The target gene PnuC (gene ID: 66264729) was synthesized by Universe Gene Technology (Tianjin) Co., Ltd. (Tianjin, China), and its sequence was codon-optimized for expression in B. subtilis. Oligonucleotide primers were synthesized by GENEWIZ Bio, Inc. (Suzhou, China). NMN was purchased from Shanghai Bide Pharmaceutical Technology Co., Ltd. (Shanghai, China) and used for high-performance liquid chromatography (HPLC) analysis. Unless otherwise specified, all organic reagents were purchased from Beijing AoBoXing Bio-Technology Co. Ltd. (Beijing, China), and all inorganic reagents were purchased from China National Medicines Corporation Ltd. (Beijing, China).

The strains and plasmids used in this study are listed in Supplementary Table S1. The protease-deficient strain B. subtilis WB600 (Wu et al., 1991) and pMA5 (Zhu et al., 2020) were used as the host strain and overexpressing vector, respectively. Compared with integrated plasmids, the shuttle plasmid PMA5 has the advantage of having a higher copy number and a stronger HpaII promoter. In addition to Bacillus mycoides PnuC, B. subtilis 168 was used as a source of genomic DNA for amplifying NMN biosynthesis genes. All primers listed in Supplementary Table S2 were designed based on the genomic sequences of B. subtilis 168.

To construct the recombinant plasmid pMA5-nadE, the primers B. N-F/R were used to amplify NadE gene from B. subtilis 168 chromosomal DNA, and the resulting fragment was cloned into NdeI/BamHI-digested pMA5 at multiple cloning site 1 (MCS1) using ClonExpress II One Step Cloning Kit (Nanjing Vazyme Biotech Co., Ltd., Nanjing, China), according to the manufacturer’s instructions. The recombinant plasmid pMA5-pncB1 was constructed as follows. The entire fragment of PncB (amplified using primers B. P1-F/R) was purified and cloned into NdeI-digested pMA5 at MCS1. The recombinant plasmid pMA5-pncB2 was constructed using PncB as mentioned above and then cloned into HindIII-digested pMA5 at MCS2. To construct a multigene coexpression plasmid, different enzyme cleavage sites were selected based on the sequence of the single-gene expression plasmids. NadE and PncB (with BamHI and HindIII restriction sites) were inserted at MCS1 and MCS2, respectively. To generate pMA5-nadE-pncB1 plasmid, pMA5 was linearized using BamHI and NdeI, and the amplification products of NadE and PncB were inserted at MCS1. The recombinant plasmid pMA5-nadE-pncB2 was constructed by inserting PncB at MCS2 of HindIII-digested pMA5-nadE. The B. mycoides PnuC sequence was synthesized by Universe Gene Technology Co., Ltd. (Tianjin) and cloned into the recombinant plasmid pMA5-pncB1 (HindIII-digested) at MCS2 to establish the recombinant vector pMA5-pncB-pnuC. The resulting plasmids (Figure 3) were used for the transformation of B. subtilis WB600 (Anagnostopoulos and Spizizen, 1961). For selection, all transformants were plated onto Luria broth (LB) agar plates containing kanamycin (100 μg/mL), and positive transformants were verified via colony PCR.

Bacillus subtilis strains were routinely cultured at 37°C in LB medium (consisting of 10 g/L tryptone, 5 g/L yeast extract, and 10 g/L NaCl) or on LB plates containing 1.5–2% (w/v) agar supplemented with appropriate antibiotics as necessary (kanamycin, 50–100 μg/mL). All recombinant B. subtilis cells precultured overnight at 37°C in LB medium containing kanamycin (50 μg/mL) under shaking conditions at 220 rpm were used as seed cultures.

For flask cultivation, the seed culture (5%) was inoculated into 500 mL baffled flasks containing 50 mL of liquid medium (12 g/L tryptone, 24 g/L yeast extract, 12.53 g/L K₂HPO₄, 2.31 g/L KH₂PO₄, and 20 g/L glucose) supplemented with kanamycin (50 μg/mL). The cultures were incubated at 37°C and 220 rpm for approximately 24 h.

Batch fermentation was performed in a 5-L bioreactor (5BG, Baoxing, China). The seed culture (5%) was inoculated into a 2-L liquid medium. The culture temperature was maintained at 37°C, and the initial pH was set at approximately 7.0. The airflow rate was set at 1 L/min, and the dissolved oxygen (DO) content was maintained at 30% by automatically varying the agitation speed (400–800 rpm). During the initial stage of fermentation, the OD₆₀₀ of the fermentation broth was similar to that of the blank culture medium, i.e., close to zero.

The inoculation amount, liquid medium composition, and DO content for fed-batch fermentation in a 5-L fermenter were identical to those for batch fermentation. The pH was automatically maintained at approximately 6.5 using 2 M HCl and NaOH solution after fermentation for 4 h. The residual glucose (RG) content was maintained at 10 g/L by adding a glucose solution (400 g/L) when the RG content was reduced to approximately 8 g/L. A biosensor (SBA-40C, Institute of Biology, Shandong Academy of Sciences) was used to quantitate RG content.

To confirm the relative transcriptional levels of PncB, NadE, and PnuC, we performed qRT-PCR. Total RNA was extracted from different B. subtilis strains cultured in a shake flask for 12 h using Bacterial RNA Kit (Omega Bio-Tek, Inc.). Reverse transcription was performed using PrimeScript^™ RT Reagent Kit (Takara Bio Inc.). qRT-PCR was performed using the StepOnePlus Real-Time PCR System (Applied Biosystems, Inc.). The specific primers designed in this study are listed in Supplementary Table S3. The 16S rRNA gene encoding the major sigma factor in B. subtilis was used as the internal control (Klappenbach et al., 2000), and relative quantification was performed using the comparative cycle threshold method (Proctor et al., 2018).

To quantify extracellular and intracellular NMN, 1 mL of cultured cells was collected and centrifuged for 5 min at 4000 rpm and 4°C. The supernatant was used as the extracellular sample. After discarding the supernatant, the bacterial pellet was resuspended in 1 mL of lysozyme solution (4 mg/mL), and the mixture was incubated for 30 min at 37°C. Cell lysis was performed on ice via ultrasonication (400 W) for 30 cycles (each for 3 s with 2 s intervals), and the suspension was centrifuged for 5 min at 6000 rpm and 4°C. The supernatant was collected as the intracellular sample. To measure total NMN titer, lysozyme was added to the culture broth, and the mixture incubated for 30 min at 37°C. The whole-cell lysate was ultrasonicated and centrifuged as described above. The supernatants were collected and used as total NMN samples.

All NMN samples were filtered through a 0.22 μm PES membrane before analysis using an HPLC instrument (Agilent 1,260 Series) equipped with a TSKgel ODS-80TS quinolinic acid (QA) column (4.6 mm × 25 cm, 5 μm, Tosoh). The mobile phase comprised methanol and ion pair aqueous solution (5.4 g of KH₂PO₄ and 1.7 g of tetrabutylammonium hydrogen sulfate were dissolved in 1000 mL of water; pH 6.0) at a ratio of 30:70, with a flow rate of 1 mL/min. The eluates were monitored at 260 nm, and the column temperature was set at 28°C (Stocchi et al., 1987). NAM, nicotinic acid (NA), NR, NAD⁺ standard sample processing methods were the same as described above. The standard NMN and NAD⁺ samples were diluted to different concentrations, and a standard curve was constructed to calculate NMN and NAD⁺ contents.

The NMN biosynthesis pathway primarily uses NAM or NR as a precursor (Figure 1), which requires high production costs. Nicotinic acid mononucleotide (NaMN) biosynthesis can be achieved via two main routes (Figure 2): (1) biosynthesis of NaMN from NA via nicotinic acid phosphate ribose transferase (PncB); (2) conversion of QA, which is biosynthesized from glucose through the TCA cycle, into NaMN. NMN synthesis via esterification of NA to NaMN and conversion of the hydroxyl group of NaMN to an amino group is a pathway for NMN production in B. subtilis.

PncB can increase the NaMN production rate (Wünsche et al., 2012), thereby promoting NMN biosynthesis. NadE catalyzes the direct synthesis of NMN from NaMN and participates in the NAD⁺ salvage pathway to further enhance NMN production. To decrease production costs, we increased the expression of PncB in B. subtilis WB600 and used glucose as a substrate to produce NMN without adding expensive substrates. The pMA5 plasmid contains two MCSs (MCS1 and MCS2). To compare the expression effects of MCS1 and MCS2, we constructed recombinant plasmids pMA5-PncB1 and pMA5-PncB2 (Figure 3), which were subsequently transformed into B. subtilis WB600, resulting in the generation of two genetically engineered strains B. subtilis PncB1 and B. subtilis PncB2. After culture for 12 h in the shake flask, the strains B. subtilis WB600, B. subtilis PncB1, and B. subtilis PncB2 produced 720.05, 1160.80, and 1097.10 mg/L NMN, respectively (Figure 4A). Subsequently, we further quantified the NAD⁺ contents of strains B. subtilis PncB1 and B. subtilis WB600 (Supplementary Figure S1). After shake-flask fermentation, the NAD⁺ content of the strain B. subtilis PncB1 (265.41 mg/L) at 12 h was 37.2% higher than that of the control strain WB600 (193.46 mg/L). These results indicate that PncB overexpression positively impacts the production of NMN and NAD⁺ by these engineered strains, and MCS1 appears to have a superior overexpression effect compared with MCS2.

As shown in Figure 1, the de novo synthesis of NMN is mediated by nadD and NadE, leading to the production of NAD⁺. After NAD⁺ generation through this pathway, nudE can degrade NAD⁺ into NMN (Rodionov et al., 2008; Osterman, 2009), and NadE participates in the direct conversion of NaMN to NMN. Sorci et al. revealed that NadE encoded by FtNadE can directly catalyze the conversion of NaMN into NMN in Francisella tularensis (Sorci et al., 2009), without the need to participate in the NAD⁺ remedial synthesis pathway. To enhance NMN biosynthesis, in addition to overexpressing PncB in B. subtilis to increase the conversion of NaMN to NMN, we attempted to overexpress NadE in B. subtilis (Figure 2). The recombinant plasmid pMA5-NadE was constructed (Figure 3) to overexpress NadE. This plasmid was transformed into B. subtilis WB600 to obtain the engineered strain B. subtilis NadE. The NMN titer of B. subtilis NadE reached approximately 969.69 mg/L after 12 h of fermentation in the flask, which was 34.67% higher than that of the host strain B. subtilis WB600. NadE and PncB were inserted at different MCSs of the plasmid pMA5 and overexpressed in the host strain B. subtilis WB600, resulting in the generation of strains B. subtilis NadE-PncB1 and B. subtilis NadE-PncB2, which produced 821.05 and 941.38 mg/L NMN, respectively. However, compared with the individual overexpression of PncB or NadE, their combined overexpression did not enhance the NMN titer (Figure 4A). The shake-flask culture results indicated that genetically engineered strains increased their NMN titers by 16–64% compared with the host strain B. subtilis WB600 (Figure 4A).

There are several pathways for the synthesis and recycling of NMN in B. subtilis cells, and intracellular NMN is utilized for bacterial growth (Magni et al., 1999). To increase the NMN titer, in addition to enhancing the expression of key enzymes in the NMN synthesis pathway, it is necessary to transport intracellular NMN to the extracellular environment to avoid degradation. Studies on transporter proteins that transport NMN to the extracellular environment (Grose et al., 2005; Shoji et al., 2021) have shown that overexpression of PnuC from Bacillus mycoides can efficiently increase the NMN titer (Sauer et al., 2004; Rodionov et al., 2008). As shown in Figure 4A, B. subtilis PncB1 exhibited the highest titer among the five engineered strains, with a 64% increase in titer relative to that in B. subtilis WB600. Therefore, PnuC was inserted into pMA5-PncB1 to generate the plasmid pMA5-PncB1-PnuC (Figure 3). In a study on transporter proteins, the control group comprised the strains B. subtilis WB600 and B. subtilis PncB1, whereas the experimental group contained the strain B. subtilis PncB1-PnuC. B. subtilis WB600, B. subtilis PncB1, and B. subtilis PncB1-PnuC were cultured under the same conditions. The titers of total NMN, extracellular NMN, and intracellular NMN produced by different strains were determined. As shown in Figure 4B, the total NMN titers of B. subtilis PncB1 and B. subtilis PncB1-PnuC were significantly higher than the total NMN titer of B. subtilis WB600. However, B. subtilis PncB1-PnuC (1160.24 mg/L) did not show a considerable increase in total NMN titer compared with B. subtilis PncB1 (1160.80 mg/L). Further analysis revealed that extracellular NMN accumulation of B. subtilis PncB-PnuC accounted for 83.69% of the total NMN titer. The extracellular NMN titers of B. subtilis WB600 and B. subtilis PncB1 reached only 22.17 and 26.32% of the total NMN titer, respectively. At the time point of the highest cumulative titer, the extracellular NMN titer of B. subtilis PncB1-PnuC was significantly higher than the corresponding intracellular titer. In contrast, the extracellular NMN titers of B. subtilis WB600 and B. subtilis PncB1 were significantly lower than the corresponding intracellular titers. To further verify that NMN was produced through strain fermentation, we measured the contents of NMN, NA, NAM, and NR in the blank culture medium. The results indicated that the culture medium contained trace amounts of NA (<29.7 mg/L) and NAM (<160 mg/L), further confirming that the engineered strains can utilize glucose to biosynthesize NMN (Supplementary Figure S2).

Taken together, the results indicate that PnuC is extremely effective in transporting intracellular NMN to the extracellular environment. PnuC-mediated excretion of NMN reduced intracellular NMN titers, thereby accelerating the conversion of NAM to NMN and preventing the undesired conversion of NMN to other compounds. Furthermore, even without exogenous NMN transporter proteins, a small amount of NMN accumulation was detected in the extracellular area of B. subtilis. However, similar results have not yet been reported for endogenous NMN transporters. The actual transmembrane transport process of wild-type bacterial NMN should be experimentally verified.

To study the effect of the transcriptional levels of PncB, NadE, and PnuC on NMN production, the expression of each gene was analyzed in recombinant and control strains of B. subtilis WB600, both of which were cultured under identical conditions for 12 h and analyzed via qRT-PCR. The results indicated that PncB transcriptional levels of the three engineered bacteria were 1.5–4.4-fold higher than those of the control strain (Figure 5A), with their NMN titers being increased to varying degrees (Figure 4A). These data support the hypothesis that PncB overexpression plays an important role in NMN biosynthesis. As shown in Figure 5A, the transcriptional level of B. subtilis PncB1 was higher than that of B. subtilis PncB2, which is consistent with the higher NMN titer of B. subtilis PncB1 (Figure 4A). This result showed that the number of terminators following MCS1 and MCS2 probably affects the gene expression intensity in B. subtilis (Figure 3). Furthermore, Lin et al. revealed that the tandem terminator can reinforce the heterologous gene expression (Lin et al., 2021a). The transcriptional levels of NadE in the three recombinant strains were significantly higher than those in the control B. subtilis WB600, indicating the successful construction of NadE-overexpressing recombinant strains (Figure 5B). However, the overexpression of NadE alone and coexpression with PncB did not significantly increase the NMN production of B. subtilis NadE, B. subtilis NadE-PncB1, and B. subtilis NadE-PncB2 (Figure 4A). As shown in Figure 2, NadE is involved in two conversion pathways from NaMN to NMN (Foster and Moat, 1980; Begley et al., 2001). This may lead to the overexpression of NadE, which is not as effective as PncB.

B. subtilis PncB1-PnuC exhibited a significant increase in PnuC expression compared with B. subtilis WB600 and B. subtilis PncB1 (Figure 5C). PnuC overexpression increased the extracellular NMN titer of B. subtilis PncB1-PnuC (Figure 4B). Overall, the results indicate that increasing the transcriptional levels of PncB and PnuC can efficiently enhance the production of NMN.

The total NMN titers of B. subtilis PncB1 and B. subtilis PncB1-PnuC obtained via flask fermentation were similar. To further verify whether the functional activity of PnuC can improve the biosynthesis of NMN, batch fermentation and fed-batch fermentation strategies were employed. The engineered strains B. subtilis PncB1 and B. subtilis PncB1-PnuC and the host strain B. subtilis WB600 were cultivated in a 5-L fermenter using the batch fermentation method. The pH, OD₆₀₀, RG content, and NMN titer were measured during fermentation (Figures 6A–C). The total NMN titer of the control B. subtilis WB600 strain was 849.31 mg/L at 16 h, whereas the NMN titers of the two engineered strains reached 1,135 (24 h) and 1,245 (24 h) mg/L, respectively. The cell concentration (OD₆₀₀) of the strains B. subtilis WB600, B. subtilis PncB1, and B. subtilis PncB1-PnuC reached 20.12, 25.66, and 26.32, respectively, after 20 h of batch culture (Figures 6A–C). NMN production in the engineered stains was slightly delayed by approximately 4 h compared with that in the control strain WB600, which corresponded to the overexpression of PncB and PncB-PnuC. After glucose consumption, pH initially decreased and then increased with the gradual increase in OD₆₀₀ values. In the 5-L bioreactor, the extracellular NMN titer of B. subtilis PncB1-PnuC relative to the total NMN titer was 81.06%, whereas the titers of B. subtilis WB600 and B. subtilis PncB1 reached only 21.17 and 25.32% of the total NMN titer, respectively (Figure 7A).

To further improve the NMN titer, the engineered strains were cultured in a 5-L fermenter via fed-batch fermentation. The OD₆₀₀, pH, RG content, and NMN titer measured during fermentation are shown in Figures 6D–F. The control strain B. subtilis WB600 exhibited the highest NMN accumulation of 1,612 mg/L at 20 h. The NMN titer of the genetically engineered strain B. subtilis PncB1 was 2,598 mg/L, whereas that of B. subtilis PncB1-PnuC reached a maximum of 3,396 mg/L at 32 h. Overexpression of PncB1 in B. subtilis PncB1 and co-overexpression of PncB1 and PnuC in B. subtilis PncB1-PnuC resulted in an increase in the total NMN titer by 61.17 and 110.67%, respectively, compared with the host strain B. subtilis WB600. The extracellular NMN titer of B. subtilis PncB1-PnuC accounted for 76.96% of the total NMN titer, and the extracellular NMN titers of B. subtilis WB600 and B. subtilis PncB1 accounted for only 19.57 and 23.12%, respectively, of the total NMN titer (Figure 7B). Under fed-batch fermentation conditions, the glucose consumption rate of the genetically engineered strains was slightly higher than that of the parental strain. The growth rates and biomass of B. subtilis PncB1 and B. subtilis PncB1-PnuC were also higher than those of the parental strain. As shown in Figures 6D–F, the NMN titers of B. subtilis PncB1 and B. subtilis PncB1-PnuC strains peaked at 36 h during fed-batch fermentation, which was delayed by 16 h compared with the WB600 strain (20 h). After entering the decay stage, as the bacterial cells autolyze, NMN production may suddenly and rapidly decrease. Consequently, it is necessary to terminate fermentation in a timely manner to obtain the optimal NMN production (Figure 6). In addition, the NMN production rate of B. subtilis PncB1-PnuC during fed-batch fermentation (106.12 mg/[L h]) increased by 30.72% compared with that during batch fermentation (81.18 mg/[L h]). Thus, the fed-batch fermentation strategy markedly increased the total and extracellular NMN titers of B. subtilis PncB1-PnuC.