Introduction

The global situation of malaria and its vectors

Malaria is one of the most serious and complex health problems faced by humanity. Besides that, it has become a threat for social and economical development in tropical and subtropical regions, specially given the decrease in work capacity of the affected victims [1]. According to the World Health Organization (WHO), approximately 207 million cases of malaria were reported in 2012, with an estimate of 627,000 deaths, with the highest incidence rates observed in Africa (80%), Asia (15%) and the Americas (14%) [2]. Among the factors contributing to this scenario, it is possible to highlight the absence of an effective antimalarial vaccine, the distribution of drug-resistant Plasmodium, the development of insecticide resistance in vector mosquitoes, as well as ecological, socio-economic and medical-sanitary factors [3, 4]. Mosquito resistance to at least one insecticide used for malaria control has been identified in 64 countries [5]. Malaria vectors are part of the Anopheles genus, including nearly 484 species, distributed in seven subgenera [6], 70 of which showing vectorial competence for human malaria [7], with 41 of them being considered as dominant vector species [8] (Table 1).

Table 1 List of 41 dominant vector species by area
Full size table

Use of insecticides against malaria vectors

The strategic tools to fight malaria are oriented towards two principal directions: (i) prevention, by means of controlling vector mosquitoes; and (ii) case management, through malaria diagnosis and treatment with effective medicines, being the former considered as the most effective [12]. The techniques for controlling vector mosquitoes are didactically classified as: mechanical (elimination of breeding sites), biological (use of predators or parasitoids) or chemical (application of synthetic insecticides) [13, 14]. The development of chemical insecticides that remain active for long periods of time was one of the most relevant breakthroughs of the 20th century [15] and nowadays they still play an important role in the control of disease vectors and plagues in agriculture.

There are four main groups of neurotoxic insecticides permitted to be used for public health purposes, classified according to their chemical nature and mode of action: organochlorines, organophosphates, carbamates and pyrethroids. The first insecticide used against anophelines was the DDT, an organochlorine firstly used in Naples in 1944 against a typhus epidemic [15]. In 1995, WHO proposed the global eradication of malaria based on the spraying of DDT inside the houses. Highly efficient and inexpensive, it was able to decimate populations of vectors on a global scale. However, the development of environmental and sanitary problems, coupled with the emergence of resistance, resulted in the prohibition of the product in many countries [16]. In spite of that, after the “Stockholm Convention on Persistent Organic Pollutants” in 2007, DDT was reestablished in restricted areas with high malaria transmission, such as in African locations [17].

The organophosphates (malathion, temephos, fenitrothion etc.) were developed in the 1940s and have been used ever since as insecticides, herbicides and plant growth regulators. Despite being biodegradable and non-cumulative, they have disadvantages, like chemical instability and high toxicity for vertebrates [18]. The carbamates, also referred to as methylcarbamates for deriving from the methylcarbamic acid [19], are compounds used as insecticides, nematicides and acaricides. They have low environmental persistence and are less toxic to living organisms than organochlorines. Due to their wide use in agriculture, they were incriminated as food, water and air contaminant agents, with adverse effects in humans and other animals [20]. Around 1970, synthetic pyrethroids were released as a class of insecticides considered more efficient and less toxic. These insecticides raised the attention for presenting higher lethal capacity against insects, requiring only small doses of the product for satisfactory effects [21]. Consequently, pyrethroids virtually substituted/supplemented the use of other classes in many pest control areas, representing nearly 23% of the chemical insecticides market, more than one fourth of the world market [22].

Pyrethroids are synthetic analogues of the chrysanthemic acid (pyrethrins I) and pyrethric acid (pyrethrins II) ester insecticides, naturally found in leaves of Chrysanthemum cinerafolis. They are chemically distinguished as type I, compounds that lack an alpha-cyano group, like permethrin, and type 2, with an alpha-cyano group, like deltamethrin [23]. They are biodegradable, non-cumulative insecticides that rarely cause acute intoxication in birds and mammals [24]. Currently, malaria control basically depends upon this insecticide class, which has been widely employed in indoor residual spraying (IRS) and also to control agricultural pests worldwide. Besides, pyrethroid is the only class approved by the World Health Organization Pesticide Scheme (WHOPES) for mosquito net impregnation (Insecticide Treated Net – ITN; Long Lasting Insecticide Treated Net - LLIN) [1, 25, 26].

IRS is a method in which residual insecticides are applied on the surface of walls and ceilings of houses [27]. Based on this strategy it is expected that the mosquitoes, after feeding on blood, rest on these surfaces and remain long enough to absorb a lethal dose of the insecticide. ITN is a mosquito net that repels, incapacitates or kills mosquitoes that come into contact with the insecticide impregnated in the net material, being both a chemical and a physical barrier against insects. There are two ITN categories: conventional nets and LLIN [27, 28].

The initial success of insecticide based strategies caused the optimistic sensation that the elimination of malaria as a public health concern would be possible through the elimination of its vectors. However, these strategies are threatened today, due to the emergence of vector populations resistant to insecticides. Since new classes of alternative, equally interesting insecticides are not yet available on the market, the selection for resistance tends to continue increasing, unless effective management strategies are implemented [29].

Review

Mode of action of pyrethroids

Pyrethroids, such as DDT and its analogues, belong to a group of neurotoxins that share a similar mode of action. They all target NaV, which is present in cells of the central and peripheral nervous systems (neurons, myocytes, endocrine cells and ovaries), changing the kinetics of propagation of nerve impulses [22]. Structurally, NaV is an integral transmembrane protein, composed of four homologous domains (I-IV), each of them composed of six helices (S1-S6) connected by loops. The segments S5, S6 and the S5-S6 P-loops form a central aqueous pore, and the S1-S4 helices of each domain unite to form four independent voltage-sensitive domains [30, 31]. The A. gambiae Na V alpha subunit gene comprises an ORF (Open Reading Frame) with 6,417 nucleotides that encodes 2,139 amino acids, resulting in a protein with a molecular mass of 240 kDa. This gene, located at the para (paralysis) loco of the X chromosome, is composed of 35 exons, including two duplicated exons, and 32 introns, which transcribes for different messenger RNAs (mRNA) through alternate splicing [31].

The effects of pyrethroids are stereospecific and two different NaV binding sites were identified. The first was proposed by O’Reilly et al. [32], in which IIS5 and IIIS6 helices would play an important role in the interaction with the insecticide molecule and the additional link in the IIS4-S5 linker would explain the higher potency of pyrethroids compared with DDT. The second was suggested by Du et al. [33], where the binding site would be a type of “pocket” formed by the IS4-S5 linker and the helices IS5 and IIS6. For both models, the selective effect of the insecticide would be explained by the non-conservation of the amino acids of these regions between arthropods and other animals.

Pyrethroid resistance mechanisms

Insecticide resistance can be defined as the ability of individuals of a species to withstand doses of toxic substances, that would be lethal for most individuals of a population [34]. It is, therefore, a milestone in the change of the genetic composition of a given population, in response to the selection pressure. This is a typical case of Natural Selection, which consists in the increase of the relative frequencies of some “pre-adapted” individuals present in a population, resulting from the constant application of the same chemical product [35]. Intensive and extensive use of chemical insecticides has selected populations resistant to these compounds [36]. The resistance phenomenon has been observed in more than 500 insect species around the world, among which more than 50 are anophelines [37]. According to WHO [5], resistance to at least one insecticide had been identified in 64 malaria-endemic countries. Resistance to pyrethroids seems to be the most widespread. Two main mechanisms are incriminated as responsible for the pyrethroid resistance: metabolic resistance and target-site insensitivity [38, 39].

Metabolic resistance occurs when high activity of one or more enzymes results in a sufficient portion of insecticide being sequestered or detoxified before reaching its target and promoting the desired effect [38]. It occurs due to the increase in the number of available molecules (genetic amplification or hyperactivation of the gene expression) or through mutations in the coding gene portion of the enzyme, producing the more efficient metabolization of the insecticide [37, 40, 41]. This mechanism is highly complex, although recent advances have been characterizing the main enzyme genes responsible for the detoxification, paving the way for the development of molecular markers for the resistance [42]. Three main enzyme superfamilies are involved in the detoxification process: Esterases, Mixed Function Oxidases (MFO, or simply P450) and Glutathione S-Transferases (GST) [37]. Colorimetric biochemical trials are widely employed to detect changes in the activity of detoxification enzymes. In this test, the enzymatic activity of a natural population is compared with the control lineages ones, using specific substrates for each enzymatic family [43].

On a transcriptional level, more recently microarray assays have gained prominence in the investigation of metabolic resistance. In this technique, the detoxification chips (or detox chips) compare the expression of virtually all genes of the families related to the metabolism of insecticides (GSTs, MFOs, Esterases), between resistant and susceptible mosquitoes. In addition to these main families, the expression of other genes are evaluated, such as some related to redox metabolism, involved in the protection against free radicals [44]. The analysis of the gene expression through detox chip in A. gambiae showed high activity of GST genes (GSTE2), P450 (CYP6Z1 and CYP325) and peroxidases in DDT resistant mosquitoes [44]. Genes with anti-oxidizing function (Superoxide dismutase, GST, Peroxidase and P450) were differently expressed in deltamethrin-resistant populations of A. arabiensis in Cameroon [45]. High expression of CYP6P3, a gene of the P450 family, was observed in permethrin-resistant populations of A. gambiae [46]. Differential expression was also observed in A. funestus, whose P450 genes (CYP6P9, CYP6M7) and COI (from the redox system) were more expressive in resistant individuals [47].

Resistance based on target-site insensitivity occurs when there is an alteration in the molecules that directly interact with the insecticide, making it less toxic or inefficient [42, 43, 48]. Since insecticide targets are structural molecules of the nervous system, highly conserved throughout evolution, few alterations are permissive without the loss of their physiological functions. Thus, it is common that the mutations selected for resistance occur at homologous sites among different insect species [49]. Target-site insensitivity is the most understood mechanism, and in many cases is the characteristic attributed to the higher portion of the genetic variation related to resistance [50]. In this sense, molecular diagnoses for detection target-site mutations are part of the strategies to monitor insecticide resistance in many malaria control programs [51].

Kdr mutations as a resistance mechanism

Many studies showed that resistance to the knockdown effect of several insect species is associated with point mutations in the Na V gene. By definition, the knockdown effect is the loss of coordination and paralysis caused by the insecticide, which are often accompanied by spasms and tremors [22]. This resistance mechanism was first observed in the housefly Musca domestica [52], where later it was suggested that the substitution of one amino acid leucine by phenylalanine in the hydrophobic segment IIS6 (L1014F) resulted in a moderate increase of DDT resistance, termed as the kdr mutation (knockdown resistance). In Anopheles the homologous L1014F kdr mutation was first identified in lineages of A. gambiae resistant to pyrethroids [53] and since then it has also been detected in a series of other anophelines [54–59]. Still in the 1014 site, another substitution, leucine by serine (L1014S), was identified in A. gambiae, also associated with the kdr phenotype [60]. The mutations L1014F and L1014S were first observed in populations of West and East Africa, respectively. Therefore, the former is sometimes referred to as kdr-w (kdr-west), and the latter, as kdr-e (kdr-east) [61]. In any case, it is noticeable that the distribution of these mutations is strongly related to sibling species of the Anopheles gambiae complex [62].

In Asian A. sinensis populations, in addition to the L1014F/S substitutions, the mutations L1014C and L1014W were reported, changing the amino acid leucine to cysteine and to tryptophan, respectively. Additionally, in the site immediately before the one of the classical kdr mutation, an N1013S substitution occurs, changing the amino acid asparagine to serine [63, 64]. In Indian A. culicifacies populations, also in addition to the L1014F/S substitutions, a new mutation in the site 1010 was described, substituting valine by leucine (V1010L) [65].

Another mutation in the NaV of Musca domestica, which substitutes methionine by threonine in 918 position, corresponding to the loop between IIS4-S5 segments, synergic to the classical L1014F mutation, was associated with high levels of DDT and pyrethroid resistance, thus being referred to as super kdr [66]. An analogous situation was observed in other insect species, such as in the horn fly Haematobia irritans[67], green peach aphid Myzus persicae [68, 69], onion thrips Thrips tabaci [70] and in the moth Tuta absoluta [71]. However, there are still no records of homologous substitutions in anophelines.

Based on the current molecular techniques, it was possible to identify and map the distribution of kdr mutations among a wide range of Anopheles species around the world. Since it was first described in 1998 [53], the identification of changes in the Na V gene in the Anopheles genus has been widely monitored, in a way that we were able to recorded about 98 references published until the end of 2013 (Table 2).

Table 2 Anopheline species with kdr mutations detected
Full size table

So far, Na V mutations were described in at least 13 different anophelines. A. gambiae was the most studied (62 records), showing three mutational variants (L1014F, L1014S and N1575Y), detected in 19 out of 54 countries in Africa (Figure 1). Following, the African A. arabiensis presented 17 records, showing two variants (L1014F and L1014S) detected in seven countries. A. sinensis was the third one, with six records. Surprisingly, it showed the highest number of kdr variants (L1014F, L1014S, L1014C, L1014W and N1013S), distributed in five Asian countries, mostly detected in China. According to Kang et al. [142], this fact is related to the high population size and to the wide geographical distribution of the species, which tends to increase the genetic variability.

Figure 1
figure 1

Distribution of kdr mutations in Anopheles mosquitoes around the world: (a) America, (b) Africa, (c) Asia.

Full size image

Among other Asian species, A. stephensi showed three records of two variants (L1014F and L1014S), detected in Dubai and India. A.subpictus (L1014F), A. culicifacies (L1014F, L1014S and V1010L) and A. vagus (L1014F) showed two records; while A. sacharovi (L1014F/L1014S), A. sundaicus (L1014F), A. aconitus (L1014F), A. paraliae (L1014S) and A. peditaeniatus (L1014F/L1014S) had just one record. The presence of kdr mutations in the Americas was observed only in A. albimanus, for the variants L1014F and L1014C in populations from Mexico, Nicaragua and Costa Rica (Table 2).

A survey on the geographical distribution of kdr mutations in African populations of A. gambiae, conducted by Pinto et al. [61], detected the presence of the variant L1014F in west countries (kdr-w), from Nigeria to Senegal, the presence of L1014S (kdr-e) in the East (Kenya), and both mutations occurring in the Midwest, comprising Angola, Gabon, Equatorial Guinea and Cameroon. This same distribution pattern was reported one year later by Santolamazza et al. [62]. The occurrence of both mutations is currently found, sympatrically, in several African countries. Exceptions were Niger, Ethiopia, Chad and Liberia, which reported the presence of L1014F only, and Burundi and Kenya with L1014S only (Figure 1).

It is noteworthy that the A. gambiae complex is composed of seven sibling species: A. gambiae s.s., A. arabiensis, A. melas, A. merus, Anopheles quadriannulatus species A, A. quadriannulatus species B and Anopheles bwambae. They are morphologically indistinguishable, however, they can be classified according to fixed and polymorphic chromosomal inversions [152]. The classical molecular forms are Savannah, Mopti, Bamako, Forest and Bissau, according to paracentric inversions of the second chromosome of A. gambiae s.s. [153]. The mutation L1014F was firstly described in the Savannah form of A. gambiae populations, also known as S form, and until mid-1999 this mutation had not occurred in sympatry with the Mopti form (M form) [154]. However, later studies identified its presence also in the M form, possibly resulting from genetic introgression from the S form [81, 155]. Introgression was also suggested by Tripet et al. [114] when the kdr allele was detected in the Bamako form. On the other hand, a new independent mutational event could explain the emergence of the kdr mutation in A. arabiensis [135].

Despite 15 years of research, some doubts still arise with respect to the kdr mutations, especially if they are indeed correlated with the resistant phenotype. One of the techniques adopted to test this association is the employment of bioassays with insecticides (WHO cones, bottle test, ITN, LLIN) followed by the genotyping of kdr mutation between dead and surviving mosquitoes after the test. In other words, it is aimed to test whether the mutation frequency is higher among resistant than the susceptible individuals. In our survey, out of the 98 studies here considered, 63 (64.3%), conducted bioassay followed by genotyping, correlating the mutation with insecticide resistance. Among them six detected the involvement of more than one mechanism of resistance (target site and metabolic alterations) [78, 82, 88, 100, 139, 140] and two only suggested their occurrence [123, 129]. On the other hand, six studies (9.5%) did not associate the occurrence of NaV mutations with knockdown resistance [58, 107, 119, 131, 136, 143]. In these cases, the lack of a “mutation versus resistance” association was suggested due to low sample size [107], mutation similarly distributed between dead and surviving individuals in the insecticide bioassay [58, 131, 136, 143] or mutation among susceptible individuals [119]. Nevertheless, in this last example the authors recognized that the bioassays were performed outside the WHO recommended standards. Lastly, in 28 studies (28.6%) only the genotyping of field samples was performed, considering the presence of the mutation as enough evidence for resistance.

Association between ITN and kdr mutation

The use of ITNs/LLINs treated with pyrethroids is an important tool to reduce morbidity and mortality caused by malaria [26]. According to a survey performed by Lengeler et al. [156], the implementation of this strategy in Sub-Saharan Africa, between 1986 and 2003, was able to reduce morbidity by 50% and the infant mortality by 17%. In Kenya, for instance, the employment of ITNs was able to prevent infant mortality in an area with high malaria transmission [25]. However, the maintenance of this efficiency is still a controversial issue nowadays, given the occurrence of highly resistant anopheline populations. There are several records indicating good results of pyrethroid treated materials where kdr mutation had been identified, such as in Nigeria [157], where the LLINs were efficient at killing or reducing the blood feeding of A. gambiae, Mali [115], Benin [108, 158] and Uganda [124]. On the other hand, a reduction in the susceptibility of A. gambiae populations subjected to ITNs was observed in Uganda [159]. Besides that, increases in kdr frequency were evidenced for this same species after the distribution of LLINs in Kenya [101], Niger [130], Senegal [122] and Benin [160].

The most recent update of WHOPES continues indicating only pyrethroids (deltamethrin, alphacypermethrin, permethrin and a combination of deltamethrin or permethrin and piperonyl butoxide – PBO) for LLINs [161]. However, given the possibility of loss of effectiveness caused by resistance, the development of mosquito nets impregnated with other classes of insecticides is a promising alternative. A study conducted with mosquito nets impregnated with chlorpyrifos-methyl (organophosphate) and lambdacyalothrin (pyrethroid), showed that, alone or combined, they were efficient at killing or reducing blood feeding of A. gambiae from the Ivory Coast, even in areas with high kdr and ace-IR mutation frequencies. This ace-1R mutant allele belongs to the acetylcholinesterase gene, conferring resistance to organophosphates [162].

Association between Plasmodium infection and insecticide resistance

Regardless of the extensive literature concerning kdr mutations and their association with resistance to insecticides, few reports have presented their impact on malaria transmission dynamics, i.e., on the ability of mosquitoes to transmit malaria. Infection rate and oocyst burden are two of the five factors that determine the vectorial capacity of mosquitoes [163]. The response to Plasmodium exposure in vectors is modulated by the mosquito’s innate immune system. In A. gambiae, for example, changes in its global gene expression patterns are expressed upon Plasmodium infection [164]. Exposure to pyrethroids, in turn, induces metabolic changes that alters the immune response [165] and may therefore affect the outcome of Plasmodium infection.

An insecticide susceptible strain of A. funestus showed greater ability to become infected with Plasmodium berghei than its resistant counterpart [166]. In A. gambiae, infection with this same parasite increased the expression level of CYP6M2, a gene related with metabolic resistance [164]. In relation to the possible impacts of kdr mutation on vector competence, few records are available and are sometimes conflicting. For instance, neither positive or negative correlation was found between the occurrence of kdr and ace-1R alleles with infection of Plasmodium falciparum in A. gambiae natural populations from Benin [109].

Other studies, however, showed that the presence of both resistant alleles could be associated with increased prevalence of Plasmodium infection in an A. gambiae resistant strain. Additionally, individuals carrying the kdr mutation had increased prevalence of sporozoites, which is likely to impact on parasite transmission [167]. Given the dissemination of kdr mutation in natural populations, similar studies should be conducted in order to better understand the impact of insecticide resistance on vector competence.

Molecular tools for KDR mutation diagnosis

The resistance phenomenon can be studied on many levels, from biological assays in order to evaluate the susceptibility/resistance status to biochemical and molecular characterizations able to infer the mechanisms and effective genes selected for resistance. Currently, the development of tools for genetic screening of natural populations on a large scale, are aimed to predict the predisposition of those populations to develop insecticide resistance.

Thus, the identification of genetic markers associated with resistance were included in the priorities of the WHO Global Plan for Insecticide Resistance Management (GPIRM) [5]. In this sense, the identification of kdr genetic markers truly associated with pyrethroid resistance, as well as the improvement of existent diagnostic assays are constantly in the course of studies in this field. DNA based genotyping techniques have as main advantages the high sensitivity and the capacity to distinguish between homo and heterozygous individuals [37]. The principal methods employed in the detection of kdr mutations are listed in Table 3, with emphasis on the equipment required for each technique.

Table 3 Molecular methods used for detecting kdr mutations
Full size table

Strategies for managing resistance

The evolution of insecticide resistance has become a great threat to chemical products-based malaria control programs due to the strong selection pressure placed on resistance genes [5]. Therefore, strategies for managing resistance to minimize operational obstacles to the use of a given product have gained prominence on the world stage. The resistance management strategies are divided into three groups: management by moderation, management by saturation and management by multiple attack [175].

Management by moderation aims to reduce the selection pressure to conserve susceptible individuals of a given population, by the use of lower dosages of insecticides, higher treatment thresholds, chemicals with shorter residual activity and maintaining unsprayed areas as refuges for susceptible individuals [176]. Even though, peculiarities have to be considered. For instance, a study evaluating the effects of sublethal doses of permethrin in an A. stephensi strain showed that lower concentrations were more efficient in increasing the mortality rates [177]. Concerning refuges, it is important to maintain susceptible alleles in a population, mainly in the case of resistant alleles, which carry a fitness cost. However, resistant alleles can also invade untreated areas. This was the case observed in a survey conducted in populations of A. gambiae from Burundi, where high frequencies of kdr allele were detected in unsprayed areas, due to migration [128].

Management by saturation involves methods that overcome resistance mechanisms present in the insect, by the use of high rates of insecticides, that should kill even resistant individuals, or by the use of chemical synergists [21]. For example, the evaluation of the dosage-dependent effect of permethrin-treated nets in experimental hut trials from Benin showed that nets treated with higher permethrin concentrations provided better blood feeding prevention against pyrethroid-resistant A. gambiae [158]. Similar efficiency against pyrethroid-resistant A. gambiae populations were observed in a net impregnated with deltamethrin-pyperonil butoxide combination [157, 178].

Finally, the management by multiple attacks involves either mixtures or rotations of insecticides to avoid resistance. This method is based on the concept that insects resistant to one insecticide will be killed by the other component of the mixture and that few insects will be resistant to the entire mixture [176]. A combination of IRS with chlorfenapyr and LLIN impregnated with deltamethrin, in an experimental hut trial from Benin, was effective to provide additional level of transmission control and personal protection against pyrethroid-resistant A. gambiae [108]. Similar results were obtained by the use of mosquito nets impregnated with chlorpyrifos-methyl and lambdacyalothrin against A. gambiae from Ivory Coast [162].

Conclusions

After 15 years of intense research, kdr mutations were recorded in 13 anopheline species, in natural populations from three continents, revealing the preponderance of this phenomenon in the process of resistance to pyrethroid insecticides, either alone or combined with other mechanisms (e.g., metabolic resistance). These alterations emerged in different species as well as within populations of the same species, and are spreading quickly, given the strong selection pressure exerted by the pyrethroids. Although compounds with new modes of action, such as neonicotinoids and pyrroles, have been introduced in public health, they are still not indicated for IRS and ITN, for instance. The availability of a new generation of environmentally friendly compounds may take as long as the implementation of advanced strategies, likewise, the use of genetically modified mosquitoes. Therefore, the use of pyrethroids has to be severely monitored in order to try to maximize their effectiveness.