Benzoylphosphonate-based photoactive phosphopeptide mimetics for modulation of protein tyrosine phosphatases and highly specific labeling of SH2 domains

Article abstract of DOI:10.1002/anie.201201475

A light switch for phosphotyrosine- recognizing proteins: Irradiation of the bioisosteric benzoylphosphonate suffices to "turn off" the activity of target proteins and to label them covalently (see scheme). Photoactive bioisosters may find applications in functional cell biology, bioanalytics, and proteome research.

Full text of DOI:10.1002/anie.201201475

Angewandte
Chemie
DOI: 10.1002/anie.201201475
Photoactive Bioisosters
Benzoylphosphonate-Based Photoactive Phosphopeptide Mimetics for
Modulation of Protein Tyrosine Phosphatases and Highly Specific
Labeling of SH2 Domains**
Andrꢀ Horatscheck, Stefan Wagner, Jutta Ortwein, Boo Geun Kim, Michael Lisurek,
Samuel Beligny, Anja Schꢁtz, and Jçrg Rademann*
The phosphorylation of tyrosine residues is an important
posttranslational modification of proteins and regulates the
activity of numerous signal transduction pathways, for
example, in highly proliferating cells. Phosphotyrosine resi-
dues are located on membrane receptors,[1] adaptor pro-
teins,[2] enzymes,[3] and transcription factors.[4] They are
recognized by protein tyrosine phosphatases (PTPs) and
phosphotyrosine binding domains exemplified by the src-
homology domains type 2 (SH2 domains) first described in
the human sarcoma protooncogene, src.[5] Temporally
resolved and spatially resolved analysis of the presence and
activity of phosphotyrosine binding sites is of fundamental
relevance for describing cellular activation properly.
Photoactivated chemical probes have been used inten-
sively to identify the interaction partners of small molecules
in biological systems.[6] Usually, such probes are composed of
a bioactive ligand component and a photoactive label, which
contains an aryl azide, a diazirine, or a benzophenone residue.
For targeting phosphotyrosine binding sites several bioiso-
steric groups have been described.[5] Benzyl phosphonates,[7]
phenyl difluoromethyl phosphonates,[8] and isothiazolidi-
nones[9] have found broader application. Recently, we have
used several phosphotyrosine-mimicking fragments for the
development of very specific PTP inhibitors[10] employing
a variation of dynamic ligation screening.[11] In addition,
covalent modifiers targeting the active sites of PTPs have
been proposed as “activity-based probes”, however, with
limited selectivity and specificity.[12]
Here, we describe a conceptually different approach
toward novel covalent protein probes targeting the active
sites of phosphotyrosine binding proteins. Benzoylphospho-
nates of general structure 1 were envisaged as bioisosters of
the phosphotyrosine residue capable of binding to phospho-
tyrosine recognition sites as phosphotyrosine mimetics.
Owing to the carbonyl functionality directly adjacent to the
phosphonic acid group, compounds of structure 1 were
supposed to be activated by irradiation with light providing
reactive radical species,[13] which might modify the target
protein covalently and thereby modulate its activity
(Scheme 1). To verify this hypothesis, benzoylphosphonates
1a,b were prepared by Arbuzov acylation of triethyl phos-
phite with the respective acyl chlorides followed by depro-
tection with trimethylsilyl bromide (TMSBr).[14] The UV
spectrum of compound 1a in water displayed the n–p*
transition at 368 nm in water with an extinction coefficient
e of 96 cmÀ1mÀ1 [15]
Irradiation of 1a,b dissolved in 2-propanol
.
and water, a solvent mixture capable of hydrogen radical
donation,[16] revealed a half-life of approximately 20 min. The
major irradiation products of 1a,b under these conditions
were the isomeric photodimers meso-2a,b and rac-2a,b,
corresponding to the meso and the racemic diols formed
from recombination of monomeric radical intermediates.
Minor by-products of the photoreaction of 1a,b included the
respective benzaldehydes 3a,b and benzoic acid derivatives
4a,b.
[*] M. Sc. S. Wagner,[+] Dipl.-Ing. J. Ortwein, Prof. Dr. J. Rademann
Medizinische Chemie, Institut fꢀr Pharmazie
Universitꢁt Leipzig
Brꢀderstrasse 34, 04103 Leipzig (Germany)
E-mail: rademann@uni-leipzig.de
Homepage: http://www.uni-leipzig.de/agrademann
Dr. A. Horatscheck,[+] Dr. B. G. Kim, Dr. M. Lisurek, Dr. S. Beligny,
Prof. Dr. J. Rademann
Abteilung fꢀr Medizinische Chemie
Without irradiation, compounds 1a,b remained stable in
2-propanol/water solution. Irradiation in a non-hydrogen-
radical-donating solvent like water led to only slow degrada-
tion of the benzoylphosphonate and provided no crosslinking
product. Addition of the tetramethylpiperidinyl N-oxide
radical (TEMPO) completely suppressed the formation of
photodimerization products in 2-propanol/water. Mercaptans
such as dithiothreitol and glutathione reduced the dimeriza-
tion efficiency.[17] Products of the photodimerization reaction
could be quantified when 4-methyl benzoylphosphonate (1b)
served as the starting material since they display to better
retention on the RP-18 column (Scheme 1, bottom). The half-
life of 1b was 15 min. After 40 min of irradiation, more than
90% of the recovered 2b was found in the dimers meso- and
rac-2b. The ratio of the two dimers was 2:1 in favor of the
Leibniz-Institut fꢀr Molekulare Pharmakologie (FMP)
Robert-Rçssle-Strasse 10, 13125 Berlin (Germany)
A. Schꢀtz
Helmholtz Protein Production Facility
Max-Delbrꢀck Center fꢀr Molekulare Medizin
Robert-Rçssle-Strasse 10, 13125 Berlin (Germany)
[+] These authors contributed equally to this work and are listed in
alphabetical order of their surnames.
[**] This work was supported by the DFG (FOR 806, SFB 765) and the
Investitionsbank Berlin, ProFit-Projekt 10143721. MptpA was
provided by Dr. Matthew Groves, EMBL-HH, STAT5b by Dr. Anne
Diehl, FMP, using a clone donated by Prof. Thorsten Berg, Leipzig.
We thank Natalja Erdmann, Janett Tischer, and Tracy Dornblut for
excellent technical assistance.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201201475.
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Figure 1. Docking experiments with benzoylphosphonate 1a in the
phosphotyrosine binding site of the protein tyrosine phosphatase from
Mycobacterium tuberculosis (MptpA) (A) and the SH2 domain of
STAT5b (modeled from STAT5a; B).
Binding predictions for the phosphatase MptpA were
verified by measuring the inhibitory activity of 1a. Employing
p-nitrophenol phosphate (pNPP) as the substrate, the enzyme
reaction of MptpA was monitored colorimetrically at 405 nm
by means of detection of the nitrophenoxide anion. Com-
pound 1a inhibited MptpA with an IC50 value of approx-
imately 1 mm. Taking into account a KM value of 2.3 mm for
pNPP, this IC50 value corresponds to a KI value of 186 mm,
assuming a competitive mode of inhibition.[18] Irradiation of
1a with the protein MptpA in tris(hydroxymethyl)amino-
methane (Tris) buffer for 45 min and subsequent measure-
ment of PTP activity resulted in a decreased IC50 value of
84 mm, indicating the deactivation of the phosphatase by
irradiated benzoylphosphonate. For comparison, irradiation
of phosphatase and substrate alone did not change the
activity.
Additionally, the binding of benzoylphosphonates 1 to
one of its target proteins was investigated by using a fluo-
rescently labeled derivative of benzoylphosphonate to visual-
ize the photocrosslinking product (Scheme 2). Starting from
4-aminomethylbenzoic acid, the pentynoic acid amide 7 of 4-
aminomethylbenzoyl phosphonate was prepared in three
steps (see the Supporting Information). Dipolar cycloaddition
with the 6-fluoresceinylcarboxamide derivative 6 of 3-azido-
prop-1-yl amine yielded triazole 8 under copper(I) cataly-
sis.[19] Triazole 8 was first investigated as a photoactive
inhibitor of PTP1B (Figure 2A). Inhibition was amplified
from an IC50 value of 1 mm to a value of 28 mm by irradiation
for 45 min. Next photocrosslinking products of 8 with protein
tyrosine phosphatase PTP1B were generated at different
concentrations in Tris buffer and were subsequently separated
Scheme 1. Top: Photoactivation and photocrosslinking of benzoyl
phosphonate 1a,b as phosphotyrosine bioisosters. Bottom: Products
of the reaction with the starting material 1b were quantified by LC–
MS. 3b: 4-methylbenzaldehyde, 4b: 4-methylbenzoic acid. ISC=inter-
system crossing.
kinetically and thermodynamically favored meso product. To
demonstrate whether crosslinking of benzoylphosphonate 1a
with peptides was feasible, the model peptide FKIAG-NH2
was irradiated in the presence of 1a for 4 h. LC/ToF-MS
measurements revealed the formation of a benzoyl adduct
only upon irradiation, which can be formed by crosslinking of
1a and subsequent elimination of phosphonic acid. It is not
clear whether elimination of phosphonic acid occurred during
irradiation in buffer or during ionization in the mass
spectrometer.
Next, benzoylphosphonate 1a was investigated in the
presence of the phosphotyrosine binding sites of Mycobacte-
rium tuberculosis protein tyrosine phosphatase A (MptpA)
and the SH2-containing transcription factor STAT5b as
model proteins. Docking experiments suggested favorable
binding interactions with the P-loop of MptpA, similar to
those predicted for phosphotyrosine (Figure 1A). Docking
experiments with the SH2 domain of STAT5b (modeled from
the crystal structure of STAT5a) also indicated a potential for
binding (Figure 1B). In both cases the carbonyl group acted
as hydrogen-bond acceptor of an amino acid side chain of
aspartic acid and asparagine, respectively.
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
These are not the final page numbers!

Angewandte
Chemie
Figure 2. A) Inhibition (IC50 values) of protein tyrosine phosphatase
PTP1B with benzoylphosphonate 8 in the enzymatic pNPP assay with
(gray) and without (black) irradiation. Bottom: B) Fluorescence of
PTP1B with 8 as detected in gel electrophoresis (PAGE) after irradi-
ation for 45 min. C) Same gel stained with Coomassie Blue. (For time-
dependent experiments see Figure S5 in the Supporting Information.)
Scheme 2. Preparation of the fluorescein-labeled benzoyl phosphonate
8. Reaction conditions: a) isobutyryl chloride, pyridine, CH2Cl2, 08C,
85%, after isomer separation the 6-isomer was obtained in 41% yield;
b) 3-azidoprop-1-yl amine, DCC, HOBt, THF, 88%; c) TFA, 608C, 93%;
d) pent-4-ynoic acid-OSu, Na2CO3, 87%; e) 1. (COCl)2, [DMF], CH2Cl2,
2. P(OBn)3,toluene, 75%; f) TFA, 08C, 92%; g) 1. 6, 7, NaAsc, CuSO4,
THF/H2O, basic buffered HPLC, 2. Amberlite IR-120 cation-exchange
resin, Na+ form, H2O, 40%. Asc=ascorbate, DCC=N,N’-dicyclohex-
ylcarbodiimide, HOBt=1-hydroxybenzotriazole, OSu=1-oxysuccini-
mide, TFA=trifluoroacetic acid.
affinity, crosslinking efficiency, and specificity of the photo-
active phosphotyrosine isoster toward its target proteins. In
order to test this hypothesis N-Fmoc-4-(O-benzyl-phospho-
nocarbonyl)phenylalanine building block 15 was designed for
the direct integration of the benzoylphosphonate into pep-
tides (Scheme 3).
Enantiomerically pure N-Boc-l-tyrosine methyl ester 9
was sulfonylated with N-phenylbis(trifluoromethylsulfonyli-
mide) yielding N-Boc-O-triflyltyrosine methyl ester 10.[20]
Compound 10 was carboxylated employing gaseous carbon
monoxide and palladium acetate with 1,3-bis(diphenylphos-
phino)propane (dppp) as a catalyst in 89% yield.[21] The
resulting N-Boc-4-carboxyphenylalanine methyl ester 11 was
N-deprotected with TFA and reprotected employing Fmoc-
succinimide (FmocOSu) to furnish 12, which was activated as
carboxylic acid chloride and used in a Michaelis–Arbuzov
acylation of tribenzyl phosphite to provide the 4-(O,O’-
dibenzyl-phosphonocarbonyl) derivative 13 of N-Fmoc-phe-
nylalanine methyl ester.[22] As the dibenzylphosphonate 13
was cleaved rapidly by piperidine[23] under the conditions used
for Fmoc removal, nucleophilic monodebenzylation of 13 was
accomplished with lithium bromide, yielding the monobenzyl
phosphonate 14.[24] The monobenzyl-protected benzoyl-
phosphonate was sufficiently stable toward the basic con-
ditions required during Fmoc cleavage. Next, the methyl ester
14 was saponified enzymatically with subtilisin Carlsberg
(“alcalase”) at pH 8.[25] The remaining protons in the two
acidic positions of the molecule were replaced by sodium ions
using sodium-loaded Amberlite IR-120 ion-exchange resin;
the resulting N-Fmoc-protected 4-(O-benzylphosphonocar-
by electrophoresis using a sodium dodecyl sulfate polyacryl-
amide gel (SDS-PAGE) (Figure 2B,C). For comparison, the
gel was stained with Coomassie Brilliant Blue in order to
visualize the entire protein. Fluorescently labeled PTP1B was
found at concentrations of triazole 8 of 125 mm and higher.
The fluorescence intensity correlated with the concentrations
of the fluorescent label and protein and confirmed the
covalent modification of the PTP1B protein. Fluorescent
protein bands were detected only after irradiation in the
presence of 15. The degree of covalent protein modification
was investigated at three irradiation times (20, 40, 60 min)
each at three concentrations (0.25, 0.5, and 1 mm), and protein
fluorescence was found to increase over time (see Figure S5 in
the Supporting Information). Prolonged irradiation times and
increased probe concentrations, however, led to a defocusing
of the protein band and therefore shorter irradiation at
moderate probe concentrations was preferable.
Phosphotyrosine(pTyr)-containing peptides are the
native and in many cases potent ligands of phosphotyrosine
receptors. Therefore, the incorporation of benzoylphospho-
nate into peptides using the unnatural amino acid 4-(phos-
phonocarbonyl)phenylalanine (pcPhe) as a building block
(Scheme 1) was hypothesized as a strategy to enhance the
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Scheme 4. Synthesis of phosphotyrosine peptides 16 and 17 and the
corresponding 4-(phosphonocarbonyl)phenylalanine derivatives 18 and
19 for use in binding assays and for covalent photomodification of
proteins.
Table 1: Binding affinities of tested phosphotyrosine- and benzoyl-
phosphonate-based peptides toward STAT5b.[a]
Peptide
KD, KI[a]
16
17
18
19
CF-GpYLSLPPW-NH2
Ac-pYLSLPPW-NH2
CF-GpcFLSLPPW-NH2
Ac-GpcFLSLPPW-NH2
0.042 (Æ0.003) mm
0.604 (Æ0.11) mm
0.896 (Æ0.44) mm
11 (Æ1.9) mm
Scheme 3. Reaction conditions: a) N-phenylbis(trifluoromethylsulfonyl-
imide), Et3N, CH2Cl2, 08C, 96%; b) dppp, Pd(OAc)2, CO(g), DIPEA,
DMF/H2O 3:1, 708C, 89%; c) TFA, CH2Cl2; FmocOSu, Na2CO3,
acetone/H2O, 08C, 87%; d) (COCl)2, DMF, CH2Cl2; P(OBn)3, toluene,
75%; e) LiBr, MeCN; f) subtilisin Carlsberg, aq. NH4HCO3, pH 8;
Amberlite IR-120, Na+ form, H2O, 89%. Bn=benzyl, Boc=tert-butoxy-
carbonyl, DIPEA=diisopropylethyl amine, Fmoc=9-fluorenylmethyl-
oxycarbonyl, FmocOSu=Fmoc-succinimide.
[a] Without irradiation. KI values were calculated from the IC50 values
with a formula by Nikolovska-Coleska et al.[31] pcF=4-phosphoncarbo-
nylphenylalanine.
of 16, peptide 17, was capable of competing with 16 for
binding of STAT5 and displayed a KI value of 604 nm.
Benzoylphosphonate 1a had an IC50 value of 840 mm which
corresponds to a KI value of 326 mm.[31] Irradiation of the
probe led to a significant increase of measured affinity over
time up to an IC50 of roughly 457 mm after 4 h (Table 2).
bonyl)phenylalanine building block 15 was suitable for use in
solid-phase peptide synthesis. Ion exchange was essential for
obtaining a stable product 15 ([a]D,25 8C =+ 25.08, c = 1.0,
H2O), which could be stored for months without racemization
or hydrolysis of the carbonyl–phosphonate bond.
As the first target protein for photoactive phosphotyro-
sine peptide mimetics, we selected the STAT5b protein, which
contains a phosphotyrosine-recognizing SH2 domain. In the
cellular context STAT5 is phosphorylated at a tyrosine
residue predominantly by JAK2 kinase,[26] which is associated
with several receptors including the erythropoietin[26,27] and
interleukin 2 receptors;[26,28,29] this reaction results in the
formation and subsequent dimerization of phospho-
STAT5.[26,29,30] Phospho-STAT5 dimers are transported into
the nucleus where they act as transcription factors. Constit-
utive activation or overexpression of STAT5 has been firmly
established for several types of leukemia. For our experi-
ments, STAT5b was recombinantly expressed conjugated with
maltose-binding protein (MBP) as a purification tag, resulting
in a protein construct of 120 kDa.[14]
Several phosphotyrosine peptides derived from STAT5-
binding proteins were synthesized and investigated as ligands
of STAT5. For binding studies the peptides were N-terminally
labeled with 5,6-carboxyfluorescein (CF). The octapeptide N-
CF-Gly-pTyr-Leu-Ser-Leu-Pro-Pro-Trp-NH2 (16) from the b-
subunit of the GM-CSF receptor was identified as the most
potent ligand of STAT5; it displayed a dissociation constant
(KD) of 42 nm (Scheme 4, Table 1). The N-acylated derivative
Table 2: IC50 values of benzoylphosphonate 1 against MptpA, PTP1B,
and STAT5b without and with irradiation.
Without irradiation
With irradiation
MptpA (1a)
STAT5b (1a)
PTP1B (1a)
PTP1B (15)
993 (Æ46) mm
840 (Æ126) mm
ca. 4 mm
1001 (Æ69) mm
84 (Æ9) mm[a]
457 (Æ47) mm[b]
8.9 (Æ0.3) mm
28 (Æ1) mm
[a] Irradiation for 45 min. [b] Irradiation for 240 min.
Next, 4-phosphonocarbonyl peptides 18 (R = N-fluores-
ceinyl-carboxy-glycinyl) and 19 (R = N-acetyl-glycinyl), the
hypothetic photoactive mimetics of phosphotyrosine peptides
16 and 17, were synthesized (Scheme 4). Fmoc-protected
amino acid building blocks were coupled on Rink amide
linker by diisopropyl carbodiimide/HOBt activation. The
protected 4-phosphonocarbonyl building block 15 and the
subsequent amino acid building block were coupled following
activation with O-(benzotriazol-1-yl)-N,N,N’,N’-tetramethyl-
uronium tetrafluoroborate (TBTU) and DIPEA; the cou-
pling proceeded with high efficiency as indicated by a negative
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
These are not the final page numbers!

Angewandte
Chemie
Kaiser test. After incorporation of 15, subsequent removal of
the Fmoc protecting groups was carried out with 2% 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) in DMF[32] (1 min,
10 min) in order to avoid partial cleavage of the phospho-
nate.[33] Peptides 18 and 19 were cleaved from the resin by
treatment with neat TFA. Thiol nucleophiles were avoided at
this point as they led to the formation of dithioketals.
Following the cleavage, TFA was removed by evaporation
and the crude peptides were dissolved in aqueous NH4HCO3.
For purification HPLC was conducted with a basic ammoni-
um bicarbonate buffer at pH 8 furnishing peptides 18 and 19
in yields of 29% and 36%, respectively. The products were
stable under mildly basic and neutral conditions.
In order to exclude intramolecular photoreactions of the
synthesized peptide probes 18 and 19, we compared their
photostability without the target protein. Samples were
irradiated in 2-(4-(2-hydroxyethyl)-1-piperazinyl)-ethansul-
fonic acid (HEPES) buffer at 48C and analyzed by HPLC
at different time points. No significant differences could be
detected. Next, binding of 18 to STAT5 protein was inves-
tigated and was found to proceed with a KD value of 0.9 mm, as
determined by a fluorescence polarization binding assay
(Table 1). The binding of peptide 19 was evaluated in
a competition experiment with ligand 16 and a KI value of
11 mm (IC50 = 28 mm) was measured for the former. Upon
irradiation a significant time-dependent decrease of the
measured IC50 could be observed, resulting in an IC50 value
of 13 mm after 4 h, while the inhibitory effect of phosphotyr-
osine peptide 17 remained unaffected.
Figure 3. Experiments to verify the specificity of the photolabeling of
MBP-STAT5b with the photoactive mimetic CF-GpcFLSLPPW-NH2 18.
A) Concentration-dependent photolabeling of equal concentrations of
STAT5b (black) and BSA (gray), respectively, with 18 as quantified
from the SDS polyacrylamide gel (for inverted fluorescence and
Coomassie pictures of the gel see Figures S8 and S9 in the Supporting
Information). B) Photolabeling of MBP-STAT5b at different concentra-
tions of 25 (in mm) without (left six lanes) and with protein denatura-
tion using 6m urea (see Figures S10 and S11 in the Supporting
Information). C) Photolabeling of BSA without (left six lanes) and with
protein denaturation using 6m urea (see Figures S10 and S11 in the
Supporting Information).
Having established the recognition of peptides 18 and 19
by the MBP–STAT5b protein construct, we investigated the
photocrosslinking potential of the ligand. For this purpose,
solutions of protein with 4-(phosphonocarbonyl)phenylala-
nine peptide at different concentrations in HEPES buffer
were irradiated in microtiter plates with 384 wells and
a transparent bottom with light at 365 nm from below using
a transilluminator. The resulting protein solutions were
analyzed by SDS polyacrylamide gel electrophoresis. Excita-
tion of the gel indicated labeling of the protein with
fluorescein in the case of irradiated samples, whereas non-
irradiated samples showed no fluorescence. When the gels
were stained with Coomassie Blue as a control, the amount of
protein and the appearance of the protein band were not
altered by irradiation. Labeling was quantified using a Lumi-
Imager and was found to be concentration-dependent with
respect to the protein and the photocrosslinking peptide 18
(see Figure 3A and Figures S6 and S7 in the Supporting
Information).
The photolabeling efficiency, given as the probe concen-
tration at half-maximum labeling, was determined to be
18 mm. Time-dependent labeling was investigated over a time
period of 120 min at different concentrations. At a low
concentration (2.5 mm) of the photoactive modifier 18 an
almost linear increase of the protein fluorescence was
observed over the entire period, whereas at high concen-
tration (40 mm) protein fluorescence went into saturation after
20–30 min (see Figures S14–S16 in the Supporting Informa-
tion).
In contrast, irradiation of 18 in the presence of bovine
serum albumine (BSA) as a control protein without phos-
photyrosine recognition sites resulted in unspecific labeling
which was not saturated and progressed linearly even at very
high concentrations of the photoreactive probe (Figure 3A
and Figures S8 and S9 in the Supporting Information). In
addition, the specificity of the photolabeling was verified by
denaturation of the target protein using 6m urea (Figure 3B,C
and Figures S10 and S11 in the Supporting Information). Only
correctly folded STAT5b was modified covalently, whereas
denatured STAT5, BSA, and denatured BSA were not
labeled. Finally, the specificity of the photolabeling was
further confirmed by using the nonfluorescent, phosphotyr-
osine-containing control peptide 17 as a competing molecule;
17 suppressed the photolabeling with peptide 18 in a concen-
tration-dependent manner while it had no effect on the
unspecific photolabeling (Figure 4 and Figures S12 and S13 in
the Supporting Information).
In summary, we have demonstrated that benzoylphosph-
onates are photoactive bioisosters of phosphotyrosine resi-
dues. Likewise, 4-(phosphonocarbonyl)phenylalanine-con-
taining peptides were proven to be photoactive mimetics of
native phosphotyrosine peptides. Both the bioisosteric frag-
ments and the peptide mimetics bind reversibly to phospho-
tyrosine recognizing protein domains. Upon irradiation with
light, photoactivation of the benzoylphosphonates resulted in
the functional deactivation and the covalent modification of
the target proteins STAT5b, PTP1B, and MptpA. The degree
of covalent labeling correlated with concentration and
irradiation time. Most significantly, phosphotyrosine peptide
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
[5] a) T. R. Burke, Z. Yao, D. Liu, J. Voigt, Y. Gao, Biopolymers
2001, 60, 32 – 44; b) M. A. T. Blaskovich, Curr. Med. Chem. 2009,
16, 2095 – 2176.
[6] a) Y. Hatanaka, Y. Sadakane, Curr. Top. Med. Chem. 2002, 2,
271 – 288; b) M. Thꢀrien, K. Skorey, R. Zamboni, C. S. Li, C. K.
Lau, T. LeRiche, V. Linh Truong, D. Waddleton, C. Ramachan-
dran, Bioorg. Med. Chem. Lett. 2004, 14, 2319 – 2322; c) M.
Sydnes, M. Kuse, M. Kurono, A. Shimomura, H. Ohinata, A.
Takai, M. Isobe, Bioorg. Med. Chem. 2008, 16, 1747 – 1755.
[7] a) I. Marseigne, B. P. Roques, J. Org. Chem. 1988, 53, 3621 –
3624; b) S. M. Domchek, K. R. Auger, S. Chatterjee, T. R.
Burke, Jr., S. E. Shoelson, Biochemistry 1992, 31, 9865 – 9870.
[8] a) 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; b) T. R. Burke, Jr., H. K. Kole, P. P. Roller, Bio-
chem. Biophys. Res. Commun. 1994, 204, 129 – 134; c) T. R.
Burke, Jr., Curr. Top. Med. Chem. 2006, 6, 1465 – 1471.
[9] a) A. Combs et al., J. Med. Chem. 2005, 48, 6544 – 6548; b) P. J.
Ala et al., J. Biol. Chem. 2006, 281, 32784 – 32795.
[10] M. F. Schmidt, M. Groves, J. Rademann, ChemBioChem 2011,
12, 2640 – 2646.
[11] a) M. Schmidt, A. Isidro-Llobet, A. El-Dahshan, M. Lisurek, J.
Tan, R. Hilgenfeld, J. Rademann, Angew. Chem. 2008, 120,
3319 – 3323; Angew. Chem. Int. Ed. 2008, 47, 3275 – 3278; b) M.
Schmidt, A. El-Dahshan, S. Keller, J. Rademann, Angew. Chem.
2009, 121, 6464 – 6467; Angew. Chem. Int. Ed. 2009, 48, 6346 –
6349; c) M. Schmidt, J. Rademann, Trends Biotechnol. 2009, 27,
512 – 521.
[12] D. Krishnamurthy, A. M. Barrios, Curr. Opin. Chem. Biol. 2009,
13, 375 – 381.
[13] For photodimerization reactions of ketophosphonate esters: K.
Terauchi, H. Sakurai, Bull. Chem. Soc. Jpn. 1970, 43, 883 – 890.
[14] G. D. Kishore Kumar, D. Saenz, G. L. Lokesh, A. Natarajan,
Tetrahedron Lett. 2006, 47, 6281 – 6284.
[15] For spectral studies of ketophosphonate esters: K. Terauchi, H.
Sakurai, Bull. Chem. Soc. Jpn. 1969, 42, 821 – 823.
[16] J. N. J. Pitts, R. L. Letsinger, R. P. Taylor, J. M. Patterson, G.
Recktenwald, R. B. Martin, J. Am. Chem. Soc. 1959, 81, 1068 –
1077.
[17] The same effect was reported for benzophenone: S. Cohen, W.
Sherman, J. Am. Chem. Soc. 1963, 85, 1642 – 1647.
[18] Y.-C. Cheng, W. H. Prusoff, Biochem. Pharmacol. 1973, 22,
3099 – 3108.
Figure 4. Competitive displacement of PAM probe 18 (PAM=photoac-
tive mimetic) by acetylated phosphotyrosine peptide 17 results in the
suppression of labeling. Photolabeling was quantified from the
inverted fluorescence image of the SDS-polyacrylamide gel. (For the
entire fluorescence picture see Figure S12, for Coomassie control
staining see Figure S13).
mimetics labeled their target proteins with high specificity for
the phosphotyrosine binding site. The photochemical ligation
reaction was saturated in a concentration-dependent manner,
and the labeling could be suppressed by protein denaturation
and by using the native phosphotyrosine peptide as a compet-
ing ligand.
To our knowledge this work constitutes the first example
of a photoactive bioisoster, a chemical moiety that combines
photocrosslinking activity and biomimicry in one bioisosteric
group. Photoactive bioisosters might find applications in
functional cell biology, bioanalytics, and proteomic investiga-
tions.
[19] a) M. Meldal, C. W. Tornøe, C. Christensen, J. Org. Chem. 2002,
67, 3057 – 3064; b) V. V. Rostovtsev, L. G. Green, V. V. Fokin,
K. B. Sharpless, Angew. Chem. 2002, 114, 2708 – 2711; Angew.
Chem. Int. Ed. 2002, 41, 2596 – 2599.
[20] J. W. Tilley, R. Sarabu, R. Wagner, K. Mulkerins, J. Org. Chem.
1990, 55, 906 – 910.
[21] Reaction conditions modified from: J. B. Grimm, K. J. Wilson,
D. J. Witter, Tetrahedron Lett. 2007, 48, 4509 – 4513.
[22] Reaction conditions modified from: J. Wrobel, A. Dietrich,
Tetrahedron Lett. 1993, 34, 3543 – 3546.
Received: February 22, 2012
Revised: May 31, 2012
Published online: && &&, &&&&
Keywords: phosphotyrosine isosters · photoaffinity labeling ·
.
[23] M. S. Sekine, M. Satoh, H. Yamagata, T. Hata, J. Org. Chem.
1980, 45, 4162 – 4167.
photochemistry · proteins · protein tyrosine phosphatases
[24] Reaction conditions as described in: R. Karaman, A. Goldblum,
E. Breuer, J. Chem. Soc. Perkin Trans. 1 1989, 765 – 774.
[25] S. T. Chen, W.-H. Huang, K.-T. Wang, Chirality 1994, 6, 572 –
576. The procedure had to be modified as reported in the SI.
[26] P. M. Grimley, F. Dong, H. Rui, Cytokine Growth Factor Rev.
1999, 10, 131 – 157.
[27] F. W. Quelle, D. Wang, T. Nosaka, W. E. Thierfelder, D.
Stravopodris, Y. Weinstein, J. E. Ihle, Mol. Cell. Biol. 1996, 16,
1622 – 1631.
[1] M. A. Lemmon, J. Schlessinger, Cell 2010, 141, 1117 – 1134.
[2] T. Pawson, J. D. Scott, Science 1997, 278, 2075 – 2080.
[3] a) A. Kamat, G. Carpenter, Cytokine Growth Factor Rev. 1997, 8,
109 – 117; b) J. D. Bjorge, A. Jakymiw, D. J. Fujita, Oncogene
2000, 19, 5620 – 5635; c) B. G. Neel, H. Gu, L. Pao, Trends
Biochem. Sci. 2003, 28, 284 – 293.
[28] Z.-Y. Zhang, Annu. Rev. Pharmacol. Toxicol. 2002, 42, 209 – 234.
[29] a) J. Hou, U. Schindler, W. J. Henzel, S. C. Wong, S. L.
McKnight, Immunity 1995, 2, 321 – 329; b) H. Wakao, N.
[4] D. E. Levy, J. E. Darnell, Nat. Rev. Mol. Cell Biol. 2002, 3, 651 –
662.
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
These are not the final page numbers!

Angewandte
Chemie
Harada, T. Kitamura, A. L.-F. Mui, A. Miyajima, EMBO J. 1995,
14, 2527 – 2535.
[32] J. D. Wade, J. Bedford, R. C. Sheppard, G. W. Tregear, Pept. Res.
1991, 4, 194 – 199.
[30] a) A. Meinke, F. Barahmand-Pour, S. Wçhrl, D. Stoiber, T.
Decker, Mol. Cell. Biol. 1996, 16, 6937 – 6944; b) N. Cella, B.
Groner, N. E. Hynes, Mol. Cell. Biol. 1998, 18, 1783 – 1792.
[31] Z. Nikolovska-Coleska, R. Wang, X. Fang, H. Pan, Y. Tomita, P.
Li, P. P. Roller, K. Krajewski, N. G. Saito, J. A. Stuckey, S. Wang,
Anal. Biochem. 2004, 332, 261 – 273.
[33] Fmoc removal with piperidine/DMF 20:80 resulted in partial
cleavage of the carbonyl–phosphonate bond and the formation
of piperidine amide, typically in the range of 20% after three
cycles of Fmoc removal.
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!

.
Angewandte
Communications
Communications
Photoactive Bioisosters
A light switch for phosphotyrosine-rec-
ognizing proteins: Irradiation of the bio-
isosteric benzoylphosphonate suffices to
“turn off” the activity of target proteins
and to label them covalently (see
A. Horatscheck, S. Wagner, J. Ortwein,
B. G. Kim, M. Lisurek, S. Beligny,
A. Schꢁtz, J. Rademann*
&&&& — &&&&
scheme). Photoactive bioisosters may
find applications in functional cell biol-
ogy, bioanalytics, and proteome research.
Benzoylphosphonate-Based Photoactive
Phosphopeptide Mimetics for
Modulation of Protein Tyrosine
Phosphatases and Highly Specific
Labeling of SH2 Domains
8
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 8
These are not the final page numbers!