Abstract

Members of the gut microbiota genus Bifidobacterium are widely distributed human and animal symbionts believed to exert beneficial effects on their hosts. However, in-depth genomic analyses of animal-associated species and strains are somewhat lacking, particularly in wild animal populations. Here, to examine patterns of host specificity and carbohydrate metabolism capacity, we sequenced whole genomes of Bifidobacterium isolated from wild-caught small mammals from two European countries (UK and Lithuania). Members of Bifidobacterium castoris, Bifidobacterium animalis and Bifodobacterium pseudolongum were detected in wild mice (Apodemus sylvaticus, Apodemus agrarius and Apodemus flavicollis), but not voles or shrews. B. castoris constituted the most commonly recovered Bifidobacterium (78% of all isolates), with the majority of strains only detected in a single population, although populations frequently harboured multiple co-circulating strains. Phylogenetic analysis revealed that the mouse-associated B. castoris clades were not specific to a particular location or host species, and their distribution across the host phylogeny was consistent with regular host shifts rather than host-microbe codiversification. Functional analysis, including in vitro growth assays, suggested that mouse-derived B. castoris strains encoded an extensive arsenal of carbohydrate-active enzymes, including putative novel glycosyl hydrolases such as chitosanases, along with genes encoding putative exopolysaccharides, some of which may have been acquired via horizontal gene transfer. Overall, these results provide a rare genome-level analysis of host specificity and genomic capacity among important gut symbionts of wild animals, and reveal that Bifidobacterium has a labile relationship with its host over evolutionary time scales.

Highlights

  • Species and strains belonging to the bacterial genus Bifidobacterium are prominent members of the gut microbiota in many animals, and are universally distributed among animals exhibiting parental care, including humans and non-human mammals, birds and social insects [1]

  • We identified 26 to new conditions following an allopatric event can result in an isolates as B. castoris, 4 isolates as B. animalis and a further 3 altered intestinal environment, to which symbiotic bacteria can isolates as B. pseudolongum (Supplementary Tables S1 and S2)

  • Given the prediction of glycosyl transferases (GTs) genes in the B. castoris glycobiome, we examined our collection of genomes for the presence of genes potentially involved in EPS biosynthesis, as previous studies have indicated that EPS may support gut colonisation and stimulate host immune responses [3]

Read more

Summary

INTRODUCTION

Species and strains belonging to the bacterial genus Bifidobacterium are prominent members of the gut microbiota in many animals, and are universally distributed among animals exhibiting parental care, including humans and non-human mammals, birds and social insects [1]. The predominant GH family, with 14.15 ± 1.29 GH genes per genome (mean ± sd), constituting on average 29.98 ± 1.88% of the GH repertoire, was GH13 whose members include enzymes acting on a very wide range of carbohydrates containing α-glucoside linkages, e.g. starches and related substrates, trehalose, raffinose, stachyose and melibiose [9, 38, 41] These results were consistent with previous studies of the type strains representative of the genus Bifidobacterium, as well as strains belonging to B. pseudolongum species, which identified this particular GH family as the most commonly detected, in particular in the genomes of strains isolated from mammals [9, 24]. We test whether predicted to horizontally acquire a GH36 family α-galactosidase a similar pattern holds true based on the analysis of genomic (Supplementary Table S9) Overall, these results suggest that HGT sequences of isolates from a single Bifidobacterium species may have contributed to the evolution of B. castoris strains and (B. castoris) and their wild murine hosts. In contrast to previous work [11, 22], we found no statistical congruence

DISCUSSION
CONCLUSION
Findings
MATERIALS AND METHODS
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call