Synthesis of docosahexaenoic acid derivatives designed as novel PPARγ agonists and antidiabetic agents

Article abstract of DOI:10.1016/j.bmc.2005.07.074

To discover novel peroxisome proliferator-activated receptor γ (PPARγ) agonists that could be used as antidiabetic agents, we designed docosahexaenoic acid (DHA) derivatives (2 and 3), which have a hydrophilic substituent at the C(4)-position, based on the crystal structure of the ligand-binding pocket of PPARγ. These compounds were synthesized via iodolactone as a key intermediate. We found that both DHA derivatives (2 and 3) showed PPARγ transactivation higher than, or comparable to, that of pioglitazone, which is a TZD derivative used as an antidiabetic agent. DHA derivatives related to these potent compounds 2 and 3 were also synthesized to study structure-activity relationships. Furthermore, 4-OH DHA 2, which shows strong PPARγ transcriptional activity, was separated as an optically pure form.

Full text of DOI:10.1016/j.bmc.2005.07.074

Bioorganic & Medicinal Chemistry 14 (2006) 98–108
Synthesis of docosahexaenoic acid derivatives designed as novel
PPARc agonists and antidiabetic agents
Toshimasa Itoh,a Itsuki Murota,b Kazuyoshi Yoshikai,b
Sachiko Yamadaa and Keiko Yamamotoa,*
aInstitute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, 2-3-10 Kanda-Surugadai,
Chiyoda-ku, Tokyo 101-0062, Japan
bCentral Research Institute, Maruha Corporation, Tsukuba 300-4295, Japan
Received 8 July 2005; revised 28 July 2005; accepted 28 July 2005
Available online 29 September 2005
Abstract—To discover novel peroxisome proliferator-activated receptor c (PPARc) agonists that could be used as antidiabetic
agents, we designed docosahexaenoic acid (DHA) derivatives (2 and 3), which have a hydrophilic substituent at the C(4)-position,
based on the crystal structure of the ligand-binding pocket of PPARc. These compounds were synthesized via iodolactone as a key
intermediate. We found that both DHA derivatives (2 and 3) showed PPARc transactivation higher than, or comparable to, that of
pioglitazone, which is a TZD derivative used as an antidiabetic agent. DHA derivatives related to these potent compounds 2 and 3
were also synthesized to study structure–activity relationships. Furthermore, 4-OH DHA 2, which shows strong PPARc transcrip-
tional activity, was separated as an optically pure form.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
acid catabolism in the liver and skeletal muscle, while
PPARc regulates fatty acid storage in adipose tissue.
Mutations of PPARc have been reported to cause type
2 diabetes, indicating the importance of PPARc in glu-
cose homeostasis.5
The incidence of type 2 diabetes has increased dramati-
cally over the past two decades, and it is now becoming
one of the biggest public health problems worldwide.
Thiazolidinediones (TZDs), such as pioglitazone and
rosiglitazone, are widely used for patients with type 2
diabetes to lower their plasma glucose level. However,
these drugs have secondary effects such as obesity, ede-
ma, and hepatotoxicity.1,2 Therefore, there is currently
a need to develop innovative agents for treatment of
type 2 diabetes without side effects.
We are now developing novel PPARc agonists as antidi-
abetic agents that do not cause side effects.6 We consid-
ered that natural products and their derivatives,
particularly biomolecules, would be appropriate for treat-
ment of chronic diseases such as diabetes because humans
have pathways for metabolism or excretion of biomole-
cules for protection against their undesirable effects. We
therefore focused on docosahexaenoic acid (DHA, 1)
and its metabolites (Fig. 1). DHA is a long-chain polyun-
saturated fatty acid present in large amounts in the adult
mammalian brain and retina.7 It is an essential fatty acid,
a major constituent of nutrients rich in n-3 polyunsaturat-
ed fatty acids, and has beneficial effects on blood choles-
terol levels and insulin sensitivity.8,9 In addition, DHA
and its metabolites would be expected not to have marked
side effects because it has been used for some time as a
functional food ingredient throughout the world. Here,
we describe the synthesis of DHA derivatives, designed
on the basis of the crystal structure of the ligand binding
domain (LBD) of PPARc, and their related compounds.
It is well accepted that peroxisome proliferator-activated
receptor c (PPARc) is the target of TZDs.3,4 Peroxisome
proliferator-activated receptors (PPARs) consist of
three subtypes—a, c, d—and are members of the nuclear
receptor superfamily of ligand-activated transcription
factors. These three PPARs play an important role in
lipid and glucose metabolism. PPARa promotes fatty
Keywords: PPARc agonist; Docosahexaenoic acid; Antidiabetic agent;
Iodolactonization.
*
Corresponding author. Tel.: +81 3 5280 8038; fax: +81 3 5280
8005; e-mail: yamamoto.mr@tmd.ac.jp
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.07.074

T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
99
dation gave the 4-oxo-compound 7. Attempts to hydro-
lyze methyl ester 7 to carboxylic acid 3 under basic
conditions resulted in production of a complex mixture
by double bond isomerization. This reaction was im-
proved (80% yield) by using lipase PS (Amano), which
works under neutral conditions. Transactivation assay
of PPARc showed that all synthetic compounds (2–7)
had higher activity than the parent compound DHA
(1) (Fig. 2a).6 It should be noted that 4-OH DHA 2
and 4-oxo-DHA 3, designed with the aid of docking
software and synthesized, showed activity that was com-
parable with, or higher than, that of pioglitazone.
O
pioglitazone
NH
S
N
O
O
COOH
DHA (1)
OH
O
COOH
COOH
2
As a next step, we evaluated the importance of the
hydroxyl or oxo group at C(4) of DHA derivatives 2
and 3. We synthesized the methyl ethers (8 and 9) and
acetate (10) as compounds protecting the 4-hydroxyl
group (Scheme 2). Ether 8 was obtained by treating 6
with CH3I in the presence of Ag2O. Transactivation as-
say showed that both methyl ethers, 8 and 9, and the
acetate 10 had significantly reduced activity (Fig. 2b),
indicating the importance of the 4-hydroxyl group
(Scheme 2). To confirm this hypothesis, we synthesized
the 4-fluorinated compound (11) and compounds with
no substituent at C(4) (14 and 15). The former (11)
was designed as a molecule capable of forming a hydro-
gen bond between the F-substituent and the hydroxyl
group of Y327, while the latter two (14 and 15) were de-
signed as counterparts not capable of forming the corre-
sponding hydrogen bond. A compound with a fluorine
substituent (11) was successfully synthesized by treating
6 with DAST.13 Attempts to reduce the 4-hydroxyl
group to yield 14 and 15 failed to afford a complex mix-
ture including isomerization products. We synthesized
these by an alternative method (Scheme 3). Wittig re-
agent 12 was derived by (1) iodolactonization of eicosa-
pentaenoic acid, (2) hydrolysis under basic conditions,
(3) oxidative cleavage with sodium periodate followed
by reduction with NaBH4, and (4) direct bromination
of alcohol in the presence of CBr4 and Ph3P, followed
by treatment with Ph3P in acetonitrile in a sealed tube.
Aldehyde 13 was derived from d-valerolactone by meth-
anolysis followed by Swern oxidation and then the
Wittig reaction with (triphenylphosphoranylidene)acetal-
dehyde. Target compound 15 was obtained by coupling
of Wittig reagent 12 and aldehyde 13. Fluorinated com-
pound 11 showed potent activity, while compounds 14
and 15, having no substituent at C(4), showed weak
3
Figure 1.
Interestingly, our designed compounds are also putative
metabolites of DHA (Fig. 1).
2. Results and discussion
The X-ray crystal structure of PPARc-LBD docked
with the TZD derivative, rosiglitazone, has been report-
ed.10 On the basis of the three-dimensional structure of
the ligand-binding pocket, we designed 4-OH DHA (2)
and 4-oxo-DHA (3) as agonists for PPARc. Our dock-
ing analysis indicated that the 4-hydroxyl and 4-oxo
groups of DHA derivatives could generate a new hydro-
gen bond with the hydroxyl group of Y327 lining the li-
gand-binding pocket of PPARc.6 Using DHA as a
starting material, 4-OH DHA (2) and 4-oxo-DHA (3)
were synthesized as shown in Scheme 1. Iodolactoniza-
tion of DHA, which is the key reaction for introducing
a functional group at C(4), was performed in dichloro-
methane in the presence of iodine together with c-colli-
dine. Treatment of iodolactone 4 with DBU afforded
dehydrohalogenation product 5, which was converted
to 4-OH DHA 2 by basic hydrolysis. Oxo-compound 3
was synthesized in two different ways. Direct oxidation
of 4-OH DHA 2 with Dess–Martin reagent11,12 gave
the desired product 3 in poor yield (25%). In the second
method, methanolysis of lactone 5 gave the 4-hydroxyl
methyl ester 6, from which Swern or Dess–Martin oxi-
I2, collidine
I
O
O
O
O
X
DBU
DHA (1)
X
2
5
4
X =
X
COOR
X
COOCH3
O
OH
3 R=H
7 R=CH3
6
Scheme 1.

100
T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
a
b
16
14
12
10
8
6
4
2
0
16
14
12
10
8
6
4
2
0
c
d
16
14
12
10
8
6
4
2
0
16
14
12
10
8
6
4
2
0
vehicle pioglit
3
2
23
24
25
26
e
12
10
8
6
4
2
0
Figure 2. Activity of DHA derivatives on human PPARc. All compounds were tested at 5 lM in the presence of 5% FBS using Cos7 cells by dual
luciferase assay. Activities are presented by fold induction of PPARc activation. Details were described in our previous paper.6
OR'
potency, as 20 and 21 showed no activity (Fig. 2c). Thus,
it was clarified that both a hydrophilic group at C(4)
and a 5E,7Z-conjugated diene structure are essential
for producing highly potent DHA derivatives.
8: R = CH3, R' = CH3
9: R = H, R' = CH3
10: R = CH3, R' = Ac
COOR
6
DAST
F
COOCH3
Since lactone 5 showed moderate PPARc transactiva-
tion potency, we synthesized lactone derivatives (23
and 24) to investigate the structure–activity relation-
ships of compounds having a lactone ring (Scheme 5).
Iodolactone 4 was hydrolyzed under basic conditions
to yield the 5-hydroxyl lactone 23, which was produced
via unstable but detectable 4-epoxy DHA 22. Reduction
of iodolactone 4 with Bu3SnH afforded lactone 24
bearing no substituent at C(5). Compounds 23 and 24
showed poor and moderate activity, respectively
(Fig. 2d). Lactol 25 and diol 26 were prepared as
pro-drug candidates for both 4-OH DHA 2 and 4-oxo-
DHA 3. These two compounds were obtained by
treating lactone 5 with DIBAL at low temperature. Inter-
estingly, both compounds showed significant activity, but
it is not known whether they would exhibit their activity
after conversion to oxidative metabolites such as 2.
11
Scheme 2.
activity (Fig. 2b). These results demonstrate that the
hydrophilic substituent at C(4) plays an important role
in binding to PPARc.
Compound 16 with a methylene substituent at C(4) was
designed to test the importance of the 4-oxo group. This
was synthesized by treating 7 with Tebbe reagent.14
Methylene compound 16 showed significantly lower
activity than 4-oxo-DHA 3, indicating the importance
of the 4-oxo group. Together, these results emphasize
the indispensability of a hydrophilic substituent at C(4).
To examine the significance of the double bond at C(5),
5-dihydro derivatives 17 and 18, and the deconjugated
compound 19 were designed and synthesized (Scheme 4).
In addition, fully saturated derivatives (20 and 21)
were synthesized by catalytic hydrogenation. Saturation
of the double bond at C(5) with hydrogen significantly
lowered the potency, as confirmed by the poor activity
of compound 17 (Fig. 2c). A similar result was obtained
by isomerization of the double bond at C(5), as com-
pound 19 showed poor activity (Fig. 2c). In addition,
complete saturation eliminated the transactivation
To clarify how PPARc distinguishes 4S-OH DHA (2a)
and 4R-OH DHA (2b), we tried asymmetric iodolacton-
ization as well as asymmetric reduction of 4-oxo-DHA
7. However, the reactions yielded products only in poor
enantiomeric excess. Therefore, we carried out optical
resolution of racemic mixture 6. We prepared the isocy-
anate 28 by reaction of dehydroabietylamine (27) with
phosgene15 and converted racemate 6 to a mixture
of diasteromeric urethane derivatives, 29 and 30
(Scheme 6). We separated these diastereomers (29 and
30) by column chromatography on silica gel with an

T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
101
EPA
O
HO
OH
I
O
OH
CO2H
12b
12c
12a
PPh3Br
OHC
CO2CH3
+
COOR
13
12d
Ph3P=CHCHO
14: R = H
15: R = CH3
OHC
CO2CH3
δ-valerolactone
COOCH3
7
16
Scheme 3.
OH
eluent of 5–8% AcOEt/hexane to obtain urethane deriv-
atives of the optically pure forms of 6. Both diastereo-
mers (29 and 30) were treated with trichlorosilane in
the presence of triethylamine to afford optically pure
alcohols, (+)-6a and (ꢀ)-6b, respectively. These two
compounds were hydrolyzed to afford (+)-2a and (ꢀ)-
2b, respectively (Scheme 6).
OH-
or
COOR
24
Et3N, MeOH
17: R = H
18: R = CH3
O
Et3N
COOCH3
COOH
The absolute configuration at C(4) was determined by
the Kusumi–Mosher method.16 Compound (ꢀ)-6b was
converted to (S)-MTPA ester 31 and (R)-MTPA ester
32. The difference in the chemical shifts of these two es-
ters shown in Figure 3 indicates that (ꢀ)-6b and 2b have
the R-configuration, and (+)-6a and 2a have the S-con-
figuration. As shown in Figure 2e, compound (2a) with
4S-hydroxyl group showed higher activity than the race-
mate (2), while compound (2b) with 4R-hydroxyl group
showed lower activity. However, it is clear that both 4S-
OH DHA (2a) and 4R-OH DHA (2b) bind to PPARc to
induce luciferase activity. This result coincides with the
docking analysis by which both 4S- and 4R-hydroxyl
groups could form a hydrogen bond with the hydroxyl
group of Tyr327 lining the ligand binding pocket of
PPARc (data not shown).
7
19
OH
Pd-C
2
H2
20
O
Pd-C
H2
COOH
3
21
Scheme 4.
HO
Bu3SnH
LiOH
O
O
O
O
O
X
X
4
X
COOH
24
22
23
OH
DIBAL
O
OH
DIBAL
OH
X
5
26
25
X =
Scheme 5.

102
T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
vacuum. Silica gel C200 (75–150 lm) was used for col-
umn chromatography, and precoated silica gel 60F254
plates (0.2 mm, Merck) were used for TLC.
3.2. 5-[(3Z,6Z,9Z,12Z,15Z)-1-Iodo-3,6,9,12,15-octadeca-
pentaenyl]dihydro-2(3H)-furanone (4)
H2NH2C
H
OCNH2C
H
28
27
To a solution of DHA (1) (425 mg, 1.30 mmol) and c-
collidine (686 lL, 5.19 mmol) in CH3CN (43 mL) was
added I2 (660 mg, 2.60 mmol) at 0 ꢁC. After being stir-
red at rt for 1 h, the mixture was quenched by addition
of 5% aqueous Na2S2O3 and extracted with ethyl ace-
tate. The organic layer was washed with 10% aqueous
HCl, water, and brine, dried, and evaporated. The resi-
due was chromatographed on silica gel (20 g, 30% ethyl
acetate–hexane) to give 4 (492 mg, 83%) as a yellow oil:
1H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 2.08 (3H, m,
H-21, 3), 2.40 (1H, m, H-3), 2.56 (1H, m, H-2), 2.69 (1H,
m, H-2), 2.80–2.88 (10H, m, H-6, 9, 12, 15, 18), 4.13
(1H, m, H-4), 4.25 (1H, m, H-5), 5.28–5.44 (9H, m, H-
8, 10, 11, 13, 14, 16, 17, 19, 20), 5.55 (1H, m, H-7);
13C NMR d 14.6, 20.8, 25.8, 25.9, 26.0, 26.1, 27.6,
28.8, 34.9, 38.0, 81.0, 127.0, 127.3, 127.6, 128.1, 128.2,
128.7, 128.9, 129.0, 131.8, 132.3, 176.4; MS m/z 454
(M+, 5), 327 (22), 257 (15), 175 (60), 119 (48), 79
(100); HRMS Calcd for C22H31IO2 (M+) 454.1369,
found 454.1358.
X
Y
CO2CH3
6
29; X=OCONHR, Y=H
30; X=H, Y=OCONHR
R=
H
31; X=H, Y=O(S)MTPA
32; X=H, Y=O(R)MTPA
Scheme 6.
OH
+0.12
-0.04 -0.08
COOCH3
+0.02
-0.03
-0.12
3.3. 5-[(1E,3Z,6Z,9Z,12Z,15Z)-1,3,6,9,12,15-Octadeca-
hexaenyl]dihydro-2(3H)-furanone (5)
Figure 3. Determination of stereochemistry at C-4 of (ꢀ)-6b by 1H
NMR analysis (Kusumi–Mosher method). The difference of chemical
shift values between S- and R-MTPA esters is shown in ppm.
A solution of 4 (280 mg, 0.62 mmol) and DBU (111 lL,
0.740 mmol) in benzene (2.8 mL) was stirred at rt for
5 h. The reaction mixture was quenched by addition of
10% aqueous HCl and extracted with ethyl acetate.
The organic layer was washed with water, dried, and
evaporated. The residue was chromatographed on silica
gel (15 g, 20% ethyl acetate–benzene) to give 5 (170 mg,
We synthesized DHA derivatives modified at C(4), of
which several showed PPARc transactivation compara-
ble to, or higher than, that of pioglitazone, which is a
TZD derivative used as an antidiabetic agent. We have
reported that one of them, the methyl ester of 4-oxo-
DHA 7, lowers blood glucose levels in diabetic animal
models without producing undesirable effects such as
obesity and hepatotoxicity.6 Thus, DHA derivatives ap-
pear to have considerable promise for use as antidiabetic
agents. We are now studying the detailed biological
activities of DHA derivatives synthesized.
1
85%): H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 2.08
(3H, m, H-3, 21), 2.40 (1H, m, H-3), 2.58 (2H, m, H-
2), 2.79–2.90 (6H, m, H-12, 15, 18), 2.97 (2H, t,
J = 6.6 Hz, H-9), 5.0 (1H, q, J = 6.9 Hz, H-4), 5.34–
5.43 (8H, m, H-10, 11, 13, 14, 16, 17, 19, 20), 5.52
(1H, m, H-8), 5.68 (1H, dd, J = 15.1, 6.9 Hz, H-5),
6.01 (1H, t, J = 11.0 Hz, H-7), 6.62 (1H, dd, J = 15.1,
11.0 Hz, H-6); 13C NMR d 14.6, 20.8, 25.9, 26.0 (2 car-
bons), 26.5, 28.9, 29.2, 53.3, 80.9, 127.3, 127.5 (2 car-
bons), 128.1, 128.4, 128.7, 128.9, 129.2, 130.3, 132.3,
132.8, 177.1; MS m/z 326 (M+, 10), 257 (8), 246 (15),
131 (62), 79 (100); HRMS Calcd for C22H30O2 (M+)
326.2246, found 326.2234. UV (95% EtOH) kmax
238 nm.
3. Experimental
3.1. General methods
NMR spectra were recorded using CDCl3 as a solvent
unless otherwise noted. 1H NMR spectra were recorded
at 400 MHz with TMS (0.00 ppm) as a reference. 13C
NMR spectra were recorded at 100 MHz with CDCl3
(77.05 ppm) as a reference. 19F NMR spectra were
recorded at 376.5 MHz using trifluorotoluene
(ꢀ63 ppm) as an external reference. Low- and high-res-
olution mass spectra were measured at an ionizing volt-
age of 70 eV. All air-sensitive reactions were run under
argon or nitrogen atmosphere, and reagents were added
through septa using oven-dried syringes. The phrase
‘‘dried and evaporated’’ indicates drying over MgSO4
followed by evaporation of the solvents under house
3.4. (5E,7Z,10Z,13Z,16Z,19Z)-4-Hydroxy-5,7,10,13,16,
19-docosahexaenoic acid (2)
A solution of 5 (93 mg, 0.285 mmol) in 5% KOH/
CH3OH–H2O (19:1, 2.9 mL) was stirred at rt for 5 h.
The reaction mixture was neutralized with 5% aqueous
HCl and then extracted with ethyl acetate. The organic
layer was washed with water, dried, and evaporated.
The residue was chromatographed on silica gel (10 g,
1
40% ethyl acetate–hexane) to give 2 (91 mg, 93%): H

T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
103
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) ; 13C NMR d 14.4, 18.2, 20.7, 25.7, 25.8, 26.3,
30.3, 31.9, 58.3, 71.8, 126.1, 127.1, 127.7, 128.01,
128.08, 128.12, 128.5, 128.76, 128.80, 130.8, 132.2,
135.4, 178.6; 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+ꢀH2O)
326.2246, found 326.2256. UV (95% EtOH) kmax
237 nm (e = 29,500).
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. UV (95% EtOH) kmax 280 nm
(e = 24,000).
3.7. (5E,7Z,10Z,13Z,16Z,19Z)-4-Oxo-5,7,10,13,16,19-
docosahexaenoic acid (3)
To a solution of Dess–Martin Periodinane (370 mg,
0.872 mmol) and Et3N (484 lL, 3.49 mmol) in
CH2Cl2 (5.8 mL) was added 2 (200 mg, 0.581 mmol)
in CH2Cl2 (5.8 mL) at ambient temperature. The mix-
ture was stirred at that temperature for 30 min and
the reaction was quenched with water. The mixture
was extracted with CH2Cl2, and the organic layer
was washed with brine, dried, and evaporated. The
residue was chromatographed on silica gel (20 g,
10% ethyl acetate–hexane) to give 3 (51 mg, 26%):
1H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 1.79–
1.93 (2H, m, H-3), 2.08 (2H, quint, J = 7.4 Hz, H-
21), 2.45 (2H, t, J = 7.4 Hz, H-2), 2.80–2.88 (6H,
m, H-12, 15, 18), 2.97 (2H, t, J = 6.6 Hz, H-9), 3.27
(3H, s, MeO), 3.64 (1H, q, J = 6.7 Hz, H-4), 5.28–
5.46 (9H, m, H-8, 10, 11, 13, 14, 16, 17, 19, 20),
5.50 (1H, dd, J = 15.2, 6.7 Hz, H-5), 6.01 (1H, t,
J = 10.9 Hz, H-7), 6.50 (1H, dd, J = 15.2, 10.9 Hz,
H-6); 13C NMR d 14.5, 20.8, 25.7, 25.8, 25.9, 26.3,
30.3, 30.5, 56.5, 81.3, 127.2, 127.9, 128.0, 128.1,
128.2, 128.8, 128.9, 130.9, 132.3, 133.2, 179.5; MS
m/z 342 (M+, 7), 324 (4), 313 (2), 288 (4), 262 (18),
255 (6), 241 (15), 207 (26), 189 (27), 153 (56), 79
(100); HRMS Calcd for C22H30O3 (M+) 342.2195,
found 342.2201.
3.5. Methyl(5E,7Z,10Z,13Z,16Z,19Z)-4-hydroxy-5,7,10,13,
16,19-docosahexaenoate (6)
A solution of 5 (170 mg, 0.52 mmol) and Et3N (217 lL,
1.56 mmol) in CH3OH (10 mL) was stirred at rt for 6 h.
The reaction mixture was evaporated and the residue
was chromatographed on silica gel (10 g, 10% ethyl ace-
1
tate–benzene) to give 6 (113 mg, 61%): H NMR d 0.97
(3H, t, J = 7.5 Hz, H-22), 1.88 (2H, m, H-3), 2.08 (2H,
quint, J = 7.4 Hz, H-21), 2.45 (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), 3.68 (3 H, s), 4.25 (1H, m, 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) ; 13C NMR d 14.4, 18.4, 20.7, 25.7, 25.8, 26.3, 30.3,
32.2, 51.9, 58.3, 71.7, 125.9, 127.2, 127.8, 128.0, 128.2,
128.5, 128.8, 130.6, 132.2, 135.8, 174.7; MS m/z 358
(M+, 5), 192 (24), 79 (100), 67 (53); HRMS Calcd for
C23H32O2 (M+ꢀH2O) 340.2402, found 340.2380. UV
(95% EtOH) kmax 238 nm.
3.8. Methyl(5E,7Z,10Z,13Z,16Z,19Z)-4-methoxy-
5,7,10,13,16,19-docosahexaenoate (8)
3.6. Methyl(5E,7Z,10Z,13Z,16Z,19Z)-4-oxo-5,7,10,13,
16,19-docosahexaenoate (7)
To a solution of 6 (100 mg, 0.279 mmol) and CH3I
(140 lL, 2.24 mmol) in CH3CN (280 lL) was added
Ag2O (129 mg, 0.558 mmol) at ambient temperature.
After being stirred for 18 h, the mixture was diluted
with ethyl acetate and washed with water. The organic
layer was dried and evaporated. The residue was chro-
matographed on silica gel (10 g, 4 % ethyl acetate–
hexane) to give 8 (61 mg, 59%): 1H NMR d 0.98
(3H, t, J = 7.5 Hz, H-22), 1.79–1.93 (2H, m, H-3),
2.08 (2H, quint, J = 7.4 Hz, H-21), 2.39 (2H, t,
J = 7.5 Hz, H-2), 2.79–2.90 (6H, m, H-12, 15, 18),
2.97 (2H, t, J = 6.5 Hz, H-9), 3.26 (3H, s, MeO),
3.64 (1H, q, J = 6.7 Hz, H-4), 3.67 (3H, s, CO2Me),
5.28–5.46 (9H, m, H-8, 10, 11, 13, 14, 16, 17, 19,
20), 5.50 (1H, dd, J = 15.2, 6.7 Hz, H-5), 6.01 (1H,
t, J = 10.9 Hz, H-7), 6.49 (1H, dd, J = 15.2, 10.9 Hz,
H-6); 13C NMR d 14.5, 20.8, 25.8, 25.9, 26.3, 30.2,
30.8, 51.8, 56.5, 81.3, 127.2, 127.8, 128.0 (2 carbons),
128.2, 128.6, 128.8 (2 carbons), 128.9, 130.7, 132.3,
133.6, 174.2; MS m/z 372 (M+, 1), 357 (1), 340 (11),
285 (8), 271 (9), 253 (8), 183 (40), 131 (100), 79
(75); HRMS Calcd for C23H32O2 (M+ꢀCH3OH)
340.2402, found 340.2394.
To a solution of (COCl)2 (97 lL, 1.12 mmol) in
CH2Cl2 (10 ml) was added a solution of DMSO
(158 lL, 2.24 mmol) in CH2Cl2 (1 ml) at ꢀ78 ꢁC and
the mixture was stirred at ꢀ78 ꢁC for 10 min. A solu-
tion of 6 (200 mg, 0.559 mmol) in CH2Cl2 (4 ml) was
added and the mixture was stirred at that temperature
for 10 min. Et3N (640 lL, 4.60 mmol) was added to
the reaction mixture at ꢀ78 ꢁC, and then the mixture
was allowed to warm to 0 ꢁC. The reaction was
quenched with water and the mixture was extracted
with ethyl acetate. The organic layer was washed with
water and brine, dried, and evaporated. The residue
was chromatographed on silica gel (20 g, 10% ethyl
1
acetate–hexane) to give 7 (152 mg, 76%): H NMR d
0.97 (3H, t, J = 7.5 Hz, H-22), 2.07 (2H, quint,
J = 7.5 Hz, 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), 3.10 (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-8), 6.15 (1H, t,
J = 11.4 Hz, H-7), 6.20 (1H, d, J = 15.5 Hz, H-5),
7.56 (1H, dd, J = 15.5, 11.4 Hz, H-6); 13C NMR d

104
T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
3.9. (5E,7Z,10Z,13Z,16Z,19Z)-4-Methoxy-5,7,10,13,16,19-
docosahexaenoic acid (9)
3.12. 6-[(3Z,6Z,9Z,12Z)-1-Iodo-3,6,9,12-pentadecatet-
raenyl]tetrahydro-2H-pyran-2-one (12a)
According to the procedure described for 2, the hydroly-
sis of the ester 8 was performed to give 9 in 87% yield:
1H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 1.79–1.93
(2H, m, H-3), 2.08 (2H, quint, J = 7.4 Hz, H-21), 2.45
(2H, t, J = 7.5 Hz, H-2), 2.80–2.88 (6H, m, H-12, 15,
18), 2.97 (2H, t, J = 6.6 Hz, H-9), 3.27 (3H, s, MeO),
3.64 (1H, q, J = 6.7 Hz, H-4), 5.28–5.46 (9H, m, H-8,
10, 11, 13, 14, 16, 17, 19, 20), 5.50 (1H, dd, J = 15.2,
6.7 Hz, H-5), 6.01 (1H, t, J = 10.9 Hz, H-7), 6.50 (1H,
dd, J = 15.2, 10.9 Hz, H-6); 13C NMR d 14.5, 20.8,
25.8, 25.9, 26.3, 30.3, 30.5, 56.5, 81.3, 127.2, 127.7,
127.9, 128.0, 128.1, 128.2, 128.3, 128.6, 128.8, 128.9,
130.9, 132.3, 133.2, 179.5: MS m/z 358 (M+, 3), 326
(11), 297 (5), 213 (10), 108 (100), 79 (95); HRMS Calcd
for C22H30O2 (M+ꢀCH3OH) 326.2246, found 326.2234.
Iodolactonization of eicosapentaenoic acid was per-
formed by the same procedure described for 4 to give
12a in 88% yield: H NMR d 0.98 (3H, t, J = 7.5 Hz,
H-20), 3.96 (1H, m, H-5), 4.10 (1H, m, H-6), 5.36–5.39
(7H, m, H-9, 11, 12, 14, 15, 17, 18), 5.56 (1H, m, H-8).
1
3.13. (3Z,6Z,9Z,12Z)-3,6,9,12-Pentadecatetraen-1-ol (12c)
A solution of 12a (2.17 g, 5.08 mmol) in 5% KOH/
CH3OH–H2O (19:1, 20 mL) was stirred at 60 ꢁC for
4 h. The reaction mixture was acidified with 5% aque-
ous HCl and then extracted with ethyl acetate. The
organic layer was washed with water, dried, and evap-
orated. The crude product 12b was obtained (1.56 g,
91%). A solution of 12b (1.56 g, 4.63 mmol) and sodi-
um periodate (1.46 g, 6.86 mmol) in THF/H2O (2:1,
13.7 mL) was stirred at 0 ꢁC for 1 h. The mixture was
extracted with ethyl acetate. The organic layer was
washed with water, dried, and evaporated. The residue
was solved in MeOH (13 mL) and treated with NaBH4
(518 mg, 13.7 mmol) at 0 ꢁC for 0.5 h. The reaction was
quenched with water at 0 ꢁC and then extracted with
ethyl acetate. The organic layer was washed with
water, dried, and evaporated. The residue was chro-
matographed on silica gel (40 g, 50% ethyl acetate–hex-
3.10. Methyl(5E,7Z,10Z,13Z,16Z,19Z)-4-acetyloxy-
5,7,10, 13,16,19-docosahexaenoate (10)
A solution of 6 (10 mg, 0.028 mmol) in pyridine (200 lL)
and Ac2O (50 lL) was stirred at 0 ꢁC for 6 h. The mixture
was neutralized with 10% aqueous HCl and extracted
with ethyl acetate. The organic layer was washed with
water, dried, and evaporated. The residue was chromato-
graphed on silica gel (1 g, 10 % ethyl acetate–hexane) to
give 10 (9 mg, 90%): 1H NMR d 0.97 (3H, t, J = 7.5 Hz,
H-22), 1.98 (2H, m, H-3), 2.05 (3H, s, CH3CO), 2.08
(2H, m, H-21), 2.36 (2H, t, J = 7.5 Hz, H-2), 2.79–2.90
(6H, m, H-12, 15, 18), 2.96 (2H, t, J = 6.9 Hz, H-9), 3.67
(3H, s, MeO), 5.28–5.46 (10H, m, H-4, 8, 10, 11, 13, 14,
16, 17, 19, 20), 5.57 (1H, dd, J = 15.1, 7.3 Hz, H-5), 5.96
(1H, t, J = 11.0 Hz, H-7), 6.55 (1H, dd, J = 15.1,
11.0 Hz, H-6); MS m/z 400 (M+, 1), 369 (4), 340 (66),
192 (73), 91 (100), 79 (90); HRMS Calcd for C23H32O2
(M+ꢀAcOH) 340.2402, found 340.2383.
1
ane) to give 12c (758 mg, 75%): H NMR d 0.98 (3H, t,
J = 7.5 Hz, H-15), 2.08 (2H, quint, J = 7.5 Hz, H-14),
2.36 (2H, q, J = 7.0 Hz, H-2), 2.80–2.88 (6H, m, H-5, 8,
11), 3.65 (2H, m, H-1), 5.24–5.62 (8H, m, H-3, 4, 6, 7, 9,
10, 12, 13); 13C NMR d 14.7, 21.0, 25.9, 26.0, 26.1, 31.2,
62.6, 126.0, 127.4, 128.2, 128.3, 128.8, 129.0, 131.6, 132.5.
3.14. Bromo[(3Z,6Z,9Z,12Z)-3,6,9,12-pentadecatetrae-
nyl]triphenylphosphorane (12d)
To a solution of 12c (450 mg, 2.05 mmol) in CH2Cl2
(20 mL) were added Ph3P (1.61 g, 6.14 mmol) and
CBr4 (2.05 g, 6.14 mmol) and the mixture was stirred
at room temperature for 1 h. The reaction mixture
was evaporated. The residue was chromatographed
on silica gel (15 g, 0–5% ethyl acetate–hexane) to give
1-brominated compound (510 mg, 88%). To a solution
of 1-brominated compound (430 mg, 1.52 mmol) in
CH3CN (3.0 mL) was added Ph3P (1.20 g, 4.57 mmol)
and the mixture was stirred at 80 ꢁC for 20 h in a
sealed tube. The reaction mixture was evaporated and
the residue was chromatographed on silica gel (26 g,
3.11. Methyl(5E,7Z,10Z,13Z,16Z,19Z)-4-fluoro-5,7,10,13,
16,19-docosahexaenoate (11)
To a solution of diethylaminosulfur trifluoride (DAST)
(54.8 lL, 0.418 mmol) in CH2Cl2 (0.5 mL) was added
dropwise 6 (100 mg, 0.279 mmol) in CH2Cl2 (0.5 mL)
at ꢀ78 ꢁC. The reaction mixture was stirred at that tem-
perature for 10 min and then at 0 ꢁC for 30 min. The
reaction was quenched with brine and the mixture was
extracted with ethyl acetate. The organic layer was
washed with 10% aqueous HCl, saturated solution of
aqueous NaHCO3 and brine, dried, and evaporated.
The residue was chromatographed on silica gel (10 g,
1
3–5% MeOH–CH2Cl2) to give 12d (762 mg, 92%): H
NMR d 0.96 (3H, t, J = 7.4 Hz, H-15), 2.04 (2H, quint,
J = 7.4 Hz, H-14), 2.57 (2H, m, H-2), 2.58 (2H, t,
J = 7.1 Hz, H-11), 2.69 (2H, t, J = 7.1 Hz, H-8), 2.75
(2H, t, J = 7.1 Hz, H-5), 3.91 (2H, m, H-1), 5.25–5.40
(7H, m, H-3, 4, 6, 7, 9, 10, 12), 5.64 (1H, m, H-13),
7.70–7.90 (15H, m, Ph).
1
3% ethyl acetate–hexane) to give 11 (42 mg, 42%): H
NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 2.07 (4H, m,
H-3, 21), 2.44 (2H, m, H-2), 2.79–2.90 (6H, m, H-12,
15, 18), 2.97 (2H, t, J = 6.7 Hz, H-9), 3.68 (3H, s), 5.01
(1H, dq, J = 48.6, 6.3 Hz, H-4), 5.27–5.42 (9H, m, H-
8, 10, 11, 13, 14, 16, 17, 19, 20), 5.69 (1H, m, H-5),
6.00 (1H, t, J = 10.8 Hz, H-7), 6.60 (1H, m, H-6); 19F
NMR d ꢀ175.55 (1F, ddt, J = 48.6, 21.1, 16.0 Hz); MS
m/z 360 (M+, 2), 340 (13), 271 (10), 260 (10), 197 (71),
131 (43), 79 (100); HRMS Calcd for C23H33FO2 (M+)
360.2465, found 360.2488.
3.15. Methyl(7Z,10Z,13Z,16Z,19Z)-5,7,10,13,16,19-doco-
sahexaenoate (15)
To a solution of phosphonium bromide 12d (30 mg,
55 lmol), aldehyde 13 (17.2 mg, 110 lmol), and HMPA

T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
105
(58 lL, 333 lmoL) in THF (330 lL) was added
LHMDS (1.0 M in THF, 66 lL, 66 lmol) at ꢀ78 ꢁC.
The reaction mixture was allowed to warm to 0 ꢁC for
2 h, and the reaction was quenched with water. The mix-
ture was extracted with ethyl acetate. The organic layer
was washed with water and brine, dried, and evaporat-
ed. The residue was chromatographed on silica gel
(6 g, 1% ethyl acetate–hexane) to give 15 (6.0 mg,
174.0; MS m/z 354 (M+, 15), 285 (12), 267 (11), 213
(13), 145 (41), 131 (48), 108 (100), 91 (92), 79 (95);
HRMS Calcd for C24H34O2 (M+) 354.2559, found
354.2564.
3.18. (7Z,10Z,13Z,16Z,19Z)-4-Hydroxy-7,10,13,16,19-
docosapentaenic acid (17)
1
33%): H NMR d 0.98 (3H, t, J = 7.5 Hz, H-22), 1.75
According to the procedure described for 2, the hydroly-
sis of lactone 24 was performed to give 17 in 82% yield:
1H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 1.73 (2H, m,
H-5), 1.84 (2H, m, H-3), 2.08 (2H, m, H-21), 2.20 (2H,
m, H-6), 2.53 (2H, t, J = 7.2 Hz, H-2), 2.79–2.90 (8H,
m, H-9, 12, 15, 18), 3.69 (1H, m, H-4), 5.29–5.45
(10H, m, H-7, 8, 10, 11, 13, 14, 16, 17, 19, 20); MS
m/z 346 (M+, 6), 328 (6), 259 (8), 175 (45), 119 (56), 79
(100); HRMS Calcd for C22H32O2 (M+ꢀH2O)
328.2402, found 328.2382.
(2H, quint, J = 7.4 Hz, H-3), 2.08 (2H, quint,
J = 7.5 Hz, H-21), 2.16 (2H, q, J = 7.4 Hz, H-4), 2.33
(2H, t, J = 7.4 Hz, H-2), 2.75–2.90 (6H, m, H-12, 15,
18), 2.94 (2H, m, H-9), 3.67 (3H, s, Me), 5.28–5.43
(9H, m, H-8, 10, 11, 13, 14, 16, 17, 19, 20), 5.65 (1H,
dt, J = 14.0, 7.4 Hz, H-5), 5.97 (1H, t, J = 11.0 Hz, H-
7), 6.35 (1H, dd, J = 14.0, 11.0 Hz, H-6); 13C NMR d
14.7, 21.0, 24.9, 25.9, 26.1, 26.5, 32.6, 33.8, 51.9, 126.7,
127.4, 128.3, 128.4, 128.5, 128.7, 128.8, 128.9, 129.0,
129.1, 132.5, 134.1, 174.4; MS m/z 342 (M+, 6), 313
(2), 288 (4), 273 (7), 241 (7), 227 (7), 145 (21), 131
(26), 119 (33), 108 (100), 91 (59), 79 (84); HRMS Calcd
for C23H34O2 (M+) 342.2559, found 342.2529.
3.19. Methyl(7Z,10Z,13Z,16Z,19Z)-4-hydroxy-7,10,13,
16,19-docosapentaenoate (18)
According to the procedure described for 6, methanoly-
sis of lactone 24 was performed to give 18 in 33% yield:
1H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 1.53 (2H, m),
1.73, 1.82 (each 1H, m), 2.07 (2H, m, H-21), 2.19 (2H,
m, H-6), 2.46 (2H, t, J = 7.4 Hz, H-2), 2.79–2.90 (8H,
m, H-9, 12, 15, 18), 3.64 (1H, m, H-4), 3.68 (3H, s, H-
Me), 5.29–5.44 (10H, m, H-7, 8, 10, 11, 13, 14, 16, 17,
19, 20); MS m/z 360 (M+, 2), 342 (1), 322 (6), 259 (7),
175 (40), 119 (55), 79 (100); HRMS Calcd for
C23H36O3 (M+) 360.2664, found 360.2636.
3.16. (7Z,10Z,13Z,16Z,19Z)-5,7,10,13,16,19-Docosahexae-
noic acid (14)
Hydrolysis of ester 15 was performed, according to the
procedure described for 2, to give 14 in 91% yield: H
1
NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 1.76 (2H, quint,
J = 7.4 Hz, H-3), 2.01 (2H, quint, J = 7.5 Hz, H-21),
2.17 (2H, q, J = 7.4 Hz, H-4), 2.37 (2H, t, J = 7.4 Hz,
H-2), 2.75–2.88 (6H, m, H-12, 15, 18), 2.94 (2H, t,
J = 5.9, H-9), 5.25–5.45 (9H, m, H-8, 10, 11, 13, 14,
16, 17, 19, 20), 5.64 (1H, dt, J = 15.1, 7.4 Hz, H-5),
5.98 (1H, t, J = 11.5 Hz, H-7), 6.35 (1H, dd, J = 15.1,
11.5 Hz, H-6); 13C NMR d 14.7, 21.0, 24.6, 25.9, 26.5,
32.5, 33.6, 126.9, 127.4, 128.2, 128.3, 128.4, 128.5,
128.7, 128.8, 128.9, 129.0, 132.5, 133.9, 179.3; MS m/z
328 (M+, 5), 299 (3), 274 (4), 241 (6), 219 (4), 145 (22),
131 (26), 108 (100), 91(68), 79 (91); HRMS Calcd for
C22H32O2 (M+) 328.2402, found 328.2428.
3.20. Methyl(6E,8E,10Z,13Z,16Z,19Z)-4-oxo-6,8,10,
13,16,19-docosahexaenoate (19)
To a solution of 7 (100 mg, 279 lmol) in CH2Cl2
(10 mL) was added Et3N (272 lL, 1.95 mmol) at 0 ꢁC
and the mixture was stirred at that temperature for
5 h. The reaction mixture was diluted with n-hexane,
washed with water, dried, and evaporated. The residue
was chromatographed on silica gel (6 g, 8.5% ethyl ace-
1
3.17. Methyl(7Z,10Z,13Z,16Z,19Z)-4-methylene-5,7,10,13,
16,19-docosahexaenoate (16)
tate–hexane) to give 19 (65 mg, 65%): H NMR d 0.98
(3H, t, J = 7.5 Hz, H-22), 2.08 (2H, quint, J = 7.5 Hz,
H-21), 2.59 (2H, t, J = 6.6 Hz, H-2), 2.77 (2H, t,
J = 6.6 Hz, H-3), 2.80–2.88 (4H, m, H-15, 18), 2.99
(2H, t, J = 6.3 Hz, H-12), 3.27 (2H, d, J = 13.3 Hz, H-
5), 3.67 (3H, s, Me), 5.30–5.46 (7H, m, H-11, 13, 14,
16, 17, 19, 20), 5.78 (1H, m, H-6), 6.03 (1H, t,
J = 11.0 Hz), 6.20 (2H, m), 6.48 (1H, m); 13C NMR d
14.4, 20.7, 25.7, 25.8, 26.3, 27.8, 36.8, 46.9, 51.9, 125.4,
127.1, 127.7, 127.8, 127.9, 128.8 (3 carbons), 130.6,
132.2, 132.4, 134.6, 173.3, 206.7.
To a solution of 7 (50 mg, 140 lmol) in THF (180 lL)
was added 0.5 M Tebbe reagent in toluene (168 lmol,
336 lL) at ꢀ20 ꢁC. After being stirred for 30 min at
ꢀ20 ꢁC, saturated solution of aqueous NaHCO3 was
added to the reaction mixture. The mixture was extract-
ed with ethyl acetate, washed with water, dried, and
evaporated. The residue was chromatographed on silica
gel (5 g, 3% ethyl acetate–hexane) to give 16 (7.1 mg,
1
14%): H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 2.06
(2H, quint, J = 7.5 Hz, H-21), 2.52–2.63 (4H, m, H-2,
3), 2.78–2.93 (6H, m, H-12, 15, 18), 3.00 (2H, t,
J = 6.3 Hz, H-9), 3.69 (3H, s), 5.00 (1H, s), 5.06 (1H,
s), 5.24–5.47 (9H, m, H-8, 10, 11, 13, 14, 16, 17, 19,
20), 6.05 (1H, t, J = 11.1 Hz, H-7), 6.24 (1H, d,
J = 15.5 Hz, H-5), 6.56 (1H, dd, J = 15.5, 11.1 Hz, H-
6); 13C NMR d 14.7, 21.0, 25.9, 26.0, 26.1, 26.7, 27.5,
33.3, 52.0, 116.4, 124.1, 127.4, 128.1, 128.3, 128.4,
128.8, 129.0, 129.1, 129.3, 131.0, 132.4, 134.8, 145.0,
3.21. 4-Hydroxydocosanoic acid (20)
A mixture of 2 (3 mg, 8.7 lmol), 10% Pd–C (5 mg), and
CH3OH (1 mL) was stirred under a hydrogen atmo-
sphere for 7 h. The reaction mixture was filtered and
evaporated. The residue was chromatographed on silica
gel (1 g, 10% MeOH–CH2Cl2) to give 20 (2.5 mg, 81%):
1H NMR (CD3OD) d 0.90 (3H, t, J = 6.7 Hz, H-22),
1.19–1.50 (34H, m, H-5ꢁH-21), 1.54–1.84 (2H, m,

106
T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
H-3), 2.39 (2H, m, H-2), 3.53 (1H, m, H-4); 13C NMR d
14.6, 23.9, 25.3, 27.0, 30.6, 30.9, (9 carbons), 31.6, 33.2,
33.6, 38.6, 50.9, 71.8, 178.0.
3.25. 5-[(1E,3Z,6Z,9Z,12Z,15Z)-1,3,6,9,12,15-Octadeca-
hexaenyl]tetrahydro-2-furanol (25)
To a solution of 5 (300 mg, 0.92 mmol) in THF (9.2 mL)
was added dropwise 0.95 M diisobutylaluminum hy-
dride in hexane (1.84 mmol, 1.94 mL) at ꢀ78 ꢁC. After
being stirred at ꢀ78 ꢁC for 3 h, brine was added and
the mixture was filtered through a Celite pad. The fil-
trate was extracted with ethyl acetate and the organic
layer was dried and evaporated. The residue was chro-
matographed on silica gel (15 g, 10% ethyl acetate–ben-
3.22. 4-Oxodocosanoic acid (21)
A mixture of 7 (27 mg, 0.076 mmol), 10% Pd–C (27 mg),
and CH3OH (1 mL) was stirred under a hydrogen atmo-
sphere for 7 h. The mixture was filtered and evaporated.
The residue was chromatographed on silica gel (20 g,
2.5% ethyl acetate–hexane) to give methyl ester of 21
1
1
(27 mg, 97%): H NMR d 0.87 (3H, t, J = 6.8 Hz, H-
zene) to give 25 (146 mg, 48%): H NMR d 0.98 (3H, t,
22), 1.18–1.32 (30H, m, H-7ꢁH-21), 1.57 (2H, m, H-6),
2.43 (2H, t, J = 7.5 Hz, H-5), 2.58 (2H, t, J = 6.7 Hz,
H-2), 2.71 (2H, t, J = 6.7 Hz, H-3), 3.67 (3H, s, MeO);
13C NMR d 14.6, 23.1, 24.3, 28.1, 29.6, 29.8, 29.9, 30.1
(10 carbons), 32.4, 37.4, 43.3, 53.2, 173.7, 209.6. Accord-
ing to the procedure described for 2, the hydrolysis of the
J = 7.5 Hz, H-22), 1.57–2.28 (6H, m, H-2, 3, 21), 2.78–
2.89 (6H, m, H-12, 15, 18), 2.96 (2H, t, J = 6.6 Hz, H-
9), 3.05 (1H, br s, OH), 4.50 and 4.73 (total 1H, each
q, J = 7.9 Hz, H-4), 5.28–5.43 (9H, m, H-8, 10, 11, 13,
14, 16, 17, 19, 20), 5.51 and 5.61 (total 1H, m, H-1),
5.62 and 5.77 (total 1H, dd, J=15.2, 7.8 Hz), 5.98 (1H,
m), 6.53 (1H, m); MS m/z 328 (M+, 1), 310 (4), 241
(3), 213 (8), 145 (21), 79 (100); HRMS Calcd for
C22H30O (M+ꢀH2O) 310.2297, found 310.2303.
1
methyl ester was performed to give 21 in 76% yield: H
NMR d 0.88 (3H, t, J = 6.8 Hz, H-22), 1.19–1.32 (30H,
m, H-7ꢁH-21), 1.58 (2H, m, H-6), 2.44 (2H, t,
J = 7.5 Hz, H-5), 2.63 (2H, m, H-2), 2.73 (2H, m, H-3);
13C NMR d 14.5, 23.1, 24.2, 27.9, 29.6, 29.8, 29.9, 30.0
(10 carbons), 32.3, 37.2, 43.2, 173.5, 209.9.
3.26. (5E,7Z,10Z,13Z,16Z,19Z)-5,7,10,13,16,19-Docosa-
hexaene-1,4-diol (26)
3.23. 5-[(3Z,6Z,9Z,12Z,15Z)-1-Hydroxy-3,6,9,12,15-
octadecapentaenyl]dihydro-2(3H)-furanone (23)
To a solution of 5 (300 mg, 0.92 mmol) in THF (9 mL)
was added 0.95 M diisobutylaluminum hydride in hex-
ane (2.76 mmol, 2.90 mL) at 0 ꢁC and the mixture was
stirred at 0 ꢁC for 3 h. The reaction was worked up
and purified, by the same procedure described for 25,
to give 26 (182 mg, 60%): 1H NMR d 0.98 (3H, t,
J = 7.5 Hz, H-22), 1.54–1.73 (4H, m, H-2, 3), 2.07 (2H,
m, H-21), 2.84 (6H, m, H-12, 15, 18), 2.97 (2H, t,
J = 6.5 Hz, H-9), 3.68 (2H, m, H-1), 4.24 (1H, m, H-4),
5.28–5.45 (9H, m, H-8, 10, 11, 13, 14, 16, 17, 19, 20),
5.72 (1H, dd, J = 15.2, 6.7 Hz, H-5), 6.00 (1H, t,
J = 11.0 Hz, H-7), 6.52 (1H, dd, J = 15.2, 11.0 Hz, H-6)
A solution of 4 (86 mg, 0.189 mmol) in 0.2 M LiOH/
THF–H2O (3:2, 5 mL) was stirred at 0 ꢁC for 5 h. After
the addition of 10% aqueous HCl, the mixture was
extracted with ethyl acetate, washed with water, and
evaporated. The residue was dissolved in chloroform
(10 mL) and the solution was stored at rt for 24 h. The
solution was evaporated and chromatographed on silica
gel (5 g, 30 % ethyl acetate–hexane) to give 23 (54 mg,
1
83%): H NMR d 0.97 (3H, t, J = 7.5 Hz, H-22), 2.08
(2H, m, H-21), 2.21 (2H, m, H-3), 2.40 (2H, m, H-2),
2.53 and 2.60 (each 1H, m, H-6), 2.79–2.90 (8H, m, H-
9, 12, 15, 18), 3.63 (1H, m, H-5), 4.47 (1H, t,
J = 7.2 Hz, H-4), 5.33–5.44 (10H, m, H-7, 8, 10, 11,
13, 14, 16, 17, 19, 20); 13C NMR d 14.5, 20.8, 24.3,
25.7 (2 carbons), 25.9, 26.0, 28.8, 31.6, 73.4, 82.2,
124.6, 127.3, 127.8, 128.1, 128.2 (2 carbons), 128.6,
128.8, 131.9, 132.3, 177.6; MS m/z 344 (M+, 12), 288
(2), 276 (10), 175 (69), 119 (51), 79 (100); HRMS Calcd
for C22H32O3 (M+) 344.2351, found 344.2365.
;
13C NMR d 14.5, 20.8, 25.8, 25.9, 26.0, 26.3, 29.0, 34.5,
63.1, 72.8, 125.7, 127.2, 127.8, 128.1, 128.2, 128.3, 128.6,
128.8, 130.6 (2 carbons), 132.3, 136.4; MS m/z 330 (M+,
1), 312 (6), 243 (8), 159 (15), 108 (70), 79 (100); HRMS
CalcdforC22H32O (M+ꢀH2O)312.2453, found312.2471.
3.27. Methyl(4S,7Z,10Z,13Z,16Z,19Z)-4-{[({[(1R,4aS)-7-
isopropyl-1,4a-dimethyl-1,2,3,4,4a,9,10,10a-octahydro-1-
phenanthrenyl]methyl}amino)carbonyl]oxy}-5,7,10,13,16,19-
docosahexaenoate (29) and methyl(4R,7Z,10Z,13Z,16Z,
19Z)-4-{[({[(1R,4aS)-7-isopropyl-1,4a-dimethyl-1,2,3,4,4a,
9,10,10a-octahydro-1-phenanthrenyl]methyl}amino)car-
bonyl]oxy}-5,7,10,13,16,19-docosahexaenoate (30)
3.24. 5-[(3Z,6Z,9Z,12Z,15Z)-3,6,9,12,15-Octadecapen-
taenyl]dihydro-2(3H)-furanone (24)
A solution of 4 (436 mg, 0.960 mmol) and Bu3SnH
(517 lL, 1.92 mmol) in THF (44 mL) was refluxed for
5 h. The mixture was evaporated and the residue was
chromatographed on silica gel (20 g, 2.5% ethyl ace-
tate–hexane) to give 24 (207 mg, 66%): 1H NMR d
0.97 (3H, t, J = 7.5 Hz, H-22), 1.66 (1H, m), 1.87 (2H,
m), 2.08 (2H, m, H-21), 2.22 (2H, m), 2.33 (1H, m),
2.54 (2H, m, H-2), 2.79–2.90 (8H, m, H-9, 12, 15, 18),
4.50 (1H, m, H-4), 5.32–5.41 (10H, m, H-7, 8, 10, 11,
13, 14, 16, 17, 19, 20); MS m/z 328 (M+, 14), 228 (10),
175 (63), 79 (100); HRMS Calcd for C22H32O2 (M+)
328.2402, found 328.2408.
A mixture of 6 (500 mg, 1.40 mmol), isocyanate 28
(1.31 g, 4.20 mmol), DMAP (342 mg, 2.80 mmol),
and CH2Cl2 (4.2 mL) in a sealed tube was stirred at
50 ꢁC for 20 h under nitrogen. Removal of solvent
in vacuo and chromatography (10% AcOEt/hexane)
on silica gel afforded urethane (29:30 = 1:1) (850 mg,
91%) along with recovered 28 (451 mg). The diastereo-
meric carbamates were separated by chromatography
on silica gel (5.0, 5.5, 6.0, 6.5, 7.0, 7.5, and 8.0%
AcOEt/hexane) to afford 30 (261 mg, 28%) and 29
(243 mg, 26%) in this order. 30:1H NMR d 0.92
(3H, s, Abieta-18), 0.97 (3H, t, J = 7.5 Hz, DHA-22),

T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
107
1.21 (3H, s, Abieta-20), 1.22 (6H, d, J = 7.5 Hz, Abi-
eta-16, 17), 1.25–1.89 (8H, m, Abieta-1, 2, 3, 6),
1.95 (2H, q, J = 7.3 Hz, DHA-3), 2.07 (2H, m,
DHA-21), 2.29 (1H, d, J = 12.9 Hz, Abieta-5), 2.34
(2H, t, J = 7.3 Hz, DHA-2), 2.74–2.96 (11H, m,
DHA-9, 12, 15, 18, Abieta-7, 15), 3.05, 3.10 (each
1H, dd, J = 13.8, 7.0 Hz, Abieta-19), 3.63 (3H, s,
CO2Me), 4.65 (1H, t, J = 6.4 Hz, NH), 5.20 (1H, q,
J = 7.3 Hz, DHA-4), 5.29–5.44 (9H, m, DHA-8, 10,
11, 13, 14, 16, 17, 19, 20), 5.56 (1H, dd, J = 15.2,
7.3 Hz, DHA-5), 5.95 (1H, t, J = 10.8 Hz, DHA-7),
6.52 (1H, dd, J = 15.2, 10.8 Hz, DHA-6), 6.89 (1H,
s, Abieta-14), 6.99 (1H, d, J = 8.0 Hz, Abieta-12),
7.19 (1H, d, J = 8.2 Hz, Abieta-11 ); 13C NMR d
14.5, 18.7, 18.8, 19.1, 20.8, 24.1 (2 carbons), 25.5,
25.8, 25.9, 26.3, 30.0, 30.1, 30.4, 33.7, 36.2, 37.6,
37.6, 38.6, 45.5, 51.8, 51.9, 74.2, 124.1, 124.4, 127.1,
127.2, 128.0, 128.1, 128.2, 128.6, 128.8, 128.9, 131.4,
(R)-methoxy(trifluoromethyl)phenylacetyl
chloride
(21 lL, 112 lmol) at 0 ꢁC. The solution was stirred at
room temperature for 20 min. The reaction mixture
was poured into ice and water, and extracted with ether.
The organic layer was washed with water, dried, and
evaporated. The residue was chromatographed on silica
gel (4 g) with 20% AcOEt-hexane to give S-MTPA ester
1
31 (28.5 mg, 89 %). H NMR d 0.97 (3H, t, J = 7.5 Hz,
H-22), 1.98–2.12 (4H, m, H-3, 21), 2.37 (2H, H-2), 2.78–
2.87 (6H, m, H-12, 15, 18), 2.91 (2H, t, J = 7.3 Hz, H-9),
3.53 (3H, br s), 3.68 (3H, s, CO2Me), 5.30–5.60 (11H, m,
H-4, 5, 6, 10, 11, 13, 14, 16, 17, 19, 20), 5.94 (1H, t,
J = 11.0 Hz, H-8), 6.59 (1H, dd, J = 14.3, 11.0 Hz, H-
7), 7.34–7.43 (3H, m), 7.45–7.56 (2H, m).
3.29. Methyl(4S,7Z,10Z,13Z,16Z,19Z)-4-{[(2S)-3,3,3-
trifluoro-2-methoxy-2-phenylpropyl]oxy}-5,7,10,13,16,
19-docosahexaenoate (32)
1
131.5, 132.3, 134.9, 145.9, 147.3, 156.3, 173.8. 29: H
NMR
d
0.92 (3H, s, Abieta-18), 0.97 (3H, t,
Compound (ꢀ)-6b (10 mg, 28 lmol) was treated with
J = 7.5 Hz, DHA-22), 1.21 (3H, s, Abieta-20), 1.22
(6H, d, J = 7.5 Hz, Abieta-16, 17), 1.25–1.77 (8H, m,
Abieta-1, 2, 3, 6), 1.95 (2H, dd, J = 14.4, 7.3 Hz,
DHA-3), 2.08 (2H, m, DHA-21), 2.25 (1H, d,
J = 12.9 Hz, Abieta-5), 2.35 (2H, dd, J = 7.5, 5.9 Hz,
DHA-2), 2.77–2.95 (11H, m, DHA-9, 12, 15, 18, Abi-
eta-7, 15), 3.00, 3.14 (each 1H, dd, J = 13.8, 6.5 Hz,
Abieta-19), 3.64 (3H, s, CO2Me), 4.66 (1H, t,
J = 6.4 Hz, NH), 5.18 (1H, q, J = 6.7 Hz, DHA-4),
5.29–5.44 (9H, m, DHA-8, 10, 11, 13, 14, 16, 17,
19, 20), 5.56 (1H, dd, J = 15.2, 6.7 Hz, DHA-5), 5.93
(1H, t, J = 10.8 Hz, DHA-7), 6.54 (1H, dd, J = 15.2,
10.8 Hz, DHA-6), 6.87 (1H, s, Abieta-14), 6.95 (1H,
d, J = 8.0 Hz, Abieta-12), 7.15 (1H, d, J = 8.0 Hz, Abi-
eta-11 ); 13C NMR d 14.7, 19.0 (2 carbons), 19.2, 21.0,
24.4 (2 carbons), 25.7, 25.9, 26.0, 26.1, 26.5, 30.2,
30.3, 30.6, 33.8, 36.4, 37.8, 38.8, 45.5, 52.0, 52.1,
74.4, 124.2, 124.6, 127.3, 127.4, 128.0, 128.1, 128.2,
128.3 (2 carbons), 128.4, 128.8, 129.0, 129.1,
131.5,131.6, 132.5, 135.2, 146.0, 147.5, 156.5, 173.8.
(S)-methoxy(trifluoromethyl)phenylacetyl chloride
according to the procedure described above to give R-
1
MTPA ester 32 (13 mg, 81%). H NMR d 0.97 (3H, t,
J = 7.5 Hz, H-22), 1.98–2.12 (4H, m, H-3, 21), 2.25
(2H, H-2), 2.78–2.87 (6H, m, H-12, 15, 18), 2.94 (2H,
t, J = 7.3 Hz, H-9), 3.54 (3H, br s), 3.66 (3H, s, CO2Me),
5.30–5.60 (11H, m, H-4, 5, 6, 10, 11, 13, 14, 16, 17, 19,
20), 5.98 (1H, t, J = 11.3 Hz, H-8), 6.67 (1H, dd, J = 14.1,
11.3 Hz, H-7), 7.34–7.43 (3H, m), 7.45–7.56 (2H, m).
(+)-(4S)-2a and (ꢀ)-(4R)-2b. Basic hydrolysis of (+)-
20
(4S)-6a and (ꢀ)-(4R)-6b afforded (+)-(4S)-2a and (ꢀ)-
20
(4R)-2b, respectively. (4S)-2a: ½aꢂ +5.9 (c 0.7, CHCl3).
D
(4R)-2b: ½aꢂ ꢀ5.9 (c 0.8, CHCl3).
D
3.30. Transfection and transactivation assay
COS-7 cells were cultured in DulbeccoÕs modified Ea-
gleÕs medium (DMEM) supplemented with 5% fetal calf
serum (FCS). Cells were seeded on 24-well plates at a
density of 2 · 104 per well. After 24 h, the cells were
transfected with a reporter plasmid containing four cop-
ies of MH100 GAL4 binding site (MH100 · 4-TK-
Luc),17 GAL4-hPPARc chimera expression plasmid
(pSG5-GAL-hPPARc),3 and the internal control plas-
mid containing sea pansy luciferase expression con-
structs (pRL-CMV) by the lipofection method as
described previously.18 After 4 h-incubation, the medi-
um was replaced with fresh DMEM containing 5% char-
coal-treated FCS (HyClone, UT, USA). The next day,
the cells were treated either with the ligand (final concen-
tration, 5 lM) or ethanol vehicle and cultured for 24 h.
Cells in each well were harvested with a cell lysis buffer,
and the luciferase activity was measured with a lucifer-
ase assay kit (Toyo Ink, Inc., Japan). Transactivation
measured by the luciferase activity was normalized with
the internal control. All experiments were done in
triplicate.
(+)-6a. To a solution of carbamate 29 (115 mg,
172 lmol) and triethylamine (120 lL, 0.859 mmol) in
benzene (1.7 mL) was added trichlorosilane (43 lL,
0.430 mmol) in benzene (0.86 mL) and the mixture was
stirred for 2.5 h. The mixture was treated with a small
amount of water (ꢁ0.2 mL) and subjected to rough
chromatography on silica gel (2 g) with AcOEt followed
by chromatography on silica gel (12–20% AcOEt/hex-
ane) to afford isomer with (+) optical density of 6
20
D
(37.5 mg, 61%). ½aꢂ +7.0 (c 2, CHCl3).
(ꢀ)-6b. Carbamate 30 (111 mg) was treated by the same
20
procedure as described above to afford (ꢀ)-isomer 6
(36 mg). ½aꢂ ꢀ7.0 (c 2, CHCl3).
D
3.28. Methyl(4R,7Z,10Z,13Z,16Z,19Z)-4-{[(2S)-3,3,3-tri-
fluoro-2-methoxy-2-phenylpropyl]oxy}-5,7,10,13,16,19-
docosahexaenoate (31)
To a solution of alcohol (ꢀ)-6b (20 mg, 56 lmol) in
dichloromethane (1.0 mL) were added triethylamine
(38 lL, 273 lmol), DMAP (38 mg, 0.31 mmol), and
References and notes
1. Stumvoll, M.; Haring, H. U. Ann. Med. 2002, 34, 217–224.

108
T. Itoh et al. / Bioorg. Med. Chem. 14 (2006) 98–108
2. Zanchi, A.; Chiolero, A.; Maillard, M.; Nussberger, J.;
Brunner, H. R.; Burnier, M. J. Clin. Endocrinol. Metab.
2004, 89, 1140–1145.
3. 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.
10. 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.
11. Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113,
7277–7287.
4. Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B.
R. J. Med. Chem. 2000, 43, 527–550.
12. Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58,
2899.
5. 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.
6. Yamamoto, K.; Itoh, T.; Abe, D.; Shimizu, M.; Kanda,
T.; Koyama, T.; Nishikawa, M.; Tamai, T.; Ooizumi, H.;
Yamada, S. Bioorg. Med. Chem. Lett. 2005, 15, 517–522.
7. Neuringer, M.; Anderson, G. J.; Connor, W. E. Annu.
Rev. Nutr. 1988, 8, 517–541.
13. Middleton, W. J. J. Org. Chem. 1975, 40, 574–579.
14. Tebbe, F. N.; Parshall, G. W.; Reddy, G. S. J. Am. Chem.
Soc. 1978, 100, 3611–3613.
15. Corey, E. J.; Hashimoto, S. Tetrahedron Lett. 1981, 22,
299–302.
16. Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J.
Am. Chem. Soc. 1991, 113, 4092–4096.
17. Kang, T.; Martins, T.; Sadowski, I. J. Biol. Chem. 1993,
268, 9629–9635.
8. Storlien, L. H.; Hulbert, A. J.; Else, P. L. Curr. Opin. Clin.
Nutr. Metab. Care 1998, 1, 559–563.
9. Rose, D. P.; Connolly, J. M. Pharmacol. Ther. 1999, 83,
217–244.
18. Yamamoto, K.; Masuno, H.; Choi, M.; Nakashima, K.;
Taga, T.; Ooizumi, H.; Umesono, K.; Sicinska, W.;
VanHooke, J.; DeLuca, H. F.; Yamada, S. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 1467–1472.