IFlZV-1 VACCINE DESIGN AND RAT CEAUXNGE MODEL

Jacqueline Arp

Department of Microbiology and hunoiogy

Facuity of Medicine

Submitted in partial fulfilment

of the reqirements for the degree of
Doctor of Philosophy

Facuity of Graduate Studies

The University of Western Ontario
Lmdon, Ontario
Aprii, 1996

Jacquelllie Arp, 1997

National Library

Bibliothque nationale
du Canada

Acquisitions and
Bibliographie S e ~ k e s

Acquisitions et
services bibliographiques

395 w

395. rue Wdcington

W O N K1AW
Cansda

d r iS m t
OrtawaON K1AiU4
Canada

The authot bas granted a nonexclusive iicence allowing the

National Lr'brary ofCanada to
reproduce, loaq distribute or seli
copies of his/her thesis by any means
and in any fomi or formaf malDng
this thesis available to interested
pemns.

L'auteur a accord une licence non

exclusive permethnt la
Bibliothque nationale du Canada &
reproduire, prter, distn'buer ou
vendre des copies de sa th& de
qyelcpe manire et sous quelcpe
forme que a soit pour mettre des
exemplaires de cette thse la
disposition des personnes inresses.

The author retaEs ownership of the

copyright in hismer thesis. Neither
the thesis nor substantial exffacts
fiom t ioay be printe or otherwise
reproduced with the author's
permission.

L'auteur conserve la proprit du

droit d'auteia qui protge sa these. Ni
la thse ni des extraits substantieis de
celle-ci ne doivent e miprims ou
autrement reproduits sans son

The widespread prevalence of human T-ceiI leukemia Vuus type 1(HTLV-9 and
the Me-threatening or debilitating diseases it induces are compehg reasons to search

for an HTLV-1 vaccine. Propeaies of the vinis, its mechanisms of infection and potential
effectiveness of the host immune respoose favour the feasl'bility of attaining
immunologicai protection agahst HTLV-1-

To ultimately test the efficacy of the HTLV-1 envelope protein as a potentiai

vaccine target antigen, the recombinant viral protein was expresseci in both the
bacuiovirus and vaccinia vin1~/T7polymerase systems. The entire HTLV-I envelope
protein, gp63, was produced in quantitative amounts in the baculovinis system, however
it was insoluble and the majority of protei. was incompletely post-translationaiiy

processed. The isolateci aggregates of envelope protein demonstrateci immunogenicity in

various mains of mice and were capable of stimulating envelope-specific T-ceiis from

naive hwnaos in vitro.

The vaccinia/T7 polymerase expression system was developed as an alternative

source of recombinant HTLV-1 surface envelope protein. The mammalian system
produced high levels of gp46 in a propetly processed and folded fom. Conformational
integrity of the recombinant protein was confirIi1ed by reactivity with sera from HTLV-1
infected patients and several conformational-dependent anti-envelope human monoclonal

iii

antibodies. Furthemore, this recombinant gp46 exhibitecl was recognized by HTLV-II

infected human sera and sera h m a Pan pmism chimpanzee infected with the distantly

related STLV-.

Highiy conserveci conformational epitopes maintaineci in the

recombinant HTLV-1 envelope protein structure mggestexi that it may serve as a usefiil
diagnostic reagent ami an effective vaccine candidate.

To enable the future evaiuation of potentid EELV-1 vaccines, an HTLV-1

infection model in adult and neonatal Fischer F344 rats was established. Through

challenge with a Live HTLV-1 infcted ceil he, Fischer F344 newborn and adult rats
were successhilly infected. In addition, the unprecedented infection of newborn rat pups

with short-term cultureci cells isolated fkom a HAM/TSP patient, was induced. To insure

redistic challenge parameters in the adult rat model, the minimum challenge dose of
HTLV-I infecteci ceils was detennined that would maintain a 100% frequency of
infection.

Keywords:

HTLV-1,

recombinant envelope glycoprotein, protein expression,

recombinant baculovinis expression system, recombinant vaccinialT7 polymerase

expression system,Fischer rat model of HTLV-1 infection, immunogenicity of HTLV-I
envelope protein, glycosylation studies, vaccine candidates, diagnostic reagents for
detection of HTLV-1 iafection

This thesis is dedicated to my parents

whose love gave me the strength.

Many significant coliaborations arose fiom this thesis research and enabled many

related facets to be explored, that wouid have been otherwise inaccessible. 1 would Wre
to thank Dr. Thomas Palker and Dr. Emily Riggs at Duke University (North Carolina,

USA) for their assistance in performing the virus neutraiization assays. 1would also Like
to ihanL

Dr. Steven Fouig and Judith (Joe) Rowe at Stanford Schwl of Medicine

(California, USA) for the use of their panel of anti-HTLV-I conformational-dependent

human monoclonal a n h i e s and anaiyzing the vTME46/vTF7-3 infected H9 and
HeLa ceils by immunofluore~cene.1 am also grateul to Dr. Fabrizio Manca at the
Advanced Biotechnology Centre (Genoa, Italy) for his interest and faith in the env-I.B.

preparation. Dr. Manca and his world-wide collaborators were instrumental in developing
the assay to test the proferative effects of env-I.B. on T-cells isolated from HTLV-1

naive individuais. My sincere appreciation goes to Dr. Judy Ball (U-W.0, Canada) for

both her knowledgeable technical assistance with the animais and her valuable intellectual
input throughout the years.

1 would Uce to extend a special thanks to all my present and past coileagues in
the lab, for thek insighthil scientific suggestions, collaborative support, and

encouragement. 1 feei very fortunate to have worked with my supervisor, Dr. Greg
Dekaban and laboratory reseachers consisting of Sheela Hota, Sumee Kim,Elaine King,

Cheryl Leystra-Lenz, Marcia Levatte, Andrew Peters, Michael Coulthart, Francois

Picard, Steve Po11and and Carol Ford - who aU share a valuable phosophy of life best
d e s c n i by George Elliot, "What do we live for, i not to make the world les difficuit
for each other?"

1 wouid like to extend a siacere thaok you to my mentor Dr. Greg Dekaban for

his guidance and perseverhg interest in my scientific career, but particularly for his

compassion and understanding of my personal cinumstanes. 1 also feel very fortunate

and blessed to have the special niendships of Sheela Hota, Christoph von Baiiestrem and
Olga Krassioukova, whose love of Me and strength of character were an inspiration to

me and my work.

Lastly, 1 am grateful to the Medical Research Council of Canada and its

scholarship candidate selection cornmittee for their support through the duration of this
snidy.

TABLE OF CONTENTS

TABLE OF CONTENTS

LIST OF FIGURES
LIST OF TABLES

...
....................................... vlii

...
..........................................xiri

...........................................xvi

LIST OF APPENDICES

......................................

xvii

LIST OF ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . wiii

CHAITER 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 General Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
1.2 Epidemiology of HTLV-1 ............................... 1
1.3 HTLV-1 Infection and Disease ........................... 2
1.3.1 Adult T+eU Leukemia/Lymphoma . . . . . . . . . . . . . . . . . . 2
1-3.2 HTLV-1 Associatecl Myelopathy/Tropical Spastic
Paraparesis ..................................
4
1.3.3 Cumulative lifetime risk of HTLV-1 associated
disease/disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Phylogenetic Ongin of HTLV-1 . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4.1 Relatedness o f primate T-ceil leukemialiymphoma vUuses
(PTLVs) .................................... 11
1.5 Molenrlar Biology of HTLV-1 .................... .
.
.
.. . . 13
1.5 .1 Genomic Organization of HTLV-1 . . . . . . . . . . . . . . . . . . 13
1.5.2 Regdatory firnctiom mediated by Tax . . . . . . . . . . . . . . . . 17
1S.3 Regulatory fiuictions of Rex ....................... 20
1S.4 Oncogenic Potential of p 12' ...................... 21
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.6 HTLV-1 trop1.7 Modes of Transmission ................................ 26
1.8 immune responses associated with protection agabt HTW-1
Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
1.8.1 Passive immunkation of animals . . . . . . . . . . . . . . . . . . . . 27
1.8.2 Naturai Immunity in Humans . . . . . . . . . . . . . . . . . . . . . . 28
O

viii

1.8.3 Possible Mechanisms of Anhibody-mediated Protection

1.8.4 Potentiai role of cytotoxic T-ceUs in protection against

....

29

HTLV-1 .................................... 33
i -9 Envelope glycoproteinas a Vaccine Target ................... 35
i .9.i Properties of the HTLV-1 Envelope Glycoprotein ........ 35
1.9.2 Genetic Variation o f the H'EV-1 Envelope Gene . . . . . . . 37
1.9.3 Active Iimunization of Recombinant HTLV-1 Envelopebased Vaccines ............................... 38
1.9.4 m e r Retroviral Envelope-based Vaccines ............. 40
1.10 Feasibility of generating a successfiil vaccine against HTLV-1 . . . . . 41
1.1 1 Thesis objectives and contents ........................... 43

CHAPTER 2 EXPRESSION AND IMMUNOGENICFI'Y OF THE

ENTIRE HUlMAN T-CELL LEUgEMIA VIRUS TYPE 1
ENVELOPE: PROTEIN PRODUCED IN A BACULOVIRUS
S Y s m ............................................. 46
2.1 htroduction ........................................ 46
2.2 Materials and Methods ................................. 49
2.2.1 Ceii lines ....................................49
2.2.2 Plasmids and vinises ............................ 49
2.2.3 Isolation of recombinant baculovims . . . . . . . . . . . . . . . . . 50
2.2.4 DNA extraction and Soutbem blot analysis . . . . . . . . . . . . . 51
2.2.5 Indirect immunofluorescence of baculovirus infected ceils . . 52
2.2.6 Detection of secreted or cel associateci HTLV-1 envelope
protein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.7 Inclusion body isolation . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.8 Solubilizatioa o f inclusion bodies for in vitro ceil assays . . . 54
2.2.9 Characterization of the recombinant HTLV-1 envelope
protein forms ................................. 55
2.2.10 Western blot assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.2.11 Ixnmunizations ............................... 57
2.2.12 Immunoprecipitation of radiolabelled proteins . . . . . . . . . . 58
2.2.13 Peptide ELISA ............................... 59
2.2.14 Syncytium inhibition assay . . . . . . . . . . . . . . . . . . . . . . . 60
2.3 Redts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.3.1 Construction and identification of recombinant
baculovirus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.3.2 Identification and isolation of the HTLV-1 envelope
glycoprotein ................................. 67
2.3.3 Characterization of env-I.B. . . . . . . . . . . . . . . . . . . . . . . . 75
2.3.4 Immunogenicity of the inclusion bodies in vivo . . . . . . . . . 76
2.3.5 Characterization of the A n t i i i y Respome to env-I.B.
. . . . 77
2.3 .6 Combined Immunization Regimens . . . . . . . . . . . . . . . . . . 83

2.3 -7 Other studies determining the mtigenicity~tmmuaogenicity

of env-1.B ................................... 91
2.4 Discussion ......................................... 95

CH-

3 .A SOURCE OF GLYCOSYLATED El'LV-1 ENVELOPE

PROTEIN= EXPRESSION OF gp46 BY THE VACCINIA/T7
POLYMERASE SYSTEM ................................
3.1 introduction .......................................
3.2 Materials and Methods ................................
3.2.1 CeUlines ...................................
3.2.2 Construction o f recombinant plasmid ................
3.2.3 Isolation of recombinant vaccinia virus . . . . . . . . . . . . . .
3 .2.4 Southern blot analysis . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.5 Recombinant vaccinia virus infitions ...............
3.2.6 Source o f confomtion-dependent human monoclonai

105
105
112
112
112
113
114
115

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Radiolabelling ....................... .
.
.. . . . . 116
Glycosylation Inhibition . . . . . . . . . . . . . . . . . . . . . . . . . 116
Imrnunoprecipitation ........................... 117
anticbodies

3 2.7
3.2.8
3 .2.9
3 .2.10 Glycosidase Digestions . . . . . . . . . . . . . . . . . . . . . . . . .
3 .2.1 1 immunoblotting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.12 Binding of HMAbs to "native" and denatured dual
vaccinia infcted cell lysates and recombinant
baculovirus-derived envelope inclusion bodies . . . . . . . . .
3 .2.13 indirect Immunofluorescence of dual recombinant
vaccinia infkcted cells ..........................

...........................................

118
119

121
122
124

Construction and characterization of recombinant vaccinia

virus

.....................................

Identification and characterization of the recombinant

proteins expressed by the dual vaccinia system: vTME46hTF7-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
N-glycosylation of the recombinant HTLV-1 envelope
protein forms ................................
Effects of vinis multipiicities on recombinant HTLV-1
envelope protein expression . . . . . . . . . . . . . . . . . . . . . .
Time course of protein yields . . . . . . . . . . . . . . . . . . . . .
Expression in H u m H.9versus HeLa ceils . . . . . . . . . . .
Epitopes of recombinant HTLV-I gp46 shared with
........................
HTLV-II aad STLVBinding of conformation-dependent anti-HTLV-1 human
monoclonal antiibodies to recombinant envelope proteins . .
Influence of oiigosaccharide complexity on antibody
binding to conformational epitopes . . . . . . . . . . . . . . . . .

124

129
129

135
138
141

144
148

162

3.4 Discussion

........................................

166

..............

177
177

C-4-RATMODELOFHTLV-IINFECTION

4.1 Introduction .......................................

4.2 Materials and Methods ................................
4.2.1 Ceii lines ...................................
4.2.2 Short-term culture o f T-ceiis h m EIA.WTSP patient ....
4.2.3 Animais and transmission of HTLV-1 ...............
4.2.4 Western blot detection of antibodies against HTLV-1

proteins

...................................

179
179

179
180
181

4.2.5 Detection of anti-envelope a n h i e s using

....................... 181
...................... 182
4.2.7 Polymerase Chain Reaction ....................
.
. 183
4.2.8 Southern blot analysis of PCR amplifieci products ....... 185
R d t s ........................................... 186
cadioimniunoprecipitation

4.2.6 Syncytium Inhiiition Assay

4.3

4.3.1 lafctious transmission of HlIZV-1 into adult rats . . . . . . . 186

4.3.2 HTLV-1 infection of newborn rat pups . . . . . . . . . . . . . . 191
4.3.3 Detennination of minimal infectious dose ............. 196
4.3-4 Detection of anti-envelope anb'bodies . . . . . . . . . . . . . . . 201
4-4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

CHAPTERS-PERSPECTWES ................................ 213

5.1 Diagnostic screening with recombinant envelope protein antigens . . . 2 14

5.2 Recombinant envelope proteins as potential vaccine candidates . . . . 218
5.2.1 Glycosylation effects on antigeaicity and immunogenicity . . 220
5.2.2 Biologicai coIlSequences of native versus denatured
conformation ................................ 227
5.2.3 Rat challenge mode1 available for testing of potential
vaccine candidates ............................ 232
5.3 Goals of vaccination: Generation of anti-viral antibody and/or
CTL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
5.4 Goals of vaccination: Target populations . . . . . . . . . . . . . . . . . . . 236
5.5 Goals of vaccination: Prevention of infection amilor disease . . . . . . 240

LIST OF FIGURES
Description

page

Neighbour-joining phylogenetic tree of HTLV and STLV isolates. . . . . . 9

Schematic representation of the viral RNA gemme structure and
the encodeci proteins of HTLV-1 . . . . . . . . . . . . . . . . . . . . . . . . -15

Gene d e r vector used to generate the recombinant baculovinis

for expression of the fITLV-1envelope glycoprotein . . . . . . . . . . . . .64

Location of synthetic peptides spanning HTLV-1envelope protein

that were used in peptide ELISA . . . . . . . . . . . . . . . . . . . . . . . . .64
Southern blot arialysis of recombinant baculovirus encoding the
entire HTLV-1envelope protein . . . . . . . . . . . . . . . . . . . . . . . . .66
Indirect immunofluorescence analysis of recombinant baculovinisinfected insect celis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
Western blot analysis of supernatants and ceii pellets followhg
recombinant baculovinis infection . . . . . . . . . . . . . . . . . . . . . . . .71
SDS-PAGE anaiysis of HTLV-1envelope proteins isolated as
inclusion bodies by recombinant baculovirus . . . . . . . . . . . . . . . . - 7 4

Western blot analysis of inclusion bodies . . . . . . . . . . . . . . . . . . . .74

Radioimmunopraipitation aiialysis used to monitor the in vivo
immunogenicity of the HTLV-1envelope inclusion bodies . . . . . . . . . 79
Peptide ELISA titres of sera from mice immunized with inclusion
bodies and various arnounts of adjuvant . . . . . . . . . . . . . . . . . . . . 8 2
Diagrammatic representations of recombinant vaccinia virus
constructs utilized in combined immunization reghem in wivo . . . . . . 85
Monitoring anti-envelope seroconversion of mice receiving
combined immunization regimens by western blot analysis . . . . . . . . . 88

xiii

Per cent reactivity of HTLV-1infcted asymptomatic and

HAM/TSP sera to env-I.B. for each of the immunoglobulin (Ig)
isotypes and IgG subtypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93
Southern blot analysis of recombinant vaccinia virus encoding the
HTLV-1 surface envelope protein . . . . . . . . . . . . . . . . . . . . . . . 126
Schematic diagram of the duai recombinant vaccinia/T7
polymerase system for expression of the HTLV-1 surface envelope
protein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128
Western blot anaiysis of HeLa cells infecteci with the dual
recombinant vaccinia vinises . . . . . . . . . . . . . . . . . . . . . . . . . . 131

N-glycosylation determination of recombinant HTLV-1 surface

envelope protein forms produced by the duai recombinant
vaccinia/T7 polymerase system . . . . . . . . . . . . . . . . . . . . . . . . . 134
Dependence of maximal gp46 expression by the dual vaccinia
system on virus multiplicities . . . . . . . . . . . . . . . . . . . . . . . . . . 137

Time course of recombinant protein yields fiom two different

vaccinia expression systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Expression of recombinant HTLV-I SlUface envelope protein in
Human K9 versus HeLa cells . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Western blot reactivity of HTLV-IC, HTLV-IICand STLVwp'
sera with recombinant HTLV-1 surface envelope protein . . . . . . . . . 146

Graphic demonstration of the dose-dependent ability of antiHTLV-1 human monoclonal anti.i.iesto intuiit HTLV-1
syncytium formation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

Radioimmunoprecipitation of dual vaccinia infecteci ceil lysates

with HTLV-I envelope-specific human monoclonal antibodies . . . . . . 155
Immunoprecipitation of native and denaturecl vTME-46/vTF7-3 inf'ted
ce11 lysates (rgp46) and baculovirus-generated recombinant HTLV-1
envelope protein extracts (env-I.B.) with various human monoclonal
antibodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 8

xiv

h d k c t immtmofluorescence of dual vaccinia iofected HeLa

celis with RTLV-1-specificsera . . . . . . . . . . . . . . . . . . . . . . . . 161
DBerential immunoprecipitation of recombinant envelope
proteins expressed in the absence or presellce of the
glycosylation inhibitors Brefeldin A and tunicamycin . . . . . . . . . . . 164
Seroreactivity of adult Fischer F344 rats with HTLV-1 antigens
fouowing challenge with HTLV-1 infected MT-2cells . . . . . . . . . . . 188
PCR amplification of HTLV-1envelope sequences fkom peripheral
blood lymphocytes of adult Fischer rats injected with HTLV-1
infkcted MT-2 celis

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

PCR amplification of HTLV-1envelope sequences h m peripheral

blood lymphocytes of family 1 rat pups that were injected with
MT-2 ceUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

PCR amplification of HTLV-I envelope sequences from peripheral

blood lymphocytes of rat family 2 pups that were iojected with
ceiis freshiy isolateci fiom a HAM/TSP patient. . . . . . . . . . . . . . . . 195
Seroreactivity of adult rats with HTLV-1 antigens foliowing
challenge with various amounts of MT-2cells to determine the
minimal infectious dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

PCR amplikation of HTLV-1 envelope sequences fkom perpheral

blood lymphocytes of adult rats injected with various amounts
of MT-2 ceils to determine the minimal infectious dose . . . . . . . . . . 200
Radioimmuaoprecipitation of recombinant envelope proteins by
rat sera foliowing challenge with a high dose of HTLV-Iinfe~kMT-2 =US . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 2 0 3
Radioimmmoprecipitation of recombinant envelope proteins by
rat sera foilowing challenge with various amounts of HTLV-Iinfected MT-2 ceils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 2 0 3

2.1

Characterization of the immune response induceci in mice immunized

with the combineci recombinant vacciaia virus and HTLV-1 envelope
protein regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -90

3.1

Characterization of the anti-IirrZV-1 human monoclonal anti'bodies

xvi

...

150

LIST OF APPENDICES

Comparative analysis of the antibody response to the HTLV-1

in HTLV-1 asymptomatic carriers and
W T S P patients: an isotype and subclass arialysis . . . . . . . . . . . 242

gag ami env proteins

Recognition of HTLV-1 envelope by human CD4+ T-cell lines

nom HTLV-1 seronegative individuais: specificity and clonal
heterogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -254

xvii

LIST OF ABBREVIATIONS
aa

amino acids

Abs
AcMNPV

Aurographa cali$omica multi-nuclear polyhedrosis virus

anti'bodies
(bacuiovinis)

AIDS
ADCC
APC(s)
ATL
ATLL
B-ceU
BCIP
BHV-1
BLV
BSA
BUdR
CaCi,

CD
CDC
CDL
OC
Ci/mmole

CMV
CNS
CO?
cpdw
CSF
CTL
CREB

CREM
dCTP
DMEM
DNA
DNAse 1
DTT
E.C.L.
EBV
EDTA
ELISA
EMC

acqyited immunodeficiency syidrome

antibody-dependentceil-mediated cytotoxicity
aatigen presenting celi(s)
aduit T-cell leukemia
aduit T-celi leukemia/lymphorna
lymphocyte of bone marmw origin (B-lymphocyte)
5-brorno4chloro-3-indolyl phosphate
bovine herpesvirus type 1
bovine ieukemia virus
bovine senun albumin
bromo-2deoxyUndine
cdcium chioride
cellular determinant
complement-dependent cytotoxicity
complementdependent lysis
Celcius (degrees)
Cumes per millimole
cytomegaloWus
central netvous system
carbon dioxide
counts per minute per microgram
cerebrospinai fluid
cytotoxic T-lymphocytes
cyclic adenosine monophosphate tesponsive element binding protein
B
cyclic adenosine monophosphate responsive element binding protein
M
deoxycysteiw triphosphate
Dulbecco's Modifiecl Eagle's tissue cuiture medium
deoxyriboonucleic acid
deoxyribonuclease type 1
dithiothreiotol
enhanced cherniluminescent detection
Epstein Barr Vus
eihylenediaminetetraacetic acid
enzyme-linked immunosorbent assay
encephalomyocarditis virus
xviii

env
ENV
env-1.B.
F-MuLV

FACS
FCS
FeLV
gag
Gag

GAM-FITC

GM-CSF
a21
gp41
gp46
~ 6 3
gpll0
gp120
gp160
EI
H+ ATPase
HC1
HLA-DR

Mab
HOS
hr .

HAM/TSP
HIV
m-112

H
HTLV
HTLV-rn
i.p.
i.v.

FA
1FN-Y
Ig
IL-2/3/5/6

IL-2R
IL-2Rd@Iy
in. Hq.
kb
KCl
kD

endonuclease H
retroviral gene encoding envelope glycoproteins
retroviral envelope proteins
recombinant HTLV-I envelope protein inclusion bodies
Friend-murine leukemia vins
fluorescent activated ceIl sorting analysis
fetal calf serum
feline leukemia-sarcoma vinis complex
netroviral gene e d i n g core proteins
retroviral core proteins
goat anti-human fluorescein isothiocyanate
granulocyte-macrophage colony-stimulating factor
HTLV-1 envelope trammembrane glycoprotein of 21 kD
HIV-1 envelope tmnsmembrane glycoprotein of 41 kD
HTLV-1envelope d a c e glycoprotein of 46 kD
HTLV-1envelope glycoprotein precursor of 63 kD
SIV envelope giycoprotein of 110 kD
HIV-1envelope surface glycoprotein of 120 kD
HIV-1 envelope glycoproteh precursor of 160 kD
human epidemiological clade
adenosine triphosphatedependent vacuolar proton pump
hydrochloride
human leukocyte antigen class II
human monoclonal antibdy
human osteosascoma celi line
hour(s)
HTLV-1 Associated MyelopathyITropical Spastic Paraparesis
human immunodeficiency viruses
human imrnunodeficiency virus type-112
human epidemiological clade
human T-cell leukemianymphoma vinws
human T-cellleukemia/lymphoma virus type-I/II
intravenous
imrnunofluoresceace of fixed cens
gamma-interferon
immunoglobulin
interleukin-2/3/96
interleukin-2 receptor
IL-2receptor alpha, beta and gamma chains
vacuum pressure measured in inches of mercury
Hobase(s)
potassium chioride
kiiodaltons
xix

NHS
N-linked
NMS
orf

Iive cell immunofluorescence assay

lymphocytic choriomeningititis vinis
retrovirai long terminal repeat
Molar
monoclonal anti'bodies
mycobacterial ceil waii extract
mligmm per millilitre
milligrams of immunoglobuiin per miiiiiitre
magnesium chioride
major histocompaaility molecules
minute(s)
mitlimolar
multiplicity of infection
messager RNA
sodium chloride
nitro-blue tetrazolium
normal chimpanzee senim
nuclear factor kappa B
normai huma0 senun
asparagine-linked glycan moiety
normal mouse serum
open reading fhme

OTG

octy l-8-D-thiogiucoside

PL

p s t infection
peripheral blood lymphocytes
peripheral blood mononuclear celis
phosphate buffered saline
polymerase c h a h reactioa
plaque-forrning units per celi
phytohemagglutinin
pheny1-methy1-suLfony1-fluoride
Glycopeptidase F
retroviral gene ewoding RNA-dependent DNA polymerase
retroviral RNAdependent DNA polymerase protein (reverse
transcriptase)
polyadenosine
primer pair
primate Tceii leukemia/lymphoma viruses
parathyroid hormone-related protein
Rex-responsive cis-acting element
recombinant HTLV-1 gp46
recombinant HIV gp120
ribonucleic acid

LCA
LCMV

LTR
M
MAbs
M m

min.
mM
moi
mRNA
NaCl

NB
NCS
NF-KB

PBLs
PBMCs

PBS
PCR
pficeil
PHA

PMSF
PNGase F
PO[

Pol

poly (A)

Pr*

PTLV

PTHrP
Rex RE
rgp46
rgpl20
RNA
RNAse

ribonuclease

R W Els

R W Elas
S

SAmS

SDS
SDS-PAGE
SB

STLV
STLVp4-1
SN
SRF
SSC
T@?
T-ce11
TGF-B 1
tk

TKtRNA

TSP
Tunc.

recombinant vaccinia virus constnicts expressing variations of the

H'l'LV-1 envelope glycoproteins
recombinant vaccinia Vuus expressing the wiid-type envelope
proteins in the sense orientation
RW El construct expressed in the anisense orientation
simian epidemiologicai clade
Simian acquired immunodeficiency syndrome
sodium dodecyl sulfate
sodium dodecyl sulfate-polyacryIamide gel electrophoresis
Spodoptera fnrgiper gut epiheiial ceii h e
simian T-ceii leukemia vinises
STLV strain pan poniscus/ype-1
simian immunodeficiency virus
senun respome factor
stamhd sodiwn citiate
Thennophilus aqwtcus
iymphocyte of thymus origin (T-lymphocyte)
tumour growth factor-81
thymidine kinase gene locus
thymidine kinase negative phenotype
transfer RNA
Tropical Spastic Paraparesis
tunicamyciptreated
unit(s)

micrograms
micrograms per millilitre
microlitres
untreated
uridine triphosphate
recombinant baculovinis (AcvMNPV) encoding the entire HTLV-1
envelope gene hgment
viral protein
recombinant vaccinia virus expressing vTME-46
recombinant vaccinia virus expressing HTLV-1 surface envelope
protein, gp46
wiid-type genotype/phenoype
times gravity
positive
per cent weight pet volume

CHAPTER 1 INTRODUCTION

1.1 General Introduction

The widespread prevalence of human T-ceii leukemia virus type 1(HTLV-I) and
the me-threatening or debilitating diseases it induces are c o m p e h g reasons to continue
the search for an HTLV-1 vaccine. Properties of the virus, its mechanisms of infection

and potentiai effectiveness of the host immune response favour the feasibility of attaining
imrnunological protection against HTLV-1. In addition, its simiiarities to the human

irnmunodeficiency virus 0 suggest that advances in the development of an HTLV-1

vaccine rnight help in the design of an HIV vaccine, especiaiiy when addressing the issue

of whether it is possible to prevent infection by ceii-associated retrovinises.

1.2 Epidemiology of EiTLV-1

Major endemic foci of HTLV-1 infection exist in ethnic subpopulations of

southern Japan, intertropical Africa, and the Caribbean basin (Hinuma et al. , 1982;
Kazedi-Kayembe et al., 1990; Blattner et al., 1982). Other endemic foci have been
identified in the Middle East (Meytes 1990), South America and the Seychelles Islands
(Arango et al., 1988; Lavanchy

et al., 1991). However, reports of sporadic HTLV-1

infection and associated diseases have now corne tom almost every major region of the

world. Relatively recentiy, HTLV-1 has been found in indigenous peoples fiom Pacifc

Rim countries and North and South America, including Ausualia and Meianesia (Saksena
et al., 1992; Bastian et al., 1993; Gessain n al., 1993). ur laboratory has recentiy
published the first observations of HTLV-1 infection and associateci diseases in North
Amencan Amerindians from coastai British Columbia (Oger et al., 1993; Dekaban et al.,

1994). Epidemiological estimates suggest that a total of 10 to 20 d o n persons are

infected with HTLV-1woridwide (De Th et al., 1993).

1.3 HTLV-1 Infection and Disease

1.3.1 Adult T-cell Leukemia/Lymphoma

Human T-celi Leukemia Virus Type 1 (HTLV-I)has been finnly established by

epidemiological and molecular studies to be the etioolgical agent of adult T-ce11
Ieukemia/lymphoma (ATL) (Robert-Guroff et al., 1982; Yoshida et al., 1984). The
existence of the malignant syndrome was first descnbed to be prevalent in clustered areas

of southwestem Japan, by Takatsuki et al. (1977). However it was not untii 1980, that
HTLV-1 virions were successfdiy isolated nom ceils of an ATL patient (Poisez et al.,
1980). ATL is a rapidly progressing malignancy, usually of mature CD4+ T
Lymphocytes, that occurs in 2-5% of HTLV-1 infected people within their lifetirne

(Yamaguchi and Takatsuki, 1993).The average interval between initial HTLV-1 Section

and the developrnent of maIignancy is 20 to 30 years (Tokudome a al., 1989).

Epidemiologic data indicate that ATL develops mrinly in individuals infected at binh and
suggest that the age of initial Wal iafction may be important to the development of
ieukemia (Takatsuki et al., 1985; M q h y et al., 1989; Yamaguchi et al., 1993;
Yamaguchi et al., 1994; Cleghorn et al., 1995).

Clinicaily, ATL is an aggressive, lethal disease that is highly resistant to

chemotherapy treatment with individuais having a mean survival time of oniy a few
months (Tajima et al., 1990). Patients with ATL often present symptoms of
hypercalcemia, skin infiltration, hepatosplenomegaly and lytic bone lesions (Takatsuki
et al., 1985; Yamaguchi et al-, 1983; Yamaguchi et al., 1994). Rare cases of

spontaneous remission of ATL have been desrribed (Shimamoto et al., 1993).

Characteristic of ATL, is the presence of large flower-like ceUs with lobulated nuclei
(Yamaguchi et al., 1983; Takatsukiet al., 1985; Yamaguchi et al. , 1994) and the display

of activation markers including CD25 (IL-2R)and HLA-DR by the leukemic cells

(Takatsuki et al., l m ) . The clinical course of ATL comprises at least four different

subtypes referred to as acute, chronic, lymphomatous and smouldering, that are

differentiated by the extent of the disease manifestations and serum calcium Ievels of the
patient (Yamaguchi et al., 1983; Yamaguchi et al., 1993). Ail leukemia subtypes are

characterized by the presence of virai DNA clonally integrated in the ceiiular DNA of
the neoplastic T-ceils (Wong-Staal et al., 1983; Yoshida et a., 1984). In fact, the
appearance in HTLV-1 infected individuais of one or more prevalent T-cell clones

4
carrying the HTLV-1provital genome represents an iocreased risk of developing fullblown disease (Yamaguchi et al., 1988). Interestingiy, in the leukemic cells of ATL
patients, the provinrs is often present in a lamu state (Franchini et ai., 1984; Felber et
al., 1985) or is defective (Korber et ai., 1991), suggesting that progression of disease

might not reque the expression of viral proteins in the leukemic ceils.

1.3.2 ffIZV-1 Associatecl Myelopathy/Tropicai Spastic Paraparesis

HTLV-1infection has also been causally Linked to a degenerative neurological

disorder known as HTLV-I Associated Myelopathy/Tropical Spastic Paraparesis,
(HAM/TSP) (Osame et al., 1986). This disease was first clinicaliy characterized in the
tropics and designated TSP (Gessain et al., 1985; Rodgers-Johnson et al., 1985); later
the same syndrome syndrome was rezognized in HTLV-1 infected Japanese patients as

HAM (Osame et al., 1986). Cases of HAWTSP have been described in every area
where HTLV-I is eodemic. This syndrome is characterized by overall hyperimmunity
maaifested as a pan B-ceU response with high antibody titres ami activated T-cells in both

the circulation and spinal fluid of patients (Osame et al., 1987; Link et al., 1989;

Iwasaki, 1990). Indeed, one of the hallmarks of lymphocyte abwrmalitites observed in

HAMITSP patients is the high degree of spontawous lymphocyte proliferation in vitro,

in the absence of stimulation (Itomaya et al. , 1988). ClinicaIly,these myelopathy patients

demonstrate spasticity and weakness of their lower limbs, paraplegia, sphincter

disturbances and various degrees of sensory nerve loss. The pathology observed in
HAMfTSP involves progressive demyelination of long motor neuron tracts in the spinai

cord which may be accompanied by axona loss, p e r i v d a r cuffing, and gliosis in the
lower thoracic spinal cord (Montgomery, 1983; Osame et al., 1989; Kermode et al.,
1990; Roman et al., 1990; Gessain ancl Gout, 1992; HoLlsberg and Hafier, 1993).

Cumulative lifetime risk of HTLV-1 aSSOciated diseaseldisorder

The cumulative lifetime risk of an HTLV-I infectexi individual developing the Lifethreatening adult T-ceil leukemia or the incapacitating tropical spastic paraparesis had
been previously estimateci to be approximately 5% (De Th and Bomford, 1993). This

cdculated fkequency was largely based on the incidence of dwase in Japaaese and
Jarnaici111 populations; however, an over-representation of HAM/TSP in HTLV-1 infected

individuals has recently been reported in several populations including those of

Colombia, Martinique, Israel and the Northwest Pacific Coastal Indians (Kaplan and
Osarne, 1990; Achiron et al., 1993; Oger et al., 1993; Rerjifo et al., 1995). For
example, in a population of Iranian-born Mashhadi Jews, close to 60% of HTLV-1
infected patients exhibit signs of myelopathy (Achiron et al., 1993). Thus the 5%

combined frequency of ATL and W T S P incidence in HTLV-1 carriers may be an

overly conservative estimate and cannot be applied to geographically distinct people.
These variations in disease frequencies suggest that the genetic background and socio-

ecoaomic status of the population may have a profourd eEect on the virus-related
morbidity and mortity (Sonoda et al., 1994; Blattner 1995). Sonoda has reported that
HAM/TSP segregates with specinc RLA-DR haplotypes @uman leukocyte antigen class
II)which are associated with a tendency to exhibit a "highimmune cesponse" to various
antigens (Sonoda et al., 1994). whiie ATL haplotypes generaliy dernonstrate a "low

immune response" (Sowda m al., 1994; Iacobson 1994).

An additional 305% of individuals infectai with HTLV-I exhibit chronic

Mmunosuppression, arthropathy (Ijichi et al., 1990). uveitis (Mochizuki m al., 1992)

and infective dermatitis (Lagrenade et al., 1990). There is also evidence fiom both the
Caribbean and the United States that dual infection with HTLV-1 and HIV accelerates the

progression to AIDS, (Bartholomew et al., 1987; Page et al., 1990) hcreasing the risk

threefold in the United States foliow-up study (Page et al., 1990). In summary, one cm
project that approximately 8 to 10% (1 million) HTLV-1 iafcted individuals will be
stricken with an overt fatal disease or overail morbidity (De Th and Bomford, 1993).
The global incidence of HTLV-1 infection, its potentiai to induce disease and the severity
of the resultant diseases, places HTLV-1 in the category of viruses for which vaccines

are made and used.

Interest in the MWbistory and evolution of HTLV-1 was stimuiated by the

isolation of genetically-related vinises in some species of nonhuman primates (Fukasawa
et al. , 1987; Komurian et al., 1991; Paine et al., 1991; Gessain et al., 1992a; Saksena
et al., 1993; Koralnik et al., 1994;Saksena et al., 1994). By the use of the polymerase

chah reaction, several Wal strains were characterized fiom human and nonhuman

primates residing in divergent geographic ar~asof the world. Relative cornparisons of the
collected sequences using various phylogenetic aoalyticai methods disthguished three

dBerent clades (groups, clusters) of HTLV-1 (Figure 1.1; Koralnik et al., 1994). The
cosmopolitan strain found in humans (clade H 1;Figure 1.1) ,which has k e n successfuiiy

transmitted to many populations in the world, represents a very homogenous group of

viruses that probably originated from West Afiica and was transported to the rest of the

world by the slave trades (Gessain et al. , 1992a; Gallo et al., 1993; Koralnik et al. ,
1994). The second cluster (clade H2; Figure 1.1) was identified in Centrai Afnca in the

Zairian basin and is referred to as HTLV-1-

(Gessain et al., 1992a). The third HTLV-1

clade (clade H3;Figure 1.1) was identified in remote inhabitants of Papua New Guinea

and the Solomon Islands and later in Abongines f'rom Austraiia (Gessain et al., 1991;
Yanagihara et al., 1991; Bastien et al., 1993; Gessain et al., 1993). Phylogenetic
analysis of the three Wal strains suggested that they arose from a cornmon ancestor with
the Melanesian clade diverging first, preceeding the evolutionary split between the

HTLV-1,

and cosmopolitan groups (Koralnik et al., 1994).

Figure 1.1
A simplified version of a neighbour-joining phylogeaetic tree of HTLV and STLV
isolates (reprinted with permission; Franchini, 1995). H l corresponds to the HTLV-1
cosmopolitan clade, H2 to the HTtV- and EI3 to the HTLVMdgroup. The simian
clades (S) are numbered as descriid in the text of the chapter.

HUMAN
PAPlO
CHIMPANZEE

LI

ri

HUMAN

-'

~ERcommcus
CHlMPANZEE

ERCOPITHECUS
PAPlO

] S2

7 S6

_r

1-

CERCOPITHECUS

MACAQUE

] S6

Similar evolutioiiary analysis of simian T-ceii leukemia virus type 1 (STLV-I)

strains of Afrcan an Man origin lead to severai conclusions (Figure 1.1; Koralnik et

al. , 1994; Franchini 1995). Interiestingiy, STLV-1 isolated from a single species of

nonhuman primate hqyently was sorteci into several genetidly distinct clades. In fact,

STLV-1 fkom Cercopithecw am*

P a T ~ ~ t(common
e s chimpanzee) and the

Papio baboon were distinguished into different phylogenetic clades, respectively into

clades S3 and S6, S2 and S5, and S 4 and S7 (Figure 1.1; Koralnik

et

al., 1994;

Franchini 1995).

These findirigs suggested there have been recent interspecies traosfers between
species within the primate gewra, including humans. In fact, the human clade H2 and
the cornmon chimpanzee clade S5 clustered together, suggestiog that the HTLV-123irr

clade rnight have resulted fiom a cross-species transmission of chimpanzee STLV-1 to

humans. Figure 1.1 summarizes bat at least three examples of interspecies transmission

that are supported by the phylogenetic analysis of STLV-1 from Afncaa primates with

different geographical ongins (Franchini. 19%). In the equatorial region of Africa, the
HTLV-1 and STLV-1 clades S2, S3, S5 and H2 are grouped by their geographkal origin
rather than by their species. In addition, viral strains obtained from a West AEncan

baboon also clustered with the cosmopolitan H l clade. Similarly, in Asia, the S1 clade
contains heterogeneous members in which the HTLV-1 Melanesia is nested.

11

These resuits affirmeci the evolution of three clades in the human species and
suggested that at least three independent human simian exchanges have occurred during
the evolution of these retrovinses. Franchini (1995) bas mggesteci an interpretation of

the naniral fiistory of these retrovinises, consistent with the phylogenetic, epidemiologic

and geographic data. The evolutionary mode1 suggests that ancestors of HTLV-1 and
STLV-1 entered primates in Asia and were transmitted to multiple species. Primates
infected with STLVs migrated to Afnca and STLVs were transmited to severai primate

genera (CerBoihecus, Papio, Pan and human). Meanwhile, somewhere in southwest

Asia, HTLV-1 emerged in Melanesia. Recent human migrating patterns, including the
slave trades. led to the dissemination of the cosmopolitan HTLV-1 worldwide.

1.4.1 Relatedness of primate TceU leukeminllymphoma &mes (PTLVs)

STLV-1 and human T-celI leukemia Wus type II (HTLV-II) are genetically

related to HTLV-1, exhibithg seyence homology of approximately 95% and 60%

respectively (Seiki et al., 1983; Shimotono et al., 1985; Watanabe et al., 1985). As
expected from the genetic similarities, STLV-1 Section has k e n associated with
hematological abnormalities in a variety of nonhuman primates (Miyoshi et al., 1982;
Sakakibara et al., 1986; Schatzi et al., 1992). Despite the genetic relatedness of HTLV-II
with HTLV-1, HTLV-il has not yet been etioiogidy liaked to disease; however,

recently HTLV-II infection has been observed in an increasing number of patients with

12

either T e l l rnalipnaltcies or a neurologicai disorder characterized by ataxia and

spasticity of the lower extrernities (hiynaraman et al., 1982; Lairmore et al., 1990;

Hjelie et al., 1992; Loughraa et al., 1992; Zucker-Franklin et al., 1992; Sheremata et
al., 1993; Iacobson et ai., 1993). Polymerase chah reactioa (Pa)and se~uencingdata

has established that the PTLVs coilectively, evolved from a common ancestor of bovine

leukemia virus (BLV);folowing this evolutionary split, HTLV-1 and STLV-1 diverged

from HTLV-II (Poiesz et al., 1993).

Not surprisingiy, naturai anti'bodes to HTLV-1, HTLV-LI ancl STLV-1 exhibit

antigenic cross-reactivi~and thus sensitive serological assays are re-d

to definitively

characterize and distinguish PTLV infections (Poiesz et ai., 1993). Despite antigenic
similarity, the HTLV-II envelope glycoprotein is significantly different nom the structure
of the HTLV-1 envelope glycoprotein, as suggested by weak neutralization of HTLV-II

syncytium formation by a n t i i i e s From HTLV-I carriers (Clapham et ai., 1984). In

addition, minimal cross-reactivity between the lioear epitopes involved in the

neutraiization of the two related vinises has been observed, despite the^ comparable
locations within the enveiope glycoprotein structures (Palker et al., 1992).

The recent description of very divergent STLVs in two species of African

primates (Gin et al. , 1994;

L
i
u

et al., 1994; Goubau et al., 1994), the pygmy

chimpanzees (Pan paniscw) and baboons fiom Ethiopia increases the complexity of the

ongin and evolution of STLVs and HTLVs. The STLVpaPpisolated from pygmy

13

chimpanzees that live excluively in Cenaal M c a (Giri et al., 1994) is more genetically

related to HTLV-II than to HTLV-1 (Giri et ai., 1994; LN et al, , 1994). The PTLV-L
strain isolated

h m the Papio timnadrya baboons appears to be phylogenetidy

equidistant between HTLV-1 and HTLV-IZ (Goubau et al., 1994). The identification of
these evolutionarily distinct viruses raises the question of the existence of other related

viruses in humans.

1.5 Molecular Biology of ETLV-1

HTLV-1 is a type-C retrovirus belonging to the BLV-HTLV gewra (Coffi,

1992), and contains a single positive stranded RNA genome 9.0 kiiobases (kb) long

(Seiki et al., 1983). Figure 1.2 depicts the genetic organjzation of the HTLV-1 genome

and the subgenomic messages produced from differential splicing of the genome. HTLV1, like other type-C retrovinws, contains three major coding regions designatted gag,pol

and env (Yosbida, 1987; Yashiki et al. , 1987). The gag (group-specif ic antige@-protease
region encodes the main structural proteins of the viral core and the virus-specific
protease which is responsible for cleaving the gag precursor protein into its mature forms
(p19, p24 and p15). The pol gene encodes both the viral reverse transcriptase and

integrase; while the envelope surface glycoprotein gp46 and its associated transmembrane

Figure 1.2
A schematic representation of the virai RNA genome structure and the encoded proteins
of HTLV-1 (reprinted with permission; Franchini, 1995).

R US

POL

16

anchor giycoprotein gp21, are encoded by the env gene. The HTLV-1 long terminal
repais (LTR),iike the LTR's of other type-C retrovinises, are located at the 5' and 3'
of the viral genome and confain virai transcriptional regulatory sequemes such as

enhancers, TATA and CAAT motifs as weil as sequences to which virai and cellular

transactivation proteins interact to reguiate transcription of HTLV-Igenes (Greene et al.,

1986;Nyborg et al., 1988). However, unlike most other type-C retrovinises, the HTLV1 genome increases its complexity by encodiag an additional region h o w n as X (Seiki

et al. , 1985). The X region contains at least four open reading fiames (off- o p ) which
appear to be utilized through alternative splicing of viral mRNA and possibiy by the use
of a recently descri'bed internai promoter (Nosaka et al., 1993). The novel regulatory Tax
and Rex proteins are encoded by the X region (Sodroski et al., 1984; Cann et al., 1985;

Felber et al. , 1985;Kiyokawa et al. , 1985; Hidaka et al., 1988;houe et al. , 1991), in
addition to a number of awciliary proteins p12', p13", p30n and p21rd (Orita et ai.,

1991;Koralnr et al., 1993) whose functions have yet to be M y elucidated. The Tax
protein has beeu cleariy demonstrated to be responsible for the transactivation of
transcription initiating from the HTLV-1 long terminal repeat (LTR) (Sodroski et al..

1984;Cann et al., 1985; Felber et al., 1985; Yoshida, 1987; for a ment review, see
Yoshida 1994). The Rex protein is responsible for positively reguiating the levels of
rnessenger RNA (mRNA)for the Gag, Pol and Env proteins while dom-regulating the

double-spliced TaxlRex mRNA (Hidaka et al. , 1988; houe et al., 1991).

1.5.2 Reguiatory functions mediated by Tax

The traiisfonnirig abity of HTLV-1 does wt depend on sites of virai insemon

nor on the expression of an oncogene (Seti et al., 1984; Yoshida, 1987; Yoshida and
Fujisawa, 1992). Lnstead, the tumomgenic potentid of HTLV-1 appears to be mediated,
at least in part, by the Tax protein, of whkh the= is no n o d cellular homologue. As

a nuclear 42-kDa phosphoprotein, Tax, is capable of binding to various host regdatory

proteins such as CREB, CREM, NF-gB p50, NF-A3 pl05 and SRF which bind to
enhancer DNA and ultimately resuit in transcriptional transactivationof not oniy HTLV-1

genes from the viral LTR,but also of various cellular genes (Suzuki et al., 1993; Zhao
et al., 1991; Hirai et al., 1992). Meed, it has been documented that Tax can activate
the expression of many cellular genes including those of interleukh-1 (IL-l), IL-2,IL-2

receptor a chah (L-~Rcx),IL-3, L 6 , c-fos, c-sis, c-myc, parathyroid hormone-related

protein (PTHrP), tumour growth factor-fl1 (TGF-fll), major histocompatibility class 1
(MHC class 1), and granulocyte-macrophage colony-stimuiating factor (GM-CSF)

(Greene et al., 1986; Iwwet al., 1986; Maruyama et al., 1987; Fuji et al., 1988; Wano

et al. , 1988; Nagata et al. , 1989; Ramer e? al. , 1989; Kim et al. , 1990; Watanabe et al. ,
1990; Nishida et al., 1991). Thus Tax appears to contribute to the transformation of

HTLV-1 infected cells which lose nonnal growth regdation by cytokines and
constituitively express U R , GM-CSF, interleukin 5 O,interleuLin 6 (IL-6) and
gamma-interferon (IFNy) (Miyatake et al., 1988; houe et al., 1986; Wano er al.,
1988). It is thus proposed that this dysregulated cytokine production together with

18

illegitimate transactivation evenfs will set up an autocrine der paracrine selfstimulating mechaniSm whereby HTLV-1 infecteci T-ceiis may uncontroiiably proLiferate
and ultimately become mimoaalized andor traasfonned (Maruyama et al., 1987;
Goebels et al., 1988). It has been suggested that one of the first mechanisms of T-celi

activatiodproliferation induced by Tax is mediateci by the autocrine lwp of IL-2liL-2R

chah activation (Maruyama et al., 1987; Goebels et al., 1988). Interestingly, Tax c m
be excreted from HTLV-1 infected celis and can stably exist extraceliularly (Marriott et

al.. 19Z). In addition, it has been demonstrateci that Tax is readily internaiized by

human periphed blood mononuclear ceiis (PBMC)and is capable of altering their

proliferation (Linciholm et al. , 1990; Marriott et al., 1991; Marriott et al. , 1992). This

observation suggests that through its extraceliular existence, Tax can exert its infuence

on celis at a proximal or distant site, which may explain some of the clinical symptoms
observed in ATL (ie. hypercaicemia by Tax-mediated activation of the parathyroid
hormone respome protein,

PTHrP;Watanabe et al., 1990). T'us the interaction of Tax

with cellular transcription factors appears to conm%ute to the pathogenesis of ATL,

HAM/TSP and other associateci diseases.

The activation/proIiferation of HTLV-1 infected T-cells is not the final event

leading to the generation of neoplastic T-celis. Events absequent to the self-stimulation

of HTLV-1 infected T-celis appear to be necessary for the development of fuli blown

acute ATL since elevated expression of IL-2R is not always accompanied by the
production of IL-2 in ATL celis (Arya et al. , 1984; Arima et al. , 1986). What these

19

subseqwnt events are remains to be elucidated, but they may require alterations in the

immune status of the irrfected individuah andlor M e r aiterations at the level of human

DNA since leukemic ATL T-ceUs have been demonstrated to contain various
chromosornal abnorrnalities (Sanada, 1986). The irrreased mutation frequency observed

in HTLV-1 iafcted cells may be attributed to the Tax protein which dom-regulates
expression of human DNA /3-polymemse, a cellular enzyme involved in host cell DNA
repair (leang et al., 1990).

The Tax-mediated cytokine dysreguiation in HTLV-1 infectecl cells may also be

involved in the neurologicai degeneration observed in HAM/TSP patients. The presence
of infiammatory cells including macrophages and T-cells in HAWTSP spinal cord

material, in addition to the upregulation of HLA and cytokine molecules (Wu et al.,
1993; Umehara et al., 1994) suggests that an increase in immunologie activity in the

central nenrous system (CNS)may be related to the pathogenesis of HAMfTSP. HTLV-1

presence and expression in the CNS has recently been c o n f h e d in three different
HAMITSP patient autopsies by RNA in situ hybridization (Lehky et al., 1995). The

rnolecular events involved in HAWTSP have still to be elucidated, however several

hypothesis have been proposed: (a) a slow virai infection of the nervous system by
HTLV-1 which might cause neuronal degeneration; (b) an immune response by cytotoxic

T-cells a g a h t HTLV-1 infecteci neuronal cells or autibodydependent cellular cytotoxicity

which induces destruction of neuronal cells; (c) demyelination caused by cytokines
secreted fiom HTLV-1 infected microglial celis; and (d) an autoimmune response in

20

which T-cells reactive to a muronai seif-antigen are activated as a result of HTLV-1

infection (Hara et al., 1994).

1.5.3 Reguiatory fmctions of Reg

Rex is a nucleolar phosphoprotein which regdates the balance of single- and

double-spliced versus unspliced viral mRNAs necessary for Wal replication (Kiyokawa
et al. , 1985; Nagashima et al. , 1986; Hidaka et al., 1988; houe et al. , 1991). Rex has
a positive effect on the expression of both the single-spLicd mRNA encoding the

envelope gene and the unspliced virai genomic RNA for the Gagmol proteins.
Altematively, Rex has a negative effect on the splicing/cransport of double-spliced
rnRNAs that encode for Rex itself, Tax, and the other altematively spliced mRNAs. The
effect of Rex on mRNA levels is exerted in tram on the Rex-responsive cis-acting

element (Rex RE), a highly stable RNA stem-loop structure in the U3lR region of the

HTLV-1 3' LTR (Yoshida et al., 1987; Seiki et al., 1988; Hanly et al., 1989; Baiiaun
et al., 1991). Rex binds directly to the Rex RE (Ballaun et

al., 1991; Unge et al., 1991;

Bogerd et al., 1992). The formation of the Rex RE stem-loop structure is also crucial for
appropriate polyadenylation of virai RNAs (Derse et al., 1987; Ahmed et al., 1991). In
addition, Rex mediates the stabiiization of mRNA for the L 2 u chah by acting in tram
on the coding seqyence of the L 2 R a chah gene (Kanamon et al., 1990), as well as

indirectly potentiating IL-2 gene expression in concert with Tax (McGuire et al., 1993).
Thus Rex not ody regulates Wal expression but it also interferes with host ce11 functions

21
by affecthg the pst-transcription and translation of ceiluiar genes. These observations
suggest that the effects of Rex in conjmction with Tax, are relevant to the pathogenesis

of HTLV-1-associated diseases.

An auxillary mechanism involved in transformation of HTLV-1 infected celis may

be attributed to the recently characterized pl2' protein (Koralnik et al., 1993). This
highiy conserveci and hydrophobie HTLV-I protein (Ciminaie m al., 1992; Ciminale et

al., 1995) has been shown to be a weak oncogene through its cooperation with the bovine

papilloma virus ES protein in the traiisformation of murine fibroblasts (Schlegel et al.,

1986; Franchini et al., 1993). In addition to interacting with the H+vacuolarATPase,

commonly referred to as the proton pump, (Nelson 1989; Franchini e? al. , l993), p12'
specifically interacts with the /3 and y, chains of the L 2 R (Franchini 1995). The
biological cooseqyences of pl2' interaction with the L 2 R g and y, chaios are not fully
understood, maialy because of the diffculty of analyzing its function in primary T-cells.
Mulloy et al. (1996) have recently observed that pl2' is capable of stabilizing the

immature forms of the IL-2w and 1~ chains, thus decreasing thek celi Wace
expression. It has also k e n suggested that the binding of pl2' to the L 2 R chaias may
alter receptor signaihg (Franchini 1995). Interference by pl2' in normal celi membrane
events has been suggested by the observation that p12' binds to the same region of the

22
IL-2w chab that intetacts with ~ 5 6a ~
lymphocyie
.
kinase involved in T-celi activation

and I L 2 signaihg (Hatakeyama et al., 1991; Franchini 1995). Interestingly,

dysregulation of IL-2-iaduced protein kinases in HTLV-1 infited ceils has been reporteci
by several research gmips (Yamiinashi et al., 1989; Koga et al., 1991; Mills et al.,
1992). Inappropriate signahg and kinase activation mediated through the IL-2R may

result from pl;?-mediateci mutual binding of the fl aad .y,chaias; whose juxtaposition has
been reporte to be crucial in kinase activation and IL2 signailing (Nakamura et al.,
1994; Nelson et al., 1994). In addition, the y, chab of IL-2R is part of at l e s t four

other receptors (IL-4, IL-7, Il-9and L I S ; Kondo et al., 1993; Kondo et ai., 1994;
Grabstein et al., 1994) and thus p12' might affect the fiinction of these receptors as well;
possibly contributing to the inappropriate cellular activation, transformation and ultimate
IL-2 independence observed in some HTLV-1 infected cells (Migone et al., 1995; Xu et
al., 1995). Thus, the interaction of p12' with the various inter1eukin receptors may have

important implications in the immune suppressive effect of HTLV-1 in vivo as wel as in

the ligand-independent HTLV-1-mediated T-cell proliferation (Mailoy et al., in press).

From these initial characterization studies of p d , a potentiai role of the protein in the
pathogenesis of ATL and HAM/TSP seems highly probable.

HTLV-1 is g e n e d y considered to be T-lymphocyte tropic and to infect

predominantiy the helper T e l l subset d e f ' by the surface antigen C D 4 Earlest
characterization of HTLV-1 infected leukemic cells in ATL patients revealed that they
were usually CDepositive (CM4) T lymphocytes (Catovsky et al., 1982; Hattori et al.,

198 1). Later in vitro d i e s confirmed that the CD4+phenotype was predominant among

vims-cafcyiog immortazed T-ce11 w s , obtained either upon cultivation of peripheral

blood mononuclear ceils (PBMC) b m HTLV-1 carriers or by cocultivation of

uninfected cord blood mononuclear celis with lethally irradiateci HTLV-I producing ceils
(Markbarn et al., 1983; Popovic et al., 1983; Sugamura et al., 1984b; Gessain et al.,
1990). Employment of similar culture techniques have sometimes yielded HTLV-1

infected B-lymphocyte nes and CD8' T-cell nes (Ruscetti et al., 1983; Longo et al.,
1984; Sugamura et al., 1984a/b), but it has not been established whether these ceiis are

targets for infection in vivo. Both cloned T-cell nes and bulk T-cell populations
expressing CD8 have been shown to be susceptible to experimental Mixtion and

immortaiizationby HTLV-1, but attempu to infect B lymphocytes in vino have generally

been u~lsuccessful(Ruscetti et al., 1983; Krichbaum-Stenger et al., 1987; Macchi et al.,
1987; Faiier et ai., 1988).

To more definitively detennine the in vivo tropism of HTLV-1, ce11 fhctionation

and polymerase chain reaction were used by Richardson et al. (1990) to establish the

24
phenotype of H'ILV-1 infectesi celis in the peripheral blood of HAM/TSP patients and

asymptomatic carriers. In aii subjects, HTL,V-1 was predomhantiy associated with CD4*
lymphocytes restricted to T-celis expressing the CWSRO antigen. Currently, the
CD45RO' subset of T-celis is thought to fepresem previowly activated or memory cells
(Smith et al., 1986; Sanders et al., 1988). The dominant presence of provirai H n V - I

in this mature cell phenotype is consistent with the recognize requirement of ceii
activation for permissive iofection by human and animal retrovinises (Varmus and

Swanstrom, 1982). In vitro snidies have shown that resting cells are not permissive for
productive infection by retrovinses (Varmus and Swanstrom, 1982; Zack et al., l99O),
apparently because of a block at the level of proviral DNA synthesis or integration which

can be overcome if the cells are stimulateci after exposure to the virus (Richardson et al.,
1990). In keeping with these observations, it has been suggested that T lymphocytes may

be infected by HTLV-1 regardless of their state of maturation, but uniess the cells are

subsequently activated (thereby becoming postive for CD45RO), the infection will be

abortive (Richardson et al. , IWO).

The ceil-type tropism of HTLV-1 does not appear to be restricted to lymphoid

lineages, as suggested by various in vitro studies. Productive HTLV-1 infection of
primary endothelid and glial cells, as wel as several non-lymphoid tumour lines has
been described (Clapham et al., 1983; Hoxie et al., 1984; Kiramam et al., 1986;

Macchi et al. , 1987; Akagi et al., 1988; Saida et al., 1988). Interestingly, HTLV-I has
been found to infect several neuronal ceii types in vitro, including oligodendrocytes, the

25

cells which secrete and maintah the myelin sheet of the spinal chord (Watabe et al.,
1989; Graziani et al., 1993; h h L y et al., 1994). Lehky et al. (1994) rtxently described

HTLV-1 infection of neuroblastoma ceUs with a concornmitant induction of MHC class

II expression on the ceii surface. This observation suggests the possibility that induction
of MHC class II m o l d e s on the surface of neuronal and glial ce& (following HTLV-1

infection) might increase theu suscephiility as a target for destruction, which may be
clinicaily significant in the pathogenesk of HAWTSP.

Despite the observation that the major target of HTLV-1 infection in vivo are
CD4' T-lymphocytes (90-99%;Richardson et al., 199), the cellular CD4 molecule does

not appear to be the viral receptor directly involveci in infection, as suggested by the
iriability of soluble CD4 moleniles or mti-CD4 monoclonal antibodies to inhibit HTLV-1

infection (Dalgleish and Richardson, 1990). As of yet, the target ceii receptor
determinhg the tropisrn of HTLV-1 has not been defined, however the gene encoding it

bas been rnapped to human chromosome 17 (Sommerfelt et al., 1988).

The species tropism of HTLV-1 is not restricted to humaos. HTLV-I is also

capable of iafecting simian, rabbit, cat, rat and hamster lymphocytes (Eguchi et al.,
1988; Seto el ai., 1987; Yoshiki et al., 1987; Nakamura et al., 1987; Lairmore et al.,
1991). This ailows monkeys, rabbits, cats and rodents to be used as animal models of

human infction by HTLV-1. Despite the ability of HTLV-1 to infect monkeys and
rabbits, these animals do not exhibit histopathological or clinicai sigm of disease

26

(Cockereli et al., 1990; Lairmore et al., 1991). On the other hand, immuuosuppressed

rats and hamsters infected with HTLV-1 are suseptiile to disease, exhibithg lymphomas
of host origin (Yoshiki et al., 1987; Eguchi et al., 1988). As descnibed in more detail

in chapter 4, it appears that the age and species of rats determines the extent of pathology
upon HTLV-1 infection (Yoshiki et al.. 1987; Eguchi et al. , 1988; Yoshiki et al., 1992;
Ishiguro et al. 1992; chapter 4).

1.7 Modes of Transmission

In vivo, HTLV-1 is principally ceil-associated and thus it is thought to be

transmitted primarily by repeated intimate contact (e.g., in utero, breastfeeding, sema1
intesourse) and direct tramfer of whole blood (e.g., tramdision, needle sharing) (
M
m

and Blattner, 1991; Blattner and Gallo, 1994). In endemic areas, the major route of

transmission is fkom mother to chiid with about 25% of children bom of HTLV-1
infected mothers becoming infected. mostly through the breast mk, but for 5% of

infants that are infected during the biahiog process or tramplacentally (Hino, 1990). In
Japan and other industrially advanced countries, the avoidance of breast feeding bas

prevented the majority of matemal transmission (Ichimani et al., 1991); however this
practice cannot be proposed in other endemic areas of deveioping tropical countries
where formula is fkequently unaffordable and maternai antibodies in the breast mille are
protective against other serious infkctious agents. Sexual transmission of HTLV-1 occurs,

27

mostly fiom men to women (Kajiyama @ al., l986), but it is not reasonable to expect

that this wiii be controiied by protected sex in most eodemic tropical countries due to

culturai and economic c o ~ t s Trammission

.
by blood trarisfusion (Okochi et al.,
1984) can potentially be eliminated by screening of blood donations, as is practised in

Japan, Caoada, the United States, France and certain islands of the West Indies; however

the cost of these diagnostic kits are a major obstacle in other endemic areas. In summary ,
the application of all aforementioned protective measures can be expected to have a

significant impact in weLldeveloped countries iike Japan, but elsewhere there is a definite
need for a vaccine.

1.8 m u n e respo~lsesassociated with protection against ETLV-1 infection

1.8.1 Passive immunhtion of a n h &

Transmission of HTLV-I by blood transfusion in adult rabbits has k e n

successfuily prevented by passive immU0i;ration with HTLV-1-immunehuman or rabbit
senun (Takehara et al., 1989; Kataoka et al., 1990; Tanaka et al., 1993). In addition,

newborn bllnnies were protected against &-borne

transmission rom their infected

rabbit mothers by immunoglobulin prophylaxis (Sawada et al., 1991). Thus circulating

antibody appears to be an efficacious defense mechanism against HTLV-1 infection, even
against ceU-associated virus devered to the oral mucosa. The success of passive

28

immunization cannot be denied; however its ernployment as a coaunonly used form of

prophylaxis in humaris is doubdul, since protection requirw infusion of the large arnounts
of HTZ,V-1 immune giobulin shortiy after exposure to the virus. For example, protection
of adult rabbits required infusion 24 to 48 hours p s t HTLV-1 challenge with each cabbit
receiving lmi of 77mg IgG/ml (Takehara et al., 1989; Kataoka et al., 1990; Tanaka et

al., 1993).

The protective effect of a n t i i i e s during mucosai challenge with HTLV-1 is

corroborated in humans by evidence that maternai anti'bodies fkom HTLV-1 infected
mothers protect infants against milk-borne transmission (Talcahashi et al., 1991). This

study of HTLV-1 endemic regions of Japan revealed that the probability of

seroconversion is related to the duration of breastfeeding. Antibodies in the mother's
breast mk are present during the f h t 6-12 months of Me, after which time a
progressive declhe in the titres are observed, with a c o c f e ~ ~ ~ o dincreased
ing
proportion
of children becoming HTLV-1 positive. If breastfeeding is restricted to the fmt 6 months
of life the probability (4.4%) is similar to that of bonle-fed children (5.7%), who have
probably acquired their Wus transplacentally; but the probabiiity rises to 14.4% in
ctiildren breastfed for 14 months or longer. The implications are that breastfeeding is safe
as long as materna1 antibodies are present, and that most milk-borne transmission of

29

HTLV-1 takes place after the age of 6 months. This suggests that active or passive

immunization to sustain neutrazing antf'bodies throughout a fil period of lactation

should prevent mdk-borne trammission, and that the first few months after birth may

provide a convenient window for vaccination.

1.8.3 Possible Mechaaisms of Antibody-mediated Rotedion

The protective effect of anti'body against HTLV-1 infection might appear

surprising in view of the fact that the virus has histoncally been considered to be
primarily eraasmitted in a ceil-associated form. Ceii-ftee aansmission of HTLV-1 had

k e n previously dismissai based on the circumstantial evidence that HTLV-1 was not
transmitted efficiently by contaminated cryopreserved factor VIII preparations (Okochi
et al., 1984) and the observation that in vin0 transmission of celi-free HTLV-I was far

less efficient than by coculairing infected ceUs with target ceiis (De Rossi et al., 1985).

However, HTLV-1 c m exist as independent virions since cell-free aansmission of HTLV1 has been accomplished in monkeys (Yamamoto et al., 1984)and rabbits (Seto et al.,

1991). It has thus been hypothesized that ceil-to-ceil transmission of HTLV-1 may be

mediated by virions with a short biological half-fe, allowing it to be successfully

neutraiized by antiody (De Th and Bomford, 1993). A potential problem may be
anticipated if the local antiaody levels induced by a vaccine f a to a point where viral
infection from ce11 to ce11 over a short distance can take place.

30
A numbet of Vinis-antiiy interactions have been demonstrated in M m , which

could act to conml virus infections, including HTLV-1, in vivo. Neutralization of free
Wus is expected to be of major imponance, both in preveoting infetion and in limiting
the spread of infection in extracellular fluids. The variety of different mechanisms by

which vinises can be neutralized, such as aggregation of virions, hihibition of attachment

of virus to the ceil, inhibition of virus penetration, inhibition of virus uncoating have

been reviewed extensively by Dimmock (1993). The spread of infection could also be
Limited by antibody which prevents the direct cell-to-cell transmission of virus, such as
antibodies to the F protein of paramyxoviruses (Men et al., 1980) or inhibits the release
of progeny vus nom infected celis, such as antibodies to the neuraminidase protein of

influenza virus (Becht et al. 1971). Antibody may also c o n m i t e to recovery fkom an
established infection by lysis of virus iafected ceiis. This may be mediated by a n t i i y -

dependent celi-rnediated cytotoxicity (ADCC) or complementdependent lysis (CDL)

involving the classicai or alternative complement pathways (Sissons and Oldstone, 1980).

The isotype of the antibody generated during an immune response appears to have
signifcantimplications on protection in vivo. h t i b o d y isotype has been found to affect
the tissue distribution of the a n t i i y , play a role in the interaction of the antibody with

ceils bearhg the appropriate Fc receptor, in the activation of complement and in the
aggregation of virions (Dimmock, 1993). For example, human IgGl and IgG3 are good

activators of the ciassical complement pathway, and only human IgGl and to a lesser

31

enent IgG3 mediate ADCC (Bruggemann et al., 1987; Lucisano Va-

and Lachman;

1991). The importance of IgG isotype in protection agaiast lymphocytic choriomeningitiis

virus (LCMV) infection of mice was suggested by the observation that di the protective

monoclonal a n h i e s (MAbs) were of the IgG2a isotype (Baidridge and Buchmeier,

1992).

The biological fiinction of anhaodes and ant'body-dependent celi-rnediated

cytotoxicity in clearing virai infections has k e n observed in severai systems. Substantial
evidence has demonsuated that ceus with Fc receptors are involveci in viral clearance of
several vinises such as LCMV (Baldridge and Buchmeier, l m ) , foot-and-mouthdisease
virus (McCullough et al., 1986), Semliki forest virus (Wust et al., 1987) Venezuelan

equine encephaiomyelitis virus (Mathews et al., 1985) and herpes simplex virus

(McKendaii, 1985). ADCC agakt HTLV-1 and STLV-1 infected cells has ais0 been

correlated with protection of pig-tailed rnacacpes against STLV-I infection, foiiowing

immunization with an inactivateci HTLV-1 virus preparation (Dezzutti et al. , 1990).

The role of complementdependent lysis in virai clearance in vivo is still in

question. CDL bas k e n observed to be effective in the in vitro destruction of virus
infected ceiis and membrane-enveloped virions such as HN-1, Sindbis and Respiratory

Syncytial virus (Schmaljohn et al. , 1983; Fuia et al. , 1992; Taylor, 1994). Although a

correlation between complement-dependentlysis and the abity of antiibody to protect has

been demonstrated in some virus infections (Schmaljohn et al., 1983; Schlesinger et al.,

32
1985), there is Iinle definitive evidence for a role of complement in antibody-mediated

protection U, vivo. Funhermore, complement does not appear to be of major importance

in virus infections of man as complement deficiencies appear to be associated with
increased suscepti'bility to certain bacterial, mther than virai, infections (Hsch, 1982;

Morgan md Walport, 1991).

When considering the prevention of milic-borne aarismission of viruses such as

HTLV-1, secretory IgA (S-IgA) amibodies are likely to play an important role in mucosal
immunity. The effectiveness of S-IgA is partiaily due to its physical existence at the sites
of initial contact of the viral pathogen with host cells (Miller et al., 1994). S-IgAlviral

binding in the mucus layer can prevent virus attachrnent to epithelial ceils and thus block
their subsequent penetration (Ogra et al., 1989). Another mechanism of v h s

neutraization by S-IgA has been observed in the infuenza virus system, where S-IgA
binding to a cei surface receptor d t s in signal transduction and ultimately blocks

transcriptional replication of the virus wiuiin the infecteci cei (Dimmock, 1984). In
addition, S-IgA has aiso ken found to be capable of blocking virus replication by
binding to nral proteins within the interior of epitheiial ceus, as demonstrated by its

blocking ability of Sendai virus replication (Mazanec et al., 1992).

1.8.4 Potentiiil role of cytotoxic Te&

in protection aganst ETLV-1

Despik extensive research into the development of HTLV-1 and HIV vaccines,
no consensus has yet been reached to clarify which components of the immune system
are effectively involvesi in coderring protection agaiost these Vuus infections. It appars
that several participants of the immune response, comprishg of both the humoral and

ceii-mediated effector arms, play synergistic roles during Wal infections. Indeed, it is
clear that virus-specific cytotoxic T-lymphocytes (CTL) are generated in moa, if not dl,
infections by human RNA and DNA viruses (as reviewed in Oldstone, 1993). A plethora
of experimental animal studies in which CTL are chemicaiiy or genetically deleted,

cornbined with reconstitution studies with CTL, attest to their role inthe control of many
viral infections (Oldstone, 1993). By d o g y to the many animal models of viral disease,
it is likely that

CTL directeci against H'LV-1 and HIV play a role as a host defense in

these infections. Indeed, activated HTLV-1-specific CTL has been observed in the

peripheral blood of healthy HTLV-1 carriers, ATL patients and individuals diagnosed
with HAM/TSP (Mitsuya et al., 1983; Kannagi et al., 1983; Parker et al., 1992;

Elovaara et al. , 1993). However the effectiveness of these circulating CTL in controllhg
HTLV-1 viremia in tbese individuals, bas yet to be determined. In addition, during the

testing of severai vaccine candidates, the involvernent of CTL in conferring protection

has been grossly overlooked (Nakamua et al., 1987; Shi& et al., 1987; Shida et al.,
1988; Tanaka et al., 1994; Franchini et al., 1995). The majonty of research studying

HTLV-1-specific CTL has been concerned with the presence of activated CD8+ T-cells

34

in the spinal cord lesions of W T S P patients (Jacobsen et al., 1991; Parker et al.,

1992; Parker et al., 1994). The presence of such a CTL response provides circumstantiai

evidence that these ceiis may contribute to the pathogenesis of H A ' T S P . Indeed, there
are several precedents in human and animal dwases where CD8+ T e i l s cause tissue
damage (mumps Wus meningitis, &th

et al., 1982; LCMV meningitis, Buchmeier et

al., 1980; hepatitis B virus hepatitis, Moriyama et al., 1990). On the other hand, the fact

that CD8+ T-cells are only found in the spinal cord lesions of late-stage W T S P , as

opposed to CD4+ T-ceils which are found eatlier, could reflect the immune system's
effort to bnit HTLV-1 h6ection thmugh CDS+ T-celis (Parker et al., 1994). Also
supporting a potentiai protective role of cytotoxic T lymphocytes in retroviral infections
such as HTLV-1, Hofnnan et al. (1991) reported that the presence of anti-virai CTL

appeared to be protective in mutine leukemia virus-induced hind-limb paralysis, if

present before disease induction. Currently, no correlations of CTL and protection

against HTW-1 infection c m be made. The lack of data strongly urges the need for
fiiture research to detenniw the precise role of the CTL respoose as it relates to

protection &or

pathogenesis of HTLV-1.

1.9 Envelope giycoprotein as a Vaccine TPrget

1.9.1 Properties of the HTLV-1 Envelope GIyeoprotein

Envelope glycoproteins of human and animal retroviruses contain antigenic

regions that play an important role in determinhg the host range, strain specificity and
neutralization pattern of the virus (Bauer, 1974). The envelope protein of HTLV-1
origiaates ft0r.n a 63

kD precursor glycoprotein (gp63) which is processed into an

extenor, glycosylated protein, gp46, and the gp21 hydrophobie transmembrane protein
which traverses the iipid bilayer of the virion aml ifected ceii (Lee et al., 1984;

Schneider er al., 1984). The envelope glycoprotein complex is essential in the early

stages of HTLV-1infection (Pique et al., 1990) and appears to be the most immunogenic
of aii the viral antigens (Lee et al., 1984). udeed, the viral envelope glycoprotein is the
major antigen recognited by the sera of HTLV-1 infecteci individuals (Lee et al., 1984;

Copland et al., 1986; Chen et al., 1989) and the anti-envelope antibodies are among the
e s t to appear following primary infection (Chen et al., 1990).

Host immunological respoases to the HTLV-1envelope glycoprotein following

natural infection have been extensively analyzed. Seroanalysis has been performed using
various heterologous recombinant envelope polypeptides expresseci in Escherichia c d i
(Chen et al., 1989; Chen et al., 1991). baculovirus (Nyunoya et al., 1990; Dekaban et
al. , 1994), vaccinia virus (Ford et al., 1992) and in yeast (Noraz et al., 1993). Synthetic

36

peptides denved h m HTLV-1 envelope sequences have also been employed to identify
immunodominant epitopes recognized by human IFnV-1-sempositive sera or by
monoclonal aatl'bodies (Paiker et al., 1989; Horai et al., 1991; Blomberg et al., 1992).
These studies revealed that the domipaot antgenic portion of the envelope protein is the
carboxy-teminai half of gp46 (amino acids (aa) 176-231) which is recognized by more

tban 95% of HTLV-1 patient sera. Analysis of the cellular immune responses of HTLV-1
infected subjects suggested that this irnmunogenc domain also encompasses a cytotoxic
TceU and helper T-ceii epitope (Jacobsen et al., 1991).

The HTLV-1 envelope protein is not only immunogenic, it also appears to be a

critical antigen for targeting the development of anti-viral imrnunity. Antibodies nom

HTLV-1 positive patients specific for the gp46 and gp21 envelope glycoproteins are

capable of inhibiting syncytium fonnation between HTLV-1 infected and susceptible

receptor-bearing cells (Nagy et al., 1983; Hoshino et al., 1983). In addition, these amienvelope anti'bodies are able to neutralize pseudotypes of vesicular stornatitis virus
bearing HTLV-1 envelope glycoproteins in vitro (Clapham et al., 1984). Tanaka and his
collegues have mapped one WUS-neutraiizing epitope of gp46 to aa 191-196 and

demonstrated that immunization of rabbits with a synthetic peptide containhg this

sequence could give rise to a gp46-specific neuaazing antibody response (Tanaka et al..
1991). Also, a human monoclonal antibady (HMAb) that reacts with aa 185-195 has been

shown to be capable of activating complement-mediated lysis and neutralzing HTLV-1

induced proliferation of lymphocytes (Matsushita et al. , 1988; Ralston et al., 1989).

37

Further investigations have suggested that the centrai region of gp46 between amino acids
186-216 contains multiple overlapping HTLV-1 neutraiization epitopes (Taoaka et al.,

1994; Sagara et al., 19%). Another neutrazing region has been mapped in the amino-

terminal part of HTLV-1 gp46 between aa 90 and 98 by Pallcer et al. (1992) using rabbit
ami-peptide sera. Most recently, a region of the trammembrane protein gp21 that

corresponds to aa 400-429 has been found to be involved in syncytium formation (Sagara

et al., 1996). Thus it appears that immune tesponses to various epitopes of the envelope

glycoprotein may be effective in inhibithg celi-to-cell spread of HTLV-1, and as

discussed below these immune responses may prove to be protective.

1.9.2 Genetic Variation of the HTLV-1 Envelope Gene

Secpence and irnmunological variation of the HTLV-1 envelope gene and its
product is very Limited between various viral strains (De Banun et al., 1991; Paine et

al., 1991). The HTLV-1 envelope gene is 97-98% conserveci among virus isolates fiom
Japau, Afiica, the Caribbean, an Centrai and South America (Gray et al., 1990;

Komurian et al., 1991; Schulz et al., 1991). However, gewticay distinct HTLV-1 types
from Melanesia have been recently identifiai with sequence divergence of 7% compared
to Japanese HTLV-1 strains (Sherman et al., 1992; Gessain et al., 1993). Undoubtedly,
these virus isolates wiil exhibit minor serological dfierences as suggested by the

observation of variation in binding of a monoclotial anhibody to two different envelope

38

proteins differing by a single amino acid ( E d o d er a1.,1994). However, despite

primary SBQuence variability, the neuttalizing domains of the cosmopolitan and

Melanesian strains of HTLV-I appear to be fiiactiodly indistinguishable as demonstrateci
by the protective effet of HTLV-1 immune globulin h m asymptomatically iafected

Japanese carriers agaiiist infection with a melanesian variant strain of HTLV-1 (Tanaka

et al., 1993). Indeed, serologicai analysis of divergent HTLV-I subtypes has revealed

that cross-reactivity and cross-neutralization does exist between human sera and
heterologous vinises nom different endemic regions (Clapham et al., 1984; Benson et
ai-, 1994). This suggests that antigenic variation of HTLV-1 will not present the
difficuities in vaccine development as those encountered with HIV (Rusche et al., 1988;

LaRosa et al. , 1990).

1.9.3 Active Jmmmizatioa of Recombinant HTLV-I Envelopbased

Subunit Vaccines

Envelope-based vaccine prototypes have conferred protection against HTLV-1

infection in monkeys, rabbits and rats. Protection of cynomolgus monkeys was attained
by injection of recombinant, nonnative Escherichia coli-derived beta-galactosidase fusion
proteins containing the amino- or carboxy-terminal halves of gp63 (Nakamura et al.,
1987). Despite success, this immunization regime cannot be utiiized in humans since the

truncated envelope proteins were delivered as an emulsion with Complete Freund's

1.9.4 Other Retrovirai Envelopcbad Vaccines

Envelope glycoproteins were first localized as spikes on the surface of avian

teukosis virions; related structures referred to as virai "knobs"were later detected on the
envelopes of mammnlirn retroviruses (Bolognesi et al., 1972; Nermut et al-, 1972). The
envelope protein of Friend-murine leukemia virus (F-MuLV) was the fim mammalian

type-Cretrovirai surface protein to be isolated and characterized (Moennig et al., 1974).

This purified envelope protein of F-MuLV was shown to exhibit type-specific as weil as
group-, species-, and interspecies-specificantigenic detenninants (Hull~mannet al. ,1974;
Fischinger et al., 1976). Immunization of mice with the putified envelope protein

(Hunsman et al., 1975) or a vaccinia-F-MuLV env recomblliant virus (Earl et al., 1986)
protected against F-MW-induced splenomegaly. Moreover F-MuLV envelope-specific
antisera could protect mice against splenomegaly and death from erythroleukemia upon

F-MuLV challenge (Hunsmann et al., 1975). In addition, protection mediated through

interspecies-specific antigenic determinants was demonstrated when immunkation with
murine F-MuLV envelope protein prevented tumour development in cats upon challenge
with the feline leukemia-sarcoma virus complex (FeLV; Schafer et al., 1977). Schafer
et al. (1977) showed that the f e l k leukemia v h s envelope protein also exhibits groupspecific determinana as suggested by the proteins's abity to elicit a n t i i i e s which

neutralize ail three types of feline leukemia virus. In contrast to demonstrating broad
spectnim neutralinng epitopes, a simian immunodeficieecy virus (SIV) envelope
glycoprotein-based subunit vaccine was able to confer protection in macaques ody when

41
the envelope glycoprotein of the challenge virus stock was genetidy closely-related or
identical to the envelope glycoprotein present in the vaccine preparation (Hu et al., 1992;
Giavedoni et al., 1993). Sunilarly, the envelope glycoprotein of a laboratory strain of
human immunodefiency virus was found to protect against challenge with the identical
but not heterogeneous virus strain(s) (Beniiaa et al., 1990;Girard et al., 1991). However
just recently, cross-protection between HIV-1 and HIV-2 has been reported by Abimiku

et al. (1995) following observations that a recombinant poxvinis vaccine expressing the

envelope protein of Hnr-1 could induce protection against HIV-2 challenge in rhesus
macaqws. The protection conferreci by these envelope-based vaccines against related
re~ovinisessuggests that employment of this same spategy may be effective in the

development of a global HTLV-1 vaccine.

1.10 Feasibiiity of generating a successful vaccine agahst HTLV-1

In summary, development of a vaccine against HTLV-1 may be attainable in

recognition of the genetic stability of the vins, the success of vaccination in various
animal models and the existence of naturai immunity in humans. Indeed, the concern of
genetic and immunologicat variation in the development of a vaccine for HIV (Rusche
et al., 1988;La Rosa et al. , IWO), may not be an obstacle in the design of an HTLV-I

vaccine where genetic variability is low (De Banun et al., 1991; Paine et al., 1991).
Phylogenetically related to HTLV-1,bovine leukemia virus (BLV)exhibits similar genetic

42

stability and causes a neoplastic disease in adult cade (Milier et al., 1969). Protection

studies with r e c o m b vaccinia

~
Wuses expressing the BLV envelope were capable of
varying degrees of protection depending on the construct used (Ohishi et al., 1991; Gatei
et al., 1993). Successful vaccines have already been developed for the relatively

geneticaily stable retrovinises aich as FeLV (Osterhaus et al., 1985; Eari et al., 1986;
Ishihara et al., 1991) and Simian a c e immunodeficiency syndrome (SAIDS; Marx
et al., 1986). Importantly, natural immunity against HTLV-I infection has repeatedly

been observeci in humam where matenial anti'bodies f3om HTLV-1 infected mothers

protect iafmts against millc-borne transmission (Takahashi et al., 1991). These protecrive
effects of anti'body are corroborated by the success of passive immunization of rabbits
with immunoglobulin obtained h m HTLV-1 infecteci asymptomatic humans or rabbits

against challenge with live HTLV-1 infected cells (Takehara et al., 1989; Kataoka et ai.,
1990; Sawada et al., 1991; Tanaka et al. , 1993). In addition, experimental data suggests
that it is possible to protect rabbits and monkeys rom HTLV-1 infection using subunit

vaccines containing E. coli produced envelope protein (Nakamura et al., 1987),

inactivateci whole virus preparations @ezzuti et al., 1990) and with recombinant vaccinia
vinises expressing the HTLV-1 envelope protein (Shida et al., 1987). Fortunately, several
animal models are available for testing of potential HTLV-1 vaccine candidates; including
rabbit, rat, hamster md monkey (Eguchi et al., 1988; Seto et al., 1987; Yashiki et al.,
1987; Nakamura et al., 1987; Laimore et al., 1991). The goal of an HTLV-1 vaccine

based on a delivery system appropriate for the developing world is worthwhe, since it
would protect the human population against a fulminant leukemia and an incapacitating

43

neurodegenerative disease. In addition, its success wouid potentially assist in hding

solutions for problems highiy relevant to the design of an AIDS vaccine.

1.11 Thesis objectives and contents

In the past, limited availability of HTLV-1 envelope protein had restrained

investigations into its role in HTLV-I infection, disease and protection potential. Previous
to the work reporteci in this thesis, the complete HTLV-1 envelope protein had not ken

successfuL1y expressed in a recombinant system in sufficient amounts to permit its

isolation and use. Only truncated or fusion proteins with smaller gene segments had been
successfully expressed in Escherichia coli (Kiyokawaet al., 1984;Samuel et al., 1984
and Chen et al., 1989). Expression of full or n w l y full length gp63 in EscheBchia coli
proved to be toxic to the hoa ceiis. HTLV-1 envelope protein expression in the

baculovinis system had been limiteci to recombinant polyhedrin fusion proteins (Nyunoya
et al. , 1990). Synthesis of the HTLV-1envelope glycoprotein by a mammalianexpression

vector had also proven to be toxic to the hoa murine cells (Vile et al., 1991). Indeed,
only very low levels of expression of the entire HTL,V-1 envelope protein had k e n
obtained in Saccharomyces cermsiae by Kuga et ai. (1986).

The goal of my research thesis was to generate recombinant envelope protein

preparations that would be immunogenic in various animai species, including humans.

In addition, it was necessary to estabsh an animal challenge mode1 to allow the rigorous

44

testing of the recombi.na.nt envelope subunit vaccine candidates for their protective
efficacy .

The foiiowing two chapters describe the expression, cbaracterization and use of
the HTLV-1 envelope protein produced in the baculovinis ami vaccinia vinis/T7
polymerase systems. The entire HTLV-1 envelope protein, gp63, was expressed in

quantitative amounts in the baculovinis system,however it was insoluble and the majority
of protein was incompletely pst-translationally processed. The isolated aggregates of
envelope protein demonstrated a considerable immunogenicity in various saains of mice

and were capable of stimulating envelope-specific T d s from naive humans in vitro.

This recombinant envelope protein was also used in a comparative isotype analysis of the
antibody response to the HTLV-1 envelope glycoprotein in HTLV-1 infected humans,

both asymptomatic and those with HAMITSP. Despite achievement in solubilizing the
baculovinisexpressed envelope protein, lack of native conformationsuggested that it was

not optimal for vaccine purposes; nor was it suitable for use in studies to identify the ceil

nuface receptor for HTLV-1.

For these reasons, the vacciniafT7 polymerase expression system was developed
as an alternative source of recombinant HTLV-1 surface envelope protein. The

mammalian system produced high levels of gp46 in a properly processed and folded
form. Conformational integrity of the recombinant protein was confinnecl by reactivity
with sera from HTLV-1 infected patients and a panel of confornation-dependent an&

45

envelope human mowcloiial a n t i i e s under non-denaturingconditions. Furthemore,

this recombinant gp46 exhibiteci mssreactivity with HTLV-II infected human sera and

sera nom a Pan paniscus chimparizee infected with the more distantly related STLVpw
suggesting that this recombinant envelope protein may serve as a useful diagnostic

reagent and potential vaccine candidate.

This thesis also includes a description of the successful HTLV-I infection of

Fischer F344 newbom aml adult rats through challenge with a live HTLV-1 infected ceU
line (Yoshn et al., 1992). In addition, the unprefedented infection of aewborn rat pups

with short-tenn cultured ceiis isolateci nom a W T S P patient, is also reported. To

insure realistic challenge parameters in the adult rat model, the minimum challenge dose
of HTLV-1 Uifected ceils was detennined Uiat would maintain a 100% frequency of
infection. With the recombinant envelope protein preparations available and the
establishment of the rat challenge model, it appears that ali the necessary reagents and

technology are available for future vaccine trials.

EXPRESSION AND IMMUNOGENICFI'Y OF THE ENTlRE

aUhlAN T-CELL LEUgEMIA VIRUS TYPE 1 ENmzOPE PROTEIN
PRODUCED IN A BACULOVIRUS SYSTEM
CaAPTER 2

In an attempt to study the role of the HTLV-1 envelope glycoprotein as an

immunogenic target and assess its vaccine potential against HTLV-1 infection, we

successfully expresseci the entire HTLV-1 envelope protein using a baculovinis vector
system. Previously, expressionof the HTLV-1 envelope protein in the baculovinis system
had k e n Limited to variant tnincations of the envelope fused to the amino-terminal

peptide of baculovirus polyhedrin protein (Nyunoya et al., 1990). The recombinant

polyhedrin fusion protein containhg the major portion of gp46 was produced in Low
yields and couid not be purifid (Nyunoya et al., 1990). To date, no attempts have k e n
made to express the entire HTL,V-1 envelope glycoprotein as a non-fusion protein in the

baculovinis system.

For expression of the entire HTLV-1 envelope glycoprotein, gp63, we employed

the baculovinis strain Aidogrtph califmica nucleopolyhedrosis virus as the recombinant

expression vector, and the ovarian ceii iine established fkom the moth Spodoptera
fnrgiperrda (SB)as the host. Nigh protein production in the baculovinis system derives

from the high efficiency of the viral prornoter of the polyhedrin gene. Polyhedrin protein
is the sole component of the crystalline matrix that acts as a protective shield for vira1

47

particles outside their insect host (Doeder and Bohm, 1986). Late in lytic iafection,
polyhedrin can account for as much as 60% of the total cellular protein in S. mgberda
ceils. Since polyhedrui is not required for virus replication and infetivity (Doerfer and

Bohm, 1986), expression vectors have been developed using the s w n g A c M W

polyhedrin promoter to control expression of foreign genes (Smith et al., 1983; Matsuura
et al., 1987). Insertion of f o ~ i g nDNA into the polyhedrin gene by homologous

recombination causes a deletion of a major portion of this gene and thus recombinant
virus c m be readily detected as polyhedrin-negative v h s in a simple plaque assay.

Since the development of the baculovinis systern for foreign gene expression by
Smith and his associates (1983), many different eukaryotic proteins have been

successfully produced (Luckow, et al. , 1988). However, the expression levels of viral
membrane proteins such as influenza vins hemagglutinin (Kwoda et al., 1983),
lymphocytic choriomeningitis Wus glycoprotein (Matsuura et al., 1987), Japanese
encephalitis virus envelope protein (Matsuura et al., 1989) and bovine herpesvbs 1

giycoprotein gN (Van D m e n Littel-Van Den Hurk et al., 1991) have generally been
lower than those of cytokines (Smith et al., 1983; 1985) and viral nucleocapsid proteins
of lymphocytic choriomeningitis virus (Matsuura et al., 1989). hepatitis B virus core

antigen (Takebara et al., 1988), rhesus rotavirus VP4 (Mackow et al., 1989) and simian
rotavirus VP6 (Estes et al., 1987). Throughout these expression studies, baculovirus

infected S. frugiperda celis were shown to perform many of the post-translational

modifications of higher eukaryotes (Miller et al., 1988) iirluding glycosylation (Smith

48

et al., 1983; Lehman et al., 1993), phosphoryation (Miyamoto et al., 1985), correct

signal peptide cleavage (Smith et a[., 1983) and removal of introns by proper splicing
(Jeang et al., 1987). These propeaies would prove to be important in attempting to
generate a recombinant HTLV-1 envelope glycoprotein that closely resembled its

authentic counterpart.

This chapter d e s c n i the successful expression of the entire envelope gene of

HTLV-1 in a baculovims non-fusion vector system. The HTLV-1 envelope protein

accumulated witbin the insect ceUs as inclusion bodies whkh aiiowed efficient recovery
of the recombinant protein. in an attempt to study the role of the HTLV-1 envelope

glycoprotein as an bunogenic target, mice were immuni;r_ed with the envelope protein

inclusion bodies (env-I.B.) in the presence or absence of an adpvant. In addition, a

combined immunization regimen was also assessed. (Mostof the findings presented in
this chapter have been published elsewhere; Arp et al., 1993).

2.2 Materials and Methods

2.2.1 CeU lines

The Spodopterafnrgperdcr (SB) celis were obtained from Dr. P. Faulkner (Dept. of
Microbiology and Immmology, Queen's University) and grown in TC10 medium
(GibcoIBRL) supplemented with 10% fetal caif senun (FCS, GibcolBRL), and 1 0 uglml

of gentamycin (Giko/BRL). The human T-lymphocyte cell ne UT-2was obtained from

Dr. L. Arthur (AIDS Vaccine Program, NCI-Frederick Cancer Research and

Development Center). Two other T-ceii lines, HTLV-I producing Cg1 PL cells and noninfected indicator C8166 cells (Dr. Tom Pdker, Duke University), were used in the
syncytium inhibition assay. Ail Tceil lines were maintaineci in RPMI 1640 supplemented

with 10% FCS and 100 unitshl of pemicillin and streptomycia

2.2.2 Plasmids and viruses

The envelope coding sequences for gp46 and gp21, were derived from the plasmid pMT2 (provided by Dr. Gallo, NCVNM, Ratner et al., 1985). The baculovinis expression
vector ~ 2 1 3 9 3(provided by Dr. M. Summers, Texas A&M Agricultural
Experimentation Station) was chosen as the vehicle for insertion of the HTLV-1 envelope
glycoprotein into the genome of Aufographa califomica multi-mclearpolyhedrosis virus
(AcMNPV, provided by Dr. P. Faulkner, Queen's University). The recombinant vaccinia

virus constmcts R W Els, R W E3s,and the anti-sense constnict R W Elas used in the
combined immunization regimen, have been described in detail (Ford et al., 1991).

50
Briefly, R W Els codes for the wd-type HTLV-I envelope precunor, with the protease

cleavage recognition sequence intact. R W E3s encodes gp46 only, with the insertion of
an in-frame stop codon a d the deletion of the remainder of the envelope gene C-termiad

to gp46. R W Elas is identical to R W Els except that the envelope gene is in the anti-

sense orientation.

2.2.3 Isolation of recombinant baculovinis

The recombinant mander vector pVLEITL encoding the entire wild-type envelope gene
fragment was transfefted to the AcMNPV genome by homologous recombination in vivo,
using a calcium phosphate precipitation technique specifically modified for insect cells
(Graham and van der Eb, 1973; Burand et al., 1980; Carstem et al., 1980; Summers and
Smith, 1987). Briefiy, the DNA precipitate was formed by the addition of calcium

chloride (CaCl,, final concentration of 0.125m.M)

to a mixture of 1 ug of AcMNPV

DNA, 2 ug of pVL3fTL DNA, 475 ul Hebs (137mMNaCI; 6mM D-glucose, 5mM KCl,
20mM HEPES; pH 7.1) and 15 ug of single-strandedcalf-thymus DNA. For transfection,
the precipitate was incubated on monolayers of Sf9 ceiis at 27C for 4 hours, at which

time the monolayea were washed with media aod allowed to incubate at 27C for 6 days.
Culture nipernatants of the advanced stage infections were hamested. Recombinant v h s
was ultimately isolated and purifid from the transfection supernataats by four rounds of

standard plaque assays involving 10-fold serial dilutions of the v i m (Summers and

Smith, 1987). Recombinant baculovhs could be readily distinguished from wild-type

virus plaques by the absence of the crystalline occlusion bodies composed of polyhedrin.

51
2.2.4 DNA extnction and Southem blot analysis

Sf9 celis were infiited with wild-type or recombinant bacuiovim at a multiplicity of

infection (moi) of 0.1 plaque forming unit5 per ceii @Wceii)and virus harvested 48

hours post infation (p.i.). Viral DNA was purifieci nom extraceiiuiar virus isolated from
the culaire supematants (Summers and Smith, 1987) and digested with restriction
endonucleases HirtdIII: or BamHI. Electrophoresed virai DNA was transferred ftom a

0.7% agarose gel to Zetaprobe nylon membrane (BIORAD) in 0.4N sodium hydroxide
using a modification of the method by Southem (Southem, 1975) as described by the
manufacturer (BIORAD). Blots were rinsed in 2X SSC (standard sodium citrate: 0.3M
sodium chloride, 30mM sodium citrate, pH 7.0) to neutralize, exposed to ultraviolet light
to crosslink the transferred DNA and then baked at 80C for 30 minutes. The blot was

incubated at 42OC for 16 hours in prehybridization buffer (Dekaban and Bali, 1984;50%
V/V formamide, 6X SSC,

10X Denhardt's solution (Denhanit, l%6), 7% w/v SDS, 1mM

ethylenediaminetetraaceticacid (EDTA;pH &O), 100 uglml tRNA and 40ug/ml sheared

herring sperm DNA). To this prehybridization buffer was added 200ng of heat-denatured

DNA probe (1.7 kb HindLII-Pst1 HTLV-1 envelope fragment) labeiled wifh r2P]-dCTP

(1000Cilmmole, Amersham) by the random primiag method (Phamiacia Otigolabeiiing

kit) with a specific activity of lO%pm/ug. The blot was innibated in the presence of the

labelled probe for 24 hours, at which time it was washed twice for 15 min. in 2X
SSC/O. 1%

SDS at room temperature and twice for 1 hour in 0.1X SSC/0.1% SDS at

55C. The prepared blot was exposed to Kodak Cronex 4A X-ray film with Lightning-

52

plus intensifyirg screens for 40 minutes at room temperature. Generation of specific

DNA hgments was observed following repeated southern blot analysis.

2.2.5 Indirect immunofluocescence of bacuioviius infectecl ceUs

Individuai monolayers of SB ceUs were infected with wild-type bacuiovinis and the two
recombinant AcMNPV clones VHBS and VHB6 at a multiplicity of infection (moi) of

0.5 pWcefl. Three days post infection, the ceis were harvested, washed and resuspended
in phosphate buffered saline (pH 7.3, PBS), at a demity of 5x106 cells/ml. Ce11
suspensions were sponed on tissue-grip (Fischer Scientific) coated glas sdes, ailowed
to air-dry and directly fixed in cold acetone for 10 minutes. Slides were rinsed twice in

PBS (pH 6.8) and rinsed once in distilled water. Noospetific bindiiig was blocked by
incubating the washed cells with a 3% solution of bovine serum afbiimin (BSA,
Boehringer Mannheim)in PBS for 1 hour at r o m temperature. CeUs were incubated a

m e r 30 minutes in this blocking solution (for second a n t i i y controis) or exposed to

HTLV-1 patient sera at a f i dilution of 150 in blocking solution. The cells were rinsed

in PBS and exposed to a goat anti-human fluorescein isothiocyaoate conjugated antibody

(GAM-FITC, Jackson TmmunoResearch) at a Final dilution of 1: 100 in blocking solution
for 30 minutes at room temperature. Ceus were niwd again before m ~ ~ with
n 2.5%
~ g
Dabco (Sigma) in glycero1:PBS 9:1, pH 8.7. Immunofluorescence data that was
consistent between two independent assays was consideml valid. Photography was done

on an Olympus BH-2 fluorescence microscope with a fluorescein filter.

53
2.2.6 Detection of secrete or d associated H'M,V-1 envelope protein

Monolayers of 2x10' Sf9 ceiis were Uifected with either dd-type AcMNPV or
recombinant baculovinis (iafections of both clones, designated VHBS and CTtlB6 were
tested) at an moi of0.2 pWceil. Infcted ceiis were innibatcd in the presence of serumfree media (ExceU 400, JRH Biosciences). After 46 hours, the infecteci celis and
supernatants were harvested by centcaugation at 3 0 x g, 15 min., 4 O C in the presence
of the protease inhibitors, 1mM penyl-methyl-sulfony1-fluonde(PMSF;Boehringer

Mannheim) and ImM EDTA. The ceil pellets were stored at -70C.To remove the

majority of extracellular Vinis fkom the supernatant, the supernatant was centrifuged at
14,000 x g, 60 min., 4C. The supernatant was then dialyzed a g d t a buffer of O. lmM

EDTA, 1 m M Tris,pH 7.4 for two days at 4C. The volume of the diaiyzed supernatant
was then reduced to 200 ul by lyophikation.

Both the supematants and their

correspondhg celi peliets were resuspended in an equal volume of Laemlli buffer, boiled

for 5 min. and electophoresed on a 12%SDS-PAGE containing 6M urea (Hayden et al.,

1986). Gels were then processed for Western blot analysis. Four attempts were made
to detect HTLV-1 envelope proteins in the tissue culture supernatant.

2.2.7 Inciusion body isolation

S B cells were iafected with recombinant baculovims at an moi of 1.0 pfu/wll. Forty-su

hours p i . , the celis were chillecf for 10 minutes and pelieted at 10,000 x g, 20 min.,
4C. The cells were washed once in cell wash bmer consisting of 50mM Tris (pH 7 3 ,

1mM EDTA, 1mM dithiothreiotol @TT), 1mM PMSF (Nyunoya et al., 1990),

54
dispensed into 1x108 cefi aliqyots and the centrfigation was repeated. The crude ce11
lysate was resuspended in TD buffer (Nyunoya et ai., 1990)and homogenized in a glas

homogenizer in the prwence of DNase 1 [lmg/ml] (Boehringer Mannheim). The

suspension was centrifugai for 20 minutes at 10,000 x g. Pellets were sonicated in TD
buffer and innibated as d e s c n i (Nyuoya et ai., 1990). Future use of the pilets

decided the m e r processing of the env-1.B. (A) For mouse immunizations and in viro
cell proliferation assays, the env-I.B. pellets undement uee washes in PBS, pH 7.2

intersperseci by centrifugation at 10,000 x g, 20 min., 4 T . Pellets containing 10 ug of

env-I.B. were storeci at -70DCfoiiowing the nnal centrifugation and removal of the
supernatants. (B) For SDS-PAGEelectrophoresis, a pellet containing approximately 120
ug of env-I.B. (protein equivalent of 5x10' cells) was resuspended by sonication, in a

Laemili buffer containing 4M urea, lOOmM DTT and 1mM PMSF.This suspension was

incubated for 1 hour at 4OC on a nutator, &er which it was aliquotted into various
required qyantities for storage at -70C.
Polyhedrin occlusion bodies iom wild-type baculovinis infected ceils were

isolated in the same manner as the HTLV-1 envelope inclusion bodies and served as a
negative conml antigen for the in miro c d i assays.

2.2.8 Sdubization of inclusion bodies for in v h cell assays

Solubilization of isolated envelope inclusion body pellets for in M m ce11 proliferation

assays required sonication in phosphate buffer (pH7.4; Deutscher, 1990) supplemented
with 0.05 % SDS w/v, foliowed by gentle agitation for 1 hour at room temperature. The

55

solubilized sampie was piaced on ice for 15 minutes in the presence of octyl+Dthioglucoside (10 ug OTG: 1 ug env-I.B. protein; Pierce) and pelleted at 10,000x g, 4C
to remove the SDS precipitate. The d t i n g supernatant was fond to contain >95%
of the initial env-I.B. protein and proved to be non-toxic at 1 ugfmi when used as antigen

in T-celi proliferation assays (Manca et al., 1995). Efficient solubilization of the

envelope inclusion bodies has become a routine procedure in our laboratory with very
consistent results.

2.2.9 Characterization of the recomblit HTLV-1 envelope protein fornrr

Extent of glycosylation of the recombinant HTLV-1 envelope proteins was studied by

Western blot analysis of recombinant baculovinis iafections in the presence or absence
of the N-glycosy1ation inhi'bitor, 1 ugfml tunicamycin (Schwarz and Datema, 1982;

Boehringer Mannheim). This assay was repeated three times. To determine if the binding
of anti-SP-7 sera to the envelope proteins was specific, the inhibitory effects of a
saturating amount of SP-7peptide (2 micromole per millilitre= 2.4 mg SP-7/ml; 20 aa.
sequence derived fkom gp21) on anti'body bindhg (0.175 mg I g / d anti-SP-7 sera) was
analyzed by Western blot.

2.2.10 Western blot assay

Inclusion body extract, solubilized in 4M Urea reduchg buffet, was electrophoresed on

a 12%SDSI6M Urea polyacrylamide gel (Hayden et al., 1986). Foliowing equilibration
of the gel in Bjemun Schafer-Nielson tramfer buffer (Bjemun and Schafer-Nielson,

56
1986), proteins were eransferred to Immobilon P membrane (nylon; Mliipore) using a

semi-dryelectrophoretic mander apparatils (BIORAD) at 15Volts for 45 min. Blots were

blocked in Blotto (5% wfv dned mik powder, 0.1% vfv Tween 20,0.01% v/v antifoam
A, 0.000196 v/v thimersol in Tris-buffered d i n e pH 7.6; a modifieci version of Johnson

et al., 1984) for 5 hours at rom temperature, with hourly changes of the blocking

solution. Blots were incubated in the appropriate primary antiboy, diluted in blocking
buffer, ovemight at 4C with rocking. Blots were then washed in blocking buffer and

exposed to goat anti-mouse, goat anti-rabbit or goat anti-human IgG conjugated to

alkalioe phosphatase (Jackson ImmunoResearch Laboratories Inc.) at a final dilution of

L:5000, for 30 min. at room temperature. The blots were washed and exposed to the
colourmetric substrates nitro-blue tetrazolium and 5-bromo4chloro-3-indoly l phosphate
accordiog to the manufacturer's instruction (NB and BCIP; Blot Detection Kit,

Amersham International, PLC). Each Western blot assay included three positive controls
of (i) 1C11, an anti-gp46 mouse monoclonai anti'body, (Paiker et al., 1989a). (ii) anti-SP-

7 rabbit polyclonai serum (SP-7peptide sequence derived from gp21, Palker et al.,
1989a), and (i) human HTLV-1 patient sera

(TSP). Unstained high molecular mass

protein markers (BIORAD) and urisfained low molecular mass protein markers
(Phannacia) were used to d e t e d e the molecular mass of the electrophoresed proteins
separated on aii polyacrylamide gels.
A detailed description of the Western blot protocol used for comparative isotype

analysis of the mtibody response to the env-I.B. of HTLV-1 asymptomatic carriers and

HAM/TSP patients can be foued in Appendk 1 (Dekaban et al., 1994). The methods

used in the comparative study @ekaban et ai., 1994; Appeldix l) were similac to those

descn'bed in this chapter's material and methods section with the exception of the horse
radish peroxidase conjugated secondary antiiies which were specific for the various

immunoglobulia isotypes and IgG subclasses. As a result of the different enyme-liuked

conjugates, development of the blots invoived the enhancPrl cherniluminescent detection

system (E.C.L.; Amersham). Human samples considenxi to be positive for env-I.B. were
1/50 senun dilutions exhi'biting a signai foilowing a 2 min. exposure.

2.2-11 Immunizations

Three different inbred mouse strains, Mblc (Charles River), C57BL/6 (Charles River),

and CFWID (Ball and McCarter, 1979) were immunized at six to eight weeks of age by
intraperitoneal injection. Two different forms of HTLV-1 envelope protein immunogens

were studied. For the adpvant titration studies, mice were injected with 10 ug of envLB. in the absence or presence of various amounts of adjuvant formulated from a

mycobacterial celi wdi exact ( M m ,BionicheNetrepharm, London, Ontario).

MCWE is a purifieci and deproteinwd cell wall extract fiom a nonpathogenic species of

mycobacterium. The env-I.B. pellet was tesuspended by sonkation in either 500 ul PBS
(pH 7.2) or in 500 ul of a 1:2 dilution of PBS:MCWE emulsion. For the combined

immunkation regimen, mice were primed with an ntrapentoneal injection of 4x106

plaque-forming units of the appropriate purifieci recombinant vaccinia vims (RW Els,
R W E3s and RW Elas) dilutecl to 100 ul with RPMI 1640. RW Elas was used as a

negative control in the evaluation of the immunogenicity of the recombinants vaccinia

58

vinises encoding the seme orientations of the envelope gene fragments. After two and

four weeb, the primeci mice were boosted with either the same vaccinia preparation or
with 10 ug of env-I.B. suspeded in 100 ui of 100 ugllml MCWE adjuvant preparation.

Animals were tenninally bled two weeks after the second boost.

HTLV-1 infecteci human UJ-2 cens were labellecl with fSS]-cysteine for

imrnunoprecipitation of HTLV-I envelope proteins by serum of mice h u n i z e d with

HTLV-1envelope inclusion bodies. 2x10' cells were washed in cysteine-free RPMI 1640
(GibcoIBRL Selectamine kit) with 1% dialyzed FCS, pelleteci and irtfubated 30 min. at

37C with gentle mixing in cysteine-keemedia plus 1% FCS. The ceiis were pelieted
and resuspended in cysteine-freemedia containhg 0.5 mCi [3SS]-cysteiw(1000 Ci/mmol;
Dupont, NEN) and incubatecl 5 hours at 37C with mixing. The radiolabeffed ceiis were
rinsed twice in senun-free RPMI 1640 and resuspended in 2 ml extraction buffer

(100mMNaCl, 1 % Triton X-100, 0.5 % wlv sodium deoxycholate, O. 1%w/v SDS, 1mM
EDTA, lOmM phosphate pH 7.6; Dekaban et al., 1984). The cell suspension was
supplemented with 2.5mM PMSF and homogenized to lyse the cells. Ceil debris was

pelleted by sequentiai centrifiigation at 4OC. The &tant

celi Iysate was precleared for

2 hours at 4OC by incubathg with pre-immune s e m preabsorbed with Protein G PluslA

agarose (Oncogene Science Inc.). The Protein G Plus/A agarose complexes were pelieted
and the resuiting precleared supernatant was aliquotted into 5x106 ceus equivalents. For

the mouse test sera, each aliquot of celi lysate was suspended in a total of 1 ml extraction

buffer conaining 20 ui immune mouse serum and 30 ul of Protein G PludA agarose.

The positive control samples consisted of an aliqwt of precleared labeiied celi lysate

incubateci with 30 ul of Protein G P b / A agarose and a mixture of rabbit polyclonal

HTLV-1 enveiope anti-peptide sera raised agahst peptides SP-2,SP-4A, SP-6 and SP-7

(Palker et al., 1989a). Immune complexes were allowed to fonn overnight at 4OC,
washed with cold extraction buffet and then resuspended in an equal volume of 2x
Laemiii buffer before loading onto a 12% SDS polyacrylamide gel (SDS-PAGE)

containing 6M tuea (Hayden et al., 1986). Gels were stained with Coomassie blue, fued

for fluorography with Entensify Solution (Dupont, NEN), dned 2 hours at 80C under
vacuum and exposed for autoradiography at -70C. Uastained high molecular mass

protein markers (BIORAD) and unstained low molecular mass protein markers
(Pharrnacia) were used to determine the molecular mass of imrnunoprecipitated proteins
separated on a i l polyacrylamide gels.

2.2.13 Peptide ELISA

Binding of senun antibodies to HTtV-1 envelope synthetic peptides by ELISA was

performed as previously described (Palker et al., 1989a) with the following exceptions:
2 ug of peptide per microtiter weil was used; and for efficient blocking, the reaction

buffer contained 2% w/v dned milk instead of 5% wlv BSA. The foUowing synthetic

peptides containing hydrophilic sequences from HTLV-1 gp46 or gp21 were chosen for
the study: SP-2(gp46, envelope aa 86-107); SP4A (gp46, aa 190-209); SP-6 (gp46, aa
296-312); and SP-7 (gp21, aa 374-392), ail of which were obtahed from Dr. T. Palker

60

and have been pceviously descnbed (Palker et al., 1989a; Figure 2.1b) . AU synthetic
peptides were synthesized on an A p p M Biosystems 43 1A Synthesizer (Foster City, Ca)
ai Duke University with chemids and program cycles provided by the manufacturer.

Endpoint titer was defined as the senun dilution at which the signal to noise ratio was

> 2.0; the mean opticai demity reading obtained with mouse sera injected with

1 0 ug

MCTKE aione, was used to estimate background readings for ELISA.

2.2.14 Sgneytium inhibition assay

Syncytium Uihibition assays were used to masure neutralizing antibody titres and were

performed by Dr. Tom Palker and bis Iab associates at Duke University, North Carolina.
This assay detects the presence of neutrahing antibodies in semm by monitoring its
effects on the ability of EfLZV-1 infkcted cells to fuse with uninfecteci target ceiis, as

readily measured by the countiog of giant ceiis (Nagy et al., 1983; La1 et al., 1991).

More specificaiiy, the syncytium inhibition assay involves the incubation of 45 ul of

HTLV-1 infect& Cg1 PL T-ceUs with uninfcted C8166 T-ceiis (each ceii line at 106
ceiis/rni in RPMI with 10% FCS)ovetnight ina tissue cuiture incubator (5% CO2,37C)

in the presence of 10 ul of seriaily-diluted test semm (heat-inactivated at 56T, 30

minutes). After 24 hours, the presence of syncytia was evaiuated on an inverted
microscope at 200x magnification and the neutralizing titer was determined as the

1st

senun dilution which inhibited syncytium formation by greater thm 90 1.Routinely, 100200 syncytia could be obtained per microtiter weU in the presence of 10%wrmal mouse

61

senun. AU mouse sera were coded prior to testing in q m q t b m inhiiition assay, and

codes were broken oniy after neutnziog titers had been measured. Neualiziag, antiHTLV-1 peptide antisera (Pmet al., 1992) and pre-immune serum served as positive

and negative controls, respectiveiy.

2.3.1 Construction and identtcafion of tecombinant bacdovirus

The entire HTLV-1 envelope coding sequence was isolated nom the plasmid pMT-2
(Ratner et al., 1985) by a BumHI-Pst1 partial digest. This 1636 base pair (bp) hgment
encoding gp46 and gp21 was inseaed into the baculovirus tramfer vector p K 1 3 9 3
downstream fcom the polyhedrin gene prornoter as indicated in Figure 2.la. In this

p W T L constmct, the translation initiation codon of the envelope gene is located 123
bp downstream fkom the nonfuIctiona1start codon of polyhedriu gene and thus wiiI result
in expression of the complete HTLV-I envelope protein in the absence of additionai

polyhedrin protein sequences. The pVLHTL piasmid was then transfected together with
AcMNPV DNA into insect tissue culture c a s (SB) and virus was isolated fiom
occlusion-negative plawes .

DNA extracted fiom wild-type and two different recombinant baculovirus isolates
(VHBS and VHB6) was digeste with Hirrdar or BmnHI and subjected to Southern blot
aoalysis. Hindm digestion of both recombinant bacuiovirus isolates resulted in the
generation of a 10.6 kb fragment, which was capable of hybridizhg with an HTLV-1
envelope specific probe upon Southem blot analysis (Figure 2.2). This 10.6 kb fragment

presumably represents the 1.6 kb HTLV-1 envelope sequeoce attached to 9.0 kb of

baculovinis sequence, as expected. BamHI digestion and Southem blot arialysis of

Figure 2.1

A. The gene transfer vecor p W T L containing the entire HTLV-I envelope

glycoprotein gene used to generate the recombinant baculovinis expressing the HTLV-1
envelope glycoprotein. Horizontal cross-hatch, 5' polyhedrin seqwnces; Vertical crosshatch, 3' multiple-cloning site seqyences; Dotted region, 3' polyhedrin seqwnces
including polyadenylation addition signal.

B. The location of the synthetic peptides used in the peptide ELISA to determiDe
antibody reactivity to various regions of the HTLV-1 envelope proteins.

Transcriptio
Start

A Signal
Pst 1

HTLW I env fragment

gp2l

AUG
(Pst 1)

Truislational
Stop

cleavage
Site

Tanslational
Start
wrption

(Barn H1)

Figure 2.2
Southem blot hybridization of Hindm or BonHI digested DNA nom extracellular wildtype (wt)and recombinant baculovirus (isolates VHBS and VHB6). The blot was probed
with a 1.7 kb HindIII-Pst1HTLV-1 envelope fragment labeiled with r2P]-dCTP. An
intemal 1.1 kb envelope gene fragment gewrated upon BcmrHZ digestion of recombinant
baculovinrs DNA is indicated with an arrow,

Hind III

BamHI

wt VHW VH66

wt VHBS VHM

67

recombinant bacuiovirusDNA (Figure 2.2) revealed an intemal envelope fragment of 1.O

kb, as expected. The 2.7 kb &anAI fragment represents the 3' portion of the HTLV-1
envelope region attachai to dowristream bacdovirus sequences. The 3.6 kb hgment
redted h m incornpiete B m H I digestion of recombinant VHBS DNA. The envelope

probe did not hybridize with wild-type bacuiovirus DNA.

2.3.2 Identification and isolation of the BTLV-1 envelope glycoprotein

To determine whether the envelope gene was expresseci by the recombinant baculovinis,
indirect ixnmunofiuorescence ushg HTLV-1 patient sera was performed. As illustrated

in Figure 2.3, normal S B insect celis and those infected with wild-type bacdovirus failed
to fluoresce. while both recombinant baculovinis isolates (VHBS and MIB6) revealed
strong positive fluorescence. There appeared to be aggregates of protein at the poles of
several envelope-expressing cells. while other recombinant bacuiovim infected cells

were stippled in appearance, suggesthg that they may be sequestering the envelope
protein within vacuoles.

In light of the induect immunofluorescence observations, it was important to

determine whether any of the HTLV-1 envelope protein was king secreted by the S B
cells. Western blots of the supernatant and correspondhg ceU peliet of a recombinant
baculovinis infection (Figure! 2.4, panel B) revealed that the HTLV-1 envelope
glycoprotein was not king secreted from the infected SB cells but was accumulating

Figure 2.3
Indirect immunofluorescence analysis with anti-HTLV-1 envelope gp46 monoclonal
antibody (1Cll). A. uninfected Spodoptera frugiperrda (SB) cells; B. wild-type
baculovinis infected Sf9 cells; C. recombinant VHBS bacuiovinis infeted S B celis; D.
recombinant VHB6 baculovirus infected Sf9 ceils.

Figure 2.4
Western blot anaiysis of supernatants and ceii pellets folIowuhg recombinant baculovims
infection.

Panels A:

supernatant and cell pellet of wild-type baculovirus infection.

Panels B:

supernatant and cell pellet of recombinant baculovirus infection.

AU blots were probed with the same series of primary antibodies: lane 1 and 2, normal
mouse controls; lane 3 anti-ppQg l C l l moue MAb; lane 4 and 5 , normal human
controls; lane 6 am1 7,human HTLV-1 patient sera (TSP patients).

72

withk the cells. The recombinant baculovirus infectai ce11 peilet contained three major

size classes of HTLV-1 envelope protein with molecular mass averaging 43, 54 and 63
kiiodaitons (kD). Al1 three size classes were recognzed by both the anti-gp46 l C l l
monoclonal anti'body and human HTLV-1 patient sera (Figure 2.4, lanes 3, 6 and 7
respectively). In addition, several d o r protein bands of lower molecular mass ranging
in size fkom 30 to 39 kD were dso observed. No specific immunoreactive proteins were
observed in the supernatant and cell peilet of a wiid-type baculoWus infection (Figure
2.4, panel A). The results of the Western blots and the indirect immunofluorescence of

recombinant baculovinis infecteci cells, combined with the requirement of denaturing

agents (4M m a , 4M guanidinium hydrochloride or SDS)for solubilization of the HTLV1 envelope protein, suggested that the recombinant baculovinis infected cells stored the
HTLV-1 envelope protein as inclusion bodies.

The accumulation of the HTLV-1 envelope proteins as inclusion bodies, ailowed

for their isolation fiom other cellular and viral proteins with greater than 80%purity as

detennined by SDS-polyacrylamide gel electrophoresis. A modification of the method

developed by Nyunoya (1990) allowed for the enrichment of these insoluble protein

aggregates as iilustrated in Figure 2.5a. In our method, the addition of DNase 1 was
critical in obtaining maKimm purification of the envelope inclusion bodies (env-I.B.).

Isolation of env-[.B. from an equivalent amount of uifected cells (Figure 2.5a. compare
lanes 3 and 4), resuited in the recovery of the three major HTLV-1 envelope protein
forms with minimal loss. A minor protein band with a molecular mass of approximately

Figure 2.5
Analysis of HTLV-1 envelope proteins produced as inclusion bodies by recombinant
baculovirus.

a) SDS-PAGEillu~fratingexpression and isolation of HTLV-1 envelope inclusion

bodies. Lane 1, uninfecteci Spodoptera fng@er& (SB) ceb; lane 2, wiid-type
baculovinis infectai Sf9 cells; lane 3, recombinant baculovinis infected SB ceiis; lane

4, purifed inclusion body preparation (env-I.B .).

b) Western blot anaiysis of inclusion bodies (env-1.B .).Lane A, mouse anti-gp46

1C11 monoclonal antibody; lane B, human HTLV-1 patient sera (TSP patient); lane C,
rabbit anti-SP-7peptide sera; lane D, cornpetition assay with SP-7 peptide and rabbit

anti-SP-7 peptide sera. No Western blot reactivity was observeci with normal mouse,
hwnan and rabbit sera.

75

30 kD was also observed. No HTLV-1 envelope proteins were detected by Western blot

anaiysis in the celi lysate supernatmts during the inclusion body isolation (data not
shown). The 43, 54, and 63 kD immmoreactive envelope proteins previously obsewed

in the total celi pellet (Figure 2.4, Panel B) were present in the same relative amounts
within the inciusion bodies (data not shown).

When an equivalent amount of elecirophoresed env-I.B. (Figure ZSa, lane 4) was

analyzed by Western blot, the majority of extracted inclusion body material proved to be
of HTLV-I envelope origin, as shown in Figure 2.5b (lanes A-D). The three major

proteins of 43, 54 and 63 kD present in the inclusion bodies proved to be

immunoreactive with both ana-#

l C l l monoclonal antibody (mAb specific for a

peptide sequence derived fkom gp46; Figure 2 3 , lane A) and anti-SP-7 peptide sera

(polyclonal sera specific for a peptide sequence denved om gp21; Figure 2 3 , lane C)
and thus suggested that they represent d . e m tforms of the HTLV-1 envelope precwsor

protein. To CO&

the specificity of the anti-SP-7 sera and thus coofirm the precursor

origin of the various envelope protein fonns, Western blot analysis was performed in the

presence of cornpetitor peptide. The cornpetition Western blot revealed that SP-7peptide
was capable of inhibithg the binding of anti-SW sera to the 43 kD and 63 kD proteins
(Figure 2Sb, compare lanes C and D). However, the SP-7peptide did not completely
inhibit the binding of the anti-SP-7 s e m to the 54 kD protein (Figure 2 S b , lane D) .

76

The reason for this was not clear. Control experiments using normal mouse and rabbit

sera, or sera raised agabst Sf9 cells iafected with unrelated recombinant bacuiovinis, did
wt possess a n t i e s capable of binding to the 54 kD protein (data not shown).

Conversely, HTLV-1 specific sera did not react with Western blots of unrelated
recombinant badovinis cell pellets (data not shown).

To determine if any of the three major envelope protein forms were the result of
glycosylation, the effects of the N-glycosylation inhibitor, tunicamycin (Schwarz and
Daterna, l982), were studied. Tunicamycin treatment resulted in the disappearance of the

63 kD protein; however it had no effect on the production of the 43 and 54 k D proteins

(data not shown). This suggested that the 63 kD protein was the glycosylated precursor
representing 5-10% of the total envelope protein, while the 43 and 54 kD proteks were
unglycosylated HTLV-1 envelope precursor fonns.

2.3.4 Immuwgenicity of the inclusion bodies in vivo

Radioimmunoprecipitation and Western blot assays revealed that injection of mice with

inclusion bodies (env-LB), in the absence of adjuvant, could stimulate humoral responses
to the HTLV-1 envelope protein. Senun nom immunized C57BW6 mice possessed

antibodies capable of immunoprecipitathg HTLV-1 envelope proteins from [US]-cysteine

metabolically-labeiled HTLV-1 infecteci UC? cells (Figure 2.6a). Normal C57BU6 sera
did not immunoprecipitate these HTLV-1 envelope proteins.

Western blot analysis

77
confirmecl the reactivity of the sera h m the imrnunized C57BU6 mice to HTLV-I
envelope proteins (&ta not shown). Sera h m Balb/c and CFWlD mice immunized with
env-I.B. aione exhiiited similar humoral responses to the HTLV-1 envelope protein, as

monitored by radioimmunoprecipitation and Western blot assays (data w t shown).

To stimulate an elevated humoral response to env-I.B.. a mycobactenal ceil wall

extract (
a

M m ,Archambault et al., 1989) was employed as an adbvmt. Since this was

new adjuvant. a titration experiment was performed to determine the optimal dose of

MCWE (0-500 ug) required to give the best a n t i i y response. From Western blot and
radioimm~11oprecipitationassays, maximal seroconversion was observeci in mice
immunized with 50 ug of MCWE adjuvant preparation (Figure 2.6b). Exceeding this 50
ug dose of MCWE resulted in a graduai decrease in mouse seroconversion with

increasing amounts of MCWE adjuvant preparation (Figure 2 . 6 ~500

; ug of MCWE).

2.3.5 Characterization of the Antibody Response to env-I.B.

In order to m e r study the effects of varying amounts of MCWE adjuvant on the

antibody response to env-I.B., sera was assayed by ELISA for the ability to bind to four
synthetic peptides; SP-2,SP-4A,SP-6 and SP-7. The locations of these peptides within
the HTLV-1 envelope proteins is shown in Figure 2.lb. The regions of the envelope

protein gp46, encompassed by the peptides SP-2 and SP4A, have been associated with

Figure 2.6
Imrnunogenicity of the HTLV-1 envelope inclusion bodies in the absence and presence
of MCWE adjuvant preparation. Radioimmunoprecipitationof HTLV-1 envelope proteins
from HTLV-1 Secteci MJ-2 celis with sera ftom individual CS7BW6 mice immmked
with env-I.B. in the presence of (a) no adjuvant; (b) 50 ug dose of MCWE adjuvant
preparation; (c) 500 ug dose of MCWE adpvant preparation. NMS, pooled serum from
normal unimmunized CS7BU6 mice; +VE,rabbit polyclonal anti-peptide serum raised
against envelope peptides SP-2,SP-4A, SP-6 and SP-7.

80

virus neutraiization, while the SP-6 peptide region of gp46 has k e n shown to
beimmunogenic in humans (Palker et al., 1989a; Tanaka et al., 1991; Horal et al.,
1991). The SP-7peptide spans another immunogenic region of the HTLV-1 envelope and

allowed us to monitor the immune respome to the trammembrane envelope protein,

gp21. Mice injectai with env-I.B. in the presence of 10 ug of MCWE adjuvant

preparation produced sera with the highest ELISA titers for aii four synthetic peptides
(Figure 2.7, Group 4) and exhiiited strong Western blot reactivity (data not shown). As

iliustrated in Figure 2.6, inoculation of higher doses of MCWE resuited in a

corresponding decrease of sera reactivity with the various peptides, with some mouse
sera from these high-dose groups completely failing to recognize any of the synthetic
peptides. Those mice that failed to generate antibody capable of remgniPng the four

synthetic peptides, also produced low levels of anti-envelope anti'body as detected by

Westem blotting and radioimmunoprecipitation(data not shown).

The peptide ELISA data helped to map the immunogenic regions of the

recombinant HTLV-1 envelope protein presented in the inclusion bodies. In ali groups,
env-I.B. gewrated the highest ELISA titers to the synthetic SP-6 peptide (Figure 2.7).

In fact, the level of SP-6-binding a n h i e s stimulated by env-1.B. injection was only

slightly influenced by the adjuvant dose received. Xntermediate antibody titers to the

synthetic SP-4A and SP-7 peptides were observeci, with the lowest antibody titers
directed to the SP-2peptide (Figure 2.7). AU env-I.B. irnmunized mouse sera were

Figure 2.7
Peptide ELISA titres of sera from mice immunized with env-I.B. and various amounts
of MCWE. Seroreactivity to the HTLV-1 envelope peptides SP-2,SP-4A, SP-6 and SP-7
was measured. AU groups with exception of group 1, received 10 ug of env-I.B. in the
presence of the appropriate amount of MCWE. Group 1, LOOug MCWE; group 2, ug
MC=; group 3, 5ug MCWE;group 4, lug MCWE;group 5, 25ug MCWE; group
6 , 5ug MCTKE; group 7, lOOug M m ; group 8, 250ug M m ; group 9, 500ug
MCWE. ELISA titre is the dilution that d t e d in an opticd density equal to, or
greater, than twice background values obtained with contml mouse sera injected with
lOug of MCWE alme (group 1). Titres l e s than 50' were considered as values of O.

Group

83

compared to sera h m mice injectai with only MCWE (group 1) to determine signicant
ELISA titers,

The various mouse sera were also screened in a syncytium inhibition assay to
determine if neutdizhg a n t i i e s were generated. Neunaluing antibody titea of 10-40
were only observed in a few mice receiving the highest doses of MCWE (250 ug and 500
ug; data not shown). These doses produced undesirable side effects in the mice, as

judged from observations of their generd health.

Previous experiments (Ford et al., 1992)had shown that R W Els expressing the native

HTLV-I envelope (gp46 and gp21; Figure 2.8), and R W E3s expressing only the
surface glycoprotein (gp46;Figure 2.8) were capable of inducing neutralizing antiboies.

In an effort to enhance neutralizing antibody titers directed to the HTLV-1 envelope, envI.B. was injected i . combination with R W Els and R W E3s. R W Elas which

contains the anti-sense version of the native envelope gene was used as a conuol.
Following priming with either R W Els or R W E3s, the mice were boosted t e e with

env-I.B. in the presence of 10 ug of MCWE adjuvant. This dose of adjuvant was chosen
because it generated optimal antibody titers to the biologidy significant SP-2and SP-4A
regions of gp46 (Figure 2.7). The redting sera were characterized by Western

Figure 2.8
Recombinant vacciaia virus constructs (RVVs) utilized in the combined immunization
regirnens in vivo. R W El encodes the wild-type HTLV-1 envelope protein precursor
with the cleavage recognition sequence intact. Post-translational cleavage of gp63 by host
ce11 proteases yields the two mature envelope pcoteins, gp46 and gp21. To coiistnict
R W E3, the coding seqyence of gp21 was removed and two termination codons were
inserted at the 3' end of the gp46 coding sequence, such that this recombinant vaccinia
virus only expresses the surface envelope glycoprotein.

86

blot anaiysis, peptide ELISA aod syncytium inhibition assay (neutraijzation assay) and
the results are swnmarized in Table 2.1. Western blot reactivity was recorded using a

grading systern as iliustrated in Figure 2.9 (weak, moderate and strong reactivity were

indicated respectively as

+, ++, +++).

As determineci by

Western blot analysis, priming with R W Els (Table 2.1,

Group B) and R W E3s (Table 2.1, Group D) when combined with boosts of env-I.B.,
increased the overail antiibody response to the HTLV-1 envelope protein, when compared
to

immunization with either R W Els or R W E3s alone (Table 2.1, Groups A and C),

or env-I.B. alone (Table 2.1, Groups E and F). This did not translate into increased

ELISA antibody titers to the specific SP-2and SP-4Aregions of gp46, when compared
to the titers elicited by env-I.B. alone (Table 2.1, compare Groups B and D with E and
F). We did not test reactivity to the SP-6 region since it is not associated with vims

neutraikation.

Priming with R W E3s did e h c e neunaliting antibody titers when used in

combinaUon with env-I.B. (Table 2.1, compare Groups C, D, E and F). The prllning
injection of R W E3s appeared to be necessary for the generation of neutralizing
antibodies; since mice prirned with the R W Elas or media alone, followed by env-I.B.
boosts, did not produce antibodies with this biological activity (Table 2.1, Groups D , E

and F). Interestingly, as previously demonstrated (Ford et al.. 1992). multiple injections

of R W Els (Table 2.1, Group A) efficientiy induced neutralizing antibody in the

Figure 2.9

Monitoring anti-envelope seroconversion. Western blot reactivity of six C57BV6 mouse

sera screened on days 14, 28 and 37 during the course of combinecl immunization with
RW E1S and env-I.B. (Group B of Table 2.1, Mice No. 7-12). Samples were taken
prior to each env-I.B. Most and upon termination. Reactivity was graded: +, weak;
+ +, moderate; + + +, strong. No Western blot reactivity was observed with any sera
collected from negative controls: Group F and G.

Characterization of the immune response hduced in mice immunized with the combineci
recombinant vaccinia virus (RW) and env-I.B. regime (number of mice with observed
reactivitylnumber of mice per group) .
Western blot reactivity was graded as the followhg:
+ + + strong.

+ weak, + + moderate,

Peptide ELlSA titres depicted in the table are the calcuiated means of positive
samples only. Individual UISA titres were determined as the dilution that
resulted in an optical density equal to, or greater, than twice background values
obtained with control mouse sera (injected with 10 ug of MCWE doue).

Neutralization titre is the highest dilution that inhibiteci HTLV-1 syncytium

formation by greater than 90% relative to nomai mouse senun controls.

Imm-Jlm

D~Y

SPt

SP4A

SI7

mtiiy
(titreC')

91

absence of signifiant antibody titers to the SP-2 and SP-4A regions. Conversely,
boosting iojections of env-I.B., foilowing a prime with R W Els (Table 2.1, Group B)
resulted in the production of ant'bodies to the SP-2,4, and -7 peptides; however this

sera was not capable of significantly inhibithg syncytium formation.

2.3.7 0th- studies determinjng the antigenicity/imm~~~ogeaicity

of env-1.B.

The demonstrated antigenicity of the env-I.B. preparation indicated that it could also be
used for the screenhg of various sera coilected nom high nsk HTLV-1 humans. In a
coilaborative study undertaken with Elaine King, a large comparative analysis was
performed to assess the anaaody response to the HTLV-1 Gag and Env proteins in 39
HTLV-1 asymptomatic carriers and 10 HAM/TSP patients (Dekaban et al., 1994; see
Appendix 1). My conm'bution to this study involved detennining aii of the

immuaoglobulin isotype and IgG subclass anhibody profiles and respective titres specific
for the HTLV-I envelope protein for each individual sera (Dekaban et al. , 1994). Elaine
King was responsible for an equal conm'bution to the characterization of the

seroreactivities to the Gag proteins (Dekaban et al. , 1994).

Figure 2.10 iilustrates some of the interesthg ciifferences in the isotype responses
observed between the two groups of individuais (Dekaban et al., see Appendix 1 for the
entire study). The IgG response was the only exception, where 100% of the

a s y m p t o d c s and the W T S P patients reacted to the envelope proteins. Ail the other
isotype respomes demonstrated marked distinctions between the two groups. In the

Figure 2.10
Per cent reactivity of HTLV-1 infecteci asymptomatic and HAWTSP sera to env-I.B. for
each of the imrnunoglobulin (Ig) isotypes and the IgG subtypes, as determined by
Western blot assay. The level of significance between the two groups is given below each
Ig isotype of IgG subclass, as determinai by the two-tailed Fishers Exact Test. Soiid
bars, HTLV-1 asymptomatic individuais; Crom-hatched bars, HAMITSP patients.

env

env

1O 0

80
CI

8
t. 60

>
*-

5
CO

40

20

O
P

lgM IgG IgA IgE

4.661 NS 0.020 eO.001

lgGllgG2 1963 1964

P

NS NS 4.003 NS

94

asymptomatic group, IgM and IgE antibody to the envelope protein was not detected;
while 60% of asymptornatics were positive for IgA. This lack of maturation of the an&

envelope a n t i i y cesponse observeci in asymptomatics, was in sharp conaast to the

anti'body profles of the HAWTSP group; all of whom had IgM, IgG, IgA, and IgE

responses specific for the envelope protein.

In another joint study ( M m et al., 1995; Appendix II), the antigenicity and
immunogenicity of the env-1.B was tested for its ability to generate specific CD4' T-cell
lines from HTLV-1 seronegative individuais. In the above study, the recombinam HTLV1 envelope protein (in both the insoluble inclusion body and solubilized f o m ) served as

target antigens. The freqwncy of T-helper ceLi precursoa specific for HTLV-1 envelope
in the naive repertoire is f a too low for detection by a conventional proliferation assay,

as seen in the case of HIV antigens (Mancaet al., 199l), but the relevant cens could be
expanded in vitro by repeated stimulation with various envelope preparations (Manca et
al., 1995; Appendix II). Resuits from this investigation revealed that T-celi lines could

be induced by in vitro immunization of the envelope protein in particulaie or soluble form

in 9 out of 10 naive individuals. Polyhedrin occlusion bodies rom wild-type bacdovirus

infected S B ceiis, isolated in the same manner as the HTLV-1 envelope inclusion bodies,
served as negative control antigen for these in vitro c d assays. This paper (Manca et al.,
1995), hcluded in Appendi~~
II of this thesis, describes in detail the specificity and clonal

heterogeneity of the C M + T-ceU Lines that recognized env-I.B..

2.4 Discussion

As d e s c n i in this chapter, expression of the entire HTLV-1 envelope glycoprotein,

consisting of gp46 and gp21, has ken successfully achieved using a recombinant
baculovUus system. Previously, HTLV-1 envelope protein expression in the baculovirus
system had ken Iimited to recombinant polyhedtin fusion proteins (Nyuwya et al.,
1990). With the newly gained ability to express a full-length protein, the role of the

HTLV-I envelope glycoprotein as an imrnunogenic target in human HTLV-1 infections

and its vaccine potential against HTLV-I infection, could be examined.

Efficient expression of the HTLV-1 envelope protein at levels up to 6 milligrams

per litre of ceil culture medium was observed in this newly constnicted recombinant

baculovim system. The protein was not secreted kom the host ceils but rather was
stored intracelluiarly as inclusion bodies. The insolubility of the recombinant protein

simplifed the isolation of the prorein aggregates. The envelope protei.obtained rom the
p w i e d inclusion bodies consisted of three major size classes of protein averaging about
43 kD,54 kD and 63 kD.

The biochemical nature of the recombinant envelope protein forms was

detemrined. Western blot analysis revealed that the 63 kD protein was of precursor origin
and was equivalent in molecular mass to the expected glycosylated envelope precursor
gp63 of HTLV-1 (Seci et al., 1983; Pique et al., 1992). The glycosylated nature of the

96
recombinant 63 kD protein was confirmecl by its shift in electrophoresis mobility upon
tunicamycin treatment. The 54 kD and 43 kD proteins were also fourd to span the
epitopes recognized by the lC1l monoclonai a n t i i y specificaiiy binds SP-4Apeptide)
and anti-SP-7 polyclonai sera. This seroreactivity revealed that the 54 kD and 43 kD
protein forms contained amino acid sequemes encompassing both m V - 1 gp46 and
gp21. Thus in combination with the twucamycin treatment results, the 54 kD and 43 kD

proteins appeared to represent unglycosylated variant forms of the envelope precusor.

From the HTLV-I mcleotide sequence (Se& et al., 1983; Ratner et al., 1985), it may
be predicted that the 54 kD protein represents the unglycosylated envelope precursor with

its leader peptide still attached. Unfortunately, the precursor origin of the recombinant
54 kD protein cannot be confirmed despite its ability to bind with anti-SP-7 peptide sera,
since SP-7 peptide could not completely inhibit the binding of the ad-SP-7 polyclonal

semm to the 54 kD protein. The reason for this lack of cornpetitive inhibition is not
known, but may result fkom a heteroclitic response to the SP-7 peptide in the peptideimmunized rabbit. Eviderre of the immunoreactivity,molecular mass and unglycosylated
nature of the 43 kD protein fom strongly suggested that this protein represents the
authentic unglycosylated HTLV-1 envelope protein precursor with the leader peptide
appropriately removed (Seiki et al., 1983; Pique et al., 1992). The identities of the
immunoreactive 30-39 kD low molecular mass proteins observed on Western blots could
not be deduced. Similar low molecular mass proteins were seen upon expression of the
HIV envelope glycoprotein in a similar baculovinis system (Hu et al., 1987). They
could represent proteolytic degradation products of the envelope precursor proteh or

97

variant glycosylated fonns of the mature protein since some fonns disappeared when
cells were treated with tunicamycin. Confirmation of the amino acid sequences of any
of the protein forms by amino-termiinal anaiysis has been hindered by the lack of

envelope protein solubility.

The accumulation of the HTLV-1 envelope protein in the form of inclusion bodies
could be the resuit of several related factors. The baculovinis expression system has been
previously fouod to be inefficient in the processing of viral glycoproteins nich as the

hemagglutinin of influenza virus (Kwoda et al., 1986) and the envelope glycoprotein of

HIV (Hu et al., 1987; Rusche et al., 1987), whose precursor proteins did not undergo
processing into their mature forms. It is mely that the inefficient cleavage of the
baculovinis-produced HTLV-1 envelope precursor protein, into the mature d a c e and
transmembrane proteins, may be due to its incomplete or irnproper glycosylation. Sf9

cells are capable of N-Linked glycosylation but are unabie to perform the complex sugar
linkages wnnally found on the HTLV-I envelope proteins m e r , 1988). This may lead
to the accumulation of the envelope precursor within the cells shce efficient proteolytic
cleavage of the HTLV-1 envelope protein precursor appears to depend on proper
glycosylation (Pique et al., 1992). Amther contributhg factor may be that, like HIV-1
gp160/120, HTLV-1 gp63146 may have a secpence that causes the retention of large

arnounts of the envelope protein in the secretory pathway (Bonifacino et al., 1991; Li et
al., 1992). It may be that when Sf9 ceils are driven to express large amounts of HTLV-I

envelope precursor, the celis are required to retain more protein than is physiologically

98

compatible. Previous difficulties in pmducing signincant levels of recombinant EITLV-1

envelope glycoprotein that bas been encountered in several other expression systems
(Kiyokawa et al., 1984; Samuel et al., 1984; Kuga et al-, 1986; Chea et al., 1989;
Nyunoya et al., 1990; Vile e? al.. 1991)suggests that expression of the glycoprotein may
place high demands on the cellular m a c b r y . d o t metabolic ewrgy, W o r may be

toxic to the host celis. It appears that the insect ceils may have overcome the unidentified
negative property of the HTLV-1 envelope protein by storing it as inclusion bodies to
maintain their cellular viability and yet be capable of expressing significant levels of the
protein. Even in mammaiian transient expression systems, unglycosylated and pamaiiy

glycosylated forms of the HTLV-1 envelope protein have been shown to accumulate
within the ceils to signifiant levels and are not properly transported to the celi surface

(Pique et al., lm).

The HTLV-1 envelope inclusion bodies (env-I.B.) isolated fkorn ou.baculovinis

system proved to be immunogenic in three straiDs of mice. Efficient stimulation of

envelope-specific anti'bodies by env-I.B. was observed in animals immunized in the
absence of adjuvant. The signifiant anti-envelope humoral response observed mggests
that the packing of the HTLV-1 envelope protein into inclusion bodies may mediate the
slow release of the immunogen at the injection site and cause a prolonged stimulation of

the animal's immune effector cells.

99

The injection of an ernulsion of env-I.B. and MCWE adjuvant preparation, proved

to significantly elevate the HTLV4 envelope-specific response in immuaued mice. A
MCWE dose of 10 ug produced maximum antiiaody titers to the four regions of envelope
represented by the synthetic peptides used in the peptide ELISA. However, 50 ug of
MCWE gave the best overaii m i y cesponse as determined by Western blot and
radioimmw10precipitationassays. Exceeding the 10-50 ug dose range of MCWE adjuvant

resulted in a decrease in peptide ELISA titers and a &op in the absolute number of mice
who seroconverted.

The above resuits suggest that MCWE can serve as an effective adjuvant in

immunization regimens as it stimuiatedgood antibody respooses agallist important regions

of the HTLV-I envelope. Studies have shown that the carboxy-terminus of gp46 is highiy

irnmunogenic in humans (Copeland et ai., 1986 and Pdker et al., 1989a). Simar results
are shown here, especially in the presence of MCWE. The C-terminal SP6 region

elicited the highest antibody titers of the four regions tested by peptide ELISA. Immune
responses against the SP-4A region are particularly signifiant since this region

encompasses a B-ceil epitope (Palker et ai., 1989a), a T-cel epitope (Kurata et al.,
1989), a cytotoxic T-ceii epitope (Jacobsen et al., 1991), and a virus neutraiizhg epitope
(Tanaka et al., 1991). In the presence of MCWE, signincant titers to the SP4A region
were elicited. The SP-2 region has also been associateci with virus neutraikation,

however this region within the envelope inclusion bodies did not elicit as high an
antibody response, as did other regions in the presence or absence of MCWE.

100

Neutmiking a n t i i e s , as detected by the syncytium inhibition assay, were

elicited by en.-I.B. ody at high adjuvant doses of M m ;thus, there was no correlation

with the titer of a n t i i y capable of binding to the SP-2and SP4A peptides and the
ability of the sera to inhibit syricytium formation. This lack of correlation suggests either
that the HTLV-1 envelope glycoprotein possesses other epitopes which are capable of
eliciting neutraliPng anh'body, or that the generation of neuttaliziag antibodies requires
the epitopes contahed withui SP-2and SP-4A be presented in a specific conformation

that is not present in the envelope inclusion bodies in sigaincant amounts. Indeed. it was

recently s h o w that incorporation of the SP-2 and SP4A peptides into chimeric

constmcts with promiscous T e l l epitopes enhanced theu immunogenicity and overcame

genetic restriction (Lairmore, 1995). The influence of overail protein conformation on
the presentation of iinear epitopes bas been previously observed with a n t i i y binding

to HIV gp120 (Nara and Goudsmit. 1990; Moore and Ho, 1993). For example, (Moore

and Ho 1993) found that a proportion of antibodies specific for h e a r peptide epitopes
bound to denatured HIV

mB gp120 with 100-foid lower affinity, compared with the

binding to the native molecule. The importance of linear epitope presentation in an

appropriate structural context in vino. suggests that sirnilar factors may have an impact
in vivo. The conformation of an injected antigen may thus influence the presentation of
iinear epitopes and ultimately determine the specificity and biological activity of the
antibodies produceci by the tecepient host. This phenornenon may explain the antibody

response observeci in mice injected with env-I.B. The resultant antibodies recognized
certain epitopes on the peptides SP-2and SP-4A,as demonstrated by ELISA; however

101

no ant'bodies specitk for SP-2 and SP4A-encodeci epitopes involveci in syncytium

formation were ptoduced. as suggested by a kick of syncytium inhiiition.

We have recently shown ihat recombinant vaccinia expressing different versions

of the HTLV-1 envelope could induce neu-g

antiiy against HTLV-1 (Ford et al.,

1992). In an effort to i n c m the neuaalizing antifbody titers. a c o m b immunization

~

regimen was devised employing both env-I.B. and R W . Priming with R W Els or U s
clearly enhanceci the anti-envelope anti'body response as compaRd to R W alone or env[.B. alone, as measured by peptide ELISA and Western blot analysis. Most notably, the
combhed immuni7stion of R W Ws, expressing gp46, and env-I.B. gewrated improved

neutrabhg antibody responses in cornparison to mice imrnunized with E3s alone.

uiterestingly, the R W Els which expresses the native HTLV-1 envelope of gp46 and
gp21, did not prime mice boosted with env-I.B. to induce higher levels of neutraiizing

antibody. The reason for this is not clear since R W Els was capable of inducing
neutraiking a n t i i y on its own. Perhaps the mannet in which the gp46 is presented to

the immune system by R W Els and R W U s is different.

Previous studies (Ford et al., 1992) revealed that R W Els infected human H-9
T-cells properly process and express the native HTLV-1 envelope proteins on the ce11
nuface to the same extent as HTLV-1 infected T-cells,

as deterrnitied by fluorescent

activated ceii sorting analysis (FACS). We would thus anticipate that the majority of the
R W Els-encoded envelope protein would be presented to immune effector cells in

102

association with MHC class I antigen (Teyton et al., 1990). However, R W Ws infected
H9 T-ceiis do not appear to retain HTLV-I gp46 envelope protein on their celi surfaces
as determinecl by FACS analysis, although R W E3s expresses higher levels of gp46 than
R W Els in infectecl ceus. Similar observations have been made when HTLV-1 gp46
(Picpe et al., 1990) and HIV a120 Weny et al., 1988) were transiently expressed in
the absence of their respective trarrPmembrane proteins, the majority of the surface

glycoproteins was released from the ceii surface. This suggests that the majonty of
exogenous gp46 secreted by R W E3s infected cek would be processed and presented
through the MHC class II pathway (Teyton et al., 1990), the same pathway expected to

process and present the majority of recombinant baculovims-produced env-I.B.. A shared

MHC class II presentation pathway between the R W Ws and recombinant bacuiovinis-

produced HTLV-1 envelope proteins suggests that similar envelope protein epitopes

would be presented to the immune system and thus may explain why boosts of

recombinant baculovitus envelope protein (env-I.B.) enhanced the neuaaliMg response

in R W E3s-primed mice. It wiii be interesting to determine if this combined
immUOj7iltion approach will prove effective and protect animals against live HTLV-1

infecteci ceil chailenge.

In addition to these animal studies, the antigenicity and immunogenicity of env-

I.B. was also studied in humans. A comparative study of the antibody respollse to envI.B. in HTLV-1 asymptomatic carriers and HAWTSP patients revealed marked

103

distinctions in seroreactivity between the two groups (Dekaban et al., 1994; Appendix

0. It appeared that HAWTSP patients have elevated anti-envelope antiaody responses

involving the IgM, IgA and IgE isotypes (Dekaban

a al.,

1994; Apperdix 1). This

suggested that env-I.B. could be useful as a target antigen, in a longitudinal study of

isotype profiles of asymptomatic individuais, to help determine whether an individual is

progressing towards the development of HAWTSP.

In another collaberative study with (Manca et al. 1995; Appendix II), env-I.B.
was found to be capable of generating specinc CD& T-celi Lines from HTLV-1
seronegative individuals. The expansion of envelope-specific T-cells from a naive
repertoire mimicks in vitro, the events that take place during the first encornter in vivo

between the T-helper cornpartment and the env-I.B. antigen, as in the case of deliberate

vaccination (Manca et al., 1995; Appendix II). These events include antigen
administration, uptake and processing by the appropriate antigen presenting cells, and
presentation to specifc T-cells that are present at low frequency in the naive repertoire.

The observations that the env-I.B. preparation is processeci and presented by human ceils
in a manner that gemrates T-helper epitopes suggests its potential use as a vaccine
candidate. T-helper cells are critical for the clonal expansion of antibody-secrethg B-

ceiis, and of specitic cytotoxic cells (Mosmann and Cofnnan, 1989; Stuhler and Walden,
1993); both king important effector mechanisms at work during viral infections

(Koszinowski et al., 1991; Sissons and Oldstone, 1980a; Sissons and Oldstow, 1980b).

104

The nxognition of env-I.B. by the majority of individuais tested thus far,

suggested its usefulness as a reagent to test for the presellce of HTLV-1 envelope-specific
cellular immunity. As previously notecl ia situations where seronegative individuais have
a history of contact with HIV-1 (Clerici et al., 1992), the identification of individuals
with previous exposure to viral antigens in the absence of seroconversion may prove to

be important. Thus, the ability to detect the presence of ceiiular immunity with env-1.B.,
in the absence of seroconversion may be a valuable diagnostic tool for the identification
of HTLV-1 infcted iodividuals who have not yet developed (or wil never develop) an

antibody response.

In summary,we have successfully expressed the entire HTLV-I envelope protein

in the baculovinis expression system. The envelope proteins were produced as inclusion

bodies which could be efficientiy recovered. The envelope inclusion bodies demonstrated
both antigenic and immunogenic properties within in vivo animal immunization protocols

and in vitro human seroreactivity and ceU proMeratiom. The results suggested that envI.B. may serve as a useN diagnostic tool and potential vaccine d i d a t e .

CHAPTER 3 A SOURCE OF GLYCOWLATEDHTLV-1 ENVELOPE PROTEIN=

EXPRESSION OF gp46 BY THE VACCINIAITT POLYMERGSE SYSTEM

The externai surface envelope protein of HTLV-1 is one of the W e s t retroviral

envelope proteins lmown and exhibits Little seguence variabiiity (De et al., 1991;
Kinoshita et al., 1991; Komurian et al., 1991). This genetic stability may be a reflection
of the limited size of the gp46 coding seqyetlce and a need for structurai conservation in
order to preseme its fiinctionality. In the past, research focused on the structural,
biochemical and immunological properties of the HTLV-1 envelope protein has k e n
greatly hindered by the limiteci arnouts of protein available. Isolation of the HTLV-1

envelope proteins from n a t u d y inf'ected cells bas oaly generated relatively smaii
quantities (Schneider et al., 1984; Copeland et al. , 1986; Palker et al., 1987). The
labile, noncovalent association of the surface envelope protein with the transmembrane
anchor is partially responsible for the low yields of gp46 (Copeland et al., 1986). The
surface envelope protein dissociates fiom viral panicles during sucrose demity gradient

purification, which resuits in the depletion of the glycoprotein fiom the density gradient
purifed virions (Copeland et al., 1986). Moreover, although supematants of HTLV-1
infected celis contain shed HTLV-1 glycoproteins and viral particles, substantially greater
amounts appear to be associatecl with virus-Sixteci cells, which increases the difficulty

in differentially extracting the envelope proteins from the other viral and cellular proteins

106

(Schneider et al., 1984). While cesults h m a nmber of studies have helped characterize
the HTLV-1 gp46 protein (Pique et al., 1990 and 1992), difnculties bave been

experienced using several different heterologous expression systems (Kiyokawa et al.,

1984; Samuel et al., 1984; Kuga et al.. 1986; Chen et al., 1989; Nyunoya et al., 1990;

Vile et ai., 1991) to produce large enough amounts of envelope protein for use in
biochernical and immunological studies.

The previous chapter d e s c n i the expression of the e n t h HTLV-1 envelope

protein, gp63, in a bacdovirus expression system (Arp et al., 1993). Although the
recombinant protein was expressed in large amounts, the majority of protein was not

processeci in an authentic fashion as suggested by its accumulation as relatively insoluble

inclusion bodies. The soluble and insoluble forms of gp63 exhibited significant
antigenicity and irnmunogenicity, as demonstrated by their ability to stimulate the
proliferation of envelope protein-specific human T-celi precwsors in vitro (Manca et al.,
1995) and high anti-envelope a n t i i y titres in mice and rabbits (Arp, et al., 1993; J.
Arp and G. Dekaban, unpubfished data). Unfomuiately, neither form of the recombinant

gp63 protein was capable of induchg neutmiking antibodes when it was injected alone,

in the absence of a recombinant vaccinia virus priming injection (Arp et al., 1993; S. Arp
and G. Dekaban, unpubfished data). The lack or absence of conformational integrity in

the baculovinis-expressed envelope protein preparations suggested that they may not be

optimal for vaccine purposes; nor did they seem suitable for use in studies to identify the
ceil surface receptor for HTLV-1.

107

Development of a mnmmaiian expression system appeared to be a potentid

solution to fulnll the need for a properly processed and soluble recombinant HTLV-1
envelope protein. Previous coiiaborativestudiesusing a recombinant vaccinia virus (RW
E3) containhg the EfTI,V-1 c&g

region for gp46 alone, revealed that the virus was

capable of produchg a confomutiody correct envelope surface protein, induchg

neutraiizing ant'bodies in mice (Ford et al., 1991; Ford et al., 1992; Arp et al., 1993)
and expressing envelope proiein at much higher levels than it did when gp21 was

concomitantly expresseci (RW El; Ford et al., 1992). h addition, expression of gp46

in the absence of gp21 couid prove to be immunogenicdiy advantageous in funw

vaccination studies. It appears that expression of the trammembrane protein might impair
the immunogenicity of the envelope protein preparation since HTLV-I gp21 shares a
short peptide sequence with other trammembrane envelope proteins that imparts

immunosuppressive properties to these other retiovinises (Ruegg et ai., 1989).Similarly,

research by Wainberg and his collegues (1987) has suggested that the gp2l protein may
be responsible for the immunosuppression observed

in HTLV-1 infected individuals

(Taguchi et al., 1991). In light of these observations, the vaccinia viruslT7 polymerase
expression system (Fuerst et al., 1987; Elroy-Stein et al., 1989) was chosen to express
the HTLV-1 envelope gp46 protein.

This hybrid vector system offered significant advamages with the utilkation of
the highiy efficient bacteriophage T7 RNA polymerase in a eukaryotic environment. T7

RNA polymerase is a single-subunit enzyme, with high catalytic activity ami strict

108

promoter specincity (Chamberlin and Ryan, 1982; Dunn and Studier, 1983) that had
previously found wide application for in Mtm synthesis of RNA and as the basis for hi&level gene expression systems in Escherichia coli (T'abor and Richardson, 1985; Studier
and Moffat, 1986). The use of the vaccinia virus as a vector for introduction of the

bacteriophage T7 polymerase into mammaiianc d cytoplasm had been utilizd by Fuerst

et al. (1986), to overcome the stringent recpkrnents of eukaryotic ceil RNA processing
and translation; otherwse, delivery of the T7 polymerase into the celI nucleus would

result in TI promoter-controUed mRNA that would not be spliced, capped, methylated

or polyadenylated. The cytoplasmic homing properties of vaccinia vinis render it suitable

as the delivery and expression vehicle for the T7 polymerase gene in eukaryotic cells.

Vaccinia virus has a large linear double-strarided DNA genome that encodes an entire
transcription system including a DNA-dependent RNA polymeme, a transcription factor,
capping/methylating enzymes and poly (A) polymerase (Moss, 1990). Additional
advantages of vaccinia virus include its large capacity for foreign DNA, genornic
stability, and its wide vertebrate host range (Smith and Moss, 1983; Moss, 1990).

In the bactenophage-vaccinia hybrid system, the 'I7 polymerase gene is under the

control of the earlyllate vaccinia virus W .S promoter such that it is continuously

expressed in recombinant vacciaia virus vTF7-3 infixted cells. A target gene under the
control of a T7 promoter, encoded within a second recombinant vaccinia virus, is
expressed upon co-infection and presence of the T7 polymerase enzyme. Initial design
of this system by Fuem et al. (1986) revealed that although large amants of RNA were

109

made, formation of the cap structure which is naturally presem at the 5' ends of

eukaryotic mRNAs and enhances nbsome bindiog, occurred inefficientiy and restricted

the amount of pmtein synthesized (Fuerst and Moss, 1989). To overcome this problem,
the encephalomyocarditis virus (EMCV) untranslated leader sequeflce was incorporated

immediately doamstrearn of the T7 promoter, enabling efficient capindependent

translation of the target gene transcripts (Eiroy-Stein et al., 1989).

In generd, proteins synthesized by vaccina Wus vectors are biologically active

and are processed and transporteci in accord wth the primary structure of the protein and
the inherent capabilities of the host ceiIs. As vaccinia virus proteins nomally undergo

a variety of pst-translational modifications (VanSlyke and Hniby, 1990) it is not

surprishg that encoded recombinant proteins undergo

N- and O-glycosylation,

phosphorylation, myristylation, proteolytic cleavage, polarized membrane and nuclear

transport, ami secretion (Moss and Renier, 1990). The dual vaccinia system had
previously demonstrateci its efficient expression of hepatitis B virus suface antigen,
measles Wus hemagglutulln protein, human prostaglandin synthases and H W gp160
(Fuerst et al., 1987; Hu et al., 1994; O'Neill et al., 1994), with higher levels of
expression obtained with the vaccinia virus-T7 systcm than with conventional
recombinant vaccinia v i . s (Fuerst et al., 1987). For smali-scale or transient
expression, an infectiodtransfectionprotocol predecessor hvolving both vinis infection
and plasmid transfction, bas proven to be effective and has been widely used (Fuerst et

a l , 1986). This transient infection/tranSfection, vaccinia/T7 poLymerase system is a

110

simplified version of the dual infection protocol since a plasmid CiIITies the reporter gene

controlled by the 'i7 promoter and does not require construction of a second recombinant
virus; however the protein yields are signincantiy less than the duai vaccinia virus system
(Fuerst et al., 1987). Despite its lower expression capacity, this infkction/transfection
protocol has been successfiilly applied to the anaiysis of interactions of CD4 with the

envelop glycoprotein nom RIV (Miaikami et al., 1988; Buonocore and Rose, 199),
the rescue of temperature-sensitive virus mutants (Li et al., 1988), epitope mapping of
monoclonal antiodes (Ke and Wagner, 1989), expression of voltage-gated channels

(Yang et al., 1991) and the cystic fibrosis transmembrane conductance regulator (Rich
et al. , 1990).

This chapter descn'bes the successful expression of recombinant HTLV-I envelope

surface protein in an authentic form, by the hybrid vaccinia virus-T7 RNA polymerase
system. Glycosylation complexity of the surface envelope protein was investigated using
funicamycin, endoglycosidase H and glycopeptidase F. Conformational integrity of the

recombinant HTLV-1 envelope protein was confimed by reactivity with sera from

HTLV-1 inf'ted patients and a panel of conformational-dependent human monoclonal

antibodies d e r non-denaturing conditions. Furthemore, this recombinant gp46 was

recognized by a series of HTLV-II infecteci human sera and sera from a Pan paniscus
chimpanzee infateci with the cstantly related STLVpan-p.EIighly conserved
conformational epitopes maintaineci in the recombinant HTLV-1 envelope protein
smcture suggests that it may serve as a useful diagnostic reagent and an effective vaccine

111

candidate in an appropriately adjuvanteci formulaton. (Most of the fiodings presented in

this chapter have been pubiished eisewhere; Arp e? al., 1996).

3.2 Materials and Methos

3.2.1 Cell lines

Mouse thymidine kinase negative (TIC) L ceis (Dr. W. FLintoff, Dept. of Microbiology
and Immunology, University of Western Ontano) and human HeLa feus (Dr. F.

Graham, Dept. of Biology, Mwaster University) were grown in Dulbecco modified

Eagle medium (DMEM) containing 10%fetal calf serum (FCS;Gibco-BRL) and 100

unitslml of penicillin G and streptomycin. HTLV-1 infecteci MT-2 (Popovic et al., 1983)

and uninfecteciII9 ceil lines were obtained h m Dr. L. Arthur (AIDS Vaccine Program,
NCI-Frederick Cancer Research and Development Center). MT-2and H9 human T-ceii
lines were rnaintained in RPMI 1640 medium supplemented with 1 96 L-glutamine, 10%

FCS and 100 unitslml of penicillin G and streptomycin.

3.2.2 Construction of recombinant plasmid

The HTLV-1 gp46 envelope coding sequences were derived from the plasmid pMT-2
(provided by Dr. R. C. Gaiio, N C m ; Ratner et al., 1985). Two translational stop

codons were placed immediately downstream of the gp46 coding region by insertion of

a &ml-EcoRT oligonucleotide. Plasmid pTME-46 was constructed by insertion of the 945

base pair (bp) fragment (NcoI-Xhd) encoding gp46 into the expression vector, pTM-1,
provided by Dr. B. Moss (NIH, MAID, LVD; ElroyStein et al. , 1989); such that the
envelope gene fragment was flanked by the T7 bacteriopbage promoter and t e e t o r
regulatory elements.

3.2.3 Isolation of recombinant vaccinia virus

The recombinant m e r vector. pTME-46, was used as the vehicle for insertion of the

TI-promoted HTLV-1 gp46 expression cassette into fk locus of the the vacciaia Wus
genome (Figure 3.2; vaccinia strain IHD-J, pmvided by Dr. S. Dales, Dept. of
Microbiology and Immunology, University of Western Ontario: Ddes and Sirninovitch,
1961). Recombinant virus was prepared by the infection of mouse TR'L ceUs with wild-

type vaccinia vins and subsequent transfection of the infecteci ceils with calcium

phosphate-precipitated pTME-46 donor DNA (Blair et al., 1980; Mackett et al., 1984;
1985). Briefly, the DNA precipitate was fomed by the addition of CaCl, (final

concentration of 250mM CaCld to a mixture of 2 ug pTME-46 plasmid DNA suspended

in a solution of 1X Hank's Buffered Saline and 40mM HEPES @H 7.O). This precipitate
was added to TK-L ceiis Mcted with wiid-type vaccinia virus at a multiplicity of 0.5

pfidceil (elapsed cime of infection was 2 hours). Overlayed ceil monolayen were
incubated at 37C for 4 hours, at which thne the ceUs were shocked with a 15%w/v

glycerol solution, washed with media and allowed to incubate at 374: for 4 days. The
ceUs of the advanced stage iafections were harvested anci TK-vaccinia virus was isolated
by four successive plaque purifications on TIC cells in the presence of bromo-2-

deoxyuridine @UdR; S i ) at a nnal concentation of 25 uglml (Mackett et al., 1984;

1985). Upon confirmation of recombinant vaccinia virus genomic structure by southern

blot anaiysis, large stocks of recombinant virus (vTME-46) were prepared under
nonselective conditions in HeLa cek .

L 14
3.2.4 Southern blot anaiysis

Monolayers of mouse L TIC- ceiis were infected with TK-putative recombinant or wildtype vaccinia virus at a multiplicity of iafection (moi) of 0.1 pWcell and virus harvested
48 hours post infection. Cellular and virai DNA was extracted h m infected cells by

resuspending celis in DNA lysis b&er

(150mM NaCI, 70mM EDTA, pH 8.0).

Foiowing a one hour digestion with proteinase K at a final concentration of 200 ug/ml,
1% sodium dodecyl suiphate (5% w/v SDS) was added to the suspension and incubated

for an additional 18 hours at 50C. Ri'bonuclease A (RN&

A) was added to 100 u g h l

and the mixture was incubated 1 how at 37C. DNA was extracted with
phenoVchloroform and precipitated with 2.5 volumes of ice-cold 95% ethanol. A 7.5-fold
excess of e n 1 restriction ettdonuclease was added to each 20 ug sample of cellu1arlv;il

DNA. The recombinant shunle plasmid vector, pTME-46, was digested with Q n I to
serve as a positive contml(0.5 ug). The digested DNA was electrophoresed on a 0.7 %

agarose gel and transferred for 90 minutes at 5 in. Hq. to Hybond-N membrane
(BIORAD) ushg a vacuum bloaer (BIORAD) as d e s c r i i by the manufacturer. The

resuiting Southem blot was hybridzed at high stnngency (65C) with a digoxigenin-UTP

(Boehringer Mannheim)-labellecl probe (1.O kb @nI-K#nI fragment spanning the 5' noncoding encephalomyowditis and HTLV-1 envelope fragment of the parent shunle

plasmid, pTME-l), as directed by the manufacturer (Boehringer Mannheim) and

described in chapter 2. FoUowing indirect cherniluminescence, the blot was exposed to

Kodak Cronex 4A X-ray film for 1 mimite.

115

3,2,5 Recombinant vaccinia virus i d d o n s

AU infections were performed with suspensions of intnceilular vaccinia virus isolated

from ceiiular mitochondrial fractions (Mackett et ai., 1984; 1985). The mitochondiral
fractions were titred twice with a standard vaccinia plaqye assay (Mackett et al., 1984;
1985), involving a methylceUulose overlay, prior to their use. The recombinant vaccinia
virus expressing T7 RNA polymerase, vTF7-3, was obtained from Dr. T. Fuem and Dr.

B. Moss (through te AlDS Research and Reference Reagent Program of AIDS, NIAID,

NIH; Fuerst et al., 1986). A previously descn'bed recombinant vaccinia virus expressing
HTLV-1 gp46 (RW E3; Ford et al., 1991; 1992) was used as a positive control. R W

E3 contaias the identical gp46 coding seyences as vTME-46; however, expression is

regulated by the vaccinia early/late promoter, W S.

3 J . 6 Source of conformation-depeadent buman monoclonal mtibodies

Human monoclonal antibodies (HMAbs) specific for the HTLV-1 envelope glycoprotein
were obtained from Dr. S. Foung (Staaford University School of Medicine, California).
Monoclonal an~r'body-prducingceil lines were generated using B lymphocytes isolated

fiom a HAMITSP patient, activated with Epstein-Barr virus (EBV), and fused to mouse-

human heteromyeloma ceii lines by electrofwion, as previously described (Foung and

Perkins, 1989; Perkins et al., 1991; PerLUis and Foung, 1995). Through various assays
perfomed by Rowe et al. (MM), the conformation-dependeenceand neutmiking activity
of the HMAbs WAl IIlFS, WAO7/2F7, WAO7/1G7, WAllRE2, WA1 WF3, and

WA04/2B10 on native HTLV-1 envelope protein and infected cells had been previously

116
characterized (Figures 3.9 and 3.10). AU envelope-specific human monocloaal anti'ibodies

were of the lgGl isotype. Human monoclonai mibody ROI, specific for
cytomegalovinis, seored as an isotype-matched negative control sera (obtained from Dr.
S. Foung ; Rowe et al., 1994).

3.2.7 Radi01abeiiing

HeLa cells were infecteci with recombinant vaccinia virus vTF7-3 done or together with
vTME-46, at a multiplicity of iafection (moi) of 4 plaque forming uni& of each virai
strain per ceii (pNceU). Twenty-four hours p s t iofection, ceils were labelled with f%]-

Lcysteine (O. 125 rnCi/Sx106 cells; NEN) for 5 hours in cysteine-free DMEM (GIBCO),

supplemented with 2%dialyzed FCS.

3.2.8 Glycosglation Iiihibiaon

Treatment with glycosylation inhiiitors involveci addition of 2 ug/mi of tunicamycin or

Brefeldin A (Schwarz and Datema, 1982; Miswni et al., 1986; Boehringer Mannheim)
to the appropriate tissue culture media (i) prior to radiolabelling (6 hours prior or 2 hours

prior, respectively); (ii) during the exposure to IfsSJL-cysteine; (i) in aiI the washing
procedures. To insure that clifferences in immunoprecipitated envelope protein amounts
were solely due to variations in a n t i i y binding and not due to d m g effects on protein
expression, the BIORAD phoshophoimager was used to quantitate the mounts of

radiolabelled envelope protein produceci in each infection treatment. This insured the
addition of equal arnounts of r2S]-cysteine-labelled envelope protein to each antibody

117
mixture for iaunu110precipitation. Immunoprecipitation of ecpivalent amoums of

recombinant envelope protein from each iafection lysate with MAb l C l l (Dr.Tom
Palker: Duke University; Pallcet et al., 1989), an aati'body specific for a linear epitope
of HTLV-1 gp46, served as controls for additionai c o ~ t i o n .

3.2.9 Immunopredpitation
Culture supernatants were supplemented with 1mM EDTA and 1mM PMSF (Boehringer

Mannheim). Ceils were lysed in cold extraction buffer (pH7.6; lOOmM NaCl, 10 mM
sodium phosphate, 1% v/v Triton X-100, 0.5% w/v sodium deoxycholate, O. 1% w/v

SDS, ImM EDTA, and 1m.M PMSF). Supernatans and cell lysates were precleared for
18 hours at 4OC by incubation with 60 ul Protein G PluslA Agarose (Oncogene Science),
which had k e n preincubated for 2h at 4C with 20 ul normal sera of the appropriate
species; volumes of agarose beads and normal sera mentioned were prepared for each
5x106 ceU equivaient. The Protein

G PluslA agarose

was peiieted and the resulting

nipernatants was divided h o 5x106 celi equivalents. hmunoprecipitation of each

precleared celi equivalent was performed at 4C for 18 hours, in the presence of 40ul of
Protein G PluslA agarose and various HTLV-1-specific antisera as specified, in a final
volume of 1.9 ml. Anti-gp46 mouse monoclonal amibody l C l 1 was provided by Dr.

Tom Paiker @uke University, North Carolina; Palker et al., 1989) as tissue culture
supernatant and diluted 1:4. HTLV-1 envelope-specific human monoclonai antibodies
(HMAbs; WA111IFS; WAO7/2F7; WA07f lG7; W A l llZE2; WA11/2F3; and

WA412B10; Rowe et al., 1994) were added at a concentration of 7.5 ug of IgG,/ml of

118
suspension. Human monoclonal antiibody R04, specific for cytomegalovinis, was used
at a concentration of 7.5 ug IgG,/ml as an immunoglobun isotype-matchai negative

control. Polyclonal hiiman sera h m W T S P patients were used at dilutions of 1:4000

to 1:20000, as specined. Human sera h m an RTLV-II positive individual was used at
a dilution of L:400 (Arnerican Red Cross reference panel of HTLV-11 positive sera;

obtained nom Dr. T.Palker). Polyclonal sera h m a pygmy chimpanzee (Panpaniscus)

infected with STLVpw was provided by Dr. G. Franchini (NCINfH; Gin er al., 1994)
and was diluted 1:200. Sera obtained fiom uninfected humans and pygmy chimps were

diluted 1:200 and served as negative controls for the appropriate polyclonal sera. Immune

complexes were washed 4 times with extraction buffer and then resuspended in the
appropnate buffer. Samples destin& for tricineSDS-polyaqlamide gel electrophoresis
(Schagger and von Jagow, 1987) on 12% poiyacrylamide SDS/6M urea gels were

resuspended in an equal volume of 2X Laemmli sample buffer and boiled for 10 minutes.
The gels were fuced for fluorography with Enteme Solution (Dupont, NEN), dned

between sheets of Biodesign gel wrap (New York) and exposed for autoradiography at -

70C.

3.2.10 Glycosidase Digestions

Following immunoprecipitation with mouse monoclonal antibody 1C11, immune

complexes were washed four tima in extraction buffer. Protein equivalents of 2.5~106
cells were then digested with endoglycosidase H (endo H)or Glycopeptidase F (PNGase

F) at 37C for 20 hours. Endo H digestion was performed in the presence of 25mM

119

sodium acetate, pH 5.0, 1mM PMSF and 12mU endo H (Boehringer Mannheim). For

PNGase F digestion, the immune complexes were ionibated in 25mM sodium phosphate,

pH 7.0, 1% Nonidet P-40, 1mM EDTA, 1mM PMSF and

1.m of PNGase F

(Boehringer Mannheim). Digestion products were compared to immunoprecipitates

obtained fiom vTME-46/vTF7-3 infecteci ceils treated with 2 ug/mi tunicamycin

(Schwarz and Datema; Boehringer Mannheim) for 8 hours prior to radiolabelling. AU the
protein products were analyzed using tricine-SDS-polyacrylamide gel electrophoresis
(Schagger and von Jagow, 1987) to enable the resolution of the variantly glycosylated
forms. Samples were

nui

on 12% polyacrylamide SDS/6M urea gels foiiowing

suspension in an equal volume of 2X Laemmli sample bwer and boiling for 10 minutes.
Enoglycosidase digestions and andysis were repeated at least three tirnes. Gels were
fixed for fluorography with Enteas@ Solution (Dupont, NEN), cirieci between sheets of

Biodesign gel wrap (New York) aod exposed for autoradiography at -70C.

3.2.11 Lmmunoblotting
Cells were infecteci at an moi of 4 pWceLI (with the exception of the viral titration
expiment) for 1 hour at nw>m temperature. Infections were allowed to progress under
conditions of 5% CO2/37r for 36 hours pnor to harvesting, unless otherwise specified.
Cell pellets were resuspended in extraction buffet and 1mM PMSF. Foliowing addition
of an equal volume of 2X Laemmli buffer, the samples were boiled for 10 minutes and

sonicated bnefly. The supematants were electrophoresed and transferred from a SDSIoM

urea 12% polyacrylamde gel (Hayden et al., 1986) to Immobilon P membrane

120
(Millipore), as previously described in chapter 2 (page 55). BLots were blocked in a

solution of 5%sLim mille powder (Johnson et al., 1984; chapter 2) at m m temperature

for 5 hours. Blots were iiinibated in the appropriate primary antl'body (dilutecl in blocking
buffer) ovemight at 4C with rocking. himary a n t i i e s included (i) lC11, an anti-gp46

mouse monoclonal a n t i i y (Palker et of. , 1989); (ii) anti-SP-3f -4A rabbit polyclonai
senun (SP-3and SP-4A peptide sequerices were derived h m gp46;Paker et ai., 1989);
(iii) human EETLV-1 patient sera (HAM/TSP); (iv) human WTL,V-II patient sera

(American Red Cross reference panel obtained from Dr. T.Palker); (v) human "HTLVindeterminate" patient sera (persoiis identifid as "HTLV-positivenby K R , but "HTLVindetenninate" according to serological tests performed by the Canadian Red Cross; sera
obtained from the Canadian Red Cross); and (vi) polyclonal senun fiom a pygmy

chimpanzee (Pmpaniscus) Xected with STLV-;

Giri et al., 1994). To insure the

conformation-dependenceof the HTLV-1 envelope-specific human monoclonal antibodies

(KMAbs; WA11IlFS; WAO712F7; WAO7/1G7; WA1112E2; WA11/2F3; and

WA0412B10; Rowe et al., 1994)' these mtibodies were screened for Western blot

reactivity at dilutions of 1:50 and 1:100 . Blots were then washed in blocking bmer and
exposed to goat anti-mouse, goat anti-rabbit or goat anti-human IgG antibodies (Abs)

conjugated to ahLine phosphatase or horse radish peroxidase (Jackson ImmunoResearch

Laboratones hc.)at a finai daution of 1:5000, for 30 min. at room temperature. The
blots were washed and exposed to substrate according to the manufacturer's instruction

(Blot Detection Kit for the alkaline phosphatase conjugated 2OAbs or Enhanced
Cherniluminescence for the horse radish conjugated 2OAbs; Amersham International,

121

PLC). Quantitation of immunoreactive envelope proteins was determined using laser

densitometry (BIORAD).

3.2.12 Binding of HMAbs to 'kative" and denatured dual vaccinia infecteai cel
lysates and recombinant bacuioviruslderivedenvelope inciusion bodies

vTME-461vTF7-3 infecteci cei lysates (rgp46) or purifml recombinant HTLV-1 envelope

protein extracts (env-I.B.) produced in the badovinis system (Arp et al., 1993) were
prepared in extraction buffer, or were denatured by boing for 5 miautes in the presence
of 50mM dithiothreiotol @TT) and 1 % SDS,followed by a 5-fold dilution in extraction
buffer and cooled to room temperature. Prepared vTME-46fvTF7-3 infkcted lysates

consisting of 6 x 106 ceii equivalents or 6.0 microgram equivalents of env-I.B. were

precleared and immunoprecipitated as previously descn'bed. Immunoprecipitationof each
protein extract was performed with the HMAbs WAO7/1G7, WAO7/2F7, and

WAW2B10 (Rowe et al., 1994) at a concentration of 7.5 ug of IgG,lml of suspension.

The HMAb 0 . k (Lipka et al., 1990; Matsushita et al.. 1986) specific for a linear
epitope of HTLV-1 gp46 was wd as a positive control a n t i i y aad was added at a
concentration of 7.5 ugfml. Immune complexes were washed and prepared for
electrophoresis and immuwblotting as aiready described. The immunoprecipiraied
HTLV-1 envelope proteins were deiected on the resultant western blots with the mouse
monoclonal antibody lCll (Palker et al.. 1989) and horse radish peroxidase conjugated
goat anti-muse IgG (Jackson TmmunoResearch Laboratories Inc.) as the primary and

122
secondary

anti'bodies,

Rspectively. Blots

were developed using Enhanceci

3.2.13 Indirect Immunofluomsnce of dual recombinant vaccinia infectecl ceils

Human H9 T-cells were infected in suspension at an moi o f 4 pfii/ceii for 1hour at room

temperature. The free viral i n 4 u m was removed ami the infection was aliowed to
progress for 24 hours. The cells were washed ihree times in phosphate buEered saline

(PBS)containhg 2% FCS and resuspended at a final concentration of 8x106 celislml.

Each spot of a 24-spot heavy tefion-coated super-cured slide @TC; Cel-Line Associates,
Inc.) received 3 ul of celi suspension ( 2 . 4 ~ 1 6cells). HeLa cells were infecteci in

suspension at an moi of 4 pfulceii for 1 hour at room temperature. The viral supernatant
was removed and the c e k were resuspended at a final concentration of 1 . 6 ~ 1 celIs/ml.
6
300 ul of celis suspension was plated per weii in permanox 6-chamber siides (Nunc) and

incubated for 24 hours. Foliowing air-drying, aii slides were fixed in acetow at room
temperature for 10 minutes. Uoinfected ceiis and those cells infectai with only one

recombinant virus, vTF7-3, were included on each slide as negative controls. Slides were
riased twice in PBS (pH7.2) supplemented with 2% FCS. Non-specific binding was

blocked by hcubating the washed c e b in a 4% solution of skim milk powder in PBS

(Blotto) for 1 hr. at rmm temperature. Cells were M e r incubated 30 min. at 37" in
1:20 dilutions of various prMary sera, uniess otherwise specified: (i) HTLV-1 envelopespecifc human monoclonal antibodies at 50 ug IgG,lml (WAllIlFS; WA0712F7;

WA0711G7; WAllf2E2; WAllRF3; and WA0412B10;Rowe et al., 1994); () human

123

monoclonai amibody R04, specific for cytomegalovinis, used as an immunoglobulin

isotype-matcbed negative control at 50 ug IgG,/ml (Rowe et al., 1994); (ii Human

monoclonal antibody 0.5-a specific for gp46 (Matsushita

et

al., 1986); (iv) human

HTLV-1 infecteci patient sera (HAMITSP). After washing, ceils were incubated for 30

min. at 37C with goat anti-human IgG fluorescein isothiocyanate conjugated antibody
(GAH-HTC, Jackson yhunoResearch) diluted in blochg solution. Rinsed ceils were
viewed on an Zeiss Universal fluorescence microscope.

3.3.1 Construction and chaiacterizationof recombinant vaccinia virus

The tramfer vector @TME-46) was constmcted by krting the HTLV-1 envelope gp46
coding sequeoces immediaiely dowmtream of the bacteriophage T7 promoter of pTM-1

(Ehoy-Stein et al. , 1989). Seqyences corresponding to the encephalomyocarditis Wus

(EMC) 5' untranslateci region were included between the bacteriophage T7 transcriptional

regdatory elements anci the gp46 coding sepnces, to increase the translational
efficiency of uncapped T7-promoted HTLV-1 env messager RNA (Elroy-Stein et al.,
1989).

Wiid-type vaccinia infecteci'IK-cells were traasf'ted with the pTME-46 plasmid

construct. Selection of TIC vaccinia virus involved several passages through mouse TIC
ceUs in the presence of bromodeoxyurid'i. Southern hybridization of digested TKcellular DNA foiiowing infection with nmrerous putative vTME-46 clones revealed that
only one bromodeoxyuridine-mistant vaccinia virai clone had the entire HTLV-1 gp46

coding seyence and the 5' non-coding bactenophage T7 promoter and EMC seqyences
integrated into its genome (Figures 3.1 and 3.2).

Figure 3.1
Blot of @KIdigested DNA was hybridized at high stringency with an HTLV-I gp46specific probe Iabeiled with digoxigenin-UTP. Foilowing indirect cherniluminescence, the
blot was exposed to Cronex Nm for 1 minute. Laue 1, uninfected TR-cells; lane 2-5,
putative vTME-46 clones infecteci into TIC*ceiis; law 6,wild-type vaccinia infected TIC
ceiis grown in nonseledon media; lane 7, pTME-46 aarisfer vector DNA .

Figure 3.2
Development of the dual recombinant vaccinialT7 polymerase system (vTME-46IvTF7-3)
for expression of recombinant HTLV-I surface envelope protein, rgp46.

129

3.3.2 Identification and characterizationO the recombinant proteins expressed by

the dual vaccinia system: vTME-46lvTFl-3

The isolated vTME-46 virai clone was capable of T7 promoter-controlied expression of

HTLV-1 gp46 upon CO-infectionwith the vaccinia virus vTF7-3 (Fuerst et al., 1986),
which encodes the highly efficient bacteriophage T7 polymerase (Figure 3-2). Western
blot d y s i s of dually infected HeLa cek revealed expression of five variant forms of

the surface envelope protein, ranghg in m o l d a r weight nom 39 to 49 kD. Their

HTLV-1 envelope origin was suggested by theu immunoreactivityto both anti-gp46 l C l l

monoclonai anhibody and polyclonal anti-SP-3/SP-4A envelope peptide sera (Figure 3.3).
Both of the HTLV-1 envelope-specific sera did not react with Western blots of uninfected

or vTF7-3 infectai HeLa cell lysates (data not shown).

3.3.3 N-giycosglation of the recombinant BTLV-1 envelope protein fonts

To determine if the five variant fonns of the surface envelope protein resulted fkom
differential glycosylation of the proteh, vTME-46/vTF7-3 infected celi lysates were
digested with endoglycosidase H (eado H) and glycopeptidase F (PNGase F). The eodo

H enzyme is only able to remove glycan branches of manwse-rich or hybrid

glycoproteins but not those of complex glycoproteins (Trimble and Maley, 1984), while
PNGase F will remove all N-nked oligosaccharides independent of complexity

( P l m e r et al., 1984). Digestion products were compared to immunoprecipitates

Figure 3.3
Westem blot analysis of vTME-46/ vTF7-3 infected HeLa ceii lysates. Primary antibody
used: 1C11, anti-gp46 mouse monoclonal antibody; SP-3/SP-4A, anti-peptide SP-3/SP4A rabbit polyclonai sera. Both of the HTLV-1 envelope-specific sera did not react with
Westem blots of uninfected or vTF7-3 infcted HeLa cell Lysates.

ICI 1

132

obtained from vTME46/VTF7-3 infected cek treated with the N-glycosylation iahi'bitor,

tunicamycin (Schwarz and Datema, 1982).

As demonstratecl in Figure 3.4. immunoprecipitation of untreated iafected cell

lysates with anti-gp46 monoclonal antibody lCll yielded the five variant forms of the
surface envelope protein. 39-49 kD.Complete Encio H digestion of the normal infected

ceil lysates produced oniy the 39 kD protein form. Digestion with PNGase F repeatedly

produced both the 43 aod 39 kD forms. Resistance of the 43 kD protein fonn to PNGase

F digestion suggests that one glycosylation site rnay be located in a region of the protein
that is not suscept%le/accessibie to PNGase F. As expected, dual infection in the

presence of ninicamycin yielded the 39 kD protein as the predominant fom upon

radioimmunoprecipitation. Some traces of the 43 kD form were observed suggesting that

inhibition of N-glycosylation was not absolute at the concentration of Nnicamycin used.

No protein species with an apparent molecular mass l a s than 39 kD were visualized
suggesting that the 39

kD form represents the completely unglycosylated envelope

protein. The 10 kD shiff in molecular mass observed following eodoglycosidase digestion

or hinicamycin treatment is consistent with the Loss of four oligosaccharrides (Kornfeld

and Kornfeld, 1985), and suggested that all four potential N-linked glycosylation sites of
the recombinant envelope protein are utilized for glycan addition (Seiki et al., 1983).

Variation in electrophoresis coaditions and molecular weight standards may

account for the discrepancy in the calculated molecular mass for the unglycosylated

Figure 3.4
N-glycosylation of recombinant HTLV-1 envelope protein forms. Lysates of 2-5~106
esSI-cysteine-labeiledceiis were immunoprecipitated with mouse monoclonai antibody
l C l l and digested with endo H or PNGase F. Untr., immunoprecipitate of untreated
vTME46/vTF7-3 infected celi lysate; EndoH and PNGF, immunoprecipitates of
untreated vTME-46/vTF7-3 infected ceii lysate digested with endonuclease H and
PNGase F respectively; Tunc., immunoprecipitate of tunicamycin-treated dual vaccinia
infection lysate.

135

recombinant IlTLV-1 envelope protein of approxmateIy 39 kD (+1- 4 kD), compared to

the expected molecular mass of 34.5 kD suggested by the gp46 coding sequeflce (313 aa;

Ramer et al., 1985).

3.3.4

Effects of virus mulopiidties on recombinant ETLV-1 envelope protein

expression

Since the T7 RNA polymerase and the HTLV-1 gp46 envelope coding sequelices were
delivered to the target cells by two independent recombinant vaccinia vinises, we wished
to determine the arnounts of each virus that were needed for op-

expression. HeLa

cells were CO-infecteciwith recombinant vinises vTF7-3 and vTME-46 at a range of

multiplicities. Upon Western blot analysis, it appeared that a multiplicity of each viral
strain of 2 pfidcell appeared to yield maximal amounts of glycosylated envelope protein

(Figure 3.5). Higher multiplicities caflging fiom 4 to 10 pWcell did not result in
increased yields of the recombinant protein species.

The dual infection system (vTME-46/vTF7-3) was compared with a single

vaccinia

V~IUS

recombinant system ( R W E3) in which expression of gp46 was under

control of the vaccinia promoter W .S. This vaccinia promoter W -5, controiiing gp46

expression in the independent virus system of R W E3, is the same promoter responsible
for regulating expression of the T7 poiymerase gene of vTF7-3 in our dual virus system.

The high efficiency of T7 polymerase/T7 promoter-dependent expression was

Figure 3.5
Dependence of maximal gp46 expression on Wus multiplicities. Western blot anaiysis
of HeLa ceiis (2.0x1@ cells) infecteci with both vTME-461vTF7-3, R W E3 alone or
vTF7-3 aione, at various virus multiplicities of 2 , 4 , 7 or 10 pWmi. HAWTSP sera was
used for immunoblot detection. Low-range molecula.mass standards designated on left
side of the figure.

138

demonstrated when vTME-46/vTF7-3 were each infecteci at a multiplicity 2 pWcell

expressed three times as much total envelope protein (totai of dl the variously

glycosylated fomis); as compared by densitometry to the amount of envelope protein

produced by infection with the Mependent, single vaccinia R W E3 at a moi of 4

pWceil (Figure 33.

3.3.5 Time course of protein yields

To M e r determine the conditions for optimal expression, a time course of HTLV-1

envelope expression in infkcted HeLa cells was examhed. CeIls dually infected with
vTME-46/vTF7-3 or singly infecteci with R W E3, were harvested at various times and
the proteins were analyzed by imrnunoblotting with sera fiom an HTLV-L infected patient

(Figure 3.6). Maximal yields of the HTLV-I surface envelope forms by the dual
vaccinia/V polymerase system appear to occur between 36 and 48 hours post infection

(p. i.) . A k d y at 12 hours p. i., the vTME-46/vTF7-3 infection was capable of produchg

more total HTLV-1 envelope protein than the single vaccinia system, R W E3, was able
to produce at 24 hours p.i. Accumulation of partialiy glycosylated and unglycosylated
protein forms were observeci relatively early in infection with vTME-46/vTF7-3 (by 24
hours p. i. );these differentially glycosylated species were only expressed by the R W E3
vaccinia expression system in trace amounts, at later harvest of 68 and 72 hours p i .
(data not shown).

Figure 3.6
Time course of recombinant protein yields nom two different vaccinia expression
systems. Western blot analysis of cell lysates (1.2 x l(Y cells) Uifected with recombinant
vaccinia at an moi of 2 pWml and harvested at the designateci thes. The blot was
incubated with HAMITSP sera.

3.3.6 Expression in Human 89 versus EeLa ceik

Human A9 T-ceils were compared io human HeLa epitheliai celis as hosts for dual
vaccinia virus infection and recombinant envelope protein expression. &th cell lines
were infecteci with the same virai stocks, at the same multiplicities and harvested at the
same tirne. Each infection was examined by both Western blot analysis and
immunofluorescence for envelope protein expression on three separate occasions.

Western Mot aaalysis and laser densitometry revealed that HeLa ceils produce
three- to four-fold more envelope protein than the same nurnber of H9 cells (Figure 3.7).
The five different envelope f o m (3949 kD)were seen upon dual vaccinia infection of

HeLa cells, while infected EW ceils expressed maialy the two most glycosylated envelope
fonns, 47 kD and 49 kD and trace amounts of the 39 kD protein (Figure 3.7). The
difference in expression levels was also apparent upon immunofluorescence analysis of
the two infecteci ceil Liws. Strong immunofluorescence of greater than 90% of the HeLa

celis in the monolayer was observed, in contrast to moderate immunofluorescenceof only

30% of the H.9cells (data not shown). Higher virus multiplicities of 8 and 10 did not

appear to increase the amount of gp46 expression by H9 cells or the percentage of

vaccinia infecteci cells (data not shown). Single infections with vTMEi-46 suggested that
this recombinant virus is restricted

(data not shom).

in its Uifectivity of H9 celis, but not of Hela cells

Figure 3.7

Expression of recombinant HTLV-1 surface envelope protein in Human H9 (Tceils)

versus HeLa (epithelial) ceiis. Western blot analysis of A9 and HeLa celi lysates (2.0xlQ
cells) CO-infectedwith vTME-46/vTF7-3 at an moi of 4 pWd. Blot was incubated with
HAM/TSP sera. No reactivity with the HTLV-1 patient sera was observeci following
Western blot analysis of H9 and HeLa cell lysates infected with vTF7-3 alone.

144

3.3.7 E p i t o p of r e c o m b i t HTLV-I gp46 shared with HlLZV-11 and STLV,,

A panel of HTLV-1 positive sera (confirmed by ELISA and PCR; data not shown) from

both asymptomatic and W T S P patients was scteened for reactivity to the recombinant
envelope proteins produced in the vTME-46fvTF7-3 infected HeLa cell system (Figure
3.8, lanes 1-4). Despite their derivation h m seropositive individuais of Mering clinical
stahls, aU HTLV-I positive sera recognized each of the five glycosylated recombinant

envelope proteins upon Western blot analysis. Figwe 3 -8 contains a representative panel
of asymptomatic HTLV-1 patient sera (lanes 1 and 2) and HAMITSP patient sera (lanes
3 and 4).

Western blot reactivity of HTLV-IIsera with the recombinant envelope proteins

produced by the duai vaccinia/T7 polymerase system was also tested. AU HTLV-II
positive reference samples tested had been previously confirmai by ELISA and PCR to
be HTLV-1 negativemTLV-II positive by the American Red Cross. Upon Westem

blotthg, a i i 11 samples of HTLV-IIinfected patient senun tested recognized one or more

of the recombinant HTLV-1 envelope protein fonns. Seroreactivity to the 47 kD envelope
protein was consistently observed in all the HTLV-II positive samples. Figure 3.8
contains a panel of four representative HTLV-II human sera which demonstrated
heterologous reactivity patterns to the various envelope forms Ornes 5-8). Some

Figure 3.8
Reactivity of HTLV-I+, HTLV-II+and STLVWf sera with recombinant HTLV-1
envelope protein. HeLa ceil lysates infected with vTME-46/vTF7-3 were immunoblotted
with normal human serum (NHS,1:500 dilution); asymptomatic HTLV-1 infected patient
sera (lanes 1 and 2, patients A and B @ 1:500); sera fiom EITLV-1 infected patients
diagnosed with HAWTSP (lane 3, patient C @ 1:10000; lane 4, patient D @ 1: 4 0 ) ;
HTLV-II infected patient sera (law 5-8, patient E-H @ 1:500 dilution); STLV,,
infected pygmy chimp senun (lane 9, 1:100; lane 10, 1:200; lane 11, 1:400dilution);
nomai pygmy chimpmzee semm (NCS, 1:100 dilution).

147

HTLV-II individuals exhibiteci seroreactivity patterns identical to sera from individuals

infected with HTLV-1 (Figure 3.8, compare lanes 5 and 6 to laaes 1-4).

Eight additional sera were obtained h m the Canadian Red Cross that had been
previously identifiai as "HTLV-indeterminate" . The label of "HTLV-ideterminate had
"

been previously assigned to describe these patients since their sera did not demonstrate
the standard pattern of reactivity to HTLV-Va proteins by analysis with the Diagnostic

Biotechnology HTLV Blot 2.3 strips, despite the ability to PCR ampli@ HTLV sequences

fiom the corresponding DNA samples. Interestbgly. a i l eight "iadetermiaatewserum

samples (100%)repeatedly exhibited Western blot reactivity to one or more of the
recombinant HTLV-1 envelope protein forms produced by the dual vaccinialT7
polymerase system. Bhdhg to the 47 kD envelope protein was obsewed in 718 cases
(data not shown). Thus for the f h t t h e . HTLV infection in these individuals could be
detected using rgp46 as the target antigen. No reactivity to any HTLV-I envelope

proteins was observed in the normal human serum samples tested thus fat (example:

Figure 3.8).

When sera fiom a pygmy chimp infectecl with a novel STLV-

s&

(Giri et

al., 1994) was screened by Western blot. predominant reactivity was observed to the
M y glycosylated 49 kD protein and the unglycosylated 39 kD HTLV-1 euvelope proteins

(Figure 3.8, laues 9-1 1). Similar patterns of reactivity were observed following
completion of three independent trials, with low level reactivity of the STLV-

sera to

148
the intemediate glycosylated f o m king observed foUowhg longer Nm exposures of

the blots. In addition to recognition of the recombinant envelope proteins in a denawed

form, the sera from the STLV-

infeted cbimpanzee and HTLV-1 infcted and EFLZV-

II iafected humans were ais0 capable of their imrunoprecipitation (Figure 3.12, panel
A: lanes 2, 3,7, 8). Repeatedly, the two most glycosylated forms of the envelope protein

(47 and 49 kD) were preferentaliy bound and precipitated by ail the c l i n i d samples
tesed.

3.3.8 Binding of contorrnation-dependmt anti-H'M,V-1 human monoclonal

antibodies to recombinant envelope proteins

A panel of human monoclonal anabodes was tested for their ability to bind to the

recombinant HTLV-I envelope proteins produced by the vTME-46/vTF7-3 system. The

human monoclonal antibodies had previously been isolated and selected in Dr. Foung's
laboratory (Rowe et al., 1994) for their ability to bind to gp46 of several dflerent
HTLV-1 infected cell liws (Table 3.1). Rowe et al. (1994) demonstrated that these six

monoclonal antibodies are of particuiar biological importance since they are capable of
inhibiting HTLV-1 syncytia formation (Table 3.1 and Figure 3.9) and specifically

recognize conformational epitopes of the envelope protein. This conformatioa-dependent

binding was suggested by their inabiiity to recognize denatured HTLV-1 envelope
proteins rom the various iafected cell sources; as demoristrated by the absence of

Western blot reactivity to viral lysates (Table 3.1; Rowe et al., 1994). The conformation-

Table 3.1
Characterization of the anti-HTtV-1 hurnan monoclonal a n t i i e s (with permission from
Dr. S. Foung and I. Rowe).
Immunofluotescence of fuced (IFA) or live &CA) HTLV-I infected MT-2 cells
were stained with the indicated H ' s foilowed by fluorescein-conjugated
secondary anti'body (goat ad-human IgG).
Western blot reactivity was tested using commercial saips of viral lysates from
Diagnostic Biotechnologies, Ltd.

Radioimmunoprecipitation analysis (RIPA) involved incubation of the HMAbs

with ceil lysates from HTLV-1 infecteci MT-2 celis radiolabelleci with 3sSmethionine and 35S-cysteioe.Syncytium inhibition was determined by incubating
HuT-102 celis (HTLV-1 positive) with M e c t e c i human osteosarcorna ceils
(HOS) in the presence or absence of HMAbs overnight, at which thne rnonolayers
were srained and syncytia were counted.

HUMAN
mAb

IMMUNOFLUORESCENCE
IFA
LCA

WESTERN BLOT

RIPA

SYNCYTIUM
INHIBITION

Figure 3.9
Graph demonstrating the dose-dependent ability of anti-HTLV-1 humaa monoclonal
antibodies to inhibit HTLV-1 syncytium formation (with permission fiom Dr. S. Foung
and J. Rowe). Syncytium inhibition was detennined by iacubating HuT-102 ceiis (HTLVI positive) with uninfecteci human osteosarcorna ceiis (HOS)in the presence or absence
of HMAbs oveniight, at which time monolayers were staiaed and syncytia were counted.

10

Concentration
--f--

WA04nB10
WAO7nG7
WA0712W

(uglml)

WA11f2E2
WA11M F5
WA11/2F3

100

153
dependence of these anticxiies was M e r supporteci in this curzent saidy. No
seroreactivity with any of the denatureci recombinant envelope proteins was observed
upon immuwblotting the ceil lysates h m vTME-46fvTF7-3 infected ceiis (data not
shown).

Conformational integrity of the recombinant EiTLV-1 envelope proteins, produced

by the dual vaccinia system, was suggested by their immunoprecipitation by ail s u
human monoclonal a n h i e s . Preferential precipitation of the 49 and 47 kD protein
forms was observed by aU the monoclonal anti'bodies of the panel (Figure 3.10, lanes 4-

9). Similar patterns of imrnunoprecipitation of the 49 and 47 kD proteins were

demonstrated by several different polyclonal sera coiiected fiom HTLV-1 infectesi patients
(HAM/TSP; Figure

3.10,

lanes

10-12). Longer tilm exposures of

the

radioimmw1oprecipitates revealed that the three HTLV-1 patient sera were also capable
of precipita~gthe 45,43 and 39 kD f o m . A representative HTLV-I envelope-specific

human monoclonal anti'body did not immunoprecipitate proteins from either uninfected
(Figure 3.10, Iane 2) or singly vTF7-3 infectecl HeLa cells (Figure 3.10, lane 3). A pool

of HTLV-1 uifected patient sera also did not immunoprecipitate any proteins fiom HeLa
ceils infected with vTF7-3 alone (Figure 3.10, lane 1).

Figure 3.10
Radioimmunoprecipitation of dual vaccinia infected cell lysates with HTLV-1 envelopespecific human monoclonal mtibodies. [%]-cysteine-labeUed proteins from vTME461vTF7-3 infectai HeLa ceils (ianes 3-10) were immunoprecipitated with equal IgG
concentrations 17.5uglm.11 of HTLV-1-specifc human monoclonal antibodies (lane 4,
WAllIlFS; lane 5, WA0712F7; lane 6 , WA07flG7; lme 7, WAllI2E.2; lane 8,
WA1112F3; and Iane 9, WA412BlO); or a . ant-CMV isotype-matched human
monoclonal antibody (lane 3, 7.5 ug IgG/ml RW); or polyclonal sera kom HTLV-1
infected patients diagnoses with HAMfTSP (lane 10, patient C @ 1:20000; lanes 11 and
12, patients D and 1 @ 1:4). Radiolabelhi lysates fiom singly vTF7-3 infected HeLa
celis were not immunoprecipitated with polyclond HTLV-1 infected patient C senun
(iane 1 @ 1 :4000) or human monoclonal antibody, WA0412B10 (lane 2,7.5 ug IgG/ml),
and served as additionai negative controls.

156

To insure that immunoprecipitation by the human monoclonal aohbodies of the

recombinant gp46 (rgp46) envelope forms, was dependent on protein conformation, we
studied the effects of denaturation on a n t i i y binding. vTME-46lvTF7-3 iafected ceil

lysates were prepared in extraction b f l e r (0.1 % SDS, no DTT; defined here as "nativen)
or were denatured by boiling for 5 minutes in the presence of 50mM DIT and 1%SDS,

followed by a 5-fold dilution in extraction b a e r to re-establish conditions compatible

with immunoprecipitation. As demonstrated in Figure 3.1 1, boiling of rgp46 in the

presence of DTT and SDS destroyed the ability of rgp46 to be immunoprecipitated by

the human monoclonai anti'bodies. The presentation of discontinuous epitopes by the

recombinant protein produced by the vaccinia/W system is in contrast to the apparent

absence of conformational integrity of the recombinant HTLV-1 envelope protein (env-

I.B.) produced in the baculovirus system (Arp et al., 1993; chapter 2; Figure 3.11). In
the presence or absence of denahiring conditions, the recombinant baculovirus-generated

HTLV-1 envelope protein was unable to bind the conformation-dependent monoclonal

antibodies. Tmmunoprecipitations were also performed with the human monoclonal

antibody

O.sa, specific for

a hear epitope of HTLV-1 gp46 (Lipka et al., 1990;

Matsushita et al., 1986; Figure 3.11). As dernonstrated by their ability to bind to the

monoclonal antibody 0.50: in both their "native" and denatured states, the HTLV-1
envelope proteins produced by both the dual vaccinia and the recombinant baculovinis
systems were capable of presenting linear epitopes for immunoprecipitation. In addition,
the recognition of both denatured HTLV-1 envelope proteins extracts by the 0 . 5 ~ ~

monoclonal antikiy confirmed that the buffer conditions following denaturation were

Figure 3.11
Immunoprecipitation of native and denatured vTME-46fvTF7-3 inf'ected ceU lysates
(rgp46) and baculovirus-generated recombinant HTLV-1 envelope protein extracts (envLB.) with various human monoclonal antibodies. Recombinant gp46 (NT7and DT7) or
env-I.B. (NE and DE) were maintained in their native conformation (NT7 and NE), or
were denatured by boiliog in the presence of DTT (DT7 and DE). Ability of the human
monoclonal antibodies WA071lG7,WA0712F7,WAW2B10 and 0.Sa to immunoprecipitate the recombinant envelope protein extracts, was detected by western blot analysis of
the precipitates using the H n V - 1 envelope-specific monoclonal anaWy l C l l as the
primary anfi'body. The mti-cytomegalovinis monoclonal antibody R04, did not
immunoprecipitate any specific envelope proteins Born either native vTME-46fvTF7-3
infected ceil lysates or recombinant bacuiovirus generated env-1.B. extracts.

159
compatible with irnmunoprecipitation. A non-specific 51 kDa protein was
immunoprecipitated consistently firom recombinant baculovinis-generated env-I.B.,
independent of protein state or human monoclonai a n t i i y used (Figure 3.11).

Immunoprecipitation of the 51 kDa protein was even observed upon incubation with the
isotype-matched anti-~ytomegaIoVinisantiibody (R04;Figure 3.1 1). The same negative

control anti'bodydid not precipitate recombinant envelope proteins from the "native"dual
vaccinia iafected cell lysate (Figure 3.1 1).

Native folding of the recombinant envelope protein was further supported by

indirect immunofluorescence assays performed with the panel of human monoclonal

antibodies. Tbree of the six confonnation-dependentmonoclonal a n h i e s were capable

of binding to vTME-46/vTF7-3 infcted HeLa celis. Monoclonal antibodies WAOW 1G7

and WA07/2F7 (Figure 3.12, Panels e and g respectively) exhibited significant

immmofluoresceace while WA11/2F3 bound to ooly a Mted number of dually infected

ceils (data w t shown). These monocloual anti'bodies did not bind to singly vTF7-3
infected ceiis (Figure 3.12, Panel d). No immunoQuorescence of vTME-46/vTF7-3
infected celis was observed with normal patient sera and an anti-cytomegalovinis isotype-

matched monoclonal antibody (Figure 3.12, Panels a and f respectively).

Figure 3.12
Indirect immunofluorescence of HeLa cells infecteci with vTME-461vTF7-3or vTF7-3
alone. Dual vaccinia infected ceiis were incubateci with normal human sera @mel a); or
HTLV-1infected patient sera (Panel b, serum h m B-ceii hybridoma source; Panel c,
serum from an asymptomatic HTLV-1 positive individuai); or HTLV-I-specific human
monoclonal antibodies (Panel e, WA0711G7; Panel g, WA0712F7); or an a&-CMV
isotype-matched human monoclonal anabody (Panel f). Singly vTF7-3 infecteci HeLa
celis did not fluoresce with HTLV-1-specific human monoclonal antibody, WA0711G7
(Panel d).

162

3.3.9 Influence of ogosoccharde cornplex&

on antibody binding to

conformationai epitopes

To determine if the binding of confonnation-dependent antibodies to the recombinant

HTLV-1 envelope glycoprotein would be influenceci by the complexity of its anached
oiigosaccarides, the duai vaccinia expression system was tteated with Brefeldin A or
tunicamycin. Brefeldin A inhi'bits transport finm the endoplasmic reticulum to the cis
Golgi apparatus and thus prevents additional trimming and modification of

oligosaccharides attacheci to the protein backbone (Misumi et al., 1986). Treatment with
Brefeldin A interferes with oligosaccharde maturation of giycoproteins if any processing
in the Golgi is required. This is in contrast to treatment with tunicamycin, which inhibits
ail potentiai N-glycosylation by interferhg with the initial step of N-glycosylation in the
endoplasmic reticuium (Schwarz and Daterna, 1982). Radioimmunoprecipitationwas used
to monitor the differentiai binding of various m i y preparatiom to the HTLV-1

envelope protein species produced d e r the diverse treatment conditions. Precautions

were undertaken to insure that equai arnounts of ~sS]-cystehe-labelledenvelope protein

from each treatment were added to equal amounts of the antibody preparatiom. For
additionai confirmation, the determined equivdent of recombinant envelope protein from
each treatment was also immunoprecipitated with MAb 1C11, specific for a linear
epitope of HTLV-1 gp46 (Figure 3.13; lane 1 of Panels A, B and C).

Figure 3.13
Expression of recombinant envelope proteins by vTME-46/vTF7-3
infecteci HeLa ceils
in the absence or presence of the glycosylation inhibitors Brefeldin A and tunicamycin.
Equivalent amounts of rsS]cysteine-labeiled lysates (vTME-46lvTF7-3
infections:Panel
A, untreated; Panel B, Brefeldin A-treated; Panel C, tunicamycin-treated) were
immunoprecipitated with the same series of polyclonal sera and monoclonai antiiodies.
Laue 1, lCll MAb; lane 2, HAMITSP patient C @ 1:20000; lane 3, HAM/TSP patient
D Q 1 :4tOOO; lane 4,WA11IIF5 [7.5 ugIgG,/ml];lane 5,WA07/1G7[7.5 ugIgG,lml];
lane 6,WAO712F7 17.5 ugIgG,/ml];lane 7,HTLV-IIinfected human sera @ 1:400; lane
8, STLVp,, infected chimp sera @ 1:200.

4s
47

4s
. ..

a * *

43
' J I

165

The anti-HTLV-1 conformationdependentHMAbs demonstrated reduced binding

efficiency to the brefelciin-treated envelope protein compared to their abiiity to
immunoptecipitate untreated HTLV-1 envelope protein (Figure 3.13. compare Panels A
and B: lanesi 4, 5, 6). In addiaon. polyclonal sera fhm individuals/cbimps iafected with

either HTLV-1, HTLV-JI or STLV,,

also demonstrated variable reactivity to brefeldin-

treated envelope protein compared to their bhding of untreated envelope protein (Figure
3.13, compare Panels A and B: lanes 2, 3, 7, 8). However, the confonnational epitopes

were not composed solely of carbohydrate residues since the monoclonal and polyclonal
antibody samples recognized and precipitated uoglycosylated envelope protein

(tunicamycin-treated; Figure 3.13, Panel C).

3.4 Discussion

In this study, we employed a v a c c W T 7 polymerase system to express the recombinant

HTLV-1 surface envelope protein in mammalian ceUs. This strategy required the

co13~t~tlctioa
of a recombinant vaccinia vins, vTME-46, encoding the HTLV-1 gp46 gene
fiagment under the control of the T7 bacteriophage promoter and terminator regulatory
elements. Co-infection with a second recombinant virus, vTF7-3, encoding the T7

polymerase gene (Fwrst

et

al., 1986) d t e d in expression of the gp46 envelope

protein.

Five differentially glycosylated forms of the surface envelope protein were

produced by vTME-46IvTF7-3 infected HeLa cells. N-glycosylation inhibition by
tunicamycin and N-giycan removal with endo H and PNGase F revealed that the 39 kD
protein was the unglycosylated fonn and the 49 k D protein was the M y glycosylated
envelope protein. Each oligosaccharide on average contri'butes approximately 2 kD to the
apparent molecular mass of the protein (Komfeld and Kornfeld, 1985); thus the observed

IO kD ladder of different recombinant envelope forms was consistent with the dHerentiai
attachment of four oligosaccharides.This suggested that all four potential N-glycosylation
sites in gp46 (Seiki et aL, 1983) were utized for oligosacchande modification in the

vacciniaR7 polymerase system.

167

The envelope glycoproteins expressed by this mnmmaiian expression system

appeared to have both mannose-rich and hybrid oligosaccharides attacbed, as deknnined
by endo H digestion and Brefelciin A treatmem. The glycosylation of the recombinant

envelope proteins resembled that of gp46 produced by HUT-102 iafected human cells,
which were also sensitive to endo H digestion ( h e et al., 194). Endo H sensitivity of

gp 120 was also observed in HIV infateci T-ceiis (Geyer et al., L988). In contrast, Piqw
et al. (1992) found recombinant HTLV-1 gp46 expressed h m m e c t e d COS ceils to
be resistant to endo H digestion. This discrepancy in endo H sensitivity is most iikely due

to the glycosylation variation between ciifternt ceLi types (Komfeld and Komfeld, 1985;
Rademacher et ai., 1988).

The dual infection system (vTME-46IvTF7-3) was compared with the single
vaccinia virus recombinant system (RW E3) in which expression of gp46 was under

control of the vaccinia promoter P7.5. The efficiency of the T7 polymerase/T7 promoterdependent expression by the duai vaccinia system was apparent in its ability to express
more total HTLV-1 M a c e envelope protein with faster kinetics and less extensive
cytopathic destruction of the hoa cells. This corroborates observations of Fuerst et al.
(1987) who described a significant eight-fold increase in gp160 yields produced by a dual
vaccinia/n polymerase system compared to a single vaccinia vims recombinant in which

expression of gp160 was under the control of the poxvinis P7.5 promoter. This dual
vinis infection protocol later demonstrated its utilty for large-scde protein production;

168

expressing yields of 11.2 mg H W gp160 per liter in Vero-ceii microcarrier cultures

(Barrett et al. , 1989)-

Optimal conditions of the new vTME-46IvTF7-3 expression system were

detetmined to a-

maximal ElTLV-1 envelope production. Best yields of the

recombinant envelope protein were observed 36-48 hours following dual infection. There

was no advantage in allowing the infection to progress m e r due to the increased

cytopathic effects and ceU death mediated by the vaccinia virus. Similac kinetics of
recombinant protein expression were described in the dual vaccinidT7 polymerase
system of Fuerst et al. (1987). m e n the effect of virus muitiplicities was studied in our

gp46 expression system, it appeared that a muitiplicity of 2 pfulceil yielded maximal

amounts of glycosylated envelope protein, with higher multipiicities of infection having

no significant effect on protein yields. The recpirement of low viral multiplicities in our
system is in contrast to what was observed in a duai vaccinia system consmicted to

express beta-galactosidase (Fuerst et al., 1987); which required viral multiplicities of 10

pWcell for maximal yields. The biochemical, structural and physical properties of the

HTLV-1 gp46 protein compared to beta-galactosidase are dramaticaiiy different (for

example presence and absence of glycosylation respectively), such that each recombinant

protein WUmost Uely have different demands for the cellular machinery. Indeed, Fuent
et al. (1987) observed large differences in the total yields of protein when they expressed

either beta-gdactosidase, Hepatitis B surface antigen or HlV gp160 in their vaccinia/T7

polymerase system. The accumulation of the semi-glycosylated HTLV-1 envelope protein

169
forms during the course of vTME-46/vTF7-3 infection suggests that processing by the

cellular glycosylation appmti may h o m e saturateci with envelope protein; a possible

causefeffect of toxicity and limitai recombinant RIZV-1 envelope protein production that
has been observed in other heterologous expression systems (Kiyokawa et al., 1984;

Samuel et al. , 1984; Kuga et ai. , 1986;Chen et al., 1989; Nyunoya et al. , 1990; Vile
et ai., 1991). In terms of overall recombinant envelope protein production, the dual

vaccinia expression system appears to pariiaiiy overcome this limitation in its ability to
c o n ~ u to
e produce partialiy glycosylated and unglycosylated envelope protein species
that are stabily maintainecl withn the host cells.

Being aware that recombinant protein expression levels c m Vary between different
cell types, we compared human II9 T-cells to human HeLa epithelial cells as hosts for

dual vaccinia infection and recombinant envelope protein expression. II9 and HeLa cells

are very different in fimction and origin. However, the apparent size of the glycosylated
(49 and 47 kD)and unglycosylatcd (39 kD)envelope protein forms produced by the two
cei types did not Vary significantly. This suggested that similar pst-translational

processing of the recombinant envelope protein occurs in different cell types of

endothelid (HeLa) and lymphocytic (Hg) origia. In terms of protein yields. however,
there was a marked clifference between the two ceIl lines. The vTME-46/vTF7-3 infected

HeLa cells produced three- to four-fold more envelope protein than the same number of
CO-infectedH9 cells. Fuerst and his colleagues also observed low levels of recombinant

HIV gp160 protein in H9 cells CO-infectedwith a vaccinialT7 polymerase system

170

compared to recombinant gp160 levels prouced in CV-1 and BSC-I cells (1986). Pique

et al. (1992) also noted differences in the extent of processiog and yields of recombinant
HTLV-1 envelope protein between ttansfected ceiI types ushg a non-vaccinia virus
expression system.

Native folding of the recombinant HTLV-1 envelope proteins produced by our

dual vaccinia system was suggested by their ability to bind severai HTLV-I envelope-

specific human monoclonal a n t i i e s . Denaturation of the recombinant envelope protein

forms completely inhibited recognition and immunoprecipitation by the human

monoclonal anti'bodies. Conformation sensitivity of the human monoclonal antibodies had

k e n previously suggested by their lack of binding to virai lysate-based western blots and
their recognition of

authentic HTLV-1 envelope protein as measured by

radioimmunoprecipitation and immduorescence of HTLV-I infectecl ceii lines (Rowe

er al. , 1994). In addition, ail of the human monoclonal a n t i i i e s had been found to be
capable of inhibithg HTLV-1 syncytium formation (Rowe et al. , 1994). Thus the ability
of these conformationdependent monoclonal a n t i i e s to bind to the recombinant

HTLV-1 envelope proteins produced in our dual vaccinia system, suggests that the
envelope proteins are king folded and pmcessed in an authentic conformation that
enables the display of discontinuous epitopes involved in HTLV-1 neutralization. hdeed,
the most glycosylated recombinant envelope protein forms (49 and 47 kD) appear to be

processed such that they share similar epitopic structure with the envelope protein
expressed by HTLV-1 infixted human ceils; since the recombinant proteins were readily

171

immunoprecipitated with the full panel of conformationdependent HMAbs, as weli as

several polyclonai sera obtained h m HTLV-1 infected human patients.

Conformational integrity of recombinant envelope protein was also suggested by

the ability of the human monoclonal a n h i e s WA07flG7, WA07/2F7ami WA1l/2F3
to bind to dual vaccinia iafected HeLa ceis as detected by indirect immunofluorescence.

The WAO7IlG7 and WAO7/2F7 monoclonai antibodies had previously exhibited (Rowe

er al., 1994) the most avid neutraliziog/syncytium inhibition properties of the six
monoclonal antibodies tested, inhibithg greater than 90% of syncytium formation at

concentrations less than 5 ug IgG,/ml. Binding of these two monoclonai antibodies to the
recombinant HTLV-1 envelope protein suggested that the protein contains conformational
epitopes that are significant in virus neutraiizatioa/syncytium inhibition.

The absence or weak immunofiuorescence exhibited by the other human

monoclonal antibodies may be due to thei. weaker affinity for the native HTLV-1
envelope protein as suggested by the higher concentrations of these antiibodies required
to attain 90% syncytium inhibition of HTLV-1 infected cells (23-90 ug IgG,/ml; Rowe

et al., 1994). Weak immunofluorescence demonstrateci by some of the monoclonal

antibodies could also be the result of sght seqyencp variation between the envelope

sequence encoded by the recombinant vaccinia virus and that of the vviral strain of the
monoclonal a n t i y hybridoma source. This minor sequence deviation may potentidly
result in antigenic variation and may affect antibody binding efficiency. For example, a

172

signifiant ciifference in monodonal a n t i i y binding affinity was observesi by Edouard

et ai. (19W) due to a single amino acid substitution at position 192 of the HTLV-1

envelope sequence. Signifiant differences in epitope topognphy due to secpence

variation of the two envelope proteins however are dikely, since aU six monoclonal
antibodies appeared to be equally capable of irnmunoprecipitating the recombinant
envelope proteins at an IgG, concentration of 7.5 ug/ml-below the 1 0 % syncytium
inhibition threshold of all six monoclonal anti'bodies.

More likely, the lack of antiody binding during immunofluorescence analysis

may r e d t eom the expression of gp46 in the absence of gp21, by the dual vaccinia

system. Without this transmembrane anchor protein, the orientation and presentation of
the recombinant gp46 (rgp46) on the membrane surface of vTME-461vTF7-3 infected

ceLls may not be identical to native gp46 found on the membrane of HTLV-1 infected

cells. If the HTLV-1 envelope protein naturally exists as an oligomer on the surface of

virions and infectai cells, we may expect the putative gp46 oligomeric complex to exhibit
differentiaiiy exposed epitopes compared to its monomeric form. Distinct differences in
the cooformation aad accessibility of various epitopes of

HIV gp120 monomea and

oligomers has been detected using several domain-specific monoclonal antibodies

(Sattentau and Moore, 1995; Stamatatos and Cheng-Mayer, 1995). Similarly, our
observations of differential binding of the various conformationdependent human
monoclonai antibodies to the celi-associatecl HTLV-1 rgp46, might be explained by the
expression of rgp46 as a monomer in the absence of the trammembrane anchor protein

173
(gp2 1) which may ultimately result in the inavailabity of specific epitopes. However.

observed immunoprecipitation of the recornbmbinant envelope protein by ail six human

monoclonai a n t i i e s , suggests that some epitopes may only become available for

a n t i i y biodig once the recombinant envelope protein monomers are fiee in solution.

Further characterization of the conformationai integrity of rgp46 suggested that

its overall structure and epitope avdability was infiuenced by the complexity of the

oiigosaccharides attacheci to it. The importance of oligosaccharide maturation was

demonstrated by the reduced recognition and immunoprecipitation of Brefeldin A-treated
envelope protein by anti-HTLV-1 conformationdependent HMAbs, polyclonal human
HTLV-I+/HTLV-II:' sera and polyclonal chimp STLV-+

sera. This suggested that a

Iack of trimming and modification (Misumi a al., 1986) of the recombinant envelope

protein's oligosaccharides within the Golgi apparatus, may remit in the masking andlot
distortion of conformational epitopes that are availabie for antibody binding on the
mature, authentidy glycosylated HTL,V-1enveiope protein. lmportantly,recognition of
the recombinant envelope protein by these various sera did not r e w direct binding to

the carbohydrate epitopes, as indicated by the observation that the monoclonal and
polyclonal sera immunoprecipitated the unglycosylated form of the envelope protein
(tuniamycin-treated). In particuiar, it appears tbat the majority of cross-reactiviiy

exhibited by the HTLV-II+ and STLVpp+ sera with the recombinant HTLV-1 envelope

protein may be attn'buted to antibody recognition of structurai similarities of the protein

backbow, and is not simply due to binding of shared carbohydrate residues/epitopes.

174

Simiiar resuits have been descrifi by Javaherian et al. (1992), who observeci that the

majority of neuaalizing a n t i i e s in SN-, infecteci macapes prllnarily bound to the

protein backbone of SN-, gpll0 and did not require the envelope protein to be
glycosylated for recognition (Javaherian a ai., 1992).

The biochemical and conformationai integrity of the recombinant HTLV-I

envelope proteins d e s c r i i in this study, impiicate their potentiai as an effective vaccine
candidate. The conserveci epitopes recognjzed by HTLV-I', HTLV-IIf and STLVp-p+
sera suggest that the recombinant HTLV-I envelope proteins may be capable of eliciting
antibodies cross-reactive with various HTLV isolates, independent of viral primary

sequence.

The preservation of native conformation in the recombinant HTLV-1 envelope

protein preparation may be crucial for its success as a protective immunogen. As

described in this chapter, the recognition of the recombinant envelope protein fonns by
the confo~mtion-dependentmonoclonal antibodies suggests that the protein forms are

presenting virus neutraliziag epitopes and thus may be capable of stimulating an effective
immune response in vivo. Indeed, prior immunization studies conducted with other viral
envelope proteins suggest that a subunit vaccine that presents conformation-dependent
neutralization epitopes may be more effective in coderring protection than a vaccine that
presents exclusively Wear determinants (Haigwood et ai., 1990; 1992; Kang et al.,
1991).

175

In addition, resuts h m this study suggest that proper glycosylation of the

recombinant HTLV-1 envelope protein is important in estabiishing the authentic
conformation of the protein. Severai researchers have observed similar effects of
glycosylation on the antigenic and immunogenic structure of other virai glycoproteins.
For example, it has been previously reporteci tbat differentally glycosylated forms of the
Sem& forest virus p62 glycopmtein (Kaluza et al.,1980) and the HIV gp120 (Benjouad
et al., 1992) exhibited distinct epitopes and altered the protein's antigenicity and

protective efficiency. Indeed, immature glycosylation bad been shown to have a rnarked
effect on the folding of nascent glycoproteins such as HIV gp120 which demonstrated an
altered ability to bind CD4 and a V3-loop specific monoclonal anhibody (Fenouillet and
Gluckman, 1991). Evidently, carbohydrate moieties of viral glycoproteins appear to play

a signifiant role in determinhg the ultimate antigenic and immuwgenic structure.

However, an optimal vaccine against HTLV-1 may reqgire inciusion of a i l five
differentially glycosylated forms of the recombinant envelope protein since each form

may present different epitopes to the immune system. This topic is disfusseci fuaher in

chapter 5.

The recombinant HTLV-1 envelope proteins produced in this study may also prove

to be useful as diagnostic reagents in Light of the fact that they were recognized by
various HTLV-IT infeted human sera and serurn from a Pan paniscus chimpanzee

infected with STLV,,

(Gin et al., 1994). Giri and coiiegues (1994) reported that the

STLV- did aot react with either HTLV-1 gp46 viral antigen, HTLV-1 envelope peptide

176

MTA-1, or HTLV-II peptide K55 foumi on HTLV-I 2.3 blot stnps (Cellular M u c t s ) .

In addition, envelope nucleotide sequence could not be amplified by PCR nom this
inf'ted primate cellular DNA by HlIZV-USTLV-1 envelope-specificprHners (Giri et al..
1994). The observation that the recombinant HTLV-1 envelope proteins, prcxuced in our

vaccinia/ polymerase system, are recognized and immunoprecipitated by sera nom the
distantly related STLVpaPp,suggesu that these proteins may also be helpful in diagnosis
of individuals demonstrating atypical HTLV-1 and HTLV-II seropositivity, such as that

observed amongst the human Pygmy m'bes of Bambuti and Bakola (Bukner et al., 1992;
Goubau et al. , 1993).

In summary, we have successfbliy expressed the HTLV-1 gp46 envelope protein

in a recombinant vaccinialT7 polymerase system. The protein was produced at high

levels in a properly processed and folded form. Glycosylation of the recombinant gp46

in this mammalian system occurs at aii four potential N-linked glycosylation sites and
resembles that produced by an HTLV-1 iafited celi. The biochemical and structurai
homology with native gp46 suggests that the recombinant envelope protein might be
useful as a vaccine in eliciting protective immune responses in vivo, and possibly aid in

identifjhg the celi surface receptor utiiized by HTLV-1 d u ~ infection.

g
In addition, this
protein may prove to be an significant diagnostic reagent for the identifkation of novel

human retroduses.

CaAPTER 4 RAT MODEL OF HTLV-1 INFECTION

4.1 Introduction

The p i e s nopism of HTLV-1 is not restricted to humaas. Evidence indicates

that HTLV-1 infats and transfomis T-cells of a wide range of host animals including

simian (Miyoshi et al., 1982), cat (Hoshiw et al., 19&1),rabbit (Miyoshi et al., 1983)
and rat (Tateno et al., 1984) lymphocytes. Despite the abiliy of HTLV-1 to infect
monkeys and rabbits, these animais do not exhibit histopathological or clinid signs of

disease (Cockerell et al., 1990; Laimore et al., 1991). On the other hand, rats infected
with HTLV-1 are susceptible to disease, exhibithg lymphomas of host ongin (Yoshiki
et al., 1987). Newbom rats and hamsters, which die fkom the HTLV-1 induced

lymphoma, provide a mode1 of acute isease. Juvenile and immunosuppressed aduit rats
have k e n found to be suscepti'ble to HTLV-1 induced cbronic lymphoma and thus may
prove useful in the study of Wal latency and progression to clinid di-

(Yoshiki et

al., 1987; Eguchi et al., 1988).

Development of a rat mode1 of HTLV-1 infection has been favoured by the

availabity of genetically chararacterized inbred saains, handling ease, cost eficiency,
and demonstrateci susceptibiiity to HTLV-1 infection. Yoshiki and collegues were the fust
to report that

HTLV-1 could infect and imrnortaze rat T-ceils (WKA saain) when co-

cultivated with 5-bromo-2'deoxyuridine-treated ATL cells (Tateno et al., 1984). Two

178

of the estabLishedcelis Lines were later found to be transplantable imo newborn syngeneic
rats (Yodoi et al.,

1985) and could transmit HTLV-1 h t o newbom and

immunosuppressed aduit WKA

rats

imrnortaiized ceii Unes fiom several

(Yoshiki et al., 1987). More recently, HTLV-1

sains

of rat were estabiisbed in vitro and upon

inoculation demonstrated the abity to transmit HTLV-1 to the correspondhg strain of

newborn and aduit rats in vivo (Yoshiki et al., 1992). The H'I'LV-1 infkcted T-ceii ine
MT-2 was also fouml to be capable of infecting several inbred saains of newborn and
aduit rats with high efficiency ('Yoshikiet al., 1992; Ishiguro et al., 1992).

Estabshment of a similar rat mode1 of HTLV-I infection in o u laboratory was

critical for the future testing of the recombinant envelope vaccines described in this thesis
and other potential HTLV-1 vaccine candidates still king developed. This chapter

describes the successful infection of addt and newborn F344 rats with the humaa T-ceIl
line MT-2, as weli as infection of a family of newbom pups with a newly isolated human
T-ceii iine ongtuthg from a HAMITSP patient. In addition, a dose titration experiment
was performed to determine the minimal criticai number of MT-2 ceils required for

challenge to achieve 1 0% HTLV-1 infetion in aduit F344 rats. These experiments were
necessary since potentiai vaccine candidates can ody be properly tested for their
protective efficacy when the challenge dose has been confirmed to be absolutely
infectious in the absence of any prophylaxis.

4.2 Materiah and Methods

4.2.1 Cell nes

The HTLV-1 infcted MT-2 and uninfected H9 ceil Lines were obtained fiom Dr. L.
Arthur

(AIDS Vaccine Program, NCI-Frederick Cancer Research and Development

Centre). M T 2 and H9 human T-ceil lines were maintained in RPMI 1640 medium

supplernented with 1% L-glutamine (Gibco-BRL), 10% FCS (Giko-BRL) and 1 0

unitslml of peniciuto G anci streptomycin (Giaco-BRL). Human HeLa cells (Amencan

Type Tissue Culture) were grown in Dulbecco modifiecl Eagle medium containhg 10%
FCS and 100 unitslml of penicillin G and streptomycin.

4.2.2 Short-term culture OC T-celis tram HAMIITSP patient

Fresh HAMITSP lymphocytes were isolated fiom heparinized blood of a HAM/TSP

patient, known as SM, using a Ficoli-Paque (Pharmacia) gradient. The lymphocyte
preparation was cesuspendeci at a density of 5xlCfIml in selection media consisting of

RPMI 1640 supplemented with 20% FCS, 2mM L-glutamine (Gibco), 0.556
penicWstreptomycin, 0.05m.M 0-mercaptoethanol (Giho), and 2 unitslml IL-2
(Gibco), at which time the ceii suspension was seeded in 12-weii plates. The cells were
stimulated with 1: 1000 dilution of phytohemagglutinin ( P m ;Gibco) in selection media
and incubated for 24 hours at 37OC, 5%CO,. T-cellswere selected fkom the lymphocyte

pool by m e r incubation at 37C for an additional 9-12 days in selection media lacking
PHA. FolIowing selection, T-ce11 culhues were maintained for an additional 10 days in

180

maintenance media consisting of RPMI 1640 supplemented with 20% FCS, 2mM Lglutamine, 0.5 % penicillinlstreptomycin, O .OSmM 8-mercaptoethanol, and 10 unitdm1

IL-2. Before ceiis were injected iato rats, examination for expression of HTLV-1
envelope protein by immunofluorescence with the mouse monoclonai antiibody l C l l
(anti-gp46; Palker et al., 1989) was performed (as d e s c n i in chapter 1).

4.2.3 Gaimals and transmission of HTLV-1

Inbred male and female Fischer F344 rats were obtained h m HarIan/Sprague Dawley

Animals. For the initial challenge experiment, 3- to 4-month old adult rats were
inoculated on day O with two injections of lx107MT-2cells, one into the intraperitoneal
cavity and the other directly into the tail vein; on &y 15, 1x1O7MT-2cells were given
intravenously (i-v.) into the tail; for a total challenge of 3x10' HTLV-1 infected ceUs per
animal. One family of newbom pups (Family 1; 10 pups) deiivered fiom an d e c t e d

F344 rnother rat were injected intraperitody (i.p.) at 3 days of age with lx107MT-2
cells, whe the 5-day old infiant rats of Family 2 (11 pups) were given 1.2xlP SJH cells
i.p. AU animals nom this initial challenge were terminally bled under anaesthesia 70 days
post challenge.

To detennine the minimri iafitious dose of MT-2 cells required for 1 0 %

infection, groups of 2 or 3 F344 adult rats (3-4 months old) were given a single i.v. tail
injection of either 1@,104, 16, le,or 10' MT-2ceiis (specifically: 2 rats received 10);
3 rats received

le;3 rats received le,3 rats received 106, and 2 rats received IO7MT-2

cells) . Periperal blood was withdrawn rom the taif vein of these rats under anaesthetic

181
100 days post inoculation. C e h prepared for rat injections were washed twice and

resuspended in komplete RPMI 1640 media. All HTLV-1 infecteci rats were maintained

in a Level III containment facility as required by the Medicai Research Council

guidelines.

4.2.4 Western blot detedion of antibodies against ETLV-1 proteins

Western blot analysis with the HTLV Blot 2.3 kit (Cellular Products) was performed
with rat sera diluted 150 and 1:l according to manufacturer's specifications with the

exception that a goat anti-rat IgG conjugated to alkaline phosphatase (Jackson

Immunoresearch Laboratones Inc.) was used as the replacement secondary antiiody for
detection of the rat seroreactivity. HTLV-I positive human sera provided in the kit was
used as a positive control. Sera from a iininfected rat served as a negative control.

4.2.5 Detection of anti-envelope antibodies using radioUnmunoprdpitation

For the detection of anti-envelope antibodies, the various rat sera were tested in duplicate
for the ability to immunoprecipitate [US]-cysteine labelleci lysates of vTME-46/vTF7-3

infected HeLa ceils. This duai vaccinia system expresses high amounts of recombinant

HTLV-1 gp46 (rgp46; see Chapter 3). Radiolabelling of vTME-46/vTF7-3 infected HeLa

cells and preparation of the resultant cell lysate were performed without modification,
as describeci in Chapter 3 (pages 116 and 117). Cell lysates were precleared for 18 hours
at 4C by incubation with 60 ul

Protein G PluslA Agarose (Oncogene Science), which

had been preincubated for 2h at 4C with 20 ul normal rat sera; volumes of agarose

182
beads and

normal sera mentioned were pqared for each 43x106 ceU equivalent. The

Protein G PludA agarose was pelleteci and the resulting supematants were divideci uito
4.5~106cell ecpivalents. Immunopredpitations were performed at 4C for 18 hours, in

the presence of 40 ul of Protein G Plus/A agame and 1:165 dilutions of the various

HTLV-1 challenged rat sera. Anti-gp46 mouse monoclonai anti'body 1Cl 1, provided by
Dr. Tom P u e r (Duke University, North Carolina; Palker et al., 1989)was diluted 1 :4

and served as a positive control. Normal rat sera @ 150 dilution, served as a negative

control. h u n e complexes were washed 4 times with extraction buffer and then
resuspended in an equal volume of 2X L a e d sample bmer and boiled for 10 minutes.
Samples were electrophoresedusing the tricne-SDS-polyacry lamide gel system developed

by Schagger and von Jagow (1987). The 12% polyacrylamide SDSf6M urea gels were
fixed for fluorography with Entense Solution (Dupont, NEN), dned between sheets of

BioDesign gel wrap (NewYork) and exposed for autoradiography at -70C.

4.2.6 Syncytium Inhibition Assay

Neunalizing antibody titres were determined using a syncytium inhibition assay, which
was performed by Dr. T. Palker's laboratory. A detailed description of the significance

and mechanics of the assay can be fou& in Chapter 2 of this thesis (page 60). Briefly,

serial dilutions of the various HTLV-1 challenged rat senun samples were tested for their
ability to inhibit syncytium formation between HTLV-1 infeted Cg1 PL T-celis and
uninfected C8166 T-celis Routinely, 100-200syncytia were obtained per microtitre well

in the presence of 10% normal rat senim. Neuaaluiog anti-HTLV-1 peptide antisera

183

(Paiker, et ai-, 1992)and pre-immune rat s e m served as positive and uegative controls,
respectively.

4.2.7 Polymerase Chain Reaction

Peripheral blood lymphocytes were Wlated fmn heparinized blood withdrawn from the

tail veins of anaesthetized rats. Erythrocytes were preferentiaily lysed h m an aliquot of

100 ul of whole blood by brief suspension and centrifugation (10,000 x g, for 12

seconds) three times in the prrsence of low Salt buffer (pH 7.2; 1OOmM Tris-HCL;
lmM EDTA). Ceii pellets were suspended in 200 ui of PCR lysis bmer (1.5mM
MgCl,, 50mM K I , lOmM Tris-HCl (pH9.0), 0.1%Triton X-100 and 0.1 mg/mL

proteinase K) and incubated for 4 hours at 37C. The samples were then boiled for 10

min. to inactivate the proteinase K and then storecl at -7(PC. As a positive control, DNA
was extracted by standard methods (Sambrook et al., 1989) fkom HTLV-1 infected MT-2
cells diluted 1:le in H9 T-cefls (negative for HTLV-1).

The reaction mixture destineci for PCR amplification contained ceii lysate
correspondhg to 2x10' PBMCs, 0.5uM of each primer set, O.lmM of each
deoxynucleoside triphosphate (Pharmacia Utrapure), l.5m.M MgCl,, 1m.M Tris-HC1

(pH9.0), S

M KCI, 0.1% Triton X-100

ami 1U of thermostable DNA polymerase

(Taq DNA polymerase; Promega). The beta actin primers (based on published sequence;

Ny Sy et al., 1985), used to insure that DNA samples were capable of supporthg PCR
amplification. were located at nucleotide positions #1867-1883 for the 5' primer (pr.29)
and #2252-2277 for the 3'primer (pr.30) to generate a 410 bp fragment upon

184

amplification at 600C. The primer pair used for singie ampfication cycle amplification
of the envelope gene m e n t (based on published sequence; Seilci et al., 1983) were

Iocated at nucleotide positions #5168-5189 for the 5' primer (pr.2246) and #6067-6102
for the 3' primer (pr.GD7) to generate a 934 base pais m e n t upon amplifkation at
65C. The nested primers used for the two rounds of env seqyence amplification (Seilci

et al., 1983) were located at nucleotide position #Ml-5427 for the 5' primer (pr.406)

and #5424-5650 for the 3'primer (pr.407) to generate a 249 base pair kagment upon
amplification at 600C.
The envelope primer base sequences and ATLL sequence coordinates (according
to the pubiished sequence of Seiki et al..

1983)are as foiiows:

2246: 5'-ACCTCCAACACCATGGGTAAGT-3
' (#5168-5 189)

GD7: S'GATCCTAGGGTGGGAACAGGTGACAAGGAAAAGGGG-3'(#6102-6047)
406: 5'-CAGCTACCATGCCACCTATTCCCTATA-3' (#SKI 1-5427)
407: 5' -GATCCTAGGGTGGGAAC AGGTGAC AAGGAAAAGGGG-3' (#5650-5624)

Deiection of envelope (primers pr.2246GD7) and beta actin sequences @rimers

pr.29130) in the addt rat samples, and beta actin seQuences in the rat pup lysates
involved 40 cycles of amplification. Envelope seqyence amplification from the rat pup
samples required wsted PCR involving two rounds of 35-cycle amplification, the fust
amplification involved the primers pr.22WGD7 fkom which a 1110 volume of the
product was removed and ampfied with a nested set of intemai primers pr.406/407.

185

Amplification cycles consistecl of incubation at 94C for 1 min.. 60-65C for 2 min., and
72C for 2 min..

PCR products were directiy aPalyzed by electrophoresison 7X Tris/Borate/EDTA

acrylamide gels foiiowed by ethidium bromide staining, or by southern blot anaLysis of
amplified DNA electrophoresed on 0.8% agame gels. Observation of PCR arnpfication

in two independent amplification assays was reqim, before a DNA sample was

considered to possess specific DNA secpence complementaxy to the primers used

(positive).

4.2.8 Southern blot analysis of PCR amplified products

The ampiified DNA was electrophoresed on a 0.8 % agarose gel and transferred for 90

minutes at 5 in. Hq. to Hybond-N membrane (Amersham) using a vacuum blotter

(BIORAD) under allraline conditions as descriaed by the manufacturer. The resulting
Southern blot was hybridized at high stringency (65C) with a digoxigenin-UTP-

(Boehringer Mannheim)-labeiiedinternai probe (envelope probe: nucleotides#5694-5730;

sequence nwnbering accordhg to Seiki et al.. 1983) as directed by the manufacturer

(Boehringer Manaheim) and as descri'bed in chapter 2 (page 51). Following indirect

cherniluminescence, the blot was exposed to Dupont Cronex 4A X-ray fiim.

4.3.1 Mecous traasmission of HTLV-1 into ad& rats

To determine if our laboratory stock of human MT-2 celis were capable of transmitting
HTLV-1 to adult Fischer F344 rats, four rats were given an intraperitoneai and

intravenous injection on &y 0, followed by an additional intravenous challenge on day

15, with 1x10' MT-2celislsite. This challenge protocol was a modifiecl version of that

employed by Yoshiki e? al. (1992),which involved two intravewus injections of lx107

MT-2celis. Sera and whole blood was coiiected for Western blot and PCR analysis on
day 70 (10 weeks). AU four rat sera demonstrated anti-HTLV-1 antibodies, with

significant and consistent reactivity to p19, p36 and pS3 and some weak and variable
reactivity to p24, p26 and recombinant gp21 (Figure 4.1, lane B-E). No seroreactivity

to these HTLV-1 proteins was demofl~ttatedby the negative control rats (Figure 4.1, lane
A)

DNA isolated from the peripheral blood mononuclear ceiis of ail four chaiienged

rats was able to support amplification of HTLV-1 envelope sequence (Figure 4.2, lanes

11-14).This PCR amplification was confirmeci by repetition. Negative control rat DNA
was not amplif'iable by the HTLV-1 envelope-specific primers (Figure 4.2; lane 10);
however the sample was capable of supporthg amplification as demonstrated by the beta

Figure 4.1
Seroreactivity of adult Fischer F344 rats with HTLV-1 antigens foilowing challenge with
3x10' MT-2 ceiis. Four F344 rats (lanes B,C,D and E) were inoculateci with HTLV-1
infected celis on day 0, with 1x10' ceiis i.p./lx107 cells i.v. ; followed by injection on
day 15 of M O 7 ceils i.v. Sera obtained from terminal bleeds on day 70 were screened
using commercialiy available western blot strips containhg whole viral lysates. Lane A,
PBS-injected control Fischer M44 rat sera; lane F, HTLV-1 infected hurnan control sera.
Labeiled small arrows indicate various sized HTLV-1 proteins obtaiaed fkom viml lysates
and recombinant gp21. Large arrows are visuai ai& to indicate HTLV-I proteins that
were recognized by the challenged rat sera.

Figure 4.2
PCR amplification of HTLV-1 envelope sequences from peripheral bloo lymphocytes
of adult Fischer rats injected with 3x10' MT-2 ceiis. Lanes 1-8 were amplifieci with betaactin prirners to ensure that each DNA sarnple wouid support amplification; a 410 bp
fragment was expected. Lanes 9-16 were ampLified with HTLV-1 envelope-specific
primers which were capable of generating a 934 bp ftagment (indicated by arrow). PCR
amplification was performed on: lanes 4-7 and 11-14, rats #1-4 injected with HTLV-1
infected MT-2 ceiis; larus 3 and 10, a control rat injected with PBS; lane 8 and 15,
HTLV-1 infectesi MT-2cellular DNA; lanes 1,2 and 9, PCR reagent blank controls.
PM2,PM2 DNA digesteci with Haem as a molecular weight marker.

191

actin fragment generated in the presence of the control set of prirners (Figure 4.2, lane
3) *

4.3.2 HTLV-1 infecfion of newborn n

t pups

To repeat andfor improve the hcpency of HTLV-I infection of newbom F344 rats
demonstrated by Yoshiki e? al. (1992). pups were injected with MT-2 cells or with a

novel mixed T-cell lymphocyte culture, SM, isolated from a HAWTSP patient. One

family of newborn pups (Family 1, 10 pups) devered h m an uninfected F344 mother

rat were injected i.p. with 1x107Enr-2 ceils, while the newbom rats of Family 2 (11

pups) were given 1.2xl(PSJH ceUs i.p. At day 70 (10 weeks), Western blot analysis
revealed that none of the rat pups exhibited seroreactivity to any HTLV-1 proteins (data
wt shown). However, HTLV-1 envelope sequence could be ampiifed in 1 0 % of the

DNA samples isolated !Yom the young rats' PBMCs.The HTLV-1 genome was detected
in PBMCs h m 10/10 rat pups from Family 1 (Figure 4.3,lanes 3-12)and 1 111 1 rats
from Family 2 (Figure 4.4; lanes 3-13). DNA nom both uninjected mother rats was not
amplifiable by the HTLV-1 envelope-specific prirners (Figure 4.3. lane 2; Figure 4.4,

lane 14);however both samples could support beta actin amplification (data not shown).

Figure 4.3
PCR amplification of HnV-1 envelope sequences fiom peripheral blood lymphocytes
of family 1 rat pups that were injected with 1x10' MT-2 cells i.p. (ianes 3-12).
Envelope-specific seqyence was detected in the lymphocyte DNA extracts using nested
HTLV-1 envelope-specific primers (249 bp products are indicated by arrow), 70 days
post challenge. Lane 2, uninjected F344 mother of pups; lane 13, HLV-1 infected MT-2
cellular DNA; lane 1, PCR ragent controls; lane 14, PM2 DNA digested with Hael&?
as a molecular weight rnarker.

Family 1
Rat pups

Figure 4.4
PCR amplification of HTLV-1 envelope sequeoces fkom peripheral blood lymphocytes
of rat family 2 pups that were delivered from an uhfkcted F344 mother rat and were
injected as newbom intraperitoueaily with 1.2~106ceiis fkeshly isolated from a
HAM/TSP patient, SJH (lanes 3-13). Envelope-specific secpence was detected in the
lymphocyte DNA extracts using nested HTLV-1 envelope-specific primea (249 bp
products are indicated by arrow), 70 days post challenge. Lane 14, uninjected F344
mother of pups; lane 16, HTLV-1 infecteci MT-2 cellular DNA; lanes 2, 15, and 17,
PCR reagent controls of phase 1 and 2 amplifications; lanes 1 and 18, PM2 DNA
digested with HaeZZI as a molecuiar weight marker.

Famiy 2
Rat pups

4.3.3 Determination of mininiai infectious dose

A titration of MT-2 ceUs was performed to detemiiire the minimum number of ceils

required to Uiduce 100%freqyericy of HTLV-1 iafection in adult F344 rats. Rats were
injected intravenowly with a single dose of either 103, 104, I d , W,or 107 HTLV-1
infected cells. Western blot arialysis of the sera collected LOO days (14.2 weeks) post
inoculation (Figure 4.5) revealed that ody rats injected with either 106 or IO7MT-2 cells

consistently exhibited anti-HTLV-1 a~1aies.The sera obtained nom adults injected

with IO7MT-2ceils contained significant a a t i y responses to gp21, p 19 and p36; with

weaker respooses to p26 and p53 (Figure 4.5, lanes 2, 3 and 4). The three rats injected
with 106 cells demonstrated seroreactivity to the pl9 antigen consistently (Figure 4.5,

lanes 5 , 6 and 7). However, rats challenged with fewer MT-2 ceLis dernonstrated
sporadic seroreactivity to pl9 and occasionally to p24, with not all animals in the group

seroconverthg (Figure 4.5, lanes 8-15).

Results of the Western blot analysis were reiterated in experiments which

demomtrated the ability of PCR to ampiify HTLV-1 genomic seQuence from the
correspondhg samples. Southemblot analysis was performed on PCR products following
amplification with HTLV-1 envelope-specific primers (Figure 4.6). AU rats that exhibited
anti-HTLV-I antibodies in their sera contained HTLV-1 amplinable envelope sequence

in their penpheral blood mononuclear cells. Thus, only DNA isolated from rats injected
with either 106or 107MT-2ceils consistently supported HTLV-1 amplification. Al1 rat

Figure 4.5
Seroreactivity of adult rats with HTLV-1 antigens foilowing challenge with various
amounts of MT-2 cells to determine the minimai infectious dose required for 10096
infection. Each rat was given a single intravenous tail injection of MT-2 ceiis in
quantities of IO7(anes 3,4); 106 (ianes 5,6,7); 10' (ianes 8,9,10);IO' (ianes ll,12,13);
and 10' (lanes l4,E). Lane 2,sera fcom previous challenge of 3x10' MT-2 ceUs (rat #3,
lane D of figure 4.1); lane 1, HTLV-1 iafected human control sera. Sera was obtained
on day 100 and screened using commercialiy available western blot strips containing
whole viral lysates. Labelled small arrows indicate various sized HTLV-1 encoded
proteins and recombinant gp21.

Figure 4.6

PCR amplification of H T W I envelope sequences fkom peripheral blood lymphocytes

of adult Fischer rats injected with various amounts of MT-2 ceiis to determine the
minimai infectious dose r e w for 100% infection. A single intcavenous tail infection
of MT-2 ceils was given to each rat of IO7 (lanes 2,3); 106 (ianes 4,5,6); 1 v ( l m
7,8,9); 104 &mes 10,11,12); l(Y (Ianes 13,14). Lane 15, HTLV-1 infectcd MT-2cellular
DNA PCR reagents control; lane 1, PCR reagents control. Cellular DNA was amplified

using HTLV-1 envelope-specific primers which were capable of generating a 934 bp

fragment (indicated by arrow). Southern blot analysis of amplified products was
performed at high stringency using an envelope-specific oligonucleotide probe, to ensure
the identity of the amplifid ftagment.

201
DNA samples could support equivalent yields of PCR amplied product in the presence
of ceil-specific beta actin priwrs (&ta not shown).

As previously

mentioned, despite Western blot reactivity to various HTLV-1 antigens,

particuiarly to the gag proteins, no anti-gp46 anti'bodies could be detected in any of the

challenged addt rat sera (Figures 4.1 and 4.5). The rat sera did not recognize the gp46

envelope protein contained withia the whole Wal lysate, nor the recombinant gp46
peptide (aa 162-209) which impregnates the Western blot strips produced by Diagnostic
Biotechnology.

For a more sensitive assay system, the sera fiom the addt rats injected with
greater than 106 MT-2 cells were tested for their abiLity to immunoprecipitate the high
levels of rgp46 available from the recombinant vaccinid T polymerase system (vTME46/vTF7-3; chapter 3). As revealed in Figure 4.7, ail adult rats receiving a challenge

dose of greater than 106 MT-2 cells were able to immunoprecipitate rgp46 at dilutions
of 1 :165. No reactivity to rgp46 was observed in the normal rat senun samples (Figure
4.7A, lane 1 and F i g w 4.7B,lane 1).

Figure 4.7
Radioimmunoprecipitation of recombinant envelope proteins by rat sera following
challenge with various amounts of HTLV-1 infecteci MT-2ceiis. f5S] -cysteine-labeiied
proteins from the dual vaccinia vTME4/vTR-3 infecteci HeLa ceils capable of
producing rgp46 were immuwprecipitated with the different semm samples. Molecuiar
masses of the variantly glycosylated forrns of the HTLV-1 envelope glycoprotei.
produced by the recombinant dual vaccinia system are indicated on the right side of each
photograph.

Figure 4.7A: Detection of anti-envelope antibodies in sera obtained from rats

injected with 3x10' MT-2ceus; receiving on day 0, ln107 ceiis i.p./lx107 celis
i.v.; foilowed by injection on day 15 of lx107 ceiis i.v. Sera was obtained from
terminal bleeds on day 70. Lanes 2-5, sera Born rats injected with MT-2 cells;
lane 6, anti-HTLV-1 gp46 mouse MAb, 1C11; lane 1, PBS-injected control
Fischer F344 rat sera.

Figure 4.78: Radioimmunoprecipitation of rgp46 by rat sera obtained nom rats

injected with lx107 MT-2 ceils (lanes 2,3) or 1x106 MT-2 ceiis (lanes 4,5,6).
Sera was obtained tom bled on &y 100. Lane 7, ami-HTLV-I gp46 mouse
MAb, ICI1 ; lane 1, PM-injecteci control Fischer F344 rat sera.

204

Despite the presence of anti-gp46 antibodies, none of the sera obtained b m the

chaiienged aduit and infant rats contained neutralizing antibody, as defined by in vino
HTLV-1 syncytium formation assays.

4.4 Discussion

In the present study, HTLV-1 trammission to Fischer F344 rats was successfLy
demonstrated. In initiai experiments, infection of 100% of adult rats was attained when
the mimals were inoculated i.p. and i.v. with a total of 3x10' MT-2ceus. This challenge
protocol differed from an earlier study of YoshiLi et al. (1992) which involveci i.p. and
i.v. inoculation of 2x10' MT-2 ceis and resulted in infection of only 80% of the F344

rats.

S k e our rat mode1 was to be used ulcimately for testing the eficacy of potential

HTLV-1 vaccine candidates, it was necessary to detemine the minimal infectious dose

of MT-2ceiis required for 100% infection. We wished to reduce the number of MT-2
ceiis used in our first challenge trial and that reporteci by Yoshiki et al. (1992) to avoid
unnatural infection parameters. Too large a challenge dose would unreaiistically

overwhelm the immune system and not effectively test the protective properties of a
vaccine. From our titration experiment, it appears that the optimai challenge dose is a
single injection of 106 to 10' MT-2 ceils, since inoculation of fewer ceiis resulted in

inconsistent infection among animais within a group.

Our Iaboratory believes that the ability of the various rat samples to support PCR
amplification with HTLV-1 envelope-specific primers, redts From a genuine infection
in which the HTLV-1 genome has passed from the inoculated HTLV-1-infected cells into

206
host rat celis, rathet than a persisteme of AIZV-1 infecteci human ceiis. It is uolikely
that human c e k couid remain uatejected for 10 to 14 weeks; as experimentally supporteci

in the previous work of Suga et al. (1991) who witnessed an absence of PCR signal with
human-specific primers in HTLV-1-infcted Fischer rat sarnpies foliowing a sirnilar

challenge protocol and incubation t h e . Attempts at PCR amplification with humanspecific primers, of the rat samples described in this chapter, are c m n t l y umierway and
wili be eventually included as necessary controls in the forthcornhg manuscript.

In a vaccine protection triai it may be useful that each immunization proiocol be

challenged with both independent inoculum doses of 106 and IO7 MT-2 cells. Each
challenge dose wiii most Likely induce different immunoiogical responses as suggested
by the variant seroreactivity observed between the two inoculation groups. Despite 100%

infectivity in both groups, as coafirmed by PCR, rats receiving IO7 cells exhibited
antibodies to several HTLV-1 antigens, while the group receiving 106 celis possessed sera
reactive with ody p l 9 and rgp46. Determination of Wus load in these two challenge

groups has wt yet been performed. However the differences in seroreactivity exhibited
by the two groups, suggest that the use of both challenge doses may be a mode1 for

human contact with body fluids h m high Wemic versus low virexnic individuais, or as
seen in transfusion settings. It may be that protection of individuais exposed to a lower
viral dose may require a different cascade of immunologicai responses than a person who
is subjected to a more infectious hoculum, in order for it to be effective.

207

In this chapter, the successfiil infection of neonatal F344 rats was descrbed.
uioculation of the rat pups with MT-2ceils d t e d in provirai HTLV-1 integtation in
their peripherai blood mononuclear celis as demoastrated by ECR analysis. Despite

evidence of HTLV-1 provirai presence, none of the pups exhiiited any detectable antiHTLV-1 antibodies at day 70. This lack of a TH2-type response in the chaienged infant

rats, suggests that their immature immune systems may have developed primarily a TH,type or tolemce respoase to the surface expressed HTLV-1 antigens. The absence of

anti-HTLV-1 a n t i i i e s in PCR-positive ne0nata.l pups has been previously described

foilowing similar challenge protocols (Yoshiki et al., 1992; Ishiguro et al., 1992;

Ibrahim et al., 1994). This induction of neonatai tolenince to the HTLV-1 proteins, is

similar to the graft acceptance observed in adult rats, if the rats are injecteci as newborns
with lymphoid cells of the future gr&

donor (Miller et al., 1989).

Unfortunateiy due to tirne constraints, we had to sacrifice these young rats at day
70. It wouid have k e n iateresting to observe whether these pups wouid have developed

"HAMfTSP-me"symptoms Like the W U rats described by Isbiguro and his coilegues

(1992). They observed that 3/3 seronegative HTLV-1 d e r rats of the WKA saah

developed a spastic paraparesis of the hind legs at 16 rnonths of age. Thus far they have
not witnessed the same symptom in other strains of rats; however theu numbers of rats

in each group were limited. These diseased WKA rats can only be tentatively described
as "HAWTSP-like" since the pathogenesis is not identical to that seen in humans

(Ishiguro et ai., 1992). Unlike W T S P patients who exhibit high titred antibodies to

208

HTLV-1 in both the cerebral spinal miid (CSF) and senun (Osame et al., 1986; 1987;
Vernant et al., 1987). "HAM-ikewrats possess no detectable anti-HTLV-1 antaodies in

their CSF or senun (Ishiguro et al., 1992). Secondiy, in contrast to the affcted spinal
cord in HAMfTSP patients (Akisuki et al., 1988; Lwasaki, 1990), "HAM-like"rats show

no lymphocytic infiltration in the white or grey matter, or in the perivascular area

throughout the whole spinal cord. Oniy foarny macrophages predominantiy infiitrate the
affecteci lesions of these WKA rats (Ishiguni et al., 1992). Whetber the absence of

Lpphocytic infiltration in the affecteci lesion of the "HAM-Wrenrat reflects: (i) the stage
of the disease process, () species difference in the host response between rats and

The high levels of rgp46 produced in our vaccinia/'L7 polyrnerase expression

system, has enabled the unprecedented detection of anti-gp46 antibodies in HTLV-I

infected Fischer F344 rats. In previous studies (Suga et al., 1991; Ishiguro et al. , 1992;
Ibrahim et al., lgW), the failure to detect anti-gp46 antibodies using commercial Western
blot strips most like1y resulted fiom the lack of a sensitive assay system and not h m an
absence of antibody. Indeed, the use of commercial Western blot strips does not appear
to be appropriate for screening Fischer rat sera for anti-gp46 antibodies. The recombinant

gp46 peptide (aa #162-209) incorporated into the Diagnostic Biotechology kit, had

previously k e n selected on the basis of its strong reactivity with human HTLV-1 sers.
Unfortunately this highly immunogenic peptide spam only 15% of the total prirnary
sequence of the envelope protein, and thus does not aUow detection of antibodies in

209

individuals/animalr possessing restricted immune responsiveness. The lack of

incorporation of the entire HTLV-1 envelope protein into the commercial Western blot
strips, may account for the abundance of false-negative resuts gathered upon screeniag
of HTLV-1 chailenged rat sera (Suga et al., 1991; Ishiguro et al., 1992; Ibrahim et al..
1994; this study)

The differentiai recognition of envelope protein epitopes by Fischer rats was

evident in this current study. The rat sera selectively immunoprecipitated rgp46 producecl
by the dual vaccinia system, but failed to exhiiit anti-gp46 peptide Western blot

reactivity; only anti-gp21 antt'bodies could be detected. In contrast, HTLV-1 infected

human sera were capable of recognizing both denaaiced and authentic forms of gp46
(Suga et al. , 1991; Dekaban et al. , 1994; chapter 3). From the coiiective observations

of this study and othen (Suga et al., 1991; Ishiguro et al., 1992; Ibrahim et al., 1994),
it appears that the Fischer F344 rat strain is limited in its abity to recognize epitopes

of the HTLV-1 envelope protein, particularly Iiaear epitopes. The reasons for this

immune restriction have yet to be determineci. However, the observation that the
detection of

anti-envelope antibodies

in the rat sera

was

limited to

tadioimmunoprecipitation analysis, suggw that these rat antibodies may be specific for

conformational epitopes of the HTLV-1 envelope protein. The restrictd seroreactivity

with rgp46 observed in the challengecl rats was smiilar to that observed in the "HTLV-

indeterminate" human s e m samples describeci in Chapter 3. These "HTLVindeterminate" humans were determined to be IFnV-positive by PCR analysis; however

210

like the chalienged rats, they did not exhibit envelope protein reactivity upon Western
blot analysis and yet were able to ready immunoprecipitate rgp46, produced by the dual
vaccinia virus system.

Despite the demomtrated presence of anti-gp46 anh'bodies in the sera of all the
rats injected with > 106 MT-2celis, none of the challenge sera was capable of inhibithg

HTLV-I syncytium formation. This lack of neutraliPng ability of the Fischer rat sera,
was also observed by Ibrahim et al. (1994) in several strains of rats including Fischer
F344, Brown Norway, BB and LRwis, and may be due to the geneticaily restricted

recognition of envelope protein epitopes that are not involved in syncytium brmation.

In addition, Mmunoprecipitation of rgp46 by the existing Fischer rat antiiaodes was

observed to be relatively weak, requiring extended exposures to autoradiographic f i .

This low-level reactivity suggests that the anti-envelope antibodies may be of low titre
or low affinity;possibly explaining theu lack of effective neutralizing activity . This poor
responsiveness of the challengeci rats to the HTLV-1envelope protein may be similar to
that seen in (BlO.AxA/WySn)F,, H-2"" mice that are genetic nonrespondea to the
envelope protein of F-MuLV (Earl et al., 1986; Ishihara et al. , 1991). If this is the case,
low immune responsiveness may hopefblly be overcome by immunjzation of the Fischer

rats with a gp46 subunit vaccine, as observed in the H - p mice upon injection with an
emulsion of denatured F-MuLV envelope protein and adjuvant (Ishihara et al., 1991).

Similady, in rabbits and various mouse strains, genetic restriction to the biologically

211

important E1TLV-1 envelope protein region, SP-4A,was overcome by its incorporation

into a chimeric peptide (Lairmore et al., 1995).

It will be interesting to fhd out if a signifiant and effectve anti-envelope

response c m be ultiniately stimulated in the Fischer rat strain. Caution must be taken to
iosure that the meamcement of anti-envelope responses is not solely based on neutralizing

antiiiody presence. Indeed, Ibrahim et al. (1994) observed a lack of neutdzhg antibody
detection in the s e m of various straias of rats despife the resistance or relatively low
susceptibility to HmV-1 infection demonstrateci by some strains, suggesting that it is
improbable that neutralizing anhaody plays a significant role in the protective immunity

of rats.

The Fischer rat model wili hopefully conm'bute to our understanding of HTLV-1
infection and disease progression, and may ultimately aid in the development of an

effective vaccine. The fact that the Fischer F344 rats are an inbred strain, will enable
their use in adoptive transfer studies; a aecessity for detennining the correlates of

protection induced by various HTLV-1 vaccine candidates. In addition, this rat strain has
k e n used by Hori et al. (1995) to develop an infection model that wiii aiiow the
distinction of intrauterine transmission of HTLV-1, from intraparttun ;insmission during
delivery. This model involves the delivery of pups ushg a cesarean section fkom HTLV-1
infected mothers. They have also described a protocol that wi ailow examination of oral
transmission of HTLV-1 through breast feeding andor matemal saliva. These newly

212
established models will pmve usehl in the development of effective vaccines that are
f'd on the prevention of prenatal ard postnatal tmsmission of HTLV-1. Their

development in our lab WU

be our next prionty.

The current establishment of the HTLV-I infection mode1 in adult and n e o ~ t a i

Fischer F344 rats, as d e s c n i in tis chapter, can aiready be implemented for
e v a l u a ~ gvaccine candidates developed in our laboratory. Variation of the HTLV-1
uifected ceH challenge dose may mimic different ciinical situations encountered in

humans and may aiiow more effective and realistic testing of the vaccines' protective
potentiai .

CsAPTER 5 PERSPECTIVES

The work descri'bed in this thesis involved the expression of the HTLV-1 envelope

protein by two different Virayceil systems. The recombinant bacuiovirus system was
employed to produce the entire HTLV-I envelope protein, gp63. In this insect cell
system, the entire envelope precmor was produceci in signiticant amounts, however it
was insoluble and the majority of protein was incompletely pst-translationaliy processed

(env-I .B.) .As an alternative source of recombinant HTLV-1 envelope protein, the surface
HTLV-1 envelope protein, gp46, was expressed in the vaccinia virus/-

polymerase

system. This mamxnaiian system produceci high levels of recombinant gp46 in a properly
processed and folded form. Glycosylation and overall protein conformation of rgp46
resembled native gp46 produwd by an HTLV-1-infectaiceii. As will be discussed in this
final chapter, variation in the conformational and biochemical structures of the two

recombinant protein preparations has endowed them with significantly different

properties. These specifk differences may prove to be important in the future use of

these antigens as diagnostic reagents and potential vaccine candidates. Indeed the
establishment of the HTLV-1 iafection mode1 in adult and neonatal Fischer F344 rats, as
described in this thesis, wi enable the future evaiuation of the different antigens for their

immunogenicity and protective properties.

214

5.1 DiiosOc srreening with recombinant envelope protein antigens

With the two expression systems avaiiable for the production of recombinant

HTLV-1 envelope protein, the utization of their products for diagnostic purposes is
eminent. Characteristic properties of each particuiar antigen preparation should be
considered and its use wili depend on the dernards of the specfic assay.

The conserved conformationai ancl glycosylation epitopes of rgp46 has enabled

the highly sensitive detection of antibodies in sera of individuals/animais infect& with

HTLV-1 and other related retrovinws. The recombinant HTLV-1 envelope proteins,

produced in the vaccinia/T7 polymerase system, were successfully recognized and

immunoprecipitated by sera fiom HTLV-II-infected individuais and chimps infected with
the distantly-reiated STLV-

rgp46 by STLV,,

(Arp et al., submiaed; chapter 3). The recognition of

chimpanzee sera was particularly significant, since Gin and collegues

(1994) had previously reporteci that STL,VpaPpSection could not be detected in this

animal by commercial western blot analysis or PCR ampiifkation. The effective use of
rgp46 in the screening of various non-human primate and human sera, as described in
this thesis, suggests that the antigen may aid in the identification of other HT'LV-related

human retrovinises and provide insights into the etiopathogenesis of other human
diseases.

215

The conformationai authenticity of rgp46 prepmtion bas also enableci the

unprecedented detection of anti-gp46 anhbodes in IiTLV-1 tofected Fischer F344 rats
(Arp et al., subrnitied; chapter 4). In previous studies (Suga et al., 1991; Ishiguro et al.,

1992; Ibrahim et al., 1994), the failure to detect anti-gp46 a n t i i e s using commercial

western blot strips, most Wrely d t e d fiom the lack of a sensitive assay system and not

from an absence of antibody. Indeed, the use of commercial western blot strips does w t
appear to be appropriate for screening Fischer rat or pan paniscus chimpanzee sera for

anti-envelope protein a n t i i e s (Suga et al., 1991; Ishiguro et al. , 1992; Ibrahim et al..
1994; Giri et al., 1994; Arp et al., subrnitted). The recombinant gp46 peptide (aa #162209) incorporateai into the Diagnostic Biotechology kit, had previously been selected on
the basis of its strong reactivity with huma. HTLV-1 sera. Unfortunately this highly

immmogenic peptide spans only 15% of the total prMary sequence of the envelope
protein, and thus does not allow detection of autibodies in individuaislanimals possessbg
restncted immune responsiveness. In addition, analyzing the immune response to gp46
using westem blotting moa likely yields information on a very miwr subset of the total
antibody population, and does not ailow the detetion of antibodies that are sensitive to
protein conformation. The predominance of conf'ormationally-dependent antibodies in

HIV-infecteci sera has been noteci by Moore and Ho (1993) who observed that less than
1% of the total anti-gpl20 antibodies of HIV-positive polyclonal sera was capable of

binding to either short envelope peptides or to denatured gp120 under western blotting

conditions. Thus it appears that conclusioas drawn from western blotting analysis are of
resmcted value, as witnessed in the STLVpan-P-infected
chimpanzee (Gin et al., 1994) and

216

the H'LV-1-infecteci rat modeis (chapter 3), when the total immune cesponse to the

retroviral envelope protein is to be considered. The sensitivity achieved using rgp46 in

radioimmunoprecipitation assays, as reporteci in this thesis, suggests its potential
anti-envelope
effectiveness for the detection of low level and confol~mationally-dependent
anti'bodies in various animal and human sera.

The env-I.B. preparation expresseci in the wct ceils has aiready been
successfidiy used as the target antigen in a comparative study of the anti'body response

in HTLV-1 asymptomatic carriers and HAMITSP patients (Dekaban et al., 1994;

Appendix I).

This analysis reveaied marked distinctions in seroreactivity between the two

groups of individuals (Dekaban et al., 1994; Appemlix I). W T S P patients had

significantly elevated anti-envelope anti'body responses involving the IgM, IgA and IgE
isotypes. R d t s nom this serologicai analysis suggest that a longitudinal study following
the asymptomatics anti-envelope isotype profile, may help to determine whether an

asymptomatic individuai is progressing towards the development of HAM/TSP. Detection

of "high-riskW
Udividuals by monitoring alterations in the isotype profile may enable

effective medical intervention by signalling a need for neurological examination and the
employment of appropriate drug therapies such as high dose steroids (Osame et al.,
1980; Kira et al., 1991) or treatment with antivuals such as AZT and LFNy (Gout et

al. , 1991; Oger, unpublished data). While the percentage of

HTLV-1 infected

asymptomatic individuals who wiil go on to develop W T S P wili be small, once full-

217

blown HAM/TSP develops. there are no usefi treatments for this aggressive disease
which is progressive in nature.

The env-I.B. preparation may prove to be the preferred antigen to carry out such
a longitudinal sady, since it is expressed in large amounts and is readily purified. The
lack of protein conformation of the env-I.B. would not be a concem since long-term

serologid analysis would most kely employ western blot as the diagnostic method,
since this protocol is more cost effective and more easily employed than
radioimmunoprecipitation anaiysis. The correlations between isotope profiies and disease

were originally made using env-I.B. and thus recommendations for medical intervention
could only be made based on seroreactivity to this target antigen. A similar extensive
isotype analysis using rgp46 would be necessary, before it could be used as the target
antigen. Due its extensive glycosylation, the western blot seroreactivity to rgp46 and the

observed isotype profdes of the various patient groups, may be very dBerent from those
observed using env-I.B.

In addition to the utilkation of the recombinant HTLV-1 envelope protein

preparations in antibody-based diagnostic assays, these antigeus could be used to detect
cellular immunity to HTLV-1 antigem. The identification of recombinant HTLV-I

envelope preparations that are recognized by T-ceils in a majority of individuals in the

population (immunodominant and universal) may prove usehl to detect a previous

exposure to HTLV-1. For example, in a collaborative shidy described in Appendix II

218

(Manca et al., 1995b), env-I.B. was readiiy recognized and induced proiiferation of
CD4+T-ceis obtained fiom naive individuais. These immunogeaic properties of env-I.B.
suggest its potential usefulness as a reagent to test for the presence of HTLV-1 envelope-

specific cellular immunity. As of yet, the recognition of rgp46 by naive human T-cells
has not been detemiined. Once both HTLV-1 envelope preparations have been assessed,
the recombinant antigens shouhi then be screened with T-cells from iadividuals who had

known previous contact with EFIZV-1. The recombinant antigen(s) that demonstrate(s)
frequent T-ceil recognition in bath human ceil populations could potentialiy be included

in routine screening protocols of cellular immunity in seronegative high-risk individuais.

5.2 Recombinant envelope proteins as potentiai vaccine candidates

The recombinant proteins may dso serve as effective vaccine candidates to confer
protection against HTLV-1 infation, either alone or as subunit boosts. Thus far, only the
immunogenicity of the env-I.B. preparation has been tested in vivo. The baculovinisproduced aggregates of envelope precursor protein were significantly immunogenic in
various strains of mice (Arp et ai., 1993; chapter 2). Despite the induction of high amienvelope antibody titres in these in vivo studies, when env-I.B. was admiaistered as the
sole immunogen, neutraking antibodies capable of inhibithg HTLV-I syncytium

formation were rarely stimulated. However, env-I.B. was effective in enhanckg the

neutralizing antibody responses in animals primed with a recombinant vaccinia vims

219

encoding gp46, R W E3s. These fimings suggest that despite the lack of authenticity of
env-I.B. to the cooformationai structure of native HTLV-1 envelope protein, env-I.B.

could be effectively employed as a boosting immunogen in a combined immunbation

regime.

The recombinant vaccinia virus R W E3s, used in our triai combined

immuxlization regime, encodes the same gene ftagment of the HTLV-1 surface gp46

envelope protein, as that expresseci in the vaccioia/T7 polymerase system. Multiple

injections of R W E3s was successful in generating HTLV-1 neutralizing antibodies (Arp

er al., 1993; chapter 2). Thus, although no immunogenicity studies of the mammaiian
cell-derived rgp46 have been performed, one can anticipate that rgp46 may dso prove
to be an effective boostbg immunogen.

Challenge experiments have yet to be performed with the HTLV-1 envelope

proteins produced in either the baculovinis or vacciniain systems. The biochemicd and

structural propertes of the two protein preparations are very different and thus one would
anticipate that their antigenicity an immunogenicity would also be significantly distinct.
As described in several other Wal antigen systems, and as will be discussed below,

glycosylation and conformation of Wal envelope proteins may be cntical to theu

protective potentiai in immunized animals. Fortuaately, the establishment of the Fischer

F344 rat challenge model, as descriaed in this thesis (Arp et al., submitted, chapter 4)

220
wili enable the h i e testing of both of these recombiaant subunit vaccine candidates for

their protective efficacy against HTLV-1 iafction.

5.2.1 GlycoBylatioa effefts on antigeniCity and immuwgenicity

Not until recenrly, bas the eEects of glycosylation on the antigenicity and
immunogeaicity of proteins been ackmwledged by the scientinc community. As
described in this section, it is necessary to cntically analyze the glycan structures

attached to antigens, particularly those destiwd for vaccine triais. Unfortunately, these

studies are dependent on the development of appropriate reagents and assays, which are
sometimes grossly l a c h g in certain biological system. The current literature suggests
that the immunogenic aml antigenic effects of the protein and its attached

oligosaccharides are protein-specific. It appears that no generalities can be drawn. Thus,

one may create hypotheses based on the similarities of various antigen systems, but

ultimately, the biologicai effects of the candidate glycoprotein/proteins must be tested

empincally.

As descn'bed in this thesis, oth the vaccuiia/T7 polymerase and baculovirus

expression systems were capable of supporthg the glycosylation of recombinant HTLV-I

envelope protein. However, one would anticipate that the oligosaccharide complexity of
the rgp46 and env-I.B. protein preparations would be significantly different, since the

221
glycosylation machieery of mammaiianceils varies drarnatially ftom that of insect celis
(Butter ancl Hughes.1981; Wojchowski et al., 1987; Kwoda et ai., 1990; Wathen et ai. ,

1991). Indeed, glycosylation of the recombinant HTLV-1 envelope protein in the

vaccinia/T7 polymerase system was extensive and resernbled the pattern obsemed in

HTLV-1 infkcted ceils (chapter 3). Glycosidase digestion revealed that the 49 kD protein
fonn produced in vTME-46/vTF7-3-infected mammaiian celis. possessed mannose-rich

a d o r hybrid oligosaccharides at all four porentiai N-glycosylation sites.

One of the envelope protein forms produced in the bacuioWus system also
exhibited N-liaked glycosylation (chapter 2). Unfominately, the complexity of the
oligosaccharides attached to this 63 k D protein couid not be determined due to the
relative bsolubility of the protein. However Born several extensive studies of insect ceUs
and their pst-translationai machuiery (Butter and Hughes, 1981; Wojchowski et al.,
1987; Kuroda et al. , 1990; Wathen et al. , 1991), it wouid be expected tbat the N-glycans

attached to this form of env-I.B. would be simple high mannose and truncated-

trimannosyl structures, lacking complex side chahs (Kuroda et al., 1990; Wathen et al.,
1991). These glycan structures are consistently observeci on glycoproteins expressed with

the baculovim system since, insect ceiis appear to lack the enzymes to process the high-

mannose precursor oligosaccharides into mammalian-typecomplex structures (Butters and

Hughes, 1981). Core hicosylation is often retained on the truncated structures made in
insect cells however no extensive branching involving outer N-acetylglucosamine and

222

galactose residues is observed, unke the cornplex structures of mammalianderived

oligosaccharides (Kuroda et ai., 1990; Wathen et ai., 1991).

The effects of oligosaccharide structural variation on the relative antigenicity and

immunogenicity of the Merentially glycosylated forms of env-I.B. and rgp46 wili have

to be determined empincaily in the firture. Affinity and plant lectin chromatography will
hopefully aiiow the selective isolation of various glycosylated protein forms without the
need for endoglycosidase treatment of the protein preparations in vitro. Immunkation
shidies involving the variantiy glycosylated envelope protein forms, as a mixture or

individually, wiU hopefully reveal which carbohydratedependent and -independent

epitopes will CO&

protection against live EITLV-1-infectecl ceU challenge. In addition,

these comparative immunhtions will enable us to determine the conm'bution of the

glycosylated trammembrane protein gp21, that is missing in the vaccinia/T7 polymerase

rgp46 preparation, for the induction of immunological protection. This transmernbrane

glycoprotein bas been associateci with the principle fsogenic property of HTLV-Iinfected cells (Delamarre et al., 1994) and thus generation of an immune response to
regions of gp21 may be important in prevenhg HTLV-I infation. Uafortunately, one

cannot predict which form of env-I.B. or rgp46 wiU be the most effective imrnunogen.
The extensiveness and maturation of glycosylation may be critical to the effectiveness of
the recombinant envelope proteins as immunogens.

223
Carbohydrate moieties of virai giycoproteins appear to play a significant role in
deteminhg thei. ultimate antigenic and immunogenic structure through involvement in

epitope-specfic interactions, overd protein conformation, stenc hinderance, protease

excasibility, and ultimately in antigen presentation. In the past ten years, researchers
have becorne increasingly aware of the criticai involvement of oligosaccharides in

moduiating the immune response to theu attachai antigen.

Carbohydrates have been shown to alter the antigenicity of several virai

glycoproteins including bovine herpesvkus type 1, H[V, Rauscher leukemia virus,

influenza A v k s and tick-bone encephalitis v h s (Aiexander and Elder, 1984;

Guirakhoo et al., 1989; van Dmen Littehan den Hurk et ai., 1990; Doe et al., 1994).
Lndeed, several reports have descn'bed that amigenic epitopes can be revealed or hidden

by the Ioss or acquisition of an ogosaccharide moiety (Muema virus, Smith et al.,

1991; rotavins, Mackow et al., 1988; HIV-1, Davis e~ al., 1990; Ho m al., 1991a;

Cordeii et al., 1991; Benjouad et al., 1992; herpes simplex 1, Muggeridge et al., 1990).
The duai role of carbohydrates in host's immune recognition and response to
glycoproteins, was exemplfied in an immunization study involving the bovine
herpesvirus type 1 (BHV-1) glycopmteins gI and gIV (van Drunen Littel-van den Hurk
et al. , 1990). Deglycosylation of the BHV-1 glycoproteins resulted in an increased

capacity to induce a delayed-type hypersensitivity response in the immunhed rabbits;

however, it aiso redted in the decrpaspd ability to eticit antibodies involved in amibody-

dependent cytotoxicity and Mnis neutralization. This is one of many instances, where

224

carbohydrates have been found to conmLbute to the antigenicity of certain epitopes but
mask other potential recognition sites (BotareIli et al., 1991; Dnimmer et al.. 1993;Doe
et ai. , 1994; Jackson et ai., 1994).

Not only may the absolute presence or absence of an oligosaccharide branch effect
the availability of an epitope or a neighbouring epitope, it appears that the complexity

of the sugars within the branch also influences epitope recognition. Experiments

descriid in this thesis, demonstrated that a lack of oligosaccharide maturation on the

HTLV-1 envelope protein reduced the ability of virus-neutraliwng antibodies to bind. As
descnbed in the dual vaccinia expression system (Arp et al., submitted; chapter 3), the
interruption of trimming and modification of the envelope proteins by Brefeldin A,
resulted in their reduced recognition by several conformationdependent HMAbs,
polyclonal human HTLV-I+I HTLV-II+ sera and polyclonal chirnp STLVm'

sera. This

decreased antibody binding suggested that immature oligosaccharide branches on the

recombinant envelope protein resulted in the masking andfor distortion of conformational
epitopes that are available for ana'body bhdhg on the mature,authentially glycosylated
HTLV-1 envelope protein. Similar results had been observed with recombinant HIV
gp120 expression where immature glycosylation had a marked effect on the folding of
nascent HIV gp120, and ultimately on CD4 and conformationaldepeedent MAb bindhg

(Fenouillet and Gluckman; 1991).

225

Glycosylation also appears to play a role in the overali conformation of various

vual envelope proteins and thus ultimately affkcts thek specific recognition by various
cei receptors and antibodes (murine leukemia virus, Pierotti et al., 1981; bovine

leukemia virus, Bruck et al., 1984; influenza virus, Sugawara et al., 1988; Sendai virus,
Vidal et al., 1989; Hnr-1, Moore et al., 1990; Ho et al., 1991ah; Li et al., 1993). Li

and his coiieagues (1993) elegantiy demonstrateci the involvement of glycosylation in the
folding of a m e n t proteh, to insure its native conformation. In this report,
nonglycosylated f o m of HIV gp120, generated either by deletion of the signai sequence
of HIV-1 gp120 or by synthesis in the presence of tunicamycin, fded to achieve proper

conformation as measured by the ability to bind CD4 receptor (Li et al., 1993). In
contrast, highly niannosylated ml20 b o n d to soluble CD4 molecules weii. Enzymatic
removal of carbohydrate chains from glycosylated gp120 by endoglycosidases had no

effect on the ability of gp120 to bind C M . Thus, this experiment revealed that the
carbohydrate chains on gpl2O were not required for the interaction between gpl2O and
CD4, but that N-linked glycosylation was essential for generation of the proper

coaformation of gp120 to provide a CD4-binding site.

In addition to its effects on overail protein conformation, giycosylation appears

to modulate antigen uptake, processing, presentation and ultimate recognition by T-cells

(Thomas et al., 1990; Botareili et al., 1991; Harding et al., 1991; Ishioka et al., 1992;

Manca et al., 1992; Michaelsson et al., 1994). For example, antigen uptake may be
inHuenceci by the presence of attached glycans, as suggested by the glycosylation-

226

dependent uptake of tW ml20 observed by Manca et al. (1992). In their saidy,

exposure of specific mannose residues on HIV gp120 was required for the exploitation

of receptor-mediated enharred uptake by piresenthg cells and recognition by T-cells

(Manca et al., 1992). Ln light of these results, one may anticipate that the high mannose

structures identifid on the recombinant H'LV-1 envelope proteins produced by the

baculovkus and vaccWT7 polymerase expression systems may target the proteins to
the antigen presenting cells via mannose receptors. The high manwse content of the

recombinant envelope protein preparations may thus ultimately facilitate theu primhg of

class 1 MHC-restricted CTL tesponses (Noguchi et al., 1991).

In addition to its influence on antigen uptake, oligosacchande complexity also

appears to effect antigen processing and presentation (Rademacher, 1992; Varki, 1993;

Manca et al., 1995a). Carohydrate structures have been found to modulate the
accessibility of cleavage sites on the protein backbone, to specific proteases (Klenk,
1980; Soora et al., 1991). For instaoce, carbohydrate-induced changes in the t h e -

dimensional conformation of the protein or direct steric hiderance may result in the

altered accessibility to proteolytic enzymes and thereby mask the T-celi determinant.
Aitemativeiy, altered protein conformation may reveal hidden cleavage sites within the
T-cell detemiiriant resulting in the destruction of the T-cell epitope following exposure

to proteases. Indeed, a peptide pool that is generated following protease degradation of

a glycosylated protein, may be very different to that generated from an unglycosylated
protein (Kwoda et al., 1990; Ishioka et al., 1992; Sjolander et al., 1996). In addition,

227

carbohydrates have been found to interfere in some systems with MHC binding a d o r
with T-ceil recognition, even if they are not involved in denning antigen-specific

recognition (Harding et al., 1991; Ishioka et al., 199; Michaelsson et al., 1994;

Thomas et al., 1940).

Thus the influence of carbohydrate on the uitimate antigenicity and

immunogenicity of viral glycoproteins cannot be denied. Molecular biologists, protein
chemists and immunologists should pay close atfention to the glycosylation pattern of

their respective protein antigeas, particularly, when the antigen will serve as a potetitial

vaccine canidate.

5.2.2 Biologid consequences of native versus denatured conformation

The preservation of native cooformation in the recombinant HTLV-I envelope

proteins preparation may also be crucial for their success as protective immunogens.
Tmmunization studies conducted with other Wal envelope proteins suggests that a subunit
vaccine that effetively presents conformation-dependent neutraiization epitopes may
protect against a broader range of Wal variants than a vaccine that presents exclusively
liwar determinants. Thus one wouid anticipate that the rgp46 produced in the

vaccinidT7 system may confer broader protection aga-

infation with various HTLV-

UHTLV-II and STLV isolates, than env-1.B. produced in the baculovims system.

228

In this thesis, the recombinant gp46 protein forms expressed in the vaccinialT7
polymerase system were shown to be readiiy immunopreipitated by a panel of
conformationdependentmonoclonai antl'bodies (Arp et al., submitted; chapter 3). These
antibodies had been previously shown to be capable of syncytium inhibition (Rowe et al.,
1994). The recognition of the most glycosylated rgp46 forms by these antiibodies suggests

that these protein forms are presenting vinis neutraliziog epitopes and thus may be
capable of stimulating an effective immune response in vivo.

In contrast, the differentiaiiy glycosylated HTLV-1 envelope protein forms

produced by the baculoWus system were not express& on the surface of the insect celis,
but were stored intracellularly as inclusion bodies (Arp et al., 1993; chapter 2). The

envelope protein forms contained within these inclusion bodies appeared to lack native
folding and conformation, as mggestecl by their sequesttation, poor solubility, and failure
to be

recognized by any of the conformationaldependent anti'bodes tested. However,

despite its lack of authentic confomiation, the env-I.B. was shown to be capable of

effectively presenting multiple linear epitopes by binding several envelope peptidespecifc antibody preparations such as lCl 1 MAb ,anti-SP-31-4A peptide sera and 0.5-a

MAb (Matsushita et al., 1986; Paiker et al., 1989; Arp et al., 1993;chapter 2).

Expression of recombinant Wal proteins that maintain the authentic conformation

of their native counterparts appears to be a necessity in current viral vaccine

development. Indeed, native conformation of recombinant protein appears to be essential

229

for the induction of cross-reactive a n h i e s that are capable of binding to native proteins

infected ceii (Haigwood et al., 1990; Rang et ai., 1991;

encountered on the virion &or

Haigwood et al., 1992). In addition, a n t i i e s dyected against coaformationai epitopes

of various Wal envelope glycoproteins, including those of SN-,,

RIV-1, herpesvirus

and cytomegalovirus (CMV)(Banks et al., 1989; Muggeridge er al. , 1990; Haigwood

et al., 1992; Javaherian et al., 1992) have exhibiteci significant abiiities in neutralizing
the infectivity of hese MNses.

native gpllO of SN-,

For example, Iavaherian et al. (1992) observed that only

was able to elicit and absorb signifiant amounts of neutralhing

antibody, while denanired gpllO, large envelope hgments and synthetic peptides could
not. Similarly, Haigwood et al. (1992) observed broad-spectrum neunalizing activity in
baboons with native glycosylated HN-1 rgpl20, which was not achieved with a
d e n a a d , nonglycosylated rgpl20. This inabity of denatured rgpl20 to generate broad

neutralizing anei'body responses confinned earlier findings, that there are no conserved

sequential neutraiization epitopes in a 1 2 0 shared berneen various HIV-1 isolates

(Haigwood et al., 199a, 1992).

The importance of immunogen conformation has also been suggested by the

observationthat antiibodies to discontinuou or conformatiody sensitive epitopes of H[V

gp120 are highiy prevalent in infected human sera (Moore and Ho, 1993). Indeed,
antibodies to b e a r epitopes present on denatured a 1 2 0 account for only a small

proportion (1-10 %) of the total anti-al20 and neutraziiig activity in HIV-1-positive

sera (Haigwood et al., 1990a; Moore and Ho, 1993).

230

In addition to the infiuence of protein conformation on specific B-ceii responses

and the production of biologically effective a n t r i e s , the three-dimensional structure
of an antigen also appears to influence specific T-ceU induction. Recentiy, in

collaboration with Manca et al. (1995a. 199Sb; Appendix LI), it was demonstrated that
T-cells discriminate between native/denatured and intact/fragmented antigen. In this
study, T-cells avaiiable in the n o d human repertoire that respond to a given epitope,

were often unable to recognize the same epitope in a different molecular context (Manca
et al., 1995db; Appendix II). T-ceil discrimination was observed between different

physical forms of both HIV gp120 and HTLV-1 gp63, with the env-I.B. as one of the
target antigens (Manca er al., 1995alb; Appendix II). For example, T-ceil lines generated
against env-1.B. from 5/5 donors did not crossreact with recombinant envelope fragments
or peptides and vice versa (Mancaet al., 199%; Appendix II). Severai other reseachers

have observed similar T-celi discrimiiiation of other protein preparations and have
suggested tbat flanking seyences and carbohydrate chains of the target antigens play an

important role (Lefkovits et al., 1979; Pallcer et al. , 1989b; Botereil et al. , 1991;Manca

et al., 1992; Manca et al., 1994a).

It is likely that T e I l discrimination between various physical fonas of a T-ce11

epitope depends on the differential processing of the target antigen (Marica et al., 1994a,
199%). This suggests that it may be necessary in fiiture immunization regimes to drive

antigen processing mechanisms towards highly efficient presenting cells (Manca et al.,
1994a). For example, varying the route of antigen administration or the use of particular

23 1
adjuvants may resuit in preferetitial antigen uptake by an optimal antigen presenting ceU

(APC).Since Mer-

APCs appear to have different abilities in handling cornplex

antigenic structures (Mana 1995b), driving antigen ont0 an optimal APC rnay ultimately

favour the presentation of relevant epitopes , T - d s and induce a potentiay protective

response. This phenomenon had been recently descn'bed by Manca et al. (199%) in
which a T-celi epitope of HIV could not be recognized by specific T-ceiis when
presented in the context of whole virions by monocytes as APCs; however, B-cells were
capable of effectvely presenting the epitope. Thus altering the deiivery of an
immunogen, may ultimately result in the exploitation of T-ceils that would be othemise

In ght of these recent observations, the env-I.B. preparation may still prove to
be an effective immuwgen as a protein boost, despite or because of its deuatured and
particdate nature. Sequestration of the recombinant HTLV-1 envelope protein as
inclusion bodies in the insect celis rnay result in its efficient phagocytosis and
presentation by MHC class 1-bearing CD8+ T-cells, in addition to the MHC class IIbearing CD4+ T-ceils(Gooding and EdwitCdS, 1980; Bevan 1976; Rock et al., 1993;

Pfeifer et al., 1993; Zhou et al., 1993). In contmst, it would be anticipated that the
soluble exogenous rgp46, producd in the dual vaccinia system, may be less efficiently
processed and presented by the class 1MHC pathway than by the MHC class II pathway,
as observed with other soluble virai antigens (Pfeifer et al., 1993). In this case, the
presentation of soluble antigen would be restricted to a limited spectrum of cells that

232
include B lymphocytes, macrophages and dendritic ceUs (Men et al., 1987). If this
redts, the expression and storage of the HTLV-1 gp63 envelope protein in the insect
celis as insoluble inclusion bodies, may

CO&

an Mmueogenic advantage over the

soluble rgp46 produced by the mPmmalian system.

5.2.3 Rat challenge mode1 avaable for testiiig of pdentiai vaccine candidates

The establishment of the HTLV-1 infection model in ad& and neonatal Fischer

F344 rats, as described in chapter 4, will enable the fiiture evaluation of the dflerent
envelope protein antigens for their immunogenic and protective properties. In addition,
the optimal challenge dose was determined in this rat model to iosure that the HTLV-1

infection mimicked one that would be encountered in a natural environment and prevent
overwhelming of the immune system.

In future vaccine protection trials, independent challenges of Fischer rats with the
determineci optimal high and low doses of 10' and 106 HTLV-I infected ceis may mimic
different clinical situations encountered in humans and may aliow more effective and
realistic t e s ~ of
g the vaccines' protective potential. The use of both challenge doses may
serve as a mode1 for human contact with body fluids fiom high viremic versus low
viremic individuals, or as experienced in napsfusion settings. Thus, the variety of rat
infection protocols described in this thesis rnay eventually provide insights into whether

233

stimulation of different immune response cascades wiU be required to prevent or control

HTLV-1 M ' o n foilowing encounter with v-g

initial inoculums, or may be

necessary at different maturatio~liilstages of the recepien. In addition, the fact that the
Fischer F344 rats are an inbred strain, will enable their use in adoptive transfer studies;

a necessity for detennining the correlates of protection inued by the HTLV-1 envelope
protein vaccine candidates. Thus it appears that the curent establishment of the Fischer
rat challenge mode1 may uitimately conmbute to our undersfanding of RTLV-1 infection

and disease progression, and aid in the development of an effective vaccine.

234
5.3 Goais of vaccination: Generation of anti-viral antifbody andlor CTL

Despite extensive researb into the development of an HTLV-1 vaccine, no

consensus has been yet reached to cl-

which components of the immune system are

efiectively involved in conferring protection against this Wus infection. Several

laboratories have documentai that protection from retroviral chaienge can be mediated
primarily by CD4+ and CDSf CTL responses, in the presence of very low or

detectable Vinis-specinc neutralizing a n t i i y activity at the t h e of challenge (Clark

et al. , 1991; Earl e? al. , 1986; Miyazawa et al., 1992; Issel et al. , 1992). On the other

hand, antibody has beeo acknowledged for its pivotai role in acheving protective

irnmunity to HIV observed in vaccinated animals, devoid of CTL activity (Fula et al.,
1992; Girard et al., 1995). Following an extensive review of various virus systems (see

chapter l), the data suggests that several participants involved in both the cell-mediated
and humoral amis of the immune response, play a synergistic role in protection against

virai infections, including HTLV-1. This synergism was exemplified in the LCMV
model, where CTLs were found to reduce viral titres by 4-5 logs (Byme and Oldstow,
1984); whe anti'bodies were also found to play a role in protection by reducing the

LCMV titres by 1-2 logs (Baldridge and Buchmeier,

lm).Sidarly, the murine mode1

of CMV and human C M 'suggested wt only the effectiveness of C m , but also

demonsaated a role for NK ceUs and antibody (Biron et al., 1989; Koszinowski et al.,
1990). Indeed, there appears to be no case where specific antibody , if present in

sufficient concentrations in vivo, does not at l e s t substantiay limit viral infection

235

(Hayder and Blanden, 1994). The extent to which antiibodies, C M +and CD8' T-ceUs

conmbute to resolving HTLV-1 infctioncan only be detenniaed through m e r research

and wiU inevitably aid the development of an effective vaccine.

In the future, once protection against HTLV-1-infcted ceii challenge has been
attained with specific immunkation protocols, adoptive traosfer experiments should be

carrieci out. Adoptive m e r experiments wiU help define the conrn%utionsof cellular
and humoral immunity to the protective response. The source of the immune cells and
semm should be obtained h m animals immunized with the most effective protocols.
Aliquots of these cellular and humoral pools should be taken and extensively tested in
vitro, with the remauider given to naive MHC-matched animals who wl subsequently
be challenged with HTLV-1. In vitro, Factionated PBLs can be tested for HTLV-1

antigen-specific T-ce11 proliferation and cytotoxk T-ceii activity; while senun can be
andyzed for ADCC, cornplement-mediateci lysis, isotype content and overaii anti-HTLV1antibody titres. Conducting these immunological assays wl help determine the immune

correlates of protection. Identifying what immune responses are most effective, may
stimulate attempts at M e r enhancing these protective responses. Alterations in the
immunization protocols, including parameters such as adjuvants, injection sites and

number of injections, may m e r improve the vaccination regime.

5.4 Goais of vaaination: Target populations

nce an effective vaccine candidate has been developed, the next important step

is to identify target populations that are at high risk of HTLV-1 infection. As aiready

mentioned in the introductory chapter, major endemic foci of HTLV-1 iofection exist in
ethnic subpopulations of southern lapan, intertropical Afnca, the Middle East, the
Seychelles Islands, South America, Can'bbean basin and coastai western Canada (Blattner
et al. , 1982; Hinuma et al. , 1982; Arango et al. , 1988; Kazedi-Kayembe et al., 1990;
Meytes 1990; Lavanchy et al., 1991; Dekaban et al., 1994). Epidemiological studies
have helped define several sub-populations within these endemic areas that would benefit

from vaccination against HTLV-1. These include HTLV-1 infectai mothers for protection
of their breastfed babies, the wives of infected men, prostitutes and intravenous h g

abusers.

Mothers exhibithg high titres of anti-HTLV-1 a o t i i e s (Hino et al., 1986),

detectable antigen in theu PBLs (Sugiyama et al. , l986), or detectable HTLV-1 positive
cells in the milk (Kinoshita et al., 1987) are more at risk of infecting their infants. Up
to 25% of chikiren bom to HTLV-1 infkcted mothers wiii also become infecteci (Hino,
1990). Asking these select mothers to refrain nom breast-feeding is rather simple and
safe in developed countries; however, this intervention canwt be suggested in developing

countries where contaminateddrinking water may cause infantile diarrhea and appropnate

milk concentrations of baby formula is ofien not rnaintained because of economical and

237

educationai problems. Rehhing fiom breast-feeding in these areas would put these
individuais at a high risk of developing other serious dwases which becorne manifest
early in life, uniike ATL and HAMfTSP which tend to be diseases of aduits.

The collective findings of several breastfkding intemention studies in lapan,

implicate that maternai a n h i e s may a c w y be protective against milk-borne
trammission of HTLV-1 in the early months of life (Ifino et al., 1990; Takahashi et al.,
1991). If breastfeeding is restricted to the first 6 months of M e the probability of HTLV1 infection is similar to that of bottie-fed children; however the probability rises an

additionai 10%in children b&ed

for 14 months or longer (Takahashi et ai., 1991).

In addition, a decreased infection rate was observeci in paaially breast-fed children

compared to those who were not breast-fed at aU (Hino, 1990). These results suggest that
active immunization of the mother, to sustain high titres of HTLV-1 neutralizing

antibodies throughout a fidi period of lactation should prevent milk-borne transmission.

To prevent &-borne

transmission, it would be most effective to vaccinate the mother,

rather than the infant, since the human immune system does not becorne fully mature

untii approximately 6 months of age; by which time the child may have already been

infected with HnV-1.

To M t semai ansmission of HTLV-1 in endemic areas, pre-pubescent females,

wives of infected men and prostitutes would be prime candidates for vaccination. Females
are of pa.rticuIar concem since sexuai transmission is predominantly fkom man to wornan

238

(Kajiyama et al. , 1986; Murphy et al., 1989). The vaccination of pre-pubescent females,
prior to becoming sexual active, wouid hopefully control HTLV-1 by preventing both
sexuai transmission and fiequent coincidence of rnother-to-chd transmission via

In developed industriai countries, the greatest threat of epidemic spread of HTLVI is among intravenous dmg abusers (Manns and Blattner, 1991; Blattner and Gallo,
1994). Thus it would be desirable to vaccinate this group despite the obvious practical

dmculties. Serious consideration should also be given to the vaccination of health

professionals in reguLar contact with HTLV-1 positive patients. Blood is a potent vector
for transmission of HTLV-1. For instance in rabbits, HTL,V-1 c m be pas& with as little

as 10 microlitres of inf'ected blood, whiie in humans, a needle-stick has resulted in the

infection of a physician (Kataoka et al., 1990).

Not only wiU an HTLV-1 vaccine control the virus's prevaience, it may help in
Limiting the disease pathologies associateci with it. Interventions to reduce infection in

early life, specificdy mother-tochild aansmission, will greatly reduce the incidence of
T-ce11 non-Hodgkin's lymphoma, including ATL, that have been strongly associated with
HTLV-1 childhood iafections in endernic populations (Cleghorn et al., 1995). Vaccination
of HIV-infect4 individuais may also be advantageous, to prevent the dual infection of
HIVIHTLV-1 which may d

t in an exacerbation of progression to acquired immune

deficiency syndrome (AIDS; Barthoiomew et al., 1987). In addition, vaccination of

239

asymptomatic HTLV-1-infkcted individuai identifiai as "ATL-proeewindividuals, may

sufficientiy stimulate the host's immune response against proiifierating HTLV-1 infecteci

celis and thus decrease the chance of progression to ATL. However, caution must be
taken to not vaccinate " W T S P "prone HTLV-1-infeted people and shouid be checked

by MHC haplotyping (Sonoda et al., 1994) and/or serological analysis (Dekaban et ai.,
1994). 1ncreasi.gspecific-HTLV-1 responses in these individuals couid possibly lead to

aggravated destruction of the spinal cord and ultimately increase the seventy of the

disease, as has been suggested by Jacobson et al. (1991).

5.5 Goals of vaccination: Revention of infection and/or sease

Since HTLV-1 and H I ' were first isolated, scientists bave been seafching for

specific vaccines that couid completely prevent infection. Because both are retrovinises.

vaccine researchers feared that if a single virus particle Mected a ceii, integration would

follow and full-blown disease would eventuaily ensw. The only safeguard was to create

vaccines that prevented infection completely, producing "sterilizing immunity " .However
there has been a shift in the philosophy of retroviral vaccine researchen since "me
sterilizing irnmunity" appears to be realistidy impossible. True "immunity" relies on

a Ievel of infection to mount a complete and effective immune response. Even if

immunization would be ultimately capable of inducing prolonged hyper-immunity to

HTLV-1, it would be realistidy impossible to prevent the infection of at least one CD4'

T-ceil during perods of high-risk interactions. Physicai bamers, such as leak-proof

condoms, seem to be the only mode of achieving tme protection from infection "sterilizing immuaity" . Indeed every virai vaccine to date protects w t against infection,
but against disease (Oldstone, 1993).

Vaccine researchers have been redefining protection since they began observing

that the immune system has a capacity to efficiently contain infection with HN (Cohen,
1993). If the immune system has this contaimnent ab-,

then a vaccine that fails to

prevent infection, may still be successfbl in delayhg or preventing progression to

disease. h the retrovirus vaccine field, "protection" may be soon redefmed as a reduced

241

viral load. A reduction in virai load may not protect aii individuals a g a . disease.

However, it may reduce the fhqpency of vinas transmission, decrease the incidence of
occurrence, and if not capable of preventing disease, it may at least slow its progression.
In terms of an HTl,V-1 vaccine, geoeration of a more effective virai-specific immune
response may prevent the successive rounds of nIZV-1 infectecl ceU proLiferabon
obsemed in asymptomatics, and thus reduce the chances of secondary transformation
events resuiting ultimately in ATL (Arima et al., 1986; Sanada, 1986). In addition, a
significant reduction in the viral load in "HAM/TSP-pronewindividuais may reduce the

pathogenic immune responses observed in the full-blown oeurological disease state (Hara
et al., 1994; Oger and Dekaban, 1995).

The development of the two HTLV-1 envelope protein expression systems

descnbed in this thesis, wili hopefdiy contribute to a greater understanding of the

immune response to this protein and HTLV-1 infection. Results presented in this thesis
suggest tbat these proteins may serve as competent diagnostic reagents. Whether they wl
contribute directly or indirectly to the evennial protection of the world population a g h t
HTLV-1 has yet to be detemiined. However, such a goal is worth striving for since a
vaccine would protect the human population agaiast a nilminant leukemia, an
incapacitating neurodegenerative disease and general morbidity. In addition, its success

may assist in fmding solutions for problems highly relevant to the design of an AIDS
vaccine.

Comparative analysis of the anti'body response to the E1TLV-1 gag and e m proteins in
HTLV-L asymptomatic carriers and HAM/TSP patients: an isotype and subclass

anaiysis

Dekaban, G - , King, E.,Am. J., Palker, T.,and Rice, G.

SC&.

J. I ~ u r r y o l40,
. 171-180 (1994).

T-ht. [ohn P.

Robarts Research
f rist itute

Robarts Research Institue

P.O. Bon 5015
100 Perth Drive

243

London, Ontario, canada

N6A SKS
Phone: 519-663-5777 (en. 4241)
Fik: 519-663-3789

May 6, 19%

Scandiuavian Journal of Immunology

Blackwell Science Limited,
Osney Mead,
Oxford, 0x2 OEL
United Kingdom
Dear SirIMadame,
I am a graduate student completing my doctoral thesis at the University of Western Ontario,
London, Ontario, Canada. I hope to acquire your permission to include within my thesis a
complete paper that has been recently pu6lished in your journal. The entire paper will be
included as an appeedix in my dissertation. 1 am a CO-authoron this paper: Dekaban, GOA.,
King, EoE.9 Am. J., Paiker, T.J., and G o P e A o Rice. Comparative mdy& of the
antibody response to the BTLV-1 gag and env proteins in HTLV-L asgmptomatic
&ers and HAM/TSP patients: an isotype and subciass anaiysio. Scmd. J. Inniunof.
40: 171-180(l994).
1 have pemnaiiy asked my laboratory coiiegues Dr. Greg Dekaban and Elaine King
for their permission, of which they have verbaUy granteci; however the Faculty of Graduate
Studies at the University of Western Ontario, Canada, requirrs that 1 receive official written
permission from your journal, b e h my thesis can be accepred for fulfiiment of my degree.
Your expedient reply would be greatly appreciated. Thank you for your time and
consideration.

JacqueLine Arp

Comparative Analysis of the Antibody Response to the HTLV-1

gag and env Proteins in HTLV-1 Asymptomatic Carriers and
HAMJTSPPatients: an Isotype and Subclass Analysis

Dr Gregory A. Dekabm, fmmmdogy Graup, The John P. Robmts h a r c h Iititute, P.0- Box 5OlJ.IOO
Perth Drive, London, Ontario, Cana& N6A 5K8

INTRODUCTION
H u m a n T cc11 lymphotropic virus type 1 (HTLV-I) was the
first human tetrovinu to be diseavcnd and associated with a
pathological conditionh o w n as adult T e l l Icukamia [I, 2).
Later HTLV-I was found to k -at&
with a second
coadition known as m V - 1associatcd mycIopathy/tropical
spastic paraparesis [3-5j. HAMJrSP has bctn described in a
number of gcographic locations where HTLV-1 is endcrnic fq.
The immune system of patients infccted with HTLV-1
appears to be dysnguiated, especially in those with HAM/
TSP [7],T cells obtaincd from HTLV-1 infted individuals
with HAMrSP appear to spontancously prolifcrate when
culturccf in vitro [8, 91. Patients with HAMflSP exhibit an
increascd level of HTLV-I ccpIication [IO] and have a greatcr
numbet of' infectcd cclls in thcir periphetal btood than

asymptomatic iodividiiols [Il). It is likely that the dysregul a d immune system obzennd in HAM/SP patients is due
to the action of the HTLV-1 T u protcin [I2, 131. The Tax
protein not ody tmrmctivata v i d geae transcription but
aIso transactivata tnnscription from a number of allular
genes involvcd in m t b g the immune system including the
genes for IL-& L 2 teccptot, IL4 and TNF (71. It appears
that Tax can be saxeteci and takca up by uninfectcd Iyrnphocytes which rnay bt subsequently indu& to prolifcratc
[l3- 151. The Tax protein appears to be su6cicnt for the
induction OC prolifetation and immortaiization of T celis
through a paracrUie/autocrinc mechanian 116, 17). Since the
regulation of antibody productioa is controllcd by T uIIs,
any disturbance in ttit mechanisms that reguiate T ctIls rnay
end up dysngulating the conuol of antibody production.
Although some aspects of the antibody rcsponse ro

17 2

G.A .

Dekaban et al-

HTLV-I proteins have b e n characterized in HAMJTSP

patients [a, 18, 191, an analysk of the immunoglobub (Ig)
isotypes and IgG subclass profies of HTLV-1 ami-gag @ore
proteins) and anti-envelopeantibodies in HAhfflSP patients
may shed some light on the role of HIZV-1 in the immunopathogenesis of HAMPSI? since different antrlbody subciasses and isotypes rnay have dinerent protective and
immunopathogenic propcrties [20-26jFurthemore, there is no information available to determine if a given asymptomatic individual is progrtsshg
towards the devebpment of HAM/TSP. An -nation
of
the isotype and IgG subciass p r o f i may be useM in this
regard. Early diagnosis of HAMflSP may be critical to
successful treatment of the discase since suggestions exist
that early marnent with stcroids may preveat further
disease deveIopment [271.
in this study we examiad the Ig isotype and IgG subclass
profile of antibodies against the HTLV-1 gag pl9 and p24
proteins and the enveiope proteins gp46 and gp21 in HAM/
TSP patients and in asymptomatic HTLV-I infctd individuals. In each case, the Ig &type and IgG subclass responses
were uuated for the gag p 19 and p24 and envelope proteins
using Western blot assays. Envdopc anti-peptide responses
were also titrated by ELISA A cornparison of the neirtralipng
antibody titres betwem HAM/TSP and asymptomaticpatients
was aiso conducted-

MATERIALS AND METHODS

Subjecu. Eight HAMfSP patients and one HTLV-I infeacd
asymptomatic individuai w m m a t immigrants to Canada from
ChiIe. Brazil. and the Larn'bean, Two HTLV-I asymptomatic
subjets were Canadians with no obvious rirk factors- Senim from
two HAM/TSP and th- asymptomatic HTLV-I infCcted British
Coiumbian aboriginal Indians provideci by Dr D.Wcrkcr (Depanment of Health Carc and Epidemiology, C, Vancouver. BC,
Canada) and Dr I. Oger (Department of Ncurology, University
Hospital, UBC site. Vancouver. BC, Canada). Twenty-aine HZV-I
positive lapanese asymptornaac scrum sampies w m providcd by
Dr S. Hino (Nagasaki Univcmty Schoat of Medicine, Nagasaki,
Japan). Controf sera w m obtaincd from normai individuais and
from thox with other ncuroiogical diJcases. -411of the abovc samples
were obtained using accepteci consrnt proccduru.
HTL V--Ip r o t e k The entire HTLV-1 cavcIopc glycoprotein was
produced by a recombinant bacuiovirus cxp-on
systcm and was
purifid as previously describeci (281. The envdope from the baculovirus expression system is p r c d ~ ~ n a n t in
i y tht prccunor foren and
appears in thne molecuIar wghts (gp63. p54 and p43) which
represcnt differcntly p r o d versions of cnvelop prtcursor. The
HTLV-1 gag protcins were obtaincd from sucrose banded HTLV-1
from CIOMJZ celis as provideci by Dr Larry Arthur (AIDS Vaccine
Prognm. NCI-FCRDC, Program Resourcs Inc./DynCorp.,
Frederick. MD. USA). Sincc the core particles werc uscd as the
source of the gag proteins. WC wcrt able to look at the antibody
responses to pl9 and p24. Therc wcre alsa diffemt p m m r fonns
of the gag proteins but ~rormctivityto these proteins was not
i ncluded in tlus analysis,

Wesrern blot mays- The envclope glycoprotetn western biots

as prevousiy dcxn'bed [28]. The gag western blots
were gentratcd with 125pg of the gag protein per blot that was
cesuspendcd in 624 of Lannlli buffer, eiectrophoresed on a 10%
SDS-poiyacry-de
gel and transftrrrd to Irnmobilon P membrane
(Milliporr, Bedford, MA, USA) using the tram-bIor semi dry
transfer ceIl (BioRad, Mississaum Ontario, Canada). Envelope
and gag blots wcre biocktd in a 5% sian milk powder solution
[Bi.
Biots werc incubated in primary sera dilutcd in blotto using the
minibtotter 2 (Immunetics, Cambridge, MA, USA) for 4 h at room
tanperam or ovcrnight at 4OC with rocking The blou were washed
in biotto and exposai to homdish proxidase conjugatd secondary aatibodics diluted in blotto as follows: Ifl0.000 for goat Fabi
isotypf-spcciic ami-human IgG, IgM, IgE, or IgA (heavy chain
specific; Cappcl, Organon Teknika Durbarn, NC, USA); 11500 for
punficd mouse monocIonal a n h i e s (MoAb) for spccific antihuman IgG ntbtype (puri6ied mouse MoAb, IgGI. 3 and 4.
Zymed San Francisco. CA, USA; 1%- I.C.N.. Costa Mesa. CA.
USA). The blois wcrt exposed to the appropriate secondary antibody foc 60min, wasbtd and processed using the enhanced chemilurnintsccnt deteaion syst#n (ECJ; Amcrsham, Oakviiie, Ontario.
Canada) aacording CO the manufacturrrs spccifi.cations. Development of ECL blots involveci a bricfwash in nibsuate and utposure ta
X-ray film for 2s to 2min. tbe titre was determincd as the last
dilution of primary serum that exhibitcd a signai foliowing a 2 min
wrposurr.
Peptrile ELISA. The HTLV-1 syntctic peptide binding assays
were pcrfonned by ELISA as descri'bed by Arp er al, [281 at 2 pg of
peptide per w d with the exception of the xcondary antibody
conjugate and deteaion systcm- 'Ine peptides choscn for these
assays arc as follows: SP2 @p#, amino aads 86-167). SP4A
(gp46, amino acids 190-209). SP6 (gp46, amino an& 296-312)
and SP7 b 2 1 , amino acids 374-392)- Tticsc peptides have been
previously dcscribed PO]- Briefiy, aftcr ovcrnght primary antib&y
incubation and wash, the plates wcrc expoxd to horseradiih
peroxidase conjugabd saondary anu'body diIuted in reaction
b u f k (containhg 2% skim dk). (as dacribc in the western blot
assay) for 60 min. 'fbe EUSA plates were developed using the TM B
Peroxidase Substrate Systtm (Mandel, Guelph, Ontario, Canada)
and the miction was stoppcd using 1M phosphoric acid. End point
titre was defincd as the highcstdilution at whic the optical dcnsity at
450 nanometers was grrater than
the ODIM of a 1/50 diluted
pool of normal human scrum (this pool king the rcsult of an equal
mix of six normal scnim sainples. The MOLECUUR DMCES s o m u
plate rcader and software systcm wtrc uscd to detcnnine the end
pointS y n c y t h hhibilioan assay, Ntuuafizjng antibody titre was detetmincd on codcd senun samples as pmvious~ydescribed [28]. Bnefly.
H'IZV-1 producing C91/PL Tells wcre incubated with C8166 Tcelis (at 10~cell/rn1
for cach e l l fine in RPMI 1640 with 10% FBS)
overnight in the prrxnce of 10pl of sm'dly dilutcd test serum (heat
inactivatcd at 56C for 30min). Aftcr 24h, the prrscnce of syncytia
was evaluatcd in an invmcd microscope at 200-fold rnagnification.
Neutralizing titre was the hightst scnim dilution which inhibited
HTLV-1 syncytium fonatiaa by 90% or mort. The code of al1
sarnples was broktn subscquent to neutmiking titre determination.
Approximateiy 150 qmytia were countcd per wcD usine the normal
human serum controls.
Srarisricai anaIysis. fht tesuIU of the Western blot titrarions.
peptide ELISA and ncutralization assays for the WLV-1 infected
were prepared

asymptomauc group and the HAMFSP group wecc compared by

either the two-tailed Fisher's Exact Test or the Maan-Whitney
I;-Test with P < 0.05 bang accepted for statisucai signiscance.

RESULTS
lsorype responses ro gag and env pro reins

Western blots were used to examine the tg isotype responscs

specific for the HTLV-1 gag proteins p24 and p19 and the
envelope protein of the HAM/TSP patients in cornparison ta
the HTLV-I infcted aqmptomatic individuals. in addition,
the western biot assay was uscd to sirnultaneously titrate the
isotype responses for the p19, p24 and envelope proteins,
Representauve examples of the Western bIot titrations are

Fig. 2. Percent rractivity of the asymptornatic and HAM;TSP

sera to the HTLV-1 gag pl9 and p24 proteins and the envelope
protcin for each of the Ig isotypes and the IgG subtypes as
determined by the Western blot assay. The Ievel of sipificance
berwccn the two groups is given below each Ig isotype or IgG
subclass as determincd by the two-tailed F ~ h e r sExact Test.

H1ZV-1asvmotomatic individuafs. F! HAMfT3P aatienrs.

Fip. 1. Western blot titration. An cxample of the Western blot

'ItrJtions for
IgG i s O t ~ ihe (A)
gag pl9 and pZ4
and the (B) envelope protein for representative HLV-1positive
asymptornatic individuals and HAMflSP patients. - Indicates the
negative semm conrrof composed of a mixture of six HTLV-1
negarive serum: + indicares a known HAMiTSP positive semm
conrrol at a 1 : 100 dilution both of which werc inciudcd on each
Western blet.

shown in Fig. 1. The results of the Ig isotype anaIysis are

shown in Figs 2 and 3.
The IgG response was the only one in which al1 the
asymptomatics and the HAM/TSP patients reacred to the
gag p19 and p24 and envelope proteins. For the other
isotmes, there was a marked distinction between the two
groups. In the asymptomatic group, igM o r IgE antibody to
the HTLV-I proteins examined was not detected, but 20% of
aqnptomatics were positive for IgA to the pl9 and p24
proteins and 60% for the envelope proteins. These results
contrasted sharply to those of the HAM,TSP group: a11 of

174

G-A. Dekaban er ai.

p 19 tg ISOTYPES

wbom had an IgM respoase to both gag and envelope pro teins
and an IgA respoase of 100%, 100% and 80% to pI9.
envelope and p24 respcctivdy. One hwidred per cent of the
HAMmP patients were IgE positive for envelope but oniy
30% to 40% were positive for the two gag proteins- Thus, the
HAM/SP patients wcre positive for al1 isotypes to the
envelope protein but were differentially positive to the gag
proteins
When the titres wctt dctcrmined for the four Ig isotlpes
to the gag and cnvclopc protcins, M e r differcaces between
the asymptomatics and the HAMJTSP patients became
evident (Fig-3). The titres for the HAM/TSP group were
significantiy higher than the titrts for asymptomatic group
for al1 the Ig isotypc responses to the gag pi9 and p24 and
to the cnvclopc prottios- Diffcrcnccs wert also observed
bctwcen the isotype responscs directcd to tbe gag proteins
and envelope proteins. he ennlopt rtsponses were higher
than for cithcr gag pl9 and p24 in both the asymptomatic
and the HAM/TSP group for ail fou Ig isotypes, but were
cspecially devatcd in the IgG respoases-

lyOOOyOOO

1-

-- ASM H A ASM H/T ASM U/T ASM H/T

IgM
IqG
1gA
IgE

10,000

t 000

- --1 I

100 -

:I;,1 Tl;

-1

ASM H/T ASM H/T ASM H/T ASM H/T

IgM
IgG
IgA
IgE

ENV tg ISOTYPES

Using the same western blot assay, an analysis of the IgG

subclass rrspooscs was aaalysad with rtspcct to the HTLV-1
p19, pz4 and the envelope protciii. The r d t s are shown in
FSgs 2 and 4. The subclass rcsponse was mtricted to the IgGl
and IgG3 with no rcsponscs observai for the I g G and I f i
subclass for aii thra of the HTLV-1 proteins examined- h e
titres of the pl9 and p24 1 6 , and IgG respoases were
significantly d8erent bctwccn the asymptomatics and HAM/
TSP patients mg. 4). The IgG, and.IgG3subclass responses
to the envelope protcin wert also significantiy higher for the
HAMITSP patients than for the -asymptomatics. The magnitude ofthe IgGI and IgG3 responses was p a t e r with respect
to the anti-envclope responsc than for the gag p l 9 and p24
proteins in the HAMITSP group. This was not observed in
the asymptomatic group whcre the ami-envelope response
was actually lower than the anti-gag pl9 and p24 responses.
fg &type and IgG subclass anaiysit of envelope anti-peptide
responses

The envelope Ig isotype and IgG subclass responses were

-

- -

Fig- 3- Scrum titres of the Ig isotype rtsponscs to the HTLV-1 gag

p 19 and p24 proteins and envdope protcin as determincd by the
titration in the Western blot assay. The geometric mcan and

AsM H/T ASM H/T ASM H/f

IQM
I ~ G
tg4

ASM H/f
IgE

standard deviarion is shown bcside each data set. The associateci P

values for the IeveI of significancc bctwecn the two groups are
as follows: For pl9 and IgM P < 0.001; IgG P < 0.001; IgA
P c 0.001; IgE P < 0.005. For p24 IgM P < 0.001: IgG P c 0.001;
IgA P < 0.00 1; IgE P < 0.01. For cnveIope I g M P < 0.00 1; IgG
P < 0.001; IgA P < 0.002; IgE P < 0.01. The Ordinate is on a log
scale.

248
HTL Y-IAntibody Isotype rinafIvsk 175

ENV fgG SUBCLASSES

exami-ned in more detail by Iooking at four different regions

of the envelope encornpassed by the synthetic peptides SP.
SP4A and SP6 from gp46, and SP7 from gp2I (Figs 5 and 6).
With the exception of SP, the majority of the HAMflSP
patients had an IgG q n s e to the other three peptides.
wbiie the majority of the asymptomatic individuals were
positive o d y to SP4A and SP6. An IgA response was
observeci for a majority of the HAM/TSP patients to SP4A
and SP6 but was not obscrvcd in the asymptomatic
group. Ncithcr tEie SP2 or SP7 peptides detected appreciable
IgA rcsponses- IgM and IgE rcsponscs to the four
peptides werc oniy o b x m d in a srnidi proportion of
individuals in both gmups. For those individuals who were
positive for a partcufar iso~ypc,the HAM/TSP group had
signifiwntIy highcr ELISA titres than the asymptornatic
group for alt the peptides exccpt SPZ, for which there was
no significant d i n i c e c e
The IgG subclass anaiysis using envelope peptides corroboratcd the envelop Western blot analysis in that the IgG
rcsponse was stiU restricttd to an IgG, and IgG3 response.
However, some variation in the mtent of the IgG, and IgG3
rcsponses was obscrved bctw#n the individuai peptides. Oniy
a smaii proportion of the HGMfTSP and asymptornatic
groups mipondad to the SP2 peptide and those that did
respondcd maialy in an IgGl restrictcd fashion. Wheu the
titres wcrt e x a m i . the lgGl response to the SP4A. SP6 and
SP7 peptide was cIcarIy higher than the IgG3 responses
observed for both the H
A
M
m and asymptomatic groups.
Due to the inrreascd scnsitivity of the ELISA technique over
that of the Wcstern blot assay, IgG rcsponses to each of the
four peptides was also observeci in a d l proportion of both
groups tes& An IgG, subclass response was not observed in
the HAM/TSP or the asymptomatic group for any of the
peptides. With the exception of the anti-SP4A IgG? response.
the anti-peptide titres of the HAM/TSP group were significantly higher in those who wcre antibody positive for an IgG
subclass with respect to the SP4A, SP6 and SP7 peptides than
those who were antibody positive in the asymptomatic group.

H TLV-I neutraiking antibody

Fig. 4. Semm titres of the IgG subclasses to the fITLV-1 gag pI9
and p24 proteins and the envelope protein as determineci by
ritration in the Western bIot assay. The geomeuic mean and
standard deviarion is shown beside each data set, The associatcd P
values for the levei of n'gnificance betwecn the two groups art as
fottows: For pl9 lgG,, P < 0.03; IgG2 P = NS (not gnifimnt);
IgG, P < 0.03; IgG4 P = NS. For p24 IgGI. P < 0.004; IgG2
P = NS; IgGS P <= 0.03; IgG4 P = NS. For cnvelope IgGl
P < 0.001; IgG2 P = NS: IgG3 P < 0.002 IgG, P = NS. The
Ordinate is on a log scale.

The HAM/TSP and asymptomatic groups were examined for

the presence of neuualia'ag antibody as defined by the
syncytium inhibition assay. Al1 of the HAMKSP patients
tested positive for neutralizing antibody, while less than 90%
of the asyrnptomatic individuaIs were positive in the assa!.
The neutralizhg aaiibody titres from the HAMiTSP patients
(422 2.8) were significantly higher (P = 0.0005) than those
of the asymptomatic group.
DISCUSSION

The most stnking finding of the results obtained was the uniform
appearance of the HTLV-1 specific IgM isotypes in 100% of the
HAMflSP patients. In addition. 40% to 100% of the HAMTTSP

SP4A ELISA ISOIYPES

l,000,000

.
L

ASM H/T

1 gM

-i

A "
I

ASM n / t ASM H/T ASM n / t

196
IgA
ig

1:

Un n h
IgM

ASM H ~ TA ~ M
H/T A ~ M
ni1

190

1gA

IgE

Fig. 5. ELISA titres for tk rmim Ig ktype ~ ~ p o a r to

c stbe HTLV-I m n l o p synthec peptides. h e gcometric man and standard
deviation WC s h o w h i d e each data ut.The associaicd P values for the Ievel of significatlcc ktwecn the two goups is shom as follows:
For SP2 al1 isorypcr P = NS. For SP4A IgM P =
IgG P = 0.002; IgA P < 0.001; IgE P = NS. For Sp6 IgM P = 0.04; IgG
P < 0.001: IgA P = 0.00 1; I# P = 0.5. For SW IgM P = 0.02: IgG P c 0. 1; IgA P c 0.00 1; IgE P = 0.05. The Ordinate is on a log
scale.

patients were positive for the IgA and IgE Wtypes depending on
the HTiV-1 protein testcd. This was significantly different from
the IgM, IgA and IgE responses obscMd in the HTLV-1
infected asymptornatic gmup- The inucarcd prcsence of IgM
and I g E antibadies in the HAMflSP patients suggests the
presen of a persistent immune response to HTLV-1 in the
HAMfSP patients. hm could k severai factors contributhg
to this phenornenon. Severai reports have doeumented grtater
numbm of HTLV-1 iiircned aiis in the peripheral bloai and
increased viral replication in HAMlfSP patients (10, 11, 31).
This could lead to chmnic immune stimulation as suggested by

the presence of IgE a n t i i y in nearly aii of the HAM/TSP

patients. IgG, whih has also been aPsaated with chronic
"mune stimulation [26) was sot obsrvcd. In addition. the Ig
isotype and IgG mbclass respolisec are known to be regulated
by T ails [32,33]. The fa* that HTLV-I can dysregulate the T
celis to proiiferak abnormally and induce the expression of
severai different lymphokines (which arc themselves capable of
"Ruachg Ig Wtype and IgG subclass rrsponsa [33] and the
major histocompatibility cmaplcr (MHC) upressjon [M.351)
may determine what type of antibody response will be generated. An ina*rre in the mlease of snral T all lymphocyte

HTL V-I Anfibo& I s o ~ p eAnuIysir

ASM H/T 4SM u/T ASM H/T

IgGl
lgG2
IgG3

l00,000'

10,000

1000

100-

ASM H/T
1gG4

l,000,000 i

100,000 -

s
=

rO,OOO-

-CI

. .

iooo-

. .

i=

:T

~OO-

'

10

1-

177

SP7 EUSA IgG SUBCLASSES

SP6 EUSA IgG SUECLASSES

1,000,000

l-

250

IO-

Y
1,

-= - -- --

&Ils
I

ASM H/T CLSMH/T ASM H/T ASM H/T

IgGI
1962
1963
1964

A;M

n i t ASMlgG2
n i r ASM n j
IgGi
igG3

ASM
~

nir

'

IgG4

ELISA titres for the scrum IgG subclass rcsponnr to che M L V - I rynthctic peptides. Tbs gcometric mean and standard deviation is
show baide each data ret The associatecl P values for the Icvel of signifierno in the two groups is as follows: For SP2 al1 IgG subclasses
P = NS;for SP4A IgGi P = 0.0 1; IgG2 P = NNS; IgGl P < 0 . 0 1; IgG4 P = NS; for SP6 IgGi P = 0.00 1: IgG1 P = 0.05: IgG3 P < 0.00 1:
IgG4 P = NS; Sp7 IgOi P < 0.001; IgG2 P = 0.04; lgG1 P c 0M)I; lgG, P = NS. h e Ordinate k on a log scale.
Fig. 6.

surface markers, such as the IL-2 mxptor. CD4 CD8 and

CD25 into the serum has aiso bcen observai [M-381. Al1 OC
these obse~atioassuggat that the mechanism which normally
control IgG isotype and Ig subcass rcsponses arc no longer
king maintaineci in HAM/TSP patients.
Chronic immune stimulation as a coowquence of HTLV-I
infection may laid to autoimmune r s p o n s r There is a
region of HTLV-Ipl9 that is homologous to a region of
rnyeIin basic protein aithough it does not correlate with
known encephalitogenic regions [39]. HTLV-I has ken

dcmonstrated to infect al1 Iints of astrocytic and glial

origin [40-421 and the HTLV-Igenome has been demonstrated to be prescnt at the major site of pathology in
the spinal cord [43, 441. Abnormal induction of the MHC
class I and 11expression in neural tek [35.40], in the presence
of HTLV-1 proteins cross reactivc with neural proteins (such
as myclin basic protein) might be enough to trigger an
autoimmune response, This process may be initiated as a
consequence of aati-viral tesponses against HTLV-Iin fected
neural ceIIs, resulting in local inflammation which attracts

251
other immune effector lls to the infecteci region, Due to the
fact that neurai tissues do not normally express MHC
molecules on their ceIl surface, and the fact that certain
neurai proteins may not rcach the thymus in SuffiCient
concentration to effect deletion or inactivation of selfreactive T d k [4q, T-cek capable of rccognizing these
neural-specific proteins may corne into incfcastd contact
with neural specific peptides ptcstllted by the HTLV-I
induced MHC and lead to an a u t o b u n e tcsponse. It has
b e n demonsmted that i n d IeveIs of interferon-gamma
(FN-) exacerbate hown ne-autoimmme
diseases [45]?lus, in the prcsetl of the appropriate MHC background,
the elevated Ievels of IFNq prtstnt in H A M m patients [7]
could enhan the dcvelopmcat of the autoimmune proccssFurthemore, the cross-nactivc cpitopes betwcen HTLV-1
proteins and neutal-@c
proteins such as myelin basic
protein may also contribute to the dcvclopment of an autoimmune proces It is clair that the devclopmeut of HAM/
TSP is a compikated interactive proccss bctween HTLV-1
and the host's immune and nervous sysems,
The anti-peptide nspanses to the cnvelopc protein
revealed significant rcactivity to the SP4A region, which
has been dcmonstratcd to contain a B ccii epitope [30], a
helper T UU epitope [46), aad a cytotoxic T ceIl epitope
[47& The SP4A region has aiso ban -atcd
with vins
neutralization [48], The anti-SP4A rcsponscs were in
marked contrast to the poor responses observeci for the
SP2 region which has aiso k e n associated with virus
neutralization [49]. Thus, the ncutraIizing antibody
responses observed in both groups wodd appear to be
pnncipally associated with the SP4A region and probably
restricted ta the IgGl and IgG3 nibclasses. However, the
assays used to detect the antibadies did not aUow for the
detection of antibodics which bind to conformational
epitopes. Thus, attempts to monitor neutraliting antibody
titres with SP-2 and SP4A peptides in ELISA will underestimate neutralizing antibody rcsponses to native gp46- It
is clear from the use of a direct neutralizatioa assay that the
titres of neutraiizng antibody were clcarly elevated in
HAM/TSP patients.
In most cases, the magnitude of the antibody responses
to the envetope was grcatcr than those cespanses directed
against the gag pl9 and p24 proteins. h i s is what one
wouId expect since the envelope protein is the principal
antigen exposed to the immune systcm and is usuaily the
first protein to which seroconversion is observcd [50]. In
addition, the distribution of responses in IgG subclasses
demonstrated agaiast the gag and the envelope proteins
were not equal suggesting that the immune responses to
these viral proteins are differeatially regulated. The IgG
antibody response was restncted to IgGl and IgG, subclass
responses for both the gag and the envelope proteins, A
similarly restricted response has been reported for other
vintses [Sl]. An enhanced antibody response to the envelope transmembrane protein. gp21, as monitored by the

anti-peptide rcsponses to the SP7 region, was not observed

when compared to the respoases to the regions encompassed by the gp46 peptides SP2, SP4A and SP6- This is in
contrast to a =nt report which suggestcd that there was
an eltvated ttsponse to the gp21 region in HAM/TSP
patients [19]. Peptides to sunilar regions were used and
the peptide ELISA was pcrformed in a similar fashion.
Perhaps this discrcpancy in rcsponsiveness rcsults frorn
variation in ethnic background of the individuds in this
study. Howevcr, ptcihinary evidcnct from MHC class II
typing indicatcs that the HAMfTSP patients in this study
have haplotypes reporteci to be prcsent in HAMlTSP
patients from Japan [52] (G- Dekaban & J. Leushner,
unpublished observations). However, we cannot rule out
diaercnccs with respect to MHC class 1 gcnesT ' e n collectively, the rcsults of this study suggest that
HAMlTsp patients have elevatcd antibody responses
involving the IgM, IgA and IgE isotypes. By CoUowing
Iongitudinay the asymptomatics anti-HTLV-I isotype
profile, it may bc possiile to dctcrmiae whether an asymptomatic individual is progrcssing towards the development of
HAM/TSP, The IgG subciass proie will not be useful in
this regard. Any alterations in the isotype pro& could
signai the naxi for ncwologicai examinatioas and the use
of appropriate drug thetapies such as high dose steroids
[27, 531 or treatmcat with antiwals such as AZT [54] for
which tri& arc begiuning (Dr S. J- F. Ogcr, Department of
Neurology, University of British Columbia, Vancouver, BC,
Canada, persona3 communication). While the percentage of
HTLV-1 iafcctcd asymptomatic indiuiduals who will go on
to develop HAM/TSP wiU be smd, once full-bIown HAM/
TSP develops, t h e ate no usefui trcatments for this disease
which is progressive in natureACKNO WLEDGMENTS
G. A. Dekaban and G. P. A- Rice are Ontario Ministry of
Heaith Carecr Scimtists. TJ. Paiker is supported by a Leukaexnia Society of America ScholarAward- J. Arp is supported

by an MRC of Canada Studcntship. This research was

supported in part by grants from the Medicd Research
Couacil of Cana&, Lcukernia Society of America and the
NTH/NCI grant CA40660. We thanlc J. Tomabuono, J. Smith,
S. Hota, and RI Mitchell for theit excellent technical assistance, and Cimdy Woag for couducthg the statistical analysis.

REFERENCES
I Poiesz BJ, Ruscctti FW, Gazdar AF, Bunn PA. Minna JD. Gallo
RC. Detection and isolation of type C retrovints particles from
fmh and cultureci lymphocytes ofa patient with EutancousT e l l
lyrnphoma. Proc Nat1 Acad Sci USA 1980;?7:74lS- 19.
2 Hinuma Y, Nagata K,Hanaoka M et al- Adolt T-cell leukemia
antigcn in ATLL cd1 nt and detaction of antildies to the
antigm in human sera. Proc Nat1 Acad Sa USA 1981:78:6746-80.

37 Shohat B, Shathi Y, Sidi Y, Nadel G, R-blatt

ID, Mytes D.
47 lacobson S, Rcadcn J, SvciIan R PalkmT-Induction of CDJ',
Elcvated saiun interleukia-2 recqtots in a newIy ident5cd
HTLV-I speciac cytotoxic T-lymphocytes from patients aith
group with bigh ntcs of human T-cell leukania virus typc 1
HAM/'TSP wgnition of an immunogcnk region of the pp16
infection ia LmwL I Mcct Dis 1991; l6M31-2
cnvclopcglycoprotein of HTLV-1, J Immundl99 1;146: 1155-62
38 Tsukada N,Matsuda M. Miyagi K,Yanagisawa N.SoiubicCW
48 Tanaka Y,2eng L, Sbiah' H,Shi& H, Tozawa H-Identificauoa
and CDS in the paiphtral blood of patients M'th multiple
of a ncuralaation @ope on the cavclope gp46 antigen of
sclcrosis and HTLV-I-assaciated mydopathy, J P~eurairnmuaol
human T e l l lcukemia virus typc 1 and the induction of
199l;39285-93neuttaking airti'body by peptide immunization- .J Immun01
39 Wcnshcnkcr 8, Riec G P k Retroviral diseue of the PCNOUS
1991;147S4-60,
systcm- III: semg u ~ i med
, ~yelinariona d Dwiytlinaao~ 49 M e r TJ, Riggs ER, Spragion DE er uf. Mapping of homoNY: Plenum Publishg Corp, 1989:14552.
logous, anbo-terminai neutalia'ng regions of human T e I I
40 Hirayama M, Miyadai T,Yokocbi T et rrL h k t i o n of human Tlyniphottopicvirus type 1and II gp46 envtlope giycoproteins. J
Iymphotropic vinu typc I td astrocytcs m vitro with induction of
=ol 1992;66:5879-8950 Sato S, Okocbi K. Transfiision transmittcd iafcction of HTLV-1
class If MHC- Neumsci Lett 1988;9234-9and iu prrnnion. In: Takatsuki K,HiDuma Y,Yoshida M. cds.
41 Macchi B, Caronti B. PaacilaM,Bonmanar E,Lam G- Encct
GANN M o n o ~ p on
h Cancer Rcsearch Vol. 39. Boca Raton,
of human T lymphotmpic rctrovus-1 expure on culturcd
Flocida- CRC naS,LW2 151-62
human giioma d
i lins, Acta Ncuropath 1991;81:670-4.
51 Sundqvist Y-& Linde A, Kurth R et ol, Rcstricted IgG subclass
42 Yamada M, Watabc K, Saida T,Kim SU-Iacreased suseeptibiiity of h m fetal amoqta to humaa T-lympbotropic virus
rrspomcs ta ~ V - l i / L A Vand to Cytomqpiovinis in patients
Mth AIDS radlympadcn~pathy~
J Mi Dis 1986;153:970-3.
type 1 in caltue J Ucuropth Exp Ncur0l1991;50397-10743 Powers C, Wcimhdct BG,Dekaban GA, Kaufmnn JCE, Kcc
52 Uruhr U, Sonoda $ OyUW M et al. HLA haplotypc-hked high
GPA Pathologkd an m o k d a r biolagical fehins of a mycioimmunenspoasivcaessagahst HTLV-1 and HLV-I-datcd
p t h y as~ociatedwith mv-r m r i - O & CM J N-01 sci
myelopathy: comp9rison with duit T d lculrcmia/iymphoma
1991;18:352-5Ann N e m l l988;US t 43-54,
44 BhiAI, Wcy CA, Wrrcbsmrn W et ai. RTLV-i-ium5ated
53 Kir0 J, FUK, Itoyama Y, Goto i,Hasuo. Leukocnccipamyclopathy elinicopathoiogc correlation with I c d h t i o a of
lopathy in HTLV-I-usOaated myelopathy~uopical
p r o W to spinai cord Newology 199l;4l:l99O-Z
paraparc&: MRI analysis and a two-ycar foilow-up study aftcr
45 Miller A, Hafier DA, Weinct HL, Tolcrancc and suppcessor
corticosteraid thcrapy. J Neurol Sci 1991;106:41-9.
mechanisms in expcchmtai autoimmune aicephalomyciitis:
54 Gout O, G
d A, Iba-Ziacn Mt. Sa!- The effect of zitovudint
implications for immtmothcrapy of human autoimmune dison chroaic myclopathy amxiatcd with HTLV-1. J- Ncurology
taxs. Tht FASEB J 1991;5-S60-6.
1991;238:108-946 KurataA, P a l k T,Strran R,SccarccR HayncsB,BcraoWryJIrnmunodaminaot sits on human Todl lcokcmia virus typc I
Receiwd 16 Decembtr 1993
envclopcprotcia for munaCTcdls.J Immunol1989;143224-30.
Aceqted m rcvircdform 8 AprilI994

Recognition of human T-leukemia virus (HTLV-1) envelope by human CD4' TceU

lines from HTLV-1 seronegative individuals: specificity and clonal heterogeneity

Manca, F., Li P h , G., Fenoglio, D., Teresa Vaiie, M., Kunkl, A., Ferraris, A..

Lancia, F., Saverino, D., Mortara, L., Balderas, R., Am. J., Dekaban, G.,
Dalgleish. A.G., Lozzi, L.,and Theo~opoulos,A.
Blood. 85, 1547-1554 (1995).

Jacqueiiae Arp
Robarts Research Institute
P.0. Box 5015
100 Perth Drive
London, Ontario, Canada
N6A SK8
519-663-5/=Fa: 519-663-378

BLOOD
Journai Permissions Dept.,
W.B. Saunders Company,
Orlando, Florida
32887,USA

REcEIVED
Y AY 0

2 1996

4241)

c'r- ;ii\lALS
?cCElVEr!

APR 3 O 1996

~ E ~ I s S WDEPARTRIENT
S

1 am a graduate sadent complethg my doctoral thesis at the University of Western Ontario,

London, Ontario, Canada. 1hope to acquire your permission to include within my thesis a
complete papa ami two figures h m another paper that have been recently pubshed in your
journal. The entire papa will be included as an appemlix in my dissertation. 1am a COauthor on this paper: Manca, Fe,Li Kra, G., Fenoglio, D., Teresa V a , M., K d , A.,
Ferraris, A., Lancia, F., Saverho, Dm,
Mortara, L., Balderas, RR.,
Am.
Deksban,
G., Daigieish. A.G., Lmzi, L., and AmNo Thcofiiopoubs. 1995. RemgMon of humui TIeukemia virus (ETLV-I) envelope by human CD4+ T-celi nes from HTLV-I
seronegative individuals: Sperincity and donal heterogeneity. Bbod. 85, 1547-1554.
J O 9

In addition, 1 wish to ask for your permission to use two figures, previously published
in your joumai, within my iatroductory chapter. The figures were included in the pubiished
paper: Franchini, G. 1995. MolccalPr medioiiipms of humon T-ce11 leukemal
Iymphotropic virus type 1idedion. B W 86,36ie3639. The figures that 1wish to use
from this paper are labekd Figure 2 a d Figure 5.

1 have personaiiy asked Dr. FabMo Maaca and Dr. Genoveffa Franchini for their
permission, of which they have verbaiiy gmted; however the Faculty of Graduate Studies at
the University of Western Ontario, Canada, requires that I receive official written permission
from your joumai, before my thesis can be accepted for hifilment of my degree. Your
expedient reply would be greatly appreciate. Thank you for your time and cornideration.

Jacqueline Arp

Recognition of Human T-Leukemia Virus (HTLV-1) Envelope by Human

CD4' T-CeIl Lines From HTLV-1 Seronegative Individuals:
Specificity and Clonal ~ e t e r o g e n e i t ~
By Fabrizio Manca, Giuseppina Li Pira. Oaniela Fenoglio, Maria Teresa Valle. Annalisa Kunkl. Anna Ferraris,
Ffavia Lancia, Daniele Saverino. Lorenzo Mortara, Robert Balderas, J a c q u e l i n e Arp, Gregory Dekaban.
Angus G. Dafgleish, Luisa Lozxi, and Argyrios N. TheofiIopoulos

HE HUMAN RETROVIRUS human T-leukcmia vinis

type 1 (-lTLV- 1)' is tesponsiblefor adult T-ce11 leukernia (ATL)"' and is associated with the nturologic disease
uopicai spastic paraparesis (TSP)'- The antibody rcsponse
to HTL,V-1 antigens can be used for the diagnosis o f the
sempositive staws7 It is also Iikeiy that an ongoing immune
response (humoral andfor ceilular) may be rrsponsible for
the slow progression of the associated discases.' or, if prop
crly driven by meam of vaccination, it may induce a protcctive response."

Neutraiizing antibodies and cytotolric ceUs are the chief

cffcctor mechanisms at work during viral infccti~ns.~'"n
both cases. Th cells" arc r e q u i d for ctoitai expansion of
ancibody-sccreting B cells and of specific cytotoxic cells.''
Therefore, the identification o f Th cells specific for fCTLV1 in the human repertoire is important for the design of
candidate vaccines' that uiciude vual T-helpct cpitopcs. Furthemore, they are useful for the identification of viral antigens that can be used as in vitro probes to test for the presence o f HTLV-I envelope-specific cellular immunity, As
shown in the human immunodeficiency virus (HIV) sysrem.'' this may be an additionai diagnostic tooi in scronegative high-risk individuals.
MATERIALS AND METHODS

The recombinant fragments of HTILV-I

envelopc werc purchascd h m Rcpligen (Cambridge. MAL Thcy
wcrc e x p d in Eriierichia col; and containcd nmino and carboxy
mils of E coli origin of 10 to 30 midues. Squcnces are bascd
on pubiished data for HTLV-1 e n ~ d o p c - Fragment
'~
RE1 included
rcsidues 26-200. fmgmcnt RE3 rcsidues 166-307. fragment R U
miducs 308-40 1. and fragment RE6 m i d u e s 166-40 1 The prcgrn[ion and purification of the cnvclope plycoprotcin pmduced in the
baculovirus expression systcm has betn described in detail." A panel
of syntheric peptides (20-mers overlapping by 10 r t s i d u a ) was produced by P. Neri (Univcnity of Siena. Sicna. Italy). Figure I illustniu the boundaries of the recombinant antipns used in this study.
RPMI 1440 medium (Flow. Irvine. UKI wris used for ihe penention and the maintenance of the lincs and for prdifention assays.
It was supplcrncnred with 10 m m o K L-gfutilrnint. 0.05 mrnoUL 2mercapioethanol. and 5% autologous heparinized plasma. Human
Anrigem and media

Blood. Vol 85. No 6 lMarch 15). 1995: po 1547-1550

recombinant inrcrfcukin-2 (IL-2). kindly donatd by R Dclla Bitw

(EuroCctus, Amsterdam-Milan. Iaiy) was uscd at n final concentration of 30 UfmL
Gmercuon of T-celI fines specijicfor HTLV-l envelope. T-ccll
Lines wtrr obtaiatd by a buic pirwocol pnviously dcscribed in
delail'*" from HTLV-1-aaive (nonimmunizcd and noninftctcd) individu* who did aot k l o n g in catcgananes
at risk because of sexual
bchavhx or
blood transfusions. Bccfly, h-nized
vcnous
blood was scpanicd on a ficol1 gradient COobtain pcriphcd blood
mononuclcar cclls (PBMO). PBMCs wue pulsed with antigens at
1 p g l d at a catl~~tration
of 10' cciWmL for 4 hours at 37CThe ceils wtte subscqucntly dilutcd to 2 X 1CPfmi and 2 m i ceIl
suspension was se#kd in a 24-wel1 plau (Costar. Cambridge, MA).
Eve days latcr, IL-2was added and the cclk wcrr split whcn confluent The cck wtn txpandcd with IL-2for IO to 14days and restimu1;ucd wilh antigcn-puisecl. 3 r a d irradiaud autologous P B M G
and IL-2for two to ihrrt cycles to obtain m b l i s h e d T c r l l line3.
Prrvious upcrimuits (dam not shown) have shown that, after four
stimulation cycles. onIy adgm-spa5fic T celk are pic?icnr in the
bulk cuitures, because al1 of ibc clones gcneratd at this stage with
a mitogcnic stimulus rrspond to the relevant antigen.

From the Lkpamenr of lrnmunologv. San Manirio Hospiral- Unr-

ver si^ of Gema. and Advonced Biorechnology Cenrer. Grma. Irai\ :

the Deparrmenr of Immunolugy. The Scripps Rescarc-11t~tsrir~ire.
h
Jolla. CA: the I ~ o l o g Gmup.
y
The 1.P- Robarrs Krsrnrci-hIn~rr
rure. London. Onrurio. C o r d a : the Division
Oncofogy. 3r
George's Hospiral Medicuf Schml. London. UK;utrd the Dcponmenr of Mulecular Biology. University of Siena. IiulySubmirred June 13. 1994: accepted Octobcr 25. IW.
Supponcd &y granrs jrom Assoeiazione Iraliana Ricrrcn Curicro.
Milan. frotn rlie ltalian Minisrries of HealrIt f6rh AIDS Prt-ycvrl uml
of University artd Scienrijic Rcsearch 140% Fun&. t i t d j r n m rlw
National Rescurch Cound (CNR, Ravie. PFCF).
Address reprirrr requesrs rn FubriLio Munca. MD. lkp)crn~rrc.trr
rnt
Immwiufogy. Sun Murtino Hospital. Viulc Benedrrro W. 1O. 16132
Gcnw. IruyTlic publication cosrs nfrlris article weie d e f y c d ai pan f>v paRr
charge pa'menr. This u n i c k musi rherefarc be Itrrubv m n r k d
"advertivmmt" in acco&tc-e ~ i r hi 8 U.S.C. SCCI~OII 17-U. d d v I ~
inrficure r11kfacr.
C 1995 by The Americrrn Socieg of /fe~tiarolo~:y.
0a064971/95B506000363.Ml/

257
MANCA ET AL

Prol$erarion asse- Rat-bonom micnnitcr plates cmtained 2

x IV T ceU. and 1@ autologous 2.500-rad icradiami PBMCs as
p m n t i n g ceIk (pn=viously pulscd wiih the rekvant antigeus at 1
pg/mL) in 200 pL culture medium pcr weU 'be cuiturcswma pilsed
after 36 hours with 0 5 pC3 'H-thymidine (5 Cimmol; Amashm
Arnmham, UK)for 6 houn- After hameshg with a Micronute 1%
apparatus. the dry Murs w m counial in a Mauix % kia counut
(Packad-Cankm, Meridan. CI')wifhout scinall9tion fluid nie
counting efficicncy of this insnument is 2 W of convcntionaI scintillation countcrs, Rcsults are gven as mean value of duplicate culums
-I SD.

EJzMre of specij5i precu~orfiequcncy~Because of the iimittd

number of cclls amilable from the same d o n a rcpcatedly bled for
the expcrimcnts and the lack of primary pliferarive rrspanse, formril lirniting dilution assayf' could not k pcrformcd to detmninc
the frqucncy of RE and of envtlop-spcific celIs in PBMCs, In
addition. because of the low prccursor frrqucncy, WC could not use
conventiond frtquency auaiysis in which the hiceIl conntration yields 1 0 % positive w e k in fact grrater than I0b-cclls pr
wel1 should be platcd but tbis rrsulrs in ovcrcrowding and no posiuve welb could bc deiMtd in preliminary assays. Thus. WC used an
alternative malysis to estirnate prrcusor kqucncy. Positive signais
wcre cnhanced in a two-sup assay- in the first sep. 24 wclis of a
microtirer plate wcrc sccdtd with IOS aniigcn-pulscd PBMCs and
IL-?was added 5 days latcr. In the second sup on &y 15, cach
well was spiir into &ce round-bottom wells. one to maintain the
clones for furthcr expansion and che othus to test sptcific pmliferauve mponse with o r without anugen. in chc -cc
of 5 x IO'
irradiated autologous PBMCs. niese wclls wtrc puIsed on day 3,
harvcstcd aftcr 16 hours. and prorrsscd as d c s c r i M Stimulation
indexes 6 1 : cpm with angcnlcpm without antigen) rangcd from 1
IO 9. Wells with an SI grtater than 2 were s c o d as positive In
fact. we have previously established thai spccific clones can be
funtier expanded from the well with an SI -ter
r h n 2 This assay.
although not as accurate as formal Iimitingdilution analysis." allows
us 10 esrimate the relative frcqucncy of naive pmursors specific for
H1ZV- 1 antigcns vcrsus p m r s o r s spcific for rctatl antigcns such
ris retanus toxoid (T13 and purilied pmtcin derivarive (PPD). as
describcd beiow.
RNAse protection assay- RNA preparation from T ccIls, riboprote design. labcling, and the RNAse protection assay have k e n
descnbed in d e ~ i lThiny
.~
Va riboprobes in three probe sets wcre
used. Probe set A was V B ? L I . 20.1. 17.1. 8.2. 10.1. 18.1. 19.1.
8.1. 16.1, 15.1. 4.1. 11.1. 14.1. 5.1. and 5.2. Probe set B was VO

6.4. 132. 3.1. 1.1, and 7.1. Robe sec C was VB Zlb. 133. 83.
13-1. 65. 121% 55, Zla IZlb. rad 5.4. AU sets w a t according
to the nomenclature rcponed in W m et ain and Plaza et ai.24
Bnefly. tocal cellular RNA was hybMtcd ovtnughc at 56C with
CB probt or with codi ~ - n d i o l a b d c dprobe ser Unhybcidizd
pobes w m digcstcd with WAse rnd pmicctd pmbdRNA duplues wmc purifid clearophorrscd on o 6% polyacrylunidt sequtncing gci, and auiondiognphcdTbtpaicctcdhybridizcdprobes
of defincd lcagth wcre idmtified on t& fih. Lberitby pcrmimng
dctinition of VB gcne usage by the Tscil IincsT-ce11 recepror TTCR) V&naiysis by pofynunrrc chuin reanion
(KR)- Analysis of TCR V@ genc segment usage by spccific Ta11 nes uns @Ormcd by PCR on reverse mnsaibed cDNA by
using a panel of oligonuclaocidc primas and foUowing rhe pmcedurc
descricd in derai1-*

RESULTS

T-cell Iines specijic for REfiagmcnts. T'ce11 lines specificfor one or more RE fragments wcre obtained from seven
of eight individuals, A specific nsponse can be detected by
the second o r third nstirnulation cycle (data n o t s h o w ) and
by the fourth or fifth cycle the Lines are cstablishedFigure 2 sumrnthe Iines geacrateci h m different
donors. The lines were testcd after four cycles, Four of seven
donors gave rise to REl-spccific iincs, six of seven to RE3spccific Iines, six of seven to RES-specific lines. and rhree
of seven to Wspecific Iines.
The lines arc spccific for the fragments used for in viuo
immunizauon. and no recognition of the E coli scqutnces
shared by the different recombinant fragments that flank the
HTLV- 1 sequences can be detccted. In fact. a representative
T-ce11 line raised against fragment RE 1 docs not proliferate
to RE3 and RE5 (sharing the same E coli sequcnces with
RE11. but proliferates to RE6 chat overlaps with RE1 frorn
residues 1 65 to 200 (Fig 3). The lack of response to fragment
RE3, which overlaps with RE1 from aa 165 t o 200 as RE6.
is hard to explain. This fragment has no toxic activity in thrit
RE3-specific Iines can be generatcd from certain donors (Fis
2 ) and it has no inhibitory activity when coadministered wirh
stimulatory fragments RE1 and RE6 *g 3).
The fine specificity of two Iines derived from the same

HTLV-1 ENV-SPECIFIC T CELLS

donor (AF)was defined by meening with the pane1 of overlapping peptides. Line REI was responsive to peptides Id1180 and 171- 190 thar overlap from i 7 1 to 180. and line RE6
was responsive to peptide 31 1-330. Because of their narmw
specificity, these lines were examned for clonal heterogeneity ( s e beiow).
T-cell fines spec$c for HTLV- I recombulanr envelupe
glycoprorein- Lines were induce by in viuo immunization

Kcpm

Kcpm

with the envelopc glycoprotcin in particulate or solublr forni

from 9 of 10 individuals- Figure 4 shows two representirivt:
lines gencrated h m individual FM by using pattiulritr or
soluble antigcn that nspond to the envelope glycopr~rrin
indepmdcntly of the paniculatc or soluble fonn. Peptide
rnapping showe specificity for the overlapping peptidrs
20 1-220 and 21 1-230 in both lines. suggesting that the response CO this in vivo-dcfined irnmunodominant regivn is a
constant fcaiwc in lines gentratcd independentiy tioni the
sarne individuai against the same antigen in two differenr
fonns- An carly rcsponse to the recombinmt RE1 ti;igiiieiir
was dctccred at the third cycle in ihe line induced with
soIuble antigcn (datanot shown), but it was lost at the fourth
stimulation cycle, in apparent association with restriction
in clonal heterogeneity (sec below). Figure 4 also show!, ;i
line generated from donor LO that rrsponds ti, frdgmcnt RE I
and recognized cwo nonoverlapping peptides. Y 1 1 10 and

Fig 2. ProTierrtiv. mgamse to mornbinant RE fmgmcmtr. Lines

were genecrted (rom swi individu*
in vitro immunaitkn with
the individual RE fragments and t @ U d
wih the eormponding h g mmts for prolifwativa nrponse afinfour cydes- KI) &dtgmund
proliferrtion; I0) antigm-indued prdifwation. REl-rpwiCi lines
were o b t r i n d from fius in&iduak REJ-, R E , andm-spmdiilines
were o b t r i n d from six, 8ix. 8nd thm indivitluab. mp.ctiuely.

141-160,
Esrimate of precursor frcyucncy. The frrqucncy ot Tce11 precurson specific for individual RE frigments w', a t r mated in individual DF (Fie 5A). The tirquencieb rmge
from 1 to 2/10" unfractionated PBMCs. In the case of entelope-specific precursors (donors FM, DF. and GB; Fig SB I.
frequencies are in the same range. These low frequencirs
support ihe notion chat the Iines have been genrrdted truni
donors that are HTLV-1 naive to the best ofour knuwlzdgs.
By cornparison, the frequency of T-and PPD-specific proliferating cells dctected in the sarne donors retiects the metti-

259

MANCA ET AL

ory state to these recall antigens (37 and U11V wiih donor

DE 48 rind 26/10b with donor FM).

Clonal hererogeneip of HTLV-I envelope-spec$c T c e k
Keeping in mind that recognition of one single peptide d m
not imply T-ceIl monoclonality.~two T-ceiI l i n s from donor A F specific for fmgments RE1 (ovcrlapping sequence
171- 180) and RE6 (peptide 3 11-330) were a n d m for use
of TCR VB gene families. An W A s e protection assay was
used on mRNA extracted from the T-cell lines." Figure 6
shows the Vp gene families that were uscd by the two TceIl lines specific for the ncombinant RE fiagrnuis. The
RE1-specific line includes 1 1 major Vp faniilics- In the case
of the RE6-specific fine. six major VP families wcrc uscd
These remlts. summarized in Table 1, indicatc tbe presence
of a large cohort of clonotypes engage- in the response to
narrow regions of RE recombinant fragments-

Fig 4. 1-cell lines rP.cific for HnV-1

~lVcopratri%
resentnive T-celI lines &tainrd Wrn i n d i d u a l FM wrrr! inducd by
in vitro immunrtion with soluble or with p.riwlate rnvdope.
When teste at the fourth rmimulrtion m e . the lines exhibitrdfull
cross-reaaivitywith the rwa rntigoniifonnr(WuMe a d ~ l n i c u h t e
cnvelopel. Recognitionof the m o m b i n m t po-n
mt.c*n. which
is associated with mvdope as 8 nruh of th* ritpnuian swtern.
was never detected. Whrn w i n n d with the pane1 of owdapping
peptides. only m a ovcnlapping p.ptid.l yie1d.d prore
iratv
ie
nspomc (201-220 and 211-2301. R.eognitian of the RE5 iragrnent rt
the thkd cyds bot rhownl wu no longer dWM.bl8 at w fwtth
cycle. Line LO induted with partieulate rnvelope was IP.tific for
nonovtriappingpeptides 91-110 and 141-160 and for fragment RE1.

A PCR assay was also used to follow-up clona1 heterogeneity of T-cell Iines responsive to peptides 201-220 and
21 1-30(21 1-220 ovcrlap) of the envclope glycoprotein, A
discrcte numbcr of TCR V/3 gene families was uscd by the
bulk line, with a l o s of -n
TCR VB fainilics after rep t e d stimulation cyclts (Table 2). The association of reduction in cIond hctuogcneity and loss of rrspansiveness
to recombinant RE fragments may be coincidental, or some
RE responsive clones may beloag to the VB families bat
wcrc eventually lost in the buik culturc.
To fiinher define clonal heterogeneity of the same T-cc11
lines spccific for cnvelopc. clones wtrt obtahed by limiting
dilution and uidividually examined by PCR to define thcir
Vfl gene usage. AI1 of the clones exhibiteci the samc peptide
specificity as the bulk lines h m which uIcy werc dtnved
Table 3 shows that most of the clones arc monoclonal in
nature. whereas 4 of 14 (28%) are not monoclonal. in
agreement with the ex+
theoretical fiequency of 25%
accorchg to Poisson distribution when miting dilution is
perfonned at O 5 cdWwelLz'

Antibodies7 and cytotoxic cclk?'= specific for CCIZV-1

have becn detccted iainfatcd patients. In addition, extensive
information is avaiiabie on B epitopes defined on cnvelope
glycoprotein with MoAbs and witb polycloaai antibodies."
Our focus in HTLV-1 -mfic human T-hclper cells stems
from the fact that both antibody production and induction
of cytotoxic ceils are dcpendcnt ou specic CD4 ~ e l l s - ~ ~ . ' '
The goals of our smdy w c h (1) to show the presence of
specific T-helper ceils in the riaive rcputoire of seronegative
individuals not bclonging to categones at risk of cxposure
to HLV-1; (2) to test rheir fine peptide spccificicy: (3) to
fic
estimate the frcquency of WI'LV-1 e n v c I ~ p e - ~ prccursors in the human naive repcrtoirq (4) to define the clonal
heterogeneity of the human primay nsponse in vim: and
( 5 ) to use the T-ce11 lhcs and clones as allular reagenrs
to define recombinant products as appropriate antigens for
protfcratiue assays to test cellular immunity in HTLV-Ipositive individuals.
The frcqucncy of Th ccll pncursor specific for HTLV-1
envelopc or recombinant fragments in the naive npcrtoire
is far too Iow for detcction by a conventional proliferation
assay, as secn in the case of HIV antigcns," but the relevant
cells can be expandeci in vim by tcpeatcd stimulation with
various antigenic prepadons." Specific prccursors are present in the repertoire of seronegave individuals, and numerous different donotypes can ~ccogniztthe retroviral antigens. The estirnated frcquency matches with the observed
frequencies for other prirnary retroviral antigens (gp120 and
p66 of HIV},as opposed to much higher frtquencics for the
conventional recall antigcns testeci here. In addition. because
the seronegative donors used for this study do not belong
to caregories at risk (because of sexual practiccs or blood
unsfusion) and belong to a geographical region in which
the incidence of m V - 1 seropositivity is negligibIe (thus
ruling out also in utero txposure). these can be cansidered
bonafide prirnary in vitro rrsponses.

261

MANCA ET AL

rt.O. d mrp.cifitwhrr,t8st.d in an intigenthivm p i d i f e u n y . The pan& show the

protiction -y
p.r(omd witfl vp probe
ut8
8. and C. bmr 1. 4, rnd 7 show the RE%
a p a d k 6n8-Li- 2 5, and 8 show the RE-speeific
Ilnc Luws 3.6. uid 9 rhow th.paritive wntiols for
dl of th.pmkr obtrinod by running the a w y with
RNA w8p8md hom hunun hymocytes The REI-litBe
contains 11differmt donotypes and the
REb..imffine x dilCmnt don0tYp.l. as summatm in T.M. t. ~ . r csignais vime r- dateetrd by
the VB 6.1 pmk with th.RE1 Iine and bv the VP
21.6.4.65.13.1, ind 13.3 p d n s with the RE6 iine.

high titer of prefonned mibodies is desinble, rather than a

delayed anamnestic response to the pathogen itself.
The different clonotypes defined according to V B gene
usage in T-ceii Iines with narrow peptide specificity rnay
underestimate the real clona1 heterogeneity. In fact. the weak
signds in the RNAse protection auq may indicate clones
present at vcry low frequency within the bulk iine. Furthermore. antipen-specific T-cell clones exhibiting the same
TCR VB gene usage may differ when the VDJ region of the
TCR gene is sequenced.
Expansion of specific T cells from ri naive repertoire mimicks in vitro the evenis thrit cake place during the fint encounter in vivo between the T-helper cornpsinment and the via1

Table 1. TCR An8ly.i~of HTLV-1-Spreific Lines

f

Cells

Probe Sel A

%be

Scr 0

probe Set C

Summary of TCR VLi gene usage by T-cell Iines specific for RE1 and
for RE6 recombtnant fragments. as defined by tne RNAse protection
assay tllustrated in Fig 5.

anrigens. as in the case of accidental infection or of deliberrite

vaccinarion.
These events include antigen administration, uptake and
processingby the appropriate APCs, and presentation to specific T cells that are prrsent at low frequency in ihe naive
repenoire. vaccination ha5 been proposed for HTLV-1." Ir
has been pointed out that, in animal models of HTLV-1
infection. the antibody response against envelope coniers
protection."" Because HTLV- 1 envelape glycoprotein is a
candidate target for vaccination. it is useful to gain more
information on its immunodominant T-helper epitopes. This
may ailow the inclusion of universal T-ceIl epiropes""' in
recombinant or synthetic vaccines.""'"' as proposed in other
viral systems.
The identification of epitopes thrit are recognized by moht
of the individuals in the population (immundominant and
universal) may iiisa suggest useful reqenw to detect a prrivious exposure to the vin1 antigens in the absence ot' beroconversion. ris it bas k e n shown in the case of contaci with
HIV- 1 in seronegative individuals." The presence of cellulrir
immunity in the absence of seroconversion rnay be a valuablc
diagnostic tool for the identification of in fecied individual
who have not yet developed (or will never develop) an antibody response. To use such an assay for diagnostic purpowh.
+

Tible 2 Fdow-p of TCR VB Gmw Usage by ftl'LV-1 -c

T-WI Lime

Stimulation

Cwe
2
3
rl

--

6~

ai4

+
+

+/+/+/+/-

11

14

te

21

&

+/-

+/-

-..

+
+

+
+

4-

The fable illustratesihe fluctuations in TCR VB gene usage by the bulk 1-cetlins FM spacific for HTLV-1 envelope glycopratain (induced with
soluble antigen) after repeatad srnulation cycles. Whereas the rnaionty
of VB gsnas are maintain&. VB 4.6H. 18. and 22 are gmdually ton.
and VB 11 claarly shows up sfmr the fou* &e of nimulation.

Clones were genentsd by lirniting dilution from lines FM rntuolubte envslope and FM anticorpuscolate envalope- As per fig 4. the
lines have the same n a m s p e c i f i , yat thsy differ in the VB gene
usage panem.

it is important to define antigenic prcpatations (wholc proreins. recombinant fragments, ot peptides) chat are rccognized by the majority of individuals in the population. The
recombinant fragments RE3 and RE5 secm to quaiify as
such. but the envelope giycoprotcin is dso a primary hmunogen in the majoricy of the individuah tcsted so farc With
this information, the ncxt stcp wii1 k COscfeen the rccombinant antigens and the panel of ovtrlapping syntheticpeptides
by using T celis or T e l l tines and clones generatcd from
individuals who had a previous contact with HTLV-I ;O
define the aatigens that are most frequently recognized, so
as to inctude them in snccning protocols for the definition
of cellular immunity in seroncgative high-risk individuals.

ACKNOWLEDGMEM
We arc graiefu1 CO R

Della Bina (Euroccnis, Amsterdam-Milan)

for donating Lhe human recombinant IL-2The secretaria1 help of
Mmo Xmoai and the tcchnical help of Franco Vigo and Giovanni
Benucci arc &O acknowledgcd.
REFERENCES
1. Broder S. Ga110 RC: Human T-ceII Ieukemia viniss (HTLV):
A untque f m i l y o f pathogenic retrovituses. Annu Rtv lmmunol
3:37I.1985
1. Poicsz BJ, Ruscctti FW. Reia MS. Kalyanaram VS. Gallo RC:
Isolation of 3 new type-C rcvovirus (HTLV) in pnmary uncriIturcd
cclls of a partent wtth Sezary T-cc11 leukemia Nature 294268. 1981
3. Haynes BF. MiIltr SE Palker TJ.Moore ID. Dunn PH. Bolog-

ncsi DP. Mcbgar RS: fdcntifiaion of human T-ccll fcukcmia virus

in a Japanex patient with duit T-ceIl leukania and cutantous
lymphomamus vasatlitk Roc Nati Acad Sci USA 802054. 1983
4. Levine PH, B h m W.
~ Clark f, Tamne R Maloney EM:
Gcognphic disai'bution of HTLV-1 and identification of a new high
risk population. Int J Cancer 427. 1988
5. Murphy EL, Hanchrd B. Figucroa JP. Gibb W. Lofters WS.
CampkH M. Goedert JJ, B i a m v W k Modclling rhe rsk of A T U
in pasions infected wtb HTLV-1. Int J Canw 43250. 1989
6. samc M. Usulni R h m o S. Ijichi N. Amitiani H. Igata A.
bfammoto M. Tara% HTLV-1wociued myclopaihy,a new clinical cntity- Inncct 1:1031.1986
7. FordCM. Arp J. RUEa TL King EE. DcbbanGA: Charactcraticm of t& mn'body rrspwse CO tbr# differenr versions of the
tTTtV-1cnvtlopeprMtm expressai by mmbinant vaccina viniscs:
foduction of ncutralinag ati'body, 1 \Tm1191:448. 1992
8. Dt 'b G, Bamford k AU HLV-1 vaccine: Why. how. for
whom? AIDS Res Hum Rmoviwes 9381. 1993
9- KosP~)wstiH. k d d b s e MI. lonijc S: The rok of CD4
and CD8 T ceils in virai inf&ctitms.C m Opin Imrnunol3:471,1991
1O. Dohercy K. Ahm W. Eicbclberger M,M n g SR: Rolcs of
a al3 and 76 T c d snbscts in viral immunity. Annu Rev immun01
10:123, 1992
11. Siuons m.
Oldstone MBA: Aatibody-mediami destruction
of v i n i s - i n f d ctlls- Adv lmmunol t9-29.1980
IL Sissons JPG. Oldstont MAB: Kilting of virus-infectcd cclk
by cytotoxic lympliocytcs. 1 Mect Dis 142114. 1980
13. Mosmann TR Coffmui RL; T)il and TH2 cclls: Diffcrcnr
paaeras of lymphduat seamion Icod io diffmnt functiond propertics- Annu Rev Imniunot 7:145. 1989
14. Stuhicr G. Waldcn P: Coaborationof heipcr and cytotoxic
T Iymphocytcs. Eur J immun01 232279. 1993
15. claici M. Giorgi S. Chou C, Gudeman V. Zack J. Bmofsky
J. ShG: Ce11 mcdiaud nmunc mponse io HIV in seroncgative
homosuuai men with rcccnt s u a i exposure J Infect Dis 165:1019,
1992
16. Moioham S. Seisukt H, Yoko H, Mitsuaici Y : Human adult
T e I I Icukemia vinas: Cornpletc nuclcoudc sequence of the provinis
genome integraud in Ieukemia ce11 DNA Prac Nad Acad Sci USA
803618. 1983
17. Arp 1, Ford CM, Pallrcr TL King EE, Oakcban GA: Exptession and immunogenicity of the entire human T cc11 leukaemra virus
type cnvclope protein producd in a baculovirus systcm. J Gcn Virol
?4:2 1 1. 1993
18. M a n a F. H a k h a w J. Dalglcish A: The naivc rcpertoire of
human T cells specific for HIV envclope glycoprorcin gp120. I
Immun01 146: 1964, 199 1
19. Manca F. Walku L Ncwell A. Celada F, Habeshaw J. Ddglcish J: Immunosupprcssive advity of HIV cnvelope glycoproccin
dominates over its angcnicity for spctific human T cells. Clin Exp
lmmunol 88: I i . i991

20. Manca F. Habtshaw J. DaIgleish J. Scrcarz E

: Roie of flanking varablc q u a c e s in the mogniuon of consensus T epitopes
on HIV gp120- Ew S b u m i l 23969. 1993
2 1. Lefkovits i. Waldmann H:fJmiting DiIution Ana(vsis of CeUs
in the Immune Systcm- Lmdon. UK. Cambridge. 1979
22 Baccala R. Kono DU. W d k S. Baldaas RS, 'Iheofilopoulos
ANr Genomicalty imposccI and somaticalIy mdwKfitiedhuman chyme
cyrc V pgenc rcpcctoksirrs
Roc Nati Acad Sci USA 88298. 1991
23. Wilson RK. &E
i Ccmcammn P. 3anh RK. Hood J? Saucnirt, organilation d polymorphism of murine and hurnan T-cc11
nxqtor a and /3 chah gene families. Immuml Rev 101:149.1988
24. Plaza A. Kono DH, Theofilopoutos AN: New human VB
variants. J tmmunol 147:4360, 1991
gents and pot-c
35. Genevc C, Diu A Nierat J, Caignard A Dieaich P-Y, Erradini L.Roman-Roman S. Trieki F. HercaidT:An upaimenfaily
validami pancl of subfamily-spscificoligonuclcotide primas (VJw29Ndl-w24. for the audy of human T ll nccplor variable V
gtne segment usage by potymnsc cban rumion, Eur J Immun01
2:1261. 1992
26. Maaca F. Ciarkc J. Miller A. S a m z EE. Shastri NA: Lirnited
region within hm-egg Iysozyme as rbc focus of a divcrsity of T ccU
clones. J lmmunol 1332075. 1984
27. Parkm CA. Dacuk S. Nightifigalt S. Banghiuu CfZM: A&vated HTLV-1-spcinc cyotoxic T-lymphocytes
fouad in
bea1t.y scmpositivcs as weU as in patients wirh nopicai spastic
paraparcsis. Virology 188:628, 1992
28. Elovaara L Koenig S. Brewlb AY, Woods RM, k h k y T.
Jacobson S: High huxaan Tccll lymphaaapic virus typ 1 (IFItV)specific
cytotonc T lymphacytc fiesucaacs in
with H t Z V - 1 - w
nturologid dixare. J Exp Mcd
in:rs6t. 1993
29. P W TL Riggs ER Spngiou D E Muir AT. Sccarce RM.
Randdi RR McAdams MW. McKnight A. aaphom P R Weiss Rk
Haynes BF: Mapping of hwralogaus. amino-mmhd neuuaiizing
ngions of hutnan T-ccii lympbmpc
virus type 1 and IT gp46
cnvelopc glycopmttins- 1 V i l 66589. 1992
29a Manca F. Fcnogio D.V a t M, Li Pin G. KunW A Fmats
A Savcrino D. Lancia F, Morwa L, Lcazi L, Pimes M. Dalgleish
A Lewis G:Human CW+ T ccfls can disaiminocc the molecular
and smcnrral coniutr of T epitopcs of HIV gpl2O ami
p66. I
AIDS (in press)
30. Scrcarz EE. Lrhrnann PV, Amturu* A, Benichou G, Millu
A. Moudgil K: hininuice and crypticity of T ceIl ontigenic detcrminuits. ANURCV lmmunol 11:729, 1993
3 1. Palker TJ. Matthews TL Luiglois A. T ~ M Q
ME. Martin ME.
Scearce RM, Kim IE, Buzofsky IA. Bolopsi DP, Hayncs BF:
Polyvalent human nmuaodeficiency vVus synthetic immunogen
compriscd of envclope gp120 T helpcr cell si- and B a11 neutraiization cpitops. J Immun01 1423612 1989

3 2 Mana F. Dcssi V, Sbsui N. Oki A. KapIan M. Clarke 1,

Miller A S a a n EE: Isolation of lysozyme spccific T cc11 clones
h t discriminate bctween native and dewcud antigen Cd1 h m u no1 154:420. 1994
33- BotarcIli P. Houtden BA. Haigwood NL. Servis C Montagna
D. Abrignani S: N-glycosyiation of HiV gpltO may constrain cccognition by T lymphocytes. J lmmunol 1473128. 1991
34. Maaca F: Galactose reccptors and prcscntation of HIV envcl o p glycoproccin to spccific human T cells. J lmmunol 148:278,
1992
35. Manca F,Li Pira G, FeaogIio D. Sun PF.Habeshaw A. Knight
S C Dalgitirh AG: Dendcitic -11s arc porent antigen prcsenting ctUs
for in vivo iaduction of primary human CM' T ceIl lines specific
for HIV gp 120. J AIDS 7: 15. 1994
36. S&ida H, Tochikura T, Sato T,KOMOT. Hirayoshi K. ho
Y, Hafada M. Hiauma Y, Sugimoto M. T a h i - N i h i m a k i F.
Manryama T,Miici h4. S d K. Monta M. Sashiyama H. Hayami
Ms Effcct of the cecombiit -a
viruses that express HTLVI cnvclope gene on HLV-1 i n f ' o n . EMBO I 63379, 1987
37. Kuaoka R Takebrn N, IwahPra Y, Sawacfa 7
.Ohlsuki Y,
k w c i Y. Hosbino H. Miyasi 1: Transmissioa of HTLV-1 by b l d
trpnsfunon rnd iis pcvcntion by passive immunitni;on in rabbio.
Blood 76:1657. 1990
38. S;rwda T. I w a h a Y. Ish K. Tagucbi H. Hoshino H, Miyoshi Ir immunogiobulinpophyiaxisagnsr mitlbornemnsmission
of humau T ccli leukania v i n s type 1 in rabbis. J Infect Dis
I64: 1193. 1991
39, F%nha-Bordignon P. Tm A Tamiitclm A Dcmoa S. Corndia Ci, frimvccchh k U n i v d y immunogm-cT U cpitopes:
R m h m s bioding to bumrn MHC clus II and pmmiscuous rccognitioa by T & Eur I Immunol 192237,1989
40. Kumar A Aroro R Kwr P.Cbauhan VS,Sharma P: Universai
T helper ceil deramiiiotsenhance immunogcaicity of a pIasmodium
fdcipanim maornite d w r mtigen Pptidc J Immun01 148: 1499.
1992
41. Hart MK.PaikerTJ. h t k w s TJ,Luiglois Ai,La&e NW,
MPrtia ME,Saarce RM. McDPaal C Bologmsi DP. Hayncs BF:
Synthctic
conuining T and B ceil epimps fiom human
immuaodcficiency virus enwlopc gp120 induce anti-HIV pliferative iftponses and high titns of neuaalizing anribodies in rhesus
monkeys. J lmmunol 145-2677, 1990
4 2 Hcmmin A Nam PL, Zweig M. M h e W . Cease KB.
Gard EA Mukm PD, h m e y SD. Daniel MD. Desrosiers RC.
Bazofsky JAr Priming with T helpr a l 1 epitop peptides enhances
the aatibody rrspoasc CO Lhe envelopc glycoptotun of HIV-1 in
primates, J immun01 146r1667. 1991
43. Milich D R Hu@ JL,McLachlan A Tbomton GE. Moriany
k Hepatitis B synthetic immunogcn compriscd of nucleocapsid Tcc11 siw and an envtlope B-ctll epitope- Roc Nat1 Acad Sci USA
85:1610. 1988

REFERENCES

Abimiku, A., Franchini, G., Tartagh, J., Aldrich, K., Myagkch, M., Maricham, P.,
Chong, P., Klein, M., Kieny, M X ,Paoleai, E., Gallo, R., and Robea-Ouroff, M.
1995. HIV-1 recombinant poxvinis vaccine ioduces cross-protection against HIV-2

chailenge in rhesus macaques. Nature Medicine. 1, 321-329.

Achiron, A.. P b - H a m i e l , J., and Duli, 1. 1993.Spastic paraparesis associated with

human T lymphotmpic virus type 1: A clinicai and genomic study in Iranian-bom
Mushhadi Jews.Ann. Neurol. 34, 670-675.
Ahmed, Y., Gilmanin, G.,Hanly, S, Nevins, J., and Greene, W. 1991. The HTLV-1
Rex response element mediates a novel form of mRNA polyadenylation. Cell. 64, 727737.

Akagi, T.,Yoshino, T., Motoi, M.,Takata, H.,Yano, S., Miyoshi, I., Oka, T., and
Ohtsuki, Y. 1988. Isolation of virus-produchg transforrnants nom human gastric cancer
cell line, HGC-27, iafected with human T-ceU leukaemia virus type 1. Jpn J. Concer
&S.

79, 836-842.

Akiniki, S., Setoguchi, M., Nakazato, O., Yoshida, S., Higuchi, Y., Yamamoto, S.,
and Okajima, T. 1988. An autopsy case of human T-lymphotropic virus type 1-associated
myelopathy. Hm. Pathol. 19, 988-992.

Alexander, S., and Elder, J. 1984. Carbohydrate dramatidy influences immune

reactivity of antisera to virai glycoprotein antigens. Science. 226, 1328-1330.
Alexander, N., Fulgham, D., and Goldberg, E. 1992. Contraceptive vaccine
development: secretory immune respome in mice ami monkeys. Vacc. Res. 1,331-345.
Allen, P., Babbitt, B., and Unanue, E. 1987. T celi recognition of lysozyme: the
biochemical basis of presentation. Immunoi. Rev. 98, 171-187.
Arango, C., Concha, M., Zaninovic, V., Biojo, R., Borrero, I., Rodgea-Joboson, P.,
Mora, C., Garmto, R., Gibbs Jr., C., and Gajdusek, D. 1988. Epidemiology of tropical
spastic paraparesis in Colombia and associated H'LV-1 infection. Ann. Newui.
23(Suppl.), S161-165.
Archambault, G. 1989. ERect of sodium diethyldithiocarbarnate, Corynebacfenum
p a m m and mycobacterial ceii wali extract on in vlno blastogenic respolws of bovine
blood lymphocytes. Comell Vet. 79, 11-24.

Arima, N. Daitoku, Y., Ohgaki, S., Fukumori, I., Taoaka, H., Yamamoto, Y.,
Fujirnoto, K., and Onoue, K. 1986. Autocrine growth of interleukin-2 producing
leukemic cells in a patient with aduIt T ceii leukemia. B l d 68, 779-82.
Arp, J., Ford, C., Palker, T., King, E., and Dekaban, G. 1993. Expression and
immunogenicity of the entire HTLV-1envelope protein produceci in a baculovirus system.
J. Gen. Vird. 74, 211-222.
Arp, J., Levatte, M., Rowe, J., Perkhs, S., King, E., Leystra-Lantz, C., Foung,
S.K.H., and Dekaban, G. 1996. A source of glycosylated HTLV-1envelope protein:
expression of gp46 by the vaccinidT7 polymerase system.J. Wrology. 70, 7349-7359.
Arp, J., and Dekaban, O. Unpublished data.

Arya, S., Wong-Staal, F., and M o , R. 1984. T-ceil growth factor gene: lack of
expression in human T-celI leukemia-lyrnphoma Wus-infected ceiis. Science. 223,10861087.
Baldridge, L, and Bucbmeier, M. 1992. Mechanisms of antiy-meiated protection
against lymphocytic choriomeningitis virus infection: mother-to-baby transfer of humoral

protection. J. Virol. 66, 4252-4257.

Bdl,
and McCarter, J. 1979. Biological charac~rizationof a leukemogenic virus
isloated from the CFW mouse. Cancer Res. 39, 3080-3088.
J e ,

Ballaun, C., King Farrington, G., Dobrovnik, M., Rusche, J., Hauber, J., and Bohnlein,
E. 1991. Functional aDalysis of human T-cell leukernia v i . type 1tex-respome element:
Direct RNA bindiog of rex protein correlates with in vivo activity. J. Wrol. 65, 44084413.

Banks, T., Huo, B., Kousoulas, K., Spaete, R., Pachl., C.,and Pereira, L. 1989. A
major neutraliPng domain maps within the carboxyl-terminal ha of the cleaved
cytomegalovirus B giycopmtein. J. Gen. Yirol. 70, 979-985.
Bamtt, N., Mitterer, A, Mundt, W., Eibl, J., Eibl, M., Gallo, R., Moss, B., and
Domer, F. 1989. Large-scale production and purification of a vaccinia recombinantdenved HIV-1 gp160 and anaiysis of is immunogenicity. AZDS Res. Hm.Retrovimes.
5, 159-170.

Bartholomew, C., Blatmer, W., and Cleghom, F. 1987. Progression to AIDS in

homosexual men CO-infectedwith HIV and HTLV-1 in Trinidad. Lancet, 2, 1469.

Bastian, I., Gardner, J., Webb, D., and Gardner, 1. 1993. Isolation of a human Tlymphotropic virus type I saain from Austraian aboriginals. J. firoi. 67, 843-851.

Bauer, H. 1974. Virion and hMot ceil antigens of C-type RNA tumor vimses. Adv.
Cancer. Res. 20, 275-341.
Becht, H., Hammeting, U., and Rott, R. 1971. Uodisairbed release of infuenza W u s
in the presence of univalent antineuraminiidase a n f i i ' e s . Wrology. 46, 337-343.
BenjouadTA., Gluckman, J.C., Rochat, A., Mootagner, L., and Bahraoui, E. 1992.
Influence of carbohydrate moieties on the immunogenicity of human immunodeficiency
virus type 1 recombinant gp160. J. Wrul. 66, 2473-2483.
Benson, Y. Tschacher, E., Gessain, A., Yanagihara. R., Gallo, R., and Franchini, G.
1994. Cross-neutmiking anti'bodies againsf wsmopolitan and Melanesian strains of
human T ceil leukemial Iymphotropic virus type 1in sera h m inhabitants of Afnca and
the Solomon Islands. AIDS Res. Hum Retrovilzcses. 10, 91-96.

Berman, P., Gregory, TmTRiddle, L, Nakamura, G., Champe, M., Porter, I., Wurm,
F., Hershberg, R., Cobb, E., and Eichberg, 1. 1990. Protection of chimpanzees nom
infection by HIV-1 after vaccination with recombinant glycoprotein gp120 but not gp160.
Nature. 345, 622625.
Berneman, Z., Gartenhaus, R., Reitz, M.,Blattner, W., Manns, A., Hanchard, B.,
Ikehara, O., Gallo, Ra, and Rlomian, M. 1992. Expression of altematively spliced
human T-lymphotropic virus type I (HTLV-I) pX mRNA in infected c d lines and in
primary uncultured cells from patients with adult T-cei leukemia/lymphoma and healthy
carriers. Proc. N d Acu. Sci. UM.89, 3005-3009.
Bevan, M. 1976. Cross-primllig for a secondary cytotoxic respome to minor H antigens
with H-2 congenic cells which do not cross-react in the cytotoxic assay. J. &p. Med.
143, 1283-1288.

Biron, C., Byron, K., and Sullivan, J. 1989. Severe herpesvinis infections in an
adolescent without naturai kiiier ceils. N. Engl. J. Med. 320, 1731-1735.
Bjemun, O., and Schaf'er-Nielsen, C. 1986. In AMIyrcui Electroplroresis, p. 315. Edited
by M. Dunn. Verlag Chemie, Weinheim.
Blair, D., McClements, W., Oskarsson, M., Fischinger, P., and Vande Woude, G.
1980. Biologicai activity of cloneci Molowy sarcoma virus DNA: Terminafiy redundant
sequences may enhance transfomationefficiency. Proc. Natl. Acad. Sci. US.77,35043508.

Blattner, W., Kalyanaraman, V., RobertGuroff, M., Lister, T., Galton, D., Sarh, P.,
Crawford, M., Catovsky, D., Greaves, M., and Gallo, R 1982. The human typeC
retrovirus, HTLV, in blacks h m the Cari'bbean region and relationsip to aduft T-cell
leukemiaflymphoma. Int. J. Cancer. 30,257-264.
Blattner, W., Nomura, A., Clark, J., Ho, G., Nakao, Y., Gallo, R., ami Robert-Guroff,
M. 1986. Modes of traosmission and eviden for Wal latency from studies of human
T-celllymphotrophic Mnis type 1in Japanese migrant populations in Hawaii. Proc. Nad.
Acad. Sei. USA. 83, 48954898,
Blattner, W., aad Gallo, R. 1994. Epidemiology of NTLV-1 aad HTLV-IIinfection. In
Adult T-ceil leukaemia, pp .&-!JO.
Edited by K. Takatsuki. Oxford University Press,
Oxford.
Blattner, W. 1995. New Developments in HTLV-VII Epidemiology.J. AlDS and Hum.
Retroviruses. 10, 257.
Blomberg, J., Robert-Guroff, M., Blattner, W., and Pipkom, R. 1992. Type- and groupspecific continous antigenic determnants of HTLV. Use of synthetic peptides for
serotyping of HTLV-1 and -II infection. J. Ac@. lnmwe Defc. Syndr. 5, 294-302.
Bogerd, H.,Tiley, L. and Cullen, B. 1992. Specific bding of the human T-ce11
leukemia virus type 1 Rex protein to a short RNA sequence located within the Rexresponse element. J. Virol. 66, 7572-7575.
Bolognesi, D., Bauer, H., Gelderblom, H. and Huper, G. 1972. Polypeptides of avian
RNA tumor viruses. IV. Components of the virai envelope. Virology. 47, 551-566.
Bonifacino, J., Cosson, P., Shah, N. and Klausner, R. 1991. Role of potentiaily charged
trammembrane residues in targeting proteins for retention and degradation within the
endoplasmic reticulum. EMBO J. 10, 2783-2793.
Botareni, P., Houlden, B., Haigwood, N., Servis, C., Montagna, D. and Abrignani, S.
1991. N-glycosylation of HIV-gpl20 may constrain recognition by T-lymphocytes. JImmunol. 147, 3128-3132.
Bnick, C., Resomet, N., Portelle, D., Cleuther, Y., Mammerickx, M., Burny, A.,
Mamoun, R., Guillemain, B., van der Maaten, M. and Ghysdael, J. 1984. Biologically
active epitopes of bovine leukaeda virus glycoprotein gp51:their dependence on protein
glycosylation and genetic variability. Wology. 136, 20-31.

Bruggemann, M., Williams, G. and Bindon, C. 1987. Cornparison of the effector

functions of human immunoglobuluis using a matched set of chimeric antibodies. J. Erp.
Med. 166, 1351-1361.

Buchmeier, M., Welsh, R., Dutko, F. and Oldstone, M. 1980. The virology and
immunology of lymphocytic choriomeningitis Wus infdon. A& Innumol. 30, 275331,
Bukner, C., Roberts, C., Foung, S., Lipka, J., Reyes, G., Hadlock, K., Chan, L.,
Gongora-Biachi, R., Hjelie, B., and Lai, R. 1992. Immune responsiveness to the
immw1odominant recombinant envelope epitopes of human T lymphotropic virus types
1 and II in diverse geographic populations. J. I n f e c Dis. 166, 1160-1163.
Buonocore, L., and Rose, 1. 1990. Prevention of HIV-1 glycoprotein transport by soluble
CD4 retained in the endoplasmic retidum. Noture. 345, 625628.
Butten, T O T and Hughes, R. 1981. Isolation and characterization of mosquito ceii
membrane glycoproteins. Biochim Biophys. Acta. 640,655-67 1.

Byme, J., and Oldstone, M. 1984. Biology of cloned cytotoxic T lymphocytes specific
for lymphocytic choriomeningitisvirus: clearance of virus in vivo. 3. Virol. 51,682686Cam, A., Rosenblatt, J., Wachsman, W., Shah, N., and Chen, 1. 1985. Identification
of the gene responsible for human T-ceil leukaemia virus transcriptional regulation.
Natrtre. 318, 571-574.

C m ,A., and Chen, 1. 1990. Human T-cell leukemia virus 1 and II. In Fields virology,
2nd edition, pp. 1501-27. Edited by B. Fields, D. Rnipe, R. Chanwk, et al. Raven
Press, New York.
Catovsky, D., Greaves, M., Rose, M., Galton, D., Goolden, A., McCluskey, D.,
White, I., Lampert, 1.. Bourikas, G., Irelaad, R., BrowneU, A., Bridges, J., Blattner,
W., and Gallo, R. 1982. Adult T ceil lymphoma-leukaemia in blacks from the West
Indies. Laneet. i, 639-643.

Chamberiin, M.,and Ryan, T. 1982. In nie Enzymes, Vol. 15, pp. 82-108. Edited by
P. Boyer . Academic Press, New York.
Chen, Y., Lee, T., Samuel, K., Okayama, A., Tachibana, N., Miyoshi, I., Papas, T.,
and Essex, M. 1989. Antibody reactivity to different regions of human T-cell leukemia
virus type 1 gp61 in infected people. J. Nrol. 63, 4952-4957.

Chen, Y., Gornez-Lucia, E., Okayama, A., Tachibana, N., Lee, T., Mueller, N., and
Essex, M. 1990. Antibody profile of early HTLV-1 infection. Lance?. 336, 1214-1216.
Chen, Y., Zhang, X., Dahl, C., Samuel, K., Schooley, R., Essex, M., and Papas, T.
1991. Delineation of type-specifc regions on the envelope glycoproteins of human T ce11
leukemia viruses. 3. Immunol. 147, 2368-2376.

Ciminaie, V., Pavlakis, G., Derse, D., Cunningham, C., and Felber, B. 1992. Complex
splicing in the huma T-ceii leukemia v h (HTLV) famiIy of retrovinises: Novel
rnRNAs and proteins produced by IFIZV-1. Virology. 66, 1737-1745.

Ciminaie, V., D'Agostho, D., Zotti, L., Franchiai, G., Felber, B., and Chieco-Bimchi,
L. 1995. Expression and characterization of proteins produced by mRNAs spliced into
X region of the human Tcell Ieukemiaflymphotropic v i n s type II. Wrology. 209, 445456.

Clapham, P., Nagy, K., Cheingsong-Popov, R., Wey, M., and Weiss, R. 1983.
Productive iafectioa and ceU-fiez transmission of human T-cell leukaemia virus in a nonLymphoid ce11 W. Science. 222, 1125-1127.
Clapham, P., Nagy, K., and Weiss, R 1984. Pseudotypes of human T-cell leukemia
virus types I and II: neutralisation by patients' sera. Proc. N d Acad. Science. USA. 81,
2886-2889.
Clark, N., Kushner, N., Barrett, C., Kensii, C., Salsbwy, D., and Cotter, S. 1991.
Efficacy and safety field trials of a recombinant DNA vaccine against feline leukemia
virus infection. J. Am. Vet.Med. hsoc. 199, 1433-1443.
Cleghom, F., Manns, A., Falk, R., Hartge, P., Hanchard, B., Jack, N., Williams, E.,
Jaffe, E., White, F., Bartholomew, C., and Blattnet, W. 1995. Effect of human Tlymphotropic Wus type I infection on non-hodgkin's lymphoma incidence. J. Nat1
Cancer Insr. 87, 1004-1014.
Clerici, M., Giorgi, J., Chou, C., Gudeman, V., Zack, J., Berzofsky, J., and Shearer,
G. 1992. CelI-mediated immune response to HIV in seronegative homosexual men with
recent se& exposure. J. Infec. Dis. 165, 1012-1019.

Cockereu, G.,Laimore, M.,De, B., Rovnak, J., Hartiey, T., and Miyoshi, 1. 1990.
Persistent infection of rabbits with HTLV-1: Patterns of anti-vkd antibody reactivity and
detection of virus by gene ampiifcation. 1'. J. Cancer. 45, 127-130.
Coffm, J. 1992. Genetic diversity and evolution of retrovinises. Cuw. TTops Micmbiol.
hmunol. 176, 143-164.
Cohen, S. 1993. A new goal: preventing disease, not infction. Science. 262, 1820-1821.
Copeland, T., Tsai, W . P . , Kim, Y., and Oroszian, S. 1986. Envelope proteias of
human T-cell Leukemia v i n s type 1: cbaracterization of antiseta to synthetic peptides and
identification of a natural epitope. J. Imnucnol. 137, 2945-295 1.

Cordell, J-, Moore, I., Dean, C., Rlasse, P., an Weiss, R. 1991. Rat monoclonal
anti'bodes to non-overlapping epitopes of human immunodeficiency virus type 1 a 1 2 0
block CD4 binding in vitro. Virology. 185, 72-79.
Daigleish, A., and Richardson, J. 1990. nTLV host range and its receptor: impcations
for pathogenesis. In Hia~nRenowtoIogy: H W , pp. 45-48. Edited by W. Blatmer.
Raven Press, New York.
Dales, S., and Siminovitch, L. 1%1. The development of vaccinia virus in Earle's L
strain cells as examineci by electron microscopy. J. Biophys. Biochm. Qtol. 10, 475503.

Davis, D., Stephens, D., Wiers, C., and Lachmaon, P. 1990. Glycosylation governs
the binding of antipeptide antibodies to regions of hypewariable amino acid seyence
within recombinant gp120 of human ~unodeficiencyvirus type 1. J. Gen. Virol. 71,
2889-2898.

De, B., L a h o r e , M., Griffs, K., Williams, L., Villinger, F., Q u h , T.,Brown, C.,
Sugimoto, M., Araki, S., and Folk, T. 1991. Comparative analysis of nucleotide
sequences of the partial envelope gene (5' d o d ) among human T lymphotropic virus
type-1 (HTLV-I) isolates. Virology. 182, 413-419.
Dekaban, G., Oger, J., Foti, D., King, E., Waters, D., Picard, F., Arp, I., Werker,
D., and Rice, G. 1994. HTtV-1 iafection associateci with disease in aboriginal Indians
fiom British Columbia: a serological and PCR anaiysis. Clin. Diag. Virol. 2, 67-78.
Dekaban, G., King, E., Arp, I., and Rice, G. 1994. Comparative analysis of the
antibody response to the HTLV-1 Gag and Env proteias in HTLV-1 asyrnptomatic carriers
and W T S P patients: an isotype and subclass analysis. Sc&. J. Immunof. 40, 171180.
Delamarre, S., Pipue, C., Pham, D., Tum, T., and Dokhelar, M.-C. 1994.
Identification of functional regions in the human T-cell leukemia virus type 1 SU
glycoprotein. J. Kr01. 68, 3544-3549.
De Libero, G., Flesch, I., and KauGnana, S. 1988. Mycobacteria Lyt-2' T cell lines.
Eur. J. Irrmu~nol.18, 59-66.
Denhardt, D. 1966. A membrane Nter technique for the detection of complementary
DNA. Biochm. BiopS>s. Res. Commun. 23, 641.
Derse, D. 1987. Bovine leukemia virus transcription is controlled by a virus-encoded
tram-acting factor and by cis-acting response element. J. W u l . 61, 2462-2471.

De Rossi, A., Aldovini, A., Franchini, G., Mann, G., Gallo, R., and Wong-Staal, F.
1985. Clonai selection of T lymphocytes infectecl by cell-fiee human T e l l leukemia/
lymphoma VUS type 1: Parameters of virus integration and expression. Wology. 163,
640-645.

De Th, G.,and Bomford, R. 1993. An HTLV-1 vaccine: Why, how, for whom? AlDS
Res. Hum, Retrovimes. 9, 381-386.
Deutscher, M. 1990. Guide to protein purification. In Mefhods in Ewmology, V i e
182, p.34. Edited by J. Abelson and M. Simon. Academc Press Inc. New York.
Deauni, C., Frazier, D., Huff, L., Stromberg, P., and Olsen, R. 1990. Subunit vaccine
protects nocaca nemestrina @ig-taiied macaque) against smiian T-cellIymphotropic virus
type 1 challenge. Cmcer Res. (Sqpl). 50, 5687s-569ls.
Dickson, C., Eisenrnan, R., Fan, F., Hmter, E., and Teich, N. 1982. Protein
biosynthesis and assembly. In RNA hcm~rvimes, pp. 513-646. Edited by R.A. Weiss,
N. Teich, H. Varmus, and S. Co&. Cold Spring Harbour Laboratory, New York.
Dimmock, N. 1984. Mechani,mis of neutraization of animal vinises. 3. Gen. Virol.
1015-1022.
Dimmock, N. 1993. Neubtalization of animal v h e s . Curr. Topics Microbiol. Immunol.
183, 1-150.
Doe, B., Steimer, K., and Walker, C. 1994. Induction of HIV-1 envelope (gp120)specific cytotoxic T lymphocyte resporws in mice by recombinant CHO cellderived
gp 120 is enhanceci by enzymatic removal of N-linked glycans. Eur. J. Inrmunol. 24,
2369-2376.
Doeder, W. 1986. Expression of the Autographa californica nuclear polyhedrosis vims
genorne in insect cells: homologous viral and heterologous vertebrate gews-the
baculovinis vector system. Curr. Top. Microbiol. Immwiol. 131, 5 1-68.
Doherty, P. and Zinkernagel, R. 1974. T-cell mediated imrnunopathology in viral
infection. Trmsplant Rev. 19, 89-120.
Dunn, J. ami Studier, F. 1983. Complete nucleotide sequence of bacteriophage T7 DNA
and the locations of T7 genetic elements. J. Mol. Biol. 166, 477-535.
Dnimmer, H., Jackson, D., and Brown, L. 1993. Modulation of CD4+ T-cell
recognition of influenza hemagglutinin by carbohydrate side chains loated outside a Tce11 determinant. VNology. 192, 282-289.

Earl, P., Moss, B., Momhn, R., Wehrly, K., Nishio, J., and Chesebro, B. 1986. Tlymphocyte priming and protection against Friend leukemia by vaainia-retrovints env
gene recombinant. Science. 234,728-730.
Eduoard, E., Legrand, E., AstierGin, T., Daiiibaa, R., Geofhe, S., Dalbon, P.,
Guiliernain, B., and Londos-Gagliardi, D . 1994. Characterization of monoclonal
anti'bodies duected against ht egp46 of human T-ceil leukemia virus type 1. Leukemia.
Apr. Suppl. 1, S60S64.
Eguchi, R., Kubonishi, I., Daibata, M.,Yano, S., mamura, J., Ohtsuki, Y., and
Miyoshi, 1.1988- Seriai transp1antation of an HTLV-1-transfomeci hamster lymphoid ce11
line into hamsters. Int. J. Cancer. 41, 868-872.

Elovaara, L., Koenig, S., Brewah, A., Woods, R., Lehlry, T., and Jacobsen, S. 1993.
High human T celi lymphotropic virus type 1 (HTLV)-specific precursor cytotoxic T
lymphocyte fieclyencies in patients with HTLV-1-associated neurological disease. J. Erp.
Med. 177, 1567-1573,
Elroy-Stein, O., Fuerst, T., and Moss,B. 1989. Capindependent translation of mRNA
conferreci by encephalomyocarditis virus 5' sequence improves the performance of the
vaccinia vimlbacteriophage T hybrid expression system. Proc. Natl. Acad. Sci. USA.
86, 6126-6130.

Estes, M., Crawford, S., Penaranda, M., Petrie, B., Burns, J., Chan, W-K.,Ericson,
B., Smith, G., and Summers, M. 1987. Synthesis and immunogenicity of the rotavirus
major capsid antigen using a bacdovirus expression system. J. Wrof.61,1488-1494.
Faller, D., Crimmins, A., and Menaer, S. 1988. Human T-ceil leukaexnia virus type 1
infection of CM' or CDS' cytotoxic T-cell clones reults in immortalization with
retention of antigen specificity. Krol. 62, 2942-2950.

'

Felber, B., Paskalis, H., Wong-Staal, F., and Pavlakis, G. 1985. The pX protein of
HTLV-I is a transcriptional activator of its long terminai repeats. Science. 229,675-679.
Fenouillet, E., and Gluckrnan, 3.-C. 1991. Effect of a glucosidase inhibitor on the
bioactivity and immunoreactiivty of human immunodeficiencey v h s type 1 envelope
glycoprotein. J. Gen. Wrol. 72, 1919-1926.
Fischinger, P.. Schafer, W., and Bolognesi, D. 1976. Neueralization of homologous and
heterologous oncomavinises by anrisera against the pl5 (E) and gp71 polypeptides of
Friend murine leukemia virus. Vkology. 71, 169-184.

Ford, C., Arp, J., King, E., Dekaban, GA., and Palluer, T. 1991. Expression and
immunogenicity of HTLV-1 envelope proteins by recombinant vaccinia virus. In Vaccines
91: Modem Apprwches to New Vaccines, pp.253-258. Edited by R.M.Chawck. Cold
Spring Harbour Laboratory, New York.
Ford, C., Arp, J., Palker, T., King, E., and Dekaban, G. 1992. Characterization of the
anti'body response to three dBerent versions of the HTLV-1 envelope protein expressed
by recombinant vaccinia vinises: Induction of aeuaazing a n t i i y. Virology. 191,448453.
Foung, S., and PedMs, S. 1989. Elecric field-induced cell fusion and human
monoclonal antibodies. J. Immunol. Merhods. 116, 117-122.
Franchini, G., Wong-Staal, F., and Gallo, R 1984. Molecular studies on human T ce11
leukemia v i n s and adult T ceU leukemia. J. Imest. Dermatol. 83, S63-S66.
Franchini, G., Mulloy, J., Korainik, I., LoMonico A., Sparkowski, J., Andresson, T.,
Goldstein, D., and Schlegel, R 1993. The human T-ceiileukemia/lymphotropic Wus
type 1 p12l protein cooperates with the ES oncoprotein and binds the ldkilodaiton
subunit of the vacuolar HfAnase. J. Virol. 67, 7701-7704.
Franchini, G., Tartaglia, J., Markham. P., Benson, J., Fullen, J., Wills, Arp, I.,
Dekaban, G., Paoletti, E., and Gallo, R. 1995. HighLy attenuated HTLV type I,,
poxvims vaccines induce protection against a ceil-associatecl HTLV type 1challenge in
rabbits. AlDS Res. Hum. Retrovimes. 11, 307-3 13.
Franchini, G. 1995. Molecular mechanisms of human T-cell leukenlia/ lymphotropic
virus type I iafection. Blood. 86, 3619-3639.
Fuerst, T., Niles, E., Studier, F., and Moss, B. 1986. Eukaryotic transient-expression
system baseci on recombinant vaccinia virus that syntheskes bacteriophage T7 RNA
polymerase. Proc. N i l . Ac&. Sei. USA. 83, 8122-8126.
Fueat, T., h l , P., and Moss, B. 1987. Use of a hybrid vaccinia virus-T7 RNA
polymerase system for expression of target genes. Mol. Ce11 Biol. 7, 2538-2544.
Fuerst, T., and Moss, B. 1989. Structure and stability of mRNA synthesized by vaccinia
virus-encoded bacteriophage T7 RNA polymefase in mammalian cells. Importance of the
5' untranslated leader. J. Mol. Biol. 206, 333-348.
Fujii, M., Sassotte-Corsi, P., and Verma, 1. 1988. C-fos promoter transactivation by the
taxl protein of human T-celi leukemia virus type 1. h c . Nml. Ac&. Sci. USA. 85,
8526-8530.

Fukasawa, M., Tsujimoto, H., Ishhwa, K., Miura, T., Ivanoff, B., Cooper, R., Frost,
E., Delaprote, E., Mingle, J., Grant, F., and Hayami, M. 1987. Human T-ceU leukemia
virus type 1 isolates from Gabon and Ghana: Comparative analysis of proviral genomes.
Virology. 161, 315-320.
Fultz, P., Nara, P., Barre-Simussi, F., Chaput, A., Greenberg, M., Muchmore, E.,
Kieny,M.-P.,and Girard, M. 1992. Vaccine protection of chimpanzees against challenge
with HIV-1-infecteci peripheral blood monomclear cells. Science. 256, 1687-1689.
Funikawa, K., Funikawa, K., and Shih, H. 1991. Alternatively spliced M A of the
pX region of human T-lymphotropic virus type I proviral genome. FHlS Lm.295, 141145.

Gaiio, R., Sliski, A., and Wong-Staal, F. 1993. Ongin of human T-cel leukaemialymphoma virus. Luncet. 2, %2.
Gatei, M.,Naif, H., Kumar, S., Boyle, D., Daniel, R., Good, M., and Lavin, M. 1993.
Protection of sheep agahst bovine leukemia virus (BLV)infation by vaccination with
recombinant vacciaia viruses expressing BLV envelope glycoproteins: correlation of
protection with CD4 T-celi response to gp51 peptide 51-70. J. Vrol. 67, 1803-1810.

Gessain, A., Barin, F., Vernant, J., Gout, O., Maurs, L., Calender, A., and de Th,
G. 1985. h t i i i e s to human T-lymphotropic vius type 1in patients with tropical spastic
paraparesis. Lancer. 5,407-410.
Gessain, A., Saal, F., Giron, M., Lasneret, J., Lagaye, S., Gout, O., de Th, G.,
Sigaux, F., and Peries, J. 1990. Cel surface phenotype and human T lymphotropic virus
type 1 antigen expression in 12 T c d LUies derived from peripheral blood and
cerebrospinal fluid of West Indian, Guayanese and A6eica.n patients with tropical spastic
paraparesis. J. Gen. Virol. 71, 333-341.
Gessain, A., Yanagihara, R., Franchini, G., Garruto, R., Jenkins, C., Ajdukiewicz, A.,
Gallo, R., and Gajdusek, D. 1991. Higbiy divergent molecular variants of human Tlymphotropic virus type I from isolated populations in Papua New Guinea in the Solomon
Islands. Proc. Nal. Acad. Sei. USA. 88, 76947698.
Gessain, A., and Gout, 0. 1992. Chronic myelopadiy associated with human Tlymphotropic virus type I (HTLV-I). Rnn. Intem. Med. 117, 933-946.
Gessain, A., Gallo, R., and Franchini, G. 1992a. Low degree of human T-ce11
leukemia/lymphoma virus type 1 genetic drift in vivo as a means of monitoring viral
transmission and movement of ancient human populations. J. Virol. 66, 2288-2295.

Gessain, A., Boer, E., Yanagiara, R., GaiIo, R., aad Franchini, G. 1993. Complete
nucleotide SeQuenceof a highiy divergent human T-ceii leukemia (lymphotropic) virus
type 1 (HTLV-T) variant nom Melanesia: Genetic and phylogenetic relationship CO
HTLV-1 strains h m other geographicai regions. J. WroL 67, 1015-1023.
Geyer, H., Holschbach, D., Hunsmann, G., and Schneider, J. 1988. Carbohydrates of
human immunodeficiency Wus: structures of oiigosaccharides linked to the envelope
glycoprotein 120. J. Biol. Chem. 263, 11760-11767.
Giavedoni, L., Planelies, V., Haigwood, N., Ahmad, S., Kluge, J., Marthas, M.,
Gardner, M., Luciw, P., and Yilma, T. 1993. Immune response of rhesus macaques to
recombinant simian immunodeficiency vinis a130 does not protect fiom challenge
infection, J. Wrol. 67,577-583.
Girard, M.,Kieny, M-P., Pinter,A., Barre-Siwussi, F., Nara, P., and Kolbe, H. 1991.
Immunisation of chimpanzees confers protection against challenge with human
immunodeficiency Vinis. hoc. Nml.Acad Sei. USA. 88, 542-546.

Giri, A., Maricham, P., Digilio, L., Hurteau, G., Gallo, R., and Francbini, G. 1994.
Isolation of a novel simian T-cel lymphotmpic virus fkom Pan paniscus that is distantly
related to the human T-cellleukemia/iyrnphoaopic virus types 1 and II. J. Viral. 68,
8392-8395.

Glasgow,L. 1970. Cellular immmity in host resistmce to viral infections.Arch Intern.

Med. 126, 125-134.

Goebels, N., Waase, 1.. Pfizenrnaier, K., and Kronke, M. 1988. IL-2 production in
human T lymphotropic virus I-infected leukemic T lymphocytes analyzed by in situ
hybridization. J. Inununol. 141, 1231-1235.

Gooding, L., and Edwards, C. 1980. H-2 antigen requirements in the in vino induction
of SV4-specific cytotoxic T lymphocytes. J. Immunol. W. 1258-1262.
Goubau, P., Liu, H., DeLange, G. Vandamme, A., and Desmyter, J. 1993. H n V - I I
seroprevalence in pygmies across Africa since 1970. AZDS Res. Hm. Retrovinises. 9,
709-713.

Goubau, P., VanBrussel, M., VandaInme, A.-M., Liu, H.-F., and Desmyter,J. 1994.
A primate T-lymphotropic virus, PTLV-L,different from human T-lymphotropic viruses
types 1and LI, in a wild-caught baboon (Pqpio hamadrym). Prcc. Natl.Acad. Sci. USA.
91, 2848-2852.

Gout, O., Gessain, A. and Mt. Sal. Iba-Zizen. 1991. The effect of zitovudine on chronic
myelopathy associated with HTLV-1. J. Neurology . 238, 108-109.

Grabstein, K., Eisenman, J., Sbanebeck, K., Rauch, C., Srhivasan, S., Fung, V.,
Beers, C., Richardson, J., Schoenborn, M., Ahdieh, M., Johnson, L., Alderson, M.,
Watson, J., Anderson, D., and Giri, 1. 1994. Cloning of a T-celi growth factor that
interacts with the /3 chain of the interleukin-2 receptor. Sdence. 264, %5-%S.
Graziani, G., Faraoni, L., Zhand, H.,Caronti, B., L a m , G.,Boamassar, E., and
Macchi, B. 1993. Transient HTLV-I infection of a human @orna celi k e foliowing cellfiee exposw. Virology. 197, 767-769.

Gray, G., White, M.,Barmian, T., and Mann, D. 1990. Envelope gene sequence of
HTLV-1 isolate MT-2 and its cornparison with other HTLV-1 isolates. Wrology. 177,
391-395.

Greene, W., honard, W., Wano, Y., Svetiik, P., Peffer, N., Sodroski S., Rosen, C.,
Goh, W., and Haseltine, W. 1986. Trans-activator gene of HTLV-II induces IL-2
receptor and IL-2 cellular gene expression. Science. 232, 877-80.
Guirakhoo, F., He&, F., and Kun, C. 1989. Epitope mode1 of tick-borne encephalitis
envelope glycoprotein E: analysis of structural properties, role of carbohydrate side chah
and conformational changes ocuring at low pH. Krulogy. 169, 90-99.
Hadlock, K., Goh, C., Bradshaw, P., Perkins, S., Lo, J., Kaplan, I., Khabbaz, R., and
Foung, S. 1995. Delineation of an irnmunodomioant and HTLV-specinc epitope within
the HTLV-1 transmembrane glycoprotein. Blood. 86, 1392-1399.
Haigwood, N., Barker, C., Higgins, K., Skiles, P., Moore, G.,Mann, K., Lee, D.,
Eichberg, J., and Steimer, K. 1990. Evidence for neutraliziag antiboies directed against
confonnational epitopes of IW-l gp120. In Vaccines W: modern appmoches to new
vaccines including prevemion of-,
pp. 313-320. Edited by F. Brown, R. Chawck,
H. Ginsberg and R. Lerner. Cold Spring Harbor Laboratory, Cold Sprhg Harbor, New

York.
Haigwood, N., Shuster, J., Moore, G., Lee, H., Skiles, P., Higgias, K., Barr, P.,
George-Nascimento, C., and Steimer, K. 1990a. Importance of hypervariable regions of
HIV-1 a 1 2 0 in the generation of virus neuttalizing antibodies. AIDS Res. Hunt.
Retroviflcses. 6 , 855-869.
Haigwood, N., Nara, P., Brooks,E., van Nest, G., Ott, G., Higgins, K., Dunlop, N.,
Scandella, C .,Eichberg,J., and Steimer, K. 1992. Native but not denaturd recombinant
human immuodeficiency virus type 1 gp120 generates broad-specmim neutralization
antibodies in baboons. J. H r d . 66, 172-182.

Hanly, S., RimsLy, L., Malim, M., Kim, J., Hauber, J., DucDodon, M., Lee, S.,
Maizel, J., Cuilen, B., and Greene, W. 1989. Comparative anaiysis of the HTLV-I rex
and HnT-l rev trans-regdatory proteins and their RNA response elements. Gmes. Dm.
3, 1534-1544.

Hara, H., Morita, M., Iwaki, T., Hatae, T. Itoyama, Y., Kitamoto, T., Akiniki, S-1,
Goto, I., and Watanabe, T. 1994. Detection of human T lymphotropbic v h s type 1
(HTLV-I) provirai DNA and aaaiysis of T ceil receptor V-eta CDR3 sequences in spinal
cord lesions of IFIZV-1-associated myelopathy/tmpid spastic paraparesis. J. Exp. Med.
180, 831-839.
Harding, C., Roof, R., Men, P., and Unmue, E. 1991. EfEects of pH and
polysaccharide on peptide binding to class II major histocompatability complex
molecules. Proc. Natl. Acad. Sci. USA. 88, 2740-2744.

Hatakeyama, M., Kono, T., Kobayashi, N., Kawahara, A., LRvin, S., Permatter, R.,
and Taniguchi, T. 1991. Interaction of the IL-2 receptor with the sarc-family ~ 5 6 ' ~ :
Identification of novel intermo1ecuia.r association. Science. 252, 1523-1528.
Hatton, T., Uchiyama, T
y Toibana, T., Takatsuki, K., and Uchino, H. 1981. Surface
phenotype of Japanese adult T ceIi leukaemia cells characterised by monoclorial
antibodies. Bload. 58, 645647.
Hayden, D., Baker, N., Percival, M., and Beckwith, P. 1986. Modification of the
Photosystem II light-harvesting chlorophyll ah protein complex in maize during chiilinduced photoinhiiition. Biochim. Biophys. Acta. 851, 86-92.
Hayder, H., and Blanden, R. 1994. Role of CDS+ cytotoxic T ceils in vital iaf6ctions.
In Smitegies of vaccine design, pp.69-82. Edited by G. Ada. R.G. Landes Co., Texas.
Hidaka, M., h u e , J., Yoshida, M., and Seiki, M. 1988. Post-transcriptional regulator
(rex) of HTLV-I initiates expression of viral structural proteins but suppresses expression
of regulatory proeins. EMBO J. 7, 519-523.

Hino, S., Doi, H., and Yoshikuni, H. 1986. HTLV-1 carrier mothers with high titer
antibody are at high risk as a source of iafction. Jpn. J. Cancer Res. (Gann). 78, 11561158.
Hino, S. 1990. Matemal-infant transmission of HTLV-1: implication for disease. In
H u m n RetroMrology: HIZV, pp.363-375. Edited by W. A. Blattner. Raven Press, New
York.

Hinuma, T., Komoda, H., and Chosa, T. 1982. Anti'bodies to aduit T-ceii leuicemia
virus-associated antigen (ATLA) in sera h m patients with ATL and controis in Japan:
a nation-wide sero-epidemiological study.?nt. J. Cancer. 29, 631-635.
Hirai, H., Fujisawa, J., Suzuki, T., Ueda, LC., Muramatsu, M.,Tsuboi, A., Ami, N.,
and Yoshida, M. Trarscriptional activator Tax of HTLV-1 binds to the NF-kappa B
ptecursor. Oncogene. 7, 1737-1742.

Hiramatsu, K.,Masuda, M., and Yoshikura, H. 1986. Mode of transmission of human

T cei leukaemia virus type 1 (HTLV-I) in a human promyelocytic leukaemia HL60 c d .
Int. 3. Cancer. 37, 601-606.
Hirsch, R. 1982. The complement system: its importance in the host response to vixai
infection. Micmbiol. Revews. 46, 71-85.
Hjelle, B., Appenzelien, O., Milis, R., Alexander, S., Torrez-Matinez, N., Jahnke, R.,
and Ross, G. 1992. Chronic neurodegenerative disease associatecl with HTLV-II
infection. Lance?. 339, 66646.
Ho, D., McKeating, J., Li, X., Moudgil, T.,Daar, E., Sun, N.-C., and Robinson, I.
1991a. Conformationai epitope on gp120 important in CD4 binding and human
immunodeficiency virus type 1 neutraIization identified by a human monoclonal antiaody.
J. Virol. 65, 489-493.

Ho, D., Fung, M., Cao, Y., Li, X., Sun, C., Chang, T., and Sun, N.-C. 1991b.
Another discontinuous epitope on gp 120 that is important in human immunodeficiency
virus type 1 neutralization is identified by a monoclonal anti'body. Proc. N d Ac&. Sci.
USA. 88, 8949-8952.
Hoffman, P., Cimino, E., and R o b h , D. 1991. Effects of viral spe~ificcytotoxic
lymphocytes on the expression of murine leukemia virus induced neurologie disease. J.
Neuruimmunol. 33, 157-165.
Hollsberg, P., and Hafier, D. 1993. Sem+
in medicine of the Beth Israel Hopsitai,
Boston: Pathogenesis of dwases induced by human lymphotropic virus type 1 infection.
N. Engl. J. Med. 382, 1173-1182.
Horal, P., Ha,W., Svennerholm, B., Lycke, J., Jeansson, S., Rymo, L., Kaplan,
M.H., and Vahlne, A. 1991. Identification of type-specific linear epitopes in the
glycoproteins gp46 and gp21 of human T-ceIl leukemia vmes type I and type II using
synthetic peptides. Proc. N d A c d Sci. USA. 88, 5754-5758.

Hoshino, H., Shimoyama, M., Miwa, M., and Sugimura, T. 1983. Detection of
lymphocytes produchg a human retrovirus asssociated with adult T-ceU leukemia by
syncytia induction assay. hoc. N d Ac& Sci. UU.80, 7337-7341.
Hoshino, H., Tanaka, H., Shimotohao, K., Miwa,M., Nagai, M., Shimoyama, M., and
Supimura, T. 1984. Imrnortalization of peripheral blood lymphocytes of cats by human
T-cell leukemia virus. I i J. Cmcer. 34, 513-517.
Hoxie, J., Matthews, D., and Cines, D. 1984. Infetion of human endothelial cells by
human T-celi leukaernia virus type 1. Proc. N i l . Acad. Sci. USA. 81, 7591-7595.

Hu, A., C a m o , R., Schwartz, S.. and Norrby, E. 1994. Role of N-Linked
oligosaccharide chains in the processing and antigenicity of measies virus haemagglutinin
protein. J. Gm. Virol. 75, 1043-1052.

Hu,S., Kosowski,

S., and Schaaf, K. 1987. Expression of envelope glycoproteins of

human imrnunoeficiency virus by an h e c t virus vector. J. Virol. 61, 3617-3620.

Hu,S-L., Abram, K., Barber, G., Moran, P., Zarlhg, I., Langlois, A., Kuller, L.,
Morton, W., and Benveniste, R. 1992. Protection of macaques against SIV infection by
subunit vaccines of SIV envelope glycopmtein gp160. Science. 255, 456-459.
H u n s m a ~ G.,
, Moennig, V., Pister, L., Seifert, E., and Schafer, W. 1974. Properties
of mouse leukemia vinises. W.
The major viral glycoprotein of Friend leukemia virus.
Seroimmunological, interferhg a d hernagglutinating capacities. Virology.62, 307-3 18.
H u n s m a ~ ,G., Moemig, V., and Scbafer, W. 1975. Properties of mouse leukemia
viruses. K. Active and passive immunization of mice against Friend leukemia with
isolated viral gp7 1glycoprotein and its comsponding antiserum. Krology -66, 327-329.
Ibrahim, F., Fiette, L., Gessain, A., Buisson, N., De Th, G., and Bomford, R. 1994.
Infection of rats with human T-cell leukemia virus type-1: Susceptiiiiity of inbred strains,
antibody respome and provirus location, Iw. J. Cancer. 58, 446-451.

Ichimani, M., Ikeda, S., Kinoshita, K., Hino, S., and Tsuji, Y. 1991. Mother-to-child
transmission of HTLV-1. Cancer Detect. Prev. 15, 177-181.
Ijichi, S., Matsuda, T., Mamyama, I., Inimibara, T., Kojima, K., Niimura, T.,
Maruyama, Y., Sonoda, S., Yoshida, A., and Osame, M. 1990. Arthritis in a human T
lymphotropic virus type (HTLV-1) c&r. AM. Rhem. Dis. 49, 718-721.

h u e , J., Seike, M., Taniguchi, T., Tsum, S., and Yoshida, M. 1986. Induction of
interleukin 2 receptor gene expression by p40x encoded by human T-cell leukemia virus
type 1. E-O
J. 5, 2883-2888.

houe, I., Ltoh, M., Akizawa, T., Toyosbima, A.,and Yoshida, M. 1991. HTLV-1 Rex
protein accumulates unspliced RNA in the nucleus as weii as in cytoplasm. Oncogene.
6, 1753-1757.
Ishiguro, N.. Abe, M., Seto, K., Sakurai, H., Ikeda, H., Wakisaka, A., Togashi, T.,
Tateno, M., and Yoshri, T. 1992. A rat mode1 of humaa T lymphocyte vins type 1
(HTLV-1) iofection. 1. Humoral antibody respoiw, provirus integration, and HTLV-1
associated myelopathyftropical spastic paraparesis-iike myelopathy in seronegative
HTLV-1 d e r rats. J. Exp. Med. 176, 981-989.
Ishihara, D .,Miyazawa, M., Nishio, J., and Chesebro, B. 1991. Induction of protective
immunity to friend muririe leukemia v i m in genetic nonresponders to virus envelope
protein. J. Imnunol. 146, 3958-3963.

Ishioka, G., Lamont, A., Thomson, D., Buibow, N., Gaeta, F., Sette, A., and Grey,
H. 1992. MHC interaction and T celi recognition of carbobydrates and glycopeptides.
J. Imm~(n01.14%.2446-2251.
Issel, D., Horohov, D., LPa, D., Adams Ir., W.,Haglus, S., McMarnis, J., AUison, A.,
and Montelaro, R. 1992. Efficacy of inactivatecl whole-viras and subunit vaccines in
preventing infection and disease caused by equine infectious anemia virus. J. YNol. 66,
3398-3408.

Itomaya, Y., Minato, S., and Kim, J. 1988. Spontaneous proliferation of peripheral
blood leukocytes increased in patients with HTLV-I associated myelopathy. Neurofogy.
38, 1302-1307.
Iwasaki, 1. 1990. Pathology of chronic myelopathy associateci with HTLV-1 infetion
(HAWTSP). J. Neurol Sei. 96, 103-123.

Jackson, C., Dnunmer, H., Urge, L., &os, J., and Brown, L. 1994. Glycosyiation of
a synthetic peptide repmnting a T-ceii determinant of infiuenza virus hemagglutinin
results in loss of recognition by C M + T-cell clones. Virology . 199, 422-430.
Jacobson, S., Reuben, J., Streilein, R., and Palker, T. 1991. Muction of CD4+ human
T lyrnphoaopic Wus type-1-specific cytotoxk T lymphocytes from patients with
HAMfTSP. J. Immunol. 146, 1155-1162.
Jacobson, S., LRhky, T., Nishimura, M., Robinson, S., McFatlin, D., and Dbib-Jalbut,
S. 1993. Isolation of HTLV-II from a patient with chronic progressive neurological
disease clinically indistinguishable fkom HTLV-1 associated myelopathy tropical spastic
paraparesis. Ann. Neurol. 33, 392-396.

Jacobson, S. 1994. Retrovinises and Cancer: US.-Japan ciinicwpidemiologicai

experiences d e r the auspices of the U.S .-lapan bilateral agreement. Leukma. 8,694704.
Jeang, K-T, Homgren-Konig, M., and Khoury, G. 1987. A bacdovirus vector can
express intron-containing genes. J. Virol. 61, 1761-1764.
Jeang, K-T., Widen, S., Semmes IV, O., and Wiison, S. 1990. HTLV-I trans-activator
protein, tax, is a trans-repressor of the hunian beta-polymemse gene. Science. 247, 10821084.

Johnson, D., Gautsch, J., Sportmian, J., and Elder, J. 1984. Improved technigue
utilipng nonfat dry milk for anaIysis of proteins and nucleic acids transferred to
nitroceilulose. Cen. Analyt. Techn. 1, 3-8.

Kajiyama, W., Kasbiwagi, S., Ikematsu, H., Hayashi, J., Homuta, H., and Okochi, K.
1986. Intra-familial transmission of adult T celi leukemia virus. J. Infect. Dis.154, 851857.

Kaiuza, G., Rott, R., and Schwarz, R. 1980. Carbohydrate-induced coafonnational

changes of Se&
forest virus glycoproteins detenniae anitgenicity. Virology. 102,286299.

Kalyanaraman, V., Sarangadharan, M.,Miyoshi, L, Blayney, D., Golde, D., and Gallo,
R. 1982. A new subtype of human T-cell leukemia virus (HTLV-II) associated with a TceIl variant of Hairy ceii leukemia. Science. 218, 571573.
Kanamori, H., S d , N., Siomi, H., Nosaka, T., Sato, A., Sabe, H., Hatanaka, M.,
and Honjo, T. 1990. HTLV-1 p27 rex stabiizes human intedeukin 2 receptor cr chah
mRNA. EMBO J. 9, 41614166.
Kang, C.-Y., Nara, P., Chamat, V., Caraili, V., Ryskamp, T., Haigwood, N.,
Newman, R., and Kohler, H. 1991. Evidence for non-V3 specific neutralizing antibodies
in HIV-1 infeted humans which interfere with gpl20/CD4 binding. Proc. Natl. Acad.
Sei. USA. 88, 6171-6175.

Kaanagi, M., Sugamura, K., Hirouki, S., Okochi, K., Uchino, H., and Hinuma, Y.
1983. Establishment of human cytotoxic T ceil liiies specific for human adult T ce11
leukemia virus bearing ceiis. J. ImmunoI. 130, 2942-2946.
Kaplan, J., and Osame, M. 1990. The risk of development of HTLV-1 associated
myelopathy/tropical spastic paraparesis among persom infected with HTLV-1. J. AIDS.
3, 1096-1101.

Kataoka, R., Takehata, N., Iwahara, Y., Sawada, T., Ohtsuki, Y., Dawei, Y., Hoshino,
H., and Miyoshi, 1. 1990. Transmission of EITLV-I by blood traosfusion and its
prevention by passive i m m d t i o n in rabbi&. Blood. 76, 1657-1661.
Kazedi-Kayembe, Goubau, P., Desmyter, L, W i e k k , R., and Carton, H. 1990, A
cluster of HTLV-1 associateci tropical s p d c paraparesis in Equateur (Zaire): Ethnic and
familial distri"bution. J. Neurol. Neurosurg. Pgch. 53, 4-10.
Keil, W., and Wagner, R. 1989. Epitope mapphg by deletion mutants and chimeras of
two vesicular stomatitis virus glycoprotein genes expressecl by a vaccinia virus vector.
Virology. 170, 392-407.
Kermode, A., Rudge, P., Thompson, A. Du Boulay, E., and McDonald, W. 1990. MRI
of thoracic cord in tropical spastic paraparesis. J. Neurol. Neurosurg. Psychiatry. 53,
1110-1111.

Kieny, M., Lathe, R., Riviere, Y., Dott, K,,Schmitt, D., Girard, M.,Montagnier, L.,
and Lecocq, J.-P. 1988. Irnproved antigenicity of the HN env protein by cleavage site
removal. Prot. Engin. 2, 219-225.

Kim,S., Kehrl, J., m o n , J., Tender, C., Jeang, K., Danielpour, D.,nievenin, C.,
Kim, K.,Spom, M.,and Roberts, A. 1990. Tramactivation of the traosforming growth
factor 81 (TGF-61) gene by human T lymphotropic nnis type 1 Tax: A potential
mechanism for the increase production of TGF-pl in aduit T cei leukemia. J- Exp.
Med. 172, 121-129.
Kinney-Thomas, E., Weber, J., McClure, J., Clapham, P., Singhal, M., Shriver, M.,
and Weiss, R. 1988. Neutralizing monoclonal altibodies to the AIDS virus. J. AIDS. 2.
25-29,

Kinoshita, K., Amagasaki,

.
. T., Hiw, S., Doi, H., Yamanouchi, K., Ban, N., Momita,
S., Ikeda, S., Kamihira, S., Ichimani, M., Katamine, S., Miyamoto, T., Tsuji, Y.,
Ishimani, T., Yamabe, T., Ito, M.,Kamura, S., and Tsuda, T. 1987. Milk-borne
Transmission of HTLV-1 fkom Carrier Mothers to Their Children. Jpn. J. of Cancer Res.
78, 674-680.
Kinoshita, T.,Tsujimoto, A., ami Shimotohno, K. 1991. Sequence variations in LTR and
env regions of HTLV-1 do not discriminate between the virus from patients with HTLV-1
associated myelopathy and adult T-cell leukemia. Int. J. Gncer. 47, 491-495.
Kim, J., Fujihara, K., Itoyama, Y., Goto, I., and Hasuo, K. 1991.
Leukoencelphalopathy in HTLV-1-associateci myelopathyltropical spastic paraparesis:
MRI analysis and a two-year foliow-up study after corticosteroid therapy. J. Neurol. Sci.
106, 41-49.

Kiyokawa, T., Yoshihua. H.,Hattori, S., Seiki, M., and Yoshida, M. 1984. Envelope
proteins of human T-celi leukemia Wus: Expression in EFclerichia coli and its
application to studies of env gene fiurctions. Proc. Nd.Acd. Sci. USA. 81,62024206.
Kiyokawa, T., Seki, M., Iwashta, S., Imagawa, K., Shimizu, F., and Yoshida, M.
1985. p27Xm and p21Xm, proteins encoded by the pX sequence of human T-cell
leukemia virus type 1. Proc. Nml.Acad. Sei. USA. 82, 8359-8363.

Klenk, H. 1980. Processing of the hemagglutlliin:Glycosylation amlproteolytic cleavage.

In Structure an variation in infruenza vina, pp. 213-222. Edited by W.Laver and G .
Air. Elsevier/North Holland, Amsterdam.
Koga, Y., Oh-Hori, N., Sato, A., Yamamoto, N-,Kimura, G., and Nomoto, K. 1989.
Absence of tmnscription of lck (lymphocyte specific protein tyrosine kinase) message in
IL-2-independent, HTLV-1 transformecl T cell lines. J. Invnunol. 142, 4493499.
Komurian. GoTPeiloquin, F., and de Th, G. 1991. In vivo genomic variability of
HTLV-1 depends more upon geography than pathologies. J. Virol. 65, 3770-3778.
Kondo, M., Takeshita, T., Ish, N., Nakamura, M.,Watanabe, S., Arai, K., and
Sugamura, K. 1993. Sharing of the interleukin 2 (IL-2) receptor 7 chah between
receptores for IL-2 and IL-4. Science. 262, 1874-1877.
Kondo, M., Takeshita, T., Higuchi, M., Nakamura, M., Sudo, T., Nishikawa, S., and
Sugamura, K. 1994. Functionai participation of the IL-2 receptor y chah in IL-7
receptor complexes. Science. 263, 1453-1454.

Koralnik, I., Gessain, A., Klotman, M., bMonico, A., Berneman, Z., and Franchini,
G. 1992. Protein isoforms encoded by the pX region of human T-ce11
leukemia/lymphotropic vims type 1. Proc. Nd. Aca. Sci. USA.89, 8813-8817.
Koralnik, I., Fullen, J., and Franchini, G. 1993. The p12', p13', and p30a porteins
encoded by human T a I l leukemianymphotropic virus type 1open reading frames 1 and
II are locazed in three different cellular compartments. I. Virol. 67, 2360-2366.

Koralnik, I., Boeri, E., Saxinger, W., LoMonico, A., Fullen, J., Gessain, A., Guo, H.G., Gallo, R., Markham, P.,Kalyanaraman, V., Hirsch, V., Man, J., Muahy, K..
Aiford, P., Slattery, J., O'Brien, S., and Franchini, G. 1994. Phylogenetic association
of human and simian T-celI leukemia/lymphotropic vins type 1 straias: Evidence for
interspecies transmission. J. Virol. 68, 2693-2707.

Korber, B., Okayama, A., Donneily, R., Tachiana, N., and Essex, M. 1991.
Polymerase chain reactioa analysis of d e f d v e human T-ceii leukemia virus type 1
provirai genomes in leukemic ceUs of patients with aduit T-ceil leukemia. J. Wrol. 65,
5471-5476.
Kornfeld, R., and Kornfed, S. 1985. Assembly of asparagine-linked ogosaccharides.
Ann. Rev. Biochem. 54, 631-664.
Koszhowski, U., Val, M., and Reddehase, M. 1990. CeMar and molecular basis of
protective immune response to cytomegalovirus infation. In QtomegaloMmes, pp. 189220. Edited by I. McDougaii. S p ~ g e rBerlin
,
Heidelberg, New York. (Current Topics
in Microbiology and h u n o l o g y , vol. 154)
Koszinowski, U., Reddehase, M., and Jonijc, S. 1991. The role of CD4 and CD8 T celis
in viral iafections. Curr. Opin. Immunol. 3, 471-475.
Kreth, H., Kress, L., Kress, H., Ott, H., and Eckert, G. 1982. Demonstriation of
primary cytotoxic T-ceUs in venous blood and cerebrospinal fiuid in children with mumps
meningitis. J. Invnunol. U8,2411-2415.
Krichbaum-Stenger, K., Poiesz, B., Keller, P., Ehrlich, G., Gavalchin, J., Davis, B.,
and Moore, J. 1987. Specific adsorption of RTLV-1 to various target human and animal
cells. Blood. 70, 1303-1311.
Kuga, T., Hattori, S., Yoshida, M., and Taniguchi, T. 1986. Expression of human Tceli leukemia vus type I envelope protein in Saccharomyces cermsiae. Gene. 44, 337340.

Kurata, A., Palker, T., Streilein, R., Scearce, R., Haynes, B., and Berzofsky, J. 1989.
Immunodominant sites on human T-cell leukemia virus type 1 envelope protein for
murine T-ceUs. J. Immunol. 143,2024-2030Kuroda, K., Hauser,C., Rott, R., Klenlc, H-D.,and Doefler, W. 1986. Expression of
the influenza virus haemagglutinin in insect cells by a baculovims vector. EMBO J. 5,
1359-1365.
Kuroda, K., Geyer, H., Geyer, R., Doerfler, W., and Kleuk, H.-D.1990. The
oligosaccharides of influenza hemagglutinin expressed in insect celis by a baculovirus
vector. Viroiogy. 174, 418-429.

Lairmore, M., Jacobson, S., Gracia, F., De, B., Castille, L., Larreategui, M., Roberts,
B., Levine, P., Blattner, W., and Kaplan, J. 1990. Isolation of human Tcell
lymphotropic virus type 2 from Guayrni I n d i i in Panama. Proc. N d Aca. Sei. USA.
87, 8840-8844.

Lairmore, M., Roberts, B., Frank, D., Rovnak, J., Weiser, M., aiidCockerell, F. 1991.
Comparative biologicai responses of rabbits infected with human T-lymphotropic virus
type 1isolates m m patients with 1ymphoproLiferativeand neurodegenerative disease. Int.
J. Chncer. 49, 1-7.
Lairmore, M., DiGeorge, A., Conrad, S., Trevino, A., Lal, R., aod Kaumaya, P. 1995.
Human T-lymphotropic vinis type 1 peptides in chimeric and multivaIent constnicts with
promiscuous T-ceil epitopes enhaoce immunogenicity ami overcome genetic restriction.
J. Virol. 69, 6077-6089.
Lagrenade, L., Harrhard, B., Fletcher, V., Cranston, B., and Blattner, W. 1990.
Infective dermatitis of Jamaican chiidren: a rnarker for HTL,V-1 iafection. Lancet. 336,
1345-1347.
Lal, R., Rudolph, D., Paiker, T., Coligan, J., and Folks, T. 1991. A synthetic peptide
elicits a n t i i y reactive with the native emrnai envelope glycoprotein of buman T-cell
lymphotropic virus Type 1. J. Gen. Viral. 72, 2321-2324.

La Rosa, G., Davide, J., Weinhold, K., Waterbury, J., ProQ, A. Lewis, J., Langlois,
A., Dreesman, G., Bosweil, R., Shadduck, P., Holley, L., Karplus, M., Bolognesi, D.,
Matthews, T., Emini, E., and Putney, S. 1990. CO EN^^ sequence and structural
elements in the HIV-1 principal neuaalinng detenninant. Science. 249, 932-935.
Lavanchy, D., Bovet, P., HoUaods, J., Shamlaye, C., Burczak, I., and Lee, H. 1991.
High seroprevalence of LFTZV-1 in the Seychelles. Lumet. 337, 248-249.
Lee, T., Coligan, J., Homma, T., McLane, M., Tachibana, N., and Essex, M. 1984.
Human T-ceii leukemia virus-associateci membrane antigeris: Identity of the major
antigens recognued after virus infection. Proc. Narl.Acad. Sci. USA.81,3856-3860.

Letkovits, I., and W d d m a ~ H.

, 1979. Limiting dilution anaiysis of cells in the immune
system. London, U.K.Cambridge Press.

Lem, T., Cowan, E., Lampson, L., ami Jacobson, S. 1994. Induction of HLA class
1 and class II expression in human T-lymphotropic virus type 1-infected neuroblastoma
cells. J. Virol. 68, 1854-1863.

L~hky,T., Fox, C., Koenig, S., Levin, M. Flerage, N., Izumo, S., Sato, E., Raine, C.,
Osame, M., anci Jacobson, S. 1995. Detection of human T-lymphottopic virus type 1
(HTLV-1) tax RNA in the central netvous system of HTLV-associated
myelopathyftropical spastic paraparesis patients by in situ hybridization. Ann. Neurol. 37,
167-175.

Lehman, D., Roof, L., Brideau, R., Aeed, P., Thomsen, D., Eihammer, A., Wathen,
M., and Homa, F. 1993. Cornparison of soluble and secreted f o m of human
parainfluema virus type 3 glycoproteirs expressed from mnmmalian and k t cells as
subunit vaccines, J. Gen. Virol. 74,459-469.
Lehner, T., Bergmeier, L., Panagiotidi, C., et ai. 1992. Induction of mucosal and
systemic immunity to a recombinant simian immunodeficiency viral protein. Science.
258, 1365-1369.
Li, Y., Luo, L., Snyder, R., and Wagner. R. 1988. Expression of the M gene of
vesicular stomatitis virus cloneci in various vaccinia virus vectors. J. Viral. 62, 776-82.
Li, Y., Luo, L., Rassol, N., and Kang, C. 1993. Glycosylation is necessary for the
correct folding of human irnmunodeficiency vins gp120 in CD4 binding. J. mol. 67,
584-588.

Lillehoj, E., Tal, C., Nguyen, A., and Alexander, S. 1989. Characterization of env and
tm encoded polypeptides of human TeU leukemia Wus type 1. Clin. Biotechn. 1,2741.

Lindholm, P., Marriott, S., Gitlin, S., Bohan, C., and Brady, J. 1990. Luduction of
nuclear NF-d3 DNA binding activity after exposure of lyrnphoid ceils to soluble Taxl
protein. New. Biol. 2, 1034-1043.
Link, H., Cruz, M., Gessain, A., Gout, O., de Th, G., ami Km-Hansen,S. 1989.
Chronic progressive myelopathy associated with HTLV-1: oligoclonal IgG and HTLV-1
IgG a n t i i i e s in cerebrospinal fluid and senim. Neurology . 39, 1566-72.

Liu, H.-F., Vandamme, A.-M., Van Brussel, M., Desmyter, J., and Goubou, P. 1994.
New retrovinses in hiiman and simian T-lymphotropic viruses. Lancet. 344, 265-266.
Longo, D., Gelmann, E., Cossman, J., Young, R., Gaiio, R., O'Brien, S., and Matis,
L. 1984. Isolation of HTLV-tmformed Blymphocyte clone from a patient witb HTLV
associated adult T-cell leukaemia. Nature. 310, 505-506.
toughran, T., Coyle, T., Sherman, M., Starkbaum, G., Ehrilich, G., Ruscetti, F., and
Poiesz, B. 1992. Detection of human T-celi leukemia/lymphoma vins type II in a patient
with LGL leukemia. Blood. 80, 1116-1119.

Lucisano Valim, Y., and Lachman, P. 1991. The effect of antibody isotype and antigenic
epitope density on the cornplement-fixing activty of immune complexes: a systematic
snidy using chimaeric anti-NP aatibodies with humaa Fc regions. Clin. Exp. Irnmunol.
84, 1-8.

Luckow, V., and S m e r s , M. 1988. Trends in the development of bacuiovirus

expression vector. Bio/Technolgy. 6, 47-55.
Macchi, B., Popovic, M., Allavena, P., Onaldo, J., Rossi, P., Gallo, R o y aod
Bonmassar, E. 1987. In MRo susepbcbility of different human T-ce subpopulations and
resistance of large granuar lymphocytes to EfTLV-1 infection. Im J. Cancer. 40, 1-6.
Mackett, M., Smith, G., an Moss, B. 1984. General method for the production and
selection of infectious vaccinia virus recombinants expressing foreign genes. J. Virol. 49,
857-864.
Mackett, M., Smith,G., and Moss, B. 1985. The construction and chatacterization of
vaccinia v i m recombinants expressing foreign genes. In DNA cloning, vol. 2, pp . 191211. Edited by D. RickwOOd and B.D. Hames. IRL Press, Washington D.C.
Mackow, E., Shaw, R., Matsui, S., Vo, P., Benfield, Dm, and Greenberg, H. 1988.
Characterization of homotypic and heterotypic neutralization sites of rhesus rotavirus.
Vir010g~.165, 511-517.
Mackow, E., Bamett, J., Chan, H.,and Greeaberg, H. 1989. The rhesus rotavirus outer
capsid protein VP4 functions as a haemaggiutinin and is antigenidy conserveci when
expressed by a bacuiovinis recombinant. J. Virol. 63, 1661-1668.

Manca, F . Habeshaw , J., ami Dalgleish, A. 1991. The naive repertoire of human T
cells specific for HIV envelope glycoprotein gp120. J. Immunol. 146. 1964-1969.
Manca, F. 1992. Galactose receptors and presentation of HIV envelope glycoprotein to
specific human T ceis. J. I'ol. 148, 2278-2282.
Manca, F., Dessi, V., Sbastri, N., Oki, A., Kaplan, M., Clarke,J., Miller, A., and
Sercarz, E. 1994a. Isolation of lysozyme specific T cell clones that discriminate between
native and denatured antigen. tell Immunal. 154, 420429.
Manca, F., Li Pira, G., Fenoglio, D., Sun, P., Habeshaw, A., Knight, S., and
Dalgleish, A. 1994b. Demiritic ceils are potent antigen presenting cells for NI &ru
induction of primary human CD4+ T ceil Lines specific for HIV gp 120. J. AIDS.7, 1523.
Manca, F., Fenoglio, D., Valie, M., Li Pira, G., KunL1, A., Ferraris, A., Saverino, D.,
Laacia, F., Mortara, L., Lazzi, L., Pierres, M., Dalgleisb, A., and Lewis, G . 1995a.
Human CD4+T cells can discriminate the molecular and structurai context of T epitopes
of HIV gp120 and HIV p66. J. AZDS. 9, 227-237.

Manca, F., Li Ph,G., Fenoglio, D., Valie, M., Kunkl, A., Ferraris, A., Mortara, L.,
Balderas, R., Arp, J., B a d a , R., Dekaban, G., Dalgleish, A., and Theofiiopoulos, A.
199%. Recognition of HTLV-I envelope protein by human CD4+ T celi lines generated
fiom HTLV-1 seronegative nividuals. Blood. 85, 1547-1554.

Manca, F., Fenoglio, D., VaLie, M., Li P h ,G., Kunkl, A., Ferraris, A., S a v e ~ oD.,
,
Lancia, F., Moriata, L., Lozzi, L., Pierres, M., Dalgleish, A., and Lewis, G. 19%.
Human CD4+ T cells c m dimirniaaie the molecular and structural context of T epitopes
of HIV gp120 and HIV p66. J. Acquir. Immune D@c- Syndt. Hm. Retrovirol. 9,227237.
Manns, A., and Blattner, W. 1991. The epidemiology of the human T-cell lymphotropic
virus type 1 and type II: etiologic role in human disease. Tramjkion. 31, 68-75.
Markham, P., Salahuddin, S., Kaiyaxmaman, V., Popovic, M., Satin, P., and GaiIo,

R. 1983. Infietion and tramformation of fresh human umbilical cord blood cells by
multiple sources of human T-ceU leukemia-lymphoma virus (HTLV). Int. J. Cancer. 31,
413-420.
Marrion, S., Lindhoh, P., Reid, R., and Brady, J. 1991. Soluble HTL,V-1 Taxl protein
stimulates proMeration of human peripheral blood lymphocytes. New Biol., 3, 678-686.

Mamion, S., Trinh, D., and Brady, J. 1992. Activation of interleukin 2 receptor alpha
expression by extraceiiular HTLV-1 Taxl protein: A potential role in HTLV-1
pathogenesis. Oncogene. 7, 1749-1755.
Maniyama, M., Shibuya, H., Harada, H., Hatakeyama, M., Seiki, M., Fujita, T.,
houe, J., Yoshida, M., and Taniguchi, T. 1987. Evidence for aberrant activation of the
interleukin-2 autocrine loop by HnV-1 encoded p40x and T3Ti complex t r i g g e ~ g .
Cell. 48, 343-350.

Marx, P., Pedersen, N., Lerche, N., Osbom,K., bwenstine, L., Lacher, A., Maul,
D., Kwang, H.-S., Kluge, J., Zaiss, C., Sharpe, V., Spinner, A., AUison, A., and
Gardner, M. 1986. Prevention of simian accpired immune deficiency syndrome with a
formalin-inactivatecl type D retrovirus vaccine. J. Wrof.60, 431435.
Mathews, J., Roehrig, J., and Trent,D. 1985. Role of complement and the Fc portion
of immunoglobulin G in immunity to Venezuelan equine encephalomyelitis virus infection
with glycoproteh-specific monoclonal aatibodies. J. Wrol. 55. 594-600.
Matsuuta, Y., Possee, R., Overton, H., and Bishop, D. 1987. Baculovinis expression
vectors : the requkements for hi@-level expression of proteins ,including glycoproteins.
J. Gen. Wrol. 68, 1233-1250.

Matsuura, Y., Mjamoto, M., Sato, T., Morita. C., and Yasui, K. 1989.
Characterizationof Japanese errephoitis virus envelope protein expressexi by recombinant
baculovinws. VIroIogy. 173, 674-682.
Matsushita, S., Robert-Gmff, M., Trepe!l, J., Cossman, I., Mitsuya, H., ami Broder,
S. 1986. Human monoclonal antiiiy directeci against an envelope glycoprotein of
human T-celi leukemia Wus type 1. PIOC. N d Acad. Sci. USA. 83, 2672-2676.
Matsushita, S., Robert-Guroff, M., Rusche, J., Koito, A., Hattori, T., Hoshino, H.,
Javaherian, K., Takatsuki, K., and Putney, S. 1988. Characterization of a human
immmodeficiency vinis neutralinng monoclonal a n t i i y and mapping of the
neutrazing epitope. J. WroL 62. 2107-2114.
Mazanec, M., Mazanec, M., Kaetzel, C., Lamm, M., et al. 1992. Intracellular
neutralizationof virus by immunoglobuiin A ant'bodies. Proc. Nd.Acad. Sci. USA. 89,
6901-0%5.

McCullough, K., Crowther, J., ami Butcher, R. 1986. Immune protection against footand-mouth disease vinis studied using virus-neutralinng and non-neutralizing
concentrations of monoclonal antl'bodies. I~lllt~(nology.
58, 421-428.
McGuire, K., Curtiss, V., Larson, E., and Haselthe, W. 1993. Influence of human Tceil leukemia vins type I tax and rex on interieukin-2 gene expression. J. Virol. 67,
1590-1599.

McKendali, R. 1985. IgG-mediated viral clearance in experwntai infection with herpes

simplex virus type 1: Role for neutraljzation and Fc-dependent fiinctions but not C'
cytolysis and CS chemoraxis. J. In$ Diseases. 151, 464-470.
Men, D., Scheid, A., and Choppin, P. 1980. Importance of antibodies to the fusion
glycoprotein of paramyxoviruses in the prevention of infection. J. Exp. Med. 151, 275288.

Meytes, D., Schochat, B., Lee, H., Naciel, G., Sidi, Y., Cerney, M., Swanson, P.,
Shaklai, M., Kilim, Y., and Elgat, M. 1990. Serological and molecular sucvey for
HTLV-1 infection in a high-risk Middle Eastern group. Luncet. 336, 1533-1535.
Michaelsson, E., Malmstrom, V., Reis, S., Engstrom, A., Burkardt, H., and Holmdahl,
R. 1994. T cell recognition of carbohydrates on type II collagen. J. W .
Med. 180, 745749.

Migone, T., Lin, J., Cereseto, A., Muiloy, J., O'Shea, J., Franchiai, G., and Leonard,
W. 1995. Codtutively activated Jak-stat pathway in T cells transformed with HTLV-1.
Science, 269, 79-81.

Milier, C.,Alexander, NoTand McChesney, M. 1994. Vaccines to prevent sexualy

transmitted diseases: The challenge of generating protective immunity at genital mucosal
surfaces. In Stmregies in Vuccinc Design, pp.193-212. Edited by O. Ada. R. G. Landes
Co., Texas.
Miller, S., Milier, L., Olson, C., and Guette, K. 1%9. Virus-like particles in
phytohaemagglun'nin-stimuiated lymphocyte cuinires with reference to bovine
lymphosarcoma. J. Nml. Concer lm. 43, 1297-1305.

Miller, J., Moraban, G., and AUWn, J. 1989. Tmmunologicai tolerance: New
approaches using transgenic mice. fmmunol. Today. 10, 53-57.
Miller, L. 1988. Badovinises as gene expression vectors. Ann. Rev. Microbiol. 42,
177-199.
Milis, G., Arima, N., May, C., Hill, M., Schmandt, R., Li, J., Miyamoto, N., and
Greene, W. 1992. Neither the lck nor the fyn kinases are obiigatory for IL-2mediated
signal transduction in EFnV-1-inf'ted human T-ceUs. Int. Immunoi. 4, 1233-1243.
Misuni, Y.,M i d , Y., Miki, K., Takatsuki, A., Tamura, G., and Ikehara, Y. 1986.
Novel blockade by Brefeldin A of intracelluiar transport of secretory proteins in cultured
rat hepatocytes. J. Biol. Chem. 261, 1l39-lI403.

Mitsuya, H., Matis, L., Megson, M., Bunn, P., Murray, C., Mann, D., Gailo, R., and
Broder, S. 1983. Generation of an HLA restricted cytotoxic T ceii line reactive against
cultured tumor ceils fkom a patient Mixteci with human T celi leukerniallymphomavirus.
J. Eqr. Med. 158, 994-999.
Miyamoto, C., Smith, G., Farrell-Towt, J., Chizzonite, R., Summers, M.,and lu, G.
1985. Production of human c-inyc protein in k t celis infected with a baculovirus
expression vector. Mol. Cell. B i d . 5, 2860-2865.
Miyatake, S., Seilci, M., Yoshida, M., and Arai, K.4. 1988. T-celi activation signals
and human T-ceII leukemia vKus type4 encodeci p40x protein activate the mouse
granulocyte-rnacrophagecolony-stimulatingfactor gene through a cornmon DNA element.
Mol. Ce11 Biol. 8, 55814587.

Miyazawa, M., Nishio, J., and Cheesebro, B. 1992. Protection against Friend retrovinisinduced leukemia by recombinant vaccinia viruses expressing the gag gene. J. Kroi. 66,
4497-4507.

Miyoshi, I., Taguchi, H., Fujishita, M., Yoshuwto, S., Kubonishi, 1.. Ohtsuki, Y.,
Shiraishi, Y.,d Akagi, T. 1982. Transformation of monkey lymphocytes with adult
T-ce11 leukaemia virus. Lance?. i, 1016.

Miyoshi, I., Yoshimoto, S., Fujishita, M., Taguchi, H., Kubomishi, 1.. Niya, R., and
Miwzawa, M. 1982. N a t d adult T-cell leukemia virus infation in Japanese rnonkeys.
Laneet. iib, 658.
Miyoshi, I., Yoshimoto, S., Taguchi, H., Kubonishi, M.,Fujishita, M., Ohtsuki, Y.,
Shiraishi, Y., and Akagi, T. 1983. TTaf3sfomntion of rabbit lymphocytes with T-celi
leukemia virus. Jpn. J. Concer Res. (GaM). 74, 14.

Minikami, T.,Fuerst, T.,Berger, A., and Moss, B. 1988. Binding region for human
immunodeficiency virus (HIV) and epitopes for HIV-blocking monoclonal anaMies of
the CD4 molecule defined by site-directedmutagenesis. Pruc. N d Aead. Sei. USA. 85,
9273-9277.
Mochiniki, M., Watanabe, T., Yamaguchi, K., Takatsuki, K., Shuao, M., Nakashima,
S., Mori, S., Araki, S., and Miyata, N. 1992. HTLV-1 uveitis: a distinct cluiical entity
caused by HTLV-1. Jpn J. of C ~ c eRes.
r
83, 236-239.
Moennig, V., Frank, H., Hurismami, G., Schneider, I., and Scbafer, W. 1974.
Properties of mouse leukemia viruses. W. The major viral glycoprotein of Friend
leukemia virus. Isolation ami physicochemical properties. Wogy . 61, 100-111.

Montgomery, R. 1983. HTLV-1 and tropical spastic paraparesis: Clinical features,

pathology and epidemiology. Trans. R Soc. Trop. Med. mg.t33,724.
Moore, J., McKeating, J., Jones, 1.. Stephens, P., Clements, G., Thompson, S., and
Weiss, R. 1990. Characterization of recombinant gp120 and gp160 fiom HN-1 binding
to monoclonal antifbodies and soluble C M .AIDS. 4, 307-315.
Moore, J., and Ho, D. 1993. A n t i i i e s io discontiouous or conformationally sensitive
epitopes on the gp120 giycoprotein of human immunodeficiency virus type 1 are highly
prevalent in sera of infccted humans. I. Viroi. 67, 863-875.
Morgan, B., and Walport, M. 1991. Complement deficiency and disease. Immunol.
To@. If, 301-306.
Moriyama, T., Guiihot, S., Klopchin, K., Moss, B., Pinkert, L., Palmiter, R., Brinster,
R., Kanagawa, O., and Chisari, F. 1990. The immunobiology and pathogenesis of
hepatoceliular injury in Hepatitis B tramgenic mice. Science. 248, 361-364.
Mosmann, T., and Cofnnan, R. 1989. TH1 and TH2 cells: Different pattern of
lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7, 145173.

Moss, B., and Flexner, C. Vaccinia Wus expression vectors. ANI. Rev. Immwiol. 5,
305-324.
Moss, B. 1990. PoxvKidae aod theu replication. In Fieldr wirology, 2nd edition, pp.
2079-21 11. Edited by B. Fields, D .Knipe, R. Chanock, et al. Raven Press, New York.
Muggendge, M., Roberts, S., Isola, V., Cohen, G., and Eisenberg, R. 1990. Herpes
simplex virus. In Iinnuinochemishy of Mruses, Volume ZI. nie & a i sfor serodagmsis arui
vaccines, pp. 459-48 1. Elsevier Science Pubkhers, Amsterdam.
Muiloy, J., Crowley, R., Fulien,
Leooard, W.,and Franchini, G. The human T-cell
Ieukemiallymphoma virus type 1p12' protein binds the IL,-2 receptor @and 7, chains and
affects their expression on the cell surface. J. Wrol. 70, 3599-3605.
J e y

Murphy, E., Figwroa, J., Gibbs, W., et al., Senial transmission of HTLV-1. Ann.
Intem. Med. 1989. 111, 555-560.
Murphy, E., Hatlchard, B., Figueroa, J., Gibbs, W., Lofters, W.,Campbell, M.,
Goedert, J., and Blattner, W. 1989. Modeling the risk of adult T-ce11
leukemiallymphoma in persons infected with human T-lymphotropic virus type 1. Iht. J.
Cancer. 43,250-253.
NagashMa, K., Yoshida, M., and Seiki, M. 1986. A single species of pX mRNA of
human T-cell leukemia virus type 1 encodes tram-activator p4V and two other
phosphoproteins. J. Wrol. 60,394-399.
Nagata, K., Ohtani, K., Nakamura, M., a d Sugamura, K. 1989. Activation of
endogenous c-fos proto-oncogene expression by human T-cell leukemia virus type Iencoded p 4 F protein in the human T-ceii l*, Iwkat. L W O ~
63,.
3220-3226.

Nagy, K., Clapham, P., Cheiosong-Popov, R., and Weiss, R. 1983. Human Tcell
leukemia virus type 1: Induction of syncytia and inhiiition by patient's sera. Inf. J.
Cancer. 32. 321-328.
Nakamura, H.,Hayami, M., Ohta, Y., Ishikawa, K., Tsujimoto, H., Kiyokawa, T.,
Yoshida, M., Sasagawa, A., and Honjo, S. 1987. Protection of cynornolgus monkeys
againfation by human T-celi leukemia vUus type-1 by immunization with viral env
gene products produwd in ~cherichiac d . I m J. Cancer. 40,403-407.
Nakamura, Y., Russell, S., Mess, S., Friedmann, M., Erdos, M., Francois, C.,
Jacques, Y., Adelstein, S., and Leonard,W. 1994. Heterodimerizationof the interleukin2 receptor 0 and a chah cytoplasmic domains is required for signaling. Narure. 369,
330-333.

Nara, P., and Goudsmit, J. 1990. Neutraikation-resistant variants of HIV-1 escape via
the hypervariable immunodominant V3 region: evidence for a confonnational
neutraiization epitope. In V i n e s PO: nodem apprwdies to new vaccines including
preveAon of ADS, pp. 297-306. Edited by R Brown, R. Cbanock, H.Ginsberg and
R. L e m . Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
Nelson, B., Lord, J., and Greenbetg, P. 1994. Cytoplasmic domains of the intetIeukin-2
receptor 6 and a chahs mediate the signal for TcelI proMeration. Nature. 369,333-336.
Nelson, N. 1989. Structure, molenilar genetics, and evolution of vacuolar H+ATPases.
J. Bioenerg. Biomembr. 21, 553-571.

Nermut, M.,Frank, H.,and Schafer, W. 1972. Pruperties of mouse leukemia vinises.

III. Electron microscopie appearance as revealed aftet conventional preparation
techniques as weli as freeze-drying and fkeeze-etchkg. Virology. 49, 345-358.

Nishida, J.9 Yoshida, M.,Arai, K., and Yokota, T. 1991. Definition of GC-rich motif
as regdatory sequence of the human IL-3 gene: Coordinate regdation of the IL-3 gene
by CLE2/GC box of the GM-CSFgene in T celi activation. Iht. Irmtuuloi. 3, 245-254.

Ng, S., G W g , P., Eddy, R., Ponte, P.. Leavitt, J., Shows, T.,and Kedes, L. 1985.
Evolution of the functional beta actin gene and its multi-pseudogene family: conservation
of non-coding regions and chromosomal dispersion of pseudogenes. Mol. Cell. Bio. 5,
2720-2732.
Noguchi, Y., Noguchi, T., Sato, T., Yakoo, T., Itoh S., Yoshida, M.,Yoshki, T.,
Akiyoshi, K., Sunamoto, J., Nakayama, E., and S b , H. 1991. Priming for in vitro
and in vivo anti-hurnan T lymphotropic Wus type 1 cellular inimunity by virus-related
protein reconstituted into liposome. J. Immunul. 146, 3599-3603.

Noraz, N.. Benichou, S., Madaule, P., Tioilais, P., Vernant, J., and Desgranges, C.
1993. Expression of HTLV-1 Env and Tax recombinant peptides in yeasts: Identification
of immunogenic domains. Wrology. 193, 80-88.
Nosaka, T., Ariumi, Y., Sakumi. M., Takeuchi, R., and Hatanaka, M. 1993. Novel
interna1 promoter/enhancer of HTLV-1 for Tax expression. Nucleic Acidr. Res. 21,51245129.
Nybor, J., Dyoan, W., Chen, I., and Wacbsman, W. 1988. Binding of host-celi factors
to DNA seyences in the long terxninai repeat of human T-celi leukemia virus type 1:
implications for vual gene expression. Proc. Nati. Acad. Sei. USA 85, 1457-1461.

Orita, S., Saiga, A., Takagi, S., Tanaka, I., O h m , A., Aono, Y.. Hinuma, Y., and
Igarashi, H. 1991. A novel altematively spced viral mRNA a n s c n i i in ceils infected
with human T-celi leukaemia virus type 1 is mainly tesponsible for expressing p21x
protein. FEBS L a . 295, 127-134.
Orita, S., Sato, S., Aono, Y., Minoura, N., Yamashita, T., Hinuma, Y., and Igaras,
H. 1993. Identification of novel singly spliced pX mRNA transcripts common to ail
human T-ceii leukemia virus type 1 related retroviruses. Viras Genes. 7, IW-2O4.
Osame, M., Igato, A., Matsumoto, M., Kohka, M., Usuku, K., and Izumo, S. 1980.
HTLV-1-associatedmyelopathy (HAM): treatment trials, retrospective survey and ciinical
and laboratory findings. Hem~tol-Rev. 3, 271-284.
Osame, M., Usuku,K., Izumo, S., Ijichi, N., Amitani, H., Igata, A., M a m o t o , M.,
and Tara, M. 1986. HTLV-1 Associateci Myelopathy, a new clinical entity. Lpncet. 1,
1031-1032.
Osame, M., Matsumoto, M., Usuku, K.,

Izumo, S., Ijichi, N., H. Amitani, Tara, M..

and Igata, A. 1987. Chronic progrssive myelopathy associated with elevated anti'bodies
to human T-lymphotrophic virus type 1and adult T-cellleukemialike ceus. A m . Neurol.
21, 117-122.
Osame, M., Igata, A., and Matsumoto, M. 1989. HTLV-1 associated myelopathy
revisited. In HTLV-l and the Nervous @stem. Newology and Biology 51, p. 213. Edited
by G. R o m , 1. Vernant, and M. Osame. Liss, New York.
Osterhaus, A., Weijer, K., Uytdehaag, F., Jarret, O., Sundquist, B., and Morein, B.
1985. Induction of protective immune response in cats by vaccination with feline
leukemia virus ISCOM. J. Immunol. 135, 591-596.
Page, J., Lai, S., Chitwood, D.Klimas, NoTSmith, P., and Fletcher, M. 1990. HTLVI/II seropositivity and death ftom AIDS among HIV-1 seropositive intravenous dmg
users. tancet. 335, 1439-1441.
Paine, E., Garcia, J., Philpott, T.,Shaw, G., a d Ratner, L. 1991. Limited sequence
variation in human T-lymphotropic virus type4 isolates from North Amencan and
Afncan patients. Virology . 182, 111-123.
Palker T., Clark, M., Samgadharan, M., and Matthews, T. 1987. Purification of
envelope glycoproteins of human T cell lymphotropic Wus type 1 (HTLV-I) by afflnity
chromatography. J. VNol. Mah. 18, 243-256.

Pique, C., P b , D.,Tursz, T., and Dokhelar, M.-C.1992. Human T-ceil LRukemia
Virus Type 1 envelope protein anturation process: Requirements for syncytium
formation. J. Wrol. 66,906-913.
Plwnmer, T., Jr., Elder, J., Alexander S., Phelan, A., and Tarentho, A. 19W.
Demonstration of peptide: N-glycosidase F activity in endo-B-N-aceiylglucoaminidaseF
preparations. J. Biol. Chem. 259, 10700-10704.
Poiesz, B., Ruscettti, F., Gazdar, A., Bunn, P., Minas, J., and Gallo, R. 1980.
Detection and isolation of type-C retrovirus particles fkom h h and culhired
lymphocytes of a patient with cutaneous T-ceii lymphoma. Proc. N d Acad. ScL USA.
77, 7415-7419.
Poiesz, B., Sherman, M., Sakseaa, N., Dube, D., hibe, S., Gavalchin, J., Fan, N.,
Lane, N., and Paul, B. 1993. The biology and epidemiology of the human T-celi
lymphomaiieukemia vhses. Ln Frontiers of Infectt*ousDiserses: Focw on IW,pp. 189205. Edited by H. Neu, J. Levy and R. Weiss. Churchill Livingstone, London.

Popovic, M., Sarh, P., Robert-Guroff, M., Kalyanaraman, V., Mann, D., Minowada,
J., and Gallo, R. 1983. Isolation and transmission of human retrovirus (human T ce11
leukaemia virus). Science. 219, 856-859.
Rademacher, T., Parekh, R., and Dwek, R. 1988. Glycobiology. AM. Rev. Biochem.
57, 785-838.
Rademacher, T. 1992. Therapeutic challenges: Does glycobiology have a role. Trends
Biotechnol. 10, 227-230.
Ralston, S., Hoeprich, P., and Akita, R. 1989. Identification and synthesis of the epitope
for a human monoclonal antibody which can neuPalUe human T-cell Lymphotropic virus
type 1. J. Biol. Chem. 264, 16343-16346.
Ramsay, A., Ruby, J., and Ramshaw, 1. 1993. A case for cytokines as effectors in the
resolution of virus infection. Imunol. Today. 14, 155-157.
Ratner, L., Josephs, S., Starcich B., Hahn,B., Shaw, G., Galio, R., and Wong-Staal,
F. 1985. Nucleotide secpence analysis of a variant Human T-cell Leukemia Virus
(HTLV-Ib) provirus with a deletion in PX-1. J. Wtol. 54, 781-790.

Ratner, L. 1989. Regulation of expression of the c-sis proto-oncogene. Nucieic Acids.

Res. 17, 41014115.

Rejifo, B., Borrero, I., and Essex, M. 1995. Tax mutation associated with tropical
spastic paraparesislhuman T-ceii leukemia virus type 1-associated myelopathy. J. WroL.
69, 2611-2616.
Rich, D.,Anderson, M., Gregory, R., Cheng, S., Paul, S., Jefferson, D., McCann, J.,
Knger, K., Smith, A., and Welsh, M. 1990. Expression of cystic fibrosis
transmembrane conductance regufator corrects defective chloride channel regdation in
cystic fibrosis airway epithelial ceiis. Noture. 347, 358.363.
Richardson, J., Edwards, A., Cruickshank, J., Rudge, P., and Dalgleish, A. 1990. In
vivo cellular tmpism of human T-ceil leukemia virus type 1. J. Yirol. 64, 5682-5687.
Roert-Guroff, M., Nakao, Y., Notake, K. Ito, Y., Sliski, A,, and Gallo, R. 1982.
Naturai antl'bodies to human retrovirus RTLV in a cluster of Japanese patients with adult
T cei leukemia. Science, 215, 975-978.

Rock, K., Rothstein, L., Gamble, S., and Fleischacker, C. 1993. Characterization of
antigen-presenting celis that present exogenous antigeas in association with class 1 MHC
moiecuIes. J. ImmunoL. 150, 438-446.
Rogers-Jonson, P., Gadjdusek, D., Morgan, O., Zaninovic, V., Sarin, P.,and Graham,
D. l98S. HTLV-1 and HTLV-IIantiibodies and tropical spastic paraparesis. Lancer. 2,
1247-1248.

Roman, G., Roman, L., and Osame, M. 1990. Human T lymphotropic virus type I
neurotropism. Prog. Me. Virol. 37, 190-210.
Rowe, J., Perkins, S., Bradshaw, P., Song, O-Y., and Foung, S. 1994. Neutralizing
HTLV-human monoclonai a n h i e s (HMAbs) to conformational epitopes. AlDS Res.
Hum. Retrovimses. 10,509.

Ruegg, C., MoneiI, C., and Strand, M. 1989. Idenmcation, using synthetic peptides,
of the minimum amino acid SeQuence for the retroviral trammembrane protein p15E
required for inhibition of lymphoproliferation and its similady to gp21 of human Tiymphotropic virus type I and II. J. W. 63, 3250-3256.
Rusceai, F., Robert-Guroff, M., Ceccherini-NelIi, L., Milwwada, J., Popovic, M., and
Gallo, R. 1983. Persistent in-vitro infection by human T-cell leukaemia-lymphoma virus
(HTLV) of normal human T-lymphocytesfrom blood relatives of patients with HTLVassociated mature T-ceil neoplasros. In?. J. Cruicer. 31, 171-180.

Rusche, J., Lynn, D., Robert-Guroff, M., Langiois, A., Lyerly, H., Carson, H., Rrohn,
K., Ranki, A., Gallo, R., Bolognesi, D., Pumy, S., and Matthews, T. 1987. Humoral
immune respo~l~e
to the entire human immw1odeficiency virus envelope glycoprotein
made in insect cells. Proc. N d Acad. Sei. USA. 84, 6924-6928.
Rusche, J., Javaherian, K., McDanal, C., Petro, J., Lynn, D., Grimaila, R., Langlois,
A., Gallo, R., Arthur, L.,Fischinger, P., Bolognesi, D.,Putney, S., and Matthews, T.
1988. Anti'bodies that inhibit fusion of human imrnunodeficiency virus-infecteci ceiis bind
a 24-amino acid setpence of the viral envelope, gp120. Proc. Natl. Acad. Sei. USA. 85,
3198-3202.
Sagara, Y., h u e , Y., Shuaki, H., Jinno, A., Hoshino, H., and Maeda, Y. 1996.
Identification and mapping of functionai domains on human T-celI lymphotropic virus
type 1 envelope proteins by using synthetic peptides. J. mol. 70, 1564-1569.
Saida, T., Saida, K., Funauchi, M.,Nishiguchi, E., Nakajima, M., Matsuda, S., Ohta,
M., Ohta, K., Nishitani, H.,and Hatanaka, M. 1988. HTLV-1 myelitis: isolation of
vins, genomic aaalysis and infection in neural celi cultures. A m . N. Y. Ac&. Sei. 540,
636-638.
Sakakibara, I., Sugimoto, Y., Sasagawa, A., Honjo, S., Tsujimoto, H., Nakamura, H.,
and Hagarni, M. 1986. Spontaneous malignant lymphoma in an African green rnonkey
namally infectai with simian lymphotropic virus (STLV) . J. Med. Primatol. 15, 3113 18.
Saksena, N., Sherman, M. Yanagihara, R., Dube, D., and Poiesz, B. 1992. LTR
sequence and phylogenetic analysis of a newly discovered variant of HTLV-1 isolated
from the Hagahai of Papua New Guinea. Virology . 189, 1-9.

Saksena, N., Herve, K., S h e m , M., Durand, J., Mathiot, C., Muller, M., Love, J.,
Sinoussi, F., Dube, D., and Poiw, B. 1993. Sequene and phylogentic anlysis of a new
SLTV-I from a nahuaiiy infected Tantalus monkey from Central M c a . Virology. 192,
3 12-320.
Saksena, N.,Herve, K., Durand, J., LeGuenno, B., Diop, O., Digoutte, J., Mathiot,
C., Muller, M., Love, J., Dube, S., Shennan, M., Benz, P., Eremoy, S., Galat-Luong,
A., Galat, G., Paul, B., Dube, D., Barre-Sinoussi, F., and Poiesz, B. 1994.
Seroepidemiologic, molecular and phylogenetic analyses of simian TceU leukemia
viruses (STLV-r) from various nahualiy iafeted monkey species ftom central and
wertern Africa. Virology. 198, 297-310.

Sambrook, J., Fritsch, E., and Maniatis, T. 1989. Molecular cloning: a laboratory
mariud, 2nd edition, p.6.53. Cold Spring Harbor Laboratory Press, Cold S p ~ Harbor,
g

N.Y.

Samuel, K., Lautenberger, J., Jorcyk, C., Joseph, S., Wong-Staal, F., and Papas, T.
1984. Diagnostic potemial for human malignancies of bac~riallyproduced HTLV-1
envelope glycoprotein. Science. 226. 1094-1097.
Sanada, I., Nakada, K., Funigen, S., Kumagai. E., Yamaguchi, K., Yoshida, M., and
Takatsuki, K. 1986. Chromosomaiabnormalities in a patient with smoldering ad& T-ceii
leukemia: evidence for a muitistep pathogenesis. 1Rukenria Res. 10, 1377-1382.

Sanders, M., Makgoba, M., and Shaw, S. 1988. Human naive and mernory T ceiis:
reinterpretation of helper-inducer and suppressor-inducer subsets. I m m ~ ~ u lTodizy
l.
. 9,
195-199.
Satow. Y., Hashido, M., Ishikawa, K., Honds, H., MiPino, M., Kawana, T., and
Hayami, M. 1991. Detection of HTLV-1 antigen in peripheral and cord blood
lymphocytes from carrier mothers. Lancer. 338, 915-916.
Sattentau, Q., and Moore, J. 1995. W-1 neutralization is determined by epitope
exposw on the gp120 otigomer. J. Exp. Med. 182, 185-196.
Sawada, T., Iwahara, Y., Ishii, K., Taguchi, Hg, Hoshino, H., and Miyoshi, 1. 1991.
Immunoglobulin prophylaxis against milk-borne anwission of human T ceii leukemia
virus type 1 in rabbits. J. Infect. Dis. 164, 1193-1196.
Schafer, W., and Bolognesi, D. 1977. Mammalian C-type oncomavUuses: relationships
between virai structurai and ce11-dace antigens and their possible significance in
immunological defense mechani~ms.Contemp. Top. Immunobiol. 6. 127-167.
Schagger, H.. and von Jagow, G. 1987. Triche-sodRun dodecyl sulfate-polyacrylamide
gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal.
Biochem. 166, 368-379.
Schatzl, F., Yakovleva, L., Lapin, V., Rose, D., Inzhia, L., Gaedigknitschko, K.,
Denbardt, F., and von der Helm, K. 1992. Detection and characterization of T-celi
leukemia vins-like proviral sequemes in PBL and tissues of baboons by PCR.Leukemia.
6, 158-160.
Schlegel, R., Wade-Glass, M., Rabson, M., and Chung Yang, Y. 1986. The ES
transforming gene of bovine papiomavirus encodes a small, hydrophobie polypeptide.
Science. 233, 464-467.
Schlesinger, J., Brandriss, M., and Walsh, E. 1985. Protection against 17D Yellow fever
encephalitis in mice by passive transfer of monoclonai antiibodies to the nonstructurai
giycoprotein gp48 and by active immunkation with gp48. J. Imnwnol. 135, 2805-2809.

Schmaljohn, A., Kokubun, K., and Cole, G. 1983. htective monoclonal antibodies
define maturational and pHdependent antigenic changes in Sindbis virus E l glycoprotein.
Virology. UO, 144-154.
Schneider, J., Yamamoto, N., Hiauma, Y., and &nsma~, G. 1984. Sera fkom adult
T-cell leukemia patients react with envelope and core polypeptides of adult T-ce11
leukemia virus. Virology. 132, 1-11.
Schulz, T., Calabro, M-L.,Hoad, J., Carrington, C., Matutes, E., Catovsky, D., and
Weiss, R. 1991. HTLV-Ienvelope SeQuences ftom Brazil, the Can'bbeaa, and Ronmia:
clustering of sequeIlces accordhg to geograpbic ongin and variabiiity in an antigenic
epitope. Virology. 184, 483-491.

Schwarz, T., and Daterna, R. 1982. Inhiiition of the dolichol pathway of protein
glycosyiation. In M e r l i d of E.ymology, Voltune 83, pp. 432-443. Edited by J. Abelson
and M. Simon. Academic Press Inc. New York.
Seiki, M., Hattori, S., Hirayama, Y., and Yoshida, M. 1983. Human adult T-ceii
leukemia virus: Complete nucleotide seqyence of the provhs genome integrated in
leukemia ceil DNA. Proc. N i l . Acud. Sci. USA. 80, 3618-3622.

Seiki, M., Eddy, R., Shows, T., and Yoshida, M. 1984. Nonspecific integration of the
HTLV provirus genome into adult T ceii leukemia ceils. Nmre. 309, 640-642.
Seiki, M., Hikikoshi, A., Tanigucbi, T., and Yoshia, M. 1985. Expression of the pX
gene of HTLV-1: General splicing rnectmism in the HTLV-famiiy. Science. 228, 15321534.
Seiki, M., houe, J., Hidaka, M., and Yoshida, M. 1988. Two cis-acting elements
responsible for pst-transcrptiod trans-regdation of gene expression of human T-cell
leukemia virus type 1. Ploc. N d Acad. Sn'. USA.85, 7124-7128.
Seto, A., Isono, T., and Ogawa, K. 1991. Infection of inbred rabbits with ceil-free
HTLV-1. Leuk. Res. 15, 105-110.
Seto, A., Kawanishi, M., Matsuda, S., Ogawa, K., Eguchi, T., and Miyoshi, 1. 1987.
Induction of preleukemic stage of adult T celi leukernia-like disease in rabbits. Jpn. J.
Cancer Res. 78, 1150-1155.
Sheremata, W., Harrington, W. Jr., Bradshaw, P., Foung, S., Raffanti, S., Berger, I.,
Snodgrass, S., Resnick, L., and Poiesz, B. 1993. Association of "tropical ataxic
neuropathy" with HTLV-II.Virus Res. 29,71-77.

Sherman, M., Sakesena, N., Dube, D., Yanagihara, R., and Poiesz, B. 1992.
Evolutionary insights on the ongin of human T-ceII iymphomal leukemia virus type 1
(HTLV-I) derived from sequence anaiysis of a w HTLV-1 variant h m Papua New
Guinea. J. Krol. 66, 255602563.
Shida, H., Tochilaira, T., Sato, T., Konno, T., Hirayoshi, K., Seki, M., Ito, Y.,
Hatanaka, M., Hiauma, Y., Sugimoto, M., Takahashi-NiShimaki, F., Mamyama, T.,
Miki, K., S a , K., Morita, M., Sashiyama, H., and Hayami, M. 1987. Effect of the
recombinant vaccinia vinises that express HTLV-1 envelope gene on HTLV-1 infection.
EMBO. J. 6, 3379-3384.
Shida, H., Hinuma, Y., Hatanaka, M., Morita, M., Kidokoro, M., Suzuki, K.,
Maruyama, T., Takahashi-NisM, F., Sugimoto, M., Kitamura, R., Miyazawa, T.,
and Hayami, M. 1988. Effects and vinilences of recombinant vaccinia viruses derived
from attenuated sains that express the human T-cell leukemia virus type 1 envelope
gene. J. VIroL. 62, 4474-4480.
Shimamoto, Y., Kaichi, M., Funai, N.,Suga, K.,Matsuzaki, M., and Yamaguchi, M.
1993. Spontaneous regression in adult T-ceii leukemia/1ymphoma. Concer. 72,735-740.
Shimotono, K., Takahashi, Y., Shimuzn, N., Gojoboni, T., Golde, D., Chen, 1.. Mieva,
M., and Sugimura, T. 1985. Complete nucleotide sequence of an infectious clone of
human T-cell leukemia virus type II: An open reading fiame for protease gene. Proc.
NatL. Acad. Sci. USA. 82, 3101-3115.
Siekevitz, M., Feinberg, M., Holbrook, N., Wong-Staal, F., and Greene, W. 1987.
Activation of hterleukin 2 and interleukin 2-receptor (tac) promoter expression by the
transactivator (tat) gene product of human T-cell leukemia virus type 1. Proc. Natl. Acad.
Sci. USA.84, 5389-5393.
Sissons, J., and Oldstone, M. 1980a. Antiiy-mediated destruction of virus-infected
cells . Adv. Immunol. 29, 209-260.
Sissom, J., and Oldstone, M. 1980b. Killing of virus-infecteci celis by cytotoxic
lymphocytes. J. Infect Dis. 142, 114-119.
Sjolander, S., Boimstedt, A., Akerblom, L., Horal, P., Olofsson, S., Morein, B., and
Sjolander, A. 1996. N-linked glycans in the CDebinding domain of human
immunodeficieucy virus type 1 envelope glycoprotein gp160 are essential for the in vivo
priming of T ceUs recognizing an epitope located in thei. vicinity. Rrohgy . 215, 124133.

Smith, C., Bamett, B., Thomas, D., and Temoltzin-Palacios, F. 1991, Structural
assignments of novel an imrnUM3dominiuit antigenic sites in the neutraluing antiiy
response of CBAKa mice to inuenza virus. J. Exp. Med. 173, 953-959.
Smith, G., and Moss, B. 1983. Infectious poxvinis vectors have capacity for at least
25,000 base pairs of foreign DNA. Gene. 25, 21-28.
Smith, G., Summers, M., and Frazer, M. 1983. Productionof human beta interferon in
insect celis infectecl with a baculovim expression vector. Mol. Cell. Biol. 3, 2156-2165.

Smith, G., Ju, G., Ericson, B., Moschera, f., Lahm, H-W.,Chizzonite, R., and
Summen, M. 1985. Modification and secretionof human interleukin 2 produced in insect
cells by by a bacuiovinis expression vector. Proc. N d Acad. Sci. USA. 82,8404-8408.
Smith, S., Brown, M., Rowe, D., Cailard, R., and Beverley, P. 1986. Functional
subsets of human helper-inducer ceils deby a new monoclonal antiiy, UCHL1.
Immunology . 58, 63-70.

Sodora, D., Cohen, O., Muggeridge, M., and Eisenberg, R. 1991. Absence of
asparagine-nked oiigosaccharides fiom glycoprotein D of herpes simplex virus type 1
results in a stnicturaliy altered by biologicaily active protein. J. Wrol. 65, 4424443 1.
Sodroski, J., Rosen, W., and Haselthe, W. 1984. Transacting transcriptional activation
of the long terminal repeat of human T lymphotropic vinises in infkcted tells. Science.
225, 381-385.
Sommerfelt, M., Williams, B., Clapham, P., Solomon, E., Goodfeliow, P., and Weiss,
R. 1988. Human T cell leukemia vimses use a receptor detemineci by human
chromosome 17. Science, 242, 15574559.
Somda, J., Yashiki, S., Fujiyoshi, T., Osame, M., Tanaka, H., Zaninovic, V., Blank,
M., Blank, A., Hayami, M., and Tajima, K. 1994. Genetic background of ATL and
HAM/TSP among Japanese and other ethnic groups. J. AZDS Research. IO, 452.
Southern, E. 1975. Detection of spocific sequerices among DNA fragments separated by
gel electrophoresis. J. Mol. Biol. 98, 503-5 17.
Stamatatos, L., and Cheng-Mayer, C. 1995. Structural modulations of the envelope
a 1 2 0 glycoprotein of human immunodeficiency virus type 1 upon oligomerization and
differential V3 loop epitope exposure of isolates displaying distinct tropism upon virionsoluble receptor binding. J. Virol. 69, 6191-6198.
Studier, F., and Moffat, B. 1986. Use of bacteriophage T7 RNA polymerase to direct
selective high-level expression of cloned genes. J. Mol. B i d . 189, 113-130.

Stuhler, G., aod Waiden, P. 1993. Coilaboration of helper and cytotoxic T lymphocytes.
EUE J. 1-01.
23, 2279-2286.
Suga, T., Rameyama, T., Konoshita, T., Mataunni, M., Tanaka, H., Kushida, S.,
Ami, Y.,Uchida, M.,Uchida, K.,and Miwa, M. 1991. Lnfction of rats with HTLV-1:
a smaii animal mode1 for HTLV-1 carriers. 1'. J. h c e r . 49, 764-769.
Sugawara, K., Kitarne, F., Nishimura, H., and Nakamura, K. 1988. Operational and
topological analyses of antigenic sites on influenza nnis glycoprotein and their
dependence on glycosylation. J. Gen. Virol. 69, 537-547.
Sugimura, K., Fujh, M., Kobayashi, N., Sakitani, M., Hatanaka, M., and Hinuma, Y.
1984a. Retrovinis-induced expression of interleukin 2 receptoa human B-celi Iineage.
Proc. Natl. Acad. Sci. USA. 81, 7441-7445.
Sugimura, K., Sakitani, M., and Hinuma, Y. 1984b.Microplate method for retrovirusinduced transformation of normal human T cells. J. Immunol. M e t h d o a73,
s 379-385.
Sugiyama, H., Doi, H., Yamaguchi, K., Tsuji, Y., Miyamoto, T., and Hino, S. 1986.
Significance of postnatal mother-to-child transmission of human T-lymphotropic virus
type4 on the development of ad& T-cell leukedymphoma. J. Med. Km!. 20, 253260.
Summen, M., and Smith, G. 1987. A Munual of Methods for Baculovirus Vectors und
Ikrect Cell Culture Procedures, Texas Agriculniral Experiment Station Bulletin No.
1555.
S d , T., Fujisawa, J., Toita, M., and Yoshida, M. 1993. The tram-activator tac of
human T-cell leukemia virus type 1 (H.TLV-1) interacts with CAMP-responsiveelement
(CRE) bindiag and CRE modulator proteins diat bind to the 21-base-pair enhancer of
HTLV-1.Proc. Natl.Acad. Sci. USA. 90, 610-614.
Tabor, S., and Richardson, C. 1985. A bacteriophage 'I7 RNA polymerase/promoter
system for controlied exclusive expression of specitic genes. Proc. Natl. Ac&. Sci. USA.
82, lO74-lO78.
Taguchi, H., Kobayashi, M., and Mjoshi, 1. 1991. Imrnunosuppression by HTLV-1
infection. Lancez. 337, 308.
Tajimi, K., and the T- and B-ceii maligancy study group and coauthors. 1990. The
fourth nation-wide study of adult T-cell leukemia/lymphoma (ATL) in Japan: estimates
of risk of ATL and its geographical and clinical features. Int. J. Cancer. 45, 237-243.

Takahashi, K., Takezaki, T., Oki, T., Kawakami, K., Yashiki, S., Fujiyoshi, T.,
Usuku, K., Mueller, N., The Mother-to-Child Transmission Study Group, Osame, M.,
Miyata, K., Nagata, Y., and Sonoda, S. 1991. hh.iiitory effect of matenial a n t i i y on
mother-to-chiid traosmission of human T-lymphotropic virus type 1. Int. J. Gzncer. 49,
673-677.

Hino, S. 1990. Matemal-infant trammission of HTLV-1: implication for disease. In

Humnn Retrovirology: HIZV, pp.363-375. Edited by W.A. Blartner. Raven Press, New

York.
TakatSulci, K., Uchiayama, T., Sagawa, K., and Yodoi, J. 1977. Adult T-cell leukemia
in Japan. In Topics in Hemmology, pp. 73-77. Edited by S. Seno, S. Takaku and S.
Irino. Excerpta Media, Amsterdam.

Takatsuki, K., Yamaguchi, K., Kawano, F., Hattori, T., Nisbimura, H., Tsuda, H.,
and Itai, Y. 1985. Clinical diversity in adult T-ce11
leukernia/lymphoma. Cbcer. Res. 45, SP4644.

Sariada, I., Nakada, K.,

Takatsuki, K.,Hinuma, Y., and Yoshida, M. 1992. Advances in adult T ceIl leukemia
and HTLV-1 research. In Mmogrqh on Gzncer Research No. 39, p. 261. Edited by K.
Takatsuki, Y. Himuiia, and M. Yoshida. Japan Scientific Societies Press, Tokyo.
Takehara, K., Ireland, D., and Bishop, D. 1988. Co-expression of the hepatitis surface
and core antigens using baculovirus multiple expresion vectors. J. Gen. Virol. 69,27632777.
Takehara, N., Iwahara, Y., Uemura, Y., Sawada, T., Ohtsuki, Y., Iwai, H., Hosbino,
H., and Miyoshi, 1. 1989. Effect of immunization on HTLV-1 infection in rabbits. IN.
J. Cancer. 44, 332-336.

Tanaka, Y., Zeng, L., Shiraki, H., Shida, H., and T o ~ i w a H.

, 1991. Identification of
a neutralization epitope on the envelope gp46 antigen of human T-ceil leukemia virus
type 1and induction of neutralizing antibody by peptide immunzation. J. Z'mnoi. 147.
354-360.

Tanaka, Y., Ish, K., Sawada, T., Ohtsuki, Y., Hoshino, H., Yanagihara, R., and
Miyoshi, I. 1993. Prophylarris against a Melanesian variant of human T-lymphotropic
virus type 1(HTLV-I) in rabbits using HTLV-1 immune globulin fkom asymptomatically
infected Japanese carriers. Blood. 82, 3664-3667.

Tanaka, Y.,Tanaka,R., Terada, E.,Koyanagi, Y.,Miyano-Kurosaki, N., Yamamoto,

N., Baba, E., Nakamura, M., and Shida, H. 1994. Induction of antibody responses that
neutralize human TceU leukemia v i m type 1 infection in vitro and in vivo by peptide
immunization. J. Virol. 68, 6323-6331.

Tateno, M., Kondo, N., Itoh, T., Chubachi, T., Togashi., T., and Yoshiki, T. 1984.
Rat Lymphoid cell iines with human T ceil leukemia Wus production. 1. Biological and
serologicai characterization. J. Exp. Med. 159, 1105-1116.
Taylor, G. 1994. The role of antibxiy in controiing and/or clearing virus infections. In
Srrategies in Vaccine Design, pp. 17-34. Edited by G. Ada. R. G. Tandes Co., Texas.
Teyton, L., O'Sullivan, D., Dickson, P.W., Lotteau, V. Sette, A., Fink, P., and
Peterson, P. 1990. Invariant chah distinguishes between the exogenous and endogenous
antigen presentation pathways. Nmure. 348, 39-44.

Thomas, S., Hodgson, H., Riska, P., and Graham, C. 1990. The role of the endoplasmic
reticulum in antigen processing. N-glycosylation of infiuenza hemagglutinin abrogates
CD4+ cytotoxic T cell recognition of endogenously processeci antigen. J. I'nol. 144,
2789-2794.

Tokudome, S., Tokunaga, O., Shimamoto, Y., Miyamoto, Y., Sumida,I., Kilruchi, MO1
Talceshita, M., Ikeda, T., Fujiawra, K.,and Yoshihara, M. 1989. Incidence of adult Tceii leukemia/lymphoma among human T-lymphotropic vinas type I carrier in Saga.
Japan. G m e r Res. 49, 226230.
Trimble, R., and Maiey, G. 1984. Optimizing hydrolysis o f N-linked high-mannose
oligosaccharides by endo-beta-N-acety1-glucosaniinidase H. Analyt. Biochem. 141, 5 15522.

Umehara, F., Izumo, S., Ronquiiio, A., Matsumufo, K., Sato, E., and Osame, M.
1994. Cytokine expression in the spinal cord lesions in HTLV-1 associateci myelopathy.
J. Neuropathol. Exp. Neurd. 53, 72-77.

Unge, T., Solomin, L., M e W , M., Derse, D.,Felber, B., and Pavlakis, G. 1991. The
rex regulatory protein of HTLV-1 biads specifically to its target within the Wal RNA.
hoc- Nat[.AC& Scie USA. 88, 7145-7149.

van Drunen Littel-van den Hurk, S., Hughes, G., and Babiuk, L. 1990. The role of
carbohydrate in the antigenic and immunogenic structure of bovine herpesvirus type 1
glycoproteins gI and gIV. J. Gen. Wtol. 71, 2053-2063.
van D m e n Littel-Van Den Hurk,S., Parker, M., Fitzpatrick, R., Zamb, T., van den
Hurk, J., Campos, M., Harland, R., and Babiuk, L. 1991. Expression o f bovine
herpesvirus 1 glycoprotein gIV by recombinant baculovinis and aaalysis of its
immunogenic properties. J. Wrol. 65, 263- 271.
VanSlyke, J., and Hruby, D. 1990. Posttranslational modification of vaccinia virus
proteus. Curr. Top. Microbiol. Tmmunol. 163, 185-206.

Varki, A. 1993. Biological roles of oligosaccharides: AU of the theores are correct.

Giycubiology. 3, 97-130.

V m u s , H.,and Swanswm, R. 1982. Repication of retrovuses. In RNA mer

vinres, vol. l . , pp. 369-512. Edited by R. Weiss, N. Teich, H. Varmus and J. Corn.
Cold Spring Harbor Laboratoty, Cold Spring Harbor, N.Y.

Vernant, J., Maurs, L., Gessain, A., Barin, F., Gout, O., Delaporte, J., Snhadji, K.,
Buisson, G., and de Th, G. 1987. Endemic tropical spastic parparesis associated with
human T-Lymphotropic vins type 1: a clinicai and seroepidemiological study of 25 cases.
Ann. Neurol. 21, 123-130.
Vidal, S., Mottet, G., Kolakofslty, D., and Roux, L. 1989. Addition of high-mannose
sugars must precede d i s i d e bond formation for proper folding of Sendai virus

glycoproteins. I. Virol. 63,892-90.

Vile, R., Schulz, T., Danos, O., Collins, M., and Weiss, R. 1991. A Murine ceil l k
produchg HTLV-1 pseudotype virions carrying a selectable marker gene. Wrology. 180,
420-424.

Wainberg, M., Blain, N., and Spira, B. 1987. Inhibition of humao lymphocyte
mitogenesis by human and other retrovinises. Differential effect of interleukin-2 in
restoration of responsiveness. AIDS. 1, 83-87.

Wano, Y., Feinberg, M., Hosking, J., Bogerd, H., and Greene, W. 1988. Stable
expression of the tm gene of type I human T-celi leukemia virus in human T ceils
ativates specific cellular geae involveci in p w t h . Proc. N d Acad. Sci. USA. 85,97339737.
Watabe, K., Saida, T.,and Kim, S. 1989. Human and simian glial ceiis infected by
human T-lymphotropic virus type 1in culture. J. Neuropahoi. EqD. Neurol. 48,610-6 19.

Watanabe, T., Seiki, M . , Tsujimoto, H., Miyoshi, I., Haymi, M., and Yoshida, M.
1985. Sequence homology of the sirnian retrovirus genorne with human T-ce11 Ieukemia
virus type 1. Virology. 144, 59-65.
Watmabe, T., Yamaguchi, K., Takatsuki, K., Osame, M . , and Yoshida, M. 1990.
Constitutive expression of parathymid hormone-related protein gene in HTLV-1 carriers
and adult T cell leukemia patients which can be pans-activated by HTLV-I t a gene. J.
Erp. Med. 172, 759-765.

Wathen, M., Aeed, P., and Elhammer, A. 1991. Characterization of oligosaccharide

structures on a chimeric respiratory syacytial virus protein expressed in k e c t cell line
S B . Biochemistry. 30, 2863-2868.

Wojchowski, D., O r b , S., and Sytkowski, A. 1987. Active human erythropoietin

expressed in iasect cells using a bacdovirus vector: a role for N-linked oligosaccharide.
Biochim. Biophys. Aca. 910,224432.
Wong-Staal, F., Hahn, B., Manzari, V., Colombini, S., Franchini, G., GelmaM, E.,
and Gaiio, R. 1983. A m ~ e of
y human leukemias for sequences of a human retrovirus.
h r e . 302, 626-628.
Wu, E., Dickson, D., Jacobson, S.. and Raine, C. 1993. Neuroaxonal dystrophy in
tropical spastic paraparesis/HTLV-1 associatecl myelopathy.Acta. Neuropathol. 86,224235.
Wust, C., Crombie, R., and Brown, A. 1987. Passive protection acmss subgroups of
alphaviruses by hyperimmune non-cross-neutralizing anti-Sindbis senim. h c . Soc. Exp.
Biol. Med. 184, 56-63.
Yamaguchi, K., Nishimura, Kawano, F., Kohrogi, H., Jono, M., Miyamoto, Y., and
Takatsuki, K. 1983. A proposal for smolderinig adult T-ceIl leukemia-Diversity in
ciinicai picaires of adult T-cell leudemia. Jpn. J. Clin. Oncol. 13, 189-199.
Yamaguchi, K., Kiyokawa, T., Nakada, K., Yul, L., Asou, N., Ishii, T., Sanada, I.,
Seiki, M., Yoshida, M., Matutes, E., Catovslq, D., and Takatsuki, K. 1988. Polyclonai
integration of HTLV-I proviral DNA in lymphocytes rom HTI,V-1 seropositive
individuals: An intermediate state between the healthy carrier state and smoldering ATL.
Br. J. Haemaol. 68, 169-174.
Yamaguchi, K., and Takatsuki, K. 1993. Adult T-celi leukaernia-lymphoma. Clin.
Haenrcltol. 6, 899-915.
Yamaguchi, K. 1994. Human T-lymphotropic virus type I in Japan. Lance?. 343, 213216.
Yamamoto, N., Hayami, M., Komuro, A., Schneider, J., Hunsmann, G., Okada, M.,
and Hinuma, Y. 1984. Experimentd infation of cynomolgus monkeys with a human
retrovims, adult T-cell leukemia virus. Med. MMicrobiol. 173, 57-64.
Yamanashi, Y., Mon, S., Yoshida, M., Kishimoto, T., houe, K., Yamamoto, T., and
Toyoshima, K. 1989. Seleck expression of a protek-tyrosine kinase, p56[w, in
hematopoietic celis and association with production of human T e l l lymphotropic virus
type 1. hoc. Natl. Acad. Sci. USA. 86, 6538-6524.
Yamashita, I., Katamine, S., Moriuchi, R., Nakamura, Y., Miyamoto, T., Eguchi, K.,
and Nagataki, S. 1994. Transactivation of the human interleukin-6 gene by human Tlymphotropic virus type l Tax protein. Bluod. 84, 1573-1578.

Yanagibara, R., Nerurkar, V., and Ajdukiewicz, A. 1991. Cornparison between saaias
of human T lymphotropic v i . type I isolateci h m inhabitants of the Solomon Islands
and Papua New Guinea. J. Infect. Dis. 164, 443-449.

Yang, X.-C., Karschin, A., Labarca, C., Elroy-Stein, O., Moss, B., Davidson, N., and
Lester, H. 1991. Expression of ion channels and receptors in Xenopus oocytes using
vaccinia v h . FASEB J. 5, 2209-2216.
Yodoi, J., Okada, M., Tagaya, Y., Teshigawara, K., Fuhii, K., Ishida. N., Ikuta, K.,
Maeda, M., Honjo, T., Osawa, K., D-in,
T., Tateno, M., and Yoshci, T.
1985. Rat lymphoid ceil lines produckg human human T celi leukemia Wus. II.
Constituitive expression of rat interleukin 2 receptor. J. &p. Med. 161, 924-934.
Yoshida, M., Sefi, M., Yamaguchi, K., and Takatsuki, K. 1984. Monoclonal
integration of human T-ceU leukemia provirus in all primary tumours of adult T-ceii
leukemia suggests causative role of human T-ceii leukemia virus in the didiase. Proc.
Nad Acad. Sci. USA. 81,2534-2537.
Yoshida, M. 1987. Expression of the HTLV-1 genome and its association with a unique
T-ceil malignancy.Biuchim. Biophys. A a a . 907, 145-16 1.
Yoshida, M., and Seiki, M. 1987. Recent advances in the molecular biology of HTLV-1:
Transactivation of viral and cellular genes. Annu. Rev. Immunol. 5, 541-549.
Yoshida, M., and Fujisawa, J. 1992. Positive and negative regulation of HTLV-1 gene
expression and their roles in leukemogenesis in ATL. In Advances in adult T cell
leukemia and HIZV-I reseurch. Monogrqph on Cmcer Research No. 39, pp. 217-235.
Edikd by K. Takatsuki, Y. Hinuma, and M. Yoshida. Japan Scientific Societies Press,
Tokyo.
Yoshida, M. 1994. Tenth anaiversary perspectives on AIDS. Host-HTLV type 1
interaction at the molecular level. Aa)S Res. Hum. Retmviwes. 10, 1193-1197.
Yoshiki, T., Kondo, N., Chubachi, T., Tateno, M., Togashi, T., and Itoh, T. 1987. IZat
Lymphoid ceil lines with HTLV-I production. m.Transmission of HTLV-1 into rats and
analysis of celi surface antigens associated with HTLV-1. Arch. h l . 97, 181-196.
Yoshiki, T., Tatem, M., Salairai, H., Ishiguro, N., Ikeda, H., and Wakisaka, A. 1992.
A rat mode1 of HTLV-1 infection. Gann. Mono. ancer Res. 39, 205-216.

Zack, J., Amgo, S., Weitsman, S., Go, A., Haislip, A., and Chen, 1. 1990. HIV-1
entry into quiescent prirnary lymphocytes: molecular analysis reveals a labile, latent virai
structure. Cell. 61, 213-222.

Zhao, L., and Giam, C. 1991. Interaction of the human T-ceU lymphotrophic v i m type
1 (HTLV-I) transcriptionai activator Tax with ceiiular factors t
hbind specificaily to the
21-base-pairrepeats in the HTLV-1 en&ancer.Proc. N d Acud. Sci. USA. 88, 114451 1449.
Zhou, X., Glas, R., Liu, T.,Ljunggren, H.G., and Jondal, M. Antigen processing
mutant T2 cens priesent virai antigen restricted through H-2Kb. Eur. J. Immunol. 23,
1802- 180%

Zucker-Franklin, D., Hooper, W.,and Eratt, B. 1992.Human lymphotropic retroviruses

associated with mycosis fungoides: Evidence that human T-cell lymphotropic virus type
II (HTL,V-II) as weli as HTLV-I may play a rote in the disease. Blood. , 1537-1545.