Background
Mosquitoes from the
Anopheles gambiae complex are considered the main vectors of malaria, responsible for transmitting the protozoan parasite
Plasmodium spp. to humans [
1‐
3]. In Africa, most of the malaria transmission is sustained by members of the
An. gambiae complex, in which
An. gambiae sensu stricto (s.s.) is the most efficient and most studied vector in the African continent [
4‐
6].
In West Africa, 374 million people are at risk of contracting malaria. In this region, malaria cases are mainly caused by
Plasmodium falciparum (almost 100%) [
7]. In the Cabo Verde archipelago, malaria arose in the fifteenth century, during the settlement of the islands, brought by immigrants from the western part of the Africa and Europe [
8]. Currently, malaria in Cabo Verde is considered irregular, with a seasonal and sporadic transmission, low endemicity and quite variable from year to year, with no more than 100 cases per year [
8,
9].
Anopheles arabiensis, a member of the
An. gambiae complex, is the only known species in the Cabo Verde Archipelago, which is associated with the transmission of
Plasmodium. This vector is well distributed in most of the islands of the country, except in Sal and Brava, and it has been responsible for major malaria outbreaks in the archipelago [
10‐
13].
Due to the lack of malaria vaccines, control of this mosquito species has been an important strategy to prevent the disease. The use of long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) is the main component of the strategies to prevent malaria. Currently, five classes of chemical insecticides are approved for using in IRS: pyrethroids (PYRs), organochlorides (OCs), organophosphates (OPs), carbamates (CMs) and pyrroles (e.g. chlorfenapyr) [
14,
15].
In Cabo Verde, control strategies are performed in an integrated manner for both dengue and malaria vectors,
Aedes aegypti and
An. arabiensis, respectively, through the use of the OP larvicide temephos and IRS with the PYR deltamethrin [
16]. In addition, adulterated gasoil and
Gambusia affinis fish have also been used as control tools against mosquito larvae [
17]. It is important to highlight that
Aedes aegypti populations from Cabo Verde already displayed resistance to temephos and deltamethrin [
18,
19]. In this context, it is necessary to investigate insecticide resistance in anopheles populations from Cabo Verde. Resistance to OP and PYR compounds have been associated with two major mechanisms: target-site insensitivity and metabolic detoxification [
20,
21]. In general, for OP insecticides, the target-site insensitivity has been related to alterations in the acetylcholinesterase gene; while for PYR, polymorphisms in the para-gated sodium channel gene (
Nav) have been implicated, which have been named
kdr mutations (knockdown resistance) [
22]. The metabolic resistance is based on the increase of the capacity of metabolizing chemical compounds, and it is generally associated with mutations in genes that codify members of three large families of detoxification enzymes: the cytochrome P450-dependent monooxygenases (MFOs or CYP450s), esterases, and glutathione
S-transferases (GSTs) [
23]. Mutations associated with insecticide resistance, such as the L119F in the
GSTE2 gene, G119S in the de
Ace-
1 gene and the L1014F/L1014S in the
Nav gene, have been reported in several populations of
An. gambiae sensu lato (s.l.) in Africa [
24‐
32].
Cabo Verde is listed as one of the 21 countries in the world that aims to eliminate malaria by 2020 [
13,
33], however, this action may be threatened due to the emergence of a high number of recent malaria cases reported in the country. For this reason, it is urgent to investigate the presence of alleles associated with resistance to chemical insecticides in natural populations of
An. arabiensis circulating in the City of Praia, Santiago Island, Cabo Verde.
Discussion
In the global context, Cabo Verde is among the 21 countries targeting the elimination of malaria by 2020 [
33]. Although different strategies are employed in vector control programs, the use of chemical insecticides is still a key component in the context of malaria elimination.
The present study conducted for the first time a screening of resistance alleles in the malaria vector
An. arabiensis from Cabo Verde. It is important to point out that this type of study had already been conducted in Cabo Verde with populations of
Aedes aegypti resistant to organophosphate and pyrethroids [
18]. Once the vector control strategies are conducted in an integrated manner, targeting both
Anopheles and
Aedes species, the same effect could be expected in the two species, once both have been exposed for the same period.
As confirmed by molecular taxonomic identification,
An. arabiensis is the only species of the
An. gambiae complex identified so far, as described by Cambournac et al. [
11] and Ribeiro et al. [
12]. However, it is necessary to perform an entomological survey with a greater spatial coverage in order to obtain a more accurate distribution panorama of
An. arabiensis in Cabo Verde.
The absence of the two of the main mutations associated with chemical insecticides resistance described in
Anopheles spp.:
Ace-
1R by insensitivity of target site and L119F-
GSTE2 by metabolic resistance, suggests a non-selection of these alleles that affects the functionality of acetylcholinesterase and
GSTE2, respectively, but does not rule out the possibility of loss of susceptibility by alteration of other detoxification systems such as the monooxygenase family, esterases or other enzymes of the glutathione transferases family [
42]. Therefore, biochemical tests in order to evaluate the activity of these detoxification enzymes could be performed in
An. arabiensis populations from Cabo Verde. Concerning GSTs, association between high GST activity and resistance to organophosphates, organochlorines and pyrethroids in many insect species has been reported [
43]. Yahouédo et al. [
30] observed elevated levels of GST in
Anopheles populations resistant to pyrethroid from Benin. Cisse et al. [
44] reported high levels of GSTs conferring resistance to pyrethroid and DDT in
An. gambiae s.l. in Mali.
In the present study, analysis of the
GSTE2 DNA sequences revealed an abundance of polymorphisms including non-synonymous mutations in the coding region of the gene. Future investigations, such as gene expression and molecular
docking, should be carried out to verify if these mutations found in the coding region could play an important role in insecticide resistance [
45]. Previous studies of molecular
docking with AgGSTE2 and DDT have shown that the residues, Glu116, Phe120, Arg112 and Leu36 are considered important components of the AgGSTE2 site-active in the direct interaction with DDT. These residues act as a pocket that interact hydrophobically with DDT [
46,
47]. The mutations observed in the present study do not correspond to any of these sites mentioned previously. On the other hand, two amino acid substitutions observed here (V18A and E19Q) are located in a highly conserved domain (between position 16 and 30 of the encoded protein) among all members of epsilon class GSTs in several species of
Anopheles [
48], which may indicate a selective process representing specific adaptations for this mosquito species. Recently, Mitchell et al. [
49], through molecular modelling, described a variant
GSTE2-114T that is significantly associated with DDT resistance in female mosquitoes of
An. gambiae from West Africa. Pontes et al. [
50] have demonstrated, in a molecular dynamics study, that AgGSTE2-F120L may be associated with DDT resistance in
An. gambiae.
The high frequency of polymorphisms observed in
GSTE2 of the study population may be associated with a great variability of phenotypes. The high genetic diversity for this gene and significant positive value for the Fu and Li test (Table
2) may indicate that the
GSTE2 gene of the population studied may be under selection. On the other hand, the Tajima test was negative, not significant. Mirabello and Conn [
51] also found negative Tajima D, but significant, in populations of
Anopheles darlingi in South America and described this result as a recent expansion in population size after a bottleneck effect. The present data is not consistent with recent expansion; however, our sample size is probably low for analysing the demographic evolution of the
Anopheles population. The selective process could be due to intensive use of insecticides in Praia, particularly in the localities of the study, among which is Fontom, considered as one of the hot spots for malaria transmission in the country and, therefore, receiving, in addition to routine integrated control carried out by the local programme, strengthening interventions made before and after the rainy season, using intra-domiciliar spraying. On the other hand, positive selection may also be associated with adaptation of larval stages to xenobiotics of the environment or other adverse conditions, such as temperature, dryness, among others as described in previous studies [
52‐
54].
Comparing the population of
An. arabiensis in Cabo Verde, based in
GSTE2 gene, with other
Anopheles spp. vectors of the African continent, similar values for the genetic diversity indices were observed, for instance in Cameroon (hd = 0.89; π = 2.9), Mozambique (hd = 0.95; π = 3.6), and Malawi (hd = 0.78; π = 4.4). However, these indices are significantly higher when compared with those observed for populations of
Anopheles funestus in Benin (hd: 0.088; π: 0.09) [
39]. This wide range of genetic diversity for the same gene of detoxification in populations of
Anopheles from the African continent can be explained by the geographical differences that exist among these populations.
Studies performed in different African countries have reported the presence of the G119S mutation in the
Ace-
1 gene in
An. gambiae s.l. populations, with allelic frequency variations ranging from 1 to 75% [
26,
38,
55‐
58]. On the other hand, in some West and Central African countries such as Ghana, Burkina Faso and Cameroon, studies with mosquitoes of the
An. gambiae complex revealed the absence of the G119S mutation in the populations studied [
25,
29,
59]. The absence of the G119S mutation in the present study may indicate that the use of temephos in vector control programmes in Cabo Verde is not selecting alleles associated with target-site resistance to organophosphates in
An. gambiae population.
Here, it was found, for the first time, the presence of the L1014S kdr allele in field populations of
An. arabiensis in Cabo Verde, with a low allele frequency and in most cases in heterozygosis in the individuals evaluated (only one was homozygous). Although bioassays to assess the resistance phenotype were not performed, the presence of this allele suggests that resistance to pyrethroid may arise. The allele found can be quickly selected if no management is done regarding the use of pyrethroids through IRS. Therefore, bioassays are necessary to measure the phenotypic resistance of these mosquitoes to insecticides of the pyrethroids class (deltamethrin) used in vector control in Cabo Verde [
60]. Recently, a brief report based on bioassays with samples of
An. gambiae s.l. collected in Praia revealed tolerance to deltamethrin, corroborating our data [
61]. Further investigation is also needed to determine the geographic distribution of the L1014S kdr allele and to evaluate its association with the use of chemical insecticides in regions where vector control is done with chemical insecticides [
62]. In
An. arabiensis populations, the L1014S allele has been found most frequently in East Africa, however, this mutation has been also reported in other West African countries such as Benin (allelic frequency ranging from 16 to 57%), Burkina Faso (range: 16% to 40%) and Kenya (range: 0.50% to 50%) [
24,
28,
62‐
65]. These findings provide evidence of the spread of the L1014S allele in
An. gambiae s.l. populations in West Africa.
Among the sites investigated in this study, Fontom was the district that presented the highest allelic frequency of L1014S. It is possible that there is an association between this high allelic frequency and the more frequent use of pyrethroids in this specific locality, unfortunately there is no public record of the use of this insecticide by locality. Other studies in African countries with anopheles populations have found a positive association between the high allelic frequency of L1014S and the frequent use of pyrethroids [
66‐
68].
The L1014F allele has been reported in several African countries in mosquitoes of the
An. gambiae complex, mainly in West Africa with allelic frequencies varying from 0.04 to 98% [
69‐
72]. However, in this study this allele was not found in the population of
An. arabiensis analysed. Likewise, in other African countries, such as Kenya and Burkina Faso, the L1014F allele was absent in mosquitoes of the
An. gambiae complex [
28,
64].
The results found in the present study are an alert for the competent authorities of Cabo Verde, since the L1014S allele was detected in the local An. arabiensis population in the archipelago for the first time. Especially at the current moment when the Cape Verdean Ministry of Health has invested in the use of impregnated mosquito nets to contain the malaria outbreak in the country.
There is an urgent need to create measures to avoid allele fixation in the analysed mosquito population. Therefore, the creation of a program to monitor the susceptibility of An. arabiensis becomes extremely important.