Benign-to-malignant B16 melanoma progression induced in two stages in vitro by exposure to hypoxia

Benign tumors apparently become malignant by generating a succession of variants with ever-greater growth potential and autonomy. Such stepwise progression has not been achieved in vitro under conditions likely to occur in developing tumors. Tumors initiated by clone G3.5 of the mouse B16 melanoma regularly generate stable variants that are more malignant. We investigated the possibility that hypoxia might promote stepwise progression along a benign-to-malignant pathway in monolayer cultures of G3.5 cells. Confluent monolayers of metastatic clone G3.5 and the nonmetastatic clone G3.15 were subjected to severe hypoxia (< 50 ppm O2) for up to 72 hours, or to moderate hypoxia (300-1200 ppm O2) for up to 12 days, and were then maintained subconfluent or at confluence for several weeks to permit emergence of progression variants. The relative malignancy of variants was assayed in vivo after subcutaneous injection into mice, by measuring tumor growth rate and counting lung metastases, and after intravenous injection, by counting lung colonies. In vitro assessment of the variants involved growth as monolayers with or without serum, growth in soft agar, and measurement of invasiveness. G3.5 cells were converted to a more malignant variant (G3.5*) by 12-48 hours of severe hypoxia, or longer periods of moderate hypoxia, when followed by maintenance at confluence for 3-5 weeks. Conversions occurred in discrete foci of morphologically-discernible cells (optimum focus formation about one in 1-2 x 10(5) cells) that rapidly expanded to dominate the cultures. The G3.5* phenotype was comparable to the conversion phenotype generated in tumors and included acquisition of growth autonomy in serum-free medium. G3.15 cells were converted to a G3.5-like phenotype by one round of exposure to hypoxia and confluence, and then to the G3.5* phenotype during a second round, at a low frequency (one focus in 5 x 10(6) cells). This behavior was consistent with a failure of all but the largest G3.15 tumors to generate G3.5* conversion cells. Progression from a relatively benign phenotype, G3.15, to a highly malignant phenotype, G3.5*, can be produced in monolayer culture in two stable stages by sequential rounds of exposure to hypoxia and confluence. The resulting conversions corresponded to phenotypes generated within tumors. Both conversions resulted in populations with enhanced growth capabilities, which could establish dominance within tumors. The stepwise conversion of B16 melanoma c