Physical and Molecular Changes during the Storage of Gluten-Free

Subscriber access provided by RYERSON UNIV

Article

Physical and molecular changes during storage of gluten-free rice and oat bread Anna-Sophie Hager, Geertrui Bosmans, and Jan A. Delcour J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/jf502036x • Publication Date (Web): 26 May 2014 Downloaded from http://pubs.acs.org on May 31, 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 29

Journal of Agricultural and Food Chemistry

1

Physical and molecular changes during storage of

2

gluten-free rice and oat bread

3

Anna-Sophie Hager*, Geertrui Bosmans, Jan A. Delcour

4 5

Laboratory of Food Chemistry and Biochemistry and Leuven Food Science and Nutrition Research

6

Centre (LFoRCe), KU Leuven, Kasteelpark Arenberg 20, B-3001 Leuven, Belgium

7 8

*Corresponding author. Phone: (+32)-16-379193. Fax: (+32)-16-321997.

9

E-mail: [email protected]

1 ACS Paragon Plus Environment


10

Abstract

11

Gluten-free bread crumb generally firms more rapidly than that of regular wheat bread. We here

12

combined differential scanning calorimetry (DSC), texture analysis and time-domain proton nuclear

13

magnetic resonance (TD 1H NMR) for investigating the mechanisms underlying firming of gluten-free

14

rice and oat bread. Molecular mobility of water and biopolymers in flour-water model systems and

15

changes thereof after heating and subsequent cooling to room temperature were investigated as a

16

basis for underpinning the interpretation of TD 1H NMR profiles of fresh crumb. The proton

17

distributions of wheat and rice flour - water model systems were comparable, while that of oat flour-

18

water samples showed less resolved peaks and an additional population at higher T2 relaxation times

19

representing lipid protons. No significant crumb moisture loss during storage was observed for the

20

gluten-free bread loaves. Crumb firming was mainly caused by amylopectin retrogradation and water

21

redistribution within bread crumb. DSC, texture, and TD 1H NMR data correlated well and showed

22

that starch retrogradation and crumb firming are much more pronounced in rice-flour bread than in

23

oat-flour bread.

24

Key words

25

Gluten-free, time-domain proton nuclear magnetic resonance, proton mobility, amylopectin

26

retrogradation, crumb firming


Page 2 of 29

Page 3 of 29


27

Introduction

28

Coeliac disease is one of the most frequent genetically based food intolerances. It affects

29

approximately 1% of the Western population.1 The only available therapy is lifelong avoidance of

30

storage proteins from wheat, rye and barley, which, in the context of this disease, are referred to as

31

gluten. While oat has been considered unsuitable for coeliac patients, recent scientific evidence has

32

allowed concluding that most gluten-intolerant people can consume oat without adverse health

33

effects.2 Nowadays, the gluten-free diet is also followed by an increasing number of individuals who

34

do not have coeliac disease but appear to benefit from avoiding gluten. These include sufferers of

35

non-coeliac gluten sensitivity3 and dermatitis herpetiformis.4,5 In addition, also the increasing number

36

of individuals following a gluten-free diet as life style choice creates an increasing demand for

37

tasteful gluten-free bakery products with a texture comparable to that of their wheat counterparts.

38

Due to the lack of a viscoelastic gluten network, production of leavened baked goods from non-

39

wheat sources is challenging and the resulting bread has low volume and a firm crumb.6 Due to its

40

wide availability, bland taste and white color,7 rice is the most commonly used gluten-free raw

41

material.8 However, rice flour contains lower levels of nutrients such as protein, fiber and minerals

42

than wheat flour.9 Hence, rice-based gluten-free products are often nutritionally inferior to standard

43

wheat products. Oat bread not only has a better nutritional quality than rice bread, it also results in

44

bread of improved volume, texture and taste.10

45

Crumb firming of wheat bread is mainly related to changes in the starch fraction that result in a

46

continuous, rigid, crystalline starch network.11-15 On a molecular scale, amylose crystallizes already

47

during cooling after baking, while amylopectin does so over a longer time span in a process referred

48

to as retrogradation.12,16-18 The amylose and amylopectin crystallites create network junction zones.

49

These serve as nucleation sites for an intermolecular and maybe even intergranular mesoscale

50

network18 that includes unfreezable water. These changes can be monitored with differential

51

scanning calorimetry (DSC) as an increase in melting enthalpy of retrograded amylopectin and a

52

concomitant decrease in amount of freezable water (FW).15 The fast-forming amylose network and 3 ACS Paragon Plus Environment


53

the thermoset gluten network are largely responsible for initial crumb firmness and resilience, while

54

amylopectin retrogradation strengthens the network during bread storage.18 In wheat-bread crumb,

55

part of the water included in the continuous starch network is withdrawn from the amorphous

56

networks, which therefore lose plasticizing water. In addition, the thermodynamic incompatibility of

57

starch and gluten causes water to diffuse from gluten to starch as shown for a model system.19 The

58

changes, which occur in the biopolymers during wheat-bread baking as starch gelatinizes and gluten

59

cross-links, and during storage as amylopectin retrogrades,20 probably cause the water diffusion

60

process to proceed until (new) equilibria are reached. Not only does water redistribute within bread

61

crumb during storage, it also migrates from crumb to crust, leading to further local reduction in

62

moisture content of the biopolymer networks. As a result, the moisture content of the gluten

63

network can drop below that needed for full plasticization during storage. The resulting increase in

64

stiffness of the gluten network then also contributes to crumb firming, mostly after a couple of days

65

of storage.15

66

Molecular mobility of water and biopolymers in wheat model systems and wheat bread has already

67

been studied with time-domain (TD) proton nuclear magnetic resonance (1H NMR).15,21-25 Six

68

populations of different mobility, depending on the local environment of the water and biopolymer

69

protons, were categorized for fresh wheat-bread crumb and interpreted by comparing the obtained

70

NMR profiles to those of heated and cooled biopolymer model systems.21 The least mobile and hence

71

most rigid fraction, contains non-exchanging CH protons of amylose crystals formed during cooling

72

and of amorphous starch and gluten not in contact with water. Proton populations of higher mobility

73

include CH protons of amorphous starch and gluten in little contact with water and exchanging

74

protons of confined water, starch and gluten. The largest population consists of mobile exchanging

75

protons of water, starch and gluten in the formed gel network. Finally, lipid protons represent a

76

population with highest T2 relaxation time.15,26-28 Molecular and physical changes that occur during

77

storage of starch-containing systems, result in changes in proton distributions. Amylopectin

78

retrogradation can be observed as an increase in area of the population containing rigid CH protons 4 ACS Paragon Plus Environment

Page 4 of 29

Page 5 of 29


79

and a decrease in areas of populations containing protons of amorphous starch due to a shift of

80

protons to populations of lower mobility. In addition, the mobility of the exchanging protons is

81

reduced, which can be attributed to strengthening of the starch network, loss of water from the

82

gluten network and moisture migration from the crumb.15 Based on this, TD 1H NMR appears

83

promising for also studying changes on molecular scale during storage of gluten-free bread. So far

84

and to the best of our knowledge, only one publication has reported on the use of 1H NMR to

85

investigate gluten-free bread. It stated, that the mobility of the most mobile proton population,

86

representing movable water, decreases during storage. However, no population assignment was

87

carried out.29

88

Against this background, the present study investigates the mechanisms underlying firming of the

89

crumb of gluten-free rice and oat bread. Storage-related changes are studied on a macroscopic (loaf

90

volume, crumb texture) and on a molecular (proton mobility, amylopectin retrogradation, FW) level.

91

Interpretation of complex NMR proton distributions in bread crumb is based on observations from

92

simple oat and rice flour-water model systems.

93

Materials and Methods

94

Materials and analysis of their composition

95

Rice flour [11.8% moisture, 7.1% wet basis (wb) protein, 70.1% wb total starch, 0.9% wb lipids, 0.46%

96

wb ash] and commercial wheat flour (Crousti) (13.4% moisture, 11.4% wb protein, 65.5% wb total

97

starch, 2.4% wb lipids, 0.61% wb ash) were obtained from BENEO-Remy (Wijgmaal, Belgium) and

98

Dossche Mills (Deinze, Belgium), respectively. Oat flour was purchased from E. Flahavan & Son

99

(Kilmacthomas, Ireland) and sifted to obtain flour particle sizes below 400 µm. The sifted flour

100

consisted of 10.7% moisture, 11.1% wb protein, 60.2% wb total starch, 6.6% wb lipids and 1.39 % wb

101

ash. Dry yeast was obtained from Puratos (Groot-Bijgaarden, Belgium). Sugar and salt were

102

purchased at a local supermarket. All reagents, solvents and chemicals were purchased from Sigma–

103

Aldrich (Bornem, Belgium) and were of analytical grade unless otherwise specified. 5 ACS Paragon Plus Environment


104

Flour and crumb moisture contents (MC) were determined according to AACC Approved Method 44-

105

15.02.30 Flour protein content was determined using an adaptation of the AOAC Official Method31

106

with an automated Dumas protein analysis system (EAS VarioMax C/N, Elt, Gouda, The Netherlands).

107

For conversion of nitrogen content to protein content, the following factors were used: 5.7 for

108

wheat, 5.8 for oat and 6.0 for rice flour.32 Starch content was calculated as 0.9 times the

109

monosaccharide glucose content, determined by gas-liquid chromatography as by Van Craeyveld et

110

al.33 Total lipid content was measured gravimetrically following extraction with water-saturated

111

butanol (WSB). Ash content was determined according to AACC Approved Method 08-01.01.34 All

112

analyses were performed in triplicate.

113

Defatting of oat flour and lipid extraction

114

Oat flour was defatted with WSB. In practice, 1.0 g of oat flour was shaken in 10 mL WSB for 60 min

115

and the solvent was removed from the residue with a rotational vacuum concentrator (RVC, Q-Lab,

116

Vilvoorde, Belgium) (60°C, 1 mbar). This cycle was repeated twice before the resulting defatted flour

117

was air dried overnight, used to prepare flour-water model systems and analyzed with TD 1H NMR

118

(see below).

119

Total lipid extract was obtained by WSB extraction of oat flour with an ASE200 (Dionex, Amsterdam,

120

The Netherlands) as described previously by Gerits et al.35 The obtained total lipid extract was then

121

further purified to remove protein and other non-lipid material by discarding the upper phase

122

obtained following addition of chloroform, methanol, and water [ratio of 1:1:0.9 (v/v/v)] to the total

123

lipid extract and vigorous shaking as described by Bligh and Dyer.36 Upon evaporation of the solvent

124

with a RVC (40°C, 1 mbar), the total lipid extract was transferred into NMR tubes and subjected to

125

analysis (see below).

126

Flour-water model systems

127

Flour-water model systems with a MC of 47% (representing a typical wheat dough MC) were

128

prepared by gently stirring flour and water to obtain homogeneous mixtures. After a rest of 30 to 6 ACS Paragon Plus Environment

Page 6 of 29

Page 7 of 29


129

120 min, all model systems were analyzed with DSC and TD 1H NMR before and after heating and

130

subsequent cooling to room temperature. Heating of the flour-water model systems was carried out

131

in sealed NMR tubes placed in an oil bath (10 min at 110 °C).

132

Bread making

133

Gluten-free formulations usually consist of complex mixtures of ingredients. To reduce complexity

134

and to allow comparison with wheat-bread recipes, simple formulations based solely on flour, water,

135

sugar, salt and yeast described earlier by Hager et al.10 were used here. Batters were prepared by

136

using 1.5% salt, 6.0% sugar, 3.0% yeast and a water addition level of 95.0% for oat and rice

137

(percentages based on flour weight). Optimal water addition levels were determined by preliminary

138

baking tests. To control mold growth during storage 0.1% calcium propionate and 1.6% potassium

139

sorbate were added to oat and rice batter, respectively (percentages based on flour weight). Yeast

140

was suspended in the water (30 °C) and rested for 10 min in a fermentation chamber (National

141

Manufacturing, Lincoln, NE, USA) at 30 °C and a relative humidity of 90%. The suspension was then

142

added to the premixed dry ingredients. Mixing was carried out for 60 s at speed 1 with a KitchenAid

143

KPM5 (KitchenAid, St. Joseph, MI, USA) equipped with a batter blade. The bowl content was scraped

144

down, followed by further mixing for 90 s at speed 4. Greased baking tins [internal dimension (width

145

× length × height), 8.0 cm × 14.5 cm × 5.5 cm] were filled with batter (350 g) and placed in the

146

fermentation chamber for 30 min. Bread was then baked in a Condilux deck oven (Hein, Strasse,

147

Luxemburg) for 45 min at 195 °C top and bottom heat. Prior to analyses, bread loaves were cooled

148

for 2 h.

149

Bread storage and crumb sampling

150

Cooled fresh bread loaves were wrapped in plastic foil and stored in sealed plastic bags for up to

151

120 h at 23 °C. Three bread loaves each were analyzed by crumb texture analysis and DSC at 2, 48

152

and 120 h after removal from the oven. Additionally, center-crumb samples were withdrawn at these

153

time points and transferred into NMR tubes, which were then sealed and analyzed (see below).



154

Loaf volume and crumb texture analysis

155

After cooling for 2 h, loaf volume was determined using a VolScan Profiler (Stable Micro Systems,

156

Surrey, UK). Firmness and springiness of crumbs were measured at 2, 48 and 120 h after removal

157

from the oven using an Instron 3342 (Norwood, MA, USA) equipped with a cylindrical probe

158

(diameter 25 mm). From the center of each of three fresh and stored bread loaves, three slices were

159

cut (25 mm thickness), from which cylindrical samples were obtained with a cookie cutter (diameter

160

35 mm). In a first test (test speed of 100 mm/min), one indentation was carried out. The force

161

required to cause an indentation of 30% in the center of the samples is referred to as firmness. In the

162

second test (same test speed), two indentations of 30% in the center of the samples were carried out

163

with a resting period of 60 s in between. The maximum force in the second indentation cycle divided

164

by the maximum force in the first indentation cycle is the springiness, which is a measure for the

165

reversibility of the imposed bread-crumb deformation.

166

Differential scanning calorimetry

167

DSC was carried out using a Q1000 DSC (TA Instruments, New Castle, DE, USA) previously calibrated

168

with indium. To determine the temperatures and enthalpies corresponding to melting of natively

169

present amylopectin crystals or such crystals formed as a result of retrogradation, 2.5-4.0 mg of

170

sample were accurately weighed in triplicate in aluminum pans (Perkin-Elmer, Waltham, MA, USA).

171

Deionized water was added in a ratio of 1:3 w/w (sample dry matter : water). Flour-water model

172

systems were analyzed as such, while bread crumb was freeze-dried and ground prior to analysis.

173

Pans were hermetically sealed and equilibrated at 0 °C together with an empty reference pan before

174

heating to 120 °C at a heating rate of 4 °C/min. Samples were measured after 2, 48 and 120 h of

175

storage. The resulting endotherms were evaluated using the TA Instruments Universal Analysis

176

software. Enthalpies were expressed in J/g sample dry base (db).

177

For determination of the portion of FW, i.e. the water that transforms into ice during cooling to -40

178

°C, fresh bread crumb (10-15 mg) was accurately weighed in quadruplicate in aluminum pans. The


Page 8 of 29

Page 9 of 29


179

samples were equilibrated at 15 °C and cooled to -40 °C at a rate of 4 °C/min. The portion of FW

180

(Equation 1) was calculated from the melting enthalpy of the sample (ΔHmelting, in J/g sample

181

measured between -6 and 0 °C), the MC of the sample (g water/g sample), and the melting enthalpy

182

of ice (ΔHice = 334 J/g ice).

183

% = × × 100

(Equation 1)

184

The FW content of fresh bread crumb was analyzed using samples withdrawn within 12 h after

185

baking. Pans with fresh crumb withdrawn after baking and stored at 23 °C were also analyzed after

186

48 and 120 h of storage.

187

Time-domain proton nuclear magnetic resonance

188

Measurements of proton distributions in model systems and bread were performed as described by

189

Bosmans et al.21 with a Minispec mq 20 TD NMR spectrometer (Bruker, Rheinstetten, Germany) at an

190

operating resonance frequency of 20 MHz for 1H (magnetic field strength of 0.47 T). An external

191

water bath maintained the temperature of the probe head at 25 ± 1 °C. Relaxation curves were

192

acquired using a single 90° pulse (free induction decay, FID) and a Carr-Purcell-Meiboom-Gill (CPMG)

193

pulse sequence with 90° and 180° pulse lengths of 2.86 μs and 5.42 μs, respectively. For the FID, an

194

acquisition period of 0.5 ms was used and 500 data points were collected. For the CPMG sequence,

195

the 90° and 180° pulses were separated by 0.1 ms and 2500 data points were acquired. For both

196

measurements, a recycle delay of 3.0 s was used and 32 scans were accumulated to increase the

197

signal-to-noise ratio. Three samples (approximately 0.3 g, accurately weighed) of a model system or

198

bread crumb were placed in three different NMR tubes (10 mm external diameter). All samples were

199

tightly compressed to a height of about 8 mm to remove air bubbles. Each tube was analyzed in

200

triplicate and the exact sample weight in each tube was taken into account during the computations.

201

The CONTIN algorithm of Provencher37 (software provided by Bruker) was applied to transform the

202

transverse relaxation curves with an inverse Laplace transformation to a continuous distribution of



203

transverse or spin-spin (T2) relaxation times, which are related to proton mobility. The area under the

204

curve of a population with a certain T2 relaxation time is proportional to the number of protons in

205

that population. The calculations were performed on 500 data points logarithmically spread between

206

T2 relaxation times of 0.007 and 0.5 ms and of 0.2 and 500 ms for the FID and CPMG measurements,

207

respectively. Because the inhomogeneity of the static magnetic field affected the output for the most

208

mobile FID population found around 0.5 ms,38,39 this fraction was not taken into account in the

209

following discussion.

210

Statistical analysis

211

Statistical analysis was performed using R version 3.0.2 "Frisbee Sailing" (The R Foundation for

212

Statistical Computing, Vienna, Austria) and the package “Agricolae”. Analysis of variance was carried

213

out, followed by Shapiro Wilk normality test and Tukey’s honest significant difference test.


Page 10 of 29

Page 11 of 29


214

Results and Discussion

215

Flour-water model systems

216

Molecular mobility of water and biopolymers in wheat, oat and rice flour-water model systems and

217

changes thereof after heating and subsequent cooling to room temperature were investigated first.

218

The obtained proton distribution patterns of unheated and heated gluten-free flour-water models

219

were comparable to those of wheat flour-water models21 and population assignment, therefore, was

220

performed in a similar way. Figure 1 shows the FID and CPMG proton distributions of unheated and

221

heated wheat, oat, and rice flour-water model systems. The unheated samples showed one

222

predominant FID population (population A) (Figure 1, left pane). Based on previous studies, this

223

population contains rigid CH protons of crystalline and amorphous starch21,40 and proteins not in

224

contact with water.21,28

225

Regarding the CPMG proton distributions, population E was most abundant in all samples (Figure 1,

226

right pane). This population contains mobile OH protons of water in the extragranular space, which

227

are in exchange with starch OH protons on the granule surface and with exchangeable protons of

228

proteins21. Population C of the wheat flour-water model system contains CH protons of amorphous

229

starch and gluten in contact with water21. Population D then contains OH protons of intragranular

230

water and starch, but also some CH protons of gluten and exchanging protons of confined water and

231

gluten.21,41 It can reasonably be assumed, that also CH protons of rice or oat protein are part of

232

populations C and D.

233

For unheated rice and wheat samples three distinct CPMG populations (C, D, E) were observed, even

234

if these were not fully resolved for wheat. The difference with the oat samples is that in the latter

235

population E was much broader and peaks C and D were not resolved. The total lipid extract from oat

236

flour showed CPMG populations with T2 relaxation times corresponding to populations C, E and F of

237

oat flour (results not shown). The lipid T2 relaxation times apparently are too similar to those of the

238

mobile and exchangeable protons of starch and protein, to permit clean separation of peaks. In the 11 ACS Paragon Plus Environment


239

case of unheated oat flour, an additional proton population F was observed with a T2 relaxation time

240

of approximately 100 ms (Figure 1b, right pane). Prior removal of lipids from oat flour by WSB

241

extraction, reduced the area of population F (Figure 2, right pane). Thus, this peak probably

242

represents protons of the lipid fraction. That such peak was present in the oat flour-water model

243

system, is in line with the higher endogenous lipid content of oat than of either wheat or rice.

244

For heated flour-water model systems, DSC did not detect endothermic peaks (results not shown),

245

confirming that heating model systems for 10 min at 110 °C fully gelatinized the starch in each flour

246

model system.

247

Heat treatment and subsequent cooling of the flour-water model system resulted in similar changes

248

in proton distributions. The area of population A decreased, while those of populations B and C

249

increased (Figure 1). These changes were related to starch gelatinization, resulting in a decrease in

250

protons in a rigid environment due to melting of the starch crystals and swelling of the previously

251

densely packed amorphous starch not in contact with water (present in population A), and an

252

increase in protons of amorphous starch in contact with water (present in populations B and C).21 The

253

area of population E, representing mobile protons in strong interaction with water, increased upon

254

heating. The mobility of this population decreased during heat treatment. The reduced mobility was

255

also observed for a starch-water model system and attributed to starch gelatinization and gelation.21

256

After heating and subsequent cooling, this population consisted of mobile exchanging protons of

257

water, starch and proteins present in the formed gel network. The reduced mobility was possibly

258

caused by more intense contact between the biopolymers and water.21 The CPMG populations of

259

heated wheat and rice flour-water model systems were well resolved, while this was not the case for

260

the heated oat flour-water model system. The broadness of the oat peaks can possibly be attributed

261

to the presence of endogenous lipids, as clearly resolved peaks were observed after heating and

262

subsequent cooling of the oat flour model system prepared with defatted flour (Figure 2). Heated

263

model systems prepared with oat or rice flour, contained an additional small population G with a T2


Page 12 of 29

Page 13 of 29


264

relaxation time of 300 ms. The population was not influenced by defatting, hence does not contain

265

protons of the lipid fraction, but possibly results from expulsion of water from the flour proteins

266

during heating. This small proton population with very high mobility was also observed for heated

267

and cooled egg28,42 and gluten.21

268

Oat and rice bread

269

Loaf specific volume was significantly higher (p

Physical and Molecular Changes during the Storage of Gluten-Free

Recommend Documents