Characterizing the molecular and metabolic mechanisms of insecticide resistance in Anopheles gambiae in Faranah, Guinea

Background In recent years, the scale-up of long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) has greatly reduced malaria transmission. However, malaria remains a global public health concern with the majority of the disease burden in sub-Saharan Africa. Insecticide resistance is a growing problem among Anopheles vector populations, with potential implications for the continued effectiveness of available control interventions. Improved understanding of current resistance levels and underlying mechanisms is essential to design appropriate management strategies and to mitigate future selection for resistance. Methods Anopheles gambiae sensu lato mosquitoes were collected from three villages in Faranah Prefecture, Guinea and their levels of susceptibility to seven insecticides were measured using CDC resistance intensity bioassays. Synergist assays with piperonyl butoxide (PBO) were also undertaken to assess the role of elevated mixed-function oxidases in resistance. Five hundred and sixty-three mosquitoes underwent molecular characterization of vector species, presence of target site mutations (L1014F kdr, N1575Y and G119S Ace-1), Plasmodium falciparum infection, and relative expression of three metabolic genes (CYP6M2, CYP6P3 and GSTD3). Results In Faranah, resistance to permethrin and deltamethrin was observed, as well as possible resistance to bendiocarb. All assayed vector populations were fully susceptible to alpha-cypermethrin, pirimiphos-methyl, clothianidin and chlorfenapyr. Plasmodium falciparum infection was detected in 7.3% (37/508) of mosquitoes tested. The L1014F kdr mutation was found in 100% of a sub-sample of 60 mosquitoes, supporting its fixation in the region. The N1575Y mutation was identified in 20% (113/561) of individuals, with ongoing selection evidenced by significant deviations from Hardy–Weinberg equilibrium. The G119S Ace-1 mutation was detected in 62.1% (18/29) of mosquitoes tested and was highly predictive of bendiocarb bioassay survival. The metabolic resistance genes, CYP6M2, CYP6P3 and GSTD3, were found to be overexpressed in wild resistant and susceptible An. gambiae sensu stricto populations, compared to a susceptible G3 colony. Furthermore, CYP6P3 was significantly overexpressed in bendiocarb survivors, implicating its potential role in carbamate resistance in Faranah. Conclusions Identification of intense resistance to permethrin and deltamethrin in Faranah, is of concern, as the Guinea National Malaria Control Programme (NMCP) relies exclusively on the distribution of pyrethroid-treated LLINs for vector control. Study findings will be used to guide current and future control strategies in the region.

Background Despite impressive progress made towards the control and elimination of malaria, this disease remains the leading cause of morbidity and mortality in the tropics, where it is estimated to have resulted in the deaths of approximately 435,000 individuals in 2017 [1]. Between 2010 and 2017, global malaria incidence has fallen by approximately 18% globally (72 to 59 cases per 1000 at risk) and by 20% in the World Health Organization African region [1], which still bears the greatest disease burden [2]. Malaria deaths have been decreasing annually, largely due to the scale-up of long-lasting insecticidal nets (LLINs) [3] and implementation of indoor residual spraying (IRS) [2]. However, progress has stalled in some areas, with an increase of 2 million cases from 2016 to 2017 [1].
Long-term intensive insecticide use to control agricultural pests and disease vectors has resulted in the selection of resistance in many insect species [4]. The widespread use of dichlorodiphenyltrichloroethane (DDT) in the 1950-1960s, followed by the recent increase in distribution of LLINs impregnated with pyrethroids, and the broad use of the same insecticides in the agricultural industry has led to the development of resistance in mosquito populations worldwide [5]. This resistance poses a major threat to malaria control [4], where vector control is reliant primarily on insecticide-based interventions [6]. Of the 80 malaria-endemic countries for which data are available from 2010 onwards, 68 countries detected decreased susceptibility to at least one insecticide among vector populations, with 57 countries reporting resistance to two or more chemical classes [1].
In Guinea, malaria remains one of the most significant diseases of public health importance, with 92% of infections caused by Plasmodium falciparum [7]. The national malaria prevalence is approximately 15%, reaching up to 25% in Faranah Prefecture [8]. Guinea's tropical climate allows for year-round malaria transmission, with peak transmission from July through to October in most areas. The Guinean national vector control strategy focuses almost exclusively on the distribution of LLINs, with IRS occurring in only 1.7% of households, primarily those of workers engaged in mining operations [9]. The United States President's Malaria Initiative (PMI) estimates that intervention coverage remains low, with only 48% of households owning at least one insecticide-treated net (ITN) per every two members [9]. Given the reliance on LLINs for malaria control, the detection of nationwide pyrethroid resistance is of concern [10]. In order to safeguard malaria control efforts in the country, current insecticide resistance levels and underlying mechanisms were characterized among vector populations, to design appropriate management strategies and mitigate future selection for resistance.

Study sites and mosquito collections
Human landing catches (HLCs) were conducted at three sites in Faranah Prefecture (Fig. 1). The villages of Balayani (10.1325, − 10.7443), Foulaya (10.144633, − 10.749717), and Tindo (9.9612230, − 10.7016560) were selected, due to their high malaria prevalence, and were visited every other day until a sufficient sample size for testing was acquired. Following consent from the household owner, two to three fieldworkers, positioned outside of the house, collected mosquitoes landing and attempting to feed on their exposed legs and feet from 18.00 to 07.00. Mosquitoes were transported back to the insectary in Faranah and provided with 10% sucrose solution prior to bioassay testing.
Larval collections were also performed in Faranah (10.042423, − 10.740980), at sites selected through active searching and/or known to have been productive in previous years. Larvae and pupae were collected using larval dippers, ladles, buckets, and pans. Collections were then transported back to the insectary in Faranah where they were fed on ground fish food (TetraMin ® Tropical Fish Food Flakes). Pupae were removed on a daily basis, placed in plastic cups, and transferred to mosquito cages. Emerging adult mosquitoes were provided 10% sucrose solution prior to bioassay testing. All mosquito samples were collected between 25th June and 20th July 2018, at the start of the long rainy season.

CDC resistance intensity and synergist bioassays
Centers for Disease Control and Prevention bottle bioassays were performed using 250 mL Wheaton bottles with adult female mosquitoes of varying ages caught in HLCs from the three villages (each village was tested separately), or 2-5-day old female mosquitoes raised from larvae in the insectary. Mosquitoes collected using different methods were tested separately in bioassays. Bioassays were performed at the Centre de Santé Marché in Faranah; mosquitoes were held in the insectary for no more than 48 h prior to testing. Following CDC guidelines, bottles were coated with alpha-cypermethrin (12.5 μg/bottle), deltamethrin (12.5 μg/bottle), permethrin (21.5 μg/bottle), and bendiocarb (12.5 μg/bottle) at 1, 2, 5, and 10 times the diagnostic dose, and with pirimiphos-methyl (20 μg/bottle), clothianidin (90 μg/bottle), and chlorfenapyr (100 μg/bottle) at the diagnostic dose. Stock solutions for all insecticides and synergists were prepared using 95-98% ethanol as a solvent for all, clothianidin also included 58.8 μg Mero (Sigma-Aldrich, USA), to circumvent issues previously experienced with bottle coating and insecticide crystallization. Approximately 20-25 mosquitoes were introduced into each assay bottle and mortality was recorded in all bottles at the start Esri, HERE, Garmin, © OpenStreetMap contributors, and the GIS user community, Source: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and the GIS User Community 0 1 50 300 75 Kilometers T i n d o T i n d o F a r a n a h F a r a n a h F o u l a y a F o u l a y a B a l a y a n i B a l a y a n i  of the assay and at 15-min intervals for up to 30 min (for  alpha-cypermethrin, deltamethrin, permethrin, bendio-carb, and pirimiphos-methyl) or 60 min (for clothianidin and chlorfenapyr). Mosquitoes exposed to chlorfenapyr were held for an additional 24 h in paper cups, with access to 10% sugar solution. In each bioassay, a control bottle, coated with 98% ethanol, was run in parallel. Mortality was defined as the inability of a mosquito to stand or fly in a coordinated manner [11]. Synergist CDC bottle bioassays were performed using piperonyl butoxide (PBO) to investigate the potential role of detoxifying enzyme families in resistance. Bottles were coated with 100 μg of PBO and mosquitoes were exposed for 60 min, followed by exposure to pyrethroid treated bottles. Adult mosquitoes reared from larvae were exposed for 30 min to permethrin and deltamethrin for comparison to wild caught adults. Multiple replicates were performed per insecticide and study village. When limited by mosquito availability, for insecticides where there was a clear lack of resistance, i.e. 100% of exposed mosquitoes dying/ becoming knocked-down very quickly into the exposure period, priority was given to testing remaining mosquitoes with doses of insecticides where initial resistance to the diagnostic dose was observed. At the end of each exposure period, separate individual surviving (resistant), knocked-down or dead (susceptible), and control mosquitoes were dipped in ethanol and then preserved in RNAlater ® (Thermo Fisher Scientific, UK) at 4 °C in Faranah, for a maximum of 3 weeks, and subsequently at − 80 °C for downstream molecular analyses at the London School of Hygiene and Tropical Medicine (LSHTM).

RNA extraction
Single mosquitoes were homogenized using a Qiagen Tissue Lyser II (Qiagen, Hilden, Germany) with 3 mm stainless steel beads and RNA was extracted using Qiagen 96 RNeasy Kits according to the manufacturer's instructions (Qiagen, Hilden, Germany). RNA was eluted in 45 μL of RNase-free water and stored at − 70 °C. Specimens were selected from all three study villages, including representative susceptible and resistant individuals per insecticide per concentration. A High Capacity cDNA Reverse Transcription kit (Applied Biosystems) was used to perform reverse transcription on eluted RNA. Reactions were performed in a Bio-Rad T100 ™ Thermal Cycler which cycled for 10 min at 25 °C, 120 min at 37 °C, and 5 min at 85 °C. The resulting cDNA was then stored at − 20 °C.

Identification of Anopheles gambiae species complex
Mosquitoes were morphologically identified as Anopheles gambiae sensu lato (s.l.) in the field prior to CDC bottle bioassay testing. A sub-set of 480 samples were further identified using an end point PCR assay developed by Santolamazza et al. [12]. This assay amplifies the SINE200 insertion, a highly repetitive ~ 200 bp element which is widespread in the An. gambiae sensu stricto (s.s.) genome [12]. Samples were prepared with forward (5′-TCG CCT TAG ACC TTG CGT TA-3′) and reverse (5′-CGC TTC AAG AAT TCG AGA TAC-3′) primers, and amplifications performed in 20 µL reactions containing 2 µL cDNA, 2 µL of each primer (

Plasmodium falciparum detection
A total of 508 whole body mosquito samples (collected in HLCs) were tested for the presence of P. falciparum using a real-time assay targeting the cytochrome c oxidase subunit 1 (cox1) mitochondrial gene of P. falciparum according to Boissière et al. [13]. This sample includes all specimens which underwent PCR for species identification and were subsequently processed for metabolic gene expression (see below). This assay is highly sensitive and specific, capable of detecting the target gene at all stages of the P. falciparum life cycle [13]. Samples were prepared with forward (5′-TTA CAT CAG GAA TGT TAT TGC-3′) and reverse (5′-ATA TTG GAT CTC CTG CAA AT-3′) primers, and amplifications performed in 10 µL reactions containing 2 µL cDNA, 1 µL of each primer (10 µM

Characterization of resistance mutations: target site mutations
A sub-sample of 60 mosquitoes was selected to be tested for the West African L1014F kdr mutation, given the high frequency of this allele and its fixation in many parts of Guinea and West Africa [10,14]. The PCR master mix was prepared according to MR4 guidelines [15]. Detection of the N1575Y mutation was carried out with 570 samples (including the 60 individuals who tested positive for L1014F kdr), using a TaqMan PCR assay developed by Jones et al. [16]. Forward primer (5′-TGG ATC GCT AGA AAT GTT CAT GAC A-3′), reverse primer (5′-CGA GGA ATT GCC TTT AGA GGT TTC T-3′), and two probes: Yprobe (5′-TTT TTC ATT GCA TAA TAG TAC-3′) and Nprobe (5′-ATT TTT TTC ATT GCA TTA TAG TAC -3′) were used to detect the presence of the wild type and the mutation allele. HEX and FAM fluorophores were used due to their different excitation wavelengths, ensuring no interference: excitation of HEX showed no mutation, while excitation of HEX and FAM at similar Ct values indicated the N1575Y mutation. 20 µL reactions containing 2 µL cDNA, 1 µL each primer (10 µM), 0.5 µL each probe, 5 µL H 2 O, and 10 µL QuantiTect Probe Master Mix were prepared in plates and run on an Agilent Technologies Stratagene Mx3005P qPCR system and cycled according to the Quantitect ™ Probe PCR Handbook guidelines (15 min at 95 °C; 35 cycles of 15 s at 95 °C and 60 s at 60 °C). Positive controls from gDNA extracted from known An. gambiae s.s. with the N1575Y mutation and without the mutation were included on each run in addition to no template controls (NTCs).
A subsample of 30 mosquitoes which were resistant or susceptible to bendiocarb were tested for the presence of the G119S Ace-1 mutation using a TaqMan PCR assay, according to Weill et al. [17]. Samples were prepared with degenerate primers Moustdir1 (5′-CCGGGNGCSACY ATG TGGAA-3′) and Moustrev1 (5′-ACGATMACG TTC TCY TCC GA-3′), and amplifications performed in 20 µL reactions containing 2 µL cDNA, 2 µL each primer (10 µM), 4 µL H 2 O, and 10 µL 2× Hot Start Taq PCR Master Mix (New England Biolabs). Samples were loaded into a Biorad T100 ™ Thermal Cycler for 3 min at 95 °C, followed by 30 cycles of 30 s at 95 °C, 30 s at 52 °C, and 1 min at 72 °C, and a final step of 10 min at 72 °C. The resulting PCR fragments were then digested with AluI restriction enzyme (Thermo Scientific) for 16 h, according to the manufacturer's instructions, and run on 2% agarose gels in an Invitrogen E-gel iBase ™ Real-Time Transilluminator. 194 bp undigested PCR products indicated the susceptible allele and 74 bp and 120 bp digested fragments indicated the presence of the resistant allele. Presence of all three product sizes indicated the sample was heterozygous. Positive controls from gDNA extracted from known An. gambiae s.s. that were homozygous susceptible (SS), homozygous resistant (RR) and heterozygous individuals (RS) for G119S Ace-1 were included in addition to no template controls (NTCs).

Characterization of resistance mechanisms: metabolic gene expression
The relative gene expression of two cytochromedependent monooxygenases: CYP6P3, CYP6M2, and glutathione-s-transferase, GSTD3, was analysed in 461 individuals from Guinea and 41 susceptible G3 individuals from a colony at LSHTM, using quantitative reverse transcriptase PCR (qRT-PCR) relative to the housekeeping gene ribosomal protein S7 (RPS7), according to Yahouédo et al. [18]. RPS7 was selected as an endogenous reference gene, commonly used to normalize RNA expression levels, between Anopheles populations. These genes were targeted based upon their significant overexpression in other neighbouring West African vector populations [18,19]. Each gene used the following primers: RPS7 forward (5′-ATT GCC GAG CGC CGC ATT CT-3′) and reverse (5′-GAC GCG GAT ACG CTT GCC GA-3′) primers, CYP6M2 forward (5′-TCG GGA TGT GTG CGT TCG GC-3′) and reverse (5′-TCG TGT CTC GCA CCG CGT TC-3′) primers, CYP6P3 forward (5′-TGT GAT TGA CGA AAC CCT TCG GAA G-3′) and reverse (5′-ATA GTC CAC AGA CGG TAC GCGGG-3′) primers, and GSTD3 forward (5′-CTA AGC TTA ATC CGC AAC ATA CCA -3′) and reverse (5′-GTG TCA TCC TTG CCG TAC AC-3′) primers. For each gene, 10 µL reactions were prepared containing 2 µL cDNA, 1 µL each primer (10 µM), 1 µL H 2 O, and 5 µL 2× Roche FastStart Essential DNA green master mix containing SYBR Green. Prepared reactions were loaded into the Agilent Technologies Stratagene Mx3005P qPCR system which cycled for 10 min at 95 °C; 35 cycles of 10 s at 95 °C, 22 s at 60 °C, and 10 s at 72 °C; followed by a melt curve. Serial dilutions were performed on selected samples for each of the four genes and relative standard curves produced using the Stratagene MxPro qPCR software (Agilent Technologies). Using the same software, sample Ct values could then be used to generate relative quantities, accounting for each assays' efficiency, and the expression level of each metabolic gene tested could be normalized to the housekeeping gene RPS7.

Data analysis
Data were recorded on pre-prepared data sheets and entered into an Excel spreadsheet. Control mortality in bioassays never exceeded 5%, thus correction using Abbott's formula was not necessary. Mosquito mortality was analysed according to WHO criteria: 98-100% mortality at 30 min of exposure indicates 'susceptibility' , 90-97% mortality suggests 'possible resistance' and < 90% indicates the presence of 'resistance' [4]. Graph-Pad Prism 7 (GraphPad Software) was used for statistical analysis (t-tests, Fisher's exact tests and Chi-squared tests). Microsoft Excel was used to calculate proportions and construct resistance graphs. Stratagene MxPro qPCR software (Agilent Technologies) was used to produce relative standard curves for genotypic analysis.
Considering HLC-collected adult mosquitoes from Faranah Prefecture as a whole, resistance was consistently observed to permethrin and deltamethrin. Permethrin gave the lowest mortality of all insecticides tested with 4% [95% CI 1%, 11%] and 15% [95% CI 9%, 24%] mosquito mortality at 1× and 2× the diagnostic doses, respectively ( Fig. 2 and Table 1). Resistance to deltamethrin was also evident, but to a lesser degree; average mosquito mortality to the diagnostic dose was 86% [95% CI 77%, 91%] (Fig. 2). Possible resistance was observed to bendiocarb, with mosquito mortality ranging between 94 and 97% at 1×, 2×, and 5× concentrations. Mosquitoes were found to be susceptible to the diagnostic doses of all other insecticides with mortalities > 98% for alpha-cypermethrin, chlorfenapyr, clothianidin, and pirimiphosmethyl ( Fig. 2   Synergist bioassays were performed on a sub-sample of mosquitoes from Balayani. Pre-exposure to PBO and subsequent permethrin or deltamethrin treatment resulted in partial or complete abolishment of resistance. Mortality to permethrin increased from 6 to 87% at 1×, and from 10 to 100% at 2×; mortality following deltamethrin exposure increased from 79 to 100% at 1×.

Characterization of resistance mechanisms: metabolic gene expression
461 samples identified as An. gambiae s.s. were tested for the expression of CYP6M2, CYP6P3, and GSTD3 genes relative to the housekeeping gene RPS7 and compared to 41 An. gambiae s.s. samples from a susceptible G3 colony. These three genes were significantly overexpressed in the majority of wild-caught An. gambiae s.s. when compared to the susceptible G3 laboratory strain (Table 4 and Fig. 3).
The greatest changes in gene expression were observed for GSTD3, with an average mean fold change of 1.19 among wild An. gambiae s.s. compared to 0.44 in G3 colony individuals; average levels of CYP6M2 and CYP6P3 were 0.84 and 0.79 compared to 0.09 and 0.12 between field and colony mosquitoes, respectively. Among Guinean vectors, a significantly higher expression of CYP6P3 was observed between individuals which survived bendiocarb exposure at 1× and 2×, compared to those that died (1.44 vs. 0.62; p = 0.0524 and 1.68 vs. 0.55; p = 0.0366, respectively) ( Table 4). However, no significant changes in gene expression of CYP6M2 and GSTD3 were apparent between wild An. gambiae s.s. which survived or died after insecticide exposure at any dose.

Discussion
The susceptibility of An. gambiae s.l. populations from Faranah, Guinea to seven public health insecticides were assessed and underlying resistance mechanisms characterized. Intense resistance to permethrin and deltamethrin was apparent, with survivors at 5× and 2× the diagnostic doses, respectively; minor heterogeneity in mosquito mortality was also evident across this restricted geographic area. By comparison, no evidence of alpha-cypermethrin resistance was observed in Faranah, with 100% mosquito mortality following exposure. Increased tolerance to particular pyrethroids in this region is not unexpected, given LLINs are the sole insecticide-based malaria control strategy implemented in Guinea. Deltamethrin LLINs were distributed nationwide in 2013 (Netprotect ® ) and 2016 (PermaNet ® 2.0 and Yorkool ® ) by the National Malaria Control Programme (NMCP) [10], thus a resulting increase in deltamethrin resistance would be anticipated. However, the higher levels of permethrin resistance demonstrated in this study could be the result of control efforts prior to 2013 or concurrent use of this insecticide (and/or others capable of facilitating cross-resistance) in agriculture activities. Alternatively, L1014F kdr, which our study indicated was fixed in Faranah in 2018, as well as in other parts of Guinea [10,20], has been reported to contribute more to resistance to type I (permethrin) vs type II (alpha-cypermethrin and deltamethrin) pyrethroids [21].
In Faranah, susceptibility to all other classes of insecticides under evaluation was confirmed, excluding the carbamate, bendiocarb, which showed the development of possible resistance with mortality ranging between 94 and 97% in the region. This emerging resistance may also be attributable to carbamate use in the agricultural setting [10], as has been documented previously in other parts of Guinea [20]. The organophosphate, pirimiphosmethyl, achieved complete mosquito mortality in this study. Currently, large-scale government-funded, widespread IRS activities are not underway in Guinea. However, knowledge of susceptibility in relation to current LLIN use will help to guide potential IRS implementation in the future [9]. Local mosquito populations were also susceptible to two new insecticides under consideration for public health use, clothianidin (neonicotinoid) and chlorfenapyr (pyrrole), strengthening the evidence that net impregnations and IRS formulations with these insecticides may be successful in future vector control efforts. Likewise, absence of alpha-cypermethrin resistance supports the deployment of Interceptor G2 ® LLINs (containing a combination of alpha-cypermethrin and chlorfenapyr) in prospective LLIN distributions.
Regarding underlying resistance mechanisms, decreased susceptibility to pyrethroids in Faranah was mediated both by target site mutations and overexpression of metabolic enzymes. The L1014F kdr mutation was detected in all tested samples, which is consistent with high frequencies of this mutation observed throughout the country [10,20]. This study did not screen for the East African L1014S kdr mutation, which is known to have variable frequencies in West Africa and may warrant surveillance in Guinea in the future [22][23][24]. A    [16]. By comparison, in Maferinyah, Guinea, this mutation was present at similar frequencies to Faranah and directly implicated in phenotypic resistance to permethrin [20]. The ability of the synergist PBO to re-establish full or partial susceptibility to both permethrin and deltamethrin indicates that mixed-function oxidases (MFOs) play some role in the resistance reported. However, the expression of metabolic enzymes examined in this study (CYP6M2, CYP6P3, and GSTD3), while significantly upregulated in wild An. gambiae s.s. compared to the susceptible G3 colony, did not statistically differ between our pyrethroid resistant and susceptible wild-caught mosquitoes. Further exploratory analyses of our wild populations are warranted to characterize additional metabolic resistance pathways. Given the influence of MFOs as a predominant resistance mechanism to pyrethroids in the area and recent findings of improved protection with permethrin and PBO impregnated nets [25], the Guinea NMCP should consider including these next-generation nets in future vector control activities.
The G119S Ace-1 mutation, which confers resistance to carbamate and organophosphate insecticides [17] was found in this study to be highly predictive of bioassay survival to bendiocarb. Furthermore, CYP6P3 was significantly overexpressed in bendiocarb survivors, suggesting it may also be responsible for decreased susceptibility to carbamates in Faranah. Similar results implicating a role for CYP6P3 in bendiocarb metabolism and resistance have been reported from Côte d'Ivoire [26] and overexpression of this CYP450 has been documented among many multi-insecticide resistant field populations [5,[27][28][29][30].
In this study, mosquitoes that were reared from larvae collected from sites in the town of Faranah showed a higher mortality rate to permethrin at 1× and 2× the diagnostic dose, when compared to the wild-caught adult population of varying physiological age. This finding contradicts previous studies which have suggested that field-collected adults have higher mortality rates due to the mixture of ages and blood-feeding statuses in these populations [31], and the reported decline of phenotypic resistance with increasing mosquito age [32]. This observation strengthens the importance of utilizing the same methods of mosquito collection and assay preparation when testing field-collected mosquitoes [31] and cautions against the use of such data interchangeably. Because mosquito availability was a limiting factor, testing priority for replicates was given to insecticides where initial resistance was observed to the diagnostic dose, to more accurately characterize levels of phenotypic resistance intensity. Additional bioassay testing of larger samples sizes is required to corroborate these results in Faranah.
Malaria prevalence in Faranah Prefecture is approximately 15%, reaching up to 25% [8]; mosquito infection rates in this study were 7.3%. Because whole mosquito bodies were screened for P. falciparum by PCR, it is also possible that, in addition to infection or infectivity, a positive sample could be indicative of a mosquito which recently fed upon an infected human blood meal. More importantly, insecticide susceptible vectors were more likely to be infected with P. falciparum, but this phenomenon could not be ascribed to a single insecticide or resistance mechanism.
Additional considerations for the interpretation of study data include that in the tested population of An. gambiae s.l., An. gambiae s.s. was the predominant species at 97.1%, with An. coluzzii representing 1.3%, and the hybrid form 1.3%. Only two samples failed to produce any result on the An. gambiae complex end-point PCR assay. Anopheles gambiae s.s. breed in temporary, rain-dependent breeding sites [33], which could account for its predominance during the rainy season. Future collections at multiple timepoints, utilizing different collection methods would be necessary to determine the principal species in the area. Regarding sampling methodology, in this study mosquitoes were primarily collected using HLCs as this is often the most efficient technique to obtain large catches for bioassay testing. Due to heavy rains in the areas, larval breeding sites can be completely washed away, rendering this sampling method unpredictable. HLCs were chosen in favour of aspirating indoor resting, blood-fed adults and forcing oviposition, to generate an F 1 population, because house collections frequently yield low numbers of mosquitoes, potentially introducing family effects as a bias. In addition, preference was given to testing mosquitoes caught in HLCs, because the latter two techniques are not appropriate for pathogen screening. Finally, sampling took place in only three villages in one area of Guinea and, therefore, these findings cannot be extrapolated across the entire country. However, this information will contribute to an understanding of resistance in this region, which will be valuable information for the NMCP [34].