as a Probe of Slow
of Short Linear Peptides
Anastasia S. Politou2
1Department of Chemistry,
University of Ioannina,
45110 Ioannina, Greece
University of Ioannina,
45110 Ioannina, Greece
3Department of Pharmacy,
University of Athens,
15771 Athens, Greece
Received 27 March 2002;
accepted 21 October 2002
Abstract: According to general belief, the conformational information on short linear peptides in
solution derived at ambient temperature from NMR spectrometry represents a population-weighted
average over all members of an ensemble of rapidly interconverting conformations. Usually the
search for discrete conformations is concentrated at low temperatures especially when sharp NMR
resonances are detected at room temperature. Using the peptide Ac–RGD–NH2(Ac–Arg–Gly–Asp–
NH2, Ac: acetyl) as a model system and following a new approach, we have been able to
demonstrate that short linear peptides can adopt discrete conformational states in DMSO-d6
(DMSO: dimethylsulfoxide) which vary in a way critically dependent on the reconstitution condi-
tions used before their dissolution in DMSO-d6. The conformers are stabilized by intramolecular
hydrogen bonds, which persist at high temperatures and undergo a very slow exchange with their
extended structures in the NMR chemical shift time scale. The reported findings provide clear
evidence for the occurrence of solvent-induced conformational exchange and point to DMSO as a
valuable medium for folding studies of short linear peptides.
Biopolymers 69: 72–86, 2003
© 2003 Wiley Periodicals, Inc.
Keywords: arginine ionic interactions; aspartic acid ionic interactions; fibrinogen inhibitor;
hydrogen bonds; NMR; peptide folding; RGD peptide; slow conformational exchange
Correspondence to: Vassilios Tsikaris; email: btsikari@
Contract grant sponsor: Greek Secretariat of Research and
Biopolymers, Vol. 69, 72–86 (2003)
© 2003 Wiley Periodicals, Inc.
Information on the favored conformational states of
short linear peptides comes mainly from x-ray diffrac-
tion, NMR studies in different solvents, and molecular
modeling calculations.1–3For instance, five different
crystal structures for Leu-enkephalin and three for
Met-enkephalin, depending on solvent crystallization
conditions, have been reported.4,5By taking advan-
tage of the solubility of peptides in a number of polar
solvents, various solvent-dependent conformational
states can be resolved. It is generally accepted that
these states are found in very fast conformational
exchange. The structures of peptides, especially those
derived from NMR, are believed to represent only
population-weighted averages over all conformers in
a given solvent at ambient temperature. Only slow
transitions in the NMR time scale conformational,
such as the amide bond cis–trans interconversion,
have been resolved to date for linear peptides by
NMR spectroscopy. To study very rapid processes,
such as the folding of peptides and proteins, new
methods had to be developed (stopped-flow fluores-
cence, stopped-flow CD, temperature jump, etc.),
which allow the earliest events in folding to be
probed.6,7Using temperature jump and photodissocia-
tion techniques, Eaton et al.8have shown that some
?-helices are formed in a few nanoseconds, whereas
others require microseconds to fold, depending on the
particular amino acid sequence. To date, intermediate
conformational states between folded and unfolded
states in peptides have not been detected.
Some recent studies have indicated that short linear
peptides can adopt different conformational states in
dimethylsulfoxide (DMSO) solutions depending on
the pH value of the aqueous solution they originated
from.9–12We have also provided experimental evi-
dence by17O-NMR spectroscopy for slow conforma-
tional exchange of Boc–[17O]Tyr(2,6-diClBzl)–OH
(Boc: tert-butoxycarbonyl; Bzl: benzyl) in DMSO
solution.13Our conclusion was based on the detection
of two, rather than one,17O resonances for the car-
boxyl group of the peptide in DMSO, most probably
due to the engagement of the carboxyl group in a
strong hydrogen-bonding interaction. In CDCl3solu-
tion, on the other hand, a single17O resonance, re-
sulting from the fast exchange between open and
hydrogen bonded states, was observed. The depen-
dence of the peptide conformation on the nature of the
reconstitution medium has also been highlighted by
the recent work of Boden et al., who studied the
three-dimensional (3D) structure of a linear 27-resi-
due peptide in lipid bilayers by Fourier transform
infrared (FTIR) absorption.14When that peptide was
reconstituted from methanol, it adopted a ?-strand
structure, while in the case of 2,2,2-trifluoroethanol it
formed initially an ?-helix, which relaxed very slowly
(within hours) to an equilibrium state between ?-helix
It appears, therefore, that the nature of the solvent
and the conditions employed in the conformational
reconstitution might influence the prevalence of a
certain peptide conformation.
The work reported here aims to develop a NMR-
based strategy that would allow us to identify directly
discrete conformational states of short linear peptides,
differentiate them from the average one, and gain new
insight into their folding process. The tripeptide Ac–
RGD–NH2(Ac–Arg–Gly–Asp–NH2, Ac: acetyl) was
used as a model compound for our NMR and molec-
ular modeling studies.15–18Solutions of this peptide in
DMSO-d6, reconstituted from aqueous solutions at
different pH values, were studied by1H-NMR spec-
troscopy at several temperatures (in the range 300–
355 K). We managed to detect discrete conforma-
tional states of the peptide in DMSO-d6, which vary
with the initial pH of the solution, and to show that
these states can be in slow exchange depending on the
reconstitution conditions. We present here our find-
ings and discuss their implications for peptide folding.
MATERIALS AND METHODS
Reagents and solvents were used without further purifica-
(TBTU), 1-hydroxybenzotriazole (HOBt), and Boc amino
acids were purchased from Neosystem (France); solvents
from Labscan Ltd. (Ireland); trifluoroacetic acid (TFA) and
diisopropylethylamine (DIEA) from Merk Schuchardt (Ger-
many); and 4-methyl-benzhydrylamine resin (MBHA) resin
from Saxon Biochemicals (Hannover, Germany). DMSO-d6
and tetramethylsilane (TMS) were purchased from Euriso-
Synthesis of Ac–RGD–NH2
This was carried out by the stepwise solid-phase procedure
on a MBHA resin following the Boc chemistry.19Arg was
introduced as Boc–Arg(Tos)–OH (Tos: toluene-4-sulfonyl)
and Asp as Boc–Asp(OBzl)–OH. Coupling reactions were
performed using the molar ratio of amino acid/TBTU/
HOBt/DIEA/resin 3/2.9/3/3/1. The ?NH2 group, after
cleavage of the Boc protecting group with TFA, was acety-
lated using an excess of Ac2O in pyridine (the ratio
Ac2O/—NH2group was 30:1). Ac–RGD–NH2was cleaved
from the resin with anhydrous HF in the presence of phenol
and anisole as scavengers. The crude material (yield 80%)
was subjected to high performance liquid chromatography
(HPLC) purification (semipreparative reverse-phase C18col-
umn) using gradient elution with the following solvents: A,
H2O/0.1% TFA; B, CH3CN/H2O/TFA (10/90/0.1). A pro-
grammed gradient elution (4 mL/min) was applied (A/B: 90/
10–A/B: 75/25), elution time 20 min (yield 60%). The purity
of the peptide was checked by analytical HPLC and the correct
spectroscopy (ESI-MS) (MW calc.: 387.40; found: 387.62).
Synthesis of Ac–RGd–NH2
This was carried out on a MBHA resin according to Boc
chemistry as described above. The purity of the peptide was
checked by analytical HPLC and the correct molecular mass
was confirmed by ESI-MS.
The NMR samples were prepared by dissolving the solid
material in H2O, and adjusting the pH to the desired value
with NaOH or HCl. The aqueous solutions were lyophi-
lized, and weighted amounts were dissolved in DMSO-d6at
concentrations ?5 mM. The NMR experiments were per-
formed at 295–355 K on Bruker AMX 400 and Avance 500
(COSY), total COSY (TOCSY), and rotating frame nuclear
Overhauser effect spectroscopy (ROESY) Bruker micropro-
grams were used. The TOCSY spectra were recorded using
a mixing time of 100 ms. The spectral width in F1 was 5600
Hz. Various ROESY experiments were performed using
mixing times of 250 and 350 ms at 300, 305, and 310 K.
The rate constants of rotation about the guanidinium N?–C?
(k?), C?–N?1(k?1), and C?–N?2(k?2) bonds were obtained by
calculating the NMR line shape for a four-site exchange pro-
cess using the general multiple site exchange matrix algo-
rithm.20The best fitted simulated spectra were obtained by
using the spectral parameters (chemical shifts, line widths, and
intensities) for guanidinium protons and varying rate constants
k?, k?1, and k?2. The natural line widths of guanidinium NH
signals required in these determinations were estimated by
measuring the line widths of nonexchanging Arg amide proton
signals at appropriate temperatures. All calculations were per-
formed using the program Muses (MUltiple Site Exchange
Simulations).21The activation parameters were evaluated from
ln?k/T? ? 23.76 ? ??H*/RT? ? ??S*/R?
?G* ? RT?23.76 ? ln(k/T?]
where R is the universal gas constant.
Structure calculation was carried out using the software DY-
ANA (DYnamics Algorithm NMR Applications).22The dis-
tance restraints used as inputs in DYANA were derived from
reconstituted in DMSO-d6after lyophilization from an aque-
ous solution at pH 4.9. The ROESY spectrum was recorded at
converted into distances using the ?1-?2 cross peak of Asp as
reference. Thirty-six upper-limit cross-peak intensities, classi-
fied as strong (up to 2.8 Å), medium (up to 3.5 Å), and weak
(up to 5 Å), were used as input restraints for the calculation.
Appropriate corrections for center averaging were added to
DYANA restraints for degenerate proton resonances.23No
lower limits were used. Constraints for ?, ?, and ?1angles
were calculated using the HABAS program of DYANA pack-
age. Six3J??2and3J?coupling constants were included with
a tolerance of 2.5 Hz. All distance and angle constraints were
assigned the default relative weight of 1. The default tolerance
of 0.05 at target function units was applied. The calculations
were performed using the standard minimization protocol and
the REDAC (REdundant Dihedral Angle Constraints) strategy
implemented in DYANA.
1H-1H ROESY spectrum of the Ac–RGD–NH2peptide
RESULTS AND DISCUSSION
Conformational State of Ac–Arg–Gly–
Asp–NH2, Reconstituted in DMSO-d6
from an Aqueous Solution at pH 2.0
The complete assignment of all proton resonances of
Ac–Arg–Gly–Asp–NH2was based on the combined
use of COSY, TOCSY, and ROESY experiments. The
1H-NMR spectrum in DMSO-d6solution of the pep-
tide reconstituted after lyophilization from an aqueous
solution at pH 2 is shown in Figure 1 (bottom). The
resonance at 12.3 ppm (not shown) confirms the pro-
tonated state of the Asp ?-carboxylic group. The high
absolute temperature coefficient values of all of the
NH protons, including the C-terminal amide protons
(?8 ppb/K), suggest that they are exposed to the
solvent. The equal3JN?and3JN??values (5.32 Hz) of
Gly indicate that there is a free rotation about the
NOC?bond of this residue,24while the Asp3J??and
3J???coupling constant values (5.27 and 8.46 Hz)
correspond to a high percentage (?80%, Table I) of
the two energetically favored C?–C? rotamers25–28(I
and II, ??
rotational restrictions about the C?OC?bond. On the
other hand, the Arg–N?H and Arg–N2
symbology indicates both N?atoms and all four pro-
tons linked to them) proton resonances, at 7.63 and
7.13 ppm, respectively, are attributed to their nonhy-
drogen-bonded states.10,11,29It must be noted that the
can be detected as two broad peaks due to chemical
exchange in the guanidinium group.10,11,30,31In this
conformational state two separate, broad peaks at 7.29
1?60°, 180°), suggesting the absence of
?H4protons under free rotational conditions
Biris et al.
displacement (RMSD) values. A classification of the
experimental restraints and a statistical analysis of the
best 10 of the calculated structures are given in Table II.
The elements of intramolecular structuring indi-
cated by the conformational analysis presented in
detail above and summarized in the model depicted in
Figure 6 are confirmed by the calculated structure
(Figure 9). The existence of a compact “pseudocy-
clic” structure stabilized by hydrogen bonding in a
head-to-tail fashion, also supported by previous work
with peptides having an Arg residue in the third
position from the C-terminus,9–12is consistent with
the pattern of hydrogen bonds resulting from the
calculations. More specifically: (a) there seems to be
a bifurcated hydrogen bond between the Asp
?-COO?and the Arg–N?and one of the Arg–N?1
protons; (b) the conformation might be even more
stabilized through an additional interaction of the
guanidinium group with the terminal carbonyl group
of Asp; (c) the trans proton of the C-terminal NH2
group is pointing to the interior of the pseudocyclic
structure in more than half of the best 10 structures,
being thus shielded from the solvent, as also indicated
by its temperature coefficient values; (d) the compact-
ness of the peptide conformation is confirmed by the
short distances in all calculated structures between the
Asp–NH proton and the Arg–N?(?3.2 Å) and Arg–
C?2H (?2.7 Å) protons, due to the restricted mobility
of these side chains.
In this study, focused on the Ac–RGD–NH2peptide
as a model system, we provided experimental evi-
dence for some important general aspects related to
the conformational states of short linear peptides in
solution. The novel information deduced from this
work can be summarized as follows:
The discrete conformational states of the peptides
in DMSO-d6can undergo very slow exchange. Thus,
the species normally observed by NMR might repre-
sent only local conformational changes, which are fast
in the NMR time scale, rather than an overall average
The approach based on the rational selection of the
conditions employed before the conformational re-
constitution of a peptide in DMSO allows a more
accurate characterization of its folded structure and of
the folding process, since the driving interactions and
forces become in this manner more discernible and
can be quite easily evaluated. Moreover, the hypoth-
esis that short linear peptides in solution are found in
fast conformational exchange in the NMR time scale
of the REDAC strategy of DYANA. Thin lines represent hydrogen bonds.
Best structure (the one with the lowest value of target function) calculated with use
is not always valid, and has to be explored and veri-
fied in each separate case.
Intramolecular hydrogen-bonding interactions can
persist and stabilize the conformation of short linear
peptides even at very high temperatures.
The folded structure in DMSO-d6, of Ac–RGD–
NH2, is characterized by an ionic bridge between Arg
guanidinium and Asp ?-carboxylate groups. Two hy-
drogen bonds involving the Arg–N?H and Arg–N?H
protons stabilize the interaction between the oppo-
sitely charged groups.
This work was supported by a grant from the Greek Secre-
tariat of Research and Technology. The authors thank V. I.
Polshakov for providing the program Muses.
1. Eggleston, D. S.; Feldman, S. H. Int J Pept Protein Res
1990, 36, 161–166.
2. Kang, Y. K.; Jhon, J. S. J Pept Res 2000, 56, 360–372.
3. Bogdanowich,-Knipp, S. J.; Jois, S. D.; Siahaan, T. J. J
Pept Res 1999, 53, 523–529.
4. Aubry, A.; Birlirakis, N.; Sakarellos-Daitsiotis, M.;
Sakarellos, C.; Marraud, M. Biopolymers 1989, 28,
5. Aubry, A.; Birlirakis, N.; Sakarellos-Daitsiotis, M.;
Sakarellos, C.; Marraud, M. J Chem Soc, Chem Com-
mun 1988, 963–964.
6. Brockwell, D. J.; Smith, D. A.; Radford, S. E. Curr
Opin Struct Biol 2000, 10, 16–25.
7. Volk, M. Eur J Org Chem 2001, 2605–2621.
8. Eaton, W. A.; Munoz, V.; Thompson, P. A.; Henry,
E. R.; Hofrichter, J. Acc Chem Res 1998, 31, 745–753.
9. Sakarellos-Daitsiotis, M.; Panou-Pomonis, E.; Sakarel-
los, C.; Cung, M. T.; Marraud, M.; Tzinia, A. K.;
Soteriadou, K.; Tsikaris, V. Lett Pept Sci 1996, 3,
10. Tsikaris, V.; Sakarellos-Daitsiotis, M.; Panou-Pomonis,
E.; Detsikas, E.; Sakarellos, C.; Cung, M. T.; Marraud,
M. Pept Res 1992, 5, 110–114.
11. Tsikaris, V.; Cung, M. T.; Panou-Pomonis, E.; Sakarel-
los, C.; Sakarellos-Daitsiotis, M. J Chem Soc, Perkin
Trans 2 1993, 1345–1349.
12. Tsikaris, V.; Sakarellos-Daitsiotis, M.; Tzovaras, D.;
Sakarellos, C.; Orlewski, P.; Cung, M. T.; Marraud, M.
Biopolymers 1996, 38, 673–682.
13. Tsikaris, V.; Moussis, V.; Sakarellos-Daitsiotis, M.;
Sakarellos C. Tetrahedron Lett 2000, 41, 8651–8654.
14. Boden, N.; Cheng, Y.; Knowles, P. F. Biophys Chem
1997, 65, 205–210.
15. Pierschbachter, M. D.; Ruoslahti, E. Nature 1984, 308,
16. Suzuki, S.; Oldberg, A.; Hayman, E. G.; Piersch-
bachter, M. D.; Ruoslahti, E. EMBO J 1985, 4, 2519–
17. Watt, K. W. K.; Cottrall, B. A.; Strong, D. D.; Doolitle,
R. F. Biochemistry 1979, 18, 5410–5416.
18. Stavrakoudis, A.; Bizos, G.; Elefteriadis, D.; Kouki, A.;
Panou-Pomonis, E., Sakarellos-Daitsiotis, M.; Sakarel-
los, C.; Tsoukatos, D.; Tsikaris, V. Biopolymers 2001,
19. Stewart, J. M.; Young, J. D. Solid Phase Peptide Syn-
thesis; Pierce Chemical Company: Rockford, IL, 1984.
20. Sandstro ¨m, J. Dynamic NMR Spectroscopy; Academic
Press: London, 1982.
21. Polshakov, V. I.; Birdsal, B.; Feeney, J. Biochemistry
1999, 38, 15962–15969.
22. Guntert, P.; Mumenthaler, C.; Wu ¨thrich, K. J Mol Biol
273 1997, 283–298.
23. Wu ¨thrich, K.; Billeter, M.; Braun, W. J Mol Biol 1983,
24. Tsikaris, V.; Detsikas, E.; Sakarellos-Daitsiotis, M.;
Sakarellos, C.; Vatzaki, E.; Tzartos, S. J.; Marraud, M.;
Cung, M. T. Biopolymers 1993, 33, 1123–1134.
25. Benedetti, E.; Morelli, G.; Nemethy, G.; Scheraga,
H. A. Int J Pept Protein Res 1983, 22, 1–15.
26. De Leeuw, F. A. A. M.; Altona, C. Int J Pept Protein
Res 1982, 20, 120–125.
27. Cung, M. T.; Marraud, M. Biopolymers 1982, 21, 953–
28. McGregor, M. J.; Islam, S. A.; Sternberg, M. J. J Mol
Biol 1987, 198, 295–310.
29. Mayer, R.; Lancelot, G. J Am Chem Soc 1981, 103,
30. Smith, R. J.; Williams, D. H.; James, K. J Chem Soc,
Chem Commun 1989, 682–683.
31. Kanamori, K.; Roberts, J. D. J Am Chem Soc 1983,105,
32. Sanderson, P. N.; Glen, R. C.; Payne, A. W. R.; Hud-
son, B. D.; Heide, C.; Tranter, G. E.; Doyle, P. M.;
Harris, C. J. Int J Pept Protein Res 1994, 43, 588–596.
33. Bogusky, M. J.; Naylor, A. M.; Pitzenberger, S. M.;
Nutt, R. F.; Brady, S. F., Colton, C. D.; Sisko, J. T.;
Anderson, P. S.; Veber, D. F. Int J Pept Protein Res
1992, 39, 63–76.
34. Gargaro, A. R.; Frenkiel, T. A.; Nieto, P. M.; Birdsall,
B.; Polshakov, V. I.; Morgan, W. D.; Feeney, J. Eur
J Biochem 1996, 238, 435–439.
35. Nieto, P. M.; Birdsall, B.; Morgan, W. D.; Frenkiel,
T. A.; Gargaro, A. R.; Feeney, J. FEBS Lett 1997, 405,
36. Mitchell, J. B. O.; Thornton, J. M.; Singh, J.; Price,
S. L. J Mol Biol 1992, 226, 251–262.
Biris et al.