In vitro resistance selections for Plasmodium falciparum dihydroorotate dehydrogenase inhibitors give mutants with multiple point mutations in the drug-binding site and altered growth

Abstract

Malaria is a preventable and treatable disease; yet half of the world's population lives at risk of infection, and an estimated 660,000 people die of malaria-related causes every year. Rising drug resistance threatens to make malaria untreatable, necessitating both the discovery of new antimalarial agents and the development of strategies to identify and suppress the emergence and spread of drug resistance. We focused on in-development dihydroorotate dehydrogenase (DHODH) inhibitors. Characterizing resistance pathways for antimalarial agents not yet in clinical use will increase our understanding of the potential for resistance. We identified resistance mechanisms of Plasmodium falciparum (Pf) DHODH inhibitors via in vitro resistance selections. We found 11 point mutations in the PfDHODH target. Target gene amplification and unknown mechanisms also contributed to resistance, albeit to a lesser extent. These mutant parasites were often hypersensitive to other PfDHODH inhibitors, which immediately suggested a novel combination therapy approach to preventing resistance. Indeed, a combination of wild-type and mutant-type selective inhibitors led to resistance far less often than either drug alone. The effects of point mutations in PfDHODH were corroborated with purified recombinant wild-type and mutant-type PfDHODH proteins, which showed the same trends in drug response as the cognate cell lines. Comparative growth assays demonstrated that two mutant parasites grew less robustly than their wild-type parent, and the purified protein of those mutants showed a decrease in catalytic efficiency, thereby suggesting a reason for the diminished growth rate. Co-crystallography of PfDHODH with three inhibitors suggested that hydrophobic interactions are important for drug binding and selectivity.

Keywords: Drug Resistance; Evolution; Infectious Disease; Malaria; Nucleotide; Pyrimidine.

SCHEME 1.

Reactions catalyzed by DHODH. DHODH catalyzes two reversible half-reactions. In the first, l-DHO is oxidized by the FMN cofactor. In the second, the reduced FMN cofactor is re-oxidized by a variety of electron acceptors such as fumarate or NAD⁺ in cytosolic class 1 enzymes or CoQ_n in mitochondrial class 2 enzymes. Humans and Plasmodium parasites both have class 2 enzymes. The reduced electron acceptor may be re-oxidized by coupling to dichloroindophenol (DCIP), an indicator dye, to indirectly measure enzyme activity. Alternatively, the production of orotate may be directly measured.

SCHEME 2.

Three types of resistance selections. A primary selection treats wild-type parasites with a wild-type drug. If resistant parasites arise, these mutant parasites can be sequentially treated with a mutant-selective drug. A simultaneous selection treats wild-type parasites with both wild-type and mutant-type-selective drugs at the same time. MT-type, mutant.

FIGURE 1.

In vitro resistance selection results. A, hypothetical resistance mechanisms include resistance by mutation or amplification of the target or a broad other category. B, observed resistance mechanisms were largely due to a SNP in the target. Target copy number changes and other mechanisms also led to resistance. Simultaneous selection with inhibitors with mutually incompatible resistance mechanisms greatly suppressed, but did not eliminate, the emergence of resistance. C, mutations in the target gene that confer resistance line the inhibitor-binding site. Image of Protein Data Bank 3o8a made in CCP4mg is shown (24). Residue 61 (for the N61S resistance mutation) is not part of the crystallography construct and so cannot be shown.

FIGURE 2.

Resistance to PfDHODH inhibitors via mutation in the target gene. A and B, mutations in the target gene pfdhodh gave resistance to the wild-type-selective PfDHODH inhibitor Genz-669178 but gave both resistance and sensitivity to DSM74. C, sensitivity to the unrelated antimalarial dihydroartemisinin was largely unchanged. Although a small effect was seen with dihydroartemisinin in several parasite lines, we do not know if nucleotide flux or PfDHODH-related mitochondrial processes are involved in persistence or resistance to artemisinin-based drugs. D–F, mutations in PfDHODH gave resistance or sensitivity to mutant-selective PfDHODH inhibitors. n.s., not significant (p > 0.5); *, p < 0.05; **, p < 0.01; ***, p < 0.0015.

FIGURE 3.

Resistance to PfDHODH inhibitors through copy number variation of PfDHODH. A and B, copy number variation of the target gene pfdhodh gave resistance to wild-type-selective PfDHODH inhibitors. Note that increased copies led to resistance for Genz-669178, an alkylthiophene, and to both resistance and sensitivity to DSM74, a triazolopyrimidine. C, copy number variation has little effect on the response to the unrelated antimalarial dihydroartemisinin. Although a small effect was seen with dihydroartemisinin in several parasite lines, we do not know if nucleotide flux or PfDHODH-related mitochondrial processes are involved in persistence or resistance to artemisinin-based drugs. D–F, copy number variation of PfDHODH gave resistance to the mutant-selective PfDHODH inhibitors IDI-6253 and GSK3, but occasionally gave sensitivity to IDI-6273. n.s., not significant (p > 0.5); *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 4.

Resistance to PfDHODH inhibitors without mutation or copy number variation in pfdhodh. A–C, unknown resistance mechanisms altered responses to wild-type-selective PfDHODH inhibitors. D, several parasites, including two that were not selected with atovaquone, gained resistance to atovaquone. Atovaquone's target, cytochrome bc₁, is important for PfDHODH function. E, sensitivity to the unrelated antimalarial dihydroartemisinin was largely unaffected. F–H, these parasites also had altered responses to mutant-selective PfDHODH inhibitors. n.s., not significant (p > 0.5); *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 5.

Evolutionary loop in pfdhodh codon 182. A, 3D7: E182D mutant and 3D7 E182D/D182E revertant were approximately equal in fitness to each other, but both were less fit than the wild-type 3D7. Note that the rate of decrease for the initial three cycles is greater than for the subsequent three cycles. B and C, purified PfDHODH protein with codon 182 mutated to aspartate showed substantially lower k_cat and K_m values than the wild-type protein. This suggests that the reduced fitness of the 3D7 E182D mutant may be due to a less catalytically efficient enzyme, but it does not explain the fitness cost seen in the 3D7 E182D/D182E revertant. We propose that the revertant has reduced fitness due to rare codon effects and/or changes outside of the pfdhodh gene.

FIGURE 6.

Enzymatic inhibition mirrors whole cell activity of PfDHODH inhibitors. A, Coomassie-stained gel indicates purity of the various purified PfDHODH proteins. B, sample data for enzymatic inhibition curves show a rightward shift (resistance) of the E182D protein to Genz-669178, which matches the whole cell data. C–F, enzymatic inhibition correlated with whole cell proliferation inhibition, which suggests but does not prove that mutation in PfDHODH causes the observed changes in sensitivity to PfDHODH inhibitors. Cutoff bars indicate an inhibition value greater than the maximum limit of detection of the assay, 2 mm.

FIGURE 7.

Structural plasticity in the PfDHODH inhibitor-binding pocket. A–C, stereo images of PfDHODH crystal structures with ligands Genz-669178 (A), IDI-6253 (B), and IDI-6273 (C) are shown (structures 4cq8, 4cq9, and 4cqa, respectively). D, two residues, histidine 185 and phenylalanine 188, have substantially altered conformations upon ligand binding. Note that Phe-188 was mutated twice in resistance selections for PfDHODH inhibitors (F188I and F188L). Images were made in CCP4mg (24).