Since the changes that accompany induced differentiation of neoplastic cells are quite extensive (Olsson et al., 1996), a number of findings have suggested a variety of potential mechanisms involved in the induction of differentiation of cells in culture. Alterations in DNA methylation have been found during the induced differentiation of Friend erythroleukemia cells (Vizirianakis and Tsiftsoglou, 1996). Interestingly, the DNA methylation inhibitor, 5-azacytidine, is not a very effective inducer of differentiation of the Friend erythroleukemia system (Christman, 1984). Other studies using these cell lines suggest that alterations in membrane structure (Lyman et al., 1976), protein kinase C activation (Aihara et al., 1991), and alterations in the expression of a cyclic AMP–dependent protein kinase (Tortora et al., 1991) are indicative of significant effects in alterations of signal transduction pathways as components of the induced differentiation pro- cess. A role for heat-shock proteins and their interaction with the signal transduction pathway as components of the differentiation program has been suggested (Whitesell et al., 1994). Sachs (1993) and his associates have studied the effects of growth factors and cytokines that are nor- mally involved in the regulation of erythroid and granulocyte maturation. These workers have suggested that reversion and induction of differentiation may occur in appropriate cell types
even in the presence of significant genetic abnormalities such as chromosomal alterations. That active oxygen radicals may be involved in this process as well is suggested by the report by Barrera et al. (1991) of the induction of differentiation of HL-60 cells by 4-hydroxynonenal, a major product of lipid peroxidation. Although all of these various items of information do not make a clear picture as yet, the establishment of the phenomenon of the induction of differentia- tion of neoplastic cells in vitro and the several model systems in which mechanisms of this effect can be studied have opened the way to the potential use of such technologies in the therapy of human neoplasia (Chapter 20).
GENETIC STUDIES OF THE NEOPLASTIC TRANSFORMATION IN VITRO
Because one of the principal mechanisms involved in the neoplastic transformation is presumed to be that of specific structural alterations in the cellular genome, cell culture potentially offers a system in which to study directly genetic alterations causally related to or associated with the neoplastic transformation. Just as studies on the genetics of cell transformation by oncogenic viruses allowed the demonstration of temperature-sensitive mutants (Chapter 4), it has also been possible to isolate such mutants in cells transformable in culture. However, the examples of such identification have thus far been in cell lines, and it is somewhat difficult to equate such findings with those involving specific mutations in viral onc genes. Both in a line of epithelial cells de- rived from rat liver (Yamaguchi and Weinstein, 1975) and in the C3H/10T1/2 cell line (Boreiko and Heidelberger, 1980), mutants have been isolated that behave like transformed cells at tem- peratures of 33° to 36°C, whereas at 40°C such alterations are lost. Colburn and associates (Ler- man et al., 1986) have identified genes, termed pro 1 and pro 2, that specify the sensitivity to the induction of transformation by TPA in a mouse epidermal cell line called JB6. Homologs of these genes have also been identified in human neoplasia, but only the pro 1 homolog was able to confer sensitivity to TPA-induced transformation and transfected into JB6 cells. Related stud- ies by Bouck and diMayorca (1982) with a mouse cell line have suggested that transformation by a papovavirus is expressed as a dominant trait, whereas transformation by chemicals is indic- ative of a recessive trait. Other studies have partially characterized or postulated genes associ- ated with neoplastic transformation in vitro in a human squamous cell carcinoma cell line (Li et al., 1995) and SHE cell transformation respectively (Preston et al., 1994). Of recent interest is the report by Hahn et al. (1999) of the induced transformation of normal human epithelial and fibroblast cells in primary culture by the transfection with retroviral vectors of three different genes, telomerase (Chapter 15), the SV40 large T antigen, and an oncogenic allele of the H-ras proto-oncogene. These authors suggested that the ectopic expression of these three genes re- sulted in the direct conversion of normal to neoplastic cells in vitro. These latter studies are some of the few that have been carried out on the genetics of the neoplastic transformation in primary cell cultures.
Somatic Cell Hybridization as a Tool to Study Mechanisms of the Neoplastic
Transformation in Vitro
With the discovery of the phenomenon of somatic cell hybridization or cell fusion, the potential for genetic studies on somatic cells, especially neoplastic cells, was realized, at least theoreti- cally. In this process (shown diagrammatically in Figure 14.8), two cells from the same or differ- ent species are fused, usually in the presence of inactivated Sendai virus, certain lipids, polyethylene glycol, or electrofusion (Široký and Cervenka, 1990) , resulting in the formation of a single cell with two nuclei. This is termed a heterokaryon when cells of two different types or
Figure 14.8 Outline of the fusion of somatic cells in the laboratory. The chromosomes of each of the two cells arising from two different species are denoted by + and λ. After fusion of the cells and subsequent mitosis, the heterokaryon is seen to contain the chromosomes of both cells, but subsequent replication re- sults in stable clones with a reduced number of chromosomes.
species are fused. If cells of the same type are fused, the product is termed a homokaryon. The binucleate hetero- or homokaryon contains all of the genetic apparatus from each of the two original cells. After DNA synthesis, mitosis occurs with a mixing of the chromosomes from each of the two nuclei and subsequent formation of a single nucleus containing most or all of the chromosomes from the two donor cells. As the hetero- or homokaryon continues to undergo successive cell divisions, generally chromosomes of one or the other of the donor cell nuclei are lost until a relatively stable karyotype is obtained, usually consisting of virtually all of the chromosomes of one donor cell and one or only a few chromosomes from the other. By suitable chromosome identification one can determine the origin of each of the chromosomes of the heterokaryon.
Ephrussi (1965) and Harris (1972) and their associates were among the first to utilize this technique in an attempt to determine whether or not the inheritance of the malignant state in such heterokaryons acted as a dominant or recessive trait. In some of their studies on the fusion between malignant and nonmalignant cells, malignancy behaved like a recessive character, al- though fusion of a variety of malignant cells failed to demonstrate any complementation of the supposed genetic trait, since all such resulting heterokaryons were malignant (Wiener et al.,
1974). When fusion of normal cells with virally transformed cells occurred, the resulting heter- okaryon was neoplastic if it possessed integrated viral information within its genome (Croce et al., 1975). While a number of other studies demonstrated that hybrids produced within the same or different species sometimes were neoplastic, the careful studies of Stanbridge and associates reproduced the earlier investigations of Ephrussi and Harris studying the fusion of normal and neoplastic cells of the same or different tissues (cf. Stanbridge, 1984). The suppression of tum- origenicity occurred even in hybrids in which the neoplastic cell possessed activated oncogenes (Geiser et al., 1986). In some instances, the suppression of malignancy in the hybrid involves the production of gene products involved in terminal differentiation (Harris and Bramwell, 1987). In an extension of these investigations, Howell and Sager (1978), using techniques of mass fusion of whole cells with enucleated cytoplasm to form “cybrids,” have shown that the suppression of tumor-forming ability may be cytoplasmically transmitted in specific cybrids. This finding was further supported by the studies of Israel and Schaeffer (1988), using normal and transformed cells derived from an original single clone.
Harris and his associates (Jonasson et al., 1977) also showed that if a normal–malignant hybrid lost many of its chromosomes, the neoplastic state would reappear; this indicates that the normal cells may contain specific suppressors of malignancy. Sachs and associates (Hitotsuma- chi et al., 1972) have suggested similar mechanisms, even relating the suppressor function to specific chromosomes within hamster cells (see above). More recently Sasabe and Inana (1991) have demonstrated that the suppression of malignancy in hybrids between a retinoblastoma cell line and a “nontransformed” NIH 3T3 resulted in cells of a nontransformed phenotype. These authors suggested that the normal retinoblastoma gene from the nontransformed cells was re- sponsible for this change.
In general, these findings using homo- and heterokaryons of normal and neoplastic cells are compatible with the presence of tumor suppressor genes in situations where the transformed or malignant phenotype is suppressed in the heterokaryon. On the other hand, cell fusion of nor- mal and neoplastic cells has found an interesting application in the production of monoclonal (uniquely specific) antibodies directed toward specific antigens (see Chapter 19). This tech- nique, pioneered by Kohler and Milstein (1976), involves the fusion of normal antibody-produc- ing cells obtained from the spleen or lymph nodes of animals immunized against certain antigens with malignant myeloma cells (cf. Westerwoudt, 1985). Specific clones of cells result- ing from the fusion can be isolated and their antibody production determined. After cloning and isolation, the resulting homokaryon (termed a hybridoma) can be grown in appropriate animals or in cell culture for the production of specific monoclonal antibodies. This technique, a by- product of attempts to determine whether neoplasia is a dominant or recessive trait by the method of cell fusion, has revolutionized our ability to produce large amounts of monospecific antibodies for research and clinical use.
CELL CULTURE AS A TOOL IN OUR UNDERSTANDING OF CARCINOGENESIS AND CANCER
The transformation of cells in culture to the neoplastic state, both spontaneously and induced by chemical, physical, and biological agents, exhibits many similarities to carcinogenesis in vivo. In this chapter an effort has been made to emphasize the more important examples of close anal- ogies between the carcinogenic process in vitro and in vivo. It is through a more careful study of the mechanisms of such analogies that a better understanding of the process of carcinogenesis may be forthcoming. On the other hand, an understanding of the mechanisms of the differences between the carcinogenic process in vivo and that in vitro may be equally rewarding.