Influence of Nucleophilic Amino Acids on Enzymatic Browning Systems

Article abstract of DOI:10.1021/acs.jafc.8b06458

In the present study the enzymatic oxidation of gallic acid and catechin catalyzed by nashi pear polyphenol oxidase (PPO) in the presence of the amino acids lysine, arginine, or cysteine was investigated for polyphenol-amino acid adducts. HPLC analyses revealed the formation of two novel dihydrobenzothiazine carboxylic acid derivatives (8-(3',4'-dihydro-2H-chromene-3',5',7'-triol)-3,4-dihydro-5-hydroxy-2H-benzothiazine-3-carboxylic acid and 7-(3',4'-dihydro-2H-chromene-3',5',7'-triol)-3,4-dihydro-5-hydroxy-2H-benzothiazine-3-carboxylic acid) from 2′-cysteinyl catechin and 5′-cysteinyl catechin in cysteine incubations, respectively. In contrast, arginine and lysine did not lead to any amino acid adducts. Target compounds were separated by high-performance countercurrent chromatography and preparative HPLC and unequivocally characterized by mass spectrometry and nuclear magnetic resonance spectroscopy. Mechanistic incubations starting from the catechin-cysteine adducts showed that both catechin and PPO are crucial components in the formation of the dihydrobenzothiazines. The cysteine incubations showed a red-brown coloration, which coincided with formation and degradation of the dihydrobenzothiazines finally leading to the formation of high-polymeric melanins. Therefore, these compounds might be the key intermediates to understand development of color during cysteine-driven enzymatic browning reactions.

Full text of DOI:10.1021/acs.jafc.8b06458

Article
pubs.acs.org/JAFC
Cite This: J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Influence of Nucleophilic Amino Acids on Enzymatic Browning
Systems
Nils Mertens, Franziska Mai, and Marcus A. Glomb*
Institute of Chemistry, Food Chemistry, Martin-Luther-University Halle-Wittenberg, Kurt-Mothes-Strasse 2, 06120 Halle/Saale,
Germany
S
* Supporting Information
ABSTRACT: In the present study the enzymatic oxidation of gallic acid and catechin catalyzed by nashi pear polyphenol
oxidase (PPO) in the presence of the amino acids lysine, arginine, or cysteine was investigated for polyphenol−amino acid
adducts. HPLC analyses revealed the formation of two novel dihydrobenzothiazine carboxylic acid derivatives (8-(3’,4’-dihydro-
2H-chromene-3’,5’,7’-triol)-3,4-dihydro-5-hydroxy-2H-benzothiazine-3-carboxylic acid and 7-(3’,4’-dihydro-2H-chromene-
3’,5’,7’-triol)-3,4-dihydro-5-hydroxy-2H-benzothiazine-3-carboxylic acid) from 2′-cysteinyl catechin and 5′-cysteinyl catechin
in cysteine incubations, respectively. In contrast, arginine and lysine did not lead to any amino acid adducts. Target compounds
were separated by high-performance countercurrent chromatography and preparative HPLC and unequivocally characterized by
mass spectrometry and nuclear magnetic resonance spectroscopy. Mechanistic incubations starting from the catechin−cysteine
adducts showed that both catechin and PPO are crucial components in the formation of the dihydrobenzothiazines. The
cysteine incubations showed a red-brown coloration, which coincided with formation and degradation of the
dihydrobenzothiazines finally leading to the formation of high-polymeric melanins. Therefore, these compounds might be
the key intermediates to understand development of color during cysteine-driven enzymatic browning reactions.
KEYWORDS: enzymatic browning, polyphenol−amino acid adducts, high-performance countercurrent chromatography,
color-dilution analysis, dihydrobenzothiazine
enzymatic oxidation of chlorogenic acid in the presence of Nα-
(tert-butyloxycarbonyl)lysine and N-acetylcysteine showed
covalent interactions between the oxidized phenolic compound
and the two amino acid derivatives, though, again, no specific
structures were isolated or elucidated.5 Investigations and
structural analyses by Yabuta et al. on the greening reaction of
caffeic esters with primary amino acids gave additional
evidence for the formation of a highly reactive benzoacridine
derivative.6 Thiol compounds such as cysteine are known to
inhibit enzymatic browning reactions by inhibiting PPOs or by
forming colorless adducts with several phenolic substrates such
as catechol,7 caffeic acid,8 and chlorogenic acid and
catechin.2,9−11 Reactions of lysine and tryptophan side chains
of myoglobin with different phenolic compounds were also
discussed.12−14 Under oxidative conditions heated solutions of
catechin with glycine resulted in the formation of various
adducts, which were tentatively characterized by mass
spectrometry.15 Thus, the specific interaction between o-
quinones and amino acids during enzymatic browning with
respect to discrete structures of resulting polyphenol−amino
acid adducts, the underlying mechanisms, and the impact on
formation of colored compounds remains barely investigated.
Therefore, the aim of the present study was to
mechanistically reinvestigate polyphenol−amino acid reactions
based on a model system for enzymatic browning reactions
INTRODUCTION
■
Texture and appearance are the two crucial attributes
determining the quality and the acceptance of fruits and
vegetables for consumers. During postharvest handling,
processing or storage tissue damage leads to discoloration
and texture changes in plant based foods. This is commonly
referred to as the enzymatic browning reaction.1 Enzymatic
browning is a complex process where enzymatic and non-
enzymatic reactions intertwine. During these reactions
phenolic compounds are oxidized by polyphenol oxidases
(PPOs) to highly reactive electrophilic o-quinones, which
subsequently react non-enzymatically with different nucleo-
philes such as other polyphenols or, supposingly, the
nucleophilic side chains of proteins or amino acids. The
formation of polyphenol−amino acid adducts is thought to
affect the color and aroma of foods.2
Spectrophotometric investigations of different enzymatically
oxidized phenolic compounds by Mason and Peterson implied
that the formed o-quinones might react with thiol-, N-terminal
and aliphatic amino groups of amino acids, whereas amide,
guanidine, and seryl hydroxyl groups showed no reactions.3
The oxidation of caffeic acid and chlorogenic acid in the
presence of different amino acids carried out by Pierpoint led
to the formation of colored products. These studies indicated
that the primary reactions between amino acids and o-
quinones involve the thiol group of cysteine, the ε-amino
group of lysine, and the α-amino groups of other amino acids.
However, the experimental approach was only manometrical
and spectrophotometrical, and no reaction products were
structurally characterized.4 Mass spectrometric studies on the
Received: November 21, 2018
Revised: January 11, 2019
Accepted: January 14, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
min; hump, tR = 68−78 min). For qualitative peak assignment the
eluent was directly coupled to the mass spectrometer. The mass
analyses were performed using an Applied Biosystems API 4000
quadrupole instrument (Applied Biosystems, Foster City, CA, USA).
MS ionization was achieved using the turbospray ionization source
operated in positive-ion mode. The settings were as follows: curtain
gas (N2) at 30 psi, ion source gas 1 at 50 psi, ion source gas 2 at 60
psi, source temperature at 550 °C, declustering potential at 40 V,
entrance potential at 10 V, and ion spray voltage at 4500 V. A scan
range between m/z 100 and 700 was used for the Q1 mode (PGA, m/
z 265 [M + H]+; TA, m/z 429 [M + H]+; A1, m/z 410 [M + H]+; A2,
m/z 410 [M + H]+; B1, m/z 392 [M + H]+; B2, m/z 392 [M + H]+).
For quantitation with HPLC−diode array detection (DAD) an
external calibration based on standard solutions of authentic isolated
references dissolved in water was used. The chromatographic hump
was quantified in reference to the calibration curve of catechin.
High-Performance Countercurrent Chromatography. The
HPCCC system (HPCCC Spectrum Series 20, Dynamic Extractions,
Tredegar, Wales) was equipped with a Jasco PU-4180 RHPLC pump
and a Jasco ultraviolet (UV) detector (UV-2075 Plus, Jasco, Gross-
Umstadt, Germany) operating at 290 nm. Eluted liquids were
collected in fractions of 15 mL with a fraction collector (LKB Ultrorac
7000, LKB-Producter AB, Stockholm-Bromma, Sweden). Chromato-
using nashi pear homogenate as PPO source and gallic acid
and catechin as substrates in the presence of amino acids. We
were able to verify two novel dihydrobenzothiazine derivatives
from isolated 2′-cysteinyl catechin and 5′-cysteinyl catechin.
Their importance as key intermediates in cysteine-driven
enzymatic browning reactions was established by color-dilution
analysis.
MATERIALS AND METHODS
■
Chemicals. All chemicals of the highest quality available were
provided by Sigma-Aldrich (Munich/Steinheim, Germany), Fluka
(Taufkirchen, Germany), Merck (Darmstadt, Germany), VWR
Chemicals (Darmstadt, Germany), ACROS Organics (Geel, Bel-
gium), and ARMAR Chemicals (Leipzig, Germany) unless otherwise
indicated. For all experiments ultrapure water was used.
Purpurogallin-4-carboxylic acid (PGA) and theaflavic acid (TA)
were isolated from incubations of gallic acid (GA) and (+)-catechin
(CA) and nashi pear polyphenol oxidase (nashi-PPO) using high-
performance countercurrent chromatography (HPCCC) and prepa-
rative HPLC. Verification of the isolated material was carried out by
NMR spectroscopy and was in line with the literature;17,18 high-
resolution mass determination (HR-MS): PGA, m/z 263.0196
(found), m/z 263.0192 (calcd for C12H7O7 [M − H]−); TA, m/z
427.0666 (found), m/z 427.0665 (calcd for C21H15O10 [M − H]−).
2′-Cysteinyl catechin (A1), 5′-cysteinyl catechin (A2), 8-(3,4-
dihydro-2H-chromene-3,5,7-triol)-3,4-dihydro-5-hydroxy-2H-benzo-
thiazine-3-carboxylic acid (B1), and 7-(3,4-dihydro-2H-chromene-
3,5,7-triol)-3,4-dihydro-5-hydroxy-2H-benzothiazine-3-carboxylic acid
(B2) were isolated from incubations of nashi-PPO, catechin, and
cysteine using HPCCC and preparative HPLC. Verification was
carried out by NMR spectroscopy, and data of A1 and A2 were in line
with literature;9 HR-MS: A1, m/z 408.0753 (found), m/z 408.0753
(calcd for C18H18O8NS [M − H]−); A2, m/z 408.0755 (found), m/z
408.0753 (calcd for C18H18O8NS [M − H]−).
Nashi Pear Homogenate (nashi-PPO). Nashi pears (Pyrus
pyrifolia) were purchased from local markets in Germany. The fruits
were peeled, and 500 g were homogenized in 500 mL of phosphate
buffer (0.2 M, pH 6.5) and filtrated over a Buchner funnel. The
homogenate was stored at −20 °C. The PPO activity was assayed
according to Mai and Glomb,16 however, with one unit activity
defined as 0.1 absorbance unit change per minute at 420 nm, and was
determined as 660 U/mg of protein.
Polyphenol−Amino Acid Model Incubation. Mixtures con-
taining GA (0.1 mM), CA (0.1 mM) in nashi-PPO, and either in the
presence of one amino acid (0.1 mM; lysine, arginine, cysteine) or no
amino acid were incubated in falcon tubes at room temperature and
pH 6.5. Aliquots of 1 mL were collected each hour. The enzymatic
reaction was stopped by mixing the aliquots with 1 mL of stopping
solution containing NaF (8 mM). Each sample was instantly stored at
−20 °C until further analyses. For preparative isolation 10 mL
aliquots were mixed with 10 mL of stopping solution after formation
of color and freeze-dried.
Analytical HPLC−Diode Array Detection−Mass Spectrome-
try. A Jasco PU-2080 Plus quaternary gradient pump with degasser
(DG-2080−54), quaternary gradient mixer (LG 2080−02), multi-
wavelength detector (MD-2015 Plus) (Jasco, Gross-Umstadt,
Germany), Waters 717 plus autosampler, and column oven (Techlab
Jet Stream np K-3) was used. Chromatographic separations were
performed on stainless steel columns (Vydac CRT, 201TP54, 250 ×
4.6 mm, RP-18, 5 μm, Hesperia, CA, USA) using a flow rate of 1.0
mL/min. The mobile phase consisted of water (solvent A) and
MeOH/water (7:3, v/v, solvent B). A 0.8 mL/L aliquot of formic acid
was added to both solvents (A and B). Samples were analyzed using a
gradient system: samples were injected at 10% B and held for 10 min.
The gradient was changed linearly to 65% B in 55 min and to 100% B
after 5 min and held at 100% B for 10 min. The column temperature
was always 25 °C. The effluent was monitored at 220, 280, and 400
nm (GA, tR = 5 min; CA, tR = 20 min; PGA, tR = 49 min; TA, tR = 52
min; A1, tR = 12 min; A2, tR = 14 min; B1, tR = 32 min; B2, tR = 42
grams were recorded on a plotter (Servogor 200, GMC, Nurnberg,
̈
Germany). The total volume of the coil was 140 mL. The HPCCC
was run at a revolution speed of 1600 rpm and a flow rate of 6 mL/
min of the upper organic phase in tail-to-head modus. Samples of 1 g
of freeze-dried incubations were dissolved in a 1:1 mixture of light
(organic) and heavy (aqueous) phase (6 mL) and injected via a 6 mL
sample loop. Incubations were separated using ethyl acetate/water
(1:1, v/v) (PGA, tR = 5−10 min; TA, tR = 5−10 min; CA, tR = 10−15
min; GA, tR = 17.5−35 min; B1, tR = 5−30 min; B2, tR = 5−30 min;
A1 and A2 were collected from the retained stationary phase). The
solvent of the collected fractions was instantly evaporated under
reduced pressure at 30 °C and freeze-dried. The dried residues were
used for preparative HPLC−UV.
Preparative HPLC−UV. A Besta pump (Besta, Wilhelmsfeld,
Germany) was used at a flow rate of 16 mL/min with a Knauer
degasser (Knauer, Berlin, Germany) and a Jasco UV detector (UV-
2075 Plus, Jasco, Gross-Umstadt, Germany) set at 280 nm. Up to 20
mg of the dried collected fractions of the HPCCC was injected.
Chromatographic separations were performed on a stainless steel
column (Vydac CRT, 218TP1022, 250 × 20 mm, RP-18, 10 μm).
The mobile phases used were solvents A and B identical to the
HPLC−DAD system. An isocratic method was chosen according to
the characteristics of the target compounds (PGA, 40% B, tR = 49−60
min; TA, 40% B, tR = 61−85 min; A1, 5% B, tR = 53−71 min; A2, 5%
B, tR = 72−100 min; B1, 30% B, tR = 22−32 min; B2, 30% B, tR = 60−
86 min). The solvents of the collected fractions were instantly
removed under reduced pressure at 30 °C, freeze-dried, and stored at
−20 °C.
High-Resolution Mass Determination. Negative ion high-
resolution ESI mass spectra were obtained from an Orbitrap Elite
mass spectrometer (Thermofisher Scientific, Bremen, 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−1. The
data were evaluated by the Xcalibur software 2.7 SP1.
Nuclear Magnetic Resonance Spectroscopy. NMR spectra
were recorded on a Varian Unity Inova 500 instrument operating at
500 MHz for 1H and 125 MHz for 13C. SiMe4 was used as a reference
for calibrating the chemical shift.
Color-Dilution Analysis. An aliquot (50 μL) of the incubations
at 3 and 8 h was analyzed by HPLC−DAD for determination of the
color-dilution factor (CD factor). Collected colored fractions of 1 mL
were then stepwise 1:1 diluted with solvent B. The test was carried
out in daylight against a white background. Each dilution was visually
monitored until the color difference between the diluted fraction and
B
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
two blanks of solvent B could just be visually detected by five different
panelists using a triangle test. This dilution was defined as the CD
factor. For assignment of the total color-dilution factor (CDtotal
factor), 50 μL of the incubation was diluted to 1 mL and treated as
described above. The contribution of a specific structure/fraction to
CD
CDtotal
the total color was calculated as follows:
× 100%.19
Statistical Analysis. Analyses were performed at least in triplicate
and resulted in coefficients of variation less than 10%.
RESULTS AND DISCUSSION
■
Polyphenol Amino Acid Incubations and Color-
Dilution Analysis. The general incubation system was chosen
based on the black tea fermentation. Therefore, gallic acid and
catechin were used as substrates because of the relatively high
amounts of catechins, catechin gallates, and gallocatechins in
fresh tea leaves.20 After mixing gallic acid and catechin with
nashi pear homogenate (nashi-PPO) the enzymatic oxidation
started immediately developing a yellow-brown discoloration.
Nashi pear was used as an economical enzyme source, because
of its high polyphenol oxidase activity but low polyphenol
content.21 In the first hours of incubation purpurogallin
carboxylic acid and theaflavic acid, known reaction products
from the fermentation of black tea,22 were formed instantly
(Figure 1A). After 3 h PGA was completely degraded and TA
explained 25% of the overall color as calculated from color-
dilution analysis (Table 1). TA was finally degraded to high-
molecular structures under the emerging chromatographic
hump, forming an intense yellow-brown coloration. The
addition of lysine and arginine, respectively, had virtually no
influence, neither on the course of reaction nor on the course
of coloration, and gave the same reaction products as without
addition. Importantly, in these incubations no formation of
polyphenol−amino acid adducts was detectable.
In contrast, incubations with cysteine showed a completely
different picture of both reaction and coloration. First, after
addition of nashi pear homogenate no change of color and no
formation of PGA and TA occurred. Instead, the two colorless
catechin−cysteine adducts A1 and A2 emerged, which were
already isolated and characterized by Richard et al. from the
enzymatic oxidation of catechin in the presence of cysteine (for
NMR data obtained herein for A1 and A2, see Supporting
Information).9 These adducts were degraded after 3 h, and two
novel colorless compounds evolved, the dihydrobenzothiazine
carboxylic acid derivatives B1 and B2 (Figure 1B). Exactly with
onset of the formation of B1 and B2 at 3 h, the development of
a red-brown color and the pronounced formation of a
chromatographic hump was initiated. This indicated that
adducts and dihydrobenzothiazine derivatives are reactive
intermediates in the color formation mechanism and are
ultimately degraded to high-molecular, yet unknown, struc-
tures under the chromatographic hump (Figure 1C).
Interestingly, the emergence of the dihydrobenzothiazine
derivatives was accompanied by the formation of PGA and
TA contributing to the overall color.
Color-dilution analyses revealed that the chromatographic
hump primarily contributed with 41% to the overall color after
8 h of incubation (Table 1), while TA gave 2%. In presence of
cysteine the time course depicted in Figure 2 clearly shows the
immediate degradation of catechin in contrast to gallic acid,
which reacted much slower. Concurrently the formation of the
catechin−cysteine adducts began in an almost quantitative
fashion, i.e., after 1 h 18% and 31% of A1 and A2, respectively,
were formed, while catechin declined by 51%. Obviously, the
Figure 1. HPLC-DAD chromatograms of incubation of gallic acid
(GA) and catechin (CA) at 3 h (A; where, for example, 1.5e+5
represents 1.5 × 105) and incubation of GA, CA, and cysteine at 3 h
(B) and at 8 h (C) at λ = 280 nm: purpurogallin-4-carboxylic acid
(PGA), theaflavic acid (TA), 2′-cysteinylcatechin (A1), 5′-cysteinyl-
catechin (A2), dihydrobenzothiazine derivative 1 (B1), dihydroben-
zothiazine derivative 2 (B2), and chromatographic hump (H).
formation of A2 was favored in comparison to A1. This
difference in formation was also reflected by B1 and B2,
suggesting a mechanistic connection between B2 and A2 and
between B1 and A1, respectively. However, the difference was
much more pronounced.
Isolation and Elucidation of the Dihydrobenzothia-
zine Carboxylic Acid Derivatives. Because of its com-
paratively minor turnover, gallic acid seemed to have no
relevant impact on the formation of A1, A2, B1, and B2 and,
thus, was left out for incubations used for structure isolation.
Incubations of catechin and cysteine were first subjected to
high-performance countercurrent chromatography to separate
B1 and B2 from A1 and A2. Afterward B1 and B2 were purified
by preparative HPLC from the organic mobile fractions, while
A1 and A2 were separated from the retained aqueous phase. It
has to be mentioned that all target compounds were of
extremely high reactivity leading very fast to colored
degradation. Thus, after removal of organic solvents at ambient
C
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
Table 1. Yields (mol %) Catechin and Color Contribution (%) of the Main Reaction Products in the Incubations of Gallic Acid
(GA) and Catechin (CA) at 3 h (1), GA and CA and Lysine at 3 h (2), GA and CA and Arginine at 3 h (3), GA and CA and
Cysteine at 3 h (4-1), and GA and CA and Cysteine at 8 h (4-2)
catechin (mol %)
color contribution (%)
TA
catechin (mol %)
color contribution (%)
hump
PGA
TA
A 1
A 2
B 1
B 2
1
2
3
4−1
4−2
0
0
0
0.3
1.9
4.8
5.2
4.0
1.7
2.4
25
25
25
0
0
0
0
18.8
0.9
0
0
0
25.2
1.9
0
0
0
5.4
0
0
0
0
20.3
6.6
0
0
0
0
2
41
identical NMR spectra for the former catechin A and C ring, as
expected from their common backbone. The missing proton
signal at C-9 (positions 2′ and 5′of the former catechin B ring,
respectively) of the dihydrobenzothiazine moiety and the ABx
system of the single α-proton H-2 at 4.20 ppm (B1) and 4.26
ppm (B2), respectively, giving two characteristic doublets of
doublets for the two non-equivalent β-protons of former
cysteine at 3.03 (J = 12.5, 7.4 Hz) and 3.17 ppm (J = 12.5, 3.3
Hz) (B1) and at 3.05 (J = 12.5, 7.5 Hz) and 3.25 ppm (J =
12.5, 3.0 Hz) (B2), respectively, verify the ring formation
between the B-ring of catechin and cysteine. These
observations are in agreement with the NMR data reported
by Richard et al. for the cysteine adducts of catechin.9 The
different constitutions of B1 and B2 are validated from the
coupling constants and the HMBC correlations. In B1 the two
aromatic protons H-6 and H-7 (6.51 and 6.55 ppm) of the
benzothiazine moiety give two doublets with a typical 3J
coupling constant of 8.1 Hz. As opposed to this, the two
aromatic protons H-6 and H-8 of B2 (6.56 and 6.56 ppm)
display a long-range coupling of 1.7 Hz. This clearly confirms
the experimentally proven origin of B1 from the 2′-substituted
and of B2 from the 5′-substituted cysteinyl catechin. The
presence of HMBC correlations of H-7 to C-2′ as well as to C-
9 and between H-6 and C-10 of B1 are in further support of
structure B1. For B2 the substitution pattern is validated by
HMBC correlations of H-8 to C-6 as well as to C-2′ (likewise
H-6 to C-8 and C-2′). The closer vicinity of the
dihydrobenzothiazine ring sulfur to the C-ring H-2′ at 5.00
ppm in B1 results in a significant downfield shift of about 0.5
ppm compared to H-2′ in B2. This downfield shift was also
confirmed for the 2′-cysteinyl substitution of catechin (A1) by
us and by Richard et al., who recorded a downfield shift of 0.73
ppm of the C-ring H-2 in comparison to H-2 for the 5′
substitution (A2).9 Thus, the structures of B1 and B2 were
unequivocally identified as 8-(3’,4’-dihydro-2H-chromene-
3’,5’,7’-triol)-3,4-dihydro-5-hydroxy-2H-benzothiazine-3-car-
boxylic acid and 7-(3’,4’-dihydro-2H-chromene-3’,5’,7’-triol)-
3,4-dihydro-5-hydroxy-2H-benzothiazine-3-carboxylic acid, re-
spectively.
Figure 2. Conversion of GA and CA in the presence of cysteine: GA
( ), CA ( ), hump ( ), absorbance of total incubation at λ = 420
●
○
▼
◆
◇
■
□
▽
▲
△
nm ( ), A1 ( ), A2 ( ), B1 ( ), B2 ( ), PGA ( ), and TA ( ).
The quantitative assessment of the hump is based on the calibration
curve of catechin.
temperature, all fractions had to be immediately freeze-dried to
give amorphous, slightly pinkish materials. The spectrometric
and spectroscopic data of the isolated cysteine−catechin
adducts A1 and A2 were in line with the above literature.9
For the novel dihydrobenzothiazine derivatives B1 showed
absorption maxima of 230 and 306 nm, while B2 gave 235 and
295 nm. The elemental composition was verified by HR−MS
to deliver a pseudo-molecular ion at m/z 390.0649 [M − H]−
for B1 and at m/z 390.0648 for B2 (m/z 390.0648 calcd for
C18H16O7NS).
Formation Mechanism of B1 and B2. To elucidate the
formation mechanism leading to B1 and B2 being isolated,
authentic A1 or A2 was added to the above incubation system
of catechin and nashi-PPO, however in the absence of cysteine.
In contrast to the presence of cysteine, the dihydrobenzothia-
zine derivative synthesis started immediately. Figure 3 clearly
shows that B1 originated solely from A1 and B2 from A2,
respectively. Importantly, this formation was accompanied by
immediate appearance of color paralleled by the development
of the chromatographic hump with several absorbance maxima
in the visible range. Additional incubations were performed
omitting either nashi-PPO or catechin. None of these
Final structural evidence was achieved by 1H NMR and 13
C
NMR-APT measurements, as well as homonuclear correlation
spectroscopy (H,H COSY), heteronuclear single quantum
coherence (HSQC), and heteronuclear multiple-bond correla-
tion (HMBC) techniques. Table 2 provides 1H NMR and 13
C
NMR data of the novel dihydrobenzothiazine derivatives as
well show selected HMBC correlations. Compared to
published data of catechin,23 both structures show virtually
D
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
1
Table 2. High-Resolution Mass and H and 13C NMR Spectroscopic Data of B1 and B2 (in CD3OD)
a
Signal was overlapped.
incubations showed the emergence of the dihydrobenzothia-
zine derivatives or the chromatographic hump. This means that
both catechin and PPO are essential for the conversion of the
cysteinyl adducts to the dihydrobenzothiazines.
moiety to the o-quinone. From literature it is known that
cysteine adducts of polyphenols are no longer recognized as
substrates by PPOs.2,9 This is in line with the present
experiments, in which A1 and A2 were stable, when only nashi-
PPO was added. However, this immediately changed when
catechin was included in the reaction mixture. Catechin as a
substrate of PPO is then oxidized to the corresponding o-
quinone, which in turn can then oxidize the cysteinyl adducts
to give back the reduced catechin to be again enzymatically
oxidized. This redox cycling is well-known from the black tea
fermentation chemistry, where, e.g., gallo substituted phenols
as epigallocatechin gallates must get oxidized in order to form
the theaflavic chromophores.24 Obviously, these redox systems
Based on these observations the formation mechanism
depicted in Figure 4 is proposed. In the first step catechin is
oxidized by nashi-PPO to the o-quinone, which reacts with the
thiol group of cysteine forming the 2′- and 5′-cysteinyl adducts
A1 and A2. The obvious steric hindrance of the attack at
position C-2′ compared to C-5′ also explains the discrim-
ination of A1 by almost 50%. For the intramolecular ring
́
closure the second nucleophilic reaction with the amino acids
primary amino function requires reoxidation of the catechol
E
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
parts B1 and B2. Most likely from the present experiments the
non-enzymatic reduction agents are catechin itself or the
cysteinyl adducts. It has to be stated that cysteine under
oxidative conditions can be oxidized easily to cystine which
represents an additional partner for redox reactions. All
isolated structures were transient intermediates finally leading
to color formation. The proposed mechanism is reminiscent of
the formation of melanins during pheomelanogenesis. In this
case the reaction starts from tyrosine via 3,4-dihydroxyphenyl-
alanine (DOPA) to give indol intermediates or in the presence
of cysteine benzothiazines. Thus, it can be hypothesized that
the present phytophenolic derivatives will show similar follow-
up reactions such as decarboxylation, ring contraction, or
polymerization and will eventually lead to high-polymeric
colored melanins.
In summary, we succeeded in gaining mechanistic insights
into the influence of nucleophilic amino acids on enzymatic
browning reactions in catechin-based model incubations.
While lysine and arginine had no impact on the general
course, cysteine incubations led to the unequivocal identi-
fication of two novel amino acid adducts in the form of
dihydrobenzothiazine derivatives. This study strongly suggests
their importance as crucial intermediates in the formation of
color. Ongoing research will now clarify the nature of the
resulting colored structures and their mechanistic background.
Figure 3. Incubation of (A; where, for example, 3.5e+5 represents 3.5
× 105) A1 or (B) A2 in the presence of CA and nashi-PPO: B1, B2,
and H.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jafc.8b06458.
are driven by the difference in oxidation potentials. This must
also be true for the final step in the mechanistic scheme in
Figure 4, as the products of ring closure, the quinone imins,
were isolated as their reduced dihydrobenzothiazine counter-
1H and 13C NMR spectroscopic data of A1 and A2 in
CD3OD (PDF)
Figure 4. Formation of dihydrobenzothiazine carboxylic acid derivatives B1 and B2 in the presence of catechin, cysteine, and PPO.
F
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Article
(13) Kroll, J.; Rawel, H. M. Reactions of plant phenols with
myoglobin: influence of chemical structure of the phenolic
compounds. J. Food Sci. 2001, 66, 48−58.
(14) Kroll, J.; Rawel, H. M.; Rohn, S. Reactions of plant phenolics
with food proteins and enzymes under special consideration of
covalent Bonds. Food Sci. Technol. Res. 2003, 9, 205−218.
(15) Guerra, P. V.; Yaylayan, V. A. Interaction of flavanols with
amino acids: postoxidative reactivity of the B-ring of catechin with
glycine. J. Agric. Food Chem. 2014, 62, 3831−3836.
(16) Mai, F.; Glomb, M. A. Isolation of Phenolic Compounds from
Iceberg Lettuce and Impact on Enzymatic Browning. J. Agric. Food
Chem. 2013, 61, 2868−2874.
(17) Collier, P. D.; Bryce, T.; Mallows, R.; Thomas, P. E.; Frost, D.
J.; Korver, O.; Wilkins, C. K. The theaflavins of black tea. Tetrahedron
1973, 29, 125−142.
(18) Imai, K.; Mitsunaga, T.; Takemoto, H.; Yamada, T.; Ito, S.;
Ohashi, H. Extractives of Quercus crispula sapwood infected by the
pathogenic fungi Raffaelea quercivora I: comparison of sapwood
extractives from noninfected and infected samples. J. Wood Sci. 2009,
55, 126−132.
AUTHOR INFORMATION
Corresponding Author
*E-mail: marcus.glomb@chemie.uni-halle.de. Fax: ++049-345-
5527341).
■
ORCID
Nils Mertens: 0000-0002-9769-1340
Marcus A. Glomb: 0000-0001-8826-0808
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
̈
We thank Dr. D. Strohl from the Institute of Organic
Chemistry, Halle, Germany, for recording NMR spectra and
Dr. A. Frolov and A. Laub from the Leibniz Institute of Plant
Biochemistry, Halle, Germany, for performing accurate mass
analysis.
(19) Hofmann, T. Characterization of the most intense coloured
compounds from Maillard reactions of pentoses by application of
colour dilution analysis. Carbohydr. Res. 1998, 313, 203−213.
(20) Graham, H. N. Green tea composition, consumption and
polyphenol chemistry. Prev. Med. 1992, 21, 334−350.
(21) Tsurutani, M.; Murata, M.; Homma, S. Comparison of
Enzymatic Browning of Japanese Pear and Apple. Food Sci. Technol.
Res. 2000, 6, 344−347.
(22) Sang, S.; Lambert, J. D.; Tian, S.; Hong, J.; Hou, Z.; Ryu, J.;
Stark, R. E.; Rosen, R. T.; Huang, M.; Yang, C. S.; Ho, C. Enzymatic
synthesis of tea theaflavin derivatives and their anti-inflammatory and
cytotoxic activities. Bioorg. Med. Chem. 2004, 12, 459−467.
(23) Hye, M. A.; Taher, M. A.; Ali, M. Y.; Ali, M. U.; Zaman, S.
Isolation of (+)-catechin from Acacia Catechu (cutch tree) by a
convenient method. J. Sci. Res. 2009, 1, 300−305.
(24) Robertson, A. Effects of physical and chemical conditions on
the in vitro oxidation of tea leaf catechins. Phytochemistry 1983, 22,
889−896.
ABBREVIATIONS USED
■
DOPA, 3,4-dihydroxyphenylalanin; PGA, purpurogallin car-
boxylic acid; PPO, polyphenol oxidase; TA, theaflavic acid
REFERENCES
■
(1) Toivonen, P. M. A.; Brummell, D. A. Biochemical bases of
appearance and texture changes in fresh-cut fruit and vegetables.
Postharvest Biol. Technol. 2008, 48, 1−14.
(2) Bittner, S. When quinones meet amino acids: chemical, physical
and biological consequences. Amino Acids 2006, 30, 205−224.
(3) Mason, H. S.; Peterson, E. W. Melanoproteins I. Reactions
between enzyme-generated quinones and amino acids. Biochim.
Biophys. Acta, Gen. Subj. 1965, 111, 134−146.
(4) Pierpoint, W. S. o-Quinones formed in plant extracts. Their
reactions with amino acids and peptides. Biochem. J. 1969, 112, 609−
616.
(5) Schilling, S.; Sigolotto, C.; Carle, R.; Schieber, A. Character-
ization of covalent addition products of chlorogenic acid quinone with
amino acid derivatives in model systems and apple juice by high-
performance liquid chromatography/electrospray ionization tandem
mass spectrometry. Rapid Commun. Mass Spectrom. 2008, 22, 441−
448.
(6) Yabuta, G.; Koizumi, Y.; Namiki, K.; Hida, M.; Namiki, M.
Structure of green pigment formed by the reaction of caffeic acid
esters (or chlorogenic acid) with a primary amino compound. Biosci.,
Biotechnol., Biochem. 2001, 65, 2121−2130.
(7) Dudley, E. D.; Hotchkiss, J. H. Cysteine as inhibitor of
polyphenoloxidase. J. Food Biochem. 1989, 13, 65−75.
(8) Bassil, D.; Makris, D. P.; Kefalas, P. Oxidation of caffeic acid in
the presence of l-cysteine: isolation of 2-S-cysteinylcaffeic acid and
evaluation of its antioxidant properties. Food Res. Int. 2005, 38, 395−
402.
(9) Richard, F. C.; Goupy, P. M.; Nicolas, J. J.; Lacombe, J. M.;
Pavia, A. A. Cysteine as an inhibitor of enzymic browning. 1. Isolation
and characterization of addition compounds formed during oxidation
of phenolics by apple polyphenol oxidase. J. Agric. Food Chem. 1991,
39, 841−847.
(10) Ali, H. M.; El-Gizawy, A. M.; El-Bassiouny, R. E. I.; Saleh, M. A.
Browning inhibition mechanisms by cysteine, ascorbic acid and citric
acid, and identifying PPO-catechol-cysteine reaction products. J. Food
Sci. Technol. 2014, 52, 3651−3659.
(11) Valero, E.; Varon, R.; Garcia-Carmona, F. A kinetic study of
irreversible enzyme inhibition by an inhibitor that is rendered
unstable by enzymic catalysis. The inhibition of polyphenol oxidase by
L-cysteine. Biochem. J. 1991, 277, 869−874.
(12) Kroll, J.; Rawel, H. M.; Seidelmann, N. Physicochemical
properties and susceptibility to proteolytic digestion of myoglobin−
phenol derivatives. J. Agric. Food Chem. 2000, 48, 1580−1587.
G
DOI: 10.1021/acs.jafc.8b06458
J. Agric. Food Chem. XXXX, XXX, XXX−XXX