Metagenome-assembled genomes of three Hepatoplasmataceae provide insights into isopod-mollicute symbiosis

Abstract

The digestive organs of terrestrial isopods harbour bacteria of the recently proposed mollicute family Hepatoplasmataceae. The only complete genome available so far for Hepatoplasmataceae is that of 'Candidatus Hepatoplasma crinochetorum'. The scarcity of genome sequences has hampered our understanding of the symbiotic relationship between isopods and mollicutes. Here, we present four complete metagenome-assembled genomes (MAGs) of uncultured Hepatoplasmataceae members identified from shotgun sequencing data of isopods. We propose genomospecies names for three MAGs that show substantial sequence divergence from any previously known Hepatoplamsataceae members: 'Candidatus Tyloplasma litorale' identified from the semiterrestrial isopod Tylos granuliferus, 'Candidatus Hepatoplasma vulgare' identified from the common pill bug Armadillidium vulgare, and 'Candidatus Hepatoplasma scabrum' identified from the common rough woodlouse Porcellio scaber. Phylogenomic analysis of 155 mollicutes confirmed that Hepatoplasmataceae is a sister clade of Metamycoplasmataceae in the order Mycoplasmoidales. The 16S ribosomal RNA gene sequences and phylogenomic analysis showed that 'Candidatus Tyloplasma litorale' and other semiterrestrial isopod-associated mollicutes represent the placeholder genus 'g_Bg2' in the r214 release of the Genome Taxonomy Database, warranting their assignment to a novel genus. Our analysis also revealed that Hepatoplasmataceae lack major metabolic pathways but has a likely intact type IIA CRISPR-Cas9 machinery. Although the localization of the Hepatoplasmatacae members have not been verified microscopically in this study, these genomic characteristics are compatible with the idea that these mollicutes have an ectosymbiotic lifestyle with high nutritional dependence on their host, as has been demonstrated for other members of the family. We could not find evidence that Hepatoplasmataceae encode polysaccharide-degrading enzymes that aid host digestion. If they are to provide nutritional benefits, it may be through extra-copy nucleases, peptidases, and a patatin-like lipase. Exploration of potential host-symbiont interaction-associated genes revealed large, repetitive open reading frames harbouring beta-sandwich domains, possibly involved with host cell adhesion. Overall, genomic analyses suggest that isopod-mollicute symbiosis is not characterized by carbohydrate degradation, and we speculate on their potential role as defensive symbionts through spatial competition with pathogens to prevent infection.

Keywords: Hepatoplasma; Mycoplasma; isopods; metagenome; mollicutes; symbiosis.

Fig. 1.

Genome diagrams of Hepatoplasmataceae members. (a) Circular genome diagrams of Hepatoplasmataceae members. Arrowheads indicate the transcriptional orientation. Outer track: protein-coding genes (blue), ribosomal RNA genes (green), transfer RNA genes (orange), repeat regions (grey). Middle track: GC skew of 100 bp sliding windows with 10 bp increments (positive: emerald, negative: purple). Inner track: Deviation of GC contents from the average, 100 bp sliding windows with 10 bp increments. (b) Linear diagrams of Hepatoplasmataceae genomes. The reference genome for ‘Candidatus Hepatoplasma crinochetorum’ is shown at the bottom as isolate ‘Av’ for comparison. TBLASTX hits (e-value: 1e-3, bitscore:50) are shown in grey.

Fig. 2.

Phylogenetic analysis of 16S rDNA gene sequences. A total of 1462 sites (model: GTR+F+R3) were used in the maximum-likelihood phylogenetic analysis using IQ-TREE v. 2.2.0.3. Values beside nodes indicate the ultrafast bootstrap support (1000 trials).

Fig. 3.

Phylogenomic analysis of Mollicutes. (a) A total of 41 single-copy proteins conserved among 155 mollicute genomes (8219 sites; model: LG+F+I+I+R10) were used in the maximum-likelihood phylogenetic analysis using IQ-TREE v. 2.2.0.3. Values beside nodes indicate the ultrafast bootstrap support (1000 trials). Clades were delineated according to Gupta et al. [51, 52]. (b) Subtree of (a) highlighting Hepatoplasmataceae. Names in bold indicate genus names in the GTDB-Tk reference data version r214 [48, 49].

Fig. 3.

Phylogenomic analysis of Mollicutes. (a) A total of 41 single-copy proteins conserved among 155 mollicute genomes (8219 sites; model: LG+F+I+I+R10) were used in the maximum-likelihood phylogenetic analysis using IQ-TREE v. 2.2.0.3. Values beside nodes indicate the ultrafast bootstrap support (1000 trials). Clades were delineated according to Gupta et al. [51, 52]. (b) Subtree of (a) highlighting Hepatoplasmataceae. Names in bold indicate genus names in the GTDB-Tk reference data version r214 [48, 49].

Fig. 4.

Summary of metabolic pathways in Hepatoplasmataceae. The diagram was drawn manually based on KEGG Pathway diagrams generated on the BLASTKOALA server (https://www.kegg.jp/blastkoala/) (53), incorporating manually curated genes in Table 5.

Fig. 5.

Large repetitive ORFs in Hepatoplasmataceae MAGs. (a) Diagrams of large, repetitive ORFs in Hepatoplasmataceae MAGs. Light blue arrows indicate the protein-coding genes and their transcriptional orientations. Numbers at the left and right ends indicate the start and end coordinates of the genome segments shown in the figure. Vertical rounded rectangles indicate the locations of repetitive motifs. (b) Schematic (left) and ColabFold 3D structure prediction (right) of beta-sandwich domains in the large repetitive ORFs of Hepatoplasmataceae MAGs. The sequences are coloured from blue at the N terminus to red at the C terminus. Arrows represent beta-strands, helices are shown as cylinders, and the loops are drawn as grey lines. Beta-strands forming Ig-like folds (HCTKY_2320 and HPPSJP_2440) are numbered alphabetically according to [111]. The numbering of beta-strands in TYPL_3910 is arbitrary.