Article pubs.acs.org/JAFC Metabolism of Nonesterified and Esterified Hydroxycinnamic Acids in Red Wines by Brettanomyces bruxellensis † ‡ § § † Lauren M. Schopp, Jungmin Lee, James P. Osborne, Stuart C. Chescheir, and Charles G. Edwards*, † School of Food Science, Washington State University, Pullman, Washington 99164−6376, United States ‡ USDA−ARS−HCRU worksite, Parma, Idaho 83660, United States § Department of Food Science and Technology, Oregon State University, Corvallis, Oregon 97331, United States ABSTRACT: While Brettanomyces can metabolize nonesterified hydroxycinnamic acids found in grape musts/wines (caffeic, p- coumaric, and ferulic acids), it was not known whether this yeast could utilize the corresponding tartaric acid esters (caftaric, p- coutaric, and fertaric acids, respectively). Red wines from Washington and Oregon were inoculated with B. bruxellensis, while hydroxycinnamic acids were monitored by HPLC. Besides consuming p-coumaric and ferulic acids, strains I1a, B1b, and E1 isolated from Washington wines metabolized 40−50% of caffeic acid, a finding in contrast to strains obtained from California wines. Higher molar recoveries of 4-ethylphenol and 4-ethylguaiacol synthesized from p-coumaric and ferulic acids, respectively, were observed in Washington Cabernet Sauvignon and Syrah but not Merlot. This finding suggested that Brettanomyces either (a) utilized vinylphenols formed during processing of some wines or (b) metabolized other unidentified phenolic precursors. None of the strains of Brettanomyces studied metabolized caftaric or p-coutaric acids present in wines from Washington or Oregon. KEYWORDS: Brettanomyces, hydroxycinnamic acids, phenolic acids, cinnamates, volatile phenols ■ INTRODUCTION pool of precursors available for conversion to volatile phenols. 21 Brettanomyces has been regarded by some as the most severe In fact, Nikfardjam et al. postulated that those grape cultivars 1 fi with high amounts of hydroxycinnamic acids could be more microbiological threat to wine quality causing serious nancial fi losses each year.2 As part of its metabolism, Brettanomyces prone to Brettanomyces infections, although speci c precursors utilizes hydroxycinnamic acids to produce ethylphenols (i.e., were not studied. As such, the objective of this study was to volatile phenols) which have been sensorily described as determine utilization of red wine hydroxycinnamic acids by − “barnyard,”“leather,”“horse sweat,” and others.3 5 Biochemi- strains of B. bruxellensis isolated from Washington and cally, the hydroxycinnamic acids are first decarboxylated into California. vinylphenols, which are subsequently reduced to ethyl- 3,6 ■ MATERIALS AND METHODS phenols. Biosynthesis is dependent on many factors including − fi the strain and energy source such as glucose or ethanol.7 13 Strains and Starter Cultures. For the rst experiment, B. 14 bruxellensis strains B1b, B5, E1, and I1a were originally isolated from As reviewed by Conde et al., hydroxycinnamic acids 22 ff Washington wines as described by Jensen et al. All strains were (ca eic, p-coumaric, or ferulic acids) are located in the vacuoles maintained in glycerol at −70 °C as well as being streaked on WL of the skin and pulp cells of grapes, primarily as esters of tartaric medium (Oxoid, Hampshire, England). Starter cultures were prepared acid (caftaric, p-coutaric, or fertaric acids, respectively). Once a by transferring a single colony into 100 mL of YM broth (Difco, must is prepared and during the course of fermentation, tartaric Sparks, MD) containing 5% v/v ethanol and adjusted to pH 3.8. After acid can be hydrolyzed to form the nonesterified acids with a week of growth, 1 mL of culture was transferred to additional YM concentrations varying according to cultivar, vintage, and broth containing 10% v/v ethanol to improve acclimation to wine − winemaking conditions.15 18 Analyzing Merlot grapes from conditions. Cells were harvested by centrifugation (4200g, 15 min) Washington, Nagel et al.16 reported that the average after an additional week of incubation, washed in 0.1% w/v peptone, and resuspended in additional peptone. Wines were inoculated, in concentrations of caftaric, p-coutaric, and fertaric acids were triplicate, at initial populations of 105 to 106 cfu/mL and monitored 59.2, 16.9, and 3.2 mg/L, respectively. Furthermore, concen- over a nine-week period by spiral plating with an Autoplate 4000 trations of caftaric and p-coutaric acids decreased 34% to 61% (Spiral Biotech, Bethesda, MD) on WL agar. All plates were incubated during fermentation and subsequent storage of Cabernet at 26 °C prior to counting. Sauvignon and Merlot wines.15 In a recent study, Ginjom et As part of a second experiment, additional strains of B. bruxellensis al.18 calculated p-coutaric/p-coumaric acids molar ratios to vary (493, 495, 496, 497, 607, 613, 614, 615, 616, 635, and 643) were from 0.6 to 2.1 depending on cultivar and fermentation provided by E.&J. Gallo Winery (Modesto, CA) and maintained in − ° temperature. glycerol stored at 80 C. Starter cultures were prepared as previously While Brettanomyces possess enzymatic activity to act on a number of esters19 including the ethyl esters of hydroxycin- Received: August 5, 2013 namic acids,20 it is not known whether tartaric acid esters can Revised: November 6, 2013 serve as substrates. If Brettanomyces are able to metabolize Accepted: November 12, 2013 tartaric acid esters, these compounds could represent a large Published: November 12, 2013 © 2013 American Chemical Society 11610 dx.doi.org/10.1021/jf403440k | J. Agric. Food Chem. 2013, 61, 11610−11617 Journal of Agricultural and Food Chemistry Article Figure 1. HPLC chromatogram of commercially prepared Pinot noir wine 63 days after no inoculation (A) or inoculation with B. bruxellensis strain E1 (B). Peak identifications: (1) caftaric acid, (2) p-coutaric acid, (3) internal standard chlorogenic acid, (4) caffeic acid, (5) p-coumaric acid, (6) ferulic acid, and (u) unknown. described except harvested cells were resuspended in 0.1% w/v aseptically collecting 5 mL samples and adding 40 mg/L chlorogenic peptone for inoculation into wine at 103 to 104 cfu/mL. acid (Acros Organics, Geel, Belgium) as the internal standard. Samples Wines. In a first experiment, commercially prepared wines of were immediately frozen at −10 °C until phenolic acid purification Cabernet Sauvignon (pH 3.79; 13.7% v/v alcohol), Merlot (pH 3.71; using neutral Bakerbond SPE C18 column (J. T. Baker, Phillipsburg, 13.4% v/v alcohol), and Syrah (pH 3.69; 13.1% v/v alcohol) were NJ) and subsequently through a preconditioned acidic SPE column. obtained from a Washington winery, while Pinot noir (pH 3.57; 12.2% The columns were preconditioned according to the methods of 23 v/v alcohol) originated from a winery located in Oregon. Residual SO2 Jaworski and Lee. Wines were adjusted to pH 7.0 using 5 N NaOH fi μ fi was removed by the addition of equimolar amounts of H2O2 (J.T. and ltered through a 0.45 m PVDF syringe lter (Whatman, Baker, Phillipsburg, NJ), while ethanol concentrations were stand- Piscataway, NJ) before 3 mL was then passed through the ardized to 12.8% v/v by the addition of Milli-Q water and/or 200- preconditioned neutral SPE column. The neutral column was washed proof ethanol. Additionally, 0.5% w/v glucose (J.T. Baker, Phillipsburg, with Milli-Q water (4 mL) with the effluent acidified to pH 2.0 using NJ), 0.5% w/v fructose (Spectrum, Gardena, CA), and 0.1% w/v yeast 0.1 N HCl prior to being passed through a preconditioned acidic SPE extract (Difco, Sparks, MD) were added to all wines to limit potential column. Phenolic acids were eluted using 2 mL of methanol (J.T. differences in nutrient composition. Finally, the pH of all four wines Baker, Phillipsburg, NJ) which was concentrated to approximately 0.5 was adjusted to 3.94 through the addition of 5 N NaOH. After mL using a Rotavapor R-210 equipped with a heating bath at 32 °C additions, wines were sterile-filtered through 0.22 μm PVDF filters (Buchi, Flawil, Switzerland). (Millipore, Bedford, MA) prior to the addition of 0.1% w/v autoclaved A HPLC-DAD unit with detection set at 320 nm (Agilent 1100 suspension cellulose (Sigma-Aldrich, St. Louis, MO). While wines series, Wilmington, DE) was used to analyze samples as described by were stirred, 100 mL aliquots were transferred into sterile milk dilution Lee and Finn.24 Phenolic compounds were identified based on UV− bottles using a varistaltic dispenser pump before inoculation. visible spectra and retention times of known standards (caftaric, Uninoculated wines serves as controls. caffeic, p-coumaric, and ferulic acids) obtained from Sigma-Aldrich (St. In a second experiment, a Pinot noir wine was produced using Louis, MO). When standards were not available (i.e., p-coutaric acid), grapes obtained from the Woodhall Vineyard at Oregon State retention times were used from Lee and Scagel25 and verified with University (Alpine, OR). Once harvested, the grapes were stored at UV−visible spectra.26 Calibration curves were prepared for caffeic, p- 4 °C overnight before being crushed/destemmed, pooled, and coumaric, and ferulic acids in 10% v/v ethanol. Calibration graphs distributed (60 L) into 100 L stainless steel tanks. No SO2 additions were prepared by plotting the ratios of the standard peak areas to the were made. Each lot was inoculated with an active dry form of internal standard peak area versus the ratios of the standard Saccharomyces cerevisiae strain VQ-15 (Lallemand, Montreal,́ Canada) concentrations to the internal standard concentration. The curves rehydrated according to manufacturer’s specifications to yield 106 cfu/ (five data points) were linear with R2 values higher than 0.999. Caftaric mL. For fermentation, the tanks were placed into a room held at 27 °C acid was quantified using the caffeic acid curve, whereas p-coutaric acid with punch downs performed twice daily.