EHPnet: Pesticides and People

S OF INVITED CONTRIBUTIONS GENETIC CONTROL OF CELL SURFACE COMPONENTS. W. F. BODMER, Genetics Laboratory, Department of Biochemistry, University of Oxford. Title Only VIRUS-ASSOCIATED CELL SURFACE PRODUCTS: RELEVANCE TO MALIGNANT TRANSFORMATION. J. WYKE, ICRF, London. Infection by viruses of several different families is associated with neoplasia, and many instances are known in which products associated with these viruses are situated at the surface of both normal and malignant cells. It has proved difficult to link these virus-specific cell-surface molecules with either the cause or the progression of malignancy. For this reason we will consider in basic terms the possible mechanisms of virusinduced transformation, and the role that virus-associated cell-surface components may play in such events, and we will attempt to assess the relevance of known virus-specific changes within this framework. Cells may become transformed by viruses whose life cycle is lytic as well as by nonlytic agents. In the former case the cell survives, either because it is non-permissive or semi-permissive for virus replication or because virus replication is defective. In the latter instance similar constraints may reduce the level of virus expression, but it is also possible for the transformed cell to yield fully infectious virus. Whatever the nature of virus-cell interaction, the transformed cell always carries virus genetic material, and viruses could thus transform by interfering with normal cell controls in several ways. Firstly, integrated viral DNA may disrupt regulated expression of the cell genome, a situation which does not require the expression of viral genes. Secondly, virus-coded products may mimic the effect, or even the detailed action, of cell growthpromoting stimuli. If the virus-specific molecules are located in or near the cell membrane they may mimic exogenous signals acting across the cell surface, but the viral products may also resemble endogenous signals for growth. Such virus-coded or virus-specific growthpromoting stimuli could be (1) structural proteins of the virus or their precursors (in their native form or modified by mutation, recombination or aberrant processing in an unusual host), (2) non-structural proteins involved in virus replication (which may also undergo modification), or (3) non-structural proteins with no apparent function in the virus life cycle. The latter class of potential transforming proteins may be particularly important in the Retroviridae, whose mode of replication and high rate of genetic recombination make it likely that the virus can acquire, by recombination, a new genetic material from either the host or another virus. The new material may comprise a clearly defined cistron, distinct from viral replicative functions, or it may be part of a hybrid gene (derived by recombination within a viral cistron) which codes for a protein with some amino acids in common with a virus protein and others of foreign origin. If retroviruses recombine with one another, the viral replicative functions could also be altered, and some oncogenic viruses seem to arise by recombination between two non-oncogenic parents. Only two virus genes have been shown, by reasonable genetic and biochemical criteria, to be involved in causing cell transformation; the early region (T-antigen complex) of papovaviruses and the src gene of sarcomainducing retroviruses. It is possible that the proteins encoded in these genes may have similar mechanisms for inducing cell transformation but, whereas the papovaviruscoded proteins are involved in virus replication, the src gene product seems unrelated to the virus life cycle and this gene is possibly of host-cell origin. The cellular location of these proteins and their detailed mode of action is not understood, and it is possible that they may show functional polymorphism. One component of the T-antigen complex is membrane associated (though not known to be on the cell surface) and the expression of both the papovavirus and the retrovirus genes in the cell is closely correlated with the presence of virus-specific tumour antigens on the cell surface. However, the relationship between these antigens and the viral transforming genes is not understood. Another group of retroviruses, typified by the erythroblastosis and myelocytoma viruses of chickens and the Abelson leukaemia virus of mice, are replication-defective agents which all encode a protein which comprises a portion 458 ABSTRACTS OF INVITED CONTRIBUTIONSS OF INVITED CONTRIBUTIONS of the retrovirus gag gene and amino acid sequences of an unknown origin. For some of these viruses there is preliminary genetic evidence that these hybrid proteins may be a cause of cell transformation, but their location in the infected cell is as yet unknown. In the case of other leukaemia-inducing retroviruses there is no clear evidence for the involvement of a virus gene in the maintenance of the transformed state, but there is ample documentation of the presence of virusassociated antigens on the leukaemic cell surface. Some of these antigens, such as the mouse Glx thymocyte antigen and Gross cellsurface antigen (GCSA) were known to be associated with leukaemia-virus-induced neoplasms, but were thought possibly to be of non-viral origin. However, it is now known that Glx antigenicity is carried on the murine leukaemic virus (MuLV) envelope glycoprotein, whilst GCSA represents glycosylated forms of the precursors to the internal proteins of MuLV encoded by the viral gag gene. With increasing awareness of endogenous virus-specific proteins, including the env and gag gene products, in normal as well as transformed cells, it is difficult to assign to these antigens a function in the causation of leukaemia It is possible that they are simply evidence of viral gene activity, which may be a prerequisite for leukaemogenesis but which may also be apparent in other stages in the life of the host. However, though we do not know whether the virus antigens on the surface of the leukaemic cells have a causal role in the neoplasm, they, like the surface tumour antigens associated with sarcoma viruses, are invariably present. If this novel antigenicity elicits an immune response from the host, the virus-associated surface proteins maybe an important factor in the progression ofthe disease. CELL-SURFACE PROPERTIES, GROWTH AND MALIGNANCY OF HAEMOPOIETIC CELL LINES. K. NILSSON, The Wallenberg Laboratory, University of Uppsala, Sweden. During the last few years, improved tissueculture methodology has allowed the establishment of a number of cell lines in vitro from both normal and neoplastic human haemopoietic tissue and peripheral blood. Analyses of the phenotypic characteristics of haemopoietic lines have revealed that most of them are polyclonal and normal diploid in early passages, carry Epstein-Barr virus (EBV) and are therefore derived from presumably nonneoplastic precursor cells. Such cell lines, tentatively termed lymphoblastoid, can be established wi,ith ease from any anti-EBV7 seropositive individual. In contrast, truly neoplastic haemopoietic cell lines (leukaemia, lymphoma, myeloma) are still rare, and have been established only with great difficulty. The neoplastic lines are always monoclonal, aneuploid and, with exception for Burkitt's lymphoma lines, EBV-genome negative. The lymphoblastoid lines (LCL) have a lymphoblastoid morphology, produce immunoglobulin (Ig) and express lymphocyte surface markers indicating a B-cell descent. The LCLs gradually evolve towards monoclonality and aneuploidy during continuous cultivation. This in vitro selection of an aneuploid clone usually occurs within one year of serial passage. From the studies with the group of malignant haemopoietic lines a few general conclusions seem justified. Firstly, all the lines have individually distinct properties, suggesting that each human haemopoietic tumour, although sharing some basic morphological and functional features with other tumours belonging to the same histopathological entity, is unique. Secondly, haemopoietic tumours continue to express most of the features of the corresponding normal cells. Thirdly, the expression of the phenotypic properties is usually stable during at least 1-2 years of continuous cultivation in vitro. The leukaemic-cell lines are most frequently lymphoblastoid in morphology and usually have a T-cell-like phenotype. A few cell lines have surface characteristics indicating a non-T, non-B lymphocyte origin; even fewer lines have a suggestive B-cell origin. Recently, the first leukaemic cell line with myeloid characteristics has been established from a myeloid leukaemia. To date, 3 myeloma lines and 3 lymphocytic lymphoma lines have been reported. The myeloma lines express plasmacell surface characteristics while the lymphocytic lymphoma lines have a B-cell phenotype. The largest number of lymphoma lines has been established from diffuse histiocytic lymphoma. The phenotypic variability among these lines is suggestive of a marked heterogeneity with respect to cell type of origin in tumours classified histologically as "histiocytic". The majority produce Ig, and thus 459 I.C.R.F. AND B.A.C.R. JOINT SYMPOSIUM seem to be derived from the B-lymphocyte series. Some lines have a lymphocyte morphology but express no lymphocyte surface inarkers, and have therefore tentatively been classified as being derived from non-T, non-B lymphocytes. Three lines have morphological, lymphocyte surface marker and enzyme characteristics, indicating descent from a monocyte type of cell, and would thus represent truly histiocytic lymphomas. The aimn of this overview is to describe some of our recent studies on cell-surface characteristics of malignant haemopoietic cell lines, with special reference to phenotypic properties assumed to be related to the neoplastic state of the cells. The main results wNill be briefly summarised in the following: (1) Exposed surface glycoproteins (GP) have been studied in a panel of neoplastic (lymphoma, leukaemia, myeloma) and nonneoplastic (LCL) lines by the galactose oxidase-tritiated sodium borohydride labelling technique. The labelled GPs were separated by polyacrylamide slab-gel electrophoresis and visualized by autoradiography. Each type of malignant haemopoietic cell line was found to have a basic, easily distinguishable surface GP pattern. In addition, each line had one or a few unique GPs. The B and T lines had most GPs in common with normal B and T cells, respectively. The GPs most constantly associated with malignancy were found in B lymphomas as double bands with apparent mol. wts of 85-87 and 69-71. The nature of these GPs has not as yet been clarified. In leukaemia and myeloma lines, only unqiue GPs were found in addition to those shared with normnal cells. In the group of histiocytic lymphoma lines 4 types of GP patterns could be distinguished: (a) a basically B-lymphocyte-like pattern in 2 lines expressing B-cell surface markers; (b) a basically T-lymphocyte-like pattern in 1 line with non-T, non-B surface phenotype; (c) a basically monocyte-like pattern in 1 line with a monocyte-like phenotype: (d) a 'stem" cell-like pattern in 2 other lines with a histiocyte-like phenotype. The surface-labelling technique thus seems to be useful in discriminating the different types of "histiocytic" lymphomas. (2) The fucose-labelled GPs of the haemopoietic lines were also characterized by gel filtration after degradation and release from the cell surface by proteolytic enzymes. The results demonstrate that the tumour lines, almost wAithout exception, contain an increased amount of fast-eluting glycopeptides than a larger normal cell. Such fast-eluting glycopeptides have previously been demonstrated in several animals and human tumours, but not in normal cells. Interestingly, the surface glycopeptides of the LCL type of lines also had this "malignant" elution