Background
The emergence and spread of CQR
Plasmodium falciparum in malaria-endemic countries has led to alterations in the anti-malarial treatment policy and the introduction of artemisinin-based combination therapies (ACT) for treatment of uncomplicated malaria [
1]. The failure of CQ to clear
P. falciparum asexual parasites in Ethiopia was first reported in 1986 in areas bordering Somalia, Kenya and Sudan. A steady increase in the number of refractory isolates was observed in all malarious areas in Ethiopia and as a result CQR
P. falciparum became a major public health threat in 1990s [
2]. In response to high levels of
P. falciparum resistance to CQ and then sulphadoxine-pyrimethamine (SP), Ethiopia adopted artemether-lumefantrine (AL or Coartem©) as first-line therapy in 2004 [
3‐
5]. According to the current national malaria diagnosis and treatment guidelines, AL and CQ are first-line treatment for uncomplicated
P. falciparum and
Plasmodium vivax infections, respectively. For all clinical infections without laboratory confirmation, AL, which is effective against both
P. falciparum and
P. vivax, is the first-line treatment. The reduced susceptibility of parasite strains to artemisinin necessitates assessment of the current status of CQR in Ethiopia. CQ is a drug of interest for study because of its efficacy, affordability, easy administration, prophylaxis potential, low toxicity and relatively few side effects [
4].
It has been recognized that CQR falciparum malaria is caused by mutations in two genes, the
P. falciparum CQ resistance transporter (
pfcrt) and multidrug resistance transporter-1 (
pfmdr1) however the former is a stronger predictor of CQR than
pfmdr1. Polymorphisms in the
pfcrt gene segregate precisely with two distinct drug response classes, considered either CQS or CQR. All
P. falciparum clinical samples that are resistant to CQ contain the K76T mutation [
6]. Amino acid change from lysine to threonine at position 76 (K76T) appears necessary for the resistance phenotype and is the most reliable molecular marker of resistance of the various
pfcrt mutations. It is unfortunate to notice that strains of
P. falciparum resistant to artemisinin have appeared in the Cambodia–Thailand border region [
7]. While CQS strains are characterized by the CVMNK haplotype, irrespective of geographic origin, two major haplotypes defined by specific mutations at amino acid positions 72-76 of
pfcrt, CVIET and SVMNT, are associated with the geographic origin of CQR [
8]. The former haplotype is predominantly found in Southeast Asia and Africa, whereas the SVMNT haplotype is characteristic of South America, Papua New Guinea [
9], and the Philippines [
10].
A decrease in drug resistant
pfcrt alleles has been reported following discontinuation of drug pressure [
11] in Africa although this process can vary greatly between countries. A long-term decline in the use of CQ can lead to resurgence of drug-sensitive populations of
P. falciparum. This raises the possibility that, in time, the drug could be re-introduced [
12]. Understanding the influence of CQ withdrawal on local malaria parasite populations has great significance. In Malawi, for instance, the frequency of the mutation declined from 85% in 1992 to 13% in 2000 in one study [
13] and from 17% in 1998 to 2% in 2000 in another study [
14] after CQ was withdrawn as the first- line drug. Studies in Kenya [
15] and China [
16] have reported complete and partial recovery of CQS parasites after the use of CQ was abandoned. This recovery was thought to be due to the re-introduction of susceptible parasites harbouring a CQS
pfcrt[
17,
18], although the possibilities of back mutation in
pfcrt at position 76 may not be underestimated. Continued surveillance of markers of resistance to withdrawn drugs has been indicated with the prospect that decreases in resistance may make it possible to reuse this safe and cheap drug [
4,
17]. The overall trend in the reversal of the resistance could be related to a country’s drug policies although data regarding the impact of CQ withdrawal on circulating parasite population is not well studied in Ethiopia.
When advantageous mutations such as drug resistance mutations spread through a population, unrelated markers that flank these loci are carried along through linkage disequilibrium by a “hitch-hiking effect”, thereby removing genetic variation from chromosomal regions surrounding the locus [
19]. Unlike in CQS alleles, there is a marked drop in variability and an increase in linkage disequilibrium with CQR associated alleles in ~40 kb region flanking the
pfcrt gene [
8]. MS repeats, occurring on the average approximately every 1 kb [
20,
21], are common in the
P. falciparum genome and have been recently used to analyse both intra- and inter-population relationships among drug-resistant
P. falciparum strains.
Drug treatment has not only resulted in positive selection at codons conferring resistance but also alleles at loci closely linked to these genes, due to limited recombination in the chromosome regions containing these genes since drug selection was imposed [
22]. If a particular mutation in the target molecule is fixed in the parasite population due to the drug pressure, its flanking MS markers will start showing minimum variation [
23]. Thus, selective sweeps for drug-resistant genotypes may have restricted the genetic diversity of this parasite [
8]. MS markers associated with drug resistance enabled identification of drug resistance origin as well as tracing spread of these mutations [
24]. Drug- resistant genotypes are characterized by reduced diversity around the major resistant alleles [
25]. Thus, the expected heterozygosity (H
e) values of the flanking MSs will be reduced among mutant parasites. Diversity of haplotypes existed in the parasite population before CQR but after the selection, parasites bearing the
pfcrt 76K allele are eliminated and those with
pfcrt 76T survive and expand in the population. Thus, in areas where CQ has been widely used, there will be homogeneity of the loci in this chromosome region of the resistant parasites around the
pfcrt whereas diversity of MS alleles will still be found in sensitive parasites. Despite high level of CQR in Ethiopia, pertinent data regarding the type of mutant
pfcrt genotype are lacking.
Applications of MS markers flanking the genes have revealed that drug resistance appears to have resulted from the migration of limited resistant lineages to many endemic regions however the resistance haplotypes present in Ethiopia is not known. Despite the migration of resistant lineages, multiple resistant lineages independently evolved in several endemic regions. Indeed, the geographic origin and spread of resistance can be estimated by the genotyping of pfcrt at positions 72-76 and MS haplotyping flanking this locus. Although Ethiopia is highly endemic for P. falciparum malaria and CQR P. falciparum isolates occur widely, there is scarcity of data regarding the pfcrt genotypes and diversity among the parasite population. Authors hypothesize that although CQ was replaced by ACT for treatment of P. falciparum malaria in the country, unlike in many African countries, the withdrawal CQ may not result in the reduction of CQR genotype for the fact that it is still the first-line treatment for P. vivax and co-infections are inevitable. The purpose of this study was, therefore, to determine the CQR haplotypes, to unveil the magnitude of P. falciparum CQR genotypes 14 years after its withdrawal and to determine the genetic diversity in CQR alleles in south-central Oromia, Ethiopia.
Discussion
The widespread resistance to CQ and other-antimalarials occurring in Ethiopia has fostered the deployment of ACT for falciparum malaria. ACT (AL/Coartem©), has been the choice of
P. falciparum therapy in Ethiopia since 2004. At the time of CQ withdrawal, the prevalence of the
pfcrt-76 T resistance marker was estimated to have been 100% [
31] in Ethiopia. Meanwhile, declining resistance to CQ observed in
P. falciparum in many malarious regions after its withdrawal suggests a possibility of future reintroduction of the drug.
In this study the prevalence of K76T mutation was 100% although CQ has been withdrawn since 1999 in Ethiopia. The fact that all isolates were mutant mean that reversal from CQR to wild-type has not occurred and the parasites seem under continuous selection pressure of CQ. The high level of
pfcrt K76T mutation observed in this study is in agreement with a study done in a rural hospital in southern Ethiopia [
32]. For various reasons, drug resistance is expected to drop after the removal of selective pressure [
33] although CQR has remained high in the study site (Ethiopia) which is in concordance with a study conducted in Yunnan Province, China [
33]. It is highly likely that the use of CQ as a first-line treatment for
P. vivax, which accounts for approximately 40% of all malaria cases, may have contributed to the persistence of the K76T mutation in Ethiopia. The use of CQ for the treatment of vivax malaria in areas with CQR
P. falciparum could increase the transmission of falciparum malaria and the selection of mutant strains. In areas where
P. falciparum and
P.vivax co-exist, mixed-species infections are common and undoubtedly, CQ selective pressure on
P. falciparum is retained in such areas. Nonetheless, some of these co-infections are often diagnosed as vivax malaria by conventional microscopic examination and treated with CQ, when cryptic CQR
P. falciparum infections emerge after treatment [
34]. For instance, using PCR,
P. falciparum was identified in 18.7% (3/16) and 35.3% (6/14) of RDT-and microscopy-confirmed
P. vivax infections, respectively in the study area (unpublished). The persistent use of CQ for the treatment of
P. vivax is, therefore, responsible for the absence of decline of
P. falciparum CQR in Ethiopia.
This is the first report in Ethiopia, to the best of author’s knowledge, to uncover the presence of a single CQR mutant
pfcrt genotype, CVIET, the most prevalent haplotype in Africa. In this study, the SVMNT mutant haplotype was not identified. Studies have shown that the withdrawal of CQ in Kenya [
15], Malawi [
13], China [
35] and Tanzania [
36], have resulted in the reversion to the ancestral state of CQS from the CVIET haplotype. In contrast, no change in the prevalence of the mutant
pfcrt allele occurred in Gabon [
37,
38] which changed its national treatment guidelines from CQ to the combination of artesunate plus aminoquinoline. But in Ghana [
39] which also changed its national treatment guidelines from CQ to ACT (the combination of artesunate plus aminoquinoline), a decrease in the prevalence of the mutant
pfcrt allele was observed in northern parts of the country. Indeed, it is hardly possible to arrive at a definitive conclusion as to whether the absence of CQ in the field or the use of ACT has accounted for the re-emergence of wild types for
pfcrt[
13,
40]. Kamugisha
et al., [
41] attributed the observed differences in the decline of CQR in Mwanza, Tanzania and Iganga, Uganda to country-specific drug change policies. Since the emergence of a new haplotype can be spontaneous, continuous monitoring is necessary as the SVMNT haplotype initially absent from Africa suddenly appeared in Tanzania in 2004 [
42].
Studies have shown that CQS parasites (
pfcrt 76K) exhibited extensive diversity at all loci while CQR strains (
pfcrt 76T) showed reduced diversity around the
pfcrt gene but high diversity for loci on other chromosomes [
39‐
42]. In this study, the CQR isolates showed reduced diversity with respect to all MS markers (two introns and two flanking). The isolates were 89.9, 100, 88.6 and 100% identical with respect to msint 2 (0 kb), msint 3 (0 kb), mscrt -2 (-2.814 kb) and mscrt -29 (-29.268 kb), respectively. Although some allelic combinations were shared among parasites collected from asymptomatic subjects and clinical samples, there are certain allelic combinations specific to each repertoire isolate types. Of the seven haplotypes identified in the study area, haplotye E1 is the most prevalent allelic type. But whether these allelic combinations could be linked to either symptomatic or asymptomatic infections needs further research. It can be inferred from this study that parasites identified from asymptomatic subjects are more diverse than those present in symptomatic infections (clinical) sample sets. Despite low malaria transmission setting in the study area, asymptomatic parasite carriage is still associated with a higher multiplicity of infection. The number of MS alleles varied among the
pfcrt introns and flanking loci successfully analyzed, with msint 2 displaying the greatest number of alleles (n = 4). On the other hand, msint 3 and mscrt -29, showed no polymorphisms. In this study, the CVIET mutant genotype showed variation in 11.1% (msint 2) and 13.4% (mscrt -2), but the isolates were 100% identical with respect to msint 3 and mscrt -29 allelic sizes. In
P. falciparum strains, very low or no heterozygosity would be expected for msint 2 and msint 3 [
8]. Diversity observed in msint 2 in this study is supported by a study in a Papua New Guinea [
43] that has indicated the continuing evolution of the region’s parasite populations. Given that msint 2 and msint 3 are similar in their length and nucleotide content, it is unclear why the mutant strains show variability in msint 2 but not in msint 3. The possible explanation for the observed reduction in MS allelic diversity in the closest proximity to the
pfcrt allele is the result of strong CQ selection pressure. Because CQ was responsible for homogenizing genetic diversity at polymorphic sites in close physical linkage with CQR-associated alleles [
8].
Genetic diversity of the malaria parasite at a given region is correlated with the local transmission intensity. It has been indicated that drug selection tends to diminish genetic diversity, especially near the resistance-determinant locus [
8]. The low heterozygocity of intronic and flanking MS alleles observed in this study may indicate that parasite population in the study area is genetically homogeneous, which could be the result of the CQ-selective sweep. Although analysis of MS loci around the
pfcrt gene provides suggestive evidence for selection around this gene, the absence of parasites bearing sensitive
pfcrt (CVMNK haplotype) alleles in this population limits the strength of this conclusion. Hence, it was difficult to compare patterns of MS variation in both sensitive and resistant
pfcrt alleles in this study. The other limitation of this study was the fact that other mutations at
pfcrt codons associated with CQR were not examined also absence of mscrt -29 variations limits the conclusive power.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LG, BE, AA and GS designed and were involved in all stages of this study. LG did the sample collection and prepared the draft manuscript. BE and AA coordinated the field work. LG, NE and GS performed molecular analysis and MS genotyping. All authors contributed to interpretation of data, writing and revising the manuscript and have approved the final version.