The entomological survey was conducted across 21 locations in Pune District, Maharashtra, India. The majority of selected sites were located within Pune city taluk, which reported the highest number of arboviral cases during the 2024 outbreak, according to official data from the Maharashtra State Health Department. Site selection was therefore prioritized based on reported case density and entomological feasibility to capture hotspots of arbovirus transmission. While the focus on Pune city enabled intensive sampling in outbreak-affected areas, we acknowledge that fewer sites were included from other talukas. This selection bias was a logistical constraint during the outbreak response phase, and future studies will aim to include broader spatial representation across the district for more generalizable findings.

The surveyed locations included urban, peri-urban, and rural settings. Urban areas comprised Bodpodi, Khadaki, Bhekrai Nagar, SBH Yerawada, Landae Wadi Bhosari, Thergaon, Vishrantwadi, Vaiduwadi Shivajinagar, Sus Goan, Punawal, Dhankawadi, Malwadi, and Maan. Rural sites included Theur, Kunjirwadi, Nere, Boore Budruk, and Morgaon. Peri-urban areas, characterized by transitional landscapes and mixed land use, included Lohegaon, Loni Kalbhor, and Takave. Although Takave is administratively classified as rural, it exhibits peri-urban characteristics due to ongoing urban expansion and infrastructure development.

This classification facilitated stratified sampling and spatial analysis of entomological indices and arbovirus detection across ecologically distinct zones. The geographical distribution of the study sites is depicted in Figure 1.

Immature and adult Aedes mosquitoes were collected from 21 locations in Pune District, Maharashtra, which were affected by a confirmed dengue and chikungunya outbreak in mid-2024. Entomological surveillance and mosquito sampling were conducted during the active transmission period from August to November 2024 (08:00–12:00 h), following notification from the Health Department of Maharashtra. This post-monsoon period is also known to coincide with peak Aedes-borne virus activity in the region. A total of 612 households were randomly surveyed using standardized entomological protocols (Padonou et al., 2023; WHO, 2016; Kumar et al., 2021; Kumar et al., 2021). Adult mosquitoes were collected from indoor and outdoor environments using oral and mechanical aspirators. Simultaneously, a larval and pupal survey was conducted by inspecting all potential water-holding containers located inside and around the surveyed households (Balaji et al., 2019; Shah and Sani, 2011; NCVBDC, 2022; NCVBDC, 2020).

Containers were considered positive if they exhibited the presence of at least one larva or pupa. A household was recorded as positive if it contained at least one such container. Larvae and pupae were collected from positive containers using a dipper or pipette and transferred into labeled 250 mL plastic containers. All samples were subsequently transported to the laboratory for further processing. To evaluate the density and spatial distribution of Aedes breeding, entomological indices were calculated for each sampling location. These included the Container Index (CI), House Index (HI), Breteau Index (BI), and Free Larva Index (FLI), following standard WHO and NCVBDC guidelines (WHO, 2016; NCVBDC, 2022; NCVBDC, 2020; Garjito et al., 2020). In the laboratory, all mosquito specimens were morphologically identified to species level using standard taxonomic keys. Adult specimens were further sorted by species and sex, fed, unfed gravid etc. For molecular analyses, mosquitoes were pooled in groups of up to 20 individuals and stored in sterile 2 mL Eppendorf tubes. Larval and pupal samples were also pooled by collection location and stored at −80 °C (Bikangui et al., 2023; Padonou et al., 2023; Kumar et al., 2021; Balaji et al., 2019; Garjito et al., 2020; Kumar et al., 2012; Badolo et al., 2022; Sanchez et al., 2006).

The pooled samples were homogenized in Minimum Essential Medium supplemented with Fetal Bovine Serum. Then the cellular debris was separated using a centrifuge at 8000 rpm for 5 min, and the supernatant was used for DNA isolation. Extraction of genomic DNA was performed according to the manufacturer’s recommendations using a QIAamp DNA Mini Kit (Qiagen). The concentration and quality of isolated DNA was tested by absorbance at 260/280 using a nanodrop spectrophotometer (Chao and Shih, 2023; Hu et al., 2020; Somia et al., 2023; Zhang et al., 2022).

The Morphologically identified Ae. aegypti samples are further confirmed by amplifying Cytochrome Oxidase I (COI), a mitochondrial gene. The PCR was performed with a total reaction volume of 50 μL comprising the following components: 25 μL of Takara Emerald Green master mix, 15 μL of d. dH2O, 2.5 μL of Forward primer, and 2.5 μL of Reverse primer, with 5 μL of template. The thermal cycle condition for the PCR is as follows: 95 °C for 5 min, 1 cycle Pre-cycle Denaturation step followed by 35 cycles of 94 °C for 30 s, 57 °C for 30 s, and 72 °C for 45 s with a final extension of 72 °C for 10 min. Species identification was achieved by sequencing the amplicon (Chao and Shih, 2023).

The viral RNA was extracted from 140 μL suspension of mosquito pools using a commercial viral RNA extraction kit. The viral RNA was subjected to a two tube real-time RT-PCR assay for detection of DENV serotypes (DENV serotypes 1–4) and CHIKV as described earlier (Alagarasu et al., 2023a,b). ZIKV was detected using a real-time RT-PCR as reported previously (Lanciotti et al., 2008).

A polymerase chain reaction (PCR) assay for amplification of the 16S rRNA gene was used for Wolbachia screening. The amplification was performed with Wolbachia-specific primers 16S WspecF (5′-CATACCTATTCGAAGGGATAG-3′) and 16S Wspec (5′- AGCTTCGAGTGAAACCAATTC-3′) as forward and reverse primers, respectively (Hu et al., 2020; Braig et al., 1998; Braig et al., 1998) which produce an amplicon spanning 438 bp within the 16S rRNA region. The PCR was performed with a total reaction volume of 10 μL comprising the following components: 5 μL of Takara Emerald Green master mix, 3 μL of dH2O, 0.5 μL of Forward primer, and 0.5 μL of Reverse primer, with 1 μL of template. The thermal cycle conditions for the PCR are as follows: 95 °C for 5 min, followed by 1 cycle of pre-cycle denaturation, and then 35 cycles of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 45 s, with a final extension at 72 °C for 10 min. 5 μL of the product was loaded into a 1.5% agarose gel and resolved by electrophoresis (Misailidis et al., 2024; Muharromah et al., 2023).

Three different wsp (Wolbachia surface protein) gene primers have been used to detect super-group A and super-group B. For super-group A, primers used: 328F: (5′-CCAGCAGATACTATTGCG-3′) and 691R: (5′-AAAAATTAAACGCTACTCCA-3′) and for Super-group B, two different primers were used (i) 81F: (TGGTCCAATAAGTGATGAAGAAAC); 522R: (ACCAGCTTTTGCTTGATA) & (ii) 183F: (AAGGAACCGAAGTTCATG); 691R: (AAAAATTAAACGCTACTCCA) (Chao and Shih, 2023; Zhang et al., 2022; Braig et al., 1998; Carvajal et al., 2019).

The PCR was performed with a total reaction volume of 50 μL comprising the following components: 25 μL of Takara Emerald Green master mix, 15 μL of d. dH2O, 2.5 μL of Forward primer, and 2.5 μL of Reverse primer, with 5 μL of template. The thermal cycle conditions for super-group A and 183F;691R are as follows: 95 °C for 5 min, followed by 1 cycle of pre-cycle denaturation, and then 35 cycles of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 45 s, with a final extension at 72 °C for 10 min. For the primers 81F;522R, the PCR mixture compositions are the same, and the thermal conditions for the primer as follows: 95 °C for 5 min, followed by 1 cycle of pre-cycle denaturation, and then 35 cycles of 94 °C for 30 s, 58 °C for 30 s, and 72 °C for 45 s, with a final extension at 72 °C for 10 min. The primers of Supergroup A are expected to amplify a product of approximately 380 base pairs, and the primers of Supergroup B are expected to produce an amplicon of approximately 501 bp and 442 bp. 5 μL of the PCR products with 100 bp ladder was loaded into a 1.5% agarose gel and resolved by electrophoresis (Kittayapong et al., 2000; Hu et al., 2020; Somia et al., 2023; Zhang et al., 2022; Misailidis et al., 2024; Muharromah et al., 2023; Muharromah et al., 2023; Carvajal et al., 2019; Nugapola et al., 2017).

Data were analysed at the cluster level (N = 21) using R software (v4.5.1) with packages including tidyverse, pheatmap, and RColorBrewer. Descriptive statistics were used to summarise entomological indices (HI, CI, BI, MHD), Wolbachia strain prevalence (wAlbA, wAlbB, wPip), and arboviral RNA positivity (DENV, CHIKV, ZIKV). Means and 95% confidence intervals were calculated for continuous variables, while proportions were used for categorical data. Pearson’s correlation coefficients assessed associations between vector indices, Wolbachia prevalence, and arbovirus detection, with significance set at p < 0.05. A clustered heatmap (Figure 2) was generated using the pheatmap package. Variables were z-score standardised, and hierarchical clustering (Euclidean distance, Ward’s method) was applied to rows (clusters) and columns (indicators). The heatmap colour gradient (blue to red) visualised patterns of variation in entomological risk, endosymbiont presence, and arboviral activity across sites.

Comprehensive entomological investigations were conducted across 21 outbreak-affected localities in Pune district to assess Aedes aegypti larval infestation and adult mosquito density. Standard WHO-recommended indices House Index (HI), Container Index (CI), and Breteau Index (BI) were computed, along with the Larval Free Index (LFI) and Man Hour Density (MHD) to estimate adult vector abundance (Figure 3). A total of 612 households were surveyed, with an average of 29.1 houses per cluster (95% CI: 24.1–34.2). Of these, 7.4 houses per cluster (95% CI: 4.9–9.9) were positive for Aedes breeding, resulting in a mean House Index (HI) of 23.7% (95% CI: 17.6–29.9). The number of inspected water-holding containers per cluster averaged 40.0 (95% CI: 31.5–48.5), with 9.7 containers (95% CI: 7.1–12.2) found positive for larvae or pupae, yielding a mean Container Index (CI) of 23.2% (95% CI: 18.9–27.5) and a mean Breteau Index (BI) of 31.7% (95% CI: 25.5–38.0).

At the site level, HI varied markedly across study areas, ranging from 4.55% in Morgaon (rural) to 58.62% in Vaiduwadi Shivajinagar (urban), with 10 localities exceeding the WHO threshold of 10%, indicating active transmission risk. Similarly, CI ranged between 3.85% (Maan) and 33.33% (Bodpodi and Vaiduwadi Shivajinagar). BI was highest in Thergaon (47.22%), Bhekrai Nagar (46.43%), and Landae Wadi Bhosari (46.15%), all urban areas exceeding the 20% risk threshold. Comparative analysis revealed that urban and peri-urban clusters demonstrated consistently higher larval indices than rural sites. However, rural sites such as Kunjirwadi and Theur also exhibited high values, indicating active breeding despite lower container density. The Larval Free Index (LFI) was lowest in Bhekrai Nagar (1.72%), Sus Goan (1.74%), and Morgaon (0.78%), suggesting widespread infestation. Only Theur (LFI = 20.3%), Vaiduwadi Shivajinagar (16.7%), and Khadaki (15.6%) recorded Larval Free Index (LFI) values exceeding the 10% threshold. Since LFI reflects the percentage of inspected households with no larval presence, higher values indicate more effective control of larval habitats and reduced breeding potential of Aedes mosquitoes.

Adult mosquito surveillance yielded a mean Man Hour Density (MHD) of 6.1 (95% CI: 2.1–10.1) across sites. On average, 48.6 adult Ae. aegypti mosquitoes per cluster were collected (95% CI: 16.5–80.7), comprising similar numbers of males (24.9; 95% CI: 5.5–44.2) and females (23.7; 95% CI: 10.5–36.9). Highest MHDs were observed in Kunjirwadi (35.63), Vaiduwadi Shivajinagar (21.38), and Theur (15.00), correlating with high larval indices and denoting sustained vector abundance. Molecular identification revealed that all adult mosquito pools (100%) tested positive for Ae. aegypti (95% CI: 83.9–100.0), with no other Aedes species detected, confirming mono-vector dominance in the outbreak zone. A detailed summary of household, container, pupal, and adult mosquito data including Wolbachia strain prevalence and arboviral screening results is presented in Table 1. Figure 3 displays the comparative entomological indices (HI, CI, BI, LFI, MHD) across all surveyed localities.

Arboviral screening of a total of 93 mosquito pools revealed dengue virus (DENV) RNA in two pools, representing 2.1% of all tested pools and 2 out of 21 study clusters 1.2% (95% CI: 0.0–3.0). Notably, one DENV-2 positive pool from Khadki (ENT 24/MSC 15-2) was also positive for Wolbachia strain wAlbB, suggesting possible co-occurrence of endosymbiont and virus within the same mosquito pool. The second DENV-2 positive pool (ENT 24/MSC 1, Bopodi) was Wolbachia-negative. No pools tested positive for chikungunya virus (CHIKV) or Zika virus (ZIKV) in the current screening. These findings are summarized in Tables 1, 2.

A total of 1,020 adult Ae. aegypti (522 males and 498 females), along with 750 larvae and 250 pupae, were collected from multiple sites and grouped into 93 pools according to developmental stage, sex, and collection locality. All mosquito pools were confirmed as Ae. aegypti through morphological examination and molecular identification based on cox1 gene amplification for one pool. Each pool was individually screened for the presence of Wolbachia using 16S rRNA gene-specific primers. Wolbachia was detected in 11.8% of mosquito pools (11 out of 93 pools tested), representing pooled samples grouped by mosquito sex, developmental stage, and location. The positive detections were distributed across developmental stages, including six male pools, three female pools, one larval pool, and one pupal pool, indicating the presence of Wolbachia across multiple life stages (Table 2).

To determine the genetic lineage, all Wolbachia-positive samples were subjected to supergroup-specific PCR assays. Among the 11 positive pools, two pools were identified as belonging to supergroup A, while the remaining nine pools were assigned to supergroup B. Further strain differentiation within supergroup B revealed that three pools were positive with wsp183F/wsp691R [B(Pip)] primers, and seven pools were positive with wsp81F/wsp522R [B(wAlbB)] primers. These findings demonstrate the coexistence of multiple Wolbachia supergroups within field populations of Ae. aegypti in Pune District, with a predominance of supergroup B, particularly the wAlbB strain. The detection of Wolbachia in both adult and immature stages supports the potential for vertical transmission and suggests natural maintenance of the endosymbiont within local mosquito populations.

Further analysis of Wolbachia prevalence revealed low but variable strain detection across clusters: wAlbA was present in 3.6% of pools (95% CI: 0.0–9.0), wPip/B in 2.7% (95% CI: 0.0–7.7), and wAlbB in 5.0% (95% CI: 0.1–9.8).

Correlation analyses revealed strong relationships among entomological indices. The House Index correlated moderately with the Container Index (r = 0.53, p < 0.05) and Breteau Index (r = 0.61, p < 0.01), while CI and BI were very strongly correlated (r = 0.92, p < 0.001). Man Hour Density correlated moderately with HI (r = 0.51, p < 0.05), suggesting that higher larval infestation was associated with increased adult mosquito abundance.

Among Wolbachia strains, wAlbA and wPip/B were highly correlated (r = 0.88, p < 0.001), indicating possible co-circulation patterns, while wAlbB showed a significant positive correlation with dengue positivity (r = 0.64, p < 0.01). Interestingly, Wolbachia infections did not correlate significantly with larval indices, suggesting their dynamics are not directly dependent on breeding site positivity (Table 3). Figure 2 illustrates a clustered heatmap depicting relationships among entomological indices, Wolbachia infection rates, and arboviral positivity across the 21 clusters. Distinct clustering patterns emerged among both clusters and variables, reflecting heterogeneity in vector infestation and infection profiles. Clusters such as Vishrantwadi, Lohegaon, Bodpodi, and Vaiduwadi Shivajinagar exhibited higher infestation levels, indicated by elevated CI, BI, and HI values (red tones), alongside detectable wAlbB strain and sporadic dengue positivity. Conversely, clusters including Takave, Malwadi, ceMorgaon, and Maan had low HI, CI, and BI values (blue tones), and absence of Wolbachia infection, and no arboviral markers. Intermediate clusters such as Thergaon, SBH Yerawada, and Landae Wadi Bhosari showed moderate entomological indices but low viral detection, indicating possible transitional transmission dynamics or partial infestation control.

Hierarchical clustering grouped HI, CI, and BI tightly, consistent with their biological association in larval and pupal infestation. These indices clustered closely with MHD, representing adult densities. Wolbachia strain proportions with weaker correlations to vector density metrics indicated that Wolbachia prevalence was not directly dependent on mosquito abundance at this stage. Overall, the heatmap highlights spatial heterogeneity in vector infestation and infection, underscoring high-risk clusters and viral transmission potential.

To further investigate the genomic relationships among Wolbachia strains, a pairwise intergenomic similarity matrix was generated based on aligned sequence identities values among 30 wsp gene sequences, including nine NIV-derived sequences and reference sequences from Ae. albopictus, Ae. aegypti, and other arthropod hosts. The resulted clustering heatmap (Figure 6) using VIRIDIC revealed three distinct genomic clusters that correspond closely with the supergroups identified by our NJ phylogeny (Figure 5).

Sequences within supergroup A (ENT-A13 and ENT-A14) shared high intergenomic similarity with each other, but exhibited low similarity (~20%–35%) to sequences belonging to supergroups B and F, consistent with their placement on a distinct phylogenetic branch. Likewise, supergroup B sequences (ENT-B15, ENT-B17, ENT-WA1B-12, ENT-WA1B-21, and ENT-WA1B-22) displayed moderate to high similarity among themselves, while showing low intergenomic similarity (<30–40%) when compared with supergroup A and F members, supporting their monophyletic clustering with the wAlbB lineage. Members of supergroup F formed a separate genomic cluster, characterized by higher within-group similarity but consistently low similarity (<30%) to both supergroup A and B sequences. Overall, inter-supergroup comparisons (A–B, A–F, and B–F) showed marked genomic divergence, directly reflecting the major phylogenetic splits in the wsp tree and confirming that genome-wide similarity patterns reinforce the inferred evolutionary relationships among Wolbachia supergroups. The VIRIDIC similarity matrix quantitatively confirms the three major supergroups (A, B, and F) identified by phylogenetic analysis. The strong agreement between similarity-based clustering and NJ topology supports the conclusion that Ae. aegypti populations in Pune contain a mix of wAlbA-related, wAlbB-derived, and highly divergent F-type Wolbachia, indicating substantial endosymbiont diversity and complex evolutionary interactions in Indian mosquito populations.

The detection of three distinct Wolbachia supergroups wAlbA, wAlbB, and a putative supergroup F in Aedes aegypti populations from Pune reveals notable genetic diversity rarely reported in Indian mosquitoes. While wAlbA and wAlbB have been documented globally, their natural occurrence in Ae. aegypti in India is uncommon. The clustering of one sequence within supergroup F typically found in non-dipteran hosts raises the possibility of horizontal transmission or an underexplored endosymbiotic lineage.

Different Wolbachia strains have varying effects on vector competence. For example, wAlbB is linked to dengue virus suppression and has been used in successful field-based population replacement programs. However, the biological role of supergroup F remains unknown. Future laboratory studies will assess the virus-inhibitory potential of these strains, focusing on their effects on dengue and chikungunya virus replication, mosquito fitness, and vertical transmission. These investigations will support the development of locally tailored Wolbachia-based biocontrol strategies and advance understanding of Wolbachia mosquito virus interactions in Indian ecosystems.

This study has several limitations that warrant consideration. First, the cluster-level sample size (n = 21) is relatively small, potentially limiting the statistical power and generalizability of the findings. Second, mosquito specimens were screened in pools rather than individually, which restricts the ability to confirm co-infection status within single mosquitoes and may underestimate low-prevalence infections. While the minimum infection rate (MIR) was used to estimate Wolbachia and virus prevalence, this method provides a conservative estimate. A maximum likelihood estimation (MLE) approach would yield more robust prevalence estimates, particularly when the number of positive pools is low, and should be considered in future studies. Third, correlation analyses conducted between Wolbachia prevalence, entomological indices, and viral positivity were exploratory in nature. Given the small sample size and multiple comparisons without false discovery rate (FDR) correction, the observed associations may include chance findings and should be interpreted cautiously.

Finally, strain-level identification of Wolbachia was based on wsp gene analysis, which, while widely used, has recognized limitations due to recombination, hypervariability, and homoplasy. These issues can obscure true phylogenetic relationships and may result in misclassification or overestimation of strain diversity. Thus, the phylogenetic clustering results should be viewed as preliminary. Future work incorporating multilocus sequence typing (MLST) or whole-genome sequencing will be essential for more accurate and definitive characterization of Wolbachia strains and their evolutionary relationships.