Pangenome analysis of Bifidobacterium longum and site-directed mutagenesis through by-pass of restriction-modification systems

Abstract

Background: Bifidobacterial genome analysis has provided insights as to how these gut commensals adapt to and persist in the human GIT, while also revealing genetic diversity among members of a given bifidobacterial (sub)species. Bifidobacteria are notoriously recalcitrant to genetic modification, which prevents exploration of their genomic functions, including those that convey (human) health benefits.

Methods: PacBio SMRT sequencing was used to determine the whole genome seqeunces of two B. longum subsp. longum strains. The B. longum pan-genome was computed using PGAP v1.2 and the core B. longum phylogenetic tree was constructed using a maximum-likelihood based approach in PhyML v3.0. M.blmNCII was cloned in E. coli and an internal fragment if arfBarfB was cloned into pORI19 for insertion mutagenesis.

Results: In this study we present the complete genome sequences of two Bifidobacterium longum subsp. longum strains. Comparative analysis with thirty one publicly available B. longum genomes allowed the definition of the B. longum core and dispensable genomes. This analysis also highlighted differences in particular metabolic abilities between members of the B. longum subspecies infantis, longum and suis. Furthermore, phylogenetic analysis of the B. longum core genome indicated the existence of a novel subspecies. Methylome data, coupled to the analysis of restriction-modification systems, allowed us to substantially increase the genetic accessibility of B. longum subsp. longum NCIMB 8809 to a level that was shown to permit site-directed mutagenesis.

Conclusions: Comparative genomic analysis of thirty three B. longum representatives revealed a closed pan-genome for this bifidobacterial species. Phylogenetic analysis of the B. longum core genome also provides evidence for a novel fifth B. longum subspecies. Finally, we improved genetic accessibility for the strain B. longum subsp. longum NCIMB 8809, which allowed the generation of a mutant of this strain.

Fig. 1

Comparative genomics of fully sequenced B. longum subsp. longum genomes. a Cluster of Orthologues (COG) classification of all ORFs from publicly available B. longum subsp. longum genomes. For each COG entry the average percentage of hits among B. longum subsp. longum has been indicated. The most abundant COG families are assigned to housekeeping functions and have been indicated. From the outer to inner circle: B. longum subsp. longum NCIMB 8809, B. longum subsp. longum CCUG 30698, B. longum subsp. longum NCC2705, B. longum subsp. longum DJO10A, B. longum subsp. longum BBMN68, B. longum subsp. longum JCM1217, B. longum subsp. longum JDM301, B. longum subsp. longum KACC91563, B. longum subsp. longum GT15 and B. longum subsp. longum BXY01. b The number of glycosyl hydrolase family members for B. longum subsp. longum NCIMB 8809 (blue) and B. longum subsp. longum CCUG 30698 (red). The number of predicted extracellular glycosyl hydrolases is indicated

Fig. 2

Pan-genome and core-genome of B. longum. a The pan-genome plot is of the accumulated number of new genes against the number of genomes added. b The core-genome plot is represented by the accumulated number of genes attributed to the core-genomes versus the number of added genomes. The resulting functions for both plots are also reported

Fig. 3

Phylogenetic analysis of B. longum. A phylogenetic supertree showing the relationship between thirty three complete and incomplete bifidobacterial strains and L. salivarius UCC118 as an outlier

Fig. 4

Carbohydrate metabolism by B. longum. Final OD600 values following 24 h growth of various B. longum strains grown on 1 % rye arabinoxylan, 1 % wheat arabinoxylan, 1 % sugar beet arabinan, 1 % xylan from beech wood, 1 % xylo-oligosaccharide (XOS), 1 % arabinose, 1 % xylose and 1 % lactose

Fig. 5

Genetic accessibility of B. longum subsp. longum NCIMB 8809. Transformation efficiency of B.longum subsp. longum NCIMB 8809 using pAM5 (purple bars) and pPKCM7 (blue bars) isolated from E.coli EC101, pWSK29-MSS.blmncI, pNZEM-M.blmncII, pNZEM-M.blmncIII and B.longum subsp. longum NCIMB 8809

Fig. 6

Phenotypic analysis of the B. longum subsp. longum NCIMB 8809 insertion mutant strain. Growth profile analysis of B.longum subsp. longum NCIMB 8809 (purple) and B.longum subsp. longum NCIMB 8809-ArfB (blue) in modified Rogosa broth supplemented with lactose, rye arabinoxylan or wheat arabinoxylan