Science Guardian

(show)

[CANCER RESEARCH 47, 1199-1220, March 1, 1987]

Perspectives in Cancer Research

Retroviruses as Carcinogens and Pathogens: Expectations

and Reality

Peter H. Duesberg

Department of Molecular Biology and Virus Laboratory, University of California, Berkeley, California 94720

Abstract

Retroviruses (without transforming genes) are thought to cause leukemia’s and other cancers in animals and humans because they were originally isolated from those diseases and because experimental infections of new-borns may induce leukemia’s with probabilities of 0 to 90%. According to this hypothesis viral cancersshould be contagious, polyclonal, and preventable by immunization.

However, retroviruses are rather widespread in healthy animals and humans where they typically cause latent infections and antiviral immunity. The leukemia risk of such infections is less than 0.1% and thus about as low as that of virus-free controls. Indeed retroviruses are not sufficient to initiate transformation (a) because of the low percentage of symptomatic virus carriers and the complete lack of transforming function in vitro; (b) because of the striking discrepancies between the long latent periods of 0.5 to 10 years for carcinogenesis and the short eclipse of days to weeks for virus replication and direct pathogenic and immunogenic effects; (c) because there is no gene with a late transforming function, since all genes are essential for replication; (d) because host genes, which do

not inhibit virus, inhibit tumorigenesis up to 100% if intact and determine the nature of the tumor if defective; and above all (e) because of the monoclonal origin of viral leukemias, defined by viral integration sites that are different in each tumor. On these bases the

probability that a virus-infected cell will become transformed is estimated to be about 10 (11th power). The viruses are also not necessary to maintain transformation, since many animal and all bovine and human tumors do not express viral antigens or RNA or contain only incomplete proviruses. Thus as carcinogens retroviruses do not or only very rarely (10 (11th power)) fulfill the third. Therefore it has been proposed that retroviruses transform inefficiently by activating latent cellular oncogenes by for example provirus integration. This predicts diploid tumors with great diversity, because integration sites are different in each tumor.

However, the uniformity of different viral and even nonviral tumors of the same lineage, their common susceptibility to the same tumor resistance genes, and transformation-specific chromosome abnormalities shared with nonviral tumors each argue for cellular transforming genes. Indeed clonal chromosome abnormalities are the only known transformation-specific determinants of viral tumors. Since tumors originate with these abnormalities, these or associated events, rather than preexisting viruses, must initiate transformation. Therefore it is proposed that transformation is a virus-independent event and that clonal viral integration sites are consequences of clonal proliferation of transformed cells. The role of

the virus in carcinogenesis is limited to the induction of hyperplasia which is necessary but not sufficient for carcinogenesis. Hyperplasia depends on chronic viremia or high virus expression which are very rare in animals outside the laboratory and have never been observed in humans. Since latent viruses, which are typical of nearly all natural infections, are neither direct nor indirect carcinogens, they are not targets for cancer prevention. Viruses are also not targets for cancer therapy, since tumors are not maintained and not directly initiated by viral genes and occur naturally despite active antiviral immunity.

Lymphotropic retrovirus has been proposed to cause AIDS because 90% of the patients have antibody to the virus. Therefore antibody to the virus is used to diagnose AIDS and those at risk for AIDS. The virus has also been suggested as a cause of diseases of the lung and the nervous system. Promiscuous male homosexuals and recipients of frequent transfusions are at high risk for infection and also at a

relatively high risk for AIDS, which averages 0.3% and may reach 5%. Others are at a low risk for infection and if infected are at no risk for AIDS. AIDS viruses are thought to kill T-cells, although these viruses depend on mitosis for replication and do not lyse cells in asymptomatic infections. Indeed the virus is not sufficient to cause

AIDS (a) because the percentage of symptomatic carriers is low and varies between 0 and 5% with the risk group of the carrier, suggesting a cofactor or another cause; (b) because the latent period for AIDS is 5 years compared to an eclipse of only days to weeks for replication and direct pathogenic and immunogenic effects; and (c) because there is no gene with a late AIDS function, since all viral genes are essential for replication. Moreover the extremely low levels

of virus expression and infiltration cast doubt on whether the virus is even necessary to cause AIDS or any of the other diseases with which it is associated. Typically, proviral DNA is detectable in only 15% of AIDS patients and then only in one of 10 (2nd power) to 10 (3rd

power) lymphocytes and is expressed in only 1 of 10 (4th power) to 10 (5th power) lymphocytes. Thus the virus is inactive or latent in carriers with and without AIDS. It is for this reason that it is not transmitted as a cell-free agent. By contrast, all other viruses are expressed at high titers when they function as pathogens. Therefore AIDS virus could be just the most common occupational infection of

those at risk for AIDS because retroviruses are not cytocidal and unlike most viruses persist as latent, nonpathogenic infections. As such the virus is an indicator of sera that may cause AIDS. Vaccination is not likely to benefit virus carriers, because nearly all have active antiviral immunity.

How often have I said to you, that when you have eliminated the impossible, whatever remains however improbable must be the truth.

– Sherlock Holmes

The irreversible and predictable courses of most cancers indicate that

cancer has a genetic basis. In 1914 Boveri (1) proposed that cancer is

caused by chromosomal mutations. This hypothesis has since received

ample support (2-4), although a cellular cancer gene has yet to be

identified (5). In the light of the spectacular discovery of RSV* in 1911,

which proved to be a direct, infectious carcinogen, the hypothesis

emerged that viruses may be a significant source of exogenous cancer

genes (6). The virus-cancer hypothesis has since steadily gained support

because retroviruses and DNA viruses were frequently isolated from

animal leukemias and other tumors, and occasionally from human

leukemias, in efforts to identify causative agents (7-16). However, once

discovered in tumors and named tumor viruses, most of these viruses

were subsequently found to be widespread in healthy animals and

humans (8, 12-18). Thus these viruses are compatible with the first but

apparently not necessarily with the third of Koch’s postulates** as viral

carcinogens. Only a few of the many tumor viruses are indeed directly

oncogenic, such as RSV and about 20 other types of retroviruses (5, 13,

19, 20), and hence compatible with Koch’s third postulate. Therefore, if

we want to assess the role of viruses in cancer, there must be a clear

separation between those viruses which are directly oncogenic and those

which are not. The directly oncogenic retroviruses owe their

transforming function to a particular class of genes which are termed onc

genes (20). These are as yet the only known autonomous cancer genes

that can transform diploid cells in vitro as well as in animals susceptible

to the particular virus (5). Since susceptible cells are inevitably

transformed as soon as they are infected, the resulting tumors are

polyclonal (13, 16). Nevertheless, directly oncogenic retroviruses have

never caused epidemics of cancer. The probable reason is that onc genes

are not essential for the survival of the virus and hence are readily lost by

spontaneous deletion or mutation (5). Indeed, onc genes were originally

discovered by the analysis of spontaneous onc deletion mutants of RSV

(21). Moreover, because onc genes typically replace essential genes

(except in some strains of RSV) these viruses cannot replicate unless

aided by a nondefective helper virus (5, 13).

The vast majority of the tumor viruses are retroviruses and DNA viruses

that do not contain onc genes. The RNA genomes of all retroviruses

without onc genes measure only 8 or 9 kilobases (13, 22). They all

encode three major essential genes which virtually exhaust their coding

capacity. These are in the 5′ to 3′ map order gag which encodes the viral

core protein, pol which encodes the reverse transcriptase, and env which

encodes the envelope glycoprotein (23, 24). Although these viruses lack

onc genes they are considered tumor viruses, because they were

originally isolated from tumors and because experimental infections may

induce tumors under certain conditions. However, in contrast to tumors

caused by viruses with onc genes, such tumors are always monoclonal

and induced reproducively only in genetically selected animals

inoculated as newborns after latent periods of over 6 months (see below).

Because of the long latent periods, these retroviruses are said to be

“slow” viruses (13, 16), although their mechanism of replication is

exactly the same as that of their fast and efficient relatives with onc

genes that transform cells as soon as they infect them (5, 19) (Table 1).

The retroviruses are also considered to be plausible natural carcinogens

because they are not cytocidal and hence compatible with neoplastic

growth and other slow diseases. Indeed, retroviruses are the only viruses

that depend on mitosis for replication (13, 25).

However, the retroviruses without onc genes are also the most common

and benign passenger viruses of healthy animals and humans probably

because of their unique noncytocidal mechanism of replication and their

characteristic ability to coexist with their hosts without causing any

pathogenic symptoms either as latent infections, which make no

biochemical demands, or even as productive infections. Based on the

permissiveness of a host for expression and reproduction, they have been

divided into exogenous viruses which are typically expressed and hence

potentially pathogenic and endogenous viruses which are typically latent

and hence nonpathogenic (16-18). Because they are so readily

suppressed in response to as yet undefined cellular suppressors (8, 11,

12, 16-18), endogenous viruses are integrated as proviruses into the germ

line of most if not all vertebrates (8, 13, 16-18). Nevertheless, the

endogenous and exogenous retroviruses are entirely isogenic and there is

no absolute biochemical or functional distinction between them except

for their response to suppressors of a particular host (13, 16-18) (Part I,

Section A). Therefore the association of these viruses with a given

disease is not sufficient even to suggest a causative role in it. Indeed

there is as yet no direct evidence that retroviruses play a role as natural

carcinogens of wild animals and humans. Thus the critical expectations

of the virus-cancer hypothesis, namely that RNA or DNA tumor viruses

would be direct carcinogens, that viral tumors would be polyclonal

because each virus-infected cell would be transformed, and above all that

viral carcinogenesis would be preventable by immunization, remain

largely unconfirmed.

Recently retroviruses without onc genes have been isolated from patients

with AIDS and those at risk for AIDS and have since been considered

the cause of AIDS (26). In contrast to other retroviruses, the AIDS

viruses are thought to act as direct, cytocidal pathogens that kill

susceptible T-cells (13, 27).

Here we discuss how the retroviruses without onc genes fit the role of

viral carcinogens or AIDS pathogens and whether these viruses are

indeed the vessels of evil they have been labeled to be. Above all we

hope to identify transformation-specific or AIDS-specific viral and

cellular determinants and functions. Since the genetic repertoire of all

retroviruses without onc genes, including that of the AIDS viruses (28),

is exhausted by genes that are essential for virus replication (13,24), a

hypothetical oncogenic or AIDS function would have to be indirect or it

would have to be encoded by one of the essential genes. In the second

case the virus would be oncogenic or cause AIDS wherever it replicates.

A survey of the best studied animal and human retroviruses demonstrates

that these viruses are not sufficient to cause tumors and not necessary to

maintain them. Most likely these viruses play a role in inducing tumors

indirectly. Indeed transformation appears to be a virus-independent,

cellular event for which chromosome abnormalities are the only specific

markers. Likewise the AIDS viruses are shown not to be sufficient to

cause AIDS, and the evidence that they are necessary to cause it is

debated.

1. Retroviruses and Cancer

A. Retroviruses Are Not Sufficient for Transformation Because Less

Than 0.1% of Infected Animals or Humans Develop Tumors

Avian lymphomatosis virus was originally isolated from leukemic

chickens (29). However, subsequent studies proved that latent infection

by avian lymphomatosis viruses occurs in all chicken flocks and that by

sexual maturity most birds are infected (30-32). Statistics report an

annual incidence of 2 to 3% lymphomatoses in some inbred flocks. Yet

these statistics include the more common lymphomas caused by Marek’s

virus (a herpes virus) (33, 34). The apparent paradox that the same virus

is present in most normal and healthy animals (30) but may be

leukemogenic in certain conditions was resolved at least in descriptive

terms by experimental and congenital contact infections. Typically

experimental or contact infection of newborn animals that are not

protected by maternal antibody would induce chronic (31, 32) or

temporal (35, 36) viremia. The probability of such animals for

subsequent lymphomatosis ranges from 0 to 90% depending on tumor

resistance genes (Section C). However, infection of immunocompetent

adults or of newborn animals protected by maternal antibody and later by

active immunity would induce latent, persistent infections with a very

low risk of less than 1% for lymphomatosis (32, H. Rubin, personal

communication.) Thus only viremic animals are likely to develop

leukemia at a predictable risk.

Viremia has a fast proliferative effect on hemopoietic cells and generates

lymphoblast hyperplasia (Fig. 1) (32, 36, 37). Hyperplasia appears to be

necessary but not sufficient for later leukemogenesis because it does not

lead to leukemia in tumor-resistant birds (36) (Section C) and because

removal of the burso of Fabricius, the major site of lymphoproliferation,

prevents development of the disease (9, 32).

The murine leukemia viruses were also originally isolated from leukemic

inbred mice (9) and subsequently detected as latent infections in most

healthy mice (8, 13, 16, 17, 38). Indeed, about 0.5% of the DNA of a

normal mouse is estimated to be proviral DNA of endogenous

retroviruses, corresponding to 500 proviral equivalents per cell (18).

Nevertheless leukemia in feral mice is apparently very rare. For instance

low virus expression, but not a single leukemia, was recorded in 20% of

wild mice (38) probably because wild mice restrict virus expression and

thus never become viremic and leukemic. However, in an inbred stock of

feral mice predisposed to lymphoma and paralysis, 90% were viremic

from an early age, of which 5% developed lymphomas at about 18

months (39).

Experimental infections of newborn, inbred mice with appropriate strains

of murine leukemia viruses induce chronic viremias. Such viremic mice

develop leukemias with probabilities of 0 to 90% depending on the

mouse strain (Section C). However, if mice that are susceptible to

leukemogenesis are infected by the time they are immunocompetent or

are protected by maternal antibodies if infected as neonates, no chronic

viremia and essentially no leukemia are observed (although a latent

infection is established) (41). Thus leukemogenesis depends on viremia

(40) as with the avian system. However, viremia is not sufficient,

because certain tumor-resistant strains do not develop leukemia even in

the presence of viremia (42) (Section C). Again viremia has an early

proliferative effect on lymphocytes which has been exploited to

quantitate these viruses in vivo within 2 weeks by the “spleen weight” or

“spleen colony” assay (18, 43-47). This hyperplasia of lymphocytes is

necessary for leukemogenesis, because the risk that an infected animal

will develop leukemia is drastically reduced or eliminated by

thymectomy, which is a major source of cells for prospective

leukemogenesis (9).

The AKR mouse is a special example in which spontaneous expression

of endogenus virus and the absence of tumor resistance genes inevitably

lead to viremia at a few weeks after birth and, in 90% of the animals, to

leukemia at 6 to 12 months of age (9, 41, 48). This also shows that

endogenous viruses can be just as pathogenic or leukemogenic as

exogenous viruses if they are expressed at a high level. Likewise,

endogenous avian retroviruses are leukemogenic in chickens permissive

for acute infection (49, 50).

The evidence that mammary carcinomas are transmissible by a milkborne

virus, MMTV, indicates that the virus is an etiological faction (51,

52). However, the same virus is also endogenous but not expressed in

most healthy mice (16, 53). Since no mammary tumors have been

reported in wild mice the natural incidence must be very low, but in mice

bred for high incidence of mammary carcinomas it may rise to 90% (13,

16, 54, 55). As with the leukemia viruses, the risk for tumorigenesis was

shown to depend on a high level of virus expression from an early age

and on the development of hyperplasias that are necessary but not

sufficient for carcinogenesis (56, 57). For example, BALB/c mice that

express over 100 mu-g virus per ml milk all develop tumors after

latencies of over 12 months, but mice that express 3 mu-g or less virus

per ml develop no tumors at all (54, 58).

Feline leukemia virus was originally isolated from cats with

lymphosarcoma (59) and subsequently from many healthy cats. It is

estimated that at least 50 to 60% of all cats become naturally infected by

feline leukemia viruses at some time during their lives (60, 61).

However, only about 0.04% of all cats develop leukemia on an annual

basis (62), which is thought to be caused by these viruses (13, 61, 63).

Most natural infections cause transient virus expression which is

followed by an immune response, after which little virus is expressed

(60, 64, 65). Such infections do not induce leukemias at a predictable

rate (61). However, 1 to 2% of the naturally infected cats become

chronically viremic (66). About 28% of the viremic cats develop

leukemias after latent periods of 2 years. Thus viremia indicates a high

risk for the development of leukemia (66). Viremia may result from a

congenital infection in the absence of maternal antibody or from a native

immunodeficiency. As in the avian and murine systems, experimental

infection of newborn, immunotolerant cats produces early viremia and

runting diseases and late leukemias at a much higher incidence than

natural infections (63, 64, 67, 68). The gibbon ape leukemia virus was

also initially discovered in leukemic apes and was later isolated from

healthy gibbons (13, 69). Again, only chronically viremic gibbons were

shown to be at risk for leukemia (70).

The bovine and human retroviruses associated with acute leukemias are

always biochemically inactive or latent (Section D). Viremia, which is

frequently associated with a leukemia of congenitally or experimentally

infected domestic chickens, cats, or inbred mice, has never been

observed in the bovine or human system. Accordingly bovine and human

leukemia viruses could be isolated from certain leukemic cells only after

cultivation in vitro away from the suppressive immune system of the host

(71, 72). In regions of endemic bovine leukemia virus infection 60 to

100% of all animals in a herd were found to contain antiviral antibody

(73, 74). However, the incidence of leukemia was reported to range only

from 0.01 to 0.4% (16, 73). Experimental infections with cell-free virus

have not provided conclusive evidence for viral leukemogenesis. As yet

only 1 of 25 animals infected with bovine leukemia virus has developed

a leukemia 7 years after inoculation (73). Additional inoculations of 20

newborn calves did not cause a single leukemia within 5 years, although

all animals developed antiviral antibody. [J. M. Miller and M. S. Van der

Maaten, personal communication.] However, 50% of newborn sheep

inoculated with bovine leukemia virus developed leukemia about 4 years

later (75). These sheep were probably more susceptible to the bovine

virus than cattle, because they would lack maternal antibody to the virus.

Indeed they could have been transiently viremic, because antibody was

detected only 4 months after inoculation (75).

HTLV-1 or ATLV was originally isolated from a human cell line derived

from a patient with T-cell leukemia (71). It replicates in T-cells (27) and

also in endothelial cells (76) or fibroblasts (77). The virus was

subsequently shown, using antiviral antibody for detection, to be

endemic as latent, asymptomatic infections in Japan and the Caribbean

(27). Since virus expression is undetectably low not only in healthy but

also in leukemic virus carriers, infections must be diagnosed indirectly

by antiviral antibody or biochemically by searching for latent proviral

DNA (Section D). Due to the complete and consistent latency, the virus

can be isolated from infected cells only after activation in vitro when it is

no longer controlled by the host’s antiviral immunity and suppressors.

Therefore the virus is not naturally transmitted as a cell-free agent like

other pathogenic viruses, but only congenitally, sexually, or by blood

transfusion, that is, by contacts that involve exchange of infected cells

(13, 27).

It is often pointed out that functional evidence for the virus-cancer

hypothesis is difficult to obtain in humans because experimental

infection is not possible and thus Koch’s third postulate cannot be tested.

However, this argument does not apply here since naturally and

chronically infected, asymptomatic human carriers are abundant. Yet

most infections never lead to leukemias and none have ever been

observed to cause viremias. Moreover, not a single adult T-cell leukemia

was observed in recipients of blood transfusions from virus-positive

donors (13, 78, 79), although recipients developed antiviral antibody

(81).

The incidence of adult T-cell leukemia among Japanese with antiviral

immunity is estimated to be only 0.06% based on 339 cases of T-cell

leukemia among 600,000 antibody-positive subjects (78). Other studies

have detected antiviral antibody in healthy Swedish donors (268) and in

3.4% of 1.2 x 10 (6 power) healthy Japanese blood donors (79). Further,

it was reported that 0.9% of the people of Taiwan are antibody positive,

but the incidence of the leukemia was not mentioned (80).

In conclusion, the tumor risk of the statistically most relevant group of

retrovirus infections, namely the latent natural infections with antiviral

immunity, is very low. It averages less than 0.1% in different species, as

it is less than 1% in domestic chickens, undetectably low in wild mice,

0.04% in cattle, and 0.06% in humans. Thus the virus is not sufficient to

cause cancer.

Moreover, since the viruses associated with all human tumors and most

natural tumors of animals are latent and frequently defective (Section D),

it is difficult to justify the claims that these viruses play any causative

role in tumorigenesis. Indeed nearly all healthy chickens, mice, cats,

cattle, and humans carry endogenous and exogenous retroviruses that are

latent and hence neither pathogenic nor oncogenic (12, 16-18, 78, 82).

Latent infections by cytocidal viruses, such as herpes viruses, are

likewise all asymptomatic (83). Nevertheless it may be argued that only

a small percentage of retroviral infections are expected to be oncogenic

because only a small percentage of all other viral or microbial infections

are pathogenic. However, the low percentage of symptomatic infections

with other viruses and microbes reflects the low percentage of acute

infections that have overwhelmed host defense mechanisms, but not a

low percentage of latent infections that cause disease. Thus there is no

orthodox explanation for the claims that some murine and avian, most

feline, and all bovine and human leukemias (Section D) are the work of

latent viruses.

Even the view that retroviruses cause leukemia or carcinoma directly in

productive infections is debatable, because indeed highly productive

infections are frequently asymptomatic. For example, despite chronic

acute viremias certain chickens, mice, or cats, inoculated experimentally

or by contact as immuno-tolerant newborns, do not develop leukemia

(see above and Section C). Further no malignant transformation has ever

been observed in cultured cells that are actively producing retroviruses,

and the probability that an infected cell of an animal will become

transformed is only 10 (11th power) (Section F). This low probability that

a productively infected cell will become transformed is a uniquely

retrovirus-specific reason for asymptomatic infections. It is for this

reason that retroviruses without onc gene can be asymptomatic for cancer

even in acute, productive infections of animals (30, 31, 36, 42, 66, 70),

although they may then cause other diseases (Section B).

Thus retrovirus infections are not only asymptomatic due to latency and

low levels of virus infiltration, like all other viruses, but are also

asymptomatic due to a particular discrepancy between acute and

productive infection and oncogenesis. To answer the question of why

some viremic animals do and others do not develop leukemia and why

tumors appear so late after infection (Section B), both tumor resistance

genes (Section C) and the mechanism of transformation must be

considered (Section H).

B. Discrepancies between the Short Latent Period of Replication and the

Long Latent Periods of Oncogenesis: Further Proof That Virus Is Not

Sufficient for Cancer

Here we compare the kinetics of virus replication and direct pathogenic

and immunogenic effects with the kinetics of virus-induced

transformation. If retroviral genes were sufficient to induce cancer, the

kinetics of carcinogenesis would closely follow the kinetics of virus

replication.

Kinetics of Replication and of Early Pathogenic and Immunologenic

Effects. The eclipse period of retrovirus replication has been determined

to be 1 to 3 days in tissue culture (Table 1) using either transforming onc

genes as markers or the appearance of reverse transcriptase or

interference with other viruses or plaque formation for viruses without

onc genes (13, 16) (see below). The incubation period following which

retroviruses without onc genes induce viremia in animals is 1 to several

weeks (9, 13, 14, 16) (Table 1). In immunocompetent animals antiviral

immunity follows infections with a lag of 2 to 8 weeks.

In animals, retroviruses without onc genes can be directly pathogenic if

they are expressed at high titers. For instance, avian retroviruses may

cause in newborn chickens diseases of polyclonal proliferative nature

like osteopetrosis, angiosarcoma, hyperthyroidism (84-87), or

hyperplastic follicles of B-cells in the bursa of Fabricius (36, 37) after

latencies of 2 to 8 weeks. The same viruses may also cause diseases of

debilitative nature such as stunting, obesity, anemia, or

immunodeficiency after lag periods of 2 to 8 weeks (88, 89). Infections

of newborn mice that cause viremia also cause polyclonal lymphocyte

hyperplasias, splenomegaly, and immuno-suppression several weeks

after infection (47) (Section A). The early appearance of hyperplastic

nodules in mammary tumor virus-infected animals prior to malignant

transformation has also been proposed to be a virus-induced,

hyperplastic effect (56, 57). Infection of newborn kittens with feline

leukemia virus causes early runting effects and depletion of lymphocytes

within 8 to 12 weeks (64, 67, 68) followed by persistent viremia in up to

80% of the animals (90). In experimentally infected adult animals mostly

transient (85%) and only a few persistent (15%) viremias are observed

(64, 68, 90). Likewise primate retroviruses such as Mason-Pfizer virus

(91) or simian AIDS virus (92) or STLV-III virus (93) may cause

runting, immuno-depression, and mortality several weeks after

inoculation if the animals do not develop antiviral immunity. These early

and direct pathogenic effects of retroviruses without onc genes depend

entirely on acute infections at high virus titers and occur only in the

absence of or prior to antiviral immunity.

Retroviruses have also been observed to be directly pathogenic by

mutagenesis via provirus integration of cellular genes (13, 16, 94, 95).

Given about 10 (6th power) kilobases for the eukaryotic genome and

assuming random integration, a given cellular gene would be mutated in

1 of 10 (6th power) infected cells (see Sections E and F). Therefore this

mechanism of pathogenesis would play a role in vivo only if mutagenesis

were to occur at a single or few cell stage of development (94) or if such

a mutation would induce clonal proliferation, as is speculated in Section

E.

Certain direct, cytopathic effects of retroviruses without onc genes are

also detectable in vitro within days or weeks after infection, although

malignant transformation has never been observed in cell culture. For

example, the avian reticulo-endotheliosis viruses fuse and kill a fraction

of infected cells during the initial phase of infection (96, 97). Certain

strains of avian retroviruses form plaques of dead primary chicken

embryo cells in culture within 7 to 12 days postinfection. This effect is

probably based on cell fusion and has been used as a reliable virus assay

(45, 98). The plaque assays of murine leukemia viruses on XC rat cells

(99) and on mink cells (101-104) also reflect fast cytopathic effects

involving fusions of infected cells (45). Cell fusion of human

lymphocytes in vitro is also typical of HTLV-I (105, 106) and of AIDS

virus (27) (see Part II). Cells are thought to be fused in vitro by crosslinking

through multivalent bonds between viral envelope antigens and

cellular receptors, a process that requires high local concentration of

virus particles (13, 16, 27, 45, 105). The fusion effect is not observed in

chronic acute or latent infections of animals or humans or in chronically

infected cell lines cultured in vitro. Therefore it appears to be

predominantly a cell culture artifact, possibly resulting from interaction

between virus receptors of uninfected cells with viruses budding from

the surface of adjacent cells. This has been directly demonstrated by

inhibition of HTLV-I-mediated fusion with antiserum from infected

individuals (105). Thus as direct pathogens the retroviruses are not

“slow” viruses, as they are frequently termed with regard to their

presumed role in carcinogenesis. The “lentiviruses” that are considered

models of slow viral pathogenesis (13), but not carcinogenesis, are no

exception. Recently an ovine lentivirus known as visna or maedi virus

was shown to cause rapid lymphoid interstitial pneumonia in newborn

sheep, several weeks after infection (269). This study pointed out that the

virus, if expressed at high titer, is directly and rapidly pathogenic. Slow

disease may reflect persistent virus expression at restricted sites.

Late Oncogenesis. Since retroviruses without onc genes do not

transform cells in culture, all measurements of the latent period of viral

oncogenesis are based on studies of infected animals or humans (Table

1). Typically, the latent periods are dated from the time of virus infection

and thus are somewhat presumptuous, in that the assumption is made that

tumors, if they appear, were initiated by the virus.

The latent period between experimental or congenital infection and

lymphomatosis in chickens ranges from 6 months to several years (13,

16, 30, 32, 36, 107). In mice congenitally or experimentally infected with

murine leukemia viruses, leukemia takes 6 to 24 months to appear (9, 39,

42, 108). The latent period of mammary carcinomagenesis in mice

infected by milk-transmitted MMTV ranges from 6 to 18 months and

typically requires several pregnancies of the mouse (16, 54). Longer

latent periods of up to 24 months are observed in mice that do not

express virus in their milk (55, 109).

The latent period between experimental infection and leukemia is 8 and

12 months in most cats, but only 2 to 3 months in some (62, 66, 90).

(The early tumors may have been hyperplasias or tumors induced by

feline sarcoma viruses.) The latent period estimated between natural

virus infection and leukemia is estimated to be 2 to 3 years in cats that

express virus and about 2 to 6 years in cats that do not express virus (63,

66, 110). By contrast, induction of antiviral immunity occurs within

several weeks after infection (64, 67).

Bovine leukemia virus-associated leukemias are never seen in animals

less than 2 years old and appear at a mean age of 6 years (16). The only

experimental bovine lymphosarcoma on record appeared 7 years (73)

and some experimental ovine leukemias appeared 4 years (75) after virus

inoculation. By contrast, antibody to viral core and envelope proteins

appears 4 and 9 weeks after infection (73). Experimental infection of

gibbon apes generated leukemia after a latent period of 1 year compared

to only 2 weeks for the appearance of antiviral immunity (16, 70).

The latent period for the development of human T-cell leukemia in

HTLV-1 positive cancers has been estimated at 5 to 10 years based on

the lag between the onset of leukemia and the first appearance of

antiviral antibodies of proviral DNA (13, 111, 112). More recently, the

latent period of HTLV-I has been raised to record heights of 30 (270)

and 40 years (271). By contrast, the latent period of infection and

subsequent antiviral immunity was determined to be only 50 days based

on seroconversion of the recipients of HTLV-I-positive blood

transfusions (81).

The 5- to 40-year latencies claimed for leukemogenesis by HTLV-I are

perhaps the most bizarre efforts in linking a virus with a disease. If

correct this means either that an infected T-cell becomes leukemic by the

time it is 5 to 40 years old or that one of its offspring becomes leukemic

in the 50th to 500th generation, assuming an average generation time of a

month (176). Clearly the role of the virus in such a process, if any, must

be highly indirect. Since all viral genes are essential for replication (13,

204), there is nothing new that the virus could contribute after one round

of infection or 24 to 48 hours. This is specifically for HTLV-I and

bovine leukemia viruses which are biochemically inactive not only

during the long latent period but also during the lethal period of the

disease (Sections A and D).

The monumental discrepancies between the long latent periods from 6

months to 10 years for leukemogenesis compared to the short latent

periods of several weeks for virus replication or direct pathogenic and

immunogenic effects are unambiguous signals that the viruses are not

sufficient to initiate leukemia and other tumors (Fig. 1). The viruses are

fast and efficient immunogens or pathogens but are either not or are

highly indirect carcinogens.

Transformation in Vitro by HTLV-I in 30 to 60 days?

Immortalization of primary human lymphocytes infected by HTLV-I or

ATLV or simian retroviruses in vitro has been suggested to be equivalent

to leukemogenic transformation in vivo (13, 27, 113, 114). If correct, this

would be the only example of a retrovirus without onc genes capable of

malignant transformation in vitro. The assay infects about 5 x 10 (6th

power) primary human lymphocytes with HTLV-I. However, less than

one of these cells survives the incubation period of 30 to 60 days, termed

“crisis” because the resulting immortal cells are monoclonal with regard

to the proviral integration site and because only 4 of 23 such experiments

generate immortal cells (115). Since no virus expression is observed

during the critical selection period of the immortal cell and since some

immortalized cells contain only defective proviruses (115),

immortalization is not a viral gene function. Further it is unlikely that the

integration site of the provirus (Sections E, G, and H) is relevant to the

process of immortalization, since different lines have different

integration sites (115). Indeed, spontaneous transformation or

immortalization of primary human lymphocytes has been reported

applying this assay to simian viruses (113). It follows that

immortalization in culture of cells infected by HTLV-I is an extremely

rare, perhaps spontaneous event.

There are several indications that in vitro immortalization and leukemic

transformation are different events and that both do not depend on

HTLV-I: (a) the latent period for immortalization is 30 to 60 days, while

that of leukemogenesis is estimated to be 5 to 10 years; (b) in vitro

immortalized cells are diploid (116), while all leukemic cells have

chromosome abnormalities (Section G); (c ) leukemic cells do not

express virus (Section D) while immortalized cells do (115); (d) cells

that are clonal with regard to viral integration sites are not necessarily

leukemic, because normal T-lymphocytes monoclonal with regard to

HTLV-I integration were observed in 13 nonleukemic Japanese carriers

(112); (e) finally immortalized cell lines with defective viruses (115) or

no viruses (113) indicate that immortalization is a virus-independent,

spontaneous event. The evidence that cat, rat, and rabbit cells are

immortalized, although they are presumably insusceptible to the human

virus (13), endorses this view. It would appear that HTLV-I is directly

involved neither in immortalization nor in transformation (Sections A, B,

G and H). Instead the assay appears to be a direct measure of cell death

of human lymphocytes, due in part to HTLV-I-mediated fusion in vitro

(105, 106), and of rare spontaneous immortalization.

C. Tumor Resistance Genes That Inhibit Tumorigenesis but not Virus

Replication

If the virus were a direct and specific cause of tumori-genesis, one would

expect that all individuals who are permissive for infection would also be

permissive for viral tumors. However, this does not appear to be so. For

example certain inbred lines of chicken like line 7 (117, 118) or line SC

(35, 107) are highly susceptible to induction of lymphoma-tosis by avian

retroviruses, whereas line 151 (32, 119, 120) is highly susceptible to

induction of erythroblastosis by the same avian retroviruses. By contrast

other lines like line 6 (118, 121), line FP (107), or line K28 (122) are

either completely or highly resistant to these leukemias but are just as

susceptible to virus infection and replication as the tumor-susceptible

lines (32, 117, 118, 122, 123). Indeed, both the lymphoma-susceptible

SC chickens and the resistant FP chickens develop early viremias and

hyperplastic B-cell follicles, but only 50% of the SC chickens develop

lymphomas (35, 36). Lymphoma resistance is dominant, indicating that

tumor suppressors are encoded (120, 124). The same genes also appear

to impart resistance to Rous sarcoma (124). By contrast resistance to

erythroblastosis is recessive (Section E).

Analogous tumor resistance genes have been observed in mouse strains.

For instance, resistance of C57BL mice to radiation leukemic virusinduced

leukemia (125) or of AKR X BALB/c mice to AKR virusinduced

leukemia (40) is controlled by the H-2D gene, which is

dominant for resistance. Inoculation of the virus into adult C57BL mice

caused polyclonal B- and T-cell hyperplasia from which most animals

died after 4 to 5 months. However, no leukemia was observed (47).

Clearly the tumor resistance genes of the C57BL mice do not suppress

virus replication but apparently proliferation of transformed cells.

Likewise the SI and the Fv-2 genes of mice inhibit leukemogenesis but

not replication of Friend leukemia virus (13, 16, 126). The fates of

DBA/2 and ST/b mice inoculated neonatally with AKR virus are another

example. After expressing virus for at least 8 months (41), only ST/b

mice show a high incidence (about 80%) of leukemia between 8 and 12

months of age, whereas DBA/2 mice show a lower incidence (about

30%) but only at 2 to 3 years of age. Furthermore, not a single

lymphomania developed during a period of 1 year in chronically viremic

CBA/N mice, inoculated as newborns with Moloney leukemia virus,

signalling an absolute resistance to leukemogenesis (42, 46). By contrast,

about 90% of viremic AKR mice develop leukemia (40, 48). The wide

range of sucsceptibilities to virus-induced leukemia among different

mouse strains inoculated with AKR virus, as originally observed by

Gross (9), probably also reflects postinfection tumor resistance genes in

addition to genes conferring resistance to virus infection and expression

(16).

The over 100-fold variation (from less than 1% to 90%) in the incidence

of mammary carcinomas among mice that are susceptible to the

mammary tumor virus and also contain endogenous MMTVs also

reflects host genetic factors that govern resistance to tumori-genesis (16,

54, 55, 58, 127-129). One set of resistance genes governs virus

expression, as for example the sex of the host, because almost only

females secrete virus and develop tumors (13, 16). Another set governs

resistance to carcinogenesis because virus-induced hyperplasia does not

necessarily lead to mammary tumors (56, 57).

Resistance to tumorigenesis in animals which are permissive for virus

replication indicates that tumors contain nonviral, cellular determinants

or tumor antigens. Moreover defects of tumor resistance genes rather

than viral genes determine tumor specificity since the nature of the tumor

induced by a given virus depends on the host and not on the virus. This

lends new support to the conclusion that viruses are not direct causes of

tumorigenesis.

D. Tumors without Virus Expression, without Complete Viruses, or

without Viruses: Proof that Virus Is Not Necessary to Maintain

Transformation

If the retroviruses encode transformation-specific functions, one would

expect that viral genes are continuously expressed in viral tumors.

However, only 50% of virus-induced avian lymphomas express viral

RNA (130). In many clonal lymphomatoses of chickens only incomplete

or truncated proviruses are found. These defective proviruses lack the 5′

half of the genome and hence are unable to express any viral gene (36,

50, 131, 132).

Moreover neither exogenous nor active endogenous retroviruses can be

detected in some lymphomas. One rare study that investigated

lymphomatosis in lymphomatosis virus-free chickens found that 10 of

about 2000 (0.5%) chickens of line 7 died from lymphomas that were

indistinguishable from viral lymphomas at the ages of 6 to 18 months

(49, 121). Thus the incidence of lymphoma in virus-free chickens is very

similar if not the same as that of chickens infected by lymphomatosis

virus with antiviral immunity (less than 1%) (Section A). Since almost

all chickens contain multiple endogenous retroviruses (16, 133), it may

be argued that these viruses were responsible for the leukemias in

animals free of exogenous virus. However, the evidence that endogenous

viruses were latent in leukemic as in nonleukemic birds indicated that the

endogenous retroviruses were not involved in these spontaneous

lymphomas (121). The existence of endogenous viruses in the

lymphatoma-resistant chickens of line 6 supports this view (121, 133). In

fact, it has been argued that endogenous viruses protect by interference

against infection by exogenous variants (13, 16, 134).

A few cases of mouse T-cell lymphomas with defective leukemia viruses

have also been observed (135-137). These findings indicate that murine

leukemia can also be maintained without expression of retroviral genes.

Expression of mammary tumor virus appears also not necessary to

maintain tumors, because no viral antigens (138) and no virus expression

are detectable in many virus-positive mammary tumors (9, 52, 139) and

because defective proviruses are observed in some tumors (140).

Moreover, in mice which lack mammary tumor virus altogether,

mammary tumors were observed that cannot be distinguished from viruspositive

tumors, indicating that the virus is not necessary to initiate

mouse mammary tumors (141). However, in the absence of virus

expression, mammary carcinomas develop at lower incidence and after

longer latent periods (9, 16, 52, 139-142).

Among virus-positive feline leukemias, some contain only defective

proviruses, as in the avian system (143-145). However, about 25 to 35%

of all feline leukemias are free of virus, viral antigens (67, 68, 110), and

proviral DNA (143-145). This is significantly higher than the percentage

of virus-free avian lymphomas. In some virus-free leukemias, the

presumably lymphotropic virus is believed to be in other cells of the cat

(65).

In provirus-positive natural bovine and experimental ovine leukemias

expression neither of virus nor of viral RNA have been detected (75,

146). This result is at odds with the proposal, based on in vitro evidence,

that the virus encodes a protein that activates virus transcription and

expression of latent cellular transforming genes (147). In addition, the 5′

half of bovine leukemia provirus is absent from 25% of bovine

leukemias (146, 148). This entirely prevents expression of all viral genes.

Other investigators have described that 30% of bovine leukemias are

virus free (72).

The proviruses of HTLV-I associated with human T-cell leukemias are

also consistently latent. For instance, no expression of viral antigens

(149) and no transcription of viral RNA are observed in freshly isolated

leukemic T-cells from (5 of 6) HTLV-I positive patients with human Tcell

leukemia (150, 151). Again, this is incompatible with the in vitro

evidence for a viral transcriptional activator that was proposed to activate

virus expression and expression of latent cellular transforming genes

(152, 153) (Section H). Moreover, about 10% of the ATLV- or HTLV-Ipositive

adult T-cell leukemias from Japan contain only defective viruses

(77, 151, 154). Since the 5′ half of the viral genome was reported to be

missing no viral gene expression is possible (77, 151, 155). Further, a

minority of Japanese ATL patients appears to be free of ATLV, based on

the serological assays that are used to detect the virus (156, 157). A

recent analysis found 5 virus-free cases among 69 Japanese ATL

patients, who lacked both HTLV-I provirus and antiviral immunity

(158). Comparisons among T-cell leukemias in Italy found only 2 of 68

(159) or 3 of 16 (160) otherwise identical cases to be HTLV-I positive. A

survey from Hungary found 2 of 326 leukemias antibody positive (161).

Other studies from the United States and Italy describe HTLV-I-free Tcell

leukemias that share chromosome abnormalities with viral leukemias

(Section H). Thus, the ratio of nonviral to viral T-cell leukemias in

humans outside Japan appears to be even higher than that of nonviral to

viral feline and bovine leukemias.

Since retrovirus expression is not observed in many virus-positive

leukemias and since only defective viruses are associated with some

leukemias it follows that viral gene products are not necessary to

maintain these leukemias. These tumors must be maintained by cellular

genes (Section H). The occurrence of “viral” leukemias of chicken, mice,

cats, cattle, and humans despite antiviral immunity (Section A) supports

this conclusion. This conclusion is also consistent with the evidence that

about 30% of the natural feline and bovine leukemias as well as many

human and some avian leukemias and murine mammary carcinomas are

virus free, yet these tumors cannot be distinguished from viral.

E. Transformation Not Dependent on Specific Proviral Integration Sites

Since retroviruses without onc genes are not sufficient to cause tumors

and do not encode transformation-specific functions (Sections A-C) but

may nevertheless induce experimental tumors (Section A), several

hypothetical mechanisms of viral carcino-genesis have been proposed

that each require a specific interaction with the host cell (Section H). One

of these postulates is that retroviruses without onc genes activate latent

cellular cancer genes, termed proto-onc genes, by site-specific proviral

integration (13, 16, 130, 162). The proposal is based on structural

analogy with retroviral onc genes, which are hybrids of sequences

derived from retroviruses and proto-onc genes (5, 19, 20). It is termed

downstream promotion hypothesis (130) because the promoter of the 3′

long terminal repeat from the provirus is thought to promote

transcription of a proto-onc gene downstream.

It is consistent with this hypothesis that leukemias and other tumors from

retrovirus-infected animals and humans are typically all monoclonal with

regard to the integration sites of the provirus in the host chromosome.

However, if one compares different monoclonal tumors of the same cell

lineage, different integration sites are found in each individual tumor.

This has been documented for retroviral lymphomas of chickens (37,

131, 132), mice (13, 163, 164), cats (143-145), cattle (146, 148), and

humans (13, 151, 154, 155, 165) and also for mammary tumors of mice

(13). It is unlikely that the mutant genes generated by provirus

integrations are transforming genes, because they are not specific and not

known to have transforming function upon transfection. Instead the

clonal proviral integration sites of individual tumors appear to be the

consequence of clonal proliferation of a single transformed cell from

which the clonal tumor originated (Section G).

Relevance of Preferred Integration Regions. Although the search for

specific proviral integration sites in viral tumors has met with no success,

preferred integration regions were observed in three systems, namely in

erythro-blastoses and lymphomas of chicken strains predisposed to these

tumors and in mammary tumors of mice bred for susceptibility to this

tumor (13, 16). For instance in erythroblastosis-prone 15I chickens that

suffer 80% erythroblastosis upon infection (120), integration upstream of

proto-erb was observed in 90% (119) and 45% (120, 122) of

erythroblastoses. Proto-erb is a proto-onc gene because it is the cellular

progenitor of the transforming gene of avian erythroblastosis virus (13,

19). This region-specific integration appears to activate proto-erb

transcription compared to certain normal controls (119). However, there

are as yet no data on activation of proto-erb translation in leukemic cells.

Unexpectedly 45% of the erythroblastoses observed in 15I chickens

contained viruses with transduced proto-erb (122). The outstanding yield

of proto-erb transductions in this line of chicken compared to others (5,

19) (Section H) suggests an altered proto-erb gene, perhaps already

flanked by defective proviral elements which would permit transduction

via homologous recombination. It is consistent with this view that in 15I

chickens susceptibility in erythroblastosis is dominant (120), while

typically resistance to tumors is dominant in chickens and mice (Section

C).

Further in about 85% of the viral lymphomas of lymphoma-prone

chicken lines (Section C) transcription of the proto-myc gene is activated

compared to certain controls (130). Proto-myc is a proto-onc gene

because it is the cellular progenitor of the transforming genes of four

avian carcinomas viruses, MC29, MH2, CMII, and OK10 (5, 13, 19).

Transcriptional myc activation ranges from 300- to 500-fold in some

lymphoma lines (RP) to 30-to 100-fold in most primary lymphomas

(85%) down to undetectable levels in a few (6%) primary lymphomas

(130). However, the activation of proto-myc translation, compared to

normal fibroblasts, was estimated as only 7-fold in one RP lymphoma

line and even lower in three other lines (166). Assuming that the same

ratios of transcriptional to translational activation apply to all

lymphomas, activation of myc translation would be only 1- to 2-fold in

most lymphomas, hardly enough to explain carcinogenesis. In 5 to 15%

of the lymphomas there is no detectable transcriptional activation of

proto-myc and the retroviruses appear to be integrated outside of and in

random orientation relative to the proto-myc genes (50, 105, 130, 132,

167, 168, 169).

Thus, in lymphomas, proto-myc transcription is frequently but not

always activated whereas proto-myc translation appears to be barely, if at

all activated. It is not known whether translation of proto-erb is activated

in viral erythroblastoses. By contrast viral myc and erb genes are

efficiently translated in all virus-transformed cells (5, 13, 16, 19, 20).

Moreover in contrast to the hypothetical lymphoma specificity of

activated proto-myc, viral myc genes typically cause carcinomas and

viral erb genes cause sarcomas in addition to erythroblastosis (5, 13).

Integration of mostly intact murine leukemia viruses into or upstream of

proto-myc is also observed in mouse and rat lymphomas. But since it

occurs only in 10 (170, 171) to 65% (172) of the cases analyzed, it is not

necessary for lymphoma-genesis. Moreover provirus integration near

murine proto-myc is also not sufficient for leukemogenesis. Virus

integrated near proto-myc was found in 15% of the hyperplastic thymus

colonies of AKR mice that appeared 35 days after infection with MCF

virus. These colonies were not tumorigenic (172). However, more

malignant lymphomas develop from cells with provirus integrated near

myc than from other cells, because in 65% of the lymphomas virus was

integrated in proto-myc.

There are also preferred regions of provirus integration for MMTV in

carcinomas of mice, termed int-1 in C3H mice and int-2 in BR6 mice

(13, 16). The int loci or genes are considered to be proto-onc genes only

because they are preferred MMTV integration sites. They have not been

progenitors of viral onc genes and there is no direct evidence that they

can be activated to cellular cancer genes. Moreover transcriptional

activation of int is observed only in some tumors (173) and there is no

evidence for viral-int hybrid mRNAs (140). It is also not known whether

the int loci are coding. The two int loci are totally unrelated to each other

and map on different chromosomes (174). Integration within the int

regions is neither site nor orientation specific with regard to the int loci

(13). Integration at int loci is also not necessary for carcino-genesis,

because integration in int-1 is found in only a fraction (22 of 26) of C3H

tumors (173) and in int-2 only in a fraction (22 of 45) of BR6 tumors

(140). Further integration in int-1 was found in benign hyperplastic

nodules that did not become malignant, proving that it is also not

sufficient for carcinogenesis (56, 57).

The hypothesis that region-specific integration generates hybrid

transforming genes that are equivalent to viral onc genes is inadequate on

several counts. (a) Region-specific integration is not necessary for

transformation, because in most systems (human, bovine, feline) it is not

observed and in all others it is not obligatory. (b) It is also not sufficient

for carcinogenesis based on the particular cases of clonal murine

leukemia virus integration into proto-myc that did not cause leukemia

(172), clonal MMTV integration into int-1 that did not cause mammary

carcinomas (56, 57), and monoclonal HTLV-I infections that did not

cause T-cell leukemia (112). The non-leukemic proto-myc integration is

incompatible with the model purporting that activated proto-myc is like

the inevitably transforming viral myc genes (5). The prediction that

native proviral-cell DNA hybrids have transforming function, like the

related retroviral onc gene models, is unconfirmed. Attempts to

demonstrate transforming function of proviral-proto-myc hybrids from

chicken lymphomas were negative but led to a DNA with transforming

function termed B-lym (13, 175). A plausible reason is that the myc

RNAs initiated from upstream viral promoters are poor mRNAs because

they start with intron sequences that are not part of normal mRNA and

cannot be spliced out, since there is no splice donor downstream of the 3′

viral long terminal repeat (Section H). (d) The prediction that the

probability of all infected cells to become transformed should be the

same as that of region-specific integration is also unconfirmed on the

basis of the following calculations (5). The proto-myc, -erb, or int

regions that are preferential proviral landing sites in viral tumors

measure about 2 and 40 kilobases, respectively (13). Since the chicken

chromosome contains about 1 x 10 (6th power) kilobases and the mouse

chromosome contains about 3 x 10 (6th power) kilobases, and since

provirus integration is random (13, 16), about 2 in 10 (6th power) or 1 in

10 (5th power) infections should generate a tumor cell, if region-specific

integration were the mechanism of carcinogenesis. Yet the probability

that an infected cell will initiate an monoclonal tumor is only about 10 (-

11th power) (Section F). In addition, the latent period of tumorigenesis

would be expected to be short because there are at least 10 (8th power)

target cells of the respective lineages and many more viruses to infect

them (Section F). Moreover, given the long latent periods of

carcinogenesis, polyclonal rather than monoclonal tumors would be

expected from integrational carcinogenesis. It may be argued that this

discrepancy reflects the work of tumor resistance genes. However,

postinfection resistance genes that suppress tumor formation by the viral

derivatives of proto-myc or erb, like MC29 or avian erythro-blastosis

virus, have never been observed in vivo or in vitro. Clearly, since tumor

resistance genes do not function in vitro it would be expected that at least

2 of 10 (6th power) cells infected in vitro would be transformed by

activation of proto-myc and 2 by activation of proto-erb. However, no

transformation by leukemia viruses has ever been observed in vitro

(Section B).

In view of this, it is more likely that region-specific integration may

provide proliferative advantages to hyperplastic cells or may initiate

hyperplasia by activating or inactivating growth control genes rather than

being the cause of malignancy. This proposal predicts that integration

into proto-myc and proto-erb precedes tumorigenesis (Fig. 1).

It is inconsistent with this proposal that murine leukemia virus

integration into proto-myc (172) and MMTV integration into int-1 (56,

57) occur prior to carcinogenesis and thus are not sufficient for

carcinogenesis. This proposal predicts also that the chicken lines that are

susceptible to lymphoma or erythro-blastosis lack genes that check

hyperplasia of lymphocytes or erythroblasts. It is consistent with this

view that the same retroviruses cause either lymphomatosis or

erythroblastosis or no tumors in different chicken lines. The exclusive

(but not absolute) usage of only one of two different int loci by MMTV,

namely int-1 in carcinomas of C3H mice and int-2 in BR6 mice, is also

more likely to reflect strain-specific activation or inactivation of

proliferative controls than two entirely different transforming genes that

would nevertheless generate indistinguishable carcinomas.

F. The Probability That a Virus-infected Cell Will Become Transformed

Is Only 10 (-11th power)

To calculate the probability that a virus-infected cell will become

transformed, we must consider the ratio of symptomatic to asymptomatic

carriers, the clonality of the viral tumors, and the long latent periods of

oncogenesis. (a) The ratio of symptomatic to asymptomatic carriers with

latent infections and antiviral immunity averages less than 10 (-3rd

power) (Section A), but that of viremic animals susceptible to

transformation may reach 0.9 (Section C). (b) Since monoclonal tumors

emerge from at least 10 (8th power) B- or T-cells (176), the probability of

an infected cell in an animal to become the progenitor of a clonal

leukemia is only about 10 (-8th power). This calculation assumes that all

of these cells are infected. This is certainly true for the mice that carry

AKR virus, radiation leukemia virus (82), or inducible mammary tumor

virus (75, 142) in their germ line, and is probably the case in congenitally

infected viremic chickens, cats, gibbons, and mice (12, 16, 31, 39, 63,

66, 70). In fact in viremic animals, the hyperplastic effect of the virus

would have enhanced the number of prospective tumor cells to at least

10 (9th power) (Sections A and B). Even if only a fraction of susceptible

cells are infected in animals or humans with latent infections and

antiviral immunity, the number of infected cells per host is estimated to

be at least 10 (6th power) to account for the immune response (Section B,

and Refs. 13, 16, 27, 31, and 63) or the proviruses that are used to

diagnose latent virus infection (Section D). Proviruses cannot be detected

biochemically unless they are present in at least 1 of 100 cells. © Finally,

the probability of an infected cell to become transformed in an animal is

a function of the number of generations of infected cells that occur

during the latent period of the disease. Given latent periods of 6 to 120

months (Section B) and assuming an average life span of 1 month for a

susceptible B- or T-cell (176), about 10 to 100 generations of infected

cells are required to generate the one transformed cell from which a

clonal tumor emerges. The corresponding probability that a generation of

cells will develop a clonal tumor would be 10 (-1 power) to 10 (-2

power). Considering the proliferative effect of the virus on hemopoietic

target cells in viremic animals, this may again be a conservative estimate.

Indeed, a mitotic rate of 1 day has been assumed for B-cells of

lymphoma-tosis virus-infected chickens (177).

Thus the probability that a virus-infected, hemopoietic cell will become

transformed in an individual with a latent infection and antiviral

immunity is about 10 (-3 power) x 10 (-6th power) x 10 (-2 power) = 10

(-11th power), and that in a viremic individual without tumor resistance

genes is about the same, namely 0.9 x 10 (-9th power) x 10 (-2nd power) =

10 (-11th power). Therefore the increased risk of viremic animals to

develop leukemia must be a direct consequence of the hyperplasia of

prospective tumor cells (Section A) (Fig. 1). In tumor-resistant animals

the probability that the infected cell will become transformed may be the

same, but the resistance genes would prevent proliferation of the

transformed cells (Section C and H). The apparent probability that virusinfected,

non-hemopoietic cells will become transformed must be lower

in both susceptible and resistant animals, because the incidence of solid

tumors is much lower than that of leukemia (9, 32).

G. Clonal Chromosome Abnormalities Are the Only Transformation-

Specific Markers of Retrovirus-infected Tumor Cells: Causes of

Transformation?

The evidence that viral tumors are monoclonal (Section E) and that

leukemogenesis by retroviruses (without onc genes) is highly dependent

on tumor resistance genes, which are different from genes that determine

susceptibility to the virus, suggest virus-independent steps in

carcinogenesis (Section C). Indeed clonal chromosome abnormalities of

virus-positive mammalian tumors provide direct evidence for cellular

events that may be necessary for carcinogenesis. (Avian cells have not

been studied because of their complex chromosome structure.) For

example, trisomies of chromosomes 15 have been observed frequently in

viral T-cell leukemias of mice (16). In addition translocations between

chromosomes 15, 17, and others have been recorded (108, 178-180,

272). In mammary carcinomas of mice, a chromosome 13 trisonomy was

observed in 15 of 15 cases including inbred GR and C3H mice (which

contain MMTV) and outbred Swiss mice (which probably also contain

the virus) (181). Clonal chromosome abnormalities have also been

observed in 30 of 34 bovine leukemias induced by bovine leukemia virus

(75). A recent cytogenetic analysis of human adult T-cell leukemias

(ATL) from Japan showed that 10 of 11 cases had an inversion or

translocation of chromosome 14 (183). Rearrangements of other

chromosomes have been detected in 6 of 6 (184), 12 of 13 (116), and 8

of 9 cases of HTLV-I-positive leukemias (185). Thus over 90% of viruspositive

T-cell leukemias have chromosome abnormalities. A survey of

all viral T-cell leukemias analyzed shows rearrangements of

chromosome 14 in 26% and of chromosome 6 in 29% (186, 187).

The chromosome abnormalities of these viral leukemias and carcinomas

are as yet the only known determinants that set apart transformed from

normal virus-infected cells. Since the chromosome abnormalities are

clonal, the origin of the tumor must have coincided with the origin of the

chromosome abnormality. Therefore chromosome abnormalities or

closely associated events must be directly relevant to initiation of

tumorigenesis. They could either be, or coincide with, a single step

mechanism of transformation or with one of several steps in

transformation, as postulated in the case of the Philadelphia chromosome

(188). It is consistent with this view that chromosome abnormalities are

found in all virus-infected tumors analyzed.

However, heterogeneity among the karyotypes of individual human or

murine leukemias of the same lineage (16, 179, 182, 189, 190, 272) and

thus heterogeneity of mutation support the view that chromosome

abnormalities are coincidental with rather than causal for transformation.

Yet this view does not take into consideration that together with the

microscopic alterations, other submicroscopic mutations may have

occurred that could have initiated the disease (108). It is consistent with

this view that tumor cells contain in addition to microscopic karyotype

changes submicroscopic deletions, detectable as restriction enzyme site

polymorphisms (191). Some of these mutations may be functionally

equivalent to the truncation-recombination mechanism that activates the

docile proto-onc genes of normal cells to the onc genes of directly

oncogenic retroviruses (5, 192). Thus specific karyotypic changes may

only be the tip of the iceberg of multiple chromosomal mutations,

referred to as “genequake,” [G. Matioli, personal communication] which

must have occurred in the same cell. One or several of these could have

initiated the tumor. Chromosome recombination sites are also postulated

to be cellular transforming genes of virus-negative tumors, as for

example in Burkitt’s lymnphoma (5) or in human leukemia with the

Philadelphia chromosome (193).

If chromosomal abnormalities are necessary for transformation of cells

infected by retroviruses without onc genes, chromosomal abnormalities

would not be expected in tumors caused by retroviruses with directly

transforming onc genes. This has indeed been confirmed for tumors

caused in mice by Rous sarcoma virus (194) or by Abelson leukemia

virus (195) which have normal karyotypes (Table 1).

The clonality of retrovirus-positive tumors is then defined in two

different ways: by a retroviral integration site (see Section E), and by a

chromosome abnormality (see Fig. 1). Each of these two clonal

chromosome alterations could then mark the origin of the tumor, while

the other must have pre-existed. Since the tumors originate late after

infection and probably from a virus-infected, normal cell, the clonal

retroviral integration site would appear to be a direct consequence of

clonal proliferation of a cell transformed by a chromosome alteration.

Indeed chromosome abnormalities are typical of tumor cells but not of

virus-infected normal cells. This view is consistent with the evidence that

retrovirus integration does not cause transformation and that

transformation is not dependent on specific integration sites. It is also

highly improbable that chromosome abnormalities are caused by the

virus, because they are not found in virus-infected normal cells and

because they are also characteristic of virus-negative tumors (Section H).

The clonal retroviral integration sites in viral tumors the chromosomes of

which have not been analyzed, as for example avian, feline, and simian

leukemias, may indeed signal as yet undetected clonal chromosome

abnormalities.

Virus-independent Transformation in Virus-positive and -negative

Tumors

Several hypotheses postulate that retroviruses play a direct role in

carcinogenesis. One reason is that viruses, seemingly consistent with

Koch’s first postulate, are associated with tumors although frequently in a

latent or defective form. In addition it appears consistent with Koch’s

third postulate that experimental infections with retroviruses may induce

leukemia under certain conditions (see Sections B and C). However,

none of these hypotheses provide an adequate explanation for the fact

that retroviruses are not sufficient to initiate (Sections A to C) and not

necessary to maintain (Sections D and E) transformation and do not

encode a transformation-specific function. Moreover none of these

hypotheses can explain why transformation is initiated with a clonal

chromosome abnormality (Section G) and why tumor specificity is

determined by the host rather than the virus (Sections C and E). The

short-comings of three of these hypotheses are briefly reviewed here.

1. The Oncogene Hypothesis. Huebner (8) and others (9, 82) have

postulated that retroviruses (without onc genes) are direct carcinogens

that include oncogenes, hence the term “oncogene hypothesis” (8). The

hypothesis was based on abundant positive correlations between

retrovirus expression and cancer incidence in laboratory mice and

domestic chickens, which indeed suggested direct viral etiology in

apparent accord with Koch’s third postulate. The hypothesis generalized

that either import of retroviruses from without, or activation of latent

viruses from within, is the direct cause of spontaneous, chemically

induced, or physically induced tumors (8, 9, 82). However, the

hypothesis failed to account for the long latent periods of oncogenesis

and for complete tumor resistance by certain animals that are highly

susceptible to the virus and for host genes that would determine tumor

specificity (Section C). Above all the hypothesis failed to account for the

monoclonality and the chromosome abnormalities of the resulting

tumors.

2. The Hypothesis That Latent Cellular Cancer Genes Are Activated

by Provirus Integration. This hypothesis has been introduced in

Section E. It holds that retroviruses act as direct, albeit inefficient

carcinogens by generating hybrid transforming genes from proviruses

joint with cellular proto-onc genes. Excepting the specific cases

described in Section E, this mechanism makes four clear predictions,

namely: (a) that different transforming genes exist in each tumor,

because each has a different proviral integration site (Section E); (b) that

therefore a large number of tumor resistance genes exist in tumorresistant

animals (Section C); (c ) that provirus-cell hybrid genes are

expressed to maintain transformation; and (d) that virus-transformed

cells exist without chromosome abnormalities, analogous to cells

transformed by retroviruses with onc genes (Section G).

None of these predictions is confirmed, (a) Contrary to the expectation

for many different transforming genes, all virus-positive tumors of a

given lineage are phenotypically highly uniform (Section A). Even virusfree

tumors are indistinguishable from virus-positive tumors of the same

lineage only by the presence of viruses. Examples are the identical

pathologies and pathogeneses of viral and nonviral murine leukemias

(196-198), chicken B-cell lymphomas (121), human T-cell leukemias

(158, 161, 186), and mouse mammary tumors (11, 139, 141, 142)

(Section D). (b) Contrary to expectation only a small set of cellular

resistance genes controls the development of viral tumors in chicken or

mice (13, 16) (Section C). Moreover apparently the same resistance

genes of chickens of line 6 suppress viral and nonviral lymphomas, and

even lymphomas induced by Marek’s virus (124). By contrast chickens

of line 7 that lack these genes are equally susceptible to both (121)

(Section D). Mice provide parallel examples such as in the CBA strain,

which is resistant to spontaneous (9) as well as to viral (46) leukemia

(Section C). (c ) Contrary to expectation for virus-cell hybrid

transforming genes, proviruses are latent or defective and biochemically

inactive in many animal and all bovine and human leukemias (Section

D). (d) Contrary to expectation for viral carcinogenesis all virus-positive

murine, bovine, and human tumors analyzed have chromosome

abnormalities. Further, similar chromosome abnormalities in viral and

nonviral tumors again suggest common cellular transforming genes. For

instance, the same chromosome 15 trisomy is observed in murine

leukemias induced by viruses, chemicals, or radiation (180, 190, 199-

201, 272). In addition virus-positive and virus-free human T-cell

leukemias have common abnormalities in chromosomes 14 and 16 (160,

183, 186, 187, 189, 202, 203). Since all human T-cell leukemias and all

bovine leukemias have chromosome abnormalities but not all are

infected by viruses (Sections D and G), it would appear more likely that

the viruses are coincidental passengers rather than causes of the disease.

3. The Hypothesis That Latent Cellular Cancer Genes Are trans-

Activated by Viral Proteins. This hypothesis postulates that certain

retroviruses directly activate latent cellular transforming genes with a

specific viral protein. This has been proposed for bovine leukemia virus

and human HTLV-I based on in vitro models (147, 152, 153) (see

Section D). However, the hypothesis is unlikely for the following

reasons. Since the putative trans-activation protein of HTLV-I is

essential for replication (204), all cells in which the virus replicates

would expect to be transformed. This is clearly not the case. Further this

gene cannot be relevant for transformation since bovine and human

leukemias in particular do not express viral RNA or protein or cannot

express RNA or protein because of defective proviruses (Section D). In

addition this hypothesis also fails to account for the chromosome

abnormalities found in all bovine and human leukemias (Section G).

Finally both the proviral insertion and the transactivation hypotheses fail

to explain the inevitably long latent periods of viral tumori-genesis

(Section B).

Therefore it is proposed that transformation is a virus-independent event

that must be due to cellular genes (Fig. 1). These genes would be

generated by chromosomal mutations for which chromosome

abnormalities are a macroscopic indicator. This explains the clonal

chromosome abnormalities that could not be predicted by any of the

virus-cancer hypotheses. In a given lineage of cells the number of

cellular genes convertible to transforming genes must be limited since

they cause highly uniform tumors which can be suppressed by a small set

of resistance genes.

Retrovirus-independent transformation resolves the apparent paradox

that tumors occur very seldom in typical natural infections of wild

animals and humans, and then only long after infection, and despite viral

latency and antiviral immunity. It is also consistent with virusindependent

transformation that the probability that an individual virusinfected

cell will become transformed is only 10 (-11th power) and that

this probability is the same in a viremic chicken with a virus-induced

hyperplasia, as in a normal chicken with a latent infection and antiviral

immunity (Section F). The low probability of virus-independent

transformation also explains directly why cells infected by retroviruses

are not transformed in culture, namely because not enough cells can be

maintained for a long enough time to observe spontaneous

transformation. Virus-independent transformation is also compatible

with tumor resistance genes that do not inhibit viral replication or growth

of normal virus-infected cells. In addition it is consistent with the notion

that defects of cellular resistance genes rather than viral genes determine

tumor specificity (Section C).

The role of the virus in tumorigenesis is then limited to the induction of

hyperplasia by activating cellular proliferative functions either from

within or from without via viral antigens or virus-induced growth factors

(13, 16, 46). For this purpose the virus must be expressed at a high titer

or it must have infected a large number of cells, if insertional

mutagenesis of proliferative genes were involved (Section E). This may

be similar to the mechanism whereby DNA viruses induce

transformation, as for example Epstein-Barr virus which is thought to

induce Burkitt’s lymphoma. Exactly like their retroviral counterparts, all

Burkitt’s lymphomas have chromosome abnormalities but not all contain

the virus (5). Thus the role of the retrovirus in carcinogenesis is as

indirect as that of chemical or physical carcinogens.

Alternatively a latent retrovirus may itself be subject to activation by

physical, chemical, or spontaneous events that can induce hyperplasis

and cancer (8, 12, 82) (Fig. 1). The physically activated radiation

leukemia virus (82) or the chemically activated endogenous retroviruses

of mice or chickens (12, 16) are examples. It is uncertain whether under

these conditions the retrovirus is just an indicator or an intermediate of

proliferative activations that may lead to carcinogenesis because

comparable studies with virus-free strains of animals are not available.

The physically or chemically inducible phages or herpes viruses may in

turn be models for this (11, 83).

Little is known about the nature of the hyperplastic cell. The existence of

viral hyperplasias in tumor-resistant animals indicates that the

hyperplastic cell is not neoplastic (Section C). Most hyperplastic cells are

polyclonal with regard to proviral integration sites (118) and are likely to

have a normal karyotype, as has been shown in some cases (47) (Section

C). Hyperplastic cells with normal karyotypes have also been observed

as precursors of radiation leukemia in mice (205). Nevertheless the

evidence for clonality with regard to a proviral integration site in T-cell

hyperplasias (172) and mammary hyperplasias (56, 57) of mice and in Tcells

of healthy humans (112) indicates clonal, possibly virus-induced

alterations that are not sufficient for carcinogenesis. One could speculate

then that hyperplastic cells fall into two classes, those which respond to

viral antigens delivered from within or without (42) and those which

respond to growth control genes altered by provirus integration (Section

E).

Notable exceptions to virus-independent transformations are infections

that generate retroviral transforming genes. However, the probability of

generating a retrovirus with an onc gene is clearly much lower than

integration into a cellular gene (10 (-6th power), Section E) and even

significantly lower than virus-independent transformation (10 (-11th

power), Section F) (273). Only about 50 such viral isolates have been

recorded in history (5, 13, 19). (The frequent erb transductions from the

chicken 15I line are an exception to this rule (Section E).) The generation

of these viruses requires two rare illegitimate recombinations to

transduce a transformation-specific sequence from a cell into a retrovirus

vector (5, 19, 20, 273). However, one illegitimate recombination that

unites the 5′ promoter, translational start sequence, and splice donor of a

retrovirus with a transformation-specific sequence from a cellular protoonc

gene would be enough to generate a functional virus-cell

recombinant onc gene that cannot be replicated. Tumors caused by such

genes are presently unknown. They will be harder to diagnose but are

probably more frequent than the rare, natural tumors containing complete

retroviruses with onc genes (273).

This raises the question of why orthodox integration of a provirus within

a proto-onc gene, like proto-myc, is not observed to transform infected

cells in vivo or in vitro with the predicted probability. Based on the

calculations described in Section E, this probability should be about 1 in

10 (4th power) considering that about 20 proto-onc genes are known from

20 viral onc genes (5, 13, 19). A possible answer is that proviruses

abutting proto-onc genes from the proviral ends rather than from within,

as in viral onc genes (273), provide neither new downstream translational

starts nor splice donors for those coding regions of the proto-onc genes

that are separated from their native start signals by the inserted provirus.

Nevertheless they can provide efficient downstream promoters (130) of

RNAs that may not be translatable.

I. Are Retroviruses a Basis for Cancer Prediction, Prevention, or

Therapy?

In assessing the tumor risk of a retrovirus-infected animal or human,

latent infections must be clearly seperated from chronic, acute, or

viremic infections. The control of virus expression in a given host is a

product of three factors: the virus; the host cell; and the animal. The viral

factor is defined by viral genes and promotors (13, 16, 206). The cellular

factor is defined by genes that encode viral receptors and unknown

suppressors (8, 9, 11-13, 16-18, 82). The animal factor is defined by

antiviral immunity.

By far the most common natural retrovirus infections are latent, chronic

infections that persist in animals and humans in the presence of antiviral

immunity presumably only in a limited number of cells (38, 40, 90, 207).

The leucemia risk of this statistically most relevant group of natural

infections avarages about less than 0.1% in different animal species

(Section A). It is possibly the same as, but certainly not much higher

than, that of uninfected controls (Sections A and D). Thus latent viruses

offer no targets for tumor prevention. The low probability that an

immunocompetent individual will develop chronic viremia and hence

leukemia also suggests that retroviruses carrying therapeutic genes are

not a significant risk as leukemogens.

By contrast the leukemia risk of a viremic animal that survives the early

pathogenic effects of the infection (Section B) can be high barring

tumor-resistance genes (Section A and C). It ranges between 0 and 90%

in different lines of chicken or strains of inbred mice and avarages about

30% in domestic cats. However, outside the laboratory chronic viremias

are very rare and have never been recorded in humans. They result either

from congenital infections in the absence of maternal antibody (Section

A) or from rare, native immunodeficiency (66).

Thus a predictable tumor risk depends entirely on high virus expression

and virus-induced hyperplasia. This risk can be reduced or prevented by

limiting or blocking lymphoblast hyperplasia as for example by

bursectomy or thymectomy (Section A). Alternatively, inoculation of

newborn AKR mice with antiviral antibody was observed to suppress

viremia and subsequent leukemia in 68% (208). It would appear more

practical, however, to breed or select animals with genes that confer

resistance either to the virus or tumorigenesis or both.

Above all, neither active nor latent viruses offer targets for tumor

therapy, since tumors are not maintained and are not directly initiated by

viral genes, and also occur despite active antiviral immunity.

Clearly the cell is the more complex variable in the as yet poorly defined

interaction between retroviruses and cells that leads to hyperplasia and

than carcinogenesis. In view of the evidence for cellular genes that

determine resistance to hyperplasia and tumorgenesis, further progress in

understanding and treating virus-induced cancer will depend on

identifying cellular determinants of carcinogenesis and the function of

hyperplasia and tumor resistance genes.

II. Retroviruses and AIDS

The isolation in 1983 of a retrovirus from a human patient with

lymphoadenopathy, a typical symptom of AIDS, led to the proposal that

the virus, now termed lymphadenopathy-associated virus, is the cause of

AIDS (26). Related viruses, termed HTLV-III, ARV, or HIV (209), have

since been isolated from about one-half of the AIDS patients that have

been sampled (210-214). In the United States about 26,000 AIDS cases

and 15,000 AIDS fatalities have been reported between 1981, when the

disease was first identified (215), and October 1986 (216). Women

represent only 7% of the AIDS cases in the United States (216). The

number of AIDS cases reported in the United States has increased from

about 100 per 6-month period in 1981 to about 5,000 during the last

three 6-month periods from January 1985 (216). At the same time the

case fatality rate has declined from a high of 88% in 1981 to 32% in

1986 (216). In absolute numbers the known deaths have declined from a

high of 2,600 in the first 6 months of 1985 to 1,800 in the first 6 months

of 1986. This suggests either that the virulence of the disease is dropping

or that other diseases were diagnosed as AIDS. Recently the virus was

also suggested to cause disease of the brain and of the nervous system

(230, 255, 268, 274) and lymphoid interstitial pneumonia (275).

Antibody to the virus is found in about 90% of AIDS patients and

correlates with chronic latent infection by the virus (217-221). Because

of the nearly complete correlation between AIDS and immunity against

the virus, the virus is generally assumed to be the cause of AIDS (13,

27). Accordingly, detection of antiviral antibody, rather than virus, is

now most frequently used to diagnose AIDS and those at risk for AIDS

(27, 217-224). This is paradoxical, since serum antibody from AIDS

patients neutralizes AIDS virus (225-227) and since antiviral immunity

or vaccination typically protects against viral disease. It is even more

paradoxical that a low antibody titer is equated with a low risk for AIDS

(228, 229).

Unlike all other retroviruses, AIDS viruses are thought to be direct

pathogens that kill their host cells, namely T-lymphocytes (13, 27), and

possibly cells of the brain (230, 255). This view is compatible with the

phenotype of AIDS, the hallmark of which is a defect in T-cells (13, 27,

215), and with experimental evidence that many but not all viral isolates

induce cytopathic fusion of T-lymphocytes under certain conditions in

vitro (Section D). Further it is incompatible with neurological disease

(231, 232, 255). However, cell killing is incompatible with the obligatory

requirement of mitosis for retrovirus replication (16, 25) and with the

complete absence of cytocidal effects in all asymptomatic infections in

vivo (Section D).

A. Infections with No Risk and Low Risk for AIDS Indicate That the

Virus Is Not Sufficient to Cause AIDS

Since their original discoveries in AIDS patients, the virus and more

frequently antibody to the virus have also been demonstrated in a large

group of asymptomatic persons (212, 214). The virus has been estimated

to occur in about 1 to 2 x 10 (6th power) or about 0.5 to 1% of all

Americans (223, 224). In the United States persons at high risk for

infection include promiscuous homosexual and bisexual men, of whom

17 to 67% are antibody positive; intravenous drug users, of whom 50 to

87% are positive; and hemophiliacs, of whom 72 to 85% are positive

according to some studies (13, 218, 223). On the basis of this particular

epidemiology, it was concluded that the virus is not transmitted as cellfree

agent like pathogenic viruses but only by contacts that involve

exchange of cells (13, 27).

In these virus-infected groups the annual incidence of AIDS was found

to average 0.3% (224) and to reach peak values of 2 to 5% (218, 223,

233). However even in these groups there are many more asymptomatic

than symptomatic virus carriers.

Other infected groups appear to be at no risk for AIDS. In Haiti and in

certain countries in Africa antibody-positive individuals range from 4 to

20% of the population, whereas the incidence of AIDS is estimated at

less than 0.01% (223, 229, 234). Several reports describe large samples

of children from Africa who were 20 (228) to 60% (221) antibody

positive and of female prostitutes who were 66 to 80% antibody positive

(221, 235), yet none of these had AIDS. Among male homosexuals and

hemophiliacs of Hungary about 5% are AIDS virus positive, yet no

symptoms of AIDS were recorded (161). Among native male and female

Indians of Venezuela 3.3 to 13.3% have antiviral immunity, but none

have symptoms of AIDS (236). Since these Indians are totally isolated

from the rest of the country, in which only one hemophiliac was reported

to be virus positive (236), the asymptomatic nature of their infections is

not likely to be a consequence of a recent introduction of the virus into

their population. Thus it is not probable that these infections will produce

AIDS after the average latent period of 5 years (Section B).

Since the percentage of virus carriers with symptoms of AIDS is low and

in particular since it varies between 0 and 5% depending on the AIDS

risk group of the carrier, it is concluded that the virus is not sufficient to

cause AIDS and that it does not encode an AIDS-specific function. The

virus is also not sufficient to cause neurological disease, since it has been

detected in the brains of persons without neurological disease and of

healthy persons who had survived transient meningitis (230-232).

Thus the virus appears only rarely compatible with Koch’s third postulate

as an etiological agent of AIDS. It may be argued that the asymptomatic

infections reflect latent infections or infections of only a small

percentage of susceptible cells, compared to presumably acute infections

with symptoms of AIDS. However, it is shown in Section C that

infections of neither symptomatic nor asymptomatic carriers are acute;

instead both are equally latent and limited to a small percentage of

susceptible cells.

Further the observations that some virus carriers are at high and others at

essentially no risk for AIDS directly argue for a cofactor (218, 237) or

else for a different cause for AIDS. The strong bias against women,

because only 2.5% (479 of 17,000 cases) of the sexually transmitted

AIDS cases in the United States are women (216), is a case in point. The

virus-positive but AIDS-negative children and prostitutes of Africa (221)

or Indians from Venezuela (236) are other examples.

B. Long Latent Period of AIDS Incompatible with Short Latent Period of

Virus Replication

The eclipse period of AIDS virus replication in cell culture is on the

order of several days, very much like that of other retroviruses (238). In

humans virus infection of a sufficient number of cells to elicit an

antibody response appears to take less than 4 to 7 weeks. This estimate is

based on an accidental needle-stick infection of a nurse, who developed

antibody 7 weeks later (239), and on reports describing 12 (240) and 1

(232) cases of male homosexuals who developed antibody 1 to 8 weeks

after infection. During this period a mononucleosis-like illness associated

with transient lymphoadenopathy was observed. In contrast to AIDS (see

below), this illness appeared 1 to 8 weeks after infection and lasted only

1 to 2 weeks until antiviral immunity was established. The same early

mononucle-osis-like disease, associated with lymphocyte hyperplasia,

was observed by others in primary AIDS virus infections (234). This is

reminiscent of the direct, early pathogenic effects observed in animals

infected with retroviruses prior to the onset of antiviral immunity (Part I,

Section B).

By contrast the lag between infection and the appearance of AIDS is

estimated from transfusion-associated AIDS to be 2 to 7 years in adults

(220, 223, 241, 242) and 1 to 2 years in children from infected mothers

(220, 223). The most likely mean latent period was estimated to be 5

years in adults (220, 223). Unexpectedly, most of the AIDS viruspositive

blood donors identified in transfusion-associated AIDS

transmission did not have AIDS when they donated blood and were

reported to be in good health 6 years after the donation (220). Likewise

there is evidence that individuals shown to be antibody positive since

1972 have not developed AIDS (228). Further 16 mothers of babies with

AIDS did not have AIDS at the time of delivery but three of them

developed AIDS years later (276). This indicates that the latent period

may be longer than 5 years or that AIDS is not an obligatory

consequence of infection.

In view of the claim that the virus directly kills T-cells and requires 5

years to cause disease, we are faced with two bizarre options: Either 5

year old T-cells die 5 years after infection or the offspring of originally

infected T-cells die in their 50th generation, assuming a generation time

of one month for an average T-cell (176). It may be argued that the virus

is biochemically inactive during the first five years of infection and then

activated by an unknown cause. However, AIDS virus is biochemically

inactive even during the acute phase of the disease (Section C).

Moreover it would be difficult for the retrovirus to become acute five

years after it had induced chronic antiviral immunity.

Because of the 5 year latency between infection and AIDS, the virus has

been likened to the lentiviruses (277), a group of animal retroviruses that

is thought to cause debilitating diseases only after long latent periods

(13) (Part I, Section B). However recently an ovine lentivirus, the visna

or maedi virus of sheep, was shown to cause lymphoid interstitial

pneumonia in 2 to 4 weeks if expressed at a high titer (269). (The same

disease is believed to be caused by AIDS virus in humans (see below)).

Therefore lentiviruses are not models for retroviruses that are only

pathogenic after long latency (Part I, Section B).

Based on the 5-year latent period of the disease and on the assumption

that virus infection is sufficient to cause AIDS, one would expect the

number of AIDS cases to increase to 1 to 2 x 10 (6th power) in the United

States in the next 5 years. The virus has reportedly reached its present

endemic level of 1 to 2 x 10 (6th power) in the United States (223, 224)

since it was introduced there, presumably, less than 10 years ago (27).

Yet the spread of AIDS from 1981 to 1986 has not followed the spread

of virus with a latent period of 5 years. Instead, recent statistics (see

above) indicate no further increases in the number of AIDS cases and a

significant decline in the number of AIDS fatalities in the United States

(216, 244).

Clearly, the long lag between infection and AIDS and the large number

of virus-positive cases in which as yet no AIDS is observed, even after

long latent periods, lead to the conclusion that the virus is not sufficient

to induce AIDS and does not encode an AIDS-specific function. Indeed,

this conclusion is directly supported by genetic evidence against a viral

AIDS gene. Deletion analysis has proved that all viral genes are essential

for replication (28, 245), which requires not more than 1 or 2 days, yet

AIDS follows infection only with an average lag of 5 years and even

then only very rarely.

C. Levels of AIDS Virus Expression and Infiltration Appear Too Low to

Account for AIDS or Other Diseases

If AIDS viruses were pathogenic by killing susceptible lymphocytes, one

would expect AIDS to correlate with high levels of virus infiltration and

expression, because uninfected cells would not be killed by viruses nor

would unexpressed or latent viruses kill cells. As yet no report on virus

titers of AIDS patients has appeared, despite the record interest in the

epidemiology and nucleic acid structure of this virus (13, 27, 223). In

view of the consistent antiviral immunity of AIDS patients and the

difficulties in isolating virus from them (213), the virus titers are

probably low. Titers have been said to range between only 0 and 10 (2nd

power) per ml blood (213, J.A. Levy, personal communication.)

Proviral DNA has been detected in only 15% (9 of 65) AIDS patients; in

the remaining 85% the concentration of provirus, if present, was

apparently too low for biochemical detection (246). Viral RNA was

detected in 50 to 80% of AIDS blood samples. However, among the

positive samples, RNA was found in only less than 1 of 10 (4th power) to

10 (5th power) presumably susceptible lymphocytes (247). The relatively

high ratios of provirus-positive (10 (-2 power) to 10 (-3 power)) to viral

RNA-positive cells (10 (-4 power) to 10 (-5 power)) of AIDS patients

indicate latent infections. Further there is no evidence that the virus titer

or the level of virus infiltration increases during the acute phase of the

disease. It is probably for this reason that cells from AIDS patients must

be propagated several weeks in culture, apart from the host’s immune

system, before either spontaneous (210-214) or chemically induced (248)

virus expression may occur. Further, the AIDS virus is completely absent

from the Kaposi sarcoma (27, 246), which is associated with 15% (216)

to 30% (249) of AIDS cases and is one of the most characteristic

symptoms of the disease.

Similar extremely low levels of virus infiltration and expression were

also recorded in AIDS virus-associated brain disease (274). Likewise, in

interstitial lymphoid pneumonia less than 0.1% of lung cells expressed

viral RNA (275). Indeed there is evidence that even latent virus may not

be necessary for AIDS, since 85% of AIDS patients lack proviral DNA

(246) and since over 10% of AIDS patients have been observed to lack

antiviral immunity (214, 221, 222, 234). Further, in a study from

Germany 3 of 91 AIDS patients were found to be virus free, based on

repeated negative efforts to detect antibody or to rescue virus. [H.

Ruebsamen-Waigmann, personal communication.]

It is concluded then that the AIDS virus infects less than 1%, and is

expressed in less than 0.01%, of susceptible cells both in carriers with or

without AIDS. This raises the question of how the virus could possibly

be pathogenic and responsible for immunodeficiency or other diseases.

For instance even if the virus were to claim its 10 (-4th power) or 10 (-5th

power) share of T-cells that express viral RNA every 24 to 48h, the

known eclipse period of retroviruses, it would hardly ever match or beat

the natural rate of T-cell regeneration (176).

All other viruses function as direct pathogens only if they are

biochemically active and expressed at high levels. For instance, the titers

that correlate with direct pathogenicity for avian retroviruses are 10 (5th

to 12th power) (31, 35, 250) and they are 10 (4th to 7th power) for murine

retroviruses (12, 38, 40, 42, 251) (Section B). Hepatitis viruses reach

titers of 10 (12th to 13th power) when they cause hepatitis (15), and latent

infections are not pathogenic (83). Further, the very low levels of AIDS

virus expression in vivo are difficult to reconcile with reports based on in

vitro studies with synthetic indicator genes that the AIDS virus encodes a

potent transcription-stimulating protein (28, 153, 245). Clearly such

activators are not at work in vivo.

The extremely low virus titers of symptomatic and asymptomatic carriers

also explain why infection by the virus in the United States is essentially

limited to contacts that involve transmission of cells (244) rather than

being transmitted as a cell-free, infectious agent like pathogenic viruses.

For instance, among 1750 health care workers with exposure to AIDS,

only 1 or 2 were found to be antibody positive (252). Another study

failed to find a single antibody-positive person among 101 family

contacts of 39 AIDS patients, all of whom had lived in the same

household with an AIDS patient for at least 3 months (253).

D. AIDS Viruses Not Directly Cytocidal

The AIDS viruses are reported to display in culture a fast cytocidal effect

on primary T-cells within 1 to 2 months after infection (13, 27, 254). The

cytocidal effect was shown to involve cell fusion (27, 238, 254). The

effect is thought to reflect the mechanism of how the virus generates

AIDS after a latent period of 5 years (27, 254).

This is debatable on several grounds: (a) above all, the in vitro assay

cannot account for the large discrepancy between the short latent period

of cell death in vitro and the 5-year latent period of the disease; (b) T-cell

fusion is not observed in vivo in chronic, asymptomatic virus carriers

and not in prospective AIDS patients during the long latent period of the

disease (255), although virus expression is not lower than during the

acute phase of AIDS; © T-cell killing is also not observed in T-cell lines

in vitro (27) and not in primary lymphocytes under appropriate

conditions (238). Further primary lymphocytes infected by AIDS virus

were shown to double every 5 days in cell culture for three weeks; at the

same time the previously latent AIDS virus was activated to high levels

of expression (278); (d) virus strains that do not cause cytopathic fusion

in vitro have been isolated from 7 of 150 AIDS patients. [H. Ruebsamen-

Waigmann, personal communication.] This demonstrates that the fusioninducing

function of the virus can be dissociated from a putative AIDS

function.

Thus T-cell killing by fusion is apparently a cell culture artifact that

depends on the virus strain and the cell used, as has been shown for

many other retroviruses including HTLV-I (Part I, Section B), and not an

obligatory feature of virus infection. As with other retroviruses, fusion

involves binding of viral envelope antigens on the surface of infected

cells with receptors of uninfected cells. Accordingly, fusion is inhibited

by AIDS virus-neutralizing antibody (256). It apparently depends on

high local virus titers that in particular in the case of AIDS are not

observed in vivo. This view of the cell-killing effect also resolves the

apparent contradiction between the postulated cytocidal effects of AIDS

viruses and the obligatory requirement of all retroviruses for mitosis in

order to replicate (16, 25). Indeed AIDS viruses have been reported to

replicate without cytocidal effects not only in T-cells but also in human

monocytes and macrophages (257, 278), which share the same virusspecific

receptors (258), and in B-cell lines (259), in fibroblasts (261) in

human brain and the lung (213, 230, 232, 257, 261).

E. No Simian Models for AIDS

Since retroviruses have been isolated from monkeys in captivity with

immunodeficiencies and since experimental viremina can depress

immune functions in monkeys, such systems are considered to be animal

models of human AIDS. For example, 42 of 68 newborn monkeys died

with a broad spectrum of diseases that included runting and

lymphadenopathy 4 to 6 weeks after inoculation with Mason-Pfizer

monkey virus (91). However, this virus has since been found in healthy

macaques (262). More recently a retrovirus termed simian AIDS or

SAIDS was isolated from monkeys with immunodeficiency (92, 262).

Inoculation of three juvenile rhesus monkeys by one isolate was reported

to cause splenomegaly and lymphoadenopathy within 2 to 5 weeks. One

animal became moribund and two others were alive with simian AIDS at

the time of publication (92). However, in another study only transient

lymphadenopathy but no lasting AIDS-like disease was observed in

macaques inoculated with this virus (263). Another simian virus that is

serologically related to AIDS virus, termed STLV-III, was isolated from

immunodeficient macaques and from one macaque with a lymphoma.

Macaques inoculated with blood or tissue samples of the viral lymphoma

died 50 to 60 days later with various diseases (93). However,

asymptomatic infections by the same virus have since been identified in

no less than 50% of wild green monkeys that did not show any

symptoms of a disease (264).

Eight chimpanzees infected with human AIDS virus had not developed

symptoms of AIDS 1.5 years past inoculation (265). However, each

animal developed antiviral immunity about 1 month after infection,

followed by persistent latent infection, as in the human cases (265). A

follow-up of chimpanzees inoculated with sera from AIDS patients in

1983 reports no evidence for AIDS in 1986 although the animals had

developed antibodies to the virus (243).

Several reasons suggest that these experimental infections of monkeys

are not suitable models for human AIDS. Above all, the human virus is

not pathogenic in animals. The diseases induced in monkeys by

experimental infections with simian viruses all occur fast compared to

the 5-year latency for AIDS. Moreover the simian viruses are never

associated with a disease in wild animals. Therefore these diseases

appear to be exactly analogous to the direct, early pathogenic effects

caused by other retroviruses in animals prior to antiviral immunity (see

Part I, Section B), and thus are probably models for the early

mononucleosis-like diseases which occur in humans infected with AIDS

virus prior to antiviral immunity (232, 234, 240) (Section B). Indeed the

persistent asymptomatic infections of wild monkeys with simian

retroviruses appear to be models for the many asymptomatic infections

of humans with AIDS virus or HTLV-I.

F. AIDS Virus as an Indicator of Low Risk for AIDS

The only support for the hypothesis that the AIDS virus causes AIDS is

that 90% of the AIDS patients have antibody to the virus. Thus it would

appear that the virus, at least as an immunogen, meets the first of Koch’s

postulates for an etiological agent. This conclusion assumes that all

AIDS patients from whom virus cannot be isolated (about 50%) (278) or

in whom provirus cannot be demonstrated (85%) and the antibodynegative

cases (about 10%) and the virus-free cases reported in one study

(3%) (Section C) are false negatives. Indeed the diagnosis of AIDS virus

by antibody has recently been questioned on the basis of false positives

(234).

At this time the hypothesis that the virus causes AIDS faces several

direct challenges. (a) First it fails to explain why active antiviral

immunity, which includes neutralizing antibody (225-227) and which

effectively prevents virus spread and expression, would not prevent the

virus from causing a fatal disease. This is particularly paradoxical since

antiviral immunity or “vaccination” typically protects against viral

pathogenicity. It is also unexpected that AIDS patients are capable of

mounting an apparently highly effective, antiviral immunity, although

immunodeficiency is the hallmark of the disease. (b) The hypothesis is

also challenged by direct evidence that the virus is not sufficient to cause

AIDS. This includes (i) the low percentage of symptomatic infections,

(ii) the fact that some infected groups are at a relatively high and others

at no risk for AIDS, (iii) the long latent period of the disease (Section B),

and (iv) the genetic evidence that the virus lacks a late AIDS function.

Since all viral genes are essential for virus replication (28, 245), the virus

should kill T-cells and hence cause AIDS at the time of infection rather

than 5 years later. (c ) The hypothesis also fails to resolve the

contradiction that the AIDS virus, like all retroviruses, depends on

mitosis for replication yet is postulated to be directly cytocidal (Section

D). (d) The hypothesis offers no convincing explanation for the paradox

that a fatal disease would be caused by a virus that is latent and

biochemically inactive and that infects less than 1% and is expressed in

less than 0.01% of susceptible lymphocytes (Section D). In addition the

hypothesis cannot explain why the virus is not pathogenic in

asymptomatic infections, since there is no evidence that the virus is more

active or further spread in carriers with than in carriers without AIDS.

In view of this it seems likely that AIDS virus is just the most common

among the occupational viral infections of AIDS patients and those at

risk for AIDS, rather than the cause of AIDS. The disease would then be

caused by an as yet unidentified agent which may not even be a virus,

since cell-free contacts are not sufficient to transmit the disease. Other

viral infections of AIDS patients and those at risk for AIDS include

Epstein-Barr and cytomegalovirus in 80 to 90% (222, 268), and herpes

virus in 75 to 100%. [D. Purtilo, personal communication.] In addition

hepatitis B virus is found in 90% of drug addicts positive for antibody to

AIDS virus (267). Among these different viruses, retroviruses are the

most likely to be detectable long after infection and hence are the most

probable passenger viruses of those exposed to multiple infectious

agents. This is because retroviruses are not cytocidal and are unsurpassed

in establishing persistent, non-pathogenic infections even in the face of

antiviral immunity. Therefore AIDS virus is a useful indicator of

contaminated sera that may cause AIDS (13, 27) and that may contain

other cell-free and cell-associated infectious agents. It is also for these

reasons that latent retroviruses are the most common nonpathogenic

passenger viruses of healthy animals and humans. For the same reasons,

they are also frequently passenger viruses of slow diseases other than

AIDS like the feline, bovine and human leukemias (see Part I) or

multiple sclerosis (268) in which latent or defective “leukemia viruses”

are occasionally found.

It is concluded that AIDS virus is not sufficient to cause AIDS and that

there is no evidence, besides its presence in a latent form, that it is

necessary for AIDS. However, the virus may be directly responsible for

the early, mononucleosis-like disease observed in several infections prior

to antiviral immunity (Section B). In a person who belongs to the high

risk group for AIDS, antibody against the AIDS virus serves as an

indicator of an annual risk for AIDS that averages 0.3% and may reach

5%, but in a person that does not belong to this group antibody to the

virus signals no apparent risk for AIDS. Since nearly all virus carriers

have antiviral immunity including neutralizing antibody (225-227),

vaccination is not likely to benefit virus carriers with or without AIDS. *

Acknowledgements

I am grateful to R. Cardiff (Davis, CA), K. Cichutek, M. Gardner (Davis, CA), D.

Goodrich, E. Humphries (Dallas, TX), J.A. Levy (San Francisco, CA), F. Lilly (New

York, NY), G. S. Martin, G. Matioli (Los Angeles, CA), E. Noah (Villingen, Germany),

S. Pfaff, W. Phares, D. Purtilo (Omaha, NE), H. Rubin, B. Singer, G. Stent, and R.-P.

Zhou for critical comments or review of this manuscript or both and R.C. Gallo (NIH

Bethesda, MD) for discussions.

For Table 1 and Figure 1 see printed publication.

* The abbreviations used are: RSV, Rous sarcoma virus; AIDS, acquired

immunodeficiency syndrome; HTLV-1, human T-cell leukemia virus; MMTV, mouse

mammary tumor virus; ATLV, adult T-cell leukemia virus; STLV-III, simian T-cell

leukemia virus; ATL, adult T-cell leukemia; MCF, mink cell focus-forming; HIV,

human immunodeficiency virus; ARV, AIDS-associated retrovirus.

** Koch’s postulates define the steps required to establish a microorganism as the cause

of a disease: (a) it must be found in all cases of the disease; (b) it must be isolated from

the host and grown in pure culture; © it must reproduce the original disease when

introduced into a susceptible host; and (d) it must be found present in the experimental

host so infected.

Received 6/2/86; revised 10/14/86; accepted 11/11/86.

(This work was) supported by (OIG) National Cancer Institute Grant CA-39915A-01

and Council for Tobacco Research Grant 1547 and by a scholarship in residence of the

Fogarty International Center, NIH, Bethesda, MD.

[Exponential power is not printed here, so the phrase "10 (3rd power)" indicates

"1,000"]

References

1. Boveri, T. Zur Frage der Entstehung maligner Tumoren. Jena, Germany: Gustav

Fischer Verlag, 1914.

2. Foulds, L. Neoplastic Development, Vols. 1 and 2. New York: Academic Press,

1969.

3. Burnett, M.F. The biology of cancer. In: H. German (ed.), Chromosomes and Cancer,

Vol. 1, p. 21. New York: John Wiley & Sons, Inc., 1974.

4. Nowell, P.C. The clonal evolution of tumor cell populations. Science (Wash. DC),

194: 23-28, 1976.

5. Duesberg, P.H. Activated proto-onc genes: sufficient or necessary for cancer?

Science (Wash.DC), 228: 669-677.

6. Rous, P. The challenge to man of the neoplastic cell. Science (Wash. DC), 157: 24-

28, 1967.

7. Stanley, W.M. Virus-induced neoplasia-outlook for the future. Cancer Res., 20: 669-

830, 1960.

8. Huebner, R.J., and Todaro, G. Oncogenes of RNA tumors viruses as determinants of

cancer. Proc. Natl. Acad. Sci. USA, 64: 1087-1094, 1969.

9. Gross, L. Oncogenic Viruses. Elmsford, NY: Pergamon Press, 1970.

10. Kaplan, H.S. On the natural history of the murine leukemias: Presidential Address.

Cancer Res., 27: 1325-1340, 1967.

11. Lwoff, A. Introduction-RNA viruses and host genome in oncogenesis, pp. 1-11. In:

P. Emmelot and P. Bentvelzen (eds.), RNA Viruses and Host Genome in Oncogenesis,

New York: American Elsevier Co., 1972.

12. Rowe, W.P. Genetic factors in the natural history of murine leukemia virus

infection. Cancer Res., 33: 3061-3068, 1973.

13. Weiss, R., Teich, N., Varmus, H., and Coffin, J. RNA Tumor Viruses. Ed. 2. Cold

Spring Harbor, NY: Cold Spring Harbor Press, 1985.

14. Tooze, J. The Molecular Biology of Tumor Viruses. Cold Spring Harbor, NY: Cold

Spring Harbor Press, 1973.

15. Fenner, F., McAuslan, B.R., Mims, C.A., Sambrook, J., and White, D.O. The

Biology of Animal Viruses. New York: Academic Press, 1974.

16. Weiss, R., Teich, N., Varmus, H., and Coffin, J. RNA Tumor Viruses. Ed. 2. Cold

Spring Harbor, NY: Cold Spring Harbor Press, 1982.

17. Levy, J.A. Xenotropic type C viruses. Curr. Top. Microbiol. Immunol., 79: 111-

123, 1978.

18. Kozak, C., and Silver, J. The transmission and activation of endogenous mouse

retroviral genomes. TIG Rev., 1: 331-334, 1985.

19. Duesberg, P.H. Retroviral Transforming genes in normal cells? Nature (Lond.),

304: 219-225, 1983.

20. Duesberg, P.H. Transforming genes of retroviruses. Cold Spring Harbor Symp.

Quant. Biol., 44: 13-29, 1980.

21. Duesberg, P.H., and Vogt, P.K. Differences between the ribonucleic acid of

transforming and nontransforming avian tumor viruses. Proc. Natl. Acad. Sci. USA, 67:

1673-1680, 1970.

22. Duesberg, P.H., Vogt, P.K., Beemon, K., and Lai, M. Avian RNA tumor viruses:

mechanism of recombination and complexity of the genome. Cold Spring Harbor

Symp. Quant. Biol., 39: 847-857, 1974.

23. Wang, L.H., Duesberg, P.H., Beemon, K., and Vogt, P.K. Mapping RNase T-1-

resistant oligonucleotides of avian tumor virus RNAs: sarcoma-specific

oligonucleotides are near the poly(A) end and oligonucleotides common to sarcoma and

transformation-defective viruses are at the poly(A) end. J. Virol., 16: 1051-1070, 1975.

24. Wang, L.H., Galehouse, D., Mellon, P., Duesberg, P., Mason, W.S., and Vogt, P.K.

Mapping oligonucleotides of Rous sarcoma virus RNA that segregate with polymerase

and group-specific antigen markers in recombinants. Proc. Natl. Acad. Sci. USA, 73:

3952-3956, 1976.

25. Rubin, H., and Temlin, H.M. A radiological study of cell-virus interaction in the

Rous sarcoma. Virology, 7: 75-91, 1958.

26. Barre-Sinoussi, F., Chermann, J.C., Rey, F., Nugeyre, M.T., Chamaret, S., Gruest,

J., Danguet, C., Axler-Blin, C., Vezinet-Brun, F., Rouzioux, C., Rosenbaum, W., and

Montagnier, L. Isolation of T-lymphotropic retrovirus from a patient at risk for acquired

immune deficiency syndrome (AIDS). Science (Wash. DC), 220: 868-870, 1983.

27. Wong-Staal, F., and Gallo, R.C. Human T-lymphotropic retroviruses. Nature

(Lond.), 317: 395-403, 1985.

28. Sodroski, J., Goh, W.C., Rosen, C., Dayton, H., Terwilliger, E., and Haseltine, W.

A second post-transcriptional trans-activator gene required for HTLV-III replication.

Nature (Lond.), 321: 412-417, 1986.

29. Ellermann, V., and Bang, O. Experimentelle leukaemie bei Huehnern. Centralblatt

Bakteriol. Parasitenk. Abt. I Orig., 46: 595-609, 1908.

30. Burmester, B.R., Gentry, R.F., and Waters, N.F. The presence of the virus of

visceral lymphomatosis in embrionated eggs of normal appearing hens. Science (Wash.

DC), 34: 609-617, 1955.

31. Rubin, H., Fanshier, L., Cornelius, A., and Hughes, W.F. Tolerance and immunity

in chickens after congenital and contact infection with avian leukosis virus. Virology,

17: 143-156, 1962.

32. Purchase, H.G., and Burmester, B.R. Leukosis sarcoma group. In: M.S. Hofstad et

al. (eds.), Diseases of Poultry, Ed. 6, pp. 502-568. Ames, IA: The Iowa State University

Press, 1972.

33. Calnek, B.W. Neoplastic diseases. Introduction. In: M.S. Hofstad et al. (eds.),

Diseases of Poultry, Ed. 6, pp. 467-470. Ames, IA: The Iowa State University Press,

1972.

34. Calnek, B.W., and Witter, R.L. Marek’s disease. In: M.S. Hofstad et al. (eds.),

Diseases of Poultry, Ed. 6, pp. 470-502. Ames, IA: The Iowa State University Press,

1972.

35. Baba, T.W., and Humphries, E.H. Avian leukosis virus infection: analysis of

viremia and DNA integration in susceptible and resistant chicken lines. J. Virol., 51:

123-130, 1984.

36. Baba, T., and Humphries, E.H. Formation of a transformed follicle is necessary but

not sufficient for development of an avian leukosis virus-induced lymphoma. Proc.

Natl. Acad. Sci. USA, 82: 213-216, 1985.

37. Neiman, P., Payne, L.N., and Weiss, R.A. Viral DNA in bursal lymphomas induced

by avian leukosis viruses. J. Virol., 34: 178-186, 1980.

38. Gardner, M.B. Type C viruses and wild mice: characterization and natural history of

amphotropic, ecotropic, and xenotropic MuLV. Curr. Top. Microbiol. Immunol., 79:

215-259, 1978.

39. Gardner, M.B., Rasheed, S., Estes, J.D., Casagrande, J., Ihle, J.N., and Stephenson,

J.R. The history of viruses and cancer in wild mice. Cold Spring Harbor Conf. Cell

Proliferation, 7: 971-987, 1980.

40. Lilly, F., Duran-Reynals, M.L., and Rowe, W.P. Correlation of early murine

leukemia virus titer and H-2 type of spontaneous leukemia in mice of the BALB X

AKR cross: a genetic analysis. J. Exp. Med., 141: 882-889, 1975.

41. Melamedoff, M., Lilly, F., and Duran-Reynals, M.L. Suppression of endogenous

murine leukemia virus by maternal resistance factors. J. Exp. Med., 158: 506-514,

1983.

42. Lee, J.C., and Ihle, J.N. Chronic immune stimulation is required for Moloney

leukemia virus-induced lymphomas. Nature (Lond.), 289: 407-409, 1981.

43. Pluznik, D.H., and Sachs, L. Quantitation of a murine leukemia virus with a spleen

colony assay. J. Natl. Cancer Inst., 33: 535-546, 1964.

44. Duesberg, P.H., and Robinson, W.S. Nucleic acid and proteins isolated from the

Rauscher mouse leukemia virus (MLV). Proc. Natl. Acad. Sci. USA, 55: 219-227,

1966.

45. Klement, V., and Nicolson, M.O. Methods for assays of RNA tumor viruses.

Methods Virol., 6: 60-108, 1977.

46. Ihle, J.N., and Lee, I.C. Possible immunological mechanisms in C-type viral

leukemogenesis in mice. Curr. Top. Microbiol. Immunol., 98: 85-100, 1981.

47. Mosier, D.E., Yetter, R.A., and Morse, H.C. Retrovirus induction of acute

lymphoproliferative disease and profound immunosuppression in adult C57BL/6 mice.

J. Exp. Med. 161: 766-784, 1985.

48. Rowe, W.P., and Pincus, T. Quantitative studies of naturally occurring murine

leukemia virus infection of AKR mice. J. Exp. Med., 135: 429-436, 1972.

49. Crittenden, L.B. New hypotheses for viral unduction of lymphoid leukosis in

chickens. Cold Spring Harbor Conf. Cell Proliferation, 7: 529-541, 1980.

50. Robinson, H.L., and Gagnon, G.C. Patterns of proviral insertion and deletion in

avian leukosis virus-induced lymphomas. J. Virol., 57: 28-36, 1986.

51. Little, C.C., per staff of Roscoe B. Jackson Memorial Laboratory. The existence of

non-chromosomal influence in the incidence of mammary tumors in mice. Science

(Wash. DC), 78: 465-466, 1933.

52. Bittner, J.J. The milk-influence of breast tumors in mice. Science (Wash. DC), 95:

462-463, 1942.

53. Cohen, J.C., and Varmus, H.E. Endogenous mammary tumor virus DNA varies

among wild mice and segregates during inbreeding. Nature (Lond.), 278: 418-423,

1979.

54. Cardiff, R.D., and Young, L.J.T. Mouse mammary tumor biology: a new synthesis.

Cold Spring Harbor Conf. Cell Proliferation, 7: 1105-1114, 1980.

55. Bentvelzen, P., Brinkhof, J., and Westenbrink, F. Expression of endogenous

mammary tumor virus in mice: its genetic control and relevance to spontaneous

mammary carcinogenesis. Cold Spring Harbor Conf. Cell Proliferation, 7: 1095-1104,

1980.

56. Cardiff, R.D., Morris, D.W., Young, L.J.T., and Strange, R. MuMTV genotype,

proto-neoplasia, and tumor progression. In: M.A. Rich, J.C. Hager, and J. Taylor-

Papadimitriou (eds.), Breast Cancer Origins, Detection and Treatment. Proceedings of

the International Breast Cancer Research Conference, London, March 24-28, pp. 156-

166. The Hague: Martinus Nijhoff Publishing, 1985.

57. Pathak, V.K., Strange, R., Young, L.J.T., Morris, D.W., and Cardiff, R.D. A survey

of Int Region DNA rearrangements in C3H and BALB/cfC3H mouse mammary tumor

systems. J. Natl. Cancer Inst., in press, 1987.

58. Altrock, B.W., and Cardiff, R.D. Mouse mammary tumor virus infection: virus

expression and tumor risk potential. J. Nat’l. Cancer Inst., 63: 1075-1080, 1979.

59. Jarrett, W.F.H., Crawford, A.M., Martin, W.B., and Davie, F. Leukemia in the cat: a

virus-like particle associated with leukemia (lymphosarcoma). Nature (Lond.), 202:

566-568, 1964.

60. Roberson, P., Jarrett, W., and Mackey, L. Epidemiological studies of feline

leukemia virus infection I. A serological survey in urban cats. Int. J. Cancer, 15: 781-

785, 1985.

61. Essex, M. Horizontally and vertically transmitted oncorna-viruses of cats. Adv.

Cancer Res., 21: 175-248, 1975.

62. Dorn, C.R., Taylor, D.O.N., Schneider, R., Hibbard, H.H., and Klauber, M.R.

Survey of animal neoplasms in Alameda and Contra Costa counties, California. II.

Cancer morbidity in dogs and cats from Alameda County. J. Nat’l. Cancer Inst., 40:

307-318, 1968.

63. Essex, M. Feline leukemia and sarcoma viruses. In: G. Klein (ed.), Viral Oncology,

pp. 205-229, New York: Raven Press, 1960.

64. Jarrett, W.F., Jarrett, O., Mackey, L., Laird, W., Hardy, W.D., and Essex, M.

Horizontal transmission of leukemia virus and leukemia in the cat. J. Nat’l. Cancer Inst.,

51: 833-841, 1973.

65. Rojko, J.S., Hoover, E.A., Quackenbush, S.L., and Olsen, R.G. Reactivation of

latent feline leukemia virus infection. Nature (Lond.), 298: 385-388, 1972.

66. Gutensohn, N., Essex, M., Francis, D.P., and Hardy, W.D. Risk of humans from

exposure to feline leukemia virus: epidemiological considerations. Cold Spring Harbor

Conf. Cell Proliferation, 7: 699-706, 1980.

67. Anderson, L.J., Jarrett, W.F.M., Jarrett, O., and Laird, H.M. Feline leukemia-virus

infection of kittens: mortality associated with atrophy of the thymus and lymphoid

depletion. J. Nat’l. Cancer Inst., 47: 807-817, 1971.

68. Hardy, W.D., Old, L.J., Hess, P.W., Essex, M., and Cotter, S. Horizontal

transmission of feline leukemia virus. Nat. New Biol., 244: 266-269, 1973.

69. Kawatami, T.G., Sun, L., and McDowell, F.S. Infectious primate type-C virus shed

by healthy gibbons. Nature (Lond.), 268: 448-450, 1977.

70. Kawatami, T.G., and McDowell, F.S. Factors regulating the onset of chronic

myelogenous leukemia in gibbons. Cold Spring Harbor Conf. Cell Proliferation, 7: 719-

727, 1980.

71. Poiesz, B.J., Ruscetti, F.W., Gazdar, A.F., Bunn, P.A., Minna, J.D., and Gallo, R.C.

Detection and isolation of type C retrovirus particles from fresh and cultured

lymphocytes of a patient with cutaneous T-cell lymphoma. Porc. Nat’l. Acad. Sci. USA,

77: 7415-7419, 1980.

72. Miller, J.M., Miller, L.D., Olson, C., and Gillette, K.G. Virus-like particles in

phytohemagglutinin stimulated lymphocyte cultures with reference to bovine

lymphosarcoma. J. Nat’l. Cancer Inst., 43: 1297-1305, 1969.

73. Miller, J.M., and van der Maaten, M.S. The biology of bovine leukemia virus

infection in cattle. Cold Spring Harbor Conf. Cell Proliferation, 7: 901-999, 1980.

74. Mussgay, M., Dietzschold, R., Lorenz, R., Matheka, H.-D., Mattheus, W., Straub,

O.C., Welland, F., Wilesmith, J.W., Frenzel, B., and Kaaden, O.R. Some properties of

bovine leukemia virus, its use in seroepidemiological studies and eradication of the

disease from infected herds. Cold Spring Harbor Conf. Cell Proliferation, 7: 911-925,

1980.

75. Kenyon, S.Y., Ferrer, J.F., McFeely, R.A., and Graves, D.C. Induction of

lymphosarcoma in sheep by bovine leukemia virus. J. Nat’l. Cancer Inst., 67: 1157-

1163, 1981.

76. Hoxie, J.E., Matthews, D.M., and Ames, D.B. Infection of human endothelial cells

by human T-cell virus. Proc. Nat. Acad. Sci. USA, 81: 7591-7595, 1984.

77. Yoshikura, H., Nishida, J., Yoshida, M., Kitamura, Y., Takaku, F., and Ikeda, S.

Isolation of HTLV derived from Japanese adult T-cell leukemia patients in human

diploid fibroblast strain IMR90 and the biological characters of the infected cells. Int. J.

Cancer, 33: 745-749, 1984.

78. Tajima, K., and Kuroishi, T. Estimation of rate of incidence of ATL among ATLV

(HTLV-I) carriers in Kyushu, Japan. Jpn. J. Clin. Oncol., 15: 423-430, 1985.

79. Maeda, Y., Furakawu, M., Takehara, Y., Yoshimura, Y., Miyumoto, K., Matsubara,

T., Morishima, Y., Tajima, K., Okochi, K., and Hinuma, Y. Prevalence of possible

adult T-cell leukemia virus-carriers among volunteer blood donors in Japan: a nationwide

study. Int. J. Cancer, 33: 717-730, 1984.

80. Pan, I.-H., Chung, C.-S., Komoda, H., Mai, J., and Hinuma, Y. Seroepidemiology

of adult T-cell leukemia virus in Taiwan. Jpn. J. Cancer Res., 76: 9-11, 1985.

81. Okochi, K., Sato, H., and Hinuma, Y. A retrospective study on transmission of adult

T-cell leukemia virus by blood transfusion: seroconversion of recipients. Vox Sang.,

46: 245-253, 1984.

82. Kaplan, H.S. Mouse leukemia: interaction of multiple components in a complex

oncogenic system. In: P. Emmelot and P. Bentvelzen (eds.), RNA Viruses and Host

Genome in Onco-genesis, pp. 143-154. New York: North Holland, American Elsevier,

1972.

83. Fenner, F., and White, D.O. Medical Virology. New York: Academic Press, 1970.

84. Carter, J.K., Procter, F.J., and Smith, R.E. Induction of angiosarcomas by ringnecked

pheasant virus. Infect. Immun., 40: 310-319, 1983.

85. Carter, J.K., and Smith, R.E. Rapid induction of hyperthyroidism by an avian

leukosis virus. Infect. Immun., 40: 795-805, 1983.

86. Smith, R.E., and Morgan, J.H. Identification of plaque isolates of an avian

retrovirus causing rapid and slow onset osteopetrosis. Virology, 119: 488-499, 1982.

87. Morgan, J.H., and Smith, R.E. Rapid induction of osteopetrosis by subgroup E.

recombinant viruses. Virology 129: 493-500, 1983.

88. Carter, J.K., Ow, C.L., and Smith, R.E. Rous-associated virus type 7 induces a

syndrome in chickens characterized by stunting and obesity. Infect. Immun., 39: 410-

422, 1983.

89. Smith, R.E., and Van Eldik, J. Characterization of the immune suppression

accompanying virus-induced avian osteopetrosis. Infect. Immun., 22: 452-461, 1978.

90. Rojko, J.S., Hoover, E.A., Lawrence, E.M., Richard, G.O., and Schaller, J.P.

Pathogenesis of Experimental Feline Leukemia Virus Infection. J. Nat’l. Cancer Inst.,

63: 759-765, 1979.

91. Fine, D.L., Landon, J.C., Pienta, R.J., Kubicek, M.T., Valerio, M.G., Leob, W.F.,

and Chopra, H.C. Responses of infant rhesus monkeys to inoculation with Mason-

Pfizer monkey virus materials. J. Nat’l. Cancer Inst., 54: 651-658, 1975.

92. Marx, P.A., Maul, D.H., Osborn, K.G., Lerce, N.W., Moody, P., Lowenstine, L.J.,

Henrickson, R.V., Arthur, L.O., Gravell, M., London, W.T., Lever, J.L., Levy, J.A.,

Munn, P., and Gardner, M.B. Simian AIDS: isolation of a type D retrovirus and disease

transmission. Science (Wash. DC), 223: 1083-1086, 1984.

93. Daniel, M.D., Letvin, N.L., King, N.W., Kannagi, M., Sehgal, P.K., Hunt, R.D.,

Kanki, P.J., Essex, M., and Desrosiers, R.C. Isolation of T-cell tropic HTLV-III-like

retrovirus from macaques. Science (Wash. DC), 228: 1201-1204, 1985.

94. Breindel, M., Harbers, K., and Jaenisch, R. Retrovirus-induced lethal mutation in

collagen I gene of mice is associated with an altered chromatin structure. Cell, 38: 9-16,

1984.

95. King, W., Patel, M.D., Lobel, L.I., Goff, S.P., and Nguyen-Huu, M.C. Insertion

mutagenesis of embryonal carcinoma cells by retroviruses. Science (Wash. DC), 228:

554-558, 1985.

96. Purchase, H.G., Ludford, C., Nazerian, K., and Cox, H.W. A new group of

oncogenic viruses: reticuloendotheliosis, chick syncytial, chick infectious anemia, and

spleen necrosis virus. J. Nat’l. Cancer Inst., 51: 489-499, 1973.

97. Keshet, E., and Temin, H.M. Cell killing by spleen necrosis virus is correlated with

a transient accumulation of spleen necrosis virus DNA. J. Virol., 31: 376-388, 1979.

98. Graf, T. A plaque assay for avian RNA tumor viruses. Virology, 50: 567-578, 1972.

99. Klement, V., Rowe, W.P., Hartley, J.W., and Pugh, W.E. Mixed culture

cytopathogenicity: a new test for growth of murine leukemia viruses in tissue culture.

Proc. Nat’l. Acad. Sci. USA, 63: 753-758, 1969.

100. Hartley, J.W., Wolford, N.K., Old, L.J., and Rowe, W.P. A new class of murine

leukemia virus associated with development of spontaneous lymphomas. Proc. Nat’l.

Acad. Sci. USA, 74:

789-792, 1977.

101. Rangan, S.R.S., Ueberhorst, P.J., and Wong, M.C. Syncytial giant cell focus assay

for viruses derived from feline leukemia and a simian sarcoma. Proc. Soc. Exp. Biol.

Med., 142: 1077-1082, 1973.

102. Rangan, S.R.S., Wong, M.C., Ueberhorst, P.J., and Ablashi, D.J. Mixed culture

cytopathogenecity induced by virus preparations derived from simian sarcoma virus

(SSV-1) infected cultures. J. Nat’l. Cancer Inst., 49: 571-577, 1972.

103. Rand, K.H., Long, C.W., Wei, T.T., and Gilden, R.V. Plaque assay for the Mason-

Pfizer monkey virus. J. Nat’l. Cancer Inst., 53: 449-452, 1974.

104. Diglio, C.A., and Ferrer, J.P. Induction of syncytia by the bovine C-type leukemia

virus. Cancer Res., 36: 1056-1057, 1976.

105. Nagy, K., Clapham, P., Cheingsong-Popov, R., and Weiss, R.A. Human T-cell

leukemia virus type I: induction of syncytia and inhibition by patient’s sera. Int. J.

Cancer, 32: 321-328, 1983.

106. Hoshino, H., Shinoyama, M., Miwa, M., and Sugimura, T. Detection of

lymphocytes producing a human retrovirus associated with adult T-cell leukemia by

syncytia induction assay. Proc. Nat’l. Acad. Sci. USA, 80: 7337-7341, 1983.

107. Baba, T.W., and Humphries, E.H. Differential response to avian leukosis virus

infection exhibited by two chicken lines. Virology, 135: 181-188, 1985.

108. Rich, M.A. Virus-induced murine leukemia. In: M.A. Rich (ed.), Experimental

Leukemia, pp. 15-49. New York: Appleton-Century-Crofts, 1968.

109. Bentvelzen, P. The putative mam gene of the murine mammary tumor virus. In: G.

Klein (ed.), Advances in Viral Oncology, pp. 141-152. New York: Raven Press, 1982.

110. Hardy, W.D., McClelland, A.J., Zuckerman, E.E., Snyder, H.W., MacEwen, E.G.,

Francis, D.P., and Essex, M. Immunology and epidemiology of feline leukemia virus

nonproducer lympho-sarcomas. Cold Spring Harbor Cancer Conf. Cell Proliferation, 7:

677-697, 1980.

111. Kinoshita, K., Hino, S., Amagasaki, T., Yamada, Y., Kamihira, E., Ichimura, M.,

Munehisa, T., and Hinuma, Y. Development of adult T-cell leukemia-lymphoma (ATL)

in two anti-ATL-associated antigen-positive healthy adults. Gann, 73: 684-685, 1982.

112. Kinoshita, K., Amagasaki, T., Yamada, Y., Kamihira, E., Ichimura, M., Kamihira,

S., Yamada, Y., Momita, S., Kusano, M., Morikawa, T., Fujita, S., Ueda, Y., Ito, N.,

and Yoshida, M. Preleukemic state of adult T cell leukemia: abnormal T lymphocytosis

induced by human adult T cell leukemia-lymphoma virus. Blood, 66: 120-127,

1985.

113. Gallo, R.C., Wong-Staal, F., Markham, P.D., Ruscetti, F., Kalyanaraman, V.S.,

Ceccherini-Nelli, L., Dalla Favera, R., Josephs, S., Miller, N.R., and Reitz, M.S. Recent

studies with infectious primate retroviruses: hybridization to primate DNA and some

biological effects on fresh human blood leukocytes by simian sarcoma virus and gibbon

ape leukemia virus. Cold Spring Harbor Conf. Cell Proliferation, 7: 753-773, 1980.

114. Miyoshi, I., Yoshimoto, S., Kubonishi, I., Taguchi, H., Shiraishi, Y., Oktsuki, Y.,

and Akagi, T. Transformation of normal human cord lymphocytes by co-cultivation

with a lethally irradiated human T-cell line carrying type C virus particles. Gann, 72:

997-998, 1981.

115. DeRossi, A., Aldovini, A., Franchini, G., Mann, D., Gallo, R.C., and Wong-Staal,

F. Clonal selection of T lymphocytes infected by cell-free human T-cell

leukemia/lymphovirus type I: parameters of virus integration and expression. Virology,

143: 640-645, 1985.

116. Nowell, P.C., Finan, J.B., Clark, J.W., Sarin, P.S., and Gallo, R.C. Karyotypic

differences between primary cultures and cell lines from tumors with the human T-cell

leukemia virus. J. Nat’l. Cancer Inst., 73: 849-852, 1984.

117. Crittenden, L.B. Two levels of genetic resistance to lymphoid leukosis. Avian Dis.,

19: 281-292, 1975.

118. Fung, Y.-K.T., Fadly, A.M., Crittenden, L.B., and Kung, H.-J. Avian lymphoid

leukosis virus infection and DNA integration in the preleukotic bursal tissues: a

comparative study of susceptible and resistant lines. Virology, 119: 411-421, 1982.

119. Raines, M.A., Lewis, W.G., Crittenden, L.B., and Kung, H.-J. c-erb B activation in

avian leukosis virus-induced erythroblastosis: clustered integration sites and the

arrangement of provirus in the c-erb B alleles. Proc. Nat’l. Acad. Sci. USA, 82: 2287-

2291, 1985.

120. Robinson, H.L., Miles, B.D., Catalano, D.E., Broles, W.E., and Crittenden, L.B.

Susceptibility to erb B-induced erythroblastosis is a dominant trait of 15I chickens. J.

Virol., 55: 617-622, 1985.

121. Crittenden, L.B., Witter, R.L., Okazaki, W., and Neiman, P.E. Lymphoid

neoplasms in chicken flocks free of infection with exogenous avian tumor viruses. J.

Nat’l. Cancer Inst., 63: 191-200, 1979.

122. Miles, B.D., and Robinson, H.L. High-frequency transduction of c-erb B in avian

leukosis virus-induced erythroblastosis. J. Virol., 54: 295-303, 1985.

123. Purchase, H.G., Gilmour, P.G., Romero, C.H., and Okazaki, W. Post-infection

genetic resistance to avian lymphoid leukosis resides in B target cells. Nature (Lond.),

270: 61-62, 1977.

124. Bacon, L.D., Witter, R.L., Crittenden, L.B., Fadly, A., and Motta, J. B-haplotype

influence on Marek’s disease, Rous sarcoma, and lymphoid leukosis virus-induced

tumors in chickens. Poult. Sci., 60: 1132-1139, 1981.

125. Meruelo, D., Nimelstein, S.H., Jones, P.P., Lieberman, M., and McDevitt, H.O.

Increased synthesis and expression of H-2 antigens on thymocytes as a result of

leukemia virus infection: a possible mechanism for H-2 linked control of virus-induced

neoplasia. J. Exp. Med., 147: 470-487, 1978.

126. Evans, L.H., Duesberg, P.H., and Scolnick, E.M. Replication of spleen focusforming

Friend virus in fibroblasts form C57BL mice that are genetically resistant to

spleen focus formation. Virology, 101: 534-539, 1980.

127. Dux, A. Genetic aspects in the genesis of mammary cancer. In: P. Heston, W.E.

Genetic factors in tumorigenesis. In: P. Emmelot and P. Bentvelzen (eds.), RNA

Viruses and Host Genome in Oncogenesis, pp. 301-308. New York: American Elsevier,

1972.

128. Heston, W.E. Genetic factors in tumorigenesis. In: P. Emmelot and P. Bentvelzen

(eds.), RNA Viruses and Host Genome in Oncogenesis, pp. 13-24. New York:

American Elsevier, 1972.

129. Moore, D.H., Long, C.A., Vaidya, A.B., Sheffield, J.B., Dion, A.G., and

Lasfargues, E.Y. Mammary tumor viruses. Adv. Cancer Res., 29: 347-418, 1979.

130. Hayward, W.S., Neel, B.G., and Astrin, S.M. Activation of a cellular onc gene by

promoter insertions in ALV-induced lymphoid leukosis. Nature (Lond.), 290: 475-480,

1981.

131. Neel, G.B., Hayward, W.S., Robinson, H.L., Fong, J., and Astrin, S.M. Avian

leukosis virus-induced tumors have common proviral integration sites and synthesize

discrete new RNAs: oncogenesis by promoter insertion. Cell, 23: 323-324, 1981.

132. Payne, G.S., Bishop, J.M., and Varmus, H.E. Multiple arrangements of viral DNA

and an activated host oncogene in bursal lymphoma. Nature (Lond.), 295: 209-214,

1982.

133. Astrin, S.M., Robinson, H.L., Crittenden, L.B., Buss, E.G., Wyban, J., and

Hayward, W.S. Ten genetic loci in the chicken that contain structural genes for

endogenous avian leukosis virus. Cold Spring Harbor Symp. Quant. Biol., 44: 1105-

1109, 1980.

134. Weiss, R.A., and Frisby, D.P. Are avian endogenous viruses pathogenic? In: D.J.

Yohn and J.R. Blakeslee (eds.), Advances in Comparative Leukemia Research 1981,

pp. 303-311. Amsterdam: Elsevier/North-Holland Biomedical Press, Inc., 1982.

135. Horak, I., Lee, J.C., Enjuanes, L., and Ihle, J.N. Characterization of a unique

defective type C-virus associated with a Moloney leukemia virus-induced splenic T-cell

lymphoma cell line. J. Virol., 30: 299-308, 1980.

136. Corcoran, L.M., Adams, J.M., Dunn, A.R., and Cong. S. Murine T lymphomas in

which the cellular myc oncogene has been activated by retroviral insertion. Cell, 37:

113-122, 1984.

137. Tsichlis, P.N., Strauss, P.G., and Kozak, C.A. Cellular DNA region involved in

induction of thymic lymphomas (Mlvi-2) maps to mouse chromosome 15. Mol. Cell

Biol., 4: 997-1000, 1984.

138. McGrath, C.M., Nandi, S., and Young, L.J.T. Relationship between organization

of mammary tumor and the ability of tumor cells to replicate mammary tumor virus and

to recognize growth-inhibitory contact signals in vitro. J. Virol., 9: 367-376, 1972.

139. Bentvelzen, P. Hereditary infections with mammary tumor viruses in mice. In: P.

Emmelot and P. Bentvelzen (eds.), RNA Viruses and Host Genome in Oncogenesis, pp.

309-337. New York: American Elsevier, 1972.

140. Peters, G., Brookes, S., Smith, R., and Dickson, C. Tumorigenesis by mouse

mammary tumor virus: evidence for a common region for provirus integration in

mammary tumors. Cell, 33: 369-377, 1983.

141. Gardner, M.B., Malone, R.W., Morris, D.W., Young, L.J.T., Strange, R., Cardiff,

R.D., Foulkin, L.J., and Mitchell, D.J. Mammary tumors in feral mice lacking MuMTV

DNA. J. Exp. Pathol., 2: 93-98, 1985.

142. Michalides, R., Van Deemter, L., Nusse, R., and Hegeman, P.C. Induction of

mouse mammary tumor virus RNA in mammary tumors of BALB/c mice treated with

urethane, X-irradiation and hormones. J. Virol., 31: 63-72, 1979.

143. Levy, L.S., Gardner, M.B., and Casey, J.W. Isolation of a feline leukaemia

provirus containing the oncogene myc from a feline lymphosarcoma. Nature (Lond.),

308: 853-856, 1984.

144. Mullins, J.L., Brody, D.S., Binasi, R.C., and Cotten, S.M. Viral transduction of cmyc

in naturally occurring feline leukemias. Nature (Lond.), 308: 856-858, 1984.

145. Neil, J.C., Hughes, D., McFarlane, R., Wilkie, N.M., Onions, D.E., Lees, G., and

Jarrett, O. Transduction and rearrangement of the myc gene by feline leukaemia virus in

naturally occurring T-cell leukaemias. Nature (Lond.), 308: 814-820, 1984.

146. Kettmann, R., Deschamps, J., Clerter, Y., Couez, D., Burny, A., and Marbaix, G.

Leukemogenesis by bovine leukemia virus: Proviral DNA integration and lack of RNA

expression of viral long terminal repeat and 3′ proximate cellular sequences. Proc. Nat’l.

Acad. Sci. USA, 79: 2465-2469, 1982.

147. Rosen, C.A., Sodroski, J.G., Kettmann, R., Burny, A., and Hazeltine, W.A. trans

activation of the bovine leukemia virus long terminal repeat in BVL-infected cells.

Science (Wash. DC), 227: 320-322, 1985.

148. Kettmann, R., Deschamps, J., Couez, D., Claustriaux, J.J., Palm, R., and Burny, A.

Chromosome integration domain for bovine leukemia provirus in tumors. J. Virol., 47:

146-150, 1983.

149. Hoshino, H., Esumi, H., Miwa, M., Shimoyama, M., Minato, K., Tobinai, K.,

Hirose, M., Natunabe, S., Inadu, N., Konishita, K., Komihira, S., Ichimara, M., and

Sugimara, T. Establishment and characterization of 10 cell lines derived from patients

with adult T-cell leukemia. Proc. Nat’l. Acad. Sci. USA, 80: 6061-6065, 1983.

150. Franchini, G., Wong-Staal, F., and Gallo, R.C. Human T-cell leukemia virus

(HTLV-I) transcripts in fresh and cultured cells of patients with adult T-cell leukemia.

Proc. Nat’l. Acad. Sci. USA, 81: 6207-6211, 1984.

151. Wong-Staal, F., Franchini, G., Hahn, B., Arya, S., Gelmann, E.P., Manzari, V.,

and Gallo, R.C. Human T-cell leukemia/lymphoma virus, in vitro transformation, and

leukemia: some molecular studies. In: R.C. Gallo, M. Essex, and L. Gross (eds.),

Human T-cell Leukemia-Lymphoma, pp. 133-139. Cold Spring Harbor, NY: Cold

Spring Harbor Press, 1984.

152. Sodroski, J.G., Rosen, C.A., and Haseltine, W.A. trans-Acting transcriptional

activation of the long terminal repeat of human T lymphotropic viruses in infected cells.

Science (Wash. DC), 225: 381-385, 1984.

153. Burny, A. More and better transactivation. Nature (Lond.), 321: 378, 1986.

154. Yoshida, M., Seiki, M., Hattori, S., and Watanabe, F. Genome structure of human

T-cell leukemia virus and its involvement in the development of adult T-cell

leukemia/lymphoma. In: R.C. Gallo, M. Essex, and L. Gross (eds.), Human T-cell

Leukemia-Lymphoma Virus, pp. 141-149. Cold Spring Harbor, NY: Cold Spring

Harbor Press, 1984.

155. Yoshida, M., Seiki, M., Yamaguchi, Y., and Takatsuki, K. Monoclonal integration

of HTLV in all primary tumors of adult T-cell leukemias suggests causative role of

HTLV in the disease. Proc. Nat’l. Acad. Sci. USA, 81: 2534-2537, 1984.

156. Takenaka, T., Nakamine, H., Oshiro, I., Maeda, J., Kobori, M., Hayashi, T.,

Komoda, H., Hinuma, Y., and Hanaoka, M. Serologic and epidemiologic studies on

adult T-cell leukemia (ATL): special reference to a comparison between patients with

and without antibody to ATL-associated antigen. Jpn. J. Clin. Oncol., 13(Suppl. 2):

257-268, 1983.

157. Tobinai, K., Nagai, M., Setoya, T., Shibata, T., Minato, K., and Shimoyama, M.

Anti-ATLA (antibody to adult T-cell leukemia virus-associated antigen), highly

positive in OKT4-positive mature T-cell malignancies. Jpn. J. Clin. Oncol., 13(Suppl.

2): 237-244, 1983.

158. Shimoyama, M., Kagami, Y., Shimotohno, K., Miwa, M., Minato, K., Tobinai, K.,

Suemasu, K., and Sugimura, T. Adult T-cell leukemia/lymphoma not associated with

human T-cell leukemia virus type I. Proc. Nat’l. Acad. Sci. USA, 83: 4524-4528, 1986.

159. Manzari, V., Gradilone, A. Basillari, G., Fani, M., Collalti, E., Pandolfi, F.,

DeRossi, G., Liso, V., Babbo, P., Robert-Guroff, M., and Frati, L. HTLV-I is endemic

in southern Italy: detection of the first infectious cluster in a white population. Int. J.

Cancer, 36: 557-559, 1985.

160. Pandolfi, F., De Rossi, G., Lauria, F., Ranucci, A., Barillari, G., Manzari, V.,

Semenzato, G., Liso, V., Pizzolo, G., and Aiuti, F. T-helper phenotype chronic

lymphocytic leukeamia and “adult T-cell leukemia” in Italy. Lancet, 2: 633-636, 1985.

161. Nagy, K., Gyuris, A., Drexler, E., Foldes, I., Fust, G., and Horvath, A. Human Tlymphotropic

virus infections in Hungary. In: Proceedings of the 14th International

Cancer Congress Budapest, August 21-27, Abstract 1861, 1986.

162. Bishop, J.M. Enemies within the genesis of retrovirus oncogenes. Cell, 23: 5-6,

1981.

163. Steffen, D., and Weinberg, R.A. The integrated genome of murine leukemia virus.

Cell, 15: 1003-1010, 1978.

164. Canaani, E., and Aaronson, S.A. Restriction enzyme analysis of mouse cellular

type C viral DNA: emergence of new viral sequences in spontaneous AKR lymphomas.

Proc. Nat’l. Acad. Sci. USA, 76: 1677-1687, 1979.

165. Seiki, M., Eddy, R., Shows, T.B., and Yoshida, M. Nonspecific integration of the

HTLV provirus genome into adult T-cell leukaemia cells. Nature (Lond.), 309: 640-

642, 1984.

166. Hann, S.R., Abrams, H., Rohrschneider, L., and Eisenman, R. Proteins encoded by

v-myc and c-myc oncogenes: identification and localization in acute leukemia virus

transformants and bursal lymphoma cell lines. Cell, 34: 789-798, 1983.

167. Shih, C.-K., Linial, M., Goodenow, M.M., and Hayward, W.S. Nucleotide

sequence 5′ of the chicken c-myc coding region: localization of a non-coding exon that

is absent from myc transcripts in most avian leukosis virus-induced lymphomas. Proc.

Nat’l. Acad. Sci. USA, 81: 4697-4701, 1984.

168. Pachel, C., Schubach, W., Eisenman, R., and Linial, M. Expression of c-myc RNA

in bursal lymphoma cell lines: identification of c-myc encoded proteins by hybridselected

translation. Cell, 33: 335-344, 1983.

169. Schubach, W., and Groudine, M. Alteration of c-myc chromatin structure by avian

leukosis virus integration. Nature (Lond.), 307: 702-708, 1984.

170. Li, Y., Holland, C., Hartley, J., and Hopkins, N. Viral integration near c-myc in

10-20% of MCF247-induced AKR lymphomas. Proc. Nat’l. Acad. Sci. USA, 81: 6808-

6811, 1984.

171. Steffen, D. Proviruses are adjacent to c-myc in some murine leukemia virusinduced

lymphomas. Proc. Nat’l. Acad. Sci. USA, 81: 2097-2101, 1984.

172. O’Donnel, P.V., Fleissner, E., Lonial, H., Koehne, C.F., and Reicin, A. Early

clonality and high-frequency proviral integration into the c-myc locus in AKR

leukemias. J. Virol., 44: 500-503, 1985.

173. Nusse, R., and Varmus, H.E. Many tumors induced by the mouse mammary tumor

virus contain a provirus integrated in the same region of the host genome. Cell, 31: 91-

109, 1982.

174. Peters, G., Kozak, C., and Dickson, C. Mouse mammary tumor virus integration

regions int-1 and int-2 map on different mouse chromosomes. Mol. Cell Biol., 4: 375-

378, 1984.

175. Cooper, G.M., and Neiman, P.E. Two distinct candidate transforming genes of

lymphoid leukosis virus-induced neoplasms. Nature (Lond.), 292: 857-858, 1981.

176. Sprent, J. Migration and lifespan of lymphocytes. In: F. Loor and G.E. Roelants

(eds.), Band I cells in immune recognition, pp. 59-82. New York: John Wiley Sons,

1977.

177. Hayward, W.S. Multiple mechanisms of oncogene activation in viral and nonviral

neoplasia. In: I.L. Weissman (ed.), Leukemia, Dahlem Konferenzen Berlin, pp. 147-

162. Berlin: Springer-Verlag, 1985.

178. Dofuku, R., Biedler, J.L., Spengler, B.A., and Old, L.J. Trisomy of chromosome

15 in spontaneous leukemia of AKR mice. Proc. Nat’l. Acad. Sci. USA, 72: 1515-1517,

1975.

179. Wiener, F., Ohno, S., Spira, J., Haran-Ghera, N., and Klein, G. Chromosome

changes (trisomies, no. 15 and 17) associated with tumor progression in leukemias

induced by radiation leukemia virus. J. Nat’l. Cancer Inst., 61: 227-237, 1978.

180. Spira, J., Wiener, F., Ohno, S., and Klein, G. Is MLV or DMBA trisomy cause or

consequence of murine T-cell leukemia development? Studies on Robertsonian

translocation in mice. Proc. Nat’l. Acad. Sci. USA, 76: 6619-6621, 1979.

181. Dofuku, R., Utakoij, T., and Matsuzawa, A. Trisomy of chromosome 13 in

spontaneous mammary tumors of GR, C3H and non-inbred Swiss mice. J. Nat’l. Cancer

Inst., 63: 651-656, 1979.

182. Hare, W.C.C., Yang, T.-J., and McFreely, R.A. A survey of chromosome findings

in 47 cases of bovine lymphosarcoma (leukemia). J. Nat’l. Cancer Inst., 38: 383-392,

1967.

183. Sadamori, N., Kusano, M., Nishimo, K., Fagawa, M., Yao, E., Yamada, Y.,

Amagasaki, T., Kinoshita, K., and Ichimara, M. Abnormalities of chromosome 14 at

band 14g11 in Japanese patients with adult T-cell leukemia. Cancer Genet. Cytogenet.,

17: 279-282, 1985.

184. Rowley, J.D., Haven, J.M., Wong-Staal, F., Franchini, G., Gallo, R., and Blattner,

W. Chromosome pattern in cells from patients positive for human T-cell

leukemia/lymphoma virus. In: R.C. Gallo, M. Essex, and L. Gross (eds.), Human T-cell

Leukemia/Lymphoma Virus, pp. 85-89. Cold Spring Harbor, NY: Cold Spring Harbor

Press, 1984.

185. Sanada, I., Tanaka, R., Kumagai, E., Tsuda, H., Nishimura, H., Yamaguchi, K.,

Kawano, F., Fujiwara, H., and Takatsuki, K. Chromosomal aberrations in adult T cell

leukemia: relationship to the clinical severity. Blood 65: 649-654, 1985.

186. Pandolfi, F. T-CLL and allied diseases: new insights into classification and

pathogenesis. Dign. Immunol., 4: 61-74, 1986.

187. Pandolfi, F., De Rossi, G., and Semenzato, G. T-helper phenotype leukemias: role

of HTLV-I. Lancet, 2: 1367-1368, 1985.

188. Fialkow, P.J., and Singer, J.W. Tracing development and cell lineages in human

hemopoietic neoplasia. In: I.L. Weissman (ed.), Leukemia, Dahlem Konferenzen

Berlin, pp. 203-222. Berlin: Springer-Verlag, 1985.

189. Ueshima, Y., Rowley, J.D., Variakojis, D., Winter, J., and Gordon, L. C

198. Haran-Ghera, N. Pathogenesis of murine leukemia. In: G. Klein (ed.), Viral

Oncology, pp. 161-185. New York: Raven Press, 1980.

199. Chang, T.D., Biedler, J.L., Stockert, E., and Old, L.J. Trisomy of chromosome 15

in X-ray-induced mouse leukemia. Proc. Soc. Am. Assoc. Cancer Res., 18: 225, 1977.

200. Klein, G. Lymphoma development in mice and humans: diversity of initiation is

followed by convergent cytogenetic evolution. Proc. Nat’l. Acad. Sci. USA, 76: 2442-

2446, 1979.

201. Spira, J., Babowits, M., Wiener, F., Ohno, S., Wirschubski, Z., Haran-Ghera, N.,

and Klein, G. Non-random chromosomal changes in Thy-1-positive and Thy-1 negative

lymphomas induced by 7, 12-dimethylbenzanthracene in S26 mice. Cancer Res., 40:

2609-2616, 1980.

202. Erikson, J., Williams, D.L., Finan, J., Howell, P.C., and Croce, C. Locus of the (a)-

chain of the T-cell receptor is split by chromosome translocation in T-cell leukemias.

Science (Wash. DC), 229: 784-786, 1985.

203. Hecht, F., Morgan, R., Kaiser-McCaw-Hecht, B., and Smith, S. Common region

on chromosome 14 in T-cell leukemia and lymphoma. Science (Wash. DC), 226: 1445-

1447, 1984.

204. Chen, I.S.Y., Slamon, D.J., Rosenblatt, J.D., Shah, N.P., Quan, S.G., and

Wachsman, W. The x gene is essential for HTLV replication. Science (Wash. DC), 229:

54-58, 1985.

205. Haas, M., Mally, M.I., Bogenberger, J.M., Bogart, M.H., Buchannan, M.A.,

Augery-Bourget, Y., Hyman, R., and Vogt, M. Autocrine growth and progression of

murine X-ray-induced T cell lymphomas. EMBO J., 5: 1775-1782, 1986.

206. Oliff, A., McKinney, M.D., and Agranovsky, O. Contribution of the gag and pol

sequences to the leukemogenicity of Friend murine leukemia virus. J. Virol., 54: 864-

868, 1985.

207. Gazzolo, L., Moscovici, C., and Moscovici, M.G. Persistence of avian oncoviruses

in chicken macrophages. Infect. Immun., 23: 294-297, 1979.

208. Schwarz, H., Ihle, J.N., Wecker, E., Fischinger, P.O., Thiel, H.-J., Bolognesi, D.P.,

and Schafer, W. Properties of mouse leukemia viruses. XVI. Factors required for

successful treatment of spontaneous AKR leukemia by antibodies against gp71.

Virology, 111: 568-578, 1981.

209. Marx, J.L. AIDS virus has new name-perhaps. Science (Wash. DC), 232: 699-700,

1986.

210. Levy, J.A., Hoffman, A.D., Kramer, S.M., Landis, J.A., Shimabukuro, J.M., and

Oshiro, L.S. Isolation of lympho-cytopathic retroviruses from San Francisco patients

with AIDS. Science (Wash. DC), 225: 840-842, 1984.

211. Gallo, R.C., Salahuddin, S.Z., Popovic, M., Shearer, G.M., Kaplan, M., Haynes,

B.F., Palher, F.J., Redfield, R., Oleske, J., Safai, B., White, F., Foster, P., and

Markham, P.D. Frequent detection and isolation of cytopathic retroviruses (HTLV III)

from patients with AIDS and at risk for AIDS. Science (Wash. DC), 224: 500-503,

1984.

212. Salahuddin, S.A., Markham, P.D., Popovic, M., Sarngadharan, M.G., Orndorff, S.,

Fladagar, A., Patel, A., Gold, J., and Gallo, R.C. Isolation of infectious T-cell

leukemia/lymphoma virus type III (HTLV-III) from patients with acquired immunodeficiency

syndrome (AIDS) or AIDS-related complex (ARC) and from healthy

carriers: a study of risk groups and tissue sources. Proc. Nat’l. Acad. Sci. USA, 82:

5530-5534, 1985.

213. Levy, J.A., Kaminsky, L.S., Morrow, M.J.W., Steiner, K., Luciw, P., Dina, D.,

Hoxie, J., and Oshiro, L. Infection by the retrovirus associated with the acquired

immunodeficiency syndrome. Ann. Intern. Med., 103: 694-699, 1985.

214. Levy, J.A., and Shimabukuro, J. Recovery of AIDS-associated retroviruses from

patients with AIDS or AIDS-related conditions and from clinically healthy individuals.

J. Infect. Dis., 152: 734-738, 1985.

215. Gottlieb, M.S., Schroff, R., Schamber, H.M., Weisman, J.D., Fan, P.T., Wolf,

R.A., and Saxon, A. Pneumocystis carinii pneumonia and mucosal candidiasis in

previously healthy homosexual men: evidence of a new acquired cellular immunodeficiency.

N. Engl. J. Med., 305: 1425-1431, 1981.

216. Centers for Disease Control. Weekly surveillance report, October 20, 1986.

217. Sarngadharan, M.G., Popovic, M., Brach, L., Schupbach, J., and Gallo, R.C.

Antibodies reactive with human T-lymphotropic retroviruses (HTLV-III) in the serum

of patients with AIDS. Science (Wash. DC), 224: 506-508, 1984.

218. Blattner, W.A., Biggar, R.J., Weiss, S.H., Melbye, M., and Goedert, J.J.

Epidemiology of human T-lymphotropic virus type III and the risk of the acquired

immunodeficiency syndrome. Ann. Intern. Med., 103: 665-670, 1985.

219. Kaminsky, L.S., McHugh, T., Stites, D., Volberding, P., Henle, G., Henle, W., and

Levy, J.A. High prevalence of antibodies to acquired immune deficiency syndrome

(AIDS)-associated retrovirus (ARV) in AIDS and related conditions but not in other

disease states. Proc. Nat’l. Acad. Sci. USA, 82: 5535-5539, 1985.

220. Peterman, T.A., Jaffe, H.W., Feorino, P.M., Haverkos, H.W., Stoweburner, R.L.,

and Curran, J.W. Transfusion-associated acquired immunodeficiency syndrome in the

United States. JAMA, 254: 2913-2917, 1985.

221. Clumeck, N., Robert-Guroff, M., Van De Perre, P., Jennings, A., Sibomana, J.,

Demol, P., Cran, S., and Gallo, R.C. Sero-epidemiological studies of HTLV-III

antibody prevalence among selected groups of heterosexual Africans. JAMA, 254:

2599-2602, 1985.

222. Gottlieb, M.S., Wolfe, P.R., Fahey, J.L., Knight, S., Hardy, D., Eppolito, L.,

Ashida, E., Patel, A., Beall, G.N., and Sun, N. The syndrome of persistent generalized

lymphodenopathy: experience with 101 patients. Adv. Exp. Med. Biol., 187: 85-91,

1985.

223. Curran, J.W., Morgan, M.W., Hardy, A.M., Jaffe, H.W., Darrow, W.W., and

Dowdle, W.R. The epidemiology of AIDS: current status and future prospects. Science

(Wash. DC), 229: 1352-1357, 1985.

224. Sivak, S.L., and Wormser, G.P. How common is HTLV-III infection in the United

States? N. Engl. J. Med., 313: 1352, 1985.

225. McDougal, J.S., Cort, S.P., Kennedy, M.S., Cabridilla, C.D., Feorino, P.M.,

Francis, D.P., Hicks, D., Kalyanaraman, V.S., and Martin, L.S. Immunoassay for the

detection and quantitation of infectious human retrovirus, lymphadenopathy-associated

virus (LAV). J. Immunol. Methods, 76: 171-183, 1985.

226. Robert-Guroff, M., Brown, M., and Gallo, R.C. HTLV-III neutralizing antibodies

in patients with AIDS and ARC. Nature (Lond.), 316: 72-74, 1985.

227. Weiss, R.A., Clapham, P.R., Cheingsong-Popov, R., Dalgleish, A.G., Carne, C.A.,

Weller, I.V.D., and Tedder, R.S. Neutralization of human T-lymphotropic virus type III

by sera of AIDS and AIDS-risk patients. Nature (Lond.), 316: 69-72, 1985.

228. Saxinger, C., Levine, P.H., Lange-Wantzin, G., and Gallo, R. Unique pattern of

HTLV-III (AIDS-related) antigen recognition by sera from African children in Uganda

(1972). Cancer Res. (Suppl.), 45: 4624-4626, 1985.

229. Bayley, A.C., Cheingsong, Popov, R., Dalgleish, A.G., Downing, R.G., Tedder,

R.S., and Weiss, R.A. HTLV-III serology distinguishes atypical and endemic Kaposi’s

sarcoma in Africa. Lancet, 1: 359-361, 1985.

230. Black, P.H. HTLV-III, AIDS, and the brain. N. Engl. J. Med., 313: 1538-1540,

1985.

231. Resnick, L., DiMarzo-Veronese, F., Schupbach, J., Tortellotte, W.W., Ho, D.D.,

Muller, F., Shapshak, P., Vogt, M., Gropman, J.E., Markham, P.D., and Gallo, R.C.

Intra-blood-brain-barrier synthesis of HTLV-III-specific IgG in patients with

neurologic symptoms associated with AIDS or AIDS-related complex. N. Engl. J.

Med., 313: 1498-1510, 1985.

232. Ho, D.D., Rota, T.R., Schooley, R.T., Kaplan, J.C., Allan, J.D., Groopman, J.E.,

Resnick, L., Felsenstein, D., Andres, G.A., and Hirsch, M.S. Isolation of HTLV-III

from cerebrospinal fluid and neural tissue of patients with neurologic syndromes related

to the acquired immunodeficiency syndrome. N. Engl. J. Med., 313: 1493-1497, 1985.

233. Goedert, J.J., Biggar, R.J., Weiss, S.H., Eyster, M.E., Melbye, M., Wilson, S.,

Ginzburg, H.M., Grossman, R.J., DiGiola, R.A., Sanchez, W.C., Giron, J.A., Ebbesen,

P., Gallo, R.C., and Blattner, W.A. Three-year incidence of AIDS in five cohorts of

HTLV-III-infected risk group members. Science (Wash. DC), 231: 992-995, 1986.

234. Biggar, R.J. The AIDS problem in Africa. Lancet, 1: 79-83, 1986.

235. Kreiss, J.K., Koech, D., Plummer, F.A., Holmes, K.K., Lightfoote, M., Piot, P.,

Ronald, A.R., Ndinya-Achola, J.O., D’Costa, L.J., Roberts, P., Ngugi, E.N., and Quinn,

T.C. AIDS virus infection in Nairobi prostitutes-spread of the epidemic to East Africa.

N. Engl. J. Med., 314: 414-417, 1986.

236. Rodriguez, L., Sinangil, F., Godoy, G., Dewhurst, S., Merino, F., and Volsky, D.J.

Antibodies to HTLV-III/LAV among aboriginal Amazonian Indians in Venezuela.

Lancet, 2: 1098-1100, 1985.

237. Levy, J.A., and Ziegler, J.L. Acquired immune deficiency syndrome is an

opportunistic infection and Kaposi’s sarcoma results from secondary stimulation.

Lancet, 2: 78-81, 1983.

238. Hoxie, J.A., Haggarty, B.S., Rachowski, J.L., Pilsbury, N., and Levy, J.A.

Persistent noncytopathic infection of human lymphocytes with AIDS-associated

retrovirus. Science (Wash. DC), 229: 1400-1402, 1985.

239. Anonymous Report. Needle-stick transmission of HTLV-III from a patient

infected in Africa. Lancet, 2: 1376-1377, 1984.

240. Cooper, D.A., Maclean, P., Finlayson, R., Michelmore, H.M., Gold, J., Donovan,

B., Barnes, T.G., Brooke, P., and Penny, R. Acute AIDS retrovirus infection. Definition

of a clinical illness associated with seroconversion. Lancet, 1: 537-540, 1985.

241. Eyster, M.E., Goedert, J.J., Sarngadharan, M.G., Weiss, S.H., Gallo, R.C., and

Blattner, W.A. Development and early natural history of HTLV-III antibodies in

persons with hemophilia. JAMA, 253: 2219-2223, 1985.

242. Feorino, P.M., Jaffe, H.W., Palmer, E., Peterman, T.A., Francis, D.E., Kalzunaran,

V.S., Weinstein, R.A., Stoneburner, R.L., Alexander, W.J., Ruersky, C., Getchell, J.P.,

Weinfeld, D., Haverkos, H.W., Kilbourne, B.W., Nicholson, J.K.A., and Curran, J.W.

Transfusion-asociated acquired immune deficiency syndrome. N. Engl. J. Med., 312:

1293-1296, 1985.

243. Sarin, P.S., and Gallo, R.C. The involvement of human T-lymphotropic viruses in

T-cell leukemia and immune deficiency. Cancer Rev., 1: p. 1-17, Copenhagen: Einar

Munksgaard, 1986.

244. Sande, M.A. Transmission of AIDS. The case against casual contagion. N. Engl. J.

Med., 314: 380-382, 1986.

245. Dayton, A.I., Sodroski, J.G., Rosen, G.A., Goh, W.C., and Haseltine, W.A. The

trans-activator gene of the human T cell lymphotropic virus type III is required for

replication. Cell, 44: 941-947, 1986.

246. Shaw, G.M., Hahn, B.H., Arya, S.K., Groopman, J.E., Gallo, R.C., and Wong-

Staal, F. Molecular characterization of human T-cell leukemia (lymphotropic) virus

type III in the acquired immune deficiency syndrome. Science (Wash. DC), 226: 1165-

1167, 1984.

247. Harper, M.E., Marselle, L.M., Gallo, R.C., and Wong-Staal, F. Detection of

lymphocytes expressing T-lymphotropic virus type III in lymph nodes and peripheral

blood from infected individuals by in situ hybridization. Proc. Nat’l. Acad. Sci. USA,

83: 772-776, 1986.

248. Folks, T., Powell, D.M., Lightfoote, M.M., Benn, S., Martin, M., and Fauci, A.

Induction of HTLV-III/LAV from a nonvirus-producing T-cell line: implications for

latency. Science (Wash. DC), 231: 600-602, 1986.

249. Centers for Disease Control. Update: acquired immuno-deficiency syndrome –

United States. Morbidity Mortality Weekly Rep., 34: 245-248, 1985.

250. Beard, J.W. Avian virus growth and their etiological agents. Adv. Cancer Res., 7:

1-127, 1963.

251. Fischinger, P.J., Ihle, J.N., Levy, J.-P., Bolognesi, D.P., Elder, J., and Schaefer, W.

Recombinant murine leukemia viruses and protective factors in leukemogenesis. Cold

Spring Harbor Conf. Cell Proliferation, 7: 989-1103, 1980.

252. Centers for Disease Control. Update: evaluation of human T-lymphotropic virus

type III/lymphadenopathy-associated virus infection in health care personnel-United

States. Morbidity Mortality Weekly Rep., 34: 575-578, 1985.

253. Friedland, G.H., Satzman, B.R., Rogers, M.F., Kahl, P.A., Lesser, M.L., Mayers,

M.M., and Klein, R.S. Lack of transmission of HTLV-III/LAV infection to household

contacts of patients with AIDS or AIDS-related complex with oral candidiasis. N. Engl.

J. Med., 314: 344-349, 1986.

254. Klatzmann, D., Barre-Sinoussi, F., Nugeyre, M.T., Dauguet, C., Vilmer, E.,

Griscelli, C., Brun-Vezinet, F., Rouzioux, C., Gluckman, J.C., Chermann, J.C., and

Montagnier, L. Selective tropism of lymphadenopathy associated virus (LAV) for

helper-inducer T lymphocytes. Science (Wash. DC), 225: 59-63, 1984.

255. Barnes, D.M. AIDS-related brain damage unexplained. Science (Wash. DC), 232:

1091-1093, 1986.

256. Lifson, G.D., Feinberg, M.B., Reyes, G.R., Rabin, L., Banapour, B., Chakrabarti,

S., Moss, B., Wong-Staal, F., Steiner, K.S., and Engleman, E.G. Induction of CD4-

dependent cell fusion by HTLV-III/LAV envelope glycoprotein. Nature (Lond.), 323:

725-728, 1986.

257. Gartner, S., Markovits, P., Markovitz, D., Kaplan, M., Gallo, R., Popovic, M. The

role of mononuclear phagocytes in HTLV-III/LAV infection. Science (Wash. DC), 233:

215-219, 1986.

258. Stewart, S.J., Fujimoto, J., and Levy, R. Human T lymphocytes and monocytes

bear the same leu-3 (T4) antigen. J. Immunol., 136: 3773-3778, 1986.

259. Montagnier, L., Gruest, J., Chamaret, S., Dauguet, C., Axler, C., Guetard, D.,

Nugeyre, M.T., Barre-Sinoussi, F., Chermann, J.-C., Burnet, J.B., Klatzmann, D., and

Gluckman, J.C. Adaptation of lymphodenopathy associated virus (LAV) to replication

in EBT-transformed B lymphoblastoid cells lines. Science (Wash. DC), 225: 63-66,

1984.

260. Levy, J.A., Cheng-Mayer, C., Dina, D., and Luciw, P. AIDS retrovirus (ARV)

replicates in human and animal fibroblasts transfected with a molecular clone. Science

(Wash. DC), 232: 998-1011, 1986.

261. Shaw, G.M., Harper, M.E., Hahn, B.H., Epstein, L.G., Gajdusek, D.C., Price,

R.W., Naria, B.A., Petito, C.K., O’Hara, C.J., Groopman, J.E., Cho, E.-S., Oleske, J.M.,

Wong-Staal, F., and Gallo, R.C. HTLV III infection in brains of children and adults

with AIDS encephalopathy. Science (Wash. DC), 277: 177-181, 1985.

262. Daniel, M.D., King, N.W., Letvin, N.L., Hunt, R.D., Seghal, P.K., and Desrosiers,

R.C. A new type D retrovirus isolated from macaques with immune-deficiency

syndrome. Science (Wash. DC), 223: 602-605, 1984.

263. Letwin, N.L., Daniel, M.A., Seghal, P.K., Chaliloux, L.V., King, N.W., Hunt,

R.D., Aldrich, W.R., Holley, K., Schmidt, D.K., and Desrosiersm, R.C. Experimental

infection of rhesus monkeys with type D retrovirus. J. Virol., 52: 683-686, 1984.

264. Kanki, P.J., Alroy, J., and Essex, M. Isolation of T-lymphotropic retrovirus related

to HTLV-III/LAV from wild-caught African green monkeys. Science (Wash. DC), 230:

951-954, 1985.

265. Fultz, P.N., McClure, H.M., Swenson, R.B., McGrath, C.R., Brodie, A., Getchell,

J.P., Jensen, F.C., Anderson, D.C., Broderson, J.R., and Francis, D.P. Persistent

infection of chimpanzees with human T-lymphotropic virus type III/lymphodenopathyassociated

virus: a potential model for acquired immunodeficiency syndrome. J. Virol.,

58: 116-124, 1986.

266. Purtilo, D.T., Kipscomb, H., Krueger, G., Sonnabend, J., Casareale, D., and

Volsky, D.J. Role of Epstein-Barr virus in acquired immune deficiency syndrome. Adv.

Exp. Med. Biol., 187: 53-65, 1985.

267. Peutherer, J.F., Edmond, E., Simmonds, P., Dickson, J.D., and Bath, G.E. HTLVIII

antibody in Edinburgh drug addicts. Lancet, 2: 1129-1130, 1985.

268. Koprowski, H., DeFreitas, E., Harper, M.E., Sandberg-Wollheim, M., Sheremata,

W., Robert-Guroff, M., Saxinger, C.W., Feinberg, M.B., Wong-Staal, F., and Gallo,

R.C. Multiple sclerosis and human T-cell lymphoytropic retroviruses. Nature, 318: 154-

160, 1985.

269. Lairmore, M.D., Rosadio, R.H., and De Martini, J.C. Ovine lentivirus lymphoid

interstitial pneumonia. Rapid induction in neonatal lambs. Am. J. Path., 125: 173-181,

1986.

270. Chen, I.Y., Wachsman, W., Rosenblatt, J.D., and Cann, A.J. The role of the X

gene in HTLV associated malignancy. Cancer Surveys 5 (2): 331-342, 1986.

271. Gallo, R.C. The first human retrovirus. Scientific American, 255 (6): 88-98, 1986.

272. Zijlstra, M., and Melief, C.J.M. Virology, genetics and immunology of murine

lymphomagenesis. Biochem. et. Biophys. Acta, 865: 197-231, 1986.

273. Duesberg, P.H. Cancer genes: rare recombinants instead of activated oncogenes.

Proc. Nat’l. Acad. Sci. USA, in press, 1987.

274. Stoler, M.H., Eskin, T.A., Been, S., Angerer, R.C., and Angerer, L.M. Human Tcell

lymphotropic virus type III infection of the central nervous system. J. Am. Med.

Assoc., 256: 2360-2364, 1986.

275. Chayt, K.J., Harper, M.E., Marselle, L.M., Lewin, E.B., Rose, R.M., Oleske, J.M.,

Epstein, L.G., Wong-Staal, F., and Gallo, R.C. Detection of HTLV-III RNA in lungs of

patients with AIDS and pulmonary involvement. J. Am. Med. Assoc., 256: 2356-2371,

1986.

276. Scott, G.B., Fischl, M.A., Klimas, N., Fletcher, M.A., Dickinson, G.M., Levine,

R.S., and Parks, W.P. Mothers of infants with the acquired immuno-deficiency

syndrome. J. Am. Med. Assoc., 253: 363-366, 1985.

277. Gonda, M., Wong-Staal, F., Gallo, R., Clements, J., Narayan, O., and Gilden, R.

Sequence homology and morphologic similarity of HTLV-III and visna virus: a

pathogenic lentivirus. Science, 227: 173-177, 1985.

278. Walker, C.M., Moody, D.J., Stites, D.P., and Levy, J.A. CD8+ lymphocytes can

control HIV infection in vitro by suppressing virus replication. Science, 234: 1563-

1566, 1986.