REPORTS ON SEMINARS

(1) Seminar on "Neural Crest Tumors"
(see "Summary of Activities")
(2) Seminar on "Adult T-cell Leukemia/Lymphoma, Role of Human Type C Retrovirus" (Proceedings of a Workshop held at Fred Hutchinson Cancer Center in Seattle, Washington, March 10 and 11, 1982)

Introduction
A brief overview was given by Dr. Essex about viruses as potential causes for leukemia and lymphoma in man, retroviruses that cause leukemia or lymphoma in animals under natural conditions, and about the general biology of retroviruses. Thus far, the only virus that has been clearly show to be reproducibly associated with any form of human leukemia or lymphoma is the Epstein-Barr herpesvirus (EBV), an agent which has been associated with Burkitt's lymphoma (BL). BL is a monoclonal malignancy of B cells that occurs at elevated rates in children in certain regions of Africa. As with other B cell tumors, the involved cells of BL have chromosomal abnormalities, usually in the form of an 8 to 14 translocation. EBV genome can usually be found in tumor cells taken from African BL patients, but its DNA is primarily in an episomal form, rather than in integrated form.
Several viruses have been shown to cause leukemia or lymphoma in animals. In chickens, mice, cats, cattle, and gibbon apes, retroviruses have been shown to cause one or more forms of lymphoproliferative malignancies under natural conditions. In chickens and cattle the lymphoid tumors are primarily of B lymphocyte origin while in cats and mice they are primarily malignancies of T lymphocytes. In most, and perhaps all, of these cases the tumors are monoclonal in origin, and analysis by Southern blotting hybridization shows integration of the retroproviral genome at a single site in all cells of the same tumor but at different sites in tumors from different individuals caused by the same virus. In all of these species the retroviruses that cause the diseases in question are acquired by horizontal or contagious transmission, rather than by genetic inheritance. In selected cases, such as in chickens, the viruses are congenitally transmitted, and in most species it appears that acquisition of the virus very early in life is most likely to result in the establishment of persistent infection.
Most retroviruses have a distinctive "C-type" appearance by electron microscopic evaluation and they bud from the cytoplasmic membrane of productively infected cells. Their genome contains two complete copies of single stranded RNA, is about 70s, and the virions are 1.14 to 1.16 grams/cc buoyant density in sucrose. The RNA replicates through a DNA proviral stage using virus-encoded reverse transcriptase and integration of proviral DNA into the DNA of the infected cell routinely occurs. The typical retroviral genome is 8 to 10 kilobases (KB) in length, and has a "long terminal repeat" (LTR) sequence at both ends which can act as an efficient promoter. Retroviruses have a structure that is characteristic of transposons, and apparently represent one of the most efficient ways for the natural activation and/or transduction of genes in animal cells. The typical genome has three genes: (a) gag, which is at the 5' end of the RNA and encodes for a polyprotein which undergoes post-translational cleavage to make the four peptides which compose the core of the virion; (b) pol, which encodes for the reverse transcriptase; and (c) env, which encodes for the major virion envelope glycoprotein which protrudes at the surface of the virus, allows adsorption to receptor containing cells, and serves as a target for virus neutralizing antibodies.
Little is known about how retroviruses cause leukemia at the molecular level. Replication defective forms of many retroviruses have been isolated and characterized. Such recombinants have usually acquired or transduced one of series of highly conserved cell "onc" (c-onc) genes, which usually allow the virus to transform cultured fibroblasts and cause polyclonal tumors in vivo after very brief latent periods. One possible mechanism by which retroviruses might cause cancer under natural conditions is by the de novo generation of recombinants by replicating competent "helper" viruses. Another possible mechanism, for which some evidence is available in the avian system, is "promoter insertion". In this case, the promoter of the replicating helper virus intergrates adjacent to a c-onc gene, transcriptionally activates that gene, and the presumed product (demonstrated to be a tyrosine-specific protein kinase in some instances) causes the phenotypic alteration characteristic for the given type of malignancy.

Development and Characterization of Human T-Cell Lines
The recognition and characterization of the new human retrovirus, which apparently causes a malignancy of mature T lymphocytes, was made possible by the identification of human T cell growth factor (TCGF) by Dr. Gallo and his colleagues. TCGF, which was also subsequently isolated in mice (where it is referred to as interleukin-2 by some investigators), will allow the continuous proliferation of mature T cells. This molecule, as purified by Mier and Gallo, is a 13,000 dalton protein which binds to receptors on mature terminal transferase (Tdt)-negative T cells of the E-rosette positive and OKT-4 (helper) positive lineage. TCGF can be used to maintain cultures from tumors of the mature T cell type including some cases diagnosed as mycosis fungoides (MF) or Sezary's syndrome (SS), and many cases described by the Japanese as adult T cell leukemia (ATL). Some of the cultures from such tumors subsequently become constitutive producers of TCGF and in such cases the molecule is both present in the media and on the cytoplasmic membrane. Receptors for TCGF are preferentially expressed after stimulation of the target cells by antigen and/or lectins. The first isolation of the human T-cell leukemia virus (HTLV) was made in Dr. Gallo's laboratory in two lymphoblastoid cell lines established from a patient with a tumor composed of mature T cells of the characteristic type (see below). Using TCGF, many lines have now been established in the Gallo lab which contain HTLV, and at least three lines have been established which are constitutive producers of TCGF.
Dr. Miyoshi described his work on the derivation and characterization of human T cell lines that contain a type C virus which is apparently the same as or in the same family as HTLV.* Several cell lines were established, usually by the procedure of co-cultivating patient's cells with normal cord blood leukocytes. Some, such as MT-1 , are of patient origin, while others, such as MT-2, are apparently derived from the infected cord blood cells. Four such lines were described (MT-1 through MT-4). Most grew efflciently as tumors when transplanted into hamsters. One that was clearly derived from the patient had a 14 q + chromosomal aberration, while another that was clearly of cord blood origin (MT-2) had a normal karyotype. This is compatible with a situation where only the tumor-derived lines have distinct karyotypic abnormalities whereas cultures derived by in vitro immortalization by HTLV would not expected to have analagous chromosomal translocations, at least not until after long-term cultivation in vitro. While neither TCGF nor mitogens were added to these cultures, it appears likely that most of them which were maintained on a long-term basis became constitutive producers of this factor. These cultures regularly contain HTLV-type antigens including ATLA, the cell surface antigen (or antigen complex) initially described by Dr. Hinuma using living-cell membrane immunofluorescence.

Isolation and Characterization of HTLV
The isolation and identification of HTLV was then discussed in further detail by Dr. Gallo. It appears that the virus is T-cell tropic, but that only a subpopulation of T cells can be infected and/or transformed by this agent. Thus, although the types of tumors associated with HTLV are invariably of the mature Tdt-negative T-cell type, many cases of tumors diagnosed as MF or SS are not typical of the tumor type that is related to HTLV. Those cases of MF or SS that are related to HTLV appear to be very similar to ATL, are highly malignant, and have frequent skin involvement. Hypercalcemia may also be present, presumably because of osteoclastic factors that are released by the tumor cells.
Features that are characteristic of HTLV include "budding" from the cell surface, a typical 70s RNA, a reverse transcriptase that prefers Mg⁺⁺ over Mn⁺⁺, and characteristic virus structural proteins of about 10,000, 13,000, 19,000, 24,000, 42,000, and 52,000 daltons. The proteins that have been best characterized are the 19,000 dalton protein designated p19, for which a monoclonal antibody is available; and the major core protein designated p24. It has apparently been quite difficult to characterize the major envelope glycoprotein of the virion; presumably because it is highly unstable. This observation may be associated with observations of low infectivity observed for typical virus activity prepared from cell culture supernatant materials.
By serological analysis of the virion structural proteins and reverse transcriptase activity as well as by nucleic acid hybridization, HTLV is clearly distinct from other known retroviruses of both ecotropic and xenotropic types. This list includes various other primate retroviruses, such as the gibbon ape leukemia viruses (GaLV) and the endogenous retroviruses of baboons (BEV). p24 and p19 are clearly viral proteins in that they co-purify with viral cores and have characteristics similar to other retroviral gag proteins such as the distinctive primary amino acid sequence which begins with proline and ends with isoleucine in the case of the p24. Antisera to p19 and/or p24 react efficiently with HTLV-infected lines, including the MT-1 line of Miyoshi, but such antibodies do not react with other non-infected human cells.
The retrovirus that is apparently the closest known relative to HTLV is the bovine leukemia virus (BLV) which causes enzootic leukosis, a B-cell tumor that occurs in cattle throughout the world. Although BLV and HTLV are clearly only related in a very distant sense, characteristics they have in common include: (a) a gag protein i.e. p24 that are comparable in size to each other but somewhat different in size from most other retroviruses; (b) a reverse transcriptase that prefers Mg⁺⁺ to Mn⁺⁺; and (C) a relative inefficiency for production as extracellular virus, at least in relation to other ecotropic (as compared to xenotropic) retroviruses. Finally, recent studies by Oroszlan and Gallo indicate distant relatedness between the primary amino acid sequences of the major core protein (p24) of HTLV and BLV.
One of the most important features of HTLV, as stressed by Dr. Gallo, is the fact that it is not genetically inherited since normal human cells (including uninfected tissues from patients) do not contain any recognizable sequences related to the virus. This implies that the virus must be acquired in a horizontal or contagious manner. The presence of antibodies to HTLV in both patients and health controls with either known exposure to cases or for endemic regions but not in people from non-endemic regions (see below) is also compatible with this observation.
Dr. Yoshida then described his studies on the isolation and identification of the HTLV-related virus found in the two cell cultures designated MT-1 and MT-2. As indicated above, these viruses all appear to be either closely related or identical. Dr. Yoshida presented comparable data on the biochemical and biophysical characteristics of the Japanese isolate MT-1, as well as preliminary data on the preparation of a molecular clone of MT-2 virus in the Charon 4a vector. An exciting observation made by Dr. Yoshida was that HTLV is apparently monoclonally integrated into the human tumor cells. Those experiments were done by Southern blotting hybridization, and they yield results that are similar to those seen when avian or bovine retroviruses cause long-latent period monoclonal lymphomas or leukemias. Using a DNA probe made from virion RNA, Dr. Yoshida found single copy integration in T-tumor cells from five cases of ATL. Although only one band was seen in each tumor, the bands varied in size from one tumor or patient to the next, indicating that the site of integration was different in each case. In some cases cells from both parents of the case were also examined and found to be negative which is compatible with the observation that the virus is not even genetically transmitted in restricted families. In most instances control (non-T) cells from the same patients did not reveal evidence of HTLV provirus, but in one case the granulocytes also appeared to be infected. The virus was found in multiple sites in the MT-1 and MT-2 cell cultures. The observation that HTLV is monoclonally integrated in only the tumor cells in most of the cases that were examined contributes substantially to the case that this virus may be the causative agent of ATL.
Dr. Reitz, from the NCI, then presented nucleic acid hybridization data on various human preparations. Using HTLV-cDNA to explore HUT-102 cells, the first T-cell tumor line of this type had 2-3 copies of proviral DNA per cell. No comparable sequences were found in B cells or other tissues from either the patient or other healthy people. mRNA is also found to be expressed in the tumor cells. Using competition kinetics, it was found that the HTLV sequences related to HUT-102 were very similar to those present in the MT-1 cell culture derived from a Japanese ATL patient by Dr. Miyoshi. The patient that yielded HUT-102 was a 28-year old black from the southeastern USA. Using nucleic acid hybridization, Dr. Reitz and his colleagues found comparable related viral sequences in tissues from several other non-Japanese cases of the same tumor type including a black male from the Caribbean, and a white male from Boston. Using both serological procedures and nucleic acid hybridization, members of Dr. Gallo's group have now identified tumor cases from various areas of the world including, in addition to Japan, the eastern U.S. and the Caribbean, also the western U.S., the Middle East, and South America. Dr. Reitz also briefly presented preliminary information from Dr. Wong-Staal in the Gallo laboratory about the molecular cloning of the 5' proximal viral sequences, representative of the gag gene region, in PBR-322.

In Vitro Immortalization
Dr. Miyoshi then reviewed his studies on "transformation" or "immortalization" of human cord blood lymphocytes with HTLV. His results were that both adult and cord blood lymphocytes could be immortalized with a very high rate of success when co-cultivated with MT-2 cells that had been lethally irradiated. The infected target cells subsequently became immortalized, but did not show karyotypic changes, at least not soon after establishment of the new cultures. Most of the newly infected cells were positive for the ATLA antigen; the proportion of cells positive for this antigen increased from only 2-3% soon after co-cultivation to nearly 100% after a month of cultivation. Whether this was due to a selective growth advantage of the cells that were initially infected or subsequent infection of uninfected cells is unclear. It was clear however, that high efficiency of infection or immortalization could only be achieved by co-cultivation as opposed to inoculation with "cell-free" virus, and the MT-2 culture appeared to be the best donor cell for these experiments. Some of the newly established cultures revealed "hairy cell" processes, suggesting that alterations in phenotype had also occurred.
Dr. Hinuma then presented related studies on in vitro immortalization of peripheral blood lymphocytes where he also found that a high degree of successful immortalization could be used with MT-2 as the donor. More than 90% of the cells were positive for HTLV antigens 2-4 weeks after co-cultivation with mitomycin-C treated donor cells. The infected recipient cells had a phenotype of positivity for Leu 3a, Leu 4, and La 1, and negativity for Leu 2a. About 50% of the cells were positive for Ia antigens cell fusions could be observed in the cultures compatible with the possibility that transmission of this virus might only occur with high efficiency during cell-to-cell contact. This manner of infection with HTLV might also be analogous to that seen with BLV, since BLV is known to be poorly infectious when in the "cell-free" form, but efficient at inducing a syncitial type of effect in sensitive target cells. It seemed likely that two key elements of the successful infection and immortalization of cord blood cells was the virus of MT-2 origin and the co-cultivation procedure. Dr. Gallo also mentioned that some of the more recent isolates in his laboratory have more potential for infectivity than the original HUT-102 isolate.

Clinical Pathological and Cytogenetic Aspects
The syndrome called ATL was initially recognized as a distinct clinical entity and described as such by Dr. Takatsuki. He briefly described this syndrome at this meeting. It is composed of mature T cells, but distinct from many cases of Sezary's syndrome. ATL occurs in adults, with a mean age of about 50. Cutaneous infiltration is frequent, the rate of disease progression is rapid, and the survival rate is low. The disease clearly occurs most frequently in Kyushu, and many Japanese cases that occur outside this area can be traced back to an origin (i.e. of virus exposure) in Kyushu. Most, but not all, cases have hepatomegaly, splenomegaly, and lymphadenopathy. Anemia is not a characteristic feature of this disease and immunoglobulins are usually within normal limits. Although the rates of ATL are not high (perhaps 100 cases/year in all of Japan), the proportion of adult leukemias that are T-cell in origin is higher in Japan than in most other areas of the world. Cytogenetic abnormalities are seen in many cases of ATL. These were described by Drs. Takatsuki, Miyoshi, and Nowell, and include a trisomy 7 in some instances and a 14 q+ in some instances. The most consistent surface phenotype is OKT1⁺3⁺4⁺10⁺11⁺5^-6^-8^- and Ia⁺/- . One possibility is that the tumor cells may be of the suppressor-inducer type.
Histopathologic and cytologic criteria for the recognition of adult T cell leukemia/ lymphoma (ATL) were discussed by Drs. Hanaoka, Jaffe, and Kadin. The presence of at least 10% polylobated or pleomorphic cells in the peripheral blood is characteristic, but a leukemic component may be absent in up to 40 percent of cases. The histopathology in lymph nodes is varied with a similar spectrum encountered in cases from Southwestern Japan, the Caribbean and the United States. Histologic subtypes of ATL documented to occur in association with positive serology for HTLV (ATLV) include: (a) pleomorphic, (b) large cell, (c) mixed cell type and (d) medium-sized cell type. At this time, however, no cytologic or histologic features pathognomonic of viral association have been demonstrated.
In nearly all instances ATL can be distinguished on clinical and pathologic grounds from the cutaneous T cell lymphomas (CTCL), i.e., mycosis fungoides and Sezary syndrome. Still there are cases unquestionably positive for HTLV, both by serology and viral isolation, in which clinically and pathologically the disease is more compatible with CTCL. Further studies correlating virology, histopathology, immunology, and clinical features will be necessary to further define the spectrum of HTLV-associated T cell malignancies.

Epidemiology
Dr. Blattner discussed seroepidemiological studies with patients and controls from various areas of the world. He described about 10 cases of ATL from the West Indies. In most, if not all cases, antibodies were found in sera from such patients that reacted with HTLV antigens. In most cases antibody titers were determined by radioimmunoassay to the p24 and p19 gag antigens and to cell surface antigens by immunofluorescence. In essentially all cases there was agreement between the different procedures and if antibodies were present at all they were usually present for all the antigens.
Drs. Gallo and Blattner subsequently presented serological data on a large series of ATL and other tumor cases as well as for numerous controls. More than 90% of the cases of typical ATL appeared to have antibodies to HTLV (32 of 35). Individuals with other forms of lyphoreticular neoplasms were sometimes positive for such antibodies, with various forms of T-cell malignancies of a similar type having the next most frequent rate of positivity. Samples from 86 U.S. people with other types of neoplasms were negative. Natural antibodies to p24 were found in about 30% of the healthy relatives of ATL patients and in about 12% of the healthy donors from endemic ATL regions of Japan. About 3 to 5% of the healthy donors from the Caribbean area had such antibodies. Sera from about 180 healthy donors from the U.S. and about 800 healthy donors from Europe were negative for such antibodies. The geometric mean antibody titers to the HTLV p24 were about the same for the antibody positive ATL cases when compared to the antibody positive healthy people, whether exposed relatives of cases or people from the endemic regions.
Dr. Hinuma also presented data on the serologic screening of ATL cases and healthy controls for antibodies to ATLA, the cell surface antigen specific for HTLV as expressed on MT-1 cells. Dr. Hinuma found that all of the sera from more than 40 cases of ATL had antibodies to ATLA and the majority of cases of other T-cell malignancies from the endemic area also had such antibodies. When serum from healthy controls over 40 years of age from various geographical areas of Japan were checked, a close geographical association was found with the Kyushu area for the presence of antibodies. While very few of the healthy people were antibody positive in most areas away from Kyushu (i.e. 0 of 74 in Sendai, 0 of 60 in Yamagata, 0 of 74 in Kyoto, 1 of 56 in Sapporo) about 25% of the healthy people from the Kyushu region had such antibodies. A limited number of samples from healthy adults of several other countries were also screened. With the exception of Taiwan, where 2 of 40 were positive, most other geographical areas in the Orient appeared negative for such antibody activity.
Finally, results presented by Drs. Miyoshi and Hinuma suggested that HTLV antigen positive cells and/or active virus was present in the small number of antibody positive people that were checked by analysis for ATLA antigen and/or their ability to infect cord blood lymphocytes by co-cultivation. Although these results were presented in a preliminary manner, it obviously raises the possibility that most antibody-positive health people may be virus carriers.

Summary and Conclusions
HTLV is clearly highly associated with the development of a highly malignant type of leukemia of mature T lymphocytes which has been classified as ATL in Japan but not as clearly and consistently classified as a distinct clinical entity elsewhere. HTLV, the agent initially discovered by Dr. Gallo and his colleagues, appears likely to be the cause of this disease. On both epidemiologic and molecular grounds the association between HTLV and this particular clinical entity appears at least as strong, and probably stronger, than the association between any other human cancer and a particular viral agent (including the association between EBV and Burkitt's lymphoma or the association between hepatitis B virus and primary liver cancer). Although the disease associated with HTLV is clustered in southern Japan and in the Caribbean, both the disease and the agent have been found in various areas of the world. The virus is clearly spread in a contagious manner, even though it has a low degree of infectivity by cell culture analysis. The exact mechanism of infection and the source of virus is unknown. The possibility of transmission by blood transfusion have been suggested but the frequency of exposure in endemic areas (up to 25% of the adult population) is obviously too high to be explained solely in this manner. The possibility that an animal reservoir might exist has also been suggested. If, however, a significant number of the antibody-positive healthy people are indeed carriers with latent infections, this would be adequate for the maintenance of HTLV in the human population without any need for a vector or an outside animal reservoir.

*Note: For simplification, the term HTLV win be used in this report for the designation of an the different human retrovirus isolates found in tumors of the mature T-cell type. While it is recognized that all of these have not been cross-checked by serologic and/or molecular hybridization, those that have been checked appear to be closely related to the original isolate which was designated HTLV. While it is also recognized that some investigators may prefer a different designation for some of the agents subsequently identified by the Japanese investigators it appears that such wording would introduce unnecessary confusion in the current report which is not intended as a policy statement on issues of nomenclature.



SEMINAR AGENDA AND PARTICIPANTS

(1) U.S.-Japan Seminar on Neural Crest Tumors
East-West Center, Honolulu, Hawaii
l-2 March, 1982

AGENDA

PARTICIPANTS

UNITED STATES
J. Bruce Beckwith, M.D.
Department of Pathology
Children's Orthopedic Hospital

Mark H. Greene, M.D.
Environmental Epidemiology Branch
4C18 Landow Building
National Cancer Institute

Alfred G. Knudson, Jr. M.D., Ph.D.
Institute for Cancer Research

Anna T. Meadow, M.D.
The Children's Hospital of Philadelphia

Robert W. Miller, M.D.
Clinical Epidemiology Branch
5A21 Landow Building
National Cancer Institute

Allan E. Rubenstein, M.D.
Department of Neurology
Mount Sinai School of Medicine

R.N. Schimke, M.D.
Department of Medicine
Kansas University Medical Center

JAPAN
Prince Masahito Hitachinomiya
Guest Researcher
Cancer Institute

Atsushi Kukita, M.D.-
Professor
Department of Dermatology
Faculty of Medicine
University of Tokyo

Jiro Matsumoto, Ph. D.
Professor
Department of Biology
Keio University

(present address)
Department of Biological Sciences
Wayne State University

Harukazu Nakamura, M.D.
Instructor
Department of Anatomy
Hiroshima University
School of Medicine

Nobuaki Sasano, M.D.
Professor
Department of Pathology
Tohoku University
School of Medicine

Shigenori Sawaguchi, M.D.
Professor
Department of Pediatric Surgery
University of Tsukuba
School of Medicine

Haruo Sugano, M.D.
Director
Cancer Institute

Kimitomo Takakura, M.D.
Professor
Department of Neurosurgery
Faculty of Medicine
University of Tokyo

Kiyomi Yamada, M.D.
Section Chief
Research Laboratory
National Medical Center



(2) U.S.-Japan Seminar on T-Cell Leukemia/Lymphoma - Role of Human Type C Retroviruses
Fred Hutchinson Cancer Center, Seattle, Washington
10-11 March 1982