Selection of highly stress-tolerant yeast strains relevant for bioethanol fermentation using predictive growth models.

INTRODUCTION

The growing demand for renewable energy together with the environmental impact of fossil fuels, have intensified global interest in sustainable bioethanol production. Saccharomyces cerevisiae is the preferred microorganism for industrial fermentation due to its productivity and stress tolerance, but cumulative stress during successive cycles reduces process efficiency. Therefore, selecting stress-tolerant strains capable of adapting to fluctuating conditions is crucial. Predictive microbiology (PM), which applies mathematical models to predict and quantify microbial responses to environmental factors, remains a valuable but still limited approach in yeast-based bioethanol production.

METHODS

In this study, diverse PM models were applied to evaluate the effects of temperature, pH, sucrose, and ethanol concentrations on S. cerevisiae strains isolated from industrial fermentations in Tucumán, Argentina. Thirteen native isolates were compared with a commercial reference strain (Calsa).

RESULTS

The primary and secondary models used achieved excellent fits in all cases (R 2 > 0.9), effectively describing and anticipating growth responses under stress conditions relevant to industrial fermentations of the evaluated strains, which displayed wide tolerance ranges suggesting potential suitability for cell recycling and high-density fermentation, pending validation under fermentation conditions. Strains T415, Le384, LF84, and SR350 emerged as promising candidates for further validation in industrial bioethanol fermentations to improve the stability and sustainability of industrial bioethanol production, matching or outperforming the control strain.

DISCUSSION: Overall, this study underscores the usefulness of PM tools for characterizing and selecting native yeast strains with enhanced stress tolerance in local bioethanol production, as they allow for anticipation of physiological behavior in growth-based assays that mimic key industrial stresses, reduction of experimental workload, and strengthen strain selection and evaluation criteria.