Light-directed protein binding of a biologically relevant β-sheet

Full text of DOI:10.1002/anie.200901933

Communications
DOI: 10.1002/anie.200901933
Biphotonics
Light-Directed Protein Binding of a Biologically Relevant b-Sheet**
Christian Hoppmann, Sabine Seedorff, Anja Richter, Heinz Fabian, Peter Schmieder,
Karola Rꢀck-Braun,* and Michael Beyermann*
b-Hairpin structures are frequently involved in protein–
protein interactions that control essential processes in cells
and are therefore interesting targets for interference. Hairpin-
forming peptides that compete with such protein interactions
are valuable tools for studying biological processes. More-
over, the incorporation of a photoswitchable unit into
appropriate b-hairpin-forming peptide ligands could allow
protein interactions in cells to be studied by light-triggered
interference. However, b-hairpin structures are rarely studied
because of the limited availability and stability of suitable
model peptides. This is because such a model peptide has to
fulfill at least three requirements: 1) the b-hairpin has to be
sufficiently stable as monomer without the tendency to self-
aggregate, 2) the photoswitchable unit incorporated must
stabilize the biologically active peptide conformation, and
domain, thus mediating the membrane association of nNOS
to skeletal muscle and inducing the production of the second
messenger nitric oxide (NO) for muscle contraction.
Crucial for binding is the internal recognition motif
-LETTF- of the extended PDZ domain of nNOS located in
the first strand of the hairpin peptide (Scheme 1), a stable
[4]
3
) disturbing the protein binding site by light-induced isomer-
ization of the photoswitch must not result in intermolecular
[
1]
association or even formation of insoluble fibrils. Herein, we
report the first example of a b-hairpin model peptide of a
biologically important protein domain that shows consider-
ably different binding affinities for the target protein that are
dependent on the isomerization state of the embedded
photoswitch.
Scheme 1. a) Sequence of the b-finger peptide of nNOS, the amino
acids in the box represent the internal recognition site, b) structure of
the photoresponsive unit A.
conformation which consists of two antiparallel strands
[5]
PDZ domains mediate the formation of a variety of
connected by a turn. Previous work was based on cyclic
peptides as mimics of the nNOS b-finger (Scheme 2, peptide
1). Structural calculations revealed a structure of 1 that is in
agreement with the original b-finger. Binding studies con-
[
2]
multiprotein complexes in the cell. Besides C-terminal
protein sequences, PDZ domains are also able to recognize
internal peptide motifs that bind at the same binding pocket
as the C-terminal ones. The best example of this type of
internal ligand recognition is found in the extended PDZ
domain of neuronal nitric oxide synthase (nNOS) which
interacts with the PDZ domain from a-1-syntrophin or the
[
6]
firmed binding at the correct site in the protein. Based on
these results, a photoswitchable w-amino acid has been
incorporated into the cyclic b-finger peptide 1 replacing the
d-Pro-Gly turn element and the two amino acids flanking
both ends (Scheme 2b). Photoswitchable w-amino acids
[
3]
second PDZ domain from PSD95. The formation of the
PDZ/PDZ heterodimer requires the b-finger structure of
nNOS (30 amino acid residues) to bind at the syntrophin PDZ
[
7]
[8]
based on either azobenzene or hemithioindigo are good
[
9]
candidates for the modulation of peptide conformations
because they undergo ultrafast photoisomerization
[10]
thus
allowing monitoring of conformational transitions in the pico-
[
*] C. Hoppmann, S. Seedorff, A. Richter, Prof. Dr. K. Rꢀck-Braun
Institut fꢀr Chemie, Technische Universitꢁt Berlin
Strasse des 17. Juni 135, 10623 Berlin (Germany)
E-mail: krueck@chem.tu-berlin.de
[
11]
[12]
to femtosecond timescale.
For biological applications,
azobenzene has been demonstrated to be very effective in
that it shows isomerization around the central N=N bond
C. Hoppmann, S. Seedorff, Dr. P. Schmieder, Dr. M. Beyermann
Leibniz-Institut fꢀr Molekulare Pharmakologie
Robert-Rꢂssle-Strasse 10, 13125 Berlin (Germany)
Fax: (+49)30-9479-3159
http://www.fmp-berlin.de
E-mail: beyermann@fmp-berlin.de
(transQcis) with high isomerization yields and remarkable
changes in geometry. In contrast to the previously reported
[9d,e]
light-switchable b-hairpins,
we embedded the azoben-
zene-w-amino acid 3-((4’-aminomethyl)phenylazo) benzoic
acid (A) as photoswitch because of the high thermal stability
of the photostationary state (pss) of the cis form
Dr. H. Fabian
Robert-Koch-Institut
[
7c]
(
Scheme 1b). The extended geometry of the trans form of
Nordufer 20, 13353 Berlin (Germany)
the azobenzene was expected to disturb the binding site in the
peptide ligand.
[
**] This work was supported by the VW Stiftung.
High isomerization yields in the pss of the cis form of 2
(90% cis content) were achieved after irradiation of the
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200901933.
6
636
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 6636 –6639

Angewandte
Chemie
photoswitching (up to ten photocycles) nor storage for two
weeks at 508C in the dark led to association or formation of
aggregates of the photoswitchable peptide 2, which reverts
completely into the initial monomeric trans form. The
separation of the two photoisomers in the pss by SEC is
indicative of different hydrodynamic volumes of the two
monomeric forms as a result of global changes in their
conformations (Figure 2). Thereby, SEC provides an alter-
Figure 2. SEC profile (220 nm) of peptide 2 in buffer, pH 7.5 after ten
photocycles and two-weeks storage in buffer at 508C displaying the
pure trans form (c), the cis/trans mixture (a) and untreated trans
form (g).
Scheme 2. Structures of synthetic cyclic peptides 1 and 2 related to the
b-finger peptide of nNOS a) 1 containing the -Val-d-Pro-Gly-His- motif
for the non-switchable form, b) 2 containing A as photoswitch. The
lysine side chain for immobilization on the sensor chip is depicted.
native method to reverse phase (RP)-HPLC for determining
cis:trans ratios in that it avoids hydrophobic or ionic
interactions of the peptide with the matrix. The cis content
in the photostationary state determined by SEC is estimated
to approximately 75%.
Using surface plasmon resonance spectroscopy (SPR) we
have investigated the influence of photoisomerization of
peptide 2 on binding to syntrophin compared to the binding of
the non-switchable peptide 1.
With the knowledge that strand I of the hairpin contains
the internal recognition motif, the ligands were immobilized
through the lysine side chain in strand II on a CM5 sensor
chip. The pure trans form of peptide 2 showed almost no
binding (solid line, Figure 3). After photoisomerization, a
thermodynamically stable trans form at 330 nm in buffer
solution at pH 7.5 (Figure 1). The thermal cis!trans isomer-
ization of 2 in aqueous solution was found to be slow (half-life
2
5 days). The repeated photochemical interconversion of the
two states was achieved without occurrence of association,
precipitation, or photobleaching.
Size-exclusion chromatography (SEC) studies under
physiological conditions (pH 7.5) have shown that neither
significant binding of syntrophin (K = 10.6 mm) was found
D
which is comparable to that of the model peptide 1 (K =
D
Figure 1. a) UV/Vis spectra in buffer, pH 7.5 of 2 in the pure trans
form (c) and in the cis form in pss (a); maxima for p–p* and n–
p* transitions are at 327 and 424 nm, respectively; b) RP-HPLC profile
Figure 3. SPR sensorgrams of the interaction of peptide 2 as pure
trans form (c) and cis form in pss (a) with syntrophin PDZ
(c=7.74 mm).
(
220 nm) of the cis form in pss (90% cis content).
Angew. Chem. Int. Ed. 2009, 48, 6636 –6639
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6637

Communications
5
.4 mm). To rule out any binding arising from the solid phase
involved in SPR, we confirmed the different binding behavior
of the two isomers of 2 to syntrophin by ITC (isothermal
titration calorimetry; Supporting Information). Again, whilst
the trans form showed no binding, the cis form of 2 bound
with a remarkable affinity.
In aqueous solution the model peptide 1 adopts an
[6]
antiparallel b-sheet structure. The correlation between the
binding affinities of either 2 in the cis form in the pss or 1 lead
us to assume similar structural features of the binding site.
The IR spectrum of the trans form of 2 (solid line in
Figure 4a) is characterized by a broad and featureless amide I
1
Figure 5. H NMR spectra of peptide 2 in D O/H O (1:9), pH 6
2
2
recorded at 600 MHz; a) cis form in pss and b) trans form, containing
traces of the cis form.
corresponding non-switchable model peptide that adopts a
structure similar to the b-hairpin in the native protein in
aqueous solution. The interaction of a biologically important
b-sheet with a protein domain has been modulated by a light-
induced conformational change without destabilizing the
system. Therefore, the peptide may serve as a suitable
model for a light-triggered b-sheet for use in cells, however
for intracellular applications peptides may be subject to
[
15]
Figure 4. FTIR spectra of 2 in D O buffer, pH 7.5; a) pure trans
form (c) and cis form in pss (a); b) cis-trans-difference spectra.
reduction and proteolytic degradation.
2
Received: April 9, 2009
Revised: May 25, 2009
ꢀ
1
Published online: August 7, 2009
band contour centered at approximately 1645 cm , typical of
[
13]
an unordered peptide structure. Photoisomerization to the
cis-azobenzene induced the formation of secondary structure,
Keywords: azobenzene · conformational switch · isomerization ·
.
photochemistry · protein–protein interactions
as indicated by
1
a
band component at approximately
ꢀ
1
615 cm (dashed line, Figure 4a). The strong low-frequency
ꢀ1
ꢀ1
band at 1613 cm and a weaker band at 1677 cm of the IR
difference spectrum (Figure 4b) indicate that some amide C=
O groups are involved in an antiparallel b-type structure
[
13,14]
[1] I. W. Hamley, Angew. Chem. 2007, 119, 8274 – 8295; Angew.
Chem. Int. Ed. 2007, 46, 8128 – 8147.
that is induced by the trans!cis photoisomerization of the
[
[
2] B. Z. Harris, W. A. Lim, J. Cell Sci. 2001, 114, 3219 – 3231.
3] a) B. J. Hillier, K. S. Christopherson, K. E. Prehoda, D. S. Bredt,
W. A. Lim, Science 1999, 284, 812 – 815; b) H. Tochio, F. Hung,
M. Li, D. S. Bredt, M. Zhang, J. Mol. Biol. 2000, 303, 359 – 370.
photoswitch in the peptide.
Both forms of peptide 2 have been investigated by NMR
spectroscopy (Figure 5). The region of the amide and
aromatic protons of the trans form and cis form are shown.
Most importantly, the sharp lines in the spectra show that
neither conformer forms aggregates, that is, the peptide is
monomeric in both photoisomeric forms even at the relatively
high concentrations required for NMR spectroscopy.
The light-triggered modulation of the binding behavior of
a b-hairpin peptide to its target protein has been successfully
demonstrated. The interaction was studied between a-1-
syntrophin and a b-hairpin peptide derived from the PDZ
domain binding site of the neuronal NO synthase. By
incorporating an azobenzene-w-amino acid as a photoswitch
in a peptide ring of appropriate size, a cyclic light-directed
ligand was developed. Its trans form shows no binding while
the cis form features overall binding comparable to the
[4] J. E. Brenman, D. S. Chao, H. Xia, K. Aldape, D. S. Bredt, Cell
995, 82, 743 – 752.
1
[
[
[
5] P. Wang, Q. Zhang, H. Tochio, J.-S. Fan, M. Zhang, Eur. J.
Biochem. 2000, 267, 3116 – 3122.
6] S. Seedorff, C. Appelt, M. Beyermann, P. Schmieder, unpub-
lished results.
7] a) L. Ulysse, J. Chmielewski, Bioorg. Med. Chem. Lett. 1994, 4,
2145 – 2146; b) R. Behrendt, M. Schenk, H.-J. Musiol, L.
Moroder, J. Pept. Sci. 1999, 5, 519 – 529; c) B. Priewisch,
Dissertation, TU Berlin, 2006; d) B. Priewisch, W. Steinle, K.
Rꢀck-Braun in Peptides 2004 (Eds.: M. Flegel, M. Fridkin, C.
Gilon, J. Slaninova), Kenes, Genf, 2005, pp. 756 – 757.
8] a) T. Schadendorf, C. Hoppmann, K. Rꢀck-Braun, Tetrahedron
Lett. 2007, 48, 9044 – 9047; b) W. Steinle, K. Rꢀck-Braun, Org.
Lett. 2003, 5, 141 – 144; c) S. Herre, W. Steinle, K. Rꢀck-Braun,
Synthesis 2005, 3297 – 3300.
[
6
638 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 6636 –6639

Angewandte
Chemie
[
9] a) L. Ulysse, J. Cubillos, J. Chmielewski, J. Am. Chem. Soc. 1995,
17, 8466 – 8467; b) R. Behrendt, C. Renner, M. Schenk, F.
Wang, J. Wachtveitl, D. Oesterhelt, L. Morder, Angew. Chem.
104, 15729 – 15734; b) J. Bredenbeck, J. Helbig, A. Sieg, T.
1
Schrader, W. Zinth, J. Wachtveitl, C. Renner, R. Behrendt, L.
Moroder, P. Hamm, Proc. Natl. Acad. Sci. USA 2003, 100, 6452 –
6457.
1999, 111, 2941 – 2943; Angew. Chem. Int. Ed. 1999, 38, 2771 –
2774; c) C. Renner, L. Moroder, ChemBioChem 2006, 7, 868 –
878; d) S.-L. Dong, M. Lꢁweneck, T. E. Schrader, W. J. Schreier,
[12] a) L. G. Ulysse, Jr., J. Chmielewski, Chem. Biol. Drug Des. 2006,
67, 127 – 136; b) A. Cattani-Scholz, C. Renner, C. Cabrele, R.
Behrendt, D. Oesterhelt, L. Moroder, Angew. Chem. 2002, 114,
299 – 302; Angew. Chem. Int. Ed. 2002, 41, 289 – 292; c) G.
Hayashi, M. Hagihara, K. Nakatani, Chem. Eur. J. 2009, 15, 424 –
432; d) M. Volgraf, P. Gorostiza, R. Numano, R. H. Kramer,
E. Y. Isacoff, D. Trauner, Nat. Chem. Biol. 2006, 2, 47 – 52; e) L.
Guerrero, O. S. Smart, C. J. Weston, D. C. Burns, G. A. Woolley,
R. K. Allemann, Angew. Chem. 2005, 117, 7956 – 7960; Angew.
Chem. Int. Ed. 2005, 44, 7778 – 7782.
[13] a) H. Fabian, W. Mꢂntele in Handbook of Vibrational Spectros-
copy, Infrared Spectroscopy of Proteins (Eds.: J. M. Chalmers,
P. R. Griffiths), Wiley, New York, 2002, pp. 3399 – 3425; b) A.
Barth, C. Zscherp, Q. Rev. Biophys. 2002, 35, 369 – 430.
[14] J. Hilario, J. Kubelka, F. A. Syud, S. H. Gellman, T. A. Keiderl-
ing, Biopolymers 2002, 67, 233 – 236.
W. Zinth, L. Moroder, C. Renner, Chem. Eur. J. 2006, 12, 1114 –
120; e) A. Aemissegger, V. Krꢂutler, W. F. van Gunsteren, D.
1
Hilvert, J. Am. Chem. Soc. 2005, 127, 2929 – 2936; f) S. Jurt, A.
Aemissegger, P. Gꢀntert, O. Zerbe, D. Hilvert, Angew. Chem.
2006, 118, 6445 – 6448; Angew. Chem. Int. Ed. 2006, 45, 6297 –
6300.
[
10] a) S. Spꢁrlein, H. Carstens, H. Satzger, C. Renner, R. Behrendt,
L. Moroder, P. Tavan, W. Zinth, J. Wachtveitl, Proc. Natl. Acad.
Sci. USA 2002, 99, 7998 – 8002; b) M. Lꢁweneck, A. G. Milbradt,
C. Root, H. Satzger, W. Zinth, L. Moroder, C. Renner, Biophys.
J. 2006, 90, 2099 – 2108; c) T. Cordes, D. Weinrich, S. Kempa, K.
Riesselmann, S. Herre, C. Hoppmann, K. Rꢀck-Braun, W. Zinth,
Chem. Phys. Lett. 2006, 428, 167 – 173; d) T. Cordes, C. Elsner,
T. T. Herzog, C. Hoppmann, T. Schadendorf, W. Summerer, K.
Rꢀck-Braun, W. Zinth, Chem. Phys. 2009, 358, 103 – 110.
[15] For reduction of azobenzene systems see: C. Boulꢃgue, M.
Lꢁweneck, C. Renner, L. Moroder, ChemBioChem 2007, 8, 591 –
594.
[
11] a) T. E. Schrader, W. J. Schreier, T. Cordes, F. O. Koller, G.
Babitzki, R. Denschlag, C. Renner, M. Lꢁweneck, S.-L. Dong, L.
Moroder, P. Tavan, W. Zinth, Proc. Natl. Acad. Sci. USA 2007,
Angew. Chem. Int. Ed. 2009, 48, 6636 –6639
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6639