Gene Amplification and Sequencing of PCR products

In this study, we amplified and sequenced IIS6 dom= ain of vgsc gene from a population = of Cx. quinquefasciatus collected from northeastern India. Genomic DNA from individual adult female �Cx. quinquefasciatus mosquitoes �was extracted using the= SDS extraction and ethanol precipitation method as described previously^{23<= /sup>. The DNA samples were air-dried and suspended in 100 μl of nuclease free water. We designed two primers, Cq1 (5'GTGGAACTTCACCGACTTC 3') and Cq2 (5' GCAAGGCTAAGAAAAGGTTAAG 3') to amplify a fragment of the IIS6 domain of the = vgsc gene, reported to contain the= kdr mutation site. The PCR reaction was carried out according to Sarkar et al.<= sup>17}. PCR reactions were replicated twice, each experiment conducted on a differe= nt day.

PCR products were purified using the QIAquick PCR purification kit (Qiagen). A total of 100 PCR samples, 50 numbers each from knockdown resistant and knockdown susceptible phenotypes were sequenced on = both strands using an ABI automated DNA sequencer. Sequences were analyzed using BLAST program (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Sequences were submitted to the National= Center for Biotechn= ology Information Open Reading Frame Finder (http://www.ncbi.nlm.nih.gov/projects/gorf/<= /a>) to determine the coding regions and introns.

Single nucleotide polymorphisms in IIS6 domain of = vgsc were detected as sequence differences in multiple alignments using CLUSTALW (http://align.genome.jp/). Electrophoregrams were visually inspected using BioEdit (http://www.mbio.ncsu.edu/BioEdit/BioEdit.html) and heterozygotes w= ere identified²⁴. SNPs were identified as transitions or transversio= ns in coding and non-coding regions. Nucleotide diversity, polymorphisms, divergence, gene flow, and genetic differentiations among the populations studied were analyzed using DnaSP 5.00.02 (http://www.ub.es/dnasp)^25.

Analysis of Data and Molecular Evolutionary Theories

Most intra-population (i.e., the single population tested for neutrality in this study) data analyses � including estimates of= DNA polymorphisms, nucleotide diversity (π) number of segregating sites (S= ), number of haplotypes (h), haplotype diversity (Hd) � were performed among 1= 00 sequences of the vgsc gene by subdividing the gene into functional domains (exons and introns). The data = were analyzed using DnaSP 5.00.02. We analyzed the neutral model/hypothesis (alternatively known as the null model; states that the vast majority of evolutionary changes at the molecular level is caused by random drift of selectively neutral mutants²⁶), to infer if natural selection is acting upon the analyzed vgsc g= ene. Tajima�s D test²⁷, Fu and Li�s D* and F*²⁸ tests and = Fu�s Fs test²⁹ were performed to determine whether the distribution of nucleotide variation within the samples was consistent with the neutral mod= el. Tajima�s D test statistics is defined as the standardized difference between two estimators of the population mutation rate parameter θ (=3D 4 N = 56; for an autosomal region of diploid individuals; where N is the effective population size, and μ is the per-gene or per-site per-generation muta= tion rate). One estimator based on the number of segregating sites and other on = the average number of pairwise nucleotide differences, which should be equal un= der the neutral mutation model and should differ when natural selection affects= the genomic region. Fu and Li�s D* test compares two estimators of the populati= on mutation rate parameter θ, based on the differences between the number= of singletons (mutations appearing only once among the sequences), and the tot= al number of mutations²⁸. Fu and Li�s F* test statistics is based on the differences between the number of singletons and the average number of nucleotide differences between pairs of sequences^{28, 30, equation 10. The Fs test statistic29, equation 1 is based on the haplotype (g= ene) frequency distribution for a given value of θ derived from the average number of pair wise nucleotide differences31, equations 19-21. We performed Coalescent simulations to estimate confidence intervals and exact P-values (10000 interactions) for Tajima�s D, Fu and Li�s D* and F* test, nucleotide diversity, linkage disequilibrium, number of haplotype and haplo= type diversity test. These simulations were conditioned on the observed number of segregating sites (S), and thus population size is not a factor.}

Results and Discussion

Bioassays were carried out using 4% DDT and mortalities were recorded after 24 hour exposure. The mortality of Cx. quinquefasciatus with DDT ranged from 11.9% to 41.25% whereas mortality in S�Lab insects was estimated at 91.2%. This bioassay result suggests that Cx. quinquefasciatus population tested is highly resistant to DDT across all study sites. The details of insecticide resistance and/or susceptibility status, and detoxif= ying enzyme profiles of this population were described in our other publication<= sup>2.

Detection of kdr mutatio= n in wild population of Culex quinquefasciatus

We randomly sampled 50 DDT-resistant and 50 susceptible mosquitoes in search of polymorphisms in the insecticide-binding segment of vgsc gene. We found = two distinct sequence variants (i.e., knockdown susceptible or kds and knockdown resistance or kdr) after sequencing 100 samples. These two distinct sequence variants were submitted in GenBank under accession number FJ182226 and FJ970025. Figure 1 displays the sequence alignment of the knockdown= -susceptible and knockdown resistant population, which confirms the presence of the polymorphic site at position 127 (TTA to TTT) that induces the substitution of leucine to phenylalanine in resistant mosquitoes, as previo= usly reported by others in the same species ^{14, 15}. However, we did not find any A to C mutation as reported by Wondji et al.¹⁵

Neutrality tests and evo= lutionary pattern of knockdown resistance

The sequences of the vgsc gene consisted of one exon and one intron region. We found four parsimony variable sites (127, 155, 157, and 176) in these sequences; = out of these one was found in the exon region (A127T, kdr mutatio= n) and three in the intron-2 region (Table 1). Polymorphic sites in the intron= ic sequences co-segregate with the kdr allele. Statistical analysis of polymorphism for the entire sequenced region and each functional domain of = the vgsc gene are summarized in Table = 1. In vgsc, nucleotide diversity in the = exon (π/site=3D0.0033) and intron (π/site=3D0.0038) was found to be al= most similar when compared with the entire gene (π/site=3D0.0037). However,= we observed a higher θ/site value for the intron region in comparison to = the exon and entire sequence (Table 1). This is probably due to higher mutation rate= per nucleotide per generation in the intronic region.

Figure 1. Alignment of nucleic acids and corresponding amino acids of IIS6 domain of the para-type sodium channel gene from= two representative sequences of knockdown susceptible (FJ182226) and resistant (FJ970025) strain of Culex quinquefasciatus. Note the A to T change at nucleotide 127 and leucine = (L) to phenylalanine (F) at amino acid at 39 in the resistant strain. Alignment also displays four polymorphic sites: at position 155 (C-A), 157 (T-C) and = 176 (A-G) in the intron-2 region.

The polymorphic sites in different intronic region= s of the vgsc gene vary depending on= the mosquito species. We observed three polymorphic sites in intron-2 of Cx. quinquefasciatus whereas Pinto et al.²⁰ reported eight polymorphic sites in intro= n-1 of the An. gambiae population t= hat explain the distinctive pattern of polymorphisms in different intronic regi= ons of vgsc in different populations.�

We performed Fisher�s exact test for independence between sites (with Bonferroni correction) and Kelly�s ZnS to test linkage disequilibrium³². Only parsimony informative polymorphic sites w= ere considered for the analysis and results are presented in Table 1; there is = no evidence of significant nonrandom association between nucleotide variants at different polymorphic sites (P > 0.1). Recombination per gene in vgsc domain-II was very low (R=3D0= .001). The minimum number of recombination eve= nts R_M=3D0. In the presence of non-significant linkage disequilibrium and very low recombination, it is not unexpected that the estimates of haplotype number = and corresponding haplotype diversity for vgsc do not deviate from the neutral expectations [number of haplotype, h=3D3; haplotype diversity, Hd=3D0.537 (�0.09) with P=3D0.316] (Table 1 and 2). Nucleotide variations in synonymous and nonsynonymous sites in vgsc are presented in Table 3. We = found no singleton and synonymous mutation (Table 3).

One method for revealing the influence of selection on a sequence is by examini= ng the distribution of segregating variants at that locus in a population and testing this distribution against the neutral model³³. We have u= sed the D statistics^{27, 28} to examine the hypothesis that all substitutions at the locus are neutral. Several possibilities were explored: using the whole region sequenced, only the coding region, or only the intro= n-2. Where appropriate, the significance of the test or parameter has been verif= ied by coalescent simulation (10000 repeats) (Table 2). The detailed results of neutrality tests including coalescent simulation are presented in Table 1 a= nd Table 2. Upon analyses of segregating variation, no statistics (D, F* or Fs) provide evidence for a significant departure from neutral evolution. The po= wer of neutrality test is limited in the presence of recombination³⁴, but because recombination was very low, this may explain the lack of significant departure from neutrality. Overall, the evolutionary pattern of intra-population distribution of variability in the domain-II of the vgsc gene is consistent with the n= eutral expectation; hence there is no evidence that positive Darwinian selection h= as recently influenced, or is currently influencing, nucleotide variation in t= his region of the genome. Hence, there is a possibility that knockdown resistan= ce (kdr), associated with polymorphism= in domain-II of the vgsc gene, may= be caused by random drift of selectively neutral mutants. Should we then concl= ude that selection did not affect the nucleotide variation in vgsc and evolution of the �kdr trait is largely due to genetic drift rather than natural selection? The present data are obviously not sufficient to tell apart the various scenarios that could have led to the present structure, because information at many independent loci is required= to make strong inference on past demographic factors that could affect the selection processes.

On the other hand, the neutral theory does not rule out the role of natural selection in certain scenarios of adaptive evolutio= n - such patterns are expected from neutrality. Hence, the molecular evolution of th= e vgsc domain may be dominated by selectively neutral evolution of segregating variants (polymorphic sites) at genomic level, but at the phenotypic level, changes in knockdown resistance were probably dominated by natural selection rather than sampling drift.

Our observations provoked a hypothesis of possible evolutionary pattern of kdr allele in Cx. quinquefasciatus. Further critical study with whole vgsc sequences is needed to addres= s this problem. At this point, we have no means to claim any definite evolutionary mechanisms controlling knockdown resistance in mosquito by using the data f= rom the present study. We also suspect that the evolutionary puzzle regarding t= he origin of knockdown resistance in mosquitoes can be addressed successfully = by thorough analysis of intronic polymorphism in the entire vgsc gene between samples from different parts of the world. The study of a molecular evolutionary pattern of the insecticide-binding domain= of the vgsc gene in Cx. quinquefasciatus, described here, has important academic implications and opens a scope for a= future study with a holistic approach to identify definite evolutionary forces at = work in knockdown resistance.