ABSTRACT
Inhibitors of dihydrofolate reductase (DHFR) have been mainstays in the treatment of falciparum malaria. Resistance to one of these antifolates, pyrimethamine, is now common in Plasmodium falciparum populations. Antifolates have not traditionally been recommended for treatment of vivax malaria. However, recent studies have suggested that a third-generation antifolate, WR99210, is remarkably effective even against highly pyrimethamine-resistant parasites from both species. Two methods were used to identify a compound that is effective against quadruple mutant alleles from P. falciparum (N51I/C59R/S108N/I164L) and from Plasmodium vivax (57L/111L/117T/173F). The first was simple yeast system used to screen a panel of WR99210 analogs. The biguanide prodrug, JPC-2056, of the 2-chloro-4-trifluoromethoxy analog of WR99210 was effective against both the P. falciparum and P. vivax enzymes, and has been selected for further development. The second method compared the analogs in silico by docking them in the known structure of the P. falciparum DHFR-thymidylate synthase. The program reproduced well the position of the triazine ring, but the calculated energies of ligand binding were very similar for different compounds and therefore did not reproduce the observed trends in biological activity. The WR99210 family of molecules is flexible due to a long bridge between the triazine ring and the substituted benzene. During docking, multiple conformations were observed for the benzene ring part of the molecules in the DHFR active site, making computer-based predictions of binding energy less informative than for more rigid ligands. This flexibility is a key factor in their effectiveness against the highly mutant forms of DHFR.
