Comparative genomics reveals immunomodulatory and anti-obesity traits of Lactococcus lactis subsp. lactis Authors: Huijin Jeong, Hyukjoon Choi, Young-Seo Park Published: 2026 ABSTRACT AbstractAbstractLactococcus lactis subsp. lactis strains have been widely recognized for their probiotic potential, including their immunomodulatory and anti-obesity effects. This study aims to determine the genomic basis of the health-promoting properties of L. lactis subsp. lactis CAB701 and WiKim0124 using whole-genome sequencing and comparative genomic analysis. Taxonomic classification using average nucleotide identity and digital DNA-DNA hybridization confirms both strains as L. lactis subsp. lactis. Genome alignment reveals high synteny among CAB701, WiKim0124, and the reference strains, with minor structural variations. Functional annotation identifies genes associated with carbohydrate metabolism, biofilm formation, and immune modulation. Pan-genome analysis shows a substantial proportion of strain-specific genes, suggesting niche-specific adaptation. Notably, L. lactis subsp. lactis CAB701 and WiKim0124 harbor unique biosynthetic gene clusters for nisin Z and sactipeptides, which may contribute to their antimicrobial and immunomodulatory functions. Additionally, genes involved in short-chain fatty acid metabolism are identified, highlighting their potential role in metabolic regulation. These findings provide valuable insights into the functional genetics of L. lactis subsp. lactis and their applicability as probiotics for immune modulation and obesity management.Graphical Abstract Key points • Comparative genomics revealed strain-specific immune- and metabolism-linked genes • CAB701 and WiKim0124 uniquely harbor nisin Z and sactipeptide biosynthetic clusters • Genomic features support L. lactis application as functional probiotics Supplementary InformationThe online version contains supplementary material available at 10.1007/s00253-026-13892-0.Keywords: Lactococcus lactis, Comparative genomics, Immunomodulation, Anti-obesity, Bacteriocin biosynthesis, Probiotics Introduction Probiotics, defined as live microorganisms that, when administered in adequate amounts, confer a health benefit to the host, have emerged as a cornerstone of the functional food and nutritional therapy field (Araya et al. 2002). These beneficial microbes are recognized for their central role in maintaining gut health, enhancing immune function, and potentially playing a role in obesity management. The health-promoting effects of probiotics are attributed to their ability to modulate the gut microbiota, improve barrier function, produce antimicrobial compounds, and interact with the host immune system (Nagpal et al. 2012). This burgeoning interest in probiotics underscores the need for a deeper understanding of their mechanisms of action and the specific contributions of different probiotic strains to human health. Advances in genomic analysis have opened new frontiers in the study of probiotics. The benefits of genomic analysis lie in its ability to provide a comprehensive view of an organism’s genetic make-up, providing insights into its potential functions and interactions within the host. Comparative genomics, a branch of genomics, involves analyzing and comparing the genomes of different organisms to understand their evolutionary relationships and functional differences. This approach is particularly useful in identifying genetic variations that may contribute to the unique properties of different strains. Pangenomics, another key concept in the field, extends this analysis by examining the entire gene set of a species, including both the core genome shared by all strains and the accessory genome specific to individual strains. This comprehensive approach allows a holistic understanding of the genetic basis of the variability and adaptability of probiotic strains (Bukhari et al. 2022; Luo et al. 2019). Recent research has revealed the functional capabilities of L. lactis subsp. lactis CAB701 and WiKim0124, particularly in the areas of immunomodulation and anti-obesity. A previous study on L. lactis subsp. lactis CAB701 demonstrated its immunostimulatory capacity. The strain significantly enhanced nitric oxide production in macrophage cells and modulated key inflammatory mediators, positioning it as a potential probiotic candidate for boosting immune responses (Jeong et al. 2023b). In addition, the strain also showed remarkable anti-obesity effects in vitro and in vivo, inhibiting adipocyte differentiation and reducing lipid levels, suggesting its potential against obesity and metabolic disorders (Jeong et al. 2023a). Other previous research investigated L. lactis subsp. lactis WiKim0124 and its impact on functional diversity. A study focused on vegetable juice fermented with the strain and its efficacy in reducing lipid accumulation in both adipocyte-differentiated cells and obese mice (Lee et al. 2023). In addition, the study demonstrated the significant anti-inflammatory effects of L. lactis subsp. lactis in LPS-stimulated RAW264.7 cells and in DSS-induced colitis in mice, mediated through the NF-kB pathway (Yun et al. 2022). The aim of this study is to use these genomic tools to elucidate the genetic basis of two specific strains of L. lactis subsp. lactis CAB701 and WiKim0124. Through detailed comparative genomic analysis, this research aims to elucidate the genetic factors that underlie these beneficial effects. The exploration of the genomic landscapes of L. lactis subsp. lactis CAB701 and WiKim0124 has deepened the understanding of the molecular mechanisms governing their interaction with the human host and underlined their potential as probiotics in the functional food industry. This study is positioned at the intersection of probiotic research and genomic science and aims to make a significant contribution to the advancement of probiotic applications in health and disease management. Materials and methods Whole-genome sequencing of L.lactis subsp. lactis CAB701 DNA extraction and quality assessment Genomic DNA (gDNA) was extracted from L. lactis subsp. lactis CAB701 using Qiagen MagAttract HMW DNA kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. gDNA integrity was assessed through agarose gel electrophoresis, and quantification was performed using a Qubit 2.0 fluorometer (Invitrogen, Carlsbad, USA). gDNA purity was determined based on the A260/A280 ratio, with acceptable values set at ≥ 1.8. To confirm taxonomic identity, the 16S rRNA gene was amplified and sequenced using an ABI 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). MiSeq library construction For short-read sequencing, gDNA was fragmented to ~ 550 bp using an M220 Focused-ultrasonicator (Covaris Ltd, Brighton, UK). Fragment size distribution was assessed using a Bioanalyzer 2100 (Agilent, Palo Alto, USA) with the DNA 7500 kit. Library preparation was conducted using the TruSeq DNA Library LT kit (Illumina, San Diego, USA), and sequencing was performed on an Illumina MiSeq platform, generating 2 × 300 bp paired-end reads with the MiSeq Reagent Kit v3 (Illumina). PacBio library construction For long-read sequencing, 5 μg of gDNA was fragmented using a Megaruptor 3 (Diagenode, Belgium) and size-selected using AMPure XP beads (Beckman Coulter, USA) to retain fragments > 3 kb. A SMRTbell library was constructed using the SMRTbell Express Template Preparation Kit 3.0 (PacBio, Menlo Park, CA, USA), followed by sequencing on the PacBio Sequel System using SMRT Cell 1 M v2 with a 10-h movie collection time. Whole genome sequencing and assembly Hybrid genome assembly was performed using Unicycler v0.4.9, integrating high-accuracy Illumina short reads with PacBio long reads. Quality control for short reads was conducted using Trimmomatic v0.36, with adapter removal using BBMap 38.32. Circularized contigs were reoriented to start at the dnaA or repA gene using Circlator v1.4.0. Genome assembly and quality evaluation were carried out at CJ Bioscience Inc. (Seoul, Korea). Comparative genomic dataset construction and taxonomic classification To construct a comprehensive dataset for comparative genomic analysis, genome sequences of L. lactis subsp. lactis CAB701 and WiKim0124 were analyzed alongside publicly available reference genomes. A taxonomic search for Lactococcus species and subspecies was conducted using the List of Prokaryotic names with Standing in Nomenclature (LPSN) on January 15, 2025 (Parte et al. 2020). This search identified 26 species and six subspecies within the Lactococcus genus. Type strains were subsequently verified, and whole-genome sequences of 28 available type strains excluding L. formosensis subsp. formosensis, which was not available in NCBI were retrieved from NCBI GenBank to construct the Lactococcus type strain dataset. Additionally, 24 reference genomes belonging to the Lactococcus genus were selected from NCBI to establish the NCBI reference dataset. A detailed list of strains included in each dataset is provided in Online Resource Table S4. The whole-genome sequence of L. lactis subsp. lactis CAB701 was obtained through de novo sequencing and assembly, as described in Section 'Whole-genome sequencing of L. lactis subsp. lactis CAB701', while the genome sequence of L. lactis subsp. lactis WiKim0124 was retrieved from NCBI GenBank. To determine the taxonomic classification of these strains, ANI and dDDH analysis was performed. ANI values were calculated using the OrthoANI Tool (OAT) command-line version (Lee et al. 2016). Pairwise whole-genome comparisons were conducted using NCBI BLAST + (v 2.16.0), with ANI ≥ 95 ~ 96% used as the species boundary threshold (Camacho et al. 2009). dDDH was performed using the Genome-to-Genome Distance Calculator (GGDC, v 3.0, https://ggdc.dsmz.de/), which employs the recommended BLAST + alignment method (Meier-Kolthoff et al. 2022). The dDDH values were calculated based on the sum of identities found in high-scoring segment pairs (HSPs) divided by the total HSP length. Genomic sequences were uploaded in FASTA format, and dDDH calculations were performed using default settings. The results were interpreted based on established species and subspecies cutoffs, with a dDDH value of ≥ 70% supporting species classification and ≥ 79% supporting subspecies classification. Based on taxonomic classification, two reference strains were selected to facilitate comparative genomic analysis alongside L. lactis subsp. lactis CAB701 and WiKim0124 (Online Resource Table S5). Genome alignment and structural variation analysis Pairwise genome alignments were conducted using NUCMER (v 4.0) to identify homologous regions among L. lactis strains (Marçais et al. 2018). Alignment results were filtered to retain high-confidence matches with a minimum length of 10,000 bp. Strand orientation was adjusted to ensure consistency across comparisons, and the processed alignments were visualized as diagonal dot plots using ggplot2 in R, with syntenic regions represented as colored segments (Wickham 2009). For whole-genome alignment and structural variation analysis, Mauve (v 2.4.0) was used to detect syntenic blocks and genomic rearrangements (Darling et al. 2004). Multiple genome alignments were performed using Progressive Mauve, while pairwise comparisons were conducted with the default mauve aligner. LCBs were identified to assess conserved genomic regions and structural variations across strains. The results were used to compare genome architecture and infer potential evolutionary relationships. Genome annotation and pangenome analysis Genome annotation was performed using Prokka (v 1.14.6) to predict CDS, tRNA, rRNA, and other genomic features (Seemann 2014). The annotated genome was visualized using Proksee, which provides an interactive platform for circular and linear genome representations (Grant et al. 2023). Functional annotation of protein-coding genes was conducted using eggNOG-mapper (v 2.1.12) with the eggNOG database (v 5.0) (Cantalapiedra et al. 2021). Protein sequences predicted by Prokka were used as input for functional annotation and default parameters were used for all analyses. Pan-genome analysis was conducted using Anvi’o (v 7) to compare gene content across L. lactis strains (Eren et al. 2021). To enhance analytical reliability and prevent fragmented gene predictions, contigs shorter than 2500 bp were removed prior to analysis. Core and accessory genes were identified based on their presence or absence across the genomes. The Anvi’o interactive interface was used to visually inspect gene presence-absence patterns among strains. Additionally, gene distribution among strains was represented using Venn diagrams generated with the ggVennDiagram package in R, providing a comparative overview of shared and strain-specific genes (Gao et al. 2024). Functional gene analysis related to health benefits To characterize metabolic and regulatory pathways associated with immune modulation and anti-obesity effects, functional annotation was conducted using multiple bioinformatics tools. CAZymes were identified using dbCAN3 (Zheng et al. 2023). Three independent search algorithms, HMMER, DIAMOND, and dbCAN_sub, were employed to predict CAZyme families. To ensure high-confidence annotation, only CAZymes detected by both HMMER and DIAMOND were considered for further analysis. The functional classification of the identified CAZymes was performed using the CAZy database (http://www.cazy.org/). The predicted enzymes were categorized into GH, GT, CE, PL, AA, and CBM. The BGCs of L. lactis subsp. lactis strains were identified using antiSMASH (v 7.1.0) and Bagel4 with default parameters (Blin et al. 2023; van Heel et al. 2018). To compare the identified BGCs across multiple strains, Clinker (v 0.0.31) was utilized to generate synteny-based visualizations. The GenBank files obtained from antiSMASH were used as input, and homologous gene clusters were aligned to assess structural variations and conserved regions. Functional gene annotation and metabolic pathway reconstruction were performed using BlastKOALA (Kanehisa et al. 2016). Protein-coding sequences from the assembled genome were uploaded to the BlastKOALA server (https://www.kegg.jp/blastkoala/) with the prokaryotes reference gene dataset for KEGG Orthology assignment. Annotated genes were mapped to KEGG pathways using KEGG Mapper to predict metabolic functions. Key pathways related to immunomodulatory and anti-obesity properties were identified for further analysis. Results Taxonomic identification based on ANI and dDDH analysis To determine the taxonomic relationship between L. lactis subsp. lactis CAB701, L. lactis subsp. lactis WiKim0124, and reference strains, pairwise average nucleotide identity (ANI) and DNA-DNA hybridization (dDDH) values were calculated. ANI analysis with the NCBI reference strain L. lactis subsp. lactis 14B4 showed high sequence similarity, with L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124 exhibiting ANI values of 97.525% and 98.851%, respectively, both exceeding the 96% threshold for species delineation (Fig. 1a). Similarly, ANI comparisons with the type strain L. lactis subsp. lactis ATCC 19435 revealed sequence similarities of 97.477% for L. lactis subsp. lactis CAB701 and 98.544% for L. lactis subsp. lactis WiKim0124. Both strains also shared ANI values of 96.967% and 98.544% with L. lactis subsp. hordniae NBRC 100933, further supporting their high sequence similarity (Fig. 1b). Heatmap of average nucleotide identity (ANI) values among Lactococcus lactis strains. ANI values were calculated using OrthoANI to compare L. lactis subsp. lactis CAB701, L. lactis subsp. lactis WiKim0124, and NCBI reference genomes (a) or type strain genomes (b) of L. lactis. The heatmap represents ANI values, with hierarchical clustering illustrating the genetic relationships among the strains In the dDDH analysis, L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124 exhibited similarity values of 79.10% and 89.90% with L. lactis subsp. lactis 14B4, respectively, both exceeding the 79% threshold for subspecies classification. L. lactis subsp. lactis CAB701 showed a dDDH value of 78.60% with L. lactis subsp. lactis ATCC 19435, slightly below the subspecies classification threshold. In contrast, L. lactis subsp. lactis WiKim0124 exhibited a dDDH value of 88.90% with L. lactis subsp. lactis ATCC 19435. For L. lactis subsp. hordniae NBRC 100933, L. lactis subsp. lactis WiKim0124 had a dDDH value of 69.40%, slightly below the 70% species threshold, while L. lactis subsp. lactis CAB701 exhibited a dDDH value of 75.60%, suggesting species-level similarity. Additionally, L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124 shared an ANI of 97.624% and a dDDH value of 80.00%, indicating a high degree of genomic similarity. Genome alignment and structural variation analysis Pairwise genome alignments revealed a high degree of synteny among L. lactis subsp. lactis CAB701, WiKim0124, and 14B4, with strong collinearity and minimal inversions or rearrangements. Syntenic regions were observed primarily along the main diagonal in dot plots, with only a few dispersed structural variations appearing as off-diagonal alignments (Online Resource Fig. S1A–B, D–E). In contrast, alignments involving L. lactis subsp. lactis ATCC 19435 showed several apparent inversions and translocations when compared to L. lactis subsp. lactis CAB701 and WiKim0124 (Online Resource Fig. S1E, F). Whole-genome collinearity analysis using Mauve revealed conserved Locally collinear blocks (LCBs) across all strains, with highly syntenic regions forming distinct segments. L. lactis subsp. lactis CAB701, WiKim0124 and 14B4 exhibited nearly identical LCB patterns, whereas L. lactis subsp. lactis ATCC 19435 displayed more extensive rearrangements compared to other strains (Online Resource Fig. S2). Genome annotation The general genomic features of L. lactis subsp. lactis strains are summarized in Online Resource Table S1. Genome sizes ranged from 2.5 to 2.6 Mb, with variations in coding sequence (CDS) counts. L. lactis subsp. lactis WiKim0124 had the smallest genome, while L. lactis subsp. lactis 14B4 exhibited the largest. Assembly continuity varied, with L. lactis subsp. lactis WiKim0124 having a single-contig genome, L. lactis subsp. lactis CAB701 and 14B4 assembled into two contigs each, and L. lactis subsp. lactis ATCC 19435 displaying a highly fragmented genome with 51 contigs. The number of ribosomal RNA (rRNA) genes was conserved across most strains, except for L. lactis subsp. lactis ATCC 19435, which had fewer. Similarly, transfer RNA (tRNA) gene counts varied among strains. A single tmRNA gene was present in all strains, and only L. lactis subsp. lactis WiKim0124 harbored a repeat region. Figure 2 provides a comparative visualization of genome organization, illustrating differences in assembly continuity. Strain-specific variations in CDS distribution, G + C content, and G/C skew suggest potential functional and structural genomic diversity. Genome circular view of Lactococcus lactis subsp. lactis strains. From the outermost to the innermost ring, the first ring represents the coding sequences annotated by Prokka, including tRNA, rRNA, and tmRNA. The second ring displays the G+C content. The third ring shows the G/C skew, with the (+) strand (green color) and the (−) strand (pink color). This figure was generated using the ProkSee web server (https://proksee.ca). a CAB701, L. lactis subsp. lactis CAB701. b Wikim0124, L. lactis subsp. lactis WiKim0124. c 14B4, L. lactis subsp. lactis 14B4. d ATCC 19435, L. lactis subsp. lactis ATCC 19435 Functional categorization Functional annotation using the eggNOG database classified genes into 25 Clusters of Orthologous Groups (COGs) across L. lactis subsp. lactis CAB701, WiKim0124, ATCC 19435, and 14B4. The total number of annotated genes ranged from 2101 in 14B4 to 2158 in ATCC 19435. Among the functional categories, Function unknown (S) was the most abundant, with similar gene counts across all strains. This was followed by transcription (K), with CAB701 containing the highest number and ATCC 19435 the lowest. Other highly represented categories included amino acid transport and metabolism (E), translation, ribosomal structure, and biogenesis (J), and carbohydrate transport and metabolism (G). The number of genes assigned to category E ranged from 167 to 177, while category J showed minimal variation. In contrast, category G exhibited greater differences, ranging from 143 to 173, suggesting potential strain-specific differences in carbohydrate metabolism. Additional abundant categories included inorganic ion transport and metabolism (P) and nucleotide transport and metabolism (F), with gene counts varying from 139 to 154 and 130 to 138, respectively. Figure 3 illustrates the distribution of genes across COG categories, highlighting conserved and strain-specific functional traits. Clusters of Orthologous Groups (COGs) functional classification of annotated genes in L. lactis subsp. lactis strains. The number of genes assigned to each COG category is displayed, representing various cellular processes and metabolic functions. WiKim0124, L. lactis subsp. lactis WiKim0124; CAB701, L. lactis subsp. lactis CAB701; ATCC 19435, L. lactis subsp. lactis ATCC 19435; 14B4, L. lactis subsp. lactis 14B4 Pangenome analysis A total of 3506 gene clusters were identified in the pan-genome analysis of four L. lactis subsp. lactis strains. Among these, 1786 clusters (50.94%) were classified as core genes, present in all four genomes (Fig. 4). A total of 443 clusters (12.64%) were identified as accessory genes, shared by at least two but not all genomes. Additionally, 1277 clusters (36.42%) were designated as singleton genes, unique to a single genome. Strain-specific gene clusters were most abundant in L. lactis subsp. lactis 14B4 (416 clusters, 12%), followed by L. lactis subsp. lactis ATCC 19435 (350 clusters, 10%), L. lactis subsp. lactis CAB701 (291 clusters, 8%), and L. lactis subsp. lactis WiKim0124 (220 clusters, 6%). Pairwise comparisons revealed varying levels of gene sharing among strains. The number of shared gene clusters between individual strains ranged from 22 to 86, while three-strain shared clusters ranged from 26 to 50 (Fig. 5). Pan-genome distribution and genomic similarity of Lactococcus lactis strains. The radial plot illustrates the distribution of 3506 pan-genomic clusters, categorized into core, accessory, and singleton genes. The inner tree represents genome clustering based on gene presence/absence patterns (Euclidean distance; Ward linkage). The heatmap shows ANI values, with red indicating high similarity and white indicating low similarity. Gene functions are annotated based on COG categories. WiKim0124, L. lactis subsp. lactis WiKim0124; CAB701, L. lactis subsp. lactis CAB701; ATCC 19435, L. lactis subsp. lactis ATCC 19435; 14B4, L. lactis subsp. lactis 14B4 Venn diagram of shared and strain-specific gene clusters in Lactococcus lactis strains. The Venn diagram, generated with ggVennDiagram, illustrates the distribution of shared and unique gene clusters among the four L. lactis subsp. lactis strains. The core genome is highlighted in green, while strain-specific genes are shaded in pink. 14B4, L. lactis subsp. lactis 14B4; ATCC 19435, L. lactis subsp. lactis ATCC 19435; CAB701, L. lactis subsp. lactis CAB701; WiKim0124, L. lactis subsp. lactis WiKim0124 Functional genomic insights into health-related properties Carbohydrate-active enzymes annotation Carbohydrate-active enzymes (CAZymes) were identified in L. lactis subsp. lactis CAB701, WiKim0124, 14B4, and ATCC 19435. The annotation results revealed a similar number of CAZyme families, including glycoside hydrolases (GH), glycosyltransferases (GT), carbohydrate esterases (CE), polysaccharide lyases (PL), auxiliary activities (AA), and carbohydrate-binding modules (CBM) (Fig. 6a). The distribution of CAZyme families among the four strains was further visualized using a heatmap (Fig. 6b). L. lactis subsp. lactis CAB701 harbored a higher number of genes in specific GT and GH families compared to the other strains, with an increased abundance of GT2, GH1, and GH32. Additionally, GT122, GH43_26, GH113, and CE7 were uniquely present in L. lactis subsp. lactis CAB701. However, L. lactis subsp. lactis CAB701 lacked GH67, CE20, and GH8, which were detected in the other strains. In contrast, L. lactis subsp. lactis WiKim0124 exhibited a higher abundance of GH170 and GH38 compared to the other strains, while GT8, GT113, GH24, GH111, GH78, and GH151 were exclusively found in L. lactis subsp. lactis WiKim0124. Comparative CAZyme profiles of Lactococcus lactis subsp. lactis strains. a Distribution of CAZymes identified by both HMMER and DIAMOND analyses. b Heatmap visualization of CAZyme gene counts. The color intensity represents the number of genes within each family. Strains included in the analysis: ATCC 19435, L. lactis subsp. lactis ATCC 19435; 14B4, L. lactis subsp. lactis 14B4; CAB701, L. lactis subsp. lactis CAB701; Wikim0124, L. lactis subsp. lactis Wikim0124. CAZyme families are classified as follows: AA, auxiliary activities; CBM, carbohydrate-binding modules; CE, carbohydrate esterases; GH, glycoside hydrolases; GT, glycosyltransferases Biosynthetic gene clusters Biosynthetic gene clusters (BGCs) in L. lactis subsp. lactis strains were identified using antiSMASH and Bagel4 (Online Resource Table S2). AntiSMASH analysis revealed the presence of streptide-like thioether-bonded ribosomally synthesized and post-translationally modified peptides (RiPPs) and β-lactone-containing protease inhibitor clusters in all strains. Type III polyketide synthase clusters were detected in all strains except L. lactis subsp. lactis 14B4. A Class I lanthipeptide cluster encoding nisin A was identified in L. lactis subsp. lactis CAB701 and WiKim0124. Bagel4 analysis further indicated strain-specific variations in bacteriocin-encoding clusters. L. lactis subsp. lactis CAB701 and WiKim0124 both harbored nisin Z, whereas lactococcin Q and lactococcin B were uniquely detected in L. lactis subsp. lactis CAB701 and ATCC 19435, respectively. Additionally, enterolysin A was identified in L. lactis subsp. lactis WiKim0124, 14B4, and ATCC 19435, with WiKim0124 carrying two enterolysin A clusters. Comparative analysis revealed that L. lactis subsp. lactis CAB701 and WiKim0124 contained two BGCs consistently detected by both tools at the same genomic loci: a Class I lanthipeptide cluster encoding nisin Z and a RiPPs cluster classified as a streptide-like thioether-bond RiPPs cluster by antiSMASH and a sactipeptide cluster by Bagel4. Notably, no other strains exhibited BGCs co-identified by both tools at the same loci. Figure 7 provides a comparative overview of BGC distribution across strains, highlighting strain-specific patterns. Comparative analysis of lanthipeptide-class-I biosynthetic gene clusters (a) and Streptide-like thioether-bond RiPPs gene clusters (b) in Lactococcus lactis subsp. lactis WiKim0124 and CAB701. The biosynthetic gene clusters were identified using antiSMASH, with functional annotations based on gene categories: biosynthetic (dark red), biosynthetic-additional (light red), transport (blue), regulatory (green), and other/hypothetical (gray). The genes involved in nisin biosynthesis, transport, and regulation are highlighted BlastKOALA Functional gene annotation and metabolic pathway reconstruction identified genes associated with immunomodulatory and anti-obesity properties in L. lactis subsp. lactis CAB701, WiKim0124, 14B4, and ATCC 19435 (Online Resource Table S3). Comparative genomic analysis revealed a highly conserved core genome, particularly in genes involved in peptidoglycan biosynthesis, cell wall modification, and organic acid metabolism. However, strain-specific variations were observed in genes related to peptidoglycan degradation, surface-associated proteins, and extracellular vesicle (EV) formation. Genes involved in peptidoglycan precursor biosynthesis were highly conserved across all strains, while those encoding peptidoglycan remodeling enzymes exhibited strain-specific variations. Notably, L. lactis subsp. lactis CAB701 had the highest copy number of cwlO, a peptidoglycan DL-endopeptidase, whereas srtA showed increased copy number in L. lactis subsp. lactis 14B4. Genes encoding peptidoglycan-anchor proteins also varied among strains. L. lactis subsp. lactis CAB701 uniquely harbored inlJ and genes encoding LPxTG motif proteins, which were absent in WiKim0124 and 14B4 but present in ATCC 19435. Genes encoding lantibiotic leader peptide-processing proteases were exclusively detected in L. lactis subsp. lactis CAB701 and WiKim0124. WxL domain proteins, involved in cell wall binding, were present in all strains, whereas tagT_U_V, which contributes to teichoic acid biosynthesis, exhibited a reduced copy number in L. lactis subsp. lactis 14B4. Exopolysaccharide biosynthesis genes (pgaC, icaA) were also present in all strains, with L. lactis subsp. lactis ATCC 19435 displaying a slightly higher copy number. Metabolic genes associated with acetate, lactate, and formate metabolism were consistently detected across all strains, suggesting conserved metabolic functions. In contrast, genes related to EV formation exhibited strain-specific variations, with L. lactis subsp. lactis WiKim0124 possessing the highest number of phage holin genes. Discussion Sequence-based comparisons placed both L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124 within the genomic boundaries of L. lactis subsp. lactis, although their relationship to L. lactis subsp. hordniae warrants careful interpretation. The dDDH values for L. lactis subsp. lactis WiKim0124 clearly exceeded the 79% subspecies threshold against all L. lactis subsp. lactis reference strains, supporting its unambiguous classification. In contrast, L. lactis subsp. lactis CAB701 yielded dDDH values approaching, but in one case falling below, this threshold, indicating that genome-based metrics alone are insufficient for definitive subspecies assignment in this taxon. The extensive genomic identity shared between L. lactis subsp. lactis and L. lactis subsp. hordniae further supports the integration of phenotypic characterization to resolve ambiguous classifications. Phenotypic profiling, particularly carbohydrate utilization patterns, provides the resolution required to discriminate L. lactis subsp. lactis from L. lactis subsp. hordniae when genomic indices alone are inconclusive. Differential acid production across a defined panel of carbohydrates in API 50 CHL analysis has been shown to reliably separate the two subspecies (Li et al. 2019). In a previous study, L. lactis subsp. lactis CAB701 exhibited a carbohydrate utilization profile closely matching that of the L. lactis subsp. lactis type strain ATCC 19435 rather than L. lactis subsp. hordniae (Jeong et al. 2023b), which complements the present genomic data and supports the classification of L. lactis subsp. lactis CAB701 as L. lactis subsp. lactis. The high syntenic conservation observed among L. lactis subsp. lactis CAB701, L. lactis subsp. lactis WiKim0124, and the closely related L. lactis subsp. lactis 14B4 indicates a stable genome architecture within L. lactis subsp. lactis, with only minor variations attributable to localized mutations or niche-specific adaptation. The apparent inversions and translocations detected in alignments involving L. lactis subsp. lactis ATCC 19435, however, should be interpreted with caution: L. lactis subsp. lactis ATCC 19435 was the only strain assembled in scaffolds rather than as a complete genome, and fragmented assemblies are known to generate artifactual structural differences (Klassen and Currie 2012). Resequencing L. lactis subsp. lactis ATCC 19435 with long-read technology would clarify the extent of genuine rearrangement. Despite this limitation, the broader pattern of conserved LCBs across all four L. lactis subsp. lactis strains supports a relatively stable evolutionary trajectory at the subspecies level. Importantly, such genomic conservation does not preclude functional divergence, as probiotic phenotypes have been shown to vary substantially among strains sharing nearly identical genome architecture (Sharma et al. 2023)—a point that motivates the strain-level functional analyses described below. Genome annotation highlighted differences in assembly quality among the four strains, most notably between the single-contig genome of L. lactis subsp. lactis WiKim0124 and the highly fragmented assembly of L. lactis subsp. lactis ATCC 19435. The reduced rRNA gene count detected in L. lactis subsp. lactis ATCC 19435 (six versus the 19 observed in the other strains) likely reflects incomplete annotation of repetitive elements rather than true genomic loss, reinforcing the importance of high-quality assemblies for downstream comparative analyses (Smits 2019). Functional categorization based on the COG database revealed broadly conserved gene distributions in core biological processes—transcription (K), translation and ribosomal structure (J), and amino acid metabolism (E)—across all four L. lactis subsp. lactis strains. In contrast, the carbohydrate transport and metabolism (G) category showed pronounced strain-specific variation that paralleled the isolation source of each strain. L. lactis subsp. lactis WiKim0124, isolated from kimchi, harbored the largest G-category repertoire, consistent with the diverse fermentable substrates—glucose, fructose, sucrose, and structural polysaccharides—encountered during kimchi fermentation (Song et al. 2020). L. lactis subsp. lactis CAB701 (cabbage-derived) and L. lactis subsp. lactis ATCC 19435 contained intermediate gene counts, while L. lactis subsp. lactis 14B4, isolated from an almond drupe with minimal free sugars (Soler et al. 1989; Tran et al. 2018), possessed the smallest set. Together, these patterns are consistent with niche-driven shaping of carbohydrate metabolism repertoires in L. lactis subsp. lactis, an adaptive feature that has been similarly reported for other lactic acid bacteria (Zhao et al. 2023). The open pan-genome structure observed in this study, with a core genome of 1786 gene clusters alongside a substantial accessory and singleton fraction, indicates that L. lactis subsp. lactis retains a stable genetic backbone while continuing to undergo genetic diversification. The relatively higher singleton counts in L. lactis subsp. lactis 14B4 and L. lactis subsp. lactis CAB701 are consistent with niche-driven gene acquisition, paralleling the strain-specific COG (G) profiles described above. Such accessory gene content represents the most informative target for linking genotype to probiotic phenotype (Golchha et al. 2024), motivating the strain-resolved functional analyses—CAZymes, BGCs, and pathway-level annotation—presented in the following sections. Strain-specific differences in the CAZyme repertoire offer mechanistic clues to the previously demonstrated functional phenotypes of L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124. L. lactis subsp. lactis CAB701 was enriched in GT2, GH1, and GH32, and uniquely encoded GH43_26, GH113, and CE7—a profile broadly aligned with biofilm formation, plant-derived oligosaccharide utilization, and hemicellulose modification (Couturier et al. 2022; Graebin et al. 2016; Holyavka and Artyukhov 2024; Mewis et al. 2016; Singh and Manoj 2016). Among these, GT2 is particularly relevant to its probiotic potential, as it encodes PNAG synthase, an enzyme implicated in exopolysaccharide-mediated biofilm formation and mucosal adhesion that has been linked to bacterial persistence and immunomodulation in lactic acid bacteria (Fukao et al. 2025). The absence of GH67, GH8, and CE20 in L. lactis subsp. lactis CAB701—enzymes involved in xylan and xyloglucan degradation—likely reflects niche-specific reduction rather than a functional limitation of probiotic relevance. L. lactis subsp. lactis WiKim0124 displayed a contrasting profile dominated by peptidoglycan-acting and complex-carbohydrate-degrading enzymes, including elevated GH170 and GH38 and the exclusive presence of GT8, GT113, GH24, GH111, GH78, and GH151. The exclusive occurrence of GH24, encoding a lysozyme-like endo-β−1,4-N-acetylmuramidase, is of particular interest: lysozyme-mediated peptidoglycan fragmentation generates muropeptides that modulate macrophage activity and dampen pro-inflammatory signaling (Tagashira et al. 2018; Wohlkonig et al. 2010), providing a plausible genomic basis for the anti-inflammatory phenotype previously reported for L. lactis subsp. lactis WiKim0124 (Yun et al. 2022). Together, these strain-resolved CAZyme signatures suggest two complementary routes by which L. lactis subsp. lactis may engage the host immune system—surface-mediated adhesion via GT2 in L. lactis subsp. lactis CAB701, and peptidoglycan remodeling via GH24 in L. lactis subsp. lactis WiKim0124. The most striking outcome of the BGC analysis was the exclusive co-occurrence of two clusters—encoding nisin Z and a sactipeptide—in L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124, but in neither of the reference strains. Nisin Z is a well-characterized class I lantibiotic with broad-spectrum antimicrobial activity (Mihaylova-Garnizova et al. 2024), and accumulating evidence indicates that it also acts on the host: it suppresses pro-inflammatory cytokine production, shapes innate immune cell activity, and promotes protective immune responses (Kindrachuk et al. 2013; Małaczewska and Kaczorek-Łukowska 2021). Animal studies have additionally linked nisin to attenuation of adiposity through the regulation of lipid metabolism and inflammatory pathways (Al-Emarah et al. 2023). Together, these activities provide a coherent genomic explanation for the immunomodulatory and anti-obesity phenotypes previously documented for L. lactis subsp. lactis CAB701 (Jeong et al. 2023a; Jeong et al. 2023b) and L. lactis subsp. lactis WiKim0124 (Lee et al. 2023; Yun et al. 2022), placing nisin Z at the intersection of these two health-promoting traits. Sactipeptides represent a less-studied class of ribosomally synthesized peptides whose reported functions include antimicrobial activity and quorum-sensing-related host interactions (Mihaylova-Garnizova et al. 2024; Rebuffat 2022); their co-presence with nisin Z in these two strains suggests a layered repertoire of bioactive peptides that may further shape colonization and host signaling. The mechanistic contributions of these clusters—particularly the sactipeptide—to host responses warrant validation through transcriptomic, metabolomic, and in vivo approaches in future work. Genes governing peptidoglycan biosynthesis and the canonical A4α-type cell wall architecture of L. lactis (Courtin et al. 2006) were conserved across all four strains, providing the structural foundation on which strain-specific surface and secretion features are layered. Among these strain-specific features, L. lactis subsp. lactis CAB701 carried the highest copy number of cwlO, encoding a peptidoglycan D, L-endopeptidase implicated in cell-wall turnover and innate immune recognition (Kawai and Akira 2009), and uniquely harbored inlJ and additional LPxTG-motif surface proteins. Although the precise role of inlJ in L. lactis remains to be defined, internalin-family LPxTG proteins have been linked to host engagement and immune modulation rather than epithelial invasion (Sabet et al. 2005), aligning with the surface-mediated immunostimulatory phenotype previously reported for L. lactis subsp. lactis CAB701 (Jeong et al. 2023b). Lantibiotic leader peptide-processing proteases—essential for bacteriocin maturation (Oman and van der Donk 2010)—were detected exclusively in L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124, complementing the strain-specific occurrence of the nisin Z BGC described above and indicating a complete genetic apparatus for nisin Z production in these two strains. Beyond surface structure, the four strains also encoded the metabolic and secretory machinery relevant to host signaling. Genes for acetate, lactate, and formate metabolism were retained in all strains, providing the genetic capacity to contribute to the gut SCFA pool, a class of microbial metabolites that engages G-protein-coupled receptors on immune cells and modulates regulatory T-cell differentiation, anorectic hormone release, and adipose lipid handling (Hevia et al. 2015; Kabat et al. 2014; Schutz et al. 2021; Włodarczyk and Śliżewska 2021). This metabolic feature is therefore consistent with the immunomodulatory and anti-obesity phenotypes previously documented for L. lactis subsp. lactis CAB701 (Jeong et al. 2023a; Jeong et al. 2023b) and L. lactis subsp. lactis WiKim0124 (Lee et al. 2023; Yun et al. 2022). The elevated holin and lysin gene counts observed in L. lactis subsp. lactis WiKim0124 may further enable enhanced EV release (Liu et al. 2022); given the recognized role of bacterial extracellular vesicles in intercellular communication and immune modulation (Dean et al. 2022; Dominguez Rubio et al. 2022; Morishita et al. 2021), this represents an additional, vesicle-mediated route by which L. lactis subsp. lactis WiKim0124 may engage the host. Together, the comparative genomic and functional analyses presented here provide a coherent genetic framework for the immunomodulatory and anti-obesity phenotypes previously documented in L. lactis subsp. lactis CAB701 (Jeong et al. 2023a, 2023b) and L. lactis subsp. lactis WiKim0124 (Lee et al. 2023; Yun et al. 2022). Whole-genome sequencing, taxonomic verification, genome alignment, and pan-genome analysis established that the two strains share a stable core genome with the broader L. lactis subsp. lactis species while harboring distinctive accessory and singleton features. Among these features, three converge on host engagement and metabolic regulation: surface-mediated routes—GT2-encoded PNAG synthase and inlJ/LPxTG surface proteins in L. lactis subsp. lactis CAB701, and GH24-encoded lysozyme-like activity together with elevated holin/lysin-driven EV potential in L. lactis subsp. lactis WiKim0124; the exclusive co-occurrence of nisin Z and sactipeptide BGCs together with their leader peptide-processing machinery in both strains; and the conserved acetate, lactate, and formate metabolic capacity that supports microbial contribution to the host SCFA pool. Collectively, these features provide a plausible genetic basis linking L. lactis subsp. lactis CAB701 and L. lactis subsp. lactis WiKim0124 to the immune- and metabolism-modulating phenotypes attributed to them, and identify candidate molecular determinants for follow-up validation through transcriptomic, metabolomic, and in vivo approaches in the development of these strains as functional probiotics. Supplementary Information Below is the link to the electronic supplementary material. (PDF 624 KB) Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2025-24523283) and the 2024 Otoki Ham Taiho Foundation Research Program. Author contributions YP conceived and designed research. HJ conducted experiments and analyzed data. HJ and YP wrote the manuscript. All authors read and approved the manuscript. Funding This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2025–24523283) and the 2024 Otoki Ham Taiho Foundation Research Program. Data availability The whole-genome sequence of L. lactis subsp. lactis CAB701 generated in this study has been deposited in the NCBI database under BioProject accession number PRJNA992956, BioSample accession number SAMN36378437, GenBank accession number GCA_030489665.1, and RefSeq accession number GCF_030489665.1. The strain CAB701 has also been deposited in the Korean Culture Center of Microorganisms (KCCM) under accession number KCCM13360P. Genome assemblies of the three publicly available L. lactis subsp. lactis strains (WiKim0124, 14B4, and ATCC 19435) used for comparative genomic analyses, along with the expanded Lactococcus genome dataset used for taxonomic classification and ANI-based phylogenomic analysis, were retrieved from the NCBI GenBank database; their accession numbers are provided in Online Resource Tables S4 and S5. All other data generated or analyzed during this study are included in this published article and its supplementary information files. Declarations Ethics approval This article does not contain any studies with human participants or animals performed by any of the authors. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials (PDF 624 KB) Data Availability Statement The whole-genome sequence of L. lactis subsp. lactis CAB701 generated in this study has been deposited in the NCBI database under BioProject accession number PRJNA992956, BioSample accession number SAMN36378437, GenBank accession number GCA_030489665.1, and RefSeq accession number GCF_030489665.1. The strain CAB701 has also been deposited in the Korean Culture Center of Microorganisms (KCCM) under accession number KCCM13360P. Genome assemblies of the three publicly available L. lactis subsp. lactis strains (WiKim0124, 14B4, and ATCC 19435) used for comparative genomic analyses, along with the expanded Lactococcus genome dataset used for taxonomic classification and ANI-based phylogenomic analysis, were retrieved from the NCBI GenBank database; their accession numbers are provided in Online Resource Tables S4 and S5. All other data generated or analyzed during this study are included in this published article and its supplementary information files.