Introduction
An increasing number of studies demonstrate ubiquity and high diversity of insect-associated microbiomes (Douglas, 2015; Engel, & Moran, 2013). These microbial communities, composed of various pathogens, commensals and random contaminants, can serve as natural sources of beneficial symbiotic bacteria. In some insects they give rise to highly specialized, maternally-transmitted mutualists called primary symbionts (P-symbionts), which contribute to the host’s metabolism (Douglas, 1989). However, depending on richness and dynamics, the microbiomes usually contain several symbiotic bacteria in various evolutionary stages. In their typical form, P-symbionts are readily recognized by several features (since they are indispensable mutualists): they are universally present in all individuals, as a rule inhabiting specialized host’s organs called bacteriomes (Baumann, 2005), and their genomes are significantly reduced with a strong AT bias (Moran, 1996). One specific feature of P-symbionts is their co-phylogeny with the host (Chen, Li, & Aksoy, 1999; Clark, Moran, Baumann, & Wernegreen, 2000; Sauer, Stackebrandt, Gadau, Holldobler, & Gross, 2000). For example, two of the most studied P-symbionts, Buchnera in aphids andWigglesworthia in tsetse flies, were acquired at the beginning of their hosts’ diversification and strictly mirror their entire phylogeny (Chen et al., 1999, Clark et al., 2000). Other P-symbionts are restricted to some of the host’s lineages, indicating that they are either recently acquired symbionts or remnants of an ancient symbiont lost in some of the host lineages (Bennett & Moran, 2013). In some insects, several different P-symbionts may coexist and/or can be accompanied by various secondary symbionts (S-symbionts). The latter are less modified, retain more free-living-like characteristics, and some are supposed to be the intermediate stages of evolution towards obligate symbionts. Wigglesworthia represents a typical example of this as it is often accompanied by the S-symbionts Sodalis glossinidiusand Wolbachia (Aksoy, 2000). The complexity of symbiotic associations is obviously due to an ongoing process of symbiont acquisition/loss/replacement, which is well known from several bacteria-insect models and has a well-developed theoretical background (Bennett & Moran, 2015). The theoretical work postulates that after a certain amount of coevolutionary time, the symbiotic bacterium becomes too degenerated and functionally inadequate, and it has to be replaced (or accompanied) by another symbiont. While it would be interesting to see how the microbiome diversity and dynamics relate to the complexity of symbiosis in different insect groups, there is very little information available today. The majority of studies on insects and their P- and S- symbionts relies on metagenomic information and phylogenetic reconstructions, likely missing a substantial part of microbiome diversity. The introduction of amplicon approaches recently demonstrated that this method can significantly improve our insight into microbiome composition, even in extensively studied model systems (Doudoumis et al., 2017; Gauthier, Outreman, Mieuzet, & Simon, 2015; Manzano-Marin, Szabo, Simon, Horn, & Latorre, 2017; Meseguer et al., 2017).
Amongst hematophagous (blood-feeding) insects which live exclusively on vertebrate blood, sucking lice of the order Anoplura, with more than 500 spp. (Light, Smith, Allen, Durden, & Reed, 2010), are the most ancient and diversified group. Accordingly, they possess a high diversity of symbiotic bacteria (Allen, Burleigh, Light, & Reed, 2016; Boyd, Allen, de Crécy-Lagard, & Reed, 2014; Boyd et al., 2016; Fukatsu et al., 2009; Hypsa & Krizek, 2007). Depending on interpretation, the 16S rRNA gene-based phylogenies for the available taxa suggest 5-6 independent symbiotic lineages. However, none of them is a universal louse symbiont distributed across the whole order (e.g. like Buchnera in aphids). The distribution of louse symbionts suggests a relatively recent origin of each lineage and hence a high rate of acquisition/loss/replacement processes. Moreover, compared to the extensively screened phytophagous groups, only a small fraction of sucking lice diversity has been investigated. The actual number of symbiotic lineages is therefore likely to be much higher. Of the currently known lineages, genomic data are only available for four; three of them showing clear signatures of P-symbionts: Riesiaspp., Puchtella pedicinophila , and Legionellapolyplacis (Table 1). Correspondingly, each of these lineages has been found in two to four related host species as a result of co-phylogenetic processes. The fourth lineage, the Sodalis -like symbiont from Proechinophthirus fluctus , possesses a significantly larger genome exceeding 2 Mbp, and GC content 50%, which the authors interpret as possible evidence of recent replacement of a more ancient and now extinct endosymbiont (Boyd et al., 2016). The diversity and distribution of the known symbionts in sucking lice thus indicate that this insect group has been undergoing particularly dynamic acquisition, loss, and replacement of symbionts. In this study, we analyze the background of these processes by combining genomic and amplicon approaches across several populations of the louse generaPolyplax and Hoplopleura . We reveal a new symbiotic lineage related to the genera Neisseria and Snodgrassella(the latter being a symbiont of bees). We show that these bacteria established their symbiotic relationships independently with the two louse genera, and we prove their intracellular localization in host’s bacteriocytes. Based on the phylogeny-dependent diversity of the microbiome profiles, we suggest rapid microbiome changes at the host population level, possibly underlying the dynamic processes of symbiont acquisition, loss, and replacement in these blood sucking insects.