Microfilaments (or actin filaments) are the thinnest filaments of the cytoskeleton found in the cytoplasm of all eukaryotic cells. These linear polymers of actin subunits are flexible and relatively strong, resisting buckling by multi-piconewton compressive forces and filament fracture by nanonewton tensile forces. Microfilaments are highly versatile, functioning in (a) cell crawling, amoeboid movement, and changes in cell shape, where one end of the actin filament elongates while the other end contractile, presumably by myosin II molecular motors (though an alternative "actoclampin"-driven expansile molecular motors exists), and (b) actomyosin-driven contractile molecular motors, where the thin filaments serve as tensile platforms for myosin's ATP hydrolysis-dependent pulling action in muscle contraction and uropod advancement.
Actin filaments are assembled in two general types of structures: bundles and networks. Bundles can be composed of polar (all barbed ends point to the same end of the bundle) or non-polar (barbed ends point towards both ends) arrays of filaments . A class of Actin-binding proteins, called cross-linking proteins, dictate the formation of these structures. Cross-linking proteins determine filament orientation and spacing in the bundles and networks. These structures are regulated by many other classes of actin-binding proteins (motor proteins, branching proteins, severing proteins, polymerization promoters, capping proteins, etc).
The thinnest fibers of the cytoskeleton (measuring approximately 6 nm in diameter), microfilaments are formed by the head-to-tail polymerization of actin monomers (also known as globular or G-actin). Actin subunits as part of a fiber are referred to as filamentous actin (or F-actin). Each microfilament is made up of two helical interlaced strands of subunits. Much like microtubules, actin filaments are polarized, with their fast-growing barbed-ends (because of their appearance in electron micrographs after binding of myosin S1 sub-fragments) and a slow-growing pointed-end (again based on the pattern created by S1 binding). The pointed end is sometimes referred to as the minus (-) end and the barbed end is sometimes referred to as the plus (+) end because of the growth rates, but this is nomenclature adapted from the microtubule field, and is not generally accepted in the actin field.
In vitro actin polymerization, nucleation, starts with the self-association of three G-actin monomers to form a trimer. ATP-actin then binds the barbed end, and the ATP is subsequently hydrolyzed with a half time of about 2 seconds and the inorganic phosphate released with a half-time of about 6 minutes, which reduces the binding strength between neighboring units and generally destabilizes the filament. In vivo actin polymerization is catalyzed by a new class of filament end-tracking molecular motors known as actoclampins (see next section). Recent evidence suggests that ATP hydrolysis can be prompt in such cases (i.e., the rate of monomer incorporation is matched by the rate of ATP hydrolysis).
ADP-actin dissociates slowly from the pointed end, but this process is greatly accelerated by ADP-cofilin, which severs ADP-rich regions nearest the (–)-ends. Upon release, ADP-actin undergoes exchange of its bound ADP for solution-phase ATP, thereby forming the ATP-actin monomeric units needed for further barbed-end filament elongation. This rapid turnover is important for the cell's movement. End-capping proteins such as CapZ prevent the addition or loss of monomers at the filament end where actin turnover is unfavourable like in the muscle apparatus.
As a result of the ATP hydrolysis, filaments elongate approximately 10 times faster at their barbed ends than their pointed ends. At steady-state, the polymerization rate at the barbed end matches the depolymerization rate at the pointed end, and microfilaments are said to be treadmilling. Treadmilling results in elongation in the barbed end and shortening in the pointed-end, so that the filament in total moves. Since both processes are energetically favorable, this means force is generated, and the energy ultimately comes from the NTP hydrolization.
Intracellular actin cytoskeletal assembly and disassembly are tightly regulated by cell signaling mechanisms. Many signal transduction systems use the actin cytoskeleton as a scaffold holding them at or near the inner face of the peripheral membrane. This subcellular location allows immediate and exquisite responsiveness to transmembrane receptor action and signal-processing enzyme cascades. Because actin monomers must be recycled to sustain high rates of actin-based motility during chemotaxis, cell signalling is believed to activate cofilin, an actin-filament depolymerizing protein which binds to ADP-rich actin subunits nearest the filament's pointed-end and promotes filament fragmentation, with concomitant depolymerization to liberate actin monomers. The protein profilin enhances the ability of monomers to assemble by stimulating the exchange of actin-bound ADP for solution-phase ATP to yield Actin-ATP and ADP. In most animal cells, monomeric actin is bound to profilin and thymosin-beta4, both of which preferentially bind with one-to-one stoichiometry to ATP-containing monomers. Although thymosin-beta4 is strictly a monomer-sequestering protein, the behavior of profilin is far more complex. Profilin is transferred to the leading edge by virtue of its PIP2 binding site, and profilin also employs its poly-L-proline binding site to dock onto end-tracking proteins. Once bound, Profilin-Actin-ATP is loaded into the monomer-insertion site of actoclampin motors (see below). Another important component in filament formation is the Arp2/3 complex, which binds to the side of an already existing filament (or "mother filament"), where it nucleates the formation of a new actin filament and creates a fan-like branched filament network.
In non-muscle cells, actin filaments are formed at/near membrane surfaces. Their formation and turnover are regulated by many proteins, including
The actin filament network in non-muscle cells is highly dynamic. As first proposed by Dickinson & Purich (Biophysical Journal 92: 622-631), the actin filament network is arranged with the barbed-end of each filament attached to the cell's peripheral membrane by means of clamped-filament elongation motors ("actoclampins") formed from a filament barbed-end and a clamping protein (formins, VASP, Mena, WASP, and N-WASP). The primary substrate for these elongation motors is Profilin-Actin-ATP complex which is directly transferred to elongating filament ends (Dickinson, Southwick & Purich, 2002). The pointed-end of each filament is oriented toward the cell's interior. In the case of lamellipodial growth, the Arp2/3 complex generates a branched network, and in filopods, a parallel array of filaments is formed.
Myosin motors are intracellular ATP-dependent enzymes that bind to and move along actin filaments. Various classes of myosin motors have very different behaviors, including exerting tension in the cell and transporting cargo vesicles.
One proposed model  suggests the existence of actin filament barbed-end-tracking molecular motors termed actoclampin. The proposed actoclampins generate the propulsive forces needed for actin-based motility of lamellipodia, filopodia, invadipodia, dendritic spines, intracellular vesicles, and motile processes in endocytosis, exocytosis, podosome formation, and phagocytosis. Actoclampin motors also propel such intracellular pathogens as Listeria monocytogenes, Shigella flexneri, Vaccinia and Rickettsia. When assembled under suitable conditions, these end-tracking molecular motors can also propel biomimetic particles.
The term actoclampin is derived from acto- to indicate the involvement of an actin filament, as in actomyosin, and clamp to indicate a clasping device used for strengthening flexible/moving objects and for securely fastening two or more components, followed by the suffix -in to indicate its protein origin. An actin filament end-tracking protein may thus be termed a clampin.
Dickinson and Purich (2002) recognized that prompt ATP hydrolysis could explain the forces achieved during actin-based motility. They proposed a simple mechanoenzymatic sequence known as the Lock, Load & Fire Model, in which an end-tracking protein remains tightly bound ("locked" or clamped) onto the end of one sub-filament of the double-stranded actin filament. After binding to Glycyl-Prolyl-Prolyl-Prolyl-Prolyl-Prolyl-registers on tracker proteins, Profilin-ATP-actin is delivered ("loaded") to the unclamped end of the other sub-filament, whereupon ATP within the already clamped terminal subunit of the other subfragment is hydrolyzed ("fired"), providing the energy needed to release that arm of the end-tracker, which then can bind another Profilin-ATP-actin to begin a new monomer-addition round.
The following steps describe one force-generating cycle of an actoclampin molecular motor:
When operating with the benefit of ATP hydrolysis, AC motors generate per-filament forces of 8–9 pN, which is far greater than the per-filament limit of 1–2 pN for motors operating without ATP hydrolysis (Dickinson and Purich, 2002, 2006; Dickinson, Caro and Purich, 2004). The term actoclampin is generic and applies to all actin filament end-tracking molecular motors, irrespective of whether they are driven actively by an ATP-activated mechanism or passively.
Some actoclampins (e.g., those involving Ena/VASP proteins, WASP, and N-WASP) apparently require Arp2/3-mediated filament initiation to form the actin polymerization nucleus that is then "loaded" onto the end-tracker before processive motility can commence. To generate a new filament, Arp2/3 requires a "mother" filament, monomeric ATP-actin, and an activating domain from Listeria ActA or the VCA region of N-WASP. Ther Arp2/3 complex binds to the side of the mother filament, forming a Y-shaped branch having a 70 degree angle with respect to the longitudinal axis of the mother filament. Then upon activation by ActA or VCA, the Arp complex is believed to undergo a major conformational change, bringing its two actin-related protein subunits near enough to each other to generate a new filament gat. Whether ATP hydrolysis may be required for nucleation and/or Y-branch release is a matter under active investigation.