p53 regulation of podosome formation and cellular invasion in vascular smooth muscle cells

Abstract

The p53 transcription factor, discovered in 1979^1;2, is well known as a potent suppressor of tumor development by inhibiting cell cycle progression, and promoting senescence or apoptosis, when the genome is compromised or under oncogenic stress³. Accumulating evidence has pointed to an alternative role of p53 in the curtailment of tumor progression and colonization of secondary sites by negatively regulating tumor cell metastasis^4;5. Recently, we have found that p53 suppresses Src-induced formation of podosomes and associated invasive phenotypes in fibroblasts and vascular smooth muscle cells (VSMC)^6;7. In this review, I will focus on some recent studies that have identified p53 as a suppressor of cell migration and invasion in general, and VSMC podosome formation and ECM degradation in particular.

A Lesser Known Role of p53: Suppression of Cell Migration and Invasion

Is p53, often dubbed “guardian of the genome,” also a suppressor of cell migration and invasion? There were hints. One arose from the identification of p53 transcriptional targets that code for regulators of cell migration and invasion, such as smooth muscle α-actin,8 vascular endothelial growth factor (VEGF),9 MMP1,10 and caldesmon.11 Another came from the finding that p53 associates with cytoskeletal structures such as actin filaments,12 microtubules13 and intermediate filaments,14 albeit without any clear physiological significance. Valuable clues can also be gleaned from studying the effects of wild-type and p53 mutants on the invasion potential of tumor cells. Over 50% of cancers are associated with p53 mutations, many of which involve point mutations clustered at the DNA-binding domain of the protein, and the mutated proteins are expressed in high level in tumors. Mutations in p53 that result in a loss of its wild-type function in cell growth suppression often cause a gain of functions in cell migration and invasion. For example, expression of two p53 mutants (R175H and R273H) drives invasion in lung cancer cells via enhanced integrin and EGFR trafficking.15 Other p53 mutants (R248W, R280H and R273H) increase cancer cell invasion by inhibiting MDM2-mediated degradation of Slug, an invasion promoter that represses E-cadherin transcription; in contrast, wild-type p53 upregulates MDM2-mediated Slug degradation, and hence suppresses invasion.16

The mechanism of regulation has not been clearly defined, but there is convincing evidence that downregulation of cell migration by p53 is mediated by RhoGTPases, the key modulators of cytoskeletal remodeling. For instance, depletion of p53 leads to upregulation of RhoGTPases, Rac 1 in particular, and induction of cell migration and cell proliferation in MEF cells.17 Upregulation of wild-type p53 prevents Cdc42-mediated formation of filopodia,18 cell polarization and cell spreading19 in MEF cells, while p53-deficient MEF cells cultured in 3D matrix acquire an enhanced invasive amoeboid phenotype that requires RhoA-ROCK signaling.20

If p53 is engaged in the regulation of cell cycle and cytoskeleton remodeling, are the dual functions of p53 coordinated or independently executed? While a definitive answer must await further investigations, recent studies have suggested that JMY (junction-mediating and regulatory protein), a positive co-factor of p53 transcriptional functions, may act as a coordinator of p53 functions in the nucleus and the cytoskeleton.21,22 JMY is primarily nuclear in undifferentiated HL-60 cells assisting transcriptional functions of p53. However, in differentiated, highly mobile cells, JMY becomes exclusively cytosolic, co-localized with actin filaments at the cell cortex, where it nucleates actin filaments in an Arp2/3-dependent and -independent manner.

p53 and Atherosclerosis

What p53 does in vitro appears to be straightforward and consistent: it inhibits cell proliferation and promotes apoptosis, and suppresses cell invasion. Based on evidence from p53-knockout mice, its role in atherosclerosis has been surprising and often contradictory, although there appears to be a general consensus that global knockout of p53, for example, in p53^-/-/apoE^-/- mice, accelerates aortic atherosclerosis.23,24 While endogenous p53 appears to protect against apoptosis, increase in p53 promotes cell apoptosis in atherosclerotic plaques.25 It is understandable that interpretation of atherosclerotic phenotypes in global p53-null mice is complicated by the fact that the role of p53 varies significantly in different cell types, such as macrophages and VSMC, which make up the complex plaque tissue. With recent findings implicating an alternative role of p53 in VSMC invasion, it would be of interest to limit p53-deletion in VSMC using a smooth muscle cell-targeted, conditional p53-knockout in adult mice to investigate the role of p53 in VSMC migration and ECM invasion in the initiation of intimal thickening and plaque formation, and subsequent plaque stability.

Migration of Vascular Smooth Muscle Cells Across Tissue Boundaries in Atherosclerosis

Cross-tissue migration and proliferation of VSMC in the arterial wall cause intimal thickening that sets the stage for the pathogenesis of atherosclerosis and plaque formation.26 Atherosclerotic plaque formation and the danger of plaque rupture are major causes of atherosclerosis-related, acute cardiovascular events such as myocardial infarction. While the accumulation of VSMC and their collagen deposits in the fibrous cap stabilize the atherosclerotic plaque, macrophages and lipid contents destabilize it. Thus, migration and invasion potentials of VSMC have a direct impact on plaque formation and stability.

Differentiated VSMC in the medial layer of the arterial wall are furnished with an array of contractile myosin-actin filaments, aligned along the longitudinal axis of the cell for maintenance of vascular tone. In response to endothelial injuries, VSMC in the medial layer of arterial walls undergo a phenotypic modulation, characterized by a loss of contractile apparatus and remodeling of the cytoskeleton to acquire a migratory phenotype.27 Before VSMC are able to migrate, however, they have to break down or invade the extracellular matrix (ECM), hence clearing a path for migration. ECM invasion is achieved by secretion of proteases, primarily metalloproteases such as MMP2, MMP9 and MT1-MMP, at discrete sites of membrane protrusions called podosomes.28

Podosomes and Vascular Smooth Muscle Cell Invasion

Podosomes are dynamic, actin-based structures that act as sites of cell adhesion and active ECM remodeling.28–32 First discovered in monocytic cells such as macrophages, dendritic cells and osteoclasts, podosomes have since been found in many cell types including epithelial cells,33,34 endothelial cells35 and VSMC.36 Src-transformed fibroblasts and VSMC spontaneously produce podosomes, some of which self assemble, with uncertain mechanisms, into more stable ring structures called rosettes that bear features resembling the sealing zones or belts of podosomes found in osteoclasts37,38 (Fig. 1). Invadopodia, which are podosome-related structures found in carcinoma cells,39 appear in fewer numbers per cell as irregular, actin-rich dots. It remains a subject of debates whether podosomes and invadopodia represent different phases of the same structure, or unique, cell type-specific invasive organelles. Nevertheless, the discovery of podosomes and podosome-like structures in an increasingly wider spectrum of invasive cell types suggests a general requirement for these less well-understood cytoskeletal organelles in cell invasion in health and diseases. This view has received considerable attention and stimulated much interest in the study of podosomes and invadopodia in invasive cells, including tumor cells in cancers, osteoclasts in bone resorption, endothelial cells in angiogenesis and VSMC in vascular wound-healing and atherosclerosis.

A typical podosome is a finger-like vertical column arising from the ventral plasma membrane at sites of integrin clusters. The core of the podosome column (0.5 to several µm in diameter), which comprises linear and branched actin filaments, houses a squad of actin-nucleating and polymerizing proteins such as N-WASP, the Arp2/3 complex, cofilin, cortactin and caldesmon, and is surrounded by an outer ring of focal adhesion proteins such as α-actinin, vinculin, talin and paxillin. There is no known protein that exclusively localizes to podosomes. However, cortactin and its Tyr-phosphorylated counterpart always exist in podosomes, and have been widely used as podosome markers. Cortactin is a major Src-substrate that stabilizes branch points of actin filament network in lamellipodia and podosomes by binding to the Arp2/3 complex, N-WASP and F-actin.40

The sequence of events leading to podosome assembly involves remodeling of focal adhesions at integrin clusters and recruitment of Src and Src-substrates such as cortactin, FAK, Tks5/FISH and N-WASP to initiate actin nucleation and polymerization at the sites of podosome biogenesis. Src induces self-assembly of podosomes into more stable bands and rosettes in VSMC. The actin cores of individual podosomes in a rosette are connected by a network of radial actin filaments, a so-called “actin cloud,” that transmits tangential forces to organize the podosome arrays in the rosette.41 The transport of appropriate MMPs to podosomes and rosettes, which endows these structures with an ECM-degrading capacity, may be facilitated by molecular motors and the microtubule cytoskeletal system.42

Podosomes are highly dynamic structures with a lifetime of a few minutes.29,31,43 Elucidation of the mechanisms that regulate the assembly and disassembly of podosomes is an active area of research that has gained much momentum in recent years. Although a detailed picture is still wanting, it is clear that podosome dynamics are regulated by a network of coordinated signaling pathways involving cell migration and cytoskeleton regulators (RhoGTPases, integrins, N-WASP, Arp2/3, AFAP, caldesmon and cortactin),44 lipid kinases [PtdIns(3)-kinase (PI3K)], protein kinases (PKCs, PAK, FAK),33,45 intracellular and secreted proteases (Calpain, MMPs),46,47 adaptor proteins (Tks5/FISH),48 oncogenes (cSrc, Stat3)48 and tumor suppressors (p53, PTEN),5,6 and undoubtedly others yet to be discovered.

p53 and Src Play Antagonistic Roles in Podosome Formation in Vascular Smooth Muscle Cells

We have identified p53 as a suppressor of Src-induced podosome formation and associated invasive phenotype in VSMC.6 p53 acts via a two-prong mechanism: (1) by upregulating the expression of caldesmon, an actin-binding protein and a negative regulator of podosome assembly; and (2) by enhancing the expression of PTEN, thereby inhibiting the Src-mediated PI3K-Akt and Stat3 pro-invasion pathways.7 In contrast, constitutively active Src (SrcY527F) suppresses endogenous p53 expression and induces podosome formation and an invasive phenotype in VSMC6 by upregulating Stat3, which suppresses p53 and its target, caldesmon.7 These data suggest the existence of mutually antagonistic cross-talks between the anti-invasion and pro-invasion pathways mediated by p53 and Src3, respectively, serving as a check and balance that dictates the outcome of either an invasive or non-invasive phenotype7 (Fig. 2).

The p53-Mediated Anti-Podosome/Invasion Pathways

p53-caldesmon.

We have demonstrated that expression of caldesmon positively correlates with the transcriptional function of p53, which in turn negatively impacts on Src-induced podosome formation.6 This suggests that caldesmon may be a transcriptional target of p53 and is consistent with data obtained from the mapping of p53-binding sites in the human genome that identifies caldesmon as a potential p53 target gene.11

Caldesmon is a Ca²⁺/calmodulin- and actin-binding protein that plays a key role in the regulation of cytoskeletal organization and cell motility of non-muscle and de-differentiated ‘synthetic’ smooth muscle cells.49,50 Caldesmon modulates cell contractility by inhibiting actin-activated myosin ATPase activity, and depending on its expression level and interaction with Ca²⁺/calmodulin, either destabilizes or stabilizes actin stress-fibers and focal adhesions.51 Caldesmon was first identified in podosomes in Rous or avian sarcoma virus-transformed cells by Sobue's group in 1993.52 More recently, caldesmon has been shown to be a prominent component of podosomes in VSMC and fibroblasts, and a repressor of podosome formation and cell invasion,6,51,53 that is in part attributable to its ability to inhibit Arp2/3-mediated actin branching.54 That caldesmon acts as an invasion suppressor is also consistent with findings that expression of caldesmon is repressed in a number of cancer and transformed cells.52 It remains to be determined if p53 modulates caldesmon expression at the DNA level, or indirectly via microRNAs or at the protein level.

p53-PTEN.

The tumor suppressor PTEN (Phosphatase and Tensin homolog deleted on chromosome 10), which is mutated in the majority of advanced, invasive tumors55 and in vascular remodeling,56 is another strong candidate for mediating p53-dependent podosome suppression. PTEN and p53 mutually enhance each other's expression and activity. Thus, PTEN has been shown to increase p53 protein levels and stability via phosphatase-dependent and -independent mechanisms.57 On the other hand, we have found that enhanced p53 increases PTEN expression in VSMC and suppresses Src-induced podosome formation.7 Together, these observations suggest that there is a positive feedback loop that regulates the expression and stability of p53 and PTEN.

PTEN is a phosphatase that exhibits both lipid- and protein-phosphatase activities in vitro. The lipid-phosphatase activity of PTEN has been well documented to play a dominant role in tumor suppression by negatively modulating the PI3K-Akt pathway.58 Upregulation of PTEN by p53 would lead to a decrease in PI(3,4,5)P₃ and PI(3,4)P₂ in the plasma membrane and a reduction in podosome/rosette formation.55 This is in line with the finding that PI(3,4)P₂ is enriched in focal adhesions and podosomes, and is responsible for recruiting the adaptor protein, Tks5/FISH, which is essential for podosome formation, to sites of podosome biogenesis.59

Recent studies using the PTEN(Y138L) and PTEN(G129E) mutants, which exhibit only lipid- or protein-phosphatase activity, respectively, have demonstrated that both activities are involved in cell motility and invasion.60–63 For example, Davidson et al.60 have shown that both the lipid- and protein-phosphatase activities are required for PTEN to inhibit invasion into the surrounding matrix when expressed in PTEN-null glioblastoma cells, but expression of either activity alone is insufficient to block invasion. We have also shown that both the protein- and lipid-phosphatase activities of PTEN contribute to p53-mediated suppression of Src-induced podosome and rosette formation in VSMC.64

A number of in vitro protein substrates of PTEN have been identified, but their physiological relevance remains to be firmly established. We have reported that shRNA-driven knockdown of PTEN increases the phosphorylation and activation of Stat3 (pY705Stat3); and that overexpression of PTEN decreases Stat3 activity, resulting in podosome suppression.7 It is not clear whether PTEN has a direct role in Stat3 dephosphorylation via its protein-phosphatase activity or an indirect role via PI3K-Akt signaling. The identification of direct protein substrates of PTEN will be critical for our understanding of its role in cell migration. Taken together, these recent findings suggest that the p53-PTEN pathway inhibits cell invasion by inactivating podosome agonists via its protein-phosphatase activity on the one hand and by suppressing the pro-invasive PI3K-Akt pathway via its lipid-phosphatase activity on the other. Much work is required, however, to address the question of whether the protein- and lipid phosphatase activities of PTEN act independently or co-operatively.

The Src-Mediated Pro-Podosome/Invasion Pathways

Activation of c-Src by receptor Tyr kinases, such as PDGFR and EGFR, induces the formation of podosomes and rosettes in various cell types including VSMC, fibroblasts, endothelial cells and epithelial cells.65,66 Many Src-substrates, such as focal adhesion kinase (FAK), Stat3,67 cortactin,68 Tks5 and Tks4,48,69 and β-dystroglycan70 are enriched in Src-induced podosomes and rosettes, and play distinct roles in the regulation of actin cytoskeleton remodeling and podosome formation. For instance, phosphorylation of cortactin by Src is involved in regulating the formation and/or turnover of podosomes in A7r5 smooth muscle cells.68 Src-dependent phosphorylation of β-dystroglycan, a cell adhesion receptor, promotes its association with Tks5 and podosome formation in myoblasts.70 Recently, it has been shown that in response to PDGF, Src inhibits p53 leading to a decrease in the expression of microRNAs, miR-143/145, in A7r5 VSMC lines, thereby upregulating podosome formation.71

Src-Stat3.

Stat3 is a member of the Signal Transducer and Activator of Transcription family, which is activated by Jak and Src family kinases.67 Stat3 is upregulated in many cancers and is implicated in the promotion of aggressive metastasis via the transactivation of MMPs.72 We have shown recently that Stat3 is a required downstream effector of Src in inducing podosome formation and related invasive phenotypes. Stat3 promotes Src-induced invasive phenotypes through the suppression of p53 and the p53-caldesmon pathway. Furthermore, Stat3 and its activated form, pY705Stat3, localize to Src-induced podosomes.7 This suggests that Stat3 is recruited to podosomes where it is activated by Src, and thus mediates a Src-induced invasive phenotype either through its transcriptional functions and/or transcriptional-independent activities.73 Identification of downstream effectors of Stat3 that participate in the regulation of podosome dynamics will be critical to further our understanding of its role in cell invasion.

Src-PI3K-Akt.

Akt (also known as Protein Kinase B, PKB) is a Ser/Thr kinase that plays a pivotal role in signaling downstream of receptor Tyr kinases (RTKs) involved in cell growth, angiogenesis and cell migration.74 PI3K is activated by either binding directly to RTKs, or indirectly through binding to Tyr-phosphorylated adaptor proteins such as IRS1. PI(3,4,5)P₃ produced by PI3K recruits Akt to the plasma membrane, where PDK1 and mTOR complex 2 (mTORC2) phosphorylate Akt at T308 in the activation loop and S473 in the hydrophobic motif, respectively, resulting in Akt activation.

We have shown that activated Src upregulates Akt expression, accompanied by enhanced podosome/rosette formation and ECM digestion in VSMC.7 However, how Akt promotes podosome/rosette formation and cell invasion remains unclear. This is further complicated by findings that the Akt1 and Akt2 isozymes, both of which are expressed in VSMC, often exhibit opposite roles in cell migration.75 Numerous potential Akt substrates have been reported, many of which are involved in cell growth and proliferation, cell survival, metabolism and angiogenesis. However, relatively few Akt substrates have been implicated in cell migration. The sodium-hydrogen exchanger (NHE1) has been shown to be a bona fide Akt substrate necessary for growth factor-mediated remodeling of actin cytoskeleton.76 Two other reports have provided evidence for a possible link between Akt-PDK1 and PAK1 (p21-Associated Kinase), a key modulator of cell migration and podosome formation in VSMC.45 First, Akt phosphorylates PAK1 and modulates its binding to the adaptor protein, Nck, and cell migration.77 Second, PAK1 acts as a scaffold to recruit Akt1 and PDK1 to PI(3,4,5)P₃ in the plasma membrane for efficient activation of Akt.78

PAK1 is a member of the PAK family of Ser/Thr kinases that are activated by Cdc42/Rac.79,80 Of the six PAK isoenzymes identified in mammalian tissues, PAK1, PAK2 and PAK3 are expressed in VSMC, with PAK1 being the predominant species. PAK1 may play a role in podosome formation by phosphorylating substrates such as LIM kinase,81 caldesmon82 and cortactin,83 which are key regulators of actin polymerization in podosome dynamics. Alternatively, PAK1-induced formation of podosomes can be mediated by forming a PAK1/PIX/GIT complex, which targets PAK1 to focal adhesions, where GIT in turn binds to paxillin and initiates focal adhesion remodeling prior to podosome assembly.84 PIX (PAK-Interacting exchange factor)85,86 is a multifunctional protein containing a number of protein interaction domains including a SH3 motif that binds to one of the Pro-rich domains of PAK1, a Pleckstrin-homology (PH) domain and a tandem Dbl homologous (DH) domain that is responsible for its role as a guanine nucleotide exchange factor (GEF) of Cdc42 and Rac, a GBD domain that binds the G protein-couple receptor kinase-interacting target (GIT).87 It is quite possible that at the initiation of podosome assembly, PAK1 acts as a scaffold for Akt1 and PDK1 and targets them to PIX-GIT and PI(3,4,5)P₃ at the sites of podosome biogenesis. This would facilitate the phosphorylation of PAK1 by activated Akt and invokes downstream events in podosome formation (Fig. 2).

Perspective and Conclusion

Cell proliferation and cell invasion are defining characteristics, not only of cancer cells, but also of VSMC in the pathogenesis of atherosclerosis and plaque formation. Accumulating data support the intriguing hypothesis that p53 plays a dual role in the regulation of cell growth, and cell migration/invasion. Much work is required to determine if and how transcriptional functions of p53 during nuclear stress may be coupled to cytoskeletal remodeling events. It is tempting to speculate that such coordination may be fulfilled by dual-functional factor(s) such as JMY that coordinates p53 nuclear functions in response to stress and cytoskeletal events in cell migration/invasion.21,22 Alternatively, it cannot be ruled out that p53 independently regulates the transcription of key regulators of cell cycle and cytoskeletal remodeling.

Figures and Tables

Figure 1 Src-induced formation of podosomes and rosettes in rat aortic smooth muscle cells that actively digest extracellular matrix. (A) A control rat aortic smooth muscle cell (RASMC) showing prominent actin-stress fibers stained with FITC-phalloidin. (B) RASMC line expressing constitutively active Src(Y527F) encoded in retroviral vectors. Note the disappearance of actin-stress fibers and formation of podosomes that self-assemble into rosettes. (C) Digestion of TRITC-labeled fibronectin (red) in ECM by rosettes appears as dark spots. Actin is stained with FITC-phalloidin (green). Scales bars represent 20 µm.

Figure 2 Proposed mechanisms for anti- and pro-podosome formation mediated by p53 and Src, respectively. p53 suppresses podosome formation in VSMC by upregulating caldesmon and PTEN. Caldesmon acts as a podosome antagonist by inhibiting Arp2/3-mediated actin polymerization. Both PTEN protein- and lipid-phosphatase activities contribute to podosome suppression by downregulating the Src-Stat3 and Src-PI3K-Akt pro-podosome and invasion pathways. Membrane receptor tyrosine kinases activate Src, which induces podosome formation and cell invasion by phosphorylating podosomal proteins such as cortactin and Tks5/FISH. Activation of PI3K by receptor tyrosine kinases or Src increases PI(3,4,5)P₃, which together with GIX-PIX recruit the Akt-PAK1-PDK1 ternary complex to sites of podosome biogenesis at the plasma membrane. Activated Akt and PAK1 proceed to modulate regulatory proteins engaged in podosome dynamics. While p53 and Src mutually antagonize each other, a positive feedback loop appears to exist between p53 and PTEN.

Acknowledgements

ASM is funded by the Canadian Institute of Health Research and Heart and Stroke Foundation of Ontario. Infrastructure supports provided by the Canada Foundation of Innovation (CFI)-funded Protein Function Discovery facilities are gratefully acknowledged. I thank Graham Côtí and Robert Eves for reading the manuscript.