**3. Evolution of insecticide resistance**

concomitantly selected mutations would be necessary in *Aedes* mosquitoes, an unlikely

Similarly, several mutations associated with PY and DDT resistance are present in distinct insect orders: the *kdr* mutations, that impair the *knockdown effect* provoked by those insecticides. The most common *kdr* (*knockdown resistance*) mutation is a leucine-to-phenylalanine substitu‐

also found (reviews presented in Rinkevich et al., 2013 [15]). Several PY-resistant populations of major arthropod pests and disease vectors were found harboring *kdr* mutations. In this sense, for diagnostic purposes, different well-established tools for *kdr* genotyping have been imple‐ mented, specific for an increasing number of insect species. This allows a rapid and accurate

The recent commercially introduced SP insecticides, which target the nicotinic acetylcholine receptors (nAChRs) [17], have been used for crop protection, animal health, and against human disease vectors. Three formulations of SP were approved by WHOPES for use in drinking water, increasing the chemical arsenal against mosquitoes [4]. However, resistance to this class of insecticides was already detected in a variety of insect species. A target-site point mutation (glycine-to-glutamate substitution G275E), for example, was identified in the nAChR of a Western flower thrips (*Frankliniella occidentalis*) in association with SP resistance [18]. Besides this single amino acid substitution, alternative splicing in the nAChRα 6 subunit seemed to be the mechanism selected in an SP-resistant population of the diamondback moth *Plutella*

As exemplified above, mutations selected for resistance in the molecular targets of insecticides generally share homologous sites among different insects. These molecules are components of the nervous system, which are highly conserved among animals. Therefore, it is expected that few mutations can be maintained without impairing the essential physiological role of that molecule [20]. Target-site-resistant alleles are increasing in frequency and rapidly spreading, as well-recorded for malaria and dengue vectors. An interactive compilation of these data, organized in time and space scales, can be currently accessed on two distinct online platforms: IR Mapper (http://www.irmapper.com) and Popbio (https://www.vectorbase.org/

Detoxifying enzymes are naturally present in living organisms with a protective function against potential damages caused by xenobiotics and endogenous metabolites. In many cases, insecticide resistance occurs due to an increased activity of such enzymes, a mechanism known as metabolic resistance. In general, this mechanism is related with the intense use of insecti‐ cides. However, other toxic compounds, such as chemical pollutants and plant toxins can also select for metabolic resistance mechanisms in insect populations. In this sense, different xenobiotics present in the environment are probably related, at least in part, with a preadap‐ tation for insecticide resistance in disease vector and agricultural pests [21, 22]. Basically, xenobiotics pass through a series of enzymatic steps that transform them in polar substances,

3 In the case of the voltage gated sodium channel (NaV), the *M domestica* aminoacid sequence is most commonly taken

access of the genetic background for PY resistance in natural populations [16].

, although serine, histidine, cysteine, and tryptophan replacements are

situation referred to as codon constraint [14].

tion in the 1014 codon3

246 Insecticides Resistance

*xylostella* [19].

popbio/).

as reference.

Insecticide resistance is an adaptive trait in which a set of genes are favorably selected to maintain the insect alive and able to reproduce under an environment exposed to pesticides. After being introduced, insecticides gradually eliminate the susceptible specimens, usually found at higher frequencies within populations. By contrast, harbors of resistant alleles, supposedly rare in the population, increase their frequencies along the time of continuous pesticide application. The importance of resistance alleles occurring prior to insecticide employment has been discussed since the 1950s with the works of Crown [38] and more recently incremented on ffrench-Constant's reviews [39, 40]. If resistance mechanisms hold elevated fitness cost in absence of insecticide (as discussed subsequently), the rareness of these alleles in nonexposed populations is then a direct assumption. In this case, the selection of resistance genes is a post adaptive response. On the other hand, pre adaptive selection of resistant alleles might have happened before the insecticide pressure, presumably if those alleles had another physiological role. Consequently, this type of resistance alleles would be less likely to carry a fitness cost [39].

The presence of insecticides in the environment is the basis for resistance selection. Operational factors, like formulation, dosage, frequency, and intensity of application, will determine the strength of that selection pressure. Likewise, environmental and intrinsic biological elements will determine the extension and velocity for the dispersion course of resistance alleles. The amount of resistance alleles and their initial frequency, as well as their dominance, penetrance, expressiveness, and interaction within the whole genetic background are the genetic compo‐ nents. In parallel, biological and ecological pieces in this scenario include the offspring size, generation turnover, mono or polygamy behaviors, together with degrees of mobility, isolation, and migration, mono or polyphagia, use of refuges, etc. [41]. Naturally, the knowl‐ edge of most of these aspects will optimize the design for more effective insect control strategies. Even considering all those parameters, insecticide application can play a strong selection pressure, able to change the profile of a population very quickly [42].

One parameter that probably has a large impact on the evolution of insecticide resistance is the side effects, usually negative, related to the resistance mechanisms. This is likely the main reason that explains the low frequency of resistance alleles in populations not exposed to chemicals. Therefore, the most common assumption is that when the use of insecticides is interrupted, the frequency of nonresistant specimens would tend to increase toward the establishment of the previous susceptibility levels of the population. This is especially what managers of campaigns against vector of pathogens anxiously look for, once the arsenal of insecticide compounds to this end is very restricted [4, 5].

The mode of insecticide application is crucial to the velocity of resistance evolution. Since the main goal of these control strategies is a prompt reduction of the targeted insect population, they often apply high dosages of insecticides, which combined with the indiscriminate use of the household or agriculture products, result in a strong selective pressure. Hence, even with a high impact on the fitness, some resistance alleles can spread among populations [43]. Besides physiological and reproductive hitched-hiked costs for resistance, a continuous pressure may favor the spread of mechanisms with lower side effects. An important factor resulting from the refining aspect of Natural Selection over the adaptation for resistance is the selection of "modifier genes", which neutralize or compensate deleterious effects [44]. The modifier genes can reduce drastic effects on the overall fitness previously induced by some resistance alleles, enhancing the adaptation to the environment with insecticides.

An emblematic example occurred in the in the Australian sheep blowfly *L. cuprina*, where a mutant allele for the carboxylesterase E3 is responsible for resistance to the OP diazinon, presenting, however, high disadvantage in environments free of insecticide. One of the effects on the overall fitness was a bilateral asymmetry in the resistant flies. With continuous use of insecticide over the resistant population, a modifier gene was subsequently selected, increas‐ ing the fitness and also neutralizing the negative effects over the asymmetry [45]. Later, it was verified that the candidate for that modifier was a gene with an important role in oogenesis, spermatogenesis, embryonic mesoderm formation, and eyes development. The authors hypothesized that the resistance allele had a broad pleiotropic effect causing developmental perturbations that affected bristles and wing development, presumably impelled by a role of the carboxylesterase E3 in cell adhesion. The selection of the modifier gene compensated these effects [46].

employment has been discussed since the 1950s with the works of Crown [38] and more recently incremented on ffrench-Constant's reviews [39, 40]. If resistance mechanisms hold elevated fitness cost in absence of insecticide (as discussed subsequently), the rareness of these alleles in nonexposed populations is then a direct assumption. In this case, the selection of resistance genes is a post adaptive response. On the other hand, pre adaptive selection of resistant alleles might have happened before the insecticide pressure, presumably if those alleles had another physiological role. Consequently, this type of resistance alleles would be

The presence of insecticides in the environment is the basis for resistance selection. Operational factors, like formulation, dosage, frequency, and intensity of application, will determine the strength of that selection pressure. Likewise, environmental and intrinsic biological elements will determine the extension and velocity for the dispersion course of resistance alleles. The amount of resistance alleles and their initial frequency, as well as their dominance, penetrance, expressiveness, and interaction within the whole genetic background are the genetic compo‐ nents. In parallel, biological and ecological pieces in this scenario include the offspring size, generation turnover, mono or polygamy behaviors, together with degrees of mobility, isolation, and migration, mono or polyphagia, use of refuges, etc. [41]. Naturally, the knowl‐ edge of most of these aspects will optimize the design for more effective insect control strategies. Even considering all those parameters, insecticide application can play a strong

One parameter that probably has a large impact on the evolution of insecticide resistance is the side effects, usually negative, related to the resistance mechanisms. This is likely the main reason that explains the low frequency of resistance alleles in populations not exposed to chemicals. Therefore, the most common assumption is that when the use of insecticides is interrupted, the frequency of nonresistant specimens would tend to increase toward the establishment of the previous susceptibility levels of the population. This is especially what managers of campaigns against vector of pathogens anxiously look for, once the arsenal of

The mode of insecticide application is crucial to the velocity of resistance evolution. Since the main goal of these control strategies is a prompt reduction of the targeted insect population, they often apply high dosages of insecticides, which combined with the indiscriminate use of the household or agriculture products, result in a strong selective pressure. Hence, even with a high impact on the fitness, some resistance alleles can spread among populations [43]. Besides physiological and reproductive hitched-hiked costs for resistance, a continuous pressure may favor the spread of mechanisms with lower side effects. An important factor resulting from the refining aspect of Natural Selection over the adaptation for resistance is the selection of "modifier genes", which neutralize or compensate deleterious effects [44]. The modifier genes can reduce drastic effects on the overall fitness previously induced by some resistance alleles,

An emblematic example occurred in the in the Australian sheep blowfly *L. cuprina*, where a mutant allele for the carboxylesterase E3 is responsible for resistance to the OP diazinon, presenting, however, high disadvantage in environments free of insecticide. One of the effects

selection pressure, able to change the profile of a population very quickly [42].

insecticide compounds to this end is very restricted [4, 5].

enhancing the adaptation to the environment with insecticides.

less likely to carry a fitness cost [39].

248 Insecticides Resistance

In *Culex* mosquitoes the *ace-1R* allele codes for the G119S mutant AChE resistant to OP, however, with 60% lower activity than the wild-type enzyme. Consequently, resistant individuals present a severe fitness cost, reflected with the decrease of the *ace-1R* allele frequency in the absence of insecticide, as observed in some *Culex pipiens* populations [47, 48]. The G119S mutation in *Anopheles gambiae* followed the same tendency [49]. The emergence of gene duplication in the *ace-1* locus containing both resistant *ace-1R* and susceptible *ace-1S* alleles not only guaranteed resistance to OP but also diminished the resistance deleterious effects, once the physiological role of the enzyme was no longer compromised [50].

Another scenario of amelioration of resistance was richly described by Labbé et al. (2009) for a gradual replacement of resistant genes in a decade's time among populations of *C. pipiens* from Montpelier, Southern France. In that study, the authors found that the *Ester1* allele (from *Ester* locus, enrolled with over production of EST) was selected for resistance to OP; however, it was later replaced by the *Ester4* allele. This newer one conferred the same advantages over insecticides, nonetheless with lower pleiotropic effects and fitness cost. Interestingly, a third allele *Ester2* with both higher advantage and fitness cost seemed to be replacing the previously selected *Ester4* . The hypothesis raised was that the first replacement (*Ester1* to *Ester4* ) occurred as a compensatory amelioration, since *Ester4* is less costly and more "generalist". On the other hand, the *Ester2* allele would be more "specialist" to insecticide-treated areas, conferring high resistance but with strong pleiotropic effects. The practices of insecticide use in different areas of Montpelier during that time certainly influenced the evolution of this *Ester* locus. If the intensity of treatment had decreased, *Ester4* would have possibly been favored over the stronger resistant *Ester2* allele, given the former's lower fitness cost [44].

Although a common class of insecticide can select the same mutation for resistance in different insects, its effects on fitness vary through the species or even among different populations of the same species. For instance, the A302S *rdl* mutation remained under high frequencies in natural populations and the resistance persisted despite the withdrawal of cyclodienes in the field for years, as reported to natural populations of *Drosophila* [51], the German cockroach [52], and to the mosquito *A. gambiae* [53]. On the other hand, a reduction in the *rdl* resistant allele without insecticide selection pressure was observed in natural populations of the horn fly *H. irritans* [13] and the Australian sheep blowfly *L. cuprina* from both field and laboratory caged strain [54]. In the same way, *rdl* mutant *A. gambiae* and *Anopheles stephensi* mosquitoes pre‐ sented reduced fertility and fecundity [55].

One has to consider that the evaluation of the overall fitness effects of a given mutation is very challenging, once it is difficult to separate their own effects from those caused by other mechanisms possibly coselected for resistance. In these aforementioned *rdl* examples, the reduced fitness might be related to the A302S mutation itself, and/or to metabolic resistance mechanisms. Similarly, the persistence of the resistance allele in an environment free of dyeldrin might be explained by the rdl cross-resistance with other insecticide that had been continually applied, as well as by the selection of modifiers genes, as previously discussed.
