6-Substituted 2,4-diaminopyrido[3,2-d]pyrimidine analogues of piritrexim as inhibitors of dihydrofolate reductase from rat liver, Pneumocystis carinii, and Toxoplasma gondii and as antitumor agents.

The synthesis and biological activity are reported for 21 6-substituted 2,4-diaminopyrido[3,2-d]pyrimidine analogues (4-24) of piritrexim (PTX) as inhibitors of dihydrofolate reductase (DHFR) and as antitumor agents. Recombinant DHFR from Pneumocystis carinii (pc) and native DHFR from Toxoplasma gondii (tg) were the target enzymes tested; these organisms are responsible for fatal opportunistic infections in AIDS patients. Rat liver (rl) DHFR served as the mammalian reference enzyme to determine selectivity for the pathogenic DHFR. The synthesis of S9-bridged compounds 4-6 was achieved by aryl displacement of 2,4-diamino-6-chloropyrido[3, 2-d]pyrimidine (27) with thiol nucleophiles. Oxidation of 4-6 with hydrogen peroxide in glacial acetic acid afforded the corresponding sulfone analogues 7-9. The N9-bridged compounds 10-24 were synthesized from their precursor 3-amino-6-(arylamino)-2-pyridinecarbonitriles via a thermal cyclization with chloroformamidine hydrochloride. Unlike the S9-bridged compounds, the arylamino side chains of the N9-bridged analogues were introduced prior to the formation of the 2, 4-diaminopyrido[3,2-d]pyrimidine nucleus. A reversed two-atom-bridged analogue (25) was also synthesized using a synthetic strategy similar to that utilized for compounds 10-24. The IC50 values of these compounds against pcDHFR ranged from 0.0023 x 10(-6) M for 2,4-diamino-6-(N-methyl-3',4'-dimethoxyanilino)pyrido[3, 2-d]pyrimidine (21), which was the most potent, to 90.4 x 10(-6) M for 2,4-diamino-6-(4'-methoxyanilino)pyrido[3,2-d]pyrimidine (12), which was the least potent. The three S9-bridged compounds tested were more potent than the corresponding sulfone-bridged compounds for all three DHFRs. N9-Methylation increased the potency by as much as 17 000-fold (compounds 15 and 21). None of the analogues were selective for pcDHFR. Against tgDHFR the most potent analogue was again 21 with an IC50 value of 0.00088 x 10(-6) M and the least potent was 12 with an IC50 of 2.8 x 10(-6) M. N9-Methylation afforded an increase in potency of up to 770-fold (compound 15 NH vs 21 N-CH3) compared to the corresponding N9-H analogue. In contrast to pcDHFR, several analogues had a greater selectivity ratio for tgDHFR compared to trimetrexate (TMQ) or PTX, most notably 2, 4-diamino-6-[(3',4'- dimethoxyphenyl)thio]pyrido[3,2-d]pyrimidine (4), 2,4-diamino-6-[(2'-methoxyphenyl)sulfonyl]pyrido[3, 2-d]pyrimidine (7), and 2,4-diamino-6-(2', 5'-dimethoxyanilino)pyrido[3,2-d]pyrimidine (14) which combined relatively high potency at 10(-7)-10(-8) M along with selectivity ratios of 3.97, 6.67, and 4.93, respectively. Several analogues synthesized had better selectivity ratios than TMQ or PTX for both pcDHFR and tgDHFR, and the potencies of the N9-methylated compounds were comparable to or greater than that of TMQ or PTX. Selected compounds were evaluated as inhibitors of the growth of a variety of tumor cells in culture. The N9-CH3 analogues were, in general, highly potent with GI50 values in the nanomolar range. The N9-H and S9 analogues were less potent with GI50 values in the millimolar to micromolar range.