Analysis of Advanced Glycation Endproducts in Rat Tail Collagen and Correlation to Tendon Stiffening

Article abstract of DOI:10.1021/acs.jafc.8b00937

Methylglyoxal is a major 1,2-dicarbonyl compound in vivo and leads to nonenzymatic protein modifications, known as advanced glycation endproducts. Especially long-lived proteins like collagen are prone to changes of the mechanical or biological function,

Full text of DOI:10.1021/acs.jafc.8b00937

Article
Cite This: J. Agric. Food Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/JAFC
Analysis of Advanced Glycation Endproducts in Rat Tail Collagen and
Correlation to Tendon Stiffening
Tobias Jost, Alexander Zipprich, and Marcus A. Glomb*
†
‡
,†
†
Institute of Chemistry−Food Chemistry, Martin-Luther-University Halle-Wittenberg, Kurt-Mothes-Strasse 2, D-06120 Halle,
Germany
‡
Department of Internal Medicine I, Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Strasse 40, D-06120 Halle, Germany
ABSTRACT: Methylglyoxal is a major 1,2-dicarbonyl compound in vivo and leads to nonenzymatic protein modifications,
known as advanced glycation endproducts. Especially long-lived proteins like collagen are prone to changes of the mechanical or
biological function, respectively, by accumulation of Maillard-derived modifications. Specifically, the resulting nonenzymatic
cross-link structures in parallel to the natural maturation process of collagen fibrils lead to complications with age or during
disease. A novel lysine−lysine amide cross-link derived from methylglyoxal, 2,15-diamino-8-methyl-9-oxo-7,10-diaza-1,16-
hexadecanedioic acid, named MOLA, was synthesized and identified in vitro and in vivo. Tail tendons of young, adult, and old
rats (3, 12, and 22 months) were enzymatically digested prior to analysis of acid-labile glycation products via liquid
chromatography−tandem mass spectrometry (LC−MS/MS). As a result, nine monovalent amino acid modifications, mostly
originating from methylglyoxal (36 μmol/mol leucine-equivalents in total), and four glycation cross-links (0.72 μmol/mol
glucosepane, 0.24 μmol/mol DODIC (3-deoxyglucosone-derived imidazoline cross-link), 0.04 μmol/mol MODIC
(
methylglyoxal-derived imidazoline cross-link), 0.34 μmol/mol MOLA) were quantitated in senescent tendon collagen. The
results correlated with increased tail tendon breaking time from 10 to 190 min and indicate that methylglyoxal is a major player
in the aging process of connective tissue.
KEYWORDS: methylglyoxal, Maillard reaction, advanced glycation endproducts, collagen, glycation cross-links,
connective tissue stiffening
9
INTRODUCTION
degradation of triose phosphates. Quantitation of MGO in
vivo was conducted comprehensively. Reported MGO levels in
human plasma range from 100 to 500 nM in healthy
■
The Maillard reaction or glycation is one crucial aspect of
protein modification during aging in vivo. In complex reaction
cascades reducing sugars, especially 1,2-dicarbonyl compounds,
interact with functional groups of lysine and arginine side
chains to form stabile amino acid modifications, named
10
nondiabetic subjects, while in rats plasma levels are around
μM. In cells and tissues, MGO levels are commonly 2−4
μM. Until now, nine MGO-related amino acid derivatives are
11
2
12
6
1
reported in the literature. Lysine modifications are CEL and N -
advanced glycation endproducts (AGEs). Long-lived proteins
of the extracellular matrix, with a turnover of several years, are
13,14
lactoyl lysine.
The methylglyoxal hydroimidazolones MG-
7
H3 and MG-H1, their open-chained form N -carboxyethyl
arginine (CEA) and the two pyrimidine structures tetrahy-
dropyrimidine (THP) and argpyrimidine are derived from
highly susceptible to structural changes due to accumulation o2f
AGEs over time resulting in tissue stiffening and dysfunction.
Collagen is the most abundant polypeptide in connective
tissues (e.g., tendon, skin, bone, cartilage, or vessels). A total of
15
arginine. Two cross-link structures are known to originate
from MGO, the acid-stable lysine−lysine imidazolium com-
pound MOLD (methylglyoxal lysine dimer), and the acid-
labile lysine-arginine amidine MODIC (methylglyoxal-derived
2
6 different types of collagen with a broad functional diversity
16
3
exist. The fibril-forming type I collagen, mainly found in skin
and tendons, is an established model for both in vitro and in
vivo studies due to its structural homology among different
organisms and, thus, has been studied intensively. Glycation
cross-link structures are assumed to contribute and to explain
17
imidazoline cross-link). The aim of the present work was
therefore to gain further insights into the mechanisms of AGE
formation in collagen and to highlight relevant structures with a
focus on MGO chemistry. Correlation of degree of chemical
modification to change of physical properties was first accessed
in an ex vivo model using 3 months old rat tail tendons
incubated with methylglyoxal (200 μM) under physiological
conditions (pH 7.4, 37 °C) for 7 days. The study was then
extended to the natural aging process using native tendons of
4
,5
collagen damage on a molecular basis. However, only a few
6
selected structures, specifically N -carboxymethyl lysine
6
(
CML), N -carboxyethyl lysine (CEL), MG-H1, and pentosi-
dine, were reported in literature as collagen modifications in
6
,7
vivo. In more recent work, glucosepane was considered as a
5
,8
major glycation cross-link in extracellular matrix protein.
One very potent glycating agent in biological systems is
methylglyoxal (MGO). MGO is a short-chained, highly reactive
Received: February 20, 2018
Revised: March 29, 2018
Accepted: March 31, 2018
1,2-dicarbonyl compound originating from numerous sources
in vivo including the Maillard triggered breakdown of longer-
chained sugars but foremost the enzymatic and nonenzymatic
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
young, adult and old rats (3, 12, and 22 months). A novel
MGO-derived amide AGE cross-link structure was identified
and quantitated in parallel to other established AGEs.
extracted with ethyl acetate. The organic layer was dried with sodium
sulfate, solvents were evaporated, and the residue was subjected to
column chromatography (silica gel, n-hexane/acetone, 80:20, v/v).
Fractions with material having an R value of 0.10 were combined and
f
solvents were evaporated to yield 5 as a colorless oil (301 mg, 53%).
−
MATERIALS AND METHODS
HR-MS: m/z 473.2864 [M − H] (found), m/z 473.2868
■
−
(
calculated for C H N O [M − H] ).
Chemicals. All chemicals of the highest quality available were
purchased from Sigma-Aldrich (Munich/Steinheim, Germany) and
Carl Roth AG (Karlsruhe, Germany) unless otherwise indicated.
Methylglyoxal was prepared freshly from its dimethyl acetal as
23 41
2
8
13
C NMR (100 MHz, CD OD) δ [ppm] = 175.6, 173.8, 158.1,
3
1
2
54.3, 83.4, 82.5, 80.4, 57.2, 55.8, 47.6, 32.4, 28.8, 28.7, 28.3, 28.0,
4.0, 17.4.
1
1
5
6
18
6
H NMR (500 MHz, CD OD) δ [ppm] = 4.03 (q, J = 7.0 Hz, 1H),
described by Klo
̈
pfer et al.1 N -carboxymethyl lysine (CML), N -
3
9
7
20
13
3.94 (s, 1H), 3.30−3.04 (m, 2H), 1.82−1.51 (m, 6H), 1.47 (d, J = 2.0
Hz, 3H), 1.47 (s, 9H), 1.45 (s, 18H).
glycoloyl lysine (GALA), N -carboxymethyl arginine (CMA),
21
6
glyoxal hydroimidazolone (G-H3), N -carboxyethyl lysine (CEL),
N -lactoyl lysine, and argpyrimidine were synthesized according to
6
14
22
1,16-Di(t-butyl)-2,15-di[(t-butyloxycarbonyl)amino]-7-(t-butylox-
ycarbonyl)-8-methyl-9-oxo-7,10-diazahexadecanedioate (6). 100
mg (0.21 mmol) of 5 were dissolved in 2 mL of absolute
tetrahydrofuran (THF) and 41 mg (0.30 mmol) of 1-hydroxybenzo-
triazole were added while stirring. 47 mg (0.30 mmol) of 1-ethyl-3-(3-
(dimethylamino)propyl)carbodiimide dissolved in 1 mL of THF were
added dropwise at 0 °C. After 30 min 64 mg (0.21 mmol) of 3
dissolved in 2 mL of THF were added to the solution. The reaction
mixture was stirred for 24 h at room temperature. Solvents were
evaporated and the resulting residue was subjected to column
chromatography (silica gel, n-hexane/acetone, 70:30, v/v). Fractions
literature. Methylglyoxal hydroimidazolones (MG-H1 and MG-H3) as
7
well as N -carboxyethyl arginine (CEA) and tetrahydropyrimidine
2
(
THP) were isolated from methylglyoxal/N -t-Boc-arginine reaction
1
5
mixtures as described previously by our working group. Preparation
23
of glucosepane, MODIC (methylglyoxal-derived imidazoline cross-
17
link)2 and DODIC (3-deoxyglucosone-derived imidazoline cross-
4
link) was carried out as described by Lederer’s group.
Synthesis of MOLA. The basic synthetic route was adapted with
19
modifications from Glomb and Pfahler. Thin-layer chromatography
TLC) was performed on silica gel 60 F254 (Merck, Germany).
(
with an R value of 0.43 were combined and solvents were evaporated
Visualization of separated material was achieved with ninhydrin (2%
ninhydrin + 5% acetic acid in n-butanol, w/v/v). Preparative column
chromatography was performed on silica gel 60 (0.06−0.20 mm,
Merck, Germany). Solvents were all chromatographic grade. From the
individual fractions, solvents were removed under reduced pressure.
f
to yield 6 as a colorless oil (150 mg, 95%).
+
HR-MS: m/z 759.5121 [M + H] (found), m/z 759.5114
+
(
calculated for C H N O [M + H] ).
38 71
4
11
2
,15-Diamino-8-methyl-9-oxo-7,10-diaza-1,16-hexadecanedioic
2
6
acid (MOLA, 7). 100 mg (0.13 mmol) of 6 were dissolved in 2 mL of
THF and 2 mL of hydrochloric acid (6.0 M) were added. The reaction
was monitored by TLC (silica gel, n-butanol/water/acetic acid/
pyridine, 4:2:3:3, v/v/v/v). Solvents were removed under reduced
pressure. The product was dried over potassium hydroxide under high
vacuum to yield 7 as a brown, amorphic material (44 mg, 98%, MOLA
t-Butyl-N -(t-butyloxycarbonyl)-N -(benzyloxycarbonyl)-L-lysi-
2
5
nate (2). 2 was synthesized according to Strazzolini et al.: 2.0 g (5.26
mmol) of N -(t-butyloxycarbonyl)-N -(benzyloxycarbonyl)-L-lysine
1) and 330 mg (2.70 mmol) of 4-dimethylaminopyridine were
2
6
(
dissolved in 10 mL of warm t-butanol. 1222 mg (5.60 mmol) of di-t-
butyl dicarbonate (Boc O) were added. The solution was stirred for 24
2
×
3 HCl).
HR-MS: m/z 345.2141 [M-H] (found), m/z 345.2143 (calculated
h at room temperature (RT). Solvents were evaporated and the
resulting residue was subjected to column chromatography (silica gel,
n-hexane/acetone, 3:1, v/v). Fractions with an R value of 0.38 were
−
−
for C H N O [M-H] ).
15 29
4
5
f
13
C NMR (100 MHz, D O): δ [ppm] = 172.99, 172.89, 170.48,
combined (TLC, same eluent), and solvents were evaporated to yield
2
1
70.39, 57.21, 53.88, 53.76, 46.76, 40.03, 39.90, 30.43, 30.29, 28.74,
2
as a colorless oil (2.2 g, 95%).
2
28.72, 26.17, 22.58, 22.49, 16.72.
t-Butyl-N -(t-butyloxycarbonyl)-L-lysinate (3). 2.2 g (5.01 mmol)
1
H NMR (500 MHz, D O): δ [ppm] = 4.14−4.07 (m, 2H), 3.98 (q,
of 2 were dissolved in 10 mL of methanol/dichloromethane (1:1) with
a small amount of palladium/activated charcoal (10%). The stirred
mixture was purged with hydrogen for 2 h. The reaction was
2
J = 7.0 Hz, 1H), 3.37−3.21 (m, 2H), 3.14−2.94 (m, 2H), 2.09−1.89
(
3
m, 4H), 1.84−1.73 (m, 2H), 1.67−1.38 (m, 6H), 1.53 (d, J = 7.0 Hz,
H).
Animals. Three and 12 months old male Wistar rats were bred in
monitored by TLC (R 0.24, silica gel, methanol/dichloromethane/
f
triethylamine, 70:29:1, v/v/v). After filtration through Celite, solvents
the Center of Medical Basic Research (ZMG), Medical Faculty,
Martin-Luther-University (Halle, Germany). Eighteen months old
male Wistar rats were purchased from Janvier Laboratories (Le
Genest-Saint-Isle, France) and kept in the ZMG until the age of 22
months. Rats were housed in standard cages in a climate room with 12
h light and dark phases and free access to food. The American
Physiological Society guide principles for the care and use of animals
were followed. All animal experiments were approved by the local
animal committee (42502-2-1123 MLU, Landesverwaltungsamt
Sachsen-Anhalt, Germany). Individuals at the age of 3 months (n =
were evaporated to yield 3 as a brown oil (1.4 g, 93%).
2
6
t-Butyl-N -(t-butyloxycarbonyl)-N -(carboxyethyl)-L-lysinate (4).
00 mg (2.31 mmol) of 3 and 203 mg (2.31 mmol) of pyruvic acid
7
were dissolved in 5 mL of absolute methanol. After addition of 435 mg
6.93 mmol) of sodium cyanoborohydride, the solution was stirred at
(
RT for 3 h. Solvents were evaporated and the resulting residue was
subjected to column chromatography (silica gel, methanol/ethyl
acetate, 30:70, v/v). Fractions with material having an R value of 0.24
f
were combined and solvents were evaporated to yield 4 as a white
powder (450 mg, 52%).
−
8
1
), 12 months (n = 8), and 22 months (n = 6) were anaesthetized with
50 mg/kg bodyweight Narcoren (Merial, France). In deep narcosis,
HR-MS: m/z 373.2339 [M − H] (found), m/z 373.2344
−
(
calculated for C H N O [M − H] ).
18
33
2
6
1
3
C NMR (100 MHz, CD OD): δ [ppm] = 174.4, 173.5, 158.1,
animals were killed by exsanguination and tails were removed.
Tendons of the same diameter were prepared from the ventral bundles
of the skinned tails.
3
8
2.6, 80.5, 59.2, 55.5, 47.1, 32.1, 28.7, 28.3, 27.0, 24.1, 16.2.
1
H NMR (500 MHz, CD OD): δ [ppm] = 4.00−3.91 (m, 1H),
3
3
.52 (q, J = 7.2 Hz, 1H), 3.05−2.90 (m, 2H), 1.85−1.48 (m, 6H), 1.48
Incubation of Tendons. Tendons at the age of 3 months were
incubated separately in 1.5 mL of phosphate buffered saline (PBS, 10
mM phosphate, 150 mM sodium chloride, pH 7.4) containing
methylglyoxal (0.2 mM) and diethylenetriaminepentaacetic acid
(DTPA, 1 mM) under deaerated conditions at 37 °C in an incubator
shaker (New Brunswick Scientific, New Jersey) for 1−7 days. A
control was incubated in PBS under the same conditions. Six tendons
were incubated for each time point. Afterward, tendons were washed
three times with PBS and kept at −20 °C until analyses.
(
d, J = 7.2 Hz, 3H), 1.46 (s, 9H), 1.44 (s, 9H).
2
6
6
t-Butyl-N ,N -bis(t-butyloxycarbonyl)-N -(carboxyethyl)-L-lysinate
5). 450 mg (1.20 mmol) of 4 were dissolved in 10 mL of a
triethylamine solution (10% in absolute methanol, v/v). 1310 mg
6.00 mmol) of Boc O were added and the reaction mixture was
(
(
2
refluxed for 30 min while stirring. Stirring was continued for 30 min at
RT, solvents were evaporated, and the resulting residue was taken up
in ice-cold 0.5 N hydrochloric acid. The emulsion was immediately
B
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
Table 1. LC−ESI-MS/MS Parameters for AGE Measurements
a
b
b
precursor ion
product ion 1
product ion 2
product ion 3
c
d
e
analyte
retention time
min]
m/z
DP
[V]
m/z
CE
CXP
[V]
m/z
CE
CXP
[V]
m/z
CE
CXP
[V]
[
[eV]
[eV]
[eV]
CML
11.6
12.2
16.0
17.1
18.6
19.5
19.8
24.2
24.4
24.4
27.4
28.6
28.8
28.8
29.3
29.7
29.8
30.3
31.0
31.3
31.7
205.1
205.2
219.2
219.1
319.2
215.1
233.1
229.2
247.1
229.2
276.1
327.2
429.3
333.2
447.4
341.3
347.5
343.3
255.3
357.3
379.2
42
40
40
54
52
48
55
45
51
55
70
60
15
45
45
45
80
20
50
25
84.1
142.1
156.2
84.1
30
20
20
33
66
20
43
21
48
23
44
51
41
54
33
33
58
32
44
45
47
14
11
8
130.2
84.1
17
36
35
18
36
38
23
45
25
43
27
31
55
26
42
52
31
44
24
35
9
14
9
56.1
56.2
59
64
17
59
33
20
22
21
24
21
27
28
55
32
53
31
37
74
28
45
9
8
GALA
6
N -lactoyl lysine
84.1
173.1
56.1
8
CEL
7
130.1
116.3
70.1
11
9
8
THP
70.2
11
8
186.4
116.2
118.2
116.1
132.1
116.1
213.2
198.1
339.2
130.2
357.2
212.3
130.2
70.2
10
10
6
G-H3
100.1
70.1
11
8
CMA
12
9
116.1
70.2
MG-H3
CEA
114.2
70.2
11
10
12
11
14
20
12
20
14
10
13
10
7
9
11
12
14
13
19
13
16
18
7
116.2
70.1
10
9
MG-H1
LNL
166.2
84.1
130.1
282.3
269.2
169.1
402.5
84.2
16
14
20
9
GOLD
glucosepane
GOLA
84.1
384.5
84.3
DODIC
MOLD
MOLA
GODIC
argpyrimidine
MODIC
pentosidine
301.3
296.3
84.1
18
17
12
11
13
15
14
173.2
183.2
140.0
312.2
316.3
298.4
70.2
7
12
14
16
192.1
267.3
84.2
197.4
187.3
20
36
8
59
a
b
c
d
MRM transition used for quantitation (quantifier). MRM transition used for confirmation (qualifier). Declustering potential. Collision energy
e
Cell exit potential.
Tail Tendon Breaking Time Assay. Measurement of tail tendon
breaking time (TTBT) of incubated tendons (3 months old) and
native tendons of different age (3, 12, and 22 months) was performed
as described by Harrison and Archer. Two centimeters of each
tendon were fixated with surgical thread at both ends with a slipknot
and a weight of 2.746 g. The tendons were immersed in a urea solution
Hydrochloric acid was removed by vacuum centrifugation. The dry
residue was taken up in 492 μL of diluted hydrochloric acid (0.05
mM) and filtered through a centrifugal tube filter (Costar Spin-X, 0.45
μm, cellulose acetate, Corning Inc., New York).
26
Enzymatic Hydrolysis. To 500 μL of above collagen solution, 25
units of pepsin were added and incubated at 37 °C for 24 h. The
solution was evaporated to dryness by vacuum centrifugation. The
residue was taken up in 480 μL of TRIS acetate buffer (TA, 100 mM
TRIS, 5 mM calcium chloride, pH 7.4). Stepwise, the following
proteases were added every 24 h: 10.0 units of collagenase, again 10.0
units of collagenase, 0.3 units of Pronase E, again 0.3 units of Pronase
E, 1.0 unit of leucine aminopeptidase (LAP), and 0.95 units of
carboxypeptidase Y. The samples were incubated in an incubator
shaker at 37 °C. A small crystal of thymol was added with the first
digestion step in TA. After the last digestion step, solutions were
filtered through a molecular weight cutoff membrane (MWCO 3000
centrifugal filters, VWR International, Germany). Efficiency of the
enzymatic hydrolysis for each sample was compared to the acid
(
7.0 M urea, 4.3 mM potassium dihydrogen phosphate, 1.4 mM
sodium tetraborate, pH 7.5) held at a constant temperature of 40 °C.
Time counting was stopped automatically when the tendon broke.
TTBT of six tendons per individual were measured for both incubated
and native tendons, respectively.
Preparation and Solubilization of Collagen. Tendons were
washed with purified water three times and then cut into small pieces.
In 2 mL tubes filled with 500 μL of water, tendons were minced with a
zirconium oxide grinding ball (5 mm in diameter) in a mixer mill
(
3
MM400, Retsch, Germany) at 30 Hz for 30 min. After lyophilization,
mg of tendon were solubilized in 1.2 mL of diluted acetic acid (20
mM) at 37 °C and 1000 rpm for 30 min (Thermomixer compact,
Eppendorf, Germany) to give a collagen solution.
SDS-PAGE. 100 μL of above collagen solution were predigested
with 25 units of pepsin at 37 °C for 1 h. The reaction was stopped by
adding 100 μL of SDS solution (160 mM TRIS, 4% (w/v) SDS) and
6
hydrolysis by LC−MS/MS analysis of the acid-stable modification N -
carboxyethyl lysine (CEL).
Preparative Fractionation of Enzymatic Hydrolysates. 150
μL of individual enzymatic hydrolysates were pooled for each age
group. The pooled samples were separated by high-performance liquid
chromatography (HPLC) on an analytical stainless steel column
packed with RP-18 material (Vydac Protein & Peptide C18, 250 mm
× 4.6 mm, 5 μm, Grace, Maryland). The HPLC system (Jasco,
Germany) consisted of a pump (PU-2080 Plus) with a degasser (DG-
2080-54) and a quaternary gradient mixer (LG-2080-04), a column
oven (Jetstream II) and an autosampler (AS-2057 Plus). The mobile
phase used was water (eluent A) and methanol with water (7:3, v/v,
eluent B). 1.2 mL/L of heptafluorobutyric acid was added to both
eluents (A and B) as ion pair reagent. Analyses were performed at a
column temperature of 25 °C using a flow rate of 1.0 mL/min and
gradient elution: 2% B (15 min isocratic), to 70% B (35 min), to 100%
B (5 min), and hold (10 min). 100 μL of pooled samples were
repetitively injected, and eluate from 37 to 52 min was collected. Eight
chromatographic runs were performed in total for each group. The
1
00 μL of nonreducing Laemmli buffer (125 mM TRIS, 20% (w/v)
glycerin, 5% (w/v) SDS, 1 mM EDTA, 0.05% (w/v) bromophenol
blue). 20 μL of this solution (20 μg of protein) were analyzed by
denaturing polyacrylamide electrophoresis using discontinuous acryl-
amide gels (resolving gel, T = 6%, C = 3.3%; stacking gel, T = 5%).
Protein bands were visualized with colloidal Coomassie staining
27
overnight as described by Dyballa.
Acid Hydrolysis. 400 μL of above collagen solution were
evaporated to dryness by vacuum centrifugation (Savant Speed Vac
Concentrator, Thermo Scientific, Germany). 200 μL of sodium
borohydride (8 mg/mL) in sodium hydroxide (0.01 mM) were added
and incubated at room temperature for 1 h. The solution was
evaporated to dryness again. After addition of 800 μL of hydrochloric
acid (6.0 M), the head space was purged with argon and the reaction
tube was sealed and kept in a drying oven at 110 °C for 20 h.
C
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
combined fraction were evaporated by vacuum centrifugation and
freeze-dried. The amorphic material was dissolved in 80 μL of water
and analyzed by LC−MS/MS.
Table 2. Limits of Detection (LOD) and Limits of
Quantitation (LOQ) for AGEs and Lysinonorleucine (LNL)
in Tendon Collagen
Ninhydrin Assay. The amino acid content of acid collagen
hydrolysates was measured by the ninhydrin method of Smuda et al.
[μmol/mol leu-eq]
2
8
with L-leucine as the reference standard. After derivatization,
absorption of the diluted samples and a standard curve was determined
at 566 nm with a microplate reader (infinite M200, Tecan,
Switzerland) using 96-well plates.
analytes
LOD
LOQ
CEL
1.4
3.6
6
N -lactoyl lysine
0.12
0.86
HPLC−MS/MS Analysis. The high-performance liquid chroma-
MG-H1
MG-H3
CEA
10
13
0.23
10.3
35
0.06
tography (HPLC) system (Jasco, Germany) consisted of a pump (PU-
2
0.14
5.2
080 Plus) with a degasser (DG-2080-02) and a quaternary gradient
mixer (LG-2080-04), a column oven (Jetstream II), and an
autosampler (AS-2057 Plus). Mass spectrometric detection was
conducted on an API 4000 QTrap LC−MS/MS system (Applied
Biosystems/AB Sciex, Germany) equipped with a turbo ion spray
source using electrospray ionization in positive ion mode: ion spray
voltage of 2.5 kV, nebulizing gas pressure of 70 psi, heating gas
pressure of 80 psi at 650 °C, and curtain gas pressure of 30 psi.
Chromatographic separations were performed on a stainless steel
column packed with RP-18 material (Xselect HSS T3, 250 mm × 3.0
mm, 5 μm, Waters, Massachusetts) using a flow rate of 0.7 mL/min.
The mobile phase used was water (eluent A) and methanol with water
THP
10
argpyrimidine
CML
0.03
1.1
3.6
2.2
7.3
GALA
0.7
CMA
3.5
G-H3
5
12
glucosepane
DODIC
MODIC
MOLA
LNL
0.21
0.04
0.005
0.10
6
0.52
0.19
0.016
0.21
(
7:3, v/v, eluent B). 1.2 mL/L of heptafluorobutyric acid were added
to both eluents (A and B) as ion pair reagent. Analyses were
10
performed at a column temperature of 25 °C using gradient elution:
2
% B (2 min) to 14% B (10 min) to 87% B (22 min) to 100% B (0.5
cross-linking was a particular focus. Interestingly, for the MGO-
derived cross-link structures MOLD (methylglyoxal lysine
dimer) and MODIC (methylglyoxal-derived imidazoline
min) and hold (7 min). For mass spectrometric detection, the
scheduled multiple-reaction monitoring (sMRM) mode was used,
utilizing collision-induced dissociation (CID) of the protonated
molecules with compound specific orifice potentials and fragment
specific collision energies (Table 1). Appropriate dilutions of the
samples were injected dependent on analyte concentration and matrix
interference. Quantitation was based on the standard addition method.
More precisely, increasing concentrations of authentic reference
compounds at factors of 0.5, 1, and 2 times the concentration of the
compounds in the sample were added to separate aliquots of the
sample. The aliquots were analyzed, and a regression of response
versus concentration was used to determine the concentration of the
analyte in the sample. Calibration with this method resolves potential
matrix interference.
High-Resolution Mass Determination (HR-MS). Positive and
negative ion high-resolution ESI mass spectra were obtained from an
Orbitrap Elite mass spectrometer (Thermofisher Scientific, Germany)
equipped with an HESI electrospray ion source (spray voltage 4 kV;
capillary temperature 275 °C, source heater temperature 40 °C; FTMS
resolution >30 000). Nitrogen was used as sheath and auxiliary gas.
The sample solutions were introduced continuously via a 500 μL
Hamilton syringe pump with a flow rate of 5 μL/min. The data were
evaluated by the Xcalibur software 2.7 SP1.
29
cross-link), there are also homologous compounds GOLD
17
and GODIC derived from glyoxal (GO). Consequently, a
counterpart for the GO-amide cross-link GOLA, which was first
19
described by Glomb and Pfahler in 2001, should be two lysine
molecules linked by MGO via an amine and an amide bond,
respectively. However, this hypothetical structure has not been
described in the literature so far, although Thornalley had
already suggested the compound as a “putative crosslink” in
30
1
994. We thus synthesized the novel AGE cross-link mainly
according to the synthetic route for GOLA by Glomb and
Pfahler using a carbodiimide catalyzed coupling strategy of
protected lysine derivatives. The final authentic reference
material as well as all intermediates were confirmed by one- and
1
13
two-dimensional NMR experiments ( H, C, H,H-COSY,
HSQC, HMBC) and by accurate mass determination. In
analogy to GOLA, we named the new compound MOLA for
methylglyoxal lysine amide. To date, GOLA and MOLA are the
only known glycation cross-links with an amide bond.
Preparation and Solubilization of Tail Tendon
Collagen. Rat tail tendons are highly hierarchical constructs
consisting of almost exclusively type I collagen. A tendon is
built of fiber bundles, whereas fibers are composed of many
fibrils. A fibril is formed from single collagen molecules, called
tropocollagen, which consist of three polypeptide chains
forming a triple helix (∼3000 amino acids) with short
nonhelical ends, called telopeptides. A high content of glycine
(33%), proline (12%), alanine (12%) and 4-hydroxyproline
Nuclear Magnetic Resonance Spectroscopy (NMR). NMR
spectra were recorded on a VXR 400 spectrometer (Varian, California)
1
13
operating at 400 MHz for H and 100 MHz for C or on a Unity
Inova 500 instrument (Varian, California) operating at 500 MHz for
1
13
H and 125 MHz for C, respectively. Tetramethylsilane was used as a
reference for calibrating the chemical shift.
Statistical Analysis. Statistical significance between two adjacent
groups was examined using Fisher’s test (F-test) with a probability
value of 95%. All significance tests were performed by two-sample
Student’s t-test or Welch’s t-test, respectively, with a probability value
of at least 95%. Limits of detection (LOD, 3 × S/N) and limits of
quantitation (LOQ, 10 × S/N) for each analyte were estimated
according to signal/noise ratio (S/N) and are given in Table 2.
(
11%) is characteristic for type I collagen and allows very tight
31
packing of helices and stabilization via hydrogen bonds. Final
mechanical and chemical stability results from enzymatic cross-
linking of collagen molecules. Especially the explicit water
insolubility required the development of an optimized protein
workup protocol. To achieve solubility, tendons were first cut
into small pieces and minced in a mixer mill to disrupt the
fibrillar packing. A fluffy and weighable material resulted after
lyophilization. As a result, the processed collagen was then
soluble in diluted acetic acid (0.1%, pH 3.5, 37 °C, 30 min).
RESULTS AND DISCUSSION
■
Synthesis of MOLA. The present paper aimed to
investigate the impact of nonenzymatic protein modification
on the physical properties of collagen. Therefore, the formation
of Maillard endproducts leading to intra- and intermolecular
D
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
Importantly, neither collagen nor amide AGEs were hydrolyzed
under these conditions. The protein solution was then
aliquoted for both acid and enzymatic hydrolyses. For acid
hydrolysis, a reduction step with sodium borohydride converted
dehydrolysinonorleucine into the stabile lysinonorleucine
times lower than the hydroimidazolones underlining that
arginine guanidino functions are the preferred glycation sites
in collagen. In the α -chain of collagen type I with a total
1
number of 1052 amino acids, there are 52 arginine and 35
31
lysine residues explaining the higher capacity of arginine over
lysine binding sites. CEL was the dominant lysine modification
which is formed nonoxidatively via isomerization from the
(
LNL) and avoided CML artifact formation from the Amadori
product. Enzymatic hydrolysis was carried out for 7 days at 37
C by adding fresh proteases every 24 h. Pepsin (25 U) was
6
°
Schiff base of the keto group of MGO and the N -amino
used in the first digestion step to generate smaller-weight
soluble peptides. Collagenase was added (20 U) for the next 48
h to further breakup the collagen peptide structure. The clear
solution was then subjected to the standard digestion protocol
published by our working group. The complete procedure led
to an average hydrolysis efficiency of 73%, which is comparable
to values obtained for soluble proteins (e.g., bovine serum
albumin).
function of lysine (Figure 2). Interestingly, in contrast to all
other MGO-derived AGEs, native tendons already contained
CEL (0 days, 1.5 μmol/mol leucine-equivalents). The lysine-
arginine cross-link MODIC was formed at the same
concentration levels as CEL until 3 days. After that, CEL
formation slowed down, while MODIC formation proceeded to
28
6
increase linearly. N -lactoyl lysine is the amide analogue of CEL
formed from the Schiff base adduct of the aldehyde group of
6
MOLA in ex Vivo Tendon Incubations. After identifying
MGO. Thus, formation of CEL and N -lactoyl lysine is
2
MOLA in a simple model system comprising of N -t-Boc-lysine
competitive and driven by the reactivity and availability of the
respective carbonyl moiety. In our experiment, CEL levels were
constantly about 10 times higher. This ratio can be explained by
the initial step of Schiff base formation. In line with the
carbonyl reactivity, free MGO (<1%) is in equilibrium with its
and MGO (data not shown), we investigated AGE formation
by MGO in an ex vivo system of 3 months old rat tail tendons.
Tendons were incubated under physiological, deaerated
conditions over a period of 7 days. Formation of MGO-derived
AGEs is shown in Figure 1. As expected, the quantitative
32
mono- (56%) and dihydrate (44%) in aqueous solution.
Thus, the monohydrate form showing a free keto group is
much more available for nucleophilic attack than the aldehyde
group of free MGO. Based on the isomerization mechanism
6
leading to N -lactoyl lysine, we postulate the formation of
MOLA. After formation of the hemiaminal at the aldehyde
group, a second lysine residue attacks the free keto group and
rearrangement leads to MOLA. Indeed, MOLA formation
6
proceeded in parallel to that of CEL and N -lactoyl lysine in our
6
system. MOLA levels were twice as high as N -lactoyl lysine
levels. In accordance with above overall amino acid content,
arginine modification increased in a linear fashion indicating
still free reaction sites while lysine modification slowed down
after 3 days indicating limitations of formation, e.g., de novo
synthesis of MODIC and CEL was almost the same until 3 days
but diverted to 11 and 8 μmol/mol leucine-equivalents after 7
days, respectively. Comparison of MODIC and MOLA
formation suggested that cross-linking in collagen is especially
favored between lysine and arginine side chains in collagen due
to sterical aspects. Gautieri et al. simulated possible cross-
linking sites in collagen fibrils in silico and indeed identified 14
possible lysine-arginine pairs of which six were intra- and eight
3
3
were intermolecular.
Figure 1. AGE formation in rat tail tendons (3 months) incubated
with methylglyoxal (200 μM).
Mechanical Properties of Incubated and Aged
Tendons. To evaluate the effect of MGO-mediated glycation
on the physical characteristics of collagen, one part of each
incubated tendon was subjected to the tail tendon breaking
7
important structures were MG-H1 and N -carboxyethyl
5
26
arginine (CEA). Condensation with the N -guanidino function
time (TTBT) assay as described by Harrison and Archer.
of arginine residues has been reported to be a dominating
Accordingly, TTBT mainly depends on covalent cross-linking
within the collagen fibrils. The results are shown in Figure 3 for
the present tendon incubations. Increase of TTBT from 10 min
for native tendons to 600 min after 7 days correlated with
increasing AGE levels, especially of the cross-link structures
MODIC and MOLA. Several prerequisites were chosen to
ensure that ex vivo incubation conditions were at least related
to conditions found during the natural aging process.
Preliminary experiments confirmed that MGO-glycation
cross-links are formed nonoxidatively. Furthermore, reactive
oxygen species like hydrogen peroxide lead to collagen cross-
6
reaction pathway of MGO. In the first step, MGO condenses
5
with the guanidino group to give the short-lived N -endocyclic
dihydroxyimidazolidine. After elimination of water, MG-H3 is
formed, which is the kinetically controlled product. Hydrolysis
converts to the open-chained intermediate CEA to form the
7
15
thermodynamically controlled N -exocyclic MG-H1. This
reaction course was reflected by the tendon model. The levels
of both MG-H1 and CEA were about 30 times higher than that
of MG-H3. THP and to a much lesser extend argpyrimidine are
secondary products resulting from the reaction of MG-H3 with
another molecule of MGO indicating an excess of MGO in our
system. All other structures were found at concentrations 1000
4
linking even in absence of glycation agents. For that reason, all
incubation solutions were purged with helium to provide
E
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
6
1
Figure 2. Postulated mechanism for the formation of MOLA according to the CEL and N -lactoyl lysine reaction cascade.
17,18
trations.
In the case of MGO for arginine modifications,
high amounts of THP and argpyrimidine and for lysine
modifications specifically the cross-link structure MOLD would
result. In the present ex vivo setup at 200 μM MOLD was never
detected. In parallel to the TTBT assay, gel electrophoresis of
the glycated tendons was conducted (Figure 4). After
predigestion with pepsin for 1 h, soluble polypeptide cleavage
Figure 3. Tail tendon breaking time of rat tail tendons (3 months)
incubated with methylglyoxal (200 μM).
strictly deaerated conditions. Second, only freshly prepared
MGO was used to avoid side reactions of contaminants in
1
5
commercial MGO solutions like formaldehyde. Third,
compared to literature a relatively low MGO concentration of
6
2
00 μM was chosen. In relation to MGO levels of 2 μM
normally found in rat plasma, this is still very high but allowed
to monitor TTBTs comparable to those found in vivo within
the given incubation time frame. Test incubations with MGO
concentrations of 1 mM and 10 mM resulted in extremely long
or not measurable TTBTs. From a mechanistic point of view,
high dicarbonyl levels will shift the AGE spectrum to structures
that are unlikely to form under physiological conditions which
are characterized by low carbonyl but high amine concen-
Figure 4. SDS-PAGE of rat tail tendons (3 months) incubated with
methylglyoxal (200 μM) and native rat tail tendons of different ages.
F
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
a
fragments with a molecular weight of around 116 kDa and over
Table 3. AGE and Lysinonorleucine (LNL) Levels in Native
Rat Tail Tendon Collagen of Different Age
200 kDa increased with incubation time up to 5 days indicating
a progress of covalent cross-linking. At 7 days these bands faded
out, as the highly glycated collagen became less digestible and
major portions could no longer enter the gel.
The TTBT assay was then performed on native tendons of
different ages (3, 12, and 22 months, Figure 5). TTBT
age [months]
analyte [μmol/mol
leu-eq]
3
12
22
MGO-Derived
CEL
1.5 ± 0.2
0.20 ± 0.03
19 ± 1
0.57 ± 0.04
7.2 ± 0.6
1.6 ± 0.2
0.32 ± 0.07**
21 ± 2**
0.65 ± 0.04**
7.4 ± 0.5
1.8 ± 0.1*
0.5 ± 0.1*
25 ± 1***
0.75 ± 0.03***
8.0 ± 0.3*
6
N -lactoyl lysine
MG-H1
MG-H3
CEA
GO-Derived
CML
4.7 ± 0.5
0.6 ± 0.1
3.1 ± 0.5
21 ± 2
6.9 ± 0.7**
0.7 ± 0.1
6 ± 1***
23 ± 2
13 ± 1**
1.1 ± 0.4*
8 ± 1*
GALA
CMA
b
G-H3
21 ± 2
c
Glycation Cross-Links
glucosepane
DODIC
n. d.
n. d.
n. d.
n. d.
<LOQ
<LOQ
<LOQ
<LOQ
0.72
0.24
0.04
0.34
Figure 5. Tail tendon breaking time of native rat tail tendons of
different age.
MODIC
MOLA
increased from 10 min of young, to 80 min for adult, to 190
min for old individuals (r = 0.986, P ≤ 0.001). These findings
clearly confirmed that cross-linking proceeds with age.
However, in vivo enzymatic cross-linking of collagen occurs
in parallel to nonenzymatic processes. In immature collagen,
lysyl oxidase (LOX)-mediated cross-links are labile aldimines,
ketoamines, or aldol condensation products. During maturation
of collagen, these bivalent structures are slowly converted into2
tri- and tetravalent cross-links between the collagen fibrils.
Interestingly, the total number of enzymatic cross-links remains
constant over time due to a decreased LOX activity and a
d
Enzymatic Cross-Link
LNL
32 ± 3
22 ± 1**
20 ± 2*
a
AGE levels are given as mean values ± standard deviation; statistical
significance between adjacent groups: *P ≤ 0.05, **P ≤ 0.01, ***P ≤
b
0
.001. Sum parameter for dihydroxyimidazolidine and carboxyme-
c
thylarginine (CMA) after acid hydrolysis. After preparative enrich-
d
ment (n. d., not detectable). After reduction and acid hydrolysis.
34
34
limited number of lysine residues in the telopeptide region.
As a probe for these processes, we determined lysinonorleucine
LNL) after reduction in acidic hydrolysates. LNL is the
properties like molecular packing but also biological function.
The MGO-lysine modifications CEL and N -lactoyl lysine were
6
(
also identified in tendon collagen at lower concentrations up to
2.3 μmol/mol leucine-equivalent with a low but still significant
increase in old tendons. Unlike to ex vivo experiments, a wide
array of other AGEs was detected due to the occurrence of
many additional 1,2-dicarbonyls in vivo. A major group were
GO-derived lysine and arginine modifications. CML was the
most prominent lysine modification reaching 13 μmol/mol
reduced form of the immature aldimine cross-link dehydroly-
sinonorleucine. In compliance to literature, LNL decreased with
age from 32 to 20 μmol/mol leucine-equivalents (Table 3). In
contrast to LOX-modified lysine residues located in the
telopeptides, lysine side chains within the triple helix remain
susceptible to glycation during aging leading to accumulation of
36
2
1
nonenzymatic cross-link structures. Consequently, AGE cross-
leucine-equivalents due to its multiple formation pathways. In
links will significantly contribute to the increased TTBT in aged
tendons. SDS-PAGE of aged tendons revealed a similar cross-
linking pattern as in MGO-glycated tendons (Figure 4),
however, to a much lesser extent.
addition to GO-isomerization reactions, the predominant
pathway in vivo is the oxidative fragmentation of fructosyl
lysine. In contrary, GALA is a GO-specific alternative lysine
modification from the CML isomerization cascade. GALA was
about 10 times lower than CML which is not according to the
ratio found for the pendants N -lactoyl lysine and CEL. CML
levels were 7-fold higher than CEL in old individuals. This
Quantitation of MOLA and Other AGEs in Vivo. To
elucidate AGE candidates responsible for collagen cross-linking
during aging, native rat tendons of different age were processed
via the optimized workup protocol. Results of AGE
quantitation are summarized in Table 3. In line with above
ex vivo MGO-incubation, MG-H1 was the major modification
in young, adult, and old tail tendon collagen. Similar findings
6
again reflects that major portions of CML must originate from
7
other sources, specifically the Amadori product. N -Carbox-
ymethyl arginine (CMA) is a GO-specific arginine modifica-
21
tion. In acidic tendon hydrolysates, imidazolinone G-H3 was
quantitated to estimate the total amount of arginine-bound GO.
G-H3 is formed from both CMA and the dihydroxyimidazo-
lidine under the harshly acidic conditions used for total protein
3
5
were made in different extracellular matrix proteins. In our
study, MG-H1, CEA, and MG-H3 increased significantly with
age totaling up to 34 μmol/mol leucine-equivalents. The ratio
between these structures reflected again the above-described
mechanistic relationship as in the ex vivo incubations. Although
CEA and the hydroimidazolones are no cross-linking structures,
monovalent modification of arginine residues will change the
isoelectric point of collagen molecules affecting physical
21
hydrolysis. Thus, G-H3 can be regarded as a sum parameter
for both structures. Interestingly, CMA increased significantly
with age while G-H3 remained unchanged. This suggests that
(I) most glyoxal is bound as dihydroxyimidazolidine at young
age and converts slowly into CMA over time and (II) that total
G
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
GO-arginine modification does not accumulate in collagen with
age.
ACKNOWLEDGMENTS
■
We would like to thank Dr. D. Stro
̈
hl from the Institute of
In the original workup of both acidic and enzymatic
hydrolysates no AGE cross-link structures were identified. For
that reason, enzymatic hydrolysates of each age group were
pooled and repetitively fractionated by analytical HPLC on a
quantitative basis. The 10-fold concentrated samples were
analyzed again and four glycation cross-links were quantitated.
Glucosepane was the dominant structure in old tendons with
Organic Chemistry, Martin-Luther-University Halle-Witten-
berg, Halle/Saale, Germany, for recording NMR spectra and
A. Laub from the Leibniz Institute of Plant Biochemistry,
Halle/Saale, Germany, for performing accurate mass determi-
nation.
ABBREVIATIONS USED
■
0
.72 μmol/mol leucine-equivalents which is in support of the
7
5
,8
AGE, advanced glycation endproduct; CEA, N -carboxyethyl
literature. Considering the amino acid composition of type I
collagen, this can be estimated as one glucosepane molecule per
6
7
arginine; CEL, N -carboxyethyl lysine; CMA, N -carboxymethyl
6
arginine; CML, N -carboxymethyl lysine; DODIC, 3-deoxy-
600 collagen molecules. This lysine-arginine cross-link is
6
glucosone-derived imidazoline cross-link; GALA, N -glycoloyl
formed from the Lederer’s glucosone resulting from degrada-
tion of the Amadori product. Amounts of MOLA and DODIC
were about half and one-third of glucosepane concentrations,
respectively. Surprisingly, MODIC, which was the dominant
structure in the ex vivo incubations, was only formed in traces.
These findings indicate in line with the above-discussed
conformational conditions in type I collagen that covalent
cross-linking of lysine and arginine residues was obviously
favored over lysine−lysine cross-linking. Second, cross-linking
was mainly mediated by the Amadori product and its
dicarbonyl follow-up structures Lederer’s glucosone and 3-
deoxyglucosone. Nevertheless, MOLA was the only quantita-
tively relevant lysine−lysine cross-link derived from MGO
found in old collagen. With a concentration of 0.34 μmol/mol
leucine-equivalents this calculates to approximately one MOLA
molecule per 1200 collagen molecules. In comparison, levels
found for CEL and MG-H1 correspond to one molecule per
lysine (glycolic acid lysine amide); G-H3, glyoxal hydro-
imidazolone; GODIC, glyoxal-derived imidazoline cross-link;
GOLA, glyoxal lysine amide; GOLD, glyoxal lysine dimer; leu-
eq, leucine-equivalents; LNL, lysinonorleucine; MG-H1/3,
methylglyoxal hydroimidazolone 1/3; MGO, methylglyoxal;
MODIC, methylglyoxal-derived imidazoline cross-link; MOLA,
methylglyoxal lysine amide; MOLD, methylglyoxal lysine
dimer; MW, molecular weight; n. d., not detectable; PBS,
phosphate buffered saline; THP, tetrahydropyrimidine; TTBT,
tail tendon breaking time
REFERENCES
■
(
1) Henning, C.; Glomb, M. A. Pathways of the Maillard Reaction
under Physiological Conditions. Glycoconjugate J. 2016, 33, 499−512.
(
2) Bailey, A. Molecular Mechanisms of Ageing in Connective
Tissues. Mech. Ageing Dev. 2001, 122, 735−755.
3) Gelse, K. Collagens - Structure, Function and Biosynthesis. Adv.
(
250 and 20 collagen molecules, respectively. Other AGE cross-
Drug Delivery Rev. 2003, 55, 1531−1546.
links, more precisely GOLA, GOLD, GODIC, MOLD, and
pentosidine were also analyzed but were not detected.
(4) Monnier, V. M.; Glomb, M.; Elgawish, A.; Sell, D. R. The
Mechanism of Collagen Crosslinking in Diabetes. A Puzzle Nearing
Resolution,. Diabetes 1996, 45, S67−S72.
In conclusion, collagen is proceedingly modified by glycation
during aging. For the first time, based on an optimized protein
workup protocol rat tail tendons were comprehensively
analyzed by LC−MS/MS for monovalent and cross-link
modifications. MGO was verified as the major glycating agent
leading up to 36 μmol/mol leucine-equivalents of mainly
arginine directed monovalent modifications. MOLA, a novel
lysine−lysine amide cross-link derived from methylglyoxal, was
established for the first time as a quantitatively relevant cross-
linking structure in senescent collagen. Chemical cross-linking
was also visualized by SDS-PAGE and correlated with increase
in collagen stiffening measured as TTBT.
(5) Monnier, V. M.; Mustata, G. T.; Biemel, K. L.; Reihl, O.; Lederer,
M. O.; Zhenyu, D.; Sell, D. R. Crosslinking of the Extracellular Matrix
by the Maillard Reaction in Aging and Diabetes. An Update on ″A
Puzzle Nearing Resolution″,. Ann. N. Y. Acad. Sci. 2005, 1043, 533−
5
44.
(6) Fessel, G.; Li, Y.; Diederich, V.; Guizar-Sicairos, M.; Schneider,
P.; Sell, D. R.; Monnier, V. M.; Snedeker, J. G.; Screen; Hazel, R. C.
Advanced Glycation Endproducts Reduce Collagen Molecular Sliding
to Affect Collagen Fibril Damage Mechanisms but Not Stiffness. PLoS
One 2014, 9, e110948.
(7) Skovgaard, D.; Svensson, R. B.; Scheijen, J.; Eliasson, P.;
Mogensen, P.; Hag, A. M. F.; Kjaer, M.; Schalkwijk, C. G.; Schjerling,
P.; Magnusson, S. P.; Couppe, C. An Advanced Glycation Endproduct
(
AGE)-Rich Diet Promotes Accumulation of AGEs in Achilles
Tendon. Physiol. Rep. 2017, 5, e13215.
8) Sell, D. R.; Biemel, K. M.; Reihl, O.; Lederer, M. O.; Strauch, C.
AUTHOR INFORMATION
■
(
Corresponding Author
M.; Monnier, V. M. Glucosepane is a Major Protein Crosslink of the
*
E-mail: marcus.glomb@chemie.uni-halle.de. Phone: +49-345-
525 784. Fax: +49-345-5527 341.
Senescent Human Extracellular Matrix. J. Biol. Chem. 2005, 280,
1
(
2310−12315.
5
9) Thornalley, P. J. Pharmacology of Methylglyoxal: Formation,
ORCID
Modification of Proteins and Nucleic Acids, and Enzymatic
Tobias Jost: 0000-0001-7871-1345
Marcus A. Glomb: 0000-0001-8826-0808
Detoxification. Gen. Pharmacol. 1996, 27, 565−573.
(
10) Kalapos, M. P. Where does plasma methylglyoxal originate
from? Diabetes Res. Clin. Pract. 2013, 99, 260−271.
Funding
(11) Nagaraj, R. H.; Sarkar, P.; Mally, A.; Biemel, K. M.; Lederer, M.
Funding was supported in part by the Deutsche Forschungsge-
meinschaft (DFG, Germany) Research Training Group 2155,
ProMoAge.
O.; Padayatti, P. S. Effect of pyridoxamine on chemical modification of
proteins by carbonyls in diabetic rats: characterization of a major
product from the reaction of pyridoxamine and methylglyoxal. Arch.
Biochem. Biophys. 2002, 402, 110−119.
Notes
(12) Rabbani, N.; Thornalley, P. J. Methylglyoxal, glyoxalase 1 and
The authors declare no competing financial interest.
the dicarbonyl proteome. Amino Acids 2012, 42, 1133−1142.
H
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
(
13) Ahmed, M. U.; Frye, E. B.; Degenhardt, T. P.; Thorpe, S. R.;
Liquid Chromatographic Fluorimetric Assay of Methylglyoxal. Anal.
Chim. Acta 1992, 263, 137−142.
(33) Gautieri, A.; Redaelli, A.; Buehler, M. J.; Vesentini, S. Age- and
Diabetes-related Non-enzymatic Crosslinks in Collagen Fibrils:
Candidate Amino Acids Involved in Advanced Glycation Endproducts.
Matrix Biol. 2014, 34, 89−95.
(34) Avery, N. C.; Bailey, A. J. The Effects of the Maillard Reaction
on the Physical Properties and Cell Interactions of Collagen,. Pathol.
Biol. 2006, 54, 387−395.
(35) Ahmed, N.; Thornalley, P. J.; Dawczynski, J.; Franke, S.; Strobel,
J.; Stein, G.; Haik, G. M. Methylglyoxal-Derived Hydroimidazolone
Advanced Glycation End-Products of Human Lens Proteins, Invest.
Ophthalmol. Invest. Ophthalmol. Visual Sci. 2003, 44, 5287−5292.
(36) Henning, C.; Liehr, K.; Girndt, M.; Ulrich, C.; Glomb, M. A.
Extending the Spectrum of α-Dicarbonyl Compounds in vivo. J. Biol.
Chem. 2014, 289, 28676−28688.
e
Baynes, J. W. N -Carboxyethyl Lysine, a Product of the Chemical
Modification of Proteins by Methylglyoxal, Increases with Age in
Human Lens Proteins. Biochem. J. 1997, 324, 565−570.
(
14) Smuda, M.; Voigt, M.; Glomb, M. A. Degradation of 1-deoxy-D-
erythro-hexo-2,3-diulose in the presence of lysine leads to formation of
carboxylic acid amides. J. Agric. Food Chem. 2010, 58, 6458−6464.
(
̈
15) Klopfer, A.; Spanneberg, R.; Glomb, M. A. Formation of
α
Arginine Modifications in a Model System of N -tert-Butoxycarbonyl
Boc)-Arginine with Methylglyoxal. J. Agric. Food Chem. 2011, 59,
94−401.
16) Nagaraj, R. H.; Shipanova, I. N.; Faust, F. M. Protein
(
3
(
Crosslinking by the Maillard Reaction: Isolation, Characterization
and in vivo Detection of a Lysine-Lysine Crosslink Derived from
Methylglyoxal. J. Biol. Chem. 1996, 271, 19338−19345.
(
17) Lederer, M. O.; Klaiber, R. G. Crosslinking of Proteins by
Maillard Processes. Characterization and Detection of Lysine-Arginine
Crosslinks Derived from Glyoxal and Methylglyoxal. Bioorg. Med.
Chem. 1999, 7, 2499−2507.
(
18) 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.
(
19) Glomb, M. A.; Pfahler, C. Amides Are Novel Protein
Modifications Formed by Physiological Sugars. J. Biol. Chem. 2001,
76, 41638−41647.
20) Odani, H.; Iijima, K.; Nakata, M.; Miyata, S.; Yasuda, Y.; Irie, S.;
2
(
ω
Maeda, K.; Fujimoto, D. Identification of N -Carboxymethyl Arginine,
as a new Advanced Glycation Endproduct in Serum Proteins of
Diabetic Patients. Int. Congr. Ser. 2002, 1245, 295−301.
(
21) Glomb, M. A.; Lang, G. Isolation and Characterization of
Glyoxal−Arginine Modifications. J. Agric. Food Chem. 2001, 49, 1493−
501.
22) Shipanova, I. N.; Glomb, M. A.; Nagaraj, R. H. Protein
1
(
Modification by Methylglyoxal. Chemical Nature and Synthetic
Mechanism of a Major Fluorescent Adduct. Arch. Biochem. Biophys.
1
(
997, 344, 29−36.
23) Lederer, M. O.; Buhler, H. P. Crosslinking of Proteins by
̈
Maillard Processes. Characterization and Detection of a Lysine-
Arginine Crosslink Derived from D-Glucose. Bioorg. Med. Chem. 1999,
7
(
, 1081−1088.
24) Biemel, K. M.; Reihl, O.; Conrad, J.; Lederer, M. O. Formation
Pathways for Lysine-Arginine Crosslinks Derived from Hexoses and
Pentoses by Maillard Processes. Unraveling the Structure of a
Pentosidine Precursor. J. Biol. Chem. 2001, 276, 23405−23412.
(
25) Strazzolini, P.; Melloni, T.; Giumanini, A. G. Selective Nitrolytic
Deprotection of N-Boc-Amines and N-Boc-Amino Acids Derivatives.
Tetrahedron 2001, 57, 9033−9043.
(
26) Harrison, D. E.; Archer, J. R. Measurement of Changes in
Mouse Tail Collagen with Age. Temperature dependence and
procedural details. Exp. Gerontol. 1978, 13, 75−82.
(
27) Dyballa, N.; Metzger, S. Fast and Sensitive Colloidal Coomassie
G-250 Staining for Proteins in Polyacrylamide Gels. J. Visualized Exp.
009, DOI: 10.3791/1431.
28) Smuda, M.; Henning, C.; Raghavan, C. T.; Johar, K.; Vasavada,
2
(
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.
(
29) Chellan, P.; Nagaraj, R. H. Protein Crosslinking by the Maillard
Reaction and diabetes. Dicarbonyl-derived Imidazolium Crosslinks in
Aging. Arch. Biochem. Biophys. 1999, 368, 98−104.
(
30) Thornalley, P. J.; Westwood, M. E.; McLellan, A. C.; Lo, T. W.
Binding and Modification of Proteins by Methylglyoxal under
Physiological Conditions. J. Biol. Chem. 1994, 51, 32299−32305.
(
31) Hulmes, D. J.; Miller, A.; Parry, D. A.; Piez, K. A.; Woodhead-
Galloway, J. Analysis of the primary structure of collagen for the
origins of molecular packing. J. Mol. Biol. 1973, 79, 137−148.
(
32) McLellan, A. C.; Thornalley, P. J. Synthesis and Chromatog-
raphy of 1,2-Diamino-4,5-dimethoxybenzene, 6,7-Dimethoxy-2-meth-
ylquinoxaline and 6,7-Dimethoxy-2,3-dimethylquinoxaline for Use in a
I
DOI: 10.1021/acs.jafc.8b00937
J. Agric. Food Chem. XXXX, XXX, XXX−XXX