Comparative effects of adenine analogs upon metabolic cooperation between Chinese hamster cells with different levels of adenine phosphoribosyltransferase activity

Colony formation by variant Chinese hamster cells highly resistant to adenine analogs and deficient in adenine phosphoribosyltransferase (APRT) activity was measured after co-cultivation with APRT⁺, CHO-K1 cells in medium containing one of three different adenine analogs. Depending upon the density of APRT⁺ cells and the specific adenine analog, large differences in the recovery of APRT^? colonies were observed. The particular adenine analog and APRT⁺ cell density were more significant factors in the recovery of APRT^? colonies than the concentration of the analog or the level of APRT activity. The number of wild-type cells (CHO-K1) required to inhibit formation of APRT^? colonies by 50% (mean lethal density; MLD₅₀) with 65 μg/ml 8-aza-adenine (AzA) as the selective drug was 8.0 × 10⁵ cells/100 mm dish (1.5 × 10⁴/cm²). With 100 μg/ml 2,6-diaminopurine (DAP) the MLD₅₀ for CHO-K1 was 4.0 × 10⁵ cells/100 mm dish (7.3 × 10³/cm²). The MLD₅₀ for CHO-K1 when the DAP concentration was decreased to 50 μg/ml was only slightly higher, 5 × 10⁵ cells/100 mm dish (9.1 × 10³/cm²). The most toxic effect was observed with 2-fluoroadenine (FA). The MLD₅₀ for CHO-K1 in 2 μg/ml FA was 4.5 × 10⁴ cells/100 mm dish (8.2 × 10²/cm²), a cell density which permits minimal direct contact between APRT⁺ and APRT^? cells. The toxic effects of FA on individually resistant, APRT^? cells were found to be mediated by metabolites released into the medium by dying APRT⁺ cells. This metabolite toxicity to APRT^? cells was also demonstrated in mixtures with cells having only 8% of wild-type APRT activity. The MLD₅₀ for these APRT⁺ (8%) cells in 2 μg/ml FA was 7.5 × 10⁴ cells/100 dish (1.4 × 10³/cm²), a small difference from the MLD₅₀ for cells with wild-type levels of APRT activity. The differences in the recovery of APRT^? colonies from mixtures with APRT⁺ cells in these three adenine analogs are critical to the design of procedures for the selection of APRT^? cells from populations of APRT⁺ cells and emphasize the importance of establishing the parameters of metabolic cooperation, not only in terms of cell density but also with regard to the particular selective agent, in any experiment designed to determine precise mutation rates or to test putative mutagens upon mammalian cells in culture.