All the strains and plasmids used in this study are shown in Table 1. E. coli DH5α was employed for recombinant plasmid construction, and its competent cells were bought from Tsingke Biotechnology Co. Ltd. (Hangzhou, China). IJAFHEBC (IC) was used as the chassis strain in this work. Plasmid pAm and pACYC184 were employed as the original materials for the plasmid recombination and as the vectors for gene synthesis.

All the PCR experiments were implemented using the Phanta Max PCR kit (Vazyme, Nanjing, China). The PCR products were extracted via a PCR cleanup kit (Axygen, MA, United States) for further molecular experiments (such as recombinant plasmid construction). The construction of recombinant plasmid was carried out via ClonExpress^® II One-step Cloning Kit (Vazyme, Nanjing, China).

Plasmid 184-medh, 184-faldh, 184-fthfl, 184-mthfs, pAm-BsBHMT, and pAm-TnBHMT were synthesized by GenScipt (Nanjing, China). Therein, the medh, faldh, fthfl, and mthfs genes were codon-optimized and synthesized along with a Tac promoter on the pACYC184, respectively. Genes BsBHMT and TnBHMT were codon-optimized and synthesized using pAm as the vector. The constructions of the plasmids, namely, 184-fl-ms, 184-FMMF, and pAm-fl-ms were carried out by the One-step Cloning Kit, and the process was shown in Supplementary Figure S2. For constructing 184-fl-ms, the vector 184-fthfl was linearized via PCR. The mthfs gene was cloned via PCR as well, which used the plasmid 184-mthfs as the template. The recombinant plasmid 184-fl-ms was then employed as the vector to construct 184-FMMF and was linearized by PCR. The medh and faldh genes were cloned by PCR respectively, which used the corresponding synthesized plasmid as the template (184-medh or 184-faldh). For the construction of pAm-fl-ms, the vector pAm was linearized by PCR, and these two target genes were cloned from the plasmid 184-fl-ms by PCR as well. Suitable RBS was chosen for these exogenous genes. The sequences of all synthesized genes, promoters, and RBS used in this work are shown in the Supplementary Material. All the primers for recombinant plasmid construction are listed in Supplementary Table S1.

PCR-based site-directed mutagenesis was implemented to construct 184-FMMF-1, 184-FMMF-2, and 184-FMMF-3. For the construction of 184-FMMF-1, 184-FMMF was as the template, and two-round site-directed mutagenesis PCR was carried out by two pairs of primers (Medh A26V F/R and Medh A169V F/R). Then 184-FMMF-2 was constructed via the PCR-based site-directed mutagenesis using the 184-FMMF-1 as the template and Medh A31V F/R as primers. For constructing 184-FMMF-3, the plasmid 184-FMMF-2 was employed as the template for site-directed mutagenesis PCR, in which Medh A368R F/R was used as the primers. The sequences of these primers are shown in Supplementary Table S1.

LB medium (5 g/L yeast extract, 10 g/L NaCl, and 10 g/L tryptone) was employed for the general culture of E. coli (including seed culture). MF medium (2 g/L yeast extract, 20 g/L glucose, 2 g/L Na₂S₂O₃, 1 g/L KH₂PO₃, 16 g/L (NH₄)₂SO₄, 1 mL/L salty solution, 0.01 g/L L-lysine, 0.2 mg/L Vb_12, and 10 g/L CaCO₃) was used to determine the methionine titer for the flask test. Therein, L-lysine, Vb_12, and CaCO₃ were sterilized individually. The salty solution contained 5 g/L FeSO₄·7H₂O, 500 g/L MgSO₄·7H₂O, 5 g/L ZnSO_4, and 5 g/L MnSO₄·8H₂O. Isopropyl β-D-1-thiogalactopyranoside (IPTG) was used to induce the regulated gene expression for methionine production at the additive amount of 100 µM. When it was necessary, kanamycin (50 mg/L), chloramphenicol (25 mg/L), and ampicillin (100 mg/L) were applied to maintain the stability of plasmids.

For the shake flask fermentation experiment, at first, the single clone of the strain was inoculated into 10 mL LB medium as the seed, and it was incubated at 37°C and 200 rpm for 11 h. Then 1 mL of the seed was added into the large shake flask (0.5 L) that contained 30 mL of MF medium, and it was cultured at 30°C and 150 rpm for 48 h. After fermentation, 1 mL of the broth was collected to determine the biomass (OD₆₀₀) and methionine titer.

The biomass (OD₆₀₀) was measured via the spectrophotometer (Ultrospec 3100, Amersham Biosciences, Uppsala, Sweden). The supernatant of fermentation broth was diluted 20 times and filtered for further detection. The titer of methionine was measured via an LA8080 amino acid analyzer (Hitachi, Japan).

RT-qPCR was implemented to compare the gene expression level between the plasmid pAm and pACYC184. IC/pAm/184-fl-ms and IC/pAm-fl-ms were cultured in 30 mL of MF medium at 150 rpm and 30°C for 18 h. After fermentation, 1 mL of the broth was centrifugated and then frozen using liquid nitrogen. The samples were sent to Tsingke Biotechnology Co. Ltd. (Hangzhou, Zhejiang, China) for RT-qPCR. The 16S rRNA was employed as the internal control. IC/pAm/184-fl-ms was the normalization of gene expression.

The exogenous methanol module was designed to enable the microbial host to use methanol as a carbon source and convert it into methionine. Methanol is an attractive carbon source for microbial fermentation because it is abundant, inexpensive, and can be easily converted into useful products. However, most microorganisms cannot utilize methanol as a carbon source due to the lack of specific metabolic pathways. Therefore, the construction of the exogenous methanol module aimed to introduce the necessary enzymes and genes from other organisms that could enable the microbial host to metabolize methanol and produce methionine.

According to the study conducted by Yu and Liao (Yu and Liao, 2018), the exogenous methanol module was explored as a means of driving methionine biosynthesis (Figure 2). Four enzymes were utilized in this module, namely, methanol dehydrogenase (Medh, from Cupriavidus necator N-1), formaldehyde dehydrogenase (Faldh, from Methylobacterium. extorquens AM1), formate tetrahydrofolate ligase (Fthfl, from M. thermoacetica), and 5,10-methylene-tetrahydrofolate synthase (Mthfs, from M. thermoacetica). These enzymes were encoded by genes that were codon-optimized and synthesized, and the sequences of nucleotides are exhibited in the Supplementary Material. To reduce the burden of pAm, a two-plasmid system was employed for methionine biosynthesis. These four exogenous genes were integrated together on the pACYC184, which could coexist with pAm, resulting in the construction of IC/pAm/184-FMMF. An empty vector of pACYC184 was also introduced into IC/pAm for comparison. The result of the shake flask experiment revealed that the initial strain IC/pAm produced 2.38 g/L methionine, whereas the two-plasmid system of the empty pACYC184 did not seem to influence the methionine titer (2.32 g/L), while the introduction of 184-FMMF reduced it slightly (2.21 g/L).

Subsequently, the effect of the exogenous methanol module on methionine production was tested by adding 0.5 g/L of methanol to the medium. However, both the control (IC/pAm) and IC/pAm/184-FMMF could not grow in the medium containing 0.5 g/L methanol due to the toxicity of methanol, and methionine was not detected in the fermentation broth (Figure 3A). An optimized experiment of methanol concentration was then carried out, and the results indicated that under a methanol concentration of 0.2 g/L, the control strain IC/pAm was still unable to grow, while IC/pAm/184-FMMF could reach 1.32 of OD₆₀₀ and accumulate 0.09 g/L methionine. When the concentration of methanol had reduced to 0.1 g/L, IC/pAm could finally grow to 0.76 of OD₆₀₀, although no methionine was detected in the fermentation broth (Figure 3B). The other strain, IC/pAm/184-FMMF, grew better (2.31 of OD₆₀₀) and accumulated more methionine (0.21 g/L) under the 0.1 g/L of methanol concentration (Figure 3B). However, this fermentation behavior was significantly different from that observed in the usual medium, indicating that the added methanol was not converted in a timely manner, and the cell was seriously influenced by the toxicity of methanol.

To enhance the conversion efficiency of methanol, modifications were made to methanol dehydrogenase (Medh). Three types of modifications were employed based on previous reports (Yu and Liao, 2018; Zhou et al., 2020) to improve Medh activity. These modifications included A26V/A169V (Medh^fbr−1), A26V/A31V/A169V (Medh^fbr−2), and A26V/A31V/A169V/A368R (Medh^fbr−3), and the corresponding mutant plasmids were named 184-FMMF-1, 184-FMMF-2, and 184-FMMF-3, respectively. The effectiveness of these modifications was evaluated by introducing the mutant plasmids into IC/pAm and conducting a fermentation experiment with a methanol concentration of 0.1 g/L. All three mutants showed increased growth and methionine biosynthesis compared to the control (IC/pAm/184-FMMF) (Figure 4A), with the highest level of biomass (3.87 of OD₆₀₀) and methionine titer (1.26 g/L) observed in Medh^fbr−3, which contained four mutant sites. As the number of mutant sites increased, the effectiveness of the modifications became more pronounced.

To further explain the results presented in Figure 4A, it is important to note that methanol is a toxic compound for many microorganisms, including E. coli. When methanol is metabolized in cells, it generates formaldehyde, which can damage cellular components such as proteins, lipids, and nucleic acids. To avoid this damage, cells have evolved several mechanisms to detoxify formaldehyde, such as the formaldehyde oxidation pathway, the formaldehyde assimilation pathway, and the formaldehyde resistance pathway. However, these pathways are not always effective, especially at high methanol concentrations.

In the experiment presented in Figure 4B, as the methanol concentration in the medium increased, the toxicity of methanol became more apparent. The growth of IC/pAm/184-FMMF-3, even with the Medh^fbr−3 modification, was hindered by the presence of methanol, and the accumulation of methionine was significantly reduced. At a methanol concentration of 0.5 g/L, the growth of IC/pAm/184-FMMF-3 was limited to 0.85 of OD₆₀₀ and the methionine titer was only 0.1 g/L, indicating that the presence of methanol was inhibiting the production of methionine. These results suggest that while the Medh^fbr−3 modification can improve the activity of Medh and enhance the production of methionine under certain conditions, the toxic effects of methanol limit its practical application as a substrate for methionine production in E. coli. Further research is needed to identify alternative substrates that can be used for methionine production without compromising the growth and survival of E. coli.

Methyl donors play a crucial role in cellular metabolism, the exogenous addition of methyl donors may have negative effects on cell growth and metabolism, and thus, a low toxicity methyl donor module was constructed to minimize these effects. Formate, a mesostate of the methanol module was found to be nearly non-poisonous and was selected for this purpose (Figure 2). The two corresponding genes, fthtl and mthfs, were introduced into the starting strain (IC/pAm) via a plasmid (184-fl-ms), and for a higher expression level, they were also integrated on pAm (pAm-fl-ms). RT-qPCR experiments confirmed that the transcription levels of these two genes on pAm were significantly higher than those on the low-copy-number plasmid pACYC184 (Supplementary Figure S1). The fermentation behaviors of these constructed strains were analyzed, and overexpression of exogenous genes caused a minor decrease in the methionine titer, regardless of whether it was in a two-plasmid system or a single-plasmid strain (Figure 5A).

To examine the methionine production ability of the exogenous formate module, 1 g/L of ammonium formate was added to the medium. For the control strain (IC/pAm), the addition of ammonium formate had almost no effect (Figure 5A). However, the two constructed strains (IC/pAm/184-fl-ms and IC/pAm-fl-ms) showed a stronger capacity for methionine biosynthesis, with titers of 2.44 g/L and 2.56 g/L, respectively. The strain with a higher expression level of the exogenous genes (IC/pAm-fl-ms) exhibited better fermentation behavior than the two-plasmid system strain (IC/pAm/184-fl-ms) and was used for an optimization experiment of ammonium formate concentration. As shown in Figure 5B, when the concentration of ammonium formate was below 2 g/L, methionine production increased as the amount of additive increased. The highest titer of methionine was observed at 2 g/L of ammonium formate, reaching 2.67 g/L. However, when the concentration of ammonium formate reached 3 g/L, the methionine titer decreased to 2.48 g/L, suggesting that the conversion efficiency of the module from formate to methionine is relatively low, despite the visible improvement in the methionine titer. These findings provide insight into the construction of low-toxicity methyl donor modules and their potential applications in industrial fermentation.

To enhance the supply of methyl donors in a more effective manner, a betaine module was developed based on prior research (Liu et al., 2022), which could directly deliver methyl to homocysteine while producing methionine. This module was powered by betaine-homocysteine methyltransferase (BHMT), which is absent in E. coli. Therefore, two exogenous genes encoding different sources of BHMTs were integrated into pAm, namely, BsBHMT (from Bacillus selenitireducens) and TnBHMT (from Thioclava nitratireducens) (Liu et al., 2022). The nucleotide sequences of these genes are provided in the Supplementary Material. The fermentation results are presented in Figure 6A and indicate that overexpression of exogenous genes had a minor impact on the methionine titer (only a 0.05 g/L decrease). However, when 5 g/L of betaine was supplemented, the strains carrying exogenous BHMTs (IC/pAm-BsBHMT and IC/pAm-TnBHMT) showed significant improvement in the methionine titer (increased to 2.63 g/L and 2.76 g/L, respectively) compared to the control strain (IC/pAm), which had negligible changes (Figure 6A).

An optimization experiment on betaine concentration was conducted using IC/pAm-TnBHMT, which had better methionine production ability. The results are shown in Figure 6B. In general, the addition of betaine to the medium increased methionine biosynthesis in the strain containing the exogenous betaine module. The maximum methionine titer was achieved at 2.87 g/L when 6 g/L of betaine was supplemented (a 20% increase compared to the initial strain). However, when the betaine concentration was further increased to 7 g/L, the methionine titer decreased to 2.7 g/L. Although the betaine module for the exogenous supply of methyl donors significantly increased the methionine titer, its conversion rate of added betaine was only 6.5%, indicating relatively low efficiency.