The Contribution of the RPV B5R protein to virus virulence

MISSING IMAGE

Material Information

Title:
The Contribution of the RPV B5R protein to virus virulence
Physical Description:
Unknown
Creator:
Stern, Richard J., 1963-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 28144882
oclc - 48925946
System ID:
AA00026800:00001


This item is only available as the following downloads:


Full Text










THE CONTRIBUTION OF THE RPV B5R PROTEIN TO VIRUS VIRULENCE












By

RICHARD J. STERN















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

UNIVERSITY OF FLORIDA

1996
























Copyright 1996

by

Richard James Stern


























To Rita, who saved me from myself and without whose love and friendship

this journey would have been unbearable.
























Its cold comfort
To the ones without it
To know how they struggled
And how they suffered about it.

If their lives were exotic and strange
They would likely have gladly exchanged them
For something, a little more plain
Maybe something, a little more sane

We each pay a fabulous price
For our visions of paradise
But a spirit with a vision,
Is a dream, with a mission.

-"Mission", Rush















ACKNOWLEDGMENTS


Although only a single name appears on this dissertation, the making

of a PhD. is a process that involves a large number of people in a dozen

different ways. Anyone who says otherwise is either lying or is blinded by

their own ignorance. In acknowledgment of this, I would like to take this

opportunity to thank those people who I feel contributed to my education in

some significant way.

First, and most importantly, I would like to thank my wife, Rita, for

her love and unending support on this flight of fancy as well as for her

patience in waiting for me to come alive again. Her belief in me carried me

through considerable periods of time when I took this all much too seriously

and lost The Way. I will forever admire the strength she showed during this

period of our lives. In addition, I would like to thank my friends and family

for their encouragement, support and for smiling politely while I explained

what it is I do. A heartfelt thanks goes to Bob and Anne Badger for the

friendship, for housing me during the finale and, well...for everything. What

more can I say?

I would like to thank Dick Moyer for the opportunity to work in his lab

as well as for his advice and guidance in this work. Oh yeah, and thanks for

v








the pretzels. In addition, I thank the remaining members of my committee,

Maureen Goodenow, Paul Gulig and Maurice Swanson for their guidance in

this work and for their deftness at turning the screws when it became

necessary. I would also like to thank Jim Thompson for his suggestions on

this project and for his considerable help getting me started on the animal

work. A special thanks goes to Parker Small for his insightful advice and

guidance during this odyssey as well as for his infectious enthusiasm for

science, teaching and life in general.

I owe a good deal of thanks to all members of the Moyer continuum for

their suggestions, advice and criticisms and for the many hours spent kicking

around ideas just to see where they would lead. In particular, I wish to thank

Michael Brooks and Pierre Musey for the many interesting and enlightening

discussions, scientific and otherwise, and for frequently calling "bullshit".

They kept me honest. Special thanks go to Michael Duke and Dorothy Smith

for keeping everything in the lab running smoothly as well as for the

friendship we shared during my stay. I would particularly like to thank Mike

Duke for the many hours of philosophical discussion, the laughter, the

homebrews and the humble sense of balance and common sense wisdom he

brought to the lab. Of course, being a Libra, he could hardly do otherwise.

I am indebted to the staff of the Department of Molecular Genetics and

Microbiology for making the process work smoothly. In particular, I would

like to thank Peggy Kidder, Vicky Parrot, Beverly Anderson and Brad Moore


vi








for placing orders "ASAP" with a smile and for treating me kindly the many

times I asked, "What form?". My utmost gratitude goes to Joyce Conner for

her excellence in making sure I was registered and in compliance with the

rules of the Graduate School. Thanks to the efforts of these people, for seven

years I was able to remain blissfully ignorant of the rules and how to follow

them.

Finally I wish to acknowledge Victor Sapirstein for encouraging me to

take this journey and George Stone and Ruth Gubitz who warned me about

the whole thing in the first place. I finally understand.




























vii














TABLE OF CONTENTS


ACKNOWLEDGMENTS ............................................ v

LIST OF TABLES ...................................................... x

LIST OF FIGURES .................................................... xi

GLOSSARY OF ABBREVIATIONS .................................. xiv

ABSTRACT ....................................................... xvii

CHAPTERS

I INTRODUCTION ......................................... 1
Introduction ............................................ 1
Poxvirus Classification ............................... 6
Poxvirus Morphology .................................. 8
The Poxvirus Genome and its Organization ............... 14
Poxvirus Biology ....................................... 19
The Replication Cycle ............................ 19
Some Musings on the Significance of Enveloped
Virus .......................................... 41
Poxvirus Pathogenesis ................................. 43
Non-specific Host Responses to Infection ........... 48
Viral Proteins that Modulate the Host Immune
Response ...................................... 64
The B5R Protein of RPV ............................... 70


II MATERIALS AND METHODS ........................... 84
Recombinant DNA Techniques ......................... 84
Virological Techniques ................................. 88
Construction of Mutant Viruses ......................... 98
Protein Techniques ................................... 110




viii










III THE INVOLVEMENT OF THE RPV B5R PROTEIN IN
HOST RANGE VIRULENCE AND INFLAMMATION ..... 117
Introduction .......................................... 117
Results ............................................... 122
Cell Culture ..................................... 122
Animal Work ................................... 135
Discussion ........................................... 178


IV THE IMPORTANCE OF THE MEMBRANE-BOUND AND
SECRETED FORMS OF B5R IN MORPHOGENESIS AND
POCK MORPHOLOGY ................................. 191
Introduction .......................................... 191
Results .............................................. 195
Construction of a Mutant Producing only a
Secreted Form of the B5R Protein ................ 195
Construction of Mutants Producing only
Membrane Bound B5R Protein ................. 221
Discussion ........................................... 241


V SUMMARY, DISCUSSION AND FUTURE DIRECTIONS .... 253
Summary ............................................. 253
Future Directions ................................... 271

LIST OF REFERENCES ............................................ 275

BIOGRAPHICAL SKETCH ........................................ 295

















ix














LIST OF TABLES


Table Page


1 Homologies of the B5R Homologues ........................ 74

2 Symptomology Resulting from Poxvirus Infection of Rabbits ... 141

3 Summary of the Pathology Produced by RPV and RPVAB5R .... 146

4 Summary of the Pulmonary Pathology in Infected Mice ........ 153

5 Virus Titers from Lungs of Infected Mice ..................... 162

6 Histopathology of the CAM Resulting from Virus Infection .... 171

7 Symptoms resulting from Poxvirus Infection of Rabbits ......... 224






















x














LIST OF FIGURES


Figure Page


1 The Structure of a Poxvirus .................................. 10

2 Organization and Structure of the Poxvirus Genome ............. 17

3 The Life Cycle of a Poxvirus .................................... 21

4 Model for Replication of the Poxvirus Genome .................. 35

5 Alignment of the Orthopoxvirus B5R Homologues .............. 73

6 Schematic Representation of the B5R Protein .................... 77

7 The B5R Protein is a Member of the SCR Superfamily ............. 79

8 Regions of the B5R Protein Sharing Homology with
Members of the SCR Superfamily ............................. 81

9 Topology of the B5R Protein of RPV ............................ 101

10 Scheme for Construction of RPVB5R-T ........................ 104

11 Location of the Primers used to Construct the RPVB5R:X
Mutants ..................................................... 108

12 Construction and Characterization of RPVAB5R ................ 124

13 Analysis of RPV and RPVAB5R Infected Cell and
Supernatant Extracts for the Presence of the B5R Protein ........ 127

14 Plaque Formation of RPV and RPVAB5R on Various Cell
Lines ....................................................... 130



xi









15 The Effect of the B5R Gene on the Ability to Form Progeny
Virus ....................................................... 134

16 RPV and RPVAB5R Pock Phenotype on the CAM ............... 137

17 Rectal Temperatures in RPV and RPVAB5R Infected Rabbits ...... 140

18 Histopathology of RPV and RPVAB5R Infected Rabbit Skin ....... 145

19 Weight Change in Mice Following Intranasal Infection ........... 149

20 Lung Sections from Intranasally Infected Mice .................. 152

21 Growth of RPV and RPVAB5R in Mouse Lung ................. 157

22 Effects of Dexamethasone on Virus Replication in Mice .......... 160

23 NBT Staining of CAMs ....................................... 167

24 Histopathology in the CAM following Infection with Virus ....... 169

25 Low Power Magnification of RPV and RPVAB5R Pocks ........... 174

26 Effect of Dexamethasone on Pock Color in Eggs ................. 177

27 Construction and Characterization of RPVB5R-T ............... 198

28 Peptide-N-glycosidase Treatment of B5R and B5R-T Proteins ...... 201

29 Western Blot Analysis of IMV, IEV and EEV ................... 204

30 Western Blot Analysis of Na2CO3 Treated RPV or
RPVB5R-T EEV ........................................... 207

31 Plaque Size and Host Range of RPVB5R-T ...................... 210

32 The Ability of RPVB5R-T to Produce Virus in RK-13 and
CEF Cells .................................................... 214

33 Behavior of RPVB5R-T on the CAM ........................... 217

34 Weight Change in RPV and RPVB5R-T Infected Mice ............ 220

35 Rectal Temperature of Infected Rabbits ......................... 223

xii










36 Estimation of the Cleavage Site of B5R .......................... 228

37 Construction and Characterization of the RPVB5R:X
Mutants .................................................... 233

38 Plaque Phenotype of the RPVB5R:X Mutants on RK-13
Cells ....................................................... 237

39 Plaque Formation on CEF Cells by the RPVB5R:X Mutants ........ 239

40 Pock Morphology of the RPVB5R:X Mutants ................... 243


































xiii














GLOSSARY of ABBREVIATIONS




AA Amino acid
bp Base pair
1B-ME Beta-mercaptoethanol
BSA Bovine serum albumin
BudR Bromodeoxyuridine
C Centigrade
CAM Chorioallantoic membrane
cm Centimeter
CPE Cytopathic effect
CPV Cowpoxvirus
CsCl Cesium chloride
dATP Deoxyadenosine triphosphate
dCTP Deoxycytosine triphosphate
dGTP Deoxyguanosine triphosphate
dTTP Deoxythymidine triphosphate
DMSO Dimethyl sulfoxide
EEV Extracellular Enveloped Virus
EDTA Disodium ethylenediamine tetracetic acid
FBS Fetal bovine serum
G Gravity
gm Grams




xiv









gpt Guanyl-ribosylphosphotransferase
HCL Hydrochloric acid
IEV Intracellular Enveloped Virus
ITR Inverted terminal repeat
IMV Intracellular Mature Virus
kb Kilobase
kbp Kilobase pairs
kDa Kilodalton
kg Kilogram
mAb Monoclonal antibody
mCi Millicuries
MEM Minimal essential medium
MeOH Methanol
mg Milligrams
m.o.i. Multiplicity of infection
ml Milliliter
mm Millimeter
mM Millimolar
MPA Mycophenolic acid
NBT Nitroblue tetrazolium
NaCl Sodium chloride
Na20D3 Sodium carbonate
NaOH Sodium hydroxide
NaN3 Sodium azide
NP40 Nonidet P-40
ORF Open reading frame
pM Picomolar
PAGE Polyacrylamide gel electrophoresis




xv









PBS Phosphate buffered saline
PEG Polyethylene glycol
pfu Plaque forming unit
PPO 2,5-diphenyl-oxazole
RPV Rabbitpox virus
SDS Sodium-dodecyl-sulfate
THAM Tris-hydroxymethyl aminomethane
Tris Tris-hydroxymethyl aminomethane
p 1 Microliter
V Volts
V V Vaccinia virus
v / v Volume-to-volume
w/v Weight-to-volume



























xvi











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



CONTRIBUTION OF THE RPV B5R PROTEIN TO VIRUS VIRULENCE
By


Richard J. Stern


August, 1996






Chairperson: Richard W. Moyer, Ph.D.
Major Department: Molecular Genetics and Microbiology



Poxviruses produce a variety of proteins that modulate the host

immune response. The predicted product of the rabbitpox virus (RPV) B5R

gene, which exists in a membrane-bound and secreted form, has homology to

complement factors involved in regulation of inflammation. This homology

suggests a mechanism by which the B5R protein might influence virus

virulence. This study was undertaken to determine if either form of the B5R

protein contributes to virulence by affecting the host inflammatory response.

xvii








To examine the overall contribution of the B5R protein to virulence, a

B5R null mutant (RPVAB5R) was constructed. In cell culture, RPVAB5R

showed a cell-line specific defect in production of the enveloped forms of

virus, correlating with an inability to form plaques on several normally

permissive cells. In animal studies, RPVAB5R was shown to be attenuated

through a mechanism not involving enhancement of the host inflammatory

response.

The roles of the individual forms of B5R were tested by construction of

mutants that solely produced either the membrane-bound or secreted form of

the protein but not both. The membrane bound form was shown to be

essential for virus growth in cell culture and virulence in animals. Mutants

failing to produce this form had a phenotype identical to the null mutant.

Mutants unable to produce the secreted form of B5R had no measurable effect

on virus growth in cell culture or on the pock color (morphology) produced

on the chorioallantoic membrane (CAM) of chicken eggs indicating that this

form of the protein is dispensable in these systems.

This work shows that although B5R has homology to regulators of

inflammation the role of the protein in virulence is likely not through

modulation of the host inflammatory response. The results strongly suggest

that the attenuation observed in B5R null mutants results from a defect in

production and release of mature enveloped virus particles, as demonstrated

in cell culture. In addition, these results demonstrate it is the membrane-


xviii









bound B5R, not the secreted form that is responsible for the activity of this

protein in these systems.












































xix















CHAPTER 1
INTRODUCTION AND BACKGROUND



Introduction


The Tao doesn't take sides; it gives birth to both good and evil.
The Master doesn't take sides; he welcomes both saints and sinners.
-Lao-tzu, translated from the Tao Te Ching by Stephen Mitchell

DNA neither knows nor cares. DNA just is. And we dance to its tune.
-Richard Dawkins


The vast majority of people living in the developed world today can

expect long lives virtually free of serious threat from infectious diseases. It is

a tribute to mankind's ingenuity and adaptability that the expectations of such

a life are often taken for granted and, in fact, have come to be considered the

normal condition. In reality, the current balance in our ecological

relationship with microscopic pathogens is a relatively recent aberration in

recorded human history, throughout which disease resulting from infection

has been a consistent fact of life (122). It has only been over the last two

centuries that advances in vaccination, simple sanitation and, most recently,

the development of antibiotics, have significantly increased both the length

and quality of life (191). While these developments have reduced the threat



1









2

of infectious disease, the rapid evolution and adaptability of micro-organisms

makes it inevitable that we will forever continue to see the emergence of

pathogens, both familiar and unknown. If the current, favorable balance is to

be maintained and improved, there must be a continuing evolution in our

understanding of the abilities of pathogenic organisms to cause disease.

Although medical research has resulted in tremendous progress in the

prevention and treatment of infections of both bacteria and viruses, the

molecular mechanisms underlying the ability of these organisms to cause

serious disease are only beginning to be understood. While the use of

antibiotics has allowed the successful treatment of most bacterial infections,

the recent emergence of resistant strains threatens to make these compounds

useless. Even more troublesome is the fact that progress in the development

of specific, effective treatments for viral infections has been both difficult and

slow with the result that many viral infections remain a significant health

threat. Understanding both the host response to a viral infection as well as

the mechanisms by which the virus is able to persist and spread in the

presence of such a response would provide great insight in the development

of new treatments for virally induced diseases.

While many viruses are known to cause disease in humans, the

poxviruses, although no longer a significant health threat, offer an excellent

model system in which to study virus host interactions. Within this family

are found viruses that naturally infect a wide range of animal species offering









3

researchers several animal models with which to study various aspects of

viral pathogenesis. Several of these viruses are known to be extremely

virulent within their natural host, dramatic examples being variola major

virus, the causative agent of smallpox in humans and myxoma virus which

causes myxomatosis in the European rabbit (128). Infection with variola

major, which is believed to have been eradicated from the wild, results in a

severe, disseminated infection and produces a mortality rate of twenty to

thirty percent (67). Another member of this family, which causes a severe

infection in rabbits, is rabbitpox virus (RPV), which was first isolated during a

spontaneous outbreak of infection in an isolated rabbit colony at the

Rockefeller Institute in 1932 (85,86). Animals infected with RPV develop

lesions ('pocks') on all parts of their skin, have high fevers and often display

signs of hemorrhage from the nares. With regard to these symptoms, the

disease resulting from infection with RPV has many feature characteristic of

smallpox in man, although a significant difference is seen in the mortality

rates, which in the case of an RPV infection is about fifty to ninety percent

(16). This ability of RPV to cause a lethal infection combined with the ready

availability of an animal model makes this virus an attractive candidate for

studying the effect of specific poxvirus genes on virulence.

The pathogenicities of variola and RPV are an indication that these

organisms are well adapted for survival in their hosts. Further, although the

infected host mounts an aggressive response to the infection, these viruses









4

are able to survive and thrive within the host animal suggesting that they

have evolved mechanisms to contend with or evade the host immune

system. This is indeed the case. Recent work has resulted in the discovery of

several poxvirus proteins that, in vitro, are able to neutralize specific aspects

of the host immune system (36). Many of these proteins either interact

directly with or have homology to critical components of the host immune

system. It is believed and in some cases has been demonstrated that the

interactions of these proteins with selective components of the host immune

system are essential to the ability of the virus to evade the host defenses.

Examples of such proteins include a secreted interleuken-f12 receptor (7,187), a

TNF receptor (50,97,178,182,204), a receptor for IFN-gamma (130,131,194,205), a

complement C4b-binding protein (99,110,111) and a serine protease inhibitor,

SPI-2, also known as crmA in cowpox virus (CPV) (142,155,158). The crmA

protein, one of the most thoroughly studied proteins of this group, has been

shown to inhibit the interleukin-19 converting enzyme (ICE), in vitro, and to

be involved in the inhibition of inflammation in vivo. The study of such

proteins will lead to a better understanding of the immune response to viral

infections resulting in improved treatment for viral diseases.

The work contained in this dissertation focuses on the B5R protein of

RPV. This recently discovered protein has homology to both the

complement proteins factor H and the C4b binding protein (57,100,119,195).

These proteins prevent activation of the complement cascade and, as a









5

consequence, inhibit the elicitation of the host inflammatory response. These

homologies suggest a mechanism by which B5R may act to enhance viral

virulence through inhibition of complement activation and, as a

consequence, inhibition of inflammation. Previous work on B5R in both

vaccinia virus and RPV has focused on the expression and localization of the

B5R protein in virus-infected tissue culture cells (57,100,119). These studies

showed that a 45 Kda form of B5R is found as a component of extracellular

enveloped virus (EEV), one form of infectious virus, but not within

intracellular mature virus (IMV) a progenitor of EEV (57,100,119). In

addition, a 38 KDa form of the protein, which is derived from the larger form

by processing, can be found in the supernatant from infected cells (119). The

work presented in chapter three focuses on the contribution of B5R to viral

virulence and, in particular, its effect on the inflammatory response to the

infection. The work presented in chapter four examines the importance of

the individual forms of the B5R protein in virus replication and virulence.

For my studies, the overall role of the B5R protein was first examined

by construction of a B5R null mutant, RPVAB5R, which was unable to

produce either form of this protein. In agreement with what has been

observed for similar mutants in vaccinia, this mutation resulted in a specific

interruption of EEV morphogenesis. Unlike what was observed with

vaccinia, however, this defect was observed to be cell-line dependent

(119)(this dissertation). RPVAB5R was also observed to be severely attenuated









6

in animals; however, this attenuation was not found to involve the host

inflammatory response. Although the growth of RPVAB5R was inhibited in

the animal, there was no increase in the inflammatory response to RPVAB5R

when compared to the response generated against RPV. In addition,

attenuation of RPVAB5R was not lessened under conditions in which the

ability of the animal to mount an inflammatory response was inhibited. The

conclusion drawn from this work is that B5R has no role in altering the host

inflammatory response and that the attenuation of RPVAB5R results from

the defect in morphogenesis observed in cell culture.

The importance of the various forms of the B5R protein was examined

by construction of mutants that produced only membrane-bound or only

secreted B5R. In all systems tested, failure to produce the membrane-bound

form of B5R resulted in a phenotype identical to that observed with

RPVAB5R. In contrast, failure to produce the secreted form of B5R had no

effect in cell culture or on pock color on the CAM and these mutants behaved

in a manner identical to RPV. These results indicate that membrane bound

B5R protein is essential for morphogenesis and virulence and that the

secreted form of B5R has no obvious function in cell culture or on the CAM.



Poxvirus Classification


The Poxviridae are an extensive family of large DNA viruses. The









7

members of this family are categorized, based on their host range, into two

sub-families, the Chordopoxviridae, which are viruses that infect vertebrates,

and the Entemopoxviridae, which are viruses that infect insects (128). The

sub-families are further divided into several genera, and individual species of

virus have historically been placed within a genus based on their shared

antigenicity as well as similarities in morphology and host range (67). More

recently, restriction fragment length polymorphism (RFLP) analysis, in

combination with complete or partial sequence from individual viral

genomes (79,102), has allowed a more precise means for measuring the

relatedness of individual viruses. These genetic analyses have supported the

earlier taxonomical assignments.

Rabbitpox virus, the virus used throughout this work, is a member of

the Orthopoxvirus genus, which contains at least seventeen members. Of all

the poxviruses, the most heavily studied members are found within this

genus. This is due in large part to the medical importance and usefulness of

two of its members, variola and vaccinia virus. These viruses represent the

infectious agent and the vaccine agent, respectively, of the disease known as

smallpox. RFLP analyses has revealed extensive homology between the

viruses in this genus with respect to the relative arrangement of restriction

endonuclease recognition sequences within the linear genome implying their

evolution from a common ancestor (60,117). A more detailed comparison

using the available DNA sequence from these viruses has revealed that RPV









8

is an extremely close relative of vaccinia virus, in particular the WR and

Copenhagen strains. A comparison of about 21.0 kb of sequence from the left-

hand of the viral genomes showed a degree-of-similarity of greater than

ninety-nine percent throughout the entire region (24). These data strongly

suggests that RPV arose from adaptation of VV(WR) or VV(Cop) or vice

versa.



Poxvirus Morphology


Because of its extensive use as a vaccine, almost all of what is known

about the structure and biology of the poxviruses has come from the study of

vaccinia virus, which is considered the prototypic poxvirus. With regard to

the structure of the virus, however, most if not all of the poxviruses have

been observed to share the same physical features. The poxviruses have the

distinction of being the largest of the animal viruses (128). The virus consists

of a complex arrangement of lipid, protein, carbohydrate and DNA. The

general appearance of the virus, when observed by electron microscopy (EM),

is that of an oval or barrel-shaped particle measuring 200 x 300 nanometers in

size (Figure 1A) (44,215). Within the center of thin-sectioned virions is seen

a dumbbell-shaped structure which is referred to as the nucleoid of the virus

(215). This membrane-bound structure contains the genomic DNA of the

virus in the form of a supercoiled, nucleoprotein complex (186). Electron-






















Figure 1. The structure of a poxvirus. (A) Cutaway view of the proposed virion structure as determined by EM using
standard fixation techniques. L, Lateral bodies. (Adapted from Fenner (65)). (B) Cutaway view of the proposed virion
structure as determined by cryoelectron microscopy. (Adapted from Dubochet (54))











10












o
-
-4 __































f '4'' "
O-
rs ^j '
w ^T _S* *S*S I*S S*~* ^9*1 9 S
M-l H IV^ '^'ZZ-Z''%1'-^f



Jn* ^********^^^^^^L"
cr ( J o\*.**.**.**.**.**'o
S *fcjj-.'j;U ~ ~ ~ jjjy
ca ^^^ kA^^WN%%^^SS""";3
c E^^^ ^w ^^////////^
ti^ rBAfww '^'^~~~~~~fa

nj ^ ^*HSr~iYV./."."/./.".'/./.'!^
l//oU LA/~l".VV..V.. I^ I
01.. Is s^ ^B?^/"./"^"-/"-/B
Y h^syOaww/a^ <
o u l^^sTtE^





CQO

////// ~ ~ ~ a>
....~fi
Fr" o.










u///


<;w









11

dense masses of unknown function are observed within the concavities of

the core. These bi-lateral structures have been named lateral bodies, and their

removal with non-ionic detergent yields an oval core, suggesting it is the

lateral bodies that give the core its characteristic dumbbell shape (56). The

core structure and lateral bodies of the mature virus are surrounded by a

second outer membrane that gives the virus its final characteristic shape.

When the particles are lightly shadowed with gold particles, this outer

membrane is observed to be covered with randomly oriented structures called

surface tubules which are believed to be composed of a single 58 KDa

polypeptide (189). The function of the surface tubules is unknown at this

time.

The viral structure presented thus far represents the traditional view of

poxvirus morphology which has been based on standard EM techniques.

However, recent work using cryoelectron microscopy (CEM) of vitrified

samples suggests a different structure (54). The newer work reports that the

dumbbell shaped core, the lateral bodies and the surface tubules are artifacts

that result from the use of standard fixation techniques. In particles viewed

by CEM, the virion is still seen as a barrel shaped particle; however, the core,

instead of being dumbbell shaped, has the same barrel shape as the rest of the

virion, and there are no lateral bodies (Figure 1B). In addition, the ST

structures are absent, and the core membrane is instead surrounded by a

regularly striated palisade.









12

The general particle described above has traditionally been called an

intracellular naked virus (INV) particle; however, in recent literature it is

referred to as an intracellular mature virus (IMV) particle. IMV particles are

fully infectious and, in fact, the majority of work in the poxvirus field to date

has been done using IMV purified from mechanically lysed cells. Although

the exact number is still unclear, IMV particles contain between 52 and 100

proteins, as determined by two-dimensional gel electrophoresis (61,139).

Twenty-one of these are known to reside in the core (139) including core

structural proteins and all of the proteins necessary for early gene

transcription. Biotinylation of purified IMV particles has shown the

envelope to contain at least ten prominent polypeptides with molecular

weights of 58 kDa, 39 Kda, 35 kDa, 32 kDa, 31 kDa, 24 kDa, 23 kDa, 22 kDa and

14 kda (83). Activities have been associated with several of these proteins,

some of which are described in the section on poxvirus biology.

In addition to the IMV particle just described, two other forms of virus

known as intracellular enveloped virus (IEV) and extracellular enveloped

virus (EEV) (hereafter collectively referred to as the enveloped forms of the

virus) are also produced during the infection (25). The three forms of virus

differ in the total number of envelopes that surround the viral core and in

the case of EEV, by its extracellular location. As described above, IMV

particles have a total of two membranes. IEV particles, which are seen within

the infected cell, are IMV that have become wrapped with two additional









13

outer membranes resulting in a total of four membranes. The final form of

virus, called EEV, is identical to IMV except that it has one additional outer

membrane giving it a total of three membranes. In addition, unlike IMV and

IEV, EEV is located outside of the unlysed cell. The additional membranes

present in the enveloped virions, give these particles a buoyant density

different than that of IMV allowing the separation of IMV from the

enveloped forms by centrifugation of the virions through an equilibrium

CsCl gradient. IMV has a density equal to 1.27 g/ml whereas the enveloped

forms have a density of 1.23 g/ml (144).

Extensive biochemical analysis has not been performed on the

outermost membrane of IEV but it is presumably similar or identical to that

of EEV, about which much is known. The earliest analysis of the protein

composition of the EEV outer membrane showed that the envelope

contained ten proteins unique to EEV, nine glycoproteins and one non-

glycosylated protein (144,145). Five of these proteins had molecular weights

of 210 kDa, 110, kDa 89 kDa, 42 kDa, 37 kDa, and five had molecular weights in

the range of 20-23 kDa The 37 kda protein was the only non-glycosylated

protein. A more recent analysis using monoclonal antibody precipitation

showed that the smallest proteins are 23-28 kDa in size and, further, showed

that these glycoproteins are actually differentially glycosylated forms of a

single protein specie, 21 kDa in size (147). In addition, this work showed that

the 220 kda protein is actually a complex of several differentially glycosylated









14

forms of the 89 kDa protein. Most recently, a previously undescribed 43-50

kda non-glycosylated protein has also been reported as being located in the

EEV envelope (143). The current model drawn from these data is that the

EEV envelope contains four glycoproteins having molecular weights of 110

kDa, 89 kDa, 42 kda, and 21 kda and two non-glycosylated proteins of

molecular weights 37 kDa and 43-50 kda.



The Poxvirus Genome and its Organization


While the specific size of the genome varies between the individual

species of poxvirus, the general characteristics and structure are shared by all

members of the family. All poxvirus genomes consist of a linear, double

stranded DNA molecule ranging in size from 85 million daltons

(approximately 129 kbp) for parapoxviruses to 185 million daltons

(approximately 280 kbp) for fowlpox virus (128). Recently, the sequences of

the entire genomes of the Copenhagen strain of vaccinia virus and the

Bangladesh-1975 strain of variola major virus have been determined, and

these viruses have genome sizes of 191,636 bp and 186,102 bp, respectively

(79,120). Computer-assisted sequence analysis predicts that vaccinia encodes

263 potential proteins of 65 amino acids or greater (79). Similar analysis for

variola reveals 187 potential proteins with eighty percent of these having

significant homology to proteins encoded by vaccinia virus (120). While the









15

size of the RPV genome has not yet been determined by sequence analysis,

RFLP analysis predicts a size of 118 million daltons (approximately 180 kbp)

(217). Comparison of the RPV HindIII restriction map with that of vaccinia

(Copenhagen strain) shows that the relative arrangement of the fragments

throughout the entire genome is identical between the two viruses

(24,117,217). However, while the order and size of most of the fragments are

identical in both genomes, the sizes of the terminal HindIII B and C

fragments differ between RPV and vaccinia, indicating a difference between

the two viruses in the coding capacity of the terminal regions. This difference

in coding capacity is likely to be responsible for differences in the host range

and virulence of vaccinia and RPV since it is believed that many of the genes

within these terminal regions influence these properties.

The structure of the vaccinia virus genome is illustrated in Figure 2.

While the majority of the linear genomic sequence is unique, the sequences

found in the 12.0 kbp left and right terminal regions are nearly identical and

are inverted with respect to each other (65,74,216). Within these inverted

repetitions and adjacent to the very ends of the genome are located tandem

repeats of 10 x 69 bp, 27 x 70 bp and 9 x 54 bp (65,79). Analysis of renaturation

kinetics and sedimentation analysis revealed that the two genomic DNA

strands are covalently cross-linked at their ends, preventing the two strands

from separating (18). Detailed sequence analysis of the terminal DNA

suggests that the terminal 104 bp exist as a single stranded loop, the sequence






















Figure 2. Organization and structure of the poxvirus genome. (A) HindIII restriction map of the poxvirus genome
showing the relative position and size of the restriction fragments. The area in gray represents the central conserved
core while the area in white represents the left and right non-essential regions. The asterisk indicates the location of
the B5R ORF. The regions containing the inverted terminal repetitions are indicated in the center of the page, while
the structure of the extreme terminal regions is shown in part B. (B) Proposed structure of the genome termini.
The vertical lines depict the clustered tandem repeats. A, B, C, A' and B' depict sequences within the 104 base-pair
imperfect palindromic sequences which form the terminal loop. (Adapted from Fenner (66)).










A.
A MNK F E OI G L J H D A B



CONSERVED CORE




12.0 kb '12.0 kb
I ................... INVERTED TERM INAL REPEATS .....................i






B. C

B A104bp
D' 10X69 bp 27X70 bp 9X54 bp
A









18

of which may be folded, using computer modeling, into an imperfect

palindrome (14,65,79).

The poxvirus genome may be divided into three regions based on the

absolute requirement for the genes found within each of these regions. The

central region contains genes essential for replication of the virus in tissue

culture, the so-called 'house-keeping' genes such as the viral RNA and DNA

polymerases. The essential nature of these genes is highlighted by the

sequence conservation exhibited within this region by the various viruses, a

feature that can easily be demonstrated by the near perfect alignment of the

core region HindllI restriction maps of the orthopoxviruses (117). The

essential core region extends from the rightmost end of the HindIII F

fragment until the rightmost end of the HindlIl A fragment and contains

approximately one-hundred open reading frames (79). The two regions

flanking either side of the core are referred to as the non-essential regions

and, as suggested by the name, contain genes not essential for replication in

tissue culture. While not absolutely essential in tissue culture, many of the

genes located in the non-essential regions are involved in determining host

range and in determining the virulence of the virus in animals. Most of the

major sequence variation observed between the orthopoxviruses is found

within the two non-essential regions which extend outward from the HindII

F and A fragments until the ends of the genome, which are found in the

HindIII B and C fragments (79).









19



Poxvirus Biology




The Replication Cycle


The replication cycle of the poxviruses is composed of a carefully

coordinated series of steps culminating in the release of newly synthesized

virions. The steps that make up this cycle, listed in order of occurrence, are

attachment, entry, early protein synthesis, DNA replication, intermediate and

late protein synthesis, viral morphogenesis and release of the virus from the

infected cell. One complete cycle of replication is illustrated in Figures 3 and

4. Each process in the cycle has been studied in great detail, and a brief

discussion of each will be given below.

Attachment, The infectious process begins with adsorption of a virion

to the surface of a host cell. Presumably, attachment of the virus to the cell is

mediated through binding of one or more viral attachment proteins to

molecules on the cell surface which serve as viral receptors. Since

interactions of the virus at the cell surface probably help to determine host

range, the broad host range of the orthopoxviruses makes it likely that they

recognize more than one cellular receptor. As of this date, no cellular

receptor molecules have been identified. The finding that vaccinia produces

a growth factor that binds to the epidermal growth factor receptor (EGFR) led





I






















Figure 3. The Life Cycle of a Poxvirus. The figure depicts one complete round of the poxvirus life cycle from entry
(upper left) to release of newly synthesized virus (lower left). IMV=intracellular mature virus. IEV=Intracellular
enveloped virus. EEV= extracellular enveloped virus. VETF=viral early transcription factor. VITF=viral
intermediate transcription factor. VLTF= viral late transcription factor.







Core
Lateral body o

IMV membrane Uncoang2


S^Intermediate
Early Early RNA
mRNA Proteins
(VITF)

Intermediate
S. .. Proteins
Gogi UC(VLTF)



late

0 0o mRNA
O


late
proteins
(VETF)

IMV




EEV EEV membrane









22

to the suggestion that this protein might serve as a receptor for the virus

(162). Although initial work seemed to support this hypothesis, it has since

been shown that adsorption of virus to the cell and subsequent replication are

unaffected by either deletion of the viral growth factor or by the use of cells

lacking the EGFR, demonstrating that these molecules do not serve as major

receptors (35,59,98). The most recent work in this area has shown that a

monoclonal antibody produced against a cellular surface protein blocks

vaccinia virus infection; however, neither the target nor the mechanism of

the anti-viral properties of this antibody have been identified (38). Future

work on this antibody may result in the first purification of a vaccinia

receptor.

The external location of the various virally-encoded membrane

proteins makes them logical candidates for virus attachment proteins.

However, although the currently known envelope proteins of both IMV and

EEV have been extensively studied, none of these proteins has been shown to

directly mediate virus attachment. Of the IMV envelope proteins, the 35 kDa,

32 kDa and 14 kDa have been the most actively investigated for their ability to

mediate virus attachment. The most recent data available for the 35 kDa

protein indicate that monoclonal antibodies directed against this protein are

able to neutralize virus infectivity by approximately seventy percent;

however, further proof of this proteins ability to mediate virus attachment is

needed (83). The 32 kDa protein has been reported to bind to cell surfaces in a









23

manner that correlated with the ability of vaccinia to bind the same cells;

however, mutants lacking this protein were unaffected in their ability to

attach to and replicate in tissue culture cells (114,136,162). Interestingly,

deletion of this gene results in decreased viral gene expression in a polarized

cell culture system and also results in severe attenuation of the virus in mice,

indicating that this protein is important for virulence (162). The 14 kDa

protein is known to be involved in the fusion of the virus envelope with the

cell plasma membrane and it is also involved in the fusion of the plasma

membrane of infected cells (51,163,164). Even though this protein is

intimately involved with early events in virus entry, antibodies to the 14 kDa

protein only block virus fusion, not binding, indicating no role for 14 kDa in

attachment (163). The above results suggest several possibilities: I) cell

attachment is mediated by several proteins so that deletion of a single protein

has little effect on attachment, II) standard cell culture systems may not

accurately reflect the binding activity observed in animals, a possibility

suggested by the results with the 32 kda protein, or III) the major attachment

protein has yet to be found.

Similar work has been conducted on the proteins of the EEV envelope

again yielding no conclusive evidence of a major viral attachment molecule.

A significant problem with attempting to measure the binding activity of EEV

particles lacking these molecules is that deletion of the genes encoding these

proteins often results in a failure to produce EEV particles. This EEV-minus









24

phenotype has been demonstrated for the 43-50 kDa, 42 kDa, 37 kDa and 21

kDa proteins of the envelope, making it difficult to assess the effect of these

molecules on virus attachment. Monoclonal antibodies directed against the

42 kDa protein have been show to inhibit spread of EEV in cell culture;

however, no clear binding activity has been shown for this protein (57). In

fact, in experiments using RPV 42 kDa null mutants, which are capable of

producing EEV only in certain cell lines, EEV particles lacking the 42 Kda

protein bound cells with kinetics identical to those observed for EEV from

wild-type RPV (Stern, R.J., Unpublished data). The 21 kDa envelope protein

has been shown to be involved in the release of newly synthesized EEV

particles from the cell surface; however, the importance of this molecule in

initial virus-cell interactions is unclear (22). Of all the EEV proteins, only the

89 kDa protein, known as the hemagglutinin (HA) protein, has been shown

to have binding activity. The HA protein ha been shown to bind sialic acid

containing molecules present on the surface of chicken red blood cells.

However, deletion of this open reading frame (ORF) has no effect on the

ability of virus to attach to cells in culture or on viral virulence in animals.

In fact, the RPV strain commonly used for laboratory study is HA- due to a

naturally occurring mutation resulting in a frameshift.

Entry. The attachment phase of the infectious cycle is followed by entry

of the virus into the cell, the primary mechanism of which is still somewhat

unclear. The earliest studies of this process using EM and analysis of the









25

effect of metabolic inhibitors suggested that endocytosis of the virus particle

was the most common route of entry (40,41). However, later studies indicated

that entry occurs by direct fusion of the viral envelope with the cell surface

(10,37). The most recent work confirms that poxviruses do indeed fuse with

the host cell in a pH and temperature-independent manner (51,81,152).

While phagocytosed virus particles were observed in these studies, particles

contained in phagocytic vesicles represented only a small fraction of the total

input virus, suggesting a minor role for this route of entry. This work also

demonstrated that EEV particles fuse with cell membrane at twice the rate of

IMV particles (51). In addition, fusion of EEV was unaffected by treatment of

cells with compounds that effect endocytosis, conditions which were shown

to severely reduce IMV fusion, suggesting that the two forms of virus may

enter the cell by different methods (152). The significance and relevance of

these observations to the disease process in the animal remain to be

determined.

Several viral proteins have been shown to be involved in virus-cell

fusion and in cell-cell fusion. The best characterized of these is again the 14

kDa fusion protein found on the envelope of the IMV particle (163,164).

Monoclonal antibodies to this protein have been show to block fusion, but

not binding, of the virus with the cell (51,164). In addition, deletion analysis

has shown that the 14 kDa protein mediates fusion of infected cells at acid pH,

during the late stages of the infection (81,163,164). Two other proteins are also









26

known to affect the cell-cell fusion phenotype, these being the 89 kDa EEV

envelope HA protein and the SPI-3 protein. Although deletion of the HA

ORF has no effect on virus-cell fusion, infection of cells with HA- virus

results in one hundred percent fusion of the infected cell, suggesting that one

of the functions of the HA protein during the infection is to prevent cell-cell

fusion. The localization of the HA protein to the envelope of EEV suggests it

may have a similar function in virus-cell fusion. The importance of this is

unclear, however, since HA- viruses are as infectious as HA+ viruses.

Deletion of the SPI-3 ORF also results in fusion of infected cells,

demonstrating some role for SPI-3 in fusion activity. The intracellular

location of this protein suggests that it may somehow affects the 14 kDa

protein; however there is no evidence for such an activity and its mechanism

of action on fusion the fusion event is unknown.

Uncoating. Following entry of the virus into the cell, the virus

undergoes a process known as uncoating (92,103,153). This process, which has

been studied both biochemically and by EM, has been divided into two well-

defined stages. In the first stage, there is a rapid degradation of some of the

outer viral proteins and virtually all of the viral phospholipid resulting in

structures that are called sub-viral cores. These cores appear similar to cores

observed by EM, and the DNA at this stage remains resistant to degradation by

DNase (103,104). The first stage of the uncoating process is independent of

new protein synthesis and will occur in irradiated cells, suggesting it is a









27

property intrinsic to the intact virion and/or pre-existing factors within the

cell. Following a lag period of about one hour, the virus enters the second

stage of uncoating during which the core is degraded and the DNA becomes

susceptible to degradation by DNase. The lag period may be shortened by

increasing the multiplicity of infection or abolished by pre-infection with any

of the other orthopoxviruses. In addition, the second stage is dependent on

viral protein synthesis (153). Although it was initially thought that a cellular

protein induced uncoating, the recent development of an in vitro system to

study this process resulted in the purification of a virally produced 23 kDa

protein that is able to initiate uncoating of the viral cores (153).

Early protein synthesis. The synthesis of viral proteins can be detected

shortly after the virus has entered the cell, even before the second stage of

uncoating has taken place. Production of the uncoating protein has been

detected as soon as 20 minutes after infection of the cell (153). The production

of early transcripts is directed by a DNA-dependent RNA polymerase complex

present within the infecting virion (129). This complex, which contains two

large and six small virus-encoded subunits has a molecular mass of around

500 kDa and contains all the activities necessary for randomly initiated RNA

transcription from a single stranded DNA template (3,129). This polymerase

activity is made specific for poxvirus early genes if one adds several early

transcription factors to the polymerase complex. One such factor is the

product of the H4 ORF, a protein known as Rap94 ( 94 kDa RNA polymerase









28

associated protein) (3,105). The addition of a second factor, composed of two

protein subunits and known as the vaccinia early transcription factor (VETF),

to this complex allows the polymerase-Rap94 complex to utilize double-

stranded DNA containing early promoters. VETF, a heterodimer made up of

an 82 kDa protein and a 70 kDa protein, has been shown to possess ATPase

activity. In addition, purified VETF has been shown to be a DNA binding

protein with specificity for sequences within the promoters of virus early

genes (219). The RNA-polymerase-Rap94-VETF complex contains all the

factors necessary for transcription of early genes from the double-stranded

viral genome.

Termination of early gene transcription begins once the polymerase

complex encounters the sequence UUUUUNU in the newly created RNA

strand (166). Actual cessation of transcription appears to occur randomly

approximately 50-70 bases downstream of the termination signal and is

known to be dependent on viral polymerase, VETF and the virus encoded

capping enzyme (166,180). The capping enzyme, which consists of two virus

encoded proteins of 97 kDa and 33 kDa, is also known to be involved in the

capping of the 5' end of the viral mRNA (118,129,134,135,181). Following

termination, the mRNA 3' ends are polyadenylated by a complex (known as

the poly-A polymerase) composed of two viral proteins with molecular

weights of 55 kDa and 33 kda (75). The poly-A polymerase seems to have no

specific recognition sequence since any transfected RNA is also









29

polyadenylated (66). These capped and polyadenylated transcripts, which

appear structurally identical to cellular eukaryotic transcripts, are now

released from the viral core into the cytoplasm of the infected cell to direct the

translation of viral polypeptides necessary for intermediate protein synthesis

and DNA replication.

Intermediate protein synthesis. Shortly after early protein synthesis

and immediately after the onset of DNA replication, transcription from a

second class of genes, known as the intermediate genes, begins (129,209). This

class is defined by genes that are dependent on trans-acting factors produced

in the cell shortly after infection (early proteins) and on DNA replication but

not on de novo protein synthesis following DNA replication. This is in

contrast to late genes which are dependent on both DNA replication and on

continued intermediate protein synthesis following DNA replication. In

addition, unlike the late genes, plasmid-borne intermediate genes are

expressed prior to DNA replication in vaccinia virus infected cells that have

been treated with inhibitors of DNA replication, demonstrating that all of the

factors necessary for intermediate gene transcription are present prior to DNA

replication. Currently, five intermediate genes (AlL, A2L, GK1, 18 and 13)

have been identified, three of these (AlL, A2L, GK1) being late transcription

factors (108,209). No consensus promoter sequence for intermediate genes has

yet been identified; however, the factors required for intermediate

transcription are known (207,208). These are the viral RNA polymerase, the









30

capping enzyme, which in this capacity has a role additional to its role in

capping mRNA (207), and two proteins known as virus intermediate

transcription factor (VITF)-1 and VITF-2 (168). VITF-1 has been shown to be

the 35 kDa rpo30 protein which is coded for by the vaccinia E4L ORF (4,168).

Rpo30, which has homology to eukaryotic transcription elongation factor SII

(TFIIS), has also been shown to be a component of the early RNA polymerase

complex suggesting a dual role in transcription for this protein (4). VITF-2

activity, which co-purifies with a 68 kDa protein, has been reported to be

cellularly encoded since its activity can be purified from uninfected cells (169).

In this regard, it is interesting that enucleation experiments indicated a role

for the cell nucleus at a late stage in infection (95,96). It has also been shown

that in RK-13 cells, which lack VITF-2 activity, mutants in the K1L gene of

vaccinia are blocked in the intermediate transcription stage of the replication

cycle (169). In vitro transcription assays detected VITF-2 activity in RK-13 cells

stably expressing K1L indicating that the K1L protein can functionally

substitute for VITF-2. The function of K1L, which has been shown to contain

several ankyrin repeats, is currently unknown (169). Further studies should

reveal the nature of the link between K1L and VITF-2.

The intermediate transcripts have been characterized and are thought

to initiate within a sequence resembling a late promoter and to have

heterogeneous 3' ends (12). It was initially been reported that these mRNAs

lacked a poly (A) head (209); however, recent studies have reported poly (A)









31

heads of up to 30 bases on these mRNAs (12). The mechanism of termination

of intermediate transcription is unknown.

Late protein synthesis. Following DNA replication, which begins

approximately 2 hours post-infection, and the accumulation of late

transcription factors produced from intermediate genes, late gene

transcription begins. The late genes encode the structural proteins of the

virus particle along with the proteins necessary for the proper assembly of the

virus particle (128). In addition, the proteins that comprise the early-gene

transcription complex found within the virion are produced late in infection.

Transcription of late genes requires a complex of polymerase and several late

transcription factors produced from intermediate genes. It is likely the

polymerase used for early gene transcription is also used for late gene

transcription since conditionally lethal mutations in the polymerase subunits

affect late gene transcription (93). In addition to the polymerase complex,

several late transcription factors are known to be required. These include the

above mentioned 30 kDa, 26 kDa and 17 kDa proteins produced from

intermediate genes (108,209) and a recently discovered factor known as P3

(112).

Comparison of the sequence of several late genes has revealed a

consensus late promoter sequence consisting of two elements (66). The first

element, consisting of the short sequence TAAAT followed by an A or a G, is

found immediately upstream of the translation initiating ATG. In fact, in









32

most cases, the initiating ATG forms the 3' end of the consensus sequence

(TAAATG). This sequence is conserved in nearly all late genes, suggesting

some, as of yet unknown, important function. Any mutations within this

sequence abolish late transcription of the corresponding gene. The second

element is found 16 to 20 bp upstream of the initiating ATG codon. This

element shows no strong conservation of sequence but usually consists of a

stretch of five to eight A or T residues that are essential for late gene

transcription.

The mRNA transcribed from late genes has several unusual properties.

The mRNA contains 35-50 non-templated polyadenylate residues at its 5' end

that are thought to result from slippage of the RNA polymerase within the

TAAATG sequence (66). This model is based on a similar model in

bacteriophage T4 for which there is considerable experimental evidence (66).

A clear function for the poly-A head is unknown however it has been

speculated that it may act as a ribosome binding site and increase the

efficiency of vaccinia mRNA translation (66).

In addition to the 5' poly-A head, late mRNAs also have heterogenous

3' ends indicating that late transcription does not terminate uniformly as does

early mRNA. To date, no termination signals have been recognized for late

mRNA transcription and, in fact, late transcription can continue

uninterrupted through several open reading frames generating transcripts

that are several kilobases in length. Since both strands of DNA are









33

transcribed, this results in the generation of a significant amount of double-

stranded RNA at late times during infection.

Replication of the poxvirus genome. Based on measurements of the

amount of accumulated DNA that hybridizes to known quantities of filter-

bound DNA, it has been determined that replication of genomic DNA begins

around 2 hours post-infection, concurrent with intermediate gene expression,

and proceeds for at least ten hours (200). This replication occurs in the

cytoplasm of the cell in electron-dense areas referred to as viral factories or

virosomes. This cytoplasmic location suggests that, as is the case with the

transcriptional apparatus, the virus must provide most if not all of the

enzymes necessary for genomic replication. Experiments using enucleated

cells indicate that, indeed, the virus provides all the necessary replication

machinery (154).

Although many of the details surrounding poxvirus DNA replication

remain unclear, the available data has resulted in a well accepted working

model which is illustrated in Figure 4 (132,200). This model is based on the

known structure of the genomic termini as well as on analysis of genome

intermediates observed during DNA replication (132). As previously

described, the termini of the poxvirus genome consist of a single-stranded

DNA containing an imperfect palindromic sequence. When the two halves

of this sequence are maximally aligned, several mismatched bases are seen. It

is within this hairpin region that replication is thought to initiate.




























Figure 4. Model for replication of the poxvirus genome. Self-complimentary
sequences within the inverted terminal repeats are depicted in upper and
lower case letters. (A, a', B, b', C, c'). The initial nicking site is denoted by the
"X". Newly synthesized DNA is depicted by dotted lines. (Adapted from
Traktman (200)).









35
ABC CBA

a' b' c'V c' b' a'

a' b'c' ABC CBA

A B Ca' b' cc' b' a'







b'5 a












ABC | c' b' a'

a' b' c' CBA






ABC c' b' a' ABC c' b' a'
S)+( )
a' b'c' CBA a' b'c' CBA









36

Replication begins with a single-stranded nick being made within the

hairpin structure. Nicking likely occurs at either or both ends of the genome,

but for ease of discussion, it will be assumed here that only one nick per

genome is made. The presence of this nick has been identified by the nicked

DNA's altered mobility and sedimentation properties (157). The necessity of

sequence specificity is argued against by the fact that any DNA transfected into

poxvirus infected cells is replicated (46). This initiating event results in a free

3'-hydroxy terminus which can now be used by the polymerase as a primer.

Elongation of this strand displaces the complementary strand and uses it as a

template. This model is supported by data showing that radiolabeled

nucleotides are first incorporated into sequences from the end of the virus

genome (156). The resulting molecule is a longer-than-genome-length

hairpin with palindromes on either end. The newly synthesized end, being a

palindrome, can fold back on itself, and continued extension of this strand

results in copying of the entire genome. The majority of evidence to date

indicates that elongation occurs by unidirectional, leading strand synthesis

(200). The product of this event is a molecule in which two complete

genomes are joined by their right-hand ends in a tail-tail arrangement.

Continued elongation yields a concatemeric molecule with multiple genomes

organized in tail-to-tail and head-to-head arrangements. Concatemers

consisting of up to four genomes have been shown to exist (45,132). These

concatemeric molecules are then resolved into unit length genomes through









37

a process known as telomere resolution. This process requires a pair of

sequences known as the telomere resolution targets, which are present in the

junction fragments, as well as an activity referred to as a resolvase (192). The

importance of the telomere resolution targets has been demonstrated by

experiments showing that plasmids containing this sequence within cloned

junction fragments are resolved into linear molecules with plasmid DNA in

the center and viral hairpin at the ends (47,48).

Several of the proteins involved in replication of the poxvirus genome

have been purified including a DNA dependent DNA polymerase, a

topoisomerase, a DNA ligase, a thymidine kinase and several DNA binding

proteins of unknown function (200). The polymerase is found as a monomer

of 110 kda and possesses both polymerase and 3'-5' exonuclease activity

(66,200). Although the polymerase shares homology with several eukaryotic

polymerases, its activity cannot be replaced by the cellular polymerase, and it

has been shown to be essential for virus replication.

Virus morphogenesis. Although vaccinia virus was one of the first

objects to be viewed by EM, the molecular mechanisms responsible for the

assembly and release of the intact virion remain obscure. Most of what is

currently known about viral morphogenesis is largely descriptive

information based on the effects of mutations and inhibitors of virus

replication on virus morphogenesis, as determined by EM. Nevertheless, a

model for the pathway of virion assembly has been constructed.









38

Assembly of the mature virus begins in granular, electron-dense

structures called viral factories that can be observed in the cytoplasm starting

about 4 hours post-infection (185). Staining of thin sections of infected cells

with DNA-specific indicators reveals large quantities of DNA present in what

has been called the "viroplasm" within these factories. The first indication of

assembling virus is the appearance within the viral factory of crescent-shaped

membranous structures that will eventually form the envelope of the IMV

particle (42). It has long been thought that these membranes are not obtained

from a cellular organelle but instead are synthesized de novo by the virus

since they have never been observed to be continuous with cellular

organelles. This hypothesis seemed to be supported by chemical analysis of

these membranes which indicated that their phospholipid composition was

vastly different than that of cellular membranes (188). However, recent

studies using CEM showed that the envelope membrane is, in fact,

continuous with cellular, membranous structures (185). Through the use of

labeled antibodies specific for proteins which localize to several subcellular

compartments, it has been determined that the envelope membrane is

derived from the intermediate compartment between the ER and the golgi

stacks. In addition, these studies showed that the crescent shaped structures

and IMV envelope are actually composed of two closely apposed membranes.

These data support the hypothesis that the envelope results from wrapping of

the core with cisternae derived from a golgi-ER intermediate compartment.









39

The crescents observed in the viral factories are covered with a layer of

spicules formed from a single polypeptide of 65 kDa of unknown function

(62). The crescents assemble around viroplasm into spherical, immature

particles containing a dense granular center. Interestingly, although virions

are formed late in infection, empty immature particles are formed in the

presence of inhibitors of DNA replication, indicating that viral proteins

required for particle formation are produced early. Following production of

the immature particle, the viroplasm contained within condenses by

unknown mechanisms into a DNA-containing core structure, resulting in an

infectious intracellular mature virus particle. During this transformation,

the spicules are replaced by the tubule structures which are composed of a 58

kDa protein (189).

As previously described, while IMV particles are fully infectious, a

small percentage become wrapped with additional envelopes and are

transported out of the cell resulting in extracellular enveloped virus. The

origin of the EEV envelope has been a subject of some controversy. Although

both golgi and early-endosomal network membranes have been postulated to

be the source, the most recent study using antibodies to proteins specific for

intracellular compartments indicates that the envelope is derived from the

trans-golgi network (TGN) (90,176,199). The cytoplasmic IMV particle is

wrapped by cisternae derived from the TGN resulting in the addition of two

membranes to the particle, yielding an IEV particle. This wrapping event has









40

been shown to be dependent upon the presence of the 42 kDa, 37 kDa, 32 kDa

and the 21 kDa envelope proteins as well as on the 14 kDa fusion protein

located in the IMV envelope (20,55,58,119,143,165,218). Deletion of any of

these genes has no effect on IMV production but results in the inability of

such mutants to produce the enveloped forms of virus (IEV and EEV). The

requirement for these enveloped proteins in the wrapping process is

currently not understood but it is thought that some of them may mediate

wrapping by interacting with IMV surface proteins. Interestingly, the

compound NI-isonicotinoyl-N2-3-methyl-4-chlorobenzoylhydrazine

(IMCBH) specifically inhibits the wrapping of the IMV particle (89,148).

While the mechanism of this inhibition is unknown, mutations in the 37

kDa envelope protein have been shown to confer on the virus resistance to

the inhibitory effects of IMCBH (177). In addition, inhibition of enveloped

virus formation may also be achieved by treatment of infected cells with

inhibitors of glycosylation or with brefeldin A, a fungal metabolite that has

been shown to inhibit the transport of proteins to post-ER compartments

(150,203).

Once the IEV particle has been formed, it moves to the cytoplasmic

membrane where the outermost virus membrane fuses with the cytoplasmic

membrane releasing the virion into the cytoplasm. The cellular apparatus

involved in transport of the particle remain uncharacterized; however

treatment of infected cells with cytochalasin D blocks release but not









41

formation of enveloped virus suggesting, a role for microfilaments in the

release of the enveloped virus particle (149). The amount of EEV produced

varies from less than one percent to around twenty-five percent of the total

virus yield, depending on the virus and cell line used (145). This variation

suggests some role for as of yet unidentified cell components in enveloped

virus release. The 21 kda envelope protein is the only EEV protein that has

been implicated in modulating the amount of virus released from the

infected cell. A single amino acid change in the lectin homology domain of

this protein has been shown to alter the amount of virus released by a factor

of ten however, the cellular target of this protein is unknown (22).


Some Musings on the Significance of Enveloped Virus


Since the IMV particle produced in the infected cell is completely self-

sufficient in its ability to initiate a complete round of infection, it may seem

wasteful for the virus to devote both energy and coding capacity to the

production of enveloped virus. However, since many of the poxviruses are

known to produce EEV, there is likely some survival advantage in this ability

beyond the interest it generates in scientists who then continue to propagate

the virus. One might think that enveloped virus production is an artifact of

cell culture, however, enveloped virions have been observed in thin sections

of infected animal tissue (113,151). Since the virus used in this study, RPV,

produces large quantities of EEV, it is worth spending some time reviewing









42

current thought on the significance of enveloped virus in cell culture and in

animal infections.

The enveloped forms of the virus, IEV and EEV, are believed to have

two distinct roles in dissemination of virus from the initial foci of infection

in cell culture. A simple 'comet' assay has historically been used to measure

the ability of virus to spread from the initial focus of infection (9). In this

assay, cells infected under conditions that yield well-isolated plaques are

incubated under liquid for several days. Virus released from a focus of

infection spreads, relatively unidirectionally, through the medium producing

comet-shaped areas of cell destruction, with the head of the comet formed by

the primary plaque. There exists a correlation between the amount of EEV

produced by a virus and its ability to form comets, suggesting EEV is

responsible for long-distance spread of virus (146). This has been confirmed

experimentally since neutralizing antibodies to EEV, but not IMV, prevent

comet formation without affecting primary plaque formation (9).

The observation that viruses can produce vastly differing amounts of

EEV yet still produce similarly-sized primary plaques led to the hypothesis

that another form of the virus, IMV or IEV, was responsible for cell-to-cell

spread and subsequent plaque formation. Mutants that produce IMV but

which are unable to form enveloped virus fail to plaque suggesting that IMV

is not sufficient for efficient cell-to-cell spread (20,165). This has led to the

current model in which IEV, which can be observed attached to the outer









43

surface of the infected cell (21), is thought to be responsible for cell-to-cell

spread and plaque formation, although there is no direct proof of this as of

yet.

The importance of EEV in the dissemination of virus throughout the

infected animal has been suspected for a long time and is derived from

several observations (9,27,146). Although the majority of virus produced in

tissue culture is IMV, it has been shown that antiserum generated against live

rabbitpox virus can neutralize both IMV and EEV in vitro and is able to

passively protect rabbits against lethal infection with rabbitpox virus (27). By

contrast, antiserum to inactivated virus although neutralizing for IMV in

vitro is non-neutralizing for EEV and is unable to passively protect rabbits

from lethal infection. In addition, the vast majority of virus purified from

the blood of infected animals is of the enveloped type. Finally, mutants

defective in their ability to produce enveloped virus (but not IMV) have been

shown to be attenuated in animals (58,143) (and this dissertation). Taken

together, these results provide strong evidence that it is the EEV particle that

is important for successful spread and the manifestation of full virulence in

the infected animal.


Poxvirus Pathogenesis


The severity of smallpox in man resulted in a search for useful animal

models with which to study poxviral pathogenesis. The study of two such









44

models, rabbitpox virus in rabbits and mouse-pox (ectromelia) virus in mice,

has led to a greater understanding of poxviral infections with regard to the

disease progression and pathology of these agents (8,19,85,86). These studies

produced nearly identical results suggesting common themes in the ability of

poxviruses to cause disease. Since I have restricted my work to the use of

rabbitpox virus (RPV), I will concentrate primarily on the pathogenesis of this

virus.

The history of RPV is an interesting story of epidemiology and the

emergence of new infectious agents. The virus was first reported in 1932

when its spontaneous appearance resulted in a smallpox-like epidemic

within a breeding colony of rabbits housed at the Rockefeller Institute for

Medical Research (84). Within two weeks one-hundred percent of the colony

had been infected and forty-six percent had died. Subsequent studies revealed

the etiological agent to be a poxvirus that was given the name rabbitpox virus

(84). While the exact origin of this virulent virus will likely never be known,

records from the time indicate that vaccine virus and a neurovirulent strain

of vaccinia were being used to infect animals in common areas of the

breeding facility (84). Epidemiological studies from the time strongly suggest

that vaccinia virus was the progenitor of RPV, a hypothesis supported by

sequence comparisons showing ninety-nine percent nucleotide homology

between the two viruses (24).

The outbreak of such a severe disease within a well regulated colony









45

resulted in excellent documentation of the symptoms and pathology of RPV

infection (85-87). Clinically, the disease resembles smallpox. Onset was

marked by malaise and fever in conjunction with enlargement of the

popliteal and inguinal lymph nodes. This was followed in two to three days

by the appearance of papules on all parts of the animal which eventually

became umbilicated and necrotic. These papules were often associated with

extensive edema. Upper respiratory infection commonly occurred and was

accompanied by bronchopneumonia resulting in labored respiration and

mouth breathing. Many of the animals also developed conjunctivitis and

released mucopurulent discharges, often stained with blood, from the nasal

passages. Death from infection resulted within eight to ten days in nearly

fifty-percent of the cases.

A detailed analysis of the progress and course of a poxvirus infection

has also been described for ectromelia in mice, resulting in a now classic

model that has since been shown to be valid for RPV as well (16,26,214).

Spread of the virus through the body is a stepwise invasion involving (1) the

primary site of infection (2) the regional lymph nodes (3) the spleen and (4)

the lungs, skin and other remote organs. The route of infection affects mostly

the LD50 and has virtually no effect on the progress of infection (16).

Experimentally, initiation of an ectromelia infection is usually accomplished

by injection (intradermal or footpad) or inhalation of aerosolized virus, both

of which have been demonstrated as feasible natural modes of transmission









46

(16,214). Most of the available pathological examinations have been

performed on animals infected by the respiratory route (16,85,86). Following

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 only found at the inoculation site, however, within 36

hours, it can be recovered from the regional lymph nodes, having likely been

delivered there by infected lymphocytes (16). Replication within the lymph

tissue is followed by a minor viremia immediately after which, on day two

post-infection, virus is detected in the spleen. In animals inoculated

intradermally, however, virus can be detected simultaneously in the lymph

nodes and the spleen, casting suspicion on the role of the lymph node as an

intermediate in the spread of the virus (16). The virus rapidly achieves high

titers within the spleen resulting in a second and massive viremia around

day three post infection. It should be noted that virus purified from blood is

associated with the lymphocytic fraction suggesting that during viremia, the

virus is not free but is instead cellularly associated (16,214). Concurrent with

this secondary viremia and signaling the end of the incubation period is the

onset of fever which likely arises due to the inability of the spleen to clear the

growing load of virus and the resulting systemic infection (16). A nasal,

mucopurulent discharge also appears around day three and increases in

severity until around day five post infection when the animals become

infectious (16,86,214). The increased levels of blood-borne virus result in









47

delivery of this virus to the remaining organs of the body, and papules begin

to appear on the skin approximately five to six days post infection. Death

from the infection usually occurs between days six and ten post infection and

is frequently preceded by a sharp drop in body temperature. If the animal

survives, the papules scab over and clear around days ten to twelve post

infection, usually leaving behind scarring.

Microscopic analysis of the papules indicates that they all originate as

vascular or perivascular processes, which is not surprising in view of the

viremic spread of the virus through the body(85). Early in the development

of the lesion, the epithelial layer becomes swollen, and there is proliferation

of the cells of the arterial wall. This hyperplasia is probably caused by the

viral growth factor, as has been demonstrated on the chorioallantoic

membrane of embryonated chicken eggs (33). Cells, polymorphonuclear

leukocytes, monocytes, and some red blood cells, become grouped in a loose

collar formation around the affected blood vessel. As the lesion progresses, it

assumes a focal form; however, in a small percentage of cases it becomes

diffuse, spreading laterally through the tissues. The diffuse form of the lesion

is more hemorrhagic than the focal form and results in the sloughing of large

areas of tissue. Within the center of the developing nodule, the cells become

necrotic and disintegrate so that cell outlines can no longer be distinguished.

The histological picture of the nodule is now a central area of necrosis

surrounded by a small area of polymorphs which itself is surrounded by a









48

zone of mononuclear cells. The mononuclear cell zone and the surrounding

tissues are edematous and occasionally hemorrhagic. Extension of the lesion

into deeper layers of the skin causes impairment of the blood supply resulting

in anemic necrosis that extends to the surface of the animal.

While the obvious pneumonia and labored breathing observed in the

late stages of infection would lead one to think death results from

pneumonia, the results from studies conducted on the cause of death suggest

otherwise (26). Physiological studies of infected animals consistently showed

a severe drop in blood pressure and the occurrence of a shock-like state

immediately prior to death. This was coupled with a dramatic rise in serum

potassium concentrations to levels that have been shown experimentally to

be lethal. From these data it has been concluded that death results from

potassium intoxication.


Nonspecific Host Responses to Infection


The host response to the infection described above consists first of a

non-specific immune response followed by the development of a specific

immune response. Since the majority of the work in this dissertation is

concerned with the very early response to infection, this section will be

limited to a discussion of the non-specific response (16,85,86,214). Although

no specific immune response can be detected until around day eight post-

infection, it is known that the non-specific components of the immune









49

system are activated immediately following infection. This non-specific

response is composed of pyrexia, interferon production, apoptosis of infected

cells, complement activation, as well as recruitment of leukocytes and NK

cells to the site of infection. A brief discussion of each of these is given below.

Interferon. One of the earliest responses to viral infection is the

production of interferon at the site of infection. The interferons (IFN's) are a

family of proteins that is divided into two groups, type I and type II

interferons (174). The type I interferons are induced in response to viral

infection and include IFN-alpha and IFN-beta which are produced by

leukocytes and fibroblasts, respectively. Type II interferon, called IFN-gamma

is produced by antigenic stimulation of T lymphocytes and NK cells.

Production of interferon in poxvirus-infected cells is believed to be due

mainly to the production of dsRNA late in infection (see Late Protein

Synthesis) and results in the establishment of an antiviral state in

neighboring cells (77,115).

The effects of interferon on poxvirus replication have been studied

both in cell-culture and in the infected animal (174). In cell culture, treatment

of cells with IFN prior to infection may cause abortion of the poxvirus

infection; however, the efficiency of the antiviral state is dependent on the

virus used, the cell type and the concentration and the type of IFN used (13).

The nature of the antiviral state has been well studied and has been shown to

result from a block at the level of mRNA translation in infected cells,









50

effectively blocking even the earliest steps in virus replication such as particle

uncoating (72,123). The mechanisms by which interferon blocks protein

production are proposed to result from two pathways dependent upon the

action of three interferon-induced enzymes: a 2',5'-oligoisoadenylate

synthetase, an RNase, and a 67-kDa protein kinase (174). In the first pathway,

the 2',5'-oligoisoadenylate synthetase catalyzes the synthesis, from ATP, of a

family of oligonucleotides with the general structure ppp(A2'p5')nA (nr2)

abbreviated 2,5A. This molecule then activates an endonuclease, known as

RNase L, which is present in a latent state within the cell. Once activated,

RNase L cleaves both cellular and viral single-stranded RNA. In the second

pathway, a 67-kDa protein known as P1/ eIF-2, the production of which is

induced by the IFN's (30,116), is activated by either double or single-stranded

RNA(116). Once activated, P1/elF-2 catalyzes the phosphorylation of the

alpha subunit of protein synthesis initiation factor eIF-2 (175). The end result

of this process is inhibition of protein synthesis at the initiation step of

translation.

Systemically, IFN-gamma is known to increase MHC-I production and

to activate natural-killer (NK) cells (see below) (1,167). The study of the effects

of IFN-gamma on poxvirus replication in the animal has primarily involved

correlating the levels of IFN's with the severity of the disease, or by

measuring the effect of injecting either IFN or inhibitors of IFN (36). In

poxvirus infected animals, increased levels of all IFN's can be detected in the









51

serum as well as in cells taken from the peritoneal cavity or from the site of

inoculation suggesting that IFNs are important in combating the infection

(71,78,193). In support of this, it was shown that treatment of infected mice

with anti-IFN antisera resulted in a significant increase in the amount of

virus recovered from these animals (101). In addition, administration of

alpha or beta IFN prior to infection has been shown to decrease the severity of

the disease and the mortality rate (36). Finally, while athymic nude mice

were able to clear an infection produced by vaccinia recombinants expressing

IL-2, injection of similarly infected mice with anti-IFN antisera resulted in

one-hundred percent mortality (107). It should be kept in mind that the

result of such studies are almost always dependent on the host and virus

species and on the routes of inoculation.

Pyrexia. Pyrexia, or fever, which can be induced by interleuken-1 (IL-1)

or by tumor necrosis factor (TNF), is a general systemic response to the

infection of a host by a variety of organisms (50). The importance of increases

in temperature on poxvirus infections has been studied to some degree in

both cell culture and in the animal. The concept known as "ceiling

temperature" was introduced in the 1960's when it was noticed that above

certain temperatures, the orthopoxviruses would not form pocks on the CAM

(17,68). At that time, it was also determined that ceiling temperatures could

be measured in cultured cells. The ceiling temperature was observed to vary

depending on the virus and thus became useful as a method to distinguish









52

between various species of orthopoxvirus. In infected animals, changes in

body temperature have been shown to effect the severity of myxomatosis. In

addition, the susceptibility of mice to infection with ectromelia was increased

100-fold by an 180C drop in their ambient temperature (160). Together these

results indicate that alterations in body temperature can effect viral disease

and, further, suggest that such alterations may play a role in limiting

poxvirus replication in the animal.

Complement. The complement system is an integral part of the non-

specific host response to infection and of the inflammatory process (80,91,133).

The system is composed of twenty plasma proteins, normally found in an

inactive state, as well as many cell-surface receptors. These receptors have

specificity for components of the complement pathway and include several

regulatory proteins which are thought to prevent accidental activation of the

system on host cell surfaces. Activation of the complement system results in

a cascading series of enzymatic activations with the final result being the

formation of a membrane attack complex (MAC) that forms pores in

biological membranes. The physiological consequences of complement

activation are opsonization of infecting organisms, cellular activation and

lysis of target cells by MAC induced osmotic swelling.

The complement cascade can be activated by different mechanisms

involving two separate pathways, the classical pathway and the alternative

pathway. Although these pathways differ in their initiation and their early









53

stages, they both result in the formation of an enzyme called C5-convertase

which is the key enzyme in activating the formation of the MAC. The

classical pathway is dependent on target-bound antibody for initiation. The

antigen-antibody complex is bound by complement component C1, resulting

in self-cleavage and activation of C1. Activated C1 catalyzes the assembly of

C3 convertase from components C2 and C4. This fluid-phase convertase

(C4b,2a), upon binding to a surface, then cleaves component C3, a fragment of

which remains associated with the C3 convertase altering it to a C5 convertase

(C4b,2a,3b). The cleavage of C5 by the C5 convertase leads to the assembly of

the MAC from components C6, C7, C8 and C9. Insertion of the MAC into the

target cell membrane results in osmotic swelling and lysis of the cell. MAC

insertion into the membrane of enveloped virions can also allow access of

degradative enzymes to the interior of the virion (91).

The alternative pathway is initiated by the reaction of component C3

with water resulting in C3(H20). This molecule can bind complement factor

B, which is then activated by factor D to give C3(H20),Bb a fluid phase enzyme

that acts as a C3 convertase. This newly formed convertase produces a

metastable C3b that covalently attaches to nearby surfaces and binds more

factor B. Activation of this surface-bound C3bB by factor D yields a surface

bound C3b convertase (C3b,Bb) that continues to convert C3 to C3b resulting

in a self-amplifying, activation loop. The addition of another C3b molecule to

the bound C3 convertase produces C3b,Bb,C3b which acts as a C5 convertase,









54

initiating assembly of the MAC as described above.

Because the action of the MAC is non-specific in terms of the

membranes it will permeabilize, and because activation of the complement

system results in numerous systemic effects, activation of this system is

tightly regulated at several points by both serum and cell-surface molecules

(133,210). Activation of the classical pathway is regulated by inhibition of C1

cleavage and by inhibition of C3 convertase formation. The latter inhibition

is mediated by the C4b binding protein (C4b BP) and factor I in the plasma as

well as by decay accelerating factor (DAF) and the C3b receptor (CR1) on the

cell-surface. Factor I and C4b binding protein, together, catabolize C4b and

promote the dissociation of the C3 convertase (C4b2b). DAF and CR1 inhibit

C3 convertase formation and promote the degradation of C4b and C3b by

factor I.

Similar control is exerted over the alternative pathway of activation by

the serum factors properdin (P), factor H (H), factor I (I) and the membrane

factors CR1 and membrane cofactor protein (MCP) (133,210). The C3

convertase (C3bBb) formed by the alternative pathway is unstable and rapidly

dissociates unless it is bound by P which stabilizes the complex and which

therefore acts as a positive regulator. Factor H is homologous with the C4b

binding protein and promotes the dissociation of C3b and Bb and also

promotes the degradation of C3b by factor I, providing negative regulation.

Whether C3b acts as binding site for factor B, resulting in the amplification









55

loop, or whether it is degraded by factor I resulting in inhibition, is

determined by the nature of the surface to which C3b is bound. Host

membranes contain CR1 and MCP that promote binding of factor H to C3b

resulting in inhibition of the cascade. Non-host surfaces on the other hand

lack these regulatory molecules and act as protected sites for C3b allowing

activation of the pathway. The protective nature of non-host surfaces also

seems to involve the carbohydrate composition of the surface in some

unknown manner.

With the cloning and sequencing of the genes coding for the

complement proteins it has become clear that many of these proteins can be

assigned to superfamilies based on structural homologies (210). It is thought

that many of these genes arose from a common progenitor by gene

duplication and exon shuffling and that these duplicated genes then evolved

in parallel maintaining similar structures and often similar functions. This

conservation of structure and function is seen in the SCR superfamily, the

members of which contain repeats of a conserved sequence known as a short

conserved repeat (SCR) element (94). This sequence, which is approximately

60 amino acids in length, is not strictly conserved but instead consists of non-

conserved sequences flanking several conserved amino acids within the

element. Many of the proteins that regulate complement activation, such as

factor H, C4b binding protein, DAF, and CR1 are members of the SCR

superfamily. While the actual tertiary structure of these proteins varies









56

widely, they all contain a variable number of SCR elements and they all share

similar functions, namely preventing the stable formation of the C3

convertase. It is tempting to speculate that the SCR domain mediates binding

to the convertase components C4a, C2a, C3b and Bb, however, two other

members of this family, the IL-2 receptor and factor XIIIb, a protein of the

coagulation pathway, show no such activity. Since all of the members of this

family do show binding activity, it is likely that the domain provides a

framework for a general binding structure, and that the specificity of binding

is determined by the variable residues located between the strictly conserved

core amino acids.

While complement activation results in MAC formation and lysis of

the target cell, it also has several other consequences, namely opsonization of

foreign particles and activation of the cellular components of the immune

system. Opsonization utilizing complement components is accomplished by

covalent attachment of factor I-mediated-degradation products of component

C3 to the foreign particle (80). This bound C3 is recognized by C3 receptors on

leukocytes enhancing phagocytosis by these cells.

In addition to mediating MAC formation and opsonization, the

proteins of the complement cascade are also the source of anaphylatoxins,

which are small biologically active peptides that affect the blood vessels,

smooth muscle and blood leukocytes (80). In vitro experiments have shown

that enzymatic treatment of C3, C4 and C5 produces the anaphylatoxins C3a,









57

C4a and C5a which, when injected into animals, result in smooth muscle

contraction and increased vascular permeability. In addition to these actions,

C5a has a dramatic effect on monocytes/macrophages and neutrophils, that is

mediated by binding to specific cellular receptors. Upon binding of C5a to its

receptor, the cells display increased oxidative metabolism, increased

adherence to surfaces, polarization and directed migration towards the C5a

source, and in the case of monocytes/macrophages, generation of IL-1, TNF-

alpha and IL-6.

The involvement of the complement pathways in poxvirus infectivity

and pathogenesis has been studied both in vitro and in vivo. Early in vitro

studies showed that antibody-mediated neutralization was greatly enhanced

by a heat-labile serum component (complement) (52). These studies have

been extended, and it has since been reported that complement can mediate

direct lysis of virions in addition to lysis of infected cells (28,29,36). While it is

thought that opsonization of virus particles is also important, this has yet to

be demonstrated. Activation of the alternative pathway in the absence of

antibody has been detected in fowlpox virus (FPV)-infected chick embryo cells

and vaccinia-infected tumor cell lines (36,137). This activation correlated

with increased killing of infected cells and decreased yields of virus progeny

suggesting that activation of this pathway may play a role in combating

infection.

In vivo, the role of complement has primarily been studied by









58

measuring complement levels in infected animals and by examining the

result of virus infections in complement-depleted animals (29,138).

Measurement of the levels of the products of complement activation has

shown that the complement system is active in FPV infected animals (106).

Injection of cobra venom factor (CFV) into animals results in a temporary

depletion of circulating complement levels. CFV treatment of chickens prior

to infection resulted in increased virus levels in primary lesions and a

decreased inflammatory response. The mortality rate for these animals was

one-hundred percent whereas it was zero percent for untreated animals.

Similar results have been obtained in chicken embryos. Together, the results

to date suggest that complement can enhance virus neutralization in vitro

and may play a role in resistance to viral infection.

NK Cells. Within the first 2-3 days following infection, non-specific

cytolytic lymphocyte (CTL) activity, attributable to natural killer (NK) cells,

can be demonstrated in the infected animal (2,36). NK cells are a bone

marrow-derived subset of lymphocytes found mainly in the blood and in the

spleen. The cells appear as large lymphocytes containing large cytoplasmic

granules and are often referred to as large granular lymphocytes. The cytolytic

activity displayed by NK cells is similar to that of the major histocompatibility

(MHC)-restricted CTL response in its mechanism; however, NK activity is not

restricted by MHC molecules. Although this activity is not specific, it is also

not random in that NK cells have the ability to kill tumor, but not normal,









59

cells and will also kill cells infected by some viruses but not by others (2,212).

Killing of target cells by NK cells involves two steps, granular excytosis

and secretion of a cell toxin (2). In the first mechanism, NK cells secrete the

contents of some of their cytoplasmic granules in areas of contact with target

cells. The consequence is the polymerization of a pore-forming protein

known as perforin that permeabilizes the target cell. The second mechanism

involves secretion of a cell toxin that enters the cell through the perforin

pores and kills the target cell by inducing it to undergo apoptosis, also known

as programmed cell death.

The importance of NK cells in combating poxvirus infections has been

studied by correlating NK activity with disease severity as well as by

measuring the effect of NK cell depletion on virus replication (32,101,190).

NK cytolytic activity in the spleen has been shown to be elevated following

intraperitoneal or intravenous inoculation of mice with vaccinia. In

addition, a strain of mice resistant to infection with ectromelia shows a

higher level of NK cell activity following inoculation with virus than do

mice that are susceptible to infection (101). These data suggest that NK cell

activity has a role in resistance to poxvirus infection. Evidence in support of

this comes from studies in which the NK cell population is depleted through

the use of antibodies against a prominent NK cell surface antigen (32,101).

Treatment of mice with anti-NK antisera prior to infection with ectromelia,

resulted in increased viral titers in the spleen and liver as well as an increase









60

in the severity of the disease following infection. Finally, while certain

strains of mice are normally resistant to infection by poxviruses, other strains

which have been derived from these mice and which contain mutations

associated with deficiencies in NK cells are able to contract lethal infections

(36). The above data strongly suggest a role for NK cells in the host response

to poxvirus infections. Recent studies suggest that it is not the cytolytic

activity of the NK cells but rather the production of IFN-gamma by these cells

that is crucial (107). These studies indicate that while NK cells are involved

in virus clearance, they are not sufficient to completely inhibit virus spread.

Inflammation. Prior to the development of a specific cellular immune

response, the host mounts a non-specific inflammatory response consisting of

vasodilation, edema and the appearance of inflammatory cells at the site of

infection (2,43). The cellular component of this response consists of

mononuclear phagocytes (monocytes/macrophages) and granulocytes, the

latter primarily of the type known as polymorphonuclear leukocytes (PMNs)

or neutrophils. These cells engulf and degrade foreign particles through the

action of degradative enzymes and reactive-oxygen intermediates. While

both PMNs and macrophages are capable of phagocytosing and degrading

foreign particles, they differ in that macrophages, but not PMNs, are involved

in antigen presentation and the generation of the late specific immune

response. Generally the PMNs are the first cells to appear, but they are soon

followed by the appearance of macrophages.









61

The movement of these professional phagocytes from the circulation

to the site of infection is directed by numerous proinflammatory mediators.

Some of these mediators serve to increase the mobility and phagocytic activity

of the inflammatory cells while others affect the vasculature near the

infection, causing vasodilation and the expression of cellular adhesion

molecules on the vascular endothelial cells. Examples of such mediators are

tumor necrosis factor (TNF), IFN-gamma and IL-1 (1,2). Still other mediators

serve as chemoattractants for the migrating phagocytic cells, forming a

concentration gradient beginning at the site of infection. Examples of this

type of mediator include IL-8, monocyte chemotactic protein-1 (MCP-1),

platelet activating factor (PAF), products of the kinin system and a product of

the complement cascade known as C5a (184). It is likely that there are many

more as-of-yet undefined proinflammatory mediators and it is also likely that

there is significant overlap between the types of mediators described above.

Like much of poxvirus pathogenesis, the exact role played by

inflammation in the inhibition and clearance of the viral infection is unclear.

Infection of mice with virulent strains of ectromelia fails to elicit a significant

inflammatory response until 7 days post-infection, a time when the

development of lesions is well under way (8,19). In contrast, inflammation

has been observed within 24 hours following inoculation of rabbits, mice and

guinea pigs with vaccinia virus or inoculation of rabbits with rabbitpox virus

(16,19,49,85,127,147,214). Such differences in the inflammatory response could









62

result from differences in the ability of the various viruses to either elicit or

suppress such a response. Perhaps the more virulent viruses are able

suppress inflammation until later in the infection. It is also likely that the

virus and host combination is important in the type of response generated.

In studies in which an inflammatory response has been reported, the initial

cell-type observed in this response has also varied between studies. This may

be due to variation in the inoculum size, routes of inoculation as well as the

above mentioned variation among the virus species.

Although inflammatory cells may be present at the site of infection, the

role these cells play in disease clearance or progression is unclear. In vitro, it

has been demonstrated that poxviruses are taken up when mixed with

phagocytic cells (11,161,213). In addition, when the virus is injected

intravenously into mice, more than ninety-percent is taken up by hepatic

littoral cells within a few minutes of inoculation (124). A point about which

there is little agreement, however, is the fate of the phagocytosed virus.

Results from some of this early work suggest that phagocytosis results in

destruction of virus or neutralization of infectivity in animals (11,52,63).

However, workers performing similar studies have also reported that

phagocytosis does not inhibit viral replication (15,69,76,161). It is possible that

macrophages and neutrophils destroy some virus following phagocytosis but

that enough virus survives to replicate. If correct, this model would mean

that leukocytes may actually aid in enhanced dissemination of the virus, an









63

idea that is supported by work showing that viable virus can be recovered

from the leukocytes of infected animals (36,69,124).

Recent studies, concerned with understanding the involvement of the

inflammatory response in poxvirus infections, have focused on the

inflammatory response generated following infection of the 11-day-old

chicken embryo. At this stage of development, the embryo lacks a mature T

and B cell response making the inflammatory response together with the

complement system the only systems of defense. In this regard, the CAM

makes an excellent system in which to study the non-specific response of the

host to a poxvirus infection. Infection of the egg at this time results in the

formation of characteristic lesions (pocks) on the chorioallantoic membrane

(CAM). RPV, cowpox virus (CPV) and certain strains of vaccinia produce red

pocks upon infection of the CAM. Microscopic examination of these pocks at

3-days post-infection reveals a region of ectodermal hyperplasia within the

pock with edema of the underlying mesoderm (68). The pocks are frequently

innervated by blood vessels which show extensive vasodilation. In addition,

necrosis and extravasation of erythrocytes are seen; however there is a

noticeable absence of inflammatory cells within the lesion. In contrast to

these red pocks, viruses containing deletions of the RPV SPI-2 (CPV crmA)

gene produce white pocks on the CAM. The SPI-2 protein has been shown to

inhibit an enzyme involved in the generation of inflammatory mediators

(see below). These white pocks produced by RPVASPI-2 show the same degree









64

of vasodilation and extravasation as the RPV red-pocks; however, in contrast

to the red pocks, the mesoderm of the white pocks is heavily infiltrated by

leukocytes. The hyperplasia and severe impaction of heterophils within the

lesion give the pock its white appearance. Similar results have been reported

for the naturally occurring white-pock mutants. The massive inflammatory

response observed in white-pocks correlated with a decreased level of virus

from within the pock suggesting that inflammation plays a role in virus

clearance in this system (142). However, even though there was a decrease in

the level of virus detected in these primary lesions, virus was still able to

travel to and infect internal organs of the embryo suggesting that the

inflammatory response by itself is not sufficient to clear the infection (36).

Recent work suggests that while inflammation by itself cannot contain the

viral infection, it is necessary for virus clearance and disease resolution (202).

In this work, treatment of mice with carrageenan, which depletes the

macrophage pool, prior to infection with ectromelia in resulted increased

virus replication. The above results, taken together, suggest that a

combination of inflammation and other host responses is necessary for

recovery from poxvirus infections.


Viral Proteins that Modulate Host Immune Responses


The success of the poxviruses as pathogens indicates that they have

evolved ways of evading or altering the host immune response to the









65

infection. Indeed, some of the most exciting work in this field in recent years

has been the discovery that the poxviruses produce several proteins that

either have homology with or interact with components of the host immune

response. To date the list includes a viral C4b binding protein (VCBP), an

inhibitor of the interleuken-l converting enzyme (crmA), a virally produced

secreted interleuken-1l receptor, secreted IFN receptors (alpha,beta,and

gamma) and a secreted TNF receptor. Those proteins for which an activity

has been described are briefly discussed below.

Viral C4b binding protein. The major secreted 35 kda protein of

vaccinia was shown to be a product of the C21L gene which is located in the

left, non-essential region of the genome (111). Sequence analysis of this gene

revealed four SCR elements covering the length of the protein, making it a

member of the SCR superfamily. Database searches using this protein

showed that it shares thirty-eight percent identity with the complement C4b

binding protein resulting in the viral protein being named the viral C4b

binding protein (VC4bBP). Subsequent studies have shown that purified

VC4bBP can bind to C4b preventing stability of the C3 convertase and

therefore inhibiting the complement cascade (110). VC4bBP was also shown

to prevent antibody-dependent complement-enhanced neutralization of

infectivity in a manner independent of complement C4b (99).

The importance of this protein during infection was demonstrated

using mutants lacking the VC4bBP ORF. The mortality rate caused by a









66

VC4bBP null mutant following intracranial injection was decreased

dramatically when compared to that of wild type virus. In addition, lesions

produced by this mutant following intradermal injection of rabbits were

smaller and healed more rapidly than those produced by wild type virus.

Together, these results indicate that inactivation of the complement system is

important for successful virus replication in these systems (36,99,110).

The crmA/Spi-2 protein. Evidence that the poxviruses could modulate

the immune response began to mount when it was realized that CPV white-

pock mutants had lost the ability to inhibit the inflammatory response on the

CAM. Further study narrowed the mutations to a single gene in the right,

non-essential region of the virus that was given the name crmA (cytokine

response modifier A) (155). Homologs have since been found in both VV and

RPV, although crmA isolated from CPV remains the most well characterized.

The product of the crmA gene is a 38 kDa cytoplasmic protein that has around

thirty percent identity with members of the serine protease inhibitor

superfamily which includes antithrombin and plasminogen activator

inhibitor-2 (PAI-2) (36,155). Biochemical analysis has shown that the crmA

protein is able to inhibit the interleuken-lf converting enzyme (ICE) which is

responsible for the production of IL-1 from pro-interleuken-l1 (158). The

failure of crmA mutants to inhibit inflammation most likely arises from the

failure to inhibit IL-1 production since IL-1 produced by macrophages is

known induce neutrophilia (50). ICE has also recently been implicated as part









67

of the apoptotic pathway, and it has been shown that crmA can inhibit

cytokine-induced apoptosis (197). Since cytotoxic-T cell killing is mediated

through apoptosis, inhibition of cell death by crmA suggests another

mechanism by which the virus can alter the immune response.

The importance of crmA in altering the host response has been well

characterized in the chicken embryo where deletion of this gene results in

heterophilia of the primary lesions (70). Interestingly, intranasal infection of

mice with viruses lacking crmA yield a mortality rate identical to that of wild

type virus. Finally, while RPV mutants of crmA still produce a severe disease

in rabbits following intradermal infection, the mortality rate is zero compared

with fifty to seventy-five percent for wild type RPV (R.Stern, M. Brooks,

Unpublished data). These differences in effect might be attributable to species

specificity or to differences in routes of inoculation.

Interleukin-l1 binding protein. Sequence analysis of DNA from the

right, non-essential region of vaccinia (WR) revealed that the predicted

protein product of the B15R gene is a member of the immunoglobulin

superfamily and shares homology with the binding domains of the human

and murine IL-1 receptors (183). This homology suggested that the viral

protein may bind IL-1, a hypothesis that was shown to be correct when IL-1

binding activity found in the supernatants of vaccinia infected cells was

abolished by deletion of the B15R ORF (187). The cloned B15R gene produces

a glycoprotein of 40 kDa that has specificity for IL-1 but not IL-la, and which









68

is able to compete with the natural IL-1 receptor for binding of IL-1 (7).

Inactivation of the viral IL-1 receptor from vaccinia virus has produced

contradictory effects in mice (7,187). Viruses lacking the B15R ORF were

attenuated when mice were inoculated intracranially but were more virulent

when inoculation was via the respiratory route. This result is not surprising

since increased IL-1 production and the subsequent immune response in the

lung may lead to increase pathological damage. This suggests that some viral

proteins can act to decrease the damage resulting from the host immune

response and may therefore act to make the virus less pathogenic.

Viral interferon receptors. The production of an IFN receptor by

poxviruses was first reported by groups studying myxoma virus but it has

since been shown that 17 of the orthopoxviruses, including vaccinia, cowpox

and rabbitpox, produce a secreted IFN receptor (5,130,205). Vaccinia has been

shown to produce two IFN receptors; an IFN-gamma receptor (IFN-yR) and

an IFN a/9 receptor (IFNa/fBR). The IFN-yR is a product of the B8R ORF from

the right non-essential region and the predicted amino acid sequence of this

gene shows an overall twenty-one percent identity with the extracellular

domain of the human and mouse IFN-yR. The activity of the viral receptors

is species specific and it has been suggested that these specificities may be

useful in determining the origins of the various viruses.

Viral TNF receptor. Open reading frames coding for proteins with

homology to the tumor necrosis factor receptor (TNFR) were first described









69

for Shope Fibroma Virus and myxoma virus (182,204). The SFV and myxoma

TNFR ORFs ( named T2) share forty and forty-three percent identity,

respectively, with the extracellular domain of human TNFR-1. Expression

from a recombinant T2 gene results in production of a glycoprotein of

approximately 55 kDa that is secreted and binds both TNF-a and TNF-B in

vitro. These recombinant viral TNFRs have also been shown to compete

with human TNFR for human TNF. Recently, the crmB gene of CPV, which

shares forty-two percent identity with the binding domain of human TNFR,

has also been shown to produce a 48 kDa protein that binds to TNF-a and

TNF-6 in vitro.

The importance of the TNF receptors in virus infection has been

examined in the chicken embryo and in rabbits. Infection of the CAM with a

CPV TNFR null mutant, resulted in the production of wild type red pocks

indicating no involvement of this protein in the pock color phenotype (97).

In contrast, deletion of the T2 ORF from myxoma virus resulted in

attenuation (204). The disease produce in rabbits following injection of this

virus was much less sever than that produced by wild type myxoma. In

addition, the mortality rate of the mutant was decreased by seventy percent in

comparison to the mortality rate of wt virus.









70



The B5R Protein of RPV


The B5R (the fifth ORF from the left end of the HindIII B fragment)

protein of RPV, the focus of this dissertation, is another poxvirus protein that

shares homology with components of the host immune system. Proteins

having nearly identical sequence have been shown to exist in both vaccinia

virus and variola virus (57,100,119,120,195). The homologies between the

various viruses and between B5R and the complement components will be

described below in detail.

The B5R ORF was first isolated from the Lister strain of vaccinia in

1991 by marker rescue of a low-neurovirulence strain of vaccinia lister known

as LC16m8. This virus is a naturally occurring variant that was of interest

because the plaques it formed on RK-13 cells were consistently smaller than

those produced by the parent strain, LC16mO. In addition, LC16m8 had an

altered host range and was unable to grow in Vero cells, unlike the parent

strain LC16mO. Marker rescue experiments resulted in the isolation of a gene

from the right-hand non-essential region of the genome that restored both

the ability to form large plaques on RK-13 cells and to grow in vero cells.

Based on these phenotypes the gene was originally called the plaque size/host

range (abbreviated ps/hr) gene. Functioning homologs of the ps/hr gene

have since been identified in the right end (HindIII B fragment) of the









71

vaccinia and RPV genomes (the B5R ORF) and of the variola genome (the B7

ORF).

The sequence of the B5R gene has been determined for several strains

of vaccinia, variola and for RPV allowing comparisons to be made between

viruses and with the protein database. Comparison of the predicted amino

acid sequence of the viral genes, as is illustrated in Figure 5, shows that

among the members of the poxviridae family for which B5R homolog

sequence is available, the protein sequence of the various B5R homologs has

been extremely well conserved. In all cases the degrees of similarity and

identity between the different homologs are greater than ninety-six percent

and ninety-three percent (Table 1), respectively suggesting the importance of

this protein to survival of the virus. The B5R protein of RPV is one-hundred

percent homologous with that of VV(Copenhagen) while it is most diverged

from the variola strains India and Bangladesh, although it should be noted

that the differences here are still extremely small, only six percent at the level

of amino acid identity. Most of the differences observed are the result of

conservative substitutions of amino acids and it is likely that they have little

effect on the overall protein structure.

Analysis of the RPV B5R amino acid sequence reveals that this protein

has N-terminal and a C-terminal hydrophobic regions that in the vaccinia

protein have been shown to function as a signal sequence and a membrane

anchor, respectively (100). In addition, the protein, which is shown




























Figure 5. Alignment of the orthopoxvirus B5R homologs. Shown is an
alignment of the complete predicted amino acid sequences from the currently
available B5R homolog genes. Boxed areas indicate differences within the
sequences of the various proteins. Black bars underline potential
glycosylation sites. UTR, Utrecht; COP, Copenhagen; WR, Western Reserve;
LIS, Lister; BAN, Bangladesh; IND, India.











73

60
RPV(UTR) MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLTSTETSFN )KVTFTC Q 3H S PNAV
VV(COP) MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLTSTETSF KN 2KVTFTC Q 31H S PNAV
VV(WR) MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLTSTETSFN 3KKVTFTCEQ3:3 S DPNAV
VV(LIS) MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLTSTETSFN 3KKVTFTCEQ3H S DPNAV
VAR(BAN) MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLTSTETSFNDK KVTFTCS3Y L PNAV
VAR(IND) MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLTSTETSFNK 2KVTFTCE S 3YL PNAV

120
RPV(Utr) CETDKWKYENPCKKMCTVSD I EL NKPLYEV STMS S ETKYFRCEEKNGNISWU
VV(COP) CETDKWKYENPCKKMCTVSDI 3EL NIPLYEW SlS NGETKYFRCEEKNGNTSW
VV(WR) CETDKWKYENPCKKMCTVSD I3ELi NPLYEVI SATI S NETKYFRCEEKNGNHSWN
VV(LIS) CETDKWKYENPCKKMCTVSD V3ELY DPLYEVI STM S NETKYFRCEEKNGINTWN
VAR(BAN) CETDKWKYENPCKKMCTVSD V3EL PLYIEVI=I TKYFRCEEKNGNTSWN
VAR(IND) CETDKWKYENPCKKMCTVSDi ELI PLYE TKYFRCEEKNGNTSW

180
RPV(UTR) DTVTCPNAEC P QI EGSCQPVKEKYSFGE I INCDVGYEVIGASYISTANSWNVIP
VV(COP) DIVTCPNAECC P QI E GSCQPVKEKYSFGEY FINCDVGYEVIGASY ISTANSWNVIP
VV(WR) DTVTCPNAECC P QI EiGSCQPVKEKYSFGEYl INCDVGYEVIGASYIS:TANSWNVIP
VV(LIS) DTVTCPNAEC P QI EGSCQPVKEKYSFGEYI TINCDVGYEVIGASYIS2TANSWNVIP
VAR(BAN) DTVTCPNAEC S QID IGSCQPVKEKYSFG H INCDVGYEVIGASY T:TANSWNVIP
VAR(IND) DTVTCPNAEC S QI EGSCQPVKEKYSFG H INCDVGYEVIGASYI TANSWNVIP

240
RPV(UTR) SCQQKCI ?SLSNGLISGSTFSIGGVIHLSCKSGF TGSPSSTCIDGKWNI V-EI VT
VV(COP) SCQQKCII SLSNGLISGSTFSIGGVIHLSCKSGF ITGSPSSTCIDGKWNIV T IV
VV(WR) SCQQKCIM SLSNGLISGSTFSIGGVIHLSCKSGFT TGSPSSTCIDGKWNI Vl I VT
VV(LIS) SCQQKCIM ?SLSNGLISGSTFSIGGVIHLSCKSG T TGSPSSTCIDGKWNI I T VS
VAR(BAN) SCQQKCII PSLSNGLISGSTFSIGGVIHLSCKSGFI TGSPSSTCIDGKWN Vl I I S
VAR(IND) SCQQKCI I?SLSNGLISGSTFSIGGVIHLSCKSGF 4TGSPSSTCIDGKWNI VlJ IS

300
RPV(UTR) NEE DP DGPDDETDLSKLSKDVVQYEQEIESLEATYHIIIVALTIMGVIFLISVIVLV
VV(COP) NEE DPDGPDDETDLSKLSKDVVQYEQEIESLEATYHIIIVALTIMGVIFLISVIVLV
VV(WR) NE EDP DGPDDETDLSKLSKDVVQYEQEIESLEATYHIIIVALTIMGVIFLISVIVLV
VV(LIS) NEK DP DGPDDETDLSKLSKDVVQYEQEIESLEATYHIIIVALTIMGVIFLISVIVLV
VAR(BAN) NEE DP DGPDDETDLSKLSKDVVQYEQEIESLEATYHIIIVALTIMGVIFLISVIVLV
VAR(IND) NE DP DGPDDETDLSKLSKDVVQYEQEIESLEATYHIIIVALTIMGVIFLISVIVLV

317
RPV(UTR) CS NNDQYKFHKLIP
VV(COP) CSCD NNDQYKFHKLIP
VV(WR) CSC DNNDQYKFHKLIP
VV(LIS) CSCD NNDQYKFHKLIP
VAR(BAN) CSN NNDQYKFHKLIL
VAR(IND) CSCEQNNDQYKFHKLIL















Table 1. Homologies of the B5R homologs. The table shows the percent identity and
similarity between the predicted amino acid sequences from the available
orthopoxviruses B5R homologs. Alignments were performed using the Bestfit
software. ( Software was provided by the Interdisciplinary Center for Biotechnology
Research as part of the Wisconsin Package).

RPV VV(WR) VV(Lis) VV(Cop) Var(Ind) Var(Ban)
RPV 100 / 100 ------------- ----------- ---------- ----------- --------------
VV(WR) 98.7 / 99.4 100 / 100 -------- ---------- -------- ---------
VV(Lis) 96.5/ 98.4 97.8 / 99.1 100/ 100 --------- -------- -------
VV(Cop) 100 / 100 98.7 / 99.4 96.5 / 98.4 100 / 100 ---------- ----------
Var(Ind) 93.4/ 96.2 93.4/ 96.2 93.0/ 95.8 93.4/ 96.2 100 / 100 ------
Var(Ban) 93.4/ 96.2 93.4/ 96.2 93.0/ 95.8 93.4/ 96.2 100/ 100 100/ 100









75

schematically in Figure 6, contains four contiguous short consensus repeat

(SCR) elements, which comprise the first seventy percent of the protein,

making it a member of the SCR superfamily. Alignment of these regions

with the SCR domains from several other members of this superfamily is

shown in Figure 7. While it is obvious that the overall sequences differ, it is

also clear that the highly conserved residues that define the SCR domain are

present in the B5R protein. These include the essential cysteine and proline

residues that begin and end each domain as well as several less well

conserved residues found at characteristic positions within the domain.

Included in this alignment are the human and viral C4b binding proteins

which illustrate how divergent two proteins may be in their sequence and yet

still share the same activity.

Comparison of the RPV B5R predicted amino acid sequence with the

current database using BLAST reveals that this protein shares homology with

several complement control proteins, with the areas of greatest homology

falling, not surprisingly, within the SCR domains. B5R shares the greatest

degree of homology with the complement components factor H and the C4B

binding protein, both of which are inhibitors of complement activation. An

illustration of the B5R regions that share this homology as well as an

indication of the degree of homology are shown in Figure 8. The regions of

homology basically cover two separate areas of the B5R protein, one region

within SCR I and a second region within SCRs II and/or III. The region of






















Figure 6. Schematic representation of the B5R protein. The 317 amino acid RPV B5R protein is represented by the
narrow black line. The filled black box at the left-hand end represents the amino terminal signal sequence region.
The four SCR domains are represented by the white boxes labeled I through IV. The hatched box represents the
membrane anchor sequence. The arrow above the illustration indicates the calculated position of the cleavage site.
The numbers above the figure indicate amino acid positions and are shown for reference.











N





















OOIAI E In, ^ I----- *

ooe TZ szzT osi s






















Figure 7. The B5R protein is a member of the SCR superfamily. The figure shows the SCR (short conserved repeat)
domains of the RPV B5R protein aligned with the SCR domains from several other members of the SCR
superfamily. Letters in bold below the sequences denote residues conserved in over 95% of the members of the
family. These conserved residues are also highlighted in yellow. Letters in plain text below the alignment denote
the residues found in approximately 60% of the members of the family. Gaps were introduced to maximize
alignment of conserved residues. F13B, human factor 13b; HumFH, human factor H; VC4bp, viral C4b binding
protein; C4bp, mouse C4b binding protein.











SCR I

F13B NLTFIIILIISGELYAEEKPCGFPHVENGRIAQYYYTFKSFYFPMSIDKI HumFH VCRKGEWVALNPLRKCQKRPCGHPGDTPFG.TFTLT... GGNVFEYGVKA .YPCNEGYQLLGEINYRE..CDT WTND. CEVVKC.LPVTAPENG
RPVB5 MKTISVVTLLCVLPAVVYSTCTVPTMNNAKLT....... STETSFNNNQKVFrCDQGYHSSDPNAV .... CETDK.WKYEN. .CKK.MCTVSDYISELY
VC4bp MKVESVTFLTLLGIGCVLSCCTIPSRPINMKFKNSVETDANA.NYNIGDTIIYLMCLPGYRKQKMGPIYAK.CTGp. WTLFNQ..CIKRRCPSPRDIDNGQ
C4bp ...................NCGPPPTLSFAAPMDITLTE...TRFKTGTIJ C P IFC GF C GW PC C P
LY Y S
V A

SCR II SCR III

B13B SD LLYKIQE ......GCAS T KDEEVVQCLS3.. WS.SQo rCRKEHETCLAPE IN YSTTQKTRKVKDKVQYECATGYYTAGGKKT
HumFH KI.VSSAMEPDREY = QAVRFVCNSG I EEMHCSD..DI..FWSKEKPKCVEIS..CKSPD.VN3PISQKII KENERFQYKCNMGYEY.SERGD
RPVB5 NKPEVNSTMTL. .....SCN.GETK ....CEE.KN ~TSWN.DTYTCPNAE..CQ.PLL H 3CQPVKEK 3FGEYMTINCDVGYEV... IGA
VC4bp LD.II GVDFGSS.II .....SCNSGYHLI SKSYCELGSTSMVWNPEApCESVK..CQSPPS I NHNGYEDFfPDGSVVTYSCNSGYSL ...IGN
C4bp VE.IKTDLSFGSQI .......SCSEGFFLII TTSRCEVQDR .VGWSHPLRCEIVK..CKPPPE JIHSGEENFfkYGFSVTYSCDPRFSL...LGH
I F C GF G C G W PC C P I G F C GF
L Y Y L Y Y
V V

SCR IV

F13B EEVECLTY.....GWSLT.PKCTKLKCSSLEI GFH V.KQTBEGrlVQFFCHEN I EGSDL.IQCYNF.GWY PSPVCEGRRNRC
HumFH ..AVCTES.....GWRPL.PSCEEKSCDNPYI. G )Y. PLRIK ITGE ITYQCRNG YP TRGNTAKCTST.GWIPM.PRCTLKP...
RPVB5 SYISCTAN..... SWNVI.PSCQQ.KCDIPSL. SFG.SI. S..T IGVIHLSCKSG I GSPS.STCIDG.KWNP/LPICVRTNEEF
VC4bp SGVLCSGG .....EWSDP.PTCQIVKCPHPqII. S GLS 3FKRS YN DFKCKYG GSSS.STCSPGNTWPELPKCVR
C4bp ASISCTVENETIGVWRPSPPTCEKITCRKPE .S 3FGPI YKIVFKCQKG GSSV.IHCDADSKW PPACEPNS..C
C W PC C PI G S F D C F L C WP PC
L Y Y
V






















Figure 8. Regions of B5R sharing homology with members of the SCR superfamily. The narrow black bar at the top
of the figure depicts the 317 amino acids of the RPV B5R protein. The SCR regions (I through IV) and the membrane
anchor domain are depicted by the white boxes and hatched box, respectively. Listed below are three members of the
SCR superfamily with which B5R shares the greatest homology. The short black bars denote the relative location
and size of the regions of the B5R protein that show this homology. The numbers shown indicate percent identity
and similarity within the homologous regions. The numbers above the figure indicate anino acid positions.









75 150 225 300
I I I L II E IIIiZ I-*Z-t IVI


---- I 1 34/59 Coagulation
18/52 30/47 34/59 Coagulation
iFactor XIIIB




Mouse C4B
34/62 42/57 40/60 48/65 binding protein
40/80


25 Complement
30/69 128/60 I 23/57 Factor H
46Factor H
45/63 40/56






0-0o




Full Text

PAGE 1

THE CONTRIBUTION OF THE RPV B5R PROTEIN TO VIRUS VIRULENCE By RICHARD J. STERN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1996

PAGE 2

Copyright 1996 by Richard James Stern

PAGE 3

To Rita, who saved me from myself and without whose love and friendship this journey would have been unbearable.

PAGE 4

Its cold comfort To the ones without it To know how they struggled And how they suffered about it. If their lives were exotic and strange They would likely have gladly exchanged them For something, a little more plain Maybe something, a little more sane We each pay a fabulous price For our visions of paradise But a spirit with a vision, Is a dream, with a mission. -"Mission", Rush

PAGE 5

ACKNOWLEDGMENTS Although only a single name appears on this dissertation, the making of a PhD. is a process that involves a large number of people in a dozen different ways. Anyone who says otherwise is either lying or is blinded by their own ignorance. In acknowledgment of this, I would like to take this opportunity to thank those people who I feel contributed to my education in some significant way. First, and most importantly, I would like to thank my wife, Rita, for her love and unending support on this flight of fancy as well as for her patience in waiting for me to come alive again. Her belief in me carried me through considerable periods of time when I took this all much too seriously and lost The Way. I will forever admire the strength she showed during this period of our lives. In addition, I would like to thank my friends and family for their encouragement, support and for smiling politely while I explained what it is I do. A heartfelt thanks goes to Bob and Anne Badger for the friendship, for housing me during the finale and, well.. .for everything. What more can I say? I would like to thank Dick Moyer for the opportunity to work in his lab as well as for his advice and guidance in this work. Oh yeah, and thanks for v

PAGE 6

the pretzels. In addition, I thank the remaining members of my committee, Maureen Good enow, Paul Gulig and Maurice Swanson for their guidance in this work and for their deftness at turning the screws when it became necessary. I would also like to thank Jim Thompson for his suggestions on this project and for his considerable help getting me started on the animal work. A special thanks goes to Parker Small for his insightful advice and guidance during this odyssey as well as for his infectious enthusiasm for science, teaching and life in general. I owe a good deal of thanks to all members of the Moyer continuum for their suggestions, advice and criticisms and for the many hours spent kicking around ideas just to see where they would lead. In particular, I wish to thank Michael Brooks and Pierre Musey for the many interesting and enlightening discussions, scientific and otherwise, and for frequently calling "bullshit". They kept me honest. Special thanks go to Michael Duke and Dorothy Smith for keeping everything in the lab running smoothly as well as for the friendship we shared during my stay. I would particularly like to thank Mike Duke for the many hours of philosophical discussion, the laughter, the homebrews and the humble sense of balance and common sense wisdom he brought to the lab. Of course, being a Libra, he could hardly do otherwise. I am indebted to the staff of the Department of Molecular Genetics and Microbiology for making the process work smoothly. In particular, I would like to thank Peggy Kidder, Vicky Parrot, Beverly Anderson and Brad Moore

PAGE 7

for placing orders "ASAP' with a smile and for treating me kindly the many times I asked, "What form?". My utmost gratitude goes to Joyce Conner for her excellence in making sure I was registered and in compliance with the rules of the Graduate School. Thanks to the efforts of these people, for seven years I was able to remain blissfully ignorant of the rules and how to follow them. Finally I wish to acknowledge Victor Sapirstein for encouraging me to take this journey and George Stone and Ruth Gubitz who warned me about the whole thing in the first place. I finally understand. vii

PAGE 8

TABLE OF CONTENTS ACKNOWLEDGMENTS v LIST OF TABLES x LIST OF FIGURES xi GLOSSARY OF ABBREVIATIONS xiv ABSTRACT xvii CHAPTERS I INTRODUCTION 1 Introduction 1 Poxvirus Classification 6 Poxvirus Morphology 8 The Poxvirus Genome and its Organization 14 Poxvirus Biology 19 The Replication Cycle 19 Some Musings on the Significance of Enveloped Virus 41 Poxvirus Pathogenesis 43 Non-specific Host Responses to Infection 48 Viral Proteins that Modulate the Host Immune Response 64 The B5R Protein of RPV 70 H MATERIALS AND METHODS 84 Recombinant DNA Techniques 84 Virological Techniques 88 Construction of Mutant Viruses 98 Protein Techniques 110 viii

PAGE 9

ffl THE INVOLVEMENT OF THE RPV B5R PROTEIN IN HOST RANGE VIRULENCE AND INFLAMMATION 117 Introduction 117 Results 122 Cell Culture 122 Animal Work 135 Discussion 178 IV THE IMPORTANCE OF THE MEMBRANE-BOUND AND SECRETED FORMS OF B5R IN MORPHOGENESIS AND POCK MORPHOLOGY 191 Introduction 191 Results 195 Construction of a Mutant Producing only a Secreted Form of the B5R Protein 195 Construction of Mutants Producing only Membrane Bound B5R Protein 221 Discussion 241 V SUMMARY, DISCUSSION AND FUTURE DIRECTIONS .... 253 Summary 253 Future Directions 271 LIST OF REFERENCES 275 BIOGRAPHICAL SKETCH 295 ix

PAGE 10

LIST OF TABLES Table Page 1 Homologies of the B5R Homologues 74 2 Symptomology Resulting from Poxvirus Infection of Rabbits 141 3 Summary of the Pathology Produced by RPV and RPVAB5R .... 146 4 Summary of the Pulmonary Pathology in Infected Mice 153 5 Virus Titers from Lungs of Infected Mice 162 6 Histopathology of the CAM Resulting from Virus Infection 171 7 Symptoms resulting from Poxvirus Infection of Rabbits 224 x

PAGE 11

LIST OF FIGURES Figure Page 1 The Structure of a Poxvirus 10 2 Organization and Structure of the Poxvirus Genome 17 3 The Life Cycle of a Poxvirus 21 4 Model for Replication of the Poxvirus Genome 35 5 Alignment of the Orthopoxvirus B5R Homologues 73 6 Schematic Representation of the B5R Protein 77 7 The B5R Protein is a Member of the SCR Superfamily 79 8 Regions of the B5R Protein Sharing Homology with Members of the SCR Superfamily 81 9 Topology of the B5R Protein of RPV 101 10 Scheme for Construction of RPVB5R-T 104 11 Location of the Primers used to Construct the RPVB5R:X Mutants 108 12 Construction and Characterization of RPVAB5R 124 13 Analysis of RPV and RPVAB5R Infected Cell and Supernatant Extracts for the Presence of the B5R Protein 127 14 Plaque Formation of RPV and RPVAB5R on Various Cell Lines 130 xi

PAGE 12

15 The Effect of the B5R Gene on the Ability to Form Progeny Virus 134 16 RPV and RPVAB5R Pock Phenotype on the CAM 137 17 Rectal Temperatures in RPV and RPVAB5R Infected Rabbits 140 18 Histopathology of RPV and RPVAB5R Infected Rabbit Skin 145 19 Weight Change in Mice Following Intranasal Infection 149 20 Lung Sections from Intranasally Infected Mice 152 21 Growth of RPV and RPVAB5R in Mouse Lung 157 22 Effects of Dexamethasone on Virus Replication in Mice 160 23 NBT Staining of CAMs 167 24 Histopathology in the CAM following Infection with Virus 169 25 Low Power Magnification of RPV and RPVAB5R Pocks 174 26 Effect of Dexamethasone on Pock Color in Eggs 177 27 Construction and Characterization of RPVB5R-T 198 28 Peptide-N-glycosidase Treatment of B5R and B5R-T Proteins 201 29 Western Blot Analysis of IMV, IEV and EEV 204 30 Western Blot Analysis of Na 2 C03 Treated RPV or RPVB5R-T EEV 207 31 Plaque Size and Host Range of RPVB5R-T 210 32 The Ability of RPVB5R-T to Produce Virus in RK-13 and CEF Cells 214 33 Behavior of RPVB5R-T on the CAM 217 34 Weight Change in RPV and RPVB5R-T Infected Mice 220 35 Rectal Temperature of Infected Rabbits 223 xii

PAGE 13

36 Estimation of the Cleavage Site of B5R 228 37 Construction and Characterization of the RPVB5R:X Mutants 233 38 Plaque Phenotype of the RPVB5R:X Mutants on RK-13 Cells 237 39 Plaque Formation on CEF Cells by the RPVB5R:X Mutants 239 40 Pock Morphology of the RPVB5R:X Mutants 243 xiii

PAGE 14

GLOSSARY of ABBREVIATIONS A A Amino acid bp Base pair 6-ME Beta-mercaptoethanol BSA Bovine serum albumin BudR Bromodeoxyuridine C Centigrade CAM Chorioallantoic membrane cm Centimeter CPE Cytopathic effect CPV Cowpoxvirus CsCl Cesium chloride dATP Deoxyadenosine triphosphate dCTP Deoxycytosine triphosphate dGTP Deoxyguanosine triphosphate dTTP Deoxythymidine triphosphate DMSO Dimethyl sulfoxide EEV Extracellular Enveloped Virus EDTA Disodium ethylenediamine tetracetic acid FBS Fetal bovine serum G Gravity gm Grams xiv

PAGE 15

("Iiianvl-ribosvlrihosnhotransf erase I-fvHrfwhlnrif* Arid TFV I LL V TntrArplliilpir Fnvplnnpd Virn^ J. 1 1 1 1 Cl V-vT I 1 U I CI L 1_ 1 1 V ClvUvU V 11 U J ITR Inverted terminal repeat TMV 1 1V1 V Tntrarellnlar Matiirp Virus 1 1 1 1 1 Cl\_Cl 1 14 1 Cl 1 tYUIlUI C V 1 1 UO kb Kilobase kbo IX L/ L/ Kilobase nairs x\LS(\ rv'il nd Alton rvi i cii i\j 1 1 kcr V'i 1 no-ram JViiv/tii din TY1 A h 11 1.TA. l* Monoclonal Antibodv mCi Millicuries MEM Minimal essential medium IVf pfh Anol 1V1C llldllUl mg WAi 111 c ra m q IVllllIiLl dlllO ill .U. I lVrt iiltirilifitv nf infpftifin ml lit X Millili tpr J V 1 11 1 I 1 J IV. 1 in in A^i lit m pip r lVlilllXIlC lei m M xVllililllOicir MPA 1V11 r\ MvconHpnoli r Arid xvi y lul/i ivi iv/iiv. nvju NBT NitrobltiP tPtra7oliuTn ] 1 111 V/ L-* 1 V4_ \* IV. 11 V H V.l 111 NaCl Sodium rhloridp Matron ^rH iuro rArmn Afp •jUUlUlli Cell LJLMldlC NaOH Sodium hvdrovidp M A \f, C7\JU1UII1 dZlUc 1\ 1 4U lNoniaet r-4U ORF Open reading frame pM Picomolar PAGE Polyacrylamide gel electrophoresis XV

PAGE 16

PBS Phosnhate buffered saline PEG Polvethvlene tdvcol nfu Plaaue forminff unit X 1UU d V X V 1 XX 111 1 C-j U Ill L PPO 2 5-diohenvl-oxazole RPV RabbitDox virus S n H i 1 1 tn H n H p rv 1
PAGE 17

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 CONTRIBUTION OF THE RPV B5R PROTEIN TO VIRUS VIRULENCE By Richard J. Stern August, 1996 Chairperson: Richard W. Moyer, Ph.D. Major Department: Molecular Genetics and Microbiology Poxviruses produce a variety of proteins that modulate the host immune response. The predicted product of the rabbitpox virus (RPV) B5R gene, which exists in a membrane-bound and secreted form, has homology to complement factors involved in regulation of inflammation. This homology suggests a mechanism by which the B5R protein might influence virus virulence. This study was undertaken to determine if either form of the B5R protein contributes to virulence by affecting the host inflammatory response. xvii

PAGE 18

To examine the overall contribution of the B5R protein to virulence, a B5R null mutant (RPVAB5R) was constructed. In cell culture, RPVAB5R showed a cell-line specific defect in production of the enveloped forms of virus, correlating with an inability to form plaques on several normally permissive cells. In animal studies, RPVAB5R was shown to be attenuated through a mechanism not involving enhancement of the host inflammatory response. The roles of the individual forms of B5R were tested by construction of mutants that solely produced either the membrane-bound or secreted form of the protein but not both. The membrane bound form was shown to be essential for virus growth in cell culture and virulence in animals. Mutants failing to produce this form had a phenotype identical to the null mutant. Mutants unable to produce the secreted form of B5R had no measurable effect on virus growth in cell culture or on the pock color (morphology) produced on the chorioallantoic membrane (CAM) of chicken eggs indicating that this form of the protein is dispensable in these systems. This work shows that although B5R has homology to regulators of inflammation the role of the protein in virulence is likely not through modulation of the host inflammatory response. The results strongly suggest that the attenuation observed in B5R null mutants results from a defect in production and release of mature enveloped virus particles, as demonstrated in cell culture. In addition, these results demonstrate it is the membranexviii

PAGE 19

bound B5R, not the secreted form that is responsible for the activity of this protein in these systems. xix

PAGE 20

CHAPTER 1 INTRODUCTION AND BACKGROUND Introduction The Tao doesn't take sides; it gives birth to both good and evil. The Master doesn't take sides; he welcomes both saints and sinners. -Lao-tzu, translated from the Tao Te Ching by Stephen Mitchell DNA neither knozvs nor cares. DNA just is. And zve dance to its tune. -Richard Dawkins The vast majority of people living in the developed world today can expect long lives virtually free of serious threat from infectious diseases. It is a tribute to mankind's ingenuity and adaptability that the expectations of such a life are often taken for granted and, in fact, have come to be considered the normal condition. In reality, the current balance in our ecological relationship with microscopic pathogens is a relatively recent aberration in recorded human history, throughout which disease resulting from infection has been a consistent fact of life (122). It has only been over the last two centuries that advances in vaccination, simple sanitation and, most recently, the development of antibiotics, have significantly increased both the length and quality of life (191). While these developments have reduced the threat 1

PAGE 21

of infectious disease, the rapid evolution and adaptability of micro-organisms makes it inevitable that we will forever continue to see the emergence of pathogens, both familiar and unknown. If the current, favorable balance is to be maintained and improved, there must be a continuing evolution in our understanding of the abilities of pathogenic organisms to cause disease. Although medical research has resulted in tremendous progress in the prevention and treatment of infections of both bacteria and viruses, the molecular mechanisms underlying the ability of these organisms to cause serious disease are only beginning to be understood. While the use of antibiotics has allowed the successful treatment of most bacterial infections, the recent emergence of resistant strains threatens to make these compounds useless. Even more troublesome is the fact that progress in the development of specific, effective treatments for viral infections has been both difficult and slow with the result that many viral infections remain a significant health threat. Understanding both the host response to a viral infection as well as the mechanisms by which the virus is able to persist and spread in the presence of such a response would provide great insight in the development of new treatments for virally induced diseases. While many viruses are known to cause disease in humans, the poxviruses, although no longer a significant health threat, offer an excellent model system in which to study virus host interactions. Within this family are found viruses that naturally infect a wide range of animal species offering

PAGE 22

researchers several animal models with which to study various aspects of viral pathogenesis. Several of these viruses are known to be extremely virulent within their natural host, dramatic examples being variola major virus, the causative agent of smallpox in humans and myxoma virus which causes myxomatosis in the European rabbit (128). Infection with variola major, which is believed to have been eradicated from the wild, results in a severe, disseminated infection and produces a mortality rate of twenty to thirty percent (67). Another member of this family, which causes a severe infection in rabbits, is rabbitpox virus (RPV), which was first isolated during a spontaneous outbreak of infection in an isolated rabbit colony at the Rockefeller Institute in 1932 (85,86). Animals infected with RPV develop lesions ('pocks') on all parts of their skin, have high fevers and often display signs of hemorrhage from the nares. With regard to these symptoms, the disease resulting from infection with RPV has many feature characteristic of smallpox in man, although a significant difference is seen in the mortality rates, which in the case of an RPV infection is about fifty to ninety percent (16). This ability of RPV to cause a lethal infection combined with the ready availability of an animal model makes this virus an attractive candidate for studying the effect of specific poxvirus genes on virulence. The pathogenicities of variola and RPV are an indication that these organisms are well adapted for survival in their hosts. Further, although the infected host mounts an aggressive response to the infection, these viruses

PAGE 23

4 are able to survive and thrive within the host animal suggesting that they have evolved mechanisms to contend with or evade the host immune system. This is indeed the case. Recent work has resulted in the discovery of several poxvirus proteins that, in vitro, are able to neutralize specific aspects of the host immune system (36). Many of these proteins either interact directly with or have homology to critical components of the host immune system. It is believed and in some cases has been demonstrated that the interactions of these proteins with selective components of the host immune system are essential to the ability of the virus to evade the host defenses. Examples of such proteins include a secreted inter leuken16 receptor (7,187), a TNF receptor (50,97,178,182,204), a receptor for IFN-gamma (130,131,194,205), a complement C4b-binding protein (99,110,111) and a serine protease inhibitor, SPI-2, also known as crmA in cowpox virus (CPV) (142,155,158). The crmA protein, one of the most thoroughly studied proteins of this group, has been shown to inhibit the interleukin-16 converting enzyme (ICE), in vitro, and to be involved in the inhibition of inflammation in vivo. The study of such proteins will lead to a better understanding of the immune response to viral infections resulting in improved treatment for viral diseases. The work contained in this dissertation focuses on the B5R protein of RPV. This recently discovered protein has homology to both the complement proteins factor H and the C4b binding protein (57,100,119,195). These proteins prevent activation of the complement cascade and, as a

PAGE 24

5 consequence, inhibit the elicitation of the host inflammatory response. These homologies suggest a mechanism by which B5R may act to enhance viral virulence through inhibition of complement activation and, as a consequence, inhibition of inflammation. Previous work on B5R in both vaccinia virus and RPV has focused on the expression and localization of the B5R protein in virus-infected tissue culture cells (57,100,119). These studies showed that a 45 Kda form of B5R is found as a component of extracellular enveloped virus (EEV), one form of infectious virus, but not within intracellular mature virus (IMV) a progenitor of EEV (57,100,119). In addition, a 38 KDa form of the protein, which is derived from the larger form by processing, can be found in the supernatant from infected cells (119). The work presented in chapter three focuses on the contribution of B5R to viral virulence and, in particular, its effect on the inflammatory response to the infection. The work presented in chapter four examines the importance of the individual forms of the B5R protein in virus replication and virulence. For my studies, the overall role of the B5R protein was first examined by construction of a B5R null mutant, RPVAB5R, which was unable to produce either form of this protein. In agreement with what has been observed for similar mutants in vaccinia, this mutation resulted in a specific interruption of EEV morphogenesis. Unlike what was observed with vaccinia, however, this defect was observed to be cell-line dependent (119)(this dissertation). RPVAB5R was also observed to be severely attenuated

PAGE 25

6 in animals; however, this attenuation was not found to involve the host inflammatory response. Although the growth of RPVAB5R was inhibited in the animal, there was no increase in the inflammatory response to RPVAB5R when compared to the response generated against RPV. In addition, attenuation of RPVAB5R was not lessened under conditions in which the ability of the animal to mount an inflammatory response was inhibited. The conclusion drawn from this work is that B5R has no role in altering the host inflammatory response and that the attenuation of RPVAB5R results from the defect in morphogenesis observed in cell culture. The importance of the various forms of the B5R protein was examined by construction of mutants that produced only membrane-bound or only secreted B5R. In all systems tested, failure to produce the membrane-bound form of B5R resulted in a phenotype identical to that observed with RPVAB5R. In contrast, failure to produce the secreted form of B5R had no effect in cell culture or on pock color on the CAM and these mutants behaved in a manner identical to RPV. These results indicate that membrane bound B5R protein is essential for morphogenesis and virulence and that the secreted form of B5R has no obvious function in cell culture or on the CAM. Poxvirus Classification The Poxviridae are an extensive family of large DNA viruses. The

PAGE 26

7 members of this family are categorized, based on their host range, into two sub-families, the Chordopoxviridae, which are viruses that infect vertebrates, and the Entemopoxviridae, which are viruses that infect insects (128). The sub-families are further divided into several genera, and individual species of virus have historically been placed within a genus based on their shared antigenicity as well as similarities in morphology and host range (67). More recently, restriction fragment length polymorphism (RFLP) analysis, in combination with complete or partial sequence from individual viral genomes (79,102), has allowed a more precise means for measuring the relatedness of individual viruses. These genetic analyses have supported the earlier taxonomical assignments. Rabbitpox virus, the virus used throughout this work, is a member of the Orthopoxvirus genus, which contains at least seventeen members. Of all the poxviruses, the most heavily studied members are found within this genus. This is due in large part to the medical importance and usefulness of two of its members, variola and vaccinia virus. These viruses represent the infectious agent and the vaccine agent, respectively, of the disease known as smallpox. RFLP analyses has revealed extensive homology between the viruses in this genus with respect to the relative arrangement of restriction endonuclease recognition sequences within the linear genome implying their evolution from a common ancestor (60,117). A more detailed comparison using the available DNA sequence from these viruses has revealed that RPV

PAGE 27

8 is an extremely close relative of vaccinia virus, in particular the WR and Copenhagen strains. A comparison of about 21.0 kb of sequence from the lefthand of the viral genomes showed a degree-of-similarity of greater than ninety-nine percent throughout the entire region (24). These data strongly suggests that RPV arose from adaptation of VV(WR) or VV(Cop) or vice versa. Poxvirus Morphology Because of its extensive use as a vaccine, almost all of what is known about the structure and biology of the poxviruses has come from the study of vaccinia virus, which is considered the prototypic poxvirus. With regard to the structure of the virus, however, most if not all of the poxviruses have been observed to share the same physical features. The poxviruses have the distinction of being the largest of the animal viruses (128). The virus consists of a complex arrangement of lipid, protein, carbohydrate and DNA. The general appearance of the virus, when observed by electron microscopy (EM), is that of an oval or barrel-shaped particle measuring 200 x 300 nanometers in size (Figure 1A) (44,215). Within the center of thin-sectioned virions is seen a dumbbell-shaped structure which is referred to as the nucleoid of the virus (215). This membrane-bound structure contains the genomic DNA of the virus in the form of a supercoiled, nucleoprotein complex (186). Electron-

PAGE 28

TO 1/3 1/3

PAGE 29

u &. o > c w a o U c id £ o D M-l — 3 a, o IN E w 3 o F\-\ \ > \ s i s s i \ \ \ \ £ o O 01 U 3 z 1/ / X X X X X x^ \\\\\\\\ /xxxxxxxxx k\\\\\\\\\\\-^ m xxxxxxxxxx/x \\\\\\\\\\\\\ /xxxxxxxxxxxxxxf x x x x x X X / i L\\\\\\\ Fx XXX/// \ \ \ \ \ \ \ X/XXXXXXX \\\\\\\\" 01

PAGE 30

11 dense masses of unknown function are observed within the concavities of the core. These bi lateral structures have been named lateral bodies, and their removal with non-ionic detergent yields an oval core, suggesting it is the lateral bodies that give the core its characteristic dumbbell shape (56). The core structure and lateral bodies of the mature virus are surrounded by a second outer membrane that gives the virus its final characteristic shape. When the particles are lightly shadowed with gold particles, this outer membrane is observed to be covered with randomly oriented structures called surface tubules which are believed to be composed of a single 58 KDa polypeptide (189). The function of the surface tubules is unknown at this time. The viral structure presented thus far represents the traditional view of poxvirus morphology which has been based on standard EM techniques. However, recent work using cryoelectron microscopy (CEM) of vitrified samples suggests a different structure (54). The newer work reports that the dumbbell shaped core, the lateral bodies and the surface tubules are artifacts that result from the use of standard fixation techniques. In particles viewed by CEM, the virion is still seen as a barrel shaped particle; however, the core, instead of being dumbbell shaped, has the same barrel shape as the rest of the virion, and there are no lateral bodies (Figure IB). In addition, the ST structures are absent, and the core membrane is instead surrounded by a regularly striated palisade.

PAGE 31

12 The general particle described above has traditionally been called an intracellular naked virus (INV) particle; however, in recent literature it is referred to as an intracellular mature virus (IMV) particle. IMV particles are fully infectious and, in fact, the majority of work in the poxvirus field to date has been done using IMV purified from mechanically lysed cells. Although the exact number is still unclear, IMV particles contain between 52 and 100 proteins, as determined by two-dimensional gel electrophoresis (61,139). Twenty-one of these are known to reside in the core (139) including core structural proteins and all of the proteins necessary for early gene transcription. Biotinylation of purified IMV particles has shown the envelope to contain at least ten prominent polypeptides with molecular weights of 58 kDa, 39 Kda, 35 kDa, 32 kDa, 31 kDa, 24 kDa 23 kDa, 22 kDa and 14 kda (83). Activities have been associated with several of these proteins, some of which are described in the section on poxvirus biology. In addition to the IMV particle just described, two other forms of virus known as intracellular enveloped virus (IEV) and extracellular enveloped virus (EEV) (hereafter collectively referred to as the enveloped forms of the virus) are also produced during the infection (25). The three forms of virus differ in the total number of envelopes that surround the viral core and in the case of EEV, by its extracellular location. As described above, IMV particles have a total of two membranes. IEV particles, which are seen within the infected cell, are IMV that have become wrapped with two additional

PAGE 32

13 outer membranes resulting in a total of four membranes. The final form of virus, called EEV, is identical to IMV except that it has one additional outer membrane giving it a total of three membranes. In addition, unlike IMV and IEV, EEV is located outside of the unlysed cell. The additional membranes present in the enveloped virions, give these particles a buoyant density different than that of IMV allowing the separation of IMV from the enveloped forms by centrifugation of the virions through an equilibrium CsCl gradient. IMV has a density equal to 1.27 g/ ml whereas the enveloped forms have a density of 1.23 g/ ml (144). Extensive biochemical analysis has not been performed on the outermost membrane of IEV but it is presumably similar or identical to that of EEV, about which much is known. The earliest analysis of the protein composition of the EEV outer membrane showed that the envelope contained ten proteins unique to EEV, nine glycoproteins and one nonglycosylated protein (144,145). Five of these proteins had molecular weights of 210 kDa, 110, kDa 89 kDa, 42 kDa, 37 kDa, and five had molecular weights in the range of 20-23 kDa The 37 kda protein was the only non-glycosylated protein. A more recent analysis using monoclonal antibody precipitation showed that the smallest proteins are 23-28 kDa in size and, further, showed that these glycoproteins are actually differentially glycosylated forms of a single protein specie, 21 kDa in size (147). In addition, this work showed that the 220 kda protein is actually a complex of several differentially glycosylated

PAGE 33

14 forms of the 89 kDa protein. Most recently, a previously undescribed 43-50 kda non-glycosylated protein has also been reported as being located in the EEV envelope (143). The current model drawn from these data is that the EEV envelope contains four glycoproteins having molecular weights of 110 kDa, 89 kDa, 42 kda, and 21 kda and two non-glycosylated proteins of molecular weights 37 kDa and 43-50 kda. The Poxvirus Genome and its Organization While the specific size of the genome varies between the individual species of poxvirus, the general characteristics and structure are shared by all members of the family. All poxvirus genomes consist of a linear, double stranded DNA molecule ranging in size from 85 million daltons (approximately 129 kbp) for parapoxviruses to 185 million daltons (approximately 280 kbp) for fowlpox virus (128). Recently, the sequences of the entire genomes of the Copenhagen strain of vaccinia virus and the Bangladesh-1975 strain of variola major virus have been determined, and these viruses have genome sizes of 191,636 bp and 186,102 bp, respectively (79,120). Computer-assisted sequence analysis predicts that vaccinia encodes 263 potential proteins of 65 amino acids or greater (79). Similar analysis for variola reveals 187 potential proteins with eighty percent of these having significant homology to proteins encoded by vaccinia virus (120). While the

PAGE 34

1? size of the RPV genome has not yet been determined by sequence analysis, RFLP analysis predicts a size of 118 million daltons (approximately 180 kbp) (217). Comparison of the RPV Hz zzdlll restriction map with that of vaccinia (Copenhagen strain) shows that the relative arrangement of the fragments throughout the entire genome is identical between the two viruses (24,117,217). However, while the order and size of most of the fragments are identical in both genomes, the sizes of the terminal Hz ndlll B and C fragments differ between RPV and vaccinia, indicating a difference between the two viruses in the coding capacity of the terminal regions. This difference in coding capacity is likely to be responsible for differences in the host range and virulence of vaccinia and RPV since it is believed that many of the genes within these terminal regions influence these properties. The structure of the vaccinia virus genome is illustrated in Figure 2. While the majority of the linear genomic sequence is unique, the sequences found in the 12.0 kbp left and right terminal regions are nearly identical and are inverted with respect to each other (65,74,216). Within these inverted repetitions and adjacent to the very ends of the genome are located tandem repeats of 10 x 69 bp, 27 x 70 bp and 9 x 54 bp (65,79). Analysis of renaturation kinetics and sedimentation analysis revealed that the two genomic DNA strands are covalently cross-linked at their ends, preventing the two strands from separating (18). Detailed sequence analysis of the terminal DNA suggests that the terminal 104 bp exist as a single stranded loop, the sequence

PAGE 35

•5 9o S -co < <-3 60 £.2 (N bp'

PAGE 37

18 of which may be folded, using computer modeling, into an imperfect palindrome (14,65,79). The poxvirus genome may be divided into three regions based on the absolute requirement for the genes found within each of these regions. The central region contains genes essential for replication of the virus in tissue culture, the so-called 'house-keeping' genes such as the viral RNA and DNA polymerases. The essential nature of these genes is highlighted by the sequence conservation exhibited within this region by the various viruses, a feature that can easily be demonstrated by the near perfect alignment of the core region Hindlll restriction maps of the orthopoxviruses (117). The essential core region extends from the rightmost end of the HindUlY fragment until the rightmost end of the Hindlll A fragment and contains approximately one-hundred open reading frames (79). The two regions flanking either side of the core are referred to as the non-essential regions and, as suggested by the name, contain genes not essential for replication in tissue culture. While not absolutely essential in tissue culture, many of the genes located in the non-essential regions are involved in determining host range and in determining the virulence of the virus in animals. Most of the major sequence variation observed between the orthopoxviruses is found within the two non-essential regions which extend outward from the Hindlll F and A fragments until the ends of the genome, which are found in the H//idIII B and C fragments (79).

PAGE 38

19 Poxvirus Biology The Replication Cycle The replication cycle of the poxviruses is composed of a carefully coordinated series of steps culminating in the release of newly synthesized virions. The steps that make up this cycle, listed in order of occurrence, are attachment, entry, early protein synthesis, DNA replication, intermediate and late protein synthesis, viral morphogenesis and release of the virus from the infected cell. One complete cycle of replication is illustrated in Figures 3 and 4. Each process in the cycle has been studied in great detail, and a brief discussion of each will be given below. Attachment. The infectious process begins with adsorption of a virion to the surface of a host cell. Presumably, attachment of the virus to the cell is mediated through binding of one or more viral attachment proteins to molecules on the cell surface which serve as viral receptors. Since interactions of the virus at the cell surface probably help to determine host range, the broad host range of the orthopoxviruses makes it likely that they recognize more than one cellular receptor. As of this date, no cellular receptor molecules have been identified. The finding that vaccinia produces a growth factor that binds to the epidermal growth factor receptor (EGFR) led

PAGE 39

o> 3 oi u es H X > H > X o CL o> T3 C 3 O 01 a re Ih <-H >^ |-H to 0 s0> c o -M Mh jx Ih o> -si 5u LO c re 2 = 5 vC5 > o> fo id O U OJ O) X H T3 o< 'Lo X r" >"J2 S. Oi £ 3 PL H > LO .5 g > 0> > II C o> Ih C3 H > Io H^ u re c o 2 w a. CI LO c re S-H £ ~ re T3 o> *t3 o_ 01 g 8 -2 g S .3. S C o> Ih 01 Ph o> .a

PAGE 41

22 to the suggestion that this protein might serve as a receptor for the virus (162). Although initial work seemed to support this hypothesis, it has since been shown that adsorption of virus to the cell and subsequent replication are unaffected by either deletion of the viral growth factor or by the use of cells lacking the EGFR, demonstrating that these molecules do not serve as major receptors (35,59,98). The most recent work in this area has shown that a monoclonal antibody produced against a cellular surface protein blocks vaccinia virus infection; however, neither the target nor the mechanism of the anti-viral properties of this antibody have been identified (38). Future work on this antibody may result in the first purification of a vaccinia receptor. The external location of the various virally-encoded membrane proteins makes them logical candidates for virus attachment proteins. However, although the currently known envelope proteins of both IMV and EEV have been extensively studied, none of these proteins has been shown to directly mediate virus attachment. Of the IMV envelope proteins, the 35 kDa, 32 kDa and 14 kDa have been the most actively investigated for their ability to mediate virus attachment. The most recent data available for the 35 kDa protein indicate that monoclonal antibodies directed against this protein are able to neutralize virus infectivity by approximately seventy percent; however, further proof of this proteins ability to mediate virus attachment is needed (83). The 32 kDa protein has been reported to bind to cell surfaces in a

PAGE 42

23 manner that correlated with the ability of vaccinia to bind the same cells; however, mutants lacking this protein were unaffected in their ability to attach to and replicate in tissue culture cells (114,136,162). Interestingly, deletion of this gene results in decreased viral gene expression in a polarized cell culture system and also results in severe attenuation of the virus in mice, indicating that this protein is important for virulence (162). The 14 kDa protein is known to be involved in the fusion of the virus envelope with the cell plasma membrane and it is also involved in the fusion of the plasma membrane of infected cells (51,163,164). Even though this protein is intimately involved with early events in virus entry, antibodies to the 14 kDa protein only block virus fusion, not binding, indicating no role for 14 kDa in attachment (163). The above results suggest several possibilities: I) cell attachment is mediated by several proteins so that deletion of a single protein has little effect on attachment, II) standard cell culture systems may not accurately reflect the binding activity observed in animals, a possibility suggested by the results with the 32 kda protein, or III) the major attachment protein has yet to be found. Similar work has been conducted on the proteins of the EEV envelope again yielding no conclusive evidence of a major viral attachment molecule. A significant problem with attempting to measure the binding activity of EEV particles lacking these molecules is that deletion of the genes encoding these proteins often results in a failure to produce EEV particles. This EEV-minus

PAGE 43

24 phenotype has been demonstrated for the 43-50 kDa, 42 kDa, 37 kDa and 21 kDa proteins of the envelope, making it difficult to assess the effect of these molecules on virus attachment. Monoclonal antibodies directed against the 42 kDa protein have been show to inhibit spread of EEV in cell culture; however, no clear binding activity has been shown for this protein (57). In fact, in experiments using RPV 42 kDa null mutants, which are capable of producing EEV only in certain cell lines, EEV particles lacking the 42 Kda protein bound cells with kinetics identical to those observed for EEV from wild-type RPV (Stern, R.J., Unpublished data). The 21 kDa envelope protein has been shown to be involved in the release of newly synthesized EEV particles from the cell surface; however, the importance of this molecule in initial virus-cell interactions is unclear (22). Of all the EEV proteins, only the 89 kDa protein, known as the hemagglutinin (HA) protein, has been shown to have binding activity. The HA protein ha been shown to bind sialic acid containing molecules present on the surface of chicken red blood cells. However, deletion of this open reading frame (ORF) has no effect on the ability of virus to attach to cells in culture or on viral virulence in animals. In fact, the RPV strain commonly used for laboratory study is HAdue to a naturally occurring mutation resulting in a frameshift. Entry The attachment phase of the infectious cycle is followed by entry of the virus into the cell, the primary mechanism of which is still somewhat unclear. The earliest studies of this process using EM and analysis of the

PAGE 44

25 effect of metabolic inhibitors suggested that endocytosis of the virus particle was the most common route of entry (40,41). However, later studies indicated that entry occurs by direct fusion of the viral envelope with the cell surface (10,37). The most recent work confirms that poxviruses do indeed fuse with the host cell in a pH and temperature-independent manner (51,81,152). While phagocytosed virus particles were observed in these studies, particles contained in phagocytic vesicles represented only a small fraction of the total input virus, suggesting a minor role for this route of entry. This work also demonstrated that EEV particles fuse with cell membrane at twice the rate of IMV particles (51). In addition, fusion of EEV was unaffected by treatment of cells with compounds that effect endocytosis, conditions which were shown to severely reduce IMV fusion, suggesting that the two forms of virus may enter the cell by different methods (152). The significance and relevance of these observations to the disease process in the animal remain to be determined. Several viral proteins have been shown to be involved in virus-cell fusion and in cell-cell fusion. The best characterized of these is again the 14 kDa fusion protein found on the envelope of the IMV particle (163,164). Monoclonal antibodies to this protein have been show to block fusion, but not binding, of the virus with the cell (51,164). In addition, deletion analysis has shown that the 14 kDa protein mediates fusion of infected cells at acid pH, during the late stages of the infection (81,163,164). Two other proteins are also

PAGE 45

26 known to affect the cell-cell fusion phenotype, these being the 89 kDa EEV envelope HA protein and the SPI-3 protein. Although deletion of the HA ORF has no effect on virus-cell fusion, infection of cells with HA" virus results in one hundred percent fusion of the infected cell, suggesting that one of the functions of the HA protein during the infection is to prevent cell-cell fusion. The localization of the HA protein to the envelope of EEV suggests it may have a similar function in virus-cell fusion. The importance of this is unclear, however, since HA viruses are as infectious as HA+ viruses. Deletion of the SPI-3 ORF also results in fusion of infected cells, demonstrating some role for SPI-3 in fusion activity. The intracellular location of this protein suggests that it may somehow affects the 14 kDa protein; however there is no evidence for such an activity and its mechanism of action on fusion the fusion event is unknown. Uncoating Following entry of the virus into the cell, the virus undergoes a process known as uncoating (92,103,153). This process, which has been studied both biochemically and by EM, has been divided into two welldefined stages. In the first stage, there is a rapid degradation of some of the outer viral proteins and virtually all of the viral phospholipid resulting in structures that are called sub-viral cores. These cores appear similar to cores observed by EM, and the DNA at this stage remains resistant to degradation by DNase (103,104). The first stage of the uncoating process is independent of new protein synthesis and will occur in irradiated cells, suggesting it is a

PAGE 46

property intrinsic to the intact virion and /or pre-existing factors within the cell. Following a lag period of about one hour, the virus enters the second stage of uncoating during which the core is degraded and the DNA becomes susceptible to degradation by DNase. The lag period may be shortened by increasing the multiplicity of infection or abolished by pre-infection with any of the other orthopoxviruses. In addition, the second stage is dependent on viral protein synthesis (153). Although it was initially thought that a cellular protein induced uncoating, the recent development of an in vitro system to study this process resulted in the purification of a virally produced 23 kDa protein that is able to initiate uncoating of the viral cores (153). Early protein synthesis. The synthesis of viral proteins can be detected shortly after the virus has entered the cell, even before the second stage of uncoating has taken place. Production of the uncoating protein has been detected as soon as 20 minutes after infection of the cell (153). The production of early transcripts is directed by a DNA-dependent RNA polymerase complex present within the infecting virion (129). This complex, which contains two large and six small virus-encoded subunits has a molecular mass of around 500 kDa and contains all the activities necessary for randomly initiated RNA transcription from a single stranded DNA template (3,129). This polymerase activity is made specific for poxvirus early genes if one adds several early transcription factors to the polymerase complex. One such factor is the product of the H4 ORF, a protein known as Rap94 ( 94 kDa RNA polymerase

PAGE 47

28 associated protein) (3,105). The addition of a second factor, composed of two protein subunits and known as the vaccinia early transcription factor (VETF), to this complex allows the polymerase-Rap94 complex to utilize doublestranded DNA containing early promoters. VETF, a heterodimer made up of an 82 kDa protein and a 70 kDa protein, has been shown to possess ATPase activity. In addition, purified VETF has been shown to be a DNA binding protein with specificity for sequences within the promoters of virus early genes (219). The RNA-polymerase-Rap94-VETF complex contains all the factors necessary for transcription of early genes from the double-stranded viral genome. Termination of early gene transcription begins once the polymerase complex encounters the sequence UUUUUNU in the newly created RNA strand (166). Actual cessation of transcription appears to occur randomly approximately 50-70 bases downstream of the termination signal and is known to be dependent on viral polymerase, VETF and the virus encoded capping enzyme (166,180). The capping enzyme, which consists of two virus encoded proteins of 97 kDa and 33 kDa, is also known to be involved in the capping of the 5' end of the viral mRNA (118,129,134,135,181). Following termination, the mRNA 3' ends are polyadenylated by a complex (known as the poly-A polymerase) composed of two viral proteins with molecular weights of 55 kDa and 33 kda (75). The poly-A polymerase seems to have no specific recognition sequence since any transfected RNA is also

PAGE 48

29 polyadenylated (66). These capped and polyadenylated transcripts, which appear structurally identical to cellular eukaryotic transcripts, are now released from the viral core into the cytoplasm of the infected cell to direct the translation of viral polypeptides necessary for intermediate protein synthesis and DNA replication. Intermediate protein synthesis. Shortly after early protein synthesis and immediately after the onset of DNA replication, transcription from a second class of genes, known as the intermediate genes, begins (129,209). This class is defined by genes that are dependent on trans-acting factors produced in the cell shortly after infection (early proteins) and on DNA replication but not on de novo protein synthesis following DNA replication. This is in contrast to late genes which are dependent on both DNA replication and on continued intermediate protein synthesis following DNA replication. In addition, unlike the late genes, plasmid-borne intermediate genes are expressed prior to DNA replication in vaccinia virus infected cells that have been treated with inhibitors of DNA replication, demonstrating that all of the factors necessary for intermediate gene transcription are present prior to DNA replication. Currently, five intermediate genes (AIL, A2L, GK1, 18 and 13) have been identified, three of these (AIL, A2L, GK1) being late transcription factors (108,209). No consensus promoter sequence for intermediate genes has yet been identified; however, the factors required for intermediate transcription are known (207,208). These are the viral RNA polymerase, the

PAGE 49

30 capping enzyme, which in this capacity has a role additional to its role in capping mRNA (207), and two proteins known as virus intermediate transcription factor (VITF)-l and V1TF-2 (168). VITF-1 has been shown to be the 35 kDa rpo30 protein which is coded for by the vaccinia E4L ORF (4,168). Rpo30, which has homology to eukaryotic transcription elongation factor SII (TFIIS), has also been shown to be a component of the early RNA polymerase complex suggesting a dual role in transcription for this protein (4). VITF-2 activity, which co-purifies with a 68 kDa protein, has been reported to be cellularly encoded since its activity can be purified from uninfected cells (169). In this regard, it is interesting that enucleation experiments indicated a role for the cell nucleus at a late stage in infection (95,96). It has also been shown that in RK-13 cells, which lack VITF-2 activity, mutants in the K1L gene of vaccinia are blocked in the intermediate transcription stage of the replication cycle (169). In vitro transcription assays detected VITF-2 activity in RK-13 cells stably expressing K1L indicating that the K1L protein can functionally substitute for VITF-2. The function of K1L, which has been shown to contain several ankyrin repeats, is currently unknown (169). Further studies should reveal the nature of the link between K1L and VrTF-2. The intermediate transcripts have been characterized and are thought to initiate within a sequence resembling a late promoter and to have heterogeneous 3' ends (12). It was initially been reported that these mRNAs lacked a poly (A) head (209); however, recent studies have reported poly (A)

PAGE 50

31 heads of up to 30 bases on these mRNAs (12). The mechanism of termination of intermediate transcription is unknown. Late protein synthesis. Following DNA replication, which begins approximately 2 hours post-infection, and the accumulation of late transcription factors produced from intermediate genes, late gene transcription begins. The late genes encode the structural proteins of the virus particle along with the proteins necessary for the proper assembly of the virus particle (128). In addition, the proteins that comprise the early-gene transcription complex found within the virion are produced late in infection. Transcription of late genes requires a complex of polymerase and several late transcription factors produced from intermediate genes. It is likely the polymerase used for early gene transcription is also used for late gene transcription since conditionally lethal mutations in the polymerase subunits affect late gene transcription (93). In addition to the polymerase complex, several late transcription factors are known to be required. These include the above mentioned 30 kDa, 26 kDa and 17 kDa proteins produced from intermediate genes (108,209) and a recently discovered factor known as P3 (112). Comparison of the sequence of several late genes has revealed a consensus late promoter sequence consisting of two elements (66). The first element, consisting of the short sequence TAAAT followed by an A or a G, is found immediately upstream of the translation initiating ATG. In fact, in

PAGE 51

32 most cases, the initiating ATG forms the 3' end of the consensus sequence (TAAATG). This sequence is conserved in nearly all late genes, suggesting some, as of yet unknown, important function. Any mutations within this sequence abolish late transcription of the corresponding gene. The second element is found 16 to 20 bp upstream of the initiating ATG codon. This element shows no strong conservation of sequence but usually consists of a stretch of five to eight A or T residues that are essential for late gene transcription. The mRNA transcribed from late genes has several unusual properties. The mRNA contains 35-50 non-templated polyadenylate residues at its 5' end that are thought to result from slippage of the RNA polymerase within the TAAATG sequence (66). This model is based on a similar model in bacteriophage T4 for which there is considerable experimental evidence (66). A clear function for the poly-A head is unknown however it has been speculated that it may act as a ribosome binding site and increase the efficiency of vaccinia mRNA translation (66). In addition to the 5' poly-A head, late mRNAs also have heterogenous 3' ends indicating that late transcription does not terminate uniformly as does early mRNA. To date, no termination signals have been recognized for late mRNA transcription and, in fact, late transcription can continue uninterrupted through several open reading frames generating transcripts that are several kilobases in length. Since both strands of DNA are

PAGE 52

33 transcribed, this results in the generation of a significant amount of doublestranded RNA at late times during infection. Replication of the poxvirus genome. Based on measurements of the amount of accumulated DNA that hybridizes to known quantities of filterbound DNA, it has been determined that replication of genomic DNA begins around 2 hours post-infection, concurrent with intermediate gene expression, and proceeds for at least ten hours (200). This replication occurs in the cytoplasm of the cell in electron-dense areas referred to as viral factories or virosomes. This cytoplasmic location suggests that, as is the case with the transcriptional apparatus, the virus must provide most if not all of the enzymes necessary for genomic replication. Experiments using enucleated cells indicate that, indeed, the virus provides all the necessary replication machinery (154). Although many of the details surrounding poxvirus DNA replication remain unclear, the available data has resulted in a well accepted working model which is illustrated in Figure 4 (132,200). This model is based on the known structure of the genomic termini as well as on analysis of genome intermediates observed during DNA replication (132). As previously described, the termini of the poxvirus genome consist of a single-stranded DNA containing an imperfect palindromic sequence. When the two halves of this sequence are maximally aligned, several mismatched bases are seen. It is within this hairpin region that replication is thought to initiate.

PAGE 53

Figure 4. Model for replication of the poxvirus genome. Self-complimentary sequences within the inverted terminal repeats are depicted in upper and lower case letters. (A, a', B, b', C, c'). The initial nicking site is denoted by the "X". Newly synthesized DNA is depicted by dotted lines. (Adapted from Traktman (200)).

PAGE 54

ABC C b' c'> C B A ) c' b' a' a'b'c'ABC CB A A B C a' b' c' c' b' a' A B C a'' b'" c' '*:']> vi -n b a n > C B A c ABC c' b' a' ABC ) + ( c' b' a' a' b'c" CB A a' b'c' CB A J

PAGE 55

36 Replication begins with a single-stranded nick being made within the hairpin structure. Nicking likely occurs at either or both ends of the genome, but for ease of discussion, it will be assumed here that only one nick per genome is made. The presence of this nick has been identified by the nicked DNA's altered mobility and sedimentation properties (157). The necessity of sequence specificity is argued against by the fact that any DNA transfected into poxvirus infected cells is replicated (46). This initiating event results in a free 3'-hydroxy terminus which can now be used by the polymerase as a primer. Elongation of this strand displaces the complementary strand and uses it as a template. This model is supported by data showing that radiolabeled nucleotides are first incorporated into sequences from the end of the virus genome (156). The resulting molecule is a longer-than-genome-length hairpin with palindromes on either end. The newly synthesized end, being a palindrome, can fold back on itself, and continued extension of this strand results in copying of the entire genome. The majority of evidence to date indicates that elongation occurs by unidirectional, leading strand synthesis (200). The product of this event is a molecule in which two complete genomes are joined by their right-hand ends in a tail-tail arrangement. Continued elongation yields a concatemeric molecule with multiple genomes organized in tail-to-tail and head-to-head arrangements. Concatemers consisting of up to four genomes have been shown to exist (45,132). These concatemeric molecules are then resolved into unit length genomes through

PAGE 56

a process known as telomere resolution. This process requires a pair of sequences known as the telomere resolution targets, which are present in the junction fragments, as well as an activity referred to as a resolvase (192). The importance of the telomere resolution targets has been demonstrated by experiments showing that plasmids containing this sequence within cloned junction fragments are resolved into linear molecules with plasmid DNA in the center and viral hairpin at the ends (47,48). Several of the proteins involved in replication of the poxvirus genome have been purified including a DNA dependent DNA polymerase, a topoisomerase, a DNA ligase, a thymidine kinase and several DNA binding proteins of unknown function (200). The polymerase is found as a monomer of 110 kda and possesses both polymerase and 3'-5' exonuclease activity (66,200). Although the polymerase shares homology with several eukaryotic polymerases, its activity cannot be replaced by the cellular polymerase, and it has been shown to be essential for virus replication. Virus morphogenesis. Although vaccinia virus was one of the first objects to be viewed by EM, the molecular mechanisms responsible for the assembly and release of the intact virion remain obscure. Most of what is currently known about viral morphogenesis is largely descriptive information based on the effects of mutations and inhibitors of virus replication on virus morphogenesis, as determined by EM. Nevertheless, a model for the pathway of virion assembly has been constructed.

PAGE 57

38 Assembly of the mature virus begins in granular, electron-dense structures called viral factories that can be observed in the cytoplasm starting about 4 hours post-infection (185). Staining of thin sections of infected cells with DNA-specific indicators reveals large quantities of DNA present in what has been called the "viroplasm" within these factories. The first indication of assembling virus is the appearance within the viral factory of crescent-shaped membranous structures that will eventually form the envelope of the IMV particle (42). It has long been thought that these membranes are not obtained from a cellular organelle but instead are synthesized de novo by the virus since they have never been observed to be continuous with cellular organelles. This hypothesis seemed to be supported by chemical analysis of these membranes which indicated that their phospholipid composition was vastly different than that of cellular membranes (188). However, recent studies using CEM showed that the envelope membrane is, in fact, continuous with cellular, membranous structures (185). Through the use of labeled antibodies specific for proteins which localize to several subcellular compartments, it has been determined that the envelope membrane is derived from the intermediate compartment between the ER and the golgi stacks. In addition, these studies showed that the crescent shaped structures and IMV envelope are actually composed of two closely apposed membranes. These data support the hypothesis that the envelope results from wrapping of the core with cisternae derived from a golgi-ER intermediate compartment.

PAGE 58

39 The crescents observed in the viral factories are covered with a layer of spicules formed from a single polypeptide of 65 kDa of unknown function (62). The crescents assemble around viroplasm into spherical, immature particles containing a dense granular center. Interestingly, although virions are formed late in infection, empty immature particles are formed in the presence of inhibitors of DNA replication, indicating that viral proteins required for particle formation are produced early. Following production of the immature particle, the viroplasm contained within condenses by unknown mechanisms into a DNA-containing core structure, resulting in an infectious intracellular mature virus particle. During this transformation, the spicules are replaced by the tubule structures which are composed of a 58 kDa protein (189). As previously described, while IMV particles are fully infectious, a small percentage become wrapped with additional envelopes and are transported out of the cell resulting in extracellular enveloped virus. The origin of the EEV envelope has been a subject of some controversy. Although both golgi and early-endosomal network membranes have been postulated to be the source, the most recent study using antibodies to proteins specific for intracellular compartments indicates that the envelope is derived from the trans-golgi network (TGN) (90,176,199). The cytoplasmic IMV particle is wrapped by cisternae derived from the TGN resulting in the addition of two membranes to the particle, yielding an IEV particle. This wrapping event has

PAGE 59

been shown to be dependent upon the presence of the 42 kDa, 37 kDa, 32 kDa and the 21 kDa envelope proteins as well as on the 14 kDa fusion protein located in the IMV envelope (20,55,58,119,143,165,218). Deletion of any of these genes has no effect on IMV production but results in the inability of such mutants to produce the enveloped forms of virus (IEV and EEV). The requirement for these enveloped proteins in the wrapping process is currently not understood but it is thought that some of them may mediate wrapping by interacting with IMV surface proteins. Interestingly, the compound Ni-isonicotinoyl-N2-3-methyl-4-chlorobenzoylhydrazine (IMCBH) specifically inhibits the wrapping of the IMV particle (89,148). While the mechanism of this inhibition is unknown, mutations in the 37 kDa envelope protein have been shown to confer on the virus resistance to the inhibitory effects of IMCBH (177). In addition, inhibition of enveloped virus formation may also be achieved by treatment of infected cells with inhibitors of glycosylation or with brefeldin A, a fungal metabolite that has been shown to inhibit the transport of proteins to post-ER compartments (150,203). Once the IEV particle has been formed, it moves to the cytoplasmic membrane where the outermost virus membrane fuses with the cytoplasmic membrane releasing the virion into the cytoplasm. The cellular apparatus involved in transport of the particle remain uncharacterized; however treatment of infected cells with cytochalasin D blocks release but not

PAGE 60

formation of enveloped virus suggesting, a role for microfilaments in the release of the enveloped virus particle (149). The amount of EEV produced varies from less than one percent to around twenty-five percent of the total virus yield, depending on the virus and cell line used (145). This variation suggests some role for as of yet unidentified cell components in enveloped virus release. The 21 kda envelope protein is the only EEV protein that has been implicated in modulating the amount of virus released from the infected cell. A single amino acid change in the lectin homology domain of this protein has been shown to alter the amount of virus released by a factor of ten however, the cellular target of this protein is unknown (22). Some Musings on the Significance of Enveloped Virus Since the IMV particle produced in the infected cell is completely selfsufficient in its ability to initiate a complete round of infection, it may seem wasteful for the virus to devote both energy and coding capacity to the production of enveloped virus. However, since many of the poxviruses are known to produce EEV, there is likely some survival advantage in this ability beyond the interest it generates in scientists who then continue to propagate the virus. One might think that enveloped virus production is an artifact of cell culture, however, enveloped virions have been observed in thin sections of infected animal tissue (113,151). Since the virus used in this study, RPV, produces large quantities of EEV, it is worth spending some time reviewing

PAGE 61

42 current thought on the significance of enveloped virus in cell culture and in animal infections. The enveloped forms of the virus, IEV and EEV, are believed to have two distinct roles in dissemination of virus from the initial foci of infection in cell culture. A simple 'comet' assay has historically been used to measure the ability of virus to spread from the initial focus of infection (9). In this assay, cells infected under conditions that yield well-isolated plaques are incubated under liquid for several days. Virus released from a focus of infection spreads, relatively unidirectionally, through the medium producing comet-shaped areas of cell destruction, with the head of the comet formed by the primary plaque. There exists a correlation between the amount of EEV produced by a virus and its ability to form comets, suggesting EEV is responsible for long-distance spread of virus (146). This has been confirmed experimentally since neutralizing antibodies to EEV, but not IMV, prevent comet formation without affecting primary plaque formation (9). The observation that viruses can produce vastly differing amounts of EEV yet still produce similarly-sized primary plaques led to the hypothesis that another form of the virus, IMV or IEV, was responsible for cell-to-cell spread and subsequent plaque formation. Mutants that produce IMV but which are unable to form enveloped virus fail to plaque suggesting that IMV is not sufficient for efficient cell-to-cell spread (20,165). This has led to the current model in which IEV, which can be observed attached to the outer

PAGE 62

43 surface of the infected cell (21), is thought to be responsible for cell-to-cell spread and plaque formation, although there is no direct proof of this as of yet. The importance of EEV in the dissemination of virus throughout the infected animal has been suspected for a long time and is derived from several observations (9,27,146). Although the majority of virus produced in tissue culture is IMV, it has been shown that antiserum generated against live rabbitpox virus can neutralize both IMV and EEV in vitro and is able to passively protect rabbits against lethal infection with rabbitpox virus (27). By contrast, antiserum to inactivated virus although neutralizing for IMV in vitro is non-neutralizing for EEV and is unable to passively protect rabbits from lethal infection. In addition, the vast majority of virus purified from the blood of infected animals is of the enveloped type. Finally, mutants defective in their ability to produce enveloped virus (but not IMV) have been shown to be attenuated in animals (58,143) (and this dissertation). Taken together, these results provide strong evidence that it is the EEV particle that is important for successful spread and the manifestation of full virulence in the infected animal. Poxvirus Pathogenesis The severity of smallpox in man resulted in a search for useful animal models with which to study poxviral pathogenesis. The study of two such

PAGE 63

44 models, rabbitpox virus in rabbits and mouse-pox (ectromelia) virus in mice, has led to a greater understanding of poxviral infections with regard to the disease progression and pathology of these agents (8,19,85,86). These studies produced nearly identical results suggesting common themes in the ability of poxviruses to cause disease. Since I have restricted my work to the use of rabbitpox virus (RPV), I will concentrate primarily on the pathogenesis of this virus. The history of RPV is an interesting story of epidemiology and the emergence of new infectious agents. The virus was first reported in 1932 when its spontaneous appearance resulted in a smallpox-like epidemic within a breeding colony of rabbits housed at the Rockefeller Institute for Medical Research (84). Within two weeks one-hundred percent of the colony had been infected and forty-six percent had died. Subsequent studies revealed the etiological agent to be a poxvirus that was given the name rabbitpox virus (84). While the exact origin of this virulent virus will likely never be known, records from the time indicate that vaccine virus and a neurovirulent strain of vaccinia were being used to infect animals in common areas of the breeding facility (84). Epidemiological studies from the time strongly suggest that vaccinia virus was the progenitor of RPV, a hypothesis supported by sequence comparisons showing ninety-nine percent nucleotide homology between the two viruses (24). The outbreak of such a severe disease within a well regulated colony

PAGE 64

resulted in excellent documentation of the symptoms and pathology of RPV infection (85-87). Clinically, the disease resembles smallpox. Onset was marked by malaise and fever in conjunction with enlargement of the popliteal and inguinal lymph nodes. This was followed in two to three days by the appearance of papules on all parts of the animal which eventually became umbilicated and necrotic. These papules were often associated with extensive edema. Upper respiratory infection commonly occurred and was accompanied by bronchopneumonia resulting in labored respiration and mouth breathing. Many of the animals also developed conjunctivitis and released mucopurulent discharges, often stained with blood, from the nasal passages. Death from infection resulted within eight to ten days in nearly fifty-percent of the cases. A detailed analysis of the progress and course of a poxvirus infection has also been described for ectromelia in mice, resulting in a now classic model that has since been shown to be valid for RPV as well (16,26,214). Spread of the virus through the body is a stepwise invasion involving (1) the primary site of infection (2) the regional lymph nodes (3) the spleen and (4) the lungs, skin and other remote organs. The route of infection affects mostly the LD50 and has virtually no effect on the progress of infection (16). Experimentally, initiation of an ectromelia infection is usually accomplished by injection (intradermal or footpad) or inhalation of aerosolized virus, both of which have been demonstrated as feasible natural modes of transmission

PAGE 65

46 (16,214). Most of the available pathological examinations have been performed on animals infected by the respiratory route (16,85,86). Following 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 only found at the inoculation site, however, within 36 hours, it can be recovered from the regional lymph nodes, having likely been delivered there by infected lymphocytes (16). Replication within the lymph tissue is followed by a minor viremia immediately after which, on day two post-infection, virus is detected in the spleen. In animals inoculated intradermally, however, virus can be detected simultaneously in the lymph nodes and the spleen, casting suspicion on the role of the lymph node as an intermediate in the spread of the virus (16). The virus rapidly achieves high titers within the spleen resulting in a second and massive viremia around day three post infection. It should be noted that virus purified from blood is associated with the lymphocytic fraction suggesting that during viremia, the virus is not free but is instead cellularly associated (16,214). Concurrent with this secondary viremia and signaling the end of the incubation period is the onset of fever which likely arises due to the inability of the spleen to clear the growing load of virus and the resulting systemic infection (16). A nasal, mucopurulent discharge also appears around day three and increases in severity until around day five post infection when the animals become infectious (16,86,214). The increased levels of blood-borne virus result in

PAGE 66

delivery of this virus to the remaining organs of the body, and papules begin to appear on the skin approximately five to six days post infection. Death from the infection usually occurs between days six and ten post infection and is frequently preceded by a sharp drop in body temperature. If the animal survives, the papules scab over and clear around days ten to twelve post infection, usually leaving behind scarring. Microscopic analysis of the papules indicates that they all originate as vascular or perivascular processes, which is not surprising in view of the viremic spread of the virus through the body(85). Early in the development of the lesion, the epithelial layer becomes swollen, and there is proliferation of the cells of the arterial wall. This hyperplasia is probably caused by the viral growth factor, as has been demonstrated on the chorioallantoic membrane of embryonated chicken eggs (33). Cells, polymorphonuclear leukocytes, monocytes, and some red blood cells, become grouped in a loose collar formation around the affected blood vessel. As the lesion progresses, it assumes a focal form; however, in a small percentage of cases it becomes diffuse, spreading laterally through the tissues. The diffuse form of the lesion is more hemorrhagic than the focal form and results in the sloughing of large areas of tissue. Within the center of the developing nodule, the cells become necrotic and disintegrate so that cell outlines can no longer be distinguished. The histological picture of the nodule is now a central area of necrosis surrounded by a small area of polymorphs which itself is surrounded by a

PAGE 67

48 zone of mononuclear cells. The mononuclear cell zone and the surrounding tissues are edematous and occasionally hemorrhagic. Extension of the lesion into deeper layers of the skin causes impairment of the blood supply resulting in anemic necrosis that extends to the surface of the animal. While the obvious pneumonia and labored breathing observed in the late stages of infection would lead one to think death results from pneumonia, the results from studies conducted on the cause of death suggest otherwise (26). Physiological studies of infected animals consistently showed a severe drop in blood pressure and the occurrence of a shock-like state immediately prior to death. This was coupled with a dramatic rise in serum potassium concentrations to levels that have been shown experimentally to be lethal. From these data it has been concluded that death results from potassium intoxication. Nonspecific Host Responses to Infection The host response to the infection described above consists first of a non-specific immune response followed by the development of a specific immune response. Since the majority of the work in this dissertation is concerned with the very early response to infection, this section will be limited to a discussion of the non-specific response (16,85,86,214). Although no specific immune response can be detected until around day eight postinfection, it is known that the non-specific components of the immune

PAGE 68

49 system are activated immediately following infection. This non-specific response is composed of pyrexia, interferon production, apoptosis of infected cells, complement activation, as well as recruitment of leukocytes and NK cells to the site of infection. A brief discussion of each of these is given below. Interferon. One of the earliest responses to viral infection is the production of interferon at the site of infection. The interferons (IFN's) are a family of proteins that is divided into two groups, type I and type II interferons (174). The type I interferons are induced in response to viral infection and include IFN-alpha and IFN-beta which are produced by leukocytes and fibroblasts, respectively. Type II interferon, called IFN-gamma is produced by antigenic stimulation of T lymphocytes and NK cells. Production of interferon in poxvirus-infected cells is believed to be due mainly to the production of dsRNA late in infection (see Late Protein Synthesis) and results in the establishment of an antiviral state in neighboring cells (77,115). The effects of interferon on poxvirus replication have been studied both in cell-culture and in the infected animal (174). In cell culture, treatment of cells with IFN prior to infection may cause abortion of the poxvirus infection; however, the efficiency of the antiviral state is dependent on the virus used, the cell type and the concentration and the type of IFN used (13). The nature of the antiviral state has been well studied and has been shown to result from a block at the level of mRNA translation in infected cells,

PAGE 69

50 effectively blocking even the earliest steps in virus replication such as particle uncoating (72,123). The mechanisms by which interferon blocks protein production are proposed to result from two pathways dependent upon the action of three interferon-induced enzymes: a 2',5'-oligoisoadenylate synthetase, an RNase, and a 67-kDa protein kinase (174). In the first pathway, the 2',5'-oligoisoadenylate synthetase catalyzes the synthesis, from ATP, of a family of oligonucleotides with the general structure ppp(A2'p5')nA (n^2) abbreviated 2,5A. This molecule then activates an endonuclease, known as RNase L, which is present in a latent state within the cell. Once activated, RNase L cleaves both cellular and viral single-stranded RNA. In the second pathway, a 67-kDa protein known as PI / eIF-2, the production of which is induced by the IFN's (30,116), is activated by either double or single-stranded RNA(116). Once activated, PI/ eIF-2 catalyzes the phosphorylation of the alpha subunit of protein synthesis initiation factor eIF-2 (175). The end result of this process is inhibition of protein synthesis at the initiation step of translation. Systemically, IFN-gamma is known to increase MHC-I production and to activate natural-killer (NK) cells (see below) (1,167). The study of the effects of IFN-gamma on poxvirus replication in the animal has primarily involved correlating the levels of IFN's with the severity of the disease, or by measuring the effect of injecting either IFN or inhibitors of IFN (36). In poxvirus infected animals, increased levels of all IFN's can be detected in the

PAGE 70

51 serum as well as in cells taken from the peritoneal cavity or from the site of inoculation suggesting that IFNs are important in combating the infection (71,78,193). In support of this, it was shown that treatment of infected mice with anti-IFN antisera resulted in a significant increase in the amount of virus recovered from these animals (101). In addition, administration of alpha or beta IFN prior to infection has been shown to decrease the severity of the disease and the mortality rate (36). Finally, while athymic nude mice were able to clear an infection produced by vaccinia recombinants expressing IL-2, injection of similarly infected mice with anti-IFN antisera resulted in one-hundred percent mortality (107). It should be kept in mind that the result of such studies are almost always dependent on the host and virus species and on the routes of inoculation. Pyrexia. Pyrexia, or fever, which can be induced by interleuken-1 (IL-1) or by tumor necrosis factor (TNF), is a general systemic response to the infection of a host by a variety of organisms (50). The importance of increases in temperature on poxvirus infections has been studied to some degree in both cell culture and in the animal. The concept known as "ceiling temperature" was introduced in the 1960's when it was noticed that above certain temperatures, the orthopoxviruses would not form pocks on the CAM (17,68). At that time, it was also determined that ceiling temperatures could be measured in cultured cells. The ceiling temperature was observed to vary depending on the virus and thus became useful as a method to distinguish

PAGE 71

between various species of orthopoxvirus. In infected animals, changes in body temperature have been shown to effect the severity of myxomatosis. In addition, the susceptibility of mice to infection with ectromelia was increased 100-fold by an 18C drop in their ambient temperature (160). Together these results indicate that alterations in body temperature can effect viral disease and, further, suggest that such alterations may play a role in limiting poxvirus replication in the animal. Complement. The complement system is an integral part of the nonspecific host response to infection and of the inflammatory process (80,91,133). The system is composed of twenty plasma proteins, normally found in an inactive state, as well as many cell-surface receptors. These receptors have specificity for components of the complement pathway and include several regulatory proteins which are thought to prevent accidental activation of the system on host cell surfaces. Activation of the complement system results in a cascading series of enzymatic activations with the final result being the formation of a membrane attack complex (MAC) that forms pores in biological membranes. The physiological consequences of complement activation are opsonization of infecting organisms, cellular activation and lysis of target cells by MAC induced osmotic swelling. The complement cascade can be activated by different mechanisms involving two separate pathways, the classical pathway and the alternative pathway. Although these pathways differ in their initiation and their early

PAGE 72

53 stages, they both result in the formation of an enzyme called C5-convertase which is the key enzyme in activating the formation of the MAC. The classical pathway is dependent on target-bound antibody for initiation. The antigen-antibody complex is bound by complement component CI, resulting in self-cleavage and activation of CI Activated CI catalyzes the assembly of C3 convertase from components C2 and C4. This fluid-phase convertase (C4b,2a), upon binding to a surface, then cleaves component C3, a fragment of which remains associated with the C3 convertase altering it to a C5 convertase (C4b,2a,3b). The cleavage of C5 by the C5 convertase leads to the assembly of the MAC from components C6, C7, C8 and C9. Insertion of the MAC into the target cell membrane results in osmotic swelling and lysis of the cell. MAC insertion into the membrane of enveloped virions can also allow access of degradative enzymes to the interior of the virion (91). The alternative pathway is initiated by the reaction of component C3 with water resulting in C3(H20). This molecule can bind complement factor B, which is then activated by factor D to give C3(H20),Bb a fluid phase enzyme that acts as a C3 convertase. This newly formed convertase produces a metastable C3b that covalently attaches to nearby surfaces and binds more factor B. Activation of this surface-bound C3bB by factor D yields a surface bound C3b convertase (C3b,Bb) that continues to convert C3 to C3b resulting in a self-amplifying, activation loop. The addition of another C3b molecule to the bound C3 convertase produces C3b,Bb,C3b which acts as a C5 convertase,

PAGE 73

54 initiating assembly of the MAC as described above. Because the action of the MAC is non-specific in terms of the membranes it will permeabilize, and because activation of the complement system results in numerous systemic effects, activation of this system is tightly regulated at several points by both serum and cell-surface molecules (133,210). Activation of the classical pathway is regulated by inhibition of CI cleavage and by inhibition of C3 convertase formation. The latter inhibition is mediated by the C4b binding protein (C4b BP) and factor I in the plasma as well as by decay accelerating factor (DAF) and the C3b receptor (CR1) on the cell-surface. Factor I and C4b binding protein, together, catabolize C4b and promote the dissociation of the C3 convertase (C4b2b). DAF and CR1 inhibit C3 convertase formation and promote the degradation of C4b and C3b by factor I. Similar control is exerted over the alternative pathway of activation by the serum factors properdin (P), factor H (H), factor I (I) and the membrane factors CR1 and membrane cofactor protein (MCP) (133,210). The C3 convertase (C3bBb) formed by the alternative pathway is unstable and rapidly dissociates unless it is bound by P which stabilizes the complex and which therefore acts as a positive regulator. Factor H is homologous with the C4b binding protein and promotes the dissociation of C3b and Bb and also promotes the degradation of C3b by factor I, providing negative regulation. Whether C3b acts as binding site for factor B, resulting in the amplification

PAGE 74

loop, or whether it is degraded by factor I resulting in inhibition, is determined by the nature of the surface to which C3b is bound. Host membranes contain CR1 and MCP that promote binding of factor H to C3b resulting in inhibition of the cascade. Non-host surfaces on the other hand lack these regulatory molecules and act as protected sites for C3b allowing activation of the pathway. The protective nature of non-host surfaces also seems to involve the carbohydrate composition of the surface in some unknown manner. With the cloning and sequencing of the genes coding for the complement proteins it has become clear that many of these proteins can be assigned to superfamilies based on structural homologies (210). It is thought that many of these genes arose from a common progenitor by gene duplication and exon shuffling and that these duplicated genes then evolved in parallel maintaining similar structures and often similar functions. This conservation of structure and function is seen in the SCR superfamily, the members of which contain repeats of a conserved sequence known as a short conserved repeat (SCR) element (94). This sequence, which is approximately 60 amino acids in length, is not strictly conserved but instead consists of nonconserved sequences flanking several conserved amino acids within the element. Many of the proteins that regulate complement activation, such as factor H, C4b binding protein, DAF, and CR1 are members of the SCR superfamily. While the actual tertiary structure of these proteins varies

PAGE 75

56 widely, they all contain a variable number of SCR elements and they all share similar functions, namely preventing the stable formation of the C3 convertase. It is tempting to speculate that the SCR domain mediates binding to the convertase components C4a, C2a, C3b and Bb, however, two other members of this family, the IL-2 receptor and factor Xlllb, a protein of the coagulation pathway, show no such activity. Since all of the members of this family do show binding activity, it is likely that the domain provides a framework for a general binding structure, and that the specificity of binding is determined by the variable residues located between the strictly conserved core amino acids. While complement activation results in MAC formation and lysis of the target cell, it also has several other consequences, namely opsonization of foreign particles and activation of the cellular components of the immune system. Opsonization utilizing complement components is accomplished by covalent attachment of factor I-mediated-degradation products of component C3 to the foreign particle (80). This bound C3 is recognized by C3 receptors on leukocytes enhancing phagocytosis by these cells. In addition to mediating MAC formation and opsonization, the proteins of the complement cascade are also the source of anaphylatoxins, which are small biologically active peptides that affect the blood vessels, smooth muscle and blood leukocytes (80). In vitro experiments have shown that enzymatic treatment of C3, C4 and C5 produces the anaphylatoxins C3a,

PAGE 76

57 C4a and C5a which, when injected into animals, result in smooth muscle contraction and increased vascular permeability. In addition to these actions, C5a has a dramatic effect on monocytes /macrophages and neutrophils, that is mediated by binding to specific cellular receptors. Upon binding of C5a to its receptor, the cells display increased oxidative metabolism, increased adherence to surfaces, polarization and directed migration towards the C5a source, and in the case of monocytes/ macrophages, generation of IL-1, TNFalpha and IL-6. The involvement of the complement pathways in poxvirus infectivity and pathogenesis has been studied both in vitro and in vivo. Early in vitro studies showed that antibody-mediated neutralization was greatly enhanced by a heat-labile serum component (complement) (52). These studies have been extended, and it has since been reported that complement can mediate direct lysis of virions in addition to lysis of infected cells (28,29,36). While it is thought that opsonization of virus particles is also important, this has yet to be demonstrated. Activation of the alternative pathway in the absence of antibody has been detected in fowlpox virus (FPV)-infected chick embryo cells and vaccinia-infected tumor cell lines (36,137). This activation correlated with increased killing of infected cells and decreased yields of virus progeny suggesting that activation of this pathway may play a role in combating infection. In vivo, the role of complement has primarily been studied by

PAGE 77

58 measuring complement levels in infected animals and by examining the result of virus infections in complement-depleted animals (29,138). Measurement of the levels of the products of complement activation has shown that the complement system is active in FPV infected animals (106). Injection of cobra venom factor (CFV) into animals results in a temporary depletion of circulating complement levels. CFV treatment of chickens prior to infection resulted in increased virus levels in primary lesions and a decreased inflammatory response. The mortality rate for these animals was one-hundred percent whereas it was zero percent for untreated animals. Similar results have been obtained in chicken embryos. Together, the results to date suggest that complement can enhance virus neutralization in vitro and may play a role in resistance to viral infection. NK Cells. Within the first 2-3 days following infection, non-specific cytolytic lymphocyte (CTL) activity, attributable to natural killer (NK) cells, can be demonstrated in the infected animal (2,36). NK cells are a bone marrow-derived subset of lymphocytes found mainly in the blood and in the spleen. The cells appear as large lymphocytes containing large cytoplasmic granules and are often referred to as large granular lymphocytes. The cytolytic activity displayed by NK cells is similar to that of the major histocompatibility (MHC)-restricted CTL response in its mechanism; however, NK activity is not restricted by MHC molecules. Although this activity is not specific, it is also not random in that NK cells have the ability to kill tumor, but not normal,

PAGE 78

59 cells and will also kill cells infected by some viruses but not by others (2,212). Killing of target cells by NK cells involves two steps, granular excytosis and secretion of a cell toxin (2). In the first mechanism, NK cells secrete the contents of some of their cytoplasmic granules in areas of contact with target cells. The consequence is the polymerization of a pore-forming protein known as perforin that permeabilizes the target cell. The second mechanism involves secretion of a cell toxin that enters the cell through the perforin pores and kills the target cell by inducing it to undergo apoptosis, also known as programmed cell death. The importance of NK cells in combating poxvirus infections has been studied by correlating NK activity with disease severity as well as by measuring the effect of NK cell depletion on virus replication (32,101,190). NK cytolytic activity in the spleen has been shown to be elevated following intraperitoneal or intravenous inoculation of mice with vaccinia. In addition, a strain of mice resistant to infection with ectromelia shows a higher level of NK cell activity following inoculation with virus than do mice that are susceptible to infection (101). These data suggest that NK cell activity has a role in resistance to poxvirus infection. Evidence in support of this comes from studies in which the NK cell population is depleted through the use of antibodies against a prominent NK cell surface antigen (32,101). Treatment of mice with anti-NK antisera prior to infection with ectromelia, resulted in increased viral titers in the spleen and liver as well as an increase

PAGE 79

60 in the severity of the disease following infection. Finally, while certain strains of mice are normally resistant to infection by poxviruses, other strains which have been derived from these mice and which contain mutations associated with deficiencies in NK cells are able to contract lethal infections (36). The above data strongly suggest a role for NK cells in the host response to poxvirus infections. Recent studies suggest that it is not the cytolytic activity of the NK cells but rather the production of IFN-gamma by these cells that is crucial (107). These studies indicate that while NK cells are involved in virus clearance, they are not sufficient to completely inhibit virus spread. Inflammation Prior to the development of a specific cellular immune response, the host mounts a non-specific inflammatory response consisting of vasodilation, edema and the appearance of inflammatory cells at the site of infection (2,43). The cellular component of this response consists of mononuclear phagocytes (monocytes/ macrophages) and granulocytes, the latter primarily of the type known as polymorphonuclear leukocytes (PMNs) or neutrophils. These cells engulf and degrade foreign particles through the action of degradative enzymes and reactive-oxygen intermediates. While both PMNs and macrophages are capable of phagocytosing and degrading foreign particles, they differ in that macrophages, but not PMNs, are involved in antigen presentation and the generation of the late specific immune response. Generally the PMNs are the first cells to appear, but they are soon followed by the appearance of macrophages.

PAGE 80

61 The movement of these professional phagocytes from the circulation to the site of infection is directed by numerous proinflammatory mediators. Some of these mediators serve to increase the mobility and phagocytic activity of the inflammatory cells while others affect the vasculature near the infection, causing vasodilation and the expression of cellular adhesion molecules on the vascular endothelial cells. Examples of such mediators are tumor necrosis factor (TNF), IFN-gamma and IL-1 (1,2). Still other mediators serve as chemoattractants for the migrating phagocytic cells, forming a concentration gradient beginning at the site of infection. Examples of this type of mediator include IL-8, monocyte chemotactic protein-1 (MCP-1), platelet activating factor (PAF), products of the kinin system and a product of the complement cascade known as C5a (184). It is likely that there are many more as-of-yet undefined proinflammatory mediators and it is also likely that there is significant overlap between the types of mediators described above. Like much of poxvirus pathogenesis, the exact role played by inflammation in the inhibition and clearance of the viral infection is unclear. Infection of mice with virulent strains of ectromelia fails to elicit a significant inflammatory response until 7 days post-infection, a time when the development of lesions is well under way (8,19). In contrast, inflammation has been observed within 24 hours following inoculation of rabbits, mice and guinea pigs with vaccinia virus or inoculation of rabbits with rabbitpox virus (16,19,49,85,127,147,214). Such differences in the inflammatory response could

PAGE 81

62 result from differences in the ability of the various viruses to either elicit or suppress such a response. Perhaps the more virulent viruses are able suppress inflammation until later in the infection. It is also likely that the virus and host combination is important in the type of response generated. In studies in which an inflammatory response has been reported, the initial cell-type observed in this response has also varied between studies. This may be due to variation in the inoculum size, routes of inoculation as well as the above mentioned variation among the virus species. Although inflammatory cells may be present at the site of infection, the role these cells play in disease clearance or progression is unclear. In vitro, it has been demonstrated that poxviruses are taken up when mixed with phagocytic cells (11,161,213). In addition, when the virus is injected intravenously into mice, more than ninety-percent is taken up by hepatic littoral cells within a few minutes of inoculation (124). A point about which there is little agreement, however, is the fate of the phagocytosed virus. Results from some of this early work suggest that phagocytosis results in destruction of virus or neutralization of infectivity in animals (11,52,63). However, workers performing similar studies have also reported that phagocytosis does not inhibit viral replication (15,69,76,161). It is possible that macrophages and neutrophils destroy some virus following phagocytosis but that enough virus survives to replicate. If correct, this model would mean that leukocytes may actually aid in enhanced dissemination of the virus, an

PAGE 82

63 idea that is supported by work showing that viable virus can be recovered from the leukocytes of infected animals (36,69,124). Recent studies, concerned with understanding the involvement of the inflammatory response in poxvirus infections, have focused on the inflammatory response generated following infection of the 11-day-old chicken embryo. At this stage of development, the embryo lacks a mature T and B cell response making the inflammatory response together with the complement system the only systems of defense. In this regard, the CAM makes an excellent system in which to study the non-specific response of the host to a poxvirus infection. Infection of the egg at this time results in the formation of characteristic lesions (pocks) on the chorioallantoic membrane (CAM). RPV, cowpox virus (CPV) and certain strains of vaccinia produce red pocks upon infection of the CAM. Microscopic examination of these pocks at 3-days post-infection reveals a region of ectodermal hyperplasia within the pock with edema of the underlying mesoderm (68). The pocks are frequently innervated by blood vessels which show extensive vasodilation. In addition, necrosis and extravasation of erythrocytes are seen; however there is a noticeable absence of inflammatory cells within the lesion. In contrast to these red pocks, viruses containing deletions of the RPV SPI-2 (CPV crmA) gene produce white pocks on the CAM. The SPI-2 protein has been shown to inhibit an enzyme involved in the generation of inflammatory mediators (see below). These white pocks produced by RPVASPI-2 show the same degree

PAGE 83

64 of vasodilation and extravasation as the RPV red-pocks; however, in contrast to the red pocks, the mesoderm of the white pocks is heavily infiltrated by leukocytes. The hyperplasia and severe impaction of heterophils within the lesion give the pock its white appearance. Similar results have been reported for the naturally occurring white-pock mutants. The massive inflammatory response observed in white-pocks correlated with a decreased level of virus from within the pock suggesting that inflammation plays a role in virus clearance in this system (142). However, even though there was a decrease in the level of virus detected in these primary lesions, virus was still able to travel to and infect internal organs of the embryo suggesting that the inflammatory response by itself is not sufficient to clear the infection (36). Recent work suggests that while inflammation by itself cannot contain the viral infection, it is necessary for virus clearance and disease resolution (202). In this work, treatment of mice with carrageenan, which depletes the macrophage pool, prior to infection with ectromelia in resulted increased virus replication. The above results, taken together, suggest that a combination of inflammation and other host responses is necessary for recovery from poxvirus infections. Viral Proteins that Modulate Host Immune Responses The success of the poxviruses as pathogens indicates that they have evolved ways of evading or altering the host immune response to the

PAGE 84

65 infection. Indeed, some of the most exciting work in this field in recent years has been the discovery that the poxviruses produce several proteins that either have homology with or interact with components of the host immune response. To date the list includes a viral C4b binding protein (VCBP), an inhibitor of the interleuken-16 converting enzyme (crmA), a virally produced secreted interleuken-16 receptor, secreted IFN receptors (alpha,beta,and gamma) and a secreted TNF receptor. Those proteins for which an activity has been described are briefly discussed below. Viral C4b binding protein. The major secreted 35 kda protein of vaccinia was shown to be a product of the C21L gene which is located in the left, non-essential region of the genome (111). Sequence analysis of this gene revealed four SCR elements covering the length of the protein, making it a member of the SCR superfamily. Database searches using this protein showed that it shares thirty-eight percent identity with the complement C4b binding protein resulting in the viral protein being named the viral C4b binding protein (VC4bBP). Subsequent studies have shown that purified VC4bBP can bind to C4b preventing stability of the C3 convertase and therefore inhibiting the complement cascade (110). VC4bBP was also shown to prevent antibody-dependent complement-enhanced neutralization of infectivity in a manner independent of complement C4b (99). The importance of this protein during infection was demonstrated using mutants lacking the VC4bBP ORF. The mortality rate caused by a

PAGE 85

66 VC4bBP null mutant following intracranial injection was decreased dramatically when compared to that of wild type virus. In addition, lesions produced by this mutant following intradermal injection of rabbits were smaller and healed more rapidly than those produced by wild type virus. Together, these results indicate that inactivation of the complement system is important for successful virus replication in these systems (36,99,110). The crmA/Spi-2 protein. Evidence that the poxviruses could modulate the immune response began to mount when it was realized that CPV whitepock mutants had lost the ability to inhibit the inflammatory response on the CAM. Further study narrowed the mutations to a single gene in the right, non-essential region of the virus that was given the name crmA (cytokine response modifier A) (155). Homologs have since been found in both VV and RPV, although crmA isolated from CPV remains the most well characterized. The product of the crmA gene is a 38 kDa cytoplasmic protein that has around thirty percent identity with members of the serine protease inhibitor superfamily which includes antithrombin and plasminogen activator inhibitor-2 (PAI-2) (36,155). Biochemical analysis has shown that the crmA protein is able to inhibit the interleuken-lG converting enzyme (ICE) which is responsible for the production of IL-16 from pro-interleuken-16 (158). The failure of crmA mutants to inhibit inflammation most likely arises from the failure to inhibit IL-1 production since IL-1 produced by macrophages is known induce neutrophilia (50). ICE has also recently been implicated as part

PAGE 86

67 of the apoptotic pathway, and it has been shown that crmA can inhibit cytokine-induced apoptosis (197). Since cytotoxic-T cell killing is mediated through apoptosis, inhibition of cell death by crmA suggests another mechanism by which the virus can alter the immune response. The importance of crmA in altering the host response has been well characterized in the chicken embryo where deletion of this gene results in heterophilia of the primary lesions (70). Interestingly, intranasal infection of mice with viruses lacking crmA yield a mortality rate identical to that of wild type virus. Finally, while RPV mutants of crmA still produce a severe disease in rabbits following intradermal infection, the mortality rate is zero compared with fifty to seventy-five percent for wild type RPV (R.Stern, M. Brooks, Unpublished data). These differences in effect might be attributable to species specificity or to differences in routes of inoculation. Interleukin-113 binding protein. Sequence analysis of DNA from the right, non-essential region of vaccinia (WR) revealed that the predicted protein product of the B15R gene is a member of the immunoglobulin superfamily and shares homology with the binding domains of the human and murine IL-1 receptors (183). This homology suggested that the viral protein may bind IL-1, a hypothesis that was shown to be correct when IL-1 binding activity found in the supernatants of vaccinia infected cells was abolished by deletion of the B15R ORF (187). The cloned B15R gene produces a glycoprotein of 40 kDa that has specificity for IL-16 but not IL-la, and which

PAGE 87

is able to compete with the natural IL-1 receptor for binding of IL-1 (7). Inactivation of the viral IL-1 receptor from vaccinia virus has produced contradictory effects in mice (7,187). Viruses lacking the B15R ORF were attenuated when mice were inoculated intracranially but were more virulent when inoculation was via the respiratory route. This result is not surprising since increased IL-1 production and the subsequent immune response in the lung may lead to increase pathological damage. This suggests that some viral proteins can act to decrease the damage resulting from the host immune response and may therefore act to make the virus less pathogenic. Viral interferon receptors. The production of an IFN receptor by poxviruses was first reported by groups studying myxoma virus but it has since been shown that 17 of the orthopoxviruses, including vaccinia, cowpox and rabbitpox, produce a secreted IFN receptor (5,130,205). Vaccinia has been shown to produce two IFN receptors; an IFN-gamma receptor (IFN-yR) and an IFN a/G receptor (IFNa/fiR). The IFN-yR is a product of the B8R ORF from the right non-essential region and the predicted amino acid sequence of this gene shows an overall twenty-one percent identity with the extracellular domain of the human and mouse IFN-yR. The activity of the viral receptors is species specific and it has been suggested that these specificities may be useful in determining the origins of the various viruses. Viral TNF receptor. Open reading frames coding for proteins with homology to the tumor necrosis factor receptor (TNFR) were first described

PAGE 88

69 for Shope Fibroma Virus and myxoma virus (182,204). The SFV and myxoma TNFR ORFs ( named T2) share forty and forty-three percent identity, respectively, with the extracellular domain of human TNFR-1. Expression from a recombinant T2 gene results in production of a glycoprotein of approximately 55 kDa that is secreted and binds both TNF-a and TNF-S in vitro. These recombinant viral TNFRs have also been shown to compete with human TNFR for human TNF. Recently, the crmB gene of CPV, which shares forty-two percent identity with the binding domain of human TNFR, has also been shown to produce a 48 kDa protein that binds to TNF-a and TNF-fi in vitro. The importance of the TNF receptors in virus infection has been examined in the chicken embryo and in rabbits. Infection of the CAM with a CPV TNFR null mutant, resulted in the production of wild type red pocks indicating no involvement of this protein in the pock color phenotype (97). In contrast, deletion of the T2 ORF from myxoma virus resulted in attenuation (204). The disease produce in rabbits following injection of this virus was much less sever than that produced by wild type myxoma. In addition, the mortality rate of the mutant was decreased by seventy percent in comparison to the mortality rate of wt virus.

PAGE 89

70 The B5R Protein of RPV The B5R (the fifth ORF from the left end of the Hindlll B fragment) protein of RPV, the focus of this dissertation, is another poxvirus protein that shares homology with components of the host immune system. Proteins having nearly identical sequence have been shown to exist in both vaccinia virus and variola virus (57,100,119,120,195). The homologies between the various viruses and between B5R and the complement components will be described below in detail. The B5R ORF was first isolated from the Lister strain of vaccinia in 1991 by marker rescue of a low-neurovirulence strain of vaccinia lister known as LC16m8. This virus is a naturally occurring variant that was of interest because the plaques it formed on RK-13 cells were consistently smaller than those produced by the parent strain, LC16mO. In addition, LC16m8 had an altered host range and was unable to grow in Vero cells, unlike the parent strain LC16mO. Marker rescue experiments resulted in the isolation of a gene from the right-hand non-essential region of the genome that restored both the ability to form large plaques on RK-13 cells and to grow in vero cells. Based on these phenotypes the gene was originally called the plaque size/ host range (abbreviated ps/hr) gene. Functioning homologs of the ps/hr gene have since been identified in the right end (Hindlll B fragment) of the

PAGE 90

71 vaccinia and RPV genomes (the B5R ORF) and of the variola genome (the B7 ORF). The sequence of the B5R gene has been determined for several strains of vaccinia, variola and for RPV allowing comparisons to be made between viruses and with the protein database. Comparison of the predicted amino acid sequence of the viral genes, as is illustrated in Figure 5, shows that among the members of the poxviridae family for which B5R homolog sequence is available, the protein sequence of the various B5R homologs has been extremely well conserved. In all cases the degrees of similarity and identity between the different homologs are greater than ninety-six percent and ninety-three percent (Table 1), respectively suggesting the importance of this protein to survival of the virus. The B5R protein of RPV is one-hundred percent homologous with that of VV(Copenhagen) while it is most diverged from the variola strains India and Bangladesh, although it should be noted that the differences here are still extremely small, only six percent at the level of amino acid identity. Most of the differences observed are the result of conservative substitutions of amino acids and it is likely that they have little effect on the overall protein structure. Analysis of the RPV B5R amino acid sequence reveals that this protein has N-terminal and a C-terminal hydrophobic regions that in the vaccinia protein have been shown to function as a signal sequence and a membrane anchor, respectively (100). In addition, the protein, which is shown

PAGE 91

Figure 5. Alignment of the orthopoxvirus B5R homologs. Shown is an alignment of the complete predicted amino acid sequences from the currently available B5R homolog genes. Boxed areas indicate differences within the sequences of the various proteins. Black bars underline potential glycosylation sites. UTR, Utrecht; COP, Copenhagen; WR, Western Reserve; LIS, Lister; BAN, Bangladesh; IND, India.

PAGE 92

73 RPV (UTR) MKTISWTLLCVLPAWYSTCTVPTMNNAKLTSTETSFrfJNt2KVTFTCEjQ W(COP) MKT I W(WR) MKT I W(LIS) MKT I VAR ( BAN ) MKT I VAR(IND) MKT I iswtllcvlpawystctvptmnnakltstetsfn sin 2kvtftce q iswtllcvlpawystctvptmnnakltstetsfn dk 2kvtftce q [swtllcvlpawystctvptmnnakltstetsfn dk 2kvtftce q [swtllcvlpawystctvptmnnakltstetsfn dk2kvtftce s' [swtllcvlpavwstcwptmnnakltstetsfnqkJ2kvtftce|s 3\ H : 1\ Y E yi y 60 s dpnav S DPNAV S DPNAV S DPNAV L DPNAV L DPNAV RPV(Utr) W(COP) W(WR) W(LIS) VAR (BAN) VAR(IND) CETDKWKYENPCKKMCTVSm! I 5ELY N KPLYEVft STM EE S NG ETKYFRCEEKNG CETDKWKYENPCKKMCTVSD'* I CETDKWKYENPCKKMCTVSD'i I CETDKWKYENPCKKMCTVSD^ V SEWi N KPLYEVIN All CETDKWKYENPCKKMCTVSD^ V 3ELi D KPLYEVIN STM IT S I NG ETKYFRCEEKNG NTS WN CETDKWKYENPCKKMCTVSD"^ELMNkPLYEV NAIll rdlttafeTKYFRCEEKNGNTSWN 3ELY N KPLYEVIS STM n S I NG ETKYFRCEEKNG NTSW N 3ELY N KPLYEVT SJTM n S I NG ETKYFRCEEKNGNTS.WH 120 3NTSWN KE ETKYFRCEEKNG NTSW N RPV (UTR) W(COP) W(WR) W(LIS) VAR (BAN) VAR(IND) DTVTCPNAECC P LQI E IGSCQPVKEKYSFGE YV. TINCDVGYEVIGASY1 S DTVTC PNAECC P 1.QI E 3GSCQPVKEKYSFGE YV. TINCDVGYEVIGASYI S DTVTCPNAECC P LQI E -JGSCQPVKEKYSFGE YV. TINCDVGYEVIGASYI S DTVTC PNAECC P L.QI E 1GSCQPVKEKYSFGE YI TINCDVGYEVIGASYI S DTVTCPNAECC S uQI D 4GSCQPVKEKYSFGE HI TINCDVGYEVIGASYI T DjrVTCPNAECqsLQdDHGSCQPVKEKYSFGgHITriNCDVGYEVIGASYIlT 180 itanswnvip itanswnvip tanswnvip :tanswnvip itanswnvip itanswnvip RPV (UTR) SCQQKCIT PSLSNGLISGSTFSIGGVIHLSCKSGF W(COP) SCQQKCIT PSLSNGLISGSTFSIGGVIHLSCKSGF W ( WR) SCQQKCI M PSLSNGLISGSTFSIGGVIHLSCKSGF W (LIS) SCQQKCI M PSLSNGLISGSTFSIGGVIHLSCKSGF VAR (BAN) SCQQKCIT PSLSNGLISGSTFSIGGVIHLSCKSGF VAR(IND) SCQQKCIT PSLSNGLISGSTFSIGGVIHLSCKSGF 240 I liTGSPSSTCIDGKWNI V.I I V • T I jTGSPSSTCIDGKWNI V jl T ) V T T LiTGSPSSTCIDGKWNI V IT I V T T LiTGSPSSTCIDGKWNI I IE T V : S I LiTGSPSSTCIDGKWNI V .11 1 • S I LiTGSPSSTCIDGKWNI V IT I : T : S rpv ( utr ) ne e ?dp\ t dgpddetdlsklskdwqyeqeiesleatyhi i w(cop) ne e ?dp\ t dgpddetdlsklskdwqyeqeiesleatyhii w(wr) ne e ?dp\ i dgpddetdlsklskdwqyeqeiesleatyhii w ( lis ) ne k fdp\ e dgpddetdlsklskdwqyeqeiesleatyhii var ( ban ) ne e fdp\ e dgpddetdlsklskdwqyeqeiesleatyhii var ( ind) ne^epdp\(fJdgpddetdlsklskdvvqyeqeiesleatyhii 300 IVALTIMGVIFLISVIVLV IVALTIMGVIFLISVIVLV IVALTIMGVIFLISVIVLV IVALTIMGVIFLISVIVLV IVALTIMGVIFLISVIVLV IVALTIMGVIFLISVIVLV RPV (UTR) W(COP) W(WR) W(LIS) VAR (BAN) VAR (IND) CSC D KNNDQYKFHKLE CSC D KNNDQYKFHKLI CSC D KNNDQYKFHKLI CSC D KNNDQYKFHKLI CSC N KNNDQYKFHKLI CSC N KNNDQYKFHKLI 17 P P P P L

PAGE 93

T3 c T3 0 CO pa a; 0> S3 S Oh 5 03 o> 3 ^> s to _, qj _o> O O CO g<3 a 2 'S 1 5 S2 ^ S 60 u o l— 1 o> O re g oi toe o o 5 5 J3 tJ o c o 2 T3 O n to mh n, o 03 -i— < "S 55 > tJ o > oi +j a C 5 eft s •rj O n U y ? oi cfl O eft CO id > T3 n > o U > > to > > PS > > > H 'to o to K o o o o > o o o o ON on CO ON > > o o o o ON ON CO On CO On Ifl <*£ ON eft • fH > > o O O O CO On Ifl v ON ON On ON O o o o a, o U > > o o o o CN| on ON CO Ifl On q cfl ON CN ON m ON. ON ON 73 id > o o o o o a o vd ON. tn ON CO ON Cfl ON

PAGE 94

schematically in Figure 6, contains four contiguous short consensus repeat (SCR) elements, which comprise the first seventy percent of the protein, making it a member of the SCR superfamily. Alignment of these regions with the SCR domains from several other members of this superfamily is shown in Figure 7. While it is obvious that the overall sequences differ, it is also clear that the highly conserved residues that define the SCR domain are present in the B5R protein. These include the essential cysteine and proline residues that begin and end each domain as well as several less well conserved residues found at characteristic positions within the domain. Included in this alignment are the human and viral C4b binding proteins which illustrate how divergent two proteins may be in their sequence and yet still share the same activity. Comparison of the RPV B5R predicted amino acid sequence with the current database using BLAST reveals that this protein shares homology with several complement control proteins, with the areas of greatest homology falling, not surprisingly, within the SCR domains. B5R shares the greatest degree of homology with the complement components factor H and the C4B binding protein, both of which are inhibitors of complement activation. An illustration of the B5R regions that share this homology as well as an indication of the degree of homology are shown in Figure 8. The regions of homology basically cover two separate areas of the B5R protein, one region within SCR I and a second region within SCRs II and / or III. The region of

PAGE 95

(U c o 'un '5b 0) -*-< to -o ^_ leava (LI c nce sen QJ epres sequ repr X -t-< m_ s0 (A 0 c C T3 o OJ o 'a O pa *CO 53 15 oj £

PAGE 96

IT) O If) IT)

PAGE 97

03 O. Cl> s-c o> rj o oi sO XI ay C_ U X 0> -t— r£ bo so o r n o> ; ^ < ^ > £ £ *4> CS O SO c • I— N £ IS |U re +3 .9 -a ai u 3 c >u > s01 e o u c "3 _£ cn Li Oi o> (A Ol s" -o o 0j C -3 *^ ^ X i/i 60 Oi ;j2 §f o oi DC 01 In QhMO re £ ^th" X 2 £ 73 3 mI o CO £ C (B O T3 0) c = £ *re £ 4> R £ .S -a T5 2 T3 '3 -o re 03 c/1 01 <8 0) 3 3 3 (0 £ 3 O V pQ CO 01 ^ C .5in 0 CO 5 CO aj J 01 X X H 6 > a; S o u a; 3 bE £ o O) cfi Li O 3 O U Oi O In to c 2 x U u 3 O £ a. Ml Xl O rj* 3-U c bD o -3 re SX

PAGE 98

Eh 2 O U C) M J > K in ft n b a a ft ei E !> < .Q rH 3 ft U <* h K K > U w cn o o fa Q tO H • • o< w W Q J ft ft ft ft ft i < cn • cn o u u u cn w H 2 cn H • W S S u ft ft • a • • • 2 h J • a w ; Ej w w CX Q 2 W O M ft Eh O H W fu W >h 2 2 3 J J > D J W W O Eh H HI H J Q M J > K m ft m pa a ft n B > rH 3 ft U b a a > o u PS ft 2 PS P C5 J PS W Eh > u o o > PS H ft ft ft a h O m PS • J >_e oD5 fa >H Eh W M • PS ft • u W fa a a ri. P cn M J _bi fa cn cn cn Eh U U CJ U CJ u > cn j cn p> < H > M w • >h o cn W • cn cn < S in ft m h pa a ft ro e > •* a rH 3 ft U b a Pi > o

PAGE 99

*8l g^ g3 g is s e c .2 ^ -o g -js a § g s ~ S.& jS (u ft ^ -Q b o *J re u o ~ g £ .a £§i g g a ^2 2j§ ^ CD ~ x J= oj es; o> aJ H £ ai £ H o £ I— I 5 £ M s g g> o ^ ^ c -5b O -c *j a> { Bfli £ o < £ o aV§) 60 C M & W o C c ~ lo £ -5 fl a> C ^ CO *! X yqi uj jj s M T3 O g 2 o o> > t>0 C 0(0 >, £ > O 0> — > •5b ts .S g a> O t/5 <*i C CJ c c

PAGE 101

82 shared homology within SCR I stretches from amino acids 18 to 71, and has percent identities ranging from thirty-four to forty-five percent depending on the alignment used while the second region of homology lies between amino acids 112 and 231 with percent identities ranging from twenty-three to fortyeight percent, again depending on the alignments used. This degree of homology may be considered significant since it is similar to the homology shared between the viral and human C4b binding proteins, two proteins that have been shown to have the same activity. The shared homology of B5R with proteins that are regulators of the complement cascade suggests that the B5R protein may also function as an inhibitor of the complement cascade. The expression and biochemistry of the W and RPV B5R proteins have been studied in some detail. The product of the B5R gene is produced late in infection and is a 45 kDa protein that has been shown to be both glycosylated and palmitated (57,100,119,147). The 45 kDa glycoprotein is found as a component of EEV but not of IMV, and it has been demonstrated to be part of the EEV envelope (119). In addition, the B5R protein is found in the supernatant of infected cells as a smaller, secreted 37 kDa protein derived by cleavage of the 45 kda form, a process believed to occur at the cell surface (119). This process is independent of viral morphogenesis indicating that there is a second pathway for secretion of this protein and suggesting a dual role for B5R.

PAGE 102

The B5R protein of vaccinia has been shown to be essential for EEV synthesis in some cells, a feature characteristic to many of the EEV proteins. To date, the importance of this protein in virulence has not been determined, and there are no data available on the relative importance of the individual forms of this protein. The goal of this work is to explore the importance of the membrane-bound and secreted forms of B5R in virus virulence. The work in chapter three will investigate the overall importance of B5R in virulence while the work in chapter four will address the importance of the individual forms of the protein.

PAGE 103

CHAPTER II MATERIALS AND METHODS Recombinant DNA Techniques Bacterial Strains and Plasmids Escherichia coli (strain UT481) was used for the propagation of all plasmids using standard culture conditions (173) unless otherwise specified. pBluescript, which was used for general cloning and construction of mutant open reading frames, was obtained from Stratagene Inc. (San Diego, Ca.). The plasmid pBluescriptrJF, in which the vaccinia virus thymidine kinase open reading frame, containing an internal polylinker site, has been cloned into the Hind III and Xho I sites of pBluescript, was obtained from Dr. Joyce Feller through coercion and the threat of physical violence. The plasmid pBS-gpt-A, which contains the £. coli. guanyl-ribosylphosphotransferase gene under the control of the poxvirus p7.5 promoter, was kindly provided by Dr. Grant McFadden. The PCR product cloning vector pCRl was obtained from Invitrogen (San Diego, CA). The plasmid pBR322 was obtained from 84

PAGE 104

85 Stratagene Inc. The plasmid pMALCR-1, for the production of maltosebinding protein fusion proteins, was obtained from New England Biolabs (Beverly, MA). Manipulation and Cloning of DNA All restriction digests of DNA were performed using restriction endonucleases and buffers obtained from New England Biolabs (Beverly, MA) using reaction conditions specified by the manufacturer. Agarose gel electrophoresis was performed as described by Sambrook and Maniatis (172), using 0.8% (w/ v) SeaKem LE agarose (FMC Bioproducts) gels buffered in TBE ( 0.089 M Tris-hydroxy methyl aminomethane (THAM) (Fisher), 0.089 M Boric acid (Fisher), 0.002 M Disodium ethylenediamine tetracetic acid (EDTA) (Fisher)) or TAE (0.04 M Tris-acetate, 0.01 M EDTA) containing ethidium bromide (2,7-diamino-l 0-ethy 1-9-pheny 1-phenanthridinium bromide) (Sigma) at a concentration of 0.5 /ig/ ml. Purification of DNA from agarose gels was accomplished using the Geneclean II Kit (Bio 101 Inc.) or by the polyethylene glycol (PEG) method of electroelution, as described by Zhen and Swank (220). All ligation reactions were incubated overnight at 16C using T4 DNA Ligase (New England Biolabs) in buffers and conditions specified by the manufacturer. The ligation products were transformed into competent E. colx. cells using the calcium chloride procedure (88), and transformants were selected on LB (1% (w/ v) tryptone 0.5% (w/ v) yeast extract, 1% (w/v) NaCl)

PAGE 105

86 agar plates (1.2% (w/v) Difco Bacto-agar) containing 100 /ig/ml ampicillin (Sigma). Plasmids were purified from bacterial cultures using the Wizard Midiprep DNA purification system (Promega) or by the standard alkaline lysis protocol (173). PCR Amplification of DNA All primers used for PCR amplification were synthesized by the University of Florida ICBR DNA synthesis core, and PCR reactions were carried out in PTC-100 Thermal Controller (MJ Research) using thin-walled tubes (USA Scientific Plastics). Unless otherwise specified, 30 cycles of amplification, consisting of denaturing (94C, 20 seconds), annealing ( 52C, 30 seconds), extension ( 72C, 60 seconds) were typically performed on 400 pg of template DNA and at least 60 pM of each primer. Amplification reactions were done in a volume of 100 \i\ using Taq DNA Polymerase (FisherBiotech) in reaction conditions specified by the manufacturer (50 mM Tris pH 9.0, 50 mM NaCl, 10 mM MgCl 2 50 j*M each of dATP, dGTP, dCTP, dTTP). For generation of the DNA fragments used in construction of the RPVB5R:X mutants (B5R non-secretors), all of the PCR amplifications were performed with Vent DNA polymerase (New England Biolabs), to minimize PCR-generated errors, using buffers and reaction conditions provided by the manufacturer (50 mM Tris pH 9.0, 50 mM NaCl, 10 mM MgCl 2 50 //M each of dATP, dGTP, dCTP, dTTP).

PAGE 106

87 Southern Blot Analysis of DNA Southern blot analysis was performed using standard protocols (31). Briefly, the agarose gel containing the separated DNA fragments was soaked for 15 minutes in 2 volumes of nicking solution (0.25 M HQ), 2 x 15 minutes in denaturation solution (1.5 M NaCl, 0.5 M NaOH) and 2 x 15 minutes in neutralization buffer (1.5 M NaCl, 0.5 M Tris-HCl [pH 7.2], 0.001 M EDTA). The DNA was transferred to a nylon membrane (Hybond-N, Amersham, Inc.) by downward capillary transfer in 0.025 M sodium phosphate (pH 6.5). The blot was dried, and the DNA fixed to the membrane by irradiation with 120 millijoules using the UV Stratalinker 1800 (Stratagene). The generation of labeled probes using 32p-dCTP (3000 Ci/mM, New England Nuclear) was performed using the random oligomer primer extension protocol (64), and unincorporated nucleotides were removed by centrifugation of the extension products through sephadex G-50 (Pharmacia). Blots were blocked by incubation for one hour at 65C in 6X SSC (0.9 M NaCl, 0.09 M sodium citrate [pH 7.0J (Sigma) containing 0.25% dry milk (Carnation Foods)) following which hybridization with the denatured, labeled probe was carried out under the same conditions for 16 hours. Blots were washed in 2X SSC (0.3 M NaCl, 0.03 M sodium citrate [pH 7.0]) containing 0.1% (w/v) SDS (Sigma) 2 X 15 minutes, once at room temperature and once at 65C. Blots were then exposed to Kodak X-Omat AR film.

PAGE 107

88 Virological Techniques Cell Culture All cells were maintained at 37C with 5% CO2 in (with the exception of chick embryo fibroblasts (CEFs) and quail carcinoma (QT-6) cells) Gibco minimal essential medium (MEM) supplemented with streptomycin (50 I.U. /ml), penicillin (50 mcg/ml), glutamine (2 mM), sodium pyruvate (1 mM), MEM nonessential amino acids (0.1 mM) (Mediatech Inc. /Fisher), and 5% (v/v) fetal bovine serum (FBS) (GIBCO BRL). QT-6 cells were maintained in Optimem medium (GIBCO BRL) with 5% (v/v) FBS. Chicken embryo fibroblasts were prepared from 11 day old embryos and maintained for 8 passages in supplemented medium F-199 (Gibco) containing MEM vitamin solution [final concentration: NaCl (170 mg/ml), D-Ca pantothenate (2 mg/ml), choline chloride 2 mg/ml), folic acid (2 mg/ml), i-inositol (4 mg/ml), nicotinamide (2 mg/ml), pyridoxal HC1 (2 mg/ml), riboflavin (0.2 mg/ml), thiamine HC1 (2 mg/ml)l (Gibco BRL) and 5% (v/v) fetal bovine serum Preparation of Viral Stocks RPV (Utrecht strain) was obtained from the American Type Culture Collection (ATCC). CPV (Brighton Red strain) was kindly provided by Dr.

PAGE 108

89 David Pickup. RPVASPI-2 and CPVASPI-2 were graciously provided by Dr. Ahmad Ali. Confluent monolayers of RK-13 cells (ATCC), grown in 150 mm dishes, were infected with virus at a multiplicity of infection (m.o.i.) of between 0.01 and 0.1 plaque forming unit (pfu) per cell and incubated at 37C until all of the cells showed cytopathic effect (CPE). The cells were scraped from the dish using a rubber policeman (Fisher Scientific), and collected by centrifugation at 10,000 RPM in a Beckman 1A-10 rotor (approx. 800 x G) for 30 minutes. The pellet was resuspended in serum-free MEM and stored at -70C until used. Virus that was used for animal studies was subject to the following purification: Infected cells were scraped in to the medium and collected by centrifugation at 10,000 RPM in a Beckman JA-10 rotor (approx. 800 x G) for 30 minutes. The resulting cell pellet was resuspended in TM buffer (10 mM Tris-HCl fpH 7.5], 5 mM MgC^), and after 10 minutes the cells were lysed by 10 strokes with a dounce homogenizer. The resulting lysate was centrifuged at 2,000 RPM in a Beckman JA-20 rotor (approx. 500 x G) to remove the nuclei. The supernatants from this centrifugation were then sonicated for 30 seconds using a Vibra-Cell probe sonicator (Sonics and Materials, Danbury, CT.) set on level 2 after which they were layered onto a 36% (w/v) sucrose pad and centrifuged at 18,000 RPM in a Beckman SW28 rotor (approx. 65,000 x G) for 90 minutes. The resulting pellet was resuspended in MEM and stored at -70 C.

PAGE 109

90 Purification of IMV. IEV and EEV IMV (intracellular mature virus), IEV (intracellular enveloped virus) and EEV (extracellular enveloped virus) particles for western blot analysis were purified using a modification of the procedure described by Payne (145). Purification started with infection of 5-10 150 mm dishes of RK-13 cells infected with the appropriate virus at an m.o.i. of 0.01-0.1. After total CPE was observed, the infected cells were scraped into the media and pelleted by centrifugation at 10,000 RPM for 30 minutes. EEV was recovered by centrifugation of this supernatant at 23,000 RPM in a Beckman SW28 rotor (approx. 100,000 x G). The EEV pellet was resuspended in 3.0 ml distilled water for purification on CsCl gradients. To obtain a crude cellular fraction of IMV and IEV, the infected cell pellet was resuspended in 3.0 ml of TM buffer (10 mM Tris-HCl [pH 7.5], 5 mM MgCl2), incubated for 10 minutes at room temperature, lysed by ten strokes with a dounce homogenizer and centrifuged at 500 x G for 5 minutes to remove the nuclei. The supernatant and cellular samples were both sonicated for 30 seconds using a VibraCell probe sonicator (Sorties and Materials, Danbury, CT) set on level 2. The samples then layered onto individual discontinuous CsCl gradients that had been prepared by layering 2.0 ml CsCl (1.3 mg/ml), 3.0 ml CsCl (1.25 mg/ml) and 4.0 ml CsCl (1.2 mg/ml) in an SW41 centrifuge tube.

PAGE 110

91 The gradients were centrifuged at 30,000 RPM in an SW41 rotor (approx. 175,000 x G) for 90 minutes after which time the resulting virus bands were collected by centrifugation at 23,000 RPM in a Beckman SW28 rotor (approx. 100,000 x G), resuspended in PBS (137 mM NaCl, 2.7 mM KC1, 4.3 mM Na 2 HP0 4 1.4 mM KH 2 P0 4 fpH 7.3]) (Fisher) and stored at -70C. Titration of Virus All titers of viral stocks were determined on RK-13 cells except where mycophenolic acid selection of recombinant virus was performed, in which case, CV-1 cells (ATCC) were used. Virus stocks were thawed, mixed thoroughly and sonicated for 60 seconds in a Vibracell sonicator with a cup attachment. The sonicated virus stock was then serially diluted in serum-free MEM and 0.4 ml. of each dilution was used to infect a confluent monolayer of PBS (Fisher) washed cells in a 35 mM dish. Infected cells were incubated at 37C for 2 hours and were rocked every 30 minutes to distribute the media. At the end of two hours, the medium was removed, and the cells were overlaid with 0.4% (w/ v) LE Seakem agarose (FMC Bioproducts) melted in supplemented MEM containing 2% (v/v) FBS Cells were then incubated at 37C for 2-3 days after which time, 3 ml of serum-free MEM containing neutral red (3-amino-7-dimethylamino-2-methylphenazine hydrochloride) (0.111 g/L) was added to each dish and incubation at 37C continued, until the cells were stained and the plaques could be easily counted.

PAGE 111

92 In experiments where mycophenolic acid selection was used, CV-1 cells were pre-treated for at least four hours with mycophenolic acid (10 /ig/ ml) prior to infection. In addition, the agarose overlay contained MPA (5 /Jg/ml), hypoxanthine (15 jig/ ml), and xanthine (250 /ig/ ml). Cells were stained with neutral red (as described above) 3-4 days following infection. When selection of thymidine kinase null mutants was performed, the virus stock was titered on Rat-2 cells, and the overlay contained bromodeoxy uridine at a concentration of 100 /ig/ml. Titration of Virus From Infected Lung s Infected mice were sacrificed by cervical dislocation. The lungs were removed, weighed and the tissue disrupted by agitation in 2.0 ml. of sterile distilled water, using a stomacher lab blender 80 (Tekmar Co., Cincinnati, Ohio) until the mixture appeared homogeneous. The tissue was then further disrupted by 10 strokes with a dounce homogenizer after which the homogenate was frozen and thawed two times. Samples were then sonicated for 60 seconds using a Vibra-Cell probe sonicator on level 2 after which titration was performed, as described above. One Step Growth Experiment Confluent monolayers of cells in a 60 mm dish were infected with the appropriate virus in serum-free MEM at an m.o.i. of 10. At two hours post-

PAGE 112

93 infection, [methyl-3H]thymidine (25 Ci/ mMol., Amersham) was added to the dish to a final concentration of 10 jiCi/ml, and incubation was continued until 24 hours post infection. The medium was then removed, clarified by centrifugation in a 2,000 RPM in a Beckman JA-20 rotor (approx. 500 x G) for 10 minutes, and the virus pelleted by centrifugation at 23,000 RPM in a Beckman SW28 rotor (approx. 100,000 x G) for 40 minutes. The viral pellet was resuspended in 2.0 ml of distilled water for analysis on a CsCl gradient. The infected cells were harvested in PBS (Fisher) by scraping, pelleted at 1,000 x G for 10 minutes and following a 10 minute incubation in distilled water at room temperature, the cells were lysed by ten strokes with a tight fitting dounce homogenizer. Nuclei were then removed by centrifugation at 2,000 RPM) in a Beckman JA-20 rotor (approx. 1,000 x G) for 10 minutes. The supernatants were sonicated for 10 seconds using a VibraCell probe sonicator at level 2 (Sonics and Materials, Danbury, CT), layered onto a 4.0 ml 36% (w/ v) sucrose pad and centrifuged at 31,000 x G for 90 minutes. The resulting viral pellet was then resuspended in distilled water for analysis on a CsCl gradient. The 2.0 ml virus samples, derived from either the medium or the cells were layered onto a discontinuous CsCl gradient (4.0 ml CsCl [1.2 mg/ml], 3.0 ml CsCl [1.25 mg/mll, 2.0 ml CsCl [1.3 mg/ml] (Sigma) ) and centrifuged at 30,000 RPM in an SW41 rotor (approx. 175,000 x G) for 90 minutes. Fractions (0.5 ml) were collected, dropwise, from the bottom of the gradient and the level of radioactivity in each fraction was determined.

PAGE 113

94 Plaque Lift Hybridization The detection of recombinant viruses by plaque-lift hybridization analysis was done as has been described (206). Briefly, cell monolayers containing virus plaques were transferred to nitrocellulose membranes, and these membrane were then soaked for 1 minute in 0.25 M NaCl, 0.1 N NaOH, 1 minute in 0.2 M Tris-HCl [pH 7.6], and 1 minute in 2X SSC (0.3 M NaCl, 0.03 M sodium citrate fpH 7.01). The membranes were air dried after which time the DNA was fixed to the membrane by baking in a vacuum oven at 80C for three hours. The membranes were then incubated in 6X SSC (0.9 M NaCl, 0.09 M sodium citrate [pH 7.0] (Sigma) containing 0.25% dry milk for one hour at 65C after which time heat denatured 32p.] a beled probe was added and the incubation continued overnight. (Radiolabeled probes were made as described in the section on southern blotting). The membranes were then washed 2 x 15 minutes with 2X SSC containing 0.1% SDS (Sigma), the first wash at room temperature and the second at 65C. The membranes were sandwiched between clear plastic wrap, placed against film (Kodak X-Omat) and stored at -70C until developed, generally 1-2 days. Transfection of Cells A confluent monolayer of CV-1 cells, grown in a 35 mm dish, was infected at an m.o.i. of 0.05 pfu/cell and incubated at 37C for 60 minutes. At

PAGE 114

1 hour post-infection, 1.0 ml of Optimem was added, and the infected cells were transfected with 5-10 ug of the appropriate plasmid DNA using Transfectase (Life Technologies), as specified by the manufacturer. Following a 48 hour incubation at 37C, cells were scraped into the medium, frozen and thawed two times and then stored at -70C. When the transfected DNA contained the Ecogpt cassette, the medium added just prior to the addition of DNA contained mycophenolic acid (5.0 f/g/ml final concentration), xanthine (250 /ig/ml final concentration) and hypoxanthine (15jig/ml final concentration) (Sigma). Infection of the Chorioallantoic Membrane The chorioallantoic membrane was infected as has been described previously (53). Embryonated, White-Leghorn chicken eggs were obtained from SPAFAS Inc. (Norwich, CT), and incubated in a Kuhl incubator (Flemington, N.J.) according to the manufacturer instructions (37.5C, 82C wet bulb relative humidity) with rocking every 30 minutes. At 11 days of incubation, two small holes were made through both the shell and the underlying membrane, the first into the air space on the large end of the egg. and the second on the long side of the egg in between visible blood vessels. The air was then drawn out of the air sac through the first hole, using a rubber bulb, resulting in the chorioallantoic membrane (CAM) separating from the shell, leaving a new air space on the side of the egg, under the

PAGE 115

96 second hole. 100 [i\ of sterile PBS, containing virus, at a concentration of either 500, 1000 or 2500 pfu/ml, was then dropped onto the CAM through the second hole, using a tuberculin syringe, after which the egg was gently rocked to distribute the virus. The holes were covered with tape and the eggs were returned to the incubator, lying on their sides with the new air space facing upward. At 72 hours post-infection, the eggs were opened, the CAM surrounding and including the infection site was carefully removed and rinsed with PBS for examination. In experiments involving dexamethasone treatment of eggs, the inoculum contained dexamethasone (Sigma) at a concentration of 8 mM. Staining of the infected CAMs with nitrobluetetrazolium (NBT) ( 2,2'-di-p-nitropheny 1-3,3' [dimethoxy -4,4'diphenylenej-ditetrazolium blue)(Sigma) was accomplished by a 30 minute incubation of the membranes at 37C in PBS containing 0.1% (w/v) NBT. Intranasal Infection of Mice Female BALB/ C mice (6-8 week old, 1520 grams) were obtained from Harlan, Sprague-Dawley Inc. (Indianapolis, In.) and housed 4 animals per cage unless otherwise specified. Intranasal infections of mice were performed as follows: Mice were first anesthetized by intraperitoneal injection of 100 /il (0.05-0.1 mg/gm body weight) of sodium pentobarbital (10 mg/ml in 17% (v/v) glycerol, 8.6% (v/v) absolute ethanol) following which, the mice were inoculated, 10 \i\ per nostril, with virus diluted in sterile PBS, using a

PAGE 116

97 microcapillary pipet. The weight of individual mice was determined using a pan balance. LD50S for the individual viral strains were determined using the method Reed and Muench (159). In experiments involving the use of dexamethasone, mice were given twice-daily intraperitoneal injections of 100 fil of sterile PBS containing dexamethasone-21-phosphate (Sigma) (500 /ig/ml). Ear Swelling Assay The anti-inflammatory activity of dexamethasone was confirmed by the use of an ear swelling assay (201). Mice were first anesthetized as described above. Pre-treatment ear thickness was then determined using an engineer's micrometer (Starrett Inc., Athol, MA.) after which, 25 fi\ of croton oil (5% (v/ v) in ethanol) (Sigma) was applied to one ear. The other ear was left untreated and served as an internal control. After four hours, ear thickness was again measured using an engineers micrometer. Ten separate readings per ear were used to calculate the average thickness for that ear. Intradermal Infection of Rabbits Adult (approximately 3 kg) female rabbits were obtained from M & P's rabbitry (Tampa, Fl.) and were housed in individual cages. To initiate the infection in these animals, virus resuspended in 100 y\ of PBS was injected intradermally into the shaved flank of each animal using a tuberculin

PAGE 117

98 syringe. Rectal temperatures were recorded using a digital thermometer (Becton Dickinson, Franklin, New Jersey). Histology Tissue samples from rabbits, mice and embryonated eggs were removed following euthanasia of the animal and stored in 10% (v/v) formalin in PBS until sectioning. All tissues were embedded and sectioned by the histopathology laboratory in the College of Veterinary Medicine at the University of Florida. Descriptions of the histopathology within tissue sections were provided by the Department of Clinical Pathology also at the College of Veterinary Medicine at the University of Florida. Construction of Mutant Viruses Construction of RPVAB5R The RPV B5R gene was amplified from RPV genomic DNA using primers RM152R, 5'-GGCTCGACATGAAAACGATTTCCGT-3', and RM151R, 5'-GGC AAGCTTT ATATTCACGGTAGCA-3'. The primers were constructed based on the published sequence of the B5R gene from the Lister strain of vaccinia and are derived from the 5' and 3' sequences of the gene, respectively. RM152R hybridizes to the first 17 nucleotides of the B5R ORF

PAGE 118

99 while RM151R hybridizes with the last 11 nucleotides of the B5R ORF. Underlined regions contain added, non-coded Sail and Hindlll sites used to clone the final PCR product into the plasmid pBR322 to produce the plasmid pBR322-ps/hr. The Ecogpt cassette was excised from pBS-gpt-A by digestion with the endonuclease EcoRI, treated with the Klenow fragment of the E. coli DNA polymerase I to obtain blunt ends, and ligated into the B5R gene at a single EcoKV restriction site located 152 bp downstream from the translation initiation codon. The location of this insertion relative to the overall protein is shown in Figure 9. The resulting plasmid was called pBR322-ps/hr-gpt. The portion of this plasmid containing the gpt-disrupted B5R gene was amplified using primers RM151R and RM152R and the linearized plasmid as the template. This PCR-generated DNA fragment was then transfected into RPV infected CV-1 cells as described above and gpt + recombinants were enriched by the addition of selection medium which contained mycophenolic acid (2.5 /ig/ml), 250 hypoxanthine (/ig/ml), and hypoxanthine (15 >/g/ml). Following an overnight incubation at 37C, recombinant virions containing the gpt cassette were identified by plaque-lift hybridization analysis of infected RK-13 cells (described above) using a probe for the gpt ORF. Gpt+ recombinants were isolated, the purification procedure repeated three more times after which virus recovered from isolated plaques was used for the generation of stocks. Construction was confirmed by southern and western blot analysis as described in this dissertation.

PAGE 119

T3 Ol IT, CD in • u re O o X -c „ HH ly a> bO > £ x n £! bo re T— I CD X -D C u eg 01 |H 0> x o> re Ih X O u bO re > re 0) u o sH re V X H T3 01 X -I— bO 2 ;=; IH 01 O K 3 pa 5 < x > "8 6 Oh w U 0> x o> X >> X -a CD U O. -c S-H re u re j O. o s3 bo o> X H >* Ch MH O c 01 o HH a in CO u X +> >^ bC O *o Oh O H cr — bJD Ph 01 u c £ o u 3 O* 3 cr o> cc so X 5 re Ol c re In X £ 0) 3 3 s4n in 3 O 1/1 s DT3 Ol HH re -3 3 O re u O) 3 O 9 X £ o> 01 in 3 • — in re 3 re •*< 6 c O) >h X o) p X >-,~ o -a ^ 3 O) 3 h re re o> _c 3 x 0> s01 | .9 > x 3 53 O a U > t x Ol ifl J(0 in U re Ol bO Ol -g X u X re u 3 3 H in re Ol 3 Ol bO S fa o T3 3 re i — i re 3 bD > O T3 PS u en ^. 3 O u_ H Ih T3 i — re O) X _ra o Sre cu X i— m CD Ol o> X u 5 Oi u 3 0> sO) J— 0> h — O u— 3 O X (H Oi sre 13 3 re ca 3 O O Oi X in — re .„ re -3 3 Ol O) X X X ^ w £ re Ol *> re u n 3 Ol **X oi 3 o 3 • — 4H o h Du U5 01 TJ Oi ss01 V Ih bO' re -a 3 01 s3 bD D s0 Oi 3 O 3 O Ih sre o X (— CO T3 41 Oi X re o> X H Ol > O X re

PAGE 121

102 Construction of RPVB5R-T RPVB5R-T, which produces a B5R protein lacking the B5R 3' putative membrane anchor region, was constructed using the scheme outlined in Figure 10. A fragment of the B5R ORF, from nucleotide 563 to nucleotide 1004, was amplified by PCR using primers RM277 (GGGGCTCGAGCCGTCTCTATCTAATGC) and primer RM278 (CGATACCGTCGACCTCGAATTTCTAACGATTCTATTTC). This amplified fragment contains the sequence immediately upstream of the membrane anchor coding sequence and in addition, replaces the codon for amino acid 275 with a stop codon (TAA) contained in primer RM278. The relative location of this newly inserted stop codon is indicated in Figure 9. After ligation into the pCRl plasmid (Invitrogen), the insert was sequenced and a clone free of unintended mutations was selected for further construction. The cloned B5R gene fragment was then excised by digestion with endonuclease EcoRI, and the 250 nucleotide B5R fragment was ligated into the EcoRI site of the plasmid pBluescript-gptR so that the p7.5 promoter and the gpt ORF were placed immediately downstream of the introduced stop codon. The result was a plasmid, called pBS-gptR-B5RL, in which the 3' end fragment of the B5R ORF is 5' of the gpt cassette (see Figure 10). A second PCR product beginning one nucleotide after the B5R stop codon and extending approximately 120 bases into the B6R ORF was generated using primers

PAGE 122

Figure 10. Scheme for construction of RPVB5R-T. The regions immediately upstream of the membrane anchor and downstream of the B5R ORF were amplified using primer pairs RM277/RM278 and RM279/RM299, respectively, to generate the PCR products called B5R-L and B5R-R. These fragments of the B5R ORF were cloned into pCRl, sequenced and then subcloned into pBS-gptR to give the plasmid pBS-gptR-B5R-T which was used in the construction of RPVB5R-T. Open boxes signify ORFs or fragments of ORFs which are identified by the type within the box. The figure at the very top of the page represents the B5R/B6R region of the RPVgenome. The hatched box represents the membrane anchor region of the B5R ORF. The small, black arrows above the figure represent the location of the primers used in construction. The number '276' refers to the corresponding amino acid position and denotes the site where a stop codon was introduced during construction of the mutant. The type enclosed within the ovals indicates the name of that particular plasmid. The speckled box labeled p7.5 denotes the p7.5 promoter. GFT guanylribosylphosphotransferase ORF; B5R-L, 250 bp fragment of the B5R ORF from immediately upstream of the membrane anchor sequence; B5R-L, 250 bp fragment of the B5R ORF from immediately downstream of the B5R ORF.

PAGE 123

RM277 RM278 RM279 RM299 B5R B5R-L In CN Ligate into pCRl i B5R-L pCRl-B5R-L B5R-L j Sequence and Subclone ^oSTOP www B5R-R pCRl-B5R-R id _SP B5R-R pBS-gptR-B5R-T Transfect STOP

PAGE 124

105 RM279 (GATCCACTAGTTCTAGACCGTAAATATAAATCCG) and RM299 (CCGATAAATGTTCTACGCGAGCTCCCCC). The PCR product was ligated into the plasmid pCRl, sequenced and a clone called pCRl-B5R-R free from mutations was isolated for further construction. The PCR-generated insert was removed by digestion of pCRl-B5R-R with the endonucleases Xbal and Sad and ligated into the corresponding sites in pBS-gptR-B5RL so that the B5R/B6R intra-cistronic region began immediately 3' of the stop cod on of the gpt ORR The resulting plasmid, pBS-gptB5R-T contains a B5R ORF (referred to as B5R-T) which has a stop codon at amino acid 275 and in which the coding sequence for the membrane anchor region has been replaced by the p7.5 driven gpt ORF. The B5R-T ORF was transferred into RPV by transfection of linearized pBS-gptB5R-T into RPV infected CV-1 cells and recombinant viruses were enriched by the addition of selection media which contained mycophenolic acid (2.5 /ig/ml), 250 hypoxanthine (/ig/ml), and hypoxanthine (15 /ig/ml). Virions resulting from this transfection were plated on RK-13 cells and mutants were identified and isolated by plaque lift hybridization using a probe for the gpt ORF as was described for the construction of RPVAB5R. Construction of RPV mutants that fail to secrete B5R The RPVB5R:X mutants, each of which contains a unique deletion of between 12 and 36 nucleotides in the presumed cleavage site region of the

PAGE 125

106 B5R coding sequence, were constructed using a recombinant PCR method described by Higuchi. The general locations of these deletions in relation to the overall protein are indicated in Figure 9, while the specific annealing sites of the primers as well as the specific amino acids deleted from the protein are indicated in Figure 11. To construct each of the mutant B5R ORFs, two PCR products comprising the sequences immediately upstream and downstream of the deleted sequence and having overlapping 3' and 5' ends, respectively, were generated. The upstream arms were generated using the primer RM260 (AACTGCAGAAGCTGGATCCTGGAG) which hybridizes to a sequence 211 nucleotides upstream of the first codon of the B5R ORF (to include the B5R promoter) along with one of the following primers: RM347 (GCGAACACATATTGGGAGTACGGG), RM349 (TACTGGATCAAATTCTTCGTT), RM351 (TGTCTCATCGTCGGGACC), RM355 (ACCATCGTCGACTGGATCAAATTC), RM365 (TTTGCTCAGATCTGTCTCATCG) all of which hybridize to sequences within the presumed cleavage site region. The downstream arms were generated using primer RM299 (CCGATAAATGTTCTACGCGAGCTCCCCC) which hybridizes to a sequence approximately 120 nucleotides downstream of the B6R initiation codon, along with one of the following primers: RM348 (CCCAATATGTGTTCGCGACGATGAGACAGATCTG), RM350 (GAAGAATTTGATCCAGTACTCTCGAAAGACGTTGTA), RM352 (CCGGACGATGAGACAGAACAAGAAATAGAATCGTTA), RM356 (CCAGTCGACGATGGTGA-

PAGE 126

T5 "0 S3 O c E a 01 X g '3 £ tO =3 c £ ea M to to • 0) cj ^ to u • — Oh 0> D 2 £ Oh 01 0) >-i a> S S 3 3 s CO > 3l o fa x o X C r ^ .2 jS to T3 at tJ S .3£ c a W a> h 03 o> is ^ (0 to c ai tO a> sD_ at sto sT3 at 3 at C* IT* u a at X -c c ai at X O 3 O +> rfl u O i— I o ^DS LO CQ E o -c So X c 3 0 c to XI £ a; at to a u v. a o a -c H c to £ x CQ > T3 G a> to .§> a -2 I! |ir tn c £ 0 .£ U o> 01 .f 3 >, — 43 T3 a = i to X £ a B e !sa> to 3 oi cr" Jc a 33 to t B T3 O to ni at X, h to & a T3 £ c .5 o Cl, to at 01 h r5 g P 8 0) ai c s(0 a, O) pfl C3 C I -c -o o t3 C 01 in 'S a £ x o *-o I I C to £ c x to 8 g ai H c X Erj2 01 i3 60 0 at *01 — 3 O jj ai C to -O O s £^ >to I o 3 .£ £ (0 an a H a, *x to *.£ 'I to 5 4— 01 01 Dm to at oi — e c to -3 oi cr h at 4— T3 a> T3 X C E 3 3 at 3 3 3 ^ ai rt > a>
PAGE 128

109 TCTGAGCAAACTCTCG), RM366 (GAGACAGATCTGAGCAAAGTTGTACAATATGAACAA) all of which hybridize to sequences within the presumed cleavage site region (see Figure 11). Pairs of arms containing overlapping complementary 5' and 3' ends were used to generate the full-length mutant B5R ORFs by recombinant PCR using the outside primers RM260 and RM299. The PCR products were digested with the restriction endonucleases Kpnl and Psfl and ligated into the Kpnl and Psfl sites of pBluescript after which the inserts were sequenced and clones that were free of unintended mutations were isolated. The inserts were removed from pBluescript by digestion with Kpnl and Psfl and were ligated into the Kpnl and Psfl sites of the plasmid pBluescript:JF which contains the pBluescript polylinker sequence within the vaccinia thymidine kinase (TK) ORF. The result was a series of plasmids (pBluescript:JFB5R:X) which contained mutant copies of the B5R ORF flanked on both sides by sequences from the vaccinia TK gene. Theses B5R ORFs were transferred into the TK gene of RPVAB5R by transfection of individual pBluescript:JFB5R:X plasmids into RK-13 cells that had been infected with RPVAB5R at an m.o.i. of 0.1. After CPE was observed, the cells were harvested and TK(-) recombinants were selected by titration of the virus on Rat-2 cells (which are unable to produce thymidine kinase) in the presence of bromodeoxyuridine (100 /ig/ml). Individual plaques were picked from the monolayer and the selection procedure was repeated twice. After the round of selection, the virus was titrated on CV-1 cells in the presence of

PAGE 129

110 mycophenolic acid (2.5 jig/ ml), 250 hypoxanthine (/ig/ml), and hypoxanthine (15 fig/ ml) and individual plaques were picked from the monolayer for the production of virus stocks. Protein Techniques Antisera Polyclonal anti-B5R sera was most skillfully produced by Dr. LuisaMartinez Pomares and Rita Stern as follows. The B5R gene was amplified from RPV genomic DNA using primers RM151R and RM 152R, described above. The PCR product was cloned into the Sail and Hindlll sites of pMALCR-1 (New England Biolabs) to produce the plasmid pMALCR-ps/hr in which the B5R ORF is joined to a maltose binding protein. This plasmid was then transformed into the TBI strain of E. coli as described above. Following induction of the transformed bacteria with IFTG (isopropyl-G-Dthiogalactopyranyside)(GibcoBRL), an abundant 84 kDa MalE-B5R fusion protein was expressed. The fusion protein was purified on an amylose resin column and was used for the immunization of female New Zealand White rabbits following standard protocols (39). The monoclonal antibody C-B5R, which recognizes a peptide corresponding to the sequence immediately upstream of the membrane

PAGE 130

Ill anchor domain of the B5R protein, was provided for use in western blot analysis as a generous gift of Dr. Lendon Payne. Immunoprecipitation of B5R Cells grown in a 35 mm dish were infected with virus at a high m.o.i. (i.e. 10) in serum-free MEM and incubated as described in the section on cell culture. At 6 hours post infection, the medium was removed, and the cells washed one time with cysteine and methionine free media (Sigma). Incubation was then continued for 30 minutes in 1 ml of cysteine and methionine-free media after which 3 5 S-Trans label (ICN Biochemicals, Inc.) was added to a final concentration of 50 jiCi/ml. Following a 30 minute incubation at 37C, the medium was removed and replaced with one ml of cysteine and methionine-free medium containing 50 fiCi/ml of 35STranslabel and 10% (v/v) MEM. Following an overnight incubation, the medium was harvested, clarified by centrifugation at 13,000 RPM in a Beckman Microfuge (approx. 12,000 x G) for 30 minutes and a 150 jul aliquot of the resulting virus-free supernatant was adjusted to 1% (v/v) Nonidet P-40, 1% (w/v) sodium deoxycholate 0.1% SDS (w/v) (Fisher), 150 mM NaCl (Fisher), 10 mM Tris-HCl [pH 7.4] (Sigma), 2 /ig/ml of aprotinin (Boehringer Manheim), 0.75% (w/v) bovine serum albumin (BSA) and 0.015% (w/v) NaN 3 (Sigma), by the addition of 50 /il of a 4x stock of RIPA buffer. To this was added 5 \i\ of undiluted polyclonal anti-B5R rabbit sera, and the sample

PAGE 131

112 was incubated at 4C for 3 hours prior to further processing for analysis of the immunoprecipitated proteins by SDS-PAGE. The infected cells were harvested by the addition of 0.5 ml of RIPA buffer (1% (v/v) Nonidet P-40, 0.5% (w/v) sodium deoxycholate, 0.1% (w/v) SDS, 150 mM NaCl, 50 mM Tris-HCl fpH 7.4]) containing 2 /ig/ml of aprotinin (Boehringer Manheim) to each dish followed by a 5 minute incubation at room temperature. The resulting cell lysate was frozen and thawed, incubated at 37 C for one hour and then centrifuged at 13,000 RPM in a Beckman Microfuge (approx. 12,000 x G) for 30 minutes. A 75 \i\ aliquot of the clarified supernatant was mixed with 125 \i\ of dissociation buffer (RIPA buffer containing .1% (w/v) BSA and 0.02% (w/v) NaN 3 ) and 5\i\ of the B5R anti-serum. The sample was then incubated at 4C for 3 hours. Following the three hour incubation of the cell and supernatant fractions, 50 /il of protein A-Sepharose (a 1:1 slurry of PBS-0.02% (w/v) NaN 3 and Sepharose (a mixture of protein A-Sepharose and Sepharose CL4 B (Pharmacia) at a ratio of 1:2)) was added to each sample and incubation was continued at 4C for one hour while the samples were continuously mixed. Following this incubation, the protein A-sepharose was pelleted by centrifugation at 12,000 x G for 1 minute and the supernatant was discarded. The pellet was washed three times with dissociation buffer (RIPA buffer plus BSA (1 mg/ml), with a 30 second centrifugation at 12,000 x G following each wash, and the final pellet was resuspended in 60 /il of IX Laemmli sample

PAGE 132

113 buffer (0.625 MTris-HCl [pH 6.8] (Fisher), 10% (v/v) glycerol (GibcoBRL), 2% (w/v) SDS (Fisher), 0.7 M 2-mercaptoethanol (Sigma)). The sample was mixed, boiled for 3 minutes, centrifuged at 12,000 x G for 1 minute to pellet the sepharose and the supernatant analyzed by SDS-PAGE. Polyacrylamide Gel Electrophoresis (PAGE) and Autoradiography Protein samples solubilized in Laemmli sample buffer were heated in a boiling water bath for three minutes and separated in 10% (w/ v) SDSpolyacrylamide slab gels according to the method of Laemmli (125). Gels were run at 3.75 V/cm. Following electrophoresis, gels to be used for autoradiography were soaked in 2 gelvolumes of fixing solution ( 15% (v/v) acetic acid, 25% (v/v) methanol in distilled water ) (Fisher) for 15 minutes followed by two 30 minute incubations in 2 volumes of DMSO. The gel was then infused with PPO (2,5-diphenyl-oxazole)(Sigma)by incubation in 13% (w / v) PPO in DMSO for 30 minutes after which time the DMSO was removed by a one hour incubation in continuously running water. Following removal of the DMSO, the gel was dried using a Savant slab gel dryer, wrapped in plastic film, placed against Kodak X-Omat film and stored at -70C. Western Blot Analysis Following separation of viral proteins by electrophoresis in a 10% (w/ v) SDS-polyacrylamide slab gel, the gel was soaked in two gel-volumes of

PAGE 133

transfer buffer (0.025 M Tris, 0.192 M glycine, 20% (v/ v) MeOH) after which the proteins were transferred to nitrocellulose following the method of Towbin (125). Briefly, the gel was held tightly against a Hybond N nitrocellulose membrane (Amersham) using a BioRad western blot gel holder cassette, and the proteins were electrophoresed from the gel to the membrane in transfer buffer at 250 mA for 3 hours using a BioRad Trans-blot Cell (BioRad). The blot was incubated for one hour at room temperature in PBS containing 0.05% (v/v) Tween-20 (polyoxyethylene-sorbitan monolaurate) (Sigma) and 5 (w/ v) % bovine serum albumin (BSA) (Sigma) followed by an overnight incubation at room temperature in PBS /Tween20/ BSA containing B5R anti-serum at a dilution of 1:5000. The PBS/Tween20 /BSA solution containing sera was removed and four washes ( 1 x 15 minutes, 3x5 minutes) were done using PBS /Tween-20 after which the blot was incubated for one hour at room temperature in PBS /Tween-20 /BSA containing alkaline phosphatase linked goat, anti-rabbit serum at a dilution of 1:5000. Following this incubation, the blot was washed five times (1 x 20 minutes, 4x5 minutes) in PBS /Tween-20, and detection of the antibody complexes was accomplished using the Enhanced Chemiluminescence Kit (Amersham) according to the manufacturer specifications.

PAGE 134

115 Glycosidase Treatment of Proteins The B5R and the B5R-T proteins were labeled and immunoprecipitated following the procedures described above; however after the final wash step, Laemmli sample buffer was not added to the final pellet. Instead the immunoprecipitated viral protein, complexed to the sepharose/protein-A bead, was resuspended in 25 \A of 0.1 M 2-mercaptoethanol/ 1% (w/ v) SDS and heated to boiling for 5 minutes. The suspension was adjusted to 0.125 M Tris-HCl [pH8.0] (Fisher), 0.1 M 2-mercaptoethanol (Sigma), 0.25% (w/v) SDS (Fisher), 1.25% (v/v) NP-40, 5 jug/ ml aprotinin (Sigma) in a final volume of 100 fih One unit of peptide-N-glycosidase-F (Boehringer Manheim) was added after which the sample was incubated at 37C for 3 hours. The suspension was then mixed with 25 ^il of 5X Laemmli sample buffer, heated to boiling for five minutes and following a 1 minute centrifugation at 12,000 x G, the proteins present in the supernatant were analyzed by SDSPAGE. Sodium Carbonate Treatment of Virus CsCl gradient-purified EEV or IEV particles were incubated in 50 /il of either 100 mM Tris-HCl [pH7.6] or 100 mM sodium carbonate (Na 2 C0 3 )[pH 11.5] for 30 minutes on ice with mixing every 10 minutes to remove nonintegral membrane proteins. Following centrifugation at 13,000 RPM in a Beckman Microfuge (approx. 12,000 x G) for 30 minutes to pellet the virus, the

PAGE 135

116 supernatant was removed and mixed with 12 /d of 5X Laemmli buffer. The viral pellet was resuspended in lx Laemmli buffer, and the samples were subject to SDS-PAGE and western blot analysis, as described above.

PAGE 136

CHAPTER III THE INVOLVEMENT OF THE RPV B5R PROTEIN IN HOST RANGE, VIRULENCE AND INFLAMMATION Introduction The poxviruses comprise an extensive family of large, cytoplasmic, DNA viruses. Several members of this family are known to cause a severe and often fatal disease upon infection of their natural host, a dramatic example being variola virus, the causative agent of human smallpox. A second virulent member of the orthopox genus is rabbitpox virus. This virus was first isolated from infected rabbits housed in the Rockefeller Institute animal breeding facility during a spontaneous, lethal epidemic that occurred in 1932 (84,86). Infection of rabbits with RPV, which is believed to have evolved from a neurovirulent strain of vaccinia virus, results in a disseminated and generally lethal infection producing symptomology and pathology similar to those seen in smallpox. The production of a lethal infection combined with the availability of a convenient host make this virus attractive for use in studying the contribution of particular proteins to poxvirus-related pathogenesis. 117

PAGE 137

118 Several animal model systems have been developed to study the involvement of individual viral proteins in viral pathogenesis, the most common systems employed being intranasal infection of mice, scarification of rabbits (or guinea pigs) and infections of the chorioallantoic membrane (CAM) of the 11-day-old chicken embryo. Intranasal infection of mice induces a viral pneumonia resulting in generalized illness, the severity of which increases with the infectious dose. Weight loss is used as convenient and objective measure with which to measure the progress of the infection since it has been shown to correlate with the severity of the disease (198). If the infectious dose is high enough, death of the animal occurs in 5-7 days due to a virally induced pneumonia. Unlike infections of the rabbit (see below), in the mouse model, there is no need for spread of the virus from the initial site of infection, in this case the lung, and death results from damage to the lung as a consequence of local growth of the virus (198). In contrast, the death of rabbits following infection with rabbitpox virus results from systemic shock as a consequence of the widespread dissemination and damage caused by replication of the virus (26). The evaluation of pock color on the CAM is also a well established system that has been invaluable for studying the effects of viral proteins on the host inflammatory response. This system was used extensively in the past for isolation and characterization of the CPV crmA protein, which inhibits the IL-lfi converting enzyme and has been implicated in inhibition of

PAGE 138

119 the host inflammatory response (142,158). In contrast to the red, hemorrhagic pocks of wild-type CPV, viruses unable to produce the crmA protein form white, inflammatory pocks on the CAM. These white pocks have been shown to be impacted with heterophils presumably as a direct result of the mutant's inability to block an IL-mediated inflammatory response. Support for this idea has come from work demonstrating phenotypic reversion of the white pocks to red pocks when infections are done in the presence of the antiinflammatory agent dexamethasone (70). The establishment of an infection in any new host is a major challenge facing the infecting virus. Not only must the virus be able to replicate in a wide variety of cell types within the host, but it must also be able to spread to uninfected cells in the presence of an increasingly hostile host response. Spread of virus from the initial site of infection both in cell culture and in the animal is believed to be mediated by the enveloped form of the virus, one of two forms of infectious virus produced during the infection (9,27,146). The enveloped virions (CEV and EEV) differ from intracellular mature virus (IMV) in that they are wrapped with either one or two extra membranes, and, as a result, the enveloped virions contain polypeptides not found in IMV. Deletion of the genes encoding these envelope proteins frequently results in a selective defect in the synthesis of enveloped virus but not of IMV. Such mutants produce normal amounts of IMV but produce minuscule plaques in cell culture and are severely attenuated in animals (20,58,177).

PAGE 139

As was mentioned above, spread of virus in the animal must occur in the presence of an ever-increasing host immune response. The severity of the infection produced by poxviruses demonstrates that these viruses have been quite successful at evolving methods for evading or modifying the host response to the infection. In recent years it has come to be appreciated that the poxviruses produce several proteins which, in vitro, are capable of neutralizing critical components of the host immune system (6,36,121). Several of these proteins have been shown to be crucial for full expression of virus virulence. Many of these virus-encoded virulence proteins either share homology with and /or interact directly with regulatory components of the immune system. Examples of such proteins include a secreted IL-lfi receptor, a secreted TNF receptor, a secreted IFN-gamma receptor, a secreted complement factor C4b binding protein and the serine protease inhibitor that inhibits the IL-16 converting enzyme (5,97,99,110,111,142,158,182,183,187). The B5R protein of RPV is yet another poxvirus protein for which sequence homology with components of the host immune system has been observed (57,100,218). The 45 kDa glycoprotein is produced late during the infection and is found as a component of CEV and EEV but not of IMV. In addition, the B5R protein is found in the supernatant of infected cells as a smaller secreted protein of around 38 kDa that is derived by cleavage of the larger cellular form (119). Examination of the predicted amino acid sequence of the B5R protein reveals four short consensus repeat (SCR) domains of

PAGE 140

121 approximately 60 amino acids each making B5R a member of the SCR superfamily. These domains are illustrated schematically in Figure 6 while an alignment of the SCR domains of B5R with domains from several other members of this superfamily is shown in Figure 7. Proteins found within the SCR superfamily typically possess regulatory activity, the majority of these proteins being involved in regulation of the complement cascade. Database analysis has shown that B5R shares regions of homology with factor H and the complement C4b binding protein both of which are proteins that act to inhibit complement activation. Since one consequence of complement activation is the generation of an inflammatory response, these proteins may be viewed as negative regulators of inflammation. Based on the shared homologies with complement control proteins, it was predicted that the B5R protein of RPV would, in addition to its role as a component of EEV, act as a virulence factor by modulation of the host inflammatory response. This hypothesis was tested by construction of a RPV B5R null mutant (called RPVAB5R) and analysis of this mutant in cell culture and in animals. The data presented within this chapter shows that while RPVAB5R is indeed attenuated in animals, the attenuation observed for the B5R null mutant does not result from an enhanced inflammatory response. Instead, the argument is made that attenuation of RPVAB5R results from a defect in the morphogenesis of the enveloped forms of virus a phenotype that can be demonstrated in cell culture. It is likely that this defect also occurs

PAGE 141

122 within the tissues of the animal resulting in failure of the virus to spread leading to low viral titers and attenuation. Results Construction of the B5R Null Mutant, RPVAB5R A detailed description of the construction of RPVAB5R is provided in the Materials and Methods chapter. Briefly, the gene encoding the £. coli. guanyl-ribosylphosphotransferase (gpt) protein, under control of the poxvirus p7.5 promoter (collectively referred to as the gpt cassette), was inserted within a cloned copy of the B5R open reading frame at a unique EcoRV restriction site located within the N-terminal region of the protein (see Figure 9). The gpt gene confers on recombinant viruses which carry it the ability to replicate in the presence of the purine biosynthesis inhibitor mycophenolic acid (MPA). This construction scheme also placed a stop codon at the B5R-gpt junction 152 base pairs into the B5R gene ensuring interruption of B5R translation. The B5R-gpt construct was then used to replace the wild-type copy of the gene in the RPV genome by homologous recombination and recombinant viruses that contained the gpt cassette were purified (see Materials and Methods). Construction of RPVAB5R was initially confirmed by southern blot analysis the result of which is shown in Figure 12. EcoRI

PAGE 142

Figure 12. Construction and characterization of RPVAB5R. (A) Schematic representation of the structure of the B5R gene in RPV and in RPVAB5R in which the B5R ORF is disrupted by the Ecogpt cassette. The open box represents the B5R ORF; the hatched box represents the Ecogpt cassette. RL EcoRI; RV, EcoRV. (B) Southern blot analysis of RPV and RPVAB5R genomic DNAs. The DNAs were digested with EcoRI and probed with radiolabeled B5R gene DNA.

PAGE 143

124

PAGE 144

125 restriction fragments from either RPV and RPVAB5R genomic DNA were separated by electrophoresis in an agarose gel, transferred to a nitrocellulose membrane and hybridized to radiolabeled B5R DNA. The B5R gene contains a single internal EcoRI site, located approximately 150 bp into the ORF and as expected, Southern analysis of DNA from RPV resulted in two fragments of 13.0 and 7.5 kbp (lane 1). Incorporation of the 2.0 kbp gpt cassette, which is free of EcoRI sites, should result in a 2.0 kbp shift in the 13.0 kbp EcoRI fragment of the B5R gene. This is exactly what was observed; the 13.0 B5R fragment was shifted to 15.0 kbp in DNA purified from RPVAB5R indicating insertion of the gpt cassette into the B5R gene (lane 2). Failure of RPVAB5R to produce the B5R protein was confirmed by western blot analysis. Figure 13 shows the results of an immunoblot analysis of RPV and RPVAB5R infected cell extracts and clarified supernatants using polyclonal anti-B5R sera. As expected, the anti-B5R antisera detected a 45 kDa protein and a 38 kDa protein in RPV infected cell extracts (lane 1) and supernatant from these cells (lane 2), respectively. The 38 kDa protein represents the smaller, secreted form of the B5R protein. In contrast to fractions from RPV infected cells, no protein was detected in either cellular or supernatant fractions (lane 3 and 4, respectively) from RPVAB5R infected cells.

PAGE 145

Figure 13. Analysis of RPV and RPVAB5R infected cell extracts and supernatants for the presence of the B5R protein. Western blot analysis was performed on cell (lanes 1 and 3) and supernatant (lanes 2 and 4) fractions from RK-13 cells infected at a high m.o.i. with either RPV (lanes 1 and 2) or RPVAB5R (lanes 3 and 4). Samples were prepared as described in Materials and Methods.

PAGE 146

50.3 kDa 33.3 kDa 28.5 kDa

PAGE 147

128 Host Range and Plaque Morphology of RPVAB5R The original work reporting the isolation of the vaccinia (Lister strain) ps/hr gene (the B5R homolog in that virus) showed that the plaques produced on RK-13 cells by viruses containing a naturally occurring frameshift mutation in this gene were smaller in size than those produced by RPV (195). In addition, these mutants were shown to be unable to plaque on Vero cells. Since the RPV B5R gene is not one-hundred percent homologous with the ps/hr protein of VV(Lister) and since proteins may behave differently in different virus backgrounds, it was of interest to see if interruption of the RPV B5R gene resulted in similar host range and plaque size alterations. To test this hypothesis, the ability of RPVAB5R to form plaques was compared with that of RPV on several cell lines including RK-13 cells, Vero cells and primary chicken embryo fibroblasts (CEFs). The results, shown in Figure 14, are in agreement with those reported for the VV B5R mutant. Not only did RPVAB5R fail to form plaques on Vero cells, it also failed to form plaques on PK-15 cells, CEFs and QT-6 (quail carcinoma) cells. In the remaining cell lines tested (RK-13, CV-1 and Rat-2 cells), RPVAB5R was able to form plaques, however, these plaques were consistently smaller than those formed by RPV in these same cell lines, again in agreement with what was reported for the W(Lister) ps/hr (B5R homolog) mutant.

PAGE 148

Figure 14. Plaque formation of RPV and RPVAB5R on various cell lines. Cell were infected with either RPV or RPVAB5R, and plaques were allowed to develop for 96 hours at 37C after which the monolayers were stained with crystal violet (0.01 % (w/v) in 5% (v/v) ethanol).

PAGE 149

130 Uninfected RPV RPVAB5R

PAGE 150

131 The B5R Protein is Necessary for Enveloped Virus Production There are several possible explanations for the failure of RPVAB5R to plaque on CEFs. One possibility is that the mutation results in a complete inhibition of all virus growth in the non-permissive cells. However, this was ruled out since although RPVAB5R was unable to form plaques on CEFs, the pattern of viral protein synthesis in RPVAB5R infected CEFs appeared normal (Martinez-Pomares, L., unpublished data). Another possibility is that loss of B5R function results in a defect in the basic pathway of virion morphogenesis. Still a third possibility was that virion production was normal, however, the virions produced were non-infectious. To determine if B5R function is linked to virus morphogenesis, RPVAB5R was examined for its ability to produce virus particles in both permissive and non-permissive cells. RK-13 cells were chosen for use as permissive cells since the plaques formed by the mutant were largest in this cell line, suggesting that deletion of B5R had only minor effects in these cells. Although RPVAB5R failed to form plaques on several cell lines, CEF cells were chosen as the non-permissive cell line. Since the plaques formed by RPV are intrinsically larger in CEF cells than in any of the other non-permissive cell lines, it was thought that the failure of RPVAB5R to plaque at all on CEF cells indicates that any defect in virus particle production would be greatest in these cells. Cells were infected at a high m.o.i., and the newly synthesized viral DNA was labeled with [ 3 H]-

PAGE 151

132 thymidine as described in Materials and Methods. At 48 hours post-infection, the virus present in the supernatant (EEV) and the cellular associated virus (IEV and IMV) were analyzed by separation of the various forms of virus on CsCl gradients. The level of virus present was estimated by quantitation of the amount of radioactive DNA present in fractions from these gradients. As can be seen in Figure 15, in the permissive cell line (RK-13 cells), RPV and RPVAB5R produced identical amounts of IMV and EEV. However, the mutant appeared to be somewhat impaired in its ability to produce IEV, judging from the difference in the height of the peaks in fraction 9 which represents IEV. Overall while there was approximately a twenty-five percent reduction in total virus yield, which correlated with the slightly decreased plaque size in this cell line, virus production was relatively normal. The picture was dramatically different in non-permissive cells (CEFs). The amount of IMV produced by RPVAB5R was roughly half that produced by RPV. More significantly, while RPV produced large amounts of enveloped virus, virtually no enveloped virus, either EEV or IEV, was formed in RPVAB5R-infected cells. Since enveloped virus is needed for cell to cell spread (21), the inability of RPVAB5R to form plaques on this cell line could be attributed to this defect in enveloped virus production. Behavior of RPVAB5R on the CAM As previously described, the predicted amino acid sequence of RPV

PAGE 152

£ T3 CO 2d 13 w 6 o < .5 Q O 2 C tu 01 If) (3 O

PAGE 153

134

PAGE 154

135 B5R shares homology with several regulatory proteins of the complement cascade (57,100,119,195). These homologies suggest that the B5R protein may inhibit complement activation in a manner analogous to that of the viral C4b binding protein (99,110,111). Since one consequence of complement activation is the generation of an inflammatory response (80,210), it was hypothesized that the B5R protein may have a role in inhibiting inflammation. If this were so, then null mutants of B5R might be expected to behave like null mutants of crmA, a known inhibitor of inflammation, and generate white pocks on the CAM (142,158). The importance of complement as a part of the antiviral response of the host in this system has been demonstrated by work showing that complement depletion of the 11 day-old chicken embryo results in an altered pock morphology on the CAM (138). To examine the involvement of the B5R protein in the abrogation of host defenses, the behavior of RPVAB5R on the was evaluated on the CAM. As can be seen in Figure 16, in contrast to the red, hemorrhagic pocks produced by RPV, RPVAB5R produced smalkwhite, non-hemorrhagic pocks similar to the pocks produced by CPV crmA mutants. This result demonstrates that the B5R protein is necessary for red, hemorrhagic pock formation and suggests that B5R, like crmA, might be involved in the modulation of inflammation. Attenuation of RPVAB5R in Rabbits The data obtained on the CAM suggested that the B5R protein has a

PAGE 155

u o> o 15 2 O o 2.S Jr. OJ 0> I J O aj i-i o> 0> II Si T2 < u u to T5 O a c 4-1 Si ^ ^* as 8..S C2 T3 If) 0> < S 2 -o si VP S Jo c o • — u 01 to o Si K N to 3 O

PAGE 157

138 role in modulation of the host inflammatory response, and as a consequence, might have a role in virulence of the virus. To gain a better understanding of the importance of this protein in virulence during infection of an animal, the ability of RPV and RPVAB5R to cause disease in rabbits was compared. Rabbits, which are considered to be the natural host for RPV, are exquisitely sensitive to infection with RPV, and a lethal infection can usually be initiated by intradermal injection of a single pfu. The progress of the infection in these animals can easily be followed by monitoring the increase in body temperature as well as by noting the secondary signs of infection which develop such as nasal discharge, dyspnea and the appearance of secondary / satellite lesions which are indicative of viremia. The data in Figure 17 compare the febrile response resulting from intradermal injection of rabbits with either RPV or RPVA B5R, while the data contained in Table 2 compare overt clinical symptoms. All animals infected with RPV developed significant induration at the site of inoculation beginning two-days postinfection, which generally increased throughout the course of the infection. RPV infected animals (Figure 17a) also showed a sharp increase in body temperature (pyrexia) beginning approximately two days post-inoculation and persisting until death or, in the twenty-five to fifty percent of the animals which survived, until approximately 10 days post-inoculation. In contrast, rabbits infected with RPVAB5R (Figure 17b ) showed no significant increase in body temperature at any time during the infection. In

PAGE 158

o o o 0> t/1 <'i 1 § £ -c <4H 0> s= > m ~ a o> in £ 2 01 — 4-1 3 (0 cx £ H X) X) H T3 01 u 01 CO T3 C > Oh 3 c o 3 < .24 > s (A R3 Oi so> > 2 O) 33 to £ OJ g 01 u 01 £ E o> 1 I O o C> a, 10 o H 3 60 3 to 01 Q iS .S

PAGE 159

140 ootcor^vom^cocNHo HOOOOOOOOOO i-li-li-IHHHr-IHHr-IH (d 0 ) dm! tb^osh o<7toor>>voin^iro(>iHo HOOOOOOOOOO HHHHHHHHHHH (d Q ) duiaj; t^ooh

PAGE 160

>> 6 2 S < >h .S a ^ 43 4-1 • 4-> re wre CJ 4-1 u c 3 j2 T3 •h re ty 01 9 o ^ a CO Q) 15 *I-i tO 3 S co c f3 O — I X £ c 0. co 4-1 re X hh c > x ™ 5 ^ x" x> t3 -g (8 m 0 So C^S 4_ 0 re .2 x! .a 01 (D C Mh 2 W 4-. .a in a S<-6 6&fl c Oh CD re 73 c re OJ X to tU 4-> U T5 C 6 CO H + < C o 'X u e .3 X 0) ^ H CO £ o 4-1 D01 o -4H o X X Oh So o > 0) X cn re Ej X T3 CD „ o 0 u 01 X bC 01 c CO 5 ? 0) 3 3 (5 s bp -g £ § W oi X X pa 141 re 01 D > o> u 0 cn u o> 01 _j C/3 01 bC re X H O S 01 X re o c O. CO >> D 01 bo re re w X re u Z .2 D 73 3 )H <4H Oh o o ID > Ph CO 3 <4H Oh O o ID > Ph 3 MH Oh O in > Oh O, CO 3 MH Oh O ID > 2 3 MH OO o ID Oh ID pa < > Ph EC po 3 MH Oh O o ID 0ID pa < > Ph 5 < 3 "*H Oh O ID ID pa < > Ph Pi pa 3 o, o ID o. ID pa < > Ph 2

PAGE 161

142 addition, there was no viremia and few if any symptoms of clinical disease in these animals (Table 2). Although there was some induration present in these animals at the site of inoculation, the indurated area was much less significant than that seen in the RPV infected animals (data not shown). In addition, unlike the induration observed in RPV infected animals which increased throughout the experiment, the induration generated by RPVAB5R began decreasing on day 6 post-inoculation and had usually resolved by day 12. The data was in agreement with that obtained from the CAM and showed that loss of B5R activity resulted in complete attenuation even in an animal that is normally extremely sensitive to infection by RPV. Histopathology of the Primary Lesions Produced by RPV and RPVAB5R The results obtained so far show that the B5R protein is indeed necessary for full virulence of RPV and further, they suggest that the protein might be involved at the level of suppression of the inflammatory response. If failure to suppress the inflammatory response was the cause of attenuation of RPVAB5R observed in rabbits, then one might expect to see an enhancement of inflammatory cell infiltrate within the lesions produced by RPVAB5R in rabbits. To test this hypothesis, a histological examination was made of the primary lesions resulting from intradermal infection of rabbits with 500 p.f.u. of either RPV or RPVAB5R. Samples of this tissue are pictured in Figure 18. while the major pathological features are summarized in Table

PAGE 162

143 3. Intradermal injection of PBS (Figure 18a) had little effect on the surrounding dermal tissue. There was virtually no cell necrosis and these sections were almost entirely free of inflammatory cells. Skin sections from RPV induced lesions however, showed major pathological changes (Figure 18b). The epidermis showed severe vacuolization together with necrosis that extended through the dermis and into the skeletal muscle bundles at the deep border of the sections. An extensive inflammatory cell infiltrate was observed throughout the lesion extending from the epidermis down to and into the skeletal muscle bundles. A considerable amount of hemorrhage, infiltrated by neutrophils and occasional macrophages, was present within the dermis, and the collagen in this area was reduced to granular fragments. In stark contrast to the pathology produced by RPV, injection of RPVAB5R failed to produce significant pathological alterations at the site of inoculation (fig. 18c). While the area showed slight induration, presumably due to a minor degree of edema, there was only occasional necrosis within this area and only a small number of neutrophils could be found within the dermis. Injection of larger numbers of virus (50000 p.f.u.) resulted in the development of a more severe pathology by both RPV and RPVAB5R; however, again, the damage and resulting inflammatory response were both greatly reduced in the RPVAB5R infected animals (Table 3). The data indicated that, contrary to what was observed on the CAM, the inflammatory response to RPVAB5R in rabbits was not enhanced in comparison to that

PAGE 163

01 o • o X 01 *i ^ — 2 £ g x 8 J! in C x C £ •l-H 2 < §1 cn 2 CD o .g CD w CQ to ^ o c c >> (ft • 2 10 T3 T3 g t5 s a c ^ 01 -c c • l—l to in B o — u a> to 0) Z £ T3 bJD Ol to § 6 'eg oi S B o £ a. M-l O 2 o.* £ DC 'S 00 o rH ^ 01 C 5 § toC O ri" *0 ra c o> > 1 -r in ai Js a, >• g to _o o Ol 01 c > 3 2 cd Oi 5c o > (0 > c .2 £ x Ol So .5 o O &H

PAGE 165

146 x: £ •I-H £ c H O re .5 i ** — T3 h ^ 0> >PS pa p. in CO Ph PS £> T3 u to tn >P c Oh re re o .a 00 > o cv 2 sn 6 •ts o O 4, 15 -n J8 X, in re t>o Cm C o .5 > o E 0) s01 re c s -Jo > re o a HrHli re S to 3 CQ Qh s rH X II Ui cc 0 Q O in II S < Z O H C J3 O O, 'x g s § O x a) CQ Ph -a oi n X c 2 Dm C o o -C a* u 2 ai 5 > o 13 | I •g -a .5 'a, x o> X 05 3 60 >-. C 0> 3) U 01 • (so o x> S o "a j C 3 T3 05 o § 2 .y so c o -3 -a t-t c c 0) 01 05 0 01 -a 511 Oh x — a a* £ -c ^ 01 0> u c 3 '£ 3 5 o > ^ o .S 05 M-i 9 O 0 P 01 ^ 0) u JC c "3 "3 c c o o I*. u u o o u m CQ > c tL x i 1= c E c C

PAGE 166

147 elicited by RPV but, in fact, was diminished in its intensity. These results could be explained by the reduced growth of the RPVAB5R virus. Attenuation of RPVAB5R in Mice The extreme sensitivity of rabbits to lethal infection with RPV, as well as their non-clonal nature, makes the determination of an LD50 for the virus in rabbits impractical. To more accurately quantitate the extent of RPVAB5R attenuation, the LD 5 q for this mutant was measured (by the method of Reed and Muench (159)) in a mouse intranasal infection model in which weight loss and the mortality rate are used to gauge the virulence and lethality of the virus (23). The results of this experiment are illustrated in Figure 19. Mice inoculated with RPV lost weight in a dose dependent manner, and death, resulting from the viral infection, occurred in animals infected with doses greater than 105 p f u yielding an LD 50 for RPV of 105.6 p f u (Fig. 19a). In contrast, inoculation of mice with identical doses of RPVAB5R resulted in either minimal or no weight loss (Fig. 19b). No deaths ever resulted from infection of mice with RPVAB5R, even following inoculation of the animals with doses as high as 108 pf u (data not shown). The LD50 for RPVA B5R can therefore be estimated to be at least one thousand fold greater than the LD 50 for RPV. As with the previous results, the data illustrates the importance of the B5R protein for full virulence.

PAGE 167

c ^ 3 •t-H *2 i in ^ to O T3 ~ g .5 co 01 o> u 01 c B co la C > 8 £ 3 2 m .S T3 > 3 o> 2 <*-< w O. 3 *0 ^ 3 O oi T-H I* 60 O 'S S u H 3 O QJ ^-5 J— .SP*2 > Uh '53 o>

PAGE 168


PAGE 169

150 Histopathology of RPV and RPVAB5R Infected Lung s The observation that RPVAB5R was severely attenuated in mice was consistent with the results obtained with this mutant in rabbits. To determine if the lungs of these animals exhibited the same lack of inflammation observed in the RPVAB5R infected rabbits tissues, a histological examination of lungs from infected mice was performed four days after intranasal infection with either RPV, RPVAB5R or RPVASPI-2 (crmA). Lungs from the ASPI-2 (crmA) infected mice were included as a positive control with which to make comparisons, since the SPI-2 protein has been shown to suppress the host inflammatory response in chickens. Sample sections from all lungs are shown in Figure 20 while the major histopathological features are summarized in Table 4. As expected, the lungs from PBS treated mice were normal and showed no evidence of necrosis, edema or inflammatory cell infiltrate (Figure 20a). However, infection with either RPV (20b) or RPVASPI-2 (Figure 20d) resulted in the development of periarteriolar, peribronchiolar and alveolar edema. In addition, necrosis of peribronchiolar epithelial cells was noted in these lungs. Both RPV and RPVASPI-2 infected lungs contained a mixed (neutrophilic and lymphocytic) inflammatory cell infiltrate; however, as was expected, the response was more severe in RPVASPI-2 infected lungs. In contrast to lungs from RPV and RPVASPI-2 infected animals, lungs from RPVAB5R infected mice showed no

PAGE 170

hi 0 U u re c 2 o "* o O c IT. {= .S £ v *I S .5 re O t o C o s pa g ^ u ^ o >, x 32 u 09 bC C 3 < > •— o LT) < re I E 2 8 0J ^ ** "S S B to ^ bO > c 2 LT) *: re

PAGE 171

152

PAGE 172

153 re E O e rO TO OJ 23 a OJ CJ 43 O <=. o c (X) CQ re £ c re ai o> 01 — 05 (B c to a; iOo T3 o C 3 GO o H c — 2 ~ s-2 "o -ri fl a> u o c c p u ta -53 T3 3 Oh h to 1? .2 c 0) £ 0_jD re re £ 0> -a o re Q O s 3 1 u O 5 0> e o. 0 O fis! 01 o> 01 T3 3 ^ '> '> 2 0> Ol "3 oop C C u s45 u c o sjD U Ol D_ "C c fl ai M Oh cn re re Ol -O CD re Ol 45 c o CO Oi IH O10 re 01 to o u Ol c 53 C — re re re ^ U Ol 43 SS 'B 5 J • •go .2 c N O o> £ Oh XI re H ^3 t/5 S o % g Sh re 2 I i fr_2 .53 o> 0> T3 -2 Ol 01 11 £ oi re ^ i K. 03 <1 > On pe5

PAGE 173

154 evidence of either edema or necrosis (Figure 20c). In addition, a mixed inflammatory cell infiltrate, similar to that observed in the RPV infected lungs, was noted in the RPVAB5R infected lungs. The degree of inflammatory cell infiltrate present in these lungs was mildly increased in comparison to RPV although it was not as severe as the infiltration observed in RPVASPI-2 infected tissue. These observations are consistent with the hypothesis that the severe attenuation of RPVAB5R in mice is not due to a large increase in the inflammatory response. Growth of RPVAB5R in the Mouse Lung The studies performed in animals up to this point suggested that the attenuation of RPVAB5R does not result from a dramatic increase in inflammation at the site of inoculation. In addition, the data from the cell culture experiments demonstrated that deletion of the B5R gene resulted in a cell line-dependent defect in extracellular enveloped virus production, resulting in an altered host range. If a similar host range restriction occurs in infected animals, the decreased viral growth could explain the severe attenuation observed. To measure the ability of RPVAB5R to grow in a localized environment within the animal, the ability of RPV and RPVAB5R to grow in the mouse lung was determined. Mice were infected intranasally with a lethal dose (106 p f u ) of either RPV or RPVASPI-2, and virus growth was followed by measuring the viral titers in the animals lung at various

PAGE 174

155 times throughout the course of the infection. As shown in Figure 21, the titers of virus in the lungs of RPV infected animals rapidly increased to yield a titer of 10 8 pfu/lung on day 5 post-infection. In contrast, the yield of virus in RPVAB5R infected animals was approximately 100 fold lower when compared to the wt RPV infected group, correlating with the diminished weight loss and morbidity observed in these animals. Although the level of RPVAB5R produced was much less than that of RPV, comparison of the final titer of RPVAB5R with that seen at time 0 in RPVAB5R infected animals shows a small but reproducible increase in titer, demonstrating that the mutant is able to grow, to some extent, in the mouse lung. Effect of an Anti-inflammatory Drug on Virus Replication in Mice The data obtained from infection of the chicken embryo demonstrated that RPVAB5R produced white pocks on the CAM. These white pocks appeared similar to the pocks produced by a SPI-2 null mutant of either RPV or CPV which have been shown to be the result of an enhanced host inflammatory response (70,141,142). However, no such inflammatory response was observed in infected rabbit tissue and only a slight increase in inflammation was observed in the mouse. To ascertain if the decreased yield of virus observed in the lung of RPVAB5R infected mice was due to the mild increase in the host inflammatory response, a determination was made of the ability of the B5R null mutant to replicate in animals in which the

PAGE 175

Figure 21. Growth of RPV and RPVAB5R in mouse lung. Mice (four per virus) were infected intranasally with 106 pf u 0 f either RPV (•) or RPVAB5R (O). At the times indicated, lungs were removed and the amount of virus present was measured by determination of titers on RK-13 cells.

PAGE 177

158 inflammatory response had been chemically suppressed. Suppression of the inflammatory response in these mice was achieved by treatment of the mice prior-to and throughout the course of the experiment twice-daily with dexamethasone, a known and potent, anti-inflammatory agent (see Materials and Methods). Suppression of inflammation was confirmed by measuring the reduction in phorbol ester-induced ear swelling (201) in dexamethasonetreated animals as is shown in Figure 22a. Mice treated with dexamethasone showed a dramatic decrease in inducible ear-swelling when compared to untreated mice indicating that the drug was acting as expected. In addition, inhibition of inflammatory cell infiltration by dexamethasone was confirmed by histological examination of treated and untreated infected lungs (data not shown). To determine the effect of dexamethasone treatment on virus growth, the level of virus present in the lungs of control and dexamethasonetreated infected animals was determined by plaque assay at 4, 6, and 8 days post-inoculation. The results of this experiment are shown in Figure 22b. The total amount of virus present in the lungs of infected, untreated animals was similar to that seen in the previous experiment, with RPV infected animals producing 2-3 orders of magnitude more virus than RPVAB5R infected animals. Dexamethasone-induced suppression of the host inflammatory response did not alter the amount of virus produced in either RPV or RPVAB5R infected animals at day 4 following infection although viral load decreased at a slower rate through day 8 in the dexamethasone

PAGE 178

Figure 22. Effects of dexamethasone in mice. (A) Change in phorbol esterinduced ear swelling (inflammation). Mice were either left untreated or were treated twice-daily with dexamethasone as described in Material and Methods. On the fourth day of treatment, ear swelling was induced in untreated (black bar) and dexamethasone treated (grey bar) mice by the application of croton oil to one ear. Ear thickness was measured four-hours after application of croton oil. (B) Effect of dexamethasone on viral titers in mice. Four mice, either untreated (open symbols) or dexamethasone treated (closed symbols) were infected with 104 5 p f u of either RPV (circles) or RPVAB5R (triangles). At the specified times, the amount of virus present in the lungs was measured by determination of the titers on RK-13 cells. Error bars are shown, although for some samples are insignificant.

PAGE 179

9 21 0123456789 10 Day Post Infection

PAGE 180

treated mice. From this data, it was concluded that suppression of the inflammatory response did not alter the ability of RPVAB5R to replicate in this environment. Measurement of Viral Yields in Co-infected Mice Although the results from the previous experiment indicated that the dexamethasone-sensitive steps of the inflammatory response are not responsible for the decreased viral yields observed in RPVAB5R infected animals, it remained formally possible that the B5R protein acted on some component of the immune system not affected by dexamethasone. This possibility was tested by determining whether B5R protein present in the mouse lung could, in some way, alter the environment of the lung and allow for increased production of RPVAB5R; in other words, could RPV achieve a physiological localized complementation of the B5R mutation. To test this possibility mice were infected with either RPVAB5R, RPV or an equal mixture of the two viruses, and the resulting yields of each virus were measured at 4 and 6 days following infection. Since the RPVAB5R mutant contains the gene for the E.coli guanyl-ribosylphosphotransferase (gpt), the levels of this virus in the co-infected lung can be distinguished from that of RPV by performing plaque assays in the presence or absence of mycophenolic acid. The results of this experiment are shown in Table 5. Lungs removed from animals infected solely with either RPV or RPVAB5R contained levels

PAGE 181

162 in v. > w oi o> U o c *o QJ M X 0) T3 > PL, > 0T 01 c .-3 3 o> S o O m M *** o (i Si-B s g T3 .3 oi > t? T3 O" O '-2 •— c — m 01 3 o> X 3 I -5 "£ x: S o > 01 X o I* ^ .2 3 o u X % C8 Oi TS .U *-x g 7 a. £^ o £ § £ i a w — in 3 60 in 3 ^ .O — 1 Q_ 3 rg a$ 01 £2 fQ — OI C — 1 a — u 3 3 o> Oi w in o oi ~ yCL, Oh CO 0 < S Q 2 IT) X) in CO < > ID ^ CO Si oi

PAGE 182

163 of virus similar to those seen in previous experiments at both 4 and 6 days following infection, when the titers were determined in the absence of mycophenolic acid. Determination of the level of virus present in the lungs of both of these singly infected groups in the presence of MPA gave the expected results: i.e. lung samples from RPV-infected animals produced no plaques in the presence of MPA whereas samples from RPVAB5R-infected animals produced the same number of plaques in the presence of MPA as were produced in the absence of the drug. When titers from lung samples of the co-infected mice were determined in the absence of MPA the level of total virus was similar to that seen in animals infected with RPV alone. Titration of these samples in the presence of MPA to determine the level of RPVAB5R virus in co-infected lungs revealed that co-infection of the lung with RPV had no effect on RPVAB5R levels. In all cases, the low level of RPVAB5R in the co-infected lung was identical to that found in the lung of animals infected with RPVAB5R alone. This result indicates that the B5R protein is unable to affect the localized environment of the lung in a way that results in enhanced growth of the RPVAB5R mutant. The data further supports the conclusion that attenuation of RPVAB5R does not result from an inability to suppress the inflammatory response.

PAGE 183

164 The Nature of the Pocks Produced by RPVAB5R The data obtained from the rabbit experiments showed that attenuation of RPVAB5R in this animal did not result from enhancement of the host inflammatory response. In addition, inflammation was also ruled out as the cause of the severe attenuation of this mutant in mice. These results were in apparent contrast to the results of work showing that RPVAB5R produced white pocks on the CAM, the formation of which by CPVASPI-2 has been shown to be the result of inflammation (70,140,142,155). This apparent discrepancy resulted in further investigation into the nature of the RPVAB5R white pocks. If the formation of white pocks by RPVAB5R resulted from failure to suppress inflammation, then these pocks should contain heterophils, as has been demonstrated for CPVASPI-2 (142). A simple test for the presence of activated heterophils involves staining membranes with nitro-blue tetrazolium (NBT) which is a compound that is reduced to formazan by free-oxygen radicals produced by heterophils present in the lesions. NBT treatment of the white pocks produced by crmA (or SPI-2) of CPV (or RPV) mutants has been shown to cause these pocks to turn dark blue as NBT is reduced by the heterophil-generated oxygen radicals (140). The ability of the RPVAB5R white pocks to react with NBT was tested. Eggs were infected with either RPV, RPVAB5R or RPVASPI-2 as previously described. After 3 days of incubation, the CAMs were removed, incubated in PBS

PAGE 184

165 containing NBT and analyzed for reduction of NBT to formazan. As can be seen in Figure 23 (upper panel), as expected, RPV produced red pocks while RPVAB5R and RPVASPI-2 produced white pocks. When the infected CAMs were incubated with NBT, a dramatic difference was observed between the two mutants (Figure 23, lower panel). The white pocks resulting from infection with RPVASPI-2 reacted with the NBT and turned dark blue, consistent with the presence of activated heterophils in these pocks (142). However, the white pocks produced by RPVAB5R remained white, showing only a small degree of staining even after prolonged periods of incubation. The RPV pocks also did not turn blue and were unaffected by treatment with NBT. These results suggest that the white pocks produced by RPVAB5R do not contain activated heterophils and are therefore fundamentally different in nature then the white pocks produced by SPI-2 mutants. Histopathology of the Chorioallantoic Membrane To confirm the results obtained by NBT detection of heterophils, a histological analysis was performed on pocks from RPV or RPVAB5R-infected CAMs. The results are shown in Figure 24 while the pathological features are summarized in Table 6. As expected, uninfected membranes showed no pathology and were free of heterophils. Infection with RPV, however, resulted in marked proliferation of both the chorionic epithelial cells and the fibroblast cells of the stroma. The epithelium in this region showed

PAGE 185

as

PAGE 186

167

PAGE 187

0) tQ

PAGE 189

170

PAGE 190

V to o ** > o o X TJ H *(0 £2 £ • O •H If) 0> 5•r c o 60 > 3 to a; U | < ** o X ** s to w TO s Q. ai o £ .2 Z hri _o H (Q hi ^ I c x .2 • 0> == Oi X 1 1 *^ j, o bO IB -C DO la y 6 c o o> -a (3 > a 3 O 01 — £ <" £ -c .g X) O O C C 01 0 c 1 1 01 3 3 = o 2 hi -a ai c ^ O" is &c E _x E 8 2 3 CO 0> <8 ii a, l*£ "r3 o | c is O to o o C y bC O 3 S. 4= Q> -i-i h 01 01 h 2 O o* 9r c 0 nj hi 01 c T3 S3 oj -a r 1 o ra > ^| S c o *J o x — 01 > u u S Oi -a > .a oi c -2 I c £ -2 00 o C X O 01 13 -c s. a. 0 u E c T3 C (3 X O — '— 57 qj § Q C It tl h oj bp "o .5 — s t3 3, "3 S 13 £ c o — x) ai oi ~ to S£ S C Io o J g E -J .2 .2 (8 CO -3 o o o> 5 o 2^ c — o a f •2 ^ u 5 •= c to x a. E LT) >

PAGE 191

ballooning degeneration along with necrosis and exfoliation of the epithelial layer. A mild heterophilic cell infiltrate was observed in the epithelium as well as in the stromal layer. Blood vessels within the stroma appeared dilated, and there was also evidence of hemorrhage. These last features were more easily observed under low power magnification of intact pocks (Figure 25). Infection of the CAM with RPVAB5R produced pocks containing many of the same features although there were several noticeable differences. These pocks were also defined by extensive proliferation of the epithelial and stromal layers. The epithelium showed similar ballooning degeneration necrosis and exfoliation as observed with RPV; however, the degree of exfoliation was less severe in the RPVAB5R pocks. Unlike the RPV pocks, the pocks generated by RPVAB5R contained large number of heterophils, most of which were contained within the stromal layer. In addition, the innervating blood vessels of the RPVAB5R lesions were not dilated, and there was a noticeable lack of hemorrhage within these pocks in comparison to the RPV pocks (Figures 24 and 25). Effect of Dexamethasone on RPVAB5R Pock Color The results obtained from RPVAB5R infection of animals demonstrated that attenuation of this mutant was not the result of inflammation and, in addition, suggested that B5R was not involved in suppression of inflammation. However, the white pocks produced by

PAGE 192

CD c o U £ o H 3 u 0 OT3 T5 C x 60 o c ax | it Pi o £ w C CQ O < •a ^ c o c < Oh >o IT) 3 ^ .SP< Uh U

PAGE 193

174

PAGE 194

RPVAB5R on the CAM contained non-activated heterophils suggesting that B5R may play a direct role in suppression of inflammation and that failure to do so results in white pock formation. Alternatively, the production of white pocks by RPVAB5R may somehow result as an indirect effect of the inability of this mutant to produce enveloped virus. Dexamethasone-induced inhibition of inflammation has been shown to cause CPVASPI-2 (or RPVASPI-2) white pocks to revert to red (70) demonstrating that the formation of white-pocks by these mutants is a direct result of an activated cell infiltrate. To determine if the white pocks of RPVAB5R are also a direct result of the inflammatory cell infiltrate, the pock morphology of this mutant was determined following inhibition of inflammation with dexamethasone. If the host inflammatory response is the cause of the RPVAB5R white pock phenotype, then suppression of inflammation in the infected egg should allow the B5R null mutant to form red pocks. To test this hypothesis, CAMs were infected with RPV, CPV, RPVASPI-2, CPVASPI-2 or RPVAB5R, with or without dexamethasone, and the resulting pocks examined 3 days postinfection. The top panel in Figure 26 shows the familiar pock phenotypes resulting from infection of the CAM in the absence of dexamethasone; i.e. RPV and CPV produced red pocks while the SPI-2 and B5R mutants produce white pocks. The lower half of Figure 26 shows the effect of adding dexamethasone to inhibit inflammation at the time of infection. Dexamethasone had no effect on the color of the pocks produced by either

PAGE 195

-a rv C En > o _, a> at a> £ c s 3 (8 J§| ** £ ii I s T3 a* t>c 5 Q> £ CB C X •FH i-c t2 Is u oo o .£ g c d S o t 5 5fl tJ m ^ .£ S o S o O J-* ex 2 u O a CuX
PAGE 197

178 RPV and CPV although the lesions appeared to be reduced in size when compared with untreated membranes. Dexamethasone treatment of either RPVASPI-2 or CPVASPI-2 infected eggs resulted in pocks that were red in contrast to the white pocks normally formed by these mutants, consistent with a dexamethasone-mediated inhibition of inflammation. However, dexamethasone had no effect on the color of the pocks produced by RPVAB5R. The resulting pocks remained white and appeared similar to those produced in the absence of the drug. It should be noted, however, that pocks formed by RPVAB5R on the dexamethasone-treated membranes were smaller and less dense than those on the untreated membranes. These results suggest that while inflammation contributes to the production of white pocks by RPVAB5R, it is not the sole cause of this phenotype as it is for RPVASPI-2 white pocks. Discussion The severity of a poxvirus infection is determined by a combination of the ability of the virus to spread to and replicate in a wide variety of cell types as well as its ability to evade the host immune system. The importance of these abilities has been demonstrated by work showing that viruses defective in either of these functions are attenuated in animals. Local and long-range spread of virus within the animal is believed to be a function of the

PAGE 198

179 enveloped form of the virus, assembly of which is dependent upon the production of several membrane-associated EEV proteins. Since deletion of the genes coding for these proteins often interrupts EEV morphogenesis it is not surprising that such mutations frequently results in attenuation of the virus in animals. The ability of the virus to alter the host immune response has also been attributed to the production of several recently discovered virusencoded proteins. Many of these proteins interact with components of the host immune system at key points and are believed to allow the virus to alter the immune response. Again, not surprisingly, deletion of the genes encoding these immune-modulators can result in attenuation of the virus. As a result of its additional and unique outermost envelope, extracellular enveloped virus contains several membrane-associated proteins not found in the IMV particle. Many of these membrane-associated EEV proteins have been studied in order to gain a better understanding of their contribution to the biology of the viral infection. A common feature shared by many of these proteins is that deletion of their respective ORFs often interrupts the terminal steps of morphogenesis, resulting in inhibition of EEV production (20,55,58,143,165,218). Since it is believed that it is the enveloped form of the virus that is responsible for virus spread, both in tissue culture and in the infected animal (9,21,27,146,150), it is not surprising that several such mutants fail to produce plaques in cell culture and are

PAGE 199

180 attenuated in animals. The product of the B5R gene has been shown to be a membraneassociated glycoprotein present in EEV but not IMV. Unlike the other membrane-associated EEV proteins, B5R is unique in that in addition to being a component of the EEV envelope, the protein is also secreted from the cell by a pathway independent of EEV morphogenesis (119). Interestingly, the B5R protein has homology to factor H and complement C4b binding protein, two regulatory components of the complement cascade. This homology coupled with its extracellular location led to the hypothesis that B5R might also belong to a group of viral proteins known to be host response modifiers and as such it might function to inhibit inflammation by inhibition of complement activation. The work presented in this chapter was undertaken to determine if the B5R protein has a role in viral pathogenesis and if so, does this role involve control of inflammation. This question was examined by construction of a B5R null mutant, RPVAB5R, which was characterized for its ability to cause disease in animals, from the data presented here, it can be concluded that while RPVAB5R is attenuated, this attenuation does not result from an enhanced inflammatory response but instead most likely arises from a defect in EEV production, a phenotype easily demonstrated in cell culture. Failure of RPV to produce the B5R protein resulted in a phenotype in cell culture identical to that reported for the vaccinia ps/hr (B5R homolog) mutant. RPVAB5R failed to form plaques on Vero cells and when tested on

PAGE 200

181 RK-13 cells, produced plaques somewhat smaller than those produced by RPV. The inability of RPVAB5R to form plaques was shown to extend to several other cell types, the most notable being chicken embryo fibroblasts (CEFs). This result is significant in that these cells have been considered extremely permissive and are commonly used to grow various mutants of the virus when other cell lines are found to be unable to do so. Hence, CEF cells were traditionally used for the isolation of white pock and host range mutants many of which contain deletions encompassing large segments of the viral genome Failure of RPVAB5R to plaque on these cells shows that they are not universally permissive and suggests that a class of white-pock mutants may have been missed in past studies because they failed to plaque on these cells. Indeed, RPVAB5R is the first mutant of any kind reported to be restricted for plaque formation on CEFs. This result should serve as a reminder that the choice of a cell line for use in the purification of mutants can skew the class of mutants that is isolated. An additional point worth noting is that since CEFs are primary cells, it is likely that the behavior of the virus in animal tissue is more accurately reflected by its behavior in CEF cells than in any of the other established cell lines. The alteration in the plaque morphology and host range of RPVAB5R led to an examination of the ability of the mutant to replicate in permissive and non-permissive cells. In permissive cells, RPVAB5R replication was nearly normal showing only a slight decrease in the levels of IEV when

PAGE 201

182 compared with RPV. It is likely that this reduced level of IEV are responsible for the small-plaque phenotype since IMV and EEV levels are equivalent in RPV and RPVAB5R infected cells. In non-permissive cells, the major defect observed for RPVAB5R was a failure to produce the enveloped forms of virus (IEV and EEV). The levels of IMV, though somewhat reduced, remained substantial. Failure of RPVAB5R to form plaques in these cells supports the work of others indicating that IMV plays little if any role in the spread of virus and that the enveloped form of virus must be produced for plaque formation. This result shows that like many of the other EEV membraneassociated proteins, B5R is necessary for production of EEV. However, in this regard it is interesting that deletion of B5R in RPV confers a host range indicating that the protein is essential in some cells but not others. This result is in contrast with results from similar studies performed with the IHD-J and WR strains of vaccinia (58,218). While deletion of B5R in RPV had no effect on EEV production in RK-13 cells, deletion of B5R gene from vaccinia abolished the ability of that virus to produce EEV in these cells. In this way, the B5R protein of RPV also differs from several of the EEV envelope proteins of vaccinia in that deletion of these proteins usually results in universal disruption of EEV production. These observations suggest that RPV but not VV may produce other viral proteins that can compensate for the loss of B5R in a cell line specific manner. RPVAB5R was shown to be severely attenuated in both mice and

PAGE 202

183 rabbits. This attenuation was most dramatically demonstrated by infection of rabbits with RPVAB5R. Whereas intradermal injection of rabbits with a single pfu of RPV is usually sufficient to result in a lethal infection, it was observed that infection of these animals with up to 50000 pfu of RPVAB5R was insufficient to cause disease and in fact, did not even result in a systemic response to the virus, as judged by a lack of pyrexia (Figure 17b). A local but contained reaction to the infection produced some induration at the site of injection, however, the extent and duration of the induration was much less than was observed in lesions resulting from infection with RPV (data not shown). In addition, RPVAB5R-infected rabbits failed to develop any secondary lesions suggesting that the virus failed to spread from the inoculation site. These observations are all consistent with the idea of limited growth of RPVAB5R at the primary site of inoculation, a situation that could result from a defect in virus morphogenesis or from an enhanced inflammatory response. Histological examination of primary lesions taken from RPV and RPVAB5R infected animals showed that the attenuation of RPVAB5R was not due to an enhanced inflammatory response. Lesions produced by RPV showed significant necrosis, hemorrhage as well as extensive infiltration of neutrophils. In contrast, RPVAB5R produced comparatively mild lesions that contained very little necrosis and only a mild to moderate degree of inflammatory cell infiltrate. These observations rule out the failure of RPVAB5R to control the inflammatory response as the

PAGE 203

184 cause of its attenuation in rabbits and are consistent with reduced growth of the virus in rabbit skin. A more extensive analysis of RPVAB5R attenuation was carried out in the mouse model. In this system, weight loss and illness result from localized growth of virus in the lung, in contrast to the rabbit system where illness and death occur due to spread of the virus from the initial site of inoculation. Similar to what was observed with the rabbits, RPVAB5R was extremely attenuated in mice as well. While infection of mice with RPV resulted in rapid weight loss and death, animals infected with RPVAB5R showed minimal weight loss and suffered no deaths resulting from the infection. It was noted that mice infected with the mutant virus did show some weight loss, especially at the highest doses used ( 10 8 pfu, data not shown). Since weight loss correlates with lung damage, this observation is consistent with limited growth of the virus at the site of inoculation in RPVAB5R infected mice. Determination of the virus titers in infected lungs confirmed that while RPVAB5R was able to replicate in the mouse lung the extent of this growth was limited. RPVAB5R titers were reduced approximately 100-fold in comparison to the levels of virus produced in RPV infected animals. It is likely that the limited growth of RPVAB5R is simply a manifestation of a failure to produce enveloped virus, as mirrored in cell culture, and does not indicate a failure to suppress the host inflammatory system based on the following arguments. First, while histological

PAGE 204

185 examination revealed that the inflammatory cell infiltrate in RPVAB5R infected animals was slightly greater than was observed in RPV infected animals, the difference was minimal and is unlikely to result in such severe attenuation. In this regard, it is worth noting that studies using RPVASPI-2 a mutant that is well known to be unable to inhibit inflammation, have shown that this mutant is not attenuated in mice. This suggests that inflammation has little effect on the course of the infection in this system and is consistent with the dramatic attenuation observed with RPVAB5R in mice resulting not from failure to suppress inflammation but from a more fundamentally disruptive defect such as failure to grow or spread. Failure of this mutant to grow efficiently is consistent with the observation that RPVAB5R infected lungs, unlike RPV and RPVASPI-2 infected lungs, were free of the edema and necrosis that results from destruction of the lung tissue. More significantly, while treatment of these animals with the anti-inflammatory drug dexamethasone completely blocked the appearance of inflammatory cells in the lung, it did not alter the yields of virus produced in the lungs of RPVAB5R infected animals. While this result does not rule out a possible interaction of the B5R protein with the host immune system, it does provide direct evidence that the attenuation of this mutant is not a result of the inflammatory response. Finally, co-infection with RPV was unable to elevate the levels of RPVAB5R in animals. This result indicates that the defect in RPVAB5R cannot be rescued in trans and rules out any general immuno-

PAGE 205

186 suppressive property as a possible function for the B5R protein. Collectively, these results show that the attenuation of RPVAB5R is not due to an increase in inflammation but is most likely due to a failure of the virus to grow efficiently due to its defect in the production of enveloped virus. While the above results show that inflammation is not the cause of RPVAB5R attenuation, neither do they exclude the possibility that B5R does not affect inflammation since work with RPVASPI-2 has suggested that the activity of known virus-encoded antiinflammatory proteins may be difficult to detect in mice and especially in rabbits (109) (and Brooks, M. and Stern, R., manuscript in preparation). Because of this, the link between B5R and inflammation was investigated in the 11-day old chicken embryo, a system in which inflammation, which is essentially the only immune response present, has been shown to have a large impact on pock phenotype (70,138). In this system, RPVAB5R produced white pocks on the CAM, similar to the white, inflammatory pocks produced by CPV (or RPV) containing deletions of SPI-2, which is known to inhibit inflammation (142). Previous studies by others have shown that the formation of white pocks by RPVASPI-2 is a direct result of a significant inflammatory cell infiltrate (70). Histological examination of the RPVAB5R white pocks revealed that analogous to the white pocks of RPVASPI-2 and unlike the pocks produced by RPV, the white pocks produced by the B5R mutant contained large numbers of heterophils. This result shows that the B5R protein does have an effect on the inflammatory

PAGE 206

187 response; however, the question arises as to whether this increase in inflammation is a primary effect of the B5R deletion or whether it is a secondary effect resulting from the inability of the mutant to produce enveloped virus (and therefore spread) in CEF cells. While it is impossible to distinguish between these two possibilities from the data contained in this dissertation, several curious observations indicating that the white pocks of RPVAB5R and RPVASPI-2 are fundamentally different give some credence to the latter interpretation. First, while inhibition of inflammation with dexamethasone did reduce the opaqueness of the RPVAB5R white pocks, it failed to restore the wild-type red pock phenotype. This result indicates that unlike the white pocks produced by RPVASPI-2, the formation of white pocks by RPVAB5R is not due entirely to inflammation and suggests that RPVAB5R is defective in a second function that is required for red pock formation. Since this mutant is known to be defective in EEV production in CEF cells, it is likely that normal growth and spread is essential to red pock formation. Second, when incubated with NBT, the RPVAB5R white pocks, unlike the white pocks produced by SPI-2 mutants, failed to reduce the NBT to formazan (i.e. failed to turn blue), indicating that although heterophils have migrated into these pocks they are not activated. Heterophil migration and activation are two separate events that are coordinately regulated to prevent inappropriate damage to the host and as a result are thought to differ in their response to the levels and types of pro-inflammatory stimulus

PAGE 207

188 (170,171,179,184). In view of this, these results might suggest that although there is enough stimulus (damage) within the pock to result in heterophil migration, due to the inability of the mutant to spread within these membranes, there is not enough stimulus to cause detectable activation of these heterophils. Consistent with this interpretation is the observation that dilation of blood vessels, one characteristic of an inflammatory response, was observed within RPV and RPVASPI-2 (not shown) but not RPVAB5R lesions suggesting that the proper level or type of stimulus to induce vasodilation is not produced in RPVAB5R infected tissue. Collectively, these observations could be taken as evidence that inflammation is not directly effected by B5R but is instead a secondary effect resulting from defective growth of the mutant. While this mutant still produces other anti-inflammatory proteins, failure of the virus to rapidly spread to uninfected cells might result in the production and amplification of pro-inflammatory signals by the infected cell and neighboring cells before a significant concentration of virus encoded antiinflammatory proteins can be produced in the local area. In support of the argument for non-involvement of B5R in the modulation of inflammation it should be noted that the domains of factor H with which B5R shares homology are different from those involved in the inhibition of complement activation as determined by the recent mapping of the inhibitory domains (82).

PAGE 208

189 The work presented here suggests that the idea that red pock generation equates with the ability to inhibit inflammation while white pock generation equates with the loss of this ability might be too simple in that pock formation is a complex process. Based on the above observations, I propose that pock formation is the result of at least three separate events; membrane hyperplasia, inflammation and hemorrhage. Both red and white pocks exhibit hyperplasia of the epithelium and stromal fibroblasts within the pock as a result of the virus encoded growth factor (33) endowing all pocks an underlying opaqueness. In red pocks, such as those produced by RPV, virus replication and spread result in severe necrosis of the cell layers along with severe vasodilation resulting in hemorrhage (155) within the stroma giving the pock its red color In these pocks, the elicitation of inflammatory cells is apparently specifically inhibited by virus encoded factors such as SPI-2 (crmA). In RPVASPI-2 white pocks, unrestricted inflammation within the pock inhibits virus growth (70), preventing spread of the virus which limits the damage and the subsequent hemorrhage. This combined with the increase in the cell density results in a white pock. The RPVAB5R white pocks also result from hyperplasia, lack of hemorrhage and inflammation. However, I believe that with this mutant, inhibition of virus growth and spread and the subsequent lack of hemorrhage is not due to inflammation but is instead the result of an innate defect in enveloped virus production. This would be consistent with this mutant failing to form a red pock in the presence of the

PAGE 209

190 anti-inflammatory agent dexamethasone since virus growth would still be restricted even in the absence of inflammation. It is suggested that labeling pocks red or white, while convenient, be considered insufficient for determining the link between virus proteins and inflammation and that histological examination should become the standard method for making this determination. In summary, the results presented in this chapter show that the B5R protein is necessary for the proper morphogenesis of the enveloped form of virus. Failure to produce B5R results in an inability to produce EEV, but not IMV, in a cell line specific manner. In addition, while deletion of the B5R gene does result in attenuation of the virus in animals this attenuation was not due to failure of the mutant to inhibit the inflammatory response. Instead, it is believed that the mutant is attenuated as a result of its inability to produce enveloped virus. While it is impossible to conclude that B5R does not directly inhibit the inflammatory response based on the result obtained on the CAM, it is believed that the B5R protein has no such activity.

PAGE 210

CHAPTER IV THE ROLE OF RPV MEMBRANE-BOUND AND SECRETED B5R PROTEIN IN MORPHOGENESIS AND POCK MORPHOLOGY Introduction The poxviruses are large, structurally complex viruses that have the unique feature of replicating entirely within the cytoplasm of the infected cell (128). Because the entire replication cycle occurs within the cytoplasm, the viral genome must encode virtually all of the proteins necessary for replication and assembly of new viral particles. A consequence of this is that the poxviruses have an exceedingly large genome consisting of approximately 180 kilobases of DNA with the potential to encoding approximately 200 proteins (79,120). Many viral proteins have been identified and characterized and can be organized into groups based on the roles they play in virus replication. One such group consists of proteins that have been shown to be involved in the basic processes of replication such as DNA synthesis (200)and RNA transcription (129). A second group comprises proteins that have been shown to be structural components of the virus particle, of which there are two forms, intracellular mature virus (IMV) and extracellular enveloped 191

PAGE 211

192 virus (EEV) (61,139,147). These two equally infectious forms of the virus represent different developmental stages in viral morphogenesis and differ only in the number of envelopes encasing the virus particle. While most virion structural proteins are found in both IMV and EEV, a subset of these proteins are unique in that they are found only in the EEV particle (144,147). This results from these EEV specific proteins, of which eight have been identified, being located in the membrane(s) unique to this form of the virus. All but one of these EEV proteins have been shown to be glycoproteins (144,147). While localization of these proteins in the virion outer membrane might suggest a role for these proteins as receptor molecules, to date the true function of these proteins remains largely unknown. A common feature shared by nearly all of these proteins, however, is that null mutations in many of the corresponding genes result in a defect in enveloped virion morphogenesis (55,58,143,177,218). Such mutants produce tiny plaques and are attenuated in animals presumably due to a failure of the virus to spread from the site of inoculation. While the proteins described so far are involved in the synthesis of virus components and the assembly of these components into virus particles, a third population of viral proteins exists which serve as virus virulence factors. Many of the proteins found within this group interact with components of the host immune system resulting in down-regulation of the host immune response to the virus infection (5,6,110,158,187). Since many of

PAGE 212

193 the regulatory and effector proteins of the immune system are extracellular, it is not surprising that many of these virus-encoded homologues are also secreted from the infected cell. An interesting feature of these proteins is that they often have significant structural homology to proteins of the host immune system suggesting that the virus has acquired them from the host during the course of the viruses evolution. To date, examples of such proteins include a secreted interleukin-lC receptor, a secreted tumor necrosis factor receptor, a secreted interferon-gamma receptor and a secreted complement factor C4b binding protein (5,97,99,110,111,142,158,182,183,187) Each of these viral proteins has been shown to have the activity suggested by its homology with a host immune protein. The B5R protein of RPV is yet another viral protein which shares homology with components of the mammalian immune system but for which a function remains obscure. This virus-encoded protein is a member of the SCR (short consensus repeat) superfamily which are involved in control of the complement cascade. Database analysis using the B5R predicted amino acid sequence has shown that it has homology to complement factor H and complement factor C4b, both of which are negative regulators of the complement cascade (133,210). The B5R protein exists as a 45 kDa glycoprotein present in the membrane of EEV but not in IMV. As has been observed for many of the EEV membrane proteins, deletion of the B5R ORF results in a defect in

PAGE 213

194 enveloped (IEV and EEV) virus but not IMV morphogenesis. In addition to the membrane-bound form of the protein, B5R is also found in the supernatant of infected cells as a 38 kDa secreted protein. This secreted form has been shown to be derived from the 45 kDa form by processing and its secretion is independent of viral morphogenesis. The homology of this protein with that of host regulatory proteins together with the finding that B5R is secreted led to the hypothesis that this protein may regulate complement activation and as a consequence, inflammation. However, the work presented in the previous chapter demonstrated that attenuation of a B5R null mutant was not a result of an enhanced inflammatory response but was likely due to a defect in morphogenesis observed with this mutant. To date, RPV, VV and CPV have all been shown to produce the secreted form of B5R. Since the sequence of the B5R gene has been extremely highly conserved by the orthopoxviruses, it is likely that all orthopoxviruses secrete this protein suggesting that its secretion is important for the virus. The goal of the work presented in this chapter was to determine the importance of the membrane-bound and secreted forms of the B5R protein in morphogenesis and in virulence. This was achieved through the construction of mutants that produced either the membrane-bound or secreted form of B5R, but not both, and the characterization of these mutants in cell culture and in animals.

PAGE 214

195 Results Construction of a Virus Producing only Secreted B5R Protein Previous work has shown that RPV produces two forms of B5R protein, a 45 kDa membrane bound form, found associated with enveloped virus particles (57,100,119), and a second, 38 kDa secreted form, which has been shown to be derived from the membrane bound form by processing (119). To determine the importance of the membrane bound form in viral virulence, a mutant virus was constructed in which the membrane anchor of the B5R protein was removed. Removal of the membrane anchor should result in the synthesis of a mutant B5R protein that fails to localize to either the virion or host-cell membrane. It would also be expected that such a mutant would be secreted in far greater amounts from the cell in comparison to the wild-type B5R protein. Previous studies have shown that a carboxylterminal hydrophobic region containing amino acids 276-300 serves as a membrane anchor for the protein and that truncation of the protein at amino acid 275 results in enhanced secretion of this truncated protein into the supernatant (100). Based on these data, a mutant, called RPVB5R-T, was constructed in which the B5R ORF of RPV was replaced by a mutant ORF encoding a B5R protein (referred to as B5R-T) truncated at amino acid 275 and

PAGE 215

196 therefore lacked the membrane anchor (details of this construction are provided in Materials and Methods). Construction of RPVB5R-T was confirmed by PCR analysis of viral DNA (data not shown) as well as by SDSPAGE analysis of immunoprecipitated proteins from infected CEF cells which is shown in figure 27. As expected, much higher levels of the truncated protein were secreted into the supernatant of infected cells when compared to the levels of wild-type B5R found in the supernatant. Immunoprecipitation of RPV infected cell extracts and supernatants from these cells yielded the 45 kDa membrane bound and the 38 kDa forms (lanes 1 and 2 respectively) of the wild-type B5R protein in agreement with previous results (119). As has been previously noted (119), the level of secreted B5R protein in RPV infected CEF cells is considerably lower than that noted for other cell lines. In RPVB5R-T infected CEF cells, the level of B5R-T secreted from the cell was much higher than the level of wild type B5R secreted from RPV infected cells (lane 4). In addition, the molecular weight of the B5R-T protein (approximately 40.5 kDa) was consistent with the size predicted by termination of the predicted amino acid sequence at amino acid 275 (lane 3). Interestingly, the protein recovered from the supernatant of RPVB5R-T infected cells had approximately the same apparent molecular weight as that of the RPVB5R-T cellular form (lanes 3 and 4). The size of the B5R-T secreted protein also appears larger than the secreted B5R protein produced by RPV. This size difference might indicate that the normal processing of the B5R protein is not at the start of the trans-

PAGE 216

Figure 27. Construction and characteriztion of RPVB5R-T. A) Construction scheme for RPVB5R-T. Depicted is a representation of the B5R gene of RPV (upper) and of RPBB5R-T (lower). In RPVB5R-T, the membrane anchor sequence (represented by the diagonally hatched box) has been replaced by the Ecogpt cassette. In addition, a stop codon was placed at the postion coding for amino acid 276. RM276, RM277, RM278, and RM299 represent the primers used in the construction of this mutant. B) Immunoprecipitation of B5R from RPV and RPVB5R-T infected cells. RPV (lanes 1 & 2) or RPVB5R-T (lanes 3 & 4) infected CEF cells were labeled with 35Strans label from 6-18 hours post-infection. Proteins from cell extracts (approximately 5.0 x 10 5 cells) (lanes 1 & 3) and from an equivalent amount of virus depleted supernatants from these cells (lanes 2 & 4) were then immunoprecipitated with polyclonal anti-B5R sera and the immunoprecipitated proteins separated by polyacrylamide-gel electrophoresis. The arrows on the left indicate the cellular (arrow A) and secreted (arrow B) forms of the protein synthesized by RPV. The arrow on the right (arrow C) indicates the B5R protein secreted by RPVB5R-T.

PAGE 217

12 3 4

PAGE 218

199 membrane region as predicted but is instead further upstream of this region. Alternatively, the unexpected larger size of the B5R-T secreted protein might result from an altered pattern of glycosylation in this protein as a result of its missing the membrane anchor region. The latter possibility was tested by comparing the decrease in molecular weights of B5R and B5R-T following removal of their sugar residues by treatment with peptide-N-glycosidase. The results, which are shown in Figure 28, indicate that the gross level of glycosylation is approximately equal in both proteins. In agreement with published results (147), N-glycosidase treatment of immunoprecipitated intracellular B5R from RPV infected cells resulted in a shift in size of approximately 3 kDa from 45 kDa to 45 kDa (lane 2) when compared to untreated protein (lane 1). Similar treatment of B5R-T also altered the apparent molecular weight by approximately 3 kDa (from 40 kDa to 37 kDa) when compared to the untreated protein (lanes 3 and 4). The similar shift in molecular weights suggests that the apparent size difference between secreted B5R and secreted B5R-T is not due to different levels of glycosylation. B5R-T Still Remains Associated with Enveloped Virions The 45 kDa form of B5R has been shown to be a membrane-associated protein located in the envelope of the EEV particle (57,100,119). Anchoring of B5R in the membrane is believed to be mediated through a hydrophobic stretch of amino acids located near the carboxyl end of the protein, a

PAGE 219

Figure 28. Peptide-N-glycosidase treatment of the B5R and B5R-T proteins. CEF cells were infected with either RPV or RPVB5R-T and the viral proteins were labeled overnight with 35S-translabel. B5R (lanes 1 & 2) and B5R-T (lanes 3 & 4) were immunoprecipitated from total cell extracts after which samples were either left untreated (lanes 1 & 3) or were treated with peptideN-glycosidase (lanes 2 & 4) prior to separation by PAGE.

PAGE 220

83kDa — 33.6 kDa — 28.1 kDa —

PAGE 221

202 hypothesis that is supported by data showing that deletion of this region results in enhanced secretion of this protein into the supernatant. In addition to causing increased secretion of this protein, removal of the membrane anchor should result in a failure of such a mutant protein to associate with the EEV particle. To test this and to further characterize the B5R-T protein, CsCl gradient purified IMV, IEV and EEV from RPV, RPVAB5R or RPVB5R-T infected RK-13 cells were analyzed for the presence of B5R or B5R-T, respectively, by western blot analysis using anti-B5R serum. The results of this experiment are shown in Figure 29. In agreement with previous work, the 45 kDa B5R protein was detected in enveloped virus, but not IMV, from RPV infected cells (lanes 1, 2 and 3) (119). As expected, no B5R protein was detected in any form of virus obtained from RPVAB5R infected cells (lanes 4, 5 and 6). Surprisingly, B5R-T was found to be present in virus purified from RPVB5R-T infected cells. The truncated protein was present in the enveloped forms of virus but not in IMV, a pattern of localization identical to that observed for the wild-type B5R protein. This result was unexpected based on previous work by others and indicated that removal of the B5R hydrophobic region does not prevent association of B5R-T with enveloped virus particles. The Nature of the Association of B5R-T with Enveloped Virions The finding that B5R-T localizes to IEV and EEV was surprising since this protein no longer contained a membrane anchor sequence. This

PAGE 222

Figure 29. Western blot analysis of IMV, IEV and EEV. RK-13 cells were infected with either RPV, RPVAB5R or RPBB5R-T and the progeny virus were purified by centrifugation in CsCl gradients (see Material and Methods). Equal amounts of viral proteins (as determined by coomassie staining of a duplicate gel) were separated by polyacrylamide-gel electrophoresis, blotted to nitrocellulose and hybridized overnight to polyclonal anti-B5R sera. IMVintracellular mature virus. IEV-intracellular enveloped virus. EEVextracellular enveloped virus.

PAGE 223

204 4 ft ir ir £ > > 2 > > i— i i— i M-l i— i i— i MJ hh i— i W > > 78 kDa — 47 kDa 31.4 kDa

PAGE 224

205 association could be due to one of three possibilities; 1) the B5R protein is normally anchored in the membrane in a fashion independent of the hydrophobic region, 2) removal of the hydrophobic region results in a second region of the protein acting as a membrane anchor, or 3) the truncated B5R protein is not an integral membrane component of the virion but is instead associated with the virus membrane through interactions (specific or nonspecific) with the particle surface, possibly involving interactions with other virion membrane proteins. The last possibility implies that the nature of the associations of B5R-T and wild type B5R with the virion membrane should be different whereas the association of these proteins with the virion should be similar if either of the first two hypotheses is correct. Sodium carbonate treatment of membranes at alkaline pH (11.5) is a technique that has been used to distinguish between integral and non-integral membrane proteins (73,211). Under these conditions integral membrane proteins remain membrane associated whereas non-integral membrane proteins are preferentially released from the membrane. The nature of the B5R-T/EEV association was examined using sodium carbonate treatment of EEV purified from either RPV or RPVB5R-T infected cells. The results of this experiment are shown in Figure 30. When virions were incubated in 100 mM Tris, pH 7.0, both B5R and B5R-T remained tightly associated with the virus and were found almost exclusively with the pelleted virus following centrifugation of the incubation mixture (lanes 1 and 3). However, incubation of these virions

PAGE 225

Figure 30. Western blot analysis of Na 2 G0 3 treated RPV or RPVB5R-T EEV. CsCl gradient purified EEV from RPV or RPVB5R-T infected cells were incubated in either lOOmM Tris pH 7.4 or lOOmM Na 2 C0 3 (pH 11.5) on ice for 30 minutes after which the virus was pelleted by centrifugation at 12,000 RPM for 30 minutes. Viral proteins present in the pellet (P) and supernatant (S) fractions were separated by polyacrylamide-gel electrophoresis, transferred to a nitrocellulose membrane and the membrane hybridized with polyclonal anti-B5R sera. P-pellet fraction, S-supernatant fraction.

PAGE 226

50.3 kDa 33.3 kDa 28.8 kDa RPV RPVB5R-T L + + + + J L J 100 mM Tris 100 mM Na 2 C0 3 pH 7.4 pH H.5

PAGE 227

208 in 100 mM sodium carbonate, pH 11.5, revealed a dramatic difference in the association of the two proteins with the virion. Whereas the association of wild type B5R with the virion was unaffected by incubation in alkaline conditions (lane 4), the association of B5R-T was apparently somewhat unstable and a significant percentage of this protein was released from the virion as indicated by its appearance in the supernatant fraction following collection of the virus by centrifugation (lane 7). These results demonstrated that while B5R-T could still associate with the virus particle, it was not associated as tightly with the membrane as was the B5R protein, confirming that the carboxyl hydrophobic domain of B5R (amino acids 276-300) which is absent in B5R-T, is necessary for proper anchoring of wild type B5R. Plaque Size and Host Range Characteristics of RPVB5R-T Deletion of the B5R gene product has been shown to alter the plaque size and host range of the virus (58,119,147,195,218), (and this dissertation). To measure the consequence of deleting the membrane anchor sequence on these characteristics RPVB5R-T was tested in cell culture for its ability to form plaques on RK-13 and CEF cells, which are permissive and non-permissive, respectively, for RPVAB5R. The results, shown in Figure 31, revealed that RPVB5R-T had a host range pattern identical to that seen with RPVAB5R and was able to form plaques on RK-13 cells but not on CEF cells. In addition the

PAGE 228

u 3 J2i

PAGE 229

210

PAGE 230

211 size of the plaques formed by RPVB5R-T in RK-13 cells were reduced when compared with those formed by wild-type RPV, again a pattern noted with RPVAB5R on these cell lines. Evaluation of a number of cell lines indicated that RPVB5R-T had a reduced host range pattern identical to that observed for RPVAB5R. Hence, the presence and enhanced secretion of the B5R protein fails to allow wild-type host range and suggests that the host range observed with B5R mutants was solely a function of failure to produce the membrane anchored form of the protein. In addition, these data indicated that although the B5R-T protein is associated with the virus particle, this association is apparently non-productive suggesting that the activity of B5R requires firm anchoring of the protein in the virus envelope. RPVB5R-T is Defective in Enveloped Virus Production in CEF Cells The ability to form plaques in cell culture is believed to be dependent upon the production of the enveloped forms of virus. The failure of RPVAB5R to form plaques in CEF cells has been shown correlate with and is believed to result from an inability of this mutant to form enveloped virus in these cells as shown in chapter three. Since RPVB5R-T has a host range identical to that of RPVAB5R it is likely that this new mutant also suffers the same defect in morphogenesis in the non-permissive CEF cells. To test this, the ability of RPVB5R-T to produce virus was measured in permissive (RK13) cells and non-permissive (CEF) cells. 3H-thymidine labeled virus was

PAGE 231

212 purified from permissive and non-permissive cells and the various forms of virus were separated on CsCl gradients. The amount of each type of virus produced was determined by quantitation of the amount of 3H-thymidine present in fractions of the gradient. The results of this experiment, shown in Figure 32, were as predicted. In permissive cells (RK-13), both RPV and RPVB5R-T produced identical amounts of IMV and EEV, however, the mutant produced roughly half the amount of IEV as that produced by RPV. In non-permissive cells (CEF), while both viruses produced similar amounts of IMV, the mutant was completely defective in its ability to produce the enveloped form of the virus. These results are identical to those obtained from similar experiments with RPVAB5R and suggest that it is the membrane-bound form of the protein that is necessary for proper morphogenesis of enveloped virus. The Behavior of RPVB5R-T on the CAM The similar behavior of RPVAB5R and RPVB5R-T in cell culture might imply that it is only the membrane bound form of the protein which is important for full expression of RPV virulence. This question was initially addressed by examining the pock morphology produced by RPVB5R-T on the CAM. The results detailed in chapter three showed that RPVAB5R produced white pocks on the CAM in contrast to the red, hemorrhagic pocks produced by RPV. Although white in color, these pocks do not contain activated

PAGE 232

CO > Oh -a 0 re 73 C — U I S > o -D 0> a re 73 O) to c •4= C 01 l/l Ol h Oh >H C 3 re C M Oi Oh in £ o H (0 5 > Oh a. a,< o Z Q 73 01 re -D a/ x H 73 0> u O c O -w u Ol c CO 73 c re bO i) x c o> 73 re H GO 0 X 73 Ol o Uh w u 73 C m i Oh C 7) 3 si o> jo ^ O O a. in sO X 00 3P o> u X Ol > _0> Ol £ o X (0 x Oh re U T3 0> c o — o ^ re cm C o c re o u 3 73 Ong 73 Ol to o> > sre v oi J H 73 01 o re O Ox 60 3 C Ol i/i Oi Sho ^3 "cL h re jto bO-3 bo c "3 're *• h= 3 C c o re H i in pa > Oh 3 "8 Oh £ £ re c O* re •— O Oh X oi re c O u 73 c re re Ol re c bow .2 <+H 2 .2 o re H re Ol X H o 3 '> Ol o 3 73 O sOh 73 re lH II lH Ol T3 Ol X • — su o> 73 tn re 3 H 73 C re U .3 c 0> 73 re lH So Oi X e — O .~ o re 060 o w X 1 2 — — £ o — I o X

PAGE 234

215 heterophils in contrast to the white pocks produced by RPV mutants of SPI-2, a virus encoded protein known to be involved in the inhibition of inflammation (142,158). To determine the role of membrane anchored B5R in pock phenotype, CAMs were infected with RPV, RPVASPI-2 or RPVB5R-T and examined for pock formation 72 hours later. The results are shown in the upper half of Figure 33. Both RPV and RPVASPI-2 behaved as expected, producing red and white pocks respectively. As might be expected from the results in cell culture, RPVB5R-T behaved like RPVAB5R and produced white pocks on the CAM, and these pocks appeared similar to the white pocks of RPVASPI-2. The results from the previous chapter showed that the white pocks of RPVASPI-2 and RPVAB5R differ in that while both contain inflammatory cell infiltrate, the activation state of the heterophils present within these pocks is different as determined by the nitroblue tetrazolium (NBT) reduction assay (158). To further compare the phenotypes of RPVAB5R and RPVB5R-T, the ability of the RPVB5R-T to reduce formazan was measured and is shown in the lower half of Figure 33. As was previously demonstrated, the white pocks generated by RPVASPI-2 reduced NBT to formazan, turning blue in the process. As was expected, the white pocks of both RPVB5R-T and RPVAB5R behaved alike, failing to turn blue when incubated in the presence of NBT, and indicating that these pocks do not contain heterophils in an activated state. These results show that on

PAGE 235

t— i ~ o S as £ u 4> .2 .CD ^3 CD (A m CD C O r0 sO u_ T3 CD X cj c CD sCD S O c O u < U c o H i es in CO > Oh CD C (0 c_ '— CD o CD CD CD CD In O '> CD k o .SP ca 60 X c CD -g SB* o c "o M u O rv CD C"> x ** *: cs

PAGE 236

217

PAGE 237

218 the CAM, RPVB5R-T behaves in a manner identical to that of RPVAB5R and together with the behavior exhibited in cell culture suggest that it is the membrane bound form of B5R alone that produces the wild-type phenotype. The Behavior of RPVB5R-T in Animals The importance of the membrane bound form of B5R in virus virulence was further explored by ascertaining the ability of RPVB5R-T to cause illness in mice and rabbits, the two animal systems we commonly employ to evaluate poxvirus pathogenesis. The virulence of RPVB5R-T was first assessed by measuring weight loss and the LD-50 of the mutant in intranasally infected mice. The data of Figure 34 showed that mice infected with RPV rapidly lost weight in a dose dependant manner, as expected. All of the mice infected with RPV in this experiment died as a result of the infection. As was expected, mice infected with similar amounts of RPVAB5R showed little if any weight loss and few signs of illness. Similar results were obtained with the RPVB5R-T group and again, no mice died. As was observed in cell culture and on the CAM, RPVB5R-T behaved in a fashion identical to RPVAB5R in the mouse model. Studies have shown that the activity of several poxviral proteins involved in modifying host immune responses are species specific (5,6,130). Since RPV is most virulent in rabbits, the possibility was considered that any activity possessed by the secreted, truncated B5R-T might be species dependent

PAGE 238

CD § .a bp P 5 TO *o C T— I T3 CD u CO 3 CD U i IT; CO > Ph •2 3 2 T5 C re > 2 c n h CD Qx 1 — 1 r"

PAGE 239

220

PAGE 240

221 and therefore most apparent in rabbits. The data in Figure 35 and in Table 7 show the results of an experiment in which rabbits were infected intradermally with virus and observed for the development of symptoms of infection, as has been described earlier in this dissertation. As previously shown (Figure 17), infection with RPV caused pyrexia within 3 days of infection (Figure 35A) and ultimately resulted in fifty percent lethality within 10 days. In addition, these animals showed many of the expected secondary signs of infection (Table 7). The infection of rabbits with RPVB5R-T (figure 35C) was attenuated and gave results identical to those obtained from the RPVAB5R infected animals (Figure 35B). Neither RPVAB5R or RPVB5R-T infected animals developed any sign of pyrexia during the course of the experiment. As was noted in animals infected with RPVAB5R, a small amount of induration was present at the site of inoculation in RPVB5R-T infected animals early in the infection but quickly resolved by approximately 10 days post-infection. In addition, these animals failed to develop any secondary signs of infection in contrast to what was seen in RPV infected animals (Table 7). In all respects, the phenotype displayed by RPVB5R-T in rabbits was identical to that of RPVAB5R. Construct ion of Mutants Producing only Membrane-bound B5R Protein While the results obtained with RPVB5R-T showed that the membrane-bound form of B5R is necessary for the proper morphogenesis of

PAGE 241

O Oh log o o w m X in | cm iH QJ > O • ~ 1= a £ -S o £ s01 -o H to to p0> 3 L0 10 £ h to 73 s e +• 0 -2 h •S? £ o C £ < to h o tO Oh qJ -o § g J ss s-g c & £ ra < S P o o o £ £ H a 7] a> 1 3 H 12 £ £ > S T3 ai N c to ^ x; boo ^ o n s Uh in biw 3 w O x;

PAGE 242

ro ooooooooo '" M rHrHrHrHrHrHrHrHrH (j 0 ) -duiax ppa^ < OOOOOOOOO r-lrHr-lrHrHrHrHrHrH (Jo) *duiax lepan

PAGE 243

03 03 *H H O CD s01 QJ -t-> ^ g* cfi c ^ 13 CD h-j re s^5 Oi CD 3 y G O O Oh nj £ I .5 ^ ^ l| 1.1 r w hd T3 5^ Oh d LO > 03 u 4CD O X H c EO 03 C re CD X. HHJ CO Oh r-m d o o O -J3 IT) o3 S d d 0) b0 e .3 re u •73 d 01 •3 ry^ — < 09 LO pa Oh o II •d -d OB ^ CD o M 3 5 Mi l Oh 5 S o. t< o LO rt > 5 (3 £4 •a b S a; O CD £ d s 5 oi Z, d O cfl u d +-> w ^ CD d X, o 53 -d bojy d £ o +-> CX co o G H to u CD 73 w re 03 X CD CD 03 >> CD s -d 03 G -t- G_ E GO 224 03 D CD > >(O •g o d ,,H cy p— I 01 6C 03 kiO CD 03 Ol d D 03 (X o Ln > fx s CQ d Oh O LD > cx es 3 <+H Oh O LO > fX CO 3 Oh o LO > (X 3 MH Oh O LT) ID CQ > (X CQ 3 UH Oh O LO H I a; LO CQ > X OS < 3 o o Ln H CSS LD CQ > C* CQ 3 O o LO CtS LO CQ > X Q£ < 3 UH Oh o LO H i ess LO CQ > X — CQ 3 Oh O LO H i OS LO CQ > X 2

PAGE 244

225 the virus particle, previous work in cell culture has shown that a significant amount of B5R protein is secreted from the infected cell by a process that is independent of viral morphogenesis. Poxviruses have been shown to secrete several proteins from the infected cell, many of which share homology with proteins involved in the host immune response and which have been shown to interact with key components of the immune system present in the serum. These observations led to the speculation that the secreted form of the B5R protein might also interact with components of the host immune system, altering the host immune response to the viral infection. While the results of chapter three showed that B5R has no major effect on inflammation in infected animals, it remained possible that B5R does interact with other components of the host immune system, altering the hosts response to the viral infection in a way that has yet to be discovered. To determine if the secreted form of B5R has any role in either plaque formation in cell culture or in viral virulence, it was necessary to construct a mutant virus that produced a functional 45 kDa membrane-bound B5R protein but which was unable to produce a secreted form of the protein. Previous studies from this laboratory have shown that the secreted form of the B5R protein is generated by cleavage of the 45 kDa cellular form, presumably at the cell surface (119). The strategy chosen for the construction a non-secreting mutant was based on the premise that deletion or alteration of the cleavage site would prevent processing of the 45 kDa protein and would therefore prevent production of the secreted form

PAGE 245

226 of the B5R protein. To eliminate the cleavage site, a small in-frame deletion would be created in the region of the protein predicted to contain the cleavage site. Since the exact location of the cleavage site is unknown, several mutants would actually be constructed, each one containing a unique deletion, although all of the deletions would be clustered around a small area thought to contain the actual site of cleavage (see Figure 37). A caveat to this strategy is that it assumes only one cleavage site and further it assumes that no cryptic cleavage sites will be utilized in the absence of the primary site. Before any mutants could be constructed, it was first necessary to determine the location of the B5R cleavage site. An approximate location for this site was estimated based on several pieces of information which are summarized and illustrated in Figure 36. Following our demonstration that RPV produces a secreted form of B5R, vaccinia was then also shown to produce a 38 kDa secreted form of B5R (119) although initial studies failed to detect it. The detection of VV B5R in these initial studies relied on western blot analysis using a monoclonal antibody called C-B5R which was made against a 14 amino acid peptide corresponding to amino acids 262-276 of the B5R protein (Figure 36a). The fact that C-B5R did not detect the secreted form of B5R suggested that the C-B5R antigenic sequence is not contained within the secreted protein implying that cleavage must occur upstream of amino acid 262. To test this hypothesis, the ability of C-B5R to detect the cellular and secreted forms of B5R was compared to that of a polyclonal anti-B5R serum

PAGE 248

229 which has been shown to detect both forms of the protein. A comparison was made by performing a western blot analysis on duplicate blots containing RPV infected cell extracts and supernatants from these cells, using either CB5R or polyclonal anti-B5R serum to detect B5R. The results of this comparison confirmed that indeed, C-B5R fails to detect the secreted form of B5R (Figure 36b). The polyclonal anti-B5R serum detected the 45 kDa membrane-bound protein in the cellular fraction (lane 3) and the 38 kda secreted protein in the supernatant (lane 4) as expected. However, a different pattern of detection was seen when the C-B5R mAb was used. While C-B5R was capable of detecting the cellular form of B5R (lane 1), it failed to detect the secreted form (lane 2) confirming that amino acids 262-276 are not contained in the secreted form of B5R. From these data, it can be concluded that the cleavage site must be located upstream of amino acid 262. The data obtained so far only excludes a small region of the protein in which the cleavage site is probably not located. Further estimation of the location of this site was based on the difference in molecular weights between the various forms of B5R. Careful measurement and comparison of the apparent molecular weights of the cellular and secreted forms of B5R revealed a difference of approximately 6.0 kDa. Assuming that the shift in size results solely from removal of the carboxyl terminal fragment of the protein, this difference represents removal of approximately 51 amino acids from the carboxyl end of the 45 kDa protein, placing the cleavage site at aa 266.

PAGE 249

230 A similar comparison of the difference in molecular weights between secreted B5R-T, which is truncated at amino acid 275, and the wild-type secreted protein shows that B5R-T is approximately 1.7 kDa larger than the wild-type secreted B5R protein. This difference translates into B5R-T having approximately 14 additional amino acids compared with the wild-type secreted B5R protein. By counting back 14 amino acids from the terminal residue (amino acid 275) of the B5R-T protein, the cleavage site was predicted to be at amino acid 262. From the combined analyses, the cleavage site was estimated to be located around amino acids 262-266. However, because of the inherent difficulty in accurately determining the molecular weight of glycoproteins by PAGE analysis, it was recognized that the actual site of cleavage might be located somewhere upstream of amino acid 262. The majority of B5R is composed of the four SCR domains (approximately 60 amino acids each) which extend from amino acid 21 to amino acid 237. Since it was thought unlikely that the cleavage site would be located within an SCR region it was hypothesized that the site of cleavage would lie between amino acids 238 and 266, a region that will be referred to as the linker region. Based on these estimations and assumptions, a series of mutant B5R genes was constructed such that each gene contained a small in-frame deletion of 12-36 nucleotides (corresponding to 4-12 amino acids) within the coding sequence for the linker region (Details of this construction are provided in Materials and Methods). A schematic map showing the location

PAGE 250

231 of the deletions in these mutants is shown in Figure 37 A while detailed information about the actual amino acids deleted can be found in Figure 11 in the Materials and Methods chapter. Individual constructs differed in the relative location of their deletion which were designed to be slightly offset and overlapping between the various mutants. In this way determination of the effect of deletions throughout the entire linker region could be made by the construction of only a few mutants. The engineered B5R genes, under control of the native B5R promoter, were then recombined into the thymidine kinase gene of RPVAB5R resulting in a series of mutants designated RPVB5R:X where X refers to the region covered by the deletion within the B5R gene (diagrammed in Figure 37A). As a control, a wild type copy of the B5R ORF was also inserted into the TK gene resulting in a mutant referred to as RPVB5R:WT. The mutants were analyzed for their ability to produce the cellular and secreted forms of B5R. Cells were infected with either RPV, RPVAB5R (as controls) or one of the RPVB5R:X mutants and a western blot analysis performed on cell extracts and supernatants from these cells. The results of this analysis are shown in Figure 37B. As expected, a 45 kda and a 38 kda protein were detected in the cell and supernatant fractions, respectively, from RPV infected cells (lanes 1 and 2). These proteins were absent from the cell extract and supernatant obtained from cells infected with RPVAB5R (lanes 3

PAGE 251

SX£ PQ

PAGE 252

233 D Z Z Qjj u:

PAGE 253

234 and 4), the parent virus of the RPVB5R:X mutants. All of the RPVB5R:X mutants tested produced the 45 kDa cellular form of B5R indicating they contained an active copy of the B5R gene. It was noted that the levels of B5R protein observed in extracts from all of the RPVB5R:X infected cells was somewhat reduced in comparison to the levels observed in RPV infected cells. However, this apparently had no obvious effect on the efficiency of cleavage since the ratio of secreted to cellular protein in RPVB5R:wt and RPVB5R:G appeared similar to that observed for RPV. Although all of the RPVB5R:X infected cells produced B5R only cells infected with RPVB5R:WT (which has a copy of the wild-type B5R gene inserted into the TK gene) and RPVB5R:G had B5R in their supernatants. The size of the protein secreted by these mutants was identical to that of wild-type B5R suggesting that the protein was being cleaved at the natural cleavage site. Cells infected with any of the other recombinants failed to secrete the B5R protein suggesting that the B5R genes contained by these mutants has lost their cleavage sites. Based on these results and on the arguments above, in can be concluded that cleavage of B5R is likely to occur between amino acid 239 (the first amino acid downstream of SCR IV) and amino acid 261 (the lysine missing in RPVB5R:G). Plaque Size and Host Range of B5R Non-secretors Previous work has shown that mutants unable to produce the B5R

PAGE 254

235 protein or which produce solely the secreted form of B5R have a reduced plaque size on RK-13 cells. In addition, such mutants are unable to form plaques on CEF cells a phenotype that correlates with and is presumably due to the inability of this class of mutants to produce enveloped virus in these cells. The experiments described above examining the properties of RPVB5RT indicated that the membrane bound form of B5R is necessary for normal plaque formation and host range expression. To determine if the secreted form of B5R has any effect on plaque formation in cell culture, the ability of the non-secreting mutants to plaque on RK-13 and CEF cells was determined. These results shown in Figure 38 (RK-13 cells) and Figure 39 (CEF cells). On RK-13 cells, RPVAB5R produced plaques smaller in size ( Figure 38b) than those produced by RPV (Figure 38a) as has been previously demonstrated. Insertion of a wild-type copy of the B5R gene into the TK gene of RPVAB5R (RPVB5R:WT) as a control for insertion of the mutant genes into the TK site resulted in the production of plaques similar, although slightly smaller, in size to those produced by RPV(Figure 38c). This control is necessary since all of the RPVB5R:X mutants are also TKas a result of the construction scheme. When tested on RK-13 cells, all of the remaining RPVB5R:X mutants produced plaques identical to those produced by RPVB5R:WT. The ability of the RPVB5R:X mutants to form plaques on CEF cells was also tested and the results are shown in Figure 39. While RPV normally

PAGE 255

.2 £ U u (0 rz

PAGE 256

237

PAGE 257

Figure 39. Plaque formation on CEF cells by the RPVB5R:X mutants. CEF cells were infected at a low m.o.i. and overlaid with methylcellulose. After four days of incubation at 37C, the overlay was removed and the cells were stained with crystal violet, panel a-RPV, panel b-RPVAB5R, panel c-RPV TK(), panel dRPVB5R:WT, panel e-RPVB5R:A, panel f-RPVB5R:B, panel gRPVB5RC, panel h-RPVB5R:E, panel i-RPVB5R:G. Plaques are shown at 40X magnification.

PAGE 259

240 forms large plaques in CEF monolayers (Figure 39a), deletion of the B5R gene results in a mutant (RPVAB5R ) which is unable to form plaques at all(Figure 39b) on these cells. The ability of this mutant to form plaques in CEF cells was restored by insertion of the B5R gene into the RPVAB5R TK gene, although these plaques were greatly reduced in size in comparison to RPV plaques (compare Figure 39a and 39d). However, the plaques produced by RPVB5R:WT were identical in size to those produced by an RPV TK-( null mutant) (Figure 39c) which were also much smaller then wild-type plaques. As was observed in RK-13 cells, failure to secrete the B5R protein had no effect on plaque formation. All of the RPVB5R:X mutants produced plaques in CEF cells and these plaques were identical in size to those produced by the RPV TK(-) virus. Pock Morphology on the CAM in the Absence of Secreted B5R Protein Previous work has shown that a mutant unable to produce either form of B5R (membrane associated or secreted) produced a white pock on the CAM, in contrast to the red, hemorrhagic pocks normally produced by wild-type virus. Work with the mutant RPVB5R-T also showed that failure to produce the membrane-bound form of B5R also generated white pocks, even though these mutants overproduced a secreted form of the protein. To determine the role, if any, of the secreted form of the B5R protein in the production of red pocks, the pock morphology of the various B5R non-secreting mutants on the

PAGE 260

241 CAM was examined. The results of this analysis are shown in Figure 40. Infection of the CAM with either RPV (Figure 40b) or RPVAB5R (Figure 40d) produced the expected red and white pocks, respectively, as was described in chapter three. The RPV TKmutant also produced red pocks, however these pocks appeared less red than the RPV pocks (Figure 40c). Insertion of the wild-type B5R gene into the TK gene of RPVAB5R restored the ability of the RPVAB5R mutant to produce red pocks (Figure 40e) as expected. Consistent with what was observed in cell culture, failure to produce the secreted form of B5R had no effect on the viruses ability to produce red pocks on the CAM. All of the non-secreting mutants produced red pocks similar to those produced by RPV TK(-). Discussion The B5R protein of RPV exists in two forms, each of which may be assigned to distinct and seemingly diametric groups of viral proteins. One form of the B5R protein is a 45 kDa membrane-bound glycoprotein found in the envelope of the EEV particle. A second form of B5R is secreted into the supernatant as a smaller 38 kDa protein. The membrane-bound form of B5R is one of six outer envelope proteins found in EEV. A feature characteristic of all but one of these proteins (the hemagglutinin protein) is that while they are not necessary for

PAGE 261

Figure 40. Pock morphology of the RPVB5R:X mutants. 11 day old embryonated chicken eggs were inoculated with either PBS (panel a), RPV (panel b), RPV TK(-) (panel c), RPVAB5R (panel d), RPVB5R:WT (panel e), RPVB5R:A (panel f), RPVB5R:B (panel g), RPVB5R:C (panel h), RPVB5R:E (panel i), RPVB5R:G (panel j) and incubated for 72 hours after which time the CAMs were removed and examined.

PAGE 263

244 morphogenesis of IMV, they are essential at least under some conditions for the proper morphogenesis of the EEV particle. The secreted form of B5R, which is derived from the 45 kDa form by proteolytic processing, is one of at least five secreted proteins produced by the poxviruses. A characteristic shared by many of these secreted proteins, including B5R, is that they share homology to proteins involved in regulation of the host immune response. In many cases these viral proteins have been shown to interact with the host immune system in a manner predicted by their homologies. The existence of two forms of the B5R protein has resulted in speculation that each form may have a unique function, however, assignment of a unique function to either the membrane-bound or the secreted from of B5R has been difficult since deletion of the B5R gene results in failure to make either form of the protein. To explore the possibility that the individual forms of B5R may each have a unique function, mutants were constructed that produced either the membrane-bound form or the secreted form of B5R but not both. These mutants were examined in cell culture and in animals to determine any phenotype associated with each form of B5R. A mutant, referred to as RPVB5R-T, was constructed which produces only the secreted form of B5R. In this mutant, the coding sequence for the membrane anchor of the B5R protein has been removed. As a result, RPVB5R-T produces a truncated 40 kDa B5R protein (referred to as B5R-T)

PAGE 264

245 which lacks the membrane anchor region. As was expected, loss of this anchor sequence resulted in a significant increase in the amount of B5R-T that was secreted from infected CEF cells demonstrating that the membrane anchor normally keeps a significant portion of B5R in association with the cell. Interestingly, the apparent molecular weight of B5R-T was unchanged upon secretion and as a result, the secreted B5R-T protein was larger than the secreted wild-type B5R protein. The difference in the sizes of the secreted forms of the wild type B5R and the B5R-T proteins was not a result of aberrant glycosylation since it was shown that removal of the sugars from both B5R and B5R-T resulted in a similar shift in size for both of these proteins. The simplest interpretation of these results is that the site at which processing normally occurs in B5R is upstream of the position at which B5RT was truncated and that processing does not occur with B5R-T. The results obtained using the mAb C-B5R show this to be true. One would then also predict that the B5R protein must be anchored in a membrane for processing to occur. Coupled with the observation that the 38 kda form of B5R is normally found in supernatant but not cellular fractions from infected cells, these data suggest that processing of B5R occurs on the cytoplasmic membrane. A major caveat of this work is the failure of B5R-T to be properly processed to the correct size for secreted B5R protein. While the data from the non-secreting mutants support the idea that secreted B5R has no activity, it is entirely possible that the extra amino acids on the secreted B5R-T protein

PAGE 265

246 could be inhibiting any possible small amount of activity this form of the B5R protein might possess. The B5R protein has been shown to be one of several glycoproteins that localize to and are anchored in the membrane of the EEV particle. This association with EEV is presumed to be mediated by the hydrophobic membrane anchor region near the carboxyl terminus of the protein. Surprisingly lack of a membrane anchor failed to prevent B5R-T from localizing to the EEV particle. It is possible that the amino acids which fail to be removed from B5R-T and result in the unexpected increase in size compared to the wild type B5R protein are responsible for this retention; however, there are no further data to support this hypothesis at the moment. Nevertheless, biochemical analysis revealed that the nature of the association of the truncated protein with EEV was significantly different than that of wild-type B5R. While wild-type B5R was firmly anchored in the EEV envelope, B5R-T was shown to be loosely associated with the virion and unlike wild type B5R could be removed under conditions known to remove non-integral membrane proteins. The wild-type B5R protein remained firmly associated under these same conditions. These results are consistent with B5R-T lacking a membrane anchor with which to properly associate with the virion. The loose association of the truncated protein with the virion suggests that this protein may be binding, not to the membrane, but to one of the other EEV glycoproteins. It is likely that such an interaction

PAGE 266

247 originates in the trans-golgi compartment, where EEV glycoproteins have been shown to accumulate, and that B5R-T is carried to the virion as a result of this association. Indeed, earlier work by others has indicated that B5R can be co-precipitated with p37, a major EEV envelope protein (119). A demonstration of the ability of B5R-T to bind other proteins in a manner analogous to wild-type B5R would provide further evidence that removal of the membrane anchor has not significantly altered the overall structure and properties of the protein. In cell culture, the B5R null mutant RPVAB5R, which is unable to produce either form of B5R, has a smaller plaque morphology on RK-13 cells and produces no plaques on CEF cells. When RPVB5R-T was examined in these cells, it also produced small plaques on RK-13 cells and failed to form plaques on CEF cells identical to the pattern observed with RPVAB5R. Furthermore, investigation of the cause of this host range restriction revealed that this mutant failed to produce enveloped virus in CEF cells, a phenotype identical to the defect observed in RPVAB5R infected CEF cells. Together these results indicate that the membrane bound form of B5R is essential for the proper morphogenesis of enveloped virus in cell culture. It is interesting that although the B5R-T is properly trafficked to the EEV particle, it is functionally inactive in the sense that the association is inadequate to allow EEV formation. This result suggests that the B5R protein must be firmly anchored in the envelope for proper function.

PAGE 267

248 The pocks produced by RPVB5R-T were examined on the CAM, and virulence was evaluated in both mice and rabbits. On the CAM, RPVB5R-T produced white-pocks that appeared identical to those produced by the RPVAB5R mutant. In addition, the RPVB5R-T mutant was attenuated in both mice and rabbits, failing to cause any secondary signs of infection, as was also observed with RPVAB5R. These results show that the membrane bound form is necessary for full RPV virulence in these systems. Given the defect in morphogenesis observed in cell culture, this result is not surprising. Since RPVB5R-T and RPVAB5R behave identically in cell culture it is likely that they are attenuated for the same reason, most likely as a result of their inability to produce enveloped virus. To determine if secreted B5R is a necessary form of the protein, a series of mutants was constructed in which each strain of virus contains a small overlapping, in-frame deletions in the region of the protein calculated to contain the cleavage site. This region, which we refer to as the linker legion, is located between amino acids 238 and 260. The result of these constructions was several mutants that produced solely the 45 kDa form of the protein but which were unable to produce the secreted form. Based on the outcome of these constructions several points can be made. First, it can be concluded that the region encompassing amino acids 238-260 is involved in proteolytic processing of B5R since deletions anywhere within this region prevented processing. It is unlikely that failure of the protein to be processed is simply

PAGE 268

249 due to a general perturbance of the protein structure since deletions within this area failed to prevent processing (see the data for RPVB5R:G). It can also be concluded that the actual site of cleavage resides upstream of amino acid 260 since deletions downstream of this site have no effect on cleavage. Additionally, sequences downstream of this site are not present in the secreted form of B5R providing further evidence that the site of cleavage is upstream of amino acid 260. Since it is unlikely that the processing site would reside within the SCR regions (which end at amino acid 237), it is predicted that the actual cleavage site will be located between amino acids 238 and 260. Failure to secrete B5R had little effect on plaque morphology and host range in cell culture. In RK-13 cells, non-secreting mutants produced plaques similar in size to RPV, although these plaques appeared slightly smaller than the wild-type plaques. We believe that this somewhat smaller plaque size can be attributed to the fact that all of the RPVB5R:X mutants lack a functional TK gene. Also, all of these mutants, including one carrying a wild type copy of the B5R ORF, produce lower levels of the B5R protein which may have an effect on plaque size. Nonetheless, all of the non-secreting mutants produced plaques identical to those produced by the recombinant carrying a wild type copy of the B5R ORF confirming that the secreted form of B5R has no effect on plaque production in this cell line. In addition to producing wild-type plaques on RK-13 cells, all of the non-secreting mutants were also able to

PAGE 269

250 form plaques on CEF cells although, it should be noted that these plaques were significantly smaller than those produced by RPV. Again, the differences are believed to result from interruption of the thymidine kinase gene of the virus since the small plaques produced by the non-secreting mutants were identical in size to plaques produced by RPV TK null mutants. From these results it was concluded that the secreted form of B5R has no significant role in virus replication in cell culture. Similar results were obtained on the CAM. While RPVAB5R has been shown to produce white pocks on the CAM, insertion of an ORF encoding a non-cleavable B5R protein into the TK gene of RPVAB5R restored the ability of RPVAB5R to form red pocks. Again, although these pocks were less red than those produced by RPV, this was attributed to interruption of the TK gene. This result shows that the secreted form of B5R does not alter the host's response to the virus and further, this protein has no discernible effect in this system. The data from cell culture and from the CAM indicated that the secreted form of B5R has no detectable activity in these systems. However, since the proteins with which B5R shares homology (complement factor H and factor C4b binding protein) are mammalian proteins that are involved in the complex regulation the immune system, it was thought that the secreted form of the protein might have some activity that would only be manifest during infection of an animal. To test this, the RPVB5R:X mutants were

PAGE 270

251 tested in mice and rabbits for their ability to cause disease. It was only after these experiments were performed that the TK gene, which was previously thought to be a non-essential gene in RPV and was therefore used as the insertion site for the mutated ORFs, was found to be essential for virulence in rabbits and mice. All of the RPVB5R:X mutants were extremely attenuated in these animals including RPVB5R:X which contains a wild type copy of the B5R ORF in the TK gene. This severe attenuation was also shown by the RPV TK null mutant, indicating that the attenuation of the non-secreting mutants could be attributed to the TK mutation in these viruses. As a result of this rather ugly fact, conclusions about the role of the secreted form of B5R in animals cannot be made using the RPVB5R:X mutants described here. Such conclusions will have to wait until a second generation of these mutants, which will have the mutant ORFs described inserted into a non-essential site, are available. In summary, the work presented in this chapter shows that although there are two forms of B5R, only the membrane-bound form appears to have any function in cell culture and in virulence. Failure to produce membrane bound B5R results in a cell line-dependent defect in the morphogenesis of enveloped virus as well as attenuation of the virus, a phenotype identical to that observed for B5R null mutants. In contrast, failure to produce the secreted form of B5R had no effect in cell culture or on pock morphology in the CAM. As a result of this work, a region of the protein has also been

PAGE 271

252 identified that is involved the processing of the cellular form of B5R to generate the secreted form of B5R. This information will prove useful in the construction of B5R mutants and may provide insight into the interactions of this protein with those of the host.

PAGE 272

CHAPTER V SUMMARY, DISCUSSION AND FUTURE DIRECTIONS Summary The poxviruses have evolved to exist as obligate intracellular parasites, and therefore rely on the replication "machinery" of the host cell for propagation of the viral genome. A consequence of this dependency is that the viruses are constantly in danger of being detected and eliminated by the host immune system. The survival of these organisms to the present day suggests that they have evolved methods for contending with the host response to the viral infection allowing them to propagate relatively unhindered. In recent years work in this field has shown that the poxviruses do indeed produce several proteins that are capable of altering the host immune response (6,36,121). An interesting feature of many of these proteins is that they share homology with proteins found in the mammalian immune system suggesting that the poxviruses have evolved a method for "capturing" host genes that increase their survival advantage. Experimental evidence has shown that although the sequence of these viral proteins is usually somewhat divergent from the host homologue, they frequently retain activity 253

PAGE 273

254 identical or similar to that possessed by the host protein. To date, examples of such viral proteins include a secreted IL-16 receptor (7,183,187), a secreted TNF receptor (97,178,182,204), a secreted IFN-gamma receptor (130,131,194,205), and a secreted complement factor C4b binding protein (vC4b bp) (99,110,111). These virus encoded proteins have all been shown to have the activity suggested by their homology. The product of the B5R gene, which is a 45 kDa glycoprotein found in the envelope of the EEV particle, has homology with complement factor H and the C4b binding protein, two proteins that prevent activation of the complement cascade (57,119,195). Since one consequence of complement activation is the elicitation of the inflammatory response, it was theorized that the B5R protein might affect virus virulence by inhibiting complement activation and as a consequence, inhibiting the host inflammatory response. The feasibility of this hypothesis was increased with the discovery that a 38 kDa form of the B5R protein is secreted from the infected cell suggesting that it might interact with soluble components of the immune system in a manner similar to the previously described vC4b binding protein (110,119). The present work was undertaken to test the hypothesis that the B5R protein affects virus virulence by suppression of the host inflammatory response. This hypothesis was tested by construction of a null mutant (RPVAB5R) and analysis of this mutant in cell culture and in animals. In addition, the role of the individual forms of B5R was evaluated by construction and analysis of

PAGE 274

255 mutants that produced either the 45 kDa membrane-bound form or the secreted form of the protein but not both. RPVAB5R was shown to be severely attenuated in both rabbits and mice and in the latter, this attenuation was shown to correlate with a decrease in virus production. Attenuation of this mutant was initially suspected to result from failure to inhibit inflammation since on the CAM, in contrast to the red, hemorrhagic pocks produced by RPV, RPVAB5R formed white pocks similar to those produced by CPVASPI-2, a mutant known to be crippled in its ability to inhibit inflammation (142,155,158). More specifically, the RPVAB5R white pocks, like those of CPVASPI-2 and unlike those of RPV, were shown to contain a large number of heterophils indicating that the B5R protein is necessary for inhibition of inflammation in this system. While these observations seemed to offer a ready explanation for the attenuation of this mutant in animals, further analysis of the behavior of RPVAB5R revealed that in fact, attenuation occurred independently of any inflammatory response. Histological examination of the inoculation site following intradermal infection of rabbits with RPVAB5R showed that unlike what was observed on the CAM, there was no change in the degree of inflammatory cell infiltrate in these tissues when compared with RPV infected tissues. While a mild increase in the level of neutrophils was observed in mice, this increase was small and was thought to be too insignificant to result in the dramatic attenuation observed with the mutant. The non-involvement of

PAGE 275

256 inflammation in the attenuation of RPVAB5R in mice was confirmed with the demonstration that inhibition of inflammation by injection of the steroid dexamethasone failed to increase virus yields or virus virulence in RPVAB5R infected mice. Collectively, these results show that while deletion of the B5R ORF results in a decrease in the ability of the virus to grow in animals resulting in attenuation, this defect is independent of the host response suggesting that it is innate to the virus. I believe that the simplest interpretation of this data is that while B5R may have some small influence on inflammation, the severe attenuation of RPVAB5R is fundamentally due to failure of the mutant to spread efficiently from initially infected cells due to a B5R-related defect in the ability of this mutant to produce enveloped virus. While I have been unable to obtain direct evidence of this defect in animal tissue, this defect was characterized in cell culture and will be discussed below in detail. In support of this interpretation, it is worth noting that when animals are infected with a RPV null mutant of SPI-2, a protein which is known to inhibit inflammation, the resulting symptoms and pathology are nearly identical to that produced by RPV (Brooks and Stern, Manuscript in preparation). This observation suggests that inflammation plays only a minor role in combating the virus infection in animals and demonstrates that failure to suppress inflammation alone does not result in the severe attenuation observed with the RPVAB5R mutant. Such severe attenuation can be easily accounted for, however, by failure of the virus to replicate and

PAGE 276

spread in the animal. With regard to this interpretation, it was also found that viruses failing to produce the membrane-bound form of B5R, but still able to produce the secreted form, were defective in morphogenesis and were also attenuated. Since one would expected that it is the membrane-bound form of B5R that is involved in morphogenesis, these results are consistent with a defect in morphogenesis being responsible for attenuation of these mutants. Finally in this vein, attenuation has been shown to be a common feature of mutants defective in the production of enveloped virus demonstrating that production of enveloped virus is essential for full virulence in animals (58,143). The results discussed above show that inflammation is not involved in attenuation of RPVAB5R and further, are strongly consistent with attenuation of RPVAB5R being attributable to an innate growth defect of the mutant. Although attempts to demonstrate such a defect in animal tissue were unsuccessful due to the enormous technical obstacles, such a defect was clearly demonstrated and characterized in cell culture. Analysis of RPVAB5R in cell culture revealed that the B5R protein of RPV is a necessary component in the production of enveloped virus. Failure to produce B5R resulted in a defect in enveloped virus morphogenesis in all cells although the effect was more pronounced in some cells than in others. In RK-13 cells, RPVAB5R had a reduced plaque size when compared to RPV, which correlated with a decrease in the amount of IEV that was produced. These results are

PAGE 277

258 consistent with the work of others suggesting that it is IEV that is responsible for direct cell to cell spread and plaque formation (21). In CEF cells, RPVAB5R was unable to produce any enveloped virus particles and was unable to form plaques on monolayers of these cells. These results are also consistent with the work of others showing that production of IMV alone is not sufficient to produce plaques in cell culture (20,21,55,177). Collectively, these results demonstrate that B5R plays an important role in the production of enveloped virus and in fact is essential for this process in some cells. In this regard, the B5R protein is similar to several other poxvirus envelope proteins, such as the p37 and p43-50 proteins of vaccinia, which have also been shown to be essential for the morphogenesis of the enveloped form of the virus (20,143,177). However, B5R differs from the other enveloped proteins in that while the majority of these proteins are absolutely necessary for EEV production regardless of the cell line, the necessity for B5R varies depending on the cell line used. This suggests that B5R must interact with a host protein to carry out its normal function making it unique among the envelope proteins and suggesting that it may play some central role in the assembly or transport of the virion. Investigations into the fate of the other enveloped proteins in the absence of B5R in permissive and non-permissive cells would be revealing in this regard. While the exact role played by B5R in virion morphogenesis is unclear, a possible role for its involvement in morphogenesis can be proposed using

PAGE 278

259 the following observations and arguments. First in RPVAB5R infected RK-13 cells, it was observed that there is a dramatic decrease in the levels of IEV but no change in the levels of EEV. This is somewhat surprising considering the currently accepted model for virion morphogenesis in which the virus particle progressively matures from IMV to IEV to EEV so that IEV serves as an intermediate between the other two forms (128). From this model, one might assume that alterations in the production of IEV should result in downstream effects on EEV. However, the data presented here indicate that this is not necessarily so. If one assumes the current model for morphogenesis to be correct, then a viable interpretation of these data is that B5R effects the rate at which virions either enter or leave the IEV stage of morphogenesis. I believe that B5R is likely to affect entry, not egress, of virions into the IEV pool based on the following observations and speculation. If B5R affected egress from the IEV pool, then the decrease in IEV observed in the absence of B5R would indicate that virions are being moved out of this pool faster than normal. However, this would result in an increase in the level of EEV, an event that is not observed in these cells. This suggests that the inability of RPVAB5R to produce IEV is at the level of production of this form of virus. Since IMV formation is apparently unaffected in RPVAB5R infected RK-13 cells, this suggest that the defect in morphogenesis is at the level of wrapping of the IMV particle. The results obtained in CEFs are consistent with this interpretation, since the lack of B5R

PAGE 279

260 resulted in a failure to produce any enveloped virus suggesting that B5R is involved in the earliest steps of enveloped virus synthesis. These observations and speculations lead me to hypothesize two separate models for the involvement of B5R in enveloped virus morphogenesis. The first and most straight forward suggests that the need for B5R in enveloped virus formation is simply as a structural component and that in its absence, the wrapping of IMV is aberrant and non-productive. In this model, one would have to assume that in permissive cells, some host function can substitute for B5R allowing IEV production although somewhat inefficiently. In nonpermissive cells, this protein would be missing resulting in an inability to produce enveloped virus. While this model has simplicity on its side, it seems unlikely that some cells would fortuitously produce a protein that can provide B5R function. A second, more complex but equally valid model involves interactions between B5R and other viral proteins and one or more host factors. In this model, B5R, located in the golgi, would play a central role in wrapping of IMV to form IEV. Since in the absence of B5R, IEV production is abolished in some cells but can occur in other cells, there is obviously some B5R-host factor interaction that is not entirely essential for morphogenesis. Based on this, I suggest that the B5R-host factor interaction is not the key interaction in the actual mechanics of virion wrapping but instead, that B5R facilitates and stabilizes the interaction of other viral envelope proteins with the host factor. Further, I suggest that in the absence of B5R, these

PAGE 280

261 interactions are unstable and are therefore somewhat inefficient, the result being a slow production of IEV. However, I believe that once this particle is formed, it is rapidly shuttled out of the cell to yield EEV, hence the low levels of IEV and the normal levels of EEV. An assumption of this model is that the IEV to EEV throughput rate is fixed and that normal IEV production exceeds this capacity resulting in an intracellular store of wrapped virus. In non-permissive cells, the host protein essential for this interaction is either missing or it is significantly different from its homolog in permissive cells that its interactions with virus proteins are so unstable that no enveloped virus is made. In regard to this model, it has been reported that B5R can be co-immunoprecipitated with p37, another virion envelope protein, although the specific details and significance of this interaction are not known (147). Still, p37 is known to interact with an IMV membrane protein called pl4, and this interaction is apparently essential in the formation of enveloped virus (165). These data suggest that it is not individual membrane proteins but rather complexes of these proteins that are important in the morphogenesis of enveloped virus. The results obtained from infection of the CAM with RPVAB5R merit some discussion. As has been mentioned above, the white pocks produced by RPVAB5R on the CAM were similar in appearance to the pocks produced by CPVASPI-2 (and RPVASPI-2) which are known to result as a direct consequence of this mutants failure to inhibit inflammation (142,155,158).

PAGE 281

262 This result indicated that B5R does indeed play some significant role in the inhibition of inflammation even though no significant effect was observed in infected animals. In this regard, it is important to note that work with RPVASPI-2 has shown the effect of inflammation on virus virulence in animals to be minimal meaning that small effects of B5R on inflammation may be missed. In contrast, in the 14 day old embryo where complement and inflammation are the only host defenses available, effects on inflammation are likely to have a significant impact on virus replication as has been shown with CPVASPI-2 and RPVASPI-2. To determine if the RPVAB5R white pocks, like those of RPVASPI-2, resulted from failure to inhibit inflammation a more extensive analysis of the RPVAB5R pocks was performed. While the results showed that B5R is indeed necessary for suppression of the inflammatory response in the CAM, several curious observations suggested that, unlike the SPI-2 protein, its involvement may not be so straightforward. Histological examination of the lesions from infected CAMs confirmed that the RPVAB5R white pocks, like those of RPVASPI-2, contained a large number of heterophils providing direct proof that B5R is indeed necessary for inhibition of the inflammatory cell infiltrate. However, unlike the heterophils found in RPVASPI-2 pocks, the heterophils present in RPVAB5R white pocks were unable to reduce the chromogenic substrate nitroblue tetrazolium (NBT). Since NBT is reduced to formazan by oxygen radicals generated by activated heterophils within the pocks, this result is taken to

PAGE 282

263 mean that heterophils found within RPVAB5R white pocks are present in an inactivated state. This differs significantly from the RPVASPI-2 white pocks in which the heterophils are activated as determined by their ability to reduce NBT. As was mentioned in chapter three, the migration and activation of heterophils are two distinct events that occur during inflammation. These events have been shown to differ in their response to different soluble factors or to different concentrations of the same factor(s) (126,170,171,179). Migration of these cells is sensitive to relatively low levels of factors (such as leukotriene B4or platelet activating factor) whereas activation requires much higher levels of these same molecules presumably as a safeguard against activation at an inappropriate location. In addition, activation of the oxidative burst phenomenon can also be triggered by the initiation of the engulfment process by these cells. In view of this, the data obtained from the CAM was interpreted to mean that either (1) B5R normally inhibits the production or action of a migration specific factor or (2) that the inflammatory response to RPVAB5R differs from that generated against RPVASPI-2 only in its severity meaning that RPVAB5R fails to induce a significant response. The observation that vasodilation of blood vessels which can be seen in RPV and RPVASPI-2 pocks is absent from the RPVAB5R white pocks leads me to favor he second of these hypotheses. If the first hypothesis were correct, this would mean that both migration and vasodilation must be under control of a single factor. Since several factors are known to affect these activities, I think it

PAGE 283

264 unlikely that B5R directly inhibits heterophil migration. Along these lines, it should also be noted that if B5R did affect inflammation, one would expected it to do so by preventing the generation of pro-inflammatory factors derived from complement components. However, since these molecules are also known to be potent stimulators of neutrophil activation (126) it is unlikely that B5R inhibits the generation of complement-derived mediators of inflammation. The simplest explanation for the results on the CAM is that infection with RPVAB5R elicits only a mild inflammatory response. In this view, damage from the infection would be provide enough stimulus to cause heterophil migration to the site but not enough to cause detectable activation. As for the cause of this limited damage, we are brought full circle to the defect in enveloped virus production possessed by RPVAB5R. Although there is no proof that this defect occurs in the CAM, the fact that the CEF cells in which the defect was first shown come from chicken embryos makes it a near certainty. Failure of this mutant to produce EEV in the CAM would result in extremely limited spread and as a result limited damage. This is sure to elicit some inflammatory response although again, the limited stimulus would produce only a small response. Compounding the situation would be the fact that only the initially infected cells, and very few others, would become infected and produce virus proteins, meaning that the local concentration of virus-encoded anti-inflammatory proteins would likely be lower, possibly too low to inhibit inflammation. Consistent with this view, inhibition of

PAGE 284

265 inflammation in the CAM with dexamethasone failed to convert the RPVAB5R pocks to red indicating that although an inflammatory cell infiltrate was present in the white pocks, it is not the sole or even the primary cause of this phenotype. Since the red color observed in RPV pocks is due to hemorrhage, these results can be interpreted to mean that even in the absence of inflammation, RPVAB5R fails to produce enough damage to cause hemorrhage. In the absence of hemorrhage, the pock appears white simply due to the thickening of the membrane due to the virus-encoded growth factor. Again, the most consistent explanation for all of these results is a defect in enveloped virus production due to a defect in a B5R dependent pathway. The work presented in this dissertation has shown that the phenotype of the virus is altered dramatically when it is no longer able to produce either the membrane bound or the secreted form of B5R. A significant finding from this work is that the activity of B5R, at least in cell culture and in the egg system, can be attributed solely to the membrane-bound form. Mutants unable to produce secreted B5R but producing the membrane-bound form, displayed a phenotype identical to that of RPV. Conversely, mutants lacking the membrane-bound form, but producing secreted form, were identical in phenotype to RPVAB5R. A caveat that must be kept in mind, however, is that the secreted form of B5R produce by RPVB5R-T (which does not produce membrane-bound B5R protein) contains several extra (approximately 20)

PAGE 285

266 amino acids compared with the wild-type secreted form and may be inactive as a result. However, the fact that mutants failing to produce the secreted form behave as wild-type virus indicates that the wild-type secreted form has no detectable activity at least in these systems. While mutant viruses that fail to produce the secreted form of B5R were tested in animals, the results were un-informative in regard to B5R since the scheme used to construct these mutants resulted in interruption of the thymidine kinase (TK) ORF. Empirically derived data showed that deletion of the TK protein resulted in complete attenuation of RPV as has been shown for vaccinia virus (34,196). As an interesting aside, it should be noted that TK null mutants also produced small plaques on CEF cells (which are primary cells) but not on other cells suggesting that primary cells in culture may be useful for predicting the behavior of mutations on virus growth in animals. Determination of the role of B5R in animal infections will have to await the construction of new mutants in which only the B5R ORF is affected. Assuming these results agree with those obtained in cell culture and in the egg, they should prove valuable to those who in the future may seek to understand this protein since it will enable them to construct hypotheses based on the activity of this protein occurring on the membrane. If these results are correct and the secreted form of B5R has no activity it leads one to wonder why this protein is cleaved to produce the secreted form. It is possible that progenitor protein for B5R contained this cleavage

PAGE 286

267 site for some purpose and the cleavage event was neither beneficial nor harmful and so the cleavage event remained unaltered. However, if this were true, one might expect that random mutations would lead to different species of poxvirus differing in their sequences in the region around the cleavage site. However, alignment of the region predicted, by this work, to be involved in cleavage shows that the B5R proteins from the various viruses are completely conserved throughout this region implying strong selective pressure in favor of cleavage. A caveat of this comparison is that the alignments shown in this work compares only the closely related orthopoxviruses since these are the only viruses for which B5R sequence is available. If the secreted form of B5R has no function, why is the cleavage event selected for? Looking at this puzzle another way, one could argue that it is not generation of the secreted form but rather removal of the membranebound form from the surface of the cell that drives the selection of this event. On the question of why this would be advantageous, two possibilities immediately come to mind. The first is related to the fact that viral replication occurs within the cell where the virus remains safely hidden from the host immune system for a period of time. Viral proteins displayed on the cell surface, which B5R is known to be, reveal the presence of the virus to the host and provide antigenic stimulation for the immune system. The obvious result of this event would select for viruses better able to mask their presence by restricting the amount of viral proteins that are displayed on the surface.

PAGE 287

268 A second possibility is that B5R located on the infected cell surface may bind to other viral proteins and keep the virus particle from spreading from the primary infected cell. This would restrict the virus to one location making it easier for the immune system to eliminate the infection. Again, survival would favor viruses that are able to spread to distal sites making neutralization of the infection more difficult. In regard to the latter possibility, it is believed that the B5R protein of vaccinia binds to the EEV envelope protein p37 since the two can be co-immunoprecipitated (147). While the work presented here has demonstrated the effects of deleting the B5R protein, it still leaves the open the grand question, namely, what is the true function of the B5R protein? It is possible that the only role for B5R is assisting in the synthesis of enveloped virus. However, I find this answer unsatisfactory for several reasons. First, the possession of an anchor sequence and sugar moieties by B5R indicate that this protein is unlikely to function as an intracellular protein. Since virus assembly is an intracellular process, I would expect proteins whose sole purpose is to aid in assembly to be designed for an intracellular existence, which clearly, B5R is not. It is possible that B5R aids in assembly simply by its presence in the membrane, acting as a structural component as it were. However, the results with RPVAB5R show that apparently perfectly normal virus can be assembled without the B5R protein. In addition, the strong conservation of a large, highly structured extracellular domain showing homology with host regulatory proteins of specific function

PAGE 288

269 strongly suggests that this protein has been acquired by the virus for a purpose other than simply acting as a girder in the virus particle. Since factor H is an inhibitor of the complement system, one possible explanation is that the B5R protein acts to protect the EEV particle from complement-mediated neutralization which would be consistent with the activity of B5R residing in the membrane bound form. However, the absence of B5R protein from the membrane of EEV particles had no effect on complement-mediated neutralization of these particles in cell culture (Martinez-Pomares, L. Personal Communication) ruling out this possibility. In fact, any activity involving the regulation of complement is not likely based on recent work mapping the inhibitory activity of human factor H (82). This work reported that the first three SCR domains of factor H are necessary and sufficient for the inhibitory activity of this protein. Since the homology between B5R and factor H for the most part falls outside of these domains, B5R cannot possess the inhibitory activity of factor H. Since B5R is located in the membrane of the EEV particle, another obvious function for it is that of a receptor. Indeed, comparison of the amounts of recoverable virus immediately (< one hour) following intranasal infection of mice with RPV or RPVAB5R revealed that the disappearance of RPV is greater than that of RPVAB5R by about one order of magnitude. In addition, the growth curves observed in the mouse lung seem to indicate a delay in the lag phase of growth during which attachment and entry of the

PAGE 289

270 virus occur. These observations are consistent with a decreased ability of RPVAB5R to attach to and /or enter cells within the animal. However, preliminary studies of binding kinetics performed on purified EEV particles failed to reveal any difference between RPV and RPVAB5R in cell culture in either permissive or non-permissive cells (data not shown). While these results appear to rule out any B5R-mediated receptor activity, it is quite possible that the cells used were not appropriate and might be lacking the receptor bound by B5R. In addition, it is also possible that any receptor activity possessed by B5R might require some soluble host protein, such as a complement component, that might be lacking in cell culture. It would be interesting to repeat these studies using a wider variety of cells, including, if possible, primary cells obtained from the mouse respiratory tract. It is also possible that the B5R protein is involved in entry of the virus into cells, an activity that would be missed in binding experiments. This is thought to be unlikely, however, since it has been observed that the appearance and processing of viral proteins produced by RPVAB5R in cell culture appear normal in relation to RPV. This suggests that in cell culture at least, the entry of RPVAB5R into cells occurs at a rate equivalent to that of RPV. Again the use of different cells in this assay might reveal a difference. I personally feel that B5R will be shown to have some type of receptor activity based on the conservation of the SCR domains. Since all of the proteins in the SCR family show some type of binding activity, it is probable that the SCR structure

PAGE 290

constitutes a general structure for binding activity while the amino acids within the SCR framework determine the specificity of binding. Future Directions While this work has provided some insight into the role of the RPV B5R protein in virus morphogenesis and virulence, like all research, it has generated many new questions. Of primary importance is a final determination of the effect of B5R on inflammation in the CAM. While I have argued that the increase in inflammation observed in RPVAB5R white pocks occurs as a secondary effect of the mutants defect in enveloped virus assembly, I have no firm proof of this. Since the effects of B5R on inflammation were originally predicted to occur through interactions with complement components, it would be informative to examine the pock morphology of RPVAB5R on CAMs depleted of complement by treatment with cobra venom factor. The presence of inflammatory cells within RPVAB5R pocks in treated CAMs would indicate that complement is not involved in the inflammatory response and would provide extremely strong evidence that B5R does not directly affect inflammation. Along these same lines, it would be interesting to examine the pock morphology of other mutants that are defective in enveloped virus assembly. If failure to grow efficiently is truly the reason for the white pock phenotype of RPVAB5R, then

PAGE 291

272 one would expect these other mutants to also produce white inflammatory pocks on the CAM. A second area demanding further investigation is the involvement of B5R in the pathway of enveloped virus assembly. I have predicted that B5R is involved in the wrapping of virions and that deletion of this ORF affects the rate of this process in RK-13 cells. Pulse / chase experiments comparing the rate at which IMV from RPV and RPVAB5R infected cells is converted to enveloped virus might be used to test this hypothesis. In addition, electron microscopy of RPV and RPVAB5R infected cells with special attention to the presence of the intermediates in morphogenesis may provide some insight into how B5R affects this process. Since I have propose that B5R may play a central role in the formation of enveloped virus, it would be interesting to examine the trafficking of the other envelope proteins to see if they are properly incorporated into virions in the absence of B5R. Along these lines it would also be useful to test if B5R co-immunoprecipitates with any of the other virus envelope proteins or perhaps even host cell proteins since the host range properties of the B5R mutants indicate that these interactions must surely occur. Although I have indicated that the absence of B5R protein apparently had no effect on binding activity of EEV, these studies were limited and for reasons already described, I feel this area deserves more intensive investigation. Again, for the reasons described earlier, it would be interesting

PAGE 292

273 to repeat these studies using other cell types, particularly primary cells from the respiratory tract. In addition, since factor H is thought to interact with other cellular complement components, it would be interesting to use cells expressing these components. Also, my preliminary studies into this area used uninfected cells and since I have theorized that B5R may interact with other viral proteins, it would be interesting to examine the binding of RPVAB5R EEV particles to infected cells. Much of this work has been based on the homology of B5R with other proteins containing SCR domains; however, there are no data available as to which regions of the protein are important. The construction of mutants lacking one or more of the SCR domains would be informative in this area. In addition, it would be interesting to test the effect of perturbing the SCR structure by changing the coding sequence at amino acid sites thought to be critical for the SCR structure. Finally, while this work has identified a region of the protein involved in cleavage, it would be interesting to localize the specific cleavage site. This might be achieved by cloning the region identified as important in cleavage into another membrane protein that is normally not cleaved and looking for cleavage of this hybrid protein. Subcloning of progressively smaller regions into the recipient protein would eventually isolate the minimum sequence necessary for cleavage. A good candidate for a recipient protein is the hemagglutinin (HA) protein of vaccinia. The HA protein is a membrane

PAGE 293

274 protein that is found on the infected cell surface. This protein is not normally cleaved from the cell and in addition, its hydrophobicity plot is very similar to that of B5R especially near the anchor sequence meaning that there is a good chance that the environment surrounding the transplanted B5R fragment may not be so dissimilar as to prevent cleavage. In conclusion, this work has demonstrated that the B5R protein of RPV is intimately involved in the pathway of enveloped virus production and is necessary for normal virulence of the virus. The evidence presented argues that the effects of B5R on virulence are through its involvement in virion morphogenesis and not through any interaction with the host immune system. Finally, it has been demonstrated that in the systems tested here, the membrane-bound form of B5R is sufficient for expression of B5R activity.

PAGE 294

REFERENCES 1. Abbas, A. K v A. H. Lichtman, and J. S. Pober. 1991. Cytokines, p. 225-243. In Cellular and Molecular Immunology. W.B. Saunders Company, Philadelphia*. 2. Abbas, A. K., A. H. Lichtman, and J. S. Pober. 1991. Effector Cells of Cell-Mediated Immunity, p. 244-258. In Cellular and Molecular Immunology. W.B. Saunders Company, Philadelphia. 3. Ahn, B., P. D. Gershon, and B. Moss. 1996. RNA polymerase-associated protein rap94 confers promoter specificity for initiating transcription of vaccinia virus early stage genes. J. Biol. Chem. 269:7552-7557. 4. Ahn, B. Y., P. D. Gershon, E. V. Jones, and B. Moss. 1990. Identification of rpo30, a vaccinia virus RNA polymerase gene with structural similarity to a eucaryotic transcription elongation factor. Mol. Cell. Biol. 10:5433-5441. 5. Alcami, A. and G. L. Smith. 1995. Vaccinia, cowpox and camelpox viruses encode soluable gamma interferon receptors with novel broad species specificity. J. Virol. 69:4633-4639. 6. Alcami, A. and G. L. Smith. 1995. Cytokine Receptors encoded by poxviruses: a lesson in cytokine biology. Immunology Today 16:474-478. 7. Alcami, A. and G. L. Smith. 1996. A soluble receptor for interleukin-lb encoded by vaccinia virus: A novel mechanism of virus modulation of the host response to infection. Cell 71:153-167. 8. Allen, A. M., G. L. Clarke, J. R. Ganaway, A. Lock, and R. M. Werner. 1981. Pathology and diagnosis of mousepox. Lab. Anim. Sci. 31:599-608. 9. Appleyard, G., A. J. Hapel, and E. A. Boulter. 1971. An antigenic difference between intracellular and extracellular rabbitpox virus. J. Gen. Virol. 13:9-17. 275

PAGE 295

276 10. Armstrong, J. A., D. H. Metz, and M. R. Young. 1973. The mode of entry of vaccinia virus into L cells. J. Gen. Virol. 21:533-537. 11. Avila, F. R., R. M. Schultz, and W. A. F. Tompkins. 1972. Specific macrophage immunity to vaccinia virus: macrophage-virus interactions. Infect. Immun. 6:9-16. 12. Baldick, C J. and B. Moss. 1993. Characterization and temporal regulation of mRNAs encoded by vaccinia virus intermediate-stage genes. J. Virol. 67:3515-3527. 13. Barban, S. and S. Baron. 1968. Differential inhibitory effects of interferon on deoxythymidine kinase induction of vaccinia-infected cell cultures. Proc. Soc. Exp. Biol, and Med. 127:160-164. 14. Baroudy, B. N., S. Venkatesan, and B. Moss. 1982. Incompletely base-paired flip-flop terminal loops link the two DNA strands of the vaccinia virus genome into one uninterupted polynucleotide chain. Cell 28:315. 15. Beard, J. W. and P. Rous. 1938. The fate of vaccinia virus on cultivation In vitro with Kupffer cells (reticulo-endothelial cells). J. Exp. Med. 67:883-911. 16. Bedson, H. S. and Marjorie J. Duckworth. 1963. Rabbit Pox: An experimental study of the pathways of infection in rabbits. J. Path. Bact. 85:1-19. 17. Bedson, H. S. and K. R. Dumbell. 1961. The effect of temperature on the growth of pox viruses in the chicken embryo. J. Hyg. 62:141. 18. Berns, K. I. and C. Silverman. 1970. Natural occurence of cross-linked vaccinia virus deoxyribonucleic acid. J. Virol. 5:299. 19. Blanden, R. V. 1971. Mechanism of recovery from a generalized viral infection: Mousepox. III. regression of infectious foci. J. Exp. Med. 133:1090-1104. 20. Blasco, R. and B. Moss. 1991. Extracellular vaccinia virus formation and cell-to-cell virus transmission are prevented by deletion of the gene encoding the 37,000-dalton outer envelope protein. J. Virol. 65:5910-5920. 21. Blasco, R. and B. Moss. 1992. Role of cell-associated enveloped vaccinia virus in cell-to-cell spread. J. Virol. 66:4170-4179.

PAGE 296

277 22. Blasco, R., J. R. Sisler, and B. Moss. 1993. Dissociation of progeny vaccinia virus from the cell membrane Is regulated by a viral envelope glycoprotein: effect of a point mutation in the lectin homology domain of the A34R gene. J. Virol. 67:3319-3325. 23. Bloom, D. C, K. M. Edwards, C Hager, and R. W. Moyer. 1991. Identification and characterization of two non-essential regions of the rabbitpox virus genome involved in virulence. J. Virol. 65:1530-1542. 24. Bloom, D. G, R. J. Stern, M. Duke, D. Smith, and R. W. Moyer. 1993. A revised Hind III map and sequence analysis of a large left-hand' non-essential region of the rabbit poxvirus genome. Virus Res. 28:125-140. 25. Boulter, E. A. 1969. Protection against poxviruses. Proceedings of the Royal Society of Medicine 62:295. 26. Boulter, E. A., H. B. Maber, and E. T. W. Bowen. 1961. Studies on the physiological disturbances occuring in experimental rabbit pox: an approach to rational therapy. Br. J. Exp. Path. 42:433-444. 27. Boulter, E. A., H. T. Zwartouw, D. H. J. Titmuss, and H. B. Maber. 1971. The nature of the immune state produced by inactivated vaccinia virus in rabbits. Amer. J. Epidem. 94:612-620. 28. Brier, A. M., C. Wohlenberg, J. Rosental, M. Mage, and A. L. Notkins. 1971. Inhibition or enhancement of immunological injury of virus-infected cells. Proc. Natl. Acad. Sci. USA 68:3073-3077. 29. Briody, B. A., N. Ledinko, and C. Stannard. 1951. Studies on vaccinia virus. II. Neutralization of vaccinia virus by normal guinea pig serum. J. Immunol. 67:413-422. 30. Broeze, R. J., J. P. Daugherty, J. Pichon, B. M. Jayaram, and P. Lengyel. 1981. Studies with pure mouse Ehrlich ascites tumor interferons alpha and beta: patterns of induction of 2-5A synthetase and a double stranded RNA dependent protein kinase in mouse cells and human cells. Journal of Interferon Research. 1:191-202. 31. Budelier, K. and J. Schorr. 1995. Preparation and anlysis of DNA, p. 2.9.1-2.9.14. In F. M. Ausubel, R. Brent, R. Kingston, D. Moore, J. G. Seidman, J. Smith, and K. Struhl (eds.), Current Protocols in Molecular Biology. John Wiley and Sons, Inc., Boston.

PAGE 297

278 32. Bukowski, J. F., B. A. Woda, S. Habu, K. Okumura, and R. M. Welsh. 1983. Natural killer cell depletion enhances virus synthesis and virus-induced hepatitis in vivo.. J. Immunol. 131:1531-1537. 33. Buller, R. M. L, S. Chakrabarti, B. Moss, and T. Fredrickson. 1988. Cell proliferative response is mediated by VGF. Virology 164:182-192. 34. Buller, R. M. L, G. L. Smith, K. Cremer, A. L. Notkins, and B. Moss. 1985. Decreased virulence of recombinant vaccinia virus expression vectors is associated with a thymidine kinase-negative phenotype. Nature 317:813-815. 35. Buller, R. M. L., S. Chakrabarti, J. A. Cooper, D. R. Twardzik, and B. Moss. 1988. Deletion of the vaccinia virus growth factor gene reduces virus virulence. J. Virol. 62:866-874. 36. Buller, R. M. L. and G. J. Palumbo. 1991. Poxvirus Pathogenesis. Microbiol. Rev. 55:80-122. 37. Chang, A. and D. H. Metz. 1976. Further investigation on the mode of entry of vaccinia virus into cells. J. Gen. Virol. 32:275-282. 38. Chang, W., J. Hsiao, C. Chung, and C. Bair. 1995. Isolation of a monoclonal antibody which blocks vaccinia virus infection. J. Virol. 69:517-522. 39. Cooper, H. and Y. Paterson. 1995. Immunology, p. 11.12.1-11.12.6. In F. M. Ausubel, R. Brent, R. Kingston, D. Moore, J. G. Seidman, J. Smith, and K. Struhl (eds.), Current Protocols in Molecular Biology. John Wiley and Sons, Inc., Boston. 40. Dales, S. 1965. Penetration of animal viruses into cells. Progress in Medical Virology 7:1-43. 41. Dales, S. and R. Kajioka. 1964. The cycle of of multiplication of vaccinia virus in Earles strain L cells: I Uptake and penetration. Virology 24:278-294. 42. Dales, S. and E. H. Mosbach. 1968. Vaccinia as a model for membrane biogenesis. Virology 35:564-583. 43. Davidson, J. M. 1992. Wound Repair, p. 809. In J. I. Gallin, I. M. Goldstein, and R. Snyderman (eds.), Inflammation: Basic Principles and Clinical Correlates. Raven Press, Ltd., New York City. 44. Dawson, I. M. and A. S. Mcfarlane. 1948. Structure of an animal virus. Nature 161:464-466.

PAGE 298

279 45. Delange, A. M. 1989. Identification of temperature sensivtive mutants that are defective in conversion of concatameric replicative intermediates to the mature linear DNA genome. J. Virol. 63:2437-2444. 46. Delange, A. M. and G. McFadden. 1986. Sequence non-specific replication of transfected plasmid in poxvirus-infected cells. Proc. Natl. Acad. Sci. USA 83:614. 47. Delange, A. M. and G. McFadden. 1987. Efficient resolution of replicated poxvirus telomeres to native hairpin structures requires two inverted symmetrical copies of a core target DNA sequence. J. Virol. 61:1957-1963. 48. Delange, A. M., M. Reddy, D. Scraba, and G. McFadden. 1986. Replication and resolution of cloned poxvirus telomeres in vivo generates linear minichromosomes with intact viral hairpin termini. J. Virol. 59:249-259. 49. Dienes, L. and H. L. Naterman. 1937. The immunological response to vaccinia in guinea pigs. J. Infect. Dis. 61:279-290. 50. Dinarello, C. A. 1992. Role of Interleukin-1 and Tumor Necrosis Factor in Systemic Responses to Infection and Inflammation, p. 211. In J. I. Gallin, I. M. Goldstein, and R. Snyderman (eds.), Inflammation: Basic Principles and Clinical Correlates. Raven Press, Ltd., New York. 51. Doms, R. W., R. Blumenthal, and B. Moss. 1990. Fusion of intraand extracellular forms of vaccinia virus with the cell membrane. J. Virol. 64:4884-4892. 52. Douglas, S. R. and W. Smith. 1930. A study of vaccinal immunity in rabbits by means of in vitro methods. Br. J. Exp. Path. 11:96-111. 53. Downie, A. W and K. R. Dumbell. 1947. The isolation and cultivation of variola virus on the chorio-allantois of chick embryos. J. Pathol. Bacteriol. 59:189-198. 54. Dubochet, J., M. Adrian, K. Richter, J. Garces, and R. Wittek. 1996. Structure of intracellular mature vaccinia virus observed by cryoelectron microscopy. J. Virol. 68:1935-1941. 55. Duncan, S. A. and G. L. Smith. 1992. Identification and characterization of an extracellular envelope glycoprotein affecting vaccinia virus egress. J. Virol. 66:1610-1621.

PAGE 299

280 56. Easterbrook, K. B. 1966. Controlled degradation of vaccinia virus in vitro : an electron microscopic study. J. Ultrastruct. Res. 14:484-496. 57. Engelstad, M., S. T. Howard, and G. L. Smith. 1992. A constitutively expressed vaccinia gene encodes a 42-kDa glycoprotein related to complement control factors that forms part of the extracellular virus envelope. Virology 188:801-810. 58. Engelstad, M. and G. L. Smith. 1993. The vaccinia virus 42-kDa envelope protein is required for the envelopment and egress of extracellular virus and for virus virulence. Virology 194:627-637. 59. Eppstein, D. A., Y. V. Marsh, A. B. Schreiber, S. R. Newman, G. J. Todaro, and J. J. Nestor Jr.. 1985. Epidermal growth factor receptor occupancy inhibits vaccinia virus infection. Nature 318:663-665. 60. Esposito, J. J., J. F. Obijeski, and J. H. Nakano. 1978. Orthopoxvirus DNA: strain differeniation by electrophoresis of restriction endonuclease fragmented virion DNA. Virology 89:53-66. 61. Essani, K. and S. Dales. 1979. Biogenesis of vaccinia: evidence for more then 100 polypeptides in the virion. Virology 95:385-394. 62. Essani, K., R. Dugre, and S. Dales. 1982. Biogenesis of vaccinia: involvement of spicules of the envelope during virion assembly examined by means of conditional lethal mutants and serology. Virology 118:279-292. 63. Fairbrother, R. W. 1933. The production of immunity to vaccinia virus by mixtures of immune serum and virus and the importance of phagocytosis in antivaccinial immunity. J. Pathol. Bacteriol. 36:55-64. 64. Feinberg, A. P. and B. Vogelstein. 1983. A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-13. 65. Fenner, F., R. Wittek, and K. R. Dumbell. 1989. Structure and Chemical Composition of the Virion, p. 29-60. In The Orthopoxviruses. Academic Press, San Diego. 66. Fenner, F., R. Wittek, and K. R. Dumbell. 1989. Replication and Morphogenesis of Vaccinia Virus, p. 61-84. In The Orthopoxviruses. Academic Press Inc., San Diego.

PAGE 300

281 67. Fenner, F., R. Wittek, and K. R. Dumbell. 1989. Historical Introduction and Overview, p. 1-28. In The Orthopoxviruses. Academic Press, San Diego. 68. Fenner, F., R. Wittek, and K. R. Dumbell. 1989. The Pathogenesis, Pathology and Immunology of Orthopoxvirus Infections, p. 85-141. In The Orthopoxviruses. Academic Press Inc., San Diego. 69. Florman, A. L. and J. F. Enders. 1942. The effect of homologous antiserum and complement on the multiplication of vaccinia virus in roller-tube cultures of blood-mononuclear cells. J. Immunol. 43:159-174. 70. Fredrickson, T. N., J. M. Sechler, G. J. Palumbo, J. Albert, L. H. Khairallah, and R. M. L. Buller. 1992. Acute inflammatory response to cowpox virus infection of the chorioallantoic membrane of the chick embryo. Virology 187:693-704. 71. Friedman, R. M., S. Baron, C. E. Buckler, and R. I. Steinmuller. 1962. The role of antibody, delayed hypersensitivity, and interferon production in recovery of guinea pigs from primary infection with vaccinia virus. J. Exp. Med. 116:347-356. 72. Friedman, R. M., D. H. Metz, R. M. Esteban, D. R. Tovell, L. A. Ball, and I. M. Kerr. 1972. Mechanism of interferon action: Inhibition of viral messenger ribonucleic acid translation in L-cell extracts. J. Virol. 10:1184-1198. 73. Fujiki, Y., A. L. Hubbard, S. Fowler, and P. B. Lazarow. 1982. Isolation of intracellular membranes by means of sodium carbonate treatment: application to endoplasmic reticulum. J. Cell Biol. 93:97-102. 74. Garon, C. F., E. Barbosa, and B. Moss. 1978. Visualization of an inverted terminal repetition in vaccinia virus DNA. Proc. Natl. Acad. Sci. USA 75:4863. 75. Gershon, P. D., B. Y. Ahn, M. Garfield, and B. Moss. 1991. Poly (A) polymerase and a dissociable polyadenylation stimulatory factor encoded by vaccinia virus. Cell 66:1269-1278. 76. Ginder, D. R. 1955. Resistance to Fibroma Virus infections, the role of immune leukocytes and immune macrophages. J. Exp. Med. 101:43-58. 77. Glasgow, L. A. and K. Habel. 1962. The role of interferon in vaccinia virus infection of mouse embryo tissue culture. J. Exp. Med. 115:503-512. 78. Glasgow, L. A. and K. Habel. 1963. Interferon production by mouse leukocytes in vitro and in vivo. J. Exp. Med. 117:149-161.

PAGE 301

282 79. Goebel, S. J., G. P. Johnson, M. E. Perkus, S. W. Davis, J. P. Winslow, and E. Paoletti. 1990. The complete DNA sequence of vaccinia virus. Virology 179:247-266. 80. Goldstein, I. M. 1992. Complement: Biologically Active Products, p. 63-77. In J. I. Gallin, I. M. Goldstein, and R. Snyderman (eds.), Inflammation: Basic Principles and Clinical Correlates. Raven Press, Ltd., New York City. 81. Gong, S., G Lai, and M. Esteban. 1990. Vaccinia virus induces cell fusion at acid pH and this activity is mediated by the N-terminus of the 14 kda virus envelope protein. Virology 178:81-91. 82. Gordon, D. L., R. M. Kaufman, T. K. Blackmore, J. Kwong, and D. M. Lublin. 1995. Identification of complement regulatory domains in human factor H. J. Immunol. 155:348-356. 83. Gordon, J., A. Mohandas, S. Wilton, and S. Dales. 1991. A prominent surface polypeptide involved in the biogenesis and function of the vaccinia virus envelope. Virology 181:671-686. 84. Greene, H. S. N. 1933. A pandemic of rabbit-pox. Proc. Soc. Exp. Biol, and Med. 30:892-894. 85. Greene, H. S. N. 1934. Rabbit pox II. Pathology of the epidemic disease. J. Exp. Med. 60:441. 86. Greene, H. S. N. 1934. Rabbit pox I. Clinical manifestations and course of disease. J. Exp. Med. 60:427. 87. Greene, H. S. N. 1935. Rabbit Pox III: report of an epidemic with especial reference to epidemiological factors. J. Exp. Med. 61:807-831. 88. Greenstein, D. and C. Besmond. 1995. E. coli., plasmids and bacteriophages, p. 1.8.1-1.8.3. In F. M. Ausubel, R. Brent, R. Kingston, D. Moore, J. G. Seidman, J. Smith, and K. Struhl (eds.), Current Protocols in Molecular Biology. John Wiley and Sons, Inc., Boston. 89. Hiller, G., H. Eibl, and K. Weber. 1981. Characterization of intracellular and extracellular vaccinia virus variants: Ni-isonicotinoyl-N2 3-methyl-4-chlorobenzoylhydrazine interferes with cytoplasmic virus dissemination and release. I. Virol. 39:903-913.

PAGE 302

283 90. Hiller, G. and K. Weber. 1985. Golgi-derived membranes that contain an acylated polypeptide are used for vaccinia virus envelopment. J. Virol. 55:651-659. 91. Hirsch, R. L. 1982. The complement system: its importance in the host response to viral infection. Microbiol. Rev. 46:71-85. 92. Holowczak, J. A. 1972. Uncoating of poxviruses I: detection and characterization of subviral particles in the uncoating process. Virology 50:216-232. 93. Hooda, D., C L. Thompson, and R. C. Condit. 1989. Detailed phenotypic characterization of five temperature-sensitive mutants in the 22and 147-kilodalton subunits of vaccina virus DNA-dependent RNA polymerase. J. Virol. 63:714-729. 94. Hourcade, D., V. M. Holers, and J. P. Atkinson. 1989. The regulators of complement activation (RCA) gene cluster. Advances in Immunology 45:381-416. 95. Hruby, D. E., D. L. Lynn, and J. R. Kates. 1979. Vaccinia virus replication requires active participation of the host cell transcriptional apparatus. Proc. Natl. Acad. Sci. USA 76:1887-1890. 96. Hruby, H, L. A. Guarino, and J. R. Kates. 1979. Vaccinia virus replication. I. Requirement for the host cell nucleus. J. Virol. 29:705-715. 97. Hu, F., G A. Smith, and D. J. Pickup. 1994. Cowpoxvirus contains two copies of an early gene encoding a soluable secreted form of the type II TNF receptor. Virology 204:343-356. 98. Hugin, A. W. and C. Hauser. 1994. The epidermal growth factor is not a receptor for vaccinia virus. J. Virol. 68:8409-8412. 99. Isaacs, S. N., G. J. Kotwal, and B. Moss. 1992. Vaccinia Virus complement-control protein prevents antibody-dependent complement-enhanced neutralization of infectivity and contribute to virulence. Proc. Natl. Acad. Sci. USA 89:628-632. 100. Issacs, S. N., E. J. Wolffe, L. G. Payne, and B. Moss. 1992. Characterization of a vaccinia virus-encoded 42-kilodalton class I membrane glycoprotein component of the extracellular virus envelope. J. Virol. 66:7217-7224.

PAGE 303

284 101. Jacoby, R. O., P. N. Bhatt, and D. G. Brownstein. 1989. Evidence that NK cells and interferon are required for genetic resistance to lethal infection with ectromelia virus. Archives of Virology 108:49-58. 102. Johnson, G. P., S. J. Goebel, and E. Paoletti. 1993. An update on the vaccinia virus genome. Virology 196:381-401. 103. Joklik, W. K. 1964. The intracellular uncoating of poxvirus DNA I: The fate of radioactively-labeled rabbitpox virus. J. Mol. Biol. 8:263-276. 104. Joklik, W. K. 1964. The intracellular uncoating of poxvirus DNA II: The molecular basis of the uncoating process. J. Mol. Biol. 8:277-288. 105. Kane, E. M. and S. Shuman. 1992. Temperature-sensitive mutations in the vaccinia virus H4 gene encoding a component of the virion RNA polymerase. J. Virol. 66:5752-5762. 106. Kapil, S., S. K. Das, and A. Kumar. 1989. Total haemolytic complement profile in chicks following fowl pox vacciniation of infection. Acta Virol. 33:87-90. 107. Karupiah, G., R. V. Blanden, and I. A. Ramshaw. 1990. Interferon gamma is involved in the recovery of athymic nude mice from recombinant vaccinia virus /interleukin 2 infection. J. Exp. Med. 172:1495-1503. 108. Keck, J. G., G J. Baldick, and B. Moss. 1990. Role of DNA replication in vaccinia virus gene expression: A naked template is required for transcription of three late trans-activator genes. Cell 61:801-809. 109. Kettle, S., N. W. Blake, K. M. Law, and G. L. Smith. 1995. Vaccinia virus serpins B13R (SPI-2) and B22R (SPI-1) ancode M(r) 38.5 and 40K, intracellular polypeptides that do not effect virus virulence in a murine intranasal model. Virology 206:136-147. 110. Kotwal, G. J., S. N. Isaacs, R. McKenzie, M. M. Frank, and B. Moss. 1990. Inhibition of the complement cascade by the major secretory protein of vaccinia virus. Science 250:827-830. 111. Kotwal, G. J. and B. Moss. 1988. Vaccinia Virus encodes a secretory polypeptide structurally related to complement control proteins. Nature 335:176-178.

PAGE 304

285 112. Kovacs, G. R., R. Rosales, J. G. Keck, and B. Moss. 1994. Modification of the cascade model for regulation of vaccinia virus gene expression: purification of a prereplicative, late-stage-specific transcription factor. J. Virol. 68:3443-3447. 113. Kristensson, K., B. Lundh, E. Norrby, L. Payne, and G Orvell. 1984. Asymmetric budding of viruses in ependymal and choroid plexus epithelial cells. Neuropathol. Appl. Neurobiol. 10:209. 114. Lai, C, S. Gong, and M. Esteban. 1991. The 32-kilodalton envelope protein of vaccinia virus synthesized in Escherichia coli binds with specificity to cell surfaces. J. Virol. 65:499-504. 115. Lampson, G. P., A. A. Tytell, A. K. Field, M. M. Nemes, and M. R. Hilleman. 1967. Inducers of interferon and host resistance: I. Double-stranded RNA from extracts of Penicilium funiculosum.. Proc. Natl. Acad. Sci. USA 58:782-786. 116. Lengyel, P. 1982. Biochemistry of interferons and their actions. Ann. Rev. Biochem. 51:251-282. 117. Mackett, M. and L. C Archard. 1979. Conservation and variation in Orthopoxvirus genome structure. J. Gen. Virol. 45:683-701. 118. Martin, S. A., E. Paoletti, and B. Moss. 1975. Purification of mRNA guanylyl-transferase and mRNA (guanine-7-) methyltransferase from vaccinia virons. J. Biol. Chem. 250:9322-9329. 119. Martinez-Pomares, L., R. J. Stern, and R. W. Moyer. 1993. The ps/hr Gene (B5R Open Reading Frame Homolog) of rabbitpox virus controls pock color, is a component of extracellular enveloped virus, and Is secreted into the medium. J. Virol. 67:5450-5462. 120. Massung, R. F., L. Liu, J. Qi, J. C Knight, T. E. Yuran, A. R. Kerlavage, J. M. Parsons, J. C. Venter, and J. J. Esposito. 1994. Analysis of the complete genome of smallpox variola major strain Bangladesh-1975. Virology 201:215-240. 121. McFadden, G., K. Graham, K. Ellison, M. Barry, J. Macen, M. Schreiber, K. Mossman, P. Nash, A. Lalani, and H. Everett. 1995. Interruption of cytokine networks by poxviruses: lessons from myxoma virus. J. Leuk. Biol. 57:731-738. 122. McNeill, W. H. 1976. Plagues and Peoples. Anchor Press /Doubleday, Garden City, N.Y..

PAGE 305

286 123. Metz, D. H. and M. Esteban. 1972. Interferon inhibits protein synthesis in L cells infected with vaccinia virus. Nature 238:385-388. 124. Mims, C. A. 1959. The response of mice to large intravenous injections of ectromelia virus I. The fate of injected virus. Br. J. Exp. Path. 40:533-542. 125. Moos, M. 1995. Analysis of proteins, p. 10.0.1-10.19.12. In F. M. Ausubel, R. Brent, R. Kingston, D. Moore, J. G. Seidman, J. Smith, and K. Struhl (eds.), Current Protocols in Molecular Biology. John Wiley and Sons, Inc., Boston. 126. Morgan, B. P. 1990. The biological effects of complement activation, p. 36-56. In Complement: Clinical aspects and relevance to disease. Academic Press, San Diego. 127. Morishima, T. and K. Hayashi. 1978. Meningeal exudate cells in vaccinia meningitis of mice: role of local T cells. Infect. Immun. 20:752-759. 128. Moss, B. 1985. Replication of Poxviruses, p. 685-703. In Bernard N. Fields, David M. Knipe, Robert M. Chanock, Joseph L. Melnick, Bernard Roizman, and Robert E. Shope (eds.), Virology. Raven Press, New York City. 129. Moss, B., B. Ahn, B. Amegadzie, P. D. Gershon, and J. G. Keck. 1991. Cytoplasmic transcription system encoded by vaccinia virus. J. Biol. Chem. 266:1355-1358. 130. Mossman, K., C. Upton, R. M. L. Buller, and G. McFadden. 1995. Species specificity of ectromelia virus and vaccinia virus interferon-gamma binding proteins. Virology 208:762-769. 131. Mossman, K., C. Upton, and G. McFadden. 1995. The myxoma virus soluable interferon-gamma receptor homolog, M-T7, inhibits interferon-gamma in a species specific manner. J. Biol. Chem. 270:3031-3038. 132. Moyer, R. W. and R. L. Graves. 1981. The mechanism of cytoplasmic orthopoxvirus DNA replication. Cell 27:391-401. 133. Miiller, E. 1992. Complement: Chemistry and Pathways, p. 33-61. In J. I. Gallin, I. M. Goldstein, and R. Snyderman (eds.), Inflammation: Basic Principles and Clinical Correlates. Raven Press, Ltd., New York City. 134. Niles, E. G., R. Condit, P. Caro, K. Davidson, L. Matusick, and J. Seto. 1986. Nucleotide sequence and genetic map of the 16-kb vaccinia virus Hidlll D fragment. Virology 153:96-112.

PAGE 306

287 135. Niles, E. G., G. Lee-Chen, S. Shuman, and B. Moss. 1989. Vaccinia virus gene D12L encodes the small subunit of the viral mRNA capping enzyme. Virology 172:513-522. 136. Niles, E. G. and J. Seto. 1988. Vaccinia virus gene D8 encodes a virion transmembrane protein. J. Virol. 62:3772-3778. 137. Ohta, H., C. Kai, V. Yoshikawa, and K. Yamanouchi. 1983. Activation of chicken alternative pathway by fowlpox virus-infected cells. Infect. Immun. 42:721-727. 138. Ohta, H., V. Yoshikawa, C Kai, K. Yamanouchi, H. Taniguchi, K. Komine, V. Ishijima, and H. Okada. 1986. Effect of complement depletion by cobra venom factor on fowlpox virus infection in chicken and chicken embryos. J. Virol. 57:670-673. 139. Oie, M. and Y. Ichihashi. 1981. Characterization of vaccinia polypeptides. Virology 113:263-276. 140. Palumbo, G. J., R. M. Buller, and W. C Glasgow. 1994. Multigenic evasion of inflammation by poxviruses. J. Virol. 68:1737-1749. 141. Palumbo, G. J., W. C Glasgow, and R. M. Buller. 1993. Poxvirus-induced alteration of arachidonate metabolism. Proc. Natl. Acad. Sci. USA 90:2020-2024. 142. Palumbo, G. J., D. J. Pickup, T. N. Fredrickson, L. J. Mclntyre, and R. M. L. Buller. 1989. Inhibition of an inflammatory response is mediated by a 38-kDa protein of cowpoxvirus. Virology 172:262-273. 143. Parkinson, J. E. and G. L. Smith. 1994. Vaccinia virus gene A36R encodes a Mr 43-50 K protein on the surface of extracellular enveloped virus. Virology 204:376-390. 144. Payne, L. G. 1978. Polypeptide composition of extracellular enveloped vaccinia virus. J. Virol. 27:28-37. 145. Payne, L. G. 1979. Identification of the vaccinia hemaglutinin polypeptide from a cell system yielding large amounts of extracellular enveloped virus. J. Virol. 31:147-155. 146. Payne, L. G. 1980. Significance of extracellular enveloped virus in the i n vitro and in vivo dissemination of vaccinia. J. Gen. Virol. 50:89-100.

PAGE 307

288 147. Payne, L. G. 1992. Characterization of vaccinia virus glycoproteins by monoclonal antibody precipitation. Virology 187:251-260. 148. Payne, L. G. and K. Kristenson. 1979. Mechanism of vaccinia virus release and its specific inhibition by N}-isonicotinoyl-N2 3-methyl-4-chlorobenzoylhydrazine. J. Virol. 32:614-622. 149. Payne, L. G. and K. Kristensson. 1982. The effect of cytochalasin D and monesin on enveloped vaccinia virus release. Archives of Virology 74:11-20. 150. Payne, L. G. and K. Kristensson. 1982. Effect of glycosylation inhibitors on the release of enveloped vaccinia virus. J. Virol. 41:367-375. 151. Payne, L. G. and K. Kristensson. 1985. Extracellular release of enveloped vaccinia virus from mouse nasal epithelial cells in vivo.. J. Gen. Virol. 66:643-646. 152. Payne, L. G. and E. Norrby. 1978. Adsorption and penetration of enveloped and naked vaccinia virus particles. J. Virol. 27:19-27. 153. Pedley, C. B. and R. J. Cooper. 1987. The assay, purification and properties of vaccinia virus-induced uncoating protein. J. Gen. Virol. 68:1021-1028. 154. Pennington, T. H. and E. A. C. Follett. 1984. Vaccinia virus replication in enucleate cells: partial production and synthesis of viral DNA and proteins. J. Virol. 13:488-493. 155. Pickup, D. J., B. S. Ink, W. Hu, C. A. Ray, and W. K. Joklick. 1986. Hemorrhage in lesions caused by cowpox virus is induced by a viral protein that is related to plasma protein inhibitors of serine poteases. Proc. Natl. Acad. Sci. USA 83:7698-7702. 156. Pogo, B. G. T., E. M. Berkowitz, and S. Dales. 1984. Investigation of vaccinia virus DNA replication employing a conditional lethal mutant defective in DNA. Virology 132:436-444. 157. Pogo, B. G. T. and M. T. O'Shea. 1977. Further characterization of deoxyribonucleases from vaccinia virus. Virology 77:56-66. 158. Ray, G A., R. A. Black, S. R. Kronheim, T. A. Greenstreet, P. R. Sleath, G. S. Salvasan, and D. J. Pickup. 1992. Viral inhibition of inflammation: Cowpox virus encodes an inhibitor of the interleukin-lb converting enzyme. Cell 69:597-604.

PAGE 308

289 159. Reed, L. J. and H. Muench. 1938. A simple method for estimating fifty per cent endpoints. The American Journal of Hygiene 27:493-497. 160. Roberts, J. A. 1964. Enhancement of the virulence of attenuated ectromelia virus maintained in a cold environment. Aust. J. exp. Biol. Sci. 42:657-666. 161. Roberts, J. A. 1964. Growth of virulent and attenuated ectromelia virus in cultured macrophages from normal and ectromelia immune mice. J. Immunol. 92:837-842. 162. Rodriguez, J., D. Rodriguez, and M. Esteban. 1992. Insertional inactivation of the vaccinia virus 32-kilodalton gene is associated with attenuation in mice and reduction of viral gene expression in polarized epithelial cells. J. Virol. 66:183-189. 163. Rodriguez, J. F., R. Janeczko, and M. Esteban. 1987. Isolation and characterization of neutralizing antibodies to vaccinia virus. J. Virol. 61:3550-3554. 164. Rodriguez, J. F., E. Paez, and M. Esteban. 1987. A 14,000 Mr envelope protein of vaccinia virus is involved in cell fusion and forms covalently linked trimers. J. Virol. 61:395-404. 165. Rodriguez, J. F. and G. L. Smith. 1990. IPTG-dependent vaccinia virus: identification of a virus protein enabling virion envelopment by Golgi membrane and egress. Nucleic Acids Research 18:5347-5351. 166. Rohrmann, G., L. Yuen, and B. Moss. 1986. Transcription of vaccinia virus early genes by enzymes isolated from vaccinia virions terminates downstream of a regulatory sequence. Cell 46:1029-1035. 167. Rosa, F. and M. Fellous. 1984. The effect of gamma-interferon on MHC antigens. Immunology Today 5:262. 168. Rosales, R., N. Harris, B. Ahn, and B. Moss. 1994. Purification and identification of a vaccinia virus-encoded intermediate stage promoter-specific transcriptio factor that has homology to eukaryotic transcription factor SII (TFIIS) and an additional role as a viral RNA polymerase subunit. J. Biol. Chem. 269:14260-14267. 169. Rosales, R., G. Sutter, and B. Moss. 1994. A cellular factor is required for transcription of vaccinia viral intermediate-stage genes. Proc. Natl. Acad. Sci. USA 91:3794-3798.

PAGE 309

290 170. Rot, A. 1992. The role of leukocyte chemotaxis in inflammation, p. 271. In Biochemistry of Inflammation. Kluwer Academic Publishers, Dordrecht. 171. Rot, A., L. E. Henderson, T. E. Copeland, and E. J. Leonard. 1987. A series of six ligands for the human formyl peptide receptor: tetrapeptides with high chemotactic potency and efficacy. Proc. Natl. Acad. Sci. USA 84:7967-7971. 172. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Gel electrophoresis of DNA, p. 6.2-6.62. In Chris Nolan (ed.), Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. 173. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Plasmid vectors, p. 1.21-1.22. In Chris Nolan (ed.), Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. 174. Samuel, C. E. 1991. Antiviral action of interferons: interferon-regulated cellular proteins and their surprisingly selective antiviral properties. Virology 183:1-11. 175. Samuel, C. E., R. Duncan, G. S. Knutson, and J. W. B. Hershey. 1984. Mechanism of interferon action:Increased phosphorylation of proteini synthesis initiation factor eIF2-alpha i ninterferon, reovirus-infected mouse L929 fibroblasts in vitro and in vivo.. J. Biol. Chem. 259:13451-13457. 176. Schmelz, M., B. Sodeik, M. Ericsson, E. J. Wolffe, H. Shi da, G. Hiller, and G. Griffiths. 1994. Assembly of vaccinia virus: the second wrapping cisterna Is derived from the trans golgi network. J. Virol. 68:130-147. 177. Schmutz, C., L. G. Payne, J. Gubser, and R. Wittek. 1991. A mutation in the gene encoding the vaccinia virus 37,000-M r protein confers resistance to an inhibitor of virus envelopment and release. J. Virol. 65:3435-3442. 178. Schreiber, M. and G. McFadden. 1994. The myxoma virus TNF-receptor homologue (T2) inhibits tumor necrosis factor-alpha in a species specific fashion. Virology 204:692-705. 179. Showell, H. J., R. J. Freer, S. H. Zigmond, E. Schiffman, S. Aswanikumar, B. Corcoran, and E. L. Becker. 1976. The structure-activity relations of synthetic peptides as chemotactic factors and inducers of lysosomal enzyme secretion for neutrophils. J. Exp. Med. 143:1154-1169. 180. Shuman, S. and B. Moss. 1989. Bromouridine triphosphate inhibits transcription termination and mRNA release by vaccinia virions. J. Biol. Chem. 264:21356-21360.

PAGE 310

291 181. Shuman, S., M. Surks, H. Furneaux, and J. Hurwitz. 1980. Purification and characterization of a GTP-pyrophosphate exchange activity from vaccinia virions. J. Biol. Chem. 255:11588-11598. 182. Smith, C. A., D. Terri, J. M. Wignall, W. S. Dins, T. Farrah, C Upton, G. McFadden, and R. G. Goodwin. 1991. T2 open reading frame from the Shope fibroma Virus encodes a soluable from of the TNF receptor. Biochem. Biophys. Res. Commun. 176:335-342. 183. Smith, G. L. and Y. S. Chan. 1991. Two vaccinia proteins structurally related to the intereukin-1 receptor and the immunoglobulin superfamly. J. Gen. Virol. 72:511-518. 184. Snyderman, R. and R. J. Uhing. 1992. Chemoattractant Stimulus-Response Coupling, p. 421-439. In J. I. Gallin, I. M. Goldstein, and R. Snyderman (eds.), Inflammation: Basic Principles and Clinical Correlations. Raven Press, Ltd., New York City. 185. Sodeik, B., R. W. Doms, M. Ericsson, G. Hiller, C. E. Machamer, W. Hof, G. van Meer, Bernard Moss, and Gareth Griffiths. 1993. Assembly of vaccinia virus: Role of the intermediate compartment between the endoplasmic reticulum and the golgi stacks. J. Cell Biol. 121:521-541. 186. Soloski, M. J., E. S. Viteta, L. Flaherty, and J. Uhr. 1981. Characterization of supercoiled nucleoprotein complex released from detergent treated vaccinia virions. J. Virol. 37:770-783. 187. Spriggs, M. K., D. E. Hruby, C. R. Maliszewski, D. J. Pickup, J. E. Sims, R. M. L. Buller, and J. VanSlyke. 1992. Vaccinia and cowpox viruses encode a novel secreted interleukin-l-binding protein. Cell 71:145-152. 188. Stern, W. and S. Dales. 1974. Biogenesis of vaccinia: concerning the origin of the envelope phospholipids. Virology 62:293-306. 189. Stern, W. and S. Dales. 1976. Biogenesis of vaccinia: isolation and characterization of a surface component that elicits antiboy suppressing infectivity and cell-cell fuson. Virology 75:232-241. 190. Stitz, L., A. Althage, H. Hengartner, and R. Zinkernagel. 1985. Natural killer cells vs cytotoxic T cells in the peripheral blood of virus-infected mice. J. Immunol. 134:598-602. 191. Strehler, B. L. 1975. Implications of aging research for society. Federation Proceedings 34:5-8.

PAGE 311

292 192. Stuart, D., K. Ellison, K. Graham, and G. McFadden. 1992. In vitro resolution of poxvirus replicative intermediates into linear minichromosomes with hairpin termini by a virally induced Holliday junction endonuclease. J. Virol. 66:1551-1563. 193. Subrahmanyan, T. P. and C. A. Mims. 1966. Fate of intravenously administered Interferon and the distribution of interferon during virus infection in mice. Br. J. Exp. Path. 47:168-176. 194. Symons, J. A., A. Alcami, and G. L. Smith. 1995. Vaccinia virus encode a soluable type I interferon receptor of novel structure and broad species specificty. Cell 81:551-560. 195. Takahashi-Nishimaki, F., S. Funahashi, K. Miki, S. Hashizume, and M. Sugimoto. 1991. Regulation of plaque size and host range by a vaccinia virus gene related to complement system proteins. Virology 181:158-164. 196. Taylor, G., E. J. Stott, Gail Wetz, and A. Ball. 1991. Comparison of the virulence of the wild-type thymidine kinase (tk)-deficient and tk+ phenotypes of vaccinia virus recombinants after intranasal inoculation of mice. J. Gen. Virol. 72:125-130. 197. Tewari, M. and V. M. Dixit. 1995. Fasand tumor necrosis factor-induced apoptosis is inhibited by the poxvirus crmA gene product. J. Biol. Chem. 270:3255-3260. 198. Thompson, J. P., P. C. Turner, A. N. Ali, B. C. Crenshaw, and R. W. Moyer. 1993. The effects of serpin gene mutations on the distinctive pathobiology of cowpox and rabbitpox virus following intranasal inoculation of BALB/C mice. Virology 197:328-338. 199. Tooze, John, Michael Hollinshead, Burkhardt Reis, Klaus Radsak, and Horst Kern. 1993. Progeny vaccinia and human cytomegalovirus particles utilize early endosomal cisternae for their envelopes. Eur. J. Cell Biol. 60:163-178. 200. Traktman, P. 1990. The enzymology of poxvirus DNA replication. Current Topics in Microbiology and Immunology 163:93-122. 201. Trottier, R. W. and F. Ogolla. 1990. In vivo plaque-forming cell suppression by methyl 20beta-dihydroprednisolonate. Steroids 55:279-282.

PAGE 312

293 202. Tsuru, S., H. Kitani, M. Seno, M. Abe, Y. Zinnaka, and K. Nomoto. 1983. Mechanism of protection during the early phase of a generalized viral infection. I. Contribution of phagocytes to protection against ectromelia virus. J. Gen. Virol. 64:2021-2026. 203. Ulaeto, D., D. Grosenbach, and D. E. Hruby. 1995. Brefeldin A inhibits vaccinia virus envelopment but does not prevent normal processing and localization of the putative envelopement receptor P37. J. Gen. Virol. 76:103-111. 204. Upton, C., J. L. Macen, M. Schreiber, and G. McFadden. 1991. Myxoma virus expresses a secreted protein with homology to the tumor necrosis factor receptor gene family that contributes to viral virulence. Virology 184:370-382. 205. Upton, C., K. Mossman, and G. McFadden. 1992. Encoding of a homolog of the IFN-gamma receptor by myxoma virus. Science 258:1369-1372. 206. Villareal, L. P. and P. Berg. 1977. Hybridization in situ of SV40 plaques: Detection of recombinant virus carrying specific sequences of nonviral DNA. Science 196:183-185. 207. Vos, J. C., M. Saskar, and H. G. Stunnenberg. 1991. Vaccinia virus capping enzyme is a transcription initiation factor. EMBO J. 10:2553-2558. 208. Vos, J. C., M. Saskar, and H. G. Stunnenberg. 1991. Promoter melting by a stage-specific vaccinia virus transcription factor is independent of the presence of RNA polymerase. Cell 65:105-114. 209. Vos, T. C and H. G. Stunnenberg. 1988. Derepression of a novel class of vaccinia virus genes upon DNA replication. EMBO J. 7:3487-3492. 210. Walport, M. 1989. Complement, p. 13.1-13.16. In Ivan Roitt, Jonathan Brostoff, and David Male (eds.), Immunology. Gower Medical Publishing, St. Louis. 211. Weidman, P. J., P. Melancon, M. R. Block, and J. E. Rothman. 1989. Binding of an N-ethylmaleimide-sensitive fusion protein to Golgi membranes requires both a soluable protein(s) and an integral membrane receptor. J. Cell Biol. 108:1589-1596. 212. Welsh, R. M. and L. A. Hallenbeck. 1980. Effect of virus infections on target cell susceptibility to natural killer cell-mediated lysis. J. Immunol. 124:2491.

PAGE 313

294 213. West, B. C, M. L. Eschete, M. E. Cox, and J. W. King. 1987. Neutrophil uptake of vaccinia virus in vitro. J. Infect. Dis. 156:597-606. 214. Westwood, J. C. N, E. A. Boulter, E. T. W. Bowen, and H. B. Maber. 1966. Experimental respiratory infection with poxviruses I: Clinical virological and epidemiological studies. Br. J. Exp. Path. 47:453-465. 215. Westwood, J. C N., W. J. Harris, H. T. Zwartouw, D. H. J. Titmuss, and G. Appleyard. 1964. Studies on the structure of vaccinia virus. J. Gen. Microbiol. 34:67-78. 216. Wirtek, R., A. Menna, H. K. Muller, D. Schumperli, P. G. Boseley, and R. Wyler. 1978. Inverted terminal repeats in rabbit poxvirus and vaccinia virus DNA. J. Virol. 28:171-181. 217. Wittek, R., A. Menna, D. Schumperli, S. Stoffel, H. K. Muller, and R. Wyler. 1977. Hind III and Sst I restriction sites mapped on rabbit poxvirus and vaccinia virus DNA. J. Virol. 23:669-678. 218. Wolffe, E. J., S. N. Isaacs, and B. Moss. 1993. Deletion of the vaccinia virus B5R gene encoding a 42-Kda membrane glycoprotein inhibits extracellular virus envelope formation and dissemination. J. Virol. 67:4732-4741. 219. Yuen, L., A. J. Davison, and Bernard Moss. 1987. Early promoter binding factor from vaccinia virions. Proc. Natl. Acad. Sci. USA 84:6069-6073. 220. Zhen, L. and R. Swank. 1993. A simple and high yield method for recovering DNA from agarose gels. Biotechniques 14:894-898.

PAGE 314

BIOGRAPHICAL SKETCH I was born in New York City, N.Y., in 1963 and after 18 months of intense sensory input, I was bundled off to the quieter, suburban landscape of Dumont, N.J., where the next 18 years of my life passed relatively uneventfully. After receiving a B.S. degree in animal science from the University of Massachusetts at Amherst in 1986, I was employed for two years in the laboratory of Dr. Victor Sapirstein at the Rockland Psychiatric Center in Orangeburg, N.Y. It was during this time that I, in an unparalleled burst of good sense and fortune, married Rita, a kindred spirit and fellow traveler on The Path. In 1988, with little forethought and a good deal of naivete, I entered graduate school, spending one year at Cornell University in N.Y.C. and then transferring to the University of Florida in Gainesville, Fl. What followed can only truthfully be described as a 7-year-long "Season in Hell" the tangible result of which is in your hands. Having survived this ordeal scarred but intact, I will be heading to Colorado State University in Fort Collins, Col., where the mycobacteria is abundant, the mountains are high and Rita and our three dogs patiently wait for me to come out and play. 295

PAGE 315

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Richard W. Moyer,/Ohair Professor of Molecular Genetics and Microbiolot I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Maureen Goodenow Associate Professor of Pathology and Laboratory Medicine I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Paul Gulig Associate Professor of Molecular Genetics and Microbiology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Maurice Swanson Associate Professor of Molecular Genetics and Microbiology

PAGE 316

This dissertation was submitted to the Graduate Faculty of the College of Medicine and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1996 Dean, College of Medicine Dean, Graduate School L