Strecker Aldehyde Formation in Wine: New Insights into the Role of

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Strecker Aldehydes Formation in Wine: New insights of the role of gallic acid, glucose and metals in phenylacetaldehyde formation Ana Rita Monforte, Sara Martins, and Antonio Cesar Silva Ferreira J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b00264 • Publication Date (Web): 26 Feb 2017 Downloaded from http://pubs.acs.org on February 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 32

Journal of Agricultural and Food Chemistry

Title: Strecker Aldehydes Formation in Wine: New insights of the role of gallic acid, glucose and metals in phenylacetaldehyde formation. A. R. Monfortea, S.I.F.S. Martinsc, d, A.C. Silva Ferreira a, b, e

a

Universidade Católica Portuguesa, CBQF - Centro de Biotecnologia e Química Fina.

– Laboratório Associado, Escola Superior de Biotecnologia, Rua Arquiteto Lobão Vital, Apartado 2511, 4202-401 Porto, Portugal. b

IWBT – DVO University of Stellenbosch, Private Bag XI, Matieland 7602, South

Africa. c

Food Quality & Design group, Wageningen University, The Netherlands

d

Unilever R&D Vlaardingen, 3130 AC Vlaardingen, The Netherlands.

e

Cork Supply Portugal, S.A., Rua Nova do Fial, 4535, Portugal.

Corresponding author: António César da Silva Ferreira Phone: +351 22 558 0001 Fax: +351 22 509 0351 E-mail address: [email protected]

ACS Paragon Plus Environment


1

Abstract

2

Strecker degradation (SD) leading to the formation of phenylacetaldehyde (PA) was

3

studied.in wine systems. New insights were gained by using two full factorial designs

4

focusing on the effect of: 1) pH and 2) temperature. In each DoE three factors:

5

glucose, gallic acid and metals at two levels (present or absence) were varied while

6

phenylalanine was kept constant. The obtained results gave a clear indication, with

7

statistical significance, that in wine conditions, the SD occurs in the presence of

8

metals preferentially via the phenolic oxidation independently of the temperature

9

(40°C or 80°C). The reaction of the amino acid with the o-quinone formed by the

10

oxidation of the gallic acid seems to be favored when compared with the SD

11

promoted by the reaction with α-dicarbonyls formed by MR between glucose and

12

phenylalanine. In fact, kinetic results showed that the presence of glucose, had an

13

inhibitory effect on PA rate of formation. PA formation was 4 times higher in the

14

control wine when compared to the same wine but with 10 g/L of glucose added. By

15

gallic acid quinone quantitation it is shown that glucose affects directly the

16

concentration of the quinone. decreasing the rate of quinone formation. This

17

highlights the role of sugar on the o-quinones concentration and consequently on the

18

impact on Strecker aldehydes formation, a promising new perspective regarding wine

19

shelf-life understanding.

20 21

Keywords: Strecker degradation, phenylacetaldehyde, phenolic oxidation, sugar,

22

phenolic, o-quinone

23


Page 2 of 32

Page 3 of 32


24

Introduction

25

In foods and in particular in wines, phenylacetaldehyde, a Strecker aldehyde, is

26

reported to be particularly important, contributing to “honey” aroma notes associated

27

to “premature oxidation”.1, 2 Phenylacetaldehyde has been identified in several types

28

of wines and reported to have an aroma threshold between 1.03 and 254 µg/L.

29

Since Hodge review (1953)5 different studies have shown that besides reducing

30

sugars, other dicarbonyls such as o-quinones formed by the oxidation of polyphenols6

31

and lipids7 can also participate in Strecker degradation (SD) to produce flavor

32

compounds.

33

The formation of Strecker aldehydes in wines is suggested to occur (i) by Strecker

34

degradation of parent amino acid with α-dicarbonyl8 or an o-quinone9, 10 (ii) by the

35

direct oxidation of the corresponding alcohol with a peroxidation mechanism11 or (iii)

36

by the release of these compounds from them adducts with sulfur dioxide - the

37

hydroxyalkylsulphonic acid2. Several intermediaries of this pathways have been

38

identified in wines such as: hydroxyalkylsulphonic acids,12 o-quinones,10 as well as 3-

39

deoxyosone13 an α-dicarbonyl generated by the degradation of the Amadori

40

rearrangement through the enolization pathway.

41

In wine conditions Strecker aldehydes formation has been reported to be mainly due

42

to the presence of metals and oxygen, during the phenolic oxidation. Metals will act

43

as electron donors by adding a singlet electron to triplet oxygen (O2). This transfer of

44

an electron leads to a cascade of reactions that ends with a very reactive oxidant, the

45

hydroxyl radical (HO●), which can abstract a hydrogen atom from organic compounds

46

to produce water the final oxygen reduction product.14, 15 Metals will also be able to

47

convert the phenolic compounds into o-quinones, which are highly reactive and can

48

combine spontaneously with nucleophilic compounds such as amino acids resulting in



49

the formation of aldehydes.16 Rizzi6 studied under phenolic oxidation conditions, the

50

formation of PA at pH 7 with K3Fe(CN)6, phenylalanine and several phenolic

51

compounds. It was observed that PA was formed from phenolics under oxidative

52

conditions, but only in the presence of a strong oxidant. More recently, Hidalgo et

53

al.17 evaluated the impact of phenolic compounds structure characteristics and their

54

ability to degrade amino acids. In that study it was shown that polyphenols are

55

capable of reacting with amino acids in the absence of metal ions independently of the

56

pH, however the polyphenol needs to have two hydroxyl groups in ortho or para

57

positions. However, in wine that is rich in polyphenols with vicinal hydroxyl

58

structures such as gallic acid as well as transition metals such as copper (Cu) and iron

59

(Fe), most studies in wine have shown that phenolic oxidation only occurs in the

60

presence of metals.14, 18

61

The interaction between Maillard reaction (MR) and phenolics oxidation (PO) has, to

62

our knowledge, not yet been reported in wine conditions. Peterson et al, has

63

investigated the role of catechin and epicatechin in preventing the formation of

64

dicarbonyls and pyrazines,19 as well as flavor compounds,20 proving that epicatechin is

65

capable to form adducts with methyl glyoxal formed from the Maillard reaction.21

66

Recently, Wilker et al studied the impact of polyphenols in a solution of

67

glucose/alanine (pH 5 and 7) at 80, 100 and 130oC in the formation of glucosone,22

68

and observed that the pro-oxidative effects of polyphenols depend on their

69

concentration, pH and temperature of the model system. These studies focused on the

70

impact of polyphenols in model systems or food matrices with pH’s between 5 and 7

71

and never in hydro alcoholic solutions such as wine.

72

In order to better understand the impact of the interaction between Maillard reaction

73

and phenolic oxidation in the formation of phenylacetaldehyde a systematic study in


Page 4 of 32

Page 5 of 32


74

wine like model systems was conducted using a well-delivered design of experiments

75

(DoE). Between several DoE techniques, full factorial design identifies experiments

76

at every combination of factor levels. In the present work two full factorial designs

77

have been studied focusing on the effect of: 1) pH and 2) temperature, when three

78

factors: glucose, gallic acid and metals at two levels (present or absence) were varied

79

while keeping phenylalanine constant. For the most relevant parameters extracted

80

from the DoE, kinetics studies were also performed and rate constants estimated for

81

both wine model and real wine systems.

82 83

Materials and Methods

84

Reagents.

85

D-Glucose were purchased from Alfa Aesar (Massachusetts, US) Gallic acid, o-

86

phenylenediamine (OPD), formic acid, 3-octanol, and phenylacetaldehyde were

87

obtained from Sigma Aldrich (St. Louis, Mo). Ethylene diaminetetraacetic acid

88

disodium salt (EDTA), D-phenylalanine, anhydrous sodium sulfate, tartaric acid,

89

sodium hydroxide, Cu(II)sulfate.5H2O, Fe(II)sulfate.7H2O and acetic acid were

90

purchased from Merck (Darmstadt, Germany). Ethanol, methanol, acetonitrile and

91

isopropanol (all LC-MS grade) were obtained from Fischer Scientific (UK).

92 93

Experimental Design

94

Two 24 full factorial designs were performed to identify the relationship existing

95

between the response functions and variables as well as to determine those conditions

96

that optimized the behavior of each response. Determination of responses was

97

conducted as a duplicated completely randomized full factorial design. DoE 1:

98

evaluating four factors (glucose, gallic acid and metals) at two levels (present or



99

absence) at pH 3.4 and 7 and DoE 2: evaluating four factors (glucose, gallic acid and

100

metals) at two levels (present or absence) at two temperatures (40ºC and 80ºC), as

101

depicted in Figure 1.

102

A total of 16 combinations were generated for each DOE’s and performed in

103

duplicated in random order. The models were built with JMP statistical software

104

version 10.0.0 (SAS Institute Inc., Cary, NC, USA). Each response was analyzed by

105

two-way analysis of variance (ANOVA) using the general linear model (GLM)

106

procedure using JMP statistical software version 10.0.0 (SAS Institute Inc., Cary, NC,

107

USA). The single and interactive effects on phenylacetaldehyde formation were

108

determined to explore the significance of differences at p-value < 0.05. Graphics were

109

created with the JMP statistical software version 10.0.0 (SAS Institute Inc., Cary, NC,

110

USA).

111 112

Preparation of model systems and storage conditions

113

Two model systems comprising 12% (v/v) aqueous ethanol and tartaric acid (0.03M)

114

with addition of phenylalanine (2.4 mM) buffered to pH 3.4 and 7 with NaOH (1M)

115

were used for all experiments. The role of tartaric acid was not only to maintain

116

acidity but in its role in coordinate preferentially to ferric ions to reduce the formal

117

reduction potential of the Fe3+/Fe2+couple14. According with the DOE conditions,

118

glucose (2.4 mM), gallic acid (2.4 mM) were added to the solution. The metal ions,

119

copper (II) and iron (II), were added to the model system at concentrations of 6.3 µM

120

and 0.1 mM respectively in the form of Cu (II) sulfate.pentahydrate and Fe (II)

121

sulfate.heptahydrate. In the solutions with no metals addition, a 50µM EDTA solution

122

was also added. The solutions were prepared in 20mL vials. The vials were closed

123

with an internal silicone septum and an external screw cap. An accelerated


Page 6 of 32

Page 7 of 32


124

“browning” reaction scheme was employed. This involved heating the samples at

125

80oC or 40ºC depending on the DOE conditions. Dissolved oxygen was measured in

126

the beginning of all experiments using a Fibox 3 LCD Trace, with Pst3 sensors

127

(Presens GmbH, Germany). Concentration of dissolved oxygen for all the solutions

128

were 9 mg/L. All samples were incubated at 80oC or 40ºC for 24 hours, samples were

129

prepared in duplicate.

130 131

Kinetics of phenylacetaldehyde formation in model solutions and in real wine

132

For the kinetic studies two equimolar (2.4mM) model systems were considered: i)

133

containing phenylalanine and gallic acid (so called PO system) and ii) containing

134

phenylalanine, gallic acid and glucose (so called PO + MR system), both with

135

addition of metal ions. Solutions were prepared in wine model system comprising

136

12% (v/v) aqueous ethanol and tartaric acid (0.03 M). The metal ions, copper (II) and

137

iron (II), has previously described were added at concentrations of 6.3 µM and 0.1

138

mM respectively. All solutions were adjusted to the pH of wine (pH 3.4) with NaOH

139

(1M). In order to evaluate the impact of the presence of sugar in the

140

phenylacetaldehyde formation in real wine systems, glucose 10 g/L has been added to

141

a young, dry white wine, produced in the Douro region (North Portugal) (pH=3.4;

142

SO2 free = 0.2 mM), following standard winemaking procedures, without any wood

143

contact. Sample solutions were put into tightly closed glass tubes and placed in an air

144

circulating, controlled oven at 40ºC. The samples were removed from the oven at

145

different sampling times (1, 3, 5, 9, 21 and 24 hours). Dissolved oxygen was

146

measured in the beginning of all experiments using a Fibox 3 LCD Trace, with Pst3

147

sensors (Presens GmbH, Germany). Concentrations of dissolved oxygen for all the

148

solutions were 9 mg/L.



149

It was assumed that PA formation follows a pseudo-first-order reaction; that is

150

C=C0exp(-kt), where C is the concentration of the reactant (µM), C0 is the initial

151

reactant concentration, k the reaction rate constant and t time (hours). Fit of the curves

152

were optimized to the data by minimizing the sum of squares with the SOLVER add-

153

in, Excel 2016 (Microsoft USA).

154 155

Analytical procedure.

156

Volatiles quantitation – GC-MS

157

Phenylacetaldehyde was analyzed by headspace (HS) – solid phase micro extraction

158

(SPME) using a SPME grey (DVB/CAR/PDMS) fiber (Supelco, Bellefonte, PA) and

159

the volatiles were analyzed using a Varian 450 Gas Chromatograph, equipped with a

160

mass spectral detector, Varian 240-MS and a CombiPAL auto sampler (CTC

161

Analytics AG, Zwingen, Switzerland). The column used was a VF-WAXms (15m x

162

0.15mm, 0.15µm) (Varian). A 5 mL sample was spiked with 20 µL of 3-octanol (50

163

mg/L) as the internal standard. Anhydrous sodium sulfate (0.5 g) was added to

164

increase ionic strength. Samples were pre-incubated in the CombiPAL oven at 40°C

165

and 500 rpm for 2 min, then the fiber was exposed on the sample headspace for 10

166

minutes at 250 rpm and 40°C for extraction. The fiber was kept 10 mm from the

167

bottom. Desorption of volatiles took place in the injector at 220°C for 15 minutes.

168

The oven temperature was 40°C (1 min), and then increased 4 °C/min to 220 °C and

169

held for 2.5 min (48.5 min). The carrier gas was Helium at 1 mL/min, with a constant

170

flow. The injector port was heated to 220°C. The injection volume was 1 µL in

171

splitless mode and the split vent was opened after 30 seconds. All mass spectra were

172

acquired in the electron impact (EI) mode (ionization energy, 70 eV; source

173

temperature, 180°C). The analysis temperatures were: trap 170°C, the manifold 50°C,


Page 8 of 32

Page 9 of 32


174

transfer line 190°C and ion source 210°C. Compound identification was achieved by

175

comparing retention times and mass spectra obtained from a sample containing pure,

176

authentic standards. Compound quantitation was performed based on standard

177

calibration curves (n=7) using m/z = 91 as quantifier and m/z=120 as qualifier. LOD

178

and LOQ of the method for PA are: 30 and 100 µg/L respectively. The recovery is

179

99.8% and the RSD was 2.5%.

180 181

Intermediaries Identification – LC-ESI-QqTOF-HRMS

182

After reaction, 3 mL of each samples of the kinetic analysis in model solutions (t=0;

183

t=1h; t=24h) were derivatized overnight in closed vials with an OPD solution (0.05

184

M) to form stable quinoxalines.23

185

These were further analyzed by LC-ESI-UHR-QqTOF-MS. The LC-ESI-UHR-

186

QqTOF-MS analysis was performed on a UltiMate 3000 Dionex UHPLC (Thermo

187

Scientific), coupled to a Ultra-High Resolution Qq-Time-Of-Flight (UHR-QqTOF)

188

mass spectrometer with 50,000 Full-Sensitivity Resolution (FSR) (Impact II, Bruker

189

Daltonics, Bremen, Germany).

190

Separation of metabolites was performed using an Acclaim RSLC 120 C18 column

191

(100mm x 2.1 mm, 2,2um) (Dionex). Mobile phases were 0.1% aqueous formic acid

192

(solvent A) and acetonitrile with 0.1% formic acid (solvent B). The gradient started

193

with 5% during increased to 95% in 7 min, which was kept constant for 2min and;

194

returned to 5% B in 1 minute and maintained at 5% B for an additional 5 min at a

195

flow rate of 0.25 mL/min. The injection volume was 1uL. Parameters for MS analysis

196

were set using positive ionization mode with spectra acquired over a range from m/z

197

20 to 1000. The parameters were as follow: capillary voltage, 4.5kV; drying gas

198

temperature, 200ºC; drying gas flow, 8.0 L/min; nebulizing gas pressure, 2bar;



199

collision RF,300 Vpp; transfer time, 120us and prepulse storage, 4us. Post-acquisition

200

internal mass calibration used sodium formate clusters with the sodium formate

201

delivered by a syringe pump at the start of each chromatographic analysis.

202

High-resolution mass spectrometry was used to identify the adduct formed between

203

the gallic acid quinone and the OPD (4-hydroxyphenazine-2-carboxylic acid). The

204

elemental composition for the compound was confirmed according to accurate masse

205

and isotope rate calculations designated as mSigma (Bruker Daltonics). The accurate

206

mass measured was within 5mDa of the assigned elemental composition and mSigma

207

values of < 20 provided confirmation. Gallic acid quinone were identified based on

208

its accurate mass [M+H]+= 241.0619. Its concentration has been expressed in gallic

209

acid equivalents in mg/L, for that samples without derivatization for times 0h, 1h and

210

24h were analyzed in the same acquisition conditions but in negative mode.

211 212

Results and Discussion

213

The interactions between Maillard reaction (MR) and phenolic oxidation (PO), in

214

phenylacetaldehyde (PA) formation in hydro alcoholic solutions, were studied using

215

two full factorial designs where temperature and pH effects were studied separately.

216

One DoE focused on the effect of pH (3.4 and 7) and the other on the effect of

217

temperature (40°C and 80°C). In each DoE three factors: glucose, gallic acid and

218

metals (Fe (II) and Cu (II)) at two levels (present or absence) were varied while

219

keeping phenylalanine constant. All the experiments were performed in (12% ethanol)

220

solutions with tartaric acid. F-test was conducted for the analysis of variance

221

(ANOVA) to evaluate the statistical significance of each model.

222

For the two studied DoEs the coefficient of determination (R2) was used to evaluate

223

the fitness of the model. The R2 of 0.97 for pH model and 0.98 for temperature model


Page 10 of 32

Page 11 of 32


224

indicates that the predicted values obtained from each model are a good fit of the

225

experimental data. The lack of fit compared the residual error to the pure error from

226

duplicated experimental design points. In the full mode, the p-value for lack of fit is

227

0.204 and 0.154 for each model, which is greater than 0.05, indicating that the lack-

228

of-fit is insignificant relative to the pure error. Also, the obtained F-value were 135.4

229

for the temperature and 48.4 for pH model, and the obtained Prob > F-value of

Strecker Aldehyde Formation in Wine: New Insights into the Role of

Recommend Documents