Functional Genomic and Phenotypic Analysis of Lactiplantibacillus pentosus P7 Isolated from Pickled Mustard Greens Reveals Capacity for Exopolysaccharide, B-Vitamin, and Lactic Acid Production

Abstract

Lactiplantibacillus pentosus is a lactic acid bacterium frequently detected in various fermented foods; however, the genomic traits related to its biotechnological potential have been underexplored. In this study, 34 catalase-negative isolates were obtained from pickled mustard greens, among which strain P7 exhibited the highest exopolysaccharide (EPS) yield (781.9 ± 14.7 mg/L) and was capable of growing in a chemically defined medium lacking riboflavin. Whole-genome sequencing revealed a 3,749,478 bp circular chromosome with 46.5% G + C content and 3389 protein-coding genes. A phylogenomic analysis identified P7 as L. pentosus. Functionally, 1 mg/mL EPS extracted from P7 demonstrated strong antioxidant activity, with DPPH and hydroxyl radical scavenging capacities of 89.8 ± 4.6% and 76.5 ± 9.5%, respectively. The use of 0.2 mg/mL EPS also protected Saccharomyces cerevisiae cells from oxidative stress. A comparative genomic analysis indicated the presence of nearly complete biosynthetic pathways for riboflavin, folate, and pyridoxine. High-performance liquid chromatography (HPLC) confirmed the production of 23.8 ± 0.4 µg/mL riboflavin, 36.6 ± 0.6 µg/mL folic acid, and 0.42 ± 0.02 µg/mL pyridoxine in the culture supernatant, which have not been previously reported. Additionally, strain P7 produced 91.2 ± 12.3 g/L of lactic acid after 24 h of incubation. These results support the potential of L. pentosus P7 as a candidate for industrial applications in the production of EPS, B-group vitamins, and lactic acid.

1. Introduction

Lactobacilli are known to be the most important group of bacteria in human and scientific history [1]. They belong to a group of Gram-positive, facultatively anaerobic, and non-spore-forming bacteria that live in a broad range of nutrient-rich environmental niches, such as human gastrointestinal systems, plant materials, and fermented foods [2,3]. In this context, members of this group are known to exhibit great biotechnological applications, including starter cultures, probiotics, and the production of bioactive metabolites, such as bacteriocins, vitamins, exopolysaccharides (EPSs), organic acids, and hyaluronic acid [4]. Lactiplantibacillus plantarum Lp01 is an important ingredient of probiotic preparations, while EPS-producing Limosilactobacillus fermentum YL-11 is recovered from fermented milk [4,5]. L. fermentum MT903311 and MT903312 not only produce antimicrobial compounds but also synthesize high amounts of folic acid and cobalamin (B12), which are not synthesized by humans and animals [6]. Hence, finding potent strains from unique environmental niches is still a promising strategy for producing value-added compounds and enhancing the nutritional properties of food products.

Among lactobacilli, Lactiplantibacillus pentosus, formerly known as Lactobacillus pentosus, is commonly found in a large set of environmental niches, especially fermented foods and beverages [7]. As health-promoting molecules, EPSs produced by L. pentosus LPB8-0 and LPB8-1 showed remarkable emulsifying activity against several edible oils and in vitro antioxidant activity against DPPH and ABTS radicals, for which there is a lack of further in vivo studies [8]. This bacterium produces lactic acid to inhibit the growth of spoilage-associated microbes in order to preserve foods [9]. Oral administration of L. pentosus isolated from table olive fermentations exerts beneficial regulation of the gut microbiota in healthy persons, while the kynurenine level is reduced, leading to improvements in cognitive function in patients [10]. B-group vitamins essential for human life, including B2 (riboflavin), B6 (pyridoxine), B9 (folic acid), and cobalamin (B12), can only be synthesized by microorganisms [11]; however, the ability of this bacterium to synthesize B-group vitamins has not been reported to date.

Whole-genome sequencing and genomic analysis provide the most effective methods to decipher the metabolic capacities and biotechnological applications of a potent bacterium. A thorough genomic study of L. pentosus CF2-10N revealed genetic evidence related to adhesion, EPS biosynthesis, tolerance to low pH and bile salts, immunomodulation, enzyme production, and safety, and thus, its excellent potential as a probiotic or starter culture [9]. Despite high similarity at the genomic level, genomic studies indicate that gene encoding for signal peptides and proteins, specifically transmembrane proteins, present in L. pentosus is higher than in L. plantarum, which may support its prevalence and adaptivity in fermentation [12]. Significant questions remain concerning the genetic traits underlying EPS, lactic acid, and the B-group vitamin production of L. pentosus.

Pickled mustard greens, one of the most popular ethnic foods in Vietnam, are a super nutrient-dense food, especially rich in vitamin B, antioxidant compounds, and beneficial probiotics. Metagenomic analysis of pickled mustard greens collected in Thailand revealed the abundance of Weissella and Lactobacillus genera [13]. Therefore, the aim of the present study was to characterize a new L. pentosus isolated from pickled mustard greens with the potential for EPS and B-group production for application in food fermentation. In silico genomic analysis with other L. pentosus genome sequences was then performed to unveil strain-specific genes and pinpoint genes related to the biosynthesis of EPS, lactic acid, and B-group vitamins.

2. Materials and Methods

2.1. Isolation and Screening of Lactic Acid-, EPS-, and Riboflavin-Producing Isolates

De Man–Rogosa–Sharpe (MRS) agar (HiMedia, Mumbai, India) was prepared to isolate lactic acid-producing bacteria. One gram of pickled mustard greens collected in Hanoi, Vietnam, was transferred into 9 mL of 0.9% (w/v) NaCl solution, serially diluted, and spread on the modified MRS agar at 37 °C for 72 h. Bacterial colonies negative for catalase were selected, re-streaked on MRS agar plates, and incubated at 37 °C for 48 h to obtain pure isolates. To screen lactic acid-producing bacteria, isolates with a clear zone formation were selected and streaked on new MRS agar plates supplemented with 1% (w/v) CaCO₃ [14]. For the screening of EPS-producing bacteria, MRS medium supplemented with 6% (w/v) sucrose was used to promote EPS yield, followed by cultivation at 37 °C for 48 h under anaerobic conditions and precipitation with absolute ice-cold ethanol (99%) at 4 °C overnight [15]. After centrifugation, the pellets were lyophilized and weighted.

Moreover, a chemically defined medium lacking riboflavin was used to evaluate the ability of isolated bacteria to produce riboflavin, as described previously [16]. The overnight cultures were cultivated at 37 °C for 48 h in the chemically-defined medium with a starting optical density of 0.05 at 600 nm. The bacterial growth was evaluated using a UV-spectrophotometer (Shimadzu, Kyoto, Japan) at 600 nm. The observations were performed in triplicate.

2.2. Whole-Genome Sequencing and Genome Annotation

After cultivation in MRS liquid medium at 37 °C for 48 h, the P7 cells were harvested by centrifugation at 10,000 rpm for 15 min and washed 2 times with PBS buffer. Total DNA was extracted using a GeneJET Genomic DNA Purification Kit (Thermo Fisher Scientific, Wilmington, DE, USA) according to the manufacturer’s instructions, and was subsequently checked via agarose gel electrophoresis and quantified using NanoDrop™ (Thermo Fisher Scientific, Waltham, MA, USA). The whole genome of P7 was sequenced using the Illumina NovaSeq platform (Illumina, San Diego, CA, USA), and the resulting raw data were processed using FastQC v0.12.1 and Trimmomatic v0.32 software. The obtained sequencing data were de novo assembled using SPAdes 3.15.3 software with default parameters [17]. In addition, QUAST was utilized to evaluate the quality of the assembled genome, and contigs with lengths shorter than 500 bp were removed [18]. The genome assembly completeness was evaluated by BUSCO v5.6.1 [19].

The P7 genome was annotated using Prokka via the public Galaxy server using default parameters. In addition, functional annotation was performed using the COG classifier v1.0.5, Kyoto Encyclopedia of Genes and Genomes (KEGG), and EggNog mapper v.2 tools. The genome of P7 was then annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP) and deposited in the GenBank database under accession number JBLMKZ000000000. The sequencing data were deposited in the NCBI database under BioProject PRJNA1304875 and BioSample SAMN50574712.

2.3. Taxonomic Identification and Comparative Genomes

The 16S rRNA gene sequence of the P7 was retrieved from its genome, and others of closely related bacterial species were obtained from the NCBI database using BLASTn. A phylogenetic tree was reconstructed using Molecular Evolutionary Genetics Analysis (MEGA) v11.0 with 1000 bootstrap replicates, and the type strain Lactobacillus huangpiensis F306-1^T was chosen as a root. The 16S rRNA gene sequence of P7 was deposited on GenBank (NCBI) under the accession number PV164899. In addition, Type (Strain) Genome Server (TYGS) (https://tygs.dsmz.de, accessed on 22 July 2025) was used to identify the P7 strain at the species level.

For comparative studies, the following 3 genome sequences of L. pentosus were downloaded from NCBI: L. pentosus BGM48 (NZ_CP016491), L. pentosus DSM20314^T (AZCU01000000), and L. pentosus L33 (JAHKRU000000000). Later, Roary was selected to elucidate the core and specific genes within L. pentosus genomes. The proteins of these genomes were uploaded into the OrthoVenn web server with default parameters for a comparison of orthologous clusters [20].

2.4. In Silico Identification of Genes Related to the Production of EPS, Folate, Riboflavin, and Pyridoxine Phosphate

The presence of the genes encoding for EPS, folic acid, riboflavin, and pyridoxine production was investigated in the L. pentosus P7 using BLASTn and RAST based on reference proteins from UniProtKB/Swiss-Prot. An e-value < 10⁻⁵, identity > 30%, and coverage > 50% were used to sort the outcomes.

2.5. EPS Production and Extraction

The isolate P7 was grown in MRS medium (HiMedia, Mumbai, India) supplemented with 60 g/L sucrose for 48 h at 37 °C under anaerobic conditions. The growth was observed by measuring the optical density (OD) at 600 nm using a spectrophotometer (Shimadzu, Kyoto, Japan). The cultures were then centrifuged by Thermo Scientific ST16 Centrifuge (Thermo Fisher Scientific, Waltham, MA, USA) to remove bacterial cells and heated at 100 °C for 10 min to inactivate enzymes. For protein degradation and removal, 4% (w/v) trichloroacetic acid (TCA) was added to cell-free supernatant at 4 °C for 12 h followed by centrifugation at 10,000 rpm at 4 °C for 15 min. The resulting supernatant was precipitated with three times the volume of absolute ice-cold ethanol (99%) and stored at 4 °C for 24 h. The crude EPS precipitate was recovered through centrifugation at 10,000 rpm for 15 min at 4 °C, dissolved in deionized water, and precipitated again with 3 volumes of absolute ice-cold ethanol (99%). The precipitate was dialyzed against distilled water overnight at 4 °C with the water changed twice. Finally, the EPS was lyophilized, weighted, and used for further experiments [15]. The experiment was performed in triplicate.

2.6. Antioxidant and Protective Potential of EPS

The antioxidant potential of the EPS (0.1–1.0 mg/mL) was assessed via different antioxidant assays, such as 1,1-diphenyl-2-picrylhydrazyl (DPPH) and hydroxyl radical scavenging tests [21]. In addition, the absorbance of the mixture was measured using a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA) at 517 nm (0.1 mM DPPH) and 624 nm (0.435 mM brilliant green, 0.5 mM FeSO₄, and 3% H₂O₂), respectively. Ascorbic acid (0.1–1.0 mg/mL) was used as the positive control. Results are from a representative experiment performed in triplicate, and error bars indicate standard deviation on triplicate samples.

The protective effect of the EPS against H₂O₂ was determined with slight modifications, as described previously [22]. Briefly, wild-type S. cerevisiae BY4742 was seeded in the Yeast Extract Peptone Dextrose (YPD) medium (HiMedia, Mumbai, India) at 160 rpm and 30 °C to reach optical density at 600 nm (OD₆₀₀) of 0.6. Subsequently, the culture was pre-treated with 0.4 and 0.8 mg/mL EPS for 1 h, and then subjected to 2 mM H₂O₂ for 1 h at 30 °C. The yeast cells with and without the addition of 0.4 mg/mL EPS were applied as controls. Ascorbic acid, as a reactive oxygen species scavenger, was used as a positive control, while only treatment with 2 mM H₂O₂ served as a negative control. Ten-fold serial dilutions were produced, and a 6 μL aliquot was spread on the YPD agar plates [23]. The experiment was carried out 3 times.

2.7. Determination of Riboflavin, Folic Acid, and Pyridoxine Phosphate Contents

2.7.1. Sample Preparation

The potent candidate was cultivated anaerobically in folic acid-free media at 37 °C for 48 h, followed by centrifugation at 10,000 rpm for 10 min and the addition of a volume of 0.1 M sodium acetate buffer (pH 4.8) supplemented with 1% (w/v) ascorbic acid equal to the volume of cell-free supernatant. For the quantitative analysis of riboflavin, strain P7 was cultivated in the riboflavin-free medium at 37 °C for 48 h and under static conditions, with a starting optical density at 600 nm of 0.1; this process was then repeated twice to obtain good growth. About 500 μL of the cell-free supernatant was subjected to the same volume of 1% (v/v) acetic acid, followed by centrifugation and boiling at 100 °C for 5 min [24]. As for vitamin B6 determination, strain P7 was grown in the synthetic medium (2% glucose, 2% peptone, 0.1% KH₂PO₄, 0.05% MgSO₄.7H₂O, 0.0005% MnSO₄, 0.0005% FeSO₄, pH 6.5) at 37 °C for 48 h without shaking. After centrifugation, proteins in the cell-free supernatant were removed using 0.7 M perchloric acid and then put on ice for 30 min. Lastly, 3 samples were filtered through a 0.2 µm filter and then stored at −20 °C until the HPLC analysis.

2.7.2. HPLC Analysis

The quantification of vitamins produced by the bacterial strain P7 was performed using an Agilent 1290 UHPLC system equipped with a Diode Array Detector (DAD) (Agilent, Santa Clara, CA, USA), which enables UV detection in the range of 195 to 600 nm. The analysis targeted three B-group vitamins, riboflavin, pyridoxine, and folic acid, in the bacterial culture supernatant. Chromatographic separation was carried out using an Agilent 5 Prep-C18 Scalar column (250 × 4.6 mm) under specific conditions for each analyte.

For riboflavin, isocratic elution was applied using a mobile phase of water and acetonitrile (85:15, v/v) at a flow rate of 0.4 mL/min, with the column temperature maintained at 30 °C [25]. Detection was performed at 270 nm, and the total run time was 20 min. Extracellular folic acid was determined using HPLC according to the previous protocol with a slight modification [26]. Folic acid was analyzed under isocratic elution with a mobile phase ratio of water to acetonitrile at 90:10 (v/v), a flow rate of 0.3 mL/min, and a column temperature of 30 °C. Detection was carried out at 290 nm with a total run time of 15 min. For pyridoxine, a gradient elution was employed starting with 10% acetonitrile (B) from 0 to 5 min, gradually increasing to 60% B from 5 to 30 min, holding at 60% B until 40 min, and then returning to 10% B from 40 to 42 min, followed by re-equilibration at 10% B until 60 min [27]. The flow rate was maintained at 0.3 mL/min, the column temperature was 30 °C, and detection was performed at 290 nm. Data are expressed as the mean of triplicate tests, with error bars representing the standard deviation.

2.8. Lactic Acid Production

The quantification of total lactic acid was carried out using reverse-phase HPLC on an Agilent 5 Prep C18 Scalar column (250 × 4.6 mm) [28]. The mobile phase consisted of 0.01 N sulfuric acid, applied in isocratic mode at a flow rate of 0.8 mL/min. The column temperature was maintained at 30 °C, and detection was performed at 210 nm using a UV detector. The sample injection volume was 10 µL, and the total run time was approximately 15 min. Prior to analysis, the samples were centrifuged and filtered through a 0.22 µm membrane. Quantification was based on an external standard curve constructed from serial dilutions of lactic acid standard. Bars represented standard deviation values of triplicates.

3. Results

3.1. Screening of Lactic Acid-, EPS-, and Riboflavin-Producing Isolates

In this study, 34 catalase-negative isolates were recovered from fermented mustard green samples, which showed morphological characteristics of lactic acid bacteria. Among them, 12 isolates produced clear zones around colonies with diameters of above 10 mm when grown on MRS agar plates containing 1% (w/v) CaCO₃. With the MRS medium supplemented with sucrose, three strains (P7, P8, and P9) appeared to be potent and high EPS producers; their yield ranged from 452.2 to 781.9 mg/L (

Table 1. Screening of lactic acid bacteria for their ability to produce lactic acid, EPS, and riboflavin.

Isolates	Lactic Acid Production	EPS Production (mg/L)	Growth on Riboflavin-Free Medium
P1	+	133.0 ± 10.6	−
P2	+	82.2 ± 15.8	+
P3	+	219.5 ± 16.8	−
P4	+	73.9 ± 16.6	−
P5	+	381.8 ± 12.2	+
P6	+	23.5 ± 12.0	−
P7	+	781.9 ± 14.7	+
P8	+	688.4 ± 11.6	−
P9	+	452.2 ± 16.2	−
P10	+	69.2 ± 23.4	−
P11	+	345.5 ± 11.6	−
P12	+	7.3 ± 2.0	+

Note: −, no growth; +, growth/positive. The bold value indicates the isolate P7, which exhibited the highest lactic acid production among all tested strains.

). In addition, 12 isolates were used to screen for riboflavin production. The results revealed that only four isolates grew well on the commercial riboflavin-free medium, indicating their ability to synthesize riboflavin. From the results, the isolate P7 was identified as the most potent candidate and selected for further studies.

3.2. Whole-Genome Sequence and Identification of P7

The whole-genome sequence of strain P7 was obtained using the Illumina NovaSeq sequencing 150PE platform, generating 5,546,849 paired-end reads with an average length of 150 bp. After quality trimming with Trimmomatic, 4,313,170 reads remained and were used for de novo genome assembly with SPAdes software. The resulting draft genome of strain P7 comprised 123 contigs, with a total length of 3,749,478 bp, a GC content of 46.5%, and no plasmids detected (Figure 1A). The completeness assessment using the BUSCO program indicated that 100% of single-copy orthologs were fully assembled, based on comparisons with the bacteria_odb10 lineage dataset. The genome contained 3460 predicted genes, of which 3389 were identified as protein-coding sequences. In addition, 64 tRNAs, 4 rRNAs, 3 ncRNAs, and 85 pseudogenes were found.

A total of 2678 genes were annotated in the COG database, classified into 21 major categories. The category “Carbohydrate transport and metabolism” was the most abundant (307 genes; 11.5%), followed by “Transcription” (280 genes; 10.5%), “Amino acid transport and metabolism” (238 genes; 8.9%), and “Translation, ribosomal structure and biogenesis” (208 genes; 7.8%) (Figure 1B). Of note, only 138 genes (5.2%) were found in “Function unknown”. The functional annotation of the P7 genome using GO showed the presence of 20 categories. The most commonly identified GO categories were “Biological process” (384 genes), followed by “Molecular function” (366 genes), “Cellular component” (296 genes), and “Metabolic process” (291 genes) (Figure S1).

A genome-based phylogenetic analysis using TYGS revealed that strain P7 only formed a distinct phylogenetic lineage with L. pentosus DSM 20314^T (dDDH d₀ 96.2%, d₄ 97.7%, and G + C difference of 0.23%), and not with Lactiplantibacillus plantarum ATCC 14917^T (dDDH d₀ 30.4%, d₄ 24.0%, and G + C difference of 1.6%) or Lactiplantibacillus plantarum DSM 20174^T (dDDH d₀ 30.7%, d₄ 24.2%, and G + C difference of 1.58%) (Figure 1C). The P7 genome exhibited a similar genome size to Lactiplantibacillus pentosus BGM48 and Lactiplantibacillus pentosus DSM 20314^T (Figure S2). In addition, the 16S rRNA gene sequence retrieved from the P7 genome had the highest identity with L. pentosus DSM 20314^T (99.7% matches). Furthermore, it was placed in the Lactiplantibacillus cluster on the phylogenetic tree, with L. pentosus as the most closely related species with a bootstrap of 99% (Figure S2). The morphological analysis confirmed that P7 was milky white with a diameter of around 2 mm, Gram-positive, non-spore-forming, and negative for catalase. These findings verified the identification of the strain as L. pentosus P7.

3.3. Comparative Genome Analysis

A comparative analysis was conducted on the genome of L. pentosus P7 in comparison with those of L. pentosus DSM 20314^T, L. pentosus L33, and L. pentosus BGM48. The results of the core genome analysis suggested that four L. pentosus strains shared a core set of 2658 orthologous protein cluster genes, while the genomes of strains DSM 20314^T, BGM48, and L33 contained 0, 6, and 17 unique orthologous protein clusters, respectively (Figure 2). Of note, two unique protein clusters, UDP-N-acetylglucosamine 2-epimerase (wecB) and UDP-galactose 4-epimerase (galE), were identified in the genome of P7.

3.4. Assessment of Protective Effect Against Oxidative Stress

The genomic analysis showed that the complete EPS biosynthesis gene cluster of L. pentosus P7 was about 16.3 kb. It included 18 genes mainly classified into 4 functional protein groups: regulation, biosynthesis of repeating units, polymerization, and export (Figure 3A). The comparative analysis of the L. pentosus P7 genome revealed that the EPS cluster exhibited homology to those of L. pentosus L33 and L. pentosus SLC13, and was highly conserved in L. pentosus L33 and L. pentosus SLC13. The detailed comparison of the length, position, and identity (%) of genes comprising the EPS cluster of L. pentosus P7 with those of L. pentosus L33 and L. pentosus SLC13 is shown in Table S2.

To highlight the potential of the EPS extracted from the P7, DPPH and hydroxyl radical scavenging assays were utilized to evaluate the antioxidant activity. The results in Figure 3B showed that the strong antioxidant potential was observed against DPPH free radicals. The highest scavenging activity was determined as 89.8 ± 4.6% at concentrations of 1 mg/mL, which was not comparable to the antioxidant compound, ascorbic acid (93.7 ± 2.6%). An increase in EPS concentration did not further promote the scavenging activity against DPPH. Similar to DPPH scavenging activity, antioxidant activity against hydroxyl radicals was EPS dose-dependent between 0.1 and 1 mg/mL, with a maximum concentration of 1 mg/mL inhibition percentage (76.5 ± 9.5%) against hydroxyl radicals. In support of these results, the growth of the yeast model S. cerevisiae was not inhibited by the EPS, and pre-treatment with 0.2–0.4 mg/mL EPS protected the yeast cells against oxidative stress provoked by 2 mM H₂O₂ (Figure 3C). In the absence of EPS, the number of cells was strongly reduced when exposed to 2 mM H₂O₂.

3.5. Molecular Mechanisms Involved in Vitamin B Production

L. pentosus P7 was also found to have a rib operon containing four genes, ribG, ribB, ribA, and ribH, supporting its ability to produce riboflavin. This was consistent with L. plantarum WCFS1 and L. plantarum P-8, but not with L. plantarum CECT 8962^T (Figure 4A). Surprisingly, L. pentosus P7 also contained a de novo biosynthesis pathway for folate, which required 16 necessary genes encoding catalytic enzymes divided into 4 modules: chorismate (aroF, aroB, aroD, aroE, aroK, aroA, aroC), pABA (padA, padC), 6-hydroxymethyl-7,8-dihydropterin pyrophosphate (DHPPP) (folE, folQ, folB, folK), and THF-polyglutamate (folP, folC, dfrA) synthesis (Figure 4A). Of note, shikimate dehydrogenase aroE was found in four copies, in contrast to L. pentosus BGM48 and L. pentosus DSM 20314^T (two copies). Compared to L. brevis, L. delbrueckii, Lactobacillus reuteri, and L. plantarum, the complete pathway for folate biosynthesis was only determined in L. pentosus.

Further genomic annotation indicated that the P7 genome was composed of the core subset of the key genes for the biosynthesis of pyridoxine phosphate. It included pyridoxine kinase pdxK, pyridoxamine phosphate oxidase pdxH, 1-deoxy-D-xylulose 5-phosphate synthase dxs, phosphoserine aminotransferase pdxF, and D-3-phosphoglycerate dehydrogenase serA (Figure 4B). The gene serA was identified as having two copies in both L. pentosus BGM48 and L. pentosus DSM 20314^T, while three serA genes were present in the P7 genome.

In agreement with the genomic analysis, the P7 synthesized 23.8 ± 0.4 µg/mL riboflavin and 36.6 ± 0.6 µg/mL folic acid in the cell-free supernatant through HPLC analysis (Figure 4C and Figure S3). In addition, the pyridoxine level was determined to be 0.42 ± 0.02 µg/mL.

3.6. Assessment of Lactic Acid Production

Following supplementation with 40 g/L sucrose, P7 rapidly produced 30.6 ± 4.2 g/L lactic acid after 12 h under static conditions. The strain then entered the stationary phase after 24 h and yielded 91.2 ± 12.3 g/L lactic acid. The increase in incubation time did not result in the production of lactic acid (Figure 5A). In support of these results, genome mining revealed the presence of two copies of L-lactate dehydrogenase (ldhL) and D-lactate dehydrogenase (ldhD). In the glycolysis pathway that converts sucrose into pyruvate, both L-and D-lactate dehydrogenase promote the conversion of pyruvate to lactate and the reverse reaction. Similar results were also observed in the genomes of L. plantarum, L. reuteri, and L. helveticus, but not in those of L. casei or L. delbrueckii (Figure 5B).

4. Discussion

L. pentosus is typically found in a wide range of vegetable fermentations, including fermented olives, fermented teas, glutinous rice dough, chili sauce, mustard pickles, stinky tofu, and dairy products, and other sources, such as human stool, sewage, and milk [29]. As a result, 136 genome sequences of L. pentosus were available on the GenBank database at the time of writing, with genome sizes ranging from 3.4 to 3.9 Mb. In our study, the genome size of P7 falls in the middle (3.7 Mb) and has no plasmid (Table S1). It is believed that the genomic size and GC content reflect bacterial lifestyle and ecological flexibility. To adapt to specific ecological niches, lactic acid bacteria such as Lactobacillus helveticus (2 Mb) and Lactobacillus acidophilus (2 Mb) tend to reduce their genome size, in contrast to free-living or nomadic strains [30]. As compared to its closest related L. plantarum, a higher number of signal peptides, signal proteins, and transmembrane proteins are found in L. pentosus [12]. Among L. pentosus genomes, a number of unique genes varied significantly, which may result from horizontal gene transfer, an open pangenome, and the nomadic lifestyle. Taken together, the larger genome makes L. pentosus more advantageous in adapting its ecological niches and fermentation.

Our first preliminary screening revealed that L. pentosus P7 also produced the greatest quantity of EPS (781.9 ± 14.7 mg/L) among the 12 tested strains. The EPS yield from P7 was 1.8-fold higher than that from EPS-producing L. pentosus SLC13 isolated from mustard pickles [7]. L. pentosus B8 produces 2 exopolysaccharides, such as LPB8-0 and LPB8-1, in MRS medium supplemented with 40 g/L sucrose, which have strong emulsifying and antioxidant activities [8]. Mannose and glucose are the main monosaccharide components of the LPB8-0, while LPB8-1 comprises mannose, glucose, and galactose, leading to different bioactivities. Using semi-defined medium containing glucose as the carbon source, the presence of glucose, glucuronic acid, and rhamnose is determined in the EPS produced by L. pentosus LPS26 [31]. These results are in agreement with the assumption that EPS’s composition and structure depend mainly on culture conditions, since genetic determinants within the species vary slightly [4,8]. In support of this, genome analysis revealed the presence of a 16.3 kb EPS biosynthesis gene cluster in P7, which is homologous to that of L. pentosus SLC13. Notably, the genes wecB and gale, which are responsible for the synthesis of EPS, were also uniquely identified along the chromosome of P7 [3,15]. This suggests that the additional genes may promote EPS production. One of the limitations of this study is that the identification of biosynthetic genes was solely based on in silico annotation, without transcriptomic or proteomic validation. Future work involving RNA sequencing, quantitative reverse transcription-PCR, and protein expression assays will be required to confirm these findings, strengthen genotype–phenotype associations, and fully elucidate the biosynthetic pathway responsible for EPS synthesis in strain P7.

In addition, EPS from L. pentosus P7 showed 89.8 ± 4.6% and 76.5 ± 9.5% antioxidant activity against DPPH and hydroxyl radicals at 1 mg/mL, respectively. In contrast, two EPS fractions, LPB8-0 and LPB8-1, secreted by L. pentosus B8 have antioxidant activity against DPPH and hydroxyl radicals ranging from 47.9 to 72.5% at 10 mg/mL [8]. The antioxidant activity of probiotic bacteria plays an important role in reducing the oxidative stress provoked by the gut environment and fermentation [22,32]. Moreover, free radicals contribute to the development of several chronic diseases and the aging process [33]. Our results proved that the treatment with 0.2 and 0.4 mg/mL EPS eliminated ROS production and protected the yeast model from oxidative stress. In contrast to in vitro antioxidant activity, only a few studies have used yeast models to test the protective effects of EPS produced by Lactiplantibacillus species. Based on the results of this study, P7 may serve as a suitable probiotic candidate and EPS producer in the pharmaceutical industry.

The first significant finding of this study was the verification of the ability of L. pentosus P7 to synthesize riboflavin at the genomic and phenotypic levels. It is widely believed that riboflavin is required as a precursor for the flavin adenine dinucleotide and flavin mononucleotide, which cannot be synthesized by humans and animals [24]. Under optimized conditions, riboflavin production of up to 26.9 mg/mL in 90 h was reported from genetically engineered Bacillus subtilis KCCM 10445^T [34], which is the highest concentration recorded in microorganisms. In lactic acid bacteria, the highest riboflavin accumulation was observed in L. plantarum HY7715 (34.5 ± 2.4 µg/mL) [35], which was higher than that of L. pentosus P7 (23.8 ± 0.4 µg/mL) (

Table 2. Concentration of B-vitamins such as riboflavin, folic acid, and pyridoxine, as produced by lactobacilli.

Bacterial Strain	Vitamin Concentrations (µg/mL)			References
	Riboflavin (B2)	Folic Acid (B9)	Pyridoxine (B6)
L. pentosus P7	23.8 ± 0.4	36.6 ± 0.6	0.42 ± 0.02	This study
L. plantarum HY7715	34.5 ± 2.4	-	-	[35]
L. delbrueckii KH1	-	100 ± 2.4	-	[36]
L. casei	-	45.4	-	[37]
L. paracasei JCM 1171T			1566.17	[38]
L. rhamnosus VKPMB-8238	-	2.09 ± 0.01	-	[39]
L. acidophilus VKPMB-2105	0.917 ± 0.010	-	4.09 ± 0.02	[39]

). The genes ribA, ribB, ribH, and ribG gene expressions resulting from a single-nucleotide mutation in the regulatory region contribute to the high riboflavin-overproducing ability of L. plantarum HY7715 under in vitro and in vivo conditions [35]. Genomic evidence proved that L. pentosus P7 contains all genes in the rib cluster, confirming the studied strain to be the first L. pentosus able to produce riboflavin. Taken together, the optimization of cultural conditions and CRISPR/Cas9-based genome editing are interesting subjects for future studies.

Given that folate and its derivatives function as cofactors for DNA and amino acid synthesis, folate-producing bacteria are required to improve the nutritional value of foodstuffs and replace chemically synthesized folic acid [38,40]. In our study, the strain P7 produced 36.6 ± 0.6 µg/mL folic acid. L. delbrueckii KH1 is reported to yield 100 ± 2.4 µg/mL folic acid, which is 1000-fold higher than that of L. sakei and L. plantarum from Japanese pickles [36,40]. The fermentation of milk by lactobacilli such as L. plantarum, L. casei, and L. acidophilus results in yogurt products containing folic acid concentrations ranging from 42.8 to 63.2 μg/mL [37]. In comparison to other lactobacilli, genetic determinants associated with the production of chorismate, PABA, DHPPP, and THF-polyglutamate of L. pentosus P7 were present in the P7 genome, which were consistent with the phenotypic analysis. In contrast, many non-producers of folate attribute this to the lack of some genes, such as folP, folQ, and dfrA [41]. Hence, L. pentosus P7 could be a promising candidate for application to functional products with folic acid production.

Last but not least, P7 synthesized 0.42 ± 0.02 µg/mL pyridoxine in the cell-free supernatant. In contrast, L. paracasei subsp. tolerans JCM 1171^T, derived from traditional Iranian yogurt, synthesized 1566.17 µg/mL pyridoxine [38] (Table 2). In addition, both L. plantarum LRCC5310 and L. pentosus P7 comprised the complete pathways for pyridoxine biosynthesis [42]. Although variations in analytical methods across different studies may contribute to discrepancies in the reported concentrations of pyridoxine, the pyridoxine-producing capability of strain P7 is insufficient to meet industrial requirements.

Using chemical analysis, L. pentosus P7 was found to produce 91.2 ± 12.3 g/L lactic acid after 24 h. Recently, lactic acid has become the necessary material to manufacture polylactic acid, which is widely used in food packaging materials, fibers, agricultural films, and biomaterials [43]. Many efforts in producing high lactic acid concentrations by lactobacilli under anaerobic conditions have been reported to date. For example, 192 g/L lactic acid is produced by L. paracasei subsp. paracasei CHB2121, and L. lactis yields 210 g/L lactic acid under the optimized condition [44,45]. The total amount of LA produced by L. casei DSMZ 20011 is 397.1 g/L after 288 h using a 5-cycle repeated batch process with hydrolyzed apple pomace [46]. The combination of low-cost, renewable raw materials and fermentation technology represents an effective strategy for scaling up P7 production to the industrial level.

5. Conclusions

This study provides the first comprehensive characterization of the biotechnological potential of L. pentosus P7, a strain isolated from pickled mustard greens. Whole-genome sequencing and comparative genomic analyses confirmed the taxonomic identity of P7 as L. pentosus and revealed the presence of genes associated with exopolysaccharide (EPS) biosynthesis, antioxidant defense, lactic acid fermentation, and the production of B-group vitamins, including riboflavin, folate, and pyridoxine. Phenotypic assays demonstrated that P7 produces high levels of EPS (781.9 ± 14.7 mg/L), shows strong antioxidant activity, and protects Saccharomyces cerevisiae cells against oxidative stress. The strain was also capable of synthesizing 91.2 ± 12.3 g/L lactic acid, along with 23.8 ± 0.4 µg/mL riboflavin, 36.6 ± 0.6 µg/mL folic acid, and 0.42 ± 0.02 µg/mL pyridoxine under the defined culture conditions.

These findings highlight the dual genomic and metabolic versatility of L. pentosus P7 and underscore its value as a promising candidate for applications in food fermentation and functional product development. The integration of omics-based analysis with phenotypic validation offers a foundation for strain optimization through classical methods or genome engineering. Future research should demonstrate a biosynthetic pathway leading to EPS and B-vitamin synthesis using structural and transcriptomic analysis.