Highly Selective Enzymatic Degradation of Undesirable Phenols in Brettanomyces and Smoke Taint Contaminated Wines. AVF Project #2025-2795.
Summary of Progress:
We have completed the third year of a multi-year project focused on developing enzymes that selectively degrade 4-ethylguaiacol (4EG) and 4-ethylphenol (4EP), the primary sensory-active compounds in Brettanomyces-contaminated wine and a subset of the broader taint phenols present in smoke-tainted wine (Licker 1998; Mirabelli-Montan 2021). A second phase of this project, beyond the scope of the current proposal period, will extend the same engineering framework to additional smoke taint marker compounds, including guaiacol, 4-methylguaiacol, cresols, syringol, and 4-methylsyringol.
In Yr 1 and Yr 2, we screened a diverse library of 46 laccases across buffered and wine-like conditions, identifying high-performing candidates and narrowing to two primary scaffolds, laccases #15 and #36. We then mapped active-site hot spot regions in both enzymes using alanine substitutions to distinguish residues essential for catalysis from positions that tolerate mutation while maintaining activity. Structural modeling and docking highlighted a key design requirement: both scaffolds have wide, accessible active sites that support taint phenol turnover but also increase the risk of off-target oxidation of desirable wine aroma chemistry, motivating active-site enclosure as the path toward specificity.
In Yr 3, we transitioned from hot spot mapping to AI-guided design and experimental validation of novel enzymes, termed Wine Specific Phenolases (WSP). We implemented a custom design pipeline that introduces targeted structural insertions at permissive hot spots, applies sequence optimization, and uses stringent structural filtering to preserve the trimeric architecture required for activity. Trimer stability emerged as a strict constraint that reduced the number of viable designs and required iterative refinement of the workflow, particularly for laccase #36. Despite this bottleneck, experimental testing validated multiple successful designs. In Round 1, six WSP designs spanning a range of active-site enclosure were produced and tested in model wine; three retained full 4EG degradation while tolerating added structure near the pocket, whereas more aggressively enclosed or heavily edited variants lost activity, establishing a clear boundary for maintaining function. In Round 2, refined WSP candidates selected from the larger design pools yielded additional successes, including one that maintained full 4EG degradation while increasing 4EP degradation to 52 percent, representing a near fivefold improvement relative to wild type under the same conditions. These results demonstrate that active-site enclosure is compatible with catalytic function and that iterative refinement can improve 4EP performance without sacrificing 4EG activity.
Over the next year, our primary focus will shift from model-wine activity to selectivity in Brett-tainted wine. We have begun testing WSPs directly in tainted wine to quantify 4EG and 4EP removal while monitoring representative non-target phenolics and aroma-linked markers, and these outcomes will serve as explicit constraints for the next design cycle by identifying and eliminating off-target risk interactions. In parallel, we will initiate staged sensory validation, starting with small-scale difference testing on top candidates to detect any negative sensory shifts, and expanding to more structured ranking once chemical selectivity and efficacy are demonstrated.
