inducingelementsbyJolliffe andco-workers8 and theuse of
photolabile auxiliaries by Smythe and co-workers.9 Re-
cently, ourlabreportedamethodtoconvertlinearpeptides
into macrocycles through the use of aziridine aldehydes
and isocyanides.10 The products of this transformation are
heterodetic cyclic peptides incorporating an N-acyl azir-
idine moiety in their backbones. Since then, we became
interested in the incorporation of the nonstandard amino
acid, aziridine-2-carboxylic acid (Azy), into cyclic peptides.
We became intrigued by the physical properties of the
aziridineÀamide bond and, as part of an ongoing program,
wanted to take advantage of a powerful one-step transfor-
mation of an aziridine into an amino acid residue.11 Herein
we report a straightforward synthesis of aziridine-contain-
ing cyclic peptide templates which allow for the late-stage
modification. A significant feature of this new method is the
facility with which R-substituted β-amino acid residues can
be introduced into tetrapeptide scaffolds.
The outcome of the end-to-end cyclization of tetrapep-
tides often depends on the site of ring closure.12 For a given
cyclic tetrapeptide (1) containing an Azy residue, there
exists four retrosynthetic disconnections for end-to-end
cyclization (Figure 1). We decided to synthesize and cyclize
linear Azy-containing tetrapeptides of the general struc-
ture 3 for the following reasons. Peptides of the general
structure 2, with an N-terminal aziridine and free car-
boxylic acid, have been studied by Moroder and co-
workers as promising irreversible inhibitors of cysteine
proteases.13 However, many of these compounds were
shown to exhibit sequence-dependent instability in both
reaction and purification steps, as well as on storage. This
is caused by self-protonation to generate a reactive azir-
idinium species that subsequently decomposes.
Furthermore, tetrapeptides of the general structure 4
were shown by Gin and co-workers to be unstable due to
an intramolecular acyl transfer as a result of the N-termi-
nus “biting back” and reacting with the distorted aziridine
amide, forming the corresponding diketopiperazinone.14
This left 3 and 5 as viable options for a linear precursor and
we chose 3 as the centerpiece of our approach.
In order to synthesize the Azy-containing tetrapeptide 3
insolutionphase, anappropriateprotecting group strategy
had to be designed. Acylated aziridines are very reactive
moieties that readily undergo ring-opening or acyl-transfer
reactions with a wide range of nucleophiles.15 Since azir-
idines are unstable to acid, a Boc-based strategy was
immediately ruled out. We also ruled out an Fmoc strategy
since the secondary amines commonly used for Fmoc-
deprotection would react with the acyl aziridine function-
ality. Furthermore, since activated aziridines are known to
take part in hydrogenolysis reactions,16 we decided that a
Cbz/benzyl-based protecting group strategy was not a
viable option. We therefore decided to pursue an allyl-
based strategy since these protecting groups can be re-
moved mildly in the presence of catalytic Pd(0) and an
appropriate scavenger. No premature aziridine ring-open-
ing was observed under these conditions.17 Our course of
action and results are shown in Scheme 1.
We proceeded to incorporate the Azy unit into the
peptide chain through a coupling reaction of N-tritylazir-
idine-2-carboxylic acid 618 with the corresponding dipep-
tide 7 containing an allyl ester to afford tripeptide 8. A
reductive trityl deprotection19 yielded NÀH aziridine tri-
peptide 9 in quantitative yield. A subsequent coupling to
the Alloc-protected amino acid furnished the protected
linear tetrapeptide 10. The Alloc and allyl ester protecting
groups were then removed by treating the protected tetra-
peptide with a catalytic amount of Pd(PPh3)4 and 2 equiv
of N,N-dimethylbarbituric acid (DMBA) as a scavenger
to afford the deprotected tetrapeptide 11. DMBA is a
weak carbon-based nucleophile that was determined to
be unreactive toward acylated aziridines in a control
experiment.20 Compound 11 was isolated by a simple
filtration from the reaction solution in high yield and
(11) Sweeney, J. B. Chem. Soc. Rev. 2002, 31, 247.
(12) (a) Pastuszak, J.; Gardner, J. H.; Singh, J.; Rich, D. H. J. Org.
Chem. 1982, 47, 2982–2987. (b) Schmidt, U.; Beutler, U.; Lieberknecht,
A. Angew. Chem., Int. Ed. Engl. 1989, 28, 333.
€
(13) Korn, A.; Rudolph-Bohner, S.; Moroder, L. Tetrahedron 1994,
50, 1717.
ꢁ
(14) Galonic, D. P.; van der Donk, W. A.; Gin, D. Y. J. Am. Chem.
Soc. 2004, 126, 12712.
(15) Nakajima, K.; Tanaka, T.; Morita, K.; Okawa, K. Bull. Chem.
Soc. Jpn. 1980, 53, 283.
(16) Davis, F. A.; Zhang, Y.; Rao, A.; Zhang, Z. Tetrahedron 2001,
57, 6345.
(17) For an example of Pd(0) insertion into aziridines, see: Wolfe,
J. P.; Ney, J. E. Org. Lett. 2003, 5, 4607.
(18) For synthesis, see: (a) Tanaka, T.; Nakajima, K.; Okawa, K.
ꢁ
Bull. Chem. Soc. Jpn. 1980, 53, 1352–1355. (b) Galonic, D. P.; Ide, N. D.;
van der Donk, W. A.; Gin, D. Y. J. Am. Chem. Soc. 2005, 127, 7359.
(19) Vedejs, E.; Klapars, A.; Warner, D. L.; Weiss, A. H. J. Org.
Chem. 2001, 66, 7542.
(20) In a control experiment, Cbz-Gly-Azy-Gly-OBn was stirred with
a variety of scavengers commonly used in allyl deprotections. After
stirring with DMBA for 60 h, the tripeptide remained completely
unreacted.
Figure 1. Possible ring disconnections for an Azy-containing
cyclic tetrapeptide.
Org. Lett., Vol. 14, No. 11, 2012
2899