The effect of silencing immunity related genes on longevity in the naturally occurring Anopheles arabiensis mosquito population of Southwest Ethiopia

Background In the fight against malaria, vector control remains the most important tool, butit is now severely constrained by the spread of insecticide or behavioral resistance by mosquito populations. Therefore, new vector control tools are warranted. Such novel tools include anti-mosquito vaccines or mosquito genetic modifications targeting the mosquito midgut homeostasis and reducing the mosquito lifespan beyond a stage they can transmit malaria. Methods We assessed the effect of RNA interference silencing of the midgut homeostasis regulators FN3D1, FN3D2, FN3D3, GPRGR9 and PGRPLC3 in populations of Anopheles arabiensis reared at nearly natural setting. We monitored the survival of gene-silenced mosquitoes and assessed the load of their midgut microbiota using flow cytometry. The effect of gene silencing was modeled by the Cox proportional hazards frailty model, and bacterial counts were first log transformed and then compared by a mixed model. Result Significantly higher mortality rates were observed for the FN3D1 (Hazard ratio =1.64, P=0.004), FN3D3 (HR=1.79, P<0.001) and GPRGr9 silenced mosquitoes (HR=2.00, P<0.001) as compared to a control group injected with dsRNA against a non-related bacterial gene LacZ. The bacterial load ratios for all target gene silenced mosquitoes compared to control mosquitoes were above 1, with the highest value for FN3D1 equal to 2.66 (95%CI: [0.94;7.57]) but no statistically significant difference could be demonstrated. Interestingly, there was a strong correlation (r=0.61) between the mortality hazard ratio and the bacterial count ratio of the gene-silenced mosquitoes. Increased mortality rates were reversed when the gene-silenced mosquitoes were treated with antibiotic mixtures suggesting that gut microbiota play a key role in the observed reduction of mosquito survival. Conclusion We demonstrate that interfering with the expression of theFN3D1, FN3D3 or GPRGr9 genes can cause a significant reduction of the longevity of An. arabiensis mosquitoes due to the disruption of the mosquito gut homeostasis.

143 (AARA003963) andPGRPLC3 (AARA002982). DsRNA for the LacZ gene that served as a control was 144 synthesized using as a template a plasmid containing the LacZ gene. The full primer sequences are 145 shown in S1. table. DsRNA synthesis was carried out using the TranscriptAid T7 High Yield Transcription 146 Kit (Thermoscientific, Lithuania). The dsRNA was then purified using the RNeasy Mini Kit (Qiagen, UK), 147 following the manufacturer's protocol. The concentration of dsRNA was determined 148 spectrophotometrically by Nano drop 1000 spectrometer at 260nm and adjusted to 3μg/μl using ultra-149 pure water. Gel electrophoresis (1%TBE agarose) was performed on a subsample of the PCR products 150 to confirm that products of the expected size were detected for each gene. Zero to two days old An.
151 arabiensis mosquitoes were injected with 69 nl dsRNA specific to a target gene or the LacZ control gene 152 following the RNA interference technique as described by [38].
153 Gene silencing efficiency was measured for each of the 5 silenced genes using qrtPCR. Quantification 154 of transcript abundance was performed on cDNA synthesized from total RNA extracted from 155 mosquitoes injected with dsRNA three days earlier and maintain on 10% sugar solution. Fast SYBR 9 157 Time PCR system (Applied Bio Systems, UK). Each target gene was quantified in duplicate. The AgS7 158 gene was used as an internal control. Primer sequences are given in S 2. Table. 159 160 Monitoring of mosquito survival 161 The dsRNA injected mosquitoes were put into their respectively labelled cups. For each gene between 162 20 and 30 mosquitoes were injected per replicate. The cups were kept in the microclimate regulatory 163 box described above. The mosquitoes were offered a 10% sugar solution daily and a blood meal every 164 fourth day by direct feeding on a goat. Survival was monitored daily for 25 days, starting 24 hours post 165 injection. For each gene 6 replicates of mosquito injection were performed.

167 Midgut microbiota analysis
168 For microbiota analysis, mosquitoes injected with the specified dsRNA were transferred into their 169 respective labelled cups. For each gene between 20 and 30 mosquitoes were injected and were kept in 170 a microclimate regulatory box. On day four post injection, the mosquitoes were allowed to feed on 171 blood or kept on sugar meal. Twenty-four (24) hours post feeding four blood-fed and four sugar fed 178 DsRNA-injected mosquitoes were placed in 6 different cups, each cup containing 20-30 mosquitoes.
179 The cups were kept in the microclimate regulatory box. On the day of dsRNA injection, the mosquitoes 180 were given a cotton ball soaked in antibiotic cocktail of Streptomycin and Norfloxacin, both a dose rate 181 of 10 μg/mL in a 10% sugar solution. On day 4post injection, the mosquitoes were blood fed on a goat.
182 The cotton balls were re-soaked with the antibiotic cocktail every fourth day for a period of 12 days (on 183 days 4, 8and 12). A blood meal was also offered on the same days. The mosquito survival was monitored     (Table 1). Gene silencing has no effect on the 228 Plasmodium falciparum infectivity of the mosquitoes (S3. Figure)  242 We further investigated whether there was a correlation between the mortality hazard ratio and the 243 bacterial count ratio of the different silenced genes, i.e., whether a high mortality hazard ratio of a 244 particular silenced gene corresponds with a high bacterial count ratio for that same gene. The 245 relationship is presented in Figure 4, and it shows correlation (r=0.61), although it fails to be significantly  276 They were also allowed to blood feed on a goat every 4 day, fully reproducing their natural habits. We 278 orGPRGr9expression significantly reduces the longevity of the An. arabiensis mosquitoes to an average 279 of 12, 13and 11days, respectively, compared to a 20-day average longevity of control mosquitoes.