Anti-tumor-promoting activity of simple models of galactoglycerolipids with branched and unsaturated acyl chains

Article abstract of DOI:10.1016/j.ejmech.2004.09.021

Six new galactoglycerolipid analogs, in which one or two 4-methylpentanoyl or trans-2-butenoyl groups are linked to the 2-O-β-d-galactosylglycerol skeleton, were tested for their anti-tumor-promoting activity using a short-term in vitro assay for Epstein-

Full text of DOI:10.1016/j.ejmech.2004.09.021

European Journal of Medicinal Chemistry 40 (2005) 69–74
www.elsevier.com/locate/ejmech
Original article
Anti-tumor-promoting activity of simple models of galactoglycerolipids
with branched and unsaturated acyl chains
Diego Colombo a,*, Laura Franchini a, Lucio Toma b, Fiamma Ronchetti a, Nami Nakabe c,
Takako Konoshima d, Hoyoku Nishino c, Harukuni Tokuda c
a Dipartimento di Chimica, Biochimica e Biotecnologie per la Medicina, Università di Milano, Via Saldini 50, 20133 Milano, Italy
b Dipartimento di Chimica Organica, Università di Pavia, Via Taramelli 10, 27100 Pavia, Italy
c Department of Internal Medicine and Molecular Biochemistry, Kyoto Prefectural University of Medicine, Kamigyo-ku, Kyoto 602-0841, Japan
d Kyoto Pharmaceutical University, Yamashina-ku, Kyoto 607-8414, Japan
Received 29 July 2004; received in revised form 9 September 2004; accepted 27 September 2004
Available online 15 December 2004
Abstract
Six new galactoglycerolipid analogs, in which one or two 4-methylpentanoyl or trans-2-butenoyl groups are linked to the 2-O-b-D-
galactosylglycerol skeleton, were tested for their anti-tumor-promoting activity using a short-term in vitro assay for Epstein–Barr virus early
antigen (EBV-EA) activation. All these compounds were more active than their linear or saturated reference compounds in inhibiting the EBV
activation promoted by the tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA), the diester 1-O-(4-methylpentanoyl)-2-O-[6-O-(4-
methylpentanoyl)-b-D-galactopyranosyl]-sn-glycerol resulting the most active glycoglycerolipid analog till now tested. Four compounds
(three butenoates and one 4-methylpentanoate), when tested in the in vivo two-stage carcinogenesis test, exhibited also inhibitory effects on
mouse skin tumor promotion.
© 2004 Elsevier SAS. All rights reserved.
Keywords: Glycoglycerolipids; Cancer chemoprevention; Lipase transesterification; Epstein–Barr virus; Two-stage skin carcinogenesis test
1. Introduction
gen (EBV-EA) activation test [5,6] and also in the in vivo
two-stage mouse skin carcinogenesis test [7]. The activity of
such compounds (1a,b–3a,b) seems strictly related to the acyl
chain length, six carbon atoms resulting in maximum effect.
The shape of the lipophilic chain is also reported to influence
the anti-tumor activity of some PKC ligands the structure of
which is related to diacylglycerols [4]. In particular their activ-
Chemoprevention is dedicated to identifying agents with
potential preventive roles in cancer [1]. At present, the action
mechanism of chemopreventive compounds is not known,
however, it has been suggested that some of them could inhibit
the tumor promotion stage by interacting with the protein
kinase C [2], a family of serine/threonine kinases named PKC
that play an essential role in mediating cellular responses to
extracellular stimuli involved in proliferation, differentiation
and apoptosis [3].
PKC is activated by the potent tumor promoter 12-O-
tetradecanoylphorbol-13-acetate (TPA) that binds to it with
very high affinity [4]. Recently, glycoglycerolipid analogs
such as 2-O-glycosylglycerols linking short to medium length
lipophilic chains, have shown anti-tumor-promoting activity
in the TPA-promoted in vitro Epstein–Barr virus early anti-
ity was related to the presence of
a 4-methyl-3-
(methylethyl)pentanoyl chain, used to improve the hydropho-
bic contacts of these compounds with PKC. Moreover similar
lipophilic residues such as trans-3,4-dimethylpent-2-enoyl or
cis-2-butenoyl chains, seems to be a common motif that deco-
rate the structure of some natural compounds exhibiting a
potent inhibitory effect on EBV-EA activation induced by TPA
[8,9].
On the basis of these evidences, which indicate the nature
of the lipophilic chain as a crucial structural feature for the
activity of many compounds, we decided to prepare simple
branched and unsaturated glycoglycerolipid analogs in order
to study the effect on the anti-tumor-promoting activity caused
* Corresponding author. Tel.: +39 02 5031 6039; fax: +39 02 5031 6036.
E-mail address: diego.colombo@unimi.it (D. Colombo).
0223-5234/$ - see front matter © 2004 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2004.09.021

70
D. Colombo et al. / European Journal of Medicinal Chemistry 40 (2005) 69–74
by introducing small modifications to the acyl chains of gly-
coglycerolipid analogs 1a,b–3a,b and, eventually, to obtain
more active compounds. So here we describe the anti-tumor-
promoting activity of compounds 1c,d–3c,d based on the
short-term in vitro assay for the inhibition of EBV-EA acti-
vation induced by TPA. The inhibitory effect of compounds
1c and 1d–3d on mouse skin tumor promotion in an in vivo
two-stage mouse skin carcinogenesis test will be also dis-
cussed.
3. Pharmacology
Epstein–Barr virus (EBV) is known to be activated by
tumor promoters to produce viral early antigens (EA), and an
evaluation of its inhibition is often used as a primary screen-
ing for in vitro anti-tumor-promoting activities [13]. The
inhibitory effect of compounds 1c–3c and 1d–3d was assayed
using a short-term in vitro assay for EBV-EA activation in
Raji cells induced by the tumor promoter TPA, as described
in Refs. [14,15].
Mouse skin tumor promotion inhibition of compounds 1c
and 1d–3d was also evaluated in an in vivo two-stage mouse
skin carcinogenesis test as reported in the experimental pro-
tocols.
2. Chemistry
Compounds 1c–3c and 1d–3d were prepared according
to a well established chemo-enzymatic approach starting
from 2-O-b-D-galactosylglycerol (4) [10]. In particular,
Pseudomonas cepacia lipase (lipase PS) catalyzed transes-
terification of 4 with the proper vinyl ester as the acyl donor
afforded compounds 2c, 3c, 1d and 3d, whereas compound
2d was obtained using Bacillus subtilis protease (Proleather
FG-F) as the catalyst. Compound 1c was obtained through
lipase PS catalyzed transesterification of the derivative 5 fol-
lowed by removal of the chloroacetyl groups of 6c with hydra-
zine acetate as reported in Ref. [11]. The assignment of the
configuration of compounds 1c–d and 3c–d was determined
as reported in Ref. [10], and the glycerol part in their struc-
tures is shown by Fischer projection formula [12] (Scheme
1).
4. Results and discussion
The three branched analogs 1c–3c of hexanoates 1a–3a
and the three unsaturated analogs 1d–3d of butyrates 1b–3b,
were prepared and tested for their anti-tumor-promoting activ-
ity using a short-term in vitro assay for EBV activation in
Raji cells induced by TPA. Table 1 shows the in vitro tumor
inhibitory activity of compounds 1c,d–3c,d together with that
of the already studied [5,6] esters 1a,b–3a,b. Only weak cyto-
toxicity against Raji cells was observed for all compounds
(70% viability at 1000 mol ratio/TPA and >80% at all the
other mol ratio/TPA, Table 1) that showed potent inhibitory
activity, as indicated by their percentage to control (7.9–
17.5% at 500 mol ratio/TPA, Table 1).
In particular, comparing the new data with those previ-
ously published [5,6], both the branched 4-methylpentanoyl
derivatives (1c–3c) and the unsaturated crotonyl derivatives
(1d–3d) resulted more active than the corresponding linear
hexanoates (1a–3a) and the saturated butanoates (1b–3b)
respectively. In fact, going from linear to branched com-
pounds, the averaged inhibitory activities increased of 22%,
14% and 36% at 500 mol ratio/TPA (see Table 1, 1c–3c vs
Scheme 1
.
Table 1
Inhibitory effects of 1a–d, 2a–d and 3a–d on TPA-induced EBV-EA activation
Concentration (mol ratio/TPA)
% to control SE (n = 3)a
IC50
1000
500
100
10
1a [5]
2a [5]
3a [6]
1c
2c
3c
1b [5]
2b [5]
3b [6]
1d
2d
3d
0
0
0
0
0
0
0
0
0
0
0
0
0
11.4 0.3
10.7 0.1
12.4 0.2
8.9 0.7
9.2 0.9
7.9 0.6
22.4 0.4
20.4 0.8
21.1 0.4
15.6 1.1
17.5 1.8
13.8 1.4
32.1 0.9
30.1 0.9
32.3 1.1
25.2 1.2
28.9 1.5
23.5 1.1
45.1 1.1
43.6 1.0
47.6 1.2
34.5 2.1
36.6 2.0
31.5 2.0
63.4 1.3
67.8 1.2
69.5 1.8
65.0 2.3
67.0 2.1
62.0 2.0
80.2 1.0
78.2 1.1
80.5 1.3
73.3 1.3
75.7 1.6
71.5 1.5
25.0
28.3
32.2
22.7
26.5
19.4
70.1
62.1
75.2
39.8
46.1
34.2
0
0
0.4
0.2
0.2
0
0
0
0.6
0.5
0.3
a Values are EBV-EA activation (%) in the presence of the test compound relative to the control (100%). Activation was attained by treatment with TPA
32 pmol. IC50 represents the mol ratio to TPA that inhibits 50% of positive control (100%) activated with 32 pmol TPA.

D. Colombo et al. / European Journal of Medicinal Chemistry 40 (2005) 69–74
71
1a-3a) and of 30%, 14% and 35% going from saturated to
unsaturated compounds (Table 1, 1d–3d vs 1b–3b at 500 mol
ratio/TPA). The comparison of the IC50 values showed more
enhanced activities for unsaturated vs. saturated compounds
than for branched vs linear chain compounds (see Table 1).
In the case of linear saturated acyl chains it has been shown
that the introduction of a second acyl group does not signifi-
cantly modify the activity of the monoesters [6]. On the con-
trary, the branched and unsaturated diesters 3c and 3d here
studied represent the most active terms in the respective series
(Table 1), the compound 3c becoming the most active gly-
coglycerolipid analog till now prepared.
Compounds 1c and 1d were also submitted to an in vivo
two-stage carcinogenesis test of mouse skin papillomas using
dimethylbenz[a]anthracene (DMBA) as an initiator and TPA
as a promoter, in order to evaluate their inhibitory effect and
compare it with that of 1a that carries the linear lipophilic
chain on the same glycosylglycerol position of 1c and 1d,
and that is the most active glycoglycerolipid analog till now
assayed in the in vivo test [7]. Also the butenoate 2d, posi-
tional isomer of 1d, and the diester 3d were tested to evalu-
ate, respectively, the influence of the position and of the num-
ber of the acyl chains on the in vivo activity. After 20 weeks
of promotion, there was no statistically significant difference
in body weights between control and any treated group. The
activities, estimated by both the incidence (percentage of mice
bearing papillomas) (Fig. 1A) and the multiplicity (average
numbers of papillomas per mouse) (Fig. 1B), were compared
with those of the control group. In the control group 100% of
the mice bore papillomas at 10 weeks of promotion, and
5.4 and 9.3 papillomas per mouse were formed, respectively,
after 10 and 20 weeks of promotion (Fig. 1). All the tested
compounds were able to inhibit the tumor promotion in this
in vivo assay showing a similar effect on reducing the per-
centage of mice bore papillomas. In fact, in all the groups
about 27–33% of mice bore papillomas at 10 weeks of pro-
motion (Fig. 1), and 87–93% of mice bearing papillomas
resulted at 20 weeks of promotion (Fig. 1). In particular the
differences in the incidence of papillomas in the treated cases
with respect to control, analyzed by the v2-test, appear sig-
nificant after 10 weeks (P < 0.005) but not after 20 weeks.
The decrease of the number of papillomas per mouse (papil-
loma multiplicity) was statistically significant (P < 0.001, Stu-
dent’s t-test for 1c and 1d–3d) in all the treated cases with
respect to the control. The differences among the groups
treated with the unsaturated esters 1d–3d were not signifi-
cant. However, a significantly more pronounced inhibitory
effect was observed in the lowered number of papillomas per
mouse produced by 1c with respect to the unsaturated esters
1d (P < 0.02) and 2d (P < 0.05). In fact 1.6, 1.9 and 1.9 pap-
illomas (30%, 35% and 35% with respect to control) were
formed by treatment, respectively, with 1c, 1d and 2d at
10 weeks of promotion, and 5.1, 6.1 and 5.9 (55%, 65% and
63% with respect to control) at 20 weeks of promotion
(Fig. 1). The comparison of compounds 1a and 1c shows that
despite of a quite similar in vitro activity (Table 1), com-
Fig. 1
.
Inhibitory effects of compounds 1c and 1d–3d (85 nmol) on DMBA-
TPA mouse skin carcinogenesis and of 1a (Ref. [7]). All mice were initiated
with DMBA (390 nmol) and promoted with TPA (1.7 nmol) twice a week
starting 1 week after initiation. A: percentage of mice with papillomas; B:
averaged number of papillomas per mouse (*, TPA alone; D, TPA + 1c;
♦ , TPA + 1d; +, TPA + 2d; *, TPA + 3d; M, TPA + 1a). At 20 weeks of
promotion the averaged number of papillomas per mouse was reduced, with
respect to the control group, by 45% (P < 0.001), 34% (P < 0.001), 37%
(P < 0.001), 40% (P < 0.001) and 71% (P < 0.001) [7], respectively, for 1c,
1d, 2d, 3d and 1a.
pound 1c (the branched hexanoate) is significantly (P < 0.05,
papilloma incidence, and P < 0.001, multiplicity) less active
than 1a (the linear hexanoate) in the in vivo test (Fig. 1), in
which the results might be also influenced by differences in
the lipophilicity of the tested compounds. Glycoglycerolipid
analogs could serve as non-specific physical barriers for TPA
to get its target in the cells, so exerting their protective action
simply by delaying the onset of papillomas induced by TPA
on mouse skin. However, the hypothesis that non specific
interaction of glycoglycerolipids with membranes due to
detergency is involved in the anti-tumor-promoting action was
ruled out by Murakami et al. [16]; moreover, the fact that 1a,
1c and 1d–3d show significantly different activities (Fig. 1)
despite very similar structures, could account for a possible
specific interaction of glycoglycerolipid analogs with some
biological target. Nevertheless, work is still required to clarify
the action mechanism of tumor promotion inhibition by gly-

72
D. Colombo et al. / European Journal of Medicinal Chemistry 40 (2005) 69–74
coglycerolipid analogs and, eventually, to identify their bio-
logical target (e.g. PKC).
merically enriched 1d) could be obtained after repeated flash
chromatographies (dichloromethane/methanol from 95:5 to
75:25, v/v).
In conclusion, six new potent anti-tumor-promoting com-
pounds have been prepared, which structures are referred to
simple branched and unsaturated galactoglycerolipid ana-
logs, the branched diester 3c resulting the most in vitro active
glycoglycerolipid analog till now studied. The little changes
made to the structures, with respect to the linear and satu-
rated chains present in the previously studied galactoglycero-
lipid analogs, have exerted a positive effect on the anti-tumor-
promoting activity of these compounds, suggesting that both
unsaturated and branched acyl chains could be used to obtain
new more potent cancer chemopreventing agents as anti-
tumor-promoters. Besides the direct effects of their struc-
tural differences, the small variations in their chain lipophi-
licity might contribute to the modulation of their activity.
5.1.2.1. 2-O-[6-O-(4-methylpentanoyl)-b-D-galactopyrano-
20
1
syl]-sn-glycerol (2c). Yield 9%; oil;
a D : +4.4. H-NMR
͓ ͔
selected signals (pyridine-d5): d 4.75 (dd, 1H, J6′a,5′ = 4.9 Hz,
J6′a,6′b = 11.2 Hz, H-6′a); 4.89 (d, 1H, J6′b,5′ = 7.7 Hz, H-6′b);
5.07 (d, 1H, J1′,2′ = 7.7 Hz, H-1′); CI-MS m/z 370 [M + NH4]+.
Anal. C15H28O9 (C, H, O).
5.1.2.2. 1-O-(4-methylpentanoyl)-2-O-[6-O-(4-methylpenta-
noyl)-b-D-galactopyranosyl]-sn-glycerol (3c). Yield 40%; oil;
20
a
͓ ͔
= –1.8. 1H-NMR selected signals (pyridine-d5):
d 4.D63–4.70 (m, 2H, H-1a and H-1b); 4.74 (dd, 1H,
J6′a,5′ = 4.9 Hz, J6′a,6′b = 11.2 Hz, H-6′a); 4.88 (dd, 1H,
J6′b,5′ = 7.7 Hz, H-6′b); 4.98 (d, 1H, J1′,2′ = 7.7 Hz, H-1′);
CI-MS m/z 468 [M + NH4]+. Anal. C21H38O10 (C, H, O).
5. Experimental protocols
5.1.2.3. 1-O-(trans-2-butenoyl)-2-O-(b-D-galactopyranosyl)-
sn-glycerol (1d). Yield 10%; foam; 91% diastereomeric purity
(by 1H-NMR); 1H-NMR selected signals (pyridine-d5): d 4.65
(dd, 1H, J1a,2 = 5.6 Hz, J1a,1b = 12.0 Hz, H-1a); 4.71 (dd, 1H,
J1b,2 = 4.9 Hz, H-1b); 5.01 (d, 1H, J1′,2′ = 7.7 Hz, H-1′); CI-MS
m/z 340 [M + NH4]+. Anal. C13H22O9 (C, H, O).
5.1. Chemistry
5.1.1. Materials
P. cepacia lipase (lipase PS, specific activity 30.5 triacetin
units/mg solid), a generous gift from Amano Pharmaceutical
Co. (Mitsubishi Italia), was supported on celite [17] and kept
overnight, under vacuum, prior to use. B. subtilis protease
(Proleather FG-F, specific activity 10 units/mg solid) was pur-
chased from Amano. Pyridine was distilled from calcium
hydride. The acyl carriers were purchased fromAldrich (vinyl
crotonate) or synthesized from 4-methylvaleric acid (Ald-
rich) and vinyl versatate (Aldrich) [18]. 2-O-b-D-
(galactopyranosyl)glycerol (4) was synthesized according to
literature procedures [19]. Evaporation under reduced pres-
sure was always effected with a bath temperature below 40 °C.
5.1.2.4. 2-O-[6-O-(trans-2-butenoyl)-b-D-galactopyranosyl]-
sn-glycerol (2d). Yield 5%; m.p. 120–121 °C (from CH2Cl2–
1
iPr2O); H-NMR selected signals (pyridine-d5): d 4.83 (dd,
1H, J6′b,5′ = 4.9 Hz, J6′b,6′a = 11.2 Hz, H-6′b); 4.87 (d, 1H,
J6′a,5′ = 7.7 Hz, H-6′a); 5.06 (d, 1H, J1′,2′ = 7.7 Hz, H-1′); MS
m/z 340 [M + NH4]+. Anal. C13H22O9 (C, H, O).
5.1.2.5. 1-O-(trans-2-butenoyl)-2-O-[6-O-(trans-2-butenoyl)-
b-D-galactopyranosyl]-sn-glycerol (3d). Yield 15%; foam;
a
20 = –1.1; 1H-NMR selected signals (pyridine-d5): d 4.65
͓ ͔
1
All the new compounds were characterized by H-NMR
(dd,D1H, J1a,2 = 5.6 Hz, J1a,1b = 12.0 Hz, H-1a); 4.73 (dd, 1H,
J1b,2 = 4.9 Hz, H-1b); 4.82 (dd, 1H, J6′a,5′ = 5.6 Hz,
J6′b,6′a = 11.2 Hz, H-6′a); 4.88 (dd, 1H, J6′b,5′ = 7.0, H-6′b);
4.98 (d, 1H, J1′,2′ = 7.7 Hz, H-1′); MS m/z 408 [M + NH4]+.
Anal. C17H26O10 (C, H, O).
analysis at 500 MHz and chemical ionization mass spectrom-
etry (CI-MS) [11]. The elemental analyses were consistent
with the theoretical ones. Optical rotations were determined
on a Perkin–Elmer 241 polarimeter in methanol solutions (c
= 1.0) in a 1 dm cell at 20 °C unless otherwise stated. Melting
points (m.p.) were recorded on a Büchi 510 capillary m.p.
apparatus and were uncorrected. Analyses of the new com-
pounds, indicated by the symbols of the elements, were within
0.4% of the theoretical values.
5.1.3. Synthesis of compound 1c
5.1.3.1. 2-O-(2,3,4,6-tetra-O-chloroacetyl-b-D-galactopy-
ranosyl)glycerol (5). 1.2g (2.76 mmol) of 1,3-di-O-benzyl-
2-O-(b-D-galactopyranosyl)glycerol [20] were dissolved in
36 ml of a dichlorometane/pyrydine 9:1 mixture and treated
with chloroacetic anhydride (3.14 g, 18.4 mmol) at 0 °C for
15 min. Usual work-up and flash chromatography (petro-
leum ether/ethyl acetate 70:30, v/v) yielded the tetrachloro-
5.1.2. General procedure for the enzymatic synthesis
of compounds 2c, 3c, 1d–3d
2-O-b-D-(galactopyranosyl)glycerol (4) (0.40 g,
1.57 mmol) was dissolved in pyridine (5 ml) and the appro-
priate vinyl ester (6 mmol) and lipase PS or Proleather FG-F
(2.50 g) were added in the order. The mixture was stirred for
4 or 18 h (using vinyl crotonate) at 45 °C and the reaction
was stopped by filtering off the enzyme and washing with
pyridine and methanol. The solvent was removed under
vacuum and the esters (pure 2c, 3c, 2d and 3d or diastereo-
acetylated derivative (1,43 g, 1.93 mmol), oil, a 20 = –1.24
͓ ͔
D
1
(c 1.0, CHCl3). H-NMR selected signals (CDCl3): d 3.83,
3.95, 3.99 and 4.14 (4s, 8H, 4CH2Cl); 4.02 (m, 1H, H-2);
4.46–4.53 (m, 4H, 2CH2Ph); 4.73 (d, 1H, J1′,2′ = 7.7 Hz, H-1′).
Anal. C31H34Cl4O12 (C, H, O). Following catalytic hydro-

D. Colombo et al. / European Journal of Medicinal Chemistry 40 (2005) 69–74
73
genolysis (10% Pd/C in methanol) yielded 0.93 g (1.7 mmol)
triplicate for each compound. The average EBV-EA inhibi-
tory activity of the test compounds was compared to that of
control experiments (100%) with butyric acid (4 mM) and
TPA (32 pM) in which EBV-EA induction was ordinarily
around 30%. The viability of the cells was assayed against
treated cells using the Trypan Blue staining method. For the
determination of cytotoxicity, the cell viability was required
to be more than 60% 2 days after treatment with the com-
pounds for an accurate result.
of 2-O-(2,3,4,6-tetra-O-chloroacetyl-b-D-galactopyranosyl)
glycerol (5), amorphous solid, a 20 = +10.8 (c 1.0, CHCl3);
͓ ͔
D
m.p. 141–142 °C. 1H-NMR selected signals (CDCl3): d 3.79
(m, 1H, H-2); 3.97 and 4.17 (2s, 4H, 2CH2Cl); 4.01–4.13 (m,
5H, 2CH2Cl and H-5′); 4.73 (d, 1H, J1′,2′ = 7.7 Hz, H-1′).
Anal. C17H22Cl4O12 (C, H, O).
5.1.3.2. 1-O-(4-methylpentanoyl)-2-O-(2,3,4,6-tetra-O-chlo-
roacetyl-b-D-galactopyranosyl)-sn-glycerol (6c). Com-
pound 5 (0.93 g, 1.7 mmol) was dissolved in dry dichlo-
romethane (20 ml) and vinyl 4-methylvalerate (15 mmol) and
lipase PS (4.8 g) were added in the order. After stirring for
40 h at 45 °C the lipase was filtered off and the solvent
removed under vacuum. Chromatographic purification (petro-
leum ether/ethyl acetate 1:1, v/v) yielded 0.76 g (1.15 mmol)
of compound 6c, 93% diastereomeric purity (by 1H-NMR),
5.2.2. In vivo two-stage mouse skin carcinogenesis test
Female ICR mice were obtained at 5–6 weeks of age from
SLC Co. Ltd. (Shizuoka, Japan). Groups of animals (15 ani-
mals per group) were housed in bunches of five in polycar-
bonate cages. Mice were permitted free access to MP solid
diet (OrientalYeast Co. Ltd., Chiba, Japan) and drinking water
at all times during the study. The back of each mouse was
shaved with surgical clippers before the first day of initiation.
Tumors on the back of the mice were initiated with DMBA
(390 nmol) in acetone (0.1 ml). One week after initiation,
they were promoted twice a week by application of TPA
(1.7 nmol) in acetone (0.1 ml). For the animals in the test
compound treated groups the mice were treated with the test
compounds (85 nmol) in acetone (0.1 ml) 1 h before each
TPA treatment. The incidence of papillomas was observed
weekly for 20 weeks. The differences in mouse skin papillo-
mas between control and experiments were analyzed by
means of the Student’s t-test at 20 weeks of promotion,
whereas the differences in tumor bearing mice were ana-
lyzed by means of the v2-test.
1
oil, H-NMR selected signals (CDCl3): d 0.89 (d, 6H, J
= 6.3 Hz, 2CH3); 1.46–1.60 (m, 3H, CH2 and CH); 2.30 (m,
2H, CH2CO); 3.88 (m, 1H, H-2); 3.97 (s, 2H, CH2Cl); 4.02–
4.18 (m, 9H, 3CH2Cl, H1a, H1b and H-5′); 4.67 (d, 1H, J1′,2′
= 7.7 Hz, H-1′). Anal. C23H32Cl4O13 (C, H, O).
5.1.3.3. 1-O-(4-methylpentanoyl)-2-O-(b-D-galactopyrano-
syl)-sn-glycerol (1c). To a solution of 6c (0.76 g, 1.15 mmol)
in a 1:1 ethyl acetate/methanol mixture (22 ml) hydrazine
acetate (1.6 g, 17.4 mmol) was added. After 4 h at room tem-
perature the solvent was removed and the crude reaction mix-
ture was submitted to repeated flash chromatografies
(dichoromethane/methanol, 80:20, v/v) yielding 0.197 g
(0.56 mmol) of compound 1c, 93% diastereomeric purity (by
1
1H-NMR), oil, H-NMR selected signals (pyridine-d5):
d 4.60–4.70 (m, 2H, H-1a and H-1b); 4.99 (d, 1H,
J1′,2′ = 7.7 Hz, H-1′); CI-MS m/z 370 [M + NH4]+. Anal.
C15H28O9 (C, H, O).
Acknowledgements
This study was supported in part by the Ministero
dell’Istruzione, dell’Università e della Ricerca, Rome, Italy
(COFIN 2003: antitumoral natural and related synthetic com-
pounds) and University of Milan (Italy) and in part by Grants-
in-Aid from the Ministry of Education, Science and Culture
and the Ministry of Health and Welfare (Japan), and also sup-
ported by grant from the National Cancer Institute (CA
17625), USA.
5.1.4. Configuration assignment at C-2 for 1c–d
and 3c–3d
The 1-O-acyl-sn-glycerols obtained by treatment of 1c and
1d with b-galactosidase from Aspergillus oryzae, were reacted
with acetone, 2,2-dimethoxypropane and p-toluenesulfonic
acid. The obtained 1-O-acyl-2,3-O-isopropylidene-sn-
glycerols were analyzed by chiral GLC and compared with
authentic analytical standards [10]. The configuration assign-
ments for the 1,6′-diesters 3c and 3d were obtained by selec-
tive LPS catalyzed transesterification of the 1-monoesters 1c
and 1d with the proper vinyl ester [10].
References
[1] L.W. Wattenberg, Cancer Res. 53 (1993) 5890–5896.
[2] G.J. Kapadia, M.A. Azuine, J. Takayasu, T. Konoshima, M. Takasaki,
H. Nishino, H. Tokuda, Cancer Lett. 161 (2000) 221–229.
[3] Y. Nishizuka, Science 258 (1992) 607–614.
5.2. Pharmacology
[4] K. Nacro, B. Bienfait, J. Lee, K.C. Han, J.H. Kang, S. Benzaria,
N.E. Lewin, D.K. Bhattacharyya, P.M. Blumberg, V.E. Marquez, J.
Med. Chem. 43 (2000) 921–944.
[5] D. Colombo, F. Compostella, F. Ronchetti, A. Scala, L. Toma,
T. Mukainaka, et al., Cancer Lett. 143 (1999) 1–4.
5.2.1. Short-term in vitro bioassay for anti-tumor
promoters
Inhibition was tested using a short-term in vitro assay for
EBV activation in Raji cells cultivated in RPMI 1640 medium
containing 10% fetal calf serum, and induced by TPA as
described previously [14,15]. The assays were performed in
[6] D. Colombo, F. Compostella, F. Ronchetti, A. Scala, H. Tokuda,
H. Nishino, Eur. J. Med. Chem. 36 (2001) 691–695.

74
D. Colombo et al. / European Journal of Medicinal Chemistry 40 (2005) 69–74
[7] D. Colombo, F. Compostella, F. Ronchetti, A. Scala, L. Toma,
T. Mukainaka, M. Kuchide, H. Tokuda, H. Nishino, Cancer Lett. 161
(2000) 201–205.
[8] A. Iida, K. Konishi, H. Kubo, K. Tomioka, H. Tokuda, H. Nishino,
Tetrahedron Lett. 37 (1996) 9219–9220.
[9] M. Okano, N. Fukamiya, K. Tagahara, H. Tokuda, A. Iwashima,
H. Nishino, et al., Cancer Lett. 94 (1995) 139–146.
[10] D. Colombo, F. Ronchetti, A. Scala, L. Toma, Tetrahedron Asymm. 9
(1998) 2113–2119.
[11] D. Colombo, F. Compostella, F. Ronchetti, A. Scala, L. Toma,
H. Tokuda, et al., Bioorg. Med. Chem. 7 (1999) 1867–1871.
[12] M.A. Chester, Pure Appl. Chem. 69 (1997) 2475–2487.
[13] A. Murakami, H. Ohigashi, K. Koshimizu, Food Rev. Int. 15 (1999)
335–395.
[14] H. Shirahashi, N. Murakami, M. Watanabe,A. Nagatsu, J. Sakakibara,
H. Tokuda, et al., Chem. Pharm. Bull. (Tokyo) 41 (1993) 1664–1666.
[15] H. Tokuda, T. Konoshima, M. Kozuka, T. Kimura, Cancer Lett. 40
(1988) 309–317.
[16] A. Murakami, Y. Nakamura, K. Koshimizu, H. Ohigashi, J. Agric.
Food Chem. 43 (1995) 2779–2783.
[17] D. Colombo, F. Ronchetti, A. Scala, I.M. Taino, F. Marinone Albini,
L. Toma, Tetrahedron Asymm. 5 (1994) 1377–1384.
[18] M.A.S. Mondal, R. Van der Meer, A.L. German, D. Heikens, Tetrahe-
dron 30 (1974) 4205–4207.
[19] P.W. Austin, F.E. Hardy, J.G. Buchanan, J. Baddiley, J. Chem. Soc.
(1965) 1419–1424.
[20] F. Compostella, D. Colombo, P. Ferraboschi, A. Scala, L. Toma,
F. Ronchetti, Eur. J. Org. Chem. (2002) 1429–1435.