Whole-genome sequencing and probiotic properties of Lactobacillus helveticus KM7 isolated from the gut of the Chinese honey bee (Apis cerana): A promising exopolysaccharide-producing strain

Chen Liu; Le Xu; Shiyu Chen; Sifan Wang; Mingkui Lv; Kun Dong; Qiuye Lin; Zhenhui Cao

doi:10.1186/s12866-025-04286-9

. 2025 Aug 23;25:530. doi: 10.1186/s12866-025-04286-9

Whole-genome sequencing and probiotic properties of Lactobacillus helveticus KM7 isolated from the gut of the Chinese honey bee (Apis cerana): A promising exopolysaccharide-producing strain

Chen Liu ^1,^#, Le Xu ^1,^2,^#, Shiyu Chen ¹, Sifan Wang ¹, Mingkui Lv ⁴, Kun Dong ¹, Qiuye Lin ^3,^✉, Zhenhui Cao ^1,^2,^✉

PMCID: PMC12374427 PMID: 40847281

Abstract

Lactobacillus helveticus is a probiotic bacterium widely used in the food industry. In this study, we evaluate the safety and probiotic properties of L. helveticus strain KM7, isolated from the gut of Apis cerana, through whole-genome sequencing and in vitro experiments. The complete genome consists of 2,164,024 bp, encoding 2192 genes with an average GC content of 36.82%. The strain lacks identifiable virulence factor genes and antibiotic resistance genes, shows no hemolytic activity, and cannot produce biogenic amines such as spermine, cadaverine, or putrescine. L. helveticus KM7 exhibited strong acid and bile salt tolerance, superior adhesion capacity, and remarkable antioxidant activity, with its genome harboring genes associated with these traits. It contains gene clusters potentially involved in the biosynthesis of bacteriocins Helveticin J and Enterolysin A, as well as encoding multiple genes of carbohydrate-active enzymes. Moreover, KM7 demonstrated excellent exopolysaccharide-producing capability, and glucose supplementation in the medium enhanced yield. These findings indicate that KM7 is a safe strain with promising probiotic properties, highlighting its potential for food applications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-025-04286-9.

Keywords: Lactobacillus helveticus, Whole-genome sequencing, Probiotic properties, Safety, Exopolysaccharide

Introduction

Probiotics are defined as living microorganisms that are beneficial to the health of the host when administered in adequate amounts, and lactic acid bacteria (LAB) are one of the most common groups [1]. Lactobacillus helveticus is a homofermentative, Gram-positive, rod-shaped, thermophilic LAB predominantly isolated from cheese, fermented milk, and other dairy products [2, 3]. Its remarkable adaptability to diverse environmental stresses, including high temperature, low pH, low osmotic pressure, and low oxygen levels, has enabled L. helveticus to be widely used as a probiotic [4, 5]. As a dietary supplement, L. helveticus has been demonstrated to improve mood, alleviate alcohol-induced liver injury, and ameliorate intestinal inflammation [6–9]. Furthermore, owing to its unique proteolytic system and capacity for exopolysaccharide (EPS) production, L. helveticus is widely employed in food fermentation [10, 11]. For example, during cheese production, it not only suppresses the growth of gas-producing microorganisms by metabolizing residual carbohydrates, but it also effectively reduces bitterness and accelerates the generation of flavor compounds [12, 13]. These multifaceted probiotic and technological properties of L. helveticus underscore its broad application potential, which has recently attracted increased research attention.

With the expanding applications of probiotics, safety assessment has become increasingly critical [14]. Systematic evaluation of strain safety and probiotic properties is a fundamental prerequisite for their use in human and animal health [15]. Beyond the ability to survive gastrointestinal transit, ideal probiotics should have effective intestinal adhesion and colonization capabilities, along with functional characteristics such as antimicrobial activity [16]. Furthermore, growing concerns regarding food safety and antibiotic resistance have underscored the importance of assessing potential virulence factors and antimicrobial resistance profiles in probiotic strains. Whole-genome sequencing (WGS) has recently emerged as an indispensable tool for strain characterization, offering unprecedented resolution for phylogenetic analysis, functional gene identification, and metabolic pathway elucidation [15]. Consequently, it has become a highly reliable approach for comprehensive safety and probiotic trait assessment.

The honeybee gut microbiota is highly specific and functionally diverse, playing crucial roles in carbohydrate metabolism, antioxidation, and immune modulation [17, 18]. In a previous study, we isolated L. helveticus KM7 from the Apis cerana gut and demonstrated its notable antimicrobial activity and antibiotic susceptibility [19]. However, most currently characterized L. helveticus strains originate from fermented dairy products, while insect-derived isolates (particularly those from honeybees) remain understudied and their potential probiotic traits and genomic features poorly characterized. Therefore, we here employ WGS combined with phenotypic assays to systematically evaluate the safety, probiotic properties and EPS-producing capacity of L. helveticus KM7. Our findings provide a theoretical foundation for developing this strain as both a probiotic candidate and a microorganism for fermented food production.

Materials and methods

Bacterial strains and growth conditions

We previously isolated L. helveticus KM7 from the intestinal tract of A. cerana, collected in Kunming, China (102°45’30.5’’E, 25°8’5.8’’N). KM7 was revived and anaerobically cultured in MRS medium (Guangdong Huankai Microbial Sci. & Tech. Co., Ltd., Guangzhou, China; cat. no. 027312) at 37℃ for 24 h. The pathogenic strains (Escherichia coli K88, Staphylococcus aureus ATCC 49521, and Salmonella typhi) were revived and aerobically cultured in LB medium (Guangdong Huankai Microbial Sci. & Tech. Co., Ltd., Guangzhou, China; cat. no. 028320) at 37℃ for 24 h.

Genome sequencing and analysis

The activated KM7 culture was inoculated into MRS broth at a 1% (v/v) concentration and incubated at 37 ℃ until reaching the mid-logarithmic growth phase (8 h). The bacteria were then harvested by centrifugation at 2,800 × g for 10 min, followed by three washes with sterile PBS (phosphate-buffered saline) buffer for subsequent DNA extraction, which was performed with a Wizard^® Genomic DNA Purification Kit (Promega Corporation, WI, USA; cat. no. A1620). Purified genomic DNA was quantified with a TBS-380 fluorometer (Turner BioSystems Inc., CA, USA). High quality DNA (OD_260/280 = 1.8 ~ 2.0, >20 µg) was used for further research.

We sequenced the KM7 genome using a combination of PacBio RS II Single Molecule Real-Time (SMRT) and Illumina platforms. For Illumina sequencing, at least 1 µg of genomic DNA per strain was sheared into 400–500 bp fragments using a Covaris M220 Focused Acoustic Shearer (Covaris, Inc., MA, USA). Sequencing libraries were prepared using the NEXTflex™ Rapid DNA-Seq Kit (Bioo Scientific, TX, USA), including end-repair, phosphorylation, A-tailing, adapter ligation, and PCR enrichment. Paired-end sequencing (2 × 150 bp) was performed on an Illumina HiSeq X Ten. For PacBio sequencing, 15 µg of DNA was fragmented using a Covaris g-TUBE (Covaris, Inc., MA, USA) at 6,000 g for 60 s. The fragments were then purified, end-repaired, and ligated with SMRTbell adapters (Pacific Biosciences, CA, USA). The library was purified three times using 0.45 × volumes of Agencourt AMPure XP beads (Beckman Coulter, Inc., CA, USA). An ~ 10 kb insert library was prepared and sequenced on one SMRT cell using standard protocols and the complete genome sequence assembled using both PacBio and Illumina reads. Raw sequence data, generated via base calling, were saved as FASTQ files containing sequence read and quality information. Quality trimming was performed to remove low-quality reads. We assembled the genome using the hierarchical genome assembly process (HGAP) and Canu, followed by manual verification to ensure circularization, resulting in a complete genome with seamless chromosomes and plasmids. Final error correction of the PacBio assembly was performed using Illumina reads with Pilon.

Coding sequences (CDSs) were predicted using Glimmer. The predicted CDSs were functionally annotated by aligning against the NR, GO, KEGG and CAZy databases using BLAST, Diamond, Hmmscan and HMMER. Annotations were assigned based on the best-matched hits with an E-value threshold of 10⁻⁵. Putative genomic islands, prophages, and CRISPR arrays were identified bioinformatically using IslandPath-DIMOB, PHASTER, and CRISPRCasFinder, respectively, applied to the complete genome sequence. Genes associated with probiotic functions were identified following the approaches and utilizing the databases described by Sabino et al. and Rocha et al. [20, 21]. Putative bacteriocin biosynthesis gene clusters were identified using the BAGEL4 web server (http://bagel4.molgenrug.nl).

Comparative genome analysis and reconstruction of the pan-core gene set

We performed comparative genome analysis and pan-core gene set reconstruction of L. helveticus KM7 and six other L. helveticus strains from different sources: DS3_8 from commercial dietary supplements (GenBank accession: GCA_003053085.1), CAUH18 from koumiss (GCA_001308285.1), CGMCC from Emmental cheese (GCA_001434945.1), LBGT 162 from sorghum beer fermentation (GCA_019455825.1), LZ-R-5 from yoghurt (GCA_009498395.1), and TMW 12,315 from rice vinegar fermentation (GCA_019455865.1). We calculated the pairwise average nucleotide identity (ANI) between KM7 and the other strains using the integrated prokaryotes genome and pan-genome analysis service (IPGA) v1.09. We reconstructed the pan-genome also using the IPGA, which incorporates the Panaroo tool (v1.3.2) for gene clustering and annotation. Orthologous gene clusters were identified using a 70% nucleotide identity threshold, and core genes were defined as those present in ≥ 95% of the analyzed strains. A stringent support value of 1.0 was applied to ensure high-confidence clustering. Finally, we reconstructed a core gene-based phylogenetic tree using VBCG v1.3 with FastTree to assess sequence divergence among strains.

Safety assessment

Identification of safety-relevant genes

We predicted potential virulence factor-related genes in the genome of KM7 using the Virulence Factor Database (VFDB) v20240301 (https://www.mgc.ac.cn/VFs/) and annotated antibiotic resistance genes via the Comprehensive Antibiotic Resistance Database (CARD) v3.2.9 (https://card.mcmaster.ca). In accordance with the Guidelines for Identification and Safety Evaluation of Strains Used in Direct Feeding of Microorganisms and Fermented Products [22], we set the screening criteria for virulence factors and antibiotic resistance genes as follows: E value < 10⁻⁵, coverage > 70%, similarity > 80%.

Hemolytic activity assessment

The activated cultures of L. helveticus KM7 and S. aureus ATCC 49521 (negative control) were streaked onto Columbia blood agar plates (Guangdong Huankai Microbial Sci. & Tech. Co., Ltd., Guangzhou, China) and incubated at 37℃ for 48 h. Hemolytic activity was assessed by observing hemolysis zones around the colonies: A clear zone surrounding the colony signifies β-hemolysis, whereas a greenish discoloration indicates α-hemolysis; in contrast, the absence of any hemolytic activity is classified as γ-hemolysis.

Assessment of the ability to produce biogenic amines

The capacity to produce putrescine, cadaverine, and spermidine was assessed using biochemical test tubes (Hangzhou Binhe Microbial Reagent Co., Ltd., Hangzhou, China, cat. nos. A139, A140, A141). KM7 was streaked onto MRS agar plates and incubated at 37℃ for 24 h. Single colonies were then inoculated into test tubes for arginine decarboxylase, lysine decarboxylase, and ornithine decarboxylase assays, along with control tubes, followed by sealing with paraffin oil. After 24 h of incubation at 37℃, color changes were observed. A negative result was recorded if both the test and control tubes turned yellow, whereas a positive result was indicated by a color change in the test tube with the control tube remaining yellow.

Probiotic characterization

Acid and bile salt tolerance

We adjusted the pH of MRS medium to different levels (2.0, 3.0, and 4.0) using 1 mol/L diluted hydrochloric acid (HCl) (Tianjin Fengchuan Chemical Reagent Technology Co., Ltd., Tianjin, China) and inoculated activated KM7 into the adjusted medium at 5% (v/v) concentration, followed by culturing at 37℃. Viable cell counts were determined by plating on MRS agar at 0, 1, 2, and 3 h, followed by survival rate calculation. For bile salt tolerance assessment, we inoculated KM7 at 5% (v/v) concentration into MRS medium containing 0.3% (w/v) bile salt (bovine) (Solarbio Biotechnology Co., Ltd., Beijing, China, cat. no. B8210) and cultured it at 37 ℃. Survival rates were determined at 0, 1, 2, and 3 h as follows:

where N_t refers to the number of viable bacteria corresponding to 1, 2 and 3 h and N₀ refers to the number of viable bacteria corresponding to 0 h.

Bile salt hydrolase activity

KM7 was cultured in MRS broth at 37℃ for 2, 4, 6, 8, 10, and 12 h. Subsequently, the cultures were centrifuged at 8,000 × g for 10 min at 4℃. The resulting pellets were washed twice with PBS (pH 7.4) and resuspended in the same buffer. The suspensions were subjected to ultrasonication using an ultrasonic cell disruptor (Ningbo Scientz Biotechnology Co., Ltd., Ningbo, China, cat. no. 950E) operated at 200 W with cycles of 5 s pulse and 5 s pause for 15 min. Cell debris was removed by centrifugation. The Bile salt hydrolase (BSH) activity in the supernatant was quantified using a commercial BSH Activity Assay Kit (Gerisi Bio-Technologies Co., Ltd., Suzhou, China, cat. no. G0933F). One unit of enzyme activity (U) was defined as the amount of enzyme required to catalyze the production of 1 µg of taurine per minute per milliliter of supernatant under the assay conditions.

Adhesion and aggregation

Cell-surface hydrophobicity

KM7 suspension cultured for 24 h was centrifuged at 4,000 × g for 10 min at 4℃. The collected cells were washed twice with PBS and resuspended in 2 mL PBS to a concentration of 10⁸ CFU/mL, followed by measurement of the optical density at 600 nm (A₀). Then, 2 mL cell suspension was mixed with 2 mL xylene (C₈H₁₀) (Tianjin Zhiyuan Chemical Reagent Co., Ltd., Tianjin, China), vortexed for 5 min, and allowed to stand for 30 min for phase separation. The optical density of the aqueous phase was measured at 600 nm (A₁). Cell surface hydrophobicity was calculated as follows:

Auto-aggregation and co-aggregation

We cultured KM7 anaerobically in MRS broth at 37℃ until the late exponential phase (OD₆₀₀ ≈ 1.0). Cells were harvested by centrifugation (5,000 × g, 10 min, 4℃), washed twice with PBS, and resuspended in the same buffer to a final concentration of 10⁸ CFU/mL. The suspension was then vortexed for 10 s and allowed to incubate at 37℃. Aliquots (200 µl) from the upper layer were collected at 0 h (A₀) and 3 h (A₂), and OD₆₀₀ measured using a spectrophotometer (METASH Co., Ltd., Shanghai, China, cat. no. UV-5100). Self-aggregation ability was calculated as follows:

The selected pathogenic strains (E. coli K88, S. aureus ATCC 49521, and S. typhi) were cultured aerobically in LB medium at 37℃ for 8 h and adjusted to a concentration of 10⁸ CFU/mL. Equal volumes (1:1) of KM7 suspension (10⁸ CFU/mL) and each pathogen suspension (10⁸ CFU/mL) were thoroughly mixed and co-cultured at 37℃. OD₆₀₀ measurements of the upper suspension were performed at identical time (5 h) intervals. Co-aggregation efficiency was determined using the formula:

where A₂ is the OD₆₀₀ (10⁸ CFU/mL) of the KM7 suspension, A₃ is the OD₆₀₀ (10⁸ CFU/mL) of each pathogen suspension, and A₄ is the OD₆₀₀ of the mixed suspension after standing.

Antioxidant capacity

The total antioxidant capacity (T-AOC), hydroxyl radical scavenging ability, 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging ability, reduced glutathione (GSH) content, and superoxide dismutase (SOD) activity of both fermentation supernatant and cell-free extract of L. helveticus KM7 were quantified using commercial assay kits (Solarbio Biotechnology Co., Ltd., Beijing, China, cat. nos. BC1310, BC1320, BC4750, BC1170, BC0170). For T-AOC determination, FeSO₄ (40 µmol/mL) was used as the standard compound; DPPH radical scavenging ability was quantified using vitamin C (10 mg/mL) as the standard reference; Reduced glutathione (10 mg/mL) served as the standard compound for GSH content measurement. Fermentation supernatant was collected by centrifugation of the bacterial culture supernatant (8,000 × g, 15 min, 4℃). To obtain cell-free extract, cell pellets were washed twice with PBS, disrupted via ultrasonication (200 W, 5 s pulse/5 s pause, 15 min) in ice-cold lysis buffer, and clarified by centrifugation (8,000 × g, 20 min, 4℃).

Extraction and quantification of crude exopolysaccharides

Crude EPS were extracted following the method of Senanayake [23]. with modifications. Bacterial cultures were centrifuged (8,000 × g, 20 min, 4℃), and the supernatant was treated with an equal volume of 10% (w/v) trichloroacetic acid (C₂HCl₃O₂) (Chengdu Kelong Chemical Co., Ltd., Chengdu, China) to precipitate proteins (homogenized at 90 rpm for 30 min). After centrifugation (8,000 × g, 20 min, 4℃), EPS was precipitated from the clarified supernatant using ice-cold ethanol (1:2 v/v, 4 ℃, 24 h), recovered by centrifugation (8,000 × g, 20 min, 4℃), and lyophilized. The crude EPS were dissolved in deionized water to prepare an aqueous solution for further analysis.

Quantification of crude EPS was performed using the sulfuric acid-phenol method. A glucose standard curve was prepared by mixing 0–2.0 mL (in 0.2 mL increments) of 100 µg/mL glucose solution with distilled water (final volume: 2 mL), followed by addition of 2 mL 6% phenol (C₆H₅OH) (Guangdong Guanghua Sci-Tech Co., Ltd., Shantou, China) and 10 mL concentrated sulfuric acid (H₂SO₄) (Tianjin Fengchuan Chemical Reagent Technology Co., Ltd., Tianjin, China). After vortexing and 30 min incubation at 25℃, absorbance was measured at 490 nm. For crude EPS quantification, 2 mL of aqueous solution of crude exopolysaccharides was similarly processed, and concentration was calculated using the standard curve. The strain was cultured statically in MRS broth (1% inoculum, 37℃), and the content of extracellular polysaccharides was measured at 4, 8, 12, 16, 20 and 24 h. Following the method of Tenea [24] with minor modifications, the strain was statically cultured on MRS broth supplemented with 20% (w/v) each of glucose (BioFroxx GmbH, Heidelberg, Germany, cat. no. 1179GR500), sucrose (Macklin Biochemical Co., Ltd., Shanghai, China, cat. no. S818046), fructose (Macklin Biochemical Co., Ltd., Shanghai, China, cat. no. D809612), and lactose (Macklin Biochemical Co., Ltd., Shanghai, China, cat. no. L812298) (1% inoculum, 37℃, 24 h). The crude EPS content was subsequently quantified.

Statistical analysis

Data were analyzed using SPSS. Group differences were evaluated by Student’s t-test or one-way ANOVA followed by Tukey’s post hoc test. Results are reported as mean ± SD. Statistical significance was defined as a two-tailed p-value < 0.05.

Results and discussion

Genome characteristics of L. helveticus KM7

The genome of L. helveticus typically consists of a single circular chromosome ~ 2 Mbp in size, with some variation among strains [3]. The KM7 genome has a total length of 2,164,024 bp with a GC content of 36.82%, distributed on one chromosome and three plasmids (Fig. 1A). It encodes 2,192 genes with an average coding sequence length of 856.32 bp (Table 1). Of these, 2,145 genes are located on the chromosome, while the remaining 30, 11, and 6 genes are found on plasmid A, B, and C, respectively. The total length of coding sequences is 1,877,052 bp, representing 86.74% of the genome. Genomic analysis identified fifteen genomic islands (GIs) in the KM7 genome, fourteen located on the chromosome and one on plasmidA (Table S1). Three CRISPR arrays were also detected, two on the chromosome and one on plasmidA (Table S2). No genes encoding Cas proteins associated with the CRISPR arrays were detected. The absence of these essential components results in a standalone, defective CRISPR structure, compromising formation of a functional CRISPR-Cas defense system. Consequently, this abolishes the capacity of the system for targeted cleavage of exogenous genetic elements, such as plasmids harboring antibiotic resistance or virulence genes [25]. Furthermore, no prophages were identified.

Fig. 1 — Genome overview of L. *helveticus* KM7 and functional annotation in different databases. A Chromosomal and plasmid genomic circular map. B GO function classification. C KEGG function classification

Table 1.

Characteristics of the L. helveticus KM7 genome

Attributes	Values
Genome Size (bp)	2,164,024
Chromosome	2,128,667
PlasmidA	23,910
PlasmidB	6,414
PlasmidC	2,192
G + C Content (%)	36.82
Chromosome	36.81
PlasmidA	38.53
PlasmidB	34.49
PlasmidC	36.7
Coding Gene Number	2,192
Coding Gene Average Length (bp)	856.32
Coding Gene Total Length (bp)	1,877,052

Open in a new tab

The Gene Ontology (GO) database offers standardized terms and hierarchical classifications to describe gene and protein functions across diverse organisms, and is continuously updated to enhance biological research [26]. GO annotation of the KM7 genome revealed that 1,498 genes (68.34%) were successfully annotated (Fig. 1B). Within the Biological Process category, 1,194 genes were annotated, with translation-related and phosphorylation-related genes being the most abundant (58 genes each). For Cellular Component, 819 genes were annotated, with membrane-associated functions representing the predominant category (250 genes). Furthermore, 2,423 genes were assigned to Molecular Function, among which ATP-binding proteins constituted the largest group (234 genes). These genes may contribute to the adhesive colonization capacity and environmental stress resistance of L. helveticus KM7.

KEGG functional annotation enables the mapping of genes, proteins, and other biomolecules to pathways and functional modules within the KEGG (Kyoto Encyclopedia of Genes and Genomes) database, elucidating their roles and interactions in biological processes such as metabolism and signal transduction [27]. Our KEGG analysis of the L. helveticus KM7 genome successfully annotated 1581 genes (72.13%), with 77, 143, 165, 87, 1,081, and 28 assigned to Cellular Processes, Environmental Information Processing, Genetic Information Processing, Human Diseases, Metabolism, and Organismal Systems, respectively (Fig. 1C). Notably, metabolism-related genes constituted the predominant category, including 145 genes involved in carbohydrate metabolism. This metabolic profile aligns remarkably well with the characteristic fermentative properties of LAB. The efficient carbohydrate metabolic capacity of strain likely drives substantial lactate production, suggesting that KM7 holds promising potential for dairy fermentation and functional food applications [28].

Comparative genomic analysis of L. helveticus KM7 and other L. helveticus strains

Average Nucleotide Identity (ANI) analysis compares homologous sequences between pairwise genomes to evaluate phylogenetic relationships at the whole-genome level, with an ANI threshold > 95% typically indicating conspecificity [29]. Our comparative analysis of KM7 and six other L. helveticus strains from diverse origins revealed that all pairwise comparisons exceeded the 95% threshold, confirming the accurate taxonomic classification of KM7 as L. helveticus at the genome level (Fig. 2). Notably, KM7 had the highest ANI value with DS3_8 (99.81%), followed by CAUH18 (98.01%) and LBGT 162 (98.26%). This suggests that these strains have a closer evolutionary relationship with KM7 and thus may have similar metabolic capabilities due to the conservation of specific functional genes.

Fig. 2 — Comparative genomic analysis of L. *helveticus* KM7 and other L. *helveticus* strains. A Heatmap of ANI analysis. B Collinearity analysis

Genomic synteny analysis enables the elucidation of genetic locus alterations caused by various mechanisms, including inversions and translocations between sequenced and reference strains, as well as sequence insertions and deletions [30]. Comparative synteny analysis of L. helveticus KM7 with the six other strains mentioned above revealed extensive collinear conservation of most genes between KM7 and both CAUH18 and LZ-R-5, with only localized genomic rearrangements (e.g., inversions, translocations, and positional variations). This conservation pattern suggests that these strains have undergone relatively constrained genomic evolution. In contrast, KM7 had more extensive genomic rearrangements when compared with DS_8, TMW 12,315, LBGT 162, and CGMCC. These differences in genomic architecture may potentially correlate with the strains’ distinct isolation sources, environmental adaptation histories, or genetic backgrounds.

Construction of the pan-core gene set

Core and pan-genome analyses have been widely employed to study bacterial evolution [31]. The pan-genome of L. helveticus KM7 and the six additional strains comprises 6,783 genes, including 1,201 core and 5,582 strain-specific genes (Fig. 3A). The rarefaction curve revealed decreasing core-gene clusters and increasing pan-gene clusters with the addition of more strains (Fig. 3B), indicating an open genome architecture and high genetic diversity [32]. Open pan-genomes are typically associated with broad environmental adaptability and frequent horizontal gene transfer (HGT), often facilitated by the acquisition of adaptive genes (e.g., related to carbohydrate metabolism and stress response) from mobile genetic elements such as phages, plasmids, or transposons in heterogeneous niches like dairy products or the host gut [33]. This genomic flexibility underlines the functional versatility of L. helveticus in industrial fermentation and probiotic applications. Our phylogenetic analysis of core gene sequences demonstrates a close evolutionary relationship between KM7 and DS_8 (Fig. 3C), suggesting that they share biological and physiological traits. As a model strain isolated from commercial dietary supplements, DS_8 is likely to have enhanced probiotic properties and metabolic capabilities for host health benefits [34]. The genetic similarity between KM7 and DS_8 implies that KM7 may have comparable beneficial traits, warranting further investigation.

Fig. 3 — Construction of the pan-core genome of L. *helveticus* KM7 and six other L. *helveticus* strains. A Core and unique genes of the seven strains. B Dilution curves of the core-genes and pan-genes clusters. C Phylogenetic tree reconstructed from core genes

Safety assessment of L. helveticus KM7

Antibiotic resistance

To ensure the safety and efficacy of probiotic strains, they must be devoid of antibiotic resistance genes (ARGs) [35]. No ARGs meeting the screening criteria of the “Guidelines for Identification and Safety Evaluation of Strains Used in Direct Feeding of Microorganisms and Fermented Products” (Ministry of Agriculture and Rural Affairs of China) were detected in our analysis of the L. helveticus KM7 genome. Although bacteria may acquire ARGs through HGT, mobile genetic elements (MGEs), or chromosomal mutations, KM7 carries only one transposon located on a plasmid, suggesting a low risk of plasmid-mediated ARG dissemination [36]. The transporter proteins detected in GIs (e.g., the MFS transporter in GI11 and the ABC transporter in GI15) are membrane-associated but show no apparent linkage to mobile genetic elements (MGEs), suggesting a low risk of ARG dissemination [37]. Moreover, consistent with the absence of annotated resistance genes in its genome, our prior phenotypic assays confirmed that KM7 is sensitive to cefotaxime, amoxicillin, cephalothin, penicillin G, ampicillin, kanamycin, and vancomycin, as well as moderately sensitive to novobiocin [19]. These results demonstrate that L. helveticus KM7 meets probiotic safety standards at both genomic and phenotypic levels, aligning with regulatory requirements for fermented products.

Virulence factors

Virulence factors play crucial roles in bacterial pathogenesis, including host invasion, colonization, and disease progression [38]. Our genomic analysis of L. helveticus KM7 against the Virulence Factor Database identified only one putative virulence factor (capsule). Importantly, many genes annotated in virulence databases may participate in routine metabolic processes or environmental adaptation rather than directly mediating pathogenicity [39]. For instance, certain bacterial adhesion-related genes in KM7 might solely facilitate intestinal colonization, which is a desirable trait for potential probiotic applications [40]. Notably, the capsule virulence factor is commonly present across Lactobacillus species. As surface polysaccharide structures, bacterial capsules in non-pathogenic strains typically contribute to environmental stress resistance and biofilm formation rather than virulence [41]. The GIs within the KM7 strain lack detectable canonical virulence genes (e.g., hemolysins, invasins) and harbor only toxin-antitoxin (TA) systems (GI02/GI03). These systems play a significant role in cellular stress responses, enhancing genomic stability and cell viability without exerting detrimental effects on the host or environment [42]. Furthermore, transposases (e.g., those in GI04/GI12) are primarily associated with metabolic genes (such as the GI06 fatty acid biosynthesis cluster) and do not carry exogenous pathogenic islands. These findings demonstrate that KM7 lacks definite virulence factors and complies with safety standards for microbial strains used in fermented food production.

Hemolytic activity

Assessment of hemolytic activity is a critical safety evaluation for probiotic strains [43]. Our hemolysis assays revealed distinct zones of β-hemolysis surrounding S. aureus colonies (positive control), indicating hemolysin production (e.g., α-toxin or staphylolysin) capable of erythrocyte lysis and systemic infection [44] (Fig. 4A). In contrast, no hemolytic zones were observed around L. helveticus KM7 colonies, demonstrating γ-hemolysis (i.e., a non-hemolytic phenotype). These results confirm that L. helveticus KM7 lacks the capacity to damage host erythrocytes, consistent with the safety profile expected of LAB strains.

Fig. 4 — Safety assessment of *L. helveticus* KM7. A Hemolytic activity assay. B Production capacity assays for spermine, cadaverine, and putrescine

Biogenic amine production

Biogenic amines (BAs) in food are primarily formed through decarboxylation of free amino acids by microbial amino acid decarboxylases, and their excessive accumulation may lead to food poisoning. For instance, in fermented foods such as cheese and sausages, certain LAB and Enterococcus species can pose safety risks by producing BAs [45]. Therefore, we evaluated the BA-producing capacity of L. helveticus KM7. According to our results, the arginine decarboxylase, lysine decarboxylase, and ornithine decarboxylase test tubes showed no color change, indicating that KM7 lacks the ability to decarboxylate arginine, lysine, or ornithine, and thus cannot produce spermine, cadaverine, or putrescine (Fig. 4B). This further shows the suitability of L. helveticus KM7 for applications in the food industry.