Development of a controlled-release mosquito RNAi yeast larvicide suitable for the sustained control of large water storage containers

Abstract

Large household water storage containers are among the most productive habitats for Aedes aegypti (Linnaeus, 1762), the primary mosquito vector for dengue and other arboviral pathogens. Increasing concerns for insecticide resistance and larvicide safety are limiting the successful treatment of large household water storage containers, which are among the most productive habitats for Aedes juveniles. The recent development of species-specific RNAi-based yeast larvicides could help overcome these problems, particularly if shelf stable ready-to-use formulations with significant residual activity in water can be developed. Here we examine the hypothesis that development of a shelf-stable controlled-release RNAi yeast formulation can facilitate lasting control of A. aegypti juveniles in large water storage containers. In this study, a dried inactivated yeast was incorporated into a biodegradable matrix containing a mixture of polylactic acid, a preservative, and UV protectants. The formulation was prepared using food-grade level components to prevent toxicity to humans or other organisms. Both floating and sinking versions of the tablets were prepared for treatment of various sized water containers, including household water storage tank-sized containers. The tablets passed accelerated storage tests of shelf life stability and demonstrated up to six months residual activity in water. The yeast performed well in both small and large containers, including water barrels containing 20-1000 larvae each, and in outdoor barrel trials. Future studies will include the evaluation of the yeast larvicide in larger operational field trials that will further assess the potential for incorporating this new technology into integrated mosquito control programs worldwide.

Fig. 1

Larvicidal activity is retained in spray-dried and lyophilized RNAi yeast formulations. Yeast preparations were produced by spray drying a yeast-water mixture (A–C). A wet formulation (WF) consisting of heat-inactivated yeast that was pelleted but not dried served as a control (A, C), as did a control yeast strain (Control) that produces shRNA with no known target in mosquitoes (all panels). WF and spray-dried (SD) formulations of the syt.427 and sema.460 larvicides induced comparable levels of significant larval death (A). The mean results from three replicate trials per treatment conducted on A. aegypti larvae are shown, and errors denote SEM (A-C, *** P < 0.001 vs. controls). Significant syt.427 and sema.460 larvicide activity was also observed (B) after storage of the yeast at room temperature for two (2W), three (3W), and four weeks (4W). Results were combined from three replicate containers per treatment (B; *** = P < 0.001 vs. control). The RNAi larvicides were prepared through spray drying of a yeast-water mixture containing ascorbic acid (AA), benzoic acid (BA), citric acid (CA), potassium sorbate (PS), or no additive (WP). Significant syt.427 activity was observed in comparison to any of these spray dried preparations, as well as to wet formulation control yeast (C) from the same preparations (*** = P < 0.001 vs. any of the WF or spray dried controls). Formulated spray dried larvicide activity was comparable to that of WF syt.427. Larvicides were also prepared through lyophilization of yeast that had been rinsed with water to remove residual culture media (D, E). Significant syt.427 and sema.460 larvicide activity was observed in comparison to control yeast following one (1W), two (2W), three (3W), and four weeks (4W) of storage in the dark at room temperature (D; mean results from three replicate trials per treatment conducted on A. aegypti larvae are shown; *** = P < 0.001 vs. treatments with the control yeast). With respect to control interfering RNA yeast, lyophilized powder syt.427 yeast prepared with benzoic acid preservative induced significant A. aegypti larval mortality following 20 °C frozen or 54 °C accelerated storage (E, *** = P < 0.001; mean results from eight replicate containers per treatment are shown). Data were analyzed with ANOVA/Tukey’s multiple comparison test. Error bars throughout the figure represent SEMs.

Fig. 2

Formulating a controlled-release tablet. A. aegypti larval mortality was observed following larval treatments with tablets containing 20% syt.427 yeast, 20% chitosan, and 60% microcrystalline cellulose (A, *** = P < 0.001 vs. tablets prepared with control yeast, paired two-tailed t-test). The syt.427 tablets containing PLA or PCL induced significant larval mortality (B, *** = P < 0.001 with respect to the control counterpart, ANOVA/Tukey’s multiple comparison test; refer to Table 1 for the compositions of each tablet). The mean results from two replicate trials per treatment are shown in (B), and error bars represent SEMs. The syt.427 tablets containing the UV protectants zinc oxide (ZO) or titanium dioxide (TiO₂) induced significant larval mortality (C, *** = P < 0.001 with respect to the control counterpart, ANOVA/Tukey’s multiple comparison test). No significant larvicidal activity was observed in control tablets containing the UV protectants. Tablets contained UV protectants in either 20:20:5:45:10 yeast:chitosan:microcrystalline cellulose:PLA:UV protectant (ZO1 and TiO2-1) or 20:20:5:45:20 yeast:chitosan:microcrystalline cellulose:PLA:UV protectant (ZO2 and TiO2-2). The mean results from three replicate trials per treatment are shown in (C), and error bars represent SEMs. Tablets stored at 54 °C for two weeks in accelerated storage (2W-AS, D) did not have significantly different activity compared to fresh tablets (OW, D). Both the OW and 2W-AS tablets killed significantly more larvae than their control counterparts (D, P < 0.001 is denoted by ***, ANOVA/Tukey’s multiple comparison test; mean results from two replicate trials per treatment are shown, and error bars correspond to SDs).

Fig. 3

Evaluating the impacts of the tablet formulation on A. aegypti oviposition choice. (A) Gravid adult female A. aegypti are not attracted to water treated with tablets containing PCL. The mean number of eggs laid by a gravid female mosquito in three replicate trials per treatment are shown. Although gravid A. aegypti females were attracted to lay eggs in containers treated with tablets containing PLA, the mosquitoes preferred to lay eggs in untreated water vs. water containing tablets with PCL (A, Error bars correspond to SEM; *** = P < 0.001 with respect to the alternative container in each two-point assay, ANOVA/Tukey’s multiple comparisons test). Gravid adult female A. aegypti laid significantly more eggs in containers treated with tablets composed of 20:20:5:45:10 yeast:chitosan:microcrystalline cellulose:PLA:titanium dioxide (B; assays were pursued after the tablets had been subjected to accelerated storage conditions; results were compiled from 12 replicate trials/treatment, and error bars correspond to SEM; *** = P < 0.001 with respect to the untreated water container in each two point assay, Mann Whitney U Test).

Fig. 4

Controlled release syt.427 tablets or briquettes retain larvicidal activity conducted for several months in indoor and outdoor trials. Controlled-release 5 g sinking briquettes killed A. aegypti larvae in simulated field trials conducted in the insectary over the course of 5–6 months (A). The mean larvicidal results from two separate (five and six month) trials, each in which five releases of 100 larvae into 7.5 L containers with 3.5 L of water are shown in A (error bars correspond to SD). The mean mortality of all 10 replicate releases per treatment is shown in (B), and error bars correspond to SEM). A floating version of the syt.427 tablets performed well in semi-field trials conducted in Indiana during the summer of 2021, resulting in significant mortality of A. aegypti larvae in 200 L barrels (C; the mean results from replicate trials for each treatment conducted on 20 larvae are shown, and error bars correspond to SD). Sustained four-month larvicidal activity of sinking syt.427 tablets was observed in semi-field outdoor roof trials conducted in a 220 L barrel in St. Augustine, Trinidad (D, E). Larvicidal activity was maintained through three releases of 150 field strain larvae into the barrels (E); the mean mortality of larvae from the three releases is shown (D); error bars represent SEM. Throughout the figure, *** corresponds to P < 0.001 with respect to control, t-test, two-sample equal variance).