Synthetic receptors for the differentiation of phosphorylated peptides with nanomolar affinities

Article abstract of DOI:10.1002/chem.200800432

Artificial ditopic receptors for the differentiation of phosphorylated peptides varying in i + 3 amino acid side chains were synthesized, and their binding affinities and selectivities were determined. The synthetic receptors show the highest binding affi

Full text of DOI:10.1002/chem.200800432

DOI: 10.1002/chem.200800432
Synthetic Receptors for the Differentiation of Phosphorylated Peptides
with Nanomolar Affinities
Andreas Grauer, Alexander Riechers, Stefan Ritter, and Burkhard Kçnig*[a]
Abstract: Artificial ditopic receptors
for the differentiation of phosphorylat-
ed peptides varying in i+3 amino acid
side chains were synthesized, and their
binding affinities and selectivities were
determined. The synthetic receptors
show the highest binding affinities to
phosphorylated peptides under physio-
logical conditions (HEPES, pH 7.5,
154 mm NaCl) reported thus far for ar-
tificial systems. The tight and selective
binding was achieved by high coopera-
tivity of the two binding moieties in
the receptor molecules. All receptors
interact with phosphorylated serine by
bis(ZnII–cyclen) complex coordination
and a second binding site recognizing a
carboxylate or imidazole amino acid
side chain functionality.
Keywords: chelates · cooperativity ·
emission spectroscopy · molecular
recognition · peptides
Introduction
tended the concept to a hybrid receptor[10] consisting of a
natural WW binding site (a protein module that binds pro-
line-rich ligands)[11] for phosphorylated peptides and a fluo-
rescent metal complex. Their hybrid receptor[12] shows en-
hanced micromolar binding affinity and selectivity towards
diphosphorylated peptides derived from sequences of the C-
terminal domain of the RNA polymerase II.
We report herein the design, synthesis and binding prop-
erties of synthetic ditopic receptors with nanomolar affinity
to phosphorylated peptides in buffered aqueous solution
under physiological pH. Their affinity depends on a second
amino acid residue present in the peptide, namely carboxyl-
ate or histidine.
Phosphorylation of proteins is a ubiquitous regulation mech-
anism in biology and plays a central role in controlling intra-
cellular signaling networks.[1] Protein phosphorylation regu-
lates enzyme activity and the reversible formation of signal-
ing complexes by specific recognition of phosphorylated
proteins by phosphoprotein binding domains.[2] To investi-
gate, monitor or specifically inhibit such phosphorylation-
dependent processes, analytical tools allowing a specific rec-
ognition of phosphorylated peptides and proteins are desira-
ble. Historically, phosphorylation detection methods relied
on radioisotopes or phosphoamino acid selective antibodies.
Chromatographic, staining and surface device techniques ex-
tended the available methods.[3] Several biosensors for pro-
tein kinase activity based on GFP-FRET probes[4–6] or syn-
thetic fluorophores[7] have been reported, which typically
signal their own phosphorylation or dephosphorylation.
However, the number of artificial systems for a specific rec-
ognition of phosphorylated peptides is still limited. Hamachi
et al.[8,9] recently reported fluorescent dinuclear zinc com-
plexes for the detection of peptide phosphorylation and ex-
As target peptides Flu-GpSAAEV-NH2 (1) and Flu-
GpSAAHV-NH2 (2) were selected from sequences of
human STAT (signal transducer and activator of transcrip-
tion) proteins[13] and prepared by standard solid-phase
synthesis methods. The peptides were N-terminally lab-
eled by fluorescein (Flu) to facilitate the binding studies. A
bisACHTNUTGRNEUNG
(ZnII–cyclen) triazine complex was used in the receptor
design as the binding site for phosphoserine. The high affini-
ty of this complex to phosphates was previously shown on
modified surfaces.[14] Peptide 1 presents, beside the phos-
phate ester, as an additional functional group for specific
molecular recognition a carboxylate in the sidechain of the
i+3 glutamic acid. Guanidinium moieties have been used as
binding sites for carboxylate binding.[15] Thus, to achieve se-
lective affinity for target peptide 1, both binding sites for
phosphoserine and the carboxylate side chain were connect-
ed by a peptidic linker giving compounds 3 and 4. Receptors
5 and 6 consist of two bis(ZnII–cyclen) triazine complexes
[a] Dipl.-Chem. A. Grauer, Dipl.-Chem. A. Riechers, Dr. S. Ritter,
Prof. Dr. B. Kçnig
Institute for Organic Chemistry
University of Regensburg, Universitꢀtsstr. 31
93040 Regensburg (Germany)
Fax : (+49)941-943-1717
E-mail: burkhard.koenig@chemie.uni-regensburg.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800432.
8922
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8922 – 8927

FULL PAPER
tethered by a peptidic linker. X-ray structure analysis of imi-
dazole–ZnII–cyclen cocrystals (see Supporting Informa-
tion)[16] and potentiometric titrations[17] have shown that the
complex can well accommodate imidazole as a guest. There-
fore binding specificity for peptide 2 was expected. The
third group of synthetic receptors 7 and 8 contains a ZnII–
NTA (NTA = nitrilotriacetic acid) complex beside the
bis(ZnII–cyclen) triazine complex. Unlike CuII- and NiII–
NTA complexes which bind imidazole and are widely used
for purification of His-tagged proteins by immobilized metal
affinity chromatography (IMAC),[18,19] the binding affinity of
ZnII–NTA complexes for imidazole is significantly re-
duced.[20] However, the NTA ligand also represents a trun-
cated EDTA motif, known to bind ZnII with an affinity of
logK=16.5.[21] Accordingly, it can be expected that a carbox-
ylate can coordinate to the unoccupied coordination sites in
the NTA complex intramolecularly, thus completing a mimic
of an EDTA coordination sphere.[22] Intermolecularly, this
interaction has already been described.[23] For receptors 7
and 8 (Figure 1), the two ZnII complexes were connected by
an alkyne azide cycloaddition[24] giving a triazole linker. All
details of the synthesis and the spectroscopic characteriza-
tion of receptors 3–8 are provided in the Supporting Infor-
mation of this article.
Materials and methods
Synthesis of receptors 3–8: For the synthesis of the com-
plexes 3 and 4 the cyclen building block precursor 11 and
the two guanidine building block precursors 12 and 13 were
coupled using standard peptide coupling chemistry with
EDC, HOBt and DIPEA as reagents. Compound 11 was
synthesized by a substitution on the literature known tri-
ACHTUNGTRENNUNG
azene 14[13a] using benzyl-2-aminoethylcarbamate (35). The
building blocks 12 and 13 were prepared by a substitution
reaction on 1,3-bis(tert-butyloxycarbonyl)-2-methyl-2-thio-
ACHTUNGTRENNUNGurea (38) with H-Gly-OMe·HCl (37) for the first and H-
Gly-Gly-OMe·HCl (41) for the latter. The resulting methyl
esters were cleaved with LiOH and after workup with an
aqueous KHSO4 solution the free acids were obtained and
used for the amide bond formation. The Boc-protecting
groups of the protected precursors of 3 and 4 were cleaved
with a saturated solution of HCl in diethyl ether. The am-
monium salts precipitated from solution quantitatively and
were deprotonated using a strongly basic anion exchanger in
its hydroxide ion form. After lyophilization, the ligands
were complexed with ZnII perchlorate and the receptors 3
and 4 were obtained after recrystallization from a methanol
water mixture.
The protected precursors of the receptors 5 and 6 were
obtained by a two-fold substitution of the previously pre-
pared bisamines 15 and 16 with the cyclen building block 14.
The amines were synthesized starting from tert-butyl 2-ami-
noethylcarbamate (45), which was coupled with Boc-Gly-
OH (44) in the first and Boc-Gly-Gly-OH (48) in the latter
case by standard peptide coupling conditions using EDC,
HOBt and DIPEA as reagents. After the deprotection with
Figure 1. Phosphorylated peptides 1 and 2, and synthetic receptors 3–8
for their binding.
Chem. Eur. J. 2008, 14, 8922 – 8927
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8923

B. Kçnig et al.
a saturated solution of HCl in
diethyl ether the precipitated
ammonium salts were used for
the substitution reaction with-
out further purification. The re-
sulting precursors of 5 and 6
were deprotected using a satu-
rated solution of HCl in diethyl
ether. The ammonium salts
were deprotonated by the use
of a strongly basic ion exchang-
er column in its hydroxide ion
form. After lyophilisation the
ligands were complexed with
ZnII perchlorate and the recep-
tors 5 and 6 were obtained after
recrystallization from a metha-
nol water mixture.
Scheme 1 shows exemplary
the synthesis of the protected
precursor of 7. For the receptor
7 the cyclen building block pre-
cursor 17 and the NTA-building
block precursor 18 were used in
a CuI-mediated Huisgen azide
alkyne
cycloaddition
with
sodium ascorbate and CuSO4.
The copper- and ascorbate salt
were not used in catalytic
amounts as this proved to give
higher yields after shorter reac-
tion times. As the reduction of
Scheme 1. Synthesis of the synthetic phosphopeptide receptors.
CuII by sodium ascorbate
caused a lowering of the pH
value, the reaction was carried out in buffered solution (ace-
tate buffer, pH 5, c=0.5m). Subsequently, the Boc-protect-
ing groups were cleaved with a saturated solution of HCl in
diethyl ether. The ammonium salts precipitated from solu-
tion quantitatively and were deprotonated using a strongly
basic anion exchanger in its hydroxide ion form. Cleavage
of the ester groups during this procedure was not observed.
The obtained aqueous solutions of the free amines were
treated with a stoichiometric amount of LiOH to cleave the
benzyl, methyl and ethyl ester functionalities. Lyophilisation
of these solutions gave the polydentate ligands in quantita-
tive yield, ready for complexation with ZnII perchlorate.
ted by the concentration range) were recorded. The stoi-
chiometries were determined by Jobꢂs plot analysis. To de-
termine the binding constants, the obtained fluorescence in-
tensities at 520 nm were volume corrected, plotted against
the receptor concentration and evaluated by nonlinear fit-
ting. Fluorescence polarization titrations were conducted
under identical conditions using an ISS K2 Multifrequency
Phase Fluorometer.
Results and Discussion
Binding constants of receptors 3–8 to peptides 1 and 2 were
determined by fluorescence emission titrations in aqueous
solution (HEPES, pH 7.5, 50 mm, 154 mm NaCl) and non-
linear fitting of the data (see Supporting Information). Jobꢂs
plots were used to determine the binding stoichiometry,
which was found to be 1:1 for all experiments. For compari-
son the affinity of complexes 9 and 10 (Figure 2), represent-
ing the single binding sites, to peptides 1 and 2 were mea-
sured. Table 1 summarizes the results, while Figures 3 and 4
illustrate the binding selectivity and show exemplary emis-
sion titrations.
Binding studies
All binding studies were conducted in buffered aqueous so-
lution (50 mm HEPES, pH 7.5, 154 mm NaCl). A Varian
Cary Eclipse Fluorometer was used for the emission titra-
tions. The cuvette with 800 mL of peptide in HEPES buffer
was titrated stepwise with small amounts (beginning with
0.13 equiv) of the receptor solution. After each addition the
solution was allowed to equilibrate for 2 min before the
fluorescence intensity and the UV spectrum (where permit-
8924
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8922 – 8927

Synthetic Receptors
FULL PAPER
Table 1. Binding affinities of complexes 3–10 to peptides 1 and 2.
Binding affinity [logK]
receptor
peptide 2
peptide 1
3
4
5
6
7
8
9
10
4.5
5.8
7.5
6.5
4.8
4.6
n.d.
4.8
8.0/8.0[a]
7.8
5.0
4.9
8.0/8.4[a]
7.9
< 3
4.8
[a] Reference values from fluorescence polarization titrations.
Figure 3. Emission response of peptides 1 and 2 in the presence of recep-
tors 3 and 5. The logarithmic x axis allows the depiction of the binding
selectivity in orders of magnitude of the concentration range.
Figure 2. Complexes 9 and 10 representing receptor substructures.
Receptors 3 and 4 show affinities for Flu-GpSAAEV-NH2
(1) of logK=8.0 and 7.8, respectively, while the binding to
peptide 2 is two to three orders of magnitude less (logK=
4.5 and 5.8, respectively). This is in accordance with our ex-
pectations, as the interaction of the guanidine binding site
with the glutamate carboxylate of peptide 1 is significantly
stronger than with the imidazole side chain of peptide 2 due
to electrostatic attraction. The reverse binding selectivity is
observed for tetra(ZnII–cyclen) receptors 5 and 6: The histi-
dine-containing peptide 2 is bound one to two orders of
magnitude stronger (logK=7.5 and 6.5, respectively) than
peptide 1 (logK=5.0 and 4.9, respectively). The interaction
of the second bis(ZnII–cyclen) triazine with the imidazole
side chain contributes to the aggregateꢂs stability, although
the selectivity is not as pronounced as in the case of com-
plexes 3 and 4. Additional entropic stabilization comes from
the bivalent structure of 5 and 6 with two identical binding
sites, in our eyes also an example of positive cooperativity in
enthalpy.[25] Bis(ZnII–cyclen)–ZnII–NTA receptors 7 and 8
again show a pronounced selectivity towards peptide 1 with
logK=8.0 and 7.9, respectively. The binding to peptide 2 is
about one thousand fold weaker (logK=4.8 and 4.6, respec-
tively) and we attribute the selectivity to the interaction of
ZnII–NTA with the glutamate carboxylate. Generally, we
find a strong cooperativity of binding in the “matched”
cases, a behavior which has been described before for artifi-
cial receptors.[26]
Figure 4. Emission titrations of receptor 3 against peptides 1 (top) and 2
(bottom). The continuous line represents the Hill equation fit.
strength (logK=4.8), which shows that the interaction of
the phosphate ester with the bis(ZnII–cyclen) binding site is
not affected by the peptide sequence. The binding affinities
of receptors 3 and 4 to peptide 2 are similar to this value.
This leads to the conclusion that interactions of the guanidi-
nium moiety to peptide 2 are negligible. The same applies to
the interaction of receptors 7 and 8 with peptide 1: The
ZnII–NTA–imidazole interaction does not contribute to the
receptor affinity as the ZnII–NTA–carboxylate binding does.
This is confirmed by the binding data of ZnII–NTA complex
A
bisACHTUNGTRENNUNG
comparison of the peptide binding affinities of
(ZnII–cyclen) triazine complex 10 to receptors 3–8 reveals
the contribution of the second binding site to the overall af-
finity. Complex 10 binds to both peptides with identical
Chem. Eur. J. 2008, 14, 8922 – 8927
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8925

B. Kçnig et al.
9 to peptides 1 and 2. A small, but significant interaction is
observed with peptide 1, while no interaction is determined
with peptide 2.
[1] a) L. N. Johnson, R. J. Lewis, Chem. Rev. 2001, 101, 2209–2242;
b) T. Hunter, Cell 2000, 100, 113–127.
[2] M. B. Yaffe, Nat. Rev. Mol. Cell Biol. 2002, 3, 177–186.
[3] a) A. Ojida, T. Kohira, I. Hamachi, Chem. Lett. 2004, 1024–1025;
b) K. Martin, T. H. Steinberg, L. A. Cooley, K. R. Gee, J. M. Bee-
cham, W. F. Patton, Proteomics 2003, 3, 1244–1255; c) E. Kinoshita,
E. Kinoshita-Kikuta, K. Takiyama, T. Koike, Mol. Cell. Proteomics
2006, 5, 749–757; d) E. Kinoshita, E. Kinoshita-Kikuta, K. Takiya-
ma, T. Koike, Mol. Cell. Proteomics 2006, 5, 749–757; e) S. Yamada,
H. Nakamura, E. Kinoshita, E. Kinoshita-Kikuta, T. Koike, Y. Shiro,
Anal. Biochem. 2007, 360, 160–162; f) E. Kinoshita-Kikuta, Y.
Aoki, E. Kinoshita, T. Koike, Mol. Cell. Proteomics 2007, 6, 356–
366; g) R. Freeman, R. Gill, I. Willner, Chem. Commun. 2007,
3450–3452; h) H.-Y. Lin, C.-T. Chen, Y.-C. Chen, Anal. Chem. 2006,
78, 6873–6878.
To verify the results from the emission titrations by an in-
dependent method, fluorescence polarization titrations were
conducted under identical conditions as the emission titra-
tions. The two values determined for receptors 3 and 7 are
in good agreement with the results from the emission titra-
tions. Fluorescence labels may falsify the binding results by
contributing to the peptide affinity. To exclude interactions
of the peptide label with the receptors, fluorescein sodium
was titrated with a receptor of each binding motif showing
no emission changes even at large receptor excess.
The length or type of the linkers connecting the two re-
ceptor binding sites has little or no influence on the binding
affinity. This is not surprising as the short peptides display
no stable secondary structure in solution and the receptors
linkers are flexible. The binding selectivity and affinity of
the receptor molecules rely on the presence of complemen-
tary functional groups for non-covalent interaction in reach
and not on their exact spatial position.
[4] S. Moritoshi, O. Takeaki, I. Kouichi, A. Tomoichiro, U. Yoshio, Nat.
Biotechnol. 2002, 20, 287–294.
[5] K. Kurokawa, N. Mochizuki, Y. Ohba, H. Mizuno, A. Miyawaki, M.
Matsuda, J. Biol. Chem. 2001, 276, 31305–31310.
[6] a) A. Y. Ting, K. H. Kain, R. L. Klemke, R. Y. Tsien, Proc. Natl.
Acad. Sci. USA 2001, 98, 15003–15008; b) J. Zhang, Y. Ma, S. S.
Taylor, R. Y. Tsien, Proc. Natl. Acad. Sci. USA 2001, 98, 14997–
15002; c) M. Sato, T. Ozawa, K. Inukai, T. Asano, Y. Umezawa, Nat.
Biotechnol. 2002, 20, 287–294.
[7] a) M. D. Shults, B. Imperiali, J. Am. Chem. Soc. 2003, 125, 14248–
14249; b) D. M. Rothman, M. D. Shults, B. Imperiali, Trends Cell
Biol. 2005, 15, 502–510; c) B. R. Sculimbrene, B. Imperiali, J. Am.
Chem. Soc. 2006, 128, 7346–7352.
Conclusion
[8] a) A. Ojida, Y. Mito-Oka, M. Inoue, I. Hamachi, J. Am. Chem. Soc.
2002, 124, 6256–6258; b) A. Ojida, M. Inoue, Y. Mito-Oka, I. Ha-
machi, J. Am. Chem. Soc. 2003, 125, 10184–10185.
[9] A. Ojida, Y. Mito-oka, K. Sada, I. Hamachi, J. Am. Chem. Soc.
2004, 126, 2454–2463.
[10] T. Anai, E. Nakata, Y. Koshi, A. Ojida, I. Hamachi, J. Am. Chem.
Soc. 2007, 129, 6232–6239.
[11] P.-J. Lu, X. Z. Zhou, Y.-C. Liou, J. P. Noel, K. P. Lu, J. Biol. Chem.
2002, 277, 2381–2384.
[12] Other ditopic receptors recognizing non-phosphorylated peptides
have been reported, for example: a) A. Hossain, H.-J. Schneider, J.
Am. Chem. Soc. 1998, 120, 11208–11209; b) L. Hernandez-Folgado;
C. Schmuck, Org. Biomol. Chem. 2007, 5, 2390–2394.
[13] C. P. Lim, X. Cao, Mol. Biosyst. 2006, 2, 536–550.
[14] a) D. S. Turygin, M. Subat, O. A. Raitman, V. V. Arslanov, B.
Koenig, M. A. Kalinina, Angew. Chem. 2006, 118, 5466–5470;
Angew. Chem. Int. Ed. 2006, 45, 5340–5344; b) D. S. Turygin, M.
Subat, O. A. Raitman, S. L. Selector, V. V. Arslanov, B. Kçnig,
M. A. Kalinina, Langmuir, 2007, 23, 2517–2524; c) H. Fujioka, T.
Koike, N. Yamada, E. Kimura, Heterocycles 1996, 42, 775–787.
[15] a) C. Schmuck, Coord. Chem. Rev. 2006, 250, 3053–3067; b) R. J.
Fitzmaurice, F. Gaggini, N. Srinivasan, K. Natarajan, D. Jeremy,
Org. Biomol. Chem. 2007, 5, 1706–1714; c) C. Schmuck, Chem. Eur.
J. 2000, 6, 709–718.
The combination of bis(ZnII–cyclen)–triazine metal complex
binding sites with guanidinium moieties or ZnII–NTA com-
plexes leads to artificial receptors for the differentiation of
phosphorylated small peptides. Using the right combination
of binding moieties, nanomolar peptide binding affinities at
physiological conditions are achieved. To the best of our
knowledge these are the highest affinities of phosphopeptide
binding by artificial receptors reported thus far. Depending
on a second functional group beside the phosphate ester se-
lectivities of up to three orders of magnitude are observed.
To better resolve the affinity of artificial receptors and pep-
tides in the distance of the involved functional groups, the
binding properties of more rigidified receptors and cyclo-
peptides are currently investigated and results will be re-
ported in due course. Furthermore, projects employing these
recognition moieties in molecular biology are under way.
Experimental Section
[16] C. Pariya, T.-Y. Chi, T. K. Mishra, C. S. Chung, Inorg. Chem.
Commun. 2002, 5, 119–123.
[17] M. Kruppa, D. Frank, H. Leffler-Schuster, B. Kçnig, Inorg. Chim.
Acta 2006, 359, 1159–1168.
Synthetic details and titration plots are available in the Supporting Infor-
mation.
[18] M. Y. Wang, Y. Y. Kau, M. S. Lee, S. R. Doong, J. Y. Ho, L. H. Lee,
Biotechnol. Bioeng. 2000, 67, 104–111.
[19] M. Kruppa, B. Kçnig, Chem. Rev. 2006, 106, 3520–3560.
[20] J. Volz, J. Chromatogr. A 1998, 800, 29–37.
[21] R. Delgado, M. de Fꢄtima Cabral, R. Castanheira, A. Zhang, R.
Herrmann, Polyhedron 2002, 21, 2265–2276.
Acknowledgements
[22] Although NTA–ZnII complexes are known to be chelated by His-
tagged proteins, these interactions are based on the chelating effect
of two imidazoles occupying both remaining coordination sites. A
single histidine residue is not sufficient for a binding event. See also
ref. [20].
We thank the Volkswagen Foundation, the Fonds der Chemischen Indus-
trie, the University of Regensburg and the Graduiertenkolleg GRK 760
for support of this research. We further thank Axel Dꢃrkop for help with
the fluorescence polarization measurements and Chiara Cabrele for
advice and valuable discussions.
8926
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8922 – 8927

Synthetic Receptors
FULL PAPER
[23] a) D. Hopgood, R. J. Angelici, J. Am. Chem. Soc. 1968, 90, 2508–
2513; b) D. L. Rabenstein, G. Blakney, Inorg. Chem. 1973, 12, 128–
132.
[26] G. J. Kirkovits, J. A. Shriver, P. A. Gale, J. L. Sessler, J. Inclusion
Phenom. Macrocyclic Chem. 2001, 41, 69–75.
[24] J. E. Moses, A. D. Moorhouse, Chem. Soc. Rev. 2007, 36, 1249–1262.
[25] D. H. Williams, M. S. Westwell, Chem. Soc. Rev. 1998, 27, 57–63.
Received: March 10, 2008
Published online: August 11, 2008
Chem. Eur. J. 2008, 14, 8922 – 8927
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8927