Properties of Wine Polymeric Pigments Formed from Anthocyanin and

Subscriber access provided by Library, Special Collections and Museums, University of Aberdeen

Article

Properties of wine polymeric pigments formed from anthocyanin, and tannins differing in size distribution and subunit composition Keren A. Bindon, Stella Kassara, Yoji Hayasaka, Alex Schulkin, and Paul A. Smith J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/jf503922h • Publication Date (Web): 30 Oct 2014 Downloaded from http://pubs.acs.org on November 14, 2014

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 46

Journal of Agricultural and Food Chemistry

1

Properties of wine polymeric pigments formed from anthocyanin,

2

and tannins differing in size distribution and subunit composition.

3

KEREN BINDON*, STELLA KASSARA, YOJI HAYASAKA, ALEX SCHULKIN AND

4

PAUL SMITH

5 6

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

7 8

* Corresponding author

9

Tel: +61-8-83136190

10

Fax: +61-8-83136601

11

E-mail: [email protected]

12 13

ACS Paragon Plus Environment

1


14

Page 2 of 46

Abstract

15

To explore the effect of tannin composition on pigment formation, model ferments of

16

purified 3-O-monoglucoside anthocyanins (ACN) were conducted either alone or in the presence

17

of two different tannins. Tannins were isolated from grape seeds (Sd) or skins (Sk) following

18

exhaustive extraction in 70% v/v acetone. The Sd and Sk tannin fractions had a mean degree of

19

polymerization of 5.2 and 25.6 respectively. The Sd fraction was highly galloylated, at 22%, but

20

galloylation was < 2% in the Sk fraction. The Sk fraction was distinguished by a high degree of

21

trixydroxylation, at 58%. After a 6 month aging period, polymeric pigments were quantified and

22

their color properties determined following isolation by solid-phase extraction. Wine color and

23

polymeric pigment was highest in the treatment containing ACN+Sd, and similar in the ACN+Sk

24

and ACN treatments. The same trend between treatments was observed for total and polymeric

25

non-bleachable pigments. Only minor changes in tannin subunit composition were found

26

following ACN incorporation, but the size distribution of polymeric pigments determined by gel

27

permeation chromatography decreased, in particular for the ACN+Sk treatment. Color

28

incorporation in the higher molecular mass range was lower for ACN+Sk wines than ACN+Sd

29

wines. Compositional differences between the two tannin fractions may therefore limit the

30

incorporation of ACN in the colored form. The results suggest that in the ACN+Sk and ACN

31

treatments, the formation of lower molecular mass oligomeric pigments was favored. In

32

polymeric pigments derived from ACN, the presence of ethyl- and vinyl-linked ACNs to the

33

level of trimers were identified using mass spectrometry.

34 35

Keywords: proanthocyanidin; anthocyanin; tannin; color; pigment; bisulfite; wine; ethyl linkage;

36

vinyl linkage; molecular mass

37 ACS Paragon Plus Environment

2

Page 3 of 46

38


Introduction

39

In red wines produced from V. vinifera grapes, monoglucoside anthocyanins (ACN) react

40

during fermentation and aging to yield a stable red-purple color. Stable wine color is conferred

41

by the presence of pigments which retain their color at wine pH and demonstrate an enhanced

42

resistance to bleaching by bisulfite. Such pigments are diverse, and can form via multiple

43

chemical pathways.

44

The reaction of ACN with condensed tannins (proanthocyanidins) extracted from the

45

grapes during fermentation is thought to contribute significantly to the formation of polymeric

46

pigments, which are thought to make a significant contribution to the stabilization of wine color1.

47

Condensed tannins in wine originate from the skin and seed components of the grape, and the

48

degree to which they are extracted depends upon the grape cultivar, environmental conditions

49

during development, grape ripeness stage, winemaking technique and wine ethanol concentration

50

2-5

51

and contains prodelphinidin as an extension subunit6 while seed tannins contain a higher

52

proportion of galloylated flavan-3-ol units7. This structural diversity in grape-derived tannin

53

potentially defines its reactivity to ACN, or other grape phenolics8.

. Skin tannin is characterized by a higher degree of polymerization (mDP) than seed tannin,

54

Studies on the formation of polymeric pigments in model systems thus far have focused

55

on the addition of different concentrations of seed tannin, in combination with purified

56

monoglucoside ACNs and monomeric flavan-3-ols1, 8. When seed tannin and ACN were

57

combined in white wine medium, polymeric pigment formation was dependent upon the ratio of

58

tannin to ACN, increasing incrementally as the concentration of tannin increased1. Interestingly,

59

ACN was found to yield colored pigment that had a similar retention time to tannin on reversed-

60

phase HPLC, suggesting a polymeric form, although the yield of polymeric pigment was lower


3


Page 4 of 46

61

than when seed tannin was present1. The medium in which ACN reactions occur has been found

62

to underpin pigment formation8, 9. It was found that polymeric pigment formation only occurred

63

in wine1, but not in model wine solution, suggesting the requirement of a wine-derived co-

64

factor8. Although direct condensation products of ACNs can exist in grapes, the important

65

contribution of pyrano-ACNs and their condensation products with flavanols (vinyl-linked ACN

66

with flavanols) has been demonstrated in red wines, and these may be intermediates in polymeric

67

pigment formation

68

products has led to the suggestion that the presence of pyruvic acid, acetaldehyde and their

69

respective reaction products in wine may present the crucial co-factors in pigment formation.

70

Indeed, the direct polymerization of ACNs has been demonstrated in the presence of

71

acetaldehyde9, 15. However, similar to studies of ACN-tannin interactions in wine medium1, the

72

degree of pigment coloration of oligomeric ACN in the presence of acetaldehyde was lower than

73

when ACN was combined with flavan-3-ols

74

found to have little effect on the kinetics of ACN loss and the color of pigments produced,9, 15

75

potentially indicating a minor role for tannin molecular mass in conferring a specific reactivity

76

with ACN.

10-14

. Detection of vinyl and ethyl linkages in these ACN and flavanol

9, 15

. The degree of flavan-3-ol polymerization was

77

Yet, as highlighted above, skin tannin has a far higher mDP than seed tannin, and the

78

relative reactivities of tannins of different size and composition for ACN have not yet been

79

studied. While the detailed identification and quantification of low molecular mass ACN

80

derivatives in model and commercial wines has been achieved using mass spectroscopy14, 16, 17

81

the pigmented tannin fraction is only partially resolved using LC-MS18. A review11 has

82

highlighted that research in this area is limited by a lack of information on the molecular mass

83

distribution of grape flavanols, and their reaction products in wine. In particular, there is limited


4

Page 5 of 46


84

information regarding grape tannins and their pigmented reaction products with ACN. A recent

85

report has indicated that the progression of red color incorporation to wine tannin is skewed in

86

favor of smaller pigments in young wines, with a gradual conversion of larger, unpigmented

87

tannin to smaller colored species during aging

88

partitioning of ACN to either complex with tannin or form low molecular mass pigments are

89

currently unknown, leaving scope for aspects of earlier experiments 1, 9, 15 to be revisited.

19

. This highlights that the factors which drive

90

It remains to be established how non-bleachable pigments, including polymeric pigments,

91

form in the presence of tannins of very different structural characteristics. Using traditional

92

techniques for colorimetric analysis of wines

93

characterize wine pigmented polymers by solid-phase extraction (SPE)

94

and gel permeation chromatography (GPC) 24 these knowledge gaps have been explored.

20, 21

in conjunction with newer methods to 22

, phloroglucinolysis

23

95 96

Materials and Methods

97

Instrumentation. An Agilent model 1100 HPLC (Agilent Technologies Australia Pty Ltd,

98

Melbourne, Australia) was used with Chemstation software for chromatographic analyses.

99 100

Grape tannin isolation. Grapes were obtained from a 10-year old commercial Vitis vinifera L.

101

cv. Cabernet Sauvignon vineyard in the Langhorne Creek growing region of South Australia.

102

Grapes were harvested corresponding with a commercial harvest point of ≅ 23 °Brix in the 2009

103

season. Skin and seed components were prepared from a 300-berry sample, and were visually

104

free of flesh material at the point of extraction. Tannin was extracted from frozen skins and seeds

105

in 70% v/v acetone:water for 18 h and then purified from acetone extracts as described

106

previously 25. Extracts were concentrated under reduced pressure at 35 °C to remove acetone and


5


Page 6 of 46

107

then extracted with hexane to remove residual lipophilic material. The aqueous fraction was then

108

recovered using a separatory funnel and made up to a final concentration of 60% v/v methanol

109

containing 0.05% v/v trifluoroacetic acid (TFA) and applied (~18.3 mL/min) to a 300 mm x 21

110

mm glass column (Michel-Miller, Vineland, NJ, USA) containing Sephadex LH20

111

chromatography resin (Amersham, Uppsala, Sweden) to a bed volume of 93 mL. Low molecular

112

weight phenolics were eluted with 300 mL 60% (v/v) HPLC grade methanol containing 0.05%

113

v/v TFA. Tannin was then eluted with 250 mL 70% v/v acetone containing 0.05% v/v TFA. The

114

column was re-equilibrated with 60% v/v methanol containing 0.05% v/v TFA after each

115

sample. The tannin fractions eluted were concentrated under reduced pressure at 35°C to remove

116

organic solvent and the aqueous fraction recovered (≅ 5 mL). The aqueous fraction was frozen at

117

-80 °C and lyophilized to a dry powder. Dried tannin isolates were stored in the dark and under

118

nitrogen at -20 °C prior to use.

119 120

Grape anthocyanin isolation. Shiraz grapes were sourced from the Coombe Vineyard, University

121

of Adelaide, South Australia at commercial ripeness (24 °Brix) in the 2011 season. Whole bunches

122

were frozen at -20°C until used. A 1 kg bunch sample was de-stemmed, allowed to thaw at room

123

temperature, crushed and partially homogenized using a blender. A 250 g sample of the coarse

124

homogenate was extracted for 18 h with 2 L of 50 % v/v MeOH, filtered and concentrated under

125

reduced pressure at 30 °C. The aqueous extract was loaded in 100 mL batches to a Toyopearl (660

126

mL bed volume) column equilibrated with 0.1% v/v aqueous TFA, and water-soluble non-

127

adsorbent compounds were eluted in a further 2 L of 0.1% v/v TFA. The 3-O-monoglucoside ACN

128

fraction was eluted with 1.5 L 30% v/v MeOH containing 0.1% v/v aqueous TFA. The column was

129

then washed with 2 L 2:1 v/v acetone:water and re-equilibrated with 2 L 0.1% v/v aqueous TFA.


6

Page 7 of 46


130

The eluate was concentrated under reduced pressure, frozen at -80°C and then lyophilized, yielding

131

≅ 200 mg of dry 3-O-monoglucoside ACNs which were stored under nitrogen at -80 °C until used.

132

Extract purity was verified by HPLC

133

monomeric flavonoids, and polymeric material. It was found to have trace quantities of early-

134

eluting material at 280 nm, but this was below the detection limit for quantification. A calibration

135

curve was established for the gravimetric concentration of the ACN isolate and the concentration

136

estimated spectrophotometrically according to 26 and is shown as supporting information (S1). This

137

was in order to approximate the concentration of ACN addition (using the spectral method) to the

138

concentration expected in grapes.

26

and was free of acetylated or coumaroylated ACNs, other

139 140

Treatments. Chemically-defined medium (CDM) was based on the compositional specifications

141

described by Schmidt et al. (2011)27. All components used to prepare CDM were from Sigma

142

Aldrich, St. Louis, MO, USA. CDM was prepared in MilliQ water as described by Contreras et

143

al. (2014)28 and contained (per L): glucose, 100 g; fructose, 100 g; citric acid, 0.2 g; malic acid, 3

144

g; potassium hydrogen tartrate, 2.5 g; K2HPO4, 1.1 g; MgSO4·7H2O, 1.5 g; CaCl2·2H2O, 0.4 g;

145

H3BO3, 0.04 g and proline, 0.84 g. To this solution, stock solutions (described below) of nitrogen

146

(20 mL), trace elements (1 mL) and vitamin (1 mL) were added, adjusted to pH 3.5 and sterile-

147

filtered. The nitrogen stock solution contained the following (per L): 30% ammonium hydroxide

148

solution, 27.7 g; alanine, 10.5 g; γ-amino butyrate, 7.2 g; arginine, 26.9 g; asparagine, 0.4 g;

149

aspartate, 3.0 g; citrulline, 0.4 g; glutamate, 6.0 g; glutamine, 8.4 g; glycine, 0.4 g; histidine, 1.2

150

g; isoleucine, 1.2 g; leucine, 1.2 g; lysine, 0.4 g; methionine, 0.4 g; ornithine, 0.4 g;

151

phenylalanine, 0.8 g; serine, 5.4 g; threonine, 6.0 g; tryptophan, 0.4 g; tyrosine, 0.4 g; valine, 2.1

152

g; and cysteine, 1.2 g. The trace elements stock solution contained (per L): MnSO4·H2O, 3.5 g;


7


Page 8 of 46

153

ZnCl2, 1.0 g; FeSO4·7H2O, 6.0 g; CuSO4·5H2O, 1.5 g; KIO3, 0.01 g; Co(NO3)2·6H2O, 0.03 g;

154

Na2MoO4·2H2O, 0.025 g; LiCl, 0.1 g; NiSO4·6H2O, 0.05 g; and RbCl, 0.7 g. The vitamin stock

155

solution contained (per L): thiamine HCl, 0.5 g; riboflavin, 0.2 g; pyridoxine HCl, 1.0 g;

156

calcium D-pantothenate, 1.0 g; nicotinic acid, 1.0 g; myo-inositol, 10 g; biotin, 0.05 g; folic acid,

157

0.05 g; and 4-amino benzoic acid, 0.05 g. Following sterile-filtration, fatty acid and sterols (both

158

1 ml) stock solutions were added. The fatty acid stock solution contained (per L absolute

159

ethanol): palmitic acids, 2.0 g; oleic acid, 1.0 g; linoleic acid, 3 g; and linolenic acid, 0.5 g. The

160

sterol stock solution contained 1 g/L β-sitosterol in absolute ethanol.

161

Purified ACN, skin (Sk) and seed (Sd) tannin extracts were prepared as separate solutions

162

in either sterile CDM or as a weakly acidified solution in 15% v/v ethanol containing 0.01% v/v

163

TFA. Triplicate treatments were prepared with or without additions according to the

164

experimental outline in Figure 1 to a final volume of 30 mL in 50 mL centrifuge tubes. Target

165

final gravimetric ACN and tannin additions in 30 mL CDM or 15% v/v acidified ethanol were to

166

achieve concentrations of 1.15 g/L and 1.5 g/L respectively. For 15% v/v ethanol 0.01% v/v TFA

167

treatments, additions were made at the same time, tubes were sealed and mixed by shaking for 24

168

h prior to analysis. For ferments in CDM, ACN additions were made as a 5 mL aliquot in CDM

169

prior to inoculation of yeast. Tannin addition was made as a 5 mL aliquot in CDM 24 h after

170

inoculation to minimize the risk of yeast inhibition by excess tannin. The difference in the timing

171

and type of additions was not found to affect fermentation kinetics. For CDM ferments, the yeast

172

added was S. cerevisiae PDM (Maurivin, Sydney, Australia) at 200 ppm. After inoculation with

173

yeast, tubes were sealed with a lid fitted with a gas release port. Ferments were carried out open

174

to atmosphere for 24 h at 22 °C in a temperature-controlled room, then sealed in plastic

175

containers fitted with an airlock to continue fermentation. Glucose concentration was tracked


8

Page 9 of 46


176

during ferment using the Keto Diabur 5000 test strips (Roche Diagnostics, NSW, Australia), and

177

following each sampling point, dry ice was added to the plastic ferment receptacles to minimize

178

oxidation. When glucose fermented to dryness, a final aliquot was removed and tested for

179

residual fructose and final ethanol concentrations by HPLC29. Fructose concentration was < 0.5

180

% w/v and ethanol was 10 % v/v for all treatments, indicating complete fermentation. Although

181

the target alcohol concentration of CDM ferments is lower than that of red wine, the approach

182

was selected in favor of bottled wine or fresh grape juice in order to conduct the experiments

183

under controlled conditions, in accordance with published experiments28, enabling comparison

184

and replication in future studies. Model ferments did not undergo malolactic fermentation, were

185

sealed under CO2 and cold-settled at 4 °C. Completed ferments were centrifuged at 1730 g for 5

186

min. Aliquots of wines and 15% v/v ethanol controls were sealed in glass vials with no ullage

187

and placed in kegs that were fitted with a PreSens Pst6 oxygen sensor (Presens, Regensburg,

188

Germany) and flushed with nitrogen. This system is described in detail as supporting information

189

(Supporting information S2). A slight positive pressure of nitrogen was maintained and the

190

oxygen concentration in the kegs maintained below 10 ppb. Samples were stored anaerobically

191

for 6 months at 22 °C.

192 193

Analysis of tannin, anthocyanin and pigments. General analyses performed are summarized in

194

Figure 1. Treatments at 0 and

195

concentration was analyzed in duplicate according to the published methyl cellulose precipitation

196

assay26. ACN and pigment analysis was only done in treatments containing ACN, except for the

197

Sk treatments. For Sk treatments with ACN addition, the Sk control was included as sample

198

blank due to a minor contribution of color (but not monomeric ACN) which resulted from the

6 months were centrifuged 5 min at 16 000 g. Tannin


9


Page 10 of 46

199

tannin preparation step. Sample aliquots were analyzed by HPLC according to the method

200

described by Mercurio et al., (2007)26. ACNs, derived pigments and polymeric pigments were

201

quantified using their peak areas at 520 nm as malvidin-3-glucoside units based on a commercial

202

standard (Polyphenols Laboratories, Norway). Vitisin A was identified according to its retention

203

time and UV-visible spectrum (200-600 nm) in comparison with a synthesized standard30

204

(Supporting information S3). Vitisin B was not detected under the chromatographic conditions.

205

The polymeric fraction, co-eluting with tannin26 as well as other non-polymeric red pigments

206

apparent by HPLC with absorbance maxima at 520 to 540 nm, and 270 nm (Supporting

207

information S3) were quantified as malvidin-3-glucoside equivalents. After 6 months, polymeric

208

pigments from triplicate experiments were purified from wine samples using SPE22 with the

209

modifications outlined in Kassara and Kennedy (2011)31. Duplicate extractions were done on 1

210

mL wine aliquots as preparation for different types of analysis. The first of the duplicate SPE

211

extracts was reconstituted in 1 mL of model wine buffer containing 0.5% (w/v) tartaric acid and

212

12% (v/v) ethanol (pH 3.4). Color characteristics of unpurified treatments and SPE isolates were

213

determined using a published protocol26, for which a brief description follows. All spectral

214

analysis was performed in duplicate in a 370 µL 96 well UV plate, and 420 nm and 520 nm

215

absorbances were measured using a SpectraMax M2 Microplate Reader (Molecular Devices,

216

Australia). Readings were corrected using an appropriate sample blank and a water constant

217

correct function to approximate a 1 cm path length. For determination of total ACN, a 20 µL

218

sample aliquot was added to 980 µL of 1 M HCL, and mixed. Samples were left to stand in the

219

dark for 1 h and a 300 µL aliquot analyzed. The 1 h period was modified from the original

220

method stipulating 3 h, as we had demonstrated that maximal ionization of ACNs was achievable

221

within this period (data not shown). A further 100 µL sample aliquot was added to 900 µL of


10

Page 11 of 46


222

model wine buffer, mixed, and a 300 µL aliquot analyzed immediately. Buffer solution was also

223

prepared containing 0.375% (w/v) sodium metabisulfite, 900 µL was combined with a 100 µL

224

sample aliquot, mixed, left to stand in the dark for 1 h. Thereafter, a 300 µL aliquot was

225

analyzed. Estimates of ACN, were determined according to the calculations outlined in the

226

published method26. The 520 nm analysis in the respective solutions were used to determine

227

pigment color at pH 3.4, in 1M HCl and in the presence of sodium metabisulfite (SO2-resistant

228

pigment). The 420 nm measure was not found to vary significantly during the limited aging

229

period, and was thereafter not included in the assessment of pigment color. Further, the SPE

230

isolate reconstituted in model wine was analyzed for protein-precipitability of non-bleachable

231

pigments using a published protocol 32.

232

The second of the duplicate SPE extracts, as well as a 1 mL aliquot of the treatments stored

233

in 15% v/v ethanol were dried under a stream of nitrogen at 30 °C and reconstituted in 100 µL

234

methanol. Lyophilized powders of purified Sk and Sd tannins (pre-ferment) were reconsituted in

235

methanol to a gravimetric concentration of 10 g/L and analyzed in duplicate, as described for the

236

other samples. In wines, a small amount of precipitate was observed after 6 months aging, which

237

was not found for tannin solutions stored in 15% v/v acidified ethanol. The wine layer was

238

carefully removed using a glass dropper in order to recover the precipitate. Wine precipitates

239

were dried under a stream of nitrogen at 30 °C and reconstituted in methanol. Methanolic

240

solutions of purified tannins, SPE isolates, dried (aged) ethanol solutions and precipitates were

241

analyzed by phloroglucinolysis and gel permeation chromatography (GPC) using published

242

protocols 25, 33. For phloroglucinolysis, (-)-epicatechin (Sigma Aldrich, St. Louis, MO, USA) was

243

the quantitative standard. To determine the proportion of wine tannin as precipitate, or retained

244

in solution, the assumption was made that conversion yield differences would not differ


11


Page 12 of 46

245

significantly between the two fractions. Therefore, the proportion (%) of total wine tannin as

246

precipitate was calculated directly from phloroglucinolysis product yields (on a mg/L wine

247

volume basis). For GPC, separate calibration curves were established for skin and seed tannins

248

using molecular mass standards reported previously 33. This decision was based on our previous

249

observation that seed and skin tannins differ in their hydrodynamic volume. In addition,

250

oligomeric ACN-derived pigments and a standard of the chloride of malvidin-3-O-glucoside of

251

molecular mass 528.9 g/mol (Polyphenols Laboratories, Norway) were analyzed by GPC and

252

nominal molecular mass values were assigned using the skin tannin calibration.

253 254

Mass spectrometric analysis. The SPE isolates of the 0 and 6 month aged ACN wines, and

255

purified ACNs were mixed with the same portion of 0.25% v/v formic acid in 50% v/v methanol,

256

and then analysed by direct infusion-MS and MS/MS. The sample was infused to an electrospray

257

(ESI) interface at a flow rate of 5 µL/min using a syringe pump (Harvard, Holliston, MA).Mass

258

spectrometric analysis was carried out using a 4000 QTrap mass spectrometer equipped with a

259

Turbo ion source (AB Sciex, Mt Waverly, Victoria, Australia). Data acquisition and processing

260

were performed using Analyst software 1.4.2 (AB Sciex). All ESI mass spectrometric data were

261

obtained in positive ion mode. Nitrogen was used as curtain and nebulizer gases set at 10 and 12

262

(arbitrary units), respectively. The electrospray and declustering potentials were set at 5000 V

263

and 120 V, respectively. Mass spectra were obtained by linear ion trap mode. Full scan spectrum

264

ranging from m/z 300 to 2500 and product ion spectrum from precursor ions of interest were

265

consecutively accumulated until approximate ion intensity was obtained. For MS/MS, nitrogen

266

was used as a collision gas setting at ‘High’ and the collision energy was optimized in the range

267

30-60 eV as appropriate.


12

Page 13 of 46


268 269

Statistical analysis. Significant differences between treatments were determined by one-way

270

analysis of variance (ANOVA). Where ANOVA data were significant at p < 0.05 and follow-up

271

statistics were required, differences between sample sets were analyzed by a post-hoc Student’s

272

T-test. Dunnett’s test was used for comparison of multiple treatments against a single pre-

273

treatment control. The JMP 5.0.1 statistical software package (SAS, Cary, NC, USA) was used

274

for all statistical analysis.

275

Results and Discussion

276 277

Tannin concentration and subunit composition. Tannin compositional information determined

278

by phloroglucinolysis is shown in Table 1. According to the respective subunit yields by

279

phloroglucinolysis, the Sd tannin extract prior to fermentation and aging had a higher conversion

280

to component subunits on a w/w basis (% mass conversion) than the corresponding Sk tannin

281

extract, at 76% and 65% respectively (Table 1). It has been shown in various studies that seed

282

and skin tannins differ in their conversion yields from phloroglucinolysis by ripeness stage and

283

grape variety

284

different, with Sd tannins having a far lower mDP as determined by phloroglucinolysis compared

285

with Sk tannin. Sd tannins were distinguished by a high degree of galloylation (Table 1), which

286

was low in Sk tannin at

Properties of Wine Polymeric Pigments Formed from Anthocyanin and

Recommend Documents