Engineering Escherichia coli for Microbial Production of Butanone

Abstract

To expand the chemical and molecular diversity of biotransformation using whole-cell biocatalysts, we genetically engineered a pathway in Escherichia coli for heterologous production of butanone, an important commodity ketone. First, a 1-propanol-producing E. coli host strain with its sleeping beauty mutase (Sbm) operon being activated was used to increase the pool of propionyl-coenzyme A (propionyl-CoA). Subsequently, molecular heterofusion of propionyl-CoA and acetyl-CoA was conducted to yield 3-ketovaleryl-CoA via a CoA-dependent elongation pathway. Lastly, 3-ketovaleryl-CoA was channeled into the clostridial acetone formation pathway for thioester hydrolysis and subsequent decarboxylation to form butanone. Biochemical, genetic, and metabolic factors affecting relative levels of ketogenesis, acidogenesis, and alcohol genesis under selected fermentative culture conditions were investigated. Using the engineered E. coli strain for batch cultivation with 30 g liter(-1)glycerol as the carbon source, we achieved coproduction of 1.3 g liter(-1)butanone and 2.9 g liter(-1)acetone. The results suggest that approximately 42% of spent glycerol was utilized for ketone biosynthesis, and thus they demonstrate potential industrial applicability of this microbial platform.

FIG 1

Schematic representation of the butanone biosynthetic pathway. Green text, heterologous enzymes from Cupriavidus necator (CN) and Clostridium acetobutylicum (CA); green box, fermentative pathway for glycerol dissimilation; yellow box, respiratory pathway for glycerol dissimilation; purple box, sleeping beauty mutase (Sbm) pathway; blue arrows, route to the C₃-fementative products; dark-orange box, C₂-fermentative pathway; blue box, C₃-fermentative pathway; light-orange box, heterologous acetone formation pathway; gray box, heterologous butanone formation pathway; blue text, relevant enzymes for production of various fermentative products as well as the enzymes of the respiratory and fermentative glycerol pathways and the Sbm pathway; red arrows, competing pathways at the pyruvate/acetyl-CoA and propionyl-CoA nodes. Abbreviations: HPr, phosphocarrier protein; P_i, inorganic phosphate; UQ, ubiquinone.

FIG 2

Synthetic biology strategies used for heterologous production of butanone in engineered E. coli. The ketone biosynthesis pathway consisted of three modules, as follows: module 1, the chromosomally activated Sbm pathway in CPC-Sbm for supply of propionyl-CoA; module 2, a set of promiscuous β-ketothiolases (i.e., PhaA and BktB) expressed independently or in tandem to generate the CoA-dependent chain elongation pathway for either homofusion of acetyl-CoA or heterofusion of acetyl-CoA and propionyl-CoA; module 3, the clostridial acetone formation pathway (i.e., CtfAB-Adc) for thioester hydrolysis of the fused CoA intermediate and subsequent decarboxylation. All three modules were assembled together to generate various ketone production strains, i.e., CPC-MEKCon1 contained modules 1, 2A, and 3, CPC-MEKCon2 contained modules 1, 2B, and 3, and CPC-MEK contained modules 1, 2C, and 3. The single-gene knockouts (i.e., adhE, pta, glpD, and dhaK) are all variants of the parent ketogenic strain CPC-MEK. Abbreviations: FRT, flippase recognition target; O, operator; RBS, ribosome binding site; T, terminator.

FIG 3

Construction of the biosynthetic pathway for ketone production in E. coli. (A and B) Final metabolite titers, including acetone and butanone, with microaerobic cultivation of the ketogenic strains CPC-MEKCon1 (A) and CPC-MEKCon2 (B), using 30 g liter⁻¹ glycerol as the major carbon source. (C and D) Final metabolite titers, including acetone and butanone, with microaerobic cultivation (C) and semiaerobic cultivation (D) of the ketogenic strain CPC-MEK, using 30 g liter⁻¹ glycerol as the major carbon source. All of the strains were induced at the start of the batch cultivation with 0.1 mM IPTG. Error bars, standard deviations (n = 2). The metabolite distribution (denoted as the percentage of spent glycerol converted to a specific metabolite) is indicated above the bars.

FIG 4

Enhancement of ketone production in engineered E. coli. Final metabolite titers with microaerobic cultivation of CPC-SbmΔadhE (A), semiaerobic cultivation of CPC-SbmΔadhE (B), semiaerobic cultivation of CPC-SbmΔpta (C), semiaerobic cultivation of CPC-SbmΔglpD (D), and semiaerobic cultivation of CPC-SbmΔdhaK (E), using 30 g liter⁻¹ glycerol as the major carbon source, are shown. All of the strains were treated with 0.1 mM IPTG at the start of the batch cultivation. Error bars, standard deviations (n = 2). The metabolite distribution (denoted as the percentage of spent glycerol converted to a specific metabolite) is indicated above the bars.