Non-Saccharomyces yeasts for fermentation of lignocellulosic biomass: tolerance to inhibitors and fermentation to ethanol

Lignocellulosic biomass has received increasing attention as carbon source for microbial fermentations. Pretreatment of the biomass is needed to liberate sugars with intense treatments yielding higher sugar concentrations (which are economically preferred). However, these intense treatments also result in the formation of several undesired compounds in concentrations which may reduce the fermentation efficiency. The main inhibitors are solutes (osmostress), product stress (e.g. ethanol, lactic acid, dicarboxylic acid,...), furans (5-hydroxymethylfurfural, furfural), weak acids (acetic acid, levulinic acid, formic acid) and phenolic and aromatic compounds originating from the lignin fraction (e.g. vanillin). Microorganisms are therefore exposed to a new, challenging and diverse fermentation medium due to the large diversity of lignocellulosic biomass sources and hydrolysis conditions. It is therefore required to align the hydrolysate with a microorganism with the best characteristics in terms of tolerance to inhibitors and sugar consumption profile. In this study, we screened more than 50 yeast isolates isolated from plant nectar and beet sugar thick juice for tolerance to weak acids, furans, osmostress and ethanol and compared it with an industrially used S. cerevisiae strain and natural strain (isolated from oak). Non-Saccharomyces yeasts showed better osmotolerance than S. cerevisiae. The species Candida bombi, Wickerhamomyces anomalus and Torulaspora delbrueckii also showed better hydroxymethylfurfural tolerance than S. cerevisiae. T. delbrueckii and Pichia kudriavzevii showed better furfural tolerance. Growth on 2% xylose as sole carbon source was better for non-Saccharomyces yeasts, with W. anomalus reaching more than 20% growth relative to 2% glucose. As a case study, we performed bioethanol fermentations mimicking high gravity fermentations (25% glucose) and artificial lignocellulose hydrolysates (with a myriad of inhibitors) using 5 selected isolates. Interestingly, among two tested S. cerevisiae strains, the natural strain isolated from an oak tree performed better than Ethanol Red, a S. cerevisiae strain which is currently commonly used in industrial bioethanol fermentations. Additionally, a W. anomalus strain isolated from sugar beet thick juice was found to have a comparable ethanol yield, but needed longer fermentation time. No xylose was fermented to ethanol. Other non-Saccharomyces yeasts yielded lower ethanol amounts.status: publishe