Phenotypic and molecular analysis of insecticide resistance in Aedes albopictus populations from Greece

Background: Aedes albopictus has a well-established presence in southern European countries, associated with recent disease outbreaks (e.g. Development of insecticide resistance in the vector is a major concern as it’s control mainly relies on the use of biocides. Data on the specie’s resistance status is essential for efficient and sustainable control. Methods: We investigated the insecticide resistance status of several Ae. albopictus populations from Greece. Bioassays were performed against diflubenzuron (DFB), B. thuringiensis var. israelensis (Bti), deltamethrin and malathion. Molecular analysis of known insecticide resistance loci was performed, i.e. voltage-gated sodium channel (VGSC) mutations associated with pyrethroid resistance; presence and frequency of carboxylesterases 3 (CCEae3a) and 6 (CCEae6a) gene amplification associated with organophosphate (OP) resistance and; chitin synthase-1 (CHS-1) for the possible presence of DFB resistance mutations. Results: Bioassays showed full susceptibility to DFB, Bti and deltamethrin, but resistance against the OP malathion. VGSC analysis revealed a widespread distribution of mutations F1534C (in all populations, with allelic frequencies between 6.6% - 68.3%), and I1532T (in 6 populations), but absence of V1016G. CCE gene amplifications were recorded in 8 out of 11 populations. Co-presence of mutation F1534C and CCEae3a amplification was reported in a subgroup of samples. No mutations at the CHS locus I1043 were detected. Conclusions: The results indicate: (i) the suitability of larvicides DFB and Bti for Ae. albopictus control in Greece, (ii) a possible incipient pyrethroid resistance due to the presence of kdr mutations and (iii) a possible reduced efficacy of OPs, in a scenario of re-introducing them for vector control. The study highlights the need for systematic resistance monitoring for developing and implementing appropriate evidence-based control programs. This study delineates the susceptibility profile of Aedes albopictus populations from Greece against commonly used insecticides. Our findings suggest the suitability of the larvicides DFB and Bti and the current effectiveness of pyrethroids for Ae. albopictus control. However, the presence of high kdr mutation frequencies raises concerns, given the dominant role of pyrethroids in mosquito control and the few alternative synthetic compounds available for public health purposes. The OPs malathion and temephos appear unreliable alternative insecticides for possible future re-introduction in mosquito control, as resistance against malathion was detected in the country. Systematic insecticide resistance monitoring is imperative for the development of insecticide resistance management programs ensuring the sustainability of the current chemical control tools.


Introduction
In the last two decades southern European countries have suffered from a number of vector-borne diseases (VBDs), such as Chikungunya (CHIKV), Dengue (DENV) and West Nile virus (WNV) outbreaks the majority of the vector control interventions implemented in large-scale (regional level) control programs. Pyrethroid insecticides, both type II (such as deltamethirn) and type I (such as permethrin and alpha-cypermethrin) formulations are registered and used in professional vector control programs and for personal protection/ household level applications [20]. A major issue associated with the intensive use of a limited number of insecticides in mosquito and agricultural pest control is the development of insecticide resistance [18]. Insecticide resistance has been reported in Ae. albopictus, yet a big knowledge gap remains in regards to the susceptibility status, geographical distribution, frequency and co-occurrence of resistance traits and underlying mechanisms in the vector populations. Bioassay experiments have recorded resistance to several insecticides including pyrethroids, DDT, temephos, malathion, etc. mainly in rural and urban central and southern Asian populations (e.g. from China, Thailand, Singapore, India), while only sporadical cases of resistance have been reported in Europe, America and western Africa [18].
The two insecticide resistance mechanisms reported so far in Aedes mosquitoes are a) target-site resistance, involving mutations at the insecticide's target site of action and b) metabolic detoxification, through overexpression or conformational changes of enzymes involved in the metabolism of the insecticide [18,21]. To date, 5 mutations in 2 loci of the voltage-gated sodium channel (VGSC) gene; V1016G (valine to glycine), I1532T (isoleucine to threonine) and F1534C/L/S (phenylalanine to cysteine/ leucine/ serine), have been reported in Ae. albopictus populations from central and southeastern Asia, European Mediterranean countries, the USA and Brazil [18].
The involvement of all these five mutations in insecticide resistance has been further examined by expressing the mutated VGSC channels in Xenopus eggs and investigating how their electrophysiological properties are affected in the presence of insecticides. All three mutations at position F1534 (F1534C/L and S), as well as the I1532T mutation have been shown to significantly reduce significantly the channel's sensitivity to type I pyrethroids, but not to type II pyrethroids [22,23]. Mutation V1016G mutation also reduces the channel's sensitivity mainly to permethrin and slightly to deltamethrin [22]. However, a synergistic effect has been reported in the presence of the triple mutant V1016G + F1534C+ S989P (a third mutation identified in Ae. aegypti populations), which substantially reduced sensitivity to both permethrin and deltamethrin [24].
Pyrethroid resistance has also been correlated to CYP6P12 over-expression conferring increased metabolic detoxification [25]. Metabolic resistance to the larvicide temephos (OP) has been functionally associated with the up-regulation of carboxylesterases CCEae3a and CCEae6a, due to gene amplification [26]. Ae. albopictus populations from Greece and Florida (USA) have been found to carry this CCEae3a gene amplification or the CCEae3a-CCEae6a co-amplification [27]. Concerning alternative insecticides, such as IGRs or Bti, no genotypic resistance data have been reported for Aedes mosquitoes, but three point mutations I1043M/L/F in the chitin synthase gene 1 (CHS-1) were identified in Cx. pipiens mosquitoes, conferring very high levels of resistance to the larvicide DFB [28][29][30].
The re-appearance of VBDs in Europe, the widespread distribution of Ae. albopictus in southern Europe, the sporadic information on the vectors' insecticide resistance status and the need for evidence-based mosquito control programs acting in advance of new outbreaks, make the monitoring and analysis of the Ae. albopictus insecticide resistance traits a necessity [31]. Here, we analysed the insecticide resistance status in a number of Ae. albopictus populations from Greece, using bioassays and molecular genotyping assays targeting known resistance markers.  (Table 1).

Study sites, sample collections and mosquito handling
Adult specimens were collected with aspiration catches and CDC-light traps baited with dry ice.
Larvae were sampled with dipping collections and eggs were collected with oviposition traps (black plastic cup with 2 wooden tongue depressors as ovipositional substrate). Both larvae and eggs were collected from at least five different sites within each locality in order to minimize the probability of including isofemale mosquitoes in the molecular analyses.
Following ovitrap collections, eggs were reared to adults in standard insectary conditions (temperature 27± 2 o C and relative humidity 70-80%), identified morphologically to species [33] and stored in ethanol at -4 o C for the subsequent molecular analysis. A subgroup of eggs from Aghios Stefanos (region of Attica), Kefalonia, Patras and Heraklion were reared to larvae or adults to use for susceptibility bioassays, as described below.

Genomic DNA extraction and molecular identification of mosquito species
Genomic DNA (gDNA) was extracted from individual larvae or adult mosquitoes and from pools of eggs (10 eggs per pool per locality; 3 localities), using DNazol reagent according to manufacturer's instructions (Invitrogen, Carlsbad, CA).
Species identification was based on the PCR amplification (KAPA Taq PCR Kit) of the nuclear ribosomal spacer gene ITS2 [34] (primers 5.8S, 28S; Additional file 1: Table S1), following an assay that discriminates between Ae. albopictus, Ae. cretinus and Ae. aegypti, by generating PCR products of 509 bp, 385 bp and 324 bp in length, respectively. The applied thermal protocol was the following: initial denaturation at 94 o C for 10 min, 40 cycles x [denaturation at 94 o C for 1 min, primer annealing at 52 o C for 1 min, primer extension at 72 o C for 1 min] and final extension at 72 o C for 10 min. The PCR products were electrophoresed on a 1.5% w/v agarose gel containing ethidium bromide.

Insecticide susceptibility bioassays a) Larval bioassays
Following the WHO guidelines for laboratory and field testing of mosquito larvicides [35], we examined the susceptibility of Ae. albopictus populations to two larvicides; the bacterial larvicide Bti (VectoBac 1200 ITU (International Toxin Units)/ mg; 11,61% w/v) and the insect growth regulator DFB (DU-DIM 15SC; 15% w/v). Both insecticides were diluted in distilled water. Bioassays were performed using Ae. albopictus third-early fourth instar larvae (F 0-F 1 generation), reared under standard insectary conditions (temperature 27±2 o C and relative humidity 70-80%). Fifteen to twenty larvae were placed in 99 ml water, to which 1 ml of the insecticide solution was added. A range of five to nine concentrations was tested for each insecticide (Bti: 0,008 -0 ,50 mg/L; DFB: 0,0004 -0,02 mg/L) in order to define a mortality range between 10 and 95% and determine the LC 50

b) Adult bioassays
Three to five days old, non-blood fed female mosquitoes were subjected to insecticide susceptibility

Genotyping of target site resistance mutations a) Detection of knock-down resistance (kdr) mutations in the VGSC gene
The VGSC domain I was investigated for the presence of the V1016G mutation and domain III for mutations I1532T and F1534L/S/C via PCR and product sequencing.

b) Analysis of the CHS locus 1043
Analysis of the CHS1 locus I1043, to identify possible conserved DFB resistance mutations that have been found in other species [28,30] was performed in pools of mixed gDNA extracted from 5 to 8 Ae.
albopictus individuals of the same locality. Available Ae. albopictus DNA samples from other countries [27] were included in the analysis and genotyped individually. A 350bp fragment of the Ae. albopictus Clean-Up Kit (Macherey Nagel, Dueren, Germany) and sequenced using the kkv F3 primer.  Table S1) and SYBR Select Master Mix (Applied Biosystems, Thermo Fischer Scientific). Histone 3 (NCBI: XM_019696438.1) was used as a reference gene for normalization (primers His3 TaqF, His3

Metabolic resistance: detection of esterase gene amplification
TaqR; Additional file 1: Table S1). The thermal parameters were: 50 o C for 2 min, 95 o C for 2 min and 40 cycles x [95 o C for 3 sec, 60 o C for 30 sec]. Melting curves were performed for reference and target genes to verify the presence of a unique specific PCR product, which was also checked on a 1% w/v agarose gel. A no-template control was included to detect possible contamination. Two replicates per sample were included. CCEae3a and CCEae6a gene copy numbers were estimated relatively to a temephos susceptible Αe. albopictus lab-strain from Greece.

Molecular identification of mosquito species
A total of 482 individual mosquitoes (larvae and adults) and 30 eggs (in pools) were identified to species by PCR discrimination; of the ITS2 genomic sequence length; 97.6% of the samples were identified as Ae. albopictus, while only 12 specimens, all from Chania-Crete, were identified as Ae.

Insecticide susceptibility bioassays a) Larval bioassays
Ae. albopictus populations from Aghios Stefanos-Attica (ATT), Patras and Heraklion, were tested for Bti resistance and found to be susceptible (Figure 1). The calculated LC 50 values were below 0.20 mg/L (corresponding to less than 0.240 ITU/mL) for all populations (  50 values of the control susceptible strains were 0.036 mg/L and 0.044 mg/L, respectively (both corresponding to more than 0.240 ITU/ mL).
All three populations were also tested for diflubenzuron resistance and showed mortality of 100% in DFB doses below 0.02 mg/L. This is significantly lower than the recommended field doses (DU-DIM Genotyping of the VGSC domain II was performed for a total of 323 larvae/adult mosquitoes (in pools of 5-7 individuals) and 20 eggs (in pools of 10). The wild type allele V1016 (codon GTA) was recorded in all cases (Figure 1, Table 4).
For the detection of mutations I1532T and F1534C/L/S in the VGSC domain III, 319 individuals were genotyped. For the first time in Greece, mutation I1532T was detected in 6 out of the 11 surveyed regions. Particularly, 23 genotyped specimens from Rodopi, Thessaloniki, Patras, Argolida, Rethymno and Chania were found to have this substitution, mostly in heterozygosis (genotype 1532I/ 1532T). In the majority of the above-mentioned regions, the mutant allele 1532T frequency was low, varying from 1.7 to 6.5% (Figure 1). The highest frequency was observed in Patra (where the only 2 homozygotes 1532T/1532T were reported) ( Table 4).
Mutation F1534C was found in all regions. The regions with the highest 1534C allele frequency were: Attica (overall allele frequency: 68.3%), Argolida (45.2%), Rethymno (48.3%), Heraklion (44.3%) and Chania (29%), all located in Central and Southern Greece (we excluded Kalamata results due to the small number of specimens collected and analyzed). In contrast, regions from Northern Greece (Rodopi and Thessaloniki) and Western Greece (Patra and Kefalonia) displayed lower 1534C allele frequencies, ranging between 6.6 and 16.7% (Figure 1). The F1534C mutation appeared mainly in heterozygosis, with the exception of Athens, where more than half of the genotyped specimens (51.6%) were homozygous for the mutation (genotype 1543C/1534C) ( Table 4). Only two individuals (sampled from Argolida and Patra) harbouring both mutations, I1532T and F1534C in heterozygosis (genotype 1532I/1532T -1534F/1534C), were recorded.

b) Chitin Synthase (CHS-1): mutations at position I1043
The CHS-1 genomic sequence of 325 Ae. albopictus mosquitoes (larvae and adults in pools of 5-7 individuals) and a total of 20 eggs (in pools of 10) were analyzed for the presence of either of the three mutations I1043L/M/F, previously linked to DFB resistance. No mutations were detected ( Figure   1, Table S2). Likewise no mutations were recorded in the 178 genotyped samples from USA, Brazil, Belize, Gabon, Switzerland, Taiwan, France, Mexico, China, Sri Lanka, Australia, Japan, Lebanon and Bangladesh (Table S3).

Esterase gene amplification
CCEae3a and CCEae6a amplification associated with temephos resistance was recorded in 8 out of the 11 surveyed regions in Greece. Two types of amplification were found: a CCEae3a amplicon and a CCEae3a -CCEae6a co-amplicon.
Amplified esterase genes were detected in specimens from Chios, Argolida, Patra, Kalamata, Attica, Chania, Rethymno and Heraklion. The CCeae3a amplicon frequency reported in the eight locations (i.e frequency of specimens harbouring the amplicon), ranged from 16.6 to 84 % and that of CCEae3a-CCEae6a amplicon from 5 to 80 %. The majority of samples with amplified esterases harboured between 2-10 gene copies. Individuals with more than 10 (11-20) and more than 20 gene copies (in the case of one individual from Rethymno) were also recorded. Attica was the region with the highest percentage of individuals carrying ≥2 copies of either CCEae3a or CCEae6a gene (84% and 80%, respectively) and Chios of individuals with >10 copies (25% and 33.3%, respectively) ( Figure 2).
No carboxylesterase gene amplification was detected in Rodopi, Thessaloniki and Kefalonia populations (Figure 1).

Discussion
This study represents an extended investigation of the insecticide resistance status of Aedes albopictus in Greece. Entomological monitoring revealed the dominant presence of Ae. albopictus over other Aedes container-breeding species and confirmed the species' widespread distribution in the country [12,42] Crete corresponding to an overall frequency of 2.4%. The absence/very low Ae. cretinus frequency in urban and peri-urban settings of the country may be correlated to the competitive and highly adaptable behavior displayed by Ae. albopictus [46].
The Ae. albopictus WHO and CDC bioassays showed full susceptibility of the tested populations against deltamethrin and the larvicides DFB and Bti, suggesting their current suitability for the vector control programs in Greece. On the contrary, mortality levels < 90 % were recorded in the malathion assays indicating resistance against this OP insecticide. Although malathion is not currently used in Greece, the observed resistance raises concerns, regarding the potential need for re-introducing OP insecticides used in the past.
Genotyping of the VGSC gene for the detection of mutations associated with pyrethroid resistance revealed a widespread distribution of the kdr mutation F1534C across the country, following initial reports of the mutation in Greece [47,48]. The recorded 1534C allele frequencies raised above 40% in populations from southern and central Greece, peaking in Attica (68.3%). Previous studies also reported relatively high mutation frequencies (reaching 66%) in populations sampled from the Attica region [47,48]. To date, kdr F1534C has also been recorded in Brazil [49], Singapore, Vietnam [50] and China [51], while other substitutions in the same genomic locus, F1534L and F1534S, also associated with pyrethroid resistance, have been identified in the USA, China, Vietnam and Italy [47,[50][51][52].
The involvement of both F1534L and F1534S mutations in resistance to the type I pyrethoids was recently further supported by in vitro functional characterization data [23]. The F1534L substitution was shown to confer similar levels of insensitivity to the previously characterized F1534C, while F1534S seemed to have an even bigger effect. Thus, monitoring the presence of these mutations in countries, like Greece, were type I pyrethroids are used for vector control is of great importance.
Mutation F1534C in the homozygous state has been associated with resistance to permethrin [22] in Aedes mosquitoes. In the analyzed populations, F1534C was observed mainly in heterozygosis, potentially accounting for low resistance levels, due to the recessive nature Nevertheless, the recorded mutation frequencies should not be undermined, as a notable raise of insecticidal pressure may lead to rapid increase in mutation selection potentially hampering the effectiveness of permethrin-based applications for Ae. albopictus control. Additionally, the investigation of P450smediated pyrethroid detoxification, which in combination with target-site mutations has been shown to confer operationally significant resistance levels [54], would be critical complementary evidence facilitating the development of efficient control programs.
Another mutation in the VGSC gene, I1532T, was found in Greece at low frequencies (< 10%) in several surveyed regions with the exception of Patras (western Greece) where the mutant allele reached a frequency of 22.7%. This mutation has been also monitored and reported at considerable frequencies in Ae. albopictus populations from Italy [47,48], Albania [48] and China [51]. Although Kasai et al., 2019 [50] demonstrated a lack of association between this mutation and resistance to pyrethroids, a recent study revealed that VGSC channels carrying this mutation have reduced sensitivity to type I pyrethoids, notably at similar levels to channels harbouring the mutation F1534S [23]. Therefore, it is important to monitor the presence and distribution of this mutation in field populations. We also identified in our samples a small number of individuals carrying both the I1532T and F1534C mutations. In some cases, combination of mutations can have an additive or even synergistic effect resulting in very high levels of resistance [21]. Although the combination of I1532T+F1534C has not been functionally characterized yet, the combinations of I1532T with F1534S and L have been recently shown to have a similar effect to the single mutants in conferring insensitivity to type I pyrethroids [23].
The kdr mutation V1016G, correlated with "stronger" resistance (than F1534C) to type I and type II pyrethroids [22,50] was not detected in any of the analyzed samples from Greece. However, it was recently recorded in Ae. albopictus populations from Hanoi-Vietnam, in Beijing-China and across Italy [48,50,55]. The systematic monitoring of this mutation in Ae. albopictus populations from Greece and elsewhere is strongly recommended, given the primary role of pyrethroids in Aedes mosquito control worldwide [18] and the possibility of low mutation frequencies and/or focal mutation distribution going undetected. Carboxylesterase CCEae3a and CCEae6a gene amplifications linked to temephos resistance [26,27] were detected in 8 out of 11 surveyed regions. Almost 35% of the total Ae. albopictus specimens analyzed, the majority of which were sampled from central and southern Greece, carried more than 1 copy of CCEae3a or both CCEae3a -CCEae6a genes indicating elevated temephos detoxification.
Interestingly out of the 156 Ae. albopictus samples commonly analyzed for the detection of the kdr mutation F1534C and CCEae3a amplification, 25% were found to harbor both mutations, denoting a potential risk for multi-resistance against pyrethroid and organophosphate insecticides.
The observed resistance against malathion may be associated with the high occurrence of amplified

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article [and its supplementary information files].

Competing interests
The authors declare that they have no competing interests   Log-dose probit-mortality data for larvicides tested against Ae. albopictus larvae. The results are compared to the susceptible laboratory Ae. albopictus control strains of other studies [37,38]. N: total number of larvae tested to a range of insecticide concentrations (4 replicates per concentration). LC 50 , LC 95 : lethal concentration (mg/ L) that kills 50% and 95% of the population, respectively. CI: Confidence Intervals. ITU: International Toxic Units; χ 2 : Chi-square testing linearity of dose-mortality response with degrees of freedom (df).

Additional Files
Additional file 1: Table S1. Primers used in this study for regular and real-time PCR. Table S2.
Genotype and allele frequencies of the CHS-1 locus 1043, Greece. Table S3

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.