Mechanisms of artemisinin resistance in Plasmodium falciparum malaria

Abstract

Artemisinin-based combination therapies (ACTs) have substantially reduced worldwide malaria burden and deaths. But malaria parasites have become resistant to artemisinins. Prior studies suggested two different molecular pathways of artemisinin-resistance. Here we unify recent findings into a single model, where elevation of a lipid, phosphatidylinositol-3-phosphate (PI3P) results in vesicle expansion that increases the engagement with the unfolded protein response (UPR). Vesicle expansion (rather than increasing individual genetic determinants of the UPR) efficiently induces artemisinin resistance likely by promoting 'proteostasis' (protein translation coupled to proper protein folding and vesicular remodeling) to mitigate artemisinin-induced proteopathy (death from global abnormal protein-toxicity). Vesicular amplification engages the host red cell, suggesting that artemisinin resistant malaria may also persist by taking advantage of host niches and escaping the immune response.

Figure 1:

Life cycle of Plasmodium falciparum and measures of artemisinin drug resistance. (a) (Adapted from Delves, M., Scheurer, C, et al. 2012 [48]). P. falciparum malaria infection in humans is initiated when an infected mosquito bites and releases sporozoites that infect liver cells and develop into merozoites. These merozoites emerge into the blood and invade erythrocytes (red blood cells), where they develop over a 48-hour asexual life cycle through morphologically defined ring, trophozoite and schizont stages. Schizont lysis leads to the release of daughter merozoites (which coincides with fever), to initiate a new asexual life cycle. A small subpopulation of ring-form parasites develops into sexual gametocyte stages which are taken up by the mosquito during a blood meal. The killing activity of artemisinins (against asexual and early sexual gametocyte stages) is dependent on the cleavage of their endoperoxide bond. Clinical resistance to artemisinins is seen at the parasite ring stage. (b) Clinical artemisinin resistance is measured in vitro by the Ring-stage Survival Assay (RSA). (c) In the RSA, ring-stage parasites 0–3 hour, are treated with maximal concentrations (700 nM) of dihydroartemisinin (DHA, the active form of all artemisinins) seen in plasma for 6 hours, after which the drug is washed out, mimicking pharmacological exposure seen in patients. Parasites are subsequently allowed to progress for another 66 hours, after which parasitemia is determined. The RSA value is calculated as shown. (d) The standard IC50 assay is carried out by exposing ring, trophozoite and schizont stages to continuous drug treatment over 72 hours. (e) Plasmodium falciparum K13 (PfKelch13) is a causal molecular marker of artemisinin resistance. It contains a single BTB domain (common to all Broad-complex, Tramtrack and Bric-a-bric domain (BTB) family proteins), which is present at the amino terminus, followed by multiple copies of β-propeller ‘Kelch’ repeats. K13 contains six β-propeller Kelch domains, mutations in which induce artemisinin resistance.

Figure 2:

Mechanisms of artemisinin resistance. (a–e) Proteomic and cellular studies of clinical and lab-engineered artemisinin resistance. (Summarized from Bhattacharjee et al. 2018 [11^●●]). (a) K13 (red star) is a marker for PI3P (black dots) in vesicles (yellow) enriched in the parasite ER (dark blue). K13-PI3P vesicles are also observed in the food vacuole (orange) and apicoplast (green). These vesicles were not enriched at the parasite mitochondria (M), nucleus or plasma membrane. (b) Biochemical isolation followed by proteomic analysis of the 3D7K13 vesicle revealed several parasite pathways of proteostasis systems enriched in reactive oxidative stress complex (ROSC), Chaperone Assisted Folding; T-complex chaperone; HT/PEXEL/PNEP (host-targeting/plasmodium export element/PEXEL-negative exported proteins) cargo; posttranslational translocation; translation; ER/Golgi/vesicle transport and quality control; plasmodium translocon of exported proteins (PTEX); proteasome mediated proteolysis of ubiquitinylated proteins. (c) Hypergeometric analyses revealed significant overlap between the K13-vesicle proteome and upregulated (but not downregulated) genes of the in vivo artemisinin resistant transcriptome (TRAC) [10]. (d) Synthetic resistance by elevation of PI3P induced independent of K13 mutation (using VPS34myc) yielded vesicular proteomes enriched in pathways of ROSC, Chaperone Assisted Folding, T- complex chaperones, HT/PEXEL/PNEP, Posttranslational translocation, Translation, Proteasome mediated proteolysis of ubiquitinylated proteins. Hypergeometric analyses revealed significant overlap between the synthetic PI3P (VPS34myc) proteome and upregulated (but not downregulated) genes of the in vivo artemisinin resistant transcriptome (TRAC; c). (e) Vesicular K13 and synthetic PI3P proteomes predict a system of vesicular proteostasis containing oxidative stress responses to protein damage, the UPR and ERAD (ER-associated degradation) pathway including (but not limited to) key ER proteins such as PDI (protein disulfide isomerase), BiP (heat shock protein 70). This model is validated by localization of K13 to vesicles (panel a). (f–i). Transcriptomic studies of in vitro lab selected artemisinin resistance (summarized from Rocamora et al., 2018 [12^●●]). (f) In vitro selection of P. falciparum clones after long term exposure to artemisinins. The 3D7 strain was exposed at mid ring (10 hour+) to short (4 hour) pulses of a clinically relevant dose (900 nM) of artemisinins continuously for two years. (g) Two in vitro-selected resistant clones 6A-R and 11C-R were generated. (h) Their transcriptional profiling showed enhancement of adaptive responses against oxidative stress and protein damage (shown in bold). (i) Normalized Gene Set Enrichment Analysis (GSEA) Score was used to depict overlap between significantly upregulated functionalities of the in vitro lab selected resistant lines and in vivo artemisinin resistant transcriptome [10]. (j) Ring-stage survival assay (RSA) values associated of indicated strains reported by Bhattacharjee et al., 2018 [11^●●], Mbengue et al., 2015 [9], Rocamora et al., 2018 [12^●●].

Figure 3:

Model unifying proteostasis in ER and cytoplasm of Plasmodium falciparum in mechanisms of artemisinin resistance. Pathways of protein quality control in the ER and the cytoplasm stimulate concerted mechanisms of protein translation, translocation, vesicular export and additional chaperone functions to enable proteostasis and thereby restore proper folding of proteins and their function in a cell. A model is proposed for proteostasis pathways to rescue Plasmodium falciparum parasites from artemisinin-induced protein damage, proteopathy and death. In artemisinin-resistant parasites, K13 mutations prevent binding, ubiquitinylation and degradation of PfPI3K (step 1, [9]). This leads to an increase in the kinase and thereby its lipid product, PI3P (step 2) causing expansion of homeostatic PI3P-vesicles of proteostasis from the ER (step 3, [11^●●]) that may underlie a mechanism of autophagy. In addition, these vesicles contain BiP, a key component of the ROSC. BiP is usually bound to UPR transmembrane receptors, but under conditions of stress disassociates from the membrane and binds to misfolded proteins, shifting equilibrium away from and activating UPR receptors. In P. falciparum, the UPR receptor, ER transmembrane sensor protein-kinase R (PKR)-like ER kinase (PERK, also known as PK4) has been shown to phosphorylate elongation initiation factor 2a (eIF2a), leading to translational repression, and a reduction of general protein synthesis in artemisinin resistant parasites (step 4, [16^●●]). Increase in BiP transcript levels in vivo artemisinin resistance (step 5) [10], may reflect a response linked to steps 4 and 3. The T-complex protein 1 (TCP1) ring complex (TRiC) chaperone transcript also increased in in vivo artemisinin resistance [10] may enable misfolded proteins in the cytoplasm to become properly folded (step 6) appears associated with step 3 [11^●●]. In the ER-associated degradation (ERAD) pathway misfolded proteins bound and unfolded by BiP (step 5), are translocated to the cytoplasm, ubiquitinylated and degraded in the proteasome (through the ubiquitin-26S proteasome pathway (step 7). Misfolded proteins in the cytoplasm that are not rescued by TRiC maybe ubiquitinylated and targeted for degradation in the proteasome (step 8, [32]).

Figure 4:

Expansion of PI3P vesicles by the major K13C580Y mutation of artemisinin resistance. (a) In Plasmodium falciparum wild type (WT) cells, PI3P vesicles (black in yellow spheres) with associated K13 (red star) are seen at parasite ER, apicoplast and food vacuole. (b) In the major mutation of artemisinin resistance K13C580Y, PI3P vesicles are amplified exported from the ER and disseminated in all organelles throughout the parasite as well as the erythrocyte, where PI3P is detected on vesicles and ‘Maurer’s clefts’ known to mediate the export of virulence determinants such as a major adhesin family (PfEMP1) to infected host cell surface. (c) Export of PfEMP1 (blue spikes) and cytoadherence of infected erythrocytes (blue circles with brown spheres) to host receptors (green line) are diminished by dihydroartemisinin (DHA) which is known to reduce levels of PI3P in artemisinin sensitive parasites [9,11^●●]. (d) Export of PfEMP1 and cytoadherence are not diminished by DHA in artemisinin resistant parasites that show elevation in PI3P vesicles in the parasite and erythrocyte [11^●●].