Background Mitigation of climate change requires that new routes for the production of fuels and chemicals be as oil-independent as possible. The microbial conversion of lignocellulosic feedstocks into terpene-based biofuels and bioproducts represents one such route. This work builds upon previous demonstrations that the single-celled carotenogenic basidiomycete, Rhodosporidium toruloides, is a promising host for the production of terpenes from lignocellulosic hydrolysates. Results This study focuses on the optimization of production of the monoterpene 1,8-cineole and the sesquiterpene α-bisabolene in R. toruloides. The α-bisabolene titer attained in R. toruloides was found to be proportional to the copy number of the bisabolene synthase (BIS) expression cassette, which in turn influenced the expression level of several native mevalonate pathway genes. The addition of more copies of BIS under a stronger promoter resulted in production of α-bisabolene at 2.2 g/L from lignocellulosic hydrolysate in a 2-L fermenter. Production of 1,8-cineole was found to be limited by availability of the precursor geranylgeranyl pyrophosphate (GPP) and expression of an appropriate GPP synthase increased the monoterpene titer fourfold to 143 mg/L at bench scale. Targeted mevalonate pathway metabolite analysis suggested that 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase (HMGR), mevalonate kinase (MK) and phosphomevalonate kinase (PMK) may be pathway bottlenecks are were therefore selected as targets for overexpression. Expression of HMGR, MK, and PMK orthologs and growth in an optimized lignocellulosic hydrolysate medium increased the 1,8-cineole titer an additional tenfold to 1.4 g/L. Expression of the same mevalonate pathway genes did not have as large an impact on α-bisabolene production, although the final titer was higher at 2.6 g/L. Furthermore, mevalonate pathway intermediates accumulated in the mevalonate-engineered strains, suggesting room for further improvement. Conclusions This work brings R. toruloides closer to being able to make industrially relevant quantities of terpene from lignocellulosic biomass.

One of the requirements for efficient biological conversion of lignocellulose to bioproducts is the compatibility of biological catalysts with the processes employed to solubilize and depolymerize the lignocellulosic components. The red yeasts Rhodosporidium toruloides and Rhodotorula mucilaginosa were evaluated for their ability to assimilate sugars and aromatic compounds extracted from two engineered lines of Arabidopsis thaliana with modified lignin or the wildtype using ionic liquid, acid or alkaline pretreatments. Differential amounts of monomeric sugars, organic acids and, in the case of the engineered lines, either 4-hydroxybenzoic or protocatechuic acid were additionally released from the biomass and found to be tolerated and consumed by both microorganisms. Genetically-engineered strains of the two red yeasts successfully converted the depolymerized products into the biofuel precursor bisabolene when cultivated on hydrolysates or synthetic media containing specific sugars, acids and aromatics found in the hydrolysates.

Summary The bacterium Streptomyces davaonensis synthesizes the antibiotic roseoflavin in the stationary phase of growth. The starting point for roseoflavin biosynthesis is riboflavin (vitamin B2) and four enzymes (RibCF, RosB, RosA and RosC) are necessary to convert a vitamin (riboflavin) into a potent, broad‐spectrum antibiotic (roseoflavin). In S. davaonensis, seven enzymatic functions are required to synthesize the roseoflavin precursor riboflavin from the central building blocks GTP and ribulose 5‐phosphate. When compared with other bacterial and in particular Streptomyces genomes the S. davaonensis genome contains an unusual high number (21) of putative riboflavin biosynthetic genes (rib genes), including a rib gene encoding an additional riboflavin synthase originating from an Archaeon. We show by complementation analyses and enzyme assays that 17 out of these 21 putative rib genes indeed encode for riboflavin biosynthetic enzymes. Biochemical analyses of selected enzymes support this finding. Transcriptome analyses show that all of the rib genes are expressed either in the exponential or in the stationary phase of growth and thus do not represent silent genes. We conclude that the Rib enzymes produced in the stationary phase represent a physiological adaptation to support roseoflavin biosynthesis.