Metabolic engineering of Bacillus subtilis for production of para-aminobenzoic acid - unexpected importance of carbon source is an advantage for space application

Abstract

High-strength polymers, such as aramid fibres, are important materials in space technology. To obtain these materials in remote locations, such as Mars, biological production is of interest. The aromatic polymer precursor para-aminobenzoic acid (pABA) can be derived from the shikimate pathway through metabolic engineering of Bacillus subtilis, an organism suited for space synthetic biology. Our engineering strategy included repair of the defective indole-3-glycerol phosphate synthase (trpC), knockout of one chorismate mutase isozyme (aroH) and overexpression of the aminodeoxychorismate synthase (pabAB) and aminodeoxychorismate lyase (pabC) from the bacteria Corynebacterium callunae and Xenorhabdus bovienii respectively. Further, a fusion-protein enzyme (pabABC) was created for channelling of the carbon flux. Using adaptive evolution, mutants of the production strain, able to metabolize xylose, were created, to explore and compare pABA production capacity from different carbon sources. Rather than the efficiency of the substrate or performance of the biochemical pathway, the product toxicity, which was strongly dependent on the pH, appeared to be the overall limiting factor. The highest titre achieved in shake flasks was 3.22 g l^-1 with a carbon yield of 12.4% [C-mol/C-mol] from an amino sugar. This promises suitability of the system for in situ resource utilization (ISRU) in space biotechnology, where feedstocks that can be derived from cyanobacterial cell lysate play a role.

Figure 1

Toxicity of pABA to the Bacillus subtilis strains 168 and SMY at different pHs. pABA concentration in the gradient agar plates ranged from 0 mM (bottom‐end of figure) to 24 mM (top‐end of figure), and tested pHs were (from left to right) 5, 6 and 7.

Figure 2

Simplified shikimate pathway and its integration into wider central metabolism, including modifications for pABA production. Knockout targets are highlighted in red, heterological genes that were overexpressed in green. The significant intermediates are glucose 6‐phosphate (G6P), erythrose 4‐phosphate (E4P), phosphoenolpyruvate (PEP), pyruvate (PYR), 3‐deoxy‐d‐arabino‐heptulosonate‐7‐phosphate (DAHP), shikimate, shikimate‐3‐phosphate (S3P), chorismate, phenylalanine (PHE), tyrosine (TYR), tryptophan (TRP) and para‐aminobenzoic acid (pABA). Important genes and the respective enzymes involved are aroA: 3‐deoxy‐d‐arabino‐heptulosonate‐7‐phosphate (DAHP) synthase/chorismate mutase, aroH: chorismate mutase, trpC: indole‐3‐glycerol phosphate synthase, pabAB: aminodeoxychorismate synthase, pabC: aminodeoxychorismate lyase.

Figure 3

pABA production by differently engineered B. subtilis strains. Production from 5 g l⁻¹ glucose (≙ 0.1665 C‐mol l⁻¹) M9 minimal medium in shake‐flask experiments (end‐point samples shown, taken 12 h after inoculation). Base strains: B. subtilis 168 (A) auxotrophic trpC ₂, (B) prototrophic trpC ⁺ and (C) B. subtilis SMY (prototrophic).

Figure 4

Fluorescence‐based detection of tagged pab enzymes in crude protein extract. Samples were taken from exponential growth phase. Lane 1–3: 168 trpC ⁺ ΔaroH pabAB ^t pabC ^t (three biological replicates). Lane 4–6: 168 trpC ⁺ ΔaroH pabAB ^t C (three biological replicates). Lane 7–10: controls (background strains 168 trpC ⁺ and 168 trpC ⁺ ΔaroH, two biological replicates each).

Figure 5

(A–C) pABA production on different carbon sources – titres, yields and rates. Strains 168 trpC ⁺ ΔaroH pabAB pabC and SMY ΔaroH pabAB pabC and their xyl⁺ derivatives were compared on glucose (GLC), sucrose (SUCR), glycerol (GLY), glucosamine (GlcN), acetyl‐glucosamine (AcGlcN) and xylose (XYL) using the equivalent of 5 g l⁻¹ glucose (0.1665 C‐mol l⁻¹). (A) Shows the maximum achieved titres. (B) The carbon yields. (C) The productivity averaged over the course of the cultivation. Yields are given as fraction of theoretical maximum: Y _real = carbon yield obtained in shake‐flask experiments with engineered strains (calculated based on supplied carbon source and final product titre), Y _max = theoretical maximum product yields determined by metabolic modelling (cf., experimental procedures).