Influence of Production Method on the Chemical Composition

Subscriber access provided by University of Newcastle, Australia

Article

Influence of Production Method on the Chemical Composition, Foaming Properties and Quality of Australian Carbonated and Sparkling White Wines Julie Culbert, Jacqui M. McRae, Bruna Condé, Leigh M. Schmidtke, Emily Nicholson, Paul A. Smith, Kate Howell, Paul Kenneth Boss, and Kerry L. Wilkinson J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.6b05678 • Publication Date (Web): 27 Jan 2017 Downloaded from http://pubs.acs.org on January 28, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Agricultural and Food Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35

Journal of Agricultural and Food Chemistry 1

Influence of Production Method on the Chemical Composition, Foaming Properties and Quality of Australian Carbonated and Sparkling White Wines

Julie A. Culbert,1,6 Jacqui M. McRae,2 Bruna C. Condé,3 Leigh M. Schmidtke,4 Emily L. Nicholson,5 Paul A. Smith,2 Kate S. Howell,3 Paul K. Boss5 and Kerry L. Wilkinson1,*

1

School of Agriculture, Food and Wine, The University of Adelaide, Waite Campus, PMB 1, Glen

Osmond, SA, 5064, Australia 2

Australian Wine Research Institute, P.O. Box 197, Glen Osmond SA 5064, Australia

3

Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Parkville, Victoria,

3010, Australia 4

National Wine and Grape Industry Centre, School of Agricultural and Wine Science, Charles Sturt

University, Wagga Wagga, New South Wales, 2678, Australia 5

Commonwealth Scientific and Industrial Research Organisation, Agriculture and Food, PMB2,

Glen Osmond SA 5064, Australia 6

Current address: Australian Wine Research Institute, P.O. Box 197, Glen Osmond SA 5064,

Australia

* Corresponding Author: Dr Kerry Wilkinson, telephone: + 61 8 8313 7360 facsimile: + 61 8 8313 7716, email: [email protected]

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

Page 2 of 35 2

1

Abstract

2

The chemical composition (protein, polysaccharide, amino acid and fatty acid/ethyl ester content),

3

foaming properties and quality of fifty Australian sparkling white wines, representing the four key

4

production methods, i.e. Méthode Traditionelle (n=20), transfer (n=10), Charmat (n=10) and

5

carbonation (n=10), were studied. Méthode Traditionelle wines were typically rated highest in

6

quality and were higher in alcohol and protein content, but lower in residual sugar and total

7

phenolics, than other sparkling wines. They also exhibited higher foam volume and stability, which

8

might be attributable to higher protein concentrations. Bottle-fermented Méthode Traditionelle and

9

transfer wines contained greater proportions of yeast-derived mannoproteins, whereas Charmat and

10

carbonated wines were higher in grape-derived rhamnogalacturonans; however total polysaccharide

11

concentrations were not significantly different between sparkling wine styles. Free amino acids

12

were most abundant in carbonated wines, which likely reflects production via primary fermentation

13

only and/or the inclusion of non-traditional grape varieties. Fatty acids and their esters were not

14

correlated with foaming properties, but octanoic and decanoic acids and their ethyl esters were

15

present in Charmat and carbonated wines at significantly higher concentrations than bottle-

16

fermented wines, and were negatively correlated with quality ratings. Research findings provide

17

industry with a better understanding of the compositional factors driving the style and quality of

18

sparkling white wine.

19 20 21 22 23 24 25

Keywords: amino acids, fatty acids, foaming, polysaccharides, proteins, sparkling wine, wine

26

quality


Page 3 of 35


27

INTRODUCTION

28

The sensory properties of different styles of sparkling white wine are diverse, ranging from

29

predominantly simple, fruity characters to more complex toasty, yeasty and bready notes,1

30

depending on the method of production. In Australia, there are four key production methods,

31

involving either direct infusion of carbon dioxide into base wine (i.e. carbonation) or generation of

32

carbon dioxide via secondary fermentation of base wine, which can occur in pressurized tanks (i.e.

33

Charmat) or in the bottle (i.e. transfer and Méthode Traditionelle).1 Fruit driven styles of sparkling

34

wine are typically derived from carbonation or the Charmat method, whereas more complex

35

sparkling wines are achieved as a consequence of the bottle fermentation and/or lees aging

36

processes employed in the transfer method and Méthode Traditionelle; the latter being analogous

37

with the Méthode Champenoise used for Champagne.

38 39

The sensory and foaming properties of sparkling wine, and therefore wine quality, are strongly

40

influenced by production method, which encompasses factors such as grape variety, the methods by

41

which fruit is harvested and processed, the yeast strains selected for primary and secondary

42

fermentation, the extent to which base wines are fined and filtered, and the duration of lees aging.2

43

During lees ageing, yeast autolysis occurs; i.e. intracellular enzymes slowly hydrolyze yeast to

44

release amino acids, peptides, proteins, polysaccharides (mannoproteins) and fatty acids,2,3 which

45

collectively influence the organoleptic properties of sparkling wine (i.e. aroma, flavor, taste and

46

mouthfeel). As a consequence, a number of studies have explored the compositional changes that

47

occur during sparkling wine production and the factors that influence mouthfeel and foaming

48

properties. Foaming is considered an especially important attribute, and an indicator of wine

49

quality,4,5 since bubbles are not only intrinsic to the sensory appeal of sparkling wine, but the first

50

characteristic observed by consumers.6

51



Page 4 of 35 4

52

In the case of amino acids, sparkling wine-related studies have typically focused on the influence of

53

yeast strain, yeast autolysis and aging during the traditional secondary fermentation process on their

54

concentrations.7–10 In contrast, wine proteins have received considerably more attention, given their

55

role in wine haze formation. Chitinases and thaumatin-like proteins are the two major haze forming

56

proteins present in wine; for which molecular weights range between 21 and 32 kDa.11 In sparkling

57

wine, previous studies have sought to understand the contribution of proteins to wine foaming

58

properties through protein fractionation and characterization,12,13 as well as changes to protein

59

composition during Champagne production14 or following the use of bentonite either for fining base

60

wines15 or as a riddling agent.16 Proteins are generally considered to positively impact wine foam

61

stability13,14,17,18 and to contribute sparkling wine body and quality.12

62 63

Wine polysaccharides also contribute to perceptions of body and mouthfeel,19 and given their

64

influence on viscosity, might also influence sparkling wine foaming properties. Polysaccharides can

65

either be grape or yeast derived, with their size and type largely determining their impact on wine

66

sensory and foaming properties.4,7,20 Grape-derived polysaccharides include type II arabinogalactan-

67

proteins (AGPs) and rhamnogalacturonans type I (RG-I) and type II (RG-II), while mannans and

68

mannoproteins (MPs) are examples of yeast-derived polysaccharides which can be released in wine

69

during fermentation or yeast autolysis.20 Polysaccharides, in particular mannoproteins, have been

70

shown to enhance foaming properties;18,20–22 but mannoproteins can also prevent protein haze23 and

71

potassium bitartrate crystallization,11 thereby improving wine quality.

72 73

Relatively few studies have investigated the occurrence and impact of lipids (fatty acids) on

74

sparkling wine foaming properties, with conflicting results published in the literature. An early

75

study reported a reduction in foam stability following the addition of octanoic and decanoic acids to

76

sparkling wine,24 whereas a subsequent study found the addition of a lipid mixture had no effect on

77

foaming, with alcohol being more influential.25 In a more recent study, octanoic (C8), decanoic


Page 5 of 35


78

(C10) and dodecanoic (C12) acids were found to be negatively correlated with foamability, while

79

ethyl hexanoate, ethyl octanoate and ethyl decanoate were positively correlated, but no effect on

80

foam stability was observed.26

81 82

To date, few studies have considered the compositional and/or organoleptic properties of Australian

83

sparkling wine. This study therefore aimed to profile the compositional variation amongst

84

Australian carbonated and sparkling white wines, and to determine to what extent production

85

method influences the compositional factors that drive sparkling wine style and quality.

86 87

MATERIALS AND METHODS

88

Chemicals. All reagents were analytical grade, unless otherwise stated. Fructose, glucose,

89

thaumatin, commercial dextrans (with molecular weights ranging from 5 to 270 kDa), amino acids,

90

hexanoic acid, octanoic acid, decanoic acid, ethyl hexanoate, ethyl octanoate and ethyl decanoate

91

were obtained from Sigma-Aldrich (Castle Hill, NSW, Australia). Deuterium-labelled internal

92

standards were sourced from CDN Isotopes (Pointe-Claire, CA), except for d5-ethyl nonanoate

93

which was synthesized as previously reported.27 Ethanol (absolute) was purchased from Merck

94

Millipore (Billerica, MA). The AccQ-Fluor Reagent Kit and AccQ-Tag Eluent A used for amino

95

acid derivitization and analysis were purchased from the Waters Corporation (Milford, MA).

96

Ultrapure water was obtained from a Milli-Q purification system (Millipore, North Ryde, NSW,

97

Australia).

98 99

Sparkling Wine Samples. Fifty Australian sparkling white wines were selected with input from an

100

industry reference group comprising of four prominent sparkling winemakers and sourced from

101

producers or retail outlets. The selected wines represented the four key production methods

102

employed in Australia: Méthode Traditionelle (n=20, hereafter designated as MT01–MT20),

103

transfer (n=10, hereafter designated as Tr01–Tr10), Charmat (n=10, hereafter designated as Ch01–



Page 6 of 35 6

104

Ch10) and carbonation (n=10, hereafter designated as Ca01–Ca10), as well as a range of wine

105

regions, price points (being AUD $5 to $90, for 750 mL bottles) and established brands. Méthode

106

Traditionelle sparkling wines comprised 15 vintage (3 x 2004; 1 x 2005; 3 x 2008; 3 x 2009; 4 x

107

2010; 1 x 2011) and 5 non-vintage wines; while 4 transfer wines were vintage (1 x 2008; 3 x 2011)

108

and 6 were non-vintage. Charmat and carbonated wines were non-vintage, with the exception of one

109

carbonated wine (Ca04, vintage 2012). Sparkling wines were predominantly made from

110

Chardonnay, Pinot Noir and Pinot Meunier (i.e. the classic varieties), or blends thereof; except for

111

Ca01, Ca02, Ca05, Ca06 and Ca07, which comprised Chardonnay blended with one or more non-

112

classic varieties (Chenin Blanc, Colombard, Sauvignon Blanc and Semillon). The vintage, varietal

113

composition, geographical origin, price and bottle weight of each sparkling wine is provided as

114

Supporting Information, Table S1. Bottle weights were measured (in duplicate, two separate

115

bottles) with an analytical balance (AUW220D, Shimadzu, Rydalmere, NSW, Australia).

116 117

Basic Wine Composition. Sparkling wine samples were degassed prior to basic compositional

118

analyses (in triplicate, from three separate bottles). Wine (~10 mL) was placed in a loosely capped

119

50 mL Schott bottle and sonicated in an ultrasonic bath (Sonorex Digitec DT 1028F, Bandelin

120

Electronic GmbH & Co. KG, Berlin Germany) for 10 minutes. pH and titratable acidity (TA,

121

expressed as g/L of tartaric acid) were determined using an autotitrator (Compact Titrator, Crison

122

Instruments SA, Allela, Spain), ethanol content (% alcohol by volume, abv) was determined with an

123

alcolyzer (Anton Paar, Graz, Austria), and glucose and fructose (i.e. residual sugar) were

124

determined enzymatically (Boehringer-Mannheim, R-BioPharm, Darmstadt, Germany) with a

125

liquid handling robot (CAS-3800, Corbett Robotics, Eight Mile Plain, Qld, Australia) and

126

spectrophotometric plate reader (Infinite M200 Pro, Tecan, Grödig, Austria). The absorbance of

127

degassed wines (at 280 nm) was measured using a spectrophotometer (GBC Scientific Equipment,

128

Melbourne, Australia) to determine total phenolics.

129


Page 7 of 35


130

Protein Analysis. The concentration of haze-forming proteins, including chitinases and thaumatin-

131

like proteins, were measured in each sparkling wine (in duplicate, from two separate bottles) using

132

ultra-high performance liquid chromatography (UHPLC), as described previously,28 with

133

modifications. Degassed wine samples (1.5 mL) were filter-sterilized (0.45 µm PVDF syringe filter,

134

Merck Millipore, Billerica, MA) prior to injection on an Agilent (Palo Alto, CA) 1260 UHPLC

135

equipped with a Prozap C18 column (10 mm x 2.1 mm, Agilent). Separation was achieved with a

136

solvent system of 0.1% TFA/H2O (Solvent A) and 0.1%TFA/ACN (solvent B) and a flow rate of

137

0.75 mL/min. The mobile phase gradient was: 0–1 min 10–20% B; 1–4 min 20–40% B; 4–6 min

138

40–80% B; 6–7 min 80% B; and 7–10 min 10% B. Proteins were detected at 210 nm, using an

139

Agilent UV/Vis detector. Identification was achieved by comparing retention times with isolated

140

standards28 and quantitation performed against an external standard curve for thaumatin; results are

141

expressed as mg/L thaumatin equivalents.

142 143

Polysaccharide Analysis. Polysaccharides were isolated and analyzed for each sparkling wine (in

144

duplicate, from two separate bottles) using HPLC and a refractive index detector (RID), as

145

previously described.29 In brief, polysaccharides were precipitated from degassed wine (1 mL)

146

following the addition of ethanol (absolute; 5 mL) and overnight chilling (at 4°C). The crude

147

polysaccharide precipitate was isolated as a pellet after centrifugation, then further purified by

148

dialysis (Pur-A-Lyzer Midi 3500, 50–800 µL, Sigma-Aldrich) over 48 hours. Following dialysis,

149

samples were freeze dried and the resulting purified polysaccharide re-suspended in buffer solution

150

(0.1 M sodium nitrate, 150 µL) for HPLC analysis. Separation was achieved with a BioSep Sec

151

2000 column (300 mm x 7.8 mm, Phenomenex, Lane Cove, NSW, Australia). Polysaccharide

152

classes were quantified as: high molecular weight (approximately 200 kDa), medium molecular

153

weight (approximately 100 kDa) and low molecular weight (with molecular distributions of 50, 20

154

and 10 kDa). Polysaccharide concentrations were estimated by comparing peak areas against a

155

standard curve for dextran; results are expressed as mg/L of 50 kDa dextran equivalents.



Page 8 of 35 8

156 157

Amino Acid Analysis. Prior to analysis, amino acids were converted to their highly fluorescent 6-

158

aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC) derivatives using the AccQ-Fluor reagent

159

kit (Waters Corporation, Milford, MA); with each sparkling wine analyzed in duplicate (from two

160

separate bottles). Wine samples (50 µL) were mixed with α-aminobutyric acid (0.5 mM in Milli-Q

161

water, 50 µL) and sodium borate buffer (0.2 M, pH 8.8, 900 µL) and a portion of the mixture (20

162

µL) was subsequently transferred to a 2 mL high recovery vial (Agilent Technologies) containing

163

AccQ-Fluor Borate Buffer (60 µL) and reconstituted AccQ-Fluor reagent (20 µL, prepared by

164

adding AccQ-Fluor reagent diluent (1 mL) to the AccQ-Fluor reagent powder, with vortexing for 10

165

sec and heating at 55°C for 10 min). The resulting mixture was incubated for 1 min at ambient

166

temperature before heating at 55°C for 10 min, and then quantitation. HPLC Method: The

167

derivatized amino acids were analyzed on an Agilent 1200 series HPLC equipped with a

168

fluorescence detector. Separation was achieved using a Phenomenex (Torrance, CA) Luna 3 µm

169

C18 column (150 x 4.6 mm) fitted with a guard cartridge. Mobile phase A consisted of AccQ-Tag

170

Eluent A (100 mL concentrate added to 1L Milli-Q water); mobile phase B consisted of 60%

171

acetonitrile in Milli-Q water. The injection volume was 10 µL (with a needle wash with water

172

between samples) and the pump flow rate was constant at 1 mL/min. The solvent gradient was as

173

follows: 0% B to 2% B (0.5 min); 2% B to 7% B (14.5 min); 7% B to 13% B (7 min); 13% B to

174

32% B (19 min); 32% B to 40% B (5 min); 40% B to 85% B (9 min); 85% B to 100% B (1 min);

175

100% B isocratic for 5 min; 100% B to 0% B (2 min); 0% B isocratic for 10 min; giving a total

176

runtime of 73 min. Fluorescence detector parameters comprised excitation at 250 nm and emission

177

at 395 nm. Amino acids were quantified against calibration solutions of known concentrations.

178 179

Fatty Acid (C6, C8 and C10) and Ethyl Ester Analysis. Fatty acids and their ethyl esters were

180

determined in each sparkling wine in triplicate (from three separate bottles). Wine samples (0.5 mL)

181

were placed in 20 mL autosampler vials containing sodium chloride (2.0 g) and Milli-Q water (4.5


Page 9 of 35


182

mL). To this was added an internal standard solution containing d4-methyl-1-butanol, d3-hexyl

183

acetate, d13-1-hexanol, d5-ethyl nonanoate, d5- phenethyl alcohol and d19-decanoic acid (10 µL), to

184

give final concentrations of 48.0, 0.5, 1.0, 0.084, 10.0 and 1.0 mg/L, respectively. Vials were sealed

185

and mixed thoroughly prior to GC-MS analysis. GC-MS Method: Samples were analyzed using a

186

7890A gas chromatogram coupled to a 5975C inert XL mass selective detector (Agilent

187

Technologies, Santa Clara, USA) and equipped with a Gerstel MPS2 Multipurpose sampler

188

(Gerstel, Mülheim an der Ruhr, Germany). The GC-MS instrument was controlled using

189

ChemStation software in combination with Gerstel Maestro software. Samples were incubated with

190

agitation for 10 min at 50°C prior to headspace-solid phase microextraction (HS-SPME) for 30 min

191

at 50°C (with agitation) using a Supelco 50/30um DVB/CAR/PDMS 1 cm SPME fiber. The SPME

192

fiber was desorbed in the GC inlet containing an ultra-inert glass SPME liner (straight taper with

193

0.75 mm i.d.) operated in splitless mode at a temperature of 240°C. The SPME fiber remained in

194

the inlet for 10 min but with a purge flow to the split vent of 20 mL/min after 3 min. Separation of

195

volatiles was achieved using a J&W DB-WAXetr capillary column (60 m x 0.25 mm i.d. x 0.25 µm,

196

Agilent Technologies) with a constant carrier gas (ultrahigh purity helium) at a flow rate of 1.5

197

mL/min. The oven program comprised: 40°C (held for 5 min); 2°C/min until 210°C (held for 5

198

min); and then 5°C/min until 240°C (held for 10 min); giving a total runtime of 111 min. The MS

199

was operated using positive ion electron impact at 70 eV in full scan mode (m/z 35–350), with MS

200

source and quadrupole temperatures of 230°C and 150°C, respectively. The MS transfer line was

201

held at 240°C. GC-MS data (3D) was exported to Excel and subjected to multivariate curve

202

resolution and alternating least squares (MCR-ALS) analysis, using methods described

203

previously.30 Briefly all chromatogram TICs were overlaid and inspected for elution time segments

204

suitable for batch processing based upon stable baseline and peak heights. Segments were processed

205

by smoothing each m/z channel; aligning the resulting TIC and applying the alignment to the three

206

dimensional data cube before feature extraction. The number of features in each segments was

207

determined from visual inspection of the data. The MCR-ALS deconvolution commenced using an



Page 10 of 35 10

208

initial estimate of the spectra for each feature obtained using SIMPLISMA.31 Peak areas of

209

identified features and spectral libraries derived from MCR-ALS were exported from MATLAB

210

(version 7.4.0.287 R2007a) in a format compatible with the National Institute of Standards and

211

Technology (NIST) mass spectral search program (version 2.0). Compound identification was

212

achieved using the NIST 05 Mass Spectral library database, as well as by comparing retention times

213

and mass spectra to those of known standards. Areas for: octanoic acid and decanoic were corrected

214

using d19-decanoic acid; hexanoic acid using d13-hexanol; and ethyl hexanoate, ethyl octanoate and

215

ethyl decanoate using d3-hexyl acetate.

216 217

Foaming Analysis. Two parameters representative of foam stability and foamability were

218

calculated, being the maximum volume of foam (Vf) and the average lifetime of foam (Lf), using a

219

robotic pourer and image analysis according to previously described methodology.32 Measurements

220

were performed (in duplicate, from two separate bottles) on chilled wines (5°C), with the exception

221

of one Méthode Traditionelle wine which could not be measured because the bottle did not fit the

222

pourer. Data for this production method therefore represents 19 wines only.

223 224

Quality Ratings. Quality ratings were measured as described previously.33 Briefly, an expert panel

225

comprising sparkling winemakers and wine show judges (n=19) rated the quality of each sparkling

226

wine (with 5 wines presented in duplicate to validate reproducibility) using the 20 point scoring

227

system employed in Australian wine shows;34 with wines presented to panelists as brackets of 5

228

freshly poured, chilled (5 ºC) wines, in random order, in three digit coded XL5 (ISO standard) wine

229

glasses.

230 231

Statistical Analysis. Data were analyzed using a combination of descriptive and multivariate

232

techniques, including Analysis of Variance (ANOVA) with post-hoc Tukey's test at P

Influence of Production Method on the Chemical Composition

Recommend Documents