Food Chemistry 127 (2011) 547555

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Evaluation of different Saccharomyces cerevisiae strains on the prole of volatile compounds and polyphenols in cherry wines
Shu Yang Sun a,, Wen Guang Jiang b, Yu Ping Zhao c
a

School of Food Engineering, Ludong University, Yantai, Shandong 264025, PR China Technology Center of Changyu Pioneer Wine Company Limited, Yantai, Shandong 264025, PR China c Institute of Food Science and Engineering, Yantai University, Yantai, Shandong 264005, PR China
b

a r t i c l e

i n f o

a b s t r a c t
Tart cherries of Early Richmond, widely grown in Shandong (China), were fermented with six different Saccharomyces cerevisiae strains (BM44, RA17, RC212, D254, D21 and GRE) to elucidate their inuence on the production of volatiles and polyphenols. Acetic acid and 3-methylbutanol were found in the highest concentrations among all identied volatiles with all six yeast strains, followed by 2-methylpropanol and ethyl lactate. RA17 and GRE cherry wines were characterised by a higher amount of esters and acids. D254 wine contained a higher concentration of alcohols. With respect to polyphenols, ve phenolic acids and four anthocyanins were identied among all tested samples, with chlorogenic and neochlorogenic acids, cyanidin 3-glucosylrutinoside and cyanidin 3-rutinoside being the major compounds. When using principal component analysis to classify the cherry wines according to the volatiles and polyphenols, they were divided into three groups: (1) RA17 and GRE, (2) RC212 and D254 and (3) BM44 and D21. 2011 Elsevier Ltd. All rights reserved.

Article history: Received 23 September 2010 Received in revised form 24 November 2010 Accepted 11 January 2011 Available online 19 January 2011 Keywords: Cherry wine Volatile compounds Polyphenols Saccharomyces cerevisiae

1. Introduction Shandong (a province situated in eastern China) is the main region of cherry production in the country. The cherry crop occupies more than 8000 ha in this province, with an estimated annual harvest of 60,000 tons, creating a sales revenue of over 2 billion Chinese yuan (Han et al., 2008). However, cherry is relatively perishable, and is often not availably served as fresh table fruit after transportation; thus, it is often processed to juice and wine, especially with the sour varieties (Prunus cerasus L.). In Shandong, for many years, the cherry wine has been an important segment of the fruit and vegetable industry. As Shandong cherry wine became more popular in China over the past decade, its characteristic and distinct avour began to receive more scrutiny from the consumers. Aroma is a most important and distinguishing characteristic of cherry wine. It is derived from hundreds of volatile compounds from cherry berry, and from wine-making by yeasts and aging process. Considering the fact that (1) most of the odorous compounds are produced during fermentation, and (2) yeasts have a signicant effect on the sensory characteristics of wines, the selection of a proper yeast strain is critical for the development of the desired cherry wine style (Patel & Shibamoto, 2002). During fermentation, the function of yeasts in the production of avour is to release varietal volatile compounds from cherry precursors, and in the meantime, to synthesise de novo yeast-derived volatile compounds
Corresponding author. Tel.: +86 0535 6695491; fax: +86 0535 6695490.
E-mail address: sysun81@yahoo.com.cn (S.Y. Sun). 0308-8146/$ - see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodchem.2011.01.039

(Molina, Swiegers, Varela, Pretorius, & Agosin, 2007; Molina et al., 2009). Today, a wide range of wine yeast strains are commercially available, which offers us the opportunity to explore one or more suitable yeasts to assure a rapid and reliable fermentation process, and give wines a consistent and predictable quality (Rodriguez, Lopes, Barbagelata, Barda, & Caballero, 2010). However, to the best of our knowledge, there have been limited studies performed to evaluate the effect of Saccharomyces cerevisiae strains on the avour proles of cherry wines. Furthermore, cherry wines are still under-researched among fruit wines. Hence, the rst objective of this study is to employ the headspace solid phase microextraction with gas chromatographymass spectrometry (HSSPME-GCMS) technique for the identication and quantication of volatiles of cherry wines inoculated with six commercial yeast strains. Phenolic compounds are also important group of cherry wine constituents, and they greatly contribute to the sensory properties by affecting the colour and taste (Czyzowska & Pogorzelski, 2002). These compounds are natural antioxidants, and possessing neuroprotective and potent cancer-preventive properties, which are considered as benecial and proved (Lee, Hur, Lee, & Lee, 2005; Yoo, Al-Farsi, Lee, Yoon, & Lee, 2010). Wine, in comparison with other sources, contains relatively high amounts of highly diversied polyphenols. Most of the phenolic compounds pass from berry to the wine during extraction and fermentation, and a few are newly formed, such as some free phenolic acids and avonic isomers (Wulf & Nagel, 1980). One of the possibilities of the occurrence of new polyphenols in wine may be transformation (e.g.

548

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

enzymatic) of substances contained in fruit. Enzymes of fruit origin and microorganisms responsible for fermentation may lead to oxidation or hydrolysis of native components (Nagel & Wulf, 1979), therefore, when fermenting the same must, the employment of whatever yeast strains are straightly associated with the composition and content of polyphenols in the nal product. Here, in the current investigation, the second purpose was to elucidate the effect of different yeast species on the proles of polyphenols in cherry wines, with special attention paid to the phenolic acids and anthocyanins. 2. Materials and methods 2.1. Chemicals FolinCiocalteus phenol reagent, chlorogenic acid, p-coumaric acid, neochlorogenic acid, p-hydroxybenzoic acid, caffeic acid, cyanidin 3-glucoside, gallic acid, anhydrous sodium sulphate, ethanol (chromatographic grade) and methanol (99%) were purchased from Sigma Chemical Co. (St. Louis, MO). Cyanidin 3-rutinoside, cyanidin 3-glucosylrutinoside and peonidin 3-rutinoside were obtained from Polyphenols Laboratories (Sandnes, Norway). All volatile standards were purchased from Aldrich (Milwaukee, Wis., USA) and Fluka (Buchs, Switzerland). All other chemicals used were of analytical or high performance liquid chromatography (HPLC) grade. 2.2. Yeast strains Six S. cerevisiae commercial strains including Lalvin BM44, RA17, RC212, D254, D21 and GRE, purchased from Lallemand (France), were used in this work. 2.3. Preparation of cherry wine Fruits of Early Richmond picked at commercial maturity during the 2009 harvest season from the menlou cherry orchard (Yantai, China) were used in this work. Cherries were crushed, manually deseeded and poured into 20 L bottles where they were treated pectinolytic preparation Lallzyme EX-V pectinase (20 mg/L) for 10 h at 20 C. Then, sucrose was added up to 210 g/L; sulphur dioxide additions were made as needed to maintain approximately 30 mg/L of free sulphur dioxide. Initial must analysis results were as follows: total sugars 165 g/L, titratable acidity 7.9 g/L, pH 4.03. Fermentation occurs under control temperature (20 C) with an association of S. cerevisiae strains (250 mg/L) as recommended by the producer. Fermentation progress was monitored by measuring the total reducing sugars (TRS) using a glucose oxidase test strip (Chemstrip, Boehringer Mannheim Corp., Indianapolis, IN) until the sugar content was reduced below 1 g/L. The nished young wine was ltered through four layers of ne cheesecloth into 750 mL wine bottles that were subsequently corked and stored in an 18 C storage chamber until the analysis. All samples were done in triplicate. 2.4. Analytical method Total reducing sugars (TRS), titratable acidity (expressed as g/L of malic acid), pH, total soluble solids (TSS), ethanol, volatile acidity (VA), and free SO2 were determined according to the ofcial methods (OIV, 2005). 2.5. Total phenolics The total content of phenolics was measured according to the FolinCiocalteu colorimetric method (Rakitzis, 1975). Briey,

1 mL of appropriately diluted sample or a standard solution was added to 9 mL of distilled water, 1 mL of FolinCiocalteu phenol reagent and 10 mL of 7% Na2CO3 solution. The mixture was then immediately diluted to a volume of 25 mL in a volumetric ask with distilled water and left to stand for 90 min at 23 C. The absorbance was read at 750 nm, and results are expressed as milligrams of gallic acid equivalents (GAE) per litre of cherry wine. 2.6. Total anthocyanins Total anthocyanins were determined by the pH differential method (Kim, Heo, Kim, Yang, & Lee, 2005). Cherry wine dissolved in 0.025 M potassium chloride buffer (pH 1.0) and 0.4 M sodium acetate buffer (pH 4.5) were measured simultaneously at 510 and 700 nm after 15 min of incubation at 23 C. The content of total anthocyanins was expressed as milligrams of cyanidin 3-glucoside equivalents (CGE) per litre of cherry wine. A molar absorptivity of 26,900 L/mol cm was used for cyanidin 3-glucoside, which has the molecular mass of 449.2 g/mol. 2.7. Fractionation of polyphenols For the easier separation of anthocyanins from nonanthocyanin phenolics during HPLC analysis, a simple fractionation was performed using two connected and preconditioned C18 Sep-Pak cartridges (Waters, Milford, MA, USA), as described by Kim et al. (2005). Briey, a load of 200 lL phenolic extract was applied onto the cartridges, and washed with 6 mL of 0.01 N aqueous HCl to get rid of sugars, acids, and other water-soluble compounds. Nonanthocyanins were obtained by rinsing cartridges with 20 mL of ethyl acetate and followed by elution. Finally, the adsorbed anthocyanins were eluted with 10 mL of methanol with 0.1% (v/v) HCl, collected, dissolved in 50% aqueous methanol and distilled water, and stored at 4 C under nitrogen gas to prevent oxidation until HPLC analysis. 2.8. HPLC analysis of polyphenols A reversed-phase HPLC system (HewlettPackard model 1100; Palo Alto, CA) consisting of a 20 lL sample loop, a photodiode array detector, a quaternary pump, and a vacuum degasser was employed for the analysis of polyphenols as reported previously (Kim et al., 2005). Separation was performed on a C18 reversed-phase Symmetry Analytical column (5 lm 250 mm 4.6 mm; Waters Corp., Milford, MA) thermostated at 30 C. The mobile phase was composed of a solvent A (0.1% H3PO4 in HPLC grade water) and solvent B (0.1% H3PO4 in HPLC grade acetonitrile). Linear solvent gradient was applied as follows (total 60 min): 92% A/8% B at 0 min, 89% A/11% B at 4 min, 87% A/13% B at 25 min, 80% A/20% B at 27.5 min, 40% A/60% B at 50 min, 92% A/8% B at 55 min, and 92% A/8% B at 60 min. The ow rate was at 1.0 mL/min, and the runs were monitored at the following wavelengths: phenolic acids at 320 nm, and anthocyanidins at 520 nm. UVvisible spectra were simultaneously recorded from 200 to 600 nm during sample running. For quantication analysis, the standard curves of individual phenolics that relate various concentrations of authentic standard solutions to the areas of their corresponding peaks were obtained. 2.9. Qualitative analysis of volatile compounds A portion of 5 mL cherry wine, 5 lL of 3-octanol (internal standard, 100 mg/L standard solution in ethanol) and 1.5 g of NaCl were added into 15 mL headspace vial. Vials were capped with a Teon septum and an aluminium cap (Chromacol Ltd., Herts, England). Volatile compounds were extracted from the headspace using a bre of 50/30 lm DVB/CAR/PDMS (Supelco, Bellefonte

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

549

PA, USA). Before extraction, the bre was conditioned by being inserted into the GC injector at the 250 C for 2 h to prevent contamination. Then, the bre was exposed to the headspace of above prepared glass vial, which was placed in a thermostatted bath adjusted to 30 C with constant agitation in a magnetic stirrer for 50 min. After extraction, the bre was inserted into the injection port of GC (250 C) to desorb the analytes for 5 min. All analysis was made in triplicate. GCMS analyses were conducted using an Agilent 6890N gas chromatograph equipped with an Agilent 5975 mass selective detector (MSD). The SPME extraction was analysed on a DB-Wax column (60 m 0.25 mm i.d., 0.25 lm lm thickness, J&W Scientic). Each extraction was injected with splitless. The oven temperature was kept at 40 C for 2 min, then raised to 230 C at a rate of 6 C/min, and held at 230 C for 15 min. The column carrier gas was helium at a constant ow rate of 2 mL/min. The mass spectrometer was operated in the electron impact mode (EI) at 70 eV scanning the range 34348 m/z, and the ion source temperature was set at 230 C. Compound identications were made by comparing mass spectral data from Database (Agilent) and conrmed by comparing Kovats RIs to those of the standards. Retention indices (RI) of unknown compounds were calculated in accordance with a modied Kovats method (Cates & Meloan, 1963). 2.10. Quantitative analysis of volatile compounds Synthetic cherry wine solution was achieved by 12 v/v% ethanolwater solution (pH adjusted at 4.10 with lactic acid). 2.10.1. Calibration of standard curves An aliquot of 5 mL of synthetic solution containing different concentrations of volatile standards was diluted stepwisely with synthetic solution using a series of 1:1 dilution. 1.5 g NaCl and 5 lL of 3-octanol (internal standard, 100 mg/L standard solution in ethanol) were placed in 15 mL autosampler vials for each diluted solution. Vials were capped with a Teon septum and an aluminium cap. The HS-SPME and GCMS conditions were set as described in Section 2.9. Selective ion-monitoring (SIM) mass spectrometry was used to quantify the volatile compounds. The ion monitored of IS in the SIM run was m/z 45. The standard curve for individual volatile compounds was built up by plotting the response ratio of target compound and IS against the concentration ratio. The limits of quantication (LOQs) were established for signal to noise ratios of 10. 2.10.2. Calculation of recovery rates The recovery rates of volatile compounds were evaluated in cherry wine. Known amounts of volatile compounds were added to cherry wine, separately. The concentrations of the volatile compounds in cherry wine before and after the volatile additions were quantied by the procedure described above. The recovery rate was calculated as:

determine signicant differences, and the model was statistically signicant with a value of P < 0.05. Principal component analysis (PCA) was applied in order to determine the main sources of variability present in the data sets and to establish the relation between samples (objects) and volatile compounds or polyphenols (variables). 3. Results and discussion 3.1. Fermentation parameters and general wine composition The six S. cerevisiae commercial wine yeast strains including BM44, RA17, RC212, D254, D21 and GRE, were grown under wine fermentation conditions at 20 C. For these strains, fermentation parameters were calculated. All wine fermentations were completed successfully to dryness but differences among the time courses to complete fermentations were observed. The rates of sugar consumption varied among the yeast inocula (data not shown). BM44 and GRE generated the fastest rate of sugar consumption, which reached below 1 g/L of TRS approximately 7 d after inoculation, while RC212 and RA17 needed almost 200 h to complete the fermentation. The other two yeast inocula, i.e., D254 and D21 had intermediate fermentation rates, showing similar maximum growth rates (data not shown). When fermentation was completed, the compositions of these young cherry wines, such as pH, ethanol concentration, titratable acidity, etc., were analysed. As shown in Table 1, a moderate amount of ethanol, ranging from 11.3% to 12.3% (v/v), was detected across the yeast inocula; a small variation of pH values (4.074.22) was recorded among the fermented wines. The amount of volatile acidity also showed slight differences among the yeast inocula. D254 and GRE contained the highest concentrations of volatile acidity of both 0.42 g/L acetic acid, while RC212 and RA17 contained the lowest concentrations of 0.38 and 0.36 g/L acetic acid, respectively. All values were within the ranges acceptable to the wine industry (OIV, 2005). The level of titratable acidity, reported as grams of malic acid per litre of cherry wine, was between 5.79 and 6.71 g/L in these wines, with content of the highest in RC212 and the lowest in GRE. It was worth noticing that a signicant reduction in the titratable acidity was observed during the fermentation in all cherry wines (average 1.62 g/L), indicating that the majority of strains of S. cerevisiae used in the current study possessed the capability to metabolise malic acid, in the presence of glucose or other assimilable carbon sources (Satora, Sroka, DudaChodak, Tarko, & Tuszynski, 2008). Since malic acid is the dominant organic acid in cherries, the introduction of these S. cerevisiae yeasts signicantly reduce the acerb taste of these cherry wines. In addition, the free SO2 percentages varied signicantly among the fermented wines, with the lowest amount being found with RC212 and D254 (13.3 and 13.7 mg/L), respectively, and the highest with GRE (16.5 mg/L). Conversely, GRE wine had the lowest level of total soluble solids in comparison to those observed from other wines.

Recovery rates

detected amount after addition detected amount before addition 100% added amount

2.11. Statistical analysis Statistical analyses were performed using the SPSS version 16.0 statistical package for windows (SPSS Inc., Chicago, Illinois). ANOVA and Duncans multiple range tests were applied to the data to

3.2. HS-SPME-GCMS analysis of volatile compounds 3.2.1. Calibration curves and performance characteristics In the synthetic cherry wine solution, nine levels of concentration were tested in duplicate and regression lines were calculated

550

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

Table 1 Composition of young cherry wines inoculated by various yeasts. BM 44a Fermentation time (h) Alcohol (% v/v) Total reducing sugar (g/L) pH Titratable acidityb (g/L) Volatile acidity (g/L) Total soluble solids (g/L) Free SO2 (mg/L) Total phenolics (mg/L) Tanins (mg/L) Total anthocyanins (mg/L)
a b

RA17 192 11.6 0.1bc 0.7 0.1a 4.13 0.01b 5.94 0.04b 0.36 0.02a 19.9 0.2c 14.7 0.3b 584.3 23.5a 788.1 19.5b 19.8 1.1a

RC212 192 11.7 0.0c 0.8 0.0ab 4.07 0.01a 6.71 0.03f 0.38 0.01ab 21.7 0.1e 13.3 0.2a 592.1 30.2a 740.5 19.2a 27.9 1.3c

D254 180 11.5 0.1b 0.9 0.0bc 4.12 0.00b 6.25 0.03c 0.42 0.00c 18.7 0.2b 13.7 0.4a 670.6 24.5b 860.3 22.2c 25.6 1.0b

D21 180 11.3 0.1a 1.0 0.1c 4.16 0.00c 6.63 0.02e 0.40 0.01bc 20.6 0.1d 15.4 0.2c 679.8 14.2b 864.6 24.1c 18.3 0.9a

GRE 168 12.1 0.0d 0.9 0.1bc 4.25 0.02d 5.79 0.01a 0.42 0.02c 17.6 0.1a 16.5 0.2d 642.4 22.2b 856.6 22.7c 17.9 0.8a

168 12.3 0.1e 0.9 0.0bc 4.16 0.01c 6.37 0.02d 0.41 0.01bc 18.5 0.2b 15.8 0.3c 742.7 20.1c 900.7 16.4d 24.4 1.2b

Values with different superscript roman letters (af) in the same row are signicantly different according to the Duncan test (P < 0.05). Expressed as g/L of malic acid.

for 44 compounds using the method of HS-SPME combined with GCMS. Linear range covered the volatile compound concentrations expected in the samples and a good regression coefcient (R2 P 0.99) was achieved in all standard curves. Detection and quantication limits calculated from the baseline of a standard solution chromatogram (as 3 and 10 times the noise of the base line, respectively) were good. Several compounds had a higher detection limit because of their poorer afnity with SPME bre, such as acetic acid, 1-propanol and 2-methylpropanol. The most compounds analysed showed very good recoveries (80110%) with relative standard deviation (RSD) less than 20%. Thus, the method proposed was successfully applied to experimental cherry wines. 3.2.2. Identication of volatile compounds Table 2 shows the volatile prole identied in the cherry wines fermented with different S. cerevisiae yeast strains. A total of 44 key volatile compounds were identied and quantied, including 14 esters, 12 alcohols, 1 aldehyde, 7 acids, 6 terpenes and 2 C13-norisoprenoids, and 2 phenols. Alcohols were the most abundant volatile group in the tested cherry wines. Normally, alcohols arise from sugar catabolism, as well as from decarboxylation and deamination of amino acids; and small amounts are produced by yeast through the reduction of the corresponding aldehydes (Hernandez-Orte et al., 2008). Five alcohols, i.e., 1-propanol, 2-methylpropanol, 3-methylbutanol, benzyl alcohol, and 2-phenylethyl alcohol were identied as major alcohol compounds (concentration P 10 mg/L) in all tested samples. Of these, 3-methylbutanol, giving a fusel-like odour (Molina et al., 2009), covered the highest concentration. Its largest amount (69.99 mg/L) was found in the D254 cherry wine. This strain, suggested for a wide range of fermentation temperatures and low formation of foams, has been used in wine-making to bring out relatively neutral avour. 2-Methylpropanol, also responsible for fusel-like avour (Molina et al., 2009), was detected in its highest amount (40.38 mg/L) with D254 as well. 1-Propanol, generating an ethanol-like avour, was found in rather high levels ranging from 54.89 mg/L (GRE) to 14.82 mg/L (RC212). The yeast GRE has been used where high green and vegetable aromas are in need of reduction, and RC212 is used in wine preparation when the winemaker desires to bring out a strong sense fruity avours. In addition, benzyl alcohol and 2-phenylethyl alcohol were two important benzene derivatives in cherry wine samples. 2-Phenylethyl alcohol is an important metabolic byproduct by yeast fermentation, and it has a typical rose-like avour (Ledauphin, Guichard, Saint-Clair, Picoche, & Barillier, 2003). Comparatively, the yeasts of D254 and D21 should have higher yield ability of 2phenylethyl alcohol according to the higher concentration (41.21 and 32.99 mg/L, respectively). As similar performance as GRE, D21 is also used in fermentation to reduce high green and vegeta-

ble aromas. Benzyl alcohol, having a oral-like avour (Sun, Jiang, & Zhao, 2010), accounted for 17.44% (RA17) to 12.71% (BM44) of total volatile alcohols. Overall, D254 generated the highest content of total volatile alcohols (204.48 mg/L), due to the presences of the highest concentrations of 2-methylpropanol, 3-methylbutanol, 1-hexanol, benzyl alcohol, and 2-phenylethyl alcohol, and then followed GRE, D21, BM44, RA17, and RC212. The acid produced in the highest relative amounts by all six yeasts was acetic acid. It is produced during alcoholic and malolactic fermentations (Mato, Suarez-Luque, & Huidobro, 2005). Based on the results of Duncans multiple range tests, signicant difference in the content of acetic acid was observed among all tested wines, with the exception of those between BM44 and D21. Its largest concentration (54.39 mg/L) was found in RA17 wine, and the lowest (29.92 mg/L) was detected with RC212. This result was in good accordance with that obtained from volatile acidity (VA) of the six cherry wines. Hexanoic acid is also one of the major acids found in cherry wines, which contributes a sweaty odour (Sun et al., 2010). In addition to the above two acids, octanoic and decanoic acids had higher concentrations in all samples as well. From the quantitative point of view, the total concentrations of volatile acids in a stepwise decreasing manner were RA17, GRE, D254, D21, BM44, and RC212. Esters were mostly formed from the esterication of alcohols with fatty acids during the fermentation and ageing process (Comuzzo, Tat, Tonizzo, & Battistutta, 2006). In the current study, among the 14 esters quantied, ethyl lactate was the most important volatile ester in all samples. This compound appears mainly in the malo-lactic fermentation during wine vinication (Fang & Qian, 2005), accounting for 54.68% (BM4 4) $77.95% (GRE) of total esters. Ethyl acetate was another major ester in the six cherry wines; its largest concentration was also found in RA17 cherry wine, while the lowest was with RC212. Ethyl acetate is mainly formed during alcoholic fermentation (Gurbuz, Rouseff, & Rouseff, 2006). The other important esters (concentration P 100 lg/L) found in most samples included isopentyl acetate (banana-like avour), ethyl hexanoate (fruity-like avour), hexyl acetate (sweet fruity-like avour), ethyl 3-hexenoate (fruity-like avour), ethyl octanoate (cooked fruity-like avour), ethyl decanoate (fruity-like avour), ethyl benzoate(oral avour), and diethyl succinate (fruity-like avour) (Fang & Qian, 2005; Gurbuz et al., 2006). The presence of many ethyl esters might be indicative of fruity-berry avour in cherry wine. RA17 strain is an ethanol-tolerant, oral yeast that is used in making wines. It produced the greatest concentrations of isopentyl acetate, ethyl hexanoate, hexyl acetate, ethyl octanoate, ethyl decanoate, ethyl benzoate, and diethyl succinate among six yeasts. According to the results of statistics analysis, ethyl decanoate had signicant differences among all tested samples; ethyl octanoate had signicant differences in most wines with

Table 2 Concentrations of volatile compounds in six cherry wines (in micrograms per litre, n = 3). RI DB-wax Compound BM 44a Average Esters 897 1057 1119 1133 1229 1269 1296 1337 1430 1635 1655 1673 1791 1808 Alcohols 1035 1082 1142 1199 1241 1245 1348 1358 1377 1542 1553 1874 1902 Acids 1451 1562 1626 1668 1846 2057 2270 Aldehydes 1508 Ethyl acetate Ethyl 3-methylbutanoate Isopentyl acetate Ethyl pentanoate Ethyl hexanoate Hexyl acetate Ethyl 3-hexenoate Ethyl lactate Ethyl octanoate Ethyl decanoate Ethyl benzoate Diethyl succinate Ethyl 2-phenylacetate 2-Phenylethyl acetate Totals 1-Propanol 2-Methylpropanol 1-Butanol 3-Methylbutanol 3-Methyl-3-buten-1-ol 1-Pentanol 1-Hexanol (E)-3-hexen-1-ol (Z)-3-hexen-1-ol 2,3-Butanediol 1-Octanol Benzyl alcohol 2-Phenylethyl alcohol Totals Acetic acid 2-Methylpropanoic acid Butanoic acid 3-Methylbutanoic acid Hexanoic acid Octanoic acid Decanoic acid Totals Benzaldehyde Totals 1603.69b 14.73a 600.30b 21.68a 92.49a 166.88c 82.90a 5079.19a 755.97d 177.51b 567.06c 81.99a 13.88a 30.81b 9289.08 29012.21c 17960.31a 753.42a 49457.21d 1535.91a 16.57d 1760.20c 83.01b 114.74a 256.80d 31.80b 17085.16a 16352.18a 134419.51 35178.99c 845.31a 284.30a 401.04a 18169.85a 5364.62a 3973.27a 64217.39 118.64b 118.64 18.95b 16.45a 24.37a 374.19c 7.73a 48.32b SD 180.18 0.74 32.02 1.58 5.62 6.34 4.15 203.96 31.30 9.18 26.35 4.30 0.70 1.40 RA17 Average 1604.94b 5.82a 773.29c 21.02a 440.40d 177.66c 95.66bc 15947.34d 1516.53e 459.40f 578.20c 159.95b <q.l. 40.82c 21821.03 19954.15b 27504.27b 743.55a 31975.91b 1540.73a 12.25a 1928.66d 84.77b 118.61a 521.75f 51.85d 22335.31c 21282.40b 128054.23 54385.89e 1642.18c 450.10b 499.46b 24160.41d 7374.16c 5488.01c 94000.22 248.83d 248.83 35.95d 18.89b 113.67b 426.05d 13.36c 79.80d SD 172.25 0.39 35.66 1.01 20.02 7.88 5.63 597.37 65.83 20.97 27.91 8.00 2.41 RC212 Average 1051.53a 7.60a 437.58a 31.92b 153.86b 115.42a 92.62bc 9143.69b 227.11a 281.02d 433.17b 88.02a 13.72a 25.45a 12102.70 14819.07a 28619.28c 906.01b 44395.52c 1529.50a 13.61ab 996.43a 66.74a 286.58c 111.99a 36.76c 20234.58b 15334.75a 127350.81 29923.24a 864.13a 263.15a 428.87a 18395.21a 5558.18a 4939.76b 60372.54 370.06e 370.06 26.58c 17.63ab 124.60c 193.01b 15.49d 73.17c SD 122.58 0.58 29.88 1.40 8.69 5.07 4.63 357.18 12.36 13.15 22.66 4.50 0.86 1.27 D254 Average 1377.99b 7.94a 569.04b 29.85b 161.06b 121.58ab 92.99bc 9521.22b 519.34b 309.35e 375.60a 87.07a 16.19b 33.04b 13222.27 18609.95b 40380.69e 875.16b 69986.35e 1536.13a 16.39d 2262.30e 85.88b 278.79c 187.00b 36.88c 29012.53d 41207.82d 204475.89 32615.61b 1508.19b 267.15a 495.20b 22779.60 cd 7569.99c 5150.46b 70386.19 369.81e 369.81 28.65c 16.45a 139.04d 442.20d 11.20b 51.61b SD 98.90 0.40 28.45 1.89 7.95 6.08b 6.63 406.06 21.97 14.47 19.78 4.25 0.96 1.18 D21 Average 1377.45b 141.68b 425.08a 20.37a 346.68c 130.28b 102.57c 4668.30a 272.08a 65.94a 558.41c 86.26a 15.40b 41.71c 8252.20 41317.55d 27668.07b 733.94 a 30483.97b 1542.03a 15.61 cd 2141.51e 65.08a 129.28a 225.26c 26.11a 22882.34c 31992.92c 159223.66 36176.64c 845.78a 263.30a 472.51b 21376.89bc 7156.39c 3819.04a 70110.56 88.33a 88.33 15.06a 17.19ab 19.37a 417.56d 8.39a 32.60a SD 108.87 6.08 20.25 2.02 17.33 6.91 5.63 203.42 14.60 3.90 27.02 5.31 0.80 2.86 GRE Average 1595.21b 234.60c 548.20b 108.95c 92.87a 125.23ab 90.60ab 15032.73c 611.40c 213.75c 498.47c 89.89a 13.65a 30.30b 19285.85 54894.93e 30540.85d 2053.32c 25448.17a 1526.85a 14.61bc 1271.25b 109.28c 200.22b 290.03e 39.34c 22352.26c 20831.64b 159572.76 46976.02d 888.18a 291.20a 412.17a 20769.52b 6439.33b 4964.50b 80740.92 171.17c 171.17 28.72c 17.34ab 116.10bc 118.41a 10.15b 68.10c SD 129.76 21.73 29.41 5.05 4.64 4.26 4.13 651.64 29.57 12.69 24.92 4.39 0.83 1.89 S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

450.61 598.02 7.67 1272.43 56.80 1.13 88.01 4.15 6.74 13.84 1.99 804.26 901.52

1297.71 375.21 57.18 300.40 70.04 0.71 96.43 4.24 6.93 25.09 2.79 836.77 855.82

740.95 430.96 40.30 219.89 66.48 0.78 49.82 4.34 15.33 7.60 1.84 811.73 915.35

530.50 219.03 33.76 298.66 71.81 1.02 113.12 5.29 14.94 10.35 2.84 1050.63 1912.08

665.88 383.40 16.70 1402.10 70.10 0.88 107.08 4.25 6.96 10.86 1.51 960.64 1095.93

744.75 527.04 112.67 1072.20 66.34 0.93 63.56 6.46 10.91 13.50 2.00 817.61 1002.32

758.95 62.27 15.22 20.15 708.49 268.23 128.66

1319.29 89.11 20.51 24.87 908.02 368.71 174.40

896.16 47.21 14.16 23.44 719.76 277.91 146.99

1030.78 70.41 13.96 24.16 1028.98 378.50 157.52

808.83 41.29 14.17 23.23 868.84 357.82 90.95

1348.80 46.41 14.06 21.61 938.48 321.97 148.22

5.93

12.44

18.50

18.49

4.42

8.56

Terpenes and C13-norisoprenoids 1546 Linalool 1604 4-Terpineol 1695 a-Terpineol 1764 b-Citronellol 1512 Nerol 1808 b-Damascenone

1.95 1.52 2.22 16.71 0.87 2.47

2.80 1.34 5.68 19.30 0.80 3.69

2.33 0.18 6.03 10.65 0.74 3.16

1.87 0.62 6.25 20.11 0.60 2.78

0.91 0.88 1.20 21.02 0.20 1.81

1.84 1.07 5.40 6.20 0.53 3.45

(continued on next page) 551

552

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

Different capital letter in the same row represents signicant difference at P < 0.05 by Duncans multiple range test.<q.l., below quantication limit.

SD

the exception of those between RC212 and D21. RA17 wine had the largest concentration of total volatile esters (136.52 mg/L), nearly three times higher than those contained in D21, suggesting that RA17 strain possessed a relatively higher capacity in ester synthesis. And the GRE, D254, RC212 and BM44 cherry wines contained an intermediate amount of them. Terpenes and norisoprenoids are mainly released from their nonodorous glycosides during alcoholic fermentation stage (Skouroumounis & Sefton, 2001). b-Citronellol and geraniol were the major terpenes in the tested samples with fruity and oral-like avour (Fang & Qian, 2005). The largest concentration of b-citronellol was found with D254 (442.20 lg/L), and the greatest amount of geraniol was found in GRE (386.29 lg/L). b-Damascenone could be important in cherry wines, due to its low threshold level (0.009 lg/L in water) (Ferrari et al., 2004). This compound mainly comes from degradation of carotenoids with a honey-like odour (Fang & Qian, 2005; Skouroumounis & Sefton, 2001). Its largest concentration was found with RA17 (79.80 lg/L), and the lowest was with D21 (32.60 lg/L). The RA17 samples contained a relatively high level of total terpenes and norisoprenoids (1019.37 lg/L), whereas D21 and BM44 wines had a signicantly lower level (676.25 and 676.47 lg/L, respectively). The former observation was associated with greatest concentrations of b-citronellol (426.05 lg/L), geraniol (331.68 lg/L) and b-damascenone (79.80 lg/L) contained in RA17 wines, suggesting that RA17 strain had a relative stronger ability in the hydrolysis of glycosides. Benzaldehyde was the only aldehyde analysed in six yeasts. Its largest concentration was found in D254 wine (370.06 lg/L), while its lowest level (88.33 lg/L) was with D21. This compound showed signicant differences among most cherry wines, except for those between RC212 and D254. Two phenols, i.e., phenol and 4-ethylphenol, were quantied in the tested wines. 4-Ethylphenol is usually considered as an off-avour compound in wine (Martorell, Marti, Mestres, Busto, & Guasch, 2002). Its largest concentration (966.50 lg/L) was found in RA17 wine, and the lowest value (61.20 lg/L) was with D21. D254 produced the greatest amount of phenol (241.27 lg/L) among the six yeast strains, whereas D21 did not produce any detectable levels. 3.3. Polyphenols in cherry wine 3.3.1. Total phenolics The nal composition of phenolic compounds in cherry wines depends on many factors; phenolics originally contained in fruit, must extraction, wine production, and chemical reactions taking place during wine ageing are four major aspects (Czyzowska & Pogorzelski, 2002). In the current investigation, most of the procedures proceeded simultaneously and consistently, with the exception of fermenting inocula, thus generating a considerable variation among the tested wines (584.3742.7 mg GAE/L). As seen in Table 1, BM44 fermented wine was found to have the highest total phenolics among the six wines tested, whereas the lowest was observed in the case of RA17. BM44 has been recommended for the stabilization of polyphenols in wines and reduction of astringency caused by tannins. D254 and D21 had about 13% and 15% higher of total phenolics than that of RC212, respectively. Previous studies have shown that the phenolics can provide healthimproving benets due to various biological activities, especially the antioxidant properties (Kim et al., 2005; Yoo et al., 2010). All tested wines contained high amounts of phenolics, which indicated that high antioxidant activity was contained in these wines. During our experiment, it was found that a detectable drop of almost 20% in phenolics content was recorded right after the wine fermentation, probably due to the adsorption of phenolics onto yeast cell walls and the reaction with cell wall proteins. Researches conducted by Czyzowska and Pogorzelski (2002) demonstrated a

SD

386.29d 745.10

18.31

GRE

Average

SD

9.08

D21

Average

166.07a 676.25

SD

221.73b 910.88

D254

Average

SD

363.86d 814.34

17.06

RC212

Average

14.82

RA17

Average

SD

186.45a 676.47

11.26

BM 44a

Average

Compound

Table 2 (continued)

RI DB-wax

Phenols 2035 2189

1857

Phenol 4-Ethylphenol Totals

Geraniol Totals

152.74a 61.20a 213.94

7.64 3.26

193.53b 966.50d 1160.04

331.65c 1019.37

9.98 38.33

194.79b 353.88b 548.66

10.74 15.69

241.27c 557.09b 798.36

11.86 21.85
a

10.09

<q.l. 66.03a 66.03

3.90

204.57b 76.89a 281.46

9.03 4.54

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

553

similar result. In the wines tested by them, the total content of polyphenols was decreased by approximately 25% when the fermentation was completed. Tannin, a typical phenolic compound, can markedly affect both the avour and the astringency of wines. Although tannin content has been shown to be closely associated with astringency, it does not impart detectable astringency to cherry juice when its concentration is lower than 1 g/L (Bernalte, Hernandez, Vidal-Aragon, & Sabio, 1999). In this study, tannins were found in moderate levels, ranging from 900.7 to 740.5 mg/L, with an average amount of 835.1 mg/L. They covered relative higher amount in the cherry wines resulting from BM44; whereas signicant lower concentration was found with RC212. 3.3.2. Total anthocyanins Anthocyanins are colouring compounds located within the cherry esh and skins, and are responsible for red colour. Considering the fact that the colour of wine is one of the most important indicators for product acceptability, thus, from the quantitative and qualitative point of view, anthocyanins are both aesthetically and economically important (Chaovanalikit & Wrolstad, 2004b). Table 1 shows the content of total anthocyanins in the six cherry wines resulting from different yeasts, evaluated by the colorimetric method. As observed on total phenolics, a wide range of concentrations of total anthocyanins was revealed, varying from 17.9 to 27.9 mg/L, with an average amount of 22.3 mg/L. RC212 fermented wine had the highest content of total anthocyanins, followed by D254. The two strains have been recommended for the elaboration of wines with a marked promotion of stabilization of the colour during fermentation. RA17 and GRE were the two lowest anthocyanins producers, and both were about two thirds of RC212. Total anthocyanins in Danube cherry wines have been reported to be 29.6 mg/L (Yoo et al., 2010), and our results were comparable with theirs. 3.3.3. HPLC analysis of polyphenols Polyphenols of cherry wines were fractionated into anthocyanins and nonanthocyanins for the HPLC analysis by the method introduced by Kim et al. (2005). According to literatures (Chandra, _ Nair, & Iezzoni, 1992; Czyzowska & Pogorzelski, 2004; Kim et al., 2005), the following anthocyanins are present in sour cherry fruits: cyanidin 3-glucoside, cyanidin 3-rutinoside, cyanidin 3-glucosylrutinoside, delphinidin 3-glucoside, delphinidin 3-rutinosid, peonidin 3-glucoside, peonidin 3-rutinoside as well as malvidin 3glucoside. Most of these phenolic compounds passed from the fruits to wines and remained active, but their prole changed and partially degraded during disintegration of raw material, fermentation, and wine aging (Chaovanalikit & Wrolstad, 2004a). In the current study, as seen in Table 3, all cherry wines resulting from different yeasts commonly contained four anthocyanins, i.e., cyanidin 3-glucoside, cyanidin 3-rutinoside, cyanidin 3-glucosylrutinoside and peonidin 3-rutinoside, with cyanidin 3-glucosylrutinoside as the major anthocyanin, composing 57.161.4% of total anthocyanins quantied by HPLC, followed by cyanidin 3-rutinoside (10.715.1 mg/L). However, the content of individual anthocyanin differed signicantly among the yeast strains, especially in the

case of the two major anthocyanins. In specic, the amount of cyanidin 3-glucosylrutinoside was almost 1.5 times higher in RC212 fermented wine than in R17. And a similar phenomenon occurred in the case of cyanidin 3-rutinoside, with the largest amount being found in RC212 (15.1 mg/L) and the least in RA17 (10.7 mg/L). Previously, in cherry wines tested by Czyzowska and Pogorzelski (2002), the content of cyanidin 3-glucosylrutinoside produced from utwka was 18.4 mg/L, indicating that this compound studied here was slightly higher than that reported. Furthermore, peonidin 3-rutinoside was found in moderate levels, with mean concentrations around 1.31 mg/L in all tested wines. Cyanidin 3-glucoside was found in very low levels, and its contribution to total anthocyanins was less than 2%. In addition, the content of total anthocyanins quantied by HPLC in RA17 wine amounted to 29.7 mg/L, while in RC212 it was 42.8 mg/L, which suggests that RC212 was the most capable in the extraction of anthocyanins from cherry esh and skins. The tendency of this result was in good agreement with that obtained from the colorimetric assays. With regard to phenolic acids, previous studies (Czyzowska and Pogorzelski, 2002) have shown that the individual phenolic acid contained in tart cherries includes neochlorogenic, chlorogenic, p-coumaric, caffeic, as well as ferulic acid. These compounds are susceptible to enzymatic browning reactions when they are contained in the cherry fruit, especially caffeic acid derivatives (Czyzowska and Pogorzelski, 2002). In our investigation, ve individual phenolic acids were identied and quantied, i.e., chlorogenic, neochlorogenic, p-hydroxybenzoic, p-coumaric, and caffeic acid (Table 4). But the content varied signicantly among the wines. Chlorogenic acid was present in the highest amount in all tested cherry wines (21.729.6 mg/L), followed by neochlorogenic acid (12.818.5 mg/L). It has been reported that neochlorogenic and chlorogenic were the major phenolic acids in utwka cherry wines (Czyzowska and Pogorzelski, 2002), and the levels were 44.5 and 61.1 mg/L, respectively, which were evidently higher than these studied here. p-Hydroxybenzoic acid had an intermediate concentration in most tested wines, with a mean concentration of around 4.31 mg/L. And a similar result was obtained in the case of p-coumaric acid which also had a moderate amount in most tested wines, with the exception of BM 44 and D21, of which this compound covered a concentration of 6.95 and 6.63 mg/L, respectively, almost two times higher than that in other wines. Evidently, the highest content of total amount of phenolic acids was achieved in BM44 wine (61.08 mg/L), and the lowest was with D254 (43.41 mg/L). From these results, it was concluded that RC212 and D254 were relative incapable in the retention and production of phenolic acids; whereas, RA12 and D21 showed intermediate capability. 3.4. Principal component analysis 3.4.1. PCA of volatile compounds When PCA was applied to normalise the relative amounts of 44 volatile compounds and 18 cherry wines, the rst two principal compounds (PC) were extracted explaining 61.15 cumulative percent (cum.%) value of the total variance of the initial data set.

Table 3 Individual anthocyanin in the cherry wines standard deviation (n = 3) measured in milligrams per litre. Anthocyanins Cyanidin 3-glucoside Cyanidin 3-rutinoside Cyanidin 3-glucosylrutinoside Peonidin 3-rutinoside
a

BM 44a 0.45 0.02b 11.4 0.55ab 20.8 1.03bc 1.22 0.06b

RA17 0.36 0.01a 10.7 0.52a 17.6 0.88a 1.08 0.05a

RC212 0.68 0.03e 15.1 0.76d 25.4 1.27e 1.64 0.08d

D254 0.56 0.03d 14.6 0.71d 23.3 1.16d 1.52 0.07c

D21 0.51 0.02c 12.3 0.62bc 21.5 1.09 cd 1.31 0.06b

GRE 0.53 0.02 cd 12.8 0.64c 19.2 0.97ab 1.06 0.05a

Values with different superscript roman letters (ae) in the same row are signicantly different according to the Duncan test (P < 0.05).

554

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

Table 4 Phenolic acids in the cherry wines standard deviation (n = 3) measured in milligrams per litre. Phenolic acid Chlorogenic Neochlorogenic p-Hydroxybenzoic p-Coumaric Caffeic
a

BM 44a 29.6 1.49c 18.5 0.94c 4.31 0.23ab 6.95 0.36c 1.72 0.09f

RA17 26.3 1.32b 14.9 0.75b 4.26 0.21a 4.05 0.21b 1.06 0.05d

RC212 23.5 1.17a 14.1 0.71ab 4.02 0.21a 3.52 0.17a 0.68 0.03b

D254 21.7 1.09a 12.8 0.64a 4.69 0.23b 3.28 0.15a 0.94 0.05c

D21 28.4 1.43bc 17.3 0.88c 4.36 0.22ab 6.63 0.34c 1.35 0.07e

GRE 22.4 1.12a 13.2 0.65a 4.21 0.21a 3.33 0.17a 0.42 0.02a

Values with different superscript roman letters (af) in the same row are signicantly different according to the Duncan test (P < 0.05).

The observation of the loading scores suggests that 29 redundant variables having coefcients magnitude < 0.8 are insufcient to adequately describe the samples according to yeast strains, and was removed from the matrix. And the new set with 15 volatiles (data matrix 18 15) accounting for 87.98 cum.% of the total variance was used. As seen in Fig. 1, the rst principal component (PC1) is plotted against the second (PC2), and the separation among different wine samples from this PC1PC2 scatter point plot is obvious. The corresponding loadings plot that establishes the relative importance of each variable and it is therefore useful for the study of relations among the volatile compounds and relations between volatiles and cherry wines. The cherry wines produced by RA17 and GRE were located at area A, the rst quadrant of scatter point plot. In comparison with GRE, RA17 wine was more inuenced by the variables related with PC1. Isopentyl acetate (0.896), hexyl acetate (0.870), ethyl octanoate (0.968), acetic acid (0.891), 2,4-butanediol (0.988), butanoic acid (0.975), and diethyl succinate (0.899) were highly oriented towards positive PC1. Wine samples from RC212 and D254 were located at the second quadrant; and benzaldehyde, linalool, aterpineol, nerol, b-damascenone, phenol, and decanoic acid are strongly associated with this quadrant. BM44 and D21 cherry wines were located at area C. D21 was in the third quadrant of scatter point plot, and only ethyl 2-phenylacetate is in accordance with this requisite. 3.4.2. PCA of polyphenols To establish which phenolic compounds were more valuable to differentiate the samples according to the yeast strains, principal component analysis was also carried out. From the nine phenolics identied present in cherry wine samples, eight compounds were selected, with the exception of p-hydroxybenzoic acid (coefcients magnitude < 0.8). When these variables were used to classify cherry wines, a data matrix (18 8) accounting for 94.64 cum.% of the total variance in the data was generated, where PC1 explains

RC212(1) D254(3)

2.0000 1.0000

RC212(2)

RC212(3) V9 V5 V8 V6 V8 V7 V7 V6 V3 D21(2) D21(1) BM 4 4(2) V4 V2 V3 A BM 4 V4 V5 V1 D21(3)1.0000 BM 4 4(3)

PC2(45.61%) PC1(28.90%)

D254(2) D254(1)

-2.0000

GRE(3) GRE(1)

-1.0000

GRE(2) RA17(3)

0.0000 0.0000 -1.0000

2.0000

4(1)

RA17(1)

RA17(2)

-2.0000 PC1(55.81%) PC1(49.03%)

Fig. 2. PC1 vs. PC2 scatter point plot of cherry wines resulting from six yeasts and eight polyphenols by PCA. Note: V1, chlorogenic acid; V2, neochlorogenic acid; V3, p-coumaric acid; V4, caffeic acid; V5, cyanidin 3-glucoside; V6, cyanidin 3rutinoside; V7, cyanidin 3-glucosylrutinoside; V8, peonidin 3-rutinoside.

2.000
RC212(3) V11 V15 RC212(2) D254(3) 1.000 V12 RC212(1) V7 V3 V14 V13 D254(1) V9 GRE(2) GRE(1) D254(2) V8 V1 V3 GRE(3) 0.000 V4 V6 V10

49.03 cum.% of the variance in the initial data set and PC2 explains 45.61 cum.%. The projections of the samples along the directions identied by the rst two PCs, is reported in Fig. 2, where PC1 of wine samples is plotted against PC2. According to the result, the six cherry wines were mainly classied into three groups: (1) BM44 and D21, (2) RC212 and D254, and (3) RA17 and GRE, which was interestingly in coincidence with that obtained from volatile compounds. Furthermore, the coefcient that denes the weight of the original variable in the PCs was investigated to understand which phenolic compounds are responsible for the ranking of wine samples. Four phenolic acids, including chlorogenic acid, neochlorogenic acid, p-coumaric acid, and caffeic acid were highly oriented towards positive PC1, whereas cyanidin 3-glucoside, cyanidin 3rutinoside, cyanidin 3-glucosylrutinoside, and peonidin 3-rutinoside were highly oriented towards positive PC2. Therefore, the cherry wines from BM44 and D21 were most inuenced by four phenolic acids related with PC1, and they both had higher concentrations of phenolic acids. Cherry wines produced by RC212 and D254 were located at the second quadrant, and four anthocyanins are strongly associated with this quadrant. The cherry wines from RA17 and GRE were located at the third quadrant (negative PC1 and PC2). 4. Conclusion

PC2(38.02%)

RA17(1)

A
1.500

RA17(3)

RA17(2)

-1.500

-0.500

0.500

D21(3)

-1.000

BM 4 4(1) BM 4 4(3) BM 4 4(2)

V2

2.500

D21(1)

D21(2)

-2.000 PC1(49.96%)

Fig. 1. PC1 vs. PC2 scatter point plot of cherry wines resulting from six yeasts and 15 volatile compounds by PCA. Note: V1, isopentyl acetate; V2, hexyl acetate; V3, ethyl octanoate; V4, acetic acid; V5, benzaldehyde; V6, 2,4-butanediol; V7, linalool; V8, butanoic acid; V9, diethyl succinate; V10, ethyl 2-phenylacetate; V11, aterpineol; V12, nerol; V13, b-damascenone; V14, phenol; V15, decanoic acid.

In the current study, cherry wines resulting from six commercial S. cerevisiae strains using fruits of tart cherry variety Early Richmond were characterised in the proles of volatile compounds and polyphenols. It was found the general wine compositions differed insignicantly in most analysed items, but the level of titratable acidity showed evident variations. When PCA was used to classify the cherry wines according to the aroma characteristics and phenolic compounds, interesting results were obtained. RA17 and GRE were found capable to produce higher amount of volatile esters and acids, and isopentyl acetate, hexyl acetate, ethyl octanoate, acetic acid, and butanoic acid were identied to be associated with the above two wines. Cherry wines from RC212 and

S.Y. Sun et al. / Food Chemistry 127 (2011) 547555

555

D254 were characterised by a higher content of volatile aldehydes and terpenes, and more inuenced by benzaldehyde, linalool, aterpineol, nerol and b-damascenone; in the meantime, both strains showed excellent capability in the extraction of anthocyanins. BM44 and D21 wines were located at same area, with ethyl 2phenylacetate in accordance, and were both extraordinary in the retention and production of phenolic acids. Acknowledgement This work was nancially supported by the Doctoral Research Foundation of Shandong Province (BS2010NY007), and the Doctoral Foundation of Ludong University (LY2010002). References
Bernalte, M. J., Hernandez, M. T., Vidal-Aragon, M. C., & Sabio, E. (1999). Physical, chemical, avor and sensory characteristics of two sweet cherry varieties grown in Valle Del Jerte (Spain). Journal of Food Quality, 22, 403416. Cates, V. E., & Meloan, C. E. (1963). Separation of sulfones by gas chromatography. Journal of Chromatography A, 11, 472478. Chandra, A., Nair, M. G., & Iezzoni, A. (1992). Evaluation and characterization of the anthocyanin pigments in tart cherries (Prunus cerasus L.). Journal of Agricultural and Food Chemistry, 40, 967969. Chaovanalikit, A., & Wrolstad, R. E. (2004a). Anthocyanin and polyphenolic composition of fresh and processed cherries. Journal of Food Science, 69, FCT73FCT83. Chaovanalikit, A., & Wrolstad, R. E. (2004b). Total anthocyanins and total phenolics of fresh and processed cherries and their antioxidant properties. Journal of Food Science, 69, FCT67FCT72. Comuzzo, P., Tat, L., Tonizzo, A., & Battistutta, F. (2006). Yeast derivatives (extracts and autolysates) in winemaking: Release of volatile compounds and effects on wine aroma volatility. Food Chemistry, 99, 217230. Czyzowska, A., & Pogorzelski, E. (2002). Changes to polyphenols in the process of production of must and wines from blackcurrants and cherries. Part I. Total polyphenols and phenolic acids. European Food Research and Technology, 214, 148154. _ Czyzowska, A., & Pogorzelski, E. (2004). Changes to polyphenols in the process of production of must and wines from blackcurrants and cherries. Part II. Anthocyanins and avanols. European Food Research and Technology, 218, 355359. Fang, Y., & Qian, M. (2005). Aroma compounds in Oregon Pinot Noir wine determined by aroma extract dilution analysis (AEDA). Flavour and Fragrance Journal, 20, 2229. Ferrari, G., Lablanquie, O., Cantagrel, R., Ledauphin, H., Payot, T., Fournier, N., et al. (2004). Determination of key odorant compounds in freshly distilled Cognac using GCO, GCMS, and sensory evaluation. Journal of Agricultural and Food Chemistry, 52, 56705676. Gurbuz, O., Rouseff, J. M., & Rouseff, R. L. (2006). Comparison of aroma volatiles in commercial Merlot and Cabernet Sauvignon wines using gas chromatography olfactometry and gas chromatographymass spectrometry. Journal of Agricultural and Food Chemistry, 54, 39903996.

Han, L. X., Huang, Z. G., Zhao, G. R., Li, M., Qi, X. J., & Li, Y. H. (2008). Development status and prospect of sweet cherry industry of China. China Fruits, 1, 5860. Hernandez-Orte, P., Cersosimo, M., Loscos, N., Cacho, J., Garcia-Moruno, E., & Ferreira, V. (2008). The development of varietal aroma from non-oral grapes by yeasts of different genera. Food Chemistry, 107, 10641077. Kim, D.-O., Heo, H. J., Kim, Y. J., Yang, H. S., & Lee, C. Y. (2005). Sweet and sour cherry phenolics and their protective effects on neuronal cells. Journal of Agricultural and Food Chemistry, 53, 99219927. Ledauphin, J., Guichard, H., Saint-Clair, J. F., Picoche, B., & Barillier, D. (2003). Chemical and sensorial aroma characterization of freshly distilled Calvados. 2. Identication of volatile compounds and key odorants. Journal of Agricultural and Food Chemistry, 51, 433442. Lee, K. W., Hur, H. J., Lee, H. J., & Lee, C. Y. (2005). Antiproliferative effects of dietary phenolic substances and hydrogen peroxide. Journal of Agricultural and Food Chemistry, 53, 19901995. Martorell, N., Marti, M. P., Mestres, M., Busto, O., & Guasch, J. (2002). Determination of 4-ethylguaiacol and 4-ethylphenol in red wines using headspace-solid-phase microextraction-gas chromatography. Journal of Chromatography A, 975, 349354. Mato, I., Suarez-Luque, S., & Huidobro, J. F. (2005). A review of the analytical methods to determine organic acids in grape juices and wines. Food Research International, 38, 11751188. Molina, A., Swiegers, J., Varela, C., Pretorius, I., & Agosin, E. (2007). Inuence of wine fermentation temperature on the synthesis of yeast-derived volatile aroma compounds. Applied Microbiology and Biotechnology, 77, 675687. Molina, A. M., Guadalupe, V., Varela, C., Swiegers, J. H., Pretorius, I. S., & Agosin, E. (2009). Differential synthesis of fermentative aroma compounds of two related commercial wine yeast strains. Food Chemistry, 117, 189195. Nagel, C. W., & Wulf, L. W. (1979). Changes in the anthocyanins, avonoids and hydroxycinnamic acid esters during fermentation and aging of Merlot and Cabernet Sauvignon. American Journal of Enology and Viticulture, 30, 111116. OIV (2005). Compendium of international methods of wine and must analysis. Paris: Organisation Internationale de la Vigne et du Vin. Patel, S., & Shibamoto, T. (2002). Effect of different strains of Saccharomyces cerevisiae on production of volatiles in Napa Gamay wine and Petite Sirah wine. Journal of Agricultural and Food Chemistry, 50, 56495653. Rakitzis, E. T. (1975). Reaction of thioureas with FolinCiocalteu reagent. Analytica Chimica Acta, 78, 495497. Rodriguez, M. E., Lopes, C. A., Barbagelata, R. J., Barda, N. B., & Caballero, A. C. (2010). Inuence of Candida pulcherrima Patagonian strain on alcoholic fermentation behaviour and wine aroma. International Journal of Food Microbiology, 138, 1925. Satora, P., Sroka, P., Duda-Chodak, A., Tarko, T., & Tuszynski, T. (2008). The prole of volatile compounds and polyphenols in wines produced from dessert varieties of apples. Food Chemistry, 111, 513519. Skouroumounis, G. K., & Sefton, M. A. (2001). The formation of b-damascenone in wine. In P. Winterhalter & R. Rouseff (Eds.), Carotenoid-derived aroma compounds (pp. 285299). Washington, DC: American Chemical Society. Sun, S. Y., Jiang, W. G., & Zhao, Y. P. (2010). Characterization of the aroma-active compounds in ve sweet cherry cultivars grown in Yantai (China). Flavour and Fragrance Journal, 25, 206213. Wulf, L. W., & Nagel, C. W. (1980). Identication and changes of avonoids in Merlot and Cabernet Sauvignon wines. Journal of Food Science, 45, 479484. Yoo, K. M., Al-Farsi, M., Lee, H., Yoon, H., & Lee, C. Y. (2010). Antiproliferative effects of cherry juice and wine in Chinese hamster lung broblast cells and their phenolic constituents and antioxidant activities. Food Chemistry, 123, 734740.