Three types of induced tryptophan optical activity compared in model dipeptides: Theory and experiment

Article abstract of DOI:10.1002/cphc.201200201

The tryptophan (Trp) aromatic residue in chiral matrices often exhibits a large optical activity and thus provides valuable structural information. However, it can also obscure spectral contributions from other peptide parts. To better understand the indu

Full text of DOI:10.1002/cphc.201200201

DOI: 10.1002/cphc.201200201
Three Types of Induced Tryptophan Optical Activity
Compared in Model Dipeptides: Theory and Experiment
[a]
[a]
[a]
[a]
Jana Hudecovꢀ,*[a, b] Jan Hornꢁcek, Milos Budesꢁnsky, Jaroslav Sebestꢁk, Martin Safarꢁk,
ˇ
ˇ
ˇ
ˇ
ˇˇ
ˇ
´
[c]
[c]
[a]
ˇ
Ge Zhang, Timothy A. Keiderling, and Petr Bour*
The tryptophan (Trp) aromatic residue in chiral matrices often
exhibits a large optical activity and thus provides valuable
structural information. However, it can also obscure spectral
contributions from other peptide parts. To better understand
the induced chirality, electronic circular dichroism (ECD), vibra-
tional circular dichroism (VCD), and Raman optical activity
(ROA) spectra of Trp-containing cyclic dipeptides c-(Trp-X)
(where X=Gly, Ala, Trp, Leu, nLeu, and Pro) are analyzed on
the basis of experimental spectra and density functional theory
(DFT) computations. The results provide valuable insight into
the molecular conformational and spectroscopic behavior of
Trp. Whereas the ECD is dominated by Trp p–p* transitions,
VCD is dominated by the amide modes, well separated from
minor Trp contributions. The ROA signal is the most complex.
However, an ROA marker band at 1554 cmꢀ1 indicates the local
c2 angle value in this residue, in accordance with previous the-
oretical predictions. The spectra and computations also indi-
cate that the peptide ring is nonplanar, with a shallow poten-
tial so that the nonplanarity is primarily induced by the side
chains. Dispersion-corrected DFT calculations provide better re-
sults than plain DFT, but comparison with experiment suggests
that they overestimate the stability of the folded conformers.
Molecular dynamics simulations and NMR results also confirm
a limited accuracy of the dispersion-DFT model in nonaqueous
solvents. Combination of chiral spectroscopies with theoretical
analysis thus significantly enhances the information that can
be obtained from the induced chirality of the Trp aromatic resi-
due.
1. Introduction
The tryptophan (Trp) residue plays an important role in pep-
tide conformational studies, especially those using chiral and
fluorescence spectroscopic techniques.[1–6] The aromatic chro-
mophore has an easily detectable spectral response. This is
both convenient and a problem, as electronic circular dichro-
ism (ECD) arising from Trp can interfere with that due to the
peptide backbone and bias secondary structure analyses, while
its fluorescence overlaps that of other aromatics. Coupling be-
tween Trp and other aromatic residues leads to a particularly
large ECD that can be used for tertiary structure analyses.[7–11]
Trp also has distinctive vibrational spectral properties.[12,13] Its
bands can be selectively enhanced in Raman spectra[2] and
used for surface-enhanced studies.[14] Although the aromatic
side chain in Trp is planar, strong Raman optical activity (ROA)
features have been identified in some peptides and confirmed
theoretically as being due to the chiral orientation of the adja-
cent covalent link.[15]
tion. The cyclic dipeptides (sometimes called 2,5-diketopipera-
zines, DKPs, after the central six-membered ring) are favored
for model studies[19–21] as they are more rigid than linear pep-
tides and are reasonably small, thereby facilitating the use of
more accurate computations.
In particular, we were interested in the interaction of the Trp
residue with the backbone and other side-chain parts of pep-
tides, as such effects are crucial for developing ROA and ECD
spectroscopic responses. As shown below, they provide com-
plementary, but not identical, information about the structure
to that obtained with NMR spectroscopy, which can be often
reconciled only with complex theoretical modeling of molecu-
ˇ
ˇ
ˇ
ˇˇ
´
ˇ
[a] J. Hudecovꢀ, J. Hornꢁcek, Dr. M. Budesꢁnsky, Dr. J. Sebestꢁk, M. Safarꢁk,
ˇ
Dr. P. Bour
Institute of Organic Chemistry and Biochemistry
Academy of Sciences
ˇ
Flemingovo nꢀmestꢁ 2, 16610 Prague 6 (Czech Republic)
Fax: (+420)220183123
E-mail: hudecova@uochb.cas.cz
bour@uochb.cas.cz
The Trp residue has many biological functions, including par-
ticipation in ion transfer[16] and providing a signal or anchor for
pores formed from transmembrane helices, which often termi-
nate in Trp. Quinacrine drugs were suggested to interact with
tryptophans.[17] Often, short secondary structures containing
these residues are stabilized by a hydrophobic collapse.[18]
In this work, a series of small Trp-containing cyclic dipepti-
des was synthesized and subjected to spectroscopic and com-
putational studies, in order to understand the side-chain role
in the ECD, vibrational circular dichroism (VCD), and ROA spec-
tra, and to monitor the Trp conformational properties in solu-
[b] J. Hudecovꢀ
Institute of Physics, Faculty of Mathematics and Physics
Charles University
Ke Karlovu 5, 12116 Prague 2 (Czech Republic)
[c] G. Zhang, Dr. T. A. Keiderling
Department of Chemistry
University of Illinois at Chicago
845 W. Taylor St., Chicago, IL 60607-7061 (USA)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cphc.201200201.
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

ˇ
J. Hudecovꢀ, P. Bour et al.
tides c-(l-Trp-l-X) were obtained from Z-l-Trp-X-OMe, whereas c-
(d-Trp-d-X) compounds were prepared from Fmoc-d-Trp-d-X-OMe,
since different blocking groups were available for l- and d-Trp. The
peptides were purified by column chromatography (Merck silica
gel 60, CHCl3/MeOH 25:1 or 10:1). Cyclization of the l series was
achieved by 5 h of hydrogenolysis[58] on Pd sponge in MeOH with
continual heating to 508C for 15–40 h. The Pd sponge was re-
moved by filtration and washed with MeOH, acetonitrile (AcCN) or
DMSO depending on the product solubility. The solvent was
evaporated, and the product was crystallized several times from
MeOH.
lar behavior. For symmetric molecules, such as the cyclic di-
peptide c-(l-Trp-l-Trp), NMR spectroscopy cannot discriminate
among the conformational variants of individual side chains,
thus making the optical techniques, in particular ROA, in prin-
ciple more suitable in this case.
However, the ROA spectral features are not necessarily local,
and the side-chain contributions are mostly overlapped with
the backbone signal.[22,23] Only some ROA bands can be associ-
ated with local molecular parts. Fortunately, for Trp, such
a band exists at 1554 cmꢀ1, and its conformational depend-
ence on the c2 side-chain dihedral angle can be reliably inter-
preted with computations.[15] For ECD, the Trp–Trp exciton cou-
pling is generally believed to indicate a fixed and close mutual
position. This was only partially confirmed by our modeling, as
a large signal was observed in mono-Trp peptides (e.g. c-(Trp-
Ala)) as well.
The cyclic d series peptides were obtained in two steps. First, the
Fmoc group was removed using 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU)/EtSH/AcCN (2:20:78) mixture[59] for 1 h, which was followed
by extraction of the amine form between 1% HCl and diethyl
ether. The aqueous solutions were adjusted to pH 8 with saturated
NaHCO3. Free amine was taken to EtOAc. After drying with Na2SO4
and evaporation of the solvent, the residue was heated in MeOH
for 24–72 h. The precipitated cyclic form was filtered off and recrys-
tallized from MeOH. In cases of low conversion, the remaining
amino component was removed with Dowex-50 in MeOH.
Peptide structural studies through chiral spectroscopies
profit from the improved reliability and performance of the
computational tools. The ECD spectra, for example, can be
nowadays routinely simulated using time-dependent density
functional theory (TDDFT)[24–26] for fairly large molecules, in-
cluding the aromatic residues in peptides.[10,11]
Because of limited solubilities of the various cyclic dipeptides, a rel-
atively wide range of solvents had to be used for the experimental
NMR spectroscopy, ECD, VCD, and ROA, as summarized in Table 1.
Also, information about local molecular structure in the vi-
brational optical activity spectra can be fully grasped only
when supported by the simulations.[27,28] For VCD theory, the
most important milestones were perhaps the development of
the magnetic field perturbation theory of Stephens,[29,30] and
its implementation within the gauge-independent atomic orbi-
tals (GIAOs)[31] and the DFT methodology.[32]
Similarly to VCD, the GIAOs should be used for ROA.[33] Al-
though implemented within DFT,[34] ROA calculations have for
a long time been hampered by the necessity to compute de-
rivatives of some tensors numerically. This obstacle was lifted
only recently by implementation of fully analytical schemes in
publicly available programs.[35–40]
Table 1. Overview of solvents used for the experimental spectra.
Compound
NMR
ECD
–
IR/VCD
DMSO
Raman/
ROA
c-(l-Trp-Gly)
[D6]DMSO, CD3OD,
DMSO
CDCl3
c-(d-Trp-Gly)
c-(l-Trp-l-Ala)
–
–
–
DMSO
–
DMSO
–
TFE,
AcCN
AcCN
c-(l-Trp-l-Trp)
c-(d-Trp-d-Trp)
[D6]DMSO, CD3OD
–
DMSO,
AcCN
DMSO,
AcCN
–
DMSO,
CH3OH
DMSO
AcCN
Our study also confirms the advantage of analysis of data
from a combination of several spectroscopic techniques to
characterize molecular behavior.[41,42] Especially, the vibrational
methods provide new insight into details of peptide secondary
structure.[43–48] However, the spectral interpretations and com-
putations are still challenging, in particular for the proper rep-
resentation of solvent involvement, conformational averaging,
and balance of dispersion force.[49–51]
c-(l-Trp-l-Leu)
c-(l-Trp-l-nLeu)
[D6]DMSO, CD3OD
–
–
CH3OH
–
TFE,
AcCN
AcCN
–
c-(l-Trp-l-Pro)
c-(d-Trp-d-Pro)
[D6]DMSO, CD3OD
–
DMSO,
AcCN
DMSO,
AcCN
–
AcCN
DMSO
As indicated in previous studies the dispersion should be
added to most DFT biomolecular studies.[52–56] This is true also
for the Trp-containing dipeptides, although the benefit of the
correction within a simplified solvent model may be limited.
NMR Spectroscopy: 1H and 13C NMR spectra of c-(l-Trp-Gly), c-(l-
Trp-l-Trp), c-(l-Trp-l-Leu), and c-(l-Trp-l-Pro) were measured on
a Bruker Avance II NMR spectrometer (1H at 600.13 and 13C at
150.9 MHz) equipped with a 5 mm cryo-probe. The spectra of all
cyclic dipeptides were measured at 300 K in DMSO, CD3OD, and
CDCl3. For structural assignment of proton and carbon signals
(using natural 13C occurrence), a combination of homo- and hetero-
nuclear 2D NMR spectra (H,H-COSY, H,C-HSQC and H,C-HMBC) was
used. The NOE contacts were determined from 2D H,H-ROESY
spectra (mixing time 300 ms).
Experimental Section
Synthesis: The synthesis started from N-protected (benzyloxycar-
bonyl, Z, and fluorenylmethoxycarbonyl, Fmoc) and C-protected
(hydrochlorides of corresponding methyl esters) amino acids ob-
tained from Merck, Czech Republic. Linear dipeptide precursors
were prepared from Z-l-Trp-OH or Fmoc-d-Trp-OH and HCl·H-X-
OMe peptides (X=Gly, Ala, Trp, Leu, nLeu, and Pro) using standard
BOP activation[57] with 3 equiv diisopropylethylamine. Cyclic dipep-
ECD Spectra: Electronic CD spectra for c-(l-Trp-l-Ala), c-(l-Trp-l-
Trp), c-(d-Trp-d-Trp), c-(l-Trp-l-nLeu), c-(l-Trp-l-Pro), and c-(d-Trp-
d-Pro) samples were measured using a Jasco J-810 spectropolarim-
eter. Samples were studied in 0.1 cm path length quartz cells,
&2
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!

Induced Tryptophan Optical Activity in Dipeptides
using concentrations of about 0.2 mgmLꢀ1 in AcCN or 2,2,2-tri-
fluoroethanol (TFE). c-(Trp-Gly) was not sufficiently soluble under
these conditions. Each spectrum was obtained as an average of six
scans taken with a band pass of 1 nm and scanning speed of
urable; the spectra are then presented as the difference (lꢀd)/2,
as for VCD.
DFT Geometries: Lowest-energy conformations of selected dipepti-
des (Tables 2 and 3) were obtained by geometry optimization with
the Gaussian program,[63] mostly using the B3LYP[64] functionals and
6-311++G** basis set. For some tests, the aug-cc-pVTZ standard
basis sets, MP2,[65] MPW2PLYP,[54,66] B3PW91,[67] and BPW91[68] meth-
ods were used as specified below (Table S1, Supporting Informa-
tion). The conductor-like polarizable continuum model (CPCM)[69]
was used with AcCN (relative permittivity, er =36), CHCl3 (er =4.7),
TFE (er =27), CH3OH (er =33), and DMSO (er =47) parameters for
computations on the c-(l-Trp-l-Trp) molecule. The solvent varia-
tions, however, had only a minor effect on the resultant spectra
and relative conformer energies, as documented in Figure S1 in
the Supporting Information. For calculations of other molecules
50 nmminꢀ1
. The experimental spectra are expressed in De
(Lmolꢀ1 cmꢀ1); for this purpose the concentration was determined
by the Trp UV absorption.
VCD Spectra: VCD spectra of the d and l forms of c-(Trp-Gly), c-
(Trp-Trp), and c-(Trp-Pro) were measured using a homemade dis-
persive instrument separately described in detail.[60] The corre-
sponding IR spectra were recorded on the same samples using
a Vertex 80 FTIR (Bruker) spectrometer.
Samples were prepared by dissolving the peptides in DMSO or
AcCN (not shown), to a concentration of about 10 mgmLꢀ1, and
placing the solutions in a sealed cell composed of two CaF2 win-
dows separated by
a 100 mm
spacer. Spectra were obtained as
the average of six scans and were
corrected by subtraction of an
identically obtained spectrum of
the solvent. Most VCD signals were
in general quite weak and compa-
rable in intensity to the instrumen-
tal artifact signals developed with
this high-refractive-index solvent.
The VCD spectra are therefore pre-
sented as the difference of enan-
tiomers, or (lꢀd)/2. Similarly, IR
spectra are presented as their sum,
or (l+d)/2.
Table 2. Selected geometry parameters,[a] relative energies,[b] and Boltzmann weights[c] of the most populated
conformers,[d] calculated at the B3LYP/6-311++G**/CPCM(DMSO) level.
Conformation[e]
c1
c2
f
y
DE
h
c-(Trp-Gly)
A
A’
B
D
ꢀ61
ꢀ60
ꢀ59
62
103
106
ꢀ87
ꢀ92
89
23
ꢀ25
22
28
25
ꢀ16
17
ꢀ16
ꢀ19
ꢀ17
0
35
32
11
11
8
0.1
0.7
0.7
0.9
C
63
c-(Trp-Ala)
A
A’
B
B’
D
C
ꢀ60
ꢀ60
ꢀ59
ꢀ57
63
103
106
ꢀ87
ꢀ83
ꢀ89
90
18
ꢀ28
18
ꢀ25
17
14
ꢀ12
19
ꢀ13
15
ꢀ12
ꢀ9
0
39
32
11
6
6
6
0.1
0.7
1.1
1.1
1.1
ROA and Raman Spectra: All cyclic
dipeptides were dissolved in
DMSO or methanol to concentra-
tions of 50–150 mgmLꢀ1, and the
spectra were measured with
a backscattering SCP BioTools m-
63
c-(Trp-Leu)
A3
Trp
Leu
ꢀ61
ꢀ63
ꢀ58
ꢀ63
ꢀ61
ꢀ167
62
ꢀ62
64
ꢀ63
103
172
ꢀ88
172
103
64
ꢀ89
174
90
15
16
18
18
20
27
16
14
13
16
ꢀ11
ꢀ11
ꢀ13
ꢀ14
ꢀ15
ꢀ21
ꢀ11
10
0
45
13
9
B3
A8
D3
C3
0.7
0.9
1.1
1.2
Chiral
Raman-2X
instrument
equipped with an Opus diode-
pumped solid-state laser operating
at 532 nm.[61,62] The laser power
was set to 50–100 mW, and power
at the sample was 30–60 mW.
Higher powers would cause
a faster degradation of the sam-
ples. Residual fluorescence was
quenched by leaving the sample
in the laser beam for a few hours
before measurement. The total ac-
quisition time was about 20 h for
each sample. For most samples,
ROA spectra from two or three in-
dependent measurements were
averaged. Solvent bands were sub-
tracted from the Raman spectra,
and minor baseline corrections
were made. In the case of strong
solvent scattering, affected wave-
number regions were deleted from
the spectra. In this work we ana-
lyzed only the ROA of c-(Trp-Trp)
and c-(Trp-Gly), where both enan-
tiomers were available and meas-
7
ꢀ9
6
173
ꢀ11
c-(Trp-Trp)
AA
Trp 1
Trp 2
ꢀ61
ꢀ61
ꢀ59
ꢀ60
ꢀ61
ꢀ58
ꢀ60
ꢀ58
ꢀ60
62
102
102
105
105
103
ꢀ88
105
ꢀ84
103
ꢀ90
16
16
ꢀ19
ꢀ19
16
ꢀ10
ꢀ10
11
0
25
22
15
10
11
A’A’
AB
0.1
0.7
1
12
ꢀ11
ꢀ10
6
15
A’B’
AD
ꢀ12
ꢀ10
16
5
ꢀ10
ꢀ11
0.9
17
c-(Trp-Pro)[f]
A’ S
B’ S
A’ N
B’ S
(P, Vm)
(302.4, 38.2)
(302.7, 38.3)
(75.9, 36.1)
(300.3, 38.1)
(303.9, 38.2)
ꢀ60
ꢀ58
ꢀ59
ꢀ63
54
108
ꢀ82
108
ꢀ22
80
ꢀ37
ꢀ36
ꢀ41
ꢀ39
ꢀ35
31
31
36
33
30
0
1
1.1
1.5
1.9
64
12
10
5
C’ S
3
[a] Torsion angles c1, c2, f, and y, in degrees, defined in Figure 1, for l,l enantiomers for DFT structures. [b] DE,
in kcalmolꢀ1. [c] Including degeneracy; in %. [d] Conformers with relative energy DE<2 kcalmolꢀ1 are specified.
[e] Notation of peptide conformation is specified in Figure 2. [f] Type of proline puckering: S or N defined by
phase P and amplitude Vm.
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
3
&
ÞÞ
These are not the final page numbers!

ˇ
J. Hudecovꢀ, P. Bour et al.
nates were allowed to relax. This
revealed four minima; the angle c1
favors values around ꢀ608 and
608, and c2 prefers ꢀ908 and 908.
Table 3. Selected geometry parameters, relative energies, and Boltzmann weights of the most populated con-
formers, calculated at the B3LYP-D/6-311++G**/CPCM(DMSO) level.[a]
Conformation
c1
c2
f
y
DE
h
c-(Trp-Gly)
D
C
c-(Trp-Ala)
D
C
Based on these scans, c-(l-Trp-Gly),
c-(l-Trp-l-Ala), c-(l-Trp-l-Leu), c-(l-
Trp-l-Trp), and c-(l-Trp-l-Pro) di-
peptide geometries were generat-
ed with starting values c1 of about
ꢀ60, 60, and 1808, c2 ꢁꢀ90 and
908 (these conformations are de-
picted in Figure 2) together with
two possible ring conformations,
fꢁꢀ258 and 258, and optimized
by energy minimization (with
B3LYP or B3LYP-D and CPCM-
(DMSO)/6–311++G**). For c-(l-
Trp-l-Trp), c-(l-Trp-l-Leu), and c-(l-
Trp-l-Pro), the conformation of the
second side chain was also investi-
gated systematically; for Leu, the
c1’ and c2’ torsion angles were set
at ꢀ608, 608 or 1808 for beginning
the optimization; similarly, we
60
61
ꢀ90
37
33
ꢀ24
ꢀ23
0
0.7
76
23
87
62
63
ꢀ90
24
21
ꢀ15
ꢀ11
0
0.1
55
43
94
c-(Trp-Leu)
D3
Trp
Leu
64
ꢀ50
65
ꢀ55
61
ꢀ176
64
ꢀ93
ꢀ175
91
178
98
54
ꢀ93
64
22
13
18
20
20
26
20
20
ꢀ18
ꢀ9
0
64
17
5
C3
C8
D2
ꢀ11
ꢀ13
ꢀ9
0.8
1.5
1.6
ꢀ15
ꢀ15
ꢀ15
4
ꢀ74
c-(Trp-Trp)
AD
ꢀ56
64
ꢀ64
66
ꢀ59
72
67
64
62
56
ꢀ57
64
103
ꢀ89
ꢀ90
ꢀ81
ꢀ96
102
92
17
22
10
23
11
13
ꢀ18
ꢀ13
ꢀ2
ꢀ16
19
ꢀ8
ꢀ13
ꢀ1
ꢀ12
ꢀ7
ꢀ4
10
6
1
12
0
66
19
4
BD
0.8
1.7
1.3
1.4
1.8
BC
C’C’
D’D’
AC
4
used the
S and N conforma-
83
tions[71–74] of the proline five-mem-
bered ring.
ꢀ83
ꢀ96
103
93
3
ꢀ10
ꢀ10
3
Generation of the Spectra: Har-
monic IR, VCD, Raman, and ROA
intensities were computed with
the Gaussian programs at the
same level of theory as for opti-
mized structures. The B3LYP func-
tional with a medium-sized basis
19
c-(Trp-Pro)
D
C
(P, Vm)
(4.0, 39.7)
(326.8, 41.9)
62
65
ꢀ91
26
21
ꢀ18
ꢀ12
0
1.4
86
8
88
[a] Symbols same as in Table 2.
only the DMSO er value (or CHCl3 value for checking of conformer
preference) was used with both the normal and dispersion-correct-
ed (B3LYP-D)[54,66,70] B3LYP functional.
Table 4. R on cyclic dipeptides.
R
peptide
A two-dimensional scan was performed for c-(l-Ala-l-Ala) to inves-
tigate the inner ring potential energy surface (PES). Torsion angles
f and y (Figure 1, Table 4) were scanned in the range from ꢀ508
ꢀH
c-(L-Trp-Gly)
c-(L-Trp-L-Ala)
c-(L-Trp-L-Leu)
c-(L-Trp-L-nLeu)
ꢀCH3
ꢀCH2CH(CH3)2
ꢀ(CH2)3CH3
c-(L-Trp-L-Trp)
set has been found very convenient for analogous spectral simula-
tions in many previous studies.[75–78] An excitation frequency of
532 nm was used for backscattered Raman and ROA dynamic (fre-
quency-dependent) polarizabilities. Relative Raman and ROA spec-
tral S(w) shapes were obtained by convoluting the calculated in-
tensities (I) with Lorentzian bands D=8 cmꢀ1 wide, and multiply-
ing by a temperature-correction factor for T=298 K, so that the
spectrum from each mode i can be represented as [Eq. (1)]:
Figure 1. Characteristic coordinates of the c-(l-Trp-X), X=Gly, l-Ala, l-Leu, l-
nLeu, l-Trp, and c-(l-Trp-l-Pro) cyclic dipeptides.
to +508 with a step of 58 at the B3LYP/CPCM(DMSO)/6-311++G**
level. Because of the relatively simple single-valley PES that result-
ed from the 2D scan, for the other dipeptides we performed a re-
laxed 1D scan along the f angle only.
h
ꢀ
ꢁiꢀ1 h ꢀ
ꢁ
iꢀ1
2
wi
1
wi
w ꢀ wi
For c-(l-Trp-l-Trp) at the B3LYP/6-311G** level, the Trp side-chain
conformation was investigated by a scan along torsion angles c1
and c2, from ꢀ1808 to +1658 with 158 increments; other coordi-
ð1Þ
SðwÞ ¼ I 1 ꢀ exp ꢀ
4
þ1
kT
D
&4
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!

Induced Tryptophan Optical Activity in Dipeptides
2. Results and Discussion
Dipeptide Ring Geometry
It is known that the cyclic dipeptidic unit is normally quite flex-
ible, and its conformation is very dependent on the amino
acid side-chain type and interactions.[83] For example, c-(l-Ala-
l-Ala) in a crystal is puckered, whereas c-(l-Ala-d-Ala) is nearly
flat.[84] The peptide ring in the c-(l-Trp-l-Trp) crystal structure is
also flat.[85] This is in agreement with the computed one- and
two-dimensional PESs for c-(l-Ala-l-Ala), displayed in Figure 3.
Figure 2. Conformational classes (A–F) of the Trp-containing dipeptides (see
also Table 2 for conformer energies). The canonical angles (c1, c2, Figure 1)
are A(ꢀ60, 908), B(ꢀ60, ꢀ908), C(60, 908), D(60, ꢀ908), E(180, 908), and F(180,
ꢀ908).
where k is the Boltzmann constant and wi the vibrational frequen-
cy. Similarly, for VCD (S(w)=De) and IR spectroscopy (S(w)=e)
[Eq. (2)]:
h ꢀ
ꢁ
iꢀ1
2
2wi
Dp
w ꢀ wi
ð2Þ
SðwÞ ¼ cJ
4
þ1
D
where S(w) is in units of molꢀ1 Lcmꢀ1, J is the rotational/dipole
strength in debyes2, and c=108 for IR and 435 for VCD. Contribu-
tions of individual conformers were averaged using Boltzmann
weights based on the sum of electronic and zero-point energies
(ZPEs).
ECD spectra were generated from the dipole and rotational
strengths calculated with Gaussian programs using the TDDFT
method (B3LYP or B3LYP-D/CPCM(DMSO)/6-311++G**), and con-
voluted with Gaussian bands 15 nm wide. Surprisingly, the CAM-
B3LYP[79] functional, sometimes claimed to be superior to B3LYP for
ECD,[80] performed much worse for our system, and was thus not
used.
Figure 3. Top: calculated 2D (f, y, B3LYP/CPCM(DMSO)/6-311++G**) PES
for the inner ring in c-(l-Ala-l-Ala). Bottom: the corresponding 1D energy
scan along the torsion angle f [B3LYP/CPCM(DMSO)].
Molecular Dynamics: An alternate solvent model was explored for
c-(l-Trp-l-Trp) and c-(l-Trp-l-Pro), by running molecular dynamics
(MD) simulations with the Amber10 program package[81] and using
the Amber99 force field. Missing force-field parameters in the
cyclic dipeptide ring were derived from a Hartree–Fock (HF)/6-
31G* calculation with Gaussian, using the “POP=MK” keyword.
DMSO and chloroform force fields were obtained from the extend-
ed Amber database (http://www.pharmacy.manchester.ac.uk/bryce/
amber). The solute molecule was surrounded by solvent molecules
(DMSO, methanol, chloroform, or water) up to a distance of 8–
14 ꢃ. A four-step equilibration[82] was carried out, followed by
a 50 ns (100 ns for c-(l-Trp-l-Trp) in water) production run, using
NpT ensembles and 1 fs integration time. Snapshots were taken
each 5 ps and divided into three groups for c-(l-Trp-l-Trp) (folded:
CC, CD, and DD; partially folded: AC, AD, BC, BD, CE, CF, DE, and
DF; and extended: AA, AB, AE, AF, BB, BE, BF, EE, EF, and FF), and
into two groups for c-(l-Trp-l-Pro) (folded: C, D; and extended: A,
B, E, F; see Figure 2), according to the values of the c1 and c2 Trp
angles. The relative conformer populations were normalized to
100%. From three independent MD runs for c-(l-Trp-l-Trp) the
population error thus obtained was estimated as 20%.
Clearly, the potential is shallow, and at 300 K (kTꢁ0.6 kcal
molꢀ1) the ring is quite flexible, thus allowing for large devia-
tions of the y and f angles. The other principal angles in the
dipeptide ring (y’ and f’) adopt similar values to y and f
during the ring deformation. Minor variations were caused by
the 6-311++G** and aug-cc-pVTZ basis set change (Figure 3,
bottom). The flat ring-deformation potential is undoubtedly
caused by a partial conjugation of the amide-bond p-electron-
ic systems and strongly anharmonic out-of-plane amide defor-
mation energy, as observed, for example, for the N-methylace-
tamide (NMA) molecule.[86]
Other peptides (c-(Gly-Gly), c-(l-Ala-l-Ala), c-(l-Ala-Gly), and
c-(l-Leu-Gly)) showed a similar dependence (other 1D surfaces
are shown in Figure S2, Supporting Information), although
a closer look reveals finer differences between the dipeptides.
Unlike c-(l-Ala-l-Ala), c-(Gly-Gly) exhibits two minima with the
same energy, which reflects the symmetry of the molecule. The
l-Leu enforces a stronger preference for the global minimum.
The relative energies are notably changed by switching on the
dispersion correction (bottom of Figure S2); if included, for ex-
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
5
&
ÞÞ
These are not the final page numbers!

ˇ
J. Hudecovꢀ, P. Bour et al.
ample, the difference in the two minima of c-(l-Ala-l-Ala) in-
creases by ꢁ0.4 kcalmolꢀ1 and the dipeptide ring becomes
more twisted.
NMR Results
NMR spectra allow for the use of more variable experimental
conditions, in particular different kinds of solvents. In most
cases NMR spectroscopy can monitor the conformation of
a cyclic dipeptide, and verify theoretical structural predic-
tions.[88,89] The analysis of the J(NH, aH) vicinal couplings and
the resultant f angles in [D6]DMSO are summarized in Table 5.
For c-(l-Trp-Gly), the relative energy differences between the
two dipeptide ring conformers (fꢁꢀ40 and 408) also strongly
depend on the conformation of the side chains, as document-
ed in Figure S3 in the Supporting Information. For some con-
formers (C and D, Figure 2), the fꢁꢀ408 minimum is missing;
for A and B this minimum is very shallow.
Table 5. Experimental J(NH, aH) coupling constants [Hz] in DMSO, the
inner-ring f angle [8] in DMSO, MD-averaged f values (for c-(l-Trp-l-Trp))
in DMSO, and Boltzmann-averaged f values from DFT and DFT-D compu-
tations (B3LYP/CPCM(DMSO)/6-311++G**).
Conformations of the Dipeptides
c-(l-Trp-Gly)
Trp Gly
c-(l-Trp-l-Leu)
c-(l-Trp-l-Pro) c-(l-Trp-l-Trp)
In Tables 2 and 3 we list the lowest-energy conformers (DE<
2 kcalmolꢀ1) for the c-(l-Trp-Gly), c-(l-Trp-l-Ala), c-(l-Trp-l-Leu),
c-(l-Trp-l-Trp), and c-(l-Trp-l-Pro) dipeptides as obtained by
a systematic conformer search. The B3LYP/CPCM(DMSO)/6-
311++G** approximation level was used with (DFT-D) and
without (DFT) correcting for the dispersion interactions.
Trp
Leu
Trp
Trp
J
2.6 2.9; 0.9 2.6
2.9
12
–
ꢁ0.8
ꢀ37
ꢀ14
ꢀ37
24
2.9
12
11
5
[a]
fexp
fMD
fDFT
9
–
7
12
–
5
9
–
16
20
18
16
fDFT-D 36
30
16
The conformers can be approximately categorized according
to the values of c1 and c2 as “extended” (marked as A, B, E, F),
where the Trp indole side ring points out from the dipeptide
ring, or “folded” (marked as C, D), where the Trp indole is
above/below the ring, potentially stabilized by interaction with
the other amino acid side chain (Figure 2).
[a] Obtained from the coupling according to ref. [100], with coefficients
A=7.0, B=ꢀ1.1, C=0.55.
Table 5 also includes averaged f values from the MD run fMD
(for c-(l-Trp-l-Trp) and c-(l-Trp-l-Pro)) and Boltzmann-weighted
DFT and DFT-D (B3LYP/CPCM(DMSO)/6-311++G**) results.
Except for c-(l-Trp-l-Pro), NMR data indicate a very flattened
boat form of the dipeptide ring (f=9–128). This agrees better
with the uncorrected DFT and MD values than with DFT-D, but
for c-(l-Trp-l-Trp), the experimental values lie between the DFT
and DFT-D results. For c-(l-Trp-l-Pro) an opposite pucker (f=
ꢀ378) was determined by NMR spectroscopy than for the
other dipeptides, in agreement with a previous observation for
a similar c-(l-Phe-l-Pro) compound (ꢀ498).[90] This value is also
nicely reproduced by DFT (f=ꢀ378), but again not as well by
DFT-D (Boltzmann average, f=248).
Rather contradictory conformational analysis results are ob-
tained with the DFT and DFT-D approaches (Tables 2 and 3),
using the CPCM solvent model with DMSO parameters. When
DMSO was replaced by chloroform in the model, only minor
changes in conformer populations appeared (mostly less than
ꢂ5%). Typically, the DFT method alone predicts that the ex-
tended conformers are most stable, with a minor but not neg-
ligible population (ꢁ20%) of the folded ones. DFT-D almost
exclusively favors the folded structures, separated from the ex-
tended ones by a wide energy margin. Interestingly, even the
c-(l-Trp-Gly) is predicted to be entirely folded by DFT-D, al-
though Gly does not possess any significant polarizable com-
ponent beyond the DKP ring. The results are nevertheless con-
sistent with previous studies, which clearly document the large
effect of including the dispersion correction, and the signifi-
cant energy changes computed with this force.[53,54,66,70,87]
For c-(l-Trp-Gly), Boltzmann populations obtained at other
approximation levels (including B3LYP, BPW91, B3PW91, MP2,
MPW2PLYP) are summarized in Table S1 in the Supporting In-
formation. The uncorrected DFT results (B3LYP, BPW91,
B3PW91) are very similar, and favor all the extended conform-
ers A and A’. Likewise, the dispersion correction always
switches the equilibrium to the folded structures C and D. The
MP2 theory provides almost the same conformer distribution
as the DFT-D methods. The MPW2PLYP results are also very
similar to those from DFT-D; yet we see that some population
of the A, A’, B, and B’ conformers (in total 20%) is allowed by
the MPW2PLYP method, unlike for MP2 and DFT-D (<1%). This
reflects the well-known fact that the plain MP2 correction
tends to overestimate the dispersion correction if compared to
HF or older DFT formulations.[53,54,66]
The amino acid side-chain conformation can be deduced
b
b
from the J(aH, H) and NOE (NH, H) values. The coupling con-
stants and resultant approximate populations of the c1 rotam-
ers are listed in Table 6; the solvents included [D6]DMSO,
CD3OD, and CDCl3 (see also Table 1). Note that two b-protons
(referred to as R and S) provide individual NMR signals. The
NMR data thus indicate a significant preference for the folded
rotamer (where fꢁ608) in c-(l-Trp-Gly) and c-(l-Trp-l-Leu). In
the case of c-(l-Trp-l-Leu) this leads to strong shielding of the
Leu aH (3.39 and 3.57 ppm) and bH protons (0.62 and
ꢀ0.05 ppm in DMSO, and 0.66 and ꢀ0.20 ppm in CD3OD) due
to a ring current effect of Trp. Folded conformers also prevail
in c-(l-Trp-l-Pro), but only at about half the population, 46–
52%; in CDCl3 extended conformers are strongly preferred.
These facts are somewhat inconsistent with the dipeptide ring
analysis, in that they agree more with the DFT-D results than
with DFT. Nevertheless, they can be explained by the overesti-
mation of the effects of dispersion in the DFT-D method. In
particular, the Trp residue is mostly folded, as predicted by
DFT-D, but it does not deform the inner ring so much. This is
&6
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!

Induced Tryptophan Optical Activity in Dipeptides
b
Table 6. Experimental J(aH, H) coupling constants [Hz] and c1-rotamer populations h [%].[a]
c-(l-Trp-Gly)
Trp
c-(l-Trp-l-Leu)
Trp
c-(L-Trp-L-Pro)
Trp
c-(L-Trp-L-Trp)
Trp
Leu
bH
J, DMSO
J, CD3OD
J, CD3Cl3
R
S
4.6
4.6
–
R
4
3.6
3.9
S
4.8
4.6
8
R
9.4
10
S
4.7
4.3
3.7
R
S
5.8
ꢁ5
10.9
R
S
6.6
7.4
8
4.7
3.8
–
4.7
ꢁ5
3.8
4.3
3.9
3.7
10.2
AB
ꢀ60
21
11
–
EF
CD
60
62
73
–
AB
ꢀ60
10
5
11
EF
180
23
21
60
CD
60
67
74
29
AB
ꢀ60
20
23
10
EF
180
34
25
90
CD
60
46
52
0
AB
ꢀ60
15
11
9
EF
180
43
53
60
CD
60
42
36
31
c1 [8]
h, DMSO [%]
h, CD3OD [%]
h, CDCl3 [%]
180
17
16
–
ꢀ60
75
83
180
17
12
6
60
8
5
84
10
b
[a] According to ref. [101], J(aH, bH)=5.86–1.86cos(t)+3.81cos(2t)+0.37sin(t), where t=a(aH, C, C, H).
also consistent with the prevailing conformation in the c-(l-
Trp-l-Trp) crystal, for example, where a T-shaped folded confor-
mation was found (indicated as AD in Tables 2 and 3), but the
inner ring is almost planar.[85] The NMR data for the Trp confor-
mation in c-(l-Trp-l-Trp) are not usable for reliable prediction;
because of the symmetry, the Trp residues are not resolved by
NMR spectroscopy and are subject to fast conformer ex-
change.
Unlike with DFT-D/CPCM (Table 3), for c-(l-Trp-l-Trp) the ex-
tended structures may be additionally present in MD. In the
low-polarity CHCl3 environment the extended forms prevail
(90%). On the other hand, more polar solvents (DMSO, H2O,
MeOH) strongly favor more compact forms, in agreement with
the hydrophobic collapse known for some Trp-containing pep-
tides. In water (first line in Table 7) even the fully stacked paral-
lel l-Trp-l-Trp folded conformers appear (7%), but the edge-
on-face l-Trp-l-Trp interaction is still more probable (92%), as
commonly found in the Trp-containing peptides and pro-
teins.[18,91–94]
The c2 Trp angle can in principle be derived from the ob-
served NOE values. However, our measured data indicate
strong NOE contacts of Trp R-bH with both HN protons, thus
g
implying a fast flipping around the bCꢀ C bond, which pre-
The MD-predicted preferences of the Trp side-chain position
for c-(l-Trp-l-Pro) are in agreement with the NMR results
(Table 6), also favoring extended conformers in CHCl3. For c-(l-
Trp-l-Trp) the results cannot be compared to NMR data be-
cause of molecular symmetry. The behavior in CH3OH is very
similar to that in water according to MD, that is, the ratio of
extended and folded conformers is close to 50:50. In DMSO
the folded conformers slightly prevail in MD (73:27); in NMR
the ratio was 46:54. Thus, we see that the MD simulations
reveal finer solvent effects than DFT-D/CPCM, which had
almost the same conformer ratios for all solvents, although the
MD results are limited by the force field inaccuracy.
vents a reliable prediction.
Molecular Dynamics
The MD simulations with explicit solvent provide an alternative
to the DFT-D/CPCM model. In spite of the large error of the
populations (ꢁ20%, see the Experimental Section), the MD re-
sults (Table 7) clearly indicate a more complicated picture.
Table 7. Conformer populations [%] for c-(l-Trp-l-Trp) and c-(l-Trp-l-Pro)
obtained by MD/Amber10 simulations in different solvents.
ECD Spectra
c-(l-Trp-l-Trp)
Population[a]
extended
partially folded
folded
The experimental spectra of c-(l-Trp-l-Ala), c-(l-Trp-l-Trp), c-(l-
Trp-l-nLeu), and c-(l-Trp-l-Pro) in AcCN are plotted in Figure 4.
Due to interfering absorbance of DMSO in the UV region, it
was necessary to employ different solvents for ECD than for
the vibrational spectra (VCD and ROA). The ECD spectra for
these molecules have the unique characteristic that they are
dominated almost completely by the Trp contributions, even
for DKPs with only one Trp residue. The differences between
the ECD in AcCN and TFE (mostly ꢁ5 nm shift, but preserving
the general shape, Figure S4, Supporting Information) were rel-
atively minor, and less than the differences between the vari-
ous molecules. Such solvent effects would be difficult to
model.[95]
water
DMSO
CH3OH
CHCl3
1
45
7
92
55
93
10
7
0
0
0
90
c-(l-Trp-l-Pro)
Population
extended conformation
folded
c1
ꢀ608
1808
608
C
30
45
23
11
A
14
8
B
21
8
15
45
E
4
8
1
0
F
2
3
1
0
D
28
28
34
3
water
DMSO
CH3OH
CHCl3
26
41
[a] Folded conformers: CC, CD, and DD; partially folded: AC, AD, BC, BD,
CE, CF, DE, and DF; and extended conformers: AA, AB, AE, AF, BB, BE, BF,
EE, EF, and FF (see Figure 2).
The c-(l-Trp-l-Pro) is the outlier, which corresponds to a dis-
tortion of the peptide ring and consequently the interaction of
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
7
&
ÞÞ
These are not the final page numbers!

ˇ
J. Hudecovꢀ, P. Bour et al.
spectra (Figure 5). Individual conformer ECD (e.g. conformer C
of c-(l-Trp-l-Ala) calculated with and without the dispersion,
see Figure 5) are quite similar; thus, the resultant spectrum is
mostly influenced by the weighting scheme dependent on the
relative conformer energies (Tables 2 and 3).
For all molecules, it is clear that the conformer averaging is
needed to obtain realistic spectral shapes and absolute intensi-
ties. It is also apparent that the balance of populated conform-
ers changes the predicted ECD band shape by shifting the
spectral band overlap, thus making the relative energetics
more critical than the spectral prediction for each conformer.
Figure 4. Experimental ECD spectra of c-(l-Trp-l-Ala), c-(l-Trp-l-Trp), c-(l-Trp-
l-nLeu), and c-(l-Trp-l-Pro), all measured in AcCN. For analogous TFE results,
see Figure S4 in the Supporting Information.
IR and VCD Spectra
The experimental VCD and IR spectra of c-(l-Trp-Gly), c-(l-Trp-
l-Trp), and c-(l-Trp-l-Pro) in DMSO are compared in Figure 6.
We also measured spectra in other solvents, but in AcCN the c-
(Trp-Trp) and c-(Trp-Pro) developed added bands which may
be indicative of aggregation at IR concentrations. Such data
are not presented. The amide I (C=O stretching, 1600–
1700 cmꢀ1) IR spectrum has a relatively sharp band at 1670–
80 cmꢀ1 for c-(l-Trp-Gly) and c-(l-Trp-l-Trp), but the c-(l-Trp-l-
Pro) exhibits a broadening due to the Trp–Pro link being a terti-
ary amide with a lower amide I frequency.[19,89] The main VCD
signal is very weak and predominantly negative in the amide I
region. It has some contributions from other underlying
modes, which arise from the aromatic Trp side chain. The c-(l-
Trp-l-Trp) molecule might have a positive couplet shape (+/ꢀ,
from lower to higher frequency), but this is not clear in the ex-
periment due to a baseline distortion.
the Trp with the rest of the molecule caused by the constraints
due to the Pro pyrrole ring. The near identical ECD for c(l-Trp-
l-Ala) and c(l-Trp-l-Leu) suggests that the Trp is not interact-
ing significantly with the aliphatic side chain, at least in these
solvents. The much larger ECD seen for c-(l-Trp-l-Trp) in AcCN
than for the other peptides (Figure 4) suggests a strong exci-
ton coupling of the Trp residues,[11] and thus indicates a signifi-
cant contribution from a stable interacting or folded conforma-
tion. This is further confirmed by the negative ECD at 290 nm,
which is not seen for the other dipeptides.
The main features of the ECD spectra are reproduced for c-
(l-Trp-l-Ala), c-(l-Trp-l-Trp), and c-(l-Trp-l-Pro) by the B3LYP/
CPCM(DMSO)/6-311++G** computations (Figure 5). The aver-
aging of contributions from thermally populated conformers
had a minor influence on the total absorption, but had a major
impact on the resultant averaged ECD, and consequently the
DFT and DFT-D methods clearly provide very different ECD
The c-(l-Trp-l-Pro) VCD is surprisingly weak, considering the
expected distortion of its peptide ring, with the amide I VCD
being above but near to our measurement limits. (Most spec-
Figure 5. Calculated (B3LYP/CPCM(DMSO)/6-311++G**, DFT, and DFT-D geometries) and experimental ECD spectra of c-(l-Trp-l-Ala), c-(l-Trp-l-Trp), and c-(l-
Trp-l-Pro). Relative abundances (Boltzmann weights) of different conformers are given in Tables 2 and 3.
&8
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!

Induced Tryptophan Optical Activity in Dipeptides
Figure 6. Calculated (B3LYP/CPCM(DMSO)/6-311++G**, DFT, and DFT-D) and experimental VCD and IR spectra of c-(l-Trp-Gly), c-(l-Trp-l-Trp), and c-(l-Trp-l-
Pro) in DMSO. Experimental intensities are only approximate due to concentration error. Relative abundances (Boltzmann weights) of different conformers are
given in Tables 2 and 3.
tra are measured with Aꢁ0.5 for the amide I, so De/e of 10ꢀ5
is above but close to our reliable measurement limit.) Such
weak VCD spectra are subject to distortion through absorption
artifacts, so that we limited our experimental VCD measure-
ments to those samples for which we have both d,d and l,l
isomers. In addition to amide I, there is a significant VCD signal
arising from CH2 motions and amide II bands (1430–
1530 cmꢀ1), although this region is broader and more complex
resulting in the IR and VCD patterns being less characteristic
and not easily separable into local modes.[96]
The positive signal at 1652 cmꢀ1 predicted by DFT might
correspond to the very weak experimental feature at
1630 cmꢀ1. DFT-D provides a conservative couplet for amide I,
in both C and D conformers, which does not reflect experi-
ment. For the region 1400–1550 cmꢀ1 (combination of CH2 and
amide II modes), on the other hand, neither method is in good
agreement, but several conformers give rise to a positive band
higher in frequency than a negative band, which is seen exper-
imentally. After averaging, the DFT-D VCD curve is perhaps in
better agreement with the experimental amide II than the DFT.
The overlap and mixing of amide II (C-N-H deformation) and
CH2 modes is difficult to reproduce correctly by computations,
as we have also found in previous model calculations.[97]
The spectral simulations (Figure 6) reproduce many of the
observed dependencies, with the IR and VCD predictions
being very good aside from small frequency shifts as are ex-
pected for DFT. For example, in c-(l-Trp-Gly) the amide I VCD
computed with the DFT calculation is predominantly negative.
While many of the individual conformers have couplet amide I
shapes, the negative lobes tend to dominate and, when
weighted by population, prove to be the larger contributions.
For c-(l-Trp-l-Trp) both DFT and DFT-D approaches provide
the basic VCD pattern correctly (mostly negative amide I and
a negative amide II region), although the calculated dispersion
of the negative intensities around 1439 cmꢀ1 is too large. In c-
(l-Trp-l-Pro) the amide I VCD is predominantly negative and
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
9
&
ÞÞ
These are not the final page numbers!

ˇ
J. Hudecovꢀ, P. Bour et al.
broader than in c-(l-Trp-Gly) and c-(l-Trp-l-Trp), which is repro-
duced by the DFT calculations but not by the DFT-D results.
served Raman spectrum well. Both compounds have very simi-
lar Raman intensity patterns, which are dominated by the Trp
modes (see the assignment in Table 8). Similar domination of
the spectra by aromatic residues was observed previously for
a model peptide.[22] The ROA spectra are more complex, but
many observed features can be explained by the calculation.
For example, the C=C five-membered Trp ring stretching band
(experimentally at ꢁ1554 cmꢀ1) exhibits a negative ROA signal.
This is, however, provided only by the dispersion model. On
the other hand, DFT-D overestimates the relative intensity.
As discussed before,[15] although coming from the nonchiral
chromophore, the ROA for this vibration is extremely sensitive
to the Trp side-chain conformation, that is, the c2 angle. In par-
ticular, a negative ROA band is associated with conformations
ROA Spectra
The ROA measurements were often hampered by the sample
fluorescence (Table 1) and known instability of the Trp com-
pounds in the (green) laser light.[98] The best experimental
Raman and ROA spectra were obtained for 150 mgmLꢀ1 (c-(l-
Trp-l-Trp) and c-(d-Trp-d-Trp)) and 50 mgmLꢀ1 (c-(l-Trp-Gly)
and c-(d-Trp-Gly)) solutions in DMSO.
In Figure 7 the ROA and Raman spectra are shown, together
with the corresponding DFT and DFT-D computations. The
computations reproduce the strongest features in the ob-
Figure 7. Calculated (B3LYP/CPCM(DMSO)/6-311++G**, DFT, and DFT-D geometries) and experimental ROA (IRꢀIL) and Raman (IR +IL) spectra of c-(l-Trp-Gly)
and c-(l-Trp-l-Trp) in DMSO. The intensity of the experimental spectra is relative, only the ratio ROA/Raman is meaningful.
&10
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!

Induced Tryptophan Optical Activity in Dipeptides
conformer ratios lying between DFT and DFT-D. Different spec-
tral types (e.g. ECD, VCD, and ROA) also reacted differently in
the dispersion correction; for VCD, for example, DFT-D provid-
ed bands that were too narrow due to the limited number of
folded conformers. To better balance these complex disper-
sion, flexibility, and solvent effects remains a challenge for the
future.
Table 8. Assignment of the most intense Raman bands.[a]
c-(Trp-
Gly)
DFT-D[b]
c-(Trp-
Gly)
Exp.
c-(Trp-
Trp)
Exp.
Vibrations
1704
1695
1652
1607
1578
1449
1377
1357
1312
1271
1241
1202
1118
1019
982
1684
1684
1620
1577
1547
1678
1678
1622
1580
1552
A I, out of phase
A I, in phase
n(C=C), n(C=N) Trp
n(C=C), n(C=N) Trp
n(C=C), Trp 5-membered ring
amide II
amide III
n(C=C), Trp
Trp, 5-membered ring breathing
n(C=C), phenyl ring in Trp
n(C=N) Trp, d(CH)
3. Conclusions
1361
1340
1303
1264
1241
1199
1128
1362
1343
1306
1259
We have systematically compared the ECD, VCD, and ROA
spectra of Trp-containing cyclic dipeptides. The results enable
us to better understand the chiral spectral response of this res-
idue in larger proteins and to characterize the link between
the spectra and molecular structure. The Trp chromophore, al-
though not intrinsically chiral if isolated from the backbone,
dominated in the ECD and ROA dipeptide spectra. Especially
surprising was the large Trp ECD signal of c-(l-Trp-l-Ala), as
this molecule contains only one Trp residue without an exciton
coupling between identical oscillators. Only for VCD can the in-
terference between the side- and main-chain signals be avoid-
ed. However, the relative flatness of the ring makes the VCD
weak and subject to artifacts.
1192
1128
ꢁ1020
955
879
761
d(CH), aliphatic
n(CꢀN) in amides
n(C=C), phenyl ring in Trp
n(CꢀC) in amides, d(CH)
n(C=C), n(C=N) Trp
n(C=C), n(C=N) Trp
n(C=C), n(C=N) Trp, out of plane
amide NH
885
767
564
881
764
576
572
528
550
543
dipeptide ring deformation
[a] Frequencies in cmꢀ1. [b] B3LYP-D/CPCM(DMSO)/6-311++G** level.
The DFT computations provided a reliable basis for spectral
interpretation. Dipeptide theoretical PESs, however, were
strongly influenced by the presence or absence of the disper-
sion correction. Stable conformers yielded about the same
spectra with DFT and DFT-D, but the dispersion energetically
favored the compact folded forms. In general, the corrected
computations also provided better spectra. Nevertheless, sev-
eral indications appeared pointing to an overestimation of the
dispersion effect within the CPCM solvent model. This was also
confirmed by the NMR data and MD simulations, which re-
vealed finer solvent effects, stemming from the solvent–solute
dispersion and hydrogen bonding, that could be only partially
included within DFT.
with c2 ꢁꢀ908, which is in agreement with the prevalent con-
former population of this isomer predicted for c-(l-Trp-Gly)
(Table 3). For c-(l-Trp-l-Trp) the prevalence of the T-shaped
folded conformers with alternate (+908, ꢀ908) c2 values
causes partial cancelation and the resultant negative signal is
smaller, which can be seen in both the theoretical and experi-
mental ROA spectra.
Around 1350 cmꢀ1 (CH bending, amide III) the predominant-
ly positive experimental ROA signal is better reproduced by
DFT-D for c-(l-Trp-l-Trp), but by DFT for c-(l-Trp-Gly). Within
1100–1250 cmꢀ1 the experimental “+ꢀ+ꢀ” pattern seems to
be better reproduced by DFT-D for both peptides. A negative
ROA signal at 920–942 cmꢀ1 is present in both experiments
and all calculations. Below 900 cmꢀ1, the differences in the
spectra provided by DFT and DFT-D are minor; nevertheless,
both computations mostly reproduced the sign pattern ob-
served experimentally. The large experimental positive ROA in-
tensity at 576 and 569 cmꢀ1 for c-(l-Trp-Gly) and c-(l-Trp-l-
Trp), respectively, is not fully reproduced by the computation,
which can be explained by an anharmonic character of the NH
out-of-plane deformation.[99]
The chiral spectroscopies, at least in principle, eliminate the
problems associated with measurements of symmetric mole-
cules in solutions and unstable conformers by NMR spectrosco-
py. However, some experiments were hampered by limited sol-
ubility and instability of the dipeptides with Trp, and artifacts
associated with the overlap with solvent vibrational bands. The
ROA spectra appeared to be the most sensitive to the Trp
side-chain conformation. Specific Trp marker bands could be
found within the entire spectral region. In particular, the
1554 cmꢀ1 ROA signal appeared useful as a unique local probe
of the c2 angle, which is also otherwise difficult to monitor by
other methods. Overall, we can conclude that the chiral spec-
troscopies provide very detailed information about the peptide
structure, which must be, however, supported by theoretical
modeling.
DFT versus DFT-D
To summarize the role of the dispersion at the spectral simula-
tions, we can conclude that adding the physically correct van
der Waals interaction significantly changes conformer equili-
bria. Most spectral features were improved when the disper-
sion was included, in line with similar investigations in the Acknowledgements
past.[52–56] However, there are also indications that only adding
the correction to the dielectric solvent model may be an over-
simplification. This is supported by the NMR data and the MD
This study was performed with support from the Czech National
Grant Agency (P208/11/0105), the Grant Agency of Charles Uni-
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
11
&
ÞÞ
These are not the final page numbers!

ˇ
J. Hudecovꢀ, P. Bour et al.
[40] J. R. Cheeseman in Calculation of Molecular Chiroptical Properties Using
Density Functional Theory, University of Groningen, Groningen, 2007,
p. INV3.
[41] P. L. Polavarapu, Chirality 2008, 20, 664–672.
[42] F. Wang, C. Zhao, P. L. Polavarapu, Biopolymers 2004, 75, 85–93.
[43] T. Weymuth, C. R. Jacob, M. Reiher, ChemPhysChem 2011, 12, 1165–
1175.
versity (126310), and the Ministry of Education (LH11033). The
work at UIC was supported in part by a grant from the National
Science Foundation (CHE 07 18543 to T.A.K.) and T.A.K. in part by
a Research Award from the Alexander von Humboldt Stiftung.
Keywords: circular dichroism · conformation analysis · density
functional calculations · peptides · Raman spectroscopy
[44] C. Herrmann, K. Ruud, M. Reiher, ChemPhysChem 2006, 7, 2189–2196.
´
ˇ
[45] S. Yamamoto, J. Kaminsky, P. Bour, Anal. Chem. 2012, 84, 2440–2451.
[46] S. Luber, M. Reiher, J. Phys. Chem. B 2010, 114, 1057–1063.
[47] C. R. Jacob, S. Luber, M. Reiher, J. Phys. Chem. B 2009, 113, 6558–6573.
[48] C. Merten, L. D. Barron, L. Hecht, C. Johannessen, Angew. Chem. 2011,
123, 10149–10152; Angew. Chem. Int. Ed. 2011, 50, 9973–9976.
[49] M. Pecul, E. Lamparska, C. Capelli, L. Frediani, K. Ruud, J. Phys. Chem. A
2006, 110, 2807–2815.
[1] H. Edelhoch, Biochemistry 1967, 6, 1948–1954.
[2] R. P. Rava, T. G. Spiro, J. Am. Chem. Soc. 1984, 106, 4062–4064.
[3] L. S. Itzhaki, P. A. Evans, C. M. Dobson, S. E. Radford, Biochemistry 1994,
33, 5212–5220.
ˇ
´
ˇ
[50] K. H. Hopmann, K. Ruud, M. Pecul, A. Kudelski, M. Dracꢁnsky, P. Bour, J.
[4] S. M. O’Malley, M. J. McDonald, J. Protein Chem. 1994, 13, 561–567.
[5] E. Bismuto, I. Sirangelo, G. Irace, E. Gratton, Biochemistry 1996, 35,
1173–1178.
[6] C. P. Pan, M. D. Barkley, Biophys. J. 2004, 86, 3828–3835.
[7] R. W. Woody, Eur. Biophys. J. 1994, 23, 253–262.
[8] R. A. Sendak, D. M. Rothwarf, W. J. Wedemeyer, W. A. Houry, H. A.
Scheraga, Biochemistry 1996, 35, 12978–12992.
[9] D. M. Rogers, D. M. Hirst, Biochemistry 2004, 43, 11092–11102.
[10] D. M. Rogers, N. A. Besley, P. O’Shea, J. D. Hirst, J. Phys. Chem. B 2005,
109, 23061–23069.
Phys. Chem. B 2011, 115, 4128–4137.
[51] B. Mennucci, C. Cappelli, R. Cammi, J. Tomasi, Chirality 2011, 23, 717–
729.
[52] I. Dabkowska, H. V. Gonzalez, P. Jurecka, P. Hobza, J. Phys. Chem. A
ˇ
2005, 109, 1131–1136.
[53] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104.
[54] S. Grimme, J. Comput. Chem. 2006, 27, 1787–1799.
ˇ
[55] K. H. Hopmann, J. Sebestꢁk, J. Novotnꢀ, W. Stensen, M. Urbanovꢀ, J.
ˇ
Svenson, J. S. Svendsen, P. Bour, K. Ruud, J. Org. Chem. 2012, 77, 858–
ˇ
[11] A. Roy, P. Bour, T. A. Keiderling, Chirality 2009, 21, E163–E171.
869.
[12] B. Hernꢀndez, F. Pfluger, A. Adenier, S. G. Kriglik, M. Ghomi, J. Phys.
Chem. B 2010, 114, 15319–15330.
[56] X. Li, K. H. Hopmann, J. Hudecovꢀ, W. Stensen, J. Novotnꢀ, M. Urban-
ˇ
ovꢀ, J. S. Svendsen, P. Bour, K. Ruud, J. Phys. Chem. A 2012, 116, 2554–
2563.
[13] S. D. Dieng, J. P. M. Schelvis, J. Phys. Chem. A 2010, 114, 10897–10905.
[14] A. E. Aliaga, I. Osorio-Roman, P. Leyton, C. Garrido, J. Carcamo, C. Can-
iulef, F. Celis, G. Diaz, E. Clavijo, J. S. Gomez-Jeria, M. M. Campos-Val-
lette, J. Raman Spectrosc. 2009, 40, 164–169.
ˇ
[57] J. Jezˇek, R. Markosovꢀ, Collect. Czech. Chem. Commun. 1994, 59, 671–
706.
[58] E. Wꢄnsch, Methoden der Organischen Chemie (Houben-Weyl), Thieme,
Stuttgart, 1974.
[59] J. E. Sheppeck, H. Kar, H. Hong, Tetrahedron Lett. 2000, 41, 5329–5333.
[60] A. Lakhani, A. Roy, M. De Poli, M. Nakaema, F. Formaggio, C. Toniolo,
T. A. Keiderling, J. Phys. Chem. B 2011, 115, 6252–6264.
[15] C. R. Jacob, S. Luber, M. Reiher, ChemPhysChem 2008, 9, 2177–2180.
[16] W. Hu, T. A. Cross, Biochemistry 1995, 34, 14147–14155.
[17] A. Saravanamuthu, T. J. Vickers, C. S. Bond, M. R. Peterson, W. N.
Hunter, A. H. Fairlamb, J. Biol. Chem. 2004, 279, 29493–29500.
[18] A. G. Cochran, N. J. Skelton, M. A. Starovasnik, Proc. Natl. Acad. Sci. USA
2001, 98, 5578–5583.
[61] W. Hug, G. Hangartner, J. Raman Spectrosc. 1999, 30, 841–852.
[62] W. Hug, Appl. Spectrosc. 2003, 57, 1–13.
ˇ
´
ˇ
ˇ
[19] P. Bour, V. Sychrovsky, P. Malon, J. Hanzlꢁkovꢀ, V. Baumruk, J. Pospꢁsek,
[63] Gaussian 09 (RevisionB01), M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratch-
ian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, T. Vreven, J. Montgomery, J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starover-
ov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E.
Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V.
Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc., Pittsburgh, PA, 2009.
[64] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
ˇˇ
´
M. Budesꢁnsky, J. Phys. Chem. A 2002, 106, 7321–7327.
[20] N. A. Besley, M. J. Brienne, J. D. Hirst, J. Phys. Chem. B 2000, 104,
12371–12377.
[21] J. Fleischhauer, J. Grçtzinger, B. Kramer, P. Krꢄger, A. Wollmer, R. W.
Woody, E. Zobel, Biophys. Chem. 1994, 49, 141–152.
[22] J. Hudecovꢀ, J. Kapitꢀn, V. Baumruk, R. P. Hammer, T. A. Keiderling, P.
Bour, J. Phys. Chem. A 2010, 114, 7642–7651.
[23] S. Yamamoto, M. Straka, H. Watarai, P. Bour, Phys. Chem. Chem. Phys.
2010, 12, 11021–11032.
[24] C. Jamorski, M. E. Casida, D. R. Salahub, J. Chem. Phys. 1996, 104,
5134–5147.
[25] F. Furche, R. Alhrichs, J. Chem. Phys. 2004, 121, 12772–12773.
[26] F. Furche, R. Ahlrichs, C. Wachsmann, E. Weber, A. Sobanski, F. Vçgtle,
S. Grimme, J. Am. Chem. Soc. 2000, 122, 1717–1724.
[27] L. A. Nafie, Annu. Rev. Phys. Chem. 1997, 48, 357–386.
[28] P. L. Polavarapu, Chem. Rec. 2007, 7, 125–126.
[29] P. J. Stephens, J. Phys. Chem. 1985, 89, 748–752.
[30] P. J. Stephens, J. Phys. Chem. 1987, 91, 1712–1715.
[31] K. L. Bak, F. J. Devlin, C. S. Ashvar, P. R. Taylor, M. J. Frisch, P. J. Stephens,
J. Phys. Chem. 1995, 99, 14918–14922.
ˇ
ˇ
[65] C. Møller, M. S. Plesset, Phys. Rev. 1934, 46, 618–622.
[66] T. Schwabe, S. Grimme, Phys. Chem. Chem. Phys. 2007, 9, 3397–3406.
[67] J. P. Perdew, K. Burke, Y. Wang, Phys. Rev. B 1996, 54, 16533–16539.
[68] A. Becke, Phys. Rev. A 1988, 38, 3098–3100.
[69] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999–3093.
[70] S. Grimme, J. Comput. Chem. 2004, 25, 1463–1473.
[32] J. R. Cheeseman, M. J. Frisch, F. J. Devlin, P. J. Stephens, Chem. Phys.
Lett. 1996, 252, 211–220.
´
ˇ
[71] J. Kapitꢀn, V. Baumruk, V. Kopecky, Jr., P. Bour, J. Am. Chem. Soc. 2006,
[33] T. Helgaker, K. Ruud, K. L. Bak, P. Joergensen, J. Olsen, Faraday Discuss.
1994, 99, 165–180.
128, 2438–2443.
[72] J. Kapitꢀn, V. Baumruk, V. Kopecky
´, Jr., R. Pohl, P. Bourˇ, J. Am. Chem.
ˇ
Soc. 2006, 128, 13451–13462.
[34] K. Ruud, T. Helgaker, P. Bour, J. Phys. Chem. A 2002, 106, 7448–7455.
[35] V. Liꢅgeois, K. Ruud, B. Champagne, J. Chem. Phys. 2007, 127, 204105.
[36] M. Pecul, K. Ruud, Int. J. Quantum Chem. 2005, 104, 816–829.
[37] K. Ruud, J. Thorvaldsen, Chirality 2009, 21, E54–E67.
[38] J. R. Cheeseman, M. S. Shaik, P. L. A. Popelier, E. W. Blanch, J. Am. Chem.
Soc. 2011, 133, 4991–4997.
[73] C. Altona, M. Sundaralingam, J. Am. Chem. Soc. 1972, 94, 8205–8212.
[74] S. J. Han, Y. K. Kang, J. Mol. Struct.: THEOCHEM 1996, 369, 157–165.
[75] E. E. Zvereva, A. R. Shagidullin, S. A. Katsyuba, J. Phys. Chem. A 2011,
115, 63–69.
[76] M. Reiher, V. Liegeois, K. Ruud, J. Phys. Chem. A 2005, 109, 7567–7574.
[77] J. Haesler, I. Schindelholz, E. Riguet, C. G. Bochet, W. Hug, Nature 2007,
446, 526–529.
[39] J. R. Cheeseman, M. J. Frisch, J. Chem. Theory Comput. 2011, 7, 3323–
3334.
&12
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!

Induced Tryptophan Optical Activity in Dipeptides
ˇ
ˇ
[78] S. Yamamoto, H. Watarai, P. Bour, ChemPhysChem 2011, 12, 1509–
[91] R. Huang, L. Wu, D. McElheny, P. Bour, A. Roy, T. A. Keiderling, J. Phys.
1518.
Chem. B 2009, 113, 5661–5674.
ˇ
[79] T. Yanai, D. Tew, N. C. Handy, Chem. Phys. Lett. 2004, 393, 51–57.
[92] P. Bour, T. A. Keiderling, J. Phys. Chem. B 2005, 109, 23687–23697.
ˇ
[80] O. Julꢁnek, M. Krupicka, W. Lindner, M. Urbanovꢀ, Phys. Chem. Chem.
[93] W. Y. Yang, J. W. Pitera, W. C. Swope, M. Gruebele, J. Mol. Biol. 2004,
336, 241–251.
Phys. 2010, 12, 11487–11497.
[81] D. A. Case, I. T. E. Cheatham, T. Darden, H. Gohlke, R. Luo, J. K. M. Merz,
A. Onufriev, C. Simmerling, B. Wang, R. Woods, J. Comput. Chem. 2005,
26, 1668–1688.
[94] S. J. Russell, T. Blandl, N. J. Skelton, A. G. Cochran, J. Am. Chem. Soc.
2003, 125, 388–395.
ˇ
ˇ
[95] J. Sebek, Z. Kejꢁk, P. Bour, J. Phys. Chem. A 2006, 110, 4702–4711.
[96] T. A. Keiderling, J. Kubelka, J. Hilario in Vibrational Circular Dichroism of
Biopolymers: Summary of Methods and Applications (Eds.: M. Braiman,
V. Gregoriou), CRC, Boca Raton, 2006, pp. 253–324.
ˇ
ˇ
ˇ
´
[82] Z. Vokꢀcovꢀ, J. Sponer, J. E. Sponer, V. Sychrovsky, J. Phys. Chem. B
2007, 111, 10813–10824.
[83] A. P. Mendham, T. J. Dines, M. J. Snowden, B. Z. Chowdhry, J. Raman
Spectrosc. 2009, 40, 1478–1497.
[84] E. Sletten, J. Am. Chem. Soc. 1970, 92, 172–177.
[85] G. D. Grant, A. L. Hunt, P. J. Milne, H. M. Roos, J. A. Joubert, J. Chem.
Crystallogr. 1999, 29, 435–447.
[97] J. Kubelka, T. A. Keiderling, J. Phys. Chem. A 2001, 105, 10922–10928.
[98] K. Huvaere, L. H. Skibsted, J. Am. Chem. Soc. 2009, 131, 8049–8060.
ˇ
[99] C. N. Tam, P. Bour, J. Eckert, F. R. Trouw, J. Phys. Chem. A 1997, 101,
5877–5884.
ˇ
´
ˇ
[86] V. Andrushchenko, P. Matejka, D. T. Anderson, J. Kaminsky, J. Hornꢁcek,
[100] J. M. Schmidt, M. Blꢄmel, F. Lçhr, H. Rꢄterjans, J. Biomol. NMR 1999, 14,
1–12.
ˇ
L. O. Paulson, P. Bour, J. Phys. Chem. A 2009, 113, 9727–9736.
ˇ
´
ˇ
´
ˇ
[87] V. Parchansky, P. Matejka, B. Dolensky, M. Havlꢁk, P. Bour, J. Mol. Struct.
[101] J. M. Schmidt, J. Biomol. NMR 2007, 37, 287–301.
2009, 934, 117–122.
[88] M. J. O. Anteunis, Bull. Soc. Chim. Belg. 1978, 87, 627–650.
ˇˇ
´
´
ˇ
´
[89] M. Budesꢁnsky, J. Symersky, J. Jecny, J. van Hecke, N. Hosten, M. J. O.
Received: March 8, 2012
Anteunis, F. Borremans, Int. J. Pept. Protein Res. 1992, 39, 123–130.
Revised: April 27, 2012
ˇˇ
´
ˇ
[90] M. Budesꢁnsky, I. Cꢁsarovꢀ, J. Podlaha, F. Borremans, J. C. Martins, M.
Waroquier, E. Pauwels, Acta Crystallogr. Sect. B 2010, 66, 662–677.
Published online on && &&, 2012
ChemPhysChem 0000, 00, 1 – 14
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&13
&
ÞÞ
These are not the final page numbers!

ARTICLES
ˇ
ˇˇ
´
J. Hudecovꢀ,* J. Hornꢁcek, M. Budesꢁnsky,
Peptides revealed: The tryptophan resi-
due is a sensitive probe in protein struc-
tural studies, but its own optical activity
can interfere with that of the protein.
The chirality is studied in the electronic
circular dichroism, vibrational circular di-
chroism, and Raman optical activity of
model dipeptides. The experimental re-
sults are related to DFT computations,
revealing unique information about the
conformational behavior of the peptides
(see picture).
ˇ
ˇ
ˇ
J. Sebestꢁk, M. Safarꢁk, G. Zhang,
ˇ
T. A. Keiderling, P. Bour*
&& – &&
Three Types of Induced Tryptophan
Optical Activity Compared in Model
Dipeptides: Theory and Experiment
&14
&
www.chemphyschem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 14
ÝÝ
These are not the final page numbers!