Niche-specific adaptation of Lactobacillus helveticus strains isolated from malt whisky and dairy fermentations

Abstract

Lactobacillus helveticus is a well characterized lactobacillus for dairy fermentations that is also found in malt whisky fermentations. The two environments contain considerable differences related to microbial growth, including the presence of different growth inhibitors and nutrients. The present study characterized L. helveticus strains originating from dairy fermentations (called milk strains hereafter) and malt whisky fermentations (called whisky strains hereafter) by in vitro phenotypic tests and comparative genomics. The whisky strains can tolerate ethanol more than the milk strains, whereas the milk strains can tolerate lysozyme and lactoferrin more than the whisky strains. Several plant-origin carbohydrates, including cellobiose, maltose, sucrose, fructooligosaccharide and salicin, were generally metabolized only by the whisky strains, whereas milk-derived carbohydrates, i.e. lactose and galactose, were metabolized only by the milk strains. Milk fermentation properties also distinguished the two groups. The general genomic characteristics, including genomic size, number of coding sequences and average nucleotide identity values, differentiated the two groups. The observed differences in carbohydrate metabolic properties between the two groups correlated with the presence of intact specific enzymes in glycoside hydrolase (GH) families GH1, GH4, GH13, GH32 and GH65. Several GHs in the milk strains were inactive due to the presence of stop codon(s) in genes encoding the GHs, and the inactivation patterns of the genes encoding specific enzymes assigned to GH1 in the milk strains suggested a possible diversification manner of L. helveticus strains. The present study has demonstrated how L. helveticus strains have adapted to their habitats.

Keywords: L. helveticus; carbohydrate metabolism; dairy fermentation; glycoside hydrolases; stress tolerance; whisky fermentation.

Fig. 1.

Ethanol tolerance (2.5, 5, 7.5 or 10 %) (a), lysozyme tolerance (1, 10 or 100 μg ml⁻¹) (b), and lactoferrin tolerance (0.1, 1 or 10 mg ml⁻¹) (c) of the whisky strains (n=5) and the milk strains (n=6). Relative growth ratios in mMRS broth supplemented with chemicals against the growth in mMRS broth without supplements after 24 and 48 h of incubation were measured. The line within each box represents the median, with the lower line as the 25 % border and the upper line as the 75 % border. The end of the upper vertical line represents the maximum data value; outliers were not considered. The end of the lower vertical line represents the lowest value; outliers were not considered. The separate dots indicate outliers. These experiments were performed in triplicate. The Mann–Whitney U test was applied to compare stress tolerances between the milk strains and whisky strains.

Fig. 2.

Genome sizes and numbers of CDSs in the whisky strains (n=5), all milk strains (n=14) and draft-sequenced milk strains (n=6). The line within each box represents the median, with the lower line as the 25 % border and the upper line as the 75 % border. The end of the upper vertical line represents the maximum data value; outliers were not considered. The end of the lower vertical line represents the lowest value; outliers were not considered. The separate dots indicate outliers. The Mann–Whitney U test was applied to compare between the whisky strains and all milk strains, and between the whisky strains and the draft-sequenced milk strains.

Fig. 3.

ANI values and ANI-value-based hierarchical clustering of L. helveticus strains. The strain names of the whisky strains are indicated in blue and those of the milk strains are in orange.

Fig. 4.

Phylogenetic relationships among 19 strains of L. helveticus based on multiple alignments of the 755 conserved genes and number of GH family proteins found in the strains. L. acidophilus NCFM was used as an outgroup in the phylogenetic analysis. Bootstrap values over 90 % are indicated in the tree. Strain names of the whisky strains are indicated in blue and those of the milk strains are in orange. The scale bar means substitution per site.

Fig. 5.

Comparison of gene numbers in each COG class for the whisky strains and milk strains. Bars and error bars indicate means and sd, respectively. The Mann–Whitney U test was used to compare between the whisky strains and all milk strains, between the whisky strains and the draft-sequenced milk strains, and between the draft-sequenced milk strains and all milk strains. Different letters on top of the bars indicate significant differences. Class names shown in bold are the classes significantly different between the milk strains and whisky strains.

Fig. 6.

Phylogenetic relationships of GH1 enzymes found in L. helveticus strains. Locus tags of proteins and strain names are shown. When ranges of locus tags are shown, amino acid sequences of the genes were concatenated and used for multiple alignments. Strain names of the whisky strains are indicated in blue and those of the milk strains are in orange. Reference GH1 enzymes in L. acidophilus NCFM are shown in red. The GH2 enzyme of L. helveticus JCM 1120^T (LHEJCM1120_13990) was used as an outgroup. Bootstrap percentages above 90% are indicated at branching points. The scale bar means substitution per site.

Fig. 7.

Phylogenetic relationships of GH13 enzymes found in L. helveticus strains. Locus tags of the proteins and strain names are shown. Strain names of the whisky strains are indicated in blue and those of the milk strains are in orange. Reference GH13 enzymes in L. acidophilus NCFM are shown in red. GH2 enzyme of L. helveticus JCM 1120^T (LHEJCM1120_13990) was used as an outgroup. Bootstrap percentages above 90% are indicated at branching points. The scale bar means substitution per site.

Fig. 8.

Phylogenetic relationships and sequences around the NDPNG motif of GH32 enzymes found in L. helveticus strains. Locus tags of proteins and strain names are shown. Strain names of the whisky strains are indicated in blue and those of the milk strains are in orange. Reference GH32 enzymes, i.e. S6PH in L. gasseri 224-1 and β-fructofuranosidase in L. acidophilus NCFM, are shown in red. GH2 enzyme of L. helveticus JCM 1120^T (LHEJCM1120_13990) was used as an outgroup. Bootstrap percentages above 90 % are indicated at branching points. Amino acid sequences around the NDPNG motif (red boxed) region are shown. The scale bar means substitution per site.

Fig. 9.

Gene arrangements of PUL-GH2-GH42 (a) and PUL-GH13-GH65 (b) in CNRZ32. Locus tags of proteins in the PULs range from GCA_000422165.1_01875 to GCA_000422165.1_01885, and from GCA_000422165.1_00292 to GCA_000422165.1_00303, respectively.

Fig. 10.

Inactivation patterns of genes encoding GH1-1 (a) and GH1-2 (b) found in the milk strains. Names of the whisky strains are indicated in blue and those of the milk strains are in orange. Dark blue lines indicate genes in each strain and grey lines indicate deletions of the gene in the region. Red lines on the genes and red numbers indicate positions of stop codons based on protein sequences in the whisky strains. Arrowheads indicate insertion of amino acid sequences (14 residues) that were not found in the protein in the whisky strains. The scale bar means substitution per site.