New folate analogs of the 10-deaza-aminopterin series Basis for structural design and biochemical and pharmacologic properties

Summary

Structural modification of the N¹⁰ position of 4-amino folates affects mediated membrane transport in mammalian cells but has little or no effect on target enzyme (dihydrofolate reductase) inhibition. Some of these modifications have been associated with differential effects on transport in various cell types in a manner which favored greater accumulation and persistence of drug in reponsive tumor cells than in normal proliferative tissue. With the aim of identifying new structures with greater potential for differential mediated accumulation, we have studied three new 10-alkyl analogs of 10-deaza-aminopterin. Two of these analogs showed therapeutic efficacy substantially greater than 10-deaza-aminopterin, an analog with antitumor properties superior to methotrexate. These analogs, the 10-methyl, 10-ethyl, and 10,10-dimethyl derivatives, were equivalent to the parent compound, 10-deaza-aminopterin, and aminopterin, and slightly more potent than methotrexate, as inhibitors of L1210 cell dihydrofolate reductase. The three new analogs, 10-deaza-aminopterin, and aminopterin exhibited similar transport properties in L1210, Ehrlich, and S180 cells. Efflux and influx V_max were similar to those of methotrexate, but influx K_m was 4- to 14-fold lower than for methotrexate. That is, substitution at N¹⁰, but not at C¹⁰, reduced influx potential in these tumor cells. These differences in transport properties among this group of analogs which determine net accumulation were reflected in the individual values for growth-inhibitory potency. In contrast to that seen in tumor cells, alkylation at both N¹⁰ and C¹⁰ reduced influx potential (increased K_m) in isolated intestinal epithelial cells from mouse small intestine. Influx was in the order aminopterin >10-deaza-aminopterin with further reduction in each series showing a magnitude in proportion to the size of the 10 substituent. Otherwise, influx V_max and efflux were similar for the group. Accumulation of polyglutamates in small intestine was greater following aminopterin administration than following administration of other analogs (10-ethyl, 10-deaza-aminopterin < methotrexate < 10-deaza-aminopterin). Polyglutamate accumulation for all the analogs was greater in tumor cells, but accumulation of each varied between the two tumors (L1210 and S180) examined. Differences among the analogs were not as great in L1210 as in S180 cells, and their metabolism was not in the same relative order. Plasma pharmacokinetics for 10-methyl and 10-ethyl derivatives were similar to those for 10-deaza-aminopterin and methotrexate. The 10,10-dimethyl analog was cleared more rapidly.

As in our prior reports, a greater selective action of some of the new analogs was associated with increased persistence of the analog in tumor versus small intestine. The greatest differential in persistence was found with analogs which had a lower value for influx K_m in tumor but a higher value in intestinal epithelium. Two analogs (10-deaza-aminopterin and 10-ethyl, 10-deaza-aminopterin) which exhibited identical transport properties but different extents of polyglutamylation had pharmacokinetics in tumor that were indistinguishable. However, the analog that was more rapidly polyglutamylated (10-ethyl, 10-deaza-aminopterin) was therapeutically more effective.