Hugo Devillers

Metschnikowia pulcherrima , an ascomycetous yeast, has been frequently isolated from various substrates associated with agriculture such as nutrient‐rich plant materials (fruit, fruit juices and fermenting grape must), soil and with various insects including fruit flies. Due to the growing consumer demand for wines with reduced levels of additives such as sulphites, there has been great interest in this yeast species for its ability to compete with other fungi. Despite having similar phenotypes, many species related to M. pulcherrima have been described in the last 20 years based on the sequences of the LSU D1/D2 domain. These species were grouped together as M. pulcherrima based on primary and secondary barcode sequences, physiological tests and hybridization experiments (Sipiczki 2022a). However, this new classification is supported by limited genetic information. In this project we performed a genomic characterization of 63 strains, including 10 type strains of current synonyms. The combination of different genomic‐scale analyses (ANI, k‐mer, Busco genes, whole‐genome genotyping) conclusively demonstrated the existence of a main cluster of strains representing the M. pulcherrima species and a small sister cluster related to M. chrysoperlae that appeared as potential triploid hybrids. We also observed a high heterozygosity in M. pulcherrima which was significantly higher than S. cerevisiae and the presence of some homozygous strains. Strong linkage disequilibrium decay also suggests active outcrossing leading to recombination in this species. This demonstrates the power of phylogenomic in resolving taxonomic issues and the ability of such approaches to provide information on the life cycle of this species.

The wine industry is constrained to adapt to new challenges to cope with new climatic and societal contexts. For this, the use of selected yeast strains, mainly from the species Saccharomyces cerevisiae, is a powerful lever, with UV mutagenesis used successfully to generate genetic diversity. From a library of 502 mutants, phenotyped in wine conditions, we identified a bimodal distribution in the production of certain primary metabolites. Here, we focus on the study of the two subpopulations, divided into low or high succinate producers, under wine-like conditions. We observed marked differences in primary metabolite profiles, particularly in the production patterns of tricarboxylic acid cycle intermediates. These patterns suggest that both TCA branches are impacted and point to a succession of pathways, with pyruvate and α-ketoglutarate produced early, and then converted into succinate during the stationary growth phase. Genomic analysis revealed that the tested low succinate producers exhibited a complete absence of the mitochondrial genome, indicating the absence of the respiratory chain. However, this finding does not exclude the possibility that low succinate production may be caused by other genetic factors. Although yeasts largely use glycolysis followed by ethanol synthesis during fermentation, our results also highlighted a crucial role of mitochondria in modulating the production of compounds of great relevance in winemaking, especially organic acids. The main hypothesis is that a perturbation of the redox balance, potentially caused by a reduced fumarase activity, jeopardised cofactors re-oxidation, disturbing the activity of tricarboxylic acid cycle branches

Maudiozyma humilis is the second most frequently encountered yeast species in sourdough bread. Despite its ecological and food relevance, little is known about its evolutionary trajectories and phenotype traits of interest. Here we investigated the genomic and phenotypic diversity of a world-wide collection of 55 M. humilis strains by combining genomic analyses and high-throughput phenotyping of fermentation kinetics and fitness. Population genomic analysis revealed six distinct clades, three diploid and three triploid, with no geographical or substrate-specific structuring. Phylogenetic, loss of heterozygosity (LOH) and allele specific analyses indicated that triploid strains originated from both recent and more ancient hybridization events involving multiple lineages. The absence of the HO gene and mating-type silent cassettes revealed the inability of M. humilis to carry mating-type switching. In addition, high linkage disequilibrium and variable LOH accumulation were observed consistent with a predominantly clonal reproduction. Last, an exceptionally high level of heterozygosity was detected, suggesting that occasional hybridization is the major driver of genetic diversity. Phenotypic characterization in a synthetic sourdough medium revealed variation in fermentation kinetics and fitness statistically associated with the genetic clades. Interestingly, genetic distance between clades and strains better explain the phenotypic variation than difference in ploidy. Altogether, our findings highlight the complex evolutionary history of M. humilis, shaped by hybridization and ploidy variation and reveal that historical contingency shapes the phenotypic landscape of this species. Beyond providing the first analysis of M. humilis evolution, our result challenges the hypothesis that increases in ploidy are necessarily beneficial in domesticated species.

La génomique comparative est l'étude des relations structurales et fonctionnelles entre des génomes appartenant à différentes souches ou espèces. Cette discipline offre ainsi la possibilité d'étudier et de comprendre les processus qui façonnent les génomes au cours de l'évolution. Dans le cadre de cette thèse, nous nous sommes intéressés à la génomique comparative des bactéries et plus particulièrement aux méthodes relatives à la comparaison des séquences complètes d'ADN des génomes bactériens. Ces dix dernières années, le développement d'outils informatiques permettant de comparer des génomes entiers à l'échelle de l'ADN est devenu une thématique de recherche à part entière. Actuellement, il existe de nombreux outils dédiés à cette tâche. Cependant, jusqu'à présent, la plupart des efforts ont été dirigés vers la réduction du temps de calcul et l'optimisation de la mémoire au détriment de l'évaluation de la qualité des résultats obtenus. Pour combler ce vide, nous avons travaillé sur différents problèmes statistiques soulevés par la comparaison de génomes complets bactériens. Notre travail se divise en deux axes de recherche. Dans un premier temps, nous nous sommes employés à évaluer la robustesse des alignements de génomes complets bactériens. Nous avons proposé une méthode originale fondée sur l'application de perturbations aléatoires sur les génomes comparés. Trois scores différents sont alors calculés pour estimer la robustesse des alignements de génomes à différentes échelles, allant des nucléotides aux séquences entières des génomes. Notre méthode a été expérimentée sur des données génomiques bactériennes réelles. Nos scores permettent d'identifier à la fois les alignements robustes et non robustes. Ils peuvent être employés pour corriger un alignement ou encore pour comparer plusieurs alignements obtenus à partir de différents outils. Dans un second temps, nous avons étudié le problème de la paramétrisation des outils de comparaisons de génomes entiers. En effet, la plupart des outils existants manquent à la fois de documentation et de valeurs par défaut fiables pour initialiser leurs paramètres. Conséquemment, il y a un besoin crucial de méthodes spécifiques pour aider les utilisateurs à définir des valeurs appropriées pour les paramètres de ces outils. Une grande partie des outils de comparaisons de génomes complets est fondée sur la détection des matches (mots communs exacts). Le paramètre essentiel pour ces méthodes est la longueur des matches à considérer. Au cours de cette thèse, nous avons développé deux méthodes statistiques pour estimer une valeur optimale pour la taille des matches. Notre première approche utilise un modèle de mélange de lois géométriques pour caractériser la distribution de la taille des matches obtenus lorsque l'on compare deux séquences génomiques. La deuxième approche est fondée sur une approximation de Poisson de la loi du comptage des matches entre deux chaînes de Markov. Ces méthodes statistiques nous permettent d'identifier facilement une taille optimale de matches à la fois pour des séquences simulées et pour des données génomiques réelles. Nous avons également montré que cette taille optimale dépend des caractéristiques des génomes comparés telles que leur taille, leur composition en base ou leur divergence relative. Cette thèse représente une des toutes premières études dont l'objectif est d'évaluer et d'améliorer la qualité des comparaisons des génomes complets. L'intérêt et les limites de nos différentes approches sont discutés et plusieurs perspectives d'évolution sont proposées.

Maudiozyma humilis is the second most frequently encountered yeast species in sourdough bread. Despite its ecological and food relevance, little is known about its evolutionary trajectories and phenotype traits of interest. Here we investigated the genomic and phenotypic diversity of a world-wide collection of 55 M. humilis strains, including 52 from sourdough, by combining genomic analysis, flow cytometry and high-throughput phenotyping of fermentation kinetics and fitness. Population genomic analysis revealed six genetically distinct clades, three diploid and three triploid, with no geographical or substrate-specific structuring. Phylogenetic, loss of heterozygosity (LOH) distribution and allele specific analysis indicated that triploid strains originated from both recent and more ancient hybridization events involving multiple diploid lineages. The absence of the HO gene, and mating-type silent cassettes, revealed that M. humilis is not able to carry mating-type switching. In addition, high linkage disequilibrium (LD) and variable LOH accumulation were observed consistent with a predominantly clonal reproduction. Last, an exceptionally high level of heterozygosity was detected, suggesting that occasional hybridization is the major driver of genetic diversity. Phenotypic characterization in a synthetic sourdough medium revealed variation in fermentation kinetics and fitness statistically associated with the genetic clades. Interestingly, genetic distance between clades and strains better explain the phenotypic variation than difference in ploidy. Altogether, our findings highlight the complex evolutionary history of M. humilis , shaped by hybridization and ploidy variation and reveal that historical contingency, more than ploidy, shapes the phenotypic landscape of this species. Beyond providing the first analysis of M. humilis evolution, our result challenges the hypothesis that increases in ploidy are necessarily beneficial in domesticated species.

Over the last decades, the yeast Saccharomyces cerevisiae has emerged as a key model for studying microbial domestication, particularly in the context of winemaking. The recent surge of population genomic data, encompassing thousands of genomes, has profoundly reshaped our understanding of the evolutionary history and adaptive potential of wine yeasts. Despite arising from a domestication bottleneck, wine yeasts form a well-structured population and display striking genome dynamism. Extensive variation in heterozygosity, aneuploidy, structural variations, and gene content may provide a reservoir of genomic variation that contributes to adaptive potential in the harsh winemaking environment. While some genetic variation and genome restructuring contribute to adaptation, gene flow through hybridization, introgression and especially horizontal gene transfer has emerged as a major driver of functional innovation. These findings establish wine yeasts as a powerful model to link genome evolution with adaptation to anthropogenic environments. They also provide a foundation for the rational improvement of industrial strains through approaches such as quantitative trait locus mapping, adaptive laboratory evolution and genome-wide association studies. Extending these frameworks to non-Saccharomyces species and integrating genomic, functional and ecological data will be key to understanding and engineering microbial communities, to face the modern winemaking challenges.

Food microorganisms have been employed for centuries for the processing of fermented foods, leading to adapted populations with phenotypic traits of interest. The yeast Monosporozyma unispora (formerly Kazachstania unispora) has been identified in a wide range of fermented foods and beverages. Here, we studied the genetic and phenotypic diversity of a collection of 53 strains primarily derived from cheese, kefir, and sourdough. The 12.7-Mb genome of the type strain CLIB 234T was sequenced and assembled into near-complete chromosomes and annotated at the structural and functional levels, with 5639 coding sequences predicted. Comparison of the pangenome and core genome revealed minimal differences. From the complete yeast collection, we gathered genetic data (diversity, phylogeny, and population structure) and phenotypic data (growth capacity on solid media). Population genomic analyses revealed a low level of nucleotide diversity and strong population structure, with the presence of two major clades corresponding to ecological origins (cheese and kefir vs. plant derivatives). A high prevalence of extensive loss of heterozygosity and a slow linkage disequilibrium decay suggested a predominantly clonal mode of reproduction. Phenotypic analyses revealed growth variation under stress conditions, including high salinity and low pH, but no definitive link between phenotypic traits and environmental adaptation was established.

Repellents play a fundamental role in vector control and prevention to keep mosquitoes away from humans. Available in limited numbers, it is absolutely necessary to find new repellents for preventing problems of resistance. QSAR (Quantitative Structure–Activity Relationship) methods are particularly suited for designing molecules with potential repellent activity. These models require that the molecules be described by physicochemical properties, topological indices, and/or structural indicators. In the former situation, QSPR (Quantitative Structure–Property Relationship) models are used for calculating physicochemical descriptors. Use of different QSPR models for the same property can lead to different values for the same molecule. In this context, the influence of the 1-octanol/water partition coefficient (log P) calculated according to two different methodologies was statistically evaluated in the modeling of 2171 molecules for which their skin repellent activity against Aedes aegypti was available. The two series of supervised artificial neural networks differed only by their input neuron coding for log P. Although both categories of classification models led to overall good statistics, we clearly showed that differences in log P values calculated for a molecule could result in very different prediction results. This was especially true for repellents. The practical implication of these differences was discussed.

Finding new marketable mosquito repellents is a complex and time-consuming process that can be optimized via modelling. In this context, a SAR (Structure–Activity Relationship) model was designed from a set of 2171 molecules whose actual repellent activity against Aedes aegypti was available. Information-rich descriptors were used as input neurons of a three-layer perceptron (TLP) to compute the models. The most interesting classification model was a 20/6/2 TLP showing 94% and 89% accuracy on the training set and test set, respectively. A total of 57 other artificial neural network models based on the same architecture were also computed. This allowed us to consider all chemicals both as training and test set members in order to better interpret the results obtained with the selected model. Most of the wrong predictions were explainable. The 20/6/2 TLP model was then used for predicting the potential repellent activity of new molecules. Among them, two were successfully evaluated in vivo.

Although Yarrowia lipolytica is a model yeast for the study of lipid metabolism, its diversity is poorly known, as studies generally consider only a few standard laboratory strains. To extend our knowledge of this biotechnological workhorse, we investigated the genomic and phenotypic diversity of 56 natural isolates. Y. lipolytica is classified into five clades with no correlation between clade membership and geographic or ecological origin. A low genetic diversity (π = 0.0017) and a pan-genome (6528 genes) barely different from the core genome (6315 genes) suggest Y. lipolytica is a recently evolving species. Large segmental duplications were detected, totaling 892 genes. With three new LTR-retrotransposons of the Gypsy family (Tyl4, Tyl9, and Tyl10), the transposable element content of genomes appeared diversified but still low (from 0.36% to 3.62%). We quantified 34 traits with substantial phenotypic diversity, but genome-wide association studies failed to evidence any associations. Instead, we investigated known genes and found four mutational events leading to XPR2 protease inactivation. Regarding lipid metabolism, most high-impact mutations were found in family-belonging genes, such as ALK or LIP, and therefore had a low phenotypic impact, suggesting that the huge diversity of lipid synthesis and accumulation is multifactorial or due to complex regulations.