Development of a scalable solution that unlocks thiols from their S-conjugate precursors in dry hopping

Details

Supervisors

Collin, Sonia ; Gros, Jacques

Faculty

Faculté des bioingénieurs

Degree label

Master de spécialisation en génie brassicole

Abstract

Flavors are one of the most important determinants of food quality and acceptance. Since the late '90s, polyfunctional thiols (PFT) in wine and beer drive discussions and interest of scientists. Those volatile molecules found in hops and grapes are characterized by very potent and pleasant aromas and by very low odor thresholds. Non-volatile precursors generate a large portion of the beers’ PFT through the release occurring in-process and beyond. A sequential activity of yeast γ-glutamyl transpeptidase (GGT), carboxypeptidase and β-lyase enzymes is mostly responsible for free volatile PFT release in the final beer. The final molar conversion rates, below 1%, remain however low. The overall objective of the present study was therefore to develop a solution (enzymatic or not) which could be applicable for dry hopping to unlock volatile thiols from their bound precursors. Several options were assessed: an inhouse malt and barley extract performed with a straightforward process to be implemented at industrial scale; commercial enzymes such as BBA Optimalt β-amylase extract purified from barley and recombinant cystathionine β-lyase (CBL); and chemical release by thermal treatment and incubation of hop extraction with beer dicarbonyls (diacetyl and 2,3-pentanedione). The performances of these options were evaluated using Amarillo or Apollo spent hop extracts as sources of thiol S-conjugates, and their specificity was further investigated using synthetic precursors. Trials were performed at lab scale, while the best solutions were also tested at semi-industrial scale. Two parameters were important to determine the performance of the enzymatic or chemical treatment: the release percentage of the five PFT studied (3SHol, 3S4MPol, 3SPol, 4S4M2Pone and 3SHA) and the organoleptic qualities of the sample. To optimize the inhouse extracts’ performances, several parameters of enzyme solubilization and incubation were first assessed. Longer enzyme solubilization times as well as coincubation gave promising results. Overall, our research pointed out 6RW malt and 6RW barley to be potential candidates for optimal enzyme sources, as opposed to 2RW malt and feed barley. However, the key conclusion highlighted that inhouse enzymatic solutions were not efficient enough to induce a significant release of PFT that was strongly noticeable in beer aroma. In contrast, the two commercial enzymes displayed significant release of all PFT when assessed on hop extract or on synthetic cysteinylated precursors. While CBL demonstrated a significantly higher activity on such precursors, BBA Optimalt did exhibit significant GGT and carboxypeptidase coupled activity towards G-3SHol. Unfortunately, at the tested dosage, the β-amylase extract also displayed unpleasant overdose flavors due to the high dosage required, which also results in high cost associated. Finally, on the non-enzymatic pathway investigated, no significant change in PFT releases was observed for both chemical release trials performed, except for a potential increase of 4S4M2Pone release when heating Amarillo extract at 80°C for 40 minutes. The main limitation of such trials was the lack of sensory analysis to evaluate the impact of such treatments on the organoleptic qualities of the beer. In conclusion, the present study has not allowed to identify a sufficiently performing solution yet. However, several options have been screened, some of which are promising enough and should be further investigated. Complementary research should first and foremost focus on investigating the use of BBA Optimalt and CBL. Before considering further purification of the former or a partnership with an enzyme supplier to run tailored extraction of the latter from a natural source, more extensive assays, as well as sensory analyses, should be performed to fine-tune the model and define the optimal conditions to maximize the potential of the solutions already evidenced in the present study.