Site-directed asymmetric quaternization of a peptide backbone at a C-terminal azlactone

Article abstract of DOI:10.1002/anie.200803661

(Chemical Equation Presented) Dual-purpose activation: Peptide C-terminal azlactones I undergo stereoselective alkylation with high efficiency by the use of a newly devised chiral tetraaminophosphonium salt as a phase-transfer catalyst, and the alkylated

Full text of DOI:10.1002/anie.200803661

Angewandte
Chemie
DOI: 10.1002/anie.200803661
Peptide Modification
Site-Directed Asymmetric Quaternization of a Peptide Backbone at a
C-Terminal Azlactone**
Daisuke Uraguchi, Yoshihiro Asai, and Takashi Ooi*
The incorporation of nonproteinogenic amino acids into
peptides is a rational and indispensable approach to the
elucidation of the diverse roles of peptides and can also
impart new functions to the peptides to facilitate their use as
pharmaceuticals.[1] In particular, the appropriate introduction
of a,a-dialkyl a-amino acids (quaternary a-amino acids) is
known to induce significant conformational constraints and
can thus be used to probe the molecular structure of receptors
or enhance the biological activity of the peptides by helping to
preorganize the optimum conformation for binding or by
inhibiting metabolic degradation.[2] For example, naturally
occurring aminoisobutyric acid (Aib) and related achiral
amino acids are often
D2-symmetric tetraaminophosphonium salt as a chiral
phase-transfer catalyst (PTC).[8] In combination with a
simple and direct ligation process,[9] this site-directed quater-
nization method enables the customized asymmetric synthesis
of nonnatural oligopeptides that contain quaternary a-amino
acid residues at the requisite positions.
Our strategy is based on the dehydrative activation of the
C terminus of a peptide as an oxazol-5-(4H)-one I, otherwise
known as an azlactone,[10] and its stereoselective alkylation
under practical biphasic conditions in the presence of an
appropriate PTC (Scheme 1).[11,12] Since an azlactone is
known to be an active methine compound as well as a
employed to study the rela-
tionship between the struc-
ture, stability, and function of
bioactive peptides.[3] In sharp
contrast, however, the use of
chiral, nonracemic quater-
nary a-amino acids has been
very limited, despite a long-
standing interest in the cor-
relation of the chirality of
Scheme 1. Strategy for the incorporation of chiral quaternary a-amino acids at specific sites of a peptide
strand. +Q*XÀ =chiral quaternary onium salt; LG=leaving group.
these amino acids with pep-
tide structure, such as the
handedness of the helix, and
peptide function.[3c,d,4] This situation is primarily due to the
lack of flexible synthetic strategies for incorporating a wide
range of chiral quaternary a-amino acids into peptide strands
at specific sites, which seems to be synonymous with the
intricacy of peptide-bond construction with independently
prepared, enantiomerically enriched quaternary a-amino
acids.[5–7] Herein, we report our own solution to this problem.
This solution involves the direct and highly stereoselective
construction of quaternary stereogenic carbon centers on
C-terminal amino acid residues of growing peptides by
alkylation under organic–aqueous biphasic conditions in the
reactive acyl donor, I can be regarded as a doubly activated
form of a peptide for two purposes, alkylation and subsequent
ligation. Thus, if we are able to utilize I selectively as the key
substrate for the first purpose, the purpose- and stereoselec-
tive introduction of an additional side chain R2 (ꢀ in
1
Scheme 1), the resulting azlactone II could be employed
directly for the second purpose, the ligation of the peptide
(ꢀ). A crucial factor in enabling the stereoselective double
2
functionalization of the peptide C-terminal azlactone is the
structure of the PTC. It would be ideal if the primary structure
of the PTC was easy to synthesize and readily tunable so that
the most suitable catalyst for a particular transformation
could be identified promptly. With this aim in mind, we
focused our attention on the structural features of tetra-
aminophophonium salts,[13,14] which can be assembled from a
variety of amines and a commercial phosphorous source.
We first attempted the allylation of azlactone 3a, which
was formed from Boc-l-Ala-d,l-Phe by dehydration with a
carbodiimide (Scheme 2), in the presence of a known achiral
aminophosphonium salt (Table 1, entry 1). Thus, a mixture of
3a, allyl bromide, and tetra(N-methylcyclohexylamino)phos-
phonium tetrafluoroborate[13b,c] (5 mol%) in a mixture of
toluene and saturated aqueous tripotassium phosphate was
stirred vigorously at 08C. The starting material was consumed
presence of
a catalytic amount of an optically pure,
[*] Dr. D. Uraguchi, Y. Asai, Prof. Dr. T. Ooi
Department of Applied Chemistry
Graduate School of Engineering, Nagoya University
Furo-cho B2-3(611), Chikusa, Nagoya 464-8603 (Japan)
Fax: (+81)52-789-3338
E-mail: tooi@apchem.nagoya-u.ac.jp
[**] This research was partially supported by the Global COE Program in
Chemistry of Nagoya University, a Grant-in-Aid from the Ministry of
Education, Culture, Sports, Science, and Technology (Japan), and
the Naito Science & Engineering Foundation.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200803661.
Angew. Chem. Int. Ed. 2009, 48, 733 –737
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
733

Communications
Scheme 2. Dehydration to the azlactone. Boc=tert-butoxycarbonyl,
DCC=N,N’-dicyclohexylcarbodiimide, WSCD·HCl=1-ethyl-3-(3-di-
methylaminopropyl)carbodiimide hydrochloride.
remained unsatisfactory (Table 1, entry 2). However, the
introduction of 3,5-bis(trialkylsilyl)benzyl substituents on
the nitrogen atoms of the catalyst led to a dramatic enhance-
ment of both the purpose- and stereoselectivities (Table 1,
entries 3 and 4): Product 4a was isolated in 78% yield with an
excellent diastereomeric ratio of 94:6 when 1c, which
contains 3,5-bis(tert-butyldimethylsilyl)benzyl substituents,
was used as the catalyst (Table 1, entry 4).
The structure of the chiral D2-symmetric P-spiro amino-
phosphonium salts with silyl substituents was verified by
single-crystal X-ray diffraction analysis of 1b (Figure 1). The
absolute configuration of the newly created quaternary
carbon center of the major isomer of 4a was determined to
be R by X-ray analysis of an amide derivative.[17]
Table 1: Effect of the catalyst structure on the purpose selectivity and
stereoselectivity under organic–aqueous biphasic conditions.[a]
Entry
1
Solvent
T
t
Yield[b] d.r.[c]
[8C] [h]
[%]
S,R/S,S
1[d]
2
3
4
5
6
7
8
9
(c-C6H11MeN)4PBF4 toluene
0
0
0
0
0
0
0
2
1.5
2
1
2
21
26
65
78
60
75
52:48
64:36
78:22
94:6
92:8
94:6
95:5
93:7
95:5
93:7
9:91
13:87
7:93
1a
1b
1c
1c
1c
1c
Et2O
TBME[e]
CPME[e]
1.5
0.75 79
1c
1c
1c
1c
1c
1c
ent-1c
25 0.25 75
À15
À15
0
2
2
86
85
10[f]
11[g]
12[g]
13[g]
14
0.75 69
25 0.25 60
À15 1.5
70
70
À15
2
8:92
[a] Unless otherwise noted, reactions were carried out on a 0.1 mmol
scale with 1.2 equivalents of allyl bromide. [b] Yield of the isolated
product. [c] The diastereomeric ratio was determined by HPLC on a
chiral stationary phase. [d] Catalyst: 5 mol%. [e] TBME=tert-butyl
methyl ether; CPME=cyclopentyl methyl ether. [f] The azlactone 3b
derived from Cbz-l-Ala-d,l-Phe was used as the substrate instead of 3a
(Cbz=carbobenzyloxy). The absolute configuration of 4b was assigned
by analogy with 4a. [g] Compound ent-3a was used instead of 3a. The
diastereomeric ratio (R,S)-4a/(R,R)-4a is given.
Figure 1. Molecular structure of the tetraaminophosphonium salt 1b.
The counterion (ClÀ), hydrogen atoms, and solvent molecules are
omitted for clarity. a) Perpendicular view of the 5,5-spirocycle; b) diag-
onal view. Purple: phosphorus, blue: nitrogen, light blue: silicon.
within 2 h; however, the allylated dipeptide derivative 4a was
obtained in just 21% yield. The observed (S,R)-4a/(S,S)-4a
ratio of 52:48 reveals the intrinsic stereoselectivity of the
transformation under these conditions. The concomitant yet
substantial formation of a diastereomeric mixture of an
undesired self-acylation product (73%) suggested the
extreme difficulty of the purpose-selective utilization of a
reactive azlactone moiety.[15]
These initial observations led us to design a series of
optically pure tetraaminophosphonium chlorides 1 as new,
chiral PTCs and to evaluate the relationship between their
structure and their selectivity in the allylation of 3a under
similar conditions.[16] Although an improvement in diastereo-
selectivity was observed with the N-benzylated aminophos-
phonium salt 1a, which was prepared from commercially
available (R,R)-1,2-diphenylethylenediamine and phospho-
rus pentachloride in two steps, the chemical yield of 4a
This biphasic alkylation system is compatible with ethe-
real solvents (Table 1, entries 5–7). Cyclopentyl methyl ether
(CPME) was found to be the solvent of choice (Table 1,
entry 7). The reaction temperature affected the stereoselec-
tivity only subtly (Table 1, entries 7–9), but self-acylation was
suppressed further (and thus the chemical yield was higher)
when the reaction mixture was cooled in an ice–salt bath
(À158C; Table 1, entry 9). The reaction of the N-carboben-
zyloxy-protected dipeptide 3b proceeded with comparable
reactivity and selectivity under the optimized conditions
(Table 1, entry 10).
To gain insight into the origin of the facial discrimination
in the present inherently diastereoselective reaction, we
carried out the allylation of the enantiomer of 3a, ent-3a
734
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 733 –737

Angewandte
Chemie
(derived from Boc-d-Ala-d,l-Phe), in the presence of 1c at
08C. The diastereomeric product (R,R)-4a was produced as
the major isomer (69%, (R,S)-4a/(R,R)-4a 9:91; Table 1,
entry 11). The yield and selectivity both increased when the
temperature was lowered (Table 1, entries 11–13). Although
the slightly lower stereoselectivity suggests that ent-3a has a
mismatched stereocenter with respect to diastereofacial
differentiation with 1c, the establishment of the quaternary
carbon center with high enantioselectivity and the absolute
configuration of this center indicate clearly that the stereo-
chemical outcome of the reaction is mainly determined by the
chirality of the catalyst. The pivotal role of the chiral catalyst
was also supported by the predominant formation of (S,S)-4a
in the allylation of 3a with ent-1c under similar conditions
(Table 1, entry 14).
(Table 2, entry 6).[18] We also probed the scope of the reation
with respect to the alkyl halide component (Table 2,
entries 9–13). A terminal triple bond, a potential functionality
for use in click chemistry,[19] was introduced with rigorous
stereochemical control (Table 2, entry 10). Alkyl halides
containing nitrile and ester functionalities also reacted
efficiently (Table 2, entries 11 and 12). The use of such
electrophiles enables the incorporation of asparagine and
aspartic acid analogues at the C terminus of a peptide.
Furthermore, a serine derivative was constructed in moderate
yield with high selectivity (Table 2, entry 13).
We next extended the asymmetric phase-transfer catalysis
of 1c to the C-terminal alkylation of tripeptides. The
allylation of azlactone 6, derived from the homotripeptide
of phenylalanine, with 1c under the standard conditions
resulted in the formation of 7 in 94% yield with 97:3
diastereoselectivity (Scheme 3). The distinct advantage of this
stereoselective alkylation strategy is that an alkylated,
diastereomerically enriched azlactone, such as 7, can be
used directly as a reactive acyl donor for a subsequent
ligation. For example, the simple treatment of 7 with a nearly
equimolar amount of l-phenylalanine methyl ester at 1208C
for 10 h furnished 8 quantitatively without detectable epime-
rization of any stereocenter. The resulting N-Boc tetrapeptide
To explore the substrate generality of the phase-transfer
allylation catalyzed by 1c, various dipeptide-derived azlac-
tones were treated with allyl bromide under the optimized
conditions (Table 2). First, the reactions of five azlactones
Table 2: Asymmetric alkylation of dipeptide-derived azlactones under
phase-transfer conditions.[a]
Entry AA1 AA2
3
R-X
4
t
[h]
Yield[b] d.r.[c]
[%]
=
1
2
3
Leu Phe 3c Br-CH2CH CH2 4c
7
8
89
94
74
97:3
99:1
93:7
91:9
97:3
91:9
97:3
95:5
95:5
98:2
93:7
93:7
95:5
Val
3d
3e
3 f
3g
3h
4d
4e 12
4 f
4g
Pro
Gly
Phe
4[d]
5
4.5 77
2.5 82
6
7
Trp
4h 12
93
97
Leu 3i
Leu Leu 3j
4i
4j
8
8[d]
9[d]
10[e]
11
12
13[e]
1.5 98
2.5 87
7
3
7
2
Br-CH2Ph
4k
4l
ꢁ
Phe Phe 3g Br-CH2C CH
Br-CH2CN
97
88
82
66
4m
4n
4o
Br-CH2CO2Me
Cl-CH2OMe
[a] Unless otherwise noted, reactions were carried out on a 0.1 mmol
scale with 1.2 equivalents of alkyl halide. [b] Yield of the isolated product.
[c] The diastereomeric ratio was determined by HPLC on a chiral
stationary phase. The absolute configuration was assigned by analogy
with 4a and 10. [d] The reaction was performed at 08C. [e] Alkyl halide:
5.0 equivalents.
with a different pendent amino acid (AA1) were examined,
and excellent stereoselectivities were uniformly observed
(Table 2, entries 1–5). Even 3 f, which contains a glycine unit,
reacted with high stereoselectivity (Table 2, entry 4), and 3g
derived from the homodipeptide of phenylalanine proved to
be a good substrate (Table 2, entry 5). Similarly, the amino
acid of the azlactone moiety (AA2) can be varied (Table 2,
entries 6–8). Notably, a tryptophan unit was used successfully
without protection of the reactive indole nitrogen atom
Scheme 3. Asymmetric alkylation of a tripeptide-derived azlactone and
construction of a tetrapeptide with two adjoining quaternary a-amino
acid residues.
Angew. Chem. Int. Ed. 2009, 48, 733 –737
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
735

Communications
methyl ester 8 was converted readily into a new azlactone 9
then transformed into the C-terminal azlactone 13, the
subsequent phase-transfer allylation of which in the presence
of 1c proceeded in high yield with high diastereoselectivity.
The resulting tetrapeptide 14 can be used as a transient
substrate for a further ligation–alkylation sequence or alter-
natively for the synthesis of oligopeptides with chiral quater-
nary a-amino acid residues incorporated at specific sites
through ligation with a preexisting, different oligopeptide
strand, as exemplified by the assembly of octapeptide 15.
In conclusion, we have developed a practical method for
the incorporation of a wide variety of chiral, nonracemic
quaternary a-amino acids at specific sites of a peptide strand.
This backbone modification of a growing peptide is based on
the dehydrative, double-purpose activation of the C terminus
as an azlactone and its highly selective quaternization under
biphasic conditions with the chiral D2-symmetric tetraamino-
phosphonium chloride PTC 1c. Since the alkylated azlactone
can be employed directly for the second purpose, that is,
peptide ligation, appropriate repetition of the operationally
simple alkylation–ligation process provides a powerful basis
for the synthesis of oligopeptides with quaternary a-amino
acid residues of the desired configuration at the requisite
positions. We hope that the results presented herein will
stimulate and even facilitate research in diverse fields in
which such nonnatural peptides serve as valuable tools.
through a saponification–dehydration sequence in prepara-
tion for another alkylation. However, when 9 was treated with
propargyl bromide in the presence of 1c under biphasic
conditions, significantly diminished diastereoselectivity was
observed. This result implied that the adjacent, sterically
congested quaternary carbon stereocenter had a considerable
influence on the stereochemical outcome of the reaction, and
that 1c may be a mismatched catalyst for the diastereofacial
differentiation of (S,S,R)-9. Indeed, a high level of stereose-
lectivity was observed with catalyst ent-1c. Thus, the straight-
forward asymmetric construction of stereochemically dense
tetrapeptides with two adjoining quaternary a-amino acid
residues is possible. The absolute stereostructure of the major
isomer of tetrapeptide 10 was determined to be S,S,R,S by
X-ray diffraction analysis after conversion into the amide
derivative 11.
To further highlight the broad scope of our strategy, we
undertook the stereoselective synthesis of a tetrapeptide with
two non-adjacent chiral quaternary a-amino acid residues.
The allylated azlactone 4b was prepared as described earlier
with the catalyst 1c (Table 1, entry 10) and elongated by two
amino acid residues through amide-bond formation with the
tert-butyl ester of the homodipeptide of l-phenylalanine
(Scheme 4). The protected tetrapeptide 12 thus obtained was
Received: July 26, 2008
Published online: November 25, 2008
Keywords: alkylation · amino acids · peptide modification ·
.
phase-transfer catalysis · phosphonium salts
[1] a) J. Gante, Angew. Chem. 1994, 106, 1780 – 1802; Angew. Chem.
Int. Ed. Engl. 1994, 33, 1699 – 1720; b) J. Venkatraman, S. C.
Shankaramma, P. Balaram, Chem. Rev. 2001, 101, 3131 – 3152;
c) L. Wang, P. G. Schultz, Angew. Chem. 2005, 117, 34 – 68;
Angew. Chem. Int. Ed. 2005, 44, 34 – 66.
[2] A. Giannis, T. Kolter, Angew. Chem. 1993, 105, 1303 – 1326;
Angew. Chem. Int. Ed. Engl. 1993, 32, 1244 – 1267.
[3] a) S. D. Bryant, R. Guerrini, S. Salvadori, C. Bianchi, R. Tomatis,
M. Attila, L. H. Lazarus, J. Med. Chem. 1997, 40, 2579 – 2587;
b) I. Sꢀgalas, Y. Prigent, D. Davoust, B. Bodo, S. Rebuffat,
Biopolymers 1999, 50, 71 – 85; c) C. Toniolo, M. Crisma, F.
Formaggio, C. Peggion, Biopolymers 2001, 60, 396 – 419; d) C.
Toniolo, F. Formaggio, B. Kaptein, Q. B. Broxterman, Synlett
2006, 1295 – 1310.
[4] a) H. Heimgartner, Angew. Chem. 1991, 103, 271 – 297; Angew.
Chem. Int. Ed. Engl. 1991, 30, 238 – 264; b) R. T. Shuman, R. B.
Rothenberger, C. S. Campbell, G. F. Smith, D. S. Gifford-Moore,
J. W. Paschal, P. D. Gesellchen, J. Med. Chem. 1995, 38, 4446 –
4453; c) M. Tanaka, Chem. Pharm. Bull. 2007, 55, 349 – 358.
[5] a) J. M. Humphrey, A. R. Chamberlin, Chem. Rev. 1997, 97,
2243 – 2266; b) Y. Fu, L. G. J. Hammarstrꢁm, T. J. Miller, F. R.
Fronczek, M. L. McLaughlin, R. P. Hammer, J. Org. Chem. 2001,
66, 7118 – 7124; c) S. Stamm, A. Linden, H. Heimgartner, Helv.
Chim. Acta 2006, 89, 1 – 15; d) A. R. Katritzky, E. Todadze, P.
Angrish, B. Draghici, J. Org. Chem. 2007, 72, 5794 – 5801.
[6] a) S. Wenglowsky, L. S. Hegedus, J. Am. Chem. Soc. 1998, 120,
12468 – 12473; b) Y. Ohfune, T. Shinada, Bull. Chem. Soc. Jpn.
2003, 76, 1115 – 1129; c) K. A. Brun, A. Linden, H. Heimgartner,
Helv. Chim. Acta 2008, 91, 526 – 558.
Scheme 4. Facile asymmetric synthesis of a tetrapeptide with two
quaternary stereocenters on non-adjacent amino acid residues and its
use as an intermediate in the synthesis of an octapeptide.
736
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 733 –737

Angewandte
Chemie
[7] a) C. Cativiela, M. D. Dꢂaz-de-Villegas, Tetrahedron: Asymmetry
1998, 9, 3517 – 3599; b) Y. Ohfune, T. Shinada, Eur. J. Org. Chem.
2005, 5127 – 5143; c) C. Cativiela, M. D. Dꢂaz-de-Villegas, Tetra-
hedron: Asymmetry 2007, 18, 569 – 623; d) H. Vogt, S. Brꢃse,
Org. Biomol. Chem. 2007, 5, 406 – 430.
structural assignment of the dimer 16, see the Supporting
Information.
[8] a) Handbook of Phase Transfer Catalysis (Eds.: Y. Sasson, R.
Neumann), Blackie Academic & Professional, London, 1997;
b) Asymmetric Phase Transfer Catalysis (Ed.: K. Maruoka),
Wiley-VCH, New York, 2008.
[9] a) F. Formaggio, Q. B. Broxterman, C. Toniolo in Houben-Weyl:
Methods of Organic Chemistry, Vol. E22c (Eds.: M. Goodman,
A. Felix, L. Moroder, C. Toniolo), Thieme, Stuttgart, 2002,
chap. 10.3; b) D. Obrecht, U. Bohdal, C. Broger, D. Bur, C.
Lehmann, R. Ruffieux, P. Schꢁnholzer, C. Spiegler, K. Mꢄller,
Helv. Chim. Acta 1995, 78, 563 – 580.
[16] The results of the allylation with other representative chiral
quaternary ammonium salts are reported in the Supporting
Information.
[17] A simple ring-opening reaction of 4a with (R)-phenylethylamine
without a solvent gave the amide 5 for X-ray crystallographic
analysis.
[10] J. S. Fisk, R. A. Mosey, J. J. Tepe, Chem. Soc. Rev. 2007, 36,
1432 – 1440.
[11] M. L. Gelmi, D. Pocar, L. M. Rossi, Synthesis 1984, 763 – 765.
[12] a) T. Ooi, E. Tayama, K. Maruoka, Angew. Chem. 2003, 115,
599 – 602; Angew. Chem. Int. Ed. 2003, 42, 579 – 582; b) K.
Maruoka, E. Tayama, T. Ooi, Proc. Natl. Acad. Sci. USA 2004,
101, 5824 – 5829.
[13] a) R. Schwesinger, J. Willaredt, H. Schlemper, M. Keller, D.
Schmitt, H. Fritz, Chem. Ber. 1994, 127, 2435 – 2454; b) R.
Schwesinger, R. Link, P. Wenzl, S. Kossek, M. Keller, Chem. Eur.
J. 2006, 12, 429 – 437; c) R. Schwesinger, R. Link, P. Wenzl, S.
Kossek, Chem. Eur. J. 2006, 12, 438 – 445.
[14] D. Uraguchi, S. Sakaki, T. Ooi, J. Am. Chem. Soc. 2007, 129,
12392 – 12393.
[15] As an azlactone is a reactive acyl donor, 3a is prone to react with
its enolate under the present conditions to give the self-acylation
product 16. The yield was calculated by considering a 100%
yield to correspond to the complete dimerization of 3a. For the
[18] Although we attempted to use histidine as a more hydrophilic
amino acid in the azlactone moiety (AA2), the corresponding
azlactone, such as that derived from Boc-l-Ala-d,l-His(p-Bom),
was not stable enough under the present biphasic conditions
(Bom = benzyloxymethyl).
[19] Y. L. Angell, K. Burgess, Chem. Soc. Rev. 2007, 36, 1674 – 1689.
Angew. Chem. Int. Ed. 2009, 48, 733 –737
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
737