Yeast complex trait genetics: beyond classical crosses – AcrossTrait

Setting up of the RTG protocol: we established two different protocols to isolate the RTG cells, (1) by selection of arginine prototrophs resulting from recombination between the heteroalleles, or (2) by dissection of unselected pairs of mother and daughter cells right after RTG division.

LOH detection: We expanded the RTG protocol to non laboratory strains. We used classical yeast genetics method to detect LOH in various cross combinations. These relied on the URA3 marker (loss of this gene can be select in 5-FOA media), ARG4 heteroalleles or loss of G418 resistance. We perform the RTG experiment in many replicas and in different cross combination to understand the timing and the frequency of the RTG process.

Detection of LOH from genome analysis: we developed a dedicated bioinformatic pipeline to analyze the genome dynamic in RTG hybrids. These pipelines rely on analysis of coverage, heterozygous SNPs and recombinant reads. A similar analysis was also applied to natural strains. We also establish efficient approaches to generate high quality de novo assemblies from PacBio genome sequence.

Phenotyping: we applied a wide range of phenotype to investigate the degree of phenotypic variation generated by the RTG process. The phenotype consist of monitoring growth in many environmental conditions as well as fermentation related phenotypes.

The results obtained so far provide an unprecedented view of the genome dynamic occurring during the RTG process and reveal how RTG can give rise to important applications in industrial settings for yeast strain improvement. Below we provide a brief summary and include a detailed report with extended methods as attachment. Our results indicate:

• RTG cells are bona fide diploid cells with 16 chromosomes pairs
• RTG cells have a broad frequency of recombinant regions, comprising 5 to 168 regions of LOH ranging from few bp to 785 kb per RTG strain.
• Reiteration of the RTG regime leads to additional recombination events
• RTG generate phenotypic variability and RTG cells can be used for the mapping of phenotypic traits
• RTG is not restricted to lab strain and can be efficiently induced in S. cerevisiae natural isolates
• LOH due to mitotic recombination can be observed in interspecies hybrids, mostly localized around the marker used for screening recombinant cells
• Our data indicated that RTG occur at lower frequencies in the interspecies hybrids, although these rare recombination events can have a remarkable impact on genome evolution
• We identified an intriguing example of a natural S. cerevisiae x S. paradoxus hybrid that is a pure diploid with an extremely high number of LOH events.

In the next 24 months of the grant, we will finalise the analysis describe above as well as move into new direction as indicated in our original proposal. We are currently refining the bioinformatic pipelines in order to be able to detect the short recombination events. Improvement of the genome assemblies of the strain used in the crosses will also allow mapping the recombination events in a more accurate way.
Furthermore, we will test how structural variation (SV) affects the recombination landscape during the RTG process. We have generated hybrid and RTG cells between using the Sc strain from the Malaysian lineage that harbor extensive SVs.
For the interspecies hybrids with industrial interest (Sc x Sk), we will phenotype enological relevant traits.
Finally, we are investigating natural hybrids for genomic signature of historical RTG. We have identified an intriguing example of a natural Sc x Sp hybrid that is a pure diploid with an extremely high number of LOH events. Spore viability and RTG analysis will be investigate in this strain and other natural hybrids that show RTG evidence.

Publications

1. Laureau R, Loeillet S, Salinas F, Bergström A, Liti G and Nicolas A. “Recombination of the yeast S. cerevisiae genome through meiotic reversion”. manuscript in preparation.
2. Marsit S. and Dequin S. 2015. Diversity and adaptive evolution of Saccharomyces wine yeast: a review. FEMS Yeast Res. In press.
3. Liti G. 2015. The fascinating, secret wild life of the budding yeast Saccharomyces cerevisiae. eLife. Mar 25; 4. doi: 10.7554/eLife.05835.
4. Liti G, Jonas Warringer and Anders Blomberg. 2015. Budding Yeast Strains and Genotype-Phenotype mapping. Book chapter in: Budding Yeast: A Laboratory Manual. Publisher: Cold Spring Harbor Protocols. In press.
5. Gibson B and Liti G. 2015. Saccharomyces pastorianus: genomic insights inspiring innovation for industry. Yeast. 32 (1),17-27.

Patent
Title: Yeast strain improvement method
Publication number: WO2014083142 A1
Publication date: Jun 5, 2014
Inventors: Alain Nicolas (Institut Curie), Gianni LITI (CNRS)
URL: www.google.com/patents/WO2014083142A1

Complex phenotypes are regulated by multiple interacting quantitative trait loci (QTLs). Dissection of the genetic mechanisms underlying the phenotypic variations remains a major conceptual and experimental challenge, due to the complex genetic architecture with many loci contributing to phenotypic effects, low penetrance, gene-gene and gene-environment interactions.
Over the past years, the budding yeast Saccharomyces cerevisiae has become an important model. This success is partially due to its intrinsic biological features, such as its short sexual generation time, high meiotic recombination rate, and small genome size. Precise reverse genetics technologies offer the unique opportunity to experimentally measure the phenotypic effect of genetic variants. Furthermore, intensive efforts have provided the genome sequence of numerous S. cerevisiae isolates and related species allowing powerful comparative genomics and new perspectives for functional and evolutionary studies. So far, nearly all of the yeast QTL mapping studies have used classical F1 meiotic segregants but novel strategies to produce powerful mapping population is needed. This proposal provides an innovative approach to link complex yeast phenotypes with high-resolution mapping of the genetics factors. We will implement an innovative strategy, named return to growth (RTG) that is rapid and does not require classical crosses to generate a large mapping population. This method takes advantage of the natural capacity of S. cerevisiae, to reverse its meiotic progression and yield diploid cells. Most importantly, our preliminary data, based on Next Generation Sequencing (NGS) of a yeast polymorphic strain, demonstrate that the process of RTG gives rise to recombined diploid cells. The repair of the meiotic DSBs upon gene conversion or crossover resolution leads to the maintenance or loss of heterozygosity (LOH) in various ratios and locations. Cells derived from RTG are all genetically different, a resource for quantitative trait analyses and mapping. Our preliminary phenotypic analysis of few multi-factorial traits (colony morphology, sporulation efficiency, growth at high temperature) revealed quantitative variation, providing a proof of concept for using the RTG process in QTL mapping. To optimize and utilize this approach for laboratory and industrial yeast strains, we will collaborate to: (i) isolate, NGS and bioinformatically determine the allelic genotype of 96 RTG diploids of the S288c/SK1 hybrid strain as well as 96 strains derived from spores, (ii) perform high throughput phenotypic analyses of recombined strains for multiple traits, map and validate the QTLs and, (iii) develop a genetic system that allows one to follow the progress of the meiotic cycle and the frequency of LOH in a large variety of strains. We will generate several sets of diploid recombined strains for identifying QTLs of medical and biotechnological relevance. Importantly, we will test if RTG is able to produce recombinant hybrids between different reproductively isolated Saccharomyces species. These experiments will allow us to understand the contribution of RTG to genome structure evolution, where inter-specific introgressions are prevalent in the Saccharomyces species. We envisage that the RTG process is likely to have profound implications in terms of genome evolution, since yeasts have to survive in fluctuating environments where they must also undergo rapid diversification. This scenario offers an attractive alternative to the full meiotic cycle where sequence divergence and chromosomal rearrangements can impair the process, resulting in poor gamete viability and therefore reducing the individual fitness. This ambitious project will bring a new dimension to QTL mapping, strengthen budding yeast as a model organism for quantitative genetics and provide a GMO-free innovation to improve the performance of industrial yeast strains.