One-Step Purification of Microbially Produced Hydrophobic Terpenes via Process Chromatography

Abstract

Novel and existing terpenes are already being produced by genetically modified microorganisms, leading to new process challenges for the isolation and purification of these terpenes. Here, eight different chromatographic resins were characterized for the packed bed adsorption of the model terpene β-caryophyllene, showing their applicability on an Escherichia coli fermentation mixture. The polystyrenic Rensa^® RP (Ø 50 μm) shows the highest affinity, with a maximum capacity of >100 g L^-1 and the best efficiency, with a height equivalent of a theoretical plate (HETP) of 0.022 cm. With this material, an optimized adsorption-based purification of β-caryophyllene from a fermentation mixture was developed, with the green solvent ethanol for desorption. A final yield of >80% and a purity of >99% were reached after only one process step with a total productivity of 0.83 g h^-1 L^-1. The product solution was loaded with a volume ratio (feed to column) of >500 and the adapted gradient elution yielded a 40 times higher concentration of β-caryophyllene. The adsorption-based chromatography represents therefore a serious alternative to the liquid-liquid extraction and achieves desired purities without the utilization of hazardous solvents.

Keywords: fermentation; preparative chromatography; process development and integration; purification; terpenes; β-caryophyllene.

Figure 1

Chemical structure of the experimentally investigated sesquiterpene β-caryophyllene, consisting solely of hydrocarbons. It is extremely hydrophobic and poorly soluble in water with an octanol-water partition coefficient logP > 5.

Figure 2

Properties (permeability, packing density, and total porosity) of the packed adsorbents in a chromatographic column, measured at 100% DI H₂O with uracil and a flowrate of 1 mL min⁻¹.

Figure 3

Performance characterization of the packed bed adsorbents in terms of HETP and reduced HETP. The values were determined at 100% DI H₂O with uracil at 1 mL min⁻¹ flow rate.

Figure 4

Henry constants from batch adsorption experiments at an initial concentration of 0.1 g L⁻¹ β-caryophyllene (A) for all investigated adsorbents at different ethanol (blue) and acetonitrile (gray) concentrations. Rensa^® RP shows the highest Henry constant with 60% ethanol. Retention factors for different adsorbents with β-caryophyllene (B,C) subject to the solvent concentration of ethanol and acetonitrile show extreme variances in the affinity.

Figure 5

(A) Adsorption isotherm for β-caryophyllene on Rensa^® RP at 70% ethanol determined with the frontal analysis (FA). (B) Dependence of the Dynamic Binding Capacity (DBC) subject to the ethanol concentration for β-caryophyllene on Rensa^® RP. Shown are the capacities for the breakthrough at 5, 10, and 20% (relatively to the initial concentration; Column volume = 0.46 mL, flow rate = 2.9 cm min⁻¹, HETP = 0.02 cm).

Figure 6

Elution chromatogram of β-caryophyllene from Rensa^® RP column (CV = 0.9 mL, HETP = 0.02 cm) after loading with LB-media or fermentation supernatant with an adapted ethanol gradient. The purity of >99% (HPLC) for β-caryophyllene could be reached by this single bind-and-elute step. The first major impurities are mostly peptides followed by a second impurity (fatty acids) at around 80% ethanol, which was separated by a factor >1.5 from the product.