Biology Reference
In-Depth Information
FIGURE 5.8
A diagrammatic representation of the problem of local maxima in a fitness landscape. The ordinate
represents fitness; the abscissa is a cross-section across what would really need to be a vastly multi-dimensional
state space, each location on which corresponds to a specific arrangement of cellular components. The wiggly line
represents the fitness with respect to arrangement. A system that goes consistently upwards without disassembling
itself somewhat and starting again risks being trapped forever in one of the suboptimal, local maxima. A system
that is less unstable has more chance of escaping a local maximum to find the global one ('optimum').
arrangements of a cell can be expressed by adding an extra axis to the phase space graph
described in Chapter 3, fitness being represented by the height along that axis. For any
complex system, the landscape up the fitness axis will be complicated and, as well as
including one large peak of fitness it will include a large number of other peaks, of much
lower fitness than the main peak but of significantly higher fitness than the states in phase
space that immediately surround them (
Figure 5.8
). These small peaks, called 'local maxima'
in the jargon of phase space, are a danger to any self-organizing system because a system that
seeks to maximize fitness can be trapped by a local maximum, unable to cross a zone of lower
fitness to a much greater peak elsewhere. Systems whose structures are unstable even when
in a fitness peak are forced to leave their local maximum from time to time anyway, and, as
they seek again a maximum, they have the chance to find the highest peak of all. Because the
optimum fitness peak tends to have a broad base (in most biological situations), systems that
have found it tend to find it again after small perturbations caused by their inherent instabil-
ities. Therefore, however counter-intuitive it may seem, instability of even optimal cellular
structures is important to the ability of a cell to find its optimum arrangement. The instability
is also very important to the ability of a cell to adapt to changing circumstances.
Naturally, the frequency with which microfilaments are nucleated from particular sites is
modulated by biochemical signalling. That is why stress fibres of fibroblasts are more
numerous in the presence of serum than in its absence.
37
Of the various regulators of the
stress fibre formation, the small GTPase, Rho, is of outstanding importance. Rho can be regu-
lated by a variety of extracellular signals and it regulates the microfilament system by two
main routes (
Figure 5.9
). It encourages the formation of new microfilaments by activating
Formins
38,39
and, acting via ROCK (
ΒΌ
Rho-dependent kinase), Rho also controls the tension
