Synthetic studies toward bioactive cyclic peroxides from the marine sponge Plakortis angulospiculatus

Article abstract of DOI:10.1021/ol016943y

(formula presented) The total synthesis of two stereoisomers of a bioactive cyclic peroxide isolated from the marine sponge Plakortis angulospiculatus has been achieved in 18 steps with an overall yield of 2.8%. Diels-Alder addition of singlet oxygen to an acyclic triene carboxylic acid precursor was used to construct the 3,6-dihydro-1,2-dioxin ring. By comparing spectral data of the synthesized compounds and the natural material, we tentatively assign the absolute stereochemistry for the natural product as 3S,6R,8S,10R.

Full text of DOI:10.1021/ol016943y

ORGANIC
LETTERS
2002
Vol. 4, No. 4
485-488
Synthetic Studies toward Bioactive
Cyclic Peroxides from the Marine
Sponge Plakortis angulospiculatus
Gang Yao† and Kosta Steliou*
Department of Chemistry, Metcalf Center for Science and Engineering,
590 Commonwealth AVenue, Boston UniVersity, Boston, Massachusetts 02215
steliou@bu.edu
Received October 20, 2001
ABSTRACT
The total synthesis of two stereoisomers of a bioactive cyclic peroxide isolated from the marine sponge Plakortis angulospiculatus has been
achieved in 18 steps with an overall yield of 2.8%. Diels−Alder addition of singlet oxygen to an acyclic triene carboxylic acid precursor was
used to construct the 3,6-dihydro-1,2-dioxin ring. By comparing spectral data of the synthesized compounds and the natural material, we
tentatively assign the absolute stereochemistry for the natural product as 3S,6R,8S,10R.
In 1990 Gunasekera and co-workers isolated two cyclic
peroxides, 1a and 2a, (Scheme 1) from the marine sponge
with IC50 values of 0.2-0.3 µg/mL in vitro against P388
mouse leukemia cells and MIC values of 1.6 µg/mL against
Candida albicans for 1a and 2a.1a Interestingly, the corre-
sponding methyl esters 1b and 2b are void of any antifungal
activity.
Snider and Shi2 reported the first total synthesis of a cyclic
peroxy-ketal member of this family of compounds. A novel
photooxygenation procedure was specifically developed for
this synthesis and then successfully applied to the preparation
of chondrillin and plakorin in racemic form. Dussault’s3
stereoselective synthesis of (-)-plakorin and (+)- and (-)-
chondrillin also relies on a similar photomediated oxygen-
Scheme 1
(1) (a) Gunasekera, S. P.; Gunasekera, M.; Gunawardana, G. P.;
McCarthy, P.; Burres, N. J. Nat. Prod. 1990, 53, 669. (b) Compagnone, R.
S.; Pin˜a, I. C.; Rangel, H. R.; Dagger, F.; Sua´rez, A.; Venkata, M.; Reddy,
R.; Faulkner, D. J. Tetrahedron 1998, 54, 3057. (c) Takada, N.; Watanabe,
M.; Yamada, A.; Suenaga, K.; Yamada, K.; Ueda, K.; Ueura, D. J. Nat.
Prod. 2001, 64, 356. (d) Campagnuolo, C.; Fattorusso, E.; Taglialatela-
Scafiti, O.; Ianaro, A.; Pisano, B. Eur. J. Org. Chem. 2002, 61.
Plakortis angulospiculatus of the family Plakinidae from the
Caribbeans.1a These natural cyclic peroxide-containing car-
boxylic acids1 have potent cytotoxic and antifungal properties
† Present address: ArQule, Inc., 19 Presidential Way, Woburn, MA
01801.
(2) Snider, B. B.; Shi, Z. J. Am. Chem. Soc. 1992, 114, 1790.
(3) Dussault, P. H.; Woller, K. R. J. Org. Chem. 1999, 64, 1789.
10.1021/ol016943y CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/22/2002

ation and peroxyl radical rearrangement. Bloodworth and co-
workers4 prepared a series of five-membered ring plakinic
acids in racemic form via sequential peroxymercuriation and
in situ reduction of the resulting mercuriated 2,4-dienic acids
with sodium borohydride. Also, Xu and co-workers5a pub-
lished the enantioselective total synthesis of yingzhaosu C,
an antimalarial peroxy-containing sesquiterpene isolated from
yingzhao.5
Scheme 3a
However, no synthesis of any member of the plakortic
acid family of compounds has, heretofore, been achieved.
In this communication, we describe the first total synthesis
of compounds 13a and 13b and, from a comparative spectral
analysis of their structures, deduce the likely absolute
configuration of the natural product 1a to be 3S,6R,8S,10R,
which is consistent with the absolute stereochemistry reported
for C3, C6, and C8 in analogous natural examples.1b-d The
salient feature in our retrosynthetic strategy6 focuses on a
singlet oxygen [4 + 2] cycloaddition reaction to an enantio-
pure form of (3E,5E,11E)-triene carboxylic acid 3, an
advanced intermediate in the synthesis, which in principle
could be derived from commercially available optically pure
methyl (S)-(+)-3-hydroxy-2-methylpropionoate 6 as shown
in Scheme 2.
a (a) TBDPSCl, imidazole, DMF, 0 °C to rt, 5 h; (b) DIBAL,
toluene, -78 °C to rt, 2 h, (97% two steps); (c) PPh3, imidazole,
I2, CH2Cl2, 0 °C to rt, 2 h (96%); (d) LDA, LiCl, THF, 0 °C to rt,
24 h (97%); (e) LiH2NBH3, THF, 0 °C to rt, 8 h (93%, de > 95%).
(f) (COCl)2, DMSO, -78 °C; Et3N, -78 °C to rt (100%).
yield with 95% de after aqueous acidic workup. Swern10
transformation of the alcohol into aldehyde 5 proceeded
smoothly in quantitative yield and without epimerization of
the C10 (13a numbering) stereocenter.
The synthesis of fragment 4 is outlined in Scheme 4.
Julia-Lythgoe11 olefination of aldehyde 5, by reaction of
Scheme 2. Retrosynthetic Analysis
Scheme 4a
a (a) C2H5CH2SO2Ph, n-BuLi, THF, -78 °C; (b) Ac2O, TEA,
DMAP; (c) Mg, HgCl2, EtOH, rt, 72% three steps, E:Z 6:1; (d)
TBAF, THF, rt, 100%; (e) I2, PPH3, imidazole. toluene, rt, 89%;
(f) NaCN, DMF, rt, 92%; (g) DIBAL, Et2O, -78 to 0 °C, 90%;
(h) PPh3, CBr4, K2CO3, CH2Cl2, 0 °C to rt, 92%; (i) n-BuLi, THF,
-78 °C, 77%.
The synthesis of diastereomer 5 is outlined in Scheme 3.
Treatment of 6 with tert-butyldiphenylsilyl chloride and
imidazole followed by DIBAL reduction7 provided alcohol
7a in 97% overall yield. Iodination of 7a using Corey’s
procedure8 gave iodide 7b (96%). Asymmetric alkylation of
the lithium enolate of butyramide 8 (derived from (1S,2S)-
(+)-pseudoephedrine) with 7b in THF at room temperature
following Myers’ method9 yielded an R,γ-skipped dialkyl
substituted amide, which after reduction with lithium amido-
trihydroborate afforded the corresponding alcohol in 93%
the lithium anion derived from phenyl propyl sulfone,
afforded a stereoisomeric mixture of the corresponding
â-hydroxyl sulfones. The â-hydroxy sulfones were isolated
and then acetylated (Ac2O, Et3N, DMAP, CH2Cl2) to give a
mixture of acetoxy sulfone diastereomers. The more direct
and typical quenching of the reaction mixture with acetic
anhydride gave a much lower yield, presumably as a result
of the high alkalinic nature of the reaction medium. Reduc-
(4) Bloodworth, A. J.; Bothwell, B. D.; Collins, A. N.; Maidwell, N. L.
Tetrahedron Lett. 1996, 37, 1885.
(9) (a) Myers, A. G.; Yang, B. H.; Chen, H.; Gleason, J. L. J. Am. Chem.
Soc. 1994, 116, 9361. (b) Myers, A. G.; McKinstry, L. J. Org. Chem. 1996,
61, 2428. (c) Myers, A. G.; Gleason, J. L.; Yoon, T.; Kung, D. W. J. Am.
Chem. Soc. 1997, 119, 656. (d) Meyers, A. I.; Wunsch, T. J. Org. Chem.
1990, 55, 4233.
(10) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. J. Org. Chem. 1978,
43, 2480.
(11) (a) Julia, M.; Paris, J.-M. Tetrahedron Lett. 1973, 49, 4833. (b)
Kocienski, P. J.; Lythgoe, B.; Ruston, S. J. Chem. Soc., Perkin Trans. 1
1978, 829. (c) Kocienski, P. J.; Lythgoe, B.; Waterhouse, I. J. Chem. Soc.,
Perkin Trans. 1 1980, 1045.
(5) (a) Xu, X.-X.; Dong, H.-Q J. Org. Chem. 1995, 60, 3039. (b) Posner,
G. H.; O’Dowd, H.; Polypradith, P.; Cumming, J. N.; Xie, S.; Shapiro, T.
A. J. Med. Chem. 1998, 41, 2164. (c) Bachi, M. D.; Korshin, E. E.; Hoos,
R.; Szpilman, A. M. J. Heterocycl. Chem. 2000, 37, 639.
(6) Steliou, K.; Milot, G.; Salama, P.; Yu, X. Presented at the 200th
National Meeting of the American Chemical Society, Washington, DC,
August 1990; ORGN 119.
(7) Jones, T. K.; Reamer, R. A.; Desmond, R.; Mills, S. G. J. Am. Chem.
Soc. 1990, 112, 2998.
(8) Corey, E. J.; Pyne, S. G.; Su, W. Tetrahedron Lett. 1983, 24, 4883.
486
Org. Lett., Vol. 4, No. 4, 2002

Scheme 5a
a (a) EtCu[Me2S]MgBr2/Et2O/Me2S (1:2:1), -25 °C, 12 h; (b) HMPA, THF, EtCOCl, Pd(PPh3)4, -30 °C to rt, 82% two steps, Z/E
1
<1:99; (c) 12, t-BuLi, THF -78 °C to rt, 73%; 3E:3Z 88:12; (d) (C6H11)2BH, THF, 0 °C; (e) H2O2, NaOH, 0 °C; 42% two steps; (f) O2,
hν, rose bengal, MeOH, CH2Cl2, 8 h; (g) CH2N2, 40% two steps.
tion of the isolated acetoxy sulfones with magnesium
amalgam (Mg, HgCl2, 0 °C to room temperature) in THF-
ethanol (1:3)12 afforded a 6:1 trans to cis ratio of the olefin
9 in 70% overall yield from 5. Separation of the geometric
isomers was not possible at this stage. The 6:1 E:Z ratio is
based on analysis of the 1H NMR data. The (E)-double bond
configuration in 9 displays a characteristic trans-coupling
constant of J ) 15.2 Hz in the 1H NMR spectra for the C11
and C12 vinyl hydrogens as expected.
Condensation of enone 11 with the lithium salt of
phosphonate 12 (1.3 equiv) in THF at -78 °C followed by
gradual warming (ca. 6 h) to room temperature yielded a
mixture of trienyne geometrical isomers (E:Z ratio 88:12)
that were readily separated by flash column chromatography
in 76% combined yield. Importantly, the substrate concentra-
tion has a profound effect on the reaction yield. An optimal
yield is achieved when the concentration is kept between
0.2 and 0.5 M. The 88:12 3E:3Z ratio was determined by
1
Desilylation of 9 with TBAF afforded the free alcohol in
quantitative yield. Conversion of the alcohol, by the Corey
method8 (PPh3, imidazole, I2, PhH, rt, 10 min), into the iodide
proceeded in 89% yield. Attempts to make the terminal
alkyne by direct nucleophilic displacement of the iodide with
lithium acetylide13 resulted only in elimination products being
formed. To overcome this difficulty, the iodide was converted
into a nitrile by reaction with NaCN in dry DMF (92%).
Reduction of the nitrile with DIBAL furnished aldehyde 10
analyzing the H NMR spectrum. The 3E,5E geometry of
the major product (3, Scheme 5) was confirmed by NOE
studies. Irradiation at the C7 methylene group produced a
10.4% NOE to the C5 vinylic proton, consistent with a 5E
geometry. Irradiation at the methylene hydrogens of the ethyl
group at C4 produced a larger NOE (14%) to the C5 vinylic
proton and a smaller NOE (4.5%) to the C3 vinylic proton,
consistent with the 3E-olefin geometry. It is noteworthy that
the larger NOE between the allylic hydrogens of the ethyl
group at C4 and the vinylic proton at C5 is indicative of the
diene being significantly populated in the s-cis conformation,
which is believed to be crucial for the success of the planned
[4 + 2] singlet oxygen cycloaddition reaction.16
Transformation of the 3E,5E-enyne into the (3E,5E,11E)-
carboxylic acid 3 was accomplished by utilizing the method
of Zweifel and Backlund.17 Thus, regioselective hydrobora-
tion of the enyne with dicyclohexylborane (1.2 equiv) at 0
°C followed by oxidation using H2O2 in aqueous basic
medium at 0 °C converted the enyne into carboxylic acid 3
in 65% yield. Reaction of 3 with singlet oxygen was then
carried out.18 The reaction was conducted by irradiating a 3
× 10-3 M solution of 3 in 19:1 CH2Cl2/MeOH with a sun
lamp for 8 h at 0 °C in the presence of rose bengal (10-4 M)
and oxygen. This afforded a 2:1 mixture of the diastereomeric
peroxy carboxylic acids, which were then esterified to give
their respective corresponding methyl esters 13a and 13b in
a combined overall yield of 40%. The separation of 13a and
13b was achieved by reverse phase HPLC (C-18, gradient,
CH3CN/H2O, 80:20 f 100:0).
14
(90%). Treatment of 10 with Ph3P and CBr4 gave the
corresponding dibromo alkene in 92% yield. Finally, expo-
sure of the dibromo alkene to n-BuLi (THF, -78 °C) and
quenching the intermediate acetylenic anion with MeOH
provided the desired alkyne 4 in good yield (77%).
The completed synthesis is shown in Scheme 5. Alkyne 4
was added to a mixture of EtCu[Me2S]MgBr2 in Et2O/Me2S
(1.2:1) at -45 °C, and the whole was allowed to react at
-25 °C for 12 h to generate the required alkenylcuprate
reagent. In situ acylation by the addition of propionyl chloride
in the presence of HMPA (8 equiv) and Pd(PPh3)4 (5 mol
%) afforded dienone 11 in 82% yield and with high
geometric purity (Z:E < 1:99).15 Although dienone 11 is
reasonably stable under mild basic conditions, it is very
sensitive to acid. Attempts to purify 11 using flash column
chromatography inevitably resulted in partial isomerization.
Nonetheless, by rapid filtration through a thin pad of alumina,
the crude ketone can be made sufficiently pure to be used
in the next step.
(12) Lee, G. H.; Lee, H. Y.; Choi, E. B.; Kim, B. T.; Pak, C. S.
Tetrahedron Lett. 1995, 36, 5607.
(13) Smith, W. N.; Beumel, O. F., Jr. Synthesis 1974, 441.
(14) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769.
(15) Jabri, N.; Alexakis, A.; Normant, J. F. Tetrahedron 1986, 42, 1369
and reference therein.
(16) Clennan, E. L. Tetrahedron 1991, 47, 1343.
(17) Zweifel, G.; Backlund, S. J. J. Am. Chem. Soc. 1977, 99, 3184.
(18) Murthy, M. S. R.; Pande, S. V. J. Biol. Chem. 1984, 259, 9082.
Org. Lett., Vol. 4, No. 4, 2002
487

The absolute stereochemistry of the newly introduced C3
and C6 stereocenters was deduced using the modified
Mosher’s empirical method19 (Scheme 6). Reduction of the
On the basis of this assignment plus additional NMR data,
the absolute configuration for the natural product 1a could
be deduced. The 1H NMR spectral data of the major isomer
13a are virtually identical with those reported for the methyl
ester of the natural material (1b) except for C7-H, C8-H,
and C9-H. Yet the 1H NMR spectral data of the minor isomer
13b is quite different from that of 1b. In the natural ester
1b, C7-H and one of C9-H overlap and absorb at 1.40-
1.38 ppm. In 13a, they are more upfield (1.34-1.24 ppm)
and also overlap with the C20 methylene protons. The 13C
NMR spectrum of 13a is also similar to that of the natural
product, and here too, the discrepancy is confined to the C7-
C10 fragment as well. Compounds 13a and 13b are therefore
diastereoisomers of the natural ester 1b.
Scheme 6a
Since the substituents at C8 and C10 have an anti
relationship in 13a and 13b, they must, therefore, be syn in
the natural peroxide. Furthermore, since the 1H NMR
spectrum of the peroxy ring in 13a is almost identical to
that of the natural peroxide, the relative stereochemistry at
C8, C3, and C6 of the natural material must respectively be
the same as that conferred on the synthetic compound 13a.
Thus, it may be concluded that natural peroxide 1b has
either an absolute stereochemistry of 3S,6R,8S,10R or the
absolute stereochemistry of its enantiomer, 3R,6S,8R,10S.
However, the optical rotation obtained for synthetic peroxide
13a, [R]D ) +50.6° (c 0.17, CHCl3), is different and opposite
in sign to the value reported for the natural peroxide 1b ([R]D
) -89° (c 0.1, CHCl3). The optical rotation obtained for
13b, [R]D ) -58.3° (c 0.06, CHCl3), is, however, the same
in sign to the value reported for natural peroxide 1b.
Coupling these findings with the fact that the optical
rotations of these compounds tend to be governed by their
peroxy ring moiety,21 it is quite possible that the peroxy ring
of 1b has the same absolute stereochemistry as that of 13b.
Comparing the structure of plakortin22 with that of 1b further
supports this assignment.1b,d Since both compounds share the
same biogenetic pathway, it is highly likely that the C10
ethyl group in 1a is also the same. Accordingly, it is fair to
suggest that 3S,6R,8S,10R is the most likely absolute
configuration of the natural product 1a.
a (a) Al/Hg; (b) DCC, DMAP, MPTA.
peroxide bond in 13a was accomplished with the use of an
aluminum amalgam20 to give a diol. Treatment of the diol
with (-)-(S)- and (+)-(R)-R-methoxy-R-(trifluromethyl)-
phenylacetic acid (MTPA), DCC, and a catalytic amount of
DMAP led to the corresponding diastereomeric (S)-14a and
1
(R)-14b esters, respectively, which were analyzed by H
NMR spectroscopy.1b,d
Specifically, by comparing the H NMR spectra of (R)-
1
and (S)-MTPA esters, it could be clearly observed that the
chemical shifts for the methoxy ester group and the C2-H
protons of 14a appear at lower field than the corresponding
chemical shifts of these groups in 14b (Scheme 6, ∆δ )
+0.064 and +0.029, respectively). In addition, the chemical
shifts for the C5 vinylic proton of 14a appear at higher field
than that of 14b (∆δ ) -0.045). The calculated ∆δ values
are, therefore, consistent with an R configuration at C3 in
13a, which is identical to the absolute configuration deduced
by Faulkner et al.1b for a closely related example.
Since singlet oxygen addition to dienoids occurs in a syn
fashion, the absolute stereochemistry at C6 in 13a must,
therefore, be S. Further, because the absolute stereochemistry
for C8 (R) and C10 (R) in 13a was secured by way of
asymmetric synthesis (vide supra), the absolute configuration
for the peroxy ester 13a is thus 3R,6S,8R,10R, respectively.
Consequently, the absolute configuration for peroxy ester
13b is correspondingly 3S,6R,8R,10R.
Acknowledgment. The authors thank Dr. S. P. Gunasek-
era of the Harbor Branch Oceanographic Institution for
providing the NMR spectra of the natural ester 1b.
Supporting Information Available: Experimental pro-
cedures for the preparation of compounds 13a, 13b, and all
compounds reported in Schemes 3-5. This material is
available free of charge via the Internet at http://pubs.acs.org.
(19) (a) Ohtani, I.; Kusumi, T.; Ishitsuka, M. O.; Kakisawa, H.
Tetrahedron Lett. 1989, 30, 3147. (b) Ohtani, I.; Kusumi, T.; Kashman,
Y.; Kakisawa, H. J. Am. Chem. Soc. 1991, 113, 4092. (c) Varoglu, M.;
Peters, B. M.; Crews, P. J. Nat. Prod. 1995, 58, 27. (d) Tanala, J.-I.; Higa,
T.; Suwanborirux, K.; Kokpol, U.; Bernardinelli, G.; Jefford, C. W. J. Org.
Chem. 1993, 58, 2999.
OL016943Y
(21) Capon, R. J.; Macleod, J. K. Tetrahedron 1985, 41, 3391.
(22) (a) Schimidazolet, E. W.; Faulkner, D. J. Tetrahedron Lett. 1996,
37, 6681. (b) Higgs, M. D.; Faulkner, D. J. Org. Chem. 1978, 43, 3454.
(20) Corey, E. J.; Jardin, D. S.; Rohloff, J. C. J. Am. Chem. Soc. 1988,
110, 3672.
488
Org. Lett., Vol. 4, No. 4, 2002