The goal of this proposal was the determination of the genetic basis of the variation in hydrogen sulfide formation observed in commercial and native isolates of Saccharomyces cerevisiae. An initial goal was the characterization of the variability in strain behavior using different synthetic media and natural juices. No correlation was found between the basal level of sulfite reductase activity and hydrogen sulfide production. While strains lacking sulfite reductase did not produce hydrogen sulfide, they did produce copious amounts of sulfite. Methionine and cysteine supplementation had variable effects on volatile sulfur formation. In general these compounds were only of limited effectiveness or lead to the formation of higher sulfur volatiles that are the degradation products of these amino acids. All yeast strains evaluated show an effect of limiting nitrogen in the increased production of hydrogen sulfide, but the level of nitrogen at which this occurs is variable and impacted by medium composition, likely due to relative differences in demand for nitrogen.
There are several major conclusions of this work. First, cytoplasmic cysteine levels are correlated with activity of the sulfate reduction pathway and the production of hydrogen sulfide. Second, cytoplasmic cysteine levels vary up to five fold in different genetic backgrounds in strains grown under identical conditions. Those strains maintaining high internal concentrations of cysteine produce lower amounts of sulfide. Third, methionine and glutathione levels were not found to be correlated with sulfate reduction and H2S production. Fourth, cytoplasmic cysteine levels were a function of the activity of cystathionine b-synthase (CYS4) in most genetic backgrounds. In one genetic background, UCD713, a defective allele of MET17, the step immediately preceding CYS4, carried a mutation reducing cysteine production. Elevation of the activity of this protein restored cysteine levels and reduced sulfide production. Fifth, elevation of CYS4 activity by transformation of strains with a plasmid leading to over-expression of the protein increased cysteine pools and reduced hydrogen sulfide formation in some strains. In other strains over-production of the enzymatic activity did not occur, and in others it was reduced, suggesting that high enzyme levels can catalyze oligomer formation. Sixth, CYS4 allele analysis revealed that one strain that consistently produces little to no H2S carries a mutation in an important regulatory domain. Thus, the differences in hydrogen sulfide formation of different enological strains of Saccharomyces are due to differences in the pool levels of cysteine, which is a function of the relative activity of the enzymes directly involved in cysteine biosynthesis. Finally, alterations of the activity and regulation of Cys4p are associated with greatly diminished hydrogen sulfide production. Allele swap technology should be useful for the generation of low sulfide producing variants of any commercial strain.