First observation of two consecutive γ turns in a crystalline linear dipeptide

Article abstract of DOI:10.1002/anie.200461230

(Figure Presented) Which way to turn? For the very first time, a linear dipeptide has been shown to accommodate two consecutive γ turns in the solid state. An aminocyclopropane carboxylic acid has overridden the tendency of proline to induce a β-turn conf

Full text of DOI:10.1002/anie.200461230

Angewandte
Chemie
Communications
The dipeptide Ac-l-Pro-d-c3Dip-NHMe, which incorporates a cyclo-
propane amino acid, adopts two consecutive g turns that are stabilized
by intramolecular hydrogen bonds. For more details on this con-
formation, which has not been seen before in crystalline short linear
peptides, see the Communication by A. I. Jimꢀnez et al. on the
following pages.
396
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200461230
Angew. Chem. Int. Ed. 2005, 44, 396 – 399

Angewandte
Chemie
Peptide Structures
are the most widely distributed and differ essentially in the
orientation of the middle peptide group. The other main
category of chain reversal is the so-called g turn,[1,4] which is
tighter and less common than the b bend. The g turn is
centered at a single residue (i + 1, Figure 1), usually with a
hydrogen bond between the CO and NH groups at positions i
and i + 2, respectively.
First Observation of Two Consecutive g Turns in a
Crystalline Linear Dipeptide**
Ana I. Jimꢀnez,* Gema Ballano, and Carlos Cativiela
Dedicated to Dr. Michel Marraud
We are involved in a research project aimed at evaluating
the relative stability of the bI and bII turns in model peptides
RCO-l-Pro-Xaa-NHR’ that incorporate different con-
strained a-amino acids at the i + 2 position (Xaa).[5] Such
terminally protected dipeptides constitute the smallest sys-
tems able to accommodate the b-turn conformation. As part
of these investigations we undertook the study of the l-Pro-
c3Dip sequence, where c3Dip denotes each enantiomer of a,b-
methanodiphenylalanine,[6] a cyclopropane analogue of phe-
nylalanine. The results of the solid-state conformational
analysis, of high structural relevance, are reported herein.
Racemic N’-methyl-2,2-diphenyl-1-aminocyclopropane-
carboxamide[7] (H-c3Dip-NHMe) was coupled to N-tert-
butoxycarbonyl-l-proline by the mixed-anhydride method
(Scheme 1). Column chromatography allowed the separation
of the resulting diastereomeric dipeptides (1a, 2a), which
Reverse turns (or bends) are elements of primary importance
both in the structure and function of peptides and proteins.[1]
They give rise to a sharp reversal in the polypeptide chain,
thus conferring on proteins their globular character. Turns
have been suggested to play a role in the initial stages of
protein folding and to be propitious sites for molecular
recognition and receptor binding. Many naturally occurring
oligopeptides have been proposed to adopt turns in their
bioactive conformation.[1,2]
Different types of bends have been recognized according
to the number and spatial arrangement of the residues
involved. The most common is the b turn,[1,3] which involves
four consecutive amino acids (i to i + 3, Figure 1) and is
Figure 1. Schematic representation of the intramolecular hydrogen
bond that stabilizes the b turn (i+3!i) and the g turn (i+2!i) in a
peptide chain. The ideal torsion angles of the central residues for the
g turn and the most common types of b turn are indicated.
Scheme 1. Synthesis of l-Pro-c3Dip dipeptides. Reagents and condi-
tions: a) iBuOCOCl, NMM, THF, ꢀ158C; b) column chromatography
(CH2Cl2/iPrOH 93/7); c) 3n HCl/EtOAc, RT; d) tBuCOCl, NMM,
CHCl3, 08C; e) Ac2O, NMM, CHCl3, RT. NMM=N-methylmorpholine.
generally stabilized by an intramolecular hydrogen bond
between the CO group at position i and the NH group at i + 3.
Several subclasses of b turns are further distinguished on the
basis of the backbone dihedral angles (f,y) associated with
the central i + 1 and i + 2 positions. Types I and II (Figure 1)
were isolated in optically pure form. The difficulties encoun-
tered in growing single crystals from these compounds
prompted us to replace the N-terminal urethane protecting
group with different amide groups. Single crystals suitable for
X-ray diffraction analysis were finally obtained from the
pivaloyl derivative 2b and the acetyl derivative 1c.
The crystalline structure of dipeptide 2b (Figure 2)
revealed an l (or S) configuration for the c3Dip residue.
The molecule adopts a b-turn conformation with the terminal
pivaloyl CO and methylamide NH groups intramolecularly
[*] Dr. A. I. Jimꢀnez, G. Ballano, Dr. C. Cativiela
Departamento de Quꢁmica Orgꢂnica, ICMA
Universidad de Zaragoza-CSIC
50009 Zaragoza (Spain)
Fax: (+34)976-761-210
E-mail: anisjim@unizar.es
ꢀ
hydrogen bonded (N···O: 2.97 ꢀ; N H···O: 1568). The
(f,y) torsion angles of the l-Pro (ꢀ63,133) and l-c3Dip
(69,ꢀ2) residues correspond to a b bend of type II (Figure 1).
It should be noted that the bI turn is energetically more
favorable for l-Pro-l-Xaa dipeptides in low-polarity sol-
vents[5a,8] but, in general, this conformation is not retained in
[**] The authors thank Dr. M. M. Zurbano for collection of the X-ray
diffraction data. Financial support from the Ministerio de Educaciꢃn
y Ciencia (project PPQ2001–1834 and predoctoral fellowship for
G.B.) and the Diputaciꢃn General de Aragꢃn is gratefully acknowl-
edged.
Angew. Chem. Int. Ed. 2005, 44, 396 –399
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
397

Communications
frequent than b bends, g turns are found to be abundant in
crystallized proteins,[10] where they are stabilized or induced
by long-range interactions. Such a source of stabilization is
not possible in small peptides and g turns are observed almost
exclusively in poor-solvating media,[1] that is, in the absence of
competing intermolecular hydrogen-bonding interactions
either with the solvent or with other peptide molecules in
the crystal. A few examples of X-ray diffraction structures
exhibiting a g turn have been reported for cyclic oligopep-
tides—mostly pentapeptides with proline at the turn center—
while this conformation is extremely uncommon in acyclic
sequences.[1,8b,11] The double g turn encountered in 1c is,
therefore, unique among crystalline short linear peptides.
Moreover, a survey of the Cambridge Structural Database[12]
indicates[13] that this is the first report in which proline
accommodates a g-turn conformation in the solid state in the
absence of the constraints imposed by backbone cyclization.
Proline is, in fact, the proteinogenic amino acid with the
highest folding propensity[10b,14] and has been shown to adopt
Figure 2. X-ray diffraction structure of Piv-l-Pro-l-c3Dip-NHMe (2b)
accommodating a bII turn. The intramolecular i+3!i hydrogen bond
is indicated by a dashed line. Only the proline Ca and the amide hydro-
gen atoms are shown.
the solid state, where the bII turn becomes preferred because
it allows the Xaa NH group to participate in intermolecular
contacts.[5a,d,8] The l-c3Dip NH site in 2b is actually hydrogen
bonded to a carbonyl group of a neighboring molecule. In
comparison, RCO-l-Pro-d-Xaa-NHR’ dipeptides are more
prone to bII folding under all environmental conditions and
have been found to invariably crystallize in the bII form.[5,8]
On the basis of the above discussion, Ac-l-Pro-d-c3Dip-
NHMe (1c) was expected to accommodate a bII turn in the
solid state. Surprisingly, a conformation consisting of two
consecutive g turns was found (Figure 3). The (f,y) torsion
the g-turn conformation in low-polarity solvents,[1b,c,15]
a
capacity that is not maintained in the solid state, where
proline is overwhelmingly found in the (ꢀ60,ꢀ30) or
(ꢀ60,140) region of the (f,y) map.[1b,5,8,10b,14,16]
Accordingly, the g-turn disposition accommodated by
proline in 1c can be ascribed to the presence of the contiguous
d-c3Dip residue, whose conformational tendencies appear to
prevail over those of proline. The latter consideration is of
special relevance given that proline is the proteinogenic
amino acid with the strongest conformational preferen-
ces.[10b,14] Cyclopropane a-amino acids have been suggested
as promising candidates for g-turn stabilization on the basis of
theoretical and experimental studies.[5c,d,9,17] However, the
fact that theoretical calculations on small peptides generally
overestimate the stability of g-folded conformers,[18] together
with the limited experimental evidence available to date,
made it premature to proclaim the capacity of such residues to
promote g-turn conformations.
The result reported here constitutes the first unequivocal
evidence of the ability of a cyclopropane a-amino acid to
adopt a g-turn disposition in a nonpropitious environment
and to influence the overall folding of the peptide chain.
Clearly, much work remains to be done to determine the
precise conditions under which c3Dip is able to nucleate
g turns and, thereafter, to establish its scope as a g-turn
inductor. The finding of an a-amino acid with a clear
preference for the g-turn conformation would be of enormous
help in the stabilization of this structural motif in bioactive
peptides, as well as in the construction of appropriate models
to study the factors that govern g folding.
Figure 3. X-ray diffraction structure of Ac-l-Pro-d-c3Dip-NHMe (1c)
exhibiting two consecutive g turns, each stabilized by an intramolecu-
lar i+2!i hydrogen bond (dashed lines). Only the proline Ca and the
amide hydrogen atoms are shown.
angles for both l-Pro (ꢀ82,61) and d-c3Dip (ꢀ72,47) lie close
to the typical values (ꢀ75,60),[1] with the larger deviation
being observed for the y angle of the d-c3Dip residue.
However, the excellent agreement between the observed
y value and that predicted from theoretical calculations[9] for
the g-turn minimum energy conformation of cyclopropane
residues bearing a phenyl substituent cis to the CO terminus
(y = 478) is remarkable, and appears to reflect a balance
between steric and hyperconjugative effects. Whatever the
origin of this deviation, it does not seem to impede the
appropriate alignment of the l-Pro CO and methylamide NH
sites, which form a strong intramolecular hydrogen bond
Experimental Section
2b: m.p. 2288C (iPr2O/CH2Cl2); [a]2D6 = + 56.1 (c = 0.42 in MeOH);
1
IR (Nujol): n˜ = 3377, 3317, 1691, 1641, 1608 cmꢀ1; H NMR (CDCl3,
400 MHz): d = 1.18–1.29 (m, 1H), 1.31 (s, 9H), 1.39–1.52 (m, 1H),
1.63–1.74 (m, 2H), 2.00 (d, J = 5.6 Hz, 1H), 2.62 (d, J = 5.6 Hz, 1H),
2.78 (d, J = 4.8 Hz, 3H), 3.42 (m, 1H), 3.57 (m, 1H), 4.09 (dd, J = 8.3,
5.0 Hz, 1H), 6.01 (brs, 1H), 7.16–7.29 (m, 6H), 7.37–7.43 (m, 4H),
7.45 ppm (brq, J = 4.8 Hz, 1H); 13C NMR (CDCl3, 100 MHz): d =
24.71, 25.73, 26.93, 27.10, 27.31, 39.10, 44.05, 45.44, 48.64, 63.20,
ꢀ
(N···O: 2.81 ꢀ, N H···O: 1478), as do the acetyl CO and
ꢀ
d-c3Dip NH groups (N···O: 2.78 ꢀ, N H···O: 1528).
The solid-state conformation exhibited by compound 1c is
of extraordinary structural significance. Although much less
398
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 396 –399

Angewandte
Chemie
126.74, 126.77, 128.12, 128.31, 128.93, 129.38, 140.42, 140.84, 168.94,
172.62, 178.20 ppm; elemental analysis (%) calcd for C27H33N3O3: C
72.46, H 7.43, N 9.39; found: C 72.61, H 7.37, N 9.47.
[7] This compound was obtained as a racemate by reaction of
diphenyldiazomethane with an a,b-dehydroalanine derivative
and subsequent elaboration of the protecting groups. G. Ballano,
A. I. Jimꢂnez, C. Cativiela, unpublished results.
1c: m.p. 1808C (Et2O); [a]2D1 = ꢀ218.3 (c = 0.27 in MeOH); IR
(Nujol): n˜ = 3369, 3347, 1676, 1660, 1629 cmꢀ1
;
1H NMR (CDCl3,
[8] a) A. Aubry, M. T. Cung, M. Marraud, J. Am. Chem. Soc. 1985,
107, 7640; b) M. Marraud, A. Aubry, Biopolymers 1996, 40, 45.
[9] a) C. Alemꢃn, A. I. Jimꢂnez, C. Cativiela, J. J. Perez, J. Casano-
vas, J. Phys. Chem. B 2002, 106, 11849; b) J. Casanovas, A. I.
Jimꢂnez, C. Cativiela, J. J. Pꢂrez, C. Alemꢃn, J. Org. Chem. 2003,
68, 7088.
[10] a) K. Guruprasad, S. Rajkumar, J. Biosci. 2000, 25, 143; b) P.
Chakrabarti, D. Pal, Prog. Biophys. Mol. Biol. 2001, 76, 1.
[11] a) E. Benedetti, Biopolymers 1996, 40, 3; b) C. Toniolo, M.
Crisma, F. Formaggio, C. Peggion, Biopolymers 2001, 60, 396.
[12] The CSD covers peptides of up to 24 residues. a) F. H. Allen,
Acta Crystallogr. Sect. B 2002, 58, 380; b) I. J. Bruno, J. C. Cole,
P. R. Edington, M. Kessler, C. F. Macrae, P. McCabe, J. Pearson,
R. Taylor, Acta Crystallogr. Sect. B 2002, 58, 389.
[13] The November 2003 version (2 updates) of the CSD was
reviewed. The proline -CO-Pro-NH- fragment was searched,
considering both l and d configurations and torsion angles
within the ranges (f,y) = (ꢀ75 ꢁ 30,60 ꢁ 30) (or their enantio-
meric values). Only proline residues included in cyclic peptides
were found to fulfill these requirements (refcodes: CGPGAP10,
ICOGAI, PAPGAP, SIKMUU10, WAWQOA, DHCMYD10,
HEBLIJ).
[14] M. W. MacArthur, J. M. Thornton, J. Mol. Biol. 1991, 218, 397.
[15] a) G. Boussard, M. Marraud, A. Aubry, Biopolymers 1979, 18,
1297; b) V. Madison, K. D. Kopple, J. Am. Chem. Soc. 1980, 102,
4855; c) G. B. Liang, C. J. Rito, S. H. Gellman, Biopolymers
1992, 32, 293; d) E. Beausoleil, W. D. Lubell, J. Am. Chem. Soc.
1996, 118, 12902; e) M. Miyazawa, K. Inouye, T. Hayakawa, Y.
Kyogoku, H. Sugeta, Appl. Spectrosc. 1996, 50, 644.
400 MHz): d = 1.51–1.59 (m, 1H), 1.64–1.80 (m, 3H), 1.76 (s, 3H),
2.21 (m, 1H), 2.56 (d, J = 5.6 Hz, 1H), 2.64 (d, J = 4.8 Hz, 3H), 2.87
(m, 1H), 3.03 (m, 1H), 4.25 (dd, J = 7.8, 1.3, 1H), 7.04–7.18 (m, 6H),
7.34–7.39 (m, 4H), 7.86 (brs, 1H), 7.90 ppm (brq, J = 4.8 Hz, 1H);
13C NMR (CDCl3, 100 MHz): d = 22.38, 23.38, 24.72, 26.55, 26.75,
46.28, 46.31, 48.09, 59.26, 126.67, 126.94, 128.00, 128.27, 129.16, 129.44,
140.22, 140.96, 169.10, 171.21, 173.43 ppm; elemental analysis (%)
calcd for C24H27N3O3: C 71.09, H 6.71, N 10.36; found: C 70.89, H 6.66,
N 10.39.
The X-ray crystal structures of 2b and 1c were solved by direct
methods using SHELXS-97.[19a] Refinement was performed using
SHELXL-97[19b] by the full-matrix least-squares technique. Hydrogen
atoms were located by calculation, with the exception of the
cyclopropane and NH protons, which were found on the E map.
CCDC-243527 (2b) and -243526 (1c) contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the
Cambridge Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB21EZ, UK; fax: (+ 44)1223-336-033; or deposit@ccdc.cam.
ac.uk).
X-ray data for 2b (C27H33N3O3·3/2H2O): Mr = 474.59, ortho-
rhombic, space group P21212, a = 11.0690(10), b = 23.498(2), c =
10.1750(10) ꢀ,
V= 2646.5(4) ꢀ3,
Z = 4;
1calcd = 1.191 gcmꢀ3
,
m(MoKa) = 0.081 mmꢀ1, T= 293(2) K, F(000) = 1020, 2qmax = 50.18;
3449 reflections collected, of which 3241 unique (Rint = 0.0308).
Final R indices (2380 observed reflections, I > 2sI): R1 = 0.0529,
wR2 = 0.1257; final R indices (all data): R1 = 0.0784, wR2 = 0.1409.
Highest residual electron density 0.19 eꢀꢀ3
.
X-ray data for 1c (C24H27N3O3): Mr = 405.49, orthorhombic, space
group P212121, a = 10.4010(2), b = 12.6760(2), c = 15.8660(3) ꢀ, V=
2091.82(7) ꢀ3, Z = 4; 1calcd = 1.288 gcmꢀ3, m(MoKa) = 0.086 mmꢀ1, T=
173(2) K, F(000) = 864, 2qmax = 538; 13835 reflections collected, of
which 4276 unique (Rint = 0.0559). Final R indices (3738 observed
reflections, I > 2sI): R1 = 0.0409, wR2 = 0.0855; final R indices (all
data): R1 = 0.0501, wR2 = 0.0892. Highest residual electron density
[16] Note that these regions correspond to the i + 1 position of a bI or
a bII turn, respectively (Figure 1).
[17] a) K. Burgess, K.-K. Ho, B. M. Pettitt, J. Am. Chem. Soc. 1994,
116, 799; b) V. N. Balaji, K. Ramnarayan, M. F. Chan, S. N. Rao,
Pept. Res. 1994, 7, 60; c) K. Burgess, C.-Y. Ke, J. Org. Chem.
1996, 61, 8627; d) C. Alemꢃn, J. Casanovas, S. E. Galembeck, J.
Comput.-Aided Mol. Des. 1998, 12, 259; e) D. Moye-Sherman, S.
Jin, S. Li, M. B. Welch, J. Reibenspies, K. Burgess, Chem. Eur. J.
1999, 5, 2730; f) C. Peggion, F. Formaggio, M. Crisma, C.
Toniolo, A. I. Jimꢂnez, C. Cativiela, B. Kaptein, Q. B. Broxter-
man, M. Saviano, E. Benedetti, Biopolymers 2003, 68, 178.
[18] a) H.-J. Bꢅhm, S. Brode, J. Am. Chem. Soc. 1991, 113, 7129;
b) H. S. Shang, T. Head-Gordon, J. Am. Chem. Soc. 1994, 116,
1528.
[19] a) G. M. Sheldrick, SHELXS-97, Program for the Solution of
Crystal Structures, University of Gꢅttingen, Gꢅttingen (Ger-
many), 1997; b) G. M. Sheldrick, SHELXL-97, Program for the
Refinement of Crystal Structures, University of Gꢅttingen,
Gꢅttingen (Germany), 1997.
0.15 eꢀꢀ3
.
Received: July 8, 2004
Keywords: b turn · conformation analysis · hydrogen bonds ·
.
peptide folding · structure elucidation
[1] a) C. Toniolo, Crit. Rev. Biochem. 1980, 9, 1; b) G. D. Rose, L. M.
Gierasch, J. A. Smith, Adv. Protein Chem. 1985, 37, 1; c) E. Vass,
M. Hollꢁsi, F. Besson, R. Buchet, Chem. Rev. 2003, 103, 1917.
[2] A. F. Spatola in Bioorganic Chemistry: Peptides and Proteins
(Ed.: S. M. Hecht), Oxford University Press, New York, 1998,
pp. 367 – 394.
[3] C. M. Venkatachalam, Biopolymers 1968, 6, 1425.
[4] G. Nꢂmethy, M. P. Printz, Macromolecules 1972, 5, 755.
[5] a) A. I. Jimꢂnez, C. Cativiela, A. Aubry, M. Marraud, J. Am.
Chem. Soc. 1998, 120, 9452; b) A. I. Jimꢂnez, C. Cativiela, J.
Gꢁmez-Catalꢃn, J. J. Pꢂrez, A. Aubry, M. Parꢄs, M. Marraud, J.
Am. Chem. Soc. 2000, 122, 5811; c) A. I. Jimꢂnez, C. Cativiela,
M. Marraud, Tetrahedron Lett. 2000, 41, 5353; d) A. I. Jimꢂnez,
M. Marraud, C. Cativiela, Tetrahedron Lett. 2003, 44, 3147.
[6] For a given a-amino acid Xaa, we designate its cyclopropane
analogue, obtained by linking the a and b carbon atoms with a
methylene group, as c3Xaa. b,b-Diphenylalanine is a phenyl-
alanine derivative known in the abbreviated form as Dip and,
accordingly, we have named its cyclopropane analogue as c3Dip.
Angew. Chem. Int. Ed. 2005, 44, 396 –399
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
399