The proof of concept chemogenomic profiling experiments

The proof of concept chemogenomic profiling experiments in P. falciparum (Fig. 3A) showed that the primary drug target can be revealed as a significant EC50 ratio. All antifolates were identified as DHFR-TS targeting drugs, despite the wide variation in antimalarial potency from the extremely potent WR99210 to PYR in the K1 background, representing a greater than 106 difference in EC50 value (Supplementary Table S2). The test of the K1_glmS parasite with PYR showed that even the mutated DHFR-TS in this strain, known to have lower affinity for PYR (Foote et al., 1990, Yuvaniyama et al., 2003), can still be identified as a PYR target in vivo, albeit with a lower EC50 ratio than the PYR-sensitive 3D7_glmS parasite. The reason for the lower EC50 ratio in the K1_glmS parasite is not understood. However, other PYR resistance mechanisms, such as increased gch1 copy number (Kidgell et al., 2006, Nair et al., 2008, Kumpornsin et al., 2014), may contribute towards the lower EC50 ratio. The ability of this chemogenomic profiling approach to identify the primary target is further highlighted by the insignificant EC50 ratio for DSM1. This drug is known to target dihydroorotate dehydrogenase (PfDHODH), an enzyme which acts upstream of DHFR-TS in the pyrimidine biosynthetic pathway (Phillips et al., 2008). The chemogenomic approach for identifying targets in vivo can be adapted for high-throughput screening, as shown by our study of the Malaria Box THZ531 library. The screening data show that attenuation of DHFR-TS has no effect on parasite sensitivity to most of this diverse set of compounds, highlighting the specificity of the approach. A conservative threshold of the EC50 ratio (six SDs from the mean) was selected for identifying DHFR-TS targeting compounds, as the method used for estimating the EC50 ratio in the screening experiment is likely to have substantial error. High-throughput antimalarial screening procedures can be completed with much smaller volumes (Plouffe et al., 2008, Gamo et al., 2010, Guiguemde et al., 2010), such that more data points could be obtained for each compound, leading to more accurate estimate of the EC50 ratio and a lower threshold for identifying target-specific compound hits. The compounds MMV667486 and MMV667487, identified from screening as DHFR-TS targeting compounds, possess the antifolate pharmacophore common with CYC, i.e. dihydrotriazine (Yuthavong, 2002). Although the antifolate pharmacophore is important for interaction with DHFR-TS, it is not sufficient for target specificity in vivo (Plouffe et al., 2008). Chemogenomic profiling therefore provides extra information of target specificity that cannot be predicted from chemical structures alone. Furthermore, it is possible that novel anti-DHFR-TS pharmacophores could be found by chemogenomic profiling of larger and more diverse compound libraries. The data from drug screening in P. berghei (Fig. 3D) provide proof of concept that chemogenomic profiling can be accomplished in P. berghei in the same manner as P. falciparum, in that significant increases in the EC50 ratio were found only for antifolate drugs in the parasite with attenuated TgDHFR-TS expression. That increased antifolate sensitivity under this condition was observed despite the presence of endogenous PbDHFR-TS indicates that chemogenomic profiling is not limited to parasites with modification of endogenous drug target only. The specificity of the approach is highlighted by the insignificant EC50 ratio for SDX, a drug targeting dihydropteroate synthase that acts immediately upstream of DHFR-TS in the folate/pyrimidine metabolic pathway (Triglia et al., 1998). The significant EC50 ratio for PGN suggests that this compound targets DHFR-TS, perhaps because conversion to the antifolate drug CYC is manifest in the P. berghei in vitro system, possibly by leukocytes. The low EC50 ratio of PGN compared with antifolate drugs, however, suggests that conversion to CYC occurs at a low level, or PGN has other significant targets. The insignificant EC50 ratio for PGN in P. falciparum suggests that this compound does not target PfDHFR-TS, in agreement with an earlier study using a different transgenic parasite approach (Fidock et al., 1998).