Wheat gluten amino acid composition analysis by high-performance anion-exchange chromatography with integrated pulsed amperometric detection

1

This is the final draft post-refereeing. 1

The publisher’s version can be found at http://dx.doi.org/ 10.1016/j.chroma.2009.05.066 2

Please cite this article as: Rombouts, I.; Lamberts, L.; Celus, I.; Lagrain, B.; Brijs, K.; Delcour, J. 3

A. Wheat gluten amino acid composition analysis by high-performance anion-exchange 4

chromatography with integrated pulsed amperometric detection. Journal of Chromatography A. 5

2009, 1216, 5557-5562. 6

7

Wheat gluten amino acid composition analysis by high-performance anion-exchange 8

chromatography with integrated pulsed amperometric detection 9

10

Ine Rombouts*, Lieve Lamberts, Inge Celus, Bert Lagrain, 11

Kristof Brijs, and Jan A. Delcour 12

13

Laboratory of Food Chemistry and Biochemistry and 14

Leuven Food Science and Nutrition Research Center (LFoRCe), Katholieke Universiteit Leuven, 15

Kasteelpark Arenberg 20, B-3001 Leuven, Belgium 16

17

*Corresponding author. Tel.: (+32)-16-321482. Fax: (+32)-16-321997. 18

E-mail: [email protected] 19

20

21

22

2

Abstract 23

A simple accurate method for determining amino acid composition of wheat gluten 24

proteins and their gliadin and glutenin fractions using high-performance anion-exchange 25

chromatography with integrated pulsed amperometric detection is described. In contrast to 26

most conventional methods, the analysis requires neither pre- or post-column derivatization, 27

nor oxidation of the sample. It consists of hydrolysis (6.0 M hydrochloric acid solution at 110 28

°C for 24 h), evaporation of hydrolyzates (110 °C), and chromatographic separation of the 29

liberated amino acids. Correction factors (f) accounted for incomplete cleavage of peptide 30

bonds involving Val (f = 1.07) and Ile (f = 1.13) after hydrolysis for 24 h and for Ser (f = 31

1.32) losses during evaporation. Gradient conditions including an extra eluent (0.1 M acetic 32

acid solution) allowed multiple sequential sample analyses without risk of Glu contamination 33

on the anion-exchange column. While gluten amino acid compositions by the present method 34

were mostly comparable to those obtained by a conventional method involving oxidation, 35

acid hydrolysis and post-column ninhydrin derivatization, the latter method underestimated 36

Tyr, Val and Ile levels. Results for the other amino acids obtained by the different methods 37

were linearly correlated (r > 0.99, slope = 1.03). 38

39

Keywords: Amino acid, Gluten, Wheat, High-performance anion-exchange chromatography 40

with integrated pulsed amperometric detection, HPAEC-IPAD 41

42

43

3

1. Introduction 44

The storage proteins of wheat represent an important fraction of the daily human protein 45

intake. These proteins form a cohesive viscoelastic dough mass when mixed with water, 46

which is referred to as gluten [1,2]. Wheat gluten can be fractionated with 70% ethanol into 47

extractable gliadins and unextractable glutenins [3]. Because of their unique properties, gluten 48

proteins play a key role in the quality of different wheat-based products, e.g. bread [2,4], 49

sugar snap cookies [5], pound cakes [6] and pasta [7-9]. Amino acid analysis is useful for 50

wheat cultivar mapping and breeding purposes [10] [11] [12], and for evaluating impact of 51

wheat processing on nutritional value of gluten proteins [13][14]. Moreover, gluten further 52

polymerizes at high temperature and/or pressure. It has a large potential for use in non-food 53

applications, such as adhesives, coatings and thermoplastic materials [15,16]. Amino acid 54

analysis of hydrothermally treated gluten samples is helpful for understanding polymerization 55

reactions. 56

Amino acid analysis starts with their release in a hydrolysis reaction. Hydrolysis induces 57

analytical errors [17,18]. Hydrolysis of food proteins in 6.0 M hydrochloric acid (HCl) 58

solution releases Asn and Gln as Asp and Glu respectively. Hence, Asx and Glx refer to the 59

sum of Asp and Asn, and Glu and Gln, respectively [17,18]. The minor amino acid Trp can 60

only be determined after hydrolysis with sulfonic acids or after alkaline hydrolysis [17]. 61

Oxidation with performic acid prior to hydrolysis avoids degradation of the free amino acids 62

Cys and Met [17]. Furthermore, a protective agent (e.g. phenol) is sometimes added to 63

prevent Tyr oxidation. It has not been tested whether those protective measures (oxidation 64

with performic acid, and addition of a protective agent) are necessary in the specific case of 65

gluten proteins. Finally, acid hydrolysis of proteins during a single time interval leads to 66

inaccurate estimation of amino acid levels due to an effect of hydrolysis time on either 67

peptide bond cleavage or amino acid degradation [19]. Correction factors compensate for 68

both errors and vary for the protein studied since they depend on amino acid sequence and 69

4

tertiary structure. Kohler and Palter [20] estimated correction factors for amino acid levels 70

upon acid hydrolysis of wheat flour in 6.0 M HCl (without phenol) at 110 °C for 24 h. 71

However, these factors are not necessarily valid when applied to gluten isolated from wheat 72

flour, or when using other hydrolysis conditions. For complete removal of HCl after 73

hydrolysis, Kohler and Palter [20] washed and evaporated hydrolyzates several times. HCl 74

can also be removed from reaction mixtures by freezing and desiccating under vacuum (ca. 75

11 h) [21]. More recent methods consume less time and work but require special test tubes or 76

apparatus to obtain vacuum (e.g. SpeedVac Concentrator, Thermo Fisher Scientific Inc., 77

Waltham, MA, USA) [13,19]. Although gluten hydrolysis is frequently performed, the 78

optimization of conventional protein hydrolysis conditions to the specific case of gluten 79

proteins has, to the best of our knowledge, not been described before. 80

After hydrolysis, the liberated amino acids are quantified by chromatography. To the best 81

of our knowledge, all current methods for chromatographic determination of gluten amino 82

acids require either pre- or post-column derivatization [17,22-24]. The majority of amino acid 83

analyzers use ninhydrin post-column derivatization [17,22], while pre-column derivatization 84

is predominantly performed with phenyl isothiocyanate [17,23]. Extensive sample 85

preparation, low stability of derivatives and interference of reagents are the main 86

shortcomings of pre-column derivatization. In post-column derivatization, the online reactor 87

needed adds complexity to the system and causes peak-broadening, additional baseline noise 88

and high maintenance cost. The major disadvantage of ninhydrin post-column derivatization 89

is the poor sensitivity of the subsequent detection method [17,24]. Although efforts have been 90

made to reduce derivatization times [25], a derivatization-free method has not yet been 91

developed for gluten amino acid analysis. In contrast, literature reports on amino acid analysis 92

for other (non-)food products without a derivatization step by the use of high-performance 93

anion-exchange chromatography (HPAEC) with integrated pulsed amperometric detection 94

(IPAD) [26-29]. This method uses single-component mobile phases at low flow rates. Studies 95

5

have focused on simultaneous quantification of free amino acids and carbohydrates in model 96

systems [30-32] and in various types of samples (cell culture and fermentation broth media 97

[26,27], green tea [28], sourdough [29], skim milk [29], lemon juice [29], and potato [29]). 98

Literature also reports on HPAEC-IPAD separation and quantification ofamino acids in 99

protein hydrolyzates [33]. In gluten proteins, however, the high Gln levels complicate the 100

determination method as the AminoPac PA10 analytical column manual describes strong Glu 101

retention, which leads to column contamination and thus inaccurate Glx quantification. For 102

that reason, adaptation of the standard gradient conditions is needed. 103

Against this background, we developed an accurate and user-friendly method for gluten 104

amino acid analysis. For the first time, correction factors were derived for the different gluten 105

amino acids to compensate for errors due to hydrolysis during a single time interval. 106

Furthermore, the impact of an oxidation step prior to hydrolysis and evaporation of gluten 107

hydrolyzates was investigated. In addition, gluten amino acid compositions were, for the first 108

time, quantified by a derivatization-free method, using HPAEC-IPAD. Adequate gradient 109

conditions were developed to avoid Glu accumulation on the anion-exchange column as a 110

result of high Gln levels in gluten proteins. We here report on the outcome of this work. 111

112

113

6

2. Material and Methods 114

2.1. Gluten and its fractions 115

Three commercial vital wheat gluten samples were obtained from Syral (Aalst, Belgium). 116

Their protein content (N x 5.7, see below) ranged from 74.2 to 77.8% on dry matter basis 117

(db). To obtain enriched gliadin and glutenin fractions, gluten samples (20.0 g) were 118

extracted several times with 70% (v/v) ethanol (250 ml), followed by centrifugation (10 min, 119

10000g). The gliadin containing supernatant after the first extraction was evaporated to 120

remove ethanol and is further referred to as gliadin fraction. The glutenin containing residue 121

after the third extraction was washed with deionized water and is further referred to as 122

glutenin fraction. Both the gliadin and glutenin fractions were freeze-dried and then ground in 123

a laboratory mill (250 µm, IKA, Staufen, Germany). Protein contents (N x 5.7, db) of gliadin 124

and glutenin fractions were 82.3 and 83.1%, respectively. 125

2.2. Chemicals and reagents 126

Four eluents were used for the gradient mobile phases: water of 18.2 M conductivity 127

(A), and solutions of sodium hydroxide (B; 0.250 M, Baker, Deventer, The Netherlands), 128

sodium acetate (C; 1.0 M, Dionex Benelux, Amsterdam, The Netherlands), and acetic acid 129

(D; 0.100 M, VWR International Europe, Leuven, Belgium). All eluents were degassed and 130

kept under slight helium overpressure to prevent accumulation of atmospheric carbon 131

dioxide. HCl, norleucine, and the amino acid standards were purchased from Sigma-Aldrich 132

(Bornem, Belgium). Phenol was from VWR International Europe. All chemicals and reagents 133

were at least of analytical-reagent grade 134

2.3. Protein content determination 135

Protein contents were determined in triplicate, using an adaptation of the AOAC Official 136

Method 990.03 [34] to an automated Dumas protein analysis system (EAS Variomax N/CN, 137

Elt, Gouda, The Netherlands). Nitrogen contents were converted to protein contents by using 138

5.7 as conversion factor. 139

7

2.4. Acid hydrolysis of gluten (fractions) 140

Hydrolysis was performed by heating proteins (in triplicate) in 6.0 M HCl containing 141

0.1% phenol at 110 °C for different times (3 to 96 h). To that end, 500 µl of 12.0 M HCl 142

solution containing 0.2% phenol (as protective agent) and 500 µl of 6.0 mM norleucine 143

solution (as internal standard) were added to sample containing 7.0 mg dry matter protein. 144

Samples were incubated under nitrogen prior to hydrolysis by flushing the headspace during 145

10 s with nitrogen gas to prevent amino acid oxidation. To investigate impact of evaporation 146

of gluten hydrolyzates on amino acid levels (see below), reaction mixtures were either diluted 147

(200-fold) or subsequently evaporated (110 °C, 180 min), resuspended, and diluted (200-148

fold). After filtration (Millex-GP, 0.22 µm, polyethersulfone, Millipore, Carrigtwohill, 149

Ireland), amino acid levels were determined by HPAEC-IPAD. 150

2.5. Chromatography 151

Analyses were performed with a Dionex BioLC system (Sunnyvale, CA, USA), as 152

described by Lamberts et al. [35]. The system was equipped with a GS50 gradient pump with 153

online degasser, an AS50 autosampler with a thermal compartment, and an ED50 154

electrochemical detector containing both a gold working electrode and a pH reference 155

electrode. Separation of 25 µl samples was performed at 30 °C with an AminoPac PA10 156

guard (Dionex Benelux, 50 x 2 mm) and an analytical (Dionex Benelux, 250 x 2 mm) column 157

at a flow rate of 0.25 ml/min. Cys is thereby oxidized to cystine. Chromeleon Version 6.70 158

software (Dionex Benelux) was used for chromatographic system control, data acquisition 159

and data analysis. Amino acid levels in gluten proteins were expressed on dry matter protein 160

(µmol/g). The gradient conditions (Table 1) were adapted from Lamberts et al. [35]. The 161

detection waveform (Table 2) was from Ding et al. [28]. 162

2.6. Amino acid analysis by AEC with ninhydrin post-column derivatization 163

Amino acid composition of three gluten samples was determined by AEC with ninhydrin 164

post-column derivatization [36]. The analysis was performed by Agrobio Laboratoires (Vezin 165

8

Le Coquet, France) and consisted of 24 h hydrolysis in 6.0 M HCl (without phenol) at 110 166

°C, followed by chromatographic separation and post-column derivatization of the amino 167

acids with ninhydrin. Degradation of Met and Cys was prevented by oxidation with performic 168

acid. This method is further referred to as the reference method. 169

2.7. Statistical analysis 170

Method performance indicators were determined by making at least five repetitive 171

injections of amino acid standards. Limit of detection (LOD) and quantification (LOQ) were 172

estimated from the minimum concentration (25 µl injection) required to produce peak area 173

signal-to-noise ratios of 3 and 10, respectively. Response factors were defined as the peak 174

area-to-concentration ratio. To evaluate reproducibility, an amino acid standard was injected 175

five times and relative standard deviations (RSDs) were calculated. Spike-and-recovery tests 176

were performed by spiking sample hydrolysates with amino acid standards at 10 to 70% of 177

the determined level. Gluten amino acid levels were compared in absence and presence of 178

amino acid standards to calculate recoveries of individual amino acids. Each spiking 179

experiment was repeated for three different gluten samples. Regression analysis was 180

performed using the Statistical Analysis System Software 8.1 (SAS Institute, Cary, NC, 181

USA). Pearson correlation coefficients (r) and 95% confidence intervals (CI) for slope values 182

were calculated. Significant differences (P < 0.05) of gluten amino acid levels were 183

determined by t test (99% CI, two tailed) using Graphpad Prism 3.0 (San Diego, CA, USA). 184

185

186

9

3. Results and Discussion 187

3.1. Optimization of acid hydrolysis of gluten: effect of hydrolysis time 188

The success of amino acid analysis largely depends on the proper course of hydrolysis. 189

Conventional hydrolysis of gluten proteins consists of heating in 6.0 M HCl for 24 h at 110 190

°C [18,24]. Some authors assume that after 24 h gluten proteins are completely hydrolyzed 191

and that no amino acid degradation occurs. Others compensate for both errors by using 192

correction factors from Kohler and Palter [20] for wheat flour. We here determined correction 193

factors for 24 h acid hydrolysis of wheat gluten. To investigate the effect of hydrolysis time 194

on gluten amino acid levels, gluten samples were hydrolyzed for different times (3, 6, 12, 20, 195

24, 48, 72 and 96 h). Amino acids were then separated by HPAEC after dilution and 196

filtration. Arg coeluted with HCl, so it could not be quantified. 197

After 24 h hydrolysis, most amino acid levels reached their maximum value (Figure 1.A), 198

indicating complete peptide bond cleavage. However, maximum Val and Ile levels were only 199

found after hydrolysis for 72 h (Figure 1.B). To estimate Val and Ile levels adequately, 200

correction factors (f = 1.07, respectively 1.13) were derived, based on Val and Ile levels 201

determined after 24 h and 72 h of hydrolysis. Our observations that peptide bonds involving 202

Val and Ile are difficult to hydrolyze are in line with Darragh et al. [19] and Albin et al. [37]. 203

Steric hindrance of aliphatic side chains explains the particular resistance of Ile-Ile, Val-Val 204

and Ile-Val bonds to hydrolysis [17]. Hydrolysis as a function of time for wheat flours [20] 205

also resulted in correction factors for Val and Ile, but the Ile correction factor for wheat flour 206

(f = 1.08) is lower than that for wheat gluten isolated therefrom. For some proteins, peptide 207

bonds involving Leu are also difficult to hydrolyze [19]. In the present case, maximum Leu 208

levels in gluten samples were found after hydrolysis for 24 h (Figure 1.A), indicating that Leu 209

is relatively accessible in gluten. Indeed, the macromolecular protein structure also affects 210

susceptibility to hydrolysis, so correction factors vary with the protein studied [24]. 211

10

Conclusions for Leu are in line with Kohler and Palter [20] who found maximum Leu levels 212

in wheat flour samples after acid hydrolysis for 24 h. 213

Cys, Met, Ser, Thr, and Tyr were noticeably susceptible to degradation when hydrolysis 214

was carried out for more than 24 h (Figure 1.C). It is well known that sulfur-containing (Cys, 215

Met) and acid labile amino acids (Ser, Thr, Tyr) are partially destroyed at longer hydrolysis 216

times [19]. Kohler and Palter [20] did not determine amino acid levels at hydrolysis times 217

shorter than 24 h. Based on the decrease of amino acid levels with time, they derived 218

correction factors for Cys (f = 1.09), Met (f = 1.03), Ser (f = 1.08) and Thr (f = 1.04), 219

assuming that degradation starts at time zero. However, our results for gluten indicate that no 220

correction is necessary for these amino acids since their maximum levels are found after 221

hydrolysis for 24 h and degradation only becomes significant at longer hydrolysis times. 222

In conclusion, all amino acids of gluten proteins but Val and Ile were completely liberated 223

after acid hydrolysis for 24 h under the specified conditions. Furthermore, no degradation 224

occurred. Thus, the use of correction factors to compensate for hydrolysis errors of gluten 225

proteins is only necessary for Val and Ile (f = 1.07 and 1.13, respectively). 226

3.2. Impact of evaporation of gluten hydrolyzates at 110 °C 227

Since HCl interferes with arg detection and may damage the gold working electrode of 228

the Dionex BioLC system, the presence of HCl in samples had to be avoided. We therefore 229

studied the impact of evaporation of gluten hydrolyzates at 110 °C on levels of the different 230

amino acids. Chromatograms of evaporated gluten hydrolyzates illustrated an additional 231

advantage of evaporation, i.e. elimination of phenol. Amino acid levels in the (non-) 232

evaporated samples were compared by a t test (99% CI, two tailed). As indicated above, Arg 233

levels are overestimated in non-evaporated samples (P < 0.05) due to interference by HCl. 234

Significant differences were also found for Ser (P < 0.05). Ser levels were 24 ± 1.2 % lower 235

in evaporated samples than in non-evaporated samples due to losses during acid hydrolysis. 236

Hence, Ser levels were determined in non-evaporated hydrolyzates, or, alternatively, could be 237

11

estimated in evaporated hydrolyzates by multiplying Ser levels with a correction factor (f = 238

1.32). Since correction is only needed for Ser, simple evaporation of hydrolyzates at 110 °C is 239

a good alternative to evaporation using a SpeedVac System. 240

3.3. Optimization of gradient conditions 241

Adjustment of the gradient conditions described by Lamberts et al. [35] avoided retention 242

of Glu on the anion-exchange column (Table 1). Thus, an acetic acid solution was used to 243

remove the small carryover peaks and to achieve accurate separation of Arg, Lys, Ala, Thr, 244

Gly, Val, Pro, Ile, Leu, Met, norleucine, His, Phe, Glx, Asx, cystine and Tyr (in that order) 245

(Figure 2). 246

Table 3 shows the method performance indicators LOD, LOQ, upper limit of linearity, 247

correlation coefficient r, response factor, RSD and recoveries using the adapted gradient 248

conditions. LOD and LOQ for the different amino acids ranged from 2 to 14 pmol and from 8 249

to 65 pmol, respectively (25 µl sample injection). Hence, all amino acids were detectable and 250

quantifiable in concentrations exceeding 0.6 µM and 2.6 µM, respectively. LODs for other 251

methods, e.g. ninhydrin post-cleanup derivatization (approximately 100 pmol [24]) were 252

generally higher, which illustrates the sensitivity of HPAEC-IPAD. For all amino acids, 253

concentration and peak area were linearly correlated up to 100 µM or more (r > 0.99), except 254

for his and arg which were linearly related up to 25 µM. Injection of an equimolar solution of 255

amino acids leads to different peak intensities (Figure 2). Response factors were calculated to 256

investigate the sensitivity of the column for different amino acids. High response factors were 257

obtained for cystine, His, Arg and Tyr (6.12, 3.77, 2.71 and 2.60 respectively). Response 258

factors for Leu, Glu, Ile, Val, and Asp were lower (0.57, 0.71, 0.71, 0.82, 0.93 and 1.26 259

respectively). Early elution of amino acids (Arg) and peak sharpening by the acetate gradient 260

(His, cystine, Tyr) led to peaks with high area-to-concentration ratio and thus high response 261

factors, as earlier observed by Clarke et al. [38]. These authors also found lower response 262

factors for the acidic amino acids Glu and Asp. RSDs (n = 5) of the different amino acids did 263

12

not exceed 4.0%. Spike-and-recovery was used to investigate the impact of gluten sample 264

matrix on analyte detection. Recoveries for the different amino acids ranged from 89.5 to 265

102.7%. Thus, the sample preparation described here is useful for quantitative amino acid 266

analysis of gluten samples. 267

3.4. Method testing 268

Amino acid compositions of gluten and its fractions were determined using the developed 269

post-separation cleanup gradient conditions (Table 4). Prior to HPAEC-IPAD separation, 270

samples were hydrolyzed for 24 h and evaporated as described above. Correction factors 271

compensated for amino acid losses during acid hydrolysis (Val, Ile) and evaporation (Ser). 272

Cystine levels were expressed as Cys levels. Glx and Pro (2 450 and 1 080 µmol/g, 273

respectively) are the major amino acids of wheat gluten. Levels of other amino acids range 274

from 100 µmol/g (Met) to 550 µmol/g (Leu). The gliadin fraction is enriched in Pro (1 080 to 275

1 270 µmol/g) and Glx (2 450 to 2 510 µmol/g), while the glutenin fraction is enriched in Gly 276

(420 to 560 µmol/g), Lys (110 to 155 µmol/g) and Ser (440 to 620 µmol/g). These results are 277

in line with average amino acid composition of gluten and its fractions described earlier [3]. 278

To evaluate the need for oxidation with performic acid prior to hydrolysis, results for 279

amino acid levels in three different gluten samples obtained by the present oxidation-free 280

method were compared with results obtained by the reference method (with oxidation step, 281

cfr. supra) with a focus on Met and Cys levels (Table 5). Results were not significantly 282

different (P > 0.05), suggesting that the time-consuming oxidation step is not required in the 283

specific case of gluten proteins. That Cys in gluten is almost exclusively present as the more 284

stable cystine linkage, explains redundancy of Cys oxidation. Tyr levels were significantly 285

underestimated (86%, P < 0.05) by the reference method, probably because of the absence of 286

phenol during acid hydrolysis. Val and Ile were significantly underestimated (82 and 87% 287

respectively, P < 0.05) by the reference method because no correction factors were used. 288

Amino acid levels excluding those of Tyr, Val and Ile, obtained by the different methods 289

13

were linearly correlated (r > 0.99, slope = 1.03). Regression analysis without results for the 290

abundant amino acid glx confirmed the linear correlation (r > 0.99, slope = 0.99). Slope 291

values never significantly differed from 1.0 (95% CI). Hence, hydrolysis of gluten proteins 292

for 24 h as previously described, followed by quantification of amino acids by HPAEC-293

IPAD, is a reliable method for determination of amino acid composition. Correct estimation 294

of Val and Ile levels requires the use of correction factors and phenol prevents Tyr oxidation 295

during acid hydrolysis. 296

297

298

14

4. Conclusions 299

A derivatization- and oxidation-free method for determination of gluten amino acid 300

composition of multiple samples was developed. It consists of hydrolysis in 6.0 M HCl for 24 301

h, followed by evaporation, dilution, filtration and quantification with HPAEC-IPAD. The 302

chosen detection method, namely IPAD, makes a derivatization step unnecessary and is for 303

the first time applied to gluten proteins. The use of well specified gradient conditions (post 304

separation-cleanup) eliminated Glu contamination on the column and thus allowed multiple 305

sample analyses. Protein hydrolysis conditions were optimized. Amino acid levels were 306

determined after hydrolysis for different times and correction factors were calculated to 307

compensate for incomplete cleavage of peptide bonds involving Val (1.07) and Ile (1.13) 308

after 24 h hydrolysis. Evaporation of gluten hydrolyzates prior to chromatographic separation 309

eliminated possibly interfering substances (HCl). During evaporation of gluten hydrolyzates 310

at 110 °C only Ser losses occurred. Our results for gluten amino acid composition were 311

consistent with results obtained by a reference method (with oxidation step), except for Tyr, 312

Val and Ile. In contrast to the general belief, oxidation prior to hydrolysis does not 313

significantly affect Cys or Met levels in gluten samples, which indicates redundancy of the 314

oxidation step and leads to method simplification. Val and Ile levels were underestimated by 315

the reference method because no correction factors were used to compensate for incomplete 316

peptide bond cleavage after 24 h hydrolysis and Tyr was partially degraded during hydrolysis 317

of the reference method since no protective agent was used. Thus, correct estimation of Val, 318

Ile and Tyr levels in gluten proteins required the use of correction factors and a protective 319

agent (e.g. phenol). We conclude that we developed a user-friendly and accurate method for 320

determination of amino acid levels in gluten and its fractions. Tests for more complex gluten-321

containing systems and even non-gluten proteins indicated that the method has a wide range 322

of applications. 323

324

15

Acknowledgements 325

This work is a part of the Methusalem programme ''Food for the Future" at the 326

K.U.Leuven.327

16

References 328

[1] H. Wieser, Food. Microbiol. 24 (2007) 115-119. 329

[2] W.S. Veraverbeke, J.A. Delcour, Crit. Rev. Food. Sci. Nutr. 42 (2002) 179-208. 330

[3] H.D. Belitz, W. Grosch, P. Schieberle, in: H.D. Belitz, W. Grosch, P. Schieberle (Eds.), 331

Food Chem., Springer-Verlag, Berlin, Heidelberg, 2004. 332

[4] B. Lagrain, B.G. Thewissen, K. Brijs, J.A. Delcour, J. Agric. Food. Chem. 55 (2007) 333

5320-5325. 334

[5] B. Pareyt, E. Wilderjans, H. Goesaert, K. Brijs, J.A. Delcour, J. Cereal Sci. 48 (2008) 335

863-869. 336

[6] E. Wilderjans, B. Pareyt, H. Goesaert, K. Brijs, J.A. Delcour, Food Chem. 110 (2008) 337

909-915. 338

[7] M.J. Sissons, N.E. Egan, M.C. Gianibelli, Cereal Chem. 82 (2005) 601-608. 339

[8] M.D. Oak, M. Sissons, N. Egan, S.A. Tamhankar, V.S. Rao, S.B. Bhosale, Int. J. Food 340

Sci. Tech. 41 (2006) 538-544. 341

[9] M.J. Sissons, H.N. Soh, M.A. Tumer, J. Sci. Food Agric. 87 (2007) 1874-1885. 342

[10] J.E. Ruckman, F.P. Zscheile, C.O. Qualset, J. Agric. Food. Chem. 21 (1973) 697-700. 343

[11] K.C. Sikka, S.K. Duggal, R. Singh, D.P. Gupta, M.G. Joshi, J. Agric. Food. Chem. 26 344

(1978) 788-791. 345

[12] L.F.G. Del Moral, Y. Rharrabti, V. Martos, C. Royo, J. Agric. Food. Chem. 55 (2007) 346

8144-8151. 347

[13] X.L. Jiang, Z. Hao, J.C. Tian, Cereal Chem. 85 (2008) 504-508. 348

[14] P.R. Shewry, R. D'Ovidio, D. Lafiandra, J.A. Jenkins, E.N.C. Mills, F. Békés, in: K. 349

Khan, P.R. Shewry (Eds.), Wheat: Chemistry and Technology, AACC International, St. 350

Paul, MN, USA, 2009, pp. 223-298. 351

[15] P. Hernandez-Munoz, A. Kanavouras, P.K.W. Ng, R. Gavara, J. Agric. Food. Chem. 51 352

(2003) 7647-7654. 353

[16] M. Pommet, A. Redl, M.H. Morel, S. Domenek, S. Guilbert, Macromolecular Symposia 354

197 (2003) 207-217. 355

[17] M. Fountoulakis, H.W. Lahm, J. Chromatogr. A 826 (1998) 109-134. 356

[18] M. Weiss, M. Manneberg, J.F. Juranville, H.W. Lahm, M. Fountoulakis, J. Chromatogr. 357

A 795 (1998) 263-275. 358

[19] A.J. Darragh, D.J. Garrick, P.J. Moughan, W.H. Hendriks, Anal. Biochem. 236 (1996) 359

199-207. 360

[20] G.O. Kohler, R. Palter, Cereal Chem. 44 (1967) 512-520. 361

[21] R. Tkachuk, G.N. Irvine, Cereal Chem. 46 (1969) 206-218. 362

[22] H. Wieser, W. Seilmeier, H.D. Belitz, Z. Lebensm. Unters. Forsch. 178 (1984) 173-178. 363

[23] M. Chandrasekharaiah, K.T. Sampath, A. Thulasi, Indian J. Anim. Sci. 73 (2003) 305-364

307. 365

[24] P. Koehler, in: P.R. Shewry, G.L. Lookhart (Eds.), Wheat Gluten Protein Analysis, 366

AACC Press, Washington, DC, USA, 2003, pp. 137-157. 367

[25] B.C. Fermin, J.A. Radinsky, R.J. Kratochvil, J.E. Hall, Y.M. Lo, J. Food Sci. 68 (2003) 368

2667-2671. 369

[26] V.P. Hanko, J.S. Rohrer, Anal. Biochem. 324 (2004) 29-38. 370

[27] P. Jandik, J. Cheng, D. Jensen, S. Manz, N. Avdalovic, J. Chromatogr. B 758 (2001) 189-371

196. 372

[28] Y.S. Ding, H. Yu, S.F. Mou, J. Chromatogr. A 982 (2002) 237-244. 373

[29] C. Thiele, M.G. Ganzle, R.F. Vogel, Anal. Biochem. 310 (2002) 171-178. 374

[30] P. Jandik, J. Cheng, N. Avdalovic, J. Biochem. Biophys. Methods 60 (2004) 191-203. 375

[31] H. Yu, S.F. Mou, J. Chromatogr. A 1118 (2006) 118-124. 376

[32] H. Yu, Y.S. Ding, S.F. Mou, P. Jandik, J. Cheng, J. Chromatogr. A 966 (2002) 89-97. 377

17

[33] P. Jandik, J. Cheng, N. Avdalovic, in: B.J. Smith (Ed.), Protein Sequencing Protocols, 378

Humana Press, Clifton, NJ, USA, 2002, pp. 155-167. 379

[34] AOAC, Official Methods of Analysis, Association of Official Analytical Chemists, 380

Washington DC, 16th, 1995. 381

[35] L. Lamberts, I. Rombouts, J.A. Delcour, Food Chem. 111 (2008) 738-744. 382

[36] S. Moore, W.H. Stein, J. Biol. Chem. 176 (1948) 367-388. 383

[37] D.M. Albin, J.E. Wubben, V.M. Gabert, Anim. Feed Sci. Technol. 87 (2000) 173-186. 384

[38] A.P. Clarke, P. Jandik, R.D. Rocklin, Y. Liu, N. Avdalovic, Anal. Chem. 71 (1999) 2774-385

2781. 386

387

388

18

FIGURE CAPTIONS 389

FIGURE 1: Effect of hydrolysis time on amino acid levels detected in gluten hydrolyzates. 390

Levels of amino acids are represented as the percentage of the maximum value. 391

FIGURE 2: Separation of amino acids [— equimolar solution of amino acids (12.5 µM) and 392

norleucine (15 µM); — amino acids in gluten hydrolyzate and norleucine (25 µM)] by 393

HPAEC-IPAD, using the post-separation cleanup gradient conditions [1 = Arg; 2 = Lys; 3 = 394

Ala; 4 = Thr; 5 = Gly; 6 = Val; 7 = Ser; 8 = Pro; 9 = Ile; 10 = Leu; 11 = Met; 12 = norleucine; 395

13 = His; 14 = Phe; 15 = Glu; 16 = Asp; 17 = cystine; 18 = Tyr]. 396

397

19

FIGURES 398

FIGURE 1: 399

A 400

70

75

80

85

90

95

100

0 20 40 60 80 100

Time of hydrolysis (h)

% o

f m

ax

imu

m a

min

o a

cid

le

ve

l

Ala

Gly

Pro

Phe

Glx

Asx

Lys

Leu

His

401

B 402

40

50

60

70

80

90

100

0 20 40 60 80 100

Time of hydrolysis (h)

% o

f m

ax

imu

m a

min

o a

cid

le

ve

l

Val

Ile

403

C 404

20

40

50

60

70

80

90

100

0 20 40 60 80 100

Time of hydrolysis (h)

% o

f m

ax

imu

m a

min

o a

cid

le

ve

l

Tyr

Thr

Ser

Met

Cystine

405

406

407

FIGURE 2: 408

0

50

100

150

200

250

300

(nC

)

409

0

50

100

150

200

250

300

0 5 10 15 20 25 30 35 40(min)

(nC

)

410

411

1 2

8

7

3

4

6

5

9 10

11

12

13

14 15

16

17

18

15

1

2

8

7

3 4

6

5

9 10

11 12

13 14

16 17

18

21

TABLES 412

TABLE 1: Gradient conditions for separation of gluten amino acids by HPAEC-IPAD, 413

including a post-cleanup step to allow multiple analyses. 414

Time

(min)

Water

18.2 M conductivity (%)

Sodium hydroxide 250 mM

(%)

Sodium acetate 1.0 M (%)

Acetic acid 0.1 M (%)

Curvea

0.0 76 24

2.0 76 24

8.0 64 36 8

11.0 64 36

27.0 40 20 40 8

47.0 40 20 40

47.1 100 8

49.1 100

49.2 20 80 8

51.2 20 80

51.3 76 24 5

76.0 76 24 a The gradient curve is the line representing the change in gradient conditions. Shapes of gradient curves are defined in the 415

GS50 Gradient Pump Operator’s Manual, p.37-38 (Dionex Document No. 031612, Revision 3). Curve 5 is linear. Curve 8 is 416 concave with 20% of change at about 60% of a time segment and 70% change at 90% of the same time segment. 417

418

419

TABLE 2: Detection waveform for amino acid analysis by HPAEC-IPAD (Ding et al. [28]). 420

Time (ms)

Potential (V) versus pH

Current Integration

0 0.13

40 0.13

50 0.33

210 0.33 Begin

220 0.60

460 0.60

470 0.33

560 0.33 End

570 -1.67

580 -1.67

590 0.93

600 0.13

421

422

22

TABLE 3: Method performance indicators for amino acid analysis by HPAEC-IPAD (25 µl 423

sample injection). 424

LOD LOQ

Upper limit r Response RSD Recovery

of linearity (n > 6) factor (n=5)

(pmol) (pmol) (µM) (µM-1

) (%) (%)

Arg 8 36 25 0.989 2.71 0.9 95.4 ± 3.3

Lys 2 8 100 0.985 1.87 1.3 100.0 ± 2.6

Ala 8 38 100 0.992 1.28 1.0 97.1 ± 3.2

Thr 6 27 200 0.987 2.34 0.4 99.4 ± 1.6

Gly 9 43 150 0.986 1.61 2.1 98.4 ± 1.4

Val 7 31 150 0.991 0.93 0.6 100.3 ± 2.2

Ser 14 65 200 0.989 2.50 1.8 102.7 ± 2.8

Pro 5 24 200 0.988 1.59 0.4 99.6 ± 4.5

Ile 10 43 200 0.993 0.71 2.3 100.5 ± 2.1

Leu 6 29 200 0.990 0.57 1.6 101.2 ± 5.2

Met 9 42 150 0.985 2.15 0.6 89.5 ± 3.1

Norleucine 5 22 100 0.999 0.82 2.0 102.0 ± 5.8

His 4 16 25 0.993 3.77 0.5 102.6 ± 4.5

Phe 8 34 100 0.988 1.96 1.4 99.0 ± 1.9

Glu 8 36 200 0.988 0.71 1.8 95.3 ± 8.4

Asp 9 40 100 0.988 1.26 1.1 95.8 ± 5.9

(Cys)2 6 28 100 0.986 6.12 3.9 94.5 ± 8.9

Tyr 7 32 100 0.994 2.60 0.4 98.5 ± 1.0

425

426

TABLE 4: Amino acid composition of gluten and its fractions obtained after acid hydrolysis 427

in 6.0 M hydrochloric acid at 110 °C for 24 h, followed by evaporation, filtration, dilution 428

and separation by HPAEC-IPAD (RSDs < 5.0%). 429

Gluten Glutenin Gliadin

mol% µmol/g mol% µmol/g mol% µmol/g

Ala 3,5 270 4,1 315 3,0 225

Arg 3,2 245 3,1 240 2,9 215

Asx 2,8 215 3,2 250 2,3 170

Cys 2,2 170 1,9 150 2,4 180

Glx 31,9 2450 28,1 2180 33,8 2510

Gly 5,4 420 7,2 560 2,9 215

His 1,7 130 1,7 135 1,7 125

Ile 4,1 315 4,0 305 4,7 345

Leu 7,2 550 7,1 550 7,8 580

Lys 1,4 110 2,0 155 0,6 45

Met 1,3 100 1,4 105 1,2 90

Phe 4,4 335 4,0 310 5,1 380

Pro 14,1 1080 12,4 965 17,1 1270

Ser 5,7 440 8,0 620 4,3 320

Thr 2,8 215 3,3 255 2,3 170

Tyr 2,8 220 3,1 235 2,6 195

Val 5,4 415 5,5 425 5,3 395

23

TABLE 5: Comparison of results for amino acid composition of three different gluten 430

samples (RSDs < 5.0%), determined (i) by HPAEC-IPAD [after hydrolysis (6.0 M HCl, 0.1% 431

phenol, 110 °C, 24 h), evaporation, filtration and dilution] and (ii) by AEC with ninhydrin 432

post-column derivatization [after oxidation and hydrolysis (6.0 M HCl, 110 °C, 24 h)]. 433

(mol%) Sample A Sample B Sample C

HPAEC AEC HPAEC AEC HPAEC AEC

Ala 3,6 3,8 3,5 3,7 3,6 3,7

Arg 3,3 2,7 3,2 2,6 3,1 2,5

Asx 3,0 3,2 2,8 3,1 2,8 3,2

Cys 2,2 2,2 2,2 2,2 2,1 2,5

Glx 31,5 32,5 31,9 33,0 32,0 32,6

Gly 5,4 5,8 5,4 5,8 5,3 5,6

His 1,7 1,7 1,7 1,7 1,7 1,7

Ile 4,2 3,6 4,1 3,6 4,1 3,7

Leu 7,1 7,1 7,2 7,1 7,2 7,1

Lys 1,5 1,5 1,4 1,4 1,4 1,4

Met 1,3 1,5 1,3 1,5 1,3 1,7

Phe 4,3 4,2 4,4 4,2 4,4 4,2

Pro 13,7 13,9 14,1 14,1 14,3 13,9

Ser 6,2 6,0 5,7 5,9 5,8 5,8

Thr 2,9 2,8 2,8 2,7 2,8 2,8

Tyr 2,7 2,4 2,8 2,4 2,7 2,3

Val 5,3 4,4 5,4 4,3 5,3 4,5

434

435

436