Dragmacidin E synthesis studies. Preparation of a model cycloheptannelated indole fragment

Article abstract of DOI:10.1021/ol0522081

(Chemical Equation Presented) The conversion of N-2,2-dichloropropionyl indole methyl ester into a tetracyclic cycloheptannelated indole model compound for the synthesis of dragmacidin E was accomplished in 10 steps. Key reactions include a Witkop cyclization to fashion a C-C bond at C(4) of the indole nucleus and a subsequent Dieckmann cyclization to deliver the desired cycloheptanoid ring.

Full text of DOI:10.1021/ol0522081

ORGANIC
LETTERS
2005
Vol. 7, No. 24
5449-5452
Dragmacidin E Synthesis Studies.
Preparation of a Model
Cycloheptannelated Indole Fragment
Ken S. Feldman* and Paiboon Ngernmeesri
Department of Chemistry, The PennsylVania State UniVersity,
UniVersity Park, PennsylVania 16802
ksf@chem.psu.edu
Received September 13, 2005
ABSTRACT
The conversion of N-2,2-dichloropropionyl indole methyl ester into a tetracyclic cycloheptannelated indole model compound for the synthesis
of dragmacidin E was accomplished in 10 steps. Key reactions include a Witkop cyclization to fashion a C
nucleus and a subsequent Dieckmann cyclization to deliver the desired cycloheptanoid ring.
−C bond at C(4) of the indole
Dragmacidin E (1) and related structures dragmacidins D
and F represent a small class of bis-indole-derived sponge
metabolites isolated from a Spongosorites sp. found in
Australian waters.1 Their novel architecture and promising
selective phosphatase inhibitory activity has fueled many
synthesis studies on these species,2 with recent preparations
of dragmacidin D3a and F3b,c representing the apotheosis of
this field at the moment. However, the synthesis challenges
posed by dragmacidin E, including (1) the timing of
guanidine introduction, (2) establishing the C(6′′′) quaternary
center with correct relative stereochemistry, and (3) forging
the pyrazinone unit in a sterically demanding environment,
have yet to be addressed. Furthermore, an asymmetric
synthesis of 1 would secure the absolute stereochemistry,
suggested as shown in Scheme 1 by Stoltz,3c and if correlated
with the known absolute stereochemistry of dragmacidins
D and F, would lend credence to the prevailing biosynthetic
hypothesis1,3a connecting these dragmacidins.
These considerations presented both an opportunity and a
challenge; using a tryptophan derivative as a starting point
might permit ready entry into the chiral series, but forming
the necessary C-C bond to C(4) of the indole moiety could
prove problematic. One tryptophan C(4) functionalization
reaction that merited consideration is the Witkop photocy-
clization of N-haloacetyl derivatives,4 a transform that was
used with great success by Moody’s group5 but has seen
only sporadic use since.6 However, whereas this reaction
affords cyclooctannelated indole products with facility,
(1) Capon, R. J.; Rooney, F.; Murray, L. M.; Collins, E.; Sim, A. T. R.;
Rostas, J. A. P.; Butler, M. S.; Carrol, A. R. J. Nat. Prod. 1998, 61, 660-
662.
(2) (a) Whitlock, C. R.; Cava, M. P. Tetrahedron Lett. 1994, 35, 371-
374. (b) Jiang, B.; Smallheer, J. M.; Amaral-Ly, C.; Wuonola, M. A. J.
Org. Chem. 1994, 59, 6823-6827. (c) Miyake, F. Y.; Yakushijin, K.; Horne,
D. A. Org. Lett. 2000, 2, 3185-3187. (d) Kawasaki, T.; Enoki, H.;
Matsumura, K.; Ohyama, M.; Inagawa, M.; Sakamoto, M. Org. Lett. 2000,
2, 3027-3029. (e) Jiang, B.; Yang, C.-G.; Wang, J. J. Org. Chem. 2001,
66, 4865-4869. (f) Yang, C.-G.; Wang, J.; Tang, X.-X.; Jiang, B.
Tetrahedron: Asymmetry 2002, 13, 383-394. (g) Kawasaki, T.; Ohno, K.;
Enoki, H.; Umemoto, Y.; Sakamoto, M. Tetrahedron Lett. 2002, 43, 4245-
4248. (h) Miyake, F. Y.; Yakushijin, K.; Horne, D. A. Org. Lett. 2002, 4,
941-943. (i) Yang, C.-G.; Wang, J.; Jiang, B. Tetrahedron Lett. 2002, 43,
1063-1066. (j) Yang, C.-G.; Liu, G.; Jiang, B. J. Org. Chem. 2002, 67,
9392-9396. (k) Kawasaki, T.; Kouko, T.; Totsuka, H.; Hiramatsu, K.
Tetrahedron Lett. 2003, 44, 8849-8852. (l) Garg, N. K.; Stoltz, B. M.
Tetrahedron Lett. 2005, 46, 2423-2426.
(3) (a) Garg, N. K.; Sarpong, R.; Stoltz, B. M. J. Am. Chem. Soc. 2002,
124, 13179-13184. (b) Garg, N. K.; Caspi, D. D.; Stoltz, B. M. J. Am.
Chem. Soc. 2004, 126, 9552-9553. (c) Garg, N. K.; Caspi, D. D.; Stoltz,
B. M. J. Am. Chem. Soc. 2005, 127, 5970-5978.
(4) (a) Yonemitsu, O.; Cerutti, P.; Witkop, B. J. Am. Chem. Soc. 1966,
88, 3941-3945. (b) Naruto, S.; Yonemitsu, O. Chem. Pharm. Bull. 1980,
28, 900-909.
10.1021/ol0522081 CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/27/2005

Scheme 1. Retrosynthesis Analysis for Dragmacidin E
Scheme 2. Key Witkop/Dieckmann Sequence To Fashion the
Cycloheptannelated Indole Core
proximity, and so, not surprisingly, irradiation of a 5.0 mM
solution of 8 in CH3CN led to reasonably efficient conversion
into the C(4)-cyclized species 9. Yields in the Wipkop
procedure rarely exceed 50%, and it is possible that the added
benefit of a second radical stabilizing chloride within 8
contributed to the relatively higher yield with this substrate.
Presumably, an unobserved intermediate C(5′′′)-methyl,
-chloride-bearing species is formed en route to 9, but a
second photochemically mediated dehydrochlorination in-
tervenes to deliver the alkene of the product. Similar
chemistry has been reported in related systems.5c Conversion
of the cyclooctanoid skeleton of 9 into the carbocyclic
cycloheptanoid bridge of the target occupied a second pivotal
role in the synthesis strategy. Although several routes could
be envisioned, after much experimentation, an efficient and
scalable sequence, which featured treatment of an N-BOC
derivative of the enone 9 with a hydride source, was
identified. Presumably, the unobserved enolate 10, derived
from conjugate addition of [H-], played a central role in this
transformation. Dieckmann closure into the proximal ester
moiety then completes the preparation of 11. This Dieckmann
cyclization lacks the usually cited driving force associated
with high-yielding conversion (deprotonation of an acidic
product), but it likely benefits from relief of steric compres-
sion as the endo-disposed ester unit is incorporated into the
molecular framework. The imide BOC group is cleaved
during this transformation, but it can be readily reinstalled
in the next step. An alternative route to 11, which involved
initial hydrogenation of the alkene within the N-BOC
derivative of 9 to afford an R-methyl ketone and then
treatment with base, was explored briefly but offered no
advantage.
attempts to prepare lower homologues largely met with
failure.5b Thus, application of this useful C(4) functional-
ization reaction to the dragmacidin E problem would require
some modification or extension of the basic transform in
order to secure the desired cycloheptannelated structure of
1.
This line of reasoning led to the retrosynthesis shown in
Scheme 1, wherein dragmadicin E (1) was envisioned to
derive from a cyclocondensation between aminoketone 2 and
the ketoamide 3. The aminoketone 2 embodies appropriate
functionality and stereochemistry along its periphery to
deliver the requisite guanidine unit of the natural product.
The C(6′′′) quaternary center of 2 might be introduced by
formal Strecker chemistry on the ketone within 4, the pivotal
cycloheptannelated indole intermediate. In the key sequence
of a projected synthesis of 2, this cycloheptane ring might
originate from an unusual Dieckmann cyclization within the
cyclooctanoid construct 6, an obvious target for Witkop
photocyclization chemistry.
Synthesis of the model dragmacidin E tetracycle 18
commenced with the dichloroamide derivative 8 of tryp-
tophan methyl ester, available by simple acylation of the
parent amine (97%) (Scheme 2). This Witkop reaction
substrate contains both the N-unprotected indole electron
donor unit and chlorocarbonyl electron acceptor in close
(5) (a) Beck, A. L.; Mascal, M.; Moody, C. J.; Slawin, A. M. Z.;
Williams, D. J.; Coates, W. J. J. Chem. Soc., Perkin Trans. 1 1992, 797-
811. (b) Beck, A. L.; Mascal, M.; Moody, C. J.; Coates, W. J. J. Chem.
Soc., Perkin Trans. 1 1992, 813-820. (c) Mascal, M.; Moody, C. J.; Slawin,
A. M. Z.; Williams, D. J. J. Chem. Soc., Perkin Trans. 1 1992, 823-830.
(6) (a) Nagata, R.; Endo, Y.; Shudo, K. Chem. Pharm. Bull. 1993, 41,
369-372. (b) Mascal, M.; Moody, C. J.; Morrell, A. I.; Slawin, A. M. Z.;
Williams, D. J. J. Am. Chem. Soc. 1993, 115, 813-814. (c) Mascal, M.;
Wood, I. G.; Begley, M. J.; Batsanov, A. S.; Walsgrove, T.; Slawin, A. M.
Z.; Williams, D. J.; Drake, A. F.; Siligardi, G. J. Chem. Soc., Perkin Trans.
1 1996, 2427-2433. (d) Bennasar, M.-L.; Zulaica, E.; Ram´ırez, A.; Bosch,
J. Heterocycles 1996, 43, 1959-1966. (e) Ruchkina, E. L.; Blake, A. J.;
Mascal, M. Tetrahedron Lett. 1999, 40, 8443-8445. (f) Bremner, J. B.;
Russell, H. F.; Skelton, B. W.; White, A. H. Heterocycles 2000, 53, 277-
296.
Two more operations of note were required to process 11
into 18: (1) installation of the spiroimidazolone ring at C(6′′′)
and (2) scission of the imide bridge. A priori, it was not
clear if any benefit attended sequencing step (1) before step
(2) or vice versa, and so the former course was selected for
study first (Scheme 3). The bridging amide 11 could be
5450
Org. Lett., Vol. 7, No. 24, 2005

Scheme 3. Synthesis of a Pentacyclic Dead End
Scheme 4. Synthesis of the Tetracyclic Dragmacidin E Model
System
converted into a single Strecker product 12 cleanly and in
good yield and then into the imidazolone-containing species
13. The relative stereochemistry indicated in 13 is tentative
and based upon the observation of an NOE signal between
the R-positioned proton on C(8′′′) and the NH of the
carbamate derived by acylation of 12. Unfortunately, all
attempts to cleave the imide bridge failed, either returning
starting material 13 or a deprotected version of same missing
one or both BOC groups. It appears that the functionality
surrounding the imide unit effectively encapsulates the
bridge’s carbonyl and deflects all nucleophiles examined
(LiOH/H2O/THF, KOH/DMSO, Cs2CO3/THF/H2O, H2SO4/
H2O, LiAlH4, LiEt3BH, NaBH4, NaBH3NMe2, LiBH3NH2,
NH2NH2, inter alia).
Examination of the alternative functionalization sequence
proved more fruitful (Scheme 4). The apparently less
hindered lactam bridge of the imide derived by BOC
protection of 11 underwent facile hydrolysis under unex-
ceptional conditions to provide the cycloheptanone product
15 in excellent yield. Thus, the imidazolone ring in 13
appeared to provide the decisive steric obstacle to nucleophile
addition for that bridged system. Decarboxylation attended
this hydrolysis, and enolate protonation led to a relatively
unbiased mixture of diastereomeric secondary methyl groups
at C(5′′′) slightly favoring the â-isomer as shown. Treatment
of partially purified samples of 15 with LiOH led to the same
∼ 2:1 mixture of isomers, suggesting that thermodymanic
equilibration prevails upon formation of the cycloheptanone
product. Epimerization at C(5′′′) might not be problematic,
however, provided that there was some exploitable preference
in the next species to be formed, the cyanoamine 16.
Exposure of the diastereomeric mixture 15 to forcing Strecker
conditions generated an intermediate imine, and then con-
densation of that imine with TMSCN furnished a single
diastereomer of the cyanoamine product 16. Molecular
mechanics calculations (Macromodel 9.0) on the intermediate
imine bearing either diastereomeric methyl appendage sug-
gested that the R-methyl species would be preferred by as
much as 3.0 kcal/mol, plausibly as a consequence of
unavoidable A1,3 interactions in the alternative â-methyl
isomer (see the Supporting Information). Furthermore,
inspection of the R-methyl isomer generated by this modeling
process held forth the promise that the kinetically preferred
trajectory of cyanide addition would follow a pathway
opposite to the pseudoaxial methyl unit (Supporting Informa-
tion). That one isomer was formed in this Strecker sequence
was intriguing, but unambiguous stereochemical assignment
had to await preparation of a derivative with more diagnostic
NMR signals. Toward this end, conversion of aminonitrile
16 into the aminocarbamate 17 afforded a species whose
relative stereochemistry could be ascertained by initial
application of HMBC/HMQC NMR techniques to identify
proton resonances and then by NOE measurements (cf. 17)
to assign the stereochemistry. Interestingly, attempts at
hydrogenation (various Pd or Pt sources/H2 at various
pressures) of the carbamate nitrile derived from 16 (ClCO2-
CH3, K2CO3) led only to recovered starting material or to a
secondary amine presumably resulting from reduction of the
N-carbomethoxyimine derived from HCN ejection. The
difference in succeptibility to hydrogenation between the
carbamate nitrile prepared from 12 and that from 16 is
striking but defies ready explanation. Fortunately, the cobalt-
Org. Lett., Vol. 7, No. 24, 2005
5451

mediated reduction methodology of Battersby,7 which failed
with the carbamate nitrile derived from 12, worked superbly
in this case.
and without complications. No evidence for oxidation
products at the alternative benzylic position C(5′′′) was
detected, and the desired C(8′′′) ketone-containing product
18 was delivered in good yield.
The possibility exists that under the harsh conditions of
the Strecker reaction, epimerization of the NHBOC-bearing
C(7′′′) center could occur, much as it does at C(5′′′).
Examination of 16 and 17 by polarimetry led to the
disappointing but incontrovertable conclusion that this
concern was borne out in practice, as largely racemic material
was isolated. In fact, unreacted ketone 15 recovered from
incomplete conversion displayed only minimal optical activ-
ity, indicating that epimerization at C(7′′′) was unavoidable
under these imine-forming reaction conditions. However,
even without preservation of the starting chirality of tryp-
tophan, further exploration of amine 17’s downstream
functionalization chemistry seemed warranted.
In summary, the development of a Witkop/Dieckmann
cyclization strategy permitted the synthesis of an advanced
but essentially racemic dragmacidin E model system from
tryptophan methyl ester in 11 steps. Efforts to apply this
chemistry to dragmacidin E itself are underway, and results
will be reported in due course.
Acknowledgment. Financial support from the National
Institutes of Health, General Medical Sciences (GM 35727),
is gratefully acknowledged.
The final two operations of this model system synthesis
involved base-promoted imidazolone closure (with concomi-
tant BOC removal from the indole ring) and DDQ-mediated
oxidation at C(8′′′). Both transformations occurred smoothly
Supporting Information Available: Experimental pro-
cedures and full characterization data for 8, 9, 11-13, and
15-18. This material is available free of charge via the
Internet at http://pubs.acs.org.
(7) Aucken, C. J.; Leeper, F. J.; Battersby, A. R. J. Chem. Soc., Perkin
Trans. 1 1997, 2099-2109.
OL0522081
5452
Org. Lett., Vol. 7, No. 24, 2005