Elevated Plasmodium infection and multiple insecticide resistance in the principal malaria vectors Anopheles funestus and Anopheles gambiae in a forested locality close to the airport of Yaoundé, the Capital city of Cameroon

Reducing the burden of malaria requires better understanding of vector populations particularly in forested regions where the incidence remains elevated. Here, we characterized malaria vectors in a locality near the international airport of Yaoundé, the capital city of Cameroon, including species composition, abundance, Plasmodium infection rate; insecticide resistance profiles and underlying resistance mechanisms. Methods Blood-fed adult mosquitoes resting indoors were aspirated from houses in April 2019 at Elende, a village located 2km from the Yaoundé-Nsimalen airport. Female mosquitoes were forced to lay eggs to generate F 1 adult progeny. WHO bioassays were performed to assess resistance profile to insecticides. The threshold of insecticide susceptibility was fixed above 98% mortality rate and mortality rates below 90% were considered indicative of confirmed insecticide resistance according to the WHO test procedures for insecticide resistance monitoring. Furthermore, the molecular basis of resistance and Plasmodium infection rates were investigated.

the high pyrethroid/DDT resistances in An. gambiae s.s. corresponded with an increase frequency of 1014F kdr allele (95%). The CYP6P9a P450 resistance allele was absent in the An. funestus s.s. population meanwhile transcriptional profiling of candidate cytochrome P450 genes reveals the over-expression of CYP6P5 , CYP6P9a and CYP6P9b.

Conclusion
The resistance to multiple insecticide classes observed in these vector populations alongside the high Plasmodium infection rate highlights the challenges that public health vector control programs encounter in sustaining the regular effectiveness of contemporary insecticide-based control interventions in forested areas. More particularly, the resistance observed against the carbamates and possible resistance against the organophosphates constitutes a major concern for IRS.

Background
Malaria is the major vector-borne disease globally and a leading public health problem [1].
In 2018, there were roughly 228 million cases of the disease and about 405,000 malariarelated deaths. Approximately 70% and 20% of deaths recorded were children aged below five years and pregnant women respectively [1]. Although, a shift in focus from malaria control to elimination was declared by the WHO in 2012, it was observed that between 2015 and 2018; no considerable progress was achieved in decreasing global malaria cases. Rather, there was a reported increase in malaria cases in 2018 compared with the previous years in ten African countries scoring the highest burden of the disease [1].
To this effect, the WHO Global Technical Strategy for Malaria (2016-2030) outlines a pathway for malaria control and elimination and designates a target for a 90% reduction in global malaria mortality rates by 2030 relative to a 2015 baseline [2]. In this vein, the recent certification of Algeria and Argentina as malaria free countries by the World Health Organization (WHO) has been a historic achievement for universal health coverage, and serving as a model in demonstrating the feasibility of malaria elimination in the Afrotropical region [1]. This success was in part largely attributed to a coordinated system of vector control interventions such as long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) among others including prompt diagnosis, effective treatment and efficient surveillance response system [3]. Despite this, the efficacy of these insecticide based-vector control tools is compromised by the growing problem of insecticide resistance widely exhibited by Anopheles vectors across the African continent [4].
In Cameroon, malaria is endemic with the entire population considered to be at risk.
Available data from health facilities reported 23.6% of cases, 68.7% of deaths among children aged less than 5 years and 16.9 % of pregnancy related deaths attributed to malaria [5]. In order to reduce the malaria burden, the Cameroonian government supports the National Malaria Control Program (NMCP) and other partners who have established a strategic plan to achieve her goal of ensuring equal access to quality and affordable tools necessary for sustaining malaria control and elimination. The malaria interventions include mass distribution of long lasting insecticide nets (LLINs), prompt and effective diagnosis, artemisinin-based combination therapy (ACTs), seasonal malaria chemoprevention (SMC) and intermittent preventive therapy (IPT) in pregnant women through administration of sulfadoxine-pyrimethamine [5]. Moreover, to further strengthen the current vector control intervention, an indoor residual spray (IRS) campaign is to be launched in Cameroon for the first time by the United States President Malaria Initiative (PMI) project. Unfortunately, the wide spread nature of insecticide resistance in malaria vectors to common insecticides used to impregnate nets and spray walls in Cameroon [6][7][8][9][10]  LLINs is the main prevention method with a coverage of around 70% [5] as the area is endemic for malaria.
Following verbal approval from the chief and household heads; blood fed indoor resting adult mosquitoes (F0s) were collected between 06.00am and 11.00am. The Prokopack electrical aspirator (John W Hook, Gainesville, FL, USA) was used for mosquito collection after which they were kept in a humid cage and later transported to the insectarium of the Centre for Research in Infectious Diseases (CRID), Yaoundé. Field collected females were placed in a cage to rest for 1 hour prior to morphological identification [13]. Basically, each live mosquito sample was aspirated from the cage in to a hemolysis tube and observed microscopically for distinct morphological differences of wings, mouthparts and size based on the Afro-tropical anopheline key [13]. Specimens of the An. funestus group and An. gambiae complex were then placed in two separate small-sized labeled cages and allowed for 4-5 days feeding on 10% sugar soaked in cotton wool for them to become fully gravid. A forced-egg laying method as previously described [14,15] was utilized for individual oviposition of females. After oviposition, all the carcasses of the F 0 were kept in separate 1.5ml tubes containing silica gel and stored at -20°C prior to molecular analysis.

Molecular species identification
The Livak method was used to extract genomic DNA from a subset of each individually oviposited field-caught female (F 0 s) [16]. DNA extracts were quantified using NanoDrop™ spectrophotometer (Thermo Scientific, Wilmington, USA). A SINE-200 PCR [17] and cocktail PCR [18] were performed to identify the different species within the An. gambiae s.l.
complex and An. funestus s.l. group respectively. The ribosomal DNA internal transcribed spacer region 2 (ITS2) was amplified to identify the undetermined species using the protocol of Hackett et al [19]. The PCR amplicon of 840bp was purified with Exonuclease I (Exo I) and Shrimp Alkaline Phosphate method (Exo-SAP) based on the New England Biolabs procedure (NEB, MA, USA) and directly sequenced commercially.

Plasmodium infection rates
Plasmodium infection was assessed in 150 An. funestus s.s. and 39 An. gambiae s.s. F0 females. Genomic DNA was extracted from the head and thorax of each specimen, and infection with P. falciparum or OVM+ (P. ovale, P. vivax and P. malariae) was detected with the TaqMan assay as described previously [20]. Sequentially, nested PCR was conducted on all the positive samples to confirm the TaqMan assay results and to specifically differentiate between the OVM species obtained [21].

WHO Insecticide susceptibility bioassays
Various insecticides employed in control of malaria vectors were tested in bioassays to assess the resistance profile of the An. funestus s.s. and An. gambiae s.s. mosquito populations, according to the WHO protocol [22]. Brand new insecticide impregnated papers were supplied by WHO reference center (Vector Control Research Unit, University Sains Malaysia, Penang, Malaysia). Two-to five-day-old, unfed F 1 female An. funestus s.s.
were exposed for 1 h to discriminating concentrations of the following insecticides: Similarly, An. gambiae s.s. female mosquitoes were exposed to all the insecticides except dieldrin and propoxur, instead fenithrothion (5%) and primiphos-methyl (1.25%) were added.
Furthermore, in order to establish the resistance intensity to pyrethroid insecticides, the An. gambiae s.s. F 1 generations from Elende were tested on permethrin and deltamethrin concentrations of 5x and 10x for 60 mins. Mortality rates were recorded 24h post exposure. The dead mosquitoes were kept in 1.5ml tubes containing silica gel while the survivors were placed in tubes containing RNALater and stored at -80˚C for molecular analyses.
A set of 20-25 mosquitoes exposed to untreated papers were used as control for each test.
The experiment was carried out at ambient temperatures of 25 °C ± 2 °C and 80% ± 10% relative humidity. A mortality rate > 98% of the mosquito populations was considered susceptible to the insecticide, meanwhile suspected resistance was considered at mortality between 90-98%, and resistant where mortality was found to be < 90%.

Piperonyl butoxide (PBO) synergist assay
Only females of An. funestus s.s. were used for this assay since An. gambiae s.s. was inadequate in the study area at the period of collection. In order to determine the possible implication of cytochrome P450s in the observed phenotypic resistance to pyrethroids, two to five days old F 1 female An. funestus s.s. were initially exposed to 4% PBO for 1 h proceeded by immediate second exposure to permethrin (0.75%) and deltamethrin (0.05%) for another 1 h interval. The mortality was determined 24 h post exposure and compared with mortality achieved for mosquitoes subjected to the pyrethroids only.
Differences in mortality among the various groups were analyzed and recorded [22]. replicates of ten unfed mosquitoes were introduced into each plastic cone attached on pieces of fresh, unused bed nets of dimension 30 cm x 30 cm: Olyset®Net (containing 2% permethrin), OlysetPlus®Net (containing 2% permethrin combined with 1% of the synergist PBO), PermaNet®2.0 (containing 1.4-1.8 g/kg±25% deltamethrin), and test, 3 min was the exposure time. After exposure, the mosquitoes were gently and immediately removed from the cones using a mouth aspirator, transferred into paper cups and fed with 10% sucrose soaked in cotton wool. The number of mosquitoes knocked-down was recorded after 1 h while mortality was calculated after 24 h of observation. The experiment was carried out at ambient temperature of 25 °C ± 2 °C and 80% ± 10% relative humidity.

Anopheles funestus s.s
An allele specific PCR was used to genotype and determine the frequency of the L119F-GSTe2 mutation in An. funestus s.s. F 0 mosquito population of Elende as previously described [24,25]. This was to investigate the role of glutathione S-transferases in DDT resistance. The presence of CYP6P9a resistant allele associated with resistance to pyrethroids was genotyped by a PCR-RFLP assay [26] while the A296S-RDL mutation known to be linked with dieldrin resistance was also genotyped by TaqMan assay [27].

Anopheles gambiae s.s.
The L1014F-kdr and L1014S-kdr mutations involved in pyrethroid and DDT resistance in An. gambiae s.s. were genotyped in F 0 Elende mosquitoes using the TaqMan assay [28]. In addition, the G119S ace-1 responsible for carbamate and organophosphate resistance in An. gambiae s.s. was also genotyped in Elende mosquitoes using a TaqMan assay as previously described [29].

Transcription profiling of candidate resistance associated genes
A quantitative Reverse Transcriptase PCR (qRT-PCR) was done to investigate the prominent role of some previously reported Cytochrome P450 detoxification genes (CYP325A, CYP6P5, CYP6P9a and CYP6P9b) in An. funestus s.s. associated to the phenotypic resistance recorded during bioassay [30-32]. Using a triplicate of 10 F 1 females each that recovered after 1h exposure to permethrin from Elende and 3 batches of unexposed 10 F 1 females that were used as control samples; total RNA extraction, cDNA synthesis and qRT-PCR reactions were performed as earlier reported [30]. Fold change and expression of each gene in resistant (R) and control (C) samples were computed according to 2-ΔΔCT method [33] following standardization with housekeeping genes RSP7 ribosomal protein S7 (AFUN007153-RA) and the Actin 5C (AFUN006819) gene.

Species identification
A total of 269 adult resting female Anopheles mosquitoes were collected indoor at Elende over a two-day period. Out of the total sample collected; 230 (85.50%) were members of
The PCR-RFLP genotyping of CYP6P9a revealed that all the mosquitoes were homozygous susceptible with a band size of about 500bp (Fig. 4a) indicating that this mutation is absent in An. funestus s.s. population of Elende.
Also, the 50 samples genotyped for A296S marker were all homozygous resistant RR (100%), revealing that the mutation is fixed in this population (Fig. 4a) in line with the high dieldrin resistance observed.

Transcriptional profiling of candidate genes
A quantitative Reverse Transcriptase PCR (qRT-PCR) was done to examine the role of some previously reported Cytochrome P450 metabolic genes in An. funestus s.s. linked to the resistance observed during bioassay. The qRT-PCR results reveal that CYP6P5, CYP6P9a and CYP6P9b genes known to be involved in pyrethroid resistance are significantly up-regulated in An. funestus s.s. population from Elende as compared to the susceptible laboratory strain FANG (Fig. 5). Both CYP6P9a and CYP6P9b exhibited a 6 fold change in Elende resistant mosquito when compared to FANG (P<0.05) whereas CYP6P5 displayed a 2.20 fold change difference in expression (P<0.05) between the wild mosquitoes and susceptible strain. On the other hand, when comparing permethrin exposed to the unexposed (control) mosquito and FANG strain, CYP325A was not significantly expressed.

Discussion
Rapid scaling up of vector control interventions is ongoing in Cameroon, where malaria is highly endemic. As such, characterization of local vectors alongside investigation of their resistance profile is essential for the effective designing and execution of successful and sustainable vector control interventions as well as for evaluating the impact of insecticides resistance. In the past, the possibility of generating a large number of F1 progeny from small numbers of field collected mosquito for molecular characterization constituted a major hindrance for colonizing An. funestus in the lab. However, this limitation has been resolved by the invention of the forced-egg laying method [14]. An easy approach to addressing this barrier is by collecting indoor resting blood fed female mosquitoes and putting each of them in a confined 1.5 ml tube to forcefully lay eggs. This method has made feasible the substantial evaluation of the susceptibility profile of this mosquito species population against different classes of insecticide. Nevertheless, for experiments involving F1 adults to be informative, it is cardinal that, the offsprings obtained by this technique should not be bias and family isolation effects must be reduced such that the progeny are typical of the overall population. In this regard, pooled egg batches were reared together and the F1 adults were randomly combined in cages for the various assays.
With the possibility of generating a large number of progeny from field collected female An. funestus and An. gambiae mosquitoes, this study therefore characterized the principal malaria vectors in a rural setting within the forested region of Cameroon; located 2 km away from the Yaoundé -Nsimalen International airport.

Mosquito Species composition in Elende
From the 09 species of An. funestus s.l. group described, An. funestus s.s., An. rivulorum and An. vaneedeni were detected in Elende, with abundance of 98.34%, 0.83% and 0.83% respectively. This result is similar to a study conducted in Tibati [34] and Gounougou Regarding the An. gambiae complex, An. gambiae was the only species found. This result is similar to previous studies demonstrating that An. gambiae s.s. was the major species in rural and semi-rural areas of the Centre and Littoral regions in Cameroon, particularly in Yaounde and Douala [41][42][43].

Roles of both vectors in malaria transmission in forested areas
This study confirms the role of An. funestus s.s. and An. gambiae s.s. in malaria transmission in this locality with sporozoite infection rate of 6.66% and 5% respectively. This result is similar to An. funestus s.s. sporozoite infectivity rate in Mibellon (5%) but higher than in Obout (3.2%) and Tibati (2.94%). In Benin (Kpome), whole mosquito extracted DNA of An. funestus s.s. population found high Plasmodium infection during the dry season (18.20% infection rate) [35].
Due to the low number of the field collected An. gambiae s.s. during the study period, infection with Plasmodium (5%) was lower compared to An. funestus s.s. (6.6%). Also, this rate is similar to previous results in Cameroon [10]. Because the location of Elende is close to the Nsimalen-International Airport and to the city of Yaoundé, efforts should be made to reduce the malaria transmission in this locality to avoid it constitute a reservoir for transmission in the city particularly as it was recently shown that mosquitoes can fly over long distances [12].

Multiple and high insecticide resistance in both major vectors constitutes a challenge for vector control in forested areas.
Insecticide resistance profile of An. funestus s.s. in Elende locality is similar to previous studies in Cameroon documented for this species where resistance against all pyrethroids and full susceptibility to organophosphates was observed [11]. The multiple insecticide resistance patterns observed in the An. funestus s.s. population to pyrethroids and DDT in Elende corresponds to the trends observed in Gounougou (2012) [9] and higher than in Obout (2016) [44] but lower than in Tibati (2018) [34]. Moreover, the high resistance pattern of An. funestus s.s. to pyrethroids observed in this locality is similar to  [45], Mozambique [46]; East African region including Uganda [47]; and West Africa in Ghana [48], Benin [35] and Nigeria [49].
The full reversal to susceptibility observed after PBO exposure to permethrin and deltamethrin, implies that cytochrome P450 genes are playing a notable role in the resistance mechanisms. This increasingly higher resistance to pyrethroids poses a remarkable challenge for malaria control programs in Cameroon and necessitates the urgent implementation of insecticide resistance management strategies so as to prevent failure of future programs directed at scaling-up distribution campaigns of pyrethroid impregnated LLINs.
Extremely high levels of resistance to several classes of insecticides, including organochlorine, pyrethroid and carbamate, were also noticed in An. gambiae s.s. population from Elende. Moreover, the intense resistance of An. gambiae s.s. to 5x and 10x concentration of permethrin and deltamethrin each suggests that the resistance is elevated in this population. This elevated resistance in An. gambiae s.s. corresponds with the high level of resistance reported in this species across various sites in Cameroon [6][7][8]34]. Furthermore, the reduced susceptibility observed against the organophosphates

Bio-efficacy of LLINs in cone assays
Freely distributed LLINs by the National Malaria Contol Programme (NMCP) constitute the central malaria vector control intervention in Elende. The dramatic drop in potency of these solely impregnated pyrethroid nets is comparable to cases reported in other localities in Cameroon [11] and Africa [35,46]. Resistance to pyrethroids in this species is linked with a tremendous decline in efficacy to all pyrethroid only LLINs as demonstrated by the diminishing mortality rates against PermaNet 2.0 (<10%) and Olyset net (<25%). In opposite, PBO-based nets demonstrated a greater efficacy with the highest reported by both PermaNet 3.0 top and Olyset plus scoring 100% mortality. This indicates that cytochrome P450 genes are probably propelling pyrethroid resistance in this locality.
The higher mortality rate observed with PBO-based nets suggest that these synergist nets including Olyset Plus and PermaNet 3.0 (roof) should be regarded as a substitution to pyrethroid-only nets in areas of increasing resistance fueled by metabolic mechanisms particularly for cytochrome P450s as it is the situation for An. funestus s.s. [29].

Elevated metabolic resistance in An. funestus differs with high levels of knockdown resistance in An. gambiae
The full susceptibility noticed for pyrethroids in An. funestus s.s. after firstexposure to PBO points out that metabolic resistance mediated by cytochrome P450s is the main mechanism [50]. This is linked to previous studies which confirm the absence of kdr target site sensitivity mutation in this species in Cameroon [9] and across Africa [51].
In the absence of voltage-gated sodium channel knockdown resistance mutations in An. funestus s.s. [51], this study demonstated that pyrethroid resistance in Elende populations of An. funestus s.s. is possibly steered by metabolic resistance machinery. Overall, the role of metabolic resistance is apparent by the marked up-regulation of the three P450 genes (CYP6P5, CYP6P9a, and CYP6P9b ) already reported as essential genes conferring Cross-resistance between DDT and pyrethroids has been demonstrated to be conferred by GSTe2. In relation to this, the frequency of the L119F-GSTe2-resistant allele in Elende field population (48%) is higher than in Mibellon (28%), Tibati (10.2%) and lower than in Gounougou (52%). Similarly, across Africa, the frequency of the DDT resistance marker was closer to that observed in Democratic Republic of Congo [52] and Ghana [48]; higher than the frequency reported in eastern part of Africa, Uganda [36,37] although lower to studies in Benin [35].
The frequency of the 296S-RDL-resistant allele is 100% which is higher as compared in the northern region; particularly in Mibellon (9.7%), Gounougou (14.6%) and Tibati (0.4%). However, this result is similar to mortality rate recorded in An. funestus s.s.
from Obout that exhibited very high level of resistance to dieldrin (4.35% mortality rate) [44]. This high frequency could be as a result of strong resistance selection due to environmental persistence of insecticide residues since its withdrawal from public and agricultural use in Cameroon.
The elevated resistance levels to pyrethroids in An. gambiae s.s. accords with the increased frequency of the 1014F kdr allele (94.9%). This correlates with past studies done in Africa where high pyrethroid resistance in An. gambiae s.l. has been coupled with almost fixed kdr allele in the population, as recently reported in DR Congo [52], or earlier in Côte d'Ivoire [53]. Consistent with previous research performed in other parts of Cameroon [8,34], this study found elevated frequencies of the kdr mutation in An. gambiae s.s. population in Elende that has almost reached fixation.
The very low frequency of the 1014S kdr allele in Elende (1.28%) is in parallel to earlier reports across Cameroon exhibiting that this target site resistance mutation, originally discovered in East Africa, is gradually spreading to Central and West Africa although still at very low occurrence [54].
The presence of the 119S ace-1 mutation in An. gambiae s.s. population is in line with the reduced susceptibility observed in this population to carbamates and organophosphates [55][56][57]. The use of carbamates and organophosphates may be regarded as an alternative for the management of this highly insecticide resistant vector population although the detection of Ace-1 is also a cause of concern.
Overall, similar studies should be conducted across different ecological settings in

Conclusion
This study reports the characterization and resistance profile of the major malaria vectors An. funestus and An. gambiae in Elende locality, situated close to a port of entry in Yaounde, the capital city of Cameroon. The high Plasmodium infection rate alongside the resistance to multiple insecticide classes observed in these vector populations highlights the challenges that public health vector control programs encounter in sustaining the regular effectiveness of contemporary insecticide-based control interventions aimed at reducing malaria transmission in forested areas. More particularly, the resistance observed against the carbamates and possible resistance against the organophosphates constitutes a major concern for IRS; while suggesting the susceptibility testing of Anopheles malaria vectors in this locality to neonicotinoids and pyrrole insecticides in preparation for indoor residual spraying campaigns with novel insecticide ingredients.
Also, this study further provides operational evidence to National Malaria Control Programs for a shift from mass distribution of pyrethroid-only LLINs to second-generation bed nets (containing synergist) in areas where high resistance is driven by metabolic mechanisms notably cytochrome P450s.

Availability of data and materials
All the data generated by this study are included in the manuscript.

This work was funded by a Wellcome Trust Senior Research Fellowship in Biomedical
Sciences to CSW (101893/Z/13/Z) and a DFG grant (BO 2494/3-1) to SB and CSW.