Biology Reference
In-Depth Information
concentrated in the leading lamella. Seminal experiments with photobleaching
of fluorescent actin (Wang, 1985) and photoactivation of caged-fluorescent
actin (Theriot and Mitchison, 1991) demonstrated that these actin filaments
assemble near the leading edge and turn over on a time scale of tens of seconds
deeper in the cytoplasm. More recent work with fluorescent speckle
microscopy (Watanabe and Mitchison, 2002) suggests that assembly and
disassembly are actually distributed processes: assembly is strongest at the
leading edge but occurs deeper as well; and filaments turn over in a broad zone
behind the leading edge. A key point is that external signals from chemotactic
attractants and repellents guide assembly temporally and spatially. On a time
scale of seconds cells can re-orientate toward a new source of attractant or
turn away from repellents (Gerisch, 1982; Bourne and Weiner, 2002).
Deciphering the molecular basis of this localized assembly and disassembly
of actin filaments might have proven impossible, given that it requires
coordinating the activities of millions of protein molecules. However, the
system has proven to be remarkably amenable to reductionist analysis coupled
with insightful microscopy. All of the major components have been identified
and their cellular concentrations measured in selected systems. A crystal
structure of each component is available. Rate and equilibrium constants for
most of the reactions are known. In vitro motility assays based on the
propulsive comet tails of bacteria allow reconstitution of the whole assembly
and disassembly process (Loisel et al., 1999). Many of the reactions of the
purified proteins have also been visualized in real time by fluorescence
microscopy (Maciver et al., 1991; Amann and Pollard, 2001; Ichetovkin et al.,
2002; Fujiwara et al., 2002). Experiments in genetically tractable organisms
allow tests for physiological functions, and mathematical models (Mogilner
and Edelstein-Keshet, 2002) of the whole cycle of assembly and disassembly
now guide mechanistic experiments.
We now have a first generation, quantitative model for the process (Figure
1.1) called the dendritic nucleation, array treadmilling hypothesis (Mullins et
al., 1998a; Svitkina and Borisy, 1999; Pollard et al., 2000). This model is an
over-simplification, since it considers only five protein components and ATP,
but these molecules are sucient to reconstitute motility of bacterial comet
tails (Loisel et al., 1999) and thus are likely to be the core components that
operate a more elaborate system in cells. I will explain our current
understanding of this machine and indicate some points that need clarification
or possible revision.
Inventory of components
In vitro reconstitution of bacterial motility from purified proteins (Loisel et al.,
1999) showed that the essential components of the system are actin, Arp2/3
