Malaria is endemic throughout much of the Angolan territory, and is by far the highest cause of morbidity and mortality particularly among children under five years old and pregnant women. Malaria continues to be responsible for 50% of all outpatient attendance and around 25% of all hospital deaths http://www.rollbackmalaria.org/wmr2005/. Recently and according to the Angola Malaria Indicator Survey, 2007, http://www.procaare.org/archive/procaare/200804 these figures have been reduced due to control programmes, but about 35% of all cases and 70% of all deaths reported annually occur in children under five years of age.

The first cases of resistance to chloroquine (CQ) were described in the eighties 1234. As a consequence of increasing rates of clinical resistance to CQ, the protocol at the Pediatric Hospital David Bernardino (HPDB) was changed, in 2006, two combined therapies artesunate with lumefantrine (Coartem^®), and amodiaquine (AQ) with sulphadoxine-pyrimethamine (SP – Fansidar^®) as first-line treatments of uncomplicated malaria.

The pfmdr1 gene (multidrug resistance) has been described as associated with CQ resistance 567, with a polymorphism resulting from the substitution of an asparagine for a tyrosine in amino acid 86 (N86Y) in Plasmodium falciparum 891011121314. CQ resistance is also associated with a mutation in the transporter gene pfcrt where the amino acid substitution at pfcrt codon 76 (K to T) has been shown to have a determinant association with the resistance phenotype 1516171819.

Plasmodium falciparum resistance to sulphadoxine and pyrimethamine (SP) is conferred by mutations of the dihydropteroate synthase (pfdhps) and dihydrofolate reductase (pfdhfr) genes, respectively 20212223. pfdhfr 108N mutation seems to be enough to confer resistance to pyrimethamine 2425. The presence of a mutation at positions 51 (N51I) or 59 (C59R), together with S108N confer a considerable increase in the resistance level to pyrimethamine when compared with mutation S108N by itself 262728.

Mutations in codons 437 (A437G) and 540 (K540E) of P. falciparum pfdhps are associated with resistance to sulphadoxine. The quintuple mutant (triple pfdhfr: 51I, 59R, 108N and double pfdhps: 437G, 540E) is considered as the molecular marker of SP treatment failure 293031323334. Studies of genetic transfection of P. falciparum, confirmed that the amino acid substitution at pfdhfr codon 108 (S→N), increases approximately ten times the resistance to pyrimethamine 35. Epidemiology studies also demonstrated that the presence of mutations 59R in pfdhfr and 540E in pfdhps are significantly associated to resistance as well as to the presence of the other above mentioned mutations 303637.

In this work were investigated the frequencies of mutations associated with chloroquine resistance (pfmdr1 N86Y and pfcrt K76T) and also screened for the two mutations associated to the resistance of P. falciparum to SP as used in populations from high malaria transmission areas.

In vivo resistance to CQ and SP have reached high levels in some regions of Angola, where resistance to these drugs is high (25%) 41. This study reveals a high frequency of drug resistance molecular markers for, CQ (pfcrt T76; 93.9% and pfmdr1 Y86; 61.3%) and SP (9% of bearers with quintuple mutant).

In this study, associations within and between the mutations that confer resistance against CQ and those considered predictive of SP treatment failure were evaluated. Obtained data revealed an association (P = 0.029) between pfcrt T76 (chromosome 7) and pfmdr1 Y86 (chromosome 5) as reported by other authors 42434445.

Presence of mutations R59 and E540 in pfdhfr and pfdhps, respectively, is considered a marker of confirmed resistance to SP and our data indicates that 9% of the children attending the HPDB, were bearers of P. falciparum parasites with the SP quintuple mutant associated usually to treatment failure. These results are in line with others from western Africa, where low prevalence of the quintuple mutant has been observed 4647. Although the predictive value of these markers for SP treatment failure has not been established in this study, these results emphasize the need for close monitoring of mutation prevalence with treatment outcome, since these antimalarials are now used as first line treatment of uncomplicated malaria in children attending the HPDB as well as prophylactic treatment in pregnant women.

Inter- and intragenic association of pfdhfr and pfdhps mutant codons was indirectly proven in other studies where SP resistance was found to be associated with double up to quintuple mutations in both genes 4849. Here, only 9% of the samples, with concurrence of the pfdhfr and pfdhps variants, R59 and E450 simultaneously were observed, which are considered to be predictive of the quintuple mutant (pfdhfr I51, R59, N108, pfdhps G437, E540) associated to SP treatment failure 3750515253. All other samples had presence of one or the other mutation suggesting that fixation of the quintuple mutant was still an on going process.

Association between mutations on pfcrt, pfmdr1, pfdhfr and pfdhps, have been reported from West Africa 5455 but reports on linkage between response to treatment with CQ and SP, in the same patients, is rare. Even though no statistic association was found, 43 double mutants for pfdhfr/pfcrt, 21 for pfdhfr/pfmdr1, 12 pfdhps/pfcrt and 6 pfdhps/pfmdr1 were detected. Regarding the association between mutations conferring resistance to SP and CQ; almost all of the isolates carrying the mutant genotype for pfdhfr (43 out of 46) or pfdhps (12 out of 14) carried at least one of the CQ associated mutations in the genes pfcrt or pfmdr1. This reflects the possible association and accumulation of at least three or four out of 7 mutations (pfcrt T76, pfmdr1 Y86, pfdhfr: N108, I51, R59 and pfdhps: G437, E540) scattered on four different chromosomes and involved in resistance to three different antimalarials, chloroquine, sulfadoxine and pyrimethamine.

As a result of this study it is not possible to comment whether the presence of pfcrt T76 favors the presence of pfdhfr and pfdhps mutations or vice versa because this requires longitudinal and long lasting studies and observations. Despite these difficulties in drawing firm conclusions, several previous findings support the hypothesis of linkage disequilibrium between mutations associated with SP resistance and CQ resistance. In a murine malaria model, CQ resistance could be induced in pyrimethamine-resistant parasites, but not in sensitive ones 56.

In field isolates, several observations indicate that it is more likely to find mutations in pfdhfr and pfdhps in isolates exhibiting already pfcrt T76 mutation than in isolates comprising pfcrt wild type parasites 575859. The reason for this apparent association between resistance to CQ and SP is obscure since both drugs have distinct modes of action and resistance to these is determined by mutations on different chromosomes 606162. The accelerated acquisition of resistance to multiple drugs (ARMD) possibly reflecting a rapid mutator phenotype 63 could be one explanation.

Multiple CQ and SP resistance mutations are thought to have higher fitness cost of the parasite asexual reproduction (reviewed in 6465). In line with the previous considerations, the reproductive capacity of mutant strains is believed to be jeopardized 646566. The results did not reflect these since gametocyte prevalence in infections with mutant-type parasites, was not higher than in infections with wild-type isolates.

This was the first molecular study carried out in this geographical area including a considerable number of samples (245) and focused in the mutations of pfmdr1, pfcrt, pfdhps and pfdhfr genes, strongly associated to CQ and SP resistance. The results of the epidemiological study on prevalence of genotypes associated with drug resistance, carried out at a HPDB in Luanda, showed high presence of CQ resistance markers, while for SP a much lower prevalence was detected. This work can be important to evaluate the implementation of new therapeutics strategies based on combinations that includes SP, like the protocol that are now implemented in HPDB (artesunate combined with SP) as first-line drug for uncomplicated malaria treatment.

The authors declare that they have no competing interests

PF carried out the molecular analyses and drafted this manuscript. FN participated in the design of the study performed the statistical analysis and helped to draft the manuscript. CB carried out the selection of children and sample collection. DL helped in molecular analysis and the draft of the manuscript. LB coordinated sample collection in PHDB as hospital director. VEdR helped in coordination and design of the study. LV coordinated the project and designing of the study. All authors read and approved the final manuscript.