Cutaneous squamous cell carcinoma (cSCC) accounts for 25% of cutaneous malignancies diagnosed in the Caucasian population. Surgical removal in combination with radio- and chemotherapy is an effective treatment; however, prognosis for patients suffering from aggressive cSCC is still relatively poor. Increasing prevalence coupled with high mortality and morbidity in aggressive metastatic forms of cSCC highlights the need for development of novel targeted therapeutics. Metastasis is a complex process requiring dramatic reorganization of the cell cytoskeleton. Recent studies have highlighted the importance of mechanical forces and actin dynamics in cancer cells’ intrinsic ability to invade adjacent tissues, intravasate into vasculature, and ultimately metastasize. Tight regulation of the biochemical and mechanical properties of the actin cytoskeleton drives cellular processes involved in cSCC progression including polarity establishment, morphogenesis, and motility. Here we will provide a short introduction to disease pathogenesis, give an overview of the role of key regulatory proteins governing the mechanical forces and actin dynamics critical to cSCC progression, and describe the contribution of actin remodeling and actomyosin signaling to cSCC progression. We will also discuss how targeting protein regulating mechanical force and actin dynamics may have clinical utility in development of novel treatment modalities for patients suffering from aggressive cSCC.
- cutaneous squamous cell carcinoma
- actin cytoskeleton remodeling
- mechanical force
- systemic therapy
Cutaneous squamous cell carcinoma (cSCC) most commonly arises in actinically damaged skin and accounts for 25% of cutaneous malignancies diagnosed in the Caucasian population . The incidence of cSCC continues to rise annually, with an estimated 50–200% increase in incidence in the last three decades in USA alone, and is predicted to increase in future years due to an aging global population . Solar ultraviolet radiation is the primary environmental extrinsic cause of cSCC. Intrinsic immunosuppression, the second most common cause, leads to the formation of aggressive cSCC in organ transplant patients, patients on immunomodulatory therapies, and those suffering from recessive dystrophic epidermolysis bullosa, a genetic skin blistering disease [3, 4, 5]. The incidence of cSCC is higher in individuals who are fair-skinned and have a sun-sensitive phenotype; however, the aggressive forms of cSCC are more common in men and the elderly . Despite its prevalence, the relatively low fatality rate of cSCC means that its health and economic burden is often substantially underestimated , albeit latest data showing that in addition to significant morbidity cSCC accounts for up to 8000 deaths per year and costs approximately $4.8 billion annually in USA alone .
cSCC generally presents as a scaly, red or bleeding abnormal lesion on sun-exposed areas, and is associated with relatively benign outcomes and a low risk of metastasis. However, cSCC can demonstrate dramatic histopathological heterogeneity, resulting in a wide range of clinical outcomes . Histopathologic subtypes of cSCC are broadly divided into low-grade SCC (Figure 1A) that are well-differentiated but have low metastatic potential (keratoacanthomas, SCC in situ and verrucous carcinoma), or high-grade SCC (Figure 1B) that are poorly differentiated, have high potential of metastasis and recurrence, and are associated with a poor prognosis for patients (desmoplastic cSCC, adenosquamous cSCC and cSCC associated with non-healing ulcers or scarring processes arising from chronic wounds) [2, 7]. Once developed, the natural history of untreated SCC is one of local invasion followed by metastasis via the lymphatic system, blood or perineural invasion, which can lead to death . In most cases, cSCCs are detected early and can be successfully eradicated by surgical excision. However, if not detected and/or left untreated, disease progression to high-grade cSCC will often lead to mortality. Clinically, the most powerful predictor of disease pathogenesis is nodal metastasis and size, followed by invasion beyond fat, location, and lastly perineural invasion. These parameters are used in clinical staging systems for cSCC . Management of cSCC is primarily surgical, with adjuvant chemoradiation approaches based on risk factors, patient and tumor features, as well as care features including access to treatment and associated costs .
With increasingly longer life expectancy, the health and economic burden of cSCC is likely to continue to increase significantly. Hence, a better understanding of factors contributing to cSCC progression and metastasis is necessary to aid development of novel therapies, aimed at combatting cSCC in the community. Despite recent advances in gene expression screening technologies that have begun to identify candidate genes commonly mutated in patients with cSCC (including TP53, CDKN2A, Ras and NOTCH1), which may be responsible for regulating motility and invasion in cSCC, a comprehensive understanding of the factors contributing to cSCC invasion including mechanical tension and actin dynamics is still emerging . One thing that is clear is that patient outcome directly correlates with the degree of local and regional invasion, and coordinated regulation of the actin cytoskeleton is critical to cell motility, invasion and metastasis . Consequently, the signaling pathways involved in mediating chemotactic cues from the extracellular environment that regulate the actin cytoskeleton and mechanical forces, guiding cancer cell invasion and metastasis, have been and continue to be an area of intense study.
Recent studies have revealed a number of proteins and molecules that are aberrantly expressed in cSCC. These proteins link cell migratory signals to the actin cytoskeleton, thereby playing an instrumental role in the ability of cancer cells to resist chemotherapy and/or metastasize . In this chapter we will describe the actin dynamics and mechanical force governing tumor cell migration, invasion and metastasis. We will outline the main signaling pathways governing the formation of invasive protrusions by cancer cells with regard to the function of key regulatory proteins involved in actin cytoskeleton remodeling in cSCC.
The metastatic spread of aggressive cancers, including cSCC, is a highly selective process involving a series of sequential and orchestrated steps in the so-called “metastatic cascade”: detachment from the primary tumor site, cell migration and invasion of the surrounding extracellular matrix (ECM), intravasation into vasculature, extravasation at a secondary site, and interaction with the extracellular environment to form metastatic tumors . Each of these steps offers the potential for design of different therapeutic approaches to combat aggressive cSCC. Indeed, a number of recent studies have identified novel therapeutic approaches including both adjuvant and neoadjuvant treatments, with clinical trials utilizing epidermal growth factor receptor inhibitors and immune checkpoint blockers (nivolumab, pembrolizumab, and ipilimumab) showing promising early results as potential treatments of cSCC . Recent trials using cytotoxic chemotherapy have, however, shown limited advances for the treatment of cSCC, and trials investigating combined immune checkpoint inhibitor and radiation therapies, which may have synergistic effects in treatment of cSCC, are still pending . This highlights the need for increased research to close the gaps in our knowledge of cSCC biology, including better understanding of the factors that lead to aggressive cSCC, the role of microbiomes and HPV infection, the role that mechanical force and actin dynamics plays in this process, prediction of clinical response to therapies including immune checkpoint blockade, and how to tailor better prevention and treatment strategies to individual risk factors and needs . Emerging evidence on the crosstalk between different components of the cytoskeleton in metastatic progression combined with clinical data illustrating strong relationships between cytoskeletal alterations and metastasis in various cancers pinpoints important opportunities for potential therapeutic targets . Later in this chapter we will describe current research that has attempted to identify the steps of the metastatic cascade suitable and most amenable for therapeutic intervention, with a focus on harnessing our knowledge of actin cytoskeleton remodeling and mechanical forces to postulate therapeutic strategies targeting cytoskeletal and cytoskeletal-associated proteins critical in cSCC.
2. Cytoskeletal dynamics and regulation during cSCC progression
The skin is exposed to and responds to a wide range of mechanical signals throughout homeostasis and through to malignancy. Mechanical forces have been shown to regulate these normal cellular processes including stem cell renewal, lineage differentiation and proliferation, wound healing, as well as transformation through changes in the actin cytoskeleton—the ability to protrude, adhere to the ECM, migrate through tissue and invade into the underlying basement membrane.
The types of mechanical forces exerted upon the skin can vary depending on the context. In homeostasis, tensile/stretch forces and compressive forces arise as a result of muscle and joint movements, and physical location—skin stretched over bone is under significantly more stress than when over fat or muscle . Tensile forces cause cells to elongate and expand, and therefore are generated at sites of wounding as epithelial cells migrate in and contract to close the wound. This can generate scarring and fibrosis, which can lead to skin cancer including cSCC . Compressive forces generate different biomechanics in skin cells compared to tensile force . Compressive forces are able to activate Rho-ROCK signaling (described below) in the skin, which has been shown to play a role in tumor progression . In melanoma, it was found that stress-bearing areas of the foot were more conducive to cancer development due to increased mechanical compressive stress . Changes in substrate stiffness underlie these mechanical signals (Figure 2), and it has been shown that stiff stroma can lead to an activation of integrin signaling and subsequent cSCC development .
Cells have the ability to sense these changes in their environment (process referred to as “mechanosensing”). The mechanical signal is then converted to a biochemical signal in a process called mechanotransduction, and the biochemical signals initiate a cascade of changes within the cell at the transcriptional, translational and post-translational levels that result in a cell that can appropriately and reciprocally respond to the extracellular signals (process referred to as “mechanoreciprocity”) . In disease states including cSCC, the heightened and/or constitutive extracellular signals generate a detrimental loop of ever-increasing mechanoreciprocal signaling, hence leading to enhanced tumor progression, invasion and eventually metastasis .
Triggered by the changes in the cell microenvironment, the actin cytoskeleton undergoes a number of changes that allows a cell to become more motile and/or invasive. The ability of a cell to undergo directed migration is essential to its ability to metastasize, and is characterized by an ordered process (Figure 3) of membrane protrusion at the leading edge (filopodia and lamellipodia) and sides (invadopodia) of the cell, contact and adhesion between the protrusion and the matrix, movement of the main cell body, and retraction of the trailing edge [21, 22]. Lamellipodia are flat membrane protrusions containing dendritic arrays of actin filaments that branch out like a sheet from the leading edge of a cell. This particular form of protrusion is thought to have a major role in cell migration as their morphology allows the cell to make multiple contacts with the underlying substrate and pull the cell forward. Filopodia are narrow protrusions made up of bundled and cross-linked actin filaments that also stretch out from the leading edge. Invadopodia, unlike lamelli- and filopodia, are protrusions branching out from the sides of a cell, which have increased membrane remodeling and matrix degradation proteins . These particular membrane protrusions are often seen in cancer cells including during cSCC progression .
Adhesions between the cancer cells and matrix are necessary for invasion and are largely mediated by integrins (discussed in detail below) . During cSCC progression, there is an increase in cellular attachment to the ECM and, concurrently, a decrease in attachment to neighboring epithelial cells signified by a reduction in levels of E-cadherin expression. Following matrix adhesion, the trailing edge of the cell contracts, allowing the cell to move forward. Myosin II is required for this actin filament contraction, and is largely regulated by signaling through the small G-protein Rho. Once the cell contracts, its tail detaches as focal adhesion complex components are cleaved . Actin remodeling proteins are essential in these processes, and often regulate cell-cell and cell-stroma attachment and turnover of focal adhesions, allowing cell traction and movement to take place. The coordination of actin polymerization and contraction allows the cSCC cell the ability to migrate through dense, stiff ECM and stroma to metastasize to lymph node or surrounding organs.
A number of different pathways are activated downstream of mechanical signals, causing changes in the actin cytoskeleton. Signaling pathways involved in cytoskeletal dynamics during cSCC progression are also activated as a result of constitutive or heightened growth factor signaling . Together, the combination of mechanical and biochemical signals can trigger a multitude of intracellular signaling cascades that ultimately affect cell morphology. In this section, we will discuss the broad families of proteins regulating actin dynamics and mechanical forces in cSCC, an overview of which is illustrated in Figure 4. The overview provided in this chapter will not cover all the pathways or actin remodeling proteins involved but focus on those most relevant to cSCC progression, invasion and metastasis.
Integrins are well-characterized as the first point of contact for mechanical signal transduction. Inactive heterodimers on the cell surface, they are partially activated by intracellular proteins (“inside-out” signaling) before full activation upon binding to extracellular ligands (“outside-in” signaling) . Binding to the extracellular ligand results in full activation of the integrin receptor and leads to the formation of either an intracellular focal adhesion complex to link the ECM to the actin cytoskeleton, or hemidesmosomes, linking the ECM to intermediate filaments. The integrin that is activated is context-dependent in regards to the particular focal adhesion complexes that are formed, broadly encompassing a range of adaptor proteins including: talin, vinculin, paxillin, Flightless I (Flii), focal adhesion kinase (FAK) and integrin-linked kinase (ILK). FAK is rapidly recruited to focal adhesions upon integrin activation and is auto-phosphorylated, driving downstream signaling. FAK phosphorylation, stimulated by integrin activation, allows binding of Src-family kinases that are then able to trans-phosphorylate FAK. This leads to activation of ERK/MAPK signaling, and this complex is therefore able to control cell shape and regulate focal adhesions . In cSCC, a step-wise increase in activation of FAK from unaffected margin skin to hyperproliferative skin and invasive cSCC  results in elevated integrin-FAK-Src signaling that stimulates keratinocyte migration  and drives progression of benign papillomas to aggressive cSCC [29, 30]. In cSCC, it has been demonstrated that FAK function is required for cancer stem cell maintenance, regulating cSCC initiation, growth, regression, and progression . Actin remodeling protein Flii, which will be discussed in detail in Section 2.4, has been shown not to directly bind integrin receptors but form focal adhesion complexes with adaptor proteins and regulate integrin activation, downstream Src/paxillin signaling and focal adhesion turnover in a Rac1 dependent manner [25, 32]. Additionally, latest research has shown that actin remodeling proteins Flii and gelsolin, which have always been thought to be intracellular, have also both been shown to be secreted where they function to sequester extracellular actin post tissue injury, modulate inflammation and affect collagen VII anchoring fibril formation. Flii is able to modulate inflammation via toll-like receptor 4 (TLR-4) signaling, as Flii leucine-rich repeat (LRR) domains have 50% similarity to LRR domains of TLR-4, by which the immune system is able to detect infection or injury. The binding of LRRs to PAMP and DAMP molecules activates intracellular TLR signaling and ultimately results in the release of proinflammatory cytokine secretion . The extracellular roles of Flii and gelsolin in respect to mechanosensing and cSCC progression are still to be examined. Nevertheless, it is clear that the coordinated activation of integrin receptors, focal adhesion complex formation and downstream signaling stimulation is essential for cytoskeletal changes that are necessary for cell migratory and invasive capability during cSCC progression.
2.2Mechanotransduction: Rho GTPases
Downstream of integrin activation are the Rho small GTPases, which are part of the Ras superfamily and key regulators of cell cytoskeletal dynamics through both actin polymerization and organization, hence driving cancer cell motility . Of this subfamily, RhoA, Rac1 and Cdc42 are the best-characterized.
RhoA is involved in actomyosin contractility, formation of actin stress fibers and assembly of focal adhesion complexes. The main regulator of cytoskeletal dynamics leading to formation of stress fibers is myosin II, and its regulatory subunit myosin regulatory light chain-2 (MLC2) can be activated by RhoA signaling leading to contraction of actin fibers. Rho-associated kinases ROCK1 and ROCK2 are serine/threonine kinases that contain a Rho-binding domain and are activated by RhoA in its active GTP-bound form, directly activate MLC2 via phosphorylation. Due to its roles in cell contractility and movement, Rho-ROCK signaling has been implicated as a driver for invasiveness during cSCC progression but also plays a positive role in physiological normal wound healing processes . In human cSCCs, ROCK is not only highly expressed but also activated in the hyperproliferative skin and invasive regions of the tumor, as shown by phosphorylation of the ROCK substrate myosin phosphatase (MYPT1) . In the skin, this ROCK-mediated actomyosin contractility is required for proliferation of the epidermis, as ROCK activation stabilizes β-catenin through phosphoinositide 3-kinase (PI3K), Akt, and inhibition of its phosphorylating kinase GSK3β . During cancer progression from normal skin through to hyperproliferative and invasive cSCC, nuclear localization of active β-catenin and inactivation of GSK3β is increased, accompanied also by a progressive increase in FAK activation [27, 36]. It has further been demonstrated that a negative regulator of ROCK signaling, 14-3-3ζ, is significantly down-regulated in human cSCCs. As genetic deletion of 14-3-3ζ results in significantly larger papillomas in the two-stage chemical carcinogenesis (DMBA-TPA) mouse model of SCC, this suggests that uncontrolled ROCK signaling can drive cSCC tumor growth .
Cdc42 is involved in formation of F-actin microspikes and filopodia in both normal and cSCC cells, by via actin polymerization at the leading edge and at the sides of the cells, contributing to cSCC invasion. Traf6 has been demonstrated to regulate Cdc42 to induce these F-actin microspikes in SCC cells . When Cdc42 is absent in keratinocytes, cells are no longer able to properly process and deposit ECM components or integrin receptors, hence halting cellular migration . Taken together, these studies demonstrate roles for Cdc42 in both skin cell migration and invasion, necessary cellular processes for progression of cSCC. Cdc42 is necessary for proper cellular polarity in normal and migratory cells , and this in turn activates G-proteins of the Rac subfamily of Rho small GTPases .
Rac1 is hyperactivated in cSCC via integrins including α3β1, and is important for keratinocyte cell proliferation [40, 41]. Rac1 stimulates polymerization of actin via multiple kinase signaling cascades including that of MAP kinase (elevating the transcription factors AP-1, NFκB and CRE). This therefore allows the cell to form a branched actin network, necessary for leading-edge lamellipodia formation and membrane ruffling . Indeed, actin remodeling protein, Flightless I (Flii), has been shown to regulate focal adhesion by inhibition of paxillin phosphorylation via a Rac1 dependent pathway . Rac1 can be activated by Tiam1, and it was shown in the two-stage chemical carcinogenesis cSCC model that genetic deletion of Tiam1 significantly reduced tumor incidence, burden and growth. However, SCC tumors that did arise in Tiam-null mice were significantly more invasive and malignant, potentially due to a loss of cell-matrix adhesion . This highlights the dual homeostatic and tumor-promoting roles that actomyosin regulatory pathways can play during cSCC progression.
Rho-associated kinase ROCK can also phosphorylate and activate LIM kinases 1 and 2, which are then able to phosphorylate and inactivate cofilin, an F-actin severing protein, resulting in F-actin filament stabilization . Accordingly, it has been shown that LIMK1 levels are increased in cSCC tumor tissue compared to normal skin, and that LIMK1 silencing can suppress cell growth and invasion in cSCC cell lines . In addition, it has been suggested that LIMK is required in the microenvironment in leading fibroblasts, to allow for efficient remodeling of the ECM and subsequent cSCC invasion . It has been demonstrated that cofilin phosphorylation can be abolished by treating cSCC cells with LIMK inhibitors. This reduces β-catenin accumulation and epidermal proliferation via reversing actomyosin contractility , however clinical trials using these inhibitors are still pending. The involvement of Rho GTPases in cellular migration and invasion in cSCC due to cytoskeletal rearrangement implicates this family of proteins as drivers of cancer initiation, progression and metastasis. Hence, targeting Rho pathway signaling, in particular that of RhoA-ROCK signaling, is an attractive therapeutic option that will be explored later in this chapter.
2.3Actin polymerization: WASP, cortactin, and Arp 2/3
Actin nucleation promoting molecules are activated downstream of Rho GTPases and growth factor receptors. Wiskott-Aldrich syndrome protein (WASP) family members, via Erk, paxillin, and Src signaling together with cortactin, act to stimulate the actin-related protein (Arp) 2/3 complex which in turn mediates actin polymerization .
The WASP family consists of WASP proteins and WASP-family verprolin-homologous (WAVE) proteins. WASP proteins interact with Rho GTPases in order to form cellular protrusions and allow the cell to migrate, for example, N-WASP is involved in formation of filopodia and invadopodia upon its activation by Cdc42. WASP proteins bind to G-actin and Arp2/3 resulting in their activation and hence triggering actin filament production. WASP and WAVE proteins also bind profilin, which transports actin monomers onto the growing ends of actin filaments, and is therefore also an important factor in cell motility . As a loss of N-WASP in keratinocytes causes epidermal hyperplasia and a reduction in epithelial cell tight junctions , this highlights the need for proper control of these processes. The formation of cellular protrusions also relies on the Src protein cortactin, which binds to and activates the Arp2/3 complex independently of WASP, thereby regulating actin filament nucleation. Cortactin binds to F-actin, stabilizing actin filaments and allowing it to properly activate Arp2/3 [49, 50]. In head and neck SCC, overexpression of cortactin increases cancer cell proliferation and increases cell survival in anchorage-independent conditions , while in oral SCC, silencing of cortactin was shown to significantly impair invasiveness and downregulate the levels of epithelial markers, indicating an epithelial to mesenchymal transition (EMT) , a process by which epithelial cells lose their adhesion to one another and acquire a migratory and invasive mesenchymal phenotype (discussed in further detail below). Indeed, Arp2/3 complex proteins are required for cell proliferation and migration in other forms of SCC [53, 54], and based on the Arp2/3 role in actin filament polymerization, it is clear that Arp2/3 is also critical for actin cytoskeletal remodeling leading to cancer cell motility and invasion.
2.4Actin remodeling: tropomyosin, Flightless I, and podoplanin
The actin cytoskeleton is composed of three distinct elements including microfilaments, microtubules and intermediate filaments. Tight regulation of cytoskeletal elements must be coordinated, and latest research has shown that interplay between actin and microtubules is bidirectional . Actin-based motility is also dependent on the balanced activity of number of specific actin remodeling proteins. In this section we will highlight main actin remodeling proteins that have been shown to have specific functions in cSCC progression, including members of the tropomyosin family of actin-associated proteins, the gelsolin family of actin remodeling proteins, and Podoplanin, a simple glycoprotein with important roles in cSCC progression.
Members of the tropomyosin family of actin remodeling proteins display a tissue- and time-specific expression, while their association with actin filaments impairs isoform-specific regulation of actin filament dynamics . Tropomyosin proteins assemble as polymers in the major groove of the polymerized actin filament and their association drives actin filament turnover, hence playing an important part in a number of cellular functions including motility and metastasis . There are over 40 different isoforms of tropomyosin and few have been described as having an important role in cSCC progression. High expression of Tm5NM1, a specific cytoskeletal tropomyosin isoform, has been shown to inhibit cell migration and invasion as well as impair normal wound healing via its effects on Src activation, focal adhesion stabilization, increased actin filament tension, and paxillin phosphorylation . Current research is examining the effect of Tm5NM1 inhibitor TR100 on cSCC progression (see Section 3). On the other hand, downregulation of tropomyosin-1 and complete loss of β-tropomyosin has been identified in human esophageal SCC, while α-tropomyosin has been shown to be preferentially expressed in keratinocytes of the multistage model of murine cSCC, collectively suggesting isoform specific functions [59, 60, 61].
The dynamic remodeling of the actin cytoskeleton is also tightly regulated by the gelsolin family of actin remodeling proteins, which includes: gelsolin, villin, adseverin, capG, advillin, supervillin, and Flightless I (Flii) . These actin binding proteins function in the cytoplasm of the cells where they control actin organization by severing pre-existing filaments, capping the fast growing filament ends and bundling filaments, enabling filament reassembly into new cytoskeletal structures that are required for cell motility, invasion and metastasis . Studies have shown that downregulation of gelsolin proteins counteracts cancer cell invasion in vitro , however in cSCC, gelsolin and Flii have been the most studied to-date. Gelsolin over-expression has been shown to promote cell growth and motility in oral SCC [64, 65], while Flii, through its effects on apoptosis, has been linked to promotion of breast cancer progression and invasion and progression of cSCC [23, 66]. Flii is an important regulator of cell adhesion, migration and proliferation and a number of previous studies have described the role of Flii protein in wound healing and demonstrated the therapeutic effect of Flii neutralizing antibodies (FnAb) in acute and chronic wounds, skin blistering diseases and inflammatory skin conditions [25, 32, 67, 68, 69, 70, 71, 72]. Flii modulates cell adhesion and paxillin signaling, and regulates actin polymerization, tight junction formation and ECM production during wound repair suggesting that similar roles may govern Flii activity in cSCC progression [23, 25, 32, 72]. Indeed, altering Flii levels both genetically and using Flii neutralizing antibodies significantly augments cSCC progression . Therapeutic approaches targeting Flii in cSCC are described in Section 4.
The expression of Podoplanin, a small mucin-like protein, has also been linked to remodeling of the actin cytoskeleton in cSCC. Podoplanin is upregulated in the invasive front of a number of human carcinomas including cSCC and has been shown to induce collective cell migration by filopodia formation, via downregulating the function of Rho small GTPases [73, 74]. Podoplanin has also been linked to an increase in the migration of cancer-associated fibroblasts as well as endothelial network formation . Collectively these findings suggest that Podoplanin is able to induce an alternative pathway of tumor cell invasion in the absence of traditional epithelial-mesenchymal transition.
Taken together, these studies highlight the important role of actin remodeling in cSCC progression and outline the importance of bidirectional stimulation of actin remodeling by both intrinsic factors and the microenvironment, critical to tumor invasion/metastasis. These findings provide a rationale for development of novel therapeutic strategies that target tumor invasion and metastasis.
3. Physiological effects of actin remodeling
Changes in the actin cytoskeletal structure result in changes to cell morphology, creating a cell more conducive to invasion. One of the commonly recognized requirements of metastasis is a cellular transition from epithelial to mesenchymal phenotype (EMT). This transition is characterized by upregulation of genes including vimentin, SNAI1 (snail), SNAI2 (slug) and Zeb1, and a downregulation of epithelial genes including cadherins, as well as concurrent loss of cell-cell junctions . For example, in head and neck SCC an increase in matrix stiffness and hence increased mechanical signaling caused an increase in EMT markers in tumor-initiating cells . Cells undergoing EMT develop an elongated spindle-like morphology, due to the enhanced membrane protrusion formation . It has been shown in A431 cells, a human epidermal SCC cell line, that loss of T-cadherin induces elongation of cells and formation of lamellipodia and multiple leading edges via changes in EGF-stimulated motility and invasion, as T-cadherin influences EGFR localization and responsiveness . Of note, RhoA activation was also increased upon the loss of T-cadherin . Likewise, Podoplanin is also capable of transforming cells to an invasive state without having to undergo EMT, due to rearrangements of the actin cytoskeleton .
The remodeling of the actin cytoskeleton also creates intracellular reciprocal forces that balance out the forces received by the cell from the extracellular microenvironment. Actin polymerization extends the filament network, and as filaments in the leading edge are compressed between transient associations with the cell membrane and the bulk of the actin cytoskeletal network behind them, intracellular force is generated. As protrusions are extended and retracted, actin filaments experience tension from transient bonds with the membrane, becoming bent or compressed . Activation of the mechanotransduction pathways described above, downstream of ECM stiffness in cSCC, can also increase the propensity for augmented interactions with the stroma, and generate a tumor-promoting environment that enhances mechanoreciprocal signaling [20, 36].
4. Therapeutic approaches targeting actin cytoskeletal regulatory pathways
Metastasis is a complex process requiring significant reorganization of the actin cytoskeleton and coordinated involvement of number of key proteins. These proteins interact directly and indirectly with both actin and microtubule networks, hence significantly influencing migratory and metastatic cell phenotypes. Strong clinical relationships between actin cytoskeletal alterations and cutaneous cancer metastasis have been previously described , offering potential opportunities for therapeutic intervention. For example, up-regulation of cortactin, an actin-binding adaptor protein in melanoma, has been directly linked to increased distal metastasis and reduced disease-free survival, while up-regulation of Ras mRNA has been directly linked to Stage III and Stage IV disease in head and neck SCC [81, 82]. The complex nature of cellular migration and invasion presents challenges in developing therapeutic approaches, as compensatory pathways may overcome the effects of specific inhibitors. This highlights the need for development of combinational and adjuvant therapies targeting multiple pathways that are involved in actin dynamics to treat aggressive cSCC. Pharmacological inhibitors of actin have failed clinical development due to non-specific effects on normal actin function in tissue, resulting in high levels of cardiotoxicity. Hence, research efforts have centered on therapeutic approaches that can modify signaling pathways regulating the actomyosin cytoskeleton and/or target cytoskeletal and cytoskeletal-associated proteins .
Increasingly it has been recognized that microRNAs (miRNAs) and long non-coding RNAs (lncRNAs) are involved in regulating cytoskeletal dynamics through regulation of gene expression. lncRNAs have been shown to regulate lamellipodia formation by downregulating integrin expression in cSCC . A number of miRNAs have also been shown to play a role in regulating cell cytoskeletal dynamics and interactions with stroma in cSCC, including: miR-340 , miR-20a , miR-31  and miR-125b . These miRNAs act via inhibiting RhoA, LIMK1, WAVE3 and matrix metalloproteinase (MMP)-13 respectively. The use of miRNAs in the clinic has clear potential, however clinical trials are yet to be undertaken.
One of the signaling pathways participating in regulation of cancer cell motility, invasion and metastasis is the ROCK signaling pathway, described in detail above. Hyperactivation of this pathway promotes cancer cell invasion in many solid tumors and studies have shown that Rho signaling through ROCK promotes the rounded bleb-associated mode of amoeboid motility, thereby promoting tumor cell metastasis . Earlier reports have shown that treatment with the selective ROCK inhibitor Y-27632 increases SCC cell adhesion, upregulates expression of E-cadherin, and decreases the phosphorylation of cofilin (thereby activating it), resulting in altered actin cytoskeleton rearrangement . In the two-stage chemical carcinogenesis model, Y27632 treatment resulted in significantly smaller and fewer papillomas, with a reduced rate of cSCC conversion. This was associated with reduced collagen deposition in the ECM, which would indicate a decrease in mechanical signaling due to ECM stiffness . This illustrates that inhibition of ROCK is a potential strategy for treatment of solid cancers including cSCC.
A potential upstream regulator of ROCK-mediated cell migration is gamma-actin. Modulation of gamma-actin changes the directional cell migration via effects on microtubule dynamics and cell polarity, hence highlighting the crosstalk between actin cytoskeleton and microtubule signaling as a potential modality for targeting specific components of the network . More recent studies, using newer generations of ROCK inhibitors and pharmacological small molecule inhibitors of the downstream effectors of ROCK that have micro-tubule stabilizing effects, are also showing some promise in regulating tumor metastasis, however no compounds are yet clinically approved . Another potential therapeutic strategy is harnessing the ability of 14-3-3ζ, a negative regulator of ROCK signaling, to moderate mechanoreciprocity in cSCC . These approaches are particularly enticing due to the negative effects that clinical targeting of ROCK itself can potentially have .
Interestingly, studies investigating the interactions between SCC cells and cancer-associated fibroblasts have shown that ROCK activity is also an important requirement for adjacent stromal fibroblasts. ROCK activity positively influences the JAK1-STAT3 signaling pathway resulting in increased actomyosin contractility and proinflammatory cytokine secretion, favoring cSCC cancer cell invasion . Consequently, these studies suggest that approaches aimed at inhibiting ROCK signaling have the potential to interrupt both intrinsic and microenvironment-derived signals during cSCC progression.
Actin remodeling proteins have long been implicated in cSCC, as a dysregulated actin cytoskeleton and an aberrant tumor microenvironment is a hallmark of aggressive cSCC [11, 93]. One particular actin remodeling protein, Flightless I (Flii), has been identified as a tumor promoter with transcriptional activity in colorectal, breast and hepatocellular carcinoma cell lines . However, recent studies have also shown that Flii is significantly increased in human and mouse cSCC tissue samples, while secreted Flii is elevated in the sera of patients with cSCC and is increased in different cSCC cell lines established from human primary, recurring and metastatic cSCC as well as immortalized keratinocytes . Human cSCC samples show positive staining for Flii in invading keratinocytes, surrounding tumor stroma and the outer hyperkeratotic layer of cSCC nodules present in the deep dermis . Together, these data suggest that Flii is not only an important regulator of the actin cytoskeleton involved in cSCC progression but also a potential therapeutic target and diagnostic marker of cSCC severity. Indeed, overexpression of Flii resulted in severe cSCC development via evasion of apoptosis, while reducing Flii expression using intradermal injections of FnAb during cSCC initiation and progression significantly reduced Flii expression in both the tumor microenvironment and in the serum, and led to significantly smaller tumor size (Figure 5) and decreased cellular sphere formation and invasion in vitro .
Remodeling and polymerization of actin filaments is critical during cSCC invasion and formation of invadopodia. Increased Flii levels have shown to weaken cell-stroma and cell-cell adhesions via alteration of GTPase and Src/paxillin signaling pathway activity  and augmented integrin-facilitated cell migration . This promotes tumor progression and facilitates invadopodia formation and subsequent tumor invasion into surrounding tissue . Indeed, Flii is significantly increased in invading cSCC and has been demonstrated to associate with cortactin at leading edges of invadopodia and to regulate the invasive properties of cSCC keratinocytes . Systemic and topical therapeutic approaches using FnAb are currently in development with FnAb as a therapy for wound healing now entering the final pre-clinical validation stage [68, 94]. Flii has been shown to colocalize with structural (Claudin-1, -4 and -6) and adaptor (ZO-1 and -2) tight junction proteins and its overexpression in keratinocytes results in an altered F-actin/G-actin ratio, which can be restored using FnAb . Therefore, taken together, these studies suggest that therapies targeting Flii may be a potential strategy for reducing the severity of cSCC in the community, however clinical trials using FnAb are still pending.
Pharmacological inhibition of actin-associated proteins aimed at compromising the survival and invasion of tumor cells may also have clinical benefit. One example of this strategy is harnessing TR100 inhibition of the tropomyosin isoform Tm5NM1. Tm5NM1 belongs to a family of actin-associated proteins that regulate the activity of several effectors of actin filament dynamics , as described above. The TR100 inhibitor has been shown to preferentially disrupt the actin cytoskeleton of tumor cells, impairing tumor cell motility and viability, and reducing melanoma growth both in vitro and in vivo. This therefore provides a pathway for development of a novel class of anti-actin compounds for the potential treatment of wide variety of cancers including cSCC .
Microtubule targeting agents of both synthetic and natural design, and microtubule stabilizing and destabilizing agents, have been the focus of anti-cancer therapy in the last decade and remain one of the most successful group of agents in the clinic [97, 98]. Their ability to regulate the tubulin-microtubule equilibrium disrupts the mitotic spindle, halting the cell cycle and resulting in cell death. They have been shown to be effective in combination with anti-angiogenic and anti-vascular properties and in some cases have demonstrated the ability to overcome multi-drug resistance, supporting their utilization as a chemotherapy . Epothilones are a new class of anti-microtubule agents currently in clinical trials. Epothilones have shown activity in cSCC cell lines and in melanoma clinically, however clinical trials on cSCC patients are still pending [100, 101]. Other examples of microtubule-targeting agents, which have shown clinical promise in different subtypes of SCC including metastatic and recurrent disease, include semisynthetic compounds docetaxel and eribulin and a natural compound called rhizoxin [102, 103, 104]. While microtubule-targeting compounds are widely used as chemotherapeutic agents, they do have variability in different cancers, cancer cells frequently develop resistance to them, and they can be toxic to normal tissue, highlighting the need for better research and refinement of these compounds as well as a need to further understand their interactions with microtubule-associated proteins . It is possible that microtubule-targeting agents also exert broader effects on tumor cell migration, invasion and metastasis and future studies should explore their effects on cSCC in combination with actin pathway inhibitors. Gaining a better understanding on the interplay of regulatory proteins governing the mechanotransduction and actin cytoskeletal remodeling involved in tumor cell migration, invasion and metastasis will lead to increased efforts to exploit therapeutic avenues targeting the actin cytoskeleton to treat aggressive cSCC.
The contribution of actin cytoskeletal remodeling and actomyosin signaling during SCC progression is significant and cannot be undervalued in the search for new treatment modalities. Recent research has identified a number of potential novel therapeutic targets within regulatory actin and microtubule signaling pathways that should be explored as potential therapeutic adjuvants to immunomodulatory therapies currently in clinical trials. A comprehensive understanding of the regulatory network of cutaneous mechanotransduction, mechanical forces and actin dynamics in cSCC, as discussed in this chapter, will facilitate the development of novel approaches to curb the incidence and progression of aggressive cSCC in the community, generating new inroads toward development of novel, individually personalized and efficient therapeutic approaches.
Dr. Zlatko Kopecki is supported by a Future Industries Institute Foundation Fellowship from University of South Australia. Dr. Sarah Boyle is supported by a Royal Adelaide Hospital Research Fund Early Career Fellowship.
Conflict of interest
The authors declare no conflict of interest.
|Akt||protein kinase B—serine/threonine-specific protein kinase|
|AP-1||activator protein 1—transcription factor that regulates various cellular processes downstream of stimuli|
|Arp2/3||actin-related proteins 2/3 complex—protein complex regulating the actin cytoskeleton|
|Cdc42||cell division control protein 42—protein involved in regulation of cell cycle|
|CDKN2A||cyclin-dependent kinase inhibitor 2A|
|cSCC||cutaneous squamous cell carcinoma|
|DAMPs||damage-associated molecular pattern molecules|
|DMBA||7,12-dimethylbenz[a]-anthracene—polycyclic aromatic hydrocarbon that acts as a carcinogen|
|EGF||epidermal growth factor|
|EGFR||epidermal growth factor receptor|
|EMT||epithelial to mesenchymal transition|
|Erk||extracellular signal-regulated kinase—kinase involved in regulation of cellular processes such as meiosis and mitosis|
|FAK||focal adhesion kinase—a focal adhesion-associated protein kinase involved in cellular adhesion|
|FnAb||Flii neutralizing antibody|
|GSK3β||glycogen synthase kinase 3-beta—serine/threonine protein kinase|
|GTPase||enzyme that binds and hydrolyzes guanosine-5′-triphosphate|
|H&E||hematoxylin and eosin|
|lncRNA||long non-coding RNA|
|MAPK||mitogen-activated protein kinase|
|MLC2||myosin regulatory light chain-2—subunit of myosin, which regulates cell contractility|
|MMP||matrix metalloproteinase—endopeptidases that degrade various ECM proteins|
|MYPT1||myosin phosphatase target subunit 1—subunit of myosin phosphatase, which dephosphorylates MLC and thereby opposes contractility|
|NFκB||nuclear factor kappa-light-chain enhancer of activated B cells—protein complex that regulates transcription and other cellular processes|
|PAMPs||pathogen-associated molecular pattern molecules|
|PI3K||phosphoinositide 3-kinase—intracellular signal transducer enzymes ROCK: Rho-associated coiled-coil containing kinases|
|SNAI1||snail—transcription factor promoting repression of E-cadherin|
|SNAI2||slug—transcription factor promoting repression of E-cadherin|
|STAT||signal transducer and activator of transcription proteins—intracellular transcription factor|
|Tiam1||T-lymphoma invasion and metastasis-inducing protein 1—regulates Rho-like proteins and transduces extracellular signals|
|TP53||tumor protein p53|
|TPA||12-o-tetradecanoylphorbol-13-acetate—tumor promoting phorbol ester|
|TLR-4||toll-like receptor 4—transmembrane receptor that activates NFκB signaling|
|Tm5NM1||tropomyosin isoform 5NM1|
|Traf6||tumor necrosis factor (TNF) receptor associated factor 6|
|WASP||Wiskott-Aldrich syndrome protein|
|Zeb1||zinc-finger E-box-binding homeobox 1—transcription factor that induces EMT by repression of E-cadherin and other genes|