Assimilation of Formic Acid and CO2 by Engineered E. coli

Review of Assimilation of formic acid and CO2 by engineered Escherichia coli equipped with reconstructed one-carbon assimilation pathways Ng Wenfa 27 July 2020 1

Paper in PNAS 2

Introduction Assimilation of formate and carbon dioxide by engineered Escherichia coli with reconstructed one carbon assimilation pathway. The approach utilises reconstructed tetrahydrofolate cycle and reverse glycine cleavage pathway. Demonstrations of the utility of the pathway came from channelling of formic acid flux to glycine, serine and pyruvate. However, the system suffers from requirement of relatively high concentration of glucose as auxiliary substrate. Introduction of formate dehydrogenase improved formic acid utilisation, and reduced glucose requirement. 3

Overview of genetic modifications in the project Overall logic of project Formate is taken up by E. coli, and condenses with THF to enter the THF cycle. Progression of the cycle results in production of serine, which could be converted to pyruvate that enters central carbon metabolism for biomass formation. A puzzle in the work revolves the need for glucose supplementation. Secondly, a more ready approach exists in using cell growth to assess the efficiency of pathway compared to the convoluted approach of using carbon-13 proteoinogenic analysis. 4

Genes and pathways that have been knocked out Genes that have been knocked out serA: phosphoglycerate dehydrogenase pflB: pyruvate formate-lyase gcvR: putative transcriptional regulator GcvR serA: phosphoglycerate dehydrogenase Catalyzes the reversible oxidation of 3-phospho-D-glycerate to 3-phosphonooxypyruvate, the first step of the phosphorylated L-serine biosynthesis pathway. pflB: pyruvate formate-lyase Catalyzes conversion of pyruvate to formate and acetyl-CoA gcvR: putative transcriptional regulator GcvR Transcription factor that regulates the gcv operon 5

Genes and pathways that have been inserted Gene Species Function Formate-THF ligase bifunctional 5,10-methylene-tetrahydrofolate dehydrogenase/ 5,10-methylene-tetrahydrofolate cyclohydrolase Converts 10-CHO-THF to 5,10-CH=THF Converts 5,10-CH=THF to 5,10-CH2-THF serine hydroxymethyltransferase ftl Methylobacterium extorquens folD Escherichia coli fch Methylobacterium extorquens mtd glyA Methylobacterium extorquens Escherichia coli lipoamide dehydrogenase aminomethyltransferase glycine cleavage system H protein glycine decarboxylase L-serine deaminase I L-serine deaminase II serine:H(+) symporter SdaC lpd Escherichia coli gcvT Escherichia coli gcvH Escherichia coli gcvP Escherichia coli sdaA Escherichia coli sdaB Escherichia coli sdaC Escherichia coli 6

Overall framework of metabolic engineering steps The overarching objective of the authors work is to build a THF cycle to afford the assimilation of formate. However, the THF needs a source of glycine to produce serine. Hence, the gcvTHF operon for glycine cleavage pathway was over-expressed. Formate assimilation ends at serine. As serine is a fringe metabolite in central carbon metabolism, this may lead to accumulation of serine without kickstarting growth. Hence, there was a need to overexpress sadABC to push serine flux to pyruvate and onto biomass synthesis. 7

Use of the rTHF cycle improves formic acid assimilation: Data on methionine Classification of different strains THF1: equipped with THF cycle THF2: equipped with rTHF cycle THF3: equipped with stronger rTHF cycle THF4: equipped with THF cycle and overexpression of folD gene Results indicated that rTHF cycle work better than THF in assimilating formic acid 8

Use of the rTHF cycle improves formic acid assimilation: Data on serine Classification of different strains THF1: equipped with THF cycle THF2: equipped with rTHF cycle THF3: equipped with stronger rTHF cycle THF4: equipped with THF cycle and overexpression of folD gene Results indicated that rTHF cycle work better than THF in assimilating formic acid Same result was obtained for both serine and methionine amino acid 9

Formic acid concentration profile Results Reductive THF cycle resulted in the greatest consumption of formic acid. THF cycle with overexpression of folD did not result in significant consumption of formic acid. Similarly, THF cycle without modification did not result in significant consumption of formic acid. 10

Growth performance from strains with different THF cycles Results Cells more capable of assimilating formic acid did not engender higher cell growth. This suggests that formic acid flux channelled to serine could not be converted to cell growth given that serine is a fringe metabolite in central carbon metabolism. Importantly, rTHF cycle induced significant metabolic burden in cells and resulted in slower growth. 11

Increasing in vivo glycine biosynthesis from formic acid and carbon dioxide Classification of different strains RG1: Knock down of repressor of gcv operon RG2: Introduction of rTHF cycle into RG1 RG3: rTHF introduced to strain whose gcvTHP is under the control of trc promoter RG4: RG3 strain with pflB knockout and lpd overexpression RG5: RG3 strain with pflB knockout and serA knockout RG6: Strain with rTHF, gcvTHF and lpd over-expression on one plasmid RG7: RG6 strain with Methylobacterium extorquens fdh gene RG8: RG6 strain with C. boidinii fdh gene Results rTHF cycle was more capable of incorporating carbon from formic acid into glycine compared to strain with endogenous THF cycle. Activation of gcv pathway was common for both RG1 and RG2 strains. 12

Increasing in vivo serine biosynthesis from formic acid and carbon dioxide Classification of different strains RG1: Knock down of repressor of gcv operon RG2: Introduction of rTHF cycle into RG1 RG3: rTHF introduced to strain whose gcvTHP is under the control of trc promoter RG4: RG3 strain with pflB knockout and lpd overexpression RG5: RG3 strain with pflB knockout and serA knockout RG6: Strain with rTHF, gcvTHF and lpd over-expression on one plasmid RG7: RG6 strain with Methylobacterium extorquens fdh gene RG8: RG6 strain with C. boidinii fdh gene Results Enhanced activation of gcvTHP in RG2 strain resulted in better assimilation of formic acid flux into serine compared to RG2 strain. Flux to serine did not show significant deviation between RG3 to RG6 strains. 13

Increasing in vivo pyruvate biosynthesis from formic acid and carbon dioxide Classification of different strains RG1: Knock down of repressor of gcv operon RG2: Introduction of rTHF cycle into RG1 RG3: rTHF introduced to strain whose gcvTHP is under the control of trc promoter RG4: RG3 strain with pflB knockout and lpd overexpression RG5: RG3 strain with pflB knockout and serA knockout RG6: Strain with rTHF, gcvTHF and lpd over-expression on one plasmid RG7: RG6 strain with Methylobacterium extorquens fdh gene RG8: RG6 strain with C. boidinii fdh gene Results Coupling of rTHF and reductive glycine pathway help improve flux to pyruvate. Knocking out of pyruvate to formate transformation helps preserve pyruvate that has been synthesized. 14

Specific consumption rates of different metabolites and C1 utilisation ratio Results The strain equipped to assimilate formic acid and carbon dioxide still needs substantial amount of glucose to power cell growth. Utilisation rate of formic acid and carbon dioxide remains low. C1 utilisation ratio remains low, which reveals that only a small proportion of cellular building blocks derive its carbon flux from formic acid and carbon dioxide. Overall, the data suggests that formic acid may not be able to serve as a sole carbon source. 15

Attempts to improve formic acid utilisation and reduce glucose consumption Results Fed-batch experiment was used with respect to formic acid. But the experiment does not provide evidence that formic acid could be the sole carbon source since glucose was used in the initial part of the experiment. The speed of glucose uptake was also inexplicably slow in the initial part of the experiment given that the engineered E. coli should have sufficient capability to resist the antibiotic used, and the lag phase should not be so long. Cultivation of RG8 strain in M9 minimal medium with glucose and sodium formate supplementation in shake flasks Blue line: Cell growth, Yellow line: Glucose conc., Grey line: sodium formate conc. 16

Attempts to improve formic acid utilisation and reduce glucose consumption Results RG6 strain without fdh gene had a shorter lag phase compared to RG8 strain with the gene. Lack of fdh gene also explains reduced formic acid utilisation. But the same problem of poor uptake of glucose in RG8 strain also appears in RG6 strain. In addition, the lag phase should not be about 24 hours as E. coli should have sufficient capability to resist antibiotic and glucose is a preferred carbon source for the bacterium. Cultivation of RG6 strain in M9 minimal medium with glucose and sodium formate supplementation in shake flasks Blue line: Cell growth, Yellow line: Glucose conc., Grey line: sodium formate conc. Data collected suggests that formic acid could not serve as sole carbon source even in engineered E. coli with formate assimilation pathways. 17

Use of carbon-13 flux tracing to assess if formate flux could be channelled to different branches of metabolism from pyruvate entry point Approach taken in this segment The objective is to use carbon-13 flux tracing to determine if carbon flux from formate could flow to intermediates at different branches of central carbon metabolism such as gluconeogenesis and tricarboxylic acid cycle. 18