Factors Affecting Conformation in Proline-Containing Peptides

Article abstract of DOI:10.1021/ol035711r

(Equation presented) NMR was used to study the thermodynamics of the cis → trans isomerization for prolyl amide bonds in the compounds shown. The magnitude of K_t/c for C-terminal esters is greater than for the corresponding amides, signifying s

Full text of DOI:10.1021/ol035711r

ORGANIC
LETTERS
2003
Vol. 5, No. 23
4413-4416
Factors Affecting Conformation in
Proline-Containing Peptides
,†
Carol M. Taylor,* Renaud Hardre´,† Patrick J. B. Edwards,† and Jae H. Park‡
Institute of Fundamental Sciences, Massey UniVersity, PriVate Bag 11-222,
Palmerston North, New Zealand, and Department of Chemistry,
UniVersity of Auckland, PriVate Bag 92019, Auckland, New Zealand
c.m.taylor@massey.ac.nz
Received September 7, 2003
ABSTRACT
NMR was used to study the thermodynamics of the cis f trans isomerization for prolyl amide bonds in the compounds shown. The magnitude
of K for C-terminal esters is greater than for the corresponding amides, signifying stronger backbone stereoelectronic effects in esters.
t/c
Increasing the steric bulk of the N-terminal residue from Ac- to Ac-Gly- favors the trans conformation. Incorporation of a Phe residue N-terminal
to Pro, however, shifts the equilibrium in favor of the cis conformation, via a stabilizing aromatic−proline interaction.
Amide bonds, including peptide bonds, have partial double
bond character, and this is often explained by invoking the
resonance contributors I and II, as illustrated for Ac-Pro-
OMe (1) (Figure 1a). In recent years, Wiberg has asked us
that the nitrogen atom inductively withdraws electron density
back through the σ-bond network and, in fact, bears a partial
negative charge.2 It is perhaps more useful to view the
planarity of the amide bond as a result of π-orbital overlap
(Figure 1b).
As a result of this π-orbital overlap, and concomitant
restricted rotation about the amide bond, there are two
conformations which correspond to energy minima attained
when the dihedral angle about the C(dO)-N bond (ω) is 0°
(cis) and 180° (trans) as illustrated in the graphical abstract.
For these two conformations to interconvert, the nitrogen
must become transiently sp3-hybridized, i.e., pyramidalized.3,4
Proline is arguably the most important amino acid, vis-
a`-vis the determination of protein structure and function.5
The cyclic nature of the side chain means that cis and trans
conformations are closer in energy for prolyl amides6 than
Figure 1. Electron distribution in prolylamides.
(2) Milner-White, E. J. Protein Sci. 1997, 6, 2477-2482.
(3) Fischer, S.; Dunbrack, R. L., Jr.; Karplus, M. J. Am. Chem. Soc.
1994, 116, 11931-11937.
(4) Kang, Y. K. THEOCHEM 2002, 585, 209-221.
(5) (a) Vanhoof, G.; Goossens, F.; De Meester, I.; Hendriks, D.; Scharpe´,
S. FASEB 1995, 9, 736-744. (b) Reiersen, H.; Rees, A. R. TIBS 2001, 26,
679-684.
(6) Lubell has defined the term prolyl amide as “a tertiary amide
composed of the pyrrolidine nitrogen of the prolyl residue and the carbonyl
of the N-terminal residue.” Halab, L.; Lubell, W. D. J. Am. Chem. Soc.
2002, 124, 2474-2484.
to consider structure III as well.1 While these representations
illustrate the distribution of π-electrons, they neglect the fact
† Massey University.
‡ University of Auckland.
(1) Wiberg, K. B.; Laidig, K. E. J. Am. Chem. Soc. 1987, 109, 5935-
5943.
10.1021/ol035711r CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/18/2003

for other peptide bonds.7 The population of molecules in the
cis conformation is significant.
each spectrum.11,12 The Van’t Hoff plots for compounds 1
and 2 are given in Figure 3. A slight positive gradient was
Herein, we describe the investigation of six proline deriv-
atives (Figure 2) to probe the relative importance of various
Figure 3. Van’t Hoff plots for Ac-Pro-OMe (1) and Ac-Pro-NHMe
(2) in D2O.
seen for Ac-Pro-OMe (1), as reported by others.9c,13 The trans
conformation is favored (Kt/c ) 5.2 at 298 K), and Kt/c
decreases with increasing temperature (Kt/c ) 3.9 at 358 K),
as more energy is available to populate the higher energy
cis species. The replacement of the methyl ester with a
methyl amide led to a significant reduction in the magnitude
of the equilibrium constant. This effect was observed
consistently in the other ester/amide pairs of compounds that
we studied (vide supra).
With the exception of theoretical studies,14,15 C-terminal
amides have rarely been used in conformational studies,16
to avoid complications arising from hydrogen bonding.
Evidence has been presented for hydrogen bonds in both cis17
and trans18 conformations (Figure 4a) in non-hydrogen-
bonding solvents.19 Such effects are less significant in an
aqueous environment and are unlikely to explain the differ-
ence in Kt/c between compounds 1 and 2.
Zimmerman and Scheraga proposed that restricted rotation
about the N-CR bond leads to a favorable CdO- - -CdO
electrostatic interaction.20 Maccallum et al. have demon-
strated that this is almost as strong (80%) as a typical
backbone hydrogen bond in a protein.21 This stabilizing
n f π* interaction involves electron donation from the
Figure 2. Proline derivatives.
factors in determining the position of the cis f trans
equilibrium.
We prepared N-acetyl-L-proline methyl ester (1) and the
corresponding methyl amide (2). These proline derivatives
have been widely investigated8 and serve as an important
baseline. We next wanted to consider the impact of replacing
the N-terminal acetyl group with an amino acid. Eberhardt
et al. have prepared the C-terminal methyl ester 3,9 and we
can compare their data with the C-terminal amide 4. Finally,
we will consider the impact of an aromatic amino acid in
the so-called (i - 1) position, by looking at dipeptides 5
and 6. It is also instructive to consider the tetrapeptide 7
which was studied by Wu and Raleigh.10
(11) Others have utilized integration of COOCH3 or CONHCH3 singlets,
or in the case of 5-tBu-Pro derivatives, the C(CH3)3 signal. Our experience
has been that resolution is not adequate to give good results.
(12) The trans/cis ratio in D2O has been shown to be independent of
concentration (ref 8).
(13) Renner, C.; Aldefelder, S.; Bae, J. H.; Budisa, N.; Huber, R.;
Moroder, L. Angew. Chem., Int. Ed. 2001, 40, 923-925.
(14) Kang, Y. K.; Jhon, J. S.; Han, S. J. J. Peptide Res. 1999, 53, 30-
40.
NMR spectra were recorded for 0.01-0.04 M solutions
of each compound in D2O over the temperature range
25-85 °C. The equilibrium constants were calculated by
integration of as many well-resolved signals as possible in
(15) Zimmerman, S. S.; Scheraga, H. A. Biopolymers 1977, 16, 811-
(7) There are also issues concerning the conformation of the pyrrolidine
ring, although this is beyond the scope of this discussion.
(8) For example, an extensive NMR analysis of Ac-Pro-NHMe: Hi-
gashijima, T.; Tasumi, M.; Miyazawa, T. Biopolymers 1977, 16, 1259-
1270.
843.
(16) For other studies on Ac-X-Pro-NHMe derivative, see: (a) ref 6.
(b) Halab, L.; Lubell, W. D. J. Org. Chem. 1999, 64, 3312-3321. (c) Halab,
L.; Lubell, W. D. J. Peptide Sci. 2001, 7, 92-104.
(17) Cox, C.; Lectka, T. J. Am. Chem. Soc. 1998, 120, 10660-10668.
(18) (a) Mizushima, S.; Shimanouchi, T.; Tsuboi, M.; Sugita, T.;
Kurosaki, K.; Mataga, N.; Souda, R. J. Am. Chem. Soc. 1952, 74, 4639-
4641. (b) Matsuzaki, T.; Iitaka, Y. Acta Crystallogr. 1971, B27, 507-516.
(19) Liang, G. B.; Rito, C. J.; Gellman, S. H. Biopolymers 1992, 32,
507-516.
(9) Specifically, the doubly 13C-labeled compound Ac-Gly-[â,δ-13C)-
Pro-OMe]: (a) Eberhardt, E. S.; Loh, S. N.; Hinck, A. P.; Raines, R. T. J.
Am. Chem. Soc. 1992, 114, 5437-5439. (b) Eberhardt, E. S.; Loh, S. N.;
Raines, R. T. Tetrahedron Lett. 1993, 34, 3055-3056. (c) Eberhardt, E.
S.; Panasik, N., Jr.; Raines, R. T. J. Am. Chem. Soc. 1996, 118, 12261-
12266.
(20) Zimmerman, S. S.; Scheraga, H. A. Macromolecules 1976, 9, 408-
416.
(10) Wu, W.-J.; Raleigh, D. P. Biopolymers 1998, 45, 381-394.
4414
Org. Lett., Vol. 5, No. 23, 2003

Figure 5. Steric interactions and Kt/c values at 298 K.
in favor of the trans conformation is a consequence of the
increase in steric bulk. The equilibrium constants are larger
for Ac-Gly-Pro-OMe (3) than for Ac-Gly-Pro-NHMe (4),
since the trans conformation is again stabilized to a greater
extent by the n f π* interaction in the ester.
Figure 4. Intramolecular forces: (a) hydrogen bonding and (b)
n f π* interaction (looking down the CR-N bond).
The Van’t Hoff plots for dipeptides 3,9 4, 5, 6, and
tetrapeptide 710 are presented in Figure 6. As observed for
oxygen lone pair of the (i - 1) amide CdO, to the
antibonding orbital of the CdO bond belonging to the Pro
(i) residue. Raines and co-workers have recently provided
evidence for the significance of this interaction, which they
describe as quantum mechanical rather than electrostatic, and
estimate a contribution of 0.7 kcal mol-1 to the stability of
the trans conformation of compound 2 at 298 K.22 Wiberg
et al. have shown that there is greater positive charge density
on the CdO carbon of esters than amides.23 We suggest
cautiously that a stronger n f π* interaction in the trans
conformation of esters (cf., amides) accounts for the differ-
ence in Kt/c values. Indeed, Raines and co-workers have
acknowledged that an amide carbon is less electron-deficient
than an ester carbon,22b thereby predicting the difference in
the strength of the n f π* interactions which we have ob-
served experimentally.
We next consider the dipeptides Ac-Gly-Pro-OMe (3) and
Ac-Gly-Pro-NHMe (4). A dipeptide represents the smallest
unit which enables us to look at a true peptide bond (i.e.,
between two amino acid residues) and is the next step up in
complexity from a proline derivative. Early studies of cis
f trans isomerism of X-Pro dipeptides were carried out with
free amino and carboxy termini. This work demonstrated that
pH, and therefore the degree of ionization, has a significant
impact on the conformation of the peptide bond.24 The
incorporation of amide blocking groups at the N- and
C-termini has therefore been employed to eliminate electro-
static interactions, making it possible to focus on more subtle
effects.
Figure 6. Van’t Hoff Plots for Ac-Gly-Pro-OMe (3),9 Ac-Gly-
Pro-NHMe (4), Ac-Phe-Pro-OMe (5), Ac-Phe-Pro-NHMe (6), and
Ac-Gly-Phe-Pro-Gly-NH2 (7)10 in D2O.
the single residue Pro derivatives (Figure 3), the dipeptides
containing glycine have a positive slope, with the two lines
running parallel. For the peptides containing a Phe-Pro amide
bond, especially dipeptide 6, there is a reduction in the
magnitude of Kt/c; there is also less temperature-dependence.
There is a high propensity for cis amide bonds when Pro
is preceded by an aromatic residue.25 Surveys of crystal-
lographic databases reveal that 5.7% of X-Pro peptide bonds
are in the cis conformation. When X is Phe this percentage
rises to 6.4% and leaps to 19.1% for Tyr;26 there is limited
data for Trp-Pro linkages.27 This is attributed to a stabilizing
Ar-Pro interaction in the cis conformation (Figure 7). Halab
The addition of a Gly residue N-terminal to the Pro residue
leads to an increase in Kt/c (Figure 5). This significant shift
(21) (a) Maccallum, P. H.; Poet, R.; Milner-White, E. J. J. Mol. Biol.
1995, 248, 361-373. (b) Maccallum, P. H.; Poet, R.; Milner-White, E. J.
J. Mol. Biol. 1995, 248, 374-384.
(25) (a) MacArthur, M. W.; Thornton, J. M. J. Mol. Biol. 1991, 218,
397-412. (b) Stewart, D. E.; Sarkar, A.; Wampler, J. E. J. Mol. Biol. 1990,
214, 253-260.
(26) For NMR studies of cis-trans isomerism in model peptides
containing Tyr-Pro linkages, see: (a) Stimson, E. R.; Montelione, G. T.;
Meinwald, Y. C.; Rudolph, R. K. E.; Scheraga, H. A. Biochemistry 1982,
21, 5252-5262. (b) Juy, M.; Lam-Thanh, H.; Lintner, K.; Fermandjian, S.
Int. J. Peptide Protein Res. 1983, 22, 437-449.
(22) (a) DeRider, M. L.; Wilkens, S. J.; Waddell, M. J.; Bretscher, L.
E.; Weinhold, F.; Raines, R. T.; Markley, J. L. J. Am. Chem.. Soc. 2002,
124, 2497-2505. (b) Hinderaker, M. P.; Raines, R. T. Protein Sci. 2003,
12, 1188-1194.
(23) Wiberg, K. B.; Hadad, C. M.; Rablen, P. R.; Cioslowski, J. J. Am.
Chem. Soc. 1992, 114, 8644-8654.
(24) (a) Grathwohl, C.; Wu¨therich, K. Biopolymers 1976, 15, 2025-
2041. (b) Grathwohl, C.; Wu¨therich, K. Biopolymers 1981, 20, 2623-2633.
(c) Mariappan, S. V. S.; Rabenstein, D. L. J. Org. Chem. 1992, 57, 6675-
6678.
(27) Poznanski et al. have gone so far as to say that nonbonding
interactions between pyrrolidine and indole rings govern the cis-trans
equilibrium in peptides containing Trp-Pro fragments: Poznanski, J.; Ejchart,
A.; Wierzchowski, K. L.; Ciurak, M. Biopolymers 1993, 33, 781-795.
Org. Lett., Vol. 5, No. 23, 2003
4415

while Phe-His-Thr-Phe-Pro was only 15% cis.30 The isomer-
ization of X-Pro bonds is normally enthalpy-driven.31
However, the Ar-Pro interaction stabilizes the cis conforma-
tion and brings it closer in energy to its trans counterpart so
that ∆H is small, accounting for the essentially flat line in
the Van’t Hoff plots for these compounds. Moreover, the
conversion of cis f trans must disrupt the “organized” cis
species and its inherent Ar-Pro interaction (Figure 7). This
is reflected in positive ∆S values for compounds 5-7.
In conclusion, we have identified significant differences
in the strength of backbone n f π* stabilizing effects
between esters and amides. This difference is tentatively
attributed to the difference in electron density at the acyl
carbon. So long as these differences are borne in mind, we
believe esters can still serve as useful model compounds.
The effect of noncovalent interactions on the relative
energies of cis and trans conformations is considerable. An
aromatic residue preceding the proline greatly stabilizes the
cis conformation via a hydrophobic interaction between the
aromatic and pyrrolidine rings. It is not clear what makes
the pyrrolidine ring different from other nonpolar side chains
(e.g., Val, Leu), other than its cyclic nature and the lack of
an amide proton.32 The impact of this Ar-Pro interaction is
tempered by steric considerations which cause the peptide
backbone to prefer an extended conformation (i.e., a trans
peptide bond).
Figure 7. Ar-Pro Interaction.
and Lubell have suggested6 that the Ar-Pro interaction is
of a cationic-pi nature, as defined by Dougherty.28 This seems
unlikely if we acknowledge the net negative charge on the
prolyl nitrogen discussed earlier.
Evidence for this nonbonding interaction is found in the
1H chemical shifts for HR of the Pro residue. In the cis
conformation, HR is shielded by the aromatic ring and a
dramatic upfield shift is observed. The data are summarized
for compounds 4, 6, and 7 in Table 1.
1
Table 1. Parameters Derived from H NMR (D2O, 298 K)
parameter
4
6
7
δHR (trans)
δHR (cis)
∆δR
4.23
4.41
+0.18
5.54
4.38
3.47
-0.91
2.56
4.42
3.82
-0.60
4.8
While the ultimate preferred conformation, or conforma-
tions, of proline-containing peptides remains the outcome
of the interplay of a number of considerations, we hope that
we have demonstrated how these work together.
Kt/c
Wu and Raleigh advocate ∆δR,29 the chemical shift
difference for Pro HR in the cis and trans conformations, as
a measure of the strength of the Ar-Pro interaction. The
low Kt/c value for dipeptide Ac-Phe-Pro-NHMe (6) reflects
the maximum impact of a preceding Phe residue. In the
tetrapeptide 7, described by Wu and Raleigh,10 the effect of
elongating the peptide chain is demonstrated. While the
Ar-Pro interaction persists, the steric demands of the extended
peptide backbone counteract it. A similar example was
reported by Wu¨therich and Grathwohl: in DMSO-d6, Thr-
Phe-Pro was 60% cis with respect to the X-Pro peptide bond,
Acknowledgment. This research was supported by the
Marsden Fund, administered by the Royal Society of New
Zealand (Grant No. 00-MAU-007). Jodie M. Johnston made
early contributions to the synthesis of some dipeptides on a
Health Research Council Summer Studentship (1997-98).
Supporting Information Available: Experimental pro-
1
cedures and H and 13C NMR spectra for compounds 1, 2,
4-6, and intermediates in their synthesis. Tables of ther-
modynamic data derived from NMR experiments. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(28) (a) Dougherty, D. A. Science 1996, 271, 163-168. (b) Gallivan, J.
P.; Dougherty, D. A. J. Am. Chem. Soc. 2000, 122, 870-874.
(29) Wu and Raleigh actually use the term “∆∆R”; however, we believe
that the second character should be a lower case δ since it refers to chemical
shift.
OL035711R
(30) Wu¨therich, K.; Grathwohl, C. FEBS Lett. 1974, 43, 337-340.
(31) Kang, Y. K.; Park, H. S. THEOCHEM 2002, 593, 55-65.
(32) Kersteen, E. A.; Raines, R. T. Biopolymers 2001, 59, 24-28.
4416
Org. Lett., Vol. 5, No. 23, 2003