Genomic saturation mutagenesis and polygenic analysis identify novel yeast genes affecting ethyl acetate production, a non-selectable polygenic trait

Abstract

Isolation of mutants in populations of microorganisms has been a valuable tool in experimental genetics for decades. The main disadvantage, however, is the inability of isolating mutants in non-selectable polygenic traits. Most traits of organisms, however, are non-selectable and polygenic, including industrially important properties of microorganisms. The advent of powerful technologies for polygenic analysis of complex traits has allowed simultaneous identification of multiple causative mutations among many thousands of irrelevant mutations. We now show that this also applies to haploid strains of which the genome has been loaded with induced mutations so as to affect as many non-selectable, polygenic traits as possible. We have introduced about 900 mutations into single haploid yeast strains using multiple rounds of EMS mutagenesis, while maintaining the mating capacity required for genetic mapping. We screened the strains for defects in flavor production, an important non-selectable, polygenic trait in yeast alcoholic beverage production. A haploid strain with multiple induced mutations showing reduced ethyl acetate production in semi-anaerobic fermentation, was selected and the underlying quantitative trait loci (QTLs) were mapped using pooled-segregant whole-genome sequence analysis after crossing with an unrelated haploid strain. Reciprocal hemizygosity analysis and allele exchange identified PMA1 and CEM1 as causative mutant alleles and TPS1 as a causative genetic background allele. The case of CEM1 revealed that relevant mutations without observable effect in the haploid strain with multiple induced mutations (in this case due to defective mitochondria) can be identified by polygenic analysis as long as the mutations have an effect in part of the segregants (in this case those that regained fully functional mitochondria). Our results show that genomic saturation mutagenesis combined with complex trait polygenic analysis could be used successfully to identify causative alleles underlying many non-selectable, polygenic traits in small collections of haploid strains with multiple induced mutations.

Polygenic Analysis in Absence of Major EffectorATF1Unveils Novel Components in Yeast Flavor Ester Biosynthesis

“…Ethyl acetate production during alcoholic fermentation has been investigated previously by random mutagenesis and pooled-segregant quantitative trait locus (QTL) analysis, with the identification of causative alleles of TPS1 , PMA1 , and CEM1 ( 9 ). CEM1 encodes a mitochondrial β-keto-acyl synthase, a homologue of the fas2 β-keto-acyl synthase subunit, that is required for respiration, but apparently does not affect mitochondrial lipid homeostasis ( 10 ).…”

QTL mapping of volatile compound production in Saccharomyces cerevisiae during alcoholic fermentation

“…The detected QTLs were compared with loci found in QTL mapping studies of similar traits [ 10 , 13 – 15 , 24 ]. Only QTL chr7@161.6, which influences ethyl lactate formation, co-localizes with PMA1 , a plasma membrane P2-type H + -ATPase that was shown by Abt et al (2016) to be the responsible gene in a QTL affecting ethyl acetate production.…”

Identification of Novel Alleles Conferring Superior Production of Rose Flavor Phenylethyl Acetate Using Polygenic Analysis in Yeast

“…Flavor production is highly variable among yeast strains, and the genetic basis of this broad phenotypic variation has remained largely unknown until recently. Quantitative trait locus (QTL) mapping studies have now revealed genes involved in production of nerolidol, 2-phenyl ethanol, and ethyl esters ( 13 ), ethyl acetate ( 14 ), and undesirable sulfur flavor compounds ( 15 ). Given the large variety of flavor compounds and the many parameters affecting their formation, most of the underlying genetic basis of the natural variation in flavor compound production remains unknown.…”

“…This low number is important for genetic analysis of nonselectable phenotypes and for traits that require an elaborate experimental setup for scoring, such as a requirement for small-scale fermentations with all individual segregants. The latter has been successfully accomplished for the nonselectable traits of low glycerol production ( 24 , 25 ), maximum ethanol accumulation capacity ( 23 ), ethyl acetate production ( 14 ), and production of multiple other flavor compounds ( 13 ). Production of flavor compounds, like ethyl acetate, is a nonselectable trait and has to be scored in individual small-scale fermentations with hundreds of segregants to obtain a pool of superior segregants large enough for efficient QTL mapping ( 13 , 14 ).…”

“…Bulk RHA has also been used in this bulk RHA; first, consecutive blocks of genes are deleted, followed by deletion of single genes in blocks where a phenotypic difference between the two hemizygous diploid strains is observed in the RHA analysis ( 22 ). For production of the flavor compound ethyl acetate, QTL mapping using pooled-segregant whole-genome sequence analysis and RHA for determination of causative genes have resulted in the identification of PMA1 , CEM1 , and TPS1 as novel genes affecting ethyl acetate production ( 14 ). The recent development of the clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 technology for site-directed genetic modification in yeast ( 26 – 30 ) allows us to exchange precisely the causative alleles for their counterparts in the superior and inferior parent, which allows evaluating the contribution of the causative alleles in the haploid parent background.…”