Rates of development of methotrexate resistance in heterogeneous mouse mammary tumor cell cultures.

Our previous studies have indicated that the expression by tumor cells of sensitivity to chemotherapeutic drugs such as methotrexate can be affected by the presence of other tumor cells; thus, otherwise methotrexate-resistant cells may respond to that drug in the presence of methotrexate-sensitive cells. In order to determine whether tumor heterogeneity also affects the emergence of drug resistance, we measured the rate of development of methotrexate resistance in mixed monolayer cultures of three mammary tumor subpopulation lines (66, 168TFAR, 4T07) that differ in degree of sensitivity to methotrexate. Cultures were treated weekly with 80 nM or 200 nM methotrexate. Each individual cell line was re-isolated from the mixture by passage in selective medium and then assayed for methotrexate sensitivity. Cultures of each of the three lines were treated and assayed in parallel. Few differences in the rate of development of methotrexate resistance were seen among cells from mixtures and cells cultured alone; line 4T07 appeared to become resistant somewhat more rapidly in mixtures. In untreated mixed cultures, line 66, the line least sensitive to methotrexate, gradually became dominant; this process was accelerated in treated cultures. One methotrexate-resistant subline from each parent cell line was tested to determine the mechanisms by which methotrexate resistance was increased. Two lines appeared to have increased levels of dihydrofolate reductase, and one exhibited decreased methotrexate transport as well. The third cell line had neither mechanism. Others have shown that tumor heterogeneity can act as a brake on the rate of development of new metastatic or immunogenic variants. Our data indicate that, at least in the model system we have tested, the rate of development of extrinsic drug resistance is substantially independent of pre-existing clonal diversity.

Tumor subpopulation interactions affecting melphalan sensitivity in palpable mouse mammary tumors.

Paired mixtures of melphalan-sensitive and relatively insensitive tumor cell subpopulation lines, originally derived from the same mammary tumor, were injected s.c. into syngeneic mice. When tumors were palpable, the mice were treated with melphalan at doses shown to be effective against the melphalan-sensitive subpopulations. Sensitivity was assessed by the loss of colony-forming ability of tumor cells harvested 1 to 14 days after treatment. When growing in tumors mixed with melphalan-sensitive line 4TO7 cells, line 66 (less sensitive) appeared much more sensitive than when it was grown alone. Line 66 tumors growing on the opposite sides of mice bearing line 4TO7 tumors were not more sensitive than when grown alone, indicating the lack of a systemic mechanism in the transfer of sensitivity from 4TO7 to 66. Furthermore, line 66 was not more sensitive when mixed with line 168TFAR (another melphalan-sensitive subpopulation) than when alone. The "transfer of sensitivity" from line 4TO7 to line 66 could be reproduced in collagen gel cultures but not in monolayer. Interestingly, line 4TO7, unlike line 168TFAR, is more sensitive to melphalan in collagen culture than in monolayer. This difference in sensitivity does not appear to be influenced by differences in cell density between the two culture systems. In collagen culture, the increased sensitivity of line 66 in the presence of line 4TO7 did not require cell contact and so appeared to act through diffusible factors(s). Collectively, these data suggest that the transfer of sensitivity is not dependent upon host factors or upon drug sensitivity per se but rather upon some mechanism requiring tumor cell-tumor cell interaction between specific subpopulation pairs. In additional studies, pH was ruled out as a factor in the transfer of sensitivity.