Structure and cytotoxic activity of enzymatic hydrolysis products of secoiridoid glucosides, isoligustroside and isooleuropein

Article abstract of DOI:10.1002/cbdv.201000092

Hydrolysis of isoligustroside (1) and isooleuropein (2), secoiridoid glucosides, in the presence of β-glucosidase provided 2-(4-hydroxyphenyl) methyl (2R,3S,4S)-3-formyl-3,4-dihydro-4-(2-methoxy-2-oxoethyl)-2-methyl-2H- pyran-5-carboxylate (3) and 2-(3,4-

Full text of DOI:10.1002/cbdv.201000092

CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
651
Structure and Cytotoxic Activity of Enzymatic Hydrolysis Products of
Secoiridoid Glucosides, Isoligustroside and Isooleuropein
by Masafumi Kikuchia)b), Yasunori Yaoitab), Nariyasu Manoa), and Masao Kikuchi*b)
a) Department of Pharmaceutical Sciences, Tohoku University Hospital, 1-1 Seiryo-machi, Aoba-ku,
Sendai, Miyagi, 980-8547, Japan
b) Department of Molecular Structural Analysis, Tohoku Pharmaceutical University, 4-4-1
Komatsushima, Aoba-ku, Sendai, Miyagi 981-8558, Japan
(phone: þ81-22-234-4181; fax: þ81-22-275-2013; e-mail: mkikuchi@tohoku-pharm.ac.jp)
Hydrolysis of isoligustroside (1) and isooleuropein (2), secoiridoid glucosides, in the presence of b-
glucosidase provided 2-(4-hydroxyphenyl)methyl (2R,3S,4S)-3-formyl-3,4-dihydro-4-(2-methoxy-2-oxo-
ethyl)-2-methyl-2H-pyran-5-carboxylate (3) and 2-(3,4-dihydroxyphenyl)methyl (2R,3S,4S)-3-formyl-
3,4-dihydro-4-(2-methoxy-2-oxoethyl)-2-methyl-2H-pyran-5-carboxylate (4), respectively. The struc-
tures of 3 and 4 were elucidated on the basis of extensive spectral analyses, including 2D-NMR experi-
ments. Compounds 3 and 4 were found to be new rearrangement products of the aglycones of 1 and 2. The
cytotoxic activities of 3 and 4 were evaluated using a disease-oriented panel of 39 human cancer cell lines
and showed moderate cytotoxic activity for 4, while 3 exhibited weaker activity compared to that of 4.
Introduction. – Lilacs (Syringa) are a genus of ca. 20 species of flowering plants in
the family Oleaceae, and they are used as antimicrobial, antipyretic, and antivirus
medicines. In a previous article, we reported the isolation and structure elucidation of
secoiridoid glucosides isoligustroside (1) and isooleuropein (2) from the leaves of S.
vulgaris L. [1]. It has been reported that the aglycones, obtained by treatment of
secoiridoid glucosides, gentiopicroside and sweroside, with b-glucosidase, exhibited
antitumor activities; however, gentiopicroside and sweroside did not display this
activity [2]. These observations indicated that these secoiridoid glucosides possess
bioactivities after only deglucosidation [3]. This prompted us to investigate the
bioactivity of the enzymatic hydrolysis products of 1 and 2. Here, we describe the
structure and cytotoxic activity of the products 3 and 4, obtained from the enzymatic
hydrolysis of 1 and 2 with b-glucosidase.
ꢀ 2011 Verlag Helvetica Chimica Acta AG, Zꢁrich

652
CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
Results and Discussion. – 1. Enzymatic Hydrolysis and Structure Elucidation.
Hydrolysis of compounds 1 and 2 in the presence of b-glucosidase provided compounds
3 and 4 in 20 and 25% yield, respectively. The reaction conditions of Guiso and Marra
[4] were adopted, i.e., a solution of the substrate in H2O was treated with an excess of b-
glucosidase at 348.
Compound 3 was obtained as an amorphous powder. Its molecular formula was
determined as C19H22O7 based on HR-EI-MS (m/z 362.1367 (Mþ ; calc. 362.1365)). The
UV spectrum suggested the presence of an enol-ether system conjugated with a CO
group (232 nm), which is typical for many iridoid and secoiridoid skeletons [5]. In
addition, absorption due to a phenol chromophore (276 nm) was also observed. The
1H-NMR spectrum of 3 (in CDCl3; Table 1) exhibited signals due to one Me group
(d(H) 1.42 (d, J¼6.6)), one MeO group (d(H) 3.66 (s)), one O-bearing CH group
(d(H) 4.43 (dq, J¼6.6, 6.2)), one C¼CꢀH moiety (d(H) 7.45 (s)), and one aldehyde H-
atom (d(H) 9.61 (d, J¼1.8)). The 1H-NMR spectrum also indicated the presence of a
p-substituted aromatic ring. The 13C-NMR spectrum of 3 (in CDCl3; Table 1) displayed
1
signals due to three C¼O groups (d(C) 166.5, 172.4, and 199.9). The H,1H-COSY
spectrum of 3 (Fig. 1) implied connectivities of HꢀC(1) to HꢀC(8)1), of HꢀC(5) to
CH2(6), of HꢀC(5) to HꢀC(8), of HꢀC(8) to HꢀC(9), of HꢀC(9) to Me(10), of
CH2(1’) to CH2(2’), of HꢀC(4’) to HꢀC(5’), and of HꢀC(7’) to HꢀC(8’). The HMBC
1
Table 1. 1H- (600 MHz) and 13C-NMR (150 MHz) Data of Compounds 3 and 4 in CDCl3 ). d in ppm,
J in Hz.
3
4
d(H)
d(C)
d(H)
d(C)
HꢀC(1)
HꢀC(3)
C(4)
9.61 (d, J¼1.8)
7.54 (s)
–
199.9
155.4
106.7
27.2
9.60 (s)
7.56 (s)
–
200.0
155.7
106.5
27.0
HꢀC(5)
3.38–3.40 (m)
2.50 (dd, J¼16.1, 9.5),
2.85 (dd, J¼16.1, 3.7)
–
2.68 (ddd, J¼6.2, 5.9, 1.8)
4.43 (dq, J¼6.6, 6.2)
1.42 (d, J¼6.6)
–
4.28–4.33 (m)
2.88–2.90 (m)
–
7.08 (d, J¼8.1)
6.77 (d, J¼8.1)
–
3.42–3.43 (m)
CH2(6)
36.8
2.55 (dd, J¼16.1, 9.9),
37.0
2.84–2.88 (m)
C(7)
172.4
54.7
70.8
19.4
166.5
64.9
–
173.0
54.5
70.7
19.4
166.7
65.1
HꢀC(8)
HꢀC(9)
Me(10)
C(11)
CH2(1’)
CH2(2’)
C(3’)
HꢀC(4’)
HꢀC(5’)
C(6’)
HꢀC(7’)
HꢀC(8’)
MeO
2.65–2.67 (m)
4.47 (dq, J¼6.6, 6.2)
1.42 (d, J¼6.6)
–
4.30–4.32 (m)
2.84–2.88 (m)
–
6.80 (s)
–
–
6.75–6.80 (m)
6.62 (d, J¼7.0)
3.69 (s)
34.3
34.4
130.02
130.04
115.4
154.3
115.4
130.04
51.8
130.6
115.3
143.5
142.7
116.1
121.2
52.1
6.77 (d, J¼8.1)
7.08 (d, J¼8.1)
3.66 (s)
1
)
Arbitrary numbering, see Formulae. For systematic names, see Exper Part.

CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
653
Fig. 1. 1H,1H-COSY (bold lines) and HMBC (full-line
arrows) correlations for 3
spectrum (Fig. 1) showed the correlations HꢀC(3)/C(4), C(9), and C(11); HꢀC(5)/
C(4); CH2(6)/C(7); CH2(2’)/C(3’); HꢀC(4’)/C(3’); HꢀC(5’)/C(6’); HꢀC(7’)/C(6’);
HꢀC(8’)/C(3’), and MeO/C(7). By considering the chemical shift of C(11) (d(C)
166.5), the linkage position of the 2-(4-hydroxyphenyl)ethoxy group was determined as
C(11), affording the constitution of 3. The relative configuration of 3 was determined as
follows. In the 1H-NMR spectrum, HꢀC(8) had a large coupling constant (J(8a,9b)¼
6.2). In an NOESY experiment, NOE correlations HꢀC(1)/HꢀC(5), HꢀC(5)/HꢀC(9),
and HꢀC(8)/Me(10), were observed revealing that 3 adopted a conformation as shown
in Fig. 2, and the CH2COOMe group at C(5), the CHO group at C(8), and Me(10)
were a-, b-, and a-oriented, respectively (Fig. 2). The absolute configuration of 3 was
determined as (5S,8S,9R)1) by the positive [a]D value [6]. Accordingly, the structure of
3 was elucidated as 2-(4-hydroxyphenyl)ethyl (2R,3S,4S)-3-formyl-3,4-dihydro-4-(2-
methoxy-2-oxoethyl)-2-methyl-2H-pyran-5-carboxylate.
Fig. 2. Selected coupling constant (dotted arrow) and
NOESY (full-line arrows) correlations for 3
1
Compound 4 was obtained as an amorphous powder. The general features of H-
and 13C-NMR spectra of 4 (in CDCl3; Table 1) closely resembled those of 3 except that
1
one of the aromatic H-atoms in H-NMR spectrum of 4 was missing. The molecular
formula was determined as C19H22O8 based on HR-EI-MS (m/z 378.1318 (Mþ ; calc.
378.1315)). Consequently, 4 was deduced to be a compound in which an aromatic H-
atom in 3 was replaced by an OH group. The resonance for C(5’) at d(C) 115.4 of 3 was
shifted downfield to d(C) 143.5 in 4, suggesting that an additional OH group was
located at C(5’) [7]. This was confirmed by the HMBC spectrum, in which the cross-
peak HꢀC(4’)/C(5’) was observed. The absolute configuration of 4 was determined as
(5S,8S,9R)1) by the positive [a]D value [6]. From these data, the structure of 4 was
elucidated as 2-(3,4-dihydroxyphenyl)ethyl (2R,3S,4S)-3-formyl-3,4-dihydro-4-(2-me-
thoxy-2-oxoethyl)-2-methyl-2H-pyran-5-carboxylate.

654
CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
A proposed pathway for the formation of 3 and 4 is depicted in the Scheme. The
aglycones 5 and 6, resulting from the enzymatic hydrolysis of the glucosidic linkage of 1
and 2, respectively, are in equilibrium with 7 and 8, respectively. Compounds 3 and 4 are
formed from 7 and 8, respectively, by Michael-type addition [6][8]. Thus, a rearrange-
ment of the dihydropyran ring of 5 and 6 occurs during the enzymatic hydrolysis.
Compounds 3 and 4 were found to be new rearrangement products of aglycones.
Scheme
2. Growth Inhibition against a Panel of 39 Human Cancer Cell Lines. To evaluate
compounds for their cell growth inhibition profile, a human cancer cell line panel
combined with a database was established by the Cancer Chemotherapy Center of the
Japanese Foundation for Cancer Research [9]. The system as a whole was developed
according to the method of the National Cancer Institute, with modifications [10]. The
cell-line panel consisted of 39 human cancer cell lines (five breast, six central nervous
system, five colon, seven lung, one melanoma, five ovary, two kidney, six stomach, and
two prostate cancers). The cancer growth inhibitory properties of 3 and 4 were

CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
655
examined using a disease-oriented panel of 39 human cancer cell lines. The results are
presented in Table 2 as log GI50 (logarithmic concentration of compound for inhibition
of cell growth at 50% relative to control) and GI50 (concentration of compound for
inhibition of cell growth at 50% relative to control) values. Compound 4 showed
significant cytotoxic activity against the lung cancer DMS273 (log GI50 ꢀ5.19 (6.4 mm))
and DMS114 (log GI50 ꢀ5.06 (8.7 mm)) cell lines, and the average logarithm of the GI50
(MG-MID) across all cell lines tested was ꢀ4.66 (22 mm). The MG-MID value of 3 was
ꢀ4.41 (39 mm). These data disclosed that 4 showed moderate cytotoxic activity, while 3
was weaker in activity compared to 4. It is noteworthy that the OH group at C(5’) of 4 is
important to enhance the cytotoxic activity. The D value (the difference in log GI50
value between the most sensitive cell and the MG-MID value) and range value (the
difference in the log GI50 value between the most and least sensitive cells) have been
employed as a criterion for the selective cytotoxic activity (effective value: Dꢁ0.5 as
well as range ꢁ1.0) [9]. As shown in Table 2, the D and range values of 4 were 0.54 and
1.19, respectively, indicating that 4 exhibited selective cytotoxic activity. On the other
hand, 3 did not show selective cytotoxic activity (D¼0.40, range¼0.81). Most
anticancer agents with the same mechanism of action display similar cell growth
inhibition profiles against a cancer cell line database. The cell growth inhibition profiles
of 3 and 4 were submitted to data analysis by the COMPARE program [10], a database
covering cell growth inhibition profiles of over 200 known anticancer agents with
various action mechanisms. This program can be used to predict the mode of action of
test compounds by assessing the correlation coefficient (r) between the cell growth
inhibition profiles mediated by such test compounds and various reference compounds
with known modes of action. The highest r values of 3 and 4 were realized with 6-
sulfanylpurine (r¼0.601), a purine antagonists, for 3, and nitrogen mustard (r¼0.574),
a DNA alkylating agent, for 4, respectively, suggesting that the modes of action of 3 and
4 are similar to those of the respective anticancer agents.
We are grateful to Dr. T. Yamori (Screening Committee of Anticancer Drugs supported by a Grant-
in-Aid for Scientific Research on Priority Area ꢀCancerꢁ from the Ministry of Education, Culture, Sports,
Science and Technology, Japan) for performing the assay for cytotoxicity using a panel of 39 human
cancer cell lines. We are also grateful to Mr. S. Satoh and Mr. T. Matsuki of Tohoku Pharmaceutical
University for NMR and MS recordings.
Experimental Part
General. HPLC: CCPD pump (Tosoh) and UV-8011 UV/VIS detector (Tosoh); tR in min. Optical
rotations: JASCO DIP-360 digital polarimeter. UV Spectra: Beckman DU-64 spectrophotometer. NMR
Spectra: JEOL JNM-LA 600 (1H: 600 MHz; 13C: 150 MHz) spectrometer; chemical shifts d in ppm, rel.
to Me4Si, J in Hz. EI-MS: JEOL JMS-DX 303 mass spectrometer; in m/z.
Substrate Material. Isoligustroside (1) and isooleuropein (2) were isolated from the leaves of Syringa
vulgaris, and characterized by spectroscopic methods as described in [1].
Enzymatic Hydrolysis of 1 and 2 with b-Glucosidase. Compound 1 (25 mg) was dissolved in H2O
(5.0 ml). The soln. was treated with an excess of b-glucosidase (from almonds; 15.3 units/mg; Toyobo
Corporation) for 1 d at 348. The soln. was then extracted with an equal amount of CHCl3 (3ꢂ ), and the
CHCl3 layer was evaporated under reduced pressure. The residue was purified by prep. HPLC (Cosmosil
5SL-II column (250ꢂ10 mm, 10 mm; Nacalai Tesque); CHCl3/hexane/MeOH/AcOH 55 :45 :4 :2, 1.0 ml/
min) to give 3 (3.5 mg, tR 24.0). By a similar procedure except the prep. HPLC conditions (Cosmosil 5SL-

656
CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
Table 2. Growth-Inhibition Activities of Compounds 3 and 4 against a Panel of 39 Human Cancer Cell
Lines
Type of cancer
Breast
Cell line
3
4
log GI50 [m]a) GI50 [mm]b) log GI50 [m] GI50 [mm]
HBC-4
BSY-1
HBC-5
MCF-7
ꢀ4.32
ꢀ4.25
ꢀ4.42
ꢀ4.44
48
56
38
36
42
ꢀ4.67
ꢀ4.71
ꢀ4.72
ꢀ4.66
ꢀ4.74
21
19
19
22
18
MDA-MB-231 ꢀ4.38
Central nervous system (brain) U251
ꢀ4.34
ꢀ4.28
ꢀ4.43
ꢀ4.56
ꢀ4.46
ꢀ4.36
46
52
38
28
34
44
ꢀ4.63
ꢀ4.67
ꢀ4.54
ꢀ4.74
ꢀ4.69
ꢀ4.62
24
21
29
18
21
24
SF-268
SF-295
SF-539
SNB-75
SNB-78
Colon
Lung
HCC2998
KM-12
HT-29
HCT-15
HCT-116
ꢀ4.51
ꢀ4.35
ꢀ4.40
ꢀ4.37
ꢀ4.44
ꢀ4.37
ꢀ4.49
ꢀ4.81
ꢀ4.21
ꢀ4.00
ꢀ4.77
ꢀ4.58
31
45
40
42
37
ꢀ4.76
ꢀ4.61
ꢀ4.50
ꢀ4.65
ꢀ4.61
ꢀ4.57
ꢀ4.62
ꢀ4.95
ꢀ4.53
ꢀ4.00
ꢀ5.19
ꢀ5.06
17
25
32
22
25
NCI-H23
NCI-H226
NCI-H522
NCI-H460
A549
43
33
16
27
24
11
62
29
>100
17
>100
DMS273
DMS114
6.4
8.7
26
Melanoma
Ovary
LOX-IMVI
ꢀ4.54
29
ꢀ4.74
18
OVCAR-3
OVCAR-4
OVCAR-5
OVCAR-8
SK-OV-3
ꢀ4.71
ꢀ4.21
ꢀ4.46
ꢀ4.44
ꢀ4.27
19
61
35
36
54
ꢀ4.74
ꢀ4.65
ꢀ4.68
ꢀ4.55
ꢀ4.58
18
22
21
28
26
Kidney
RXF-631L
ACHN
ꢀ4.53
30
48
ꢀ4.62
24
20
ꢀ4.32
ꢀ4.70
Stomach
St-4
ꢀ4.47
ꢀ4.44
ꢀ4.40
ꢀ4.39
ꢀ4.31
ꢀ4.40
34
36
40
41
48
40
ꢀ4.55
ꢀ4.78
ꢀ4.61
ꢀ4.57
ꢀ4.65
ꢀ4.62
28
16
25
27
22
24
MKN1
MKN7
MKN28
MKN45
MKN74
Prostate
DU-145
PC-3
ꢀ4.07
85
44
ꢀ4.56
28
26
ꢀ4.35
ꢀ4.59
MG-MIDc)
Dd)
Rangee)
ꢀ4.41
0.40
0.81
ꢀ4.66
0.54
1.19
a) Logarithmic concentration of compound for inhibition of cell growth at 50% relative to control.
c
b) Concentration of compound for inhibition of cell growth at 50% relative to control. ) Mean value of
log GI50 over all cell lines tested. d) Difference in log GI50 value between the most sensitive cell and the
MG-MID value. e) Difference in log GI50 value between the most and least sensitive cells.

CHEMISTRY & BIODIVERSITY – Vol. 8 (2011)
657
II column (250ꢂ10 mm, 10 mm, Nacalai Tesque); CHCl3/hexane/MeOH/AcOH 55 :45 :4 :1, 1.0 ml/min),
4 (9.8 mg, tR 30.6) was obtained from 2 (55 mg).
2-(4-Hydroxyphenyl)ethyl (2R,3S,4S)-3-Formyl-3,4-dihydro-4-(2-methoxy-2-oxoethyl)-2-methyl-2H-
pyran-5-carboxylate (3). Amorphous powder. [a]2D5 ¼ þ30.9 (c¼0.35, CHCl3). UV (MeOH): 232 (4.2),
276 (3.3). 1H- and 13C-NMR: see Table 1. EI-MS: 362 (Mþ ). HR-EI-MS: 362.1367 (Mþ, C19H22Oþ7 ; calc.
362.1365).
2-(3,4-Dihydroxyphenyl)ethyl (2R,3S,4S)-3-Formyl-3,4-dihydro-4-(2-methoxy-2-oxoethyl)-2-methyl-
2H-pyran-5-carboxylate (4). Amorphous powder. [a]2D7 ¼ þ31.0 (c¼0.98, CHCl3). UV (MeOH): 231
(4.5), 281 (3.8). 1H- and 13C-NMR: see Table 1. EI-MS: 378 (Mþ ). HR-EI-MS: 378.1318 (Mþ,
C19H22Oþ8 ; calc. 378.1315).
Cell Growth-Inhibition Assay. This experiment was carried out at the Cancer Chemotherapy Center
of the Japanese Foundation for Cancer Research. The screening panel consisted of the following 39 human
cancer cell lines: breast cancer HBC-4, BSY-1, HBC-5, MCF-7, and MDA-MB-231; brain cancer U251,
SF-268, SF-295, SF-539, SNB-75, and SNB-78; colon cancer HCC2998, KM-12, HT-29, HCT-15, and
HCT-116; lung cancer NCI-H23, NCI-H226, NCI-H522, NCI-H460, A549, DMS273, and DMS114;
melanoma LOX-IMVI; ovarian cancer OVCAR-3, OVCAR-4, OVCAR-5, OVCAR-8, SK-OV-3; renal
cancer RXF-631L and ACHN; stomach cancer St-4, MKN1, MKN7, MKN28, MKN45, and MKN74;
prostate cancer DU-145 and PC-3. The GI50 values for these cell lines were determined by using the
sulforhodamine B colorimetric method [9].
REFERENCES
[1] M. Kikuchi, Y. Yamauchi, C. Yanase, I. Nagaoka, Yakugaku Zasshi 1987, 107, 245.
[2] K. Ishiguro, M. Yamaki, S. Takagi, Y. Ikeda, K. Kawakami, K. Ito, T. Nose, Chem. Pharm. Bull. 1986,
34, 2375.
[3] K. Konno, C. Hirayama, H. Yasui, M. Nakamura, Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 9159.
[4] M. Guiso, C. Marra, Nat. Prod. Res. 2005, 19, 105.
[5] L. J. El-Naggar, J. L. Beal, J. Nat. Prod. 1980, 43, 649.
[6] P. Gariboldi, G. Jommi, L. Verotta, Phytochemistry 1986, 25, 865.
[7] S. P. Shrestha, Y. Narukawa, T. Takeda, J. Nat. Med. 2007, 61, 73.
[8] T. Tanahashi, Y. Takenaka, N. Okazaki, M. Koge, N. Nagakura, T. Nishi, Phytochemistry 2009, 70,
2072.
[9] T. Yamori, A. Matsunaga, S. Sato, K. Yamazaki, A. Komi, K. Ishizu, I. Mita, H. Edatsugi, Y.
Matsuba, K. Takezawa, O. Nakanishi, H. Kohno, Y. Nakajima, H. Komatsu, T. Andoh, T. Tsuruo,
Cancer Res. 1999, 59, 4042.
[10] K. D. Paull, R. H. Shoemaker, L. Hodes, A. Monks, D. A. Scudiero, L. Rubinstein, J. Plowman,
M. R. Boyd, J. Natl. Cancer Inst. 1989, 81, 1088.
Received April 6, 2010