Loss of controlled nuclear division in BHK21 cells passed in vivo.

Low-passage BHK21/C13 cells respond to cytochalasin B (CB) by undergoing limited or controlled nuclear division. These cells respond to CB as normal cells do since nuclear division usually occurs only once. Premature chromosome condensation, a result of mitoses in highly multinucleate cells, occurs in less than 0.5% of the cells. When they are inoculated into weanling hamsters, s. c. tumors appear within 3 to 4 weeks with as few as 10(3) cells. When these cells are returned to cell culture they respond to CB with uncontrolled nuclear division and premature chromosome condensation. All cultured tumors respond in this manner regardless of the number of cells originally inoculated into animals. This suggests at least two possibilities: (a) that loss of controlled nuclear division in BHK cells is closely associated with or required for tumorigenicity, with in vivo passage selecting for a rare tumorigenic variant or (b) that loss of controlled nuclear division is secondary to tumorigenicity and results when cells are passed in vivo, i.e., in vivo passage has the direct effect of causing cells to lose controlled nuclear division. If the first possibility is operating, then it would be expected that a very small fraction of BHK21/C13 cells show uncontrolled nuclear division (approximately 1 of 1000 CB-treated cells). Also, clones of C13 should be nontumorigenic if only 1 of 1000 cells is tumorigenic. Extensive examination of CB-treated C13 cells shows 1 of 1250 cells to be highly multinucleated although not as highly multinucleated as tumor cells. This provides some evidence in support of the first possibility. However, three separate clones of C13 cells were found to be tumorigenic providing evidence supporting the second possibility. BHK21/C13 and various BHK21 tumors all appeared to grow to concentration densities markedly higher than hamster embryo fibroblasts. However, the tumor cells usually grew to the same density as did BHK21/C13 or only slightly higher. This suggests that loss of contact inhibition is not sufficient for loss of controlled nuclear division. It also suggests that the hypothetical 1 of 1000 tumorigenic cells and 1 of 1250 cells with uncontrolled nuclear division do not overgrow the "normal" cells because all cells grow to similar densities. The relationship between the ability to grow in agar and uncontrolled nuclear division was also examined. Approximately 1 of 2500 C13 cells were able to form colonies in agar. All three colonies isolated showed normal control of nuclear division. These results show that the ability to grow in agar may be separate from the expression of uncontrolled nuclear division. They also suggest a fundamental difference between tumor cells and cells that have been grown in agar.