A single nucleotide polymorphism in the Plasmodium falciparum atg18 gene associates with artemisinin resistance and confers enhanced parasite survival under nutrient deprivation
- 836 Downloads
Artemisinin-resistant Plasmodium falciparum has been reported throughout the Greater Mekong subregion and threatens to disrupt current malaria control efforts worldwide. Polymorphisms in kelch13 have been associated with clinical and in vitro resistance phenotypes; however, several studies suggest that the genetic determinants of resistance may involve multiple genes. Current proposed mechanisms of resistance conferred by polymorphisms in kelch13 hint at a connection to an autophagy-like pathway in P. falciparum.
A SNP in autophagy-related gene 18 (atg18) was associated with long parasite clearance half-life in patients following artemisinin-based combination therapy. This gene encodes PfAtg18, which is shown to be similar to the mammalian/yeast homologue WIPI/Atg18 in terms of structure, binding abilities, and ability to form puncta in response to stress. To investigate the contribution of this polymorphism, the atg18 gene was edited using CRISPR/Cas9 to introduce a T38I mutation into a k13-edited Dd2 parasite. The presence of this SNP confers a fitness advantage by enabling parasites to grow faster in nutrient-limited settings. The mutant and parent parasites were screened against drug libraries of 6349 unique compounds. While the SNP did not modulate the parasite’s susceptibility to any of the anti-malarial compounds using a 72-h drug pulse, it did alter the parasite’s susceptibility to 227 other compounds.
These results suggest that the atg18 T38I polymorphism may provide additional resistance against artemisinin derivatives, but not partner drugs, even in the absence of kelch13 mutations, and may also be important in parasite survival during nutrient deprivation.
KeywordsArtemisinin resistance Autophagy Fitness Drug resistance atg18
apicoplast ribosomal protein S10
area under the curve
artemisinin-based combination therapy
autophagy-related gene 18
autophagy-related protein 8
chloroquine resistance transporter
genome-wide association study
green fluorescent protein
half-maximal inhibitory concentration
human dihydrofolate reductase
maximum dose response
Mechanism Interrogation Plate
National Center for Advancing Translational Sciences
multidrug-resistance protein 2
P. falciparum autophagy-related protein 18
protospacer adjacent motif
single guide RNA
single nucleotide polymorphisms
Tracking Resistance to Artemisinin Collaboration
transmission electron microscopy
unfolded protein response
WD-repeat protein Interacting with Phosphoinositides
whole genome sequencing
yeast dihydroorotate dehydrogenase
Artemisinin (ART)-resistant Plasmodium falciparum has been reported in the Greater Mekong Subregion since 2007. Artemisinin-based combination therapy (ACT), which pairs a short-acting ART derivative with a long-acting partner drug, is the mainstay of anti-malarial treatment and is likely to have been partly responsible for significantly decreasing malaria-related morbidity and mortality over the past 15 years. The slow parasite clearance rates following ACT suggest resistance to ART derivatives. This resistance places increasing selective pressure for variants or traits that confer resistance to ACT partner drugs and has led to the rapid failure of several artemisinin-based combinations, including dihydroartemisinin-piperaquine in Cambodia [1, 2]. Single nucleotide polymorphisms (SNPs) in kelch13 (k13) have been associated with slow parasite clearance rates in response to ACT  and increased in vitro parasite survival in response to ART derivatives [4, 5].
In clinical isolates, k13 polymorphisms appear on a genetic background comprising polymorphisms in apicoplast ribosomal protein S10 (arps10), chloroquine resistance transporter (crt), ferrodoxin (fd), and multidrug-resistance protein 2 (mdr2) . Although parasites with k13 mutations have been found elsewhere in the world, including Africa, they are not always associated with long parasite clearance half-life in response to ACT [7, 8, 9, 10, 11, 12]. Therefore, the genetic background specific to some areas in Southeast Asia may be responsible for some of the drug resistance phenotype. Alternatively, these background mutations may be important in transmission or confer a survival advantage over parasites without this array of SNPs. Indeed, overcoming the cellular stress response following ART treatment may underlie the parasite resistance mechanism. Autophagy is one such cellular stress response that may be employed by a parasite and therefore may not only be modified by genetic variants that promote survival but also may represent a potential target pathway for novel anti-malarial compounds.
Autophagy, an intracellular process that degrades and recycles damaged organelles, has been well-characterized in many organisms, but is not well-described in P. falciparum. A limited set of putative autophagy-related proteins is encoded by the P. falciparum genome, but the functions of most of these proteins have yet to be defined. Two of these autophagy-related proteins, autophagy-related protein 8 (PfAtg8) and autophagy-related protein 18 (PfAtg18), have been shown to co-localize with the apicoplast and to be involved in apicoplast inheritance . Homologous Atg8 and Atg18 proteins in yeast and mammals have been shown to form puncta during the upregulation of autophagy . Autophagy was investigated as a possible mechanism related to ART resistance due to several connections between an autophagy-like pathway in P. falciparum and known mechanisms of ART action and resistance . Indeed, ART can damage cells by reactive oxygen species (ROS)  and ROS are potent activators of autophagy . This would indicate that ARTs, even in the absence of resistance, could induce an autophagy-like pathway. Several possible resistance mechanisms have been posited wherein the resistant isolate is able to withstand the deleterious effects of ART based on an ability to withstand oxidative stress. Another hint at a connection to autophagy is through increased levels of phosphatidylinositol 3-phosphate (PI3P), a lipid regulating autophagy, that occur in ART-resistant isolates . Lastly, the upregulation of the unfolded protein response (UPR), a process that induces autophagy, is associated with resistance . Therefore, an ostensible connection between an autophagy-like pathway and ART resistance in P. falciparum was investigated.
Several polymorphisms were found in autophagy-related genes that associate with drug-resistant phenotypes, most interestingly a T38I SNP in autophagy-related gene 18 (atg18), which associates with long parasite clearance half-life following ACT. The protein encoded by this gene has similar structural domains, binding capabilities, and behaviour as mammalian and yeast homologues. A T38I CRISPR-edited line was created that determined that the SNP confers increased parasite survival in a nutrient-limited setting, as well as differential susceptibility to over 200 compounds in a 72-h assay that measures the half-maximal inhibitory concentration (IC50) values for ART derivatives.
Bioinformatics of autophagy-related genes
Putative autophagy-related genes in the P. falciparum genome were found through the conversion of a previously published list of human autophagy-related genes from Behrends et al.  and supplemented with autophagy-related genes appearing on the PlasmoDB and Malaria Parasite Metabolic Pathways (MPMP) websites.
A sub-analysis of a previously performed genome-wide association study (GWAS) on 782 isolates from Southeast Asia was performed to determine if any SNPs in genes involved in autophagy were associated with slow parasite clearance rates following ART treatment or partner drug resistance (IC50 values). The resistance phenotype was a quantitative trait in a linear mixed model. A Bonferroni correction was applied, placing statistical significance at p-values less than 5E−6.
IC50 experiments for chloroquine, piperaquine, quinine, artesunate, and DHA were performed ex vivo in Cambodia on a subset of isolates from the GWAS, namely those isolates from in the NIH Cambodian research sites, as previously described  and used a SYBR green fluorescence readout , assessed using a FLUOstar OPTIMA. IC50 calculations were performed using the online IVART IC50 analysis, Worldwide Antimalarial Resistance Network (http://www.wwarn.org/tools-resources/toolkit/analyse/ivart).
These data were pooled and matched with sequencing data from whole genome sequencing (WGS). Samples were excluded from the IC50 analyses if data were missing for either atg18 (n = 265) or k13 (n = 311) or both (n = 196) or if the final sequence was heterogeneous for that locus (atg18: n = 22; k13: n = 37). In total, 334 isolates were excluded, making the final cohort of 448 samples. Data were analysed using GraphPad Prism 6, unpaired t test.
Plasmodium falciparum line production
The homology region sequences were designed to include the atg18 SNP, changing the 38th amino acid from a threonine to an isoleucine (T38I). The protospacer adjacent motif (PAM) guide sequence was edited to include multiple synonymous SNPs to prevent cutting of the new sequence via the PAM-directed Cas9. PciI sites flank the sequences for later infusion reactions. The pL6 plasmid with human dihydrofolate reductase (hDHFR) was modified as described by Ghorbal and colleagues using In-Fusion cloning (Clontech) to include the desired single guide RNA (sgRNA) . A DNA construct for the homology repair region was made by gene synthesis (GENEWIZ) and added to the pL6 plasmid via In-Fusion. sgRNA and homology region sequences were confirmed using DNA sequencing (GENEWIZ). The pUF1-Cas9 plasmid was used for Cas9 expression with yeast dihydroorotate dehydrogenase (yDHODH) resistance cassette .
The plasmid backbone for pDC2 AttP BSD was obtained from Dr. Richard Eastman at the National Center for Advancing Translational Sciences (NCATS). The sequence for atg18 with an N-terminal green fluorescent protein (GFP) was made by Integrated DNA Technologies and inserted into pDC2 AttP plasmid with a BSD resistance cassette and ef1α promoter. Plasmid sequences were confirmed by PCR and DNA sequencing (GENEWIZ).
Uninfected erythrocytes were electroporated in the presence of plasmids using a Bio-Rad Gene Pulser Xcell and Percoll-isolated late-stage parasites were allowed to invade as previously described [23, 24, 25]. Selection for pUF1-Cas9 plasmids was with 1.5 μM DSM1, selection for pL6 was with 2 nM WR99210, and selection for pDC2 AttP BSD was with 4 μM blasticidin (BSD). Drug selection media was changed every day for 1 week until all parasites appeared dead by microscopy. Media was then changed three times a week and blood was added once weekly. Smears were checked three times a week for parasites until day 60 post-transfection. Parasites were detected by Giemsa-stained smears and were screened for editing by PCR. Cultures with positive PCR products associated with editing were sequenced, then cloned by limiting dilution and sequence verified (GENEWIZ). The atg18 T38I locus of Dd2R539T, a laboratory line (Dd2) with a R539T Kelch13 mutation , was edited to create the mutant Dd2R539T/T38I.
Isogenic parasite lines Dd2R539T and Dd2R539T/T38I were screened at NCATS against Mechanism Interrogation Plate (MIPE) 4.1 library  and NPACT library (https://ncats.nih.gov/preclinical/core/compound/npact). MIPE 4.1 contained 1978 compounds, which were screened against the lines at 11 concentrations. NPACT contained 5632 compounds, which were screened at seven concentrations. Because NPACT and MIPE contain some of the same compounds, the total screened were 6349 unique compounds. Concentrations screened were three-fold dilutions for MIPE from 29 µM to 0.5 nM, or five-fold dilutions for NPACT from 29 µM to 1.86 nM. Parasite lines were diluted to 0.3% parasitaemia at 4% Hct and dispensed using a Multidrop Combi in 5 µL volume and 3 µL complete media into 1536-well black clear cyclo-olefin polymer plate plates for a final hematocrit of 2.5%. Compounds were plated with 23 nL of each compound using a pin tool . Plates were incubated for 72 h at 37 °C to allow for 1.5 generations of intraerythrocytic parasite growth .
Cells were lysed with SYBR Green I in lysis buffer and incubated overnight (approximately 18 h) in the dark [29, 30]. Fluorescence was measured using an EnVision plate reader at 485/14 nm excitation and 535/25 nm emission. The dose response data was fitted to the 4-parameter Hill equation using a grid-based algorithm developed in house . For compounds with valid curve fits, an IC50 value was obtained. In addition, fits were classified into curve classes, a heuristic scheme that allows one to easily categorize fits as good quality (i.e., well defined upper and lower asymptotes, greater than 80% efficacy), inconclusive (missing an asymptote or displaying activity at the highest dose) or inactive (no dose response). Statistical analysis to identify statistically significant differences in the responses at the maximum dose (MaxR) was done using ANOVA followed by Tukey post hoc test. Determination of class enrichment was assessed using Fishers test with Benjamini–Hochberg correction for multiple comparisons.
Atg18 alignment with homologues
Genes for proteins in the WIPI family were aligned using MegAlign version 12.2.0 using protein sequences from Homo sapiens WIPI1 (AAH39867.1), WIPI2a-d (NP_056425.1, NP_057087.2, NP_001028690.1, NP_00102891.1) and P. falciparum Atg18 (XP_001347411.1). Analysis was performed using the Clustal V method, although relevant findings did not differ when Clustal W or Jotun Hein methods were used.
Using the sequence from PlasmoDB to the currently available crystal structure of Hsv2 in K. lactis, a model of PfAtg18 was created with UCSF Chimera version 1.10.1. The sequence for PI3P binding, FRRG, was highlighted in addition to the SNP of interest, T38I, and Atg16 binding site. The phosphorylation site score of T38 was determined using NetPhos 3.1 server to predict the likelihood of the threonine being a phosphorylation site for one of 17 kinases .
PIP binding assay
The ability of Dd2R539T to bind to various lipids, including PI3P, was assessed using a PIP Strip (Echelon). A GFP-trap_M (Chromotek) was used to isolate GFP-tagged Atg18 from Dd2R539T parasites with Atg18-GFP, according to the manufacturer’s instructions. The PIP Strip membrane was then prepared according to the manufacturer’s instructions using the isolated Atg18-GFP, then stained with anti-GFP antibody (Abcam, ab290) at 1:2000 for 1 h at room temperature. Anti-rabbit HRP secondary antibody was used at 1:2000 for 1 h at room temperature followed by incubation with K-TMBP (Echelon) for development for 10 min. The membrane was then imaged to detect binding. The resulting image was processed using ImageJ. Briefly, the image was converted to 32-bit grayscale, background was subtracted with a radius set to 25 pixels, and the image was inverted. A circular selection was created and the integrated density of the center of each dot on the membrane was measured. The value of the blank on the membrane was subtracted out and anything below the blank was recorded as zero.
Confocal slide production
Dd2R539T lines with GFP-tagged Atg18 were used for confocal imaging studies. Cultures were washed three times with PBS, then were fixed with 0.0075% glutaraldehyde, 2% paraformaldehyde in PBS at room temperature for 30 min, washed with PBS three times, and permeabilized with 0.1% Triton X-100 (Sigma) in PBS for 10 min at room temperature. Parasites were washed with PBS three times and then stained with anti-GFP rabbit polyclonal (Abcam ab290) primary antibody overnight at 4 °C, washed with PBS three times, then stained with goat polyclonal to rabbit secondary antibody (Abcam ab150079, AF647) for 2 h at room temperature. Cells were washed three times with PBS and slides were made using Prolong Gold mounting solution with 4′,6-diamidino-2-phenylindole (DAPI) (ThermoFisher). Coverslips were placed on top of each slide, slides were covered with tin foil, and left to dry overnight at room temperature. Slides were stored at 4 °C before imaging. Images were collected on a Leica SP5 inverted confocal microscope with a 63 × oil immersion objective NA 1.4 (Leica Microsystems, Buffalo Grove, IL). Post-processing and image analysis were performed using Huygens (SVI imaging, Nederland) and Imaris software (Bitplane Inc., South Windsor, CT).
Parasite culture Dd2R539T with GFP-labelled Atg18 were synchronized to trophozoite and schizont stages using a Percoll gradient separation and then exposed to 700 nM DHA for 10 min, 1 h, or left unexposed. Cultures were pelleted and washed with PBS then pelleted and resuspended in 4% paraformaldehyde and 0.1% glutaraldehyde. Samples were stored at 4 °C before being shipped to Rocky Mountain Laboratories for processing and imaging.
For transmission electron microscopy of immune-labelled sections, samples were prepared essentially as previously described  with several adaptations as follows. For embedment, LR White acrylic resin (Electron Microscopy Sciences, Hatfield, PA) was used rather than Araldite epoxy resin. Silver sections were mounted on nickel grids, then dried overnight at 45 °C. The treatments with formic acid and sodium meta-periodate were omitted. Before labeling, grids were immersed in 100% ethanol for 10 s, then quickly rinsed with deionized water, and placed into blocking buffer. Primary and secondary antibodies were used at 1:50 dilution in blocking buffer. A rabbit anti-eGFP polyclonal primary antibody (Thermofisher OSE00003G) was used overnight at 4 °C following microwave irradiation.
Nutrient deprivation assays
The ability of parasite lines Dd2, Dd2R539T, and Dd2R539T/T38I to survive under nutrient deprivation was assessed using growth studies. Parasitaemia of the three lines were assessed by FACS and then diluted down to approximately 0.5% parasitaemia. Lines were then plated in flat-bottomed plates in various nutrient restriction conditions at 5% Hct and incubated in a gassed incubator. Complete media was titrated with a basal salt solution (Additional file 1) to create 100%, 85%, 75%, 65%, 50%, and 20% complete media with and without glucose. Parasitaemia was assessed at 48 h by FACS.
Ring-stage survival assays to dihydroartemisinin or piperaquine pulse
Ring-stage survival assay (RSA) and piperaquine survival assay (PSA) were performed based on the methodologies described by Kite et al. and Duru et al. respectively [2, 34]. Briefly, Dd2, Dd2R539T, and Dd2R539T/T38I parasites were cultivated in 20 mL cultures at 5% hematocrit cultures and synchronized every 46–48 h by a 10 min incubation with 5% sorbitol at 37 °C. After at least two sorbitol synchronizations, 0–3 h ring-stage 1 mL cultures at 2% hematocrit and 1% initial parasitaemia were incubated with either DMSO (Sigma, vehicle), 700 nM DHA (Sigma) in DMSO, 3% methanol (Sigma, vehicle), or 200 nM piperaquine in 3% methanol (kindly provided by Erin Coonahan). DMSO and DHA cultures were washed with 10 mL complete RPMI twice (2500 rpm, 3 min, 9/10 acceleration, 4/10 deacceleration) after 6 h incubation; methanol and piperaquine cultures were washed after 48 h incubation. Thin blood films were prepared from all cultures after a total of 72 h from beginning of each drug or vehicle incubation. Percentage parasite survival was calculated dividing the parasitaemia from a drug-treated culture by the parasitaemia of the culture incubated with the respective drug vehicle, then multiplying the result by 100. To determine parasitaemia, an estimated 10,000 red blood cells from methanol-fixed and 20% Giemsa-stained blood films were counted with oil immersion in a light microscope.
A SNP in atg18 is associated with ART-resistance phenotypes
SNPs in autophagy-related genes from the Tracking Resistance to Artemisinin Collaboration (TRAC) study
Amino acid change
Kelch protein K13
Parasite clearance half-life
Autophagy-related protein 18 (Atg18), putative
Parasite clearance half-life
Conserved Plasmodium protein, unknown function
Parasite clearance half-life
Autophagy-related protein 18 (Atg18), putative
Conserved Plasmodium protein, unknown function
Parasite clearance half-life
Conserved Plasmodium protein, unknown function
ThiF family protein, putative
ThiF family protein, putative
PfAtg18 is similar in structure and behaviour to yeast and mammalian homologues
Homologous Atg18 proteins bind PI3P as one of the first steps in initiating an autophagy cascade; therefore, in determining possible functions of PfAtg18, it is important to assess if this protein also has these binding capabilities. An Echelon phosphatidylinositol phosphate (PIP) Strip membrane was used to assess binding of PfAtg18 with eight phosphoinositides and seven other lipids. The protein was detected in 10 of the 15 spots, including seven of the eight phosphoinositides (Fig. 1d). Similarly to other Atg18 proteins, PfAtg18 bound PI5P and PI3P at high levels .
An atg18 SNP is associated with faster growth in nutrient-limited conditions
An atg18 mutation associates with ex vivo resistance even in the absence of k13 mutations
An atg18 mutant parasite line shows differential susceptibility to over 200 compounds
The proposed mechanisms of ART resistance suggest that an autophagy-like mechanism could play a role in modulating resistance in P. falciparum. Several SNPs in autophagy-related proteins were associated with drug resistance phenotypes; however, this study chose to focus on atg18 based on the level of significance of its SNP in the GWAS. A second report associated a similar SNP in atg18 with ART derivative and piperaquine resistance phenotypes in an independent cohort in the border region of China and Myanmar, confirming the potential relevance of this SNP in conferring anti-malarial resistance .
PfAtg18 is likely a homologue of WIPI/Atg18 with a similar binding capacity and behaviour pattern following stress. The ability of PfAtg18 to bind to PI3P is significant. Homologous Atg18/WIPI proteins also bind PI3P and, therefore, the conservation of binding suggests potential preservation of part of the autophagy pathway. Further, PI3P been implicated in drug resistance in P. falciparum from work by Mbengue and colleagues , who suggested there is a relationship between PI3P and artemisinin resistance. While autophagy-related proteins are involved in apicoplast inheritance , the results of this study indicate that autophagy-related proteins may have alternative functions outside of the apicoplast, as the only apparently essential function of the apicoplast is to provide isopentyl pyrophosphate (IPP) .
If PfAtg18 is involved in an autophagy-like pathway in Plasmodium, this SNP may modulate its activity. A number of mutations in homologous WIPI proteins have been described in various cancers . The diversity of mutations found in WIPI proteins in cancer suggests that these mutations provide a survival advantage. For a mutation to be advantageous in cancer, it must facilitate cancer cell growth, for example through nutrient acquisition, dealing with cellular stress or several other pro-survival mechanisms [45, 46]. Thus, the plasmodial atg18 T38I mutation may similarly offer a survival benefit. Given the ability of the Dd2R539T/T38I mutant to survive better in nutrient-limited settings than either its parent Dd2R539T line or the original Dd2 line, the T38I SNP may therefore represent a fitness advantage by growing more effectively in the nutrient-limited setting of parasite infection and acute malaria. This alone or in addition to a drug-resistance phenotype, could select for the T38I SNP. Despite allowing for increased survival of early ring-stage parasites, k13 polymorphisms have been shown to cause growth deficits in vitro . The T38I mutation could, therefore, compensate for the loss of fitness caused by k13 or other unknown polymorphisms in the parasite population. DHA-resistant isolates have been shown to grow more rapidly in stressful conditions, as evidenced by a decreased growth rate for sensitive parasites at high parasitaemia . At higher parasitaemia, growth media, which is normally supra-physiologic, may become rapidly depleted, thus creating a nutrient-limited media. Therefore, these results align with the finding that efficient growth in nutrient deprivation may be driving selection. Further, these stressful, oxidative stress-inducing conditions would typically slow parasite growth . One possible explanation is that if a parasite is better equipped to handle damage by oxidative stress, perhaps by employing autophagy-like mechanisms, that parasite would likely outcompete other parasites in a population.
Growth of Plasmodium parasites appears to be susceptible to changes in the immediate host environment. Dietary restrictions in mice have been shown to be protective against cerebral malaria by Plasmodium berghei [50, 51]. Refeeding of patients with P. falciparum in the hospital setting has been associated with an increase in parasitaemia . In culture supplemented with sera from calorie-restricted mice as compared to fed mice, parasites formed fewer merozoites during schizogony, thus reducing overall parasitaemia . Thus, parasitaemia is intricately tied to host nutrition status. The serine/threonine kinase KIN, which is putatively homologous to AMP-activated kinase (AMPK), has been implicated in nutrient sensing in Plasmodium in the absence of TOR proteins . AMPK is another regulator of autophagy along with mTOR . An autophagy-like pathway may therefore play a role in nutrient sensing in Plasmodium and may modulate parasite virulence based on the ability of the parasite to obtain limited nutrients.
Many anti-malarial drugs target metabolic pathways . An improvement in a parasite’s ability to survive low nutrient conditions not only benefits the parasite in terms of its growth rate, but also enables a parasite to evade drugs that may target various metabolic pathways. While a drug may block one metabolic pathway, another pathway that was previously redundant may become essential. The ability to grow more efficiently in nutrient-limited media suggests changes in the metabolism of parasites carrying this mutation, which could allow the parasite to survive drug treatment. Again, the ability to survive a drug pulse based on the parasite’s ability to employ alternate metabolic pathways may appear inconsequential in a drug screen in complete media, where nutrient limitation is not a factor. Experiments examining parasite responses to drugs in a variety of media are ongoing.
The drug screen revealed that the mutant line is differentially susceptible to > 200 compounds, including a number of compounds that would likely modulate autophagy, including vincristine , sapitinib [57, 58], and gefitinib [59, 60], whereas others modulate ROS, including elesclomol [41, 61], olmesartan , and paeoniflorin . Despite these effects on the parasite’s susceptibility to other compounds, the atg18 T38I SNP did not change IC50 values for the tested anti-malarial compounds. This was surprising given the results of the ex vivo IC50 experiments and GWAS. It is possible that, like k13 polymorphisms, the contribution of the atg18 polymorphism to resistance cannot be detected using a 72-h assay [64, 65]. Thus, studies are underway to determine if this SNP modulates several other phenotypes including those measured by dormancy assays. Additionally, the effects of the SNP may be important only in the context of other background mutations, which are absent in the Dd2R539T line.
It is possible that the T38I SNP changes the function of PfAtg18 to promote more efficient nutrient acquisition through an autophagy-like pathway. This could allow for either increased fitness leading to expansion within the parasite population or the ability of a parasite to survive drug treatment by utilizing its autophagy-like pathway to survive in the absence of a blocked metabolic pathway. Further studies, including a more detailed characterization of the function of PfAtg18, are necessary to determine if this is indeed the case.
KFB, RMF, DR, and AKS conceived of the project. KFB, RE, PL, CK-T, CM, and JMS performed the experiments. KFB, SG and DWD performed imaging. RA and RG performed statistical and bioinformatics analyses. KFB wrote the manuscript with significant edits by DR, CT, and AKS. All authors read and approved the final manuscript.
We are grateful to study participants and clinical staff that made this study possible. We thank David Ito for advice in the production of mutant parasites and Thomas Wellems and Chanaki Amaratunga for advice and support of this work. We thank David Fidock and Jose Juan Lopez-Rubio for use of materials. Whole genome sequencing was done by the Wellcome Sanger Institute (WSI) as part of the MalariaGEN Plasmodium falciparum Community Project, and coordinated by the MalariaGEN Resource Centre. We thank the staff of the WSI Sample Logistics, Sequencing, and Informatics facilities for their contribution.
The authors declare that they have no competing interests.
Availability of data and materials
The WGS dataset analyzed during this work is available on the MalariaGEN repository (https://www.malariagen.net/data). The drug screen data generated and analyzed during this work is available from the corresponding author on reasonable request. All other data generated or analyzed during this study are included in this published article and its additional files.
Consent for publication
Ethics approval and consent to participate
Written informed consent to participate was obtained from all subjects. Patient samples and data approval was obtained from the Cambodian National Ethics Committee for Health Research and the NIAID Institutional Review Board.
This study was funded by the Intramural Research Programs of the National Institute of Allergy and Infectious Diseases and the National Center for Advancing Translational Sciences, the Wellcome Trust, Bill & Melinda Gates Foundation, Medical Research Council, and the NIH OxCam Scholars Program.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 2.Duru V, Khim N, Leang R, Kim S, Domergue A, Kloeung N, et al. Plasmodium falciparum dihydroartemisinin-piperaquine failures in Cambodia are associated with mutant K13 parasites presenting high survival rates in novel piperaquine in vitro assays: retrospective and prospective investigations. BMC Med. 2015;13:305.CrossRefGoogle Scholar
- 14.Klionsky DJ, Abdelmohsen K, Abe A, Abedin MJ, Abeliovich H, Acevedo Arozena A, et al. Guidelines for the use and interpretation of assays for monitoring autophagy 3rd ed. Autophagy. 2016;12:1–222.Google Scholar
- 22.Bacon DJ, Latour C, Lucas C, Colina O, Ringwald P, Picot S. Comparison of a SYBR green I-based assay with a histidine-rich protein II enzyme-linked immunosorbent assay for in vitro antimalarial drug efficacy testing and application to clinical isolates. Antimicrob Agents Chemother. 2007;51:1172–8.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.