Understanding early yeast lipid consumption in wine fermentation through lipidomics: new perspectives for fermentation management

Abstract

Lipids play a pivotal role in yeast stress resistance, fermentation kinetics, and aroma production. However, while nitrogen nutrition is well characterized, the consumption of the diverse lipid fraction present in grape musts during fermentation remains poorly understood. This study applies a comprehensive targeted lipidomics approach, combining GC-MS and LC-MS/MS, to investigate the consumption kinetics of 94 lipid species in two grape musts (Chardonnay and Gewurztraminer) by three commercial yeast strains. We demonstrate that yeasts consume a much broader range of exogenous lipids than previously reported. Significant concentration changes were observed for 66 lipids during fermentation. In addition to the previously reported consumption of Free Fatty Acids (FFAs) and phytosterols, we demonstrate that yeasts also actively consume Sterol Esters (SE) and Diglycerides (DG). Adsorption tests confirmed that these decreases during fermentation are primarily driven by metabolic consumption rather than cell wall binding, although minor phytosterol adsorption was observed. Lipid uptake occurred principally during the yeast exponential growth phase. We quantified functional requirements for specific consumption: phytosterol uptake (4.5 ± 1.6 mg/L/10⁸ cells) and DG consumption were constant across strains and musts. Conversely, FFA and SE requirements were matrix-dependent. Notably, strong linear correlations were found between initial concentrations and specific consumption for FFAs (R2 = 0.91), phytosterols (R2 = 0.81), SE (R2 = 0.94), and DG (R2 = 0.93). This indicates a dose-response effect where uptake is largely proportional to initial abundance. These results highlight that yeast lipid demand is significantly greater than previously thought. The established quantitative data and linear models offer new opportunities for optimizing nutrient supplementation and developing predictive models for precision oenology.