Stafia-1: a STAT5a-Selective Inhibitor Developed via Docking-Based Screening of in Silico O-Phosphorylated Fragments

Article abstract of DOI:10.1002/chem.201904147

We present a new approach for the identification of inhibitors of phosphorylation-dependent protein–protein interaction domains, in which phenolic fragments are adapted by in silico O-phosphorylation before docking-based screening. From a database of 10 369 180 compounds, we identified 85 021 natural product-derived phenolic fragments, which were virtually O-phosphorylated and screened for in silico binding to the STAT3 SH2 domain. Nine screening hits were then synthesized, eight of which showed a degree of in vitro inhibition of STAT3. After analysis of its selectivity profile, the most potent inhibitor was then developed to Stafia-1, the first small molecule shown to preferentially inhibit the STAT family member STAT5a over the close homologue STAT5b. A phosphonate prodrug based on Stafia-1 inhibited STAT5a with selectivity over STAT5b in human leukemia cells, providing the first demonstration of selective in vitro and intracellular inhibition of STAT5a by a small-molecule inhibitor.

Full text of DOI:10.1002/chem.201904147

A Journal of
Accepted Article
Title: Stafia-1: a STAT5a-selective inhibitor developed via docking-
based screening of in silico O-phosphorylated fragments
Authors: Kalaiselvi Natarajan, Daniel Müller-Klieser, Stefan Rubner,
and Thorsten Berg
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201904147
Link to VoR: http://dx.doi.org/10.1002/chem.201904147
Supported by

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
Stafia-1: a STAT5a-selective inhibitor developed via docking-
based screening of in silico O-phosphorylated fragments
Kalaiselvi Natarajan, Daniel Müller-Klieser#, Stefan Rubner# and Thorsten Berg*
Abstract: We present a new approach for the identification of
inhibitors of phosphorylation-dependent protein-protein interaction
domains, in which phenolic fragments are adapted by in silico O-
phosphorylation before docking-based screening. From a database of
10,369,180 compounds we identified 85,021 natural product-derived
phenolic fragments, which were virtually O-phosphorylated and
screened for in silico binding to the STAT3 SH2 domain. Nine
screening hits were then synthesized, eight of which showed a degree
of in vitro inhibition of STAT3. After analysis of its selectivity profile,
the most potent inhibitor was then developed to Stafia-1, the first small
molecule shown to preferentially inhibit the STAT family member
STAT5a over the close homologue STAT5b. A phosphonate prodrug
based on Stafia-1 inhibited STAT5a with selectivity over STAT5b in
human leukemia cells, providing the first demonstration of selective in
vitro and intracellular inhibition of STAT5a by a small-molecule
inhibitor.
no STAT5a inhibitors[3, 9] with selectivity over STAT5b have been
disclosed to date.
Here, we present virtual (in silico) O-phosphorylation of
preselected phenolic fragments of natural products,[10] followed by
docking-based virtual screening, as a novel methodology for the
identification of inhibitors of phosphotyrosine-dependent protein-
protein interaction domains. The initial virtual compound library
was downloaded from the ZINC data base[11] as a collection of
10,369,180 structures. Filtering this database for structural
elements described by the structural classification of natural
products (SCONP) tree[10] identified 799,335 compounds (Figure
1A, step 1, Figure S1, and Supporting Methods). Further filtering
for fragments with a phenol moiety and a molecular weight below
500 g/Mol, and removal of certain reactive moieties (Figure 1A,
step 2, and Supporting Methods), narrowed down the selection to
85,021 compounds, which were then virtually O-phosphorylated
on their phenolic moiety by altering their SMILES string (Figure
1A, step 3).[12] Virtual screening of the O-phosphorylated
compounds against the STAT3 SH2 domain (PBD ID: 1BG1)[13]
with AutoDock Vina[14] resulted in 1,114 compounds which fulfilled
predefined criteria for the distances between the molecules’
phosphate groups and the crucial STAT3 SH2 domain residues
Arg609 and Lys591 (Figure 1A, step 4, and Figure S2).[13] After
visual inspection of the binding poses, 9 molecules (1-9) were
selected (Figure 1A, step 5, Table S1), which display a variable
degree of resemblance to natural products, depending on the size
of the underlying natural product-derived structural element from
the SCONP tree.[10] Molecules 1-9 were synthesized via O-
phosphorylation of commercially available or pre-synthesized
phenolic precursors, by a two-step phosphorylation/debenzylation
process (Figure 1A, step 6, Table S1, and Supporting Information),
and tested in a fluorescence polarization (FP) assay against the
STAT3 SH2 domain (Figure 1A, step 7).[15] 8 of the O-
phosphorylated molecules 1-9 showed a degree of STAT3
inhibition, with 1-3 inhibiting STAT3 by more than 40 % at 100 µM
(Table S1). 1 docked into the STAT3 SH2 domain with its
phosphate group in close proximity to Arg609 and Lys591 (Figure
1B). Although screening had been performed with the aim of
identifying inhibitors of the STAT3 SH2 domain, analysis of
specificity profiles revealed that several compounds, including 1,
were more active against STAT5a and STAT5b[16] than against
STAT3 (Table S1). This suggests that the docking approach may
not be sensitive enough to clearly discriminate between the STAT
family members. Since selective STAT5a inhibitors have not yet
been reported, we decided to optimize 1 for binding to STAT5a,
rather than STAT3. Compound 1 was chosen as a starting point
for inhibitor development because it lacked reactive functional
groups, and its m-terphenyl scaffold should allow for flexible
modifications via Suzuki couplings.
Protein-protein interactions mediate most biological processes,
and their functional modulation by small molecules offers vast
opportunities for basic research and drug development.[1]
However, protein-protein interactions represent challenging
targets for small molecules, and design approaches for inhibitor
development are rare.[2]
Phosphorylation-dependent protein-protein interactions are
mediated by the phosphorylated side chains of tyrosine, serine,
and threonine residues, and play an important role in signal
transduction. We recently proposed O-phosphorylation of
preselected natural products as an approach for the development
of non-peptidic and non-reactive ligands of phosphorylation-
dependent protein-protein interactions.[3] We used this approach
to develop catechol bisphosphates[4] as the first chemical entities
that inhibit the phosphotyrosine-dependent Src homology 2 (SH2)
domain of the transcription factor STAT5b with high selectivity
over the close homologue STAT5a.[5] Both STAT5 proteins are
constitutively activated in numerous human tumors.[6] Selective
inhibition of either STAT5 protein is desirable for the functional
analysis of the non-redundant functions of STAT5a and
STAT5b,[7] and would offer flexibility in tailoring the antitumor
treatment strategy to individual human tumors. Small molecule
STAT5a inhibitors with selectivity over STAT5b could also serve
as therapeutic modalities for age-related osteoporosis.[8] However,
[a]
Dr. K. Natarajan, D. Müller-Klieser[#], S. Rubner[#], Prof. Dr. T. Berg*
Leipzig University, Institute of Organic Chemistry, Johannisallee 29,
04103 Leipzig, Germany. E-mail: tberg@uni-leipzig.de
http://home.uni-leipzig.de/tberg
[#]
These authors contributed equally.
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
Table 1: Structures of synthesized compounds and activities against STAT5a,
STAT5b, and STAT3.
No.
Structure
STAT5a
Ki (µM) or
STAT5b
Ki (µM) or
STAT3
Ki (µM) or
inhibition (%)
at 200 µM
inhibition (%)
at 200 µM
inhibition (%)
at 200 µM
1
10
11
12
12a
13
14
15
16
17
18
19
20
21
22
23
19.7 ± 0.9
20.9 ± 1.6
23.7 ± 2.1
13.8 ± 0.8
no inhibition
16.0 ± 2.8
43.9 ± 3.0
16.5 ± 2.5
22.3 ± 4.8
17.1 ± 2.2
14.0 ± 2.1
11.1 ± 1.9
10.9 ± 1.8
55.2 ± 4.9
10.8 ± 1.1
53.0 ± 2.1
24.5 ± 1.6
18.7 ± 0.0
48.4 ± 2.0
38.5 ± 0.2
no inhibition
43.1 ± 1.5
78.6 ± 0.4
59.2 ± 5.0
18.9 ± 0.9
48 ± 1 %
inhibition
77.1 ± 5.8
no inhibition
51 ± 2 %
inhibition
Figure 1. A) Selection criteria and procedures applied for docking-based
screening of virtually O-phosphorylated natural product-related libraries.
B) Docking pose of
1 in the STAT3 SH2 domain with predicted distances
between the phosphate group of
1 and STAT3 Lys591 and Arg609. The
28 ± 1 %
inhibition
figure was generated using PyMOL.[17]
Dose-dependent analysis of 1 in FP assays already showed a
slight preference of 1 for STAT5a over STAT5b, which was lost
by 4,4’’-dichloro-substitution (compound 10, Tables 1 and S2), but
improved for the unsubstituted m-terphenyl phosphate 11 and the
4,4’’-dimethoxy derivative 12. Activity of 12 was phosphorylation-
dependent, as indicated by the lack of activity of the
unphosphorylated precursor 12a. Given the beneficial effects of
the methoxy groups, we carried out a positional scanning
approach of di-, tetra-, and hexa-methoxy substitution on the
symmetrical m-terphenyl phosphates. In the disubstituted series
12-14, 4,4’’-disubstitution (12) was preferable over 3,3’’-
disubstitution (compound 13) or 2,2’’-disubstitution (compound
14). In the tetrasubstituted series 15-17, the 3,3’’,4,4’’-substituted
compound 15 was more selective for STAT5a than the 2,2’’,3,3’’-
substituted isomer 16 and the 3,3’’,5,5’’-substituted isomer 17,
and also more selective for STAT5a than 12. Attempts to improve
STAT5a selectivity of 15 by rigidifying the alkoxy substituents
(compounds 18 and 19) were only moderately successful. In the
hexasubstituted series 20-21, the 3,3’’,4,4’’,5,5’’-substituted
compound 20 was significantly more active against STAT5a than
21. Compound 20, which was synthesized in three synthetic steps
(Figure 2A), inhibited STAT5a (IC50 = 22.2 ± 3.6 µM, Ki = 10.9 ±
1.8 µM) with at least 9-fold selectivity over STAT5b (36 ± 5 %
inhibition at 200 µM, the highest concentration tested) and higher
selectivity against other STAT family members (Figure 2B).
Exchanging the methoxy substituents for fluorine (compound 22)
maintained activity against STAT5a but reduced selectivity.
Deletion of one phenyl ring of 20 caused a 5-fold drop in activity
against STAT5a (compound 23), demonstrating the necessity of
the m-terphenyl phosphate moiety for STAT5a inhibition.
44 ± 3 %
inhibition
43 ± 2 %
inhibition
40 ± 2 %
inhibition
15 ± 2 %
inhibition
36.5 ± 2.5
28.0 ± 0.5
35.4 ± 2.8
50 ± 3 %
inhibition
63.8 ± 2.3
51.2 ± 2.0
37 ± 5 %
inhibition
27 ± 2 %
inhibition
28 ± 2 %
inhibition
15 ± 3 %
inhibition
23.3 ± 0.4
50.4 ± 4.2
39 ± 1 %
inhibition
40 ± 1 %
inhibition
This article is protected by copyright. All rights reserved.

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
context of STAT5b, as represented by the mutant STAT5b
Arg566Trp/Gln636Pro/Met639Asn/Phe640Leu/Met644Lys/Asn6
64Ser/Tyr679Phe (dubbed STAT5b 7M),[18] almost restored
binding (Ki = 21.4 ± 4.4 µM, Figure 3A and Table S3). These data
indicate that recognition of 20 by STAT5a depends on both the
SH2 domain and Trp566. The remaining twofold activity
difference of 20 between STAT5a (Ki = 10.9 ± 1.8 µM) and
STAT5b 7M (Ki = 21.4 ± 4.4 µM, Figure 3A and Table S3) may be
mediated by allosteric cross-communication with divergent amino
acid positions in the linker domain,[20] the DNA binding domain,[21]
or the coiled-coil domain.[22]
Figure 2. A) Synthesis of 20. a) Na2CO3, Pd(PPh3)4, H2O/MeOH, 64%; b)
(BnO)2P(O)H, CCl4, DIEA, DMAP, CH3CN, 58 %; c) Pd/C, H2, EtOH; 98%.
B) Activities of 20 against the SH2 domains of STAT proteins and Lck.
The SH2 domains of STAT5a and STAT5b are 93% identical on
the amino acid level, with only 6 of 91 amino acids differing
(Figure S3). To investigate the molecular origin of the specificity
of 20 for STAT5a over STAT5b, we used wild-type and point
mutant STAT5 proteins in FP assays. The activity of 20 against
the
point
mutant
STAT5b
Gln636Pro/Met639Asn/Phe640Leu/Met644Lys/Asn664Ser/Tyr6
79Phe (dubbed STAT5b 6M),[18] in which the six amino acids of
the STAT5b SH2 domain which differ from those in the STAT5a
SH2 domain have been replaced by their STAT5a counterparts
(Figure S3), was only marginally increased (44 ± 5 % inhibition at
200 µM) compared to wild-type STAT5b (36 ± 5 % inhibition at
200 µM), and thus approximately 10-fold lower than the activity
against wild-type STAT5a (Ki = 10.9 ± 1.8 µM, IC50 = 22.2 ± 3.6
µM, Figure 3A and Table S3). This indicated that factors outside
the SH2 domain must play a significant role for binding of 20 to
STAT5a.
We recently described the divergent amino acids in position
566 of the STAT5 linker domain (Trp in STAT5a, Arg in STAT5b,
Figure S3), adjacent to the SH2 domain, as the crucial
determinant for selective STAT5b inhibition by small
molecules.[18-19] To investigate the role of STAT5a Trp566, we
tested 20 against the crossover point mutant STAT5a Trp566Arg,
and found its activity to be reduced by 7-fold (Ki = 74.0 ± 7.0 µM,
Figure 3A and Table S3) as compared to wild-type STAT5a. While
this suggests a role for Trp566 in binding to 20, it is not the sole
determinant of specificity, since the reverse mutant STAT5b
Arg566Trp (43 ± 7 % inhibition at 200 µM) was only marginally
more inhibited than wild-type STAT5b (36 ± 5 % inhibition at 200
µM) by 20 (Figure 3A and Table S3). However, the combined
presentation of the STAT5a SH2 domain and Trp566 in the
Figure 3. A) Activity of 20 against wild-type and mutant STAT5 proteins.
n.a.: not applicable. B) Docking pose of 20 into the X-ray structure of
murine STAT5a (PBD 1Y1U).[23] SH2 domain shown in yellow, linker
domain in blue; the side chains of amino acids defined as flexible in the
docking are shown with carbon atoms in green. Divergent amino acids in
the SH2 domain defined as rigid are shown with carbon atoms in yellow.
Asn664 is replaced by Ser in human STAT5a. The figure was generated
using PyMOL.[17]
Docking of 20 into the crystal structure of STAT5a[23] using
AutoDock FR,[24] with the side chains of amino acids in the
phosphotyrosine binding pocket or its immediate vicinity (Trp566,
Lys600, Arg618, Asn639, and Lys644) defined as flexible, placed
one of the terminal phenyl rings of 20 near STAT5a Trp566
(Figure 3B), consistent with -stacking interactions and with
hydrogen bonding between the proton attached to the indole
nitrogen of Trp566 and at least one of the methoxy groups of 20.
The phosphate group of 20 is predicted to form electrostatic
interactions with Lys600 and Arg618, which are identical in
This article is protected by copyright. All rights reserved.

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
STAT5a and STAT5b, in
a manner similar to that of
phosphotyrosine-containing ligands.[13] Selectivity-conferring
electrostatic interactions may arise from interaction with the side
chain of STAT5a Lys644, which is replaced by methionine in
STAT5b. The relevance of STAT5a Lys644 for binding of 20 was
demonstrated by reduced inhibition of the crossover double
mutant STAT5a Trp566Arg/Lys644Met (44 ± 4 % inhibition at 200
µM) as compared to STAT5a Trp566Arg (Ki = 74.0 ± 7.0 µM,
Figure 3A and Table S3). Conversely, the activity of 20 against
the crossover mutant STAT5b Arg566Trp/Met644Lys was
enhanced (Ki = 63.1 ± 4.2 µM) as compared to STAT5b
Arg566Trp (43 ± 7 % inhibition at 200 µM, Figure 3A and Table
S3).
The
triple
cross-over
mutant
STAT5b
Arg566Trp/Met639Asn/Met644Lys (dubbed STAT5b-3M, Ki =
49.0 ± 5.4 µM, Figure 3A and Table S3) was somewhat more
susceptible to inhibition by 20, suggesting a small contribution of
Asn639 (present in STAT5a) to binding. The remaining 2.5-fold
activity gap between the triple mutant STAT5b-3M and the 7-fold
mutant STAT5b-7M (Ki = 21.4 ± 4.4 µM) points towards an
allosteric contribution by one or more of the remaining four
STAT5a/b SH2 domain divergent amino acids, which are further
removed from the putative binding site of 20.
We note that in the available crystal structure of STAT5a,[23]
the position of the side chain of Lys600 is not suitable for
phosphate group binding (Figure S4). Lys600 is required for
binding of phosphotyrosine in peptidic SH2 domain ligands, and
the unsuitable orientation is presumably a consequence of the
fact that this STAT5a structure derives from the protein without a
bound ligand. In contrast, in the crystal structure of tyrosine-
phosphorylated STAT3 homodimers,[13] the important amino acid
side chains adopt a suitable conformation for phosphate binding
(Figure S4). The virtual screening of STAT3 described in this
study was achieved using AutoDock Vina,[14] which in our
experience works best with rigid proteins. In contrast, STAT5a
docking requires an approach with flexible amino acid side chains,
such as AutoDock FR,[24] which was used for rationalizing the
binding mode of 20 (Figure 3B). However, AutoDock FR is less
adaptable to virtual screening of large chemical libraries than
AutoDock Vina.
We synthesized the phosphonates 24-26, which are not
susceptible to phosphatase-catalyzed cleavage, as derivatives of
the phosphate 20 (Figure 4A, Figures S5-S7).[25] The methylene
phosphonate 24 (Ki = 75 ± 13 μM) was slightly less active against
STAT5a than the difluoromethylene phosphonate 25 (Ki = 52 ± 3
μM, Figure 4B). The most potent derivative was the
monofluoromethylene phosphonate 26 (Ki = 28 ± 3 μM), which
retained selectivity over STAT5b and other STAT proteins (Figure
Figure 4. A) Structure of phosphonates 24-26. B) Activities of 24-26 against
STAT5a in FP assays. C) Selectivity profile of 26 against STAT proteins.
Phosphonates are typically not cell-permeable, being
negatively charged at physiological pH. In order to mask its
negative charges, 26 was converted to the pivaloyloxymethyl
prodrug 27 (Figure 5A). Pivaloyloxymethyl prodrugs of organic
phosphates[4a-c] and phosphonates[27] are cleaved by intracellular
esterases, releasing the bioactive parent compounds. 27 was
tested for its ability to prevent phosphorylation of STAT5a in K562
cells, a human chronic myelogenous leukemia cell line. In these
cells, both STAT5a and STAT5b are constitutively
phosphorylated on Tyr694 (STAT5a) and Tyr699 (STAT5b),
respectively, by Bcr-Abl. Since phosphorylation is dependent on
the function of the SH2 domain, a STAT5 SH2 domain inhibitor
will reduce the ability of STAT5 to be tyrosine phosphorylated by
Bcr-Abl (Figure 5B). Commercially available antibodies that
recognize STAT5a/b only in their tyrosine phosphorylated state
do not differentiate between the two STAT5 proteins, so we
ectopically expressed fusion proteins of either STAT5a or
STAT5b and GFP to distinguish unambiguously between the two
STAT5 proteins.[4a-c] In the presence of the prodrug 27,
phosphorylation of STAT5a Tyr694 in the STAT5a-GFP fusion
protein was inhibited in a dose-dependent manner (Figure 5C, E).
In contrast, phosphorylation of Tyr699 in the STAT5b-GFP
construct was only minimally affected (Figure 5D, F), showing that
the selectivity of 26 for STAT5a over STAT5b is maintained in the
cellular environment. In contrast, the STAT5 inhibitor AC-4-
130,[28] for which no preference for either STAT5 protein has been
reported, and the Bcr-Abl inhibitor imatinib[29] do not discriminate
between STAT5a and STAT5b in this assay (Figure S8).
Phosphorylation of endogenous STAT5a/b was reduced to a
substantially lesser degree than that of STAT5a-GFP (Figure S9)
by 27. This is consistent with our previous observations, which
4B, C). This represents
a
rare case in which
a
monofluoromethylene phosphonate displays higher activity
against an SH2 domain than the corresponding difluoromethylene
phosphonate,[26] despite the introduction of an uninduced chiral
center by fluorine monosubstitution.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
indicate that the majority of endogenous phospho-STAT5 in K562
cells is STAT5b.[4a-c]
demonstration that selective targeting of STAT5a over STAT5b is
feasible both in vitro and in cells.
Acknowledgements
This work was generously supported by the Deutsche
Forschungsgemeinschaft (BE 4572/4-1 and INST 268/281-1
FUGG), and the European Union and the Free State of Saxony,
European Regional Development Fund. We extend our thanks to
Nagarajan Elumalai, Julian Gräb and Barbara Klüver for
experimental support, and to Andreas Rost for support with virtual
screening-related computations.
Keywords: biological activity • inhibitors • protein-protein
interactions • SH2 domains • transcription factors
[1]
[2]
a) J. A. Wells, C. L. McClendon, Nature 2007, 450, 1001-1009; b) L. G.
Milroy, T. N. Grossmann, S. Hennig, L. Brunsveld, C. Ottmann, Chem.
Rev. 2014, 114, 4695-4748.
a) P. M. Cromm, J. Spiegel, T. N. Grossmann, ACS Chem. Biol. 2015,
10, 1362-1375; b) C. G. Cummings, A. D. Hamilton, Curr. Opin. Chem.
Biol. 2010, 14, 341-346; c) J. Zaminer, C. Brockmann, P. Huy, R. Opitz,
C. Reuter, M. Beyermann, C. Freund, M. Muller, H. Oschkinat, R.
Kuhne, H. G. Schmalz, Angew. Chem. Int. Ed. 2010, 49, 7111-7115.
M. Gräber, W. Janczyk, B. Sperl, N. Elumalai, C. Kozany, F. Hausch, T.
A. Holak, T. Berg, ACS Chem. Biol. 2011, 6, 1008-1014.
Figure 5. A) Synthesis of prodrug 27. B) Tyrosine phosphorylation of
STAT5a/b by Bcr-Abl is inhibited by a ligand of the SH2 domain. C) Effect
of 27 on phosphorylation of STAT5a in STAT5a-GFP-transfected K562
cells, and D) on phosphorylation of STAT5b in STAT5b-GFP-transfected
K562 cells. E, F) Quantitation of the data shown in C) (n=4), and D) (n=3),
respectively. Phospho-STAT5a/b-GFP levels are normalized against total
STAT5a/b-GFP. All error bars represent standard deviations (s.d.).
[3]
[4]
a) N. Elumalai, A. Berg, K. Natarajan, A. Scharow, T. Berg, Angew.
Chem. Int. Ed. 2015, 54, 4758-4763; b) N. Elumalai, A. Berg, S.
Rubner, T. Berg, ACS Chem. Biol. 2015, 10, 2884-2890; c) N. Elumalai,
A. Berg, S. Rubner, L. Blechschmidt, C. Song, K. Natarajan, J. Matysik,
T. Berg, Sci. Rep. 2017, 7, 819; d) N. Elumalai, K. Natarajan, T. Berg,
Bioorg. Med. Chem. 2017, 25, 3871-3882.
In conclusion, we present docking-based screening of in
silico O-phosphorylated natural product fragments as a novel
method for identifying lead structures for the development of
[5]
[6]
L. Hennighausen, G. W. Robinson, Genes Dev. 2008, 22, 711-721.
G. Miklossy, T. S. Hilliard, J. Turkson, Nat. Rev. Drug Discov. 2013, 12,
611-629.
inhibitors
of
phosphorylation-dependent
protein-protein
interaction domains, which are of crucial importance to cellular
signaling. While virtual screening of phosphonates against a
phosphorylation-dependent protein-protein interaction domain
has been reported,[30] our work represents the first case in which
the virtual screening library itself is generated by in silico O-
phosphorylation. Application of this concept to the STAT3 SH2
domain resulted in the moderate STAT3 inhibitor 1, which was
then discovered to preferentially target STAT5. Analysis of
structure-activity relationships led to the development of 20, the
first inhibitor of STAT5a which displays high selectivity over
STAT5b and other STAT family members. We dubbed 20 Stafia-
1 (STAT five a inhibitor 1). The use of wild-type and point mutant
STAT5 proteins demonstrated that both the SH2 domain and
Trp566 in the adjacent linker domain contribute to selective
recognition of Stafia-1 by STAT5a. The cell-permeable prodrug
27, based on the Stafia-1-derived monofluoromethylene
phosphonate 26, inhibited tyrosine phosphorylation of STAT5a
with selectivity over STAT5b in cultured human leukemia cells,
and represents a valuable tool to define the non-redundant
molecular functions of the two highly homologous transcription
factors in tumor cells.[7] Selective inhibition of STAT5a by 27,
especially in direct comparison with selective inhibition of STAT5b
by catechol bisphosphate-based prodrugs such as Pomstafib-
2,[4c] would allow for dissection of the target genes of STAT5a and
STAT5b with high temporal control.[7] Our data provide the first
[7]
[8]
B. Basham, M. Sathe, J. Grein, T. McClanahan, A. D'Andrea, E. Lees,
A. Rascle, Nucleic Acids Res. 2008, 36, 3802-3818.
K. M. Lee, K. H. Park, J. S. Hwang, M. Lee, D. S. Yoon, H. A. Ryu, H.
S. Jung, K. W. Park, J. Kim, S. W. Park, S. H. Kim, Y. M. Chun, W. J.
Choi, J. W. Lee, Cell Death Dis 2018, 9, 1136.
[9]
a) J. Müller, B. Sperl, W. Reindl, A. Kiessling, T. Berg, ChemBioChem
2008, 9, 723-727; b) A. A. Cumaraswamy, A. M. Lewis, M. Geletu, A.
Todic, D. B. Diaz, X. R. Cheng, C. E. Brown, R. C. Laister, D. Muench,
K. Kerman, H. L. Grimes, M. D. Minden, P. T. Gunning, ACS Med.
Chem. Lett. 2014, 5, 1202-1206; c) Z. Liao, L. Gu, J. Vergalli, S. A.
Mariani, M. De Dominici, R. K. Lokareddy, A. Dagvadorj, P.
Purushottamachar, P. A. McCue, E. Trabulsi, C. D. Lallas, S. Gupta, E.
Ellsworth, S. Blackmon, A. Ertel, P. Fortina, B. Leiby, G. Xia, H. Rui, D.
T. Hoang, L. G. Gomella, G. Cingolani, V. Njar, N. Pattabiraman, B.
Calabretta, M. T. Nevalainen, Mol. Cancer Ther. 2015, 14, 1777-1793;
d) A. Berg, T. Berg, Bioorg. Med. Chem. Lett. 2017, 27, 3349-3352; e)
E. L. Wong, E. Nawrotzky, C. Arkona, B. G. Kim, S. Beligny, X. Wang,
S. Wagner, M. Lisurek, D. Carstanjen, J. Rademann, Nat. Commun.
2019, 10, 66.
[10] M. A. Koch, A. Schuffenhauer, M. Scheck, S. Wetzel, M. Casaulta, A.
Odermatt, P. Ertl, H. Waldmann, Proc. Natl. Acad. Sci. U. S. A. 2005,
102, 17272-17277.
[11] J. J. Irwin, T. Sterling, M. M. Mysinger, E. S. Bolstad, R. G. Coleman, J.
Chem. Inf. Model. 2012, 52, 1757-1768.
[12] D. Weininger, J. Chem. Inf. Comput. Sci. 1988, 28, 31-36.
[13] S. Becker, B. Groner, C. W. Müller, Nature 1998, 394, 145-151.
[14] O. Trott, A. J. Olson, J. Comput. Chem. 2010, 31, 455-461.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
[15] J. Schust, T. Berg, Anal. Biochem. 2004, 330, 114-118.
[16] J. Müller, J. Schust, T. Berg, Anal. Biochem. 2008, 375, 249-254.
[17] Schrodinger, LLC, 2015.
[25] T. R. Burke, Jr., Z. J. Yao, D. G. Liu, J. Voigt, Y. Gao, Biopolymers
2001, 60, 32-44.
[26] T. R. Burke, Jr., M. S. Smyth, A. Otaka, M. Nomizu, P. P. Roller, G.
Wolf, R. Case, S. E. Shoelson, Biochemistry 1994, 33, 6490-6494.
[27] P. K. Mandal, F. Gao, Z. Lu, Z. Ren, R. Ramesh, J. S. Birtwistle, K. K.
Kaluarachchi, X. Chen, R. C. Bast, W. S. Liao, J. S. McMurray, J. Med.
Chem. 2011, 54, 3549-3563.
[18] J. Gräb, A. Berg, L. Blechschmidt, B. Klüver, S. Rubner, D. Y. Fu, J.
Meiler, M. Gräber, T. Berg, ACS Chem. Biol. 2019, 14, 796-805.
[19] A. Berg, B. Sperl, T. Berg, ChemBioChem 2019, 20, 2227-2231.
[20] A. T. Namanja, J. Wang, R. Buettner, L. Colson, Y. Chen, J. Mol. Biol.
2016, 428, 579-589.
[28] B. Wingelhofer, B. Maurer, E. C. Heyes, A. A. Cumaraswamy, A.
Berger-Becvar, E. D. de Araujo, A. Orlova, P. Freund, F. Ruge, J. Park,
G. Tin, S. Ahmar, C. H. Lardeau, I. Sadovnik, D. Bajusz, G. M. Keseru,
F. Grebien, S. Kubicek, P. Valent, P. T. Gunning, R. Moriggl, Leukemia
2018, 32, 1135-1146.
[21] C. Mertens, B. Haripal, S. Klinge, J. E. Darnell, Proc. Natl. Acad. Sci. U.
S. A. 2015, 112, 14811-14816.
[22] a) T. Zhang, W. H. Kee, K. T. Seow, W. Fung, X. M. Cao, Mol. Cell.
Biol. 2000, 20, 7132-7139; b) M. B. Minus, W. Liu, F. Vohidov, M. M.
Kasembeli, X. Long, M. J. Krueger, A. Stevens, M. I. Kolosov, D. J.
Tweardy, E. A. Sison, M. S. Redell, Z. T. Ball, Angew. Chem. Int. Ed.
2015, 54, 13085-13089.
[29] R. Capdeville, E. Buchdunger, J. Zimmermann, A. Matter, Nat. Rev.
Drug Discov. 2002, 1, 493-502.
[30] P. Thiel, L. Roglin, N. Meissner, S. Hennig, O. Kohlbacher, C. Ottmann,
Chem. Commun. 2013, 49, 8468-8470.
[23] D. Neculai, A. M. Neculai, S. Verrier, K. Straub, K. Klumpp, E. Pfitzner,
S. Becker, J. Biol. Chem. 2005, 280, 40782-40787.
[24] P. A. Ravindranath, S. Forli, D. S. Goodsell, A. J. Olson, M. F. Sanner,
PLoS Comput Biol 2015, 11, e1004586.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904147
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents
COMMUNICATION
Kalaiselvi Natarajan, Daniel Müller-
Klieser#, Stefan Rubner# and Thorsten
Berg*
Page No. – Page No.
Stafia-1: a STAT5a-selective inhibitor
developed via docking-based
screening of in silico O-
phosphorylated fragments
Creating and searching the virtual haystack: In silico O-phosphorylation of preselected natural product-related fragments from the
SCONP (structural classification of natural products) tree, virtual screening and chemical derivatization enabled the development of
Stafia-1, the first molecule shown to inhibit the transcription factor STAT5a with selectivity over the close homologue STAT5b.
This article is protected by copyright. All rights reserved.