A broad analysis of resistance development in the malaria parasite

Abstract

Microbial resistance to chemotherapy has caused countless deaths where malaria is endemic. Chemotherapy may fail either due to pre-existing resistance or evolution of drug-resistant parasites. Here we use a diverse set of antimalarial compounds to investigate the acquisition of drug resistance and the degree of cross-resistance against common resistance alleles. We assess cross-resistance using a set of 15 parasite lines carrying resistance-conferring alleles in pfatp4, cytochrome bc1, pfcarl, pfdhod, pfcrt, pfmdr, pfdhfr, cytoplasmic prolyl t-RNA synthetase or hsp90. Subsequently, we assess whether resistant parasites can be obtained after several rounds of drug selection. Twenty-three of the 48 in vitro selections result in resistant parasites, with time to resistance onset ranging from 15 to 300 days. Our data indicate that pre-existing resistance may not be a major hurdle for novel-target antimalarial candidates, and focusing our attention on fast-killing compounds may result in a slower onset of clinical resistance.

Figure 1. Chemically diverse compound set.

Hierarchical clustering of the 50 compound set. Compounds were clustered by a maximum substructure similarity Tanimoto coefficient. In vitro selections that were successful in yielding resistant parasites are highlighted in blue, whereas compounds where resistance development was unsuccessful are highlighted in dashed red.

Figure 2. Cross-resistance fold shifts observed in compound set.

(a) A total of 15 resistant strains were tested with each MMV compound to identify potential pre-existing cross-resistance. Calculating the fold shifts between each clone and either a corresponding parent or a drug sensitive 3D7 strain generated the heatmap. To normalize conferred resistance and sensitivity, the natural log of each fold shift is displayed. Fold shifts instead of raw data were used as multiple assays were run with varying times and detection indicators. Incomplete cross-resistance assays are depicted in white. All assays were run in triplicate. For one compound (BRD1095), a close analogue (BRD3444) was analysed for TM90C2A and PfATP4-Mut1–3. SMILEs for all compounds are listed in Supplementary Data 1. (b) Chemical structures of the two MMV compounds (MMV009108 and MMV028038) with increased efficacy against one or more pfatp4 mutated clones. Both compounds displayed low structural similarity to a number of other known pfatp4 inhibitors. (c) Chemical structures of the two MMV compounds (MMV019066 and MMV008149) with decreased efficacy against one or more cytochrome bc1 mutated clones. Atovaquone and decoquinate, two other cytochrome bc1 inhibitors, were structurally significantly different.

Figure 3. Killing rate trends.

(a) Proportion of fast (CQ), moderate (PYR or PYR/CQ), and slow (ATQ) compound killing rates in our compound set (49 compounds in analysis). (b) Proportion of failed and successful compound selections sorted by killing rate (47 compounds in analysis). (c) Successful selections further proportioned out based on amount of time required to result in resistant parasites. Trend in selection success and killing rate was found to be statistically significant (P=0.0022) by a one-tailed Fisher's exact test.