Whole-Cell Production of Patchouli Oil Sesquiterpenes in Escherichia coli: Metabolic Engineering and Fermentation Optimization in Solid-Liquid Phase Partitioning Cultivation

Abstract

Patchouli oil is a major ingredient in perfumery, granting a dark-woody scent due to its main constituent (-)-patchoulol. The growing demand for patchouli oil has raised interest in the development of a biotechnological process to assure a reliable supply. Herein, we report the production of patchouli oil sesquiterpenes by metabolically engineered Escherichia coli strains, using solid-liquid phase partitioning cultivation. The (-)-patchoulol production was possible using the endogenous methylerythritol phosphate pathway and overexpressing a (-)-patchoulol synthase isoform from Pogostemon cablin but at low titers. To improve the (-)-patchoulol production, the exogenous mevalonate pathway was overexpressed in the multi-plasmid PTS + Mev strain, which increased the (-)-patchoulol titer 5-fold. Fermentation was improved further by evaluating several defined media, and optimizing the pH and temperature of culture broth, enhancing the (-)-patchoulol titer 3-fold. To augment the (-)-patchoulol recovery from fermentation, the solid-liquid phase partitioning cultivation was analyzed by screening polymeric adsorbers, where the Diaion HP20 adsorber demonstrated the highest (-)-patchoulol recovery from all tests. Fermentation was scaled-up to fed-batch bioreactors, reaching a (-)-patchoulol titer of 40.2 mg L^-1 and productivity of 20.1 mg L^-1 d^-1. The terpene profile and aroma produced from the PTS + Mev strain were similar to the patchouli oil, comprising (-)-patchoulol as the main product, and α-bulnesene, trans-β-caryophyllene, β-patchoulene, and guaia-5,11-diene as side products. This investigation represents the first study of (-)-patchoulol production in E. coli by solid-liquid phase partitioning cultivation, which provides new insights for the development of sustainable bioprocesses for the microbial production of fragrant terpenes.

Figure 1

Engineering of the biosynthetic pathway for (−)-patchoulol production in E. coli. Heterologous genes appear in blue. Mevalonate (MEV) pathway genes: Acetyl-CoA acetyltransferase (AtoB), HMG-CoA synthase (HMGS), truncated HMG-CoA reductase (tHMGR), mevalonate kinase (MK), phosphomevalonate kinase (PMK), mevalonate diphosphate decarboxylase (PMD), isoprenyl diphosphate isomerase (IDI), and farnesyl diphosphate synthase (ISPA). (−)-Patchoulol pathway: (−)-patchoulol synthase isoform (PTS).

Figure 2

Production of (−)-patchoulol and growth kinetics from the Mev, PTS, and PTS + Mev E. coli strains during 72 h of cultivation. (A) (−)-Patchoulol titer and (B) (−)-patchoulol yield, Y_P/X. (C) Cell growth shown by the optical density (OD₆₀₀). Plots correspond to the mean of three independent tests with standard deviation (SD) error bars.

Figure 3

Gas chromatography–mass spectrometry (GC–MS) chromatograph comparing the terpene profile between the patchouli oil (A) and the PTS + Mev E. coli strain (B). (C) Mass spectra from identified terpenes from PTS + Mev E. coli strain. A full comparison of the mass spectra of terpene products between the patchouli oil and the PTS + Mev strain is found in Table S2.

Figure 4

Optimization of cultivation conditions for the whole-cell production of (−)-patchoulol. Factorial design of the temperature and pH in culture media: (A) (−)-Patchoulol titer and (B) (−)-patchoulol yield, Y_P/X. Evaluation of defined media: (C) (−)-Patchoulol titer and yield Y_P/X. (D) Growth kinetics described by the cell dry weight (CDW) and optical density (OD₆₀₀). PT: (−)-patchoulol. Data are the mean of three independent tests with SD error bars and analyzed by analysis of variance (ANOVA) (α = 0.05). Letters that differ are significantly different.

Figure 5

Comparison of the (−)-patchoulol recovery from different polymeric adsorbers. (A) (−)-Patchoulol titer recovered from adsorbers, cell-free media, and cells. (B) Growth kinetics described by the cell dry weight (CDW) and optical density (OD₆₀₀). Control (−) test without extractants. Liquid–liquid phase partitioning cultivation (LLPPC) test with an organic solvent overlay. Data are the mean of three independent tests with SD error bars. Data analyzed by ANOVA (α = 0.05). Letters that differ are significantly different.

Figure 6

(−)-Patchoulol production in a 2 L bioreactor from the SH-PTS + Mev (A) and PTS + Mev (B) E. coli strains, represented as the (−)-patchoulol titer recovered from adsorbers (green circle filled), cell-free media (■), and cells (⧫). Growth kinetics depicted by the optical density (OD₆₀₀) (magenta circle filled), glucose (■), and acetate (⧫) from the SH-PTS + Mev (C) and PTS + Mev (D) E. coli strains. Data are the mean of three sample replicates with SD error bars.