Photoswitched cell adhesion on surfaces with RGD peptides

Article abstract of DOI:10.1021/ja053648q

Coating of surfaces by RGD peptides is well-known. Herein we describe the possibility to switch cell adhesion properties by changing the distance and orientation of the RGD peptides to the surface. A set of RGD peptides of the type cyclo(-RGDfK-) was synt

Full text of DOI:10.1021/ja053648q

Published on Web 10/29/2005
Photoswitched Cell Adhesion on Surfaces with RGD Peptides
†
†
†
‡
J o¨ rg Auernheimer, Claudia Dahmen, Ulrich Hersel, Andreas Bausch, and
,
†
Horst Kessler*
Contribution from the Department Chemie, Lehrstuhl II f u¨ r Organische Chemie, Technische
UniVersit a¨ t M u¨ nchen, Lichtenbergstr. 4, D-85747 Garching, Germany, and Department Physik,
E22, Technische UniVersit a¨ t M u¨ nchen, James-Franck-Str., D-85747 Garching, Germany
Received June 3, 2005; E-mail: kessler@ch.tum.de
Abstract: Coating of surfaces by RGD peptides is well-known. Herein we describe the possibility to switch
cell adhesion properties by changing the distance and orientation of the RGD peptides to the surface. A
set of RGD peptides of the type cyclo(-RGDfK-) was synthesized containing the photoswitchable 4-[(4-
aminophenyl)azo]benzocarbonyl central unit as spacer between the acrylamide anchor and the RGD peptide.
PMMA (poly methyl methacrylate) surfaces were coated with these peptides. Control of adhesion stimulation
by irradiation with 366 or 450 nm light could be achieved.
Introduction
One of the most commonly used classes of substances
containing reversible photoinduced isomerization units are
Coating of surfaces with cell-adhesive molecules provides a
strong mechanical contact between cells and the surface. Cell
adhesion is mediated by integrins, a class of heterodimeric
transmembrane cell receptors that bind selectively to different
proteins of the extracellular matrix (EMC). Cellular binding
sites, like RGD peptides, have been reported to play a major
role in mediating cell adhesion through integrins, which
transduce information to the nucleus through cytoplasmic
signaling pathways. In this study, we have used tailor-made
cyclic RGD peptides, developed in our group, which bind
azobenzenes because of the pronounced changes in geometry
upon its light-induced Z-E isomerization, the high (photo)-
stability, the high quantum yields, as well as of the extremely
1
7
2
fast and fully reversible isomerization processes. These are
completed within 10 ps for the E-Z and within 1 ps for the
8
3
Z-E direction. Conversely, thermal Z-E relaxation is a slow
7
process, but leads to 100% E isomer.
9
Azobenzene derivatives, such as 4-(phenylazo)phenylalanine,
1
0
4
4-[(4-aminophenyl)azo]benzoic acid,
4-(4-aminomethyl)-
phenyl)azobenzoic acid, and 3-[(3-aminomethylphenyl)azo]-
phenylacetic acid, have been incorporated into peptides and
biopolymers to change their structure by photoisomeriza-
1
1
specifically to Rvâ3 and Rvâ5 integrins, but to RIIbâ3 with
low affinity. The Rv integrins are known to adhere to vitronectin.
Only the Rvâ3 integrin is found in focal contacts and leads to
1
2
9
,10,12,13
5
tion.
In the field of RGD peptides, 4-[(4-aminomethyl)-
spreading and migration of (endothelial) cells on vitronectin.
phenylazo]benzoic acid has been incorporated into the backbone
One promising approach to generate cell adhesive and cell
repulsive areas on material surfaces is coating with photochemi-
cally regulated molecules. This method provides surfaces
irreversibly coated with specific photolithographically accessible
patterns. However, a transient photochemical alteration of the
(
6) (a) Irie, M. Chem. ReV. 2000, 100, 1685-1716. (b) Yokoyama, Y. Chem.
ReV. 2000, 100, 1717-1739. (c) Feringa, B. L.; van Delden, R. A.;
Koumura, N.; Geertsema, E. M. Chem. ReV. 2000, 100, 1789-1816. (d)
Willner, I.; Rubin, S. Angew. Chem., Int. Ed. 1996, 35, 367-385. (e)
Willner, I. Acc. Chem. Res. 1997, 30, 347-356.
binding properties of biomolecules can only be achieved by
(7) Rau, H. In Studies in Organic Chemistry: Photochromism, Molecules and
Systems; D u¨ rr, H., Bonas-Laurent, H., Eds.; Elsevier: Amsterdam, 1990;
pp 165-192.
light-induced structural changes.6
(8) (a) Lednev, I. K.; Ye, T.-Q.; Hester, R. E.; Moore, J. N. J. Phys. Chem.
†
Department Chemie.
Department Physik.
1996, 100, 13338-13341. (b) Wachtveitl, J.; N a¨ gele, T.; Puell, B.; Zinth,
‡
W.; Kr u¨ ger, M.; Rudolph-B o¨ hner, S.; Oesterhelt, D.; Moroder, L. J.
Photochem. Photobiol. A: Chem. 1997, 105, 283-288. (c) N a¨ gele, T.;
Hoche, R.; Zinth, W.; Wachtveitl, J. Chem. Phys. Lett. 1997, 272, 489-
495. (d) Lednev, I. K.; Ye, T.-Q.; Matousek, P.; Towrie, M.; Foggi, P.;
Neuwahl, F. V. R.; Umapathy, S.; Hester, R. E.; Moore, J. N. Chem. Phys.
Lett. 1998, 290, 68-74. (e) Fujino, T.; Arzhanstsev, Y. S.; Tahara, T. J.
Phys. Chem. A 2001, 105, 8123-8129. (f) Satzger, H.; Sp o¨ rlein, S.; Root,
C.; Wachtveitl, J.; Zinth, W.; Gilch, P. Chem. Phys. Lett. 2003, 372, 216-
223.
(
(
(
1) Eble, J. A. Integrins, A Versatile and Old Family of Cell Adhesion
Molecules. In Integrin-Ligand Interaction; Eble, J. A., K u¨ hn, K., Eds.;
Springer-Verlag: Heidelberg, 1997; pp 1-40.
2) K u¨ hn, K. Extracellular Matrix Constituents as Integrin Ligands. In Integrin-
Ligand Interaction; Eble, J. A., K u¨ hn, K., Eds.; Springer-Verlag: Heidel-
berg, 1997; pp 41-83.
3) (a) Pierschbacher, M. D.; Ruoslahti, E. J. Nature 1984, 309, 30-33. (b)
Okamoto, K.; Matsuura, T.; Hosokawa, R.; Akagawa, Y. J. Dent. Res. 1998,
7
7, 481-487. (c) Matsuura, T.; Hosokawa, R.; Okamoto, K.; Kimoto, T.;
(9) (a) Goodman, M.; Falxa, M. L. J. Am. Chem. Soc. 1967, 89, 3863-3867.
(b) Liu, D.; Karanicolas, J.; Yu, C.; Zhang, Z.; Woolley, G. A. Bioorg.
Med. Chem. Lett. 1997, 7, 2677-2680. (c) Hamachi, I.; Hiraoka, T.;
Yamada, Y.; Shinkai, S. Chem. Lett. 1998, 6, 537-538.
(10) Behrendt, R.; Renner, C.; Schenk, M.; Wang, F.; Wachtveitl, J.; Oesterhelt,
D.; Moroder, L. Angew. Chem., Int. Ed. 1999, 38, 2771-2774.
(11) Ulysse, L.; Chmielewski, J. Bioorg. Med. Chem. Lett. 1994, 4, 2145-
2146.
Akagawa, Y. Biomaterials 2000, 21, 1121-1127. (d) Hersel, U.; Dahmen,
C.; Kessler, H. Biomaterials 2003, 24, 4385-4415.
(
4) (a) Aumailley, M.; Gurrath, M.; M u¨ ller, G.; Calvete, J.; Timpl, R.; Kessler,
H. FEBS Lett. 1991, 291, 50-54. (b) Haubner, R.; Gratias, R.; Diefenbach,
B.; Goodman, S. L.; Jonczyk, A.; Kessler, H. J. Am. Chem. Soc. 1996,
1
18, 7461-7472.
5) (a) Wayner, E. A.; Orlando, R. A.; Cheresh, D. A. J. Cell Biol. 1991, 113,
19-929. (b) Leavesley, D. I.; Ferguson, G. D.; Wayner, E. A.; Cheresh,
D. A. J. Cell Biol. 1992, 117, 1101.
(
9
(12) Aemissegger, A.; Kr a¨ utler, V.; van Gunsteren, W. F.; Hilvert, D. J. Am.
Chem. Soc. 2005, 127, 2929-2936.
10.1021/ja053648q CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 16107-16110
9
16107

A R T I C L E S
Auernheimer et al.
For the second set of peptides (5-7), N-acryloylglycine,20 N-acryloyl-
δ-amino valeric acid,21 and N-acryloyl-ꢀ-amino hexanoic acid were
preactivated with HATU/HOAt/collidine for acylation of 4-(4-ami-
nophenyl)azobenzoic acid.
Coating. Twenty microliter solutions of peptides 1-7 in 2-propanol/
DMSO were placed on PMMA disks (Palacos R, 1 cm)ssix for each
peptidesfollowed by irradiation at 254 nm for 2 h. After standing
overnight in the dark, every disk was rinsed with 2 mL of PBS buffer
of cyclic peptides, and the effects of light-switched conforma-
tional preferences on the binding affinities to integrin have been
21
analyzed.14
15
Surface-immobilized photoswitchable 4-arylazopyridine and
1
6
azobenzene can be used to form self-assembled monolayers
SAMs) on gold surfaces and Langmuir-Blodgett films (LB
(
films) on glass and silicon. Thereby, it was shown that surface-
positioned photoisomerizable substances retain their light-
responsiveness. In the dark, at equilibrium, azobenzene is
predominantly in the more stable E form, and its switching
kinetics was found to be altered by immobilization, but both
diastereomers could be generated by irradiation with appropriate
(pH 6.0) to remove uncoated peptide. Three disks from every peptide
were irradiated at 450 nm for 3 h and stored for 5 days in the dark to
ensure that thermal equilibrium is reached even on bound material.
The other three disks were irradiated at 366 nm for 12 h directly before
seeding with cells. All disks were transferred to cell culture plates (48
wells).
Cell Adhesion Assay. The cell adhesion assays were performed as
described by Landegren.22 MC3T3 E1 mouse osteoblasts were seeded
on the BSA blocked substrate at a density of 50 000 cells per well.
The cells were allowed to adhere for 1 h under standard tissue culture
wavelengths of light. On the other hand, previous studies from
our laboratory and others17 clearly revealed that cell adhesion
on different surfaces depends on the spacer length applied
between the ligand peptide and the surface. From these
experiments, it was compelling to analyze the effect of photo-
responsive units in the spacer for anchoring cyclic RGD peptides
to surfaces with the working assumption that a shorter spacer
would not mediate cell adhesion whereas a longer one would
do it. If such concept is realized, photolithographic structuring
of surfaces in cell adhesive and cell repulsive areas would
become possible.
2
conditions (37 °C, 5% CO ) in serum-free culture medium (DMEM)
containing 1% BSA (w/v). The wells were washed three times with
PBS (pH 7.4) to remove nonadherent cells. Attached cells were
quantified by an ELISA detection of the activity of the lysosomal
enzyme hexosaminidase. p-Nitrophenol-N-acetyl-â-D-glucosaminide
was cleaved by the enzyme, and the amount of colored p-nitrophenol
was measured with an ELISA reader (Dynatech Laboratories, MRX)
at 405 nm. Results are given as the percentage of the total number of
cells seeded (which is considered as 100% of cell adhesion). In all
experiments, the mean value of each point given in the figures is the
result of triplicates; the error bars represent standard deviations.
Materials and Methods
Peptide Synthesis. The cyclic pentapeptide cyclo(-RGDfK-) was
4
b,17d
synthesized as described previously
and coupled in solution to the
spacer/anchor construct with HATU/HOAt/collidine. In one set of
peptides (1-4), the azobenzene moiety was placed near the anchor
group and thus in proximity of the solid surface, while in a second set
Results and Discussion
For this study, 4-[(4-aminophenyl)azo]benzoic acid was
chosen as the light switch as it was shown to retain all typical
properties of azobenzene derivatives in terms of photoisomer-
(
5-7), it was grafted directly to the cyclic RGD peptide. Correspond-
18,19
ingly, for compounds 1-4, 4-(4-aminophenyl)azobenzoic acid
was
acylated with acryloyl chloride and then C-terminally elongated by
coupling to glycine, γ-amino butyric acid, and ꢀ-amino hexanoic acid
attached to TCP-resin with TBTU/HOBt/DIEA.
13c
ization and photostability.
As known for this and other
azobenzene derivatives, at thermodynamic equilibrium in the
dark, the E isomer, which is about 3 Å longer than the Z isomer,
is obtained in almost quantitative yield. It can be switched to
the Z form by irradiation with light at 360 nm, but at the
photostationary equilibrium, conversion to the Z isomer proceeds
(
13) (a) Ulysse, L.; Cubillos, J.; Chmielewski, J. J. Am. Chem. Soc. 1995, 117,
8
466-8467. (b) Renner, C.; Behrendt, R.; Sp o¨ rlein, S.; Wachtveitl, J.;
Moroder, L. Biopolymers 2000, 54, 489-500. (c) Renner, C.; Cramer, J.;
Behrendt, R.; Moroder, L. Biopolymers 2000, 54, 501-514. (d) Pieroni,
O.; Fissi, A.; Angelini, N.; Lenci, F. Acc. Chem. Res. 2001, 34, 9-17. (e)
Cattani-Scholz, A.; Renner, C.; Cabrele, C.; Behrendt, R.; Oesterhelt, D.;
Moroder, L. Angew. Chem., Int. Ed. 2002, 41, 289-292. (f) Sp o¨ rlein, S.;
Carstens, H.; Satzger, H.; Renner, C.; Behrendt, R.; Moroder, L.; Tavan,
P.; Zinth, W.; Wachtveitl, J. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 7998-
7
only with yields of about 70-90%. Therefore, in cell adhesion
assays, the lack of 100% E isomer has to be taken into account.
However, if the density of RGD is relatively low, the depletion
of RGD peptides which are exposed to be recognized by the
integrins can switch the recognition from “binding” to essentially
“no stimulated binding”. It is known from our previous studies
that cells need oligomeric binding with a maximum distance of
8
002. (g) Hugel, T.; Holland, N. B.; Cattani, A.; Moroder, L.; Seitz, M.;
Gaub, H. E. Science 2002, 296, 1103-1106. (h) Bredenbeck, J.; Helbing,
J.; Sieg, A.; Schrader, T.; Zinth, W.; Renner, C.; Behrendt, R.; Moroder,
L.; Wachtveitl, J.; Hamm, P. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 6452-
6
457. (i) Renner, C.; Kusebauch, U.; L o¨ weneck, M.; Milbradt, A. G.;
Moroder, L. J. Pept. Res. 2005, 65, 4-14.
2
3
(
14) (a) Milbradt, A. G.; L o¨ weneck, M.; Krupka, S. S.; Reif, M.; Sinner, E.-K.;
Moroder, L.; Renner, C. Biopolymers 2005, 77, 304-13. (b) Sch u¨ tt, M.;
Krupka, S. S.; Milbradt, A. G.; Deindl, S.; Sinner, E.-K.; Oesterhelt, D.;
Renner, C.; Moroder, L. Chem. Biol. 2003, 10, 487-490.
about 65 nm between RGD peptides.
An acrylamide anchor on PMMA was chosen for im-
mobilization of the cyclo(-RGDfK-), as the influence of spacer
length on the integrin-mediated cell adhesion is already known
(
15) Cook, M. J.; Nygård, A.-M.; Wang, Z.; Russell, D. A. Chem. Commun.
2
002, 1056-1057.
17d
(
16) (a) Evans, S. D.; Johnson, S. R.; Ringsdorf, H.; Williams, L. M.; Wolf, H.
Langmuir 1998, 14, 6436-6440. (b) Walter, D. G.; Champbell, D. J.;
Mirkin, C. A. J. Phys. Chem. B 1999, 103, 402-405. (c) Tamada, K.;
Akiyama, H.; Wei, T. X. Langmuir 2002, 18, 5239-5246. (d) Sidorenko,
A.; Houphouet-Boigny, C.; Villavicencio, O.; McGrath, D. V.; Tsukruk,
V. V. Thin Solid Films 2002, 410, 147-158.
for this system. For an effective spacer, two ꢀ-amino hexanoic
acids (Ahx) and an acrylamide as anchor was required, which
corresponds to a length of 1.7 nm (distance between both
external carbonyl carbon atoms), while the acryloyl-Ahx spacer
with 0.9 nm length was found to be too short to mediate cell
adhesion. Taking this information into account, a set of peptide
constructs was designed (Figure 1) containing the photoswit-
(
17) (a) Beer, J. H.; Springer, K. T.; Coller, B. S. Blood 1992, 79, 117-128.
(b) Craig, W. S.; Cheng, S.; Mullen, D. G.; Blevitt, J.; Pierschbacher, M.
D. Biopolymers 1995, 37, 157-175. (c) Kantlehner, M.; Finsinger, D.;
Meyer, J.; Schaffner, P.; Jonczyk, A.; Diefenbach, B.; Nies, B.; Kessler,
H. Angew. Chem., Int. Ed. 1999, 38, 560-562. (d) Kantlehner, M.;
Schaffner, P.; Finsinger, D.; Meyer, J.; Jonczyk, A.; Diefenbach, B.; Nies,
B.; H o¨ lzemann, G.; Goodman, S. L.; Kessler, H. ChemBioChem 2000, 1,
(20) Korte, F.; St o¨ riko, K. Chem. Ber. 1960, 93, 1033-1042.
(21) Pless, D. D.; Lee, Y. C.; Roseman, S.; Schnaar, R. L. J. Biol. Chem. 1983,
258, 2340-2349.
(22) Landegren, U. J. Immunol. Methods 1984, 67, 379-388.
(23) Arnold, M.; Cavalcanti-Adam, E. A.; Glass, G.; Bl u¨ mmel, J.; Eck, W.;
Kantlehner, M.; Kessler, H.; Spatz, J. ChemPhysChem 2004, 5, 383-388.
1
07-114.
(
18) Sch u¨ ndeh u¨ tte, K. H. In Houben-Weyl; Thieme Verlag: Stuttgart, 1965;
Vol. 10/3, p 340.
(
19) Behrendt, R.; Schenk, M.; Musiol, H.-J.; Moroder, L. J. Pept. Sci. 1999,
5
, 519-529.
16108 J. AM. CHEM. SOC.
9
VOL. 127, NO. 46, 2005

Photoswitched Cell Adhesion on Surfaces with RGD Peptides
A R T I C L E S
Figure 1. Set of cyclic RGD peptides containing a photoswitchable 4-[(4-
aminophenyl)azo]benzocarbonyl unit.
Table 1. Estimated Maximal Spacer Lengths of the Reversibly
Photoswitchable Peptides 1-7 in Z and E Configuration
Compared to Acrylamide Linkers with and without ꢀ-Amino
Hexanoic Acid Spacer (the distance between both external
carbonyl carbon atoms of the spacer is given)
Figure 2. UV/vis absorption spectra from 2 (0.1 mM in DMSO) in E
configuration, (a) during irradiation with UV light at 366 nm and (b)
following thermal relaxation at room temperature in the dark. (Light source
was changed at 350 nm while spectra were recorded. Spectra were smoothed
according to the Savitzky-Golay method (2 × 13 data points) to reduce
noise.)
spacer
length [nm]
Acryl-Ahx-c(-RGDfK-)
Acryl-Ahx-Ahx-c(-RGDfK-)
0.856
1.723
E isomer
1
2
3
4
5
6
7
1.25
1.73
1.82
2.06
1.65
2.01
2.13
monitored by UV spectra. Figure 2 shows the isomerization
behavior of peptide 2. The maximum population of Z isomers
is reached 5 min after irradiation of the E isomer. As known
for azobenzene, thermal relaxation is much slower than the
photoinduced isomerization. When a sample is kept in the dark
for more than 3 days, the initial state is regained (pure E isomer).
The absorption maximum of the E isomers of all peptides is at
about 370 nm, which allows a relatively selective irradiation
of the E configuration at 366 nm. Reversibility of the photo-
chemical-induced isomerization was demonstrated by repeated
switching (at least two times), followed by spectroscopic
characterization. The photostability of the peptides in solution
Z isomer
1
2
3
4
5
6
7
0.69
0.95
1.16
1.38
1.10
1.35
1.46
1
was verified by H NMR, ESI-MS, and HPLC.
chable azobenzene, and the maximum length of the correspond-
ing Z and E isomers was estimated by using the X-ray structure
of cis-azobenzene24 and the standard averaged bond lengths of
CH2-CH2 and amides (Table 1).
The reversible light-switching of all azo compounds in DMSO
solution was confirmed by UV/vis and H NMR spectroscopy.
By confirming the slow kinetics of thermal relaxation for all
peptides compared to the time scale of cell adhesion assays, at
least from this point of view, all peptides are potentially suitable
for the reversible photocontrol of cell adhesion on artificial
surfaces.
1
Peptides 1-7 were immobilized on PMMA disks by irradia-
tion with UV light at 254 nm. Then one part of the disks was
irradiated at 450 nm for 3 h and then stored in the dark for 5
days to switch the peptides completely into the E configuration,
while the other part was irradiated overnight at 366 nm directly
before the cell adhesion test. All peptide-coated surfaces showed
an increase of cell adhesion after irradiation at 450 nm and
storage in the dark (Figure 3). Disks irradiated at 366 nm showed
lower enhancements of cell adhesion, but only for peptide 5
was cell adhesion by light-induced E,Z isomerization reduced
to the same level as that for uncoated material. On the other
hand, peptide 1, which has the shortest spacer in Z configuration,
still exhibits increased cell adhesion compared to uncoated
PMMA, even after irradiation at 366 nm. A rational explanation
The E:Z ratio at photostationary equilibrium was calculated from
the different chemical shifts of the olefinic anchor protons
1
measured by H NMR. The Z isomer dominates but in addition
between 20 and 30% E configuration was present. After light-
protected storing of the samples for 5 days, the pure E isomers
were characterized.
For UV/vis measurements, the samples were stored in the
dark for 5 days at room temperature to obtain pure E isomer
and were then irradiated with UV light at 366 nm for 60 min.
After 1, 5, 20, and 60 min, the UV/vis spectra were taken to
monitor kinetics of the EfZ isomerization; in a similar manner,
thermal relaxation at room temperature in the dark was
(24) Hampson, G. C.; Robertson, J. M. J. Chem. Soc. 1941, 409-413.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 46, 2005 16109

A R T I C L E S
Auernheimer et al.
isomerization. All peptides lead to enhanced cell adhesion on
PMMA disks in their E form, whereas the plating efficiency
was decreased after irradiation at 366 nm, which shortens the
distance of the RGD-containing peptides to the surface by E,Z
isomerization. For acryloyl-Gly-[4-(4-aminophenyl)azo]benzo-
carbonyl-c(-RGDfK-) 5, the cell adhesion in the Z configu-
ration could be reduced to almost the same level as that on
uncoated PMMA disks, while in the E conformation, the plating
efficiency was still enhanced by 17%. Therefore, photochemical
control of cell adhesion was achieved.
Acknowledgment. The authors thank Gabi Chmel (Technical
University of Munich, Physik Department, E22) for carrying
out the cell adhesion tests, Biomet Deutschland GmbH (Darm-
stadt, Germany) for the kind supply of Palacos R and a sample
of MC3T3 E1 mouse osteoblasts. This work was supported by
the DFG, Sonderforschungsbereich 563 (C3 and B14).
Figure 3. Cell adhesion of MC3T3 E1 mouse osteoblasts on PMMA. The
peptide concentration in the coating solution was 1 mM. Samples were
irradiated for 3 h at 450 nm and overnight at 366 nm, respectively.
of this phenomenon is difficult, even taking into account the
changes observed in photoisomerization of azobenzene deriva-
tives on solid surfaces.15,16
Supporting Information Available: Synthetic procedures and
spectroscopic data. This material is available free of charge via
the Internet at http://pubs.acs.org.
Summary
The distance of the RGD ligand from the PMMA surface
could be switched in the peptides 1-7 by light-induced E,Z
JA053648Q
16110 J. AM. CHEM. SOC.
9
VOL. 127, NO. 46, 2005