Substituted azabicyclo[3.1.0]hexan-1-ols from aspartic and glutamic acid derivatives via titanium-mediated cyclopropanation

Article abstract of DOI:10.1016/j.tetlet.2008.08.114

The Kulinkovich cyclopropanation reaction has been used to synthesize azabicyclo[3.1.0]hexanols from amino acid derivatives containing two ester moieties.

Full text of DOI:10.1016/j.tetlet.2008.08.114

Tetrahedron Letters 49 (2008) 6512–6513
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Substituted azabicyclo[3.1.0]hexan-1-ols from aspartic and glutamic
acid derivatives via titanium-mediated cyclopropanation
*
Catherine A. Faler, Madeleine M. Joullié
Department of Chemistry, University of Pennsylvania, Philadelphia, PA 19104-6323, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The Kulinkovich cyclopropanation reaction has been used to synthesize azabicyclo[3.1.0]hexanols from
amino acid derivatives containing two ester moieties.
Received 1 August 2008
Revised 26 August 2008
Accepted 29 August 2008
Available online 3 September 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Kulinkovich reaction
Cyclopropanol
Azabicycle
Small heterocycles have great potential for biological activity.
Constrained derivatives of naturally occurring heterocycles can
have important pharmaceutical applications and their syntheses
are desirable. For these reasons, our research group has studied
the synthesis of azabicycles from amino acid derivatives via the
Kulinkovich reaction.1,2 Prior to our work, it was shown that bi-
cyclic cyclopropanols could be made from esters in good yields
with an intramolecular variant of this reaction.3 This reaction
was extended to N-alkylated amino acid esters, allowing incorpo-
ration of nitrogen into the ring.4 Although many variations of the
Kulinkovich reaction have been well-investigated, cyclizations of
substrates with two esters are less explored.
O
O
HO
c or d
HO
MeO
a,b
HO
HO
OMe
OH
O
N
O
NH
N
2
1
2
d
c
O
O
O
O
O
+
O
N
O
O
N
N
OH
3
We attempted to apply the Kulinkovich reaction to the synthe-
sis of a novel cyclopropane-substituted indolizidine (2). The double
cyclopropanation of N,N-diallyl aspartic acid dimethyl ester (1)
would give a compound similar to a synthetic analog of the anti-
cancer agent, Swainsonine.5 Diester 1 was exposed to a stoichiom-
etric amount of ClTi(Oi-Pr)3 and to an excess of c-C5H7MgBr, which
is a slight modification to the original Kulinkovich conditions and
is more often applied to the cyclization of amides.6 Rather than
the desired indolizidine, almost exclusive formation of
[3.1.0]cyclopropanol 3 was observed (Fig. 1).7 Other variations of
the Kulinkovich reaction conditions were tried and it was found
that when Ti(Oi-Pr)4 was used, mixtures of piperidinols and pyrro-
lidinones were produced. It was also noted that transesterification
of the methyl ester with an isopropoxide ligand from titanium oc-
curred, which is a common Kulinkovich side reaction.8
Figure 1. Cyclization of diallylated aspartic acid methyl ester. Reagents and
conditions: (a) SOCl2, MeOH, quant.; (b) allyl bromide, K2CO3, Bu4NI, MeCN, 90%;
(c) ClTi(Oi-Pr)3, c-C5H7MgBr, THF, 35%; (d) Ti(Oi-Pr)4, i-PrMgCl, THF.
To improve the yield and applicability of our aspartic acid-
derived cyclopropanol, N-benzyl-N-allyl aspartic acid methyl ester
was made and cyclized. This substrate showed similar selectivity
for the [3.1.0] system, and transesterification was observed (Fig. 2).
Additionally, an azabicyclic propionic acid ester was obtained
via cyclization of a glutamate derivative (Fig. 3). Not surprisingly,
only [3.1.0]cyclopropanol was formed.
Independently, Ollivier and co-workers performed competitive
cyclizations of succinate derivatives, and found a clear preference
for bicyclo[3.1.0]hexanol formation over bicyclo[2.1.0]hexanol.9
In addition, they exposed a monoallylated isopropyl aspartate
derivative to the Kulinkovich cyclopropanation (Fig. 4). This reac-
tion, however, did not result in isolation of the cyclopropanol,
but rather a pyrrolidinone. The researchers hypothesized that
* Corresponding author. Tel.: +1 215 898 3158; fax: +1 215 573 9711.
E-mail address: mjoullie@sas.upenn.edu (M. M. Joullié).
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.08.114

C. A. Faler, M. M. Joullié / Tetrahedron Letters 49 (2008) 6512–6513
6513
O
O
O
O
HO
MeO
MeO
a,b
R
R
a or b
OMe
O
OMe
R
N
N
N
O
NH
O
N
2
Bn
Bn
Bn
R'
Bn
4
R = Ph, CO2Me, CO2i-Pr
HO
R' = OH, Cl
c
Figure 5. Cyclopropanol ring-opening. Reaction conditions: (a) silica gel, O2; R0 =
OH; (b) FeCl3, pyridine, DMF; R0 = Cl.
O
N
Bn
5
Figure 2. Synthesis of azabicyclo[3.1.0]hexanol. Reagents and conditions: (a)
benzaldehyde, Na(OAc)3BH, CH2Cl2, 95%; (b) allyl bromide, K2CO3, Bu4NI, MeCN,
86%; (c) ClTi(Oi-Pr)3, c-C5H7MgBr, THF, 64%.
Kulinkovich mechanism, must have given rise to the unexpected
pyrrolidinone.
In conclusion, we have synthesized three azabicyclo[3.1.0]hex-
anols (3, 5, and 7). The [3.1.0] system is formed preferentially over
other cyclic systems, and slight changes in cyclization conditions
give much lower yields and monocyclic side products.
O
O
O
O
a,b
HO
OH
MeO
OMe
NH
N
2
Bn
Acknowledgments
6
O
HO
c
We thank the NSF (CHEM-0515443), Wyeth Research, and the
University of Pennsylvania for the financial support. We are also
grateful to Dr. Rakesh Kohli for technical assistance and Dr. George
Furst for NMR assistance, especially with the HMBC experiments.
MeO
N
Bn
7
Figure 3. Bicyclo[3.1.0]hexanol from glutamic acid. Reagents and conditions: (a)
SOCl2, MeOH, quant.; (b) benzaldehyde, Na(OAc)3BH, CH2Cl2, 90%; (c) allyl bromide,
K2CO3, Bu4NI, MeCN, 85%; (c) ClTi(Oi-Pr)3, c-C5H7MgBr, THF, 65%.
Supplementary data
Experimental procedures, characterization, and spectra for
compounds 1, 3, 4, and 5 are available. Supplementary data associ-
ated with this article can be found, in the online version, at
doi:10.1016/j.tetlet.2008.08.114.
HO
CO i-Pr
i-Pr O C
2
C H MgCl
2
6
11
i-Pr O C
2
N
N
Ti(O i-Pr)
4
Bn
Bn
References and notes
O
i-Pr O C
2
1. (a) Cao, B.; Xiao, D.; Joullié, M. Org. Lett. 1999, 1, 1799–1801; (b) Faler, C. A.;
Cao, B.; Joullié, M. Heterocycles 2006, 67, 519–522.
2. Kulinkovich, O.; Sviridov, S.; Vasilevskii, D.; Pritytskaya, T. Zh. Org. Khim. 1989,
25, 2244–2245.
3. (a) Kim, S.; Sung, M.; Cha, J. Org. Synth. 2003, 80, 111–119; (b) Kasatkin, A.; Sato,
F. Tetrahedron Lett. 1995, 36, 6079–6082.
N
Bn
Figure 4. Ollivier’s cyclization of N-allyl-N-benzylaspartate.
4. Okamoto, S.; Kobayashi, M.; Iwakubo, K.; Sato, F. J. Am Chem. Soc. 1997, 119,
6984–6990.
5. Maggini, M.; Prato, M.; Ranelli, M.; Scorrano, G. Tetrahedron Lett. 1992, 33,
6537–6540.
the observed pyrrolidinone resulted from a rearrangement of
cyclopropanol.
6. (a) Lee, J.; Cha, . J. K. J. Org. Chem. 1997, 62, 1584–1585; (b) Chaplinski, V.;
deMeijere, A. Angew. Chem., Int. Ed. 1996, 35, 413–414.
7. Stereochemistry was proved by a HMBC experiment because the [4.1.0] system
is predicted to have similar 1H shifts.
8. (a) Seebach, D.; Hungerbühler, E.; Naef, R.; Schnurrenberger, P.; Weidmann, B.;
Zuger, M. Synthesis 1982, 138–141; (b) Lewis, N.; Ribas, C.; Wells, A. Synlett
1999, 957–959; (c) Eisch, J. J.; Adeosun, A. A.; Gitua, J. N. Eur. J. Org. Chem. 2003,
4721–4727.
9. Garnier, J.-M.; Jida, M.; Ollivier, J. Synlett 2006, 2739–2742.
10. (a) Ito, Y.; Fujii, S.; Saegusa, T. J. Org. Chem. 1976, 41, 2073–2074; (b) Booker-
Milburn, K.; Thompson, D. Synlett 1993, 592–594; (c) Iwasawa, N.; Hayakawa,
S.; Funahashi, M.; Isobe, K.; Narasaka, K. Bull. Chem. Soc. Jpn. 1993, 66, 819–
827.
The preferential formation of five-membered rings observed by
Ollivier is in accord with our results. However, all of our attempts
to convert the cyclopropanol into a pyrrolidinone resulted in
decomposition or recovery of starting material (Fig. 5). Only a
few oxidative reagents would react with the cyclopropanol, and
those experiments led to a ring-expanded product, which is consis-
tent with the literature precedent.10
As a pyrrolidinone was isolated from the cyclopropanation
reaction, but not directly from the cyclopropanol, we believe that
another intermediate, possibly a titanacycle from the accepted