Pseudo-prolines as a molecular hinge: Reversible induction of cis amide bonds into peptide backbones

Article abstract of DOI:10.1021/ja962780a

Serine, threonine-derived (4S)-oxazolidine-4-carboxylic acid, and cysteine-derived (4R)-thiazolidinecarboxylic acid, denoted pseudo-proline (Xaa[Ψ(R₁,R₂)pro]), serve as structure disrupting, solubilizing building blocks in peptide sy

Full text of DOI:10.1021/ja962780a

918
J. Am. Chem. Soc. 1997, 119, 918-925
Pseudo-Prolines as a Molecular Hinge: Reversible Induction of
cis Amide Bonds into Peptide Backbones
Pascal Dumy,* Michael Keller, Declan E. Ryan, Barbara Rohwedder,
Torsten Wo1hr, and Manfred Mutter
Contribution from the Institute of Organic Chemistry, UniVersity of Lausanne,
BCH-Dorigny, CH-1015, Lausanne, Switzerland
ReceiVed August 8, 1996X
Abstract: Serine, threonine-derived (4S)-oxazolidine-4-carboxylic acid, and cysteine-derived (4R)-thiazolidinecar-
boxylic acid, denoted pseudo-proline (Xaa[ΨR ,R pro]), serve as structure disrupting, solubilizing building blocks in
peptide synthesis. Variation of the 2-C substituents within the heterocyclic system results in different physicochemical
and conformational properties. NMR studies of a series of pseudo-proline (ΨPro)-containing peptides reveal a
pronounced effect of the 2-C substituents upon the cis to trans ratio of the adjacent amide bond in solution. 2-C
unsubstituted systems show a preference similar to that of the proline residue for the trans form, whereas 2,2-
dimethylated derivatives adopt the cis amide conformation in high content. For 2-monosubstituted ΨPro, the cis-
trans distribution depends on the 2-C chirality. For the 2-(S)-diastereoisomer, both forms are similarly populated in
solution, whereas the 2-(R)-epimer adopts preferentially the trans form. The results are supported by conformational
energy calculations and suggest that, by tailoring the degree of substitution, pseudo-prolines may serve as a temporary
proline mimetic or as a hinge in peptide backbones.
1
2
Introduction
as protein folding and protein recognition6,7,8 has led to the
development of numerous mimetics10,11 and substituted-proline
analogues12,13 intended to constrain and control the peptide
backbone in reverse turn motifs or to alter the imide cis-trans
ratio.14 For example, substitution with â-alkyl proline12 results
in a Ψ angle constrained by the syn-â substituent around 0°
while (S)-R-methyl-substituted proline12 (PMe) induces prefer-
entially the trans form of the Xaa-Pro bond due to unfavorable
steric interaction between the methyl group and the R-proton
of Xaa in the cis form.
The proline residue plays a peculiar role in peptide and protein
secondary structure formation1,2 and thereby represents an
important means in nature to switch and direct peptide chains
into favorable topologies. The cyclic nature of the proline
residue confers unique conformational properties to the peptide
or protein backbone when compared to the common proteino-
genic amino acids. For instance, on the one hand, the proline
ring serves to intrinsically restrict its Φ dihedral angle around
-60° ( 15, while on the other hand, the imidic bond formed
with the preceding residue (Xaa-Pro) is readily subject to cis-
trans isomerization. Accordingly, proline residues are usually
encountered in loop or turn1,2 motifs with the utmost preference
at the i + 1 position for â-turn type I or type II when the Xaa-
Pro imide bond is trans (ωi ) 180°) or at the i + 2 position of
turn type VI (ωi+1 ) 0°) in the cis form.1,3,4 In addition, the
characteristic low activation barrier for isomerization combined
with the small free energy difference5 between the two Xaa-
Pro peptide bond isomers provides a rationale for the putative
role of proline in the limiting steps of the protein folding
pathway.6 This is supported by the discovery of the ubiquitously
occurring family of prolyl-peptidyl cis-trans isomerases such
as cyclophilins7 and FKBPs8 which catalyze the cis-trans
isomerization of peptidyl-prolyl bonds in ViVo9 and in Vitro.7,8
The prevalence of proline residues in biological processes such
Recently we reported that Ser- or Thr-derived 2-substituted
oxazolidine-4(S)-carboxylic acid and Cys-derived 2-substituted
thiazolidine-4(R)-carboxylic acid, denoted as pseudo-prolines
(ΨPro’s) (Xaa[ΨR ,R pro], Figure 1), exert a pronounced effect
upon the peptide backbone, preventing peptide aggregation and
self-association during solid phase conditions, thus improving
the efficacy of peptide synthesis.15-18 Such building blocks are
readily accessible by cyclocondensation of Ser, Thr, or Cys with
aldehydes or ketones18 (Figure 1) and represent useful protecting
techniques for the parent amino acids. Furthermore, variation
of the substituents R1 and R2 at position 2 affects the ring
stability, resulting in pseudo-proline (ΨPro) of differential
chemical stability.17
1
2
In this paper, we describe the imide cis-trans distribution
upon variation of the 2-C substituents (R1, R2) by NMR
(9) Schmid, F. X.; Mayr, L. M.; Mu¨cke, M.; Scho¨nbrunner, E. R. AdV.
Prot. Chem. 1993, 44, 25-66.
X Abstract published in AdVance ACS Abstracts, January 1, 1997.
(1) Rose, G. D.; Gierasch, L. M.; Smith, J. A. AdV. Prot. Chem. 1985,
37, 1-109.
(2) Mu¨ller, G.; Gurrath, M.; Kessler, H. Proteins: Struct., Funct., Genet.
1993, 15, 235-251.
(3) Richardson, J. S.; AdV. Protein Chem. 1981, 34, 116-339.
(4) Smith, J. A.; Pease, L. G. CRC Crit. ReV. Biochem. 1980, 8, 315-
399.
(5) MacArthur, M. W.; Thornton, J. M. J. Mol. Biol. 1991, 218, 397-
412.
(6) Kim., P. S.; Baldwin, R. L.; Annu. ReV. Biochem. 1982, 51, 459-
489.
(10) Zabrocki, J.; Dunbar Jr., J. B.; Marshall, K. W.; Toth, M. V.;
Marshall, G. R. J. Org. Chem. 1992, 57, 202-209.
(11) Gramberg, D.; Weber, C.; Beeli, R.; Inglis, J.; Bruns, C.; Robinson,
J. A. HelV. Chim. Acta 1995, 78, 1588-1606.
(12) Delaney, N. G.; Madison, V. J. Am. Chem. Soc. 1982, 104, 6635-
6641.
(13) Chalmers, D. K.; Marshall, G. R. J. Am. Chem. Soc. 1995, 117,
5927-5937 and references cited therein.
(14) Grathwohl, C.; Wu¨thrich, K. Biopolymers 1981, 20, 2623-2633.
(15) Mutter, M.; Nefzi, A.; Sato, T.; Sun, X.; Wo¨hr, T. Pept. Res. 1995,
8, 145-153.
(16) Haack, T.; Mutter, M. Tetrahedron Lett. 1992, 33, 1589-1592.
(17) Wo¨hr, T.; Wahl, F.; Nefzi, A.; Rohwedder, B.; Sato, T.; Sun, X.;
Mutter, M. J. Am. Chem. Soc. 1996, 118, 9218-9227.
(18) Wo¨hr, T.; Mutter, M. Tetrahedron Lett. 1995, 36, 3847-3848.
(7) Fischer, G. Angew. Chem., Int. Ed. Engl. 1994, 33, 1415-1436 and
references cited therein.
(8) Chen, J. K.; Schreiber, S. L. Angew. Chem., Int. Ed. Engl. 1995, 34,
953-969 and references cited therein.
S0002-7863(96)02780-1 CCC: $14.00 © 1997 American Chemical Society

Pseudo-Prolines as a Molecular Hinge
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 919
Table 1. cis to trans Ratio of Peptides
Ac-Ala-Xaa[ΨR ,R pro]-NH-Me 1-6 and
1
2
succinyl-Val-Xaa[ΨR ,R pro]-Phe-NH-pNA 7-10
1
2
compd
R1
R2
Xaa
X
cis (%)a
1
2
3
4
5
6
7
8
9
Me
Me
Ser
Thr
Cys
Ser
Pro
Ser
Ser
Ser
Ser
Ser
O
O
S
O
CH2
O
O
O
O
∼100
∼100
∼100
∼33
∼18
∼63
∼15
> 95
∼10
∼50
Me
Me
H
Me
Me
H
biphenyl
H
Me
H
PMP
H
H
Me
PMPb
H
10
O
a Determined by 1H NMR in DMSO at 300 K. b PMP: p-methoxy-
phenyl.
data. Complete proton resonance assignments were made using (DQF)
COSY experiments, HMQC was used to assign unambiguously the
carbon resonances. J coupling constants were directly measured from
high-resolution 1H-1D spectra ((0.1 Hz) or simulated for complex or
overlapped systems. Prochiral assignments for the ΨPro â and Me
protons were achieved, in some cases, on the basis of relative interproton
ROE intensities and J coupling constant values.
Computational Methods. All the simulations were carried out in
Vacuo using the CVFF24 force field as implemented in Discover.23 The
calculations used a value of 1 for the macroscopic dielectric constant
and an infinite cutoff for nonbonded interactions. Structure minimiza-
tions for each isomer were carried out using the Newton-Raphson
algorithm until the gradient was less than 0.001 kcal/mol. Ramachan-
dran plots25 were obtained by mapping the energy of the structure upon
incrementation of the Φ/Ψ dihedral angles of alanine. Contour levels
are plotted with an interval of 10 kcal/mol.
Figure 1. Reaction of Ser, Thr, or Cys (Xaai)-containing peptide I
with carbonyl compounds resulting in the formation of pseudo-prolines
(Xaa[ΨR ,R pro]i) exhibiting a cis (ωi-1 ) 0°, II) or trans (ωi-1 ) 180°,
III) amide bond, depending on the nature of R1 and R2. Cleavage of
the cyclic system with acid (H+) transforms the constrained peptide I
backbone to a regular all-trans amide conformation.
1
2
spectroscopy. This distribution is essentially driven by the
number and steric requirements of the 2-C substituents which
destabilize one of the two rotamers. We demonstrate that
variation of these substituents at the position 2 of the ring system
provides an efficient synthetic tool to tailor the cis to trans ratio
of the Xaai-1-Xaai[ΨR ,R pro] amide bond in peptides.
1
2
Results
Methods
In order to evaluate the effects exerted by the substituents in
position 2 within the heterocycle of ΨPro on the cis-trans
isomerization of the peptide bond, we prepared a number of
Peptides Synthesis. The ΨPro-containing peptides were synthesized
according to methodologies17,18 described earlier. Unsubstituted ΨPro’s
4 and 7 (Figure 1) were obtained by cyclocondensation with formal-
dehyde on the corresponding free amino acid. 2-Alkylated ΨPro’s
were obtained by cyclocondensation with the corresponding ketal (2,2-
dimethoxypropane for 2,2-dimethyl-ΨPro compounds 1-3 and 8, 2,2-
dimethoxyanisaldehyde for compounds 9 and 10) at the dipeptide stage
according the post-insertion method18 and subsequent peptide chain
prolongation.
peptide derivatives (1-6, Table 1) of the types Ac-Alai-1
-
-
Xaai[ΨR ,R pro]-NHMe(Figure1)andsuccinyl-Vali-1-Xaai[ΨR ,R
1
2
1
2
pro]-Phe-NH-pNA (compounds 7-10). Both N and C peptide
termini were capped to minimize intramolecular electronic
interactions, which have been shown to alter the cis-trans ratio
of the Xaa-Pro bond.14 As relevant spectral parameters on
thiazolidine are available in literature, we focused mainly on
the 2-substituted oxazolidine-4(S)-carboxylic acid derivatives.
NMR Spectroscopy. Spectra were recorded at 400 MHz using a
Bruker ARX spectrometer at 300 K. Samples of around 25 mg were
dissolved in 0.4-0.5 mL of DMSO-d6. Chemical shifts were calibrated
with reference to the residual DMSO signal (1H, 2.49 ppm; 13C, 39.5
ppm). 2D experiments were typically acquired using 2K × 512
matrices over a 4000 Hz sweep width in both dimensions. Quadrature
detection in the indirect dimension was achieved by using the TPPI
procedure.19 Scalar connectivities were recovered from 2D double
quantum filtration (DQF) COSY experiments.20 Dipolar connectivities
were obtained by either the conventional NOESY sequence21 or the
The conformational preference of the Xaai-1-Xaai[ΨR ,R
1
2
-
1
pro] imide bond was investigated by H NMR and 13C NMR
spectroscopy. Examination of the 1D spectra in DMSO of
model compounds 1-10 immediately reveals two spectroscopic
characteristics: 2,2-dimethyl-ΨPro derivatives (1-3 and 8)
exhibit a predominant set of resonances for the cis conformer,
while for the monosubstitued compounds 6, 9, 10, the unsub-
stituted compounds 4 and 7, and the proline containing
compounds 5, two sets of resonances were obtained due to a
comparable amount of cis and trans conformers in solution.
Table 1 summarizes the cis-trans ratios of the Xaa-ΨPro bond
measured by 1H NMR for different 2-C substituents of the five-
membered ring. For the proline analogue 5, the occurrence of
both isomers in solution is readily confirmed by 13C NMR on
the basis of the well-documented shift differences observed for
γ and â carbons.26,27
ROESY sequence22 with mixing times from 150 to 200 ms.
A
randomization of the mixing length ((5%) was introduced in the
NOESY experiments in order to minimize coherence transfer. The spin
lock mixing interval of the ROESY sequence was applied by coherent
CW irradiation at γB2/2π ) 2 KHz. Experimental data processing was
performed using the Felix software package.23 The standard sinebell
squared routine was employed for apodization with a shift range of
60-90° and zero filling in both dimensions before 2D transformations
were applied to end up with square matrices of 2K × 2K real point
(19) Drobny, G.; Pines, A.; Sinton, S.; Wertekamp, D.; Wemmer, D.
Faraday DiV. Chem. Soc. Symp. 1979, 13, 48-58.
(20) Piantini, V.; Sørensen, O. W.; Ernst, R. R. J. Am. Chem. Soc. 1982,
104, 6800-6801.
2-C Disubstituted ΨPro-Containing Peptides. The obser-
vation of a highly predominant set of resonances in the 1H NMR
(21) Jeener, J.; Meier, B. H.; Bachmann, P.; Ernst, R. R. J. Chem. Phys.
1979, 71, 4546-4553.
(24) Hagler, A. T.; Huler, E.; Lifson, S. J. Am. Chem. Soc. 1974, 96,
5319-5327.
(22) Bothner-By, A. A.; Stephen, R. R-L.; Lee, J., Warren, C. D.; Jeanloz,
R. W. J. Am. Chem. Soc. 1984, 106, 811-813.
(25) Ramachandran, G. N.; Sasisekharan, V. AdV. Protein Chem. 1968,
23, 283-437.
(23) Biosym Technologies Inc., San Diego, CA.
(26) Dorman, D. E.; Bovey, F. A. J. Org. Chem. 1973, 38, 2379-2383.

920 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dumy et al.
a
Table 2. ΨPro Relevant Spectral Parameters for Compounds 1-10 in DMSO-d6
entry
compd
R (3J RH-âH)b
â (2J âH-âH)b
2-H(Me) (2JHH
)
1
2
3
4
Ser(ΨMe,Mepro)
Thr(ΨMe,Mepro)
Cys(ΨMe,Mepro)
Ser(ΨH,Hpro)
4.44 (6.1pro-R/-)
4.17/4.05pro-R (8.8)
4.12/1.29
1.52/1.43pro-S
1.52/1.50pro-S
1.93/1.85pro-S
5.17/4.96 (3)
4.88
4.03 (5.7)
5.34 (6.4pro-R/-)
4.31
3.62 pro-R/3.20 (11.8)
4.09/3.83
4 (cis)
7
8
9 (trans)
10 (trans)
10 (cis)
Ser(ΨH,Hpro)
4.48
4.18/4.03
Ser(ΨH,Hpro)
4.485 (7.3 pro-R/5.0)
4.64 (6.5 pro-R/1.9)
4.65 (7.4 pro-R/5.6)
4.704 (7.2 pro-R/3.1)
4.853 (6.3 pro-R/1.9)
4.174 pro-R/3.818 (8.7)
4.102 pro-R/4.00 (9.1)
4.122 pro-R/4.005 (8.9)
3.979 pro-R/3.864 (9.3)
4.29 pro-R/4.02 (8.9)
5.289 (3.7)/4.95
Ser(Ψ Me,Mepro)
2-(R)-Ser(ΨH,PMPpro)
2-(S)-Ser(ΨPMP,Hpro)
2-(S)-Ser(Ψ PMP,Hpro)
1.43
6.396
6.32
6.17
a Chemical shifts (δ in ppm; (0.01). b R stands for the 4-H, â for the 5-H, J (in Hz ( 0.2).
2
1
Figure 2. Expanded region of the D H NMR ROESY spectrum at
400 MHz (τm ) 200 ms) of 1, in DMSO at 300 K. The boxed peaks
indicate the RHi-1-RHi and âHi-1-RHi (right) connectivities between
Alai-1 and Xaai[ΨMe,Mepro] typically observed for the cis conformer.
2
1
Figure 3. Expanded region of the D H NMR ROESY spectrum of
10 showing the assignment of the cis and trans isomer (400 MHz, τm
) 200 ms, in DMSO at 300 K). The cis and trans isomers are denoted
by a small superscript letter (c or t) in the 1D NMR spectrum; exchange
ROE cross peaks are labeled with an e on the map (same phase as
diagonal).
and 13C NMR spectra of peptides 1-3 upon varying both solvent
(H2O/D2O, CDCl3, or DMSO-d6) and temperature is consistent
with the presence of a major conformation in solution. From
2D 1H NMR ROESY experiments we observe the typical
pattern28 expected for a cis amide bond (Figure 2, boxed peak),
i.e. RHi-1-RHi and âHi-1-RHi ROE cross-peaks between
Alai-1 or Vali-1 and Xaai[ΨMe,Mepro] reflecting the spatial
proximity of the corresponding protons in the cis form. For
peptide 8, a set of minor resonances is also detected (less than
5% as estimated by 1D integrals) and is assigned to a minor
conformation in slow exchange with the major one on NMR
time scale. However, neither chemical shifts criteria (Vide
supra) nor ROE measurements allow for the assignment of the
latter minor species to the trans form (Vide supra). Stereospe-
cific assignment is realized for the ΨPro â-protons on the basis
of their relative ROE cross-peak intensities with the ΨPro R
proton and their 3J RH-âH values (Table 2). The pro-R
â-protons (anti to the carboxyl group, Figure 1, R3 ) H) are
shifted downfield (except for 1) and exhibit a greater coupling
constant (Table 2) with the R-proton compared to that of the
pro-S â-proton (syn to the carboxyl group). On the basis of
the latter assignment, the upfield 2-methyl group (syn to the
carboxyl, Figure 1, R2 ) H) is assigned to pro-R by the presence
of a ROE cross-peak with the pro-S â-proton. Interestingly,
upon substitution of Ala for Gly or D-Ala in peptide 2, no
significant effect on the backbone conformation is observed by
NMR (data not shown), indicating a pronounced preference of
the cis conformer in the 2,2-dimethyl-ΨPro-containing peptides
1-3.
2-C Unsubstituted ΨPro-Containing Peptides. For pep-
tides 4 and 7, the occurrence of exchange cross peaks in ROESY
experiments between each observed set of resonances and their
coalescence upon warming points to the presence of cis and
trans conformers in slow exchange is similar to that of proline-
containing peptides26,27 and thiazolidine derivatives.29-31 The
more abundant trans isomer is characterized by a RHi-1-2-Hi
and âHi-1-2-Hi ROE cross-peak between Alai-1 or Vali-1 and
2-Hi in analogy to the short RHi-1-δHi distance typically
observed in trans-proline peptides.28 The cis isomer is only
characterized by an Ri-Ri exchange cross-peak, the R- and
â-protons being broadened and deshielded compared to those
of the trans form, while the protons at position 2 are found
upfield in agreement with previous reports on related thiazolidine
derivatives.31
2-C Monosubstituted ΨPro-Containing Peptides. The
presence of the two conformers in solution is confirmed by the
same type of ROE cross peaks (Figure 3) as mentioned for the
2,2-dimethyl and unsubstituted compounds (Vide infra).
A striking difference between the two diastereoisomers with
respect to the cis-trans population is observed in the 1D spectra.
For the 2-(S) epimer (Vide supra), the distribution of the 2-H
signals (cis and trans) is nearly identical, while in the second
case, a ratio of 90:10 is measured (Figure 4).
Stereospecific assignments of the ΨPro â-protons (Table 2)
were performed as exemplified above for the 2-dimethyl
(27) Kessler, H. Angew. Chem., Int. Ed. Engl. 1982, 21, 512-523. For
1
the proline containing peptide 5, the H and 13C NMR spectra display the
(29) Goodman, M.; Niu, G. C.-C.; Su, K.-C. J. Am. Chem. Soc. 1970,
92, 5219-5220, 5520-5222.
(30) Samanen, J.; Zuber, G.; Bean, J.; Eggleston, D.; Romoff, T.; Kopple,
K.; Saunders, M.; Regoli, D. Int. J. Pept. Protein Res. 1990, 35, 501-509.
(31) Borremans, F. A. M.; Nachtergaele, W. A.; Budesinsky, M.;
Anteunis, M. J. O. Bull. Soc. Chim. Belg. 1980, 89, 749-758.
two sets of resonances expected for the cis and trans conformers in
equilibrium. This is confirmed by the 13C chemical shift difference measured
between γ- and â-carbons of proline and also by ROESY experiments.
(28) Wu¨thrich, K. NMR of Proteins and Nucleic Acids; John Wiley &
Sons, Inc.: New York, 1986; pp 117-125.

Pseudo-Prolines as a Molecular Hinge
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 921
(trans in that case) is much less pronounced. Interestingly, for
compound 6, the two rotameric forms have nearby the same
energy values along the contour plot for the (S)-epimer (Figure
5, top) while differing by up to 3 kcal/mol for the (R)-form
(Figure 5, bottom).
Discussion
In general, steric interactions between the C(δ) pyrrolidine
and the preceding residue provide some rational for the cis-
trans ratio found in proline-containing peptides and proteins.14
The trans peptide bond introduces steric interactions between
the δ-CH2 in Pro and the preceding residue resulting in an
increase of the net free energy difference between the cis and
trans conformers. In ΨPro-containing model peptides, the
substituents at the 2-Ci of the cyclic system interact severely
with the side chain Ri-1 or the peptide backbone of the preceding
residue (Figure 1, III) in the trans form, whereas in the cis
conformer, the corresponding interactions stem from the car-
bonyl group of residue i - 1. Consequently, in compounds
1-3 and 8, the two 2-C methyl groups (pro-R and pro-S) result
in more severe steric restriction in the trans conformer compared
to the cis. These findings are confirmed by comparing the
energy differences between the two rotamers along the Ram-
achandran plot. Starting from a given value of the cis (Φ, Ψ)
angles, rotation of 180° around ωi-1 results in the trans
conformer with up to 15 kcal/mol higher in energy. This is
consistent with the substitution of Alai-1 for Gly or for D-Ala
which did not result in detectable amounts of the trans
conformer. Furthermore, substitution of Ala for Gly also
indicates that not only the side chain but also the peptide
backbone accounts for steric interactions with the 2-C substit-
uents of ΨPro. Despite the strong thermodynamic preference
for the cis in 2,2-dimethyl-containing ΨPro derivatives, the
presence of a small amount of trans conformer cannot be ruled
out by NMR measurements (Vide infra). Preliminary experi-
ments based on the chymotrypsin-coupled assays developed by
Fischer7 show that in compound 8 about 8% of the trans isomer
is present in aqueous solution. A similar preference for cis has
been observed for a compound based on δ-dimethyl-substitued
proline.33 Notably, strong preference for the cis conformation
in solution has been reported for N-acetylated 2,2-methyl-
thiazolidine on the basis of chemical shift criteria and 1D-NOE
analysis.34 Crystallographic data on the protected dipeptide
Fmoc-Ala-Cys(ΨMeMepro)-OH reveal also the cis geometry35
for the peptide bond. Most notably, an inversion of that cis-
trans preference has been reported for N-formyl-2,2-methyl-
thiazolidine,36 supporting the findings that steric hindrance is
the major driving force for adapting a high cis content.
Figure 4. Expanded region of the 1D NMR of 9 (bottom) and 10
showing the 2-H signal of the trans (left downfield) and cis (right
upfield).
Table 3. Amide Geometry and cis-trans Energy Calculated for
Compounds 1, 4, 5, and 6 (R/S)
peptide
energy (trans/cis) (kcal/mol)
1
4
-44.9/-48
20.1/19.9
5
27.5/29.9
124.7/126.6
120.4/123.6
6 (S)
6 (R)
compounds. The stereochemistry of the 2-C position inferred
for 9 and 10 is on the basis of ROE cross peaks between the
2-C substituents, namely the ortho protons of the p-methox-
yphenyl (PMP) group or the 2-H, with the â-protons of the
ΨPro. For instance, the 2-(S) epimer (Figure 1, R1 ) PMP, R2
) H) is set by the ROE cross peaks between the 2-H and the
pro-S â-proton (cis relationship) while a ROE between the PMP
ortho protons and the pro-S â-proton is of diagnostic value for
the 2-(R) epimer. For compound 6, the 2-(S) center can also
be assigned on the basis of the 2-H and the â-(pro-S)-proton
ROE cross peak, in harmony with the X-ray data of the
compound.32 Figure 3 shows the assignment of all the cis and
trans protons of compound 10. The chemical shift values show
that R- and â-protons are shifted downfield while the 2-H proton
is upfield in the cis form, supporting the corresponding
observations for the unsubstituted compounds 4 and 7. Dif-
ferences in the 3J RH-âH values are also indicative: the pro-S
â-protons range from 2 Hz in the cis to 3-6 Hz in the trans
form, whereas for the pro-R â-protons a constant value of 6-7
Hz regardless of the nature of the isomer is observed.
Beside a significant amount of the cis conformer in solution
(Table 1), the unsubstituted peptide 4 adopts a preference for
the trans conformation with a value of about 0.5 kcal/mol for
the free energy difference of the cis-trans equilibrium similar
to that of the proline-containing peptide 5. Similarly, unsub-
stituted 2-thiazolidine-4-carboxylic acid analogues also display
a strong preference for the trans form and have been used as a
proline surrogate into biological peptides37 and polymers.29 The
structural similarity of compound 4 with proline is also observed
in the solid state as delineated by their X-ray structures.32
Conformational Energy Calculations. Table 3 summarizes
the conformational energies found after minimization of com-
pounds 1, 4, 5, and 6 (both 2-C epimers) in the cis or trans
form. While in general these values do not perfectly match
the experimental data found in solution, they still reflect the
general tendencies of the cis and trans forms.
Although the same dihedral angle values are found qualita-
tively along the contour lines in the Ramachandran map,25 the
analysis of the energy difference corresponding to a given couple
of (Φ, Ψ) for ωi-1 ) 0 and ωi-1 ) 180° is instructive. For
compound 1, the preference for cis amounted to a value up to
15 kcal/mol, whereas for proline-containing peptide 4 and for
the unsubstituted derivative 5, the preference for one conformer
(33) Magaard, V. W.; Sanchez, R. M.; Bean, J. W.; Moore, M. L.
Tetrahedron Lett. 1993, 34, 381-383.
(34) Savrda, J. Peptides: Chemistry and Biology; ESCOM: Leiden, The
Netherlands, 1976; pp 653-656.
(32) Evaluation of the rate constant and thermodynamic parameters of
the cis-trans equilibrium in compounds 4-7, 9, and 10 as well as the
solid state conformation for the peptides 4-6 will be subject of a separate
publication.
(35) Nefzi, A.; Schenk, K.; Mutter, M. Protein Pept. Lett. 1994, 1, 66-
69.
(36) Toppet, S.; Claes, P.; Hoogmartens, J. Org. Magn. Res. 1974, 6,
48-52.

922 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dumy et al.
Figure 5. Ramachandran plot of compounds 9 (bottom) and 10 calculated for each rotameric form.
The presence of a single substituent leads to cis-trans ratios
depending on the 2-C chirality. As for the 2,2-dimethyl-ΨPro,
the contour levels in the Ramachandran plot (Figure 5) reveal
that steric factors determine the preferential geometry of the
imide bond. When the substituent is anti to the carboxyl group
(2-(S), Figure 1, R2 ) H), the two forms have nearly the same
free energy (∆G°t-c ≈ 0), whereas a difference in energy of
about 1.3 kcal/mol is found for the 2-(R) epimer by NMR. The
anti and syn diastereoisomers of N-acetyl-5-methylproline N′-
methylamide are reported38 to display a similar cis to trans
distribution in solution which again can be rationalized invoking
steric effect of the 5-substituents. Such a dependency on the
chirality at the 2-C position is reported in the literature for
thiazolidine congeners, but no unequivocal assignments of the
cis or the trans form have been provided so far.39 It is important
to note that the chirality is usually inferred by the sum of the
coupling constants between the R- and â-protons of the
thiazolidine ring. This sum is calculated for temperatures above
the coalescence of the two isomers on the basis of the 1D NMR
spectrum in order to simplify the corresponding spectra and to
overcome the problems of resonance overlapping.39,40 A sum
less than 10 Hz is of diagnostic value for the (S)-epimeric center,
whereas for values exceeding 10 Hz, the (R)-form is inferred.
The isomer weighted sum results in 9.2 Hz for compound 10
and 13 Hz for compound 9 (Table 2), resulting according to
the above rule in the (S) and (R) stereochemistry assignments
for the 2-C center in agreement with our NMR assignments.
Thus, the 2-C chirality of the ΨPro compounds can be
1
determined directly by H NMR using the coupling constant
values and the cis to trans distribution. The cis and trans isomer
can be easily assigned by 1D NMR on the basis of their chemical
shift differences and coupling constant.
The differences in coupling constant values in both epimers
also provide information on the puckering of the heterocycle.32
Table 2 shows that the 3J RH-âH values are conserved within
each isomeric form irrespective of the nature of substitution.
3
In the cis conformation, the sum of the J RH-âH coupling
constants is always less than 9 Hz (Table 2), corresponding to
a staggered conformation along the CR-Câ bond. A greater
sum is only consistent with a conformation where the R-proton
is gauche to the pro-R â-proton while the pro-S â-proton is
pseudo-axial. On the basis of X-ray data30,35 and similar
conclusions31 established for thiazolidine, these arrangements
lead to the cis form with a preferential â-exo envelope
conformation. The fact that one arrangement of the substituent31
corresponds to one isomer form may reflect an influence of the
puckering for the cis-trans ratio. Although the puckering
contribution alone cannot be easily decoupled and evaluated
from the steric effects,36 it still may explain the observed
difference between 9 and 10. In the 2-(S) diastereoisomer, steric
hindrance between PMP and the preceding Val does not
destabilize the pseudo-equatorial (cis) form relative to the
pseudo-axial position (trans) form whereas the PMP group in
a pseudo-equatorial position destabilizes the cis in the 2-(R)
(37) Kiso, Y. Biopolymers (Pept. Sci.) 1996, 40, 235-244. Samanen,
J.; Narindray, D; Cash, T.; Brandeis, E; Adams, W., Jr.; Yellin, T.;
Eggleston, D.; DeBrosse, C.; Regoli, D. J. Med. Chem. 1989, 32, 466-
472.
(38) Delaney, N. G.; Madison, V. Int. J. Pept. Protein Res. 1992, 19,
543-548.
(39) Szilagyi, L.; Gyo¨rgydeak, Z. J. Am. Chem. Soc. 1979, 101, 427-
432.
(40) Benedeti, F.; Ferrario, F.; Sala, A.; Sala, L.; Soresinetti, P. R. J.
Heterocyclic Chem. 1994, 31, 1343-1347.

Pseudo-Prolines as a Molecular Hinge
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 923
Mass spectra were obtained by electron spray ionization (ESI-MS) on
a Finnigan LC710 or by chemical ionization (CI-MS) on a Nermag
R10-10C. 1H-NMR spectra were obtained on a Bruker-WH250 or
Bruker DPX-400 with trimethylsilane as the internal standard for
intermediate compounds. Infrared spectra were recorded on a Perkin-
Elmer 1430 spectrometer using the KBr disc method. Elemental
analysis ((0.4%, C, H, N) has been performed at the Mikroanalytisches
Laboratorium (Kronach, Germany). Melting points are uncorrected
values measured with a Bu¨chni 510 point apparatus.
General Procedure for the Preparation of ΨPro-Containing
Dipeptide. Fmoc-Ala-ΨPro-OH was obtained by reacting Ser, Thr,
or Cys with formaldehyde to give the unsubstituted compounds or by
reacting Ser- or Thr-containing dipeptides with the corresponding ketal
according to published procedures of the post-insertion method.17,18 In
the latter case, the reaction was carried out in the presence of catalytic
amounts of pyridyltoluene-4-sulfonate (PPTS) and 2,2-dimethoxy-
propane (DMP) in THF.
Ac-Ala-Ser(ΨMe,Mepro)-NHMe (1). Z-Ala-Ser(ΨMe,Mepro)-OH17
(1.35 g, 4.0 mmol) was dissolved in 30 mL of THF/DCM (1:1; v/v)
and reacted with 1.14 g (5.54 mmol) of DCC and 1.05 g (6.65 mmol)
of HOBt at room temperature. After the suspension was stirred for 20
min, 1.0 mL of a 8.02 M solution of methylamine in ethanol was added.
The mixture was stirred for 10 min at 0 °C and further 30 min at room
temperature before filtering off the unsoluble solid and evaporating
the solvent. The resulting material was taken up in 100 mL of ethyl
acetate, washed [with a 5% aqueous sodium carbonate solution (3 ×
50 mL), a 5% aqueous solution of citric acid (3 × 50 mL), and brine
(3 × 50 mL)], and dried over magnesium sulfate. The solvent was
evaporated and the residue purified on silica gel (flash, eluent CHCl3/
MeOH, 100:4), resulting in 1.29 g (89%) of Z-Ala-Ser(ΨMe,Mepro)-
NHMe as a white powder. C18H25N3O5 (363.41). Mp 134-136 °C.
Rf (CHCl3/MeOH, 20:1) 0.29. tR ) 17.5 min (20-80% B, 30 min,
C18). CI-MS (NH3): 364 (1, [M + 1]+), 305 (23, [M - acetone]+),
206 (8), 143 (5), 127 (11), 91 (100). Pd (98 mg) on charcoal was
added to a solution of 980 mg of Z-Ala-Ser(ΨMe,Mepro)-NHMe in 30
mL of methanol under H2 atmosphere. After complete disappearance
of the starting material, the mixture was filtered through Celite and
the solvents were removed under reduced pressure. The acetylation
was performed in 10 mL of acetic anhydride/water (1:1; v/v) within 2
h at room temperature. The solvents were removed under reduced
pressure, and the crude product was purified by column chromatography
(eluent EtOAc/MeOH, 10:1) to give 630 mg (98%) of Ac-Ala-Ser-
(ΨMe,Mepro)-NHMe as a colorless solid. C12H21N3O4 (271.32). Mp
60-70 °C. tR ) 14.1 min (5-50% B, 30 min, C18). CI-MS (NH3):
272 (2, [M + 1]+), 214 (23, [M - acetone]+), 185 (8), 156 (1), 143
(16), 127 (20), 114 (43, [AcAla]+), 91 (100), 86 (60). IR (KBr): 3274
(m), 3059 (w), 2988 (w), 2942 (w), 1641 (vs), 1552 (s), 1424 (s), 1383
(m), 1340 (m), 1306 (m), 1253 (m), 1232 (m), 1203 (w), 1145 (s),
1093 (w), 1057 (s), 1008 (w), 954 (w). Anal. (C12H21N3O4) C, H, N.
Ac-Ala-Thr(ΨMe,Mepro)-NHMe (2). The compound was synthe-
sized starting from Z-Ala-OSu and L-threonine by following the above
procedure for Ac-Ala-Ser(ΨMe,Mepro)-NHMe: C13H23N3O4 (285.34).
Mp 47-50 °C. tR ) 15.6 min (5-50% B, 30 min, C18). CI-MS
(NH3): 286 (86, [M + 1]+), 228 (70, [M - acetone]+), 199 (10), 141
(11), 114 (100), 99 (13), 86 (30). IR (KBr): 3312 (m), 3084 (w),
2985 (w), 2939 (w), 1640 (vs), 1551 (s), 1420 (s), 1376 (s), 1250 (m),
1216 (w), 1170 (s), 1113 (m), 1094 (m), 989 (m), 954 (m). Anal.
(C13H23N3O4) C, H, N.
isomer. In the same vein, the spatial arrangement of the
substituents should also have a subtle effect in 2-monosubstituted
compound on the anti and syn transition41 state by decreasing
one state relative to the other as found for 5-methyl-substitued
proline derivatives.38 The determination of the corresponding
kinetic parameters are currently under investigation in our
laboratory.32
Due to the described structural properties, the incorporation
of a ΨPro residue into peptide backbone provides also a
powerful means to overcome some fundamental problems of
peptide synthesis.17 For example, ΨPro residues have been
shown to disrupt â-sheet structure formation during stepwise
peptide synthesis, resulting in increased solvation of the peptide
chain. The cis-trans isomerization process described here
provides a rational for these experimental findings. The cis-
trans isomerization induced by ΨPro results in a kink confor-
mation of the peptide backbone, thus preventing the onset of
peptide self-aggregation. ΨPro therefore may act as a molecular
hinge within the peptide chain.
Conclusion
Incorporation of a pseudo-proline residue into the peptide
backbone results in variable cis-trans ratios of the preceding
amide bond depending on the nature of the 2-C substituents of
the ring system. The conformational preference is essentially
driven by the number and steric requirements of the 2-C
substituents which destabilize one of the two conformers. Due
to the comparable structural properties with proline and their
high chemical stabilities, unsubstituted Ser-, Thr-, and Cys-
derived ΨPro’s represent useful proline surrogates. In contrast,
the unusually high cis amide content of 2-C disubstituted ΨPro
offers a simple and attractive means to constrain selectively
peptide bonds, i.e. to induce a cis amide in the peptide backbone
especially for the temporary insertion of â-turns into polypeptide
chains. Most notably, type VI â-turns with 2,2-methyl-ΨPro
residue at position i + 2 can be induced.42 Ring opening under
acidic conditions allows to reconstitute the regular all-trans
peptide. More subtle effects on the imide bond geometry are
observed for 2-C monosubstituted ΨPro. In this context, ΨPro
represent a versatile tool for investigating the influence of ring
substitution upon the isomerization of the imide bond. On the
basis of these results and their ready synthetic access, ΨPro
can be used as a molecular hinge to explore and control the
peptide backbone conformation. Thus, pseudo-prolines offer
a wide range of applications as in peptide-based drug and
prodrug design, molecular recognition studies, or protein folding
and self-aggregation processes.
Experimental Section
Materials. All protected amino acids were purchased from Cal-
biochem-Novabiochem AG (La¨ufelfingen, Switzerland); reagents and
solvents were purchased from Fluka (Buchs, Switzerland) and used
without further purification. HPLC was performed on Waters equip-
ment using columns packed with Vydac Nucleosil 300 Å 5 µm C18
particles unless otherwise stated. The analytical column (250 × 4.6
mm) was operated at 1 mL/min and the preparative column (250 × 21
mm) at 18 mL/min, with UV monitoring at 214 nm. Solvent A
consisted of 0.09% TFA and solvent B of 0.09% TFA in 90%
acetonitrile. TLC was performed on silica gel plates Merck 60 F254,
visualized by UV or 5% vanillin in concentrated sulfuric acid. Flash
chromatography was performed on Merck silica gel 60 (40-63 mesh).
Ac-Ala-Cys(ΨMe,Mepro)-NHMe (3). Fmoc-Ala-Cys(ΨMe,Mepro)-
OH17 (100 mg, 220 µmol) was reacted with a solution of dimethyl-
formamide (DMF) containing 30% diethylamine for 60 min. After
the completeness of the reaction (TLC), the solvent was removed and
the residue treated with CHCl3 and then filtered to afford 32 mg (63%)
of a white solid. Subsequent acetylation was performed with Ac2O
(12.3 µL) in DMF in the presence of N,N-diisopropylethylamine (24
µL) to provide 35 mg of the desired acetylated compound. Subse-
quently, the compound was reacted with 80 mg of pyBOP in a saturated
solution of methylamine in DMF (3 mL) for 1 h. The solvents were
removed under reduced pressure, and the crude product was purified
by column chromatography (eluent CHCl3/MeOH/AcOH, 92:6:2) to
give 30 mg (98%) of Ac-Ala-Cys(ΨMe,Mepro)-NHMe as a colorless
solid.
(41) Fischer, S.; Dunbrack, Jr.; Karplus, M. J. Am. Chem. Soc. 1994,
116, 11931-11937.
(42) Ryan, D. E.; Dumy, P; Eggleston, I. M.; Mutter, M. InnoVation
and PerspectiVes in Solid Phase Synthesis; Epton, R., Ed.; Mayflower
Worldwide Ltd.: Birmingham, U.K., in press.

924 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dumy et al.
Ac-Ala-Ser(ΨH,Hpro)-NHMe (4). To a solution of 5.0 g (12.2
mmol) of Fmoc-Ala-Ser(ΨH,Hpro)-OH,17 were added 1.44 g (12.2
mmol) of N-hydroxysuccinimide, 70 mL of dioxane/ethyl acetate (6:1,
v/v), and 2.76 g (13.4 mmol) of DCC at 0 °C. The suspension was
stirred overnight at room temperature. After filtration the solvent was
evaporated. The residue was dissolved in 150 mL of ethyl acetate and
washed with a 5% solution of sodium carbonate (2 × 80 mL) and brine
(2 × 80 mL). The combined organic phases were dried over
magnesium sulfate and evaporated. The crude product was purified
by precipitation from dichloromethane (DCM)/hexane to yield 4.76 g
(9.37 mmol, 78%) of Fmoc-Ala-Ser(ΨH,Hpro)-OSu as a white powder.
Fmoc-Ala-Ser(ΨH,Hpro)-OSu: C26H25N3O8 (507.50). Mp 98-100 °C.
Rf (CHCl3/MeOH/AcOH, 100:1:1) 0.18. tR ) 14.6 min (40-100% B,
30 min, C18). 1H-NMR (250 MHz, DMSO-d6, 49 mM): 7.90-7.29
(m, 8 arom. H), 7.83 (d, 3J ) 7.1, HN), 5.13 (br., 2 H, H2-C(2)), 5.00
(d, J ) 7.0, H3-CâAla). CI-MS (NH3): 326 (0.2), 266 (0.5),178 (100).
IR (KBr): 3312 (m), 2980 (m), 1718 (vs), 1639 (vs), 1528 (m), 1450
(s), 1247 (s), 1074 (m), 741 (s). The transformation of 3.20 g (7.83
mmol) of Fmoc-Ala-Pro-OH to Ac-Ala-Pro-NHMe was performed by
following the procedure for the synthesis of Ac-Ala-Ser(ΨH,HPro)-
NHMe. The final product was crystallized from DCM/cyclohexane
to yield 840 mg (3.48 mmol) of pure Ac-Ala-Pro-NHMe. C11H19N3O3
(241.29). Mp 170-172 °C. Rf (CHCl3/MeOH, 85:15) 0.44. tR ) 20.5
min (0-15% B, 30 min, C18). CI-MS (NH3): 242 (2, [M + 1]+), 210
(1), 183 (3,[M - HNCOMe]+), 155 (2), 129 (19), 114 (51), 86 (9), 72
(100). IR (KBr): 3329 (vs), 2987 (s), 1653 (vs), 1544 (vs), 1418 (s),
1309 (s), 1256 (m), 1199 (m), 972 (m). Anal. (C11H19N3O3) C, H, N.
Ac-Ala-Ser(Ψ(2S)Biphe,Hpro)-NHMe (6). The compound was syn-
thesized starting from Z-Ala-OSu and L-serine by following the above
procedure for Ac-Ala-Ser(ΨMe,Mepro)-NHMe and using biphenyl-4-
carboxaldehyde dimethyl acetal instead of 2,2-methoxypropane (DMP)
in the acid-catalyzed acetalization step. After chromatography purifica-
tion and crystallization, the pure (S)-isomer was obtained. The 2-C
chirality is inferred on the basis of X-ray data obtained on the precursor
Z-Ala-Ser(Ψ(2S)Biphe,Hpro)-OH.32 C22H25N3O4 (395.46). Mp > 180 °C
(dec). Rf (EtOAc/MeOH 17:3) 0.33. tR ) 17.9 min (20-80% B, 30
min, C18). CI-MS (NH3): 396 (1, [M + 1]+), 214 (27, [M -
biphenylcarboxaldehyde]+), 310 (28), 281 (35), 224 (31), 152 (13).
Synthesis of Suc-Val-Ser(ΨH,Hpro)-Phe-pNA (7). Fmoc-Val-
Ser(ΨH,HPro)-OH (0.35 g, 0.8 mmol) was dissolved in DCM (10 mL)
and isobutyl chloroformate (1.05 equiv, 120 µL) and NEM (1.05 equiv,
78 µL) added to give a white suspension which was stirred for 30 min
under nitrogen at -10 °C. H-Phe-pNA (1.05 equiv, 239 mg) dissolved
in DCM (2 mL) was added to give a clear solution which was stirred
for 4 h to give Fmoc-Val-Ser(ΨH,Hpro)-Phe-pNA in 60% yield. DCM
was evaporated and the residual substance taken up in DMF/morpholine
(13 mL of a 5.4% solution). The already slightly yellow solution was
stirred under nitrogen for 16 h to give a deep yellow solution. DMF
was removed and replaced by ethyl acetate to give a white precipitate,
210 mg (54%) of the white substance (H-Val-Ser(ΨH,Hpro)-Phe-pNA:
ESI-MS 484.5 [M + 1]+). The free amino compound (100 mg) was
dissolved with DMF (3 mL) before adding succinic anhydride (2 equiv,
42 mg) and N-ethylmorpholine (NEM) (2 equiv, 42 mg) and stirred
for 12 h. The desired product was then purified by a preparative C18
Sep-Pak column (isocratic 30% A) to give Suc-Val-Ser(ΨH,Hpro)-Phe-
pNA. Mp 124-125 °C. ESI-MS: 584.2 [M + 1]+. tR ) 9.59 min
(40-100% B, 15 min, C18). Anal (C28H33N5O9) C, H, N.
Synthesis of Suc-Val-Ser(ΨMe,Mepro)-Phe-pNA (8). Fmoc-Val-
Ser-OBzl (0.7 g, 1.36 mmol), PPTS (100 mg, 0.3 equiv), and DMP in
toluene/DMF (46 mL/4 mL) under nitrogen atmosphere were heated
under reflux during 15 h to give a slightly yellow solution. After
evaporation of the solvent, the remaining yellow oil was taken up in
CHCl3/MeOH/HOAc (100:3:1, 3 mL) and passed through a silica
column using the same eluent to give 370 mg (49%) of Fmoc-Val-
Ser(ΨMe,Mepro)-OBzl. ESI-MS 557.7 [M + 1]+. tR ) 20.43 min (40-
100% B, 15 min). The dipeptide was dissolved in methanol (20 mL)
and hydrogenated with Pd/C during 2 h to give Fmoc-Ser(ΨMe,Mepro)-
OH in quantitative yield. The product was purified by preparative
HPLC (C18) to give 260 mg (0.56 mmol) of the expected compound.
The activation of the carboxyl group was achieved with cyanuric
fluoride (600 mg) in DCM (10 mL) in the presence of pyridine (44
mg, 0.56 mmol) by heating the solution under reflux for 2 h. To the
yellowish solution was added water (10 mL) dropwise while keeping
the temperature at 0 °C. The organic layer was dried over MgSO4
and the solvent filled up to 10 mL with DCM. H-Phe-pNA (160 mg,
1 equiv), and DIEA (145 mg, 1 equiv) was added and stirred for 1 h.
The solvent was removed and replaced by with a solution of DBU in
DCM (20%) to immediately turn to a deep yellow solution which was
stirred for 30 min. After removal of the solvent, the brown oil was
purified by preparative HPLC to give 219.3 mg (76%) of H-Val-Ser-
3
3
(dd, J ) 3.4, J ) 7.3, H-C(4)), 4.38-4.11 (m, 6 H, H-CFmoc, H2-
C
Fmoc, H-CRAla, H2-C(5)), 2.80 (s, 4 H, H2-CSu), 1.20 (d, J ) 7.0, H3-
CâAla). CI-MS (NH3): 508 (0.7, [M + 1]+), 392 (1, [M + HOSu]+),
279 (3), 178 (100). A 0.80 M solution of methylamine in ethanol (14.7
mL, 9.31 mmol) was added dropwise over a period of 45 min at 0 °C
to a solution of 4.50 g (8.87 mmol) of Fmoc-Ala-Ser(ΨH,HPro)-OSu
in 30 mL of chloroform. The mixture was stirred for another 20 min
at room temperature, diluted with 90 mL of chloroform and washed
with 60 mL of a 5% solution of sodium carbonate, 60 mL of a 0.1 M
of hydrochloric acid, and brine (3 × 60 mL). The combined organic
phases were dried over magnesium sulfate, and the solvent was
evaporated. The resulting material was purified by column chroma-
tography on silica gel (eluent EtOAc/AcOH, 50:1) to give 3.27 g (87%)
of pure Fmoc-Ala-Ser(ΨH,Hpro)-NHMe: C23H25N3O5 (423.47). Mp
122 °C. Rf (EtOAc/AcOH, 50:1) 0.21. tR ) 10.7 min (40-100% B,
30 min, C18). 1H-NMR (250 MHz, DMSO-d6, 59 mM, two conform-
ers: major (75%), minor (25%)): 7.89-7.28 (m, 10 H, 8 arom. H, 2
HN), 5.19/4.88 (br., AB, 2-H), 5.12/4.93 (AB, J ) 3.4, ∆ν ) 46, 2-H),
4.50-4.07 (m, 6 H, H-CFmoc, H2-CFmoc, H-C(4), H2-C(5)), 3.82 (quint.,
J ) 4.3, H-CRAla), 2.57 (d, J ) 4.6, H3-CNH), 1.20 (d, J ) 6.9, H3-
CâAla), 1.13 (br., H3-CâAla). CI-MS (NH3): 424 (3, [M + 1]+), 279
(3), 202 (2, [M - Fmoc]+), 178 (100). Purified Fmoc-Ala-Ser(ΨH,H
-
pro)-NHMe (2.09 g, 4.94 mmol) was dissolved in 50 mL of piperidine/
DMF, 1:4 (v/v), and stirred for 15 min at room temperature. After
evaporation of the solvent, the yellow residue was purified by column
chromatography on silica gel (eluent CHCl3/MeOH, 10:1, and CHCl3/
MeOH/triethylamine, 50:50:1) to yield 915 mg (92%) of crude H-Ala-
Ser(ΨH,Hpro)-NHMe. Quantitative acetylation was performed by
treatment with acetic anhydride/water (1:1, v/v, 12 mL) for 2 h at room
temperature. The solvent was evaporated and the crude product purified
by chromatography on silica gel (eluent CHCl3/MeOH, 10:1) and
crystallized from DCM/cyclohexane to afford 863 mg (78%) of
crystalline Ac-Ala-Ser(ΨH,Hpro)-NHMe. C10H17N3O4 (243.26). Mp
148-150 °C. Rf (CHCl3/MeOH, 10:1) 0.26. tR ) 17.8 min (0-15%
B, 30 min, C18). CI-MS (NH3): 244 (18, [M + 1]+), 185 (3, [M -
HNCOMe]+), 159 (8), 131 (42), 114 (51), 86 (100), 72 (32). Anal.
(C10H17N3O4) C, H, N. The carbon combustion analysis was found to
be higher than the expected value; this is supported by the X-ray data
which show that the above compound cocrystallized with molecules
of CH2Cl2 in the solid state.
Ac-Ala-Pro-NHMe (5). L-Proline (9.43 g, 81.9 mmol) was added
to 10 mL of an aqueous solution and the pH adjusted to 9 by adding
∼0.1 g of Na2CO3. A 1 M solution of Fmoc-Ala-Cl (4.50 g, 13.7
mmol) in acetone was added dropwise over a period of 90 min while
periodically adjusting the pH to 8-9 with sodium carbonate (∼2 g).
The reaction mixture was then cooled to 0 °C and 3 N hydrochloric
acid (30 mL) added to adjust the pH to 2. The aqueous suspension
was extracted with ethyl acetate (3 × 100 mL), and the organic phases
were combined, washed with brine (2 × 50 mL), and dried over
magnesium sulfate. Purification by column chromatography on silica
gel (eluent EtOAc/MeOH/AcOH, 100:2:2) resulted in 4.11 g (74%) of
pure Fmoc-Ala-Pro-OH. C23H24N2O5 (408.45). Mp 102-105 °C
(lyophilized material). Rf (CHCl3/MeOH/AcOH, 100:5:1) 0.29. tR )
11.3 min (40-100% B, 30 min, C18). 1H-NMR (250 MHz, DMSO-
d6, 65 mM): 7.89-7.28 (m, 8 arom. H), 7.61 (d, J ) 7.5, HN), 4.35-
4.16 (m, 5 H, H-CRAla, H-CRPro, H-CFmoc, H2-CFmoc), 3.62-3.46 (m, H2-
CδPro), 2.10 (m, H-CâPro), 1.94-1.78 (m, 3 H, H-CâPro, H2-CγPro), 1.19
(Ψ
Me,Mepro)-Phe-pNA (ESI-MS: 512 [M + 1]+). The free amine
containing product (88 mg), succinic anhydride (18.1 mg, 1.05 equiv)
and NMM (18 mg, 19 µL, 1 equiv) were stirred for 30 min to give
Suc-Val-Ser(ΨMe,Mepro)-Phe-pNA quantitatively. The product was
purified by preparative HPLC (20-100% B, 30 min) with 20% yield
(68 mg). Mp 186-189 °C. ESI-MS: 612.5 [M + 1]+. tR ) 11.71
min (40-100% B, 15 min, C18). Anal. (C30H37N5O9) C, H, N.

Pseudo-Prolines as a Molecular Hinge
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 925
Synthesis of Suc-Val-Ser(Ψ(2R)H,PMPpro)-Phe-pNA (9) and Suc-
Val-Ser(Ψ(2S)PMP,Hpro)-Phe-pNA (10). Fmoc-Val-Ser-OBzl (0.5 g,
0.968 mmol), PPTS (73 mg, 0.3 eq), and anisaldehyde dimethyl acetal
(880 mg, 5 equiv) in THF (dry, 30 mL) were heated under reflux for
2 h. The THF was evaporated and replaced with methanol (30 mL)
before a stream of hydrogen (H2) was bubbled through the solution in
the presence of Pd/C catalyst (10%) until full deprotection of the
carboxyl was reached (2 h). The black suspension was filtered on Celite
and the solvent evaporated (ESI-MS 515.3 [M + 1]+). The coupling
with phenylalanine-4-nitroanilide (H-Phe-pNA; 276 mg, 1.05 equiv)
was carried out in DCM (25 mL) according the mixed anhydride method
(-10 °C, 30 min) using tert-butylchloroformate (126 µL) in the
presence of N-methylmorpholine (NMM) (106 µL) to give Fmoc-Val-
Ser(ΨPMP,Hpro)-Phe-pNA (89%). Deprotection of the Fmoc protection
group was achieved without further workup using DBU (200 µL) in
DCM (10 mL) within 30 min. After evaporation of the solvent and
the DBU, the yellow oil was taken up in 5 mL of acetonitrile/water
(1:1) and purified by preparative HPLC(C18) to give the two epimers
of H-Val-Ser-(ΨPMP,Hpro)-Phe-pNA. Acetylation of the mixture with
succinic anhydride (100 mg, 1.1 equiv) in DCM (10 mL) in the presence
of NMM (35 µL, 1.1 equiv) gave the desired Suc-Val-Ser-(ΨPMP,H
-
pro)-Phe-pNA as an epimeric mixture. Purification by preparative
HPLC (20-100% B, 30 min) gave the minor (2R) (compound 9) and
the major (2S) (compound 10) epimers separately with 30% and 70%
yield respectively. Mp 214-216 °C. ESI-MS: 690.7 [M + 1]+. tR
) 12.77 min (compound 10) and tR ) 13.3 min (compound 9) (40-
100% B, 15 min, C18). Anal. (C35H39N5O10) C, H, N.
Acknowledgment. The present work was supported by the
Swiss National Science Foundation.
Supporting Information Available: 1H and 13C NMR tables
and Ramachandran maps of the pseudo-proline-containing
peptides (13 pages). See any current masthead page for ordering
and Internet access instructions.
JA962780A