Insecticide resistant Anopheles from Ethiopia but not Burkina Faso show a microbiome composition shift upon insecticide exposure.

doi:10.21203/rs.3.rs-5150012/v1

Download PDF

Research Article

Insecticide resistant Anopheles from Ethiopia but not Burkina Faso show a microbiome composition shift upon insecticide exposure.

https://doi.org/10.21203/rs.3.rs-5150012/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Malaria remains a key contributor to mortality and morbidity across Africa, with the highest burden in children under five. Insecticide based vector control tools, which target the adult Anopheles mosquito are the most efficacious tool in disease prevention. Due to the widespread use of these interventions, insecticide resistance is now ubiquitous across Africa. Understanding the underlying mechanisms contributing to this phenotype is necessary to both track the spread of resistance and to design new tools to overcome resistance. Here, we show that the microbiome composition of insecticide resistant populations of An. gambiae, An. coluzzii and An. arabiensis originating in Burkina Faso from field caught populations and colonies across two countries show no changes in microbial composition compared to their susceptible counterparts. In contrast, An. arabiensis from Ethiopia demonstrates clear differences in microbiome composition in those dying and surviving insecticide exposure. To further understand resistance in this An. arabiensis population, we performed RNAseq and saw differential expression of detoxification genes associated with insecticide resistance and changes to respiration, metabolism and synapse-related ion channels. Taken together, these results indicate that in addition to changes to the transcriptome, the microbiome can contribute to insecticide resistance in certain settings.

Anopheles

malaria

microbiome

insecticide resistance

transcriptomics

Malaria, caused by the Plasmodium parasite and transmitted by Anopheles mosquitoes, remains one of the world’s most deadly diseases, with 249 million cases and 609,000 deaths in 2022 [1]. Despite the gains made in malaria control since the implementation of concerted intervention strategies, the downwards trend plateaued from 2015 and since the recent COVID-19 pandemic there is a clear increase in case numbers [1]. Challenges in reversing this trend and making progress towards elimination have resulted in the World Health Organisation calling for an increase in funding and implementation of new approaches.

Insecticide-based vector control tools are cornerstone of malaria control programs, of these, insecticide treated bed nets are the most effective [2]. These interventions directly target the mosquito vector. In Africa, majority of malaria cases are transmitted by the An. gambiae species complex (An. gambiae, An. coluzzii and An. arabiensis) and An. funestus [3, 4]. There are currently a limited number of insecticide classes available for malaria control, the most important of which are the pyrethroid class, which are used on all insecticide treated bed nets [1]. The selection pressure imposed by pyrethroid insecticides has led to widespread and intense resistance to this chemistry [5]. Indeed, the strength of resistance is such that in some countries, exposure to pyrethroids has no effect on the longevity of the mosquito vector [6]. In response to escalating resistance, next generation bed nets have been developed, recommended, and deployed, containing pyrethroids and a second chemistry [7, 8]. Further, new ways to administrator insecticides are being explored, such as eaves tubes [9] and attractive targeted sugar baits [10] and new chemical classes for indoor residual spraying are now available [11, 12].

Insecticide resistance is a complex phenotype, including increased expression of transcripts involved in metabolic detoxification, single nucleotide polymorphisms reducing the efficacy of the insecticide, reduced penetration through thickening of the cuticle and sequestration of the insecticides [13]. Up-regulation of metabolic enzymes involves the cytochrome p450 class which have been shown to directly metabolise several licensed insecticides, including pyrethroids [14–16]. Numerous P450s have been shown to upregulated in multiple populations across Africa, including CYP6M2, CYP6P3, CYP9K1 and CYP6AA1 in An. gambiae and An. coluzzii, CYP6P4 in An. arabiensis and CYP6P9a/b in An. funestus [14, 17–20]. In addition to cytochrome P450s, GSTs [21], ABC-transporters [22] and most recently UGTs [23] have been implicated in metabolic resistance. Other transcriptomic changes include up-regulation of putative insecticide binders, such as the d7 salivary gland proteins [24] and the chemosensory proteins [25]. Outside of transcriptomic changes, mutations to the target sites of the insecticide are well characterised. In the An. gambiae species complex, mutations in the voltage-gated sodium cannels have been consistently linked with pyrethroid resistance; this includes ‘traditional’ kdr, L995F [26], and ‘new’ kdr V402L-I1527T [27]. A second known mutation in the acetylcholine esterase gene ACE-1 [28] is linked to resistance to both carbamates and organophosphate insecticides.

Although resistance is typically described as a genetic component of the Anopheles vector, recent work has shown that the microbiome changes upon insecticide exposure, hinting at a role of these organisms in resistance. Indeed, work on the South and Central American vector, An. albimanus showed significant changes in the cuticle microbiome after pyrethroid exposure and changes on the overall microbiota after organophosphate exposure [29]. In Africa, pyrethroid resistance has been linked to changes in microbiome composition in Cameroon, Kenya and Côte D’Ivoire, whilst laboratory-adapted An. coluzzii have been shown to have a changed microbiome after selection to pyrethroid resistance [30–34]. Furthermore, experimental treatment with antibiotics and spiking with known bacteria show increased pyrethroid tolerance in a laboratory colonised An. arabiensis [35]. Taken together, these data indicate that insecticide resistance may be linked to the mosquito microbiome and that this facet of insecticide resistance could be manipulated for vector control.

In this study we use 16S sequencing to determine whether the microbiome of lab-colonised mosquitoes and field caught mosquitoes from West and East Africa differs between mosquitoes surviving and dying after exposure to the pyrethroid deltamethrin. The lab colonies originated from Burkina Faso and were reared at two separate European laboratories (An. coluzzii, Germany and An. arabiensis, UK), mosquitoes from Burkina Faso represent all three species of the Anopheles gambiae species complex, whilst those from Ethiopia were An. arabiensis. Here, we show that An. arabiensis from Ethiopia have distinct microbiome profiles in surviving and dead mosquitoes, whilst those from Burkina Faso and the two laboratories do not. To further understand insecticide resistance in this An. arabiensis population we performed RNAseq and see up-regulation of insecticide-resistance related genes and overall changes to respiration. Integration of this data with a second population from Ethiopia highlights key detoxification genes in this region. Further comparisons with An. arabiensis from Brukina Faso show large difference in transcriptomic profile but nevertheless highlights four core detoxification genes overexpressed across Africa.

Species identification of field samples

Species identification was performed on a total of 441 individual mosquitoes collected in Tiefora, Burkina Faso and Bahirdar, Ethiopia, consisting of 248 and 107 samples respectively. The Burkina Faso collections consisted of 134 (54%) An. arabiensis, 21 (8.5%) An. coluzzii, 92 (38.1%) An. gambiae and 1 (0.4%) An. gambiae x An. coluzzii hybrid. In contrast, the Bahirdar samples were all An. arabiensis (107 samples). Table 1 shows the number of control unexposed mosquitoes and those alive and dead following pyrethroid exposure.

Table 1

**Breakdown of the mosquito samples by species, mortality status and country.** Table 1 shows the country and species for control unexposed mosquitoes, alive, dead and total and those that were exposed to 5XDD Deltamethrin through WHO tube bioassay, alive, dead and total. The final column shows total number of mosquitoes tested for each row. The final row shows the total across both countries.
Country/Species	Control		Control Total	Exposed		Exposed Total	Total
Country/Species	Alive	Dead	Control Total	Alive	Dead	Exposed Total
Burkina Faso	58		58	88	102	190	248
An. arabiensis	48		48	18	68	86	134
An. coluzzii	4		4	12	5	17	21
An. gambiae	6		6	57	29	86	92
Hybrid (An. gambiae x An. coluzzii)				1		1	1
Ethiopia	32	2	34	15	58	73	107
An. arabiensis	32	2	34	15	58	73	107
Grand Total	90	2	92	103	160	263	355

Resistance levels

Resistance to the pyrethroid deltamethrin (Fig. 1A) was highest in field-caught An. gambiae with 66.3% of the mosquitoes surviving 5XDD exposure, in contrast to 20.9% survivorship of An. arabiensis. An. coluzzii from Burkina Faso also showed similarly high survivorship (70.6%) but with only 17 samples. An. arabiensis samples from Ethiopia showed similar resistance levels to An. arabiensis in Burkina Faso with 20.5% surviving 5XDD. The lab colonies were exposed to 1XDD of deltamethrin, with 41% An. arabiensis surviving and 67.3% of An. coluzzii. Each field-caught individual used for subsequent microbiome work was then assessed for the presence of kdr-L995F and ACE-1 mutations (Fig. 1B); kdr-L995F was present in 42.6% of the An. arabiensis from Ethiopia (29.4% in dead and 33.3% in alive) and ACE-1 at 6.4% (3% in dead and 10% alive). In Burkina Faso, kdr-L995F was present at 54.4% in An. arabiensis (66.7% alive and 36.7% dead), and was significantly associated with survival (p_χ2 = 0.0379); An. coluzzii at 60% (75% alive and 30% dead, p_χ2 = 0.045) and An. gambiae at 62.5% frequency (63.3% alive and 61.9% dead, p_χ2 = 0.88). L995S was not tested here, although is likely present in An. arabiensis from Burkina Faso [36]. ACE-1 was at low frequency across all Burkina samples, with 4.4% in An. arabiensis, 3.3% in An. coluzzii and 19.5% in An. gambiae.

Microbiome diversity

16S sequencing was then performed on alive and dead mosquitoes from pools of each species from each location. No differences in alpha diversity were observed, indicating no difference in operational taxonomic unit (OTU) richness; however, beta diversity was significantly different between the countries and the interaction term of countries and alive/dead status (P_ANOVA = 0.001 and 0.008 respectively) signifying diversity between these factors. No differences were observed for beta dispersion across the countries, demonstrating similar variances between groups. A Bray-Curtis multidimensional scaling plot (Fig. 2A) shows that the samples largely separate on location, with some notable exceptions. Several samples from the An. arabiensis colony and one sample from the An. coluzzii colony overlap with the Burkina Faso samples, whilst one An. arabiensis pool from Burkina Faso overlaps with those from the An. arabiensis colony; this may indicate some conservation of original microbiome stochastically in individuals across many generations and across multiple localities. There is no clear distinction between the samples from Burkina Faso based on species; this lack of separation might be expected, as the individuals, despite originating from different larval environments, were later pooled. Indeed, sharing a niche during the aquatic stage has been shown to significantly influence microbial composition [37]. Interestingly, whilst the samples from Burkina Faso and the colonies show no separation of those mosquitoes surviving or dying after insecticide exposure, there is clear separation of the Ethiopian samples indicating that microbiome composition is linked to insecticide resistance in these An. arabiensis samples. These samples were collected from the same collection site, potentially explaining the clearer assoaciation.

The abundance plots (Fig. 2B) show 2–6 highly abundant genera across each population, An. arabiensis from Ethiopia are dominated by Erwiniaceae, which align to an unspecified taxon, Salmonella and Pantoea. Six OTUs are above 10% abundance in the colonised An. coluzzii including Asaia, Salmonella, an unspecified Erwiniaceae, Swaminathania, Elizabethkingia anophelis and Enterobacter amnigena whilst An. coluzzii from Burkina Faso are dominated by Swaminathania, with Asaia and Salmonella having > 5% abundance. Similarly, An. gambiae from Burkina have these three OTUs over 5% abundance but uniquely have Tanticharoenia. An. arabiensis from Burkina Faso again shows similar dominant taxon, overlapping entirely with the other species. The An. arabiensis colony is dominated by two taxa: Elizabethkingia anophelis and an unspecified Swaminathania.

Bacteria associated with survival to pyrethroid exposure

To explore any association with insecticide resistance, the pools were split by country and association with survival assessed for OTUs representing > 1% of the overall abundance. Field caught samples from Burkina Faso had one genus associated with survival, Sphingomonas, which was at significantly higher abundance in survivors (p_ANOVA = 0.029). Similarly, the An. arabiensis colony had a low abundant unspecified gammaproteobacteria which was at higher abundance in dead mosquitoes (p_ANOVA = 0.03). Unlike these tenuous associations, the field caught An. arabiensis from Ethiopia had clear associations with bacterial genera showing significant association with survivorship. As with Burkina Faso, Sphingomonas was significantly associated with exposure (p_ANOVA = 0.045); however, it was found at higher abundance in dead mosquitoes. Perhaps the most interesting is a highly significant association of Pantoea with survival (p_ANOVA = 0.0067) at > 12% abundance in alive mosquitoes compared to ~ 2% in dead mosquitoes. Similarly, the unspecified Erwiniaceae is at ~ 66% abundance in alive mosquitoes and 12% in dead mosquitoes (p_ANOVA = 0.0082), whilst Salmonella is at ~ 14% in alive mosquitoes and 47% in dead mosquitoes, although this is not significant due to high variation (p_ANOVA = 0.11) (Fig. 3). The other significant genera at significant higher abundance in survivors include Escherichia and Kosakonia, whilst Rhizobium and Methylbacterium-Methylorbrum are at higher abundance in dead mosquitoes; however, the abundance of these bacteria is on average < 1%.

Characterisation of insecticide resistance in An. arabiensis from Ethiopia

To determine the mechanisms of genetically-driven pyrethroid resistance in the field caught An. arabiensis from Ethiopia, RNAseq was performed. PCA demonstrated separation of field mosquitoes from Bahirdar and the lab-susceptible Moz (Supplementary Fig. 1). In total, 2666 genes (26.3%) were significantly overexpressed and 2958 (29.1%) were significantly down-regulated (Fig. 4A, Supplementary Table 1). GO terms show significant enrichment numerous terms, including cellular respiration (p = 6.07e-7), ATP synthase (p = 6.15e-3), oxidoreduction activity (p = 2.09e-5), ion transport (p = 1.31e-9) and synapse (p = 1.21e-3). Four MetaCyc pathways showed significant enrichment, including lipoprotein post-translational modification (p = 1.25e-5), aerobic respiration (p = 4.07e-3), Fe(II) oxidation (p = 6e-3) and plasmalogen biosynthesis (p = 1.47e-2) (Supplementary Table 2). Of the detoxification genes, 23 ABC transporters, 33 cytochrome P450s, 10 GSTs, 5 COEs and 2 UGTs are overexpressed (Figs. 4B-C). These genes include CYP9K1, CYP6P4, CYP6AA1, CYP6Z3 and CYP6M2, which are all known pyrethroid metabolisers in the An. gambiae complex [18, 19, 38, 39] and CYP4G16 linked to cuticular thickening [40]. Additionally, the syntenic ortholog of ABCH2 [41], 3 CSPs [25], 3 alpha-crystallins, 3 hexamerins [14] and 24 cuticular genes are overexpressed (Supplementary Fig. 2). AARA016988 is the homolog of the hexamerin AGAP001345 [14] and is the second most highly significantly over-expressed gene at 175x, whilst AARA016998, the homolog of SAP2 [25], is the fourth with a fold change of 18.45. ABCA7 and ABCG2 homologs are present in the top 20 significantly overexpressed genes, whilst 5 of the top 20 are serine proteases.

GO term enrichments of significantly down regulated genes include nucleic acid binding (p = 2.19e-17), gene expression (p = 8.55e-31), cellular response to stress (p = 5.17e-13) and ribosome biogenesis (p = 8.07e-4). The KEGG pathway related to biotin metabolism (p = 1.88e-3) is also significantly enriched (Supplementary Table 2). Taken together, these indicate large metabolic changes between these populations.

Comparison of An. arabiensis from Ethiopia and Burkina Faso

A prior dataset of permethrin resistant An. arabiensis from Asendabo, approximately 425km further south [42] was compared with the results from Bahirdar. 1443 genes were significant across both sites, 607 are consistently up-regulated and 412 consistently down regulated (Supplementary Table 3). Consistently up-regulated genes include CYP6P4, CYP6M2, CYP6P3, CYP9J5 and CYP9K1, all previously implicated in pyrethroid resistance [38, 39, 43, 44]. Furthermore, GSTE2, GSTE7, GSTD3 a chemosensory protein homolog and two ABCG transcripts are included in this list. Enrichments include oxidoreduction-driven active transmembrane transporter activity (p = 2.09e-17), oxidoreductase activity (p = 1.61e-12), electron transfer activity (p = 3.82e-16), cellular respiration (p = 1.82-20) and mitochondrion (p = 2.89e-11). Consistently down regulated genes include response to stress (p = 1.61e-2).

Similarly, a previously published data is available for An. arabiensis from Gaoua, located in southwest Burkina Faso [36] and this was compared to the Ethiopian samples (Supplementary Table 4). A pairwise comparison with the samples from Bahirdar shows 2762 genes significant across both sites, of these 209 are commonly up-regulated and 176 are commonly down regulated. Consistently up-regulated genes include CYP6AA2, CYP6AG2, CYP6AK1, CYP6P1 and CYP6P4. Of the other detoxification family members, ABCA2, GSTE7, GSTS1, GSTMS3 and a UGT (AARA006222). Enriched GO terms relate to transporter activity (p = 4.95e-2).

A three-way comparison shows 726 genes commonly differentially expressed, of which 55 are commonly up-regulated and 22 genes that are commonly down-regulated (Supplementary Table 5). GSTE7, CYP6AK1, CYP6P4, a UGT (AARA006222), one COE (AARA016325) and multiple serine protease-related proteins are commonly over expressed. The consistent overexpression in these vastly different populations indicates a key role in insecticide resistance in these populations.

Resistance to pyrethroid insecticides is a complex phenotype and whilst metabolic and target site changes are relatively well understood, changes in the microbiome are just now beginning to be explored. In this study we show that microbiome changes may be more important in specific populations, we further characterise resistance in An. arabiensis from Bahirdar and show changes to known resistance-associated transcripts as well as large changes to respiration-related genes, which are also evident in a prior RNAseq study on an An. arabiensis population from Ethiopia. Taken together, these results indicate that these mosquitoes potentially rely on a mix of genetic and non-genetic basis of insecticide resistance.

Bioassay results here show that Ethiopian An. arabiensis have similar levels of pyrethroid resistance to the Burkinabe counterparts but significantly lower that An. gambiae or An. coluzzii. These findings are in line with published literature showing that An. arabiensis generally has lower levels of resistance [45, 46], likely due to a lower propensity for anthrophilic behaviour. Kdr-West (L995F) was present at high frequency in both countries, with higher levels in Burkina Faso; however, kdr-East (L995S) was not assessed here and so correspondingly high levels of this genotype in Bahirdar cannot be ruled out. Data collected in 2023, indicates a frequency of the L995S is 5% in An. arabiensis collected from the Tiefora site, Burkina Faso (Sanou, unpublished). Colony mosquitoes from both the UK and Germany are maintained under regular selection pressure [47], and thus high levels of resistance are expected. Interestingly, An. arabiensis from Burkina Faso showed a significant association of L995F and survivorship which was lacking in the samples from Ethiopia, in line with previous data [48].

The microbial composition observed in the populations of this study aligns well with previous publications, featuring symbionts commonly found in Anopheles mosquitoes such as Pantoea, Elizabethkingia anophelis, Asaia and Serratia commonly seen in Anopheles mosquitoes [30–34, 49]. Several bacterial genera are found across all populations here, likely due to previous reports of selectivity in bacterial colonisation of the adult gut and natural occurrence of soil bacteria [37, 49]. Furthermore, previous reports have demonstrated that mosquitoes from similar habitats share a portion of their microbiome [49, 50]. The populations described in this study are dominated by one or two OTUs, with many having < 1% abundance, in agreement of a deep-sequencing study across multiple mosquito species in Kenya [50]. Surprisingly, several pools of mosquitoes from the UK and Germany overlap with Burkina Faso samples, indicative of a stochastic maintenance of field-like microbiome; however, this requires further investigation as it contradicts the expectation that microbiome is largely determined by larval environment [51].

Here we show that An. arabiensis from Ethiopia have a significant association of pyrethroid survivorship with microbiome composition. Erwiniaceae, Pantoea, Kosakonia, and Escherichia are at increased abundance in mosquitoes surviving pyrethroid exposure. These OTUs are thus potentially involved in conferring some level of pyrethroid resistance in An. arabiensis from Ethiopia. Unlike previous studies, we found no association of Serratia or Asaia with pyrethroid resistance in this population [32, 34] which could potentially be due to species or locality differences. Pantoea is a known insecticide degrading bacteria [52] and was shown to be at significantly higher levels in those mosquitoes surviving exposure, as seen previously [29]. Similarly Escherichia has previously been linked to fenitrothion [30] and deltamethrin [32] resistance and has been shown to naturally metabolise carbamate insecticides [53]. Kosakonia hasn’t previously been linked to insecticide resistance in pests but is associated with rice paddy fields and organophosphate remediation [54] and has been shown to inhibit Trypanosome infection in Tsetse flies [55].

Mosquitoes from Bahir Dar have previously been shown to be resistant to pyrethroid insecticides [48] but the underlying mechanisms remained unstudied. Here, we show that the mechanisms of insecticide resistance are consistent with other study sites in Ethiopia and broadly across Africa. For example, the most highly overexpressed genes include a hexamerin [14] and a SAP2 homolog [25], indicating that these families may be important in An. arabiensis populations in addition to An. coluzzii. Furthermore, overexpression of key pyrethroid-metabolising cytochrome p450s such as CYP6P4 [19], CYP9K1 [38] and CYP6P3 [43] have previously been demonstrated in resistant An. arabiensis species. Interestingly, ABCH2, which has previously been shown to reduce uptake of pyrethroid insecticides in An. coluzzii [41] is also overexpressed in this population. The importance of the over-expression of these candidates is underlined through integration of RNAseq data available for An. arabiensis from a second site in Ethiopia [42], where similar patterns of gene expression are seen. In contrast, very few genes are commonly over-expressed with the Burkina Faso An. arabiensis [36] suggesting local adaptation to insecticide pressure. Just 55 genes are commonly over-expressed in these populations, which may be unsurprising given the extreme geographical distance and thus likely lack of inbreeding. Nevertheless, key detoxification genes such as CYP6P4, CYP6AK1, GSTE7 and a carboxylesterase may display convergent evolution of over-expression, indicating a key role in pyrethroid resistance in An. arabiensis. Temporal separation of these samples (2021 here, 2017 in the second Ethiopia study and 2018 in Burkina Faso) may confound this data as insecticide resistance is likely a continuously evolving trait.

Recent work has linked increased respiration with insecticide resistance [33], which is consistent with enriched ontology terms both within the Bahir Dar RNAseq produced here and in the integrated Ethiopian data; this again suggests that resistance results in higher respiratory rate either causally or as a result of this phenotype. If these changes result in differences in underlying reactive oxygen and nitrogen species, as previously shown [56], this could impact both vector competence and microbial colonisation.

The data presented here demonstrates that the microbiome may play a role in insecticide resistance in certain settings. We further demonstrate that An. arabiensis from within Ethiopia show similar transcriptomic changes resulting in insecticide resistance; however, the number of genes consistent with An. arabiensis from Burkina Faso are few. Several caveats remain with the conclusions drawn here and other studies. Firstly, a causal relationship between pyrethroid survival and the microbiome must be shown utilising metabolism studies or microbiome transplant. Further, validation of key candidates across multiple populations from vastly different geographies should be demonstrated. Nevertheless, taken together, our data demonstrates both site-specificity and cross-country commonalities in resistance, underlining the necessity to test new insecticide products across multiple localities.

Field Collections

Mosquito collections were carried out in both Ethiopia and Burkina Faso respectively using larval sampling. Mosquito larvae were collected from 7th July to 2nd August 2021 in Tiefora in Burkina Faso (Long. :10.62411667, lat.: -4.55335) characterised by high use of insecticides in agriculture and from 28th July to 6th in Bahirdar (Long.: 11.588790, lat.: 37.3888119) in Ethiopia.

Mosquito Rearing

Mosquitoes were reared under standard insectary conditions of 28^oC ± 2^oc with a 12:12 hour light:dark cycle with one hour dawn:dusk. Larvae were fed ground fish food and upon emergence transferred to cages and maintained on 10% sucrose solution. Banfora (An. coluzzii) and Gaoua (An. arabiensis) were reared in insectaries at Heidelberg University Hospital (UKHD) and Liverpool School of Tropical Medicine (LSTM) respectively; each strain was originally colonised from Burkina Faso [57]. The field caught mosquitoes were reared in insectaries at University of Gondar (An. arabiensis) and Centre National de Recherche et de Formation sur le Paludisme (CNRFP) (An. gambiae, An. coluzzii and An. arabiensis).

Bioassays

Mosquitoes were reared and at 3–5 days old, presumed mated, exposed to 1X, 5X and 10X diagnostic dose of 0.05% Deltamethrin for one hour using standard WHO tube tests [58] to determine the dose that gave ~ 70% mortality at 24 hours. After exposure to appropriate dose (5XDD in field samples, 1XDD at LSTM and UKHD) mortality was recorded 24 hours later. Mosquitoes were then individually stored at -20^oc and their live and dead phenotype recorded. Mosquitoes from University of Gondar were shipped to CNRFP for further processing to ensure consistency of field samples. Extractions were done separately at CNRFP, UKHD and LSTM following the below protocol.

Sample processing

Mosquitoes were individually thawed and 250µL of 70% alcohol added followed by vortexing for 10 seconds to remove the surface microbiome. 250µL of sterile distilled water was then added and the samples were vortexed for a further 10 seconds. The samples were then gently rinsed and allowed to dry. Once dry, the head and thorax were separated from the abdomen and each stored in separate tubes for further processing.

Species identification and molecular analysis

DNA was extracted from the heads and thorax by boiling in STE buffer for 15 minutes at 95^oc. The subsequent DNA from individual mosquitoes underwent species identification through SINE200 PCR following a previously published protocol [59]. To determine the frequency of kdr-L995F [60] and ACE-1 [61] additional PCRs were performed following standard protocols.

DNA extraction for 16S sequencing

Mosquitoes abdomens were then pooled into groups of 5 by species, alive:dead phenotype and location. DNA was then extracted from individual mosquitoes using the LIVAK protocol as previously published [62]. Briefly, LIVAK buffer pH7.5 (1.6ml 5M NaCl, 1.7ml 0.5M EDTA, 2.4ml 1M Tris pH 7.5, 5ml 10% SDS and up to 100ml ddH₂O) was heated to 65^oC, mosquitoes were then homogenised in 100µL LIVAK buffer and heated for 30 minutes at 65^oC. 14µL of 8M potassium acetate was added, mixed, and left on ice for 30 minutes. After centrifugation, the supernatant was collected and 200µL of 99.9% Ethanol added. DNA was then precipitated and resuspended in 50µL of ddH₂O. DNA was then shipped to BIOMES GmbH for 16S sequencing, sample information can be found in Supplementary Table 6.

16S Analysis

16S abundance tables were provided by BIOMES GmbH and analysed using the vegan package [63] in R following previously published methodology [33]. Briefly, alpha diversity was analysed using an ANOVA on the Shannon index followed by Tukey multiple comparisons for country, species and alive/dead status. Beta diversity was calculated using Brays Curtis to account for OTU abundance, following by a permuted ANOVA (n = 1000) using country and alive/dead groupings. Beta dispersion was explored using a permutation test (n = 1000) on nMDS for country pools.

RNA extraction

Total RNA was extracted from field mosquitoes from Ethiopia (ETH), as well as from susceptible An. arabiensis maintained at UKHD (Moz) [57]. RNA was extracted using a PicoPure RNA extraction kit following manufacturer’s instructions. Total RNA quantity and quality was checked on a NanoDrop and TapeStation respectively before being submitted to Eurofins Genomics for polyA enrichment-based RNA sequencing.

RNAseq analysis

The fastq files were aligned to An. arabiensis DONGALA assembly using Hisat2 [64] with default parameters. featureCounts [65] was then used to extract read counts for each gene. Count files were then analysed using the DESeq2 package [66] in R as previously described [33]. GO and KEGG enrichments were performed within VectorBase [67] with p-values taken with Bonferroni correction. Fastq files for RNAseq of An. arabiensis from Burkina Faso were retrieved from SRA (PRJNA780362). Families were extracted from VectorBase using PFAM IDs as follows: PF00067, PF00201, PF00135, PF00005, PF12848, PF03392, PF00011, PF03722, PF00372 and PF01395.

Contributions

NWM and AS carried out field collections and bioassay exposures and AS carried out the molecular work for all field collected samples. OS and MT were involved in field and molecular work in Burkina Faso. LW participated in field collection and bioassay exposures in Ethiopia. MM and NV performed bioassays and DNA extractions at LSTM and UKHD respectively. NWM, AS, MM, FC and NV performed all mosquito rearing for DNA work. JH reared mosquitoes and carried out the RNA extractions and subsequent RNA sequencing analysis. VAI analysed the RNA sequencing and 16S results and drafted the manuscript. VAI, NWM, AS, MG and TBA conceived the study. All authors reviewed the manuscript.

Acknowledgements

We thank Prof Hilary Ranson for providing mosquito extractions from the UK. We also thank Liverpool Insect Testing Establishment (part of iiDiagnostics Ltd.) and LSTM for providing the mosquito strains to VAI. We would additionally like to thank Dr Kai Stieger at BIOMES for discussions on best extraction practices across multiple sites.

Funding Declaration

This study was funded through an ERC Starting Grant (101075634, ReMVeC) and the Deutsches Zentrum für Infektionsforschung (DZIF, TTU03.705) awarded to VAI. Activities in Burkina Faso were supported by Wellcome Trust International Training fellowship (222019/Z/20/Z) awarded to AS.

Data Availability

Data is provided within the manuscript or supplementary information files.

Organization, W.H., World Malaria Report. 2022, WHO.
Bhatt, S., et al., The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015. Nature, 2015. 526: p. 207–211.
Sinka, M.E., et al., The dominant Anopheles vectors of human malaria in Africa, Europe and the Middle East: occurrence data, distribution maps and bionomic précis. 2010. p. 117.
Coetzee, M., et al., Anopheles coluzzii and Anopheles amharicus, new members of the Anopheles gambiae complex. 2013. p. 246–274.
Ranson, H., et al., Insecticide Resistance in African Anopheles Mosquitoes: A Worsening Situation that Needs Urgent Action to Maintain Malaria Control. 2016, Elsevier. p. 187–196.
Hughes, A., et al., Anopheles gambiae populations from Burkina Faso show minimal delayed mortality after exposure to insecticide-treated nets. 2020. p. 17.
Messenger, L.A., et al., Effects of next-generation, dual-active-ingredient, long-lasting insecticidal net deployment on insecticide resistance in malaria vectors in Tanzania: an analysis of a 3-year, cluster-randomised controlled trial. The Lancet Planetary Health, 2023. 7(8): p. e673-e683.
HIEN, S.A., et al., Evidence supporting deployment of next generation insecticide treated nets in Burkina Faso: bioassays with either chlorfenapyr or piperonyl butoxide increase mortality of pyrethroid-resistant Anopheles gambiae s.l. 2021, Research Square.
Knols, B.G.J., et al., Eave tubes for malaria control in Africa: an introduction. Malaria journal, 2016. 15: p. 1–7.
Fiorenzano, J.M., P.G. Koehler, and R.-D. Xue, Attractive toxic sugar bait (ATSB) for control of mosquitoes and its impact on non-target organisms: a review. International journal of environmental research and public health, 2017. 14(4): p. 398.
Ngufor, C., et al., Indoor residual spraying with a mixture of clothianidin (a neonicotinoid insecticide) and deltamethrin provides improved control and long residual activity against pyrethroid resistant Anopheles gambiae sl in Southern Benin. 2017, Public Library of Science. p. e0189575.
Fongnikin, A., et al., Efficacy of Fludora® Fusion (a mixture of deltamethrin and clothianidin) for indoor residual spraying against pyrethroid-resistant malaria vectors: laboratory and experimental hut evaluation. Parasites & vectors, 2020. 13: p. 1–11.
Ingham, V.A., L. Grigoraki, and H. Ranson, Pyrethroid resistance mechanisms in the major malaria vector species complex. Entomologia Generalis, 2023. 43(3): p. 515–526.
Ingham, V.A., S. Wagstaff, and H. Ranson, Transcriptomic meta-signatures identified in Anopheles gambiae populations reveal previously undetected insecticide resistance mechanisms. Nature Communications, 2018. 9: p. 5282.
Yunta, C., et al., Pyriproxyfen is metabolized by P450s associated with pyrethroid resistance in An. gambiae. 2016, Elsevier. p. 50–57.
Yunta, C., et al., Cross-resistance profiles of malaria mosquito P450s associated with pyrethroid resistance against WHO insecticides. 2019.
Ingham, V. and S. Nagi, Genomic profiling of insecticide resistance in malaria vectors: Insights into molecular mechanisms. Research Square, 2024.
Njoroge, H., et al., Identification of a rapidly-spreading triple mutant for high‐level metabolic insecticide resistance in Anopheles gambiae provides a real‐time molecular diagnostic for antimalarial intervention deployment. Molecular ecology, 2022. 31(16): p. 4307–4318.
Ibrahim, S.S., et al., The cytochrome P450 CYP6P4 is responsible for the high pyrethroid resistance in knockdown resistance-free Anopheles arabiensis. 2015.
Mugenzi, L.M.J., et al., Cis-regulatory CYP6P9b P450 variants associated with loss of insecticide-treated bed net efficacy against Anopheles funestus. 2019, Nature Publishing Group. p. 1–11.
Aravindan, V., S. Muthukumaravel, and K. Gunasekaran, Interaction affinity of Delta and Epsilon class glutathione-s-transferases (GSTs) to bind with DDT for detoxification and conferring resistance in Anopheles gambiae, a malaria vector. Journal of vector borne diseases, 2014. 51(1): p. 8.
Pignatelli, P., et al., The Anopheles gambiae ATP-binding cassette transporter family: phylogenetic analysis and tissue localization provide clues on function and role in insecticide resistance. 2018. p. 110–122.
Logan, R.A.E., et al., Uridine diphosphate (UDP)-glycosyltransferases (UGTs) are associated with insecticide resistance in the major malaria vectors Anopheles gambiae s.l. and Anopheles funestus. Scientific Reports, 2024. 14(1): p. 19821.
Isaacs, A.T., et al., Genome-wide transcriptional analyses in Anopheles mosquitoes reveal an unexpected association between salivary gland gene expression and insecticide resistance. 2018. p. 225.
Ingham, V.A., et al., A sensory appendage protein protects malaria vectors from pyrethroids. Nature, 2019. 577: p. 376–380.
Martinez-Torres, D., et al., Molecular characterization of pyrethroid knockdown resistance (kdr) in the major malaria vector Anopheles gambiae s.s. 1998. p. 179–184.
Clarkson, C.S., et al., The genetic architecture of target-site resistance to pyrethroid insecticides in the African malaria vectors Anopheles gambiae and Anopheles coluzzii. 2021, John Wiley & Sons, Ltd. p. 5303–5317.
Weill, M., et al., The unique mutation in ace-1 giving high insecticide resistance is easily detectable in mosquito vectors. 2004, Blackwell Publishing Ltd. p. 1–7.
Dada, N., et al., Pyrethroid exposure alters internal and cuticle surface bacterial communities in Anopheles albimanus. 2019. p. 2447–2464.
Dada, N., et al., Whole metagenome sequencing reveals links between mosquito microbiota and insecticide resistance in malaria vectors. 2018. p. 2084.
Omoke, D., et al., Western Kenyan Anopheles gambiae showing intense permethrin resistance harbour distinct microbiota. Malaria Journal, 2021. 20(1): p. 1–14.
Pelloquin, B., et al., Overabundance of Asaia and Serratia bacteria is associated with deltamethrin insecticide susceptibility in Anopheles coluzzii from Agboville, Côte d’Ivoire. 2021. p. 2021.03.26.437219.
Ingham, V.A., et al., Integration of whole genome sequencing and transcriptomics reveals a complex picture of the reestablishment of insecticide resistance in the major malaria vector Anopheles coluzzii. PLOS Genetics, 2021. 17: p. e1009970.
Djondji Kamga, F.M., et al. Contrasting Patterns of Asaia Association with Pyrethroid Resistance Escalation between the Malaria Vectors Anopheles funestus and Anopheles gambiae. Microorganisms, 2023. 11, DOI: 10.3390/microorganisms11030644.
Barnard, K., et al., The contribution of gut bacteria to insecticide resistance and the life histories of the major malaria vector Anopheles arabiensis (Diptera: Culicidae). 2019.
Williams, J., et al., Sympatric populations of the Anopheles gambiae complex in Southwest Burkina Faso evolve multiple diverse resistance mechanisms in response to intense selection pressure with pyrethroids. Insects, 2022. 13(3): p. 247.
Gimonneau, G., et al., Composition of Anopheles coluzzii and Anopheles gambiae microbiota from larval to adult stages. Infection, Genetics and Evolution, 2014. 28: p. 715–724.
Vontas, J., et al., Rapid selection of a pyrethroid metabolic enzyme CYP9K1 by operational malaria control activities. Proceedings of the National Academy of Sciences, 2018. 115(18): p. 4619–4624.
Yunta, C., et al., Cross-resistance profiles of malaria mosquito P450s associated with pyrethroid resistance against WHO insecticides. Pesticide biochemistry and physiology, 2019. 161: p. 61–67.
Balabanidou, V., et al., Cytochrome P450 associated with insecticide resistance catalyzes cuticular hydrocarbon production in Anopheles gambiae. Proceedings of the National Academy of Sciences, 2016. 113(33): p. 9268–9273.
Kefi, M., et al., ABCH2 transporter mediates deltamethrin uptake and toxicity in the malaria vector Anopheles coluzzii. PLOS Pathogens, 2023. 19(8): p. e1011226.
Messenger, L.A., et al., A whole transcriptomic approach provides novel insights into the molecular basis of organophosphate and pyrethroid resistance in Anopheles arabiensis from Ethiopia. Insect biochemistry and molecular biology, 2021. 139: p. 103655.
Müller, P., et al., Field-Caught Permethrin-Resistant Anopheles gambiae Overexpress CYP6P3, a P450 That Metabolises Pyrethroids. 2008, Public Library of Science. p. e1000286.
Mitchell, S.N., et al., Identification and Validation of a Gene Causing Cross-Resistance Between Insecticide Classes in Anopheles gambiae From Ghana. 2012. p. 6147–6152.
Mawejje, H.D., et al., Characterizing pyrethroid resistance and mechanisms in Anopheles gambiae (s.s.) and Anopheles arabiensis from 11 districts in Uganda. Current Research in Parasitology & Vector-Borne Diseases, 2023. 3: p. 100106.
Nash, R.K., et al., Systematic review of the entomological impact of insecticide-treated nets evaluated using experimental hut trials in Africa. Current Research in Parasitology & Vector-Borne Diseases, 2021. 1: p. 100047.
Williams, J., et al., Characterisation of Anopheles strains used for laboratory screening of new vector control products. Parasites & vectors, 2019. 12(1): p. 1–14.
Messenger, L.A., et al., Insecticide resistance in Anopheles arabiensis from Ethiopia (2012–2016): a nationwide study for insecticide resistance monitoring. Malaria Journal, 2017. 16(1): p. 469.
Guégan, M., et al., The mosquito holobiont: fresh insight into mosquito-microbiota interactions. Microbiome, 2018. 6(1): p. 49.
Osei-Poku, J., et al., Deep sequencing reveals extensive variation in the gut microbiota of wild mosquitoes from Kenya. Molecular Ecology, 2012. 21(20): p. 5138–5150.
Brettell, L.E., et al., Mosquitoes reared in distinct insectaries within an institution in close spatial proximity possess significantly divergent microbiomes. bioRxiv, 2024: p. 2024.08.28.610121.
Ramya, S.L., et al., Degradation of acephate by Enterobacter asburiae, Bacillus cereus and Pantoea agglomerans isolated from diamondback moth Plutella xylostella (L), a pest of cruciferous crops. 2016, Triveni Enterprises. p. 611.
Kulkarni, A.G. and B.B. Kaliwal, Bioremediation of Methomyl by Escherichia coli, in Toxicity and Biodegradation Testing, E.D. Bidoia and R.N. Montagnolli, Editors. 2018, Springer New York: New York, NY. p. 75–86.
Dash, D.M. and W.J. Osborne, Rapid biodegradation and biofilm-mediated bioremoval of organophosphorus pesticides using an indigenous Kosakonia oryzae strain-VITPSCQ3 in a Vertical-flow Packed Bed Biofilm Bioreactor. Ecotoxicology and Environmental Safety, 2020. 192: p. 110290.
Weiss, B.L., et al., Colonization of the tsetse fly midgut with commensal Kosakonia cowanii Zambiae inhibits trypanosome infection establishment. PLOS Pathogens, 2019. 15(2): p. e1007470.
Oliver, S.V. and B.D. Brooke, The role of oxidative stress in the longevity and insecticide resistance phenotype of the major malaria vectors Anopheles arabiensis and Anopheles funestus. PloS one, 2016. 11(3): p. e0151049.
Williams, J., et al., Sympatric Populations of the Anopheles gambiae Complex in Southwest Burkina Faso Evolve Multiple Diverse Resistance Mechanisms in Response to Intense Selection Pressure with Pyrethroids. 2022.
Organization, W.H., Test procedures for insecticide resistance monitoring in malaria vector mosquitoes. 2016, World Health Organization.
Santolamazza, F., et al., Insertion polymorphisms of SINE200 retrotransposons within speciation islands of Anopheles gambiae molecular forms. 2008, BioMed Central Ltd. p. 163.
Bass, C., et al., Detection of knockdown resistance (kdr) mutations in Anopheles gambiae: a comparison of two new high-throughput assays with existing methods. 2007, BioMed Central. p. 1–14.
Bass, C., et al., Development of high-throughput real-time PCR assays for the identification of insensitive acetylcholinesterase (ace-1R) in Anopheles gambiae. Pesticide biochemistry and physiology, 2010. 96(2): p. 80–85.
Livak, K.J., Organization and mapping of a sequence on the Drosophila melanogaster X and Y chromosomes that is transcribed during spermatogenesis. Genetics, 1984. 107(4): p. 611–634.
Dixon, P., VEGAN, a package of R functions for community ecology. Journal of vegetation science, 2003. 14(6): p. 927–930.
Kim, D., et al., HISAT: a fast spliced aligner with low memory requirements. 2015, Nature Publishing Group. p. 357–360.
Liao, Y., et al., featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. 2014, Oxford University Press. p. 923–930.
Love, M.I., et al., Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. 2014, Springer. p. 550.
Giraldo-Calderón, G.I., et al., VectorBase: an updated bioinformatics resource for invertebrate vectors and other organisms related with human diseases. 2015. p. D707-13.

No competing interests reported.

floatimage5.png
Supplementary Figure 1: PCA of RNAseq samples. PCA showing principal components (PC) one (x-axis) and two (y-axis) for read counts of Bahirdar field samples (CL, red) and lab susceiptbile (Moz, blue). Variance for each PC is shown on the axis.
SupplementaryFigure2.png
Supplementary Figure 2: Significantly differential cuticular proteins. Normalised read counts for Bahirdar (CL) and lab susceptible (Moz) samples. Key is shown, with red demonstrating higher read counts and blue lower. The histogram on the key represents number of samples/genes within that Z-score. Histogram on the left shows distances of gene expression patterns.
TableS1.xlsx
TableS2.xlsx
TableS3.xlsx
TableS4.xlsx
TableS5.xlsx
TableS6.xlsx

Download PDF

Editorial decision: Revision requested
03 Nov, 2024
Reviews received at journal
31 Oct, 2024
Reviews received at journal
27 Oct, 2024
Reviews received at journal
16 Oct, 2024
Reviewers agreed at journal
15 Oct, 2024
Reviewers agreed at journal
14 Oct, 2024
Reviewers agreed at journal
14 Oct, 2024
Reviewers invited by journal
12 Oct, 2024
Editor assigned by journal
26 Sep, 2024
Submission checks completed at journal
26 Sep, 2024
First submitted to journal
25 Sep, 2024

You are reading this latest preprint version

Insecticide resistant Anopheles from Ethiopia but not Burkina Faso show a microbiome composition shift upon insecticide exposure.

Status:

Version 1

Abstract

Figures

Introduction

Results

Species identification of field samples

Resistance levels

Microbiome diversity

Bacteria associated with survival to pyrethroid exposure

Discussion

Methods

Field Collections

Mosquito Rearing

Bioassays

Sample processing

Species identification and molecular analysis

DNA extraction for 16S sequencing

16S Analysis

RNA extraction

RNAseq analysis

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1