Control of protein-protein interactions: Structure-based discovery of low molecular weight inhibitors of the interactions between Pin1 WW domain and phosphopeptides

Article abstract of DOI:10.1021/jm0500119

Interactions involving phosphorylated Ser/Thr-Pro motifs in proteins play a key role in numerous regulatory processes in the cell. Here, we investigate potential ligands of the WW binding domain of Pin1 in order to inhibit protein-protein interactions between Pin1 and phosphopeptides. Our structure-based strategy implies the synthesis of analogues of the Ac-Thr(PO₃H₂)-Pro-NH₂ dipeptide and relies on high resolution NMR spectroscopy to accurately measure the affinity constants even in the high micromolar range.

Full text of DOI:10.1021/jm0500119

J. Med. Chem. 2005, 48, 4815-4823
4815
Control of Protein-Protein Interactions: Structure-Based Discovery of Low
Molecular Weight Inhibitors of the Interactions between Pin1 WW Domain and
Phosphopeptides
Caroline Smet,† Jean-Fre´de´ric Duckert,† Jean-Michel Wieruszeski,† Isabelle Landrieu,† Luc Bue´e,‡
Guy Lippens,† and Benoˆıt De´prez*,†
CNRS-University of Lille 2 UMR 8525, Institut Pasteur de Lille, Molecular Drug Discovery, 3 rue du Professeur Laguesse,
59000 Lille, France, and Inserm U422, place de Verdun, 59000 Lille, France
Received January 6, 2005
Interactions involving phosphorylated Ser/Thr-Pro motifs in proteins play a key role in
numerous regulatory processes in the cell. Here, we investigate potential ligands of the
WW binding domain of Pin1 in order to inhibit protein-protein interactions between Pin1
and phosphopeptides. Our structure-based strategy implies the synthesis of analogues of the
Ac-Thr(PO3H2)-Pro-NH2 dipeptide and relies on high resolution NMR spectroscopy to accurately
measure the affinity constants even in the high micromolar range.
Introduction
of the WW Ser16 inhibits its binding role and inacti-
vates the whole protein. In breast cancer cells, the WW
domain is dephosphorylated, leading to the active form
of the protein.14
Protein-protein interactions have an important role
in most biological processes. A number of small domains
present in various proteins with different biological
functions are recurrent in protein-protein interactions
and stand as such in the core of the cellular protein
network.1 Interactions mediated by phosphorylated
moieties on their targets are of particular interest, since
crucial regulation processes involve a balance between
phosphorylation and dephosphorylation pathways al-
lowed by kinase and phosphatase activities.2
Pin1 is a peptidyl-prolyl cis/trans isomerase consti-
tuted by two domains, a catalytic prolyl cis/trans
isomerase domain and a WW binding module.3 Both
share a specificity for phospho-Thr/Ser-Pro motifs. At
the cellular level, Pin1 is shown to be an essential
protein involved in cell cycle regulation, since Pin1
inhibition prevents entry into mitosis and cell division.4,5
Overexpression of Pin1 in numerous human cancers
indicates its implication in several oncogenic path-
ways.6,7 Therefore, Pin1 is proposed as a therapeutic
target for cancer treatment.8 Pin1 also interacts with
the neuronal microtubule-associated tau protein when
the latter is phosphorylated at specific motifs, and
thereby restores its ability to polymerize the tubulin into
microtubules.9
Structural data on Pin1 or isolated WW domain in
complexes with phosphopeptides indicate an extremely
limited interface between both partners, centered on
phosphoThr-Pro dipeptide and a poor contribution of
other sequential residues surrounding the dipeptide.15,16
At the molecular level, the WW domain has a flexible
loop encompassing Ser16-Ser19 residues that binds the
phosphate moiety on Ser or Thr residues and an
aromatic clamp constituted by its Tyr23 and Trp34 side
chains that binds the prolyl residue in the +1 position
of the phosphoSer/Thr (Figure 1).15,16 These two points
of interaction are of particular interest and are shared
by other binding modules that display either the phos-
pho specificity, distinctly for phosphoTyr (SH2,17 PTB18)
and phosphoSer/Thr (14-3-3 and FHA19), or affinity
for proline-rich sequences like SH3 and other classes
of WW domains (WW type I, II/III).20,21 The phosphate
moiety brings in strong hydrogen bonds toward the
guanidine side chain and backbone amide groups, and
is as such a real challenge to the development of potent
mimetics.22 On the other hand, proline displays a unique
pattern among the 20 natural amino acids because its
side chain is comprised in a five membered ring allowing
a critical role in protein structuration (formation of
â-turn or PP II helix in proline-rich peptides) or in
protein-protein interactions.21 Although steric and
stereoelectronic effects of aryl-, alkyl-, and heteroatom-
substituted prolines have been used to study factors that
affect prolyl peptide conformation and activity,23 the
impact of such modifications on the recognition of the
side chain itself is receptor dependent. Some examples
have been described in the case of protease inhibi-
tors.24,25
Inhibition of Pin1 activity with small molecules
targeting the enzymatic domain blocks cell proliferation
in cancer cell lines. Juglone was the first of this category
of irreversible ligands, and the synthesis and charac-
terization of several derivatives has been described.10,11
Recently, conformationally constrained isosteres of phos-
phoSer-cis/trans-Pro were described in inhibition of
cis/trans catalytic activity of Pin1.12 The WW domain
of Pin1, however, is equally essential for its in vivo
activity.5,13 Moreover, in healthy tissue, phosphorylation
The present study is an exploration of both functional
groups in the context of the Pin1 WW domain, using
NMR spectroscopy throughout to evaluate the indi-
vidual compounds.
* To whom correspondence should be addressed. Phone: +33/(0)
320964024.
Fax:
+33/(0)
320964709.
E-mail:
bdeprez@
pharma.univ-lille2.fr.
† CNRS-University of Lille 2 UMR 8525.
‡ Inserm U422.
10.1021/jm0500119 CCC: $30.25 © 2005 American Chemical Society
Published on Web 06/25/2005

4816 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15
Smet et al.
high dissociation constant of 150 µM and a similar mode
of binding as larger peptidic substrates that derive from
cell cycle-related proteins like RNA Pol II15 and Cdc25
or the neuronal tau protein.16,26 On the basis of the
molecular structures of the WW domain and the ligand
(1a), we investigate here potential ligands for the Pin1
WW binding domain, by sequentially considering com-
pounds that contain proline or phosphate mimics, and
further extending these compounds with additional
hydrophobic moieties (Figure 2). High-resolution NMR
proved a sensitive assay even for very weak ligands,27,28
with cryogenic probe technology limiting appropriately
the amount of protein and ligand.29 Moreover, structural
information could be obtained at every step through a
simple mapping of chemical shift variations of the WW
domain.
Figure 1. Structure of the dipeptide Ac-Thr(PO3H2)-Pro-NH2
(1a) bound to the Pin1 WW domain. The WW domain is
represented as ribbon and the substrate as thick sticks. Side
chains of WW residues implicated in dipeptide binding are
represented as thin sticks and annotated. Hydrogen bonds
between phosphate and Ser16 hydroxyl group and Arg17
guanidinium function are indicated by lines and distances. The
proline moiety is encompassed in an aromatic clamp consti-
tuted by side chains of Tyr23 and Trp34.
Results and Discussion
Synthesis. First, we have investigated separately the
replacement of proline and the replacement of phospho-
threonine in two series of molecules described in the
template Figure 2: (i) proline mimics coupled to phos-
phothreonine through an amide bond in the compounds
2-9a and (ii) polar moieties replacing phosphate sup-
ported by various backbone coupled to prolinamide in
the compounds 10-13. Then, introduction of hydropho-
bic moieties in compounds 1b-1g, 9d, 9g, 9h, 14, and
15 was investigated in order to improve compound
affinities for the WW module. Compounds 1-11 were
synthesized by solid phase chemistry on polystyrene
solid support functionalized by a rink amide linker
(Scheme 1). The chemistry involved is very similar to
solid phase peptide synthesis. Amine functions were
protected by a Fmoc group and deprotected by suspend-
ing the resin in a 20% piperidine solution in DMF.
Carboxylic acid activation was performed by a mixture
of 1 equiv of HOBt/1 equiv of HBTU and 3 equiv of
DIPEA in DMF. Cleavage of compounds from resin
beads and deprotection of polar moiety (phosphate,
carboxylate) were performed under high concentrations
of TFA. Compounds 12-15 were synthesized by clas-
sical homogeneous phase synthesis by coupling amino
acid derivatives of various chain lengths (â-alanine,
ornithine, lysine) on prolinamide and functionalizing the
amine group of these compounds with polar moieties
(squarate, trifluoromethylsulfonyl) (Scheme 2).
Pin 1 WW Domain Expression, Purification, and
NMR Characterization. The Pin1 WW binding mod-
ule [Ala2-Asn40] was cloned in a pET28 plasmid fu-
sioned to an N-terminal His6 tag and overexpressed in
Escherichia coli that allows uniform isotopic labeling
of the protein for NMR studies. The protein was purified
from the soluble cell lysate using two chromatography
systems: chromatography based on nickel affinity and
reverse phase high pressure liquid chromatography. The
protein was obtained at a degree of purity superior to
95% (SDS-PAGE) and characterized by matrix-assisted
laser desorption ionization time-of-flight (MALDI-TOF)
mass spectrometry and high resolution NMR. After
purification, the Pin1 WW domain spontaneously folded
when dissolved in NMR buffer, and bound correctly to
phosphorylated tau peptides as assessed by NMR or
FRET experiments.26 No detectable interaction was
observed with nonphosphorylated peptides. Initially, the
Figure 2. Structures of synthetic WW domain-binding ana-
logues. Structure details of proline mimics and polar moieties
are given in Table 1 and Table 2, respectively, and structures
of hydrophobic moieties in Table 3.
The Ac-Thr(PO3H2)-Pro-NH2 dipeptide 1a (Figure 2)
was assessed by NMR as the minimal requirement for
binding to the Pin1 WW module exhibiting a relatively

Discovery of Potent Pin1 WW Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15 4817
Scheme 1. a General Scheme of Synthesis for Compounds 1-11
a Reagents and conditions: (i) (a) piperidine 20%, DMF, room temperature, 25 min; (b) N-Fmoc-(L)-Pro-OH or mimetics, HOBt, HBTU,
3 equiv of DIPEA, DMF, room temperature, 1 h; (ii) (a) piperidine 20%, DMF, room temperature, 25 min; (b) N-Fmoc-(L)-Thr(PO(OBn)OH)-
OH or N-Fmoc-(L)-Glu(tBu)-OH, HOBt, HBTU, 3 equiv of DIPEA, DMF, room temperature, 1 h; (iii) (a) piperidine 20%, DMF, room
temperature, 25 min; (b) Ac2O, 3 equiv of DIPEA, DMF, room temperature, 30 min; (c) TFA 95%, TIS 2.5%, water 2.5%, room temperature,
5 h; (iv) TFA 95%, TIS 2.5%, water 2.5%, room temperature, 5 h; (v) (a) piperidine 20%, DMF, room temperature, 25 min; (b) carboxylic
acid precursor, HOBt, HBTU, 3 equiv of DIPEA, DMF, room temperature, 1 h; (c) TFA 95%, TIS 2.5%, water 2.5%, room temperature,
5 h; (vi) (a) piperidine 20%, DMF, room temperature, 25 min; (b) glutaric anhydride or succinic anhydride, 3 equiv of DIPEA, DMF, room
temperature, 1 h; (c) TFA 95%, TIS 2.5%, water 2.5%, room temperature, 2 h; (vii) (a) piperidine 20%, DMF, room temperature, 25 min;
(b) 3-(dibenzyloxyphosphoryl)-propionic acid, HOBt, HBTU, 2 equiv of DIPEA, DMF, room temperature, 6 h; (viii) (a) TFA 95%, TIS
2.5%, water 2.5%, room temperature, 5 h; (b) H2 at atmospheric pressure, Pd/C 10%, MeOH, room temperature, 48 h.
protein was assigned using conventional tridimensional
NMR experiments (HNCA, HNCO, and HN(CO)CA) on
a 500 µM 15N/13C-labeled protein sample on a 600 MHz
spectrometer (Bruker, Karlsruhe, Germany)).
Discovery of Novel Reversible Ligands for the
Pin1 WW Binding Module. In an initial phase, we
screened compounds by rapid 1D NMR experiments at
a 20-fold excess of ligand to WW domain. The WW
domain at 20 µM was 15N-labeled, allowing the efficient
filtering of the ligand resonances. The signal of the HNꢀ
of the Trp34 (at 10.2 ppm) was used as a sensitive probe
to detect perturbations of the Trp side chain chemical
environment upon binding of the proline ring. When
interaction was detected, we performed an additional
titration study with a 50 µM 15N-labeled WW sample
and increasing amounts of ligands (from 0.5 to 20 equiv).

4818 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15
Smet et al.
Scheme 2. a General Scheme of Synthesis for Compounds 12-15
a Reagents and conditions: (i) (a) DCC/DCM; (b) prolinamide/DCM, 0 °C, 2 h; (ii) HCl/dioxane, 0 °C then room temperature, 3 h; (iii)
(a) NEt3/EtOH; (b) diethylsquarate/EtOH, room temperature, 12 h; (iv) HCl/DCM room temperature, 24 h; (v) (a) NEt3/DCM; (b)
trifluorosulfonyl chloride/DCM, room temperature, 48 h.
Gradual chemical shift perturbations with increasing
ligand concentrations were used to derive the dissocia-
tion constant (Figure 3).
prolyl ring binding pocket remain essential and appear
as a minimal requirement for ligands to adapt the
domain since the replacement of phosphothreonine by
a glutamic acid in dipeptides carrying the same hydro-
phobic moieties did not significantly bind the WW
domain (data not shown).
The aromatic clamp was found to accept Pro mimics
with structural variations of the pyrrolidine ring. De-
spite the relatively low affinities, NMR titrations yielded
reliable SAR results for analogues of Ac-Thr(PO3H2)-
Pro-NH2 (2-9a). Compounds where the pyrrolidine ring
is replaced by an azetidine or a piperidine bind the WW
domain without significant changes of affinity. The
same observation is made for the epimer at the CR of
proline. Substitutions have contradictory effects on
binding since the replacement of proline by 4-trans-
hydroxyproline, 4-cis-fluoroproline, or thiazolidine-4-
carboxylic acid isostere decreases dramatically the
affinity for the WW domain. However replacement of
proline with a 4-trans-fluoroproline improves signifi-
cantly the binding as well as the introduction of an
unsaturation at position 3-4 of the pyrrolidine ring
(Table 1). Contrary to the proline mimics, polar groups
used to replace the phosphate moiety all exhibited
dissociation constants as low as 2 mM, pointing out an
extreme specificity of the flexible loop for phosphate
binding (Table 2).
Faced with restricted possibilities for structural modi-
fications around the two major points of interaction of
dipeptide 1a with the WW domain, we have examined
further WW hydrophobic patches in the near proximity
of the main binding site in order to gain in affinity. A
potential hydrophobic surface constituted by Phe25 and
Ala31 was targeted with a phenyl ring anchored to the
Thr(PO3H2)-Pro dipeptide via aliphatic chains of various
lengths in compounds 1b-1f and with the Fmoc group
in compound 1g. A significant increase of binding
affinity was observed for these compounds with a
maximum for the 4-phenylbutyric acid precursor and
for the Fmoc group that confirms the agreement with
the 4-carbon-length chain (Table 3). Nevertheless, in-
teractions with the phosphate binding loop and the
Consistency of Binding Mode in the Series.
Chemical shift interaction mapping was performed for
each compound of significantly improved affinity, to
verify the consistency of the binding mode. Very similar
modes of binding were observed for the various proline
mimics in comparison with the dipeptide 1a. The 3,4-
dehydroproline moiety induces larger chemical shift
perturbations for both phosphate and proline binding
sites, and especially a stronger effect on the Y23 amide
peak (Figure 4A), together with a significant line
broadening concerning most of the interacting residues.
The binding site of the hydrophobic moiety that replaces
the acetyl group in the phospho-Thr-Pro dipeptide is
given by the chemical shift mapping of compounds
1b-1g. Larger Ala31 amide shifts (Figure 4B) and a
dependence of the Ala31 variations with the chain
length carrying a phenyl group (1b-1e) that is not
shared by Trp34 or Arg17 verify our initial structural
hypothesis. It cannot be ruled out that observed chemi-
cal shift variations could also be due to conformational
changes of the domain upon ligand binding. However,
since little variations are observed at remote amino
acids from the binding pocket, we reasonably make the
assumption that, in the series, the conformation of the
WW domain in its ligand bound state is not strongly
affected.
Combining the data on chain length and nature of the
aromatic group with the previously determined optimal
proline mimics provides a further perspective for im-
proving the affinity. The combination of the three
essential motifs was further investigated in compounds
9d, 9g, 9h on the basis of the 3,4-dehydroproline
analogue (Figure 2). In the proline series (1b-1g), the

Discovery of Potent Pin1 WW Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15 4819
1
Figure 3. NMR titration of the 15N-labeled WW domain with the ligand 1g. Overlaid H-15N HSQC spectra acquired on a 50
µM WW domain 15N-labeled sample with increasing amounts of compound 1g. In the left top corner, a titration curve plotting the
chemical shift variations against the concentration ratio 1g:WW.
replacement of the acetyl group by ω-arylalkanoyl
chains or a Fmoc group increased affinity as expected.
Surprisingly, except for 4,4-bis(4-hydroxyphenyl)pen-
tanoyl (9h), most of the hydrophobic moieties induce a
decrease of affinity (Table 3) in the dehydroproline
series, suggesting different structural constraints in the
free ligand of both series.
Revisiting Phosphate Mimics on Improved Lig-
ands. The ligands containing an additional point of in-
teraction with the WW domain through the hydrophobic
moiety display affinities in the low micromolar range.
This level of affinity allows us to revisit possible phos-
phate isosteres. Squaramide is singly charged at physi-
ological pH and thanks to delocalization of this charge
can accept strong hydrogen bonds at each of its oxygen
atoms. Compounds carrying this group instead of the
phosphate were synthesized, and their binding was
studied by NMR. In the Fmoc series, compounds 14 and
15 displayed measurable affinities (respectively 910 and
540 µM), thereby providing new entry points in series
where the phosphate moiety could be replaced by an
isosteric group. This provides new perspectives for
improvement of metabolic stability and cell membrane
permeation.
Perspectives in Disrupting Protein-Protein In-
teractions and Compound Activity in a Cell-Based
Assay. To confirm that our small ligands can actually
disrupt interactions between the WW binding module
and its proline-containing substrates, we have con-
ducted competition experiments using fluorescence
energy transfer. Our ligands were tested for their
ability to displace a tetradecapeptide centered on the
Thr212(PO3H2)-Pro213 of Tau protein26 from the WW
domain. In this setup, our best ligands were shown to
displace the peptide with Ki consistent with their Kd
(data not shown).
We have tested some of our lead compounds in a
cellular model of SY5Y neuroblastoma cells with induc-

4820 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15
Smet et al.
Table 1. Structures of Proline Mimics and Dissociation
Constants of Resulting Ligands (2-9a)
Table 2. Structures of Polar Moieties and Dissociation
Constants of Resulting Ligands (10-13)
fluorescence polarization or energy transfer. It allowed
initial screening, and generated reliable SAR at low
levels of affinity, as well as structural information on
the binding mode of our ligands. The exact mode of
action of our ligands is currently being studied.
Experimental Section
Protected natural amino acid, coupling reagents (HOBt and
HBTU), and resin (polystyrene cross-linked with 2% divinyl-
benzene carrying a rink amide MBHA linker) were purchased
from Novabiochem. Purity was checked by HPLC and
LC-MS systems. LC-MS data were recorded on a Varian
1200 MS system using a TSKgel OSH supelco column. Water
(A) and acetonitrile (B) containing 0.1% and 0.12% formic acid,
respectively, were used as eluents. After sample injection, the
column is equilibrated in solution A for 2 min and a 0-100%
B 7.30 min-gradient is performed at a flow rate of 1 mL/min
to elute compounds that are detected at 215 and 254 nm.
HPLC data are collected on a C18 nucleosil column equili-
brated in an aqueous solution containing 0.05% TFA (solution
A). The eluent (solution B) contains 80% acetonitrile and 0.05%
TFA. After sample injection, the column is washed with
solution A for 5 min and then a 0-100% B 30 min gradient is
performed at a flow rate of 1 mL/min. Compounds are de-
tected at 215 nm. Characterization of compounds dissolved in
DMSO-d6 was also done with 1H and 13C NMR spectra
recorded at 300 K on a 300 MHz Bruker spectrometer equipped
with a BBI probehead.
General Procedure (A) for Synthesis of Compounds
1 to 11. Removal of Fmoc Protecting Group. Cleavage of
the Fmoc group was performed with 20% piperidine in DMF
(v/v). The resin beads were mixed with the previous solution
(3 mL for 300 mg resin), and the reaction was performed at
room temperature for 25 min. The solution was removed by
filtration under vacuum, and the resin beads were extensively
washed with DMF.
Coupling of Amino Acids. A mixture of 3 equiv of Fmoc
protected amino acid, 3 equiv of HOBt, 3 equiv of HBTU
dissolved in DMF, and 9 equiv of DIPEA were added to the
resin beads. The reaction was performed at room temperature
for 1 h. The solution was removed by filtration under vacuum,
and the resin beads were extensively washed with DMF.
ible overexpression of Pin1. This cellular assay is based
on the monitoring of cyclin D1 level as described.30 In
that model, juglone is used as a reference compound for
the inhibition of Pin1 function and induces a marked
diminution of the intracellular concentration of cyclin
D1. Our results with WW ligands are at this moment
very preliminary. Compound 1g has produced the same
effect as juglone at similar concentrations and poses the
intriguing question whether it acts directly on Pin1 or
also on other target(s). Further assays will be performed
in the future to identify the cellular target(s) of our
compounds.
Conclusion
Our NMR-based strategy has led to the conception,
synthesis, and testing of new ligands for the Pin1 WW
domain. Whereas the proline in the initial phospho-
Thr-Pro dipeptide could be successfully replaced by a
3,4-dehydro- or 4-trans-fluoroproline, the specificity for
the phosphate moiety proved more challenging. On the
basis of the structure of the WW domain, we extended
these ligands with an N-terminal aromatic group,
resulting in a new series of ligands with low micromolar
affinity. NMR proved crucial throughout this develop-
ment, as a technique to bridge the gap between very
low affinity hit compounds and conventional tools to
measure affinity of small molecules to protein, such as

Discovery of Potent Pin1 WW Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15 4821
Table 3. Structures of Hydrophobic Moieties and Dissociation Constants of Resulting Ligands Based on Proline (1b-1g) or on
3,4-Dehydroproline (9d, 9g, 9h) Derivatives
Figure 4. NMR chemical shift mapping of the 15N-labeled WW domain. Comparison of the mapping of the WW domain in the
presence of 20 equiv of compound 1a in light gray and (A) compound 9a or (B) compound 1g in dark gray. Arrows along the x-axis
indicate â-sheet structural elements. In the structure, the ribbon is darkened as a function of chemical shift variations of compound
9a (A) and compound 1g (B). Amino acids that display a variation superior to 0.3 ppm are labeled.
Amine Acetylation. Five equivalents of acetic anhydride
was dissolved in DMF, and 10 equiv of DIPEA was added. The
solution was mixed with the resin beads, and the mixture was
agitated at room temperature for 30 min. The solution was
removed by filtration under vacuum, and the resin beads were
extensively washed with DMF.
Coupling of Hydrophobic Moieties. A mixture contain-
ing the carboxylic acid derivative (5 equiv), HOBt (5 equiv),
and HBTU (5 equiv) and DIPEA (10 equiv) in DMF was added
to the resin beads, and the coupling reaction was performed
at room temperature for 1 h. The solution was removed by
filtration under vacuum, and the resin beads were extensively
washed with DMF.
TFA Cleavage of Resin-Bound Compounds and Benzyl
Protecting Group. Resin beads were extensively washed
with DMF, DCM, and 10% TFA in DCM and then mixed with
a TFA solution containing 2.5% triisopropylsilane (v/v) and
2.5% water (v/v) (5 mL for 300 mg of resin). The heterogeneous
mixture was placed under agitation for 5 h at room temper-
ature. The solution was then filtered under vacuum, and the

4822 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15
Smet et al.
resin was washed twice with TFA (2.5 mL each). The combined
filtrates were collected in 200 mL of a cold mixture of ether/
pentane (v/v). A white precipitate was formed, and the
precipitation was achieved at -20 °C overnight. The precipi-
tate was isolated by filtration, washed with ether, and then
dissolved in water and lyophilized.
pET28a plasmid under control of T7 promoter with a sequence
encoding a N-terminus His6 tag. Cells were grown at 37 °C in
a LB medium containing kanamycin (20 mg/L) until O.D. at
600 nm reached about 0.6, allowing a rapid growth to high
cell density. Then cells were harvested by centrifugation. The
pellet was resuspended in a M9 medium containing 4 g/L
glucose, 1 g/L 15NH4Cl, 1 mM MgSO4, 20 mg/L kanamycin,
and MEM vitamin cocktail (Sigma) and incubated at 37 °C
for 1 h. Induction was performed with 0.5 mM IPTG (at final
concentration) at 31 °C for 3 h. Cells were harvested by
centrifugation, and the pellet was resuspended in 50 mM
Na2HPO4/NaH2PO4 pH 8.0, 300 mM NaCl buffer (buffer A)
containing 10 mM imidazole, 1 mM DTT, 0.1% NP40 and
complemented with a protease inhibitor cocktail (Roche). The
soluble lysate was obtained by sonication and centrifugation.
The soluble extract was charged on a nickel-nitrilotriacetic
affinity column (Chelating Sepharose Fast Flow, Amersham).
Unbound proteins were washed with 20 mM imidazole in
buffer A, and the WW domain was eluted with 250 mM
imidazole in buffer A (pH 7.8). The protein was then purified
by RP-HPLC on a Source 15RPC column (Amersham) equili-
brated in a 0.05% TFA aqueous solution (solution A). Separa-
tion of proteins was carried out at room temperature at a flow
rate of 1 mL/min using an acetonitrile linear 20 min gradient
from 0 to 60% solution B (85% acetonitrile:15% water-0.05%
TFA). Homogeneous fractions as checked by SDS-polyacryl-
amide gel electrophoresis were pooled and lyophilized.
NMR Screening. NMR experiments were performed on a
Bruker DMX 600 MHz spectrometer at 293 K equipped with
a cryogenic probehead. 15N-labeled WW was dissolved in 25
mM deuterated Tris-HCl (pH ) 6.8) aqueous buffer containing
50 mM NaCl, 5 mM DTT, and 2.5 mM EDTA and 10% D2O.
The spectra were referenced to the signal of sodium 3-tri-
methylsilyl-d(3,3′,2,2′)-propionate (TSPA-d4). A solution of
15N-labeled WW at 200 µM (60 µL) was added to a 2 mM
solution of ligand (120 µL) to reach final concentrations (in
600 µL) of 20 µM WW and 400 µM ligand in the NMR sample
buffer.
General Procedure (B) for Synthesis of Compounds
12 to 15. Coupling of Carboxylic Acid Derivatives on
Prolinamide. Three equivalents of carboxylic acid derivative
(N-R-Fmoc-ornithine/lysine(Boc) or N-Boc-â-alanine) was dis-
solved in DCM (15.8 mmol in 8 mL) and added to a solution
of 1.5 equiv of DCC in DCM at 0 °C (7.9 mmol in 2 mL). The
precipitate was filtered off, and the filtrate was added to a
solution of 1 equiv of prolinamide in DCM (5.25 mmol in 3
mL). The reaction mixture was stirred for 2 h at 0 °C. The
crude mixture was evaporated and dissolved in DCM. The
precipitate was removed by filtration, and the organic layer
was washed twice with a saturated NaHCO3 aqueous solution
and then dried over MgSO4 and evaporated. Products were
isolated by purification on silica column in DCM/EtOH 95/5.
Removal of Boc Protecting Group. One equivalent of
the Boc protected compound was dissolved in DCM (1.37 mmol
in 14 mL), and the mixture was cooled in an ice bath. A
solution of 4 N HCl in dioxane was added dropwise to the
previous solution at 0 °C. The mixture was stirred at room
temperature for 3 h, and then the crude mixture was evapo-
rated. The deprotected product is obtained in quantitative
yield.
Coupling of Diethylsquarate. A solution of 1.2 equiv of
diethylsquarate in ethanol (1.64 mmol in 5 mL) was prepared.
One equivalent of â-alanine or N-R-Fmoc-ornithine/lysine
coupled to prolinamide was dissolved in ethanol (1.37 mmol
in 8 mL), and 1.5 equiv of triethylamine was added. The
mixture was added dropwise to the diethylsquarate solution,
and the reaction was performed at room temperature for 12
h. The crude mixture was evaporated, dissolved in DCM, and
purified on a silica column in DCM/MeOH 9/1.
Removal of Ethyl Protecting Group. The ethylsquarate
derivative was dissolved in THF (1.084 mmol in 6 mL) and
stirred with a solution of 4 N HCl in THF (12 mmol in 3 mL)
at room temperature for 20 h. The crude mixture was
evaporated. The powder was dissolved in DCM and extracted
four times with a 0.3 N solution of ammonium carbonate. The
aqueous solution was lyophilized.
Coupling of Trifluoromethylsulfochloride. Triethyl-
amine (1.2 equiv, 1.044 mmol, 146 µL) was added to 1 equiv
of â-alanine coupled to prolinamide dissolved in DCM. Tri-
fluorosulfonyl chloride (1.2 equiv, 1.044 mmol, 110 µL) was
dissolved in 10 mL of DCM and added to the previous solution
dropwise. The mixture was stirred at room temperature for
48 h. The crude mixture was evaporated, dissolved in DCM,
and purified on a silica column in DCM/EtOH 8/2.
Peptide Synthesis (for IC50 Determination). The pep-
tide sequence was isolated from the Tau protein around the
Thr212-Pro213 epitope as described.26 The phosphorylated
peptide was obtained by solid phase synthesis with introduc-
tion of selectively phosphorylated threonine using appropriate
building block (Novabiochem). The dansyl fluorescent probe
(Sigma) was introduced selectively at the N-terminus by
reaction of dansyl chloride (5 equiv) and DIPEA (10 equiv) in
DMF and followed by a TFA cleavage. Peptide was isolated
by precipitation of the TFA solution in 200 mL of a cold
mixture of ether/heptane (v/v) and centrifugation. The peptide
was dissolved in water and purified by RP-HPLC on a C18
nucleosil column equilibrated in a 0.05% TFA aqueous solution
(solution A). Separation of peptides in the crude mixture was
carried out at 50 °C at a flow rate of 2 mL/min using an
acetonitrile linear 60 min gradient from 0 to 50% solution B
(80% acetonitrile:20% water-0.05% TFA). Homogeneous frac-
tions as checked by RP-HPLC on a C18 nucleosil column (0 to
100% B linear 30 min gradient) and MALDI-TOF MS were
pooled and lyophilized.
NMR Dissociation Constant Calculation. Lyophilized
aliquots of ligand were prepared from a concentrated solution
(2 mM) to get a concentration range of 0.025, 0.05, 0.1, 0.15,
0.25, 0.5, and 1 mM (at final concentration in 600 µL of protein
solution). A solution of 50 µM 15N-labeled WW was added to
lyophilized aliquots of ligands. Quantification of chemical shift
perturbations was calculated according to eq 1.
∆δ(ppm) ) ∆δ(1H)2 + 0.2∆δ(15N)2
(1)
x
∆δ(1H) and ∆δ(15N) are the chemical shift variations of amide
proton (HN) and amide nitrogen (NH), respectively, of WW
domain residues the upon ligand addition.
Dissociation constants were calculated by fitting on a
graphical representation the theoretical curve obtained from
eq 2 with experimental points.
KD
∆δ(ppm) ) 0.5∆δmax 1 + X +
-
(
[protein]0
2
KD
1 + X +
- 4X (2)
)
(
x
)
[protein]0
X is the molar ratio of ligand on protein, [protein]0 the initial
concentration of protein, and ∆δ the chemical shift variation
calculated using eq 1. ∆δmax was set at the ∆δ value at the
protein saturation level.
FRET Competition Experiments for IC50 Determina-
tion. FRET experiments were achieved on a PTI fluorescence
spectrometer by exciting tryptophan residues of WW domain
at 295 nm. As for NMR titrations, a 20 µM solution of WW in
protonated Tris 50mM pH 6.8 buffer containing 50 mM NaCl
and 5 mM DTT/EDTA and various concentrations of dansy-
lated peptide were added on lyophilized aliquots of ligand. The
dissociation constant of the WW-peptide complex was evalu-
Protein Expression and Purification. The WW domain
was produced in E. coli BL21(DE3) strain, carrying the

Discovery of Potent Pin1 WW Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 15 4823
ated to 100 µM according to eq 2 by replacing ∆δ with ∆I334
,
(6) Wulf, G.; Ryo, A.; Liou, Y. C.; Lu, K. P. The prolyl isomerase
Pin1 in breast development and cancer. Breast Cancer Res. 2003,
5, 76-82.
the intensity variation of the fluorescence emission signal
measured at 334 nm.26 IC50 determination was calculated as
described for dissociation constants determined by NMR with
eq 2 by replacing ∆δ with the inhibition percentage given by
eq 3 and KD by IC50.
(7) Wulf, G. M.; Ryo, A.; Wulf, G. G.; Lee, S. W.; Niu, T.; et al. Pin1
is overexpressed in breast cancer and cooperates with Ras
signaling in increasing the transcriptional activity of c-Jun
towards cyclin D1. Embo J. 2001, 20, 3459-3472.
(8) Lu, K. P. Prolyl isomerase Pin1 as a molecular target for cancer
diagnostics and therapeutics. Cancer Cell 2003, 4, 175-180.
(9) Lu, P. J.; Wulf, G.; Zhou, X. Z.; Davies, P.; Lu, K. P. The prolyl
isomerase Pin1 restores the function of Alzheimer-associated
phosphorylated tau protein. Nature 1999, 399, 784-788.
(10) Uchida, T.; Takamiya, M.; Takahashi, M.; Miyashita, H.; Ikeda,
H. et al. Pin1 and Par14 peptidyl prolyl isomerase inhibitors
block cell proliferation. Chem. Biol. 2003, 10, 15-24.
(11) Hennig, L.; Christner, C.; Kipping, M.; Schelbert, B.; Rucknagel,
K. P.; et al. Selective inactivation of parvulin-like peptidyl-prolyl
cis/trans isomerases by juglone. Biochemistry 1998, 37, 5953-
5960.
I
334 - I3034
% inhibition )
(3)
I3034 - I334
100
I334 is the fluorescence intensity at 334 nm upon ligand
addition, I 3034, the fluorescence intensity without inhibition,
and I 313040, the fluorescence intensity at maximal inhibition.
(12) Wang, X. J.; Xu, B.; Mullins, A. B.; Neiler, F. K.; Etzkorn, F. A.
Conformationally locked isostere of phosphoSer-cis-Pro inhibits
Pin1 23-fold better than phosphoSer-trans-Pro isostere. J. Am.
Chem. Soc. 2004, 126, 15533-15542.
Acknowledgment. The authors thank the Institute
for the Study Of Aging (ISOA, the United States) for
financial support. We thank Professor Andre´ Tartar for
fruitful discussions, Ge´rard Montagne, Herve´ Drobecq,
and Eric Diesis for their technical contribution to this
work and Dr. Dragos Horvath for molecular modeling.
The 600 MHz facility used in this study was funded by
the Re´gion Nord-Pas de Calais (France), the CNRS, and
the Institut Pasteur de Lille.
(13) Lu, P. J.; Zhou, X. Z.; Shen, M.; Lu, K. P. Function of WW
domains as phosphoserine- or phosphothreonine-binding mod-
ules. Science 1999, 283, 1325-1328.
(14) Lu, P. J.; Zhou, X. Z.; Liou, Y. C.; Noel, J. P.; Lu, K. P. Critical
role of WW domain phosphorylation in regulating phosphoserine
binding activity and Pin1 function. J. Biol. Chem. 2002, 277,
2381-2384.
(15) Verdecia, M. A.; Bowman, M. E.; Lu, K. P.; Hunter, T.; Noel, J.
P. Structural basis for phosphoserine-proline recognition by
group IV WW domains. Nat. Struct. Biol. 2000, 7, 639-643.
(16) Wintjens, R.; Wieruszeski, J. M.; Drobecq, H.; Rousselot-Pailley,
P.; Buee, L. et al. 1H NMR study on the binding of Pin1 Trp-
Trp domain with phosphothreonine peptides. J. Biol. Chem.
2001, 276, 25150-25156.
(17) Songyang, Z.; Shoelson, S. E.; Chaudhuri, M.; Gish, G.; Pawson,
T. et al. SH2 domains recognize specific phosphopeptide se-
quences. Cell 1993, 72, 767-778.
(18) Yan, K. S.; Kuti, M.; Zhou, M.-M. PTB or not PTB- that is the
question. FEBS Lett. 2002, 73, 67-70.
(19) Yaffe, M. B.; Elia, A. E. H. Phosphoserine/threonine-binding
domains. Curr.Opin. Cell Biol. 2001, 13, 131-138.
(20) Macias, M. J.; Wiesner, S.; Sudol, M. WW and SH3 domains,
two different scaffolds to recognize proline-rich ligands. FEBS
Lett. 2002, 513, 30-37.
(21) Kay, B. K.; Williamson, M. P.; Sudol, M. The importance of being
proline: the interaction of proline-rich motifs in signaling
proteins with their cognate domains. FASEB J. 2000, 14, 231-
241.
(22) Burke, T. R.; Lee, K. Phosphotyrosyl mimetics in the develop-
ment of signal transduction inhibitors. Acc. Chem. Res. 2003,
36, 426-433.
(23) Jeannotte, G.; Lubell, W. D. Large Structural Modification with
Conserved Conformation: Analysis of Delta(3)-Fused Aryl Pro-
lines in Model beta-Turns. J. Am. Chem. Soc. 2004, 126, 14334-
14335.
(24) Grellier, P.; Vendeville, S.; Joyeau, R.; Bastos, I. M.; Drobecq,
H. et al. Trypanosoma cruzi prolyl oligopeptidase Tc80 is
involved in nonphagocytic mammalian cell invasion by trypo-
mastigotes. J. Biol. Chem. 2001, 276, 47078-47086.
(25) Willand, N.; Joossens, J.; Gesquie`re, J.-C.; Tartar, A. L.; Evans,
M. D. et al. Solid and solution phase synthesis of the 2-cyan-
opyrrolidide DPP-IV inhibitor NVP-DPP728. Tetrahedron 2002,
58, 5741-5746.
(26) Smet, C.; Sambo, A. V.; Wieruszeski, J. M.; Leroy, A.; Landrieu,
I.; et al. The peptidyl prolyl cis/trans-isomerase Pin1 recognizes
the phospho-Thr212-Pro213 site on Tau. Biochemistry 2004, 43,
2032-2040.
(27) Widmer, H.; Jahnke, W. Protein NMR in biomedical research.
Cell. Mol. Life Sci. 2004, 61, 580-599.
(28) Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. Discovering high-
affinity ligands for proteins. Science 1997, 278, 497, 499.
(29) Hajduk, P. J.; Gerfin, T.; Boehlen, J. M.; Haberli, M.; Marek,
D.; et al. High-throughput nuclear magnetic resonance-based
screening. J. Med. Chem. 1999, 42, 2315-2317.
(30) Hamdane, M.; Smet, C.; Sambo, A. V.; Leroy, A.; Wieruszeski,
J. M.; et al. Pin1: a therapeutic target in Alzheimer neurode-
generation. J. Mol. Neurosci. 2002, 19, 275-287.
Appendix
Abbreviations. MBHA, 4-methylbenzhydrylamine;
HOBt, 1-hydroxybenzotriazole; HBTU, 2-(1H-benzotri-
azol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophos-
phate; Fmoc, 9-fluorenylmethoxy-carbonyl; KD, dissocia-
tion constant; IC50, inhibition constant at 50%; Boc, tert-
butoxycarbonyl; Bzl, benzyl; DCM, dichloromethane;
DMF, dimethylformamide; EtOH, ethanol; DMSO, di-
methyl sulfoxide; DIPEA, diisopropylethylamine; TFA,
trifluoroacetic acid; DTT, dithiothreitol; EDTA, ethyl-
enediaminetetraacetic acid; HSQC, heteronuclear sim-
ple quantum correlation; Tris, 2-amino-2-hydroxy-
methyl-1,3-propanediol; BBI, inverse broad band (probe);
MALDI-TOF, matrix assisted laser desorption ioniza-
tion time-of-flight; RP-HPLC, reverse phase high pres-
sure liquid chromatography; SDS-PAGE, sodium do-
decyl sulfate-polyacrylamide gel electrophoresis.
Supporting Information Available: Synthetic proce-
dures and spectral characterization of compounds 1a to 15.
This material is available free of charge via the Internet at
http://pubs.acs.org.
References
(1) Pawson, T.; Raina, M.; Nash, P. Interaction domains: from
simple binding events to complex cellular behavior. FEBS Lett.
2002, 513, 2-10.
(2) Yaffe, M. B.; Smerdon, S. J. The Use of in Vitro Peptide-Library
Screens in the Analysis of Phosphoserine/Threonine-Binding
Domain Structure and Function. Annu. Rev. Biophys. Biomol.
Struct. 2004, 33, 225-244.
(3) Ranganathan, R.; Lu, K. P.; Hunter, T.; Noel, J. P. Structural
and functional analysis of the mitotic rotamase Pin1 suggests
substrate recognition is phosphorylation dependent. Cell 1997,
89, 875-886.
(4) Lu, K. P.; Hanes, S. D.; Hunter, T. A human peptidyl-prolyl
isomerase essential for regulation of mitosis. Nature 1996, 380,
544-547.
(5) Winkler, K. E.; Swenson, K. I.; Kornbluth, S.; Means, A. R.
Requirement of the prolyl isomerase Pin1 for the replication
checkpoint. Science 2000, 287, 1644-1647.
JM0500119