Actin Cell Cortex: Structure and Molecular Organization

Abstract

The actin cytoskeleton consists of structurally and biochemically different actin filament arrays. Among them, the actin cortex is thought to play key roles in cell mechanics, but remains a poorly characterized part of the actin cytoskeleton. The cell cortex is typically defined as a thin layer of actin meshwork that uniformly underlies the plasma membrane of the entire cell. However, this definition applies only to specific cases. In general, the cortex structure and subcellular distribution vary significantly across cell types and physiological states of the cell. This review focuses on our current knowledge of the structure and molecular composition of the cell cortex.

Cell Mechanics and the Actin Cortex

Cell mechanics is increasingly appreciated to be an important factor in various human pathologies, including cancer and cardiovascular, liver and renal diseases [1]. The shape and mechanical properties of the cell are largely defined by the actin cytoskeleton – a highly polymorphic and multifunctional cellular system that consists of actin filaments (F-actin) organized into various higher-order arrays capable of dynamic remodeling [2]. The most important function of the actin cytoskeleton is to generate force. For pushing, actin filaments typically assemble into branched networks assembled by the Arp2/3 complex [3], while for contraction they form composite networks or bundles with bipolar filaments of myosin II [4, 5].

The cell cortex is usually defined as a plasma membrane-associated part of the actin cytoskeleton, which is thought to be mainly responsible for cell mechanics [6, 7]. The actin cell cortex was the first cytoskeleton component discovered in non-muscle cells [8]. It was initially revealed in large cells, such as free-living amoeba and invertebrate eggs, using electron microscopy (EM) of thin sections of plastic-embedded cells and represented a thick (3–5 µm) and dense cortical layer of actin filaments. These early findings understandably had a great impact in the field, but also provoked an idea that a uniform thick cortex is a universal cellular feature. This idea is still lingering, even though Bray et al. pointed out already in 1986 that a uniform thick cortex was rarely seen in vertebrate cells, except for a much thinner version in some leukocytes, whereas most other cells exhibited only intermittent cortical actin structures [9]. However, heterogeneity of cell cortices across cell types and conditions remains largely unexplored.

The actin cortex remains a rather enigmatic subset of the actin cytoskeleton. At present, it is mostly studied by evaluating its mechanical properties, such as stiffness, viscoelasticity and tension (reviewed in [6, 7]), reconstituting artificial cortices in vitro and modeling the actin cortex in silico (reviewed in [10]). A common constraint for all these approaches is insufficient knowledge of the high resolution structure of the cortex, which is instrumental for guiding reconstitution and modeling studies and interpreting biophysical data.

This review focuses on our current knowledge of the structural and molecular organization of the cell cortex. The available data suggest that rather than being a strictly defined and universally present actin cytoskeleton component, the cell cortex varies broadly in its occurrence, composition, architecture and properties across cell types, subcellular locations and physiological states of the cell.

Cortex Definition(s)

The term “cortex”, in general, refers to an outer layer of an object. In biology, it can refer to a cell, an organ or an organism. The cell cortex is primarily an actin-based structure, although it is also interlinked with other cellular polymers, such as intermediate filaments [11, 12], microtubules [13], septins [14], clathrin lattices [15], and ESCRT complexes [16]. In the cell biology literature, the term “cortex” is used very loosely. On one extreme, it is simply synonymous with the cell boundary. Such practice is convenient if the molecular nature of the cell boundary is not essential. On the other extreme, the cell cortex can be explicitly defined, usually, as a thin layer of F-actin meshwork with specified properties. Although such precise definition(s) could be applicable to specific cases, there are clearly many unfitting examples, which raises a question of whether a universal and explicit definition of the cell cortex is feasible.

A common defining feature of the cell cortex could be its association with the plasma membrane. However, many actin structures in the cell are associated with the plasma membrane. They include actin filaments in lamellipodia (see Glossary), filopodia or microvilli, actin-myosin II bundles anchored to the plasma membrane by various adhesions, the actin-spectrin membrane skeleton, etc. Strictly speaking, all these structures belong to the cell cortex and this is how they are indeed treated in some biophysical studies with a reasonable rationale to average the properties of various cortex subsets in a composite material with uniform properties. However, these defined actin structures are often excluded from consideration in other cortex-focused studies. In such contexts, the cell cortex can be defined as “the total plasma membrane-associated actin cytoskeleton minus its well-defined and previously known components”. A common way of studying such a narrowly defined cortex is by using cells or subcellular regions that lack these “well-defined” F-actin structures. Typical model systems for this purpose are cells that are rounded up for mitosis or after deadhesion, animal eggs or early embryos, some amoeboid cells, plasma membrane blebs and cortical regions away from other actin structures in cells with a heterogeneous cytoskeleton. The question remains whether findings from these specialized systems are broadly applicable.

Cortex Thickness and Density

Quantitative parameters of the cell cortex, such as its thickness and density, have been determined in systems with a relatively homogeneous cortex, such as rounded cells and regions of spread cells away from actin stress fibers. Determination of an average cortex thickness by advanced fluorescence microscopy approaches yielded a range of 50–400 nm depending on the cell type, degree of cell spreading, and stage of cell cycle, but also on the measurement approach [17–21]. This range roughly matches the cortex thicknesses seen in EM images [22, 23], except for ~10–20 nm thick cortex in plasma membrane blebs [24], which would correspond to only 1–3 actin filament layers, suggesting that this actin cortex might not be mechanically strong. Analysis of bleb cortex assembly over time by correlative platinum replica EM (PREM) revealed that the bleb cortex is indeed very thin at the beginning of bleb retraction, but thickens as the retraction progresses, suggesting that the bleb cortex gains contractile strength in parallel as it thickens [25]. Typically, the cell cortex was thicker in less spread cells and maximal in rounded cells [19], although mitotic cells had thinner cortex than cells rounded upon deadhesion [17, 19, 20].

The cortex density in similar systems was determined by scanning EM [17, 24, 26] of detergent-extracted cells, PREM of mechanically ruptured (unroofed) cells [27], and superresolution fluorescence microscopy (SRFM) [21, 28], and atomic force microscopy (AFM) of intact cells [27, 29–31]. In most cases, these studies revealed cortical networks with mesh sizes of 100–200 nm. However, the mesh size could be as small as 20 nm, for example, in mitotic HeLa cells [24], or exceed 300 nm in some interphase cells [21, 30, 31].

Subcellular Distribution of the Cell Cortex

By contrast to uniform cortices in rounded cells, cortical F-actin in spread vertebrate cells in culture exhibits significant heterogeneity (Figure 1, Key Figure). Early studies by thin-sectioning EM revealed only intermittent cortical actin structures in cultured tissue cells [9, 22]. Heterogeneity of the cell cortex on a whole cell level can be better appreciated by PREM, which provides high resolution surface views of the cytoskeleton, and thus primarily reveals the cell cortex [32]. PREM analyses of various normal cultured cells revealed the presence of a dorsal cortex (termed “microfilament or endoplasmic sheath” in these studies) only over the cell body (“endoplasm”) [33–36] (Figure 1A–D). The dorsal cortex became progressively sparser as it approached the peripheral lamella near the cell leading edge, whereas the lamella per se was filled with a 3D network of individual and bundled actin filaments, microtubules and intermediate filaments with no obvious cortical layer of actin filaments. A dorsal cortex covering the entire upper cell surface was found only in epithelial cells that had no free edges and formed cell-cell adhesions around their entire perimeter within cohesive cell colonies [36]. The dorsal actin cortex was even less uniform, sparse and fragmented in neoplastically transformed cells (Figure 1E–H) of both mesenchymal and epithelial origin, in which the degree of cortex abnormalities appeared to correlate with other features of neoplastic transformation [33, 37].

Figure 1. Key Figure. Structure of the Cell Cortex in Normal and Cancer Cells.

(A) Cytoskeleton organization in normal fibroblasts. A relatively continuous cortex (grey-red pattern) formed by long bundled actin filaments covers the dorsal surface of the cell body, but not lamellae and lamellipodia.

(B–D) PREM of the cytoskeleton of a cultured REF52 cell (immortalized rat embryo fibroblast). (B) Overview of a front half of the cell. Dense actin cortex is present in the cell body (lower left). The peripheral lamella has a much sparser cytoskeleton, whereas the cytoskeleton is dense again in lamellipodia. Stress fibers extend from the dorsal cortex toward the cell periphery and terminate in the lamella. Yellow frame is enlarged in C. (C) Transitional zone between the dense dorsal cortex (lower left) and the peripheral lamella (upper right corner) with a sparse cytoskeleton. (D) Densely packed and aligned long actin filaments in the dorsal cortex.

(E) Cytoskeleton organization in neoplastically transformed cells. The cell cortex (grey-red pattern) is sparse and fragmented. Fragments have meshwork-like organization of actin filaments.

(F–H) PREM of the cytoskeleton of a HeLa cervical carcinoma cell in culture. (F) Overview of a front half of the cell. Lamellipodia are small; lamellae are narrow; no stress fibers are seen on the dorsal surface. Fragmented actin cortex extends almost to the leading edge of the cell. Yellow frame is enlarged in G. (G) The dorsal cortex has irregular structure with many holes. (H) The dorsal cortex has patchy structure with the internal cytoskeleton visible through numerous holes. Upper layers of the cytoskeleton that could be considered to represent the cell cortex (arrows) consist of randomly oriented actin filaments of variable lengths.

Cortex Architecture

High resolution analyses of cortex architecture in rounded cells [17] or plasma membrane blebs [24, 25] revealed isotropic filamentous networks of relatively low density. At the other extreme, the dorsal cortices of spread primary fibroblasts [33–35] and normal epithelial cells [36] consisted of long and densely packed actin filaments, which could be mutually aligned and bundled over large distances (Figure 1C,D), or radiate from nodes thus forming large asters of bundles, but also could be randomly intertwined. At a local scale, this organization closely resembled stress fibers, and the cortex split at the cell body/lamella boundary into individual stress fibers that terminated at focal adhesions in the lamella [34] (Figure 1B). In the light of these PREM data, some actin structures observed by fluorescence microscopy, such as dorsal stress fibers that extend from focal adhesions in the lamella toward the dorsal cell surface [38] and contractile actin “caps” located above the nucleus [39], could be a part of the dorsal actin cortex. Local cortical regions containing aligned bundles of actin filaments, as well as asters of converging bundles, have been also detected by cryo EM [40], AFM [29, 31], and SRFM [28], usually in cells of a non-cancerous origin.

The two extreme types of the cortex structure – anisotropic bundles and isotropic meshworks – are not mutually exclusive. As shown by high-resolution time-lapse AFM, cortical bundles and meshworks can form composite arrays and even undergo gradual remodeling in either direction in dorsal cell cortices of live NIH 3T3 cells [29]. However, a ratio between two types of cortex architecture depends on a cell type. Thus, the structure of dorsal cortices was found to be increasingly shifted toward more isotropic actin networks with greater amounts of short misaligned actin filaments in transformed mesenchymal or epithelial cells (Figure 1G,H) in parallel with a degree of neoplastic transformation in these cell lines [33, 37]. These findings raise a possibility that disorganization of the cell cortex architecture could be a signature of neoplastic transformation.

Molecular Organization of the Cell Cortex

The two main actin machineries – protrusive and contractile – are characterized by different designs and rely on distinct sets of actin-interacting proteins [2]. Recent studies indicate that both actin machineries participate in cell cortex assembly. Contractile machinery often uses relatively long actin filaments assembled by the formin family of actin-nucleating and -elongating proteins [41]. The main F-actin partners for contraction are bipolar filaments of nonmuscle myosin II (NMII), which can pull on, cross-link, and organize actin filaments, often eventually leading to the formation of aligned bundles [42]. The key actin partner for protrusion is the Arp2/3 complex, an actin filament nucleator that makes a new actin branch on a pre-existing “mother” filament [41]. Because of the ~70° branching angle, repetitive rounds of Arp2/3-dependent nucleation produce actin filaments oriented at a variety of angles. These filaments are also very short, because their growth is blocked by capping protein soon after nucleation. As a result, branched actin networks are often dense and nearly isotropic and consist of relatively short filaments [3, 43].

Contractile Cortical Cytoskeleton

The cell cortex is traditionally assumed to be a contractile structure. Consistent with this idea, functional approaches provided evidence for the role of formins and NMII in structure and functions of the cell cortex in various cell models [44–48].

The best studied contractile structures in cells are stress fibers, which are formed by mutually aligned and tightly packed actin and NMII filaments [5, 49]. Close structural similarity between stress fibers and highly anisotropic dorsal cortices of polarized primary fibroblasts suggests that the mechanisms of their assembly could be similar and involve NMII-driven reorganization of long formin-polymerized actin filaments into aligned bundled arrays [42, 50, 51]. A high degree of filament alignment under tension requires secure anchorage of actin filaments at cell-matrix or cell-cell adhesions [39, 52]. If such anchor points are weak or more stochastically distributed, the mutual alignment of actin filaments can be limited. For example, strong focal adhesions in polarized fibroblasts can maintain alignment of actin filaments in the dorsal cell cortex nearly at a whole-cell scale [34], whereas smaller and more scattered focal adhesions in non-polarized spreading cells of the same type maintain a complex pattern of local mutual filament alignment [53]. These data suggest that structural organization of the cell cortex, even when it apparently relies on the same molecular machinery, depends on the state of cell adhesion and polarization and can change over time.

The nearly isotropic cortices, such as those in round cells or membrane blebs, also exhibit contractile properties [10, 17, 54] and require activities of NMII and/or formins [44, 47, 48]. A distinctive feature of such cortices is that they are associated with the plasma membrane that is not anchored to extracellular scaffolds, such as a substrate or another cell. Extending the above concept about a positive relationship between anchorage and mutual alignment of actin filaments within the actin cortex, isotropic organization of actin filaments in such cortices can be explained by relatively labile interactions of the cortex with the plasma membrane [55, 56]. In this case, actin filaments are expected to detach under large NMII-generated forces before their significant alignment can occur, although such forces can still be sufficient for isotropic network contraction. Additionally, frequent disruption of actin-membrane bonds can increase actin filament turnover, which would further randomize the filament orientation. The nature of F-actin attachments to a nonadherent plasma membrane is not fully known, but likely depends on the ERM family of membrane-actin linkers [24, 57] or relies on the spectrin-based membrane skeleton as an intermediate [58, 59]. In any scenario, these linkages are clearly weaker than integrin- or cadherin-based adhesions supporting well-aligned stress fibers and equivalent actin-NMII bundles, thus corroborating the idea that fast dynamics of cortex-membrane linkages can favor a relatively isotropic actin filament arrangement in the cortex.

Arp2/3 Complex-Dependent Cortical Cytoskeleton

Recent biochemical and functional studies have shown that cell cortices are not exclusively contractile actin arrays, but also depend on the nucleation of branched actin filaments by the Arp2/3 complex. In fact, localization of the Arp2/3 complex all around the cortex of Acanthamoeba cells was reported when the Arp2/3 complex was originally discovered [60]. The presence of the Arp2/3 complex and its activator, the WAVE complex [41], in the cortex has been subsequently demonstrated by proteomic analysis of isolated cortex-containing plasma membrane blebs and confirmed by light microscopy of blebbing M2 melanoma cells and mitotic HeLa cells [26]. Interference with Arp2/3 complex functions resulted in a reduction in the amount of F-actin in the cortex [26, 61]. Cofilin and capping protein – common components of branched actin networks in other cellular compartments [2] – were also found to regulate cortex thickness [17]. Downregulation of the Arp2/3 complex weakened the coupling between the cortex and the plasma membrane leading to the formation of plasma membrane blebs [62, 63]. Recruitment of the Arp2/3 complex to the cell cortex was enhanced by hypertonic shock [61, 64], whereas Arp2/3-deficient cells failed to properly regulate the cell volume in response to osmotic challenges [61]. These data suggest that cortical Arp2/3 complex not only contributes to the cortex structure, but also plays an important role in the cortex functions during cell shape regulation.

The Arp2/3-dependent machinery functions in the cortex in parallel with the formin/NMII-dependent mechanism. Downregulation of either formin mDia1 or Arp2/3 complex in constitutively blebbing M2 melanoma cells demonstrated that both proteins contribute roughly equal amounts of F-actin to the cortex of blebs [26]. A study that employed a different measurement approach estimated that formin-mediated actin assembly contributed 20–25% of actin filaments to the nonblebbing parts of the cortex in M2 cells and to homogenous cortical regions of HeLa cells, while the Arp2/3 complex was responsible for the rest [45]. At a structural level, interference with Arp2/3 complex in cultured mammalian cells diminished the amount of actin meshworks and increased the proportion of long filaments in the cortex, whereas perturbation of NMII or formin functions decreased the amount of fibrillar cortical components leaving meshwork-like subsets largely unaffected [26, 29, 45]. Notably, formin functions in some of these studies were assayed using the SMIFH2 inhibitor, which was initially thought to be specific for formins, but recently found to also inhibit NMII [65]. However, since formins and NMII cooperate in the formation of the same subset of the cortical cytoskeleton, the results of SMIFH2 treatment are sufficiently interpretable in this context.

The relative contribution of two actin assembly machineries to the cortex structure varies among cell types. The Arp2/3 complex was found to be grossly dispensable for the formation of the cell cortex of mitotic epithelial cells in the developing Drosophila notum [47], primary Sertoli cells [28], and early Caenorhabditis elegans embryos [48], although a small subset of Arp2/3 complex-dependent actin structures was subsequently detected in the latter system [44]. Furthermore, the actin cortex in mouse embryonic stem cells (mESCs) was found to be largely NMII-independent, but required Arp2/3 complex activity to maintain a proper cortex density [21]. It remains presently unknown whether the fibrillar versus meshwork-like organization of the dorsal cortex in normal and transformed cells, respectively [33, 37], reflects greater contribution of formins in normal cells and the Arp2/3 complex in transformed cells. However, this idea seems an intriguing possibility, especially because cortices composed of branched actin networks are softer than those formed by NMII-containing contractile arrays [66], while cancer cells are often softer than normal cells [67, 68].

Spatial Relationship of Two Actin Machineries in the Cortex

The known contractile and propulsive actin machineries in the cell are often spatially segregated, even though they can partially overlap and remodel into each other. For example, NMII filaments rarely enter deeply into branched actin networks in lamellipodia, but can use lamellipodial filaments at the lamellipodial rear to generate stress fibers [4, 35, 69, 70]. Conversely, most actin filaments in stress fibers are associated with tropomyosins [38], which are long coiled-coil proteins that decorate actin filaments along the length. Since tropomyosin binding prevents actin filaments from interacting with the Arp2/3 complex [71], actin filaments in stress fibers lose an ability to serve as mother filaments for Arp2/3-dependent nucleation. The relationship between these two actin arrays in the cell cortex remains poorly understood, largely because direct visualization of actin branches in cells is challenging. So far, occasional actin branches in the cell cortex were directly detected by cryo EM in Dictyostelium cells [40] and by AFM in cultured mammalian cells [30]. However, lower resolution light microscopy techniques gained important insights into this question by taking advantage of fluorescent labeling and functional perturbations of key components of each actin array.

In large syncytial Drosophila embryos, the actin cortex is compartmentalized into repeating domains located above individual nuclei [72, 73]. These domains consist of a central actin cap, which is largely formed by the Arp2/3 complex. The cap is encircled by another set of actin filaments that are formed by the Dia formin and associated with NMII. Remarkably, the centrifugal growth of the Arp2/3-dependent cap physically displaced the surrounding actin-NMII ring, whereas contraction of this ring, reciprocally, counteracted the cap expansion [73]. Thus, protrusive and contractile actin machineries are spatially segregated in these large cortices, but involved in bidirectional interplay. Moreover, such interplay was also detected at an even finer scale within the central caps, where Arp2/3-dependent actin-rich puncta were scattered among small Dia-dependent actin bundles and could push these bundles apart [72].

A combination of Arp2/3-dependent actin puncta and formin-dependent actin bundles was also revealed in the cortex of early C. elegans embryos [44], mESCs [21], and natural killer cells forming an immunological synapse [74]. The SRFM analysis of the mESC cortex showed that the Arp2/3-dependent puncta were shaped as small asters, underwent fast assembly and disassembly and contained other typical components of branched actin networks, such as capping protein, cortactin, coronin 1B, VASP, and WAVE2. [21]. Dynamic Arp2/3-dependent asters were also found in cortices of cultured HeLa and rat basophilic leukemia (RBL) cells [75, 76]. Interference with the Arp2/3 complex in all these systems eliminated actin puncta, but not the linear actin subsets. If tested, linear subsets were found to be sensitive to downregulation of formins and/or NMII. Considering that aster-like morphology of cortical F-actin can also result from NMII activity [77, 78], it is important to use functional, biochemical, and ultrastructural tests to properly characterize the morphological subsets of the cortical actin cytoskeleton.

Concluding Remarks and Future Perspectives

The main difficulty in the “cell cortex” field is an ambiguity of the term itself. At the dawn of cytoskeleton research, when no other actin structures were known, the cell cortex appeared to be an obvious and universal structure. However, its definition was based on a very limited dataset and, as we know now, is not universally applicable. In certain situations, that can justify the original definition of the term, the actin cytoskeleton underneath the plasma membrane indeed is significantly denser than in the cell interior, probably, because most actin nucleators are activated by signaling molecules located at the plasma membrane. However, caution needs to be exercised regarding how broadly this definition can be applied. Furthermore, it is not obvious at present whether the cell cortex (defined as a denser cytoskeleton layer under the plasma membrane) represents a distinct cytoskeleton component with specific molecular and/or structural signatures, as in the case of other well-categorized actin-based structures, such as lamellipodia, filopodia, stress fibers, membrane skeleton, etc. Instead, available data indicate that the cell cortex, when it is present, appears to be a highly polymorphic system. Because of cortex heterogeneity, as well as differences in goals of individual studies, many other definitions of the cell cortex can be rationalized. However, considering the existing ambiguity of the cell cortex concept, it is advisable to explicitly clarify the terminology in each particular study.

The cell cortex can be relatively unambiguously defined only in a limited set of cell models, which are selected specifically based on this feature. Even in these simplified systems, the cell cortex remains a poorly understood subset of the actin cytoskeleton. Recent advances in this area are mostly focused on mechanical and biophysical characterization of cortex properties. However, available data suggest that the cell cortex defined as a submembranous enrichment of actin filaments can be highly heterogeneous, both spatially and temporally, in terms of its structure, molecular organization, and dynamics. A challenge for future studies is to fully understand this heterogeneity, link it to the cell type, a physiological state of the cell, subcellular locations, cell mechanics, and disease (see Outstanding Questions).

Another important issue is that the cell cortex is often treated, explicitly or implicitly, as being physically separate from the internal actin cytoskeleton. In extreme cases, existence of the internal cytoskeleton is totally unappreciated. This view is largely shaped by images of fluorescently labeled F-actin, in which the cytoplasm appears nearly empty if the image contrast is optimized for visualization of a much brighter cortex. This biased perception is exacerbated by greater difficulties of preserving sparse and fragile internal actin filaments, as compared with a more robust cortex. However, improved fixation and imaging protocols recently allowed visualizing F-actin in the internal cytoplasm by fluorescence microscopy [79]. An integrated nature of the actin cytoskeleton can be even better appreciated in PREM images [2, 25, 32, 80]. This integration allows the cell to respond globally to locally applied force [81]. The concept of an integrated cytoskeleton should be seriously considered in future studies of the cell cortex (see Outstanding Questions).

Highlights.

• Actin cell cortex is a heterogeneous and non-ubiquitous actin cytoskeleton component

• Diverse actin filament arrays can mix and match within the cortex in different combinations

• Contractile and protrusive actin machineries cooperate and compete within the cell cortex

Glossary

Adhesion/Deadhesion

processes of attachment/detachment of a cell to an appropriate adhesive surface. Adhesion depends on interaction between the cellular receptors, such as integrins and cadherins, and ligands on the adhesive surface. This interaction initiates cell signaling that causes cell spreading over the surface, as well as strengthening of adhesions. Deadhesion occurs when the receptor-ligand interactions are disrupted, for example, by chelating divalent ions and/or by treatment with proteases. After deadhesion, cells detach and round up

Blebs

a special type of cell protrusion that are driven by intracellular pressure rather than by actin polymerization, as other protrusions, such as lamellipodia and filolodia. During blebbing, the plasma membrane locally detaches from the underlying cell cortex and inflates forming a spherical protrusion called a bleb. Over time, an actin cortex assembles with the bleb and drives bleb retraction in a myosin II-dependent manner

Focal adhesions

sites of strong attachment between the cell and an adhesive substrate, where transmembrane adhesion receptors of the integrin family link the extracellular matrix to the actin cytoskeleton. On the cytoplasmic side, the actin-integrin interaction is mediated by adaptor proteins, primarily by talin and vinculin. Newly formed focal adhesions are small and linked to non-bundled actin filaments. The size and strength of focal adhesions grow with increasing force applied by myosin II via adhesion-associated actin filaments. Most mature focal adhesions are associated with the tips of stress fibers

Isotropic/Anisotropic

in general, these terms refer to equal/unequal properties of a material in different directions. In relation to actin filament organization in the cell cortex, they indicate either a random distribution of actin filament orientations (isotropic) or existence of a preferential orientation (anisotropic)

Lamella

a relatively thin peripheral region of a spread cell that typically lacks large membrane organelles, such as mitochondria. Lamellae are usually located behind the leading edge of the cell, where the cell forms actin-based protrusions. There is usually no lamella at the cell edges that lack protrusive activity

Lamellipodia

flat actin-rich protrusions driven by polymerization of branched actin networks generated by the Arp2/3 complex. When activated by upstream signaling, the Arp2/3 complex binds a pre-existing “mother” actin filament and forms a new “daughter” actin filament as a branch oriented at 70° angle relative to the mother filament. After a short period of elongation, the filaments can be capped by capping protein. Repeated rounds of nucleation, elongation and capping generate dense branched actin networks in lamellipodia, which push the plasma membrane forward

Stress fiber

a mixed bundle of actin and myosin II filaments in non-muscle cells, which possesses contractile properties and is usually anchored at one or both of its tips at focal adhesions