Multilayer Engineering of an Escherichia coli-Based Biotransformation System to Exclusively Produce Glycolic Acid from Formaldehyde

Abstract

A whole-cell biocatalysis was investigated for the selective synthesis of the industrially relevant C2 chemicals (e.g., glycolic acid (3)) from formaldehyde (1). Escherichia coli cells were engineered to overexpress a carboligase and an aldehyde dehydrogenase from E. coli K-12. Moreover, the side reactions, which dissipate formaldehyde and glycolic acid, were removed to produce glycolic acid to a high conversion. Host cell engineering to apply relatively chemical tolerant E. coli strains as well as substrate engineering to avoid the toxic effects of formaldehyde to the host cells allowed production of glycolic acid up to 27 mM in the reaction medium with a conversion of 85%. This study will contribute to valorization of C1 gas (e.g., CH 4 , CO 2 , and CO) to industrially relevant C2 chemicals in a sustainable way.

Identification of 2-Hydroxyacyl-CoA Synthases with High Acyloin Condensation Activity for Orthogonal One-Carbon Bioconversion

“…Remarkable consistency in yield from substantially different modes of fermentation and scales could attribute to the orthogonality of FORCE pathways from the host metabolism, which could indicate that further scale up to a higher scale in a bioreactor could be possible without the need for extensive process optimizations. To the best of our knowledge, the titer, rate, and yield of C1 to GA bioconversion reported in this study are the highest reported, exceeding the best result in the literature by more than 2.5-fold and 4-fold in titer and flux, respectively.…”

“…The best HACS variants identified from this study exhibit up to a 7.4-fold improvement in catalytic efficiency ( k cat / K M ) toward formaldehyde and a 7.5-fold improvement toward formyl-CoA as substrates compared to the starting reference enzyme, RuHACS (Table ). These are the best kinetic parameters reported for enzymes that catalyze C1–C1 condensation utilizing formaldehyde as the substrate, with orders of magnitude higher compared to engineered FLS and related variants − ,, and fold improvements from previously reported native or engineered HACSs , (Figure ). We were able to demonstrate that the superior kinetic parameters from in vitro enzyme characterization translates well into the in vivo bioconversion system as represented by the substantially improved pathway flux and selectivity compared to other studies on formaldehyde-to-GA bioconversion. ,, However, it is also important to note that among high performing HACS variants identified from this study, k cat / K M values do not necessarily represent the specific activity of the enzyme under the conditions tested.…”

“…These are the best kinetic parameters reported for enzymes that catalyze C1–C1 condensation utilizing formaldehyde as the substrate, with orders of magnitude higher compared to engineered FLS and related variants − ,, and fold improvements from previously reported native or engineered HACSs , (Figure ). We were able to demonstrate that the superior kinetic parameters from in vitro enzyme characterization translates well into the in vivo bioconversion system as represented by the substantially improved pathway flux and selectivity compared to other studies on formaldehyde-to-GA bioconversion. ,, However, it is also important to note that among high performing HACS variants identified from this study, k cat / K M values do not necessarily represent the specific activity of the enzyme under the conditions tested. This is because under the conditions tested, where substrate (formaldehyde) concentration is not significantly below the K M of enzymes, k cat becomes the dominant factor determining the “catalytic efficiency,” as seen from CfhHACS in comparison with ApbHACS, DhcHACS, and DhcHACS* (Figure and Table ).…”

“…26 For example, the engineered FLS that catalyzes C−C coupling of two or three formaldehyde molecules into glycolaldehyde or dihydroxyacetone (Figure 1) demonstrated a turnover frequency (k cat ) of only up to 1.5 s −1 and apparent Michaelis−Menten constants (K M ) for formaldehyde exceeding 100 mM, even after several rounds of directed evolution. [20][21][22][23]27 Considering the high toxicity of formaldehyde, 11,28 these enzymes are not suitable for use not only in actively growing cultures but also in resting cells. Efforts to substantially improve affinity of FLS toward formaldehyde (K M = 18 mM) led to an extremely poor turnover (k cat = 0.1 s −1 ), which requires high cell density (protein loading) to demonstrate product synthesis at the desired rate.…”

“…To address these issues, researchers developed different synthetic metabolic pathways, which are built upon newly discovered or engineered C1–C1 coupling enzymes, such as formolase (FLS) and associated variants − and 2-hydroxyacyl-CoA synthase (HACS). , However, because these enzymes often exploit a natural substrate promiscuity, many of them suffer from suboptimal kinetic parameters, even after extensive engineering efforts . For example, the engineered FLS that catalyzes C–C coupling of two or three formaldehyde molecules into glycolaldehyde or dihydroxyacetone (Figure ) demonstrated a turnover frequency ( k cat ) of only up to 1.5 s –1 and apparent Michaelis–Menten constants ( K M ) for formaldehyde exceeding 100 mM, even after several rounds of directed evolution. − , Considering the high toxicity of formaldehyde, , these enzymes are not suitable for use not only in actively growing cultures but also in resting cells. Efforts to substantially improve affinity of FLS toward formaldehyde ( K M = 18 mM) led to an extremely poor turnover ( k cat = 0.1 s –1 ), which requires high cell density (protein loading) to demonstrate product synthesis at the desired rate …”