Resistance of Anopheles gambiae sensu lato to Pirimiphos-methyl Insecticide in Kakamega County, Highlands of Western Kenya

Background Insecticide treated bed nets and Indoor residual spraying remains the principal interventional malaria control strategies. To achieve malaria disease eradication, vector control programmes that monitor insecticide resistance profiles are necessary. Objective The study evaluated pirimiphos-methyl susceptibility of Anopheles gambiae sensu lato in Kakamega County, western Kenya. Methods Adult Anopheles gambiae sensu lato mosquitoes were assayed using World Health Organization tube bioassay against 0.25% pirimiphos-methyl. Susceptible and non-susceptible populations were characterized to species-level using Polymerase Chain Reaction. Susceptible and resistant mosquitoes were further subjected to G119S Acetylcholisterase (ace 1R) mutation detection. Results Anopheles arabiensis was the predominant species in all study population Mumias east (62%), Malava (68%), Ikolomani (77%) and Lurambi (82%). Results showed phenotypic susceptibility to pirimiphos-methyl. Mortality was low in Mumias east (80.6%) and high in Lurambi (89.0%). G119S mutations ranged from 3.0% to 8.9% in Anopheles arabiensis whereas G119S mutations were relatively low ranging from 0.0% to 3.1% in Anopheles gambiae s.s populations. Study populations tested were consistent with Hardy-Weinberg equilibrium (P>0.05). Conclusion We observed pirimiphos-methyl resistance in Anopheles arabiensis and Anopheles gambiae s.s. study populations. Results showed G119S mutation in resistance population. Resistance monitoring and management are urgently required.

ance to insecticides. Pyrethroids are the only recommended class of insecticide for use in both IRS and ITNs possibly because of limited effects to users 9 while carbamates, organochlorides, anorganophosphates are only applied in IRS 10 . Pyrethroids and organochlorines are neurotoxins that act by prolonging the voltage-gated sodium channel activation whereas the organophosphates act on the central nervous system by inhibiting the expression of acetylcholinesterase enzyme 11,12 . Pyrethroids resistance in malaria vectors has mostly been studied in the major malaria vector An. gambiae s.l of Africa [12][13][14][15][16][17] . The two major mechanisms associated with insecticide resistance are; a) Target site insensitivity in the sodium channel, causing a change in affinity between the insecticide and its binding site; b) metabolic detoxification of insecticides before reaching the target site 18. Several studies have confirmed the existence of pyrethroids and organochlorines insecticides resistance in some parts of western Kenya, with knockdown resistance (kdr) mutations at position 1014 in the sodium channel gene being commonly reported 8,19,20 . However, there is limited information on the status of pirimiphos-methyl insecticide resistance amongst the An. gambiae s.l complex in Kenya.
Owing to the increased resistance of malaria vectors to available insecticides, several alternative insecticides have been developed to control and eliminate mosquitoes. Pirimiphos-methyl (p-methyl) an organophosphate insecticide, has a fast action with minimal toxicity in both humans and environment 21 . The application of pirimiphos-methyl has been effective in Killing malaria vectors 22,23 . This however, could be because of its low eminence and high vapor pressure 24,25 . Indoor residual spraying controlled trial studies indicate the presence of high levels of toxicity of p-methyl to mosquitoes 26,27 . This led World Health Organization Pesticides Evaluation Scheme (WHOPES) to recommend use of p-methyl application as IRS for malaria vector control 28 . In Africa, resistance to insecticides such as pyrethroids and Dichloro-diphenyltrichloroethane (DDT) has frequently been associated with pesticide usage in agricultural farms 29,30 . For instance An. gambiae females lay their eggs in breeding sites around agricultural settings 29 , this consequently suggests that mosquito larvae may undergo selection pressure from agricultural pesticides favouring the occurrence of resistance to insecticides by malaria vectors 31-34. This study evaluated pirimiphos-methyl insecticides susceptibility status among Anopheles gambiae sensu lato in four randomly se-lected sub-counties in Kakamega County, highlands of western Kenya.

Study area
A cross-sectional survey was conducted in four highlands located in Kakamega County of western Kenya namely; Ikolomani; (E:00.16556, N:034.73194), Mumias East; (E:00.34120, N:034.54727), Lurambi; (E:00.31806, N:034.75222) and Malava; (E:00.32957, N:034.74701 ) (Figure 1). The study sites experienceseasonal malaria transmission , with high peaks occurring at the end of long (early April to late July) and short (October to November) rainy seasons. The main economic activity in the study areas is agriculture. The main food crops are sugarcane, maize farming and short-season farming such as vegetable farming. Multiple classes of pesticides such as organophosphate, pyrethroids and carbamates are used in controlling crop pests. The used pesticides end up seeping into the breeding habitats exposing the mosquito population to chemical residues.

Mosquito sampling and rearing
Anopheles gambiae s.l larvae were sampled in their natural breeding habitats using a standard 350-milliliter dipper from October 2018 to March 2019. Mosquito larvae were randomly collected per habitat and identified morphologically to Anopheline species using a key developed by Gillies 35. Collected larvae were pooled together in each study site then transferred to an insectary at Masinde Muliro University of Science and Technology for rearing at 80 ± 10% RH, 25 ± 2 °C, 12h:12h light/dark cycle. Larvae were cultured in spring water in small trays and fed on Tetramin™ baby fish food (TetraMin tropical flakes, Blacksburg, VA, USA) every day. Individual pupated larvae were collected and transferred into a 1ft square cage containing a net. Emerged adults were prvided with 10% sugar solution until they were 3 to 5 days old. The Kisumu strain of An. gambiae sensu stricto, a reference strain susceptible to all insecticides, was reared simultaneously under the same conditions and used as a control for insecticide bioassays..

WHO susceptibility bioassay
Emerged Adult females of An. gambiae s.l aged 3-5 days were tested for susceptibility to pirimiphos-methyl using the standard WHO tube bioassay (36) with 0.25% diagnostic dose as recommemded by WHO. For each population, 100 active female mosquitoes were tested. Paraffin oil-treated papers without insecticide were used as control. The knockdown time was recorded every 10 minutes within one hour exposure period. After 1 hour of exposure to the diagnostic concentration of pirimiphos-methyl (0.25%), experimental mosquitoes were transferred to recovery cups and maintained on 10% sucrose solution for 24 hours. Mortality was recorded at 24 hours post-exposure time. Knocked down mosquitoes after 1 hour and those that were alive after the 1 hour exposure period and still surviving 24 hours later were collected and stored individually in 90% alcohol for further molecular analysis. A susceptible mosquito was defined as a mosquito that died within 24 hours of recovery period or knocked down during the 1 hour exposure time while a resistant mosquito was referred as a mosquito that survived after the end of the 24 hours recovery period.

Molecular characterization of Anopheles gambiae s.l
Dead and Survived mosquitoes preserved after the bioassays were identified to species level using species-specific polymerase chain reaction (PCR) (PCR Model FTC 3105D, TECHNE, DUXFORD CAMBRIDGE, UK) assay previously described by Scott 37 . Deoxyribonucleic acid (DNA) was extracted from legs and wings of each mosquito using the ethanol precipitation method (38). The PCR technique was used to distinguish between the two sibling species of Anopheles gambiae s.l species complex native to western Kenya namely; An. gambiae s.s and An. arabiensis 38 .

Molecular detection of G119S Ace-1R mutation
Dead and survived samples of An. anopheles s.l post exposure were subjected to G119S Ace-1R analysis. The polymerase chain reaction was done using a protocol described by Essandoh and collaborators39. In summary, each reaction constituted a volume of 50ul containing 10 picomoles of each primers Ex2Agdir1 (5' AGG TCA CGG TGA GTC CGT ACG A 3') and Ex4A-grev2 (5' AGG GCG GAC AGC AGA TGA AGC GA 3'), 10mM dNTPs, ddH2O, 5X HF Phusion buffer and 1ul of Phusion Taq polymerase (Fermentas). The cycle parameters were: 1 cycle at 98oC for 4 mins, followed by 35 cycles of 98oC for 30 sec, 64 oC for 15 sec and 72oC for 30 secs with final extension at 72oC for 5 mins

Data management and analysis
Collected data were entered in MS-Excel spreadsheet, checked, cleaned after which it was coded and imported into SPSS version 19.0 (IBM SPSS Statistics for Windows, Version 19.0. Armonk, NY: IBM Corp) for analysis. The mortality rates of mosquitoes in the standard WHO tube resistance bioassay were calculated and adjusted using Abbott's formula 40. Susceptibility status of mosquito population was classified according to the WHO criteria (98-100 % mortality indicates susceptibility, 90-97 % mortality suggests the possibility of resistance that needs to be confirmed, and <90 % mortality suggests resistance) 36 . Heterozygous and homozygous mutation rates of Ace-1R gene loci (G119S) were calculated. To determine if these genotypes were under selection, the Hardy-Weinberg equilibrium test for Ace-1R genotypes was performed, and the Chisquare test used to determine the significance of the departure from Hardy-Weinberg equilibrium.

Ethics Statement
This study was approved by the Maseno University Ethical Review Committee (MUERC) under the scientific steering committee (MUERC/0061/18). The current study did not involve endangered or protected species.

Susceptibility status of the study populations
Mosquito populations collected from Mumias east and Malava showed phenotypic resistance to 0.25% p-methyl with mortality rates of 87% and 88% respectively ( Figure 2). There was suspected insecticide resistance in Ikolomani and Lurambi mosquito populations with both populations recording 91% mortality rate. At the species level (Table 1), the study observed phenotypic resistance in Anopheles arabiensis with mortality ranging from 80.6% to 89.0% in all the study populations. On the other hand, Anopheles gambiae s.s population from Ikolomani showed resistance to p-methyl (86.9%). There was suspected p-methyl resistance (90.6%) in Malava mosquito population while Mumias east showed mortality of 97.4% whereas the Lurambi population was 100% susceptible. WHO tube bioassay showed tat Kisumu strain was 100% susceptible, this confirms the quality of the insecticide-impregnated papers used.

Anopheles gambiae s.l sibling species characterization
All the Anopheles gambiae s.l exposed to p-methyl were subjected for sibling species identification using PCR. The two species of the An. gambiae complex in Kakamega County of western highlands, that is; An. arabiensis and An. gambiae s.s were found to be sympatric in the four surveyed sites with different frequencies.
Ace-1R G119S genotyping G119S mutations ranged from 3.0% to 8.9% in Anopheles arabiensis dominant populations, whereas G119S mutations were relatively low ranging from 0.0% to 3.1% in Anopheles gambiae s.s populations. Additionally, most of the mutations were heterozygous in Anopheles arabiensis (Table 2). No mutations were observed in Lurambi population. Hardy-Weinberg analysis showed that for the four An. gambiae s.s populations tested, only one population (Lurambi) had zero percent mutation whereas only Mumias East had only one rear homozygous G119S in An. arabiensis species ( Table 2). All the populations tested were consistent with Hardy-Weinberg equilibrium as the populations had P>0.05. After exposure to pirimiphos-methyl, the G119S mutation was also found to be present in survivors ( Table 3). The survival of Anopheles gambiae (s.l) populations from the three surveyed areas was associated with the presence of the mutation.

Discussion
The present study evaluated the resistance status of Anopheles gambiae sensu lato to pirimiphos-methyl insecticide. We observed that the predominant species in our study population was Anopheles arabiensis. This corroborates with similar studies in some parts of western Kenya and East African 13,41-44 . Phenotypic susceptibility tests revealed that An. arabiensis populations showed reduced mortality to p-methyl insecticide, according to the susceptibility threshold level 10 . This agrees with a previous study that reported the emergence of resistance among An. arabiensis to p-methyl in some regions of Ethiopia 45 .
Similarly, the findings mirror previous studies in Tanzania that reported low frequency of p-methyl resistance 23 . Additionally, resistance was observed in An. gambiae s.s from Ikolomani sub-county study population and suspected resistance in Malava sub-county. The observed scenario in this particular species may have been as a result of fewer numbers of An. gambiae s.s mosquitos in study populations. The observed resistance could be due to the application of agrochemicals pesticides and herbicides in agricultural farms that belongs to organophosphates class of insecticides/pesticides 33,39,[46][47][48] . It is approximated that 60% of farmers in western Kenya use pesticides for pest control 49 . Consequently, during the application of pesticides, residues seep into mosquito breeding habitats. This might have driven the observed emergence of p-methyl insecticide resistance in An. gambiae s.l 54,55 . However, the findings of this study are inconsistent with previous studies in Migori County that reported 100% susceptibility of An. gambiae s.l to p-methyl 27,56,57 . Migori County is one of the counties where IRS is in use. This implies that continuous application of insecticide is likely to be associated with resistance in mosquitoes 58 .   The study observed mutations in phenotypically resistance mosquitoes whereas some of the resistant mosquitoes did not exhibit the mutations. Possibly suggestion of other existence resistance mechanisms that may be driving resistance of Anopheles gambiae s.l. Therefore, the low frequency of G119S mutations is an alert for an exigent need to create measures to avoid allele fixation in the mosquito population

Conclusion
The study showed pirimiphos-methyl resistance in Anopheles arabiensis and Anopheles gambiae s.s. study populations. Results indicate G119S mutation in resistance population. Furthermore, there were wildtype phenotypically resistant populations that did not harbor G119S mutations suggesting other resistance mechanisms. Resistance monitoring and management of insecticide resistance are urgently required.

Competing interests
The authors declare that they have no competing interests.
Author's contributions NK, BG, DM and BN conceived and designed the study. KM, NK performed laboratory investigations and statistical analysis. NK, BG and BN co-drafted the manuscript. NS, MGM and MK critically revised the manuscript. All authors read and approved the final manuscript.

Consent to publish
Not applicable

Availability of data and materials
The data used to support the findings of this study are included within the article.

Funding
The study received partial funding from National Research Fund (NRF) Kenya, NRF/PhD/02/93.