In the Democratic Republic of East Timor (DRET), chloroquine (CQ) resistance in Plasmodium falciparum was first reported in the 1980s [1, 2]. Despite this, until 1999, CQ plus primaquine continued to be used as the first-line treatment of uncomplicated malaria, with sulphadoxine-pyrimethamine (SP) as second-line, severe malaria being treated with quinine [3].

In 1992, high levels of resistance to chloroquine were reported [4]. In vitro and in vivo resistance to choroquine, amodiaquine and SP was documented in the district of Lospalos, where more than 67% treatment failure to CQ was reported between 1999 and 2000 [3, 5–8]. Consequently, in early 2000, the policy for the first-line treatment of uncomplicated malaria was changed. SP was introduced for the treatment of uncomplicated falciparum malaria and chloroquine was used for the treatment of Plasmodium vivax only. Currently, SP has been replaced by artemether-lumefantrine for treating P. falciparum, but chloroquine is still the recommended treatment for vivax malaria.

Genetic variation associated with both CQ and SP resistance can be monitored with specific molecular markers. The K76T mutation at the pfcrt is considered a reliable genetic marker for CQ resistance. Polymorphisms in pfmdr 1, which encodes the P. falciparum P glycoprotein homologue 1, modulate chloroquine resistance in mutant pfcrt-harboring parasites in vitro [9], although their role in vivo has not been sufficiently substantiated [10].

Molecular mechanisms of antifolate resistance in P. falciparum have been explored in detail [11]. Specific point mutations in the parasite's dihydrofolate reductase (dhfr and dihydropterate synthase (dhps) genes are associated resistance to pyrimethamine-sulphadoxine [12–14]. The first study of SP efficacy for the treatment of uncomplicated falciparum malaria, conducted in 2001, reported 81.6% single pfdhfr 108N or double C59R/S108N mutants, but none of the isolates harboured mutations in dhps, and the drug was confirmed to be efficacious [15].

The aim of the present study was to investigate the proportions and distribution of molecular polymorphisms in the parasite Plasmodium falciparum dihydrofolate reductase (pfdhfr), dihydropterate synthase (pfdhps), chloroquine resistance transporter (pfcrt) and multi-drug resistance (pfmdr 1) genes, from samples collected four to five years after the replacement of CQ by SP as the recommended first-line treatment.

A total of 900 samples were checked for presence of malaria. Mean age of sampled individuals was 21 ± 17.04 (SD) of which 51% and 49% were males and females, respectively. Prevalence of falciparum, vivax and mixed malaria infections as determined by optical microscopy (OM) was underestimated in all cases, with PCR allowing detection of a significantly higher number of cases (X² = 15,02; p = 0,001). Thus, infection prevalence detected by OM vs PCR, for P. falciparum, P. vivax and mixed infections, was 7.5% vs 11.4%, 3.7% vs 6.6% and 0% vs 1.0%, respectively.

More than 97% of the parasite population inspected exhibited the pfdhfr core mutation 108N, and an increase in mutant alleles from the first to second period in two others loci examined (pfdhfr 51 and pfdhfr 59) was detected. All isolates presented the Pfdhfr 108N mutation in combination with dhfr 51I and/or dhfr 59R. In particular, the pfdhfr triple mutation was seen in the great majority of samples (Table 1). There was a significant difference (X² = 31.1; p = 0.001) in the prevalence of the pfdhps 437G mutation which increased from 60% to 100% from the first to second periods, respectively. The quadruple mutation [triple dhfr (51I, 59R and 108N) + dhps 437G ] was found in 82.3% of isolates. None of the samples contained quintuple mutations. There was no significant association between any particular genotype prevalence and patient age or sex.

The present study evaluated mutation prevalence in the Pfcrt, Pfmdr 1, Pfdhfr and Pfdhps genes among natural parasite populations of East Timor. Overall, the frequency of mutant alleles progressively increased from the first to second period, including those suggested to be involved in chloroquine resistance. These data is in contrast with previous studies where the frequency of the Pfcrt 76T mutation decreased progressively with abolition of chloroquine for treatment for P. falciparum malaria [20, 21]. The explanation for the persistence of the K76T mutation in East Timor P. falciparum populations may entail several factors. First, despite the discontinuation of CQ, the drug was still recommended for vivax malaria infections. Vivax malaria accounts for 20 – 40% of all malaria cases [7]. Furthermore, there is the problem of underestimation of mixed infections, as the diagnosis of malaria species in East Timor is made by conventional microscopy alone. Second, anti-malarial drugs are still widely available commercially in East Timor and most people use drugs acquired from parallel markets. The indiscriminate or inappropriate use and unsupervised drugs use for malaria infections is likely to be implicated in the maintenance of CQ-resistant parasites. Third, widespread use of quinine and amodiaquine in East Timor, which have been associated to certain extent to pfcrt 76T and pfmdr 1 86Y mutations [22, 23], may impose positive selection, maintaining CQ resistance. Fourth, there has been extensive population movement between East Timor and West Timor-Indonesia, where malaria is highly endemic and CQ resistance levels are high. Thus, dissemination of resistant genotypes is likely to play an important role in maintaining CQ resistance in the region. Last, the reduced genetic diversity and lower recombination rates in south-east Asian parasites (when compared to Africa) may help maintain predominant genotypes even if mutations carry a fitness cost.

Point mutations in pfmdr 1 N86Y and D1246Y, occasionally cited as potential contributors to chloroquine resistance [24] were also inspected. The pfmdr 1 86Y allele was slightly predominant (56% prevalence) as previously observed both in Africa [25] and in some regions in Indonesia [26].

Since SP replaced CQ as recommended first line drug, the putative efficacy of SP was inferred by inspecting mutations in key codons of P. falciparum dhfr and dhps genes. Results indicated that mutation prevalence in those genes have increased steadily over a short period of time in contrast to an earlier study by [15]. Additionally, all isolates exhibited the pfdhfr 108N mutation in combination with dhfr 51I and/or dhfr 59R mutations, indicative of increased levels of pyrimethamine resistance [24, 27–29]. Also, triple mutants in the Pfdhfr gene (51I+59R+108N) accounted for 85.5% of all isolates further re-enforcing that high levels of pyrimethamine resistance are present [27, 30].

The most common mutations polymorphism in the pfdhps gene was the 437G; only one case presented a 540E mutation. These observations are consistent with pfdhps 437G being the first to emerge as result of pressure by sulpha drugs [11, 31, 32], which confers resistance to sulphadoxine in P. falciparum [17, 33, 34]. It was also verified that a high number isolates (82.3%) carried 4 mutations distributed among the dhfr and dhps genes {quadruple mutation: (51I + 59R + 108N + 437G)}, a pattern inferred in the presence of high levels of resistance to the pyrimethamine and in some cases, clinical resistance to SP [35].

Pfdhfr is present in chromosome 4 and pfdhps lies present in chromosome 8 of P. falciparum. Thus, both genes are physically distant and in conditions of normal transmission any particular haplotypes of each gene should be found randomly in human hosts. However, the present data highlights that resistance-associated alleles encoding both the pfdhfr 51 and 59 mutants are non-randomly associated with pfdhps 437 mutants, therefore indicating that both loci are under strong directional selection by sulphadoxine-pyrimethamine (SP), suggestive of selective advantage conferred by the presence of the two resistant alleles.

This work suggests that SP resistance may already exist in the East Timor, and that the continuous use of the drug will contribute to higher patterns of inefficacy in the treatment of falciparum malaria. Nevertheless, these findings could be complemented with in vivo data as to reflect more closely the therapeutic efficacy of CQ and SP patterns in the epidemiological scenario. Persistence of CQ-resistant mutants even after supposed drug withdrawal re-enforces the need of its use against P. vivax exclusively. The association between pfdhfr 51/pfdhps 437 and pfdhfr 59/pfdhps 437 alleles indicates that these are undergoing concomitant positive selection.