Production of fatty acids in Ralstonia eutropha H16 by engineering β-oxidation and carbon storage

Abstract

Ralstonia eutropha H16 is a facultatively autotrophic hydrogen-oxidizing bacterium capable of producing polyhydroxybutyrate (PHB)-based bioplastics. As PHB's physical properties may be improved by incorporation of medium-chain-length fatty acids (MCFAs), and MCFAs are valuable on their own as fuel and chemical intermediates, we engineered R. eutropha for MCFA production. Expression of UcFatB2, a medium-chain-length-specific acyl-ACP thioesterase, resulted in production of 14 mg/L laurate in wild-type R. eutropha. Total fatty acid production (22 mg/L) could be increased up to 2.5-fold by knocking out PHB synthesis, a major sink for acetyl-CoA, or by knocking out the acyl-CoA ligase fadD3, an entry point for fatty acids into β-oxidation. As ΔfadD3 mutants still consumed laurate, and because the R. eutropha genome is predicted to encode over 50 acyl-CoA ligases, we employed RNA-Seq to identify acyl-CoA ligases upregulated during growth on laurate. Knockouts of the three most highly upregulated acyl-CoA ligases increased fatty acid yield significantly, with one strain (ΔA2794) producing up to 62 mg/L free fatty acid. This study demonstrates that homologous β-oxidation systems can be rationally engineered to enhance fatty acid production, a strategy that may be employed to increase yield for a range of fuels, chemicals, and PHB derivatives in R. eutropha.

Keywords: -oxidation; Acyl-CoA ligase; Biofuel; Metabolic engineering; Ralstonia.

Figure 1. Schematic of fatty acid production and uptake by Ralstonia eutropha H16.

Heterotrophic and autotrophic growth of R. eutropha H16 both result in production of the central metabolite acetyl-CoA (bold arrow). During fatty acid synthesis, derivatives of acetyl-CoA are iteratively condensed on an acyl carrier protein (ACP) scaffold to yield a series of acyl-ACPs of increasing length for lipid synthesis. Heterologous expression of chain-length-specific acyl-ACP thioesterases (tes) enables the hydrolysis of acyl-ACPs and release of fatty acids, some of which are capable of crossing the lipid membrane. Fatty acids can be re-consumed through the β-oxidation pathway, which iteratively removes two carbons from the fatty acid to yield acetyl-CoA (top dashed line). Entry of fatty acids into β-oxidation is catalyzed by one or more acyl-CoA ligases (fadD homologues). Multiple homologues of each step of the β-oxidation pathway have been identified in the genome of R. eutropha H16, with the number of homologues indicated in parentheses. Acetyl-CoA may also be converted to polyhydroxybutyrate (PHB) through the action of three enzymes (phaCAB) under periods of carbon excess but limitation of nutrients such as nitrogen or phosphate. β-oxidation and PHB synthesis share common metabolic intermediates and may be co-regulated. The dashed line indicates the potential for 3-hydroxyl-acyl-CoAs to enter PHB synthesis pathway and be incorporated into the polymer.

Figure 2. Production of lauric acid by engineered R. eutropha.

(A) Total free fatty acids (FFA) produced by ReH16 grown on rich broth (RB) and expressing UcFatB2 from plasmids (top row) of different predicted RBS strengths (RBS Calculator; bottom row). FFAs were detected in culture supernatant 48 h after arabinose induction by the fatty acid half-micro assay (Methods). Negligible amounts of free fatty acid (0.04 ± 0.06 mg/L) were detected for ReH16 expressing pJC025. (B). Chain length distribution of saturated fatty acids extracted from total culture of ReH16-pJC024 at 48 h post-induction via GCMS. For all experiments in this figure, error bars represent the standard error of the mean (S.E.M.) from N ≥ 3 independent experiments.

Figure 3. β-oxidation limits fatty acid yield in engineered R. eutropha.

(A) Time course of free fatty acid (FFA) production by different R. eutropha mutants in RB following arabinose induction (0 h time point). (B) Total FFA production at 24 h as a function of genetic background and UcFatB2 expression plasmid; each plasmid has a different RBS and predicted expression strength. (C) Time course of fatty acid uptake by R. eutropha of different genetic backgrounds incubated with 120 mg/L lauric acid in RB. For all experiments in this figure, FFA concentration was measured using the fatty acid half-micro assay (Methods), and error bars represent S.E.M. for N ≥ 3 independent experiments.

Figure 4. Three additional acyl-CoA ligases limit FFA yield.

(A) Schematic diagram of β-oxidation in R. eutropha. The first enzymatic step of β-oxidation is carried out by an acyl-CoA ligase (ACL) followed by an acyl-CoA dehydrogenase (ACD), enoyl-CoA hydratase (ECH), 3-hydroxyacyl-CoA dehydrogenase (HCD) and β-ketoacyl-CoA thiolase (BKT). Each β-oxidation gene has multiple annotated homologues in the R. eutropha genome (number indicated in parentheses). (B) Time course of Re2061 growth on fructose or laurate minimal medium. Error bars represent S.E.M. from N = 3 independent experiments. The grey arrow indicates the point at which cultures were sampled for RNA-Seq analysis. (C) Expression levels of four acyl-CoA ligases upregulated during growth on laurate, measured by RNA-Seq and normalized by the upper quartile method (Methods). Numbers above bars indicate fold-increase in laurate media as compared to fructose media. Error bars represent S.E.M. of the normalized expression levels for three independent experiments. Differences in expression between fructose and laurate conditions are significant for all four genes after correcting for multiple hypothesis testing (q-value < 0.001). (D) Total FFA production 96 h after arabinose induction by R. eutropha knockout mutants, as measured by the fatty acid half-micro assay (top). OD600 96 h after induction for each mutant. Error bars represent S.E.M. from N = 4 independent experiments in both panels. ^∗ indicates a significant difference in mean FFA production with p < 0.05, evaluated using the one-tailed student’s t-test.