5034
E. Jao et al. / Tetrahedron Letters 44 (2003) 5033–5035
Acknowledgements
We thank Drs. T. M. Chan and A. Evans for HSQC,
HSQC-TOCSY experiments and NOE studies, Dr. B.
Pramanik, Mr. P. Bartner for high resolution mass
spectral data and Drs. J. Wong, J. Kong, and Mr. T.
Meng for the preparation of lactam 1.
References
1. Trost, B. M.; Chan, D. M. T. J. Am. Chem. Soc. 1983, 105,
2315–2325.
2. For some reviews, see: (a) Trost, B. M. Angew. Chem., Int.
Ed. Engl. 1986, 25, 1; (b) Lautens, M.; Klute, W.; Tam, W.
Chem. Rev. 1996, 96, 49; For an example where TMM
entity is incorporated into a five-membered ring, see: (c)
Romero, J. M. L.; Sapmaz, S.; Fensterbank, L.; Malacria,
M. Eur. J. Org. Chem. 2001, 767–773.
3. Binger, P.; Freund, A.; Wedemann, P. Tetrahedron 1989,
45, 2887–2894.
4. For a preparation, see: Baldwin, J. E.; Moloney, M. G.;
Shim, S. B. Tetrahedron Lett. 1991, 32, 1379–1380.
5. Langlois, N.; Van Bac, N.; Dahuron, N.; Delcroix, J. M.;
Deyine, A.; Griffart-Brunet, D.; Chiaroni, A.; Riche, C.
Tetrahedron 1995, 51, 3571–3586.
Scheme 2.
6. Zhang, R.; Mamai, A.; Madalengoitia, J. S. J. Org. Chem.
1999, 64, 547–555.
7. Compound 2:
yield. The reaction did provide the desired 5,5,5-fused
ring system 27 but was accompanied by cycloadduct 38
where the double bond had migrated in the endocyclic
position (Scheme 1). Encouraged by this first attempt,
we then investigated different conditions aimed at
improving the yield of this sequence. The best result
was obtained with the use of Pd(P(OiPr)3)4 (generated
in situ with palladium(II) acetate (20 mol%) and triiso-
propyl phosphite (160 mol%)). The reaction of 1 pro-
ceeded cleanly in refluxing toluene to afford exclusively
the desired cycloadduct 29 in 80% isolated yield with no
detectable amount of byproduct 3.
The skeleton of the molecule is readly traced by HSQC and
HSQC-TOCSY experiments. The trans nature of the
cyclization is shown by the presence of a 5-3Hb NOE and
by the larger 6Ha-4Ha than 5-4 interaction. If the addition
had been cis the 6-4Ha interaction would probably be
absent as would the 5-3Hb NOEs. Also, 5H has a large 6b-5
whereas 4H has large 6Ha-4Ha consistent with 4H and 5H
being on opposite sides of the molecule. 1H NMR (DMSO,
300 MHz): l 7.39–7.30 (m, 5H); 6.04 (s, 1H); 4.84 (bs, 2H);
4.20 (dd, J=8.5, 6.0 Hz, 1H); 3.69–3.63 (m, 1H); 3.70–3.48
(m, 1H); 3.17–3.09 (m, 1H); 2.84–2.76 (m, 1H); 2.68–2.54
(m, 2H); 2.30 (dd, J=22.8, 16.5 Hz, 2H). 13C NMR
(DMSO, 125 MHz): l 180.8, 149.6, 139.5, 128.4, 128.3
(2C), 125.9 (2C), 107.3, 87.0, 70.2, 65.2, 48.5, 39.4, 38.0,
35.9. HMRS calculated for C16H18NO2 (M+H)=256.1338,
found 256.1340. Rotation: [h]D=+132° (c 1, CHCl3).
8. Compound 3:
With intermediate 2 in hand, we turned our efforts
towards preparation of 5,5-fused proline analogs 4.
Synthesis of proline analog 4-b is illustrated in Scheme
2. Simmons–Smith10 cyclopropanation of 2, accom-
plished via the modified procedure of Shi,11 led to
tetracyclic intermediate 5 in 71% yield. Alcohol 6,
quantitatively obtained from 5 by acidic deprotection
of the N,O-benzylidene acetal was reduced with excess
LAH in refluxing THF to give amino alcohol 712 in
71% yield. Protection of 7 under Schotten–Bauman
conditions led to the N-Boc prolinol 8 in 97% yield.
The latter was further oxidized with Jones’ reagent to
give quantitatively the novel 5,5-fused proline 4-b13 in
an overall yield of 40% from lactam 1. Other prolines
(Fig. 1) have been synthesized with similar yield and
full details regarding their preparation will be published
elsewhere.
In conclusion, we have shown that lactam 1 was an
excellent
substrate
for
a
palladium-assisted
trimethylene-methane (TMM) [3+2] cycloaddition reac-
tion. This efficient process was used to quickly assemble
the desired framework for the preparation of novel
5,5-fused proline surrogates.