of RCA

molecules are restricted to complement components of the same species. In addition to

these membrane bound regulatory proteins, soluble factors exist which serve to bind and

sequester activated complement components.

Complement has been shown to be an important host response to poxvirus

infection. In vitro studies have demonstrated that vaccinia virus can be neutralized by the

addition of a heat labile component of serum (complement) following the binding of

antibody to the virus (65,90). In a cell culture model, activation of the alternative

complement cascade has been demonstrated in fowlpox virus (FPV) infection of chicken

embryo cells as early as three hours post infection, which was associated with an

increased killing of FPV-infected cells and a decrease in virus CPE and progeny virus

yield (191). Activation of the alternative pathway has also been demonstrated following

vaccinia virus infection of murine tumor cell lines (305) and a human melanoma cell line

(193). In particular, the alternative complement pathway appears to play an important

role in combating poxvirus infection in vivo. By-products of the alternative pathway

have been demonstrated at 6 days following infection of chickens with FPV. Depletion

of circulating factor C3 by pre-treatment of chickens with cobra venom factor prior to

FPV infection resulted in an increase in the level of progeny virus recovered from

lesions, a decrease in the lesion-associated inflammatory response, and a 100% mortality

rate (192). These results clearly indicate that complement, especially the alternative

pathway, is a host response to poxvirus infection which is important in combating


Several members of the orthopoxvirus genera have been reported to encode

proteins with homology to cellular complement regulatory proteins. The most widely

studied of these poxvirus gene products is the vaccinia virus complement control protein

(VCP). VCP is an abundant soluble protein which is secreted at late times post-infection

(113). The protein shares 38% identity with the first half of mammalian C4-BP, a soluble

protein that regulates the classical pathway by binding and sequestering complement

component C4b (128). Unlike C4-BP, VCP is able to bind C3b as well as C4b and

therefore is able to inhibit both the classical and alternative pathways (162). The protein

has been demonstrated to play a role in vaccinia virus pathogenesis in vivo (113). VCP

homologues have been identified in cowpox virus (CPV inflammation modulatory

protein, IMP) and variola virus (variola smallpox inhibitor of complement enzymes,

SPICE), and initial studies indicate that these proteins are also effective inhibitors of

complement (169,170,226,249).

A second strategy of poxvirus complement evasion has recently been

demonstrated for vaccinia virus. The virus has been shown to acquire cellular RCA

molecules into the outer membrane of progeny EEV virions (297). These results help

explain previous observations that EEV, but not IMV, is resistant to neutralization by

antibody and complement (111,295,297).

Inflammation. Prior to the development of a specific immune response, the host

reacts to a virus infection by mounting an inflammatory response consisting of

vasodilation, edema and the influx of inflammatory cells (79). The first cells stimulated

to the site of an infection in a naYve animal are polymorphonucleocytes (PMNs), followed

approximately 24 hours later by cells of the monocyte/macrophage lineage. Each of

these cells are capable of engulfing and degrading foreign organisms. In addition, cells

of the monocyte/macrophage lineage are then able to present foreign antigen on their

surface for recognition by antigen-specific CTL, a step important for the generation of the

late specific immune response. Leukocytes can be attracted to the site of an

inflammatory response following a gradient of chemoattractants which are produced

locally by resident macrophages, infected cells, or as a result of vascular reactions.

Inflammatory cells, once present at the site of infection, are capable of mounting a

number of responses able to limit virus spread including phagocytosis, the generation

oxygen-dependent or independent antimicrobial products, and the release of cytokines.

Cytokines are a family of related proteins important for the complex interplay of the

elements of the immune system, including components of the inflammatory system.

Cytokines are produced by many types of cells including monocytes/macrophages and

lymphocytes, and can act in an autocrine or paracrine fashion by binding to specific

receptors on the surface of target cells and stimulating signalling cascades. In vivo,

cytokines play an integral role as mediators of natural immunity and function in the

activation, growth and differentiation of lymphocytes, as regulators of immune-mediated

inflammation, and as stimulators of leukocyte growth and differentiation.

Many studies have explored the effect of the inflammatory response on the

progression of poxvirus infections. In general, the primary inflammatory response to

infection was found to vary among the reported studies and was likely due to differences

in the viruses used, inoculation methods, and size of the inoculum. Inflammation has

been detected in the lesions following infection by vaccinia virus, shope fibroma virus

(SFV), and some attenuated vaccine strains of FPV, but not virulent forms of ectromelia

virus or MCV (38). In the case of ectromelia virus, no inflammatory response was

detected in the lesions of the liver until 7 days post infection, a time when the infection

was well under way (11). In contrast, intradermal inoculation of guinea pigs with

vaccinia virus resulted in redness and an infiltration of PMNs and mononuclear cells by

24 hours post infection (67). Similar results have been reported for vaccinia and

rabbitpox virus infection of mice and rabbits (174).

Historically, the inflammatory response to poxvirus infection has been explored

by studying the effects of infection of 12 day old chicken embryos, which lack mature B

and T cells but possess other mediators of the inflammatory response as well as

components of the complement system. Poxvirus infection of the egg at this time results

in the formation of lesions (pocks) on the chorioallantoic membrane (CAM). RPV, CPV,

and certain strains of vaccinia virus each produce red pocks following infection of the

CAM. Examination of these pocks reveal areas of ectodermal hyperplasia, an increase in

vascularization, and extensive vasodilation of the blood vessels with a complete absence

of inflammation (198). During infection of the CAM, white pocks have been shown to

occur spontaneously at a rate of 1% (75,81). Closer inspection of these white pocks

revealed that the degree of vasodilation and congestion of the blood vessels was

equivalent to that seen in the red pocks but in addition, a massive inflammatory cell

influx was present comprised of activated heterophils (neutrophils), mononuclear cells

and associated chemoattractants (49,198). The presence of the inflammatory response in

the white pocks correlated with greatly reduced levels of virus antigen and progeny virus

in the lesions compared to wild type virus pocks (198). However, despite the presence of

a massive inflammatory response to infection by the white pock mutants, virus was still

able to travel to and infect the internal organs of the embryo indicating that while

important, inflammation alone is not sufficient to clear a poxvirus infection (38). Genetic

studies on the spontaneous white pock mutants have revealed deletions within the

terminal regions of the genome (182). Researchers have since used the CAM assay as a

tool to investigate the ability of specific gene products to block inflammation. Using this

in vivo assay, researchers have discovered several poxvirus factors which are necessary

for red pock formation on the CAM, including SPI-2/crmA and a TNF receptor

homologue (197).

Many genes within the nonessential regions of the poxvirus genome are devoted

to encoding factors important in countering the host inflammatory response. In

particular, most poxviruses produce several proteins which function by binding host

cytokines and chemokines, thereby inhibiting the activation and recruitment of effector

cells to the site of infection. Several members of the poxvirus family encode proteins

which function as secreted, soluble receptors to interferons (discussed above). Similarly,

genes encoding secreted, soluble receptors of the cytokine tumor necrosis factor (TNF)

have been demonstrated in several leporipoxvirus (MYX, SFV) and orthopoxvirus (CPV,

VV, variola) family members (106,107,155,161,257,291). TNF is a potent cytokine

produced by activated macrophages and T cells which functions during the inflammatory

response by binding to specific receptors on the surface of target cells and initiating

signalling cascades. The poxvirus proteins share homology with the mammalian TNF

receptors, but lack the domains necessary for membrane insertion and intracellular

signalling. Consequently, the poxvirus receptor homologues are able to efficiently bind

and sequester the cytokine prior to its engagement of the target cell receptor. Several

lines of evidence suggest that TNF is an important host target during poxvirus infection.

Deletion of the myxoma TNF receptor homologue T2 caused significant attenuation

following rabbit infection (291). Furthermore, although the ITRs of variola virus are

significantly smaller than the corresponding regions in other poxvirus members, the

conservation of the TNF receptor homologue by the virus suggests a strong selective

pressure for retention of the protein.

Interleukin-1 (IL-1) is a cytokine produced mainly by activated macrophages in

response to infection and tissue injury. The pleiotropic cytokine mediates its effects by

binding to distinct receptors on the surface of target cells resulting in signal transduction.

The majority of the gene products which are expressed following IL-I receptor binding

have a direct role in the inflammatory and immune processes. Two forms of IL-I exist,

termed IL- Ia and IL- 13, which produce similar biological effects but are mediated by

interaction with distinct cellular receptors. In particular, regulation of IL-13 activity

appears especially important for productive poxvirus infection given that the members of

the virus family use two distinct strategies to inhibit the activity of IL-113 in vivo. The

B 15R gene of vaccinia and CPV encodes a secreted receptor that specifically binds and

sequesters IL- 113 (7,264). Like other poxvirus secreted cytokine receptors, the virus

protein contains only the extracellular ligand binding domains and lacks the motifs

necessary for intracellular signalling. In vitro studies have demonstrated that vaccinia

B 1 5R is an effective IL-113 inhibitor, as the protein is able to prevent IL-i 13-induced

lymphocyte proliferation (7,264). However, published studies have reported conflicting

evidence regarding the role of the vaccinia B15R protein during infection in vivo.

Intranasal infection of mice with a vaccinia B 1 5R negative virus resulted in increased

morbidity but decreased mortality relative to wild type virus (7). In contrast, a second

study reported that intracranial inoculation of mice with a vaccinia B 15 null mutant

resulted in a significant reduction in mortality relative to infection with wild type

vaccinia (264). Still another study demonstrated that mice infected intranasally with a

vaccinia B I 5R deficient virus exhibited an increased fever response compared to animals

infected with the wild type virus, suggesting that IL-I 3 is responsible for the fever

response during vaccinia virus infection (9).

Another method used to regulate IL- I P activity during poxvirus infection is the

inhibition of interleukin- I P converting enzyme (ICE, caspase 1) by the orthopoxvirus

serpin crmA (discussed in more detail in a later section). ICE/caspase 1 is a cysteine

proteinase which is responsible for the proteolytic activation of IL-1 I3 from its inactive

pro-form. CrmA acts as a pseudosubstrate for ICE/caspase 1, forming a stable inhibitory

complex with the proteinase which renders the enzyme inactive and thus unable to

activate IL- I 3 (217). In addition, because ICE/caspase 1 is responsible for the proteolytic

activation of IL-18, this cytokine is also regulated by crmA (85). IL-18, also known as

IFNy-inducing factor, is a cytokine that functions to stimulate the production of IFNy and

to augment the lytic activity of NK cells. The inhibition of ICE/caspase I activity by

crmA and the effects of crmA mutations on infection in vivo will be discussed below.

Chemokines represent a subfamily of the larger cytokine family. Members of this

class of proteins share the ability to stimulate leukocyte movement (chemokinesis) and

directed movement (chemotaxis). Historically, chemokines have been divided into

groups, termed C, CC, CXC, and CXXXC (where C=cysteine and X=any amino acid),

based on the spacing of conserved cysteine residues important for function. Like other

cytokines, chemokines mediate their function by binding to specific receptors on the

surface of target cells. These serpentine chemokine receptors span the cell surface seven

times and are coupled to G proteins in the cytoplasm. Unlike other cytokine receptors

which are specific for a single cytokine, chemokine receptors generally bind to all

members of a chemokine subfamily.

Poxviruses use at least two methods to inhibit chemokine action during infection.

Myxoma, RPV, CPV, variola, racoonpox, and the Lister strain of vaccinia each contain

genes encoding soluble chemokine binding proteins (CBPs) (92). Initial studies indicate

that while the proteins are able to bind both CC and CXC classes of chemokines, the

binding affinities are much higher for the CC chemokines (92,258). The poxvirus CBPs

have been shown to block the interaction of CC chemokines with their cognate cell

receptors in in vitro assays (131,258). In vivo, deletion of the gene encoding the RPV

35kDa CBP was associated with a greater inflammatory influx of monocytes and

lymphocytes within the primary lesion than that seen during wt RPV infection (92).

Genes encoding proteins with homology to mammalian serpentine chemokine

receptors have been demonstrated in swinepox, capripox and fowlpox virus genomes

(3,42,156). Though no study to date has demonstrated a role for the virus cell-associated

chemokine receptors in vivo, the proteins are predicted to compete with cellular

chemokine receptors during infection by acting as chemokine sinks and preventing the

induction of chemokine-mediated signalling cascades. A similar strategy is used by

several members of the herpesvirus family, including human cytomegalovirus,

herpesvirus saimiri, and Kaposi's sarcoma associated herpesvirus.

A third method of poxvirus inhibition of chemokine-mediated inflammation has

recently been proposed. Genomic sequencing of MCV revealed the presence of a gene

encoding a protein with homology to the CC family of chemokines (MC 148R) (246).

The MC 148R protein contains the motifs necessary for effective binding to cellular

chemokine receptors, but lacks the amino-terminal domain necessary for chemokine

activation of cells and was shown to serve as a broad-spectrum CC and CXC chemokine

antagonist (57). Current models predict that the MCV protein likely functions by

competing with host chemokines for binding to target cell receptors. Analysis of the

complete genomic sequence of fowlpox virus (FPV) revealed the presence of 4 predicted

proteins sharing homology with CC chemokines. Though still undetermined, a function

for the FPV proteins similar to the MCV chemokine antagonist is proposed.

A number of genes encoding members of the serpin family of serine proteinase

inhibitors have been demonstrated in orthopoxvirus, leporipoxviruses, suipoxviruses, and

avipoxviruses. Studies have shown that many of these poxvirus serpins are important

virulence factors and are able to interfere with the host inflammatory response in vivo.

The mechanism of serpin action and the role of the poxvirus gene products in thwarting

the host immune response to infection will be discussed in a later section.

NK cells. NK (natural killer) cells are a class of large granular lymphocytes

which mediate lysis of infected cells and tumor cells in an MHC-independent manner

(102). NK cells are activated by interferons and are capable of producing gamma

interferon after exposure to interleukin-2 (99). Within the first 2 to 3 days following viral

infection, an inflammatory response typically occurs at the site of inoculation with an

influx of leukocytes, followed by the production of interferon, activation of NK cells, and

finally NK cell-mediated lysis of virus-infected target cells (311).

In vivo studies indicate that NK cell activity is an important host response to

poxvirus infection (312). Some evidence suggests that at least 50% of the cytolytic

activity against vaccinia virus infected cells in the hamster is due to NK cells rather than

CTL (47). NK cell activity has been shown to be elevated in the spleen and blood

following infection of mice with vaccinia (267). Depletion of NK cells prior to poxvirus

infection correlates with an increased level of vaccinia virus recovered from infected

mouse spleen and liver as well as an increase in the size of the virus lesions (37).

Another study showed that C56BL/6 mice, a strain naturally resistant to infection by

ectromelia virus, produces higher levels of NK cell-mediated lysis of infected cells

compared to the susceptible DBA/2 strain of mice (114). In contrast, C57BL/6 mice

carrying the bg/bg mutation, which is associated with NK cell deficiency, contract lethal

infections when exposed to ectromelia, demonstrating a crucial role for NK cells in the

defense of the host (38). Recent studies suggest that it is not the cytolytic activity of NK

cells, but rather their production of gamma interferon that is important in combating

poxvirus infection (120). These studies indicate that while NK cells are important in

virus clearance, they are not sufficient to completely inhibit virus spread.

Although no study to date has provided experimental evidence for a poxvirus

gene product specifically designed to target NK cells, many of the proteins described

above function to inhibit the cytokines that activate NK cells and augment their lytic

function. In addition, poxviruses encode many gene products that function to inhibit

apoptosis, the form of cell death which is activated by NK cell association with target

cells. Apoptosis is a complex process, and the exact mechanisms by which it takes place

have yet to be elucidated. Though a complete review of the current models of apoptosis

is beyond the scope of this study, the key features of this form of cell death which are

relevant to the understanding of poxvirus infection will be summarized, along with a

description of the poxvirus gene products which are important in modulating the host

apoptotic response during infection.


Apoptosis is a form of programmed cell death by which organisms eliminate

unwanted cells. It is the most common form of cell death and occurs during organismal

development, tissue remodeling, cell homeostasis and immune system regulation. In

addition, dysregulation of apoptosis has been demonstrated to be involved in several

human pathologies including degenerative and autoimmune diseases, cancer and AIDS


Biochemically and morphologically, apoptotic cell death is distinct from necrosis,

a form of cell death which occurs following cell trauma (Fig. 3) (reviewed in (216)). In

apoptosis, cells release contact with neighboring cells and become detached from

surrounding tissue. Marked condensation occurs in both the nucleus and the cytoplasm,

leading to a significant decrease in cell size. Mitochondrial integrity is compromised due

to the opening of tightly-regulated pores in the organellar membrane resulting in the

destabilization of the mitochondrial inner membrane potential. Cytochrome c, a

component of the electron transport chain and normally confined to the mitochondrial

Figure 3. Important features of the apoptotic cascade. Important features of the
apoptotic cascades initiated by Cytotoxic T Lymphocytes (CTL), binding to death
receptors, or the introduction of external stimuli are highlighted. Abbreviations: FasL,
Fas ligand; TNF, tumor necrosis factor; DD, death domain; DED, death effector
domain; P.T., permeability transition; Pro, proform; C2, caspase 2; C3, caspase 3; C7,
caspase 7; C8, caspase 8; C9, caspase 9; C1O, caspase 10; Cyt c, cytochrome c; Ph.S.,
phosphatidylserine. For other abbreviations, see text.


0 Perforin
0 ~ ProD
pore 0 C8


gO of initiator
gO*0 caspases

Activation of

to outer membrane PA
Formation of apoptotic bodies ai
Cellular collapse

(FasL,TNF) External

Receptor Stimuli
(Fas,TNFR1) Cytotoxic
drugs, UV


4 Bax
4-X BcI-2


intermembrane space, is released into the cytoplasm where it functions to amplify the

apoptotic signal (discussed below). As a result of the uncoupling of the electron transport

chain, reactive oxygen species accumulate in the apoptotic cell. Within the nucleus,

chromatin condenses and is eventually cleaved intranucleosomally into -180 bp

fragments. The nuclei of apoptotic cells become invaginated and fragment to form

membrane bound vesicles often containing condensed chromatin. The plasma

membrane also begins to bleb, packaging the cellular contents into membrane bound

vesicles known as apoptotic bodies. Phosphatidylserine, normally expressed on the inner

layer of the plasma membrane bilayer, becomes translocated to the extracellular

membrane where it acts to signal nearby cells. The end result of apoptosis is the

engulfment of the dying cell by nearby cells or professional phagocytes with little or no

leakage of cellular contents or generation of an inflammatory response. In contrast to

apoptosis, necrosis is a disorderly form of cell death. Cells dying by necrosis swell rather

than condense, causing the plasma membrane to lyse and release the cell contents into the

extracellular environment. Swelling of cellular organelles leads to their lysis as well.

DNA degradation occurs, but chromatin cleavage is random and does not result in the

formation of discrete sized fragments. Because necrosis results in the disruption of the

cell membrane and leakage of intracellular contents, this form of cell death is associated

with inflammation and significant damage to surrounding cells.

Important to the process of apoptosis are caspases, a family of proteinases which

to date is comprised of 14 members (144). Caspases (cysteine-dependent Mpartate-

specific proteases) are cysteine proteinases that share an unusual and strict requirement

for substrate cleavage following aspartic acid residues. The substrate specificities of the

proteinases are further defined by the three amino acids NH2-terminal to the caspase

cleavage site (188). The larger caspase family has been further divided into three

subfamilies, based on substrate specificity, phylogenetic comparisons, sequence

similarity and function (125). Interleukin-1 13 converting enzyme (ICE, caspase 1) is the

prototype of the ICE family of proteinases, which include caspases 1, 4 and 5. It is

thought that members of the ICE caspase family serve mainly to regulate maturation of

pro-inflammatory cytokines including IL-I 13 and IL- 18 (discussed above) and any

involvement in apoptosis is minor. Caspase 2 is the sole member of the ICH-1

subfamily, and is an important regulator of neuronal apoptosis (266). Caspases 3, 6, 7, 8,

9, and 10 belong to the CPP32 subfamily, and are the major proteinases involved in the

execution of apoptosis. This third subfamily is further divided into two groups: initiator

caspases and effector caspases. The initiator caspases, including caspases 8, 9, and 10,

are the apical proteinases in the apoptotic cascade and are responsible for the proteolytic

activation of the downstream effector caspases. Caspases 3, 6, and 7 are the major

effector proteins and their proteolytic activity is responsible for many of the biochemical

and morphological changes that characterize apoptotic cell death. The list of known

substrates of the effector caspases, the so-called "death substrates", is constantly growing

and includes structural proteins such as the nuclear lamins, enzymes involved in DNA

repair including poly (ADP)-ribose polymerase (PARP), and proteins involved in cell

cycle regulation such as the retinoblastoma protein (188,216). Cleavage of these and

other specific caspase substrates results in the systematic and orderly disassembly of the

dying cell.

Although there are many different signals that activate apoptosis, in many cases

the basic pattern of events is the same. Caspases exist as inactive zymogens in healthy

cells which have not been induced to undergo apoptosis and must themselves be

proteolytically activated by cleavage at a conserved aspartic acid residue. Once the

apoptotic signal is received by the cell, the appropriate initiator caspase proform

associates with its cognate adaptor protein via shared domains (DED, death effector

domain; CARD, caspase recruitment domain) present in both the caspase zymogen and

the adaptor protein. Current data supports a model in which the adaptor protein serves to

bring two initiator caspase proforms in close proximity in order to allow intermolecular

proteolytic activation. Once activated, the initiator caspase is able to cleave and activate

downstream effector caspase zymogens enabling them to attack the apoptotic death


One means of apoptotic induction important in the host response to virus infection

is mediated through death receptor-ligand interactions (186,216,308). In the Fas pathway,

effector cells such as NK cells and cytotoxic T lymphocytes use the Fas ligand on their

surface to bind the Fas receptor of the infected target cell. Engagement of the Fas

receptor induces the trimerization of the receptor and its association with the FADD (Fas-

associating protein with death domain) adaptor protein. Pro-caspase 8 molecules then

associate with FADD, leading to the proteolytic activation of the proteinases. Activated

caspase 8 is then able to proteolytically activate other downstream caspases in a cascade-

like manner. A similar mechanism of caspase activation takes place following the

engagement of the cytokine TNF by its cognate receptor on target cells.

A related but different method of caspase activation takes place in response to cell

damage by cytotoxic drugs, serum deprivation, UV irradiation or the action of p53 (216).

In this pathway, cell damage leads to a collapse of the mitochondrial inner

transmembrane potential, leading to the opening of the permeability transition (PT) pore.

As a result, proteins which are normally localized to the intermitochondrial space are

released into the cytoplasm. Included in this group of proteins is cytochrome c which

normally functions as a component of the electron transport chain. Within the cytoplasm,

cytochrome c acts with dATP to activate the adaptor protein Apaf-1 which is then able to

interact with pro-caspase 9 molecules by means of a caspase recruitment domain

(CARD), leading to caspase 9 activation. Caspase 9 is an initiator caspase and once

activated is then able to activate downstream members of the caspase proteolytic cascade.

In addition, a second mitochondrial protein termed AIF (apoptosis inducing factor) is

released following the mitochondrial permeability transition. The protein shares with

caspases the ability to cleave substrates after aspartic acid residues, and studies have

shown that AIF is able to proteolytically activate pro-caspase 3 in vitro (271), providing

an additional pathway for effector caspase activation.

Important in the regulation of the mitochondria-dependent form of apoptosis are

members of the Bcl-2 family of proteins, which are divided into two groups. Bcl-2-like

proteins, such as Bcl-2 and Bcl-xL, promote cell survival while the Bax-like proteins,

such as Bax and Bid, are pro-apoptotic molecules. Structurally, Bcl-xL appears similar to

the bacterial pore-forming toxins such as the A subunit of diphtheria toxin (183) and

several Bcl-2 family members have been demonstrated to form ion channels in lipid

membranes (237). Channels generated by Bcl-2 like proteins have a preference for

monovalent cations and show optimal conductance at acidic pH. In contrast, Bax-like

proteins preferentially conduct anions and function over a wide range of pH. Current

models propose that pro- and anti-apoptotic Bcl-2 family members fumction in opposite

fashions to regulate polarization of the intermitochondrial space, permeability transition,

and the release of cytochrome c and AIF. In this model, the endogenous ratio of Bcl-2-

like to Bax-like proteins within a given cell is thought to predispose the cell to either

survival or apoptosis. Interestingly, members of the Bcl-2 and Bax subfamilies are each

substrates for several effector caspases. However, while caspase cleavage of Bcl-2-like

proteins has been demonstrated to destroy their ability to inhibit apoptosis, cleavage of

Bax subfamily members has been shown to promote their pro-apoptotic activity.

In addition to the methods of caspase activation mentioned above, caspases can be

activated directly following interaction with granzyme B (58,154). The serine proteinase

granzyme B is a major constituent in the cytolytic granules of CTL and NK cells.

Following interaction of CTL with the appropriate target cell, the cytolytic granules are

released into the extracellular space in a calcium-dependent fashion. One granule

component, perforin, is a pore-forming protein which directs the formation of pores in

the target cell membrane allowing the entry of granule cytolytic proteinases, including

granzyme B, into the target cell. Like members of the caspase family, granzyme B

cleaves substrates following aspartic acid residues and has been shown to cleave the

proforms of caspases 3 and 8 in vivo (13,163,320), leading to their activation and the

induction of the apoptotic cascade. A recent study revealed that granzyme B is also able

to proteolytically activate the pro-apoptotic molecule Bid. Once activated, Bid is able to

affect mitochondrial permeability transition and the release of cytochrome c and AIF,

thus providing a second method of caspase activation in cells exposed to granzyme


While there is no doubt that caspases play crucial roles in many forms of

apoptosis, recent studies reporting that several features of apoptosis can occur in the

absence of caspase activation and proposing that caspase-independent pathways of

apoptosis exist have evoked controversy. Apoptosis has been reported to occur in the

presence of z-VAD, a general inhibitor of all caspases, and the baculovirus p35 protein,

an effective inhibitor of caspases 1, 3, 6, 7, 8, and 10 (39,194,318). Several other studies

have shown that when apoptosis is induced in the presence of broad-range caspase

inhibitors, membrane blebbing, chromatin condensation, and nuclear compaction are

observed in the absence of concomitant nuclear fragmentation and DNA laddering,

suggesting that caspases are responsible for some but not all of the hallmarks of apoptosis

(158,194,318). Additionally, a recent study demonstrated that by proteolytically

activating the pro-apoptotic molecule Bid, granzyme B introduced into target cells by

CTL can promote mitochondrial collapse and cell death in the absence of caspase

activation (19). In fact, several studies have indicated that mitochondrial permeability

transition, cytochrome c release, and accompanying cell death can occur in the presence

of caspase inhibitors (101,158,172,317). However, debate exists over whether the cell

death that results in the absence of caspase activation represents a caspase-independent

form of apoptosis, necrosis, or a form of cell death which can not be correctly defined

using the current criteria.

Poxvirus Inhibitors of Apoptosis

Apoptosis is a common host response to viral infection. Expression of viral

antigens in association with MHC I molecules on the surface of an infected cell can

signal CTL or NK cell recognition. These cytolytic cells can then induce apoptosis in the

target cell by the Fas- or granzyme B- dependent mechanisms mentioned above. In

addition, infected cells themselves are able to induce their own suicide following virus

infection. Direct cellular damage caused by infection or virus replication products such

as double stranded RNA can act as triggers for a cell to undergo apoptosis. If initiated

soon after infection, elimination of infected cells by apoptosis leads to a reduction in

virus production and limits the spread of progeny virus in the host.

Because apoptosis is a common host response to viral infection, many viruses

have evolved to encode inhibitors of apoptosis. These anti-apoptotic factors are designed

to either inhibit apoptosis completely, or to delay its induction until sufficient numbers of

progeny virion have been produced. In particular, the large DNA viruses including

members of the Herpesvirus, Adenovirus and Poxvirus families have been demonstrated

to contain many genes devoted to the control of apoptosis (21,54,299).

To date, 9 poxvirus genes have been shown to function as inhibitors of apoptosis

(recently reviewed in (287)). All but one of the genes are found in the terminal regions of

the genome, and each have been demonstrated to be nonessential for virus growth in

infected tissue culture cells. However, deletion or mutation of 6 of the 9 genes results in

a restricted host range phenotype, with the mutant virus losing the ability to productively

infect certain cell types. Of the three poxvirus anti-apoptosis genes that have not been

linked to virus host range, two are found in MCV which can not be studied in cell culture.

All 9 gene products are made early during infection, prior to DNA replication. In

addition, a number of genes which are proposed to function in the inhibition of apoptosis

have recently been discovered following the complete genomic sequencing of several

poxviruses and these genes will be discussed as well.

E3L. The vaccinia virus E3L protein shares homology to eukaryotic transcription

factors (Fig. 4). It is a double stranded RNA binding protein (310) and is thought to be a

component of the vaccinia RNA polymerase (4). Vaccinia virus E3L deletion mutants

are unable to replicate in HeLa, Vero or L929 cells (23,24). In HeLa cells, the restricted

host range of VV E3L mutants is associated with the induction of apoptosis (137). In

each of the restrictive cell lines, the reduced host range can be corrected by replacing the

deleted E3L gene with other genes encoding double stranded RNA binding proteins,

indicating that it is the ability of E3L to bind double stranded RNA which is essential for

virus growth (23,132,199). The current model proposes that E3L is necessary for the

sequestration of double stranded RNA produced during late gene transcription, thereby

preventing the interferon-induced double stranded RNA-dependent serine/threonine

kinase (PKR) from phosphorylating eIF-2 to inhibit protein synthesis (46,219,314).

Furthermore, by preventing activation of PKR, E3L inhibits the kinase from activating

NF-kB, a transcription factor responsible for stimulating transcription of many pro-

apoptotic genes.

CP77. The CP77 gene is intact and functional only in cowpox virus, and is

fragmented in other orthopoxviruses. The gene encodes a 669 amino acid cytoplasmic

protein that contains ankyrin repeats, motifs important in facilitating protein-protein

interactions, particularly among integral membrane and cytoskeletal proteins. Cowpox

Figure 4. Poxvirus inhibitors of apoptosis. Poxvirus gene products that interfere with
the apoptotic cascade are depicted. Abbreviations: DD, death domain; DED, death
effector domain; TNF, tumor necrosis factor; P.T., permeability transition; Pro,
proform; Cyt c, cytochrome c; C2, caspase 2; C3, caspase 3; C7, caspase 7; C8, caspase
8; C9, caspase 9; C10, caspase 10. For other abbreviations, see text.





of initiator

S Activation of
iiI executioner
: caspases

virus, unlike vaccinia, is able to productively replicate in Chinese hamster ovary (CHO)

cells. Vaccinia virus infection of CHO cells is characterized by a rapid shut-off of viral

and host protein synthesis and apoptosis (213,214). Vaccinia virus recombinants

engineered to express the CP77 gene delay apoptosis until after progeny virions are

formed, leading to a productive infection (112,262). Although the exact function of

CP77 during infection is not known, some studies suggest that it may function in a

manner analogous to Bcl-2 family members in controlling apoptosis (112).

M-T5. The product of the myxoma M-T5 gene is a 483 amino acid protein with

homology to CP77. Like CP77, M-T5 contains ankyrin repeats. Myxoma M-T5 deletion

mutants display a reduced host range and are unable to productively infect RL-5 cells, an

immortalized CD4+ rabbit T-cell line, as well as peripheral blood lymphocytes.

Following infection of these cells, host and viral protein synthesis is rapidly inhibited and

apoptosis occurs. In vivo, M-T5 mutants are severely attenuated and virus does not

spread from the initial site of inoculation (178). Histological analysis of infected lesions

showed a significant reduction in edema and an increased infiltration of heterophils

(neutrophils) relative to wild type lesions. The current model proposes that M-T5 is

somehow necessary for inhibiting protein synthesis shutdown during lymphocyte

infection. In the absence of M-T5, apoptosis prevents productive infection of

lymphocytes which are necessary for virus spread through the body.

M-T2. The M-T2 gene is found in Shope fibroma (290) and myxoma viruses

(291). A similar gene has been found in variola and cowpox virus, but not in vaccinia

where the homologue is fragmented. The gene encodes a secreted TNF receptor that most

resembles TNFR2 and has been shown to bind with high affinity to TNFa (239,257).

Localization studies showed that the myxoma T2 protein is expressed not only as a

secreted glycoprotein, but as an endoglycosidase H-sensitive cytoplasmic glycoprotein as

well (240). The protein is made throughout infection, but is secreted predominately at

early times post infection. Myxoma virus M-T2 deletion mutants are severely attenuated

in rabbits. In addition, M-T2 deletion mutants have a reduced host range and are unable

to replicate in RL-5 cells. The host range restriction is associated with the induction of

apoptosis in infected cells. Several studies have implicated the intracellular form of M-

T2 as important for inhibiting apoptosis in RL-5 cells by a process distinct from TNF

binding (241,244). One model proposes that the intracellular form of the protein is able

to form heterodimers with a member of the TNF receptor family, inhibiting the ability of

the receptor to oligomerize and effectively signal downstream members of the apoptotic

cascade (241).

M-11L. The myxoma M-1 IL gene encodes an early protein which shares no

homology with any cellular protein. Like M-T2 mutants, myxoma M- I1L mutants are

unable to productively infect rabbit RL-5 cells. Rather, an abortive infection is observed

which is associated with the induction of apoptosis (148). Deletion of the M-1 IL gene

from myxoma greatly attenuates the virus in rabbits (195). Histological analysis of

primary lesions isolated from rabbits infected with the myxoma M- IlL mutant virus

revealed the presence of large numbers of infiltrating lymphocytes relative to wild type

lesions, suggesting that the function of M- I1L is to inhibit the inflammatory response in

vivo (195). Examination of the M- IIL amino acid sequence revealed a stretch of 18

hydrophobic amino acids near the carboxy terminus of the protein which was predicted to

allow insertion into infected cell plasma membranes. Indeed, indirect

immunofluorescence analysis of nonpermeablized cells infected with myxoma virus

demonstrated the presence of M- I1L protein on the cell surface (93). Furthermore,

myxoma virus variants containing engineered mutations in the putative transmembrane

region of M- 11 L were as attenuated as the M- I1L null mutant in rabbits, leading the

authors to conclude that the ability of M-1 IL to localize to the cell membrane is essential

for function (93). However, a recent study demonstrated that while a minority of M-1 L

may associate with the plasma membrane of infected cells, the majority of the protein

localizes to the outer mitochondrial membrane during infection(69). When transiently

expressed in uninfected cells, M- I1L was shown to behave in a Bcl-2-like fashion by

preventing mitochondrial permeability transition following cell exposure to the

apoptosis-inducing drug staurosporine, though only when M- 11L was expressed at the

mitochondrial membrane (69). Remarkably, while M- 11 L, M-T2 and M-T4 have all

been demonstrated to prevent apoptosis during infection of the RL-5 rabbit lymphocyte

cell line (20,93,244,253), only the M-1 IL gene was necessary to prevent apoptosis in

infected primary rabbit monocytes (69). This suggests expression of M-1 IL during

myxoma infection is necessary to inhibit apoptosis by preventing the mitochondrial

permeability transition, particularly in cells of the monocyte/macrophage lineage.

MC159L. MC159L, along with MC160L and MC066L, are three molluscum

contagiosum virus ORFs with anti-apoptosis activity that are not found in any other

poxvirus. MCI 59L and MCI 60L are thought to have arisen from gene duplication, thus

only MCI 59L will be discussed. The MCI 59L gene product shares sequence similarity

with the death domains (DD) of the Fas/TNFR1 signaling components FADD and

caspase 8 (29,108,245). Studies using the yeast two-hybrid system demonstrated that

MC159L is able to interact with FADD, blocking both the Fas and TNFR1 signalling

pathways of apoptosis (29,108). Because MCI 59L inhibits both TNFR1 and Fas-

mediated apoptosis but not other forms of apoptosis mediated by expression of active

caspase 8 (108), it is likely that MC159L blocks signalling upstream of caspase 8

activation. Inhibition likely occurs by MCI 59L binding to the Fas and TNFR1 adaptor

proteins through the shared death domain regions, preventing their subsequent

association with pro-caspase molecules. Because no cell culture or animal model is able

to support MCV replication, the effects of MCI 59L mutation on virus host range or

virulence in vivo have not been studied.

Selenium-containing proteins. Until recently, the product of the MCV MC066L

gene was assumed to be the only known example of a viral encoded selenium-containing

protein (252). The protein shares 75% homology to mammalian glutathione peroxidase,

an antioxidant protein responsible for reducing cytotoxic peroxides. The properties of the

MC066L protein were studied by expressing the protein in cells via a transient expression

system. The protein was demonstrated to be capable of protecting human keratinocytes

from apoptosis induced by UV irradiation and hydrogen peroxide, consistent with a role

in inactivating cytotoxic radicals (252). However, the protein was unable to protect cells

from apoptosis induced through the Fas- or TNF-mediated pathways, suggesting that

MC066L is only able to inhibit apoptosis triggered by cellular exposure to peroxide or

UV irradiation. The recent sequencing of the fowlpox virus genome revealed the

presence of a homologue of MC066L referred to as FPV064 (3). The predicted amino

acid sequence of FPV064 is reported to contain the glutathione peroxidase signature

sequence including the active site selenocysteine codon (3). The authors of the study

propose that FPV064 may serve to protect infected cells from cytotoxic effects caused

environmental stress in a manner similar to MC066L in order to allow successful virus

replication in the absence of apoptosis (3).

FPV039. Analysis of the complete FPV genome revealed a gene with homology

to the Bcl-2 family of anti-apoptosis proteins (3). The predicted amino acid sequence is

proposed to share homology with both MCLI, a protein induced during

monocyte/macrophage differentiation in myeloid leukemia cell lines, as well as BFLl, an

apoptosis inhibitor produced exclusively in the bone marrow, spleen and thymus (3).

Though no studies to date have investigated the role of FPV039 during infection, the

protein is presumed to prevent a cellular apoptotic response to infection by fowlpox virus.

In addition to the genes discussed above, two of the poxvirus genes important for

regulating apoptosis during infection belong to the serpin family of seine proteinase

inhibitors and are named Serine Proteinase Inhibitor (SPI)-I and 2. Since the research

detailed in this thesis is focused on SPI-l, a description of serpins and their mode of

action precedes a discussion of SPI-1 and SPI-2 and presentation of the research.


Serpins are a family of serine proteinase inhibitors found throughout nature. They

comprise the third major protein component of blood plasma after albumin and the

immunoglobulins (204). They are single chain proteins that share a conserved domain of

370-390 amino acids and a similar tertiary structure composed of three 1-sheets and nine

a-helipes (109). A distorted c-helix extends from P3-sheet A and contains the serpin

reactive site loop (RSL) which interacts directly with the proteinase target. The primary

function of members of the serpin family is to govern serine proteinase activity, and

serpins are involved in regulating a number of key biological processes including blood

coagulation, fibrinolysis, cell migration, and extracellular matrix remodeling (228). In

addition to inhibitory serpins, there is also a class of noninhibitory serpins, such as

ovalbumin and angiotensinogen, which are devoid of any inhibitory activity and have

evolved to fill roles other than proteinase inhibition (211).

Mechanism of Serpin Action

The inhibitory reaction of a serpin and a proteinase proceeds through a branched

pathway, wherein the serpin functions as a suicide substrate (Fig. 5). The serpin RSL,

located at the C-terminal region of the protein, closely mimics the natural substrate of the

target proteinase. The RSL comprises amino acid residues designated P15 to P5', where

proteolysis occurs between residues P1 and P1' (84). Upon recognition of the serpin RSL

as a suitable substrate by the target proteinase, an initial noncovalent, reversible

Michaelis complex is formed. Nucleophilic attack of the scissile bond between serpin

residues P1 and PI' by the catalytic serine of the proteinase results in the formation of a

covalent complex between the two proteins which is thought to exist as an acyl-enzyme

intermediate (134). At this branch point, two outcomes are possible. The ratio at which

the two outcomes occur is defined as the stoichiometry of inhibition and varies with

reaction conditions. In the substrate reaction, hydrolysis of the acyl-enzyme releases a

cleaved, inactivated form of the serpin and active proteinase. Conversely, the favored

model of inhibitory serpin action proposes that concomitant with formation of the acyl-

Figure 5. Serpin Structure and Mechanism of Action. A) The branched pathway
representing the reaction mechanism of serpins is shown. E, proteinase; Sa, active, intact
form of the serpin; ES, the noncovalent Michealis complex; E-Sa, the acyl-enzyme
intermediate present prior to proteinase translocation; E-S*, the kinetically trapped
covalent acyl-enzyme inhibitory complex; Si, the inactive, cleaved form of the serpin. B)
A diagram of the structural layout of a typical serpin is depicted with the region
corresponding to the serpin reactive site loop (RSL) highlighted. The orientation of
amino acid residues comprising the RSL from position P15 to P5' is displayed while the
active site for nucleophilic attack by the target serine proteinase is boxed.

E + Sa 4 E S ----,E-Sa



.P15-P 14-P 13-P I2-P iI-P1 O-P9-P8-P7-P6-P5-P4-P3-P2 PiP'P2'-P3'-P4'-P5S'..





enzyme complex, the amino-terminal portion of the RSL inserts into the backbone of the

serpin in a process known as strand insertion, resulting in the separation of the P1 and P1'

amino acids by 70 A (44,268). Strand insertion induces a conformational change in the

serpin which produces a form of the serpin which is much more thermodynamically

stable than the native form and is proposed to result from an increase in hydrogen

bonding within the molecule (84). Because the proteinase within the complex is

covalently bound to the serpin P 1 residue through an ester linkage, strand insertion results

in a substantial change in the orientation of the proteinase as well, involving translocation

of the proteinase by more than 70 A (135,268). The distortion of the proteinase active

site that takes place as a result of its translocation in large part explains the inhibitory

nature of the serpin-proteinase complex (164,268). The inhibitory complex is stable to

boiling in SDS, and has a long half life ranging from hours to days. The extraordinary

stability and slow decay of the of the acyl-enzyme intermediate accounts for the

inhibitory properties of the serpin. Eventually, hydrolysis of the complex occurs which

releases the cleaved, inactivated form of the serpin. Additionally, because the serpin RSL

is exposed on the surface of the molecule, it is susceptible to proteolysis by nontarget

proteinases but this consistently occurs in the absence of stable complex formation.

While the overall structure of serpins clearly plays a role in the inhibitory

mechanism of serpin action, two regions of the serpin are especially important in

determining the nature and outcome of a serpin-proteinase interaction. The first is the

reactive site of the serpin which contains the target proteinase cleavage site. Within the

reactive site, it is the P1 residue that largely dictates the target enzyme specificity of the

serpin. At the most basic level, this specificity reflects the existence of three types of

serine proteinases: trypsin-like, chymotrypsin-like, and elastase-like, which preferentially

cleave after basic, aromatic and neutral amino acids, respectively. For example, serpins

that inhibit trypsin-like proteinases often have an arginine or lysine at the P1 position,

while serpins that target chymotrypsin-like proteinases typically have methionine,

phenylalanine or tryosine at the reactive site. Serpins are remarkably specific for the

proteinases that they inhibit. The importance of the P 1 residue and serpin specificity is

best illustrated in several human diseases. In the serpin mutant acI-antitrypsin Pittsburgh,

the wild type P1 methionine residue is mutated to arginine, resulting in the inability of the

serpin to inhibit elastase. Instead, the mutant serpin gains the ability to inhibit the

trypsin-like enzymes thrombin, kallikrein, factor Xa and plasmin, resulting in a severe

bleeding disorder in the affected individual (43,84,211).

Several studies have shown that residues within the region from P3 to P3', in

addition to the P I residue, contact the target proteinase and likely play a role in serpin

specificity. In one study, the single P1 amino acid or residues P3 to P3' of a,-

antichymotrypsin (ACT) were replaced with the corresponding residues from the natural

human neutrophil elastase (HNE) inhibitor PI, converting ACT from a substrate to an

inhibitor of elastase (229). Furthermore, the ACT P3-P3' variant inhibited elastase with a

greater second order inhibition constant than was seen with the ACT P1 single mutant.

Inhibitory complexes between the P3-P3' variant and HNE were also more stable than

complexes between elastase and the ACT P1 mutant. Still, the stability of the complexes

formed between the ACT P3-P3' mutant and elastase were much less than complexes

between HNE and PI, implying a role for additional regions of the serpin in inhibitory

activity. The importance of serpin domains outside of the reactive site in determining

inhibitory activity has also been demonstrated in other studies (136,210) (33). Taken

together with modelling experiments, these results suggest that both reactive site residues

as well as the serpin scaffold are important in defining the inhibitory characteristics of a


A second region important for serpin inhibitory activity is found at the hinge of

the serpin. The serpin hinge, which includes residues P14 through P10, is the part of the

serpin molecule which bends allowing reactive site loop insertion following reaction with

the target proteinase. In inhibitory serpins, the amino acid at the P14 position is typically

seine, threonine or valine. In contrast, noninhibitory serpins such as ovalbumin or

angiotensinogen, which are devoid of inhibitory activity, have arginine at the P14

location. It is thought that the presence of the large, charged arginine residue at the hinge

of the noninhibitory serpins prevents strand insertion due to steric hindrance (84).

Without strand insertion, the interacting proteinase does not translocate across the serpin

molecule and thus stable complex formation between these serpins and potential target

proteinases can not occur. Consequently, reaction of noninhibitory serpins with seine

proteinases results in cleavage within the RSL followed by release of the proteinase. The

importance of a small, uncharged amino acid at the P14 residue for inhibitory serpin

activity has been demonstrated in several studies in which the native residue of an

inhibitory serpin was altered to arginine. In each of these studies, the resulting mutant

serpin lost its ability to inhibit its target proteinase and instead acted solely as a substrate

(135) (242). In addition, residues at positions P12 and P1O are also highly conserved

among inhibitory serpins. While alanine is typically found at both positions in inhibitory

serpins, valine/glycine and proline/glutamine residues comprise these sites in the

noninhibitory serpins ovalubumin and angiotensinogen, respectively. The importance of

uncharged amino acids at the P 10 and P12 positions is revealed by the finding that

natural point mutations at these positions in inhibitory serpins have been associated with

loss of inhibitory activity and the development of disease (208,256).

Poxvirus Serpins

Although prevalent throughout nature, poxviruses are the only virus family

known to encode functional serpins. A gene encoding a protein with homology to the

serpin family but lacking a functional RSL has been found in gammaherpesvirus 68

(300). To date, genes encoding proteins with homology to serpins have been found in

four different poxvirus genera: orthopoxvirus, leporipoxvirus, suipoxvirus, and

avipoxvirus (Table 2). Each of the serpin genes is located in the nonessential regions of

the genome. Poxvirus serpins were presumably originally derived and captured from

host cell genes by an unknown mechanism.

CrmA/SPI-2. The first discovered and most widely studied poxvirus serpin is the

orthopoxvirus crmA, known in vaccinia as ORF B13R and in RPV as SPI-2. It is a 38

kDa cytoplasmic protein which is expressed early during infection. CrmA was initially

described as a protein necessary for the production of red, hemorrhagic pocks during

infection of the CAM (discussed above) (209). CPV crmA mutants produce white pocks

on the CAM which consist of massive numbers of inflammatory cells. It was determined

that the effect of crmA on inflammation was mediated by inhibition of interleukin-IB-

converting enzyme (ICE), an enzyme responsible for processing of pro-interleukin-1 3

into its active form (217). Later studies revealed that ICE was a member of a much larger

Table 2. Properties of the poxvirus serpins
Name Genus Size(kDa) Class I ocation Function

SPI-1 Orthopoxvirus 41 Early Cytoplasmic Host range factor
SPI-2/ Orthopoxvirus 38 Early Cytoplasmic Caspase and Granzyme
crmA B inhibitor
SPI-3 Orthopoxvirus 42/50* Early Membrane- Prevents cell-cell fusion
Associated? Inhihit-,trypsin-like proteinases
SPI-7 Suipoxvirus 37 Early Cytoplasmic Virulence factor?
SerpI Leporipoxvirus 42/55 Late Secreted Inhibits trypsin-like proteinases
Prevents inflammation in vivo
Serp2 Leporipoxvirus 34 Early Cytoplasmic Prevents inflammation and
Apoptosis in vivo
Serp3 Leporipoxvirus 30** ? ? Truncated
FPVOIO Avipoxvirus 38** ? ? Undetermined
FPV040 Avipoxvirus 38** ? ? Undetermined
FPV044 Avipoxvirus 41 ** Undetermined
FPV204 Avipoxvirus 39** ? ? Undetermined
FPV251 Avipoxvirus 17** ? ? Truncated
* Size of glycosylated protein
** Predicted size of unmodified protein

family of cysteine proteinases (caspases) responsible for the regulation of apoptosis. In

fact, much of what is currently known about the apoptotic cascade is due to studies

involving crmA, which was the first natural caspase inhibitor ever discovered. CrmA has

been shown to inhibit apoptosis induced by a variety of means (34,63,78,278,306).

CrmA, which possesses an aspartic acid as its P 1 residue, has been demonstrated to

inhibit caspases 1 (ICE), 4,5, 8 and 10 (265,274,324). Because caspases 8 and 10 are

thought to be the apical caspases in the apoptotic cascade initiated by engagement of the

Fas receptor, it is thought that the ability to inhibit this pathway accounts for much of the

anti-apoptotic activity of crmA. In addition, crmA is able to inhibit the CTL serine

proteinase granzyme B (212) which, like members of the caspase family, has a preference

for cleaving after aspartic acid residues.

Within the context of a poxvirus infection, crmA is able to inhibit CTL-mediated

apoptosis mediated via the Fas receptor (147,279). In addition, researchers have

demonstrated that infection of the LLC-PK1 pig kidney cell line with a CPV crmA

deletion mutant results in the induction of apoptosis. However, the induction of

apoptosis occurs late in infection after the formation of progeny virions and does not

affect the ability of the virus to have a productive infection (218). Surprisingly, infection

of LLC-PK1 cells with wtRPV results in the induction of apoptosis, even though RPV

encodes a functional SPI-2 gene which is able to inhibit caspase I as efficiently as crmA

(146). This same study demonstrated that apoptosis was inhibited when LLC-PKI cells

were co-infected with wtRPV and wtCPV, in contrast to mixed infections between

wtRPV and a CPV crmA null mutant where programmed cell death did occur. These

results suggest that despite 93% amino acid identity between SPI-2 and crmA, the slight

differences within the serpin reactive site and scaffold regions lead to a variation in

inhibitory activity between the two proteins. In vivo, crmA does not appear to have an

effect on virulence in an mouse intranasal model of infection (123,281).

Serp2. The myxoma virus serpin Serp2 is a 34 kDa cytoplasmic protein which is

made at both early and late times during infection. Myxoma Serp2 mutants are severely

attenuated in rabbits as judged by the clinical course of the disease as well as the survival

rates of the infected rabbits relative wild type infected rabbits (167). Histologically,

lesions from animals infected with the myxoma Serp2- virus contained a massive

inflammatory response which was not seen following infection with the wild type virus.

In addition, lymphocytes in lymph nodes from animals infected with the Serp2 mutant

virus which were shown to rapidly undergo apoptosis. Based on these studies, a model

proposing that Serp2 is responsible for impairing the inflammatory response and

preventing apoptosis in lymphocytes following infection in vivo was postulated (167)

Serp2, like crmA, contains aspartic acid at the P 1 position, suggesting that Serp2

and crmA may play similar roles during infection. In support of this, Serp2 expressed in

vitro was reported to have some inhibitory activity against caspase I (ICE) (207).

However, a second study was unable to demonstrate stable complex formation between

Serp2 and caspase 1 (288). In addition, purified, recombinant Serp2 was shown to have

some activity against ICE, but inhibition was much less than was seen with crmA

produced in the same manner (Serp2 Ki of 80nM compared to 4 pM for crmA). Serp2

was demonstrated to form a stable complex with granzyme B which could be detected

following SDS-PAGE. However, enzymatic studies revealed that Serp2 was only a weak

,inhibitor of granzyme B (Ki = 420 nM) (288). In order to determine whether Serp2 could

function in place of crmA during poxvirus infection, a CPV recombinant was designed

which was deleted for crmA but expressed Serp2 (288). The mutant virus was then

assayed for the ability to inhibit apoptosis in LLC-PK 1 cells, a cell line which has been

shown to undergo apoptosis following infection with CPV crmA mutants (146,218).

Unlike cells infected with wtCPV, apoptosis was readily observed in cells infected with

the CPV Serp2 recombinant, indicating that Serp2 is unable to function in place of crmA

during CPV infection.

SPI-7. The 37 kDa SPI-7 protein is the only serpin known to date that is encoded

by the suipoxvirus swinepox. It is a cytoplasmic protein which is expressed early during

infection (185). Like crmA and Serp2, SPI-7 contains an aspartic acid at its P1 position,

suggesting that it too may have some activity against the caspases or granzyme B.

However, extensive studies have been unable to detect activity against granzyme B or

any of the caspases despite the fact that human caspase 1 is able to cleave SPI-7 in a

fashion consistent with cleavage at the P1 residue (N. Thornberry, P.C. Turner, and R.W.

Moyer, unpublished results). Replacement of crmA with SPI-7 in CPV generated a

recombinant which produced white pocks on the CAM and was unable to prevent

apoptosis during infection of LLC-PKI cells, indicating that SPI-7 is not able to function

in place of crmA within the context of CPV infection (185).

Preliminary studies indicate that SPI-7 may play a role in swinepox virulence.

Genomic sequencing of two strains of swinepox isolated from wild pigs revealed the

presence of identical SPI-7 genes, indicating conservation of the gene in nature and

suggesting a possible role in pathogenesis of the virus (185). In addition, an SPV SPI-7

deletion mutant was engineered and assayed for attenuation relative to wild type

swinepox in pigs. The mutant virus caused less extensive gross symptoms as well as

histopathology at early times post infection relative to the wild type virus and fewer

secondary lesions were seen in animals infected with the mutant (185) Infection with the

SPI-7 negative virus was also associated with a trend toward a more vigorous cellular

immune response (185). Preliminary studies suggest that SPI-7 may inhibit the function

of the proteosome, a protein complex responsible for degrading cellular and foreign

antigens for presentation on the cell surface in association with MHC I molecules (185).

These in vitro results correlate with the appearance of fewer primed SPV-specific

lymphocytes in animals infected with the SPI-7 deletion mutant relative to the wild type

virus (185).

SPI-3. SPI-3 is found in members of the orthopoxvirus genus, including vaccinia

virus, variola, CPV, RPV and raccoonpox virus (260). Unlike the majority of poxvirus

serpins, SPI-3 is N-glycosylated and contains three predicted membrane-spanning

domains although the exact location of SPI-3 in infected cells is unknown. SPI-3 is not

essential for virus growth in cell culture, and does not appear to be necessary for the

virulence of CPV, VV or RPV in Balb/c mice following intranasal inoculation (133,281).

In addition, VV, RPV and CPV SPI-3 mutants are not attenuated relative to wild type

virus following infection of the CAM and pock color was not affected.

Several studies have revealed that SPI-3 is necessary for preventing cell to cell

fusion following infection (133,286,323). Infection of tissue culture cells with CPV or

VV SPI-3 deletion mutants results in the formation of multinucleated cells, known as

syncytia. Though SPI-3 shares homology with members of the serpin superfamily,

mutation of the serpin reactive site loop had no effect on the ability of SPI-3 to inhibit

cell fusion, indicating that the ability of SPI-3 to inhibit cell fusion does not rely on its

ability to function as a serpin (286).

A recent study has revealed that though the ability of SPI-3 to inhibit cell fusion is

not related to serpin function, SPI-3 is able to act as a serine proteinase inhibitor (285).

SPI-3 was shown to form strong SDS-stable complexes with several trypsin-like

proteinases, including plasmin, urokinase, tissue plasminogen activator (tPA), as well as

weaker complexes with thrombin and factor Xa. In addition, SPI-3 protein expressed and

purified in vitro was able to directly inhibit the enzymatic activity of plasmin, urokinase

and tPA. The ability of SPI-3 to react with each of these proteinases was dependent on

the serpin motifs of SPI-3, as the RSL site-directed mutant described above was inactive

in complex formation and enzymatic assays. Taken together, these results indicate that

SPI-3 is a bifunctional protein, with separate domains responsible for serpin activity and

inhibition of infected cell fusion.

Serpl. The leporipoxvirus Serpl gene is found as one copy in malignant rabbit

fibroma virus (MRV) and as two copies in myxoma virus (MYX), but is present only as a

single fragmented copy in the attenuated shope fibroma virus (SFV) (292). Unlike other

poxvirus serpins, Serpl is expressed late during infection and is the only poxvirus serpin

which is secreted (149). Like SPI-3, SerpI is N-glycosylated and shares a P1 aspartic

acid residue. In fact, Serpl and SPI-3 have nearly identical proteinase inhibitory profiles

in vitro. Like SPI-3, Serp I is able to form SDS-stable complexes with and inhibit the

proteinases plasmin, urokinase, tPA, thrombin and factor Xa (143,187,285). Because

Serpl is a secreted protein, any one of these extracellular proteinases may be the true

target of Serp 1 during infection. Indeed, Serp 1 has been demonstrated to have both

potent anti-inflammatory activity in animal models of arthritis (151) and to inhibit

restenosis following angioplasty (145).

Serp 1 is essential for full virulence of both myxoma and malignant rabbit

fibroma viruses in rabbits (292). Infection of rabbits with a MYX SerpI mutant resulted

in significant attenuation of the virus relative to wild type MYX, with more than half of

the infected animals recovering from an otherwise lethal infection. Therefore, within the

context of MYX infection, Serpl appears to be necessary for suppressing the

inflammatory response.

Because the activities of Serp 1 and SPI-3 are nearly identical in vitro,

experiments were performed to determine whether the two serpins could substitute for

one another in the context of virus infection (307). A CPV derivative, engineered to

contain the Serp 1 gene in place of SPI-3 and under the control of the SPI-3 promoter was

assayed for the ability to inhibit cell to cell fusion in infected tissue culture cells. Unlike

cells infected with wt CPV, cells infected with the CPV Serpi recombinant displayed

significant cell to cell fusion indicating that Serp I is not able to function as a cell fusion

inhibitor. In the same study, a MYX recombinant was generated which contained SPI-3

in place of both copies of Serpl and under the control of the native Serpl promoter.

Surprisingly, analysis of protein expression in cells infected with the MYX SPI-3

recombinant revealed that SPI-3 was secreted during infection rather than remaining

intracellular as is normally observed in orthopoxvirus infections. The MYX SPI-3

recombinant was assayed for virulence in a rabbit model of infection and was found to be

as attenuated as a MYX Serp I deletion mutant. Taken together, these results indicate

that despite a nearly identical reactivity profile in vitro, SPI-3 and SerpI are not able to

substitute for one another in the context of viral infection and implies that the two serpins

may have different natural proteinase targets in nature.

Other poxvirus serpin-like genes. A recent article reporting the complete

genomic sequence of myxoma virus, strain Lausanne revealed that the virus encodes a

third gene with homology to the serpin family which they designate Serp3 (41). The

authors of the study propose that Serp3 may function as a serpin during infection, as the

predicted amino acid sequence contains conserved serpin structural elements (41).

However, analysis of the predicted Serp3 ORF reveals that the carboxy-terminal region

of the protein including the serpin RSL is truncated, suggesting that the protein is not

likely a functional proteinase inhibitor.

Fowlpox virus, a member of the avipoxvirus genus, has recently been shown to

contain five genes with homology to the serpin superfamily (3). However, analysis of the

predicted protein sequences of each of the genes demonstrates that only 4 of the 5

proposed FPV serpins are full length (FPVO 10, FPV040, FPV044, and FPV204) (Table

2). The remaining FPV gene encodes a protein with a predicted molecular weight of 17

kDa that lacks the region corresponding to the serpin reactive site loop. To date, no study

has demonstrated any role for the FPV serpins during infection.

The Orthopoxvirus Serpin SPI-1

When the spontaneous white pocks mutants of CPV and RPV were discovered

following infection of the CAM nearly forty years ago, it was noted that several of the

RPV mutants had a reduced host range and were unable to form plaques on PK- 15 pig

kidney cell monolayers (75,8 1). Genetic complementation studies revealed that the gene

necessary for restoration of the plaquing phenotype on PK- 15 cells was SPI- 1, known as

ORF B22R in vaccinia virus (10). Sequencing revealed the presence of the SPI-1 gene in

other members of the orthopoxvirus genus, including CPV and variola (123). The SPI-1

and SPI-2/crmA genes are approximately 45% identical, leading some researchers to

propose that the two genes were derived from a common ancestral viral serpin gene by

duplication and divergence within the virus genome (10). The SPI-1 gene encodes a

protein with a predicted molecular mass of 41 kDa that is expressed early during

infection and is predicted to remain within the cytoplasm of infected cells (260).

Database screening of the predicted SPI-1 amino acid sequence reveals that the protein

shares homology to the serpin family of serine proteinase inhibitors.

Studies have shown that deletion of the SPI-1 gene from CPV or VV has no effect

on pock color on the CAM or virulence in mice following intranasal inoculation

(10,123,281). In contrast, RPV SPI-1 mutants produce white pocks on the CAM, but

don't appear to be attenuated in a mouse intranasal model of infection (10,281). In

addition, RPV SPI-1 null mutants have a restricted host range and are unable to

productively infect the pig kidney PK-15 and LLC-PKI cell lines or A549 human lung

carcinoma and HeLa cervical carcinoma cells (10) and (Kristin Moon, unpublished

results). In contrast, the ability of the mutant virus to productively infect and plaque on

other cell lines, including CV-I (African green monkey kidney) and RK-13 (rabbit

kidney) cells remains unchanged.

Detailed analysis of A549 and PK-15 cells infected with an RPV SPI-1 null

mutant was performed to determine the defect in virus infection (36). The study

determined that the growth and spread of the RPV SPI-1 mutant in RK-13 cells was

indistinguishable from wtRPV, but mutant virus production from A549 and PK-15 cell

lines was less than 5% of wild type levels in the corresponding cell lines. A comparison

of virus protein synthesis and processing in A549 cells revealed no differences following

infection with either wtRPV or the RPV SPI-l mutant. In addition, both the timing and

levels of viral DNA production were similar between the wild type and SPI-1 mutant of

RPV in A549 cells. However, while the wild type and mutant viruses produced similar

numbers of virus particles in RK-13 cells, the RPV SPI-1 mutant virus was found to

produce little mature virus in A549 cells when analyzed at 24 hours post-infection, in

contrast to wtRPV. Still, transmission electron microscopy of restrictive cell lines

infected with the RPV SPI- I negative virus revealed the presence of immature progeny

virions at various stages of development at earlier times post-infection (12 hours post-

infection), from the initial membrane crescents to the IMV and IEV forms of the virus,

though little mature virus was observed. Thus, it appeared as if the initial stages of virus

morphogenesis were taking place, but that virion maturation and release of the virus was

somehow impeded in the restrictive cells infected with the SPI-1 mutant virus. The same

electron microscopy experiments also revealed that the restrictive A549 and PK-15 cells

infected with the RPV SPI-1 mutant displayed many of the morphological features of

apoptosis, including chromatin condensation and margination, marked nuclear

invagination, and karyohexis. By 24 hours post infection, approximately 94% of the

nonpermissive cells infected with the SPI-1 mutant were reported to exhibit the apoptotic

morphology. In addition to apoptotic morphology, at least 92% of A549 cells infected

with the SPI-1 mutant were demonstrated to contain the DNA fragmentation typical of

cells undergoing apoptosis. When nonpermissive cells were infected with the RPV SPI-1

mutant in the presence of cytosine arabinoside, an inhibitor of viral DNA replication and

late gene expression, no evidence of apoptosis was seen, suggesting that the infection

must proceed beyond the early phase of gene expression before apoptosis is induced.

Based on these findings, it was proposed that the inability of the RPV SPI-1 null mutant

to form plaques on nonpermissive cell monolayers was due to the induction of apoptosis

in the infected cells, which served to degrade progeny virions prior to their release from

the cells (36).

A second study has demonstrated a role for SPI-1 in inhibiting apoptosis within

the context of a virus infection (147). Cells were infected with wild type CPV or RPV, or

CPV or RPV recombinants containing mutations in SPI-1, SPI-2/crmA or both genes.

The infected cells were then assayed for their ability to inhibit apoptosis directed via the

Fas- or granule-mediated pathways. Both wild type CPV and RPV were able to resist

cytolysis employed by either mechanism. Whereas mutation of the crmAISPI-2 was

necessary to relieve inhibition of Fas-mediated cytolysis, in some cell types, mutation of

SPI-I in addition to crmA/SPI-2 was necessary to completely abrogate inhibition. In

contrast, viral inhibition of granule-mediated apoptosis was dependent on mutation of

both SPI-1 and cnnA/SPI-2. These results indicate that, like crmA/SPI-2, SPI-1 also

plays a role in inhibiting apoptosis directed by the Fas- and granule-mediated pathways

of CTL and suggests that SPI-1 and SPI-2 may work together to inhibit cytolysis of

infected cells in vivo.

Though SPI-1 has been demonstrated to confer virus host range and to inhibit

apoptosis during infection of tissue culture cells, the exact mechanism by which SPI-1

functions is unknown. While SPI-1 shares homology to the serpin family and is

presumed to act as a serine proteinase inhibitor, no biochemical activity has been

attributed to SPI-1 to date. In this study, the biochemical properties of SPI-1 were

explored by expressing the protein in vitro and assaying it for serpin activity. An RPV

SPI-l deletion mutant was constructed, along with RPV recombinants containing site-

directed mutations in the SPI-1 RSL, and the ability of each of these viruses to

productively infect permissive and nonpermissive cell lines was determined. Finally,

RPV variants containing mutations in SPI-1 alone or in parallel with mutations in SPI-2

were assayed in a rabbit model of infection to more completely characterize the effects

of serpin mutations on the virulence of RPV in vivo.


Virological Techniques

Cell Culture

CV-l (ATCC CCL-70), RK-13 (ATCC CCL-37), PK-15 (ATCC CCL-33), A549

(ATCC CCL-185) and Rat-2 (ATCC CRL-1764) cells were routinely grown in GIBCO-

BRL minimum essential medium (MEM) with Earle's salts supplemented with 5% fetal

bovine serum (FBS), 2 mM glutamine, 50 U penicillin G per ml, 50 ug of streptomycin

per ml, 1 mM sodium pyruvate, and 0.1 mM MEM nonessential amino acids (GIBCO

BRL, Grand Island, NY). LLC-PKI cells (ATCC CL-101) were grown in Medium 199

(GIBCO) supplemented with 10% FBS, 2 mM glutamine, 50 U penicillin G per ml, 50 ug

of streptomycin per ml, 1 mM sodium pyruvate, and 0.1 mM MEM nonessential amino


Production and Titration of Virus Stocks

RPV-Utrecht (ATCC-VR-157), CPV-Brighton Red (ATCC-VR-302),

CPVAcrmA (10), VV-WR (ATCC VR-I 19), vTF7.3 (ATCC VR-2153), RPV SPI-1 (-

)(10) and their derivatives were grown in either CV- 1 or RK- 13 cells. Virus stocks were

produced by infecting cells at an MOI of 0.01 in medium without serum for 2 hours with

constant rocking at 37C. When cytopathic effect was maximal, infected cells were

scraped into medium using a rubber policeman. Cells were pelleted for 5 minutes at 1000

X g. Infected cell pellets were resuspended in medium without serum and stored at

-800C until used.

Virus stocks were titered by plaquing serial ten-fold dilutions on confluent 35 mm

dishes of the appropriate cell line under a 1:1 mixture of 1.2% SeaKem LE agarose (FMC

Bioproducts, Rockland, ME) and 2X GIBCO MEM. Cells were stained at 72 hours post

infection by the addition of 0.11 mg/mi neutral red (GIBCO-BRL) diluted in medium

without serum and plaques were counted.

Sucrose Pad Purification of Virus Stocks for Use in Animal Infections

Virus used for animal studies was prepared using the following protocol. Infected

RK-13 cells grown in 150 mm dishes were scraped into the medium and collected by

centrifugation at 500 X g for 10 minutes. The infected cell pellet was resuspended in TM

buffer (10 mM Tris-HC1 pH 7.5, 5 mM MgC12, 0.2 ml per 150 mm dish) and incubated

for 10 minutes at room temperature. Cells were then lysed by 20 strokes using a Dounce

homogenizer with a tight pestle. The resulting lysate was centrifuged at 500 X g for 10

minutes to pellet nuclei and cell debris. The resulting pellet was then resuspended in TM

buffer (0.35 ml per 150 mm dish) and the cells were subjected to two more rounds of

Dounce homogenization and centrifugation. Following centrifugation, supernatants were

sonicated for 60 seconds using a Vibra-cell probe sonicator (Sonics and Materials,

Danbury, CT) set on level 2. The lysates were then layered onto a 36% (w/v) sucrose pad

and centrifuged in a Beckman SW28 rotor at 65,000 X g for 90 minutes. The resulting

pellet was resuspended in PBS (0.2 ml per 150 mm dish) and the virus stock was divided

into 200 ul aliquots which were stored at -800C until used. Accurate virus titers were

obtained by plating serial dilutions from three separate aliquots of virus as described

above. Individual aliquots were thawed once, used, and then discarded.

Construction of RPV SPI-1 Deletion Mutant (RPVASPI-1)

Semiconfluent (80%) CV-1 cells grown in 35 mm dishes were infected with wt

RPV at an MOI of 0.05 and virus was adsorbed for 2 hours at 37"C. The inoculum was

removed and the cells were washed twice with GIBCO MEM medium without serum.

1.5 ml of medium without serum was added to the cells followed by 5-10 ug of the

pKMSPI-1-gpt plasmid (described below) conjugated with Lipofectin (GIBCO-BRL).

Specifically, the DNA was diluted in water to a volume of 50 ul in an Eppendorf tube,

while 12 ul of Lipofectin was diluted with 38 ul of water in a polystyrene tube. The

diluted DNA and Lipofectin were then mixed and added dropwise to the infected cells

which were incubated at 37C. At 24 hours post infection, 1.5 ml of GIBCO-MEM

supplemented with 5% FBS was added to the cells which were incubated for another 24

hours at 37C. Transfected cells were harvested at 48 hours post infection by scraping

into the media. Crude viral stocks were subjected to three cycles of freezing and thawing

and tenfold serial dilutions were plated under a 1:1 mixture of 1.2% agarose and 2X

GIBCO MEM on CV-1 cell monolayers which had been pretreated for 6 hours in the

presence of 2.5 ug/ml mycophenolic acid, 250 ug/ml xanthine, and 15 ug/ml

hypoxanthine. At 72 hours post infection, cell monolayers were stained with neutral red

to visualize plaques. Six separate plaques containing virus resistant to mycophenolic acid

were picked and placed in I ml MEM without serum. Picked plaques were sonicated for

120 seconds and serial dilutions of the picked plaque suspensions were plated under

agarose on CV-1 cell monolayers which had been pretreated with mycophenolic acid,

xanthine, and hypoxanthine for 12 hours as described above. In this manner,

mycophenolic resistant virus was purified three times before amplification in CV-1 cells.

Deletion of the SPI-1 ORF and purity of the final virus stocks were confirmed by PCR

and immunoblot assay. A single viral isolate (RPVASPI-1) was chosen for expansion

and all future work.

Construction of RPV SPI-1 Site-Directed Mutant Viruses

Semiconfluent (80%) CV-1 cells grown in 35 mm dishes were infected with

RPVASPI-1 at an MOI of 0.05 and virus was adsorbed for 2 hours at 37C. The

inoculum was removed and the cells were washed two times with medium without serum.

1.5 ml of medium without serum was added to the cells followed by 5-10 ug of the

appropriate plasmid DNA (pKMSPI-1wt, pKMSPI-1 N321A1F322A/S323A, pKMSPI-1

F322A, pKMSPI-1 T309R) conjugated with Lipofectin in a 100 ul volume. After 24

hours at 370C, 1.5 ml of medium containing 5% FBS was added to the cells which were

further incubated at 370C for another 24 hours. Transfected cells were harvested at 48

hours post infection.

RPV recombinants containing the wild type or mutant SPI-1 genes in place of the

Escherichia coli (E.coli) eco-gpt selectable marker gene at the SPI- I locus were selected

following plaque hybridization using a 32p-labeled randomly primed probe specific for

the SPI-1 ORF. Specifically, Rat-2 monolayers grown in 100 mm dishes and infected

with approximately 1000-1500 pfu of virus were stained at 72 hours post-infection with

neutral red to visualize plaques. Each stained monolayer was then covered with an 82

mm nylon membrane (Magna). Membranes were overlayed with a 3mm Whatman paper

saturated in 50 mM Tris-HCl (pH 8.0), 1.5 M NaCI and the monolayer was transferred to

the membrane using gloved fingers. A second buffer- saturated nylon membrane was

then placed on top of the first membrane, and a dry Whatman paper was placed on top of

the filter sandwich. Capillary action allowed the transfer of the monolayer to the second

nylon membrane, forming a replica of the original master membrane. Replica filters

were wrapped in aluminum foil and stored at -20'C. Master filters were processed for

plaque hybridization by incubating first in 0.5 M NaOH, 1.5 M NaCl for 5 minutes,

followed by a 5 minute incubation in 0.5 M Tris-HCl (pH 7.4), 1.5 M NaCl. Master

filters were air dried after which the DNA was fixed to the membrane by baking in a

vacuum oven at 80'C for one hour. Filters were then incubated in hybridization solution

(6X SSC, 0.25% (w/v) dried milk) for two hours at 650C after which time heat denatured,
32p-labeled, randomly primed probe specific for the SPI-1 gene (see DNA techniques for

details on probe preparation) was added and the incubation continued overnight. Filters

were washed 4 times with 2 X SSC, 0.1% SDS for 5 minutes followed by two washes for

30 minutes with 0.1 X SSC, 0.1% SDS at 60'C with constant agitation. Air dried filters

were sandwiched between clear plastic wrap exposed to film. Plaques from replica filters

which corresponded to positively hybridizing plaques from the master filters were

aseptically removed from the filter and placed in 500 ul MEM without serum. Plaque

suspensions were sonicated twice for 60 seconds using a Vibra-Cell probe sonicator on

setting 2, and dilutions of 10 0, 10-1 and 10-2 were plated on 100 mm dishes, and the

above protocol repeated until a pure virus stock was obtained. Replacement of eco-gpt

with the wild type or site-directed SPI-1 genes was confirmed by PCR, sequencing, and

immunoblot analysis.

Construction of RPV SPI-2 Insertion Mutant Viruses

CV-1 cells grown in 35 mm dishes were infected with virus (wt RPV, RPVASPI-

1, or RPV SPI-1 F322A) at an MOI of 0.05 for two hours at 37C with constant rocking.

Cells were washed twice with MEM without serum. 1.5 ml of medium without serum

was added to the cells followed by 5-10 ug of RPVASPI-2::lacZ DNA conjugated with

Lipofectin. At 24 hours post infection, 1.5 ml of MEM supplemented with 5% FBS was

added to the cells. At 48 hours post infection, cells were harvested as described

previously. Serial tenfold dilutions of the crude virus stocks were plated on CV-1 60 mm

dishes under a 1:1 mixture of 1.2 % agarose and 2X MEM. At 48 hours post-infection,

cells were stained with X-gal (300 ug/ml, diluted from a 20 mg/ml stock prepared in

DMF). Well-isolated blue plaques were picked and resuspended in 1 ml of PBS-AM.

Plaque suspensions were sonicated for 120 seconds using a Vibra Cell Sonicator with cup

attachment (Sonics and Materials, Danbury, CT), and 10"', 10-2 and 10"3 dilutions of the

samples were plated on CV-1 60 mm dishes. This procedure was repeated 3-4 times until

all of the plaques produced from an isolate stained blue after which time the isolate was

expanded to produce a high titer virus stock. Insertional inactivation of the SPI-2 gene

was confirmed by PCR and immunoblot analysis.

Virus Host Range Assay

RK-13, A549 and PK-15 cell monolayers grown in 60 mm dishes were infected

with approximately 100 pfu of virus in MEM without serum. After two hours of

incubation at 37C, the virus inoculum was removed and the infected cells were overlaid

with a 1:1 mixture of 2X GIBCO MEM and 1.2% agarose. At 72 hours post-infection,

the agarose overlays were removed and the infected cell monolayers were stained with

crystal violet to visualize plaques.

DAPI Staining of Infected Cells

A549 cells were grown in LabTek eight-well chamber slides (Nunc, Naperville,

IL) to 60% confluence and infected with virus at an MOI of 10 in 100 ul of MEM without

serum. Virus was adsorbed for 2 hours at 370C. After removal of the inoculum, cells

were washed with 300 ul of medium without serum and placed at 37C. At 18 hours

post-infection, medium was removed and cells were washed once with PBS. Cells were

fixed by incubating for 20 minutes in PBS containing 3.5% paraformaldehyde at room

temperature. Cold methanol was then added to permeabilize the cells which were further

incubated for 10 minutes at room temperature. Cells were rinsed twice with PBS

followed by the addition of PBS containing 0.5 ug/ml DAPI (4'6-diamidino-2-

phenylindole) and incubation for 30 minutes at room temperature. Cells were washed

three times with PBS. DAPI-stained DNA was visualized using a fluorescent lamp and

DAPI filter and fluorescent cells were photographed using Fuji 400 ASA film.

Intradermal Infection of Rabbits

Adult, -5 pound female New Zealand White rabbits were obtained from Myrtle's

Rabbitry (Thompson Station, TN ) and were housed in individual cages in infectious

isolation cubicles at the Animal Resource Facility at the University of Florida. Virus

inoculums were diluted in 100 ul of PBS and injected intradermally into the shaved right

or right and left flanks of each rabbit using a 27-gauge needle and tuberculin syringe.

Each animal was weighed on a daily basis and examined for clinical signs of disease.

The diameter and induration of the primary site of inoculation was measured daily and

photographed periodically. Rectal temperatures were measured using a digital

thermometer (Becton-Dickinson, Franklin, New Jersey). Rabbits were sacrificed by

intravenous overdose of Xylezene.

Histology of Infected Lesions

Primary lesions were removed following animal sacrifice and placed in 10%

buffered formalin. 5 mm sections were made through the center of the lesion, leaving a

border of normal skin adjacent to the infected tissue, placed in tissue cassettes, and

embedded in paraffin. The embedded lesions were sectioned at 5 gxm and stained with

hematoxylin and eosin. Each section was examined by two impartial veterinary


Recombinant DNA Techniques


pGEM 3ZF::PARP. A cDNA clone of human poly-(ADP-ribose) polymerase

(PARP) (kindly provided by Dr. Alexander Burkle, German Cancer Research Center,

Heidelberg, Germany) was cloned by M. Teresa Baquero into the plasmid pGEM 3ZF (-)

using Sma I, oriented such that PARP could be expressed from the PT7 promoter.

pALTER Exl::Lamin A. A cDNA clone of human Lamin A ( a gift from Dr. B.

Burke, University of Calgary, Calgary, Alberta, Canada) was subeloned into the Eco RI

and Xba I sites of the plasmid pALTER-Exl oriented such that Lamin A could be

expressed from the PT7 promoter and was kindly provided by M. Teresa Baquero.

pALTER Exl::SPI-1. This clone was constructed to enable expression of the

RPV SPI-1 gene under the control of the PT7 promoter. The SPI-l gene was amplified

from RPV genomic DNA by PCR using primers SM 227 (5' ggg gcc atg gat atc ttt aaa

gaa cta atc 3') and SM 228 (5' ggg gat cct tat tgc gga tag cag tat ttc 3'). The PCR

product DNA was digested with Nco I and Barn HI and cloned into the corresponding

restriction sites in pALTER-Ex] plasmid DNA.

pKMSPI-1. The shuttle vector pKMSPI-1 was designed so that the SPI-1 gene of RPV

could be replaced by any gene cloned between the SPI-l left and right flanks contained

within the plasmid (Figure 6). Briefly, the 235 base pairs (bp) directly upstream of the

SPI-l ORF (SPI-1 left flank) were amplified from wtRPV genomic DNA

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EGZR4YQY5_DZCIH2 INGEST_TIME 2014-06-17T18:18:18Z PACKAGE AA00022098_00001