The basic ingredients of lager beer are water, barley malt, hops, and yeasts. All contribute to the volatiles and the non-volatiles, influencing the aromas, flavors, and sensory properties of the final product. Among volatile compounds, higher alcohols and esters are key aroma components derived from the Ehrlich pathway – an amino acid catabolic pathway in yeasts. The current study aims to increase the concentrations of higher alcohols and esters in lager beer by increasing the precursor amino acids in wort via protease application in mashing. The selection of proteases is based on their activity and selectivity, and our understanding of barley malt endogenous enzymes and inhibitors, and protease products. We conducted three rounds of selections to find optimal combinations of endo- and exoproteases and tested the effectiveness to release the chosen amino acids – leucine, isoleucine, valine, and phenylalanine. Study 1 documented the best candidates favoring the release of leucine, a precursor to the two volatiles described as banana and fruity: 3-methylbutan-1-ol (i.e., isoamyl alcohol) and 3-methylbutyl acetate (i.e., isoamyl acetate). Fermentation of wort derived from protease-treated mash led to about 10% increase of 3-methylbutan-1-ol and 17% increase of 3-methylbutyl acetate.

The enhanced biosynthesis of 3-methylbutan-1-ol and 3-methylbutyl acetate is the result from 1) yeast leucine metabolism and 2) protease optimization of wort amino acid profile. To compare the influences from these two factors, we compared the aroma output from two mutagenized lager yeast strains from Trinity College Dublin and that from protease treatments in Study 2. The two strains from Trinity College Dublin included a

Weihenstephan 34/70 strain (termed TCD34/70) and a CBS 1538 strain (termed TCD1538), both of which are group II/Frohberg yeasts overexpressing genes related to the Ehrlich pathway. In Study 2, we employed protease combinations that could release leucine and phenylalanine in mashing, and subsequently tested fermentation by TCD34/70, TCD1538, and their mutants. Overall, we confirmed the general belief that yeast strain is the main determinant of beer aromas. The mutagenized TCD34/70 yeasts, termed TCD34/70-9.7, were able to maintain their aroma production patterns despite different amino acid profiles of wort before and after protease treatment. On the other hand, the mutagenized TCD1538 – termed TCD1538-11.1, produced less leucine and phenylalanine-derived higher alcohols and acetates despite the enriched amino acids present in wort after protease treatment.

In Study 1 and 2, we observed that the amino acid molar fraction affected beer aroma profile more than amino acid concentration. Considering this, we further used design of experiments to maximize leucine molar fraction in wort in Study 3. The variables of interest were endoprotease dosage, exoprotease dosage, and temperature. We managed to increase leucine molar fraction from 8% in control to 13% after protease treatment in small-scale mashing. This protease method was tested in wort production for fermentation to make low alcohol beer by two commercially important non-Saccharomyces strains: Saccharomycodes ludwigii and Pichia kluyveri.