Skip to main content
NCBI home page
Microbial Biotechnology logoLink to Microbial Biotechnology
. 2024 Nov 9;17(11):e70048. doi: 10.1111/1751-7915.70048

De novo biosynthesis and nicotinamide biotransformation of nicotinamide mononucleotide by engineered yeast cells

Yanna Ren 1, Bei Han 1, Shijie Wang 1, Xingbin Wang 1, Qi Liu 1,✉, Menghao Cai 1,2,3,✉
PMCID: PMC11549688  PMID: 39520312

Abstract

β‐Nicotinamide mononucleotide (NMN) is a precursor of NAD+ in mammals. Research on NAD+ has demonstrated its crucial role against aging and disease. Here two technical paths were established for the efficient synthesis of NMN in the yeast Pichia pastoris, enabling the production of NMN from the low‐cost nicotinamide (NAM) or basic carbon sources. The yeast host was systematically modified to adapt to the biosynthesis and accumulation of NMN. To improve the semi‐biosynthesis of NMN from NAM, nicotinamide phosphoribosyltransferases were expressed intracellular to evaluate their catalytic activities. The accumulation of extracellular NMN was further increased by the co‐expression of an NMN transporter. Fine‐tuning of gene expression level produced 72.1 mg/L NMN from NAM in flasks. To achieve de novo biosynthesis NMN, a heterologous biosynthetic pathway was reassembled in yeast cells. Fine‐tuning of pathway nodes by the modification of gene expression level and enhancement of precursor generation allowed efficient NMN synthesis from glucose (36.9 mg/L) or ethanol (57.8 mg/L) in flask. Lastly, cultivations in a bioreactor in fed‐batch mode achieved an NMN titre of 1004.6 mg/L at 165 h from 2 g NAM and 868 g glucose and 980.4 mg/L at 91 h from 160 g glucose and 557 g ethanol respectively. This study provides a foundation for future optimization of NMN biosynthesis by engineered yeast cell factories.

Metabolic engineering of P. pastoris achieves NMN production by semi‐biosynthesis and de novo biosynthesis, which represents the highest yield of de novo synthesis of NMN from basic carbon sources yet reported.

graphic file with name MBT2-17-e70048-g005.jpg

INTRODUCTION

The β‐nicotinamide mononucleotide (NMN) is a key intermediate of nicotinamide adenine dinucleotide (NAD+), which is an essential coenzyme for cellular redox reactions in human bodies and other mammals. NAD+ plays a pivotal role in diverse biological processes, encompassing gene expression, DNA repair, energy generation, etc. It also serves as a regulatory factor in metabolic pathways, specifically the tricarboxylic acid cycle and glycolysis (Yoshino et al., 2021). Recent studies revealed that administering NMN significantly mitigated age‐related physiological deterioration in mice and reversed associated mitochondrial dysfunction (Gomes et al., 2013; Mills et al., 2016). Besides, it has been proved to be an efficacious treatment for type II diabetes resulting from a high‐fat diet (Yoshino et al., 2011). Recent clinical trials have demonstrated that the addition of NMN can enhance glucose metabolism in human skeletal muscle (Yoshino et al., 2021). Given its profound effects, the NMN has been recognized as a popular raw material for pharmaceutical, food and cosmetic additives. Thus, it is of good value to develop a cost‐effective and efficient bioproduction process of NMN that meets biosafety concerns.

The natural biosynthesis of NMN exhibits distinct pathways that can be grouped into: (1) In mammals (Guo et al., 2017; Lin et al., 2016) and some bacteria (Lee et al., 2017), NMN is synthesized by nicotinamide (NAM) phosphoribosyltransferase (Nampt), using NAM and 5‐phosphoribosyl 1‐pyrophosphate (PRPP) as starting materials; (2) For bacteria lack of Nampt, NMN is predominantly produced via the degradation of NAD+ (Göckel & Richert, 2015); (3) In bacteria lack of both of the above routes, Francisella tularensis as an example, NMN synthase (FtNadE) facilitates the conversion of nicotinic acid mononucleotide (NaMN) to NMN (Sorci et al., 2009); (4) In certain mammals, NMN is synthesized from nicotinamide riboside (NR) via a phosphorylation process catalysed by nicotinamide riboside kinase (Nrk) (Yoshino et al., 2018).

Recently, NMN production by biotransformation from key intermediates has emerged as a preferred strategy as compared to chemical synthesis. Liu et al. (2021) and Maharjan et al. (2021) synthesized NMN of 496.2 and 771.5 mg/L, respectively, from NAM through whole‐cell catalysis. Shoji et al. (2021) further identified an NMN transporter (BMpnuC) from Bacillus mycoides which improved NMN production to 6.79 g/L from NAM via whole‐cell catalysis. Huang et al. (2022) achieved the fermentation semi‐synthesis of NMN by metabolic engineering of the semi‐biosynthetic pathway in Escherichia coli. A nicotinamide phosphoribosyltransferase (Nampt) from Vibrio bacteriophage KVP40 (VpNadV) was identified and a remarkable NMN yield of 16.2 g/L was achieved with NAM feedings. Besides, researchers also reported the semi‐synthesis of NMN from an alternative substrate nicotinic acid (NA), but it is non‐preferred because of the high cost of NA and its low transformation efficiency.

Although high‐level NMN production was achieved by semi‐biosynthesis in E. coli, it poses safety concerns about the potential presence of endotoxins in food applications and entails a strict purification process. Therefore, FDA generally recognized as safe (GRAS) strains could be used as alternative hosts. Pichia pastoris (syn. Komagataella phaffii), is a GRAS yeast host that has been widely used for the production of drugs and food additives. It has emerged as a versatile industrial strain with the advantages of simple medium, mild post‐translational modifications, strong secretion capabilities and good multi‐enzyme co‐expression capacities in cells (Meehl & Stadheim, 2014; Peña et al., 2018; Safder et al., 2018). In contrast, the reported technical routes relied on semi‐biosynthetic production of NMN that needs NAM as the essential substrate. Currently, NAM is mainly produced via chemical synthesis methods, which makes the NMN semi‐biosynthesis actually a chem‐bio coupling strategy. Furthermore, NMN can be naturally generated via its inherent NAD+ salvage pathway, and this process can be capitalized upon as an alternative NMN biosynthetic route. The de novo NMN biosynthesis route from aspartate acid (Asp) was explored, and 153 mg/L NMN was obtained from 1.00 g/L‐Asp substrate in recent study (Wang et al., 2024). However, the endogenous de novo biosynthesis pathway of NMN is typically initiated from proteinogenic amino acids and tightly regulated in microbial cells (Begley et al., 2001). Most recently, an NAD+ biosynthetic pathway starting from chorismate has been constructed to achieve high‐level NAD(H) in E. coli (Ding et al., 2021). Chorismate serves as an exceptional precursor due to its inherent metabolic versatility as a natural branching point for cell metabolism (Dosselaere & Vanderleyden, 2001). This inspires us to construct a heterologous chorismite‐quinolinic acid (QA) pathway in yeast cells, followed by QA‐NaMN conversion by endogenous QA phosphoribosyl transferase (NadC) (Figure 1), which can be further fused to a downstream NaMN‐NMN step catalysed by NMN synthase. As a result, a de novo synthesis of NMN could be realized by engineered cell culture with basic carbon sources.

FIGURE 1.

FIGURE 1

Open in a new tab

Metabolic engineering strategies for the biosynthesis of NMN. Route 1, semi‐biosynthesis of NMN by feeding NAM during cell culture; Route 2, de novo biosynthesis of NMN by construction of heterologous NMN synthesis pathway. PEP, phosphoenolpyruvate; E4P, erythrose 4‐phosphate; R5P, ribose‐5‐phosphate; DAHP, 3‐deoxy‐D‐arabino‐heptulosonate‐7‐phosphate; EPSP, 5‐enolpyruvylshikimate 3‐phosphate; CHO, chorismate; AIDC, 2‐amino‐2‐deoxyisochorismate; DHHA, 2,3‐dihydro‐3‐hydroxyanthranilic acid; ACMS, 2‐amino‐3‐carboxymuconate semialdehyde; 3‐HAA, 3‐hydroxyanthranilic acid; PRPP, 5‐phosphoribosyl diphosphate; NaMN, nicotinic acid mononucleotide; QA, quinolinic acid; NA, nicotinic acid; NR, nicotinamide riboside. PhzE, ADIC synthase; PhzD, DHHA synthase; DhbX, DHHA dehydrogenase; NbaC, 3‐HAA 3,4‐dioxygenase; NadC, QA phosphoribosyl transferase; NadE, NMN synthase; Nampt, nicotinamide phosphoribosyltransferase; NiaP, transporter proteins for NAM; PnuC, transporter proteins for NMN; Aro4 and Aro3, DAHP synthetase; Aro1, pentafunctional aromatic protein; Aro2, chorismate synthase; PRS, PRPP synthase; 0361, nicotinamidase encoded by PAS_chr3_0361; 0598, hypothetical protein encoded by PAS_chr2‐1_0598; 0359, ribosylnicotinamide kinase encoded by PAS_chr2‐2_0359; 0685, nucleotidase encoded by PAS_chr2‐1_0685.

In this study, we aim to enable semi‐biosynthesis and de novo biosynthesis of NMN in engineered P. pastoris and evaluate the production efficiencies of the two different technical routes (Figure 1). Systematic metabolic engineering strategies were involved to deplete the sinking branches of NMN and strengthen the generating pathways. Finally, an equivalent titre of NMN (~1000 mg/L) was obtained for either semi‐biosynthesis with NAM addition or de novo biosynthesis from basic carbon sources by yeast cells in 3‐L bioreactors. Our study provides insightful references for future advancement in the total biosynthesis of NMN.

EXPERIMENTAL PROCEDURES

Strains, media and growth conditions

The plasmids and strains used in this study are listed in Table S1 and Table S2. Escherichia coli Top10 was used for plasmids storage and propagation. The expression vectors of pGAPZ B and pUC18 were stored in our laboratory. All chemical reagents were purchased from Tansoole (Shanghai, China). The E. coli and yeast strains were cultured as previously described (Ren et al., 2020). For plate culture, agar of 2% [w/v] was added. First, 20 μL storage strain (−80°C) was inoculated into 3 mL YPD medium in a 20‐mL serum bottle and cultured for 36–48 h. Then 100 μL cell broth was inoculated into 10 mL YPD medium and cultured to an OD600 of 6.0. The cells were harvested by centrifugation at 5000g for 3 min and resuspended in sterile water, and inoculated in fermentation medium at a final density (OD600) of 1.0. The cell culture was carried out in a 6‐well deep plate (10 mL medium in each well of 50 mL) for fermentation production of NMN.

The fermentation medium consists of 2% tryptone, 1% yeast extract, 2% glucose, 0.1% MgSO4·7H2O, 0.6% KH2PO4 and 1.64% K2HPO4, 0.1% sodium citrate and 0.05% NAM (w/v for all). In addition, glucose was added at a final concentration of 2% [w/v] every 24 h. For fed‐batch fermentation, strains were cultured in 200 mL YPD medium in 1‐L shake flasks for seed preparation. After incubation for 12–16 h, the seed culture was transferred into a 3‐L fermenter containing 2 L fermentation medium. The pH was maintained at 4.5 during the batch phase and shifted to 7.0 during the fed‐batch phase by the automated control of NH4OH feeding (50%, v/v). Dissolved oxygen (DO) was maintained at above 30% by automatical control of aeration and agitation.

Construction of plasmids and strains

Abundant of plasmids and recombinant strains were constructed in this work. The construction details of plasmids and strains were described in the file of Supporting information. All the plasmids and strains with corresponding genotype information were listed in Table S1 and Table S2, respectively. Primers used for plasmid construction are listed in Table S3. Gene manipulation was carried out by our previously developed CRISPR‐Cas9 editing tools (Liu et al., 2019a). Information on enzymes and transporter proteins for expression in P. pastoris are listed in Table S4.

Analytical methods

NMN content in a 6‐well deep plate was analysed referring to the reported fluorometric derivatization method (Marinescu et al., 2018). The broth sample of 69 μL was dispensed into 96‐well plates, followed by addition of 28 μL acetophenone solution (dissolved in DMSO by 20%, v/v) and 28 μL KOH solution (2 mol/L). After incubating on ice for 2 min, 125 μL of formic acid solution (88%, v/v) was added into each well. Subsequently, the plates were subjected to a 10‐min incubation at 37°C. The UV emission intensity at 445 nm was then measured using microplate reader (BioTek Synergy 2), with an excitation wavelength of 382 nm. NMN production in a bioreactor was analysed by high‐performance liquid chromatography as previously described (Huang et al., 2022). A reverse‐phase column (YMC‐Triart C18, 4.6 × 250 mm) was eluted by 30 mM KH2PO4: methanol (95:5, v/v) at a flow rate of 1 mL/min, and the detector wavelength was set at 254 nm. Cell wet weight was analysed according to Xu et al. (2019).

Statistical analysis

The data were obtained from three biological replicates assayed in triplicate and presented as mean ± standard deviation. Experiments were carried out at least twice. The independent samples Student's t‐test was performed to determine the differences among grouped data. Statistical significance was assessed at p < 0.05.

RESULTS

Construction of heterologous pathway for the semi‐biosynthesis of NMN

As P. pastoris lacks enzymes for NAM‐to‐NMN transformation in native cells, three different Nampt enzymes derived from mammals and bacteria were first introduced into this host (Figure 2A). Codon‐optimized genes encoding Nampt‐HD from Haemophilus ducreyi (Marinescu et al., 2018), Nampt‐MM from Mus musculus (Marinescu et al., 2018) and Nampt‐CP from Chitinophaga pinensis (Shoji et al., 2021) were synthesized and expressed in the Δku70 mutant of P. pastoris (partially deficient in non‐homologous DNA end‐joining, Liu et al., 2019a). Western blot analysis conducted on the cell lysates of each transformant proved successful expression of all the tested Nampt enzymes, albeit with variations in their expression levels (Figure 2B). The NMN production capability of each strain is roughly correlated with its Nampt expression level. The Nampt‐CP from C. pinensis allowed the highest NMN production (Figure 2C,D), which was also in accordance with the previous study (Shoji et al., 2021). The titre of the Nampt‐CP expression strain reached up to 13.3 mg/L at 96 h (fed with 0.5 g/L NAM), which was increased by 71% compared to the parent strain Δku70 (Figure 2C). In addition, cell growth of the Nampt‐CP expression strain increased slightly in comparison to the Δku70 (Figure S1). Besides, the parent strain seemed to produce NMN at a very low level by an endogenous de novo synthetic pathway (Δku70 with and without NAM addition in Figure 2C). Overall, the Nampt‐CP expression strain with the highest NMN titre was selected for further experiments.

FIGURE 2.

FIGURE 2

Open in a new tab

Semi‐biosynthesis of NMN by rewiring P. pastoris cells with NAM feeding. (A) Basic engineering scheme for NMN semi‐biosynthesis; (B) Western blot analysis of Nampt enzymes from different microorganisms expressed in P. pastoris. Nampt‐MM from M. musculus, 56.7 kDa; Nampt‐HD from H. ducreyi, 55.8 kDa; Nampt‐CP from C. pinensis, 53.6 kDa; Nampt‐VP from V. bacteriophage, 55.6 kDa; (C) NMN titres of strains with expression of Nampt enzymes and transporter proteins from different sources; (D) NMN titres of strains constructed from (c) but with NMN / NAM sinking pathways inactivated. (E) NMN titres of strains with improved expression of Nampt enzymes and transporter proteins. Δku70, deficiency of KU70 which is a critical gene for non‐homologous‐end‐joining (NHEJ) repairing in P. pastoris; Nampt‐MM, Nampt‐HD, Nampt‐CP, Nampt‐VP, CP, NiaP and PnuC were generally overexpressed by the constitutive P GAP promoter, unless the stronger P OcA promoter system was used with special illustrations. CP‐NiaP, co‐expression of Nampt‐CP and NiaP; CP‐NiaP‐PnuC (CNP), co‐expression of Nampt‐CP, PnuC and NiaP; Δ0361, knockout of PAS_chr3_0361; Δ0598, knockout of PAS_chr2‐1_0598; Δ0359, knockout of PAS_chr2‐2_0359; Δ0685, knockout of PAS_chr2‐1_0685; 4Δ, knockout of PAS_chr3_0361, PAS_chr2‐1_0598, PAS_chr2‐2_0359 and PAS_chr2‐1_0685; Functions of these genes or proteins are described in Figure 1. Statistical significance values of NMN‐producing strains at different time intervals are shown; the experiments were performed 3 times as described; the mean values ± SD are presented. Bars with the same letters are not significant (p < 0.05) ‐, not deleted or not expressed; +, deleted or expressed.

To increase the production of NMN, it is crucial to enhance NAM uptake and NMN effluxion mediated by transporter proteins. Thus codon‐optimized genes for effective NAM and NMN transporters, NiaP derived from Burkholderia cenocepacia and PnuC from Bacillus mycoides (Shoji et al., 2021), were synthesized and co‐expressed in the Nampt‐CP expression strain, resulting in strains of CP‐NiaP (CN) and CP‐NiaP‐PnuC (CNP), respectively. After 96 h culture, the CN and CNP strains produced 12.9 and 14.5 mg/L NMN, respectively, with 0.5 g/L NAM feeding during culture (Figure 2C). That is, expression of NiaP did not promote the accumulation of NMN in the Nampt‐CP strain, while further expression of PnuC (CNP) increased NMN production by 9%. Besides, cell growth of the CNP strain also increased slightly (Figure S1).

Although the CNP strain well transformed NAM to NMN, it was observed that the exogenously added NAM was easily converted into NA (Figure S2). In our engineered cells, NA directly compete with NMN for the substrate of NAM (Figure 2C). Also, microbial cells commonly incorporate complicated metabolic networks to sustain stable levels of intracellular NAD+/NADH. Therefore, it is necessary to rewire metabolic pathways for the accumulation of NMN. The NAM and NMN metabolic pathway genes in P. pastoris were then identified from the KEGG and Genbank databases, that is, PAS_chr3_0361 (0361), PAS_chr2‐1_0598 (0598), PAS_chr2‐2_0359 (0359) and PAS_chr2‐1_0685 (0685). These genes were knocked out, resulting in strains of CPNΔ0361, CPNΔ0361Δ0598, CPNΔ0361Δ0598Δ0359 and CPNΔ0361Δ0598Δ0359Δ0685 (CPN4Δ). Knockout of these genes did not show growth damage on cells (Figure S3). Of note, the titre of CPNΔ0361Δ0598Δ0359 and CPN4Δ was achieved as 52.2 and 56.9 mg/L at 96 h, which was 2.6 and 2.9 folds higher than that of the CNP strain respectively (Figure 2D).

To improve the efficiency of NMN synthesis, the expression levels of different genes were further optimized. The native constitutive promoter P GAP and a synthetic constitutive transcriptional device CSAD (7.2 times stronger than P GAP on glucose, Liu et al., 2022) with a hybrid promoter of P OcA were employed to enhance enzyme expression. A Nampt‐VP from the V. bacteriophage KVP40 well‐improved NMN synthesis in E. coli (Huang et al., 2022). Thus an Nampt‐VP expression cassette under the control of P GAP was introduced into the CPN4Δ to generate a strain of CPNΔ0361Δ0598Δ0359Δ0685‐VP (CPN4Δ‐VP), for which the NMN titre reached 72.1 mg/L at 96 h and was further increased by 26.7% (Figure 2E). In addition, the expression of Nampt‐VP did not affect the cell growth of the recombinant strain (Figure S4).

As overexpression of Nampt functioned well, its expression level was further improved by introducing Nampt‐CP and Nampt‐VP expression cassettes under the control of P OcA into the CPN4Δ‐VP strain. Further expression of Nampt‐VP showed a minor impact on cell growth, but further expression of Nampt‐CP led to biomass decrease (Figure S4). Although efforts were made to promote the expression of Nampt‐CP and Nampt‐VP further, the NMN synthesis was not improved but weakened as compared to the CPN4Δ‐VP strain (Figure 2E). It was speculated that transport of substrates and products might be limited. Therefore, we further enhanced the expression of the transporter proteins of NiaP and PnuC. Unfortunately, the NMN synthesis was still not improved but somehow decreased (Figure 2E). On the contrary, overexpression of NiaP, PnuC and Nampt‐CP under the control of P OcA was introduced into Δku70Δ0361Δ0598Δ0359Δ0685 to enhance the expression of the functional proteins. The NMN titre of this strain reached up to 75.2 mg/L at 96 h, which was increased by 32.2% in comparison to the CPNΔ0361Δ0598Δ0359Δ0685 strain (Figure 2E). Nevertheless, the NMN titre was similar to that of the CPN4Δ‐VP strain (Figure 2E). In addition, cell growth of this strain became worse (Figure S4) after 72 h culture comparing with the CPN4Δ‐VP strain. Besides, overexpression of Nampt‐CP or Nampt‐VP by the P OcA were tested in the Δku70Δ0361Δ0598Δ0359Δ0685, it seemed only the Nampt‐CP positively affected NMN synthesis in contrast to the parent strain (Figure 2E). In total, these results indicated that a moderate expression of the functional proteins was preferred for NMN synthesis and the CPN4Δ‐VP was chosen as the final strain for bioreactor fermentation.

Design and construction of the de novo biosynthesis of NMN

The semi‐biosynthesis of NMN needs NAM as the direct substrate, which costs much higher than the basic carbon sources. Also, NAM for the industrial use is mainly obtained by chemical steps. To decrease cost and realize full bioproduction, we also attempted to develop a de novo biosynthetic method for NMN production from basic carbon sources. The endogenous de novo synthetic pathway of NMN in P. pastoris starts from amino acids and is tightly regulated by the cells. We, therefore, aim to establish a heterologous de novo synthetic pathway that initiates from the abundant chorismate in cells, and achieves efficient synthesis and accumulation of NMN (Figure 3A).

FIGURE 3.

FIGURE 3

Open in a new tab

De novo biosynthesis of NMN by equipping P. pastoris cells with heterologous synthetic pathways. (A) Endogenous and heterologous biosynthesis pathways of NMN in engineered P. pastoris cells; (B) Western blot analysis of enzymes in NMN de novo synthetic pathway expressed in P. pastoris. NadE, 29 kDa; NbaC, 21 kDa; PhzE, 69 kDa; NadC, 33 kDa; PhzD, 23 kDa; DhbX, 27 kDa; (C) Improved NMN production systematic metabolic engineering strategy. Δku70, Δ0361, Δ0598, Δ0359 and Δ0685 were the same as that in Figure 2. PhzE, PhzD, DhbX, NbaC, NadC, NadE, Nampt, PnuC, Aro4, Aro3K222L and PRS were expressed by the constitutive P GAP promoter except for the expression of an additional Aro3K222L illustrated specially. Functions of these genes or proteins are described in Figure 1. Other metabolites were with the same full name as illustrated in Figure 1. Statistical significance values of NMN‐producing strains at different time intervals are shown; the experiments were performed 3 times as described; the mean values ± SD are presented. Bars with the same letters are not significant (p < 0.05). ‐, not deleted or not expressed; +, deleted or expressed. L‐Try, L‐tryptophan, L‐Asp, L‐aspartate; Imino‐Asp, iminoaspartate; N‐formyl‐L‐Kyn, N‐formyl‐L‐kynurenine; L‐Kyn, L‐kynurenine; 3‐hydroxy‐L‐Kyn, 3‐hydroxy‐L‐kynurenine; NaAD, deamido‐NAD.

The Δku70Δ0361Δ0598Δ0359Δ0685 strain with four genes knocked out, that is, PAS_chr3_0361, PAS_chr2‐1_0598, PAS_chr2‐2_0359 and PAS_chr2‐1_0685 (Figure 2E), was selected for construction of the de novo synthesis host. Deficiency of the four genes did not affect cell growth of the strain (Figure S5), but increased endogenous NMN production from 7.1 to 11.1 mg/L (Figure 3C). As the previous data revealed the better effects of moderate enzyme expression, the P GAP was then selected for gene expression in this construct. A heterologous pathway from chorismate to QA (Ding et al., 2021) was introduced into P. pastoris (Figure 3A). The key genes of this pathway, including PhzE, PhzD, DhbX and NbaC, were codon optimized and expressed in the Δku70Δ0361Δ0598Δ0359Δ0685 strain. Simultaneously, P. pastoris NadC was overexpressed to convert QA to NaMN. Also, NMN synthase (FtNadE) from F. tularensis was expressed to convert NaMN to NMN (Black et al., 2020; Sorci et al., 2009). As the other precursor, PRPP reacts with NaMN to synthesize NMN. To improve NMN synthesis, it is necessary to increase the supply of PRPP. In a previous report, BaPRS with an L135I mutation (BaPRSL135I) from Bacillus amyloliquefaciens was introduced to eliminate the negative feedback allosteric inhibition (Zakataeva et al., 2012). Therefore, the BaPRSL135I was further introduced in P. pastoris to improve PRPP supply. Western blot analysis conducted on the cell lysate of each enzyme‐expression strain confirmed successful expression of the enzyme in this host (Figure 3B). With these efforts, the NMN titre reached up to 18.7 mg/L, which was increased by 163% and 68.5% in comparison to Δku70 and Δku70Δ0361Δ0598Δ0359Δ0685, respectively (Figure 3C). Also, biomass accumulation was not affected compared with the parent strains (Figure S5).

To improve the efficiency of NMN synthesis, one more copy of NadE was overexpressed. It led to an improved NMN titre of 31.9 mg/L (Figure 3C) at 24 h, which was 70.6% higher in comparison to the previous strain. The results indicated that NadE was a key enzyme and its high‐level expression was important for the NMN synthesis. Moreover, the transport protein PnuC was subsequently co‐expressed. However, the NMN titre was not increased further (Figure 3C), which differed from that in the semi‐synthesis case (Figure 2E). Then the precursor supply pathway of chorismate was enhanced. Previous reports documented that the overexpression of DAHP synthase and feedback‐insensitive mutants of chorismate mutase Aro4K229L increased the yield of derivatives in baker's yeast (Liu et al., 2019b). Moreover, the mutant form of Aro4's isoenzyme Aro3K222L also benefited the synthesis of the downstream products (Liu et al., 2018). These studies provide a good reference for the optimization of the chorismate synthetic pathway. The Aro3K222L was expressed under the control of two constitutive promoters of P GAP and P TEF1 . The titre of NMN reached 26.4 mg/L and 17.9 mg/L, respectively (Figure 3C). In contrast, expression of Aro4K229L was also beneficial for the synthesis of NMN and the titre of NMN reached 27.5 mg/L, which was similar to that of the Aro3K222L expression strain. Finally, the Aro3K222L was expressed in the constructed NMN high production strain, and the obtained construct (Δ0361Δ0598Δ0359Δ0685‐NbaC‐NadC‐PhzE‐PhzD‐DhbX‐2NadE‐PRS‐PnuC‐Aro3K222L, 4ΔNNPPD2NPPA) produced NMN of 36.9 mg/L at 48 h. It represented 4.9 times higher than that of the Δku70 strain without any NMN pathway modification (Figure 3C). Moreover, ethanol serves as a preferable carbon source to glucose. It promoted NMN production to 57.9 mg/L, which was 60% higher than on glucose (Figure S6).

NMN production by semi‐synthesis and de novo biosynthesis in bioreactor

To further evaluate the NMN‐producing performance, the semi‐biosynthesis strain CPN4Δ‐VP and the do novo biosynthesis strain 4ΔNNPPD2NPPA were cultured in a 3‐L bioreactor with a fed‐batch strategy, respectively.

For the NMN semi‐biosynthesis by CPN4Δ‐VP, pH was controlled at ~4.5 during batch phase and adjusted to 7.0 during the fed‐batch phase subsequently. It was cultured with 40 g/L glucose in the batch culture, and the glucose was fed at a rate of 7.5 g/L/h after the batch culture until the wet cell weight reached ~200 g/L. Then the feeding rate of glucose was reduced to 2.5 g/L/h and NAM was added as 2 g/L in the broth. As a result, NMN gradually accumulated until 165 h, which finally produced 1004.6 mg/L NMN by consuming about 1 g/L NAM and 868 g glucose (Figure 4A).

FIGURE 4.

FIGURE 4

Open in a new tab

Production of NMN by P. pastoris in a 3‐L bioreactor. (A) Semi‐biosynthesis of NMN by the CPN4Δ‐VP strain with NAM feeding; (B) De novo biosynthesis of NMN by the 4ΔNNPPD2NPPA strain from basic carbon sources of glucose and ethanol. The broth pH was controlled at 4.5 in the batch phase and adjusted to 7.0 in the fed‐batch phase subsequently. E, ethanol; G, glucose; WCW, wet cell weight.

For the NMN do novo biosynthesis by 4ΔNNPPD2NPPA, pH control was kept the same as that in the semi‐synthesis culture. It was cultured with 40 g/L glucose in the batch culture, the glucose was fed at a rate of 7.5 g/L/h after the batch culture until wet cell weight reached ~200 g/L. Then the feedstock medium was gradually changed to ethanol and kept at a feeding rate of 4.5 g/L/h thereafter. NMN gradually accumulated until 91 h, which finally produced 980.4 mg/L NMN from 160 g glucose and 557 g ethanol (Figure 4B). At the same time, 120 mg/L NAM was synthesized as a major byproduct. As the price of ethanol lower than glucose and NAM (NAM: ~20,000 $/ton; ethanol: 250–350 $/ton; glucose: 300–400 $/ton, Huang et al., 2022; Sun et al., 2024), the do novo biosynthesis of NMN could be a promising technical path in comparison to the semi‐biosynthesis of NMN.

DISCUSSION

NMN, a pivotal compound renowned for its physiological significance in furnishing cells in mammals with NAD+, has garnered widespread recognition, resulting in the swift commercialization of NMN‐based products (Yoshino et al., 2021). This study develops two promising strategies, that is, semi‐biosynthesis and de novo biosynthesis, for the production of NMN via one‐step living cell fermentation by engineered P. pastoris. Systematic metabolic engineering methods were performed on P. pastoris for efficient NMN biosynthesis, including block of branch pathways, expression and evaluation of functional enzymes and transporters, construction of heterologous pathways, expression balancing of the synthetic pathway, etc. With these efforts, a NMN titre of ~1000 mg/L was achieved for either semi‐biosynthesis with NAM addition or de novo biosynthesis from basic carbon sources in bioreactor fermentation.

Enzyme or whole‐cell catalysis is currently a preeminent method for the biosynthesis of NMN, yet the process is inherently intricate and entails considerable costs (Shoji et al., 2021). Most recently, an NAM feeding based fermentation method produced high‐level NMN with a simple process, high productivity and high conversion rate was achieved in E. coli (Huang et al., 2022), which opened up a new avenue for the semi‐biosynthesis of NMN and its derivatives. Nevertheless, the producing host E. coli was not a preferable strain applied in food areas. In the present study, a similar semi‐biosynthesis method of NMN was achieved in a safe (GRAS) strain P. pastoris with a simple process and low‐cost substrates. More importantly, we also realized a new technical path of de novo biosynthesis of NMN in an engineered P. pastoris strain by fermentation from basic carbon sources like glucose and ethanol. The de novo biosynthesis reached a comparable titre to the semi‐biosynthesis by the yeast host but without any NAM feeding and in a much shorter time. Generally, it offers new enlightenment for future improvement in the full biosynthesis of NMN by safe microbial cell factories.

As was reported, Nampts from different sources well supported the construction of E. coli strains for the high‐level semi‐synthesis production of NMN. This inspired us to adapt various Nampts with the yeast host for the efficient conversion of NAM into NMN in this work. The Nampt‐CP from C. pinensis was well expressed and facilitated NMN production, which was also consistent with the results in other hosts in the previous study (Shoji et al., 2021). However, the Nampt‐VP from the V. bacteriophage KVP40 that allowed the reported highest level of NMN production (16.2 g/L, Huang et al., 2022) in E. coli functioned in P. pastoris as well, that a strong promoter P OcA was used for protein expression. It seemed that the Nampt‐VP could not adapt well to this yeast host, despite codon‐optimized and expression improvements were performed. The translation and folding of this enzyme might be incorrect when expressed in P. pastoris. Further enzyme engineering endeavours should be involved to further improve its activity in P. pastoris, in view of its impressive enzymatic activity and catalytic efficiency in E. coli.

Intracellular accumulation of NMN in cells may hinder physiological processes. Shoji et al. (2021) highly increased extracellular NMN accumulation to 6.79 g/L by introducing an NMN transporter PnuC from B. mycoides in engineered E. coli. Although overexpression of this PnuC in P. pastoris promoted NMN production as well, the improvement effects were somehow limited. In‐depth exploration of more efficient transporters will be a useful strategy to further improve NMN biosynthesis. Nonetheless, 1004.6 mg/L NMN was finally obtained at the bioreactor level by semi‐biosynthesis strategy, which represents the highest yield of NMN from yeast. Future cell rewiring and culture optimization should be involved to continuously promote its production capacity as it still falls far behind compared with the achievements in E. coli.

In contrast, the de novo biosynthesis via a hybrid pathway by yeast fermentation from basic carbon sources represents an attractive path for NMN production. By reassembly of heterologous chorismite‐to‐NMN pathway, block of branch pathway and improvement of metabolic precursor accumulation, a comparable titre (980.4 mg/L) to semi‐biosynthetic production was achieved in bioreactor level but in a much shorter time. Moreover, ethanol serves as a preferable carbon source to glucose. The reasons may lie in the fact that the metabolic pathway of ethanol is inherently simpler than that of glucose, and ethanol exhibits a superior energy density in comparison to glucose, thereby providing more energy for the product synthesis (Sun et al., 2024). Deep metabolic rewiring and bioprocess optimization should be conducted to enhance the production further in future. The endogenous de novo biosynthesis pathway of NMN in E. coli from aspartate was improved and 153 mg/L NMN was obtained in a recent study (Wang et al., 2024). However, it still requires the additional supplementation of aspartate as a substrate. Generally, our study reports the first case of the construction of a de novo synthetic pathway for the production of NMN by yeast fermentation from basic carbon sources. It shows the highest yield of NMN by de novo synthesis from basic carbon sources yet reported. It also represents a new production mode of NMN compared with the current strategies, exhibiting good potential for possible industrial use in the future.

AUTHOR CONTRIBUTIONS

Yanna Ren: Writing – original draft; formal analysis; data curation; methodology. Bei Han: Data curation; formal analysis. Shijie Wang: Data curation; formal analysis. Xingbin Wang: Formal analysis. Qi Liu: Funding acquisition. Menghao Cai: Conceptualization; data curation; formal analysis; funding acquisition; project administration; writing – review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Supporting information

Data S1.

MBT2-17-e70048-s001.docx (273.5KB, docx)

ACKNOWLEDGEMENTS

This study was supported by the National Natural Science Foundation of China (32201206), China Postdoctoral Science Foundation (2022M711146) and the National Key R&D Program of China (2023YFA0914102).

Ren, Y. , Han, B. , Wang, S. , Wang, X. , Liu, Q. & Cai, M. (2024) De novo biosynthesis and nicotinamide biotransformation of nicotinamide mononucleotide by engineered yeast cells. Microbial Biotechnology, 17, e70048. Available from: 10.1111/1751-7915.70048

Contributor Information

Yanna Ren, Email: 15026455240@163.com.

Bei Han, Email: 1322564050@qq.com.

Shijie Wang, Email: Y20230054@mail.ecust.edu.cn.

Xingbin Wang, Email: 1716893621@qq.com.

Qi Liu, Email: liuqi0496@ecust.edu.cn.

Menghao Cai, Email: cmh022199@ecust.edu.cn.

DATA AVAILABILITY STATEMENT

Data supporting the conclusions of this article will be made available without undue reservation by the authors.

REFERENCES

  1. Begley, T.P. , Kinsland, C. , Mehl, R.A. , Osterman, A. & Dorrestein, P. (2001) The biosynthesis of nicotinamide adenine dinucleotides in bacteria. Vitamins and Hormones, 61, 103–119. Available from: 10.1016/s0083-6729(01)61003-3 [DOI] [PubMed] [Google Scholar]
  2. Black, W.B. , Zhang, L. , Mak, W.S. , Maxel, S. , Cui, Y. , King, E. et al. (2020) Engineering a nicotinamide mononucleotide redox cofactor system for biocatalysis. Nature Chemical Biology, 16(1), 87–94. Available from: 10.1038/s41589-019-0402-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ding, Y. , Li, X. , Horsman, G.P. , Li, P. , Wang, M. , Li, J. et al. (2021) Construction of an alternative NAD+ de novo biosynthesis pathway. Advanced Science, 8(9), 2004632. Available from: 10.1002/advs.202004632 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Dosselaere, F. & Vanderleyden, J. (2001) A metabolic node in action: chorismate‐utilizing enzymes in microorganisms. Critical Reviews in Microbiology, 27(2), 75–131. Available from: 10.1080/20014091096710 [DOI] [PubMed] [Google Scholar]
  5. Göckel, A. & Richert, C. (2015) Synthesis of an oligonucleotide with a nicotinamide mononucleotide residue and its molecular recognition in DNA helices. Organic & Biomolecular Chemistry, 13(41), 10303–10309. Available from: 10.1039/c5ob01714a [DOI] [PubMed] [Google Scholar]
  6. Gomes, A.P. , Price, N.L. , Ling, A.J. , Moslehi, J.J. , Montgomery, M.K. , Rajman, L. et al. (2013) Declining NAD+ induces a pseudohypoxic state disrupting nuclear‐mitochondrial communication during aging. Cell, 155(7), 1624–1638. Available from: 10.1016/j.cell.2013.11.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Guo, J. , Lam, L.T. , Longenecker, K.L. , Bui, M.H. , Idler, K.B. , Glaser, K.B. et al. (2017) Identification of novel resistance mechanisms to NAMPT inhibition via the de novo NAD+ biosynthesis pathway and NAMPT mutation. Biochemical and Biophysical Research Communications, 491(3), 681–686. Available from: 10.1016/j.bbrc.2017.07.143 [DOI] [PubMed] [Google Scholar]
  8. Huang, Z. , Li, N. , Yu, S. , Zhang, W. , Zhang, T. & Zhou, J. (2022) Systematic engineering of Escherichia coli for efficient production of nicotinamide mononucleotide from nicotinamide. ACS Synthetic Biology, 11(9), 2979–2988. Available from: 10.1021/acssynbio.2c00100 [DOI] [PubMed] [Google Scholar]
  9. Lee, J.Y. , Li, Z. & Miller, E.S. (2017) Vibrio phage KVP40 encodes a functional NAD+ salvage pathway. Journal of Bacteriology, 199(9), e00855. Available from: 10.1128/JB.00855-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Lin, J.B. , Kubota, S. , Ban, N. , Yoshida, M. , Santeford, A. , Sene, A. et al. (2016) NAMPT‐mediated NAD+ biosynthesis is essential for vision in mice. Cell Reports, 17(1), 69–85. Available from: 10.1016/j.celrep.2016.08.073 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Liu, Q. , Shi, X. , Song, L. , Liu, H. , Zhou, X. , Wang, Q. et al. (2019a) CRISPR‐Cas9‐mediated genomic multiloci integration in Pichia pastoris . Microbial Cell Factories, 18(1), 144. Available from: 10.1186/s12934-019-1194-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Liu, Q. , Song, L. , Peng, Q. , Zhu, Q. , Shi, X. , Xu, M. et al. (2022) A programmable high‐expression yeast platform responsive to user‐defined signals. Science Advances, 8(6), eabl5166. Available from: 10.1126/sciadv.abl5166 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Liu, Q. , Yu, T. , Li, X. , Chen, Y. , Campbell, K. , Nielsen, J. et al. (2019b) Rewiring carbon metabolism in yeast for high level production of aromatic chemicals. Nature Communications, 10(1), 4976. Available from: 10.1038/s41467-019-12961-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Liu, Y. , Xu, Y. , Ding, D. , Wen, J. , Zhu, B. & Zhang, D. (2018) Genetic engineering of Escherichia coli to improve L‐phenylalanine production. BMC Biotechnology, 18(1), 5. Available from: 10.1186/s12896-018-0418-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Liu, Y. , Yasawong, M. & Yu, B. (2021) Metabolic engineering of Escherichia coli for biosynthesis of β‐nicotinamide mononucleotide from nicotinamide. Microbial Biotechnology, 14(6), 2581–2591. Available from: 10.1111/1751-7915.13901 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Maharjan, A. , Singhvi, M. & Kim, B.S. (2021) Biosynthesis of a therapeutically important nicotinamide mononucleotide through a phosphoribosyl pyrophosphate synthetase 1 and 2 engineered strain of Escherichia coli . ACS Synthetic Biology, 10(11), 3055–3065. Available from: 10.1021/acssynbio.1c00333 [DOI] [PubMed] [Google Scholar]
  17. Marinescu, G.C. , Popescu, R.G. , Stoian, G. & Dinischiotu, A. (2018) β‐nicotinamide mononucleotide (NMN) production in Escherichia coli . Scientific Reports, 8(1), 12278. Available from: 10.1038/s41598-018-30792-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Meehl, M.A. & Stadheim, T.A. (2014) Biopharmaceutical discovery and production in yeast. Current Opinion in Biotechnology, 30, 120–127. Available from: 10.1016/j.copbio.2014.06.007 [DOI] [PubMed] [Google Scholar]
  19. Mills, K.F. , Yoshida, S. , Stein, L.R. , Grozio, A. , Kubota, S. , Sasaki, Y. et al. (2016) Long‐term administration of nicotinamide mononucleotide mitigates age‐associated physiological decline in mice. Cell Metabolism, 24(6), 795–806. Available from: 10.1016/j.cmet.2016.09.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Peña, D.A. , Gasser, B. , Zanghellini, J. , Steiger, M.G. & Mattanovich, D. (2018) Metabolic engineering of Pichia pastoris . Metabolic Engineering, 50, 2–15. Available from: 10.1016/j.ymben.2018.04.017 [DOI] [PubMed] [Google Scholar]
  21. Ren, Y. , Liu, Q. , Liu, H. , Zhou, X. , Zhang, Y. & Cai, M. (2020) Engineering substrate and energy metabolism for living cell production of cytidine‐5'‐diphosphocholine. Biotechnology and Bioengineering, 117(5), 1426–1435. Available from: 10.1002/bit.27291 [DOI] [PubMed] [Google Scholar]
  22. Safder, I. , Khan, S. , Islam, I.U. & Kazim, M. (2018) Pichia pastoris expression system: a potential candidate to express protein in industrial and biopharmaceutical domains. Biomedical Letters, 4(1), 1–13. Available from: 10.1007/s00253-014-5732-5 [DOI] [Google Scholar]
  23. Shoji, S. , Yamaji, T. , Makino, H. , Ishii, J. & Kondo, A. (2021) Metabolic design for selective production of nicotinamide mononucleotide from glucose and nicotinamide. Metabolic Engineering, 65, 167–177. Available from: 10.1016/j.ymben.2020.11.008 [DOI] [PubMed] [Google Scholar]
  24. Sorci, L. , Martynowski, D. , Rodionov, D.A. , Eyobo, Y. , Zogaj, X. , Klose, K.E. et al. (2009) Nicotinamide mononucleotide synthetase is the key enzyme for an alternative route of NAD biosynthesis in Francisella tularensis . Proceedings of the National Academy of Sciences of the United States of America, 106(9), 3083–3088. Available from: 10.1073/pnas.0811718106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Sun, M. , Gao, A.X. , Liu, X. , Bai, Z. , Wang, P. & Ledesma‐Amaro, R. (2024) Microbial conversion of ethanol to high‐value products: progress and challenges. Biotechnology for Biofuels and Bioproducts, 17(1), 115. Available from: 10.1186/s13068-024-02546-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Wang, P. , Ma, Y. , Li, J. , Su, J. , Chi, J. , Zhu, X. et al. (2024) Exploring the de novo NMN biosynthesis as an alternative pathway to enhance NMN production. ACS Synthetic Biology, 13, 2425–2435. Available from: 10.1021/acssynbio.4c00115 [DOI] [PubMed] [Google Scholar]
  27. Xu, Q. , Bai, C. , Liu, Y. , Song, L. , Tian, L. , Yan, Y. et al. (2019) Modulation of acetate utilization in Komagataella phaffii by metabolic engineering of tolerance and metabolism. Biotechnology for Biofuels, 12(1), 61. Available from: 10.1186/s13068-019-1404-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Yoshino, J. , Baur, J.A. & Imai, S.I. (2018) NAD+ intermediates: the biology and therapeutic potential of NMN and NR. Cell Metabolism, 27(3), 513–528. Available from: 10.1016/j.cmet.2017.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Yoshino, J. , Mills, K.F. , Yoon, M.J. & Imai, S. (2011) Nicotinamide mononucleotide, a key NAD(+) intermediate, treats the pathophysiology of diet‐ and age‐induced diabetes in mice. Cell Metabolism, 14(4), 528–536. Available from: 10.1016/j.cmet.2011.08.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Yoshino, M. , Yoshino, J. , Kayser, B.D. , Patti, G.J. , Franczyk, M.P. , Mills, K.F. et al. (2021) Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women. Science, 372(6547), 1224–1229. Available from: 10.1126/science.abe9985 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Zakataeva, N.P. , Romanenkov, D.V. , Skripnikova, V.S. , Vitushkina, M.V. , Livshits, V.A. , Kivero, A.D. et al. (2012) Wild‐type and feedback‐resistant phosphoribosyl pyrophosphate synthetases from Bacillus amyloliquefaciens: purification, characterization, and application to increase purine nucleoside production. Applied Microbiology and Biotechnology, 93(5), 2023–2033. Available from: 10.1007/s00253-011-3687-3 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1.

MBT2-17-e70048-s001.docx (273.5KB, docx)

Data Availability Statement

Data supporting the conclusions of this article will be made available without undue reservation by the authors.

Articles from Microbial Biotechnology are provided here courtesy of Wiley

RESOURCES