Actin is a ubiquitously expressed protein in both eukaryotes and prokaryotes. Monomers of actin polymerize into actin filaments and represent one of the three major components of the cellular cytoskeleton, also including microtubules and intermediate filaments. The discovery of the protein actin preceded the discovery of actin filaments and is dated back to the end of the 1800s, in terms of its earliest detection. The initial detection of actin was performed using vertebrate muscle tissue wherein it is in great abundance. The complete sequence of actin was initially described at a Cold Spring Harbor Symposium in 1971 and subsequently published in 1973. The solution of the structure of actin greatly benefited from the unexpected observations that actin formed high-affinity complexes with deoxyribonuclease I (DNase I) and profilin, the latter being a protein with now well-understood roles in the regulation of actin biology, while the significance of the former interaction remains unresolved. Because actin bound to DNase I in monomeric form and the interaction prevented the polymerization of actin into filaments, these observations allowed for crystallization and the analysis of its structure in monomeric form. The high-resolution structure of actin was initially published in 1990. For a further review on the history of actin, the reader is directed to Schoenenberger et al. (2011). Based on a PubMed analysis of citations using the term actin, there have been upward of 5,000 publications in one way or another regarding actin for each year between 2012 and 2016. Indeed, as further described below, unveiling the complexity of the actin cytoskeleton is an extremely active area of research with far-ranging significance spanning understanding basic cellular mechanisms and pathophysiology.
The Structure of Actin and Actin Filaments
Actin filaments are helical structures, and their ends are structurally distinct and exhibit different dynamics. The end of the filament at which polymerization is greatest is referred to as the “barbed” end of the filament. In contrast, the opposite end is referred to as the “pointed” end. This nomenclature is derived from the initial structural studies of filaments that involved decoration of filaments with the filament-binding heads of myosin II resulting in the two ends of filaments having different appearances when imaged using electron microscopy. However, it must be noted that of recent, some authors have taken to referring to the + and – ends of filaments for the barbed and pointed ends, respectively. The lengths of filaments in cells vary significantly depending on cellular location and context and can range from submicron to ten of more microns (e.g., in long filopodia). Actin in a physiological salt solution spontaneously nucleates and polymerizes. However, as discussed below, in cells the formation and polymerization of actin filaments are under strict regulation by a large cohort of regulatory proteins that act on either G- or F-actin.
Regulation of Actin Filament Turnover
The nucleation of the filament from a pool of actin monomers
The subsequent elongation of the filament through barbed end mediated polymerization
The depolymerization of the filament from the pointed end and recycling and reloading of the monomers with ATP for reuse in subsequent bouts of steps 1–3
Collectively, this process is referred to as the turnover of actin filaments. It has been estimated that up to 50% of cellular ATP is consumed by the process of actin filament turnover (Bernstein and Bamburg 2003). Moreover, up to 5–10% of total cellular protein has been estimated to consist of actin and its regulatory proteins in diverse cell types, and in muscle cells this estimate is approximately 40% (Amos and Amos 1991; Pollard 2016). Thus, the actin cytoskeletal system is a prominent component of cellular proteomics and can have a large role in cellular bioenergetics. Below is an introduction to the major aspects of actin filament turnover, but due to space limitations it is not exhaustive.
Actin monomers are presumed to be mostly cytosolic and freely diffusing. However, in the cytosol actin monomers are usually bound to accessory proteins that regulate monomer availability for nucleation and polymerization. Profilin binds actin monomers and also promotes the exchange of ADP for ATP on actin monomers priming them for utilization in polymerization. Profilin-actin complexes are recruited to the polymerizing ends of filaments and promote polymerization (see below). Thymosin beta-4 also associates with actin monomers and serves to sequester them from the pool available for polymerization. Sequestered actin monomers are then considered to be released through the action of other actin-associated proteins (e.g., gelsolin and formins discussed further below; Mannherz et al. 2010).
In addition to formins, the polymerization of the barbed end is promoted by additional proteins. However, in order to understand the functions of these additional regulators, it is also important to appreciate the role of proteins that cap the barbed ends of filaments (i.e., capping protein, villin, brevin, severin, adseverin, Esp8, Capz, winfilin, IQGAP, and CapG). The binding of these proteins to the barbed ends inhibits polymerization. Ena/VASP and VopF/VopL promote barbed end polymerization in part by antagonizing the effects of capping proteins while also promoting monomer addition to the barbed end.
In order to undergo turnover, filaments must depolymerize which can also occur in conjunction with the severing of filaments along their lengths. Actin depolymerizing factor (ADF) and cofilin are related proteins that preferentially bind ADP-actin subunits within filaments (i.e., not the most recently polymerized subunits at the barbed end that have not yet hydrolyzed ATP; Fig. 1b). ADF and cofilin both sever filaments and promote depolymerization of the pointed end of the filament, thereby promoting filament turnover.
Organization of Actin Filament Superstructures
In cells actin filaments are organized into a multitude of higher-order structures by accessory actin-binding proteins (Fig. 2). In protrusive structures at the surface of cells, filaments can have very different organizations. In filopodia and microvilli, fingerlike projections, the actin filaments are tightly bundled with their barbed ends directed toward the tip of the filopodium and their minus ends directed toward the cytoplasm (Fig. 2a). The bundling is mediated by proteins like fascin, villin, drebrin, and more. These proteins contain multiple actin filament-binding domains allowing them to link and align filaments. In contrast, in the veil-like lamellipodia found at the leading edge of migrating cells, actin filaments are organized in meshworks and dendritic networks (Fig. 2b). The dendritic networks arise through the action of the Arp2/3 complex described above. The polymerization of the barbed ends of actin filaments just below the plasma membrane is considered to provide the force that pushes the membrane outward thereby allowing cellular protrusion. In contractile structures (e.g., sarcomeres in muscle cells and stress fibers), actin filaments are arranged in antiparallel arrays wherein sets of filaments have their barbed ends directed toward the pointed ends of adjacent filaments (Fig. 2c). This allows the contractile motor protein myosin II, which in muscle cells forms filamentous structures intercalated between filaments, to slide the antiparallel filaments relative to each other and thus generate a contractile actin apparatus. Additional forms of actin filament organization include ringlike structures, as recently described along the axons of neurons, the cytokinetic contractile ring that drives cytokinesis, plasma membrane-associated meshworks (often referred to as cortical actin), and circumferential bundles at the apical portions of epithelial cells where they have a role in cell-cell junctions. For further consideration of the wide variety of actin filament-based structures, the reader is directed to the review by Köster and Mayor (2016).
Functions of Actin Filaments
Actin filaments have a wide array of functions and are necessary for the maintenance of cellular physiology. Their generalized importance is underscored by the consideration that a variety of toxins that target the dynamics of the actin cytoskeleton are made by organisms to poison predators. Cytochalasins are cell permeable fungal-derived toxins that bind the barbed end of filaments and inhibit polymerization and at high doses can also sever filaments. In contrast, phallotoxins are toxins derived from the death cap mushroom and bind the sides of actin filaments blocking their depolymerization and ultimately shutting down turnover as the majority of the available actin becomes polymerized. Both of these classes of toxins are used extensively in research as the first line of investigation probing the functions of the actin cytoskeleton.
A major role of actin filaments is to control cellular morphology and cell motility. The polymerization and organization of filaments strictly underlie the generation of protrusive structures like filopodia and lamellipodia (Fig. 2a, b). The polymerization of actin filaments underneath the plasma membrane generates the force that drives the membrane outward. Another important aspect of protrusive activity is the retrograde flow of actin filaments. Following polymerization underneath the membrane, actin filaments are actively moved centripetally toward the center of the cell by the action of myosin II. Based on the organization of actin filaments in protrusive structures and the localization of myosin II, the current best model for how myosin II drives retrograde flow is the dynamic contraction model (Fig. 2d; Svitkina et al. 1997). In this model myosin II acting at the base of the F-actin network aligns antiparallel filaments and through its motor activity pulls them together. This then in turn results in net movement of the network centripetally away from the leading edge due to the interconnected nature of the filaments. The behavior of the leading edge of the cell is thus largely determined by the outward force generated by polymerization at the membrane and the rate of retrograde flow. If the former is greater than the latter, then the leading edge advances, and conversely if retrograde flow is faster than the polymerization at the membrane, the edge retracts toward the cell body. Actin filaments also have roles in the formation and maintenance of substratum attachment points during cell migration and allow myosin II to generate contractile forces on the substratum contributing to the ability of the cell to migrate. One final component of the mechanism that regulates the behavior of leading edges is described by the clutch hypothesis that has garnered experimental support in recent years (Case and Waterman 2015). The clutch is a mechanism whereby transmembrane proteins (e.g., receptors for extracellular matrix proteins or cell adhesion molecules) bind to the underlying actin cytoskeleton through their cytoplasmic domains. When these transmembrane proteins bind their extracellular ligands, they become physically immobilized. This interaction in turn serves to generate a clutch that counters the efforts of myosin II in driving retrograde flow by retraining the movement of the filaments while also imparting the intracellular forces generated by myosin II on the actin cytoskeleton onto the extracellular environment.
Actin filaments also have organizational roles in the cell serving as scaffolds for other proteins and regulating organelle distribution and dynamics. For example, actin filaments tether subsets of mitochondria in neuronal axons, serve as the substratum for the myosin motor-dependent redistribution of mitochondria during fission in yeast, and along with myosin II activity are also a required component of the mechanism of mitochondrial fission. One of the classical roles for cortical actin filaments is the regulation of the organization of spectrin filaments and maintenance of membrane structure. Cortical actin filaments have also been involved in endo- and exocytosis. An emerging novel theme is that actin filaments also regulate the function of membrane channels including channel gating (Sasaki et al. 2014). While the list of the functions of actin is more extensive than the examples provided, the emerging role of actin in the nucleus deserves mention. Actin has no nuclear localization sequence but is instead imported in association with importin 9 and possibly cofilin. However, actin does have two nuclear export sequences, and it has been documented to traffic in and out of the nucleus. Nuclear actin has been involved in transcriptional activation, nuclear export and editing of mRNAs, chromatin remodeling, DNA repair, and nuclear envelope assembly. The functions of actin in the nucleus can be due to either its G or F forms. G-actin is a component of multiple protein complexes in the nucleus, and it is considered that through its ATPase activity may serve as a regulator of these complexes. For further information, the reader is directed to a recent review on this topic (Kristó et al. 2016).
In vertebrates there are three classes of actin isotypes: α-actin, β-actin, and γ-actin. α-Actin is expressed in cells of muscle lineage, and different isoforms are expressed in skeletal, smooth, and cardiac muscle. γ-Actin has two isoforms, the ubiquitously expressed cytoplasmic isoform and one expressed in smooth muscle. β-Actin is ubiquitously expressed in most cells. However, the relative levels of expression of isoforms vary between cell types. Most of the differences between isoforms reside in the first N-terminal 20 amino acids, and the greatest sequence divergence between actins is of 7%. The functional significance of different isoforms is still not completely understood, but evidence has been presented for differences in polymerization dynamics (under varied temperatures and ionic contexts), the binding of G-actin isoforms to monomer-binding proteins (e.g., thymosin β-4), and the actin filament-dependent activation of myosin II motor function (Herman 1993). β-Actin is likely to have specialized functions in cell motility as it has been reported to preferentially incorporate in filaments at the leading edges of cells and overexpression of β-actin but not γ-actin promotes cell migration (Cheever and Ervasti 2013). However, other studies have not found evidence for the specific subcellular distribution or function of these two isoforms, and this remains an area of intriguing investigation. Differences in isoform polymerization kinetics as a function of the ionic environment may underlie some of these discrepancies (Cheever and Ervasti 2013).
Co- and Posttranslational Modifications of Actin
Actin undergoes numerous co- and posttranslational modifications including acetylation, N-terminal arginylation, ADP-ribosylation, methylation, oxidation, nitration, nitrosylation, phosphorylation (at 35 different serine, threonine, and tyrosine sites), ubiquitylation, O-GlcNAcylation, and others. Actin methylation has been reported to regulate actin structural folding and maintenance and regulate actin-myosin II contractility and actin metabolism. ADP-ribosylation can inhibit or promote actin polymerization depending on the residue that underwent modification. β-Actin N-terminal arginylation promotes leading edge actin filament assembly in migrating cells. O-GlcNAcylation has been reported to modulate actin-myosin II-based contractility and also actin filament levels, which may have implications for disease states (e.g., diabetic pathology). Redox-related modifications are generally considered to impair actin filament polymerization and maintenance. The effects of phosphorylation on actin polymerization are dependent on the specific sites phosphorylated. For example, protein kinase C- and protein kinase A-mediated phosphorylation promotes and inhibits polymerization, respectively. For further information on actin co- and posttranslational modifications, the reader is directed to Terman and Kashina (2013).
Cytoplasmic Subcellular-Localized Translation of Actin
The localized translation of a wide array of mRNAs in subcellular cytoplasmic domains is appreciated to play important roles in development and cellular physiology (Richter 2001; Jung et al. 2014). mRNAs are actively transported in cells incorporated within ribonucleoprotein particles (RNPs) that also serve to control the translation of the cargo mRNAs. β-Actin mRNA binds a number of proteins found in RNPs with zipcode-binding protein 1 (ZBP1) being the most widely investigated. The association of β-actin mRNA, and indeed many other subcellularly targeted mRNAs, with components of RNPs is through the 3′ untranslated region. β-Actin mRNA was initially found to target to the leading edge of migrating cells, and its local translation is considered to have a role in the promotion of cell migration. Similarly, the axonal translation of β-actin mRNA is reported to have important roles in the morphogenesis and regeneration of neuronal axons. In neurons signals that regulate axonal morphology not only promote the localized axonal translation of actin but also the localized translation of a host of actin regulatory proteins. Collectively, these considerations indicate that when studying the translation of actin, cytoplasmic mRNA pools and their subcellular distribution warrant consideration.
Actin Mutations and Disease
Disease states have been linked to single amino acid mutations in multiple actin isoforms, the majority of which are of an autosomal dominant nature. These include skeletal muscle myopathies and cardiac myopathies due to single-site mutations in skeletal and cardiac α-actin, respectively. Mutations in smooth muscle α-actin are linked to vascular disorders. Mutations in smooth muscle γ-actin have been associated with familiar visceral myopathy. Mutations of cytoplasmic γ-actin have been shown to result in deafness, and it is notable that γ-actin is the predominant isoform in the cochlear hair cells responsible for hearing. Mutations in both cytoplasmic γ-/β-actin have been linked to Baraitser-Winter syndrome, a developmental disorder characterized by aberrant nervous system development, deafness, facial abnormalities, decreased stature, and impaired mobility. As the mutations are isoform specific, the effects are likely to be exhibited in cells in which the mutated isoform predominates, as noted above for γ-actin and deafness. For an in-depth review and discussion of the possible mechanistic consequences of these disease-linked actin mutations, the reader is directed to Rubenstein and Wen (2014).
In summary, although we have learned volumes about actin, understanding the full spectrum of the biology of actin remains a fundamental problem. The complexity of the actin cytoskeleton is daunting and requires an understanding not just of actin itself but also of the multitude of proteins that control actin in both monomer and filamentous form. Furthermore, each of these proteins, including actin, is under multiple levels of regulation ranging from the transcriptional/translation to the posttranslational levels. As evidenced by the highly localized regulation of the actin cytoskeleton at the subcellular level, often over domains as small as a half a micron or less, it will also be necessary to understand the dynamics and organization of upstream regulatory mechanisms (e.g., localized signaling events). Recent breakthroughs in the understanding of the role of actin in the nucleus are opening a new venue of investigation in the biology of actin, and the implications are just beginning to emerge. This chapter provides an introduction to the most salient aspects of actin and the actin cytoskeleton, and the reader is directed to more in-depth treatment of each topic throughout.
Finally, one issue not discussed in this chapter is the interplay between actin filament-binding proteins. The reader should be aware that decoration of a filament with one actin-binding protein can promote or inhibit the association of a different protein with filaments through either shared binding sites, steric hindrance, or the modulation of the helical structure of the filament. This last point serves to underscore the notion that to understand the function of the actin-based cytoskeleton requires a system-level approach considering the myriad of actin monomer or filament-binding proteins in a subcellular context.
- Pollard TD. Actin and actin-binding proteins. Cold Spring Harb Perspect Biol. 2016;8(8). pii: a018226. doi:10.1101/cshperspect.a018226. PMID:26988969.Google Scholar