The reproducibility of the IC measurements in the

The reproducibility of the IC50 measurements in the Pf-dhfr yeast was far better than comparable measurements in P. falciparum in vitro [1]. This is probably due to the extremely low level of folate and para-aminobenzoic MDL 12330A hydrochloride in the yeast growth medium. This is in contrast to culture of P. falciparum where the standard RPMI medium contains 1 mg l−1 of folic acid and para-aminobenzoic acid. The variation in the levels of folate in the human serum and red cells also contirbute to the variation in culture conditions of the P. falciparum. Data were compiled from a number of different publications where an IC50 was determined for laboratory strains of P. falciparum. The values given were determined in standard RPMI, and all labs used the standard 3H-hypoxanthine uptake assay [42]. The relative sensitivity of the pyrS D6 and the pyrR strains, Honduras 1 and its parent line, HB3 (that carry the mutation to Asn108) was compared with the sensitivity of 7G8 and It.D12, two strains that carry the Ile51 and Asn108 mutations that were observed in the dhfr allele from Mikenga. In each case, the relative resistance is the ratio of the IC50 of the sensitive strain compared with the IC50 for the resistant strain. It is striking that the values reported by different labs can differ by more than an order of magnitude. For example, the reported relative resistance of HB3 ranges from 260 to 5258 15, 18, 19. Variability of this kind is not clinically important; all groups agree in the assignment of strains to the pyrS and pyrR categories. However, Table 3 shows that in the yeast system, the measurement of relative sensitivity to any of these drugs is easy and reproducible. This precision will allow direct assessment of the contribution of each amino acid change to the overall resistance of the enzyme to a particular drug. WR99210 is the active metabolite of PS-15, one of several promising drugs that are potent inhibitors of Pf-dhfr in vitro [4]. WR99210 is effective even against strains that are extremely resistant to pyrimethamine and cycloguanil, like the Honduras 1 and Mikenga strains used here. In fact, both of these pyrimethamine- and cycloguanil-resistant strains were more sensitive to WR99210 than was the pyrimethamine-sensitive D6. This collateral hypersensitivity has been observed for other drug targets where a series of related drugs was compared [38]. The major difference between the D6 and the two pyrR strains is the presence of Asn rather than Ser at the 108 position. Detailed binding and structural studies will be required to identify the cause of the increased susceptibility of these two strains to WR99210. In any case, if strains that express the Asn108 are generally more susceptible to this new drug, it raises the hope that PS-15 may be particularly effective against the pyrimethamine-and cycloguanil-resistant strains that already prevail in large regions of Southeast Asia and South America and are developing rapidly in Africa 2, 13, 14. At this point, the effectiveness of WR99210 has been demonstrated in vitro, but it is not yet known whether all of its action is directed against DHFR 4, 5, 6. When we tested the drug against the yeast transformants expressing each dhfr allele, WR99210 efficiently killed all of them, but the IC50 was about 10−8, almost an order of magnitude higher than the comparable measurement in P. falciparum (Milhous, W. WRAIR internal document; JMW, unpublished). One explanation of this discrepancy could be that WR99210 inhibits more than one function in P. falciparum. Further experiments will be required to test this idea, but the yeast system affords a quick, quantitative way to determine whether a drug really does target the Pf-dhfr and if it does, to separate that action from inhibition of other functions in P. falciparum. The basic strategy used here may be applicable to the study of resistance to a number of other antimalaria drugs. The requirements are clear. First, the drug must inhibit a particular P. falciparum enzyme, either as a competitive inhibitor of the enzyme or by binding to a region other than the active site. Second, the homologous enzyme activity must also operate in S. cerevisiae. Third, the mechanism of resistance must depend on mutations in the gene, not on increases in gene expression or changes in other genes that affect the cell biology of P. falciparum. Fourth, the S. cerevisiae must be permeable to the drug in question and the subcellular localization of the enzyme and the drug must be the same in both systems. A number of malaria enzymes appear to conform to these requirements. Dihydropteroate synthase, the target of sulfa drugs [40], topoisomerases I and II, the target of quinolones 41, 42 and ornithine decarboxylase, the target of difluoromethylornithine and novel tetraamines 43, 44 are all possible candidates for studies of this sort.