Synthesis of 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one from maltol and its taste identification

Article abstract of DOI:10.1016/j.foodchem.2021.130052

2,3-Dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one (DDMP) exists in many foods, and its effect on taste is controversial. The aim of this study was to clarify whether DDMP has bitter taste or not. For this purpose, DDMP was synthesized from maltol instead of from glucose for the first time. In contrast, DDMP derived from glucose was also prepared and further purified. Their structures were identified by NMR and MS, and considered to be the same substance. The sensory analysis showed that DDMP derived from maltol was tasteless. Further studies indicated that some impurities in Maillard reaction made DDMP derived from glucose taste bitter.

Full text of DOI:10.1016/j.foodchem.2021.130052

Food Chemistry 361 (2021) 130052
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Synthesis of 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one from
maltol and its taste identification
,
,
Zhifei Chena 1, Gaolei Xia 1, Yufeng Fua, Qingfu Wanga, Lili Caia, Zhiwei Zhaoa, Qiang Liua,
Bing Baib *, Yuping Maa
,
,*
a Technology Center, China Tobacco Henan Industrial Co., Ltd., Zhengzhou 450016, PR China
b School of Food & Biological Engineering, Zhengzhou University of Light Industry, Zhengzhou 450002, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Maillard reaction
DDMP
2,3-Dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one (DDMP) exists in many foods, and its effect on taste is
controversial. The aim of this study was to clarify whether DDMP has bitter taste or not. For this purpose, DDMP
was synthesized from maltol instead of from glucose for the first time. In contrast, DDMP derived from glucose
was also prepared and further purified. Their structures were identified by NMR and MS, and considered to be
the same substance. The sensory analysis showed that DDMP derived from maltol was tasteless. Further studies
indicated that some impurities in Maillard reaction made DDMP derived from glucose taste bitter.
Maltol
Taste
Bitter
1. Introduction
Jiang & Peterson, 2013; Li, Wu, Tang, & Yu, 2019b; Li et al., 2019a). In
these papers the DDMP as reference substance was derived from glucose
Maillard reaction leads to large number of different reactive in-
termediates or final products, including 2,3-dihydro-3,5-dihydroxy-6-
methyl-4H-pyran-4-one (DDMP). DDMP was first isolated from stored
orange juice powder (Tatum, Shaw, & Berry, 1967; Mills, Weisleder, &
Hodge, 1970). Later, DDMP was discovered in various foods, such as
garlic oil (Sun et al., 2019), Ipomoea staphylina (Padmashree et al.,
2018), rose tea (Qin et al, 2019), heated pear (Lee et al., 2013), mango
through the Maillard reaction (Van Den Ouweland & Peer, 1970). It is
accepted that the product mixtures derived from Maillard reaction are
complex, which makes separation and purification difficult. The so-
called purified DDMP may be accompanied by trace impurities, which
cause deviations in taste judgment. Therefore, it is necessary to syn-
thesize DDMP through a step-by-step reaction to clarify academic con-
troversies. In this paper, maltol was used as the starting material to
synthesize DDMP through a 5-step reaction, and its taste was assessed.
ˇ
´
(Ali et al., 2012), prune (Cechovska et al., 2011), etc. In the Maillard
reaction, DDMP is generally considered to be converted from 1-DG, and
can further form furan and pyran compounds (Voigt & Glomb, 2009;
Voigt, Smuda, Pfahler, & Glomb, 2010; Zhou et al., 2011; Li et al., 2019).
In terms of physiological activity, although DDMP is thought to generate
free radicals to damage DNA through self-oxidation (Hiramoto et al.,
1997), more evidence shows that DDMP has a strong antioxidant effect
2. Materials and methods
2.1. Reagents and chemicals
Maltol (99%) and 5% Pd/C were purchased from J&K Scientific Ltd.
Pb(OAc)4 was purchased from Acros Organics. Lipase from Candida
rugosa, lipase from porcine pancreas and cation exchange resin were
purchased from Sigma-Aldrich. Novozym 435 was purchased from
Novozymes Biologicals, Inc. NaHCO3, NaCl, anhydrous Na2SO4, Na2CO3
and the solvents were of analytical grade.
ˇ
´
(Kanzler et al., 2016; Yu et al., 2013a, 2013b; Cechovska et al., 2011; Lee
et al., 2011) and other activities (Ban et al., 2007; Beppu et al., 2012;
Teoh, Mashitah, & Ujang, 2011).
So far there has been considerable research on DDMP, yet DDMP is
still a controversial compound. Shaw reported that DDMP is odorless
(Shaw, Tatum, & Berry, 1971), but recent reports claimed that DDMP
has a strong bitter taste and is one of the main bitter substances in food
processing (Bin, Jiang, Cho, & Peterson, 2012; Bin & Peterson, 2016;
TLC was performed on GF254 precoated plates and visualized using
phosphomolybdic acid dip or ultraviolet light. Column chromatography
was performed on silica gel 60.
* Corresponding authors.
E-mail addresses: baibing82@163.com (B. Bai), mayp159659@126.com (Y. Ma).
1
These authors contributed equally to this work.
https://doi.org/10.1016/j.foodchem.2021.130052
Received 14 March 2021; Received in revised form 2 May 2021; Accepted 4 May 2021
Available online 13 May 2021
0308-8146/© 2021 Elsevier Ltd. All rights reserved.

Z. Chen et al.
Food Chemistry 361 (2021) 130052
2.2. Synthesis of DDMP from maltol
storage in the refrigerator. [
α
]
D + 132.3 (c 0.60, CHCl3). 1H NMR (600
MHz, CDCl3) δ 4.49 (dd, J = 10.5, 5.9 Hz, 1H, H-2a), 4.44 (dd, J = 12.1,
5.9 Hz, 1H, H-3), 4.06 (dd, J = 12.1, 10.5 Hz, 1H, H-2b), 2.12 (s, 3H, H-
7). 13C NMR (150 MHz, CDCl3) δ 188.1 (C-4), 160.3 (C-6), 131.3 (C-5),
70.9 (C-2), 67.1 (C-3), 15.8 (C-7). HR-ESI-MS m/z calcd for C6H8O4 [M
+ H]+ 144.0495 , found 145.0494.
Maltol (10 g, 80 mmol) was dissolved in acetic anhydride (15 mL),
the mixture was heated at 90 ◦C for 6 h, then cooled and ethyl acetate
(200 mL) was added. The ethyl acetate solution was washed with
saturated aqueous NaHCO3 (1 × 100 mL), then aqueous NaCl (2 × 200
mL), dried over anhydrous Na2SO4 then evaporated under reduced
pressure to give 2 as a yellow oil 13.2 g, yield 98%. 1H NMR (600 MHz,
CDCl3) δ 7.69 (d, J = 5.7 Hz, 1H, H-2), 6.40 (d, J = 5.7 Hz, 1H, H-3), 2.34
(s, 3H, -OCH3), 2.27 (s, 3H, H-7). 13C NMR (150 MHz, CDCl3) δ 172.0 (C-
2.3. Synthesis and further purification of DDMP from glucose
DDMP was synthesized via Maillard reaction (Van Den Ouweland &
Peer, 1970; Kanzler et al., 2016) with slight modifications. A mixture of
0.2 mol D-glucose, 0.2 mol piperidine, and 150 mL ethanol was refluxed
for 1.5 h. Then, 0.2 mol of acetic acid in 30 mL of ethanol were added
slowly, and the mixture was heated at 75 ◦C for 32 h. After ethanol was
evaporated under reduced pressure, the residue was taken up in 100 mL
of water, and the precipitate was formed immediately. The mixture was
filtered to obtain the precipitate and the filtrate, and then the filtrate was
extracted with ethyl acetate (6 × 150 mL). The combined organic layer
was dried over anhydrous Na2SO4, and then evaporated under reduced
pressure. The crude product was purified by polyamide 6 column
chromatography (petroleum ether /EtOAc = 1/1) and silica gel 60
column chromatography (CH2Cl2/CH3OH = 50:1, v/v) in sequence, and
then recrystallized from ether/light petroleum to obtain DDMP as a
beige solid. In the process, a large amount of piperidinohexose reduc-
tone 8 was obtained from the precipitate.
–
4), 167.6 (-O-C O), 159.1 (C-2), 154.3 (C-6), 138.7(C-5), 116.8 (C-3),
–
20.3 (-OCH3), 15.0 (C-7). HR-ESI-MS m/z calculated for C8H8O4 [M +
H]+ 169.0495, found 169.0494.
2 (9 g, 54 mmol), 5% Pd/C (1 g), and 120 mL of ethyl acetate were
added to a round-bottomed flask. The mixture was stirred under a
hydrogen atmosphere at room temperature for 3 h, then the solution was
filtered and evaporated under reduced pressure. The residue was puri-
fied by silica gel 60 column chromatography. Elution was conducted
with petroleum ether petroleum ether/EtOAc = 3/1, and 40-mL frac-
tions were collected; fractions 40–75 (product by TLC analysis, petro-
leum ether/EtOAc = 2/1, Rf value 0.45) were combined and the solvent
was removed to give a colorless oil 6.7 g, yield 75%. 1H NMR (600 MHz,
CDCl3) δ 4.50 (t, J = 6.8 Hz, 2H, H-2), 2.68 (t, J = 6.8 Hz, 2H, H-3), 2.27
(s, 3H, -OCH3), 1.99 (s, 3H, H-7). 13C NMR (150 MHz, CDCl3) δ 183.9 (C-
–
4), 168.5(-O-C O), 166.9(C-6), 129.3 (C-5), 67.7(C-2), 35.6(C-3), 20.3
–
(-OCH3), 16.0 (C-7). HR-ESI -MS m/z calculated for C8H10O4 [M + H]+
171.0652, found 171.0655.
The above beige DDMP was chromatographed on an LH-20 gel col-
umn (bed volume 60 mL). Elution was conducted with 1:1 dichloro-
methane/methanol, and 2-mL fractions were collected after one bed
volume; fractions 1–8 were combined and the solvent was removed to
give F-I fraction as brown viscous, yield 5%; Fractions 11–13 (product
by TLC analysis) were combined and the solvent was removed to give F-
II fraction as a white solid, yield 94%. The F-II fraction was identified as
high-purity DDMP.
Pb(OAc)4 (31 g, 70 mmol) was added to a solution of 3 (6 g, 35
mmol) in dry toluene (150 mL) under N2. The reaction mixture was
heated to 90 ◦C and stirred for 15 h, then cooled and washed with
aqueous NaCl (3 × 150 mL), dried over anhydrous Na2SO4. The solvent
was removed under reduced pressure and the residue was purified by
silica gel 60 column chromatography. Elution was conducted with
CH2Cl2/EtOAc = 50/1, and 30-mL fractions were collected; fractions
80–120 (product by TLC analysis, CH2Cl2/EtOAc = 20/1, Rf value 0.38)
were combined and the solvent was removed to give a colorless oil 2.7 g,
yield 35%. 1H NMR (600 MHz, CDCl3) δ 5.43 (dd, J = 8.6, 4.8 Hz, 1H, H-
3), 4.52 (dd, J = 12.0, 4.8 Hz, 1H, H-2a), 4.45 (dd, J = 12.0, 8.6 Hz, 1H,
2.4. Synthesis of DDMP-5-camphorsulfonate
(+)-Camphorsulfonyl chloride (262 mg, 1.05 mmol) was added to a
solution of DDMP (144 mg, 1.0 mmol) and Et3N (121 mg, 1.2 mmol) in
dry CH2Cl2 (10 mL) under N2. The reaction mixture was stirred at room
temperature for 6 h, then cooled and washed with aqueous NaCl (3 ×
150 mL), and dried over anhydrous Na2SO4. The solvent was removed
under reduced pressure and the residue was purified by silica gel 60
column chromatography. Elution was conducted with petroleum ether/
EtOAc = 4/1, and 6-mL fractions were collected; fractions 25–34
(product by TLC analysis, petroleum ether/EtOAc = 2/1, Rf value 0.40)
were combined and the solvent was removed to give a colorless solid
310 mg, yield 87%. 1H NMR (600 MHz, CDCl3) δ 4.62 (dd, J = 11.0, 6.2
Hz, 1H, H-2a), 4.41 (dd, J = 13.4, 6.2 Hz, 1H, H-3), 4.16 (dd, J = 13.3,
11.0 Hz, 1H, H-2b), 4.01 (d, J = 15.0 Hz, 1H, -SO3-CH2–), 3.66 (d, J =
15.0 Hz, 1H, -SO3-CH2–), 2.46–2.43 (m, 1H, H-6‘a), 2.42–2.40 (m, 1H,
H-3‘a), 2.22 (s, 3H, H-7), 2.15–2.13 (m, 1H, H-4‘), 2.11–2.06 (m, 1H, H-
5‘a), 1.97 (d, J = 18.5 Hz, 1H, H-3‘b), 1.80–1.75 (m, 1H, H-6‘b),
1.48–1.44 (m, 1H, H-5‘b), 1.14 (s, 3H, C-7‘-CH3), 0.93 (s, 3H, C-7‘-CH3).
13C NMR (150 MHz, CDCl3) δ 213.7 (C-2‘), 186.7 (C-4), 173.3 (C-6),
126.9 (C-5), 70.8 (C-2), 66.9 (C-3), 58.0 (C-7‘), 49.7 (-SO3-CH2–), 47.9
(C-1‘), 42.7 (C-4‘), 42.3 (C-3‘), 26.7 (C-5‘), 25.0 (C-6‘), 19.6 (C-7‘-CH3),
19.5 (C-7‘-CH3), 17.3 (C-7). HR-ESI-MS m/z calculated for C16H22O7S
[M + H]+ 359.1159, found 359.1163.
H-2b), 2.27 (s, 3H, -OCH3), 2.16 (s, 3H, -OCH3), 2.04 (s, 3H, H-7). 13
C
–
–
–
–
NMR (150 MHz, CDCl ) δ 179.9 (C-4), 169.6(-O-C O), 168.3(-O-C O),
3
167.9 (C-6), 128.2 (C-5), 69.3(C-2), 68.0(C-3), 20.7(-OCH3), 20.2
(-OCH3), 16.3 (C-7). HR-ESI-MS m/z calculated for C10H12O6 [M +
NH4]+ 246.0972, found 246.0972.
Lipase (50 mg, derived from Candida rugosa) was added to a solution
of 4 (2 g, 9 mmol) in water (20 mL). The mixture was stirred at room
temperature for 3 h, then the aqueous phase was extracted with CH2Cl2
(3 × 20 mL). The combined organic layer was dried over anhydrous
Na2SO4, and then evaporated under reduced pressure. The residue was
purified by silica gel 60 column chromatography. Elution was conducted
with petroleum ether petroleum ether/EtOAc = 3/1, and 15-mL frac-
tions were collected; fractions 50–80 (product by TLC analysis, petro-
leum ether /EtOAc = 1/1, Rf value 0.42) were combined and the solvent
was removed to give enantiomer (+)-5 as a colorless oil 0.8 g, yield 48%
(unreacted enantiomer (–)-4 was recovered with a yield 49%). [α] +
D
74.3 (c 0.75, CHCl3). 1H NMR (600 MHz, DMSO) δ 4.43 (dd, J = 11.6,
4.6 Hz, 1H, H-2a), 4.20 (dd, J = 11.6, 8.8 Hz, 1H, H-2b), 4.12 (dd, J =
8.8, 4.5 Hz, 1H, H-3), 2.21 (s, 3H, -OCH3), 1.95 (s, 3H, H-7). 13C NMR
–
–
(150 MHz, DMSO) δ 185.7 (C-4), 168.6 (-O-C O), 166.8 (C-6), 127.4
(C-5), 72.4 (C-2), 67.5 (C-3), 20.5 (-OCH3), 16.2 (C-7). HR-ESI-MS m/z
calcd for C8H10O5 [M + H]+ 187.0601, found 187.0601.
2.5. Characterization
Na2CO3 (0.29 g, 2.7 mmol) was added to a solution of (+)-5 (0.5 g,
2.7 mmol) in water (10 mL) at 0 ◦C. The reaction mixture was warmed
up to room temperature and stirred for 4 h. When the reaction was
complete, Cation exchange resin was added and stirred for another 5
min then filtered. The filtrate was removed under reduced pressure to
give 1 as colorless oil 0.35 g, yield 90%. The product became solid after
2.5.1. NMR analysis
The NMR analyses were performed on a Bruker 600 MHz spec-
trometer with a CryoProbe 5 mm QCI probe. D2O, DMSO‑d6, MeOH‑d4
and chloroform-d were used as solvents, and tetramethylsilane was the
internal standard.
2

Z. Chen et al.
Food Chemistry 361 (2021) 130052
2.5.2. HR-MS analysis
hydrogenation.
The HR-ESI-MS analyses were performed on an AB Sciex Triple TOF
6600. Spectra were acquired in turbospray ESI mode at a flow rate of 0.2
mL minꢀ 1. The mass spectrometer was operated in the full-scan mode,
monitoring positive ions from m/z 70 to 450.
Firstly, the commonly used catalytic systems Pd/C, Raney nickel, tris
(triphenylphosphine) rhodium chloride (Wilkinson’s catalyst) and so-
dium borohydride-nickel chloride (NaBH4-NiCl2) were tested in the
hydrogenation. The results indicated that Wilkinson’s catalyst cannot
catalyze the reduction of maltol acetate; And when Raney Nickel and
NaBH4-NiCl2 systems were used as catalysts, the major products were
not 3, but over-reduced products; Encouragingly, the Pd/C system could
convert half of the substrate into dihydromaltol acetate. Therefore, the
Pd/C system was more suitable for hydrogenation of maltol acetate.
Further optimization of the reaction conditions was conducted by
examining the effect of solvents and catalyst amounts. Methanol,
ethanol and acetonitrile were not suitable for the reaction because they
produced a lot of over-reduced products. Using acetic acid as an additive
fails to inhibit the formation of over-reduction products. Ethyl acetate
provided good result that overall, 80% dihydromaltol acetate was
formed. Surprisingly, when the catalyst amount was reduced to 5%, the
complete conversion time was extended to 5 h, and the proportion of
dihydromaltol acetate was also slightly declined.
2.5.3. GC–MS analysis
The GC–MS analyses were performed on an Agilent 7890B–5975 N
gas chromatograph–mass spectrometer. The column used was a fused
silica column (DB-5MS, 30 m, 0.25 mm i.d., 0.25 μm film thickness); the
carrier gas was helium at 1.2 mL/min; the temperature program was
◦
◦
◦
heating for 2 min at 50 C, 50 to 280 C at 5 C/min; injection tem-
◦
◦
perature, 280 C; transfer line temperature of 240 C; ion source tem-
perature, 120 ◦C; ionization energy, 70 eV; mass scan range, m/z
30–450.
2.5.4. Optical rotation analysis
The optical rotation analyses were performed on an SGWzz-1
◦
polarimeter. The sample tube length, 20 cm; temperature, 20 C; sol-
vent, CHCl3.
From the results obtained, it seemed that the hydrogenation of
maltol acetate carried out in ethyl acetate with 10% Pd/C was optimal.
Under these conditions, maltol acetate was about 80% converted into
the target product, and the separation yield was about 75%.
2.6. Sensory Analyses.
Training of the sensory panel and determination of taste followed the
methods previously described (Ottinger, & Hofmann, 2001; Bin, &
Peterson, 2016): Assessors were trained to evaluate bitter taste intensity
using quinine hydrochloride (0.05 mmol/L). Sensory analyses were
performed in a panel room at 22–25 ◦C in three different sessions. The
detection thresholds of taste compounds were determined in a triangle
test using tap water (pH 6.5) as the solvent. The samples were presented
in order of increasing concentrations (serial 1:1 dilutions), and the
threshold values evaluated in three different sessions by six panelists
whose data were averaged.
3.1.2. Acetoxylation of dihydromaltol acetate
The 13C NMR chemical shift of carbonyl group from dihydromaltol
acetate is about δ 184, which indicates that the α position is inactive and
is unlikely to participate in
failed direct -oxidation reaction, we turned our attention to the ace-
toxylation reaction. Commonly used reagents for carbonyl -position
α-oxidation or α-acetoxylation. After the
α
α
acetyl acetoxylation are manganese triacetate (Adachi, Hasegawa,
Katakawa, & Kumamoto, 2017), lead tetraacetate (Teng et al., 2006)
and potassium permanganate/acetic acid system (Marín-Barrios et al.,
2014). Only the lead tetraacetate system was effective. However, Pb
(OAc)4 failed to completely convert the substrate 3, and the target
product 4 was further eliminated to produce maltol acetate 2. For this
reason, we have optimized the amount of lead tetraacetate, solvent,
temperature and reaction time.
3. Results and discussion
3.1. Synthesis of DDMP
The optimization results are summarized in Table 1. At the outset we
studied the acetoxylation of dihydromaltol acetate in toluene. It pro-
ceeded with 40% conversion to 4. Cyclohexane and benzene provided
neither suitable yield nor selectivity. Thus, toluene was used to test the
effect of the other conditions. The 1.5 equivalents of lead tetraacetate
produced the target product with highest yield in 15 h at 90 ◦C, and the
target product 4 remained, even after reaction periods of 30 h and the
use of more catalyst and higher temperature. Under these conditions,
about 40% of dihydromaltol acetate was converted into 4, and the
The structure of DDMP and maltol are highly similar, so it is feasible
to synthesize DDMP from maltol. The conjugated structure of maltol
makes its properties very stable, and it is impossible to synthesize DDMP
through the hydroboration-oxidation reaction directly. Therefore, the
strategy we adopted is to selectively reduce the double bond at C-2,3
positions, then perform acetylation or oxidation at the C-3 position, and
finally obtain the target product through hydrolysis (Scheme 1).
Although an attempt by Mills et al. to utilize the strategy was defeated, it
is still worth taking advantage of new synthetic technology to try again.
3.1.1. Hydrogenation of maltol acetate
Table 1
Maltol acetate has a structure with two double bonds, which is
difficult to maintain in the form of dihydromaltol acetate during
reduction. Therefore, based on the poor selectivity, it is necessary to
optimize the catalytic system and reaction conditions of the
Acetoxylation conditions.
Entry
Pb(OAc)4 (equiv.)
Solvent
Time
T (◦C)
4/3/2 a
1
2
toluene
benzene
cyclohexane
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
30
30
30
30
30
30
30
30
25
20
15
10
15
15
15
100
80
4/4/2
3/4/4
0.5/9.5/0
2/6/2
3/5/2
4/4/2
4/4/2
4/4/2
4/4/2
4/4/2
4/4/2
3/6/1
3/6/1
4/4/2
4/4/2
2
2
3
2
100
100
100
100
100
100
100
100
100
100
80
4
0.5
1
5
6
1.5
2.5
3
7
8
9
1.5
1.5
1.5
1.5
1.5
1.5
1.5
10
11
12
13
14
15
90
110
Scheme 1. Synthesis route of DDMP.
Note: a. The ratio of compound 4, 3 and 2 was determined by GC–MS.
3

Z. Chen et al.
Food Chemistry 361 (2021) 130052
isolated yield of the target product was 35%.
3.1.3. Hydrolysis of DDMP diacetate
Generally, sodium methoxide/methanol, potassium carbonate/
methanol, sodium hydroxide/methanol–water and other alkali-
catalyzed systems can hydrolyze ester groups quickly and efficiently.
Unfortunately, attempts to convert the diacetate 4 into 1 by hydrolysis
with base or acid were not successful (Table 2). In the catalytic system
using methanol as the solvent, sodium methoxide, potassium carbonate,
and DBU can quickly catalyze the ester hydrolysis. However, the target
product would convert into by-product, and the complex mixture could
not be separated by column chromatography. When using water as the
solvent, there was almost no measurable amount of deacetylated 4
formed. It may be because the 3-acetyl group in the structure of DDMP
diacetate is unstable, and it is easy to eliminate reaction under alkaline
conditions and further generate various impurities.
Lipase-catalyzed ester synthesis and hydrolysis have been widely
used in industrial fields, and are environmentally friendly and easy to
operate. Therefore, the enzymatic hydrolysis of diacetate 4 was
considered to be an effective way to replace the failed chemical
methods. Among the selected enzymes, lipase from Candida rugosa could
hydrolyze the 3-acetyl group completely, but had no effect on 5-acetyl
group, while lipase from porcine pancreas had no effect on either. It is
noteworthy that the lipase from Candida rugosa displayed high enan-
tioselectivity, and the optically pure enantiomer (+)-5 was obtained as
with an enantiomeric excess value of up to 98%. Novozym 435 could
hydrolyze both 3-acetyl group and 5-acetyl group slowly; however, the
poor selectivity led to separation difficulties.
The hydrolysis of DDMP-3-acetate was carried out with sodium
carbonate in water, and DDMP was obtained with a 90% yield through a
simple separation.
3.2. Identification of DDMP from different precursors
D-Glucose has been a widely used precursor of DDMP. The 1H NMR
Fig. 1. 1H NMR and 13C NMR spectra (in methanol‑d4) of DDMP.
spectrum in Fig. 1 showed signals for three coupling protons (δH 4.08,
4.18, 4.33), one independent methyl group (δH 2.04, H-7), while the 13
C
stereoisomers have different sensory properties, for examples D- and L-
isomers of certain amino acids taste different. Perhaps the differences in
reported DDMP sensory properties may be due to differences in the
stereochemistry of the products. So, it was necessary to determine
whether the absolute configuration of DDMP obtained by different
methods is the same.
NMR spectrum exhibited 6 carbon signals including one unsaturated
ketone carbon, two olefinic carbons. COSY correlations established the
subunits from H-2 to H-3, while their connectivities were completed by
detailed analysis of HSQC correlations from the protons H-2a (δH 4.08)
and H-2b (δH 4.33) to the secondary carbon C-2 (δ 72.6), from H-3 (δ
C
H
4.18) to tertiary C-3 (δC 69.0). We compared the NMR and MS spectrum
of DDMP derived from maltol with DDMP derived from glucose, and
found that there was no difference between the two. At the same time,
we used different deuterated solvents to perform NMR tests on maltol-
derived DDMP, and the results were consistent with those reported in
the literature (in methanol‑d4, Kanzler, et al., 2016; in deuterium oxide,
(Li et al., 2019a); in chloroform-d, Mills et al., 1970).
In the preparation of DDMP from D-glucose, the chirality of the C-5
position of D-glucose will be directly transferred to the C-3 position of
DDMP. We also observed that the optical rotation of DDMP is (+)
rotatory direction (DDMP from L-glucose is (–) rotatory direction,
respectively). At the same time, after the condensation of DDMP with
(+)-camphorsulfonyl chloride, the sulfonate product is a single com-
pound rather than epimers, indicating that the DDMP from glucose is a
single (+)-isomer rather than racemate. In addition, 1H NMR spectra
showed that there was no H/D exchange in the C-3 position of DDMP in
protic solvents (D2O and methanol‑d4), which indicated that enolization
cannot occur in this position, and its chirality is very stable.
In the synthesis of DDMP from maltol, due to the good enantiose-
lectivity of lipase for the hydrolysis of 3-ester group of racemic diacetate
4, the final product DDMP also has optical activity. The (+) optical
rotation in the experiment was identical with that of DDMP derived from
glucose, so these two compounds were confirmed to be the same abso-
lute configuration.
3.3. Stereochemistry
DDMP contains a chiral center in the C-3 position. It is known that
Table 2
Hydrolysis of DDMP-3-acetate.
Entry
Catalyst
Time (h)
Conversion (%)
1
2
4
5
6
7
8
9
CH3ONa/CH3OH
K2CO3/CH3OH
0.2
1
complex a
complex
complex
complex
complex
100% b
DBU/ CH3OH
2
Na2CO3/H2O
0.5
0.2
1
3.4. Taste evaluation
NaOH/H2O
Lipase from Candida rugosa
Lipase from porcine pancreas
Novozym 435
10
12
no reaction
We conducted a taste test on the DDMP obtained by the two methods
(Table 3). The results showed that the DDMP derived from glucose
exhibited a bitter taste at a threshold of 1 mmol/kg of water, while the
100% b
Note: a. Conversion was not determined. b. Single enantiomer conversion.
4

Z. Chen et al.
Food Chemistry 361 (2021) 130052
Table 3
Taste quality and taste threshold.
Entry
Compounds
Taste
Taste threshold (mmol/kg,
in water)
quality
1
2
3
4
5
DDMP-M a
tasteless
bitter
–
DDMP-G b
1.2
0.08
–
F-I
bitter
F-II
tasteless
bitter
pyrrolidinohexose
0.5 d
reductone 6c
Fig. 2. Structures of pyrrolidinohexose reductone (6), pyrrolidinohexose
reductone (7) and piperidinohexose reductone (8).
6
7
7c
8c
bitter
bitter
0.06d
0.8
Declaration of Competing Interest
Note: a. DDMP derived from maltol. b. DDMP derived from glucose. c. The
structure is displayed in Fig. 2. d. The taste threshold was taken from Ottinger
et al.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
DDMP derived from maltol was tasteless even at 50 mmol/kg. This
contradiction leads us to find the answer from the Maillard reaction
which generates extremely complex products. The bitter compound
pyrrolidinohexose reductone 6 (Fig. 2), formed by Maillard reactions
from glucose and proline, showed low bitter taste threshold. Another
reductone 8, together with DDMP, both formed from glucose and
piperidine, also showed bitter taste. There are good reasons for thinking
that glucose-derived DDMP is mixed with other substances, which
causes bitterness.
Acknowledgments
This work was partially supported by Technology Center, China
Tobacco Henan Industrial Co., Ltd..
References
Adachi, K., Hasegawa, S., Katakawa, K., & Kumamoto, T. (2017). Total synthesis of (+)-
blennolide C and (+)-gonytolide C via spirochromanone. Tetrahedron Letters, 58(47),
4479–4482.
Accordingly, Sephadex LH-20 chromatography was used to further
purify the glucose-derived DDMP, and the brown F-I fraction and white
F-II fraction were obtained. The F-I fraction flowed out firstly, and its
amount was rare, but exhibited an intense bitter taste at a low threshold
of 0.07 mmol/kg of water, which was close to the threshold of bis
(pyrrolidino)hexose reductone 7 (0.06 mmol/kg of water, Ottinger, &
Hofmann, 2001). The F-II fraction afterwards was pure DDMP without
color and bitterness. The results indicated that pure DDMP is a tasteless
compound.
Ali, M. R., Yong, M. J., Gyawali, R., Mosaddik, A., Ryu, Y. C., & Cho, S. K. (2012). Mango
(Mangifera indica L.) peel extracts inhibit proliferation of HeLa human cervical
carcinoma cell via induction of apoptosis. Journal of the Korean Society for Applied
Biological Chemistry, 55(3), 397–405.
Ban, J. O., Hwang, I. G., Kim, T. M., Hwang, B. Y., Lee, U. S., Jeong, H. S., … Hong, J. T.
(2007). Anti-proliferate and pro-apoptotic effects of 2,3-dihydro-3,5-dihydroxy-6-
methyl-4H-pyranone through inactivation of NF-KB in human colon cancer cells.
Archives of Pharmacal Research, 30, 1455–1463.
Beppu, Y., Komura, H., Izumo, T., Horii, Y., Shen, J., Tanida, M., … Nagai, K. (2012).
Identification of 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one isolated from
lactobacillus pentosus strain S-PT84 culture supernatants as a compound that
stimulates autonomic nerve activities in rats. Journal of Agricultural and Food
Chemistry, 60, 11044–11049.
So far several types of bitter tastants formed during thermal food
processing from Maillard reactions have been reported (Frank, Jezussek,
& Hofmann, 2003; Wakamatsu, Stark, & Hofmann, 2016), such as pyr-
rolidinohexose reductone, quinizolate and (2R)-3-(allylthio)-2-((4S)-4-
(allylthiomethyl)-6-formyl-3-oxo-3,4-dihydropyrrolo-[1,2-a]pyrazin-2
(1H)-yl) propanoic acid. Typically, most of the bitter tastants are ni-
trogen compounds. As for the F-I fraction, its composition was too
complex to be fully clarified, although some piperidine substituents
were detected by NMR spectra and GC–MS.
Bin, Q., Jiang, D. H., Cho, I. H., & Peterson, D. G. (2012). Chemical markers for bitterness
in wheat bread. Flavour and Fragrance Journal, 27, 454–458.
Bin, Q., & Peterson, D. G. (2016). Identification of bitter compounds in whole wheat
bread crumb. Food Chemistry, 203, 8–15.
ˇ
´
ˇ
ˇ
Cechovska, L., Cejpek, K., Konecný, M., & Velísek, J. (2011). On the role of 2,3-dihydro-
3,5-dihydroxy-6-methyl-(4H)-pyran-4-one in antioxidant capacity of prunes.
European Food Research and Technology, 233(3), 367–376.
Frank, O., Jezussek, M., & Hofmann, T. (2003). Sensory activity, chemical structure, and
synthesis of Maillard generated bitter-tasting 1-oxo-2,3-dihydro-1H-indolizinium-6-
olates. Journal of Agricultural and Food Chemistry, 51(9), 2693–2699.
Hiramoto, K., Nasuhara, A., Michikoshi, K., Kato, T., & Kikugawa, K. (1997). DNA strand-
breaking activity and mutagenicity of 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-
pyran-4-one (DDMP), a Maillard reaction product of glucose and glycine. Mutation
Research, 395(1), 47–56.
4. Conclusion
DDMP exists in many foods and is undoubtedly a very important
compound. For the first time, DDMP was synthesized from maltol and
obtained as a pure substance, and enzymatic synthesis played an
important role in this process. Pure DDMP is tasteless with no bitter
attribute. DDMP derived from glucose tastes bitter due to impurities
produced in the Maillard reaction.
Jiang, D., & Peterson, D. G. (2013). Identification of bitter compounds in whole wheat
bread. Food Chemistry, 141(2), 1345–1353.
Kanzler, C., Haase, P. T., Schestkowa, H., & Kroh, L. W. (2016). Antioxidant properties of
heterocyclic intermediates of the Maillard reaction and structurally related
compounds. Journal of Agricultural and Food Chemistry, 64(41), 7829–7837.
Lee, S. H., Oh, S. H., Hwang, C. R., Woo, K. S., Jeong, H. S., Kim, H. Y., … Kim, D. J.
(2011). Isolation and identification of an antioxidant substance from heated onion
(Allium cepa L.). Journal of the Korean Society of Food Science and Nutrition, 40,
470–474.
CRediT authorship contribution statement
Lee, S. H., Woo, K. S., Jeong, H. S., Kim, H. Y., Lee, J. S., & Hwang, S. G. (2013). Isolation
and identification of the antioxidant DDMP from heated pear (Pyrus pyrifolia Nakai).
Preventive Nutrition and Food Science, 24, 21–33.
Zhifei Chen: Conceptualization, Methodology, Formal analysis,
Investigation, Writing - original draft, Writing - review & editing,
Visualization. Gaolei Xi: Conceptualization, Methodology, Formal
analysis, Resources, Data curation, Writing - original draft, Writing -
review & editing. Yufeng Fu: Formal analysis, Investigation, Resources.
Qingfu Wang: Validation, Data curation. Lili Cai: Methodology, Visu-
alization. Zhiwei Zhao: Formal analysis, Investigation. Qiang Liu:
Formal analysis, Data curation. Bing Bai: Conceptualization, Supervi-
sion, Writing - review & editing. Yuping Ma: Supervision, Project
administration, Conceptualization.
Li, H.e., Tang, X.-Y., Wu, C.-J., & Yu, S.-J. (2019). Formation of 2,3-dihydro-3,5-dihy-
droxy-6-methyl-4(H)-pyran-4-One (DDMP) in glucose-amino acids Maillard reaction
by dry-heating in comparison to wet-heating. LWT-Food Science and Technology, 105,
156–163.
Li, H.e., Wu, C.-J., Tang, X.-Y., & Yu, S.-J. (2019a). Insights into the regulation effects of
certain phenolic acids on 2,3- dihydro-3,5-dihydroxy-6-methyl-4(H)-pyran-4-one
formation in a microaqueous glucose-proline system. Journal of Agricultural and Food
Chemistry, 67(32), 9050–9059.
Li, H.e., Wu, C.-J., Tang, X.-Y., & Yu, S.-J. (2019b). Determination of four bitter
compounds in caramel colors and beverages using modified QuEChERS coupled with
liquid chromatography-diode array detector-mass spectrometry. Food Analytical
Methods, 12(7), 1674–1683.
5

Z. Chen et al.
Food Chemistry 361 (2021) 130052
Marín-Barrios, R., García-Cabeza, A. L., Moreno-Dorado, F. J., Guerra, F. M., &
Massanet, G. M. (2014). Acyloxylation of cyclic enones: Synthesis of densely
oxygenated guaianolides. Journal of Organic Chemistry, 79(14), 6501–6509.
Mills, F. D., Weisleder, D., & Hodge, J. E. (1970). 2,3-Dihydro-3,5-dihydroxy-6-methyl-
4H-pyran-4-one, a novel nonenzymatic browning product. Tetrahedron Letters, 11,
1243–1246.
Teng, W. D., Huang, R., Kwong, C. K. W., Shi, M., Patrick, H., & Toy, P. H. (2006).
Influence of Michael acceptor stereochemistry on intramolecular Morita-Baylis-
Hillman reactions. Journal of Organic Chemistry, 71, 368–371.
Teoh, Y. P., Don, M. M., & Ujang, S. (2011). Media selection for mycelia growth,
antifungal activity against wood-degrading fungi, and GC-MS study by Pycnoporus
sanguineus. Bioresources, 6(3), 2719–2731.
Ottinger, O. F. H., & Hofmann, T. (2001). Characterization of an intense bitter-tasting
1H,4H-quinolizinium-7-olate by application of the taste dilution analysis, a novel
bioassay for the screening and identification of taste-active compounds in foods.
Journal of Agricultural and Food Chemistry, 49, 231–238.
Voigt, M., & Glomb, M. A. (2009). Reactivity of 1-deoxy-d-erythro-hexo-2,3-diulose: A
key intermediate in the Maillard chemistry of hexoses. Journal of Agricultural and
Food Chemistry, 57, 4765–4770.
Voigt, M., Smuda, M., Pfahler, C., & Glomb, M. A. (2010). Oxygen-dependent
fragmentation reactions during the degradation of 1-deoxy-d-erythro-hexo-2,3-
diulose. Journal of Agricultural and Food Chemistry, 58, 5685–5691.
Wakamatsu, J., Stark, T. D., & Hofmann, T. (2016). Taste-active Maillard reaction
products in roasted garlic (Allium sativum). Journal of Agricultural and Food
Chemistry, 64(29), 5845–5854.
Padmashree, M. S., Ashwathanarayana, R., Raja, N., & Roopa, B. (2018). Antioxidant,
cytotoxic and nutritive properties of Ipomoea staphylina Roem & Schult. plant
extracts with preliminary phytochemical and GCMS analysis. Asian Journal of
Pharmacy and Pharmacology, 4, 473–492.
Qin, H., Li, B.-C., Dai, W.-F., Xiang, C., Qin, Y.i., Jiao, S.-Y., & Zhang, M.i. (2019). Rapid
determination of antioxidant molecules in volatiles of rose tea by gas
chromatography–mass spectrometry combined with DPPH reaction. Journal of Food
Science and Technology, 56(9), 4009–4015.
van den Ouweland, G. A. M., & Peer, H. G. (1970). Synthesis of 3,5-dihydroxy-2-methyl-
5,6-dihydropyran-4-one from aldohexoses and secondary amine salts. Recueil des
Travaux Chimiques des Pays-Bas, 89(7), 750–754.
Shaw, P. E., Tatum, J. H., & Berry, R. E. (1971). 2,3-Dihydro-3,5-dihydroxy-6-methyl-4H-
pyran-4-one, a degradation product of a hexose. Carbohydrate Research, 16(1),
207–211.
Yu, X., Zhao, M., Liu, F., Zeng, S., & Hu, J. (2013a). Identification of 2,3-dihydro-3,5-
dihydroxy-6-methyl-4H-pyran-4-one as a strong antioxidant in glucose–histidine
Maillard reaction products. Food Research International, 51(1), 397–403.
Yu, X., Zhao, M., Liu, F., Zeng, S., & Hu, J. (2013b). Antioxidants in volatile Maillard
reaction products: Identification and interaction. LWT - Food Science and Technology,
53(1), 22–28.
Sun, J., Sun, B. G., Ren, F. Z., Chen, H. T., Zhang, N., & Zhang, Y. Y. (2019). Influence of
different frying processes on the flavor characteristics and sensory profile of garlic
oil. Molecules, 24, 4456–4470.
Tatum, J. H., Shaw, P. E., & Berry, R. E. (1967). Some compounds formed during
nonenzymic browning of orange powder. Journal of Agricultural and Food Chemistry,
15(5), 773–775.
Zhou, Z. L., Xu, Z. Q., Shu, J. S., She, S. K., Sun, W. F., Yin, C. Y., … Zhong, F. (2011).
Influence of various factors on formation of 2,3-dihydro-3,5-dihydroxy-6-methyl-4
(H)-pyran-4-one (DDMP) in a solid-state model system of Maillard reaction. European
Food Research and Technology, 239, 31–40.
6