All strains and plasmids used in this study are listed in Table 1. E. coli DH5α strains were employed as the cloning hosts for plasmid construction. E. coli DH5α containing genome editing plasmid was cultivated in Luria-Bertani medium supplemented with 50 μg mL⁻¹ kanamycin (Kan). L. plantarum WCFS1 was routinely cultured on MRS media without agitation. Then, 10 μg mL⁻¹ erythromycin (Em) and 10 μg mL⁻¹ chloramphenicol were added as needed. An appropriate quantity of NAM was introduced during the culture of L. plantarum WCFS1 to increase the biosynthesis of NMN.

All primers used in this study are listed in Table 2, and the successful plasmids were validated by DNA sequencing. The pair of primers was used to amplify LLNZ00315, orf1015 and lp2514 genes from L. lactis NZ9000, Streptococcus thermophilus S-3 and WCFS1, respectively, and inserted into pKLH32 (digested with Bam HI/Eco RI) to create plasmids for NMN production. To further enhance lp2514 expression, multi-copy expression cassettes were constructed. A second copy of lp2514 was amplified and inserted into a separate expression region of the pKLH32 backbone, generating pLXY09 (dual-copy lp2514). To explore the effect of additional copies, pLXY10 was constructed by inserting three tandem copies of lp2514.

The homologous arms of lp2514 were amplified from WCFS1. The specific guide RNA (gRNA), composed of a 20 bp protospacer sequence, was cloned from pLCP using gRNA-lp2514-F/gRNA-lp2514-R primers (Huang et al., 2019). The plasmid pLCP (CRISPR/Cas9 system) was digested with the restriction enzymes Apa I and Xba I, and then ligated to the fragments using the ClonExpress one-step cloning kit, resulting in the creation of recombinant plasmid pLXY07 for lp2514 gene knockout. The pHH13 plasmid was constructed as described above.

The competent cells of L. plantarum WCFS1 were generated in accordance with previous report (Huang et al., 2019). The helper plasmid pLH01 was transformed into L. plantarum WCFS1, and the resulting positive transformants were screened on Cm plates for the preparation of the competent cells. The plasmid pLXY07 was subsequently transferred into WCFS1 via electroporation, and cultivated for 72 h at 37°C. To confirm positive recombinants, colony PCR analysis was performed using the primers (lp2514-confirm-F/lp2514-confirm-R; lp2514-inter -F/lp2514-inter -R) that targeted the homologous regions of the lp2514 gene. After successful genome editing, the recombinant strain was serially passaged without antibiotics for 2–3 generations to facilitate plasmid curing. Colonies were screened for the loss of antibiotic resistance, and plasmid loss was further confirmed by colony PCR.

The plasmids pLXY04, pLXY09, pLXY10 and pKLH32 (as a control) were introduced into L. plantarum WCFS1. The engineered L. plantarum WCFS1 and WCFS1Δlp2514 strains were grown at 37°C. Cell growth was quantified by measuring optical density at 600 nm (OD 600 nm) with a UV–Vis spectrophotometer. The fermentation broths were collected by centrifugation (12,000 g, 10 min, 4°C), and the cells were resuspended in buffer solution for disruption using a low temperature ultrahigh-pressure continuous flow cell disrupter. The resulting cell lysate was then centrifuged (12,000 g, 3 min, 4°C), and the supernatants were decanted for the NMN assay. NMN was derived from L. plantarum and detected with reported fluorometric methods (Kong et al., 2023; Marinescu et al., 2018).

Total RNA was extracted from L. plantarum WCFS1/pKLH32, WCFS1/pLXY04, WCFS1/pLXY09, and WCFS1/pLXY10 using the RNAiso Plus reagent (TaKaRa, Japan) following the manufacturer’s protocol. First-strand cDNA was synthesized from 1 μg of total RNA using a reverse transcription kit. GAPDH was used as the internal control, and relative transcript levels were calculated using the 2^{^−ΔΔCt} method.

Total RNA was extracted from L. plantarum strains WCFS1/pLXY04 (lp2514 overexpression with 0.1% NAM) and WCFS1/pKLH32 (empty vector control with 0.1% NAM) using RNAiso Plus reagent (TaKaRa, Japan) following the manufacturer’s instructions. RNA purity and concentration were assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific, United States), and RNA integrity was verified via agarose gel electrophoresis and an Agilent 2,100 Bioanalyzer (Agilent Technologies, United States). Sequencing was conducted on an Illumina HiSeq 2000 platform (Majorbio Bio-Pharm Technology Co., Ltd., Shanghai, China). Clean reads were mapped to the reference genome of L. plantarum WCFS1, and gene expression levels were quantified as fragments per kilobase of transcript per million mapped reads (FPKM). Differentially expressed genes (DEGs) between the two groups (N04 vs. N32) were identified using DESeq2 software with thresholds of |log₂Fold Change| ≥ 1 and adjusted p-value (FDR) < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed to investigate functional categories and metabolic pathways significantly affected by lp2514 overexpression.

All the experimental data were independently repeated in triplicate. Statistical analyses of the data were performed with GraphPad Prism and are presented as the means ± standard deviations (*p < 0.05 and **p < 0.01).

The efficient biosynthesis of NMN by LAB using NAM as the substrate was shown to be an effective synthesis strategy in our previous study (Kong et al., 2023). The translocation of the NAM substrate into the host cell to increase NMN production requires a robust NAM transporter (Figure 1). However, there have been no reports on the endogenous proteins responsible for transporting NAM in LAB. Native proteins are more favorable for expression within the host. NiaP from Burkholderia cenocepacia significantly improved NAM uptake leading to an enhance in the synthesis of NMN from 185 mg L⁻¹ to 231 mg L⁻¹ (Shoji et al., 2021). In silico analysis revealed the presence of putative NAM transporters in LAB, including L. plantarum, L. lactis and S. thermophilus. Using BCNiaP as a query sequence for BLAST analysis with the target genomes, the genes encoding orf1015 from S. thermophilus S-3, LLNZ00315 from L. lactis NZ9000 and lp2514 from L. plantarum WCFS1 were identified. Phylogenetic tree analysis showed that the NiaP from different sources clustered into distinct branches, and that lp2514 from L. plantarum WCFS1 was positioned in close proximity to BCNiaP (Figure 2A). The above genes were subsequently inserted into the pKLH32 plasmid, and the resulting plasmids were transferred to L. plantarum WCFS1 (Figure 2B). Among these strains, strain WCFS1/pLXY04, which expressed lp2514 could biosynthesize 89.6 μmol L⁻¹ of NMN intracellularly, resulting in 62.3% increase in NMN levels, compared with those of strain WCFS1/pKLH32, which contained only the empty plasmid of pKLH32 (Figure 2C). These results clearly demonstrate that the expression of NiaP (lp2514) facilitates the production of NMN in L. plantarum, which is consistent with the phylogenetic tree. The lp2514 gene markedly affected the NMN yield and was applied for subsequent experiments.

We performed domain analysis of the NAM transporter using the search tool for conserved domains from the NCBI, and reported that BCNiaP and lp2514 possessed the same domains as MFS_SV2_like, UhpC and MFS_1, suggesting that lp2514 has the potential to conjugate NAM to NMN. The MFS transporter (UniProt A0A2R3JML5; seq similarity 53%; coverage 98%) was selected as template for homology modeling of lp2514 (Figure 3A). The Ramachandran plot revealed that all the residues were predominantly in the most favored region (96.4%) and in an additional allowable area (3.6%), suggesting that the conformation of the generated model adheres to principles of stereochemistry (Supplementary Figure S1). The structural models of lp2514 could be applied to simulate molecular docking with the substrate NAM molecule being docked in the substrate binding pocket of lp2514 using a docking server. The key residues of ASP140, TRP19, TRP318, and ILE322 were located close to the substrate NAM (Figure 3B). These findings suggest that the lp2514 protein has strong binding affinity for the NAM substrate.

Considering the potential role of the lp2514 gene in promoting NMN biosynthesis via NAM uptake, it was worthwhile to verify its functionality by knockout of lp2514 via CRISPR/Cas9 technology in L. plantarum WCFS1 (Figure 4A). A pLXY07 plasmid was constructed to develop the gene deletion mutant of lp2514, which was subsequently delivered into WCFS1 (Supplementary Figure S2). One positive mutant (WCFS1Δlp2514) was successfully generated from eight transformants via colony PCR (Figure 4B). The acquisition of a deletion was verified by sequencing the PCR amplification product of the lp2514 mutant (Figure 4C). When 1 g L⁻¹ NAM was added, the intracellular NMN production of WCFS1Δlp2514 was decreased by 17% compared with that of the wild-type WCFS1 (Figure 4D). Interestingly, when lp2514 was overexpressed in the WCFS1Δlp2514 strain, NMN production was restored. These findings revealed that the lp2514 gene plays a role as NAM transporter affecting the biosynthesis of NMN, which agrees well with our above results. This is the first report on newly developed NAM transporter screening for NMN biosynthesis in LAB.

To evaluate whether WCFS1/pLXY04 could increase in NMN production, various substrate concentrations of NAM and fermentation durations were tested. Overexpression of lp2514 did not affect the growth of the recombinant strains WCFS1/pKLH32 and WCFS1/pLXY04 (Figure 5A). Kong et al. (2023) reported that various NAM concentrations had an impact on the growth of L. lactis NZ9000, under high concentration conditions (50 g L⁻¹), NAM inhibited strain growth. The utilization of 50 g L⁻¹ NAM exhibited toxicity toward recombinant E. coli (Marinescu et al., 2018). The growth of WCFS1/pKLH32 and WCFS1/pLXY04 was not limited by the addition of 1 g L⁻¹ or 5 g L⁻¹ NAM, although a slight decrease in the growth rate was observed with 10 g L⁻¹ NAM. This finding is consistent with the findings of our previous study (Kong et al., 2023). Moreover, the growth rates are closely associated with the consumption of glucose (Figure 5B). NAM addition is a limiting factor for NMN biosynthesis, and various researchers have added different doses of NAM when is utilized NiaP (Jeanguenin et al., 2012; Shoji et al., 2021; Su et al., 2024). Compared with the control WCFS1/pKLH32, WCFS1/pLXY04 achieved significant NMN production, where varying amounts of NAM (1, 5 and 10 g L⁻¹) were supplied to the MRS medium (Figure 5C). The maximum intracellular NMN titer of WCFS1/pLXY04 (155.5 μmol L⁻¹; 1 g L⁻¹ NAM) was synthesized after 12 h of fermentation, which was 182.7% greater than that of the control (55 μmol L⁻¹) without NAM supplementation. Marinescu et al. (2018) reported that the addition of 1 g L⁻¹ NAM resulted in the greatest NMN accumulation, which is consistent with our findings. These findings revealed that relatively high concentrations of NAM may not be conducive to NMN accumulation.

To determine the appropriate fermentation time for enhanced NMN production, the effects of fermentation for various time periods (12, 24 and 36 h) on NMN biosynthesis were evaluated, and the results indicated that fermentation times of 24 h (168.9 μmol L⁻¹) or longer were was suitable for the biosynthesis of NMN (Figure 5D). Sugiyama et al. (2021) reported that Fructobacillus tropaeoli RD012354 presented the highest NMN production after 24 h cultivation. These findings revealed that the recombinant strain WCFS1/pLXY04, when cultured under optimized conditions (1 g L⁻¹ NAM and 24 h cultivation), resulted in a 207% increase in the NMN yield, reaching 168.9 μmol L⁻¹.

The multicopy expression cassette strategy represents a robust approach to increase natural product yields by increasing recombinant protein abundance, albeit with potential metabolic trade-offs. Cheng et al. (2023) employed a multicopy integration strategy to increase the production of germacrene A, which resulted in a 7.7-fold increase over the control. An optimal number of gene copies should be carefully studied because of the metabolic pressure associated with protein overexpression. To enhance the expression of lp2514, multicopy expression plasmids were constructed based on the pKLH32 vector. The lp2514 gene was amplified using overlap extension PCR and inserted sequentially into the Bam HI/Eco RI sites of pKLH32 to generate plasmids harboring one (pLXY04), two (pLXY09), and three (pLXY10) tandem copies of lp2514 under the control of the same promoter (Figure 6). Compared with that of WCFS1/pKLH32, the OD₆₀₀ value of WCFS1/pLXY09 significantly decreased during 36 h cultivation, accompanied by a simultaneous reduction in the rate of glucose consumption (Figures 6A,B). The expression of multiple lp2514 imposed a certain burden on the growth of the strain. Notably, an obviously greater quantity of NMN (203 μmol L⁻¹) was substantiated by WCFS1/pLXY09 expressing two copies of lp2514, which increased by 37 and 85% compared with those of WCFS1/pKLXY04 and WCFS1/pKLH32, respectively (Figure 6C). Kong et al. (2020) reported that dual-copy families of the SOD, not four-copy families, presented the greatest activity, with a similar conclusion. The NMN production of WCFS1/pLXY10 also increased by 42% compared with that of the control strain (WCFS1/pKLH32), but slightly decreased compared with that of WCFS1/pLXY09. It is possible that lp2514 overexpression in multiple copies may cause metabolic stress, which subsequently affects NMN biosynthesis. Compared with that of the control WCFS1/pKLH32 without NAM, the final NMN titer was increased by 269% via a series of regulatory mechanisms. The accumulation of NMN in WCFS1/pLXY09 was approximately 33.9-fold greater than that in Fructobacillus durionis RD011727 (Table 3). Moreover, the transcriptional analysis revealed a significant increase in the expression of lp2514 in multicopy recombinants, exceeding that of the control strain by more than 1,000-fold (Figure 6D). The gene copy number is strongly correlated with the transcriptional level of lp2514, rather than with NMN production. This achievement represents the highest reported yield of NMN using NAM as substrate in L. plantarum, marking a significant advancement in the field of NMN biosynthesis by LAB. Although the yield of NMN synthesized has reached g L⁻¹ levels (17.2 g L⁻¹) by E. coli, which are much greater than those of LAB (Maharjan et al., 2023), our L. plantarum platform retains distinct benefits for functional-food applications. Specifically, L. plantarum has GRAS status, an intrinsic probiotic effect (supporting gut health and immune modulation), and the ability to perform food-grade fermentations without extensive downstream purification. These properties make LAB particularly well suited for developing NMN – enriched fermented products that can be marketed directly to health – conscious consumers. This study establishes lp2514 multicopy engineering as a transformative strategy for LAB-based NMN bioproduction, bridging synthetic biology with functional food innovation.

To validate the reliability of the RNA-seq data, a subset of DEGs (upregulated, downregulated) identified between N04 and N32 was selected for RT-qPCR analysis (Table 2). Genes were chosen on the basis of their fold change values, involvement in NAD⁺ metabolism, and relevance to membrane transport processes as indicated by GO and KEGG enrichment analyses. The RT-qPCR results were strongly concordant with the RNA-Seq data (Spearman’s r = 0.93, p < 0.0001), confirming the upregulation of key transport-related genes and NAD⁺ biosynthesis genes in the N04 strain (Table 4). Notably, lp3556 (encoding L-ribulokinase) exhibited a substantial upregulation (RNA-seq = 3.47; RT-qPCR = 3.66), suggesting a shift toward ribulose metabolism potentially linked to enhanced precursor availability for NMN biosynthesis. The key upregulated genes included ribosomal subunits (lp1032, RNA-seq = 1.61; RT-qPCR = 1.89) and NAD⁺ salvage enzymes (lp2830 = 2.71; RT-qPCR = 2.54), which directly support enhanced translational capacity and aspartate-derived precursor supply for NMN biosynthesis. Strikingly, lp2703, which is involved in aspartate carbamoyltransferase activity, was markedly downregulated (RNA-seq = −2.27; RT-qPCR = −1.74), possibly reflecting alterations in pyrimidine biosynthesis under lp2514-mediated conditions. Together, the RT-qPCR results were consistent with the RNA-seq data, confirming the transcriptional regulation of key metabolic nodes.

To further validate the influence of DEGs on NMN biosynthesis, we targeted cinA, encoding nicotinamide-nucleotide amidase, which was downregulated in the lp2514-overexpressing strain and implicated in NMN turnover (Table 4). A cinA deletion mutant (WCFS1ΔcinA) was constructed (Supplementary Figures S3, S4) and assayed alongside wild-type WCFS1 for NMN production. As shown in Figure 10A, WCFS1ΔcinA accumulated significantly greater amounts of NMN than did the parental strain 29%, which was consistent with the removal of the enzyme responsible for NMN degradation. Zhang et al. (2024) similarly showed that deletion of the cinA in Bacillus subtilis enhanced NMN synthesis, consistent with our findings. RT-qPCR analysis (Figure 10B) revealed the absence of the cinA transcript in WCFS1ΔcinA and the upregulation of key salvage pathway genes (e.g., aspA) and ribosomal components (rpsJ), mirroring the global enhancements in precursor flux and translational capacity identified by RNA-seq. These data substantiate a dual mechanism-enhanced NAM uptake via lp2514 and prevention of NMN hydrolysis through cinA deletion-underscoring the value of combining transporter engineering with targeted gene knockouts to maximize NMN biosynthesis. Together, these findings highlight the robustness of the sequencing data and suggest that lp2514 overexpression, in conjunction with NAM supplementation, modulates central carbon and nucleotide metabolism in L. plantarum. Such regulatory shifts are likely instrumental in enhancing NMN biosynthetic potential, laying a molecular foundation for further metabolic engineering aimed at boosting NAD⁺ cofactor synthesis in Lactobacillus.