Identification of putative metabolites of docosahexaenoic acid as potent PPARγ agonists and antidiabetic agents

Article abstract of DOI:10.1016/j.bmcl.2004.11.053

We found that putative metabolites of docosahexaenoic acid (DHA) are strong PPARγ activators and potential antidiabetic agents. We designed DHA derivatives based on the crystal structure of PPARγ, synthesized them and evaluated their activities in vitro and in vivo. The efficacy of 5E-4-hydroxy-DHA 2a as a PPARγ activator was about fourfold stronger than that of pioglitazone. Furthermore, the 4-keto derivative (10b) showed antidiabetic activity in animal models without producing undesirable effects such as obesity and hepatotoxicity.

Full text of DOI:10.1016/j.bmcl.2004.11.053

Bioorganic & Medicinal Chemistry Letters 15 (2005) 517–522
Identification of putative metabolites of docosahexaenoic acid
as potent PPARc agonists and antidiabetic agents
Keiko Yamamoto,a,* Toshimasa Itoh,a Daijiro Abe,a,b Masato Shimizu,b
Tomoatsu Kanda,c Takatoshi Koyama,c Masazumi Nishikawa,d Tadakazu Tamai,d
Hiroshi Ooizumie and Sachiko Yamadaa,b,
*
aInstitute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, 2-3-10 Kanda-Surugadai,
Chiyoda-ku, Tokyo 101-0062, Japan
bSchool of Biomedical Science, Tokyo Medical and Dental University, Tokyo 101-0062, Japan
cGraduate School of Allied Health Sciences, Tokyo Medical and Dental University, Tokyo 113-8519, Japan
dCentral Research Institute, Maruha Corporation, Tsukuba 300-4295, Japan
eGraduate School of Biostudies, Kyoto University, Kyoto 606-8501, Japan
Received 26 October 2004; revised 21 November 2004; accepted 22 November 2004
Available online 16 December 2004
Abstract—We found that putative metabolites of docosahexaenoic acid (DHA) are strong PPARc activators and potential antidi-
abetic agents. We designed DHA derivatives based on the crystal structure of PPARc, synthesized them and evaluated their activi-
ties in vitro and in vivo. The efficacy of 5E-4-hydroxy-DHA 2a as a PPARc activator was about fourfold stronger than that of
pioglitazone. Furthermore, the 4-keto derivative (10b) showed antidiabetic activity in animal models without producing undesirable
effects such as obesity and hepatotoxicity.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Peroxisome proliferator-activated receptors (PPARa, c,
d) are nuclear receptors, which act as biosensors of fatty
acids and play key roles in fatty acid metabolism and
homeostasis. Disorders of lipid metabolism are a major
threat to human health, leading to obesity, atherosclero-
sis, cardiovascular diseases, insulin resistance, and type
2 diabetes. The incidence of type 2 diabetes has in-
creased dramatically over the past two decades, and it
is now becoming one of the biggest public-health prob-
lems worldwide. It is well accepted that PPARc is a tar-
get of antidiabetic agents such as thiazolidinediones
(TZDs).1,2 In agreement with this, mutations in the li-
gand-binding domain (LBD) of PPARc in subjects
exhibiting severe insulin resistance and early-onset type
2 diabetes mellitus have been reported.3 Although TZDs
are widely used, they have secondary effects such as
obesity, edema and hepatotoxicity.4,5 Recently, much
attention has been focused on alternatives such as car-
boxylic acid derivatives, which are agonists of both
PPARa and c,6,7 since the high PPARc selectivity of
TZDs is thought to be a cause of their side effects.8,9
Based on the idea that, for chronic diseases, compounds
that are closely related to natural substances would be
the drugs of better choice, we became interested in devel-
oping fatty acid derivatives as potent antidiabetic agents
that would target PPARc. Some arachidonic acid
metabolites, such as 15-deoxy-D12,14-prostaglandin J2
(PGJ2), are natural ligands with moderate affinity for
PPARc.10 However, they have not been reported to
have antidiabetic activity in vivo. We focused our atten-
tion on docosahexaenoic acid (DHA) derivatives, be-
cause DHA (1) has long been used as a functional
food worldwide and is known to have beneficial effects
in several human diseases including atherosclerosis,
asthma, cardiovascular disease, cancer, and depres-
sion.11 Recently, various oxidative metabolites of
DHA have been isolated and reported to exert anti-
inflammatory effects.12 However, their potency in trans-
activation of PPARc remains unknown. In this paper
we report the design of various DHA derivatives that
bind to PPARc, based on the crystal structure of
PPARc/agonist complex. We found that several of these
Keyword: PPARc agonist.
*
Corresponding authors. Tel.: +81 3 5280 8038; fax: +81 3 5280 8005
(K.Y.); e-mail addresses: yamamoto.mr@tmd.ac.jp; yamada.mr@
tmd.ac.jp
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.11.053

518
K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 517–522
compounds show higher activity in transactivation of
PPARc than PGJ2 and that some have even higher
activity than a TZD derivative, pioglitazone. These
DHA derivatives are putative metabolites or their
related compounds.12–14 We further demonstrated that
at least one of these derivatives has useful antidiabetic
activity in vivo.
with its TZD ring oriented towards the AF2 helix and
forming hydrogen bonds with five residues, Q286 (H3),
S289 (H3), H323 (H4), H449 (H11), and Y473 (H12)
(Fig. 2a). TZD derivatives are poor PPARa agonists,
because the TZD head group creates a steric clash with
Y314, which corresponds to H323 in PPARc.6,7 Our
docking analysis indicated that parent DHA (1) is easily
accommodated in the PPARc–LBP, and that the carb-
oxyl group forms hydrogen bonds with four of the five
residues that form hydrogen bonds with the TZD ring
(S289, H323, H449, and Y473) (Fig. 2b). In order to de-
sign DHA derivatives as potent PPARc agonists, we
searched for hydrophilic residues that would face
DHA when it is docked into the LBP and would be able
to form the fifth hydrogen bond. We paid attention to
Y327 in helix 4/5 whose hydroxyl group faces the C(4)
of DHA. Thus, we designed a key compound, 5E-4-
hydroxydocosahexaenoic acid (4-OHDHA, 2a), which
has a hydroxyl group at C(4). As shown in Figure 2c,
the 4-hydroxyl group of (4S)-2a does indeed form a
hydrogen bond with Y327. Lactones, 3 and 6, are pre-
cursors of 4-OHDHA (2a) in the synthesis from DHA,
and 6 may serve as a pro-drug of 2a. Lactol 7 and diol
8 were also designed as possible pro-drugs of 2a, since
these are expected to be converted to 2a by oxidation
in vivo. The other compounds (4, 5, and 9–13) were de-
signed to allow structure–activity relationship studies.
We designed DHA derivatives (Fig. 1) that might have
higher activity than the parent compound based on the
crystal structure of the PPARc–LBD/rosiglitazone com-
plex15 and using docking software FLEXX (Tripos, St.
Louis). The crystal structure indicated that rosiglitazone
docks into the ligand-binding pocket (LBP) of PPARc,
H
COOH
15-deoxy-∆12,14-PGJ2
O
O
NH
pioglitazone
S
S
N
O
O
O
O
rosiglitazone
NH
N
N
O
A total of 19 DHA derivatives were synthesized as
shown in Scheme 1. We used iodolactonization of DHA
as the key reaction to introduce functional group at
C(4). Thus, all the compounds except for the 5E-D5-
DHA derivatives (13a and 13b) were synthesized from
DHA via the key compound 3. Compound 13a and
13b were synthesized by coupling of Wittig reagent 14
derived from EPA with aldehyde 15. Compounds with
asymmetric carbon at C(4) (2, 4, 6, 8, 9, and 12) are race-
mic mixtures and those with two asymmetric carbons (3,
5, and 7) are a mixtures of possible four isomers. Struc-
tures of newly synthesized compounds were confirmed
by spectral analyses.16 Details of the synthetic proce-
dures are described in a separate paper in preparation.
COOH
COOR
DHA (1)
OR'
2a: R = H, R' = H
2b: R = CH3, R' = H
2c: R = H, R' = CH3
2d: R = CH3, R' = CH3
2e: R = CH3, R' = Ac
I
O
O
X
X
COOCH3
F
X
9
3
O
O
COOR
O
The transcriptional activities of the 19 DHA derivatives
were evaluated in Cos7 cells by dual luciferase assays
using a reporter plasmid containing four copies of
MH100 GAL4 binding site (MH100 · 4-TK-Luc),17
GAL4-hPPARc chimera expression plasmid (pSG5-
GAL-hPPARc)1 and an internal control plasmid con-
taining sea pansy luciferase expression constructs
(pRL-CMV).18 The relative activities of the DHA deriva-
tives compared with DHA, PGJ2, and pioglitazone are
summarized in Table 1. All compounds tested had higher
potency than the parent fatty acid (1) and several (2a, 9,
10a, and 10b) showed activity comparable to or higher
than that of pioglitazone. It should be noted that the
compound with the highest activity, 4-OHDHA (2a),
is a putative metabolite of DHA; DHA has been re-
ported to yield some 4-hydroxylated metabolites in vitro
under the influence of 5-lipoxygenase.12 4-Oxo-DHA
10a may also be a metabolite produced by dehydrogena-
tion of 2a, since oxidation of 5-HETE to 5-oxo-ETE has
been reported under similar conditions.19
X
4
10a: R = H
HO
10b: R = CH3
O
O
X
X
X
COOCH3
COOR
5
O
11
O
X
6
OH
O
OH
X
12a: R = H
12b: R = CH3
7
X
COOR
X
OH
OH
13a: R = H
13b: R = CH3
8
X =
Figure 1. Structures of DHA derivatives.

K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 517–522
519
Figure 2. Structures of PPARc–ligand complexes. Hydrogen bonds are depicted by red dotted lines: (a) crystal structure of the PPARc–LBD and
rosiglitazone complex.13 The TZD ring forms hydrogen bonds with five residues, Q286, S289, H323, H449, and Y473; (b) model of the PPARc
and DHA (1) complex. The carboxyl group forms hydrogen bonds with only four residues, S289, H323, H449, and Y473; (c) model of the PPARc
and (4S)-4-OHDHA 2a complex. A fifth hydrogen bond can be formed between the 4-hydroxyl group and Y327; (d) superimposition of (4S)-4-
OHDHA 2a and rosiglitazone in the PPARc–LBP. The two ligands occupy almost the same positions in the pocket.
I2, collidine
aq KOH
I
O
O
X
COOR
Bu3SnH
O
O
Et3N, MeOH
X
X
DHA (1)
OH
4
3
12a: R = H
12b: R = CH3
DBU
X
COOR
O
O
X
COOR
HO
X
O
O
X
OH
2a: R = H
2b: R = CH3
OR'
aq KOH or
6
Et3N, MeOH
2d: R = CH3, R' = CH3
2e: R = CH3, R' = Ac
2c: R = H, R' = CH3
5
Dess-Matin
DIBAL
aq
or Swern
oxidation
KOH
DAST
O
OH
X
X
OH
X
COOR
Tebbe reagent
X
COOCH3
OH
8
X
COOCH3
7
DIBAL
O
F
9
10a: R = H
11
10b: R = CH3
PPh3Br
X
COOR
OHC
CO2Me
+
X =
13a: R = H
15
13b: R = CH3
14
Scheme 1.
The dose–response results for the potent compounds 2a,
10a, and 10b are interesting (Fig. 3a). Compounds 2a
and 10a showed different patterns from both PGJ2
and pioglitazone. 4-OHDHA 2a exhibited weaker activ-
ity than pioglitazone at 1 lM, but was stronger at 5 lM.
It is known that a large proportion (>99%) of the free
fatty acids in plasma is bound to serum albumin. Since
the assay solution contained 5% FBS, we suppose that
4-OHDHA 2a might bind to albumin and be unable
to enter cells readily at low concentration. Therefore,
we re-evaluated the activity of 2a in the absence of
FBS, and found that it showed higher activity than

520
K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 517–522
Table 1. Activity of DHA derivatives on human PPARca
pioglitazone even at 1 lM. Indeed, 2a was more than
four times as effective as pioglitazone (Fig. 3b). These re-
sults indicate that, once they have been transported into
cells, at least some of the DHA derivatives have much
higher activities than those observed in the presence of
5% FBS (Table 1).
Compd
Activityb
Vehicle
PGJ2
Pioglitazone
1.0
5.9
6.9
1.2
13.4
3.8
1.7
2.3
1.9
1.4
3.5
1.4
2.5
5.7
5.1
6.0
7.8
6.6
1.5
2.0
1.7
1.2
1.3
DHA 1
2a
2b
2c
Methyl ether 2c and acetate 2e showed significantly low-
er activity than 2a, indicating the importance of the 4-
hydroxyl group. As shown in Table 1, this hydroxyl
group is replaceable with fluorine (9) or keto group
(10a and 10b), but not with hydrogen (13a) or methylene
(11). The results emphasize the importance of a hydro-
philic substituent at C(4). It should also be noted that
saturation of the double bond at C(5) with hydrogen
significantly lowered the potency, as the poor activity
of compound 12 shows. This may be explained by the
entropic effect of the rigid 5,7-conjugated diene sys-
tem.20 Thus, it is clear that both a hydrophilic group
at C(4) and a 5E,7Z-conjugated diene structure are
important in producing highly potent DHA derivatives.
Lactol 7 and diol 8 showed significant activity, but it is
not known whether these two compounds exhibit their
activity after conversion to oxidative metabolites such
as 2a.
2d
2e
3
4
5
6
7
8
9
10a
10b
11
12a
12b
13a
13b
a All compounds were tested at 5 lM in the presence of 5% FBS using
Cos7 cells by dual luciferase assay.
b Activities are presented by fold induction of PPARc activation.
The in vivo antidiabetic activities of 4-OHDHA (2a) and
keto ester 10b were evaluated using male db/db mice and
Zucker diabetic fatty (ZDF) rats. The db/db mouse
bears a defect in the leptin receptor gene, which leads
to an uncontrolled appetite to develop obesity-induced
diabetes. ZDF rats are also deficient in leptin receptors.
4-OHDHA (2a) had little effect in db/db mice (data not
shown), probably because of poor availability to the tar-
get tissues. Preliminary experiments in normal rats
showed that absorption of 2a into the lymph and serum
is poor. On the other hand, as seen in Table 2, 10b sig-
nificantly reduced blood glucose (BG) levels in db/db
mice at a dose of 10 mg/kg. The antidiabetic effect of
10b was slightly lower than that of pioglitazone. Inter-
estingly, no dose-dependent effect on the BG levels
was apparent at a higher dose (100 mg/kg). In contrast
to pioglitazone, 10b had little effect on serum triglyceride
(TG) levels. A similar activity profile was observed in the
ZDF rats, in which 10b also reduced BG but not TG lev-
els. Thus, keto ester 10b has an activity profile distinct
from that of pioglitazone. Furthermore, in preliminary
high-dose toxicity tests in normal rats (1, 0.2, and
0.04 g/kg for 14 days), 10b produced neither obesity
nor liver toxicity (data not shown). The distinct activity
profile of 10b in vivo may mean that it shows antidia-
betic activity without producing the undesirable effect
of obesity.
2
0.1 µM
1.8
1 µM
FBS (+)
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
5 µM
10 µM
(a)
-
PGJ2 pioglit
FBS (-)
DHA
2a
10a
10b
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
(b)
-
PGJ2
pioglit
DHA
2a
In the docking analysis, all of the potent compounds
(2a, 9, 10a,b) were accommodated appropriately in
the PPARc–LBP and hydrogen-bonded with Y327.
Figure 2d shows a superimposition of (4S)-2a and rosigli-
tazone in the PPARc-LBP. Compound (4S)-2a adopts
the same overall shape as rosiglitazone in the LBP. The
tail of (4S)-2a is extended a little more toward the open
channel between helix 3 and the b-sheet in comparison
with rosiglitazone. The hydrogen bond between Y327
Figure 3. Dose–response relationship of the most potent compounds
(2a, 10a, and 10b) with respect to PPARc activation. Compounds were
tested in the presence (a) or absence (b) of 5% FBS using Cos7 cells by
dual luciferase assay: (a) in the presence of 5% FBS, 2a and 10a showed
different dose–response patterns from both PGJ2 and pioglitazone.
Compound 2a showed weaker activity than pioglitazone at 1 lM but
stronger activity at 5 lM; (b) in the absence of FBS, 2a showed higher
activity than pioglitazone even at 1 lM and the efficacy of 2a was four-
times stronger than that of pioglitazone.

K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 517–522
Table 2. In vivo activity of 10b in animal models
521
Compound
Dose (mg/kg)
Blood glucose (mg/dL)
Triglyceride (mg/dL)
Body weight gain (g)
db/db Mousea
Vehicle
––
945 122
432 74**
653 66**
514 70**
193 81
73 26**
188 66
138 66
5.4 0.6
15.5 3.0**
3.2 1.3
Pioglitazone
10b
10b
100
100
10
3.7 0.9
ZDF ratb
Vehicle
Pioglitazone
10b
––
30
30
3
364 80
819 204
109 67**
765 157
958 309
56.2 16.3
96.6 9.5**
61.9 8.6
60.1 9.9
174 70**
228 49**
262 72*
10b
a Six db/db mice/group at 8 weeks old were orally administered once daily with vehicle or 10b in vehicle. After 4 weeks of dosing, mice were
anesthetized with pentobarbital and blood was collected.
b Six ZDF rats/group at 9 weeks old were orally administered once daily with vehicle or 10b in vehicle. After 2 weeks of dosing, rats were anesthetized
with pentobarbital and blood was collected.
*P < 0.05.
**P < 0.01.
6. Cronet, P.; Petersen, J. F.; Folmer, R.; Blomberg, N.;
Sjoblom, K.; Karlsson, U.; Lindstedt, E. L.; Bamberg, K.
Structure 2001, 9, 699–706.
7. Xu, H. E.; Lambert, M. H.; Montana, V. G.; Plunket, K.
D.; Moore, L. B.; Collins, J. L.; Oplinger, J. A.; Kliewer,
S. A.; Gampe, R. T., Jr.; McKee, D. D.; Moore, J. T.;
Willson, T. M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
13919–13924.
8. Hallakou, S.; Doare, L.; Foufelle, F.; Kergoat, M.;
Guerre-Millo, M.; Berthault, M. F.; Dugail, I.; Morin,
J.; Auwerx, J.; Ferre, P. Diabetes 1997, 46, 1393–1399.
9. Mukherjee, R.; Hoener, P. A.; Jow, L.; Bilakovics, J.;
Klausing, K.; Mais, D. E.; Faulkner, A.; Croston, G. E.;
Paterniti, J. R., Jr. Mol. Endocrinol. 2000, 14, 1425–1433.
10. Kliewer, S. A.; Lenhard, J. M.; Willson, T. M.; Patel, I.;
Morris, D. C.; Lehmann, J. M. Cell 1995, 83, 813–819.
11. Horrocks, L. A.; Yeo, Y. K. Pharmacol. Res. 1999, 40,
211–225.
12. Hong, S.; Gronert, K.; Devchand, P. R.; Moussignac, R.
L.; Serhan, C. N. J. Biol. Chem. 2003, 278, 14677–14687.
13. Corey, E. J.; Shih, C.; Cashman, J. R. Proc. Natl. Acad.
Sci. U.S.A. 1983, 80, 3581–3584.
14. Fischer, S.; Schacky, C. v.; Siess, W.; Strasser, T.; Weber,
P. C. Biochem. Biophys. Res. Commun. 1984, 120, 907–
918.
15. Nolte, R. T.; Wisely, G. B.; Westin, S.; Cobb, J. E.;
Lambert, M. H.; Kurokawa, R.; Rosenfeld, M. G.;
Willson, T. M.; Glass, C. K.; Milburn, M. V. Nature
1998, 395, 137–143.
and 4-hydroxyl group must be essential because only
compounds with a hydroxyl, carbonyl or fluorine group
at C(4) activate PPARc. As noted above, the planar
5E,7Z-conjugated diene structure is also important for
high biological potency. This diene moiety fits into a
narrow, hydrophobic tunnel formed by C285 (H3),
L330 (H5), and I326 (H5), and overlaps with the middle
aromatic ring of rosiglitazone, as seen in Figure 2d.
PPARc plays an important role in energy accumulation,
whereas PPARa is involved in energy consumption.
Therefore, dual agonists of PPARc and a are now
thought to be good antidiabetic agents that will not give
rise to undesirable effects such as obesity and edema.21
We identified putative metabolites of DHA that strongly
activate PPARc, of which 4-keto ester 10b showed anti-
diabetic activity in vivo without producing any side
effects. These DHA derivatives may function as dual
agonists, as they possess a carboxyl group but not a
TZD ring.6,7 Thus, we present here novel potential leads
for treating type 2 diabetes. Although they are not as ac-
tive as pure synthetic aromatic agents in vivo at the pres-
ent time, they may be nontoxic agents that would not
cause obesity during long-term treatment. Further stud-
ies are now needed to prove the potential of these com-
pounds and to clarify their behavior in vivo.
16. Spectral data. (5E,7Z,10Z,13Z,16Z,19Z)-4-Hydroxy-
5,7,10,13,16,19-docosahexaenoic acid (2a): 1H NMR: d
0.97 (3H, t, J = 7.5 Hz, H-22), 1.88 (2H, m, H-3), 2.08
(2H, m, H-21), 2.48 (2H, t, J = 7.3 Hz, H-2), 2.80–2.91
(6H, m, H-12, 15, 18), 2.97 (2H, t, J = 6.6 Hz, H-9), 4.25
(1H, dd, J = 12.6, 6.6 Hz, H-4), 5.32–5.43 (9H, m, H-8, 10,
11, 13, 14, 16, 17, 19, 20), 5.68 (1H, dd, J = 15.1, 6.5 Hz,
H-5), 6.01 (1H, t, J = 11.0 Hz, H-7), 6.54 (1H, dd,
J = 15.1, 11.0 Hz, H-6); MS m/z 344 (M+, 1), 327 (3),
326 (7), 297 (4), 246 (14), 187 (16), 117 (50), 108 (55), 79
(100); HRMS Calcd for C22H30O2 (M+ÀH2O) 326.2246.
Found 326.2256. Methyl(5E,7Z,10Z,13Z,16Z,19Z)-4-
References and notes
1. Lehmann, J. M.; Moore, L. B.; Smith-Oliver, T. A.;
Wilkison, W. O.; Willson, T. M.; Kliewer, S. A. J. Biol.
Chem. 1995, 270, 12953–12956.
2. Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B.
R. J. Med. Chem. 2000, 43, 527–550.
3. Barroso, I.; Gurnell, M.; Crowley, V. E.; Agostini, M.;
Schwabe, J. W.; Soos, M. A.; Maslen, G. L.; Williams, T.
D.; Lewis, H.; Schafer, A. J.; Chatterjee, V. K.; OÕRahilly,
S. Nature 1999, 402, 880–883.
1
oxo-5,7,10,13,16,19-docosahexaenoate (10b): H NMR: d
4. Stumvoll, M.; Haring, H. U. Ann. Med. 2002, 34, 217–
224.
5. Zanchi, A.; Chiolero, A.; Maillard, M.; Nussberger, J.;
Brunner, H. R.; Burnier, M. J. Clin. Endocrinol. Metab.
2004, 89, 1140–1145.
0.97 (3H, t, J = 7.5 Hz, H-22), 2.07 (2H, m, H-21), 2.66
(2H, t, J = 6.7 Hz, H-2), 2.77–2.88 (6H, m, H-12, 15, 18),
2.92 (2H, t, J = 6.7 Hz, H-3), 3.01 (2H, t, J = 7.3 Hz, H-9),
3.69 (3H, s, CO2Me), 5.28–5.49 (8H, m, H-10, 11, 13, 14,
16, 17, 19, 20), 5.87 (1H, m, H-6), 6.15 (1H, d, J = 11.4 Hz,

522
K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 517–522
H-8), 6.20 (1H, d, J = 15.5 Hz, H-5), 7.56 (1H, dd,
VanHooke, J.; DeLuca, H. F.; Yamada, S. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 1467–1472.
19. Powell, W. S.; Gravelle, F.; Gravel, S. J. Biol. Chem. 1992,
267, 19233–19241.
20. Yamamoto, K.; Sun, W.-Y.; Ohta, M.; Hamada, K.;
DeLuca, H. F.; Yamada, S. J. Med. Chem. 1996, 39, 2727–
2737.
J = 15.5, 11.4 Hz, H-7); 13C NMR: d 14.3, 20.6, 25.6,
25.7, 25.8, 26.7, 28.0, 35.5, 51.9, 126.4, 127.0, 127.1, 127.7,
128.6, 128.7, 129.5, 129.7, 132.1, 137.1, 140.1, 173.4, 198.3;
MS m/z 356 (M+, 5), 276 (7), 189 (22), 167 (25), 137 (25),
115 (100), 79 (43); HRMS Calcd for C23H32O3 (M+)
356.2351. Found 356.2335.
17. Kang, T.; Martins, T.; Sadowski, I. J. Biol. Chem. 1993,
268, 9629–9635.
18. Yamamoto, K.; Masuno, H.; Choi, M.; Nakashima, K.;
Taga, T.; Ooizumi, H.; Umesono, K.; Sicinska, W.;
21. Yajima, K.; Hirose, H.; Fujita, H.; Seto, Y.; Fujita, H.;
Ukeda, K.; Miyashita, K.; Kawai, T.; Yamamoto, Y.;
Ogawa, T.; Yamada, T.; Saruta, T. Am. J. Physiol.
Endocrinol. Metab. 2003, 284, E966–E971.