Unraveling the polygenic basis of HMF and furfural tolerance in yeast for optimization of lignocellulose-based processes

One of the major challenges society faces today is global warming, mainly driven by greenhouse gas emissions. To reduce fossil fuel consumption and related CO2emissions in the transportation sector, second-generation (2G) bioethanol - produced from the abundantly available lignocellulosic biomass (i.e. waste products or energy crops) - is proposed as an alternative transport fuel. Moreover, the use of different plant-based waste materials holds a great promise to produce a huge amount of other biofuels and bio-based chemicals. To obtain an economically viable lignocellulose-based production process, the fermentation of sugars present in the lignocellulosic hydrolysate by Saccharomyces cerevisiaerequires improvement. Major challenges include enzyme cost, the inability of natural S. cerevisiaestrains to ferment D-xylose, and stress factors and fermentation inhibitors present in the lignocellulose hydrolysate that inhibit the fermentation performance of 2G yeasts. Several of these hurdles have been overcome, for example by metabolically engineering industrial S. cerevisiaestrains with a xylose isomerase gene, enabling these yeasts with the ability to ferment D-xylose. Nevertheless, D-xylose fermentation by 2G yeasts is much slower compared to that of D-glucose, especially in lignocellulose hydrolysates due to presence of inhibitory compounds that affect the fermentation process. With regard to obtaining cost-efficient lignocellulose-based processes, the use of cheap methods to pretreat the lignocellulosic materials is required, but in general, this results in a high level of fermentation inhibitors. In this research project, we described for the first time the effect of the major inhibitors present in lignocellulose hydrolysates for industrially relevant concentrations, in different lignocellulose hydrolysates of different biomass origin, on multiple relevant 2G yeast strains. Here, furfural was identified universally as the major inhibitor negatively affecting not only D-xylose, but also D-glucose fermentation by 2G yeasts. In addition, we observed that fermentation performance by these yeasts was reduced for industrially relevant concentrations of HMF and to a lesser extent formic acid and acetic acid. Several inhibitor tolerant S. cerevisiaestrains or non-conventional yeasts have been identified by other research groups, but to our knowledge, not in a very extensive study. Therefore, we screened an S. cerevisiaecollection of 2527 strains, and identified, for each phenotype analyzed, only few strains with superior tolerance to either levulinic acid, formic acid, HMF, furfural, or a combination of these inhibitors for growth and fermentation performance. In addition, we identified from 17non-conventional yeast species, those most tolerant to furfural or HMF, or those that showed growth in presence of high concentrations of D-xylose or L-arabinose. All these yeast strains identified could be useful for strain improvement of relevant traits of 2G yeasts. With regard to strain improvement, we performed WGT to improve HMF or furfural tolerance in several industrially relevant host strains with genomic DNA of HMF or furfural tolerant S. cerevisiae strains or non-conventional yeasts as donor DNA. This resulted, in all cases, in stable whole-genome transformants displaying either improved HMF or furfural tolerance, for which other traits crucial for 2G fermentations were not affected. Moreover,our data supports the hypothesis that one of the proposed WGT mechanisms confers temporary protection to the host strain by uptake of DNA from the donor strain, that temporarily remains transiently present, allowing the host strain to generate a spontaneous rescuing mutation under the high stress applied. Furthermore, whole-genome sequence and bio-informatics analysis identifiedAST2N406Ias solely causative for the high HMF tolerance obtained after WGT of MD4 with gDNA of an HMF tolerant C. glabratastrain. AST2N406Iwas shown to improve HMF, furfural, vanillin, and acetic acid tolerance, in multiple inhibitor-rich media and in a lignocellulose hydrolysate, and in multiple industrial genetic backgrounds. In addition, this mutation improved ethanol production, reduced acetaldehyde accumulation and reduced biomass formation. Our data indicates the role of AST2N406Iin preferential conversion of acetate into acetaldehyde, and subsequently into ethanol, favored over acetate conversion into acetyl-coA, required for lipid biosynthesis. Furthermore, preliminary data suggests that AST1D405Isupports the function of AST2N406Iin conferring inhibitor tolerance to yeast strains, yet only at high inhibitor levels present and in yeast strains that contain the AST2N406Imutation. Our data strongly suggests the role of AST2, and possibly AST1, as aldehyde reductases, catalyzing the reduction of toxic aldehydes (HMF, furfural, vanillin and acetaldehyde) into their corresponding, lesser toxic alcohols, and as such entailing in-situ detoxification which confers increased tolerance to these inhibitors. Finally, preliminary data suggests that ABP1P467S, ASG1S899Pand genes in S. cerevisiae obtained via homologous gene transfer (i.e. a GRE2-like aldehyde reductase and a putative drug transporter) might contribute to high furfural tolerance in yeast. The generation of a 2G yeast strain with high furan aldehyde tolerance, and high inhibitor tolerance in general would be highly relevant to increase fermentation performance of such a yeast strain. This would further enable cost-efficient production of 2G bioethanol, and other biofuels and bio-based chemicals from lignocellulose-based processes. As such, this would further contribute to shifting from a fossil-fueled based economy to a bio-based economy.

Publication year:2020

Accessibility:Open