Synthesis of digalactosyl diacylglycerols and their structure-inhibitory activity on human lanosterol synthase

Article abstract of DOI:10.1016/j.bmcl.2004.10.013

Digalactosyl and monogalactocyl diacylglycerols, were synthesized. DGDG with two myristoyl groups at both sn-1 and sn-2 positions showed the most potent inhibitory activity on human lanosterol synthase. Digalactosyl and monogalactocyl diacylglycerols (DGDG and MGDG), which were identified as anti-hyperlipemia active components in Colocasia esculenta (Taro), were synthesized. The inhibitory activity of DGDG, MGDG and related compounds on human lanosterol synthase was evaluated as anti-hyperlipemic activity. DGDG with two myristoyl groups at both sn-1 and sn-2 positions and with an oleoyl group at the sn-1 position showed the most potent activity.

Full text of DOI:10.1016/j.bmcl.2004.10.013

Bioorganic & Medicinal Chemistry Letters 15 (2005) 159–162
Synthesis of digalactosyl diacylglycerols and their
structure–inhibitory activity on human lanosterol synthase
Rie Tanaka,a Yuichi Sakano,b Akito Nagatsu,c Masaaki Shibuya,b
Yutaka Ebizukab and Yukihiro Godaa,*
aNational Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-Ku, Tokyo 158-8501, Japan
bGraduate School of Pharmaceutical Sciences, the University of Tokyo, 7-3-1 Hongo, Bunkyo-Ku, Tokyo 113-0033, Japan
cGraduate School of Pharmaceutical Sciences, Nagoya City University, 3-1 Tanabe-dori, Mizuho-Ku, Nagoya 467-8603, Japan
Received 10 August 2004; revised 5 October 2004; accepted 6 October 2004
Available online 22 October 2004
Abstract—Digalactosyl and monogalactocyl diacylglycerols (DGDG and MGDG), which were identified as anti-hyperlipemia
active components in Colocasia esculenta (Taro), were synthesized. The inhibitory activity of DGDG, MGDG and related com-
pounds on human lanosterol synthase was evaluated as anti-hyperlipemic activity. DGDG with two myristoyl groups at both
sn-1 and sn-2 positions and with an oleoyl group at the sn-1 position showed the most potent activity.
Ó 2004 Elsevier Ltd. All rights reserved.
Lanosterol synthase is located in the middle stage of the
biosynthetic pathway of the cholesterol of mammals,
and is a more selective target for suppression of chole-
sterol biosynthesis. We established an in vitro assay sys-
tem of lanosterol synthase using recombinant human
enzyme and [14C] labeled oxydosqualene, to search for
the anti-hyperlipemia active component.1 In the screen-
ing study the ethanol extract of Colocasia esculenta
(Taro) exhibited the most potent activity among 60
kinds of vegetable extracts. Therefore, we investigated
the active principal from Taro and isolated monogalacto-
syl diacylglycerol (MGDG, 1) and digalactosyl diacyl-
glycerol (DGDG, 2) as the active components, after
repeated chromatography.2 MGDG and DGDG, a class
of glycolipids, are the major constituents of the chloro-
plast membrane in the plant kingdom and have at-
tracted much attention in recent years because of their
biological activities, such as anti-tumor-promoting
activity,3a DNA polymerase inhibition,3b activity of vio-
laxanthin de-epoxidase in liposomes3c and hemolytic
activity.3d It was reported that MGDG and DGDG
with various acyl pairs had been isolated from natural
resources. The variations of acyl groups in MGDG
and DGDG were supposed to reflect the strength of
inhibitory activity against human lanosterol synthase
from the viewpoint of structure–activity relationship.
However, it is difficult isolate MGDG or DGDG with
different acyl groups from Taro, because they have al-
most the same polarity and size with each other. Thus,
we attempted to synthesize selectively MGDG and
DGDG with different acyl groups. In this paper, we re-
port the first synthesis of DGDG with the desired acyl
groups at the desired positions and elucidation of their
inhibitory activity on human lanosterol synthase (Fig. 1).
MGDG was prepared in the same manner as reported
before.8 The synthetic route to DGDG (2) is shown in
CH OH
2
HO
O
OH
CH OH
2
O
CH
OH
HO
O
CH
2
2
HO
O
2
H
O-COR
OH
O
CH
2
HO
O
2
1
H
O-COR
H C O-COR
2
1
OH
H C O-COR
2
OH
MGDG (1)
DGDG ( 2)
Keywords: Colocasia esculenta; Glycolipid; Digalactosyl diacylglyce-
rols; Lanosterol synthase; Cholesterol biosynthesis.
1
2
R ,R =acyl residue
Figure 1. Chemical structures of MGDG and DGDG.
*
Corresponding author. Fax: +81 3 3707 6950/3700 9165; e-mail:
goda@nihs.go.jp
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.10.013

160
R. Tanaka et al. / Bioorg. Med. Chem. Lett. 15 (2005) 159–162
CH OBn
CH OH
2
2
BnO
O
O
O
OBn
O
F
+
O
a
O
BnO
3
4
CH OAc
CH OBn
2
2
AcO
BnO
O
O
OAc
OBn
CH OH
C OBn
b, c, d, e
2
O
AcO
CH
O
O
CH
O
2
2
AcO
BnO
H
O
O
+
O
CH OBn
2
OAc
Br
OAc
O
O
7
6
5
f
CH OH
CH OAc
2
2
HO
AcO
O
O
g, i
OH
OAc
O
CH
O
CH
2
2
HO
AcO
CH
O
CH
HO
AcO
2
2
O
O
H
OH
H
OBn
OH
OAc
H C OH
2
H C OBn
2
8
16
OH
OAc
g, h
CH OR
CH OAc
2
2
AcO
RO
O
O
OR
OAc
l
i, j, k
CH
H
2
O
CH
O
CH
2
2
AcO
RO
OY
O
CH
RO
O
O
CH
AcO
2
2
OH
O
H C OX
2
H
H
O
OR
OAc
H C O
H C OH
2
11 : X=TBDPS, Y=H
2
m
n
OAc
OR
12 : X=TBDPS, Y=THP
13 : X=H, Y=THP
9
10 : R=MPM
o
1
2
1
(COR =COR )
14 : X=COR , Y=THP
o
p, o
CH OR
2
CH OH
2
RO
HO
O
O
q
OR
OH
O
CH
2
O
CH
2
RO
HO
O
CH
RO
2
O
CH
2
HO
O
O
2
2
H
O-COR
OR
H
O-COR
OH
1
1
H C O-COR
2
H C O-COR
2
OR
OH
DGDG (2)
13 : R=MPM
Scheme 1. Reagents and conditions: (a) SnCl2, AgClO4, MS4A, Et2O, À30°C, 48h; (b) H2, 10% Pd–C, EtOH, rt, 22h; (c) Dowex 50WX8–100, H2O,
60°C, 12h; (d) Ac2O, Py, rt, 5h; (e) 30% HBr–AcOH, CH2Cl2, À10°C, 0.5h; (f) HgO, HgBr2, ClCH2CH2Cl, rt, 4h; (g) H2, 10% Pd–C, EtOH, rt, 3h;
(h) Me2C(OMe)2, p-TsOH, DMF, rt, 2h; (i) NaOMe, MeOH, rt, 10min; (j) MPM-Cl, NaH, DMF, rt, 17h; (k) p-TsOH, MeOH, rt, 2h; (m) TBDPS-
Cl, pyridine, rt, 3h; (m) DHP, PPTS, CH2Cl2, rt, 6.5h; (n) TBAF, THF, rt, 4h; (o) fatty acid, DMAP, EDCI–HCl, CH2Cl2, rt, 4h; (p) PPTS, MeOH,
rt, 12h; (q) CAN, CH3CN–H2O, rt, 45min.
Scheme 1. We selected linolate, linolenate, oleic, and
palmitate as introduction acyl groups, because they are
components of DGDG isolated from Taro,2 and myris-
trate, that is, a major acyl component of DGDG iso-
lated from microalgae.4 Compound 35 as a mixture of
a- and b-anomers was reacted with 4 in the presence
The benzyl groups and isopropylidene acetals were
cleaved followed by protection of the resulting hydroxyl
groups as acetate to afford 6. Dibenzyl glycerol (7)7
derived from D-mannitol was reacted with 6 in the
presence of HgO and HgBr2 to give 8 (64%) as the b-
anomer.
6
of SnCl2 and AgClO4 to give 5 as a 10:9 mixture of
a- and b-anomers. The mixture was separated by silica
gel chromatography to obtain the a-anomer (44%).
The route to DGDG from 8 was carried out in essen-
tially similar way to the synthesis of MGDG previously

R. Tanaka et al. / Bioorg. Med. Chem. Lett. 15 (2005) 159–162
161
reported.8 After the benzyl groups on the glycerol moi-
ety of 8 were converted to isopropylidene acetal groups,
the acetyl groups at the sugar moiety were cleaved fol-
lowed by protection of the resulting hydroxyl groups
as MPM ethers. Then, the isopropylidene acetal groups
were removed to afford 10. The hydroxyl groups at sn-1
and -2 were protected individually as t-butyldiphenylsi-
lyl (TBDPS) ether and tetrahydropyranyl (THP) ether,
respectively, to give 12 (98%). The TBDPS ether was
cleaved to 13 (97%), then acyl groups were incorporated
at the sn-1 position to give the corresponding 14 (79–
98%). After removal at the THP ether of 14, the corre-
sponding acyl groups were introduced at the sn-2 posi-
tion to afford 15 (81–98%). Finally, the MPM ethers
of 15 were removed by treatment with ceric ammonium
nitrate (CAN), and DGDG with different acyl groups
(2c–j, 2l, 2n–o) were obtained (11% from 3 for 18 steps).9
others except for that with two myristoyl groups at both
sn-1 and sn-2 positions. Both dioleoyl and dimyristoyl
DGDG showed potent inhibitory effect, probably be-
cause the distance between the terminal methyl and car-
bonyl groups of the oleoyl group, which is bent double
at cis-olefine, is close to that of the linear myristoyl
group. Digalactosyl glycerol (16) showed no inhibition
against human lanosterol synthase. From these data, it
was revealed acyl groups in the digalactosyl glycerol
were essential for the inhibitory activity, and the whole
molecular conformation of DGDG seemed to affect
the inhibitory activity on human lanosterol synthase.
DGDG with myristoyl group(s) were not reported as the
component of higher plants, but were isolated from
microalgae such as Chlorella and Spirulina species,
which are commercially available health food. DGDG
with any acyl groups as well as MGDG are known as
the component of not only Taro but also of most daily
ingestive vegetables and cereals such as spinach, wheat,
runner bean, cucumber, etc.11 Our result suggests that
the taking these vegetables should be effective and prac-
tical for avoiding hypercholesterolemia, because even
DGDG with acyl groups other than a myristoyl group
indicated moderate activity. As there were few reports
about the acyl moiety of DGDG in each vegetable,11 it
may be important to know the acyl composition of
DGDG in the vegetables for evaluating the efficacy of
them against hypercholesterolemia.
DGDG with the same acyl groups at the sn-1 and -2
positions (2a,b,k,m) were obtained by acylation of 10
followed by cleavage of the MPM ethers. Digalactosyl
glycerol (16) was obtained by removal of the protective
groups of 8. In total, we prepared 25 kinds of DGDG
(2) and digalactosyl glycerol (16).
Inhibitory activity on human lanosterol synthase of
MGDG and DGDG was determined by using the same
method as described before.2 The result is shown in Ta-
ble 1. MGDG showed 23% inhibition (ca. 0.4lmol/as-
say) or less, while most DGDG showed higher activity
than MGDG. DGDG with two myristoyl groups (2a)
showed the most potent inhibitory activity, 72% at the
concentration of 0.5l mol/assay. The dioleoyl derivative
(2b) also potent and showed 57% inhibition at the same
concentration. DGDG with an oleoyl group at sn-1
(2c,f,i) tend to inhibit enzyme more effectively than the
In summary, we have established the first sufficient syn-
thetic route to DGDG (2) with the desired acyl groups
at the desired positions, and obtained 25 kinds of
DGDG. As the result of the evaluation of their inhibi-
tory activity on human lanosterol synthase, we clarified
that the myristoyl group contributed the most to the
Table 1. Inhibitory activity of synthetic MGDG (1p–r)a, DGDG (2a–o)b and digalactosyl glycerol (16) on Human Lanosterol Synthase
Compounds
Acyl group
Inhibition (%)
sn-1 (COR1)
sn-2 (COR2)
ca. 0.4lmol/assay
1p
1q
1r
Myristoyl
Linoleoyl
Palmitoyl
Linoleoyl
Oleoyl
Linoleoyl
23
19
16
0.25lmol/assay
0.5lmol/assay
2a
2b
2c
2d
2e
2f
Myristoyl
Oleoyl
Oleoyl
Myristoyl (14:0)
Oleoyl (18:1)
Palmitoyl (16:0)
Linolenoyl (18:3)
Palmitoyl
51
42
24
18
35
9
72
57
46
45
45
40
38
31
29
27
26
26
24
21
20
n.d.
Linoleoyl
Myristoyl
Oleoyl
Linoleoyl (18:2)
Myristoyl
Oleoyl
2g
2h
2i
Linoleoyl
Linoleoyl
Oleoyl
22
8
Linolenoyl
Palmitoyl
Linoleoyl
7
2j
Linoleoyl
Linoleoyl
Linolenoyl
Palmitoyl
Palmitoyl
Palmitoyl
H
3
2k
2l
n.d.
5
Palmitoyl
Palmitoyl
2m
2n
2o
16
4
Oleoyl
Linoleoyl
4
n.d.
n.d.
H
a MGDG showimg more than 15% inhibition (ca. 0.4lmol/assay) are shown. n.d.: not detected.
b DGDG showing more than 20% inhibition (0.5lmol/assay) are shown.10

162
R. Tanaka et al. / Bioorg. Med. Chem. Lett. 15 (2005) 159–162
Parrish, C. C.; Bodennec, G.; Gentien, P. Phytochemistry
1998, 47, 783.
4. Murakami, N.; Morimoto, T.; Imamura, H.; Ueda, T.;
Nagai, S.; Sakakibara, J.; Yamada, N. Chem. Pharm. Bull.
1991, 39, 2277.
activity of DGDG among various acyl groups we tested,
and a dimyristoyl derivative (2a) and DGDG with an
oleoyl group at sn-1 (2b,c,f,i) indicated the strong
activity.
5. (a) Ferrier, R. J.; Furneaux, R. H. Methods Carbohydr.
Chem. 1980, 8, 251; (b) Nicolaou, K. C.; Dolle, R. E.;
Papahatjis, D. P.; Randall, J. L. J. Am. Chem. Soc. 1984,
106, 4189.
6. Mukaiyama, T.; Murai, Y.; Shoda, S. Chem. Lett. 1981,
43.
Acknowledgements
This work was supported by a grant from the Japan
Health Sciences Foundation.
7. (a) Mannock, D. A.; Lewis, R. N. A. H.; McElhaney, R.
N. Chem. Phys. Lipids 1987, 43, 113; (b) Takano, S.;
Ogasawara, K. Yuki Gosei Kagaku Kyokaishi 1987, 45,
1987; (c) van Boeckel, C. A. A.; Visser, G. M.; van Boom,
J. H. Tetrahedron 1985, 41, 4557.
8. Nagatsu, A.; Watanabe, M.; Ikemoto, K.; Hashimoto, M.;
Murakami, N.; Sakakibara, J.; Tokuda, H.; Nishino, H.;
Iwashima, A.; Yazawa, K. Bioorg. Med. Chem. Lett. 1994,
4, 1619.
References and notes
1. (a) Sakano, Y.; Shibuya, M.; Matsumoto, A.; Takahashi,
ˆ
Y.; Tomoda, H.; Omura, S.; Ebizuka, Y. J. Antibiot. 2003,
56, 817; (b) Sakano, Y.; Shibuya, M.; Yamaguchi, R.;
ˆ
Tomoda, H.; Omura, S.; Ebizuka, Y. Antibiot. 2004, 57,
564.
9. All compounds were characterized by NMR and MS
spectra.
2. Sakano, Y.; Mutsuga, M.; Tanaka, R.; Suganuma, H.;
Inakuma, T.; Toyoda, M.; Goda, Y.; Shibuya, M.;
Ebizuka, Y. Biol. Pharm. Bull., submitted.
10. The following DGDGs were also synthesized and assayed,
but inactive (<20% inibition at 0.5lmol/assay): (sn-1
(COR1), sn-2 (COR2)) = (myristoyl, oleoyl), (myristoyl,
linoleoyl), (myristoyl, linoleoyl), (oleoyl, myristoyl), (lino-
lenoyl, myristoyl), (linolenoyl, oleoyl), (linolenoyl, linole-
noyl), (palmitoyl, myristoyl), and (palmitoyl, linolenoyl).
11. (a) Allen, C. F.; Good, P.; Davis, H. F.; Fowler, S. G.
Biochem. Biophys. Res. Commun. 1964, 15, 424; (b)
Sugawara, T.; Miyazawa, T. Lipids 1999, 34, 1231.
3. (a) Shirahashi, H.; Murakami, N.; Watanabe, M.; Naga-
tsu, A.; Sakakibara, J.; Tokuda, H.; Nishino, H.; Iwa-
shima, A. Chem. Pharm. Bull. 1993, 41, 1664; (b)
Murakami, C.; Kumagai, T.; Hada, T.; Kanekazu, U.;
Nakazawa, S.; Kamisuki, S.; Maeda, N.; Xu, X.; Yoshida,
H.; Sugawara, F.; Sakaguchi, K.; Mizushina, Y. Biochem.
Pharmacol. 2003, 65, 259; (c) Latowski, D.; Kostecka, A.;
Strzalka, K. Biochem. Soc. Trans. 2000, 28, 810; (d)