Although the gut microbiota varies markedly between different vertebrates, some bacterial species are found in various hosts. Frese and colleagues, writing in PLoS Genetics, have identified bacterial factors that allow Lactobacillus reuteri to colonize a particular host, providing insight into the process of bacterial colonization in gut ecosystems.

L. reuteri strains can be divided into several phylogenetic clusters that correlate broadly with the hosts they colonize. To investigate how host specific these strains are, Frese and colleagues tested whether L. reuteri strains isolated from various hosts could colonize mice in which native lactobacilli had been removed through antibiotic treatment but there was an otherwise functional gut microbiota owing to restoration with bacterial cultures and bacteria harvested from other mice. Only strains isolated from rats or mice (referred to here as rodents) could establish themselves to a level similar to that seen in conventional mice; strains from humans, pigs or chickens were either not detected or were present at a level that was several orders of magnitude lower.

Frese and colleagues... have identified bacterial factors that allow Lactobacillus reuteri to colonize a particular host

To identify bacterial factors that determine host specificity, the authors sequenced L. reuteri str. 100–23, which was isolated from a rat, and compared that with the sequence of the human faecal isolate L. reuteri str. F275. There were 633 genes present only in the rat-derived strain, including genes that encode urease and the auxiliary secretion system SecA2, whereas the human-derived strain contained 352 genes that were not found in the rat isolate, including genes that are involved in cobalamin (vitamin B12) synthesis. Furthermore, each strain contained strain-specific genes encoding cell wall proteins, membrane-bound proteins and glycosyltransferases.

Using microarrays based on the two sequenced genomes, the authors determined the occurrence of the strain-specific genes among additional isolates from rodents and from human faecal samples. After excluding genes involved in transposition, this analysis reduced the number of genes specific for rodent-derived and human-derived L. reuteri to 93 and ten, respectively. Genome sequence comparisons and PCR analysis revealed that several of the rodent-specific genes were also rare in isolates from pigs, chickens and turkeys. The authors then individually inactivated eight genes in L. reuteri str. 100–23 that were representative of certain groups of rodent-specific genes. Deletions of genes encoding a two-component system, two different ABC-type transporters, proteins of the SecA2 secretion system and a serine-rich surface antigen yielded strains that were significantly less able to colonize Lactobacillus-free mice when competing with a wild-type strain. Although the exact function of these genes is unknown, they may have specific roles in the rodent host. For example, rodent strains form biofilms in the forestomach and, accordingly, several rodent-specific genes have putative roles in biofilm formation, adherence and resistance to acidic conditions. By contrast, in the human gut L. reuteri is likely to colonize the intestine, where cobalamin may be required for the utilization of 1,2–propanediol and for production of the antibiotic reuterin. Together, this study provides new insight into the factors that allow bacterial colonization of specific vertebrate hosts, an area that is currently only poorly understood.