Comparative Genomic Study of Lactobacillus jensenii and the Newly Defined Lactobacillus mulieris Species Identifies Species-Specific Functionality

OBSERVATION

The Lactobacillus species L. crispatus, L. gasseri, L. iners, and L. jensenii are predominant members of the “healthy” female urogenital microbiota (1, 2). L. jensenii has been shown to be a protective species, reducing growth of uropathogenic Escherichia coli and sexually transmitted infections (3–6). Recently, Rocha et al. (7) presented a new Lactobacillus species, L. mulieris. The type strain for this new species is L. mulieris c10Ua161M, isolated from a urine sample. Rocha et al. (7) found that the L. mulieris c10Ua161M genome, with eight L. jensenii strains, had an average nucleotide identity (ANI) with the L. jensenii type strain genome below the 95% species threshold (8). Coinciding with the Rocha et al. publication (7), we deposited and published two new Lactobacillus urinary isolates (9, 10); while matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) classified these strains as L. jensenii, ANI analysis suggested they were members of the new L. mulieris species. This prompted us to investigate 43 publicly available L. jensenii and L. mulieris genomes (see Table S1 in the supplemental material).

First, we examined the L. jensenii and L. mulieris 16S rRNA gene sequences, which were ≥99.54% identical (Fig. 1A). The two species can be distinguished by just two nucleotides in the 16S rRNA gene sequence (see Fig. S1 in the supplemental material). The L. mulieris clade included the type strain, L. mulieris c10Ua161M, as well as strains identified by Rocha et al. (7) as L. mulieris based upon pheS and rpoA gene sequence trees. Our two recent sequences UMB7784 and UMB9245, as well as six other strains, form a clade with the L. mulieris type strain, suggesting that they too represent this new species.

• Open in new tab
• Download powerpoint

FIG 1

Comparison of L. jensenii and L. mulieris genomes. (A) 16S rRNA gene sequence comparison. L. gasseri ATCC 33323 = JCM 1131 (GenBank accession no. NC_008530) and L. crispatus ST1 (NC_014106) are included as an outgroup. The range of pairwise identity between groups is shown on the branches. (B) Phylogenetic tree of single-copy-number genes in the core genome. The range of pairwise identity between groups is shown on the branches. In the legend or key, the colors used for the species designation determined in this study and the isolation site are indicated for all genomes in both trees.

Next, the pangenome and set of single-copy genes in the core genome of the 43 L. jensenii and L. mulieris genomes were identified using the tool anvi’o (11) (see Text S1 in the supplemental material). Phylogenomic analysis is based upon the concatenated protein sequences of the 453 single-copy core genes found, and sequence identities are reported for these concatenated sequences (Fig. 1B). This tree showed clear distinction between genomes of the two species. Pairwise amino acid sequence comparisons between L. jensenii and L. mulieris core genomes ranged from 95.05 to 95.36% sequence identity. In contrast, the core genome sequence of the 26 genomes that form a clade with the L. jensenii type strain are 99.60% to 100% identical. On the basis of this core genome analysis, we propose that 17 strains should be classified as members of the species L. mulieris (Table S1).

The variation within the L. mulieris clade of the core genome tree is primarily due to one isolate—L. mulieris UMB7784 (10). The core genome of strain UMB7784 is less similar to other L. mulieris strains than they are to each other. Using JSpeciesWS (12), we compared the two species’ type strains to L. mulieris UMB7784. L. mulieris UMB7784 had an ANI value of 87.81% to L. jensenii DSM 20557 and 96.29% to L. mulieris c10Ua161M. Excluding UMB7784, the L. mulieris strains have an average ANI value of 99.66% (Table S2). While UMB7784 is best grouped with L. mulieris, additional sequencing of L. mulieris strains may reveal a third group to which it belongs.

The accessory genome for these strains includes 1,738 gene clusters (Fig. 2), which is larger than the core genome. L. mulieris UMB7784 contains the most genes (n = 83) that are unique to a single genome (singleton genes). Prior genome analyses of lactobacilli found that the genus has an open pangenome (13, 14). Species that colonize diverse habitats and/or coexist with other microbes in large communities, such as many Lactobacillus species, typically have open pangenomes (15). While the majority of comparative genome studies of lactobacilli have focused on those species most relevant to the dairy industry, a recent analysis of L. paragasseri and L. gasseri considered isolates from the human gut microbiota (16). In addition to inhabiting the gut microbiota, L. paragasseri and L. gasseri are common inhabitants of the female urogenital tract, and Zhou et al. (16) found that they also have an open pangenome. Our pangenome analysis of the L. jensenii and L. mulieris strains finds that they too have an open pangenome.

• Open in new tab
• Download powerpoint

FIG 2

Pangenome analysis of 26 L. jensenii (blue) and 17 L. mulieris (pink) genomes. Each ring corresponds to a single genome, and each radial extension in the ring corresponds to a particular gene. If the gene is present in the genome, the color is darker than if it is absent from the genome. The outer ring includes the percentage of the genomes including the gene. The three largest number of singleton genes from the center out are L. jensenii UMB0077, L. jensenii UMB8651, and L. mulieris UMB7784. The inset shows the increase in the pangenome size as new genomes are examined.

Last, we investigated the genetic content that distinguishes these two species. We examined each species individually and identified genes conserved among all genomes of that species that were not present in genomes of the other species. This analysis found 32 genes for L. jensenii (Table S3) and 62 genes for L. mulieris (Table S4). Some L. jensenii genes support the findings of Rocha et al. (7). For instance, L. jensenii can use ribose and trehalose, while L. mulieris cannot; only the L. jensenii genomes encode a ribose transporter and trehalose operon repressor (Table S3). Rocha et al. (7) also noted that only L. jensenii can use arbutin. Our genome analysis identified a L. jensenii locus for disaccharide catabolism. Further investigation would be needed to ascertain whether this is arbutin. We also found new signs of functional specificity, including a conserved respiratory chain pathway present only in L. jensenii (Table S3) and multiple cell surface and secreted proteins in L. mulieris (Table S4). While none of these L. mulieris-specific genes are found in a L. jensenii strain, some of these genes do have homologs within L. psittaci, L. crispatus, and L. paragasseri (based on blastp to the nr database).

While other lactobacilli of the urinary tract have been associated with or without lower urinary tract symptoms (LUTS), L. jensenii has been found in communities regardless of urinary symptoms (17). As our 16S rRNA gene sequence analysis suggests, distinguishing between L. jensenii and L. mulieris by 16S amplicon sequencing surveys is error-prone. Thus, whole-genome sequencing of the microbiota or directed searches for the genes identified here (Tables S3 and S4) is likely the best way to distinguish the two species in a community. With this knowledge, we can determine whether L. jensenii and/or L. mulieris have any association with LUTS. Furthermore, previous studies associating L. jensenii with the benefit of inhibiting the growth of urogenital pathogens must be revisited to assess whether both species have this beneficial role. The distinction of these two species may explain reported phenotypic variation among L. jensenii isolates, including H₂O₂ production (18, 19). Twenty-three of the 43 genomes examined here are from our own collection of urinary isolates, and none of our 6 L. mulieris strains were isolated from women without LUTS (Table S1).

Recently, the entire Lactobacillus genus has been reevaluated given the availability of numerous complete and draft genome sequences (20, 21). Our phylogenetic and phylogenomic analyses support the reclassification of the eight L. jensenii genomes identified by Rocha et al. (7) and eight additional genomes into the new L. mulieris species. Furthermore, we found that the L. mulieris UMB7784 genome is distinct from other L. mulieris strains. Looking at the 83 genes unique to this strain (Table S5), we found homologs in other lactobacilli, including L. salivarius, L. crispatus, L. hominis, L. johnsonii, and L. psittaci. UMB7784-specific genes include genes that encode the following metabolism-related proteins or pathways: glucose phosphotransferase system (PTS), arginase, glucocerebrosidase, diaminopropionate ammonia-lyase, and a phospholipase and a phosphate-selective porin. The UMB7784-specific genes also include genes encoding proteins related to survival: dTDP-4-dehydrorhamnose 3,5-epimerase, a bacteriocin, and a protein G-related albumin-binding molecule. L. mulieris UMB7784 suggests that the urogenital tract may contain additional lactobacilli subspecies or species. Further investigation of lactobacilli from the urogenital tract will provide critical insight into the genomic diversity of these two species and potential associations with urogenital symptoms and infections.