Diagnostic MS/MS fragmentation patterns for the discrimination between Schiff bases and their Amadori or Heyns rearrangement products

Article abstract of DOI:10.1016/j.carres.2020.107985

Schiff bases, the Amadori and Heyns rearrangement products are the most important isomeric intermediates involved in the early Maillard reaction; distinguishing between them by analytical mass spectroscopic techniques remains a challenge. Here we demonstrate that MS/MS fragmentation patterns can be used for the discrimination between glucose derived Schiff bases, Amadori, and Heyns compounds with glycine. An ESI-qTOF-MS system operated in the positive mode under both acidic and neutral conditions was employed to generate unique MS/MS fragmentation patterns of the molecules. Analysis of the MS data has indicated that acidic medium is suitable for generating characteristic and diagnostic ions. At high collision energy (20 eV), the spectrum of Schiff base was largely uninformative, whereas both Amadori and Heyns compounds undergo characteristic fragmentations with high diagnostic value. At low collision energy values (10eV), we observed formation of prominent diagnostic ions from the Schiff base precursor, as well as extensive dehydration reactions of all three molecules. Under acidic conditions, the diagnostic fragmentation pattern of the Amadori compound featured consecutive dehydration reactions. At higher values (20 eV) it underwent the α-fission at the carbonyl group and produced a prominent diagnostic ion [AA + H + CH₂]⁺ at m/z 88. The Schiff base was found to preferentially undergo the retro-aldol degradation and produce diagnostic ions at m/z 118 [AA + H + diose]⁺ and m/z 140 [AA + Na + diose]⁺, together with their sugar complements at m/z 85 [tetrose + H–2H₂O]⁺ and m/z 143 [tetrose + Na]⁺. In the case of Heyns compound, several diagnostic ions were also detected, including the ions at m/z 154 [M + H–2H₂O–C₂H₄O₂]⁺, m/z 170 [AA + Na + triose]⁺ and m/z 142 [AA + H + Furan]⁺.

Full text of DOI:10.1016/j.carres.2020.107985

Journal Pre-proof
Diagnostic MS/MS fragmentation patterns for the discrimination between Schiff bases
and their Amadori or Heyns rearrangement products
Haoran Xing, Valeri V. Mossine, Varoujan Yaylayan
PII:
S0008-6215(19)30580-4
DOI:
https://doi.org/10.1016/j.carres.2020.107985
CAR 107985
Reference:
To appear in: Carbohydrate Research
Received Date: 2 October 2019
Revised Date: 2 March 2020
Accepted Date: 12 March 2020
Please cite this article as: H. Xing, V.V. Mossine, V. Yaylayan, Diagnostic MS/MS fragmentation patterns
for the discrimination between Schiff bases and their Amadori or Heyns rearrangement products,
Carbohydrate Research (2020), doi: https://doi.org/10.1016/j.carres.2020.107985.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.

1
2
3
4
Diagnostic MS/MS fragmentation patterns for the discrimination between Schiff bases and
their Amadori or Heyns rearrangement products
5
6
7
Haoran Xinga, Valeri V. Mossineb , Varoujan Yaylayana*
8
9
a Department of Food Science & Agricultural Chemistry, McGill University, 21111 Lakeshore,
Ste Anne de Bellevue, Quebec, Canada, H9X 3V9. varoujan.yaylayan@mcgill.ca
10
11
12
b Department of Biochemistry, University of Missouri, Columbia, MO 65211, USA
*Author to whom correspondence should be addressed

13
Highlights:
14
15
16
17
18
19
20
21
22
23
•
The main diagnostic ion for the Schiff base during MS/MS fragmentation is generated
through the retro-aldol degradation reaction yielding a diose attached to the amino acid
moiety;
•
The diagnostic ions of the Amadori compound result from a series of dehydrations of the
intact molecule at low collision voltages, and a residual CH2 group from the sugar
attached to the amino acid moiety at the high voltage;
•
•
The diagnostic ion of the Heyns compound forms through the retro-aldol degradation
yielding a triose attached to the amino acid moiety;
The MS/MS fragmentation patterns are consistent under both acidic and neutral
conditions, but neutral condition provides more stability to the fragment ions.

24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Abstract:
Schiff bases, the Amadori and Heyns rearrangement products are the most important isomeric
intermediates involved in the early Maillard reaction; distinguishing between them by analytical
mass spectroscopic techniques remains a challenge. Here we demonstrate that MS/MS
fragmentation patterns can be used for the discrimination between glucose derived Schiff bases,
Amadori, and Heyns compounds with glycine. An ESI-qTOF-MS system operated in the positive
mode under both acidic and neutral conditions was employed to generate unique MS/MS
fragmentation patterns of the molecules. Analysis of the MS data has indicated that acidic
medium is suitable for generating characteristic and diagnostic ions. At high collision energy (20
eV), the spectrum of Schiff base was largely uninformative, whereas both Amadori and Heyns
compounds undergo characteristic fragmentations with high diagnostic value. At low collision
energy values (10eV), we observed formation of prominent diagnostic ions from the Schiff base
precursor, as well as extensive dehydration reactions of all three molecules. Under acidic
conditions, the diagnostic fragmentation pattern of the Amadori compound featured consecutive
dehydration reactions. At higher values (20 eV) it underwent the α-fission at the carbonyl group
and produced a prominent diagnostic ion [AA+H+CH2]+ at m/z 88. The Schiff base was found to
preferentially undergo the retro-aldol degradation and produce diagnostic ions at m/z 118
[AA+H+diose]+ and m/z 140 [AA+Na+diose]+, together with their sugar complements at m/z 85
[tetrose+H-2H2O]+ and m/z 143 [tetrose+Na]+. In the case of Heyns compound, several
diagnostic ions were also detected, including the ions at m/z 154 [M+H-2H2O-C2H4O2]+, m/z
170 [AA+Na+triose]+ and m/z 142 [AA+H+Furan]+.
45
46
Keywords: Diagnostic ions, Schiff base, Amadori compound, Heyns compound, tandem mass
spectrometry (MS/MS), Collision induced fragmentations, Maillard reaction

47
48
49
50
51
52
53
54
55
56
57
58
1. Introduction:
The Maillard reaction is a common term for a broad array of reactions that typically involve
reducing sugars, such as glucose, and aliphatic amino groups of biological molecules. It has been
the focus of great interest for researchers in areas as diverse as human pathology and food
chemistry for more than 100 years.[1, 2] The complex reaction cascade is initiated by the
formation of a Schiff base (or glycosylamine) between a reducing sugar and an amino acid. In
presence of a nucleophilic catalyst, glycosylamine derived from an aldose can rearrange into a
more stable 1-amino-1-deoxy-2-ketose; such reaction is termed the Amadori rearrangement.
When glycosylamine is derived from a ketose sugar, it can rearrange into a 2-amino-2-
deoxyaldose, or Heyns rearrangement product (HRP).[2] The breakdown of the Amadori or the
Heyns rearrangement products ultimately yields a complex mixture of volatile carbocyclic and
heterocyclic compounds, as well as characteristic yellow-brown oligomers and polymers.[3]
59
60
61
62
63
64
65
66
Although the initial stage of the Maillard reaction is well-documented, the analytical
discrimination of the early intermediates, i.e. isomeric Schiff base and ARP (or HRP), remains a
challenge. As a result, the critical role played by Schiff bases for initiating and propagating the
Maillard reaction perhaps may have been undervalued. Besides rearranging into ARP and HRP,
Schiff bases can undergo various transformations, such as the base-catalyzed transamination,[4-6]
the intramolecular cyclization leading to decarboxylated isomeric imines via oxazolidin-5-one
intermediate,[7, 8] the decarboxylative transamination,[9] the Namiki oxidative fragmentation
pathway,[10, 11] and the acid-catalyzed Pictet-Spengler condensation pathway.[12, 13]
67
68
69
Structural identification of small organic molecules have been one of the cornerstone
applications of mass spectrometry over the last decades, especially with the development of the
electrospray ionization in combination with a collision cell (ESI-MS/MS).[14] It can also be

70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
considered as one of the key analytical techniques that is capable of generating large datasets in
the emerging “omics” fields. Recently, the Maillard reaction has also been investigated in this
regard. [15, 16] As a complex reaction cascade, hundreds and thousands of ions were detected in
various Maillard reaction mixtures represented simply by their m/z values where the same
molecular weight or a given elemental composition could represent different isomeric structures.
The main goal of this study was to develop MS/MS based tools to distinguish the isomeric
intermediates from the Maillard reaction such as Amadori compounds and Schiff bases without
lengthy separation and purification steps. The MS/MS fragmentation of Amadori compounds
under protonated positive ion mode has been studied extensively and systematically by LC-ESI-
MS/MS.[17-19] The potential of distinguishing between proline derived Amadori and Heyns
compounds,[20] or distinguishing between hexose versus pentose-derived Amadori compounds
by their MS/MS fragmentation patterns have been already reported. [17] However, to the best of
our knowledge, there are no published studies on the MS/MS behaviour of the Schiff bases, nor
on their chemical behaviour under different collision energies. To fill this knowledge gap, we
ventured to identify specific and diagnostic collision-induced dissociations (CID) of
glucose/glycine derived Schiff base and ARP and of fructose derived HRP by LC-ESI-MS/MS
under acidic (formic acid) and ESI-MS/MS under neutral (sodium formate) conditions using
variable collision energies.
88
89
2. Material and Methods
2.1 Materials:
90
91
92
All reagents and chemicals were purchased from Sigma-Aldrich Chemical Co. (Oakville, ON,
Canada) and used without further purification. Synthesis and characterization of N-
carboxymethyl-D-glucosylamine (Schiff base), 1-carboxymethylamino-1-deoxy-D-fructose

93
94
95
96
97
(Amadori compound), 2-carboxymethylamino-2-deoxy-D-glucose (Heyns compound), 1-[(2'-
carboxy)pyrrolidinyl]-1-deoxy-D-fructose (proline Amadori compound), and N-[(2'-
carboxy)pyrrolidinyl]-D-glucosylamine (proline Schiff base) were performed according to
published procedures. [21-23] See Table S1 for 13C-NMR chemical shifts and the supplementary
material for more details on their preparation.
98
2.2 ESI/MS/MS under formic acid condition
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
Samples were analyzed using an Agilent 1290 Infinity II LC system coupled to the 6545 qToF -
MS (Agilent Technologies, Santa Clara, USA). The LC separation was conducted on a
Poreshell120 EC-C18 analytical column (Agilent Technologies; 2.7 µm × 3 mm × 100 mm)
connected with a Poreshell120 EC-C18 guard column (Agilent Technologies; 2.7 µm × 3 mm ×
5 mm). The elution condition was at a flow rate of 0.4 ml/min with the mobile phase as the
mixture of 0.1% (V/V) aqueous formic acid (40%) with methanol (60%). The injection volume
was 1 µL, and the column temperature was set to 20 °C. The samples were analyzed in positive
electrospray ionization mode using product ion (MS/MS) experiment. The drying gas
temperature was at 275 °C with a flow of 10 mL/min, the sheath gas temperature was at 300 °C
with a flow of 12 mL/min, the pressure on the nebulizer at 45 psi, the capillary voltage at 4500 V,
the fragmentor voltage at 1000 V, the skimmer voltage at 50 V, and the nozzle voltage at 2000 V.
The Tandem mass spectrometry (MS/MS) data was collected by scans between m/z 50 and 1000
at a scan rate of 3 spectra/s for four different collision energies (0 V, 10 V, 15 V, and 20 V) for
the ions at [M+H]+ 238. The data sets were processed with MassHunter Profinder B.08.00
software (Agilent Technologies).
114

115
2.3 ESI/MS/MS under sodium formate condition
116
117
118
119
120
121
122
123
124
The diluted sample solutions (1 µL) in methanol/ water (10%/ 90%) were supplied to the source
directly via a syringe. The analysis was on a Bruker Maxis Impact quadrupole time of flight
mass spectrometer (Bruker Daltonics, Bremen, Germany) operated in positive ion mode.
Instrument calibration was performed using sodium formate clusters. The electrospray interphase
settings were as follows: nebulizer pressure, 0.6 bar; drying gas, 4 L/min; temperature, 180 °C;
and capillary voltage, 4500 V. The scan range was from m/z 50 to 800. Tandem mass
spectrometry (MS/MS) was carried out in MRM mode using 10.0 eV collision energy for the
ions at [M+Na]+ 260. The data were analyzed using Bruker Compass Data Analysis software,
version 4.2.
125
126
3. Results and discussion
127
128
In-source fragmentations are usually observed under ESI conditions depending on the structural
features of the analytes (e.g. presence of basic or acidic functional groups) and on the
environmental conditions (e.g. solvents or calibration compound used for a specific MS system);
all the three analytes exhibited minor in-source fragmentations mainly under acidic conditions
(Figures 1-3). Under acidic analytical conditions, samples were introduced as they were eluted
from a reverse-phase LC column. To facilitate protonation of the samples, water/methanol
mixture containing 0.1% formic acid was used as the mobile phase. Schiff base, ARP, and HRP
showed similar retention times (1.101, 1.070 and 1.080 min respectively) as shown in Figure S1.
The MS spectra of these peaks indicated the occurrence of acid catalyzed hydrolysis especially
for the Schiff base (see Table 1 Entry 1 & Figure 1) which was identified as the most labile
129
130
131
132
133
134
135
136
137

138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
molecule compared to the ARP and HRP generating free glycine and free sugar, the latter being
the most abundant peak, while the intact protonated Schiff base represented only 11.2% of the
total ion intensity. Furthermore, dehydrations were also observed for ARP and HRP, especially
for the former. (Table 1 Entries 5, 9 & Figures 2 & 3) The environmental conditions could play a
critical role in distinguishing the Schiff base from the ARP, since the former, as indicated, can
undergo hydrolysis and/or rearrange under acidic conditions to ARP and consequently undergo
acid catalyzed dehydrations as shown in Table 1 (Entries 2 & 3) and Figure 1. Because of this
concern, the Schiff base, ARP and HRP were also analyzed under neutral conditions in sodium
formate cluster calibrated system. In this setup, samples were introduced as dilute solutions in
water/methanol to the source directly via a syringe, resulting in the formation of sodiated ions
(Figure 4). The extra stability gained by the analytes under neutral conditions was reflected in the
intensities of molecular ions generated at [M+Na]+ = 260 as shown in Figure 4 (100% relative
peak intensity in all cases). Some in-source fragmentation still observed for the Schiff base.
(Table 1 Entry 13 & Figure 4) However, the overall stability was greatly enhanced relative to the
acidic condition, which is the most adopted MS/MS analytical conditions from literature. In this
study, MS/MS under both acidic and neutral condition were investigated, and the product ions
were screened elementally and reported in Table 2, the relative abundances of the product ions
and their proposed structure can be found in supplementary material (Tables S2 & S3).
156
157
158
159
160
3.1 Diagnostic MS/MS fragmentations under acidic conditions
Under acidic conditions, the MS/MS fragmentation pattern of ARP was in accordance with those
reported in the literature,[17, 19] where series of dehydration reactions lead to the formation of
characteristic [M+H-n(18)]+ peaks in the MS/MS spectrum under low CID voltage (10eV),
whereas high CID voltage (20eV) resulted in α-fission between the C-1 and C-2 of sugar

161
162
163
164
165
166
167
168
169
170
171
172
backbone carbon atoms.[18, 20] This fragmentation pattern, where dehydrated ion series has
been observed as the most abundant peaks in the spectrum, is typical of most of the Amadori
compounds of eighteen amino acids with glucose studied in the literature.[19] Under all
investigated conditions reported in the literature, the cleavage of the sugar moiety during MS/MS
fragmentations of Amadori compounds was only observed between C1-C2 or C5-C6 of the sugar,
leaving a -CH2 unit attached to an amino acid or leading to a loss of formaldehyde [-H2CO] from
the terminal carbon of the hexose, consistent with data reported in Table 1 (Entries 6-8 & Figure
2). Briefly, under conditions of low collision energy, dehydrated products were the most
abundant peaks with 100% relative intensity for [M+H-H2O]+ (m/z 220) or [M+H-2H2O]+ (m/z
202) at 10eV or 15eV respectively. At higher collision energy values (20 eV), the glycine ARP
generated [AA+H+CH2]+ (m/z 88) through α-fission as the most abundant peak, as well as the
complementary sugar portion at m/z 97 (proposed as a pyrylium ion). (Scheme 1)
173
174
175
176
177
178
179
180
181
182
183
Contrary to the ARP, the Schiff base showed a different behaviour under MS/MS fragmentation,
especially under lower energy conditions (10 and 15eV) generating through C2-C3 retro-aldol
cleavage a diagnostic ion at (m/z 118) incorporating intact amino acid and two carbon moiety
from the sugar in addition to its complementary 4-carbon sugar fragment at m/z 85 (see Table 1
Entries 2, 3 & Scheme 1). Overall, the Schiff base of glycine and glucose appears to be less
stable during MS/MS fragmentation compared to its Amadori counterpart. Even when submitted
to low collision voltage (10 eV), the Schiff base generated the most abundant peak [AA+H]+ at
m/z 76 representing the free glycine released during the fragmentation process. It appears that
the free glycine was released from doubly dehydrated Schiff base precursor at m/z 202 leaving
behind the sugar residue at m/z 145 (see Table 2). At 15 eV, the spectrum does not differ much
from the one operated at 10 eV. However, at 20 eV the ion count intensity was too low (×102) to

184
185
186
187
188
189
190
191
192
provide useful information, and most of the ions could not be identified elementally. (Figure 1)
This property of Schiff bases can be exploited to detect their presence in mixtures containing
various amounts of Amadori products. When the ion at m/z 238 [M+H]+ in five different
mixtures of glycine Schiff base and its ARP (molar ratios of 1:0, 1:2, 1:1, 2:1, 0:1) was
fragmented using collision energy of 20 eV where Schiff bases no longer generate any fragments,
the intensity of ions generated from the ARP predominates the spectrum and its diagnostic ion at
m/z 88 increased proportionately to the increased content of ARP. On the other hand, the
intensity of the diagnostic ion originating from the Schiff base at m/z 118 was only detected in
trace abundances as expected (see Figure S4).
193
194
195
196
197
198
199
200
201
202
203
204
205
206
Although Heyns rearrangement product is structurally different from ARP, its MS/MS
fragmentation pattern was partially similar to that of ARP and partially to that of the Schiff base
(see Scheme 1). Its resemblance to the behaviour of ARP was manifested through the presence of
dehydration peaks and its similarity to the behaviour of the Schiff base was indicated through
detection of the ion at m/z 130 (Scheme 1). From the elemental composition of the ion at m/z
130 (C5H8NO3) it can be concluded that it incorporates intact glycine with three carbon unit
originating from the sugar. Such a structure can only be generated through retro-aldol cleavage
of the Schiff base of fructose with glycine at C3-C4 as shown in Scheme 1. This implies that
open form of Heyns product is prone to isomerization to produce fructose/glycine Schiff base
adduct as shown in Scheme 1. This diagnostic ion is generated only under acidic conditions,
although its precursor ions m/z 170 was also detected under neutral conditions. Furthermore,
detection of the high intensity protonated ion at m/z 118 and the same ion at m/z 140 under
sodiation with common molecular formula of C4H8NO3 indicates the predominance of the open
form of the Heyns product under MS/MS analytical conditions, since this ion can be generated

207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
only through C2-C3 retro-aldolization of the open form of the Heyns product. This is equivalent
to the process observed in the case of Schiff base shown in Scheme 1. The diagnostic ion at m/z
130 was absent from both Schiff base and the ARP and it is the most abundant ion at higher
collision voltages of the HRP indicating its stability and therefore the tendency of accumulating
relative to the other fragments. Its three-carbon sugar complement was detected only under
sodiation condition at m/z 113. (Table 1 Entry 18; Scheme 2) Interestingly, Yuan et al. (2016)[20]
reported a characteristic fragment ion from MS/MS of proline Heyns compound at m/z 182
formed by the loss of two water molecules and a two-carbon sugar moiety [M+H-2H2O-
C2H4O2]+, this fragment ion was absent from the MS/MS of its Amadori compound and can be
considered as a diagnostic ion for Heyns products. The glycine equivalent of this ion was also
identified in its Heyns product at m/z 142 (C6H8NO3) (Scheme 1) along with its two-carbon split
moiety at m/z 61 (C2H5O2). The ion at m/z 142 was nearly absent from MS/MS fragmentation of
glycine Schiff base and Amadori compound making it a good candidate as diagnostic ion for
HRP. Furthermore, we have identified another fragment ion at m/z 154 [M+H-3H2O-H2CO]+
that also preferentially formed only in Heyns compound, that can also be considered as a
diagnostic ion. (Figure 3) To illustrate the generality of the diagnostic fragmentation patterns
proposed above, another set of Amadori and Schiff base of glucose this time using proline was
studied under the same conditions (see Table 1, Figures S3 and S4). The product ion from the
Amadori compound at m/z 128 was identified as the diagnostic fragments ([AA+H+CH2]+)
which is the equivalent to m/z 88 of glycine ARP. Proline Schiff base, on the other hand,
produced the diagnostic fragment at m/z 158 equivalent to m/z 118 of glycine Schiff base (see
Scheme S1). Furthermore, both proline Schiff base and ARP behaved the same way compared to
glycine Schiff base and ARP in response to increased CID energy at 10, 15, and 20 eV. Both

230
231
232
ARPs showed dehydration as the major product ions at low CID energy values and tended to
undergo α-fission and results in accumulation of the diagnostic ion at m/z 88 (glycine ARP) and
m/z 128 (proline ARP) at higher energies.
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
3.2 Diagnostic MS/MS fragmentations under neutral condition
When the mass spectrometric run was performed in presence of sodium formate, both molecular
ions and their major fragments were detected exclusively as sodium adducts, (Figure 4). There
was no evidence that the Schiff base underwent the Amadori rearrangement under such
conditions. Additionally, direct infusion of the analytes compared to prior LC separation also
reduced the time that Schiff base spent in aqueous environment and minimized its hydrolysis.
Under neutral conditions, dehydration of the sugar moiety and the dehydration of its subsequent
retro-aldolization products were also not significant, thus allowing for the detection of intact
triose and tetrose species at m/z 113 and 143 respectively (Scheme 2). Under acidic conditions,
these molecules underwent dehydration and cyclization into furan and pyran derivatives, as
shown in Scheme 1. Under neutral conditions, in the MS/MS spectrum of the Heyns compound,
a peak at m/z 170 (see Scheme 2) was detected which was not observed under acidic conditions.
This ion can be considered as diagnostic ion for Heyns product under sodiation condition.
Taking together all the data, it seems that the retro-aldolization of C2-C3 sugar chain was the
main pathway of fragmentation for all the three intermediates studied under neutral conditions;
in the spectra of the Schiff base and Heyns product the respective peaks were relatively more
intense, as compared to the Amadori product (Figure 4). We assume that the observed retro-
aldolization of ARP at C2-C3 carbon bond occurred after retro-isomerization of ARP into Schiff

252
253
base. (Scheme 2) Furthermore, only the spectra of the Amadori product contained peaks, albeit
in low abundances, that were ascribed to its dehydration under neutral conditions.
254
255
256
257
258
259
260
261
262
263
264
265
4. Conclusion
We have demonstrated that ESI-qToF-MS/MS could be used to distinguish between
glucose/glycine Schiff base and two most relevant isomers, Amadori and Heyns rearrangement
products. Analysis performed under neutral conditions (using sodium formate clusters) is more
suitable for identification of the intact sodiated molecular ions, due to lower hydrolysis rates of
the Schiff base. On the other hand, MS/MS analysis under acidic conditions (such as in presence
of formic acid) offered more informative patterns of characteristic and diagnostic ions for each
molecule, thus allowing to discriminate the Schiff base from its corresponding Amadori and
Heyns rearrangement products. At higher collision energy (20 eV), both Amadori and Heyns
products undergo unique fragmentations that do not involve dehydration. In contrast, the Schiff
base generates useful diagnostic ions under low collision energy value (10-15 eV). Loss of water
from all three molecules is prominent at energy levels between 10-15 eV.
266
267
268
269
270
271
272
273
Conflicts of interest
There are no conflicts of interests to declare
Acknowledgements
The authors acknowledge funding for this research from Natural and Engineering Research
Council of Canada (NSERC), the FRQNT - Fonds Nature et technologies, and the China
Scholarship Council (CSC).

274
275
Figure Captions

276
Figures and Tables
277
278
279
Table 1: MS and MS/MS spectral data of Schiff base, Amadori and Heyns compounds and their
significant and diagnostic fragment ions.
#
Compound
Precursor
ion
[M+X]+
(m/z)
CID voltage Significant fragment ions in MS and
(eV)a MS/MS spectra [M+X]+ (m/z) c
Acidic conditions (ESI +ve)
1
Glycine Schiff base [M+H]+ 238 0
10
238, 203
238, 220, 202, 118 b, 85, 76
2
3
15
20
220, 202, 127, 118, 85, 97, 76
4
n/a
5
Glycine Amadori
Glycine Heyns
[M+H]+ 238 0
238, 220
6
7
10
15
20
220, 202
220, 202, 97, 88, 76
202, 97, 88, 76
8
9
[M+H]+ 238 0
238, 220, 202
10
11
10
220, 202, 154, 130
202, 156, 154, 142, 130, 126, 118,
117, 101, 99, 97, 84, 76
278, 116
15 and 20
12 Proline Schiff base
13
[M+H]+ 278 0
10 and 15
278, 260, 242, 232, 214, 158, 116, 85,
70
14
20
116, 70
15 Proline Amadori
16
[M+H]+ 278 0
278, 260
10, 15 and 20 278, 260, 242, 232, 214, 128, 116,
100, 70
Neutral conditions (ESI +ve)
17 Glycine Schiff base [M+Na]+
0
260, 203, 143, 140
260
18
10
0
143, 140
260
19 Glycine Amadori
[M+Na]+
260
20
10
0
260, 242, 224
260
21 Glycine Heyns
[M+Na]+
260
22
10
260, 170, 164, 143, 140, 113, 108

280
281
282
283
284
285
a: CID voltage at 0 represents ions found in the MS spectra without collision (fragmentation
occurs during ionization).
b: The characteristic ions are in bold, where they are either uniquely present in a sample or
significantly differ in their relative abundances.
c: Proposed structural assignments and relative abundances are reported in the supplementary
material (see Tables S1 and S2)

286
287
288
Table 2: Elemental composition of important product ions generated for glycine/glucose Schiff
base, Amadori and Heyns compounds during MS/MS under acidic and neutral conditions.
Acidic Condition
Neutral Condition
#
m/z
MF
Error (ppm)
#
m/z
MF
Error
(ppm)
Fragments derived from the adduct
Fragments derived from the adduct
1
2
3
4
5
6
7
8
238.0914
220.0812
202.0710
184.0605
174.0757
126.0546
156.0652
154.0491
C8H16NO7
C8H14NO6
C8H12NO5
C8H10NO4
C7H12NO4
C6H8NO2
C7H10NO3
C7H8NO3
5.783
0.964
1.225
5.883
3.636
6.374
6.204
4.662
1
2
3
4
5
6
7
260.0735
242.0631
224.0522
216.0842
198.0729
180.0628
162.0516
C8H15NNaO7
C8H13NNaO6
C8H11NNaO5
C7H15NNaO5
C7H13NNaO4
C7H11NNaO3
C7H9NNaO2
4.312
3.953
5.767
2.741
6.702
4.792
9.245
Fragments containing glycine
Fragments containing glycine
9
76.0388
C2H6NO2
C3H6NO2
C4H8NO3
C3H6NO
C5H8NO3
C4H8NO2
C4H6NO
11.224
8.558
6.93
8
9
98.0206
C2H5NNaO2
C3H5NNaO2
C4H7NNaO3
C5H9NNaO4
C4H7NNaO
C6H7NNaO3
12.223
8.163
9.732
7.219
0.615
8.308
10 88.0389
11 118.0495
12 72.0443
13 130.0501
14 102.0546
15 84.0421
110.0209
10 140.0310
11 170.0417
12 108.0426
13 164.0310
8.867
6.291
8.853
9.981

16 142.0495
C6H8NO3
5.759
Sugar fragments
Sugar fragments
14 185.0424
17 163.0598
18 145.0498
19 127.0384
20 109.0279
21 97.0279
22 99.0438
23 117.0551
24 101.0209
25 85.0283
26 69.0326
27 61.0288
C6H11O5
C6H9O4
C6H7O3
C6H5O2
C5H5O2
C5H7O2
C5H9O3
C4H5O3
C4H5O2
C4H5O
5.203
1.956
7.235
6.919
9.836
9.132
4.862
11.572
11.224
6.371
2.53
C6H10NaO5
C6H8NaO4
C5H10NaO5
C5H8NaO4
C4H8NaO4
C3H6NaO3
1.044
6.756
6.317
0.183
8.588
11.181
15 167.0309
16 173.0415
17 155.0320
18 143.0308
19 113.0202
C2H5O2

289

290
291
292
Figure 1: MS and MS/MS spectra acquired at 0, 10, 15, 20 eV of glycine Schiff base under acidic condition.
293

294
295
Figure 2: MS and MS/MS spectra acquired at 0, 10, 15, 20 eV of glycine Amadori compound under acidic condition.
296

297
298
299
300
Figure 3: MS and MS/MS spectra acquired at 0, 10, 15, 20 eV of glycine Heyns compound under acidic condition; the diagnostic ions
are indicated in the graph.

301
302
303
Figure 4: MS and MS/MS spectra at 0 and 10eV of glycine Schiff base, Amadori, and Heyns compound under neutral condition; the
diagnostic ions are indicated in the graph.
304

305
306
307
308
309
310
311
312
Scheme 1: Proposed diagnostic collision-induced fragmentations of the glycine Schiff base and
Amadori or Heyns rearrangement products under acidic conditions. The characteristic loss of
water molecules from ARP is not shown.

313
314
315
316
317
318
319
320
321
Scheme 2: Proposed diagnostic collision-induced fragmentations of the glycine Schiff base and
Amadori or Heyns rearrangement products under neutral sodiation conditions.

322

323
Reference:
324
325
326
327
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
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
1.
2.
3.
Maillard, L., Action of amino acids on sugars. Formation of melanoidins in a methodical way.
Compte-Rendu de l'Academie des Sciences, 1912. 154: p. 66-68.
Nursten, H., The Maillard Reaction: Chemistry, Biochemistry and Implications. 2005, Cambridge,
UK: The Royal Society of Chemistry. 214.
Hodge, J.E., Dehydrated foods, chemistry of browning reactions in model systems. Journal of
agricultural and food chemistry, 1953.
1(15): p. 928-943.
4.
5.
Høltermand, A., The browning reaction. Starch
‐Stärke, 1966. 18(10): p. 319-328.
Herbst, R. and L. Engel, A reaction between α-ketonic acids and α-amino acids. Journal of
Biological Chemistry, 1934. 107(2): p. 505-512.
Wnorowski, A. and V.A. Yaylayan, Monitoring carbonyl-amine reaction between pyruvic acid and
α-amino alcohols by FTIR spectroscopy a possible route to Amadori products. Journal of
agricultural and food chemistry, 2003. 51(22): p. 6537-6543.
6.
7.
8.
Chu, F.L. and V.A. Yaylayan, Post‐Schiff Base Chemistry of the Maillard Reaction: Mechanism of
Imine Isomerization. Annals of the New York Academy of Sciences, 2008. 1126(1): p. 30-37.
Chu, F.L. and V.A. Yaylayan, FTIR monitoring of oxazolidin-5-one formation and decomposition in
a glycolaldehyde–phenylalanine model system by isotope labeling techniques. Carbohydrate
research, 2009. 344(2): p. 229-236.
Granvogl, M., S. Bugan, and P. Schieberle, Formation of Amines and Aldehydes from Parent
Amino Acids during Thermal Processing of Cocoa and Model Systems:ꢀ New Insights into
Pathways of the Strecker Reaction. Journal of Agricultural and Food Chemistry, 2006. 54(5): p.
1730-1739.
Hayashi, T., S. Mase, and M. Namiki, Formation of the N, N'-Dialkylpyrazine Cation Radical from
Glyoxal Dialkylimine Produced on Reaction of a Sugar with an Amine or Amino Acid. Agricultural
and Biological Chemistry, 1985. 49(11): p. 3131-3137.
Glomb, M.A. and V.M. Monnier, Mechanism of protein modification by glyoxal and
glycolaldehyde, reactive intermediates of the Maillard reaction. Journal of Biological Chemistry,
1995. 270(17): p. 10017-10026.
Manini, P., M. d'Ischi, and G. Prota, An Unusual Decarboxylative Maillard Reaction between l-
DOPA and d-Glucose under Biomimetic Conditions:ꢀ Factors Governing Competition with
Pictet−Spengler Condensaꢁon. The Journal of Organic Chemistry, 2001. 66(15): p. 5048-5053.
Manini, P., A. Napolitano, and M. d’Ischia, Reactions of d-glucose with phenolic amino acids:
further insights into the competition between Maillard and Pictet–Spengler condensation
pathways. Carbohydrate Research, 2005. 340(18): p. 2719-2727.
Demarque, D.P., et al., Fragmentation reactions using electrospray ionization mass spectrometry:
an important tool for the structural elucidation and characterization of synthetic and natural
products. Natural product reports, 2016. 33(3): p. 432-455.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Hemmler, D., et al., Evolution of complex maillard chemical reactions, resolved in time. Scientific
reports, 2017.
Hemmler, D., et al., Insights into the Chemistry of Non-Enzymatic Browning Reactions in
Different Ribose-Amino Acid Model Systems. Scientific reports, 2018. (1): p. 1-10.
7(1): p. 1-6.
8
Davidek, T., et al., Analysis of Amadori compounds by high-performance cation exchange
chromatography coupled to tandem mass spectrometry. Analytical chemistry, 2005. 77(1): p.
140-147.
Hau, J., S. Devaud, and I. Blank, Detection of Amadori compounds by capillary electrophoresis
coupled to tandem mass spectrometry. Electrophoresis, 2004. 25(13): p. 2077-2083.
18.

369
370
371
372
373
374
375
376
377
378
379
19.
20.
21.
Wang, J., et al., Electrospray positive ionization tandem mass spectrometry of Amadori
compounds. Journal of Mass Spectrometry, 2008. 43(2): p. 262-264.
Yuan, H., et al., The Comparison of the Contents of Sugar, Amadori, and Heyns Compounds in
Fresh and Black Garlic. Journal of Food Science, 2016. 81(7): p. C1662-C1668.
Mossine, V.V., G.V. Glinsky, and M.S. Feather, The preparation and characterization of some
Amadori compounds (1-amino-1-deoxy-d-fructose derivatives) derived from a series of aliphatic
ω-amino acids. Carbohydrate Research, 1994. 262(2): p. 257-270.
22.
23.
Mossine, V.V., et al., Molecular and crystal structure of N-(2-deoxy-
D-aldohexos-2-yl)-glycines
(Heyns compounds). Carbohydrate Research, 1996. 284(1): p. 11-24.
Mossine, V.V., et al., Superoxide free radical generation by Amadori compounds: the role of
acyclic forms and metal ions. Chemical Research in Toxicology, 1999. 12(3): p. 230-236.
380