Many arthropod hosts are infected with bacterial endosymbionts that manipulate host reproduction, but few bacterial taxa have been shown to cause such manipulations. Here, we show that a bacterial strain in the genus Rickettsiella causes cytoplasmic incompatibility (CI) between infected and uninfected hosts. We first surveyed the bacterial community of the agricultural spider Mermessus fradeorum (Linyphiidae) using high throughput sequencing and found that individual spiders can be infected with up to five different strains of maternally inherited symbiont from the genera Wolbachia , Rickettsia , and Rickettsiella . The Rickettsiella strain was pervasive, found in all 23 tested spider matrilines. We used antibiotic curing to generate uninfected matrilines that we reciprocally crossed with individuals infected only with Rickettsiella . We found that only 13% of eggs hatched when uninfected females were mated with Rickettsiella -infected males; in contrast, at least 83% of eggs hatched in the other cross types. This is the first documentation of Rickettsiella , or any Gammaproteobacteria, causing CI. We speculate that induction of CI may be much more widespread among maternally inherited bacteria than previously appreciated. Further, our results reinforce the importance of thoroughly characterizing and assessing the inherited microbiome before attributing observed host phenotypes to well-characterized symbionts such as Wolbachia .

Acknowledgements:We thank J. Harwood and J. Dryer for providing and collecting 22 specimens.Abstract 31 Maternally inherited bacterial endosymbionts are common in arthropods, but their distribution 32 and prevalence is poorly characterized in many host taxa. For example, spiders (Araneae) have 33 received little attention, but initial surveys suggest that vertically transmitted symbionts may be 34 common. Here, we characterized endosymbiont infection in a community of agricultural spiders. 35Using a combination of diagnostic PCR and high-throughput sequencing of the bacterial 36 microbiome, we evaluated symbiont infection in 267 individual spiders representing 14 species 37 in 3 families. We found 27 Operational Taxonomic Units (OTUs) that are likely endosymbiotic, 38 including several strains of Wolbachia, Rickettsia and Cardinium, all of which are vertically 39 transmitted and frequently associated with reproductive manipulation of arthropod hosts. 40Seventy percent of spider species had individuals that tested positive for one or more 41 endosymbiotic OTUs, and specimens frequently contained multiple symbiotic strain types. The 42 most symbiont-rich species, Idionella rugosa, had eight endosymbiotic OTUs, with as many as 43 five present in the same specimen. Individual specimens within infected spider species had a 44 variety of symbiotypes, differing from one another in the presence or absence of symbiotic 45 strains. Our sample included both starved and unstarved specimens, and dominant bacterial 46OTUs were consistent per host species, regardless of feeding status. We conclude that spiders 47 contain a remarkably diverse symbiotic microbiota. Spiders would be an informative group for 48 investigating endosymbiont population dynamics in time and space, and unstarved specimens 49 collected for other purposes (e.g., food web studies) could be used, with caution, for such 50 investigations. 51 52 53The majority of terrestrial arthropod species are infected by inherited bacterial symbionts. 54The most common of such bacteria, Wolbachia, is estimated to infect approximately half of 55 arthropod species, albeit often at a low frequency of infected individuals within a given species 56 [1-3]. These bacteria are primarily transmitted vertically, but horizontal transfers among host 57 taxa occasionally occur through a variety of ecological mechanisms such as shared host 58 substrates or shared enemies [4-8]. Following successful establishment in a new host lineage, the 59 endosymbiont then promotes the continued existence of that infected lineage, either by providing 60 some sort of fitness benefit to the host [9,10], or by manipulating host reproduction such that 61 infected mothers produce more daughters than uninfected mothers [11]. Over generations, the 62 result is an infection that can spread through the host population via selection, rather than 63 contagion. 64 It is inevitable that this process of infection and spread sometimes occurs in the context 65 of other vertically-transmitted endosymbionts, ...

Maternally inherited bacterial endosymbionts are common in arthropods, but their distribution and prevalence is poorly characterized in many host taxa. For example, spiders (Araneae) have received little attention, but initial surveys suggest that vertically transmitted symbionts may be common. Here, we characterized endosymbiont infection in a community of agricultural spiders.Using a combination of diagnostic PCR and high-throughput sequencing of the bacterial microbiome, we evaluated symbiont infection in 267 individual spiders representing 14 species in 3 families. We found 27 Operational Taxonomic Units (OTUs) that are likely endosymbiotic, including several strains of Wolbachia, Rickettsia and Cardinium, all of which are vertically transmitted and frequently associated with reproductive manipulation of arthropod hosts.Seventy percent of spider species had individuals that tested positive for one or more endosymbiotic OTUs, and specimens frequently contained multiple symbiotic strain types. The most symbiont-rich species, Idionella rugosa, had eight endosymbiotic OTUs, with as many as five present in the same specimen. Individual specimens within infected spider species had a variety of symbiotypes, differing from one another in the presence or absence of symbiotic strains. Our sample included both starved and unstarved specimens, and dominant bacterial OTUs were consistent per host species, regardless of feeding status. We conclude that spiders contain a remarkably diverse symbiotic microbiota. Spiders would be an informative group for investigating endosymbiont population dynamics in time and space, and unstarved specimens collected for other purposes (e.g., food web studies) could be used, with caution, for such investigations.

No abstract

No abstract