Engineering efficient D-xylose fermentation capacity in industrial Saccharomyces cerevisiae for advanced bioethanol production

SummaryBioethanol production using non-food biomass such as agricultural wastes, energy crops, and forest residues as substrates, is advancing to commercial production scale. However, economically competitive ethanol production from such lignocellulosic biomass remains a great challenge. One of the reasons is the lack of microorganisms that efficiently convert all the sugars present in the lignocellulose hydrolysates into ethanol under industrial conditions. The baker s yeast Saccharomyces cerevisiae is still the dominant organism for industrial ethanol production owing to its high rate of fermentation of hexose sugars, very good tolerance to ethanol and to inhibitors in lignocellulosic hydrolysates. Unfortunately, baker s yeast is unable to metabolize pentose sugars, particularly D-xylose, which accounts for up to 35% of total sugars in lignocellulosic biomass. Efficient utilization of D-xylose, in addition to inhibitor tolerance, is required in organisms to be used for cost effective and sustainable production of ethanol from lignocellulosic material. Our general objective was therefore to construct a robust yeast strain that efficiently converts D-xylose into ethanol and is tolerant to inhibitors present in lignocellulosic hydrolysates. For that purpose, this project started with one of the most widely used first generation bioethanol production yeast strains (Ethanol Red), in which a Clostridium phytofermentans xylA based D-xylose and an L-arabinose gene cassette had previously been inserted. Despite the presence of all known genes required for D-xylose and L-arabinose utilization in the genome, the recombinant industrial strain was unable to utilize any D-xylose or L-arabinose. We applied a systematic evolutionary engineering approach (random mutagenesis, genome shuffling followed by selection in a D-xylose-enriched lignocellulose hydrolysate, and adaptive evolution in D-xylose) to establish rapid D-xylose fermentation capacity in the recombinant strain. Consequently, we were able to develop a robust D-xylose fermenting strain with a productivity and yield of ethanol that was higher than that of any reported recombinant industrial strain of S. cerevisiae. The evolved strain GS1.11-26 demonstrated substantial tolerance to inhibitor-containing lignocellulosic hydrolysates producing ethanol with a yield close to the maximum theoretical yield. However, the evolved strain GS1.11-26 showed a partial respiratory defect causing a reduced aerobic growth rate and it also had a slightly reduced inhibitor tolerance compared to the original strain. To enhance the direct industrial applicability of the strain, we further eliminated the negative properties of GS1.11-26 (slow growth rate, and reduced inhibitor tolerance) using a meiotic recombination tool. The method combines the superior genetic elements from the genomes of the D-xylose utilizing strain GS1.11-26 with either a haploid segregant of Ethanol Red or a highly inhibitor tolerant diploid strain Fseg25. In this way, we successfully developed three robust industrial yeast strains that combine efficient D-xylose utilization with high inhibitor tolerance for use in bioethanol production with lignocellulose hydrolysates. From the economic point of view, the use of these strains will significantly contribute to the reduction of the overall ethanol production cost, especially from second generation feedstocks.Using whole genome sequence comparison between the evolved and the original parent strain, amplification of the xylA gene has been identified and experimentally verified as a crucial but not the only genetic change responsible for the high D-xylose utilization rate in GS1.11-26. In addition, we showed that an extra-chromosomal circular DNA (eccDNA) carrying xylA has been formed in the process of evolutionary adaptation. Chromosomal integration of the eccDNA over the course of the evolutionary adaptation steps as tandem repeat resulted in amplification of xylA and was stable for several generations without selection. The high copy number of xylA correlated with the high activity of D-xylose isomerase in GS1.11-26. Quantitative trait loci (QTL) mapping using a modified pooled segregant whole genome sequence analysis resulted in the identification of at least three genomic loci that are linked to the fast D-xylose fermentation rate in GS1.11-26. After evaluation of one of the QTLs by reciprocal hemizygosity analysis, a mutation in a gene that hadn t been associated previously with D-xylose fermentation was found to be linked with the high D-xylose utilization rate in GS1.11-26. The genetic changes identified in this study, together with other genetic factors yet to be identified in the other QTLs may be transferred to other target industrial strains to endow efficient D-xylose fermentation capacity. We also showed for the first time that QTL mapping can be performed directly in diploid or aneuploid strains without the need to isolate haploid derivatives. This will facilitate the polygenic analysis of important industrial traits in aneuploid or polyploid industrial strains. nrpages: 161status: publishe