Petrosaspongiolides M-R: New potent and selective phospholipase A2 inhibitors from the new Caledonian marine sponge Petrosaspongia nigra

Article abstract of DOI:10.1021/np9704922

Five new bioactive sesterterpenes (1-5) have been isolated from the New Caledonian marine sponge Petrosaspongia nigra Bergquist and named petrosaspongiolides M-R. Their chemical structures were determined from 1D and 2D NMR studies and MS data. All compounds inhibited different preparations of phospholipase A₂ (PLA₂) by irreversibly blocking these enzymes (particularly human synovial and bee venom, see Table 3), with IC₅₀ values in the micromolar range. Interestingly, these compounds displayed a much lower activity (or no activity at all) toward porcine pancreas and Naja naja venom PLA₂ enzymes. The most potent compound, 1 (IC₅₀ 1.6 and 0.6 μM for human synovial and bee venom PLA₂ enzymes, respectively), was slightly more active than manoalide (6) (IC₅₀ 3.9 and 7.5 μM) under our experimental conditions. Compound 3 is more selective, inhibiting human synovial PLA₂ to a greater extent than bee venom PLA₂.

Full text of DOI:10.1021/np9704922

J . Nat. Prod. 1998, 61, 571-575
571
P etr osa sp on giolid es M-R: New P oten t a n d Selective P h osp h olip a se A2
In h ibitor s fr om th e New Ca led on ia n Ma r in e Sp on ge P etr osa spon gia n igr a
Antonio Randazzo,‡ Ce´cile Debitus,§ Luigi Minale,†,‡ P. Garc´ıa Pastor, Maria J . Alcaraz, M. Paya´, and
Luigi Gomez-Paloma*,‡
Dipartimento di Chimica delle Sostanze Naturali, Universita` degli Studi di Napoli “Federico II”, via D. Montesano 49, 80131
Napoli, Italy, ORSTOM, Centre de Noume´a, B.P. A5 Noume´a Cedex, New Caledonia, and Department of Pharmacology,
Faculty of Pharmacy, University of Valencia, Av. V. Andre´s Estelles s/ n 46100, Burjassot, Valencia, Spain
Received November 3, 1997
Five new bioactive sesterterpenes (1-5) have been isolated from the New Caledonian marine
sponge Petrosaspongia nigra Bergquist and named petrosaspongiolides M-R. Their chemical
structures were determined from 1D and 2D NMR studies and MS data. All compounds
inhibited different preparations of phospholipase A2 (PLA2) by irreversibly blocking these
enzymes (particularly human synovial and bee venom, see Table 3), with IC50 values in the
micromolar range. Interestingly, these compounds displayed a much lower activity (or no
activity at all) toward porcine pancreas and Naja naja venom PLA2 enzymes. The most potent
compound, 1 (IC50 1.6 and 0.6 µM for human synovial and bee venom PLA2 enzymes,
respectively), was slightly more active than manoalide (6) (IC50 3.9 and 7.5 µM) under our
experimental conditions. Compound 3 is more selective, inhibiting human synovial PLA2 to a
greater extent than bee venom PLA2.
Sesterterpenes, a group of pentaprenyl terpenoids
quite common in marine sponges of the order Dictyo-
ceratida1 have received special attention from the
scientific community after the discovery of manoalide2
(6), a potent antiinflammatory monocarbocyclic deriva-
tive (its in vitro activity is nearly equivalent to that of
hydrocortisone), that binds and blocks the enzyme
phospholipase A2 (PLA2) through the formation of a
covalent inhibitor-enzyme adduct.3 More recently, sev-
eral other sesterterpenes, like luffolide4 and cacospon-
gionolide,5 have shown similar activities. All these
antiinflammatory compounds share some structural
features, like the presence of a γ-hydroxybutenolide
moiety and often an additional masked or free aldehyde
functionality, and several investigations have been
devoted to the identification of a common pharmacoph-
ore in these derivatives.6,7
In our continuing search for bioactive substances from
New Caledonian marine organisms, we have had occa-
sion to analyze the extracts of the sponge Petrosaspongia
nigra Bergquist 1995 sp.nov. (Dictyoceratida, Spongidae)
collected off New Caledonia, from which we isolated five
new sesterterpenes (1-5) named petrosaspongiolides
M-R, along with a number of other cytotoxic sestert-
erpenes (petrosaspongiolides A-L),8 all possessing a
cheilantane skeleton. Herein we report their isolation,
structure elucidation (see Tables 1 and 2 for NMR
assignments), and the data from their in vitro tests on
the inhibition of human sinovial, Naja naja venom, bee
venom, and porcine pancreas PLA2 (Table 3).
Resu lts a n d Discu ssion
Freeze-dried specimens (1 kg, 12 kg fresh), collected
off the southern coral reef of New Caledonia were
* To whom correspondence should be addressed. Tel.: +39-81-
7486553. Fax: +39-81-7486552. E-mail: gomez@cds.unina.it.
† Deceased May 11, 1997.
§ Dipartimento di Chimica delle Sostanze Naturali, Naples, Italy.
§ ORSTOM, Noume´a, New Caledonia.
Department of Pharmacology, Valencia, Spain.
S0163-3864(97)00492-8 CCC: $15.00
© 1998 American Chemical Society and American Society of Pharmacognosy
Published on Web 04/28/1998

572 J ournal of Natural Products, 1998, Vol. 61, No. 5
Randazzo et al.
Ta ble 1. 1H (500 MHz) and 13C (125 MHz) NMR Chemical Shifts and HMBC Correlations of the Sestertepenes 1, 3, and 5
1 (CDCl3)
3 (CD3OD)
5 (CD3OD)
a
a
a
position
δH
δC
HMBCb
δH
δC
HMBCb
δH
δC
HMBCb
1
2
3
4
5
6
7
1.72-1.09
1.53-1.29
1.22-1.39
40.3
18.6
42.0
33.4
56.6
18.6
39.9
1.77-1.88
1.29-1.47
1.22-1.41
41.0
19.5
43.1
34.3
57.8
0.8-1.7
1.32-1.56
1.39-1.19
41.1
19.5
43.2
34.2
57.9
19.6
0.89
1.40-1.61
0.83-1.98
0.90
1.43-1.50
1.81
0.89
1.52-1.42
19.5 C5
41.6 C6, C8, 1.21
C9, C23 1.60
37.7
60.9
38.8
20.5 C9
C9
29.4 C11
C24
41.8
53.5
43.2 C5, C15
1.11
8
9
10
11
36.6
59.5
37.5
19.5
38.8
60.1
39.0
21.1
0.90
0.92
0.96
1.29-1.63
1.37
2.09
1.03
2.07
1.33
1.18
1.65-1.40
1.57
2.02 dd (4.3;16.8)
2.46 ddd (10.7;10.7;6.1) 49.8 C12, C14
12R
12â
13
1.08
1.63
27.8
31.9 C9
1.47 ddd (13.6;13.6;4)
1.75
46.3
37.1
14
2.11 t (10.7)
60.4 C8, C13,
C23
15R
15â
16
1.85 br d (13.0)
1.33
4.72 ddd (12.0; 1.5; 1.5) 67.3
166.7
28.4 C16
1.87
1.24
30.0 C16
6.47 dd (10.7; 17.1)
146.1 C16, C17
4.39 ddd (11.1; 2.2; 2.2) 72.2 C13, C14 6.36 d (17.1)
163.7
122.9 C14, C15, C17, C18, C25
164.4
17
18
19
20
21
22
23
6.05 d (1.5)
118.6 C16, C19 5.05 d (3.1)
117.8
173.1
33.7
21.8
16.8
5.86 s
115.5 C16, C17, C19
174.1
33.8 C3, C4, C5, C20
21.9 C3
16.4
170.5
33.4
21.4
14.6
16.4
0.85 s
0.82 s
0.84 s
0.83 s
0.85 s
0.92 s
0.94 s
0.93 s
0.88 s
0.94 s
0.79 s
1.03 s
15.3
16.9
24
25
6.01 d (4.0)
6.20 br s
93.9 C13, C16 4.37 d (8.1)
101.7 C12, C16
94.0
183.5
100.1 C16, C19
102.7
21.1
6.15 br s
6.24 s
CH3CO 2.11 s
CH3CO
170.5
a
b
2,3
Coupling costants are in parentheses and given in Hz.1H assignments aided by COSY experiments. HMBC optimized for
J
)
CH
10.0 Hz.
sequentially extracted with n-hexane and dichloro-
hemiacetal oxygens connecting C-24 and C-16. The
methane in a Soxhlet apparatus and then with metha-
nol at room temperature. The dichloromethane-soluble
material was fractionated by Si gel flash chromatogra-
phy (MPLC) eluting with CHCl3-MeOH mixtures with
increasing amounts of MeOH, followed by reversed-
phase HPLC to yield petrosaspongiolides M-R (1-5)
(yields: 58.3 mg, 2.7 mg, 47.8 mg, 2.5 mg, and 3.2 mg,
respectively).
presence of γ-hydroxybutenolide moiety was indicated
by the proton signals at δH 6.20 (1H, br s, H-25) and δH
6.05 (1H, d, H-18) and by the broad carbon resonances9
at δC 102.7 (d, C-25), 118.6 (d, C-18), 166.7 (s, C-17),
and 170.5 (s, C-19). The connection of the quaternary
C-17 of the γ-hydroxybutenolide portion to ring D was
supported by an intense HMBC cross peak between
H-18 and C-16 and by the long-range coupling between
H-18 and H-16. The S* configuration at C-24 (24R
acetoxy group) was assigned on the basis of the small
coupling constant observed between the axial H-13 and
H-24 (J ) 4.0 Hz), implying that the latter is an
equatorial proton. In the same way, the shape of H-16
(dt, J ) 12.0, 1.5 Hz) indicated that it must be an axial
proton. Finally, these relative stereochemical assign-
ments (13S*, 14R*, 16R*, 24S*) were also corroborated
by NOE effects observed between H-16 and H-14 and
between H-13 and H-24, hence completing the structure
elucidation of 1. During the preparation of this paper
we found that De Rosa et al. had isolated the 24-epimer
of petrosaspongiolide M, named cavernosolide, from the
Mediterranean sponge Fasciospongia cavernosa.10
Petrosaspongiolide N (2), had the composition C29H42O8
Petrosaspongiolide M (1), a white amorphous solid,
showed a weak parent peak at m/z 460 (M+) in the
EIMS together with two intense fragment peaks at m/z
400 (C25H36O4, M+ - CH3COOH) and m/z 191 (C14H23+).
The latter fragment, typically observed in terpenoids
that possess a tricyclic carbon framework in their
structure,8 suggested that 1 is a terpene. 1H and 13C
NMR spectra9 in CDCl3 also confirmed the terpenoid
nature of 1, with four methyl singlet signals (δH 0.82,
0.83, 0.84, and 0.85), a complex array of overlapping
signals between δH 1.0 and 2.0, and an acetyl group (δH
2.11, 3H, s; δC 21.1 and 170.5). Inspection of COSY,
HMQC, and HMBC spectra allowed us to obtain full
assignments for 1 in CDCl3 solution (Table 1). COSY
correlations showed connectivities for the fragment C-9/
C-18 (with H-16 and H-18 sharing a long-range coupling
through the quaternary carbon C-17) and C-16/C-24
through C-15/C-14/C-13, indicating that a fourth cycle
(ring D) was fused to the tricyclic terpenoid system. A
key HMBC correlation between H-24 (δH 6.01) and the
oxygen-bearing C-16 (δC 67.3), together with the lowfield
chemical shift of C-24 (δC 93.9), completed ring D
characterization. Thus, an O-acetyl-hemiacetal func-
tionality was present at C-24 with one of the two
1
(HREIMS). Its H and 13C NMR spectra (CDCl3, 500
and 125 MHz) clearly showed that 2 is closely related
to 1, with an additional acetyl group (δH 2.01, 3H, s; δC
20.9 and 171.3) and one tertiary methyl proton signal
replaced by an AB system at δH 3.92 and 4.20 (2H, AB
system, J ) 9.2 Hz, H2-21). These data can be explained
by 2 being an acetoxy-derivative of 1, in agreement with
the molecular weight difference of 58 mass units ob-
served between 1 and 2. The stereospecific location of

Petrosaspongiolides M-R
J ournal of Natural Products, 1998, Vol. 61, No. 5 573
Ta ble 2. Selected 1H (500 MHz) and 13C (125 MHz) NMR
Chemical Shifts and HMBC Correlations of the Sestertepenes
2, 3c, and 4
hemiacetal function was present in the molecule. NMR
data also supported this view with two lowfield proton
(δH 4.37 and 6.15 ppm) and carbon resonances (δC 101.7
and 94.0 ppm) in the spectrum of 3. Analysis of both
proton-proton and proton-carbon connectivities of 3
and 3c through the aid of COSY, HMQC, and HMBC
experiments (Table 1) revealed close structural similari-
ties between 3 and 1. The free hemiacetal function (the
masked aldehyde) was thus located at C-13, as expected,
and C-16 was an oxygenated methine (δC 72.2) con-
nected to the usual γ-hydroxybutenolide unit. Interest-
ingly, this system gave rise to an HMBC cross peak
between H-16 (δH 4.39) and C-13 (δC 41.8), indicating
2 (CDCl3) 4 (CD3OD)
3c (CDCl3)
a
position
δC
δC
δH
δC
HMBCb
1
39.9
18.1
36.6
37.0
57.2
18.6
40.7
37.1
59.6
37.5
19.7
40.6
20.8
37.4
37.6
58.4
20.8
42.0
36.6
61.1
38.2
61.1
1.77-0.75
1.29-1.54
1.10-1.41
39.9
18.5
42.0
33.3
56.6
18.3
40.2
37.6
59.6
36.4
19.5
2
3c
4
5
6
7
8
9
10
11
0.80
1.43-1.50
0.80-1.65
0.81
4
the presence of an unusual J CH coupling.
1.37
2.09
1.03
2.07
1.47
1.20
1.33
1.75
4.38
These data are consistent with 3 being the 24-O-
desacetyl derivative of 1, as confirmed by acetylation
of 1 and 3 in dry pyridine (room temperature) and
subsequent comparative 1H NMR analysis of the ac-
etates (see Experimental Section). Concerning the
stereochemistry of 3, both 24-epimers were present in
solution (as in the case of the 25-epimers in the
butenolide portion of these molecules). NMR analysis
of intact 3 indicated that the 24S* epimer (24â hydroxy
group) is the major form present at the equilibrium.12
This latter stereochemistry was assigned on the basis
of the large proton-proton coupling constant (J ) 8.1
Hz) between the axial H-13 and H-24 (H-24 is axial)
and was supported by an intense ROESY (τm ) 400 ms)
cross peak between H-24 and H-12R (δH 1.03 ppm).
Similar arguments led to assignment of the stereochem-
istry of C-16 as 16R*. In fact, the pattern of dipolar
effects around rings C/D (i.e., ROESY cross peaks H-14/
H-24, H-14/H-12R, H-24/H-12R, H-16/H-14) indicated
the axial nature of H-16.
12
27.7
19.2
27.7
13
14
15
46.3
37.5
28.4
41.0
53.7
29.6
46.2
52.1
28.9 C17
16
17
18
19
20
21
22
23
24
67.2
166.7
118.0
172.1
27.3
66.9
14.5
16.9
93.7
72.1
166.1
117.8
173.0
27.6
68.3
15.4
17.3
101.9
71.2 C17
165.7
6.13 d (3.8) 118.4 C19, C25
169.6
0.85 s
0.80 s
0.82 s
33.3
21.4
14.5
16.4
0.79 s
5.35 d (8.3)
93.4 C13, C16,
CH3CO
92.4 C19
169.3
20.7
168.8
25
97.0
170.6
21.1
171.3
20.9
95.8
6.90 s
2.17
CH3CO
CH3CO
CH3CO
CH3CO
173.3
20.8
2.12
21.0
a
Coupling costants are in parentheses and given in Hz.1H
b
assignments aided by COSY experiments. HMBC optimized for
J
to the positions that are directly affected by the presence of a 21-
acetoxyl group.
) 10.0 Hz. c 13C NMR chemical shifts in bold are referred
2,3
CH
Petrosaspongiolide Q (4), had the composition C27H40O7
(HREIMS and negative ion FAB). Both NMR and MS
spectral data indicated that 4 is a 21-acetoxy-derivative
of 3. As already discussed for 2, the stereospecific
location of this additional acetoxy group at C-21 was
evident from the 13C NMR data of 4 by comparison with
those observed for compound 3 (chemical shift values
for C-3, C-4, C-5, Me-20, and C-21 are particularly
diagnostic; Table 2).11 As in 2, this stereochemical
assignment was confirmed by a ROESY peak between
H2-21 and the angular Me-22.
the acetoxy group at C-21 (i.e., the â methyl carbon
attached to C-4) was inferred by the replacement of the
C-21 methyl signal at δC 21.4 with a CH2O- carbon
resonance at δC 66.9 and by the chemical shift differ-
ences observed in the 13C NMR spectrum for C-3, C-4,
C-5, and Me-20 (Table 2)sall pointing to an oxidation
of the â geminal methyl group at C-4.8,11 Further
evidence for this stereospecific assignment comes from
ROESY (τm ) 400 ms) data that showed dipolar coupling
between H2-21 and the angular Me-22.
Petrosaspongiolide R (5) had the composition C25H36O5
as determined by negative-ion mode FABMS m/z 415,
13C- and DEPT-135 NMR measurement. The strong
peak at m/z 415 [(M - H)-] observed in the FABMS
(negative ion mode), together with a quaternary carbon
resonance at δC 183.5 in the 13C NMR spectrum (CD3-
OD, 125 MHz), indicates that 5 is a carboxylic acid.
COSY, HMQC, and HMBC data (see Table 1) were
consistent with the presence of the usual tricyclic carbon
system and the γ-hydroxybutenolide ring. In particular,
the R,â-unsaturated γ-lactone ring was further conju-
gated in 5 with a C-15/C-16 trans double bond [UV
(MeOH) λmax (log ꢀ) 264 (4.15) and NMR δH 6.47, dd, J
) 17.1 and 10.7 Hz; 6.36, d, J ) 17.1 Hz for H-15 and
H-16 and δC 146.1, 122.9 for C-15, C-16, respectively].
The location of the carboxylic functionality at C-13 was
documented by the chemical shifts of H-13 at δH 2.46
and C-13 at δC 49.8. Finally, the stereochemistry at
C-13 and C-14 (13S*, 14R*) was based on the large
Petrosaspongiolide P (3), had the composition C25H38O5
1
(HREIMS and negative ion FAB). Analysis of the H
and 13C NMR spectra of 3 revealed the presence of at
least two interconverting species. Unfortunately, in this
case, variable temperature experiments did not help
because, in contrast with what was observed for 1,9 no
significant sharpening of proton NMR peaks could be
observed even at 330 K. Acetylation of 3 in pyridine at
room temperature yielded two monoacetylated (3a and
3b) and four diacetylated (3c, 3d , 3e, and 3f) derivatives
(see Experimental Section). The diacetylated products
displayed no resonance broadening in their NMR spec-
tra at room temperature, and therefore the major
compound 3c was selected for spectral studies (Table
2). The behavior of 3 in solution could then be explained
by assuming that, besides the typical hemiacetal func-
tionality of the γ-hydroxybutenolide ring, an additional

574 J ournal of Natural Products, 1998, Vol. 61, No. 5
Randazzo et al.
a
Ta ble 3. Effect of Compounds 1-5 on Secretory PLA2
bee venom
N. naja venom
porcine pancreas
human synovial
% I (10 µM)
IC50 (µM)
% I (10 µM)
% I (10 µM)
% I (10 µM)
IC50(µM)
1
2
3
3a
3c
4
71.0 ( 1.8c
43.9 ( 2.2c
37.9 ( 3.2c
34.8 ( 2.5b
10.6 ( 3.5
12.5 ( 2.1
18.8 ( 5.6
0.6
11.5 ( 3.0
6.8 ( 3.0
3.0 ( 1.3
5.6 ( 2.8
6.9 ( 3.2
4.2 ( 2.7
1.0 ( 0.8
12.3 ( 6.0
11.6 ( 1.8
0.0 ( 0.0
0.0 ( 0.0
0.0 ( 0.0
0.0 ( 0.0
0.8 ( 0.8
68.6 ( 2.7c
44.0 ( 2.7c
60.9 ( 4.4c
59.0 ( 2.4c
32.1 ( 2.4b
30.1 ( 3.7b
7.1 ( 3.2
1.6
N.D.
3.8
N.D.d
N.D.
N.D.
N.D.
N.D.
N.D.
3.7
N.D.
N.D.
N.D.
5
a
b
d
Mean ( S.E.M. (n ) 6). p < 0.05. c p < 0.01. N.D. ) not determined for those compounds that did not reach 50% inhibition at 10
µM. Manoalide showed IC50 values of 7.5 and 3.9 µM for bee venom and human synovial PLA2, respectively.
proton-proton coupling constant between H-13 and
H-14 (J ) 10.7 Hz).
example, compound 3c), confirming again the participa-
tion of this portion in the inhibition of PLA2.
A thorough literature seach revealed that pharma-
cological data relative to compounds structurally related
to manoalide, luffolide, and their analogues are often
not easily comparable, due to the high variability of
experimental conditions used in different laboratories
for measuring inhibition of PLA2 enzymes.13 In our
case, the IC50 values obtained for compounds 1-5 can
be compared with those that we measured for manolide
(6) under the same experimental conditions.
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. NMR spec-
tra: Bruker AMX-500 (1H at 500 MHz, 13C at 125 MHz),
δ (ppm), J in Hz, spectra referred to CDCl3 and CD3OD
as internal standard; EIMS and FABMS [in glycerol or
glycerol-thioglycerol (3:1) matrix; Cs+ ions bombard-
ment] were obtained on VG AUTOSPEC mass spec-
trometer; optical rotations were measured on a Perkin-
Elmer 141 polarimeter; reversed-phase HPLC, C18
µ-Bondapak column (300 × 7.8 mm i.d.; flow rate 5 mL
min-1) Waters model 6000 A or 512 pump equipped with
U6K injector and a differential refractometer, model
401.
Biologica l Ma ter ia l. Petrosaspongia nigra (order
Dictyoceratida, family Spongidae) was collected in
November 1987, and December 1988, in the waters of
the southern coast of New Caledonia. Taxonomic
identification was performed by Dr. P. R. Bergquist, and
reference specimens are on file (ref 321) at the ORSTOM
Center of Noume´a. A specimen of Petrosaspongia nigra
is kept at the Queensland Museum in Brisbane (QMG
304685).
As already pointed out, extensive studies led to the
identification of a pharmacophore model that relates
chemical functionalities of these compounds with their
antiinflammatory properties.6,7 Our biological data are
in agreement with what has been observed for similar
compounds, both active2,4,5 and inactive14 and bring
some new information to the current pharmacophoric
model. For example, the importance of the tricyclic
system (corresponding to the linear portion in manoalide)
that allows a large hydrophobic area of contact in the
enzyme-inhibitor recognition process7 is confirmed by
the significantly lower activity of our 21-acetoxy deriva-
tives 2 and 4. This result would suggest that substitu-
tion in this position leads to a fall of bioactivity,
presumably because of the increased steric demand or
the greater polarity of such compounds or both. The
aldehyde functionality at C-24, either free or masked
(hemiacetal), is also known to be a key structural
feature for the bioactivity. This is indicated not only
by the high activity of luffolide (7) possessing a free
aldehyde at C-24, and correspondingly of petrosaspon-
giolide M (1), but also by the significant fall of activity
of the acid petrosaspongiolide R (5) and the spongiono-
lides A, B, and C,14 which possess an alcoholic function
at C-24. Such a pharmacophore model, though, does
not explain the high activity of cacospongionolide,5 a
molecule characterized by an ethereal functionality
replacing the usual hemiacetal. In addition, the higher
degree of selectivity displayed by compounds 3 and 3a
toward human synovial PLA2 (they inhibit human
synovial PLA2 to a greater extent than they do bee
venom PLA2) is also difficult to explain in terms of this
model.
Isola tion . The organism was freeze-dried and the
lyophilized material (1.0 kg) was extracted with n-
hexane and CH2Cl2 in a Soxhlet apparatus, and then
with MeOH (3 × 2 L) at room temperature. The CH2-
Cl2 extracts were filtered and concentrated under
reduced pressure to give 30.0 g of a brown oil. The crude
CH2Cl2 extract was chromatographated by MPLC on a
Si gel column (150 × 2 g) using a solvent gradient from
CHCl3 to CHCl3-MeOH 9:1. MPLC fractions were
further purified by HPLC on a semipreparative (7.8 ×
300 mm) µ-bondapak-C18 column (flow rate 5 mL/min)
eluting with MeOH-H2O mixtures to afford pure com-
pounds 1-5. The purity of each compound was judged
1
to be > 90% by HPLC and H NMR.
P etr osa sp on giolid e M (1): white amorphous solid,
[R]D -28.8° (c 0.02, CHCl3); UV (CH3OH) λmax (log ꢀ)
224 (3,58); IR (CHCl3) νmax 3330 (br), 1785, 1750, 1740,
1230 cm-1; tR 14.3 min, eluting with MeOH-H2O 85:
1
15; H and 13C NMR, see text and Table 1; EIMS m/z
460 (M+, C27H40O6), 400 (M+ - CH3COOH, C25H36O4),
m/z 191 (C14H23 +); HREIMS m/z 400.2623 (M+ - CH3-
COOH; calcd for C25H36O4, 400.2613).
The low (but still measurable) activity of petrosaspon-
giolide R (5) could be interpreted on the basis of the
known fact that the γ -hydroxybutenolide moiety alone
(which, interestingly is also a masked aldehyde) does
not allow a covalent ligand-enzyme interaction.6 In
addition, the blockade of the γ-hydroxyl group in the
γ-lactone ring results in a lower activity (see, for
P etr osa sp on giolid e N (2): white amorphous solid,
[R]D -23.0° (c 0.001, CHCl3); tR 5.7 min, eluting with
MeOH-H2O 85:15; 1H and 13C NMR, see text and Table
2; HREIMS m/z 518.2863 (M+, calcd for C29H42O8,
518.2879).
P etr osa sp on giolid e P (3): white amorphous solid,
[R]D +13.8° (c 0.001, CH3OH); tR 14.8 min, eluting with

Petrosaspongiolides M-R
J ournal of Natural Products, 1998, Vol. 61, No. 5 575
1
MeOH-H2O 80:20; 1H and 13C NMR, see text and Table
1; HREIMS m/z 418.2734 (M+; calcd for C25H38O5,
418.2719).
P etr osa sp on giolid e Q (4): white amorphous solid,
[R]D +5.8° (c 0.001, CH3OH); tR 14.0 min, eluting with
MeOH-H2O 70:30 (analytical column); 1H and 13C
NMR, see text and Table 2; HREIMS m/z 476.2787 (M+,
calcd for C27H40O7, 476.2774).
P etr osa sp on giolid e R (5): white amorphous solid,
[R]D -15.6° (c 0.003, CH3OH); UV (MeOH) λmax (log ꢀ)
264 (4.15); tR 3.8 min, eluting with MeOH-H2O 85:15;
1H and 13C NMR, see text and Table 1.
Aceta te 3e: For H NMR data (CDCl3), see acetate
1a .
Aceta te 3f: For 1H NMR data (CDCl3), see acetate
1b.
P h a r m a cologica l Assa ys. For pharmacolocical as-
says see R. Cholbi et al.15
Ack n ow led gm en t. This contribution is part of the
Marine Science and Technology (MAST-III) European
project named Bioactive Marine Natural Products in the
Field of Antitumoral, Antiviral, and Immunomodulant
Activity. Therefore, we thank the European Community
for financial support (contract MAS3-CT95-0032). We
wish to thank Prof. Vincenzo Amico (University of
Catania, Italy) for insightful suggestions and criticisms
to this work. We also thank the undergraduate student
Fabio Dell’Aquila for his kind collaboration. The work
of P. G. P., M. J . A. and M. P. was supported by the
grant SAF95-1046 (CICYT). NMR and MS spectra were
recorded at C.R.I.A.S. “Centro Interdipartimentale di
Analisi Strumentale”, Faculty of Pharmacy, University
of Naples “Federico II”.
Acetyla tion of P etr osa sp on giolid e M (1). A solu-
tion of petrosaspongiolide M (1, 6 mg) was kept in dry
pyridine (0.5 mL) and Ac2O (0.1 mL) with stirring at
room temperature for 12 h. After quenching with CH3-
OH, the excess reagents were removed under reduced
pressure to obtain a crude oil containing two products,
1a and 1b, which were separated by HPLC on diol Si
gel column (Phenomenex Spherex 5 diol), petroleum
ether-ethylic ether 80:20 as eluent, to obtain acetates
1a (3.9 mg) and 1b (2.4 mg).
1
Aceta te 1a : H NMR (CDCl3) δ 6.93 (1H, s, H-25),
6.12 (1H, br s, H-18), 6.07 (1H, d, J ) 4.0, H-24), 4.59
(1H, br d, H-16), 2.11 (3H, s, COCH3), 2.18 (3H, s,
COCH3), 0.87 (3H, s, Me-23), 0.86 (3H, s, Me-20), 0.86
(3H, s, Me-22), 0.84 (3H, s, Me-21).
Refer en ces a n d Notes
(1) (a) Crews, P.; Naylor, S. Prog. Chem. Org. Nat. Prod. 1985, 48,
203. (b) Hanson, J . R. Nat. Prod. Rep. 1986, 3, 123-132; Nat.
Prod. Rep. 1992, 9, 481-489; Nat. Prod. Rep. 1996,13, 529-
535. (c) Faulkner, D. J . Nat. Prod. Rep. 1996, 13, 75-125, and
earlier reviews cited therein.
1
Aceta te 1b: H NMR (CDCl3) δ 6.98 (1H, s, H-25),
6.05 (1H, br s, H-18), 5.93 (1H, d, J ) 4.0, H-24), 4.61
(1H, br dd, H-16), 2.12 (3H, s, COCH3), 2.10 (3H, s,
COCH3), 0.86 (3H, s, Me-23), 0.86 (3H, s, Me-22), 0.82
(3H, s, Me-20), 0.80 (3H, s, Me-21).
(2) de Silva, E. D.; Scheuer, P. J . Tetrahedron Lett. 1980, 21, 1611-
1612.
(3) (a) Glaser, K. B.; J acobs, R. S. Biochem. Pharm. 1986, 35, 449.
(b) Glaser, K. B.; J acobs, R. S. Biochem. Pharm. 1987, 36, 2079-
2086.
(4) Kernan, M. R.; Faulkner, D. J .; Parkanyi, L.; Clardy, J .; de
Carvalho, M. S.; J acobs, R. S. Experientia 1989, 45, 388-390.
(5) De Rosa, S.; De Stefano, S.; Zavodnik, N. J . Org. Chem. 1988,
53, 5020.
(6) (a) Glaser, K. B.; de Carvalho, M. S.; J acobs, R. S.; Kernan, M.
R.; Faulkner, D. J . Mol. Pharmacol. 1989, 36, 782-788. (b)
Albizati, K. F.; Holman, T.; Faulkner, D. J .; Glaser, K. B.; J acobs,
R. S. Experientia 1987, 43, 949.
(7) Potts, B. C. M.; Faulkner, D. J .; de Carvalho, M. S.; J acobs, R.
S. J . Am. Chem. Soc. 1992, 114, 5093-5100.
Acetyla tion of P etr osa sp on giolid e P (3). A solu-
tion of petrosaspongiolide P (3, 23.0 mg) was kept in
dry pyridine (1 mL) and Ac2O (0.4 mL) with stirring at
room temperature for 12 h. After quenching with CH3-
OH, the excess reagents were removed under reduced
pressure to obtain a crude oil containing two monoac-
etates, 3a and 3b, and four diasteromeric diacetates,
3c, 3d , 3e and 3f, which were separated by HPLC on
diol Si gel column (Phenomenex Spherex 5 diol), petro-
leum ether-ethylic ether 80:20 as eluent, to obtain
acetates 3a (1.5 mg), 3b (0.5 mg), 3c (12.0 mg), 3d (6.1
mg), 3e (1.9 mg), and 3f (0.3 mg). The monoacetates
3a and 3b (24 epimers) possess 24R* and 24S* configu-
rations, respectively. These compounds still have a
hemiacetal function at C-25, so the configuration at this
carbon cannot be assigned. As for the diacetates, each
pair of compounds (3c/3d and 3e/3f) contains deriva-
tives with configurations that are homogeneous at C-24
(24R* and 24S *, respectively) and epimeric at C-25.
Aceta te 3a : 1H NMR (CDCl3) δ 6.11 (1H, br s, H-25),
6.02 (1H, br s, H-18), 5.32 (1H, d, J ) 8.6, H-24), 4.45
(1H, br d, H-16), 2.13 (3H, s, COCH3), 0.85 (3H, s, Me-
21), 0.85 (3H, s, Me-23), 0.83 (3H, s, Me-22), 0.80 (3H,
s, Me-20).
(8) (a) Lal, A. R.; Cambie, R. C.; Rickard, C. E. F.; Bergquist, P. R.
Tetrahedron Lett. 1994, 35, 2603-2606. (b) Gomez-Paloma, L.;
Randazzo, A.; Minale, L.; Debitus, C.; Roussakis, C.; Tetrahedron
1997, 53, 10451-10458.
(9) Both 1H and 13C NMR spectra of 1 contained several broad
resonances, because of a slow interconversion of the two possible
C-25 epimers at an intermediate rate on the NMR time scale.
This behavior has been already observed for other related
compounds (see, for instance, Butler, M. S.; Capon, R. J . Aust.
J . Chem. 1992, 45, 1705). All 1H and 13C NMR assignments have
been confirmed through variable temperature NMR experiments
in pyridine-d5.
(10) De Rosa, S.; Crispino, A.; De Giulio, A.; Iodice, C.; Tommonaro,
G.; J . Nat. Prod. 1997, 60, 844-846.
(11) Fang, J . M.; Lang, C.I.; Chen, W. L. Cheng, Y. S. Phytochemistry
1991, 30, 2793-2795.
(12) The reader should note that, due to a difference in C. I. P.
priorities, opposite configurations at C-24 in compounds 1 (24R
acetoxy group) and 3 (24â hydroxyl group) lead to the same 24S*
stereochemical notation. Moreover, the presence in 3 of two
interconverting C-24 epimers (via the free aldehyde) was also
suggested from the spectral data of the diacetylated products
3c and 3e. A comparative 1H NMR analysis, in fact, revealed
that 3c and 3e were identical to the acetylated derivatives of
cavernosolide (the 24-epimer of 1) and 1, respectively.
(13) For instance, depending on how long the enzyme is preincubated
with the inhibitor, IC50 can vary as much as 1000 times.
Therefore, data are fully comparable only if produced under the
same experimental conditions.
Aceta te 3b: 1H NMR (CDCl3) δ 6.20 (1H, br s, H-25),
6.06 (1H, d, J ) 1.5, H-18), 6.01 (1H, d, J ) 4.0, H-24),
4,73 (1H, ddd, J ) 1.5, 1.5, 12.5, H-16), 2.10 (3H, s,
COCH3), 0.85 (3H, s, Me-20), 0.84 (3H, s, Me-22), 0.83
(3H, s, Me-23), 0.82 (3H, s, Me-21).
(14) He, H.; Kulanthaiavel, P.; Baker, B. J . Tetrahedron Lett. 1994,
35, 7189-7192.
(15) Cholbi, R.; Ferrandiz, M. L.; Terencio, C.; De Rosa, S.; Alcaraz,
M. J .; Paya, M. Naunyn-Schmiedeberg’s Arch. Pharmacol. 1996,
354, 677-683.
1
Aceta te 3d : H NMR (CDCl3) δ 7.02 (1H, s, H-25),
6.02 (1H, br s, H-18), 5.22 (1H, d, J ) 8.7, H-24), 4.35
(1H, dd, J ) 11.4, 1.5, H-16), 2.13 (3H, s, COCH3), 2.10
(3H, s, COCH3), 0.86 (3H, s, Me-23), 0.86 (3H, s, Me-
20), 0.83 (3H, s, Me-22), 0.80 (3H, s, Me-21).
NP9704922