Design, synthesis, and taste evaluation of a high-intensity umami-imparting oxazole-based compound

Article abstract of DOI:10.1080/09168451.2017.1345613

Umami taste is imparted predominantly by monosodium glutamate (MSG) and 5′-ribonucleotides. Recently, several different classes of hydrophobic umami-imparting compounds, the structures of which are quite different from MSG, have been reported. To obtain a novel umami-imparting compound, N-cinnamoyl phenethylamine was chosen as the lead compound, and a rational structure-optimization study was conducted on the basis of the pharmacophore model of previously reported compounds. The extremely potent umami-imparting compound 2-[[[2-[(1E)-2-(1,3-benzodioxol-5-yl)ethenyl]-4-oxazolyle]methoxy]methyl]pyridine, which exhibits 27,000 times the umami taste of MSG, was found. Its terminal pyridine residue and linear structure are suggested to be responsible for its strong activity. The time taken to reach maximum taste intensity exhibited by it, as determined by the time-intensity method, is 22.0 s, whereas the maximum taste intensity of MSG occurs immediately. This distinct difference in the time-course taste profile may be due to the hydrophobicity and strong receptor affinity of the new compound.

Full text of DOI:10.1080/09168451.2017.1345613

Bioscience, Biotechnology, and Biochemistry
ISSN: 0916-8451 (Print) 1347-6947 (Online) Journal homepage: http://www.tandfonline.com/loi/tbbb20
Design, synthesis, and taste evaluation of a
high-intensity umami-imparting oxazole-based
compound
Yusuke Amino, Yu-ki Tahara, Kei Yamada, Masakazu Nakazawa, Uno Tagami,
Takaho Tajima & Motonaka Kuroda
To cite this article: Yusuke Amino, Yu-ki Tahara, Kei Yamada, Masakazu Nakazawa, Uno Tagami,
Takaho Tajima & Motonaka Kuroda (2017): Design, synthesis, and taste evaluation of a high-
intensity umami-imparting oxazole-based compound, Bioscience, Biotechnology, and Biochemistry,
DOI: 10.1080/09168451.2017.1345613
To link to this article: http://dx.doi.org/10.1080/09168451.2017.1345613
Published online: 10 Jul 2017.
Submit your article to this journal
Article views: 10
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tbbb20
Download by: [Purdue University Libraries]
Date: 13 July 2017, At: 23:17

Bioscience, Biotechnology, and Biochemistry, 2017
Design, synthesis, and taste evaluation of a high-intensity umami-imparting
oxazole-based compound
Yusuke Amino1,*, Yu-ki Tahara1, Kei Yamada1, Masakazu Nakazawa2, Uno Tagami1,
Takaho Tajima3 and Motonaka Kuroda3
2
1Institute for Innovation, Ajinomoto Co., Inc., Kawasaki, Japan; Research Institute for Bioscience Products & Fine
3
Chemicals, Ajinomoto Co., Inc., Kawasaki, Japan; Institute of Food Research and Technologies, Ajinomoto Co.,
Inc., Kawasaki, Japan
Received April 19, 2017; accepted June 13, 2017
https://doi.org/10.1080/09168451.2017.1345613
Umami taste is imparted predominantly by mono-
sodium glutamate (MSG) and 5′-ribonucleotides.
Recently, several different classes of hydrophobic
umami-imparting compounds, the structures of
which are quite different from MSG, have been
reported. To obtain a novel umami-imparting com-
pound, N-cinnamoyl phenethylamine was chosen as
Kodama et al. and Kuninaka,6,7) and 5′-ribonucleotides
were shown to have strong synergistic effects with MSG
by Kuninaka and Yamaguchi.7,8) MSG fully activates the
human T1R1/T1R3 receptor and shows synergism with
IMP. Therefore, with reference to human psychophysical
data7,8) T1R1/T1R3 was established as the umami
receptor.9,10)
the lead compound, and
a
rational structure-
In recent years, several different classes of hydrophi-
lic and hydrophobic umami-tasting compounds, the
structures of which are quite different from MSG and
5′-ribonucleotides, have been reported.11−23) These
compounds have been termed “umami (taste) com-
pounds” or “umami (enhancing) molecules” elsewhere
in the literature. In this paper, such compounds are ter-
med “umami-imparting compounds,” and their taste is
more accurately termed “umami-imparting activity.”
Some of these compounds were discovered by isolation
from food ingredients or by classical studies on struc-
ture–taste relationship in industry and academia, and
others were discovered by high-throughput screening
using the human T1R1/T1R3 umami receptor.11−17,24)
Several representative synthetic high-intensity,
umami-imparting compounds that are registered for use
in food and included in the list of compounds that are
generally recognized as safe by the Flavor and Extract
Manufactures Association (FEMA GRAS) are shown
in Fig. 1. For example, Servant et al. reported that
compound 3 at a concentration of only 3 μM has an
umami-imparting activity equivalent to that of 12 mM
MSG.25)
The naturally occurring umami-imparting N-cin-
namoyl phenethylamine compounds rubemamine (9)
and rubescenamine (10) (Fig. 2) were identified and
reported by Backes et al. and Frerot et al. and are also
listed in the FEMA GRAS list.26,27) They are similar in
hydrophobicity and molecular size to the synthetic
high-intensity, umami-imparting molecules depicted in
Fig. 1, but their basic N-cinnamoyl phenethylamine
structures are different from those of the other synthetic
umami-imparting compounds. Preliminary studies on
the structure–taste relationship in 9 and 10 have been
optimization study was conducted on the basis of
the pharmacophore model of previously reported
compounds. The extremely potent umami-imparting
compound 2-[[[2-[(1E)-2-(1,3-benzodioxol-5-yl)ethenyl]-
4-oxazolyle]methoxy]methyl]pyridine, which exhibits
27,000 times the umami taste of MSG, was found.
Its terminal pyridine residue and linear structure
are suggested to be responsible for its strong
activity. The time taken to reach maximum taste
intensity exhibited by it, as determined by the time-
intensity method, is 22.0 s, whereas the maximum
taste intensity of MSG occurs immediately. This dis-
tinct difference in the time-course taste profile may
be due to the hydrophobicity and strong receptor
affinity of the new compound.
Key words: umami; N-cinnamoyl amine; structure-
optimization; taste evaluation; time-
intensity method
Umami taste is imparted by a number of substances,
predominantly monosodium glutamate (MSG) and 5′-ri-
bonucleotides such as 5′-inosinate (IMP) and 5′-guany-
late (GMP).1−4) In 1908, Ikeda discovered that the broth
of Laminaria japonica seaweed contains glutamate and
that glutamate is responsible for the umami taste sensa-
tion.5) For many years, only four basic tastes were recog-
nized, i.e. sour, sweet, salty, and bitter. However, umami
taste became generally accepted as the fifth basic taste
approximately 100 years after its discovery.1−4) IMP
from dried bonito fish and GMP from dried shiitake
mushroom were also found to impart umami taste by
*Corresponding author. Email: yusuke_amino@ajinomoto.com
© 2017 Japan Society for Bioscience, Biotechnology, and Agrochemistry

2
Y. Amino et al.
Fig. 1. Chemical structures of representative umami-imparting compounds registered for use in food.
R1
H
N
R2
O
O
O
R3
R4
N
N
O
O
9 : rubenamine; FEMA4310;
15
11
R1, R2, R3, R4 = OCH3
10
: rubescenamine; FEMA 4773;
R1, R2 = OCH3, R3, R4 = OCH2O
O
O
N
N
O
CH3O
O
H
N
OCH3
O
12
CH3O
H
N
N
O
CH3O
O
H
N
N
14
O
CH3
O
13
Fig. 2.
Structural optimization for rubemamine (9) and rubescenamine (10), and discovery of the high-intensity, umami-imparting oxazole
derivative (11).
performed by their discoverers, but compounds that
have much stronger umami-imparting activity than the
original compounds have not yet been found.26−29)
A primary goal in the field of taste chemistry is to
reveal the relationships between the taste intensity and
chemical structure of taste-active molecules.30) With
this in mind, we conducted studies on the structure–
umami-imparting activity relationships in N-cinnamoyl
amines to find high-intensity umami-imparting mole-
cules by referring to the structural features obtained
by pharmacophore modeling of the known umami-
imparting compounds shown in Fig. 1. As a result, the
novel oxazole-based compound 11 (Fig. 2), which has
high umami-imparting activity and excellent stability,
was discovered.
Another subject of great interest in the field of taste
chemistry is analysis of the taste characteristics of com-
pounds by sensory evaluation. Several publications have
reported the taste intensities or taste-recognition thresh-
old concentrations of hydrophilic umami-imparting
compounds.31,32) However, there are very few detailed
reports on the taste characteristics of hydrophobic high-
intensity umami-imparting compounds (other than
patents).26,27) The taste of hydrophobic high-intensity

Umami-imparting compound
3
umami-imparting compounds is expected to be quite
long-lasting and persistent, like those of high-intensity
sweeteners, as they are more hydrophobic than MSG.
Furthermore, owing to their structures being signifi-
cantly different from that of MSG, they are expected to
bind to a different site on the T1R1/T1R3 umami recep-
tor. In fact, the proposed binding site for MSG in the
umami receptor is located in the Venus flytrap domain
(VFTD) of the extracellular domain of the T1R1 sub-
unit, while that of compounds 2 (designated as S807 in
the literature) and 3 (designated as UE-1 or S336 in the
literature) is suggested to be the heptahelical transmem-
brane domain (TMD) of the T1R1 subunit.25,33)
To ascertain the difference in taste characteristics of
the high-intensity, umami-imparting compounds and
MSG, they were evaluated by comparing their time-
intensity profiles. The time to reach maximum taste
intensity (Tmax) of MSG is too small to determine
exactly, but Yamaguchi and Ninomiya estimated that it
is less than approximately 5 s.34) However, the Tmax of
our novel high-intensity, umami-imparting compounds
is 20 s or more.
Therefore, based on this prediction, we conducted a
structure-optimization study on compound 10 focused
on the amine moiety.
Structure–umami activity relationship study
The structural optimization was conducted by vary-
ing the substituents systematically while evaluating the
taste of the resultant compounds. Only the compounds
that elicited high taste intensity in the preliminary taste
test were evaluated further, and their structures are
shown in Fig. 2. Of all the newly synthesized com-
pounds evaluated, 11–15 show an intrinsic umami qual-
ity without evoking other taste sensations such as
bitterness or astringency. The taste intensity of a com-
pound in comparison with 0.3% MSG was determined
as the point of subjective equality (PSE), as described
in the experimental section. The results obtained are
summarized in Table 1. Note that the umami-imparting
activities of compounds 1–8 in Fig. 1 under identical
evaluation conditions were unavailable.
There have been some reports on the umami-impart-
ing activity of FEMA GRAS compounds. For example,
Backes et al. reported that rubemamine (9) and
rubescenamine (10) elicit a strong intrinsic umami taste
in water at 50 and 10 ppm, respectively.26) Frerot et al.
reported that 9 and several of its related compounds
have a clean umami taste at 20 ppm in water.27) In our
sensory evaluation, 13.4 mg/L (ppm) of 10 elicits the
same umami-imparting intensity as a 0.3% MSG solu-
tion, indicating that the taste intensity of 10 is approxi-
mately 220 times that of MSG (Table 1). Our data for
compound 10 is essentially consistent with those of
Backes et al. and Frerot et al.
In this paper, the rational structural optimization of
N-cinnamoyl phenethylamines and the taste evaluation
of synthesized compounds, which were performed to
explore intense umami-imparting compounds, are
described. Furthermore, the difference in the time-
course taste profiles of the high-intensity umami-
imparting molecules and MSG is investigated.
Results and discussion
Computational analysis
The ligand-based pharmacophore modeling approach
is often employed as a powerful and fast search tool
for optimizing lead structures. We built a pharma-
cophore model based on the structures of known high-
intensity, umami-imparting molecules and utilized the
features of the pharmacophore model to optimize the
structure of the N-cinnamoyl phenethylamines rube-
mamine (9) and rubescenamine (10).35,36)
The high-intensity, umami-imparting compounds (1–
10) (Figs. 1 and 2) have the diverse structures but also
have the common features, such as two nonpolar,
medium-sized substituents; substituted phenyl rings; a
pyridine ring; or branched alkyl groups. Furthermore,
they have a central linkage, such as an amide, a car-
bonyl group, or a heterocycle.
To generate pharmacophore models for the umami-
imparting compounds, a set of four structurally diverse
compounds (3, 6, 7, and 10 in Figs. 1 and 2) were
used. Further details of this process are given in the
experimental section. Three-dimensional (3D) pharma-
cophore models generated for 3, 6, 7, and 10 are
shown in Fig. 3. The pharmacophore model consists of
the following three common key features: one
hydrophobic group, one hydrogen-bond acceptor, and
one hydrogen-bond donor, which are shown in Fig. 3.
These features are likely to be crucial for the interac-
tion between the compound and the receptor. In the
right-hand area of the models, the aromatic rings are
arranged in different directions, indicating the
possibility of structural modification and optimization.
Compounds 11–15 present equivalent potency con-
centrations of 0.11 mg/L (27,000 times that of MSG),
1.3 mg/L (2,300 times), 0.53 mg/L (5,700 times),
0.12 mg/L (25,000 times), and 0.51 mg/L (5,900
times), respectively (Table 1).
Studies on the structure–taste relationship in analogs
of the N-cinnamoyl phenethylamines 9 and 10 were con-
ducted by Backes et al. and Frerot et al., respec-
tively.26,27) However, in their studies, no derivative was
found with an umami-imparting activity that improved
by more than one order of magnitude. Since the
structures of the amine moiety tested in their structure–
taste-relationship studies were limited to linear
2-arylethylamines or benzyl amines that exhibited a sim-
ilar structural motif to phenethylamines known to occur
in nature, most of the amines used had substituents on
the phenyl ring. However, unnatural structures with sub-
stituents on the methylene chain were not evaluated. It
seems that the amine moiety is more tolerant to struc-
tural modification than the N-cinnamoyl moiety.
Another structural modification of the amine moiety
in N-cinnamoyl compounds was described in a patent
filed by Tachdijan et al., in which a series of acry-
lamide derivatives with branched amines that are able
to activate the externally expressed T1R1/T1R3 human
umami receptor are described.24) The basic structure of
these acrylamide derivatives is the same as that of
N-cinnamoyl phenethylamine, except for their having
a branched amine moiety. The similarity of these

4
Y. Amino et al.
Fig. 3. Pharmacophore model mapping with compounds 3, 6, 7, and 10.
Note: Pharmacophoric determinants are the following: red: hydrogen-bond acceptor, blue: hydrogen-bond donor, and green: hydrophobic group.
Table 1. Summary of umami taste properties of compounds 10–15.
Compound
Equivalent potency concentration to 0.3% MSG (mg/L, ppm)
Relative potency to MSG (times)
Peak time (Tmax, s)
10
11
12
13
14
15
13.5
0.11
1.3
0.53
0.12
0.51
220
27,000
2300
5700
25,000
5900
24.0
22.0
29.0
28.2
29.7
30.0
Note: Taste intensity of MSG reaches the maximum value immediately after oral exposure.34).
compounds to compounds 9 and 10 was also noted in
a patent filed by Ley et al., who examined the
umami-imparting activity of derivatives of 9 and
10.28) Therefore, we hypothesized that N-cinnamoyl
phenethylamines could be modified into umami-
imparting compounds with higher activity by replacing
natural phenethylamines with unnatural amines. The
validity of this strategy was also supported by the
aforementioned results of the computational analysis.
A hydrophobic methoxymethyl group was introduced
to the 1-position of the phenethylamine moiety of a
N-cinnamoyl derivative to obtain compound 12, which
shows approximately 10 times the umami-imparting
activity of compound 10 (Fig. 2, Table 1). Compounds
12–14 have an asymmetric carbon center, and the (R)-
form molecules are more active than the (S)-form mole-
cules. As long as the substituent on the benzene ring of
the N-cinnamoyl moiety is hydrophobic (e.g. 4-meth-
oxy, 3,4-dimethoxy, and 3,4-methylenedioxy), the
umami-imparting activity is retained.
and N-[(2E)-3,7-dimethyl-2,6-octadien-1-yl]-N-[2-(2-
pyridyl)ethyl]ethanediamide.38) Therefore, the introduc-
tion of the 2-pyridine-containing fragment to an
appropriate position of compound 12 was expected to
improve its umami-imparting activity. This prediction
was also supported by the aforementioned results of the
computational analysis, in which the 2-pyridine rings of
compounds 3, 6, and 7 are not superimposed but are all
located in the right-hand side space of the pharma-
cophore model in Fig. 3. As a result, the umami-
imparting activity of compound 13, in which the
methoxymethyl group of compound 12 was replaced
with the 2-pyridylmethyloxymethyl group, is approxi-
mately twice that of 12.
Compound 14, in which the sterically hindered ben-
zyl group of compound 13 is replaced with a flexible
alkyl chain to reduce bulkiness, shows activity that is
improved to approximately five times that of the parent
compound. However, although the umami-imparting
activity of compound 14 is sufficiently high, its resid-
ual rate following stability testing in aqueous DMSO at
100 °C for 1 h was approximately 85%. Therefore, its
stability is unsatisfactory.
A pyridine substructure is a common and indicative
characteristic scaffold of some FEMA GRAS umami-im-
parting molecules, such as that in compounds 1, 3, 6, 7,
and 8. Moreover, there have been several patents pub-
lished on umami-imparting compounds that contain a 2-
pyridyl moiety, such as 2-(2-phenylthio)ethyl)pyridine37)
Oxazole peptidomimetics have been investigated
with the aim of promoting their stability and probing
their structure–activity relationships at the peptide

Umami-imparting compound
5
terminus.39) Compound 11, which has the structure of
compound 14 with an alkyl chain removed and the
amide bond linkage replaced by an oxazole isostere,
was synthesized, and its umami-imparting activity is
stronger than that of compound 14, i.e. approximately
27,000 times that of MSG. Therefore, compound 11 is
the strongest umami-imparting compound that we have
evaluated to date. Moreover, compound 11 is quite
stable in heated aqueous solution compared to com-
pound 14, with no decomposition being observed upon
heating at 100 °C for 3 h in aqueous DMSO.
Interestingly, the activity of compound 15 (Fig. 2), a
structure in which the vinyl group of compound 11 is
absent, is substantially decreased to one-fifth of that of
compound 11. These differences in an activity
indicate the significant role of the double bond in
compound 11.
Most of the analogs of compound 11 with different
substituents on the phenyl ring other than the 3,4-
methylenedioxy groups, such as 4-methoxy-, 4-iso-
propoxy-, 4-isobutyl-, and 4-tert-butylcinnamoyl
groups, have umami-imparting activities comparable to
that of compound 11. These results suggest that the
substitution pattern on the benzene ring does not signif-
icantly affect the umami-imparting activity. Of these
compounds, only 11 is crystalline. Therefore, 11 has
advantageous physical properties for isolation, purifica-
tion, and storage.
Finally, overlaying the structure of the newly discov-
ered intense umami-imparting compound 11 with the
pharmacophore model was examined. However, com-
pound 11 lacks a hydrogen-bond donor, one of the
pharmacophoric determinants, at the center of the mole-
cule and unexpectedly does not superimpose well with
the pharmacophore model constructed with compounds
3, 6, 7, and 10. The oxazole ring in the center of com-
pound 11 may function only as a linker of the terminal
aromatic rings to allow their orientation to adopt the
most suitable arrangement. However, the two suitably
located aromatic ring segments, i.e. the substituted phe-
nyl ring and 2-pyridyl ring at the ends of the com-
pound 11, promote the strong umami-imparting activity
of this compound. In addition, the linearity and length
of the molecule are likely to play critical roles in
umami-imparting activity. It is suggested that com-
pound 11 may have a different binding mode to the
T1R1/T1R3 receptor from that of the umami-imparting
compounds 3, 6, 7, and 10.
were subjected to a condensation/dehydration in xylene
at 85 °C to yield the chloromethyloxazole compound
19 in modest yield. Finally, the oxazole derivative 11
was obtained in 80% isolated yield as a white
solid by typical alkylation of compound 19 with
2-pyridylmethanol using sodium hydride as a base in
DMF.
Continuous umami-imparting activity analysis of
MSG and synthetic molecules
Although the taste quality of compound 11 is very
similar to that of MSG, the time-course profiles (the
change in taste intensity with time) of 11 and MSG are
significantly different, probably due to their differences
in chemical structure, hydrophobic/hydrophilic charac-
teristics, and binding site on the T1R1/T1R3 receptor.
To characterize the time-course profiles of the newly
identified umami-imparting compounds in comparison
with that of MSG at equivalent intensity (0.3% MSG),
time–taste response curves were established according
to the time-intensity method described in the experi-
mental section. The results obtained are summarized in
Table 1.
Umami sensation for compound 11 starts immedi-
ately upon oral administration, but increases gradually
and reaches maximum intensity at 22.0 s after exposure
and then decreases over a 70 s period. Thus, the rate of
taste increase up to Tmax for compound 11 is quite
slow, whereas, as previously discussed, that of MSG is
extremely fast.34) The times required to reach Tmax for
the other umami-imparting molecules, i.e. 10
(Tmax = 24.0 s), 12 (29.0 s), 13 (28.2 s), 14 (29.7 s),
and 15 (30.0 s), are quite long, regardless of their con-
centrations. Such a difference of time-intensity profiles
between MSG and our synthetic compounds corre-
sponds with that between sucrose and high-intensity
sweeteners.44−47)
The time-intensity profile of a compound may be
related to its taste intensity, i.e. its affinity to a receptor,
or chemical structure. It is interesting that the taste
intensities of compounds 11 and 14 are almost equal,
whereas the Tmax value for 11 (22.0 s) is much shorter
than that of compound 14 (29.7 s). Thus, the time-
intensity profiles of these compounds may not be asso-
ciated with their taste intensities but may instead be
affected by their structural features, such as whether
they are linear or branched and whether they contain
an oxazole linkage or an amide linkage. Furthermore,
Tmax of compound 11 is shorter than that of its short
chain homolog 15 (Tmax = 30.0 s), which suggests that
the molecular size and taste profile are independent; i.e.
the smaller molecule does not necessarily exhibit a
Synthesis of compound 11
The synthesis of 2-phenyl-4-chloromethyloxazole is
reported in several papers,40,41) whereas the synthesis
of 2-phenylvinyl-4-chloromethyloxazole, a key interme-
diate for compound 11, appears only in patents by
Tasaka et al. and Bossenmaier et al.41,42) Compound 11
was synthesized through the procedure shown in
Fig. 4. Briefly, by activating 3-(1,3-benzodioxol-5-yl)-
(2E)-propionic acid (16) with oxalyl chloride in toluene
at 100 °C, acid chloride 17 was obtained quantitatively.
Compound 17 was converted into the corresponding
amide 18 by adding 17 to a mixed solvent of 28%
ammonia and N,N-dimethylformamide (DMF) in quan-
titative yield. Amide 18 and 1,3-dichloro-2-propanone
shorter Tmax
.
It is well known that the time-intensity profiles of
MSG and NaCl are similar and that using MSG to
increase the umami taste helps to reduce NaCl levels.48)
Dewis et al. reported that umami-imparting compound
4 (Fig. 1) at 5 and 10 μM significantly increased the
taste quality ratings of pure 80 mM NaCl solutions rel-
ative to a control in human psychophysical tests.16)
However, the time-course profiles of the high-intensity,
umami-imparting compounds, such as compound 11,
are significantly different from that of MSG because of

6
Y. Amino et al.
O
O
O
O
O
O
a
b
Cl
OH
NH2
O
O
O
16
17
18
O
O
O
c
d
N
N
N
O
O
Cl
O
O
19
11
Fig. 4. Synthesis of oxazole derivative (11).
Notes: Conditions: (a) SOCl2/toluene, reflux; (b) aqueous NH3, DMF; (c) 1,3-dichloro-2-propanone/xylene, 85 °C; (d) 2-pyridinemethanol, NaH/
DMF, room temperature.
their low rate of taste increase and large Tmax, thus
making them less likely to be adopted as alternatives
Experimental
General. 1H NMR and 13C NMR spectra were
for MSG for NaCl reduction. Indeed, in our prelimi-
nary taste evaluation, it was found to be difficult to
replace, even partially, MSG with compound 11 to
reduce the usage of NaCl (data not shown).
Positive modulatory (synergistic) effects for 9 and 10
on the umami taste of MSG have been reported.26)
Further taste evaluation is required to investigate the
modulatory effects of our novel high-intensity, umami-
imparting compounds with MSG or 5′-ribonucleotides
and thus further understand their taste mechanism and
mode of action.
recorded using a Bruker AVANCE400 (400 MHz)
nuclear magnetic resonance apparatus and electron
spray ionization (ESI)-MS spectra in positive and nega-
tive ion modes were measured using a Thermo Quest
TSQ700 mass spectrometer.
Synthesis of 2-[[[2-[(1E)-2-(1,3-benzodioxol-5-yl)ethe-
nyl]-4-oxazolyle]methoxy]methyl]pyridine (11). Oxalyl
chloride (4.0 mL, 52.0 mmol) was added dropwise to a
suspension of 3-(1,3-benzodioxol-5-yl)-(2E)-propenoic
acid (16, 1.0 g, 5.20 mmol) in toluene (10 mL) at 0 °C.
The solution was then stirred at 100 °C for 2 h. The result-
ing solution was cooled to room temperature (r.t.) and then
concentrated under reduced pressure to give crude
3-(1,3-benzodioxol-5-yl)-(2E)-propenoyl chloride (17)
quantitatively.
Conclusion
As the result of a rational structure-optimization
study of the natural umami-imparting compounds rube-
mamine (9) and rubescenamine (10) based on a phar-
macophore modeling study of previously reported
synthetic umami-imparting molecules, the extremely
potent oxazole-based compound 11 was obtained
(Fig. 2). Compound 11 exhibited an umami taste that is
27,000 times more intense than that of 0.3% MSG in a
taste evaluation study (Table1). Compound 11 did not
fit the pharmacophore model that was constructed using
compounds 3, 6, 7, and 10 (Fig. 3), because 11 lacks a
hydrogen-bond donor at the center of the molecule.
However, the structural features of compound 11, i.e.
the aromatic ring segments at each end of the molecule
(substituted phenyl ring and 2-pyridine) and the oxa-
zole linkage at the center of the molecule, may be
responsible for its strong umami-imparting activity.
Furthermore, the stability of compound 11 was such
that no degradation was observed after heating it at
100 °C for 3 h in aqueous DMSO. Another important
finding from our work is that compound 11 exhibited
remarkable features in its time-course taste profile as
compared to MSG. Tmax for compound 11 is 22.0 s
(Table 1), whereas that of MSG is reported
approximately less than 5 s.34) Further research is
required to thoroughly elucidate the taste characteristics
of compound 11. However, the findings presented here
will assist in the continued investigation of the
structural requirements of novel taste-imparting
compounds.49−52)
Aqueous ammonia (28%, 40 mL) was added to the
crude 17 in DMF (10 mL) at 0 °C. The solution was
then stirred for 16 h at 40 °C. The resulting solution
was extracted with AcOEt (30 mL, twice). The com-
bined organic layers were washed with brine, dried
over anhydrous MgSO4, and concentrated under
reduced pressure to give crude 3-(1,3-benzodioxol-5-
yl)-(2E)-propenamide (18) quantitatively. ESI-MS m/z:
192.2 (M + H)+.
The crude 18 (approximately 5.20 mmol) and 1.3-
dichloro-2-propanone (762 mg, 6.00 mmol) were dis-
solved in xylene (10 mL), and the solution was heated
for 16 h at 125 °C. The resulting solution was cooled
to r.t. and then concentrated under reduced pressure.
The residue was purified by column chromatography
(silica gel 50 g, n-hexane/AcOEt) to give 2-[(1E)-2-
(1,3-benzodioxol-5-yl)ethenyl]-4-(chloromethyl)oxazole
(19, 913 mg, 3.47 mmol) in 57.8% yield. ESI-MS m/z:
264.2 (M + H)+.
Sodium hydride (60% dispersion in paraffin liquid,
272 mg, 6.80 mmol) was added to a solution of 19
(913 mg, 3.47 mmol) and 2-pyridinemethanol (340 mg,
3.12 mmol) in DMF (5 mL) at r.t. The solution was stir-
red for 16 h at r.t., quenched with cold water (15 mL),
and 30-mL extraction with AcOEt was done twice. The
combined organic layers were washed with brine, dried
over anhydrous MgSO4, and concentrated under
reduced pressure. The residue was purified by column
chromatography (silica gel 50 g, n-hexane/AcOEt) to

Umami-imparting compound
7
give compound 11 (933 mg, 2.77 mmol) as a white
solid in 80.1% yield.
MS(ESI) m/z: 311.2 (M + H)+. FAB-MS m/z:
311.1032 (M + H) (Calcd. for C17H14N2O4: 311.1032).
mp: 81–82 °C.
NMR (DMSO-d6) δH: 4.52 (2H, s), 4.64 (2H, s),
6.07 (2H, s), 6.94–7.04 (2H, m), 7.16 (1H, dd), 7.29–
32 (1H, m), 7.42–7.48 (3H, m), 7.79–7.83 (1H, m),
8.10 (1H, s), 8.51–8.53 (1H, m). NMR (CDCl3) δC:
64.8, 73.0, 101.4, 105.9, 108.6, 111.9, 121.9, 122.7,
123.1, 129.9, 135.9, 136.2, 137.3, 138.8, 148.4, 148.5,
148.7, 157.8, 162.1.
Computational analysis.For the ligand-based pharma-
cophore modeling, a set of four umami-imparting com-
pounds (3, 6, 7, and 10) from the FEMA GRAS list
were collected from recently published literature.18−23)
The two-dimensional (2D) chemical structures of the
compounds were sketched and saved in Molecular
Design Limited mol file format. They were imported
into Maestro 11 (Schrödinger, LLC), converted into the
corresponding standard 3D structures, and protonated
by Ligprep using OPLS3 as force field. These struc-
tures were utilized as starting conformations for confor-
mation generation.35)
ESI-MS m/z: 337.1 (M + H)+. FAB-MS m/z:
337.1204 (M + H) (Calcd. for C19H16N2O4: 337.1188).
mp: 96–98 °C.
(E)-N-((R)-1-Methoxy-3-phenylpropan-2-yl)-3-(4-
methoxyphenyl)acrylamide (12). NMR (DMSO-d6) δH:
2.64–2.91 (2H, m), 3.25–3.42 (5H, m), 3.78 (3H, s),
4.12–4.23 (1H, m), 6.49 (1H, d, J = 15.8 Hz), 6.91–
7.02 (2H, m), 7.14–7.38 (6H, m), 7.44–7.56 (2H, m),
8.01 (1H, d, J = 8.4 Hz). NMR (DMSO-d6) δC: 36.9,
50.0, 55.2, 58.3, 73.4, 114.4, 114.4, 119.7, 126.1,
127.4, 128.2, 128.2, 129.1, 129.1, 129.1, 129.1, 138.3,
138.8, 160.3, 164.8.
Next, Phase (Schrödinger, LLC)36) was used to gen-
erate a pharmacophore model for the prediction of
common features such as H-bond acceptor (A), hydro-
gen-bond donor (D), hydrophobic group (H), negatively
charged group (N), positively charged group (P), and
aromatic ring (R). Common pharmacophore features of
compounds 3, 6, 7, and 10 were identified. 3D spatial
relationship and geometric parameters of the generated
pharmacophore model are shown in Fig. 3. The phar-
macophore model consisted of three features: one
hydrophobic group, one hydrogen-bond acceptor, and
one hydrogen-bond donor, as shown in Fig. 3.
Sensory evaluation. Seven trained male panelists
aged 27–52 years participated in the sensory evaluation.
All panelists were employees of Ajinomoto Co., Inc.
and provided written informed consent. The protocol
for sensory evaluation was approved by the ethics
board of the Institute of Food Sciences and Technolo-
gies, Ajinomoto Co., Inc. Sensory evaluation was
conducted in a partitioned booth at 25 °C in an
air-conditioned sensory evaluation room.
Determination of the time to reach maximum inten-
sity (Tmax). The time to reach the maximum intensity
(Tmax) value of each compound was determined by a
time-intensity method as described by Lawless &
Heymann53) and described in ASTM E-1909-11.54) For
the determination of Tmax, compounds 10–15 were
evaluated at concentrations of 10.0, 0.125, 1.0, 0.53,
0.11, and 1.0 mg/L, respectively. Panelists were
instructed to swirl 10 mL of each sample for 70 s and
evaluate the taste intensity. In this evaluation, taste
intensity was evaluated on a scale from 0 to 100
points, and 50 points were defined as equivalent to the
intensity of 0.3% (w/w) MSG at Tmax. The results
obtained are summarized in Table 1.
Determination of taste intensity of each compound at
Tmax. The taste intensity of each compound was deter-
mined by measuring the PSE though the 2-AFC method-
ology. The PSE was measured using a method described
by Furukawa et al.55) with a slight modification. In this
evaluation, the panelists were instructed to evaluate the
taste intensity at Tmax of each sample and compare it to
the taste intensity of 0.3% MSG at Tmax through the 2-
AFC methodology.56) The data were analyzed by a pro-
bit analysis method,55) and the equivalent concentration
of each compound to 0.3% MSG was calculated.
MS(ESI) m/z: 326.1 (M + H)+. FAB-MS m/z:
326.1764 (M) (Calcd. for C20H23NO4: 326.1756). mp:
115–118 °C.
(E)-N-((R)-1-((Pyridin-2-yl)methoxy)-3-phenylpropan-
2-yl)-3-(4-methoxyphenyl)acrylamide
(13).
NMR
(DMSO-d6) δH: 2.71–2.99 (2H, m), 3.42–3.56 (2H, m),
3.70–3.82 (3H, m), 4.19–4.34 (1H, m), 4.49–4.67 (2H,
m), 6.52 (1H, d, J = 15.8 Hz), 6.91–7.00 (2H, m),
7.12–7.54 (10H, m), 7.73–7.85 (1H, m), 8.05–8.15
(1H, m), 8.49–8.57 (1H, m). NMR (DMSO-d6) δC:
36.9, 50.2, 55.2, 71.6, 73.3, 114.4, 114.4, 119.7, 121.3,
122.6, 126.1, 127.4, 128.2, 128.2, 129.1, 129.1, 129.1,
129.1, 136.7, 138.4, 138.7, 148.9, 158.1, 160.3, 164.9.
MS(ESI) m/z: 403.5 (M + H)+, 400.9 (M-H)−. FAB-
MS m/z: 403.1999 (M + H) (Calcd. for C25H25N2O3:
402.2022). mp: 77–79 °C.
(E)-N-((R)-1-(2-(Pyridin-2-yl)methoxy)pentane-3-(4-
methoxyphenyl)acrylamide (14). NMR (DMSO-d6) δH:
0.87 (3H, t, J = 4.2 Hz), 1.24–1.33 (4H, m), 1.42–1.45
(1H, m), 1.58–1.61 (1H, m), 3.42–3.53 (2H, m), 3.79
(3H, s), 4.01–4.07 (1H, m), 4.57 (1H, d, J = 3.0 Hz),
6.53 (1H, d, J = 15.7 Hz), 6.97 (2H, d, J = 5.0 Hz),
7.27–7.30 (1H, m), 7.36 (1H, d, J = 15.8 Hz), 7.44
(1H, d, J = 7.8 Hz), 7.50 (2H, d, J = 6.9 Hz), 7.79 (1H,
t, J = 7.7 Hz), 7.90 (1H, d, J = 8.6 Hz), 8.51 (1H, s).
NMR (DMSO-d6) δC: 14.0, 22.1, 27.7, 30.9, 48.3,
55.3, 72.5, 73.2, 114.4, 119.9, 121.2, 122.5, 127.5,
129.1, 129.1, 136.7, 138.3, 138.3, 148.8, 158.3, 160.3,
165.0.
MS(ESI) m/z: 369.3 (M + H)+, 367.1 (M-H)−. FAB-
MS m/z: 369.2193 (M + H) (Calcd. for C22H28N2O3:
369.2178). mp: 83–85 °C.
2-[[[2-(1,3-Benzodioxol-5-yl)-4-oxazolyl]methoxy]
methyl]pyridine (15)
NMR (DMSO-d6) δH: 4.55 (2H, s), 4.65 (2H, s),
6.13 (2H, s), 7.07 (1H, d, J = 8.1 Hz), 7.24–7.59 (4H,
m), 7.77–7.85 (1H, m), 8.17 (1H, s), 8.46–8.56 (1H,
m). NMR (CDCl3) δC: 64.8, 73.1, 101.6, 106.8, 108.6,
121.3, 121.6, 121.8, 122.6, 136.0, 137.2, 138.7, 148.1,
148.6, 149.6, 157.9, 161.9.

8
Y. Amino et al.
[4] Kurihara K. Umami the fifth basic taste: history of studies on
receptor mechanisms and role as a food flavor. BioMed Res
Inter. 2015;ID189402.
[5] Ikeda K. New seasonings. Chem Senses 2002;27:847–849.
[Translation of J Chem Soc Jpn. 1909;30:820–836.]
[6] Kodama S. Separation methods of inosinic acid. J Chem Soc
Tokyo. 1913;34:751–757.
In addition, based on the equivalent concentration of
each compound, the relative potency to MSG was calcu-
lated. The results obtained are summarized in Table 1.
Abbreviations
[7] Kuninaka A. Research on taste function of the nucleotides. J
Agric Chem Soc Japan. 1960;34:489–492.
[8] Yamaguchi S. The synergistic taste effect of monosodium gluta-
mate and disodium 5’-inosinate. J Food Sci. 1967;32:473–478.
[9] Nelson G, Chandrashekar J, Hoon MA, et al. An amino-acid
taste receptor. Nature. 2002;416:199–202.
[10] Li X, Staszewski L, Xu H, et al. Human receptors for sweet and
umami taste. Proc Natl Acad Sci USA. 2002;99:4692–4696.
[11] Winkel C, de Klerk A, Visser J, et al. New developments in
umami (enhancing) molecules. Chem Biodivers. 2008;5:1195–
1203.
[12] Suess B, Brockhoff A, Meyerhof W, et al. Structures, orosensory
activity, and T1R1/T1R3 receptor activation of amides generated
by lactone aminolysis using food-related processing conditions.
Eur Food Res Technol. 2013;237:57–70.
[13] Backes M, Paetz S, Vössing T, et al. Synthesis and sensory
studies of umami-active scaffolds. Chem Biodivers.
2014;11:1782–1797.
[14] Rubin M, Firmenich SA. Taste modifying product. WO Patent
WO 2009/122319 A2. 2019 Oct 8.
[15] Dewis ML, Phan THT, Ren ZJ, et al. N-geranyl cyclopropyl-
carboxamide modulates salty and umami taste in humans and
animal models. J Neurophysiol. 2013;109:1078–1090.
[16] Miyazawa Y, Ohkubo Y, Haraguchi K, et al. T. Hasegawa Co.,
Ltd. Flavor modulator having pyridine derivative or salt thereof
as active ingredient. WO Patent WO 2016/167153 A1. 2016 Oct
20.
IMP
5′-inosinate
5′-guanylate
GMP
MSG
T1R1
T1R3
FEMA
monosodium glutamate
taste receptor type 1 member 1
taste receptor type 1 member 3
Flavor and Extract Manufactures
Association of the United State
generally recognized as safe
Venus flytrap domain
transmembrane domain
the time to reach maximum taste
intensity
point of subjective equality
ethyl acetate
N,N-dimethylformamide
dimethyl sulfoxide
American Society for Testing and
Materials
alternative forced choice
GRAS
VFTD
TMD
Tmax
PSE
AcOEt
DMF
DMSO
ASTM
AFC
Author contributions
All the authors designed the research plan after con-
ducting an appropriate discussion; Y. T., K. Y., M. N.,
and Y. A. performed the design and synthesis of new
compounds. T. T. and M. K. performed the sensory
analysis, and U. T. contributed to the computational
analysis. Y. A. wrote the manuscript with the assistance
of M. K. and U. T., and Y. A. supervised the research.
[17] Behrens M, Meyerhof W, Hellfritsch C, et al. Sweet and umami
taste: natural products, their chemosensory targets, and beyond.
Angew Chem Int Ed. 2011;50:2220–2242.
[18] Smith RL, Cohen SM, Doull J, et al. GRAS flavoring substances
22. Food Technol. 2005;59:24–62.
[19] Waddell WJ, Cohen SM, Feron VJ, et al. GRAS flavoring sub-
stances 23. Food Technol. 2007;61:22–49.
[20] Smith RL, Waddell WJ, Cohen SM, et al. GRAS flavoring
substances 24. Food Technol. 2009;63:46–105.
[21] Smith RL, Waddell WJ, Cohen SM, et al. GRAS flavoring sub-
stances 25. Food Technol. 2011;65:44–75.
[22] Marnett LJ, Cohen SM, Fukushima S, et al. GRAS flavoring
substances 26. Food Technol. 2013;67:38–56.
[23] Cohen SM, Fukushima S, Gooderham NJ, et al. GRAS flavoring
substances 27. Food Technol. 2015;69:40–59.
[24] Tachdjian C, Patron AP, Adamski-Werner SL, et al. Senomyx,
Inc. Flavors, flavor modifiers, tastants, taste enhancers, umami or
sweet tastants, and/or enhancers and use thereof. United States
Patent US 8,895,050 B2. 2014 Nov 25.
[25] Servant G, Tachdjian C, Li X. Small molecule modulators of
family-C taste GPCRs. Igaku no Ayumi. 2010;233:881–891.
[26] Backes M, Obst K, Bojahr J, et al. Rubemamine and Rubesce-
namine, two naturally occuring N-cinnamyl phenethylamines
with umami-taste-modulating properties. J Agric Food Chem.
2015;63:8694–8704.
Acknowledgments
We would like to thank Toshihisa Kato, Yuzuru Eto,
Chiaki Nosaka, Tomohiko Kodera, and Toshimi Mizu-
koshi for their encouragement and continued support.
We are grateful to Yumiko Suzuki, Megumi Kaneko,
Kayo Matsumoto, and Yayoi Kawato for their helpful
discussions and assistance. We are also grateful to
members of the taste panel at the Institute of Food
Research and Technologies in Ajinomoto Co., Inc.
Disclosure statement
[27] Frerot E, Neirynck N, Cayeux I, et al. New umami amides:
structure?taste relationship studies of cinnamic acid derived
amides and the natural occurrence of an intense umami amide in
zanthoxylum piperitum. J Agric Food Chem. 2015;63:7161–
7168.
[28] Ley JP, Reichelt K, Paetz S, et al. Symrise A.G. Cinnamamides
as savory flavorings. United States Patent US 9,144,248 B2.
2015 Sep 29.
No potential conflict of interest is reported by the authors.
References
[1] Ninomiya K. Science of umami taste: adaptation to gastronomic
culture. Flavour. 2015;4:Article ID: 13.
[29] Billat-Rossi M, Aeberhardt K, Firmenich SA. Taste-modifying
combinations. United State Patent US 2016/0205979 A1. 2016
Jul 21.
[2] Uneyama H, Kawai M, Sekine-Hayakawa Y, et al. Contribution
of umami taste substances in human salivation during meal. J
Med Investig. 2009;56:197–204.
[30] Shallenberger RS. Taste Chemistry. Glasgow: Blackie Academic
& Professional, an imprint of Chapman & Hall; 1993.
[3] Beauchamp GK. Sensory and receptor responses to umami: an over-
view of pioneering work. Am J Clin Nutr. 2009;90:723S–727S.

Umami-imparting compound
9
[31] Festing D, Hofmann T. Discovery of N2-(1-carboxyethyl)guano-
sine 5′-monophosphate as an umami-enhancing maillard-
modified nucleotide in yeast extracts. J Agric Food Chem.
2010;58:10614–10622.
[32] Suess B, Brockhoff A, Degenhardt A, et al. Human taste and
umami receptor responses to chemosensorica generated by
maillard-type N2-alkyl- and N2-arylthiomethylation of guanosine
5′-monophosphates. J Agric Food Chem. 2014;62:11429–11440.
[33] Zhang F, Klebansky B, Fine RM, et al. Molecular mechanism
for the umami taste synergism. Proc Natl Acad Sci USA.
2008;105:20930–20934.
[34] Yamaguchi S, Ninomiya K. Umami and food palatability. J Nutr.
2000;130:921S–926S.
[35] Harder E, Damm W, Maple J, et al. OPLS3: a force field pro-
viding broad coverage of drug-like small molecules and proteins.
J Chem Theory Comput. 2015;12(1):281–296. DOI:10.1021/
acs.jctc.5b00864
[36] Dixon SL, Smondyrev AM, Rao SN. PHASE: a novel approach
to pharmacophore modeling and 3D database searching. Chem
Biol Drug Des. 2006;67:370–372.
[37] Brooks MP, Givaudan SA. Organic compounds. WO Patent WO
2015/000900 A2. 2015 Jan 8.
[38] Kaouas A, Winkel C, Givaudan SA. Oxalamide derivative as
umami flavouring agent. United States Patent US 8,470,384 B2.
2013 Jun 25.
[39] Wales SM, Hammer KA, Somphol K, et al. Synthesis and antimicro-
bial activity of binaphthyl-based, functionalized oxazole and thia-
zole peptidomimetics. Org Biomol Chem. 2015;13:10813–10824.
[40] Einsiedel J, Thomas C, Hübner H, et al. Phenyloxazoles and
phenylthiazoles as benzamide bioisosteres: synthesis and dopa-
mine receptor binding profiles. Bioorg Med Chem Lett.
2000;10:2041–2044.
[41] Ye XY, Chen S, Zhang H, et al. Synthesis and structure-activity
relationships of 2-aryl-4-oxazolylmethoxy benzylglycines and
2-aryl-4-thiazolylmethoxy benzylglycines as novel, potent
PPARα selective activators- PPARα and PPARγ selectivity
modulation. Bioorg Med Chem Lett. 2010;20:2933–2937.
[42] Tasaka A, Naito K. Takeda Pharmaceutical Company Limited.
Heterocyclic compounds, oxazole derivatives, process for prepa-
ration of the same and use thereof. United States Patent US
6,984,653 B2. 2006 Jan 10.
[43] Bossenmaier B, Friebe WG, Friess T, et al. Hoffman-La Roche
Inc. Oxazole derivatives. United States Patent US 7,205,325 B2.
2007 Apr 17.
[44] Pfeiffer JF, Boulton RB, Noble AC. Modeling the sweetness
response using time-intensity data. Food Qual Prefer.
2000;11:129–138.
[45] Janusz JM, Young PA, Hiler GD, et al. Time-intensity profiles of
dipeptide sweeteners. In: Walters DE, Orthoefer FT, DuBois GE,
editors. Sweeteners: Discovery, molecular design, and chemore-
ception. Washington (DC): ACS Symposium Series 450,
American Chemical Society; 1991. p. 277–289.
[46] DuBois GE. Sweeteners and sweetness modulators: Require-
ments for commercial viability. In: Weerasinghe DK, DuBois
GE, editors. Sweetness and sweeteners: Biology, chemistry, and
psychophysics. Washington (DC): ACS Symposium Series 979,
American Chemical Society; 2008. p. 444–462.
[47] Prakash I, Bishay IE. Development of new, no calorie commer-
cial sweetener neotame. In: Weerasinghe DK, DuBois GE, edi-
tors. Sweetness and sweeteners: Biology, chemistry, and
psychophysics. Washington (DC): ACS Symposium Series 979,
American Chemical Society; 2008. p. 492–510.
[48] Yamaguchi S, Takahashi C. Interaction of monosodium gluta-
mate and sodium chloride on saltiness and palatability of clear
soup. J Food Sci. 1984;49:82–85.
[49] Amino Y, Mori K, Tomiyama Y, et al. Development of new,
low calorie sweetener: new aspartame derivative. In: Weeras-
inghe DK, DuBois GE, Editors. Sweetness and sweeteners: biol-
ogy, chemistry, and psychophysics. Washington (DC): ACS
Symposium Series 979, American Chemical Society; 2008,
p. 463–480.
[50] Ohsu T, Amino Y, Nagasaki H, et al. Involvement of the cal-
cium-sensing receptor in human taste perception. J Biol Chem.
2010;285:1016–1022.
[51] Amino Y, Nakazawa M, Kaneko M, et al. Structure-CaSR-
activity relation of Kokumi γ-glutamyl peptides. Chem Pharm
Bull. 2016;64:1181–1189.
[52] Amino Y, Kawahara S, Mori K, et al. Preparation and character-
ization of four stereoisomers of monatin. Chem Pharm Bull.
2016;64:1161–1171.
[53] Lawless HT, Heymann H. Time-intensity methods. New York
(NY): Springer Science+Business Media, LLC; 2010.Sensory
Evaluation of Food; p. 179–201.
[54] ASTM E-1909-11, Standard guide for time-intensity evaluation
of sensory attributes. West Conshohocken (PA): ASTM Interna-
tional; 2012.
[55] Furukawa H, Saso H, Maeda S, et al. The taste test of organic
acids (1). Measurement of point of subjective equalities (P.S.E.)
on sourness of nine organic acids permitted as food additives.
Nippon Shokuhin Kogyo Gakkaishi. 1969;16:63–68.
[56] ASTM E-2164-08, Standard test method for directional
difference test. West Conshohocken (PA): ASTM International;
2012.