Background

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 [1–4]. 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 [5–7], with a polymorphism resulting from the substitution of an asparagine for a tyrosine in amino acid 86 (N86Y) in Plasmodium falciparum [8–14]. 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 [15–19].

Plasmodium falciparum resistance to sulphadoxine and pyrimethamine (SP) is conferred by mutations of the dihydropteroate synthase (pfdhps) and dihydrofolate reductase (pfdhfr) genes, respectively [20–23]. pfdhfr 108N mutation seems to be enough to confer resistance to pyrimethamine [24, 25]. 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 [26–28].

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 [29–34]. 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 [30, 36, 37].

In this work were investigated the frequencies of mutations associated with chloroquine resistance (pfmdr 1 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.

Materials and methods

Sample collection

Blood samples included in this study were collected from children between 1 and 16 years, at the HPDB Hospital in Luanda, Angola, and individually spotted on Whatman n.°4 filter paper, after microscopic confirmation of P. falciparum infection. Parent's informed consent was obtained before inclusion of the blood samples in the study which was reviewed and approved by the Ethical Committees from the Ministry of Health of Angola and of the HPDB. Original study was associated to quinine treatment and published elsewhere [38].

Genetic characterization of the parasites

Parasite DNA was extracted from dried blood spots, using Chelex as described elsewhere [39]. P. falciparum mutations associated with resistance to CQ and SP were typed by PCR-RFLP as described elsewhere [39, 40], primers sequences, amplification cycles and restrictions enzymes are described in table 1 and 2, respectively. In this study, the following codons and polymorphisms were analysed: pfmdr1 86, pfcrt 76; pfdhfr 59 and pfdhps 540. Amplicons and fragments were separated on 2% or 3% agarose gels stained with ethidium bromide and visualized under UV.

Table 1 PCR Programme for the genes pfmdr 1 86, pfcrt 76, pfdhps 540, pfdhfr 59, primer and product size
Full size table
Table 2 Fragments length, clones and enzymes used for digestion of the codons 86 Pfmdr 1, 76 Pfcrt, 540 Pfdhps and 59 Pfdhfr.
Full size table

Statistical analysis

Associations between the different mutations were tested by Fisher's exact test.

Results

Patients and parasites

From the 245 isolates of P. falciparum parasites used in this study, 199 samples were successfully typed by PCR-RFLP for pfmdr 1, for pfcrt 245, for pfdhps 221 and for pfdhfr 224. Mutant alleles were detected in four loci: pfcrt 76T, pfmdr1 86Y, pfdhps 540E and pfdhfr 59R (Figure 1).

Figure 1
figure 1

Agarose gels showing pfmdr 1 86, pfcrt 76, pfdhfr 59 and pfdhps 540 PCR-RFLP products of control and field-collected samples of P. falciparum . pfmdr1: 2-N86; 3-Y86; 4-N/Y86; 5-3D7 (N86); 6-Dd2 (Y86). pfcrt: 2-k76; 3-T76; 4-K/T76; 5-K1 (K76); 6-3D7 (T76). pfdhfr: 2-C59; 3-R59; 4-R/C59; 5-N3 (C59); 6-T9/94 (R59). pfdhps: 2-E540; 3-K540; 4-K/E540; 5-HB3 (E540); 6-K1 (K540). In all gels lanes 1 and 7 shows the molecular weight marker.

Full size image

pfmdr1 and pfcrt

As shown in Table 3, the frequency of the pure mutant allele pfcrt T76 was 93.9%. Only one isolate carried a mixed population KT for this allele (0.41%). The frequency of the mutant pfmdr1 Y86 allele was 61.3% and mixed pfmdr1 N and Y allele was detected in 10.1% of the isolates.

Table 3 Prevalence of mutations conferring resistance to chloroquine and sulphadoxine/pyrimethamine in Plasmodium falciparum isolates from Angola
Full size table

To study the associations between the loci pfcrt K76T and pfmdr1 N86Y, 176 single infection isolates were analysed. Among the 165 isolates with the pfcrt T76 mutation, 115 (70%) also carried the mutant pfmdr1 Y86 allele. Whether or not the mixed infections were excluded from the analysis, the association between the two mutant alleles was significant (P = 0.029, mixed infections excluded).

pfdhps and pfdhfr

One locus of pfdhps (K540E) and one of pfdhfr (C59R) were investigated (Table 3). For the pfdhps 88.3% of the isolates carried the mutant allele E540 and 5.4% carried mixed (K and E) alleles. For pfdhfr 60.6% were wild type (C59) and 18.8% were mixed populations. To study the associations between the pfdhps E540 and pfdhfr R59 loci, only 61 single infection isolates were analysed, and there was no association between pfdhps E540 and pfdhfr R59 (P = 0.159).

Discussion

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 [42–45].

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 [46, 47]. 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 [48, 49]. 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 [37, 50–53]. 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 [54, 55] 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, pfmdr 1 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 [57–59]. 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 [60–62]. 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 [64, 65]). In line with the previous considerations, the reproductive capacity of mutant strains is believed to be jeopardized [64–66]. 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.

Conclusion

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 pfmdr 1, 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.