Lysine-Derived Protein-Bound Heyns Compounds in Bakery Products

Article abstract of DOI:10.1021/acs.jafc.7b04172

Fructose and dicarbonyl compounds resulting from fructose in heated foods have been linked to pathophysiological pathways of several metabolic disorders. Up to now, very little has been known about the Maillard reaction of fructose in food. Heyns rearrang

Full text of DOI:10.1021/acs.jafc.7b04172

Subscriber access provided by READING UNIV
Article
Lysine-Derived Protein-bound Heyns Compounds in Bakery Products
Stephanie Treibmann, Anne Hellwig, Michael Hellwig, and Thomas Henle
J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b04172 • Publication Date (Web): 07 Nov 2017
Downloaded from http://pubs.acs.org on November 8, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Agricultural and Food Chemistry is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 32
Journal of Agricultural and Food Chemistry
Lysine-Derived Protein-bound Heyns Compounds in
Bakery Products
Stephanie Treibmann, Anne Hellwig, Michael Hellwig, Thomas Henle
Chair of Food Chemistry, Technische Universität Dresden, Dꢀ01062 Dresden, Germany
Corresponding author:
T. Henle
Tel.: +49ꢀ351ꢀ463ꢀ34647
Fax: +49ꢀ351ꢀ463ꢀ34138
Email: Thomas.Henle@chemie.tuꢀdresden.de
1
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 2 of 32
1
Abstract
2
3
Fructose and dicarbonyl compounds resulting from fructose in heated foods have been linked
to pathophysiological pathways of several metabolic disorders. Up to now, very little is
known about the Maillard reaction of fructose in food. Heyns rearrangement compounds
(HRCs), the first stable intermediates of the Maillard reaction between amino components and
fructose, have not yet been quantitated as proteinꢀbound products in food. Therefore, the
HRCs glucosyllysine and mannosyllysine were synthesized and characterized by NMR.
Proteinꢀbound HRCs in cookies containing various sugars and in commercial bakery products
were quantitated after enzymatic hydrolysis by RPꢀHPLCꢀESIꢀMS/MS in the multiple
reaction monitoring mode through application of the standard addition method. Proteinꢀbound
HRCs were quantitated for the first time in model cookies and in commercial bakery products
containing honey, banana and invert sugar syrup. Concentrations of HRCs from 19 to
287 mg/kg were found, which were similar to or exceeded the content of other frequently
analyzed Maillard reaction products, such as Nꢀεꢀcarboxymethyllysine (10–76 mg/kg), Nꢀεꢀ
carboxyethyllysine (2.5–53 mg/kg) and methylglyoxalꢀderived hydroimidazolone 1 (10–
218 mg/kg) in the analyzed cookies. These results show that substantial amounts of HRCs
form during food processing. Analysis of proteinꢀbound HRCs in cookies is therefore useful
to evaluate the Maillard reaction of fructose.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Keywords
20
21
Maillard reaction; glycation; fructose; Heyns compound; bakery product; HPLCꢀMS/MS
2
ACS Paragon Plus Environment

Page 3 of 32
Journal of Agricultural and Food Chemistry
22 Introduction
23
During storage and thermal processing of food, reducing sugars and amino components may
react to ultimately form aroma and color compounds due to the Maillard reaction. Aldoses,
such as glucose, react with amino groups to Amadori rearrangement compounds (ARCs) as
the first stable intermediates. The formation of ARCs is well described, and quantitation of
lysine derivatization is widely used to estimate the extent of early stage Maillard reactions in
food and in physiological systems.1 By contrast, the reaction of amino components and
ketoses in food has hardly been investigated. Analogous to ARCs, the epimeric Heyns
rearrangement compounds (HRCs, figure 1) Nꢀεꢀglucosyllysine and Nꢀεꢀmannosyllysine are
the first stable intermediates of the Maillard reaction of fructose and lysine.2 Both HRCs and
ARCs degrade to 1,2ꢀdicarbonylꢀcompounds, which react with amino components under
formation of advanced glycation endproducts (AGEs). Nꢀεꢀcarboxymethyllysine (CML) is a
major AGE on lysine side chains.3 It is formed via oxidative degradation of ARCs and
proposedly of HRCs4 and during the reaction of lysine and glyoxal.5 The reaction of lysine
and methylglyoxal leads to the formation of Nꢀεꢀcarboxyethyllysine (CEL). A typical AGE on
arginine side chains is methylglyoxalꢀderived hydroimidazolone 1 (MGꢀH1) which originates
from the reaction of arginine and methylglyoxal.6
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Heyns was the first to describe the synthesis of the products named after him.2 Free HRCs
were found in stored lyophilized apricots,7 tobacco,8 raw licorice,9 tomato powder10 and
garlic11 via paper chromatography,7,8 thin layer chromatography,9 cation exchange
chromatography with refractive index detection10 and HPLCꢀMS/MS.11 Reduced proteinꢀ
bound HRCs were found in human ocular lens protein after acid reduction and hydrolysis
using RPꢀHPLC.12 By adding Nꢀαꢀhippuryllysine to a cookie dough, it was shown that
peptideꢀbound HRCs form under baking conditions.13 However, the direct analysis of proteinꢀ
bound HRCs in food has yet to be described. Since the use of high fructose corn syrup
3
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 4 of 32
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
(HFCS) has replaced sucrose in the United States by almost 50% in the last 40 years, free
fructose has become far more available in foods.14 Thus, the potential of fructoseꢀinduced
Maillard reaction in food has increased. To evaluate the extent of the early stage of the
Maillard reaction in foods containing honey, HFCS and fruits, analysis of proteinꢀbound
HRCs is necessary.
The essential amino acid lysine is not available for nutrition, when bound in ARCs and AGEs,
however, the human colonic microbiota is able to degrade fructosyllysine, CML and
pyrraline.15 Dietary AGEs are discussed to have adverse metabolic effects through their
accumulation in human tissue and the postulated interaction with AGEꢀreceptors, which leads
to inflammation.16 Thus, dietary HRCs and other fructose derived MRP could have
pathophysiological effects due to the loss of lysine and because of the inflammatory effects of
dietary AGEs.
It was therefore the purpose of this study to obtain information about the role of fructose in
glycation, more precisely to quantitate proteinꢀbound HRCs in food.
62 Materials and methods
63
64
65
66
67
68
69
70
71
Chemicals. Ethanol and HPLC gradient grade acetonitrile were purchased from VWR
Prolabo (Leuven, Belgium). Nonafluoropentanoic acid (NFPA), N,Nꢀdimethylformamide
(DMF, <0.01% water), deuterium oxide, hydrochloric acid, ammonia solution (25%),
prolidase (138 U/mg protein), pepsin (3555 U/mg protein), and leucine aminopeptidase
(18 U/mg protein) from SigmaꢀAldrich were employed (Steinheim, Germany) and acetic acid,
fructose, glucose, and sucrose from Roth (Karlsruhe, Germany). Nꢀαꢀtertꢀ
butyloxycarbonyllysine (Bocꢀlysine), petroleum ether (boiling point range: 40–60 °C), and
tris(hydroxymethyl)aminomethane (TRIS) were obtained from Fluka (St Gallen, Switzerland)
and triphenyl tetrazolium chloride from Kallies (Sebnitz, Germany). Pronase E (4000 PU/mg
4
ACS Paragon Plus Environment

Page 5 of 32
Journal of Agricultural and Food Chemistry
72
protein) from Merck (Darmstadt, Germany) was used and pyridine from Acros (Geel,
Belgium). Formic acid and sodium hydroxide were purchased from Grüssing (Filsum,
Germany) and ninhydrin from Serva (Heidelberg, Germany). The water used for the
preparation of solutions and buffers was prepared by a Purelab Plus purification system (USF
Elga, RansbachꢀBaumbach, Germany).Water for HPLCꢀMS/MS was obtained by a Bi 18 E
double distillation system (QCS, Maintal, Germany). Fructosyllysine,17 CML18, CEL18 and
MGꢀH118 (Figure 1) were synthesized and characterized in our laboratory according to the
literature stated. The substances met the spectroscopic and chromatographic characteristics
published in the respective protocols.
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
Synthesis and Isolation of Nꢀε-(2-Deoxy-
D-glucosyl)-L-lysine (Glucosyllysine) and Nꢀε-(2-
Deoxy-D-mannosyl)-L
-lysine (Mannosyllysine). The compounds were synthesized2,13 and
isolated17,19 according to literature methods with minor modifications. Bocꢀlysine (0.62 g,
2.5 mmol) and anhydrous fructose (4.5 g, 14 mmol) were refluxed in 37 mL of dry
dimethylformamide for one hour. After evaporation of the solvent, the residue was dissolved
in 100 mL of water. Black precipitate was removed by filtration. The pH value was adjusted
to 2.0 with 6 M hydrochloric acid. For removal of the Boc protecting group and uncharged
byꢀproducts, the solution was applied to a 200x25 mm glass column filled with 80 mL of
cation exchange resin Lewatit S 100 (20–50 mesh) in the H+ form, previously equilibrated
with 250 mL of 6 M hydrochloric acid and 250 mL of water. The sugar was eluted with
250 mL of water and after one night of incubation, material bound to the exchanger was
eluted with 250 mL of 2 M ammonia. The eluate was evaporated to dryness and dissolved in
20 mL of 0.1 M pyridinium formate buffer, pH 3.0. The pH value was adjusted to 3.0 with
formic acid. The solution was applied to a 750x15 mm semipreparative ionꢀexchange column
filled with DOWEX 50 WꢀX8 (100ꢀ200 mesh) in the pyridinium form, previously
5
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 6 of 32
97
98
equilibrated with 360 mL of 6 M hydrochloric acid, 360 mL of water, 360 mL of 2 M aqueous
pyridine, 360 mL of water and 360 mL of 0.1 M pyridinium formate buffer (pH 3.00).
Separation was achieved with elution by gravity of 360 mL of 0.25 M pyridinium acetate
buffer (pH 4.05) and 360 mL of 0.3 M pyridinium acetate buffer (pH 4.05) at a flow rate of
0.4 mL/min. A fraction collector (RediFrac, Pharmacia Biotech, Sweden) was used and
fractions of 10 mL were collected. Of each fraction, 1 µL was spotted on two TLC plates.
Spraying with 1% triphenyl tetrazolium chloride in 1 M sodium hydroxide and 0.1%
ninhydrin in ethanol, respectively, indicated that the HRCs eluted with 70–220 mL of the last
buffer. An aliquot of those fractions (5 µL) was diluted in 495 µL of the loading buffer and
subjected to amino acid analysis. Fractions eluting with 70–190 mL of 0.3 M pyridinium
acetate buffer (pH 4.05) contained glucosyllysine, and mannosyllysine eluted with 130–
220 mL of this buffer. The fractions that contained only one HRC were pooled, dried in
vacuo, repeatedly dissolved in water and evaporated, lyophilized and characterized.
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Nꢀεꢀ(2ꢀdeoxyꢀDꢀglucosyl)ꢀLꢀlysine (glucosyllysine): HPLCꢀMS/MS: see Figure 2 A;
elemental analysis: C12H24N2O7 (MW = 308.32), calculated: C 46.75%, H 7.85%, N 9.09%;
found: C 39.59%, H 6.84%, N 7.75%; content = 76.3%, based on nitrogen and
chromatographic purity from LCꢀMS/MS (ratio of glucosyllysine to mannosyllysine and
fructosyllysine). Yield = 281 mg (molar yield = 27.8%). NMR data: see Table 1 and
supporting information Figure S1–S5.
Nꢀεꢀ(2ꢀdeoxyꢀDꢀmannosyl)ꢀLꢀlysine (mannosyllysine): HPLCꢀMS/MS: see Figure 2 B;
elemental analysis: C12H24N2O7 (MW = 308.32), calculated: C 46.75%, H 7.85%, N 9.09%;
found: C 41.19%, H 7.44%, N 6.91%; content = 64.5%, based on nitrogen and
chromatographic purity from LCꢀMS/MS (ratio of mannosyllysine to glucosyllysine and
fructosyllysine). Yield = 23 mg (molar yield = 2.1%). NMR data: see Table 2 and supporting
information Figure S6–S10.
6
ACS Paragon Plus Environment

Page 7 of 32
Journal of Agricultural and Food Chemistry
1
123
Characterization of Glucosyllysine and Mannosyllysine. H and 13C NMR spectra were
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
recorded on an Avance III HDX 500 MHz Ascend instrument from Bruker (Rheinstetten,
Germany) at 500.13 MHz and 125.75 MHz, respectively. Glucosyllysine (7.1 mg) and
mannosyllysine (5.4 mg) were each dissolved in 770 µL of deuterium oxide and stored at
room temperature overnight to allow mutarotation. All chemical shifts are given in parts per
million (ppm), those of protons relative to the internal HOD signal (4.70 ppm), those of
1
carbon atoms relative to external standard tetramethylsilane. Assignments of H and 13C
1
1
signals are based on Hꢀ1H correlation spectroscopy (COSY), Hꢀ13C heteronuclear single
1
quantum coherence (HSQC); Hꢀ13C heteronuclear multiple bond correlation (HMBC) and
13C distortionless enhancement by polarization transfer (DEPT) experiments. Elemental
analysis was performed on a vario MICRO cube CHNS elemental analyzer (Elementar,
Hanau, Germany).
Bakery Products. All bakery products, wheat flour, margarine, honey, bananas and baking
powder were purchased in local retail stores. Cookie doughs were prepared from 10.00 g of
wheat flour, 5.90 g of margarine, 0.30 g of baking powder, 7.5 mL of water and 5.00 g of
either fructose, glucose, or sucrose. For a honey cookie dough, 6.60 g of honey and 5.5 mL of
water and for a banana cookie dough, 10.00 g of banana and 3.20 g of sucrose were added to
wheat flour, margarine and baking powder. Cookies were formed manually to a thickness of
about 5 mm and a diameter of about 5 cm. The cookies were baked in a laboratory drying
oven for 10 min at 175 °C. After cooling, the cookies were homogenized by mortar and
pestle. A sample of 160 mg of a baked or purchased cookie was weighed into a 2 mL plastic
tube and defatted using 1 mL of petroleum ether as described previously.13 The sample was
mixed vigorously by vortex, centrifuged (10,600 x g, 10 min) and the petroleum ether was
discarded. After repeating this procedure, 500 µL of water was added to the precipitate. The
suspension was mixed by vortex and dialyzed against water at 4 °C for 2 days (cellulose
7
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 8 of 32
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
tubings, MWCO, 14 kDa, SigmaꢀAldrich). The water was changed twice per day. The
retentates were lyophilized.
Enzymatic Hydrolysis. Samples containing 2–3 mg of protein (8–40 mg of lyophilized
cookie, depending on protein content) were used according to the literature.20 Hydrochloric
acid (1.05 mL, 0.02 M) containing 356 U of pepsin was added. After 24 h incubation in a
drying chamber at 37 °C, 300 µL of 2 M TRIS buffer, pH 8.2 containing 400 PU of pronase E
was added. Following 24 h of further incubation, 22 µL of 2 M TRIS buffer containing 0.3 U
of leucineꢀaminopeptidase and 2 U of prolidase was added. After 48 h, the samples were
frozen. Centrifugation (10,600 x g, 10 min) and membrane filtration (0.45 µm) of thawed
samples were performed prior to HPLC analysis. An aliquot of the hydrolysate (150 µL) was
suspended in 450 µL of the loading buffer for amino acid analysis (0.12 N lithium citrate, pH
2.20). After membrane filtration (0.45 µm), samples were subjected to amino acid analysis.
Amino Acid Analysis. Amino acids separation was performed on a PEEK column filled with
the cation exchange resin LCA K07/Li (150 mm × 4.6 mm, 7 µm) using the amino acid
analyzer S 433 (Sykam, Fürstenfeldbruck, Germany) with a gradient program utilized
previously.21 Loading and running buffers for this lithium system were purchased from
Sykam. After postꢀcolumn derivatization with ninhydrin, the absorbance of the effluent was
recorded with a twoꢀchannel photometer simultaneously working at 440 nm and 570 nm,
respectively. A commercial amino acid mixture was obtained (SigmaꢀAldrich, Steinheim,
Germany) and used for external calibration.
High-Pressure Liquid Chromatography (HPLC) with Tandem Mass Spectrometric
Detection. Proteinꢀbound MRPs were analyzed as described before15,21,22 on a high pressure
gradient system 1200 Series (Agilent Technologies, Böblingen, Germany), consisting of a
8
ACS Paragon Plus Environment

Page 9 of 32
Journal of Agricultural and Food Chemistry
175
binary pump, an online degasser, an autosampler, a column oven, and a diode array detector.
Chromatographic separation was performed on a Zorbax SB Cꢀ18 column (50 mm x 2.1 mm,
3.5 µm, Agilent), with 10 mM NFPA in doubleꢀdistilled water (solvent A) and 10 mM NFPA
in acetonitrile (solvent B) as the mobile phases. The injection volume was 2 µL. Analysis was
conducted at 35 °C using a flow rate of 0.25 mL/min. A gradient of 5% B (0 min) to 32% B
(15 min), then to 85% B (16–19 min), and finally to 5% B (20 min) was used. An
equilibration time of 8 min before the next run was implemented. The HPLC was coupled to
the mass spectrometer 6410 Triple Quad (Agilent) working in the positive mode with a source
temperature of 350 °C and a capillary voltage of 4000 V. Nitrogen was used as the nebulizing
gas (nitrogen generator 5183ꢀ2003, Agilent) with a flow rate of 11 L/min and a pressure of
35 psi. The multipleꢀreaction monitoring mode (MRM) was used for mass spectrometric
detection with the conditions listed in table 3. These conditions were optimized by scan and
product ion scan measurements of the respective standards (Figure 2). Data acquisition and
evaluation was conducted with the software Mass Hunter B.02.00 (Agilent).
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
The MRPs were quantitated by the standard addition method with three runs. For each run,
100 µL of enzymatic hydrolysate was mixed either with 20 µL of water, or with 10 µL of
water and 10 µL of a standard solution, or with 20 µL of standard solution. The standard
solution contained Nꢀεꢀglucosyllysine (19.0 µg/mL), Nꢀεꢀfructosyllysine (70.1 µg/mL), CML
(5.0 µg/mL), CEL (5.0 µg/mL), and MGꢀH1 (15.0 µg/mL) dissolved in water. Limits of
detection (LOD) and quantitation (LOQ) for the compounds are shown in table 4. For a
matrix calibration, the standard solution was added to an enzymatic hydrolysate of water.
Concentration ranges of the calibration are shown in table 4.
Statistical Treatment. All samples were analyzed at least in duplicate. Baking experiments
were done one time. Limits of detection (LOD) and quantitation (LOQ) were calculated as the
concentrations of MRPs necessary to show peaks at signalꢀtoꢀnoise ratios of 3 and 10,
9
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 10 of 32
201
202
203
204
205
respectively. Coefficients of determination were calculated for standard addition regression
lines. The intercept accuracy was calculated as the percentage difference between the area of
the sample peak without addition of standards and the intercept of the regression line. The
linearity of the matrix calibration was determined with the Mandel’s fitting test for α = 0.05.
206 Results and discussion
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
Synthesis of Heyns Rearrangement Compounds (HRCs). The aim of this study was to
quantitate proteinꢀbound HRCs in bakery products containing fructose in order to evaluate the
early stage of the Maillard reaction with fructose. Therefore, the epimeric HRCs
glucosyllysine and mannosyllysine (Figure 1) needed to be synthesized and characterized as
reference material for chromatographic analysis. Protocols for synthesis of HRCs were
adapted2,13 with some modifications, whereas isolation and separation were performed as
described for fructosyllysine17 and formyline.19 The peaks of glucosyllysine and
mannosyllysine overlapped during semiꢀpreparative ionꢀexchange chromatography as
determined by amino acid analysis. Only fractions containing one product were pooled and
processed further. Oneꢀ and twoꢀdimensional NMR spectroscopy (Tables 1 and 2, supporting
information Figures S1–S10) confirmed the structural identity of the compounds. Residual
acetic acid from the chromatography eluent was identified as an impurity. Chemical shifts of
glucosyllysine were in accordance with those published in the literature.13,23,24 In pyranose
sugars, coupling constants J1,2 of two vicinal axial protons are higher than 7 Hz, while the
coupling constants of vicinal protons of which at least one is in equatorial position are smaller
than 5 Hz.25 Mannosyllysine had similar chemical shifts and coupling constants as
glucosyllysine. However, the pyranosyl form of glucosyllysine has an axial proton at C2 and
therefore the coupling constant of the protons at C2 and C3 is 10.5 Hz and that of the protons
at C2 and C1 of the βꢀpyranosyl moiety is 8.5 Hz. In mannosyllysine those coupling constants
10
ACS Paragon Plus Environment

Page 11 of 32
Journal of Agricultural and Food Chemistry
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
were 4.6 and 1.7 Hz, indicating a proton in equatorial position at C2. Based on this findings
and by comparison to similar structures,26 mannosyllysine and glucosyllysine were identified.
Glucosyllysine was found in αꢀ and βꢀpyranose isomers (ca. 75% and 25%). In addition to αꢀ
and βꢀpyranose isomers (ca. 46% and 25%), mannosyllysine also formed considerable
1
amounts of αꢀ and βꢀfuranose isomers (ca. 18% and 9%). The Hꢀ1 H NMR signal of the αꢀ
furanose form of mannosyllysine overlapped with the Hꢀ1 signal of emerging glucosyllysine
which may have been formed during the period of time between dissolution of the compound
and the NMR measurement.27 Therefore, the percentage of αꢀfuranose might be
overestimated.
The reaction mixture before ion exchange chromatography contained the HRCs
glucosyllysine and mannosyllysine, but also the ARC fructosyllysine. After isolation, 76% of
glucosyllysine, 3% of mannosyllysine and 6% of fructosyllysine were present in the
glucosyllysine isolate as deduced from amino acid analysis, HPLCꢀMS/MS and elemental
analysis. The mannosyllysine isolate contained 65% of mannosyllysine, 6% of glucosyllysine
and 6% of fructosyllysine. Though concentrations of mannosyllysine were higher than those
of glucosyllysine in the synthesis mixture after heating (molar yield glucosyllysine: ca. 30%,
mannosyllysine: ca. 50%), concentrations of mannosyllysine in the eluted fractions and the
yield of isolated mannosyllysine were more than 10 times smaller than those of glucosyllysine
(molar yield glucosyllysine: 27.8%, mannosyllysine: 2.1%). Sugars with lower shares of the
pyranose forms, such as mannosyllysine, show higher concentrations of the openꢀchain form
and are therefore more reactive via the Maillard reaction and the Lobry de BruynꢀAlberda van
Ekenstein transformation.28 Mannosyllysine, therefore, is less stable than glucosyllysine
during the isolation process. Yields and purity of both substances were high enough for
chromatographic analysis.
11
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 12 of 32
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
Determination of HRCs in Bakery Products. A method to quantitate the HRCs
glucosyllysine and mannosyllysine, as well as the ARC fructosyllysine and the AGEs CML,
CEL, and MGꢀH1, in bakery products was established using enzymatic hydrolysis and LCꢀ
MS/MS. Enzymatic hydrolysis has already been used for the quantitation of different MRPs
such as fructosyllysine, pyrraline and CML.21,29 The rate of release after enzymatic digestion
was at least 80% of the rate after acid hydrolysis for aliphatic amino acids in roasted peanuts
and dried pasta.29,30 In eye lens protein, the rate of release after enzymatic digestion compared
to acid hydrolysis was 52–91% for CML.31 The quantitative data presented in this study might
therefore be slightly underestimated. LCꢀMS/MS analysis using the standard addition method
of enzymatically hydrolyzed standard solutions of glucosyllysine and mannosyllysine showed
that mannosyllysine is prone to interconversion during enzymatic hydrolysis. Of
mannosyllysine, only 29 ± 8% (SD, n=3) could be recovered unchanged. The larger part of
63 ± 7% had been transformed to glucosyllysine and 9 ± 1% were missing. On the other hand
glucosyllysine is stable with a recovery of 88 ± 3%. The missing 12 % of glucosyllysine had
been transformed to mannosyllysine. Traces of fructosyllysine were detected, when the
glucosyllysine standard was subjected to enzymatic hydrolysis, but CML was not detectable
in enzymatic hydrolysates of both HRCs. Thus, the HRCs isomerize during enzymatic
hydrolysis, with glucosyllysine as the favored product. As explained above, mannosyllysine is
more prone to Lobry de BruynꢀAlberda van Ekenstein transformation, leading to the
formation of the more stable glucosyllysine. In the present study, mannosyllysine was only
found in traces in bakery products, either due to isomerization to glucosyllysine or due to its
absence in the products. Nonetheless, as the quantification of glucosyllysine probably
encloses mannosyllysine present in the food product, the results will be given as the sum of
both compounds and designated glucosyl/mannosyllysine (LysGlc/Man) this study. Because
only trace amounts of mannosyllysine were found in bakery products the 9% loss of
mannosyllysine during enzymatic hydrolysis can also be assumed to be negligible.
12
ACS Paragon Plus Environment

Page 13 of 32
Journal of Agricultural and Food Chemistry
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
LysGlc/Man analysis was performed via an HPLCꢀMS/MS method used previously for ARCs
and AGEs.15,21,22 Product ion spectra of cookie samples were compared to those of standard
solutions (Figure 2 A, D). Transitions that were both selective and abundant were chosen for
quantitation (Table 3). With this HPLCꢀMS/MS method, simultaneous quantitation of the two
lysine HRCs and the ARC fructosyllysine was enabled (Figure 2 E, F). Fructosyllysine and
glucosyllysine also had minor peaks at the transitions of the other compound at their peak
maxima, but were separated via their retention time (Figure 2 A, C). This led to ca. 15 µg
falsely detected glucosyllysine per mg fructosyllysine and ca. 70 µg falsely detected
fructosyllysine per mg glucosyllysine. Consequently, for the ratios glucosyllysine to
fructosyllysine found in this study, up to 12% of the quantitated amounts might be
overestimated. In addition to HRCs and fructosyllysine, the AGEs CML and CEL as major
lysine derivatives and MGꢀH1 as a major arginine derivative6 were quantitated. The standard
addition method was used because stable isotope labeled standards were not available for
HRCs. Linearity for the concentration ranges of the standard additions was verified with
matrix calibrations of the analytes (concentrations see Table 4). Coefficients of determination
of the standard addition regression lines were r2 > 0.99 for 96% of the samples and r2 > 0.98
for all samples. The intercept accuracy was between 85 and 115% for 96% of all samples and
between 75 and 125% for all samples. Such intercept accuracy may be caused by overlapping
peaks of the fructosyllysine and glucosyllysine transitions. LODs and LOQs were sufficiently
low to enable quantitation of all analytes in most samples (Table 4).
The Maillard Reaction of Fructose in Cookies Baked with Fructose, Glucose, Honey and
Banana. In order to simulate the formation of ARC and HRC under baking conditions, model
cookies containing either fructose or glucose were baked. In both cookies, proteinꢀbound
LysGlc/Man and fructosyllysine were quantitated (Table 5). Of total lysine, 6.8% were
modified to fructosyllysine in glucoseꢀcontaining cookies, while only 1.6% of total lysine
13
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 14 of 32
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
were modified to LysGlc/Man in fructoseꢀcontaining cookies. In cookies baked in the
presence of hippuryllysine, 33% of hippuryllysine were derivatized to hippurylꢀglucosyllysine
in a fructoseꢀcontaining cookie and 63% of hippuryllysine were derivatized to hippurylꢀ
fructosyllysine in a cookie containing glucose.13 While the ratio between ARCs and HRCs is
similar, the yields are 5 times higher in the hippuryllysine cookies. When hippuryllysine is
used as a model amine, all lysine amino groups can be glycated, whereas in protein, some
lysine side chains are not or less available for derivatization with fructose, causing lower
yields of HRCs.4 In raw cookie dough and in sucroseꢀcontaining cookies, fructosyllysine was
found in traces and LysGlc/Man was not detectable or found only in traces. Thus, proteinꢀ
bound HRC form under baking conditions in cookies from fructose in significant
concentrations.
Fructosyllysine was also found in fructoseꢀcontaining cookies and LysGlc/Man in cookies
containing glucose with yields of 0.8% and 0.2%, respectively. The formation of ARCs in the
presence of fructose and of HRCs in the presence of glucose was described before and was
explained via the addition of a second amino moiety to the product followed by elimination of
the first.27,32 These subsequent Amadori and Heyns rearrangement reactions were also
observed, when ARCs and HRCs were incubated with additional amino compounds.27 On the
other hand, the isomerization of glucose to fructose and vice versa through the Lobry de
BruynꢀAlberda van Ekenstein transformation could lead to this observation (Figure 3). This
transformation is described to take place not only in basic solutions, but also in neutral
aqueous glucose solutions33,34 and cookies.35 Fructose concentrations in heated glucose
solutions rose up to 38% of the initial glucose concentrations.36 This means that HRCs also
form from glucose and that they play a minor role in the Maillard reaction with glucose as
well. Additionally, the extent of Maillard reaction with fructose can be overestimated through
this pathway.
14
ACS Paragon Plus Environment

Page 15 of 32
Journal of Agricultural and Food Chemistry
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
To obtain information about the formation of HRCs from fructose in the presence of glucose
and to simulate realistic baking conditions, additional cookies containing honey and banana
were baked under the same conditions and the same water content as the cookies with
fructose and glucose. While in honey cookies the sugar content of the fructose and glucose
cookie was completely replaced by honey saccharides, in banana cookies only 36% of the
sugar was replaced by banana saccharides, since otherwise the dough would have been too
fluid. The remaining sugar was replaced by sucrose. In honey and banana, fructose and
glucose occur nearly in the ratio 1:1. Differences between reactivity of glucose and fructose
were also found when the molar conversion of the parent saccharides was compared: Per mol
of added fructose, the fructose cookie contained 0.21 mmol LysGlc/Man, while the honey
cookie had 0.15 mmol and the banana cookie had 0.36 mmol LysGlc/Man. In comparison to
that, the glucose cookie contained 0.83 mmol fructosyllysine per mol glucose, the honey
cookie 1.36 mmol and the banana cookie 1.42 mmol fructosyllysine per mol glucose.
Apparently, fructose has a lower potential to form early stage MRPs. Additionally, the HRC
formation from fructose occurs in systems containing both fructose and glucose like in
cookies with honey or banana and is similar to the HRC formation in systems without
glucose.
The AGEs CML, CEL and MGꢀH1 as important lysine and arginine AGEs were analyzed in
all cookies to study differences in the advanced Maillard reaction with fructose and glucose.
The CML contents were similar in the cookies containing fructose, glucose, honey and
banana (Table 5). This indicates that glucose and fructose do not differ significantly in the
formation of CML under baking conditions. An explanation might be that fructosyllysine and
glucosyllysine similarly degrade to CML.4 By contrast, the highest MGꢀH1 and CEL
concentrations were found in the cookies with the highest fructose content. In aqueous sugarꢀ
lysine model systems at above 100 °C, the formation of methylglyoxal from fructose was up
to three times higher than the formation from glucose.37,38 Higher concentrations of
15
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 16 of 32
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
methylglyoxal explain the increased levels of the methylglyoxalꢀarginine adduct MGꢀH1 and
the methylglyoxalꢀlysine adduct CEL. Preliminary experiments have shown that the
percentage of arginine modified to MGꢀH1 and lysine modified to CEL correlates with the
concentrations of fructose or HRCs (data not shown). We are currently working on a further
exploration of the formation of MGꢀH1 and CEL from fructose which could be an indicator
for the “fructose pathway” of the Maillard reaction. The total loss of detectable lysine was ca.
15% in the glucoseꢀcontaining cookie and ca. 60% in the fructoseꢀcontaining cookie. While
the formation of fructosyllysine, LysGlc/Man, CML, and CEL in the glucose cookie accounts
for ca. 48% of the total lysine loss, the formation of the same products accounts for not more
than 4% of the total lysine loss in the fructose cookie. Therefore other derivatization or
fragmentation pathways have to be of major importance in the Maillard reaction with
fructose. Additionally, the nutritional loss of the essential amino acid lysine is much higher in
the fructose cookies than in the glucose cookies.
The Maillard Reaction of Fructose in Commercial Bakery Products. To evaluate HRC
formation and the degree of fructoseꢀinduced Maillard reaction in commonly consumed food,
commercial bakery products containing fructose were analyzed. The different cookies, cakes
and gingerbreads contained honey, invert sugar syrup and fruits as sweeteners. Proteinꢀbound
HRCs could be quantitated in all fructose containing bakery products in the range of 19 to
287 mg/kg. As expected, the lowest concentration of HRCs was found in the banana bakery
products which had the lowest amount of fructose (Table 5). The highest HRC level was
found in a cookie containing invert sugar syrup and honey. The amounts measured lead to a
consumption of 0.6 to 9 mg LysGlc/Man per 30 mg serving of the bakery product. This equals
16–90% of the ARC consumption with those bakery products.
Fructosyllysine, CML, CEL, and MGꢀH1 were also analyzed in the commercial bakery
products. Contents of fructosyllysine (1500–6400 mg/kg protein), CML, CEL and MGꢀH1
16
ACS Paragon Plus Environment

Page 17 of 32
Journal of Agricultural and Food Chemistry
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
were in similar ranges as stated in the literature for bakery products (fructosyllysine in bakery
products: 400–13000 mg/kg protein, calculated from furosine;17,39 CML in biscuits: 10–
26 mg/kg; CEL in biscuits: 4.9–34; MGꢀH1 in biscuits: 40–369 mg/kg).40 Contents of HRC
exceeded the contents of CML and CEL in all and those of MGꢀH1 in most analyzed cookies.
Therefore, substances of the early stage of Maillard reaction are of major importance in
cookies, and ARCs and HRCs are useful indicators for Maillard reaction with glucose and
fructose, respectively. To evaluate the relevance of the Maillard reaction with fructose in
other fructose containing foods and in pathophysiological pathways, further analysis of HRCs
is important. Furthermore, indicators to differentiate the advanced stage of the Maillard
reaction with fructose and glucose need to be found.
Fructose, either of dietary origin or formed in the polyol pathway, is discussed as a causative
factor for diabetic complications, nonalcoholic fatty liver disease and the metabolic
syndrome.41,42 Because the formation of HRCs under physiological conditions could be
observed,4,12 the Maillard reaction with fructose and fructose metabolites is discussed to play
a role in the pathophysiological pathways of the diseases mentioned above.41,43 The
established protocol for analysis of HRCs adapted for pathophysiological material could bring
insight into the role of the fructose pathway of the Maillard reaction in those diseases.
With the established method and the synthesized HRCs, protein bound HRCs could be
quantitated. Analysis of proteinꢀbound HRCs was used to evaluate the early stage of the
Maillard reaction with fructose in bakery products. We could show that HRCs form under
baking conditions not only in cookies baked with fructose but also in cookies with glucose or
both sugars and in commercial bakery products containing honey, banana and invert sugar
syrup. Proteinꢀbound HRCs were quantitated in similar or higher levels (19–287 mg/kg) than
the AGEs CML (10–76 mg/kg), CEL (2.5–53 mg/kg) and MGꢀH1 (10–218 mg/kg). Analysis
17
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 18 of 32
405
406
407
of proteinꢀbound HRCs is therefore suitable to evaluate the Maillard reaction of fructose in
cookies.
408 Supporting information
1
409
410
411
412
413
414
415
1D/2D NMR spectra of glucosyllysine and mannosyllysine. Figure S1: H NMR spectrum,
Figure S2: 13C NMR spectrum, Figure S3: COSY correlations, Figure S4: HSQC correlations,
1
Figure S5: HMBC correlations of glucosyllysine. Figure S6: H NMR spectrum, Figure S7:
13C NMR spectrum, Figure S8: COSY correlations, Figure S9: HSQC correlations, Figure
S10: HMBC correlations of mannosyllysine. This material is available free of charge via the
Internet at http://pubs.acs.org.
416 Abbreviations used
417
418
419
420
421
422
423
424
AGE, advanced glycation end product; ARC, Amadori rearrangement compound; CEL, Nꢀεꢀ
carboxyethyllysine; CML, Nꢀεꢀcarboxymethyllysine; DMF, N,Nꢀdimethylformamide; HFCS,
high fructose corn syrup; HPLC, high pressure liquid chromatography; HRC, Heyns
rearrangement compound; LOD, limit of detection; LOQ, limit of quantitation; LysGlc/Man,
glucosyl/mannosyllysine; MGꢀH1, methylglyoxalꢀderived hydroimidazolone, MRM, multiple
reaction monitoring; MRP, Maillard reaction product; MS, mass spectrometry; NFPA,
nonafluoropentanoic acid; RP, reversedꢀphase; TRIS, tris(hydroxymethyl)aminomethane
425 Acknowledgments
426
427
428
429
We are grateful to the members of the Chair of Inorganic Molecular Chemistry (Prof. J. J.
Weigand), namely Dr. Kai Schwedtmann, for recording the NMR spectra and Phillip Lange
for the elemental analysis. We thank Karla Schlosser, Chair of Food Chemistry, for
performing the amino acid analysis.
18
ACS Paragon Plus Environment

Page 19 of 32
Journal of Agricultural and Food Chemistry
430
431 Notes
432
433
The authors declare no competing financial interest.
S.T. thanks the Studienstiftung des deutschen Volkes for financial support.
434
19
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 20 of 32
435 References
436
437
(1)
(2)
Hellwig, M.; Henle, T. Baking, ageing, diabetes: a short history of the Maillard
reaction. Angew. Chemie ꢀ Int. Ed. 2014, 53, 10316–10329.
438
439
440
Heyns, K.; Noack, H. Die Umsetzung von DꢀFructose mit LꢀLysin und LꢀArginin und
deren Beziehung zu nichtenzymatischen Bräunungsreaktionen. Chem. Ber. 1962, 95,
720–727.
441
442
443
444
(3)
Weenen, H.; Tjan, S. B.; Devalois, P. J.; Bouter, N.; Pos, A.; Vonk, H. Mechanism of
pyrazine formation. In Thermally generated flavors ꢀ Maillard, microwave, and
extrusion processes; Parliament, T. H., Morello, M. J., McGorrin, R. J., Eds.; American
Chemical Society, Washington, USA, 1994; Vol. 543, 142–157.
445
446
(4)
(5)
Hinton, D. J. S.; Ames, J. M. Site specificity of glycation and carboxymethylation of
bovine serum albumin by fructose. Amino Acids 2006, 30, 425–433.
447
448
449
Glomb, M. A.; Monnier, V. M. Mechanism of protein modification by glyoxal and
glycolaldehyde, reactive intermediates of the Maillard reaction. J. Biol. Chem. 1995,
270, 10017–10026.
450
451
452
(6)
(7)
Henle, T.; Walter, A. W.; Haeßner, R.; Klostermeyer, H. Detection and identification
of a proteinꢀbound imidazolone resulting from the reaction of arginine residues and
methylglyoxal. Z. Lebensm. Unters. Forsch. 1994, 199, 55–58.
453
454
Anet, E. F. L. J.; Reynolds, T. M. Reactions between aminoꢀacids, organic acids and
sugars in freezeꢀdried apricots. Nature 1956, 177, 1082–1083.
455
(8)
(9)
Wahl, R. Aminozucker im Tabak. Z. Lebensm. Unters. Forsch. 1972, 148, 94–96.
456
457
Vondenhof, T.; Glombitza, K.ꢀW.; Steiner, M. Lösliche Kohlenhydrate in Rohlakritzen
(Succus liquiritiae). Z Leb. Unters Forsch 1973, 152, 345–347.
458
459
460
461
(10) Van den Ouweland, G. A. M.; Peer, H. G.; Tjan, S. B. Occurrence of Amadori and
Heyns rearrangement products in processed foods and their role in flavor formation. In
Flavor of foods and beverages; Charalambous, G., Inglett, G. E., Eds.; Academic
Press: New York, 1978, 131–143.
462
463
464
(11) Yuan, H.; Sun, L.; Chen, M.; Wang, J. The comparison of the contents of sugar,
Amadori, and Heyns compounds in fresh and black garlic. J. Food Sci. 2016, 81,
C1662–C1668.
465
466
(12) McPherson, J. D.; Shilton, B. H.; Walton, D. J. Role of fructose in glycation and crossꢀ
linking of proteins. Biochemistry 1988, 27, 1901–1907.
467
468
(13) Krause, R.; Schlegel, K.; Schwarzer, E.; Henle, T. Formation of peptideꢀbound Heyns
compounds. J. Agric. Food Chem. 2008, 56, 2522–2527.
469
470
(14) Marriott, B. P.; Cole, N.; Lee, E. National estimates of dietary fructose intake increased
from 1977 to 2004 in the united states. J. Nutr. 2009, 139, 1228S–1235S.
471
472
473
(15) Hellwig, M.; Bunzel, D.; Huch, M.; Franz, C. M. A. P.; Kulling, S. E.; Henle, T.
Stability of individual Maillard reaction products in the presence of the human colonic
microbiota. J. Agric. Food Chem. 2015, 63, 6723–6730.
20
ACS Paragon Plus Environment

Page 21 of 32
Journal of Agricultural and Food Chemistry
474
475
(16) Kellow, N. J.; Coughlan, M. T. Effect of dietꢀderived advanced glycation end products
on inflammation. Nutr. Rev. 2015, 73 (11), 737–759.
476
477
478
(17) Krause, R.; Knoll, K.; Henle, T. Studies on the formation of furosine and pyridosine
during acid hydrolysis of different Amadori products of lysine. Eur. Food Res.
Technol. 2003, 216, 277–283.
479
480
481
482
(18) Hellwig, M.; Geissler, S.; Matthes, R.; Peto, A.; Silow, C.; Brandsch, M.; Henle, T.
Transport of free and peptideꢀbound glycated amino acids: synthesis, transepithelial
flux at CACOꢀ2 cell monolayers, and interaction with apical membrane transport
proteins. ChemBioChem 2011, 12, 1270–1279.
483
484
(19) Hellwig, M.; Henle, T. Formyline, a new glycation compound from the reaction of
lysine and 3ꢀdeoxypentosone. Eur. Food Res. Technol. 2010, 230, 903–914.
485
486
487
(20) Hellwig, M.; Henle, T. Quantification of the Maillard reaction product 6ꢀ(2ꢀformylꢀ1ꢀ
pyrrolyl)ꢀLꢀnorleucine (formyline) in food. Eur. Food Res. Technol. 2012, 235, 99–
106.
488
489
490
(21) Hegele, J.; Buetler, T.; Delatour, T. Comparative LCꢀMS/MS profiling of free and
proteinꢀbound early and advanced glycationꢀinduced lysine modifications in dairy
products. Anal. Chim. Acta 2008, 617, 85–96.
491
492
493
(22) Hellwig, M.; Witte, S.; Henle, T. Free and proteinꢀbound Maillard reaction products in
beer: method development and a survey of different beer types. J. Agric. Food Chem.
2016, 64, 7234–7243.
494
495
496
(23) Srinivas, S. M.; Harohally, N. V. Improved synthesis of lysineꢀ and arginineꢀderived
Amadori and Heyns products and in vitro measurement of their angiotensin Iꢀ
converting enzyme inhibitory activity. J. Agric. Food Chem. 2012, 60, 1522–1527.
497
498
499
(24) Jakas, A.; Katić, A.; Bionda, N.; Horvat, Š. Glycation of a lysineꢀcontaining
tetrapeptide by Dꢀglucose and Dꢀfructoseꢀinfluence of different reaction conditions on
the formation of Amadori/Heyns products. Carbohydr. Res. 2008, 343, 2475–2480.
500
501
502
(25) Bubb, W. A. NMR spectroscopy in the study of carbohydrates: characterizing the
structural complexity. Concepts Magn. Reson. Part A Bridg. Educ. Res. 2003, 19, 1–
19.
503
504
505
(26) Mossine, V. V.; Barnes, C. L.; Glinsky, G. V.; Feather, M. S. Molecular and crystal
structure of Nꢀ(2ꢀdeoxyꢀDꢀaldohexosꢀ2ꢀyl)ꢀglycines (Heyns compounds). Carbohydr.
Res. 1996, 284, 11–24.
506
507
(27) Heyns, K.; Müller, G.; Paulsen, H. Quantitative Untersuchungen der Reaktion von
Hexosen mit Aminosäuren. Justus Liebigs Ann. Chem. 1967, 703, 202–214.
508
509
510
(28) Douglas Hayward, L.; J. Angyal, S. A symmetry rule for the circular dichroism of
reducing sugars, and the proportion of carbonyl forms in aqueous solutions thereof.
Carbohydr. Res. 1977, 53, 13–20.
511
512
(29) Wellner, A.; Nußpickel, L.; Henle, T. Glycation compounds in peanuts. Eur. Food Res.
Technol. 2012, 234, 423–429.
513
514
(30) Resmini, P.; Pellegrino, L. Occurrence of proteinꢀbound lysylpyrrolaldehyde in dried
pasta. Cereal Chem. 1994, 71, 254–262.
21
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 22 of 32
515
516
517
(31) Smuda, M.; Henning, C.; Raghavan, C. T.; Johar, K.; Vasavada, A. R.; Nagaraj, R. H.;
Glomb, M. A. Comprehensive analysis of Maillard protein modifications in human
lenses: effect of age and cataract. Biochemistry 2015, 54, 2500–2507.
518
519
520
521
(32) Rewicki, D.; Kersten, E.; Helak, B.; Nittka, C.; Tressl, R. Mechanistic studies on the
formation of Maillard products vom [1ꢀ13C]ꢀDꢀfructose. In Maillard Reactions in
Chemistry, Food and Health; Labuza, T. P., Reineccius, G. A., Monnier, V., O’Brien,
J., Baynes, J., Eds.; Royal Society of Chemistry: Cambridge, 1994; 61–68.
522
523
524
525
(33) Kocadağlı, T.; Gökmen, V. Effect of sodium chloride on αꢀdicarbonyl compound and
5ꢀhydroxymethylꢀ2ꢀfurfural formations from glucose under caramelization conditions:
a multiresponse kinetic modeling approach. J. Agric. Food Chem. 2016, 64, 6333–
6342.
526
527
528
(34) Nguyen, H. T.; van der FelsꢀKlerx, H. J.; van Boekel, M. a J. S. Kinetics of N(ε)ꢀ
(carboxymethyl)lysine formation in aqueous model systems of sugars and casein. Food
Chem. 2016, 192, 125–133.
529
530
531
(35) Nguyen, H. T.; Van Der FelsꢀKlerx, H. J.; Peters, R. J. B.; Van Boekel, M. A. J. S.
Acrylamide and 5ꢀhydroxymethylfurfural formation during baking of biscuits: part I:
effects of sugar type. Food Chem. 2016, 192, 575–585.
532
533
534
(36) Liang, Z.; Li, L.; Qi, H.; Zhang, Z. X. X.; Li, B. Kinetic study on peptideꢀbound
pyrraline formation and elimination in the Maillard reaction using singleꢀ and multipleꢀ
response models. J. Food Sci. 2016, 81, C2405–C2424.
535
536
537
(37) Novotný, O.; Cejpek, K.; Velíšek, J. Formation of αꢀhydroxycarbonyl and αꢀdicarbonyl
compounds during degradation of monosaccharides. Czech J. Food Sci. 2007, 25, 119–
130.
538
539
540
(38) Chen, X.ꢀM.; Kitts, D. D. Identification and quantification of αꢀdicarbonyl compounds
produced in different sugarꢀamino acid Maillard reaction model systems. Food Res. Int.
2011, 44, 2775–2782.
541
542
(39) Henle, T.; Zehetner, G.; Klostermeyer, H. Fast and sensitive determination of furosine.
Z Leb. Unters Forsch 1995, 200, 235–237.
543
544
545
546
(40) Scheijen, J. L. J. M.; Clevers, E.; Engelen, L.; Dagnelie, P. C.; Brouns, F.; Stehouwer,
C. D. A.; Schalkwijk, C. G. Analysis of advanced glycation endproducts in selected
food items by ultraꢀperformance liquid chromatography tandem mass spectrometry:
presentation of a dietary age database. Food Chem. 2016, 190, 1145–1150.
547
548
549
(41) Schalkwijk, C. G.; Stehouwer, C. D. A.; van Hinsbergh, V. W. M. Fructoseꢀmediated
nonꢀenzymatic glycation: sweet coupling or bad modification. Diabetes. Metab. Res.
Rev. 2004, 20, 369–382.
550
551
(42) Wiernsperger, N.; Geloen, A.; Rapin, J.ꢀR. Fructose and cardiometabolic disorders: the
controversy will, and must, continue. Clinics 2010, 65, 729–738.
552
553
554
555
(43) Mastrocola, R.; Collino, M.; Rogazzo, M.; Medana, C.; Nigro, D.; Boccuzzi, G.;
Aragno, M. Advanced glycation end products promote hepatosteatosis by interfering
with SCAPꢀSREBP pathway in fructoseꢀdrinking mice. Am. J. Physiol. Gastrointest.
Liver Physiol. 2013, 305, G398–407.
556
22
ACS Paragon Plus Environment

Page 23 of 32
Journal of Agricultural and Food Chemistry
Figure captions
Figure 1. Structures of the Maillard reaction products analyzed in this study.
Figure 2. Product ion spectra of glucosyllysine (A), mannosyllysine (B), and fructosyllysine
(C) obtained during HPLCꢀMS/MS analysis of standard solutions, and of glucosyllysine in a
hydrolyzed cookie sample (D). Operating conditions were those for the transition m/z 309 ꢀ
273 (Table 3) For (D) 5 µL of NFPA were added to 100 µL of the sample for better separation
of glucosyllysine and fructosyllysine. Precursor (italic numbers), quantifier (bold numbers)
and qualifier ions are denoted. HPLCꢀMS/MS chromatograms of a mixture of glucosyllysine
and fructosyllysine (1) and a hydrolyzed cookie sample (2) with the transitions m/z 309 ꢀ
210 (E, glucosyllysine) and m/z 309 ꢀ 273 (F, fructosyllysine).
Figure 3. Reactions of glucose and fructose with amino components: 1 Lobry de Bruynꢀ
Alberda van Ekenstein transformation, 2 formation of Schiff base and Amadori rearrangement
compound, 3 formation of Schiff base and Heyns rearrangement compound, 4 formation of
Schiff base and subsequent Heyns and Amadori rearrangement compounds
23
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 24 of 32
Table 1.13C and 1H NMR Spectroscopic Data of Nεꢀ(2ꢀdeoxyꢀ
Dꢀglucosyl)ꢀLꢀlysine
C Atom
δC [ppm]
δH [ppm]
J(x,y) [Hz]
1Hꢀ1H
COSY
1Hꢀ13C
HMBC
(figure 1)
coupling coupling
glucose moiety
1
88.00[a]/93.26 5.46[a]/4.87[b]
[b] (CH)
d
3.6 [a]/8.5[b] (1,2)
2[a]
3; 5
2
59.97[a]/2.48[b 3.16[a]/2.75[b] dd
] (CH)
3.6[a]/8.5[b] (1,2);
10,6 (2,3)
1; 3[a]
1; 4; 12
3
4
5
6
69.99 (CH)
69.68 (CH)
71.26 (CH)
60.28 (CH2)
3.83 dd
3.39 dd
3.80 m
3.70 d
10.6 (2,3); 9.1 (3,4) 2; 4
2; 5
9.1 (3,4); 8.6 (4,5)
3
6
5
2; 3; 5; 6
3; 4; 6
5
ꢀ
5.4 (5,6)
lysyl moiety
7
8
9
174.55 (Ci)
ꢀ
ꢀ
ꢀ
ꢀ
54.46 (CH)
29.84 (CH2)
3.66 t
1.82 m
6.2
ꢀ
9
7; 9; 10
7; 8; 10;
11
8; 10
10
21.57 (CH2)
1.40 m
ꢀ
9; 11
8; 9; 11;
12
11
12
25.33 (CH2)
45.35 (CH2)
1.68 m
3.09 t
ꢀ
10; 12
11
9; 10; 12
2; 10; 11
7.5
Data of [a] αꢀpyranosylꢀconformation (75%) and [b] βꢀpyranosylꢀconformation (25%) of
glucosyllysine; percentage of conformation calculated from the peakꢀarea of the proton at C1
24
ACS Paragon Plus Environment

Page 25 of 32
Journal of Agricultural and Food Chemistry
Table 2.13C and 1H NMR Spectroscopic Data of Nεꢀ(2ꢀdeoxyꢀ
D
ꢀmannosyl)ꢀLꢀlysine
C Atom
δC [ppm]
δH [ppm]
J(x,y) [Hz]
1Hꢀ1H
1Hꢀ13C
(figure 1)
COSY
HMBC
coupling coupling
mannose moiety
1
89.58[a]/91.95[b]/ 5.36[a]/5.12[b]/5.47 1.7[a]/1.7[b]/5. 2[a]
3[a]; 5[a]
98.75[c]/
[c]/5.43[d] d
3[c]/5.5[d]
(1,2)
93.88[d]/ (CH)
61.09(CH)
67.38 (CH)
2[a]
3[a]
3.30 m
4.06 dd
1; 3
3
4
4.6 (2,3); 9.5 2; 4
(3,4)
4[a]
66.41 (CH)
3.53 dd
9.5 (3,4); 9.5 3; 5
(4,5)
3; 5; 6
5[a]
6[a]
71.90 (CH)
60.27 (CH2)
3.80 m
3.76 d
ꢀ
4; 6
5
4
3.9 (5,6)
4; 5
lysyl moiety
7
174.62 (Ci)
ꢀ
ꢀ
ꢀ
ꢀ
8
54.53 (CH)
29.94 (CH2)
21.69 (CH2)
26.37 (CH2)
47.21 (CH2)
3.68 t
1.82 m
1.40 m
1.68 m
2.95 t
6.2
ꢀ
9
7; 9; 10
7; 8; 10; 11
8; 9; 11; 12
9: 10: 12
10: 11
9
8; 10
9; 11
10; 12
11
10
11
12
ꢀ
ꢀ
7.7
Data of [a] αꢀpyranosylꢀconformation (46%), [b] βꢀpyranosylꢀconformation (25%), [c] αꢀ
furanosylꢀconformation (18%) and [d] βꢀfuranosylꢀconformation (9%) of mannosyllysine;
percentage of conformation calculated from the peakꢀarea of the proton at C1
25
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 26 of 32
Table 3. Transitions Recorded during MRM Measurement of MRPs in bakery products.
Time
Compound
Precursor Product Fragmentor Collision Dwell Q/q[a]
frame
ion [m/z] ion
[m/z]
84
voltage [V] energy
[eV]
time
[ms]
70
4–8.3 min CML
CEL
205
205
219
219
100
100
100
100
100
100
100
100
90
20
10
20
10
20
10
20
10
20
20
Q
q
130
84
70
70
Q
q
130
180
210
110
273
114
166
70
Glucosyllysine 309
100
100
100
100
200
200
q
309
Fructosyllysine 309
309
Q
q
Q
Q
q
8.5–19.5
min
MGꢀH1
229
229
90
[a] Q, transition used for quantitation; q, transition used for the confirmation of the presence
of the analyte.
26
ACS Paragon Plus Environment

Page 27 of 32
Journal of Agricultural and Food Chemistry
Table 4. Performance Parameters of the HPLCꢀMS/MS Method used for the Analysis of
MRPs in bakery products.
MRP
LOD[a]
LOQ
cV
calibration
range
[mg/kg cookie] [mg/kg cookie] [%]
[mg/L]
Fructosyllysine
2.1
5.2
3.7
0.4–20.1
0.08–38.9
Glucosyllysine[b] 1.3
2.0–20.4 0.02–10.4
CML
0.6
0.3
1.0
1.1
0.7
2.6
0.5–19.7 0.005–1.4
0.8–19.9 0.005–2.7
1.2–20.3 0.02–8.2
CEL
MGꢀH1
[a] Limits of detection (LOD) and limits of quantitation (LOQ) are given in mg/kg cookie for
proteinꢀbound MRPs based on the average protein content of all analyzed cookies.
Coefficients of variation (cV) were determined by triplicate analysis of different cookies.
27
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 28 of 32
Table 5. Concentrations of proteinꢀbound MRPs in cookie dough, baked cookies and
purchased baked products in mg/kg cookie
sweetener
sucrose[a]
glucose
fructose
honey
Fructosyllysine LysGlc/Man CML
CEL
nd
MGꢀH1
nd–tr
tr
dough
baked
tr
tr
nd
tr
nd–tr
1.71 ±
0.34
nd
cookies
459 ± 37
11.28 ± 0.61 7.9 ± 1.5 1.541 ±
0.013
8.8 ± 1.7
57.7 ± 1.1
312 ± 63
118 ± 24
40.4 ± 1.3
15.7 ± 1.7
6.22 ±
0.71
4.76 ±
0.87
137.5 ±
7.3
6.940 ±
0.057
7.67 ±
0.18
2.97 ±
0.15
27.7 ±
2.5
banana
63.65 ± 0.29
2.45 ±
0.12
14.7 ±
3.0
purchased honey (7)[c] 108–469
19–208
13–76
10.5 ±
1.5
2.5–32
23.5 ±
4.2
10–93
60.7 ±
3.4
bakery
banana (1)
61.3 ± 4.4
24.9 ± 1.5
products[b]
invert sugar 116–717
syrup (4)[d]
82–287
32–42
12–53
45–218
[a] data are mean ± SD (n ≥ 2) for one kind of product, data are given as ranges for groups of
different bakery products, the number of different bakery products is stated in brackets. [c] 4
products also contained glucose syrup [d] 2 products also contained small amounts of honey;
nd, not detected (below LOD); tr, traces (between LOD and LOQ)
28
ACS Paragon Plus Environment

Page 29 of 32
Journal of Agricultural and Food Chemistry
Figures
Fig. 1 (oneꢀcolumn figure)
29
ACS Paragon Plus Environment