Adult An. gambiae s.s. colonized at KEMRI-Wellcome Trust Research Programme insectary in 2011 from larvae collected in Mbogolo village of Kilifi county in Kenya were used for this study. This strain is fully susceptible to pyrethroids as it is routinely characterized quarterly for phenotypic resistance to pyrethroids using WHO tube tests. All experiments were carried out under insectary conditions at 28 ± 2 °C and 80 ± 10% relative humidity (RH) for rearing adults and at 30 ± 2 °C and 80 ± 10% RH for rearing larvae, under a 12 h light:12 h dark photoperiod. Adults were maintained on 10% glucose provided by soaking cotton wool pads in 10% glucose solution. Female adult mosquitoes, aged 2–5 days, were fed human blood through an artificial membrane to obtain eggs, which were the 86th filial generation since colonization. The eggs were collected on filter paper and immersed in deionized water for hatching. For first/second-instar larvae, hereafter referred to as L1/L2, sorting and counting of the larvae was done 2 days post-hatching, while for the third/fourth-instar larvae, hereafter referred to as L3/L4, sorting and counting was done 6 days post-hatching.
Ivermectin stock solution
For all experiments, 1% injectable Ivermectin (Ivermet®) was used. It was diluted in deionized water to prepare 50 ml of a stock ivermectin solution of 1000 ng/ml. All stock solutions were freshly prepared during the experiments.
About 100 larvae were counted and sorted into larval rearing trays and 250 ml deionized water containing the appropriate final concentration of ivermectin: 0.01, 0.001, 0.0001, or 0.00001 ppm was added. For controls, hereafter referred to as 0 ppm, only deionized water was added. This was done in triplicate for every concentration of ivermectin and replicated twice using two different batches of eggs yielding three technical replicates per biological replicate. In total, for every ivermectin treatment, ~ 600 L1/L2 and ~ 600 L3/L4 larvae were used. The larvae were maintained under normal rearing conditions at 30 ± 2 °C and 80 ± 10% relative humidity and maintained on equal weight of baby fish food (Tetramin®). At L1/L2 stages, larvae were fed 0.05 g fish food once daily, while at L3/L4 stages, larvae were fed 0.05 g fish food twice daily. Larval survival was monitored for 15 days, and pupae were collected daily for 3 days following the emergence of the first pupae. The daily number of pupae collected was counted, and the pupae were placed inside a 20 × 20 × 20-cm netted cage to eclose. To ensure that all the adults used in subsequent experiments were within the range of 3–5 days old, the pupation rate was censored at day 3 for all experimental conditions. Only pupae that emerged within the first 3 days were used in subsequent experiments. All pupae from the same larval rearing tray collected during the 3 day collection period were placed in the same cage. The emerging adult mosquitoes were maintained on 10% glucose solution. Blood feeding of the emerging adults was done 2 days after the last pupae were placed inside the cage; this ensured that the blood-fed adults were 3 to 5 days old. For all cages, blood feeding was done using 1 ml human blood in a membrane glass feeder.
Oviposition and fecundity assay
Oviposition cups were prepared by layering the base of a 250 ml plastic cups with a cotton wool pad moistened with deionized water. The cotton wool pad was covered with a filter paper, and the cup opening/mouth was covered using an insecticide-free net fastened with a rubber band. Gently, gravid females were transferred to individual oviposition cups. The mosquitoes were maintained on a 10% glucose solution and oviposition monitored daily for 10 days. After confirming the presence of eggs, mosquitoes were removed and killed by placing them in a −20 °C freezer for about 5 min. The filter paper retrieved from the oviposition cup with eggs was moistened with deionized water, and the eggs were allowed to hatch. Using a two-place denominator counter, the eggs were counted under a stereo microscope; eggs were counted to obtain the number of eggs hatched on one denominator and non-hatched eggs on the other denominator. The hatched eggs were identified by their dislodged operculum . The numbers of hatched and non-hatched eggs were recorded for each oviposition cup.
Adult susceptibility bioassay
Female adult mosquitoes, aged 3–5 days old, emerging from larvae exposed to ivermectin and their control of larvae not exposed to ivermectin were placed in 1-l mosquito holding cups previously described in . Mosquitoes from every larval exposure replicate were placed in a separate holding cup. The number of mosquitoes per holding cup ranged from 20 to 49. The mosquitoes were starved for 5 h before being membrane-fed on human blood spiked with ivermectin at a concentration of 11 ng/ml. For the non-exposed adult control, a similar amount of PBS was added to the blood in place of ivermectin. Mosquitoes were allowed to blood feed for about 45 min, after which, mosquitoes that were not fully engorged and dead mosquitoes were removed and recorded. As the amount of ivermectin taken by the mosquito is relative to the amount of blood consumed, only fully engorged mosquitoes were kept. This was done to standardize the amount of ivermectin consumed. Fully engorged mosquitoes were visually identified by a distended abdomen and presence of blood in the whole abdomen.
The adults were maintained at standard insectary conditions, 28 ± 2 °C and 80 ± 10% relative humidity. The adults were provided with 10% glucose and mortality monitored daily for 10 days. Dead mosquitoes were removed, counted, and recorded daily. For analysis, larvae not pre-exposed to ivermectin and not exposed to ivermectin as adults were used as the negative control. For the positive control, larvae not pre-exposed to ivermectin but exposed to ivermectin as adults were used.
All data were entered in MS Excel. Kaplan-Meier survival analyses were performed using R software, version 4.1.0. (R Core Team 2020). Log-rank test using a 5% significance was used to compare overall survival. For pairwise comparisons, Bonferroni was used to correct for multiple comparisons. Generalized negative binomial model with a log link was fitted to determine the relationship between time (days) taken to oviposition and ivermectin concentration and biological replicates (Table 2). The relationship between oviposition and predictors (ivermectin concentration and replicates) was performed using multivariable logistic regression analysis (Table 3). The influence of ivermectin concentration on total eggs laid by individual adult mosquitoes was assessed using generalized linear regression (Table 4). Generalized linear logistic regression was used to fit the relationship between the proportion of eggs hatched as the outcome variable and independent factors (ivermectin concentration, replicate) (Table 5). Predictors with p < 0.05 were considered independently associated with the outcome variables. Statistical analyses were performed using Stata version 15.0 (Stata Corp., College Station, TX).