Harold E. Varmus

Page 1

Career Summary

   Like many other physicians who now practice molecular biology,  I was a
relative late-comer to the laboratory. When I entered college twenty years
ago, I assumed, out of.filia1 devotion to my father's profession, that I would
ultimately practice medicine. I was, however, soon seduced out of my good
intentions into the study of English literature and philosophy; although I
managed to complete my pre-medical requirements, I spent most of my time
reading Dickens and running the college newspaper, with little time or patience
for laboratory science. During my first post-graduate year, as a graduate student
in English literature at Harvard, I developed a nagging sense that I was leaving
something(s) important behind by failing to go to medical school - a connection
with a more exciting wbrld, whatever talent I had for science, a conception of
work as an altruistic endeavor. So, with the aim of merging my verbal and
scientific interests, presumably in psychiatry, I entered medical school the
folbdng year, only to discover that I was as enchanted by the "hard" sciences
of the curriculum (especially neuroanatomy and biochemistry) as by the "softl'.
By the time of my third year, I was feeling quite committed to a career in
academic internal medicine, with research interests in endocrinology or hematology,
although my laboratory experience was still confined to a couple of rather
unsatisfactory summers (e.g., ref. 111).

My attitude was, however, shifted once again shortly after my arrival at

the NIH in 1968. I had planned, like many in my medical school generation,
to prepare for a clinically-oriented academic career as a Public Health Service
trainee at NIH. Before starting my two years as a medical house staff officer
at Presbyterian Hospital, I had been happily matched to work with Ira Pastan,
presumably on regulation of secretion of thyroid hormones. However, by the time
I was ready to move to Bethesda, Pastan and Perlman had discovered that cyclic
AMP reverses glucose repression of beta-galactosidase synthesis in E. coli.
As a consequence, I inadvertantly entered one of the most exciting areas in
molecular biology in the late 1960'~~ the mechanism of positive, pleiotropic
regulation of genes by cyclic AMP. This meant that I had to learn a large amount
of enzymology, nucleic acid biochemistry, :and bacterial genetics rather quickly,
an effort rewarded by experiments which showed that the regulation occurred at
the level of transcription of DNA into RNA, both -- in vivo and in a reconstructed
system (see refs. 2-12). To my surprise, I loved working in the laboratory,
was excited by ansqers to prob'lems I had known nothing about a year earlier,
and found my interest in clinical medicine dwindling rapidly.

Nevertheless, I felt my background and proclivities were more appropriate

to the study of animal cells. Stimulated by accounts in courses taken at NIH
of the disorderly conception of how RNA tumor viruses replicate, I sought a
postdoctoral position in animal virology. Ironically, just as I was moving in
the summer of 1970 to join the new group formed by Leon Levintow, Mike Bishop,
and Warren Levinson in San Francisco, reverse transcriptase was discovered by
Baltimore and Temin. Their discovery, however, not only injected some order
into the field, it also infused it with money, competition, and (most importantly)
experimental possibilities.

For the past seven years, working principally with the avian sarcoma and

Harold E. Varmus

Page 2

mouse mammary tumor viruses, I have watched these possibilities grow, rather
than wither. Although many have tried (generally unsuccessfully (cf ref.38))
to use RNA tumor viruses as models to search for human tumor viruses, I have
been more interested iil exploiting their utility for the investigation of
several fundamental problems in cell biology - DNA synthesis, integration,
transcriptional regulation, evolution, the origin of viruses, and even the
genetic basis of human disease.
to the questions surrounding the synthesis and integration of proviral DNA;
we have shown that viral DNA is made in the cytoplasm in infected cells, that
it enters a covalently closed form in the nucleus; and that it integrates co-
valently into the host cell genome. These observations and closely related
work with several virubes are recorded primarily in references 14,20,28,36,41,
42,43,49,58,59,66, 71,80,85,88,and 89. I have been interested in the expression
of viral genes (refs. 27,61,65,83,84,87,90), particularly in non-permissive
cells infected by avian sarcoma virus (refs. 40,43,67) and in the regulation
of mouse mammary tumor virus genes by glucocorticoid hormones (refs. 50,54,70,
86).
DNA (refs. 14,18,31,56,60,69,72) and to underctand the origin and function of viral
genes, particularly the transforming gene of avian sarcoma virus (refs. 53,55,
,64,75, and 82) have been relevant to the new field of, molecular evolution.
Some'of the technology required for these studies has, in addition, returned
me to old hematological interests, and I have collabarated with Y.W. Kan and
his colleagues on the analysis of the genetic bases of the thalassemias (refs.
47,51, and 63).

A large portion of my wo7rk. has .been devoted

Our efforts to follow the evolution of viral genes endogenous to normal

As the scope of my work has grown, so has the size of our laboratory group

and the effort involved in sustaining its momentum. Between us, Mike Bishop
and I now closely supervise the activities of about 16 postdoctoral fellows
and students and 4 technicians. I am the principal investigator for two
research grants (from NIH and ACS), and co-investigator for another research
grant and two training grants. Beyond these responsibilities, my obligations
to the medical campus have involved me in several areas outside my immediate
research interests - as a teacher of medical microbiology (bacteriology and
immunology, as well as virology) and as a committee member (for lecture series,
student research, biosafety, the M.D.-Ph.D. program, curriculum, promotions,
and others). My life has been further complicated by extramural functions -
as a speaker at
of manuscripts (officially for - Cell and Virology, unofficially for several other
journals); and, most exhausting of all, as a reviewer of grant applications
(particularly as a member of the NIH Virology Study Section). When I review
my current way of life in this peculiar fashion, I am amazed that I can still
read books, fish for trout, play the oboe, talk with my family, and, yes, work
at the bench. But sometimes I do.

611 too frehuent) meetings and on many campuses;  as an editor

Harold E. Varmus

Proposal

My principal objectives in planning a sabbatical year are not unusual:

to work in a new scientific and cultural environment, after eight years in
the same one; to shed, even temporarily, an increasing number of distractions
from laboratory work; and to learn experimental approaches that will permit at
least modest readjustments of my long-range goals in research. Since I do not
wish to make fundamental changes in the kind of research I do, and since.the
experiments I wish to do involve scientific risks, polikical rules, and prag-
matic barriers to the acquisition of reagents, I have decided to plan several
kinds of experiments and to affiliate myself with a multi-disciplinary group
that could satisfy my needs with any of those experiments.  I will' therefore
be working at the ICRF in London, under the direct sponsorship of Dr. Michael
Stoker, with Drs. Mike Fried, Bob Kamen, and/or Alan Smith, depending upon the
success I have in preparing for the various experiments.

In the main, I wish to extend our on-going studies of the replication of

avian sarcoma and mouse mamary tumor viruses in new directions, using new
tools of genetic engineering (creation of deletion mutants and DNA cloning, in
particular), and to begin to study the genetic organization and replication of
hepatitis B virus.
the sabbatical year are described below.

Three projects on which I hope to make some progress during

1. Genetic organization of hepatitis B virus (HBV). I have recently

decided to commit a portion of my laboratory's efforts to the study of HBV, a
medically important virus, as yet very incompletely described, which promises
to be amenable to the sorts of techniques we have employed in the study of RNA
tumor viruses. My immediate objectives are to prepare radiolabeled HBV DNAand to
use it in the analysis of virus-specific nucleic acids in infected liver cells.

Because of logistical, ethical, and biohazard considerations, I am attempting

to gather the appropriate reagents for these experiments prior to my departure.
This will require preparation of viral DNA from a suitable donor who produces
Dane particles (the putative infectious virus); amplification of viral DNA in
vitro using a DNA replication system, derived from T4 bacteriophage, which we are
using for studies of RNA tumor virus-specific DNA in collaboration with Dr. Bruce
Alberts; and extraction of both DNA and RNA from either (a) liver biopsies obtained
from experimentally-infected chimpanzees, or (b) livers obtained at autopsy from
patients with active hepatitis B. If some or all of these reagents can be
assembled, I will proceed during my sabbatical to develop a more satisfactory
restriction map of Dane particle DNA; to further define the structure and hetero-
geneity of the molecule (which now appears to be a gapped double-stranded circle
with
daltons); to measure the amount and strand of origin of the virus-specific RNA
in cells;  to characterize and translate the RNA synthesized from viral DNA in
vitro by various RNA polymerases; to translate RNA from infected cells, using
preparative hybridization to enrich for viral RNA and appropriate antisera to
test for synthesis of known viral antigens; and to `examine cellular DNA for the
existence of integrated viral DNA and for the specificity of'any integration
sites (using restriction endonucleases and the DNA transfer procedure, with DNA
labeled in vitro_ as hybridization reagent).
with several of the technical procedures involved, there are areas (e.g., detailed

at least moderate sequence heterogeneity and a weight of about 1.6 x lo6

Although I am obviously familiar

Harold E. Varmus
Page 4

restriction mapping, in vitro synthesis of RNA and proteins, immunoprecipitation)
in which I expect to be helped a great deal by my new colleagues. Obviously,
limitations on materials arelikely to restrict the number of questions I can
satisfactorily answer in this area; for that reason, I expect to work on the
problems discussed below as well.

2. Characterization of the sites of integration of RNA tumor virus DNA.

A major goal of my laboratory for the past few years has been a detailed charac-
terization of the mechanism of synthesis and integration of viral DNA in cells
infected by RNA tumor viruses.  Recently,' using restriction endonucleases,
hybridization reagents specific for various portions of the viral genomes,  and
the DNA transfer procedure developed by Southern, we have described the location
and orientation of integrated (proviral) DNA in several kinds of cells infected
by avian sarcoma virus (ASV) and mouse mammary tumor virus (MMTV); in addition,
we have similarly studied the ASV- and MNTV-related DNA endogenous  to avian
and murine cells. In both systems, we generally find proviral DNA to be co-
linear with the viral RNA genome, rather than permuted; in some cases (e.g.,
some strains of ASV in chicken cells) we have found preferred cellular sites
for integration, whereas, in other cases, preferred sites, if they exist, have
yet to be defined. In all cases, however, tife are now able to identify fragments
of cellular DNA containing only a few virus-specific nucleotides (corresponding
to a terminus of the viral genome) plus the adjacent cellular sequences. For
purposes of understanding both the integration mechanism and the regulation of
viral gene expression at the transcriptional level, it is necessary to determine
the nucleotide sequence of the adjacent cellular DNA. This approach requires
the preparation of cellular DNA enriched for the appropriate fragments (using
reverse phase chromatography or R-loop formation in conjunction with preparative
gel electrophoresis), followed by cloning and detecting the fragment in a
suitable vector and host. I hope to develop such clones (though not to sequence
the sites) during my sabbatical year. Since the experiments involve the use of
recombinant DNA, I am in the process of applying to the committee which governs
such research in Britain for permission to do them. A large number of possible
combinations will be offered €or considerqtion; in general, the experiments
emphasize the inclusion of very small 5' terminal regions of the viral genomes,
outside the genetically-defined areas and distant from known transforming genes;
the use of defective polyoma DNA as a vector; and the study of endogenous viral
sequences. The ICGF is well-equipped with facilities for such experiments (with
a containment area more stringently designed than our P3 facilities) and Mike
Fried would be an expert guide for me in conducting these studies since he is
currently using his polyoma mutants as cloning vehicles. Obviously, however,
execution of these experiments is dependent upon favorable action on my applications.

3. Generation of new deletion mutants of avian RNA tumor viruses for dis-

section of oncogenic functions. The use of deletion mutants has been a powerful
tool in virology, particularly with the recent development of methods to produce
small deletions in defined regions of papova virus genomes.
virus-specific DNA is available in much smaller quantities, we have identified
and prepared closed circular viral DNA from ASV-infected cells in amounts suitable
for restriction endonuclease mapping and for infectivity studies, and my reading
of the literature suggests that creation and isolation of deletion mutants should
be feasible. This approach has obvious utility in confirming the genetic map of

Although RNA tunor

Harold E. Varmus

Page 5

ASV and in attempting to identify new genes. Although transformation of cul-
tured fibroblasts (and production of sarcomas) appears to be controlled by a
single, well-studied gene (called 31,  there is thus far no genetic assignment
for the capacity of many RNA tumor viruses to  cause leukemias.  Leukemogenesis could
be dependent upon a known gene (e.g., env),lpmanas yet undiscovered gens or upon
some consequence of the positioning of the leukosis viral DNA in the host cell
genome. I would like to begin to investigate this question by preparing a
series of deletion mutants of an avian leukosis virus (ALV). A.restriction
endonuclease map of an ALV strain will have been prepared prior to my departure,
as well as an adequate amount of closed circular DNA from cells acutely-infected
with the parental strain.
with enzymes shown to pave a single recognition site in viral DNA (or with
other enzymes in the presence of intercalating dyes which prevent more than
one scission per molecule);  the ends of the resultant linear DNA will be trimmed
with an appropriate exonuclease; and the DNA will be used to infect permissive
cultures, in the presence and absence of a non-oncogenic helper virus (e.g.,
Rous associated virus-0).
restriction mapping and cloned, if necessary. My objective during the sabbatical
year will be the development of a series of such mutants, with emphasis upon
mutations in the unmapped regions of the genome, particularly the region between
the,env gene and the 3'  terminus.
tested for leukemogenic activity by any of several investigators who have helped us
with such problems in the past. Any mutants which have lost their leukemogenic
potential will be subjected to detailed analysis of their sites of integration
in various cell types, and we will attempt to generate mutants with similar
phenotypes by making lesions in the same and neighboring sequences. This project
could thus lead to new directions in the context of our present research program.
There would be several advantages to starting such work at the ICRF: Mike Fried
has had remarkable success in producing andkscribing a wide variety of deletion
mutants of polyoma virus; in addition, Robin Weiss, John Wyke and their colleagues
would be available for consultation on the genetics and biology of avian leukosis
viruses.

-

To make the mutants, circular DNA will be cleaved

Newly-acquired mutant viruses will be confirmed by

I would then arrange. to have these mutants