Enantioselective synthesis of the pyrroloquinoline core of the martinellines

Article abstract of DOI:10.1021/ol0057030

(Matrix Presented) The first enantioselective synthesis of the martinelline core (-)-3 is reported. The synthesis of (-)-3 from N-allyl-N-(benzyloxycarbonyl)-2-iodoaniline (12) proceeded in seven steps and 23% overall yield. In addition, the preparation o

Full text of DOI:10.1021/ol0057030

ORGANIC
LETTERS
2000
Vol. 2, No. 10
1395-1397
Enantioselective Synthesis of the
Pyrroloquinoline Core of the
Martinellines
James A. Nieman and Michael D. Ennis*
Structural, Analytical & Medicinal Chemistry, Pharmacia & Upjohn, Inc.,
Kalamazoo, Michigan, 49001
michael.d.ennis@am.pnu.com
Received February 22, 2000
ABSTRACT
The first enantioselective synthesis of the martinelline core (−)-3 is reported. The synthesis of (−)-3 from N-allyl-N-(benzyloxycarbonyl)-2-
iodoaniline (12) proceeded in seven steps and 23% overall yield. In addition, the preparation of a carbocyclic model system is described.
In 1995, scientists at Merck reported the isolation of two
natural products from a family of tropical plants that have
long been used by Amazon Indian tribes for medicinal
purposes.1 These new products, martinelline (1) and marti-
nellic acid (2), were found to possess antibacterial activity
as well as affinity for adrenergic, muscarinic, and bradykinin
receptors.1 The relative stereochemical assignments for the
martinellines are based entirely on spectral observations, and
the absolute configurations are unknown. Although the
unusual pyrroloquinoline nucleus of the martinellines has
attracted the attention of several research laboratories, only
syntheses of the racemic pyrroloquinoline core have been
reported to date.2 Herein we describe our approach to these
We have previously described general methodology for
interesting natural products and the first enantioselective
synthesis of 3, the pyrroloquinoline core of the martinellines.
the preparation of chiral, cis-fused bicyclic pyrrolidines such
as 6 (Scheme 1).3 Condensation of the appropriate γ-ketoester
(1) Witherup, K. M.; Ransom, R. W.; Graham, A. C.; Bernard, A. M.:
Salvatore, M. J.; Lumma, W. C.; Anderson, P. S.; Pitzenberger, S. M.;
Varga, S. L. J. Am. Chem. Soc. 1995, 117, 6682.
Scheme 1
(2) Kang, S. K.; Park, S. S.; Kim, S. S.; Choi, J.-K.; Yum, E. K.
Tetrahedron Lett. 1999, 40, 4379. Snider, B. B.; Ahn, Y.; Foxman, B. M.
Tetrahedron Lett. 1999, 40, 3339. Batey, R. A.; Simoncic, P. D.; Lin, D.;
Smyj, R. P.; Lough, A. J. Chem. Commun. 1999, 651. Lovely, C. J.;
Mahmud, H. Tetrahedron Lett. 1999, 40, 2079. Hadden, M.; Stevenson, P.
J. Tetrahedron Lett. 1999, 40, 1215. Ho, T. C. T.; Jones, K. Tetrahedron
1997, 53, 8287. Gurjar, M. K.; Pal, S.; Rao, A. V. R. Heterocycles 1997,
45, 231. Frank, K. E.; Aube´, J. J. Org. Chem. 2000, 65, 655.
10.1021/ol0057030 CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/20/2000

on the readily accessible tetralone 7a (Scheme 2).4 Unlike
our earlier studies, ester 7a failed to react with phenylglycinol
even under forcing conditions. Fortunately, the derived
carboxylic acid 7b participated nicely in the desired chem-
istry. Condensation of 7b with (R)-(-)-phenylglycinol in
refluxing toluene generated the lactam 8 in a 76% isolated
yield as a single stereoisomer. Analysis of a series of
differential NOE experiments supported the stereochemical
assignment depicted.5 Although 8 could be stereoselectively
reduced to the pyrrolidine 9b with a variety of reducing
agents (LiAlH4, BH3‚THF, and DIBAL-H), the presence of
two benzylic carbon-nitrogen bonds in 9b complicated
efforts to selectively cleave the pendant 2-hydroxy-1-
phenethyl moiety. However, successful removal was realized
through tricyclic-lactam 9a, readily obtained by stereoselec-
tive reduction of 8 with triethylsilane in the presence of
titanium tetrachloride.3 Treatment of 9a as a solution in
DMSO with lithium hydroxide at elevated temperatures
generates 10, which upon acid-catalyzed hydrolysis provides
the known lactam 11a in 76% overall yield from 8.6 The
coupling constant observed for the ring-fusion protons (J )
6.5 Hz) was consistent with the reported values for the cis-
isomer (lit.,7,8 J ) 6.4 and 6.5 Hz). Finally, reduction of 11a
with lithium aluminum hydride generated the pyrrolidine 11b
in 85% yield.
Scheme 2
6
Encouraged by the successful outcome of these model
reactions, we turned our attention to the pyrroloquinoline
core of the martinellines (3). Toward this end, we required
the carboxylic acid 13b (Scheme 3). We prepared this
material via a modification of the palladium-catalyzed
carbonylative cyclization procedure reported by Negishi et
al.9 Treatment of the Z-protected N-allyl-2-iodoaniline 12
4 with phenylglycinol generates the lactam 5, in which the
absolute stereochemistry of the ring-junction carbons is
established. To test the viability of this approach toward the
synthesis of the martinellines, we examined this chemistry
Scheme 3
1396
Org. Lett., Vol. 2, No. 10, 2000

with catalytic dichlorobis(triphenylphosphine)palladium(II)
under 1600 psi carbon monoxide at 60 °C in the presence
of methanol provided the quinolone ester 13a in 60% isolated
yield.10 Hydrolysis of 13a provided the acid 13b,11 which
upon condensation with (R)-(-)-phenylglycinol generated
tetracyclic lactam 14 in 77% isolated yield as a single isomer.
Treatment of 14 with triethylsilane in the presence of titanium
tetrachloride gave a 3:1 ratio of the cis-fused pyrrolidinones
15a and 15b in 85% yield. An X-ray crystal structure of
15a confirmed the indicated absolute stereochemistry of the
pyrroloquinoline ring-fusion. Removal of the 2-hydroxy-1-
phenethyl moiety from 15a and 15b was readily ac-
complished using the two-step elimination/hydrolysis pro-
tocol described above to give the lactam 16 in 67% yield.12
Finally, reduction of 16 with lithium aluminum hydride
afforded the tricyclic diamine 3 in 92% yield.13
The preparation of the pyrroloquinolone 3 represents the
first enantioselective synthesis of the heterocyclic core of
the martinellines. We are currently applying the methodology
described herein toward the full construction of these
interesting natural products.
(3) Ennis, M. D.; Hoffman, R. L.; Ghazal, N. B.; Old, D. W.; Mooney,
P. A. J. Org. Chem. 1996, 61, 5813. For pioneering work in this area see
the A. I. Meyers references cited within.
(4) For the synthesis of 7a and 7b, see: Ravina, E.; Fueyo, J.; Teran,
C.; Cid, J.; Garcia Mera, G.; Orallo, F.; Bardan, B. Pharmazie 1992, 47,
574. Barlocco, D.; Pinna, G. A.; Carboni, L.; Cipolla, P. Farmaco 1989,
44, 967. Fontenla, J. A.; Osuna, J.; Rosa, E.; Castro, M. E.; G-Ferreiro, T.;
Loza-Garc´ıa, I.; Calleja, J. M.; Sanz, F.; Rodr´ıquez, J.; Ravin˜a, E.; Fueyo,
J.; F-Masaguer, C.; Vidal, A.; de Ceballos, M. L. J. Med. Chem. 1994, 37,
2564.
Acknowledgment. We thank Randy M. Jensen for
performing the NOE difference experiments on compound
8. We also thank Fusen Han for solving the X-ray crystal
structure of compound 15a.
(5) The key finding in support of structure 8 was an observed NOE
between the phenylglycinol-derived aromatic ring and the indicated H6
proton.
(6) For a related, three-step procedure for the removal of a 2-hydroxy-
1-phenethyl side-chain, see Fains, O.; Vernon, J. M. Tetrahedron Lett. 1997,
38, 8265.
(7) Oppolzer, W. J. Am. Chem. Soc. 1971, 93, 3834.
(8) Koot, W.-J.; Hiemstra, H.; Speckamp, W. N. Tetrahedron Lett. 1992,
33, 7969.
Supporting Information Available: Experimental pro-
cedures and full characterization of compounds 3, 8, 9a, 9b,
11a, 11b, 13a, 13b, and 14-16 are included.
(9) Tour, J. M.; Negishi, E. J. Am. Chem. Soc. 1985, 107, 8289. Negishi,
E.; Cope´ret, C.; Ma, S.; Mita, T.; Sugihara, T.; Tour, J. M. J. Am. Chem.
Soc. 1996, 118, 5904.
OL0057030
(10) We have applied this palladium-catalyzed carbonylative cyclization
chemistry to the synthesis of related heterocycles. Further details of this
interesting transformation will be reported shortly.
(11) Although the application of H2O as the nucleophile instead of
methanol in the carbonylative cyclization reaction with 12 generated 13b
directly, purification from other carboxylic acid byproducts proved to be
problematic.
(12) Compounds 15a and 15b can be subjected separately or as a crude
mixture to the elimination protocol. When the crude mixture was used, an
improved three-step yield of 67% for 16 was realized.
(13) As with compounds 11a and 11b, compounds 16 and 3 displayed
the ring-fusion proton coupling constant typical for a cis-fused product
(J ) 6.8 and 6.3 Hz, respectively). See also ref 1.
Org. Lett., Vol. 2, No. 10, 2000
1397