Theopalauamide, a Bicyclic Glycopeptide from Filamentous Bacterial Symbionts of the Lithistid Sponge Theonella swinhoei from Palau and Mozambique

Article abstract of DOI:10.1021/jo9718455

Theopalauamide (1) is the major bicyclic peptide from the symbiotic filamentous eubacteria that are found in the interior of Theonella swinhoei from Palau. It was also isolated from T. swinhoei from Mozambique. The structure of theopalauamide was determined by analysis of spectroscopic data, and its stereochemistry was determined by chemical degradation and analysis of the products by chiral GC-MS. The minor peptide, isotheopalauamide (2), was shown to be a stable conformational isomer that was formed by TFA-catalyzed equilibration during the isolation procedure.

Full text of DOI:10.1021/jo9718455

1254
J . Org. Chem. 1998, 63, 1254-1258
Th eop a la u a m id e, a Bicyclic Glycop ep tid e fr om F ila m en tou s
Ba cter ia l Sym bion ts of th e Lith istid Sp on ge Th eon ella swin h oei
fr om P a la u a n d Moza m biqu e
Eric W. Schmidt, Carole A. Bewley, and D. J ohn Faulkner*
Scripps Institution of Oceanography, University of California at San Diego,
La J olla, California 92093-0212
Received October 6, 1997
Theopalauamide (1) is the major bicyclic peptide from the symbiotic filamentous eubacteria that
are found in the interior of Theonella swinhoei from Palau. It was also isolated from T. swinhoei
from Mozambique. The structure of theopalauamide was determined by analysis of spectroscopic
data, and its stereochemistry was determined by chemical degradation and analysis of the products
by chiral GC-MS. The minor peptide, isotheopalauamide (2), was shown to be a stable conforma-
tional isomer that was formed by TFA-catalyzed equilibration during the isolation procedure.
The metabolites of lithistid sponges, which include the
A lyophilized specimen of T. swinhoei from Palau was
extracted sequentially with hexanes, dichloromethane,
ethyl acetate, and 1:1 acetonitrile/water, as described
previously.4 The aqueous acetonitrile extract was par-
tially evaporated to obtain a peptide fraction as a white
solid. Initially, the peptides were purified by chroma-
tography on a reversed phase C18 silica Sep Pak column
using an acetonitrile/0.05% aqueous TFA gradient fol-
lowed by reversed-phase HPLC again using acetonitrile/
0.05% aqueous TFA as eluant to obtain both theopalaua-
mide (1) and isotheopalauamide (2). We then realized
that theopalauamide (1) was undergoing isomerization
during the purification process. We suspected that the
isomerization might be acid-catalyzed but elimination of
TFA from the HPLC eluant gave broader peaks and
genera Theonella, Discodermia, Aciculites, Microsclero-
derma, and Callipelta, are among the most diverse found
in any order of sponges and have often been attributed
to symbiotic microorganisms.1 Our studies of the cellular
location of the metabolites of Theonella swinhoei from
Palau led to the discovery that the cytotoxin swinholide
A was associated with a mixture of unicellular bacteria
and that the major peptide, which was referred to as
P951, was located in filamentous eubacteria that are
found only in the interior of the sponge.2 At the time
that the localization studies were performed, the struc-
ture of P951, now called theopalauamide (1), was tenta-
tively assigned but the presence of a minor peptide
complicated the assignment. We now present the struc-
tural elucidation of theopalauamide (1) and define its
relationship with the minor peptide isotheopalauamide
(2).3
rendered preparative-scale separation impractical.
A
compromise separation scheme used on the Mozambique
specimen employed an aqueous acetonitrile gradient as
eluant for reversed-phase Sep Pak separations and 38%
acetonitrile in 0.01% aqueous TFA for the final reversed-
phase HPLC separation to obtain almost pure theopa-
lauamide (1), in which only a trace of 2 could be detected
1
by H NMR spectroscopy (see below). The same condi-
tions were used on a 1:1 mixture of peptides from the
Palau specimen of T. swinhoei to separate pure samples
of theopalauamide (1), which was identical to the mate-
rial from the Mozambique sponge, and isotheopalaua-
mide (2).5
Theopalauamide (1) was obtained as a white powder
that decomposed before melting. The molecular formula,
C76H99BrN16O27, was deduced from high-resolution mass
measurement and analysis of the 13C NMR spectrum,
which contained 76 carbon signals. The molecular
formula differs from that of theonegramide (3, C75H97
-
BrN16O26) by addition of “CH2O”, and that difference was
compatible with the replacement of the pentose sugar (D-
arabinose) in 3 by a hexose in 1. The sugar was identified
1
as galactose on the basis of H and 13C NMR data and
(1) (a) Bewley, C. A.; Faulkner, D. J . Angew. Chem., Int. Edit. Engl.
Submitted. (b) Faulkner, D. J . Nat. Prod. Rep. 1997, 14, 259-302 and
previous reports in this series.
by analysis of key 1H NMR coupling constants (see Table
(2) Bewley, C. A.; Holland, N. D.; Faulkner, D. J . Experientia 1996,
52, 716-722.
(3) The name isotheopalauamide is used because compound 2 is a
stable conformational isomer of theopalauamide (1) with different
physical and spectral properties. There does not appear to be a special
prefix to describe this situation.
(4) Bewley, C. A.; Faulkner, D. J . J . Org. Chem. 1994, 59, 4849-
4852.
(5) We eventually found that isomerization occurred during removal
of the solvent after HPLC separation and that isomerization could be
suppressed by addition of ammonia to neutralize the TFA.
S0022-3263(97)01845-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/30/1998

Theopalauamide
J . Org. Chem., Vol. 63, No. 4, 1998 1255
Ta ble 1. 1H (DMF -d 7) a n d 13C (DMF -d 7) NMR Da ta for Th eop a la u a m id e (1).
amino acid
C
δC
δH
mult, J (Hz)
amino acid
C
δC
δH
mult J (Hz)
allo-Thr
1
2
3
4
OH
NH
173.4
59.1
69.1
21.7
4
NH
NH2
171.3
4.58
3.86
1.16
6.08
7.87
t, 10
m
d, 5.5
8.30
6.97
7.70
d
br s
br s
d, 8
â-OHAsp
1
2
171.3
54.6
5.73
4.21
br t, 5
br
Ser-1
1
2
3
170.1
57.8
62.3
3
4
NH
NH2
OH
1
2
3
4
72.2
170.8
4.80
3.70
3.84
8.97
m
m
m
d, 10
8.69
7.46
4.79
d, 10
br s (2H)
NH
1
2
Phe
172.4
54.9
39.8
BMPA
174.2
59.1
38.7
4.88
2.77
2.97
q, 8
m
dd, 13.5, 7
4.76
3.75
1.19
m
m
d, 5
3
17.9
4
137.8
128.9
130.0
127.2
5
141.6
129.3
131.6
120.6
5,9
6,8
7
NH
1
7.29
7.29
7.23
8.28
s
6,10
7,9
8
NH
1
7.37
7.14
d, 8
d, 8
d, 8
t, 8
d, 8
8.06
d
AHMP
172.7
37.9
i-Ser
171.9
70.4
44.3
2
2.34
2.70
4.41
4.48
5.43
br d, 13.5
m
m
m
2
3
4.29
3.12
4.00
7.44
br
br d, 10
m
3
4
5
6
7
8
9
53.7
68.9
NH
d, 7.5
134.1
135.7
134.1
128.2
138.3
127.0
131.6
128.0
13.0
d, 8.5
R-AAA
1
2
173.7
35.1
6.83
6.60
d, 15.5
d, 15.5
1.82
2.27
1.00
1.40
1.66
1.82
4.36
m
m
m
m
m
m
m
3
4
22.2
31.8
10,14
11,13
12
15
NH
7.54
7.38
7.26
1.79
8.49
d, 7
t, 8
t, 8
s
5
6
51.9
174.1
d, 9.5
NH
7.68
d, 8.5
Ser-2
Ala
1
2
3
171.9
57.5
61.5
4.00
3.71
3.94
8.18
m
m
m
br s
His
1
2
3
170.1
54.9
27.2
5.05
3.05
3.60
m
dd, 16, 3
t, 16
NH
4
5
7
NH
131.7
124.8
137.5
1
2
3
169.8
51.4
50.7
7.47
9.49
8.92
br s
br s
d, 8
5.26
4.43
5.10
8.61
br d, 10
m
br d, 9
d, 10.5
NH
Gal
1
2
3
4
5
6
90.1
70.9
74.6
71.1
78.9
63.5
5.47
3.76
3.53
4.01
3.93
3.98
d, 8.5
m
dd, 11.5, 2
m
m
m
Asp
1
2
3
172.1
52.6
37.1
4.70
2.49
m
br d, 13.5
1). The coupling constant of the anomeric proton at δ
5.47 (d, 1 H, J ) 8.5 Hz) indicated that H-1 and H-2 were
both axial and those of the H-3 proton at 3.53 (dd, 1 H,
J ) 11.5, 2 Hz) required H-3 to be axial and H-4 to be
equatorial. The H-4, H-5, and H-6 signals were overlap-
ping, but comparison of the corresponding 13C chemical
shifts with literature values6 strongly suggested the
galactose stereochemistry. Mild acid hydrolysis of theopa-
lauamide (1) produced D-galactose, which was identified
by derivatization and comparison with authentic samples
of both D- and L-galactose using GC-MS with a chiral
support.
Having identified the sugar residue, it was essential
to confirm that the bicyclic peptide portion of theopa-
lauamide (1) was identical to that of theonegramide (3).
This was most easily accomplished by performing a
complete structural elucidation. The DQCOSY and
TOCSY spectra were used to determine the identity of
most of the amino acids. The HMQC and HMBC experi-
(6) Stothers, J . B. Carbon-13 NMR Spectroscopy; Academic Press:
New York, 1972; p 461.

1256 J . Org. Chem., Vol. 63, No. 4, 1998
Schmidt et al.
Ta ble 2. Com p a r ison of NMR Da ta for th e P h eyn yla la n in e a n d AHMP Resid u es of Th eop a la u a m id e (1) a n d
Isoth eop a la u a m id e (2)
1
2
amino acid
C
δC
δH
mult, J (Hz)
δC
δH
mult, J (Hz)
Ser-1
1
2
1
2
3
170.1
57.8
172.4
54.9
170.4
57.9
172.2
55.1
4.80
m
4.70
m
Phe
4.88
2.77
2.97
q, 8
m
dd, 13.5, 7
4.73
2.70
2.81
m
m
m
39.8
39.7
4
137.8
128.9
130.0
127.2
138.3
129.9
128.8
128.0
5,9
6,8
7
NH
1
7.29
7.29
7.23
8.28
s
7.23
7.24
7.25
8.26
s
m
m
d, 9.5
d, 8
t, 8
d, 8
AHMP
172.7
37.9
173.1
37.8
2
2.34
2.70
4.41
4.48
5.43
br d, 13.5
m
m
m
2.50
2.84
4.43
4.42
5.64
m
m
m
m
3
4
5
6
7
8
9
53.7
68.9
53.8
70.4
134.1
135.7
134.1
128.2
138.3
127.0
131.6
128.0
13.0
d, 8.5
134.4
136.4
134.0
128.6
138.3
127.0
131.6
128.2
13.3
d, 8.5
6.83
6.60
d, 15.5
d, 15.5
6.92
6.65
d, 15.5
d, 15.5
10,14
11,13
12
15
NH
2
7.54
7.38
7.26
1.79
8.49
4.00
9.49
d, 7
t, 8
t, 8
s
d, 9.5
m
7.54
7.38
7.26
1.92
8.53
3.95
9.32
m
m
m
s
d, 12
m
br s
Ser-2
Hist
57.5
137.5
57.7
137.8
7
br s
ments confirmed these assignments and allowed identi-
fication of the aromatic amino acids. The amino acid
sequence was determined from the HMBC and ROESY
(mixing time ) 300 ms) data and was identical to that of
theonegramide. The absolute configuration of each amino
acid in theopalauamide was shown to be identical to that
in theonegramide by employing the same series of
chemical degradations, followed by chiral GC-MS analy-
sis of the fragments, that had been used previously.4
These experiments established that theopalauamide (1)
and theonegramide (3) differed only in the identity of the
sugar unit.
Isotheopalauamide (2) has the same molecular formula
as theopalauamide (2), and both have identical IR and
UV spectra. The differences between the 1H and 13C
NMR data for 1 and 2 are mainly confined to signals
associated with the phenylalanine and 3-amino-4-hy-
droxy-6-methyl-8-phenylocta-5,7-dienoic acid (AHMP)
units (Table 2). The most noticeable differences in the
1H NMR spectra were associated with the vinyl methyl
signal (δ 1.79 in 1 vs 1.92 in 2) and the olefinic proton
signals on the AHMP residue. In fact, the strength of
the δ 1.92 signal could be used to determine the purity
of samples of both theopalauamide (1) and, interestingly,
theonegramide (3), which appeared to have isomerized
to a lesser extent. Analysis of the DQCOSY, TOCSY,
HMQC and HMBC data indicated that the primary
structures of 1 and 2 were identical. The GC-MS data
for 1 and 2 were identical except that different retention
times were observed for the hydrogenation product of
AHMP. From these experiments, however, it was not
possible to determine whether the differences were due
to the stereochemistry of the AHMP residue or to epimers
at C-6 caused by hydrogenation occurring from different
faces of the diene in peptides 1 and 2.
spectral differences were caused by acid-catalyzed epimer-
ization of the C-4 allylic alcohol in the AHMP residue.
To test this hypothesis, a 1:1 mixture of peptides 1 and
2 was hydrogenated and hydrolyzed to obtain a mixture
of two lactones arising from the AHMP residue. In
theory, four isomers should have been produced but we
were not concerned at that point because Matsunaga et
al.7 had noted in their elegant studies of theonellamide
F that hydrogenation of the AHMP residue appeared to
produce predominantly one epimer at the C-6 carbon
bearing the secondary methyl group. The lactones were
then converted into their N-2,4-dinitrophenyl derivatives,
which were separated using HPLC to obtain lactones 4
and 5 in a ratio of 1.7:1. The NOESY spectrum of the
major (3S,4S)-lactone 4, which was identical to that
reported by Matsunaga et al.,7,8 showed a correlation
between H-3 and H-4 as expected. However, the 1H NMR
and NOESY spectra of lactone 5 were almost identical
to those of lactone 4, particularly with regard to those
signals associated with the lactone ring (Figure 1). A
careful analysis of the NOESY spectra of lactones 4 and
5 revealed that both compounds had the same relative
stereochemistry about the lactone ring. Since the major
chemical shift differences were associated with the meth-
yl signals, we concluded that the lactones 4 and 5 were
epimers at C-6 that had been generated during the
hydrogenation step. After reviewing all of the spectra
generated during the reaction sequence described above,
(7) Matsunaga, S.; Fusetani, N.; Hashimoto, K.; Walchli, M. J . Am.
Chem. Soc. 1989, 111, 2582-2588.
(8) Matsunaga et al.7 reported that hydrogenation of theonellamide
F followed by hydrolysis of the peptide and 2,4-DNP derivatization
gave a 5:1 mixture of epimers, of which the major isomer was identical
to lactone 4. When we hydrogenated theonellamide F and performed
the standard hydrolysis and derivatization for GC-MS analysis, we
found that the retention time of the minor peak resulting from the
AHMP residue was approximately the same as that of the major peak
produced by hydrogenation, hydrolysis, and derivatization of iso-
theopalauamide (2).
Having first ruled out the possibility that 2 was simply
the TFA salt of 1, we initially sought to show that the

Theopalauamide
J . Org. Chem., Vol. 63, No. 4, 1998 1257
Exp er im en ta l Section 11
Isola tion of Th eop a la u a m id e (1), Isoth eop a la u a m id e
(2), Sw in h olid e A, a n d Th eon ella ster ol. Mixtures of
theopalauamide (1) and isotheopalauamide (2) were obtained
from T. swinhoei from Palau and from the filamentous bacteria
as described previously.2 The mixture was separated by
reversed-phase HPLC on a C18 preparative column using 35%
acetonitrile in water containing 0.01% TFA to obtain pure
samples of theopalauamide (1) and isotheopalauamide (2).
Theopalauamide (1) and swinholide A were obtained from
a specimen of T. swinhoei from Mozambique (MOZ 95-004)
using a slightly modified procedure. The sponge (211 g dry
weight) was lyophilized and sequentially extracted with 1:1
hexanes/CH2Cl2 (3 × 1 L), EtOAc (3 × 1 L), and 1:1 EtOAc/
acetone (3 × 1 L). The sponge was then exhaustively extracted
with 1:1 acetonitrile/water until no further peptide was
detected in the crude extracts by TLC. The aqueous extracts
were dried by rotary evaporation until most CH3CN was
removed and a white precipitate had formed. The suspension
was centrifuged, and the supernatant removed. The precipi-
tate was applied directly to the top of a C18 Sep Pak (Waters)
column and subjected to reversed-phase chromatography using
an acetonitrile/water gradient (0-100% CH3CN in 10% incre-
ments). Fractions eluting with 40-60% acetonitrile were
enriched with a single peptide, which was purified by reversed-
phase HPLC (38% CH3CN, 0.01% TFA) to obtain theopalaua-
mide (1, 55.7 mg, 0.026% dry weight). Fractions eluting with
80% acetonitrile were combined to yield impure swinholide A,
which was subsequently repurified by silica flash chromatog-
raphy (100% EtOAc) to obtain swinholide A (21.3 mg, 0.01%
dry weight). Additional swinholide A was present in the
EtOAc/acetone extract. The hexane/CH2Cl2 extract was par-
titioned between hexane and methanol (8 mL each). The
hexane fraction was evaporated under vacuum to obtain a
residue that was purified by flash chromatography on silica
gel using a hexane/ethyl acetate gradient (20-100% EtOAc).
The 30% EtOAc fraction contained pure theonellasterol (28.2
mg, 0.013% dry weight).12
F igu r e 1. Key NOE correlations for lactones 4 and 5, which
are epimeric at C-6. NOESY spectra (500 MHz) were virtually
identical for both compounds. Strong NOE correlations are
shown as solid lines and a weak correllation as a dashed line.
we could find no evidence of epimerization at C-4 or at
any other chiral center.
In the ROESY spectrum of isotheopalauamide (2),
there is a correlation between the NH protons of the Phe
and AHMP residues that is absent in the ROESY
spectrum of 1. Conversely, the correlation between the
NH proton of the AHMP residue and the R-proton of the
Phe residue was observed in 1 but not in 2. The lack of
correlations between the R-protons of the Phe and AHMP
residues in 2 suggests that the geometry of the amide
bond is anti as usual. Thus the only remaining explana-
tion of the difference between the two peptides is that
they are different stable conformers of the same com-
pound. Since NMR chemical shift differences in ex-
changeable proton signals indicate the degree of intramo-
lecular hydrogen bonding,9 the 1H NMR spectra of
compounds 1 and 2 were acquired every 5 °C from 25 to
40 °C. Analysis of chemical shift differences showed that
the -NH protons on both the AHMP and Phe residues
were more strongly hydrogen bonded in 2 than in 1,
which is consistent with the ROESY data, while most
other chemical shift differences were roughly equal. In
1, the AHMP and Phe -NH proton signals had ∆δ/∆T
(ppb/K) values of 6.5 and 7.1, respectively, while the same
protons in 2 had ∆δ/∆T values 3.1 and 1.5, respectively.
The NH-CHR coupling constants for Phe and AHMP
were slightly larger in 2 than in 1, but the differences
are not great enough to indicate a major change in the
geometry about these bonds. Thus it appears that the
major conformational difference between 1 and 2 is a
rotation of the bond between C-1 and C-2 in the Phe
residue, as shown in Figure 2, with minor adjustments
to the geometry of other bonds in the ring. Space-filling
models of theopalauamide indicate that such a rotation
moves the phenyl ring of Phe and the AHMP side chain
further apart in 2 than in 1, which is consistent with both
the differences in NMR data for the two molecules (Table
2) and the hydrogenation data that suggests that one face
of the AHMP diene system is sterically hindered in 1 but
not in 2. Having determined that theopalauamide (1)
and isotheopalauamide (2) were simply conformational
isomers, we were surprised at their apparent stability
for we were unable to find any other conditions to
isomerize the pure compounds other than during chro-
matography in the presence of TFA and subsequent
evaporation of the solvents.10
Th eop a la u a m id e (1): white powder; [R]D ) +19° (c ) 0.4,
MeOH); UV (MeOH) 203 nm (ꢀ 19,600), 276 (ꢀ 9400), 285 (ꢀ
9900), 304 nm (ꢀ 4500); IR (AgCl) 3300, 2920, 1660, 1540 cm-1
;
1H NMR (500 MHz, DMF-d7) see Table 1; 13C NMR (75 MHz,
DMF-d7) see Table 1; HRFABMS m/z 1769.5944 (M + Na)+,
calcd for C76H9979BrN16O27Na, 1769.5947.
Isoth eop a la u a m id e (2): [R]D ) +35° (c ) 0.04, MeOH);
UV and IR data were identical to those of 1; 1H (500 MHz,
DMF-d7) and 13C NMR (75 MHz, DMF-d7) data were identical,
within the limits of error, to those reported for 1 in Table 1
except for the signals listed in Table 2; HRFABMS m/z
1747.6233 (M + H)+, calcd for C76H10079BrN16O27, 1747.6127.
Hyd r olysis a n d Der iva tiza tion of Hyd r olysa tes. The
glycopeptides 1, 2, or 3 (100-500 µg) were dissolved in 6 N
HCl (500 µL) and heated to 110 °C in tightly sealed 1 mL
conical vials for 15 h. HCl was removed under a stream of
nitrogen. The hydrolysates were dissolved in 2-propanol (400
µL), to which was added acetyl chloride (100 µL). The vials
were quickly capped, and the solutions were heated to 100 °C
for 1 h. Excess reagents were removed under nitrogen, and
the residue was redissolved in CH2Cl2 (400 µL). Pentafluo-
ropropionic anhydride (400 µL) was added, and the reaction
mixture was heated to 100 °C for 15 min. Reagents were
removed under nitrogen, and the derivatized hydrolysates
were redissolved in EtOAc for GC-MS analysis.
Id en tifica tion of D-Ga la ctose. Both theopalauamide (1,
500 µg) and isotheopalauamide (2, 500 µg) were separately
dissolved in 4 N HCl (500 µL) and heated to 70 °C for 12 h.
The solvent was removed, and the hydrolysates were deriva-
tized as described above to obtain a single peak in the GC-MS
that corresponded to D-galactose.
Both theopalauamide (1) and isotheopalauamide (2)
inhibited the growth of Candida albicans. In the stan-
dard paper disk assay, 1 was active at 10 µg/disk while
2 was active at 50 µg/disk.
(9) Morita, H.; Kayashita, T.; Takeya, K.; Itokawa, H.; Shiro, M.
Tetrahedron 1997, 53, 1607-1616.
(10) A search of the marine natural product literature failed to
provide other examples of the isolation of stable conformers of cyclic
peptides.
(11) For general experimental procedures, see: Schmidt, E. W.;
Harper, M. K.; Faulkner, D. J . J . Nat. Prod. 1997, 60, 779-782.
(12) Kho, E.; Imagawa, D. K.; Rohmer, M.; Kashman, Y.; Djerassi,
C. J . Org. Chem. 1981, 46, 1836-1839.

1258 J . Org. Chem., Vol. 63, No. 4, 1998
Schmidt et al.
F igu r e 2. Schematic drawings and space-filling models of Phe and AHMP residues in theopalauamide (1) and isotheopalauamide
(2). Key ROESY correlations are indicated in the top drawings. In the space-filling models, atom types are indicated by
patterning: C, dark shading, H, white, N, light shading, O, dark pattern. These models are derived from PCMODEL v 5.0 (MMX)
minimized fragments of theopalauamide and isotheopalauamide, restrained on the basis of coupling constants and ROESY data,
and are therefore qualitative.
Ozon olysis of Th eop a la u a m id e (1) a n d Isoth eop a la u a -
m id e (2). In separate experiments, samples of peptides 1, 2,
and 3 (1.2 mg of each) were dissolved in MeOH (0.5 mL) and
a stream of ozone in oxygen was bubbled through the cooled
solution at -78 °C for 15 min. To the resulting ozonides were
added 50% H2O2 (10 drops), and the mixture was allowed to
stand at room temperature for 1 h. The solvent was removed
under a stream of nitrogen, and the ozonolysis products were
hydrolyzed and derivatized as described above.
Hyd r ogen a tion of 1 a n d 3 for GC-MS An a lysis. Pal-
ladium (5%) on charcoal catalyst (3-10 mg) was added to
solutions of peptides 1, 2, or 3 (1-10 mg in 5 mL of MeOH),
and the solutions were stirred under an atmosphere of
hydrogen gas for 24 h. The solutions were filtered through
Celite, dried under vacuum, and subjected to the hydrolysis/
derivatization sequence described above.
) 17.5, 7 Hz), 2.70 (m, 1 H), 2.64 (dd, 1 H, J ) 17.5, 3 Hz),
2.56 (m, 1 H), 1.95 (ddd, 1 H, J ) 15, 10, 5 Hz), 1.77 (m, 1 H),
1.67 (m, 2 H), 1.43 (ddd, 1 H, J ) 15, 10, 5 Hz), 1.02 (d, 3 H,
J ) 6.5 Hz).
La cton e 5: 1H NMR (CDCl3) δ 9.17 (d, 1 H, J ) 2.5 Hz),
8.81 (d, 1 H, J ) 7.5 Hz), 8.33 (dd, 1 H, J ) 9.5, 2.5 Hz), 7.1-
7.3 (m, 5 H), 6.80 (d, 1 H, J ) 9.5 Hz), 4.83 (dt, 1 H, J ) 9, 5
Hz), 4.44 (dddd, 1 H, J ) 7.5, 6.5, 5, 3 Hz), 3.04 (dd, 1 H, J )
17.5, 6.5 Hz), 2.70 (m, 1 H), 2.66 (dd, 1 H, J ) 17.5, 3 Hz),
2.54 (m, 1 H), 1.88 (m, 1 H), 1.73 (m, 1 H), 1.69 (m, 2 H), 1.48
(m, 1 H), 1.07 (d, 3 H, J ) 6.5 Hz).
Ack n ow led gm en t. The specimen from Mozambique
was collected by Dr. Brad Carte´, National University
of Singapore, and Dr. Philip Coetzee, University of Port
Elizabeth, and was provided to us by Dr. Mike Davies-
Coleman, Rhodes University. We also thank the Re-
public of Palau and the State of Koror for Marine
Research Permits. We thank Dr. J ay Siegel for his valu-
able insights concerning stereochemistry. This research
was supported by grants from the National Science
Foundation (CHE95-27064), the National Institutes of
Health (CA 49084), and the California Sea Grant
College Program (NOAA NA36RG0537, project R/MP
60). The Varian Inova 300 MHz NMR spectrometer was
purchased using a grant from the National Science
Foundation (CHE95-23507).
Hyd r ogen a tion a n d Hyd r olysis To P r od u ce La cton es
4 a n d 5. Palladium (5%) on charcoal catalyst (250 mg) was
added to a solution of a 1:1 mixture of 1 and 2 (120 mg) in
MeOH (10 mL), and the mixture was stirred under an
atmosphere of hydrogen gas for 24 h. The product was filtered,
and the solvent was evaporated to obtain the de-brominated,
1
tetrahydrogenated compounds as indicated by H NMR data.
The mixture was dissolved in 6 N HCl (6 mL), and the solution
was heated for 15 h at 110 °C. The product was lyophilized,
and the residue was partitioned between EtOAc and water.
The EtOAc layer was evaporated to obtain an inseparable
mixture (5.6 mg) of lactones that was analyzed by 1H NMR
spectroscopy without separation.
A solution of lactones in ether (0.1 mL) containing 2,4-
dinitrofluorobenzene (14 µL) and TEA (4 µL) was stirred for 2
h at room temperature. The solution was filtered through
silica gel using 3:2 EtOAc-hexane as eluant, and the mix-
ture of products was separated by HPLC on silica using 1:1
EtOAc-hexane as eluant to obtain pure samples of lactone 4,
which had identical 1H NMR data to literaure values,7 and
lactone 5.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
and COSY spectra of theopalauamide (1) and isotheopalaua-
mide (2), an expansion of the H NMR spectrum showing the
signals due to galactose, and expansions of the ROESY spectra
that show the differences between 1 and 2 (9 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
1
La cton e 4: 1H NMR (CDCl3) δ 9.17 (d, 1 H, J ) 2.5 Hz),
8.80 (d, 1 H, J ) 7.5 Hz), 8.33 (dd, 1 H, J ) 9.5, 2.5 Hz), 7.1-
7.25 (m, 5 H), 6.79 (d, 1 H, J ) 9.5 Hz), 4.83 (dt, 1 H, J ) 10,
5 Hz), 4.40 (dddd, 1 H, J ) 7.5, 6.5, 5, 3 Hz), 3.04 (dd, 1 H, J
J O9718455