Evolutionary engineering of methylotrophic E. coli enables fast growth on methanol

Abstract

As methanol can be derived from either CO₂ or methane, methanol economy can play an important role in combating climate change. In this scenario, rapid utilization of methanol by an industrial microorganism is the first and crucial step for efficient utilization of the C1 feedstock chemical. Here, we report the development of a methylotrophic E. coli strain with a doubling time of 3.5 hours under optimal conditions, comparable or faster than native model methylotrophs Methylorubrum extorquens AM1 (T_d~4hr) and Bacillus methanolicus at 37°C (T_d~5hr). To accomplish this, we develop a bacterial artificial chromosome (BAC) with dynamic copy number variation (CNV) to facilitate overcoming the formaldehyde-induced DNA-protein cross-linking (DPC) problem in the evolution process. We track the genome variations of 75 cultures along the evolution process by next-generation sequencing, and identified the features of the fast-growing strain. After stabilization, the final strain (SM8) grows to 20 g/L of cell mass within 77 hrs in a bioreactor. This study illustrates the potential of dynamic CNV as an evolution tool and synthetic methylotrophs as a platform for sustainable biotechnological applications.

Fig. 1. Copy number and tandem repeat patterns of ddp-BAC.

a A schematic showing BAC with and without the ddp operon. b The copy number of gfp measured by droplet digital PCR (ddPCR). Each strain was tested with at least 3 individual colonies from a LB plate. Only ddp-BAC with the ddp operon in the recA⁺ host existed in high copies (BW25113 strain with ddp-BAC::gfp, n = 6; BW25113 strain with BAC::gfp and  ΔrecA strain with BAC::gfp or ddp-BAC::gfp, n = 3; all the data are biological repeats; data are presented as mean values ± SD). c Nanopore sequencing of ddp-BAC. The patterns shown are raw intact nanopore reads mapped to BAC. The orange label represents the ddp operon, while the blue arrows represent the BAC backbone with gfp. Only ddp-BAC in the recA⁺ host showed tandem repeats with a distribution of copy numbers. Source data are provided as a Source Data file.

Fig. 2. Copy number distribution of ddp-BAC.

a Time-lapse live cell imaging of GFP fluorescence. Colonies from ddp-BAC::gfp showed varied fluorescence levels both among different colonies and within the same colony. The arrows highlight the progression of the same colony formation from the initial time point (0 min) to 200 min in each column. The red and blue arrows indicate low and high fluorescence, respectively. In contrast, colonies from strains containing BAC::gfp (magenta arrows) or the high-copy pUC19::gfp (orange arrows) demonstrated uniform fluorescence intensities. GFP represents the view of green fluorescent protein fluorescence, PH represents the view of phase contrast optical microscopy, and Merge is the composite image combining both the GFP and the PH views. b Flow cytometry analysis of BW25113 containing various versions of BAC or pUC19. Results show that only ddp-BAC::gfp exhibited a bimodal distribution. The x-axis represents the fluorescence intensity in arbitrary units (a.u.) on a logarithmic scale, and the y-axis represents the count of cells. The green dashed line indicates the lowest threshold determined by WT BW25113 strain in autoclaved water for positive signals of GFP-expressing cells. Source data are provided as a Source Data file and also available at Figshare.

Fig. 3. Debottlenecking evolution using ddp-BAC::RHTTP.

a Formaldehyde concentration in SM1 cultures at an OD₆₀₀ of around 1.0, grown in cell culture tubes with low (400 mM, M400) and high (1000 mM, M1000) methanol concentrations. (n = 5 for each group, biological repeats; data are presented as mean values ± SD). b Doubling time of the SM1 strain evolved up to passages 77 showed minor improvement. c SM1p with various plasmids harboring rpe, hps, tkt, tal, phi (RHTTP) grown and passed in 1000 mM methanol. Each strain was tested with at least 3 colonies from a LB plate. d The growth curves and the doubling time of SM7 grown in 400 mM methanol. The different colors represent different biological repeats (n = 3). Source data are provided as a Source Data file.

Fig. 4. The evolution process towards SM8.

a The copy number of ddp-BAC and growth rate changes along the evolution line towards SM8. The copy number of ddp-BAC was determined from Illumina sequencing. The purple line corresponds to the copy number of ddp-BAC. The red line indicates the growth rate in 400 mM methanol, while the gray line represents the growth rate in high methanol (1000 mM or 1200 mM). The red dashed line represents the growth rate where mutS is not restored. (Each passage is defined as a subculturing step, growing from OD₆₀₀ 0.05–0.1 to OD₆₀₀ 1). kb refers to kilobases, where it depicts the unit size of IS5-flanked tandem repeats derived from the original 70 kb tandem repeats. b The overall accumulation of different kinds of mutations throughout the evolution line, including SNP, INDEL, and Silent mutations. The numbers depict the passage number of the strain. All dots represent samples that were sequenced by NGS. Source data are provided as a Source Data file.

Fig. 5. Chromosomal or plasmid changes along the evolution trajectory.

The upper shows the evolution trajectory with branching evolution lines. The numbers depict the passage number of the strain. All dots represent samples that were sequenced by NGS. The numbers in boxes depict the unit size of IS5-flanked tandem repeats derived from the original 70 kb tandem repeats. kb refers to kilobases. The lower part marks all changes in the genome along the evolution trajectory. Only changes greater than 25% variant frequency are shown. The genomic changes shown in SM1 are relative to its parental strain BW25113. The “+” & “-” signs represent mutations appeared or disappeared along the evolution trajectory, relative to the previous strain. The IS2-inserted mutS inactivation that occurred at passage 258 coincided with the increase of mutation events. The font colors in the graph are defined as follows: red for important events, green for SNPs, purple for INDELs, and gray for changes that were not retained along the evolutionary trajectory. Source data are provided as a Source Data file.

Fig. 6. Characterization of DPC in SM7.

a Time courses of SM1 and SM7 grown in 400 mM MeOH. Samples were collected at time points I to VI for SM1, and a to f for SM7 respectively, for extracellular formaldehyde measurement using LC-MS/MS. b Transmission electron microscopy (TEM) images of DPC products extracted from both SM1 and SM7 at different growth stages by negative staining. The experiment was repeated twice independently with similar results. c Enzymatic assays of Mdh and enzymes in the RuMP cycle in SM1 and SM7. Mdh activity was measured by the colorimetric Nash assay. The remaining enzymes were measured using coupled enzyme assays with a NADPH readout at 340 nm (n = 3, biological repeats; data are presented as mean values ± SD). Source data are provided as a Source Data file.

Fig. 7. Proteomics analysis of SM7 and SM1.

a Abundance ratio of the 44 E. coli ribosomal subunit proteins in SM7, SM1 (grown in 400 mM methanol), and wild-type parental strain BW25113 (grown in 1% glucose). The boxplot is generated with their whiskers defining the maximum and minimum value, while the box encloses the first (25%) and third (75%) quartiles, with an additional line at the second quartile marking the median. Error bars represent SD. b Functional enrichment analysis identified biological processes by comparing SM7 to SM1 grown with 400 mM methanol until OD₆₀₀ reached 1. The gene ratio on the x-axis represents the proportion of genes from a given gene set that are found to be differentially expressed or enriched in the analyzed dataset. A higher gene ratio indicates that a larger proportion of the gene set is involved in the indicated biological pathway. The adjusted P-value with Benjamini-Hochberg Procedure are shown with the color code (p. adjust). The data is generated by using GSEA with a permutation-based two-tailed test with 1000 permutations, with a FDR calculated P-value of 0.05 cutoff. c Metabolic protein abundance ratios. The left and right boxes represent SM7/WT and SM1/WT log₂ ratios, respectively. The circles represent the log₂ ratios of SM7/SM1 non-native proteins (Mdh, Hps, and Phi). Hps was cloned from either Methylomicrobium buryatense 5 GB1S (5G) or Bacillus methanolicus (BM). Abbreviations are listed in Supplementary Table 4. All data were done with n = 3, biological repeats. Source data are provided as a Source Data file.

Fig. 8. Characterization of synthetic methylotrophic E. coli.

a Complementation of mutated genes with BAC harboring a wild-type copy of the indicated genes in SM7n showed that gltA, ihfA, and rhlB mutations were important to the high growth rate of SM7. Growth rates of various strains with and without BAC complementation were measured in 400 mM methanol (n = 4 for SM7n strains with ihfA, ompR, cra, and gadW expression, n = 3 for SM7n strains with gltA, rhlB, gtrS, quuD, ylcG, gatY and rusA expression; All are biological repeats; data are presented as mean values ± SD; significance was tested by 2-side Welsh-test; *P < 0.05, **P < 0.01). b High-cell-density cultivation of SM8 (SM7 with WT mutS restored) in 500 mM methanol performed in a 2.0 L bioreactor or a 50 mL Falcon tube. The strain was cultivated in a fed-batch mode with continuous supply of methanol and NH₄OH. The M9 and MOPS media have been modified with an adjustment of an initial NH₄Cl concentration of 25 mM. SM8 reached 20 g DCW L⁻¹ within 77 h in the bioreactor with M9-methanol medium. The orange line represents SM8 grown in the bioreactor with M9-methanol medium, the blue line represents growth in the bioreactor with MOPS-methanol medium, and the red line represents growth in the Falcon tube with MOPS-methanol medium. c Growth of SM8 in a logarithmic scale. The doubling time of SM8 grown in a 50 mL falcon tube with modified-MOPS minimal medium containing 400 mM methanol reached 3.6 h. d The profile of fermentation products of SM8 in the bioreactor with the modified M9 methanol medium. (n = 3, sampling repeats; data are presented as mesan values ± SD) (e) Growth of SM7 in Falcon tubes containing MOPS minimal medium with 400 mM methanol (red line), 1% glucose (dark yellow line), 0.1% glucose (blue line), 400 mM methanol plus 0.1% glucose (purple line). All data were done with n = 3, biological repeats. Source data are provided as a Source Data file.