Highly efficient cell adhesion on beads functionalized with clustered peptide ligands

Article abstract of DOI:10.1039/b911440h

Resin beads were functionalized with either clustered peptide ligands or individual peptide ligands at various ligand densities and then the beads were evaluated in a cell binding assay.

Full text of DOI:10.1039/b911440h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Highly efficient cell adhesion on beads functionalized with clustered
peptide ligands†
St e´ phanie Foillard, Pascal Dumy* and Didier Boturyn*
Received 11th June 2009, Accepted 27th July 2009
First published as an Advance Article on the web 5th August 2009
DOI: 10.1039/b911440h
Resin beads were functionalized with either clustered peptide
ligands or individual peptide ligands at various ligand den-
sities and then the beads were evaluated in a cell binding
assay.
Multivalent interactions are a ubiquitous strategy that has evolved
in biology for a wide range of functions including signal transduc-
1
tion pathways triggering cell adhesion and morphology. These
multivalent interactions are more potent than the analogous
2
monovalent interactions. When designing ligands, the most
effective strategy is to use chemical entities that display multiple
3,4
copies of a recognition unit from a central scaffold. A variety of
scaffolds was exploited to construct multivalent ligands varying in
3,5
Fig. 1 General structure of tetrameric RGD-containing peptide.
size, shape and physical characteristics. We reason that clustered
ligands can improve the adhesion of cells to surfaces especially
when the concentration of ligands on the surface is decreased. This
strategy could be particularly attractive not only for designing
surfaces for tissue engineering but also for selectively capturing
cells from biological fluids. Previous efforts to study the benefit
of multivalent ligands have used clustered RGD (Arg-Gly-Asp)
as a negative control compound for biological studies. We used
the oxime bond formation to construct biomolecular assemblies
since this chemoselective reaction has proved to be particularly
9
10
efficient for preparing artificial protein, chemical microarray,
11
and other bioconjugates. Our chemical approach involves the
incorporation of building blocks encompassing a protected
aminooxy function within the peptide chain (see ESI†). The
peptides that specifically home to cancer tissue through the a
v
b
3
6,7
integrin receptor. Herein we evaluated the effects on cancer cell
adhesion of a surface functionalized with clustered or individual
RGD peptide ligands. Additionally, a fundamental issue regarding
the density of ligands on the surface necessary for cell adhesion
was addressed.
1
-ethoxyethylidene (Eei) group was chosen to protect the
aminooxy function (dotted frame in Scheme 1) since we recently
established that Eei is entirely compatible with the solid-phase
12
peptide synthesis.
Briefly, compounds 1–3 were synthesized from PEG-containing
resin beads suitable for use in cell adhesion assays. We pre-
pared resins containing different levels of ligand loading (from
Our approach basically consists in decorating resin beads with
peptide ligands via a simple chemical method and in assessing
the cell adhesion to the bead using microscopy. For this study,
we selected a tetravalent RGD-containing peptide endowed with
0
.2 mmol/g to 0.2 pmol/g). This was easily accomplished by
adjusting the amount of D-glutamic acid on the resin (see ESI†).
The latter was essential to ensure the b-turn directed head-
to-tail cyclization within the cyclodecapeptide. To conduct a
comparison of resin beads functionalized with compounds 1–3,
it was mandatory to begin the synthesis via a common
tetrapeptide intermediate 4 (Scheme 1). Afterwards, the one-
pot aminooxy deprotection and the chemoselective ligation of
the RGD units were carried out under the mild acid conditions
that are essential to prevent peptide release from the resin.
Compounds 1–3 were finally analyzed and characterized from
resin bead samples by means of RP-HPLC and mass spectroscopy
7,8
desirable biological properties (Fig. 1). The latter is composed of
a cyclic decapeptide scaffold that presents in a spatially controlled
manner two independent functional domains: a clustered ligand
domain for integrin recognition conferring cell targeting and a
domain devoted to a supplementary function (cancer monitoring
and/or drug delivery). To study the contribution of the RGD
cluster to cell adhesion, we decided to prepare compounds
1
and 2 displaying respectively four and one recognition element
by means of fully solid-phase syntheses (Scheme 1). Compound
bearing nonsense RbAD (Arg-bAla-Asp) peptides was used
3
(
see ESI†).
D e´ partment de Chimie Mol e´ culaire, UMR CNRS/UJF 5250, ICMG FR
607, 301, rue de la chimie, BP53, 38041 Grenoble cedex 9, France.
The ability of resin beads to selectively bind cells through
2
RGD–a
v
3
b integrin recognition was next evaluated using human
E-mail: didier.boturyn@ujf-grenoble.fr; Fax: +33 4 76 51 49 46; Tel: +33 4
7
6 51 45 23
embryonic kidney cells HEK 293(b3) (high level of a b integrins)
and HEK 293(b1) (a b negative but expressing a b ). We first ex-
amined beads displaying a high density of ligand (0.2 mmol/gresin).
Cells were incubated with resin beads for 30 minutes at 37 C. As
v
1
3
3
†
Electronic supplementary information (ESI) available: Details of the
v
3
v
1
solid phase peptide synthesis, syntheses and spectroscopic characterisation
of compounds 1–3, cell adhesion procedure, optical images of cell adhesion
on beads (ligand loading from 0.2 mmol/g to 0.2 pmol/g), and vinculin
staining. See DOI: 10.1039/b911440h
◦
expected, compound 3 did not lead to cell adhesion whereas
This journal is © The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4159–4162 | 4159

View Article Online
0
Scheme 1 Solid-phase syntheses of 1–3. (a–b) standard Fmoc/t-Bu solid phase peptide synthesis; (c) Pd (PPh
3
)
4
, PhSiH
3 2 2
, CH CH /DMF;
(
d) PyAOP, DIPEA, DMF; (e) 0.4 mmol/gresin c[-Arg-X-Asp-D Phe-Lys(-CO-CHO)-], CH Cl /TFA/H
2
2
2
O (94/3/3). Fmoc = 9-fluorenylmethoxy–
carbonyl, DMF = N,N -dimethylformamide, PyAOP = (7-azabenzo-triazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate, DIPEA =
diisopropylethylamine, TFA = trifluoroacetic acid.
2
16
beads exhibiting compounds 1 and 2 bound strongly positive
HEK 293(b3) cells (Fig. 2a and b, see also ESI†). This result
indicates that the clustered ligand architecture does not make
an important contribution to ligand–integrin interactions. The
multivalency effect essentially comes from the high level of resin
substitution. Surprisingly, the non specific adhesion of HEK
100 receptors/mm for spread cells. Since the majority of ligands
17
in the interior of the resin bead (~90%) is not reached by cells,
we estimate the cell accessibility to be nearly 40–80 ligands/mm
at the lower level of resin substitution (2 nmol/gresin). These
data suggest that the minimum density of ligand 1 required for
cell adhesion matches with the amount of accessible integrin
receptors. For instance, approximately 50–100 fibronectin natural
3
18
293(b1) was slightly observed on beads carrying 1 and 2 (Fig. 2c
2
19
and d) signifying that the density of ligand on the surface is crucial
for preserving the selectivity for the target.
ligands/mm are necessary for cell adhesion, while more than
200–5000 molecules/mm are required for the linear peptide
2
20
To discriminate the multivalency effect of clustered RGD 1 from
that of the resin bead, we next assessed the cell adhesion under
stringent conditions using beads containing very low densities
of RGD ligands 1 and 2 (2 nmol/gresin). In this context, only
beads functionalized with nanoscale clustered RGD peptides 1
GRGDSPK.
Finally, it was interesting to exploit the resin beads for selectively
capturing a -expressing cells from a biological mixture since
most tumour metastases express the a integrin. Methodologies
for selective tumour cell capture from biological samples are
of significant importance for cancer monitoring. Their ma-
jor barriers to use arise from limited capture efficiency and
lack of standardization. We therefore applied the resin bead 1
(20 mmol/gresin) to capture HEK 293(b3) cells from a mixed
v
b
3
v
b
3
generated the adhesion of a
while few HEK 293(b3) cells were attached via the RGD ligand
(Fig. 2f). This trend was also observed in the presence of
v
b
3
-expressing HEK cells (Fig. 2e),
2
beads displaying RGD ligands 1 and 2 at intermediate ligand
densities (2–20 mmol/gresin): we observed an enhanced cell adhesion
using 1 until the limit of 2 pmol/gresin where we did not detect
any cells (see ESI†). We confirmed our previous finding that
cell suspension containing 3LL cells (a
v
3
b negative Lewis lung
carcinoma). As expected, some HEK 293(b3) cells were trapped
onto the beads (Fig. 3a). The cells that did not adhere to the
beads were readily removed (see ESI†) and the beads subsequently
a
v
b integrin–ligand interactions can be improved through the
3
7,14,15
◦
clustered ligand domain of the cyclodecapeptide scaffold.
Additionally, beads displaying ligand 1 did not induce HEK
93(b1) cell adhesion when using low resin loading (see Fig. S29
incubated in cell culture medium for 96 hours at 37 C. As shown
in Fig. 3b (see also Fig. S34–S37 in the ESI†), the resin beads
were massively coated with cells, signifying a high rate of cell
division on beads displaying ligand 1. Interestingly, the latter
mediates morphological differentiation with a highly extended and
flattened cellular morphology (see Fig. S38 and S39 in the ESI†)
suggesting that an intracellular signalling pathway is activated by
ligand 1.
2
and S30 in the ESI†). This result indicates that control of the
ligand density on the surface may be required to ensure the cellular
selectivity.
Integrin expression level was determined to be approx-
5
imately 10 receptors/cell, yielding an integrin density of
4
160 | Org. Biomol. Chem., 2009, 7, 4159–4162
This journal is © The Royal Society of Chemistry 2009

View Article Online
Fig. 2 Optical microscopy images of cell adhesion to 120 mm resin beads.
◦
Cells (1 million) were treated with beads for 30 min at 37 C then fixed
Fig. 3 Optical microscopy images of cell adhesion to 120 mm resin beads.
A mixed HEK 293(b3)/3LL cell suspension (1/9; 1 million) was treated
with beads 1 (20 mmol/gresin) for 30 min at 37 C, (a) then fixed and stained
and stained with methylene blue. HEK 293 (b3) were incubated with
a) 1 (0.2 mmol/gresin), (b) 2 (0.2 mmol/gresin), (e) 1 (0.2 nmol/gresin),
f) (0.2 nmol/gresin). HEK 293 (b1) were incubated with
◦
(
(
(
2
with methylene blue; (b) beads were incubated in cell culture medium for
◦
c) 1 (0.2 mmol/gresin), (d) 2 (0.2 mmol/gresin).
96 hours at 37 C in a humidified 5% CO
stained with methylene blue.
2
atmosphere, then fixed and
These results indicate that beads displaying a nanoscale dis-
(
Institut Albert Bonniot, INSERM U823, La Tronche, France)
tribution of compound 1 mimic an extracellular matrix protein
such as vitronectin, the natural ligand of the a integrin. They
extend our previous observations, obtained with multivalent RGD
for providing us with all cell lines.
v
b
3
7,8,13,14
peptides (Fig. 1),
that the clustered architecture improves
Notes and references
integrin binding. We hypothesize that the observed multivalent
effect arises from a statistical rebinding of ligand 1 due to the high
local concentration of RGD elements.
1
M. Mammen, S. K. Choi and G. M. Whitesides, Angew. Chem., Int.
Ed., 1998, 37, 2754.
2 T. K. Dam and C. F. Brewer, Biochemistry, 2008, 47, 8470.
3
4
(a) L. L. Kiessling, J. E. Gestwicki and L. E. Strong, Angew. Chem.,
Int. Ed., 2006, 45, 2348; (b) R. Haag and F. Kratz, Angew. Chem., Int.
Ed., 2006, 45, 1198.
In addition, our approach is not limited to a
v
b integrin-
3
expressing cells: by adapting the corresponding domain on the
scaffold (i.e. intermediate 5 in Scheme 1), resin beads may be
easily exploited to adhere to various cells due to the increasing
number of selective ligands selected in vivo. At the present time,
we are extending our study to selectively target CD20-expressing
cells such as lymphoma.
For elegant examples, see: (a) P. I. Kitov, G. L. Mulvey, T. P. Griener,
T. Lipinski, D. Solomon, E. Paszkiewicz, J. M. Jacobson, J. M.
Sadowska, M. Suzuki, K. Yamamura, G. D. Armstrong and D. R.
Bundle, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 16837; (b) V. M.
Hern a´ ndez-Rocamora, B. Maestro, B. de Waal, M. Morales, P. Garcia,
E. W. Meijer, M. Merkx and J. M. Sanz, Angew. Chem., Int. Ed., 2009,
4
8, 948.
5
6
C. W. Cairo, ACS Chem. Biol., 2007, 2, 652.
(a) H. D. Maynard, S. Y. Okada and R. H. Grubbs, J. Am. Chem.
Soc., 2001, 123, 1275; (b) R. J. Kok, A. J. Schraa, E. J. Bos, H. E.
Acknowledgements
´
Moorlag, S. A. Asgeirsd o´ ttir, M. Everts, D. K. F. Meijer and G.
This work was supported by the Association pour la Recherche
contre le Cancer (ARC No 3741), the R e´ gion Rh oˆ ne-Alpes (No
Molema, Bioconjugate Chem., 2002, 13, 128; (c) G. Thumshirn, U.
Hersel, S. L. Goodman and H. Kessler, Chem.–Eur. J., 2003, 9, 2717.
0
0
301372501 and 0301372502), the Canc e´ rop oˆ le (No 032115 and
42259), the Universit e´ Joseph Fourier, the Centre National de
la Recherche Scientifique (CNRS) and NanoBio (Grenoble).
We gratefully acknowledge D. Desplanques and Dr. J.-C. Coll
7 D. Boturyn, J.-L. Coll, E. Garanger, M.-C. Favrot and P. Dumy, J. Am.
Chem. Soc., 2004, 126, 5730.
(a) J. Razkin, V. Josserand, D. Boturyn, Z. Jin, P. Dumy, M. Favrot,
J.-L. Coll and I. Texier, ChemMedChem, 2006, 1, 1069; (b) S. Foillard,
Z. Jin, E. Garanger, D. Boturyn, M.-C. Favrot, J.-L. Coll and P. Dumy,
8
This journal is © The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4159–4162 | 4161

View Article Online
ChemBioChem, 2008, 9, 2326; (c) S. Foillard, L. Sancey, J.-L. Coll, D.
Boturyn and P. Dumy, Org. Biomol. Chem., 2009, 7, 221.
K. Rose, J. Am. Chem. Soc., 1994, 116, 30.
15 We measured the constant of affinity of analogous compounds of
1 and 2 for a
v
b
3
integrins: KD = 4 nM and 42 nM respectively.
9
L. Sancey, E. Garanger, S. Foillard, G. Schoehn, A. Hurbin, C. Albiges-
Rizo, D. Boturyn, C. Souchier, A. Grichine, P. Dumy and J.-L. Coll,
Mol. Ther., 2009, 17, 837.
1
1
0 J. R. Falsey, M. Renil, S. Park, S. Li and K. S. Lam, Bioconjugate Chem.,
2
001, 12, 346.
1 (a) X. Chen, G. S. Lee, A. Zettl and C. R. Bertozzi, Angew. Chem., Int.
Ed., 2004, 43, 6112; (b) G.-A. Cremer, N. Bureaud, D. Leli e` vre, V. Piller,
F. Piller and A. Delmas, Chem.–Eur. J., 2004, 10, 6353; (c) A. Dirksen
and P. E. Dawson, Bioconjugate Chem., 2008, 19, 2543; (d) D. Forget, D.
Boturyn, E. Defrancq, J. Lhomme and P. Dumy, Chem.–Eur. J., 2001,
16 S. K. Akiyama and K. M. Yamada, J. Biol. Chem., 1985, 260, 4492.
17 J. V a´ gner, G. Barany, K. S. Lam, V. Krch nˇ a´ k, N. F. Sepetov, J. A.
Ostrem, P. Strop and M. Lebl, Proc. Natl. Acad. Sci. U. S. A., 1996, 93,
8194.
18 To calculate the ligand density within the resin bead, we considered
that beads are perfect spheres with a uniform distribution of functional
sites. The size of beads varies between 100 and 120 mm, and 1 g of resin
contains 2.9 ¥ 106 beads (information provided by the furnisher). We
7
, 3976; (e) S. Grigalevicius, S. Chierici, O. Renaudet, R. Lo-Man,
E. Deriaud, C. Leclerc and P. Dumy, Bioconjugate Chem., 2005, 16,
149.
1
3
1
1
1
2 S. Foillard, M. Ohsten Rasmussen, J. Razkin, D. Boturyn and P. Dumy,
J. Org. Chem., 2008, 73, 983.
3 E. Garanger, D. Boturyn, Z. Jin, P. Dumy, M.-C. Favrot and J.-L. Coll,
Mol. Ther., 2005, 12, 1168.
4 E. Garanger, D. Boturyn, J.-L. Coll, M.-C. Favrot and P. Dumy, Org.
Biomol. Chem., 2006, 4, 1958.
then estimated the density of ligands to be 600 ± 200 molecules/mm
within a 0.2 nmol/g-bead.
19 A. R. Asthagiri, C. M. Nelson, A. F. Horwitz and D. A. Lauffenburger,
J. Biol. Chem., 1999, 274, 27119.
20 L. Y. Koo, D. J. Irvine, A. M. Mayes, D. A. Lauffenburger and L. G.
Griffith, J. Cell Sci., 2002, 115, 1423.
4
162 | Org. Biomol. Chem., 2009, 7, 4159–4162
This journal is © The Royal Society of Chemistry 2009