Total syntheses of (+)- and (-)-cacospongionolide B: New insight into structural requirements for phospholipase A2 inhibition

Article abstract of DOI:10.1021/ja026899x

The first total synthesis of the antiinflammatory marine sponge metabolite (+)-cacospongionolide B has been accomplished in 12 linear steps. The pivotal transformations include a three-step sequence coupling the two main regions of the natural product as well as generating the side chain dihydropyran ring. The activity of the synthetic analogues against bee venom phospholipase A₂ suggests that cacospongionolide B has an enantiospecific interaction with the enzyme that is independent of the γ-hydroxybutenolide moiety. Copyright

Full text of DOI:10.1021/ja026899x

Published on Web 09/06/2002
Total Syntheses of (+)- and (-)-Cacospongionolide B: New Insight into
Structural Requirements for Phospholipase A2 Inhibition
Atwood K. Cheung and Marc L. Snapper*
Department of Chemistry, Merkert Chemistry Center, Boston College, 2609 Beacon Street,
Chestnut Hill, Massachusetts 02467
Received May 14, 2002
The natural product (+)-cacospongionolide B (1), isolated from
the marine sponge Fasciospongia caVernosa by De Rosa and co-
workers in 1995 was shown to possess antimicrobial and cytotoxic
activities.1 Moreover, the natural product’s potent antiinflammatory
activity2 was attributed to its inhibition of secretory phospholipase
A2 (sPLA2). Given the role of inflammation in diseases such as
asthma, psoriasis, and rheumatoid arthritis,3 identifying and devel-
oping potent inhibitors of sPLA2 continues to be of importance.
Figure 1. Retrosynthesis of cacospongionolide B.
The biological activities of cacospongionolide B have been
attributed to its γ-hydroxybutenolide moiety, which is thought to
act as a masked aldehyde. Manoalide,4 (2) with a similar side chain,
has been shown to bind covalently to sPLA2 through the formation
of a Schiff base between the γ-hydroxybutenolide-derived aldehyde
and a lysine at the enzyme-lipid interface.5 Accordingly, (+)-1
has been tested and shown to have inhibitory activity similar to
manoalide against several phospholipases.2
As an initial step toward exploring further the biological activities
of (+)-1, we required access to the natural product, as well as
structural variants. Given its interesting biological activity, it is not
surprising that several reports have appeared describing synthetic
approaches toward cacospongionolide B; however, none have yet
resulted in a total synthesis.6 Herein we report the first total
syntheses of (+)- and (-)-cacospongionolide B,7 as well as our
preliminary studies on its antiinflammatory actiVity and proVide
preliminary insight into the structural requirements for sPLA2
inhibition.
generated at a later stage in the synthesis through a selective ring-
closing metathesis.
The synthesis commences with the preparation of ketone 7, which
is generated in four steps from the known homoallylic alcohol 6
(Scheme 1). Brown’s asymmetric allylboration provided alcohol 6
in 73% yield and 94% ee.9 Alcohol 6 was then converted to ester
8 through an etherification with ethyl bromoacetate in 84% yield
(based on 76% conversion). The saponification of ester 8 and
conversion to the Weinreb amide proceeded in 86% yield for the
two steps.10 Addition of vinyl Grignard into the Weinreb amide
occurred in 83% yield and completes the preparation of vinyl ketone
7 in an enantiomerically enriched fashion.
Scheme 1. Synthesis of Vinyl Ketone 7a
From a retrosynthetic perspective, cacospongionolide B com-
prises two regions connected through an ethylene linker (Figure
1). The trans-decalin portion is coupled through a quaternary carbon
(C9) to the heterocyclic side chain. Given the distance between
relevant stereocenters, absolute stereocontrol is required in the
preparation of each fragment. In our initial analysis, fragments 3
and 4 were to be coupled through an established reductive alkylation
protocol (Plan A).8 While homoallylic iodide 4 is prepared readily
through an enyne metathesis of compound 5, followed by a
hydroboration/oxidation of the resulting diene, all attempts to
alkylate 4 were problematic in that elimination of HI was competi-
tive with alkylation using enolate 3. To circumvent this problem,
we explored a Michael addition of enolate 3 into enone 7 (Plan
B). In this approach the natural product’s dihydropyran ring is
a Reagents: (a) NaH, THF, 0 °C-rt; BrCH2CO2C2H5, 0 °C (84% at 76%
conv.); (b) LiOH, MeOH/H2O (3:1), rt; (c) MeO(Me)NH‚HCl, DIC, DMAP,
TEA, CH2Cl2, 0 °C-rt (86%, two steps); (d) H2CdCHMgBr, THF, -78
to 0 °C, (83%).
As illustrated in Scheme 2, Li/NH3 reduction of the enone 911
followed by trapping of the resulting enolate 3 with vinyl ketone 7
in Et2O provided compound 10 as the major diastereomer with the
desired relative decalin stereochemistry in 70% yield.12 Bis-
olefination of diketone 10 with Ph3PdCH2 formed triene 11 in 60%
yield. With the desired triene 11 in hand, a selective ring-closing
metathesis to install the dihydropyran ring of the side chain in lieu
of the other possible five- and ten-membered ring closures was
examined. Not surprisingly, Grubbs’ metathesis catalyst13generated
dihydropyran 12 selectively in 91% yield.
* To whom correspondence should be addressed. E-mail: marc.snapper@BC.edu.
9
11584
J. AM. CHEM. SOC. 2002, 124, 11584-11585
10.1021/ja026899x CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2. Total Synthesis of Cacospongionolide Ba
Moreover, the inhibition is notable for the synthetic precursor
possessing the furan ring in place of the γ-hydroxybutenolide moiety
(+)-13. These results suggest that the γ-hydroxybutenolide is not
the sole structural feature of the natural product inVolVed in sPLA2
inhibition.
In summary, the first total syntheses of (+)- and (-)-cacospon-
gionolide B have been accomplished in 12 linear steps from
commercially available starting materials. The pivotal transforma-
tions include a three-step sequence that couples the two fragments
of the natural product and generates the dihydropyran ring. The
activity of the analogues against bee venom sPLA2 suggests that
cacospongionolide B has an enantiospecific interaction with the
phospholipase that is independent of the γ-hydroxybutenolide
moiety. To identify improved antiinflammatory agents, efforts are
underway to prepare more potent inhibitors of sPLA2 using the
cacospongionolide scaffolding.
a Reagents: (a) Li/NH3, t-BuOH, Et2O, -33 °C; 7, Et2O, -78 °C (70%);
(b) Ph3PdCH2, DMSO, 75 °C (60%); (c) Cy3P(iMes)(Cl)2RudCHPh,
CH2Cl2, rt (91%); (d) [(COD)RhCl]2, (R)-BINAP, H2 (3 atm), CH2Cl2, 30
°C (91%, 3.3:1 dr); (e) 1 N HCl/THF (1:2), rt (90%); (f) Ph3PdCH2, DMSO,
75 °C (84%); (g) O2, Rose bengal, i-Pr2NEt, 150 W tungsten filament lamp,
CH2Cl2, -78 °C (69%).
Acknowledgment. We thank Dr. Salvatore De Rosa for
providing authentic samples of cacospongionolide B. We also thank
Mr. Jarred Blank for assistance with X-ray crystallographic studies.
The NIH (CAR01-66617) is acknowledged for the financial support
of this research.
With most of the side chain in place, the next challenge was
selective reduction of the decalin C8 exo-olefin. Heterogeneous
hydrogenation catalysts (i.e., Pd/C, Rh/C, Ir-black, Pt-black) proved
nonselective under a variety of conditions, reducing both isolated
olefins, as well as the furan ring. Homogeneous hydrogenation
catalysts can be more discriminating, and in that regard, Wilkinson’s
catalyst reduced selectively the desired disubstituted olefin in high
yield, however with only modest diastereoselectivity (1.7:1) favoring
the desired compound.14 Improved results were obtained with a
chiral rhodium catalyst.15 (R)-BINAP/Rh(I) catalyst (20 mol %) in
CH2Cl2 at 30 °C under 3 atm of H2 provided the desired isomer in
91% yield (based on 76% conversion) as a 3.3:1 mixture of
diastereomers.16 The double diastereoselectivity imparted by the
chiral ligand was slight; the (S)-BINAP ligand offered a 3:1
diastereomeric ratio, again favoring the desired stereochemistry.
With all stereocenters of the target molecule in place, completion
of the synthesis required conversion of the protected C4 ketone
into an exo-olefin. Removal of the acetal under acid conditions in
the compound resulting from hydrogenation of 12, followed by a
Wittig olefination generated furan 13 in 76% overall yield for the
two steps. Finally, photooxidation of the furan moiety under basic
conditions17 unmasked the desired γ-hydroxybutenolide functional-
ity in 69% yield and completed the first total synthesis of (+)-
cacospongionolide B. In an analogous fashion, the enantiomer of
the natural product was also prepared using the appropriate
enantiomeric catalysts and reagents.
Supporting Information Available: Experimental procedures and
data on new compounds (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) (a) De Rosa, S.; Crispino, A.; De Giulio, A.; Iodice, C. J. Nat. Prod.
1995, 58, 1776-1780. (b) Soriente, A.; Crispino, A.; De Rosa, M.; De
Rosa, S.; Scettri, A.; Scognamiglio, G.; Villano, R.; Sodano, G. Eur. J.
Org. Chem. 2000, 947-955.
(2) Garcia Pastor, P.; De Rosa, S.; De Giulio, A.; Paya´, M.; Alcarza, M. J.
Brit. J. Pharm. 1999, 126, 301-311.
(3) Heller, A.; Koch, T.; Schmeck, J.; van Ackern, K. Drugs 1998, 55, 487-
496.
(4) De Silva, E. D.; Scheuer, P. J. Tetrahedron Lett. 1980, 21, 1611-1614.
(5) (a) Potts, B. C. M.; Faulkner, D. J. J. Nat. Prod. 1992, 55, 1701-1717.
(b) Potts, B. C. M.; Faulkner, D. J.; De Carvalho, M. S.; Jacobs, R. S. J.
Am. Chem. Soc. 1992, 114, 5093-5100.
(6) (a) De Rosa, M.; Solladieˆ-Cavallo, A.; Scettri, A Tetrahedron Lett. 2000,
41, 1593-1596. (b) De Rosa, M.; Soriente, A.; Sodano, G.; Scettri, A.
Tetrahedron 2000, 56, 2095-2102. (c) De Rosa, M.; Dell’Aglio, R.;
Soriente, A.; Scettri, A Tetrahedron: Asymmetry 1999, 10, 3659-3662
and references therein.
(7) Reported at the 223rd ACS National Meeting, Orlando, FL, April 7-11,
2002; American Chemical Society: Washington, DC, 2002; ORGN-434.
(8) (a) Smith, A. B., III; Mewshaw, R. J. Org. Chem. 1984, 49, 3485-3489.
(b) Larsen, S. D.; Monti, S. A. J. Am. Chem. Soc. 1977, 99, 8015-8020.
(9) Racherla, U. S.; Liao, Y.; Brown, H. C. J. Org. Chem. 1992, 57, 6614-
6617.
(10) Piscopio, A. D.; Minowa, N.; Chakraborty, T. K.; Koide, K.; Bertinato,
P.; Nicolaou, K. C. J. Chem. Soc., Chem. Commun. 1993, 617-618.
(11) Enone 9 was prepared in >99% ee after recrystallization using a
D-phenylalanine-mediated enantioselective Robinson annulation, followed
by protection of C-4 ketone. (a) Uma, R.; Swaminathan, S.; Rajagopalan,
K. Tetrahedron Lett. 1984, 25, 5825-5828. (b) Hagiwara, H.; Uda, H. J.
Org. Chem. 1988, 53, 2308-2311.
(12) (a) Stork, G.; Rosen, P.; Goldman, N.; Coombs, R. V.; Tsujji, J. J. Am.
Chem. Soc. 1965, 87, 275-286. (b) Stork, G.; Ganem, B. J. Am. Chem.
Soc. 1973, 95, 6152-6153. (c) Boeckman, R. K. J. Am. Chem. Soc. 1973,
95, 6867-6869. (d) Boeckman, R. K. J. Am. Chem. Soc. 1974, 96, 6179-
6181. The minor diastereomer resulting from the enantiomer of 7 was
not detected by 1H NMR (i.e., < 3%).
(13) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953-
956.
(14) Harmon, R. E.; Gupta, S. K.; Brown, D. J. Chem. ReV. 1973, 73, 21-52.
(15) Ohta, T.; Ikegami, H.; Miyake, T.; Takaya, H. J. Organomet. Chem. 1995,
502, 169-176.
Figure 2. Inhibition of bee venom sPLA2.
(16) Stereochemistry confirmed through single-crystal X-ray analysis.
(17) (a) Kernan, M. R.; Faulkner, D. J. J. Org. Chem. 1988, 53, 2773-2776.
(b) Demeke, D.; Forsyth, C. J. Org. Lett. 2000, 2, 3177-3179. Synthetic
material was identical to samples provided by Dr. Salvatore De Rosa.
Inhibition of sPLA2 with synthetic variants of cacospongionolide
B revealed several important aspects (Figure 2). The inhibition is
enantioselective; the natural product (+)-1 is a more potent inhibitor
of bee venom sPLA2 than the unnatural enantiomer (-)-1.
JA026899X
9
J. AM. CHEM. SOC. VOL. 124, NO. 39, 2002 11585