Assessment of beneficial effects and identification of host adaptation-associated genes of Ligilactobacillus salivarius isolated from badgers

Background Ligilactobacillus salivarius has been frequently isolated from the gut microbiota of humans and domesticated animals and has been studied as a candidate probiotic. Badger (Meles meles) is known as a “generalist” species that consumes complex foods and exhibits tolerance and resistance to certain pathogens, which can be partly attributed to the beneficial microbes such as L. salivarius in the gut microbiota. However, our understanding of the beneficial traits and genomic features of badger-originated L. salivarius remains elusive. Results In this study, nine L. salivarius strains were isolated from wild badgers' feces, one of which exhibited good probiotic properties. Complete genomes of the nine L. salivarius strains were generated, and comparative genomic analysis was performed with the publicly available complete genomes of L. salivarius obtained from humans and domesticated animals. The strains originating from badgers harbored a larger genome, a higher number of protein-coding sequences, and functionally annotated genes than those originating from humans and chickens. The pan-genome phylogenetic tree demonstrated that the strains originating from badgers formed a separate clade, and totally 412 gene families (12.6% of the total gene families in the pan-genome) were identified as genes gained by the last common ancestor of the badger group. The badger group harbored significantly more gene families responsible for the degradation of complex carbohydrate substrates and production of polysaccharides than strains from other hosts; many of these were acquired by gene gain events. Conclusions A candidate probiotic and nine L. salivarius complete genomes were obtained from the badgers’ gut microbiome, and several beneficial genes were identified to be specifically present in the badger-originated strains that were gained in the evolution. Our study provides novel insights into the adaptation of L. salivarius to the intestinal habitat of wild badgers and provides valuable strain and genome resources for the development of L. salivarius as a probiotic. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-023-09623-8.


Background
Animals in the wild usually exhibit specific adaptation traits, such as the safe consumption of pathogeninfected, poisonous foods and tolerance and resistance to various diseases and microbial pathogens [1,2]; these traits can be partly attributed to the presence of adequate numbers of beneficial microbes in the gut microbiota [3].For example, many intestinal lactic acid bacteria (LAB) are found to exert beneficial effects on the host in numerous ways and are recognized as an important source of probiotics [4,5].Badger (Meles meles) is an omnivorous animal species widely distributed in Eurasia [6].Badgers are found to have a complex feeding habit and consume many types of plant-derived foods, such as wheat, persimmon, and hawthorn fruits, as well as animal-derived foods, such as annelids, mollusks, amphibians, and reptiles [7].In addition to being a sylvatic repository of zoonotic infectious diseases [6], it is a vital reservoir of health-beneficial probiotics.Stedman et al. isolated and identified multiple LAB strains from badger feces, and identified 40 LAB strains that exhibited significant antimicrobial activities against Mycobacterium smegmatis, a species that serves as an indicator for screening antagonistic bacteria against pathogenic Mycobacterium species [8].They then assessed the therapeutic potential of these LAB strains through in vitro immunobiological assessments and comparative genomic analyses and identified several Lactobacillus and Pediococcus strains that could act as probiotics against infectious diseases in wild animals through various mechanisms, such as modulation of proinflammatory phagocytic responses associated with protection from pathogens [9].
Among LAB species, Ligilactobacillus (L.) salivarius is a widely distributed species that has been identified in a wide range of niches, including the oral cavity, the intestinal tract of humans and several domesticated animals such as pigs and chickens, and fermented foods [10][11][12][13][14].Many members affiliated with L. salivarius isolated from humans are demonstrated to exhibit several beneficial properties, such as inhibiting the growth and reproduction of pathogenic bacteria [5,15,16], alleviating inflammatory responses [17], reducing pathogen adhesion to host cells [18], improving the absorption of calcium in the intestinal tract [19], and prolonging the lifespan of the host [20].Furthermore, multiple L. salivarius-affiliated members are found to exhibit good capability to adapt to the challenging intestinal environments, e.g., resistance to acid and bile stresses, which is the fundamental prerequisite for probiotics to exert their beneficial effects [21].However, majority of the available animal-origin L. salivarius strains are from domesticated animals, such as pigs and chickens [22], while data on strains originating from wild animals are lacking.
Comparative genomics is a powerful approach for identifying modified, acquired, or lost genetic features and understanding how they facilitate the evolution and adaptation of strains to specific environmental niches within the same species [23].Comparative genomic analysis of L. salivarius strains mainly isolated from swine, chickens, and humans identified several host-specific genes, which may help L. salivarius in adapting to its hosts [22,24].However, to the best of our knowledge, no wild animal-originated L. salivarius genome has been studied so far.Given that badgers have adapted to consume complex foods including diverse fruits and animals, and harbor zoonotic pathogens in the gut microbiome but do not exhibit disease symptoms [6,7], the L. salivarius strains colonizing the intestinal tract of badgers may harbor specific genomic features that facilitate host adaptation such as the digestion of complex plant materials, and thus isolation of L. salivarius strains and exploring their genomic contents can probably benefit the L. salivariusbased probiotics development process such as screening of strains with high capability of nutrient conversion and pathogen inhibition and directed strain improvement.Furthermore, most bacterial genome sequencing projects so far have employed short-read-based sequencing technology; the genome sequences assembled from short reads usually cannot unambiguously resolve repeat sequences that may play important biological roles, such as conferring antimicrobial resistance [25,26], and may lack some accessory genes, core genes, and plasmids [27,28].A combination of long-and short-read sequencing technologies can facilitate the assembly process and generate complete and highly accurate genome sequences [29].In this study, the complete genome sequences of nine L. salivarius strains isolated from the feces of badgers were determined, and comparative genomic analysis was performed using 15 complete genomes of L. salivarius available in the NCBI database.The analysis focused on the identification of properties contributing to the adaptation to hosts, particularly badgers.This study not only assessed and enriched the strain and genome reservoirs of L. salivarius but also provided valuable insights into the host specificity and evolution of L. salivarius in different hosts.

Some L. salivarius strains originating from badgers exhibited beneficial properties
Probiotics, such as L. salivarius, have to overcome acidic pH and bile acid stresses in the gastrointestinal environments after entering the mouth before they can colonize the intestinal tract [30].The nine L. salivarius strains in our study exhibited high tolerance to acid stress.Eight of these strains showed a > 96% survival rate at pH 3 even after 4 h of challenge (Fig. 1a), which is the regular period for which food stays in the stomach [31].The L. salivarius strains also exhibited good tolerance to 0.15% ox gall (Fig. 1b).
The anti-Salmonella effects of the L. salivarius strains were then assessed using S. Derby 14T as an indicator strain.All nine L. salivarius strains exhibited antagonistic activity against S. Derby 14T, with strains S32, S35, S99, and S103 showing significantly higher inhibitory activities compared with the other five strains (P < 0.05, oneway ANOVA, Tukey post hoc test) (Fig. 1c).LAB can form a barrier that hinders the colonization and infection of pathogens in the intestine through auto-aggregation [32].Strain S32 exhibited the highest auto-aggregation activity (58.84%), while S35 and S99 ranked to be the second higher group (55.29% and 54.79% for S35 and S99, respectively) (P < 0.05, one-way ANOVA, Tukey post hoc test) (Fig. 1d).Successful adhesion to intestinal epithelial cells is a key step for Salmonella species to colonize the intestine and cause further infection [33].Therefore, controlling the adherence of Salmonella to intestinal epithelial cells is crucial for preventing and managing Salmonella infections.The anti-adhesion ability of the nine L. salivarius strains against the adhesion of S. Derby 14T to intestinal epithelial cells was assessed in this study.A significant difference was observed in the anti-adhesion ability of the nine L. salivarius strains, with strains S40 and S32 reducing the adhesion of S. Derby 14T by 36.29% and 33.53%, respectively, and strains S96 (2.35%) and S99 (0%) exhibiting no obvious anti-adhesion effect (Fig. 1e).Moreover, coaggregation of L. salivarius strains with pathogens can avoid the colonization of intestinal pathogens by preventing pathogens from attaching to epithelial cells.All nine L. salivarius strains exhibited certain coaggregation activities with S. Derby 14T, with S32 exhibiting the highest coaggregation capacity of 15.76%, which was significantly higher than that of the remaining eight strains (P < 0.05, one-way ANOVA, Tukey post hoc test), while S35 and S99 showing the lowest coaggregation capacity of 3.12% and 2.84%, respectively (Fig. 1f ).

General genomic features of L. salivarius strains
Complete genomes of the nine L. salivarius strains were generated using long and short reads.The nine strains harbored one chromosome and one to five plasmids.The genome length was 2.10-2.28Mb, the GC content was 32.80%-32.92%,and 2027-2219 CDSs were present (Table 1).Comparative genomic analyses were performed using the nine generated badger-originated L. salivarius genomes and the 15 complete genomes available in the NCBI (as of March 10 th , 2022) (Table 1) of L. salivarius strains isolated from humans (six strains), swine (four strains), chickens (four strains), and horses (one strain) in order to identify the genomic features specifically present in the strains originating from badgers or in other groups.The ANI values across the Fig. 1 The probiotic property of the nine L. salivarius isolates.a-f Acid tolerance, bile tolerance, anti-Salmonella activity, auto-aggregation abilities, anti-adhesion abilities against S. Derby 14T, and coaggregation abilities of nine L. salivarius isolates, respectively.Note: Data shown are mean SD of triplicate values of independent experiments.The compact letter display indicates significant differences in pairwise comparisons, and strains with different letters are significantly different (P < 0.05, one-way ANOVA, Tukey post hoc test) (c-f) 24 strains were higher than 95%, which is recognized as the species boundary, indicating that the strains were affiliated with the same species.Notably, the strains from the same host exhibited much higher ANI values than those from other hosts, reflecting the host specificity of the L. salivarius strains (Fig. 2a).Furthermore, the genome size (Kruskal-Wallis test, P = 0.001, the horse group that only contained one strain was excluded from the statistical analysis, same herein) (Fig. 2b), number of CDSs (Kruskal-Wallis test, P = 0.004) (Fig. 2c), and GC content (Kruskal-Wallis test, P = 0.003) (Fig. 2d) significantly differed among the strain groups from different hosts.In particular, the genome size and number of CDSs in the strains originating from swine and badgers were significantly higher than those in the strains originating from chickens and humans (Mann-Whitney test, all P < 0.05).Moreover, the strains originating from badgers and swine harbored a significantly higher number of COG-annotated genes compared with those from chickens and humans (Kruskal-Wallis test with Mann-Whitney post hoc test, all P < 0.05) (Fig. 2e).Of the COG-annotated genes, the proportion of genes belonging to the COG categories "metabolism" (Kruskal-Wallis test, P = 0.02), "information storage and processing" (Kruskal-Wallis test, P = 0.002), and "cellular processes and signaling" (Kruskal-Wallis test, P = 0.008) significantly differed among the strains from different hosts (Fig. 2f ).Specifically, the swine-originated strains harbored fewer metabolism-related genes, and higher information storage and processing-related ones compared with the remaining animal hosts; the badger-originated strains harbored significantly fewer cellular processes and signaling related genes compared with those from humans and chickens (Mann-Whitney test, P < 0.05 for both comparisons).

Pan-genome analysis of L. salivarius strains
To identify the genomic regions associated with the host adaptation and/or specificity of L. salivarius strains, comparative genomic analyses were performed using the 24 L. salivarius genomes.The pan-genome contained 3272 orthologous clusters, whereas the core genome contained 1165 orthologous clusters.The pan-genome of the 24 L. salivarius strains was closed, and the pangenes slowly increased and tended to reach a plateau when the genome number reached 21 (Fig. 3a), suggesting that the 24 genomes could comprehensively represent the total gene repertoire of the L. salivarius population from the given hosts.The pan-genome phylogenetic tree constructed based on the gene presence/absence patterns demonstrated that the nine strains originating from badgers differed from each other in terms of the gene contents (Fig. 3b).The three strains, S35, S39, and S40, which were isolated from the same badger individual exhibited relatively conserved genomic contents and formed a cluster in the phylogenetic tree, and S103, S91, and S96 that were isolated from another badger individual also formed a separate cluster; however, the two strains isolated from different individuals, S32 and S92, formed a cluster, and S99 harbored a relatively divergent genomic content compared with the other eight strains.Of note, the badger-originated strains formed a distinct clade from the other strains, clearly suggesting the presence of badger-specific gene families (Fig. 3b).To identify genes that may contribute to the observed host range differences among the strains, gene gain/loss analysis was performed using the GLOOME server.In total, 3413 gene gain events and 392 gene loss events were identified across the phylogenetic tree (Fig. 3b).Notably, while only a minute fraction of gene gain events occurred in the strains originating from badgers, their common ancestor (internal node N16) gained 412 genes but lost only nine genes, suggesting the importance of these gained genes in its adaptation to this wild host.
A total of 270 out of the 412 genes obtained functional annotations, and these COG-annotated genes were mainly assigned to the COG categories "L" (replication,  The gene gain and loss events identified by GLOOME analysis.The red colored numbers labeled on the nodes denote the gain events, while the blue colored numbers denote the gene loss events.The tree was constructed based on the gene presence/absence matrix of the pan-genome, and the scale bar denotes 100 gene differences recombination, and repair; 12.22%), "K" (transcription; 11.85%), "M" (cell wall/membrane/envelope biogenesis; 12.22%), and "G" (carbohydrate transport and metabolism; 11.48%) (Table S1).Of note, more than half of the genes in the category "M" (57.58%) encoded glycosyl transferases (GTs), which are involved in the metabolism and transport of carbohydrate substrates.In total, 51.61% of the genes in the category "G" encoded phosphoenolpyruvate-phosphotransferase systems (PEP-PTSs), which are associated with transporters responsible for the uptake of many types of carbohydrate nutrients, including sucrose, fructose, and glucose.We speculated that these gained functional genes could contribute to the adaptation of L. salivarius to the intestinal environment of badgers, in which diverse glycan-rich resources are present because of the complex food consumption behavior of badger individuals in the wild.
Positive selection is known as an important driving force for microbes in the niche adaptation process [34].We further identified the positive selection-affected genes among the badger clade-gained genes using BUSTED [35], and 14 genes were found to be affected by positive selection (P < 0.05).Interestingly, two GTencoding genes among the gained genes that belonged to the COG category "M" were under positive selection, indicative of the importance of the gained GT-encoding genes for the adaptation of L. salivarius to wild badgers.Furthermore, cas9, whose product is a crucial component in the IIA system and is responsible for recognizing and cleaving subsequent invading nucleic acids with sequences identical to that of the spacer [36], was found to be gained by the badger clade and affected by positive selection.

CAZymes
Carbohydrate utilization-associated genes were then annotated from the L. salivarius genomes using dbCAN2, and one auxiliary activity (AA) family, three carbohydrate esterase (CE) families, six carbohydrate-binding module (CBM) families, 26 glycoside hydrolase (GH) families, and 18 GT families were identified.Genes belonging to GHs and GTs were the most abundant gene families in the 24 L. salivarius strains.Notably, the number of annotated CAZyme-encoding genes significantly differed among the strains from different hosts.The strains originating from badgers harbored the highest number of CAZyme-encoding genes, which was significantly higher than that of the human and chicken groups (Fig. 4a) (Kruskal-Wallis test with Mann-Whitney post hoc test, both P < 0.01).GHs are responsible for catalyzing the hydrolysis of glycosidic bonds between carbohydrates or carbohydrates and non-carbohydrates [37].GH13, GH23, and GH25 were the most abundant families in 24 genomes.Interestingly, the amounts of GH13-, GH109-, and GH73-encoding genes significantly varied among the strains from different hosts (Kruskal-Wallis test, all P < 0.05).While greater amounts of GH13-and GH109encoding genes were identified in the strains from badgers than in those from other hosts, a significantly lower amount of GH73-encoding genes was identified in the strains from badgers than in those from other hosts (Fig. 4b) (all P < 0.01, Mann-Whitney test).
GTs are enzymes that catalyze the formation of various glycoconjugates [38].Of these, GT2, GT4, GT8, and GT113 represent the most predominant GT family members in the 24 L. salivarius strains (Fig. 4c).Notably, the number of GT-encoding genes significantly differed among the strains from different hosts (Kruskal-Wallis test, P = 0.001), with the strains from badgers harboring the highest number of GTs (all P < 0.01, Mann-Whitney test).In particular, the amounts of GT2-, GT8-, and GT113-encoding genes significantly differed among the strains from different hosts (Kruskal-Wallis test, all P < 0.01), with the strains from badgers harboring the highest amount of all three gene families (all P < 0.05, Mann-Whitney test).

MGEs and the CRISPR-Cas system
MGEs, such as plasmids and prophages, play a prominent role in shaping a strain's genome [39] and serve as an important driving force in the evolution of organisms through inter-organism exchanges [40].Zero (strain DJ-sa-01) to six plasmids were identified from the 24 genomes (Fig. 5).Although the number of plasmids did not vary much among the strains from different hosts (on average, 2-3.5 plasmids per group), the total length of plasmids in the strains from badgers was significantly longer than that in the strains from humans and chickens [346.71 ± 53.53 kb (mean ± SD), 258.40 ± 49.54 kb, and 230.41 ± 159.97 kb for the strains from badgers, humans, and chickens, respectively] (Mann-Whitney test, P < 0.05 for both badgers vs. human and badgers vs. chicken strain  The plasmid sequences for the strains were grouped into clusters using the mob_type method implemented in mob-suite, and the clusters were named with the primary_cluster_id values deposited in the mob-suite database.The number of CDSs predicted in the plasmid sequences of each strain was also shown, and different letters denote significant differences (P < 0.05, Kruskal-Wallis test with Mann-Whitney post hoc test).Middle panel, the colored cells denote presence of each of the prophage clusters.The prophage sequences were grouped into clusters using cd-hit-est with parameter -c 0.9.Lower panel, the colored cells denote presence of each of the three CRISPR-Cas systems comparisons).However, the total length of plasmids in the strains from badgers did not differ significantly from that in the strains from swine (377.21 ± 97.05 kb) (Mann-Whitney test, P > 0.05).The number of plasmidoriginating CDSs in the strains from badgers was also significantly higher than that in the strains from humans and chickens (Mann-Whitney test, both P < 0.05) and did not differ dramatically from that in the strains from swine (P > 0.05).The 65 plasmid sequences from the 23 strains (DJ-sa-01 did not harbor plasmid) were grouped into 11 plasmid clusters, with cluster "AF479" observed in the majority of the 23 strains and cluster "AF075" identified in eight out of the nine badge-originated strains but rarely observed in strains originating from other hosts.Of the 412 gene families gained by the badger clade, 110 families originated from plasmids, which included several carbohydrate utilization-associated genes.Among them, an iol operon that is involved in the utilization of the important plant metabolite myoinositol (MI), including iolRABCDG2G3E, was identified in the plasmids belonging to cluster "AF479" of the nine strains from badgers but was absent in the other strains, except for strain IBB3154, a chicken isolate.The iol operon in the nine badger-originated strains was located in a genomic region affected by horizontal gene transfer (HGT) as revealed by the Alien Hunter analysis (Fig. S1).
In total, 52 prophage sequences were identified from the 24 L. salivarius genomes, ranging from zero (strain DJ-sa-01) to four prophages identified in the 24 genomes (Fig. 5).The number of prophage sequences significantly varied among the strains from different hosts (Kruskal-Wallis test, P = 0.018).The strains from swine (3.75 ± 0.5) were found to harbor significantly more prophage sequences than those from chickens (1.25 ± 1.26), humans (2.17 ± 0.98), and badgers (1.89 ± 0.78) (Mann-Whitney test, all P < 0.05), while the number of prophages in the three groups did not differ significantly among each others (P > 0.05).These prophage sequences were then clustered into 32 prophage groups (Fig. 5).Notably, prophage cluster 4 was present in all nine strains originating from badgers and cluster 16 was present in seven of the nine strains; however, the two prophage clusters were absent in the remaining 15 strains (Fig. 5), indicating that these two prophages were specifically acquired by the last common ancestor of the badger population after it was divergent from that of the remaining 15 strains.The remaining prophage clusters were found to be strain-specific and highly diverse, suggesting that the prophage acquisition events occurred independently after these strains diverged from their last common ancestor.In total, 83 genes were predicted from the prophage sequences among the 412 genes gained by the common ancestor of badgers; these included certain genes that were beneficial to the bacterial host.For example, a UDP-N-acetyl-Dmannosamine dehydrogenase-encoding gene, which is involved in the biosynthesis of UDP-acetamido sugars, was identified.Furthermore, a cysteine protease-encoding gene, which plays a major role in the degradation of proteins into peptides, amino acids, ammonia, and other forms of nonprotein nitrogen [41], was identified.
The CRISPR-Cas system confers adaptive immunity to host bacteria to resist the insertion of foreign genes, such as phages and plasmids [42].The CRISPR-Cas system was identified in 79.17% (19/24) of the strains (Fig. 5).Three CRISPR-Cas system types (types I, II, and III) were detected, and one subtype (IE, IIA, and IIIA) was identified for each type.Only one out of the four swineoriginated strains was found to harbor the CRISPR-Cas system, which was consistent with the observation that the strains originating from swine harbored a significantly higher number of prophages, as mentioned above.Notably, the type IIA system was identified in 14 out of the 24 strains, and all nine badger-originated strains and four out of six human-originated strains harbored the type IIA system.Furthermore, two genes involved in the CRISPR-Cas system, including cas2 and cas9, were found to be gained by the badger group and affected by positive selection.Spacers are small fragments of foreign DNA incorporated into CRISPR loci to confer adaptive immunity to the host [43].In total, 412 spacer sequences were obtained from all CRISPR loci of the 24 strains, and 0-89 spacers were identified in each isolate.The number of spacers harbored by the strains did not exhibit obvious host-specific patterns but instead exhibited highly isolate-specific patterns.The spacers that shared identical sequences were defined as a spacer operational taxonomic unit (OTU); 290 OTUs were obtained (Table S2).Of the 290 OTUs, 78.97% (229/290) contained only one spacer sequence.Of the remaining 61 OTUs, only 25 OTUs contained spacers derived from more than two different strains from the same host, badgers.Furthermore, these OTUs grouped the nine badger-originated strains into two clusters (Fig. S2 and Table S2).In cluster 1, four strains from different badger individuals (S91, S92, and S99 from the same badger individual and S32 from another badger individual) shared identical spacer distribution patterns, and S35 exhibited different spacer contents from the other four strains.In cluster 2, S39 and S40, which were isolated from the same badger individual harbored identical spacer contents, and S96 and S103 that were isolated from another individual harbored the same spacer contents (Fig. S2).Subsequently, spacers from all CRISPR loci were extracted and further mapped to the 52 predicted prophage sequences, and only 26 of the 290 OTUs were paired with the prophage sequences (Table S3).The majority of spacers from badgers, as well as spacers from other hosts, did not match their prophage contents, indicating the presence of highly diverse foreign L. salivarius phages in the intestinal environments.

Discussion
LAB are gut commensals that contribute to the maintenance of gut homeostasis by maintaining a beneficial microbial balance [9].Of the LAB species, L. salivarius is an important taxon in the intestinal tracts of mammals.Many LAB species have exhibited promising probiotic properties, such as preventing and controlling infectious diseases [5,18], modulating the immune system [17,44], and enhancing the quality of livestock products [45].Compared with domesticated animals, the gut microbiota of wild animals probably contains certain probiotic members that can balance the negative effects of the consumed pathogen-infected, poisonous foods and resist the invasion of pathogens, thereby maintaining host health [3].Therefore, wild animals can serve as a promising reservoir to identify probiotic bacteria, such as those affiliated with L. salivarius.In this study, all nine L. salivarius strains originating from badgers exhibited good intestinal adaptation ability, as revealed by acid and bile tolerance assays and auto-aggregation assays.Furthermore, several strains were effective in controlling the negative effects of intestinal bacterial pathogens on the hosts by reducing the population of pathogens in the intestinal environment through their antagonistic activities (anti-Salmonella assay) and coaggregation activities and by preventing the adhesion of pathogens to epithelial cells, which is the first step in pathogen infection [46] (Fig. 1).The aforementioned beneficial effects varied among the nine strains.Overall, strain S32 exhibited good probiotic properties and may thus have an application prospect for controlling infective pathogens, such as Salmonella, in animal breeding and raising practices.
Given that badger individuals exhibit a complex feeding habit and consume many types of plant-derived foods as well as other animal-derived foods in the wild [7], which probably contain more complex carbohydrate and other nutrients compared with that are consumed by domesticated animals and humans, the genomes of L. salivarius strains originating from badgers were compared with those originating from domesticated animals and humans, aiming to reveal potential beneficial genes in badger-originated L. salivarius strains, such as those associated with complex carbohydrate digestion.The whole genome sequences of the nine L. salivarius strains originating from badgers were obtained based on a combination of long and short reads and were a significant addition to the number of genomes available in the public database.Certain bacterial species that are widely distributed in various environments usually have an open pan-genome, in which new genes can confer beneficial effects for environmental adaptation [47].On the other hand, species that tend to occupy limited habitats frequently have a closed pan-genome because they do not need to respond to many environmental changes [48,49].Our results demonstrated that L. salivarius species harbor a closed pan-genome, similar to that in L. ruminis [50].Previous analyses of Lactobacillus species, such as L. reuteri [51] and L. ruminis [50], consistently suggested that strains from different hosts usually form separate phylogenetic clusters, which are highly reflective of the host source.Notably, the strains originating from badgers harbored a relatively larger genome size and more CDSs and formed a distinct clade from the strains originating from other hosts.This finding indicates that the presence of clade-specific genes contributes to the adaptation of L. salivarius to wild badgers, which has a complex feeding habit.
The degradation of complex carbohydrate substrates, such as starch, cellulose, and MI, to simple sugars, is crucial for the metabolism of the bacteria inhabiting the intestine and the hosts and is mainly catalyzed by carbohydrate utilization-associated enzymes [52].In this study, L. salivarius strains from the same host showed highly consistent carbohydrate metabolism behavior, as revealed by CAZyme analysis.Five major CAZymes (AAs, GHs, GTs, CEs, and CBMs) were annotated in the 24 strains.GHs and GTs were the most abundant gene families in L. salivarius, with GT2, GT4, GH13, and CBM50 being the largest families, which is consistent with the results in other Lactobacillus species [43].The diet of the host is a key evolutionary force that shapes the gut microbiota and influences the evolutionary trend of gut symbionts [53].Notably, significantly lower numbers of GH73-encoding genes and higher numbers of GH13and GH109-encoding genes were noted in the strains from badgers than in those from other hosts (Fig. 4).GH13 and GH109 are known to be crucial for the degradation of several carbohydrates, such as starch, sucrose, and O-glycan [54].In contrast, GH73 is an important enzyme family involved in the degradation of beta-glucan [54], which is an abundant ingredient in cereals [55].These findings collectively reflect the difference in food sources between wild badgers and domesticated animals and humans.Gene gain events were suggested to play an important role in the acquisition of complex carbohydrate degradation-associated genes in the evolutionary history of strains from badgers (Table S1).For instance, the iol operon, which is involved in the utilization of the important plant metabolite MI, is not widely distributed in the Lactobacillus genus [56]; however, all nine strains from badgers harbored the plasmid-derived cargo genes iolRABCDG2G3E, which were probably acquired by these strains via HGT and can benefit the bacteria and the host by digesting the plant-derived MI [56].GT2 and GT8 families are involved in the biosynthesis of exopolysaccharides and capsular polysaccharides, respectively [57].Polysaccharide production is known to be an important beneficial trait of probiotic Lactobacillus species [58].The higher abundance of GT2 and GT8 families in the strains from badgers may benefit the host by regulating gut microbiota homeostasis through the production of polysaccharides [59].Besides plasmids, prophages are an important source of the gene gain events which play a critical role in bacterial population evolution [49].Most of the prophage sequences identified in these badgeroriginated strains did not show high similarity to other analyzed L. salivarius genomes (Fig. 5) and genomes deposited in the NCBI database, these sequences are probably acquired from the microbial members in the badger gut microbiome, which has not been explored yet.Several cargo genes that were probably beneficial for L. salivarius to adapt to the badger host, such as the UDP-acetamido sugar biosynthesis-associated gene and cysteine protease-encoding gene, were identified in the prophages specifically harbored by the badger-originated strains.UDP-acetamido sugar biosynthesis-associated gene is considered to be important for bacteria to adapt to extreme environments [60].The cysteine proteaseencoding gene, whose product plays a major role in the degradation of proteins into peptides, amino acids, ammonia, and other forms of nonprotein nitrogen [41], could benefit the badger host by contributing to the utilization of protein contents in the consumed complex foods.Overall, the genomes of L. salivarius strains originating from badgers serve as valuable resources to identify probiotic-associated genes, such as those involved in complex carbohydrate substrate degradation and exopolysaccharide production.The CRISPR-Cas system serves as an immune system to prevent the bacterial host from infections caused by foreign DNA, such as phages [42].In our study, 79.17% (19/24) of the L. salivarius strains contained at least one CRISPR-Cas system (Fig. 5).Notably, the L. salivarius strains that lacked the CRISPR-Cas system contained more prophages than those possessing the CRISPR-Cas system; a similar result has been reported in L. ruminis [50], highlighting the important role of the CRISPR-Cas system in preventing the invasion of exogenous DNA in Lactobacillus species.The type II CRISPR-Cas system could play a prominent role in preventing the invasion of foreign DNA during the evolution of L. salivarius, particularly for the badger group (all nine strains containing the IIA system).The presence of highly diverse spacers in the L. salivarius genomes and the fact that the cas9 gene (which encodes the crucial component of the type II system and is responsible for recognizing and cleaving subsequent invading nucleic acids with sequences identical to that of the spacer [36]) was under positive selection, collectively indicate the presence of fierce competitions and coevolution between the phages and L. salivarius strains.

Conclusions
Overall, we obtained nine L. salivarius strains from the fecal samples of wild badgers individuals with certain probiotic traits and identified an isolate exhibiting overall good performance.Genome sequencing and comparative genomic analysis demonstrated that the strains originating from badgers harbored a larger genome size and a higher number of CDSs than those originating from chickens, humans, and horses and had significantly more genes responsible for the degradation of complex carbohydrate substrates.To some extent, these differences reflect the niche adaptation of L. salivarius in badgers through gene acquisition.However, our study also has certain limitations raised by the small number of strains from each host species (1 to 9 strains) due to the limited amount of complete L. salivarius genomes available in the NCBI database, and further large-scale comparative genomics analysis is needed in future studies to reveal more comprehensive results.Nevertheless, our findings are an important addition to the L. salivarius database and provide new insights into the adaption of L. salivarius to distinct hosts from the perspective of evolution, as well as contribute to the probiotic development practices.

Sampling collection
Fresh fecal samples from eight wild badgers individuals were collected from Shanghai, China (30.92 N, 121.46 E), on August 20 th , 2020.Badgers usually defecate in their fixed locations from midnight to early morning.Only the central inner layer of the feces was collected using aseptic sticks to avoid environmental contamination.The samples were collected into 50 mL sterile tubes and immediately transported to the laboratory on ice.

Isolation of LAB from fecal samples
In total, 20 g of each fecal sample was homogenized in phosphate-buffered saline (PBS) to prepare serial dilutions.Specifically, the homogenized fecal samples were diluted to 10 −3 , 10 −4 , and 10 −5 , and 200 μL of each dilution was coated onto MRS agar plates (Guangdong Huankai Microbial Technology Co., Ltd, Guangdong, China) for LAB isolation.After 24 h of incubation at 37°C, different types of colonies were selected for further purification based on the colony size, morphology, color, and surface roughness.Purified single colonies were subjected to Gram staining and microscope-based cell morphology determination.The checked purified strains were stored at − 80°C using 25% glycerin as the cryoprotective agent.

Taxonomic identification of LAB by 16S rRNA gene sequencing
DNA was extracted from each isolate using a DNA extraction kit (Tiangen Biotech Co. Ltd., Beijing, China), according to the manufacturer's protocol.The DNA concentration and quality were assessed using NanoDrop (Thermo Scientific, USA).The 16S rRNA gene was amplified using the universal primer set 27F (5′-AGA GTT TGA TCC TGG CTC AG-3′) and 1492R (5′-GGY TAC CTT GTT ACG ACT T-3′) [61].The PCR products were further verified by 1% agarose gel electrophoresis, and those with correct size (approximately 1,500 bp) were subjected to 16S rDNA sequencing (Genscript, Nanjing, China).The taxonomic affiliation of these LAB strains was determined based on 16S rRNA gene data of type strains deposited in EzBioCloud database [62].Of the 91 strains obtained from badger feces, nine strains identified as L. salivarius were selected for further experiments and genome sequencing.The nine strains were isolated from two individuals, while strains S32, S35, S39, and S40 were isolated from the same individual, S91, S92, S96, S99, and S103 were from another individual.

Antagonistic activities of L. salivarius strains against Salmonella species
The anti-Salmonella effects of L. salivarius strains were determined using the agar spot test, as described by Schillinger [63], with a pathogenic strain S. Derby 14T originating from swine serving as an indicator [64].After incubation in MRS liquid medium at 37°C for 24 h, 10 µL of each L. salivarius culture was inoculated in the center of MRS agar plates, air dried, and incubated at 37°C for an additional 24 h.Following this, 10 µL of overnight grown S. Derby 14T culture was added to Luria-Bertani (LB) semi-solid medium and mixed well.The semi-solid medium was then spread on MRS agar plates containing LAB until coagulation.The plates were incubated at 37°C for 24 h, and the diameters of inhibition zones were then measured by subtracting the diameter of the L. salivarius colony from the diameter of the measured inhibition zone.
The data from different strains were standardized using the Z-score method.The Z-score was calculated using the formula z = (x − µ)/σ, where x is the diameter of the inhibition zone of each L. salivarius isolate; μ denotes the mean value of the overall inhibition zone diameter; and σ is the standard deviation of the overall inhibition zones.

The anti-adhesion ability of L. salivarius strains against the adhesion of S. Derby 14T
The mouse colon cancer epithelial cell line MC38 was used in this assay, and it was purchased from Hunan Fenghui Biotechnology Co., Ltd.(Hunan, China) (catalog no.CL0203).The cells were seeded into 24-well plates to a concentration of 4 × 10 5 cells per well and were cultured overnight at 37°C in the presence of 5% CO 2 .Then, 1 × 10 7 L. salivarius cells were added into each well and incubated for 1 h.The cells were then washed with PBS and transferred to a new medium.S. Derby 14T cells were added to L. salivarius-treated and untreated MC38 cells at a multiplicity of infection (MOI) of 20:1 and incubated for 1 h.The cultured cells were washed using PBS thrice and then solubilized with 0.1% Triton X-100 for 5 min.The solution was then serially diluted, and 100 μL of the 10 −3 and 10 −4 diluted samples were evenly coated onto preconfigured sterile LB plates.The plates were incubated at 37°C for 12-16 h to calculate the number of adhered S. Derby 14T strains.Three replicates were performed for each treatment.The adhesion rate reduction ratio (%) of S. Derby 14T cells to MC38 cells in the presence of L. salivarius strains was calculated as follows: [(the number of S. Derby 14T cells adhered in the absence of L. salivarius strains − the number of S. Derby 14T cells adhered in the presence of L. salivarius strains)/ the number of S. Derby 14T cells adhered in the absence of L. salivarius strains] × 100%.

Acid and bile tolerance assays of L. salivarius strains
The nine L. salivarius strains were grown in 5 mL MRS broth and anaerobically incubated at 37°C for 24 h.The cells were then centrifuged at 4,000 rpm for 10 min, and the obtained bacterial pellet was washed thrice and resuspended to a concentration of OD 600 0.8 in sterile PBS.Then, the pH of the bacterial suspensions was adjusted to 3 using 5M HCl, and these suspensions were incubated at 37°C for 0, 2, and 4 h, respectively.Following this, 100 μL aliquots of the bacterial suspensions were serially diluted, and 100 μL of the diluted samples were spread on MRS agar plates and anaerobically incubated at 37°C for 24 h.The number of surviving cells was calculated using the plate counting method [65].Three replicates were performed for each L. salivarius isolate.
For the bile tolerance assay, L. salivarius strains were cultured in MRS broth containing 0.15% ox gall.The culture without ox gall was used as the control.The cultures were anaerobically incubated at 37°C for 0, 2, and 4 h.The number of surviving cells was calculated using the plate counting method as mentioned above.Three replicates were performed for each L. salivarius isolate.
The tolerance rate of L. salivarius strains to acidic pH and bile (%) was calculated using the following formula: [the number of viable bacteria at 2 and 4 h (CFU/mL)/the number of viable bacteria at 0 h (CFU/mL)] × 100%.

Auto-aggregation ability
The auto-aggregation ability of L. salivarius strains was assessed as described previously, with slight modification [66,67].The L. salivarius strains were grown in MRS broth and anaerobically incubated at 37°C for 24 h.The cells were then centrifuged at 4,000 rpm for 10 min, and the bacterial pellets were washed thrice and resuspended to a concentration of OD 600 0.8 in sterile PBS.Following this, 1 mL samples of the bacterial suspensions were well mixed by vertexing, and then were incubated at 37°C without shaking for 5 h. 100 μL of the upper suspension was carefully recovered from the bacterial suspensions and transferred to microtiter plates.Its absorbance was measured using a spectrophotometer (Eppendorf SE Hamburg, Germany) at 600 nm.Three replicates were performed for each L. salivarius isolate.The auto-aggregation rate (%) was calculated using the following formula: 1 − (A t /A 0 ) × 100%, where A t is the absorbance at time t = 5 h and A 0 is the absorbance at time t = 0 h.

Coaggregation ability with Salmonella strains
The coaggregation ability of L. salivarius strains with Salmonella strains was assessed as reported previously, with slight modification [68].Overnight cultures of L. salivarius and S. Derby 14T were collected by centrifugation at 13,000 rpm for 5 min, and the precipitate was washed thrice and resuspended in sterile PBS.L. salivarius and S. Derby 14T suspensions were mixed in equal volumes, vortexed for 5 min, and incubated at room temperature [69].The upper part of the mixture was carefully removed, and its OD 600 value was measured at 5 h.Single suspensions of each strain were considered as controls.Three replicates were performed for each L. salivarius isolate.The coaggregation rate (%) was calculated using the following formula: [(A x + A y )/2 − A (x + y) ]/ (A x + A y )/2 × 100%, where A x , A y, and A (x + y) represent the OD 600 values of S. Derby 14T strains, L. salivarius strains, and the S. Derby 14T and L. salivarius mixture, respectively.

Genomic DNA extraction, whole genome sequencing, and genome assembly
The genomic DNA of nine L. salivarius strains was extracted using the QIAamp ® PowerFecal ® DNA Kit (Qiagen, Germany), according to the manufacturer's protocol.The DNA concentration and integrity were assessed using NanoDrop and 0.8% agarose gel electrophoresis, respectively.Then, the DNA concentration was measured using Qubit 4.0 (Thermo Fisher Scientific, USA) to ensure that it satisfied the library preparation.
Whole-genome sequencing was performed on an Illumina HiSeq platform and a Nanopore platform to generate both short and long reads.High-quality long reads (Q > 7 and length > 1 kb) were assembled using Flye (v2.8.3) [70].The generated assemblies were then polished using NextPolish (v2) [71], with the short reads being used as inputs.The genomic sequences of the nine L. salivarius isolates have been deposited in the NCBI database under the accession numbers CP114503-CP114529 (Table 1).

Comparative genomics analysis
Fifteen completely sequenced L. salivarius genome sequences available in the NCBI database (as of March 10 th , 2022) were downloaded, and the nine de novo assembled genomes and 15 downloaded genomes were simultaneously annotated using Prokka (v1.14.6) with default parameters [72].Following this, the Clusters of Orthologous Groups of proteins (COG) annotations were generated using eggNOG-mapper (v2.1.7)with parameters -itype proteins and -m diamond based on the eggNOG orthology database (v5.0.2) [73].The carbohydrate-active enzyme (CAZyme)-encoding genes in the L. salivarius genomes were annotated using dbCAN (v2) with parameter -t all, and the Prokka-generated protein sequences were used as inputs [74].The CRISPR-Cas systems and spacer sequences were identified using CRISPRone with default parameters (https:// omics.infor matics.india na.edu/ CRISP Rone/ check.php? id= K4KfC Lvy) [75].Moreover, prophages in the L. salivarius genomes were identified using ProphET with default parameters, and the genome sequences and the Prokkagenerated gff files were used as inputs in the analysis [76].The information of the identified prophages in the genomes were listed in Table S5.The prophage sequences were clustered to groups using cd-hit-est with parameter -c 0.9 [49,77].The plasmids were clustered to groups using the mob_typer method implemented in mob-suite with default parameters [78].The acquired genomic contents in the genome sequences via horizontal transfer were predicted using Alien Hunter with default parameters [79].
The average nucleotide identity (ANI) values between the L. salivarius strains were calculated according to the methods recommended in FastANI (v1.32) with default parameters [80].The pan-genome and core genome of the 24 strains were constructed using the OMCL algorithm implemented in the get_homologues package (v 3.3.2) with parameters -M -S 75 -t 0 -e [81].The pan-genome size, as well as the fitted models of the pan-genome and core genome, were determined using the distance guide algorithm implemented in PanGP (v1.0.1) based on the generated orthologous groups by the get_homologues analysis [82].The fitted models of the pan-genome and core genome profiles were depicted as y1 = Ax B + C and y2 = De Ex + F, respectively, where x is the genome number, y1 is the pan-genome size, y2 is the core genome size, and A, B, C, D, E, and F are fitted parameters.The phylogenetic signal-informative core genes were identified using the get_phylomarkers package [83] and were concatenated using Gblocksb0.91[84].The phylogenetic tree was then constructed using IQ-tree [85].The gene gain/loss analysis was performed using GLOOME with default parameters, and the presence/absence matrix of the gene families and the phylogenetic tree served as inputs, and the probability of gain/loss events (PP ≥ 0.8) was estimated using the stochastic mapping method [86].The results were visualized using iTol [87].BUSTED [35] in the HYPHY2.5.31 package [88] was used to identify the orthologous groups affected by positive selection (P < 0.05), and the analysis was performed with default parameters.
• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year

•
At BMC, research is always in progress.

Learn more biomedcentral.com/submissions
Ready to submit your research Ready to submit your research ?Choose BMC and benefit from: ? Choose BMC and benefit from:

Fig. 2 Fig. 3 a
Fig. 2 The genomic characteristics of 24 L. salivarius strains from different hosts.a Heatmap of ANI values based on the sequences of 24 L. salivarius strains.b-f The comparison of genome size, CDSs, GC content, COG-annotated genes, and COG categories of L. salivarius strains from different hosts, respectively.Kruskal-Wallis test with Mann-Whitney post hoc test was used for comparisons (b-f).The compact letter display indicates significant differences in pairwise comparisons, and groups with different letters are significantly different (P < 0.05).The horse group which contains only one strain was excluded from the statistical anaysis.*, P < 0.05; **; P < 0.01; ***, P < 0.001 as revealed by Kruskal-Wallis test (f)

Fig. 4 Fig. 4 (
Fig. 4 Genes encoding CAZymes in the genome of L. salivarius strains among different hosts.a Distribution and abundance of CAZymes categories among different hosts.Kruskal-Wallis test with Mann-Whitney post hoc test was used for comparisons.The compact letter display indicates significant differences in pairwise comparisons, and groups with different letters are significantly different (P < 0.05).b Heatmap of the number of specific CAZymes categories in the genome of L. salivarius strains from different hosts.** denotes P < 0.01 and *** denotes P < 0.001 based on Kruskal-Wallis test, and the Mann-Whitney post hoc test results were shown in TableS4.The horse group which contains only one strain was excluded from the statistical analysis (See figure on next page.)

Fig. 5
Fig.5 Summary of mobile elements (plasmids and phages) and CRISPR-Cas systems present in L. salivarius strains.Upper panel, the colored cells denote the presence of each of the plasmid clusters, and a number is shown in the cell if a strain harbor more than one plasmid in the same group.The plasmid sequences for the strains were grouped into clusters using the mob_type method implemented in mob-suite, and the clusters were named with the primary_cluster_id values deposited in the mob-suite database.The number of CDSs predicted in the plasmid sequences of each strain was also shown, and different letters denote significant differences (P < 0.05, Kruskal-Wallis test with Mann-Whitney post hoc test).Middle panel, the colored cells denote presence of each of the prophage clusters.The prophage sequences were grouped into clusters using cd-hit-est with parameter -c 0.9.Lower panel, the colored cells denote presence of each of the three CRISPR-Cas systems

Table 1
Summary of L. salivarius complete genomes obtained by sequencing and NCBI