Synthesis of axinohydantoins

Article abstract of DOI:10.1021/jo020063v

A short synthesis of the hydantoin-containing marine sponge metabolites axinohydantoins is described. A key feature of the synthesis is a putative biomimetic, intramolecular cyclization of α-functionalized imidazolone 5, which affords the tricyclic pyrroloazepinone framework comprising 6. In addition, the conversion of imidazolones to α,β-unsaturated hydantoins is outlined and represents a new approach to these heterocyclic systems.

Full text of DOI:10.1021/jo020063v

4
498
J . Org. Chem. 2002, 67, 4498-4500
Syn th esis of Axin oh yd a n toin s
Ana Carolina Barrios Sosa, Kenichi Yakushijin, and David A. Horne*
Department of Chemistry, Oregon State University, Corvallis, Oregon 97331
horned@ucs.orst.edu
Received J anuary 28, 2002
A short synthesis of the hydantoin-containing marine sponge metabolites axinohydantoins is
described. A key feature of the synthesis is a putative biomimetic, intramolecular cyclization of
R-functionalized imidazolone 5, which affords the tricyclic pyrroloazepinone framework comprising
6
. In addition, the conversion of imidazolones to R,â-unsaturated hydantoins is outlined and
represents a new approach to these heterocyclic systems.
4
5
Over the last two decades, a series of tricyclic pyrrole
cocyamidine unit. In previous work from this lab, the
6
natural products have been isolated from various genera
of marine sponges.1 Each share in common a fused
bicyclic pyrrolo[2,3-c]azepine core to which either a
synthesis and stereoselective conversion of hymenin to
(Z)-hymenialdisines was reported (eq 1). Formation of the
thermodynamically less stable (E)-isomer was not ob-
served. In the present study, a similar approach for
installing the hydantoin unit of axinohydantoins 1 and
2 was envisioned from hymenin analogues 6. The prob-
lem of accessing the hydantoin nucleus from an imida-
zolone precursor as well as forming the less thermody-
namically stable (E)-axinohydantoins by this approach
remained an open question. Herein, we report the first
syntheses of (E)- and (Z)-debromoaxinohydantoins 1b and
2b and (E)-and (Z)-bromoaxinohydantoins 1c and 2c.
2
-aminoimidazole, glycocyamidine, or hydantoin unit is
2
appended. In 1990, Pettit and co-workers isolated (E)-
axinohydantoin (1a ) from the sponge Axinella, and its
structure was determined by X-ray crystallography.
Subsequently, the isolation of (Z)-axinohydantoin (2a )
and (Z)-debromoaxinohydantoin (2b) from Stylotella au-
rantium and Hymeniacidon sp., respectively, has been
reported by two groups.3 Axinohydantoins have been
shown to inhibit protein kinase C.3
a
The biosynthetic pathway to axinohydantoins is un-
known but plausible pathways may involve linear pre-
cursors such as R-functionalized imidazolones 5. An
(
4) (a) Sharma, G. M.; Buyer, J . S.; Pomerantz, M. W. J . Chem. Soc.,
The hydantoin-containing axinohydantoins are struc-
turally related to hymenialdisines which contain a gly-
Chem. Commun. 1980, 435-436. (b) Cimino, G.; DeRosa, S.; DeStefano,
S.; Mazzarella, L.; Puliti, R.; Sodano, G. Tetrahedron Lett. 1982, 23,
67-768. (c) Kitagawa, I.; Kobayashi, M.; Kitanaka, K.; Kido, M.;
7
Kyogoku, Y. Chem. Pharm. Bull. 1983, 31, 2321-2328. (d) Supriyono,
A.; Schwarz, B.; Wray, V.; Witte, L.; Muller, W. E. G., van Soest, R.;
Sumaryono, W.; Proksch, P. Z. Naturforsch. 1995, 50c, 669-674. (e)
Williams, D. H.; Faulkner, D. J . Nat. Prod. Lett. 1996, 9, 57-64. (f)
Eder, C.; Proksch, P.; Wray, V.; Steube, K.; Bringmann, G.; van Soest,
R. W. M.; Sudarsono; Ferdinandus, E.; Pattisina, L. A.; Wiryowidagdo,
S.; Moka, W. J . Nat. Prod. 1999, 62, 184-187.
(
1) Faulkner, D. J . Nat. Prod. Rep. 2001, 18, 1-49 and references
therein.
(2) Pettit, G. R.; Herald, C. L.; Leet, J . E.; Gupta, R.; Schaufelberger,
D. E.; Bates, R. B.; Clewlow, P. J .; Doubek, D. L.; Manfredi, K. P.;
R u¨ tzler, K.; Schmidt, J . M.; Tackett, L. P.; Ward, F. B.; Bruck, M.;
Camou, F. Can. J . Chem. 1990, 68, 1621-1624.
(3) (a) Patil, A. D.; Freyer, A. J .; Killmer, L.; Hofmann, G.; J ohnson,
(5) (a) Xu, Y.-z.; Yakushijin, K.; Horne, D. A. J . Org. Chem. 1997,
62, 456-464. (b) Barrios Sosa, A. C.; Yakushijin, K.; Horne, D. A. J .
Org. Chem. 2000, 65, 610-611.
R. K. Nat. Prod. Lett. 1997, 9, 201-207. (b) Inaba, K.; Sato, H.; Tsuda,
M.; Kobayashi, J . J . Nat. Prod. 1998, 61, 693-695. Although axino-
hydantoins were subsequently referred to as spongiacidins, we opted
to use the name originally given as axinohydantoins.
(6) Kobayashi, J .; Ohizumi Y.; Nakamura H.; Hirata, Y.; Waka-
matsu, K.; Miyazawa, T. Experientia 1986, 42, 1064-1065.
1
0.1021/jo020063v CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/10/2002

Synthesis of Axinohydantoins
Sch em e 1
J . Org. Chem., Vol. 67, No. 13, 2002 4499
Sch em e 3
Sch em e 4
Sch em e 2
dibromopyrrole carboxamide 5a to trifluoroacetic acid
produced only a minor amount (10%) of the desired
product 6a . The major product of the reaction is spiro-
lactam 7 which was isolated in 60% yield. The relative
stereochemistry of 7 was determined from NOE data
1
which suggested a cis orientation of H and the imida-
zolone ring. Hydrolysis of the bromoimine functionality
of proposed intermediate A (R ) Br) is thought to account
for the formation of spirolactam 7. In an attempt to
increase the yield of desired tricyclic azepinone frame-
work of 6, cyclization of R-debromopyrrole analogues 5b
and 5c was pursued. These analogues lack an R-bro-
mopyrrole substituent and should not undergo hydrolysis
to the spirolactam. Thus, intermediate A (R ) H) could
undergo potential rearrangement to the pyrroloazepinone
skeleton either by a 1,2-alkyl shift or through a reversible
ring opening-cyclization process. When R-debromo ana-
logues 5b and 5c were subjected to the cyclization
conditions, a modest 2-3-fold increase in yield of azepi-
nones 6b and 6c was observed.
intramolecular cyclization event would establish the
formation of the pyrroloazepinone core (Scheme 1). Initial
investigations focused on the preparation of 5. Starting
from commercially available ornithine as its methyl ester,
aminopropyl imidazolone 3 was obtained in good yield
after a two-step, single-pot sequence that involves an
Akabori reduction followed by condensation with potas-
In an alternative approach for assembling the tricyclic
core of 6a , the heterodimerization of commercially avail-
able 2-imidazolone and bicyclic olefin 8 was pursued
(Scheme 3). This approach is patterned after a similar
reaction of 8 with 2-aminoimidazole, which resulted in
7
sium cyanate. Introduction of the R-methoxy substituent
5
an efficient synthesis of hyemin. In the present case,
was achieved by treatment of 3 with NCS in methanol.
This produced oxidized imidazolone 4 in good yield.8
Acylation of 4 with the requisite trichloroacetyl pyrrole9
gave carboxamides 5a -c.
however, the use of 2-imidazolone was only partially
successful. 2-Imidazolone, unlike 2-aminoimidazole, is
prone to self-dimerization under acidic conditions, and
this process is accelerated in strong acids such as CH
SO H. Using trifluoroacetic acid, however, a 37% yield
of azepinone 6a was obtained from the reaction of 8 and
-imidazolone. Although the yield is modest, unreacted
3
-
Next, the acid-facilitated intramolecular cyclization of
3
5
to 6 was investigated (Scheme 2). This key step forms
the basis of our approach to the tricyclic pyrroloazepinone
framework that comprises axinohydantoins. Exposure of
2
starting material 8 can be recovered and recycled for
further use.
(
7) Originally performed with thiocyanate, see: (a) Akabori, S. Ber.
With tricyclic azepinones 6 in hand, the conversion of
the imidazolone unit to the requisite hydantoin was
pursued (Scheme 4). Using related conditions developed
for the installation of the R,â-unsaturated imidazolidi-
1
1
933, 66, 151-165. (b) Lawson, A.; Morley, H. V. J . Chem. Soc. 1955,
695-1698.
(8) Barrios Sosa., A. C.; Yakushijin, K., Horne, D. A. Org. Lett. 2000,
2
, 3443-3444.
9) Bailey, D. M.; J ohnson, R. E. J . Med. Chem. 1973, 16, 1300-
302.
(
5
b
1
none (glycocyamidine) functionality in hymenialdisines,

4
500 J . Org. Chem., Vol. 67, No. 13, 2002
Barrios Sosa et al.
Sch em e 5
moaxinohydantoin (1b) and the mono-reduction product,
(E)-isoaxinohydantoin (10). Spectroscopic data of syn-
thetic 2b was in agreement with data reported for the
3
natural material. Attempts to drive the debromination
to completion using longer reaction times (8 h) led to
competing reduction of the C10-C11 double bond which
produced 11 as a 4:1 mixture of diastereomers. An
explanation for the observed difference in reactivity of
isomers 1c and 2c remains elusive at this time. One
possibility is steric interactions between the (E)-hydan-
toin appendage and the catalyst may hinder complete
removal of the 3-bromine substituent prior to over-
reduction.
Finally, (E)-bromoaxinohydantoin (1c) was found to
undergo slow isomerization to the corresponding Z-isomer
2
c in DMSO-d
6
at room temperature (T ) 6 d, ∼30%
isomerization). Heats of formation derived from AM1
semiempirical calculations revealed that the Z-isomer 1c
is approximately 6 kcal/mol lower in energy than the
corresponding E-isomer 2c. Similar findings were ob-
served between the (E)- and (Z)-isomers of monobromi-
nated and debrominated analogues 10 and 1b. These
data suggest that presence of bromine atoms on the
pyrrole (steric factors) is not a predominant factor in
determining the energy differences between the (E)- and
treatment of imidazolone 6b with 2 equiv of bromine in
a sodium acetate/acetic acid medium produced a mixture
of mono- and dibrominated unsaturated pyrroloazepino-
nes 9a and 9b in 35% and 40% yields, respectively. Small
amounts (<10%) of the desired (E)- and (Z)-bromoaxino-
hydantoins 1c and 2c could be detected. These results
suggest that pyrrole bromination effectively competes
with bromination of the imidazolone unit. Increasing the
amount of bromine to 3 equiv produced the desired
products, (E)- and (Z)-bromoaxinohydantoins 1c and 2c
in 35% and 45% yields, respectively, after flash chroma-
tography. Similar treatment of dibromoazepinone 6a with
(
Z)-isomers.
In summary, a short synthesis of (E)- and (Z)-axino-
hydantoins 1b, 1c, 2b, and 2c has been achieved. The
established route provides access to a number of previ-
ously unprecedented axinohydantoin derivatives. The
R-functionalized cyclization of linear precursor 5 to the
tricyclic azepinone framework of 6 may, in fact, model a
biosynthetic cyclization step common to these natural
products and related pyrroloazepinone metabolites (e.g.,
hymenin). Furthermore, we have shown that the 2-imi-
dazolone nucleus, which is readily available from R-amino
acids, serves as a useful precursor to the hydantoin
structural unit.
2
2
equiv of bromine afforded (E)- and (Z)-isomers 1c and
c. The (E)- and (Z)-geometry of the tetrasubstituted
double bond at C-10 was deduced from comparison of the
reported carbon3b and proton chemical shift of C-9.
Interestingly, in contrast to the oxidation of hymenin to
bromohymenialdisine5b which produced only the (Z)-
glycocyamidine isomer, both (E)- and (Z)-hydantoin
isomers were obtained from oxidation of imidazolones 6a
and 6b. These results suggest that hydantoin-containing
4e
Ack n ow led gm en t. Financial Support from the Na-
tional Institutes of Health and Chugai Pharmaceutical
Co. is gratefully acknowledged.
(E)-axinohydantoins are configurationally more stable
than their glycocyamidine-containing (E)-hymenialdisine
counterparts.
Next, chemoselective hydrogenation of 1c and 2c was
investigated (Scheme 5). While reduction of (Z)-bro-
moaxinohydantoin (2c) produced (Z)-debromoaxinohy-
dantoin (2b) in good yield, the reduction of (E)-isomer
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
1
13
cedures, H and C NMR spectra for compounds 1b, 1c, 2b,
2
material is available free of charge via the Internet at
http://pubs.acs.org.
c, 3, 4, 5a -c, 6a -c, 7, 9a , 9b, 10, and 11 are reported. This
1
c afforded, surprisingly, a 1:1.3 mixture of (E)-debro-
J O020063V