Mitochondrial genomes of Anopheles arabiensis, An. gambiae and An. coluzzii show no clear species division

Introduction

Historically, mtDNA sequence has been used in taxonomy as a source of species diagnostic markers (Cronin et al. (1991); De Barba et al. (2014); Pegg et al. (2006)) or in population genetics and evolutionary studies (Fu et al. (2013); Harrison (1989); Llamas et al. (2016)). One advantage of using mitochondrial over nuclear DNA for such studies is that the mutation rate of mtDNA is about 10 times faster than nuclear DNA (Brown et al. (1979); Haag-Liautard et al. (2008)), hence amplifying the evolutionary trajectory of populations and species. In addition, mtDNA is easy to amplify, because there are more copies of mitochondrial DNA relative to nuclear DNA. Also, universal primers can be applied to a wide range of species. Widely used universal primers target the cytochrome b and cytochrome oxidase 1 genes (Tahir et al. (2016)), because both have conserved and highly variable regions. In addition to these, other genes as described in De Mandal et al. (2014), can also be used as markers. However, phylogenetic trees based on mtDNA can deviate from the ones that are derived from nuclear DNA (Phillips et al. (2013); Shaw (2002); Sota & Vogler, 2001).

The Anopheles gambiae species complex consists of eight morphologically identical species that can only be distinguished with molecular markers (Scott et al. (1993); Coetzee et al., 2013) or, for some of the species, by cytological examination of polytene chromosomes (Green, 1972; Pombi et al., 2008). The currently used molecular markers to distinguish between An. coluzzii and An. gambiae (Lee et al., 2014) are located within genomic islands of divergence located proximal to the centromeres (Turner et al. (2005)). Monitoring additional species-specific markers on mitochondrial DNA (mtDNA) could increase the ease of application and accuracy of species detection assays. In addition, mtDNA markers could enhance our understanding of divergence times among taxa within the complex.

Previous studies showed that there is a high amount of interspecific gene flow in mtDNA between An. coluzzii, An. gambiae and An. arabiensis specimens (Besansky et al., 2003; Besansky et al., 1997; Donnelly et al., 2004). Although these data suggested no evidence for clear species division among the various species, the studies only focused on the ND5 loci (Besansky et al., 2003; Donnelly et al., 2004) or included also cytochrome b and ND1 loci (Besansky et al., 1997). In our study we use the complete mitogenome for comparison, which would make the analysis more robust. In addition, we specifically included the different chromosomal forms in our analysis. These chromosomal forms are genetically diverged from each other and display strong assortative mating in the An. gambiae chromosomal forms (Touré et al., 1998). The An. coluzzii chromosomal forms differ from each other in their ecology: An. coluzzii-Mopti is found in dry areas whereas the An. coluzzii-Forest restrtict themselves to a wet climate (Lee et al., 2009).

In this study we wished to identify species-specific markers within the mtDNA for Anopheles arabiensis, An. coluzzii and An. gambiae, including among the chromosomal forms currently subsumed under the designations An. gambiae and An. coluzzii, with the goal of adding these to our existing Anopheles species detection assay (Lee et al. (2014)). We sequenced the whole mitogenomes of 70 individual mosquito specimens collected throughout Sub-Saharan Africa. The raw Illumina sequencing reads were mapped to the AgamP4 reference sequence, which included both nuclear and mitochondrial sequences. We explore the relationship among An. arabiensis, An. coluzzii, An. gambiae and four of the sub-specific chromosomal form mitogenome sequences.

Methods

Results and Discussion

We identified a total of 783 single nucleotide polymorphisms (SNPs) over the entire mitogenome. The majority of these (58.7%) were singletons (found on one of the 70 mitogenomes). We did not identify any SNPs unique to the species or chromosomal forms (Supplementary Table S1) and therefore conclude that mtDNA is not suitable for Anopheles gambiae complex species identification.

The lack of species-specific markers is also reflected in the phylogenetic tree (Figure 1). An. arabiensis, An. coluzzii and An. gambiae did not cluster separately, which is consistent with previous reports that compared mitochondrial genome sequence data from specimens originating from Kenya, Senegal and South Africa (Besansky et al. (1997)) and Burkina Faso, Cameroon, Kenya, Mali, South Africa, Tanzania and Zimbabwe (Fontaine et al. (2015), Supplemental material).

Figure 1. Phylogenetic tree inferred from mtDNA genome sequence data.

The phylogenetic tree fails to reveal a clear division of the operational taxonomic units included in this analysis. Colors indicate the species or chromosomal form and numbers at the branches indicate the accuracy of the inferred branches on a scale of 0–1, where 1 represents the highest confidence. The three An. arabiensis lineages are previously reported by Maliti and co-workers (Maliti et al., 2016).

Our data may indicate that there is no divergent selection in mitogenome among An. gambiae complex. Since mitochondrial genomes have a higher (1–10 times) substitution rate than nuclear genomes (Havird & Sloan, 2016; Lynch & Walsh, 2007), one might expect some level of divergence in the mitogenome in the absence of selection if the taxa have been separated by reproductive barrier even if they are in sympatry just as people have observed in nuclear genome. Therefore, our data showing lack of any species-specific markers on the mitogenome may due to the results of episodic hybridizations occurred between two species. Of note, 36 of the samples that we used in our study originated from Kela (Mali). Kela is located near the village of Selinkenyi, where previous studies have shown a history of hybridization and introgression between An. gambiae and An. coluzzii (Lee et al. (2013); Main et al. (2015); Norris et al. (2015)), which may have resulted in shared polymorphisms in their mitochondrial genomes. Shared polymorphisms in their mitochondrial genomes, where history has not been reported, also appeared to have occurred in Mutengene (Cameroon), where both An. gambiae and An. coluzzii occur sympatrically. Hybridization between either An. coluzzii or An. gambiae with An. arabiensis yields sterile males (Slotman et al. (2004)), but phylogenomic analysis of these species show patterns of introgression between all of them (Fontaine et al. (2015)), which could be the reason that we do not find any species-specific markers on the mitogenome. Our mitochondrial genome study does not provide conclusive evidence for hybridization and introgression among the taxa under study. However, our data suggest that this is a possibility and it would be consistent with results reported by (Fontaine et al., 2015) and (Besansky et al., 1997). Future modeling work may illuminate the likely contribution of different evoluationary forces that shapes mitogenome and nuclear genome evolution.

Data availability

Aligned sequences were submitted to the National Center for Biotechnology Information (NCBI) Accession number: MG930826 - MG930896

Dataset 1. Aligned FASTA file of mitogenome samples 10.5256/f1000research.13807.d192892 (Hanemaaijer et al., 2018)