Since the principal characteristic of the stage of progression in morphological terms is the ap- pearance of benign and malignant neoplasms, an analogous stage in cell culture would require some evidence that such cells are truly neoplastic. Thus, to be in the stage of progression cells in the process of cell transformation in vitro would have as an absolute characteristic the formation of neoplasms in vivo.
However, as noted in Chapter 9, the principal cellular and molecular characteristic of the stage of progression is that of evolving karyotypic instability and accompanying aneuploidy. Thus, one may reasonably argue that cells in culture undergoing transformation and exhibiting karyotypic instability are in the stage of progression. At our present state of knowledge this is probably an oversimplification. There are a number of aneuploid cell lines with relatively stable karyotypes that do not appear to develop into neoplasms when inoculated into an appropriate host. This also is not a true test unless one always reinoculates the transformed culture presumed to be in the stage of progression into the host from which it arose. Since this is not always possi- ble, the lack of response or takes on reinoculation of “transformed” cells in vivo may be the result of immune responses in the host, the lack of appropriate circumstances in the host to allow growth of the inoculated cells, or the production by the cultured cells of a product detrimental to the host or to which the host responds excessively. The immunobiology of the host–tumor rela- tionship is discussed later in the text (Chapter 19); there are very few neoplastic cell populations producing materials toxic to the species from which they arose. On the other hand, it has been pointed out earlier that the use of a variety of supports such as sponges, plastic films, etc., does allow the growth of cells in vivo as neoplastic cells when otherwise they would not grow (Boone et al., 1979; Sanford et al., 1980).
In parallel with the evolution of the stage of progression in vivo, one would expect to find some degree of karyotypic evolution in cell transformation as well. Table 14.10 gives a collec- tion of several studies indicating that, both in chemically induced and spontaneous transforma- tion, a time-dependent pattern of karyotypic changes evolves in the cells. Thus, one could argue by analogy that such cells are in the stage of progression, exhibiting its primary characteristic, karyotypic instability. Further evidence for this characteristic of the stage of progression in cell transformation in culture can be seen from the studies by Tlsty and her associates (1989) demon- strating that the rate of spontaneous gene amplification in transformed cells is significantly higher than that seen in nontransformed cells. Even in cell lines, it is possible to demonstrate an evolution of genomic instability when one compares nontransformed 3T3 mouse cells to those transformed by chemicals, radiation, or spontaneously (Honma et al., 1994). Mamaeva (1998) has emphasized the significance of karyotypic evolution in cells in culture, arguing that there are certain generalities that one may apply to the karyotypes of transformed and permanent cell lines. These include the nonrandom character of numerical and structural chromosome changes in cell lines of different histogenesis, the loss of one of the sex chromosomes during prolonged
cultivation, a similarity of the total chromosome material in all cells of the line despite their karyotypic heterogeneity, and retention of, as a minimum, disomy in all autosomes in most cell lines. Several of these characteristics echo the earlier findings of Kraemer et al. (1972) in com- paring karyotypes of an established neoplastic cell line with those of normal cells (Chapter 9).
It is also possible to identify “progressor” agents in cell transformation. Hojo et al. have reported that cyclosporine induces many of the characteristics of the stage of tumor progression in a nontransformed cell line (Hojo et al., 1999). The SV40 T antigen of itself is capable of driving karyotypic instability, which precedes neoplastic transformation in human diploid fibro- blasts (Ray et al., 1990). Furthermore, Li et al. (1997) demonstrated that transformation of pri- mary cultures of Chinese hamster embryo cells by chemicals or spontaneous transformation resulted in essentially 100% of transformed cells exhibiting aneuploidy. These authors have ar- gued that in this transformation process aneuploidy is the cause rather than a consequence of the transformation event itself. Thus, it would appear that, while the initial transformation event as originally described by Berwald and Sachs (1965) and further delineated by Barrett and his col- leagues may not involve in all instances karyotypic alterations, events following transformation almost always lead to a stage in in vitro carcinogenesis that is completely analogous to the stage of progression in vivo. The alternative is the reversion of the transformed cell to a normal pheno- type or its loss by apoptosis, which is considered below.
Loss or Modification of Malignant Potential as an Alternative Stage in the
Natural History of Neoplastic Transformation in Vitro
An interesting finding resulting from studies of the neoplastic transformation in vitro is the phe- nomenon of the reversion from the transformed to the normal state after the initial exposure of cells to a carcinogenic agent. Rabinowitz and Sachs (1970) found that the incidence of the rever- sion from the morphologically transformed state to that of a morphologically normal cellular appearance and life span in chemical or radiation transformation was exceedingly high, usually in the range of 81% to 93%, although such reversion occurred in only 5% of polyoma-trans- formed SHE cells. Many of these reversion variants reacquire characteristics of normal cells, such as a limited life span in vitro (Rabinowitz and Sachs, 1970). Such a high rate of reversion did not support the concept that the transformation process was the result of a single gene muta- tion, but it did not rule out the role of chromosomal changes in the genesis of cell transformation in culture. Such revertants occur even in virus-transformed cells, and the reverted cells still pos- sess viral information within their DNA. Sachs and associates (Hitotsumachi et al., 1972) dem- onstrated that most if not all reverted cells exhibit significant karyotypic abnormalities, usually characterized by some degree of polyploidy. They proposed a model that implies that the bal- ance of gene dosage for the expression and suppression of the neoplastic transformation is criti- cal in the formation of revertants. These concepts may be considered a forerunner of our knowledge of tumor suppressor genes, and later studies by Koi and Barrett (1986) supported this concept in suggesting that loss of tumor-suppressive function was then involved in the develop- ment of the stage of progression in SHE cell transformation. In addition, a number of instances of “transient” reversion of the transformed phenotype have been seen with the application of a variety of agents to specific cell culture systems. Examples of this include cycloheximide in hu- man cell transformation (Cho and Rhim, 1979), chemically defined media in C3H/10T1/2 cells (Tomei and Bertram, 1978), interferon treatment of C3H10T1/2 cells (Brouty-Boyé and Gresser,1981), and variation in cell density of different cell lines (Brouty-Boyé et al., 1980; Bempong and Myers, 1985). Reversion of the transformed phenotype in transformed NIH3T3 cell line was accomplished by the tumor-promoting agent okadaic acid (Sakai et al., 1989). Within 1 week of removal of the okadaic acid, the morphology of the cells reverted to the malignant phenotype. Suggesting that signal transduction is involved in some of these transient alterations, several re- ports have demonstrated the morphological reversion of transformed cells in vitro by the addi- tion of cyclic AMP or its congenors (Johnson et al., 1971; Krystosek and Puck, 1990).
In addition to the “spontaneous” or transient reversion of cells transformed in culture, in- duction of such reversions or, in most cases, terminal differentiation has been described for a number of neoplasms cultured in vitro. Table 14.11 lists several such examples, together with the various chemicals and culture conditions inducing such reversion or terminal differentiation. We have already noted the effect of TPA in inducing and inhibiting differentiation of various cells in culture. The examples given in Table 14.11 extend this to other agents, several of which are not promoting agents, demonstrating a greater variety of chemicals inducing this effect in cell culture.
Even more extensively studied examples of the induction of terminal differentiation in neoplastic cells in culture are those of various leukemic cells, both primary and cell lines ob- tained from both human and animal sources. In Table 14.12 may be seen the rather extensive list of chemical agents capable of inducing differentiation in a human myeloid cell line (HL-60) and in the Friend erythroleukemia cell line from the mouse. The HL-60 cell line has been
many laboratories (cf. Hozumi, 1983; Collins, 1987), while the Friend erythroleukemia cell line from the mouse has also been extensively investigated (cf. Rifkind et al., 1978). The HL-60 cell line has the interesting variation that cells may differentiate either to macrophages or granulo- cytes, some agents inducing transformation of the former (e.g., TPA), others the latter or both. HL-60 cells induced to differentiate to granulocytes apparently died subsequently via apoptosis (Martin et al., 1990).
It is significant that a number of chemotherapeutic drugs are among those inducing differ- entiation of leukemic and other neoplastic cell types. These include topoisomerase inhibitors (Nakaya et al., 1991), actinomycin D, methotrexate, and several base analogs. The potential importance of such findings with respect to the chemotherapy of cancer are discussed in Chapter 20.