VOLUME 89, NUMBER 2 PHYSICAL REVIEW LETTERS 8JULY 2002
Dynamic Force Spectroscopy to Probe Adhesion Strength of Living Cells
K. Prechtel,
1
A. R. Bausch,
1
V. Marchi-Artzner,
2
M. Kantlehner,
3
H. Kessler,
3
and R. Merkel
4
1
Lehrstuhl für Biophysik, E22, Technische Universität München, D-85747 Garching, Germany
2
Laboratoire de Chimie des Interactions Moleculaires UPR 285, College de France,
11 Place Marcelin Berthelot, 75231 Paris Cedex 05, France
3
Lehrstuhl 2 für Organische Chemie, Technische Universität München, D-85747 Garching, Germany
4
Institut für Schichten und Grenzflächen, ISG 4, Forschungszentrum Jülich GmbH, 52425 Jülich, Germany
(Received 3 September 2001; published 20 June 2002)
We studied the mechanical strength of the adhesion of living cells to model membranes. The latter
contained a RGD lipopeptide which is a high affinity binding site for a cell adhesion molecule (integrin
a
V
b
3
). Cells adhered specifically to the vesicles. We used micropipette aspiration for breaking this ad-
hesion with well defined forces. Systematic variation of the rate of force application revealed pronounced
kinetic effects. The dependence of the detachment forces on the loading rate was well described by a
power law (exponent 0.4), in agreement with recent theoretical work.
DOI: 10.1103/PhysRevLett.89.028101 PACS numbers: 87.16.Dg, 87.15.By, 87.15.Kg
Cell-matrix adhesion plays an important role in many
physiologically important processes such as tumor devel-
opment and growth or wound healing [1]. Each cell bears a
multitude of different cell adhesion molecules (CAMs) on
its surface. A cell adheres to a given matrix only if some
of its CAMs recognize a complementary part on the ma-
trix. The chemical bonds between complementary CAMs
exhibit comparatively low bond energies (0.5 eV), yet
astonishing specificity, i.e., the ability to distinguish be-
tween very closely related molecules such as blood group
antigens. Therefore these bonds are aptly called “specific
bonds.” To achieve stable cell attachments nature compen-
sates the low energies of these single bonds by using many
of them in parallel. The mechanical strength of local cell
adhesion is a key issue for many physiological processes as
witnessed by the example of cell locomotion where bonds
must be formed and broken in a tightly controlled process.
In recent studies on single specific bonds it was shown
that the dissociation rate of such bonds is dramatically in-
creased upon force application [2]. However, the impli-
cations of this finding for typical physiological situations
where many molecular bonds share the overall mechanical
load remains to be explored. In such a situation one single
broken bond may be reestablished because the other bonds
still hold the cells in place [3]. This implies that bond dis-
sociation and failure of adhesion are no longer synony-
mous as is the case in experiments on single bonds. Thus
we expect entirely new phenomena.
Experiments on enforced separation of adherent cells are
very difficult to interpret because usually more than one
type of CAMs participates, the cell surface is highly cor-
rugated, and the mechanical properties of cells vary dras-
tically as function of time and location within the cell [4].
These complications make it close to impossible to de-
termine the distribution of force between different bonds
and the time course of force application. In this Letter we
introduce a novel approach that allows quantitative mea-
surements of the mechanical strength of specific cellular
adhesion mediated through multiple bonds, a situation
most relevant for most physiological processes. The key
step in this approach is replacing the matrix by a well de-
fined functionalized model membrane, in our case a giant
unilamellar vesicle (GUV). The latter was converted into
a bioadhesion target by incorporation of a specially syn-
thesized lipopeptide [5]. The peptide part of this mole-
cule closely mimics a cell adhesion motif from vitronectin,
a major structural protein of the extracellular matrix [6].
This peptide is recognized by a member of the integrin
receptor family, a
V
b
3
. This cell adhesion molecule is
a receptor for vitronectin and is present at a density of
100 molecules per mm
2
on the surface of cultivated hu-
man endothelium cells [7]. Other CAMs exhibit very poor
binding to this specific peptide. This design of the experi-
ment ensures that just one type of specific bonds is probed.
Moreover, model membranes are homogeneous and their
mechanical properties are very well known [8]. Therefore
we could analyze the mechanical load on the adhesion zone
via the deflection of the model membrane. Hence, the
model membrane served two purposes at the same time:
it presented cell adhesion molecules and acted as a soft
spring for force measurements.
In our experiments we used endothelium cells from
human umbilical cord. Cells were grown in endothelium
cell medium containing 10% fetal calf serum on collagen
beads. Unilamellar giant vesicles containing the lipopep-
tide [9] were injected into the measurement chamber on
the stage of a light microscope. Collagen beads bearing
cells were added. Experiments were performed in phos-
phate buffered saline at a temperature of 37
±
C to ensure
physiological conditions for the cells [10]. In the actual
experiments a collagen bead on which a single endothelial
cell was adhered was picked up by a micropipette [11].
With the opposing pipette a vesicle was collected and
gently brought into contact with an endothelium cell. All
028101-1 0031-9007 02 89(2) 028101(4)$20.00 © 2002 The American Physical Society 028101-1