Asymmetric total synthesis of (-)-agelastatin A using sulfinimine (N-sulfinyl imine) derived methodologies

Article abstract of DOI:10.1021/ol047634l

(Chemical Equation Presented) The asymmetric synthesis of the cytotoxic marine metabolite (-)-agelastatin A (1) has been achieved from the C-ring intermediate 4,5-diamino cyclopenten-2-enone (-)-2. This key intermediate was efficiently prepared from the s

Full text of DOI:10.1021/ol047634l

ORGANIC
LETTERS
2005
Vol. 7, No. 4
621-623
Asymmetric Total Synthesis of
)-Agelastatin A Using Sulfinimine
(−
(N-Sulfinyl Imine) Derived Methodologies
Franklin A. Davis* and Jianghe Deng
Department of Chemistry, Temple UniVersity, Philadelphia, PennsylVania 19122
fdaVis@temple.edu
Received November 17, 2004
ABSTRACT
The asymmetric synthesis of the cytotoxic marine metabolite (−)-agelastatin A (1) has been achieved from the C-ring intermediate 4,5-diamino
cyclopenten-2-enone (
metathesis.
−
)-2. This key intermediate was efficiently prepared from the sulfinimine-derived -diamino ester 4 using ring-closing
r,â
The aim of new synthetic methodologies is to solve important
problems in synthesis. In this regard we have been engaged
in developing new sulfinimine-based methodologies for the
asymmetric synthesis of biorelevant nitrogen compounds.1
Recently we introduced a new procedure for the asymmetric
synthesis of (+)-4-aminocyclopentenone, an important chiral
building block for the synthesis of antitumor carbocyclic
nucleosides, from a sulfinimine-derived N-sulfinyl amino
â-ketodiene using ring-closing metathesis (Figure 1).2 An-
other procedure concerned a new method for the asymmetric
synthesis of syn- and anti-R,â-diamino acids via the addition
of N-protected glycine enolates to enantiopure sulfinimines
(Figure 1).3 It occurred to us that these two methods could
be employed in the synthesis of the key cyclopentane C-ring
core of (-)-agelastatin A (1) and result in an expeditious
total asymmetric synthesis of this novel antitumor agent
(Figure 1).
of (-)-1, along with two minor congeners, from the West
Australian sponge Cymbastela sp.5 Importantly, at low
concentration this alkaloid exhibits potent cytotoxicity against
L1210 in mice and human KB nasopharyngeal tumor cell
lines.6 To date the mechanism of antitumor activity has not
been elucidated. (-)-Agelastatin A (1) is also reported to
selectively inhibit GSK-3â (glycogen synthase kinase-3â)
at low concentrations and could play a role in preventing
Alzheimer’s disease7,8 and inhibiting neuronal apoptosis after
stroke. This alkaloid might also function as an insulin
mimetic.8 Potent insecticidal activity against beet army worm
larvae and corn rootworm has also been reported.5
(4) (a) D’Ambrosio, M.; Guerriero, A.; Debitus, C.; Ribes, O.; Pusset,
J.; Leroy, S.; Pietra, F. J. Chem. Soc., Chem. Commun. 1993, 1305. (b)
D’Ambrosio, M.; Guerriero, A.; Chiasera, G.; Pietra, F. HelV. Chim. Acta
1994, 77, 1895.
(5) Hong, T. W.; Jimenez, D. R.; Molinski, T. F. J. Nat. Prod. 1998,
61, 158.
(6) D’Ambrosio, M.; Guerriero, A.; Ripamonti, M.; Debitus, C.; Waike-
dre, J.; Pietra, F. Helv. Chim. Acta 1996, 79, 727.
(7) Meijer, L.; Thunnissen, A.-M. W. H.; White A. W.; Garnier, M.;
Nikolic, M.; Tsai, L.-H.; Walter, J.; Cleverley, K. E.; Salinas, P. C.; Wu,
Y.-Z.; Biernat, J.; Mandelkow, E.-M.; Kim, S.-H.; Pettit, G. R. Chem. Biol.
2000, 7, 51.
(-)-Agelastatin A (1) is an architecturally unique cytotoxic
tetracyclic alkaloid. Pietra and co-workers first isolated (-)-1
from the axinellid marine sponge Agelas dedromorpha in
1993,4 and Molinski et al. have recently reported the isolation
(1) Zhou, P.; Chen, B.-C.; Davis, F. A. Tetrahedron 2004, 60, 8030.
(2) Davis, F. A.; Wu, Y. Org. Lett. 2004, 6, 1269.
(3) Davis, F. A.; Deng, J. Org. Lett. 2004, 6, 2789.
(8) For a discussion of and leading references on selective inhibition of
GSK-â, see: Hale, K. J.; Domostoj, M. M.; Tocher, D. A.; Irving, E.;
Scheinmann, F. Org. Lett. 2003, 5, 2927.
10.1021/ol047634l CCC: $30.25
© 2005 American Chemical Society
Published on Web 01/20/2005

Scheme 1
1) would be employed to prepare (-)-2 from diamino
ketodiene (-)-3 using ring-closing metathesis (RCM).2 Hale
et al. also used RCM in the construction of their C-ring
intermediate.8 Addition of the enolate of (dibenzylamino)-
acetate to an acrolein-derived sulfinimine was expected to
furnish (-)-4.3 Finally, we envisioned that the conversion
of (-)-2 to (-)-1 could be accomplished using chemistry
similar to that reported by Weinreb and co-workers in their
racemic synthesis of 1.9
Figure 1. New sulfinimine-derived methodologies.
Isolated from natural sources, the scarcity of this biologi-
cally significant alkaloid makes a total enantioselective
synthesis of (-)-1 of prime importance for further biological
evaluation and analogue synthesis. Weinreb and co-workers
employed a hetero Diels-Alder cycloaddition reaction and
a Sharpless/Kresze allylic amination protocol in the first
racemic synthesis of 1.9 The key step in the Feldman and
Saunders enantioselective synthesis of (-)-1 was a unique
vinylcarbene C-H insertion sequence for preparation of the
C-ring core.10 A formal asymmetric synthesis of (-)-1 was
accomplished by Hale et al. in their enantioselective synthesis
of Weinreb’s C-ring intermediate from a Hough-Richardson
aziridine.8 This group recently described the total synthesis
of (-)-agelastatin A (1) from a chiral bicyclic cyclopentene
oxazolidinone intermediate.11 In each of these syntheses
construction of central C-ring core from a bicyclic cyclo-
pentene oxazolidinone proved to be the critical step.12 Our
total asymmetric synthesis of (-)-1 differs primarily in the
synthesis of the C-ring cyclopentene and is presented below.
In our retrosynthetic route to (-)-1, which draws on the
Weinreb,9 Feldman,10 and Hale8,11 syntheses, we chose to
construct a 4,5-diamino cyclopent-2-enone (-)-2 as our
C-ring intermediate (Scheme 1). The methodology we
devised for the synthesis of 4-aminocyclopentenone (Figure
Our synthesis began with the preparation of the requisite
unsaturated R,â-diamino ester (-)-4, by addition of the
acrolein-derived sulfinimine (R)-(-)-6 to 5.0 equiv of the
preformed lithium enolate of ethyl (dibenzylamino)acetate
(5). Three of the four possible diastereoisomers were detected
in a ratio of 18:1:5 with the major syn diastereoisomer (-)-4
being isolated in 73% yield (Scheme 2). Treatment of ester
(-)-4 with 5.0 equiv of lithium N,O-dimethylhydroxylamine
gave the corresponding Weinreb amide (-)-7 in 89% isolated
yield. Deprotection of the N-sulfinyl amino group (TFA/
MeOH) gave the amine, which was not isolated but im-
mediately reacted with pyrrole-2-carboxylic acid, the cou-
pling reagent HBTU, and DIPEA (Hunig’s base) to afford
amide (+)-8 in 88% isolated yield for the two steps. Next
the amide was treated with 2 equiv of allylmagnesium
bromide at 0 °C to give, presumably, an intermediate γ,â-
unsaturated ester (not shown), which was isomerized with
Et3N/EtOH to afford the diamino ketodiene (-)-3 in 85%
yield for the two-step sequence (Scheme 2).13 Refluxing
(-)-3 in DCM with 20 mol % of Grubb’s second generation
catalyst 9 for 12 h resulted in 4,5-diamino cyclopenten-2-
enone (-)-2 in 87% yield, our key C-ring intermediate.
With our C-ring intermediate in hand we planned to
construct the B-ring of (-)-1 by an intramolecular Michael
cyclization in a fashion similar to that described in earlier
syntheses of agelastatin A (1).9,11 Indeed, with an intermediate
similar to 2, Weinreb and co-workers reported that the
cyclization occurred in 61-64% yield in the presence of Cs2-
CO3/MeOH.9 On the other hand, the Hale group, with a
(9) (a) Stein, D.; Anderson, G. T.; Chase, C. E.; Koh, Y.-H.; Weinreb,
S. M. J. Am. Chem. Soc. 1999, 121, 9574. (b) Anderson, G. T.; Chase, C.
E.; Koh, Y.-h.; Stein, D.; Weinreb, S. M. J. Org. Chem. 1998, 63, 7594.
(10) (a) Feldman, K. S.; Saunders, J. C. J. Am. Chem. Soc. 2002, 124,
9060. (b) Feldman, K. S.; Saunders, J. C.; Laci Wrobleski, M. J. Org. Chem.
2002, 67, 7096.
(11) Domostoj, M. M.; Irving, E.; Scheinmann, F.; Hale, K. J. Org. Lett.
2004, 6, 2615.
(12) For an asymmetric synthesis of a C-ring intermediate, see: Baron,
E.; O’Brien, P.; Towers, T. D. Tetrahedron Lett. 2002, 43, 723.
(13) TLC indicated the presences of a new spot that was converted to
(-)-3 on exposure to Et3N.
622
Org. Lett., Vol. 7, No. 4, 2005

Scheme 2
Scheme 3
to debenzylate 12 have been unsuccessful, which suggests
that 12 is not an intermediate in the formation of 13. Twelve-
hour bromination of (-)-13 with NBS in THF, according to
the Feldman protocol, afforded (-)-agelastatin A (1) in 69%
isolated yield.10
In summary, the total asymmetric synthesis of the novel
cytotoxic tetracyclic marine alkaloid (-)-agelastatin A (1)
has been accomplished using new sulfinimine-based meth-
odologies in approximately 11 steps under eight operations
(9% overall yield) from sulfinimine (-)-6 and commercially
available materials. Highlights of our synthesis include the
sulfinimine-mediated enantioselective synthesis of syn-R,â-
diamino ester (-)-4, ring-closing metathesis of diamino
ketodiene (-)-3 to give the C-ring core intermediate (-)-2,
and the D-ring formation by the addition of methyl isocyanate
to (-)-2 under reductive conditions.
species similar to 3, was unable to affect the cyclization using
the Weinreb conditions.11 Nevertheless we found that treat-
ment of (-)-2 with 10 equiv of Cs2CO3/MeOH for 16 min
resulted in a 68% yield of the desired tricyclic ring system
(-)-10 along with trace amounts of cyclopentenone (-)-11.
Longer reaction times, 2 h, produced the enone as the major
product.
Next it was necessary to remove the N-benzyl protecting
groups, treat the free amine with methyl isocyanate to give
the D-ring, and finally brominate to complete the synthesis
of (-)-agelastatin A (1). Unfortunately, hydrogenation under
a variety of conditions (Pd(OH)2, Pd/C, various solvents)
failed to produce characterizable products. Hydrogenation
followed by treatment of the crude reaction mixture with
methyl isocyanate gave similar results. Fortunately, we
discovered that if (-)-10 and methylisocyanate were hydro-
genated together in one pot, N-benzyl debromoagelastatin
A (12) and debromoagelastatin (13) were isolated in 32%
and 47% yields, respectively (Scheme 3). To date attempts
Acknowledgment. This work was supported by a grant
from the National Institute of General Medical Sciences
57870.
Supporting Information Available: Detailed experi-
mental procedures and characterization data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL047634L
Org. Lett., Vol. 7, No. 4, 2005
623