A broad-spectrum Chromobacterium biolarvicide formulation disrupts mosquito development through gut damage and apoptotic cell death and compromises vector competence

Abstract

Mosquito resistance to chemical insecticides, the cornerstone of current vector control strategies, continues to undermine disease prevention efforts. We have developed a stable non-live larvicidal microbial biopesticide pellet formulation of Chromobacterium species Panama (Csp_P) that effectively attracts larvae and demonstrates potent activity against a broad range of mosquito disease vectors by exerting toxicity to all stages from eggs to adults upon exposure in the larval aquatic environment. Larval ingestion of Csp_P pellets causes irreversible midgut tissue damage, disrupting development from the larval stage and reducing the fitness and vector competence of surviving adults. The observed midgut pathology differs from that induced by conventional biopesticides and is accompanied by molecular markers of stress-mediated apoptosis and autophagy. These findings provide mechanistic insights into Csp_P-induced larval killing and highlight the potential of this formulation as a biologically derived, scalable tool for integrated mosquito management programs.

Chromobacterium Csp_P guar gum-based pellets induce irreversible gut damage, apoptosis, and high larval mortality across mosquito vectors, causing developmental collapse, reduced adult fitness, and decreased malaria and dengue transmission potential.

Introduction

Mosquito vector-transmitted parasite, viral, and filarial pathogens pose a significant threat to public health worldwide. Although extensive disease control measures implemented over the past 30 years have led to a reduction in disease burden, the toll on human lives remains substantial¹. Recent malaria statistics indicate an alarming 247 million reported cases in 2021, resulting in approximately 593,000 deaths, with a significant number of cases in children under the age of 5 years^1,2. Mosquito-transmitted arboviral pathogens also significantly affect human health, in particular dengue, chikungunya, and Zika viruses, which are transmitted by Aedes and Culex mosquitoes and cause millions of infections every year. The burden of arboviral diseases has risen significantly over the last 15 years as a result of climate change, mosquito population dispersal through global trade and transport, and urbanization^3–5.

The lack of effective vaccines and drugs for most arboviral pathogens and the emergence of malaria parasite resistance to drugs, along with their cost and logistical complexities of administration, lead to chemical insecticide-based vector control becoming a major approach for suppressing these diseases^6,7. Presently, pyrethroids and organophosphates are the most commonly used insecticides, and they generally kill adult mosquitoes by attacking their nervous systems⁸. Unfortunately, successes achieved with insecticidal control of mosquitoes have been greatly dampened by the emergence of resistance^9,10. More recently, the implementation of transgenic technology to release insects carrying a Release of Insects carrying a Dominant Lethal (RIDL)¹¹ gene or mosquitoes harboring the endosymbiont Wolbachia, designed to spread cytoplasmic incompatibility among their progeny, has been used to suppress mosquito populations and pathogen transmission¹². However, these mosquito control measures are still at an early stage of development and can only be implemented in certain disease-endemic settings^13,14.

These concerns, together with the environmental impact of chemical insecticides, have sharpened the current focus on the development of eco-friendly and cost-effective natural mosquitocidals^15–17. The development of Chromobacterium sp. Panama (Csp_P) as a biologically derived larvicide aligns with the principles of environmental safety and Integrated Vector Management (IVM), as recommended by the World Health Organization¹⁸. Its natural origin, targeted mode of action, and effectiveness against insecticide-resistant mosquitoes make it an attractive addition to vector control programs¹⁸. Moreover, as part of an IVM strategy, Csp_P-based interventions can be integrated with other tools such as source reduction, chemical insecticides, and community-level participation to enhance sustainability and mitigate insecticide resistance¹⁸. Only a few biologically derived insecticides or biopesticides are available for mosquito control, such as Bacillus thuringiensis israelensis (Bti)¹⁹ and Lysinibacillus sphaericus, which have proven effective in reducing the risk of mosquito-borne diseases because they can easily be administered in stagnant water in both rural and urban settings¹⁹. The highly stable spores of these bacteria form a crystal protein that shreds the mosquito gut when the spores are ingested. Notably, the toxins produced by L. sphaericus show a greater specificity to mosquitoes of the Culex and Culiseta genera, making it less effective against Aedes or Anopheles species^20,21. Entomopathogenic fungi such as Metarhizium anisopliae that can kill a variety of mosquito species are also at an exploratory stage for mosquito control²². In addition to microbial larvicides, insect growth regulators (IGRs)—such as Methoprene and pyriproxyfen—are widely used chemical-based larvicides that interfere with larval development by mimicking or inhibiting juvenile hormones²³.

Emerging insecticide resistance in mosquito populations is of particular concern because no single insecticide can be used for long-term control, severely complicating ongoing disease control efforts^24–26. Microbial biopesticides offer a promising new alternative to chemical insecticides because they frequently have lesser non-target effects and rapidly break down in the environment.

Chromobacterium species Panama (Csp_P) is a gram-negative soil bacterium that displays several unique and useful features for development as a mosquito control agent²⁷. Ingestion of a sub-lethal dose of Csp_P by adult An. gambiae mosquitoes has led to a reduced susceptibility to P. falciparum infection through the production of a Plasmodium-killing metabolite²⁷. The Csp_P biopesticide also produces an aminopeptidase that inhibits dengue virus in the Ae. aegypti vector through digestion of the viral envelope E protein²⁸. Critically, ingestion of higher doses of Csp_P by adults through an attractive toxic sugar bait (ATSB) results in potent killing of various mosquito vectors, including those that are resistant to all major classes of chemical insecticides¹⁸. Interestingly, the ATSB formulation of Csp_P showed synergist activity with currently used chemical insecticides and would render insecticide-resistant mosquitoes susceptible to these insecticides upon exposure to sub-lethal Csp_P doses¹⁸. Caragata et al.²⁹ demonstrated the larvicidal efficacy of Chromobacterium sp. Panama (Csp_P) using gelatin- and agarose-based pellet formulations²⁹. While promising, these formulations had practical limitations; gelatin pellets dissolved rapidly in water, and agarose-based pellets are costly and require a complex preparation method involving heating. In this study, we build upon these earlier findings by developing a food-grade guar gum-based pellet formulation of non-live Csp_P. This binder offered a simpler preparation method and improved aqueous stability and shelf-life without compromising larvicidal efficacy across multiple mosquito species. Additionally, our study also addresses gaps in the understanding of Csp_P’s sub-lethal and developmental effects. We found that the presence of Csp_P pellets in larval aquatic environment inhibited egg hatching, and ingestion of sub-lethal doses of Csp_P pellets by larvae resulted in reduced pupation rates, decreased adult longevity, and diminished fecundity and fertility in all tested mosquito species. Additionally, adults emerging from exposed larvae exhibited significantly reduced vector competence for both Plasmodium falciparum and dengue virus. Mechanistic analyses revealed that larval ingestion of Csp_P induces midgut tissue damage and activates stress-mediated apoptotic and autophagic pathways. Together, these findings expand upon prior work by revealing both the extended physiological impacts of Csp_P exposure and the cellular mechanisms underpinning its mosquitocidal effects.

Results

Larvicidal pellet formulation

Selection of binder

Our goal was to formulate a non-live Csp_P-based larvicide as environmentally friendly pellets that would retain their integrity and activity for sustained periods, both under ambient storage conditions and after addition to water bodies. For this purpose, we tested the ability of several natural binders to form a cohesive, stable pellet with the non-live Csp_P powder and an appropriate larval attractant (Fig. 1A). We also evaluated the effectiveness of various larval baits in attracting larvae to feed on the pellets in an aqueous environment. In experiments assessing the water stability of various binders, we found that pellets containing 400 mg/mL (per total pellet volume) of non-live Csp_P powder, 100 mg/mL of filtered fish meal suspension (as an attractive bait, described below), and 200 mg/mL of guar gum (as a binder) remained stable in water for over a month at both 28 °C and 40 °C. Pellets composed of 60% (600 mg/mL) candelilla wax and 40% (400 mg/mL) paraffin wax were stable in water for 2 months at 28 °C but became unstable at 40 °C (Supplementary Table 1). In contrast, other tested binders, including various gums and waxes such as xanthan gum, gum arabic, edible gum, candelilla wax, carnauba wax, soy wax, and paraffin wax, maintained stability for only 7–10 days at both temperatures (Supplementary Table 1). Based on the prolonged stability at both 28 °C and 40 °C, we selected a 200 mg/mL guar gum formulation for the preparation of Csp_P pellets.

Fig. 1. Schematic methodology for preparing the non-live Csp_P pellets and assessing larval feeding behavior toward pellets.

A Non-live Csp_P powder was mixed with a constant concentration of various attractive larvicide baits. Both Csp_P powder and larvicide baits were then formulated with different natural binders (waxes/gums) into small-sized pellets, followed by air drying and assessment of their stability in aqueous environment and their larvicidal activity. B Bait choice assays showing the majority of Ae. aegypti larvae are attracted and respond to filtered fish meal/Csp_P pellets as compared to other baits, indicating that fish meal is an efficient, attractive larvicide bait. Data points corresponding to independent biological replicates are overlaid on the graphs shown in Supplementary Fig. 1a. Raw data are presented in Supplementary Table 1. C Images show the feeding of Cx. quinquefasciatus larvae on Csp_P pellets at various time points, and images of fluorescein (green) in the abdomen of Cx. quinquefasciatus and Ae. aegypti larvae while feeding on fluorescein-labeled Csp_P pellets at various times. Parts of illustrations were created using BioRender (BioRender.com; https://BioRender.com/qp46r62).

Selection of attractive bait

To identify a potent larval attractant, we performed bait choice assays that involved monitoring larval preferences by placing pellets formulated with different attractants in the same larval pan containing 300 larvae and observing their responses over a 30- to 40-min period. Assessment of the attractiveness of various attractants to Ae. aegypti and Culex quinquefasciatus larvae in 200 mg/mL guar gum pellets revealed that a filtered (0.2 micron-filtered) water suspension of fishmeal exhibited the greatest attractiveness over a 40-min period, compared to the other attractants tested, including chemical attractants (0.2% phenyl ethanol and 0.1% octanol) and non-chemical attractants (Tetramin fish food, dried cat food, poly hydrogels, guar gum alone, and fish oil) (Fig. 1B and Supplementary Fig. 1).

Ingestion of Csp_P pellets at concentrations of 50 and 100 mg/mL was confirmed by incorporating 0.1% fluorescein dye into the pellets and visualizing fluorescent content in the larval gut by microscopy (Fig. 1C). Based on these observations, we decided to use a filtered fish meal supernatant as the attractive bait.

Killing dose determination for Anopheles, Aedes, and Culex mosquito species

Next, we evaluated the larvicidal efficacy of non-live Csp_P pellets against Ae. aegypti Liverpool, permethrin-resistant Ae. aegypti, An. gambiae, An. stephensi, and Culex quinquefasciatus larvae (Fig. 2A–D). In each of these experiments, pellets containing guar gum (200 mg/mL), filtered fish meal supernatant (100 mg/mL), and various concentrations (50 mg/mL, 100 mg/mL, 200 mg/mL, and 400 mg/mL) of non-live Csp_P powder were fed to L2-L3 instar larvae. Control pellets contained guar gum (200 mg/mL), filtered fish meal (100 mg/mL), and a non-live E. coli powder preparation (400 mg/mL).

Fig. 2. Larval bioassays showing Csp_P pellets’ larvicidal activity against various mosquito species.

Small-scale larval survival assays (n = 30 larvae per group) were performed for five mosquito strains: Anopheles stephensi (A), An. gambiae (B), Aedes aegypti (C), Culex quinquefasciatus (D), and a permethrin-resistant Ae. aegypti line (E). Larvae were exposed to pellets containing 50, 100, 200, or 400 mg/mL of Csp_P powder. Survival was recorded daily for 5 days. Each assayed curve point represents the mean ± SEM of three independent biological replicates. Error bars represent SEM. Large-scale larval bioassays (n = 250–500 larvae per group per replicate) were conducted in 5–8 L trays using An. stephensi (F), An. gambiae (G), Ae. aegypti (H), and Cx. quinquefasciatus (I). Percent mortality was calculated 72 h post-treatment. Bars show mean ± SEM across three independent biological replicates. Individual replicate values are overlaid as dots. Data were analyzed using the Kruskal–Wallis test followed by Dunn’s multiple comparisons test against control; **p = 0.005, ***p = 0.0003, and ****p < 0.0001. J–M Irreversible larvicidal activity of Csp_P pellets was assessed by removing pellets after 24 h and continuing larval observation for 5 days. Percent mortality is shown for An. stephensi (J), An. gambiae (K), Ae. aegypti (L), and Cx. quinquefasciatus (M). Sample sizes ranged from 30 to 90 larvae per group, as indicated. Bars represent mean ± SEM from three biological replicates, with individual replicate data points shown. Data were analyzed using unpaired two-tailed t-tests (p = 0.005). All assays used independent biological replicates (separate larval batches on separate days). Error bars represent SEM throughout. Nonparametric tests were applied to survival and mortality data, which represent binary outcomes. Data points corresponding to independent biological replicates are overlaid on the graphs shown in Supplementary Fig. 2. Raw data are presented in Supplementary Data.

In initial small-scale larvicidal bioassays (n = 30 larvae per cup), An. stephensi and An. gambiae larvae exhibited the highest mortality rates (≥80%) over a 2–4-day period, whereas Ae. aegypti and Cx. quinquefasciatus larvae reached similar mortality levels within a 3–5-day period (Fig. 2A–D). Notably, An. gambiae and An. stephensi larvae exhibited 50-55% mortality when exposed to a 50 mg/mL Csp_P dosage, with an increased 70-95% mortality at 100 and 200 mg/mL Csp_P powder concentrations, respectively. A lower mortality rate of 50-70% was recorded at the highest concentration of 400 mg/mL at 3 days, suggesting reduced larvae feeding propensity at this high dosage (Fig. 2A, B). Similarly, a dose-dependent mortality was observed for Ae. aegypti and Cx. quinquefasciatus larvae, with mortality rates ranging from 80 to 90% at 50 mg/mL, 75 to 95% at 100 and 200 mg/mL, and 80 to 100% at 400 mg/mL Csp_P powder 3 days after exposure (Fig. 2C, D). No significant mortality was observed in larvae exposed to control E. coli pellets for up to 6 days. To examine the dose efficiency and pellet performance in larger water volumes, we performed large scale larval bioassay. In these larvicidal assays (n = 250–500) after 48 h of exposure, An. gambiae and An. stephensi larvae exhibited 70–100% mortality when the larvae were provided with two pellets containing 100 mg/mL, 200 mg/mL, or 400 mg/mL of non-live Csp_P powder, and about 60% mortality when provided with pellets of 50 mg non-live Csp_P powder/mL (Fig. 2F, G). Similarly, after 48 h, Ae. aegypti and Cx. quinquefasciatus larvae showed a 40% mortality at 50 mg/mL, 60–80% at 100 mg/mL and 200 mg/mL, and 95% mortality at 400 mg/mL doses of Csp_P powder in pellets (Fig. 2H, I). Overall, we observed significant larval mortality with all the four doses of Csp_P pellets when compared to control E. coli pellets.

We also examined the larvicidal activity of Csp_P pellets against Ae. aegypti Liverpool permethrin-resistant larvae and found 90-100% mortality across all tested concentrations (50, 100, 200, and 400 mg/mL) by day 4 of exposure (Fig. 2E).

Larvicidal activity is irreversible

We examined whether larvae that survived after ingesting 50 mg/mL dose of Csp_P pellets could resume their normal development when transferred to standard rearing conditions with regular larval food pellets. Despite 50-70% mortality caused by the 50 mg/mL Csp_P pellets, we collected 30-50% surviving larvae at the end of the 3-day exposure in multiple independent experiments. These surviving larvae were then allowed to feed on regular Tetramin fish food. Notably, 75% of Ae. aegypti, 65% of Cx. quinquefasciatus, 80% of An. gambiae and 75% of An. stephensi larvae were found dead after 2 days in the Tetramin fish food-containing larval aquatic environment (Fig. 2J–M). These results indicate that the larvicidal effect of Csp_P pellets is irreversible.

Sub-lethal exposure compromises development of surviving larvae

We next examined the pupation rate of pools of 250–500 larvae exposed to 50 mg/mL Csp_P pellets. Interestingly, a significant number of surviving larvae failed to pupate in any of the tested mosquito species after exposure to Csp_P pellets. In contrast, normal pupation occurred among larvae treated with the control E. coli pellets, between day 8 and day 10 post-exposure (Fig. 3). The mean pupation time was also delayed across all tested mosquito species exposed to Csp_P pellets. For instance, the mean time to pupation was 10 and 12.5 days for surviving An. stephensi and An. gambiae larvae, respectively, when exposed to 50 mg/mL Csp_P pellets, whereas it was 9 days for the control group (Fig. 3A, B). Similarly, surviving Ae. aegypti and Cx. quinquefasciatus larvae exposed to 50 mg/mL Csp_P pellets showed mean pupation times of 11.5 days and 9.5 days, respectively, as compared to 7–8 days for the controls (Fig. 3C, D). In the case of all tested mosquito species, the mean pupation rate of surviving larvae also varied according to the dosage of Csp_P pellets used. Exposure to Csp_P pellets at 50 mg/mL resulted in pupation rates of 37.5 ± 16.4 and 57 ± 11.4% in pools of 500 larvae of An. stephensi and An. gambiae, respectively, as compared to a control rate of 97 ± 2.5% (Fig. 3A, B). However, doses of 100 mg/mL or 200 mg/mL of Csp_P pellets resulted in less than 15% pupation for both surviving Anopheles larvae (Fig. 3A, B). Similarly, the percent pupation of Csp_P pellet-treated surviving Ae. aegypti larvae was drastically reduced by various dosages, as observed with Anopheles larvae. In groups of 500 exposed larvae, we observed 66 ± 5.9, 18 ± 6.5, and 2.8 ± 1.8% pupation with 50, 100, and 200 mg/mL doses of Csp_P pellets, respectively (Fig. 3C). Likewise, with a dosage of 50 mg/mL Csp_P pellets, pupation was 47 ± 3.7%, as compared to 92.9 ± 4.6% for control pellets; at concentrations above 50 mg/mL, pupation rates below 30% were found for Cx, quinquefasciatus (Fig. 3D).

Fig. 3. Csp_P ingestion by larvae of various mosquito species affects the rate of pupation and development of adult mosquitoes.

A–D Low and delayed pupation in the larvae of multiple mosquito species when exposed to Csp_P pellets; n = 250–500 larvae/cup; (One-way ANOVA test; p = 0.01 in case of An. stephensi). E–H Pupal maturation to adult mosquitoes in Csp_P-exposed pupae, n = 250–500 larvae/cup. Countable pupae formed in the Csp_P pellet-treated group, but most of them died and did not develop into adults. Data are representative of three independent experiments, each with two replicates. Data points corresponding to independent biological replicates are overlaid on the graphs shown in Supplementary Fig. 2. Raw data are presented in Supplementary Data.

Since we observed significantly reduced pupation after larval exposure to Csp_P pellets in all tested mosquito species, we further investigated the emergence of adult mosquitoes from the surviving pupae collected in each replicate of independent experiments. In these experiments, we collected live pupae resulting from larval exposure to each of the four dosages (50, 100, 200, and 400 mg/mL) of Csp_P pellets; these pupae were allowed to emerge for 48 h. Notably, for An. gambiae, 60 ± 19.1, 82.4 ± 4, 90 ± 2.8, and 96.4 ± 2.5% pupal mortality were observed at 50, 100, 200, and 400 mg/mL dosages, respectively, as compared to 14 ± 4.53% for the control group (Fig. 3E). Similarly, in An. stephensi, pupal mortality rates of 50.9 ± 11.42, 89.5 ± 2.78, 98.1 ± 3.2, and 92.3 ± 3.22% were recorded at 50 mg/mL, 100, 200, and 400 mg/mL dosages, respectively, as compared to 10 ± 6.3% for the control group (Fig. 3F). Likewise, in Ae. aegypti, 37 ± 9.6, 71.4 ± 10, 87.9 ± 12.1, and 92.1 ± 9% of the pupae were found dead at 50, 100, 200, and 400 mg/mL dosages, respectively, as compared to controls (1.6 ± 0.81%) (Fig. 3G). In the case of Cx. quinquefasciatus, 27.3 ± 9.6, 61 ± 6.95, 78.0 ± 9, and 95.6 ± 5.8% pupal mortality was noted at 50, 100, 200, and 400 mg/mL, respectively (Fig. 3G, H). Furthermore, the pupal mortality resulted in a markedly reduced number of emerged male and female mosquitoes across all tested species (Fig. 3E–H). For instance, at 50 mg/mL, adult emergence rates were 49 ± 15.8% in An. gambiae, 42 ± 19.3% in An. stephensi, 62 ± 11.2% in Ae. aegypti, and 72 ± 15.8% in Cx. quinquefasciatus. In addition, fewer than 15 ± 6.5% of the adult mosquitoes were observed across all species at the higher doses of 100, 200, and 400 mg/mL of Csp_P pellets (Fig. 3E–H). Notably, a significant portion of adult mosquitoes that emerged at doses of 100, 200, and 400 mg/mL appeared to be impaired in their ability to fly.

Larval exposure compromises adult fitness

We also evaluated the fitness of the adult mosquitoes that emerged from larvae exposed to Csp_P pellets, in terms of their longevity, blood-feeding propensity, egg laying capacity (fecundity), and fertility. Longevity assays revealed that larval exposure to Csp_P pellets significantly reduced the longevity of all tested mosquito species when compared to the control group (Fig. 4A–D). For instance, adult Anopheles mosquitoes lived an average of 9 and 11 days, while Ae. aegypti and Cx. quinquefasciatus lived for 11 and 15 days, respectively, after emergence. These survival times were significantly lower than those of the control group (Fig. 4A–D).

Fig. 4. Larval exposure to Csp_P pellets consistently reduces adult longevity, blood-feeding propensity, fecundity, and fertility.

Life history traits were assessed for larvae that had ingested Csp_P pellets. A–D In the case of all four mosquito species, the longevity of adults emerging from the larvae that ingested Csp_P was significantly decreased when compared to control adult mosquitoes. Survival graphs represent the mean ± SEM across three independent biological replicates. The number of mosquitoes used in each independent biological replicate was 30–40 in both control and Csp_P treated group. Log-rank (Mantel–Cox) test; p < 0.0005. E–H No significant difference between the blood-feeding propensity of controls and the Csp_P-exposed larvae was observed in the case of Anopheles, Ae. aegypti and Cx. quinquefasciatus mosquitoes; Data were analyzed using a nonparametric Mann–Whitney U test; p = 0.1. Bars show mean ± SEM across three independent biological replicates. Individual replicate values are overlaid as dots. I–L Fecundity was assessed after a blood meal obtained from a mouse. For all four mosquito species, the Csp_P-exposed group had a significantly reduced fecundity in the survived female mosquitoes as compared to the controls. Mann–Whitney test; ***p = 0.0002. M–P Fertility was assessed for mosquitoes that had laid eggs in the fecundity assay above. Again, the fertility of the mosquitoes exposed to Csp_P pellets was significantly lower than that of the controls. Mann–Whitney test; ****p < 0.0002. Data are representative of three independent experiments, each with two replicates. Q Ae. aegypti eggs were allowed to hatch in presence of Csp_P pellets. Notably, egg hatching was significantly reduced after 72 h of exposure. Unpaired t- test; **p = 0.005. R Ae. aegypti egg hatching was remarkably reduced in water previously exposed to Csp_P pellets, as indicated by the significantly low number of live larvae that were counted. Unpaired t-test, **p = 0.006. Data points corresponding to independent biological replicates are overlaid on the graphs shown in Supplementary Fig. 2. Raw data are presented in Supplementary data. Bars show mean ± SEM across three independent biological replicates. Individual replicate values are overlaid as dots. Illustrations were created using BioRender (BioRender.com; https://BioRender.com/qp46r62).

No significant difference in blood-feeding propensity was observed between Csp_P-exposed and control groups in adult female mosquitoes (Fig. 4E–H).

Adults emerging from Csp_P pellet-exposed larvae laid a significantly lower number of eggs than did their respective control groups. For An. stephensi and An. gambiae mosquitoes, the median number of eggs laid per mosquito was 12 and 0, respectively, for the Csp_P pellet-treated groups, as compared to 45 for the An. stephensi control group and 37.5 for the An. gambiae control group (Fig. 4I–J). Similarly, individual Ae. aegypti and Cx. quinquefasciatus mosquitoes laid significantly fewer eggs, with medians of 45 and 0, as compared to control groups that laid 78 and 114.5, respectively (Fig. 4K, L). These results indicate that Csp_P pellet ingestion at the larval stage significantly reduces the fecundity of adult mosquitoes.

The fertility of adults that emerged from Csp_P pellet-exposed larvae was also compromised and produced median numbers of live larvae of 11, 11, 13, and 57 for An. stephensi, An. gambiae, Ae. aegypti, and Cx. quinquefasciatus, respectively (Fig. 4M–P). Conversely, for the control groups, the median numbers of live larvae were 76, 44, 69, and 109 for An. stephensi, An. gambiae, Ae. aegypti, and Cx. quinquefasciatus, respectively (Fig. 4M–P). These results indicate significant differences in female fertility between the Csp_P pellet-exposed and control groups of mosquitoes.

Sub-lethal exposure decreases Ae. aegypti larval emergence from eggs

To investigate the effect of Csp_P pellets in the larval aquatic environment on the emergence of larvae from eggs, we placed a strip of Ae. aegypti Liverpool eggs in a tray with 500 mL water containing a 100 mg/mL Csp_P pellet that is expected to release insecticidal factor (s) into the water. The live larvae were manually counted and showed significant differences after 48 h of Csp_P pellet exposure, as compared to control E. coli pellet exposure. Specifically, we observed that only 10 ± 4.3% of eggs hatched in the presence of Csp_P pellets, whereas in control, 90 ± 1.2% of eggs hatched (Fig. 4). Similarly, larval emergence from eggs was significantly affected in larval aquatic environment pre-exposed to Csp_P pellets (Fig. 4R). Only 25 ± 2.1% of the eggs hatched in water pre-exposed to Csp_P, as compared to 82 ± 2.3% of the control eggs. These data indicate that exposure to Csp_P pellets in the aqueous environment significantly impairs egg eclosion in Ae. aegypti, likely through diffusible Csp_P-produced factors.

Larval exposure decreases vector competence

To investigate whether Csp_P pellet ingestion at the larval stage influences permissiveness of Anopheles. sp to P. falciparum and of Ae. aegypti to dengue virus DENV2, we conducted infection assays using mosquitoes that developed from larvae exposed to the lowest dose (50 mg/mL) of Csp_P pellets. Although only a small number of adult mosquitoes were obtained in each of the replicates from large-scale larval bioassays (n = 250–500 larvae/tray) in multiple experiments, we were able to collect 10–30 surviving female mosquitoes that would also feed on pathogen-infected blood. We found that the surviving An. gambiae and An. stephensi and the surviving Ae. aegypti displayed a significantly lower vector competence for P. falciparum and DENV2, respectively (Fig. 5A–C). The inhibition of P. falciparum infection in An. gambiae was significant (p = 0.01), with lower oocyst counts (median, 9.5 oocysts per midgut) being observed in the Csp_P-treated group, as compared to the control group (median, 44 oocysts per midgut) (Fig. 5A). Similarly, a median of 2.5 oocysts per midgut was found in An. stephensi mosquitoes, as compared to a median of 10.5 oocysts per midgut for the control mosquitoes (Fig. 5B). Moreover, the infection prevalence in both Anopheles species was significantly reduced (Fig. 5D, E). Plaque assays indicated a significantly lower level (p = 0.001) of plaque-forming units (pfu’s) per mosquito in the Csp_P pellet-exposed Ae. aegypti group (median, 3.77) than in the control group (median, 3.95), mirroring the reduced infection prevalence (Fig. 5C, F).

Fig. 5. Larval exposure to Csp_P pellets suppresses Plasmodium falciparum infection and DENV2 virus infection in Anopheles and Ae. aegypti mosquitoes.

A, B At 7 days after an infectious blood meal, Plasmodium falciparum infection intensity was significantly reduced (as determined by low oocyst counts per midgut) in both An. stephensi and An. gambiae after exposure to Csp_P pellets, as compared to the control group. Mann–Whitney test; **p = 0.005, ***p = 0.0006. Solid red lines indicate treatment medians. The parasite and viral loads were measured in individual mosquitoes, and statistical tests reflect three biological replicates. C Surviving Ae. aegypti mosquitoes that had a lower DENV2 infection load. Mann–Whitney test; **p = 0.007. D, E Bar graphs represent the significantly reduced P. falciparum infection prevalence in An. stephensi and An. gambiae, respectively, while F shows the DENV2 prevalence in Ae. aegypti following treatment with Csp_P pellets during the larval stage. Fisher’s exact test: ****p < 0.0001. For (A–C), each data point represents a count from a single mosquito. The data are pooled from three independent replicates. Raw data are presented in the Supplementary Data. Data represent three biological replicates, with each replicate derived from independently infected mosquito cohorts, and technical replicates were averaged within each biological replicate. G Illustration of sequential survival, egg laying, pathogen infection, and offspring development in control and 50 mg/ml exposure Csp_P groups. Percentages indicate the proportion of individuals progressing through each stage relative to the preceding stage. Dashed lines delineate key biological transitions, including egg-laying success, pathogen infection prevalence (infection data are based on figures 5E and 5F), and offspring outcomes conditional on egg laying. Compared with controls, the Csp_P-treated group shows pronounced reductions in survival, egg laying, offspring development, and infection prevalence across stages. Illustrations were created using BioRender (BioRender.com; https://BioRender.com/qp46r62).

Sequential mosquito life-history impacts of Csp_P exposure

A stage-wise life-history analysis was used to assess the cumulative impact of Csp_P pellet exposure on mosquito development, reproduction, and vector competence (Fig. 5G). Across all tested mosquito species, larval exposure to Csp_P resulted in pronounced losses at successive developmental and reproductive stages, leading to near-complete collapse of infection success at the population level. In An. gambiae, controls progressed with high success (pupation 94%, adult survival 92%, blood-feeding 87%, oviposition 73%) and showed 78% infection prevalence, whereas Csp_P exposure collapsed performance at each step—pupation 18%, surviving adults 12%, fed adults 10%, egg-laying 3%—and reduced infection prevalence to 3%. Progeny of exposed An. gambiae also underperformed (larvae 38%, pupae 28%, adults 17%) compared with control progeny (99%, 93%, 80%), indicating trans-generational detrimental effects. Similarly, Ae. aegypti controls showed robust progression (pupation 96%, adults 94%, naïve blood feeding 92%, oviposition 80%) with 61% infection prevalence, while Csp_P exposure reduced these to 19%, 18%, 12%, and 9%, respectively, and lowered infection prevalence to 6%. Offspring of exposed Ae. aegypti exhibited reduced survival to adulthood (36 vs 65% in controls), with intermediate survival at larval (77 vs 87%) and pupal (69 vs 70%) stages. Collectively, these findings demonstrate that Csp_P exposure imposed the strongest bottlenecks at metamorphosis and survival to adulthood, diminished feeding and fecundity, and markedly suppressed malaria parasite and dengue infection, with effects persisting into the subsequent generation.

Additionally, we statistically evaluated egg-laying and pathogen infection  between consecutive stages  by first determining how many mosquitoes successfully laid eggs or became infected. In An. gambiae, 77% of control females laid eggs, with an average of 38 ± 19.6 eggs per female, whereas only 36% of Csp_P-exposed females laid eggs, with a reduced mean of 11.3 ± 12.7 eggs per female. Similarly, in Ae. aegypti, 85% of control females laid eggs (67.0 ± 40.0 eggs per female), compared to 42% of Csp_P-exposed females (42.1 ± 29.4 eggs per female). These results indicate that Csp_P strongly reduces the probability of egg laying and lowers fecundity among females that do lay eggs (Supplementary Table 3). For infection outcome, in An. gambiae, 94% of control females became infected, with an average of 47.0 ± 29.4 oocysts per infected mosquito, whereas only 74% of Csp_P-treated females were infected, with a reduced mean of 13.4 ± 22.8 oocysts. In Ae. aegypti, infection prevalence dropped from 73% in controls (2.9 ± 0.4 oocysts) to 47% in Csp_P-exposed females (1.7 ± 0.3 oocysts) (Supplementary Table 4). Overall, these data indicate that Csp_P exerts its strongest effect by reducing the probability of egg laying and infection, with additional effects on fecundity and pathogen burden among the mosquitoes that do progress through these stages.

The mode of action of Csp_P pellets involves cardia/ventriculus damage, interference with digestion, and midgut damage

To investigate the cause of larval mortality mediated by Csp_P, we examined the phenotype of deceased larvae following exposure to Csp_P pellets. Larvae that ingested Csp_P pellets displayed pale, emaciated bodies with internal organ damage, specifically to the cardia and ventriculus (Fig. 6A, C). In contrast, larvae exposed to a known bio-larvicide, Bti pellets, exhibited blackened bodies but no internal organ damage visible by microscopy (Fig. 6B). Also, we observed the deposition of Csp_P pellet particles in the guts of the exposed larvae, indicating their inability to digest the pellets they had ingested (Fig. 6C). This observation prompted us to further investigate larval digestion over time by examining the midguts. For this purpose, larvae were exposed to 50 mg/mL of Csp_P pellets for 48 h. The pellets were then removed from the rearing water, and the surviving larvae were transferred to standard rearing conditions, receiving regular larval food. Despite the significant mortality that resulted, we successfully dissected the guts of a sufficient number of surviving larvae up to 6–8 days after exposure by conducting multiple separate experiments. We observed that Csp_P pellet particles persisted in the larval gut for up to 6 days (Fig. 6D), indicating an inability to digest the material. This condition may have resulted in cardiac and ventriculus damage, ultimately leading to larval mortality.

Fig. 6. Larvae that have ingested Csp_P pellets show internal tissue damage, with distorted midgut integrity and deposition of Csp_P pellets in the midgut.

A Image of healthy larvae indicating the various parts of the midgut (MG) tissue, including cardia (C), and ventriculus (V) or stomach. B Images of dead larvae that had been exposed to Csp_P pellets or Bti pellets (known larvicide). Unlike those exposed to Bti pellets, larvae exposed to Csp_P pellets were thin, with breakage in the ventriculus and deposits of ingested Csp_P pellet particles. Arrow in panels (a–c) indicates the breakage in the ventriculus, while panels (d–f) shows an ingested Csp_P pellet particle. Images were taken 24 h post-exposure (C, D). Larval midguts were dissected at different points, and images were recorded to examine the deposition of Csp_P pellet particles. Arrow in upper panels (a–f) showed the midgut of control larvae with less or no food pellets over different time points, while arrows lower panels (g–l) showed midgut of Csp_P pellet ingested larvae. Images showing deposition of Csp_P pellet particles in the larval ventriculus and midgut over time suggest impaired digestive capacity and gut transit, which aligns with disrupted midgut epithelial function. E Midgut were dissected post 24 h exposure for both control and Csp_P-treated larvae. Phalloidin and DAPI staining were performed on dissected larval guts, specifically focusing on the posterior midgut region, where pronounced epithelial damage and F-actin disorganization were observed following Csp_P pellet ingestion. Arrows indicated the disrupted actin bundles and junctional disorganization in guts of Csp_P treated compared to control group.

To confirm the internal tissue damage caused by the ingestion and deposition of Csp_P pellet particles, we examined the integrity of larval midguts after pellet ingestion. At 24 h post-exposure, the larvae were starved for 2 days, after which their midguts were dissected, fixed, and stained with fluorescent phalloidin reagent to label F-actin filaments. As shown in Fig. 6E, the midguts of larvae that ingested Csp_P pellets exhibited noticeable distortion, and the F-actin filaments did not display the same organization as observed in the control larvae (Fig. 6E).

Csp_P exposure induces apoptosis and autophagy

We next analyzed larval expression of genes associated with apoptosis, autophagy, and the innate immune response after Csp_P pellet ingestion. At 48 h after exposure to 100 mg/mL Csp_P pellets, surviving larvae were collected for quantitative real-time PCR -based gene expression analyses. The whole-body samples were used for gene expression due to the technical challenge of obtaining a sufficient number or consistently dissected midgut RNA/protein from larvae, and to capture systemic responses beyond localized gut damage. We found that the mRNA abundance of apoptosis-related genes (Inhibitor of apoptosis protein 1/ 5 or IAP1/IAP5; Caspase-8 or CASPS 8; ribosomal protein L8 or RPL8, and p53) was significantly (p < 0.0001) higher in the larvae treated with Csp_P pellets than the controls (Fig. 7A). Similarly, four autophagy-related genes (Autophagy-related genes or ATG1, ATG 5, ATG4A, and ATG7) and five immune genes (Defense repressor-1 or Dnr1, RNA-binding protein (Imp), apoptotic activator or Decbl 1, buffy, and Baculoviral inhibitor of apoptosis or bruce) also showed increased mRNA abundance (P < 0.002) in larvae that had ingested Csp_P pellets, as compared to controls (Fig. 7B).

Fig. 7. Analysis of apoptotic, autophagic, and immune responses in mosquito larvae exposed to Csp_P pellets.

A, B Relative mRNA expression levels of apoptotic and autophagy-associated genes in Ae. aegypti larvae 24 h after ingestion of Csp_P pellets (blue bars), compared to control larvae fed with E. coli pellets (gray bars). RNA was extracted from whole-body homogenates of 10 pooled live larvae per group. Statistical significance was assessed using two-way ANOVA with Dunnett’s multiple comparisons test. ****p < 0.0001. C, D Caspase-8 activity in whole-body homogenates of live Ae. aegypti and Cx. quinquefasciatus larvae at 0–72 h post-Csp_P pellet exposure. Three biological replicates were used per time point. This assay assessed progressive activation of apoptosis in surviving individuals. Statistical analysis used two-way ANOVA with Dunnett’s post hoc test. *p = 0.03, **p = 0.004, ***p = 0.0003, ****p < 0.0001. E, F Caspase-8 activity in moribund/deceased Ae. aegypti and Cx. quinquefasciatus larvae collected ~24 h post-Csp_P exposure. These data represent caspase activity at terminal stages of intoxication. G, H Phenoloxidase (PO) activity in homogenates of Ae. aegypti and Cx. quinquefasciatus larvae after Csp_P pellet ingestion. PO activity was measured as an indicator of systemic immune activation, regardless of external melanization phenotype. Statistical significance: ****p < 0.0001. Raw data is presented in the Supplementary Data. Each data point represents the mean ± SEM of three independent biological replicates.

Based on the increased expression of apoptotic gene markers, we assessed effector caspase-8 activity in larval lysates after Csp_P pellet ingestion. Caspase-8 activity, measured by luminescence-based detection of caspase-3 and caspase-7 cleavage of a target peptide, was consistently higher in Ae. aegypti (p < 0.0001) and Cx. quinquefasciatus (p < 0.001) larvae exposed to Csp_P pellets (Fig. 7C–F). Notably, control larvae that had ingested E. coli pellets did not exhibit significant caspase activity when compared to the Csp_P-treated larvae (Fig. 7C–F). Both mosquito species showed elevated caspase-8 activity at various time points post-exposure. In addition, caspase-8 activity was detected in the lysates of dead larvae from both mosquito species (Fig. 7E, F).

Next, we measured phenoloxidase (PO) activity in larvae exposed to Csp_P pellets. Remarkably, PO activity, measured by changes in absorbance at 490 nm upon L-DOPA addition, was significantly elevated in Ae. aegypti (p < 0.001) and Cx. quinquefasciatus (p < 0.001) larvae that had ingested Csp_P pellets when compared to controls, with increased activity observed throughout the experiment up to 18 h, coinciding with larval death (Fig. 7G, H).

To investigate whether Csp_P pellet ingestion induces apoptotic or necrotic cell death, we dissected midguts from Ae. aegypti larvae exposed to Csp_P pellets and performed staining with Apopxin Green (apoptotic cells), 7-AAD (late apoptotic and necrotic cells), and DAPI (nuclei). In Fig. 8A, brightfield and fluorescence images of dissected midguts (excluding the alimentary canal) show a marked increase in Apopxin Green staining in Csp_P-treated larvae compared to E. coli controls, with prominent signal observed in epithelial filaments extending from the anterior midgut. Higher magnification images confirmed that this staining localized to the midgut epithelium, suggesting extensive apoptosis following Csp_P exposure. Further analysis of intact midgut and alimentary canal (Fig. 8B) showed co-localization of apoptotic and necrotic markers, with Apopxin Green staining the midgut epithelium and 7-AAD labeling necrotic regions within the alimentary canal, confirming mixed modes of cell death. Control tissues showed minimal signal for either marker. To quantify these findings, we conducted flow cytometry on dissociated midgut cells stained with Annexin V-FITC and propidium iodide. As shown in Fig. 8C, exposure to 100 mg/mL Csp_P pellets resulted in a significant increase in early apoptotic cells (Annexin V⁺/PI⁻), rising from 9.2% in controls to 71% in treated larvae. Late apoptotic or necrotic cells (Annexin V⁺/PI⁺) also increased, while the proportion of live cells declined. These data confirm that Csp_P ingestion triggers widespread midgut epithelial apoptosis and necrosis in mosquito larvae. Raw images with gated cell populations are presented in Supplementary Fig. 3.

Fig. 8. Detection of apoptosis and necrosis in larval midgut tissue following Csp_P pellet exposure.

A Low magnification brightfield and fluorescence images of dissected Ae. aegypti larval midguts (minus the alimentary canal) 24 h after exposure to Csp_P pellets (100 mg/mL). Tissues were stained with DAPI (blue, nuclei), Apopxin Green (AG, green, apoptotic cells), and 7-AAD (red, necrotic cells). Control midguts showed no apoptotic signal, while Csp_P-treated midguts exhibited strong Apopxin Green labeling in epithelial filaments extending from the anterior region. A magnified region (boxed) is shown at right, confirming apoptotic labeling in midgut epithelium. Scale bar: 200 μm. B High-magnification images of intact midgut, including the alimentary canal from larvae exposed to Csp_P (100 mg/mL). Red arrows indicate anatomical regions. Co-staining revealed apoptotic cells (AG) in the midgut epithelium and necrotic cells (7-AAD) in the alimentary canal. Merged images demonstrate spatial overlap of apoptosis and necrosis. Scale bar: 200 μm. C Flow cytometry analysis of dissociated midgut cells from control and Csp_P-exposed larvae (50 and 100 mg/mL). Cells were stained with Annexin V-FITC and propidium iodide (PI) to quantify live (Annexin−/PI−), early apoptotic (Annexin+/PI−), late apoptotic (Annexin+/PI+), and necrotic (Annexin−/PI+) populations. The percentage shown represents gated quadrant distributions from three biological replicates. Data represent three biological replicates for each experiment.

Since chromatin condensation and DNA fragmentation are hallmarks of apoptotic cell death, we used TUNEL analysis to measure DNA fragmentation in the midguts of larvae exposed to Csp_P pellets. Notably, the guts of larvae that had ingested Csp_P showed increased DNA fragmentation when compared to those that had ingested E. coli control pellets (Fig. 9A, B). FACS analysis revealed that 62% of the gut cells in Csp_P pellet-treated larvae exhibited DNA fragmentation, as compared to only 2.5% for the control-treated larvae (Fig. 9C). Raw images with gated cell populations are presented in Supplementary Fig. 4. Altogether, these results indicate that ingestion of Csp_P pellets induces apoptosis-like cell death with DNA fragmentation.

Fig. 9. Detection of DNA fragmentation in mosquito larval midgut cells following Csp_P pellet exposure using TUNEL staining and flow cytometry.

A Confocal microscopy images of dissected larval midguts from control (E. coli pellet-fed) and Csp_P-treated Aedes aegypti larvae after 24 h exposure. Nuclei were stained with DAPI (blue), and DNA fragmentation was detected using the TUNEL assay with TdT-labeled Alexa Red (red). Merged images show co-localization of DNA damage in the midgut tissue. The control midgut shows a minimal TUNEL signal, while the Csp_P-exposed midgut displays widespread DNA fragmentation. B Enlarged views of the posterior midgut region from Csp_P-treated larvae, showing intensified TdT-labeled staining in epithelial cells, indicative of apoptosis-associated DNA fragmentation. White boxes highlight regions shown at higher magnification to the right, allowing visualization of individual nuclei and red, fluorescent TUNEL-positive signals. C Flow cytometry analysis of midgut cell suspensions stained with TdT-Alexa Red (x-axis, apoptotic signal) and propidium iodide (y-axis, necrotic marker). Quadrants indicate populations of live (lower left), early apoptotic (lower right), necrotic (upper left), and late apoptotic/necrotic (upper right) cells. Exposure to 50 or 100 mg/mL Csp_P pellets led to a marked increase in DNA-fragmented cells (TUNEL positive) compared to the control. Data represents one of three biologically independent experiments.

Discussion

Formulation development for improved Csp_P pellet stability

Building on prior formulations developed by Caragata et al.²⁹, we sought to improve the field applicability of Csp_P-based larvicides by addressing limitations related to binder choice and environmental stability. Caragata et al. demonstrated that while gelatin- and agarose-based pellets containing crude fish meal powder as an attractant, were effective, they posed practical challenges gelatin degraded rapidly in water, and agarose, although more stable, involved higher production costs and potential sustainability concerns²⁹. To overcome these constraints, we developed a guar gum-based, food-grade pellet formulation that offers advantages in cost, environmental safety, stability, and operational simplicity. Guar gum is a significantly more affordable (USD 1.28–3.20/kg) than gelatin (USD 5–10/kg) and agarose (USD 110–1200/kg), plant-derived and biodegradable binder with low aquatic toxicity. Unlike gelatin, which melts above 35 °C, our guar gum pellets remained structurally intact and retained activity even after heat-treatment at 70 °C for 2–5 days, and also retained larvicidal activity for over a month in water at 40 °C. Although the shelf-life of our current formulation was 6–8 months at room temperature, slightly shorter than the >1-year shelf-life reported for Csp_P powder alone—this could potentially be extended through improved stabilizer systems and storage under cooler, dry conditions (8–12 °C). The use of a filtered fish meal water suspension instead of crude fish meal powder also excluded non-essential biomass from the pellets. Together, these findings confirm that guar gum serves as a practical, scalable binder for Csp_P pellets, enabling robust larvicidal performance while improving affordability and environmental compatibility for large-scale vector control.

Efficacy and dose dependence of Csp_P pellet formulation

Although a larvicidal formulation using non-live Csp_P powder for field implementation may require further development, our prototype Csp_P-based larvicide pellet formulation already exhibits several desirable properties; it is fast-acting and efficient, killing mosquito larvae within an average of 2–3 days after exposure in laboratory assays. This result is consistent with our previous findings, in which larvae that survived ingestion of agarose-based Csp_P pellets that were not suitable for commercialization experienced delayed and halted development^30,31. The rate of pupation was significantly reduced and delayed upon exposure to Csp_P pellets in the case of all tested mosquito species. We further found that only two pellets (weighing approximately 0.1–0.4 g) containing 100–200 mg/mL of Csp_P were sufficient to kill 500–800 larvae in 5–8 L of larval rearing water under laboratory conditions. This low effective dose highlights the formulation’s potential as cost-effective larval control.

The Csp_P pellets demonstrated broad larvicidal activity against a wide range of mosquito species, including those of high epidemiological importance, such as An. gambiae, An. stephensi, Ae. aegypti, and Cx. quinquefasciatus. The pellets effectively caused mortality in Ae. aegypti and Cx. quinquefasciatus larvae, with a similar average time-to-death after treatment as observed in other mosquito species. These mosquito species are prominent vectors of dengue virus, West Nile virus, and Japanese encephalitis virus, and of nematodes causing filariasis, all of which pose substantial health threats to humans³². Furthermore, pellets were effective against the chemical insecticide permethrin-resistant Ae. aegypti NR-48830 line²⁹. The widespread use of pyrethroid-based insecticides in mosquito control programs has led to high levels of resistance, causing current mosquito control interventions to fail. Our results indicate that the Csp_P-based pellets could represent an appropriate candidate as an alternative natural insecticide for targeting these resistant populations. We have previously shown that exposure of ten consecutive Ae. aegypti generations to Csp_P did not result in development of resistance³¹, suggesting that our current pellet formulations are likely to maintain their effective larvicidal activity upon continuous exposure to mosquito populations. Our results also demonstrated that the Csp_P pellets significantly reduced larval emergence from eggs when Ae. aegypti eggs were exposed to it, likely as the result of the diffusion of Csp_P -produced factors into the aquatic environment water. Previous studies have shown that bacteria-produced factors can inhibit egg viability and hatching^33,34. By impairing egg viability, these pellets disrupted the life cycle of the mosquito at its earliest stage, thereby reducing larval development and emerging adult mosquito populations. This targeted approach aligns with the principles of integrated vector management (IVM), emphasizing sustainable and precise interventions to control vector populations effectively. However, these experiments were conducted under controlled laboratory conditions, which may not fully capture environmental variability such as water quality, microbial competition, or temperature fluctuations that occur in natural larval aquatic environments. This limitation necessitates testing under realistic field conditions.

Post-exposure impact on mosquito longevity, fecundity, and vectorial capacity

The impacts of sub-lethal Bacillus thuringiensis var. israelensis (Bti) exposure on mosquito development and fitness have also been addressed, but to a lesser extent. For instance, Hare and Nasci observed that exposure of Aedes aegypti larvae to sub-lethal concentrations of Bti resulted in increased larval development time and reduced adult size, indicating developmental delays and potential fitness costs³⁵. Similarly, studies demonstrated that Anopheles coluzzii mosquitoes exposed to sub-lethal Bti doses during larval stages exhibited reduced adult survival rates³⁶.

Extending these reports, we show that adults derived from Csp_P-exposed larvae consistently showed decreased longevity, which varied between mosquito species. Considering the importance of longevity in vectorial capacity, any exposed mosquito larvae that survived to adulthood would be expected to have a reduced likelihood of transmitting human pathogens. Other fitness effects noticed were decreased fecundity and fertility across all tested mosquito species. Furthermore, larval ingestion of Csp_P pellets significantly affected the vectorial capacity of adult Anopheles and Ae. aegypti mosquitoes. The infection intensity and prevalence of P. falciparum and DENV2 were significantly reduced in Anopheles and Ae. aegypti mosquitoes, respectively, that had emerged from Csp_P-exposed larvae; this is another crucial feature that decreases pathogen transmission by surviving mosquitoes.

Mode of action

The majority of larvae that ingested a sub-lethal dose of Csp_P pellets were unable to survive even when subsequently provided with regular food pellets, likely indicating irreversible cell and tissue damage; a feature that enhances the cost-effectiveness of the Csp_P pellets, as a single exposure is adequate to mediate larval killing.

Although the precise mechanism of Csp_P larvicidal activity has not been studied previously, but other bio-larvicides have shown induction of stress-mediated apoptotic cell death. For example, Bacillus thuringiensis endotoxin induces programmed cell death in polytene midguts cells of Culex pipiens larvae^37,38. A study by Pereira et al. demonstrated that Csp_P biopesticide formulations are effective against economically important corn rootworm pests, including multiple Bti-resistant strains, suggesting that Csp_P acts through a mechanism distinct from Bti toxins³⁹. Larval mortality caused by Metarhizium brunneum is largely mediated through protease-induced stress followed by apoptotic cell death. Whether the predatory larvae-eating Toxorhynchites brevipalpis larvae are susceptible to the Csp_P biopesticide has not been tested⁴⁰. A recent study has demonstrated that a yeast-encapsulated orange oil larvicide kills Ae. aegypti larvae through acute midgut cell damage and apoptotic cell death⁴¹. Other studies have also shown that bio-larvicides induce stress-mediated cell death by the production of toxic factors^42–45.

Consistent with these studies, the phenotype of larvae exposed to Csp_P pellets was manifested as pale yellowish, emaciated bodies with damage to internal organs such as the cardia and ventriculus (Fig. 6A, C). This phenotype differs from that observed following Bti exposure, which induces midgut damage primarily through melanization, often without preservation of the underlying gut architecture⁴⁶, limiting a direct comparison to Csp_P-mediated death (Fig. 6B). Moreover, the presence of intact, undigested Csp_P pellets in the gut suggests impaired digestive capacity and gut transit, which aligns with disrupted midgut epithelial function. Thus, it appears that following ingestion, Csp_P pellets prevent the activation of digestive proteases required to digest the pellets. We also observed substantial damage to the midgut epithelia, including disordered actin filaments, breakage in cellular junctions, and loss of cytoplasmic contents, indicating that Csp_P pellet killing mechanisms are likely based on its toxicity to midgut cells (Fig. 6E).

We observed increased expression of CASP8 following Csp_P exposure, indicating a response likely aimed at repairing the damaged midgut epithelia. In addition, the upregulation of IAP1 and IAP5 suggested their involvement, as effector caspases, in the larvae’s protective mechanisms against excessive cell death. These findings highlight the activation of apoptotic pathways and the potential regulatory role of IAPs in maintaining cellular homeostasis during Csp-P-induced stress⁴⁷. IAPs have been shown to interact and regulate the expression of the initiator caspase Dredd and the effector caspases CASPS 7 and CASPS 8^48,49, which cause proteolysis and fragmentation of proteins during cell death. The expression of other apoptotic genes (Mx, buffy, bruce, Decb1, Dnr1, Aesps3, etc.) was also significantly increased in Csp_P-exposed larvae (Fig. 7A, B). Consistent with these gene expression changes, caspase enzymatic activity increased progressively over time in surviving larvae and peaked in moribund/dead larvae 24 h post-exposure, highlighting a time-dependent intensification of apoptotic stress that could lead to exacerbated apoptosis as described by Butt et al.⁴², eventually leading to death. PO’s are known to play roles in stress-response as they are responsible for wound healing and immunity^50,51. Notably, phenoloxidase (PO) activity rose significantly in Csp_P-exposed larvae, supporting an immune stress response. While PO is canonically associated with melanization, our data suggest that its systemic activation in this context does not necessarily result in visible pigmentation, possibly due to gut-localized responses or limitations in substrate availability. While PO is known to contribute to stress-response, wound repair, and microbial defense, studies have demonstrated that PPO can be activated to produce cytotoxic intermediates (quinones, reactive oxygen species) that kill pathogens without forming macroscopically visible melanin deposits^52,53. It is possible that the observed gut damage upon Csp_P ingestion may expose larvae to their endogenous microbiota, triggering immune activation leading to PO production. The midguts of larvae that had ingested Csp_P pellets larvae were found to be positive for Apopxin Green/Annexin V staining, as illustrated by microscopy and FACS analysis (Fig. 8). Altogether, these results indicate that the ingestion of Csp_P pellets is likely associated with cellular cytotoxicity affecting midgut integrity and inducing the apoptosis expression cascade. Previous studies have demonstrated that the expressions of ATG1, ATG6, and ATG8 are linked with V-ATPase expression in Ae. aegypti larvae exposed to the plant-derived fatty acid molecule squamocin^41,54. Upon ingestion of Csp_P pellets, the expressions of ATG1, ATG5, ATG4A, and ATG7 were also significantly increased when compared to that in larvae that had ingested control E. coli pellets. Also, autophagic cell death in the midgut has been linked to apoptosis markers, such as DNA fragmentation, in D. melanogaster⁵⁵.

In our study, typical patterns of necrotic cell death (Fig. 8) and DNA degradation (Fig. 9) were observed through 7-AAD (a marker of necrosis) staining and TUNEL analysis of the larval alimentary canal and midgut, respectively. Although the correlation between autophagic and apoptotic pathways is not fully understood, positive necrotic and apoptotic staining in the midguts was observed following Csp_P ingestion by mosquito larvae, as shown in our confocal microscopic images. These results indicate that the autophagic mechanism induced by Csp_P pellets may act in concert with apoptosis. The mechanistic insights into larval midgut damage and apoptotic pathways were primarily based on imaging and microscopy, at distinct time points and larval states (live, moribund, deceased), limiting reconstruction of a clear pathological timeline. Gene expression and enzymatic assays (Fig. 7) were time-resolved, whereas imaging studies (Figs. 6, 8 and 9) captured fixed endpoints, typically 24 h post-exposure, to highlight peak effects. Future studies will incorporate synchronized time-course analyses across all methods to better define the sequence of pathological events.

In summary, we demonstrate that our guar-gum based Csp_P pellet formulation exerts strong larvicidal activity by disrupting larval development and impairing key adult traits such as fecundity, fertility, pupation, emergence of larvae from eggs, and vector competence. This ingestible larvicide acts through mechanisms involving acute midgut damage and apoptosis, ultimately leading to larval death (Fig. 10). The pellets are low-cost with a shelf-stable for up to 8 months at room temperature and maintain their structural integrity in larval aquatic environments for over a month, even at elevated temperatures. Larvae-attracting pellets, in contrast to dispersed liquids or powders, minimize ecological impact by reducing exposure to non-target aquatic organisms. The Csp_P formulation is currently undergoing further evaluation in field trials and non-target safety testing. While ingestion was confirmed using fluorescein dye, the quantification of intake within larvae and its correlation with mortality remains to be explored in greater detail. Furthermore, we note that insect-killing properties are not unique to Csp_P, and several Chromobacterium spp. have been reported to exhibit insecticidal activity⁵⁶. However, most studies reporting such activities of other Chromobacterium species and strains have utilized live bacteria or fresh culture supernatants, and it is therefore not clear whether insect mortality was due to infection (entomopathogenic activity) or stable insecticidal metabolites^57–61. While phylogenetic analyses place Csp_P within a broader lineage of insect-associated Chromobacteria⁶², a comparative phylogenetic assessment of insecticidal traits across this group is beyond the scope of the present study and is the subject of ongoing work. Our previous work demonstrated that live Csp_P produces cyanide that dose-dependently kills mosquito larvae⁶³. However, the formulations used in the present study involve a non-live bacterial preparation that has been shown to be devoid of cyanide²⁹, and the insecticidal activity is therefore mediated by other bioactive compounds that remain stable after bacterial inactivation. Additionally, the long-term ecological impact and activity of Csp_P toward non-target aquatic organisms were not assessed in this study and will require dedicated field-based evaluations. Addressing these gaps will be essential for advancing Csp_P as a sustainable and scalable vector control intervention.

Fig. 10. Summarized graphical depiction of the larvicidal activity of Csp_P pellets and their mode of action.

Illustrations were created using BioRender (BioRender.com; https://BioRender.com/qp46r62).

Materials and methods

Ethics statements

All animal procedures were performed in strict accordance with the recommendations in the Guide for Care and Use of Laboratory Animals of the National Institutes of Health (NIH), USA, and those of the Animal Care and Use Committee of Johns Hopkins University (permit no. M006H300). The Animal Care and Use Committee of Johns Hopkins University approved all animal experiments performed in this study. The BalBc mice were maintained in the sterile animal facility at the Johns Hopkins Bloomberg School of Public Health, following all vendor-provided guidelines for animal housing and care. We have thereby complied with all relevant ethical regulations for animal use. Commercially available anonymous human blood was used for parasite culture and mosquito feeding assays. The blood was obtained from Johns Hopkins Hospital (East Baltimore Campus) and contained no identifiable donor information and has therefore been exempted by the Johns Hopkins University IRB committee from human research. Hence, all ethical regulations relevant to human research participants were followed.

Mosquito rearing

The Aedes aegypti Liverpool strain, the permethrin-resistant Ae. aegypti strain (BEI Resources, NR-48830), the Cx. quinquefasciatus JHB strain (originally isolated in Johannesburg, South Africa; BEI Resources, NR -4305), the Anopheles stephensi LISTON strain, and the Anopheles gambiae Keele strain were reared under standard laboratory conditions (temperature, 27 °C ± 1 °C, relative humidity 80% ± 10%), with a 14:10-h day-night cycle. Mosquito eggs were hatched in distilled water supplemented with Tetramin fish/cat pellets (Tetrawerke, Melle, Germany) as a food source²⁹. For maintaining the mosquito lines, BalBc mice were utilized for the mosquito feeding, in which mice were anesthetized using ketamine (100 µL per 20–25 g weight) and mosquitoes were allowed to feed on immobilized animals for 15–30 min.

Larvicide pellet formulations

For pellet preparation, all the pellet ingredients (non-live Csp_P /E. coli powder, stabilizing agent, gums/waxes, and attractants) were thoroughly mixed using a KitchenAid mixer (Amazon, ASIN No. B07GK77Z7C). Once homogenous, the mixture was made manually into small 5 mm in diameter-sized pellets. We screened a broad range of natural gum and wax as a stabilizing agent for formulating these pellets. A list of the various gums and wax used in the formulations is given in Table 1. Similarly, various larval attractants used in the formulations are also summarized in Supplementary Fig. 1. After preparation, the pellets were allowed to dry in an oven at 40–60 °C for 6–7 days.

Table 1.

Summary of Csp_P pellet effects by mosquito species

Mosquito species	Larvicidal activity (% mortality, 400 mg/mL dose)	Pupation (delayed)	Adult longevity (days)	Fecundity/Fertility	Vector competence (infection prevalence)	Egg hatchability
An. stephensi	92	4 days	9	45/30%	40%	Not tested
An. gambiae	100	5 days	7	40/33%	30%	Not tested
Ae. aegypti	95	3 days	13	38/25%	45%	20%
Cx. quinquefascias	95	4 days	15	47/34%	Not tested	Not tested

Assessment of pellet stability and residual activity

To assess their residual activity and stability, Csp_P and control E. coli pellets containing a percentage of various gum/ waxes were added to 300 mL deionized (DI) water and incubated at two temperatures, 28 °C and 40 °C, for 30 days, undisturbed. During this period, the pellets were monitored every other day to determine whether they were still intact. After a month, the pellets were fed to Ae. aegypti larvae, and survival was monitored daily, as described above. The stability and integrity of the various pellet formulations in water over various periods of time are summarized in Supplementary Table 1.

Assaying the larvicidal activity of non-live Csp_P pellets

All experiments utilized 50–400 mg/mL of non-live Csp_P powder, prepared according to the protocol described by Caragata et al.²⁹. Pellets containing the biopesticide were dried and then fed to Ae. aegypti Liverpool larvae, Cx. quinquefasciatus larvae, permethrin-resistant Ae. aegypti larvae, An. gambiae, and An. stephensi larvae to assess whether the formulation was effective across multiple mosquito vector species.

For small-scale laboratory-conducted larval bioassays, groups of 30 L2-L3 larvae were transferred to mosquito breeders (height: 19.5 cm, base diameter: 11 cm; Bioquip, catalog no. 1425) containing 300 mL of DI water. Three replicate breeders were established for each treatment. Pellets containing non-live Csp_P powder and control pellets containing non-live E. coli powder were added to the respective breeders. We also conducted large-scale larval bioassays with high larval densities: 250–500 larvae were placed in trays with 2.5–5.0 L of DI water and treated with varying concentrations of Csp_P pellets. Larval survival was monitored daily for 3–5 days, and the rates of pupation and adult emergence were recorded for each group.

We further evaluated whether the larvicidal activity of the Csp_P pellets was reversible or irreversible in various mosquito species. In these experiments, 250–500 larvae were exposed to 50 mg/mL Csp_P pellets for 2–3 days. After the treatment period, the surviving larvae were transferred to new trays containing regular Tetramin fish food and maintained under standard insectary conditions. Larval survival was monitored 48 h after transfer, and the results were recorded. All larval assays were performed in three to four independent experiments, each containing two to three technical replicates.

Impact of Csp_P pellet exposure on mosquito fitness

Fitness of adult mosquitoes after exposure to Csp_P pellets during the preceding larval stage was examined by addressing four life history traits: longevity, fecundity, fertility, and blood feeding propensity. Approximately 250–500 larvae were treated with the lowest (50 mg/mL) dose of Csp_P, and the development of the surviving larvae was assessed. For fecundity and fertility assays, mosquitoes in 25 × 25 × 25-cm cages, each containing adult male and female mosquitoes from one of the treatment groups, were allowed to mate freely. At 5 days post-eclosion, the mosquitoes were allowed to feed on anesthetized mice. Fully engorged females were counted and separated from unfed females. Three days after blood feeding, the individual females were transferred to single tube containing a damp cone of filter paper for oviposition, except in the case of Cx. quinquefasciatus females, for which small, half-filled water cups were used for egg laying. The egg papers were removed after 48–72 h and photographed, and then the eggs were manually counted using ImageJ software and its pointer tool. Tubes in which no females laid eggs during those 3 days were discarded and recorded as having zero egg production.

Fertility assays were then performed using the eggs from the fecundity experiments described above. Oviposition papers were dried individually in small plates, and once they had dried completely, they were submerged in hatching water. Three to four days later, the total number of live larvae per paper was counted manually, and the hatching rate was calculated as the percentage of total eggs laid. The longevity of the adults that emerged was compared between the control and Csp_P-treated groups of mosquitoes. Within 24 h of eclosion, pools of 20 mosquitoes were transferred to paper cups. The cups were provided with cotton balls soaked in 10% sucrose, which were changed every other day. Dead mosquitoes were counted every second day and removed from each cup. Survival was monitored until all mosquitoes died. The cups were changed weekly to avoid exposing the mosquitoes to microorganisms that might grow in sucrose droplets or mosquito excreta.

Impact of Csp_P pellet treatment on the vectorial capacity of female mosquitoes

P. falciparum infection was established in Anopheles mosquitoes by membrane feeding as described earlier⁶⁴. In brief, P. falciparum NF54 strain was cultured as described earlier⁶⁵. The mature gametocyte culture was centrifuged and diluted with fresh human RBCs and serum. Mosquitoes (n = 20–40) that had been starved overnight were allowed to feed on membrane feeders with parasite-containing blood for 15–30 min. The mosquitoes that had fed were separated and kept at 26 °C for 7–8 days. At 7–8 days after feeding, the mosquito midguts were dissected and stained with 0.1% mercurochrome^27,28, and the number of oocysts per midgut in the control and Csp_P-treated groups was counted under a light microscope. More than three independent experiments needed to be performed for each group because of the insufficient number of female mosquitoes in the Csp_P-treated group. Similarly, a dengue virus serotype 2 (DENV2) infection was established in Ae. aegypti mosquitoes. A frozen stock of DENV2 was mixed with commercial human red blood cells and serum and used for infection²⁷. Starved female Ae. aegypti mosquitoes were fed a DENV2-infected blood meal through membrane feeders. Seven days post-infection, mosquitoes were homogenized in 300 µL of DMEM using a Bullet Blender (Next Advance Inc., Averill Park, NY, USA) with glass beads. The homogenate was then centrifuged at 10,000 × g for 3 min, and 20 µl of the filtered supernatant was collected for plaque assays^27,28. In brief, the BHK-21 cells were seeded in 24-wells at a 1:10 dilution and allowed to grow at 37 °C and 5% CO₂ for 24 h or until the cells were 70–80% confluent. The next day, infected mosquito homogenates were serially diluted in fresh DMEM (four 10-fold dilutions), and then 100 µL of each dilution was added to a monolayer of BHK cells (~80% confluence) in 24-well plates. The plates were rocked at room temperature for 15–20 min and then incubated at 37 °C for 45 min. Then, 1 mL of pre-warmed methylcellulose overlay (DMEM + 2% Gibco heat-inactivated FBS with 1% l-glutamine, 1%penicillin-streptomycin and 5 μg/mL plasminogen with 0.8% methylcellulose; 0.2 micron-filter sterile) was added to each well. At 5–7 days after infection, the 24-well plates were fixed and stained with crystal violet. Plaques (indicating cytopathic effect) were counted and analyzed.

Emergence of Ae. aegypti larvae from eggs in the presence of Csp_P pellets

A strip of Ae. aegypti egg (~80–150) was added to a 1.5-L DI water tray containing 200 mg/mL Csp_P pellets. A tray with 200 mg/mL E. coli pellets was also used as a control. Both trays were then kept under standard insectary conditions for 3 days. After that time, the number of larvae was counted, and the hatch rate was calculated as described above. In a second experiment, a 200 mg/mL dose of Csp_P pellets was added to the hatching water and kept under standard insectary conditions for 48 h. The Csp_P pellets were then removed, a strip of ~100–150 eggs were added, and the eggs were allowed to hatch for 48–72 h. Afterwards, the number of live larvae was recorded, and emergence of larvae from eggs was calculated.

Phalloidin staining

Ae. aegypti L3 larvae were incubated with 100 mg/mL Csp_P pellets for 24 h and then starved for 2 days. Midguts were dissected and fixed for 45 min in 4% paraformaldehyde (Invitrogen, USA) solution in 1X PBS and then washed three times for 15 min each with 1X PBS + 0.1% Triton to allow permeabilization of the guts. This step was followed by blocking with 1% BSA in 1X PBS with 0.1% Triton for 45 min at room temperature. Phalloidin coupled with AlexaFluor^TM 488 (Invitrogen USA; cat no. A12379) was added for 1 h (1:500 in 1X PBS with 0.1% Triton and 1% BSA), followed by three washes. After removal of the phalloidin, the slides were mounted in antifade medium containing DAPI (Thermo Fisher Scientific; cat no: P36935), and coverslips were sealed with colorless nail polish^66,67.

RNA isolation, cDNA synthesis, and qRT-PCR

TRIzol (Invitrogen, USA) followed by Turbo DNase I treatment⁶⁸ was used to extract total RNA from a pool of ten Ae. aegypti larvae that had been exposed to 100 mg/mL of Csp_P pellets for 48 h. The whole-body samples were used for gene expression due to the technical challenge of obtaining a sufficient number of consistently dissected midgut RNA/protein from larvae, and to capture systemic responses beyond localized gut damage. Complementary DNA (cDNA) was synthesized from 1 µg of total RNA using oligo dT primers and Moloney murine leukemia virus (MMLV) reverse transcriptase (Promega, USA). Real-time PCR was done using SYBR Green PCR Master Mix (Applied Biosystem) in a final volume of 15 µl in a StepOne real-time PCR system (Applied Biosystem). For all assays, the expression of selected genes was normalized to the expression of RPS7. The relative expression of selected genes was calculated as the 2^−ΔΔCT value. The primer sequences are summarized in the Supplementary Table 2, and all primers were procured from Eurofins.

Caspase and phenoloxidase activity

Caspase-3 and phenoloxidase (PO) activities were measured at multiple time points in whole-body homogenates of Ae. aegypti and Cx. quinquefasciatus larvae (n = 10 per treatment group), with each biological sample analyzed in triplicate wells. Whole-body homogenates were used due to the technical difficulty of consistently obtaining sufficient midgut tissue and to capture systemic responses beyond localized gut damage. For both assays, larvae were collected from treatment groups, flash-frozen in liquid nitrogen, homogenized using a manual homogenizer, and centrifuged at 14,000 × g for 10 min at 4 °C to obtain clarified supernatants, with buffer composition adjusted according to assay requirements (RIPA buffer (Invitrogen) for caspase-3; 1× PBS, pH 7.6, for PO). For caspase-3 activity, 35 µL of supernatant was mixed with an equal volume of Caspase-Glo® 3 reagent (Promega; cat no. G8090) in a white 96-well plate according to the manufacturer’s instructions; luminescence was measured after 1 h incubation at room temperature^40,42,69.

For PO activity, 20 µL of supernatant was added to 180 µL of L-DOPA solution (4 mg/mL; Sigma) in triplicate wells, followed by incubation at room temperature for 30 min and absorbance measurement at 490 nm. Protein concentrations in the larval homogenates were determined using Bradford’s method⁴⁰.

Evaluation of apoptosis: Apopxin Green, 7-AAD, CytoCalcein violet 450, and Annexin V-FITC/PI labeling

Two days after exposure to Csp_P pellets, larvae were washed with 1x PBS, and midguts were dissected and pooled (n = 10) from each group for analysis. Apoptosis and necrosis were assessed using the Apoptosis/Necrosis Assay Kit (Abcam; cat no. ab176749): Midguts were stained with Apopxin Green (for apoptosis), 7-AAD (for late apoptosis and necrosis), and Cytocalcein Violet 450 (for viable cells). Prior to staining, the midgut tissues were fixed in 4% paraformaldehyde, followed by three washes (10 min each) in 1x PBS and two additional washes (5 min each) in 1x PBS with 0.1% Triton X-100 (PBST). Samples were mounted on slides using ProLong™ Gold Antifade Reagent with DAPI (Thermo Fisher Scientific; cat no: P36935) and visualized under a ZEISS LSM 800 confocal microscope (Carl Zeiss AG)^41,70,71. Quantitative analysis of the apoptotic/necrotic cell populations was performed using flow cytometry (BD Pharmingen, San Diego). For this purpose, the midgut tissue was lysed with trypsin-EDTA (0.05% trypsin in EDTA, Thermo Scientific) and manually ground with a Teflon pestle to isolate single cells. The cells were then harvested by centrifugation and resuspended in 1x PBS. The isolated cells were stained with Annexin V-FITC and PI using an Annexin V-FITC Apoptosis Detection Kit (BD Pharmingen, San Diego; Cat no:556570) according to the manufacturer’s instructions. The binding of Annexin V was quantified via flow cytometry using the FL1 channel (Ex/Em = 490/525 nm), while PI staining was detected using the FL3 channel (Ex/Em = 550/650 nm). Data were analyzed by first gating the total cell population while excluding debris and non-specific signals based on forward and side scatter parameters. From this total population, single cells (singlets) were selected using FSC-A vs FSC-H plots to eliminate doublets. Within the singlet gate, cells were further analyzed for apoptotic status using Annexin V and propidium iodide (PI) staining. Populations were classified as live (Annexin V⁻/PI⁻), early apoptotic (Annexin V⁺/PI⁻), and late apoptotic/necrotic (Annexin V⁺/PI⁺) to quantify cell death phenotypes.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) analysis

Ae. aegypti larvae were exposed to 100 mg/mL Csp_P or E. coli (control) pellets for 24 h. Midgut tissues were dissected and fixed in 4% paraformaldehyde. Following fixation, the tissues were washed three times (10 min each) with 1x PBS, then twice (5 min each) in 1x PBS containing 0.1% Triton X-100 (PBST) to permeabilize the cells. DNA fragmentation was detected using TdT enzyme staining according to the manufacturer’s instructions (TUNEL Assay Kit, 640 nm; Cell Signaling Technology; cat no. 64936S). The tissues were mounted on slides with ProLong™ Gold Antifade Reagent with DAPI (Thermo Fisher Scientific) to visualize the nuclei, and the samples were imaged using a ZEISS LSM 800 confocal microscope (Carl Zeiss AG)⁴¹. Furthermore, DNA fragmentation was quantified by isolating single cells from the midgut tissue, following the same procedure as described earlier using flow cytometry. To assess DNA fragmentation, 7-AAD was used as a counterstain along with BrdU-Red as the primary indicator. Quantitative analysis was performed by flow cytometry, with BrdU-Red detected at an excitation/emission of 576 nm and 7-AAD detected at an excitation/emission of 488/655 nm. This approach allowed for precise measurement of DNA fragmentation and apoptotic cell populations in the midgut tissue.

Statistics and reproducibility

Statistical analyses were performed using GraphPad Prism. All experiments included three independent biological replicates using separate larval batches conducted on different days. Survival was monitored daily for 5 days. Data are presented as mean ± SEM unless otherwise indicated, and individual replicate values are shown in the figures. Sample sizes ranged from 30 to 500 larvae per group, as specified in the respective figure legends.

Survival and mortality data, representing binary outcomes, were analyzed using nonparametric tests. Differences between two groups were assessed using unpaired two-tailed t-tests (p = 0.005). Pupation rate data were analyzed using one-way ANOVA (n = 20–50 per mosquito species). Fitness parameters of control and Csp_P-treated mosquitoes were evaluated using a nonparametric Mann–Whitney U test (p = 0.1), with sample sizes of 30–40 mosquitoes per group. Bars represent mean ± SEM across three biological replicates, and individual replicate values are displayed as dots.

Fecundity, fertility, and larval emergence rates were compared using Mann–Whitney tests. Gene expression, caspase-8 activity, and phenoloxidase activity were analyzed using two-way ANOVA with Dunnett’s multiple comparisons test. Infection prevalence and intensity in vector competence assays were evaluated using Fisher’s exact test and Mann–Whitney tests, respectively. Statistical significance was defined as p < 0.05.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

Conceptualization: V.V., G.D. Methodology: V.V., G.D. Investigation and data analysis: V.V., M.P.K. Visualization: V.V. Writing—original draft: V.V. Review and editing: V.V., G.D.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: Johannes Stortz.

Data availability

The raw data generated in this study are provided in the Supplementary Data. Supplementary data are provided with this paper. All other data are available from the corresponding author on request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-10091-5.