Studies on the Formation of Maillard and Caramelization Products

Subscriber access provided by UNIV OF MISSISSIPPI

Article

Studies on the formation of Maillard and caramelization products from glucosamine incubated at 37#C Yuliya Hrynets, Maurice Ndagijimana, and Mirko Betti J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.5b02664 • Publication Date (Web): 26 Jun 2015 Downloaded from http://pubs.acs.org on June 30, 2015

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 41

1 2

Journal of Agricultural and Food Chemistry

Studies on the formation of Maillard and caramelization products from glucosamine incubated at 37˚C

3 4

Yuliya Hrynets, Maurice Ndagijimana and Mirko Betti*

5 6 7

Department of Agricultural, Food and Nutritional Science, 4-10 Ag/For Building, University of Alberta, Edmonton, Canada T6G 2P5

8 9 10 11 12 13

*Corresponding author: Tel: 780.248.1598; E-mail: [email protected]

14 15 16 17 18 19 20 21 22 23 24 25

26 1 ACS Paragon Plus Environment


Page 2 of 41

27

ABSTRACT

28

This experiment compared the in vitro degradation of glucosamine, N-acetyl-glucosamine and

29

glucose in the presence of NH3 incubated at 37°C in phosphate buffer from 0.5 to 12 days. The

30

reactions were monitored with UV-Vis absorption and fluorescence emission spectroscopies, and

31

the main products of degradation, quinoxaline derivatives of α-dicarbonyl compounds and

32

condensation products, were determined using UHPLC-UV and Orbitrap mass spectrometry.

33

GlcN produced two major dicarbonyl compounds: glucosone, and 3-deoxyglucosone, ranging

34

from 709 to 3245 and 272 to 4535 mg/kg GlcN, respectively. 3,4-dideoxyglucosone-3-ene,

35

glyoxal, hydroxypyruvaldehyde, methylglyoxal and diacetyl were also detected in lower amounts

36

compared to glucosone and 3-deoxyglucosone. Several pyrazine condensation products resulting

37

from the reaction between dicarbonyls and GlcN were also identified. This study determined that

38

GlcN is a significantly unstable molecule producing high level of degradation products at 37 ºC.

39 40

KEYWORDS: glucosamine; Heyns compound; Maillard reaction, α-dicarbonyl compounds (α-

41

DCs); o-phenylenediamine (OPD), fructosazine

42 43 44

45 46 47 48

2 ACS Paragon Plus Environment

Page 3 of 41


49

INTRODUCTION

50

N-acetyl-glucosamine (GlcNAc, 2-acetamido-2-deoxy-D-glucose) is an essential component of

51

mucopolysaccharides and chitin. Glycosaminoglycans (GAGs; mucopolysaccharides) are large

52

complexes of negatively charged carbohydrate chains that are incorporated into mucous

53

secretions, connective tissue, skin, tendons, ligaments and cartilage.1 Deacetylation of GlcNAc

54

leads to the formation of glucosamine, (GlcN, 2-amino-2-deoxy-D-glucose), an amino

55

monosaccharide used by North American consumers as a dietary supplement to both reduce

56

osteoarthritis pain and improve joint function.2 It is believed that GlcN supplementation may

57

stimulate synovial production of hyaluronic acid (HA)3, a type of GAG responsible for the

58

lubricating and shock-absorbing properties of synovial fluid in cartilage.4 For the industrial

59

production of GlcN, chitin5 or chitosan, by-products of the shellfish processing industry, can be

60

used as a starting material.6 Different products are available on the market; the most common are

61

GlcN sulfate (GlcN·SO4), GlcN-hydrochloride (GlcN·HCl) and N-acetyl-glucosamine. In nature,

62

amino monosaccharides biosynthesis is controlled by the hexosamine biosynthesis pathway

63

(HBP), where GlcNAc is produced and used in several metabolic processes such as post-

64

translational modification of proteins and GAGs synthesis.7 In mammalian species, for instance,

65

glucosamine-6-phosphate (GlcN-6-P) is produced from the reaction between fructose-6-

66

phosphate, obtained through glycolysis, and a glutamine residue due to the action of

67

glutamine:fructose-6-phosphate-aminotransferase (GFAT). With exogenous GlcN, it is rapidly

68

phosphorylated by hexokinase yielding GlcN-6-P. Following phosphorylation, GlcN-6-P is then

69

N-acetylated to N-acetyl-GlcN-6-P via acetyl-CoA-mediated reaction and consequently

70

converted to UDP-GlcNAc, a precursor in synthesis of GAGs, proteoglycans and glycoproteins

71

(i.e. O- and N-linked glycosylation).8 On the other hand, commercial production of chitosan



Page 4 of 41

72

oligosaccharide (COS) and GlcN are normally obtained by the deacetylation and hydrolysis of

73

chitin, where the natural amide form is converted into an amino saccharides where the –NH2 is

74

free. Hence, chemically deacetylated GlcN is an equivalent aldosamine to the one firstly

75

obtained by the German chemists Heyns and Koch in the early 1950’s from the reaction of D-

76

fructose with ammonia.9 The Maillard reaction produces Amadori and Heyns rearrangements

77

forming fructosamine (1-amino-1-deoxy-ketose) or glucosamine (2-amino-2-deoxy-aldose),

78

depending if the initial sugar is an aldose (i.e. glucose) or a ketose (i.e. fructose), respectively. It

79

is known that Amadori and Heyns compounds are unstable products undergoing enolization or

80

oxidation reactions.10 On the contrary, GlcNAc, the amide derivative of GlcN, is a less reactive

81

molecule due to the stabilizing effect of the acetylation on the free amino group.

82

Due to the popularity of GlcN to treat osteoarthritis, research on this compound has dramatically

83

increased to demonstrate other activities. For instance, in addition to its chondro-protective

84

action, several studies have reported an inhibitory effect of GlcN on different types of tumors in

85

both humans and animals.11,12 The anti-tumor activity is postulated to be a selective effect on

86

nucleic acid synthesis in vivo, since RNA and DNA synthesis in treated tumors were

87

significantly inhibited, or alternatively, due to the alteration of glycosylation on the surface of

88

target proteins.13

89

In food applications GlcN has been used to generate antimicrobial, antioxidant14 and taste-

90

enhancing compounds;15 to enrich beverages16 or milk17 and to modify proteins to enhance their

91

functionality in food systems. For instance, Hrynets et al.18 took advantage of GlcN reactivity to

92

create novel glycoproteins at moderate temperatures resulting in an enhanced solubility and

93

emulsification capacity.


Page 5 of 41


94

At the same time, studies have reported a link between GlcN supplementation and diabetes

95

complications19,20 presumably due to GlcN-induced beta cell apoptosis and dysfunction.21

96

Furthermore, Candiano et al.22 indicated that GlcN possesses reactivity at physiological

97

temperatures producing fluorescent pyrazines compounds derived from the cyclocondensation

98

between two GlcN molecules, which were later reported to cause DNA strand breakage.23

99

Some of the GlcN`s effects, for instance taste-enhancing and antimicrobial properties as well as

100

protein modifications, are likely related to its high reactivity; however, to date no plausible

101

explanations for this phenomenon have been provided. It is also surprising that only a few

102

studies have focused on GlcN degradation at moderate temperatures (i.e. 37ºC)24,25 and, to best

103

of our knowledge, no research has evaluated the formation of α-dicarbonyls (α-DCs) from this

104

popular dietary supplement. It was expected that GlcN, a Heyns/Amadori-like compound, would

105

rapidly form reactive intermediates like α-DCs and several advanced and stable condensation

106

products as compared to GlcNAc. Hence, the main objective of this study was to investigate the

107

chemistry of GlcN degradation in vitro (non-enzymatic degradation) and to compare it to

108

GlcNAc and a reaction mixture containing Glc+NH3 (Glc-NH3). To meet this objective GlcNAc,

109

Glc-NH3 and GlcN were incubated at 37°C and α-DC were analyzed over time using liquid

110

chromatography and Orbitrap mass spectrometry. Condensation products were also evaluated

111

over time using both spectroscopic and mass spectrometric techniques.

112

MATERIALS AND METHODS

113

Chemicals

114

N-acetyl-D-glucosamine

115

hydrochloride (≥99% purity), potassium phosphate monobasic and dibasic, sodium azide, HPLC-

116

grade solvents (acetonitrile, methanol, formic acid), glucosone (2-keto-D-glucose; ≥98.0%

(≥99%

purity),

D-glucose

(≥99.5%

purity),

D-glucosamine



Page 6 of 41

117

purity; MW 178.14 Da), glyoxal (ethanedial; 40% in H2O; MW 58.04 Da), methylglyoxal (2-

118

oxopropanal; 40% in H2O; MW 72.06 Da), diacetyl (butane-2,3-dione; ≥95.0% purity; MW 86.09

119

Da) and 1,2-diaminobenzene were purchased from Sigma-Aldrich (St. Louis, MO). 3-

120

deoxyglucosone (3-Deoxy-D-erythro-hexosulose; ≥95% purity; MW 162.14 Da) was obtained

121

from Cayman Chemical (Ann Arbor, MI). Ammonium hydroxide was purchased from Fisher

122

Scientific. SPE tC-18 Sep-Pak Vac 6 cc columns were obtained from Waters (Milford, MA).

123

Filtration membranes (0.22 µm) were from Millipore (Billerica, MA). Standard ion calibration

124

solution (Pierce LTQ) for electrospray ionization in positive mode was purchased from Thermo

125

Scientific (Rockford, IL, USA). All buffers and reaction mixtures were prepared with Milli-Q

126

purified distilled water (Millipore, Bedford, MA, USA).

127

Experimental Design

128

In this study the non-enzymatic degradation of GlcN, GlcNAc and Glc-NH3 was evaluated over

129

time at 0, 0.5, 1, 2, 3, 6 and 12 d in phosphate buffer at 37°C using UV-Vis and fluorescence

130

spectroscopic analyses as well as mass spectrometric techniques. In the first part of the study,

131

UV-Vis and fluorescence profiles were assessed over time to understand the major changes

132

occurring during the degradation reaction of the monosaccharides. In this regard, a total of 42

133

tubes (2 tubes × monosaccharide × incubation time) were randomly placed within the incubator

134

and the reaction was monitored over time. Three independent trials were conducted resulting in 6

135

replicates per treatment.

136

Following this exploratory test, a dedicated study on α-DCs identification and production was

137

conducted over time using the same design as reported before. Liquid chromatographic

138

separations and mass spectrometry techniques were used to identify and quantify the major

139

degradation products. Finally, the last part of this study focused on the production of fluorescent 6 ACS Paragon Plus Environment

Page 7 of 41


140

and stable condensation products obtained by the reaction between the main α-DCs produced

141

during GlcN degradation and GlcN. In this regard, selected α-DCs (glucosone (G), 3-

142

deoxyglucosone (3-DG), glyoxal (GO) and methylglyoxal (MGO)) were incubated with GlcN at

143

equimolar concentrations for 0, 4, 8, 12, 24 and 48 h at 37°C, and their UV-Vis and fluorescence

144

profiles recorded. Three tubes per selected α-DCs were incubated over time resulting in 6

145

replicates.

146

Incubation of GlcNAc, Glc-NH3 and GlcN Aliquots

147

15 mg/mL solutions of GlcNAc, Glc-NH3 (equimolar mixture) and GlcN were prepared in 50

148

mM phosphate buffer. The pH of the final solutions, 7.4, was adjusted (Denver Instrument

149

UltraBasic pH Meter) if necessary. NaN3 was added to a final concentration of 0.02% to prevent

150

microbial growth. The reaction mixtures were transferred into plastic screw-cap tubes, randomly

151

arranged within an incubator (New Brunswick Scientific, Edison, NJ) and incubated at 37˚C.

152

Samples aliquots were collected at 0.5, 1, 2, 3, 6 and 12 days. The controls were reaction

153

solutions at zero hours. The pH value of the solution over time was checked, but not adjusted

154

during the reaction to enable a natural time course generation of reaction’s reactive

155

intermediates. Decrease in pH within 0.3-0.45/0.5 units was found for GlcN (6-12 reaction days,

156

respectively) and 0.1 unit Glc-NH3 solutions (12 reaction days).

157

Preparation of α-Dicarbonyls and Glucosamine Reaction Mixtures

158

Equimolar (2.5 mM) concentrations of GlcN and commercially available standards of G, 3-DG,

159

GO and MGO were solubilized in 50 mM phosphate buffer (pH 7.4), containing 0.02% NaN3.

160

Controls included fresh and incubated solutions of GlcN and each α-DCs separately. All reaction

161

mixtures were incubated at 37°C for 4, 8, 12, 24 and 48 h.

162

Spectrophotometric measurements 7 ACS Paragon Plus Environment


Page 8 of 41

163

UV-Vis absorbance and fluorescence spectra of 15 mg/mL solutions of GlcNAc, Glc-NH3 and

164

GlcN were recorded using the Spectramax M5 multi-mode microplate reader (Molecular

165

Devices, Sunnyvale, CA). For both analyses readings were taken with 1-cm quartz cuvettes and

166

samples diluted (1:20, v/v) with 50 mM phosphate buffer (pH 7.4). The absorbance was recorded

167

in the range of 200 to 500 nm. For fluorescence, the excitation wavelength was set at 347 and an

168

emission was recorded from 364 to 600 nm. A blank (buffer) was subtracted from all spectra

169

automatically. Absorbance (200-600 nm) and fluorescence (λexc = 347, λem = 364-600) of α-DCs

170

standards separately or in the mixture with GlcN were measured in the same way at the

171

concentration of 2.5 mM. For all GlcN-α-DCs mixtures, controls of incubated GlcN and

172

respective incubated α-DCs for each time periods were substracted.

173

Free α-Dicarbonyl Compounds: Extraction, Identification and Quantitation

174

Solid-Phase Extraction (SPE) and Derivatization Procedure

175

The extraction procedure was based on a method reported by Papetti et al.26 with slight

176

modification. Briefly, collected over time aliquots of GlcNAc, Glc-NH3 and GlcN were spiked

177

with 0.006 g of 1,2-diaminobenzene (o-phenylenediamine (OPD)), the pH was adjusted to 3.00 ±

178

0.02 with 4 M HCl and derivatized at 37°C for 1 h. The derivatized solution was passed through

179

a preconditioned SPE cartridge (tC-18 Sep-Pak, Waters, Milford, MA, USA). The cartridge was

180

washed with 2 mL of water, and quinoxalines were eluted with 4 mL of a MeOH/H2O mixture

181

(90/10, v/v) at a flow rate close to 2 mL/min. The first 1 mL was discarded, whereas the next 2

182

mL were used for analysis. The controls were treated exactly as collected sugar aliquots except

183

no derivatizing agent was added. To ensure complete derivatization, a preliminary test was

184

performed, where derivatization times of 0.5, 1, 2, 4 and 8 h were tested. The maximum peak


Page 9 of 41


185

area was reached after a derivatization for 1 h without further increase during longer

186

derivatization times.

187

UHPLC–UV Analysis of Quinoxalines Derivatives Formed from α-Dicarbonyl Compounds

188

and o-Phenylenediamine

189

α-DCs analysis was conducted using an Ultrahigh Performance Liquid Chromatography

190

apparatus (Shimadzu, Columbia, MD) consisting of two pumps LC-30AD, an SPD-M20A photo

191

diode array detector (PDA), DGU–20A5 degasser, SIL–30AC autosampler (operated at 4°C) and

192

CTO-20 AC column oven. The UHPLC system was controlled by a personal computer equipped

193

with LabSolutions software operating in a Windows XP. The separation was achieved following

194

the method described by Papetti et al.26 using Ascentis Express ES-C18 column (150 × 4.6 mm,

195

2.7 µm particles; Sigma-Aldrich, St. Louis, MO) with a UHPLC pre-column filter (UltraShield

196

Analytical Scientific Instruments, Richmond, CA) operating at 25.0 ± 0.5°C and a flow rate of

197

0.3 mL/min. The binary mobile phase consisted of (A) 0.1% formic acid in water and (B)

198

methanol. The 120 min gradient is described as follows: 0–5 min (90–85% A), 5–13 min (85–

199

80% A), 13–40 min (80% A), 40–65 min (80–70% A), 65–90 min (70–50% A), 90–100 min

200

(50–0% A), 100–105 min (0% A), and 105–110 min (0–90% A); the column was then re-

201

equilibrated with the initial mobile phase for 10 min. The injection volume was 5 µL. The PDA

202

detector was set at 314 and 335 nm to record the peaks, and UV spectra recorded from 215 to

203

420 nm. Before injection sample solutions were filtered with PVDF syringe filter (13 mm, 0.22

204

µm; Millipore Millex, Billerica, MA, USA).

205

Identification and Quantitation of Quinoxaline Derivatives

206

The identification of major α-DCs was based on comparison of their retention times of the

207

reference compounds (quinoxalines derivatives of G, 3-DG, GO, MGO and DA) and spectral 9 ACS Paragon Plus Environment


Page 10 of 41

208

characteristics. For unequivocal identification of the major α-DCs, ten UHPLC peaks were

209

manually collected and subjected to mass spectrometry analyses. The accurate mass and MS/MS

210

fragmentation patterns were compared to authentic standards and to the reference data. Mass

211

spectrometry analyses were performed in duplicates.

212

For quantitation external calibration was used. Each quinoxaline derivative was diluted to the

213

final concentrations ranging from 60 - 2000 (G), 30 - 2000 (3-DG), 4.5 - 50 (GO), 1.0 - 10

214

(MGO), and 1.25 - 7.5 µM for DA. Calibration curves were prepared for each trial and each

215

concentration was analyzed in triplicate. The peak area was plotted against concentration and the

216

regression equations were calculated. The correlation coefficients for all calibration curves were

217

R

218

(G), 0.27 ± 0.00 (3-DG), 0.13 ± 0.00 (GO), 0.09 ± 0.00 (MGO) and 0.18 ± 0.00 µM (DA). The

219

average limit of quantification (LOQ) was determined as 6.52 ± 0.21 (G), 0.81 ± 0.01 (3-DG),

220

0.39 ± 0.01 (GO), 0.27 ± 0.01 (MGO), and 0.55 ± 0.02 µM (DA) by assuming a signal-to-noise

221

ratio (S/N) 3:1 for LOD and S/N 10:1 for LOQ. The following ranges of sample concentrations

222

(in µM) were extrapolated from the calibration curves: 169 - 861 (G), 67 - 1289 (3-DG), 7.9 - 32

223

(GO), 2.5 - 6.8 (MGO) and 1.9 - 3.7 (DA). These values were then expressed as mg/kg GlcN.

224

Direct Infusion Orbitrap Mass Spectrometry Analyses (DIMS)

225

Collected α-DCs fractions from UHPLC and GlcN aliquots (diluted in 50:50 (methanol:water,

226

v/v)) were subjected to DIMS. All Orbitrap measurements were carried out by using the Ion Max

227

electron spray ionization (ESI) source (Thermo Fisher Scientific) mounted on a LTQ Orbitrap

228

XL (Thermo Scientific, San Jose, CA, USA). The Orbitrap mass analyzer was calibrated with

229

standard ion calibration solutions (Pierce LTQ, Thermo Scientific). Following parameters were

230

applied: sheath gas flow 15 arbitrary units, auxiliary gas flow 5 arbitrary units, and capillary

2

≥0.999. The average (from three trials) limit of detection (LOD) was calculated as 2.15 ± 0.07


Page 11 of 41


231

temperature of 275°C. Samples (3 µl) were injected manually performing direct infusion using

232

100 µl syringe (Hamilton, Reno, NV, USA) and the on board syringe pump. DIMS analyses

233

were performed at a mass resolution of 60.000 at m/z 400 and spectra were acquired over the

234

range of m/z 50 - 2000. The automatic gain control (AGC) target was set to 1e6 and the

235

maximum injection time to 250 ms. For MS/MS experiments, ions of interest were isolated with

236

an isolation width of 1-2 Da and collision induced dissociation was conducted at varying

237

voltages which were optimized for each sample. An activation time of 30 ms was used with a

238

tube lens voltage of 110 V and a capillary temperature of 275°C. Data acquisition and processing

239

were performed using Xcalibur software (Thermo Scientific).

240

Statistical Analysis

241

Spectroscopic profiles obtained from UV-Vis and fluorescence analyses from the three different

242

independent trials were averaged and reported as graphs. α-DCs production over time was

243

subjected to the analysis of variance (ANOVA) using the PROC MIXED procedure of SAS (v.

244

9.3, SAS Institute Inc., Cary, NC). The model tested incubation time as a fixed variable and used

245

the trial replication as a random variable. Tukey’s honestly significant difference (p < 0.05)

246

multiple range test was conducted to determine differences between means.

247

RESULTS AND DISCUSSION

248 249

UV-Vis and Fluorescence Profiles of Non-Enzymatic Modification of GlcNAc, Glc-NH3 and GlcN over time.

250

UV-Vis and fluorescent profiles can be a useful and simple tool to understand the major changes

251

occurring during a reaction. Solutions of GlcNAc, GlcN and Glc-NH3 were incubated at 37°C in

252

phosphate buffer from 0 to 12 days and their UV-Vis absorbance and fluorescence profiles

253

(Figure 1 A and B) were recorded to monitor their stability over time. No major changes in the

254

absorbance over time were obtained for GlcNAc, except a slight increase at λ = 220-235 at 12 11 ACS Paragon Plus Environment


Page 12 of 41

255

days. Minor absorbance increase was observed in Glc-NH3 mixture. On the contrary, a steady

256

increase of maxima of 274 and 320-360 were obtained for GlcN, suggesting a formation of

257

products which absorb in the UV region. The increase in absorbance at 320 nm was lower than at

258

274 nm, however the formation of a light yellow color was apparent after 0.5 day incubation.

259

The increase in UV absorbance at 274 and 320 nm was also reported by Zhang et al.27 who

260

incubated 0.02 M GlcN at 37°C for 50 h in phosphate buffer (pH 7.4). Candiano et al.22 and

261

Horowitz25 indicated that the increase in UV absorption at around 274 nm is due to the formation

262

of pyrazine compounds, which derive from the cyclocondensation between GlcN molecules;

263

which have been confirmed by MS analysis (see the section about condensation products). At the

264

same time increase at 320 nm was attributed to the formation of soluble pre-melanoidins,28

265

which was also confirmed by the production of light yellow-brown color over incubation time

266

(Figure 1 3A, inset). In addition, this hyperchromicity at 320-340 nm was apparent at longer

267

incubation times. No color production was observed for GlcNAc (Figure 1 1A inset); Glc-NH3

268

mixture changed from slightly yellow to yellow color over 6 to 12 days (Figure 1 2A inset),

269

whereas the time course of color development by GlcN was changed from yellowish to brown.

270

The fluorescence profile showed a progressive increase in the emission (λexc. 347/λem. 364-

271

600 nm) intensity over time for GlcN. A broad peak was observed with a maximum intensity

272

ranging from 420 to 450 nm. Glc-NH3 also showed an increased fluorescence, however with

273

significantly lower intensity and reaction velocity; at 12 days fluorescence emission was around

274

three times lower as compared to GlcN. On the other hand, GlcNAc fluorescence profile did not

275

change significantly over time. The wavelengths observed were typically used to assess the

276

progress of the Maillard reaction, and more specifically, the development of advanced glycation

277

products (AGEs).29 Zhang et al.24 reported the increase in fluorescence intensity (λexc. 340/λem. 12 ACS Paragon Plus Environment

Page 13 of 41


278

420 nm) during GlcN incubation at 37°C, and suggested an increase due to GlcN

279

cyclocondensation and rearrangements. Candiano et al.22 came to the same conclusion and

280

further proved that the increase in fluorescence emission at 430 nm (λexc = 362 nm) in lysine-

281

GlcN reaction mixture was due to GlcN cyclocondensation. The same authors indicated that

282

GlcN cyclocondensation products possessed optical and fluorescent properties of browning

283

compounds, which in turn could explain an increase in color intensity over time (Figure 1 3A

284

inset). These preliminary results suggest that with prolonged heating time GlcN

285

cyclocondensation and a variety of rearrangements occur, thus the formation of higher molecular

286

weight colored compounds.

287

Identification and Quantitation of the Major α-Dicarbonyls Produced from GlcN

288

Degradation by UHPLC and Mass Spectrometry Analyses.

289

In the previous section, UV and fluorescence analyses were discussed to evaluate the formation

290

of the final stable products of GlcN degradation (i.e. GlcN-GlcN autocondensation products).

291

However, GlcN, being a Heyns-like compound, we hypothesized the production of α-DCs,

292

which in turn have the potential to generate a variety of condensation products. α-DC are highly

293

reactive compounds, thus they were derivatized (condensation with OPD) to form stable

294

qunoxalines, followed by UHPLC and MS analyses. Separation of quinoxalines achieved with

295

chromatography26 allowed for identification of analytes. There were no α-DCs identified in

296

GlcNAc samples up to 12 days of incubation, whereas lower production was observed in Glc-

297

NH3 samples at 12 days of reaction as compared to GlcN (see Figure S1 in the Supporting

298

Information). Since the production of α-DCs at 37°C from Glc-NH3 was lower as compared to

299

GlcN, the main focus was to identify and quantify α-DCs derived from GlcN over time. Figure 2

300

A illustrates the representative UHPLC chromatogram of standard α-DCs compounds and α13 ACS Paragon Plus Environment


Page 14 of 41

301

DCs produced from GlcN incubated for 1 day. For confirmation of the peaks for which

302

commercial standards were available and the identification of unknowns, the major peaks were

303

collected and further subjected to MS analyses. Eight peaks in the UV chromatograms (Figure

304

2A (II)) were identified by their pseudomolecular peaks ([M + H]+, as detected by MS (Figure

305

2B). Individual peaks 1-8 were identified as being (1) glucosone, m/z = 251.1032; (2)

306

unidentified α-dicarbonyl, m/z = 235.0973; (3) 3-deoxyglucosone, m/z = 235.1081; (4) glyoxal,

307

m/z = 131.0611; (5) 3-hydroxy-2-oxo-propanal (hydroxypyruvaldehyde, HPA); m/z = 161.0720;

308

(6) 3,4-dideoxyglucosone-3-ene, m/z = 217.1707; (7) methylglyoxal, m/z = 145.0767 and (8)

309

diacetyl, m/z = 159.0925. Based on molecular weight and the reference data26 the peaks 9 and 10

310

were tentatively assigned to the quinoxaline derivatives of 2-amino-2-deoxy-4-glycosyl-5,6-

311

dihydroxy-2-oxohexanal (4-GlcN-3-DG) and 4-glycosyl-5,6-dihydroxy-2-oxohexanal (4-Glc-3-

312

DG), respectively (see precursor ions in Figure S2 in Supporting Information). The peaks

313

denoted as (a), (b) and (c) showed the m/z in the range of 317 to 355 (see Figure S3 in the

314

Supporting Information). The nature of these peaks is not clear, however it is suspected that they

315

derived from a variety of condensation products of GlcN itself or GlcN with α-DCs (i.e.

316

fructosazines). Hence, they are likely non-OPD condensed products, which are highly

317

hydrophilic fructosazines missing the benzene portion of OPD, and despite their high molecular

318

weight elute at early retention times in the hydrophilic part of the gradient. The product ions and

319

structures of quinoxaline derivatives of the seven identified peaks are shown in Figure 3. For Gqx

320

the molecular ion and subsequent losses of three water molecules were observed. Mittelmaier et

321

al.30 found very similar fragmentation pattern with the loss of three water molecules, except the

322

fragment ion at m/z 173.0 was dominant. Products ion spectra of 3-DGqx and its dehydration

323

product, 3,4-DGEqx showed similar fragment ions, apart from the latter missing fragments at m/z 14 ACS Paragon Plus Environment

Page 15 of 41


324

235.1081 and 171.0918. The MS/MS of the unidentified α-dicarbonyl (peak 2) showed two

325

similar fragment ions (m/z 145.0760/1 and 157.0761) with 3-DGqx, and could be an indication of

326

the presence of 1-deoxyglucosone (1-DG), 3-deoxygalactosone or 4-deoxyglucosone. However,

327

the presence of DA would support the presence of 1-DG, being a precursor of DA41 (Figure 4 A).

328

There was a minimal tendency of fragmentation for GOqx and MGOqx, with three and five

329

fragments, respectively, along with the highest intensity for precursor ions and fragment at m/z

330

77.0164 (77.1615 in MGO) typical for quinoxaline skeleton fragmentation.31 As well as for

331

DAqx four major MS/MS fragment ions were found. Under the described MS/MS conditions

332

m/z 161.0709, 143.0602, 133.0759, 131.2694 and 119.0602 were identified as the main

333

fragments of the parent molecule ion (m/z 161.0720) of HPAqx. The fragments ions from eight

334

identified α-DCs were in accordance with previously reported quinoxalines in various foods,

335

such as high-fructose corn syrup,32 coffee, barley and soy sauce,26 and honey.33 The MS/MS

336

fragmentation and proposed structures of 4-GlcN-3-DG and 4-G-3-DG are reported in the Figure

337

S3 in Supporting Information. After identifying the α-DC presence, five of them were quantified

338

and the concentrations expressed in mg/kg of GlcN are given in Table 1. It was found that G, 3-

339

DG and GO were already present in control samples, suggesting the partial degradation of GlcN

340

while stored in powder. Even though a freshly ordered GlcN was used in the analyses, rapid

341

development of α-DCs could be a possible explanation for their presence in the non-heated

342

sample. The elevated temperatures (90°C for hydrolysis, 45°C for evaporation, etc.) used during

343

GlcN’s extraction could induce its degradation and thus dicarbonyl compounds development. At

344

the beginning of the incubation G was the most abundant α-DC compound followed by 3-DG.

345

However, after 1 day incubation the concentration of G decreased significantly (p < 0.05) from

346

3245 to 1131 mg/kg GlcN, along with a significant increase in the amount 3-DG. At 0.5 and 1 15 ACS Paragon Plus Environment


Page 16 of 41

347

day of incubation G concentration was around six times higher than 3-DG. After heating for 12

348

days concentration of 3-DG was four times higher than G. G is a product of oxidative

349

degradation of Amadori, Heyns compound (see proposed pathway in Figure 4) or autoxidation of

350

sugar (i.e. glucose), while 3-DG is formed non-oxidatively through enolization pathway.34 These

351

results suggest the prevalence of an oxidative pathway during the first days of incubation,

352

whereas over time the enolization process dominates. In addition, slight pH drop at 6 and 12 days

353

could aid the 1,2-enolization to produce 3-DG.35 The concentrations of the other α-DCs: GO,

354

MGO and DA were significantly lower in comparison to G, 3-DG. Opposite relationship was

355

observed between GO and MGO formation over time. While the concentration of GO

356

significantly (p < 0.05) decreased from 34.9 mg/kg at 1 day to 12.6 mg/kg GlcN after 12 days,

357

the amount of MGO significantly (p < 0.05) increased from 4.5 to 10.4 mg/kg GlcN. DA reached

358

the highest concentration at 2 days of incubation (6.9 mg/kg GlcN), dropped almost twice by 3

359

days and was not detected with prolonged incubation. Obtained results confirm our hypothesis

360

that GlcN is an unstable amino sugar that degrades faster in phosphate buffer at 37°C, while

361

GlcNAc is stable in the condition used in this study.

362

Proposed Pathways of α-Dicarbonyls Formation from the Degradation of GlcN

363

Figure 4 reports the possible pathways involved in the production of α-DC identified in the

364

present study. This scheme can be divided into “oxidation” (A1), “enolization” (A2, A5),

365

“enolization/oxidation” (A3) and “interconversion” pathways. Glucosone, for instance, is most

366

likely produced from an oxidative pathway (A1) with a mechanism similar to the one proposed

367

by Hofmann et al.36 and later verified by Gobert and Glomb.37 Specifically, the imine or eminol

368

form of GlcN would be directly oxidized to form mono-alkylimines which upon deamination

369

will produce glucosone (Figure 4 A1). Subsequently, through a retro-aldol cleavage between C-3 16 ACS Paragon Plus Environment

Page 17 of 41


370

and C-4, G is proposed to form HPA, while the C-2/C-3 bond retroaldolization produces

371

glyoxal.36, 38

372

3-DG is proposed to be formed through the classical 1,2-enolization typical for the Heyns

373

compound;39 it is known that this compound can only go through the 1,2 enolization, as opposed

374

to the Amadori compound, which can also undergo 2,3-enolization, where 2,3-enediol is

375

deaminated to yield 1-DG.40 However, Yaylayan and Huyghues-Despointes39 reported that

376

Heyns products can isomerize to Amadori compounds through an interconversion reaction,

377

similar to Lobry de Bruyn-Alberda van Ekenstein transformation of fructose into glucose. In

378

Figure 5 A4 is a proposed reaction scheme where GlcN interconverts into Amadori compounds

379

via two possible condensation reactions: the first between two GlcN molecules and the second

380

between GlcN and NH3 formed during deamination of GlcN (see Figure S4 in Supporting

381

Information). The proposed interconversion pathway would also explain the presence of DA

382

through 1-DG degradation41 (Figure 4 A5). However, other new chemical pathways might be

383

involved in DA production. Increase in 3-DG was accompanied with increasing concentrations

384

of MGO over time (Table 1). Several studies42,43 suggested 3-DG as the precursor for MGO

385

formation, which is based on the non-oxidative fragmentation of the retroaldolization pathway.

386

In the model experiment of Thornalley et al.44 0.09 µM MGO was formed from 50 µM 3-DG,

387

which accounts for 0.18%. In the current study, at 0.5 d, MGO accounted for 1% of 3-DG,

388

whereas with incubation up to 12 days the percentage decreased to 0.23. GO is also proposed to

389

be formed through oxidative (Figure 4 A1) and enolization/oxidation pathways (Figure 4 A3).

390

Indeed, several pathways are proposed in the case of GO formation. The first, postulated by

391

Hoffman et al.,36 involves GO production from G as reported in Figure 4 A1. The other potential

392

mechanism is via enolization/oxidation pathways pathway, where GO might be produced from 17 ACS Paragon Plus Environment


Page 18 of 41

393

3-DG (Figure 4 A3) via vinylogous retroaldolization forming glycoaldehyde, which

394

subsequently oxidizes to form GO.45 The mechanism reported in Figure 4 A1, is unique, where

395

GO is proposed to originate from direct C-2/C-3 retroaldolization cleavage of GlcN in its

396

aldosamine form, forming erythrose and a two-carbon enaminol compound that will

397

subsequently undergo oxidation and deamination forming GO. This would be a similar oxidative

398

pathway to the one proposed by Hayashi and Namiki,46 with the only difference that in this

399

specific case the retro-aldol type cleavage would directly occur on the aldosamine form of GlcN

400

and not on its Schiff base (imine) form. In the Namiki type of rearrangements, GO is produced

401

before and independently of the G precursor. Hofmann and Heuberger47 reported GO production

402

in glucose/L-alanine mixture (pH 7.0, phosphate buffer, 95°C) at the early stages of the Maillard

403

reaction, prior to deoxyosone formation. A steady decrease in the amount of G was associated

404

with a decrease in GO concentrations (Table 1), therefore, it is plausible to propose that both

405

pathways described above as major routes for GO formation in the current study. HPA can also

406

be produced in a similar manner as previously described for GO, with the only difference that

407

retroaldol degradation is proposed to occur at C-3/C-4 of the Schiff base form of GlcN (Figure 4

408

A1).

409

Condensation Products from Non-enzymatic Modification of GlcN

410

As presented in Figure 1, incubation of GlcN results in an increased absorbance and

411

fluorescence, suggesting products are generated over time. To further investigate, aliquots of

412

GlcN incubated for half (Figure 5 A) and three days (Figure 5 B) were directly injected into a

413

MS to obtain a compound profile. As shown in Figure 5 A and B, several molecular ion species

414

were observed. GlcN showed peaks at m/z 180.0860 corresponding to its protonated molecular

415

ion [M+H]+, m/z 162.0755 and 144.0649 corresponding to mono- and didehydrated GlcN, 18 ACS Paragon Plus Environment

Page 19 of 41


416

respectively. The m/z 218.0418 is likely due to GlcN`s potassium adduct [M+K]+ while m/z

417

359.0836 corresponding to [2M+H]+, possibly due to N-glycoside formation between two GlcN

418

molecules. The m/z 361.1008 can also be due to an N-glycoside formation, but from a reduced

419

GlcN molecule (i.e. 2-amino-2-deoxy-D-glucitol) and GlcN. Remarkably, several other species

420

with pseudo molecular weights higher than m/z 300 were present. These species are the result of

421

two types of reactions: cyclocondensation (Figure 6), which results in pyrazine ring formation

422

and acyclic condensation (Figure 7) where no pyrazine rings are formed. One of the products of

423

GlcN

424

tetrahydroxybutyl)dihydropyrazine)], was identified at m/z 323.1451 (Figure 5 A). This

425

compound was originally identified and isolated by Candiano et al.22 during the incubation of

426

lysine in the presence of GlcN at 37°C. The products at m/z 321.1286 and 305.1336 believed to

427

be

428

deoxyfructosazine, respectively.23 Similar products of GlcN incubation at m/z 321 and 323 were

429

reported previously by Kashige et al.23 while studying the possible involvement of GlcN-derived

430

dihydrofructosazine in breaking DNA. It is also apparent that many other cyclocondensation

431

products may be produced as a consequence of the reaction between GlcN and α-DCs via the

432

mechanism proposed by Klinger.48 The proposed reaction mechanisms between the most

433

abundant α-DCs found in this study and GlcN are reported in Figure 6. The term “symmetric

434

cyclocondensation” indicates a reaction scheme that involves two identical GlcN (Figure 6 A1)

435

or α-DC molecules (Figure 6 A2) and “asymmetric cyclocondensation” refers to a reaction that

436

occurs between a specific α-dicarbonyl and GlcN (Figure 6 A3 and A4). Theoretically, the

437

formation of all masses reported in the Figure 6 is possible. The mass spectra of underivatized

438

GlcN (Figure 5 A and B) at 0.5 and 3 days show the presence of the peaks which correspond to

the

cyclocondensation,

products

of

namely

dihydrofructosazine

dihydrofructosazine

rearrangement,

producing

[2,5-bis(arabino-

fructosazine

and



Page 20 of 41

439

both “symmetric” (Figure 6 A1, A2) and “asymmetric” (Figure 6 A3, A4) cyclocondensation

440

products and in some cases the molecular weights of these products coincide (i.e. protonated m/z

441

305.1348; 321.1298; 323.1454). These data are also corroborated by the Figure 8, which

442

represents the UV-Vis (Figure 8 A) and fluorescence emission (Figure 8 B) profile of equimolar

443

mixture of G, 3-DG, GO and MGO with GlcN over a period of incubation from 0 to 48 h. In all

444

cases absorbance and fluorescence emission intensities increased over time owing to the

445

formation of cyclocondensation reaction products. Hence, GlcN possesses a trapping effect to the

446

nascent α-DCs produced from its degradation. Fluorophore formation was also reported24 to

447

represent the formation of heterocyclic compounds (i.e. pyrazines). Among four α-DCs tested, 3-

448

DG appears to be the most reactive towards GlcN, judging from an extensive accumulation of

449

fluorescence (240 units). 3-deoxyosones predominantly form pyrazines and imidazoles with

450

ammonia, which in turn are responsible for the sugar coloring.40 Therefore, the accumulation of

451

3-DG (Table 1) and its reaction products with GlcN (Figure 8) over time could be partially

452

responsible for GlcN browning with extended incubation periods (Figure 1 3A inset).

453

An example of acyclic condensation is the peak at m/z of 343.0889 identified at 3 incubation

454

days. It is proposed that this molecular ion corresponds to a reduced Schiff base from a

455

condensation between two molecules of GlcN. The possible mechanisms of formation are

456

illustrated in Figure 7 where the reduction of the GlcN-GlcN Schiff base would be promoted

457

from the dehydrogenation of the dihydrofructosazines to fructosazines. Supporting this

458

phenomenon is the observed decrease in intensity of m/z 323.1451 during incubation from 0.5 to

459

3 days with concomitant increase intensity of m/z at 321.1286 (Figure 5). Shimamura et al.49 also

460

reported the reducing capacity of dihydrofructosazine.


Page 21 of 41


461

In summary, this research demonstrated that GlcN, and not GlcNAc, is the unstable compound,

462

producing 3 - 6 mg of α-dicarbonyl compounds per gram of GlcN depending on incubation time.

463

The cause of this reactivity lays in the nature of GlcN, which is a Heyns (aldosamine) compound

464

with the potential to generate many reactive α-dicarbonyls, precursors of fluorescent

465

heterocyclic compounds such as pyrazines. By contrast, acetylation of GlcN decreases its

466

reactivity by blocking the nucleophilicity of the free -NH2 group. Considering the ubiquities of

467

GlcNAc in animal tissues, enzymatic acetylation appears to play a crucial role in minimizing the

468

less stable GlcN, which has the potential to form endogenous reactive α-DCs. Also, research

469

interests on GlcN application in food products are dramatically increasing. For instance, Jia et

470

al.50 proposed the bioconversion of GlcN to deoxyfructosazine and fructosazine in order to

471

produce flavour compounds for the food industry. Uzzan et al.17 enriched milk with GlcN with

472

the hope to produce a functional food product; however, the experiment was compromised due to

473

aggregation of milk protein upon pasteurization. This may indicate the cross-linking effects of α-

474

DCs generated from GlcN degradation (unpublished results from our research group also support

475

GlcN-induced myoglobin aggregation at moderate temperatures). In North America, GlcN has a

476

legislative status of “Natural Health Product” (or dietary supplement) and not of a food

477

ingredient; for instance a specific dose of 1.5 g daily is recommended to treat osteoarthritis. The

478

USDA does not recognize GlcN as “generally recognized as safe” due to lack of clinical studies

479

that its consumption is safe in young children, pregnant women and diabetics.16 Hence, more

480

studies are necessary for the use of GlcN as a food ingredient for both technological and safety

481

reasons.

482

ABBREVIATIONS USED



Page 22 of 41

483

α-DC, alpha-dicarbonyl compounds; AGEs, advanced glycation end-products; ANOVA, analysis

484

of variance; COS, chitosan oligosaccharide; DA, diacetyl; 3-DG, 3-deoxyglucosone; 3,4-DGE,

485

3,4-dideoxyglucoson-3-ene;

486

glycosaminoglycans; GFAT, glutamine:fructose-6-phosphate-aminotransferase; Glc, glucose;

487

GlcN, glucosamine; GlcN-6-P, glucosamine-6-phosphate; GlcNAc, N-acetyl-glucosamine; GO,

488

glyoxal;

489

hydroxypyrualdehyde; LOD, limit of detection; LOQ, limit of quantification; MGO,

490

methylglyoxal; MS, mass spectrometry; MS/MS, tandem mass spectrometry; nd, not detectable;

491

OPD, o-phenylenediamine; PDA, photo diode array detector; RA, retroaldolization; SPE, solid

492

phase extraction; UHPLC, ultrahigh performance liquid chromatography; USDA, United States

493

Department of Agriculture.

494

ACKNOWLEDGMENTS

495

The authors would like to thank Jack Moore at the Institute of Biomolecular Design (University

496

of Alberta, Edmonton) for technical assistance in mass spectrometry experiments.

497

FUNDING SOURCES

498

This research was funded by grant from Alberta Innovates-Bio Solutions (Al Bio) and Natural

499

Sciences and Engineering Research Council (NSERC) of Canada to Dr. M. Betti.

500

SUPPORTING INFORMATION

501

Supplemental UHPLC-UV chromatograms of quinoxaline derivatives of α-dicarbonyl

502

compounds produced from GlcNAc and Glc-NH3 incubated for 12 days; Precursor ions and

503

MS/MS spectra of the peaks 9 and 10 (from Figure 2 II); Production of ammonia during

504

incubation of GlcN solutions from 0 to 12 days. This material is available free of charge via the

505

Internet at http://pubs.acs.org.

HA,

hyaluronic

ESI,

acid;

electrospray

HBP,

ionization;

hexosamine

G,

glucosone;

biosynthesis

pathway;

GAGs,

HPA,


Page 23 of 41


506

REFERENCES

507

(1) Anderson, J.W.; Nicolosi, R.J.; Borzecella, J.F. Glucosamine effects in humans: a review of

508

effects on glucose metabolism, side effects, safety considerations and efficacy. Food Chem.

509

Toxicol. 2005. 43, 187-201.

510 511

(2) Block, J.A.; Oegema, T.R.; Sandy, J.D.; Plaas, A. The effects of oral glucosamine on joint health: is a change in research approach needed? Osteoarthr. Cartil. 2010, 18, 5-11.

512

(3) Uitterlinden, E.J.; Koevoet, J.L.; Verkoelen, C.F.; Bierma-Zeinstra, S. M.; Jahr, H.; Weinans,

513

H.; Verhaar, J.A.; van Osch, G.J. Glucosamine increases hyaluronic acid production in

514

human osteoarthritic synovium explants. BMC Musculoskelet. Disord. 2008, 9, 120-127.

515

(4) Qu, C-J.; Karjalainen, H.M.; Helminen, H.J.; Lammi, M.J. The lack of effect of glucosamine

516

sulphate on aggrecan mRNA expression and (35)S-sulphate incorporation in bovine primary

517

chondrocytes. Biochim. Biophys. Acta. 2006, 4, 453–459.

518 519

(5) Einbu, A,; Varum, K.M. Characterization of chitin and its hydrolysis to GlcNAc and GlcN. Biomacromolecules, 2008, 9, 1870–1875.

520

(6) Sibi, G.; Dhananjaya, K.; Ravikumar, K. R.; Mallesha, H.; Venkatesha, R. T. Preparation of

521

glucosamine hydrochloride from crustacean shell waste and it's quantitation by RP-HPLC.

522

American-Eur. J. Sci. Res. 2013, 8, 63-67.

523 524

(7) Schleicher, E. D.; Weigert, C. Role of the hexosamine biosynthetic pathway in diabetic nephropathy. Kidney Int. Suppl. 2000, 77, S13-8.

525

(8) Salazar, J.; Bello, L.; Chavez, M.; Anez, R.; Rojas, J.; Bermudez, V. Glucosamine for

526

osteoarthritis: biological effects, clinical efficacy, and safety on glucose metabolism.

527

Arthritis. 2014, 2014, 432463.



528 529

Page 24 of 41

(9) Heyns, K.; Koch, W. Z. Naturforsch. Über die bildung eines aminozuckers ausd-fruktose und ammoniak. Z Naturforsch [B]. 1952; 7B, 486-488.

530

(10) Mottram, D.S.; Elmore, J.S. Control of the Maillard reaction during the cooking of food. In

531

Controlling Maillard Pathways To Generate Flavors, edition 3; Mottram, D. S., Taylor, A.

532

J., Eds.; ACS Symposium Series 1042. American Chemical Society: Washington, DC, 2010;

533

pp. 143-155.

534

(11) Song, K.-H.; Kang, J.-H.; Woo, J.-K.; Nam, J.-S.; Min, H.-Y.; Lee, H.Y.; Kim, S.-Y.; Oh,

535

S.-H. The novel IGF-IR/Akt–dependent anticancer activities of glucosamine. BMC Cancer.

536

2014, 14, 31.

537 538

(12) Chesnokov, V.; Gong, B.; Sun, C.; Itakura, K. Anti-cancer activity of glucosamine through inhibition of N-linked glycosylation. Cancer Cell Int. 2014, 14, 45-55.

539

(13) Kantor, E. D.; Lampe, J.W.; Peters, U.; Shen, D.D.; Vaughan, T.L.; White, E. Use of

540

glucosamine and chondroitin supplements and risk of colorectal cancer. Cancer Cause

541

Control. 2013, 24, 1137–1146.

542

(14) Wu, S.; Hu, J.; Wei, L.; Du, Y.; Shi, X.; Zhang, L. Antioxidant and antimicrobial activity of

543

Maillard

reaction

products

from

xylan

with

chitosan/chitooligomer/glucosamine

544

hydrochloride/taurine model systems. Food Chem. 2014, 148, 196-203.

545

(15) Katsumata, T.; Nakakuki, H.; Tokunaga, C.; Fujii, N.; Egi, M.; Phan, T.H.; Mummalaneni,

546

S.; DeSimone, J.A.; Lyall, V. Effect of Maillard reacted peptides on human salt taste and the

547

amiloride-insensitive salt taste receptor (TRPV1t). Chem. Senses, 2008, 33, 665–680.

548

(16) Kralovec, A.; Barrow, C.J. Glucosamine production and health benefits. In Marine

549

Nutraceuticals and Functional Food, edition1; Barrow, C., Shahidi, F, Ed.; CRC Press:

550

Boca Raton, FL, 2008; pp.198-227. 24 ACS Paragon Plus Environment

Page 25 of 41

551 552


(17) Uzzan, M.; Nechrebeki, J.; Labuza, T.P. Thermal and storage stability of nutraceuticals in a milk beverage dietary supplement J. Food Sci. 2007, 72, E109-114.

553

(18) Hrynets, Y.; Ndagijimana, M.; Betti, M. Non-enzymatic glycation of natural actomyosin

554

(NAM) with glucosamine in a liquid system at moderate temperatures. Food Chem. 2013,

555

139, 1062-1072.

556

(19) Nakamura, M.; Barber, A.J.; Antonetti, D.A.; LaNoue, K.F.; Robinson, K.A.; Buse, M.G.;

557

Gardner, T.W. Excessive hexosamines block the neuroprotective effect of insulin and induce

558

apoptosis in retinal neurons. J. Biol. Chem. 2001, 276, 43748-43755.

559

(20) Hawkins, M.; Barzilai, N.; Chen, W.; Angelov, I.; Hu, M.; Cohen, P.; Rossetti, L. Increased

560

hexosamine availability similarly impairs the action of insulin and IGF-1 on glucose

561

disposal. Diabetes. 1996, 45, 1734–1743.

562 563 564

(21) Zhang, X.; Tong, N.; Li, Y. Concurrently using rosiglitazone prevents glucosamine-induced islet and beta Cell apoptosis and dysfunction. Cell Biochem. Funct. 2010, 28, 509-514. (22) Candiano, G.; Ghiggeri, G.M.; Gusmano, R.; Zetta, L.; Benfenati, E.; Icardi, G. Reaction of

565

2-amino-2-deoxyglucose

and

lysine:

Isolation

and

characterization

566

bis(tetrahydroxybutyl)pyrazine. Carbohydr. Res. 1988, 184, 67-75.

of

2,5

567

(23) Kashige, N.; Yamaguchi, T.; Mishiro, N.; Hanazono, H.; Miake, F.; Watanabe, K. Possible

568

involvement of dihydrofructosazine in the DNA breaking activity of D-glucosamine. Biol.

569

Pharm. Bull. 1995, 18, 653–658.

570

(24) Zhang, X.; Ma, Y.; Liu, H.; De Sa, P.F.; Brown, P.R.; Dain, J.A. Capillary electrophoresis

571

separation of autocondensation glycation products of glucosamine. J. Capill. Electrophor.

572

Microchip. Technol. 2003, 1-2, 33-37.



573 574 575 576

Page 26 of 41

(25) Horowitz, M.I. Nonenzymatic modifications of amino sugars in vitro. Arch. Biochem. Biophys. 1991, 288, 317-323. (26) Papetti, A.; Mascherpa, D.; Gazzani, G. Free α-dicarbonyl compounds in coffee, barley coffee and soy after in vitro digestion. Food Chem. 2014, 164, 259-265.

577

(27) Zhang, X.; Ma, Y.; Liu, H.; De Sa, P.F.; Brown, P.R.; Dain, J.A. Monitoring the formation

578

of Maillard reaction products of glucosamine with fibrinogen and human serum albumin

579

using capillary electrophoresis. Anal. Biochem. 2004, 325, 255-259.

580

(28) Fogliano, V.; Monti, S.M.; Musella, T.; Randazzo, G.; Ritieni, A. Formation of coloured

581

Maillard reaction products in a gluten-glucose model system. Food Chem. 1999, 66, 293-

582

299.

583 584

(29) Ferrer, E.; Alegria, A.; Farre, R.; Clemente, G.; Calvo, C. Fluorescence, browning index, and color in infant formulas during storage. J. Agric. Food Chem. 2005, 53, 4911-4917.

585

(30) Mittelmaier, S.; Fünfrocken, M.; Fenn, D.; Fichert, T.; Pischetsrieder, M. Identification and

586

quantification of the glucose degradation product glucosone in peritoneal dialysis fluids by

587

HPLC/DAD/MSMS. Anal. Bioanal. Chem. 2011, 401, 1183–1193.

588

(31) Pfeifer, Y.V.; Kroh, L.W. Investigation of reactive alpha-dicarbonyl compounds generated

589

from the Maillard reactions of L-methionine with reducing sugars via their stable

590

quinoxaline derivatives. J. Agric. Food Chem. 2010, 58, 8293-8299.

591

(32) Gensberger, S.; Mittelmaier, S.; Glomb, M.A.; Pischetsrieder, M. Identification and

592

quantification of six major α-dicarbonyl process contaminants in high-fructose corn syrup.

593

Anal. Bioanal. Chem. 2012, 403, 2923-2931.


Page 27 of 41


594

(33) Mavric, E.; Wittmann, S.; Barth, G.; Henle, T. Identification and quantification of

595

methylglyoxal as the dominant antibacterial constituent of Manuka (Leptospermum

596

scoparium) honeys from New Zealand. Mol. Nutr. Food Res. 2008, 52, 483-489.

597

(34) Kerler, J.; Winkel, C.; Davidek, T.; Blank, I. Basic chemistry and process conditions for

598

reaction flavours with particular focus on Maillard-type reactions. In Food Flavour

599

Technology, edition 2; Taylor, A. J., Linforth, R. S. T., Eds.; Blackwell Publishing Ltd:

600

Chichester, UK, 2010; Vol. 2, pp. 51–81.

601

(35) Sena, C.M.; Metafome, P.; Crisostomo, J.; Rodrigues, L.; Fernandes, R.; Pereira, P.; Seica,

602

R.M. Methylglyoxal promotes oxidative stress and endothelial dysfunction. Pharmacol. Res.

603

2012, 65, 497-506.

604

(36) Hofmann, T.; Bors, W.; Stettmaier, K. Studies on radical intermediates in the early stage of

605

the nonenzymic browning reaction of carbohydrates and amino acids. J. Agric. Food Chem.

606

1999, 47, 379–390.

607 608 609 610

(37) Gobert, J.; Glomb, M.A. Degradation of glucose: reinvestigation of reactive α-dicarbonyl compounds. J. Agric. Food Chem. 2009, 57, 8591–8597. (38) Degen, J.; Hellwig, M.; Henle, T. 1, 2-dicarbonyl compounds in commonly consumed foods. J. Agric. Food Chem. 2012, 60, 7071-7079.

611

(39) Yaylayan, V.A., Huyghues-Despointes, A. Chemistry of Amadori rearrangement products:

612

analysis, synthesis, kinetics, reactions and spectroscopic properties. Crit. Rev. Food Sci.

613

1994, 34, 321–369.

614

(40) Belitz, H.-D; Grosch, W;· Schieberle, P. Carbohydrates. In Food Chemistry, edition 4;

615

Belitz, H.-D., Grosch, W., Schieberle, P., Eds.; Springer-Verlag Berlin and Heidelberg

616

GmbH: Germany, 2009; pp. 248-337. 27 ACS Paragon Plus Environment


617 618 619 620 621 622 623 624 625

(41) Shibamoto, T. Diacetyl: occurrence, analysis, and toxicity. J. Agric. Food Chem. 2014, 62, 4048−4053. (42) Weenen, H. Reactive intermediates and carbohydrate fragmentation in Maillard chemistry. Food Chem. 1998, 62, 393–401. (43) Yaylayan, V.A.; Keyhani, A. Origin of carbohydrate degradation products in L-alanine/D[13C]glucose model systems. J. Agric. Food Chem. 2000, 48, 2415–2419. (44) Thornalley, P.J.; Langborg, A.; Minhas, H.S. Formation of glyoxal, methylglyoxal and 3deoxyglucosone in the glycation of proteins by glucose. Biochem. J. 1999, 344, 109-116. (45) Nursten, H. The chemistry of nonenzymic browning In The Maillard Reaction: Chemistry,

626

Biochemistry and Implications, edition 1; Nursten, H., Ed.; Royal Society of Chemistry:

627

Cambridge, UK, 2005; pp. 5-30.

628 629

Page 28 of 41

(46) Hayashi, T.; Namiki, M. Role of sugar fragmentation in the Maillard reaction. Dev. Food Sci. 1986, 13, 29–38.

630

(47) Hofmann, T.; Heuberger, S. The contribution of coloured Maillard reaction products to the

631

total colour of browned glucose/l-alanine solutions and studies on their formation. Z

632

Lebensm Unters Forsch A. 1999, 208, 17–26.

633

(48) Klinger, K.M.; Liebner, F.; Fritz, I.; Potthast, A.; Rosenau, T. Formation and ecotoxicity of

634

N heterocyclic compounds on ammoxidation of mono- and polysaccharides. J. Agric. Food

635

Chem. 2013, 61, 9004−9014.

636

(49) Shimamura, T.; Takamori, A.; Ukeda, H.; Sawamura, M. Reduction mechanism of

637

tetrazolium salt XTT by a glucosamine derivative. Biosci. Biotechnol. Biochem. 2003, 67,

638

295-299.


Page 29 of 41


639

(50) Jia, L.; Wang, Y.; Qiao, Y.; Qia, Y.; Hou, X. Efficient one-pot synthesis of

640

deoxyfructosazine and fructosazine from D-glucosamine hydrochloride using a basic ionic

641

liquid as a dual solvent-catalyst. RSC Adv. 2014, 4, 44253–44260.

642

FIGURE CAPTIONS

643

Figure 1. UV-Vis absorbance (A) and fluorescence spectra (B) profiles of GlcNAc (1A, 1B),

644

Glc-NH3 (2A, 2B) and GlcN (3A, 3B) as a function of heating time from 0 to 12 days. The

645

values are expressed in arbitrary units (A.U.). Inset images (left) show photographs of the

646

corresponding incubated sugar solutions. The images were taken for non-diluted experimental

647

solution incubated at 15 mg/mL.

648 649

Figure 2. UHPLC and MS analyses of quinoxaline derivatives of α-dicarbonyl compounds

650

produced from GlcN incubation over time. (A) Chromatograms of (I) a reference quinoxaline

651

mixture of glucosone (G), 3-deoxyglucosone (3-DG), glyoxal (GO), methylglyoxal (MGO) and

652

diacetyl (DA). (II) Representative chromatogram of GlcN incubated for 1 day, derivatized with

653

OPD and acquired by UHPLC with UV detection at 314 nm. Numbers indicate the peaks of the

654

quinoxalines of (1) G, (2) unidentified α-DC, (3) 3-DG, (4) GO, (5) HPA, (6) 3,4 - DGE, (7)

655

MGO, (8) DA, (9) 4-GlcN- 3-DG, (10) 4-G-3-DG and a, b, c peaks corresponding to non-OPD

656

derived GlcN condensation products. Inset shows a zoomed-in view of the peaks eluted at 60-90

657

min. (B) Precursor ions and retention times of the peaks corresponding to the same ten numbers

658

as indicated above in the part II.

659 660

Figure 3. MS/MS spectra and structures of quinoxaline α-dicarbonyl compounds produced

661

during GlcN degradation: (1) glucosoneqx, (2) unidentified α-DCqx, (3) 3-deoxyglucosoneqx, (4) 29 ACS Paragon Plus Environment


Page 30 of 41

662

glyoxalqx, (5) hydroxypyruvaldehydeqx, (6) 3,4-dideoxyglucosone-3-eneqx, (7) methylglyoxalqx

663

and (8) diacetylqx. The respective quinoxaline precursor ions are shown in the Figure 2 B.

664 665

Figure 4. Proposed pathways of generation of α-dicarbonyl compounds from GlcN. G,

666

glucosone;

667

dideoxyglucosone-3-ene; GO, glyoxal; MGO, methylglyoxal; HPA, hydroxypyruvaldehyde; RA,

668

retroaldolization.

3-DG,

3-deoxyglucosone;

1-DG,

1-deoxyglucosone;

3,4-DGE,

3,4-

669 670

Figure 5. Mass spectrum of GlcN incubated at 0.5 (A) and 3 (B) days and directly infused into

671

the LTQ Orbitrap – MS instrument. Inset: enlargement of the respective spectra in the range of

672

m/z 300-380.

673 674

Figure 6. Proposed structures and theoretical monoisotopic molecular weights (Mw) of

675

cyclocondensation products of GlcN only or GlcN and α-dicarbonyl compounds. Structures are

676

shown as neutral molecules.

677 678

Figure 7. Proposed structures and theoretical monoisotopic molecular weights (Mw) of products

679

deriving from GlcN acyclic condensation. Structures are shown as neutral molecules.

680 681

Figure 8. Absorbance (A) and fluorescence emission (B) spectra of control and incubated

682

equimolar mixtures of GlcN and α-dicarbonyl compounds. Part (A): GlcN incubated with (a) G,

683

(b) 3-DG, (c) GO and (d) MGO. Part (B): GlcN incubated with (i) G, (ii) 3-DG, (iii) GO and (iv)

684

MGO. 30 ACS Paragon Plus Environment

Page 31 of 41


Table 1. Concentrations of five different α-dicarbonyl compounds found in GlcN incubated over time. All values are expressed as mg of α-dicarbonyl compound per kg of GlcN. The mean ± standard deviation of three independent experiments (n = 6) is shown. α-dicarbonyls

Free α-dicarbonyl compounds (mg/kg of GlcN) α 0.5 d 1d 2d 3d 6d

0d Glucosone 3-Deoxyglucosone Glyoxal

709 ± 68e

2634± 87b

3245 ± 90a

f

e

de

2710 ± 134b cd

2443 ± 85bc c

12 d

p-value

2132 ± 89c

1131 ± 68d

**

b

a

272 ± 19

451 ± 32

571 ± 27

583 ± 33

707 ± 63

1905 ± 42

4535 ± 82

**

11.2 ± 1.2e

26.6 ± 2.1bc

34.9 ± 3.4a

28.4 ± 2.1b

24.9 ± 1.6c

17.7 ± 0.9d

12.6 ± 1.1e

**

de

cd

a

e

c

b

Methylglyoxal

nd

4.5 ± 1.3

5.5 ± 1.1

6.4 ± 1.0

7.3 ± 0.8

8.7 ± 0.9

10.4 ± 2.3

**

Diacetyl

nd

nd

6.1 ± 1.1b

6.9 ± 0.9a

3.7 ± 1.0c

nd

nd

**

The values on the same line with different uppercase superscript letters differ significantly (p < 0.05). nd - not detectable. ** Indicates p < 0.0001.



Page 32 of 41

Figure 1

A 1A

0d 0. 1 2 6 12 5 d d d d C 0.5d 1d 2d 3d 6d 12d

B

1B

0d

2A

2B C 0.5d 1d 2d 3d 6d 12d

3A

3B C 0.5d 1d 2d 3d 6d 12d



Figure 2

3-DGqx

A

I

Gqx

Absorbance at 314 nm

Page 33 of 41

MGOqx

GOqx

DAqx

1

II 4 5 a

2

6

7

8

3 b c

910 Retention time (min)

B

1 tR = 21.5 min

5 tR = 71.8 min

4

2

3

tR = 28.5 min

tR = 33.2 min

6

7

8

tR = 79.9 min

tR = 87.4 min

tR = 76.9 min

tR = 67.1 min



Page 34 of 41

2

1 Glucosoneqx

Unidentified

H

+

m/z = 251.1032

3

4

3-Deoxyglucosoneqx

Glyoxalqx H

H

+

+

m/z = 131.0611

m/z = 235.1081


Page 35 of 41


6

5 Hydroxypyruvaldehydeqx

3-Dideoxyglucosone-3-eneqx

H

+

m/z = 161.0720

H

+

m/z = 217.1707

8

7

Diacetylqx

Methylglyoxalqx H H

+

+

m/z = 159.0925 m/z = 145.0767



Page 36 of 41

Figure 4 A5`Ènolization

A1

C-2/C-3 RA

`Òxidation pathways`` C-3/C-4 RA

-

MGO

GO Erythrose

Oxidation C-3/C-4 RA G

G-monoalkylimine

3-DG Glyceraldehyde

HPA

Glyceraldehyde

C-3/C-4 RA

1,2-Enolization

A4

Oxidation

Schiff Base

Dihydroxyacetonealkylimine

Glyceraldehyde

-

HPAmonoalkylimine

HPA

C-2/C-3 RA -

Oxidation

`Ìnterconv ersion pathway``

Amadori product

Schiff base

2,3-Enolization

Heyns product (Glucosamine)

Glycoaldehydealkylimine

Erythrose

-

GOmonoalkylimine

1,2-Enolization

A2 C-4/C-5 RA

-

`Ènolization

DA C-3/C-4 RA

-

Glyceraldehyde

1-DG

GO

Diacetylformoin

MGO

GO 3-DG

: R = 2-deoxyglucose

-

A3 `Ènolization/oxidation pathway``

R=H

C-4/C-5 RA

Oxidation

GO

Glycoaldehyde 3,4-DGE


Page 37 of 41


Figure 5

A

[M+H]+ [2M-2H2O]+

[M+K]+

[2M+H]+

+

[M+H-1H2O]

[M+H]+

B [2M+H]+ [2M-2H2O-2H]+

[M+H-1H2O]+

[M+K]+ [M+H-2H2O]+



Page 38 of 41

Figure 6

-

A1

Mw = 304.1270 -2

-2H Glucosamine Glucosamine

Mw = 322.1376

A2 NH NH

Mw = 320.1220 3

-3

3

3-Deoxyglucosone Mw = 304.1270

3-Deoxyglucosone

A3 NH

3

-3

3-Deoxyglucosone Mw = 304.1270

Glucosamine

A4 NH 3

-3

Glucosone Glucosamine

Mw = 320.1220


Page 39 of 41


Figure 7

-H2O

Mw = 358.1588

Schiff base Mw = 340.1482 Dihydrofructosazine

Dihydrofructosazine R=

+2H

+2H Fructosazine

Fructosazine

Reduced Schiff base Mw = 342.1638



Page 40 of 41

Figure 8

A

a

Absorbance (A.U.)

0.5

b

0.5 G-GlcN, 4 h

3-DG-GlcN, 4 h

G-GlcN, 8 h

3-DG-GlcN, 8 h

G-GlcN, 12 h

3-DG-GlcN, 12 h

G-GlcN, 24 h

3-DG-GlcN, 24 h

G-GlcN, 48 h

200

400

600

400

600

MGO-GlcN, 4h MGO-GlcN, 8 h MGO-GlcN, 12h MGO-GlcN, 24 h MGO-GlcN, 48 h

0

0 200

d

0.5

GO-GlcN, 4 h GO-GlcN, 8 h GO-GlcN, 12 h GO-GlcN, 24 h GO-GlcN, 48 h

3-DG-GcN, 48 h

0

0

c

0.5

200

400

600

200

400

600

Wavelength (nm)

B

i

Fluorescence emission intensity (A.U.)

250

ii G-GlcN, 4 h

250

3-DG-GlcN, 4 h

G-GlcN, 8 h

200

G-GlcN, 12 h

3-DG-GlcN, 8 h

200

G-GlcN, 48 h

GO-GlcN, 4 h GO-GlcN, 8 h GO-GlcN, 12 h GO-GlcN, 24 h GO-GlcN, 48 h

200

3-DG-GlcN, 12 h

G-GlcN, 24 h

150

iii 250

3-DG-GlcN, 24 h

150

150

3-DG-GcN, 48 h

iv 250 200 150

100

100

100

100

50

50

50

50

0

0

0 364

452

540

364

452

540

MGO-GlcN, 4 h MGO-GlcN, 8 h MGO-GlcN, 12 h MGO-GlcN, 12 h MGO-GlcN, 48 h

0 364

452

540

364

452

540

Wavelength (nm)


Page 41 of 41


99x47mm (300 x 300 DPI)

ACS Paragon Plus Environment

Studies on the Formation of Maillard and Caramelization Products

Recommend Documents