Efficient total synthesis of ammosamide B

Article abstract of DOI:10.1021/ol101215x

A total synthesis of ammosamide B, a metabolite of the marine-derived Streptomyces strain CNR-698, has been executed in nine steps and 6.9% overall yield. The key step involves the condensation of a 4,6-diBoc-protected 1,3,4,6-tetraaminobenzene derivative with dimethyl 2-ketoglutaconate, which effectively constructs the pyrrolidinone ring and the quinoline ring in a single step. This contributes a unique approach to the synthesis of pyrroloquinoline alkaloids that offers the advantages of brevity and relatively high overall yield.

Full text of DOI:10.1021/ol101215x

ORGANIC
LETTERS
2010
Vol. 12, No. 13
3112-3114
Efficient Total Synthesis of
Ammosamide B
P. V. Narasimha Reddy, Biplab Banerjee, and Mark Cushman*
Department of Medicinal Chemistry and Molecular Pharmacology, School of
Pharmacy and Pharmaceutical Sciences, and the Purdue UniVersity Center for Cancer
Research, Purdue UniVersity, West Lafayette, Indiana 47907
cushman@purdue.edu
Received May 26, 2010
ABSTRACT
A total synthesis of ammosamide B, a metabolite of the marine-derived Streptomyces strain CNR-698, has been executed in nine steps and
6.9% overall yield. The key step involves the condensation of a 4,6-diBoc-protected 1,3,4,6-tetraaminobenzene derivative with dimethyl
2-ketoglutaconate, which effectively constructs the pyrrolidinone ring and the quinoline ring in a single step. This contributes a unique approach
to the synthesis of pyrroloquinoline alkaloids that offers the advantages of brevity and relatively high overall yield.
The ammosamides A-C (1-3, Figure 1) are pyrroloquino-
line natural products isolated from the marine Streptomyces
strain CNR-698.1-3 Interest in the ammosamides has been
stimulated by their cytotoxicities in cancer cell cultures, as
well as their abilities to influence tubulin and actin dynamics
through myosin targeting.2,4 More specifically, microtubule
depolymerization and an increase in actin filaments were
observed after administration of a fluorescent ammosamide
B conjugate to HCT-116 cells, and histological staining
suggested that the conjugate bound to several myosin
families.4
The ammosamides and structurally related alkaloids such
as lymphostin (4)5 and the related pyrroloiminoquinone
alkaloids,6 including isobatzelline D (5),7 present syntheti-
cally challenging, densely packed arrays of functional groups.
(1) Guadeˆncio, S. P.; MacMillan, J. B.; Jensen, P. R.; Fenical, W. Planta
Figure 1. Ammosamides A-C and related pyrroloquinolines.
Med. 2008, 74, 1083
(2) Hughes, C. C.; MacMillan, J. B.; Gaudencio, S. R.; Jensen, P. R.;
Fenical, W. Angew. Chem., Int. Ed. 2009, 48, 725–727
.
.
(3) Hughes, C. C.; Fenical, W. J. Am. Chem. Soc. 2010, 132, 2528–
Syntheses of these compounds have generally involved first
construction of quinoline systems followed by elaboration
of the pyrrole derivative,8,9 or they have started from indole
derivatives followed by construction of the quinoline
2529
.
(4) Hughes, C. C.; MacMillan, J. B.; Gaudencio, S. P.; Fenical, W.; La
Clair, J. J. Angew. Chem., Int. Ed. 2009, 48, 728–732
.
(5) Nagata, H.; Ochiai, K.; Aotani, Y.; Ando, K.; Yoshida, M.;
Takahashi, I.; Tamaoki, T. J. Antibiot. 1997, 50, 537–542.
10.1021/ol101215x  2010 American Chemical Society
Published on Web 06/01/2010

Scheme 1
.
Retrosynthetic Analysis of Ammosamide B
Scheme 2. Total Synthesis of Ammosamide B
rings.3,10 To date, only one synthesis of the ammosamides
has been reported, which produced ammosamide B in 17
steps from 4-chloroisatin in an overall yield of 2.7%.3 In
the case of lymphostin, the synthesis proceeds in 21 steps
with an overall yield of 2.0%.10
The present ammosamide B synthesis was predicated on
the hypothesis that the pyrroloquinoline ring system could
proceed with construction of both the pyrrole and the
quinoline rings in a single step through condensation of a
symmetrical, diprotected 1,3,4,6-tetraaminobenzene 6 with
a diester of 2-ketoglutaconic acid (a variation of the
Skraup-Doebner-Von Miller quinoline synthesis) (Scheme
1). If successful, this would constitute a fundamentally
different approach to the synthesis of pyrroloquinoline
alkaloids that could conceivably offer advantages in terms
of overall yield and the number of steps involved. It would
also constitute a formal synthesis of ammosamides A and
C, which have been prepared from ammosamide B.3
Investigation of the strategy outlined in Scheme 1 eventu-
ally resulted in the total synthesis of ammosamide B as
outlined in Scheme 2. Subjection of the commercially
available starting material 8 using ammonia in ethylene
glycol at 140 °C for 3 h afforded the expected diamine 9 as
previously reported.11 Treatment of intermediate 9 with 5
equiv of Boc2O and DMAP in DMF at room temperature
for 12 h afforded the tetra-Boc compound 10 in 80% yield
along with the undesired tri-Boc product in 10% yield, which
were separated chromatographically and characterized. Depro-
tection of the tetra-Boc intermediate 10 with TFA in
methylene chloride at 0 °C for 4 h provided the desired di-
Boc substance 11. Catalytic reduction of the dinitro com-
pound 11 over Pd/C in ethyl acetate at room temperature
for 12 h produced the diamine 12. Condensation of inter-
mediate 12 with dimethyl 2-ketoglutaconic acid in the
presence of PTSA and Cu(OAc)2 in refluxing chloroform
for 8 h, with the mixture open to air, provided the pyrrolo-
quinoline system 14 in the lowest yield of the synthesis
(50%).12 Chlorination of intermediate 14 with N-chlorosuc-
cinimide occurred regioselectively at 60 °C in DMF over a
period of 30 min to afford compound 15. Removal of the
two Boc protecting groups from 15 was carried out with TFA
at room temperature for 6 h, producing the diamine 16 as a
highly polar compound displaying the intense purple color
of ammosamide B. Deprotonation of the lactam 16 with 1.2
equiv of sodium hydride in DMF for 30 min provided the
anion, which was regioselectively alkylated using 1.5 equiv
of methyl iodide to afford the methylated lactam 17, the
penultimate intermediate of the synthesis. The conversion
of 17 to ammosamide B (2) was readily carried out in the
presence of aq ammonium hydroxide in THF at room
temperature for 24 h.13 The spectroscopic and chromato-
graphic properties of synthetic ammosamide B were identical
to those of an authentic sample of the natural product that
was kindly provided by Professor Fenical.
(6) Antunes, E. M.; Copp, B. R.; Davies-Coleman, M. T.; Samaai, T.
Nat. Prod. Rep. 2005, 22, 62–72.
(7) Sun, H. H.; Sakemi, S.; Burres, N.; McCarthy, P. J. Org. Chem.
1990, 55, 4964–4966.
(8) Peat, A. J.; Buchwald, S. L. J. Am. Chem. Soc. 1996, 118, 1028–
1030.
(9) Roberts, D.; Joule, J. A.; Bros, M. A.; Alvarez, M. J. Org. Chem.
1997, 62, 568–577.
(12) Carrigan, C. N.; Bartlett, R. D.; Esslinger, C. S.; Cybulski, K. A.;
Tongcharoensirikul, P.; Bridges, R. J.; Thompson, C. M. J. Med. Chem.
2002, 45, 2260–2276.
(10) Tatsuta, K.; Imamura, K.; Itoh, S.; Kasai, S. Tetrahedron Lett. 2004,
45, 2847–2850.
(11) Boyer, J. H.; Buriks, R. S.; Toggweiler, U. J. Am. Chem. Soc. 1960,
82, 2213–2215.
(13) Makela, M.; Zhang, L. A.; Zetterberg, K.; Hansson, S. Synth. Comm.
1992, 22, 2811–2814.
Org. Lett., Vol. 12, No. 13, 2010
3113

Several aspects of the synthesis deserve further comment:
(1) The condensation of intermediate 9 with dimethyl
2-ketoglutaconate (13) was tried without success, presumably
because of the reduced nucleophilicities of the two amines
resulting from the presence of the two electron-withdrawing
nitro groups.12 (2) The overall yield for formation of
intermediate 11 from 9 could likely be improved by using
the crude, unseparated mixture of 10 and the tri-Boc side
product instead of pure 10. (3) The di-Boc protection of 9
to yield 11 directly was tried by limiting the number of
equivalents of Boc2O, but undesired mixtures of products
were always obtained. (4) The methylation of the sodium
anion derived from 16 had to be carefully monitored by TLC
and the reaction time limited in order to avoid undesired
methylation of the amino groups. (5) Several attempts were
made to prepare 1,2,4,5-tetraaminobenzene from 9 and
employ it as a potential intermediate, but unfortunately, it
proved to be too unstable to be of any practical value.
The ammosamide synthesis outlined in Scheme 2 proceeds
in nine simple steps from a commercially available starting
material, with an overall yield of 6.9%. The general approach
does therefore offer certain advantages over the previous
synthesis of the ammosamides3 and related pyrroloquinoline
natural products. The synthesis appears to be general, and it
should allow access to the preparation of a variety of
structural analogues for investigation of pharmacological
structure-activity relationships. There is general interest in
the design and synthesis of myosin inhibitors because of the
possibility that fine-tuning their structures might allow the
preparation of compounds that could selectively target only
one myosin as opposed to general targeting of multiple
myosin families. This could possibly provide an effective
method for defining the roles of selected myosins in
cytoskeletal dynamics and structuring.
Acknowledgment. This investigation was made possible
by the National Institutes of Health (NIH) through support
with Research Grant No. P01 CA48112. We are grateful to
Bin Sun for early exploratory work on several synthetic
pathways that ultimately proved to be disappointing.
Supporting Information Available: Complete experi-
mental section describing the preparation and characterization
of all synthetic intermediates and ammosamide B, as well
as a complete set of 1H and 13C NMR spectra. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL101215X
3114
Org. Lett., Vol. 12, No. 13, 2010