Wound Healing from an Actin Cytoskeletal Perspective

Abstract

Wound healing requires a complex cascade of highly controlled and conserved cellular and molecular processes. These involve numerous cell types and extracellular matrix molecules regulated by the actin cytoskeleton. This microscopic network of filaments is present within the cytoplasm of all cells and provides the shape and mechanical support required for cell movement and proliferation. Here, an overview of the processes of wound healing are described from the perspective of the cell in relation to the actin cytoskeleton. Key points of discussion include the role of actin, its binding proteins, signaling pathways, and events that play significant roles in the phases of wound healing. The identification of cytoskeletal targets that can be used to manipulate and improve wound healing is included as an emerging area of focus that may inform future therapeutic approaches to improve healing of complex wounds.

Wound healing is an essential process that ensures body fluid preservation, hemostasis, thermoregulation, and protection against physical injuries and environmental insults (Adzick 1992). It is a highly orchestrated process consisting of four overlapping phases: hemostasis, inflammation, proliferation, and remodeling (Shaw and Martin 2009), which, in healthy individuals, leads to the restoration of tissue function. While the gross physiological end point is a healed piece of skin, this is the culmination of a complex cascade of highly controlled and conserved cellular and molecular processes involving numerous cell types, proteins, extracellular matrix (ECM) molecules, and sensing of environmental cues, all of which involve or are regulated by the cytoskeleton (Abreu-Blanco et al. 2012).

The actin cytoskeleton is a microscopic network of filaments present within the cytoplasm of cells, which gives them shape and provides mechanical support required for cell movement and proliferation. In eukaryotic cells, the cytoskeleton consists of three types of filaments that have different size, dynamics, and protein content: microfilaments, intermediate filaments, and microtubules all of which are regulated by a number of cytoskeletal remodeling proteins (Fletcher and Mullins 2010). This review will overview the processes of wound healing from the cellular perspective particularly in relation to the role of the actin cytoskeleton. The focus will primarily be related to actin and its binding proteins, which are known to play significant roles in the wound repair process.

CYTOSKELETAL BASICS

Cytoskeletal Filaments

The cytoskeleton is composed of three major types of filaments that are involved in cell division, cell motility, and cytoplasmic streaming. Actin filaments are composed of globular actin (G-actin) monomers, which interact with other G-actins to polymerize and create double-strand helix actin filaments (F-actin) (Hohmann and Dehghani 2019). Differences in the rate of actin polymerization (elongation) between the two ends of actin filaments create polarity, and the end with the highest rate of actin polymerization is called the plus or barbed end, whereas the end with slower polymerization rates is called the pointed or minus end (Fig. 1A; Pollard 2016). Intermediate filaments include keratins and vimentin and can be found in homopolymers and heteropolymers with each other (Fig. 1B). Emerging evidence demonstrates a vital role of these filaments in mechanically introduced signal transduction, cell proliferation, and migration (Sanghvi-Shah and Weber 2017). Microtubules are large polar and dynamic cytoskeletal structures consisting of α- and β-tubulin heterodimers (Goodson and Jonasson 2018). Similar to microfilaments, microtubules also have polymerization and shrinkage states in their ends; however, through a process called dynamic instability, these ends can switch from a polymerizing, extension end to a depolymerizing, shrinkage end, and vice versa (Fig. 1C). Microtubules are involved in cell motility, intracellular transport, mitosis, and maintenance of cell shape (Horio and Murata 2014).

Figure 1.

Sketch of the three main components of the actin cytoskeleton. (A) Microfilaments: F-actin filaments consisting of G-actin monomers. Actin bundles are created from F-actin filaments cross-linked by fimbrin or α-actinin. Actin networks are composed of F-actin filaments cross-linked by filamin or Arp2/3. (B) Intermediate filaments (IFs): IFs are built from IF monomers that bind together to form dimers, tetramers, octamers, and long IFs. (C) Microtubules: Microtubules are built from α- and β-tubulin dimers. Microtubules are dynamic structures that have polymerization and shrinkage states at their ends.

Actin-Binding Proteins

The assembly and disassembly of actin filaments is regulated by different groups of proteins termed actin-binding proteins (ABPs) (Pollard 2016). Actin nucleation/elongation factors such as formins, Arp2/3, and profilins promote the nucleation of actin monomers and the elongation of actin filaments by facilitating the addition of G-actins to the plus ends of the filaments. They also act as nucleotide-exchange factors and are involved in the removal of capping proteins (Chesarone and Goode 2009). These proteins assist lamellipodia and filopodia formation to facilitate cell migration. Actin nucleation is promoted by several nucleation-promoting factors such as the Wiskott–Aldrich syndrome protein (WASP) and WASP-family verprolin-homologous (WAVE) family protein, which interact with actin nucleation factors and regulate their function (Stradal et al. 2004). Conversely, cytosolic protein thymosin β4 inhibits actin filament assembly (Goldschmidt-Clermont et al. 1992).

The actin-depolymerization factor (ADF)/cofilin family of ABPs regulate actin filaments by severing and breaking them into shorter filaments. These severing proteins are responsible for actin turnover and are able to create new plus ends for actin polymerization (Pavlov et al. 2007). Severing proteins may stay physically located at the plus end of actin filaments forming a cap that restricts further elongation and promoting branched filament networks (Edwards et al. 2014). Actin filaments can also be capped by other proteins such as CapZ protein and tropomodulin. These do not sever the filaments but only cap the plus and minus end, respectively, and stabilize the actin filaments (Bao et al. 2012). When actin filaments are created, they mainly assemble into either packed actin bundles or semisolid gel-like actin networks. In bundles, the actin filaments are cross-linked by small cross-linking proteins such as α-actinin and fimbrin (Klein et al. 2004). In networks, the actin filaments are loosely cross-linked by large protein families such as filamin (Stossel et al. 2001).

Gelsolin is known as the most abundant and important ABP and heads the superfamily to which it lends its name, which consists of seven other proteins: Flightless I, supervillin, villin, advillin, CapG, adseverin, and adseverin D5 (Fig. 2; Silacci et al. 2004). Gelsolin is a Ca²⁺-dependent, actin-nucleating, capping, severing, and monomer-sequestering protein that regulates actin cytoskeleton formation and reorganization (Gremm and Wegner 2000). Its function is regulated by PIP2 levels, temperature, and pH (Badmalia et al. 2017). Gelsolin has three isoforms that arise by alternative splicing from the same gene: intracellular or cytoplasmic gelsolin (cGSN), extracellular or plasma gelsolin (pGSN), and gelsolin 3 (Vouyiouklis and Brophy 1997). They all have multiple intra- and extracellular functions and are involved in cellular interactions (Li et al. 2012). pGSN has an extension of 23 amino acids from its amino terminal, which leads to a slightly higher molecular weight (about 83 kD) than cGSN (80 kD) (Li et al. 2012). All gelsolin isoforms consist of six repeats of gelsolin domains (G1–6) sequencing from the amino terminus, and each contains a Ca²⁺-binding site (Burtnick et al. 1997). However, just three of these domains, G1–3, are Ca²⁺-dependent, whereas the function of the G4–G6 domains does not require Ca²⁺. Other members of the gelsolin superfamily also contain a variety of copies of gelsolin domains. CapG has three gelsolin-type domains, supervillin has five, and each of the others have six repeats (Fig. 2; Nag et al. 2013).

Figure 2.

Structure and function of the gelsolin family. Schema showing gelsolin family of actin-binding proteins with conserved gelsolin domains (G1–6) and additions including a leucine-rich repeat (LRR), nuclear localization signal (NLS), or a carboxy-terminal head piece domain (H).

Flightless I (Flii) is a unique member of the family with multifunctional roles in wound healing (Cowin et al. 2007; Kopecki and Cowin 2008; Strudwick and Cowin 2020). It has all six of the gelsolin-type domains, but only two Ca²⁺-binding sites (Nag et al. 2013) and has an additional amino-terminal leucine-rich repeat (LRR) domain enabling interactions with other proteins and lipids (Fig. 2; Liu and Yin 1998). Whereas Flii binds both globular (G)- and filamentous (F)-actin through the gelsolin domain (Liu and Yin 1998) inhibiting polymerization and capping the barbed end of F-actin, it does not possess actin-severing ability (Mohammad et al. 2012). Flii localizes with actin- and β-tubulin-based structures at the cell periphery (Davy et al. 2001), acting as a regulator of migration in multiple cell types (Strudwick and Cowin 2020). Similar to gelsolin, Flii is also found in both intracellular and extracellular compartments and plasma Flii (pFlii) is found in plasma and wound fluid of human patients (Ruzehaji et al. 2014). Flii further influences cytoskeletal dynamics, enhancing the formation of actin into filaments by Rho-induced formins, activating them by preventing their autoinhibition and recruiting G-actin and preformed F-actin (Goshima et al. 1999; Higashi et al. 2010).

Rho GTPase Family

Acting upstream of the ABPs, Rho GTPases (Rho, Rac, and Cdc42) act as key regulators of the actin cytoskeleton, controlling signal transduction pathways by linking cell surface receptors to intracellular responses (Raftopoulou and Hall 2004). Rho GTPases broadly influence many aspects of cell behavior including microtubule dynamics, vesicular trafficking, and gene expression (Raftopoulou and Hall 2004). Stress fiber formation and cell contraction is more specifically controlled by Rho, the formation of lamellipodial protrusions and membrane ruffles by Rac and filopodial extensions to control cell polarity by Cdc42 (Hall 2005). Cdc42 is also a regulator of filopodia formation, which mediates WASP and Arp2/3 factors and stimulates actin polymerization (Tomasevic et al. 2007).

Rho-associated protein kinase (ROCK) is the major effector protein of Rho, which phosphorylates several substrates on threonine and serine and regulates cell proliferation, migration, and adhesion (Amano et al. 2010). Active Rho also interacts with the mDia protein isoforms (mDia1-3), members of the formin family of actin-nucleating factors, which together with ROCK mediate Rho-induced actin reorganization (Lammers et al. 2008; Kühn and Geyer 2014). The Rho GTPases affects several additional signal transduction pathways, including those mediated by the transcription factor NF-κB in turn, contributing to microtubule dynamics, gene transcription, cell–cell adhesion, cell-cycle progression, and enzyme transport, among other intracellular mechanisms (Van Aelst and D'Souza-Schorey 1997; Tong and Tergaonkar 2014).

Caveolin1 appears to play a critical role linking Rho GTPases and signaling molecules localized together at the lipid-rich microdomains at the plasma membrane known as caveolae, with changes in its expression being linked with impaired healing (Jozic et al. 2021). Thus, the dynamic control of the cytoskeleton is critically involved in the regulation of cell shape, migration, and proliferation, processes that are fundamental to all phases of wound healing.

CYTOSKELETAL INVOLVEMENT DURING HEMOSTASIS AND INFLAMMATION

In response to injury, disruption of blood vessels leads to leukocytes being released into the wound site and activation of a chain of events, which result in coagulation, platelet aggregation, and fibrin clot formation (Laurens et al. 2006). Platelets undergo a shape reformation to enable adhesion, activation, and aggregation, a process that is facilitated by cytoskeletal reorganization (Shin et al. 2017). Upon wounding, the actin cytoskeleton in platelets communicates with the ECM through the dystrophin–glycoprotein complex (DGC) and regulates cell morphology depending on the environmental geometry (Cerecedo 2013). Platelets adopt a spherical shape after adhesion to the matrix through reorganization of the cytoskeleton to filopodia and lamellipodia structures (Escolar et al. 1986). These actin structures assist platelet attachment and activation, enabling them to spread and sense the environment. Finally, to perform clot contractions, actin filaments organize into higher-order structures including stress fibers and contractile rings, which are composed of bundles of actin filaments and myosin II, allowing nonmuscle cells to apply contractile forces (Bearer et al. 2002).

The recruitment, circulation, and trafficking of innate immune cells, such as neutrophils, monocytes, and adaptive immune cells, such as B and T cells, to the wound bed, are crucial for a rapid and appropriate response to potential invading pathogens (Gonzalez et al. 2016). Leukocyte extravasation involves sequential and discrete stages including capture, rolling, adhesion, crawling, transendothelial migration (TEM), and migration through ECM (Muller 2013). The binding of L-selectin on leukocytes with E- and P-selectin on activated endothelial cells (ECs) results in the process of “leukocyte capture,” which slows the movement of the cells within the blood vessels. Subsequent breakage of these selectin tethers at the front and rear of the cells enables them to roll along the disrupted blood vessel wall (Anderson et al. 2000). An intact actin cytoskeleton plays a vital role in persistent rolling (Anderson et al. 2000), likely because the L-selectins on leukocytes are supported through their interactions with α-actinin, whereas E-selectin on ECs interacts with a complex of ABPs, including cortactin, filamin, vinculin, α-actinin, paxillin, and focal adhesion kinase (Dwir et al. 2001).

Rolling provides further activation of leukocytes by chemokines and other proinflammatory agents leading to the expression of integrins on both leukocytes and ECs (Zarbock et al. 2012). Integrins are heterodimer receptors composed of α and β subunits, which, upon activation through ligand binding, interact with the actin cytoskeleton through the β subunit cytoplasmic domain (Delon and Brown 2007). Integrin-mediated adhesions activate several tyrosine kinases, including FAK and Src, which eventually leads to the activation of Rac and Cdc42 on opposing sides of the cell to promote actin polymerization and cell polarization (DeMali et al. 2003). The activation of small GTPase Rac proteins and Cdc42 at the front edge leads to the activation of WASP, which organizes actin polymerization via the Arp2/3 complex, resulting in the formation of lamellar F-actin filaments and pseudopods around the leading edge (Innocenti 2018). Furthermore, GTPase Rac and Cdc42 inactivate both myosin light chain (MLC) kinase (MLCK) and myosin heavy chain II, which reduces contraction in the leading edge leading to its extension in migrating cells (Infante and Ridley 2013). Simultaneously, GTPase Rho protein activation at the rear of the cell leads to the formation of the uropod, an area poor in F-actin and rich in MLC and adhesion molecules, that is required for retraction of the trailing edge (Worthylake et al. 2001). This polarization provides a pushing force for fast, amoeboid-like motility, allowing cells to crawl along the endothelium in search of a proper location to breach the endothelial layer (Fig. 3; Ren et al. 2019).

Figure 3.

Overview of the multiple steps of leukocyte extravasation. (A) Hematoxylin and eosin-stained tissue showing leukocyte extravasation. Arrows show neutrophils that are transmigrating from a blood vessel to the extracellular matrix (ECM). (B) Schematic overview of the multiple steps of leukocyte extravasation. Capture and rolling of leukocytes along the endothelium are mediated through interactions between E-selectins on endothelial cells (ECs) and L-selectins on leukocytes. The expression of integrins on leukocytes and ECs establish the firm adhesion between these cells. Leukocyte transendothelial migration (TEM) is facilitated by the adhesion molecule, platelet/EC adhesion molecule 1 (PECAM-1), intercellular adhesion molecule 1 (ICAM-1), and vascular cell adhesion molecule 1 (VCAM-1). TEM occurs through EC–cell junctions that are predominantly composed of adhesion molecules including PECAM-1, the junctional adhesion molecules (JAMs), and vascular endothelial (VE)-cadherin.

Transmigration mainly occurs through EC–cell junctions, which are predominantly composed of adhesion molecules including the platelet/EC adhesion molecule 1 (PECAM-1), the junctional adhesion molecules (JAMs), and vascular endothelial (VE)-cadherin (Reglero-Real et al. 2016). Firm adhesion of leukocytes induces cytoskeletal remodeling in ECs, mainly via depolarization of actin cytoskeleton, and EC contraction to form a gap that acts as an open transmigratory pore (Cook-Mills et al. 2004). The firm adhesion of leukocytes to the endothelial layer is enabled by enrichment of intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) on the uropod of leukocytes as they crawl. Both ICAM-1 and VCAM-1 are located over F-actin-rich vertical microvilli projections that form before TEM and called “transmigratory cups” (Carman and Springer 2004; Yang et al. 2006), but are also clustered in an endothelial actin-rich docking structure containing other cytoskeletal components such as α-actinin, vinculin, and vasodilator-stimulated phosphoprotein (VASP), which anchors the leukocyte during TEM (Barreiro et al. 2002; Carman and Springer 2004; Schnoor 2015).

Once within the wound site, inflammatory cells use phagocytosis, exocytosis, and NETosis to fight pathogens and clear cellular debris. Phagocytosis of invading pathogens or cellular debris requires a coordinated adhesion between target and phagocytic cells as well as a dramatic change in cell shape driven by actin remodeling, which may be dependent upon either immunoglobulin receptors (FcRs) or complement receptors (Vicente-Manzanares and Sánchez-Madrid 2004). Activation of FcRs on phagocytic cells initiates a signaling pathway that activates SRC family kinases (SFKs) and spleen tyrosine kinase (SYK) and stimulates Rac, which promotes actin polymerization by Arp2/3 complex and stimulates the assembly of actin-rich “phagocytic cup” to form pseudopods that engulf the target, closing around in a zipper-like mechanism (Crowley et al. 1997). Cdc42 is also recruited to the phagocytic cups, which activates WASP, promoting actin polymerization leading to membrane protrusion and particle engulfing (Dart et al. 2012). The severing of actin filaments plays a vital role in inducing actin polymerization due to the increase in the concentration of barbed filament ends. This is where the addition of actin monomers takes place, with cGSN the major actin-severing protein in phagocytes localizing in phagosomes (Serrander et al. 2000), and playing a key role in FcR and integrin-mediated phagocytosis (Li et al. 2012). Macrophages may use different modes of engulfment in response to spatial constraints, either lamellipodial phagocytosis facilitated by the formation of lamellipodia by Arp2/3, or filopodial phagocytosis that involves the extension of filopodia by Dia and/or Ena (Davidson and Wood 2020). Cytoskeletal reorganization is also a key process in the formation of pathogen-capturing neutrophil extracellular traps (NETs) released by neutrophils within the wound bed (Thiam et al. 2020b). NETosis, the formation of these web-like DNA structures of decondensed chromatin coated with histones and other proteins including cytoskeletal proteins (Vorobjeva and Pinegin 2014; Pires et al. 2016), is initiated with rapid actin depolymerization, followed closely by remodeling of microtubule and vimentin cytoskeletons to facilitate the expulsion of DNA, with NET release blocked by inhibition of actin disassembly (Thiam et al. 2020a). NETosis has been shown to impair wound healing and its inhibition results in healing improvement (Wong et al. 2015).

The release of actin from damaged cells is a common feature in all types of injury. The conditions of ionic strength and composition, temperature, as well as the pH in the extracellular environment (such as blood plasma) favors the polymerized form of actin (Li et al. 2012). The presence of F-actin, released from dying cells into the bloodstream, may be fatal as it can cause microthrombi in the vasculature of both the surrounding tissue, and pulmonary system (Haddad et al. 1990) as well as contributing to inflammation. To combat this, the extracellular actin scavenging system (EASS) comprising two key proteins, pGSN, together with vitamin D–binding protein (DBP), has evolved to rapidly clear actin from the circulation (Li et al. 2012). Working together, pGSN scavenges and severs actin filaments released in the extracellular space following injury into G-actin (Lind et al. 1986), which allows DBP to rapidly form DBP–actin complexes and clear extracellular actin via the liver (Dahl et al. 2003; Delanghe et al. 2015; Kew 2019). Serious injuries such as burns and traumatic injuries results in large amounts of extracellular actin that consumes both DBP and pGSN, reducing their concentration in plasma (Grinnell et al. 1993; Dahl et al. 2003). Reduced pGSN levels in patient blood has also been used to predict high-risk morbidities including sepsis following burn injury (Mounzer et al. 1999). Thus, both within the cell, and in the surrounding extracellular space, actin and its associated binding proteins are critical to the early stages of wound repair.

CYTOSKELETAL REMODELING DURING THE PROLIFERATIVE PHASE

The generation of new tissue occurs during the proliferative phase of healing and includes reepithelialization of the wound and formation of new granulation tissue formation, which includes ECM deposition and angiogenesis, but also fibroblast proliferation/migration, lymphangiogenesis, nerve sprouting, etc.

Formation of Granulation Tissue and Reepithelialization

In response to growth factors released by residential fibroblasts, platelets, monocytes, and keratinocytes, fibroblasts from surrounding tissues proliferate and migrate into the wound bed where they produce and deposit new ECM (Bainbridge 2013; Gonzalez et al. 2016; desJardins-Park et al. 2018). Reepithelialization usually starts even before the migration of fibroblasts and occurs through keratinocytes proliferating and migrating from around the wound edges along the injured dermis and across the newly formed granulation tissue until they meet and fuse (Jacinto et al. 2001; Abreu-Blanco et al. 2012). The actin cytoskeleton plays an integral role in this part of the proliferative phase, which involves multicellular adhesion, migration, and proliferation.

Cell Adhesion and Spreading

Cell adhesion and spreading requires the formation of Rho-dependent stress fibers and focal adhesions that instigate sustained extracellular signal‐regulated kinase (ERK) signaling to induce cyclin D1 production (Welsh et al. 2001). Proteins that link cells to the ECM, or neighboring cells, such as integrins and cadherins, are subject to and act in response to forces found in 3D matrices. These scaffolds hold cells in space and use mechanical sensing through focal adhesions to regulate proliferation (Pruitt and Der 2001; Provenzano and Keely 2011), and it is now clear that integrins are both regulated by and control the activation of GTPases including Rho, Rac, and cdc42 to affect not only adhesion and migration but also proliferation (Schwartz and Shattil 2000; Schwartz and Assoian 2001; Hou et al. 2016).

Proliferation

Once proliferation is activated, cytoskeletal remodeling plays a key role in the generation of the two daughter cells and is reliant upon the coordinated assembly and action of the actomyosin contractile ring (Cheffings et al. 2016). The activation of Rho in turn activates nucleation of F-actin by formin resulting in the formation of long, unbranched actin filaments and ROCK promotes myosin II activation to stimulate ring formation (Matsumura 2005; Goode and Eck 2007). The contractile stress required for cleavage is driven by myosin motors sliding actin filaments to constrict the ring and form a membrane furrow (Glotzer 2009). Cortical actin flow may also contribute to contractile-ring assembly and the stabilization of the actin network through cross-linking; however, this process must remain reversible and dynamic to facilitate not only the generation and positioning of the contractile ring, but also its rapid disassembly during constriction to allow separation of the daughter cells (Salbreux et al. 2009; Gerien and Wu 2018). The partition required to divide the two new daughter cells relies on the formation of a new membrane, with exocytosis relying upon vesicle transport to the membrane surface predominantly along the microtubule and actin (Noordstra and Akhmanova 2017; Papadopulos 2017). Shorter, branched F-actin accumulates at the poles of the cells with bundles of microtubules forming the central spindle that regulates the formation of the cleavage furrow to complete cytokinesis (Glotzer 2009). Distinct molecular signatures differentiate epidermal stem cells that provide the proliferative pool of newly formed cells further back from the wound edge, while those at the leading edge are altered to assume a nonproliferative, migratory phenotype (Aragona et al. 2017).

Migration

The migration of fibroblasts and keratinocytes involve the formation of lamellipodia and filopodia at the leading edge of the cells (Fig. 4). Lamellipodia are broad, sheet-like projections and are essential for cellular migration (Millard and Martin 2008). Filopodia are highly dynamic thin structures, rich in long parallel actin and able to sense the substratum ahead of the leading edge. Intensive actin reorganization is required to create these leading-edge protrusions, which are regulated by a range of ABPs and the Rho GTPase family. Rac and Cdc42 are abundant at the leading edge, facilitating actin cycling, filopodia, and lamellipodia formation (Ridley 2015). Arp2/3/WASP complex supports the nucleation of new filaments extended from preexisting filaments and the creation of branched filament network (Egile et al. 2005; Rodnick-Smith et al. 2016). Actin filament elongation is modulated by actin depolymerization through capping and severing proteins, mainly, ADF/cofilin and gelsolin family members (Mazur et al. 2010; Gross 2013). Tropomyosins are actin-associating proteins that stabilize actin filaments and have an important role in temporal regulation of cell migration during wound healing (Phillips et al. 1979; Lees et al. 2013). Cross-linking proteins like actinins have been found in the lamellipodia region, increasing the stability of actin filaments (Hamill et al. 2013). On the other pole, Rho at the trailing edge regulates ROCK, which in turn promotes stress fiber formation by inhibiting actin filament depolymerization and phosphorylating the MLC leading to actin contraction and rear retraction (Huveneers and Danen 2009).

Figure 4.

Schematic overview of cytoskeleton involvement in cell migration. Cells acquire morphological polarity due to external signals. At the leading edge, actin assembly leads to the creation of membrane protrusions including lamellipodia (mainly actin networks) and filopodia (mainly actin bundles). Cdc42 is the main GTPase contributing to filopodium extension, while Rac induces lamellipodium extension. Actomyosin contraction at the trailing edge mediates cell detachment at the rear of the cell and drives the cell forward. These contractions at the trailing edge are mediated by Rho. (ECM) Extracellular matrix.

When Rac1 is activated by the guanine nucleotide exchange factor (GEF) P-Rex-1 at the leading edge, thin membrane protrusions rich in polymerized actin form and cells assume an elongated migratory phenotype. In contrast, Tiam1 activation of Rac1 inhibits migration with increased membrane ruffling and localization of actin at cell–cell contacts (Marei et al. 2016). The ABP Flii preferentially binds active Rac1 and its activator P-Rex1, mediating cell contraction and migration through the phosphorylation of MLC and activation of myosin II (Marei et al. 2016). Flii negatively regulates migration through the promotion of stress fiber formation and impaired focal adhesion turnover (Kopecki et al. 2011; Strudwick and Cowin 2020). It increases the number and size of focal adhesions forming around the periphery of the cells and increases total F-actin levels (Kopecki et al. 2011). When Flii levels are high, impaired spreading and a reduction of filopodial processes is observed (Kopecki et al. 2009; Arora et al. 2015), with increased stable focal complexes and impaired turnover due to inhibition of paxillin phosphorylation, increased α-actinin expression, and suppression of the Src and p130Cas signaling pathways (Kopecki et al. 2009, 2011). Conversely, low levels of Flii reduce formation of focal adhesions containing vinculin and activated β1 integrins are reduced, with greater incorporation of G-actin into nascent filaments at focal adhesions (Mohammad et al. 2012). As wound healing progresses and wound size is reduced, the frequency with which contralateral protrusions make contact with each other increase, and filopodia tether together in neighboring cells to mechanically close the wound in a zipper-like manner at the leading edge, without the use of an actin cable (Wood et al. 2002; Abreu-Blanco et al. 2012). Unlike adult wounds in which keratinocytes migrate over the granulation tissue using lamellipodia to close the gap, embryonic wounds are closed by actin contraction in a contractile purse string (Martin and Lewis 1992; Cowin et al. 2003).

Angiogenesis

Concurrent with keratinocyte and fibroblast migration into the wound bed, ECs activated by hypoxia and proangiogenic factors invade into the newly deposited ECM and form tube-like structures that mature to create blood vessels (Pugh and Ratcliffe 2003). Angiogenesis relies on EC migration and proliferation. Upon wounding, ECs embrace two distinct phenotypes classified as stalk and tip cells (Potente et al. 2011). Tip cells migrate into the wound bed and extend sprouts toward chemoattractants (mainly VEGF and FGF-2), which extend outward as gradients from existing vessels (Norton and Popel 2016). The filopodia of these tip cells sense the chemoattractants and cellular extension occurs by the formation of lamellipodia at the leading edge. These protrusions attach to the ECM through focal adhesions and stress fiber–mediated contraction, allowing forward movement while MLC phosphorylation at the rear and stress fiber–mediated traction forces allow rear retraction (Lamalice et al. 2007; Chen et al. 2019). Following tip cell migration, stalk cells, with fewer protrusions, proliferate and elongate to form a vascular lumen and support the elongation of new sprouts (Sauteur et al. 2014). ECs establish cell–cell junctions (mainly VE-cadherin) with their neighboring cells to guide and extend the sprout (Szymborska and Gerhardt 2018). These cells attach to the ECM with focal adhesion. Adhesive complexes of ECs including adherens and focal adhesions interact with the intracellular F-actin network. Proteins in actin-tethered junctional complexes regulate the actin cytoskeleton remodeling required for the migration, proliferation, and elongation of ECs (Zankov and Ogita 2015). VE-cadherin's indirect association with the intracellular F-actin network has been shown to promote F-actin polymerization and ECs elongation. (Sauteur et al. 2014). When filopodia of tip cells contact each other sprouts fuse by anastomosis to form new vascular loops. Once the contact is established, VE-cadherin-containing junctions consolidate the connection. Cell–cell adhesion suppresses migration and sprouting by increasing actomyosin contractility at cell junctions (Abraham et al. 2009). Eventually, the new vessel is stabilized by recruitment of pericytes and deposition of ECM that continues into the remodeling phase of wound healing (Szymborska and Gerhardt 2018).

CYTOSKELETAL CONTROL OF TISSUE REMODELING

Toward the end of the proliferative phase of wound repair, fibroblasts that have migrated into the provisional matrix generate traction forces that facilitate collagen reorganization (Harris et al. 1981; Gov 2009). This force is generated by myosin II heavy chain activity and is regulated by MLCK and ROCK (Tomasek et al. 2002; Beningo et al. 2006). The fibroblasts develop contractile microfilament bundles (stress fibers) under mechanical stress, and become proto-myofibroblasts (Tomasek et al. 2002). The increase in the mechanical tension of ECM and the presence of TGF-β lead to the creation of differentiated myofibroblasts characterized by the expression of α-smooth muscle actin (αSMA), stress fibers, and large focal adhesions (Hinz 2016). Initially, myofibroblasts secrete matrix metalloproteinases (MMPs) that dissolve preexisting ECM and deposit new ECM proteins, mainly collagen I and III, glycoproteins, and proteoglycans (Zwetsloot et al. 2012). The myofibroblasts also synthesize contractile proteins including myosin and αSMA, which orientate themselves into horizontal patterns within the dermis and facilitate wound contraction and remodeling of the new dermal tissue (Fig. 5; Baum and Duffy 2011).

Figure 5.

The stages of myofibroblast differentiation. Fibroblasts in unwounded skin do not show any stress fibers or focal adhesion but may contain cortical actin. Due to the mechanical tension in the wound bed, fibroblasts first differentiate into proto-myofibroblasts, which contain stress fibers and focal adhesion complexes but generate low contractile force. The presence of TGF-β1 and increased mechanical tension result in further differentiation of proto-fibroblasts to differentiated myofibroblasts. Differentiated myofibroblasts are characterized by the expression of α-smooth muscle actin (αSMA), stress fibers, and super focal adhesions. Differentiated myofibroblasts generate greater contractile force than proto-myofibroblasts, which is reflected by a higher organization of collagen fibers and wound contraction. (ECM) Extracellular matrix.

Myofibroblast stress fibers are composed of γ- and β-cytoplasmic actin and nonmuscle myosin (myosin II). In addition, αSMA is strongly expressed in myofibroblasts as a result of TGF-β signaling (Zent and Guo 2018), leading to contraction and the maturation of super focal adhesions (Hinz et al. 2003; Goffin et al. 2006), which support the transmission of myofibroblast contractile forces to the ECM (Darby et al. 1990; Hinz et al. 2001; Hinz and Gabbiani 2003). In turn, ECM-to-cell mechanical transduction through focal adhesions controls the assembly of stress fibers with α-SMA (Goffin et al. 2006).

As the cell-rich and loose ECM is gradually substituted by ECM with high collagen density (Xue and Jackson 2015), the contraction of the myofibroblast population in granulation tissue formation leads to the collagen-matrix remodeling that results in tissue contracture (Tomasek et al. 2002). This contraction force relies upon myosin II–based contraction along stress fibers, a process that is regulated by Rho signaling, MLCK, and MLC phosphatase (Parizi et al. 2000). Transmission of the force to the surrounding ECM occurs through specialized focal adhesions containing transmembrane integrins (Larsen et al. 2006).

Scar Formation

In a normal healing process, once contracture occurs and tissue heals, myofibroblasts are either deactivated or eliminated from ECM by apoptosis (Wilson et al. 2007; Talele et al. 2015). Prolonged myofibroblast activities can result in accumulation and contraction of collagenous ECM leading to excessive scar formation (Van De Water et al. 2013). Nevertheless, myofibroblasts provide the major contraction force that promotes wound contraction and ECM reorganization. One of the most significant differences between normal tissue and scar tissue is the orientation of the ECM fibrils, which are mainly composed of collagens. Changes in stress of the ECM during remodeling leads to different types of contraction by myofibroblasts (Van De Water et al. 2013). Strong isometric contraction mediated by Rho-ROCK develops slackness in the fibrous collagens, which releases the tension of these fibers. Collagens are subsequently straightened, being pulled in by the myofibroblasts with micro contractions to restore tension, albeit on shorter fibrils (Castella et al. 2010). Random basketweaves of collagen fibers that occur in normal skin, are replaced by large and single-direction collagen fibers running parallel in the skin resulting in the formation of scar tissue (Ehrlich and Krummel 1996; Wolfram et al. 2009). Contractile forces initiated by myofibroblasts are transmitted to the ECM through integrins and develop a strained and more compacted ECM (Larsen et al. 2006). Myofibroblasts also mediate collagen fibrillogenesis by secreting decorin, which integrates into the collagen fibrils assuring uniform spatial fibril arrangement (Zhang et al. 2007). This impact of myofibroblasts results in less flexible dermis, a characteristic of scar formation.

Apoptosis

In the final stage of the remodeling process, macrophages, myofibroblasts, and ECs are eliminated by apoptosis resulting in tissue comprised of a small number of cells and a large amount of ECM proteins, mainly collagen fibers (Velnar et al. 2009; Gonzalez et al. 2016). The actin cytoskeleton is an initiator and mediator of apoptosis with several studies demonstrating its role in triggering apoptosis upstream of caspases through extrinsic and intrinsic pathways (Desouza et al. 2012) and by acting as a substrate for caspase cleavage (Mashima et al. 1999). The destruction of the actin cytoskeleton, through a reduction in actin turnover, induces cell death via apoptotic pathways (Gourlay et al. 2004; Gourlay and Ayscough 2005b), and the depolymerization of actin with cytochalasin D results in the inhibition of Fas-mediated apoptotic signaling (Parlato et al. 2000). The phosphorylation of MLCII with Rho GTPase leads to the contraction of the cortical actin ring and the formation of apoptotic blebs, which is followed by chromatin condensation and nuclear fragmentation (Coleman and Olson 2002). Like bleb formation, apoptotic nuclear disintegration requires ROCK activity and actin-myosin contractile force. At the final stage of apoptosis, the actin cytoskeleton depolymerizes, and the apoptotic bodies are cleared by phagocytosis (Ndozangue-Touriguine et al. 2008). The ABP gelsolin has been shown to inhibit apoptosis by preventing caspase activation and enhancing actin depolymerization (Gourlay and Ayscough 2005a).

TARGETING THE ACTIN CYTOSKELETON TO IMPROVE WOUND HEALING

In vitro and in vivo studies using knockout and overexpression mouse models have provided genetic evidence and biological perspectives regarding the manipulation of actin cytoskeletal proteins emphasizing their vital role in wound healing. A number of these studies is listed in Table 1, demonstrating the mode of manipulation and its outcome on wound healing both in vivo and in vitro. This has focused attention upon identifying cytoskeletal targets that may be manipulated to improve wound outcomes. Following these initial investigations, a number of synthetic peptides, small molecule inhibitors, and neutralizing antibody therapies have been developed for preclinical testing. Some examples are provided in Table 1.

Table 1.

Selected in vivo, ex vivo, and in vitro studies investigating actin cytoskeletal protein effects upon wound-healing processes

Cytoskeleton protein	Study type	Mode of manipulation	Outcome	References
Gelsolin	In vivo	Generation of GSN-null mice	Defects in hemostasis and platelet activation, inflammatory response and leukocyte motility, and defects in fibroblast function	Witke et al. 1995
Gelsolin	In vitro	Addition of recombinant gelsolin (rGSN) to mouse fibroblast GSN siRNA knockdown in human corneal fibroblast	Enhanced migration and increased antioxidant effect, decreased SMA synthesis	Wittmann et al. 2018; Vaid et al. 2020
Gelsolin	Ex vivo	Addition of rGSN to mice cornea defect model	Elevated cell proliferation and increased wound closure rate	Wittmann et al. 2018
Flightless I	In vivo	Generation of Flii heterozygous and overexpressing mice	Improved wound healing, decreased scarring, decreased inflammation, and improved angiogenesis in heterozygous mice	Adams et al. 2009; Cameron et al. 2014; Chan et al. 2014; Thomas et al. 2020
Flightless I	In vitro	Flii siRNA knockdown in human and mouse fibroblast and epithelial cells	Increased migration, increased proliferation, and collagen deposition	Cowin et al. 2007; Mohammad et al. 2012; Jackson et al. 2020
Tropomyosin	In vivo	Deletion of tropomyosin gene in mice	Elevated cell migration and faster healing response	Lees et al. 2013
N-WASP	In vivo	Conditional knockout of N-WASP in mice	Epidermal hyperproliferation in aged mice, increased fibroblast and collagen deposition, and enhanced wound closure	Jain et al. 2016
Arp 2/3	In vitro	Development of Arp2/3-depleted mouse embryonic fibroblasts	Disappearance of lamellipodia and defects in migration	Wu et al. 2012
Filamin	In vitro	Filamin knockdown of filamin by short hairpin RNA in mouse fibroblasts	Deformation of collagen matrix, and defect in wound contraction	Gurtner and Wong 2015
Filamin	In vivo	Conditional knockout of filamin in mice	Delayed wound closure and deformation of collagen matrix	Gurtner and Wong 2015
mDia 1	In vitro	Inhibition of cellular mDia1 activity, mDia1 siRNA knockdown in mouse and human epithelial cells	Dissociation of leader cells in migration, reduced F-actin level, and impaired wound repair	Rao and Zaidel-Bar 2016
Thymosin b4	In vivo	Addition of Thymosin b4 to db/db diabetic mice and aged mice	Improved dermal wound repair	Philp et al. 2003a

Local delivery of ROCK inhibitor Y-27632 to murine excisional wounds delayed healing with reduced formation of αSMA stress fibers but minimal effect on local inflammation or reepithelialization (Tholpady et al. 2014). Conversely, in another study, healing was significantly faster after topical treatment with Y-27632, with a keratinocyte-specific action upon growth and migration (Gandham et al. 2013). The delivery method and dose of ROCK inhibitor was, however, different between the two studies, soaked sponges (20 µmol/L Y-27632) versus daily injection (100 µL of 10−4 M Y-27632) (Tholpady et al. 2014; Turner et al. 2016), suggesting that further investigation of ROCK inhibition in wound healing is warranted. In 14-3-3ζ^−/− mice, unrestrained ROCK signaling at wound margins elevated ECM production and reduced ECM remodeling, increasing dermal stiffness and causing rapid wound healing. Inhibition of 14-3-3ζ using a novel pharmacological agent, RB-11, accelerated wound healing twofold, and as patients with chronic nonhealing wounds overexpress 14-3-3ζ, this may be a potential approach for improving wound repair (Kular et al. 2015). The therapeutic infusion of recombinant gelsolin (rGSN) has been investigated for the treatment of large area burns (Rothenbach et al. 2004). Rats with significant burn injuries (40% total body surface area) were treated with rGSN immediately before and after injury, resulting in the decreased permeability of pulmonary microvasculature. Similarly, in a model of acute respiratory distress syndrome that commonly occurs following traumatic injury, treatment with rGSN prevented infiltration of neutrophils in bronchoalveolar fluid, partially preventing the exudative response to injury (Christofidou-Solomidou et al. 2002). Gelsolin family member Flii has also been the target of therapeutic intervention, using siRNA and neutralizing antibodies to prevent its function in mouse and pig models of wound healing (Jackson et al. 2012; Kopecki et al. 2013; Turner et al. 2016, 2017). Injection of porous silicon nanoparticles (pSiNPs) containing Flii siRNA to murine wounds leads to improved healing outcomes (Turner et al. 2016). Similarly, topical application of Flii-neutralizing antibodies, either by intradermal injection to the wound margins or via delivery in a cream formulation, reduced inflammation, lowered the numbers of αSMA-positive myofibroblasts, and reduced scar formation (Jackson et al. 2012; Kopecki et al. 2013; Ruzehaji et al. 2014; Haidari et al. 2017; Turner et al. 2017). The topical treatment of wounds with ABP, thymosin β4, accelerated healing in both diabetic and aged mice, with increased wound contracture and collagen deposition (Philp et al. 2003b) being observed with the pro-healing functions of thymosin β4 confirmed using a seven amino acid synthetic peptide of the actin-binding domain (Philp et al. 2003b).

CONCLUDING REMARKS

While the healing of wounded skin is in itself a gross physiological event, every aspect of this process is exquisitely controlled at a subcellular level by the actin cytoskeleton. Without a functional cytoskeleton, wounds fail to heal making the targeting of these proteins an important and potentially critical avenue for therapeutic development. Although much is known about the integral role these proteins play in cellular responses, targeting individual components may be insufficient to improve the gross and complex physiological responses that are required to treat chronic wounds, burns, or hypertrophic scarring. Further studies are still required to identify how our understanding of the cytoskeleton can be used to successfully improve healing outcomes for complex wounds.