Yeast Strain Construction Using CRISPR/Cas9: A Tool for Research and Strain Optimisation

Abstract

Saccharomyces cerevisiae is well known as the workhorse of many fermented beverage industries underpinned by its strong capability to create alcohol and sensory metabolites through the fermentation of sugars. Whilst S. cerevisiae is generally reliable, there is a drive to improve characteristics of available strains. The production of new and improved strains has been reliant on time-consuming techniques considered non-recombinant, in order to avoid their designation as genetically modified organisms (GMOs). While genetic engineering presents a more targeted and reliable approach, in many jurisdictions the use of such techniques is restricted. The recent emergence of CRISPR/Cas9, however, may change this, with some countries classifying CRISPR-edited strains as non-GMO when used in certain circumstances. This study investigated the use of CRISPR/Cas9 to generate improved strains of the commercial Saccharomyces wine strain, Lalvin EC1118. The effects of previously reported gene disruptions, as well as novel QTLs and SNPs derived from other yeast backgrounds were used as case studies. This study also evaluated the potential of generating gene disruptions via the ‘error prone’ non-homologous end joining (NHEJ) repair pathway. First, CRISPR/Cas9 was used to produce a disruption mutation in ECM33, by introducing a premature stop codon (Gly61stop). Previously, deletion of the ECM33 open reading frame lead to decreased fermentation duration in a haploid wine yeast derivative. In the present study this mutation disrupted gene function and improved fermentation performance as before. However, homozygous disruption in EC1118, lead to the discovery of another novel phenotype, cellular aggregation, which was masked by the typical flocculation of previously used haploid yeasts. Second, novel QTLs and SNPs were introduced, a method considered non-GMO in the United States and Japan. Two different phenotypes were chosen, slower growth associated with a loss of function QTL in SER1 derived from a sake yeast (Ser1p G78R), and proline accumulation by mutations identified from a chemical mutagenesis screening of baking yeast (Pro1p I150T, P247S and E415K). In co-inoculated fermentations, the slowed growth phenotype allowed for non-Saccharomyces yeasts (Metschnikowia pulcherrima or Lachancea thermotolerans) to begin fermentation, with the slow growing S. cerevisiae eventually completing fermentation. An additional use of a slow growing S. cerevisiae was identified, i.e., sequestering sulfite, enabling the growth of SO2 sensitive Lachancea thermotolerans in mixed culture in sulfured juice. Proline accumulation in yeast is associated with resistance to baking related stresses but the effects in wine yeast are unknown. EC1118 with these mutations similarly accumulated more proline, although there was no obvious improved tolerance towards ethanol or SO2, with minor resistance to osmotic stress observed. Finally, NHEJ was investigated with a view to developing non-GMO mutants relevant for Australian regulations. Targeting of the CAN1 gene (encoding arginine permease) to reduce urea production resulted in two different frame-shift mutants. In synthetic grape must with arginine as sole nitrogen source, the mutants grew slower and reduced urea as expected. However, in Chardonnay juice, the mutants behaved as the unmodified EC1118. This research demonstrates the applicability of CRISPR/Cas9 to industrial yeast strain modification and highlights the unbound possibilities for mutant strain production.