Pre-senescence harvest of switchgrass inhibits xylose utilization by engineered yeast

Document Type


Publication Date



© 2018 Ong, Shinde, da Costa Sousa and Sanford. Proper timing of switchgrass harvest for bioenergy is important to maximize yield and optimize end use conversion. Proposed windows range from peak biomass to the following spring after overwintering in the field. There are various pros and cons associated with harvest timing: earlier harvests maximize yield but can remove nutrients from the field that may require replacement, while later harvests have reduced biomass yields due to weathering but maximize nutrient resorption in belowground tissues. Switchgrass composition changes during the harvest period, with losses of potential fermentation nutrients (amino acids and minerals), and sources of pretreatment-derived inhibitors (soluble sugars), which could affect downstream conversion by microorganisms. For this work we investigated whether switchgrass harvest could be timed to maximize beneficial impacts on fermentation. Switchgrass samples were harvested from five replicate field plots in Wisconsin, roughly every 2-3 weeks from peak biomass (Aug. 20) until after the killing frost (Nov. 7). Cell wall composition showed little consistent variation with harvest date while bulk biomass analysis showed a relative increase in cell wall content (lignin and structural sugars) and loss of extractives (minerals, protein, soluble sugars, and others). Following high or low severity AFEX pretreatment and high solids enzymatic hydrolysis (6% glucan loading), two field replicates were fermented using Saccharomyces cerevisiae 424A, a strain engineered to utilize xylose in addition to glucose. For both pretreatment severities, S. cerevisiae 424A grown in hydrolysates from the three earlier harvests utilized only a small fraction of available xylose, while almost complete utilization occurred within 96 hr for the last three harvest dates. Detailed analysis of the hydrolysate low molecular weight aromatics did not indicate any compounds potentially responsible for the inhibition, with most of the observed variation in their concentration due to pretreatment severity. Amino acid composition also did not appear to be limiting. Current indications point to a plant-generated compound that degrades during senescence, which future work will attempt to identify. Ultimately this work demonstrates that, although an attractive option to maximize yield, harvesting switchgrass before it begins senescing could have a negative effect on downstream conversion processes.

Publication Title

Frontiers in Energy Research