Total synthesis of (-)- and (+)-membrenone C.

Article abstract of DOI:10.1021/ol034367v

[reaction: see text] A synthesis of the polypropionate marine defense substance (+)-membrenone C and its enantiomer that starts from (S)-2-methyl-3-(tert-butyldimethylsilyloxy)propanal is described. Key steps include (1) additions of chiral allenylmetal reagents to effect both chain homologation and the concomitant introduction of four stereo centers, (2) a bis-intramolecular hydrosilylation-oxidation sequence to install beta-hydroxy ketone subunits, and (3) a bis-intramolecular aldol reaction to construct the two dihydropyrone termini.

Full text of DOI:10.1021/ol034367v

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
ORGANIC
LETTERS
2003
Vol. 5, No. 10
1729-1732
Total Synthesis of (−)- and
(+)-Membrenone C
James A. Marshall* and Keith C. Ellis
Department of Chemistry, UniVersity of Virginia, CharlottesVille, Virginia 22904
jam5x@Virginia.edu
Received March 2, 2003
ABSTRACT
A synthesis of the polypropionate marine defense substance (+)-membrenone C and its enantiomer that starts from (S)-2-methyl-3-(tert-
butyldimethylsilyloxy)propanal is described. Key steps include (1) additions of chiral allenylmetal reagents to effect both chain homologation
and the concomitant introduction of four stereo centers, (2) a bis-intramolecular hydrosilylation−oxidation sequence to install â-hydroxy
ketone subunits, and (3) a bis-intramolecular aldol reaction to construct the two dihydropyrone termini.
Opisthobranchs (marine nudibranchs) are small, usually
brightly colored snail-like creatures that lack a protective
shell. Their survival in the hostile marine environment is
thought to depend on chemical defense compounds that
render them unpalatable to potential predators. Three such
compounds, membrenones A-C, were isolated in 1993 from
the skin of the marine mollusc Pleurobranchus membrana-
ceus (Montagu, 1815) (Figure 1).1 The two-dimensional
rone (ent-D) that proved to be an isomer of membrenone C
(Scheme 1).2 This compound was one of four (Figure 2,
A-D) with relative stereochemistry consistent with the
reported NMR data (coupling constants for H6/H6 ) 13.7
Scheme 1. Synthesis of an Isomer of Membrenone C2
Figure 1. Partial structures for membrenones A-C.
structures of these three compounds were deduced from
spectral data, but only the relative stereochemistry at C6 and
C7 could be assigned.
Some years later, Perkins and Sampson reported a
bidirectional synthesis of a cis,anti,syn,trans-bis-dihydropy-
a Conditions: (a) HF-pyridine; (b) TFA.
(1) Ciavatta, M. L.; Trivellone, E.; Villani, G.; Cimino, G. Tetrahedron
Lett. 1993, 34, 6791.
10.1021/ol034367v CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/18/2003

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Figure 2. Structure possibilities for membrenone C.
Hz and H9/H10 ) 2.6 Hz).1 Subsequently, these authors
prepared the other three isomers by analogous routes and
found that the cis,syn,syn,trans isomer C exhibited spectral
data in close agreement with that reported for membrenone
C.3 Although significantly lower in magnitude than the
published value, the optical rotation of the synthetic material
was of the same sign as that of natural membrenone C ([R]D
) -581 vs -283, respectively).
We were attracted to the novel structure and potential
bioactivity of membrenone C and felt that it could be readily
assembled through application of allenylmetal and intramo-
lecular hydrosilylation reactions.4,5 The overall plan was a
simple one starting from a chiral aldehyde derived from (S)-
2-methyl-3-hydroxypropionic acid (Figure 3).6
Figure 3. Synthetic plan for (-)-membrenone C.
THF at 55 °C, as previously reported,5 led to a mixture of
compounds. However, by increasing the temperature to 75
°C, we were able to isolate the bis-cyclic siloxane 14 in
quantitative yield.
Initial attempts at direct oxidation of the siloxane 14 to
the diketone met with several problems. The siloxane is quite
Various practical considerations led us to employ the
allenylstannane (P)-1 and the allenylindium reagent (M)-7,
rather than the methylated homologues E and F depicted in
Figure 3, for the allenylmetal additions to aldehyde 2 and
subsequently 6 (Scheme 2). The BF3‚OEt2-promoted reaction
of (P)-1 led to the syn,syn adduct 3 in >95:5 dr and in 82%
yield. Silylation of the secondary alcohol and hydrolysis of
the primary OTBS ether afforded alcohol 5. The derived
aldehyde 6 was subjected to propargylation by the allenylin-
dium reagent (M)-7 prepared from the mesylate of (S)-4-
trimethylsilyl-3-butyn-2-ol, this alcohol being readily ob-
tained through lipase resolution of commercially available
racemic 4-TMS-3-butyn-2-ol.7 The propargylic adduct 8 of
96:4 dr was obtained in 97% yield. Removal of the silyl
groups with TBAF and resilylation of the diol 9 with
TESOTf afforded the diyne 10 in 98% overall yield. The
terminal alkynyl methyl groups were introduced by methy-
lation of the lithiated diyne in quantitative yield. Desilylation
was effected in 98% yield to afford the diol 12.
Scheme 2a
Treatment of diol 12 with tetramethyldisilazane afforded
the bis-silyl ether 13. Subjecting this silane to H2PtCl6 in
a Conditions: (a) BF3‚OEt2, CH2Cl2, -78 °C (82%); (b) TESOTf,
2,6-lutidine (87%); (c) PPTS, EtOH (92%); (d) TBAF, THF (98%);
(e) TESOTf, 2,6-lutidine (100%); (f) BuLi, THF and then (MeO)2SO4
(100%); (g) TBAF, THF (98%); (h) (Me2SiH)2NH, 85 °C (100%);
(i) H2PtCl6, THF, 75 °C (100%).
(2) Perkins, M. V.; Sampson, R. A. Tetrahedron Lett. 1998, 39, 8367.
(3) Perkins, M. V.; Sampson, R. A. Org. Lett. 2001, 3, 123.
(4) Marshall, J. A. Chem. ReV. 2000, 100, 3163.
(5) Marshall, J. A.; Yanik, M. M. Org. Lett. 2000, 2, 2173.
(6) Goodhue, C. T.; Schaeffer, J. R. Biotechnol. Bioeng. 1971, 13, 203.
(7) Marshall, J. A.; Chobanian, H. R.; Yanik, M. M Org. Lett. 2001, 3,
3369.
1730
Org. Lett., Vol. 5, No. 10, 2003

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
polar and required methanol as a cosolvent for the oxidation
reaction. However, the oxidation product, diketo diol 16, is
readily soluble in water and methanol. Hence, product
recovery was problematic. In addition, attempted purification
of the diketone 16 by chromatography on silica gel led to
partial epimerization. For these reasons, we explored an
alternative plan whereby the siloxane would be cleaved with
a hydride nucleophile and the liberated diol converted to the
bis-propionic ester, which would then be subjected to
oxidative removal of the vinylsilanes. We were able to effect
the desired cleavage in high yield with DIBAL-H in toluene
(eq 1), but the diol 15 proved to be difficult to acylate. No
product was obtained after prolonged treatment with propi-
onic anhydride or propionyl chloride with added Et3N and
DMAP. In a model system (see Supporting Information),
the propionate could be obtained, but Tamao oxidation of
the vinylsilane (H2O2, KF, THF-MeOH) caused cleavage
of the propionic ester.5,8
Figure 4. Plan for the synthesis of enantiomeric stereopentads.
We therefore revisited the direct oxidation of the cyclic
siloxane 14 (Scheme 3). By carrying out an exhaustive
presence of Hunig’s base afforded (-)-membrenone C (18)
in 51% yield. The spectral data and rotation closely matched
the published data of Perkins and Sampson.3
As we approached the final stages of the foregoing
synthetic sequence, those investigators reported experiments
that led to their reversal of the assigned configuration for
natural membrenone C.10 Although the configuration of their
synthetic material was secure, their findings suggested to
them that the initial investigators had mistakenly reported
the sign of rotation for both membrenone B and C as minus
rather than plus. Accordingly, the sign agreement between
their synthetic material and the reported rotation for the
natural material was not a valid confirmation of absolute
configuration. As we hoped to evaluate the potential anti-
tumor activity of membrenone C, we decided to prepare both
enantiomers. Initially, we considered repeating our synthetic
sequence with the (R)-enantiomer of aldehyde 2 and the
enantiomeric allenylmetal reagents of Scheme 2. However,
upon further reflection, we formulated a more interesting plan
involving a reordering of the allenylmetal additions (Figure
4).
Thus, instead of forming the 9,10-syn bond first with the
(P)-allenylstannane (P)-1 followed by the 6,7-anti bond with
the (M)-allenylindium (M)-7 reagent, we decided to construct
the 6,7-anti bond first and the 9,10-syn bond second by using
the (P)-allenylindium and (M)-allenylstannane reagents (P)-7
and (M)-1, respectively. In this way, the (S)-aldehyde 2 could
be utilized as the starting material for both enantiomers. This
alternative plan would also extend the scope of the allenyl-
metal reactions to additional aldehyde substrates.
Scheme 3a
a Conditions: (a) MeOH-THF (2:1), KHCO3, KF, H2O2; (b)
CH3CH2CO2H, DMAP, DCC, CH2Cl2.
extraction of the polar diketo diol 16 with ethyl acetate and
directly acylating the crude diol with DCC and propionic
acid, we were able to isolate the diketo dipropionate 17 in
46% overall yield. Interestingly, both propionic anhydride
and propionyl chloride in the presence of Et3N and DMAP
were less effective acylating agents.
The final step of the (-)-membrenone C synthesis entailed
a double aldol cyclization. Experiments with model systems
were conducted by using LiHMDS, NaHMDS, and NaH as
basic catalysts. None produced more than a trace of dihy-
dropyrone. However, the use of TiCl49 proved to be effective.
Slow addition of TiCl4 to diketo diester 17 at -78 °C in the
Accordingly, the (P)-allenylindium reagent (P)-7, derived
from (R)-4-TMS-3-butyn-2-ol mesylate,7 was added to the
(8) Tamao, K.; Kumada, M.; Maeda, K. Tetrahedron Lett. 1984, 25, 321.
Review: Jones, G. R.; Landis, Y. Tetrahedron 1996, 52, 7599.
(9) Oppolzer, W.; Rodriguez, I. HelV. Chim. Acta 1993, 76, 1275.
(10) Sampson, R. A.; Perkins, M. V. Org. Lett. 2002, 4, 1655.
Org. Lett., Vol. 5, No. 10, 2003
1731

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
aldehyde 22, which reacted with the (M)-allenylstannane
(M)-1 to give the homopropargylic alcohol 23 in 95:5 dr
and in 76% yield. Desilylation with TBAF afforded diol ent-9
quantitatively. This diol was identical to that prepared in
Scheme 2 except for the optical rotation, which was opposite
in sign but nearly equal in magnitude ([R]D ) -2.3 for 9
and +2.1 for ent-9). The remaining steps of the synthesis
were carried out as outlined in Scheme 3. The final product
was, as expected, identical to the previous sample of (-)-
membrenone C except for the optical rotation, which was
equal in magnitude but of opposite sign ([R]D ) -22.6 and
+23.5).
Scheme 4a
The foregoing syntheses of (-)- and (+)-membrenone C
reconfirm the relative stereochemistry as assigned by Perkins
and Sampson,3 but as samples of the natural material are no
longer available, the absolute configuration remains unsub-
stantiated.
Acknowledgment. Support for this research was provided
by grant R01 CA90383 from the National Cancer Institute.
a Conditions: (a) (R)-TMSCtCCH(OMs)Me, Pd(OAc)2, PPh3,
InI, THF-HMPA (75%, dr ) 99:1); (b) TBSOTf, CHCl2, 2,6-
lutidine (97%); (c) PPTS, EtOH (90%); (d) BF3‚OEt2, CH2Cl2, -78
°C (76%, dr ) 95:5); (e) TBAF, THF (100%).
Supporting Information Available: Experimental pro-
1
cedures for all compounds and selected H and 13C NMR
spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
(S)-aldehyde 2 affording the anti adduct 19 in 99:1 dr and
in 75% yield (Scheme 4). Silylation and selective desilylation
followed by Swern oxidation of the primary alcohol led to
OL034367V
1732
Org. Lett., Vol. 5, No. 10, 2003