Identification and quantitation of new glutamic acid derivatives in soy sauce by UPLC/MS/MS

Article abstract of DOI:10.1002/cbdv.201300150

Glutamic acid is an abundant amino acid that lends a characteristic umami taste to foods. In fermented foods, glutamic acid can be found as a free amino acid formed by proteolysis or as a non-proteolytic derivative formed by microorganisms. The aim of the present study was to identify different structures of glutamic acid derivatives in a typical fermented protein-based food product, soy sauce. An acidic fraction was prepared with anion-exchange solid-phase extraction (SPE) and analyzed by UPLC/MS/MS and UPLC/TOF-MS. α-Glutamyl, γ-glutamyl, and pyroglutamyl dipeptides, as well as lactoyl amino acids, were identified in the acidic fraction of soy sauce. They were chemically synthesized for confirmation of their occurrence and quantified in the selected reaction monitoring (SRM) mode. Pyroglutamyl dipeptides accounted for 770 mg/kg of soy sauce, followed by lactoyl amino acids (135 mg/kg) and γ-glutamyl dipeptides (70 mg/kg). In addition, N-succinoylglutamic acid was identified for the first time in food as a minor compound in soy sauce (5 mg/kg). Copyright

Full text of DOI:10.1002/cbdv.201300150

1842
CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
Identification and Quantitation of New Glutamic Acid Derivatives in Soy
Sauce by UPLC/MS/MS
by Eric Frerot* and Ting Chen1)
Firmenich SA, Corporate R&D Division, 1 Route des Jeunes, P.O. Box 239, CH-1211 Geneva-8
(phone: þ41-227803566; fax: þ41-227803334; e-mail: eric.frerot@firmenich.com)
Glutamic acid is an abundant amino acid that lends a characteristic umami taste to foods. In
fermented foods, glutamic acid can be found as a free amino acid formed by proteolysis or as a non-
proteolytic derivative formed by microorganisms. The aim of the present study was to identify different
structures of glutamic acid derivatives in a typical fermented protein-based food product, soy sauce. An
acidic fraction was prepared with anion-exchange solid-phase extraction (SPE) and analyzed by UPLC/
MS/MS and UPLC/TOF-MS. a-Glutamyl, g-glutamyl, and pyroglutamyl dipeptides, as well as lactoyl
amino acids, were identified in the acidic fraction of soy sauce. They were chemically synthesized for
confirmation of their occurrence and quantified in the selected reaction monitoring (SRM) mode.
Pyroglutamyl dipeptides accounted for 770 mg/kg of soy sauce, followed by lactoyl amino acids (135 mg/
kg) and g-glutamyl dipeptides (70 mg/kg). In addition, N-succinoylglutamic acid was identified for the
first time in food as a minor compound in soy sauce (5 mg/kg).
Introduction. – In the 1970s, Japanese researchers studied the small peptides in soy
sauce [1–3]. Using gel filtration, ion-exchange chromatography, and paper chroma-
tography, coupled with amino acid analysis, they identified many acidic dipeptides
(Glu-Glu, Asp-Glu, Glu-Ala, etc.) considered to contribute to the umami taste of soy
sauce [1]. More recent studies [4–7] questioned the contribution of small peptides, and
revealed that amino acids and sodium salts were responsible for this umami taste.
Toelstede and Hofmann [8][9] identified a series of a-glutamyl and g-glutamyl
dipeptides in cheeses, and Toelstede et al. [9] identified many g-glutamyl dipeptides in a
study on the maturation of Gouda cheese. The food industry is actively seeking
substitutes for monosodium glutamate that either occur naturally in foods or plants, or
can be produced artificially [10].
At Firmenich, Frerot and Escher [11] studied the contribution of synthetic glutamyl
tripeptides and lactoyl amino acids to the taste of cheese-like solutions. Chemically
synthesized succinoyl amino acids such as N-succinyl glutamate were also found to be
taste-active [12], but were never reported to occur in foods.
In this work, we analyzed a commercial and widely distributed soy sauce, focusing
on glutamic acid derivatives, in order to show the natural occurrence of many amino
acid derivatives.
1
)
Present Address: Firmenich Aromatics, Shanghai Firmenich Aromatics (China) Co., Ltd., No. 3901
JinDu Rd., Shanghai, P. R. China.
ꢀ 2013 Verlag Helvetica Chimica Acta AG, Zꢁrich

CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
1843
Results and Discussion. – The study was performed in four steps: preparation of an
acidic fraction of soy sauce, identification of the peptides and amino acid derivatives,
synthesis of unavailable reference products for confirmation, and calibration and
quantification of identified compounds.
For the preparation of an acidic fraction of soy sauce, solid-phase extraction (SPE)
was performed using a mixed-mode polymeric reversed-phase anion-exchanger (Oasis
MAX). These SPE cartridges were used for microbial natural-product detection and
they showed consistency and high loading capacity [13]. Although the ꢂLess Saltꢃ soy
sauce contained less sodium chloride, it had to be diluted 167ꢀ in H2O before loading.
After the loading, rinsing, and elution, the acidic fraction was obtained as a 5%
HCOOH solution that was 20-fold diluted in relation to the genuine soy sauce. It was
used as such for both qualitative and quantitative analyses.
The acidic fraction of soy sauce was analyzed by ultra-performance liquid
chromatography/mass spectrometry (UPLC/MS). As current UPLC columns can
sustain pure aqueous solvent, a good separation was obtained even for the most polar
compounds. The first analyses were performed on ion-trap and time-of-flight (TOF)
instruments because of their high scanning speed, which was suitable for UPLC. Fig. 1
shows the negative full-scan chromatogram obtained by electrospray ionization (ESI).
The first identification of peaks was achieved from the positive- and negative-ion mode
MS/MS spectra obtained with the ion-trap mass spectrometer, in addition to the
molecular formula calculated from TOF data. In Table 1, the identities of the peaks and
the spectral data that led to structural determination, as recommended by the
International Organization of the Flavor Industry (IOFI) [14], were compiled. These
compounds belong to four families of amino acid derivatives, as shown in Fig. 2. Their
identification was confirmed by comparison with the spectral and chromatographic
data of reference products synthesized by classic methods. Pyroglutamyl dipeptides
Fig. 1. Negative ESI full-scan trace of soy-sauce acidic extract

1844
CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)

CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
1845
Fig. 2. Dipeptides and amino acid derivatives found in soy-sauce acidic extract
(Fig. 2) were the major compounds. In 1970, pGlu-Glu (7) and pGlu-Pro (9) were
identified in the edible mushroom Agaricus campestris [15]. pGlu-Gly and pGlu-Gln
were recently described in soy sauce [6]. We could not identify these two compounds by
the full-scan experiments (Fig. 1). g-Glutamyl dipeptides are ubiquitous compounds in
foods [8][9][16][17]. Lactoyl amino acids were first discovered by Firmenich during
the analysis of Parmesan [18] and patented for their taste properties [19]. A thorough
study of Parmesan and other Italian cheeses revealed that pyroglutamyl dipeptides,
lactyl amino acids, and g-glutamyl dipeptides accumulated during ripening, probably
because they could no longer be recognized by hydrolytic enzymes [20]. While the four
families of amino acid derivatives identified in this work were already known, the most
polar compounds had never been previously described. In this work, pGlu-Asp (4),
pGlu-Val (11), and Lac-Glu (8) were unambiguously identified in a food product for
the first time.
Other taste-active glutamic acid derivatives were already described in foods: aGlu-
Glu (1) in cheeses [8][9], and N-acetylglutamic acid (5) in soybean seeds [21] and in
fermented tuna [22]. The two products 1 and 5, as well as N-succinoylglutamic acid (6)
[12], could not be identified by the first full-scan experiments using the ion-trap MS, but
were detected by the TOF-MS. Their occurrence in soy sauce was firmly established by
determining the MS transition ratios in the selected reaction monitoring mode (SRM),
using the triple quadrupole mass spectrometer. The ratio of the peak area
corresponding to two transitions was determined at different concentrations for the
synthetic products. This ratio was subsequently compared with that actually found in
soy sauce to ensure correct identification of the compounds, following the IOFI
guidelines [14]. Thus, the presence of aGlu-Glu (1), Suc-Glu (6), Ac-Glu (5), and Lac-
Glu 8 (Fig. 3) in soy sauce could be evidenced by SRM recordings in both the positive-
and the negative-ionization modes, and it exceeded IOFI recommendations.
Fig. 3. Additional dipeptides and amino acid derivatives in-soy sauce acidic extract detected by UPLC/
MS/MS in the SRM mode

1846
CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
The concentrations of the identified compounds were determined by UPLC/MS in
the SRM mode. The external calibration curves were obtained from the synthetic
products (0.05–50 ppm) using two mass transitions. As there was no anticipated
sensitivity issue, the MS conditions were not fully optimized, and the same collision
energy was used for all compounds. A total of 16 compounds were quantified in an SPE
Oasis MAX extract of soy sauce. The recoveries during the SPE sample-preparation
step were not investigated, but were described as nearly quantitative for weak acids
[23]. For most of these compounds, the quantitative results obtained from positive- and
negative-ion ESI SRM experiments were in good agreement, thereby minimizing the
possibility of biased results due to ion-suppression effects. In Table 2, the results
expressed as the concentration in parts per million (mg/l) in the original soy sauce are
collected. The quantified pyroglutamyl dipeptides accounted for a total of 770 ppm,
lactoyl amino acids for 135 ppm, and g-glutamyl dipeptides for only 70 ppm. Suc-Glu
(6) accounted for less than 10 ppm.
Table 2. Concentration of Acidic Amino Acid Derivatives in Soy Sauce
Product
Concentration in soy sauce [mg/l]
Negative-ion modea)
Positive-ion modeb)
Glu-Glu (1)
5.0
18.4
207.5
8.1
6.6
25.2
245.7
10.3
8.0
gGlu-Glu (2)
pGlu-Asp (4)
N-acetyl-Glu (5)
Suc-Glu (6)
5.4
pGlu-Glu (7)
Lac-Glu (8)
344.8
82.1
121.8
11.3
11.6
30.3
40.0
48.0
46.3
1.4
339.6
85.5
144.3
13.4
14.0
31.5
n.d.d)
53.6
44.1
3.0
pGlu-Pro (9)
gGlu-Ilec) (12)
gGlu-Leu (13)
gGlu-Phe (14)
Lac-Val (15)
pGlu-Ilee) (16)
pGlu-Leu (17)
pGlu-Phe (18)
Lac-Leu (20)
12.5
20.7
a) Quantified from negative-ion ESI-MS. b) Quantified from positive-ion ESI-MS. c) Same calibration
curve as for gGlu-Leu used. d) n.d., Not determined. e) Same calibration curve as for pGlu-Leu used.
During fermentation of soy and wheat proteins, the so-called Koji fermentation, a-
glutamyl dipeptides are formed by proteolytic enzymes present in the Aspergillus
microorganism [5], whereas g-glutamyl dipeptides are formed by g-glutamyl trans-
ferases [8]. Pyroglutamyl dipeptides are assumed to form from the corresponding a-
glutamyl dipeptides (a-Glu-Aaa) or from a-glutaminyl dipeptides (a-Gln-Aaa). No
enzyme is needed, since heating was shown to be sufficient to perform the cyclization
reaction [24]. The enzymatic coupling of pyroglutamic acid to a free amino acid was
also proposed [15]. The formation of lactyl amino acids in Parmesan cheese was briefly
discussed, and a coupling reaction between l-lactic acid and free l-amino acid was

CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
1847
proposed [20]. It should be mediated by an enzyme, since only l-lactyl-l-amino acids
were found in cheese [20]. In our work, the synthetic product Lac-Glu (8), synthesized
from l-lactic acid and l-glutamic acid ester [19], contained a small amount of a minor
diastereoisomer, likely d-Lac-l-Glu. We verified that this diastereoisomer was not
present in soy sauce.
Suc-Glu (6) was already known as an intermediate in the arginine catabolic
pathway of bacteria such as Pseudomonas aeruginosa [25]. A succinyl transferase
enzyme (EC 2.3.1.109) is involved and is considered to be specific to arginine (Arg) and
ornithine, according to the KEGG database [26][27]. However, it leaves the possibility
open that other succinyl amino acids may occur in fermented products. Thus, a
commercial mixture of the 20 amino acids was reacted with succinic anhydride and
analyzed by UPLC/MS. In the negative-ion ESI mode, all succinyl amino acids showed
a loss of 100 Da (loss of the succinyl residue). Taking advantage of this fragmentation,
we rapidly developed an SRM method to analyze diluted soy sauce without any sample
preparation. In addition to the two major compounds Suc-Arg and Suc-Glu (6), we
detected significant amounts of succinylated histidine, serine, valine, leucine, isoleu-
cine, phenylalanine, and tryptophane (data available as Supplementary Material). Even
if the identification of many succinyl amino acids remains tentative, we can postulate
that Suc-Arg and Suc-Glu (6) may arise from the arginine catabolic pathway, but that
the succinyl transferase present in Aspergillus oryzae may also act on other free amino
acids.
Conclusions. – During this work, we showed the natural occurrence of many
derivatives of glutamic acid, many of which are claimed to be taste-active. In particular,
Suc-Glu (6) and Lactyl-Glu (8) were shown to occur in a food product for the first time.
The authors thank Robert Brauchli for the NMR signals assignment, Daniel Grenno for LC/ToF
recordings, and Dr. Elise Sarrazin for preliminary experiments. This article is dedicated to Dr. Sina
Escher, who first identified many compounds described here in Parmesan cheese, and to FranÅois Benzi,
flavorist, for his continuous support.
Experimental Part
General. ꢂLess Saltꢃ soy sauce (salt 8.4%, Kikkoman Corp., D-Dꢁsseldorf) was purchased from a
local grocery store. High-performance liquid chromatography (HPLC)-grade MeCN and Milli-Q H2O
were used, and all other reagents were of anal. grade. UPLC/MS-Grade solvents were purchased from
BioSolve (NE-Valkenswaard).
Solid-Phase Extraction (SPE) Preparation of Soy Sauce. An Oasis MAX cartridge (6 ml/500 mg,
60 mm, Waters, Milford, MA, USA) was conditioned and equilibrated with MeOH (10 ml) and then
deionized (DI) H2O (10 ml). Soy sauce (0.3 ml) was diluted in DI H2O (50 ml), and then loaded onto the
cartridge with a flow of 5–8 ml/min. Then 5% aq. NH3 soln. (20 ml) was percolated at 1.5–2.5 ml/min to
bind the acidic components to the resin. The cartridge was then rinsed with DI H2O (80 ml) and dried by
suction. It was rinsed with CH2Cl2/MeOH 95 :5 (v/v; 6 ml) and carefully dried. The acidic components
were eluted with 5% HCOOH (6 ml) in H2O. The eluent was injected into the UPLC/MS.
Instruments. A system consisting of an Acquity UPLCꢀ (Waters) and a LXQ mass spectrometer
(Thermo, San Jose, CA, USA) was used for the full-scan and product ion scan experiments. A system
consisting of an Acquity UPLCꢀ and a TSQ Quantum Ultra AM mass spectrometer (Thermo) was used
for SRM experiments.

1848
CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
UPLC/MS Method (Qual., LXQ). The column was a Waters Acquity UPLC BEH C18 (1.7 mm 100ꢀ
2.1 mm I. D., 258). Solvent A was 0.1% HCOOH in H2O, and solvent B was 0.1% HCOOH in MeCN,
eluted at 0.3 ml/min in gradient mode: 0% B isocratic, 0–0.5 min; 0–10% B, 0.5–4.0 min; 10–100% B,
4.00–9.00 min; 100% B isocratic, 9.0–10.0 min; equilibration for 2 min at 0% B. The injection volume
was 3 ml. MS: source-heated electrospray ionization (H-ESI) mass spectrometer operated in the positive-
and the negative-ion mode. In the positive mode: source voltage, 4000 V, vaporizer temp., not heated,
sheath gas flow rate, 50 arb. units; aux. gas flow rate, 5 arb. units; sweep gas flow rate, 0 arb. units; heated
cap. temp., 3258; source CID, 5 V. In the negative-ion mode: source voltage, 5000 V; vaporizer temp., not
heated; sheath gas flow rate, 50 arb. units; aux. gas flow rate, 5 arb. units; sweep gas flow rate, 0 arb. units;
heated cap. temp., 3258; source CID, 5 V. Scan event 1, full scan [130–800]; scan event 2; data-dependent
MS/MS of the 1st most intense ion from scan event 1; collision energy, 35 V.
UPLC/MS Method (Quant., TSQ). The column was a Waters Acquity UPLC HSS T3 (1.8 mm 100ꢀ
2.1 mm i.d., 258). Solvent A was 0.1% HCOOH in Milli-Q H2O. The eluting conditions were the same as
by the qual. method, except for the solvent gradient: 0% B isocratic, 0–0.5 min; 0–10% B, 0.5–4.0 min;
10–25% B, 4.00–6.00 min; 25–100% B, 6.00–9.00 min; 100% B isocratic, 9.0–10.0 min, equilibration for
2 min at 0% B. Injection volume was 3 ml. The acidic fraction as obtained from SPE was injected. A
dilution factor of 20 was applied for the calculation regarding soy sauce.
MS: Source H-ESI operated in the negative-ion mode, spray voltage, 3900 V; vaporizer temp. 3008;
sheath gas pressure, 40 arb. units; aux. gas pressure, 20 arb. units; ion sweep gas pressure, 2.0 arb. units;
heated cap. temp., 3258; source CID, 5 V; Q2 Ar gas pressure, 0.8 mTorr. Data acquisition: Q1 and
Q3 peak width, 0.7 (nominal mass resolution); 0.5-Da scan width; scan time, 0.05 s; and 25-V collision
energy. Transitions used for quantitation in the negative mode: 275!128/146 (Glu-Glu, gGlu-Glu),
243!109/127 (pGlu-Asp), 188!102/128 (Ac-Glu), 257!128/82 (pGlu-Glu), 246!98/128 (Suc-Glu),
218!88/128 (Lac-Glu), 225!82/112 (pGlu-Pro), 259!130/128 (gGlu-Ile, gGlu-Leu), 293!164/128
(gGlu-Phe), 241!82/141 (pGlu-Ile, pGlu-Leu), 275!127/109 (pGlu-Phe), 202!143/43 (Lac-Leu).
Source H-ESI operated in positive-ion mode; spray voltage, 3700 V; vaporizer temp., 1208; sheath
gas pressure, 30 arb. units; aux. gas pressure, 20 arb. units; ion sweep gas pressure 2.0 arb. units; heated
cap. temp., 3508; source CID, 5 V; Q2 Ar gas pressure, 0.8 mTorr. Data acquisition: Q1 and Q3 peak
width, 0.7 (nominal mass resolution); 0.5-Da scan width, scan time, 0.05 s, and 30-V collision energy.
Transitions used for quantitation in the positive mode: 277!84/130 (Glu-Glu, gGlu-Glu), 245!84/74
(pGlu-Asp), 190!84/130 (Ac-Glu), 259!84/130 (pGlu-Glu), 248!138/84 (Suc-Glu), 220!84/130
(Lac-Glu), 227!84/70 (pGlu-Pro), 261!86/84 (gGlu-Ile, gGlu-Leu), 295!120/84 (gGlu-Phe), 243!
86/84 (pGlu-Ile, pGlu-Leu), 277!120/84 (pGlu-Phe), 204!86/84 (Lac-Leu).
The calibration curves were obtained from solns. of 0.1–100 ppm of the synthetic reference products
(prepared as described below) and prepared by consecutive dilutions in DI H2O. Quadratic fitting and 1/
ꢀ weighting gave r2 >0.995 using XCalibur software (Thermo).
Evidence for the New Natural Occurrence of aGlu-Glu (1), Ac-Glu (5), Suc-Glu (6), and Lac-Glu
(8) (TSQ). The ratio between the transitions for compounds 1, 5, and 6 was determined on the TSQ
instrument from the calibration injections at 0.1–100 ppm (8 points) using XCalibur software. For
compound 1 in the negative-ion mode, with the transition 275!128 being normalized to 100 in intensity,
the transition 275!146 represented 59ꢂ3 for the eight injections. In soy sauce extract, it was at 63. In
the positive-ion mode, 277!84 (100), 130(43ꢂ4); in soy sauce 277!84 (100), 130 (42). aGlu-Glu (1)
was clearly present in soy sauce according to the guidelines of the IOFI [14]. Compound 5: negative-ion
mode 188!102 (100),128 (46ꢂ3); in soy sauce 188!102 (100), 128 (48); positive-ion mode: 190!84
(100), 130 (24ꢂ2); in soy sauce 190!84 (100), 130 (25). Ac-Glu (5) was clearly present in soy sauce
according to IOFI guidelines. Compound 6: negative-ion mode 246!98 (100), 128 (84ꢂ7); in soy sauce
246!98 (100), 128 (92); positive-ion mode: 248!138 (100), 84 (95ꢂ2); in soy sauce 248!138 (100),
84 (84). Suc-Glu (6) was clearly present in soy sauce according to IOFI guidelines. Compound 8:
negative-ion mode 218!128 (100), 88 (30ꢂ5); in soy sauce 218!128 (100), 128 (30); positive-ion
mode: 220!84 (100), 130 (12ꢂ2); in soy sauce 248!84 (100), 130 (13). Lac-Glu (8) was clearly present
in soy sauce according to IOFI guidelines.
TOF-MS Recordings. A 1200 Series HPLC system and an Agilent G1969A ToF MS were used with
multimode source. The column was the same HSS T3 as described above (1.8 mm 100ꢀ2.1 mm i.d., 608).

CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
1849
Solvents A and B were the same as described above. Flow rate was 0.5 ml/min in gradient mode: 0% B
isocratic, 0–1.0 min; 0–100% B, 1.0–10.0 min; equilibration, for 2 min at 0% B; 0.5 ml/min gradient,
100% A, 1 min, to 100% B in 8 min; B 100% at 10 min; to 0% at 12 min post equilibration, 1 min.
Injection of 1 ml. MS: Source-operated in ESI mode. High resolution: 3 ppm accuracy. Fragmentor, 140.
Scan range, 103–1100; online standard for mass adjustment.
Reference Compounds. All amino acid derivatives were of l-configuration. l-Lactic acid was
purchased from Fluka (CH-Buchs).
Products aGlu-Glu (1), gGlu-Glu (2), pGlu-Val (11), gGlu-Leu (13), gGlu-Phe (14), and pGlu-Phe
(15) were purchased from Bachem (CH-Bubendorf); citric acid (3) and Ac-Glu (5) were purchased from
Sigma-Aldrich (CH-Buchs). The synthesis of Suc-Glu (6) and its spectral data were described by Frerot
and Benzi [12]. The synthesis of l-Lac-Glu (8) and l-Lac-Leu (20), and their spectral data were
described by Frerot and Escher [19].
The syntheses of pyroglutamyl dipeptides pGlu-Asp (4), pGlu-Glu (7), pGlu-Pro (8), pGlu-Ile (9),
and pGlu-Leu (10) were carried out in two steps, as exemplified by the preparation of 7.
pGlu-Glu (¼ 5-Oxoprolyl-l-glutamic Acid; 7). Step 1 (coupling): N-[(benzyloxy)carbonyl]-pyro-
glutamic acid (Z-pGlu-OH; 2.63 g, 10 mmol; Novabiochem, Lꢄufelfingen), glutamic acid dibenzyl ester
p-toluenesulfonate salt (H-Glu(OBzl)-OBzl; 1.1 equiv.; Novabiochem), and PyBOPꢅ (1 equiv.;
Novabiochem) were diluted in CH2Cl2 (100 ml) before EtNiPr2 (Sigma Aldrich) was added. The mixture
was stirred overnight. AcOEt (250 ml) was added, and the mixture was washed with 5% KHSO4, 5%
NaHCO3, and brine, and dried and evaporated to give a pale-yellow viscous oil. Flash chromatography
over silica gel (SiO2; Puriflash, 300 g; Interchim, F-MontluÅon) was performed with cyclohexane/AcOEt
and gave pure Z-pGlu-Glu(OBzl)-OBzl as a white solid (3.63 g, 63%). Step 2 (hydrogenolysis): Z-pGlu-
Glu(OBzl)-OBzl was dissolved in CH2Cl2/EtOH/H2O 10 :12 :4 (26 ml/mmol), and 5% Pd/C (1 g, Fluka)
was added. The mixture was shaken for 2–6 d under H2. H2O (ca. 100 ml) was added, and the mixture
was washed 3ꢀwith CH2Cl2. The hydroalcoholic soln. was concentrated in vacuo, H2O (200 ml) was
added, and the soln. was freeze-dried to yield a white powder (1.62 g, 98%).1H-NMR (400 MHz, D2O):
1.90–2.00 (m, 1 H); 2.07–2.19 (m, 2 H); 2.37 (t, J¼7.6, 2 H); 2.41–2.48 (m, 2 H); 2.50–2.60 (m, 1 H);
4.21–4.25 (m, 1 H); 4.34–4.39 (m, 1 H). 13C-NMR (125 MHz, D2O): 28.0 (t); 29.1 (t); 32.2 (t); 33.5 (t);
56.2 (d); 59.9 (d); 177.7 (s); 179.3 (s); 180.5 (s); 185.4 (s).
pGlu-Asp (¼ 5-Oxoprolyl-l-aspartic Acid; 4). Overall yield: 82%. 1H-NMR (400 MHz, D2O): 2.06–
2.15 (m, 1 H); 2.35–2.48 (m, 2 H); 2.50–2.60 (m, 1 H); 2.79 (dd, J¼16.5, 7.7, 1 H); 2.90 (dd, J¼16.5, 5.0,
1 H); 4.35 (dd, J¼9.1, 5.0, 1 H); 4.59 (dd, J¼7.7, 5.0, 1 H). 13C-NMR (100 MHz, D2O): 28.0 (t); 32.2 (t);
40.0 (t); 54.0 (d); 60.0 (d); 177.2 (s); 178.8 (s); 179.1 (s); 185.3 (s).
pGlu-Pro (¼ 5-Oxoprolylproline; 9). Overall yield: 91%. 1H-NMR (360 MHz, D2O): 2.00–2.09 (m,
4 H); 2.28–2.36 (m, 1 H); 2.44 (~t, J¼7.9, 2 H); 2.57–2.68 (m, 1 H); 3.58–3.67 (m, 1 H); 3.68–3.75 (m,
1 H); 4.46 (dd, J¼7.9, 4.8, 1 H); 4.70 (dd, J¼9.1, 5.0, 1 H). 13C-NMR (90 MHz, D2O): 26.8 (t); 27.5 (t);
31.6 (t); 32.1 (t); 49.9 (t); 58.3 (d); 62.5 (d); 175.8 (s); 178.8 (s); 185.0 (s).
pGlu-Ile (¼ 5-Oxoprolylisoleucine; 16). Overall yield: 90%. 1H-NMR (360 MHz, D2O): 0.88 (t, J¼
7.4, 3 H); 0.93 (d, J¼6.9, 3 H); 1.13–1.26 (m, 1 H); 1.39–1.51 (m, 1 H); 1.85–1.96 (m, 1 H); 2.04–2.14
(m, 1 H); 2.40–2.46 (m, 2 H); 2.48–2.60 (m, 1 H); 4.22 (d, J¼6.1, 1 H); 4.40 (dd, J¼8.9, 5.1, 1 H).
13C-NMR (90 MHz, D2O): 13.6 (q); 18.0 (q); 27.6 (t); 28.1 (t); 32.2 (t); 39.4 (d); 59.8 (d); 62.0 (d); 177.5
(s); 179.7 (s); 185.3 (s).
1
pGlu-Leu (¼ 5-Oxoprolylleucine; 17). Overall yield: 86%. H-NMR (360 MHz, D2O): 0.91 (br. d,
J¼6.0, 3 H); 0.93 (br. d, J¼6.0, 3 H); 1.58–1.65 (m, 3 H); 2.06–2.16 (m, 1 H); 2.41–2.47 (m, 2 H); 2.50–
2.62 (m, 1 H); 4.21–4.28 (m, 1 H); 4.36 (dd, J¼9.1, 5.0, 1 H). 13C-NMR (90 MHz, D2O): 23.6 (q); 25.3
(q); 27.5 (d); 28.0 (t); 32.2 (t); 43.3 (t); 56.8 (d); 59.9 (d); 177.1 (s); 182.6 (s); 185.2 (s).
Lac-Val (15) was synthesized in two steps according to Frerot and Escher [19] from l-lactic acid and
H-Val-OBzl tosylate salt (Bachem), with an overall yield of 56% after hydrogenolysis. 1H-NMR
(400 MHz, D2O): 0.91 (d, J¼6.9, 3 H); 0.94 (d, J¼6.9, 3 H); 1.39 (d, J¼6.9, 3 H); 2.17 (m, 1 H); 4.15 (d,
J¼5.7, 1 H); 4.30 (q, J¼6.9, 1 H). 13C-NMR (100 MHz, D2O): 20.0 (q); 21.6 (q); 22.7 (q); 33.4 (d); 62.2
(d); 70.8 (d); 180.0 (s); 180.3 (s).

1850
CHEMISTRY & BIODIVERSITY – Vol. 10 (2013)
REFERENCES
[1] S. Oka, K. Nagata, Agric. Biol. Chem. 1974, 38, 1195.
[2] S. Oka, K. Nagata, Agric. Biol. Chem. 1974, 38, 1185.
[3] S. Arai, M. Yamashita, M. Fujimaki, Agric. Biol. Chem. 1972, 36, 1253.
[4] H. N. Lioe, K. Takara, M. Yasuda, J. Food Sci. 2006, 71, S277.
[5] H. N. Lioe, J. Selamat, M. Yasuda, J. Food Sci. 2010, 75, R71.
[6] S. Kaneko, K. Kumazawa, O. Nishimura, Biosci., Biotechnol., Biochem. 2011, 75, 1275.
[7] H. N. Lioe, A. Apriyantono, K. Takara, K. Wada, H. Naoki, M. Yasuda, J. Agric. Food Chem. 2004,
52, 5950.
[8] S. Toelstede, T. Hofmann, J. Agric. Food Chem. 2009, 57, 3738.
[9] S. Toelstede, A. Dunkel, T. Hofmann, J. Agric. Food Chem. 2009, 57, 1440.
[10] C. Winkel, A. de Klerk, E. de Rijke, J. Bakker, T. Koenig, H. Renes, Chem. Biodiversity. 2008, 5,
1195.
[11] E. Frerot, S. D. Escher, to Firmenich S.A., PCT Int. Appl. WO 9704667 (1997).
[12] E. Frerot, F. Benzi, to Firmenich S.A., PCT Int. Appl. WO 2004075663 (2004).
˚
[13] M. Mansson, R. K. Phipps, L. Gram, M. H. G. Munro, T. O. Larsen, K. F. Nielsen, J. Nat. Prod. 2010,
73, 1126.
[14] IOFI Working Group on Methods of Analysis, Flavour Fragrance J. 2010, 25, 2.
[15] M. R. Altamura, R. Andreotti, M. L. Bazinet, L. Long Jr., J. Food Sci. 1970, 35, 134.
[16] A. Dunkel, J. Kçster, T. Hofmann, J. Agric. Food Chem. 2007, 55, 6712.
[17] F. Roudot-Algaron, L. Kerhoas, D. Le Bars, J. Einhorn, J. C. Gripon, J. Dairy Sci. 1994, 77, 1161.
[18] E. M. Frerot, S. D. Escher, F. Nꢄf, in ꢂBook of Abstractsꢃ, 210th ACS National Meeting, Chicago, IL,
August 20–24 (1995), (Pt. 1), AGFD-216.
[19] E. Frerot, S. Escher, to Firmenich S.A., US Pat. 5,780,090 (1998).
[20] S. Sforza, V. Cavatorta, G. Galaverna, A. Dossena, R. Marchelli, Int. Dairy J. 2009, 19, 582.
[21] A. O. Hession, E. G. Esrey, R. A. Croes, C. A. Maxwell, J. Agric. Food Chem. 2008, 56, 9121.
[22] G. Haseleu, E. Lubian, S. Mueller, F. Shi, T. Koenig, J. Agric. Food Chem. 2013, 61, 3205.
`
´
[23] N. Fontanals, B. C. Trammell, M. Galia, R. M. Marce, P. C. Iraneta, F. Borrull, U. D. Neue, J. Sep. Sci.
2006, 29, 1622.
[24] T. Kasai, T. Nishitoba, S. Sakamura, Agric. Biol. Chem. 1983, 47, 2647.
[25] A. Jann, V. Stalon, C. Vander Wauven, T. Leisinger, D. Hass, Proc. Natl. Acad. Sci. U.S.A. 1986, 83,
4937.
[26] M. Kanehisa, S. Goto, Y. Sato, M. Furumichi, M. Tanabe, Nucleic Acids Res. 2012, 40, D109.
[27] M. Kanehisa, S. Goto, Nucleic Acids Res. 2000, 28, 27.
Received May 7, 2013