Impaired fitness of drug-resistant malaria parasites: evidence and implication on drug-deployment policies

Review

Impaired fitness of drugresistant malaria parasites: evidence and implication on drug-deployment policies

ro

Author for correspondence Biochemistry Department, Faculty of Medicine, Sultan Qaboos University, Alkhod, PO Box 35, Muscat, Oman Tel.: TEL Fax: FAX [email protected]

A

ut

ho

†

Malaria, a leading parasitic disease, inflicts an enormous toll on human lives, and is caused by protozoal parasites belonging to the genus Plasmodium. Antimalarial drugs, targeting essential biochemical processes in the parasite, are the primary resources for management and control. However, the parasite has established mutations substantially reducing the efficacy of these drugs. First-line therapy is faced the with consistent evolution of drug-resistant genotypes carrying these mutations. However, drug-resistant genotypes are likely to be less fit than the wild-type, suggesting that they might disappear by reducing the volume of drug pressure. A substantial body of epidemiological evidence confirmed that the frequency of resistant genotypes wanes when active drug selection declines. Drug selection on the parasite genome that removes genetic variation in the vicinity of drug resistant genes (hitch-hiking) is common among resistant parasites in the field. This can further disadvantage drug-resistant strains and limit their variability in the face of a mounting immune response. Attempts to provide unequivocal evidence for the fitness cost of drug resistance have monitored the outcomes of laboratory competition experiments of deliberate mixtures of sensitive and resistant strains, in the absence of drug pressure, using isogenic clones produced either by drug selection or gene manipulation. Some of these experiments provided inconclusive results, but they all suggested reduced fitness of drug-resistant clones in the absence of drug pressure. In addition, biochemical ana­lysis provided clearer information demonstrating mutation of some antimalarial-targeted enzymes lower their activity compared with the wild-type enzyme. Here we review current evidences for disadvantage of drug-resistance mutations and discuss some strategies of drug deployment to maximize the cost of resistance and limit its spread.

rP

Hamza A Babikera†, Ian M Hastings and Göte Swedberg

of

Expert Rev. Anti Infect. Ther. 7(5), xxx–xxx (2009)

Keywords : chloroquine • drug-resistant genotype • fitness cost • Plasmodium chabaudi • Plasmodium falciparum • pyrimethamine • sulfadoxine

Introduction: fitness of drug-resistant mutant pathogens

Antipathogenic drugs target essential metabolic and structural proteins or processes that are essential for growth and multiplication. Therefore, genetic changes (mutations) in genes whose products are associated with these proteins/processes can encode resistance but often have deleterious effects [1] . Consequently, the development of drug resistance may incur a fitness cost in the absence of drug selection. This has been demonstrated in many pathogens, including parasites [2] , bacteria [3] and viruses [4] . For example, in the HIV-1 virus, it has been shown that mutations in a protease gene causing www.expert-reviews.com

10.1586/ERI.09.29

resistance to a protease inhibitor, can markedly reduce the virus-replication capacity [4] . Drugresistant mutant forms of HIV, have been found to infect patients approximately 20% less than might be expected [5] . Fitness costs may manifest as reduced withinhost growth and/or duration of infection and/ or increased clearance of resistant pathogens. Ultimately, the cost is paid by reduced transmission between hosts [2,6,7] . Fitness cost is often measured via controlled in vitro laboratory experiments in culture or animal models in the absence of drugs. Despite their limited fitness, resistant pathogens do not always revert to wild-type when serially passaged in drug-free

© 2009 Expert Reviews Ltd

ISSN 1478-7210

1

Babikera, Hastings & Swedberg

Malaria parasite: genetics & within host evolution of drug resistance

A

ut

ho

rP

Antimalarial drugs are the primary resources for the management and control of malaria burden, including the most vulnerable groups (children and pregnant women). The most virulent human malaria parasite Plasmodium falciparum has developed resistance to all existing antimalarial drugs, with the exception of the newly introduced artemisinin and its derivatives, which have only recently been deployed on a large scale [9,10] . Mutations associated with resistance to commonly used antimalarial drugs have been well characterized [11–14] . Resistances to some drugs, such as pyrimethamine, are controlled by mutations in a single gene, while others (e.g., chloroquine [CQ]) may require, or be modulated by, unlinked genes on different chromosomes. Mutations leading to drug resistance are adaptive changes in the face of drug pressure. If drug resistance is caused by a single gene then a mutant allele – carrying a fitness cost – can survive if its cost in the absence of the drug is outweighed by its benefit, and it is important to note that the cost–benefit balance is highly dependent of the level of drug use (Figure 1) [15] . Continuous drug pressure would not just put the mutant pathogens at a selective advantage but it can also favor mutations that compensate for the possible fitness cost of resistance [1] . When drug resistance is controlled by multiple mutations, then recombination can bring together or break down the relevant gene combinations, and give rise to lineages displaying variable levels of resistance and fitness. Therefore, the extent of crossing and recombination, which may either break down or build up such gene combinations, can have different outcomes in the presence or absence of drug pressure [16–19] . The blood stages of malaria parasites that cause the disease are haploid [20] , and more than one genetically distinct parasite clone of a given species often exist in a single host, including resistant and sensitive clones. An important biological feature of the malaria parasite that has significant impact on evolution of drug resistance is the occurrence of within-host multiplicity of clones. Multiple-clone infections are common, particularly in areas of high transmission intensity, such as in sub-Saharan Africa [21] . The parasite is extremely diverse in nature: there are multiple alleles of many genes, including those encoding antimalarial-targeted proteins. The high genetic diversity and allelic polymorphism has

been attributed to frequent crossing and recombination [22–24] . When a mosquito takes up gametocytes from two genetically distinct clones, cross-fertilization results in the formation of heterozygotes [22–24] . Genetic recombination then occurs during meiosis of heterozygotes, leading to the generation of novel haploid genotypes in the next generation of the parasite. Multiple parasite genotypes within the same infection can result in competitive interaction between these clones in attempts to secure resources [17,25] . Many laboratory studies and field surveys have demonstrated that multiple-clone infections comprise majority and minority clones [25–28] . When a multiple infection is composed of multiple drug-resistant and -sensitive clones, the resistant clone may be competitively suppressed by the sensitive clone in the absence of drug pressure [29–30] . Therefore, in-host competition may result in a competitive advantage and positive selection for virulent (majority) clones [25,29,30] . It has been suggested that if the dynamics of multiple P. falciparum clones co-existing in the same infection are not independent, and intra-host competition occurs, this may enhance transmission of drug-resistant clones [31] . Following treatment, drug-resistant parasites can replace the eliminated susceptible ones. The mutant clones can therefore increase in density, depending on the extent of multiplicity of infection [17,19] . Recent studies have confirmed that this type of intra-host competition occurs in the rodent malaria parasite Plasmodium chabaudi [30] . Support for this hypothesis has also come from ana­lysis of genes associated with resistance to CQ and sulfadoxine–pyrimethamine (SP) in P. falciparum in sites of different transmission intensity in Uganda [31,32] . Thus, many features of malaria parasite biology (intra-host dynamics, gametocytogenesis and recombination rate) are critical for understanding the evolution of drug resistance.

of

conditions. However, additional compensatory mutations can be acquired and ameliorate the fitness cost associated with antimicrobial resistance [7] . With regard to malaria parasites, the question of whether drugresistant mutants can be competitively suppressed by the sensitive forms, in the absence of the drug, was first considered decades ago [8] . However, unambiguous experimental evidence of competitive superiority of sensitive parasites is limited due to the laborious nature of laboratory competition ana­lysis and ethical issues associated with adequate study design in the field. Here we review accumulating epidemiological and laboratory evidence consistent with fitness deficit in drug-resistant malaria parasites and discuss its implications for the spread and management of drug resistance.

ro

Review

2

Within-host competition of resistant & sensitive malaria parasites

Controlled laboratory competition experiments using the rodent malaria model have provided primary evidence for a biological cost of resistance. Deliberate mixtures of drug-resistant and sensitive forms were examined for differences in growth rates and mosquito transmission [33] . To ascribe these differences to changes in drug-resistance genes only, and eliminate the possible involvement of other genes, the above studies used P. chabaudi isogenic clones, where a mutant clone was generated from a wild-type by drug selection. Initial studies produced conflicting results: a CQ-resistant mutant grew better than the wild-type sensitive form [8] but a pyrimethamine-sensitive P. chabaudi clone outcompeted the mutant derived from it, suggesting that the sensitive clone is more fit [33] . Using Plasmodium yoelii, Chawira et al. noted the disappearance of resistance to qinghaosu (artemisinin) following removal of drug pressure; however, it was regained rapidly when drug pressure was reapplied [34] . Attempts were also made to examine differences in transmission success. Shinondo et al. showed that a Plasmodium berghei pyrimethamine-resistant clone grew more slowly during its sporogenic cycle in mosquitoes, passing its sporozoites into the salivary glands several days later than the sensitive clones [35] . A Expert Rev. Anti Infect. Ther. 7(5), (2009)

Impaired fitness of drug-resistant malaria parasites

0.5 s = 0% s = 1% s = 5% s = 10% s = 20%

Rate of spread

0.4 0.3 0.2 0.1 0 -0.1

0

0.1

-0.2

0.3

Drug treatment rate

of

-0.3

0.2

ro

Figure 1. A simple example of the relationship between drug use and the spread of a resistance mutation and how this is affected by natural selection acting against the mutation in untreated hosts; based on Equation 11 of Hastings [22] with some simplifying assumptions*. The strength of natural selection against the mutation in untreated hosts was assumed to be 5% as estimated by Hastings and Donnelly [100] for the pfcrt mutation in Malawi. Treatment rate is the proportion of human infections that are drug-treated. Cutting treatment rate twofold from 30 to 15% results in a threefold reduction in the rate of spread (0.38–0.12) while cutting treatment threefold from 30 to 10% results in a sixfold reduction (0.38–0.06). There is an unstable threshold at treatment rate of 0.045: treatment rates above 0.045 are sufficient to drive the mutation through the population while treatment rates below 0.045 results in its loss (a ‘negative’ rate of spread) . This threshold was also identified by Koella and Antia [15] using an epidemiological approach based on differential equations. For illustration, we also show the relationship when natural selection pressures are 0, 1, 10 and 20%. * The revised equation is DP= {[d + (1-d)(1-s)]/(1-d)}-1 where DP is the change in frequency of the resistance gene per generation, d is the proportion of human infections that are drug-treated and s is the strength of natural selection against the mutation in untreated infections. It is assumed that competition between clones co-infecting the same human is absent, that only a single mutation is required to encode complete drug resistance, and that the frequency of resistance allele is rare (as will be the case in the initial stages of its spread).

A

ut

ho

rP

more recent study examined mosquitoes fed on gametocytes of a mixture of pyrimethamine-resistant P. chabaudi and its sensitive progenitor. The proportion of the sensitive genotype increased in mosquitoes relative to its density in asexual forms in two experiments, but decreased in a third experiment [36] . The contrasting outcome of the above competition experiments may be attributed to the lengthy nature of the protocol used and possible development of compensatory mutations during the process of selection for resistance and in vivo drug test. It is also important to note that mice are not a natural host for neither P. chabaudi nor P. yoelii. To eliminate some of these confounding factors, recent studies have monitored molecular markers associated with resistance to pyrimethamine using quantitative techniques [37,38] . These techechniques determine the proportion of resistant and sensitive alleles of the dhfr gene in an infection of resistant and sensitive parasites. A pyrimethamine-resistant P. chabaudi clone was found to grow more slowly than the sensitive clone from which it originated. However, several passages of this mutant clone through mice and mosquitoes in the absence of pyrimethamine produced a new strain, which exhibited better growth rate than the sensitive parent clone. This suggests that compensatory mutation(s) had probably occurred [36] . In addition, some studies have examined genetically diverse clones of P. chabaudi originally isolated from thicket rats in same area at the same time [39] . These studies used real-time PCR, to examine the proportion of sensitive and resistant clones [38] . Three independent studies unanimously showed that the resistant clone is competitively suppressed by the sensitive clone in the absence of drug pressure [29,30,40] . Similar competition experiments with P.  falciparum, have been limited by the difficulties of selecting drug-resistant clones in vitro. Nonetheless, the first drug resistant P. falciparum clones generated by in vitro drug selection, were atovaquone-resistant K1 clones, with double (M133I, G280D) and single (M133I) mutations in the cytochrome b gene [41] . The double mutant clone (M133I, G280D) was found to have lost 5–9% in fitness compared with the wild-type. The observed loss of fitness has been attributed to reduced binding of ubiquinone to cytochrome b as a result of the G280D mutation, which is close to a putative atovaquone-binding site. A different approach for selection of drugresistant P. falciparum clones, and generation of isogenic resistant and sensitive clones, has used a transfection method, in which a gene (pfmdr1) [12] conferring resistance to CQ was substituted by the wild-type gene [42] . An in vitro competition was then carried out between CQ-resistant and sensitive strains of P. falciparum. The transfected sensitive strain was found to have an increased growth rate, outgrowing the resistant parasite. In the absence of drug selection, mutations at codons 1034, 1042 and 1246 in the Pfmdr1 gene, which modulate levels of CQ resistance, gave rise to a substantial fitness cost. The loss of fitness incurred by these mutations was estimated at 25% with respect to the parent clone, in which wild-type allele had been substituted. The broader results of the above competition experiments, in rodent malaria parasites and in vitro P. falciparum, demonstrated that drug-sensitive parasites often grow and transmit better than the mutant ones. During an infection, the resistant genotype with

Review

www.expert-reviews.com

lower fitness can be targeted by natural selection, favoring the evolution of wild-type parasites in the absence of drug selection. As a result, the frequency of resistance will decline in the population. It is not clear how the fitness cost observed in laboratory competition experiments correlates with that seen in the field. However, such correlation has been proven to hold true for some pathogens [6] . It is therefore reasonable to extrapolate the above laboratory findings to the field situation; however, the magnitude of cost can be ameliorated by multiple inter-relating factors in the field. Field observations on fitness of drug-resistant P. falciparum

Three lines of evidence argue for fitness deficits in resistant P. falciparum clones in the absence of drug pressure in the field: 3

Babikera, Hastings & Swedberg

• Seasonal fluctuations in the frequency of drug-resistant genotypes in areas of marked seasonal transmission and drug usage; • The marked reduction in variation in large chromosomal segments flanking drug-resistance genes, which can purge genetic variation, rendering the parasite vulnerable to natural selection. Reduction of drug-resistant parasites following waning drug pressure Pyrimethamine & sulfadoxine

A

ut

ho

rP

An early opportunity to monitor the effect of discontinued therapy on the survival of drug resistant P. falciparum parasite arose after the withdrawal of pyrimethamine-based medicated salt as a feasible effective control measure in 1950s. Following cessation of the programme in Tanzania and withdrawal of the drug, the prevalence of pyrimethamine-sensitive P. falciparum parasites revived significantly [43] . However, the resistant parasite remained in the area for several years after withdrawal of the drug, demonstrating that some resistant parasites had persisted in the absence of the drug. In this case, a competitive advantage of pyrimethaminesensitive parasites was noticeable; nonetheless, the persistence of the resistant clones can be explained by development of stabilizing compensatory mutation. Broad epidemiological observations suggest that, once developed, drug-resistant forms of P. falciparum can persist despite intermittent use of therapy. However, their prevalence parallels the magnitude and pattern of drug usage. Large-scale molecular surveys in Mali, demonstrated that pyrimethamine-resistant genotypes concurrent with higher SP usage were high in urban areas where the drug is readily accessible and low in rural areas with limited drug dispersal [44,45] . These results illustrate the possible selective disadvantage of drug resistance in an environment of reduced drug pressure or, more simply, that selection is less intense in rural areas so that resistance is spreading more slowly. Similar findings have been reported from southern Mozambique, in an area neighbouring KwaZulu-Natal, South Africa. Longitudinal surveys have revealed a clear parallel between the prevalence of dhf and dhps mutations and deployment of SP. The frequency of dhps double mutations in the area peaked in 2001 (0.22) but declined to baseline levels (0.07) by 2004. Similarly, parasites with both dhfr triple and dhps double mutations had increased in 2001 (0.18) but decreased by 2004 (0.05). The peaking of SP-resistance mutations in 2001 coincided with a SP-resistant malaria epidemic in neighbouring KwaZulu-Natal, while the reduction in mutant genotypes corresponded with replacement of SP with artemether– lumefantrine in KwaZulu-Natal [46] . Additional associations between community usage of SP and prevalence of resistant genotypes came from the outcomes of an effective vector control trial. The use of insecticide-treated bednets (ITNs) in villages in Tanzania has been associated with significant decreases in P. falciparum-resistant genotypes - presumably because ITNs reduces transmission, disease incidence, drug usage and selective advantages of resistant parasites. The

prevalence of mutant dhfr and dhps genotypes in children less than 5 years old living in the above villages reduced dramatically between 1998 to 2000 following ITNs use. Conversely, the prevalence of the dhfr wild-type has increased significantly [47] . This matches prediction of drug resistance models, which suggest that lowering transmission can reduce the spread of drug resistance [17–19] . Although the interaction can be complex, with minimal selection at intermediate levels of transmission, depending on the underlying genetics of resistance [31] . Further evidence for a reduced fitness of the pyrimethamineresistant parasite, in the face of lower drug pressure, has been inferred from the observation of an inverse relationship between age and prevalence of resistant dhfr genotypes [48] . Since infections in older children are generally asymptomatic, it could be speculated that the survival advantage of mutant genotype populations is limited in the absence of therapy. In infants and younger children with a higher incidence of malaria-resistant genotypes can have a selective advantage and may consequently be more prevalent in this age group. On the other hand, the nature of frequent asymptomatic parasite carriage among older semi-immune children and adults can present an unfavorable environment for the resistant parasite. Among asymptomatic parasite carriages, the selective disadvantage of resistant parasites can be amplified by continuous asexual parasite multiplication over time. Such a pattern has been seen among asymptomatic parasite carriers, during the dry season, in areas with seasonal transmission [49] .

of

• The significant revival of sensitive parasites following withdrawal/ease of drug pressure;

ro

Review

4

Chloroquine

Ethical considerations limit approaches to obtain unequivocal data on the fitness cost of resistance, in the absence of drug pressure, as the drugs to which resistance has developed generally continue to be used in most countries. Nonetheless, the recent adoption of artesunate combination therapy (ACT) – backed by the WHO – as the first-line therapy in Africa provided an opportunity to monitor resistant forms in the absence of drugs replaced by ACT. Early observation on lower fitness of CQ-resistant P. falciparum parasites was obtained from southeast Asia, where resistance has developed in late 1950s. Following withdrawal of CQ for treatment of Falciparum malaria in Thailand, in vitro CQ sensitivity tests on Thai P. falciparum isolates collected between 1978 and 1986 demonstrated a gradual increase in drug sensitivity [50] . A similar decrease in CQ-resistant P. falciparum parasites was noted in Vietnam between 1995 and 1998, following withdrawal of the drug [51] , and China [52] . In a similarly study in Gabon, in vitro tests showed a significant decline in CQ resistance between 1992 and 1998, following changes in malaria treatment policy [53] . Recent replacements of CQ by ACT in many African countries has prompted large scale surveillance of the primary genetic determinant of CQ resistance, Pfcrt [14] and Pfmdr1 [12] , which augments the level of resistance in P. falciparum. Kublin et al. showed a marked decline in the prevalence of parasites containing the pfcrt76T allele between 1992 and 2000 in Malawi [54] . This decline correlated with the Expert Rev. Anti Infect. Ther. 7(5), (2009)

Impaired fitness of drug-resistant malaria parasites

ro

of

Abdel-muhsin et al. examined the frequencies of alleles of pfcrt, pfmdr1, pfdhfr and pfdhps in P. falciparum infections in a cohort of inhabitants of a village in eastern Sudan through the dry seasons of 1998 and 2000, when no transmission was occurring and drug usage was very limited [49,67] . At the end of each dry season, there was a decline in the frequency of both pfcrt76T, conferring CQ resistance [14] , and the pyrimethamine-resistance marker pfdhfr108N [11] , suggesting that parasites with these alleles were at a selective disadvantage compared with those with sensitive alleles in the absence of drug usage [49] . This pattern was consistent over two consecutive dry seasons in the same village [67] . A similar pattern in fluctuation of frequency of drug-resistance genes was found during the dry season in The Gambia, again implying a fitness deficit of resistance in absence of therapy [68] . The observed reduction in drug-resistant parasites over several months in the dry season can be a reflection of cumulative fitness cost during successive mitotic propagation in the dry season [69,70] . In a more recent longitudinal study in eastern Sudan, we noticed re-emergence of drug-sensitive parasites among asymptomatic parasite carriers in the dry season in the absence of treatment. A cohort of P. falciparum-infected patients was recruited in the transmission season, treated with a dose of CQ (25mg/ kg), and then followed monthly for 13 months. We used PCR to detect parasites that persisted at sub-patent levels in the dry season, and examined alleles of genes controlling the response of P. falciparum to CQ, the CQ-resistance transporter (pfcrt) and multidrug-resistance gene (pfmdr1). Mutant alleles of pfcrt reached fixation following CQ treatment and remained high in the transmission season. However, at the start of the dry season, wild-type alleles of both genes started to emerge and increased significantly in frequency as the season progressed (p = 0.03 for pfcrt and p = 0.003 for pfmdr1) (Babiker H, Unpublished Data) . These findings suggest that CQ-resistant P. falciparum clones were competitively suppressed by the drug-sensitive clone in absence of therapy. Therefore, in areas of seasonal drug pressure, sensitive parasites are likely to enjoy a selective advantage until transmission restarts, at which time clinical cases reappear and drug therapy rates increase [68,69] . However, the onset of drug treatment in the transmission season removes the competitive advantages of the drug-sensitive clones, and allows the resistant clone to fill the space occupied by susceptible clones, giving it a substantial and additional fitness benefit [30,71] . The magnitude of such a fitness benefit varies depending on the extent of multiplicity. For example, if a host is infected with three clones, the resistant clone can replace the two sensitive clones cleared by drugs, achieving a gain of 200% [71] . The observed cyclical revival of sensitive parasites at the end of the dry season in areas with seasonal malaria transmission agrees with the general pattern of epidemiology of drug resistance in these areas. Therefore, the discontinuous drug usage in the dry season, and possible accompanied costs, can result in slower spread of drug resistance compared with areas with high transmission and drug usage [69,72] . In these areas, the prevalence of drug-resistant parasites peaks with high levels of annual rains and transmission intensity, which results in high drug usage [69] .

A

ut

ho

rP

withdrawal of CQ from the area of study in 1993. Hastings and Donnelly estimated this fitness disadvantage at 5% per parasite generation and contrasted this with positive selection in the presence of the drug [55] , quantified elsewhere at approximately 15% per parasite generation, indicating that selective disadvantages may be relatively large. A decline in the frequency of pfmdr1 alleles associated with high CQ resistance was also noted, although this was less marked than that seen for pfcrt 76T. More recent studies demonstrated that the decline in the pfcrt76T and Pfmdr1 continues in Malawi [56,57] , and in vitro test confirmed that most isolates in the region are now sensitive to CQ and none is CQ-resistant, as predicted by Pfcrt genotype [57] . In addition, CQ has achieved a 99% rate of adequate clinical and parasitological response in children treated for uncomplicated P. falciparum malaria [58] . Following withdrawal of CQ in 2001 in Tanzania, similar surveys between 2002 and 2004 have supported the findings from Malawi. Prevalence of mutant alleles of both Pfcrt and Pfmdr1 decreased from 64.5 to 16% and 46.6 to 2.7%, respectively. By contrast, isolates with wild-type alleles of both Pfcrt and Pfmdr1 increased significantly [59] . These findings are consistent with the loss of CQ mutants following withdrawal of the drug, and a subsequent increase of CQ-susceptible parasiMefloquine Similar to SP and CQ, withdrawal of mefloquine in displaced camps on the northwest border of Thailand has led to decline of mefloquine-resistant parasites. Nosten et al. [60] assessed the incidence of P. falciparum malaria and responses to mefloquine treatment over 13  years in the study area. During this time, the artesunate and mefloquine combination was introduced as a first-line treatment for uncomplicated P. falciparum malaria. This has led to reduced incidence of P. falciparum malaria and mefloquine-resistant parasites. However, these changes have also been attributed to other factors, such as population movements and heterogeneity of transmission intensity, rather than being solely driven by reduced fitness of mefloquine-resistant parasites in the face of reduced drug pressure. Similar findings from this area have also been reported by Brockman et al., supporting the hypothesis that the combination of artesunate and mefloquine has reduced malaria transmission and parallel drug pressure, halting the decline in mefloquine sensitivity [61] .

Review

Interrupted drug usage & extended periods without therapy

Another set of data that provides evidence for a fitness deficit in drug-resistant P. falciparum has been obtained from studies in areas where malaria transmission and the pattern of antimalarial drugs usage are distinctly seasonal. In this unique transmission settings, a long dry period is followed by a brief period of annual rains and the appearance of Anopheles mosquitoes [62,63] . The infections established during the short, wet transmission season can be tracked throughout the lengthy, dry non-transmission season in the absence of new infections. This allows monitoring of the survival length of drug-resistant P. falciparum clones and their ability to produce gametocytes in the absence of drugs over a period of 7–8 months [64–66] . www.expert-reviews.com

5

Babikera, Hastings & Swedberg

The extent of hitchhiking is also affected by the prevailing rate of cross-mating and recombination among parasite populations. In parasite populations where self-fertilization predominates and lower levels of recombination occurs, drug selection will result in reduced genetic variation through a longer region of the chromosome. A clearer result of the strong selection and hitchhiking of mutant alleles of the above genes has been exemplified in appearance of single haplotypes of high-level pyrimethamine resistance in many countries [77,87] . This could mean that certain resistance alleles carried fitness costs, and thus parasites containing such alleles might have been at a selective disadvantage and therefore disappeared from the population [36] . Other explanations for the low rate of origin of mutations encoding drug resistance are discussed in [88] . Biochemical ana­lysis of folate pathway enzymes to explain possible fitness cost of mutations associated with resistance to antifolate drugs

To validate the above laboratory and field evidences for fitness costs of drug resistance, a powerful informative approach compared the kinetics of drug-targeted enzymes isolated from drugresistant and -sensitive malaria parasites. This has been well demonstrated in folate pathway inhibitors. Folate inhibitors are basically enzyme inhibitors, and ana­lysis of fitness effects of drugresistance mutations can be analyzed by the effect a mutation has on enzyme kinetic parameters. Specifically, detailed analyses of dihydrofolate reductase variants have been performed. Combinations between sulfadoxine and dihydrofolate reductase (dhfr) inhibitors like pyrimethamine show a marked synergistic effect. Part of the reason may be that pyrimethamine affects the assimilation of folates in the parasite, although direct measurement of folate uptake showed no evidence for this. [89] Another observation that is hard to explain is that a mutation in dhfr that leads to pyrimethamine or cycloguanil resistance also makes the parasite more dependent on de novo synthesis of folates. This is evident both from an increased dependence on p-aminobenzoic acid (PABA) during cultivation and from increased susceptibility to sulfadoxine inhibition [90] On the other hand, kinetic studies on purified dhfr have shown very limited effects on the enzyme’s Kcat, Vmax and K m values, at least for the commonly found Ser108Asn mutation, which appears to be the initial mutation in a sequence of dhfr mutation encoding gradually increasing [91] . However, one problem with this and similar studies is that only the dhfr part of the bifunctional dihydrofolate reductase-thymidylate synthase (dhfr-ts) complex was used for the assays, which were carried out with recombinant protein expressed in E. coli. Actually, older studies performed on native enzyme directly isolated from cultured parasites, show that the single 108N mutation suffers a high degree of functionality loss, which is partly restored by adding the 51I mutation [92] . These older kinetic data are easier to reconcile with data obtained from genetic crosses and field observations where the single 108N mutation seems to suffer a fitness defect, which is compensated for by additional mutations; both double and triple mutations are thus much more common than single 108 mutations [93,94] . A strong negative effect was seen for the corresponding mutation in P. chabaudi dhfr, 106N, which was

A

ut

ho

rP

The genome of malaria parasites contains 14 chromosomes, ranging in size from 0.643–3.290 Mb [73] . Sequencing the genome of P. falciparum has revealed that key ‘house-keeping’ genes occur in the central portions of each chromosome, whereas regions close to the end are highly variable, containing genes encoding polymorphic antigens and surface-related molecules [73] . The resistanceassociated genes (pfcrt, pfmdr1, dhfr and dhps) are spread over four different chromosomes [74,75] . The genetic consequences of drug selection, as a result of continuous use of antimalarials, on mutations associated with resistance has been recently addressed using a large number of polymorphic markers across the parasite genome [76] . Directional selection can increase the frequency of beneficial alleles to fixation, called a ‘selective sweep’. When a mutant gene is established in a parasite population, alleles at neutral and/or slightly deleterious loci in regions flanking the gene ‘hitchhike’ along with the selected drug-resistance locus and spread across the parasite population. This processing leads to increased linkage disequilibrium and removal of genetic variation, where balancing selection was previously maintaining several antigenic forms, close to mutant alleles in resistant isolates compared with sensitive ones [77] . There are currently no data to support this plausible hypothesis; however, deployment of genome wide tools for ana­lysis of field isolates can shed light on this. The process of selective sweep around drug-resistance genes has been noted among P.  falciparum parasite populations in all countries where such ana­lysis has been performed, indicating the power of drug selection [78–83] . For example, Nair et al. have observed reduced variation in chromosome 4 flanking the dhfr gene of approximately 100 kb (6 cM), indicative of strong selection [82] . A similar selective sweep has been seen around the CQ-resistance transporter (pfcrt) locus on chromosome 7 [78] . The reduction in variation around pfcrt was larger (200 kb) than that around dhfr. Removal of variation in genes linked to drug-resistance loci as a result of continuous reliance on chemotherapy for management and control can have debilitating effect on the ability of malaria parasites to survive human immunological responses. P. falciparum has extreme levels of allelic polymorphisms on surface antigen genes, indicative of balancing or frequency-dependent selection [84,85] . Removal of antigenic variation as a result of selection on neighboring drug-resistance genes might, therefore, reduce the ability of the parasite to persist and transmit ‘fitness’ in the face of immune selection [82] . For example, it has been suggested that there are 22 putative genes, in the hitchhiked area around the dhfr gene (~100‑kb) that are likely to have reduced variation [86] . Similar to dhfr and pfcrt, selective sweeps of resistant dhps alleles and linkage disequilibrium around the gene have been seen in chromosome 8 [80] and on chromosome 5 around pfmdr1 [83] . Although the size of the affected region near pfmdr1 is comparable to that of pfcrt and dhfr, the extent of linkage disequilibrium is modest compared with that seen around chromosome 4 and chromosome 7 [80,83] .

of

Signature of drug selection on parasite genome

ro

Review

6

Expert Rev. Anti Infect. Ther. 7(5), (2009)

Impaired fitness of drug-resistant malaria parasites

Do costs inevitable occur?

ro

of

It is generally assumed that Plasmodium metabolism has evolved to effectively meet its physiological requirements. It is therefore axiomic amongst evolutionary biologists that mutations that alter the structure of the gene product (such as mutations encoding drug resistance) will reduce the metabolic efficiency of the parasite carrying that mutation, and therefore the parasite suffers a fitness penalty. There have been two drugs deployed worldwide over extended periods of time, CQ and SP, and the genetic basis of resistance to both drugs is relatively well characterized [101– 104] . Interestingly, they suggest qualitatively different costs of resistance. Resistance to CQ is the most clear-cut example. Resistance is encoded by mutations in the pfcrt gene and particularly in position 76, giving rise to the pfcrtK76T mutation. The frequency of this mutation fell rapidly after CQ was withdrawn in both Malawi [101] and China; the selective penalty in Malawi was estimated at approximately 5% per parasite generation [105] . Resistance to antifolates such as SP is encoded by mutations in the pfdhfr and pfdhps genes and again is relatively well-characterized (op. cit.). There are two lines of evidence that suggest that the selective penalties associated with mutations in these enzymes may be small. First, the frequencies of dhfr mutations did not fall once SP use was discontinued in SE Asia [89] . Second, when SP was introduced as a first-line drug in areas of eastern Africa, it was noted that mutations in dhfr were already present, one plausible explanation being that the mutations had been selected by antifolate deployments several decades beforehand, and that natural selection pressures have been so small that these mutations could persist for extended periods in the absence of any antifolate drug pressure. Small selective pressures are not unexpected from a population genetics viewpoint. It is known from genetic knock-out experiments that most genes can be silenced with no apparent impact on the phenotype of the animals [106–108] . In addition, metabolic control theory (MCT) suggests that changes in most enzymes can be effectively buffered by other enzymes in that particular part of metabolism [109–111] . Obviously the targets of antimalarial drugs will not be typical because effective drugs, by definition, must target essential elements of Plasmodium metabolism; however, the same general principles applies. In the case of dhfr, the enzyme structure need be altered only to reduce or prevent binding of pyrimethamine. Mutations reducing drug binding may arise through steric interference and can presumably occur away from the active site of the enzyme; consequently, they may have a relatively small impact on the enzyme’s catalytic activity (discussed in more detail previously). In vivo measures of enzymes reveal that the dhfr mutations that encode pyrimethamine resistance have a relatively small impact on enzyme activity. The general properties of enzyme systems identified by MCT suggest these could be easily compensated for by other enzymes in the pathway. Another factor that may mitigate the costs of the mutation is ‘compensatory’ mutations elsewhere in the folate pathway that may compensate for the changes induced by dhfr mutations [100] .

A

ut

ho

rP

outcompeted by wild-type parasites upon co-infection [36] . The negative effects were, however, reversed by repeated passaging through mice and mosquitoes. The reversal is thought to be due to compensatory mutations. However, no attempts to identify these compensatory mutations have been reported. There is a similar discrepancy in the ana­lysis of the dhfr quadruple mutation that includes 164L. Most enzyme kinetic ana­lysis indicates this to be a relatively well-functioning enzyme. The latest publication on the issue showed no significant difference in the kinetics of the quadruple mutation in comparison with the wild-type enzyme [95] In this case, both dhfr and TS were present in the assay, expressed as separate polypeptides from two plasmids. However, complementation studies in yeast, as well as the apparent lack of this mutation in Africa, suggest a fitness disadvantage in vivo [96,97] . On the other hand, the quadruple mutation is fairly common in South-East Asia and even very common in the Nicobar Islands in India [98] . An explanation may be found in the recently detected copy number mutation affecting the gene encoding GTP cyclohydrolase (GTPCH), the first step in folate biosynthesis. The effect was first observed in cultured isolates during a screen for transcriptional changes in parasites [99] and was followed up by an extensive study in Thailand showing a genetic linkage between GTPCH copy number and the presence of the allele 164L carried by the quadruple mutants [100] . A general problem with the published results from kinetic ana­lysis of dhfr is that only the monoglutamate form of folate was used in the assays, while recycled folates are predominantly in the form of polyglutamates. In fact, the above observations would be readily explained if there is a difference in the efficiency of dhfr that is dependent on the form of folate used as substrate. Reduction of DHF during de novo synthesis could well be performed in the monoglutamate form, while recycling always involved polyglutamate forms. If polyglutamates are used less efficiently by the mutated dhfr, this could then explain that recycling is inefficient and hence that de novo synthesis becomes the predominant supply of reduced folates. Furthermore, the necessary interactions between dhfr and TS could be affected by the mutation, also leading to a diminished efficiency of recycling. The enzyme kinetic work could thus be extended to take these considerations into account, with the use of polyglutamate folates as substrates for the different variants of dhfr and with the fulllength bifunctional enzyme.

Review

Impact of fitness cost on the dynamics of drugresistant parasite & drug-deployment strategies

There are three main questions that arise when considering the impact of fitness costs on the evolution of resistance and how this may determine optimal drug-deployment strategies. First, do these costs inevitably arise or are they likely to be important in some cases and negligible in others? Second, if significant fitness costs do arise, how are the costs likely to be paid? In other words, when and how will natural selection act against them? Third, how will the presence or absence of costs affect the dynamics of the evolution of resistance and what are the implications for public-health policies designed to reduce the evolution of resistance? www.expert-reviews.com

7

Babikera, Hastings & Swedberg

given the rather weak correlation between asexual parasitemia and sexual parasitemia (and hence, presumably, infectivity). There is however evidence that competition occurs between separate clones co-infecting the same host. This would undoubtedly create the conditions required for ‘soft’ selection to occur. There is some epidemiological evidence that this occurs in human [31,115] and in a laboratory model system [3] . There is additional evidence that density-dependence occurs in the mosquito stage [116] which would also allow for competition and soft selection in this phase of the lifecycle. So there is evidence for potential competition to occur at the mosquito stage but, it is important to note, no evidence that it actually does occur.

of

How costs affect the dynamics of how resistance evolves

If costs are present, they have several consequences. If costs manifest as reduced duration of infection, or reduced ability to establish an infection, then this may result in a lower proportion of people being infected (although the impact will probably be slight in high-transmission areas). It may also result in reduced multiplicity of infections within humans, which in turn affects recombination rates, which are one of the chief drivers of resistance (discussed in more detail elsewhere) [117] . Mutations encoding resistance pay a cost in humans (and possibly mosquitoes), which are not drug treated, but this needs to be balanced against their advantage in resisting the drug in treated humans. As pointed out by Koella and Antia [15] . this introduces a threshold effect: if drug use is below a critical threshold, the disadvantage of the mutation in untreated hosts exceeds their advantage in the (small number) of treated hosts, hence resistance will not spread. A related effect was noted by Hastings and Donnelly [55] who discussed the nonproportionality of reducing drug treatment. As a crude example, imagine the overall cost of resistance is 5% and overall advantage in the presence of the drug is 30%: this gives an advantage to resistance of 30–5 = 25%. Halving drug use has a disproportionate effect, reducing fitness to (30/2)-5 = 10%; reducing use by a third gives fitness of (30/3)-5 = 5%, and so on (see Figure 1 for more explicit calculations).This effect was also identified in a recent paper advocating the simultaneous use of multiple antimalarial therapies: providing costs of resistance were present, then this strategy reduces overall drug pressure and hence the rate at which resistance evolves [118] . The above arguments are largely robust to whether selection is ‘hard’ or ‘soft’. If selection is ‘soft’ then it can have interesting effects on the dynamics of resistance. Recall that soft selection occurs when separate clones co-infecting the same host compete for transmission. When the frequency of resistance is low then most resistant clones will be competing with the fitter wild-type and costs may be high. Conversely, when the frequency of resistance is high, most co-infecting clones will be of the same less-fit resistant form so costs may be much less [119] . This introduces frequency–dependent effects that are familiar in classical population genetics as a force stabilizing mutation frequencies around an equilibrium point [120] . In the case of antimalarial drug resistance, this effect (together with frequency-dependent effects of recombination and intrahost competition) can stabilize the frequency

How are the costs paid?

ho

rP

Thus, it is relatively simple to construct an argument that the costs of resistance may be effectively negligible (although they do become important when incremented over extended evolutionary timescales, as in the classical mutation/selection balance). The significant costs associated with the pfcrtK76T mutation could then be regarded as exceptional and explained by its different role in metabolism: it is a transmembrane transporter molecule so it would not benefit from the buffering inherent in pathways of enzymes, and its fundamental role in transporting metabolites may make any changes affect numerous aspects of metabolism, thereby making mutations highly deleterious. Interestingly, the literature on insecticide resistance draws a clear distinction between mutations in the drug targets (e.g., dhfr) and mutations in drug-transporter systems that may affect access of the drug to its target (e.g., pfcrt and pfmdr1). These are known as ‘target site resistance’ and ‘metabolic resistance’, respectively [112] , and the malaria community may well be advised to heed this distinction. These are, of course, post-hoc arguments constructed to explain the data based on two resistance mechanisms, so data on more resistance mutations is urgently needed. It is important to make the argument that selective disadvantages may be negligible. The well-publicized disappearance of CQ resistance once the drug was removed has led to the general expectation that all resistance will exhibit similarly large magnitudes of selective disadvantage. This will become important once public-health policy recommendations are made, which are based on the assumption of high values of selective disadvantage.

ro

Review

A

ut

Natural selection against the mutant can be either ‘hard’, in which case the cost is paid irrespective of the presence/absence of other malaria parasites, or selection can be ‘soft’, in which case the costs only become apparent in the presence of competition from wild-type parasites [113] . There appear to be three main stages of the malaria life cycle in which these costs may arise. First, the mutations may suffer a disadvantage in the initial human liver stage that reduces the parasites’ chance of establishing a viable blood-stage infection or, if replicative efficiency is compromised in the liver, then the parasites may spend longer in the prepatent hepatic stage. This would be very difficult to detect and we are unaware of any data supporting this. Second, the selective disadvantage may be expressed as a reduced duration of infection and/or reduced infectivity. Schoepflin and colleagues recently presented evidence that resistant infections tend to have a shorter duration of infection [114] . Their methodology would not distinguish ‘hard’ from ‘soft’ selection. There is evidence that the prevalence of resistance falls during periods when drug pressure is lowered, typically during the dry seasons when transmission is reduced (see previously), which is generally, and plausibly, interpreted as evidence of a fitness cost [62–79,82] . Importantly, in the absence of transmission and hence liver-stage costs, this change in prevalence must be due to a reduced duration of infection. Third, the cost may be expressed through reduced transmission from the human host. Hard selection against resistance (i.e., a consistently lower transmission potential) would be very hard to detect, 8

Expert Rev. Anti Infect. Ther. 7(5), (2009)

Impaired fitness of drug-resistant malaria parasites

• Simultaneous monotherapy, where two or more monotherapies are deployed in the same region at the same time;

The introduction of ACTs, backed by the WHO, has encouraged the health authorities in many African countries to abandon the use of existing drugs, which have been used for decades and to which resistance mutations have accumulated. Such a move has circumvented an ethical dilemma for researchers wishing to examine the fitness cost of resistance by allowing monitoring of prevalence of resistance mutations to these drugs among asymptomatic parasite carriers. Asymptomatic infections represent the common parasite reservoir, and are of major importance from a public-health perspective. Recent technical developments allow quantitative monitoring of drug-resistance mutations, even when the wild and mutant genotypes co-exist in the same infection. In addition, it has also become possible to quantify genotype-specific investment into gametocytes when co-infecting genotypes that vary in their competitive ability. Transmission success is not only related to the density of infectious gametocytes, but also to their sex ratio (proportion of gametocytes that are male). Modulation of sex ratio can, therefore, allow each genotype within an infection to maximize its genetic representation in the next generation. The above tools and approaches will allow us to work out better estimates of the fitness cost of drug-resistance mutations. In view of the diverse nature of drug regimens and history of resistance prior to ACTs in many African countries, the coming decade provides an opportunity to monitor and examine the fate of these mutations in the absence of drugs. The use of advanced molecular tools will allow more realistic estimates of the relative cost of resistance (in terms of growth and transmission capacity) to each drug.

A

ut

ho

rP

• Combination therapy, where two or more drugs are given to patients at the same time. The dynamics by which these different strategies drive resistance are complex, but combination therapy appears to be the best strategy, except in very exceptional circumstances. The presence and magnitude of costs of resistance, therefore, have quantitative impacts on the spread of resistance to combinations, rather than a qualitative role in deciding between different deployment strategies. However, there is evidence that the current batch of combination therapies based around artemisinins are starting to fail [121] so the other strategies may have to be considered and re-evaluated. In this case, fitness costs do generate qualitative difference between the different monotherapy strategies (A ntao & H astings IM, Unpublished Data) so it seems appropriate that the fitness costs associated with resistance should be quantified as a matter of some urgency.

Five-year view

of

• Spatial deployment of different monotherapies (or mosaics), where different regions use different monotherapies;

CQ and SP means that mutations conferring resistance to these drugs may be subjected to natural selection and fall in frequency at a rate determined by their relative fitness cost. Field data has demonstrated that CQ-resistance mutations can be eliminated following extended period of absence of the drug. The fitness cost of CQ mutations has been estimated at 5% per parasite generation. There is currently a lack of similar estimates of the relative fitness cost of resistance to some antimalarial drugs, such as sulfadoxine and pyrimethamine. Such information should lead to the design of improved models to describe evolution of drug resistance that can guide rational drug-deployment policies.

ro

of resistance around either a stable point, or can force periodic, seasonal oscillation around a long-term equilibrium [119] . Finally, it is important to note that we have been considering individual mutations in isolation. There is evidence that separate mutations in different genes may interact (‘epistasis’) to modify each others’ ability to encode resistance. The arguments developed above are generally not affected by this phenomenon but it remains a pressing research question to elucidate just how much of the variation in drug resistance levels can be explained by mutations in known resistance-encoding genes and the level of epitasis that exists between them. The question of how best to deploy antimalarial drugs has been the subject of intense speculation and modeling. There are four main strategies: • Sequential monotherapies, where a single drug is deployed and replaced by another monotherapy once it starts to fail;

Review

Expert commentary

Current concerted efforts to control malaria lead by the Role Back Malaria (RBM) program of the WHO, may reduce transmission of the parasite and malaria incidence. Parallel to RBM, WHO has initiated a campaign to change first-line, low-efficacy drugs, such as CQ and SP, to artemisinin-based combination therapies (ACTs). In view of the high efficacy and shorter half-life of artimisinin derivatives, it may take time for resistance to these drugs to develop. However, their counterparts are less effective and resistance to some of them is already high in the field. Therefore, ACTs can serve as an effective first-line drugs for some years to come for management and control strategies. However, in view of the weak health infrastructure in malaria-endemic countries, the parasite is expected to resurge if or when control efforts are relaxed. Meanwhile, the withdrawal of the antimalarials such as www.expert-reviews.com

Acknowledgements

We thank Aysha AlGhazali and Taruna Duff for help with preparation of the manuscript. We also thank three anonymous referees for their helpful comments. Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. 9

Review

Babikera, Hastings & Swedberg

Key issues

Papers of special note have been highlighted as: • of interest •• of considerable interest 1

Levin BR, Perrot V, Walker N. Compensatory mutations, antibiotic resistance and the population genetics of adaptive evolution in bacteria. Genetics 154, 985–997 (2000). Ma C, Li C, Ganesan L et al. Mutations in a-tubulin confer dinitroaniline resistance at a cost to microtubule function. Mol. Biol. Cell. 18, 4711–4720 (2007).

3

Björkman J, Hughes D, Andersson DI. Virulence of antibiotic-resistant Salmonella typhimurium. Proc. Natl Acad. Sci. USA 95, 3949–3953 (1998).

5

Leigh Brown AJ, Frost SD, Mathews WC et al. Transmission fitness of drug-resistant human immunodeficiency virus and the prevalence of resistance in the antiretroviral-treated population. J. Infect. Dis 187, 683–686 (2003).

6

Andersson DI. Persistence of antibiotic resistant bacteria. Curr. Opin. Microbiol. 6, 452–456 (2003).

7

Andersson DI, Levin BR. The biological cost of antibiotic resistance. Curr. Opin. Microbiol. 2, 489–493 (1999).

8

9

10

Bosman A, Mendis KN. A major transition in malaria treatment: the adoption and deployment of artemisinin-based combination therapies. Am. J. Trop. Med. Hyg. 77, 193–197 (2007).

Hastings IM, Watkins WM. Intensity of malaria transmission and the evolution of drug resistance. Acta Trop. 94, 218–229 (2005).

19

Mackinnon MJ, Hastings IM. The evolution of multiple drug resistance in malaria parasites. Trans. R. Soc. Trop .Med. Hyg. 92, 188–195 (1998).

20

Walliker D. Malaria, in Thompson RCA (Ed.). Molecular Epidemiology of Infectious Diseases. Arnold, London, UK 93–112 (2000).

21

Babiker HA, Walliker D. Current views on the population structure of Plasmodium falciparum: implications for control. Parasitol. Today 13, 262–267 (1997).

22

Babiker HA, Ranford-Cartwright L, Currie D et al. Random mating in a natural population of the malaria parasite Plasmodium falciparum. Parasitology 109, 413–421 (1994).

23

Paul RE, Packer MJ, Walmsley M et al. Mating patterns in malaria parasite populations of Papua New Guinea. Science 269, 1709–1711 (1995).

24

Conway DJ, Roper C, Oduola AM et al. High recombination rate in natural populations of Plasmodium falciparum. Proc. Natl Acad. Sci. USA 96, 4506–4511 (1999).

25

de Roode JC, Culleton R, Bell AS, Read AF. Competitive release of drug resistance following drug treatment of mixed Plasmodium chabaudi infections. Malar. J. 3, (2004).

26

Daubersies P, Sallenave-Sales S, Magne S et al. Rapid turnover of Plasmodium falciparum populations in asymptomatic

Peterson DS, Walliker D, Wellems TE. Evidence that a point mutation in dihydrofolate reductase-thymidylate synthase confers resistance to pyrimethamine in falciparum malaria. Proc. Natl Acad. Sci. USA 85, 9114–9118 (1988).

12

Reed MB, Saliba KJ, Caruana SR, Kirk K, Cowman AF. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum. Nature 403, 906–909 (2000).

ut

Borman AM, Paulous S, Clavel F. Resistance of human immunodeficiency virus type 1 to protease inhibitors: selection of resistance mutations in the presence and absence of the drug. J. Gen. Virol. 77, 419–426 (1996).

A

4

11

18

ho

2

10

prevention of antimalarial drug resistance: rationale for combination chemotherapy for malaria. Parasitol. Today 399–401 (1996)

rP

References

ro

of

• The most pathogenic human malaria parasite Plasmodium falciparum has acquired resistance to all available antimalarial drugs, with the exception of artimisinin derivatives. • Lessons learn from other pathogenic organisms demonstrated that mutations that render an organism resistant to chemotherapeutic agents can be metabolically costly, affecting the pathogen’s ability to survive, propagate and transmit. • Controlled laboratory experiments that used isogenic strains, which vary only with regard to their response to an antimalarial drug, have demonstrated that changes that cause drug resistance lead to lower fitness of the resistant parasites in the absence of drugs. • A large body of epidemiological data demonstrated that the frequency of resistant genotypes wanes when active drug selection declines. • The high usage of antimalarial drugs for management and control of malaria impose active selection on the parasite. Analysis of chromosomal regions close to drug-resistance genes show evidence of reduced genetic variation (hitch-hiking). This can lead to reduction of variation of antigenic polymorphism and ability of the parasite to overcome immune responses. • Mutations associated with some drugs, such as chloroquine, show large magnitudes of selective disadvantage while others may not. • More data is needed to estimate the relative cost of resistance to different drugs. • A more realistic epidemiological model of drug resistance may allow a better understanding of the role of selective disadvantage in different transmission settings on which public-health policy recommendations can be made.

13

14

Wang P, Lee CS, Bayoumi R et al. Resistance to antifolates in Plasmodium falciparum monitored by sequence analysis of dihydropteroate synthetase and dihydrofolate reductase alleles in a large number of field samples of diverse origins. Mol. Biochem. Parasitol. 89, 161–177 (1997). Fidock DA, Nomura T, Talley AK et al. Mutations in the P. falciparum digestive vacuole transmembrane protein PfCRT and evidence for their role in chloroquine resistance. Mol. Cell 6, 861–871 (2000).

15

Rosario VE. Genetics of chloroquine resistance in malaria parasites. Nature 261, 585–586 (1976).

Koella JC, Antia R. Epidemiological models for the spread of anti-malarial resistance. Malar. J. 2, (2003).

16

Mackinnon MJ. Drug resistance models for malaria. Acta Trop. 94, 207–217 (2005).

White NJ, Olliaro PL. Strategies for the

17

Hastings IM. A model for the origins and spread of drug-resistant malaria.

Parasitology 115, 133–141 (1997).

Expert Rev. Anti Infect. Ther. 7(5), (2009)

Impaired fitness of drug-resistant malaria parasites

28

Bell AS, de Roode JC, Sim D, Read AF. Within-host competition in genetically diverse malaria infections: parasite virulence and competitive success. Evolution 60, 1358–1371 (2006).

29

30

de Roode JC, Culleton R, Cheesman SJ, Carter R, Read AF. Host heterogeneity is a determinant of competitive exclusion or coexistence in genetically diverse malaria infections. Proc. Biol. Sci. 22, 1073–1080 (2004). Wargo AR, Huijben S, Roode J CD, Shephard J, Read AF. Competitive release and facilitation of drug-resistant parasites after therapeutic chemotherapy in a rodent malaria model. Proc. Natl Acad. Sci. USA 104, 19914–19919 (2007). Hastings IM. Malaria control and the evolution of drug resistance: an intriguing link. Trends Parasitol. 19, 70–73 (2003).

32

Talisuna AO, Langi P, Mutabingwa TK et al. Intensity of transmission and spread of gene mutations linked to chloroquine and sulphadoxine–pyrimethamine resistance in falciparum malaria. Int. J. Parasitol. 33, 1051–1058 (2003).

Rosario VE, Hall R, Walliker D, Beale GH. Persistence of drug-resistant malaria parasites. Lancet 1, 185–187 (1978).

34

Chawira AN, Warhurst DC, Peters W. Qinghaosu resistance in rodent malaria. Trans. R. Soc. Trop. Med. Hyg. 80, 477–480 (1986).

A

Shinondo CJ, Lanners HN, Lowrie RC, Wiser MF. Effect of pyrimethamine resistance on sporogony in a Plasmodium berghei/Anopheles stephensi model. Exp. Parasitol. 78, 194–202 (1994).

36

Walliker D, Hunt P, Babiker H. Fitness of drug-resistant malaria parasites. Acta Trop. 94, 251–259 (2005).

37

Hunt P, Fawcett R, Carter R, Walliker D. Estimating SNP proportions in populations of malaria parasites by sequencing: validation and applications. Mol. Biochem. Parasitol. 143, 173–182 (2005).

38

39

Thaithong S, Suebsang L, Rooney W, Beale GH. Evidence of increased chloroquine sensitivity in Thai isolates of Plasmodium falciparum. Trans. R. Soc. Trop. Med. Hyg. 82, 37–38 (1988).

51

41

Peters JM, Chen N, Gatton M et al. Mutations in cytochrome b resulting in atovaquone resistance are associated with loss of fitness in Plasmodium falciparum. Antimicrob. Agents Chemother. 46, 2435–2441 (2002).

Thanh NV, Cowman AF, Hipgrave D et al. Assessment of susceptibility of Plasmodium falciparum to chloroquine, quinine, mefloquine, sulfadoxine-pyrimethamine and artemisinin in southern Vietnam. Trans. R. Soc. Trop. Med. Hyg. 95, 513–517 (2001).

42

Hayward R, Saliba KJ, Kirk K. pfmdr1 mutations associated with chloroquine resistance incur a fitness cost in Plasmodium falciparum. Mol. Microbiol. 55, 1285–1295 (2005).

52

Liu DQ, Liu RJ, Ren DX et al. Changes in the resistance of Plasmodium falciparum to chloroquine in Hainan, China. Bull. World Health Organ. 73, 483–486 (1995).

43

Clyde DF. Treatment of multidrug resistant malaria. East Afr. Med. J. 44, 231–327 (1967).

44

Plowe CV, Djimde A, Bouare M, Doumbo O, Wellems TE. Pyrimethamine and proguanil resistance-conferring mutations in Plasmodium falciparum dihydrofolate reductase: polymerase chain reaction methods for surveillance in Africa. Am. J. Trop. Med. Hyg. 52, 565–568 (1995).

45

ut

33

35

Wargo AR, de Roode JC, Huijben S, Drew DR, Read AF. Transmission stage investment of malaria parasites in response to in-host competition. Proc. Biol. Sci. 274(1625), 2629–2638 (2007).

Plowe CV, Djimde A, Wellems TE, Diop S, Kouriba B, Doumbo OK. Community pyrimethamine-sulfadoxine use and prevalence of resistant Plasmodium falciparum genotypes in Mali: a model for deterring resistance. Am. J. Trop. Med. Hyg. 55, 467–471 (1996).

ho

31

50

40

of

Paul RE, Diallo M, Brey PT. Mosquitoes and transmission of malaria parasites – not just vectors. Malar. J. 3, 39 (2004).

Cheesman SJ, de Roode JC, Read AF, Carter R. Real-time quantitative PCR for analysis of genetically mixed infections of malaria parasites: technique validation and applications. Mol. Biochem. Parasitol. 131, 83–91 (2003). Beale GH, Carter R, Walliker D. Genetics.

www.expert-reviews.com

falciparum drug resistance genes in Sudan. Parasitology 126, 391–400 (2003).

53

Schwenke A, Brandts C, Philipps J, Winkler S, Wernsdorfer WH, Kremsner PG. Declining chloroquine resistance of Plasmodium falciparum in Lambaréné, Gabon from 1992 to 1998. Wien. Klin. Wochenschr. 113, 63–64 (2001).

54

Kublin JG, Cortese JF, Njunju EM et al. Reemergence of chloroquine-sensitive Plasmodium falciparum malaria after cessation of chloroquine use in Malawi. J. Infect. Dis. 187, 1870–1875 (2003).

55

Hastings IM, Donnelly MJ. The impact of antimalarial drug resistance mutations on parasite fitness, and its implications for the evolution of resistance. Drug Resist. Updat. 8, 43–50 (2005).

56

Mita T, Kaneko A, Lum JK et al. Recovery of chloroquine sensitivity and low prevalence of the Plasmodium falciparum chloroquine resistance transporter gene mutation K76T following the discontinuance of chloroquine use in Malawi. Am. J. Trop. Med. Hyg. 68, 413–415 (2003).

57

Nkhoma S, Molyneux M, and Ward S. Molecular surveillance for drug-resistant Plasmodium falciparum malaria in Malawi. Acta Trop. 102, 138–142 (2007).

58

Laufer MK, Thesing PC, Eddington ND et al. Return of chloroquine antimalarial efficacy in Malawi. N. Engl. J. Med. 355, 1959–1966 (2006).

59

Temu EA, Kimani I, Tuno N, Kawada H, Minjas JN, Takagi M. Monitoring chloroquine resistance using Plasmodium falciparum parasites isolated from wild mosquitoes in Tanzania. Am. J. Trop. Med. Hyg. 75, 1182–1187 (2006).

60

Nosten F, van Vugt M, Price R et al. Effects of artesunate-mefloquine

ro

27

In: Rodent Malaria. Killick-Kendrick R, Peters W (Eds.). Academic Press, London, UK, 213–245 (1978).

rP

individuals living in a high transmission area. Am. J. Trop. Med. Hyg. 54, 18–26 (1996).

Review

46

Raman Raman J, Sharp B, Kleinschmidt I et al. Differential effect of regional drug pressure on dihydrofolate reductase and dihydropteroate synthetase mutations in southern Mozambique. Am. J. Trop. Med. Hyg. 78(2), 256–261 (2008).

47

Alifrangis M, Lemnge MM, Rønn AM et al. Increasing prevalence of wildtypes in the dihydrofolate reductase gene of Plasmodium falciparum in an area with high levels of sulfadoxine/pyrimethamine resistance after introduction of treated bed nets. Am. J. Trop. Med. Hyg. 69, 238–243 (2003).

48

Mockenhaupt FP, May J, Eggelte TA et al. High prevalence and imbalanced age distribution of the Plasmodium falciparum dihydrofolate reductase gene Asn108 mutation in an area of low pyrimethamine usage in Nigeria. Am. J. Trop. Med. Hyg. 61, 375–377 (1999).

49

Abdel-Muhsin AA, Mackinnon M, Ali E et al. Local differentiation in Plasmodium

11

Babikera, Hastings & Swedberg

64

65

66

67

Babiker HA. Unstable malaria in Sudan: the influence of the dry season. Plasmodium falciparum population in the unstable malaria area of eastern Sudan is stable and genetically complex. Trans. R. Soc. Trop . Med. Hyg. 92, 585–589 (1998). Nassir E, Abdel-Muhsin AM, Suliaman S et al. Impact of genetic complexity on longevity and gametocytogenesis of Plasmodium falciparum during the dry and transmission-free season of eastern Sudan. Int. J. Parasitol. 35, 49–55 (2005).

Nwakanma D, Kheir A, Sowa M et al. High gametocyte complexity and mosquito infectivity of Plasmodium falciparum in the Gambia. Int. J. Parasitol. 38, 219–227 (2007). Abdel-Muhsin AA, Mackinnon M, Awadalla P et al. Evolution of drugresistance genes in Plasmodium falciparum in an area of seasonal malaria transmission in Eastern Sudan. J. Infect. Dis. 189, 1239–1244 (2004).

68

Ord R, Alexander N, Dunyo S et al. Seasonal carriage of pfcrt and pfmdr1 alleles in Gambian Plasmodium falciparum imply reduced fitness of chloroquineresistant parasites. J. Infect. Dis. 196, 1613–1619 (2007).

69

Babiker HA, Satti G, Ferguson HM, Bayoumi R, Wallikef D. Drug resistant Plasmodium falciparum in an area of seasonal transmission. Acta Trop. 94, 260–268 (2005).

70

Felger I, Beck HP. Fitness costs of resistance to antimalarial drugs. Trends Parasitol. 24, 331–333 (2008).

71

Hastings IM. Malaria control and the evolution of drug resistance: an intriguing

12

74

75

76

77

Gardner MJ, Shallom SJ, Carlton JM et al. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 419, 498–511 (2002). Plowe CV, Kublin JG, Doumbo OK. P. falciparum dihydrofolate reductase and dihydropteroate synthase mutations: epidemiology and role in clinical resistance to antifolates. Drug Resist. Updat. 1, 389–396 (1998).

78

Figueiredo P, Benchimol C, Lopes D et al. Prevalence of pfmdr1, pfcrt, pfdhfr and pfdhps mutations associated with drug resistance, in Luanda, Angola. Malar. J. 17, 236 (2008). Su X-Z, Ferding MT, Huang Y et al. Genetic map and recombination parameters of the human malaria parasite P. falciparum. Science 286, 1351–1353 (1999).

Nair S, Brockman A, Paiphun L, Nosten F, Anderson TJC. Rapid genotyping of loci involved in antifolate drug resistance in Plasmodium falciparum by primer extension. Int. J. Parasitol. 32, 852–858 (2002). Wootton JC, Feng X, Ferdig MT et al. Genetic diversity and chloroquine selective sweeps in Plasmodium falciparum. Nature 418, 320–323 (2003).

79

Pearce R, Malisa A, Kachur SP, Barnes K, Sharp B, Roper C. Reduced variation around drug-resistant dhfr alleles in African Plasmodium falciparum. Mol. Biol. Evol. 22, 1834–1844 (2005).

80

McCollum AM, Mueller K, Villegas L, Udhayakumar V, Escalante AA. Common origin and fixation of Plasmodium falciparum dhfr and dhps mutations associated with sulfadoxine-pyrimethamine resistance in a low-transmission area in South America. Antimicrob. Agents Chemother. 51, 2085–2091 (2007).

81

McCollum AM, Mueller K, Villegas L, Udhayakumar V, Escalante AA. Hitchhiking and selective sweeps of Plamsodium falciparum sulfadoxine and pyrimethamine resistance alleles in populations from Africa. Antimicrob. Agents Chemother. 52, (2008).

82

selective sweep driven by pyrimethamine treatment in southeast asian malaria parasites. Mol. Biol. Evol. 20, 1526–1536 (2003). 83

Nair S, Nash D, Sudimack D et al. Recurrent gene amplification and soft selective sweeps during evolution of multidrug resistance in malaria parasites. Mol. Biol. Evol. 24, 562–573 (2007).

84

Conway DJ. Natural selection on polymorphic malaria antigens and the search for a vaccine. Parasitol Today. 13 26–29 (1997).

85

Ferreira MU, Zilversmit M, Wunderlic G. Origins and evolution of antigenic diversity in malaria parasites. Curr. Mol. Med. 7, 588–602 (2007).

of

Hamad A, Nugud AD, Arnot D et al. Marked seasonality of malaria transmission in two rural sites in eastern Sudan. Acta Trop. 83, 71–82 (2002).

73

Brasseur P, Kouamouo J, Moyou-Somo R, Druilhe P. Multi-drug resistant falciparum malaria in Cameroon in 1987–1988. I. Stable figures of prevalence of chloroquineand quinine-resistant isolates in the original foci. Am. J. Trop. Med. Hyg. 46, 1–7 (1992).

ho

63

Babiker HA, Lines J, Hill WG, Walliker D. Population structure of Plasmodium falciparum in villages with different malaria endemicity in east Africa. Am. J. Trop. Med. Hyg. 56, 141–147 (1997).

72

ut

62

Brockman A, Price RN, van Vugt M et al. Plasmodium falciparum antimalarial drug susceptibility on the north-western border of Thailand during five years of extensive use of artesunate-mefloquine. Trans. R. Soc. Trop. Med. Hyg. 94, 537–544 (2000).

A

61

link. Trends Parasitol. 19, 70–73 (2003).

86

Nash D, Nair S, Mayxay M et al. Selection strength and hitchhiking around two anti-malarial resistance genes. Proc. Biol. Sci. 272, 1153–1161 (2005).

ro

combination on incidence of Plasmodium falciparum malaria and mefloquine resistance in western Thailand: a prospective study. Lancet 356, 297–302 (2000).

rP

Review

Nair S, Nash D, Sudimack D et al. A

87

Roper C, Pearce R, Nair S, Sharp B, Nosten F, Anderson TJ. Intercontinental spread of pyrimethamine-resistant malaria, Science 305, 1124 (2004).

88

Hastings IM. The origins of antimalarial drug resistance. Trends Parasitol. 20, 512–518 (2004).

89

Wang P, Wang Q, Sims PFG, Hyde JE. Characterisation of exogenous folate transport in Plasmodium falciparum. Mol. Biochem. Parasitol. 154, 40–51 (2007).

90

Hayton K, Ranford-Cartwright LC, Walliker D. Sulfadoxine–pyrimethamine resistance in the rodent malaria parasite Plasmodium chabaudi. Antimicrob. Agents Chemother. 46, 2482–2489 (2002).

91

Sirawaraporn W, Sathitkul T, Sirawaraporn R, Yuthavong Y, Santi DV. Antifolate-resistant mutants of Plasmodium falciparum dihydrofolate reductase. Proc. Natl Acad. Sci. USA 94, 1124–1129 (1997).

92

Chen GX, Mueller C, Wendlinger M, Zolg JW. Kinetic and molecular properties of the dihydrofolate reductase from pyrimethamine-sensitive and pyrimethamine-resistant clones of the human malaria parasite Plasmodium falciparum. Mol. Pharmacol. 31, 430–437 (1987).

93

Hyde JE. Exploring the folate pathway in Plasmodium falciparum. Acta Trop. 94, 191–206 (2005).

94

Wang P, Wang Q, Aspinall TV, Sims PF, Hyde JE. Transfection studies to explore essential folate metabolism and antifolate drug synergy in the human malaria parasite Plasmodium falciparum. Mol. Microbiol. 51, 1425–1438 (2004).

Expert Rev. Anti Infect. Ther. 7(5), (2009)

Impaired fitness of drug-resistant malaria parasites

98

99

Wooden JM, Hartwell LH, Vasquez B, Sibley CH. Analysis in yeast of antimalaria drugs that target the dihydrofolate reductase of Plasmodium falciparum. Mol. Biochem. Parasitol. 85, 25–40 (1997). Ahmed A, Das MK, Dev V, Saifi MA, Wajihullah, Sharma YD. Quadruple mutations in dihydrofolate reductase of Plasmodium falciparum isolates from Car Nicobar Island, India. Antimicrob. Agents Chemother. 50, 1546–1549 (2006). Kidgell C, Volkman SK, Daily J et al. A systematic map of genetic variation in Plasmodium falciparum. PLoS Pathog. 2, e57 (2006). Nair S, Miller B, Barends M et al. Adaptive copy number evolution in malaria parasites. Plos Genetics 4(10), e1000243 (2008).

101

Plowe CV, Cortese JF, Djimde A et al. Mutations in Plasmodium falciparum dihydrofolate reductase and dihydropteroate synthase and epidemiologic patterns of pyrimethaminesulfadoxine use and resistance. J. Infect. Dis. 176, 1590–1596 (1997).

106

Brookfield JFY. Genetic redundancy. Adv. Genet. 36, 137–155 (1997).

107

Vavouri T, Semple JI, Lehner B. Widespread conservation of genetic redundancy during a billion years of eukaryotic evolution. Trends Genet. 24, 485–488 (2008)

108

Frank SA. Evolutionary dynamics of redundant regulatory control. J. Theor. Biol. 255, 64–68 (2008).

109

Kacser H, Burns JA. The control of flux. Symp. Soc. Exp. Biol. 32, 65–104 (1973).

110

Kacser H, Burns JA, Fell DA. The control of flux. Biochem. Soc. Trans. 23, 341–366 (1995).

111

Fell DA. Understanding the Control of Metabolism. Portland Press, London, UK (1997).

112

113

114

ut

Sibley CH, Hyde JE, Sims PFG et al. Pyrimethamine-sulfadoxine resistance in Plasmodium falciparum: what next? Trends Parasitol. 17, 582–588 (2001).

A

102

Hastings IM, Donnelly MJ. The impact of antimalarial drug resistance mutations on parasite fitness, and its implications for the evolution of resistance. Drug Resist. Updat. 8, 43–50 (2005).

Hemingway J, Ranson H. Insecticide resistance in insect vectors of human disease. Ann. Rev. Entomol. 45, 371–391 (2000). Wallace B. Polymorphism, population size and genetic load. In: Population Biology and Evolution. Lewontin RC (Ed.). Syracuse University Press, NY, USA 87–108 (1968).

ho

100

105

103

Wellems TE, Plowe CV. Chloroquineresistant malaria. J. Infect. Dis. 184, 770–776 (2001).

104

Plowe CV. Monitoring antimalarial drug

www.expert-reviews.com

115

Parasitol. 33, 1051–1058 (2003). 116

Sinden RE, Dawes EJ, Alavi Y et al. Progression of Plasmodium berghei through Anopheles stephensi is density-dependent. PLoS Pathog. 3, e195 (2007).

117

Talisuna AO, Bloland P, D’Alessandro U. History, dynamics, and public health importance of malaria parasite resistance. Clin. Microbiol. Rev. 17, 235–254 (2004).

118

Boni MF, Smith DL, Laxminarayan R. Benefits of using multiple first-line therapies against malaria. Proc. Natl Acad. Sci. USA 105, 14216–14221 (2008).

119

Hastings IM. Complex dynamics and stability of antimalarial drug resistance. Parasitology 132, 615–624 (2006).

of

97

Nzila A, Ochong E, Nduati E et al. Why has the dihydrofolate reductase 164 mutation not consistently been found in Africa yet. Trans. R. Soc. Trop. Med. Hyg. 99, 341–346 (2005).

resistance: making the most of the tools at hand. J. Exp. Biol. 206, 3745–3752 (2003).

ro

96

Sandefur CI, Wooden JM, Quaye IK, Sirawaraporn W, Sibley CH. Pyrimethamine-resistant dihydrofolate reductase enzymes of Plasmodium falciparum are not enzymatically compromised in vitro. Mol. Biochem. Parasitol. 154, 1–5 (2007).

rP

95

Review

Schoepflin S, Marfurt J, Goroti M, Baisor M, Mueller I, Felger I. Heterogeneous distribution of Plasmodium falciparum drug resistance haplotypes in subsets of the host population. Malar. J. 7, 78 (2008). Talisuna AO, Langi P, Mutabingwa TK et al. Intensity of transmission and spread of gene mutations linked to chloroquine and sulphadoxine–pyrimethamine resistance in falciparum malaria. Int. J.

120

Hartl DL, Clark AG. Principles of Population Genetics (4th edition). Sinauer Associates Inc., USA (2007).

121

Enserink M. Signs of drug resistance rattle experts, trigger bold plan. Science 322, 1776 (2008).

Affiliations •

Hamza A Babiker Biochemistry Department, Faculty of Medicine, Sultan Qaboos University, Alkhod, PO Box 35, Muscat, Oman Tel.: TEL Fax: FAX [email protected]

•

Ian M Hastings Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, UK Tel.: TEL Fax: FAX EMAIL

•

Göte Swedberg Department of Medical Biochemistry and Microbiology, Uppsala University, Sweden Tel.: TEL Fax: FAX EMAIL

13