Multiple insecticide resistance in Culex quinquefasciatus populations from Guadeloupe (French West Indies) and associated mechanisms

West Nile (WN) virus has been detected in Guadeloupe since 2002. Even if no WN human cases have been detected so far, mosquitoes from Culex genus especially Culex quinquefasciatus are recognized as potential WN vectors in Guadeloupe. To evaluate the impact of local vector control activities on this mosquito species we assessed the resistance levels of Cx. quinquefasciatus populations from three different sites from Guadeloupe (Abymes, Saint François and Gourbeyre) to malathion, temephos and deltamethrin. In addition, the frequencies of the L1014F kdr and the G119S ace-1 mutations were established in Cx. quinquefasciatus populations, as well as the constitutive expressions of five cytochrome P450 genes. Mosquito populations tested displayed high resistance to deltamethrin, moderate resistance to malathion (Abymes, Gourbeyre) and low resistance to temephos (Abymes et Gourbeyre). Molecular analyses revealed high frequencies of both L1014F kdr and G119S ace-1 mutations in Cx. quinquefasciatus populations, as well as overexpression of cytochrome P450 genes CYP9J45, CYP9J40 and CYP6AA7. Finally, deltamethrin resistance and knock-down rates were strongly correlated with the frequency of the resistant kdr and ace-1 alleles, as well as with CYP9J40 overexpression. These results should be taken into account to refine the current vector control strategies to prevent the appearance of Cx. quinquefasciatus-borne diseases in Guadeloupe.


Introduction
During the past 30 years, the Guadeloupe islands (French West Indies) have been facing arboviral outbreaks with increasing frequency, number of cases and fatalities. The largest dengue outbreak ever described in this territory was reported in 2010 involving around 44,000 dengue-like syndromes [1]. Since then, a Chikungunya epidemic (80,000 suspect cases) hit the PLOS  Guadeloupian archipelago between 2013 and 2015 [2], shortly followed by a Zika virus epidemic that led to 28,345 suspect cases [3,4], 67 severe neurological complications and 14 foetal neurological pathologies, including microcephaly [5].
As the arboviruses causing dengue (DENV), chikungunya (CHIKV) and Zika (ZIKV) are mainly transmitted in the Caribbean by the urban mosquito Aedes aegypti (Linnaeus 1762) [6][7][8][9], vector control activities in Guadeloupe focus mainly in this mosquito species. Nevertheless, since 2002 West Nile virus (WNV; Flavivirus) has also been detected in Guadeloupe during several serological investigations performed on horses and chickens [10,11]. WNV is the cause of the most widespread vector-borne disease in the world and one of the largest outbreak of neuroinvasive disease ever observed [12]. WNV circulates in nature via an enzootic cycle between ornithophilic mosquitoes (i.e. Culex genus), and susceptible birds [12][13][14]. In Guadeloupe, no human WNV cases have been reported so far [15], but the possibility of WNV emergence in urban settings remains a real concern. Indeed, the southern house mosquito Culex quinquefasciatus (Say 1823) is the most abundant mosquito species in urban environments of French West Indies after Ae. aegypti [16][17][18], and is a highly anthropophilic species that have been proven to be competent to transmit WNV in the Americas [19][20][21].
Vector control activities in Guadeloupe have been historically carried out by a Vector Control Agency from the French Ministry of local Health (Agence Régionale de Santé; ARS). The ARS ensures larval reduction mainly through the mechanical elimination of breeding sites, the use of larvivorous fish Poecilia reticulata and larvicide treatments, while adult densities reduction is also directly conducted by spraying chemical adulticides outdoor and indoor [22]. The insecticides commonly used in the past have been the larvicide temephos and the adulticides deltamethrin and malathion [23,24]. However, temephos and malathion have been withdrawn from the list of vector control insecticides following European Union recommendations in 2009 and 2010 respectively, and have been since then completely replaced by the biological larvicide Bacillus thuringiensis var. israelensis (Bti) and the deltamethrin [22,25].
The deployment of insecticides to control mosquito vectors has allowed the selection of resistance around the world [26,27]. The two main insecticide resistance mechanisms in mosquitoes are the enzymatic detoxification (also known as metabolic resistance) and the modification of the molecular site targeted by the neurotoxic insecticide chemical (also known as target-site modification) [28,29]. The major classes of enzymes involved in metabolic resistance are esterases, glutathione-S-transferases and cytochrome P450 oxidases [28,29]. All these classes are involved in resistance to different insecticides families, but cytochrome P450 oxidases play a major role in resistance to pyrethroids such as deltamethrin and permethrin [29][30][31]. Concerning the target-site modifications, mutations in voltage-gated sodium channels (VGSC) have been found to confer resistance to pyrethroids, while mutations in the synaptic acetylcholinesterase ace-1 gene confer resistance to organophosphates and carbamates [32,33]. Ae. aegypti populations from Guadeloupe are multiple resistant to insecticides (temephos, malathion and deltamethrin) as a result of a combined insecticide pressure exerted by agricultural and vector control activities [24]. Nevertheless, little is known about the resistance status of Cx. quinquefasciatus mosquitoes from Guadeloupe as the only insecticide resistance assessment for this species have been performed in 1999 and no mechanisms of resistance were investigated so far [18]. As Cx. quinquefasciatus and Ae. aegypti can share habitats and coexist in urban areas [34], it is probable that the large deployment of insecticides to control Ae. aegypti mosquitoes have also secondarily impacted the insecticide resistance levels of neighboring populations of Cx. quinquefasciatus, even if this latter species was not primarily targeted by the local vector control activities in Guadeloupe (i.e. deltamethrin spraying). In this study, we assessed the resistance levels of three Cx. quinquefasciatus populations from Guadeloupe to malathion, temephos and deltamethrin. We have also characterized for the first time resistance mechanisms in Cx. quinquefasciatus populations from Guadeloupe. Among them, the frequency of most commons target-site modifications described in Cx. quinquefasciatus: (i) the L1014F mutation in the VGSC gene, conferring resistance to pyrethroids and DDT (Dichlorodiphenyltrichloroethane insecticide) [33,35,36], and (ii) the G119S mutation in the ace-1 gene, conferring resistance to temephos and malathion [32]. In addition, we have determined the constitutive expression levels of five cytochrome P450s that have been shown to be involved in insecticide detoxification in Cx. quinquefasciatus [30,31,37]. Finally, we have performed association analysis between the insecticide resistance phenotypes and the resistance mechanisms explored.

Ethics statement
This study has been approved by the internal ethics committee of the Pasteur Institute of Guadeloupe. Anubis Vega-Rúa (author of the study) provided a written consent for blood donation to artificially feed mosquitoes.

Mosquito collection and rearing
Larvae of Cx. quinquefasciatus were collected from March to April 2016 in three different locations of Guadeloupe: Gourbeyre (N 16˚15'15,7"W 61˚19'58,1"), Saint François (N 16˚14'46,4"W 61˚19'56,2") and Abymes (N 16˚12'33,0"W 61˚29'57,6"). These mosquito collection sites are located in zones that often undergo a contrasted number of insecticide treatments, because of their differences in terms of human population size and density (Fig 1). Mosquito samplings were performed around private houses randomly chosen and where the residents gave their permission for the larvae collections. Larvae and pupae were then brought to the insectaries and split into 150-200 individuals per container of dechlorinated tap water supplemented with rabbit food. Adults that emerged were maintained in cages at 27˚±1˚C, 80% of humidity, 12h:12h light:dark cycle (inverted) and supplied with a 10% sucrose solution. Human blood from volunteers was provided to female mosquitoes using Hemotek 1 feeding system in order to produce eggs (first generation or F 1 ). The F 1 mosquitoes were used for all the experiments and molecular biology investigations. Slab susceptible mosquito strain was graciously provided by the Institut Pasteur in Paris [38]. Slab strain was used as a reference for all insecticide susceptibility tests.

Larval bioassays
Larval bioassays were performed using temephos and malathion insecticides following the WHO (World Health Organization) recommendations [39]. The late third and early fourthinstar larvae were used for each mosquito population. Four replicates per concentration and five to six concentrations were tested with 25 larvae per replicate and concentration. The insecticide alcohol dilutions tested were provided by WHO Universiti Sains Malaysia manufacturer. Temephos and malathion "stock" solutions received were at 156.25μg/ml and 31.25μg/ml. For testing the field mosquito samples, the concentrations were adjusted to include different percentages of survival/mortality, varying from 0% to 100%. The concentrations tested for field samples were 0.001, 0.003, 0.005, 0.007, 0.01μg/ml for temephos and 0.01, 0.025, 0.05, 0.1, 0.3μg/ml for malathion, while concentrations tested for the reference Slab strain were 0.0001, 0.0005, 0.001, 0.003, 0.005, 0.01μg/ml for temephos and 0.0075, 0.01, 0.025, 0.05, 0.1μg/ml for malathion. The results were analyzed with XLSTAT-Biomed software (XLSTAT 2015) to determine the lethal concentration for 50% (LC 50 ) of the populations. Resistance ratio (RR 50 ) was calculated using LC 50 rates from Cx. quinquefasciatus field sampled populations compared with the LC 50 rate of the susceptible Slab strain. RR 50 1 indicates susceptibility to the tested insecticide, while RR 50 >1 indicates insecticide resistance. The resistance levels were also ranked in 3 categories according to Mazzarri & Georghiou (1995) [40] and Bisset et al. (2011) [41] as follows: low resistance for RR 50 <5, moderate resistance for 5 RR 50 10 and high resistance for RR 50 >10.

Adult bioassays
Adult bioassays were performed following adapted WHO recommendations [42,43]. Deltamethrin impregnated papers were used at different concentrations: 0.01%, 0.025%, 0.05%, 0.08% for field samples and 0.0001%, 0.001%, 0.01%, and 0.025% for the reference Slab strain. For each mosquito population and each concentration, four replicates and two negative controls of 25 females (2-5 days old) were tested. Whatman n˚1 papers were impregnated in the laboratory with deltamethrin PESTANAL 1 (Sigma-Aldrich, Inc.) at the desired concentration in a solution containing 2/3 Acetone (Sigma-Aldrich, Inc.) and 1/3 silicon oil (VWR 1 ). The impregnation was conducted following WHO procedures [42], using 2ml of insecticide solution to impregnate one 12x15cm paper that was allowed to dry overnight and stored at 4˚C. Adult mosquitoes were exposed by tarsal contact for one hour to deltamethrin impregnated papers. The knock-down (after 1h exposure) and the mortality rates (after 24 hours) were recorded.

Genotyping of target-site mutations
Genomic DNA was extracted from 30 Cx. quinquefasciatus females from each of the three populations studied using CTAB technique [44]. The extracted DNA was used to genotype L1014F kdr mutation of the VGSC and G119S ace-1 mutation. The L1014F was genotyped using the double PCR-based assay described by Martinez-Torrez et al. [33] using a Dreamtaq green PCR master mix (Thermoscientific 1 ) and the following PCR conditions: 5 min at 94˚C followed by 40 cycles of 94˚C for 30 sec, 55˚C for 30 sec and 72˚C for 30 sec, with a final extension of 7min at 72˚C. The G119S ace-1 mutation was investigated using the PCR-RFLP (PCR Restriction Fragment Length Polymorphism) test developed by Weill et al. [32]. The PCR conditions were slightly modified as follows: 5 min at 94˚C, 30 cycles (94˚C for 30 sec, 52˚C for 1 min and 72˚C for 1 min) and a final extension of 7 min at 72˚C. Primers used are shown in Table 1.

Expression of cytochrome P450 genes
RNA was extracted from three pools of 16 females for each of the three different mosquito populations using trizol/chloroform (Invitrogen, Carlsbad, CA, USA) method and cDNA was

Statistical analysis
Statistical analysis were performed with XLSTAT 2015 (Addinsoft, France) software. XLSTAT was used (i) to perform log-probit logistical regression to investigate dose-effect relationship regarding insecticide tests data, (ii) to compare adult KD/mortality (Deltamethrin) and larval mortality (temephos and malathion) between mosquito populations via ANOVA and Kruskal-Wallis tests, (iii) to perform pairwise comparisons (Students t-test) of gene expression ratios between Slab strain and the field populations and (iv) to analyze the association between the obtained variables via a Principal Component Analysis (PCA). The frequency data from the ace-1 and VGSC genes were analyzed using the Genepop software [45] to check Hardy-Weinberg equilibrium for each mosquito population. The overall significance of multiple tests was estimated with Fisher's combined probability test and the p-value corrected with the sequential Bonferroni procedure [46].

Insecticide resistance profile of Cx. quinquefasciatus for temephos, malathion, and deltamethrin
Larval and adult bioassays conducted revealed multiple resistances in Cx. quinquefasciatus populations from Guadeloupe ( Table 2). The lowest resistance ratios (RR 50 ) were observed for Insecticide resistance in Culex quinquefasciatus from Guadeloupe temephos (RR 50 ranged between 1-2), indicating low resistance for Cx. quinquefasciatus populations except for mosquitoes from Saint François which are still susceptible for this insecticide (RR 50 = 1). This latter population also displayed significant higher mortality (Table 3) than the populations from Abymes and Gourbeyre when exposed to this insecticide (Kruskal-Wallis Htest: p<0.0001). Concerning malathion, mosquito populations from Abymes and Gourbeyre displayed moderate resistance with RR 50 of 5.38 and 9.75 respectively, while the population from Saint François was found to be still susceptible to this insecticide (RR 50 = 0.75). In addition, the mortality of three populations was statistically different for all the malathion doses except 0.3μg/ml (Kruskal-Wallis H-test: p<0.0001). The higher mortality after exposure to malathion was displayed by the population of Saint François, followed by Abymes and Gourbeyre (

Genotyping of L1014F kdr and G119S ace-1 mutations
A high frequency of the L1014F kdr mutation was detected in all the three Cx. quinquefasciatus populations from Guadeloupe. Whatever the mosquito population, the frequency of the mutant and resistant allele (F) was higher than the wild "susceptible" allele (L): 0.66 for Gourbeyre, 0.68 for Saint François and 0.75 for Abymes. The resistant homozygote genotype (F/F) was found more frequently in the populations Abymes and Saint François (50% of individuals) than in Gourbeyre (32% of individuals) (Fig 2). The highest proportion of heterozygote genotype (L/F) was found in Gourbeyre (69%), while the lowest proportion was observed in Table 3. Mortality and Knock-down rates of Culex quinquefasciatus populations from Guadeloupe exposed to deltamethrin (adults), malathion (larvae) and temephos (larvae).

Population Abymes Saint François Gourbeyre
Insecticide Dose (μg/ml) Ã   Regarding the G119S ace-1 mutation, the frequencies of the resistant allele (S) were lower than those of the wild "susceptible" allele (G) in all populations: 0.35 for Abymes, 0.27 for Saint François and 0.12 for Gourbeyre. In addition, no homozygote resistant genotype (S/S) was detected in any of the populations (Fig 2). Nevertheless, the proportion of individuals with the heterozygote genotype harboring the resistant mutation (G/S) was very high in the population Abymes (70%), followed by Saint François (53.3%). The majority of individuals from Gourbeyre (75.9%) still harbor the wild homozygote susceptible (G/G) genotype. Significant deviations from Hardy-Weinberg expectations (Hardy-Weinberg exact test: p = 0.014) were detected for this locus only in the Abymes population due to an excess of hererozygotes (Hardy-Weinberg exact test for alleles: p = 0.0038). Neither kdr nor ace-1 mutation was found in the susceptible Slab strain.

Transcriptional profiling of selected cytochrome P450s expression levels in Cx. quinquefasciatus
The transcription profiles of the five candidate detoxification genes potentially involved in metabolic resistance to insecticides in the three Cx. quinquefasciatus populations from Guadeloupe were compared to those of the susceptible Slab strain (Fig 3). Genes with transcription ratio (TR)!2 and P-value 0.05 (according to the T-student test) compared to the Slab strain were considered as significantly over-transcribed. Only the CYP9J45, CYP9J40 and CYP6AA7 were significantly over-transcribed. The CYP6AA7 was over-expressed in mosquitoes from Abymes (TR = 6.08; p = 0.0049) and Saint François (TR = 3.29; p = 0.05), while CYP9J45 and

Association between insecticide resistance levels, target-site mutations and expression levels of detoxification enzymes
A Principal Component Analysis (PCA) was performed for all Cx. quinquefasciatus populations using 12 variables such as (i) larval and adult resistance levels (RR 50 ) for temephos, malathion and deltamethrin, (ii) deltamethrin knock-down rates for each concentration used, (iii) adult transcription ratios of cytochrome P450s CYP6AA7, CYP9J40 and CYP9J45 and finally (iv) the allelic frequencies of kdr and ace-1 mutations. The PCA results are summarized in the Fig 4. The two first PCA axes represented 89.93% of the starting information with respectively 71.42% and 18.51%, while the third axis represented only 10.07%. Positive but no significant correlations were found between the resistance ratios of adult mosquitoes for deltamethrin and those obtained with mosquito larvae for temephos (Correlation: r(132) = 0.78, p>0.05) and malathion (Correlation: r(132) = 0.65, p>0.05). Resistance to deltamethrin was strongly correlated with the frequency of the resistant kdr allele (F) (Correlation: r(132) = 0.97, p = 0.031), while resistance to temephos was found to be correlated to the expression fold change of CYP9J40 (Correlation: r(132) = 0.965, p = 0.035). The knock-down rates found for deltamethrin displayed strong negative correlations with the resistant allele frequencies of ace-1 mutation (Correlation: r(132) = -0.99, p = 0.013), with the expression fold change of CYP9J40 (Correlation: r(132) = -0.98, p = 0.013), with deltamethrin resistance (Correlation: r(132) = -0.96, p = 0.043) and to a lesser extent with the resistant allele frequency of kdr mutation (Correlation: r(132) = -0.94, p = 0.061). No significant associations were detected between the expression fold levels of CYP9J45 and CYP6AA7 and any of the tested variables.

Discussion
In this study we investigated the resistance status of Cx. quinquefasciatus mosquitoes from Guadeloupe against malathion, temephos and deltamethrin: three insecticides that have been historically used for chemical control of the mosquito vector Ae. aegypti in the island. Malathion and temephos were respectively used in Guadeloupe for adult and larval vector control from the late 1960s to 2009-2010, while deltamethrin has been used from the early 1980s until the present day (Pasteur Institute archives). Bioassays performed on the three Cx. quinquefasciatus populations from Guadeloupe revealed high resistance levels to deltamethrin (20<RR 50 <29), moderate resistance to malathion (0.75<RR 50 <9.75) and a low resistance to temephos (1<RR 50 <2). Cx. quinquefasciatus populations from Guadeloupe have been already found to be resistant to deltamethrin since the late 1990s [18], but the resistance to malathion had never been described for this mosquito species in Guadeloupe until now. However, the resistance to malathion has already been reported for Cx. quinquefasciatus populations from Latin America with RR 50 values sometimes up to 200 [34]. The differences on the resistance levels observed for these two adulticides in the studied populations from Guadeloupe are in agreement with a more recent and higher deltamethrin selective pressure, as this insecticide is still used for vector control activities in the island (last massive spraying in 2014 during the chikungunya outbreak), while malathion has not been not used since 2010 because the insecticide was withdrawn from the list of vector control insecticides [25]. The resistance observed in Cx. quinquefasciatus mosquitoes regarding these two insecticides is possibly due to the lack of specificity of indoor and outdoor spraying used for Ae. aegypti control in Guadeloupe. This "secondary resistance" developed by a non-target organism has been often reported for different insect species [47]. In the case of Cx. quinquefasciatus, this finding is not surprising as this mosquito is a domestic species highly adapted to the urban environment that often co-exists with Ae. aegypti (especially the adult stages) in human habitations [48]. Interestingly, Cx. quinquefasciatus populations presented more heterogeneous resistance levels (0.75<RR 50 <9.75) when compared to Ae. aegypti (1.68<RR 50 <4.39) [24], which could be possibly explained by contrasted or unequal levels of co-habitation with Ae. aegypti in the sampled zones (i.e. greater Cx. quinquefasciatus-Ae. aegypti co-habitation in zones were breeding-sites are more scarce or insufficient).
The lowest resistance levels were obtained for temephos. A slight resistance was observed in populations from Abymes (RR 50 = 2) and Gourbeyre (RR 50 = 1.67), while the population from Saint François was susceptible to this insecticide (RR 50 = 1). These results are in agreement with a study conducted in Brazil were RR 50 values reported were low (RR 50 <2) [49]. As previously mentioned, temephos has been used in Guadeloupe for larval vector control for about 40 years (Pasteur Institute archives) and the first resistance to this insecticide in the island was reported as early as 1984 for Ae. aegypti mosquitoes (Pasteur Institute archives). In 1999, resistance to temephos was detected for the first time in Cx. quinquefasciatus mosquitoes from Guadeloupe with resistance ratios comprised between 4.47 and 32.58 [18]. Despite the difficulty to compare insecticide resistance assessments that were performed with 19 years of difference using different mosquito populations from Guadeloupe, the resistance levels obtained in the present study are consistent with the withdrawal of temephos for the vector control activities [25] and thus, with a lower temephos-associated selective pressure. In addition, temephos treatments in the last decades were principally used for Ae. aegypti breeding-sites and not for those infested with Culex spp, which may have also contributed to a faster decrease of temephos resistance levels in Cx. quinquefasciatus populations.
Concerning the insecticide resistance mechanisms, the kdr mutation L1014F was more frequent in Cx. quinquefasciatus populations than the ace-1 mutation G119S (Fig 2). Homozygote resistant individuals (F/F genotype) were observed for the kdr mutation only, and almost no homozygote susceptible genotypes (L/L) have been found among the analyzed mosquitoes, suggesting a higher selective pressure for this particular mutation than for ace-1. The L1014F kdr mutation has been shown to confer resistance to pyrethroids in Cx. quinquefasciatus [33,35], which can explain the high resistance values obtained for the adulticide deltamethrin (still used in Guadeloupe) for all the studied populations (RR 50 !20). This statement is also supported by the PCA analysis that revealed strong positive correlations between deltamethrin resistance and the frequency of the resistant kdr allele [F] (Correlation: r(132) = 0.97, p = 0.031). The highest deltamethrin resistance and the highest frequency of L1014F kdr mutation were both found in the Abymes population, suggesting that a higher deltamethrin-selective pressure has been exerted in this population for decades. The spraying conducted during the chikungunya outbreak in 2014 reflects this issue, as the number of deltamethrin spraying in Abymes was 53 while in Gourbeyre only 5 insecticide spraying events have been recorded in the same period (Fig 1) (see Annex 8 in [50]). The G119S ace-1 mutation was less frequent in the populations when compared to L1014F, but it was found in all Cx. quinquefasciatus populations as well, which could explain the moderate malathion resistance and the low temephos resistance observed in this study. As explained above, the G119S ace-1 mutation has shown to confer resistance to organophosphorous insecticides [32,51] and has also been associated with malathion resistance in Cx. quinquefasciatus [52]. Malathion spraying conducted in the past decades and before 2010 to target adult mosquito stages could have acted as the main source of selective pressure for this target-site mutation in Cx. quinquefasciatus populations, as the highest frequency of heterozygote individuals for ace-1 mutation harboring the resistant allele for this gene was found in Abymes, a locality considered as a transmission "hot spot" for arboviruses that has been historically more treated with insecticides when compared to Saint François and Gourbeyre (Figs 1 and 2) [50]. However, it is possible that other mechanisms nonexplored in this study such as malathion detoxification by carboxylesterase enzymes [28,[53][54][55] could have been involved in malathion resistance, as Gourbeyre population displayed the highest resistance to malathion and the lowest frequencies of the resistant allele for the ace-1 mutation. Taken together, these results suggest an association between the resistance levels observed in Cx. quinquefasciatus for the tested insecticides, the frequency of both target-site mutations studied, and the insecticide selective pressures exerted by the past and present vector control activities ( Table 2, Figs 1 and 2).
Resistance mechanisms associated to the insecticide detoxification by cytochrome P450s have also been explored for the first time in Cx. quinquefasciatus populations from Guadeloupe. Constitutive over expression of P450 genes like CYP6AA7, CYP9J40, CYP9J45, CYP4H34 and CYP9M10 has been linked to the development of resistance to pyrethroids (i.e. permethrin, deltamethrin) in Cx. quinquefasciatus [30,56]. Among the five cytochrome P450s studied, only three have found to be over-transcribed in the local Cx. quinquefasciatus populations when compared to the susceptible Slab strain: CYP9J45, CYP9J40 and CYP6AA7. Again, the mosquito population "Abymes" had the highest number of over-transcribed cytochromes with the highest transcription levels, which is in agreement with the important resistance levels to deltamethrin observed for this population. In addition, the Principal Component Analysis revealed significant negative correlations between the expression fold change of cytochrome P450 gene CYP9J40 and the deltamethrin-induced knock-down rates (Correlation: r(132) = -0.98, p = 0.013). The CYP9M10 and CYP4H34 were not over-transcribed in any tested population, which can be explained by the use of adult mosquitoes for the experiments for these two genes. Indeed, several studies have found that the constitutive over-expression of CYP9M10 and CYP4H34 genes is very high in the 4 th -larval stage but subsequently drops in pupae and adult mosquitoes [30,56]. Nevertheless, further studies should be conducted to compare the constitutive expression of CYP9M10 and CYP4H34 genes in larvae and adult mosquitoes to verify if the same expression decrease from larvae to adult is also observed in field-collected mosquitoes. In the same way, the over-expression by insecticide induction of the studied cytochromes and other metabolic enzymes should be also deeply investigated to better characterize their role in the insecticide resistance of Cx. quinquefasciatus [31].

Conclusions
Cx. quinquefasciatus mosquitoes from Guadeloupe are multiple resistant to insecticides even if this mosquito species is not the main target of the vector control activities in the region. The studied populations displayed high resistance to deltamethrin, moderate resistance to malathion and were susceptible-to-slightly resistant to temephos. The resistance levels observed in Cx. quinquefasciatus populations seem to be associated with the kdr and ace-1 mutations, as well as with the overexpression of cytochrome P450 genes CYP9J40 CYP6AA7, and CYP9J45. The population of Abymes is the better example of the effects of insecticide selective pressure in Cx. quinquefasciatus mosquitoes, as this population has historically been prioritized for the insecticide treatments, which had led to the appearance and the selection of multiple resistances and the associated mechanisms. Indeed, the Abymes mosquito population had the highest frequency of both kdr and ace-1 mutations, and the highest constitutive expression levels of the studied cytochrome P450s. Cx. quinquefasciatus mosquitoes are resistant to the only chemical insecticide that is still used in vector control strategies locally. Thus, the challenge is to manage this resistance in order to prevent the emergence of Cx. quinquefasciatus-borne diseases in Guadeloupe.