7
012
M.-P. Brun et al. / Tetrahedron Letters 44 (2003) 7011–7013
4b
wherein acetate is the source of carbon. As depicted
in Scheme 2, the latter compound was subsequently
deprotected by means of hydrogenolysis in the presence
of 10% palladium on activated charcoal in glacial acetic
acid, this solvent being chosen in order to favor the
further cyclisation step. The amine obtained was not
isolated but immediately led to a nucleophilic attack
onto the ketone, providing the intermediate imine 6
(
not isolated) which was further reduced in situ during
the process. After evaporation of the solvent, the
residue was taken-up in water and one equivalent of
HCl and evaporated, in order to obtain the hydrochlo-
ride salt 1a instead of the acetate. This one-pot synthe-
1
sis proceeded smoothly with a 85% yield, H NMR and
HPLC showing a 98:2 ratio for both cis:trans
diastereoisomers obtained. The stereoselectivity of the
reduction was consistent with that stated in the
1
literature and can be easily explained by the hydro-
genation of the less-hindered face of the cyclic imine.
The absolute (2S) configuration was postulated accord-
ing to the configuration of the starting glutamic acid.
Assignment of the relative configuration was achieved
by means of NOESY and TOCSY experiments, a com-
Scheme 2. Preparation of (2S,5R)-5-ethoxycarbonylmethyl-
pyrrolidine-2-carboxylic acid. Reagents and conditions: i: car-
bonyldiimidazole, THF, 2 h; ii: (EtOOCCH COO) Mg, rt,
2
2
1
13
overnight, 62% overall; iii: 1 atm H , Pd/C, CH COOH, 85%;
plete H and C assignment being listed in Table 1.
2
3
iv: H O, HCl (1 equiv.), evaporation.
2
NOEs were observed between Hd and He/e%, Hc%, Hb%
and Ha, confirming the trans configuration of this
compound. Moreover, no NOE appeared between Hd
and Hc, nor Hb.
having already the C2 chiral carbon. As cyclisation of
the nitrogen with the carbon of the side-chain car-
boxylic acid was required, we first transformed this
group into a more reactive b-ketoester.
Conclusion
This transformation was achieved owing to the efficient
This synthesis provides a very short route to an exam-
ple of 5-substituted proline, starting from cheap pro-
tected glutamic acid, (L) and (D) enantiomers being
5
method described by Masamune, which is well docu-
6
mented even in the case of amino acids. Starting from
commercially available Cbz-Glu-OBn (Z-Glu-OBzl,
Bachem), we first activated the side-chain carboxylic
acid by treatment with carbonyl diimidazole. The
resulting species was subsequently subjected to conden-
sation with the magnesium salt of monoethyl malonic
acid, prepared from the commercial potassium salt, via
acidification at low temperature and treatment with
magnesium ethoxide. Pure b-ketoester 5 was obtained
with a 62% yield after column chromatography on
silica-gel. Interestingly, the elongation of glutamic acid
side-chain with two carbons at the same time is also
postulated in the biosynthesis of these compounds,
commercially available. This constitutes an alternative
synthesis which does not require strong nucleophiles for
the ring opening of pyroglutamic acid. The high yield
and the diastereoisomeric excess obtained for the reduc-
tion step makes it a method of choice for the prepara-
tion of both enantiomers of these compounds.
Acknowledgements
Marie-Priscille Brun is grateful to the M.E.R.T. for a
grant (A.M.N.)
Table 1. NMR assignment of compound 1a