.v+7.(740\

. 23. .1111. 1.,
‘ ..
I
J

2...}. Ly
.4 5 ‘ L-
h

4...» IL}; I! .4
.11 2. (3.3!! .

v9 . 0...!”
Inuit,"
.9:I-v).

)1

. Yn..lciatl.\'
?.!. 3,...»832‘tf
,flwn. 523: '1‘: .i.

a $27I.~ ‘ .

2. ii. ..
V) ..’hh.
.t4néL.v:.Z A , . . 3...}. 3 tan

 

. . frr L ..3:..\.
; .3131 . ‘ .

, ‘v...5...\ (‘¢ ,
v tux. :w‘ ...J. . n\.l.r!v :

3.. : Ls».-.

 

 

MICHIGAN STAT

Mimi 1M 71m ‘IIWIHHI

)
43144

RARIES

I!

 

 

 

 

 

 

 

 

I iii/r»

w!
i.
3193009

 

This is to certify that the

dissertation entitled

Molecular Characterization of Oncogenic Marek's Disease
Virus Attenuation: Genomic Mutations and
their Affect on Viral Growth and Gene Expression.

presented by

Melinda R. Wilson

 

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Animal Science

 

LV— 4

Major professor

 

Date 3/3‘I‘i3

MSU is an Affirmative Action/Equal Opportunity Institution 0 A 12771

 

 

v V
LIBRARY
Michigan State
University

\ J

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
cmmui

 

 

 

 

MOLECULAR CHARACTERIZATION OF ONCOGENIC MAREK'S DISEASE
VIRUS ATTENUATION: GENOMIC MUTATIONS AND THEIR AFFECT ON
VIRAL GROWTH AND GENE EXPRESSION.

By

Melinda R. Wilson

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

DOCTOR OF PHILOSOPHY

Department of Animal Science

1 993

ABSTRACT
MOLECULAR CHARACTERIZATION OF ONCOGENIC MAREK’S DISEASE
VIRUS ATTENUATION: GENOMIC MUTATIONS AND THEIR AFFECT ON
VIRAL GROWTH AND GENE EXPRESSION.
BY

Melinda R. Wilson

Marek’s Disease virus (MDV), an oncogenic avian herpesvirus, produces
malignant lymphomas and nerve demyelination in infected hosts. Serial passage
of oncogenic Marek’s Disease virus in cultured avian cells results in attenuation of
viral oncogenicity. Due to emergence of very virulent (v.v.) MDV in vaccinated
flocks, analysis of viral attenuation is of considerable interest. However, the tightly
cell-associated in vitro growth Characteristics of MDV present unique problems
when attempting to purify, analyze, and manipulate MDV genomes. To facilitate
molecular characterization of MDV, contour-clamped homogeneous electric fields
electrophoresis (CHEF) was used to purify infectious MDV genomes. CHEF
techniques were optimized for evaluation of total genome size and alterations in
structure which occur during attenuation of oncogenic MDV. Using these
techniques, a 200 bp deletion in attenuated genomes of the v.v. MDV. strain Mdll
was identified. The 200 bp deletion, located in the BamHl-L fragment, is separate
and distinct from the previously reported 132 bp amplified region. An attenuated
variant which contains 132 bp amplified regions but has the wild type BamHl-L
fragment, was utilized to determine if deletions in the BamHl-L fragment affect

biological properties of attenuated Md11. Deletions in the BamHl-L region of

attenuated Md11 were associated with increased viral growth and altered
immediate early gene expression in vitro. In vivo studies indicate that the BamHl-L
region does not directly contribute to MDV-induced oncogenicity. Analysis of RNA
transcripts suggests that this region may be involved in latent infection of T-
Iymphocytes by MDV. Another prominent characteristic of attenuated MDV is
reduced expression of a predominant glycoprotein antigen, 90. To investigate the
mechanism by which MDV gC expression is reduced during attenuation of the v.v.
MDV strain Md11, protein levels, gene structure, steady state RNA levels, and
transcription rates of MDV 90 from oncogenic and attenuated isolates of Md11
were determined. The 96 promoter, or MDV regulatory protein(s) that interact
with the MDV gC promoter, are altered during attenuation of MDV strain Md11.
Specific mutations within attenuated MDV genomes may be involved in separate

steps of the pathway leading to MDV-induced oncogenicity.

Copyrighted by
Melinda Ricarda Wilson

1 993

To my husband, Michael F. Daeschlein
for his love, support, and patience.
and
to my daughter, Amber M. Wilson-Daeschlein,

for giving me the motivation to finish.

ACKNOWLEDGMENTS

A special thanks to my advisor, Paul M. Coussens, for his guidance,
friendship, encouragement, financial support, and for creating a research
environment in which I could grow creatively, scientifically, and intellectually. I am
grateful to members of my guidance committee, Drs. S. Triezenberg, L. Velicer,
J. Ireland, and M. Vandehaar for their helpful suggestions and discussions.

I gratefully acknowledge all the people I have worked with in Dr. Coussens
lab for their helpful conversations and technical prowess. I especially thank R.
Southwick for his expertise with computers and excellent technical assistance.

A sincere thanks to my parents Wanda Wilson and David Jordan; my sisters
Wendy Middleton and Sandy Geiswite; and my mother-in-law Marge Daeschlein

for there continuous encouragement.

vi

Table of Contents.

Chapter 1.
LITERATURE SURVEY ...................................... 1
I. HERPESVIRUSES .................................. 2
1. Introduction. .................................. 2
2. Herpesvirus classification. ........................ 5
3. Temporal Classification of herpesvirus genes. .......... 6
4. Herpesvirus proteins and their functions. ............. 6
a. HSV I IE proteins and their functions. ......... 6
b. VZV and EBV IE genes. ................... 7
c. Early and Late genes. ...................... 8
d. ”Non-essential" herpesvirus genes. ............ 9
II. MOLECULAR BASIS FOR VIRAL ATTENUATION. .......... 10
1. The poliovirus story. ........................... 10
2. Other RNA viruses. ............................ 14
3. Vaccinia virus. ................................ 16
4. Herpesviruses. ............................... 18
III. HISTORY OF VIRUS INDUCED CANCERS ................ 21
1. RNA viruses. ................................. 21
2. DNA tumor viruses. ............................ 22

vii

ITI‘

3. Herpesvimses. ................................ 23

IV. MAREK’S DISEASE VIRUS. .......................... 25
1. Pathology of Marek's Disease virus. ................ 25

2. Biology of MDV. ............................... 29

3. MDV genome structure. ......................... 31

4. MDV Proteins: general information. ................ 34

5. Molecular biology of MDV oncogenicity. .............. 35

a. Serotype comparisons. ..................... 37

b. MDV transformed cell lines ................... 38

6. MDV vaccines. ................................ 41

7. Attenuation of MDV. ............................ 45

a. Genome structure. ....................... 45

b. 90 expression. ........................... 47

C. Attenuation of v.v. MDV. ................... 48

Chapter 2. .............................................. 51

PURIFICATION AND CHARACTERIZATION OF INFECTIOUS MAREK'S
DISEASE VIRUS GENOMES USING PULSED FIELD

ELECTROPHORESIS.

ABSTRACT ......................................... 52
I. INTRODUCTION. .................................. 53
II. MATERIALS AND METHODS ......................... 55
III RESULTS ....................................... 59
IV. DISCUSSION .................................... 63
IV. ACKNOWLEDGMENTS ............................. 65

viii

Chapter 3. ............................................... 78

STRUCTURAL ALTERATIONS IN THE REPEAT REGIONS OF THE
ATTENUATED MAREK’S DISEASE VIRUS GENOME AFFECTS
GROWTH AND IMMEDIATE EARLY GENE TRANSCRIPTION IN

VITRO.
ABSTRACT ......................................... 79
I. INTRODUCTION. ................................... 80
II. MATERIALS AND METHODS .......................... 83
III. RESULTS ........................................ 87
IV. DISCUSSION ..................................... 95
ACKNOWLEDGMENTS ................................. 1 00
Chapter 4. ............................................... 119

MOLECULAR ANALYSIS OF THE GLYCOPROTEIN C-NEGATIVE
PHENOTYPE OF ATTENUATED MAREK’S DISEASE VIRUS.

ABSTRACT ......................................... 120
I. INTRODUCTION .................................... 121
II. MATERIALS AND METHODS .......................... 123
III. RESULTS ........................................ 128
IV. DISCUSSION ..................................... 133
ACKNOWLEDGMENTS ................................. 1 38
Chapter.5 ................................................ 150
DISCUSSION
I. SUMMARY OF RESULTS AND CONCLUSIONS. ............ 150

II. THEORETICAL RELATIONSHIPS OF MDV ATTENUATION
AND MDV-INDUCED ONCOGENICITY: WORKING
HYPOTHESES TO BE TESTED. ..................... 155

III. FUTURE RESEARCH DIRECTIONS. .................... 158

ix

Technique development. .......................... 158

MDV attenuation. ............................... 159

Vaccine development. ............................ 162

APPENDIX ............................................... 164
DESCRIPTION OF METHODS

General information. ................................... 164

CHEF. ............................................. 164

Immunoprecipitation of proteins. ........................... 165

Nuclear run off transcription assay. ........................ 165

LIST OF REFERENCES ..................................... 171

LIST OF TABLES

Chapter 2.

TABLE 1. Genome Sizes of MDV Strains as Determined from

High Resolution CHEF Gels. .................................. 73
Chapter 3.

Table 1. Abundance of viral transcripts which hybridize to the MDV BamHI-L
DNA fragment. ............................................ 1 12
Table 2. Temporal classification of viral transcripts which hybridize to the MDV
BamHI-L DNA fragment. ..................................... 112
Table 3. Incidence of Marek’s Disease with oncogenic and attenuated Md11
and Md11p83-L. ........................................... 118
Chapter 4.

Table 1. Abundance of viral transcripts in cells infected with LP and HP MDV
strain Md11. ............................................. 146

xi

 

LIST OF FIGURES

Chapter 1.

Figure 1. Pathway to MDV transformation of T-Iymphocytes ............ 27
Figure 2. Comparison of HVT and HSV genomic segment sizes to serotype 1
MDV genomic segment sizes. ................................. 33
Figure 3. Location of MDV genes on the MDV genome. .............. 36
Chapter 2.

Figure 1. Separation of oncogenic and attenuated cell-free MDV genomes from
chicken chromosomes using PFE. .............................. 68

Figure 2. Identification and localization of cell-free MDV genomes and chicken
Chromosomes on CHEF gels. ................................. 70

Figure 3. Infectivity of cell-free MDV genomes after isolation from LMP agarose
CHEF gels ................................................ 72

Figure 4. Cell-free genome size comparison of oncogenic, attenuated, and
naturally occurring non-oncogenic MDV serotypes using high resolution CHEl-75

Figure 5. Characterization of alterations in attenuated MDV genome structure

using CHEF purified call-free MDV genomes as hybridization probes. . . . . 77
Chapter 3.
Figure 1. Identification of an attenuated MDV strain Md11 variant. ....... 102

Figure 2. Identification of the altered DNA fragment in the attenuated Md11
variant as BamHI L. ........................................ 104

xii

Figure 3. Comparison of plaque morphology of oncogenic and attenuated Md11
to Md11p83-L . ........................................... 106

Figure 4. Panel A: comparison of plaque formation of oncogenic and
attenuated Md11 to Md11p83-L in CEF. Panel B: comparison of viral DNA

replication of oncogenic and attenuated Md11 to Md11p83-L in CEF. . . . . 108
Figure 5. Comparison of rates of plaque formation and DNA replication (day 1-
3) of oncogenic and attenuated Md11 to Md11p83-L in CEF. .......... 110
Figure 6. Restriction mapping to determine sub-localization of the 200 bp
deletion in the attenuated Md11 genome. ........................ 114
Figure 7. Sub-localization of the 200 bp deletion in the BamHI-L fragment of
attenuated Md11 genomes. ................................... 116
Chapter 4.

Figure 1. Immunoprecipitation of 35S-labeled viral proteins from cell Iysates and
media of DEF infected with oncogenic or attenuated MDV strain Md11. . . . 140
Figure 2. RFLP analysis of the MDV strain Md11 gene locus and flanking
regions. ................................................. 142
Figure 3. MDV strain Md11 gC steady state RNA levels in DEF infected with
oncogenic and attenuated isolates. ............................. 144
Figure 4. Transcription rates of MDV strain Md11 genes. ............. 148
Chapter 5.

Figure 1. Hypothetical relationships between mutations of attenuated Md11 and
the pathway to MDV transformation of T-Iymphocytes ................. 157
Appendix

Figure 1. Clamped homogeneous electric fields electrophoresis. ........ 167
Figure 2. Resolution of large DNA molecules using CHEF. ............ 168

xiii

Figure 3. Immunoprecipitation of proteins from MDV infected cells. ...... 169

Figure 4. Nuclear run off transcription assay. ...................... 170

xiv

 

pCl/ml
281 MI/1
a.a.
ANOVA
DP
CEF
CHEF
CHX
cm
CPM/ul
DEF
DNA

EtBr

fMoles

GA

96

LIST OF ABBREVIATIONS

microcuries pre milliliter
serotype 2 MDV strain

amino acid

analysis of variance

base pairs

chick embryo fibroblasts
clamped homogeneous electric fields
cycloheximide

centimeter

counts per minute per microliter
duck embryo fibroblasts
deoxyribonucleic acid

eany

ethidium bromide

femtamoles

serotype 1 MDV strain
glycoprotein B

glycoprotein C

XV

HSV
HP
HVT
ICP
Ics
IE
IRL
IR:
JM
kbp

LMP
LP

MD
Md1 1
MDNA
MDV
"'99
min
nm

nt.

°C

herpes simplex virus
high passage
herpesvirus of turkeys
infected cell protein
immune chicken sera
immediate early
internal repeat long
internal repeat short
serotype 1 MDV strain
kilobase pairs

late

low melting point agarose
low passage

mole

Marek’s disease
Serotype 1 MDV strain
Marek's disease nuclear antigen
Marek’s disease virus
Marek’s EcoRl 0
minute

nanometer

nucleotide

degrees celsius

xvi

PAA
PFU
9938
RaA
RFLP

RNA

S.E.M.

sa-1
8J1
SORF
TBE
TRL
T33
uFD

v.v.
VIA

VP

phosphonoacetic acid

plaque forming unit

phosphoprotein of 38 molecular weight
rabbit-anti-A (gC)

restriction fragment length polymorphism analysis
ribonucleic acid

standard error of the mean

serotype 2 MDV

standard deviation

specific open reading frame
tris-borate-EDTA buffer

terminal repeat long

terminal repeat short

microFarad

microgram

unique long

unique short

volt

very virulent

viral internal antigen

virion protein

xvii

Chapter 1.

LITERATURE SURVEY

 

l. HERPESVIRUSES
1. Introduction.

Herpesviruses are a large group of DNA viruses disseminated throughout
the animal kingdom. These ubiquitous viruses cause disease in most organisms
infected, thus they are intensely studied. In humans, herpesviruses are the
causative agent of many common illnesses. Human herpesvirus infections cause
cold sores (Herpes simplex virus I, HSV I), genital herpes (Herpes Simplex II, HSV
ll), mononucleosis (Epstein barr virus, EBV), Chicken pox and shingles (Varicella-
Zoster virus, VZV) (Volk, 1982). Many human herpesviruses can be life
threatening to infants. Cytomegalovirus (CMV) is associated with congenital
conditions such as mental retardation, microcephaly (small head), seizures,
deafness, and jaundice (Volk, 1982). HSV II can be sub-Clinical or fatal to a
newborn if an active infection occurs during childbirth. In some cases only mild
Clinical symptoms occur in adults, with the exception of Herpes B virus. B virus
is fatal in 75% of cases (Volk, 1982).

Some herpesviruses have been implicated as the causative agent of some
types of cancers. EBV is a suspect in human lymphomas. Lucke frog virus is
associated with adenocarcinomas in frogs. HSV I is implicated in adeno- and
fibrosarcomas. Herpesvims saimiri and H. Ateles are suspected as being the
agent for lymphomas of monkeys. Marek’s disease virus (MDV) is the cause of
lymphomas in chickens (Grodzicker and Hopkins, 1981).

Herpesviruses that cause disease in cattle, sheep, goats, horses, swine,

chickens, and turkeys have been identified (Review- FraenkeI-Conrat et al, 1988).

3

Two herpesviruses that infect cattle (BHV) cause rhinotracheitis, pustular
vulvovaginitis, mammillitis, pseudolumpy skin disease and spontaneous abortion
in cattle. Equine herpesvirus (EHV) can also cause abortion in horses and
rhinopneumonoitis. Pseudorabies virus (PRV) infects swine and muses Aujeszky’s
disease. MDV infects chickens and causes paralysis and lymphoma, commonly
termed Marek's disease (MD) (Payne, 1985).

Herpesviruses are relatively large in size compared to other viruses
measuring 120 - 300 nm in diameter (Roizman, 1990). They contain a core in
which the viral DNA resides. The core is encased by a capsid which is, in turn,
surrounded by tegument. The entire virion is protected by an envelope. The
envelope contains viral-encoded glycoprotein spikes protruding outward.

Herpesviruses infect host cells by first attaching to the cell membrane. This
initial step requires the interaction of viral glycoprotein spikes and the cell surface.
After attachment, the virus penetrates the cell surface by a combination of
phagocytosis and fusion of the virion envelope with the cell membrane, a step
which also requires viral-encoded glycoproteins. The virus must then uncoat and
get to the nucleus where, with the help of host cell machinery, it initiates
transcription of viral genes and DNA replication (Review- Volk, 1982). All
herpesviruses encode a variety of enzymes involved in nucleic acid metabolism,
DNA synthesis, and transcriptional regulation (Roizman, 1990). Herpesvirus
transcripts are transported to the cytoplasm where protein synthesis occurs (Volk,
1982). Viral glycoproteins are inserted into the nuclear membrane and the virion

is assembled within the nucleus (Volk, 1982). Assembled virions then bud from

4

the nuclear membrane. The nuclear membrane containing virus encoded
glycoproteins forms an envelope around the virus. Eventually, the cell dies, Iyses,
and enveloped viruses are released to invade neighboring cells (Volk, 1982).

During the course Of infection, herpesviruses "hide out" or become "latent".
In latent herpesvirus infections, only a limited number of viral genes are expressed,
and, in most cases, the genome exists as a Closed circular molecule. During a
latent infection, no clinical signs of disease are evident. However, upon
reactivation of latent herpesviruses, due to stress or immunosuppression, the virus
embarks on a cytolytic life cycle described above. Due to the lack of viral protein
expression during latency, the immune system is ineffective at ridding an organism
of latent herpesviruses. Thus, in most cases, an organism that is infected with a
herpesvirus is infected for life (Review— Roizman, 1990).

The biological characteristics of a disease process caused by herpesviruses
depends upon the interaction of virus encoded proteins with its host cell and host
encoded proteins. To understand the disease process of herpesviruses, molecular
biology techniques have been utilized to identify viral-encoded proteins and study
their interaction with the host cell. The most intensely studied herpesvirus and
guide for studying other rat-herpesviruses is HSV l. The following sections provide
an overview of what is presently understood about HSV proteins at the molecular
level. Other herpesviruses will be discussed when relevant.

2. Herpesvirus classification.
Herpesviruses are classified based upon genome structure and biological

properties. Genome structure classification is based on the location and

5

arrangement of repeated sequences within the genome. Six classes of genome
structure exist within herpesviruses (Roizman, 1990). In most cases, genome
structure Classification correlates with classification based on biological properties.
Classification of herpesviruses based on biological properties has led to some
controversy over classification of viruses which have genome structure similarities
to one virus group but biological properties similar to another group. Such is the
case with MDV. Previously, MDV was Classified as an y-herpesvirus. However,
due to recent information on MDV genome structure, MDV is presently unclassified
(Roizman, 1992). Classification based on biological properties is summarized

below (Roizman, 1990):

(II-herpesviruses are characterized by wide host range, short replication cycle,
and usually latent infection of sensory root ganglia (HSV l, HSV II, VZV,
BHV, EHV, PRV, and B virus).

8-herpesvlruses are characterized by relatively restricted host range, long
replication cycle, and latency in secretory glands, lymphoreticular cells,
kidney and other tissues (CMV).

y-herpesvlruses are notable for infecting only T and B lymphocytes. Latency is

confined to lymphoid tissues (EBV, H. Saimiri, and H. Ateles).

3. Temporal classification of herpesvirus genes.
After herpes simplex virus infects a cell, expression of its genes occurs in

a tightly regulated, temporal fashion (Review- Wagner, 1991). The virion carries

 

 

 

 

6

within it a protein called VP16. VP16 is a transactivator that, in conjunction with
cellular proteins, initiates transcription of immediate early (IE) genes . IE genes
encode regulatory proteins required for the synthesis of early (E) and late (L)
genes. E genes primarily encode proteins involved in viral DNA replication, while
most L genes are expressed only after the onset of viral DNA replication.
However, DNA synthesis is not a strict requirement for L gene expression. L
genes encode structural proteins for virion assembly and VP16. IE genes are
expressed immediately upon infection, and IE gene transcripts will accumulate in
the absence of viral protein synthesis (Roizman and Sears, 1990). E gene
transcripts can be detected as short as 30 minutes after infection and will
accumulate if DNA synthesis is blocked by chemical inhibitors (Wagner, 1991).
Maximal expression of L genes can be detected at 3 hours post-infection (Wagner,
1991). Sequence analysis of the promoter regions of representative IE, E, and L
genes indicate that, in most cases, each class has specific sequence and

structural motifs that define the gene expression class.

4. Herpesvirus proteins and their functions.

a. HSV l IE proteins and their functions. There are five known HSV I IE
proteins: ICPO, ICP4, ICP27, ICP22, and ICP47 (Roizman and Sears, 1990). Of
this group, the first four are known to affect herpesvirus gene expression. ICPO
is non-essential in cell culture, is a promiscuous transactivator acting through
another protein (Sandri-Goldin, 1991 ), and is hypothesized to be under antisense

regulation during latency (Stevens et al, 1987). In addition, virus mutants which

 

7

do not express ICPO exhibit an larger particle to PFU ratio as compared to wild
type viruses, suggesting that ICPO may confer a growth advantage to HSV l
(Sandri-Goldin, 1991). ICP4 is essential for activation of E and L gene expression.
In addition, ICP4 down-regulates its own promoter and other IE gene promoters.
ICPO and ICP4 act synergistically to enhance E and L gene expression during
transient assays (Sandri-Goldin, 1991). ICP27 is essential for viral growth,
negatively regulates IE and E genes, is non-essential for DNA replication, and is
required for L gene expression. ICP27 may be involved in the switch from E to L
gene expression (Sandri-Goldin, 1991). ICP22 is essential for L gene expression
in some cell types and may be required for repression of cellular proteins (Smiley
et al, 1991).

b. VZV and EBV IE genes. Regulation of VZV gene expression appears
to be similar to HSV I gene regulation (Ruyechan et al, 1991). However, only two
VZV regulatory proteins have been well characterized thus far. ORF62 is the
functional homolog to HSV ICP4 and ORF61 is functionally analogous to HSV
ICPO (Moriuchi, et al, 1992). EBV gene regulation during lytic infection has not
been well characterized due to lack of appropriate in vitro culture systems
(Hayward and Hardwick, 1991). However, three IE genes have been identified:
Zta, Mta, and Rta. Zta (or ZEBRA) appears to be essential for induction of lytic
cycle gene expression (Hayward and Hardwick, 1991). During latent EBV
infection, EBNA 1 regulates at the level of DNA replication, and EBNA 2 is
essential for growth transformation of lymphocytes (Raab-Traub and Gilligan,

1991). Since temporal classification of herpesvirus proteins are determined during

 

 

8

lytic infection, EBNA 1 and 2 are not Classified as IE proteins, but are considered
regulatory proteins. EBNA 1 and 2 transcripts contain multiple splice sites. EBNA
2 has up to 11 intervening sequences spanning 25 kb (Raab-Traub and Gilligan,
1991)

c. Early and Late genes. Early genes encode enzymes and proteins
involved in DNA metabolism and replication. Their expression is dependent upon
activation by IE proteins (Roizman, 1990). The early proteins which have been
well characterized are viral ribonucleotide reductase, dUTPase, major DNA binding
protein, thymidine kinase, and DNA polymerase (Roizman and Sears, 1990).

Late genes encode structural components of the virion and envelope. As
with early genes, late gene expression is dependent upon activation by IE proteins,
especially ICP4, ICPO, and ICP27 (Blair and Snowden, 1991). The well
characterized late genes are those which encode VP16 and glycoproteins. VP16
as described above, activates expression of IE genes and initiates the cascade of
herpesvirus gene expression in lytic infections (Everett et al, 1991). The VP16
protein is produced late during infection and is bound to the tegument during virion
assembly (Roizman, 1991).

Glycoproteins encoded by L genes are the major antigenic determinants of
herpesviruses. These are the molecules which are most often recognized by the
host immune system. For the virus, glycoproteins function in virus attachment and
entry into the cell. Six HSV glycoproteins have been well Characterized: 98, 9C,
gD, gE, 9H, and gI (Roizman and Sears, 1990). Attempts to elucidate the role of

glycoproteins during infection have yielded the following conclusions (Roizman and

9

Sears, 1990): 98 binds to heparin sulfate and functions during penetration and cell
fusion. Neutralizing antibodies and cell-mediated immune response are produced
against QB. 98 is the most highly conserved glycoprotein among a, B, and y
herpesviruses. 90 also has an affinity for heparin sulfate and may function during
attachment of herpesvirus to its host cell. The host will produce antibody against
90, however, it is unclear if anti-gC antibodies are protective. Oddly, 90 can bind
the c3b component of complement and accelerate its decay. Functional
significance of C3b binding by 9C is unclear. 9C is also conserved among cr-
herpesviruses. Glycoprotein D functions during penetration and cell fusion and is
the most potent inducer of neutralizing antibodies among herpesvirus
glycoproteins. When complexed with gl, gE forms a high affinity receptor for host
IgG. As with gC-c3b interactions, the functional significance of gE-gI-lgG binding
is unclear. gE is not, itself, a potent inducer of humoral immune response. gH
functions to stabilize attachment of virions to the host cell during penetration and
cell fusion and can induce neutralizing antibodies. gH is also highly conserved
among herpesviruses.

d. "Non-essential" herpesvirus genes. Experiments utilizing herpesvirus
mutants and antibody blocking studies have indicated that some HSV genes are
not required for growth In vlfro. Some of these non-essential genes encode: gE,
gl, gC, ribonucleotide reductase, dUTPase, and ICPO (Roizman and Sears, 1990).
Although they are not required in cell culture (a basic system), and therefore are
often considered non-essential, they may, in fact, function in a cell-specific or

tissue-specific manner in the host organism. In addition, these proteins may play

10

a role in the characteristic manifestation of a specific disease process.

ll. MOLECULAR BASIS FOR VIRAL ATTENUATION.

Viral attenuation leads to reduced ability of viruses to induce disease in their
natural host. The most notable attenuating mutations affect the interactions of
viral-encoded proteins with host cells (Review- Tyler and Fields, 1990). In this
section, literature related to analysis of the molecular mechanisms of attenuation
of viral pathogenesis will be reviewed. Discussion will start with poliovirus, one of
the smallest viruses known to cause disease, and end with vaccinia virus and
herpes virus, the largest viruses known to cause disease.

1. The poliovirus story.

Paralytic poliomyelitis has afflicted humankind for centuries. The earliest
known record of poliomyelitis is found on an Egyptian stone carving dating 1500
BC (Melnick, 1982). From ancient times to the end of the 19'" century, polio was
reported as sporadic cases in endemic proportions (Melnick, 1982). During the
late 19‘" and early 20'" centuries, urbanized industrial centers of Europe and the
United States experienced polio epidemics of increasing frequency and severity.
Although concern for this disease was great, research on polio advanced slowly,
with isolation of the causative agent in 1909 leading to development of effective
vaccines in the mid 1950s (Salk, 1953; Sabin, 1957). During the twelve years after
introduction of the inactivated (Salk) and live (Sabin) polio vaccines, a 99%
decrease in reported poliomyelitis was observed (Melnick, 1982). However, polio

cases continued to be reported and 17% of reported cases were in children who

11

had received the vaccine (Melnick, 1982). Presently, a world wide residual level
of poliomyelitis of 0.84 cases/million of population/year continues to exist, and is
associated with vaccinated populations (Melnick, 1982).

The causative agent of paralytic poliomyelitis is poliovirus, a small virus
containing a single stranded RNA genome of 7450 nucleotides (Putnak and
Phillips, 1981). The RNA genome is translated as a single polycistronic protein.
Extensive splicing and proteolytic cleavage occurs after transcription. The
poliovirus genome encodes four structural proteins (VP1-VP4), a protease, a
polymerase, and VPg (protein primer for RNA synthesis) (Phillips, 1981). There
are three distinct serotypes of poliovirus, and each serotype causes identical
disease symptoms. Serotype 3 poliovirus is associated with vaccine breaks at a
higher frequency than serotypes 1 or 2 (Collingham et al., 1978). The residual
level of poliomyelitis and sporadic episodes of polio associated with vaccinated
populations prompted attempts to' understand the molecular mechanisms of
attenuation of the Sabin polio vaccine. In 1984, Stanway et al. reported on the
comparison of the complete nucleotide sequence of serotype 3 wild type polio and
its attenuated vaccine derivative. Their data shows that only 10 point mutations
occurred during attenuation and the 3’ and gained a single base (G). 01 these ten
differences, 7 do not result in amino acid (a.a.) changes. The 3 Changes which
did result in a.a. substitutions were located within VP3, VP1, and the non-
structural protein. Two mutations reside within the 5’ noncoding region of the RNA
genome. Based on sequence comparisons, these authors concluded that only

mutations resulting in a.a. substitutions were responsible for attenuation of

12

serotype 3 poliovirus neurovirulence. Further clarification of mutations responsible
for attenuation of neurovirulence came from comparative studies of the genome
sequences of serotype 3 wild type, its vaccine derivative, and a vaccine revertant
(Pa/119) recovered from a patient with a fatal case of vaccine-associated
poliomyelitis (Cann et al., 1984). P3/119 differed from the vaccine strain at 7
nucleotide locations. One change in the 5' non-coding (nt. 472) was a true
reversion to the wild type nucleotide, three Changes resulted in a.a. substitutions
in VP1 and VP2, and one change resulted in loss of two nucleotides at the 3’ end
just prior to the polyadenylation tract. Evidence of true reversion at nt. 472 to wild
type prompted studies on partial sequence characterization of vaccine strains
recovered from infants after primary vaccination (Evans et al., 1985). In all
vaccine vims isolates tested, nt. 472 reverted to the wild type nt. 48 hours after
viral replication in the gut of infants that received the live vaccine. Direct evidence
for the involvement of nt. 472 in attenuation of neurovirulence came from studies
of Westrop et al. (1989). They constructed infectious Chimeric cDNAs between
wild type and vaccine strains or wild type, revertant and vaccine strains. After
transfection in cell culture to produce infectious vims, virus was injected into
monkeys to determine the level of neurovirulence for each poliovirus recombinant.
Mutations strongly correlated with attenuation of neurovimlence were located in the
5’ non-coding region (nt. 472) and a single a.a. substitution in VP3 (nt. 2034),
which results in a serine to phenylalanine substitution. However, each mutation
alone did not have a pronounced affect on attenuation, thus these mutations were

additive. To elucidate the biological significance of nt. 472, recombinant viruses

 

 

13

were tested fortheirtranslational efficiency in a cell-free translation system (Svitkin
et al., 1990). Recombinant viruses containing the attenuated nt. 472 displayed
reduced translational efficiency compared to the wild type parent. In addition,
recombinants containing the wild type base at nt. 472 in attenuated background
displayed higher translational efficiency compared to attenuated virus.
Translational efficiencies directly correlated with level of neurovirulence in monkey
tests. During analysis of a protease-negative mutant, functional significance of
the 5’ non-coding region was elucidated. Andino et al. (1990) determined that
interaction of stem-loop structures with 30pro (protease Cleavage product of the
nonstructural protein) plays an essential role in RNA replication. Disruption of
stem loop structures encompassing the attenuating nt. 472 results in increased
temperature-sensitivity (a marker for attenuated poliovirus) indicating that RNA
secondary structure in the 5’ noncoding region is important for poliovirus
attenuation (Macadam et al., 1992).

Biological significance of the single a.a. change in VP3 is less well
understood. The same wild type/vaccine strain recombinants described by
Westrop et al. (1989) were tested for their ability to induce III-interferon in human
leukocytes in vitro (Hovi et al, 1991). Their results indicate that attenuated
serotype 3 poliovirus induces less interferon than its wild type counterpart. In
addition, induction of or-interferon is directly associated with the VP3 capsid
protein.

Similar studies were carried out for attenuated serotypes 1 and 2 poliovirus.

In contrast to attenuated serotype 3 poliovirus, attenuated serotype 1 and 2

14

poliovirus contain 56 and 22 nucleotide changes, respectively (Christodoulou et al.,
1990; Moss et al., 1989). For serotype 1, nucleotide changes correlated with
attenuation of neurovirulence are located in the 5’ noncoding region (nt. 480),
polymerase, and the capsid region (Christodoulou et al., 1990). For serotype 2,
nucleotide changes associated with attenuation of neurovirulence are located
within the 5' noncoding region (nt. 481) and the capsid protein VP1 (Moss et al.,
1989). Differences in cell culture and in viva passage histories leading to the
attenuated phenotype (Sabin and Boulger, 1974), may be related to the differences
in the attenuating nucleotides within the coding regions of all three serotypes. The
consistency of the attenuating nucleotide in the 5' noncoding region would be
expected as this region has the highest degree of homology among all three
serotypes (Toyoda et al., 1984).

In summary, studies on poliovirus indicate that attenuation of a viral induced
disease state can occur by perturbation of viral transcriptional or translational
mechanisms and alteration of interactions between capsid proteins and the
poliovirus receptor (Moss and Racaniello, 1991).

2. Other RNA viruses.

A survey of other RNA viruses containing polycistronic RNA genomes
indicates that attenuation by mutations in the 5’ non-coding region is a common
theme. Two separate groups have reported sequence data comparisons of
virulent and attenuated Japanese encephalitis virus (Aihara, et al, 1991;
Nitayaphan at al, 1990). Although their results differ in location of changes within

the coding region, both groups found mutations within the 5’ non-coding region of

15

attenuated isolates to be important for reduced virulence. In addition, Pritchard et
al.(1992) has reported that a single nucleotide Change located in the 5’ non-coding
region of Theiler’s virus attenuates neurovirulence in mice. The Change alters
predicted stem loop structures in the 5’ non-coding region.

Kuhn et al. (1992) reported that defined mutations in the 5’ and 3’ non-
coding regions of Sindbis virus results in attenuation of virulence. They also
demonstrated that growth properties of mutant viruses were different in mouse
cells compared to chicken or mosquito cells. They hypothesize that cell-specific
host proteins which interact with the non-coding regions are different among cell
types, resulting in altered growth properties. From in vivo studies, Kuhn et al.
concluded that changes in the 5’ and 3’ regulatory regions may have tissue-
specific as well as host-specific affects.

Aside from alterations in regulatory regions of attenuated RNA viruses, a
few mechanisms of attenuation affecting glycoprotein genes have been reported.
Sequence analysis of an attenuated Rift valley fever virus indicates that a new
AUG codon, upstream of that which initiates translation of a viral glycoprotein
precursor, was evident. Translation initiation at this new AUG codon may alter
glycoprotein structure (T akehara et al., 1989). Lobigs et al. (1990) attenuated
Murry valley encephalitis virus by 10 passages in human cells. Sequence analysis
of the E glyCOprotein gene of 6 attenuated isolates revealed an alteration in the
RGD sequence (RGD: arginine-glycine—asparginine sequence found in the leader
peptide of adhesion molecules) of ng in 5 of the 6 isolates. Since the RGD

sequence is highly conserved among flaviviruses, the authors conclude that the

16

RGD sequence is an important general determinant for flavivirus pathogenicity.

Thus, there may be a variety of molecular mechanisms of viral attenuation.
Most notable are alterations in regulatory regions which may affect RNA
transcription or replication and alter the interaction of tissue or cell-specific proteins
with these regulatory regions. Mutations within the coding sequence of
glycoprotein genes can alter their structure affecting initial virus-host interactions.
3. Vaccinia virus.

As with poliovirus, poxviruses are ancient with a 3000 year recorded history
of human disease (Volk, 1982). Poxvirus is the largest and most complex animal
virus and was the first and only virus to be seen in the light microscope. Human
poxvirus causes smallpox, a deadly disfiguring disease characterized by large pox-
Iike lesions on the skin. Smallpox vaccines were the first vaccines to be
developed. In 1878, E. Jenner used poxvirus isolated from cows to produce a
mild form of disease in humans. This mild infection resulted in effective
immunization against human smallpox (Review— Fraenkel-Conrat et al., 1988).
Since 1977, smallpox is considered extinct due to massive immunization programs
using the related vaccinia virus strain directed by the World Health Organization
(Volk, 1982).

Poxviruses share many common features with herpesviruses: large size,
genome structure, and type of genes encoded. Poxviruses contain a linear
double-stranded genome of 130 to 300 kbp (Moss, 1990). As with herpesviruses,
the poxvirus genome encodes genes for macromolecular synthesis. In contrast

to herpesviruses however, poxvirus transcription and virus replication occur in the

 

17

cytoplasm. The poxvirus genome consists of a unique long region with ends
flanked by inverted terminal repeats. Although the unique long segment is highly
conserved among poxviruses, the size of terminal repeats can vary from 2.3 to
12.5 kbp (Moss, 1990).

Development of poxviruses as vectors for foreign genes and recombinant
vaccines has raised concern about retention of virulence within poxvirus vectors,
especially vaccinia virus. This concern has prompted researchers to identify
virulence factors and attenuating regions within the vaccinia virus genome.
Attenuation of vaccinia vims by serial passage in chick embryo cells has two major
effects on poxvirus genome structure. As a result of intermolecular recombination,
large deletions occur within the unique long region, near the ends (Review-Moss,
1990). In addition, 70 bp tandem repeats within the terminal repeat regions
become amplified by tandem reiteration. Amplification by tandem reiteration is
thought to occur by homologous recombination between flanking short repeats
(Moss, 1981). Presently no phenotypic alteration has been assigned to
amplification within terminal repeat regions.

Site-directed deletions within the unique region of the vaccinia virus genome
have been characterized with respect to attenuation of pock formation in mice (Lee
et al., 1992). Most notable, mutations or deletions in the ribonucleotide reductase
gene or the hemagglutinin gene resulted in attenuated phenotypes. Deletions in
the thymidine kinase gene did not affect pock formation with some deletion
mutants, but other mutants showed decreased ability to form pocks. This

discrepancy could have been due to strain differences among mutants. An 11 kb

 

‘1.
l

18

deletion, located between the host range determinant gene and the ribonucleotide
reductase gene, resulted in the complete absence of pock formation. However,
the mutant virus grew to parental virus titers, indicating a highly attenuated
phenotype. In addition, a 20 kb deletion located at the left end of the unique long
region greatly reduced pock formation. Genes encoded within the 11 kb and 20
kb deletions are presently unknown (Lee et al, 1992).

Virulence determinants of vaccinia virus, proposed by other investigators
have been located in the VVGF (a protein that binds epidermal growth factor),
thymidine kinase, and a serine protease inhibitor homolog (located within the right
inverted terminal repeat region) (Tyler and Fields, 1990).

Although poxviruses were the first viruses identified and used for vaccines,
the molecular mechanisms of attenuation are difficult to determine due to its
extremely large size.

4. Herpesviruses.

As with vaccinia viruses, study of herpesvirus attenuation is difficult due to
their large genome size and the many genes they encode. Presently, studies
related to attenuation of herpesvirus virulence are carried out using defined
mutants (HSV, PRV) and attenuated vaccine viruses (PRV). Some recent findings
will be discussed in this section. Most reports, with one exception, are at the level
of identifying genes involved in virulence. One paper (Robbins et al, 1984) reports
results which hint at a possible mechanism of attenuation.

Pseudorabies virus, an alphaherpesvirus which infects swine, causes an

infection similar to that of Herpes simplex virus. Due to its economic importance,

 

19

an attenuated PRV was developed and is currently used for vaccination of swine
in Europe (Card at al., 1992). Virulent PRV are neurotropic and produce lytic
Infection in the central nervous system of swine. Attenuated PRV is also
neurotropic, but does not cause pronounced lytic infection. The mechanism of
attenuation remains to be elucidated, but evidence indicates that alterations of
glycoproteins in the viral envelope may be important. The genome of attenuated
PRV contains mutations in the UL and Us regions. A large deletion in the Us has
removed coding sequence for gl and gp63 genes. These two glycoproteins form
a complex required for neurovirulence. Mutations in the UL region effect glll and
the BamHl-4 fragment which encodes 4 genes thought to be involved in capsid
assembly. Functional significance of these mutations has been investigated.
Mettenleiter et al. (1987a, 1987b) reported that absence of 91 affects release of
virus from rabbit kidney cells and plays a major role in neurovirulence. Robbins
et al. (1989) investigated the significance of mutations in the gill gene of
attenuated PRV genomes. To eliminate the contributions of other deletions in
attenuated PRV, they constructed recombinants with the glll gene from attenuated
PRV in a wild type background. They showed that, although steady state RNA
levels were the same with attenuated glll as compared to wild type, protein levels
from in vitro translation and purified virions were greatly reduced. However, levels
of glll from immunoprecipitations of infected cell lysates were the same for
attenuated glll as compared to wild type. Sequence analysis indicated that a
nucleotide change in the attenuated glll signal sequence caused a change from

leucine to proline. The authors conclude that mutation in the signal sequence

20

resulted in improper protein localization, and that glll accumulates in the cytoplasm
of infected cells instead of being inserted into the virus envelope.

Card at al. (1992) constructed a series of recombinant viruses containing
deletions in the glll plus gl genes, glll alone, or gl alone. These mutants were
then tested in viva using the rat visual neuronal system. Patterns of infection and
specific cell types infected were able to be followed in this system. glll-negative
mutants did not attenuate neurovirulence, however gl-negative mutants were
substantially attenuated for neurovirulence. glll-negative plus gI-negative mutants
were more attenuated than the single gene mutants. In addition, they determined
that gl-negative mutants were unable to infect dorsal root ganglia but were able
to infect retinal ganglia. Also, glll was not required for infection of either cell type.
The authors conclude that gl has a significant impact on neurovirulence and PRV
glycoproteins confer cell type tropism to the virus.

gl, glll and gp63 are functional homologues of gE, g0 and gl, respectively
of HSV. Interestingly, these proteins are non-essential for growth in vitra, but
seem to be major virulence determinants in viva, at least for PRV.

Studies to define genes responsible for HSV virulence have used defined
mutants and clinical isolates. The thymidine kinase gene was considered to be
involved in neurovirulence and latency reactivation. However, some investigators
argue that thymidine kinase mutants still retain neurovirulence. (Review— Fields,
1990). In the mouse model, additional genes which attenuate neurovirulence are
gE, gG, IE22, ribonucleotide reductase, and protein kinase. Recently, Robertson

et al. (1992) have reported on a neurovirulence gene located in the terminal

21

portion of the long repeat region. A mutant containing a 759 bp deletion in this
gene fails to grow in dorsal root ganglia, but is capable of replicating in the foot
pads of mice. These results suggest that the NV gene confers cell or tissue
specificity. In addition, the NV-negative mutant was able to establish latency but
was impaired for its ability to reactivate.

Another recent report has implicated the nan-essential gene, dUPTase in
being a neurovirulence determinant in the mouse model. dUTPase mutants were
able to replicate in the peripheral nervous system and establish latency, but
reactivated from latency with reduced efficiency. (Pyles et al, 1992). As with PRV,

HSV non-essential genes in vitra seem to be important for virulence in viva.

Ill. HISTORY OF VIRUS INDUCED CANCERS.

To prove the etiology of a disease, rules were adapted from postulates of
the late 19th century scientist, Robert Koch. Koch's postulates are summarized
below (Review— Volk, 1982):

1. The same organism must be found in all cases of a given disease.

2. The organism must be isolated and grown in pure culture from the

infected patient.

3. The organism from the pure culture must reproduce the disease when

inoculated into a susceptible animal.

4. The organism must then be isolated in pure culture from the

experimentally infected animal.
1. RNA viruses.

The role of viruses in the etiology of cancer was not appreciated until Rous
sarcoma virus (RSV), a single stranded RNA virus was shown to cause cancer in

experimental chickens. In 1911, Peyton Rous was able to induce tumors in

22

Chickens by injecting filtered extracts of tumor tissue isolated from a cancerous
chicken, thus establishing that a virus was the etiologic agent. After Rous’s
original report, many other RNA viruses were similarly suspected of playing a role
In cancer (Review— Weiss, 1984). The discovery that a virus can induce cancer
lead to speculation that the virus may encode a gene responsible for tumor
induction. It was not until 1970 that Martin reported that there was a genetic basis
for tumor induction by RSV. This viral genetic component was found to be a single
gene, src, encoded within RSV. Since then, many viral oncogenes have been
identified in RNA viruses (Review— Bishop, 1987). Most of these oncogenes are
structurally and functionally similar to transcription factors, receptors, hormones,
or metabolic messengers (Bishop, 1987). The final outcome of an active
oncogene in a cell is stimulation of cells to divide indefinitely.

2. DNA tumor viruses.

The involvement of DNA viruses in the etiology of cancer began with the
demonstration that papillama virus, using similar techniques of Rous, is the
musative agent of warts (Shop, 1932). Since then other DNA viruses (palyoma
virus, simian virus 40, adenoviruses, JC, and BK) have been blamed as the
culprits of cell transformation. The viral proteins implicated in cell transformation
by polyoma and SV40 are involved in stimulating viral DNA replication (T -antigens)
(Grodzicker and Hopkins, 1981). Similarly, regions of viral DNA associated with

transformation have been identified in adenovirus (Flint, 1981).

23

3. Herpesviruses.

The involvement of herpesvirus in cancer was difficult to establish. The
first hint that herpesviruses were associated with tumors came from studies by
Fawcett in 1956. Using electron microscopy, Fawcett was able to prove the
existence of herpesvirus (Lucké virus) particles in tumor tissue of frogs injected
with filtered extracts of naturally occurring tumors (zurHausen, 1981 ). However,
some tumors were free of any signs of herpesviruses. Thus, attempts to satisfy
Koch's first postulate were unsuccessful.

Herpes simplex virus I and II are associated with cervical carcinomas,
although the evidence is indirect. Many women suffering from cervical carcinoma
have antibody to HSV ll. However, virus particles or viral DNA in tumors have not
been reproducibly detected (Spear and Roizman, 1981). The ability of a certain
restriction endonuclease fragment to transform cells in vitra has been reported
(Camacho and Spear, 1978). CMV is also associated with human cancers in a
similar fashion as HSV. Patients suffering Kaposi's sarcoma have high antibody
titers to CMV, however CMV virions or DNA have never been detected in tumor
tissue (zurHausen, 1981). Due to lack of an appropriate in viva model system for
human herpesviruses, testing of Koch’s postulates is impossible. Other
herpesviruses which infect primates and are considered oncogenic are H. Saimiri
and H. Ateles. However these viruses are innocuous to their natural host, but
highly oncogenic to other primate species (zurHausen, 1981).

The human herpesvirus which is most strongly implicated as a causative

24

agent in cancer, and has prompted the most concern, is Epstein-Barr virus (EBV).
EBV was initially isolated from cultured Iymphoblastoid cells which originated from
a Burkitt’s lymphoma patient (Epstein et al, 1964). In addition, high antibody titers
to EBV are found in Burkitt’s lymphoma patients. Burkitt's lymphoma is a relatively
rare cancer occurring in children 5-12 years old only in certain regions of East
Africa (Burkitt, 1962) and New Guinea. Data which prompted concern was that
antibodies to EBV were found in 80% of healthy adults all over the world (Henle
and Henle, 1966). Although many investigators have attempted to find direct
evidence for the link between EBV and human cancers, the data remains only
circumstantial. However this Circumstantial evidence is strong: human cells can
be transformed by EBV in vitra; EBV can induce lymphoid tumors in non-human
primates; Most but not all Burkitt's lymphoma and nasopharyngeal carcinoma cells
carry EBV DNA. Again, attempts to satisfy Koch’s postulates failed.

Thus far, all herpesviruses discussed are associated with some form of
cancer or tumor formation, however testing of Koch's postulates has failed or
cannot be done in humans. Recovery of herpesvirus from only some cancer cells
may be due to coincident infection with oncogenic viruses. Failure to find virus
particles or viral DNA in some human tumor cells may be related to the
characteristics of herpesvirus latency. Herpesvirus particles are absent during
latency, and very few copies (undetectable) of herpesvims DNA are present in
latently infected cells. In addition, some herpesviruses may transform a cell by a
"hit and run" mechanism (zurHausen, 1981). The idea that herpesviruses can

cause cancer was proved by three studies. In 1967, Churchill and Biggs isolated

 

25

a herpes-like virus in cell culture and subsequently injected the virus infected cells
into Chickens. The experimental chickens developed lymphoid tumors (acute
Marek's disease). In 1970, development of Marek’s disease from injection of
purified cell-free MDV virions into experimental chickens also lead to MD (Nazerian
and Witter, 1970). Additional proof came with the development of effective
vaccines against Marek's disease. These were pivotal observations for the

involvement of herpesviruses in the etiology of cancer.

IV. MAREK’S DISEASE VIRUS.
1. Pathology of Marek’s Disease virus.

Clinical signs of Marek's Disease (MD) were firstidescribed by Joseph
Marek in 1907 (Payne, 1985). There are two distinct pathological forms of MD:
classical and acute (Payne, 1985). The classical form of MD is characterized by
impairment of neural function and cytolytic infection. The acute form of MD is
marked by lymphoproliferation and tumor formation and includes classical form
symptoms. A Chicken infected with MDV displays visible signs of nerve function
impairment, beginning with partial and ending with complete paralysis. A common
characteristic of Classical MD is uneven gait or one leg stretched forward and the
other leg stretched backward. Closed eyelids, lameness, and droopy wings are
also characteristic of nerve involvement (Purchase, 1985). This Classical form of
MD was considered the most common in the field, historically. Presently, there is
a higher prevalence for the acute form of MD (Payne, 1985). In some cases the

bird recovers, however in most instances, MDV infects lymphoid tissue and

I“- . '0', 2.1
a. In.
5‘
I.»
I

26

lymphocyte proliferation leads to the acute form of MD (Purchase, 1985). In
addition, the virus becomes latent in T-Iymphocytes, persisting throughout the
bird's lifetime (Schat, 1985). The cellular pathway to MDV transformation of T-
lymphocytes is summarized in Figure 1 and described in the following paragraph.

MDV is contracted from the environment via the respiratory system. MDV
travels from lungs to lymphoid tissue by a yet undetermined mechanism. The first
evidence (3-6 days post inoculation) of virus infection is in lymphoid organs
(spleen, thymus, and bursa) (Payne, 1985). Primarily B-lymphocytes, and a few
T-lymphocytes, are infected (Calnek et al, 1984). Initial infection of lymphoid tissue
is cytolytic in Character but does not result in the release of cell-free virions. This
"semi-productive” infection is marked by the presence of viral internal antigens
(VIA) and naked virus particles (without envelopes) within lymphoid tissues (Payne,
1985). Tissue Changes which accompany this initial infection are infiltration of
macrophages and granulocytes and reticular cell hyperplasia. The net effect is

enlargement of the Spleen and atrOphy of the bursa and thymus (Payne, 1985).

27
MDV INFECTED CELL

 

.
A. ' 2
-

 

CYTOLYTIC PHASE
LATENT PHASE
fa"
a-T :-
Al"
“'12 (3::

Figure 1.

Pathway to MDV transformation of T-lymphocytes.

 

28

From 5 to 7 days post infection, the type of infection changes from cytolytic
infection of primarily B-lymphocytes to latent infection of predominantly T-
lymphocytes. In cells latently infected with MDV, VIA are absent, Marek's
associated tumor surface antigen (MATSA: activated T-Cell marker) is present, and
virus can be recovered only after co-cultivation with permissive cells (McCall et al,
1987; Powell et al, 1974). Evidence for antibody and cell-mediated immune
response is found coincident with the switch from predominately cytolytic to
predominantly latent MDV infection (Sharma, 1985). About 9 days post infection,
transient immunosuppression occurs, but after a 2 week recovery period,
permanent immunosuppression ensues (Payne, 1985). Latently infected T-
Iymphocytes are disseminated via the Circulatory system and localized cytolytic
lesions are present in various organ and tissue systems during the
immunosuppression recovery period (9 days to 3 weeks) (Payne, 1985). Infiltration
of macrophages and lymphocytes in peripheral nerves ultimately leads to
demyelination and impairment of nerve function. Only in feather follicle epithelium
does the cytolytic infection result in release of cell-free virions (Witter, 1970). The
final outcome of acute MDV infection is transformation of T-lymphocytes and
infiltration of transformed cells into visceral organs, skin, muscle and nerves,
establishing tumorous masses which disrupt organ function. Severity and tissue
distribution of tumors of acute MD depends upon MDV strain and genetic
susceptibility, age, sex, and immune status of the host chicken (Payne, 1985).
Infection of turkeys with MDV also results in lymphoproliferative lesions, with

similar pathology as chickens. However, tumor cells are predominately of the B-

 

29

cell origin (Nazerian and Silva, 1988) and only some strains of serotype 1 MDV are

able to establish infection (Elmubarak, 1981).

2. Biology of MDV.

The herpesvirus responsible for development of MD was initially isolated in
chick kidney cell cultures (Churchill and Biggs, 1967). They described a highly
cell-associated herpes-like virus particle in cells displaying cytopathic effects (CPE:
cellrounding and syncytia formation). Evidence from electron microscopy studies
by Nazerian and Bunnester (1968) showed icosahedral, herpes-like particles 95-
100 um in diameter within the nucleus. However, enveloped viruses within the
cytoplasm were found at low frequency and not detected in culture media. The
cell-associated nature of MDV in vitra is most likely related to the absence of
enveloped virions within cells or growth fluid.

MDV is classified into 3 serotypes based primarily an antigenic similarities
between virus isolatesand ability to form lymphoid tumors in susceptible Chickens.
Serotype 1 MDV strains are oncogenic whereas serotypes 2 and 3 are
non-oncogenic (Payne-review, 1982). Serotype 3 MDV or herpesvirus of turkeys
(HVT) was originally isolated as an innocuous virus from turkeys (Witter, 1970).
Serotypes 2, 3 and attenuated serotype 1 strains have been used as effective
vaccines against oncogenic serotype 1 MDV (Payne, 1982; Witter, 1983).
Nucleotide sequence analysis and comparison of banding patterns generated by
restriction endonuclease digestions of MDV DNA indicates that MDV genome

structure differs among serotypes and within each serotype (Coussens et al., 1989;

30
Hirai et al., 1979; 1981; 1984; lgarashi et al., 1987; Ross et al., 1984; Silva and

Witter, 1985).

Serotype 1 strains are further subdivided based upon severity of disease
they cause. Vegy highly oncogenic strai_ns_ (very virulent: v.v.) are oncogenic in
genetically resistant and HVT- vaccinated birds. This group also causes acute
lethal cytolytic infections in MD susceptible birds. Highly oncogenic m (acute)
cause a high incidence of visceral and neural tumors in genetically susceptible but
low incidence of tumors in genetically resistant birds. Moderately to mildly
oncogenic strains (classic) induce low incidence of mainly neural MD in susceptible
birds. Minimally oncogenic strains cause minimal lesions only in very susceptible
birds. Serotype 1 strains of varying pathotype are not distinguishable using
conventional or monoclonal antibody methods. Serotype 1 pathotypes have been
distinguished by their ability to induce immunosuppression: highly oncogenic
strains are more immunosuppressive than less oncogenic strains (Witter et al,
1980). In addition, cell lines established from less oncogenic strains contain more
viral antigen than cell lines derived from highly oncogenic strains (Review- Schat,
1985)

The tightly cell-associated in vitra growth characteristics of MDV (Schat,
1985) have hampered progress on the molecular analysis of MDV genomes.
Separation of MDV DNA from host cell DNA by density-gradient centrifugation
results in extremely low yields of pure MDV DNA (0.1 to 10 ug MDV DNA per 10"
infected cells). Often, these preparations are contaminated with host cell DNA

(Kato and Hirai, 1985), as buoyant density of MDV DNA (1.705 g/cm’) (Lee et al.,

31
1971) is similar to that of host cell genomic DNA (1.700 g/cm"). In addition,

physical integrity of MDV genomes is often compromised during conventional
isolation procedures. The cloning and restriction mapping of MDV and HVT DNA
(Fukuchi el at, 1987: lgrashi et al, 1988) have greatly enhanced our ability to
analyze the molecular biology of MDV and HVT. The BamHl restriction maps of
MDV and HVT are presently used as standard references for location of genes on
the MDV or HVT genomes (Figure 3). In addition, adaptation of pulsed field gel
electrophoresis technology to the study of MDV has facilitated detection of novel
Changes in genome structure of various MDV isolates (lsfort et al, 1990; This
Dissertation Ch. 3).

3. MDV genome structure.

Total size of the MDV genome, derived from contour mapping electron
microscopy or sedimentation values of MDV DNA, was determined to be 180 kb
for serotype 1 MDV and 160 kb for HVT (Cebrian et al, 1982; Hirai et al., 1979;
Lee et al., 1971). Structure of the MDV genome was elucidated using electron
microscopy of partially denatured DNA (Cebrian et al .1982). These studies
revealed a structure closely related to HSV I (type E genome structure), containing
a long unique region (UL), and a short unique region (Us). Both UL and Us are
flanked by terminal repeat (T RL and TRs) and internal inverted repeat sequences
(IRL and ms) (Figure 2). Cebrian et al (1982) also compared the unit length of
MDV DNA regions with that of HSV and HVT. In all regions, MDV DNA was larger
than HSV or HVT DNA, with the exception of the Us. MDV Us is 3 kb smaller than

the Us of HSV, but 3 kb larger than the Us of HVT. The MDV TRL is 7.6 kb and

32
6.1 kb larger than HVT and HSV, respectively. The MDV UL is also 7.6 kb larger

than UL of HVT or HSV. In addition, the MDV TRs is 1.5 kb and 6.1 kb larger than
the TRs of HVT and HSV, respectively (Figure 2). Evidence for gene co-Iinearity
on the MDV genome compared to a-herpesvirus genes was supplied by
Buckmaster et al. (1988). Using random sequence analysis, Buckmaster et al.
localized 35 putative VZV homolog genes in a co-linear array on the MDV genome.
In addition, the entire Us region was completely sequenced by Brunovskis and
Velicer (1992) and they have determined that MDV Us genes which are
homologous to HSV genes are arranged in a colinear fashion with the HSV
genome. Brunovskis and Velicer also identified 3 open reading frames (SORF 1-4)

which are unique to MDV and show no homology to existing herpesvirus

 

Baum 3958 0.88% >92 P 8296.... 2 moufi EcEmom 0.89.8 >w... can ._.>_.. .o comcmano

 

a 2.6...
9. to- 9. a.a.. 9. ms- 9. 5. >9.
9. m. 7 9. Rm- 9. ms- 9. ms- 5...
HUI—“HI I >92
mm» m: a... .m. .2 .5.
09 m2. amp m9 our mop om mm om me 9. o

 

 

34

sequences in current databases. They also identified a fowlpox virus homolog
(SORF 2) (Brunovskis and Velicer, 1992). Total size of the 3 SORFs and the
fowlpox homolog nearly equals the difference in size of the MDV Us compared to
the HVT Us. Proteins encoded within these SORFs may be important in the
Characteristic pathological differences between MDV and HVT.

4. MDV Proteins: general Information.

There are at least six antigenically active viral proteins found in all three
serotypes of MDV (lkuta et al., 1983; Silva and Lee, 1984). The genes and protein
products of two of these antigens, MDV A (90) and B (98) have been extensively
Characterized. MDV 96 and 98 are homologous to herpes simplex virus-1 (HSV-
1) glycoproteins C and B, respectively (Binns and Ross, 1989; Buckmaster et al.,
1988; Chen and Velicer, 1991b; Coussens et al., 1990; Ross et al., 1989). In
HSV-1, 90 is non-essential for viral replication in vitra whereas gB, highly
conserved among alphaherpesviruses, is required for virus attachment to host cells
(Roizman, 1991). MDV 90, an abundant viral glycoprotein antigen, is found in
sera of infected birds, infected cell culture fluids, and to a lesser extent associated
with the plasma membrane of infected cells (lsfort et al., 1986). Evidence from
pulse-chase experiments and tunicamycin (inhibits protein glycosylation) treated
cells suggests that MDV 9C is synthesized on the endoplasmic reticulum and
processed through the Golgi via two pathways. Secretion of N-linked glycosylated
MDV 90 occurs through a major pathway while an unglycosylated form is secreted
through a minor pathway (lsfort et al., 1986). Approximately 95% of serotype 1

MDV go is secreted while the remaining 5% is associated with plasma membranes

35

of infected cells (lsfort et al., 1986). In comparison, MDV 9B is found primarily in
infected cell Iysates and the glycoprotein processing pathway is similar to that of
other herpesvirus gB homologues (Chen and Velicer, 1991b; Ross et al., 1989;
Velicer et al., 1978).

The gene for one MDV IE protein, MDV ICP4, has been localized to the
MDV genome (Anderson et al 1992). The most extensively characterized genes
are illustrated in Figure 3. Regions of the MDV genome implicated in playing a
role in MDV pathogenesis are all, with one exception (90), located within the IR
and TR, the most divergent region of the MDV genome as compared to or-

herpesvirus genomes. These genes will be discussed in the next section.

5. Molecular biology of MDV oncogenicity.

A variety of approaches have been utilized to identify regions of the MDV genome
which may encode genes important in MDV-induced disease processes. One
approach is comparison of proteins and gene sequences from non-oncogenic
serotypes 2 and 3 to oncogenic serotype 1. Another approach is analysis of
transcripts and their products in MDV transformed lymphoblastoid cell lines. Yet
another approach is comparison of proteins, genome structure, and transcription
patterns of oncogenic strains to their attenuated, non-oncogenic counterparts (the
subject of section 7). Serotype comparisons and MDV transformed cell lines will

be addressed in separate sections below.

 

.mEocoo >05. 2.. co mecca >92 3 5.88.. .n 2:2“.

 

 

 

 

Q...
I
a.E-.§.3=§9_3
we: can:
. _ _ . I .
.zz _ _ _ Soonefi s.\.
I, N a 33' \‘\\
X. o _ 2....
ll \\.\
I \
I \\
I \\
6 II iii
3 II \.\\
I .\\..
I 0.
” \\.\
llmll .. . I
am; 3 mm. .m. a .E.
2.0. or; On an
Iv IXIr Ivhv
p __ _ : _ b L__:: :_ _ __ _ _ b. _ _ : b : .
_ __ q 2 d _ «_~d: :4 _ __ _ d _ d A q : _ ._J
t .n. .{< .7... z v.2_z/ m i..:/m _/~v. n. o o._ 0 ~__.
\ 3...... «o m. kmm .O A.... a. «.3. o No
\ ll.
\ I’ll. —:Eam
.. 8:8 m:
\S III}!
S ll
AAA A A v v A A A v

mo .8 on 3E8 v... «m: numOm 2m: 5: «ED» {mow

37
a. Serotype comparisons.

96. Although 90 is antigenically related among serotypes, serotype 2 9C
is lower molecular weight than serotypes 1 or 3 and serotype 3 9C is more alkaline
than serotype 1 90 (vanZaane et al, 1982). Sequence data comparison of
serotype 1 and 3 90 genes indicates overall 66% and 81 % homologies at the DNA
and protein levels, respectively. However, the 5’ coding and flanking regions are
quite divergent (Coussens et al, 1990). Differences in the 5' regions between
serotypes 1 and 3 would suggest a possible role in different pathologies of
Serotype 1 and 3. Thus, it would be of interest to further investigate the divergent
5’ regions in relation to MDV pathogenesis.

pp38. By comparison of proteins among serotypes that are
immunoprecipitated with a panel of monoclonal antibodies, Silva and Lee (1984)
identified serotype 1 specific phosoproteins of 44, 38, and 24 kd. Subsequently,
the gene encoding the 38 kd phosphoprotein (pp38) was localized to the UleL
junction of the MDV genome (Chen and Velicer, 1992; Cui et al, 1991). pp38 is
found predominantly in the cytoplasm of infected cells and is not considered a
DNA binding protein (Cui et al, 1991). In addition, pp38 produces only non-
neutralizing antibodies in viva (Nazerian et al, 1992). Also, pp38 may be an IE or
E protein based upon its expression in the presence of phosphonoacetic acid
(Chen and Velicer, 1992). Data by Pratt et al (1992) suggests that pp38 may be
transactivated by the MDV ICP4. Because pp38 is expressed in MDV transformed
cell lines and MDV-induced tumor lesions in viva, pp38 is implicated as playing a

role in MDV-induced oncogenicity (Chen and Velicer, 1992; Cui et al, 1991).

. an-

 

 

38

However, pp38 is also expressed in cells lytically infected with oncogenic as well
as attenuated MDV strains (Chen and Velicer, 1992; Silva and Lee, 1984). In
addition, the original pp38 monoclonal antibody, H19, was generated from spleens
of mice that were injected with the attenuated Md11 strain Md11/750 (Lee et al,
1984). Thus, pp38 may not be the transforming protein of MDV, but may be
involved in a multi-step process leading to transformation. Clearly, further

investigation on the functional role of pp38 is required.

b. MDV transformed cell lines.

MDV tumor cell lines have been established by cultivation of MDV
lymphoma cells in vitra. Most cell lines characterized thus far are activated T-
lymphocytes (Schat et al, 1991). Virus can be rescued from most cell lines by co-
cultivation with susceptible primary cultures. After prolonged in vitra cultivation, the
ability to rescue virus is lost. These cell lines are Classified as non-producer cell
lines. Viral antigens are not detectable in non-producer cell lines. Producer cell
lines are defined as cells from which virus can be rescued after co-cultivation.
Producer lines are further subdivided as expression or non-expression based on
amount of viral antigen produced. 5-iodo-2-deoxyuridine (IUDR) treatment of non-
expression cell lines induces expression of viral antigen (Review-Schat, 1989).

MDNA. A 28 kd nuclear protein (MDNA) was found to bind to specific
sequences located in the TRs and Us regions of the MDV genome (Wen et al,
1988). MDNA was not expressed in lytically infected cells, but was expressed in

two producer cell lines (an expression and a non-expression cell line). Although

 

 

39
the authors conclude that MDNA is analogous to EBNA-1 of EBV (EBV nuclear

antigen that binds near the origin of replication in latently infected cells), two
important experiments are lacking. First, binding specificity was not conclusively
shown. Second, no evidence was given that MDNA is a viral-encoded protein. In
addition, it is questionable whether the cell lines used were truly latently infected.
Perhaps MDNA is a lymphocyte specific factor, expressed in activated
lymphocytes.

meq. Analysis of viral specific RNA expression in producer and non-
producer cell lines revealed abundantly expressed transcripts located within the IRL
of the MDV genome (Tillotson et al, 1988). In addition, these transcripts were not
expressed in lytic infections of attenuated MDV strain Md11/75C but were
expressed at low levels in virulent MDV strain JM infected cells (T illotson at al,
1988; Jones et al, 1992). The IRL transcripts were sub-localized to the EcoRI Q
fragment (spans the BamHl-I2 and BamHl-Q2 fragments). Sequence analysis of
the EcoRI O fragment revealed a putative 40 kd protein containing a leucine zipper
repeat and an upstream domain rich in basic amino acids, suggesting a protein
similar to the fos/jun family of transcriptional activators. Antiserum raised against
a synthetic peptide corresponding to the leucine zipper region was used to detect
a 40 Rd protein in MDV transformed cell lines, but not lytically infected MDV strain
GA cells (Jones et al, 1992). Further work is required to determine if meq
(Marek's EcoRl O) is a direct acting oncogene, as its physical structure would
suggest.

Transcriptional activity In MDV transformed cell lines. Because MDV

 

40

cell lines are established directly from tumor cells , transcriptional activity in MDV
transformed cell lines is of interest. Three groups have investigated transcription
patterns in various MDV transformed cell lines. Maray et al (1988) found
transcriptional activity dispersed throughout various regions of the MDV genome
in MSB-1 cells (a producer, expression line). Schat et al (1989) compared
transcriptional activity in a variety of MDV transformed cell lines. This group also
found transcriptional activity dispersed throughout various regions of the MDV
genome in an expression (producer) cell line, but non-producer and non-
expression cell lines expressed only a limited number of immediate early
transcripts from discrete genomic regions. These IE transcripts are clustered in
the IRL (BamHI-Is and -L fragments), le (BamHl-A fragment) and the unique short
(non-expression only, BamHI-P1). After treatment with IUDR, transcripts which
hybridize to a BamHl-H probe (also in IRL) were detected in non-producer cell
lines. Similar treatment with IUDR of non-expression cell lines yielded transcripts
located in BamHl-H as well as other transcripts dispersed throughout the MDV
genome. In addition, a 0.6 kb IE transcript located within BamHI-L is truncated by
0.3 kb in the non-producer cell line.

Sugaya et al (1990) also investigated transcription patterns in MDV
transformed cell lines. Using a non-expression cell line, they found transcripts in
the same regions as Schat et al (1989). However, in this particular cell line, they
detected additional transcripts located within the BamHI-H region. The presence
of MDV transcripts originating from a limited region of the MDV genome is

reminiscent of transcription patterns in transformed cells latently infected with EBV

41

(Schat et al, 1989). Comparison of transcription patterns in various cell lines
suggests that MDV infection is frozen in different stages of latency. Thus, non-
producer cell lines may be considered latently infected, containing an aberration
in MDVs ability to reactivate. Whereas, MDV in expression and non-expression
cell lines may be frozen in earlier stages of latency. Gene products from the IR._
and IRS may play a role in the initiation and maintenance of MDV transformation,

and/or initiation, maintenance, and reactivation from latency.

6. MDV vacclnes.

Prior to the development of vaccines against MDV, 20 to 30% mortality
rates in commercial flocks were common (Pattison, 1985). In 1970, vaccination
against MDV using HVT was introduced (Witter, 1970). During the 12 year
period following HVT vaccination, from 1970 to 1982, national condemnation of
young chickens due to MD declined by 95% (PurChase, 1985). Total cost to the
US. poultry industry for vaccination is $44.4 million dollars/year. Although MD
vaccines have been effective at reducing mortality and condemnation rates, there
continues to be occasional outbreaks or vaccine breaks, which contributes further
cost to the poultry industry. Economic loss in the US. due to mortality,
condemnations, and loss of egg production is $115.9 million year. In 1984, total
worldwide cost of MD was estimated at $943 million a year. Thus, control of MD
is of economic importance to the poultry industry (Purchase, 1985).

The most widely used MDV vaccines in the us. and other countries are

cell-associated and cell-free HVT preparations (Pattison, 1985). HVT produces no

 

 

42
clinical symptoms in chickens or turkeys (Payne, 1985). In turkeys, HVT is highly

contagious. ln chickens, however, spread of HVT by contact is minimal (Payne,
1985). Antigenic cross-reactivity of HVT and MDV antigens prompted its
development as a MD vaccine (Pattison, 1985). Attenuated mildly virulent
serotype 1 vaccines have been widely used in The Netherlands (Witter, 1992a).
This type of vaccine readily spreads by contact, but spread does not occur rapidly
enough to provide passive immunization of hatchmates (Witter, 1992b).
Attenuated mildly virulent serotype 1 vaccines have not been used extensively due
to their residual pathogenicity. Serotype 2 vaccines have also been used as
vaccines against MD (Witter, 1992a). The serotype 2 strain in current use, SB—1,
spreads readily by contact and is efficacious against most virulent MD strains
(Review- Witter, 1985). Recombinant MDV vaccines are presently being
developed and two vaccine candidates are worth noting. First, a fowlpox vaccine
expressing MDV 98 was 100% effective against challenge with GA and M05
(virulent and v.v. MDV strains, respectively), but only 90% protective against
challenge with v.v. MDV strain RBIB (Nazerian et al, 1992). Second, a bivalent
recombinant vaccine against MDV and Newcastle disease virus (NDV) was
developed by Morgan et al. (1992). These HVT recombinants expressing two
vimlence determinants of NDV gave 100% protection against NDV. However the
vaccine was only 64-76% protective against challenge with v.v. MDV strain RBIB
(Morgan et al, 1992).

Since 1988, the application of a bivalent vaccine, a combination of 88-1 and

HVT, has increased in the US (Witter,1992). The bivalent vaccine efficacy rivals

43

or exceeds that of HVT alone (Pruthi et al, 1989). Although this vaccine is
extremely efficacious, concern was raised when Bacon et al. (1989) reported that
SB-1 could augment lymphoid Ieukosis (avian retrovirus induced disease) in white
leghom chickens. The mechanism behind this possible interaction between MDV
and retroviruses was suggested to occur via MDV transactivator proteins (T ieber
et al., 1990). 88-1 can enhance retroviral replication in vitro and retroviral
transformed B-cells can harbor SB-1 DNA after co-infections in vivo (Pulaski et al.,
1992; Fynan et al, 1992). In addition, retroviral integration into the MDV genome
has been shown to occur in vitro (lsfort et al., 1992). Presently, there is little
concern for these interactions occurring under field conditions due to the low level
of lymphoid Ieukosis in commercial flocks (Witter, 1992a). However, published
reports on the interaction of two viruses producing an undesirable outcome in vitro
and in vivo are extensive. Possible interference with vaccine efficacy is illustrated
by reports on PRV. Recombination of two PRV vaccine strains during replication
in swine resulted in vimses with restored virulence (Katz et al., 1990; Henderson
et al, 1991).

During the course of vaccination against Marek’s Disease, some flocks
experience the emergence of MD characterized by a more severe pathotype.
These v.v. MDV strains are associated with, and considered to be the cause of,
MD outbreaks in vaccinated flocks (\Mtter, 1992a). During the first quarter of
1992, Delaware experienced a 75% increase in incidence of MD over the previous
year. Assuming vaccination and management practices have not changed, this

increase may be due to the emergence of v.v. MDV.

 

44

The increasing virulence of v.v. MDV has prompted attempts to develop
vaccines against v.v. MDV (Witter, 1992). These attempts have met with limited
success. Residual pathogenicity is associated with attenuated, mildly virulent,
serotype 1 vaccines. One vaccine, CV1988/Rispens, was found to be the most
effective vaccine against v.v. MDV with efficacy of only 80-85% (Witter, 1992b).
CV1988/Rispens has been used with success in many countries, but is not
currently licensed in the US.

The possibility of providing better protection against v.v. MDV strains
prompted development of an attenuated vaccine using v.v. MDV strain Md11
(Witter et al, 1982). MDV strain Md11 was originally isolated from an HVT-
vaccinated commercial flock experiencing vaccine breaks (Witter, 1983). Md11 is
characterized by inducing severe early bursa and thymus atrophy, lymphoid cell
necrosis, and early death in the absence of lymphomas or nerve enlargement
(Payne, 1985). The Md11/750 vaccine provided 69% protection in chickens that
did not have homologous maternal antibodies. However, vaccine protection in
chickens with homologous maternal antibodies was reduced to 14% (Witter, 1984).
A trivalent test vaccine containing HVT, 88-1, and Md11/75C was 100% effective
against challenge with vimlent and v.v. MDV (Witter, 1984). Practical
considerations on use of this trivalent vaccine, such as growth and preparation of
trivalent vaccine mixtures and the residual pathogenicity associated with the
Md11/75C component, have most likely prevented its use by the poultry industry.

Due to the increasing prevalence of v.v. MDVs isolated from vaccinated

flocks, development of vaccines against v.v. MDV is a priority concern (Witter,

 

45

1992a). However, basic understanding of the attenuation process, and genes
involved in virulence would aid in designing effective strategies for vaccine
development. Different virus strains become attenuated at different rates or
passages (Schat, 1985b). MDV strain Md11 is only partially attenuated at passage
52, further passage reduces in vivo lesions, but attenuated Md11 has an
associated residual pathogenicity of approximately 2% (Ch. 4, this dissertation).
This may be due to the presence of oncogenic sub-populations. The continued

and increasing frequency of vaccine breaks makes understanding virus attenuation

a priority.

7. Attenuation of MDV.

Attenuation of serotype 1 MDV occurs after 30 - 55 serial in vitro passages
(Schat et al., 1985b). Coincident with loss of oncogenicity in viva, attenuated MDV
fails to induce early cytolytic infection, have reduced ability to infect or replicate in
lymphocytes, and lose the ability to spread by contact, as compared to oncogenic
MDV (Schat et al., 1985; Witter,1984). in vitro studies indicate that attenuated
MDV grow faster in cell culture, are able to infect a wider variety of cultured cells,
display reduced expression of glycoprotein C expression, and contain altered
genome sthcture and sizes compared to oncogenic MDV (Nazerian, 1979; Schat,
1985; Review— Ross, 1985; Wilson and Coussens, 1991).

a. Genome structure.
Comparison of genome structure of attenuated isolates to their oncogenic

counterparts revealed gross changes in limited regions of the MDV genome (Hirai

46
et al., 1979; 1981; 1984; lgarashi et al., 1987; Ross et al., 1984; Silva and Witter,

1985). These changes were attributed to expansion, by tandem reiteration, of a
132 bp repeat region located within the TRL and IR, (BamHI-D and -H fragments)
of the serotype 1 MDV genome (Maotani et al, 1986). Although a protein product
encoded within this region has yet to be identified, transcripts encompassing the
132 bp repeat region have been intensely studied. Comparison of transcripts
located near the 132 bp repeats in oncogenic and attenuated MDV indicated that
a 1.8 kb transcript family becomes tmncated as a result of in vitro attenuation
(Bradley et al, 1989). Nucleotide sequence analysis and generation of cDNAs
from the BamHl-H fragment by several investigators have yielded diverse results.
Although exact location of transcripts differ among investigators, it is apparent that
a family of spliced and unspliced RNAs can be detected (Bradley et al, 1989;,
Chen and Velicer, 1991; Peng et al, 1992; lwata et al, 1992). Predicted amino
acid sequence homologies with various regions of the BamHl-H region to gag
(retrovirus antigen), avian fps (kinase related transforming protein), v-fms
(epidermal growth factor-like oncogene), mouse TLM oncogene, and fes/fps
oncogenes have been found (Bradley et al, 1989; Peng et al, 1992; lwata et al,
1992). In addition, transcripts are initiated at multiple sites in a bi-directional
manner (Chen and Velicer, 1991). In some transcripts, the 132 bp repeats are
contained within introns (Peng et al, 1992). Interestingly, the 132 bp repeats are
located downstream of a bi-directional promoter containing a putative origin of
replication, CAAT and Sp1 binding site motifs (Bradley et al, 1989). Kawamura et

al (1991) have shown that an antisense oligonucleotide complementary to a splice

47

donor sequence in one of the BamHl-H transcripts will inhibit proliferation of an
MDV transformed cell line (producer, expression line). Although the BamHl-H
region continues to be the subject of intensive investigation, functional significance
of this region remains to be elucidated.

b. 90 expresslon.

In all MDV serotypes tested, expression of 9C is greatly reduced after 30-50
serial passages in vitro (Bulow and Biggs, 1975; Churchill et al., 1969; lkuta et al.,
1983a; lkuta et al., 1983b; Nazerian, 1980; Purchase at al., 1971; Silva and Lee,
1984). Coincident with reduced gC expression, MDV serotypes 2 and 3 lose their
vaccine efficacy and serotype 1 becomes attenuated with respect to oncogenicity
(Calnek and Witter, 1991; Churchill et al., 1969; Kato and Hirai, 1985; Silva and
Witter, 1985). M DV gB expression does not change significantly after serial in vitro
passage (Silva and Lee, 1984).

Despite an apparent correlation between MDV gC expression and MDV
oncogenicity, isolation of naturally occurring non-oncogenic (serotype 2) MDV
which express 90 and clonal isolates of oncogenic (serotype 1) strains which do
not express gC (Bulow and Biggs, 1975a; Bulow and Biggs, 1975b; Glaubiger et
al, 1983; Purchase et al., 1971), suggests that gC does not directly affect the
oncogenic potential of MDV. However, in vivo data indicates that attenuation of
MDV reduces the efficiency of lymphocyte infection (Schat et al., 1985).
Therefore, MDV 90 may play a role in lymphocyte targeting. Regulation of MDV
gC may have important implications for vims-host interactions, and have important

consequences in the process of viral attenuation.

48
Although the gC-negative phenotype of attenuated MDV has been

documented for more than 20 years, the molecular basis for this alteration has
remained obscure. The inability of attenuated MDV to produce 90 may be due to
mutations or deletions of the MDV 90 gene or promoter. Alternatively, decreased
90 expression may result from mutations in MDV encoded or induced trans-acting
factors which interact with the QC promoter. Analysis of the mechanism for
reduced expression of 9C is presented in this dissertation (Ch. 2).

c. Attenuation of v.v. MDV.

Due to the emergence of v.v. MDV in vaccinated flocks, and limited
success with developing effective vaccines against v.v. MDV, analysis of the
attenuation process of v.v. MDV strains is of interest. One such v.v. MDV, strain
Md11, is characterized by causing acute cytolytic infection leading to early mortality
(Witter, 1984). After 57 serial in vitro passages, v.v. MDV strain Md11 became
attenuated with respect to oncogenicity. However, microscopic nerve lesions were
detectable (Witter, 1982). After passage 75, nerve lesions were not detected.
Thus, complete attenuation was not achieved until passage 75. Also, attenuated
Md1 1 p75 was shown to induce considerably less interferon in chick embryos than
that of oncogenic strains (Sharma, 1989). Decreased expression of glycoprotein
C in attenuated Md11 has also been established (Witter and Lee, 1984) and the
molecular basis of reduced 90 expression was investigated in detail (Wilson et al.,
submitted; this Dissertation Ch. 2). In addition, Tieber et al. (1989) have
determined that attenuated Md11 is less able to transactivate an exogenous RSV

promoter in vitro as compared to oncogenic Md11. Genome stmcture studies

49

indicate that amplification of the 132 bp repeats also occurs during attenuation of
Md11 (Silva and Witter, 1985). Recently, our laboratory has identified an
additional change in the genome of attenuated Md11(Wilson and Coussens, 1991;
this Dissertation Ch. 3). Using cell-free MDV genomic probes, we identified a 200
bp deletion, also located within the IRL and TRL of the attenuated Md11 genome.
This change, however is separate and distinct from the 132 bp amplified region.
132 bp amplification occurs in the BamHl-D and -H fragments (Silva and Witter,
1985), and the 200 bp deletion was mapped to the BamHI-L fragment (Wilson and
Coussens, 1991). Further characterization of the 200 bp deletion in BamHl-L is

presented in Chapter 4 of this Dissertation.

The subject of this dissertation is analysis of the attenuation mechanisms
of MDV. Research focus was directed toward two observations of attenuated MDV
isolates: genome structure alterations and reduced expression of a major
glycoprotein antigen. During preliminary investigations, it became clear that
development of techniques to isolate cell-free MDV genomes would aid in the
analysis of attenuated MDV isolates. Development of these techniques and
characterization of attenuated MDV genomes is the subject of chapter 2. During
the course of developing these techniques, novel changes in the genome of
attenuated isolates of the v. v. MDV strain Md11 were identified and further
investigated as described in chapter 3. The attenuation of this particular virus
strain is of interest because of its highly oncogenic nature and potential for vaccine

development of its attenuated counterpart. Thus, molecular analysis of the

 

50

reduced expression of a major glycoprotein antigen in v. v. MDV strain Md11 is
presented in chapter 4. Attenuation of MDV as it relates to oncogenicity and

vaccine development will be discussed in chapter 5.

 

Chapter 2.

PURIFICATION AND CHARACTERIZATION OF INFECTIOUS MAREK'S
DISEASE VIRUS GENOMES USING PULSED FIELD ELECTROPHORESIS.

Melinda R. Wilson and Paul M. Coussens

Molecular Virology Laboratory, Department of Animal Science
Michigan State University
East Lansing, Michigan 48824

Virology. 185(2):673-680

51

if!

G.

1h
«in

)

fr.
.I ,{

ABSTRACT

Marek’s Disease virus (MDV) is an acutely oncogenic avian herpesvirus.
The tightly cell-associated in vitro growth characteristics of MDV present unique
problems when attempting to purify, analyze, and manipulate MDV genomes. To
facilitate molecular characterization of MDV, contour-clamped homogeneous
electric fields electrophoresis (CHEF) was used to purify infectious M DV genomes.
CHEF techniques were optimized for evaluation of total genome size and
alterations in structure which occur during in vitro attenuation of oncogenic MDV.
Our results indicated that genomes of attenuated serotype 1 MDV strain JM may
contain deletions totaling 15 kbp while high passage serotype 2 non-oncogenic
MDV strains 88-1 and 281Ml/1 were 5 kbp and 3 kbp larger, respectively, than
their low passage counterparts. Using cell-free CHEF purified MDV genomes as
hybridization probes, we identified a 200 bp deletion in attenuated genomes of the
very virulent MDV strain Md11. Presently, it is unclear if this 200 bp deletion is
related to mutations which lead to loss of oncogenicity or pathogenicity in Md11.
This study is the first report which describes procedures for purification of
infectious herpesvirus genomes from pulsed field gels. Our results demonstrate
that pulsed field purified viral DNA will facilitate molecular characterization of M DV

and other cell-associated herpesviruses.

52

I. INTRODUCTION.

Marek’s Disease virus (MDV), an acutely oncogenic gammaherpesvirus,
causes lymphoproliferation and nerve demyelination in domestic chickens
(Payne-review, 1982). MDV is classified into 3 serotypes based primarily on
antigenic similarities between virus isolates and ability to form lymphoid tumors in
susceptible chickens. Serotype 1 MDV strains are oncogenic whereas serotypes
2 and 3 are non-oncogenic (Payne- review, 1982).

The MDV genome, co-Iinear with alphaherpesvirus genomes, consists of
unique long (UL) and unique short (Us) regions flanked by terminal repeats (T RL,
TRs) and internal repeats (IRL, IRS) (Buckmaster et al, 1988; Cebrian et al., 1982).
Nucleotide sequence analysis and comparison of banding patterns generated by
restriction endonuclease digestions of MDV DNA indicates that MDV genome
structure differs among serotypes, within each serotype, and after serial in vitro
passage of MDV (Coussens et al., 1989; Hirai et al., 1979; 1981; 1984; lgarashi
et al., 1987; Ross et al., 1984; Silva and Witter, 1985). A prominent change in
attenuated serotype 1 MDV strains is amplification by tandem reiterations of a 132
bp sequence located in the IRL of MDV genomes (Maotani et al., 1986; Bradley et
al., 1989; Chen and Velicer, 1991).

Progress on the molecular analysis of MDV genomes is hampered by the
tightly cell-associated in vitro growth characteristic of MDV (Schat, 1985).
Separation of MDV DNA from host cell DNA by density-gradient centrifugation
results in extremely low yields of pure MDV DNA (0.1 to 10 ug MDV DNA per 10°

infected cells). Often, these preparations are contaminated with host cell DNA

53

I'I

f K .r I . mm [mm by
Infl. .MV int NW .II he C If r. . «r5

{:1
VI.

u: 5 .V

54
(Kato and Hirai, 1985), as buoyant density of MDV DNA (1.705 g/cm3) (Lee et al.,

1971) is similar to that of host cell genomic DNA (1.700 g/cm"). In addition,
physical integrity of MDV genomes is often compromised during conventional
isolation procedures. Thus, development of alternative procedures to obtain
purified, intact, and infectious MDV DNA is of critical importance to the study of
MDV molecular biology.

In 1984, pulsed field electrophoresis (PFE) technology was introduced to
analyze large DNA molecules (Schwartz and Cantor, 1984). Separation of large
viral DNA genomes from host cell chromosomes by PFE has been used for
Chlorella vims (Rhozinski et al., 1989), poxvirus (Bostock, 1988), Epstein-Barr
virus (Harris and Bentley, 1988), and more recently for MDV (lsfort et al., 1990).
Although these viral DNAs were separated from cellular DNA using PFE, infectivity
of intact viral genomes isolated from pulsed field gels has not been demonstrated
in any viral system.

In this report, we present procedures developed to recover infectious
cell-free MDV genomes using contour-clamped homogeneous electric fields
(CHEF; Chu et al., 1986) electrophoresis. MDV DNA, separated from chicken
chromosomes by PFE, was given the designation "cell-free" MDV genomes.
CHEF was used to investigate genome structure differences among and within
MDV serotypes, and after serial in vitro passage of various MDV strains. Our
results indicate that oncogenic serotype 1 cell- free MDV genomes are 17-18 Kbp
larger than non-oncogenic serotype 2 and 3 cell-free genomes and that the

genome of attenuated serotype 1 MDV strain JM is 10 Kbp smaller than its

55

oncogenic counterpart. Using full length, cell-free MDV genomes as hybridization
probes, we have identified a 200 bp deletion in attenuated genomes of the very
virulent (v.v.) MDV strain Md11. Techniques and results presented in this report
will contribute to our understanding of MDV pathogenicity and are directly

applicable to studies of other large genome viruses.

II. MATERIALS AND METHODS
Cells and Viruses.

Preparation, maintenance, and infection of duck and chick embryo
fibroblasts (DEF and CEF, respectively) were performed as described (Glaubiger
at al., 1983). MDV strains GA, JM, Md11, SB-1,281Ml/1 and HVT strain F0126,
have also been described (Coussens et al., 1989; Tieber et al., 1990; Carter and
Silva, 1990). Low passage (GAp5, HVTp5, Md11 p10, JMp14, SB-1p20, and
281Ml/1p15) and high passage (Md11p80, JMp213, SB-1p100, and 281 MI/1 p96)
MDV isolates were kindly provided by the USDA-Avian Diseases and Oncology

Laboratory, East Lansing, Michigan, USA.

Preparation of MDV infected cells for CHEF.

MDV infections of CEF and DEF were allowed to progress until cell
monolayers appeared fully infected and cells began to detach from the substratum
(5-7 days). Infected cells were washed from culture plates in phosphate buffered
saline (Sambrook et al., 1989), pelleted, and resuspended in L Buffer (0.1 M
EDTA, 0.01 M Tris-HCI pH 7.6, 0.02 M NaCl) (Sambrook et al., 1989) at an

nah-:3
h

56

approximate concentration of 1x107 cells per ml. Cells in L Buffer were warmed
to 42°C and an equal volume of 42°C 1.2% low melting point (LMP) agarose
(Bethesda Research Laboratories, Gaithersburg, MD.) was mixed with cell
suspensions. Cell/agarose mixtures were pipetted into sample plug molds
(Bio-Rad, Inc., Richmond, CA.) (250 Lil per plug) and allowed to harden for 5 min
at -20°C or 15 min at 4°C. Agarose embedded cells were lysed for 48 hr in 2
changes of L buffer (3 agarose plug volumes each) containing 1% (v/v) Sarkosyl
and 0.1% (w/v) proteinase K at 50°C. Following Iysis, agarose plugs were washed
three times for 15 min each in TE (10 mM tris-HCI, pH 8.0, 1 mM EDTA) at 50°C
and stored in 0.5 M EDTA at 4°C until used. Agarose blocks were stored in this

manner for six months with minimal degradation of DNA.

Pulsed Field Electrophoresis.

Lysed cells were analyzed by CHEF using the hexagonal electrode array
developed by Chu et al. (1986). Agarose and LMP agarose gels were prepared
in 0.5x Tris-Borate-EDTA buffer (TBE) (Sambrook et al. 1989). Prior to
electrophoresis, excess EDTA in agarose plugs was removed by soaking plugs in
10 volumes of TE, ph 8.0 at 22°C for 20 min. Agarose plugs were placed in
agarose gel sample wells and sealed with 1% LMP agarose. Yeast chromosomes
(Bio-Rad, Inc., Richmond, CA.) or concatermized lambda DNA (Sambrook et al.,
1989) were used as DNA size markers during electrophoresis.

Electrophoresis was carried out in a CHEF-Dry II system (Bio-Rad, Inc.,

Richmond, CA.) gel chamber at 16°C in recirculating 0.5x TBE. Gels were stained

57
for 30 min in 0.5 ug per ml ethidium bromide (EtBr) in 0.5x TBE followed by

destaining for 30 min in the same buffer. DNA bands were visualized and
photographed with a UV light source. Sizes of DNA bands were determined from
standard curves based on either yeast chromosome or lambda DNA concatemer

standards.

Preparation of cloned MDV fragment and cell-free MDV genomic probes.

To prepare full length MDV genomic probes, gel slices containing cell-free
MDV genomes were excised from LMP agarose CHEF gels and DNA was isolated
by melting LMP agarose followed by phenol extractions as described (Sambrook
et al., 1989). Concentrations of viral DNA recovered from CHEF gels were
determined spectrophotometrically. Yields of cell-free MDV DNA varied from 130
ng for strain GAp7 to 1 ug for strain SB-1p101 per 10° infected cells.

Cell-free MDV genomes were labeled with [alpha-”P1dCTP using a random
primed labeling kit (Boehringer Mannheim Biochemicals, Indianapolis, IND.) as per
manufacturer’s specifications. Due to probe complexity, incubation times for
labeling reactions were one hour. Excess unincorporated nucleotides were
removed using Sephadex G-50 spin columns as described (Sambrook et al.,
1989). MDV fragment specific probes were prepared from a MDV strain GA DNA
clone library (a generous gift of M. Nonoyama, Showa University Research
Institute for Biomedicine in Florida, St. Petersberg, FL). BamHl-A, -H, -l2, -L, and
-P, MDV insert DNA fragments were gel purified and labeled with [or-32P1dCTP as
described above.

 

r“

t)

I F

453

58

Southern Blots.

DNA in EtBr stained agarose gels was nicked by irradiation with UV light for
3-5 min, transferred to nylon membranes (Zeta- Probe, Bio-Rad, Inc., Richmond,
CA.) under alkaline conditions (0.4 N NaOH), and hybridized to radiolabeled DNA
probes as described (Budowle and Baechtel, 1990). CHEF gels containing DNA
in the megabase size range were allowed to transfer for 24 hours. Conventional
agarose gels containing DNA in the 50 to 0.1 kilobase pair (kbp) size range were

transferred for 16 hours.

Transfection of CHEF-purified MDV DNA.

CHEF-purified, cell-free MDV genomes were introduced into CEF cells by
electroporation (Neumann et al., 1982). Typically, 50-75 ng of CHEF-purified MDV
DNA were used per 60 mm culture dish. Primary CEF cells were trypsinized,
counted and pelleted. Cells were resuspended in Leibovitz-McCoy media
supplemented with 0.3 M sucrose at a concentration of 5x10° cells per 0.8 ml.,
transferred to electroporation cuvettes (Bio-Rad, Inc., Richmond, CA), and warmed
to 37° for 5 min.

LMP agarose slices containing cell-free MDV genomes were melted at 65°,
cooled to 37°, mixed with cell suspensions, and incubated at 37° for 10 min.
Electroporation was carried out using a Gene Pulsar and capacitance extender
(Bio-Rad, Inc., Richmond, CA.). Different combinations of voltage and capacitance
settings were evaluated. In one set of experiments, voltage was held constant at

450v while capacitance settings were varied (125, 250, 500, and 960 uFD). In

59

another set of experiments, capacitance was held constant at 960 uFD and
voltages of 150v, 250v, 350v, and 450v were used. Following electroporation,
samples were incubated for 10 min at 37° and transferred to 4 ml Leibovitz-McCoy
media containing 4% calf serum in 60 mm culture dishes. Electroporated cells
were maintained at 37°C in an atmosphere supplemented with 5% 002. Culture
media was changed to Leibovitz-McCoy supplemented with 1% calf serum, 48 hr

after electroporation.

III. RESULTS
Separation and Identification of MDV genomes and chicken chromosomes
using CHEF.

On EtBr stained CHEF gels containing samples from high and low passage
Md11 infected cells, four DNA bands of 2,100, 1,400, 660, and 180 Kbp were
visible (Figure 1). A broad smear of less than 180 kbp was observed in all lanes.
Lanes containing uninfected CEF displayed all EtBr stained bands found in
infected cell lanes except the 180 kbp band, suggesting this band was of viral
origin (Figure 1).

Southern transfer analysis was used to determine whether particular bands
in CHEF gels of MDV infected cells were of viral or cellular origin. Radiolabeled
cloned MDV DNA fragments hybridized primarily to the 180 kbp band (Figure 2,
A), indicating that this band was viral DNA. Hybridization of MDV probes to DNA
retained in sample walls was also evident (Figure 2, A), and addressed previously

(lsfort et al, 1990). In lanes containing Md11p80 DNA, an additional DNA band

60
of 125 kbp which hybridized to the MDV DNA probe was evident below the more

intense 180 kbp band (Figure 2, A). Hybridization of the 2,100, 1,400, and 660
kbp bands to CEF DNA probes was observed in uninfected CEF, Md11p10, and
Md11p80 (after longer exposure to X-ray film ) sample lanes (Figure 2, B),
suggesting that these bands were chicken chromosomes which begin to resolve
under CHEF conditions.

To distinguish between cell-associated MDV DNA retained in sample wells
and MDV DNA which migrated at 180 kbp on CHEF gels, we designated the 180

kbp band as "cell-free" MDV DNA.

Infectivity of cell-free CHEF purified MDV DNA.

Although analysis of CHEF gels indicated that a 180 kbp band was the full
length MDV genome, we were not certain this DNA band retained biological
activity. To assess infectivity and physical integrity of cell-free CHEF purified MDV
genomes, electroporation of CHEF-purified MDV DNA into CEF cells was
performed as described in Materials and Methods.

Typical MDV viral plaque formations were evident 6 days post
electroporation on CEF cells electroporated with 450v or 250v at 960uFD (Figure
3, A 3). Resultant viral plaques were purified and expanded 2 passages to obtain
sufficient quantities of DNA for further studies. Analysis of BamHl restriction
digests of viral DNA obtained from cells transfected with CHEF- purified MDV DNA
verified that viral plaque formation resulted from electroporation of CHEF purified

MDV DNA. No major genome rearrangements were visible in viral DNA of cells

61
transfected with CHEF purified MDV DNA (Figure 3, B).

Transfection efficiency (20 PFU/ug viral DNA) from electroporation of
cell-free MDV genomes was similar to transfection efficiencies of total DNA
isolated from MDV infected cells using the polybrene/DMSO procedure described
by Kawai and Nishizawa (1984). However, transfection efficiencies from
electroporation of cell-free MDV DNA were 5 times lower than that obtained by
electroporation of total DNA isolated from MDV infected cells (100 PFU/ug viral

DNA).

Detection of Genome Size Differences Among MDV Serotypes.

To determine if serotype classification could be correlated to cell-free MDV
genome size differences among MDV strains, and to evaluate use of CHEF as a
diagnostic aid to MDV serotyping, cells infected with MDV serotypes 1 (GAp7,
Md11p11, Md11p81, JMp15, and JMp214), 2 (SB-1p21, SB-1p101, 281MI/1p16,
and 281 MIN p97), and 3 (HVTp7 strain F0126) were analyzed on high resolution
CHEFgms

Faster migration of low passage MDV serotypes 2 and 3 cell-free genomes
indicated lower size genomes of SB-t, 281Ml/1, and HVT as compared to low
passage serotype 1 strains (T able 1). Results for serotypes 1 and 3 are consistent
with previously reported genome sizes derived from sedimentation values of MDV
DNA from serotype 1 (180 kbp) and 3 (167 kbp) (Hirai et al., 1979; Lee et al.,
1971). An estimate of MDV serotype 2 (281Ml/1) genome size, derived from the

sum of Neil restriction fragment sizes, is 161.8 kbp (D. Reilly, personal

 

62

communication), consistent with results from CHEF gels (Table 1).

To determine if repeated passage in culture resulted in altered mobility of
cell-free MDV genomes on CHEF gels, we compared low and high passage pairs
of serotype 1 and 2 samples on high resolution CHEF gels (Figure 4, Table 1).
High passage serotype 1 strains were either smaller (JMp214) or remained the
same (Md11p81) in cell-free genome size as compared to low passage
counterparts (Figure 4, Table 1). High passage serotype 2 strains (SB-1 p101 and
281/Mlp97) were larger in cell-free genome size as compared to low passage

counterparts (Figure 4, Table 1).

Characterization of alterations in attenuated MDV genomes using full length
cell-tree MDV genomes as a hybridization probe.

Attenuated v.v. MDV strain Md11 contains 0.6 to 5.4 Kbp of 132 bp repeat
units in tandem reiteration (Silva and Witter, 1985). However, results from high
resolution CHEF gels indicated that size of cell-free Md11 genomes remained
unchanged after attenuation (Figure 4, Table 1).

To investigate the lack of genome size difference between oncogenic and
attenuated MDV strain Md11, we performed Southern transfer analysis on gels
containing restriction digests of total DNA isolated from cells infected with
oncogenic (Md11p11) and attenuated (Md11p81) MDV isolates. Using CHEF
purified, cell- free Md11 p10 genomes as a hybridization probe viral bands
(contained within the mixture of cellular and viral DNA immobilized on membranes)

were detected (Figure 5, A). BamHI and Bgll restriction fragments of viral DNA

 

_——---u_

63

were assigned letter designations according to published maps of the MDV
genome (Fukuchi et al., 1984). Amplified 132 bp repeat sequences located within
BamHl-D, -H, and Bgll-F viral DNA fragments were evident (Figure 5, A). These
results are consistent with previous findings (Hirai et al., 1981; Maotani et al.,
1986; Silva and Witter, 1985). In addition, BamHI-L and Bgll-J viral DNA
fragments were absent in attenuated Md11 DNA (Figure 5A, lane H). Instead,
attenuated Md11 DNA contained fragments 200 bp smallerthan the BamHl-L and
Bgll J fragments of oncogenic Md11 DNA (Figure 5, A). According to published
maps (Fukuchi et al., 1984), BamHl-L and Bgll J are overlapping fragments located
within the IRL region near the IRL/ IRs junction.

To verify changes assigned to specific MDV BamHl fragments using CHEF
purified cell-free Md11p10 genomic probes, membranes were stripped and
hybridized to cloned BamHl-L (Figure 5, B) and BamHl-H (Figure 5, C) MDV DNA
fragments. Both probes hybridized to identical bands identified using CHEF

purified cell-free Md11p10 genomic probes (Figure 5, B and C).

IV. DISCUSSION
Difficulties intrinsic to preparation of cell-free MDV and its cell-associated
nature during in vitro cultivation have hampered efforts to characterize MDV at the
molecular level. As described in this report, these difficulties can be circumvented
by using PFE to isolate infectious cell-free MDV genomes.
Correlation of genome size to serotype classification of MDV strains was

evaluated by comparing mobilities of cell-free MDV genomes of each serotype on

64
CHEF gels. Oncogenic serotype 1 cell-tree MDV genomes were 17-18 kbp larger

than non-oncogenic serotypes 2 and 3 MDV DNA. Extra genetic material in the
larger genomes of serotype 1 may code for factors involved in MDV-induced
lymphoid proliferation. Comparison of attenuated MDV serotype 1 JMp214 relative
to its low passage, oncogenic counterpart (JMpt 4) supports this hypothesis. MDV
strain JMp214 (an attenuated, non-oncogenic MDV isolate) contains at least 40
copies of the 132 bp tandem repeat (M. R. Wilson, unpublished results). These
repeats would increase the JMp214 genome by 5.28 kbp. However, cell-free
JMp214 genomes purified by CHEF are approximately 10 kbp smaller than
genomes of their oncogenic counterparts (JMp14). Thus, total size of deleted
genetic material in the JMp214 genome may exceed 15 kbp.

It appears that deletions and expansion of tandem repeats may occur
concurrently during attenuation of oncogenic MDV strains. Deletions, which may
be similar in size to expanded 132 bp repeats, may account for the lack of
detectable cell-free genome size difference between oncogenic and attenuated v.v.
MDV strain Md11. Minor differences in genome sizes may go undetected due to
resolution limits of PF gels. Using CHEF purified cell- free Md11p10 genomic
probes, we identified a 200 bp deletion located in the IRL region of attenuated
Md11 DNA. Presently, it is unclear if this deletion is related to mutations which
lead to loss of oncogenicity in attenuated MDV strain Md11. While MDV strain
Md11 lost the ability to form lymphoid tumors after 57 serial in vitro passages, it
retained the ability to produce typical MD microscopic nerve lesions in chickens.

At passage 75, this ability is lost (Witter and Lee, 1984). It is possible that the 200

65

bp deletion in the BamHI-L fragment effects neuroinvasive mechanisms of MDV.
Further comparison of this region isolated from oncogenic and attenuated Md11
strains is in progress.

In contrast to serotype 1 MDV, high passage serotype 2 cell-free genomes
increased in size by 5 kbp (SB-1) and 3 kbp (281Ml/1) relative to their low passage
counterparts. Increased copy number of tandem repeats may contribute to
increased size of serotype 2 genomes after serial in vitro passage. Expansion of
approximately 15 repeats is observed in SB-1p100 (M. R. Wilson, unpublished
results). However, similar expansion of repeat sequences in the 281Ml/1 p96
genome has not been observed (H. Camp, unpublished results).

Infectivity of viral genomes, isolated from pulsed field gels, has not been
assessed previously. Results presented in this report represent the first example
of procedures for recovery of purified infectious viral DNA from pulsed field gels.
Procedures described in this report may be applied to other large DNA viruses for

which isolation of pure intact viral genomes is difficult.

IV. ACKNOWLEDGMENTS
We thank Dr. Robert Bull and Dr. John Gerlach for use of a pulsed field
apparatus employed in early investigations and for many helpful discussions. We
also thank Dr. L. F. Velicer and Dr. S. Triezenberg for critical review of the
manuscript and Mr. Ronald Southwick for excellent technical assistance. This

work was supported by the Research Excellence Fund, State of Michigan, by the

66
Michigan Agricultural Experiment Station and Grants no. 88-37266-3983 and

90-34116-5329 awarded to P. M. Coussens under the Competitive and Special
Research Grants Programs, respectively, administered by the U. S. Department
of Agriculture. M. R. Wilson was supported by a Michigan State University

Biotechnology Research Fellowship award.

67

Figure 1. Separation of oncogenic and attenuated cell-free MDV genomes from
chicken chromosomes using PFE. Cells infected with oncogenic (Md11p10) and
attenuated (Md11p80) serotype 1 MDV isolates were lysed in agarose plugs and
electrophoresed on a 1% agarose CHEF gel for 24 hours at 50-90 sec. ramped
pulse times, followed by staining with EtBr. CEF: uninfected chicken embryo
fibroblasts. Yeast chromosomes and lambda concatemers were used as size
markers. Sizes, in kilobase pairs, are indicated on the right.

68

 

Figure 1.

4660

<I8O
448.5

69

Figure 2. Identification and localization of cell-free MDV genomes and chicken
chromosomes on CHEF gels. Cells infected with oncogenic (Md11p10) and
attenuated (Md11p80) serotype I MDV isolates were lysed in agarose plugs,
electrophoresed. on a 1% agarose CHEF gel for 24 hours at 50-90 sec. ramped
pulse times, transferred to nylon membranes, hybridized to 3"’-P labeled probes
under high stringency conditions, and autoradiographed. Panel A: cloned MDV
DNA BamHl fragments -A, -l2, -H and -P, mixed probe. Panel B. hybridization of
total chicken DNA probe to stripped membrane from panel A. Sizes in, kilobase
pairs, are indicated to the right of each panel.

70

A. MDV PROBE B. CEF PROBE

0‘"? $9
0 0
V} 0 v)? 0
V o o
0 \ ‘b 0 \ ‘b
o x? e e o v e e
-” --
l I
fi‘ . ’6
" "f .9 <660
q” <180
3
«48.5

Figure 2.

71

Figure 3. Infectivity of cell-free MDV genomes after isolation from LMP agarose
CHEF gels. MDV cell-free genomic bands were excised from LMP agarose CHEF
gels, melted at 65°C and directly electroporated into CEF cell suspensions.
Photomicrographs were taken 7 days after electroporation. Resultant viral plaques
were expanded, lysed in agarose plugs and cell-free MDV genomes were isolated
on LMP agarose CHEF gels. Cell-free MDV genomes from electroporated cells
were digested with BamHI, electrophoresed on 0.8% agarose gels, transferred to
nylon membranes, and hybridized under high stringency conditions to cell-free 32-P
labeled MDV genomic probes. Panel A: Photomicrographs at 100x magnification,
1. untreated CEF monolayer control; 2. Typical MDV plaque formation after
infection of CEF with cell-associated MDV; 3. MDV viral plaque formation on CEF
monolayers after electroporation of cell-free MDV genomes. Panel B: Comparison
of BamHl digested MDV parent genomes (lane 1) and cell-free MDV genomes
recovered after electroporation (lane 2). Viral bands were assigned letter
designations according to published maps of the MDV genome (Fukuchi et al.,
1984). Sizes of radiolabeled Hindlll digested Lambda DNA are located to the right
in kilobase pairs.

72

 

.32

2
4

.03

42

Figure 3.

73

TABLE 1
Genome Sizes of MDV Strains as Determined from
High Resolutiona CHEF Gels.

 

 

Serotype Strain kbp" iSD"
1 GAp7 178 n.d.°
Md11p11 178 1.2
Md1 1 p81 178 1.5
JMp15 177 2.1
JMp214 167 1.7
2 SB—1p21 160 0.6
SB-1p101 165 1.1
281Ml/1p16 160 0.6
281 Ml/1 p97 163 0.6
3 HVTp7(FC126) 161 1 .7

 

'High resolution CHEF conditionszelectrophoresis of 1% agarose gels for
32.5 hr at 10-20 sec ramped pulse times in 0.5X TBE at 14°C.

°kbp, kilobase pairs estimated from plots of the log of lambda DNA
concatemer sizes as a function of mobility on CHEF gels.

°Sample standard deviations (SD) were calculated from three independent
CHEF gels used for genome size determinations.

‘n.d., not determined.

74

Figure 4. Cell-free genome size comparison of oncogenic, attenuated, and
naturally occurring non-oncogenic MDV serotypes using high resolution CHEF.
CEF infected with serotype 1 oncogenic (JMp15) and attenuated (JMp214),
serotype 2 low passage (SB-1 p21, 281MI/1 p16) and high passage (SB-1p101,
281 Ml/1 p96), and serotype 3 (HVTp7 F0126) MDV strains were lysed in agarose
plugs and electrophoresed on 1% agarose CHEF gels for 32.5 hours. Resolution
was optimized in the 100-200 kbp range using 10-20 sec. ramped pulse times.
MDV strains are as indicated above each lane. Sizes of lambda concatemers
standards are indicated to the right in kilobase pairs.

75

4339.5
4291.0

4242.5

4194.0

4145.0

497.0

 

Figure 4.

76

Figure 5. Characterization of alterations in attenuated MDV genome stmcture
using CHEF purified cell-free MDV genomes as hybridization probes. Total DNA
was isolated from CEF infected with oncogenic and attenuated serotype l MDV
isolates, digested to completion with excess restriction endonucleases,
electrophoresed on 0.75% agarose gels, transferred to nylon membranes,
hybridized to (132-P labeled probes under high stringency conditions, and
autoradiographed. Panel A: Md11p10 CHEF purified cell-free genomic probe.
Viral DNA bands were assigned letter designations according to published maps
of the MDV genome (Fukuchi, 1984). Panel B: stripped blot from panel A
hybridized to a cloned MDV BamHl-L probe. Panel C: cloned MDV BamHl-H
probe. L: low passage, oncogenic Md11p10, and H: high passage, attenuated
Md11p80 MDV isolates. Restriction endonucleases are indicated above lanes
containing low and high passage pairs. Sizes, in kilobase pairs, of radiolabeled
Hindlll digested Lambda DNA standards are indicated to the left of panel B and
the right of panels A and C.

77

II III: I i.

0 5|:
.2 If?! .
in. u. .
“Duh- Ill-HO
: ._ z ._
Ian _: Eng

0

.m 2:2“.

end

mad
and

00.?
cud

«v.0
or...“

z ._
Ian

I .—
=.. Eon

‘ 1m
1.; HM

- .‘ III.—

In
1N0
8d _ . mm

«QN v

I v.
on... 1..
n...
and \m
«ed mm
9.8 Na

 

_ .00 EEum

Chapter 3.

STRUCTURAL ALTERATIONS IN THE REPEAT REGIONS OF THE
ATTENUATED MAREK’S DISEASE VIRUS GENOME AFFECTS GROWTH
AND IMMEDIATE EARLY GENE TRANSCRIPTION IN VITRO.

Melinda R. Wilson , K. A. Schat“, and Paul M. Coussens

Molecular Virology Laboratory,
Department of Animal Science,
Michigan State University,

East Lansing, Michigan, U.S.A., 48824

To be submitted for publication.

 

' Department of Avian and Aquatic Animal Medicine
New York State College of Vetinerary Medicine
Cornell University, Ithaca, NY

78

ABSTRACT

Marek's Disease virus (MDV), an oncogenic avian herpesvirus, produces
malignant lymphomas and nerve demyelination in infected hosts. MDV is
effectively controlled by vaccination with serotypes 2, 3, and attenuated serotypes
1 strains. However, due to emergence of very virulent (v.v.) MDV in vaccinated
flocks, analysis of the viral attenuation process is of considerable interest.
Genome structure studies, reported by others, indicates that amplification of 132
bp repeats occurs in v.v. MDV strain Md11. Recently, our laboratory has
identified a 200 bp deletion located within the IRL and TR, of the attenuated Md11
genome. This change is separate and distinct from the 132 bp amplified region,
as 132 bp amplification occurs in the BamHl-D and -H fragments, whereas the 200
bp deletion was mapped to the BamHl-L fragment. In this report, we describe the
isolation of a naturally occurring attenuated Md11 variant which contains 132 bp
amplified regions but has the wild type BamHl-L fragment. The attenuated variant
was utilized to determine if deletions in the BamHI-L fragment affect biological
properties of attenuated Md11, and assign a biological function to this region. Our
results indicate that deletions in the BamHl-L region of attenuated Md11 affected
viral growth and immediate early gene expression in vitro. However, comparison
of data from the attenuated Md11 variant indicates that the deletion in BamHl-L is
not solely responsible for enhanced growth and DNA replication of attenuated
Md11. In vivo studies indicate that the BamHl-L deletion does not directly
contribute to MDV-induced oncogenicity. Analysis of RNA transcripts suggests that

this region may be involved in latent infection of MDV-transformed cells.

79

I. INTRODUCTION.

Marek’s Disease virus (MDV), an oncogenic avian herpesvirus, produces
malignant lymphomas and nerve demyelination in infected hosts (Calnek and
Witter, 1991). The MDV genome and many of the genes it encodes is co-linear
with alphaherpesvirus genomes (Buckmaster et al, 1988; Binns and Ross, 1989;
Chen and Velicer, 1991b; Coussens et al., 1990; Ross et al., 1989; Brunovskis and
Velicer, 1992) The MDV genome consists of unique long (UL) and unique short
(Us) regions, flanked by terminal repeats (T RL, TR.) and internal repeats (IRL,
IRS) (Cebrian et al., 1982). MDV is classified into three serotypes: oncogenic
serotype 1, non-oncogenic serotype 2, and non-pathogenic serotype 3 (Payne,
1982). Serotypes 2, 3, and attenuated serotype 1 strains have been used as
vaccines against serotype 1 MDV (Payne, 1982; Witter, 1983). During the course
of vaccination against Marek’s disease (MD), some flocks experience the
emergence of MD characterized by a more severe pathotype. These very virulent
(v.v.) MDV strains are associated with, and considered to be the cause of, vaccine
breaks (Witter, 1983).

Attenuation of serotype 1 MDV occurs after 30 - 55 serial in vitro passages
(Schat et al., 1985). Coincident with loss of oncogenicity in viva, attenuated MDV
tail to induce early cytolytic infection, have a reduced capacity to infect or replicate
in lymphocytes, and lose the ability to spread by contact as compared to
oncogenic MDV (Schat et al., 1985; Witter, 1984). In vitro studies indicate that
attenuated MDV grow faster in cell culture, are able to infect a wider variety of

cultured cells, display reduced expression of glycoprotein C, and contain altered

80

81

genome structure and size as compared to oncogenic MDV (Nazerian, 1970;
Schat, 1985; Bulow and Biggs, 1975; Churchill et al., 1969; lkuta et al., 1983a;
lkuta et al., 1983b; Nazerian, 1980; Purchase et al., 1971 ; Wilson and Coussens,
1991). Genome structure analysis of attenuated isolates revealed changes in
limited regions of the MDV genome (Maotani et al, 1986; Bradley et al, 1989).
These changes were attributed to expansion, by tandem reiteration, of a 132 bp
repeat region located within the TR, and IR, of the serotype 1 MDV genome.
Due to emergence of v.v. MDV in vaccinated flocks, analysis of the
attenuation process is of considerable interest. One such v.v. MDV, strain Md11,
is characterized by severe cytolytic infection leading to early mortality (Witter,
1984). After 57 serial in vitro passages, v.v. MDV strain Md11 became attenuated
with respect to oncogenicity. However, microscopic nerve lesions were still
observed in infected birds through passage 75 (Witter, 1982). Nerve lesions were
not detected after passage 75. Decreased expression of glycoprotein C in
attenuated Md11 has also been established (Witter and Lee, 1984) and the
molecular basis of reduced gC expression was investigated in detail (Wilson et al.,
submitted) and found to involve transcriptional regulation of the gC gene. In
addition, Tieber et al. (1989) has determined that attenuated Md11 is less able to
transactivate an exogenous RSV promoter in vitro as compared to oncogenic
Md11. Taken together, these results suggest that attenuation of Md11 may
perturb a viral gene regulatory element or mechanism. Genome structure studies
indicate that amplification of the 132 bp repeats also occurs in v.v. Md11 (Silva

and Witter, 1985). Recently, our laboratory has identified an additional change in

82

the genome of attenuated Md11(WiIson and Coussens, 1991). Using cell-free
MDV genomic probes, we identified a 200 bp deletion, also located within the IR,
and TR, of the attenuated Md11 genome. This change, however is separate and
distinct from the 132 bp amplified region. 132 bp amplification occurs in the
BamHl-D and -H fragments (Silva and Witter, 1985), and the 200 bp deletion was
mapped to the BamHI-L fragment (Wilson and Coussens, 1991).

Recent advances in purification of cell-free infectious MDV genomes makes
isolation of naturally occurring mutant viruses possible. In this report, we describe
the isolation of an attenuated Md11 variant which contains 132 bp amplified
regions but has the wild type BamHI-L fragment. Additional differences in genome
structure between the attenuated Md11 variant and attenuated Md11 were not
detected. The attenuated variant was utilized to determine if deletions in the
BamHI-L fragment affect biological properties of attenuated Md11, and assign a
biological function to this region. Growth rates, viral DNA replication, and
transcriptional activity of oncogenic, attenuated, and the attenuated Md11 variant
on chick embryo fibroblasts were compared. In viva studies were performed to
determine if the BamHl-L region contributed directly to oncogenicity. Our results
indicate that deletions in the BamHl-L region of attenuated Md11 directly affected
viral growth and immediate early gene expression. However, comparison of data
from the attenuated Md11 variant indicates that the deletion in BamHl-L is not
solely responsible for enhanced growth and DNA replication in attenuated Md11.
In viva studies indicate that the BamHl-L deletion does not directly contribute to

MDV-induced oncogenicity. Analysis of RNA transcripts suggests that this region

83

may be involved in latent infection of MDV-transformed cells.

II. MATERIALS AND METHODS
Cells and viruses.

Preparation, maintenance and vims infection of duck and chick embryo
fibroblasts (DEF and CEF respectively) have been described (Glaubiger et al.,
1983). MDV strain Md11 has also been described (Silva and Witter, 1985; Tieber
et al., 1990; Witter et al., 1983). Oncogenic MDV strain Md11 at passage level 9
and attenuated Md11 at passage level 80 were kindly provided by the USDA.
Avian Diseases and Oncology Laboratory, East Lansing, Michigan, U.S.A.
Md1 1 p15 and Md11p80 to p85 are oncogenic and attenuated Md11, respectively.
Isolation of an attenuated Md11 variant (Md11p83-L) was perlorrned by
electroporating Md11p83 cell-free genomes into CEF as described (Wilson and

Coussens, 1991).

Growth rate and viral DNA replication experiments.

Growth studies were performed as follows: multiple 60 mm plates containing
1x10‘3 freshly plated CEF were infected with 200 plaque forming units (PF U) / plate
in media containing 4% calf serum. 24 hours after infection, media was replaced
with fresh media containing 1% calf serum. Under these conditions, cells were
essentially quiescent. At specific time points (days for PFU, hours for viral DNA),
triplicate cultures were trypsinized and titered on duplicate CEF cultures to

determine PFUfinput PFU. For viral DNA replication studies, cultures were

84

trypsinized and total infected cell DNA was isolated as described (T ieber et al,
1989). Slot blot assays were used to determine molar quantities of viral DNAfinput
viral DNA. Standards used in the slot blot assay were eight 1:4 serial dilutions of
an equi-molar mixture of MDV BamHl-E, -l,, -K, -L, and -M plasmids. Standards
ranged from 6 pM to 0.36 fM. Probe DNA was an equi-molar mixture of
radiolabeled, gel purified MDV specific fragments isolated from the same plasmids

used for standards. pBR322 was spotted as a negative control.

Immunoprecipitations.
Immunoprecipitations of proteins from cells infected with oncogenic and

attenuated Md1 1 and Md1 1p83-L were performed as described (lsfort et al., 1986).

Plasmids and radiolabeled probes.

BamHI fragment DNA clones of MDV strain GA (Fukuchi et al., 1984) were
kindly provided by M. Nonoyama (Showa University Research Institute for
Biomedicine in Florida, St. Petersberg, FL.). The BamHl-L fragment of the MDV
GA clone was subcloned into pUC19 and used for all experiments. cDNA L, was
a generous gift from K. Ohashi (Cornell University, Ithaca, NY). or-“P labeled
DNA probes were prepared using a random prime DNA labeling kit (Boehinger
Mannheim, Indianapolis, IN) according to manufacturer's instructions. All
radiolabeled DNA probes were prepared from gel purified inserts of plasmids

described above.

85
DNA Isolation and Southern Blots.

Infections of DEF with oncogenic, attenuated, and Md1 1p83-L were allowed
to progress until cell monolayers appeared fully infected (4-10 days). Total
infected cell DNA was isolated as described (Sambrook et al., 1989). The amount
of viral DNA in total infected cell DNA was determined by slot-blot assays as
described above. Similar amounts of total infected cell DNAs were digested with
excess restriction enzymes for 16 hrs according to manufacture’s specifications,
electrophoresed through agarose gels and transferred to nylon membranes in 0.4
N NaOH as described (Budowle and Baechtel, 1990). Radiolabeled probes were
hybridized to nylon filters at 65°C as described (Budowle and Baechtel, 1990).

Filters were exposed to XAR5 film for 30 min to 48 hrs at room temperature.

RNA Isolation and northern blots.

Infections were initiated with 3000 PFU/5x1 0° freshly trypsinized CEF in 100
mm plates. Infected cells were allowed to settle on plates for 6 hours. To obtain
immediate early (IE) transcripts, media was removed from triplicate sets of plates
6 hours post infection, and fresh media containing 100 ug/ml cycloheximide (CHX),
4% calf serum was added. Additional cultures were treated with 100 ug/ml
phosphonoacetic acid (PAA) in the same manner as described above. PAA
treatment leads to accumulation of IE and early (E) transcripts. 24 hours post
treatment, media was replaced with fresh media containing CHX or PAA and 4%
calf serum. Total infected cell RNA was isolated 48 hours post-infection using the

proteinase K/SDS method (Sambrook et al., 1989) followed by digestion with

86
RNAse free DNAse I (Promega, Inc., Madison WI). 20 ug of total RNA was loaded

per well, electrophoresed through formaldehyde gels and transferred to nylon
membranes (Hybond N, Amersham Corporation, Arlington Heights, Illinois)
according to manufacturer's specifications and as described (Sambrook et al.,
1989). Radiolabeled MDV BamHl-L probes were hybridized to filters using high
stringency conditions (Sambrook et al., 1989). RNA band sizes were determined
by comparison to RNA standards (Bethesda Research Labs, Bethesda MD) run
on the same gel. RNA blots were stripped of probe DNA and hybridized with

chicken B-actin gene specific probes (Cleveland et al, 1980).

In viva experiments.

Two day old P2a @939) chicks (Weinstock and Schat, 1987) were obtained
from the departmental flocks of the Department of Avian and Aquatic Animal
Medicine, Cornell Univ., Ithaca, NY These flocks are free of all known avian viral
infections and are genetically susceptible to Marek's Disease. Chicks were divided
at random in 3 groups and infected at 2 days of age by intraabdaminal inoculation
with 375 PFU of Md1 1p14 (oncogenic), 225 PFU of Md1 1p83 (attenuated), or 1500
PFU of Md11p85-L (variant). Chicks were kept in isolation facilities with ad Iibitum
feed and water. All chickens were debeaked at 62 days of age. All birds showing
signs of distress were euthanized by CO2 Inhalation and examined for gross MD
lesions. Survivors were euthanized at 92 and 93 days of age and examined for

pathological signs of MD.

87
Data analysis.

Intensity of bands on autoradiographs were analyzed on a FB910
Densitometer using Zeineh 1-D Videophoresis II software (FisherBiotech). Graphs
were generated using Sigma Plot versions 3.1 and 5.0 (Jandel Scientific, Corte
Madera, CA). Where appropriate, statistical analysis was performed with ANOVA

using the general linear models procedure of SAS (1985).

III. RESULTS
Isolation of an MDV strain Md11 variant.

Attenuated Md11p83 cell-free genomes were electroporated into CEF as
described in Materials and Methods. Resultant viral plaques were purified and
expanded to obtain cell-free genomes for analysis of viral genome structure by
Southern blot hybridization. Though most plaques contained MDV genome
structure indistinguishable from that of parental virus, one plaque contained a
BarnHl-L fragment 200 base pairs larger than that of parental virus DNA (Figure
1, lane 7). This variant vian was designated Md11p83-L. Further comparison of
DNA from Md1 1p83-L to oncogenic Md1 1 and attenuated Md11 DNA revealed that
the BamHl-L fragment of Md1 1p83-L was identical in size to the BamHl-L fragment
of oncogenic Md11 (Figure 2A; compare LP vs. V). In the oncogenic Md11 lane
(LP), a minor species which co-migrated with the band in the attenuated Md11
lane (HP) was also visible. In addition, Md1 1p83-L contained the expanded region

found in the BamHl-D and -H fragments (Figure 2B) of attenuated Md11.

88

Preliminary observations on viral growth characteristics suggested that Md1 1p83-L
grew slower than parental Md11p83 virus. Thus, we used Md11p83-L in
comparative experiments to study the biological consequences of the 200 bp

deletion in the BamHl-L fragment of attenuated Md11.

Growth characteristics of oncogenic, attenuated, and Md11 p83-L on CEF and
DEF.
Plaque morphology In chick and duck embryo fibroblasts.

Oncogenic Md11 formed characteristic microscopic plaques on DEF and
CEF (Figure 3, row 1). Attenuated Md11 and Md11p83-L also formed plaques but
these plaques were approximately 4 times larger than in oncogenic Md11 infected
cells (Figure 3, rows 2 and 3). On DEF, attenuated Md11 and Md11p83-L caused
formation of syncytia and multinucleated cells (Figure 3B, row 2 and 3). Syncytia
formation was not observed with oncogenic Md11 on DEF or CEF (Figure 3, row
1). Qualitative differences in plaque morphologies were consistent with previous

observations of other oncogenic and attenuated MDV strains (Nazerian, 1970).

Viral replication In chick embryo fibroblasts.

During initial isolation of Md1 1p83-L from CEF, preliminary results indicated
that Md11p83-L grew slower than its parent strain, Md11p83. To substantiate
these observations and determine if viral growth in vitra was affected by mutation
of the BamHl-L fragment, quantitative growth studies on CEF were initiated.

Growth curves of oncogenic and attenuated Md11, and Md11p83-L are illustrated

89
in Figure 4A. In CEF, attenuated Md11 and Md11p83-L exhibited a lag phase of

approximately 1 day, followed by exponential increase in plaque numbers. In
contrast, oncogenic Md11 did not display an exponential increase in plaque
numbers (Figure 4A). Consistent with previous results (Nazerian, 1970), plaque
formation of oncogenic Md11 was low. The attenuated variant, Md11p83-L,
displayed growth characteristics intermediate to those of oncogenic and attenuated
Md11, but was more similar in growth to attenuated than oncogenic Md11. PFU
on day 4 of Md11p83-L were 2-fold lower (p<0.01) than PFU on day 4 of parental

Md1 1 p83.

Viral DNA replication in chick embryo fibroblasts.

To determine if differences in plaque formation were related to viral DNA
replication, total infected cell DNA was isolated from cultures at various time points
from 0 to 69 hours post infection. Very little viral DNA replication occurred for
oncogenic Md11, whereas replication of attenuated Md11 and Md11p83-L viral
DNA was evident (Figure 4B). After 69 hours in culture, amount of viral DNA/Input
viral DNA for attenuated Md11 was approximately 2-fold higher than that of
oncogenic Md11 (0.05<p<0.01). Amount of Md11p83-L viral DNAfinput viral DNA,
after 69 hours in culture, was not significantly different from attenuated Md11
(p>0.5), and only slightly higher than oncogenic Md11 viral DNAfinput viral DNA
(0.1<p<0.2) (Figure 4B). Due to presence of residual, non-replicating DNA in the
cell-associated virus inoculum, decrease in viral DNA content in infected CEF

cultures was observed from 0 to 27 hours for attenuated Md11 and Md11p83-L

90

and from 0 to 45 hours for oncogenic Md11. Thus, non-replicating viral DNA from
cell-associated inoculums interfered with accurate assessment of viral DNA

replication from 0 to 27 hours.

Rates of plaque formation and viral DNA replication In chick embryo
fibroblasts.

Due to differences of plaque formation, growth rates in CEF of all three
strains were compared for day 1-3 (Figure 5). Md11p83-L grew approximately 2-
fold faster than oncogenic Md11 (p<0.01), but grew slower than attenuated Md11
(p<0.01). The cell associated nature of MDV precludes determination of virus
particle/ PFU ratios. Thus, DNA/PFU ratios for all strains were determined on day
4. Day 4 was chosen because virus was in log phase growth and contributions of
input DNA from cell-associated inoculums were minimal. For CEF infected
cultures DNA/PFU ratios for oncogenic, attenuated and Md11p83-L were 0.44 :t
0.009 S.E.M., 0.05 :I: 0.002 S.E.M., and 0.08 i 0.002 S.E.M. fMoles DNA/PFU,
respectively. DNA/PFUfor oncogenic Md11 was 8- to 5-fold higherthan DNA/PFU
of attenuated Md11 (p<0.05) or Md11p83-L (p<0.05), respectively. Differences in
DNA/PFU of attenuated Md11 as compared to Md11p83-L was not significant
(p>0.5).

Rates of viral DNA replication from 27 to 69 hours post infection in CEF
were compared for all 3 strains. Md11p83-L rate of viral DNA replication was
intermediate to oncogenic and attenuated Md11, consistent with data for rates of

plaque formation (Figure 5). Rate of viral DNA replication for attenuated Md11

91
was slightly higher than that of Md11p83-L (p<0.05). Rate of viral DNA replication

for oncogenic Md1 1 was significantly lowerthan that of attenuated Md1 1 (p<0.001)

or Md11p83-L (p<0.001).

Transcriptional activity and expression of major glycoprotein antigens.

Attenuation of MDV strain Md11 results in a severe reduction of 9C
expression, where as expression of other major MDV antigens (98, and p79)
remains unchanged (Wilson et al., submitted). To determine if the 200 bp deletion
within the BamHl-L fragment affected the expression of major MDV antigens,
immunoprecipitations were performed on oncogenic and attenuated Md11, and
Md11p83-L. Using convalescent chick sera (immunoprecipitates QB, 90 and p79)
or gC specific rabbit sera, there were no differences observed between amounts
of -viral antigens expressed in cells infected with attenuated Md11 or Md11p83-L
(data not shown). Thus, presence of the wild type BamHl-L fragment in the
Md11p83-L genome did not restore gC expression levels to that of oncogenic
Md11. These results suggest that the 200 bp deletion in the BamHI-L fragment
did not significantly alter the expression of the major antigens of MDV.

To determine if transcripts within or partially within the BamHI-L fragment
were affected by the 200 bp deletion in attenuated Md11, northern blot analysis
of total infected cell RNA was performed. Hybridization to a MDV BamHl-L
fragment probe was used to compare transcriptional activity in the BamHI-L region
of oncogenic and attenuated Md11, and Md11p83-L. Three major transcripts,

which hybridized to the BamHl-L probe, of 2.8, 1.7, and 0.8 Kb are present in all

92
strains (T able 1). In addition, a 10 kb transcript present only in CHX treated

groups was detectable (data not shown). These data were similar to previous
reports using RNA from other oncogenic MDV strains (Schat et al, 1989; Sugaya
et al, 1990). In attenuated Md11, the 0.8 kb transcript was expressed at higher
levels than that of oncogenic Md11 or Md11p83-L (Table 1). Relative abundance
of specific transcripts varied among strains and transcriptional class. Data for
transcripts expressed in oncogenic Md11 indicates that the 2.8, 1.7, and 0.8 kb
transcripts were expressed as IE genes (Table 2). In attenuated Md11 infected
cells treated with CHX, the 1.7 and 0.8 kb transcripts were not detected (T able 2)
even though these transcripts were expressed in the untreated and FAA treated
groups. In Md11p83-L infected cells treated with CHX, the 1.7 kb transcript was
not detected, but was expressed in the untreated and PAA treated groups.
However, the 0.8 kb transcript in Md11p83-L infected cells was expressed in a
similar fashion to that in oncogenic Md11 infected cells (T able 2). These data
indicate that patterns of expression of the 0.8 kb IE transcript was affected by the
200 bp deletion in the BamHl-L fragment of attenuated Md11. Data also indicates
that the 0.8 kb IE transcript was expressed as an E transcript in attenuated Md11.
In addition, attenuation of Md11 affects patterns of expression of the 1.7 kb IE
transcript. As with the 0.8 kb IE transcript, data indicates that the 1.7 kb IE
transcript is expressed as an E transcript in attenuated Md11. However, deletion
in the BamHl-L fragment did not have a significant affect an altered expression of
the 1.7 kb IE transcript, as expression patterns for the 1.7 kb IE transcript in

Md11p83-L were similar to that of attenuated Md11.

93
To confirm that only IE and E genes were detected, and CHX or PAA

treatment effectively blocked late gene expression, membranes were stripped and
re-hybridized with a late gene (MDV gC) specific probe. Only untreated lanes

containing oncogenic Md11 hybridized to the 9C probe (data not shown).

Sub-localization of the 200 bp deletion within the BamHI-L fragment.

Previous reports indicate that 0.6-0.8 kb immediate early (IE) transcripts
which hybridize to the BamHl-L fragment probe are expressed in MDV transformed
lymphocytes (Schat et al, 1989; Ohashi and Schat, 1992). Recently the same
group generated a cDNA (termed L,) derived from RNA isolated from MDV
transformed lymphocytes. L1 hybridizes to IE transcripts encoded within or partially
within the BamHl-L fragment (Ohashi and Schat, 1992). Physical mapping studies
were initiated to determine where in the BamHl-L fragment the 200 bp deletion
occurred in attenuated Md11 DNA and if this deletion was within or near the L1
cDNA region.

DNA isolated from oncogenic, attenuated, and Md1 1p83-L infected cells was
double digested with BamHI and Hpall, Haelll, or Hinil and subjected to Southern
transfer analysis. Initially, membranes were probed with a full length BamHI-L
DNA fragment. 1.13 and 0.36 kb Haelll, 1.05 and 0.67 kb Hpall, and 0.95 kb Hinll
BamHl-L sub-fragments in attenuated Md11 DNA were decreased in size as
compared to oncogenic Md11 and Md11p83-L DNA (Figure 6A). Fragment sizes
in oncogenic Md11 and Md11p83-L DNA were identical. Fragment sizes in

oncogenic Md11 and Md11p83-L DNA corresponding to fragments affected by the

94
deletion in attenuated Md11 DNA were 1.25 and 0.37 kb Haelll, 1.15 and 0.745

kb Hpall and 1.05 kb Hintl sub fragments of BamHl-L. L, cDNA, which was
previously mapped to the 5' end of MDV BamHl-L fragment (Ohashi, 1992), was
used as a probe for stripped membranes of panel A (Figure 68). L, hybridized
strongly to the 1.13 kb Haelll and 1.05 kb Hpall fragments in attenuated Md11
lanes (Figure 6B). Very weak hybridization was visible with the 0.95 kb Hintl
fragment. These results indicated that the 200 bp deletion was located within the
0.95 kb Hintl fragment. This fragment was adjacent to the 3’ end of L, cDNA
(Figure 7). To confirm this observation, membranes were stripped and probed with
the 1.0 kb Hintl sub-fragment of the MDV strain GA clone. As predicted, the 1.0
kb Hintl probe hybridized strongly to the 1.13 and 0.36 kb Haelll fragments, 1.05
and 0.67 kb Hpall fragments, and the 0.95 kb Hintl fragment of attenuated Md11
DNA (Figure 60). Weak hybridization was evident in bands which were not altered
in attenuated Md11 DNA. Sub-localization of the deletion in relation to location of
L, cDNA is illustrated in Figure 7.

In oncogenic Md11 DNA, minor populations of DNA fragments which
co-migrated with DNA fragments containing the 200 bp deletion of attenuated
Md11 DNA were evident (Figure 6A, B, and C). Conversely, in attenuated Md11
DNA, minor populations of DNA fragments which co-migrated with DNA fragments
of oncogenic Md11 DNA were also evident, but was not detected in Md11p83-L
DNA. These data indicate the presence of minor sub-populations in oncogenic

and attenuated Md11 DNA.

95

Assessment of oncogenicity in viva.

Physical mapping data of the deletion in the BamHI-L fragment indicated
that the BamHI-L fragment in Md11p83-L was identical in structure to that of
oncogenic Md11. Therefore it was of interest to determine if the BamHI-L region
contributed directly to oncogenicity. MD susceptible chicks were inoculated with
oncogenic, attenuated Md11 or Md11p83-L. As expected, a high incidence of MD
(95%) was found in oncogenic Md11 (T able 3). In the attenuated Md11 group,
only 1 birds displayed visible signs of MD, whereas no visible MD lesions were
found in the Md11p83-L group (Table 3). In this study, there was a high

prevalence of tumors in heart (58%) and spleen (37%) induced by oncogenic

Md11. However, the MDT bird infected with attenuated Md11 had tumorous liver

and spleen tissue. Conversely, no liver tumors were found in MD"' birds infected

with oncogenic Md11.

IV. DISCUSSION

Attenuation of oncogenic MDV offers a unique opportunity to examine
mechanisms involved in herpesvirus-induced neoplasia in a natural host. Major
changes in attenuated MDV genome structure, coincident with loss of oncogenicity,
are well documented (Bradley et al., 1989; Chen and Velicer, 1991; Maotani et al.,
1986; Silva and Witter. 1985). Disruption of RNA transcription due to amplification
of 132 bp repeat sequences within the IR, region of the MDV genome may be
related to MDV attenuation (Bradley et al, 1989). With the recent adaptation of

pulsed field gel electrophoresis technology to the study of MDV (lsfort et al., 1990;

96

Wilson and Coussens, 1991), deletions in the BamHI-L fragment of the attenuated
Md11 genome were identified (Wilson and Coussens, 1991). Subsequently, we
were able to recover a purified population of an attenuated Md11 variant
(Md11p83-L) from a heterogeneous stock of attenuated Md11. Data in this report
indicates that the Md11p83-L genome contains only one of the two changes thus
far identified in attenuated Md11. DNA physical mapping studies indicate that the
deletion in the BamHI-L fragment of attenuated Md11 occurred early during serial
passage in culture, since the oncogenic virus used in these studies, Md11 passage
15, contained a minor population of the deleted BamHl-L fragment, whereas earlier
passages (Md11p10) did not contain detectable sub-populations of the deleted
BamHI-L fragment (Wilson and Coussens, 1991). Md11p83-L was utilized in
comparative studies to determine if deletions in the BamHl-L fragment of
attenuated Md11 affected biological properties of the virus.

in vitro, deletions in the BamHl-L fragment of attenuated Md11 genomes
contributed to alteration of growth characteristics of the virus. Since plaque
formation and DNA replication ofMd11p83-L was intermediate between oncogenic
and attenuated Md11, deletion in the BamHl-L region was only partially responsible
for increased plaque formation and DNA replication of attenuated Md11. Our
results further suggest that amplification of the 132 bp repeat sequences, within
IR,, also contributed to enhanced viral replication in vitro. DNA/PFU ratios for
oncogenic Md11 were 8- and 5-fold higher than DNA/PFU ratios of attenuated
Md11 and Md11p83-L, respectively. This difference may suggest that oncogenic

Md11 was less able to spread to neighboring cells, leading to accumulation of non-

97

viable virus in vitra. Small size of oncogenic Md11 plaques as compared to
attenuated Md11 plaques supports this nation. In addition, similarity of plaque size
in Md11p83-L and attenuated Md11 infected cells suggests that the 200 bp
deletion in the BamHl-L region may not contribute to initial virus spreading
efficiency in vitro. However, preliminary results from our laboratory indicate that
during extended in vitro cultivation (5 to 9 days), the efficiency of secondary plaque
formation of oncogenic Md11 and Md11p83-L is higher than that of attenuated
Md11. Studies designed to differentiate between vims spread to neighboring cells
and virus replication within a single cell would help clarify these observations.
Analysis of RNA transcripts indicated that temporal expression of genes
encoded or partially encoded within the BamHI-L fragment, specifically the 0.8 kb
IE transcript, were directly affected by deletion in the BamHI-L fragment of
attenuated Md11 DNA. In addition, temporal class switching of the 0.8 kb IE
transcript suggest that this deletion may reside within the promoter of an IE gene.
Deletion within the BamHl-L fragment did not directly contribute to reduction in the
expression of the 1.7 kb transcript of attenuated Md11. The 2.8 kb IE transcript
was also affected by attenuation of Md11. However, data in this report indicates
that the 200 bp BamHI-L deletion was not a major factor for alteration of the 2.8
kb IE transcription pattern. Lack of detection of the 2.8 kb IE transcript in PAA
treated, attenuated Md11 or Md11p83-L infected cells suggests that attenuation
of Md11 may affect RNA stability of the 2.8 kb IE transcript. Alternately,
expression of the 2.8 kb IE transcript may have been subject to repression by a

cellular or viral-encoded regulatory protein in attenuated Md11 and Md11p83-L

98

infected cells.

Previous reports indicate that transcripts, similar in size to the 0.8 kb IE
transcript affected by attenuation, are expressed in producer as well as non-
producer MDV transformed cell lines (Schat et al, 1989). In addition, this transcript
was tmncated by approximately 300 bp in a non-producer MDV transformed cell
line. Since non-producer MDV transformed cells arise after prolonged in vitro
cultivation, our results, taken together with previous results (Schat et al, 1989)
suggest that the BamHI-L region is affected by prolonged in vitro cultivation of
MDV.

The observation that alteration of IE transcripts, DNA replication, and growth
properties in vitra are related to mutation in the BamHl-L fragment of attenuated
Md11 genomes and the ca-linear location of MDV genes with VZV and HSV
(Buckmaster et al, 1988; Binns and Ross, 1989; Chen and Velicer, 1991b;
Coussens et al., 1990; Ross et al., 1989; Brunovskis and Velicer, 1992) leads to
speculation that a protein functionally similar to HSV-l ICPO is encoded within the
BamHI-L region affected by attenuation. This is substantiated by previous results
which indicate that mutation in the HSV-l ICPO gene affects lytic growth,
expression of early and late proteins, and particle/PFU ratios of HSV-l in vitra (Cal
and Schaffer, 1992).

In vivo studies indicated that the deletion in the BamHI-L fragment of
attenuated Md11 DNA did not contribute directly to MDV-induced neoplasms.
However, MDV-induced oncogenicity is most likely a complex multi-step process

involving the viruses ability to specifically infect lymphocytes, induce cellular and

99

viral transactivator proteins, and maintain lymphocytes in a transformed state.
Expression of BamHl-L immediate early transcripts in MDV transformed cell lines
(Schat et al, 1989; Sugaya et al, 1990), similar in size to the 0.8 kb transcript
affected by Md11 attenuation, suggests that this region may be involved in latent
infection of MDV transformed cells. In viva studies, utilizing Md11p83-L, to
determine if the deletion in the BamHI-L fragment of attenuated Md11 genomes
affects latent MDV infection of lymphocytes or early cytolytic infection are in
progress.

Comparison of the BamHl-L fragment and transcripts encoded within the
BamHl-L fragment of other oncogenic MDV strains to their attenuated counterparts
would be of interest to determine if deletions occurring in this region is a general

phenomenon of attenuation or specific to v.v. MDV attenuation.

100

ACKNOWLEDGMENTS

We thank Jessica Marcus for excellent technical assistance. We also thank
S. J. Triezenberg for careful review of this manuscript and gratefully acknowledge
Telmo Oleas and John Gill for helpful discussions and assistance with statistical
analysis. This work was supported by the Michigan Agricultural Experiment
Station, the Research Excellence Fund, State of Michigan, and by Grant Nos. 88-
37266-3983, 90-34116-5329, and 92-37204-802 awarded to P. M. Coussens
under the Competitive Research Grants Program administered by the US.
Department of Agriculture. M. Wilson was supported in part by a Michigan State

University Biotechnology Research Fellowship.

101

Figure 1. Identification of an attenuated MDV strain Md11 variant. Attenuated
Md11 cell-free genomes were electroporated into CEF cell suspensions as
described (Wilson and Coussens, 1991). Resultant viral plaques were expanded,
lysed in agarose plugs and cell-free MDV genomes were isolated on low melting
point agarose CHEF gels as described (Wilson and Coussens, 1991). Cell-free
MDV genomes from electroporated cells were digested with BamHI,
electrophoresed on 0.8% agarose gels, transferred to nylon membranes, and
hybridized under high stringency conditions to cell-free or- 32P labeled MDV
genomic probes. BamHl digests of attenuated Md11 parent genomes (first lane)
and cell-free MDV genomes recovered after electroporation (CHEF: 6 and 7). Dot
on right denotes altered BamHl-L fragment in the CHEF 7 lane. Viral bands were
assigned letter designations according to published maps of the MDV genome
(Fukuchi et al., 1984). Sizes of radiolabeled Hinolll digested Lambda DNA are
located to the right in kilobase pairs.

102

CHEF
6 7

PARENT

 
  

A\ -<
B, 23.13
C?
E/
F/ ‘942
e/ .
H’ h «6 56
I- ‘ can.
J-

4.36
K.— o

O
L— " ‘
M...

«2.32
N; ' «2.03
p-
2:
S"
T..—

Figure 1.

103

Figure 2. Identification of the altered DNA fragment in the attenuated Md11
variant as BamHl-L. Cell-free MDV genomes of oncogenic (LP), attenuated (HP)
and the attenuated variant (V) were isolated, digested with BamHI, and subjected
to Southern transfer analysis as described in the Figure 1 legend. Left panel:
autoradiographs of membrane hybridized to a radiolabeled MDV BamHl-L DNA
fragment. Right panel: membrane depicted in the left panel was stripped and
hybridized to a radiolabeled MDV BamHl-H DNA fragment.

104

LPV HP

LPVHP

 

 

Figure 2.

105

Figure 3. Comparison of plaque morphology of oncogenic (1) and attenuated (2)
Md11 to Md11p83-L (3). Photomicrographs at max magnification were taken 7
days after infection of chick (A) or duck (B) embryo fibroblasts. Infections were
performed as described in Materials and Methods.

106

 

Figure 3.

107

Figure 4. Panel A: comparison of plaque formation of oncogenic and attenuated
Md11 to Md11p83-L in CEF. Growth curves forindividual strains were determined
as described in Materials and Methods. Panel B: Viral DNA replication of
oncogenic and attenuated Md11, and Md11p83-L in CEF. Isolation of total
infected cell DNA, slot bolt assays, and replication curves were performed as
described in Materials and Methods. On all graphs, each data point represents the
average of three separate determinations.

mmDZJU 7: mEDOI

.e 959....

 

 

 

 

 

mm mv vm O
.-l--. -- .r II 8.0
.II./#
. CI 0
a \ p .mfio
«I. /.z m
\ Goo; N
M
d
\\ m
.3, o
N
V
. .00—
a
33 £2 {E m
mm_.;-.-m.z_mm301
me me 4m 0
.l I - -I...Il ii? 000
.mmd
Iolo .w
OGQ: /
. M
d
m.
.2; a
N
uv
eon.—
v—QFFUE 62m

Nb

mdiI

«.9.

manssu z. mmDOI
we 3N

blmma__32

mmDZDU z. m><o

m
4
C

N

0““

4Ir4 513.: .6:
CI 0 853232
OIO 928820

. o
o
\“IIIC

VNCI lfldNI/VNO

.om...

refim

 

.ON
.oe
.om
.Ow
Too—
.ON—
.0»;
.00,

:2 <

ma lndNI/njd

 

109

Figure 5. Rates of plaque formation and DNA replication during day 1-3 of
oncogenic and attenuated Md11, and Md11p83-L in CEF. Error bars represent
standard error of the ratio: day 3/day 1. Comparison of attenuated Md11 to
Md11p83-L: p<0.01 and p<0.02 (t-test, ANOVA) for rate of plaque formation and
rate of DNA replication, respectively.

110

.m 059“.

Toma _ :5 oEész‘

0.2mooozo

 

 

 

 

We

H

 

 

 

 

 

 

 

 

 

a

 

 

wk

 

 

 

§<zo a
D Dun.

 

 

 

 

 

.m2_<w:m :22 no ZO:.<0_r_n_mm_ <20

oz<

ZO:.<§mOm MDOSQ .._O mug»;

OF

N—

C—I AVG BSVBEIONI O‘IOJ

111

Table 1. Abundance of viral transcripts which hybridize to the MDV BamHI-L DNA
fragment. Isolation of total infected cell RNA, northern transfer analysis,
hybridization with a full length radiolabeled MDV BamHl-L DNA fragment probe,
or a radiolabeled chicken B-actin DNA fragment, and densitometry was performed
as described in Materials and Methods. Transcript levels were normalized to
chicken B-actin expression.

Table 2. Temporal classification of viral transcripts which hybridize to the MDV
BamHl-L DNA fragment. Treatment of CEF infected with oncogenic or attenuated
Md11 or Md11p83-L with cycloheximide or phosphonoacetic acid, Isolation of total
infected cell RNA, northern transfer analysis, hybridization with a full length
radiolabeled MDV BamHI-L DNA fragment probe, or a radiolabeled chicken B-actin
DNA fragment, and densitometry was performed as described in Materials and
Methods. Abundance of transcript (normalized to chicken B-actin expression) are
expressed as % of transcript levels from untreated infections.

112

Table 1.
Abundance of viral transcripts which hybridize to the MDV BamHI-L DNA
fragment.

ONCOGENIC ATTENUATED
Md11 Md1 1

 

 

 

 

Table 2.
Temporal classification of viral transcripts which hybridize to the MDV BamHl-L
DNA fragment.

ONCOGENIC ATTENUATED

            

 

TREATMENT srze Md11p83-L
(1(a) Md11 Md11
CHX 2.8 290 100 150
1.7 150 .. ..
0.8 so - 4o
PAA 2.3 110 - .-
1.7 70 so 110
0.8 30 8 4o

 

 

 

113

Figure 6. Restriction mapping to determine sub-localization of the 200 bp deletion
in the attenuated Md11 genome. Total DNA was isolated from CEF infected with
oncogenic or attenuated Md11 or Md11p83-,L double digested with BamHI and
Hpall (1), BamHI and Haelll (2), or BamHI and Hintl (3), electro haresed on1.2%
agarose gels, transferred to nylon membranes, hybridized to or- 2P labeled probes
under high stringency conditions, and autoradiographed. Panel A: Full length MDV
BamHl-L fragment probe. Panel B: Stripped blot from panel A hybridized to the
BamHI - Xhal sub-fragment of L, cDNA (Ohashi and Schat, 1992). Panel C:
Stripped blot from panel B hybridized to the 1.0 kb Hintl sub-fragment of MDV
BamHl- L. Sizes, in kilobase pairs, of BstEll digested Lambda DNA standards are
indicated to the right of panel C.

114

.m 959".

0 . m

«adv ‘1‘ . I
_
ONOV a. O I -
. O 3
on; .. .. .3 ’. D‘ D .l
50;" .
004V
N0.NV
$0.97
0 N —. a N _. O N —. 0 N _. n N F M N F
ti .0 O 4 b O
0 AV 0 0 AI 0
‘40 .00 49 \Isn. #0 49
09( Ol 09¢ O/

 

115

Figure 7. Sublocalization of the 200 bp deletion in the BamHl-L fragment of
attenuated Md11 genomes. -Q2 and -A are BamHI fragments adjacent to the
BamHl-L fragment. Location of L, cDNA (Ohashi and Schat, 1992) is depicted by
a black box. Gray shaded line highlights the 1.0 kb Hintl fragment which contains
the deleted region in attenuated Md11. Parentheses around Hinfl indicates that

location of this site is approximate.

BESS :22 33895 so .859: ._-_.._Emm as E 8:28 8 8m 2: 3 22353858

K 959“.

8.8.8 8 8w
9:58:00 c202“.

=8: \IIKfIIJ :8:

=5: as;

 

 

 

 

Eon:
_IEmm iEmm

116

 

 

 

2.553 m Lm ._ _ ..
>92 E. _ m; m. _ E

 

117

Table 3. Incidence of Marek’s Disease with Md11 strains. In vivo studies were
performed as described in Materials and Methods. % Marek’s Disease was
determined from number of birds which displayed gross lymphoma formation.

118

Table 3.
Incidence of Marek's Disease with Md11 strains.

STRAIN NO. BIRDS AGE AT % MAREK’S
AND TERMINATION DISEASE
PASSAGE (WEEKS)

Md11p14

Md11p83

Md11p83-L

 

 

 

 

Chapter 4.

MOLECULAR ANALYSIS OF THE GLYCOPROTEIN C-NEGATIVE
PHENOTYPE OF ATTENUATED MAREK'S DISEASE VIRUS.

Melinda R. Wilson, James T. Pulaski, Ronald A. Southwick,
Virginia L. Tieber', Yu Hong, and Paul M.Coussens.

Molecular Virology Laboratory
Department of Animal Science
Michigan State University
East Lansing, Michigan 48824

Submitted to Virology 2/17/93.

 

' Present Address: Department of Biochemistry, Michigan State University.

119

ABSTRACT

Serial passage of oncogenic serotype 1 Marek’s Disease vian (MDV) in
cultured avian cells results in attenuation of viral oncogenicity and pathogenicity.
Coincident with attenuation, expression of MDV glycoprotein C (90) is significantly
reduced. Regulation of MDV gC may have important implications for virus-host
interactions, and have important consequences in the process of viral attenuation.
To investigate the mechanism by which MDV gC expression is reduced during
attenuation of the very virulent serotype 1 MDV strain Md11, protein levels, gene
structure, steady state RNA levels, and transcription rates of MDV 90 from
oncogenic and attenuated isolates of Md11 were determined. Comparison of
these data with similar studies on MDV proteins whose expression is not altered
during attenuation indicate that reduced expression of MDV 90 is directly related
to reduction in transcription rate of the MDV gC gene in attenuated Md11.
Reduced transcription rates and lack of any gross structural alterations in the
attenuated MDV 90 gene suggests that MDV regulatory protein(s), which interact
with the MDV gC promoter are altered during attenuation of MDV strain Md11.
These regulatory proteins may be encoded within highly polymorphic regions of the
attenuated MDV strain Md11 genome. Further study of MDV antigens and
transcription factors at the molecular level may provide important clues for
understanding virus-host interactions and immune response to MDV and other

herpesvirus infections.

120

I. INTRODUCTION

Marek’s Disease virus (MDV), an oncogenic avian herpesvirus, produces
malignant lymphomas in infected hosts (Calnek and Witter, 1991). MDV is
classified into three serotypes: oncogenic serotype 1, non-oncogenic serotype 2,
and non-pathogenic serotype 3 (Payne, 1982). Serotypes 2 and 3 have been used
as vaccines against serotype 1 MDV (Payne, 1982; Witter, 1983). There are at
least six antigenically active viral proteins found in all three serotypes of MDV
(lkuta et al., 1983; Silva and Lee, 1984). Two of these, MDV A (90) and B (98)
antigens are homologous to herpes simplex virus-1 (HSV-1) glycoproteins C and
B, respectively (Binns and Ross, 1989; Buckmaster et al., 1988; Chen and Velicer,
1991b; Coussens et al., 1990; Ross et al., 1989). In HSV-1, 90 is non-essential
for viral replication in vitra whereas 98, highly conserved among
alphaherpesviruses, is required for virus attachment to host cells (Roizman, 1991).
In all MDV serotypes tested, expression of 90 is greatly reduced after 30-50 serial
passages in vitra (Bulow and Biggs, 1975; Churchill et al., 1969; lkuta et al.,
1983a; lkuta et al., 1983b; Nazerian, 1980; Purchase at al., 1971; Silva and Lee,
1984). Coincident with reduced gC expression, MDV serotypes 2 and 3 lose their
vaccine efficacy and serotype 1 becomes attenuated with respect to oncogenicity
(Calnek and Witter, 1991; Churchill et al., 1969; Kato and Hirai, 1985; Silva and
Witter, 1985). MDV gB expression does not change significantly after serial in vitra
passage (Silva and Lee, 1984).

MDV gC, an abundant viral glycoprotein antigen, is found in sera of infected

birds, infected cell culture fluids, and to a lesser extent associated with the plasma

121

 

an.
M

h

.‘

 

122

membrane of infected cells (lsfort et al., 1986). Evidence from pulse-chase
experiments and tunicamycin treated cells suggests that MDV 9C is synthesized
on the endoplasmic reticulum and processed through the Golgi via two pathways.
Secretion of N-Iinked glycosylated MDV gC occurs through a major pathway while
an unglycosylated form is secreted through a minor pathway (lsfort et al., 1986).
Approximately 95% of serotype 1 MDV 90 is secreted while the remaining 5% is
associated with plasma membranes of infected cells (lsfort et al., 1986). In
comparison, MDV 98 is found primarily in infected cell lysates and the glycoprotein
processing pathway is similar to that of other herpesvirus gB homologues (Chen
and Velicer, 1991b; Ross et al., 1989; Velicer et al., 1978).

Despite an apparent correlation between MDV gC expression and MDV
oncogenicity, isolation of naturally occurring non-oncogenic (serotype 2) MDV
which express gC and clonal isolates of oncogenic (serotype 1) strains which do
not express gC (Bulow and Biggs, 1975a; Bulow and Biggs, 1975b; Glaubiger et
al, 1983; Purchase at al., 1971), suggests that 90 does not directly affect the
oncogenic potential of MDV. However, in viva data indicates that attenuation of
MDV reduces the efficiency of lymphocyte infection (Schat et al., 1985).
Therefore, MDV gC may play a role in lymphocyte targeting. Regulation of MDV
90 may have important implications for virus-host interactions, and have important
consequences in the process of viral attenuation.

Although the gC-negative phenotype of attenuated MDV has been
documented for more than 20 years, the molecular basis for this alteration has

remained obscure. The inability of attenuated MDV to produce 90 may be due to

123

mutations or deletions of the MDV 90 gene or promoter. Alternatively, decreased
90 expression may result from alterations of MDV encoded or induced trans-acting
factors which interact with the 9C promoter. To distinguish between these
possibilities we investigated the mechanism by which MDV gC expression is
reduced during attenuation of the very virulent (v.v.) serotype 1 MDV strain Md11.
Protein levels, gene structure, steady state RNA levels, and transcription rates of
MDV gC from oncogenic and attenuated isolates of Md11 were determined.
Comparison of these data with similar studies on MDV proteins whose expression
is not altered during attenuation indicated that reduced expression of MDV 9C is
directly related to reduction in transcription rate of the MDV 96 gene in attenuated
Md11. Reduced transcription of the MDV 90 gene in attenuated Md11 are not due
to major structural alterations of the 90 promoter or coding sequence. Reduced
transcription rates and lack of any gross sthctural alterations in the attenuated
MDV gC gene suggests that MDV regulatory protein(s), that interact with the MDV
gC promoter are altered during attenuation of MDV strain Md11. These regulatory
proteins may be encoded within highly polymorphic regions of attenuated MDV
such as the BamHI-D, -H, and -L fragments located in repeat regions of the MDV

strain Md11 genome.

II. MATERIALS AND METHODS
Cells and virus.
Preparation, maintenance and infection of duck embryo fibroblasts (DEF)

have been described (Glaubiger et al., 1983). MDV strain Md11 has also been

 

124
described (Silva and Witter, 1985; Tieber et al., 1990; Witter et al., 1983).

Oncogenic MDV strain Md11 at passage level 9 (Md11p9) and attenuated Md11
at passage 80 (Md11p80) were kindly provided by the USDA. Avian Diseases
and Oncology Laboratory, East Lansing, Michigan. Md11p9 to p12 and Md11p80
to p83 are designated low passage (LP) and high passage (HP), respectively. All
infections were initiated with 7.0 x 10‘ PFU per 1x106 DEF in 100 mm tissue
culture dish. Because high passage MDV yields higher virus titers in culture than
low passage MDV, end point titers were determined at the time of harvest for

evaluation of protein, RNA, and transcription rates.

Plasmids and Radiolabeled DNA Probes.

MDV BamHl plasmids, derived from serotype 1 MDV strain GA (Fukuchie
et al., 1984), were a generous gift from M. Nonoyama (Showa University
Research Institute for Biomedicine in Florida, St. Petersberg, FL). BamHl-l3
contains coding sequence of the MDV 93 gene (Chen and Velicer, 1991b; Ross
et al., 1989). BamHI-K2 contains coding sequence for the MDV TK gene (Ross, .
personal communication; Scott et al., 1989). p19MDA2.35 contains the promoter
and complete coding sequence of the MDV gC gene (Coussens and Velicer,
1988). lzlE is a cDNA clone containing partial coding sequence of an immediate
early gene located within genomic repeat regions (Hang, manuscript in
preparation). pA/X6.1 was derived from the MDV BamHl-A clone and contains the
MDV ICP4 gene (Coussens, manuscript in preparation). pAl contains the chicken

B-actin gene (Cleveland at al., 1980). pBR322 was obtained commercially

 

125

(Bethesda Research Laboratories Life Technologies, Inc., Gaithersburg, MD). Full
length cell-free MDV genomes were isolated as described (Wilson and Coussens,
1991) with the following modifications: 0.8% gels were electrophoresed for 16
hours at 180 volts, 50-90 sec. ramp. Viral bands were excised and electroeluted
in an IBI electroelution apparatus, as per manufacturer's specifications (IBI, New
Haven, CT). 5’gC and 3’gC gene fragments were isolated from p19MDA2.35. The
5'gC fragment spans nucleotides # 300 to 735. The 3’gC fragment spans
nucleotides It 736 to 2350. or-“P labeled DNA probes were prepared using a
random prime DNA labeling kit (Boehinger Mannheim, Indianapolis, IN). All
radiolabeled DNA probes were prepared from gel purified inserts of plasmids

described above.

Immunoprecipitations.

Cells infected with either LP or HP Md11 and uninfected control cells were
labeled for 16 hours with 50 uCi/ml of [”Sjmethionine (EXPRE3533‘S protein
labeling mix, 1232.7 Ci/mmol, NEN, Boston, MA) 72 hours post infection.
Collection of cells and immunoprecipitations were performed as described (lsfort
et al., 1986). Immune chicken sera (ICS) and rabbit antisera against MDV gC
(RorA) were generous gifts from L. F. Velicer (Michigan State University, East
Lansing, MI). Both sera have been described previously and have been used
extensively to characterize MDV antigens (Glaubiger et al., 1983; lsfort et al.,
1986; Long et al., 1985; Silva and Lee, 1884). CPM/ul of 3"’S-Iabeled Iysate or

media were used to normalize sample volumes between uninfected, high and low

 

 

126
passage MDV strain Md11.

Immunoprecipitations were analyzed using 10% denaturing polyacrylamide
gels (Laemmli, 1970), electrophoresed in tris-glycine buffer, fixed and enhanced
using Amplify (N EN, Boston, MA) according to manufacturer's specifications. Dried
gels were exposed to X-ray film (Kodak, Rochester, NY) for 24 hrs to 6 days at -
70°C. “C-Iabeled protein standards were used as molecular size markers

(Bethesda Research Labs, Bethesda, MD).

DNA extractions and restriction fragment length polymorphism (RFLP)
analysis.

LP and HP Md11 infections of DEF were allowed to progress until cell
monolayers appeared fully infected (4-10 days). Total infected cell DNA was
isolated as described (Sambrook et al., 1989). The amount of viral DNA in total
infected cell DNA was determined by slot-blot assays as described (T ieber et al.,
1990). Similar amounts of LP and HP viral DNAs were digested with excess
restriction enzymes for 16 hrs according to manufacture’s specifications,
electrophoresed through agarose gels and transferred to nylon membranes in 0.4
N NaOH as described (Budowle and Baechtel, 1990). Radiolabeled probes were
hybridized to nylon filters at 65°C as described (Budowle and Baechtel, 1990).
Filters were exposed to X-ray film for 30 min to 48 hrs at room temperature.
Radiolabeled lambda DNA digested with Hind III or BsiEll was used as size

markers.

 

 

127
RNA extractions and Northern blots.

Total RNA from LP and HP Md11 infected and uninfected cells was isolated
96 hours post-infection using the proteinase K/SDS method (Sambrook et al.,
1989) followed by DNAse l digestion. RNA samples were treated with DNAse I
and 20 ug of total RNA was loaded per well, electrophoresed through
formaldehyde gels and transferred to supported nitrocellulose (Schleicher and
Scull, Keen, NH) as described previously (Sambrook et al., 1989). Radiolabeled
probes were hybridized to filters using high stringency conditions (Sambrook et al.,
1989). RNA band sizes were determined by comparison to RNA standards
(Bethesda Research Labs, Bethesda MD) run on the same gel. RNA blots were

stripped of probe DNA and hybridized with chicken B-actin gene specific probes.

Nuclear run-off transcription assays.

Nuclei were collected 96 hours post-infection from uninfected and LP or HP
Md11 infected DEF. Labeled RNA transcript probes were prepared and hybridized
to nitrocellulose (Schleicher and Schull, Keene, NH) strips as described (Pulaski
et al., 1992; Stewart of al., 1987). Each nitrocellulose strip contained triplicate
spots (5 pg each) of linearized plasmids which contained the genes for MDV 90,
93, TK, 5’90, 390, chicken B-actin, pBR322, and cell-free MDV strain Md11
genomes. Specific plasmids are defined in the Plasmid and Radiolabeled Probe

section.

1 28
Data analysis.

Intensity of bands on autoradiographs were analyzed on a FB910
Densitometer using Zeineh 1-D Videophoresis ll software (FisherBiotech). Graphs
were generated using Sigma Plot version 3.1 (Jandel Scientific, Corte Madera,
CA). Where appropriate, statistical analysis was performed with ANOVA using the

general linear models procedure of SAS (1985).

III. RESULTS
Comparison of protein levels from oncogenic and attenuated MDV strain
Md11.

A decrease in MDV gC protein levels after serial in vitro passage is well
documented for many MDV strains and serotypes (lkuta et al., 1983a; lkuta et al.,
1983b; Silva and Lee, 1984). To determine if reduction in 90 is also observed in
v.v. MDV strain Md1 1, immunoprecipitations were performed using media and cell
Iysates from DEF infected with oncogenic (LP) or attenuated (HP) MDV strain
Md11. Immune chicken sera (ICS) contains polyclonal antibodies reactive to gB
(gpt 00/60/49), p79 (Silva and Lee, 1984), and 9C (gp57-65). The four most
predominant proteins visible in immunoprecipitations with ICS from infected cell
lysates are gB (gp100/60/49) and p79 (Chen and Velicer, 1991b; Silva and Lee,
1984). Analysis of optical densities from autoradiographs indicated that similar
amounts of 9B and p79 were detected in LP and HP Md11 infected cell lysates,

suggesting that relative amounts of these proteins are not significantly affected by

 

 

129
attenuation of Md11 (Figure 1A: A 4-fold higher end point titer for HP Md11

resulted in higher protein amounts in the H lane.) MDV gC was not detected in ICS
immunoprecipitates of media from HP Md11 infected cells (Figure 1A). In contrast,
significant amounts of MDV 90 protein were detected in immunoprecipitates of
media from LP Md11 infected cells (Figure 1A). To determine if an accumulation
of intracellular 90 occurs in HP Md11, rabbit antisera against purified MDV gC
(RorA) was used to specifically immunoprecipitate 90 from both media and cell
Iysates. The rabbit antisera prepared against gC will also recognize DEF proteins
(gC was purified from DEF infected cells for rabbit challenges). Thus, DEF specific
bands are visible in all lanes (Figure 13). Levels of MDV 90, associated with
infected cells, were 9-fold lower in HP Md11 infected cells than in LP Md11
infected cells (Figure 1B). In addition, comparison of RorA immunoprecipitations
of media from HP and LP infected cells indicated a 20-fold reduction in QC protein
levels of HP relative to LP infected cell media. These results are consistent with
data obtained using ICS, and data for other MDV strains examined previously
(Chen and Velicer, 1991 b; Glaubiger et al., 1983; lsfort et al, 1986; Silva and Lee,

1984).

RFLP analysis of LP and HP MDV strain Md11.

Although both ICS and RM antisera contain polyclanal antibodies reactive
to many continuous and discontinuous epitopes on the 90 molecule, gross
structural alterations of the MDV gC gene in HP Md11 may cause dismption of

major antigenic epitopes resulting in an inability to react with ICS or RorA. To

 

rv-“—‘~1'

 

130

determine if Md11 gC gene structure was altered during attenuation, restriction
fragment length polymorphism (RFLP) analysis were performed. Comparison of
restriction fragments from LP and HP Md11 DNA by enzymes which cut outside
flanking regions of the Md11 90 gene locus (EcoRl, Bgll, Smal, Hind III, and Xhol)
indicate no alteration or polymorphism of 9C flanking regions in HP Md11 (fig.
2B).

Evaluation of restriction sites within the Md11 gC gene locus were based
upon frequency of cutting, within the GA strain gC gene (Coussens and Velicer,
1988), of a variety of enzymes (Figure 2A). Comparison of BamHl, Smal and
Hindlll restriction fragments of LP MDV strain GA and LP Md11 DNA were
identical (data not shown), suggesting that overall structure of LP Md11 and LP GA
DNA are very similar. A total of 27 restriction sites spanning the 90 gene were
assayed using this strategy. This type of analysis can detect changes at the level
of 10-50 nucleotides. No polymorphisms were detected within the 9C gene locus
of HP Md11 relative to LP Md11 (Figure 23). Data indicated that no gross
structural alterations outside of, or within, the Md11 gC gene locus occurred during
attenuation of Md11. These data are consistent with previous reports that 90
protein structure was unchanged after attenuation of MDV (Van Zaane et al.,
1982).

Additional blots were probed with full length MDV genomic probes. For
each enzyme used (same as above) multiple changes in banding patterns of HP
viral DNA were observed relative to LP viral DNA. All restriction fragments which

were altered in HP Md11 DNA map to repeat regions which flank the unique long

131

region of MDV DNA (data not shown). These results were consistent with previous

analyses of genome structure changes in HP MDV (Silva and Witter, 1985; Wilson

and Coussens, 1991).

Steady State 90 RNA levels In LP and HP Md11.

To evaluate expression of Md11 gC at the RNA steady state level and
determine if steady state RNA levels of Md11 gC were altered in HP Md11 infected
cells, northern blots of total RNA isolated from LP and HP Md11 infections were
probed with a full length gC gene fragment. Strong hybridization to a 1.8 kb
transcript was clearly evident in RNA from LP Md11 infected cells (Figure 3).
Weak hybridization to a larger transcript, which is proposed to be a read through
message of 9C (Coussens and Velicer, 1988), was also visible. Analysis of optical
densities from autoradiographs indicated a 20-fold reduction in steady state 90
RNA in HP infected cells relative to LP Md11 RNA (Figure 3, Table 1). After
prolonged exposure of filters to X-ray film, no size differences between LP and HP
Md11 90 RNA were observed (Figure 3).

To determine if reduction in steady RNA is a general phenomenon of
attenuation or specific to gC RNA, additional northern blots of RNA isolated from
LP and HP Md11 infected cells were probed with 98, ICP4, and IZIE gene
fragments. Analysis of optical densities from autoradiographs indicated a 2-fold
reduction of 9B RNA in HP Md11 infections. Steady-state RNA levels of the
immediate early genes, ICP4 and lzlE, were also slightly lower in HP Md11 relative

to LP Md11 (Table 1).

 

132

Nuclear run off transcription assays.

To test whether the rate of transcription across the 9C gene is reduced in
HP Md11 infected cells, nuclear run off transcription assays were performed using
nuclei isolated from uninfected, and Md11 LP or HP infected DEF. For each gene
tested, differences in transcription rates between LP and HP Md11 genes were
statistically significant. Transcription rate of the 9C gene in HP Md11 infected cell
nuclei showed the most significant decrease, 6-fold lower than that in LP Md11
infected cell nuclei (f-test, p=0.0001) (Figure 4). In addition, comparison of 5’ and
3’ 9C transcription rates in LP and HP Md11 infected cell nuclei indicated a
significant decrease in 5’ (p=0.0001) as well as 3' (p=0.0001) transcriptional
efficiencies in HP Md11 infected cell nuclei. 9B and TK transcription rates were
2.4- (p=0.002) and 2.5- (p=0.04) fold lower, respectively, in HP Md11 than in LP
Md11 infected cell nuclei. Analysis of total genome transcription revealed a 1 .3-
fold decrease (p=0.02) in HP Md11, relative to LP Md11 infected cell nuclei.

To determine if premature termination of transcription was occurring in HP
Md11 infected cell nuclei, transcription efficiencies of the first 485 bp (5’90) and
downstream 916 bp of 90 RNA were tested separately. Theoretically, if premature
termination of transcription was absent, % transcription efficiency
[%TNS=(O.D.IO.D. 5'gC +3'gC )x100)] would be 35% for 5’gC and 65% for 3’gC.
If premature termination of transcription was present, %TNS of 5’ 90 would be
greater than 35% and %TNS for 3’gC would be less than 65%. In LP Md11
infected cell nuclei, %TNS for 5'gC and 3’gC was 31% and 69% respectively. In

HP Md11 infected cell nuclei, %TNS for 5'gC and 3'gC was 35% and 65%

133

respectively. These data indicated that premature termination of transcription of

the 90 gene did not occur in HP Md11 infected cell nuclei.

IV. DISCUSSION

Attenuation of oncogenic MDV by serial passage in culture offers a unique
opportunity to examine mechanisms involved in herpesvirus induced neoplasia and
viral attenuation. Changes in MDV genome structure, coincident with attenuation,
are well documented. In particular, expansion of a 132 bp repeat sequence within
the BamHl-H and -D fragments and a 200 bp deletion in the BamHl-L fragment (all
located within repeat regions flanking the unique long) may be important for viral
attenuation (Bradley et al., 1989; Chen and Velicer, 1991a; Maotani et al., 1986;
Wilson and Coussens, 1991).

Additional changes within the MDV genome, coincident with attenuation
have not been documented. It has been assumed that reduced expression of
MDV 90 during attenuation results from changes in genome structure within or
near the MDV gC gene locus. Results presented in this report indicate that, in
attenuated MDV strain Md11, no gross structural alterations in genome structure
occur within or near the MDV gC gene. In addition, RFLP analysis of DNA from
MDV strains JM (serotype 1) and SB-I (serotype 2) indicate no stmctural alteration
in the MDV gC gene locus of attenuated isolates (Wilson and Coussens,
unpublished data). Thus, it seems unlikely that loss of MDV gC expression in
attenuated MDV is due to gross structural alterations within or near the MDV gene

locus. Minor changes in nucleotide sequences, however, cannot be completely

 

134

ruled out on the basis of RFLP analysis presented in this report.

Given the similarity of MDV gC gene structure in HP and LP Md11, it seems
likely that MDV gC protein expression in HP Md11 was reduced by another
mechanism. Possible mechanisms include: gC promoter mutations, reduced
protein or RNA stability, altered translational efficiency or mRNA transcription.
Analysis of RNA from cells infected with HP and LP MDV strain Md11 indicated
a decrease in steady state levels of the 1.8 kb 90 transcript in HP Md11 infected
cells, suggesting that reduced MDV gC expression results from a mechanism
affecting RNA stability or transcription efficiency.

To distinguish between mRNA stability and transcription efficiency, nuclear
run off assays were performed. Nuclear run off assays using various M DV genes
suggests that HP Md11 may exhibit an overall reduced efficiency of gene
transcription, relative to LP Md11. Although 98 protein levels in HP Md11 infected
cells were similar to 98 protein levels of LP Md11 infected cells, a 2.5-fold
decrease in gB transcription rates was observed in HP Md11 infected cell nuclei.
A similar decrease in steady state 98 RNA in HP Md11 infected cells suggests that
decreased 98 transcription in HP Md11 infected cell nuclei was responsible for
decrease in steady state gB RNA in HP Md11 infected cells. Discrepancies
between gB protein levels and 9B transcription rates was due to higher virus titers
in HP Md11 infected cells used for immunoprecipitation studies, thus higher viral
protein levels resulted from these infections, and protein level differences were not
readily apparent. A 6-fold decrease in MDV gC transcription rate suggests that

transcription of the MDV gC gene is adversely affected by serial passage of MDV

 

135

strain Md11. Our nuclear run off assays utilized a full length MDV gC probe as
well as 3' and 5' regions of the MDV gC gene. Thus, even prematurely thncated
gC mRNA species would have been detected. A 6-fold decrease of the HP Md11
gC transcription rate did not account for the 20-fold decrease observed for 90
protein and 9C steady state RNA levels in HP Md11 infected cells. This
discrepancy suggests that decreased RNA stability may also contribute to
decreased gC expression in HP Md11 infected cells. Reduced gC transcription
rates observed in HP Md11 infected cell nuclei then suggest that reduced 90
expression was related to alteration of 9C RNA stability as well as alteration of 90
gene transcription rates. These alterations would be consistent with minor
mutations within the 90 promoter or coding sequence as well as perturbation of
an essential MDV encoded regulatory factor which interacts with the 90 gene
promoter. The overall reduction in gene expression in HP Md11 infected cells,
relative to LP Md11 infected cells further supports the hypothesis that a critical
regulatory element is perturbed in HP Md11.

The model that alteration of MDV encoded transcription factors affect
reduced expression of 90 in HP Md11 is consistent with previous reports that
transactivation of an exogenous RSV promoter-CAT construct by HP MDV is 3-fold
lower relative to LP Md11 (T ieber et al., 1990). Preliminary results from our
laboratory indicate that HP Md11 is unable to transactivate expression of the CAT
gene linked to 3 LP MDV 90 promoter.

Although RFLP analysis of the 9C gene locus suggests a lack of gross DNA

structural alterations, there may be single nucleotide changes in the HP Md11 gC

 

136

gene locus which affect promoter activity causing a reduction in QC transcription
rates. Alternately, well characterized mutations within distinct regions of the HP
Md11 genome (Bradley et al., 1989; Chen and Velicer, 1991a; Hirai et al., 1981;
Maotani at al., 1986; Silva and Witter, 1985; Wilson and Coussens, 1991) may
adversely affect factors important to MDV 90 gene expression. Single nucleotide
changes in the HP Md11 gC coding region are unlikely since 1) low levels of
unaltered 90 protein, recognized by RorA and ICS, are present in HP Md11. Van
Zanne et al. (1982) has established that the decrease in 90 protein levels is a
quantitative as opposed to a qualitative difference, 2) low levels of full length gC
RNA are present, and truncated RNA were not detected, in HP Md11 infected
cells. Transcriptional efficiency of 5’gC and 3'gC in attenuated Md11 are the same
as theoretical values and 3) RFLP analysis strongly suggests that the 90 gene
locus is unaltered in HP Md11. Ideally, a direct comparison of nucleotide
sequence of the LP and HP Md11 90 gene would mle out any contribution of
nucleotide base changes in reduction of 90 expression in HP Md11. However,
due to the tightly cell-associated nature of MDV and presence of heterogeneous
vims populations in HP, attenuated MDV (T anaka, et al., 1980), DNA clones
containing the HP 90 gene locus may not represent the attenuated phenotype.
[A few plaque isolates of HP MDV strain JM express high levels of gC (Zalinskis
and Coussens, unpublished abservations)].

Our suggestion that MDV encoded transcription factors are altered during
in vitro attenuation suggests several hypothesis. Alteration of MDV encoded

transcription factors may be responsible for loss of oncogenicity as well as

 

final. '

 

137

contribute to reduced 90 expression in attenuated serotype 1 MDV. In addition,
non-oncogenic serotypes 2 and 3 MDV may produce a slightly different
transcription factor which is non-oncogenic but functions to enhance gC
transcription. Lastly, the putative transcription factor responsible for efficient
expression of 90 may be encoded within viral repeat regions known to be altered
during serial in vitro passage of MDV.

Elucidation of the precise mechanism responsible for decreased
transcription of the 90 gene in attenuated Md11 must await development of
efficient methods to purify clonal isolates of MDV, efficient systems to produce site-
directed MDV mutants, and characterization of M DV encoded transcription factors.
Isolation and characterization of MDV encoded transcription factors and cellular
factors induced by MDV will have a significant impact on our understanding of

MDV attenuation and possibly on MDV-induced neoplasia.

 

I
r.
i

 

138

ACKNOWLEDGMENTS

We thank Jessica Marcus for excellent technical assistance. We also thank
S. J. Triezenberg for careful review of this manuscript and gratefully acknowledge
Telmo Oleas and John Gill for helpful discussions and assistance with statistical
analysis. This work was supported by the Michigan Agricultural Experiment
Station, the Research Excellence Fund, State of Michigan, and by Grant Nos. 88-
37266-3983, 90-34116-5329, and 92-37204-802 awarded to P. M. Coussens
under the Competitive Research Grants Program administered by the US.
Department of Agriculture. M. Wilson was supported in part by a Michigan State

University Biotechnology Research Fellowship.

 

139

Figure 1. Immunoprecipitation mass-labeled viral proteins from cell lysates and
media of DEF infected with oncogenic or attenuated MDV strain Md11.
Immunoprecipitation, denaturing acrylamide gel electrophoresis and flourography
were performed as described in Materials and Methods. Antisera used were
immune chicken sera (ICS) (Panel A) and rabbit anti-MDVgC (RorA) (Panel B).
“C-labeled protein standards were Myosin-Heavy Chain (200 kDa), Phosphorylase
B (97.4 kDa), Bovine Serum Albumin (68 kDa), Ovalbumin (43 kDa), and Carbonic
Anhydrase (29 kDa). Location of standards are indicated to the right of each
panel. D: uninfected DEF, L: low passage (oncogenic) MDV strain Md11 infected
DEF, and H: high passage (attenuated) MDV strain Md11 infected DEF. Bands
corresponding to previously characterized MDV proteins are indicated by dots to
the left of each panel.

140

ONV

QOV

3v
89 I I L I

r... 4 OZ...— ._ o—

ooar 3’»
AV. 9%

<8: .m—

 

37 f!
. ‘ BIG

 

 

 

.— 3.5m."—
auv
a?! oovnn
I a.
QC!
2..
NOV ”GO—hm
ocuv

_I ._ OZ: 4 n:

GI. 92%.

mo. .<

 

141

Figure 2. RFLP analysis of the MDV strain Md11 gene locus and flanking regions.
Panel A: Location of restriction sites within the MDV gC gene. Bold horizontal line
illustrates the MDV gC gene, sizes are in base pairs. Restriction sites are
numbered from the BamHI site. Panel B. Autoradiographs from Southern transfer
analysis of 0.8% (left) and 1.2% (right) agarose gels. Total DNA from DEF
infected with oncogenic low passage (L) and attenuated, high passage (H) MDV
strain Md11 was isolated, digested with enzymes indicated, electrophoresed,
transferred to nylon membranes, and hybridized to a radiolabeled full length MDV
gC gene fragment as described in Materials and Methods. Enzymes are indicated
above each low and high passage pair. Fragment sizes (kbp) of radiolabeled
BstEll or Hind Ill digested lambda DNA standards are indicated by arrows to the
right of each autoradiograph.

Ins.
line
are
sler
)EF
IDV
;ed.
IDV
tled
aled
the

142

 

o 250 500 750 1000 1250 1500 1750- 2000
CAP ATG
BamHl 5‘3“
I r IF’ I I l I I l I
310
H II | |
pa 399 818 1536
Hae III I I I
109 1031 1095
Hinfl I I I
347 812 854 1243 1616 1654 1925
Fokl I I
262 434 564 1086 1393 1745
Hhal | l
76 455 457 790 1492
Clal I
1543
TOTAL II I IIII I III III | I IIIII I |
B.
\
\ \ \\ \ \ \\ \\‘ \ \ \
\ \ \ \
o \ 0 .(P o e o o o e 1- 0
4° 9° 9“ x» o‘ 9‘ +“ 9° 3° 8 4° 9“

 

'L H' T117 r1. Hi Ii. 117 IL 11' ll. HI

.- ‘1 I’ Q - .1

11

‘. I. .’

423.13

49.42

-. 46.56

44.36

C.

-4232
. -.<2.03

I Figure 2. ‘

 

 

IL HI IL HI IL HI IL HI IL HI IL "1

I.
s ' . * .
'. I
. . -1
‘ _ u
, _ ‘ , v

       

-- up - 143.68
42.32
41.93
0. t
41. 7
, .13.
" - up- - -
- - g ”40.70
a... 3“}: .' a?
:- fl
0' Q t“ , n
.2 1"
40.22
_<o.12

 

 

 

143

Figure 3. MDV strain Md11 gC steady state RNA levels in DEF infected with
oncogenic and attenuated isolates. Uninfected DEF RNA was included as a
negative control. Total RNA was isolated, electrophoresed, transferred to
supported nitrocellulose, and hybridized to a MDV 90 gene fragment as described
in Materials and Methods. D: uninfected DEF RNA, L: Low passage MDV strain
Md11 infected cell RNA and H: High passage MDV strain Md11 infected cell RNA.
The primary (1.8 kb) MDV gC transcript is indicated by an arrow to the right.

 

F iiiiiii

 

145

Table 1. Abundance of viral transcripts in cells infected with LP and HP MDV
strain Md11. RNA isolations, electrophoresis, northern transfers, and
hybridizations were as described in Materials and Methods. Optical densities were
normalized to chicken B-actin RNA levels.

 

146

Table 1.
Abundance of viral transcripts in cells infected with LP and HP MDV strain
Md11. , --‘

TRANSCRIPT SIZE ONCOGENIC ATTENUATED FOLD

(KB) Md11p14 Md11p83 DECREASE
IDV gC 1.8 0.7 0.03 20
and
were
98 2.7 1.6 0.6 2
ICP4 4.5 1.1 0.8 1

 

1,15 1.6 0.8 0.4 2

 

 

 

 

147

Figure 4. Transcription rates of MDV strain Md11 genes. Infection of DEF with
low and high passage MDV stain Md11, isolation of nuclei from infected and
uninfected cells, nuclear am off transcription assays, and hybridizations were
performed as described in Materials and Methods. All data points for Md11 gene
transcription rates were plotted relative to B-actin gene transcription rates. Error
bars represent standard errors of the means, n=3. Genome: full-length cell-free
Md11 viral DNA. TK, gB, gC, 5’90, and 3’gC are described in the text.

 

148

 

 
    

 

II 6

oo , 9‘

3‘3 /ID

:5;

2: 0
or

7”

 

IEF with
ted and
ns were
11mm
sEmr
wMfi

O

557/////////./////////////A°’

Figure 4.

 
  

557////////////////M

 

TK

   

:55:5/////////////////.

   

5{’////////////////////A

 

TRANSCRIPTION RATES OF MDV STRAIN Md11 GENES

GENOME

 

,, Nliov '0'0/‘0'0

 

 

 

Chapter 5.

DISCUSSION

 

 

149

I. SUMMARY OF RESULTS AND CONCLUSIONS.

Attenuation is a process resulting in the reduced ability of a virus to cause
disease. Attenuation is accomplished by serial passage of viruses in an unnatural
host (most commonly cell culture). During passage in vitra, mutations accumulate
in viral genomes. Viruses that develop beneficial mutations which confer an
advantage during growth in an unnatural host will outgrow the parental, wild type
virus. This is the process of natural selection at the most basic level. When these
mutant viruses are reintroduced into their natural host, their ability to cause
disease is diminished. The universal molecular basis for viral attenuation is
genomic mutations. All attenuated viruses contain genome mutations. However,
the molecular mechanisms of viral attenuation, or how the genomic mutations alter
viral pathogenesis is not well understood. Determining pathogenic significance of
mutations in attenuated viral genomes is complicated by: 1) Not all genomic
mutations in attenuated viruses are related to viral pathogenicity. 2) Many
mutations are silent, leading to neutral amino acid changes or no change at all in
the affected proteins. Attenuating mutations which occur in regulatory regions may
be overlooked. 3) Model systems to assess biological significance of attenuating
mutations must be used for human viruses of interest. Thus, dependence on
model systems makes it difficult to extrapolate data to the natural host. 4)
Genomes of many pathogenic viruses are extremely large and genome analysis
is bulky and tedious. For most large vimses studied (fowlpox and herpes),

analysis of attenuation and virulence determinants is limited to restriction enzyme

150

 

 

 

151

mapping and in viva characterization of site-specific mutants. These types of
analysis cannot rule out single base changes which may have a significant affect
an attenuation of viral pathogenesis.

Investigation of the attenuating mechanisms of Marek's disease virus would
aid in understanding the process of herpesvirus induced oncogenicity as well as
enhance our ability to develop more efficacious vaccines, especially against v.v.
MDV strains. In addition, using MDV to study viral attenuation eliminates the
reliance on model systems. MDV itself can be considered a model system for
other herpesviruses, as attenuation studies can be conducted in MDVs natural
host. However, the genome of MDV is large, and its cell-associated growth
properties creates technical difficulties when attempting to analyze genome
structure and growth parameters.

Thus, development of alternative procedures to obtain purified, intact, and
infectious MDV DNA is of critical importance to the molecular analysis of MDV
attenuation. In chapter 2, procedures developed to recover infectious cell-free
MDV genomes using contour-clamped homogeneous electric fields (CHEF)
electrophoresis were described. CHEF was used to investigate genome structure
differences among and within MDV serotypes, and after attenuation of various
MDV strains. Results indicate that oncogenic serotype 1 cell-free MDV genomes
are 17-18 Kbp larger than non-oncogenic serotype 2 and 3 cell-free genomes and
that the genome of attenuated serotype 1 MDV strain JM is 10 Kbp smaller than
its oncogenic counterpart. Extra genetic material in the larger genomes of

serotype 1 may code for factors involved in MDV-induced lymphoid proliferation.

 

 

 

152

Presently it is unclear if large deletions observed in attenuated MDV strain JM is
a general phenomena of attenuation or specific to JM. A recent report indicates
that retroviral integration into the JM genome occurred as a result of accidental
virus stock contamination (lsfort et al., 1992). Thus, the integration event may
have resulted in deletion of sequences in the JM genome which are not related to
the attenuation process.

In contrast to serotype 1 MDV, high passage serotype 2 cell-free genomes
increased in size by 5 kbp (SB-1) and 3 kbp (281 M l/1) relative to their low passage
counterparts. Increased copy number of tandem repeats may contribute to
increased size of serotype 2 genomes after serial in vitro passage. Expansion of
approximately 15 repeats is observed in high passage SB-1(M. R. Wilson,
unpublished results). However, similar expansion of repeat sequences in the high
passage 281Ml/1 genome has not been observed. Instead, defective interfering
particles which contain concatemeric units of the region homologous to the
amplified region in high passage SB-1 was described (Carter and Silva, 1990).
This region was determined to be a functional origin of MDV replication (Camp at
al., 1991). The amplification of repeat sequences in oncogenic as well as non-
oncogenic MDV serotypes (with the exception of HP 281Ml/1), suggests that
amplification of repeats near the origin of MDV replication is a general phenomena
of attenuation.

Using CHEF purified cell-free Md1 1 p10 genomic probes, a 200 bp deletion
located in the IR, region of attenuated MDV strain Md11 DNA was identified. This

change is separate and distinct from the 132 bp amplified region of serotype 1

 

'U‘ ‘7

 

 

 

153

MDV. 132 bp amplification occurs in the BamHI-D and -H fragments (Silva and
Witter, 1985), and the 200 bp deletion was mapped to the BamHI-L fragment.

In chapter 3, the isolation of an attenuated Md11 variant which contains 132
bp amplified regions but has the wild type BamHl-L fragment, is described. This
attenuated variant was utilized to determine if deletions in the BamHl-L fragment
affect biological properties of attenuated Md11, and assign a biological function to
this region. Growth rates, viral DNA replication, and transcriptional activity of
oncogenic, attenuated, and the attenuated Md11 variant on chick embryo
fibroblasts were compared. In vivo studies were performed to determine if the
BamHI-L region contributed directly to oncogenicity. Results indicate that deletions
in the BamHl-L region of attenuated Md11 directly affected viral growth and
immediate early gene expression. However, comparison of data from the
attenuated Md11 variant indicates that the BamHl-L deletion is not solely
responsible for enhanced growth and DNA replication in attenuated Md11. Other
mutations such as the 132 bp amplified region could have contributed to enhanced
viral growth of attenuated Md11. In viva studies indicate that the BamHl-L deletion
does not directly contribute to MDV-induced oncogenicity. RNA transcription
analysis suggests that the deletion in the BamHI-L region may be involved in latent
infection of MDV transformed cells.

Another prominent characteristic of attenuated MDV is reduced expression
of 90. It has been assumed that reduced expression of MDV 90 during
attenuation results from changes in genome structure within or near the MDV gC

gene locus. Results presented in chapter 4 indicate that, in attenuated MDV strain

 

p], ‘v- T
I

 

 

 

154

Md11, no gross alterations in genome structure occur within or near the MDV gC
gene. Given the similarity of MDV 90 gene structure in HP and LP Md11, it
seems likely that MDV 90 protein expression in HP Md11 was reduced by another
mechanism. Possible mechanisms include: gC promoter mutations, a reduction
in protein stability, mRNA stability, translational efficiency, or mRNA transcription.
Data in chapter 4 indicates that alterations in MDV transcription factors which
interact with the 90 gene promoter or minor changes in the 90 promoter contribute
to reduced expression of 9C in attenuated Md11. The model that alteration of
MDV encoded transcription factors contribute to reduced expression of 9C in
attenuated Md11 is consistent with previous reports that transactivation of an
exogenous RSV promoter-CAT construct by attenuated MDV is 3-fold lower
relative to oncogenic Md11 (T ieber at al., 1990). Preliminary results from our
laboratory indicate that HP Md11 is unable to transactivate expression of the CAT
gene linked to a LP MDV gC promoter. Taken together, these results suggest
that attenuation of Md11 may perturb a viral gene regulatory element or
mechanism.

Survey of the literature in chapter 1 indicates that there may be a variety of
molecular mechanisms of viral attenuation. Attenuation of a viral induced disease
state can occur by disruption of the viral transcriptional or translational mechanism
and alteration of the interaction of tissue or cell-specific proteins with regulatory
regions of viral genomes. Data in this dissertation indicates that attenuated MDV
display alterations in the temporal expression of some IE genes, a general

reduction in transcriptional efficiency, and a general reduction in steady state RNA

 

 

155

levels. Alterations of the transcriptional apparatus suggests that alteration of MDV
encoded regulatory factors and mutations within promoter regions of MDV genes

contributes to attenuation of MDV-induced oncogenicity.

II. THEORETICAL RELATIONSHIPS OF MDV ATTENUATION AND MDV-
INDUCED ONCOGENICITY: WORKING HYPOTHESES TO BE TESTED.
The establishment of MDV-induced tumors is most likely a complex, multi-
step process. Since latent infection of T-Iymphocytes occurs before tumor
formation in viva, establishment and maintenance of latency may be a pre-requisite
for tumor formation. Vims encoded proteins which: 1) may function to get the vims
into a lymphocyte, 2) contribute to establishment or maintenance of latency, 3)
trans-activate transcription of host proto-oncogenes, or 4) stimulate the host cell
cycle resulting in continuous lymphoproliferation, are likely to be important for the
oncogenic potential of MDV. In addition, since suppressive influence of the host
immune system is involved in the establishment of latency, genes which normally
function in the lytic phase of MDV replication that interact with host cell immune
system proteins may also be involved in the pathway leading to MDV-induced
oncogenicity. Alternately, there may be a single direct acting oncogene encoded
within MDV DNA that is responsible for MDV-induced tumor formation.
Attenuation of MDV is defined by loss of MDV-induced oncogenicity. A
distinction must be made between attenuating mutations and silent mutations.
Presently, no in vivo data exist on the relationship between mutations in attenuated

MDV genomes and their role, if any, in MDV-induced oncogenicity. Therefore, we

 

‘.

 

 

 

156

must rely on what we have Ieamed during the course of investigations and
circumstantial evidence of others to hypothesize relationships between specific
mutations in attenuated MDV genomes and MDV-induced oncogenicity.
Hypothetical relationships between specific alterations of attenuated MDV and
MDV-induced transformation is summarized in Figure 1.

Reduction in the expression of gC is a prominent marker for MDV
attenuation. Because some gC negative mutants are able to cause tumors, 9C is
not considered a major determinant of MDV oncogenicity. However, virus stocks
of 90 negative mutants contain a sub-population of viruses which express gC
(Zalinskis and Coussens, unpublished results). Therefore, the role of 9C in MDV-
induced oncogenicity should be re-evaluated. Since attenuated viruses are unable
to spread by contact and display reduced efficiency for infecting or replicating in
lymphocytes in viva, and level of QC expression is different in chick vs. duck
embryo cells in vitro, gC may function in an epithelial cell-specific or lymphocyte
specific manner. If 9C is required to specifically infect lymphocytes, the lack of 90
expression would block this step in the pathway leading to transformation of
lymphocytes. Alternately, if 90 is not required for cell-specific MDV infection, then
the suggestion that reduced expression of 9C in attenuated MDV is related to
alteration of an MDV regulatory protein, indirectly implicates the putative regulatory
protein as playing a role in transactivation of cellular or viral proteins involved in
maintaining ongoing lymphocyte proliferation. Data from attenuation studies on
other viruses indicate that both hypothesis on the function of 90 are plausible.

Other mutations in attenuated MDV genomes are most likely involved in

 

 

 

 

157

MDV INFECTED CELL

  

 

 

132 BP
AMPLIFICATION
oNA
— REPLICATION
ac ?
LYMPHOCYTE
TARGETING
?
CYTOLYTIC PHASE “'—
LATENT PHASE
' a-T Ir
l 7.3"
(“fl (“3 MDV REGULATORY
3'1"; 4:1) PROTEINS:
K. ACTIVATION OF
(‘R (14‘ LYMPHOCYTE ~
Tu , Tu PROLIFERATION ?
\p-«‘ U
/ \ M. DELETION IN
‘1. t T \ Tu BamHI-L:
\Tg} K“ 2 Tu 4‘ LATENCY
"' " '1' REACTIVATION 7
Figure 1.

Hypothetical relationships between mutations of attenuated Md11 and the
pathway to MDV transformation of T-Iymphocytes.

 

 

 

 

158

separate steps in the pathway leading to MDV-induced transformation.
Amplification of 132 bp repeats has been strongly implicated as being involved in
attenuation of MDV-induced oncogenicity. Amplification of repeat sequences near
the origin of MDV replication in oncogenic as well as non-oncogenic MDV
serotypes (with the exception of HP 281Ml/1), suggests that this mutation may
directly affect viral DNA replication as opposed to cellulartransformation. Reduced
DNA replication during initial cytolytic infection in viva of attenuated serotype 1
MDV would supply an early block in the pathway leading to MDV-induced tumor
formation.

Significance of deletions in the BamHl-L region is suggested by results
presented in chapter 3. In addition, the BamHl-L region is expressed in MDV
transformed cell lines. In vivo data indicates that the region which encompasses
the BamHl-L deletion is not required for establishment of tumors. Therefore, it
may be involved in later stages of the MDV life cycle. Specifically, the BamHI-L
region may be required for reactivation of MDV replication during latent infection.
This suggestion is supported by the observation that a non-producer MDV
transformed cell line contains a tmncated RNA transcript encoded within the
BamHl-L region, and virus cannot be recovered from non-producer MDV

transformed cell lines.

lll. FUTURE RESEARCH DIRECTIONS.
Technique development.

Within the past five years there has been an explosion in the amount of

 

 

 

 

159

information on MDV molecular biology. Many genes have been defined and
mapped to the MDV genome. Testing the function of specific genes in vivo is now
possible. Development of efficient systems to produce recombinant MDV and
subsequent analysis of specific MDV mutants in vivo is important to assess the
pathological function of specific genes. Due to the cell-associated nature of MDV,
current technology to produce site-specific MDV mutants is inefficient. Although
various MDV recombinants have been constructed, the time and work required
was substantial. Procedures described in Chapter 2, in combination with using
selectable antibiotic marker genes, could be utilized in screening and identification
of recombinant MDV genomes and development of in vitro recombination systems.
Development of an in vitra recombination system may be feasible Since isolation
of infectious cell-free MDV genomes is now possible and most of the enzymes
involved in DNA replication have been identified and are commercially available.
It may simply require mixing cell-free MDV genomes with an insertion vector and
Rec A protein (promotes strand exchange between homologous sequences) and
subsequent transfection into chick or duck embryo fibroblasts. This may at least
enhance the recombination efficiency. In addition, isolation of purified populations
of attenuated MDV or naturally occurring mutants using procedures described in
Chapter 2 would be useful in further analysis of genomic regions altered during

attenuation, as exemplified by experiments described in Chapter 3.

MDV attenuation.

Regions of the MDV genome and genes affected during attenuation are the

 

 

 

160
132 bp repeats in the IR, , the BamHl-L fragment, meq, and 90. The ultimate task

is to determine if these mutations are related to MDV-induced oncogenicity, as
others have postulated. Presently there is no in vivo data which implicates
expansion of the 132 bp repeat sequences as a factor for reduced oncogenicity
of attenuated MDV. Constnrction of site specific MDV mutants and in viva analysis
of mutants would address this issue. Specific mutants to test would be oncogenic
viruses containing the amplified repeat region of attenuated or serotype 2 viruses.
Conversely, attenuated vimses containing the analogous repeat region from
oncogenic or non-oncogenic viruses could be tested. Similar types of analysis
could be performed with the meq gene; replacing or inactivating the meq gene and
subsequent analysis in viva would aid in understanding its function.

Further in vitro characterization of the BamHl-L deletion in attenuated MDV
strain Md11 is required. Nucleotide sequence analysis of the 1.0 kb Hintl sub-
fragment of BamHl-L from oncogenic, attenuated, and Md11p83-L would clarify
location of the deletion and identify the putative promoter effected. Isolation of
clones from oncogenic and attenuated Md11, and Md11p83-L may be the difficult
part of this analysis. Sequence analysis could be accomplished by utilizing
anchored polymerase chain reaction (PCR) amplification of cell-free MDV genomes
and sequencing resultant amplified products Since the deletion in Md11 was sub-
localized within the BamHl-L fragment and this mutation affects biological
properties of the virus, identification and characterization of the affected gene(s)
should be pursued. Transcripts homologous to the BamHI-L fragment are

expressed in MDV transformed cell lines. Thus, it would be important to determine

 

 

 

161

if the region encompassing the 200 bp deletion is expressed in MDV transformed
cell lines. In addition, comparison of the BamHI-L fragment and transcripts
encoded within the BamHI-L fragment of other oncogenic MDV strains to their
attenuated counterparts would be of interest to determine if deletions occurring in
this region are a general phenomenon of attenuation or specific to v.v. MDV
attenuation. In viva studies to determine the role of the BamHI-L region in MD
pathology should also be pursued. However, this may require isolating additional
mutants, using the procedures described in Chapter 2. Since the deletion in
BamHl-L occurred early in culture, it may be possible to isolate a low passage
MDV containing the BamHl-L deletion without amplified 132 bp repeats.
Comparison of the low passage variant to Md11p83-L, oncogenic and attenuated
Md11 in vivo, and subsequent analysis of resultant transformed cell lines, could
aid in understanding the pathological role, if any, of the BamHI-L region.

The only protein consistently associated with attenuation of MDV is 90. The
existence of 9C and its relation to attenuation have been documented for more
than 20 years. As described in chapter 4, molecular mechanism for reduced
expression of 9C was hypothesized to entail alteration of an MDV encoded
regulatory protein. However, single base changes in the 9C promoter region were
not completely ruled out. Therefore, additional in vitro experiments are required.
With the aid of procedures described in Chapter 2, sequence analysis of PCR
amplified promoter regions of oncogenic and attenuated MDV cell-free genomes
is possible. In addition, level of transactivation, by an exogenous MDV ICP4, of

the endogenous gC promoter in oncogenic and attenuated MDV infected cells

 

 

 

162

should be tested. Level of transactivation by ICP4 can be assayed by
immunoprecipitation of gC. This could further clarify the mechanisms involved in
reduced 90 expression in attenuated MDV. It would be of interest to pursue the
hypothesis that the altered regulatory protein responsible for 90 expression is
encoded within genomic regions affected by attenuation. This could be
investigated by transfecting low passage MDV DNA clones or fragments (which
encompass the regions altered in attenuated MDV) into cells infected with
attenuated MDV and analyzing 90 levels using immunoprecipitation. In viva
studies to investigate the function of 9C could be performed using site specific 90
negative mutants derived from oncogenic MDV. However, construction of such
mutants has been attempted previously (Wilson, unpublished results). Insertional
inactivation of 9C resulted in unstable MDV recombinants. It is possible that the
structural integrity of the QC loci is required for MDV replication. Therefore,
construction of intertypic recombinants may minimize disruption of the QC loci.
Since the HVT 90 gene has been characterized at the molecular level, the HVT
90 gene could be inserted into the serotype 1 9C gene loci, and vice versa.

These mutants could then be tested in viva.

Vaccine development.

Live MDV vaccines must maintain a high level of immunogenicity with very
low associated virulence. Ideally, a live vaccine which could be transmitted
horizontally would provide passive immunization to other members of a flock.

Alteration of the MDV genome in vitra to produce MDV vaccines may be feasible.

 

 

 

 

163

However, which regions to alter is not clearly understood. Therefore it is important
to analyze attenuated MDV and understand the underlying mechanisms involved
in MDV attenuation as well as elucidate the pathway leading to transformation by
MDV. For example, if 90 is required to specifically infect lymphocytes and is
required for horizontal transmission of MDV, a candidate "designer" vaccine could
contain a 90 gene under control of an epithelial cell-specific promoter. These
types of vaccines must await further understanding of MDV transformation and
attenuation.

Of immediate concern is development of vaccines against v.v. MDV.
Attenuated v.v. MDV strain Md11 is currently being evaluated as a vaccine.
However, previous reports indicate that the Md11 vaccine candidate is not as
efficacious as existing vaccines, cannot spread by contact, and is poorly protective
in chickens with homologous maternal antibodies. It is quite possible that the
Md11 vaccine contains a residual sub-population of oncogenic MDV. Using
procedures described in this dissertation, isolation of a purified population of
attenuated MDV is now feasible. This is exemplified by the isolation of Md1 1p83-L
(Chapter 3). In vivo oncogenicity tests with Md11p83-L indicate that this virus is
highly attenuated compared to its parent strain. In vitra studies indicate that
Md11p83-L is a purified papulation of MD11. Thus, the attenuated Md11p83-L
variant described in this dissertation is a potential vaccine candidate against v.v.

MDV. Testing the vaccine efficacy of Md11p83-L should be rigorously pursued.

 

 

 

 

 

l
I

APPENDIX

 

APPENDIX

DESCRIPTION OF METHODS.

General Information. Descriptions of basic molecular biology techniques
used for this dissertation research are found in Sambrook et al. (1989) where the
most extensively utilized laboratory procedures are discussed in detail. Three
additional procedures that were used for dissertation research will be briefly
described here.

CHEF. The CHEF technique is summarized in Figures 1 and 2.
References to CHEF can be found in Chapter 2. CHEF is an abbreviation for
clamped homogeneous electric fields electrophoresis. CHEF is a modification of
pulsed-field gel electrophoresis for the separation and analysis of DNA molecules
larger than 50 kbp. The CHEF apparatus contains a gel chamber with electrodes
arranged in a hexagon. A computer driven electrical pulse converter will cause the
polarity of electrodes to switch from + to - at pre-determined time intervals (Figure
1). Polarity switches constantly throughout the electrophoresis procedure. DNA
molecules move through an agarose gel in a snake-like motion termed "reptation".
Under electrophoresis conditions where electrodes are arranged in parallel array
(conventional), DNA larger than 50 kbp becomes intangeled within the gel matrix,
resulting in retardation of size dependent migration (Figure 2). Under CHEF
conditions, DNA molecules must change direction in response to changing polarity.
The continual turning motion of DNA molecules results in Size dependent migration

and resolution of DNA fragments larger than 50 kbp (Figure 2).

164

 

 

 

165

Immunoprecipitation of proteins. Reference to this technique can be
found in Chapters 3 and 4. Analysis and purification of proteins using
immunoprecipitation methods is illustrated in Figure 3. Proteins are metabolically
labeled with 35S-methionine (1). Cell lysates are prepared (2) and culture media
is harvested (3). Culture media is tested if the protein of interest was secreted
from cells. Cell lysates or culture media are then mixed with specific antisera or
specific monoclonal antibodies resulting in the formation of protein-antibody
complexes (4). Protein antibody complexes are then mixed with forrnalin-fixed
Staph A. Surface of the Staph A bacteria contains a protein called ”Protein A“.
Protein A will bind to the Fc portion of antibody molecules, Specifically IgG. After
protein-antibody complexes are mixed with Staph A, protein-antibody-Staph A
rosette-like complexes of will form (5). Cell debri is washed away by a series of
centrifugation and rinsing steps. The complex is then disassociated by heat
treatment (6). The disassociated mixture is analyzed on denaturing polyacrylamide
gels (7). After autoradiography, visible bands on X-ray film represent proteins that
were able to bind to the specific antibodies. [Prior to treatment of proteins with
specific antibodies, cell Iysates and culture media are preadsorbed to non-specific
antisera and treated with Staph A. This is done to remove cross-reactive, non-
specific proteins which may bind to specific antibodies]

Nuclear run off transcription assay. Reference to this technique is found
in Chapter 4. Nuclear run off transcription assays are used to quantify the in situ
amount of RNA produced from transcription of genes in intact nuclei. This method

is illustrated in Figure 4. Intact nuclei are isolated from cells of interest (1 ). Within

 

 

 

166

the nuclei, genes are being actively transcribed. Di-nucleotidetriphosphates (dATP,
dCTP, dGTP) and radiolabeled dUTP is mixed with nuclei. The dNTPs are
incorporated into the growing RNA molecule (2). Only RNA become radioactively
labeled by dUTP because dUTP does not become incorporated into DNA.
Radioactive RNA is then isolated (3) and used as an hybridization probe for slot
blot membranes (4). The slot blot membranes have the bound gene fragments of
interest. After autoradiography, transcription can be quantified by scanning x-ray

film images, from slot blots, with a densitometer.

 

 

167

CLAMPED HOMOGENEOUS ELECTRIC
FIELDS
ELECTROPHORESIS

CHEF

‘HEXAGONIAL ELECTRODE ARRAY

Am 804
+ + + + + +
’/—__—\" *r“““x\
’/ \" f/ @1 \\
// \X $/ \\
I
\ /x x\ /
*\ /$ x\ /
,\ A, ,,\ /,
X\_______/x x\______/$
+ T '1' + + + + +
[FLOW J'FLOW
NET MIGRATION

Figure 1.

 

 

 

168

RESOLUTION OF LARGE DNA MOLECULES

 

 

    

 

 

CONVENTIAL CHEF
>50 KB: SIZE INDEPENDENT >50 KB: SIZE DEPENDENT
MIGRATION MIGRATION Pr
.0... 0.... f
g Q Q o O O O O '
Q . . . REPTATION O . Q Q ‘
. .
o g g P O o o
o O o o o O o o
CONVENTIAL CHEF
- = = = = = = = C:
- _ - - -
_
-
_

 

 

 

 

 

 

Figure 2.

 

 

169

IMMUNOPRECIPITATION OF PROTEINS FROM
MDV INFECTED CELLS

p5S-methionine
16 hrs
ME IA /

MDV INFECTED CELLS ®
\ CE "I

LL LYSATE

ff} a:

SPECIFIC ANTISERA .
AQ/ I y
'1' 7* STAPHA

 

 

........

.........
22's.".

  

 

 

 

SPIN, RINSE, @
HEAT
@ F—l I——1 =

Ax°k ....
j“.

  

 

 

 

 

Figure 3.

 

170

 

NUCLEAR RUN OFF TRANSCRIPTION ASSAY

NRO

MDV INFECTED CELLS

/-

 

%- ,

dATP, dCTP, dGTP,
radiolabeled dUTP
ISOLATE RNA

W

 

 

ff

6)

SLOT BLOT

 

m o

HYBRIDIZATION

 

 

 

AUTORADIOGRAPHY AND DENSITOMETRY
Figure 4.

 

LIST OF REFERENCES

[Lt-r. .r
“d

 

LIST OF REFERENCES

Aihara,S., C. M. Rao, Yu-Yx, T. Lee, K. Watanabe, T. Komiya, H. Sumiyoshi, H.
Hashimoto, and A. Nomoto. (1991). Identification of mutations that occurred on the
genome of Japanese encephalitis virus during the attenuation process. Virus
genes. 5:95-109.

Anderson, A. S., A. Francesconi, and R. W. Morgan. (1992). Complete nucleotide
sequence of the Marek’s disease virus ICP4 gene. Virology 189:657-667.

Andino, R., G. E. Reickhof, D. Trono, and D. Baltimore. (1990). Substitution in the
protease (3PM) gene of poliovirus can suppress a mutation in the 5’ noncoding
region. J. Virology 64:607-612.

Bacon, L. D., R. L. Witter, and A. M. Fadly. (1989). Augmentation of retrovirus
induced lymphoid Ieukosis by Marek's Disease herpesvirus in white leghom
chickens. J. Virol. 63:504-512.

Binns, M. M. and N. L. J. Ross. (1989). Nucleotide sequence of the Marek's
disease virus (MDV) B-1 B A-antigen and the identification of the MDV A antigen
as the herpes simplex vims-1 glycoprotein C homologue. Virus Research. 12:371-
382.

Bishop, J. M. (1987). Trends in oncogenes. In Oncogenes and growth factors, (R.
A. Bradshaw and S. Prentis, eds.), Elsevier Science Publ. Co., New York.

Blair, E. D. and B. W. Snowden. (1991). Comparative analysis of the parameters
that regulate expression from promoters for two late HSV-1 gene products. In
Herpesvirus transcription and its regulation. (E. K. Wagner, ed.) CRC Press, Inc.,
Florida.

Bostock, C. J. (1988). Parameters of field inversion gel electrophoresis for the
analysis of pox virus genomes. Nucl. Acids Res. 16:4239-4252.

Bradley, G., G. Lancz, A. Tanaka, and M. Nonoyama. (1989). Loss of Marek’s
disease virus tumorigenicity is associated with truncation of RNAS transcribed
within BamHl- H. J. Virology 63:4129-4135.

171

 

172

Bradley, G., M. Hayashi, G. Lancz, A. Tanaka, and M. Nonoyama. (1989).
Structure of the Marek's Disease virus BamHI-H gene family: Genes of putative
importance for tumor induction. J. Viral. 63:2534-2542.

Buckmaster, A. E., S. D. Scott, M. J. Sanderson, M. E. G. Boursnell, N. L. J. Ross,
and M. M. Binns. (1988). Gene Sequence and mapping data from Marek's disease
virus and herpesvirus of turkeys: Implications for herpesvirus classification. J. gen.
Virol. 69: 2033-2042.

Budowle, B. and F. S. Baechtel. (1990). Modifications to improve the effectiveness
of restriction fragment length polymorphism typing. Appl. Theor. Electra. 1 :181-187.

Bulow, V. van, and P. M. Biggs. (1975). Precipitating antigens associated with
Marek's disease virus and a herpesvirus of turkeys. Avian Pathol. 4:147-162.

Bulow, V. van, and P. M. Biggs. (1975). Differentiation between strains of Marek’s
disease virus and turkey herpesvirus by immunofluorescence assays. Avian Pathol.
4:133-146.

Burkitt,D. (1962). A children’s cancer dependent upon climatic factors. Nature
(London) 194:232.

Cai, W. and P. A. Schaffer. (1992). Herpes simplex virus type 1 ICPO regulates
expression of immediate-early, early, and late genes in productively infected cells.
J. Viral. 66:2904-2915.

Calnek, B. W., K. A. Schat, L. J. N. Ross,W. R. Shek, and C. L. H. Chen. (1984).
Further characterization of Marek’s Disease virus infected lymphocytes. I. In vivo
infection. lntl. J. Cancer 33:389-398.

Calnek, B. W., and R. L. Witter. (1991). Diseases of Poultry: Marek's disease, pp.
342-385. (B. W. Calnek, H. J. Barnes, C. W. Beard, W. M. Reid, and H. W. Yoder,
Jr., eds.) Iowa State University Press, Ames, Iowa.

Camacho, A. and P. G. Spear. (1978). Transformation of hamster embryo
fibroblasts by a specific fragment of the herpes simplex virus genome. Cell 15:993.

Cann, A. J., G. Stanway, P. J. Hughes, P. D. Minor, D. M. A. Evans, G. C. Schild,
and J. W. Almond. (1984). Reversion to neurovirulence of live-attenuated Sabin
type 3 oral poliovirus vaccine. Nucleic Acids Res. 12:7787-7792.

Card, J. P., M. E. Whealy, A. K. Robbins, and L. W. Enquist. (1992). Pseudorabies
virus envelope glycoprotein gl influences both neurotropism and virulence during
infection of the rat visual system. J. Virology 66:3032-3041.

 

 

 

 

173

Carter, J. K. and R. F. Silva. (1990). Cell culture amplification of a defective
Marek’s Disease virus. Virus Genes 4:225-237.

Cebrian, J., C. Kaschka-Dierich, N. Berthelot, and P. Scheldrick. (1982). Inverted
repeat nucleotide sequences in the genome of Marek’s Disease Virus and the
herpesvirus of the turkey. Proc. Natl. Acad. Sci. USA 79:555-558.

Chen, X. and L. F. Velicer. (1991). Multiple bi-directional initiations and
terminations of transcription in Marek's disease virus long repeat regions. J. Virol.
65 :2445-2451 .

Chen, X. P. J. Sondermeijer and L. F. Velicer. (1992). Identification of a unique
Marek’s disease virus which encodes a 38 kilodalton phosphoprotein and is
expressed in both lytically infected cells and latently infected Iymphoblastoid tumor
cells. J. Viral. 66:85-94.

Chen, X. an L. F. Velicer. (1993). The Mark’s disease virus B antigen complex (gp
100, gp 60, gp 49) is the homolog of the herpes simplex virus glycoprotein B (98).
(in press).

Christodoulou, C., F. Colbere-Garapin, M. Macadam, L. F. Taffs, S. Marsden, P.
Minor, and F. Horaud. (1990). Mapping of mutations associated with
neurovirulence in monkeys infected with Sabin 1 poliovirus revertants selected at
high temperature. J. Viral. 64:4922-4929.

Chu, G., D. Vollrath, and R. W. Davis. (1986). Separation of large DNA molecules
by contour-clamped homogeneous electric fields. Science 234:1582-1585.

Churchill, A. E. and P. M. Biggs. (1967). Agent of Marek's disease in tissue
culture. Nature (London) 215:528-530.

Churchill, A. E., R. C. Chubb, and C. W. Baxendale. (1969). The attenuation with
loss of oncogenicity of the herpes type virus of Marek’s disease (strain HPRS-16)
on passage in cell culture. J. gen. Virology. 4:557-564.

Cleveland, D. W., M. A. Lopata, R. J. MacDonald, N. J. Cowan, W. J. Rutter, and
M. W. Kirschner. (1980). Number and evolutionary conservation of alpha- and
beta-tubulin and cytoplasmic beta- and gamma- actin genes using specific cloned
cDNA probes. Cell 20:95-105.

Collingham, K. E., T. M. Pollock, and M. O. Roebuck. (1978). Paralytic
poliomyelitis in England and Wales. Lancet i:976-977.

Coussens, P. M., V. L. Tieber, C. S. Mehigh, and M. Marcus. (1991). Identification
of a novel transcription factor, ACF, in cultured avian fibroblast cells that interacts

174

with a Marek's disease virus late gene promoter. Virology 185:80-89.

Coussens, P. M. and L. F. Velicer. (1988). Structure and complete nucleotide
sequence of the Marek’s disease herpesvirus gp 57-65 gene. J. Viral. 62:2373-
2379.

Coussens, P. M., M. R. Wilson, H. Camp, H. Roehl, R. J. lsfort, and L. F. Velicer.
(1990). Characterization of the gene encoding herpesvirus of turkeys gp 57-65:
Comparison to Marek’s disease virus gp 57-65 and herpes simplex vims
glycoprotein C. Virus Genes 32291-307.

Cui, 2., L. F. Lee, J-L Liu, and H-J Kung. (1991). Structural analysis and
transcription mapping of the Marek's disease virus gene encoding pp38, an antigen
associated with transformed cells. J. Viral. 65:6509-6515.

Dallo, S. and M. Estbam. (1987). Isolation and characterization of attenuated
mutants of vaccinia virus. Virology 159:408-422.

Elmubarak, A. K., J. M. Sharma, R. L. Witter and V. L. Sanger. (1981). Failure of
herpesvirus of turkeys to protect turkeys against Marek's Disease. Am. J. Vet. Res.
42:2122-2128.

Epstein, M. A., B. G. Achong, and Y. M. Barr. (1964). Virus particles in cultured
lymphoblasts from Burkitt’s lymphoma. Lancet i:702.

Evans, D. M. A., G. Dunn, P. D. Minor, G. C. Schild, A. J. Carin, G. Stanway, J.
W. Almond, K. Currey, and J. V. Maizel. Jr. (1985). Increased neurovirulence
associated with a single nucleotide change in a non-coding region of the Sabin
type 3 poliovaccine genome. Nature (London) 314:548-550.

Everett, R. D., C. M. Preston, and N. D. Snow. (1991). Functional and genetic
analysis of the role of VMW110 in herpesvirus replication. In Herpesvirus
transcription and its regulation. (E. K. Wagner, ed.) CRC Press, Inc., Florida.

Flint, S. J. (1982). Transformation by adenoviruses. In RNA tumor viruses 2""
Edition, Cold Spring Harbor laboratory, New York.

Fraenkel-Conrat, H., P. C. Kimball, J. A. Levy. (1988). Virology. 2"" Edition,
Prentice-Hall, Inc., New Jersey.

Fukuchi, K. M. Sudo, Y. -S. Lee, A. Tanaka, and M. Nonoyama. (1984). Stmcture
of the Marek’s disease virus DNA: Detailed restriction enzyme map. J. Viral.
51:102-109.

Fynan, E., T. M. Block, J. DuHadaway, W. Olson, and D. L. Ewart. (1992).

 

 

175

Persistence of Marek’s Disease virus in a subpopulation of B cells that is
transformed by avian leukosis virus, but not normal bursal cells. J. Viral. 66:5860-
5866.

Glaubiger, C., K. Nazerian, and L. F. Velicer. (1983). Marek's disease herpesvirus:
IV. Molecular characterization of Marek’s disease herpesvirus A antigen. J. Viral.
45:1228-1234.

Grodzicker, T. and N. Hopkins. (1981). Origins of contemporary DNA tumor virus
research. In DNA tumor viruses 2"6 Edition, Cold Spring Harbor laboratory, New
York.

Harris, A. and D. R. Bentley. (1988). Separation of episomal Epstein-Barr virus
from Burkitt’s lymphoma host cell DNA in pulsed field gels. Nucl. Acids Res. 16:
4172.

Hayward, S. D. and J. M. Hardwick. (1991). Epstein-barr virus transactivators and
their role in reactivation. In Herpesvirus transcription and its regulation. (E. K.
Wagner, ed.) CRC Press, Inc., Florida

Henderson, L. M., R. L. Levings, A. J. Davis, and D. R. Sturtz. (1991).
Recombination of pseudorabies virus vaccine strains in swine. Am. J. Vet. Res.
52:820-825.

Henle, G. and W. Henle. (1966). Studies on cell lines derived from Burkitt’s
lymphoma. Trans. N. Y. Acad. Sci. 29:71-79.

Hirai, K., K. lkuta, K. Maotani, and S. Kata. (1984). Evaluation of DNA homology
of Marek's disease virus, herpesvirus of turkeys and Epstein-Barr virus under
varied stringent hybridization conditions. J. Biochem. 95:1215-1218.

Hirai, K., K. lkuta, and S. Kata. (1981). Structural changes of the DNA of Marek’s
disease virus during serial passage in cultured cells. Virology 115:385-389.

Hirai, K., K. lkuta, and S. Kata. (1979). Comparative studies on Marek’s Disease
virus and herpesvirus of turkey DNAs. J. Gen. Virol. 45:119-131.

Hirai, K., M. Yamada, Y. Arao, S. Kato, and S. Nii. (1990). Replicating Marek's
disease virus (MDV) serotype 2 DNA with inserted MDV serotype 1 DNA
sequences in a Marek’s disease lymphoblastoid cell line MSBt-41C. Arch. Virol.
114:153-165.

Hirt, B. (1967). Selective extraction of polyoma DNA from infected mouse cell
cultures. J. Mol. Biol. 26:365-369.

176

Hovi, T., A. Pitkaranta, A. Macadam, P. Minor, and J. Almond. (1991). Covariance
of lowered capacity to induce interferon in human Ieucocytes and temperature-
sensitivity of type 3 poliovirus. J. Interferon Res. 11:105-110.

lgarashi, T., M. Takahashi, J. Donovan, J. Jessip, M. Smith, K. Hirai, A. Tanaka,
and M. Nonoyama. (1987). Restriction enzyme map of herpesvirus of turkeys and
its collinear relationship with Marek’s disease virus DNA. Virology 157:351- 358.

lkuta, K. S., S. Ueda, S. Kato, and K. Hirai. (1983). Monoclonal antibodies
reactive with the surface and secreted glycoproteins of Marek's disease virus and
herpes virus of turkeys. J. gen. Virol. 64: 2597-2610.

lkuta, K. S., S. Ueda, S. Kato, and K. Hirai. (1983). Most virus-specific
polypeptides in cells productively infected with Marek’s disease virus or
herpesvirus of turkeys possess cross-reactive determinants. J. gen Virol. 64:961-
965.

lsfort, R. J., R. A. Stringer, H. -J. Kung, L. F. Velicer. (1986). Synthesis, processing
and secretion of the Marek's disease herpesvirus A antigen glycoprotein. J. Viral.
57:464-474.

lsfort, R., D. Jones, R. Kost, R. Witter, and HJ Kung. (1992). Retrovims insertion
into herpesvirus in vitro and in viva. P. N. A. 8. USA 89:991-995.

lsfort, R. J., D. Robinson, H.-J. and Kung. (1990). Purification of genomic sized
herpesvirus DNA using pulsed-field electrophoresis. Journ. Virol. Methods
27:311-318.

lwata, A., S. Ueda, A. lshihama, and K. Hirai. (1992). Sequence determination of
cDNA clones of transcripts from the tumor-associated region of the Marek's
disease virus genome. Virology 187:805-808.

Jones, D., L. Lee, J-L Liu, H. J. Kung, and J. K. Tillotson. (1992). Marek’s disease
virus encodes a basic-leucine zipper gene resembling the fos/jun oncogenes that
is highly expressed in Iymphoblastoid tumors. P. N. A. S. USA 89:4042-4046.

Kato, S. and K. Hirai. (1985). Marek’s disease vims. In "Advances in Virus
Research" vol. 30, pp. 225-277 (K. Maramorosch, F. A. Murphy, and A. J. Shatkin,
eds. ), Academic Press.

Katz, J. B., L. M. Henderson, and G. A. Erickson. (1990). Recombination in viva
of pseudorabies vaccine strains to produce new virus strains. Vaccine 8:286-288.

Kawai, S. and M. Niahizawa. (1984). New procedures for DNA transfection with
polycation and dimethylsulfoxide. Mol. Cell. Biol. 4:1172-1174.

177

Kuhn, R. J., D. E. Griffin, H. Zhang, H. G. Nesters, and J. H. Strauss. (1992).
Attenuation of sinibis virus neurovirulence by using defined mutations in non-
translated regions of the genome RNA. J. Viral. 66:7121-7127.

Laemmli, U. K. (1970). Cleavage of stmctural proteins during the assembly of the
head of bacteriophage T4. Nature (London) 227: 680-685.

Lee, L. F., X. Liu, and R. L. Witter. (1983). Monoclonal antibodies with specificity
for three different serotypes of Marek’s disease virus in chickens. J. Immunol.
130:1003-1006.

Lee, M. S., J. M. Roos, L. C. McGuigan, K. A. Smith, N. Corrnier, L. K. Cohen, B.
E. Roberts, and L. G. Payne. (1992). Molecular attenuation of vaccinia virus:
mutant generation and animal characterization. J. Viral. 66:2617-2630.

Lee, L. F., E. D. Kieff, S. L. Bachenheimer, B. Roizman, P. G. Spear, B. R.
Burmester, and K. Nazerian. (1971). Size and composition of Marek's disease
virus deoxyribonucleic acid. J. Virol. 7:289-294.

Lobigs, M., R. Usha, A. Nestorowicz, I. D. Marshall, R. C. Weir, and L. Dalgarno.
(1990). Host cell selection of Murray Valley encephalitis virus variants altered at
an RGD sequence in the envelope protein and in mouse virulence. Virology
1 76:587-595.

Long, P. A., J. L. Clark, and L. F. Velicer. (1975). Marek’s disease herpesvirus. II.
Purification and further characterization of Marek’s disease herpesvirus A antigen.
J. Viral. 15:1192-1201.

Macadam, A. J., G. Ferguson, J. Burlison, D. Stone, R. Skuce, J. W. Almond, and
P. D. Minor. (1992). Correlation of RNA secondary structure and attenuation of
Sabin vaccine strains of poliovirus in tissue culture. Virology 189:415-422.

Maotani, K., A. Kanamori, K. lkuta, S. Ueda, S. Kato, and K. Hirai. (1986).
Amplification of tandem direct repeats within inverted repeats of Marek's disease
virus DNA during serial in viva passage. J. Viral. 58:657-660.

Maray, T., M. Malkinson, and Y. Becker. (1988). RNA transcripts of Marek's
disease virus (MDV) serotype-1 in infected and transformed cells. Virus genes
2:49-68.

Martin, G. S. (1970). Rouse sarcoma virus: a function required forthe maintenance
of the transformed state. Nature (London) 227:1021-1023.

McCall, K. A., B. W. Calnek, W. V. Harris, K. A. Schat, and L. F. Lee. (1987).
Expression of a tumor-associated surface antigen on normal versus Marek’s

178

disease virus-transformed lymphocytes. J. Natl. Cancer Inst. 79:991-1000.

Melnick, J. L. (1982). Towards the eradication of poliomyelitis. in Medical Virology,
pp 261 -299 (L. Maza and E. Peterson, eds. ) Elsevier Science Publishing Co., Inc.
Mettenleiter, T. C., C. Schreurs, F. Zuckermann, and T. Ben-Porat. (1987a). Role
of pseudorabies virus glucoprotein gl in virus release from infected cells. J. Virol.
61 :2764-2769.

Mettenleiter, T. C., L. Zsak, A. S. Kaplan, T. Ben-Porat, and B. Lomniczi. (1987b).
Role of a structural glycoprotein of pseudorabies in vims virulence. J. Virol.
61 :4030-4032.

Mikiko, K., M. Hayashi, T. Fumichi, M. Nonoyama, E. Isogai, and S. Namioka.
(1991). The inhibitory effects of oligonucleotides, complimentary to Marek's disease
virus mRNA transcribed from the BamHI-H region, on the proliferation of
transformed Iymphoblastoid cells, MDCC-MSBI. J. Gen. Virol. 72:1105-1111.

Morgan, R. W., J. Gelb Jr., C. S. Schreurs, and P. J. A. Sondermeijer. (1992). A
live recombinant HVT vaccine expressing the newcastle disease virus fusion
proteins protects chickens from both newcastle disease and Marek's Diseases.
Proc. XIX World’s Poultry Congress, vol. 1, Pansen and Looijen, Wageningen, The
Netherlands.

Moriuchi, H., M. Moriuchi, H. A. Smith, S. E. Straus, and J. I. Cohen. (1992).
Varicella-zoster virus open reading frame 61 protein is functionally homologous to
herpes simplex virus type 1 ICPO. J. Viral. 66:7303-7308.

Moss, B. (1990). Poxviridae and their replication. In Fundamental Virology. 2"‘1
Edition, (B. Fields and D. M. Knipe et al., eds.) Raven Press, Ltd, New York.

Mass, B., E. Winters, and N. Cooper. (1981). Instability and reiteration of DNA
sequences within the vaccinia virus genome. P. N. A. S. USA 78:1614-1618.

Moss, E. G., R. E. O’Neill, and V. R. Racaniello. (1989). Mapping of attenuating
sequences of an avirulent poliovirus type 2 strain. J. Viral. 63:1884-90.

Moss, E. G. and V. R. Racaniello (1991). Host range determinants located on the
interior of the poliovirus capsid. EMBO J. 10:1067-1074.

Nazerian, K. (1970). Attenuation of Marek’s disease virus and study of its
properties in two different cell cultures. J. Natl. Cancer Inst. 44:1257-1267.

Nazerian, K. (1980). Viral Oncology: Marek's disease: a herpesvirus induced
malignant lymphoma of the chicken. p. 665-682. (G. Kein, ed.), Raven Press. New
York.

 

 

 

179

Nazerian, K., and R. F. Silva. (1988). Properties of producer and non-producer
clones of a Marek's disease turkey Iymphoblastoid cell line. Avian diseases
32:486-493.

Nazerian K. and R. L. Witter. (1970). Cell-free transmission and in viva replication
of Marek’s disease virus (MDV). J. Virol. 5:388-397.

Nazerian K. and B. R. Burmester. (1968). Electron microscopy of a Herpesvims
associated with the agent of Marek’s disease in cell culture. Cancer research
28:2454-2462.

Nazerian, K., L. F. Lee, N. Yanagida, and R. Ogawa. (1992). Protection of Marek's
Disease by a fowlpox virus recombinant expressing the glycoprotein B of Marek’s
disease virus. J. Viral. 66:1409-1413.

Neumann, E., M. Schaefer-Ridder, Y. Wang, and P. H. Hofschneider. (1982).
Gene transfer into mouse lyoma cells by electroporation in high electric fields.
EMBO J. 1:841-845.

Nitayaphan, S., J. A. Grant, G. J. Chang, and D. W. Trent. (1990). Nucleotide
sequence of the virulent SA-14 strain of Japanese encephalitis virus and its
attenuated vaccine derivative, SA-14-14-2. Virology 177:541-542.

Ohashi, K. and K. A. Schat. (1992). cDNA clones derived from the Marek’s disease
virus tumor cell line MDCC-CU41. Proc. XIX World’s Poultry Congress, vol. 1,
Pansen and Looijen, Wageningen, The Netherlands.

Pattison, M. (1985). Control of Marek’s disease by the poultry industry: Practical
considerations. In Marek’s Disease, scientific methods and basis of control. (L. N.
Payne, ed), Martinus Niijhoff Publishing, Boston, Mass.

Payne, L. N. (1985). Historical review. In Marek’s Disease, scientific methods and
basis of control. (L. N. Payne, ed.), Martinus Niijhoff Publishing, Inc., Boston,
Mass.

Payne, L. N. (1982). Biology of Marek’s disease virus and the herpesvirus of
turkeys. In "The Herpesviruses" (B. Roizman, ed.), Vol. 1 pp. 347-431. Plenum,
NY.

Payne, L. N. (1985). Pathology. In Marek’s Disease, scientific methods and basis
of control. (L. N. Payne, ed), Martinus Niijhoff Publishing, Inc., Boston, Mass.

Peng, F., G. Bradley, A. Tanaka, G. Lancz, and M. Nonoyama. (1992). Isolation
and characterization of cDNAs from the BamHI-H gene family RNAS associated
with tumorigenicity of Marek’s disease virus. J. Virol. 66:7389-7396.

 

 

180

Powell, P. C., K. Howes, A. M. Lawn, B. M. Mustill, L. N. Payne, M. Rennie, and
M. A. Thompson. (1984) Marek's disease in turkeys: the induction of lesions and
the establishment of lymphoid cell lines. Avian Pathol. 13:201-214.

Powell, P. C., L. N. Payne, J. A. Frazer, and M. Rennie. (1974). T-lymphoblastoid
cell lines from Marek’s disease Lymphomas. Nature (London) 251 :79-80.

Pratt, W. D., R. W. Morgan, and K. A. Schat. (1992). Characterization of
reticuloendotheliosis vims-transformed avian T Iymphoblastoid cell lines infected
with Marek's disease vims. J. Virol. 66:7239-7244.

Pritchard, A. E., M. A. Calenoff, S. Simpson, K. Jensen, and H. L. Lipton. (1992).
A single base deletion in the 5’ non-coding region of Theiler’s virus attenuated
neurovirulence. J. Virol. 66:1951-1958.

Pruthi, A. K., R. K. Gupta, and J. R. Sadana. (1989). Studies on the pathology of
Marek’s Disease following challenge in chicks vaccinated with three different
vaccines. J. Comp. Pathol. 101:295-305.

Pulaski, J. T., V. L. Tieber, and P. M. Coussens. (1992). Marek’s disease virus-
mediated enhancement of avian leukosis virus gene expression and virus
production. Virology 186:1 13-121.

Purchase, H. C., B. R. Burmester, and C. H. Cunningham. (1971). Responses of
cell cultures from various avian species to Marek's disease virus and herpesvirus
of turkeys. Am. J. Vet. Res. 32:1821-1823.

Purchase, H. G. (1984). Clinical disease and its economic impact. In Marek’s
Disease, scientific methods and basis of control. (L. N. Payne, ed.,) Martinus
Niijhoff Publishing, Boston, Mass.

Putnak, J. R. and B. A. Phillips. (1981). Picornaviral structure and assembly.
Microbiol. Rev. 45:287-315.

Pyles, R. B., N. M. Sawtell, and R. L. Thompson. (1992). Herpes simplex virus
type 1 dUTPase mutants are attenuated for neurovirulence, neuroinvasiveness,
and reactivation from latency. J. Virol. 66:6706- 6713.

Rhozinski, J., L. E. Girton, and J. L. Van Etten. (1989). Chlorella viruses contain
linear nonpermutated double-stranded genomes with covalently closed hairpin
ends. Virology 168:363-369.

Raab-Traub, N. and K. Gilligan. (1991). Comparison of EBV transcription in
lymphoid and epithelial cells. In Herpesvirus transcription and its regulation. (E. K.
Wagner, ed.) CRC Press, Inc, Florida.

181

Reychan, W. T., P. Ling, P. R. Kinchington, and J. Hay. (1991). The correlation
between varicella-zoster virus transcription and the sequence of the viral genome.
In Herpesvirus transcription and its regulation. (E. K. Wagner, ed.) CRC Press,
Inc., Florida.

Robbins, A. K., J. P. Ryan, M. E. Whealy, and L. W. Enquist. (1989). The gene
encoding the glll envelope protein of Pseudorabies virus vaccine strain BaItha
contains a mutation affecting protein localization. J. Virol. 63:250-258.

Robertson, L. M., A. R. MacLean, and S. M. Brown. (1992). Peripheral replication
and latency reactivation kinetics of the non-neurovirulent herpes simplex virus type
1 variant 1716. J. Gen. Virol. 73:967-70.

Roizman, B. (1992). The family Herpesviridae: an update. Arch. Virol. 123:425-449.

Roizman, B. . (1991). Fundamental Virology 2nd Edition: Herpes simplex viruses
and their replication. pp. 849-895. (B. N. Fields, D. M. Knipe, eds.), Raven Press,
Ltd., New York.

Roizman, B. (1990). Herpesviridae: A bn’ef introduction. In Fundamental Virology,
2"‘1 Edition, (B. Fields and D. M. Knipe et al., eds.) Raven Press, Ltd, New York.

Roizman, B. and A. E. Sears (1990). Herpes simplex viruses and their replication.
In Fundamental Virology, 2"“ Edition, (B. Fields and D. M. Knipe et al., eds.) Raven
Press, Ltd, New York.

Ross, L. J. N., M. Sanderson, S. D. Scott, M. M. Binns, T. Doel, and B. Milne.
(1989). Nucleotide sequence and characterization of the Marek’s disease virus
homologue of glycoprotein B of herpes simplex virus. J. gen. Virol. 70:1789-1804.

Ross, L. J. N., Milne, B., and Schat, K. A. (1984). Restriction enzyme analysis of
MDV DNA and homology between strains. In ”Intemational Symposium on Marek's
Disease“ (B. W. Calnek and J. L. Spencer, eds), pp. 51-67. American Association
of Avian Pathologists, Kennett Square, PA.

Sabin, A. B. and L. R. Boulger. (1973). History of Sabin attenuated poliovirus oral
live vaccine strains. J. Biolo. Standardization 1:115-118.

Sabin, A. B. (1957). Present status of attenuated live virus poliomyelitis vaccine.
Bull. NY Acad. Med. 33:17-39.

Salk, J. E. (1953). Principels of immunization as applied to poliomyelitis and
influenza. J. Public Health 43:1384-1398.

Sambrook, J., E. F. Fritsch, and T. Maniatis. (1989). ”Molecular Cloning: A

 

182
Laboratory Manual. " Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Sandri-Goldin, R. M. (1991). Analysis of the regulation activities of the HSV-l a
protein ICP27. In Herpesvirus transcription and its regulation. (E. K. Wagner, ed.)
CRC Press, Inc., Florida.

SAS User's Guide: Statistics, version 5. (1985). SAS lnst., Inc., Cary, N. C.

Schat, K. A., C-L. H. Chen, B. W. Calnek, and D. Char. (1991). Transformation of
T-Iymphocyte subsets by Marek's disease herpesvirus. J. Virol. 65:1408-1413.

Schat, K. A. (1985). Characteristics of the virus. In Marek's Disease, Scientific
Basis and Methods of Control (L. N. Payne, ed.) Martinus Nijhoff Publishing, Inc,
Boston, Mass.

Schat, K. A., A. Buckmaster, and L. J. N. Ross. (1989). Partial transcription map
of Marek’s disease herpesvims in lytically infected cells and lymphoblastoid cell
lines. lntl. J. Cancer 44:101-109.

Schat, K. A., B. W. Calnek, J. Fabricant, and D. L. Grahm. (1985). Pathogenesis
of infection with attenuated Marek's disease vims strains. Avian pathol. 14:127-
146.

Schwartz, D. C. and C. R. Cantor. (1984). Separation of yeast chromosome-sized
DNAs by pulsed field gradient gel electrophoresis. Cell 37:67-75.

Scott, S. D., N. L. J. Ross, and M. Binns. (1989). Nucleotide and predicted amino
acid sequence of the Marek's Disease virus and turkey herpesvirus thymidine
kinase genes; comparison with thymidine kinase genes of other herpesviruses.
J. gen. virol. 70:3055-3065.

Sharma, J. M. (1989). In situ production of interferon in tissues of chickens
exposed as embryos to turkey herpesvirus and Marek’s disease virus. Am. J. Vet.
Res. 50:882-886.

Shek, W. R., B. W. Calnek, K. A. Schat, and C. L. H. Chen. (1983).
Characterization of Marek's Disease virus infected lymphocyteszdiscrimination
between cylolytically and latently infected cells. J. Natl. Cancer Inst. 70:485-491

Shope, R. E. (1932). A Filterable virus causing tumor-like conditions in rabbits and
its relationship to virus myxomatosum. J. Exp. Med. 56:803.

Silva, R. F. and L. F. Lee. (1984). Monoclonal antibody-mediated
immunoprecipitation of proteins from cells infected with Marek's disease virus or
turkey herpesvirus. Virology 136:307-320.

183

Silva, R. F. and R. L. Witter. (1985). Genomic expansion of Marek’s disease vims
DNA is associated with serial in vitro passage. J. Virol. 54:690-696.

Smiley, J. R., B. Panning, and C. A. Smibert. (1991). Regulation of cellular genes
by HSV products. In Herpesvirus transcription and its regulation. (E. K. Wagner,
ed.) CRC Press, Inc., Florida.

Spear, P. and B. Roizman. (1981). Herpes simplex viruses. In DNA tumor viruses
2"‘1 Edition, Cold Spring Harbor laboratory, New York.

Stevens, J. G., E. K. Wagner, G. Devi-Rao, M. L. Cook, and L. Feldman. (1987).
RNA complementary to herpesvirus a-gene mRNA is predominant in latently
infected neurons. Science 235:1056-1059.

Stewart, C. J., M. Ito, S. E. Conrad. (1987). Evidence for transcriptional and post-
transcriptional control of the cellular thymidine kinase gene. Mol. Cell. Biol. 7:1 156-

1163.

Sugaya, K., G. Bradley, M. Nonoyama, and A. Tanaka. (1990). Latent transcripts
of Marek’s disease virus are clustered in the short and long repeat regions. J.
Virology 64:5773-5782.

Svitkin, Y. V., N. Cammack, P. D. Minor, and J. W. Almond. (1990). Translation
deficiency of the Sabin type 3 poliovirus genome: association with an attenuating
mutation C 472 -- U. Virology 175:103-109.

Takehara, K., M. K. Min, J. K. Battles, K. Sugiyama, V. C. Emery, J. M. Dalrympal,
and D. H. Bishop. (1989). Identification of mutations in the M RNA of a candidate
vaccine strain of Rift Valley fever vims. Virology 169:452-457.

Tanaka, A., Y-S. Lee, and M. Nonoyama. (1980). Heterogeneous populations of
Virus DNA in serially passaged Marek's Disease virus preparations. Virology
103:510-513.

Tieber, V. L., L. L. Zalinskis, R. F. Silva, A. Finkelstein, and P. M. Coussens.
(1990). Transactivation of the Rous Sarcoma virus long terminal repeat promoter
by Marek's disease vims. Virology 179:719-727.

Tillotoson, J. K., H-J Kung, and L. F. Lee. (1988). Accumulation of viral transcripts
coding for a DNA binding protein in Marek’s disease tumor cells. In Advances in
Marek’s disease virus research (S. Kato, T. Horiuchi, T. Mikami, and K. Hirai,
eds.), Japanese Assoc. on Marek’s disease, Osaka, Japan.

Tooze, J. (1981). DNA tumor viruses. Cold Spring Harbor laboratory, New York.

184

Toyoda, H., M. Kohara, Y. Kataoka, T. Suganuma, T. Omata, N. Imura, and A.
Nomoto. (1984). Complete nucleotide sequence of all three poliovirus serotype
genomes. J. Mol. Biol. 174:561-585.

Tyler, K. L. and B. Fields. (1990). Pathogenesis of viral infections. In Fundamental
Virology, 2"‘1 edition, (B. Fields and D. M. Knipe et al., eds.) Raven Press, Ltd, New
York.

vanZaane, D., J. M. A. Brinkhof, A. L. J. Gielkens. (1982). Molecular-biological
characterization of Marek’s disease virus. ll. Differentiation of various MDV and
HVT strains. Virology 121:133-146.

Velicer, L. F., D. R. Yager, and J. L. Clark. (1978). Marek's disease herpesviruses.
llI. Purification and characterization of the Marek’s disease herpesvirus B antigen.
J. Virol. 27:205-217.

Volk, W. A. (1982). Essentials of Medical Microbiology. J. B. Lippincott Co., Phila.,
Pa.

Wagner, E. K. (1991). Herpesvirus transcription; general aspects. In Herpesvirus
transcription and its regulation. (E. K. Wagner, ed.) CRC Press, Inc., Florida.

Weinstock, D., and K. A. Schat. (1987). Virus specific syngenic killing of
reticuloendotheliosis vims transformed cell line target cells by spleen cells. In Avian
Immunology, pp253-263, (W. T. Weber and D. L. Ewert, eds.), Alan R. Liss, New
York.

Weiss, R., N. Teich, H. Varmus, J. Coffin. (1984). Origins of contemporary RNA
tumor virus research. In RNA tumor viruses 2"“ Edition, Cold Spn'ng Harbor
laboratory, New York.

Wen, L. T., A. Tanaka, and M. Nonoyama. (1988). Identification of Marek's disease
virus nuclear antigen in latently infected Iymphoblastoid cell lines. J. Virol. 62:3764-
3771.

Westrop, G. D., K. A. Wareham, D. M. Evans, G. Minor, D. I. Magrath, F. Taffs,
S. Marsden, M. A. Skinner, G. C. Schild, and J. W. Almond. (1989). Genetic basis
of attenuation of the Sabin type 3 oral poliovirus vaccine. J. Virol. 63:1338-1344.

Wilson, M. R. and P. M. Coussens. (1991). Purification and characterization of
infectious Marek’s disease virus genomes using pulsed field electrophoresis.
Virology 185:673-680.

Witter, R. L. (1992a). Recent developments in the prevention and control of
Marek’s disease. Proc. XIX World’s Poultry Congress, vol. 1, Ponsen and Looijen,

185

Wageningen, The Netherlands.

Witter, R. L. (1992b). Safety and comparative efficacy of the CV1988/Rispens
vaccine strain. Proc. XIX World’s Poultry Congress, vol. 1, Ponsen and Looijen,
Wageningen, The Netherlands.

Witter R. L. and L. F. Lee. (1984). Polyvalent Marek's disease vaccines: safety,
efficacy, and protective synergism in chickens with maternal antibodies. Avian
Pathol. 13:75-92.

Witter, R. L. (1983). Characteristics of Marek’s disease viruses isolated from
vaccinated commercial flocks: association of viral pathotype with lymphoma
frequency. Avian diseases. 27:113-132.

Witter R. L., E. A. Stephens, J. M. Shanna, and K. Nazerian. (1975).
Demonstration of a tumor associated-associated surface antigen in Marek’s
Disease. J. Immunol. 115:177-183.

Witter, R. L. (1985). Principels of vaccination. In Marek's Disease, Scientific Basis
and Methods of Control (L. N. Payne ed.) Martinus Nijhoff, Inc

Witter, R. L., J. M. Sharma, and A. M. Fadly. (1980). Pathogenicity of variant
Marek’s disease virus isolates in vaccinated and unvaccinated chickens. Avian
Diseases 24:210-232.

Witter. R. L. (1982). Protection by attenuated and polyvalent vaccines against
highly virulent strains of Marek’s disease virus. Avian Pathol. 11:49-62.

zurHausen, (1981). Oncogenic herpesviruses. In DNA tumor viruses 2"cl Edition,
Cold Spring Harbor laboratory, New York.

”IC

IIIIIIIIIIIIIIIIIIIIIII