Environmental Microbiology (2016) 18(8), 2646–2659 doi:10.1111/1462-2920.13411 Genomic characterization of symbiotic mycoplasmas from the stomach of deep-sea isopod bathynomus sp Yong Wang,1 Jiao-Mei Huang,1 Shao-Lu Wang,1 Introduction Zhao-Ming Gao,1 Ai-Qun Zhang,1 Deep-sea isopods are scavengers in the benthic environ- Antoine Danchin2* and Li-Sheng He1* ment. They survive in resource-poor conditions for a long 1Institute of Deep-Sea Science and Engineering, period. A deep-sea isopod died after what, from an anthro- Chinese Academy of Sciences, Sanya, Hainan, China. pocentric perspective, looked like a five-year hunger strike 2Hopital^ de la Pitie-Salp etrie^ `re, Institute of in a Japanese aquarium (Ginn et al., 2014). This phenom- Cardiometabolism and Nutrition, 47 boulevard de enon attracted interest in their distinctive nutrient storage l’Hopital,^ Paris 75013, France. capabilities and maintenance of low metabolic activity. More interestingly, deep-sea isopods are much larger than ABSTRACT their coastal relatives, highlighting an example of gigantism and introducing an enigma for deep-sea biologists Deep-sea isopod scavengers such as Bathynomus sp. (Timofeev, 2001). The first comprehensive survey of the are able to live in nutrient-poor environments, which is distribution of giant deep-sea isopods was conducted in likely attributable to the presence of symbiotic 1879 in the Gulf of Mexico (Briones-Fourzan and Lozano- microbes in their stomach. In this study we recovered Alvarez, 1991). Subsequently, mostly because of sampling two draft genomes of mycoplasmas, Bg1 and Bg2, difficulties, studies investigating deep-sea isopods have from the metagenomes of the stomach contents and been rather sparse. As a result, the origin and evolution of stomach sac of a Bathynomus sp. sample from the deep-sea isopod are still hotly debated (Schultz, 1979; South China Sea (depth of 898 m). Phylogenetic trees Raupach et al., 2009). revealed a considerable genetic distance to other The deep-sea isopod Bathynomus giganteus is found at mycoplasma species for Bg1 and Bg2. Compared with depths ranging roughly from 310 m to 2140 m in global terrestrial symbiotic mycoplasmas, the Bg1 and Bg2 oceans (Holthuis and Mikulka, 1972). Due to its wide distri- genomes were enriched with genes encoding bution in deep-sea areas, B. giganteus is an excellent phosphoenolpyruvate-dependent phosphotransferase model for studies of adaptation, evolution and ecological systems (PTSs) and sodium-driven symporters importance in deep-sea ecosystems. The food sources of responsible for the uptake of sugars, amino acids and deep-sea animals are largely dependent on organic par- other carbohydrates. The genome of mycoplasma Bg1 ticles that descend from the ocean surface. For isopods contained sialic acid lyase and transporter genes, living on the seafloor below a depth of 300 m, food is scarce potentially enabling the bacteria to attach to the stom- due to the consumption of labile organic matter in the over- ach sac and obtain organic carbons from various cell lying water column. The cellulose, agarose, chitin, and walls. Both of the mycoplasma genomes contained recalcitrant nutrients that reach the benthic sediments are multiple copies of genes related to proteolysis and oli- usually the main food sources of B. giganteus, in addition to gosaccharide degradation, which may help the host fresh food that is occasionally captured (Briones-Fourzan survive in low-nutrient conditions. The discovery of and Lozano-Alvarez, 1991). The ability of B. giganteus to the different types of mycoplasma bacteria in the survive in such poor conditions is probably at least partially stomach of this deep-sea isopod affords insights into attributable to the activity of symbionts present in their gut. symbiotic model of deep-sea animals and genomic For example, using scanning electron microscopy, Boyle plasticity of mycoplasma bacteria. et al. observed a high density of microbes in the gut of B. giganteus (Boyle and Mitchell, 1982). The symbiotic bacte- ria in terrestrial isopods are not obligate symbionts based on evidence establishing their environmental transfer rather Received 17 May, 2016; accepted 6 June, 2016. *For correspon- dence. E-mail [email protected] and antoine.danchin@normal- than vertical transfer (Wang et al., 2004; 2007). However, esup.org; Tel. (86) 898-88380060; Fax (86) 898-88222506. whether deep-sea isopods have similar bacterial species in VC 2016 Society for Applied Microbiology and John Wiley & Sons Ltd Mycoplasma in deep-sea isopod stomach 2647 their gut and maintain an obligate symbiotic relationship contents and all 34 sequences from the stomach sac were with microbes remains unanswered. nearly identical (>99%), demonstrating the highest similar- The evolution of symbiotic relationships between marine ity with the 16S rRNA gene of “Ca. Hepatoplasma isopods and bacteria may provide additional lessons about crinochetorum” (88%) (Supporting Information Table S1). the origins of isopods. The phylogenetic relationships Remarkably, these observations are consistent with the observed in the Mycoplasma bacteria in their digestive conception that the stomach, particularly on the surface of tracts were consistent with those of the isopods (Fraune the stomach sac, was dominated by mycoplasmas, with and Zimmer, 2008). These findings suggest a co-evolution hardly any other bacteria present. A second species, of isopods with their associated symbionts, as well as the Hyphomonas jannaschiana (Alphaproteobacteria), was presence of isopod species-specific symbiotic bacteria also relatively abundant (23%) in the stomach contents. (Fraune and Zimmer, 2008). The significance of the gut The phylogenetic relationships of the rRNA genes with symbiotic bacteria in deep-sea isopods may be exposed previously known relatives revealed a comparatively distant by deciphering the genomes of the symbionts. Unfortu- position with mycoplasma symbionts residing in the guts of nately, these gut symbionts are unculturable, and the deep-sea shrimps, crabs and terrestrial isopods (Hepato- limited amount of bacterial genomic DNA makes a chal- plasma group) (Fig. 1). Clades comprising symbiotic lenge for investigations of genomic features. The terrestrial Mycoplasma from different hosts may be discerned in the isopod Porcellio scaber carries the bacterial symbionts phylogenetic tree. However, some of the corresponding “Candidatus Hepatoplasma crinochetorum” in its midgut maximum-likelihood bootstrap values (<70) did not support gland (Leclercq et al., 2014). Leaves and wood have been the grouping of the clades with strong confidence (Fig. 1). found to be digested in the midgut of P. scaber,probably Intriguingly, among the closest relatives, some were isolat- with the help of “Ca. Hepatoplasma crinochetorum.” In a ed from sediments and a sea snail near hydrothermal recent study, the full-length genome of “Ca. Hepatoplasma vents. In addition to the marine environments and hosts, crinochetorum” was obtained from P. scaber and Armadilli- the remaining Mycoplasma species shown in Fig. 1 were dium vulgare (Leclercq et al., 2014; Collingro et al., 2015). associated with terrestrial hosts. The Mycoplasma species Subsequent phylogenomic analyses revealed that “Ca. that formed two branches associated with the outgroup Hepatoplasma crinochetorum” belonged to the Mollicutes species in Acholeplasma-Anaeroplasma group and Spiro- class. However, its genome is an isolated example that plasma group were largely animal pathogens. Altogether, has not yet been compared with the genome of other iso- the phylogenetic tree exhibited lineage congruencies that pod symbionts to highlight genomic features that are were determined by both hosts and environments. common or specific to different environments. To our Although the mycoplasma 16S rRNA genes from the iso- knowledge, the genomes of Mycoplasma from deep-sea pod stomach were almost identical, minor discrepancies zones have not yet been reported. Furthermore, differen- between the genes were identified. A pairwise comparison ces between the distribution and adaptive strategies of of the 16S rRNA sequences showed that the sequences deep-sea Mycoplasma, compared with their terrestrial rel- from the stomach sac were likely from a common organ- atives, remain unexplored. In this study, we captured ism, whereas the 16S rRNA sequences from the stomach several Bathynomus sp. samples using a baited trap in the contents were more diverse. The pairwise dissimilarity (on South China Sea. Using next-generation sequencing and average 0.3%) between the sequences from the sac con- a subsequent bioinformatics pipeline, we obtained two tents was significantly higher than the dissimilarity (on draft genomes of symbiotic mycoplasmas from the stom- average 0.04%) for the stomach sac (Mann-Whitney test; ach of Bathynomus sp. By comparing the genomes with P < 1025). The significance of this difference was substanti- those of close relatives, we obtained a consistent picture in ated by multidimensional scaling (MDS) analysis. The 16S which the mycoplasmas evolved as local scavengers that rRNA sequences from the stomach sac were concentrated allowed better utilization of food sources in the stomach of in a small region of the MDS plot, whereas the sequences the isopod and deep-sea benthic floor, providing insights from the stomach contents were widely scattered (Support- into the importance of deep-sea symbiotic mycoplasmas ing Information Fig. S1). This result is consistent with the