in cells treated by BDM or transfected with caldesmon (Riveline et al., 2001).
Thus, individual focal adhesions function as miniature mechanosensors: they
respond to an increase of applied tension force by activation of assembly, and
to a reduction of such force by disassembly (Figure 5.2).
The mechanism of such unusual behaviour of focal adhesions is not yet
clear, and its understanding requires more detailed knowledge of the
organization and dynamics of these structures than we presently have.
Perhaps mechanical force applied to the focal adhesion induces alteration(s) in
the conformation of some of its components and/or in their mutual positions
(Geiger and Bershadsky, 2001, 2002). This stretching could increase the
probability of incorporation of new components in such a way that the focal
adhesion assembles, preserving its self-similarity. The observation that
relaxation of tension by a variety of myosin II inhibitors leads not just to
cessation of growth, but to a rapid disassembly of focal adhesions suggests
that these structures are intrinsically dynamic: they continuously loose 'old'
subunits and incorporate the 'new' ones in a tension-dependent fashion.
Recent direct observations of b3-integrin dynamics in focal adhesions using
the technique of fluorescence recovery after photobleaching (FRAP), indeed,
revealed high turnover rate (Ballestrem et al., 2001) and dependence of
fluorescence recovery on the myosin II activity (Tsuruta et al., 2002). Among
other factors, the protease calpain specifically localized at the focal adhesions
might be involved in their rapid disassembly (Bhatt et al., 2002). Unlike focal
adhesions, other types of integrin-mediated adhesion structures such as focal
complexes and fibrillar adhesions do not disassemble following inhibition of
myosin II contractility (Riveline et al., 2001; Zamir et al., 2000).
The dependence of focal adhesion assembly on applied force creates a
positive feedback loop amplifying the growth of these structures in cells
attached to a solid substrate. In fact, since several components of focal
adhesions are actin-associated proteins that can bind and perhaps even
nucleate actin filaments, an increase in the focal adhesion plaque size will lead
to an increase in the number of actin filaments associated with it. Increasing
the actin filament number will lead, via their interactions with myosin II
motors, to augmentation of the centripetal tension force applied to this focal
adhesion. Force augmentation will in turn promote further focal adhesion
growth (via additional activation of the focal adhesion mechanosensor),
further increasing the bundle of associated actin filaments, and so on. Thus, if
this mechanism were not regulated, focal adhesions would grow indefinitely,
being limited only by a possible rupture (if the final force applied to it by the
cell exceeds a certain threshold).
Obviously, the cell needs a physiological mechanism(s) that can attenuate
this positive feedback loop and can weaken or disrupt focal adhesions, as
necessary. In particular, such mechanisms are needed to enable focal
adhesion-forming cells to migrate along the substrate (otherwise they simply