Effect of feeder cell-released substances on the survival of clonogenic 9L cells after treatment with antimetabolites.

Exponentially growing monolayer cultures of 9L rat brain tumor cells were treated with either 5-fluorouracil or methotrexate. The surviving fraction of cells was determined by a colony formation assay. After 5-fluorouracil treatment, 2 to 5 X 10(5) feeder cells were required to maximize surviving fractions for each drug concentration and to generate a biphasic dose-response curve. If only 1 X 10(4) feeder cells were used, the dose-response curve was steep. Uridine added to the dishes containing 1 X 10(4) feeder cells restored the biphasic shape, while uridine and thymidine added to the dishes yielded the same curve obtained with 2 X 10(5) feeder cells. After methotrexate treatment, the surviving fraction of cells was dependent on feeder cell number when the medium in the dishes was supplemented with dialyzed fetal bovine serum, but it was not dependent on feeder cell number when nondialyzed fetal bovine serum was used. Biphasic dose-response curves were generated from methotrexate-treated cells plated in medium supplemented with either dialyzed or nondialyzed serum, but the drug was more toxic to cells plated in medium containing dialyzed serum. This additional toxicity could be reduced if either thymidine or N-5-formyltetrahydrofolate were added to the dishes and eliminated if 1 X 10(4) feeders were added. These results suggest that any cell culture system used to evaluate antimetabolites should be optimized for possible feeder cell and serum effects.