Photoactive ligands probing the sweet taste receptor. Design and synthesis of highly potent diazirinyl d-phenylalanine derivatives

Article abstract of DOI:10.1016/j.bmcl.2009.12.029

Some d-Amino acids such as d-tryptophan and d-phenylalanine are well known as naturally-occurring sweeteners. Photoreactive d-phenylalanine derivatives containing trifluoromethyldiazirinyl moiety at 3- or 4-position of phenylalanine, were designed as swee

Full text of DOI:10.1016/j.bmcl.2009.12.029

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1081–1083
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Photoactive ligands probing the sweet taste receptor. Design and synthesis
of highly potent diazirinyl D-phenylalanine derivatives
Katsuyoshi Masuda a, Ayako Koizumi b, Takumi Misaka b, Yasumaru Hatanaka c, Keiko Abe b,
Takaharu Tanaka a, Masaji Ishiguro a, Makoto Hashimoto d,
*
a Suntory Institute for Bioorganic Research, 1-1-1 Wakayamadai, Shimamoto-cho, Mishima-gun, Osaka 618-8503, Japan
b Department of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
c Faculty of Pharmaceutical Sciences, Toyama University, 2360 Sugitani, Toyama 930-0194, Japan
d Division of Applied Science, Graduate School of Agriculture, Hokkaido University, Kita 9, Nishi 9, Kita-ku, Sapporo 060-8589, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Some
sweeteners. Photoreactive
3- or 4-position of phenylalanine, were designed as sweeteners for functional analysis with photoaffinity
labeling. The trifluoromethyldiazirinyl -phenylalanine derivatives were prepared effectively with
D
-Amino acids such as
D-tryptophan and D-phenylalanine are well known as naturally-occurring
Received 24 August 2009
Revised 13 November 2009
Accepted 7 December 2009
Available online 11 December 2009
D-phenylalanine derivatives containing trifluoromethyldiazirinyl moiety at
D
chemo-enzymatic methods using
the human sweet taste receptor.
L-amino acid oxidase and were found to have potent activity toward
Keywords:
Sweet taste
Ó 2009 Elsevier Ltd. All rights reserved.
D-Amino acid
Photoaffinity label
Diazirine
Sweet taste receptor is a member of class C G protein-coupled
receptors (GPCRs) and forms heterodimeric structure with T1R2
and T1R3 subunits. Each subunit has a large amino-terminal do-
main (ATD) linked by a cysteine-rich domain (CRD) at the extracel-
lular site to a seven transmembrane helical domain (TMD).1
Human heterodimeric sweet taste receptor (hT1R2-hT1R3) re-
sponds to a wide variety of chemical substances including natu-
homology modeling and complex structures with some sweetners
have been built through ligand-docking in the putative ligand-
binding site,10 although complex models of
D-amino acids are not
available. Combining the mutational results on T1R2 and T1R3,
complex structure models would give a reasonable binding mode
of ligands. However, they should be confirmed by appropriate
experiments.
rally-occurring sugars, glycosides,
D
-amino acids, and artificial
Photoaffinity labeling is one of the methods used in the study of
the interactions of low molecular bioactive compounds with bio-
molecules.11–15 It is suitable for the analysis of biological interac-
tions because it is based on the affinity of the bioactive compound
for biomolecules. Selection of photophore for photoaffinity labeling
is very critical to obtain better results, but there are no universal
photophore selections.12 Various photophores, such as benzophe-
none, arylazide, and 3-(trifluoromethyl)phenyldiazirine, are used
for the elucidation of the ligand–receptor or substrate–enzyme
interactions. In this study, we focused on synthesizing photoreactive
chemical compounds such as sucralose, aspartame, and saccharin.2
Although these sweeteners have various chemical structures, all of
the compounds bind to the same sweet taste receptor.3,4 Since the
receptor distinguishes
D
- and
L
-amino acids, preferring
D
-amino
-amino acids and
-amino acid deriv-
acids,5–7 the structural relationships between
D
other sweeteners and the structural features of
D
atives that favor the activation of the sweet taste receptor have
been studied with conformational analysis by crystallography,
NMR analysis and molecular modeling of these ligands.8 However,
it is still obscure to understand the structural relations among the
sweeteners, since neither the receptor structure nor the structures
of the ligands complexed with the receptor is not available so far.
Since T1R2 and T1R3 are orthologs of metabotropic glutamate
receptor1 (mGluR1),9 the crystal structure of which is known,
structural models of T1R2 and T1R3 have been constructed with
D-phenylalanine derivatives, having trifluoromethyldiazirine at 3-
or 4- position of phenylalanine benzene ring, based on our modeling
work of the ligand–receptor complex, showing that the ligand-bind-
ing site has a space for binding the trifluoromethyldiazirine moiety
on the phenyl ring of D-phenylalanine. Since preparing the carbene
precursor (3-trifluoromethyl)phenyldiazirinyl three-membered
ring requires somewhat complicated process, fewer applications
have been reported in biomolecular studies than other photo-
* Corresponding author. Tel./fax: +81 11 7063849.
E-mail address: hasimoto@abs.agr.hokudai.ac.jp (M. Hashimoto).
phores13–15 and neither does the synthesis of diazirinyl
D-amino acid
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.12.029

1082
K. Masuda et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1081–1083
derivatives. A few cases of asymmetric synthesis of (3-trifluoro-
methyl)-diazirinyl
-phenylalanine derivatives are reported.16–20
-selective asymmetric synthesis has not been reported.
-amino acid derivatives are prepared by enzymatic reso-
lution of N-acyl derivatives using acylase, the residual N-acyl D-ami-
no acid compounds were recovered with less effectiveness.
N
N
N
N
F3C
F3C
L
i)
N
C(C6H5)2
+
However,
Although
D
COOt-Bu
L
Br
3
(C6H5)2C
N
COOt-Bu
1 4-
2 3-
4 4-
5 3-
L
-Amino acid oxidase is an enzyme for converting
-keto acid but does not work on the -amino acid compounds at
all.21 Thus, the enzymatic resolution of
-amino acid from racemic
mixture using the oxidase is a useful method and thus we tried to re-
solve the trifluoromethyldiazirinyl -phenylalanine derivatives
from racemic mixture using -amino acid oxidase. The -phenylala-
L-amino acid to
ii)
a
D
N
N
N
N
D
F3C
F3C
iii)
D
H
L
D
H2N
D-6 4-
D-7 3-
COOH
H2N
COOH
nine derivatives were then subjected to calcium imaging assay using
hT1R2-hT1R3 expressing cells.
DL-6 4-
DL-7 3-
A complex model of D-phenylalanine in the ATD of the hT1R2 has
been constructed using the crystal structure of the glutamate bound
closed form of the ATD of the metabotropic glutamate receptor type-
Scheme 1. Synthesis of diazirinyl
Reagents and conditions: (i) CsOH, Bu4NBr, CH2Cl2, rt, 12 h, 60–68%; (ii) TFA, rt,
2 h, 91–95%; (iii) -amino acid oxidase, pH 7.0, 37 °C, 12 h, 40–45%.
D-phenylalanine derivatives D-6 and D-7.
1(mGluR1, PDB code: 1EWK).9,22
D-Phenylalanine was docked into
L
the binding site with a guide of the salt bond between the amino
group of the ligand and Glu302 of the receptor, which is conserved
and crucial in binding in mGluR1 and also suggested to be crucial
for binding of some sweeteners such as aspartame by mutational
experiments.3 The phenyl group bound a hydrophobic pocket which
in the neutral pH condition at 37 °C25 to give the
D-amino acid
derivative (Scheme 1).
also binds a larger indole moiety of
D
-tryptophan. The docking of the
The enzymatic product was monitored with chiral HPLC column
(CHIROBIOTIC T, with 10% ethanol–water) at 350 nm detection.
Although L-4 was still detected within 2 h, it was completely con-
designed diazirinyl -phenylalanines at the same site suggested that
D
the receptor accepts the diazirinyl moiety at the binding pocket
(Fig. 1). Thus, the photoaffinity labeling of the receptor with this
photophore would validate the complex structure model.
sumed after 12 h. The keto acid derived from the L-form (Fig. 2B,
ꢀ6 min), which is too unstable to isolate in pure form, was re-
The 4-diazirinyl benzyl bromide derivatives 124 was reacted
moved from the reaction mixture by acidification of the mixture,
with N-(diphenylmethylene)glycine tert-butyl ester
3
in the
followed by washing with ethyl acetate. The desired diazirinyl D-
presence of CsOH and Bu4NBr as phase transfer catalyst to afford
phenylalanine skeleton 4 at room temperature, followed by depro-
tection with acidic conditions to afford racemic diazirinyl phenyl-
amino acid derivative D-626 was then obtained as a powder after
lyophilization in moderate yield (Fig. 2C). The positional isomer,
3-diazirinyl D
-phenylalanine derivative D-7,27 was also prepared
alanine DL-6. The racemate was treated with
L
-amino acid oxidase
starting from 3-diazirinyl benzyl bromide derivative 228 with the
identical process in a moderate yield (Scheme 1).
The activity of the diazirinyl phenylalanine derivatives D-6 and
D-7 is shown in Figure 3. The sweetness activity of the
D-amino
acid derivatives was measured by using HEK293T cells transiently
expressing hT1R2-hT1R3 and chimeric G-protein, G
a
16-gust4429
,
Figure 1. A complex model of the phenylalanine derivative at the ligand-binding
site of hT1R2. The residues at the binding site within 5 Å from the ligand (ball-and-
stick model) are shown by stick model and Connolly surface.23 Residue numbers are
given by one letter code. The nitrogen, oxygen, fluorine and carbon atoms are
colored by blue, red deep green and green, respectively. Hydrogen atoms are
omitted for clarity.
Figure 2. Enzymatic resolution of diazirinyl amino acid derivative (DL-6) with
amino acid oxidase. (A) Before enzymatic resolution, (B) after incubation with
L
-
-
L
amino acid oxidase at 37 °C for 12
h and (C) purified D-6. Chiral HPLC was
performed with Chirobiotic T (4.6 Â 250 mm, 10% ethanol, 1 ml/min) at 350 nm
detection.

K. Masuda et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1081–1083
1083
a Priority Area, 18032007, for Scientific Research on Innovative
Areas, 20200038 and for Scientific Research (C), 19510210,
21510219 (M.H.) and 16108004 (K.A.), and the Japan Society for
the Promotion of Science (JSPS) (to A.K.). We also acknowledge
the support by the grant from Research and Development Program
for New Bio-industry Initiatives and JST, CREST. M.H. also thanks
the Fugaku Foundation and Research for Promoting Technological
Seeds for financial support for the study.
References and notes
1. Chandrashekar, J.; Hoon, M. A.; Ryba, N. J.; Zuker, C. S. Nature 2006, 444, 288.
2. Li, X.; Staszewski, L.; Xu, H.; Durick, K.; Zoller, M.; Adler, E. Proc. Natl. Acad. Sci.
U.S.A. 2002, 99, 4692.
3. Xu, H.; Staszewski, L.; Tang, H.; Adler, E.; Zoller, M.; Li, X. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 14258.
4. Cui, M.; Jiang, P.; Maillet, E.; Max, M.; Margolskee, R. F.; Osman, R. Curr. Pharm.
Des. 2006, 12, 4591.
5. Kaneko, T. Nippon Kagaku Kaishi 1938, 59, 433.
6. Kaneko, T. Nippon Kagaku Kaishi 1939, 60, 531.
7. Finley, J. W.; Friedman, M. J. Agric. Food Chem. 1973, 21, 33.
8. De Capua, A.; Goodman, M.; Amino, Y.; Saviano, M.; Benedetti, E. ChemBioChem
2006, 7, 377.
9. Kunishima, N.; Shimada, Y.; Tsuji, Y.; Sato, T.; Yamamoto, M.; Kumasaka, T.;
Nakanishi, S.; Jingami, H.; Morikawa, K. Nature 2000, 407, 971.
10. Morini, G.; Bassoli, A.; Temussi, P. A. J. Med. Chem. 2005, 48, 5520.
11. Brunner, J. Annu. Rev. Biochem. 1993, 62, 483.
12. Gillingham, A. K.; Koumanov, F.; Hashimoto, M.; Holman, G. D. Detection and
Analysis of Glucose Transporters using Photolabelling Techniques. In
Membrane transport:
A Practical Approach; Baldwin, S. A., Ed.; Oxford
University Press: Oxford, 2000; p 193.
13. Hatanaka, Y.; Nakayama, H.; Kanaoka, Y. Rev. Heteroatom Chem. 1996, 14, 213.
14. Tomohiro, T.; Hashimoto, M.; Hatanaka, Y. Chem. Records 2005, 5, 385.
15. Hashimoto, M.; Hatanaka, Y. Eur. J. Org. Chem. 2008, 2513.
16. Nassal, M. J. Am. Chem. Soc. 1984, 106, 7540.
17. Beewen, L.; Bayley, H. Anal. Biochem. 1985, 144, 132.
18. Fishwick, C. W. G.; Sanderson, J. M.; Findlay, J. B. C. Tetrahedron Lett. 1994, 35,
4611.
19. Hashimoto, M.; Hatanaka, Y.; Sadakane, Y.; Nabeta, K. Bioorg. Med. Chem. Lett.
2002, 12, 2507.
Figure 3. Sweet-tasting effect assay. (A) Representative ratiometric images of
calcium ion indicator fura-2 loaded HEK293T cells coexpressing hT1R2-hT1R3 and
G16-gust44 with 1.25 mM D-6 (1) and D-7 (3). The color scale indicates the F340/
F380 ratio, where F340 is fluorescence intensity at 340 nm. Scale bar, 50 lM.
20. Nakashima, H.; Hashimoto, M.; Sadakane, Y.; Tomohiro, T.; Hatanaka, Y. J. Am.
Chem. Soc. 2006, 128, 15092.
Competitive inhibitions were performed by co-incubation with 1.25 mM lactisole (2
for D-6 and 4 for D-7, respectively). Red color images represent responding cells for
sweet potential. (B) Sweetness potential for known chemicals and synthetic
photoreactive compounds. Cells were defined as responding positively when the
21. Meister, A.; Wellner, D. Flavoprotein Amino Acid Oxidase. In The Enzymes;
Boyer, P. D., Ed., 2nd ed.; Elsevier: Amsterdam, 1963; p 609.
22. The initial structural model of the active form of hT1R2 derived from the
crystal structure of the closed form of the ATD of mGluR1 (1EWK) was energy-
minimized with a molecular mechanics program, Discover 3 (Accerlys Inc., San
F340/F380 increased above 0.15 after addition of
a tastant. Number of cell
responses, which trigger with 10 mM aspartame, were set as standard. The degree
of cell response as the number of positively responding cells at 1.25 mM of each
chemical against standard are represented to normalize the response. Each column
represents the mean SE of three independent experiments.
Diego CA, USA), until the RMSD became less than 0.1 kcal/mol/Å. D-Trp was
then docked into the binding cleft assuming an electrostatic interaction
between Glu302 and the amino group of the ligand. The initial complex
structure model was energy-minimized and then optimized with molecular
dynamics using Discover 3. The molecular dynamics calculation was
performed for 0.2 ns, sampling the structure every 20 ps. Each structure was
energy-minimized as described above. The energetically most stable structure
was selected as the complex structure model.
with Ca2+ imaging analysis as described by Nakajima et al.30 The
cells, which responded to the ligands, showed red color in the
pseudocolor images. The responses to the diazirinyl phenylalanine
derivatives D-6 and D-7 were significantly reduced by addition of
lactisole, an antagonist against hT1R2-hT1R331 (Fig. 3). Both the
23. Connolly, M. L. J. Mol. Graphics 1992, 11, 139.
24. Hashimoto, M.; Kanaoka, Y.; Hatanaka, Y. Heterocycles 1997, 46, 119.
25. Typical procedure for enzymatic resolution; DL-6 was suspended in water
(20 ml) and the suspension was adjusted to pH 7 with NH4OH and formic acid.
photoreactive
ness activity than
D
-amino acids, D-6 and D-7, exhibited higher sweet-
-tryptophan and -phenylalanine, and a similar
L-amino acid oxidase (Sigma A-9378) was added to the suspension, The
D
D
reaction mixture was incubated at 37 °C for 12 h, then made acid with 1 N
formic acid and extracted with ethyl acetate. The aqueous layer was
concentrated and the residue was subjected to chiral HPLC with Chirobiotic
T (4.6 Â 250 mm, 10% ethanol, 1 ml/min) at 350 nm, then lyophilized to afford
colorless amorphous mass.
activity as aspartame had at the same dose (1.25 mM). Since the
trifluoromethyldiazirinyl moiety may occupy the binding site for
the phenyl ring of D-tryptophane, the ligand-binding site would
have a significantly large space for accepting hydrophobic moieties
of ligands.
26. D-6: [
a
]
D (c 0.2, MeOH) +69.0, literature for L-6 [
a
]
D (c 0.115, MeOH) À70.011).
1H NMR (500 MHz, CD3OD) d 7.46 (d, 2H, J = 7.9 Hz), 7.27 (d, 2H, J = 7.9 Hz),
3.84 (dd, 1H, J = 8.2, 4.8 Hz), 3.38 (m, 1H), 3.10 (dd, 2H, J = 14.5, 8.2 Hz).
These results indicate that the preparation of the diazirinyl
27. D-7: [
a
]
(c 0.2, MeOH) +8.9, 1H-NMR (500 MHz, CD3OD)
d 7.37 (d, 1H,
D
D
-phenylalanine derivatives through the enzymatic resolution with
J = 8.2 Hz), 7.18 (d, 2H, J = 8.2 Hz), 7.05 (s, 1H), 3.67 (m, 1H), 3.22 (m, 1H), 2.95
(m, 1H).
L-amino acid oxidase is effective and that these photoreactive com-
28. Blencowe, A.; Caiulo, N.; Cosstick, K.; Fagour, W.; Heath, P.; Hayes, W.
Macromolecules 2007, 40, 939.
29. Ueda, T.; Ugawa, S.; Yamamura, H.; Imaizumi, Y.; Shimada, S. J. Neurosci. 2003,
23, 7376.
pounds have enough affinity with the sweet taste receptor to elu-
cidate the binding site for the ligands in the sweet taste receptor.
30. Nakajima, K.; Asakura, T.; Maruyama, J.; Morita, Y.; Oike, H.; Shimizu-Ibuka, A.;
Misaka, T.; Sorimachi, H.; Arai, S.; Kitamoto, K.; Abe, K. Appl. Environ. Microbiol.
2006, 72, 3716.
31. Jiang, P.; Cui, M.; Zhao, B.; Liu, Z.; Snyder, L. A.; Benard, L. M.; Osman, R.;
Margolskee, R. F.; Max, M. J. Biol. Chem. 2005, 280, 15238.
Acknowledgments
This research was partially supported by Ministry of Education,
Science, Sports and Culture Grant-in-Aid for Scientific Research on