Background

Of the five Plasmodium species infecting humans, Plasmodium vivax is the most widely distributed species and the cause of 25-40% of malaria cases worldwide [1], and substantial morbidity associated with vivax malaria has been reported [2-4]. Despite the public health importance, P. vivax malaria has received little attention and limited funds for research and control, since it usually produces less severe symptoms than falciparum malaria [2,5,6]. Current treatment for vivax malaria relies primarily upon two anti-malarial drugs, chloroquine (CQ) and primaquine (PQ), with the latter being the only effective drug against the hypnozoite stage. Indeed, the emergence of drug resistance in P. vivax particularly to the only class of compounds available for killing the dormant liver stage is alarming and of high priority for research [7-10]. It is worth noting that inadequate surveillance tools delayed the detection and containment of CQ-resistant P. falciparum resulting in increased morbidity and mortality. If a repetition is to be avoided, the threat of emerging CQ-resistant P. vivax needs to be acknowledged quickly and widely and substantial resources need to be allocated to validate and standardize tools necessary for characterization of drug-resistant P. vivax [10]. In Indonesia, East Timor and Papua New Guinea, CQ-resistant vivax malaria has already reached an alarming prevalence [11]. Furthermore, P. vivax CQR has occurred in at least three Latin American countries (Guyana, Peru and Brazil) [12]. The four clinical trials carried out in Asia (Thailand and Pakistan) and Africa (Ethiopia), for instance, showed that CQ alone (25 mg/kg over 3 days) is less effective against P. vivax asexual blood stages than CQ (25 mg/kg over 3 days) co-administered with PQ (15 mg of PQ base/day for 14 days) over 28 days of follow-up [9]. To unveil the current knowledge regarding the molecular mechanisms of P. vivax resistance to CQ and the prospects for developing and standardizing reliable molecular markers of drug resistance, Goncalves et al. [13] reviewed the available data by combining published in vivo and in vitro studies.

Unlike in P. falciparum, the molecular mechanism of P. vivax CQR remains elusive [7]. This is because, previous studies focusing on genes known to be main determinants of CQR in P. falciparum have failed to demonstrate a strong correlation between pvcrt-o and pvmdr-1 genotypes and the CQR phenotype in P. vivax. Melo et al. [14], on the other hand, showed the association of expression levels of pvcrt-o and pvmdr-1 with CQR and severe P. vivax malaria, because parasites from patients with CQR presented up to 6.1-fold and 2.4-fold increase in pvcrt-o and pvmdr-1 expression levels, respectively, compared to the susceptible group in the Brazilian Amazon.

Drug resistance in P. vivax is becoming more widespread, hindering management of clinical cases and posing a huge threat to the health of millions of people exposed to the risk of vivax malaria. Analysis of the single nucleotide polymorphisms (SNPs) in drug resistant genes has proved to be useful and important in monitoring drug resistance in malaria endemic countries [15]. Mutations in multidrug resistance 1 (pvmdr-1) and K10 insertion in the pvcrt-o genes have been identified as possible molecular markers of CQR in P. vivax [16,17]. Few data are available on the possible relationship between the pvcrt-o and pvmdr-1 genes and CQR [18]. Nevertheless, there are a number of contradicting reports regarding the association between pvcrt-o and pvmdr-1 polymorphisms and CQR. Some reports suggest the Y976F mutation in pvmdr-1 to be associated with an increase in CQ IC50 value of P. vivax isolates in vitro [19]. Non-synonymous amino acid mutations in codons Y976F and F1076L of the pvmdr-1 have been reported to have correlation with CQR although much work remains to link these mutations irrefutably with CQR [16,18,20]. The role of Y976F mutation in pvmdr-1 gene suggested reduced susceptibility to CQ [20]. Recent experiments have shown that the expression of pvcrt-o in transgenic lines of P. falciparum modulates CQ response [17]. A study by Fernandez-Becerra et al. [21] demonstrated up to 21-fold and up to three-fold increases in transcript levels of pvcrt-o and pvmdr-1, respectively, in severe vivax malaria cases compared to isolates from non-severe vivax malaria patients. Another study in India showed the predominance of the wild-type pvmdr-1 and pvcrt-o alleles [22] although one isolate had the Y976F mutation in the pvmdr-1 gene, which could suggest the beginning of a trend towards decreased CQ sensitivity. In Thailand and Indonesia, where CQR is common, the pvmdr-1 (Y976F and F1076L) polymorphisms were also identified in P. vivax samples [16]. In Latin America, where P. vivax CQR remains relatively uncommon, the Y976F and F1076L polymorphisms are relatively infrequent [23,24].

Presently, Ethiopia maintains a species-specific treatment policy: CQ without PQ is the first-line treatment for P. vivax and artemether-lumefantrine (AL) for P. falciparum. Unlike in many malaria endemic countries in Africa, both P. falciparum and P. vivax substantially contribute to malaria morbidity in Ethiopia in relative proportions of 60 and 40%, respectively [25,26]. In 1996, Ethiopia published its first report of CQR, with 2% (5/255) of study patients on CQ with persistent parasitaemia on day 7 [27] although 13% of treatment failures and subsequent reports CQR have been documented [28-30]. Indeed, data on the presence and prevalence of mutations in pvmdr-1 and pvcrt-o genes are limited in Ethiopia. The study was, therefore, initiated to determine the SNPs in the pvmdr-1 and pvcrt-o genes.

Methods

Study area, samples collection and diagnosis

The samples for this study were collected from West Arsi district from November to December 2012. Malaria transmission is seasonal and unstable in this area. Study participants were patients seeking malaria diagnosis at the Aje Health Centre, located at 0382146.3 E, 071734.2 N and 1,852 m above sea level. Malaria diagnosis was confirmed by microscopy of Giemsa-stained blood films and the species of Plasmodium were recorded. Finger-prick blood samples were collected from patients and used for thick and thin blood film preparation. Slides were considered negative after examination of 100 high-power fields. Patients showing positive results for P. vivax infection were treated with CQ. Blood samples spotted on filter paper were used for molecular analyses.

Amplification and determination of pvmdr-1 and pvcrt-o polymorphisms

DNA was extracted from blood spots on filter paper using Chelex extraction methods as described elsewhere [31]. The pvcrt-o (K10 insertion) and pvmdr-1 (Y976F and F1076L) genes were amplified by nested PCR using gene-specific primers (Table 1). The outer and nested PCR conditions for pvmdr-1 was as follows: 94°C, 2 min; 33 cycles of 94°C, 15 sec; 56°C, 30 sec; 72°C, 1 min; 72°C, 7 min. The outer PCR condition was performed in 94°C, 2 min; 30 cycles of 94°C, 15 sec; 52°C, 30 sec; 72°C, 1 min; 72°C, 7 min, while the nested PCR was performed under the following conditions: 94°C, 2 min; 30 cycles of 94°C, 15 sec; 57°C, 30 sec; 72°C, 1 min; 72°C, 7 min. In both the pvmdr-1 and pvcrt-o loci, the nested forward as well as the reverse sequencing primer were used. PCR amplicons were analysed by nucleotide sequence determination at Uppsala Genome Center. Sequencing reactions were run with AB BigDye Terminator v3.1 and spin-column based clean-up. Sequencing samples were separated by capillary electrophoresis on the ABI3730XL DNA Analyzer (Applied Biosystems).

Table 1 Primers used for amplifications of pvcrt -o and pvmdr -1 marker genes
Full size table

Ethical approval

Study protocol was reviewed and approved by Institutional Review Boards of Aklilu Lemma Institute of Pathobiology, Addis Ababa University and of the Armauer Hansen Research Institute as well as the National Research Ethics Review Committee.

Data analysis

Since this study was a preliminary exploratory study, a power calculation of sample size was not done. Data were entered, validated and analysed in Microsoft Excel 2010. Allele proportions were calculated for codons of interest by dividing the number of samples with a particular allele to the number of samples with an identifiable allele at that position.

Results

The pvmdr-1 gene was successfully amplified and sequenced in 78% (28/36) of the P. vivax isolates. Two pvmdr-1 mutant alleles were identified: Y976F alone and Y976F-F1076L. The prevalence of pvmdr-1 Y976F mutation was 75% (21/28) (Table 2). All (100%) isolates carried the pvmdr-1 F1076L mutation.

Table 2 pvmdr -1 gene mutations and tandem repeat genotypes in pvcrt -o gene
Full size table

The pvcrt-o gene was successfully sequenced in 92% (33/36) of the isolates. Of the 33 samples successfully amplified for the pvcrt-o, the majority of the samples (93.9%) were wild type, without K10 insertion (Table 2). Synonymous mutations or insertions (in introns) were found in 6.1% (2/33) of the isolates.

Discussion

CQ continues to be used for the treatment of P. vivax infection in Ethiopia despite reports of CQR from various studies in the country [28-30,32]. It is, therefore, important to investigate the prevalence of drug-resistance associated markers in P. vivax clinical isolates in this country. In P. falciparum, mutations in the pfcrt and pfmdr-1 genes have been linked to CQR but in P. vivax the picture is still unclear regarding the possible relationship between the pvcrt-o and pvmdr-1 genes and CQR. However, the Y976F substitution in the pvmdr-1 gene is thought to be involved in CQR in P. vivax [19] because the geometric mean 50% inhibitory concentration of CQ was shown to be significantly higher in P. vivax isolates carrying the Y976F mutation than in isolates with the wild-type allele. On the other hand, the ubiquitous presence of Y976F in all patients presenting to a clinic in Papua, where CQ resistance P. vivax is both at high and prevalent, precluded correlation with ex vivo drug susceptibility to CQ [10]. In the Thai isolates, the Y976F substitution was associated with a 1.7-fold higher IC50 to CQ [10]. Unlike the Y976F mutation, Suwanarusk et al. [20] found the pvmdr-1 F1076L mutation in all the isolates (wild type and mutants).

In the present study, 75% of the P. vivax isolates had the Y976F mutation in pvmdr-1. Sequencing results of the pvmdr-1 fragment showed the presence of two non-synonymous mutations at positions 976 and 1076. The Y → F change at codon 976 (TAC → TTC) was observed in 26 (75%) of 28 isolates. The second F → L change (at codon 1076), which was due to a single mutation (TTT → CTT), was observed in 100% of isolates. Whether isolates carrying the pvmdr-1 Y976F mutation responded to CQ treatments differently from those isolates with the wild-type sequence necessitates further in vivo therapeutic efficacy study in Ethiopia, but reports from Indonesia and Thailand suggest this to be the case [19]. The difference in the prevalence of pvmdr-1 Y976F in areas where CQR P. vivax prevails versus CQ remain efficacious may indicate the correlation between CQR and sequence polymorphisms in pvmdr-1. In Papua Indonesia, where CQR P. vivax is present at high prevalence (>65%) and high level [19], the pvmdr-1 Y976F mutation was present in all patients presenting to a clinic. In contrast, the sequence polymorphism in pvmdr-1 conferring Y976F was identified in only 25% of Thai isolates from an area where CQ remains efficacious. Ninety-six percent of Indonesian isolates (where clinical resistance to CQ prevails) had Y976F mutation, compared to 25% of Thai isolates where CQ sensitivity was almost uniform [19]. The fact that that all parasites with the Y976F substitution in Ethiopia also carried the F1076L mutation, as originally described by Brega et al. [22] the F1076L mutation could be a background mutation that precedes the Y976F substitution and could potentially provide an early warning on emerging CQR.

In Ethiopia, CQR P. vivax has been reported from various studies. Given the high prevalence of the Y976F mutations in pvmdr-1 in Southeast Asia where CQR prevails [19], the high prevalence of pvmdr-1 Y976F mutations identified in this study may be associated with the CQ treatment failure reported in Ethiopia. But the exact role of this mutation needs to be determined by a combination of in vitro and clinical observation studies in this country. The fact that all isolates carried the pvmdr-1 F1076L mutation, substitution in this codon may be less involved in the modulation of P. vivax susceptibility to CQ than the pvmdr-1 Y976F mutation given that CQ is still effective and widely used in Ethiopia. Indeed, the presence of pvmdr-1 F1076L mutation in all susceptible and mutant isolates challenged the role of pvmdr-1 polymorphisms in modulating CQ responses in P. vivax [33]. On the other hand, pvmdr-1 polymorphisms have been recently suggested to be associated with CQR in Southeast Asia [19] unlike polymorphisms in pvcrt-o that have not been associated with CQR in P. vivax. The limitation of this study was that it did not determine drug resistance phenotype (either in vivo or in vitro) for the isolates undergoing molecular characterization at the pvmdr-1 and pvcrt-o genes. Indeed, withdrawal of a given drug is recommended when 10% of infections are not responding to treatment, although in practice, governments of poor countries leave it longer [34]. The fact that CQ treatment failure reported earlier in Ethiopia did not exceed the level to withdrawal CQ, periodic assessment of the current status of CQR P. vivax has great public health significance.

Conclusion

Despite the fact that P. vivax accounts for about 40% of malaria cases, little attention has been given to the urgent public health need to detect and to closely monitor the progression of CQ-resistant vivax malaria in the country. This study has observed a high prevalence of the pvmdr-1 976 F allele, which is believed to be associated with CQR in P. vivax. In view of reports from elsewhere, the high prevalence of pvmdr-1 Y976F mutation identified in this study may be associated with the reported CQ treatment failure Ethiopia. However, determination of the exact role of this particular mutation in P. vivax CQ responses, as well as the roles of other identified pvmdr-1 and pvcrt-o gene mutations needs further research, involving a combination of in vitro and clinical observation studies in this country.