Growth of the branched actin filament network
Once a branch is initiated, its free barbed end will grow at a rate [10 mM 71 s 71
(actin-profilin)] limited by diffusion of actin-profilin to the leading edge.
Mogilner estimates that this reaction creates an actin-profilin sink at the
leading edge, with a concentration about half that deeper in the cytoplasm
(Mogilner and Edelstein-Keshet, 2002). Even so, elongation is expected to be
fast, in the order of 100 subunits (0.25 mm) per second, enough to account for
the observed expansion of the leading edge in the fastest cells. Theoretical
calculations (Mogilner and Edelstein-Keshet, 2002; Mogilner and Oster, 1996;
Carlsson, 2002) show that the concentration of growing filaments is su cient
to produce a force to push forward the membrane.
Electron microscopy (Svitkina et al., 1997) indicates that the branches are
short, on the order of 0.5 mm or less, so elongation must be terminated after a
few seconds. This capping rate is consistent with the rate of capping by
heterodimeric capping protein estimated from its concentration and rate
constant for binding barbed ends (Schafer et al., 1996).
Remarkably, most filaments in these branched networks have their barbed
ends orientated toward the front of the cell. Maly and Borisy (2001) proposed
a Darwinian model to explain this bias. Their model is based on the
assumption that interactions of forward pointing filaments with the inside of
the plasma membrane inhibits capping, whereas filaments pointing away from
the membrane are capped rapidly and irreversibly. This assumption is
plausible, since PIP 2 dissociates capping protein from filament ends (Schafer et
al., 1996). The remarkable flatness of a typical leading lamella means that
growth forward is strongly favoured relative to growth in the dorsal direction.
No one has proposed a mechanism to account for this bias.
Assembly of the branched network appears to be self-organizing - an inert
plastic bead coated with a nucleation-promoting factor su ces to induce a force-
producing comet tail in a cell extract (Cameron et al., 1999). If the expanding
network at the leading edge is anchored to the substrate via transmembrane
attachments, growth of the filaments pushes the plasma membrane forward. If
the actin network is not anchored to the substrate, its expansion at the leading
edge results in the whole network sliding as a unit toward the cell centre. Some
cells exhibit a mixture of forward motion of the cell and rear-ward motion of the
network. The molecular clutch presumably consists of links between the network
and cell adhesion molecules, but they still need to be identified.
Filament ageing, remodelling and disassembly
The zone of short, branched filaments at the leading edge is narrow, less than
1 mm wide. Actin filaments further from the front are long and unbranched