Ptaquiloside Z, a new toxic unstable sesquiterpene glucoside from the neotropical bracken fern Pteridium aquilinum var. caudatum

Article abstract of DOI:10.1016/S0968-0896(98)00168-0

Reversed-phase HPLC fractionation, monitored by brine shrimp bioassay, led to the isolation of a new illudane-type sesquiterpene glucoside, ptaquiloside Z (2), as well as the known bracken carcinogen ptaquiloside (1), from a bioactive aqueous extract of the neotropical bracken fern Pteridium aquilinum var. caudatum (Pteridaceae). The structure of ptaquiloside Z (2) was confirmed by spectroscopic analyses and chemical degradation. Both compounds exhibited similar toxicity (LC₅₀ 62.5μg/ml at 24h and LC₅₀ 7.8μg/ml at 48h) toward brine shrimp. Copyright (C) 1997 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(98)00168-0

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 2229±2233
Ptaquiloside Z, a New Toxic Unstable Sesquiterpene Glucoside
from the Neotropical Bracken Fern Pteridium Aquilinum Var.
Caudatum
a,
a
b
Uvidelio Francisco Castillo, * Makoto Ojika, Miguel Alonso-Amelot
a
and Youji Sakagami
aDepartment of Molecular Biosciences, Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya 464-8601, Japan
bDepartamento de Quõ Âm ica, Facultad de Ciencias, Universidad de Los Andes, MeÂrida 5101, Venezuela
Received 29 June 1998; accepted 13 August 1998
AbstractÐReversed-phase HPLC fractionation, monitored by brine shrimp bioassay, led to the isolation of a new
illudane-type sesquiterpene glucoside, ptaquiloside Z (2), as well as the known bracken carcinogen ptaquiloside (1),
from a bioactive aqueous extract of the neotropical bracken fern Pteridium aquilinum var. caudatum (Pteridaceae). The
structure of ptaquiloside Z (2) was con®rmed by spectroscopic analyses and chemical degradation. Both compounds
exhibited similar toxicity (LC50 62.5 mg/ml at 24 h and LC50 7.8 mg/ml at 48 h) toward brine shrimp. # 1998 Elsevier
Science Ltd. All rights reserved.
Introduction
bioassay for toxic compounds, was used to test the iso-
lated fractions.17 We describe herein the isolation of a
new ptaquiloside analogue, ptaquiloside Z (2), from the
aqueous extraction of young fronds of P. aquilinum var.
caudatum. The toxicity of 2 was the same as that of 1 in
the brine shrimp assay.
Pteridium aquilinum (L. Kuhn) (Pteridaceae) is con-
sidered to be one of the most successful and widely dis-
tributed organisms of the plant kingdom.1,2 It has been
recognized that the consumption of bracken fern by
cattle induces bladder and intestinal carcinomas3,4 and
also causes a number of diseases in other farm ani-
mals.5,6 Chemical investigation of more than 30 species
Results and Discussion
7
of the Pteridaceae has been reported. An active princi-
ple responsible for the carcinogenic and mutagenic
activity was isolated simultaneously by Japanese8,9 and
Dutch10 groups as an unstable norsesquiterpene gluco-
side named ptaquiloside (1) (aquilide A) (Fig. 1). Owing
to the considerable economic importance of bracken as
a farm animal hazard11 and the discovery of the bracken
carcinogen 1 in milk from bracken-fed cows,12,13 we
were stimulated to continue searching for other toxic
components of this hazardous fern. To this end, we have
conducted additional phytochemical studies of P. aqui-
linum var. caudatum,14±16 one of the most commonly
found neotropical varieties of bracken fern in the
Northern Andes and Central America. The brine shrimp
assay, which is known to be a convenient and short-term
Puri®cation of a freeze-dried, hot-water extract of P.
aquilinum var. caudatum croziers on ODS open
column, using a water-to-acetonitrile gradient, yielded 19
fractions, from which fractions No. 5, 8, and 9 were
observed to be the most toxic in the assay. Fractions
No. 5, 8 and 9 were puri®ed by reversed-phase pre-
parative HPLC. Fractions No. 5 and 8 were found to
contain known compounds, isoptaquiloside13 and pta-
8
quiloside (1), respectively. Puri®cation of fraction No.
9 by HPLC yielded two toxic components. One of the
two was again identi®ed by NMR analysis as ptaquilo-
1
side (1) (11.5 mg, t
R
=51 min). From H and 13C NMR
spectral data (Table 1), the other component 2 (5.6 mg,
tR=72 min) was presumed to be a new analogue of 1
and was named ptaquiloside Z. The obvious di€erence
between the two compounds was that the secondary
*
Corresponding author.
968-0896/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(98)00168-0
0

2
230
U. F. Castillo et al./Bioorg. Med. Chem. 6 (1998) 2229±2233
Table 1. 13C and 1H NMR chemical shifts (d, ppm) and
coupling constants (Hz) determined for ptaquiloside Z (2) and
its tetraacetate 3.a
Ptaquiloside Z (2)
Ptaquiloside Z
tetraacetate (3)
Atom
13Cb
1Hb
1Hc
1
2
209.4 (s)
46.4 (s)
51.1 (t)
3
3
4
5
6
7
8
8
9
1
1
1
1
1
a
b
2.15 d (13.4)
2.18 d (13.4)
2.12 d (13.4)
2.16 d (13.4)
81.4 (s)
124.0 (d)
142.5 (s)
29.6 (s)
71.2 (s)
5.70 s
5.68 (s)
-OH
4.48 s
2.75 s
4.26 s
2.65 s
1.06 s
0.95 s
1.51 s
0.82 m
0.68 m
61.9 (d)
26.6 (q)
25.7 (q)
19.0 (q)
5.5 (t)
0
1.08 s
0.95 s
Figure 1. Structures of ptaquiloside (1) and ptaquiloside Z (2).
1
2
1.49 s
0.81 m
3a
3b
methyl group of 1 was replaced with a geminal dimethyl
fragment (d 0.95, 1.08) in 2. The molecular formula of 2
0.66 ddd
(9.6, 6.5, 4.1)
0.53 ddd
(8.9, 6.5, 4.1)
0.84 m
was determined to be C21
FABMS and supported that 2 was a methylated analo-
H
32
O
8
by high-resolution
14a
9.4 (t)
0.56 m
1
14b
0.86 m
1.08 s
gue of 1. Also, the following H NMR signals similar to
those of 1 were observed: two methyl signals (d 1.20 and
1
1
2
5
26.6 (q)
98.2 (d)
74.5 (d)
1.20 s
4.54 d (7.7)
3.08 dd
0
4.83 d (8.1)
4.77 dd
(9.4, 8.1)
5.14 dd
1
.49), one ole®nic signal (d 5.70), a cyclopropylidene
0
0
0
0
0
signal (d 0.53 to 0.82), an anomeric signal (d 4.54), and a
(
8.3, 7.7)
1
3
complex cluster of signals (d 3.25 to 3.28). C NMR
data showed the presence of a ketone (d 209.4) and an
anomeric carbon signal (d 98.2) as in the case of 1. The
assignment of the protonated carbons was performed by
HMQC spectrum. The connectivity of the quaternary
carbons and oxygen substituents was determined by
HMBC data as shown in Figure 2. From these ®ndings
we were able to propose a gross structure of 2 (Fig. 2).
The presence of the b-d-glucoside portion, though not
evidenced by COSY and HMBC data, was unambiguously
clari®ed by the following derivatization experiments.
Acetylation of 2 yielded ptaquiloside Z tetraacetate (3).
From the vicinal proton±proton coupling constants of
the acetylated hexose moiety of 3, we concluded that the
hexose was glucose and it attached to the aglycon via a
b-glycoside bond (Fig. 3). The tetraacetate 3 was then
subjected to acidic methanolysis to yield two indanones,
pterosin Z (4) and pterosin I (5), and methyl a- and b-d-
glucosides, which were identi®ed after acetylation and
separation as methyl 2,3,4,6-tetra-O-acetyl-a-d-gluco-
pyranoside and its b-anomer. These results also con-
3
4
5
6
77.4 (d)
3.28 m
3.25 m
3.25 m
3.75 brd
(9.6, 9.4)
4.87 dd
71.3 (d)d
(10.0, 9.6)
3.74 ddd
(10.0, 6.3, 2.4)
3.98 dd
76.5 (d)d
62.5 (t)
(
11.3)
.59 dd
11.3, 3.0)
(11.2, 6.3)
4.10 dd
(11.2, 2.4)
3
(
a
b
c
3
Measured in CD CN. Coupling constants are in parenthesis.
Measured at 150 MHz for 13C and at 600 MHz for 1 H.
Measured at 400 MHz.
Data interchangeable.
d
protons of the cyclopropane ring, H-3, and the geminal
dimethyl groups were also assigned from the NOE cor-
relations. The absolute stereochemistry of 2 was ®nally
examined by applying the octant rule to ptaquiloside Z
tetraacetate (3). We assumed that 3 possessed the same
con®guration as that of ptaquiloside tetraacetate (6),
which was determined previously by X-ray crystalo-
graphic analysis.18,19 The hydrogen atom at C-2 in the
crystal structure of 6 was thus replaced with a methyl
group and energy minimization with MM2 force ®eld
was performed to obtain a new molecular model of 3
®rmed the gross structure of 2. Utilizing NOESY and
ROESY NMR data, we were able to propose the relative
stereochemistry of the aglycon part of 2. The correlations
0
of H-9/H-1 and H-9/H-15 revealed that the ring A/B
fusion was cis, and the b-glucosyl group, H-9 and H-15
adopted the a orientation (Fig. 4). The enantiotopic

U. F. Castillo et al./Bioorg. Med. Chem. 6 (1998) 2229±2233
2231
Figure 2. Gross structure from HMBC correlations (arrows)
for ptaquiloside Z (2).
Figure 4. NOE correlations obtained from NOESY and
ROESY for ptaquiloside Z (2).
(
Fig. 5). Since there were many more atoms in the minus
areas of the back octants than in the plus areas, this
model was expected to display a negative cotton e€ect.
Actually, the CD spectrum of 3 (Fig. 6) showed a nega-
tive cotton e€ect at ca. 300 nm, which corresponds to
the n� p* transition of the ketone group (C-1). The CD
spectrum of ptaquiloside tetraacetate (6) also showed a
negative cotton e€ect (Fig. 6). From these ®ndings, we
were able to conclude that the absolute stereochemistry
of 2 was the same as that of 1. Compound 2 showed
toxicity as high as that of 1 after 24 and 48 h (LC50
Ubest-50 UV/VIS spectrometer. IR spectra were recorded
on a JASCO FT/IR-7000S. Mass spectra were recorded
on a JEOL Mstation JMS-700 mass spectrometer using
sodium iodide/glycerol as a matrix in positive mode.
One and two-dimensional (HMBC, HMQC, COSY,
NOESY and ROESY) NMR spectra were obtained for
CD CN and CDCl solutions using a Bruker AMX-600
or a Bruker ARX-400 spectrometer. Chemical shifts are
reported relative to TMS. 13C NMR signal multiplicity
was determined using DEPT 135 sequence. CD spectra
were recorded on a JASCO J-720 spectropolarimeter.
MM2 calculations in Chem3D ver. 3.5.1 were per-
formed on an Apple Macintosh 7600/200 computer.
3
3
6
2.5 mg/ml and 7.8 mg/ml, respectively) against the brine
shrimp Arteima salina LEACH.
1
7
Experimental
Plant material
Instruments
Young fronds of bracken fern were collected between
8±12 days after emergence14 at the site of ``El Cerro
Â
La Bandera'' in Merida, Venezuela, during February
1996. They were identi®ed using the key of Ortega20 as
Pteridium aquilinum var. caudatum. A voucher specimen
(UVI 95-002) has been deposited in the herbarium of the
HPLC was performed on a JASCO high-pressure gra-
dient system with PU-980 pumps and detection was
carried out using a JASCO UV-970 at 220 nm. Optical
rotations were measured with a JASCO DIP-370 digital
polarimeter. UV spectra were recorded on a JASCO
Figure 3. Ptaquiloside Z tetraacetate (3) degradation into two indanones 4 and 5 and methyl glucosides.

2
232
U. F. Castillo et al./Bioorg. Med. Chem. 6 (1998) 2229±2233
Figure 5. A plausible 3D conformation of 3 obtained from MM2 calculations. Projections were done on the plane of the keton group,
viewed from the C ˆ O axis (a) and from the side of the C ˆ O group (b). The C-13 and C-14 carbons are located in a plus area of the
front octants in (a).
ꢀ
Faculty of Pharmacy, Universidad de Los Andes, Me
Â
ida, Venezuela.
r-
� 30 C. A portion of this material (20 g) was ®rst frac-
tionated by ODS (Cosmosil 75 C18- OPN, 140 g, Naca-
lai Tesque) open-column chromatography and eluted
2 3
using 100% H O to 100% CH CN to a€ord 19 frac-
Extraction and puri®cation procedures
tions. Immediately, toxicity test was performed on each
fraction. Fraction No. 9 (47.5 mg) was further puri®ed
by ODS preparative HPLC [column: Develosil ODS-
HG-10 (10Â250 nm), Nomura Chemical] using 37:63
Fresh fronds (1600 g) were mashed in hot boiling water
using a high-power blender that reduced the plant tissue
to an amorphous mass in which no ®bers could be dis-
cerned under a mid-power scope. After blending, the
MeOH:H O as an elution solvent. From this fractiona-
2
ꢀ
mixture was cooled to 0 C, ®ltrated and centrifuged.
The supernatant was lyophilized to give a 58.4 g of solid
material. This solid material was stored with silica gel at
tion we obtained 2 compounds. The ®rst compound
R
(11.5 mg) (t =51 min) was identi®ed as ptaquiloside (1),
2
4
ꢀ
1
24
d
ꢀ
[a] � 184 (c 0.256, CH CN) [Lit. [a] � 188 (c 1.00,
d
3
8
MeOH)], H NMR (400 MHz, CD CN) d 0.46 (ddd, J
3
9.5, 6.2, 4.0 Hz, 1H), 0.63 (ddd, J=9.5, 5.8, 4.0 Hz, 1H),
0.76 (ddd, J=9.5, 6.2, 4.1 Hz, 1H), 0.84 (ddd, J=9.5,
5.8, 4.0 Hz, 1H), 1.04 (d, J=6.9 Hz, 3H), 1.21 (s, 3H),
1.49 (d, J=1.1 Hz, 3H), 1.87 (dd, J=12.4, 12.4 Hz, 1H),
2.23 (m, 1H), 2.38 (dd, J=12.4, 8.1 Hz, 1H), 2.54 (d,
J=1.3 Hz, 1H), 3.08 (dd, J=7.9, 8.2 Hz, 1H), 3.26 (m,
3
2
H), 3.58 (dd, J=11.7, 5.0 Hz, 1H), 3.76 (dd, J=11.7,
.1 Hz, 1H), 4.39 (s, OH), 4.61 (d, J=7.9 Hz, 1H), 5.68
(
s, 1H). The second component (5.6 mg) (t =72 min)
R
2
4
ꢀ
was named ptaquiloside Z (2), [a] � 168 (c 0.221,
d
CN), IR (KBr) 3407 (br), 1717, 1645 cm 1, UV
�
CH
(
4
3
3
CH CN) lmax 203 nm (e 10500), HRFABMS m/z
+
35.2005 [M+Na] , calcd for C H O Na 435.1982.
21 32 8
Ptaquiloside Z tetraacetate (3) and its degradation
Acetylation of 2 (2.06 mg) was done by standard acet-
ylation procedure (Ac O, pyridine) to yield 3 (1.64 mg,
2
5
3
6.5%): CD (CH CN) l 316 (Áe� 1.35), 313 (� 1.28),
3
ext
1
06 (� 1.77), 298 (� 1.33) nm. H NMR data are shown
Figure 6. CD spectra of 3 and 6 in CH
3
CN.
in Table 1.

U. F. Castillo et al./Bioorg. Med. Chem. 6 (1998) 2229±2233
2233
A mixture of 3 (1.64 mg) and concd. H
MeOH (1 ml) was re¯uxed for 12 h, cooled to room
temperature, neutralized with solution of 1 M
NaHCO (160 ml) and evaporated under reduced pres-
sure to give a solid. The solid was dissolved in water
3 ml) and extracted with CHCl (10 ml). The CHCl
2
SO
4
(10 ml) in
2. Page, C. M. Bot. J. Linn. Soc. 1976, 73, 1±34.
3
. Fenwick, G. R. J. Sci. Food Agric. 1995, 46, 147±173.
. Smith B. L. In Bracken Biology and Management; (Eds.)
a
4
3
Thomson, J. A.; Smith, R. T.; AIAS Occasional Publication,
Australia, 1990; No. 40, pp 227±232.
(
3
3
5. Evans, W. C. In Bracken Ecology, Land Used and Control
Technology; (Eds.) Smith, R. T.; Taylor, J. A.; Parthenon pub-
lishing, Carnforth, Lancashire, U.K; 1986, pp 121±132.
extract was concentrated to give an oily residue, which
was separated by TLC on silica gel (6:1 CCl :acetone) to
4
give two known compounds, pterosin Z (4) (0.18 mg,
1%) and pterosin I (5) (0.41 mg, 25%) which were
6
. Evans, I. A. Res. Vet. Sci. 1979, 26, 339±348.
1
7. Murakami, T; Tanaka, N. Prog. Chem. Org. Nat. Prod.
readily identi®ed by comparison of their reported NMR
spectral data.21,22 The aqueous portion was evaporated
under reduced pressure to dryness, which was dissolved
1
988, 53, 125±135.
8. Ojika, M.; Wakamatsu, K.; Niwa, H.; Yamada, K; Hirono,
I. Tetrahedron Lett. 1983, 24, 4117±4120.
. Ojika, M.; Wakamatsu, K.; Niwa, H.; Yamada, K. Tetra-
hedron 1987, 43, 5261±5274.
0. van der Hoeven, J. C. M.; Lagerweij, W. J.; Posthumus, M.
2
in a mixture of Ac O (0.5 ml) and pyridine (0.5 ml), and
9
the mixture was stirred at room temperature for 5 h and
concentrated. The resulting residue was dissolved in
1
water (3 ml) and extracted with CHCl
CHCl extract was concentrated to give an oily residue,
3
(10 ml). The
A.; van Veldhuizen, A.; Holterman, H. A. Carcinogenesis 1983,
4, 1587±1592.
3
4
which was separated by TLC on silica gel (6:1 CCl :a-
11. Alonso-Amelot, M.E.; Rodulfo-Baechler, S.; James-
Espinoza, R. Biochem. Syst. Ecol. 1995, 23, 709±716.
1
cetone) and identi®ed by H NMR and optical rotation
as methyl 2,3,4,6-tetra-O-acetyl-a-d-glucopyranoside
12. Alonso-Amelot, M. E.; Castillo, U.; De Jongh, F. Le Lait
2
4
ꢀ
20
(
+
0.37 mg, 23%), [a] +140 (c 0.019, CHCl
3
) [Lit.[a]d
ꢀ
3
130.5 (CHCl )], and its b-anomer (0.09 mg, 5%).
d
1992 73, 323±332.
23
13. Alonso-Amelot, M. E.; Castillo, U.; Smith, B. L.; Lauren,
D. R. Nature 1996, 382, 587.
14. Alonso-Amelot, M. E.; Pe
Jaimes-Espinoza, R.; Castillo, U. J. Chem. Ecol. 1992, 18,
Bioassay
Â
rez-Mena, M.; Calcagno, M. P.;
All fractions were tested for toxicity utilizing brine
shrimp (Artemia salina LEACH). The testing samples
were made by serial dilution from 256 to 8 mg/ml in each
well of a 24-well micro plate in 2 ml of arti®cial sea
water. A suspension of nauplii containing 10 organisms
1405±1420.
15. Castillo, U. F.; Wilkins, A. L.; Lauren, D. R.; Mc Naughton,
D. E.; Smith, B. L. J. Nat. Prod. 1995, 58, 1889±1891.
16. Castillo, U. F.; Wilkins, A. L.; Lauren, D. R.; Smith, B. L.;
Towers, N. R.; Alonso-Amelot, M. E.; Jaimes-Espinoza, R.
Phytochemistry 1997, 44, 901±906.
(
10 ml) was added to each well, and the covered plate
ꢀ
was incubated at 25±27 C for 48 h. Plates were observed
every 24 h under a binocular microscope (Â4) and the
number of dead (non-motile) napulii in each well was
counted. The 50% lethal concentrations (LC50) at 24
and 48 h were measured by serial dilution media.
17. Meyer, B. N.; Ferrigni, N. R.; Putman, J. E.; Jacobsen, L.
B.; Nichols, D. E.; McLaughlin, J. L. Planta Medica 1982, 45,
3
1
1±34.
8. Niwa, H.; Ojika, M.; Wakamatsu, K.; Yamada, K.; Ohba,
S.; Saito, Y.; Hirono, I.; Matsushita, K. Tetrahedron Lett.
983, 24, 5371±5372.
1
Acknowledgements
19. Ohba, S.; Saito, Y.; Hirono, I.; Niwa, H.; Ojika, M.;
Wakamatsu, K.; Yamada, K. Acta Cryst. 1984, C40, 1877±
The authors are grateful to Professor Wolfgang Kraus
È
(Universitat Hohenheim) for providing us with infor-
mation on the brine shrimp assay.
1
2
2
879.
0. Ortega, F. Biollania 1990, 7, 47.
1. Fukuoka, M.; Kuroyanagi, M.; Yoshihira, K.; Natori, S.
Chem. Pharm. Bull. 1978, 26, 2365±2385.
References
22. Hayashi, Y.; Nishizawa, M.; Hirata, S.; Sakan, T. Chem.
Lett. 1972, 375±378.
1
. Taylor, J. A. In Bracken Biology and Management; (Eds.)
Thomson, J. A.; Smith, R. T.; AIAS Occasional Publication,
Australia, 1990; No. 40, pp 3±19.
23. Dictionary of Organic Compounds, 5th Edition; Bucking-
ham, J. Ed.; Chapman and Hall, 1982; Vol. 4, 3866.