High pressure treatments accelerate changes in volatile composition of sulphur dioxide-free wine during bottle storage

Accepted Manuscript High pressure treatments accelerate changes in volatile composition of sulfur dioxide-free wine during bottle storage Mickael C. Santos, Cláudia Nunes, M. Angélica M. Rocha, Ana Rodrigues, Sílvia M. Rocha, Jorge A. Saraiva, Manuel A. Coimbra PII: DOI: Reference:

S0308-8146(15)00715-3 http://dx.doi.org/10.1016/j.foodchem.2015.05.002 FOCH 17546

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

11 December 2014 28 April 2015 1 May 2015

Please cite this article as: Santos, M.C., Nunes, C., Rocha, M.A., Rodrigues, A., Rocha, S.M., Saraiva, J.A., Coimbra, M.A., High pressure treatments accelerate changes in volatile composition of sulfur dioxide-free wine during bottle storage, Food Chemistry (2015), doi: http://dx.doi.org/10.1016/j.foodchem.2015.05.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

High pressure treatments accelerate changes in volatile composition of sulfur dioxide-free wine during bottle storage

Mickael C. Santos a, Cláudia Nunes a*, M. Angélica M. Rocha a, Ana Rodrigues b, Sílvia M. Rocha a, Jorge A. Saraiva a, Manuel A. Coimbra a

a

QOPNA, Department of Chemistry, University of Aveiro, 3810-193 Aveiro, Portugal

b

Dão Sul –Sociedade Vitivinícola, S.A., 3430-909 Carregal do Sal, Portugal

* Author to whom correspondence should be addressed. Phone:+351 234 372581 Fax: +351 234 370084 E-mail: [email protected] (Cláudia Nunes)

1

ABSTRACT

2

The impact of high hydrostatic pressure (HHP) treatments on volatile

3

composition of sulfur dioxide-free wines during bottle storage was studied. For this

4

purpose, white and red wines were produced without sulfur dioxide (SO2) and, at the

5

end of the alcoholic fermentation, the wines were pressurised at 500 and 425 MPa for 5

6

min. Wine with 40 ppm of SO2 and a wine without a preservation treatment were used

7

as controls. More than 160 volatile compounds, distributed by 12 chemical groups, were

8

identified in the wines by an advanced gas chromatography technique. The pressurised

9

wines contained a higher content of furans, aldehydes, ketones, and acetals, compared

10

with unpressurised wines after 9 months of storage. The changes in the volatile

11

composition indicate that HHP treatments accelerated the Maillard reaction, and alcohol

12

and fatty acid oxidation, leading to wines with a volatile composition similar to those of

13

faster aged and/or thermally treated wines.

14 15 16 17

Keywords: Sulfur dioxide-free wines; High hydrostatic pressure; GC×GC–ToFMS;

18

Maillard reaction; fatty acids oxidation; Principal components analysis;

1

19

1. Introduction

20

During the last decade, the use of high hydrostatic pressure (HHP) as a non-thermal

21

technology for food preservation and modification has increased substantially. Foods can be

22

submitted to high pressures, ranging from 400 to 600 MPa, in order to destroy

23

microorganisms and inactivate enzymes with minimal effects on their sensorial and nutritional

24

properties (Buzrul, 2012; Santos, Nunes, Cappelle, et al., 2013). The application of HHP in

25

winemaking has been studied as an alternative process for preservation of wine (Buzrul, 2012;

26

Santos, Nunes, Cappelle, et al., 2013; Santos, Nunes, Rocha, et al., 2013; Tabilo-Munizaga et

27

al., 2014), allowing to produce wines with lower amounts of sulfur dioxide (SO2), since some

28

consumers are intolerant to SO2-derived compounds, namely sulfites (Vally & Misso, 2012).

29

Some studies, using pressures between 200 and 500 MPa for 1 to 20 min, showed the

30

inactivation of fungi, yeasts, and lactic acid bacteria in wines without causing significant

31

changes in wine sensorial characteristics (Buzrul, 2012). However, severe high pressure

32

treatments (650 MPa for 1 h and 2 h) changed the physicochemical and sensorial

33

characteristic of red wine, namely reduction of colour intensity and phenolic compounds

34

content. In terms of sensorial properties, sour and fruity odour of wine became less intense

35

after 2 h of pressurization, whereas the intensities of several attributes, including astringency,

36

and alcoholic and bitter tastes, were slightly enhanced (Tao et al., 2012). Recently, studies

37

demonstrated that moderate HHP treatments, 425 and 500 MPa for 5 min, influenced long

38

term physicochemical characteristics of red and white wines (Santos, Nunes, Cappelle, et al.,

39

2013; Santos, Nunes, Rocha, et al., 2013), namely more orange-red colour, and reduced

40

antioxidant activity and total phenolic content. Pressurised wines possessed a higher cooked-

41

fruit aroma and lower fruity and floral aromas than the unpressurised wines. Nevertheless,

42

these effects are perceptible only after at least 6 months of storage. These results, together

43

with the lower content of free amino acids and higher content of furans in pressurised wines,

2

44

suggest an effect of HHP treatments in the acceleration of Maillard reactions occurring during

45

the wine storage period (Santos, Nunes, Rocha, et al., 2013). However, the effects of HHP on

46

the volatile composition of the wine during storage are still largely unknown.

47

Some studies showed that HHP treatments promotes alterations in volatile

48

composition of some food products processing, changing their flavour (Oey, Lille, Van Loey,

49

& Hendrickx, 2008). Porretta, Birzi, Ghizzoni, and Vicini (1995) found that the

50

concentrations of hexanal and cis-3-hexenal increased in tomato samples treated with 500

51

MPa for 3 min due to the free fatty acid oxidation after pressure treatments. Also, pressurised

52

(400 MPa at ambient temperature for 20 min) strawberry purees stored for 30 days at 4 ºC,

53

were reported to have increased content of methyl butyrate, 2-methylbutyric acid, hexanoic

54

acid, ethyl butyrate, ethyl hexanoate, 1-hexanol, and linalool (Navarro, Verret, Pardon, & El

55

Moueffak, 2002). They also found that residual lipoxygenase activity was observed after

56

pressurisation, explaining some of the behaviour of these aroma compounds.

57

As aroma is one of the most important quality parameters of wine, the aim of this

58

work was to study the effect of high hydrostatic pressure treatments on the volatile

59

composition of sulfur dioxide-free red and white wines. This study will increase the

60

fundamental knowledge about the effect of HHP on wine, particularly the feasibility of using

61

HHP for wine long-term preservation. In order to obtain a deeper characterisation of the

62

chemical groups potentially affected by HHP treatments, comprehensive two-dimensional gas

63

chromatography coupled to mass spectrometry with a high resolution time of flight analyzer

64

(GC×GC-ToFMS) combined with headspace solid-phase microextraction (HS-SPME) was

65

used. This technique is the most suitable gas chromatography technique for untargeted

66

analysis of complex samples, such as wine (Welke, Manfroi, Zanus, Lazarotto, & Alcaraz

67

Zini, 2012). GC×GC-ToFMS offers superior separation capabilities afforded by high peak

68

capacity, selectivity, structural chromatographic peak organization, and sensitivity

3

69

enhancement in comparison to 1D-GC ( Rocha et al., 2013). GC×GC has been used in the

70

determination of volatile compounds in different grape and wine varieties, including Cabernet

71

Sauvignon (Robinson et al., 2011), Fernão-Pires (Rocha, Coelho, Zrostlíková, Delgadillo, &

72

Coimbra, 2007), Madeira (Perestrelo, Barros, Câmara, & Rocha, 2011), Pinotage

73

(Weldegergis, Villiers, et al., 2011), and Marsala (Dugo et al., 2014) wines.

74 75

2.

Materials and methods

76

2.1

Wine samples and high pressure treatments

77

Encruzado and Touriga-Nacional (Vitis vinifera L.) grapes harvested in 2010 in Dão

78

Appellation were used to prepare white and red wine samples, respectively. The wines were

79

produced by Dão Sul SA (Carregal do Sal, Portugal) without the addition of SO2. For white

80

wine, after destemming and crushing in a pneumatic press, the free-running juice was cooled

81

to 5 ºC and transferred to a stainless steel vessel with the addition of commercial pectolytic

82

enzymes. The must was allowed to settle for 24 h, after which it was decanted into a stainless-

83

steel vessel. The must was inoculated with a commercial active dry Saccharomyces cerevisiae

84

and fermented at 15 to 18 ºC for three weeks. For red wine, the grapes were destemmed and

85

crushed in a stainless-steel vessel and commercial maceration enzymes were added. After 24

86

h of cold pre-fermentation maceration at 15 °C, the must was inoculated with a commercial

87

active dry Saccharomyces cerevisiae preparation. The alcoholic fermentation occurred for 10

88

days at 20 to 25 °C. After the beginning of the alcoholic fermentation, the must was punched

89

down for 20 min every 3 h and was submitted to a rack and return program for 30 min each

90

day. At the end of alcoholic fermentation, the free-running wine was transferred to another

91

stainless steel vessel for spontaneous malolactic fermentation.

92

After fermentation, part of the wine was transferred into a 500-L stainless steel tank

93

without addition of SO2, and then transferred to 250-mL screw-capped, flexible and high-

4

94

pressure-resistant polyethylene bottles, stoppered, and pressurised for 5 min at 20 °C at 425

95

MPa or 500 MPa, conditions that assure microbiologically safe wines (Buzrul, 2012), in a

96

hydrostatic press from Avure Technologies (Model 215L-600; Erlanger, KY), giving origin to

97

samples 425 MPa and 500 MPa, respectively. Pressurising water was used at a controlled

98

temperature of 15 °C. Pressure build-up took place at a compression rate of about 300

99

MPa/min (adiabatic heating caused an increased in temperature of about 4.0 °C), while

100

decompression was nearly instantaneous. As polyethylene bottles can have some impact on

101

the sensorial properties of the white wine (Ghidossi et al., 2012), two lots of the same wines

102

were also bottled in the polyethylene bottles, one with addition of 40 mg/L of SO2 (added in a

103

closed loop as a 7% sulfur dioxide solution), the typical amount used in the wine industry

104

(sample named as SO2), and another with no addition of SO2 and no high pressure treatment

105

(untreated).

106

All the oenological parameters were determined using the methods described by the

107

Office International de la Vigne et du Vin (2006). The white wine contained 9 mg/L free SO2

108

and 18 mg/L total SO2, and the red wine contained 8 mg/L free SO2, and 18 mg/L total SO2.

109

The oenological parameters of the wines at the beginning of storage were not altered by the

110

pressure treatments (Table S1, supplementary data). All wines were stored at 80% relative

111

humidity in the absence of light at around 10 °C.

112 113

2.2

Volatile composition analyses

114

The volatile composition of the red and white wines samples was analysed (three

115

independent aliquots) by HS-SPME combined with a GC×GC–ToFMS after 2 and 9 months

116

of storage (Petronilho, Coimbra, & Rocha, 2014).

117

The solid-phase microextraction (SPME) holder for manual sampling and fibre were

118

purchased from Supelco (Bellefonte, PA). The SPME device included a fused silica fibre

5

119

coating partially cross-linked with 50/30 µm divinylbenzene/Carboxen/ polydimethylsiloxane

120

(DVB/CAR/PDMS) coating. Prior to use, the SPME fibre was conditioned at 270 °C for 60

121

min in the GC injector, according to the manufacturer's recommendations. Subsequently, the

122

fibre was conditioned daily for 10 min at 250 °C. For the HS-SPME assay, aliquots of 3.0 mL

123

of the sample were placed into a 9-mL glass vial. After the addition of 0.6 g of NaCl each vial

124

was capped with a PTFE/silicone septum (Supelco). The vial was placed in a thermostated

125

bath adjusted at 40.0 ± 0.1ºC with stirring (1.5 × 0.5 mm bar) at 400 rpm, and the SPME fibre

126

was manually inserted into the sample vial headspace for 20 min. Blanks, corresponding to

127

the analysis of the SPME fibre not submitted to any extraction procedure, were run between

128

sets of three analyses.

129

After the extraction/concentration step, the SPME fibre was manually introduced into

130

the GC×GC-ToFMS injection port at 250 ºC and kept for 30 s for compound desorption. The

131

injection port was lined with a 0.75 mm I.D. splitless glass liner. The LECO Pegasus 4D

132

(LECO, St. Joseph, MI) GC×GC-ToFMS system consisted of an Agilent GC 7890A gas

133

chromatograph (Agilent Technologies, Inc., Wilmington, DE), with a dual stage jet cryogenic

134

modulator (licensed from Zoex) and a secondary oven, and mass spectrometer equipped with

135

a high resolution ToF analyser. The detector was a highspeed ToF mass spectrometer. An HP-

136

5 column (30 m × 0.32 mm I.D., 0.25 µm film thickness; J & W Scientific Inc., Folsom, CA)

137

was used as the first-dimension column, and a DB-FFAP (0.79 m × 0.25 mm I.D., 0.25 µm

138

film thickness; J&W Scientific Inc.) was used as the second-dimension column. The carrier

139

gas was helium at a constant flow rate of 2.5 mL/min. The primary oven temperature was

140

programmed from 40 (1 min) to 230 ºC (2 min) at 10 ºC/min. The secondary oven

141

temperature was programmed from 70 (1 min) to 250 ºC (3 min) at 10 ºC/min. The MS

142

transfer line temperature was 250 ºC, and the MS source temperature was 250 ºC. The

143

modulation time was 5 s; and the modulator temperature was kept at 20 ºC offset (above

6

144

primary oven). The ToFMS was operated at a spectrum storage rate of 125 spectra/s. The

145

mass spectrometer was operated in the EI mode at 70 eV using a range of m/z 33‒350 and the

146

detector voltage was -1786 V.

147

Total ion chromatograms (TIC) were processed using the automated data processing

148

software ChromaTOF® (LECO) at a signal-to-noise threshold of 100. Two commercial

149

databases (Wiley 275 and US National Institute of Science and Technology (NIST) V. 2.0,

150

Mainlib and Replib) were used. A mass spectral match factor, the majority (86%) of the

151

tentatively identified compounds showed similarity matches >850, was set to decide whether

152

a peak was correctly identified or not. Furthermore, a manual inspection of the mass spectra

153

was done, combined with the use of additional data, such as the retention index (RI) value,

154

which was determined according to the Van den Dool and Kratz RI equation (Van den Dool

155

& Kratz, 1963). For the determination of the RI, a C8–C20 n-alkanes series was used, and

156

these values were compared with values reported in the literature for chromatographic

157

columns similar to those used in the present work (Ansorena, Astiasarán, & Bello, 2000;

158

Campeol et al., 2003; Cardeal, de Souza, da Silva, & Marriott, 2008; Eyres, Dufour, Hallifax,

159

Sotheeswaran, & Marriott, 2005; Jordán, Margaría, Shaw, & Goodner, 2002; Leffingwell &

160

Alford, 2005; Perestrelo et al., 2011; Petronilho, Maraschin, Delgadillo, Coimbra, & Rocha,

161

2011; Pino, Mesa, Muñoz, Martí, & Marbot, 2005; Rocha et al., 2007). The DTIC

162

(Deconvoluted Total Ion Current) GC×GC area data were used as an approach to estimate the

163

relative content of each volatile component in wine, and were expressed as arbitrary units (a.

164

u.). Reproducibility was expressed as relative standard deviation (RSD).

165 166

2.3

Statistical analysis

7

167

Statistical data analysis was performed using analysis of variance (ANOVA) using

168

Statistica6.1 (Statsoft Inc., Tulsa, OK). Tukey’s HSD Test was used as a comparison test when

169

significant differences were observed by ANOVA (p < 0.05).

170

Principal components analysis (PCA) was applied to the auto-scaled areas of all

171

volatile compounds identified by HS-SPME/GC×GC–ToFMS in the pressurised and

172

unpressurised wines after 2 and 9 months of storage. The goal of this approach was to extract

173

the main sources of variability and hence to characterise the dataset.

174 175

3. Results and discussion

176

All the wine samples were analysed after 2 and 9 months of bottle storage in order to

177

observe a possible effect of the high pressure treatments on the volatile composition of the

178

wines.

179

Tables S2 and S3 shown as Supplementary data gives detailed information for each

180

compound, including GC peak area, RSD, and RI experimentally calculated as well as

181

reported in the literature, for the white and red wines, respectively. The reproducibility,

182

expressed as RSD, of the different identified volatile compounds ranged from 1% to 58%,

183

which is a common range for natural products. The highest variability was usually observed

184

for the compounds identified in trace amounts.

185 186

3.1. Volatile composition of the wines after 2 months of storage

187

The volatile composition analysis, after 2 months of storage, revealed for white wine

188

samples the presence of 167, 172, 163, and 157 compounds in the untreated, SO2, and

189

pressurized at 425 MPa and 500 MPa, respectively. In the red wine samples, 157, 163, 166

190

and 167 compounds were detected in the untreated, SO2, 425 MPa and 500 MPa samples,

191

respectively. These compounds were grouped into 12 chemical families: acids, esters,

192

alcohols, volatile phenols, aldehydes, ketones, furans, lactones, acetals, thiols and other 8

193

sulphur compounds, norisoprenoids, and terpenic compounds. The ester group contained the

194

highest number of identified compounds (62/63 in white/red wines), followed by alcohols

195

(30/35 in white/red wines), and terpenic compounds (15 in white wines and 23 in red wines)

196

(Tables S2 and S3). These results are in accordance with studies conducted on Pinotage wines

197

(Weldegergis, Villiers, et al., 2011), South Africa red wines (Weldegergis, Crouch, Górecki,

198

& de Villiers, 2011), and Brazilian Merlot wines (Welke, Manfroi, Zanus, Lazarotto, &

199

Alcaraz Zini, 2012).

200

The total peak areas for each chemical group identified in the white and red wine

201

samples, after 2 months of storage, are presented in Fig. 1A and 1B, respectively. After 2

202

months of storage, the impact of the two pressure treatments on the volatile composition of

203

both white and red wines, was minimal, but statistically significant for some chemical groups

204

(p < 0.05), namely for esters and acids in the case of white wine, and acids and norisoprenoids

205

for red wine.

206

After 2 months of storage the pressurised white wines contained a lower content of esters

207

than the unpressurised white wines (p<0.05) (Fig 1A). This lower content of esters is mainly due

208

to the lower content of the aliphatic ethyl esters, such as a 2-fold lower content of ethyl octanoate

209

and 6 to 9-fold lower amount of ethyl decanoate than the unpressurised white wines. These two

210

esters are frequent products of fermentation, with fruity and floral odours (Weldegergis,

211

Villiers, et al., 2011).

212

Both pressurised red wines possessed a lower content of acids (Fig. 1B), mainly due to

213

the 2 to 3-fold lower area of the acetic acid peak (peak number 1, Table S3) than the

214

unpressurised wines. Acetic acid is one of the dominant acids in red wines, based on their

215

peak area, in agreement with previous reports (Weldegergis, Crouch, et al., 2011;

216

Weldegergis, Villiers, et al., 2011), contributing negatively to the wine bouquet (Fang &

217

Qian, 2005). Since this compound is produced during fermentation, the lower content of

9

218

acetic acid in pressurized wines could indicate that the pressure treatments stopped the

219

fermentation of the wine in a more effective way than the addition of SO2. The pressurised red

220

wines also possessed a higher content of norisoprenoids (p < 0.05) when compared with the

221

SO2 and untreated samples. The higher content of norisoprenoids in pressurised wines, after 2

222

months of storage, was mainly due to the presence of geranyl acetone that was only identified

223

in the pressurised wines samples. The C13 norisoprenoids have been related to complex wine

224

flavours, described as grassy, tea, lime, honey, and pineapple, and rose (Pino et al., 2005;

225

Weldegergis, Crouch, et al., 2011; Weldegergis, Villiers, et al., 2011). These compounds,

226

similar to the monoterpenes, occur in grapes largely as non-bound carotenoid precursors,

227

while geranyl acetone may result from the oxidative cleavage of squalene (Ribéreau-Gayon,

228

Glories, Maujean, & Dubourdieu, 2006).

229

Overall, despite some differences observed in the volatile composition of the

230

pressurised wines, the impact of the pressure treatments was minimal after 2 months of

231

storage. This result is in agreement with previous results showing that high pressure

232

treatments (400–500 MPa for few minutes) do not alter significantly white and red wine

233

physicochemical and sensorial properties in the first months of storage (Santos, Nunes,

234

Cappelle, et al., 2013; Santos, Nunes, Rocha, et al., 2013; Tao et al., 2012).

235 236

3.2 Volatile composition of wines after 9 months of storage

237

After 9 months of storage a lower number of compounds was detected (up to 15%

238

less) in both white and red wines when compared with the same wine samples after 2 months

239

of storage (Tables S2 and S3). This may be due to an increase of the interaction between

240

volatile compounds and other compounds present in wine, namely polyphenols, during wine

241

aging (Ribéreau-Gayon, Glories, Maujean, & Dubourdieu, 2006;). In white wine samples 148,

242

146, 148 and 148 compounds were detected in unpressurised, SO2 and pressurised 425 MPa

10

243

and 500 MPa, respectively. In the red wine samples, 150, 151, 141 and 142 compounds were

244

detected in the samples untreated, SO2, 425 MPa, and 500 MPa, respectively. As observed in

245

the wines with 2 months of bottle aging, esters presented the higher number of identified

246

compounds (59/69 in white/red wines), followed by alcohols (26/24 in white/red wines), and

247

terpenic compounds (13/16 in white/red wines) for 9 months of bottle aging (Tables S2 and

248

S3).

249

The total peak areas for each chemical group identified in the white and red wines,

250

after 9 months of storage, are presented in Fig. 2A and 2B, respectively. It can be noticed that,

251

contrary to the wine samples with 2 months of bottle aging, the pressurised wine samples

252

presented a volatile composition remarkably different to the unpressurised, indicating a large

253

impact of the pressure treatments on the volatile composition of both white and red wines. In

254

particular, the pressurised wine samples had a higher content of acetals, ketones, furans, and

255

aldehydes.

256

In order to reduce the dimensionality of the data set, to study the main sources of

257

variability of the data set and detect differences/similarities among wine samples, principal

258

component analysis (PCA) was performed using, as analytical variables, the GC peak areas of

259

all volatile compounds of white and red wine samples with 9 months of bottle storage. This

260

showed the effect of the different treatments on the wines’ volatile composition during storage

261

as well as the relationships/correlations between wine samples and compounds.

262 263

3.2.1 White wine

264

Fig. 3A shows a biplot reporting the score plots combined with the loadings plots of

265

the two first principal components (which explain 77% of the total variability of the data set)

266

for the white wine samples. The loadings establish the relative importance of each volatile

267

compound for the observed sample distribution. PC1, which explains 60% of the total

11

268

variability, separated wines treated with high pressure (425 MPa and 500 MPa) from the

269

untreated and SO2 ones. PC2, explaining 17% of the total variability, shows the distribution

270

of the wines according to the presence of sulfur dioxide. The pressurised wines were

271

negatively located in relation to PC1 and positively located in relation to PC2. These samples

272

are characterised mainly by ketones, acetals, furans, and aldehydes.

273

The

ketones

3-pentanone,

3-penten-2-one,

1-(ethenyloxy)-3-methylbutane,

3-

274

octanone, 3-nonanone, and 2,5-octanedione were only identified in the pressurised wine

275

samples (Table S2). Ketones are reported to result from the direct oxidation of fatty acids

276

(Campo, Ferreira, Escudero, Marques, & Cacho, 2006; Weldegergis, de Villiers, McNeish,

277

Seethapathy, Mostafa, Gorecki, et al., 2011) and are mainly described to have “buttery” and

278

“fatty” odours (Rocha, Rodrigues, Coutinho, Delgadillo, & Coimbra, 2004; Schneider,

279

Baumes, Bayonove, & Razungles, 1998). The presence of these ketones in the pressurised

280

wines indicates the occurrence of fatty acid oxidation with pressure. Previous studies have

281

shown that HHP treatments enhance lipid oxidation in foods (Medina-Meza, Barnaba, &

282

Barbosa-Cánovas, 2014).

283

In the acetal family, 1,1-diethoxypentane and 1-(1-ethoxyethoxy)butane were only

284

identified in the pressurised wines and the content of 1,1-diethoxyethane and 1-(1-

285

ethoxyethoxy)-pentane were 40 to 49% and 65 to 68% higher, respectively, in these samples

286

when compared with the SO2 wine. These acetals are reported to have “caramel” and “dried

287

fruit” odours and their presence is common in wines submitted to oxidative aging, as well as

288

in sherry wines (Schneider, Baumes, Bayonove, & Razungles, 1998). These results are in

289

agreement with literature; sulfur dioxide-free wines possessed higher “cooked fruit” aroma

290

after pressure treatments (Santos, Nunes, Cappelle, et al., 2013; Santos, Nunes, Rocha, et al.,

291

2013). Since acetals are formed by the reaction of aldehydes (mainly acetaldehyde) with

12

292

alcohols, it seems that the HHP treatments accelerated the occurrence of this reaction during

293

wine storage.

294

The importance of furans and aldehydes in the differentiation of pressurised wines

295

from unpressurised (Fig. 3A) is due to the higher content of these compounds in the 425 MPa

296

and 500 MPa wines (Table S2). The higher content of furans in pressurised wines samples

297

was mainly due to the 10- and 5-fold higher content of 2-furfural in the samples pressurised at

298

425 MPa and 500 MPa, respectively. Moreover, 5-methylfurfural and2-acetyl-5-methylfuran

299

were only detected in the pressurised wine samples. The higher content of aldehydes in the

300

pressurised wine samples was mainly due to the higher content of benzaldehyde, 10- and 15-

301

fold higher content in the samples pressurised at 425 MPa and 500 MPa, respectively, when

302

compared with the untreated and SO2 white wines. Both 2-furfural and benzaldehyde are

303

considered Maillard reaction-derived volatile compounds, as 2-furfural can be formed by the

304

dehydration of sugars through the Maillard reaction (Perestrelo et al., 2011) and benzaldehyde

305

by the Strecker degradation of amino acids as a result of the Maillard reaction (Pripis-Nicolau,

306

de Revel, Bertrand, & Maujean, 2000). Benzaldehyde may also be formed through the

307

shikimic acid pathway, having phenylalanine as intermediate (Ribéreau-Gayon, Glories,

308

Maujean, & Dubourdieu, 2006). The results obtained infer that HHP treatments accelerated

309

Maillard reactions during the wine storage period. These conclusions are also supported by

310

previous studies that showed that pressurised white wines possessed, at least after 6 months of

311

storage, a more brownish colour, lower content of free amino acids, and higher content of

312

furans (Santos, Nunes, Rocha, et al., 2013). Also, several studies conducted in model systems

313

containing amino acids and sugars demonstrated that high pressure treatments can accelerate

314

the formation of Amadori rearrangement compounds (Moreno, et al., 2003; Schwarzenbolz,

315

Klostermeyer, & Henle, 2002).

13

316

According to Figure 3A, the untreated white wine is characterised (PC1 and PC2

317

positive) by 4-ethyphenol, 4-ethylguaiacol, isobutyl butyrate, propyl hexanoate, hexyl 2-

318

methylbutyrate, and isophorone (Table S2). The ethylphenols are normally produced by

319

spoilage of Brettanomyces/Dekkera spp. yeasts involving cinnamic, coumaric, and ferulic

320

acids, free or esterified with tartaric acid (Larcher, Puecher, Rohregger, Malacarne, &

321

Nicolini, 2012) These compounds are responsible for a particularly unpleasant sensory defect

322

known as ‘mousy off-flavour’ (Romano, Perello, de Revel, & Lonvaud-Funel, 2008).

323

Therefore, these compounds indicate wine spoilage in the untreated samples (Romano,

324

Perello, de Revel, & Lonvaud-Funel, 2008). In order to verify the microbiological stability of

325

wines during their storage, a simple microorganism enumeration was performed by

326

inoculating serially diluted wine samples on plates containing the specific culture media for

327

bacteria and yeast (data not shown). All the wine samples submitted to HHP or to which SO2

328

was added, showed no microorganism growth, contrary to the untreated wine, confirming the

329

results. The presence of the esters isobutyl butyrate, propyl hexanoate, and hexyl 2-methyl-

330

butyrate may also be due to the presence of microorganisms in the untreated wine, since these

331

compounds can result from fermentation occurring during wine ageing (Weldegergis, Villiers,

332

et al., 2011).

333

In the wine sample with addition of sulfur dioxide, geraniol (Table S2) is the principal

334

contributor to its location in PC1 positive and PC2 negative (Fig. 3A). The content of geraniol

335

in this sample was 63% higher when compared with the untreated samples and was not

336

identified in the pressurised wines (Table S2). Monoterpene alcohols, such as geraniol, which

337

contribute to the wine varietal characteristics, belong to the most relevant flavour compounds

338

of several white wine varieties and are responsible for their characteristic floral aroma

339

(Ribéreau-Gayon et al., 2006). Sulfur dioxide was reported to have a protective effect on these

340

volatiles (Roussis & Sergianitis, 2008), which explains the higher concentration of geraniol in

14

341

the SO2 wine sample, when compared with the other samples. In addition, geraniol content

342

decreases with wine ageing and is usually present in trace amounts after two or three years in

343

the bottle (Pedersen, Capone, Skouroumounis, Pollnitz, & Sefton, 2003). This compound can

344

undergo several reactions during wine storage (easily isomerises and oxidises, forming oxides

345

and aldehydes), induced by the time of storage and relatively low pH (Dziadas & Jeleń,

346

2010).

347 348

3.2.2 Red wines

349

Fig. 3B shows the biplot reporting the score plots combined with the loadings plots of

350

the two first principal components (which explain 65% of the total variability of the data set)

351

of the PCA performed for the red wine samples. As observed for the white wine samples,

352

PC1, which explains 55% of the total variability, distinguishes the wines as a function of the

353

pressure treatments, and PC2, explaining 10% of the total variability, differentiated the wines

354

according to the presence of sulfur dioxide. Pressurised wines were negatively located on PC1

355

and positively located on PC2, and no differences were observed between the samples 425

356

MPa and 500 MPa (Fig 3B). These results show that the difference in the pressure value

357

between the two pressures applied (425 MPa and 500 MPa during 5 min) had no significant

358

effect on the volatile composition of the red wines. As observed for the white wines, the

359

pressurised red wines were characterized mainly by ketones, acetals, furans, and aldehydes.

360

The ketones responsible for pressurised wine discrimination were 3-pentanone, 2,3-

361

pentanedione, 2,3-hexanedione, 3-octanone and 3-nonanone, since these compounds have

362

only been identified in these wines (Table S3). In addition, acetoin, 2-heptanone, and 2-

363

nonanone are presented at higher concentrations in the pressurised wines (up to 78%, 82%,

364

and 86%, respectively), compared to both unpressurised red wine samples (Table S3). These

15

365

ketones are described to have “buttery” and “fatty” odours (Schneider, Baumes, Bayonove, &

366

Razungles, 1998) and resultant from fatty acid oxidation.

367

The acetals that characterised the pressurized red wines (Fig 3B) are 1,1-diethoxy-2-

368

methylpropane, 1-(1-ethoxyethoxy)butane and 1,1-diethoxy-3-methylbutane (Table S3). 1-(1-

369

Ethoxyethoxy)-butane was only identified in the pressurized wines and the content of 1,1-

370

diethoxy-2-methylpropane and 1,1-diethoxy-3-methylbutane were up to 70% and 68% higher,

371

respectively, in these samples when compared with the SO2 wine. These results show that, as

372

observed for white wines, the formation of acetals is accelerated with the pressure treatments,

373

increasing the content of compounds with “dried fruit” odours. Hexanal is one aldehyde that

374

characterised the pressurised red wines, since these wines presented around 5-fold higher

375

content of this compound than unpressurised wines. This result infers that the oxidation of

376

some alcohols, such 1-hexanol, can also be accelerated by pressure treatments.

377

The pressurized red wines, as observed for white wines, were also characterised by the

378

presence of a higher content of Maillard volatile compounds, namely 2-furfural,

379

benzaldehyde, and phenylacetaldehyde (Table S3). In fact, the pressurized wines presented 5-

380

to 11-fold higher furfural content, and 2-fold higher benzaldehyde and phenylacetaldehyde

381

content when compared with the unpressurised samples. These results infer that pressure

382

treatments accelerated Maillard reactions during the storage period of red wines. However,

383

the difference in Maillard volatile compounds between pressurised and unpressurised wines

384

was lower in red wines than white wines. This behaviour can be due to the higher content of

385

polyphenols in red wine, when compared with white wine, that reduce the rate of Maillard

386

reactions, due to their higher antioxidant activity, and consequently decrease the formation of

387

Maillard reaction-derived volatile compounds (Oliveira, Ferreira, De Freitas, & Silva, 2011).

388

As observed in Fig 3B, the untreated and SO2 red wine samples are separated by PC2,

389

explaining 10% of the total variability. These results show that, contrary to the white wines

16

390

where the untreated white wine was well separated from the SO2, due to the presence of

391

volatile compounds possibly originating from microorganism contamination in the untreated

392

white wines, the separation between the two unpressurised red wines was not so evident.

393

Consequently, the absence of microorganism contamination in red wines was the main factor

394

for the lower value (10%) of variability across PC2 for the untreated and SO2 samples.

395

Therefore the main separation in the red wine samples is due to the pressure treatments.

396 397

3.3 Evolution of ketones, aldehydes, furans, and acetals profile during wine storage

398

Since the pressurised wines with 9 months of storage were mainly characterised by

399

ketones, acetals, furans, and aldehydes, it was necessary to understand the impact of the

400

pressure treatments on these chemical groups, after 2 months of storage, and their evolution

401

during storage. For that, a heat map (Fig 4), a logarithmic normalisation of the GC peak area,

402

was prepared, for a direct and rapid interpretation of the relative abundance of each aldehyde,

403

ketone, furan, and acetal compound for the different white and red wines (with three

404

independent assays) at 2 and 9 months of storage. White (Fig.4A) and red wines (Fig. 4B)

405

after 2 months of storage revealed a similar profile among the samples with different

406

treatments, since the relative abundance of the chemical groups are homogenous for all the

407

wines. These results are in accordance with preview ones that show that after 2 months of

408

storage, the impact of the two pressure treatments on the volatile composition of both white

409

and red wine was minimal. However, after 9 months of storage it is possible to observe that,

410

for both pressurised white and red wines, the volatile profiles of each aldehyde, ketone, furan,

411

and acetal compound were very different when compared with the unpressurised wines. These

412

results confirm that the impact of pressure treatments in both white and red wines was only

413

noticeable after several months of storage.

17

414

Acetals, ketones, and Maillard volatile compounds, such as furfural and benzaldehyde,

415

have a tendency to increase linearly during oxidative conditions of aging due to the

416

occurrence of the Maillard reaction, and the oxidation of alcohols and fatty acids, and are

417

reported as potential age markers of sherry (Sun et al., 2013) and Madeira wines (Perestrelo,

418

Barros, Camara, & Rocha, 2011). Therefore, it seems that the pressurised wine samples

419

possess a volatile composition characteristic of faster aged/thermally treated wines. These

420

results indicate that the HHP treatment influences the white and red wine long-term volatile

421

composition and seems to accelerate their evolution during storage, this being particularly

422

evident for longer storage periods.

423 424

4.

Conclusion

425

The results obtained in this work demonstrate that high pressure treatments with

426

processing time around 5 min and pressures between 400 and 500 MPa influence white and

427

red wine volatile composition However, the effect is only perceptible after some months of

428

storage, changing the wine aroma characteristics. The two pressure treatments studied showed

429

similar effects in both white and red wines. The changes on the volatile composition of the

430

pressurised wines, namely the increase of furans, aldehydes, ketones, and acetals content,

431

indicate that the HHP treatments accelerate the Maillard reaction, and the oxidation of

432

alcohols and fatty acids, leading to wines with a volatile composition characteristic of faster

433

aged and/or thermally treated wines.

434

These aspects should be taken into consideration in the implementation of HHP

435

treatments for wine preservation as an alternative to SO2. These findings also open new

436

opportunities to create wines with distinct and novel characteristics.

437

Even though the approach followed in this work provides a broad perspective into

438

complex chemical reactions, in order to fully understand the effect of HHP on volatile

18

439

composition of wine, further attention should be given to the following aspects: (i)

440

quantification of the selected discriminant components and their relation to the sensorial

441

analysis of the wines and (ii) the effect of HHP treatments on the volatility and perception of

442

wine aroma compounds in model wine solutions.

443 444

Acknowledgements

445

The authors thank Fundação para a Ciência e a Tecnologia (FCT, Portugal), European

446

Union, QREN, FEDER and COMPETE for funding the reseach unit 62/94 QOPNA (project

447

PEst-C/QUI/UI0062/2013; FCOMP-01-0124-FEDER-037296), QREN Project nº3462, and

448

FCT

449

(SFRH/BPD/46584/2008 and SFRH/BD/70066/2010). The authors also thank Frubaça-

450

Cooperativa de Hortofruticultores CRL for high pressure treatments of the wine samples.

project

(PTDC/AGR-ALI/101251/2008)

and

FCT

for

the

grants

451 452 453

References

454

Ansorena, D., Astiasarán, I., & Bello, J. (2000). Influence of the Simultaneous Addition of the

455

Protease Flavourzyme and the Lipase Novozym 677BG on Dry Fermented Sausage

456

Compounds Extracted by SDE and Analyzed by GC-MS. Journal of Agricultural and

457

Food Chemistry, 48(6), 2395–2400.

458 459

Buzrul, S. (2012). High hydrostatic pressure treatment of beer and wine: A review. Innovative Food Science & Emerging Technologies, 13, 1–12.

460

Campeol, E., Flamini, G., Cioni, P. L., Morelli, I., Cremonini, R., & Ceccarini, L. (2003).

461

Volatile fractions from three cultivars of Olea europaea L. collected in two different

462

seasons. Journal of Agricultural and Food Chemistry, 51(7), 1994–1999.

19

463

Cardeal, Z. L., de Souza, P. P., da Silva, M. D. R. G., & Marriott, P. J. (2008).

464

Comprehensive two-dimensional gas chromatography for fingerprint pattern recognition in

465

cachaça production. Talanta, 74(4), 793–799.

466

Dugo, G., Franchina, F. a, Scandinaro, M. R., Bonaccorsi, I., Cicero, N., Tranchida, P. Q., &

467

Mondello, L. (2014). Elucidation of the volatile composition of Marsala wines by using

468

comprehensive two-dimensional gas chromatography. Food Chemistry, 142, 262–268.

469 470

Dziadas, M., & Jeleń, H. H. (2010). Analysis of terpenes in white wines using SPE-SPMEGC/MS approach. Analytica Chimica Acta, 677(1), 43–49.

471

Eyres, G., Dufour, J. P., Hallifax, G., Sotheeswaran, S., & Marriott, P. J. (2005).

472

Identification of character-impact odorants in coriander and wild coriander leaves using

473

gas chromatography-olfactometry (GCO) and comprehensive two-dimensional gas

474

chromatography-time-of-flight mass spectometry (GC x GC-TOFMS). Journal of

475

Separation Science, 28, 1061–1074.

476

Fang, Y., & Qian, M. (2005). Aroma compounds in Oregon Pinot Noir wine determined by

477

aroma extract dilution analysis (AEDA). Flavour and Fragrance Journal, 20(1), 22–29.

478

Ghidossi, R., Poupot, C., Thibon, C., Pons, A., Darriet, P., Riquier, L., Mietton Peuchot, M.

479

(2012). The influence of packaging on wine conservation. Food Control, 23(2), 302–311.

480

Jordán, M. J., Margaría, C. A., Shaw, P. E., & Goodner, K. L. (2002). Aroma Active

481

Components in Aqueous Kiwi Fruit Essence and Kiwi Fruit Puree by GC-MS and

482

Multidimensional GC/GC-O. Journal of Agricultural and Food Chemistry, 50(19), 5386–

483

5390.

484

Leffingwell, J. C., & Alford, E. D. (2005). Volatile constituents of Perique tobacco.

485

Electronic Journal of Environmental, Agricultural and Food Chemistry, 4(2), 899–915.

20

486

Medina-Meza, I. G., Barnaba, C., & Barbosa-Cánovas, G. V. (2014). Effects of high pressure

487

processing on lipid oxidation: A review. Innovative Food Science & Emerging

488

Technologies, 22, 1–10.

489

Moreno, F. J., Molina, E., Olano, A., & López-Fandiño, R. (2003). High-pressure effects on

490

Maillard reaction between glucose and lysine. Journal of Agricultural and Food

491

Chemistry, 51(2), 394–400.

492

Navarro, M., Verret, C., Pardon, P., & El Moueffak, A. (2002). Changes in Volatile Aromatic

493

Compounds of Strawberry Puree Treated by High-pressure During Storage. High Pressure

494

Research, 22(3-4), 693–696.

495

Oey, I., Lille, M., Van Loey, A., & Hendrickx, M. (2008). Effect of high-pressure processing

496

on colour, texture and flavour of fruit- and vegetable-based food products: a review.

497

Trends in Food Science & Technology, 19(6), 320–328.

498 499

500 501

OIV. (2006). Recueil des Méthodes Internationales d’Analyse des Vins et Moûts. Paris: Organisation International de la Vigne et du Vin. Oliveira, C. M., Ferreira, A. C. S., De Freitas, V., & Silva, A. M. S. (2011). Oxidation mechanisms occurring in wines. Food Research International, 44(5), 1115–1126.

502

Pedersen, D. S., Capone, D. L., Skouroumounis, G. K., Pollnitz, A. P., & Sefton, M. A.

503

(2003). Quantitative analysis of geraniol, nerol, linalool, and alpha-terpineol in wine.

504

Analytical and Bioanalytical Chemistry, 375(4), 517–522.

505

Perestrelo, R., Barros, A. S., Câmara, J. S., & Rocha, S. M. (2011). In-depth search focused

506

on furans, lactones, volatile phenols, and acetals as potential age markers of Madeira wines

507

by comprehensive two-dimensional gas chromatography with time-of-flight mass

508

spectrometry combined with solid phase microextraction. Journal of Agricultural and

509

Food Chemistry, 59(7), 3186–3204.

21

510

Petronilho, S., Coimbra, M. A., & Rocha, S. M. (2014). A critical review on extraction

511

techniques and gas chromatography based determination of grapevine derived

512

sesquiterpenes. Analytica Chimica Acta, 846(1) 8-35

513

Petronilho, S., Maraschin, M., Delgadillo, I., Coimbra, M. A., & Rocha, S. M. (2011).

514

Sesquiterpenic composition of the inflorescences of Brazilian chamomile (Matricaria

515

recutita L.): Impact of the agricultural practices. Industrial Crops and Products, 34(3),

516

1482–1490.

517

Pino, J. A., Mesa, J., Muñoz, Y., Martí, M. P., & Marbot, R. (2005). Volatile components

518

from mango (Mangifera indica L.) cultivars. Journal of Agricultural and Food Chemistry,

519

53(6), 2213–2223.

520 521

Porretta, S., Birzi, A., Ghizzoni, C., & Vicini, E. (1995). Effects of ultra-high hydrostatic pressure treatments on the quality of tomato juice. Food Chemistry, 52(1), 35–41.

522

Pripis-Nicolau, L., de Revel, G., Bertrand, A., & Maujean, A. (2000). Formation of Flavor

523

Components by the Reaction of Amino Acid and Carbonyl Compounds in Mild

524

Conditions. Journal of Agricultural and Food Chemistry, 48(9), 3761–3766.

525

Ribéreau-Gayon, P., Glories, Y., Maujean, A., & Dubourdieu, D. (2006). Handbook of

526

Enology Volume 2: The Chemistry of Wine Stabilization and Treatments (2nd edn.).

527

Chichester: John Wiley & Sons Ltd.

528

Robinson, A. L., Boss, P. K., Heymann, H., Solomon, P. S., & Trengove, R. D. (2011).

529

Development of a sensitive non-targeted method for characterizing the wine volatile

530

profile using headspace solid-phase microextraction comprehensive two-dimensional gas

531

chromatography time-of-flight mass spectrometry. Journal of Chromatography. A,

532

1218(3), 504–517.

533 534

Rocha, S. M., Coelho, E., Zrostlíková, J., Delgadillo, I., & Coimbra, M. A. (2007). Comprehensive

two-dimensional

gas

chromatography

with

time-of-flight

mass

22

535

spectrometry of monoterpenoids as a powerful tool for grape origin traceability. Journal of

536

Chromatography. A, 1161(1-2), 292–299.

537

Rocha, S. M., Freitas, R., Cardoso, P., Santos, M., Martins, R., & Figueira, E. (2013).

538

Exploring the potentialities of comprehensive two-dimensional gas chromatography

539

coupled to time of flight mass spectrometry to distinguish bivalve species: Comparison of

540

two clam species (Venerupis decussata and Venerupis philippinarum). Journal of

541

Chromatography. A, 1315, 152–161.

542

Romano, A., Perello, M. C., de Revel, G., & Lonvaud-Funel, A. (2008). Growth and volatile

543

compound production by Brettanomyces/Dekkera bruxellensis in red wine. Journal of

544

Applied Microbiology, 104(6), 1577–1585.

545

Santos, M. C., Nunes, C., Cappelle, J., Gonçalves, F. J., Rodrigues, A., Saraiva, J. A., &

546

Coimbra, M. A. (2013). Effect of high pressure treatments on the physicochemical

547

properties of a sulphur dioxide-free red wine. Food Chemistry, 141(3), 2558–2566.

548

Santos, M. C., Nunes, C., Rocha, M. A. M., Rodrigues, A., Rocha, S. M., Saraiva, J. A., &

549

Coimbra, M. A. (2013). Impact of high pressure treatments on the physicochemical

550

properties of a sulphur dioxide-free white wine during bottle storage: Evidence for

551

Maillard reaction acceleration. Innovative Food Science & Emerging Technologies, 20,

552

51–58.

553

Schneider, R., Baumes, R., Bayonove, C., & Razungles, A. (1998). Volatile Compounds

554

Involved in the Aroma of Sweet Fortified Wines (Vins Doux Naturels) from Grenache

555

Noir. Journal of Agricultural and Food Chemistry, 46(8), 3230–3237.

556

Schwarzenbolz, U., Klostermeyer, H., & Henle, T. (2002). Maillard reaction under high

557

hydrostatic pressure: studies on the formation of protein-bound amino acid derivatives.

558

International Congress Series, 1245, 223–227.

23

559

Sun, S. Y., Che, C. Y., Sun, T. F., Lv, Z. Z., He, S. X., Gu, H. N., Gao, Y. (2013). Evaluation

560

of sequential inoculation of Saccharomyces cerevisiae and Oenococcus oeni strains on the

561

chemical and aromatic profiles of cherry wines. Food Chemistry, 138(4), 2233–2241.

562

Tabilo-Munizaga, G., Gordon, T. A., Villalobos-Carvajal, R., Moreno-Osorio, L., Salazar, F.

563

N., Pérez-Won, M., & Acuña, S. (2014). Effects of high hydrostatic pressure (HHP) on the

564

protein structure and thermal stability of Sauvignon blanc wine. Food Chemistry, 155,

565

214–220.

566

Tao, Y., Sun, D.-W., Górecki, A., Błaszczak, W., Lamparski, G., Amarowicz, R., Jeliński, T.

567

(2012). Effects of high hydrostatic pressure processing on the physicochemical and

568

sensorial properties of a red wine. Innovative Food Science & Emerging Technologies, 16,

569

409–416.

570 571

Vally, H., & Misso, N. L. A. (2012). Adverse reactions to the sulphite additives. Gastroenterology and Hepatology from Bed to Bench, 5(1), 16–23.

572

Weldegergis, B. T., Crouch, A. M., Górecki, T., & de Villiers, A. (2011). Solid phase

573

extraction in combination with comprehensive two-dimensional gas chromatography

574

coupled to time-of-flight mass spectrometry for the detailed investigation of volatiles in

575

South African red wines. Analytica Chimica Acta, 701(1), 98–111.

576

Weldegergis, B. T., Villiers, A. De, McNeish, C., Seethapathy, S., Mostafa, A., Górecki, T.,

577

& Crouch, A. M. (2011). Characterisation of volatile components of Pinotage wines using

578

comprehensive two-dimensional gas chromatography coupled to time-of-flight mass

579

spectrometry (GC×GC–TOFMS). Food Chemistry, 129(1), 188–199.

580

Welke, J. E., Manfroi, V., Zanus, M., Lazarotto, M., & Alcaraz Zini, C. (2012).

581

Characterization of the volatile profile of Brazilian Merlot wines through comprehensive

582

two dimensional gas chromatography time-of-flight mass spectrometric detection. Journal

583

of Chromatography A, 1226, 124–139.

24

Figure captions Fig 1. GC×GC–ToFMS peak area of the chemical groups identified in white (A) and red (B) wines after 2 months of storage. Different letters indicate significant differences according to ANOVA followed by a Tukey test (p < 0.05). Fig 2. GC×GC–ToFMS peak area of the chemical groups identified in white (A) and red (B) wines after 9 months of storage. Different letters indicate significant differences according to ANOVA followed by a Tukey test (p < 0.05). Fig 3. Biplots in the PC1×PC2 plane combining score plots and loadings plots of the different white (A) and red wines (B), after 9 months of storage, related to the volatile compounds. Attribution of the peak number is shown in Tables S1 for white wines and S2 for red wines in Supplementary Material. Fig 4. Heatmaps (logarithmic normalisation of the GC peak area) for white (A) and red (B) wines of the aldehyde, ketone, furan and acetal compounds. Different intensities correspond to the normalised GC peak areas of each compound (3 replicates). For better readability, only the compounds mentioned in the text are identified in the figure.

Fig 1. GC×GC–ToFMS peak area of the chemical groups identified in all white (A) and red (B) wines after 2 months of storage. Different letters indicate significant differences according to ANOVA followed by a Tukey test (p<0.05).

Fig 2. GC×GC–ToFMS peak area of the chemical groups identified in all white (A) and red (B) wines after 9 months of storage. Different letters indicate significant differences according to ANOVA followed by a Tukey test (p<0.05).

Fig 3. Biplots in the PC1×PC2 plane combining score plots and loadings plots of the different white (A) and red wines (B), after 9 months of storage, related to the volatile compounds. Attribution of the peak number is shown in Tables S1 for white wines and S2 for red wines in Supplementary Material.

Fig 4. Heatmaps (logarithmic normalization of the GC peak area) for white (A) and red (B) wines of the aldehyde, ketone, furan and acetal compounds. Different intensities correspond to the normalized GC peak areas of each compound (3 replicates). For better readability, only the compounds mentioned in the text are identified identified in figure.

38 39 40 41 42 43 44

-

High hydrostatic pressure (HHP) treatments influence long-term white and red wine volatile composition. Pressurised wines had a higher content of furans, aldehydes, ketones, and acetals. HHP treatments promote the Maillard reaction, and alcohol and fatty acid oxidation. HHP enhances characteristics associated with aged and/or thermally treated wines.

45 46

30