You are viewing the site in preview mode

Skip to main content
Download PDF

Rewiring Escherichia coli to transform formate into methyl groups

Microbial Cell Factories volume 24, Article number: 55 (2025) Cite this article

Abstract

Background

Biotechnological applications are steadily growing and have become an important tool to reinvent the synthesis of chemicals and pharmaceuticals for lower dependence on fossil resources. In order to sustain this progression, new feedstocks for biotechnological hosts have to be explored. One-carbon (C1-)compounds, including formate, derived from CO2 or organic waste are accessible in large quantities with renewable energy, making them promising candidates. Previous studies showed that introducing the formate assimilation machinery from Methylorubrum extorquens into Escherichia coli allows assimilation of formate through the C1-tetrahydrofolate (C1-H4F) metabolism. Applying this route for formate assimilation, we here investigated utilisation of formate for the synthesis of value-added building blocks in E. coli using S-adenosylmethionine (SAM)-dependent methyltransferases (MT).

Results

We first used a two-vector system to link formate assimilation and SAM-dependent methylation with three different MTs in E. coli BL21. By feeding isotopically labelled formate, methylated products with 51–81% 13C-labelling could be obtained without substantial changes in conversion rates. Focussing on improvement of product formation with one MT, we analysed the engineered C1-auxotrophic E. coli strain C1S. Screening of different formate concentrations allowed doubling of the conversion rate in comparison to the not formate-supplemented BL21 strain with a share of more than 70% formate-derived methyl groups.

Conclusions

Within this study transformation of formate into methyl groups is demonstrated in E. coli. Our findings support that feeding formate can improve the availability of usable C1-compounds and, as a result, increase whole-cell methylation with engineered E. coli. Using this as a starting point, the introduction of additional auxiliary enzymes and ideas to make the system more energy-efficient are discussed for future applications.

Graphical abstract

Background

The majority of industrial processes depends on fossil resources for the synthesis of compounds such as natural products, fuels, polymers, or fine chemicals including pharmaceuticals [1, 2]. With depleting fossil resources and the emerging effects of climate change, defossilisation is an important strategy to create a greener and more sustainable chemical industry [1]. Development of biochemical applications utilising alternative and renewable feedstocks to support or replace existing chemical processes is key to support this transformation [3,4]. One promising approach is the application of CO2-derived feedstocks in chemobiohybrid processes, i.e. initial (electro)chemical reduction of the greenhouse gas to produce reduced one-carbon (C1-)compounds such as formate and their subsequent valorisation through biotechnological efforts [5,6,7]. Additionally, enzymatic strategies for the less energy-intensive reduction of CO2 are investigated which will allow utilisation of CO2-derived formate for one-step whole-cell biocatalysis without prior electrochemical reduction in future applications [8, 9]. Formate has the benefit of stability, biodegradability, ease of handling and low toxicity, which means that it poses a low risk towards humans and the environment and is in accordance with many principles of green chemistry [7, 10,11,12]. For this reason, the development of biotechnological strategies based on the C1-compound formate is an emerging field [7, 10, 13].

In nature, one important strategy to transfer C1-compounds onto various molecules is S-adenosylmethionine (SAM)-dependent methylation catalysed by methyltransferases (MTs, Fig. 1a) [14,15,16]. SAM-dependent methylation is involved in important processes such as DNA or histone methylation for epigenetic regulation of gene expression [17, 18]. Various natural products with biological activity are methylated and methylation of small molecular drugs at specific positions has been shown to increase their potency by up to three orders of magnitude, called the “magic methyl effect” [19]. This makes selective methylation an important step in lead structure optimisation during development of new pharmaceuticals (Fig. 1b). Chemical introduction of methyl groups is in many cases challenging to direct and commonly mediated under energy-consuming conditions by the use of toxic and environmentally harmful methylation agents such as methyl iodide or diazomethane [20]. In comparison to chemical methylation, MTs introduce the methyl group stereo-, regio- and chemoselectively, only synthesising one specific product; by including radical SAM MTs, chemically inaccessible sp3-hybridised carbons could be targeted as well [14, 21]. While chemical methylation is not the major driving force for the use of fossil resources, it is a hazardous technique necessary for countless syntheses, and would benefit from substitution by a strategy which is safer, more environmental-friendly and, hence, more in accordance with the principles of green chemistry; an example for such a substitution strategy is enzymatic, SAM-dependent methylation [12, 15, 16, 22, 23].

Fig. 1
figure 1

SAM-dependent methylation. a Reaction mechanism of S-adenosylmethionine (SAM)-dependent transmethylation of a nucleophilic substrate by methyltransferases (MTs) creating the byproduct S-adenosylhomocysteine (SAH). b Pharmaceuticals and natural compounds carrying “magic methyl” groups, which significantly influence potency. The relevant methyl in tazemetostat was chemically introduced during lead optimisation, C-methylation of gentamicin is mediated by a radical SAM MT, other methyl groups are introduced by conventional MTs

Full size image

Intracellular methylation activity is often a major limiting factor when natural compounds are produced in biotechnological hosts, emphasising the need to develop strategies to improve conversion rates to synthesise methylated products [22,23,24,25]. Optimisation of intracellular SAM availability and acceleration of the cofactor’s regeneration can improve the formation of methylated products using whole-cell methylation [22,23,24,25,26]. One strategy which proved successful is increasing concentrations of the SAM precursor methionine, in particular its regeneration from homocysteine [23, 26,27,28]. In E. coli, homocysteine is methylated to methionine through transfer of a methyl group from methyl-tetrahydrofolate (methyl-H4F) catalysed by one of the methionine synthases EcMETE and EcMETH (Fig. 2) [29]. Majorly, the C1-H4F pool is fuelled from serine and glycine, catalysed by serine hydroxymethyltransferase (EcGLYA) and the glycine cleavage system (EcGCS), respectively [30, 31]. The C1-H4F pool supplies C1-carbon units for the synthesis of nucleobases and amino acids such as methionine or histidine, as well as for formylation of methionine, making it essential for protein production and cell growth [6, 32, 33]. Okano et al. reported that increasing the availability of C1-compounds to improve methyl-H4F production increases the formation of methylated products in E. coli. [28] We surmised that the C1-compound formate could be an ideal supplier of C1-building blocks to fuel the C1-H4F metabolism of biotechnological hosts and to drive whole-cell methylation forward.

Fig. 2
figure 2

Metabolic network to convert formate into transferable methyl groups. a Established formate assimilation machinery consisting of MexFTFL, MexMCH and MexMTDA from M. extorquens (together abbreviated as “FCM”) as it is implemented for the generation of synthetic formatotrophic E. coli. b Metabolic pathway for the native generation of C1-H4F metabolites from serine by EcGLYA and glycine by EcGCS for the supply of methyl groups to the SAM regeneration cycle. Heterologous introduction of an MT and supply of the MT substrate, here RgANMT with 2,5-ANP, enable formation of the methylated product and the byproduct SAH, which can be degraded to homocysteine either by EcMTAN and EcLUXS, or by heterologously introduced SAHH. The connection between formate assimilation and C1-H4F supply for the SAM cycle via methylene-H4F is highlighted in light blue; genes deleted in this study are labelled in the corresponding colour of the strains shown in (c). 2,5-ANP - 2,5-aminonitrophenol, ATP - adenosine-5´-triphosphate, ADP - adenosine-5´-diphosphate, C1 - one carbon, DHPD - 4,5-dihydroxy-2,3-pentanedione, H4F - tetrahydrofolate, NAD(P)H - nicotinamide adenine dinucleotide (phosphate), M-2,5-ANP - N-methyl-2,5-aminonitrophenol, SAM - S-adenosylmethionine, SAH - S-adenosylhomocysteine, EcGCS - glycine cleavage system from E. coli, EcGLYA - serine hydroxymethyltransferase from E. coli, EcLUXS - S-ribosylhomocysteine lyase from E. coli, EcMAT - methionine adenosyltransferase from E. coli, EcMETE/EcMETH - methionine synthase from E. coli, EcMETF - methylene-H4F reductase from E. coli, EcMTAN - S-methyl-5´-thioadenosine/SAH nucleosidase from E. coli, EcSERA - 3-phospho-glycerate dehydrogenase from E. coli, EcSERB - phosphoserine phosphatase from E. coli, EcSERC - phosphoserine aminotransferase from E. coli, MexFTFL - formate-H4F ligase from M. extorquens, MexMCH - formyl-H4F cyclohydrolase from M. extorquens, MexMTDA - methylene-H4F dehydrogenase A from M. extorquens, RgANMT - anthranilate N-MT from Ruta graveolens

Full size image

In previous publications, strategies to assimilate formate into the metabolism of E. coli have been established [13, 34]. Introduction of the formate-H4F ligase from Methylorubrum extorquens (MexFTFL) into E. coli allows synthesis of formyl-H4F from formate and H4F in an adenosine-5´-triphosphate (ATP)-dependent reaction (Fig. 2a) [13, 35, 36]. Further, introduction of formyl-H4F cyclohydrolase (MexFCH) and methylene-H4F dehydrogenase A (MexMTDA), both from M. extorquens, cyclise and reduce formyl-H4F to methylene-H4F, the central metabolite of the C1-H4F metabolism, which can be used to further introduce formate’s carbon into central metabolites [35,36,37]. This formate assimilation strategy composed of MexFTFL, MexFCH and MexMTDA (together referred to as FCM) proved powerful and was used for full assembly of all carbon scaffolds from formate and CO2 in synthetic formatotrophic E. coli (Fig. 2a) [13, 34, 38].

In order to harness this established formate assimilation strategy for SAM-dependent methylation, the generated methylene-H4F has to be reduced once more by the NADH-dependent methylene-H4F reductase (EcMETF, Fig. 2) to methyl-H4F [39]. EcMETE or EcMETH transfer the methyl group onto homocysteine to produce the proteinogenic amino acid methionine, the precursor of SAM [29]. Previously reported introduction of FCM in engineered E. coli strains allowed all methyl groups of methionine to derive from formate when 13C-labelled formate was fed, emphasising the potential to use formate for the synthesis of methyl groups [40]. Growth of the synthetic formatotrophs solely on formate is approximately 10-times slower than on established feedstocks [13, 38], making a strategy with growth from established feedstocks and supply of formate to push methylation of selected building blocks an interesting option.

To activate the methyl group, native methionine adenosyltransferases (EcMAT, Fig. 2b) convert ATP and methionine to SAM as a first step of the SAM regeneration cycle [41]. The cofactor is subsequently utilised by MTs for methylation of their substrate, thereby, introducing the C1-compound into the product. Byproduct of the methylation reaction is S-adenosylhomocysteine (SAH), a known inhibitor of MTs [14]. In native E. coli metabolism, SAH is degraded in an irreversible reaction to adenine and S-ribosylhomocysteine by methyl thioadenosine/SAH nucleosidase (EcMTAN) [42]. S-ribosylhomocysteine is cleaved by S-ribosylhomocysteine lyase (EcLUXS) to homocysteine and the quorum sensing precursor 4,5-dihydroxypentane-2,3-dione (DHPD) [43]. Another option to degrade SAH are SAH hydrolases (SAHH, Fig. 2b) that cleave SAH to adenosine and homocysteine [44]. Transfer of a methyl group from methyl-H4F regenerates methionine from homocysteine allowing repetition of methylation with another C1-building block.

In order to monitor transformation of formate into the methyl groups of MT-derived products, a model MT reaction is needed, such as the anthranilate N-MT from Ruta graveolens (RgANMT, Fig. 2b). RgANMT transfers a methyl group in a SAM-dependent reaction onto the amino group of the unnatural substrate 2,5-aminonitrophenol (2,5-ANP) producing N-methyl-2,5-ANP (M-2,5-ANP, Fig. 2b) [26, 45]. The unnatural substrate and product undergo minimal degradation in E. coli and can freely permeate in and out of the bacterial cell, making this system a well-suited model for investigations on improving whole-cell methylation. Conversion rates of the model MT-substrate system were shown to be strongly dependent on the intracellular availability of methionine, resulting in a great potential to use it to explore the influence of formate on whole-cell methylation [26].

Recent research towards the application of formate mainly focussed on the incorporation of formate-derived carbon into central metabolites, either to allow growth or the production of small organic compounds [13, 46]. Reports on improving whole-cell methylation, on the contrary, focus on improving efficiency of the SAM regeneration cycle and intracellular availability of methionine, mostly omitting the supply of C1-units from the C1-H4F metabolism [23, 25, 26, 47]. In this study we combine these two separated fields of research to investigate formate as a supplementary source of C1-units to increase regeneration of methionine and, therefore, whole-cell methylation with E. coli. We first investigated the sole whole-cell biotransformation of formate into transferred methyl groups using the described formate assimilation machinery FCM in combination with different MTs. We then continued using engineered E. coli strains (Fig. 2c) to allow higher formation of a methylated product by the supplementation of formate.

Methods

Chemicals

All chemicals were purchased in the highest available purity from certified vendors: 13C-formate (> 99% 13C-enriched) was obtained as sodium formate from Sigma Aldrich, unlabelled sodium formate was obtained from Carl Roth chemicals; both were according to our knowledge not synthesised by electrochemical reduction. 2,5-Aminonitrophenol (2,5-ANP), 2,4-methoxy nitroaniline (2,4-MNA), 2,7-dihydroxynaphtalene (DHN), perchloric acid (70%), and formic acid (LC–MS grade) were purchased from Sigma Aldrich; NaCl, NH4Cl, MgSO4, KH2PO4, K2HPO4, CaCl2, agarose, LB-agar, LB-medium, d-( +)-glucose, d-( +)-xylose, glycerol, iso-propyl-β-thiogalactopyranosid (IPTG), kanamycin, chloramphenicol, and glycine were purchased from Carl Roth; 2 × Phusion DNA polymerase master mix, acetonitrile (HPLC grade or LC–MS grade) were purchased from Thermo Fisher Scientific. Ethidium bromide was purchased from Merck KGaA, In-Fusion 5 × master mix was purchased from Takara Bio Inc.; high fidelity restriction enzymes, CutSmart buffer, T4-DNA Ligase, 10 × T4-DNA ligase buffer, and 1 kb Plus DNA ladder were purchased from New England Biolabs; spectinomycin was purchased from VWR International.

Strains and plasmids

See Table 1.

Table 1 Strains and plasmids used in this study (Gene sequences are given at the end of the SI)
Full size table

E. coli strain engineering

All strains and plasmids used in this study are listed in Table 1. The strain C1-opt was constructed from the published C1-Aux (C1A∆3) strain based on E. coli strain SIJ488 [50], a derivative of MG1655, engineered to carry inducible genes for λ-Red recombineering and flippase in its chromosome and C1S was directly constructed from E. coli strain SIJ488 [50]. Genome engineering was performed by λ-red recombineering [52] and P1 phage transduction [50,52]. In brief, the genes indicated in Table 1 for C1S and C1-opt were deleted using PCR-amplified linear DNA fragments carrying kanamycin or chloramphenicol resistance markers flanked by FRT-sites based on the pKD4 or pKD3 plasmids, respectively [52]. Resistance markers were amplified with primers carrying 50 bp overhangs serving as homologous regions flanking the target gene [56]. For removal of resistance markers from the chromosome, flippase was induced by addition of l-rhamnose as described previously [50]. Gene deletions and removal of resistance cassettes were verified by PCR.

For chromosomal insertion of the genes encoding the enzymes for formate assimilation (MexFTFL, MexFCH and MexMTDA, together referred to as FCM) [13,34], we used a previously described genome insertion protocol [53] and vector [13] containing the desired genes (“pDM4:SS9-C1M”, Table 1). In brief, a non-replicative plasmid (pDM4, R6K ori) was introduced into the respective recipient strain via conjugation from an E. coli ST18 donor strain, as described previously [48,49,53]. Kanamycin resistance was used to select for chromosomal insertion of the synthetic FCM operon (including the kanamycin resistance gene) via native homologous recombination based on 600 bp homology regions, with subsequent levansucrase (sacB) counter-selection.

For chromosomal insertion of the arabinose-inducible T7-RNA polymerase to create C1S and C1-opt, we transferred the corresponding genomic locus from BL21-AI (Invitrogen) using P1 phage transduction. The transferred locus consists of an insertion replacing the araB gene, introducing genes encoding a tetracycline efflux pump (a selectable tetracycline resistance marker) and T7-RNA polymerase [57]. The latter is controlled by the endogenous ParaBAD promoter, thus, achieving arabinose-inducible expression of T7-RNA polymerase. Successful transductants were selected via tetracycline resistance and the insertion site was verified by PCR.

Plasmid construction

C1-opt already contains a kanamycin resistance, which is why the anthranilate N-MT from Ruta graveolens (rganmt) was cloned with the restriction enzymes NdeI and HindIII into an ampicillin resistance-containing pET22b( +) vector: Separately, 1 µg of the vectors pET28a::rganmt and pET22b( +) were digested with NdeI and HindIII at 37 °C for 2 h followed by separation by agarose gel electrophoresis (1% agarose, 100 V, 50 min) and clean-up of the excised bands of the cut rganmt insert and the cut pET22b( +) vector using the Promega Wizard™ Plus DNA Purification System. 50 ng of the linear vector were combined in a 1:3 ratio with the respective gene to be inserted and 2 µl of 5 × In-Fusion buffer in a total volume of 10 µl nuclease free water. The mixture was incubated at 50 °C for 20 min followed by standard chemical transformation into E. coli Stellar cells, as recommended for the Takara Bio In-Fusion Cloning. The identity of each vector was confirmed by sequencing.

For co-overexpression on pET28a vector with BL21, the ecmetF gene and the SAH hydrolase gene from Mus musculus (mmsahh) were cloned behind rganmt by digesting pET28a::rganmt with XhoI and cloning of the respective insert into the linearised vector using the seamless In-Fusion Cloning technique (Table 2). When co-overexpression was conducted with pET28a and pCDF-DUET-1 vector in C1S, ecmetF was cloned into the first MCS of pCDF-DUET-1 (Table 2). For this purpose, the respective inserts were amplified from a pET28a vector and the pCDF-DUET-1 vector was linearised via polymerase chain reaction (PCR, Table 2):

Table 2 Utilised oligonucleotides
Full size table

Primers with 15 bp overhangs complementary to the vector were ordered at Eurofins as lyophilised powder. The PCR was set up with 3 µl fwrd primer (c = 10 pM), 3 µl rev primer (c = 10 pM, Table 2), 1 µl of template (pET28a vector with the gene of interest or yeast gDNA), 18 µl nuclease free water, and 25 µl 2 × Phusion master mix (Thermo Fisher). The PCR was conducted with 5 min initial denaturation at 98 °C, followed by 29 cycles of 20 s denaturation at 98 °C, 20 s annealing at 55 °C and 35 s elongation at 72 °C. Afterwards, 5 min terminal elongation at 72 °C followed. The size of the gene was confirmed by agarose gel electrophoresis (1% agarose, 100 V, 50 min), followed by extraction of the band and clean up with the Promega Wizard™ Plus DNA Purification System. For insertion of the gene, the vectors were linearised with the respective primer pairs or with XhoI (Table 2). 50 ng of the linear vector were combined in a 1:3 ratio with the respective gene to be inserted and 2 µl of 5 × In-Fusion buffer in a total volume of 10 µl nuclease free water. The mixture was incubated at 50 °C for 20 min followed by standard chemical transformation into E. coli Stellar cells, as described for the Takara Bio In-Fusion Cloning. Identity of each vector was confirmed by sequencing.

Assay procedure to perform whole-cell methylation

For whole-cell methylation experiments, the E. coli strains were chemically (co-)transformed with the respective plasmids and then grown overnight at 37 °C on agar plates supplemented with the required antibiotics. A single colony was picked and used to inoculate 5 ml LB-medium (+ antibiotic) followed by overnight incubation at 37 °C, 170 rpm. The preculture was used to inoculate 25 ml LB-medium (+ antibiotic) in a 50 ml falcon tube in a ratio of 1:100. The culture was grown to an OD600 of 0.5 – 0.7, overexpression was induced with 1 mM IPTG and took place at 20 °C, 140 rpm for 20 h (the lower temperature and shaking velocity was chosen to lower the stress level during protein production).

After overexpression, the cultures were chilled on ice for 10 min followed by centrifugation for 15 min, 4 °C, 2,500 g. The pellet was washed with 5 ml of ice cold M9-medium (48 mM Na2HPO4, 22 mM KH2PO4, 8.6 mM NaCl, 1.8 mM NH4Cl, 2 mM MgSO4, 0.1 mM CaCl2, and 22 mM glucose as carbon source. NOTE: for experiments conducted for Figure SI 5 & SI 6 20 mM glycerol or 13 mM xylose were used as carbon source instead) followed by centrifugation (15 min, 4 °C, 2,500 g). The pellet was then resuspended to an OD600 of approximately 4.5 in ice cold M9-medium containing 1 mm IPTG and required antibiotics. Assays were then pipetted on ice in 1.5 ml reaction tubes in a total volume of 1 ml followed by incubation at 37 °C and 170 rpm for 24 h. The whole-cell catalysts were used in a final OD600 of 3.0 with 0.75 mm of the substrate to be methylated (2,5-ANP or DHN), 2.5–75 mm 13C-formate and optionally 2 mm glycine. To terminate the assay after 24 h, 100 µl of the assay mixture were taken and mixed with 35 µl of 10% perchloric acid followed by centrifugation at 4 °C and 18,000 g for 30 min. The supernatant was taken and stored at −20 °C until analysis. For determination of conversion rates, HPLC–UV (λ = 280 nm) analysis was used and the share of 13C-labelled methylated product was elucidated with LC–MS/MS.

Statistics

A t-test was performed to compare the mean product formation within 24 h across different conditions. A two-sided independent samples t-test assuming equal variances was applied. Statistical significance was defined as p < 0.05. All experiments were conducted in at least three biological replicates.

HPLC–UV analysis

HPLC–UV analysis (λ = 280 nm) was used to detect MT substrates and products generated during whole-cell methylation experiments and to calculate product formation. For analysis, 10 µl of the sample were injected into an Agilent 1260 infinity II System equipped with a quaternary pump and the samples were separated on an ISAspher 100-3-C18 column (150 × 4.0 mm) with an elution gradient method using 0.1% formic acid in MilliQ water as mobile phase A and acetonitrile as mobile phase B at a flow rate of 1.0 ml *min−1 (s. Table SI 2). The method was initiated at 3% mobile phase B and held for 4 min, followed by an increase to 70% mobile phase B within the next 3 min. Within the following 0.1 min, mobile phase B was increased to 100% and kept at this level for another 2.9 min, followed by a decrease to 3% mobile phase B within 1 min and equilibration for another 3 min. Conversion was calculated by Eq. 1.

Equation 1. Calculation of conversion to methylated products from HPLC-peak area (area under curve, AUC).

$${\text{Conversion}} \,\left[ {\text{\%}} \right]{ = }\, \frac{{\text{AUC\,(product)}}}{{{\text{AUC}} \,\left( {{\text{product}}} \right){\text{ + AUC (substrate)}}}}\,{*}\,{100}$$
(1)

LC–MS/MS analysis

LC–MS/MS was used for analysis of the share of 13C-labelled products. For analysis, 1 µL of the 1:10 diluted sample was injected using a SCIEX 5500 + triplequad LC–MS/MS and the samples were separated on an ISAspher 120-3-C18 Aq column (50 × 2.1 mm) with the same mobile phases and elution gradient method as for HPLC–UV analysis but with a flow rate of 0.2 ml *min-1 (s. Table SI 3). MS conditions and mass shifts were used as described in Table SI 4. The share of 13C-labelled product was calculated by Eq. 2.

Equation 2. Calculation of share of 13C-labelled product from LC–MS/MS peak area (area under curve, AUC).

$${\text{Share }}\left[ {\text{\%}} \right] = { }\frac{{{\text{AUC }}(^{13} {\text{C}} - {\text{prod}})}}{{{\text{AUC }}(^{12} {\text{C}} - {\text{prod}}) + {\text{AUC }}(^{13} {\text{C}} - {\text{prod}})}}{*}100$$
(2)

Results

N-Methylation with formate-derived methyl groups using E. coli BL21-fcm-rganmt

We started by chemical transformation of E. coli BL21 (Gold)DE3 (abbreviated as BL21 in the following) with a plasmid carrying the fcm genes under the control of a constitutive promoter (pFCM, s. Table 1) and a pET28a vector carrying the rganmt gene, resulting in BL21-fcm-rganmt (Table SI 1). To analyse conversion rates, a biotransformation approach was chosen, where the MT is first overproduced together with FCM in rich medium (lysogeny broth, LB) during the overexpression phase, followed by harvest, and resuspension in glucose containing minimal medium (M9) to normalise the cell densities (final OD600 = 3.0) and allow energy regeneration. Subsequently, the biotransformation phase is initiated by adding the methyl acceptor substrate (2,5-ANP) for the methylation reaction together with formate [59,60].

Comparison of BL21-fcm-rganmt to a strain without formate assimilation machinery (BL21-rganmt) showed that introducing pFCM already improves product formation without addition of formate (Fig. 3). A similar increase in product formation was observed when biotransformation experiments were conducted with BL21-rganmt in combination with an empty pCDF-DUET-1 vector carrying a spectinomycin resistance in a previous study [26]. Accordingly, the increase in product formation detected here is suggested to be caused through introduction of the spectinomycin resistance encoded on pFCM, an adenylyltransferase [26]. Hereby enhanced ATP regeneration increases ATP availability for SAM synthesis and, hence, improves product formation [26]. Supplementation with 5 mM 13C-formate did not yield higher conversion for BL21-rganmt or BL21-fcm-rganmt (Fig. 3). Analysis with LC–MS/MS, however, revealed that 67% of the produced M-2,5-ANP were 13C-labelled in the experiment with BL21-fcm-rganmt (Fig. 3, Figure SI 1). This incorporation of formate-derived methyl groups did not take place in BL21-rganmt highlighting the necessity of pFCM for the assimilation of formate.

Fig. 3
figure 3

Whole-cell methylation with BL21-rganmt supplying 13C-formate. a Given is the conversion of 2,5-ANP to M-2,5-ANP (black) and the share of 13C-labelled product ± SD (red) under different conditions and co-overexpression of genes involved in formate to methyl group transformation or SAM regeneration. b Scheme of the metabolic pathway from formate to the methylated product with overproduced enzymes highlighted. Experiments were performed in biological triplicates. n.d. = not detected. Experiments were conducted in M9-medium (22 mm glucose) with a final OD600 of 3.0, 0.75 mM 2,5-ANP and 0–5 mM 13C-formate at 37 °C, 170 rpm for 24 h. NOTE: About 8% of M-2,5-ANP is naturally 13C-labelled, therefore shares given are indicators not absolutes for formate-derived methyl groups. 2,5-ANP - 2,5-aminonitrophenol, H4F - tetrahydrofolate, M-2,5-ANP - N-methyl-2,5-aminonitrophenol, SAM - S-adenosylmethionine, SAH - S-adenosylhomocysteine, EcLUXS - S-ribosylhomocysteine lyase from E. coli, EcMAT - methionine adenosyltransferase from E. coli, EcMETE/EcMETH - methionine synthase from E. coli, EcMETF - methylene-H4F reductase from E. coli, EcMTAN - S-methyl-5´-thioadenosine/SAH nucleosidase from E. coli, MexFTFL - formate-H4F ligase from M. extorquens, MexMCH - formyl-H4F cyclohydrolase from M. extorquens, MexMTDA - methylene-H4F dehydrogenase A from M. extorquens, MmSAHH - SAH hydrolase from M. musculus, RgANMT - anthranilate N-MT from R. graveolens.

Full size image

Despite introduction of formate-derived methyl groups and the, therefore, increased availability of C1-building blocks in BL21-fcm-rganmt fed with 13C-formate, conversion rates did not improve as expected from literature reports [28]. As shown in our previous study, supplementation of BL21-rganmt with 5 mM methionine strongly increased product formation; therefore, neither MT acceptor substrate permeation into the bacterial cell, nor activity of RgANMT seemed to be the major limiting factors. We rather suspected a kinetic bottleneck either in the SAM regeneration cycle or in the reduction of methylene-H4F to methyl-H4F through EcMETF. We co-overexpressed ecmat and the SAH hydrolase gene from Mus musculus (mmsahh), which should improve SAH degradation through cleavage to adenosine and homocysteine; as well as the ecmetF gene together with rganmt encoded on a pET28a vector. Curiously, BL21-fcm-rganmt-ecmat did not produce any methylated product. An increased metabolic burden through overproduction of two enzymes could be one reason for the loss of methylation activity of this strain. Co-overexpression of mmsahh in BL21-fcm-rganmt-mmsahh lowered conversion rates slightly in combination with a decreased share of labelled methylated product. A small increase in conversion with BL21-fcm-rganmt-ecmetF (21% conversion) supplemented with 5 mM 13C-formate compared to BL21-fcm-rganmt (19% conversion) was observed (Fig. 3). Reduction of methylene-H4F to methyl-H4F by EcMETF, hence, seems to be a minor bottleneck for whole-cell biotransformation of formate to methyl groups. Nevertheless, the increase in methylated product was rather small, which is why we omitted co-overexpression in the following experiments. These first results demonstrated straight-forward conversion of formate into transferable methyl groups by whole-cell biotransformation using E. coli.

Chemoselective C- and O-methylation with formate-derived methyl groups

We further analysed, whether other MTs could also be used to methylate with formate-derived SAM. The caffeic acid O-MT from Prunus persica (PpCaOMT) also accepts 2,5-ANP, however, catalyses an O-methylation to form 2,4-methoxy nitroaniline (2,4-MNA, Figure SI 2). While classic chemical methods struggle to selectively N- or O-methylate 2,5-ANP, the here described MTs were demonstrated to be highly selective [45]. As a third model reaction, the C-MT from Streptomyces rishiriensis (CouO) was chosen for the transfer of a methyl group onto the unnatural substrate 2,7-dihydroxynaphtalene (DHN), producing 1-methyl-DHN (M-DHN, Figure SI 2). The selectively methylated, functionalised MT-products are interesting as building blocks for chemical synthesis of small molecules such as active pharmaceutical ingredients. After the biotransformation experiment with supply of 5 mM 13C-formate, shares of 13C-labelled 2,4-MNA and M-DHN were 51% and 81%, respectively; again, without a substantial change in conversion rates (Figure SI 2). In a previous study, co-overproduction of EcMAT and CouO was found to strongly increase conversion rates; this was also the case here: Co-overproduction of EcMAT together with FCM and CouO increased product formation about 3.0-fold with and without the addition of formate. When 5 mM 13C-formate were supplied, a share of 68% of M-DHN was 13C-labelled (Figure SI 2). These results show the flexible applicability of whole-cell methylation with formate-derived methyl groups, using BL21-fcm in combination with different MTs.

Selecting engineered E. coli strains for improving methylation through the supply of formate

Introduction of formate-derived carbon into methyl groups worked successfully; however, conversion rates did not improve by the supply of formate. Initially, we suspected regulation through the transcription repressor EcMETJ to hamper reduction of formate to methyl groups by decreasing the expression of ecmetF, ecmetE, ecmetH, ecmat and genes involved in the synthesis of homocysteine [61,62]. Experiments with BL21ΔmetJ-fcm in combination with the MTs did not result in an increase of conversion rates when formate was supplemented (Figure SI 3). Furthermore, growth of the cultures and overall conversion decreased in comparison to BL21-fcm, suggesting that BL21ΔmetJ-fcm was not the optimal host to continue the investigations of transforming formate into methyl groups. Instead, we engineered strains based on E. coli K-12 MG1655. The strains’ design was performed in accordance with the results of a previous study where Yishai et al. [40] (i) introduced the fcm genes on a plasmid under control of a constitutive promoter, and (ii) deleted genes involved in the native generation of C1-H4F compounds, leading to excellent results in converting formate into the methyl group of methionine [40]. Yishai et al. achieved a C1-auxotrophy in E. coli by deletion of the glycine cleavage system (EcGCS, encoded by ecgcvTHP) and serine hydroxymethyltransferase (ecglyA), which represent the main supply pathways for C1-H4F metabolites, forcing assimilation of formate for the generation of C1-H4F compounds [40]. In this study, ecglyA and ecgcvTHP were deleted to construct the C1-auxotrophic strain C1-opt. Additionally, the serine-C1 auxotrophic strain C1S was constructed, in which ecgcvTHP was deleted, ecglyA is still intact but the de novo biosynthesis of serine is not possible due to deletion of the 3-phospho-glycerate dehydrogenase (ecserA) [40]. For both strains, formate assimilation is enabled by genomic introduction of the fcm genes under control of a constitutive promoter (PSTRONG, Table 1), a genomic introduced arabinose-inducible T7 RNA-polymerase system allows overproduction of MTs from the pET28a vector (Fig. 2, Table 1) [57].

To elucidate basic conversion rates, we focussed on N-methylation with RgANMT and compared BL21-fcm-rganmt, C1-opt-rganmt and C1S-rganmt. Following the overexpression phase in LB-media, all strains methylate 2,5-ANP to M-2,5-ANP without the addition of formate as analysed with HPLC–UV and LC–MS/MS (Fig. 4, Table SI 1). Product formation with C1S-rganmt was 1.5-fold higher than with BL21-rganmt, while with C1-opt-rganmt only about 2% of the substrate were converted to the methylated product (Fig. 4, Table SI 1). Growth of the cultures in LB-medium presumably allowed a certain part of metabolites (e.g. methionine, SAM, or serine for C1S) to be carried over or reside in the cells enabling whole-cell methylation without the addition of formate. Supplying C1-opt-rganmt with 5 mM 13C-formate led to a twofold increase of methylation with a share of 54% 13C-labelled product. As conversion rates were still only about 5%, this strain was not further investigated, as it was unable to compete with BL21-fcm-rganmt in terms of share of formate-derived methyl groups or total product formation. Product formation of C1S-rganmt increased 1.4-fold (42% conversion) when 5 mM 13C-formate were added, clearly outperforming BL21-fcm-rganmt-ecmetF (21% conversion). To our surprise, in the experiment with C1S-rganmt only 15% of M-2,5-ANP were 13C-labelled, indicating only a fraction of transferred methyl groups to be derived from formate. Due to the clear increase in conversion rate, a much higher share was expected. Still, in comparison to BL21-fcm-rganmt, whole-cell methylation with C1S-rganmt benefitted from the addition of formate which is why we continued focussing on this C1-auxotrophic strain.

Fig. 4
figure 4

Whole-cell methylation by BL21-fcm-rganmt, C1-opt-rganmt and C1S-rganmt. HPLC–UV (λ = 280 nm) chromatogram of whole-cell methylation experiments using BL21-fcm-rganmt, C1-opt-rganmt and C1S-rganmt without (unbroken line) and with the addition of 5 mm 13C-formate (dotted line). Experiments were conducted in M9-medium (22 mM glucose) with a final OD600 of 3.0, 0.75 mM 2,5-ANP, and 5 mM 13C-formate at 37 °C, 170 rpm for 24 h. 2,5-ANP  - 2,5-aminonitrophenol, M-2,5-ANP - N-methyl-2,5-aminonitrophenol, fcm formate assimilation machinery genes from M. extorquens, rganmt - anthranilate N-MT gene from R. graveolens

Full size image

Improving methylation with formate-derived methyl groups using C1S-rganmt

To evaluate the influence of formate concentration on product formation during the biotransformation phase, formate concentrations ranging from 2.5 mM to 75 mM were screened (Fig. 5, Figure SI 4). Substantial changes in conversion started from 2.5 mM and peaked at 10 mM 13C-formate supply with a 2.1-fold higher product formation compared to BL21-fcm-rganmt-metF supplemented with 5 mM 13C-formate (Table SI 1). Interestingly, shares of labelled product with 10 mM 13C-formate did not exceed 38%. Further increase of 13C-formate supply lowered product formation with C1S-rganmt whereas the shares of labelled product increased, indicating that the reported toxic effects of formate outweigh its benefit at such high concentrations [63,64]. When 50 mM 13C-formate were supplied, 31% conversion with 76% 13C-labelled methylated product were detected, giving the highest total amount of labelled product.

Fig. 5
figure 5

Increasing whole-cell methylation with formate in C1S-rganmt. Given is the conversion of 2,5-ANP to M-2,5-ANP by C1S-rganmt (black) and the corresponding share of 13C-labelled product (red) after supply with 2.5–75 mm 13C-formate with and without the addition of 2 mm glycine. Experiments were conducted in M9-medium (22 mm glucose) with a final OD600 of 3.0, 0.75 mm 2,5-ANP, 2.5–75 mm 13C-formate, and optionally 2 mm glycine at 37 °C, 170 rpm for 24 h. NOTE: About 8% of M-2,5-ANP is naturally 13C-labelled, therefore respective shares are indicators, not absolutes for formate-derived methyl groups. No significant difference was found between supply of the same formate concentration with and without the addition of glycine. M-2,5-ANP - N-methyl-2,5-aminonitrophenol, rganmt - anthranilate N-MT gene from R. graveolens

Full size image

In the C1S strain, EcGLYA is still present and could use methylene-H4F to synthesise serine from glycine when an excess of glycine is supplemented. On one hand, this reaction could withdraw C1-compounds from the C1-H4F metabolism, leading to decreased availability of C1-carbon units for whole-cell methylation. On the other hand, allowing serine synthesis could increase cysteine concentrations and through the transsulfuration pathway also methionine availability, supporting whole-cell methylation [65]. To test this, we repeated the screening with 2.5–75 mm 13C-formate in combination with supply of 2 mM glycine during the biotransformation phase in M9 media (Fig. 5). Co-feeding of 13C-formate and glycine increased conversion rates slightly (2–5% higher conversion, not statistically significant) for every condition up to 20 mm formate. With further increase of the C1-compound concentration, the small benefits of supplied glycine subsided. In most conditions, the share of 13C-labelled product was slightly lower when glycine was added, indicating that some of the 13C-labelled C1-compound was indeed withdrawn from the C1-H4F metabolism. While the differences observed were not statistically significant, the supply of 10 mm 13C-formate and 2 mM glycine still yielded the highest total conversion rate within this study (46% conversion with a share of 34% 13C-M-2,5-ANP). We assume that residing methionine from LB-media of the overexpression phase is responsible for the non-linear increase of labelled product with the addition of 13C-formate. Increasing the share of formate-derived methyl groups using moderate formate concentrations by exchanging media and carbon sources during overexpression did, however, not improve the total amount of formate-derived methyl groups (Figure SI 5 & Figure SI 6). We therefore focussed on improving product formation with C1S-rganmt and 50 mm 13C-formate to maintain a high incorporation of formate into the methylated product.

To do so, the genes of the SAM regeneration cycle as well as ecmetF were co-overexpressed on a separate pCDF-DUET-1 vector in C1S-rganmt (Fig. 6, Figure SI 7). Increasing SAM synthesis by co-overexpression of ecmat lowered conversion, as was seen during the experiments with BL21-rganmt. Similar results were obtained for the co-overexpression of ecmetF. Co-overexpression of ecmtan and ecluxS led to a slight increase in product formation (~ 2% higher conversion), indicating that removal of SAH and regeneration to homocysteine could be a minor factor limiting whole-cell methylation in C1S-rganmt.

Fig. 6
figure 6

Co-overexpression of SAM regeneration cycle genes in C1S-rganmt. a Co-overexpression of rganmt and adjuvant genes encoded on a pCDF-DUET-1 vector in C1S. Given are the conversion rates (black) and the relative share of 13C-labelled product (red). b Scheme of the metabolic pathway from formate to methyl groups with co-overproduced proteins highlighted. Experiments were conducted in M9-medium, with 22 mm glucose, a final OD600 of 3.0, 0.75 mm 2,5-ANP and 50 mm 13C-formate at 37 °C, 170 rpm for 24 h. NOTE: About 8% of M-2,5-ANP is naturally 13C-labelled, therefore respective shares are indicators, not absolutes for formate-derived methyl groups. *p < 0.05, 2,5-ANP - 2,5-aminonitrophenol, H4F - tetrahydrofolate, M-2,5-ANP - N-methyl-2,5-aminonitrophenol, SAM - S-adenosylmethionine, SAH - S-adenosylhomocysteine, EcLUXS - S-ribosylhomocysteine lyase from E. coli, EcMAT - methionine adenosyltransferase from E. coli, EcMETE/EcMETH - methionine synthase from E. coli, EcMETF - methylene-H4F reductase from E. coli, EcMTAN - S-methyl-5´-thioadenosine/SAH nucleosidase from E. coli, MexFTFL - formate-H4F ligase from M. extorquens, MexMCH - formyl-H4F cyclohydrolase from M. extorquens, MexMTDA - methylene-H4F dehydrogenase A from M. extorquens, RgANMT - anthranilate N-MT from R. graveolens.

Full size image

Discussion

In this study we report that E. coli can be used to transform formate into methyl groups by introducing the formate assimilation machinery of M. extorquens. Introduction of formate-derived carbon into methylated products was highly efficient in BL21-fcm-rganmt supplied with 5 mM formate while conversion was not improved. Product formation with C1S-rganmt, in contrast, benefited from the addition of 5–10 mM formate, while the shares of formate-derived methyl groups peaked at 50 mM formate supply with slight decrease in conversion. Due to the energy-intensive production of renewable formate, low supply of the C1-compound is desirable, with the aim to (i) improve product formation and to (ii) incorporate formate-derived methyl groups into the products. The so far best performing strains C1S-rganmt or C1S-rganmt-ecmtan-ecluxS will be the starting point for future adjustments towards application.

We suspect carbon source-derived unlabelled C1-compounds to be the cause for the non-linear incorporation of formate-derived methyl groups with increased formate supply using C1S. Suppressing the formation of unlabelled C1-compounds should, hence, improve efficiency of whole-cell biotransformation of formate to methyl groups. One source of unlabelled formate could be the native pyruvate formate-lyase (EcPFL). EcPFL is an oxygen-sensitive enzyme and native expression of ecpfl occurs under anaerobic conditions providing a further electron sink for fermentation by cleavage of pyruvate to acetyl-CoA and formate [66]. Another possibility could be the production of intracellular formaldehyde from the added carbon source, which can be used by the unspecific threonine aldolase (EcLTAE) to synthesise unlabelled serine, representing a substrate for the production of unlabelled methylene-H4F by EcGLYA [67]. Deletion of the respective genes is planned for follow-up experiments to improve efficiency of formate incorporation at low formate concentrations using C1S.

The highest total conversion achieved with C1S-rganmt was 46%, which has to be improved in order to use the strain on a larger scale. In future experiments, time courses showing the conversion of the MT substrate to the (13C-)methylated product in combination with analysis of the labelled and unlabelled C1-H4F metabolites, methionine, and SAM will be an important prerequisite to carry out flux analyses for identification of substantial bottlenecks. Based on this, overexpression or introduction of auxiliary pathways to improve product formation can be further explored.

The assimilation of formate, its reduction to methyl groups and the synthesis of SAM is an energy-consuming process, demanding two equivalents of ATP and two equivalents of NAD(P)H. Synthesis of homocysteine from aspartate further requires three equivalents of ATP and two equivalents of NADH. This increased demand for energy can stress E. coli and induce an imbalance in cofactor supply [27]. Increasing NADH regeneration by implementation of, e.g. formate dehydrogenase could improve the intracellular availability of reducing equivalents in E. coli and, as a result, also conversion [13].

Conclusion

Our study demonstrates the introduction of the renewable C1-compound formate into SAM-derived methyl groups. Utilisation of pFCM in BL21-fcm-rganmt-ecmetF allowed methylation with 67% of 13C-labelled product, using the MTs PpCaOMT and CouO more than 50% of introduced methyl groups were 13C-labelled. BL21-fcm-mt offers a simple opportunity to introduce formate’s carbon into value-added small molecules. Supply of formate to the engineered serine-dependent C1-auxotrophic E. coli strain C1S-rganmt successfully increased conversion by 57% with a share of 34% of 13C-labelled product; with this, formate can be used as a source for methyl groups. Starting from this, further bottlenecks of whole-cell methylation (e.g. time courses and flux analysis) can be investigated, and by the addition of auxiliary enzymes (e.g. formate dehydrogenase), and other optimisation steps (e.g. deletion of ecpfl or ecltae), it should be possible to further increase the performance of the strains. In summary, we describe a simple strategy on how to engineer E. coli to use the renewable C1-compound formate to support SAM-dependent methylation for future applications.

Availability of data and materials

All data generated or analysed during this study are included in this published article and its supplementary information files. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

References

  1. Kloo Y, Nilsson LJ, Palm E. Reaching net-zero in the chemical industry - a study of roadmaps for industrial decarbonisation. Renew Sustain Energy Transit. 2024;5:100075.

    Google Scholar 

  2. Wollensack L, Budzinski K, Backmann J. Defossilization of pharmaceutical manufacturing. Curr Opin Green Sustain Chem. 2022;33:100586.

    Article  CAS  Google Scholar 

  3. Bazzanella A, Krämer D. Technologies for sustainability and climate protection – chemical processes and use of CO2. DECHEMA eV. 2017.

  4. Sheldon RA, Woodley JM. Role of biocatalysis in sustainable chemistry. Chem Rev. 2018;118(2):801–38.

    Article  PubMed  CAS  Google Scholar 

  5. Schaub T. CO2-based hydrogen storage: CO2 hydrogenation to formic acid, formaldehyde and methanol. Phys Sci Rev. 2018;3(3):3.

    Google Scholar 

  6. Germer P, Andexer JN, Müller M. Biocatalytic one-carbon transfer – a review. Synthesis. 2022;54(20):4401–25.

    Article  CAS  Google Scholar 

  7. Yishai O, Lindner SN, Gonzalez De La Cruz J, Tenenboim H, Bar-Even A. The formate bio-economy. Curr Opin Chem Biol. 2016;35:1–9.

    Article  PubMed  CAS  Google Scholar 

  8. Ünlü A, Duman-Özdamar ZE, Çaloğlu B, Binay B. Enzymes for efficient CO2 conversion. Protein J. 2021;40(4):489–503.

    Article  PubMed  Google Scholar 

  9. Calzadiaz-Ramirez L, Meyer AS. Formate dehydrogenases for CO2 utilization. Curr Opin Biotechnol. 2022;73:95–100.

    Article  PubMed  CAS  Google Scholar 

  10. Nattermann M, Wenk S, Pfister P, He H, Lee SH, Szymanski W, et al. Engineering a new-to-nature cascade for phosphate-dependent formate to formaldehyde conversion in vitro and in vivo. Nat Commun. 2023;14(1):2682.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  11. Anastas PT. Green chemistry and the role of analytical methodology development. Crit Rev Anal Chem. 1999;29(3):167–75.

    Article  CAS  Google Scholar 

  12. Anastas PT, Warner JC. Green chemistry: theory and practice. Oxford: Oxford University Press; 2000.

    Book  Google Scholar 

  13. Kim S, Lindner SN, Aslan S, Yishai O, Wenk S, Schann K, et al. Growth of E. coli on formate and methanol via the reductive glycine pathway. Nat Chem Biol. 2020;16(5):538–45.

    Article  PubMed  CAS  Google Scholar 

  14. Struck AW, Thompson ML, Wong LS, Micklefield J. S-adenosyl-methionine-dependent methyltransferases: highly versatile enzymes in biocatalysis, biosynthesis and other biotechnological applications. ChemBioChem. 2012;13(18):2642–55.

    Article  PubMed  CAS  Google Scholar 

  15. Mordhorst S, Andexer JN. Round, round we go – strategies for enzymatic cofactor regeneration. Nat Prod Rep. 2020;37(10):1316–33.

    Article  PubMed  CAS  Google Scholar 

  16. Abdelraheem E, Thair B, Varela RF, Jockmann E, Popadić D, Hailes HC, et al. Methyltransferases: functions and applications. ChemBioChem. 2022. https://doiorg.publicaciones.saludcastillayleon.es/10.1002/cbic.202200212.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Jones PA, Takai D. The role of DNA methylation in mammalian epigenetics. Science. 2001;293(5532):1068–70.

    Article  PubMed  CAS  Google Scholar 

  18. Cheung P, Lau P. Epigenetic regulation by histone methylation and histone variants. Mol Endocrinol. 2005;19(3):563–73.

    Article  PubMed  CAS  Google Scholar 

  19. Schönherr H, Cernak T. Profound methyl effects in drug discovery and a call for new CH methylation reactions. Angew Chem Int Ed. 2013;52(47):12256–67.

    Article  Google Scholar 

  20. Kirman CR, Sweeney LM, Gargas ML, Kinzell JH. Evaluation of possible modes of action for acute effects of methyl iodide in laboratory animals. Inhal Toxicol. 2009;21(6):537–51.

    Article  PubMed  CAS  Google Scholar 

  21. Broderick JB, Broderick WE, Hoffman BM. Radical SAM enzymes: Nature’s choice for radical reactions. FEBS Lett. 2023;597(1):92–101.

    Article  PubMed  CAS  Google Scholar 

  22. Galman JL, Parmeggiani F, Seibt L, Birmingham WR, Turner NJ. One-pot biocatalytic in vivo methylation-hydroamination of bioderived lignin monomers to generate a key precursor to L-DOPA. Angew Chem Int Ed. 2022;61(8):e202112855.

    Article  CAS  Google Scholar 

  23. Liu Q, Lin B, Tao Y. Improved methylation in E. coli via an efficient methyl supply system driven by betaine. Metab Eng. 2022;72:46–55.

    Article  PubMed  CAS  Google Scholar 

  24. David B, Wolfender JL, Dias DA. The pharmaceutical industry and natural products: historical status and new trends. Phytochem Rev. 2015;14(2):299–315.

    Article  CAS  Google Scholar 

  25. Luo ZW, Cho JS, Lee SY. Microbial production of methyl anthranilate, a grape flavor compound. Proc Natl Acad Sci. 2019;116(22):10749–56.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  26. Mohr M, Benčić P, Andexer JN. Doping in vivo alkylation in E. coli by Introducing the direct sulfurylation pathway of S. cerevisiae. Angew Chem Int Ed. 2024. https://doiorg.publicaciones.saludcastillayleon.es/10.1002/anie.202414598.

    Article  Google Scholar 

  27. Lv Y, Chang J, Zhang W, Dong H, Chen S, Wang X, et al. Improving microbial cell factory performance by engineering SAM availability. J Agric Food Chem. 2024;72(8):3846–71.

    Article  PubMed  CAS  Google Scholar 

  28. Okano K, Sato Y, Inoue S, Kawakami S, Kitani S, Honda K. Enhancement of S-adenosylmethionine-dependent methylation by integrating methanol metabolism with 5-methyl-tetrahydrofolate formation in. Catalysts. 2020. https://doiorg.publicaciones.saludcastillayleon.es/10.3390/catal10091001.

    Article  Google Scholar 

  29. Gonzalez JC, Banerjee RV, Huang S, Sumner JS, Matthews RG. Comparison of cobalamin-independent and cobalamin-dependent methionine synthases from Escherichia coli: two solutions to the same chemical problem. Biochem. 1992;31(26):6045–56.

    Article  CAS  Google Scholar 

  30. Simic P, Willuhn J, Sahm H, Eggeling L. Identification of glyA (Encoding Serine Hydroxymethyltransferase) and its use together with the exporter ThrE to increase l-threonine accumulation by corynebacterium glutamicum. Appl Environ Microbiol. 2002;68(7):3321–7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  31. Kikuchi G. The glycine cleavage system: Composition, reaction mechanism, and physiological significance. Mol Cell Biochem. 1973;1(2):169–87.

    Article  PubMed  CAS  Google Scholar 

  32. Locasale JW. Serine, glycine and the one-carbon cycle: cancer metabolism in full circle. Nat Rev Cancer. 2013;13(8):572–83.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  33. Sahr T, Ravanel S, Rébeillé F. Tetrahydrofolate biosynthesis and distribution in higher plants. Biochem Soc Trans. 2005;33(4):758–62.

    Article  PubMed  CAS  Google Scholar 

  34. Wenk S, Rainaldi V, Schann K, He H, Bouzon M, Döring V, et al. Evolution-assisted engineering of E. coli enables growth on formic acid at ambient CO2 via the serine threonine cycle. Metab Eng. 2025;88:14–24.

    Article  PubMed  CAS  Google Scholar 

  35. Kim S, Lee SH, Seo H, Kim KJ. Biochemical properties and crystal structure of formate-tetrahydrofolate ligase from Methylobacterium extorquens CM4. Biochem Biophys Res Commun. 2020;528(3):426–31.

    Article  PubMed  CAS  Google Scholar 

  36. Ochsner AM, Sonntag F, Buchhaupt M, Schrader J, Vorholt JA. Methylobacterium extorquens: methylotrophy and biotechnological applications. Appl Microbiol Biotechnol. 2015;99(2):517–34.

    Article  PubMed  CAS  Google Scholar 

  37. Vorholt JA, Chistoserdova L, Lidstrom ME, Thauer RK. The NADP-dependent methylene tetrahydromethanopterin dehydrogenase in Methylobacterium extorquens AM1. J Bacteriol. 1998;180(20):5351–6.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  38. Yilmaz S, Kanis B, Hogers RAH, Benito-Vaquerizo S, Kahnt J, Glatter T, et al. System-level characterization of engineered and evolved formatotrophic E. coli strains. bioRxiv. 2024. https://doiorg.publicaciones.saludcastillayleon.es/10.1101/2024.11.11.622960.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Sheppard CA, Trimmer EE, Matthews RG. Purification and properties of NADH-dependent 5, 10-methylenetetrahydrofolate reductase (MetF) from Escherichia coli. J Bacteriol. 1999;181(3):718–25.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  40. Yishai O, Goldbach L, Tenenboim H, Lindner SN, Bar-Even A. Engineered assimilation of exogenous and endogenous formate in Escherichia coli. ACS Synth Biol. 2017;6(9):1722–31.

    Article  PubMed  CAS  Google Scholar 

  41. Berger B, Knodel M. Characterisation of methionine adenosyltransferase from Mycobacterium smegmatis and M. tuberculosis. BMC Microbiol. 2003;3(1):12.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Parveen N, Cornell KA. Methylthioadenosine/S-adenosylhomocysteine nucleosidase, a critical enzyme for bacterial metabolism: involvement of MTA/SAH nucleosidase in bacterial metabolism. Mol Microbiol. 2011;79(1):7–20.

    Article  PubMed  CAS  Google Scholar 

  43. Winzer K, Hardie KR, Burgess N, Doherty N, Kirke D, Holden MTG, et al. LuxS: its role in central metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone. Microbiology. 2002;148(4):909–22.

    Article  PubMed  CAS  Google Scholar 

  44. Shimizu S, Shiozaki S, Ohshiro T, Yamada H. Occurrence of S-adenosylhomocysteine hydrolase in prokaryote cells. Eur J Biochem. 1984;141(2):385–92.

    Article  PubMed  CAS  Google Scholar 

  45. Jockmann E, Subrizi F, Mohr MKF, Carter EM, Hebecker PM, Popadić D, et al. Expanding the substrate scope of N- and O-methyltransferases from plants for chemoselective alkylation**. ChemCatChem. 2023;15(22):e202300930.

    Article  CAS  Google Scholar 

  46. Reiter MA, Bradley T, Büchel LA, Keller P, Hegedis E, Gassler T, et al. A synthetic methylotrophic Escherichia coli as a chassis for bioproduction from methanol. Nat Catal. 2024. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41929-024-01137-0.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Kunjapur AM, Hyun JC, Prather KLJ. Deregulation of S-adenosylmethionine biosynthesis and regeneration improves methylation in the E. coli de novo vanillin biosynthesis pathway. Microb Cell Fact. 2016;15(1):61.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Jackson SA, Fellows BJ, Fineran PC. Complete Genome Sequences of the Escherichia coli Donor Strains ST18 and MFDpir. Microbiol Resour Announc. 2020. https://doiorg.publicaciones.saludcastillayleon.es/10.1128/mra.01014-20.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Thoma S, Schobert M. An improved Escherichia coli donor strain for diparental mating. FEMS Microbiol Lett. 2009;294(2):127–32.

    Article  PubMed  CAS  Google Scholar 

  50. Jensen SI, Lennen RM, Herrgård MJ, Nielsen AT. Seven gene deletions in seven days: fast generation of Escherichia coli strains tolerant to acetate and osmotic stress. Sci Rep. 2015;5(1):17874.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  51. He H, Noor E, Ramos-Parra PA, García-Valencia LE, Patterson JA, Díaz de la Garza RI, et al. In vivo rate of formaldehyde condensation with tetrahydrofolate. Metabolites. 2020;10(2):65.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci. 2000;97(12):6640–5.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  53. Wenk S, Yishai O, Lindner SN, Bar-Even A. An engineering approach for rewiring microbial metabolism. In: Scrutton N, editor. Methods in enzymology. Cambridge: Academic Press; 2018. p. 329–67.

    Google Scholar 

  54. Milton DL, O’Toole R, Horstedt P, Wolf-Watz H. Flagellin A is essential for the virulence of Vibrio anguillarum. J Bacteriol. 1996;178(5):1310–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  55. Bassalo MC, Garst AD, Halweg-Edwards AL, Grau WC, Domaille DW, Mutalik VK, et al. Rapid and efficient one-step metabolic pathway integration in E. coli. ACS Synth Biol. 2016;5(7):561–8.

    Article  PubMed  CAS  Google Scholar 

  56. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, et al. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol. 2006. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/msb4100050.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Bhawsinghka N, Glenn KF, Schaaper RM. Complete genome sequence of Escherichia coli BL21-AI. Microbiol Resour Announc. 2020. https://doiorg.publicaciones.saludcastillayleon.es/10.1128/mra.00009-20.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Braatsch S, Helmark S, Kranz H, Koebmann B, Ruhdal JP. Escherichia coli strains with promoter libraries constructed by red/ET recombination pave the way for transcriptional fine-tuning. Biotechniques. 2008;45(3):335–7.

    Article  PubMed  CAS  Google Scholar 

  59. Lin B, Tao Y. Whole-cell biocatalysts by design. Microb Cell Fact. 2017;16(1):106.

    Article  PubMed  PubMed Central  Google Scholar 

  60. de Carvalho CCCR. Whole cell biocatalysts: essential workers from Nature to the industry. Microb Biotechnol. 2017;10(2):250–63.

    Article  PubMed  Google Scholar 

  61. Marincs F, Manfield IW, Stead JA, Mcdowall KJ, Stockley PG. Transcript analysis reveals an extended regulon and the importance of protein–protein co-operativity for the Escherichia coli methionine repressor. Biochem J. 2006;396(2):227–34.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  62. Shoeman R, Coleman T, Redfield B, Greene RC, Smith AA, Saint-Girons I, et al. Regulation of methionine synthesis in Escherichia coli: Effect of metJ gene product and S-adenosylmethionine on the in vitro expression of the metB, metL and metJ genes. Biochem Biophys Res Commun. 1985;133(2):731–9.

    Article  PubMed  CAS  Google Scholar 

  63. Nicholls P. Formate as an inhibitor of cytochrome c oxidase. Biochem Biophys Res Commun. 1975;67(2):610–6.

    Article  PubMed  CAS  Google Scholar 

  64. Warnecke T, Gill RT. Organic acid toxicity, tolerance, and production in Escherichia coli biorefining applications. Microb Cell Fact. 2005;4(1):25.

    Article  PubMed  PubMed Central  Google Scholar 

  65. Aitken SM, Lodha PH, Morneau DJK. The enzymes of the transsulfuration pathways: active-site characterizations. Biochim Biophys Acta BBA - Proteins Proteomics. 2011;1814(11):1511–7.

    Article  PubMed  CAS  Google Scholar 

  66. Sawers G, Böck A. Novel transcriptional control of the pyruvate formate-lyase gene: upstream regulatory sequences and multiple promoters regulate anaerobic expression. J Bacteriol. 1989;171(5):2485–98.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  67. He H, Höper R, Dodenhöft M, Marlière P, Bar-Even A. An optimized methanol assimilation pathway relying on promiscuous formaldehyde-condensing aldolases in E. coli. Metab Eng. 2020;60:1–13.

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

We thank Aaliya Afandi Lee and Adelheid Nagel for their skilful technical support. We thank Prof. Dr Nicholas Turner for providing the E. coli BL21ΔmetJ strain. We thank Dr Hai He for providing progenitor strains used in the construction of C1-opt and C1S.

Funding

Open Access funding enabled and organized by Projekt DEAL. This work was funded by a PhD scholarship of the Deutsche Bundesstiftung Umwelt (DBU, Grant Number 20021/742) to MKFM and the European Research Council (ERC, Grant Number 716966). JNA acknowledges funding through a Heisenberg Grant of the Deutsche Forschungsgemeinschaft (DFG/527572100).

Author information

Authors and Affiliations

  1. Institute of Pharmaceutical Sciences, University of Freiburg, Albertstr. 25, 79104, Freiburg, Germany

    Michael K. F. Mohr & Jennifer N. Andexer

  2. Max Planck Institute for Terrestrial Microbiology, Karl-Von-Frisch-Straße 10, 35043, Marburg, Germany

    Ari Satanowski & Tobias J. Erb

  3. Max Planck Institute of Molecular Plant Physiology, Am Mühlenberg 1, 14476, Potsdam-Golm, Germany

    Ari Satanowski

  4. Department of Biochemistry, Charité Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin and Humboldt-Universität, Charitéplatz 1, 10117, Berlin, Germany

    Steffen N. Lindner

  5. LOEWE Center for Synthetic Microbiology (SYNMIKRO), Philipps University of Marburg, Marburg, Germany

    Tobias J. Erb

Authors
  1. Michael K. F. Mohr
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Ari Satanowski
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Steffen N. Lindner
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Tobias J. Erb
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Jennifer N. Andexer
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

MKFM and JNA designed the research idea, AS, SLM and TJE further refined the research idea, MKFM performed and analysed in vivo methylation experiments, AS prepared the T7 RNA polymerase arabinose inducible C1-opt and C1S strains. MKFM wrote and together with JNA revised the manuscript, all authors contributed together to the finalisation of the manuscript.

Corresponding author

Correspondence to Jennifer N. Andexer.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Mohr, M.K.F., Satanowski, A., Lindner, S.N. et al. Rewiring Escherichia coli to transform formate into methyl groups. Microb Cell Fact 24, 55 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12934-025-02674-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12934-025-02674-4

Keywords

Microbial Cell Factories

ISSN: 1475-2859

Contact us