Phylogenomics and functional annotation of 530 non- Saccharomyces yeasts from winemaking environments reveals their fermentome and flavorome

Abstract

The winemaking industry faces unprecedented challenges due to climate change and market shifts, with profound commercial and socioeconomic repercussions. In response, non-Saccharomyces yeasts have gained attention for their potential to both mitigate these challenges and enhance the complexity of winemaking. This study builds upon our previous cataloguing of 293 non-Saccharomyces yeast species associated with winemaking environments by rigorously analysing 661 publicly available genomes. By employing a bioinformatics pipeline with stringent quality control checkpoints, we annotated and evaluated these genomes, culminating in a robust dataset of 530 non-Saccharomyces proteomes, belonging to 134 species, accessible to the research community. Employing this dataset, we conducted a comparative phylogenomic analysis to decipher metabolic networks related to fermentation capacity and flavor/aroma modulation. Our functional annotation has uncovered distinctive metabolic traits of non-Saccharomyces yeasts, elucidating their unique contributions to enology. Crucially, this work pioneers the identification of a non-Saccharomyces 'fermentome', a specific set of six genes uniquely present in fermentative species and absent in non-fermentative ones, and an expanded set of 35 genes constituting the complete fermentome. Moreover, we delineated a 'flavorome' by examining 96 genes across 19 metabolic categories implicated in wine aroma and flavour enhancement. These discoveries provide valuable genomic insights, offering new avenues for innovative winemaking practices and research. Citation: Franco-Duarte R, Fernandes T, Sousa MJ, Sampaio P, Rito T, Soares P (2025). Phylogenomics and functional annotation of 530 non-Saccharomyces yeasts from winemaking environments reveals their fermentome and flavorome. Studies in Mycology 111: 1-17. doi: 10.3114/sim.2025.111.01.

Keywords: bioinformatics; fermentation; fungi; genomics; non-conventional yeasts; phylogeny.

Fig. 1

Bioinformatic pipeline used to analyze the 661 genomes of non-Saccharomyces yeasts. The schematic representation of the pipeline includes all the quality control checkpoints used throughout the work, together with the software and the intermediate datasets obtained. For details about each software and parameters, see the Methods section.

Fig. 2

Detailed phylogeny of the 530 genomes belonging to 293 non-Saccharomyces yeast species. Phylogeny was obtained after the alignment of 1209 core proteins, and maximum likelihood analysis with the JTT model of amino acid evolution and gamma-distributed rates (four rates) with 10 000 bootstrap replicates. Bootstrap values were omitted from the branches due to being > 99 %. Rhizophagus irregularis was used as outgroup to root the tree. Branches were coloured and annotated according to the taxonomic classification available in Supplementary Table S2, and the remaining annotations/colourings were done using results of genomes annotation (see methods for details).

Fig. 3

PCA visualization of strains distribution, considering the number of genes associated to each metabolic KEGG category (Supplementary Tables S4 and S5), and coloured according to taxonomic level. Green – Phylum Ascomycota; Blue – Phylum Basidiomycota. Symbols represent different orders according with the legend. PC1 – 18.9 %; PC2 – 6.6 %.

Fig. 4

Cumulative percentage of proteins annotated to each metabolic KEGG category (detailed results in Tables S4 and S5), divided by taxonomic level (from left to right: Phyla, Class, Order, Family).

Fig. 5

Analysis of the flavorome of non-Saccharomyces yeasts, in particular emphasizing the presence and absence of the 96 genes related with aroma and flavour production during wine fermentation. A. Heatmap displaying the adjusted p-values from the Kruskal–Wallis one-way analysis of variance, which assesses the presence/absence of genes within each taxonomic group. Significant p-values are indicated with colour coding: < 0.05 in orange, and < 0.01 in pink. B. Average percentage of strains within each subphylum that encode genes from the indicated categories: AcEs – “Acetate esters - alcohol acetyl transferases”; pA – “p-ABA synthesis - synthesis of p-ABA from chorismite”; VOA – “Volatile organic acids - acetic acid”; AaaS – “aromatic amino acid synthesis - synthesis of chorismate, phenyalanine, tryptophan and tyrosine”; C. Percentage of strains within each family that possess genes ABZ1 and ABZ2; D. Percentage of strains within each family that possess genes AAD3, AAD4, AAD10, AAD14, AAD15 and AAD16; E Percentage of strains within each family that possess genes SLI1, ATF1 and ATF2; F Scatter plot illustrating the distribution of each of the 96 genes in fermentative versus non-fermentative strains, with individual genes coloured based on overarching categories as listed in Table 2. For panels (C), (D) and (E), families represented by fewer than 4 strains were not included in the analysis.