Differential Gene Expression and Allele Frequency Changes Favour Adaptation of a Heterogeneous Yeast Population to Nitrogen-Limited Fermentations.

Eduardo I Kessi-Pérez,Jennifer Molinet,David Figueroa,Alvaro Díaz-Barrera,Camila Bastías,Jing Li,Belén Ponce,Camila Baeza,Gianni Liti,Marco Gaete,Francisco Salinas,Claudio Martínez

doi:10.3389/fmicb.2020.01204

Abstract

Alcoholic fermentation is fundamentally an adaptation process, in which the yeast Saccharomyces cerevisiae outperforms its competitors and takes over the fermentation process itself. Although wine yeast strains appear to be adapted to the stressful conditions of alcoholic fermentation, nitrogen limitations in grape must cause stuck or slow fermentations, generating significant economic losses for the wine industry. One way to discover the genetic bases that promote yeast adaptation to nitrogen-deficient environments are selection experiments, where a yeast population undergoes selection under conditions of nitrogen restriction for a number of generations, to then identify by sequencing the molecular characteristics that promote this adaptation. In this work, we carried out selection experiments in bioreactors imitating wine fermentation under nitrogen-limited fermentation conditions (SM60), using the heterogeneous SGRP-4X yeast population, to then sequence the transcriptome and the genome of the population at different time points of the selection process. The transcriptomic results showed an overexpression of genes from the NA strain (North American/YPS128), a wild, non-domesticated isolate. In addition, genome sequencing and allele frequency results allowed several QTLs to be mapped for adaptation to nitrogen-limited fermentation. Finally, we validated the ECM38 allele of NA strain as responsible for higher growth efficiency under nitrogen-limited conditions. Taken together, our results revealed a complex pattern of molecular signatures favouring adaptation of the yeast population to nitrogen-limited fermentations, including differential gene expression, allele frequency changes and loss of the mitochondrial genome. Finally, the results suggest that wild alleles from a non-domesticated isolate (NA) may have a relevant role in the adaptation to the assayed fermentation conditions, with the consequent potential of these alleles for the genetic improvement of wine yeast strains.

Highlights

The yeast Saccharomyces cerevisiae is one of the most intensively studied microorganisms and a biological model with diverse biotechnological applications, being the first eukaryote to be completely sequenced more than 20 years ago (Goffeau et al, 1996)
The results showed a higher specific growth rate for the North American (NA) strain in SM60 compared to the Wine European (WE) strain, and no differences in SM300 (Table 1 and Supplementary Figure 1)
To corroborate that fermentative conditions were achieved, different fermentation metabolites were analysed by HPLC at three different time points of a continuous culture fermentation carrying a mixture of NA and WE strains, with special interest in confirming that proline was not consumed as evidence of anaerobic metabolism (Table 2)

Summary

Introduction

The yeast Saccharomyces cerevisiae is one of the most intensively studied microorganisms and a biological model with diverse biotechnological applications, being the first eukaryote to be completely sequenced more than 20 years ago (Goffeau et al, 1996). Despite MA strain, which is reproductively isolated from the other strains (Cubillos et al, 2009), the other four representative strains have become a powerful resource for studying the genetic basis of S. cerevisiae natural variation (Cubillos et al, 2011; Liti and Louis, 2012; Villalobos-Cid et al, 2019) These strains have been extensively used as founder (parental) strains for recombinant yeast populations, which have been used in genetic studies addressing the causative alleles of phenotypic variation through various approaches, such as quantitative trait loci (QTL) mapping, selection experiments and RNAseq (Parts et al, 2011; Kessi-Perez et al, 2016; Cubillos et al, 2017; Quispe et al, 2017)

Methods

Results

Conclusion