Synthesis and evaluation of cytotoxic effects of hanultarin and its derivatives

Article abstract of DOI:10.1016/j.bmcl.2011.09.014

One of the known cytotoxic lignans is (-)-1-O-feruloyl-secoisolariciresinol designated as hanultarin, which was isolated from the seeds of Trichosanthes kirilowii. In this Letter, we described the first synthesis of 1-O-feruloyl-secoisolariciresinol, 1,4-O-diferuloyl-secoisolariceresinol and their analogues. The cytotoxicities of these compounds were evaluated against several cancer cell lines. Interestingly, we found that the feruloyl diester derivative of secoisolariciresinol was the most active cytotoxic compound against all the cancer cells tested in this experiment. The IC50 values of the1,4-O-diferuloyl-secoisolariceresinol were in the range of 7.1-9.8 lM except one cell line. In considering that both ferulic acid and secoisolariciresinol are commonly found in many plants and have no cytotoxicity, this finding is remarkable in that simple covalent bonds between the ferulic acid and secoisolariciresinol cause a cytotoxic effect.

Full text of DOI:10.1016/j.bmcl.2011.09.014

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6245–6248
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of cytotoxic effects of hanultarin and its derivatives
⇑
Eunyoung Lee, V.S. Jamal Ahamed, Mahto Sanjeev Kumar, Seog Woo Rhee, Surk-Sik Moon, In Seok Hong
Department of Chemistry, Kongju National University, 182 Shinkwandong, Gongju 314-701, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 June 2011
Revised 17 August 2011
Accepted 5 September 2011
Available online 10 September 2011
One of the known cytotoxic lignans is (ꢀ)-1-O-feruloyl-secoisolariciresinol designated as hanultarin,
which was isolated from the seeds of Trichosanthes kirilowii. In this Letter, we described the first synthesis
of 1-O-feruloyl-secoisolariciresinol, 1,4-O-diferuloyl-secoisolariceresinol and their analogues. The cyto-
toxicities of these compounds were evaluated against several cancer cell lines. Interestingly, we found
that the feruloyl diester derivative of secoisolariciresinol was the most active cytotoxic compound against
all the cancer cells tested in this experiment. The IC50 values of the1,4-O-diferuloyl-secoisolariceresinol
Keywords:
Cytotoxic lignans
Hanultarin
Cytotoxicity
1,4-O-Diferuloylsecoisolariciresinol
Immunocytochemical analysis
were in the range of 7.1–9.8 lM except one cell line. In considering that both ferulic acid and secoisola-
riciresinol are commonly found in many plants and have no cytotoxicity, this finding is remarkable in
that simple covalent bonds between the ferulic acid and secoisolariciresinol cause a cytotoxic effect.
Ó 2011 Elsevier Ltd. All rights reserved.
Lignans are polyphenols found in plants and their precursors are
found in a wide variety of plant-based foods, including seeds, whole
grains, legumes, fruits, and vegetables.1,2 Lignans have attracted
much interest over the years on account of their broad range of
biological activities such as antifungal,3 antioxidant activity,4 and
cytotoxicity.5 Secoisolariciresinol is one of the common lignans con-
tained in food plants.6,7 A racemic mixture of secoisolariciresinol is
biosynthesized in some plants.8 It has been reported that secoisola-
riciresinol was converted to enterodiol and enterolactone, which are
compounds that have been postulated to have anti-carcinogenic
properties.9,10 The Yamauchi group reported first stereo-selective
synthesis of meso-secoisolariciresinol and its biological activity
was compared with (+) and (ꢀ)-secoisolariciresinol.11 The optically
active compounds, (+) and (ꢀ)-secoisolariciresinol, stimulated the
IgM production of HB4C5 cells. However, meso-secoisolariciresinol
had no effect of IgM production-stimulating activity toward the
same cells.
In contrast to the well known biological activity of the secoisol-
ariciresinol, its simple derivatives such as an ester were not well
studied. Recently, a novel cytotoxic lignan was isolated from the
seeds of Trichosanthes kirilowii and fully characterized as shown
in Figure 1.12 The isolated lignan was identified as mono ferulic
acid ester of (ꢀ)-(2R,3R)-secoisolariciresinol and designated as
hanultarin. The known 1,4-O-diferuloyl-secoisolariciresionol was
also isolated from the seeds and fully characterized. Both com-
pounds showed a comparable cytotoxic effect to the cisplatin
against several cancer cell lines.
Interestingly, hanultarin has an inhibitory effect on actin poly-
merization, which was confirmed by an immune-fluorescence
analysis.12
In this letter, we report first synthesis of hanultarin and its ana-
logues and evaluate the cytotoxic activity of the obtained deriva-
tives against several cancer cell lines. Further, we discussed the
origin of cytotoxic effects of hanultarin and its analogues based
on the structural characteristics of these molecules.
The secoisolariciresinol compound 7 is a key intermediate for
the synthesis of hanultarin. One classical synthetic route toward
the lignin structure utilizes the direct oxidative coupling reaction
with structurally simple precursors.13,14 However, the coupling
reaction of phenol derivatives generates complex coupling prod-
ucts due to several radical intermediates.15 Stereo-selective synthe-
sis of (+) and (ꢀ)-secoisolariciresinol was achieved by multi-step
synthesis using a chiral erythro-aldol product.4,11,16
Although the first synthesis of 2,3-dibenzylidenesuccinates via
a double Stobbe condensation was achieved with the vanillin and
an unprotected phenolic group by the Brown group,17 other groups
failed to synthesize 2,3-dibenzylidenesuccinates via Stobbe con-
densation without protection of the phenol hydroxyl group.15,18
In our study, a large scale synthesis of the secoisolariciresinol
was carried out by a double Stobbe condensation, starting from a
succinate ester and the vanillin without protection of the phenolic
hydroxyl groups as shown in Scheme 1.19
Briefly, the first Stobbe reaction with the vanillin 1 and the
dimethyl succinate in the presence of lithium produced the con-
densation product 2. Compound 2 was esterified with anhydrous
MeOH in the presence of catalytic H2SO4 prior to the next Stobbe
condensation. Then, the second Stobbe condensation of 3 was car-
ried out with additional vanillin and followed esterification. Yields
⇑
Corresponding author. Tel.: +82 41 850 8498; fax: +82 41 856 8613.
E-mail address: ishong@kongju.ac.kr (I.S. Hong).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.014

6246
E. Lee et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6245–6248
MeO
HO
O
O
MeO
HO
OH
OH
OH
MeO
MeO
OH
OMe
OMe
OH
OH
TBDMSO
TBDMSO
b
a
6
MeO
MeO
MeO
OH
MeO
OH
(+)-(2S,3S)-secoisolariciresinol
(-)-(2R,3R)-secoisolariciresinol
OTBDMS
OTBDMS
8
9
O
MeO
HO
OMe
MeO
HO
MeO
O
OR1
OR2
OR1
OR2
OH
OH
TBDMSO
d
c
MeO
OH
(-)-(2R,3R)-1-O-feruloylsecoisolariciresinol
Hanultarin
MeO
MeO
OH
OTBDMS
10a R1 = O-TBDMS-feruloyl, R2 = H
10b R1 = R2 = O-TBDMS-feruloyl
11a R1 = feruloyl, R2 = H
11b R1 = R2 = feruloyl
Figure 1. Structures of ( )-secoisolariciresinol and hanultarin.
Scheme 2. Synthesis of hanultarin and its derivatives. Reagents and conditions: (a)
TBDMS-Cl, imidazole, DMF, 3 h; (b) LiAlH4, dry THF, ꢀ15 °C, 1 h; (c) DCC, DMAP, O-
TBDMS-ferulic acid, DCM, 4 h; (d) TBAF, dry THF, 2 h.
CHO
O
O
O
O
MeO
HO
MeO
HO
OMe
OH
OMe
OMe
a
b
OMe
OH
MeO
MeO
HO
OR1
OR2
homovanilloyl
OH
1
O
2
3
O
O
O
O
3-(4-hydroxy-3-methoxyphenyl)-propionyl
MeO
HO
MeO
HO
OMe
O
OMe
OH
OMe
MeO
OMe
OH
c
d
OH
12a R1 = homovanilloyl, R2 = H
MeO
12b R1 = R2 = homovanilloyl
MeO
4
OH
5
13a R1 = 3-(4-hydroxy-3-methoxyphenyl)-propionyl, R2 = H
13b R1 = R2 = 3-(4-hydroxy-3-methoxyphenyl)-propionyl
OH
O
MeO
HO
MeO
HO
OH
OH
OMe
OMe
Figure 2. Structures of mono and diester analogues of the ( )-secoisolariciresinol.
f
e
O
protected ferulic acid. As shown in Scheme 2, t-butyldimethyl-silyl
group was selected for the protection of phenolic alcohols and
conversion to the compound 8 was almost quantitative. Then,
the compound 9 was synthesized by the reduction of 8 with
LiAlH4. The DCC coupling reaction of 9 with the TBDMS-protected
ferulic acid (1.5 equiv) yielded the mixture of mono and diester
derivatives. The isolated yields of mono and diester derivatives
were 23% and 50%, respectively. Finally, TBAF treatment of 10a
produced the target hanultarin 11a. Both hanultarin (11a) and
1,4-O-diferuloyl-secoisolariciresionol (11b) were fully character-
ized by 1H NMR, 13C NMR, FT-IR and HR-Mass spectra20 and
perfectly matched to the isolated structures.12 However, the
specific optical rotation of both compounds (11a and 11b) were
determined to be almost zero, which means the formation of the
racemic mixture of hanultarin and 1,4-O-diferuloyl-secoisolaricire
sionol.
The feruloyl moiety of the hanultarin plays a key role in the
cytotoxic effect against the cancer cell lines because of the relative
low cytotoxicity of the secoisolariciresinol.12 Therefore, further
modification of the hanultarin focused on the feruloyl moiety as
shown in Figure 2. First, the feruloyl group was replaced by one
carbon short homovanilloyl moiety, which has a carboxy methyl
group bound to the phenyl ring (12a, 12b). Second modified com-
pounds were more flexible derivatives without a double bond in
the feruloyl group (13a, 13b). Mono and diester derivatives of
these analogues were prepared by the same method of the
hanultarin.
MeO
MeO
OH
7 (+)-secoisolariciresinol
OH
6
Scheme 1. Synthesis of ( )-secoisolariciresinol. Reagents and conditions: (a)
dimethyl succinate, lithium, dry MeOH, reflux, 48 h; (b) H2SO4, dry MeOH, reflux,
24 h; (c) vanillin, lithium, dry MeOH, reflux, 48 h; (d) H2SO4, dry MeOH, reflux, 24 h;
(e) 10% Pd/C (5:1 substrate/catalyst), dry MeOH, H2 gas, 24 h; (f) LiAlH4, dry THF,
ꢀ15 °C, 1 h.
of the sequential Stobbe condensation were 94% and 50%, respec-
tively. Compound 5 was subjected to the catalytic hydrogenation
with 10% Pd/C and produced the hydrogenated compound 6. In this
hydrogenation step, theoretically, three stereo isomers can be pro-
duced such as meso and one pair of enantiomer. In a similar reac-
tion carried out by the Wang group15, the hydrogenation of a
similar structure containing the t-butyl group on phenyl rings gen-
erated 40% of meso derivative and 59% of the racemic mixture.
However, the hydrogenation of 5 yielded only the racemic mixture
of 6, which was confirmed by the analysis of the 1H and 13C NMR.19
The final ( )-secoisolariciresinol was produced by the reduction of
compound 6 with LiAlH4 and the formation of racemate was also
confirmed by specific rotation, which was almost zero.19
For the synthesis of hanultarin, the phenolic alcohols of com-
pound 6 should be protected prior to the reduction to avoid gen-
eration of the side products in coupling reaction with the

E. Lee et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6245–6248
6247
Table 1
As shown in Table 1, the homovanilloyl mono ester 12a and the
3-(4-hydroxy-3-methoxyphenyl)-propionyl mono ester 13a
showed two to four times less toxicity than the feruloyl mono ester
11a. However, diester derivatives 12b and 13b were more active
than ( )-hanultarin 11a and their mono ester derivatives 12a and
13a. The diester 12b and 13b were a little less active than the feru-
lic acid diester 11b. This suggests that the cytotoxicity of these
compounds originates from the diester structure of the secoisola-
riciresinol. The diferuloyl ester derivative 11b was the most active
compound against all the cell lines. It is not common that the
Cytotoxicity assay of hanultarin and its analogues
Compounds
Cell line (IC50
,
l
M)
SKOV3d SKMel2e
>110.4 >110.4
17.3(5.6)
8.3(14.0)
>76.0
A549a
B16F10b
HCT15c
7
11a
11b
12a
12b
13a
>110.4
>110.4
>110.4
29.3
8.3
52.2
15.9
49.6
10.0
41.3
29.5(5.6)f
21.4(24.1)
9.8(16.8)
71.8
63.9
21.7
>76.0
30.8
7.1(16.8)g
65.3
17.5
65.7
12.1
13.3
68.4
12.8
14.2
>74.0
35.9
64.0
10.3
13b
Cisplatin
derivatives originating from dietary plants show the
l molar con-
37.3(33.9) 31.3(16.8)
27.3
28.0(23.5)
centration of the IC50 values against the cancer cell lines.
The diferuloyl ester derivative 11b was the most active com-
pound against all the cell lines. It is not common that the derivatives
a
b
c
d
e
f
Human small lung cancer cell.
Mouse melanoma cell.
Human colon cell.
originating from dietary plants show the
l molar concentration of
Human ovary cell.
Human skin cell.
the IC50 values against the cancer cell lines.
In parenthesis, IC50 of (ꢀ)-11a.12
An immunocytochemical analysis of A549 lung cancer cells was
carried out to determine the influence of the cellular structure on
exposure to these derivatives. After treatment of each compound
against the A549 cell lines at the dose range of the IC50 value, the
cells were sequentially exposed to microtubule, actin and nuclei-
specific antibodies.19 As shown in Figure 3, all the actin and micro-
tubule fibers were disrupted to form aggregate-like structures at a
dose of near the IC50 value of each compound, except the nucleus,
which were stable in these treatments. An inhibitory effect of the
hanultarin on actin polymerization was reported via an immune-
fluorescence analysis using normal epidermal keratinocytes.12
However, it was difficult to discriminate the inhibitory effect be-
tween antin and microtubule fibers against the A549 cells exposed
to the synthetic ( )-hanultarin.
All the diester derivatives including the diferuloyl ester of seco-
isolariciresinol showed a similar pattern to the hanultarin. At this
stage, we cannot confirm whether these cell shrinkages originated
from the process of the normal cell death or the inhibitory effect of
the polymerization of actin or microtubule fibers. It might be
proved in a specific actin or microtubule depolymerization assay
in vitro.
In parenthesis, IC50 of (ꢀ)-11b.12
g
The cytotoxicities of secoisolariciresinol, hanultarin and its ana-
logues were evaluated against human lung cancer (A549), mouse
melanoma cell (B16F10), human colon cell (HCT-15), human ovary
cell (SKOV3) and human skin cell (SKMel2) lines using a sulforhoda-
mine B (SRB) assay method.21 Triplicate experiments were carried
out against all the cell lines at a dose range of 1.56–50 l
g/mL.19
The IC50 values of each compound are summarized in Table 1.
The ( )-secoisolariciresinol 7 was inactive against all the cell
lines at the given ranges of dose concentrations.12 The IC50 values
of the ( )-hanultarin 11a were less than that of the control cis-
platin against all the cells except the SKOV3 cell. In comparison
to the isolated (ꢀ)-hanultarin, the cytotoxicities of ( )-11a were
five times less toxic against the A549, three times less against
the SKMel2 and comparable to the (ꢀ)-hanultarin against the
B16F10 cell line. This means that (+)-derivative of the ( )-hanulta-
rin is less active against all the cell lines. However, the ( )-1,
4-O-diferuloyl-secoisolariciresinol 11b was two times more active
than the isolated (ꢀ)-1,4-O-diferuloyl-secoisolariciresinol. This
indicates that (+)-11b is more cytotoxic than the (ꢀ)-11b deriva-
tive. Interestingly, the diester derivative 11b was two to four times
more active than the mono ester 11a against all the cell lines used
in this study.
In conclusion, we described the first synthesis of the feruloyl
mono ester and the diferuloyl ester of secoisolariciresinol. These
structures were fully characterized and well matched to the known
spectral data except for the formation of a racemic mixture.
Figure 3. Immunocytochemical analysis of A549 lung cell lines after treatments of hanultarin and its analogues. Microtubules and actins were labeled with anti-b tubulin
primary antibodies and rhodamine-conjugated phalloidin respectively. The nuclei were labeled with biotin-conjugated antibodies, and detected by FITC-conjugated
streptavidin. The human small lung cancer cell lines were treated with each compound at about IC50 (12.5 lg/mL).

6248
E. Lee et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6245–6248
5. Middel, O.; Woerdenbag, H. J.; Uden, W.; Oeveren, A.; Jansen, J. F. G. A.; Feringa,
B. L.; Konings, A. W. T.; Pras, N.; Kellogg, R. M. J. Med. Chem. 1995, 38, 2112.
6. Raffaelli, B.; Hoikkala, A.; Leppälä, E.; Wähälä, K. J. Chromatogr. B 2002, 777, 29.
7. Wang, L. Q. J. Chromatogr. B 2002, 777, 289.
8. Suzuki, S.; Umezawa, T.; Shimada, M. Biosci. Biotechnol. Biochem. 2002, 66, 1262.
9. Axelson, M.; Setchell, K. D. R. Fed. Eur. Biochem. Soc. Lett. 1981, 123, 337.
10. Thompson, L. U. Bailliere’s Clin. Endocrinol. Metab. 1998, 12, 705.
11. Sugahara, T.; Yamauchi, S.; Kondo, A.; Ohno, F.; Tominaga, S.; Nakashima, Y.;
Kishida, T.; Akiyama, K.; Maruyama, M. Biosci. Biotechnol. Biochem. 2007, 71,
2962.
Additionally, the analogues of the hanultarin were also prepared to
determine the structural effect on the cell toxicity. The cytotoxicity
experiments revealed that the diester derivatives were more active
than the mono ester derivatives against various cancer cell lines.
The IC50 values of the ( )-1,4-O-diferuloyl-secoisolariciresinol were
in the range of the several
l molar concentration, which is remark-
able in that simple covalent bonds between the ferulic acid and
secoisolariciresinol cause a cytotoxic effect. The immunocyto-
chemical analysis of the A549 cells after treatment of the mono
or diester derivatives showed the cell shrinkages, which might be
originated from the inhibitory effect of the polymerization of actin
or microtubule fibers.
12. Moon, S.; Rahman, A. A.; Kim, J.; Kee, S. Bioorg. Med. Chem. 2008, 16, 7264.
13. Ward, R. S. Chem. Soc. Rev. 1982, 11, 75.
14. Ward, R. S.; Hughes, D. D. Tetrahedron 2001, 57, 2057.
15. Wang, Q.; Yang, Y.; Li, Y.; Yu, W.; Hou, Z. J. Tetrahedron 2006, 62, 6107.
16. Yamauchi, S.; Sugahara, T.; Matsugi, J.; Someya, T.; Masuda, T.; Kishida, T.;
Akiyama, K.; Maruyama, M. Biosci. Biotechnol. Biochem. 2007, 71, 2283.
17. Daugan, A.; Brown, E. J. Nat. Prod. 1991, 54, 110.
Now further modification of the hanultarin derivative and a
detailed mechanistic analysis of the cytotoxicity are in progress.
18. Datta, P.; Yau, D.; Hooper, T. S.; Yvon, B. L.; Charlton, J. L. J. Org. Chem. 2001, 66,
8606.
19. See the Supplementary data for the detailed experimental.
20. Full characterization data:
Acknowledgments
( )-hanultarin (11a): IR
mmax: 3419, 2917, 2850, 1697, 1602, 1515, 1269, 1156
cmꢀ1 1H NMR (400 MHz, CD3OD) d 1.92 (1H, m), 2.19 (1H, m), 2.56 (2H, d,
:
This research was partly supported by the Basic Science
Research Program through the National Research Foundation of
Korea (NRF) funded by the Ministry of Education, Science and
Technology (2009-0071513).
J = 7.6 Hz), 2.65 (2H, dd J = 7.2, 13.6 Hz), 3.49 (1H, dd, J = 6.0, 14.4 Hz), 3.62 (1H,
dd, J = ), 3.66 (3H, s), 3.67 (3H, s), 3.82 (3H, s), 4.04 (1H, dd, J = 6.4, 11.2 Hz),
4.28 (1H, dd, J = 6.0, 11.2 Hz), 6.30 (1H, d, J = 16 Hz), 6.51 (3H, m), 6.57 (1H, d,
J = 2 Hz), 6.62 (2H, dd, J = 2, 8 Hz), 6.75 (1H, d, J = 8 Hz), 7.00 (1H, dd, J = 1.6,
8.0 Hz), 7.11 (1H, d, J = 2.0 Hz), 7.51 (1H, d, J = 16.0 Hz). 13C NMR (100 MHz,
CD3OD) 35.53, 36.01, 40.78, 56.33, 56.61, 62.87, 66.13, 111.80, 113.40, 113.53,
115.70, 116.00, 116.05, 116.63, 122.81, 122.85, 127.83, 133.39, 133.87, 145.70,
145.79, 147.00, 148.99, 149.02, 149.54, 150.80, 169.45; ESI-MS: Calcd
Supplementary data
538.2203. Found: 537.2130 [MꢀH]; observed ½a D20
ꢁ
ꢀ0.15 (c 0.1, MeOH).
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.09.014.
( )-1,4-O-di-feruloyl-secoisolariciresionol (11b): IR mmax
:
3418, 2926, 1698,
1514, 1267, 1157 cmꢀ1 1H NMR (400 MHz, CDCl3) d 2.21 (2H, m), 2.72 (4H,
;
m), 3.77 (6H, s), 3.91 (6H, s), 4.21 (2H, dd, J = 5.2, 11.2 Hz), 4.39 (2H, dd, J = 5.6,
11.2 Hz), 5.89 (2H, s), 6.28 (2H, d, J = 16.0 Hz), 6.52 (2H, d, J = 2.0 Hz), 6.61 (2H,
dd, J = 2.4, 8.0 Hz), 6.80 (2H, d, J = 8.0 Hz), 6.90 (2H, d, J = 8.0 Hz), 7.01 (2H, d,
J = 2.0 Hz), 7.06 (2H, dd, J = 1.6, 8.0 Hz), 7.58 (2H, d, J = 16.0 Hz); 13C NMR
(100 MHz, CDCl3) d 35.26, 40.12, 55.74, 55.96, 64.44, 109.44, 111.24, 114.11,
114.71, 115.18, 121.71, 123.07, 126.82, 131.69, 143.90, 145.13, 146.43, 146.76,
148.05, 167.23; ESI-MS: Calcd 714.2676. Found: 713.2604 [MꢀH]; observed
References and notes
1. Whiting, D. A. Nat. Prod. Rep. 1990, 7, 349.
2. Ward, R. S. Nat. Prod. Rep. 1993, 10, 1.
3. Figgitt, D. P.; Denyer, S. P.; Dewick, P. M.; Jackson, D. E.; Williams, P. Biochem.
Biophys. Res. Commun. 1989, 160, 257.
4. Yamauchi, S.; Hayashi, Y.; Nakashima, Y.; Kirikihira, T.; Yamada, K.; Masuda, T.
J. Nat. Prod. 2005, 68, 1459.
½ ꢁ ꢀ0.01 (c 0.1, MeOH).
a 2D0
21. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.; Vistica, D.;
Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst. 1990, 82,
1107.