A direct and efficient total synthesis of the tubulin-binding agents ceratamine A and B.; use of IBX for a remarkable heterocycle dehydrogenation

Article abstract of DOI:10.1021/ol900709n

The total synthesis of the tubulin-binding agents ceratamine A and B is reported, along with des-methyl analogs, via a synthetic route that is high-yielding and operationally efficient. The synthetic route involved a Beckmann rearrangement to form an azep

Full text of DOI:10.1021/ol900709n

ORGANIC
LETTERS
2009
Vol. 11, No. 10
2133-2136
A Direct and Efficient Total Synthesis of
the Tubulin-Binding Agents Ceratamine
A and B; Use of IBX for a Remarkable
Heterocycle Dehydrogenation
Robert S. Coleman,* Erica L. Campbell, and Daniel J. Carper
Department of Chemistry, The Ohio State UniVersity, 100 West 18th AVenue,
Columbus, Ohio 43210
coleman@chemistry.ohio-state.edu
Received April 3, 2009
ABSTRACT
The total synthesis of the tubulin-binding agents ceratamine A and B is reported, along with des-methyl analogs, via a synthetic route that is
high-yielding and operationally efficient. The synthetic route involved a Beckmann rearrangement to form an azepine ring precursor, a Knoevenagel
condensation to install the benzylic side chain, and an effective imidazole annulation onto an r-aminoketone precursor with a protected
S-methylisothiourea. Final dehydrogenation proved remarkably facile using IBX.
The marine natural products ceratamine A (1a) and B (1b)
(Figure 1) were isolated from the New Guinean sponge
Pseudoceratina sp. by Roberge and co-workers and were
shown to possess antimitotic activity in a cell-based assay.1
These unusual heterocyclic compounds join a large family
of marine natural products that exhibit antimitotic activity2,3
and were shown4 to arrest cells in mitosis and stimulate
Figure 1. Structures of the ceratamines.
microtubule polymerization in the absence of normally
associated proteins.
The ceratamines do not compete with taxol for binding to
ꢀ-tubulin, whereas taxol, the epothilones, discodermolide,
and the eleuthesides bind to a common site,5 which itself is
distinct from the binding site of the Vinca alkaloids.6 Thus,
the ceratamines may exhibit a unique spectrum of antitumor
activity.4 Preclinical studies of the ceratamines would rely
(1) Manzo, E.; van Soest, R.; Matainaho, L.; Roberge, M.; Andersen,
R. J. Org. Lett. 2003, 5, 4591.
(2) Beckers, T.; Mahboobi, S. Drugs Future 2003, 28, 767
.
(3) Jordan, M. A.; Wilson, L. Nat. ReV. Cancer 2004, 4, 253
(4) Karjala, G.; Chan, Q.; Manzo, E.; Andersen, R. J.; Roberge, M.
.
(5) Nogales, E.; Wolf, S. G.; Khan, I. A.; Luduena, R. F.; Downing,
K. H. Nature (London) 1995, 375, 424.
Cancer Res. 2005, 65, 3040.
(6) Bai, R.; Pettit, G. R.; Hamel, E. J. Biol. Chem. 1990, 265, 17141.
10.1021/ol900709n CCC: $40.75
Published on Web 04/22/2009
 2009 American Chemical Society

Scheme 1
.
Synthesis of Key Intermediate Hydroazepinedione 5
Scheme 2. Construction of the Imidazo[4,5-d]azepine System
on synthetic material as a result of the low natural abundance
of these agents, typical for sponge-derived secondary me-
tabolites.
The ceratamines possess an unusual imidazo[4,5-d]azepine
ring system that presumably results from the oxidative
coupling of a brominated tyrosine and a histidine. Dibro-
motyrosine-containing natural products are typical of sponges
of the order Verongida, which includes the genus Pseudocer-
atina, and many such metabolites possess interesting biologi-
cal activity.7 There have been two reports by Andersen and
co-workers8 on attempted syntheses of the ceratamines,
which although unsuccessful provided important analogs and
information on structure activity relationships. (Waly was
the first to report the synthesis of the imidazo[4,5-d]azepine
ring system.9) Herein, we report a direct and efficient
synthesis of ceratamine A and B that uses a synthetic strategy
that involved the initial construction of a functionalized
aminohydroazepinone skeleton, condensation with S-meth-
ylisothiourea to install the aminoimidazole ring, and a
straightforward dehydrogenation to the azepine ring
system.
6 by displacement of the C10 bromide with a nitrogen-based
nucleophile such as a protected guanidine, to directly install
the 2-aminoimidazole ring, or azide for a more indirect route.
These nitrogen-based nucleophiles underwent conjugate
addition to C11 of the benzylidene ꢀ-ketoamide of 6, in
preference to displacement of the bromide. In an attempt to
block conjugate addition, 6 was treated with sodium thiophe-
nolate, which displaced the C10 bromide. In light of this
lose-lose scenario, we elected to reduce 6, although this
would necessitate a subsequent reoxidation at a later stage.
The Hantzsch ester13 proved to be the most suitable
reagent for conjugate reduction of the double bond of 6 and
provided the desired product 7 in excellent yield (25 °C,
12 h) without over-reduction of the R-bromide. Attempted
imidazole annulation with R-bromoketone 7 using N-Boc
guanidine led to an unprecedented Favorskii rearrangement
leading to the corresponding tetrahydropyridone carboxylic
acid guanidide,14 details of which will be communicated
separately.
Key synthetic intermediate hydroazepinedione 5 (Scheme
1) could be synthesized in good yield by an efficient series
of transformations starting from known ε-lactam 2, readily
prepared from 3-ethoxy-2-cyclohexenone by Beckmann
rearrangement.10 Acid-promoted hydrolysis of the ethyl enol
ether (25 °C, 1 h) revealed the ꢀ-ketoamide of 3 in
quantitative yield. Knoevenagel condensation of 3 with 3,5-
dibromoanisaldehyde (4)11 (25 °C, 1 h) provided 5 in
quantitative yield.
At this stage, acid-promoted bromination12 of 5 occurred
readily (Amberlyst-15, 25 °C, 24 h) to provide the R-bro-
moketone 6 (Scheme 2). Initially, we planned to functionalize
However, displacement of the C10 bromide of 7 with azide
occurred cleanly (25 °C, 1 h) and provided azidoketone 8 in
quantitative yield. Although metal-catalyzed hydrogenolysis
of the azide was not the most obvious method for reduction
because of the potentially labile aryl bromides, using the
more typical reductant Ph3P was unsuccessful because the
adjacent ketone intercepted the intermediate phosphinimine
(7) Fistularins: Aiello, A.; Fattorusso, E.; Menna, M.; Pansini, M.
Biochem. Syst. Ecol. 1995, 23, 377. Psammaplins: Pin˜a, I. C.; Gautschi,
J. T.; Wang, G.-Y.-S.; Sanders, M. L.; Schmitz, F. J.; France, D.; Cornell-
Kennon, S.; Sambucetti, L. C.; Remiszewski, S. W.; Perez, L. B.; Bair,
K. W.; Crews, P. J. Org. Chem. 2003, 68, 3866. Ma’edamines: Hirano, K.;
Kubota, T.; Tsuda, M.; Watanabe, K.; Fromont, J.; Kobayashi, J. Tetra-
hedron 2000, 56, 8107.
(8) Nodwell, M.; Riffell, J. L.; Roberge, M.; Andersen, R. J. Org. Lett.
2008, 10, 1051. Nodwell, M.; Pereira, A.; Riffell, J. L.; Zimmerman, C.;
Patrick, B. O.; Roberge, M.; Andersen, R. J. J. Org. Chem. 2009, 74, 995.
(9) Waly, M. A. J. Prakt. Chem. 1994, 336, 86.
(13) Hantzsch, A. Ber. Dtsch. Chem. Ges. 1881, 14, 1637. For recent
use, see: Garden, S. J.; Guimaraes, C. R. W.; Correa, M. B.; de Oliveira,
C. A. F.; Pinto, A. C.; de Alencastro, R. B. J. Org. Chem. 2003, 68, 8815.
(c) Zolfigol, M. A.; Safaiee, M. Synlett 2004, 5, 827. (d) Yang, J. W.;
Hechavarria Fonseca, M. T.; List, B. Angew. Chem., Int. Ed. 2004, 43,
6660.
(10) Collison, C. G.; Chen, J.; Walvoord, R. Synthesis 2006, 2319.
(11) Lewin, A. H.; Szewczyk, J.; Wilson, J. W.; Carroll, F. I. Tetrahe-
dron 2005, 61, 7144.
(12) Meshram, H. M.; Reddy, P. N.; Sadashiv, K.; Yadav, J. S.
Tetrahedron Lett. 2005, 46, 623.
(14) Coleman, R. S.; Campbell, E. L. Unpublished observations.
2134
Org. Lett., Vol. 11, No. 10, 2009

Scheme 3
.
Dehydrogenation to Fully Oxidized Ring System
Scheme 4. Completion of the Total Synthesis of the Ceratamines
to form a stable oxazaphospholidine (not shown).15 Under
carefully controlled conditions, however, hydrogenolysis over
Pd/BaSO4 catalyst (1 atm H2, 25 °C, 2 h) afforded amine 9
in good yields without appreciable hydrogenolysis of the aryl
bromides. Annulation of the 2-aminoimidazole ring of the
natural products was achieved in two steps by reaction of
the amine of 9 with bis-Boc-protected S-methylisothiourea16
(0 °C, 1 h) to afford 10. This reaction was facilitated by the
addition of a thiophilic metal such as Hg or Ag. Acid-
promoted deprotection was accompanied by a concomitant
condensation reaction sequence (25 °C, 1 h) to provide the
trifluoroacetate salt of 11 in quantitative yield, as a mixture
of tautomers (major tautomer shown). The sequence of
transformations from ε-lactam 2 to 11 occurred in an overall
yield of 41% for eight synthetic operations.
role that the electron-rich imidazole ring plays in the second
oxidation is unclear. There are several mechanistic possibili-
ties including oxidation at the benzylic position followed by
elimination,8 oxidation at C5 R to the carbonyl group, or
oxidation of the imidazole ring. In any case, tautomerization
occurs to form 12, the structure of which is evident in the
1H NMR spectrum in the downfield shifts of the benzylic
C11-H2 (δ 4.15, s) and C8-H2, which change from
diastereotopic multiplets at 3.43 and 3.51 ppm in 11 to an
apparent triplet at 7.18 ppm in 12 (J8,9 ) 8.8 Hz).
Additional confirmation of the structure of 12 came from
examination of the 13C NMR spectrum. Comparing data for
12 with that reported for ceratamine B1 showed that the most
significant differences in chemical shifts occurred at C2 (∆δ
1.1 ppm), C4 (∆δ 0.8 ppm), and C10 (∆δ 0.5 ppm), closest to
the site of the absent N-methyl group, whereas other differences
were trivial (∆δ < 0.2 ppm for C5, C6, C8, C9, and C11).
Most notably, the 13C NMR chemical shifts of carbons
C2, C4, and C10 in 11 dramatically change from those
characteristic of a 2-aminoimidazole (δ 146.5, 117.0, and
120.9) to those of a nonaromatic heterocycle in 12 (δ 176.4,
162.0, and 171.3).
Methylation of the C2-amino group at the fully oxidized stage
of 12 or 13 was impractical using reductive alkylation conditions
due to the sensitivity of the fully oxidized heterocyclic ring
system toward reduction. However, methylation at the stage of
11 could be achieved by quantitative formation of the ethyl
formimidate (80 °C, 4 h) followed by reduction with NaBH418
(0-40 °C, 1 h) to afford 14 (Scheme 4). Oxidation of 14 with
IBX (35 °C, 2 h) afforded ceratamine B (1b). Methylation of
the lactam nitrogen (-10 °C, 1 h) afforded ceratamine A (1a)
in modest yield. Both 1a and 1b provided spectral data identical
with that reported for the natural products.1
The most pressing synthetic issue remaining involved
oxidation to the fully unsaturated ring system of the
ceratamines, in addition to issues surrounding the installation
of the correct methylation pattern of the targets. There is
good precedent for the use of iodine(V) reagents to form
enamides,17a imines,17b and R,ꢀ-unsaturated carbonyl
compounds,17c but there is no precedent for formation of
the C5-C4 R,ꢀ-unsaturated amide needed for the cerata-
mines. Nonetheless, given the number of potentially oxidiz-
able sites in 11, it was expected that some method(s) for the
formal removal of four hydrogen atoms could be developed.
In practice (Scheme 3), a remarkable, high-yielding double
dehydrogenation occurred upon treatment of 11 with IBX
in DMSO (35 °C, 1 h), and the fully oxidized imidazo[4,5-
d]azepine ring system was formed cleanly to provide des-
methyl ceratamine B (12). Methylation of the lactam nitrogen
with iodomethane (-10 °C, 1 h) afforded des-methyl
ceratamine A (13) in modest yield.
The formation of the C8-C9 enamide double bond of 12
using IBX had been examined in model systems, but the
(15) Legters, J.; Thijs, L.; Zwanenburg, B. Tetrahedron Lett. 1989, 30,
4881.
(16) Guo, Z. X.; Cammidge, A. N.; Horwell, D. C. Synth. Commun.
2000, 30, 2933.
The total synthesis of ceratamine A and B was accomplished
by a direct synthetic route starting from ε-lactam 2 in 11 and
(17) (a) Nicolaou, K. C.; Mathison, C. J. N. Angew. Chem., Int. Ed.
2005, 44, 5992. (b) Nicolaou, K. C.; Mathison, C. J. N.; Montagnon, T.
Angew. Chem., Int. Ed. 2003, 42, 4077. (c) Nicolaou, K. C.; Montagnon,
T.; Baran, P. S. Angew. Chem., Int. Ed. 2002, 41, 993.
(18) Roberts, R. M.; Vogt, P. J. J. Am. Chem. Soc. 1956, 78, 4778.
Crochet, R. A., Jr.; Blanton, C. D., Jr. Synthesis 1974, 55.
Org. Lett., Vol. 11, No. 10, 2009
2135

12 steps in an overall yield of 28% and 12%. This synthesis
also provides ready access to many interesting analogs, notably,
the des-methyl ceratamines 12 and 13. A remarkable and high-
yielding double dehydrogenation was the key step for introduc-
tion of unsaturation in the imidazo[4,5-d]azepine ring system
characteristic of the natural products.
(Universite´ Bordeaux 1) for preliminary studies and OBIC
for funding for a Bruker MicrOTOF MS.
Supporting Information Available: Experimental pro-
cedures and characterization of compounds. This mate-
rial is available free of charge via the Internet at http://
pubs.acs.org.
Acknowledgment. E.L.C. was an NIH CBI Predoctoral
Trainee (T32 GM08512). We thank Franc¸ois-Xavier Felpin
OL900709N
2136
Org. Lett., Vol. 11, No. 10, 2009