Article pubs.acs.org/JAFC
1,8-Cineole in French Red Wines: Evidence for a Contribution Related to Its Various Origins Xavier Poitou,*,† Cécile Thibon,‡ and Philippe Darriet*,†,‡ †
Univ. Bordeaux, ISVV, EA 4577 Œnologie, F-33140 Villenave d’Ornon, France INRA, ISVV, USC 1366 Œnologie, F-33140 Villenave d’Ornon, France
‡
S Supporting Information *
ABSTRACT: The aromatic descriptor “green”, reflecting grape unripeness in French red wines, is frequently associated with the levels of 3-alkyl-2-methoxypyrazines, particularly 3-isobutyl-2-methoxypyrazine (IBMP), which has bell pepper nuances. Nevertheless, not all green aromatic expressions in red wines correlate with 3-alkyl-2-methoxypyrazine concentrations. This study considered sensory and chemical approaches using Cabernet Sauvignon wines obtained from grapes harvested at one-month intervals during the 2014 and 2015 vintages to investigate other volatile odoriferous compounds. Semipreparative HPLC fractionation of wine extracts revealed a fraction with specific green aromas in the early harvest wines. Its sensory impact was confirmed by omission and reconstitution tests. Then, multidimensional gas chromatography coupled with olfactometry and mass spectrometry (MDGC-O-MS/TOF) was used for molecular characterization of the aroma compounds associated with the green aromas. Surprisingly, eucalyptol (1,8-cineole), with menthol odor was highlighted and assayed at concentrations sometimes above its olfactory detection threshold in Cabernet Sauvignon and Fer Servadou wines. Sensory tests confirmed its impact at several concentrations detected in French red wines (up to 2.61 ± 0.03 μg/L) on the menthol nuance and overall green perception, particularly via an additive effect with IBMP. Quantitation of 1,8-cineole in Cabernet Sauvignon and Merlot grapes during berry development in 2015 revealed its varietal origin with abundant concentrations in unripe berries and decrease during grape maturation. Moreover, the implication of an invasive plant (Artemisia verlotiorum) growing in certain vineyards was shown to be responsible for increased 1,8-cineole concentrations in some wines. KEYWORDS: French red wines, green aromas, 1,8-cineole, grape ripeness, Artemisia verlotiorum
■
Franc,13,14 Merlot,15 and Carmenere16 relatively to its low detection threshold in red wines.13 Also, high concentrations of IBMP in grapes at harvest, generally associated with unripeness,17−19 have a negative impact on wine quality. The link between the IBMP content and grape ripeness has been confirmed, decreasing significantly after veraison.13,20 It is commonly accepted that “fresh green” nuances (cut grass, menthol, green bell pepper) in red wines have a varietal origin, linked to grape ripeness. To date, IBMP is the only green volatile compound associated with grape ripeness, as contradictory results have been reported concerning varying concentrations of C6 alcohols.21−23 However, Preston et al.24 reported that IBMP levels in some wines were not sufficiently consistent to explain perceived variations in green notes. It has also been hypothesized that other odorous volatile compounds contribute to these nuances, alone or in combination with IBMP. The main objective of this study was to characterize other volatile odoriferous compounds that contribute to fresh green notes in French wines, in addition to IBMP, and particularly linked to grape ripeness. A strategy combining both analytical and sensory study was developed and applied to experimental wines obtained at two levels of ripening. Volatile compounds
INTRODUCTION Wine is a complex mixture of molecules originating from grapes, fermentation processes, and aging. Its aroma is the result of the simultaneous perception of hundreds of volatile compounds from a variety of chemical families such as thiols, monoterpenes, carbonyls, pyrazines, lactones, furanones, etc.1 These molecules are present at very low concentrations (μg or ng/L) and contribute to the individual nuances of wines by their relative proportions in the mixture. Perception is modulated by several phenomena, including masking effects, additivity, and even combinatorial phenomena at the brain level.2 Relative to the categorization of wine aromas in several families,3 the “vegetative” or “green” family is important to consider due to its perceptive antagonism with the fruity nuances of red wines.4,5 In this family, several volatile compounds have been cited as contributors to vegetal nuances, e.g., C6 alcohols for herbaceous nuances,6 4-mercapto-4methylpentan-2-one for box-tree aromas,7 dimethyl sulfide for olive notes in Syrah,8 3-methylthio-propan-1-al for cooked vegetable flavors,9 and 1,8-cineole for eucalyptus aromas.10 In terms of specific molecules associated with these nuances, 3alkyl-2-methoxypyrazines, particularly 3-isobutyl-2-methoxypyrazine (IBMP), initially identified in bell pepper (Capsicum annum var. grossum) by Buttery et al.,11 has been identified as a key compound. Many studies have revealed the major contribution of IBMP to the green aromas of red wines made from different varieties: Cabernet Sauvignon,12 Cabernet © XXXX American Chemical Society
Received: August 1, 2016 Revised: November 23, 2016 Accepted: November 27, 2016
A
DOI: 10.1021/acs.jafc.6b03042 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
from a Cognac vineyard were macerated in a 100 mL flask in hydroalcoholic solution (50% v:v) for 4 days. Winemaking Conditions. Grapes corresponding to each level of ripeness were destemmed mechanically (Bellot, Gradignan, France) then manually crushed and vinified in two 20 L stainless-steel tanks. Grape juices were inoculated with a commercial strain of Saccharomyces cerevisiae (FX10; Laffort Œnologie, Bordeaux, France) at 20 g/hL. To reproduce the winemaking conditions prevalent in the Bordeaux region, maceration lasted 3 weeks. During alcoholic fermentation (around 7 days), the cap was punched down twice per day. After alcoholic fermentation, maceration continued for 15 days. After running off, the wine underwent spontaneous malolactic fermentation in 5 L glass containers in a chamber maintained at 20 °C. The wine was then transferred into bottles and supplemented with 50 mg/L sulfur dioxide solution (6% v/v; Laffort, Bordeaux, France). Classical wine analysis (volatile acidity, sugar, and ethanol content) were done by Fourier transform Infrared spectroscopy (OenoFoss), an equipment regularly calibrated using Bordeaux wine standards. Wine Extraction for HPLC Fractionation. A 750 mL wine sample was extracted using 80, 50, and 50 mL of dichloromethane for 10 min each with magnetic stirring (700 rpm) and separated in a funnel. The organic phases were collected, frozen overnight at −20 °C to remove the emulsion, and then dried over sodium sulfate and concentrated to around 20 mL using a Buchi R-114 rotary evaporator (Buchi, Rungis, France) at room temperature (20 ± 1 °C) as mentioned by Stamatopoulos et al.26 The extract was filtered using a 0.45 μm membrane (PTFE) syringe and concentrated under a nitrogen flow (close to 100 mL/min) in a graduated glass tube (Atelier Jean Premont, Bordeaux) to obtain 500 μL of wine extract. Wine Extract Fractionation by Semipreparative HPLC. Reverse-phase (RP) HPLC was performed on the raw wine extract, using the Ultimate 3000 semipreparative HPLC system (Dionex, Courtaboeuf, France) and a Novapak C18 column (300 mm × 7.8 mm internal diameter (i.d.), 6 μm, Waters, Saint Quentin, France) with a guard column of the same phase. The procedure was based on the method described by Ferreira et al.27 Chromatographic conditions included a 250 μL injection volume with a flow rate of 1 mL/min. The linear program gradient was as follows: phase A, water; phase B, ethanol; 0−2 min, 100% A, 0−50 min, linearly programmed until 100%. Then fifty 1 mL effluent fractions were collected and evaluated for their sensory properties. Selected fractions were mixed with the next or previous fraction on the basis of their odor to form a 2 mL fraction group. Each was then diluted with ultrapure water to obtain 12% ethanol (v/v) and then reextracted three times with 10% (v/v), 5% (v/v), and 5% (v/v) dichloromethane, respectively, for 10 min each time. The organic phases were combined and concentrated to 100 μL under nitrogen flow before GC-O analysis and heart-cut, multidimensional GC-O-MS. Heart-Cut, Multidimensional Gas Chromatography Coupled to Olfactometry and Mass Spectrometry (MDGC-O-MS) Analysis. Multidimensional gas chromatography separation was performed on two Agilent 7890 GC (Agilent Technologies, Palo Alto, CA, USA), connected via a heated transfer line at 230 °C. The carrier gas used was helium N 60 (Air Liquide). Linear retention indices (LRI) were obtained by simultaneous injections of samples and a series of alkanes (C8−C20, Sigma-Aldrich, St Quentin Fallavier, France). The First Dimension (1D Column). This consisted of a fused-silica polar capillary column (BP20, SGE, 50 m, 0.25 mm i.d., 0.22 μm film thickness). The injector temperature was set at 230 °C in splitless mode (purge time = 1 min). To maintain a constant flow of 1 mL/ min, the pressure gradient was set as follows: 314 kPa for 1 min, increased to 1.89 kPa/min until 389 kPa, and maintained for 10 min. Oven 1 was programed as follows: 45 °C for 1 min, then increasing by 3 °C/min to 200 °C, then by 5 °C/min to 230 °C, and maintained at this temperature for 15 min. The total cycle lasted 73.6 min. The outlet of the first column was connected to a crosspiece (Gerstel) with five entries and 10% of the total flow was transferred into a deactivated, fused-silica column connected to an ODO-1 sniffing port (SGE, Ringwood, Australia). A Gerstel MCS 2 (multicolumn
were identified by a multiple-step approach, combining HPLC fractionation and MDGC-O/TOF-MS. Analyses were followed by sensory tests to study the impact of the target compounds on French red wines.
■
MATERIALS AND METHODS
Chemicals. Dichloromethane (99.9%) was supplied by VWR Chemicals (Fontenay-sous-Bois, France). Sodium chloride (NaCl) was supplied by Supelco (Bellefonte, PA, USA). Ultrapure water was obtained from a Milli-Q Plus water system (Millipore, Saint-Quentinen-Yvelines, France). All volatile chemicals were obtained from commercial sources by Sigma-Aldrich (Saint-Quentin-Fallavier, France): 3-isobutyl-2-methoxypyrazine (IBMP) (≥99% purity), 3isopropyl-2-methoxypyrazine (IPMP) (≥99%), 1,8-cineole (≥99%), (−)-α-thujone (≥96%), safranal (≥88%), and ethyl benzoate (≥99%). The labeled [2H3]-IBMP was synthesized as described in Guillaumie et al.25 Stock blended solutions of 100 mg/L (1,8-cineole) or 1 mg/L (IBMP) were prepared in HPLC-grade absolute ethanol (99.9%) (Merck, Fontenay-sous-Bois, France) and stored in the dark at +4 ◦C. Wines. All wines selected are listed in Table 5. These wines were used for IBMP, 1,8-cineole, and α-thujone quantitation. Grapes and Plant Materials. Sampling Grapes for Experimental Wines. To constitute an experimental of red wines with pronounced differences in aromatic constitution related to grape ripeness, 30 kg of Cabernet Sauvignon bunches were harvested from healthy plants, using identical rows in a vineyard plot, one month apart, in 2014 and 2015 (Bordeaux, Haut-Medoc). In 2014, the first grapes were harvested on September 12 (Ripeness−) and the second batch on the day of the winery’s main harvest on October 12 (Ripeness+). In 2015, the grapes were harvested on September 1 and October 1. The same harvesting strategy was applied to Merlot grapes in 2014 (September 5, October 3) (Table 1).
Table 1. Harvest Date, Technological Parameters, and Sensory Descriptions of Merlot (M) and CabernetSauvignon (CS) Experimental Wines from the 2014 and 2015 Vintages sample
variety
harvest date
% TAV
pH
aromas
Ripeness−
M
05/09/14
9.7
3.25
Ripeness−
CS
12/09/14
9.5
3.51
Ripeness+
M
03/10/14
13.0
3.53
Ripeness+
CS
12/10/14
12.0
3.74
Ripeness−
CS
01/09/15
9.5
3.42
Ripeness+
CS
01/10/15
11.9
3.70
fresh red fruits: cherry, raspberry strong green, green bell pepper, fresh red fruits: cherry, raspberry, strawberry fresh black fruits: blackcurrant, blackberry; green fresh fruits: raspberry, blackcurrant red and black fruits: blackberry, blackcurrant
Preparation for Berry Cycle Analysis. For IBMP and 1,8-cineole analysis, 100 berries of Cabernet Sauvignon and Merlot were randomly sampled each week from the same healthy, marked plants in rows of grapevines grown at ISVV (Institut des Sciences de la Vigne et du Vin, Bordeaux vineyard) during the 2015 growing and ripening season (July 17 to October 03). Grape berries without pedicels (n = 30) were frozen in liquid nitrogen and then introduced into a mechanical ball mill for 2 min and ground into a fine powder. The powder was directly introduced into a 20 mL headspace amber for analysis. Each modality was performed in duplicate (n = 2). Maceration of Plants. Ten grams of fresh Artemisia verlotiorum collected from a vineyard plot in Pauillac where the PT09, PT10, and PT14 wines are elaborated (Table 5) and Artemisia vulgaris collected B
DOI: 10.1021/acs.jafc.6b03042 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry switching system) maintained a constant flush at 30 mL/min throughout the analysis, except during the “heart-cut”, when the analytes were transferred into the second dimension column (2D column). Cryofocusing at −50 °C was performed on the start of the second column. The Second Dimension (2D Column). This consisted of a fusedsilica, nonpolar, capillary column (BP1, SGE, 50 m, 0.25 mm i.d., 1 μm film thickness). The flow of the second column was regulated by the Gerstel MCS 2 module. The gradient of oven 2 began just after the “heart-cut”, when the flow from the Gerstel MCS2 was restored. A pressure ramp program was applied to the crosspiece as follows: 1 min at 179.80 kPa, then 1.5 kPa/min up to 245 kPa. Oven 2 was initially set at 45 °C for 1 min, then raised to 180 °C at 3 °C/min, to 230 °C at 10 °C/min, and held at this temperature for 20 min. The end of the second column was split (1:1) via a crosspiece (Gerstel) between a time-of-flight (TOF) mass spectrometer (JEOL Europe, Croissy sur Seine, France) and a sniffing port (ODP III, Gerstel). The transfer between the second crosspiece and the ionization source of the mass spectrometer was via a deactivated, fused-silica column heated at 230 °C. Electron impact mode was used at 70 eV. The mass spectrometer was set to full scan mode with a m/z 40−250 range. Instrument setting, data acquisition, and processing were controlled by Mass Center System (v2.4.0), Chromatogram Viewer (v2.3.0.1000), and ChemStation (vB.04.01) software. Chemical identification was confirmed with the help of the NIST 2005 library (Gaithersburg, MD, USA). Sensory Analysis. General Conditions. Sensory analyses were conducted by judges (enologists, researchers, or students) with a good experience of wine-tasting (tasting wine several times a week), particularly the sensory evaluation of green character. Orthonasal sensory evaluations took place in a temperature-controlled room (ISO 8589:2007) maintained at 20 ± 1 °C, equipped with individual boxes. Samples were presented in random order, in black tasting glasses, coded by three-digit numbers, and covered with plastic caps. All samples were exclusively evaluated orthonasally. Characterization of Fractions. Sensory evaluation of fractions (50) from the early- and normal-harvest wine was initially performed by three experienced judges using a free-vocabulary task. Fraction Omission Experiments. A panel of 14 judges (five women, nine men) participated in this session of sensory analysis. These experiments were conducted on HPLC fractions obtained from wine extract produced from unripe grapes (Ripeness−) (750 mL). Each reconstitution was diluted in Ultrapure water to obtain 12% ethanol (v/v). The total aromatic reconstitution (TAR), containing all 50 fractions (F1 to F50), constituted the reference sample. Three reconstitutions were tested: TAR without group A (F31+F32), i.e., TAR-A, TAR without group B (F35+F36), i.e., TAR-B, and TAR without group A or group B, i.e., TAR-[A+B]. A 35 mL sample of each reconstitution was poured into dark glasses and subjected to orthonasal evaluation by the panel, focusing on the intensity of green character descriptors, using a 10 cm unstructured scale. The main green aroma descriptors (bell pepper, cut grass, leaf, and menthol) were listed on the analysis chart. Fraction Supplementation in Wine. The supplementation sensory test was performed during the same session with HPLC fractions from the Ripeness− wine extract. The Ripeness+ wine was supplemented with selected fractions from Ripeness− wine. Five wine samples were considered: Ripeness− (R−), Ripeness+ (R+), Ripeness+ supplemented with A (R+ [+A]), B (R+ [+B]), or both (R+ [+A+B]). Each 35 mL wine sample was supplemented with 0.2 μL of group A and group B obtained from 750 mL wine. 1,8-Cineole and IBMP Supplementation in Wine. The wine used for all experiments was a red, nonbarrel-aged, commercial Bordeaux. This wine had neutral aromas and contained a very low concentration of 1,8-cineole (0.1 μg/L) and IBMP (
