On the role of flexibility in protein-ligand interactions: The example of p53 tetramerization domain

Article abstract of DOI:10.1002/asia.201000938

The recognition of protein surfaces by designed ligands has become an attractive approach in drug discovery. However, the variable nature and irregular behavior of protein surfaces defy this new area of research. The easy to understand "lock-and-key" model is far from being the ideal paradigm in biomolecular interactions and, hence, any new finding on how proteins and ligands behave in recognition events paves a step of the way. Herein, we illustrate a clear example on how an increase in flexibility of both protein and ligand can result in an increase in the stability of the macromolecular complex. The biophysical study of the interaction between a designed flexible tetraguanidinium-calix[4]arene and the tetramerization domain of protein p53 (p53TD) and its natural mutant p53TD-R337H shows how the floppy mutant domain interacts more tightly with the ligand than the well-packed wild-type protein. Moreover, the flexible calixarene ligand interacts with higher affinity to both wild-type and mutated protein domains than a conformationally rigid calixarene analog previously reported. These findings underscore the crucial role of flexibility in molecular recognition processes, for both small ligands and large biomolecular surfaces. Copyright

Full text of DOI:10.1002/asia.201000938

DOI: 10.1002/asia.201000938
On the Role of Flexibility in Protein–Ligand Interactions: the Example of p53
Tetramerization Domain
Susana Gordo,[a] Vera Martos,[b, c] Marta Vilaseca,[a] Margarita Menꢀndez,[d]
Javier de Mendoza,*[b, c] and Ernest Giralt*[a, e]
Abstract: The recognition of protein
surfaces by designed ligands has
become an attractive approach in drug
discovery. However, the variable
nature and irregular behavior of pro-
tein surfaces defy this new area of re-
search. The easy to understand “lock-
and-key” model is far from being the
ideal paradigm in biomolecular interac-
tions and, hence, any new finding on
how proteins and ligands behave in
recognition events paves a step of the
way. Herein, we illustrate a clear exam-
ple on how an increase in flexibility of
both protein and ligand can result in
an increase in the stability of the mac-
romolecular complex. The biophysical
study of the interaction between a de-
signed flexible tetraguanidinium-cal-
ix[4]arene and the tetramerization
domain of protein p53 (p53TD) and its
natural mutant p53TD-R337H shows
how the floppy mutant domain inter-
acts more tightly with the ligand than
the well-packed wild-type protein.
Moreover, the flexible calixarene
ligand interacts with higher affinity to
both wild-type and mutated protein do-
mains than a conformationally rigid
calixarene analog previously reported.
These findings underscore the crucial
role of flexibility in molecular recogni-
tion processes, for both small ligands
and large biomolecular surfaces.
Keywords: ligands · microcalorim-
etry · molecular recognition · NMR
spectroscopy
interactions
·
protein–protein
Introduction
ficult to establish general rules for the selective attachment
of suitable ligands. Most of the many rationally designed
molecules able to recognize surfaces or disrupt protein inter-
faces[3] arise from computational methods.[4] Notably, some
of these ligands do not follow the classic concept of drug-
like molecules, and display large multivalent structures that
cover wide extensions, establishing multiple simultaneous in-
teractions and achieving a higher avidity.[5] In this context,
The recognition of protein surfaces with designed ligands is
a current pursuit in drug discovery,[1] and formidable chal-
lenge as well. The lack of characteristic features, the large
extension, the planarity or the discontinuity of protein surfa-
ces and protein–protein interfaces, coupled with their con-
formational flexibility and dynamic behavior,[2] makes it dif-
[a] Dr. S. Gordo,+ Dr. M. Vilaseca, Prof. E. Giralt
Institute for Research in Biomedicine
Baldiri Reixac 10, 08028 Barcelona (Spain)
Fax : (+34)93-403-7126
[e] Prof. E. Giralt
Department of Organic Chemistry
Universitat de Barcelona
Martꢂ i Franquꢅs 1, 08028 Barcelona (Spain)
[+] Current address:
E-mail: ernest.giralt@irbbarcelona.org
[b] Dr. V. Martos,++ Prof. J. de Mendoza
Institut Catalꢀ d’Investigaciꢁ Quꢂmica
Av. Paꢃsos Catalans 16, 43007 Tarragona (Spain)
Fax : (+34)977-920-926
Department of Cancer Immunology and AIDS
Dana-Farber Cancer Institute - Harvard Medical School
450 Brookline Ave, Boston, MA 02284 (USA)
[
++] Current address:
E-mail: jmendoza@iciq.es
Department of Medicinal Chemistry
Leibniz-Institut fꢆr Molekulare Pharmakologie (FMP)
Robert-Rçssle-Str.10, 13125 Berlin (Germany)
[c] Dr. V. Martos,++ Prof. J. de Mendoza
Department of Organic Chemistry
Universidad Autꢁnoma de Madrid
28049 Cantoblanco, Madrid (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/asia.201000938.
[d] Dr. M. Menꢄndez
CSIC - Instituto Quꢂmica Fꢂsica Rocasolano
Serrano 119, 28006 Madrid (Spain)
Chem. Asian J. 2011, 6, 1463 – 1469
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1463

FULL PAPERS
calixarenes have been de-
scribed as efficient multivalent
platforms for biological targets,
such as peptides[6] and protein
surfaces.[7] Most often, ligand
candidates are designed as
rigid molecules to minimize
the entropic penalty associated
with the loss of degrees of
freedom upon binding.[8] How-
ever, there are many other er-
ratic variables in the equation
that can eventually lead to an
unpredictable outcome in the
macromolecular
recognition
event,[9] and flexibility might
not always be the penalizing
factor.
The tetramerization domain
of protein p53 (p53TD) and its
oligomerization-defective natu-
ral mutants constitute a rele-
vant model to study molecules
capable of recovering and sta-
bilizing the oligomeric state of
wild-type proteins. Protein p53
is a tumor suppressor tran-
scription factor[10] that is mu-
tated in a large number of
human cancers (ca. 50%). Pro-
Figure 1. Calix[4]arene ligands for p53TD. a) and b) Molecular formula of 1, 2, and 3. c) Binding model show-
ing two molecules of the rationally designed ligand 2 bound to the two hydrophobic cavities of p53TD.
tein tetramerization is required for the p53 correct opera-
tion[11] and indeed, several mutations affecting the structural
stability of the tetramerization domain have been reported
in human tumors.[12] In particular, the structural stability of
the germ-line mutation R337H (mainly associated with
adrenocortical carcinoma in children from southern
Brazil)[13] is extremely sensitive to pH in the physiological
range.[14] This trend correlates with the protonation state of
His337 side chain. At low pH, the histidine is protonated
and roughly behaves like the native arginine–aspartate
R337–D352 ion pair, a key interaction for the stability of
the tetramer. However, the hydrogen-bonded ion pair van-
ishes as the pH reaches the upper limit of physiological
values, thus destabilizing the protein structure, compromis-
ing its function, and increasing the chance to develop tissue-
specific tumors.
structures (Figure 1a). This design was based on the previ-
ously reported calix[4]arene 2 (Figure 1b,c),[16] the bis-
crown loops of which at the lower rim were meant to en-
hance the conical rigidity of the macrocycle, thereby pre-
venting its collapse into the pinched-cone conformations
and reducing the entropic penalty upon protein interactio-
n.[8a] Interestingly, in the previous study we observed that
ligand 2 interacted more tightly with mutant p53TD-R337H
than with the well-packed and stable wild-type tetramer.
Such affinity differences were attributed to the structural
flexibility of the mutant, which might enable the protein to
adopt an optimal conformation for ligand accommoda-
tion.[17] Thus, the question arising was whether a more flexi-
ble calixarene (i.e. 1) would result in an even stronger inter-
action or, on the contrary, the entropic penalties would
weaken or preclude complexation.[8b,9f]
The flexible multivalent calix[4]arene ligand 1 (Figure 1a)
was rationally designed to fit its four propyl chains at the
lower rim into one of the two hydrophobic cavities present
on the native tetrameric structure of p53TD. Simultaneously,
its four guanidiniomethyl moieties hanging from the upper
rim would chelate the carboxylate groups from residues
Glu336 and Glu339 from two distinct monomer chains ex-
posed on the protein surface. The four conformationally
free propyl chains are long enough to prevent internal rota-
tion of the aromatic rings (cone inversion),[15] but still allow
the cone to freely oscillate between two distorted, pinched
Results and Discussion
Effect of 1 on the Thermal Stability of the Proteins
Differential scanning calorimetry (DSC) experiments pro-
vided the first answers to our question. The presence of 1
triggered a significant increase in the thermal stability of
both wild-type p53 and mutant R337H tetramerization do-
mains. The shift in the transition peak for the stable wild-
type tetramer was moderate, with a slight increase in the un-
1464
www.chemasianj.org
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 1463 – 1469

Flexibility in Protein–Ligand Interactions
addition, considering the strong dependence of p53TD-
R337H stability on the protonation state of His337,[14] the
high stabilization derived from the binding of 1 could indi-
cate that the complex formation might promote protonation
of the His337 side chain and hence, tetramer stabilization.
DSC results were also reproduced by circular dichroism
unfolding curves, thus proving the specificity of the interac-
tion (see Figure S2 in the Supporting Information).
Detection of the Noncovalent Complex
The stoichiometry of the complex with 1 for both the wild-
type and the mutant protein was confirmed by mass spec-
trometry using soft electrospray ionization (ESI-MS). ESI-
MS experiments not only detected the noncovalent tetramer
of p53TD (Figure 3a) but also the noncovalent complexes
with one and two calix[4]arene-bound molecules (Fig-
ure 3c). These protein–ligand complexes were more evident
in the case of mutant p53TD-R337H (Figure 3b,d). Al-
though these results are consistent with the DSC data, the
intensity of the species detected in the gas phase could not
be directly related to the affinity in solution, given that ESI-
MS has a bias for electrostatic interaction; under vacuum
conditions (in the absence of solvent) hydrophobic interac-
tions are weakened, while electrostatic interactions are
strengthened.[19] However, the differences detected between
the two proteins suggest that the forces stabilizing the non-
covalent complexes may not be the same. The hydrophobic
effect is probably the ruling force for the binding of 1 to
wild-type p53TD; hence, the complex is hardly detected. On
the contrary, the predominant detection of the ligand-bound
Figure 2. Effect of 1 on the thermal stability of the proteins by DSC.
a) Wild-type p53TD and b) mutant p53TD-R337H thermograms record-
ed for samples with 25 mm of protein (tetramer) in the presence of in-
creasing amounts of the ligand, in water at pH 7.0. For clarity, intermedi-
ate ligand ratios in a) are not displayed.
folding enthalpy (Figure 2a). The change was substantially
larger for the tetramerization-defective mutant p53TD-
R337H: at 25 mm p53TD-R337H (concentration considering
a tetramer) and pH 7.0, the maximum of the DSC endo-
therm shifted from 628C to nearly 898C upon addition of
200 mm 1 (Figure 2b). The unfolding enthalpy was also sig-
nificantly increased, although a small exothermic event at
the end of the melting profile for some of the samples
(probably due to the unspecific
aggregation of the unfolded
protein at these high tempera-
tures) prevented an accurate
analysis of the experimental
data.
The unfolding of p53TD-
R337H took place in a bipha-
sic mode at intermediate
ligand concentrations, whereas
for the wild-type p53TD
a
single endotherm was observed
under similar conditions. This
behavior suggests that the af-
finity of 1 for the mutant pro-
tein is larger than for the wild-
type tetramer (at least at the
melting temperature). More-
over, it also indicates that the
dissociation constant of the
mutant complex is below the
protein concentration used in
DSC experiments and/or that
ligand dissociation takes place
at a slower rate in comparison
with the DSC time scale.[18] In
Figure 3. Noncovalent tetrameric protein–ligand complexes detected by ESI-MS. Masses labeled as T (tetra-
mer) correspond to the free protein; TL, to the single-bound complex; TL2, to the double-bound complex.
The charge state of the tetramer is indicated above the peaks. Samples for a) p53TD and b) p53TD-R337H
correspond to tetramer (12.5 mm) in NH4Cl (10 mm) at pH 7.0. Ligand 1 at 125 mm was present in the samples
of c) p53TD and d) p53TD-R337H.
Chem. Asian J. 2011, 6, 1463 – 1469
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
1465

E. Giralt and J. De Mendoza et al.
FULL PAPERS
species for mutant p53TD-R337H, which is energetically
more favored, suggests that electrostatic contributions are
larger than for the wild-type p53TD (in the gas phase), al-
though hydrophobic interactions can be also relevant in
complex formation.
it is possible that the binding of the first ligand altered the
second binding site, thereby increasing its affinity. The ex-
perimental data can be reasonably interpreted by this model
(see Figure S4 in the Supporting Information), as the values
derived for the thermodynamic constants indicate: a) disso-
ciation constants for the wild-type p53TD (KD1: ~120 mm
and KD2: ~30 mm) are higher than for the mutant (KD1
:
Cooperative Binding
~50 mm and KD2: ~15 mm), b) the first binding event im-
proves the affinity for the second site (KD1 >KD2), and c) as
expected from DSC results, both proteins display lower dis-
sociation constants for ligand 1 than for ligand 2 (reported
The titration of 15N-labeled protein with 1 followed by
1H-15N-HSQC spectroscopy (Figure 4a,b; HSQC=hetero-
nuclear single quantum coherence) revealed that the ligand
binds to the protein in a slow chemical exchange process in
the NMR time scale, which usually happens for high-affinity
constants. As expected for a slow exchange, at the beginning
of the titration (i.e. at the lowest ligand concentrations) the
HSQC peaks experienced a decrease in intensity until com-
pletely disappearing; then a new set of unidentified resonan-
ces emerged. For the wild-type protein, the new set of reso-
nances remained unchanged as the concentration of ligand
increased. For p53TD-R337H, however, some of the emerg-
ing resonances were slightly shifted as more ligand was
added.
The observations that almost every resonance was affect-
ed by the presence of 1, the slow exchange regime, the lost
trace of the former resonances, and the different behaviors
displayed by the new ones strongly suggest that the proteins
underwent some structural rearrangements upon ligand
binding. For p53TD-R337H, some of the newly emerged res-
onances could be explained by long-distance structural ef-
fects occurring at different rates within the protein.[20]
No single-bound intermediate (namely, an asymmetric
tetramer) was clearly detected by NMR spectroscopy. Con-
sidering the conformational rearrangements on the protein,
affinities for 2: KD: ~280 mm for wild-type p53TD; KD1
:
~130 mm and KD2: ~65 mm for p53TD-R337H).[16]
Protein Binding Site
The complexity of the above-mentioned NMR results (slow
exchange, loss of track in resonances, structural rearrange-
ments) did not allow mapping of the binding site of ligand 1
on the protein. Therefore, we tested a potentially lower af-
finity tetrapropylcalix[4]arene ligand, 3, in which the four
guanidinium groups at the upper rim were replaced by pri-
mary ammonium moieties (Figure 1a).[16]
The evolution of the HSQC spectra of the proteins during
the titration with the calix[4]arene 3 (Figure 4c,d) con-
firmed the lower affinity of the ligand, in particular, for the
wild-type p53TD. Nonetheless, the binding to p53TD-
R337H remained at an intermediate-slow exchange rate,
thereby highlighting the role of the hydrophobic lower rim
in the recognition process. Assuming a similar cooperative
binding model, the dissociation constants were estimated as
KD1: ~200 mm and KD2: ~50 mm for the wild-type p53TD, and
KD1: ~80 mm and KD2: ~15 mm for the mutant p53TD-R337H
(see Figure S4 in the Support-
ing Information).
The large chemical shifts ex-
perienced by some of the reso-
nances of the native p53TD
when titrated with
3 (Fig-
ure 5a) were also consistent
with the previously suggested
protein structural rearrange-
ments induced by ligand bind-
ing. Mapping the chemical
shifts on the p53TD structure
for the first binding step—fast
enough to track the progress
of
the
resonances
(Fig-
ure 4c)—revealed that the pro-
tein areas defining the hydro-
phobic pocket and the hydro-
phobic core were the most sen-
sitive to ligand 3 (Figure 5b,c).
Such a map indicates that the
calixarene enters the hydro-
phobic pocket and that this
also perturbs the highly stable
Figure 4. Evolution of the 1H-15N-HSQC resonances of the tetrameric protein (100 mm) during the titration
with the calix[4]arene ligands in water (pH 7.0, 298 K). a) Wild-type p53TD and b) p53TD-R337H titrated
with ligand 1. c) Wild-type p53TD and d) p53TD-R337H titrated with ligand 3. (See also the Supporting infor-
mation.)
1466
www.chemasianj.org
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 1463 – 1469

Flexibility in Protein–Ligand Interactions
groups and the protein carboxylates, which might not be
strong enough as to promote the same structural rearrange-
ments to that of 1 upon complex formation. In conjunction
with the lower affinity displayed by the amino ligand, this
finding highlights the relevance of the guanidinium-carbox-
ylate chelation in ligand binding.
Discussion
Flexibility has been traditionally considered a penalizing
factor, as it involves a loss of freedom for both host and
guest upon binding.[8a,23] However, flexibility can often favor
protein–protein interactions and this has been an object of
major attention. For example, many homodimers assemble
from highly disordered monomers and interact tightly in the
final complex. This has also been observed in some hetero-
dimers. However, this behavior has little precedent in drug–
protein interactions. The rationale behind this is that flexible
molecules display a range of energetically close conforma-
tions, and conformers with preferred complementary shapes
are favored to bind to each other through a conformational
population shift.[24]
The case of ligands 1 and 2 complexing p53TD illustrates
this behavior for (flexible) ligand–protein interactions.
Indeed, our experiments show that the specific and coopera-
tive interaction of two molecules of the flexible ligand 1
with p53TD tetramer is stronger than that of the previously
reported rigid ligand 2, as evidenced by the induced thermal
stabilization of the tetrameric assembly for both the wild-
type and the mutant proteins in the presence of 1. In addi-
tion, interaction with 1 promotes a rearrangement in the
protein structure. Likely, the flexible propyl chains of the
calixarene deeply penetrate into the binding pocket, thus es-
tablishing additional tight hydrophobic contacts with the
core of the tetramer. Moreover, ligand flexibility allows the
guanidinium moieties on the upper rim of the calix[4]arene
to be poised in an optimal orientation for chelation and ion-
pairing with the protein carboxylates. This is even more
striking in the case of the flexible and “adjustable” mutant
tetramer (as suggested by mass spectrometry data). Never-
theless, hydrophobicity is the major driving force as it re-
veals the outstanding binding displayed by ligand 3, the
amino analog of guanidinium 1 with reduced hydrogen
bonding and electrostatic interactions.
Although stabilization of the weakly assembled core of
the p53TD-R337H mutant protein is easy to rationalize
from the above considerations, it is remarkable that the
highly robust hydrophobic core of wild-type p53TD[21] is
also further stabilized in the presence of 1. Likely, the loss
of entropy upon protein–ligand binding is overbalanced by
the new hydrophobic and electrostatic interactions created,
with the consequent release of ordered shielding water mol-
ecules to the bulk solvent.
Figure 5. p53TD mapping for the first binding of 3. a) Chemical shift
mapping at 800 mm of 3 (sample of 100 mm tetrameric wild-type p53TD).
Dashed red lines represent the cut-off limits used to categorize the de-
grees of chemical shift for mapping the p53TD structure in b) and c): in
red, residues with changes above 0.1 ppm (mean shift plus one standard
deviation); in yellow, those changing above 0.053 ppm (the mean shift).
In b) the terminal tails of the protein have been removed to better appre-
ciate the hydrophobic cavity. The representation in c) shows an upper
view of the hydrophobic cavity.
hydrophobic core of the protein. This observation was unex-
pected, because it would imply that the stabilizing motif of
the tetramer structure (namely the hydrophobic core)[21] was
being modified.[22] In such a case, it should be concluded
that the additional interactions established between ligand
and protein are favorable enough to overcome the energetic
penalties associated to the whole binding event.
Notably, the final spectra for each protein bound to 1 and
3, albeit similar, were not identical. The differences could
result from a looser interaction between the ligand amino
Determining the true binding mode of ligands 1 and 3 at
a molecular level would provide a deeper understanding of
the interaction mechanism. Unfortunately, the slow chemical
Chem. Asian J. 2011, 6, 1463 – 1469
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
1467

E. Giralt and J. De Mendoza et al.
FULL PAPERS
mass spectrometer (Waters) equipped with a NanoMate automated nano-
electrospray sample dispenser (Advion BioSciences, Ithaca, NY, USA).
The spray voltage was set to 1.7 kV and the delivery pressure at 0.3 psi.
The mass detection was carried out in positive mode at the source, at
208C. The sampling cone voltage was 40 mV, the collision energy in the
trap was set to 10 mV and in the transfer to 8 mV.
exchange prevents a more detailed NMR structural charac-
terization, and obtaining suitable crystals for X-ray structure
determination was unsuccessful (see Figures S6, S7, and S8
in the Supporting Information).
Each macromolecular system features different energetic
and structural characteristics. Some might be estimated
through simplified models, but most of them are difficult to
settle in advance. Consequently, unpredictable behaviors
can be found when moving into the assay tube. In our
system, for instance, the proteins underwent deep structural
rearrangements (unexpected beforehand) to accommodate
and optimally interact with a highly flexible ligand. All ener-
getic penalties were overcome by stabilizing factors, thus
leading to an interaction of higher affinity than with the
rigid ligand. From this perspective, flexibility could indeed
be a way to increase the binding of a ligand to achieve nano-
molar range binding; however, at this stage binding specific-
ity should also be taken into account.
1H-15N-HSQC NMR
Samples containing 15N-labeled protein (100 mm tetramer) in 90% water
with 10% D2O at pH 7.04 were used for the HSQC titration. To prevent
dilution, ligands 1 and 3 were added lyophilized. 1H-15N-HSQC spectra
were recorded at 298 K on a Bruker Digital Advance 600 MHz spectrom-
eter equipped with a cryoprobe. All spectra were processed with the
package NMRPipe/NMRView[27] and visualized and analyzed with the
NMRViewJ software (One Moon Scientific, Inc.).
Dissociation Constant Analysis
For details see the Supporting information
In conclusion, our results highlight an important, though
often neglected, aspect of protein–ligand recognition pro-
cesses, namely, the role of the flexibility of both partners to
achieve optimal complementarity. Thus, use of ligand flexi-
bility as a design tool could be eventually extrapolated for
new paradigms in drug discovery.
Acknowledgements
This work was supported by the Spanish Ministry of Science and Educa-
tion (MEC) (projects BIO2005-00295 and CTQ2005-08948-C02/BQU),
Consolider Ingenio 2010 (Grant CSD2006-0003), and the ICIQ Founda-
tion. S.G. and V.M. acknowledge the Generalitat de Catalunya AGAUR,
and the MEC for predoctoral fellowships.
[1] a) M. R. Arkin, J. A. Wells, Nat. Rev. Drug Discovery 2004, 3, 301–
317; b) P. Chꢅne, ChemMedChem 2006, 1, 400–411; c) M. Strong, D.
Eisenberg, Prog. Drug Res. 2007, 64, 193–215; d) S. Gordo, E.
Giralt, Protein Sci. 2009, 18, 481–493.
[2] a) S. Jones, J. M. Thornton, Proc. Natl. Acad. Sci. USA 1996, 93, 13–
20; b) K. V. Radha Kishan, Curr. Protein Pept. Sci. 2007, 8, 376—
380.
Experimental Section
Chemical Synthesis of the Calix[4]arene
In analogy to compound 2,[16] ligand 1 was synthesized by guanidylation
of the tetra-O-propyl-tetraaminomethylcalix[4]arene 3[25] with N,N’-bis-
(tert-butoxycarbonyl)guanidine N’’-triflate.[26] The resulting bis
ACHTUNGTRENNUNG
1
[3] a) P. A. Nygren, M. Uhlen, Curr. Opin. Struct. Biol. 1997, 7, 463–
469; b) S. Fletcher, A. D. Hamilton, Curr. Opin. Chem. Biol. 2005, 9,
632–638; c) W. J. Cooper, M. L. Waters, Curr. Opin. Chem. Biol.
2005, 9, 627–631; d) H. Yin, A. D. Hamilton, Angew. Chem. 2005,
117, 4200–4235; Angew. Chem. Int. Ed. 2005, 44, 4130–4163;
e) J. K. Murray, S. H. Gellman, Biopolymers 2007, 88, 657–686; f) P.
Block, N. Weskamp, A. Wolf, G. Klebe, Proteins Struct. Funct.
Bioinf. 2007, 68, 170–186.
under standard acidic conditions. For full details on the synthesis, see the
Supporting Information.
Protein Expression and Purification
As previously described.[16]
Differential Scanning Calorimetry (DSC)
[4] a) D. Gonzalez-Ruiz, H. Gohlke, Curr. Med. Chem. 2006, 13, 2607–
2625; b) Y. Che, B. R. Brooks, G. R. Marshall, J. Comput. Aided
Mol. Des. 2006, 20, 109–130; c) A. S. Ivanov, O. V. Gnedenko, A. A.
Molnar, Y. V. Mezentsev, A. V. Lisitsa, A. I. Archakov, J. Bioinf.
Comput. Biol. 2007, 5, 579–592; d) C. McInnes, Curr. Opin. Chem.
Biol. 2007, 11, 494–502.
[5] a) M. Mammen, S. K. Choi, G. M. Whitesides, Angew. Chem. 1998,
110, 2908–2953; Angew. Chem. Int. Ed. 1998, 37, 2754–2794; b) E.
Fan, Z. Zhang, W. E. Minke, Z. Hou, C. L. M. J. Verlinde, W. G. J.
Hol, J. Am. Chem. Soc. 2000, 122, 2663–2664; c) M. A. Fazal, B. C.
Roy, S. Sun, S. Mallik, K. R. Rodgers, J. Am. Chem. Soc. 2001, 123,
6283–6290; d) J. E. Gestwicki, C. W. Cairo, L. E. Strong, K. A.
Oetjen, L. L. Kiessling, J. Am. Chem. Soc. 2002, 124, 14922–14933;
e) S. N. Gradl, J. P. Felix, E. Y. Isacoff, M. L. Garcia, D. Trauner, J.
Am. Chem. Soc. 2003, 125, 12668–12669.
DSC thermograms were obtained in a VP-DSC microcalorimeter (Micro-
cal, Northhampton, MA, USA), scanning from 108C to 1208C, at
308C hÀ1 and at a constant pressure of 2 atm. Samples containing 25 mm
of protein (tetramer) and ligand 1 were prepared in water (adjusting pH
to 7.0) and degassed under vacuum at 188C for 30 min before filling the
calorimeter cell. The reference cell contained the corresponding degassed
sample with only ligand 1. The sample cell was systematically washed be-
tween samples and water blanks were recorded before loading the pro-
tein–ligand sample. Experimental data were analyzed with Microcal
Origin 7.0 software. After subtracting the buffer baseline from the raw
thermogram, the excess heat capacity function for the unfolding transi-
tion, <CPtr >, was obtained by subtracting a progressive baseline traced
between the signals corresponding to the native and the unfolded states.
Electrospray Ionization Mass Spectrometry (ESI-MS)
[6] a) L. Baldini, A. Casnati, F. Sansone, R. Ungaro, Chem. Soc. Rev.
2007, 36, 254–266; b) M. Torvinen, R. Neitola, F. Sansone, L. Baldi-
ni, R. Ungaro, A. Casnati, P. Vainiotalo, E. Kalenius, Org. Biomol.
Chem. 2010, 8, 906–915.
[7] V. Martos, P. CastreÇo, J. Valero, J. de Mendoza, Curr. Opin. Chem.
Biol. 2008, 12, 698–706.
Protein samples were extensively desalted by ultracentrifugation (3 kDa
cut-off) with at least three volumes of Milli-Q water MS-grade and then
lyophilized. Proteins were dissolved in NH4Cl (10 mm; MS grade) at
pH 7.0 to a final concentration of 12.5 mm (tetramer). The stock of ligand
1 was dissolved in NH4Cl (10 mm) at pH 7.0 (although it could not be de-
salted). Protein–ligand samples were preincubated for 5 min at room
temperature. ESI-MS measurements were performed in a Synapt HDMS
[8] a) C. E. Chang, W. Chen, M. K. Gilson, Proc. Natl. Acad. Sci. USA
2007, 104, 1534–1539; b) G. Klebe, H. J. Bçhm, J. Recept. Signal
1468
www.chemasianj.org
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 1463 – 1469

Flexibility in Protein–Ligand Interactions
Transduction Res. 1997, 17, 459–473; c) K. P. Murphy, E. Freire,
Pharm. Biotechnol. 1995, 7, 219–241.
[17] a) S. J. Teague, Nat. Rev. Drug Discovery 2003, 2, 527–541; b) J. S.
Marvin, H. W. Hellinga, Nat. Struct. Biol. 2001, 8, 795–798.
[18] a) J. F. Brandts, L. N. Lin, Biochemistry 1990, 29, 6927–6940; b) A.
Shrake, P. D. Ross, J. Biol. Chem. 1990, 265, 5055–5059.
[19] C. V. Robinson, E. W. Chung, B. B. Kragelund, J. Knudsen, R. T.
Aplin, F. M. Poulsen, C. M. Dobson, J. Am. Chem. Soc. 1996, 118,
8646–8653.
[20] a) T. Mittag, B. Schaffhausen, U. L. Gunther, J. Am. Chem. Soc.
2004, 126, 9017–9023; b) M. Pellecchia, D. L. Montgomery, S. Y.
Stevens, C. W. Vander Kooi, H. P. Feng, L. M. Gierasch, E. R. Zui-
derweg, Nat. Struct. Biol. 2000, 7, 298–303.
[21] C. R. Johnson, P. E. Morin, C. H. Arrowsmith, E. Freire, Biochem-
istry 1995, 34, 5309–5316.
[22] A. Weichsel, W. R. Montfort, Nat. Struct. Biol. 1995, 2, 1095–1101.
[23] K. P. Murphy, D. Xie, K. S. Thompson, L. M. Amzel, E. Freire, Pro-
teins Struct. Funct. Bioinf. 1994, 18, 63–67.
[24] a) S. Kumar, C. J. Tsai, R. Nussinov, Biochemistry 2003, 42, 4864–
4872; b) A. Barzilai, S. Kumar, H. Wolfson, R. Nussinov, Proteins
Struct. Funct. Bioinf. 2004, 56, 635–649; c) D. D. Boehr, R. Nussinov,
P. E. Wright, Nat. Chem. Biol. 2009, 5, 789–796; d) D. D. Boehr, D.
McElheny, H. J. Dyson, P. E. Wright, Science 2006, 313, 1638–1642.
[25] A. Casnati, L. Pirondini, N. Pelizzi, R. Ungaro, Supramol. Chem.
2000, 12, 53–65.
[26] K. Feichtinger, H. L. Sings, T. J. Baker, K. Matthews, M. Goodman,
J. Org. Chem. 1998, 63, 8432–8439.
[27] F. Delaglio, S. Grzesiek, G. W. Vuister, G. Zhu, J. Pfeifer, A. Bax, J.
Biomol. NMR 1995, 6, 277–293.
[9] a) P. R. Connelly, R. A. Aldape, F. J. Bruzzese, S. P. Chambers, M. J.
Fitzgibbon, M. A. Fleming, S. Itoh, D. J. Livingston, M. A. Navia,
J. A. Thomson, Proc. Natl. Acad. Sci. USA 1994, 91, 1964–1968;
b) A. Cooper, Biophys. Chem. 2005, 115, 89–97; c) J. E. Ladbury,
Chem. Biol. 1996, 3, 973–980; d) C. Clarke, R. J. Woods, J. Gluska,
A. Cooper, M. A. Nutley, G. J. Boons, J. Am. Chem. Soc. 2001, 123,
12238–12247; e) Z. Li, T. Lazaridis, J. Phys. Chem. A J. Phys. Chem.
B 2005, 109, 662–670; f) F. Yin, R. Cao, A. Goddard, Y. Zhang, E.
Oldfield, J. Am. Chem. Soc. 2006, 128, 3524–3525.
[10] A. J. Levine, Cell 1997, 88, 323–331.
[11] E. Shaulian, A. Zauberman, J. Milner, E. A. Davies, M. Oren,
EMBO J. 1993, 12, 2789–2797.
[12] A. Petitjean, E. Mathe, S. Kato, C. Ishioka, S. V. Tavtigian, P. Hai-
naut, M. Olivier, Hum. Mutat. 2007, 28, 622–629.
[13] R. C. Ribeiro, F. Sandrini, B. Figueiredo, G. P. Zambetti, E. Michal-
kiewicz, A. R. Lafferty, L. DeLacerda, M. Rabin, C. Cadwell, G.
Sampaio, I. Cat, C. A. Stratakis, R. Sandrini, Proc. Natl. Acad. Sci.
USA 2001, 98, 9330–9335.
[14] E. L. DiGiammarino, A. S. Lee, C. Cadwell, W. Zhang, B. Bothner,
R. C. Ribeiro, G. Zambetti, R. W. Kriwacki, Nat. Struct. Biol. 2002,
9, 12–16.
[15] A. Arduini, M. Fabbi, M. Mantovani, L. Mirone, A. Pochini, A.
Secchi, R. Ungaro, J. Org. Chem. 1995, 60, 1454–1457.
[16] S. Gordo, V. Martos, E. Santos, M. Menendez, C. Bo, E. Giralt, J. de
Mendoza, Proc. Natl. Acad. Sci. USA 2008, 105, 16426–16431.
Received: December 29, 2010
Chem. Asian J. 2011, 6, 1463 – 1469
ꢇ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
1469