Current understanding of the molecular basis of chloroquine-resistance in Plasmodium falciparumH Jiang, DA Joy, T Furuya, X-z Su
Laboratory of Malaria and Vector Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892-8132, USA
Correspondence Address: Source of Support: None, Conflict of Interest: None PMID: 17102545
Source of Support: None, Conflict of Interest: None
Chloroquine (CQ) is the most successful antimalarial drug ever discovered. Unfortunately, parasites resistant to the drug eventually emerged after its large scale use and are now widespread. Although great progress in our understanding of the mechanisms of CQ action and CQ resistance (CQR) has been achieved over the past two decades, including the identification of the molecules responsible for CQR (e.g., Plasmodium falciparum chloroquine resistant transporter, PfCRT) many questions remain unanswered. Here we highlight recent advances in our understanding of the genetics and molecular mechanisms of CQR, with particular emphasis on the role of genes such as pfcrt and pfmdr1 in the resistance to CQ and other drugs. New drug development and applications will undoubtedly benefit from a better understanding of CQR, eventually leading to more effective malaria control measures.
Keywords: Chloroquine, chloroquine resistance, malaria, Plasmodium falciparum, pfcrt, pfmdr1
Plasmodium falciparum malaria is a deadly infectious disease in tropical and subtropical developing countries, killing an estimated 1-2 million people annually. Because no effective vaccine is yet available, malaria prevention and control relies primarily on antimalarial drugs and antimosquito measures such as insecticides and bed nets. Unfortunately, P. falciparum has developed resistance to nearly all classes of antimalarial drugs available, leading to a resurgence of malaria mortality and morbidity in most endemic areas., Because chloroquine (CQ) has been the major antimalarial drug in the past few decades, resistance to this drug has received great attention and has been the focus of much research activity. This brief review highlights recent advances in our understanding of the mechanisms of CQ resistance (CQR). Readers interested in broader coverage of the subject are encouraged to read other excellent reviews.,,,,
CQ and CQR
CQ, a synthetic 4-aminoquinoline, is a lysosomotropic weak base compound. Until recently, it was the first line antimalarial drug for treating malaria due to its low cost, low incidence of side-effects and high efficacy. The P. falciparum parasite obtains some of its nutrients by digesting hemoglobin, a process that releases heme (ferriprotoporphyrin IX) into the parasite's acidic digestive vacuole (DV). To avoid heme toxicity, parasites polymerise it into nontoxic hemozoin. It is generally believed that CQ inhibits heme polymerisation, resulting in the accumulation of toxic heme and/or a CQ heme complex that eventually kills the parasite.,,
Regrettably, the widespread use of CQ has led to the emergence of CQR parasites. CQR was first reported in Southeast Asia and South America in the late 1950s, Papua New Guinea (PNG) in the 1960s, East Africa in the late 1970s and throughout Africa in the1980s-1990s., A recent molecular study has shown multiple CQR founder mutations and a large scale CQR selective sweep from Southeast Asia to Africa as well as other sweeps across the Amazon in South America and in PNG.
Despite multiple independent origins of CQR worldwide, CQR parasites share some common phenotypes: (1) increased IC 50 , the drug inhibition concentration at which 50% of the parasites are killed, measured by in vitro drug assays (an IC 50 of 100 nM in vitro has been used as a CQR threshold because it corresponds to ~40% to 50% of in vivo clinical CQ treatment failures); (2) chemosensitisation, for example, the calcium channel blocker verapamil (VP), can reverse CQR back to CQS; (3) reduced CQ accumulation in the acidic DV;, (4) reduced pH in DV and (5) universally shared point mutations in a putative transporter gene termed P. falciparum chloroquine resistance transporter ( pfcrt ), which is the major determinant of CQR.
From 16 progeny of a genetic cross between the CQS parasite HB3 (Honduras) and the CQR parasite Dd2 (Indochina), Wellems et al, first defined a ~400 kb DNA segment on chromosome 7 that was linked to CQR. This CQR locus was later narrowed to a ~36 kb region using high density microsatellite markers. Further analysis of this region led to the discovery of pfcrt , a ~3.1 kb gene with 13 exons encoding a transmembrane protein PfCRT (424 amino acid [ aa ], 48.6 kDa); and multiple nucleotide substitutions in the gene were found to be associated with CQR in parasites from Asia, Africa and South America. PfCRT is predicted to have ten transmembrane domains and is located on the membrane of the parasite DV.,
Mutations in pfcrt and CQR
Twenty point mutations have been found in the pfcrt gene to date from CQR field isolates or laboratory clones through drug selection. Association studies have shown that substitution of threonine (T) for lysine (K) at position 76 (K76T) is the hallmark of CQR parasites worldwide.,, Although a few exceptions (e.g ., parasites with 76T allele of pfcrt were cleared by CQ treatment or parasites survived the CQ treatment while carrying 76K allele) have been reported recently,,,,, they could be due to limited drug tests in the field, mixed infections in vivo and/or new mutations in vitro such as S163R, which may counteract K76T substitution.,,, Additionally, A220S and other substitutions are highly prevalent among CQR parasites.,,
Two major haplotypes defined by specific mutations at aa positions 72-76, CVIET from Asian and African isolates and SVMNT from South America and PNG isolates, represent the two major CQR selective sweeps in Southeast Asia and South America, respectively. However, the CVIET haplotype has also been found in South America, and the SVMNT haplotype found in Southeast Asia appeared to be associated with a decreased efficacy of amodiaquine (AQ)., Of interest, the predominant haplotype found in a study in India was SVMNT, although the Southeast Asia haplotype CVIET was also present. The 76T and other mutations in pfcrt have been employed as molecular markers for predicting CQR in field surveys,,, although CQR is not always linked to mutant pfcrt alleles in field isolates due to host immunity and other factors.
Experimental evidence showing that mutations in pfcrt lead to the CQR phenotype was provided by in vitro drug selection and allelic exchange studies. When selected by lethal doses of CQ, a CQS parasite (106/1) carrying six mutations in pfcrt (except 76T) commonly found in Southeast Asian CQR parasites became CQR. Although 106/1 was originally isolated from Sudan, it has a genome background closely related to parasites of South Asian origin, particularly FCB., Sequence analysis of the CQ selected parasites revealed two new mutations at codon position 76 (76I or 76N). Additionally, replacement of the wild type pfcrt allele in a CQS progeny GC03 of the HB3xDd2 cross with up to eight pfcrt alleles from CQR parasites originating from Asia, Africa and South America consistently produced the CQR phenotype and reduced VP reversibility that is always associated with CQR and substitution at 76 aa position of PfCRT.
Is PfCRT a transporter?
The original functional assignment of PfCRT as a putative transporter was based on its predicted ten transmembrane domains (TMD) and its localization to the DV membrane. Recent bioinformatics analyses suggested that PfCRT belongs to a drug/metabolite transporter superfamily., By comparison with other proteins within this family, the N- and C-termini of PfCRT were predicted to be located on the parasite's cytoplasmic side of the DV membrane., PfCRT was also predicted to dimerise within the DV membrane, with TMDs 1, 2, 3, 6, 7 and 8, functioning mainly in substrate discrimination and recognition; TMDs 4 and 9 in substrate binding and translocation; and TMDs 5 and 10 in the formation and/or stabilization of homodimeric structure.
Some experimental evidence supporting PfCRT transporter activity has come from expression of the gene in heterologous systems. Expression of PfCRT in yeast resulted in an increased proton (H+) gradient across the vesicular membrane with reduced pH inside, which was thought to be due to either the effect of PfCRT on chloride (Cl-) transport or ATPase activity. PfCRT was also found to bind specifically to CQ at physiological concentrations, although no significant difference was observed in binding affinity of CQ to PfCRT from wild type (CQS) and mutant (CQR) parasites. In contrast to observations in yeast, no differences in Cl- conductance between control and PfCRT expressing oocytes were observed. Xenopus laevis oocytes expressing PfCRT showed reduced intracellular membrane potential and an alkaline pH relative to controls, which were attributed to the activation of a nonselective cation transporter (Gcat) and an endogenous Na+-H+ exchanger (NHE), consistent with results from a previous study. Because these two transmembrane pathways are independent, they do not support the role of PfCRT as a direct CQ transporter. From these studies, it was proposed that PfCRT expression resulted in activation of the endogenous transporter systems: H+-ATPase in yeast and Gcat and NHE in oocytes. However, Naude et al recently expressed PfCRT in the slime mold Dictyostelium discoideum and found small reductions in pH and significantly reduced VP reversible intravesicular CQ accumulation in K76T CQR mutants relative to the control and wild type PfCRT cells. This reduced CQ accumulation was most likely due to a PfCRT mediated energy dependent efflux mechanism rather than to the small intravesicular changes in pH and CQ uptake.
Sanchez et al, measured CQ accumulation inside and outside of the DV and CQ uptake kinetics in P. falciparum parasites. Comparing several CQS and CQR parasites, they found that only in CQR parasites was there a trans-stimulated CQ accumulation inside the DV, a phenomenon that is energy-dependent and not due to simple passive diffusion through channels or pores., They also associated this CQ stimulation phenomenon in CQR parasites with pfcrt CQR alleles and suggested that pfcrt is directly or indirectly involved in a transporter mediated CQ efflux system in CQR parasites.,
Although PfCRT gene expression levels have yet to be carefully compared among CQS and CQR field isolates, a study of 40 Cambodian isolates did not find any correlation between the expression level of pfcrt , evaluated by real-time RT-PCR and CQR in vitro . However, in P. falciparum transfection studies, Sidhu et al showed that reduced PfCRT expression caused a reduction in CQ IC 50 of all pfcrt -modified clones when compared with the non-transformed CQR controls. Although no successful pfcrt gene knockouts or RNA interference (RNAi) experiments have yet been reported in P. falciparum , a recent in vitro study showed that pfcrt 'knock down' clones (originating from CQR 7G8) with a ~30% to 40% decrease in PfCRT protein expression led to an increase in pH inside the DV relative to the CQR parent and a ~40% reduction in CQ IC 50 . These studies all indicate that PfCRT plays a key role in transporting CQ and/or regulating other endogenous transporters. However, more studies are necessary to conclusively elucidate the function of PfCRT and the roles of mutations and gene expression in conferring CQR to P. falciparum parasites.
Is PfCRT a multiple drug transporter?
PfCRT appears to play some role in parasite response to other antimalarial drugs, including quinine (QN), quinidine (QD), mefloquine (MQ), halofantrine (HF), artemisinin (ART) and an antiviral agent amantadine (AM). Substitutions at position 76 (K76I or K76N) not only confer greater resistance to CQ and QD, but also render the parasite more sensitive to QN, HF, MQ and ART. Replacement of the wild type pfcrt allele in the CQS GC03 parasite with mutant alleles from CQR parasites Dd2, 106/1(with K76I) and 7G8 converted the parasite into a CQR parasite and affected its susceptibility to QN, HF, MQ and ART. Using both drug selection and allelic exchange methods, Johnson et al recently found that PfCRT could modulate parasite responses to multiple drugs. For example, relative to CQS parasites (3D7 and GC03), parasites with CQR PfCRT haplotypes were more sensitive to AM, HF and MQ. Two AM and HF selected strains (K1AM and K1HF) carrying a new PfCRT mutation, S163R, became sensitive to CQ but more resistant to MQ and QN (K1HF). When the K76T mutation was replaced with the wild type allele in CQR strains Dd2 and 7G8, the parasites not only lost VP reversible CQR but also became more sensitive to QD and AQ, although the response to AM and MQ was strain- dependent. It is clear that pfcrt point mutations not only determine a parasite's response to CQ, but also significantly affect its susceptibility to other drugs. However, how PfCRT might facilitate the transport of multiple drugs, as implied from CQR studies and bioinformatic predictions,, requires more studies.
P. falciparum multiple drug resistance gene 1 (pfmdr1)
P. falciparum CQR shares some common features with multidrug resistance (MDR) in mammalian tumor cells, e.g., reduced drug accumulation in DV and VP reversibility. Pfmdr1 , a parasite homologue of the mammalian MDR gene, was characterized and found to be associated with CQR., The protein encoded by pfmdr1 is PfPgh1 (P glycoprotein homologue, 162 kDa), a member of the ATP-binding cassette transporter superfamily that is located on the parasite DV membrane.
Varying degrees of association between PfPgh1 mutations (particularly N86Y) and reduced susceptibility to CQ and other drugs have been found both in vitro and in vivo .,,,, Other studies, however, detected no or weak association between pfmdr1 mutations and CQR parasites worldwide.,,,,,, Two studies using allelic exchange methods showed that pfmdr1 triple mutations (i.e., S1034C/N1042D/D1246Y, CDY) contributed significantly to parasite susceptibility to multiple drugs, such as QN, MQ, HF and ART,, which is in agreement with other studies.,, However, the effect of this triple mutation on CQR may be parasite strain-dependent:, (1) introduction of CDY to CQS parasites (originating from Asia, Africa and PNG) did not significantly change their CQ susceptibility;, (2) replacement of the triple mutation CDY with the wild type pfmdr1 allele significantly reduced the CQ IC 50 in South American CQR parasites but not in a CQS PNG parasite carrying the three mutations.
Gene amplification (increased copy number) and expression of pfmdr1 have also been correlated with parasite responses to CQ and other drugs such as MQ. Foote et al reported increased pfmdr1 transcript levels and amplified pfmdr1 in some CQR parasites. In vitro selected MQ resistant lines showed amplified and overexpressed pfmdr1 and an inverse relationship between CQ IC 50 and pfmdr1 gene copy number. However, other reports showed no difference in gene expression levels or copy numbers between CQS and CQR parasites.,, Gene copy number should have no effect if the corresponding gene is not expressed. Studies measuring both copy number and RNA or protein expression at the same time would clarify the contribution of gene copy number to drug responses. Unfortunately, most studies did not include follow up measurements of RNA or protein expression levels.
It is not clear how pfmdr1 affects parasite response to CQ. Functions similar to other P glycoproteins have been speculated and tested. In P. falciparum parasites, PfPgh1 was found to bind ATP and was phosphorylated extensively on serine and threonine; but neither CQ nor compounds that modulate CQR had any effect on PfPgh1 phosphorylation. The nucleotide binding domain of PfPgh1 was also found to be functional and involved in a nucleotide regulated transport across the DV membrane. Mutations S1034C and N1042D have been predicted to reside on the hydrophilic side of PfPgh1 transmembrane domain 11,, which was postulated to participate in CQ recognition and transport, thereby indirectly affecting intracellular drug concentration. Although CQ efflux is phenotypically similar to the efflux of anticancer drugs from mdr expressing mammalian cells, whether pfmdr1 plays a role in the efflux mechanism in P. falciparum is still an open question. More recent studies have suggested that pfmdr1 (e.g, copy number) is the major determinant for MQ and QN resistance, which was often found to be inversely related to CQ sensitivity.,, Thus, a 'modulator' role of pfmdr1 in CQR was proposed.,
pfcrt, pfmdr1 and other genes
Although pfcrt and pfmdr1 are located on chromosomes 7 and 5, respectively, several recent studies have found a strong association or co-selection between them.,,,,,. In terms of specific mutations, significant associations have been found between PfPgh1 N86Y and PfCRT K76T in field isolates.,,,,, Similar to mutations in pfcrt , mutations in pfmdr1 correlated with responses to CQ also affect parasite responses to MQ, HF and ART.,, It is clear that pfcrt has a causal effect on CQR, while pfmdr1 appears to act as a secondary modulator.,,
Genes other than pfcrt and pfmdr1 are likely to contribute to CQR. For instance, field isolates with the same pfcrt and/or pfmdr1 genotypes often displayed a wide range of IC 50 values to CQ., CQR phenotypes and genotypes were also found to vary among different geographical locations. Indeed, several putative transporter genes have been significantly associated with decreased sensitivity to CQ, although this finding was not confirmed by a recent study of parasites from Thailand. These discrepancies could be due to parasite population structures or to limited replicates in drug tests.
Various hypotheses have been proposed to explain how P. falciparum parasites become CQR. The majority of studies have focused on regulation of DV physiological changes (pH, ion exchange and CQ concentration) and CQ efflux. DV has a pH of 5.2 to 5.8, much lower than the surrounding cytoplasm (pH ~7.3). This lowered pH is expected to increase the accumulation of protonated CQ 2+ within the DV, according to weak-base theory. However, observations of lower pH coupled with decreased CQ accumulation in the DV of CQR parasites contradict this theory.,,,, Subsequently, the suitability of acridine orange fluorescence assay to measure pH was questioned. In a recent study, Bennett et al confirmed their initial observation of a more acidic pH (5.2) in CQR parasites relative to CQS parasites (pH = 5.7-5.8) using an improved method. A more acidic pH inside the DV of CQR parasites is believed to directly or indirectly increase the aggregation of heme and reduce its capacity to bind CQ, resulting in fewer CQ-heme complexes inside the DV and reduced toxicity to the parasite. They also found that VP reversible CQR was directly linked to greater DV acidity and mutations in pfcrt alleles from both Dd2 (Asian/African) and 7G8 (South American/Oceanic) type.
Other studies have suggested that PfCRT may affect Cl- transport across the DV,, thereby influencing the movement of other ions (including protons) and eventually affecting heme crystallisation, CQ heme-binding kinetics and/or CQ partitioning. Sanchez et al originally suggested that a plasmodial NHE in CQR parasites may be responsible for reduced CQ import compared with CQS lines. However, steady-state CQ accumulation in the DV was later demonstrated to be due solely to the binding affinity of CQ to heme in the DV rather than to regulation by NHE. Comparing CQS and CQR parasites, it appears that the number of heme-binding sites does not differ,, but the affinity of saturable CQ heme-binding at equilibrium was greatly reduced in CQR parasites.,, Technical challenges due to the sensitivity and dynamics of pH measurements in small cell compartments such as the DV make it difficult to reconcile all the contradictory results from different laboratories.
Decreased CQ accumulation could also be due to changes in CQ influx or efflux across the DV membrane. Currently, many studies support the CQ efflux hypothesis, in which increased CQ efflux leads to decreased CQ accumulation inside the DV of CQR parasites.,,,,, Krogstad et al originally proposed that rapid CQ efflux caused the lower CQ accumulation in the DV of CQR parasites. Later, CQ efflux was proposed to be an energy-dependent mechanism based on observations that increased ATP production after glucose feeding was temporally associated with decreased CQ accumulation inside the DV in CQR but not CQS parasites., Recently, others have argued that CQ efflux is not an energy-dependent process but due to a process called "charge-leak efflux". According to this model, protonated CQ 2+ in the DV 'leaks' out along its electrochemical gradient through PfCRT, a possible anion channel. Positively charged aa (e.g. wild type pfcrt K76 in CQS parasites or mutant R163 in a CQS parasite carrying K76T) might repel CQ 2+ from moving through the PfCRT channel, leading to higher CQ 2+ accumulation at its site of action. In contrast, PfCRT in CQR parasites carrying pfcrt mutations that lack the positively charged residues (e.g . K76 to T or I or N) may allow CQ to leak outside of the DV, resulting in reduced accumulations of CQ 2+. This hypothesis could also explain VP reversibility in CQR parasites, because VP may reintroduce a positive charge to a putative pore of the PfCRT protein. An assumption of this model is that PfCRT can transport drugs directly, as predicted by recent bioinformatics analyses., However, evidence for this function is not yet conclusive.
Although extensive work has been done in the past two decades, particularly following the discovery of pfcrt, we are far from a complete understanding of the molecular mechanisms underlying CQ action and CQR. It is clear, however, that mutations in the pfcrt can convert clinical parasites from CQS to CQR, while pfmdr1 may 'modulate' parasite responses to CQ and interact with other genes in response to multiple drugs.
We thank Drs. Thomas E Wellems and Rick M Fairhurst for valuable suggestions and NIAID intramural editor Brenda Rae Marshall for assistance.