(732g) Understanding the Metabolic Shift of Scheffersomyces Stipitis from Aerobic Growth to Oxygen-Limited Fermentation at Genome-Scale

Matthew Hilliard, Thomas Jeffries, Q. Peter He and Jin Wang

Xylose fermentation is essential in making lignocellulosic biofuel processes economically feasible. Scheffersomyces stipitis is a strain with the highest native capability of converting xylose to ethanol by fermentation. Because s. stipitis is a Crabtree negative yeast strain, oxygen availability is one of the key factors that affect ethanol production due to redox (im)balance in its xylose assimilation pathways. Many studies have examined the effects of oxygenation conditions on ethanol yield and productivity. As most of them were conducted under batch conditions without accurate control of oxygen supply, different conclusions have been drawn regarding the optimal oxygenation condition for ethanol production. In addition, although the cellular transcriptomic profiles of s. stipitis were compared for conditions under aerobic growth and oxygen-limited fermentation, up to now, none of the studies have examined the transcriptomic shifts at genome-scale during the transition stage from aerobic growth to oxygen-limited fermentation.

In order to gain fundamental understandings on the metabolic shift and regulatory mechanism of s. stipitis during the transition from aerobic growth to oxygen-limited fermentation, transcriptomic profiles of time-series samples collected during the transition period were obtained through RNAseq at a sampling interval of 10 minutes. Specifically, 4 batches of experiments were conducted under controlled chemostat condition to achieve aerobic growth and 2 samples were taken for the aerobic growth condition; then, oxygen supply was reduced to achieve a predetermined carbon-to-oxygen ratio (C/O) that optimizes ethanol production. At the same time, dilution rate was adjusted correspondingly to maintain constant cell density within the bioreactor. Once the oxygen supply was reduced, samples were taken every 10 minutes, with the aim to capture the metabolic shift during the transition via transcriptomic profiling. In addition, xylose, acetic acid, and xylitol concentration, as well as biomass density, oxygen consumption, and CO2 production were measured for each sampling point.

To the best of our knowledge, the obtained transcriptomic data set is the very first time-series data set that captures the transition process of s. stipitis from aerobic growth to oxygen-limited condition. Such a data set could provide important understanding on how cells adapt to oxygen-limited condition, and offer key insights on how to design mutants with improved ethanol production performance. Common bioinformatics tools^1,2 will be applied to analyze the data, but more importantly, as cellular metabolism shares many similarities with chemical plants, this work will aim to adapt some of the process engineering tools^3,4, developed for monitoring complex dynamic chemical plants, to analyze the data set. Through these analyses, we expect to identify the key pathways that are up- or down-regulated during the transition, as well as identify pathways that are co-regulated. Such information will be used to generate hypotheses on gene-regulatory mechanisms. Our preliminary results indicate that during the transition, cells induce heat shock response although the temperature was maintained as a constant. In addition, certain genes show initial up-regulation, and subsequently, down-regulation. This behavior suggests that a dynamic model is necessary to capture the cellular metabolic shift during the transition.

[1] Li, B., Dewey, C. N., RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome, BMC Bioinformatics, (2011).

[2] Robinson, M. D., McCarthy, D. J., Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data, Bioinformatics (2010).

[3] G.H. Dunteman, Principal component analysis (Quantitative applications in the social sciences), (1989).

[4] Hua, S., Sun, Z., Support vector machine approach for protein subcellular localization prediction, Bioinformatics, (2001).