Evaluation of prevalence's of pfdhfr and pfdhps mutations in Angola

According to the Angolan National Malaria Control Programme (NMCP), malaria is the major cause of morbidity and mortality in Angola, with four million clinical cases and 20 thousand deaths reported in 2005, accounting for 35% of the overall mortality in children under five years old and 25% of the maternal deaths [1‐3]. Malaria is endemic throughout the Angolan territory, Plasmodium falciparum being the predominant species [4]. Due to the high prevalence of P. falciparum strains resistant to chloroquine [5‐8], therapeutic regimens for treatment of uncomplicated P. falciparum infection were changed in 2006 [9] and, currently, the first-line treatment for uncomplicated malaria is Coartem^® (artemether-lumefantrine) followed by the amodiaquine-artesunate alternative therapy.

Additionally, in all Angolan endemic areas the strategy to protect mothers during their pregnancy includes the use of an intermittent preventive treatment (IPT) [10, 11]. This intervention has been introduced in Angola since 2006, using sulphadoxine/pyrimethamine (SP) at the second trimester of pregnancy.

In other African countries, SP - ITP has been also introduced in children as a control measure to reduce malaria morbidity and mortality in this most vulnerable population [12] and has been evaluated in a number of clinical trials in these countries, with success [12‐17]. Thus, Angolan NMCP intends in the near future to introduce this control measure in Angola. However, due to the wide use of SP combination in this country together to reported cases of SP treatment failure, it was decided to obtain further information about SP resistance in Angola, during a surveillance study carried out in 2007, before the introduction of such a control measure.

It is well known that mutations at the dihydropteroate synthase (pfdhps) and dihydrofolate reductase (pfdhfr) genes are associated with resistance of P. falciparum to SP, respectively [18‐21]. In pfdhfr, point mutations at positions 51, 59, 108, and 164 are associated with pyrimethamine resistance [22, 23]. Similarly, mutations in codons 437 (437G) and 540 (540E) of pfdhps are associated with resistance to sulphadoxine [24‐29].

Thus, to determine the polymorphism of pfdhps and pfdhfr genes, infected blood samples were collected in different representative endemic regions of the whole country (Uíge, Kwanza Norte, Malanje, Cabinda and Huambo) and the prevalence of five mutations of the pfdhfr (N51I, C59R and S108N) and pfdhps (A437G and K540E) genes was investigated.

For pfdhfr gene, all 452 samples were successfully characterized by PCR-RFLP for 51, 423 and 59 codons and 430 samples for codon 108. Regarding pfdhps gene, 438 samples were characterized for codon 437 and 448 samples for codon 540. The analysis of pfdhfr showed that 90,3% (408 out 452) of the isolates carried the mutant allele 51I, while 7,5% (34 out 452) carried a mixed infection (N and I); for 59 codon 51% (213 out 423) were wild type (C59), 29,2% (122 out 423) were mixed infections (C and R) and 19,9% (83 out 423) carried the mutant allele 59R. Concerning the pfdhfr gene codon 108, 99,1% (426 out 430) of isolates harbored the mutant allele (N). For pfdhps 83,1% (364 out 438) were mutant type (437G), 11% (48 out 438) were mixed populations and 87% (390 out 448) of studied isolates were wild type for codon 540 (K) (Table 1).

A mixture of infections with wild-type and mutant alleles was also observed. These mixed infections were seen for pfdhfr gene in positions 51 (34/452), 59 (122/416) and 108 (1/430), and in pfdhps gene in positions 437 (45/438) and 540 (29/441). All mixed infections were excluded from subsequent analysis. Therefore, successful characterization of all five molecular markers was obtained only in 241 samples out of a total of 452.

Only 25% (72) of the 400 isolates which were successfully characterized for the studied pfdhfr codons harboured the triple pfdhfr 51-59-108 mutations and one isolate carried the combination of three wild type codons. The more frequent pfdhfr genotype was ICN, which was found in 200 isolates (Table 2). Regarding the pfdhps genotype GK, double mutation (437 and 540 codons) showed the highest frequency (91,1%). 225 isolates harboured mutations on position 108 of pfdhfr gene and 437 of pfdhps gene that were reported as the initial mutations for pyrimethamine and sulphadoxine resistance, respectively [23‐36].

In pfdhfr gene, five mutant genotypes, NCN, NRN, ICN, IRN and NCS (amino acids at positions 51, 59 and 108) were found confirming the major diversity of this gene (Table 2). Among the studied isolates, 74% were double mutants (ICN or NRN), most of them being type ICN, and the triple mutant IRN was detected in 25% of the samples. Only one isolate was a single mutant (ICS). In pfdhps, three allele combinations GK, GE and AK (amino acids at positions 437 and 540) were detected nearly 3% being the double mutant GE and 91% of the isolates were GK and 6% were wild-type (AK). Considering the two studied genes, 12 different genotype combinations were found: NRN GK, ICN GK, IRN GK, IRN GE, ICN AK, IRN AK, ICN GE, NCN AK, NRN AK, ICS GK, NCN GK and NCS GK (51, 59 and 108 for pfdhfr gene and 437 and 540 for pfdhps gene). From a total of 241 isolates, 63% were ICN GK, 25% IRN GK, 3,7% were ICN AK, ICN GE. NRN GK and IRN AK were detected with same frequency of the 1,7%, 1,2% were NRN GK and ICN GE, all other combinations were found with a very low frequency (Table 2).

In a comparison evaluation between provinces, the same pattern was found except for Cabinda, where the most frequent genotype was IRN (Tables 1 and 2).This province is geographically separated from the rest of the country.

The monitoring of SP resistance is relevant in order to guide national malaria treatment policies before introduction of SP as IPT in children at Angola. In this light, this study was designed to assess the pfdhfr and pfdhps mutations associated with SP chemoresistance. For this purpose, the mutations at pfdhfr (N 51I, C 59R and S 108N) and pfdhps (A 437G and K 540E) genes, considered predictive of SP treatment failure [25‐27], were assessed in five provinces of Angola. Four of them - Uíge, Kwanza Norte, Malanje and Cabinda - are hyperendemic areas, whereas Huambo, is a mesoendemic stable. It this way, it was observed the presence of the mutations 51I and 59R jointly with 108N, which enhances the level of resistance to pyrimethamine when compared with single mutation in codon 108 [35‐37]. Similarly, mutations in 437 (437G) and 540 (540E) pfdhps codons are associated with resistance to sulphadoxine [24‐29]. In pfdhfr gene, the mutation at position 108 (S108N), which is believed to be the initial mutation causing pyrimethamine resistance, was observed in almost all isolates successfully characterized for this codon (426/430) (Table 1). In pfdhps gene, among the 438 characterized isolates, 364 presented the mutant allele at position 437, which has been reported as the initial mutation for sulphadoxine resistance in many endemic regions.

From 430 P. falciparum isolates characterized in this study, 99,1% carried the pfdhfr 108N mutation. The results also showed that 27% of P. falciparum isolates presented double mutations at codons 59 and 108, indicating the development of resistance against antifolates in Angola. Another double mutant (51I 108N) was observed in 96,7% of Angolan P. falciparum isolates; these results were consistent with the results obtained in a similar study carried out at Uíge province in 2009 [38].

A 25% of pfdhfr triple mutant prevalence (51I/59R/108N) was also noticed. The prevalence of these mutations in association with high prevalence of mutation at position 437G in pfdhps (83,1%) may indicate that these P. falciparum parasite populations have the potential to evolve into pfdhfr/pfdhps quintuple mutant in the near future, a mutant that is considered a molecular marker of SP treatment failure [25‐27].

In addition, the simultaneously presence of 59R pfdhfr and 540E pfdhps variants is considered predictive of the presence of quintuple mutant (pfdhfr 51I, 59R, 108N, pfdhps 437G, 540E) [25‐40]. In fact, despite high frequencies of mutations in this report, only one isolate was found harbouring the quintuple mutant associated with high level of resistance to SP. This finding is in accordance with the results obtained in other similar studies carried out in Angola, Republic of Congo and Gabon [31, 28, 29].

The predominant pfdhfr haplotype in the present work was 51I 59C 108N (71,9%), corroborating the results reported with Brazilian samples by Gama and collaborators [41]. The triple mutant 51I 59R 108N (24,8%) was of low prevalence in the examined isolates, similarly to previous data reported in Sri Lanka [42] and Papua New Guinea [43], but different from the isolates from Malaysia [44], Brazil [45] and India [46] where this triple mutant was the predominant haplotype. In Africa, the Republic of Congo [28] and Gabon [29] also shows differences when compared with these Angolan data, except in the province of Cabinda. The differences found between Cabinda isolates and the rest of studied isolates may be due to geographical location of this province and its proximity to the neighboring countries Gabon and Congo and the movement of people between these regions. Most of the P. falciparum isolates (91,1%) revealed the haplotype 437G 540K for pfdhps gene with low prevalence of mutation at codon 540 (3,2%).

The triple pfdhfr/pfdhps (59R 108N/437G) mutant haplotype was found in 27% of isolates, the 51I 108N/437G was the mutant haplotype more prevalent in studied isolates.

The results obtained in the present study are in line with those obtained with Iranian isolates where 51I mutation seems be a good molecular marker for the triple mutant pfdhfr/pfdhps[47]. By the other hand, theses results seem to be in contrast with the data obtained in a study carried out in Mozambique which claimed that the mutations at codon 59 in pfdhfr and codon 437 in pfdhps were enough to predict SP treatment failure [48], as well as in Burkina Faso [49] where the results showed that pfdhfr 59R is more relevant than the 51I as a marker of SP treatment failure.

Finally, this is the first molecular study carried out in Angola including a large number of samples (452) from five different provinces of the country, and where five mutations of pfdhps and pfdhfr genes, predictive of SP therapeutic failure were screening showing high frequencies of 51I, 108N pfdhfr alleles, and 437G of pfdhps gene with an almost absence of the quintuple mutation for SP.

The high frequencies of mutations and haplotypes linked to antimalarial treatment failure (for example the 88% frequency of isolates carrying pfdhfr 51I 108N - pfdhps 437G triple mutant allele, critical to SP resistance) herein described, highlight the need to revaluate the strategy of SP introduction as an IPT in children as well as the current use of SP for pregnant women IPT purposes, in Angola.