Monoterpene production by the carotenogenic yeast Rhodosporidium toruloides

Abstract

Background: Due to their high energy density and compatible physical properties, several monoterpenes have been investigated as potential renewable transportation fuels, either as blendstocks with petroleum or as drop-in replacements for use in vehicles (both heavy and light-weight) or in aviation. Sustainable microbial production of these biofuels requires the ability to utilize cheap and readily available feedstocks such as lignocellulosic biomass, which can be depolymerized into fermentable carbon sources such as glucose and xylose. However, common microbial production platforms such as the yeast Saccharomyces cerevisiae are not naturally capable of utilizing xylose, hence requiring extensive strain engineering and optimization to efficiently utilize lignocellulosic feedstocks. In contrast, the oleaginous red yeast Rhodosporidium toruloides is capable of efficiently metabolizing both xylose and glucose, suggesting that it may be a suitable host for the production of lignocellulosic bioproducts. In addition, R. toruloides naturally produces several carotenoids (C40 terpenoids), indicating that it may have a naturally high carbon flux through its mevalonate (MVA) pathway, providing pools of intermediates for the production of a wide range of heterologous terpene-based biofuels and bioproducts from lignocellulose.

Results: Sixteen terpene synthases (TS) originating from plants, bacteria and fungi were evaluated for their ability to produce a total of nine different monoterpenes in R. toruloides. Eight of these TS were functional and produced several different monoterpenes, either as individual compounds or as mixtures, with 1,8-cineole, sabinene, ocimene, pinene, limonene, and carene being produced at the highest levels. The 1,8-cineole synthase HYP3 from Hypoxylon sp. E74060B produced the highest titer of 14.94 ± 1.84 mg/L 1,8-cineole in YPD medium and was selected for further optimization and fuel properties study. Production of 1,8-cineole from lignocellulose was also demonstrated in a 2L batch fermentation, and cineole production titers reached 34.6 mg/L in DMR-EH (Deacetylated, Mechanically Refined, Enzymatically Hydorlized) hydrolysate. Finally, the fuel properties of 1,8-cineole were examined, and indicate that it may be a suitable petroleum blend stock or drop-in replacement fuel for spark ignition engines.

Conclusion: Our results demonstrate that Rhodosporidium toruloides is a suitable microbial platform for the production of non-native monoterpenes with biofuel applications from lignocellulosic biomass.

Keywords: 1,8-Cineole; Biofuel; Monoterpene; Rhodosporidium toruloides.

Fig. 1

Schematic diagram of R. toruloides native mevalonate MVA pathways. The native MVA pathway provides the opportunity for diverting the intermediate GPP from the MVA pathway toward the biosynthesis of heterologous monoterpene bioproducts

Fig. 2

Cineole production in R. toruloides. a Binary vector containing the cineole synthase designed for genome integration in R. toruloides through the ATMT transformation method, which could also be used as template for PCR amplification to transform R. toruloides through the electroporation transformation method. b The comparison of cineole production in R. toruloides strains generated by the two transformation methods. c GC chromatographs of R. toruloides engineered for expression of the cineole synthase Hyp3 using a dodecane overlay, and d SSCG_00536 CnsA SPME sample

Fig. 3

GC chromatographs of SPME samples from engineered R. toruloides. Expression of a the Ocimene synthase ama0e23, b limonene synthase Ag10, c pinene synthase PT30, d Ag3.18, e sabinene synthase SabS1 and f carene synthase TpsB. Compounds that could be identified with authentic standards are labeled above the peaks of GC chromatograph. Peaks that might be from a monoterpene based on mass spectrum patterns, but lacking authentic standards, are labeled as unknown compounds

Fig. 4

The strawberry-derived bifunctional enzyme FaNES1 was expressed in R. toruloides to examine the in vivo GPP and FPP metabolite pools. a FaNES1 can covert GPP into linalool and FPP into nerolidol. b Only nerolidol was produced in R. toruloides, with no detectable linalool

Fig. 5

1,8-cineole production from DMR-EH hydrolysate in shake flasks. Cineole was produced in R. toruloides strain 17-3 in DMR-EH hydrolysate or a mock hydrolysate medium containing the same glucose and xylose concentrations (a). Cells grown in YPD medium served as the control (a). Comparison of cell growth and cineole production over 13 days in 75% DMR-EH with SD as the nitrogen source (b), 90% DMR-EH with SD as the nitrogen source (c), 75% DMR with yeast extract as the nitrogen source, and d 75% DMR-EH with SD and ammonium sulfate as the nitrogen source (e)

Fig. 6

1,8-cineole production in DMR-EH hydrolysate in a 2L bioreactor. R. toruloides was cultivated in b 75% DMR-EH hydrolysate or a mock medium with the same amounts of glucose and xylose, each supplemented with 10 g/L yeast extract as a nitrogen source