One of the critical points in the study of transformation in vitro is that it is obviously impossible to utilize the usual in vivo criteria of carcinogenesis to identify neoplastic cells. The morpho- logic changes seen initially by Berwald and Sachs (1963) were claimed to be the result of onco- genesis; however, such claims could be substantiated only if the cells could grow as neoplasms in a suitable host. Thus, there has been much discussion of the criteria for the neoplastic trans- formation in vitro. No universal criteria have been generally accepted, although many scientists in the field may have agreed on a number of characteristics (Katsuta and Takaoka, 1973). The more important of these criteria are listed in Table 14.1.
Few if any neoplastic cell populations in vitro exhibit all of the criteria indicated in Table
14.1. Furthermore, after initiation in vitro, transformed cells may acquire one or more of the characteristics shown as a function of time in culture after initiation. Barrett and Ts’o (1978) demonstrated in hamster embryo fibroblasts that the temporal acquisition during transformation in vitro of many of the characteristics listed in Table 14.1 was not simultaneous. Rather, morpho- logical transformation (criterion 4) and the production of plasminogen activator (Chapter 16) occurred soon after the treatment with the carcinogen, whereas the growth of transformed cells in soft agar (criterion 3) usually occurred more than 30 population doublings after exposure to
Table 14.1 Criteria for Neoplastic Transformation in Vitro
1. Production of biologically malignant neoplasms in vivo by inoculation of
106 or fewer cells into syngeneic or immunodeficient hosts, in the absence of neoplasms produced by inoculation of comparable numbers of cells not
treated with the “transforming” agent. Some “transformed” cell lines do not conform to this criterion, and some embryonic cells injected into im- munosuppressed or syngeneic hosts will grow to the size of a gross tumor. The time of growth of transformed cells to detectable size in the syngeneic host may vary tremendously.
2. “Immortality” of transformed cells in culture. This is characteristic of many biologically neoplastic cells, although in some instances those hav- ing immortality in vitro do not give rise to tumors in vivo.
3. Growth of transformed cells in soft agar. With the exception of some mouse cell strains, e.g., Heidelberger’s strain C3H/10T1/2 and trans- formed mouse prostate cells, transformed cells exhibiting this characteris- tic also produce neoplasms on inoculation into a suitable host. On the other hand, a number of biologically neoplastic tissues grown in vivo will not grow in soft agar in culture.
4. Colonies of transformed cells exhibit morphologic and growth characteris- tics in culture different from those of normal cells grown in culture. Non- transformed cells grow in an “ordered” way, whereas transformed cells tend to “pile up,” with criss-cross patterns and a higher degree of pleomor- phism. However, this criterion relates only to fibroblastic cells grown in culture. So few epithelial cells have been transformed in culture that mor- phological criteria of transformation for epithelial cells have not been de- termined accurately.
5. Loss of “contact inhibition” of cell replication under specific conditions of media and plating and increase in saturation density by transformed cells. A significant number of nontransformed cells in culture demonstrate no contact inhibition.
6. Reduced requirement for serum and growth factors compared with un- transformed cells. This characteristic is less pronounced in transformed human cells since normal human cells can grow in lower serum levels.
7. Transformed cells in many, but not all, instances may be agglutinated by plant lectins. Not all agglutinated cells, however, demonstrate biological neoplasia in vivo.
8. Cells transformed by chemicals or viruses in culture exhibit antigenic al- terations. “Spontaneous transformants” show no antigenic alterations.
9. Transformed cells may show karyotypic changes. However, cell lines that produce no tumors in vivo may be quite aneuploid, as are many “rever- tants” in culture.
10. Transformed cells usually have a greater efficiency of cloning than non- transformed cells.
11. Transformed cells usually exhibit a loss of bordered cytoplasmic actin and myosin, as exhibited by a loss of parallel arrays of microfilaments.
1. Same criteria as for mesenchymal cells.
2. Same criteria as for mesenchymal cells.
3. Many but not all epithelial cells transformed in vitro exhibit the capacity to grow in soft agar.
4. Often epithelial cells transformed in vitro do not exhibit morphological differences that are dramatically different from normal, but in many in- stances a greater degree of pleomorphism is seen in transformed epithelial cells compared with their nontransformed counterparts.
5. This characteristic is inconsistently seen in transformed epithelial cells, de- pending on the cell line and the tissue of origin; in some cell lines the pil- ing up of cells occurs, whereas in others the monolayer growth pattern is maintained.
6. Decreased requirement compared with normal epithelial cells in liver; has not been sufficiently studied in a variety of systems.
7. Same criteria as for mesenchymal cells.
8. Many, but not all, epithelial cells transformed in vitro exhibit alterations in their antigenicity.
9. Same criteria as for mesenchymal cells.
10. Transformed epithelial cells usually have a much greater capability of growth in vitro from single or a few cells.
11. Most epithelial cells transformed in vitro exhibit alterations in their cyto- skeletal structure, but this finding is inconsistent and not characteristic of all epithelial cells transformed in vitro.
the carcinogen. In a sense, this indicates that an analogy can be drawn between the progression of carcinogenesis in vivo and the full expression of cell transformation in vitro (Table 14.2).
Most investigators in oncology would probably agree that the criterion in Table 14.1 that is paramount for proving that a cell has been transformed to the neoplastic state in vitro is the formation of a neoplasm in vivo on inoculation of the putative transformed cultured cells (crite- rion 1). Other than this criterion, the characteristic most frequently correlated with the neoplastic potential of cells in vivo that had been transformed in vitro is their ability to grow in soft agar (Freedman and Shin, 1974; Weinstein et al., 1976) or exhibit “anchorage independence.” As noted above, a major exception to this are human cells that may be “transformed” to anchorage independence but yet do not exhibit neoplastic potential in vivo (Peehl and Stanbridge, 1981). Furthermore, Kaplan and Ozanne (1983) argued that anchorage-independent growth is a function of the total concentration of growth factors in the medium to which these cells can respond. Thus, this criterion becomes a function of both the concentration of growth factors in the medium and the presence of their receptors on the surface of the cell, transformed or nontransformed.
Immortalization of cells in culture, i.e. the ability to continue replication beyond a finite number of doublings, a characteristic of normal cells (Hayflick, 1997), is seen with many, but not all, cells transformed in vivo or in vitro. The finite life span of most normal cells in vitro may be in some ways analogous to aging and the finite life span in vivo (Macieira-Coelho, 1993; Smith and Pereira-Smith, 1996). The process of immortalization in vitro probably involves spe- cific genetic alterations linked to chromosomal changes, although “epigenetic” mechanism(s) cannot be ruled out in some instances (Trott et al., 1995).
Several workers have shown that “normal” cells or those transformed in vitro but exhibit- ing only a few of the criteria (Table 14.1) may grow as neoplasms in vivo when inoculated to- gether with supporting structures—for example, glass beads (Boone, 1975), plastic plates (Sanford et al., 1980), or gelatin sponge (Okada et al., 1992). Barrett (1980) found that primary cultures of hamster embryo cells may be transformed to a stage in which neoplasms are pro- duced in vivo only when the cultured cells are inoculated with polymer supports (Table 14.2). Several model cell lines that are aneuploid but commonly used in transformation studies in vitro, for instance C3H/10T1/2 and 3T3 mouse lines, exhibit such behavior (Boone and Jacobs, 1976); this indicates that all or some of these cell lines are beyond the stage of initiation. Furthermore, such cell lines may be transformed by alterations in the density of the cell culture (Haber and Thilly, 1978; Rubin et al., 1995), alkalinity of the medium (Oberleithner et al., 1991), polyamines (Tabib and Bachrach, 1998), anoxia (Anderson et al., 1989) and glass fibers (Whong et al., 1999), conditions that are unlikely to or have not induced transformation in primary cul- tures. On the basis of such findings, studies of the neoplastic transformation in vitro should be
Table 14.2 Characteristics of Stages in Syrian Hamster Embryo (SHE) Cell Transformation to Neoplasia
carried out with cells of known genotype and phenotype, ideally shortly after primary explanta- tion and while still euploid.
It is clear that the ultimate standard of the neoplastic transformation is still the establish- ment of malignant neoplasms in vivo, but all of the characteristics listed above (Table 14.1) ap- ply to transformed cells in vitro, and, taken together, they strongly support the identification of a clone of cells transformed in vitro as biologically neoplastic. Still, with the possible exception of neoplasms induced on inoculation in vivo, no single criterion can mark a cell transformed in culture as biologically neoplastic.