A formal total synthesis of the marine diterpenoid diisocyanoadociane

Article abstract of DOI:10.1021/ol061228f

A formal total synthesis of diisocyanoadociane, a marine diterpenoid with potent antimalarial properties, has been completed. The synthesis begins with a phenanthrenoid precursor that is transformed into a pyrene-derived intermediate by means of an intram

Full text of DOI:10.1021/ol061228f

ORGANIC
LETTERS
2006
Vol. 8, No. 15
3395-3398
A Formal Total Synthesis of the Marine
Diterpenoid Diisocyanoadociane
Kelly A. Fairweather and Lewis N. Mander*
Research School of Chemistry, Australian National UniVersity,
Canberra, ACT 0200, Australia
mander@rsc.anu.edu.au
Received May 19, 2006
ABSTRACT
A formal total synthesis of diisocyanoadociane, a marine diterpenoid with potent antimalarial properties, has been completed. The synthesis
begins with a phenanthrenoid precursor that is transformed into a pyrene-derived intermediate by means of an intramolecular Michael reaction.
Nitrogen functionality is introduced via a double Curtius reaction.
Numerous diterpenes based on a wide range of carbon
skeletons possessing one or more isonitrile substituents have
been isolated from marine sponges.1 Many have been shown
to have significant in vitro anti-malarial activity with the most
potent being diisocyanoadociane (1), isolated originally from
an Amphimedon species2 and subsequently from Cymbastela
hooperi.3 The combination of biological activity4 with the
unusual pyrene-based structure 2 has aroused the interest of
synthetic chemists, and an elegant enantioselective synthesis
(60% ee) based on sequential Diels-Alder cycloadditions
has been reported by Corey and Magriotis.5 In the final stages
of their synthesis, however, the isonitrile groups were
introduced without diastereochemical control. This intriguing
molecule had also captured our attention, and we now
disclose an alternative formal synthesis of (()-1 in which
all 10 stereogenic centers have been introduced with
complete diastereoselectivity.6 An outline of our strategy is
provided in Scheme 1 and begins with the proposed
construction of the phenanthrene derived structure 3 as
illustrated.7
Thus, Birch reductive alkylation, then Lewis acid induced
cyclization, following earlier precedents, could be expected
to furnish 3 in good yield.8 We envisaged completion of the
pyrene structure by means of the intramolecular Michael
reaction 7 f 8, and in anticipation of this step we hoped to
introduce a propionyl side chain into 3 by means of a
(1) Chang, W. C. Progress in the Chemistry of Natural Products;
Springer, Wien: New York, 2000; Vol. 80.
(2) Baker, J. T.; Wells, R. J.; Oberha¨nsli, W. E.; Hawes, G. B. J. Am.
Chem. Soc. 1976, 98, 4010.
(3) Wright, A. D.; Ko¨nig, G. J. Nat. Prod. 1996, 59, 710.
(4) Wright, A. D.; Wang, H.; Gurrath, M.; Ko¨nig, G. M.; Kocak, G.;
Newmann, G.; Loria, P.; Florey, M.; Tilley, L. J. Med. Chem. 2001, 44,
873.
(6) The numbering system of the parent structure, namely “isocycloam-
philectane” (2) is based on its presumed biogenetic relationship to the
amphilectane skeleton (cf. ref 1) and will be used throughout this Letter.
(7) (a) Hook, J. M.; Mander, L. N.; Urech, R. Synthesis 1979, 374. (b)
Cossey, A. L.; Gunter, M. J.; Mander, L. N. Tetrahedron Lett. 1980, 21,
3309. (c) Rogers, D. H.; Frey, B.; Roden, F. R.; Russkamp, F.-W.; Willis,
A. C.; Mander, L. N. Aust. J. Chem. 1999, 52, 1093.
(8) Crabtree, S. R. Ph.D. Dissertation, Australian National University,
1990.
(5) Corey, E. J.; Magriotis, P. A. J. Am. Chem. Soc. 1987, 109, 287.
10.1021/ol061228f CCC: $33.50
© 2006 American Chemical Society
Published on Web 06/22/2006

Scheme 1
Friedel-Crafts-type acylation of the styrene double bond.
was carried out at -78 °C; at higher temperatures, the newly
formed carbonyl function underwent intramolecular cycliza-
tion onto the aromatic ring B. To reduce the styrene bond,
4 was submitted to reduction by lithium in liquid ammonia,
affording hydroxy ketone 9, a somewhat surprising result
given the absence of a proton source (other than ammonia).
The all-cis structure was provisionally assigned as indicated
from extensive NMR studies and is consistent with subse-
quent analyses. After masking the hydroxyl as a MOM ether,
the carbonyl function was reduced (LiAlH4) to give an
epimerically pure alcohol of unspecified configuration that
was protected as the benzoate 10. Hydroboration proved to
be unsatisfactory for the introduction of oxygen at C-20, so
we examined epoxidation with a view to subsequent rear-
rangement to the 20-one. When the MOM group was found
to interfere with epoxidation, it was replaced by acetate, and
then the reaction proceeded smoothly to afford 11 as a
mixture of epimers. BF3‚Et2O-induced rearrangement then
gave 12, again as a mixture of diastereomers. However,
subsequent base-catalyzed hydrolysis of the acetate function
with concomitant loss of the superfluous angular substituent
(retro-aldol as planned) led to a single diastereomer, pre-
sumed to be 13 in the expectation that the product would
possess the more stable trans-fused ring system with the 16-
methyl group equatorial. To elaborate the quaternary center
at C-20, a Wittig reaction was carried out with methoxy-
methylenetriphenyl phosphorane, followed by hydrolysis to
aldehyde 15. C-Methylation of the aldehyde was unsuccess-
ful, however, and so we abandoned the original strategy
(Figure 1a) and submitted the enol ether 14 to a modified
Simmons-Smith methylenation.11 The resulting cyclopro-
pane derivative 16 was then treated with acid at reflux to
give the desired aldehyde 17.12 NMR analysis again estab-
lished the stereochemical relationships between the substit-
uents on ring-D, the most salient feature of which was the
cis relationship between the formyl group and H-12 as
established by NOE difference spectra. It was thus estab-
lished that methylenation had taken place on the more
exposed exo face, leading to the desired diastereomer.
To maintain stereochemical control in the introduction of
the isonitrile groups, we planned to establish the quaternary
centers at C-7 and C-20 by stereochemically controlled
R-methylation of methoxycarbonyl or formyl substituents at
these loci and then apply the Curtius rearrangement to the
corresponding acyl azides.9 As illustrated in Figure 1, we
Figure 1. Stereocontrolled elaboration of the C-7 and C-20
quaternary centers.
expected simple alkylation at C-20 to proceed along the
equatorial vector (steric control), whereas to achieve “axial”
alkylation at C-7, we anticipated that the involvement of the
C-6 carbonyl group (stereoelectronic control) would be
necessary.10
Elaboration of these two centers would necessarily be
under kinetic control, as would be the stereochemistry at C-8
following Li/NH3 reduction of enone 6. The remaining
stereogenic centers, however, should be subjected to ther-
modynamic control by virtue of their relationship to carbonyl
groups at C-2, C-6, and C-20, respectively, at the appropriate
stage of the sequence.
The first stage of the synthesis is outlined in Scheme 2.
Thus, Birch reduction of methyl 2-methoxy-5-methylben-
zoate with lithium and in situ alkylation followed by BF3‚
Et2O-induced cyclization afforded 3 in good yield. Then,
AlCl3-catalyzed acylation of the styrene alkene bond with
propionyl chloride afforded 4, provided that this last reaction
(9) Cf. Piers, E.; Llinas-Brunet, M. J. Org. Chem. 1989, 54, 1483.
(10) Eliel, L. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic
Compounds; Wiley-Interscience: New York, 1994; p 901.
(11) Yang, Z.; Lorenz, J. C.; Shi, Y. Tetrahedron Lett. 1984, 39, 8621.
(12) Wenkert, E. Acc. Chem. Res. 1980, 13, 27.
3396
Org. Lett., Vol. 8, No. 15, 2006

Scheme 2
Scheme 3
procedure.15 Subsequent acylation with methyl cyanofor-
mate,16 followed by C-methylation, then proceeded in
excellent yield to afford 24. From model studies, we expected
that removal of the TBDMS group with TBAF would almost
certainly result in cyclization of the liberated hydroxyl onto
the enone function. The enone was therefore reduced with
17
NaBH4/CeCl3 prior to deprotection. After desilylation, a
double oxidation with the Dess-Martin periodinane18 af-
forded dione 25 in preparation for the pivotal Michael
reaction (Scheme 4).
An inspection of molecular models indicated that the
Michael reaction was unlikely to take place with 25 since it
would require the D-ring to adopt a boat conformation to
bring the participating functions within bonding distance, and
even then orbital overlap between the reacting centers would
be poor. It appeared that an R-configured side chain could
assume the necessary geometry, however. When treatment
of 25 with a variety of bases returned starting material, it
seemed that the axially oriented side chain was more stable
in the â-configuration, despite its axial nature. Nevertheless,
one could expect there to be a sufficient amount of the
R-epimer in equilibrium to undergo the desired cyclization.
After extensive experimentation, it was found that heating a
solution of 25 in ethanediol and triethylamine for 4 days at
70 °C effected cyclization, conditions that were selected in
The next stage of the synthesis, as outlined in Scheme 3,
involved extensive functional group manipulation, including
repeated lithium in ammonia reductions and establishment
of the second quaternary center at C-7. Thus, aldehyde 17
was oxidized to acid 18 (NaClO2, 80% yield from 14)13 and
the benzoate group replaced with a TBDMS function,
following which, Birch reduction and hydrolysis in acetic
acid of the dihydroanisole product afforded the expected â,γ-
unsaturated enone which could be isomerized by anhydrous
HCl in THF to enone 20. A second lithium/ammonia
reduction with trapping of the resulting lithium enolate by
methyl cyanoformate14 afforded a modest yield of the â-keto
ester 21, but C-methylation of this intermediate using a
variety of bases gave none of the desired product.
Alternatively, enone 20 was simply reduced with Li/NH3
to 22 and the isomeric enone 23 prepared using the Saegusa
(15) Ito, Y.; Hirao, T.; Saegusa, T. J. Org. Chem. 1978, 43, 1011.
(16) Mander, L. N.; Sethi, P. A. Tetrahedron Lett. 1983, 24, 5425.
(17) Luche, J. L. J. Am. Chem. Soc. 1978, 100, 2226.
(13) Dalcanale, E.; Montanari, F. J. Org. Chem. 1986, 51, 567.
(14) Crabtree, S. R.; Chu, W. L. A.; Mander L. N. Synlett 1990, 169.
(18) Dess, D. B.; Martin, J. C.J. Org. Chem. 1983, 48, 4155.
Org. Lett., Vol. 8, No. 15, 2006
3397

groups in 28 was effected with sodium propanethiolate,22
while the deoxygenated ester from 26 was treated with
lithium hydroxide to hydrolyze the C-7 ester function before
liberating the C-20 carboxyl with propanethiolate. Preparation
of the diacyl chloride and thence the diacyl azide 29
proceeded smoothly, as did the Curtius rearrangement to 30,
which was effected by heating in toluene at reflux.23
Hydrolysis in concd HCl then afforded diamine 31, which
Scheme 4
1
gave identical mass spectra as well as H and 13C NMR
spectra (measured on the TFA salts) to those obtained from
an authentic sample obtaineded by hydrolysis of 1. Since
31 has been reconstituted to 1,24 the formal total synthesis
of diisocyanoadociane (1) is complete. While all 10 stereo-
genic centers have been introduced with the correct relative
stereochemistry, we have yet to address the question of
enantioselectivity. However, the initial Birch reductive
alkylation, if carried out on an appropriate chiral benzamide,25
should resolve this remaining stereochemical issue.
Acknowledgment. We are indebted to Professor Mary
Garson (University of Queensland) for the provision of
authentic samples of diisocyanoadociane (1) and diamine 31;
to Dr Jamie Simpson (Monash University) for helpful advice;
to Bruce Twitchin and Tony Herlt (ANU) for technical
assistance; to Tony Willis (ANU) for X-ray studies; and to
Chris Blake (ANU) for assistance with high-field NMR
spectra. Funding for an Avance 800 MHz NMR spectrometer
by the Australian Research Council (LIEF grant LE0346876)
and the provision of an Australian Postgraduate Award plus
an Alan Sargeson Ph.D. Supplementary Scholarship in
Chemical Sciences to K.A.F. are also gratefully acknowl-
edged.
view of those reported by Corey et al. for a difficult
intramolecular Michael reaction during the synthesis of
longifolene.19 No other solvent was effective, and unfortu-
nately, trans-esterification of the 7-ester function with the
glycol had occurred, resulting in a 4:3 mixture of 26 and
27, complicating subsequent manipulations. After confirming
the structure of the latter compound by single-crystal X-ray
analysis,20 the diketone was converted into the deoxy
derivative 28 by applying the Barton-McCombie deoxy-
genation procedure to the dixanthate derived from the 2,6-
diol.21 After protecting 26 as the MOM ether, this product
was similarly deoxygenated. Demethylation of both ester
Supporting Information Available: Experimental details
and copies of selected 1H and 13C NMR spectra. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL061228F
(21) Barton, D. H. R.; McCombie, S. W. J. Chem. Soc., Perkin Trans.
1 1975, 1574.
(22) Bartlett, P. A.; Johnson, W. S. Tetrahedron Lett. 1970, 4459.
(23) Smith, P. A. S. Org. React. 1946, 3, 337
(19) Corey, E. J.; Ohno, M.; Mitra, R. B.; Vatakencherry, P. A. J. Am.
Chem. Soc. 1964, 86, 478.
(20) Willis, A. C. Unpublished results.
(24) Simpson, J. S.; Garson, M. J. Org. Biomol. Chem. 2004, 2, 939.
(25) Schultz, A. G.; Macielag, M.; Podhorez, D. E.; Suhadolnik, J. C.;
Kullnig, R. K. J. Org. Chem. 1988, 53, 2456.
3398
Org. Lett., Vol. 8, No. 15, 2006