An optimised small-molecule stabiliser of the 14-3-3-PMA2 protein-protein interaction

Article abstract of DOI:10.1002/chem.201103761

Modulation of protein-protein interactions (PPIs) is a highly demanding, but also a very promising approach in chemical biology and targeted drug discovery. In contrast to inhibiting PPIs with small, chemically tractable molecules, stabilisation of these interactions can only be achieved with complex natural products, like rapamycin, FK506, taxol, forskolin, brefeldin and fusicoccin. Fusicoccin stabilises the activatory complex of the plant H ⁺-ATPase PMA2 and 14-3-3 proteins. Recently, we have shown that the stabilising effect of fusicoccin could be mimicked by a trisubstituted pyrrolinone (pyrrolidone1, 1). Here, we report the synthesis, functional activity and crystal structure of derivatives of 1 that stabilise the 14-3-3-PMA2 complex. With a limited compound collection three modifications that are important for activity enhancement could be determined: 1) conversion of the pyrrolinone scaffold into a pyrazole, 2) introduction of a tetrazole moiety to the phenyl ring that contacts PMA2, and 3) addition of a bromine to the phenyl ring that exclusively contacts the 14-3-3 protein. The crystal structure of a pyrazole derivative of 1 in complex with 14-3-3 and PMA2 revealed that the more rigid core of this molecule positions the stabiliser deeper into the rim of the interface, enlarging especially the contact surface to PMA2. Combination of the aforementioned features gave rise to a molecule (37) that displays a threefold increase in stabilising the 14-3-3-PMA2 complex over 1. Compound 37 and the other active derivatives show no effect on two other important 14-3-3 protein-protein interactions, that is, with CRaf and p53. This is the first study that describes the successful optimisation of a PPI stabiliser identified by screening. Copyright

Full text of DOI:10.1002/chem.201103761

FULL PAPER
DOI: 10.1002/chem.201103761
An Optimised Small-Molecule Stabiliser of the 14-3-3–PMA2 Protein–
Protein Interaction
Anja Richter,[b] Rolf Rose,[a] Christian Hedberg,[b] Herbert Waldmann,*[b, c] and
Christian Ottmann*[a]
Abstract: Modulation of protein–pro-
tein interactions (PPIs) is a highly de-
manding, but also a very promising ap-
proach in chemical biology and target-
ed drug discovery. In contrast to inhib-
iting PPIs with small, chemically tracta-
ble molecules, stabilisation of these
interactions can only be achieved with
complex natural products, like rapamy-
cin, FK506, taxol, forskolin, brefeldin
and fusicoccin. Fusicoccin stabilises the
activatory complex of the plant H+
-ATPase PMA2 and 14-3-3 proteins.
Recently, we have shown that the sta-
bilising effect of fusicoccin could be
mimicked by a trisubstituted pyrroli-
none (pyrrolidone1, 1). Here, we
report the synthesis, functional activity
and crystal structure of derivatives of
1 that stabilise the 14-3-3–PMA2 com-
plex. With a limited compound collec-
tion three modifications that are impor-
tant for activity enhancement could be
determined: 1) conversion of the pyrro-
linone scaffold into a pyrazole, 2) intro-
duction of a tetrazole moiety to the
phenyl ring that contacts PMA2, and
3) addition of a bromine to the phenyl
ring that exclusively contacts the 14-3-3
protein. The crystal structure of a pyra-
zole derivative of 1 in complex with 14-
3-3 and PMA2 revealed that the more
rigid core of this molecule positions the
stabiliser deeper into the rim of the in-
terface, enlarging especially the contact
surface to PMA2. Combination of the
aforementioned features gave rise to
a molecule (37) that displays a threefold
increase in stabilising the 14-3-3–
PMA2 complex over 1. Compound 37
and the other active derivatives show
no effect on two other important 14-3-
3 protein–protein interactions, that is,
with CRaf and p53. This is the first
study that describes the successful opti-
misation of a PPI stabiliser identified
by screening.
Keywords: chemical synthesis
·
crystal structures · protein–protein
interactions · structure–activity rela-
tionships · X-ray crystallography
Introduction
karyotic adapter proteins influence a plethora of physiologi-
cal processes by binding to hundreds of diverse protein part-
ners.[2,3] They bind their targets through short phosphorylat-
ed sequences and modulate their partnersꢀ subcellular locali-
sation, enzymatic activity or their ability to interact with
further proteins.[3–5] The seven human isoforms (b, g, e, h, s,
t, and z) regulate the activity of, for example, the kinase
CRaf,[6,7] the cell-cycle phosphatase Cdc25,[8,9] the transcrip-
tional modulator YAP,[10,11] and stabilise the tumour suppres-
sor p53.[12,13] The 14-3-3 proteins have been implicated in
a variety of human diseases. In addition to their participa-
tion in diverse cancers,[14] they have been associated with the
development of neurodegenerative diseases[15] or virulence
of human pathogenic organisms.[16,17]
Targeted drug discovery mainly concentrates on the devel-
opment of enzyme active site inhibitors. Although the
highly promising field of protein–protein interaction (PPI)
modulation has been elaborated with molecules that disrupt
PPIs,[1] the reverse strategy to stabilise PPIs with small,
drug-like molecules is underrepresented in medicinal
chemistry research.
An interesting case of PPIs is typified by the 14-3-3 pro-
tein family. These highly conserved, small (25–30 kDa) eu-
[a] Dr. R. Rose,+ Dr. C. Ottmann
Chemical Genomics Centre of the Max-Planck-Society
Otto-Hahn-Strasse 15, 44227 Dortmund (Germany)
E-mail: christian.ottmann@cgc.mpg.de
The natural product fusicoccin A (FC-A)[18] produced by
the phytopathogenic fungus Phomopsis amygdali is a natural
modulator of a 14-3-3 protein–protein interaction and tar-
gets the complex of the plant plasma membrane H+
-ATPase (PMA) and 14-3-3 proteins.[19] FC-A binds to the
interface of the activated complex and enhances the affinity
of the two protein partners approximately 90-fold.[20] This
stabilisation induces permanent activation of the proton
pump, which translates into the maximal opening of the
plantsꢀ gas-exchanging stomatal pores and subsequent wilt-
ing.[21] Hence, FC-A is a potent herbicidal compound, but its
complex chemical structure and the potentially high costs of
[b] Dr. A. Richter,+ Dr. C. Hedberg, Prof. Dr. H. Waldmann
Department of Chemical Biology
Max Planck Institute of Molecular Physiology
Otto-Hahn-Strasse 11, 44227 Dortmund (Germany)
E-mail: herbert.waldmann@mpi-dortmund.mpg.de
[c] Prof. Dr. H. Waldmann
Fakultꢁt Chemie, Fachbereich Chemische Biologie
Technische Universitꢁt Dortmund
Otto-Hahn-Strasse 6, 44227 Dortmund (Germany)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103761.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

Table 1. Derivatives of 1 prepared by Doebner condensation.
production most likely would prevent a successful practical
application. Therefore, by screening a compound library we
set out to identify synthetically more accessible molecules
that show stabilising activity towards 14-3-3–PMA.
Cmp[a]
R1
R2
R3
R4
Yield[b] [%]
1
2
3
4
5
6
7
8
OH
OH
H
COOH
OH
H
COOH
H
OH
H
COOH
H
OH
COOH
H
H
OH
H
COOH
H
H
COOH
H
H
H
H
OH
H
COOH
H
COOH
COOH
COOH
H
COOH
COOH
H
tetrazo-5-yl
COOH
COOH
H
NO2
H
H
H
H
COOH
COOH
COOH
NO2
NO2
NO2
6
11
9
12
56
33
9
26
25
30
25
18
32
64
58
42
38
27
23
21
7
As a result we recently reported that epibestatins and pyr-
rolinones stabilise the PMA (isoform 2)–14-3-3 interac-
tion.[22] The crystal structures of the corresponding ternary
complexes could be solved and explained the different dy-
namics of their mode of action. These molecules bind to the
rim of the interface of 14-3-3 and PMA2 establishing con-
tacts to both protein partners and show promising physio-
logical activity in a stomata-opening assay. Consequently,
they might be good starting points for the development of
novel herbicides with an as yet not explored mode of action.
Here, we report the synthesis of a pyrrolinone collection
starting from the initial screening hit 1.[22] We determined
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
Br
Br
Br
Br
Br
Br
H
H
H
NO2
9
COOH
COOH
COOH
COOH
thiazo-2-yl
thiazo-2-yl
NO2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
H
COOH
H
COOH
tetrazo-5-yl
COOH
COOH
H
tetrazo-5-yl
COOH
H
tetrazo-5-yl
tetrazo-5-yl
COOH
tetrazo-5-yl
COOH
COOH
H
NO2
COOH
COOH
COOH
COOH
thiazo-2-yl
thiazo-2-yl
thiazo-2-yl
NO2
thiazo-2-yl
thiazo-2-yl
thiazo-2-yl
thiazo-2-yl
thiazo-2-yl
thiazo-2-yl
NO2
16
5
H
OMe
OMe
Br
Br
Br
45
29
34
19
33
32
20
tetrazo-5-yl
H
Br
H
COOH
their binding mode by solving the crystal structure of one
compound of this collection in complex with 14-3-3 and the
last C-terminal 30 amino acids of PMA2 harbouring a phos-
phothreonine-mimicking aspartate as penultimate residue
(PMA2-CT30YDI). One pyrazole analogue of the guiding
heterocycle showed enhanced stabilising activity towards the
14-3-3–PMA2 complex and is selective for this 14-3-3 PPI.
[a] All compounds were at least 80% pure; [b] chemical synthesis yield.
Table 2. Pyrazole derivatives of 1.
Cmp[a]
R1
R2
R3
R4
Yield[b] [%]
32
33
34
35
36
37
OH
OH
COOH
OH
COOH
H
COOH
COOH
H
COOH
H
NO2
Br
NO2
thiazo-2-yl
NO2
H
H
H
H
OMe
Br
72
75
77
78
77
71
Results and Discussion
tetrazo-5-yl
NO2
Synthesis of derivatives of 1: In order to identify possible
chemical derivatisations that enhance the stabilising activity
of the initial hit 1 a small focused compound library of pyr-
rolinones and pyrazoles with different substitution patterns
was prepared (Tables 1 and 2). In detail, the hydroxyl group
at R1 was either preserved (1, 2, 5, 9, 13, 17, 27, 32, 33, 35),
changed to a carboxyl (4, 7, 11, 14, 19, 22, 29, 31, 34, 36) or
omitted (3, 6, 8, 10, 12, 15, 16, 18, 20, 21, 23–26, 28, 30, 37).
The carboxyl group at R2 was kept in 18 out of 37 deriva-
tives. In 8, 16, 20, 23, 24, 26, 30, and 37 it was replaced by
a tetrazole moiety, and in 4, 7, 11, 12, 14, 19, 22, 29, 31, 34,
and 36 it was changed to a proton. For R3 the nitro group
was replaced by carboxylic acid (2–4, 9–12, and 17–20), thia-
zole (13, 14, 21–23, 25–30, and 35) or bromine (33). At R4
either a methoxy substituent (5–14, 25, 26, and 36) or bro-
mine (15–20, 27–30, and 37) was introduced. Conversion of
the corresponding pyrrolinone into a pyrazole rigidifies the
molecule, potentially enhancing binding affinity by reducing
the entropy penalty upon binding to the protein complex.
[a] All compounds were at least 80% pure; [b] chemical synthesis yield.
Initially all pyrrolinones listed in Table 1 and all pyrazoles
listed in Table 2 entries 1–5 were prepared.
Derivatives of 1 stabilise the 14-3-3–PMA2 interaction: To
investigate the stabilising activity of the synthesised com-
pound collection we employed a surface-based assay. In
brief, the binding of green fluorescent protein (GFP)-tagged
14-3-3 to surface-immobilised PMA2-CT66 (the C-terminal
66 amino acids of PMA2) fused with gluthatione S-transfer-
ase (GST) was monitored in the presence of the candidate
compounds. To compare the stabilising activity of the tested
compounds all fluorescence values were normalised to the
activity of 1, which was set to 100% (Figure 1).
Of the alterations described above (Tables 1 and 2), all
were detrimental in terms of stabilising activity with the ex-
ception of the pyrazole derivative (32) of 1 (Figure 1).
&
2
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Small-Molecule Stabiliser of Protein–Protein Interaction
FULL PAPER
Figure 1. Compounds stabilising the PMA2–14-3-3 interaction. Structures and activities of compounds from the library of derivatives of 1. The stabilising
activities of the candidate molecules were normalised to the HTS-Hit 1. Binding of the GFP-fused tobacco 14-3-3 protein T14-3e to immobilised PMA2-
CT66YDI in the presence of the compounds was measured by using a surface-based assay (see the Experimental Section). NC: no compound control.
Nonetheless, compounds 34, 15, 16, and 24 showed normal-
ised activities above 50% of 1. Notably, of these, compound
34, like 32, is also a pyrazole. The other derivatives that still
showed activity were characterised by either a bromine at
R4 (15 and 16) or/and a tetrazole moiety at R2 (16 and 24).
Of these compounds EC50 values were determined
(Figure 2). As expected from the single concentration meas-
urements these values were quite similar, ranging from 286-
A
E
U
(Æ82) mm (34). In order to elucidate the mechanistic
basis of the influence of the chemical modifications on the
protein complex, we pursued the co-crystallisation of the
compounds with 14-3-3e and PMA2-CT30YDI and succeed-
ed in the case of 34.
Figure 2. Activity of PMA2–14-3-3 stabilising compounds compared to 1.
Binding of GFP–T14-3c to immobilised GST–CT52YDI in response to
the concentration of the stabilising molecules. For comparison the bind-
ing curve for fusicoccin is shown in black.
Structure of 34 in complex with T14-3e–CT30YDI: Crystals
of 34 in complex with the tobacco 14-3-3 protein T14-3e and
the PMA2 C-terminal construct CT30YDI appeared after
5–7 days in CHES (0.1m, pH 9.5), LiCl (1.0m) and sucrose
(30% w/v) at 48C and grew to dimensions of 500ꢃ500ꢃ
300 mm within two weeks. They diffracted to 3.4 ꢄ and the
electron density allowed building of 225 out of 231 residues
of T14-3e and the complete 30 residues of CT30. Details of
the statistics of the crystal structure are summarised in
Table 3.
Despite the rather low resolution, the solvent flattened
electron density for 34 was good (Figure 3a, b) and allowed
the determination of the binding of 34 to the complex of
T14-3e and PMA2-CT30YDI (Figure 3c, d, f). The 14-3-3
protein displayed the typical dimeric form with one PMA2-
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

C. Ottmann, H. Waldmann et al.
Table 3. Crystallographic data.[a] (PDB ID: 4DX0)
cell dimensions [ꢄ]
CT30YDI peptide bound to each of the monomers (Fig-
ure 3c).
98.450, 98.450,
216.710
The pyrazole 34 occupies a pocket that is concomitantly
formed by both proteins (Figure 3d, f) and largely matches
that of 1 (Figure 3e). As with 1 most of the contact surface
is contributed by the 14-3-3 protein; this might explain the
rather weak activity of both compounds with respect to sta-
bilising the T14-3e–PMA2-CT30YDI complex. Nevertheless,
in comparison with 1, 34 is localised nearer to a narrow cleft
between T14-3e and PMA2-CT30YDI (Figure 3 f). This po-
sitions the molecule deeper into the rim of the protein–pro-
tein interface, which is desirable for a PPI stabilising com-
pound. The fact that 34 is weaker than 1 in stabilising the
14-3-3–PMA2 complex might be due to the carboxylate at
R1 and a missing substituent at R2. Compound 32, for exam-
ple, stabilises the complex more strongly and bears a hydrox-
yl group at R1 and a carboxylate at R2.
space group
resolution [ꢄ]
observations
unique Reflections
completeness [%]
I/s
Rsym [%]
Rcryst [%]
Rfree [%]
protein atoms
ligand atoms
R.m.s.d. of bond length, count; RMS; weight [ꢄ]
I4122
3.4 (3.6–3.4)
45606 (7284)
7386 (1131)
96.1 (97.6)
13.98 (3.77)
10.3 (43.0)
32.4
36.2
2020
33
2073; 0.007; 0.022
R.m.s.d. of isotropic thermal factor, count; RMS;
weight [ꢄ2]
m.c. bond
m.c. angle
s.c. bond
1248; 0.370; 1.500
2001; 0.676; 2.000
825; 0.436; 3.000
794; 0.791; 4.500
s.c. angle
[a] Data for the outermost shell are shown in parentheses.
Derivatives with enhanced stabilising potency: The structur-
al analysis of 34 in complex with T14-3e–PMA2-CT30YDI
prompted us to elongate 34 in R1 to fill the remaining space
Figure 3. Crystal structure of 34 in complex with T14-3e and PMA2-CT30YDI (PDB ID: 4DX0). a) Structure of 34. b) Final 2Fo-Fc electron density
(blue mesh, contoured at 1s) of 34 (magenta sticks) binding to the interface of T14-3e (green, semitransparent surface) and PMA2-CT30YDI (blue, sem-
itransparent surface). c) Overview of the T14-3e dimer (dark and light-green, semitransparent surface) complexed with PMA2-CT30YDI (dark and
light-blue, semitransparent surface) and 34 (magenta sticks). d) Binding of 34 (magenta sticks) to a T14-3e monomer (green surface) complexed with
PMA2-CT30YDI (blue surface). e) Superimposition of 1 (yellow sticks, PDB ID: 3M51) and 34 (magenta sticks) bound to the T14-3e–PMA2-CT30YDI
complex. f) Binding of 1 (yellow sticks and semitransparent surface) and 34 (magenta sticks and semitransparent surface) to the rim of the interface of
T14-3e (green surface) and PMA2-CT30YDI (blue surface). NC: no compound control.
&
4
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Small-Molecule Stabiliser of Protein–Protein Interaction
FULL PAPER
in the interface of the two proteins (Figure 3 f). This was at-
tempted by the generation of amides (Table S1 in the Sup-
porting Information). The resulting molecules derived from
34 were expected to form additional interactions with the
PMA2-CT30YDI surface (Figure 3 f, blue surface). Howev-
er, none of these amides showed any stabilising effect on
the 14-3-3–PMA2 complex in the surface-based assay (data
not shown). In a second approach we combined all chemical
modifications represented in the initial library that were
compatible with stabilising the 14-3-3–PMA2 interaction.
These are: 1) conversion into a pyrazole, 2) a tetrazole
moiety at R2, and 3) a bromine at R4. This combination gave
rise to molecule 37 (Table 2, entry 6, Figure 4).
Figure 5. Enhanced stabilising activity of 37 compared to 1. Binding of
GFP–T14-3c to immobilised GST–CT52YDI in response to the concen-
tration of the stabilising molecules. For comparison the binding curve for
fusicoccin is shown in black.
ure 6a) and 34 (Figure 6b) bound to the T14-3e–CT30YDI
complex allowed some explanation of the results from the
surface-based stabilisation assay with the compounds from
the first round of synthesis (Figure 1) and provided a possi-
ble rationale for the enhanced activity of 37 (modelled in
Figure 6c). One crucial point in enhancing the stabilisation
the 14-3-3–CT30YDI complex by derivatives of 1 is to
strengthen the binding of the small molecule to CT30YDI.
The extreme C-terminal Ile956 of CT30YDI plays a special
role in this context. In the T14-3c–CT52–FC-A complex the
side chain and the free C terminus of this residue make the
main contacts between CT52YDI and FC-A.[23] Further-
more, the crystal structure of 34 bound to the T14-3e–
CT30YDI binary complex (this study) revealed a relatively
large pocket near Ile956 conjointly formed by both protein
partners. Modelling of 37 into the binding pocket of 34 in
the experimentally solved structure suggests that the tetra-
zole moiety at R2 could at least partially fill this pocket; this
would establish simultaneous contacts to Ile956 of
CT30YDI as well as Ser52, Ile175 and Ile226 of T14-3e (Fig-
ure 6c). Since compounds 15 and 16 showed that bromine at
R4 was compatible with stabilisation of the complex
(Figure 1) we introduced this substitution into 37 (Figure 4).
As the modelling suggests, the hydrophobic contact surface
to Phe126 of T14-3e should be increased by this modifica-
tion (Figure 6c) most probably accountings for a positive
overall effect on the stabilising activity.
Figure 4. The optimised compound 37 stabilises the PMA2–14-3-3 inter-
action more strongly than 1. The stabilising activities of the candidate
molecules were normalised to the HTS-Hit 1. Binding of the tobacco,
GFP-fused 14-3-3 protein, T14-3e, to immobilised PMA2-CT52YDI in
the presence of the compounds was measured by using a surface-based
assay (see the Experimental Section).
Compound 37 was tested in the surface-based assay at
a single concentration (100 mm) and was found to be signifi-
cantly more active than 1 (Figure 4 and Figure 5). Here, es-
pecially interesting is the fact that the pyrazole 37 was about
three-times more active than its pyrrolinone counterpart 16;
this supports the pyrazole scaffold as being more suited than
the pyrrolinone. Another way to compare the compoundꢀs
stabilising activity towards the 14-3-3–PMA2 complex is the
determination of the EC50 (Figure 5). Here, formation and
endurance of the GST–PMA2-CT66–GFP–T14-3c complex
is monitored at different concentrations of compounds. As
a reference we determined the EC50 of the natural product
Specificity of derivatives of 1 for the 14-3-3–PMA2 complex
stabilisation: Since 14-3-3 proteins interact with hundreds of
proteins in eukaryotic organisms,[3] specificity of pharmaco-
logical intervention is an important issue. We employed our
surface-based assay to test a potential stabilising effect of
derivatives of 1 towards the 14-3-3 PPI with CRaf and p53.
Protein 14-3-3 binds to CRaf through the phosphorylation
sites pS233, pS259 and pS621.[6] Protein p53 is bound to 14-
3-3 through its C-terminal phosphorylation sites pS378 and
FC-A to 498
bilising activity of 101
33(Æ4) mm.
N
A
A structural explanation for the enhanced activity of 37
over 1 and 34: Comparison of the crystal structures of 1 (Fig-
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

C. Ottmann, H. Waldmann et al.
Figure 6. Details of the binding modes of 1 and derived molecules. Upper row: stereo figures of the compounds in their protein target pockets. Lower
row: schemes of the binding mode produced with MOE (http://www.chemcomp.com/). a) Compound 1 bound to T14-3e–CT30YDI (PDB ꢅDID: 3M51).
b) Compound 34 bound to T14-3c–CT30YDI (this study). c) Compound 37 modelled into the 34 binding pocket.
pT387.[12] As shown in Figure 7, all six compounds display
a stabilising effect towards binding of 14-3-3 proteins to
GST–PMA2-CT52YDI without showing any interference
the groups of Yao, Botta, and Fu.[26–28] Complementary to in-
hibition, also the small-molecule stabilisation of PPIs is
a feasible approach. Hence, adding to these successes, we
have identified small molecules that are able to specifically
stabilise a 14-3-3 protein–protein interaction.[22] In this at-
tempt to find small molecules stabilising the PMA2–14-3-3
interaction, we have shown that the activity of a complex
natural product (FC-A) can in part be substituted by syn-
thetically much more accessible molecules.
with GST–p53319–393 or GST–CRaf229–268
.
Here we show how an initial primary screen hit (1) could
be used as the starting point to establish a feasible structure-
activity relationship (SAR) for stabilisation of a 14-3-3 pro-
tein–protein interaction. Crystallographic studies showed
that replacement of the pyrrolinone backbone by a pyrazole
orients one of the phenyl substituents closer into the rim of
the 14-3-3–PMA2 interface; this is a promising binding site
not addressed by 1 (Figure 3 f).
Although the potency of these molecules definitely needs
further improvement, their structure provides proof-of-con-
cept for the possibility of developing a rationale for the opti-
misation of small-molecule stabilisers of 14-3-3 PPIs. Apart
from the complex natural products FC-A[18–23] and cotyleni-
n A[30,31] and the recently published epibestatin and pyrroli-
done1[22] small molecule stabilisers of 14-3-3 PPIs have not
been identified.
Figure 7. Specificity of 37 and other derivatives of 1 to the PMA2–14-3-3
protein–protein interaction. The specificity of compounds from this study
was tested by measuring their stabilising potential towards the 14-3-3
protein–protein interaction with the 14-3-3 binding motif from CRaf and
p53, both of which were fused to GST.
The versatile and widespread biological functions of 14-3-
3 proteins have given rise to an increasing interest in this
protein family as a novel target class for pharmacological in-
tervention.[5,14,15,24–28] Since 14-3-3 proteins display no enzy-
matic activity and harbour no active site, a direct pharmaco-
logical approach has to include PPI modulation. Very re-
cently, this possibility became a realistic option when the
first small molecule 14-3-3 PPI inhibitors were reported by
In general, rational design and optimisation of small mol-
ecules with a PPI stabilising mode of action is in its infancy.
Exceptions are the immunosuppressants FK506[32] and rapa-
mycin[33] with their derivatives (AP1903, rapalogues)[34,35]
and other chemically induced dimerisers frequently used in
chemical biology basic research.[36,37] A strategy to success-
&
6
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Small-Molecule Stabiliser of Protein–Protein Interaction
FULL PAPER
scribed before.[23] The C-terminal regions of PMA2 encompassing 30 or
66 amino acids (CT30, amino acids 927–956; CT66, amino acids 890–956)
and characterised by the C-terminal tripeptide YDI were expressed as
intein fusion peptides (IMPACT-CN, NEB) or GST fusions and purified
as described before.[22, 23]
fully and specifically stabilise 14-3-3 PPIs would considera-
bly expand the options available for active ingredient devel-
opment in agriculture and the treatment of human disease,
for example, in oncology, neurodegenerative diseases, cardi-
ovascular diseases, obesity and diabetes.
Surface-based assay: A solution of GST–PMA2-CT66YDI (0.02 mgmLÀ1
)
in Tris/HCl (25 mm, pH 7.5), MgCl2 (2 mm) and NaCl (150 mm) was im-
mobilised for 1 h at room temperature in a Costar 3925 96-well plate in
a volume of 200 mL per well. Wells were blocked with Tris/HCl (25 mm,
pH 7.5, 200 mL), MgCl2 (2 mm), NaCl (150 mm), Tween-20 (0.05%) and
BSA (1% w/v) for 1 h at room temperature followed by four washing
steps at room temperature with Tris/HCl (25 mm, pH 7.5, 200 mL), MgCl2
(2 mm), NaCl (150 mm) and Tween-20 (0.05%). Subsequently, wells were
incubated for 2 h at 48C with Tris/HCl (25 mm, pH 7.5, 200 mL), MgCl2
(2 mm), NaCl (150 mm) and Tween-20 (0.05%) BSA (0.1% w/v) contain-
ing T14-3c–EGFP (0.02 mgmLÀ1) and test compound (100 mm). Four wash-
ing steps were performed as described above and EGFP fluorescence
was measured by using the InfiniteꢆF500 plate reader from Tecan.
Experimental Section
Chemical synthesis: All pyrrolinones were prepared by using a one-pot
Doebner condensation[38] of a methylene component and a Schiff base.
Pyrazoles were obtained by the treatment of the corresponding pyrroli-
nones with hydrazine (Scheme 1). The general procedure for the prepara-
tion of the methylene component is as follows. The acetophenone deriva-
tive (1.00 equiv) followed by diethyl oxalate (1.00 equiv) were added to
a solution of sodium ethoxide (1.00m, 1.10 equiv) in EtOH at 08C. The
resulting suspension was stirred, overnight, at room temperature. Pentane
was added and the mixture was filtered. The residue was washed with
pentane and purified by recrystallisation.
Protein crystallography: T14-3e and PMA2-CT30 were mixed in a 1:1.4
ratio in HEPES (20 mm)/NaOH (pH 7.5), MgCl2 (2 mm), DTT (2 mm) at
a final concentration of 10 mgmLÀ1 and incubated, overnight, in the pres-
ence of the stabilising molecule 34. Protein crystals were grown in hang-
ing drops at 48C in CHES (0.1m, pH 9.0), Na-citrate (1.0m) and sucrose
(30% w/v) by mixing equal volumes of protein complex and reservoir so-
lution. After 2 days crystals appeared and grew within one week to di-
mensions of 300ꢃ200ꢃ200 mm. Crystals were directly transferred into
liquid nitrogen for flash-cooling.
Data collection, structure determination and refinement: Data collection
was performed in-house by using a Cu rotating anode (Rigaku) and was
processed with XDS.[39] The previously solved structure with 1 (PDB ID:
3M51) was used for a first round of refinement. The obtained model was
subjected to iterative rounds of model building and refinement by using
the programs COOT and REFMAC.[40,41] Figures were prepared with
PyMOL (www.pymol.org).
Acknowledgements
Scheme 1. Doebner condensation for the preparation of pyrrolinones and
the formation of pyrazoles.
C.O. thanks BayerCrop Science, Bayer Healthcare, MerckSerono, and
MSD for support.
General procedure for the preparation of pyrrolinones (Table 1): The
benzaldehyde derivative (1.00 equiv) followed by the aniline derivative
(1.00 equiv) were added to a suspension (~0.06m) of the methylene com-
ponent (1.00 equiv) in acetic acid. The reaction mixture was stirred, over-
night, at 908C. After being cooled to room temperature the mixture was
diluted with Et2O and filtered. The residue was washed with Et2O to give
the desired pyrrolinone derivative after drying (method A). Alternatively,
the filtrate was evaporated and the residue was taken up in 1,4-dioxane
and stirred for 45 min at 1208C. After being cooled to room temperature
the resulting suspension was filtered. The residue was washed with diox-
ane and dried, in vacuo, to give the desired pyrrolinone derivative (meth-
od B). The purity of the products was confirmed by RP-HPLC: C18
column (3 mm, CC 125/4), 0.5 mLminÀ1, starting from ACN/water+0.1%
TFA 20:80 (2 min) to 100:0 in 40 min.
[1] J. A. Wells, C. L. McClendon, Nature 2007, 450, 1001–1009.
[2] D. Bridges, G. B. Moorhead, Sci STKE 2005, 296, re10.
[3] C. Johnson, S. Crowther, M. J. Stafford, D. G. Campbell, R. Toth, C.
MacKintosh, Biochem. J. 2010, 427, 69–78.
[4] M. B. Yaffe, K. Rittinger, S. Volinia, P. R. Caron, A. Aitken, H.
Leffers, S. J. Gamblin, S. J. Smerdon, L. C. Cantley, Cell 1997, 91,
961–971.
[5] D. K. Morrison, Trends Cell Biol. 2009, 19, 16–23.
[6] W. J. Fantl, A. J. Muslin, A. Kikuchi, J. A. Martin, A. M. MacNicol,
R. Gross, L. T. Williams, Nature 1994, 371, 612–614.
[7] M. Molzan, B. Schumacher, C. Ottmann, A. Baljuls, L. Polzien, M.
Weyand, P. Thiel, R. Rose, M. Rose, P. Kuhenne, M. Kaiser, U. R.
Rapp, J. Kuhlmann, C. Ottmann, Mol. Cell. Biol. 2010, 30, 4698–
4711.
[8] D. S. Conklin, K. Galaktinov, D. Beach, Proc. Natl. Acad. Sci. USA
1995, 92, 7892–7896.
[9] C. Y. Peng, P. R. Graves, R. S. Thoma, Z. Wu, A. Shaw, Science
1997, 277, 1501–1505.
[10] A. Vassilev, K. J. Kaneko, H. Shu, Y. Zhao, M. L. DePamphilis,
Genes Dev. 2001, 15, 1229–1241.
[11] B. Schumacher, M. Skwarczynska, R. Rose, C. Ottmann, Acta Crys-
tallogr. Sect. F Struct. Biol. Cryst. Commun. 2010, 66, 978–984.
[12] S. Rajagopalan, R. S. Sade, F. M. Townsley, A. R. Fersht, Nucleic
Acids Res. 2010, 38, 893–906.
General procedure for the preparation of pyrazoles (Table 2): Hydrazine
monohydrate (6.00 equiv) was added to a solution (0.06m) of the corre-
sponding pyrrolinone (1.00 equiv) in acetic acid. The reaction mixture
was stirred for 2 h at 858C. After being cooled to room temperature
water was added and the mixture was extracted three times with EtOAC.
The combined organic layers were washed with brine, dried over MgSO4
and the solvent was evaporated, in vacuo, to give the desired product.
Protein preparation: The tobacco 14-3-3e isoform deleted from its C-ter-
minal 18 residues (T14-3eDC, amino acids 1–242 with a N-terminal RGS-
ACHTUNGTRENNUNG[His]6-tag, eGFP or GST-tag) was expressed in E. coli and purified as de-
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

C. Ottmann, H. Waldmann et al.
[13] B. Schumacher, J. Mondry, P. Thiel, M. Weyand, C. Ottmann, FEBS
Lett. 2010, 584, 1443–1448.
[14] H. Hermeking, Nat. Rev. Cancer 2003, 3, 931–943.
[15] D. Berg, C. Holzmann, O. Riess, Nat. Rev. Neurosci. 2003, 4, 752–
762.
[16] H. Fu, J. Coburn, R. J. Collier, Proc. Natl. Acad. Sci. USA 1993, 90,
2320–2324.
[17] C. Ottmann, L. Yasmin, M. Weyand, J. L. Veesenmeyer, C. Diaz,
R. H. Palmer, M. S. Francis, A. R. Hauser, A. Wittinghofer, B. Hall-
berg, EMBO J. 2007, 26, 902–913.
[18] A. Ballio, E. B. Chain, P. De Leo, B. F. Erlanger, M. Mauri, A.
Tonolo, Nature 1964, 203, 297.
[19] C. Oecking, C. Eckerskorn, E. W. Weiler, FEBS Lett. 1994, 352,
163–166.
[20] M. Wꢇrtele, C. Jelich-Ottmann, A. Wittinghofer, C. Oecking,
EMBO J. 2003, 22, 987–992.
[21] C. Oecking, M. Piotrowski, J. Hagemeier, K. Hagemann, Plant J.
1997, 12, 441–453.
[22] R. Rose, S. Erdmann, S. Bovens, A. Wolf, M. Rose, S. Hennig, H.
Waldmann, C. Ottmann, Angew. Chem. 2010, 122, 4223–4226;
Angew. Chem. Int. Ed. 2010, 49, 4129–4132.
[23] C. Ottmann, S. Marco, N. Jaspert, C. Marcon, N. Schauer, M.
Weyand, C. Vandermeeren, G. Duby, M. Boutry, A. Wittinghofer,
J. L. Rigaud, C. Oecking, Mol. Cell 2007, 25, 427–440.
[24] G. Tzivion, V. S. Gupta, L. Kaplun, V. Balan, Semin. Cancer Biol.
2006, 16, 203–213.
[25] R. K. Andrews, X. Du, M. C. Berndt, Drug News Perspect. 2007, 20,
285–292.
[26] H. Wu, J. Ge, S. Q. Yao, Angew. Chem. 2010, 122, 6678–6682;
Angew. Chem. Int. Ed. 2010, 49, 6528–6532.
[28] J. Zhao, Y. Du, J. R. Horton, A. K. Upadhyay, B. Lou, Y. Bai, X.
Zhang, L. Du, M. Li, B. Wang, L. Zhang, J. T. Barbieri, F. R. Khuri,
X. Cheng, H. Fu, Proc. Natl. Acad. Sci. USA 2011, 108, 16212–
16216.
[29] P. Thiel, M. Kaiser, C. Ottmann, Angew. Chem. 2012, 124, 2052–
2059.
[30] T. Sassa, T. Tojyo, K. Munakata, Nature 1970, 227, 379.
[31] C. Ottmann, M. Weyand, T. Sassa, T. Inoue, N. Kato, A. Wittinghof-
er, C. Oecking, J. Mol. Biol. 2009, 386, 913–919.
[32] J. Liu, J. D. Farmer, W. S. Lane, J. Friedman, I. Weissman, S. L.
Schreiber, Cell 1991, 66, 807–815.
[33] E. J. Brown, M. W. Albers, T. B. Shin, K. Ichikawa, C. T. Keith, W. S.
Lane, S. L. Schreiber, Nature 1994, 369, 756–758.
[34] T. Clackson, W. Yang, L. W. Rozamus, M. Hatada, J. F. Amara, C. T.
Rollins, L. F. Stevenson, S. R. Magari, S. A. Wood, N. L. Courage, X.
Lu, F. Cerasoli, M. Gilman, D. A. Holt, Proc. Natl. Acad. Sci. USA
1998, 95, 10437–10442.
[35] J. H. Bayle, J. S. Grimley, K. Stankunas, J. E. Gestwicki, T. J. Wand-
less, G. R. Crabtree, Chem. Biol. 2006, 13, 99–107.
[36] M. A. Farrar, J. Alberol-Ila, R. M. Perlmutter, Nature 1996, 383,
178–181.
[37] S. J. Kopytek, R. F. Standaert, J. C. Dyer, J. C. Hu, Chem. Biol. 2000,
7, 313–321.
[38] T. M. Kamenecka, S. J. Danishefsky, Chem. Eur. J. 2001, 7, 41–63;
b) A. Martꢈn, M. Atlabe, S. Filippone, A. Martꢈn Domenech, L.
Echegoyen, C. M. Cardona, Angew. Chem. 2006, 118, 116–120;
Angew. Chem. Int. Ed. 2006, 45, 7430–7433.
[39] W. Kabsch, J. Appl. Crystallogr. 1993, 26, 795–800.
[40] P. Emsley, K. Cowtan, Acta. Cryst. D 2004, 60, 2126–2132.
[41] G. N. Murshudov, A. A. Vagin, E. J. Dodson, Acta. Cryst. D 1997,
53, 240–255.
[27] V. Corradi, M. Mancini, F. Manetti, S. Petta, M. A. Santucci, M.
Botta, Bioorg. Med. Chem. Lett. 2010, 20, 6133–6137.
Received: November 30, 2011
Published online: && &&, 2012
&
8
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Small-Molecule Stabiliser of Protein–Protein Interaction
FULL PAPER
Glued together: An initial screening
hit that stabilises a 14-3-3 protein–pro-
tein interaction has been optimised by
means of chemical synthesis. The
resulting compound is three-times
more active than the initial hit and dis-
plays selectivity for the interaction of
14-3-3 proteins with the plant proton
pump PMA2 (see figure).
Protein–Protein Interactions
A. Richter, R. Rose, C. Hedberg,
H. Waldmann,*
C. Ottmann* .................. &&&& — &&&&
An Optimised Small-Molecule Stabil-
iser of the 14-3-3–PMA2 Protein–
Protein Interaction
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!