Expression of V3 Versican by Rat Arterial Smooth Muscle Cells Promotes Differentiated and Anti-inflammatory Phenotypes*

Background: Arterial smooth muscle cells (ASMCs) undergo phenotypic changes during pathophysiological processes. Results: V3 expression increases contractile SMC markers while decreasing a broad range of proinflammatory chemokines and transcription factors. Conclusion: V3 expression reprograms rat ASMCs promoting differentiated and anti-inflammatory phenotypes. Significance: Modifying ECM components via V3 expression could provide a potential therapeutic intervention against vascular and other diseases. Arterial smooth muscle cells (ASMCs) undergo phenotypic changes during development and pathological processes in vivo and during cell culture in vitro. Our previous studies demonstrated that retrovirally mediated expression of the versican V3 splice variant (V3) by ASMCs retards cell proliferation and migration in vitro and reduces neointimal thickening and macrophage and lipid accumulation in animal models of vascular injury and atherosclerosis. However, the molecular pathways induced by V3 expression that are responsible for these changes are not yet clear. In this study, we employed a microarray approach to examine how expression of V3 induced changes in gene expression and the molecular pathways in rat ASMCs. We found that forced expression of V3 by ASMCs affected expression of 521 genes by more than 1.5-fold. Gene ontology analysis showed that components of the extracellular matrix were the most significantly affected by V3 expression. In addition, genes regulating the formation of the cytoskeleton, which also serve as markers of contractile smooth muscle cells (SMCs), were significantly up-regulated. In contrast, components of the complement system, chemokines, chemokine receptors, and transcription factors crucial for regulating inflammatory processes were among the genes most down-regulated. Consistently, we found that the level of myocardin, a key transcription factor promoting contractile SMC phenotype, was greatly increased, and the proinflammatory transcription factors NFκB1 and CCAAT/enhancer-binding protein β were significantly attenuated in V3-expressing SMCs. Overall, these findings demonstrate that V3 expression reprograms ASMCs promoting differentiated and anti-inflammatory phenotypes.


Arterial smooth muscle cells (ASMCs) undergo phenotypic changes during development and pathological processes in vivo and during cell culture in vitro. Our previous studies demonstrated that retrovirally mediated expression of the versican V3 splice variant (V3) by ASMCs retards cell proliferation and migration in vitro and reduces neointimal thickening and macrophage and lipid accumulation in animal models of vascular injury and atherosclerosis. However, the molecular pathways induced by V3 expression that are responsible for these changes
are not yet clear. In this study, we employed a microarray approach to examine how expression of V3 induced changes in gene expression and the molecular pathways in rat ASMCs. We found that forced expression of V3 by ASMCs affected expression of 521 genes by more than 1.5-fold. Gene ontology analysis showed that components of the extracellular matrix were the most significantly affected by V3 expression. In addition, genes regulating the formation of the cytoskeleton, which also serve as markers of contractile smooth muscle cells (SMCs), were significantly up-regulated. In contrast, components of the complement system, chemokines, chemokine receptors, and transcription factors crucial for regulating inflammatory processes were among the genes most down-regulated. Consistently, we found that the level of myocardin, a key transcription factor promoting contractile SMC phenotype, was greatly increased, and the proinflammatory transcription factors NFB1 and CCAAT/en-hancer-binding protein ␤ were significantly attenuated in V3-expressing SMCs. Overall, these findings demonstrate that V3 expression reprograms ASMCs promoting differentiated and anti-inflammatory phenotypes.
Versican is a large chondroitin sulfate (CS) 3 proteoglycan that is present at low levels in quiescent tissues but accumulates during embryonic development, inflammation, cancer, cardiovascular, and other diseases (1)(2)(3)(4)(5)(6). Versican expression is induced by a number of inflammatory stimuli (7,8) and has been shown to play a key role in leukocyte adhesion (9,10). Versican exists in at least four major splice variants in humans as follows: V0, V1, V2, and V3. V3, the smallest splice variant, encodes a polypeptide that lacks the glycosaminoglycan binding domain, and thus contains no CS chains. Because versican CS chains can interact with a number of inflammatory mediators, such as MCP1, CD44, PSGL-1, and TLR2 (11)(12)(13)(14)(15), and as V3 lacks ADAMTS cleavage sites present in the larger isoforms of versican, V3 has been expected to behave very differently from the larger isoforms. We and others have shown that V3 expression induces significant changes in cell phenotypes. For example, V3 expression significantly decreased proliferation of melanoma cells and cardiomyocytes (16 -19) and increased cell-cell association (19). In melanoma cells, this reduction in cell proliferation was partly due to the inhibition of epidermal growth factor (EGF)-dependent signaling pathways (16,17). In Splotch mice, which have mutations in the Pax3 gene, versican overexpression was associated with defective neural crest migration characterized by altered expression of V3 (20,21). In chondrocytes, V3 expression reduced the expression of aggrecan, the principal proteoglycan present in cartilage (22). Simi-larly, we have shown that controlled expression of V3 alters phenotypes of fibroblasts and arterial smooth muscle cells (ASMCs) such that cell growth and migration are reduced while cell adhesion is enhanced (23)(24)(25). Moreover, V3 expression in ASMCs induced a significant change in the extracellular matrix (ECM) composition, including a reduction of the larger isoforms of versican, V0/V1, and hyaluronan as well as corresponding hyaluronan synthase expression, although elastic fiber deposition was greatly enhanced (23)(24)(25). Furthermore, V3-expressing ASMCs exhibited a reduced capacity to bind monocytes in vitro and in vivo (26) by altering TGF␤-, EGF-, and NFB-dependent signaling pathways (27). However, the molecular pathways by which V3 expression induces phenotypic switches in ASMCs, and the key mediator(s) that promote such changes in cell behavior, are not known.
In this study, a microarray analysis was employed to provide an unbiased, top-down approach to examine the changes induced by V3 expression in global gene expression in ASMCs.
Our findings indicate that V3 expression reprograms ASMCs to a differentiated, anti-inflammatory phenotype.

Experimental Procedures
Cell Culture-Fischer rat ASMCs were obtained and cultured as described previously (28). Retrovirally transduced cells were maintained in high glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS, sodium pyruvate, nonessential amino acids, glutamine, and penicillin/streptomycin (Invitrogen). Cells were used between 4 and 8 passages after the initial transduction.
Expression of V3 in ASMCs-The versican V3 splice variant was expressed in Fischer rat ASMCs using retroviral vectors, as described previously (23)(24)(25). Briefly, the rat V3 sequence was inserted into the BamHI site of the retroviral vector LXSN (courtesy of Dr. A. D. Miller, Fred Hutchinson Cancer Research Center, Seattle, WA). The retroviral vector containing the V3 coding sequence (LV3SN), as well as the empty vector control (LXSN), were used to infect Fischer rat ASMCs using PA317 packaging cells as described previously (24,28).
Microarray-Total RNA was first isolated by the Single Step Method by solubilizing the cells cultured for 3 weeks in solution D and 2.0 M sodium acetate (29). The soluble phase from subsequent phenol/chloroform/isoamyl alcohol extraction was precipitated with isopropyl alcohol. Pelleted RNA was further purified using an Agilent total RNA isolation mini kit (Agilent Technologies) according to the manufacturer's specifications. Microarray analysis was performed at the Medical University of South Carolina Proteogenomics Facility (Charleston, SC) using established methods (30 -32). Briefly, total RNA quality was evaluated by Agilent 2100 Bioanalyzer, and samples were then converted to biotin-labeled, fragmented cRNA using the 3Ј IVT Express Plus kit (Affymetrix). Hybridizations were performed using Rat Genome 230A GeneChips (Affymetrix), and posthybridization washing, staining, and scanning were done with the Affymetrix Hybridization, Wash and Stain kit, and a 7G scanner according to Affymetrix protocols. Raw data (CEL files) were normalized by Robust Multichip Average (RMA) (33) using Affymetrix Expression Console software. Significant difference in gene expression was defined by absolute fold change of Ͼ1.5 and p Ͻ 0.05 (Student's paired t test). Using these criteria, the false discovery rate (FDR), estimated by iterative permutation of sample groupings, was estimated to be ϳ1.7%. Gene ontology (GO) analysis and functional annotation clustering were performed using Database for Annotation, Visualization, and Integrated Discovery (DAVID) 6.7 (NIAID, National Institutes of Health). The total probe list of the microarray was used as the background gene set for the analysis. Microarray data collected for this study have been deposited in NCBI Gene Expression Omnibus in accordance with MIAME conventions (34) under accession number GSE66624.
Transcriptional Regulation Element Promotor Analysis and Interaction Network-Over-represented transcription factor response elements (TRE) were identified within promoter sequences (500 bp upstream of transcription start site(s)) of significantly up-and down-regulated genes using the PAINT version 4.0 (35,36). For the 261 genes that showed significant down-regulation, 232 corresponding unique upstream sequences were found, with 164 TREs detected based on a core similarity threshold of 1.0. For the 260 genes that showed robust up-regulation, 176 unique upstream sequences were found, with 167 detected TREs. TRE enrichment was determined based on a p value threshold of 0.05 and an FDR threshold of 0.3, relative to the default PAINT reference set of randomly selected Rattus norvegicus promoters.
Quantitative Polymerase Chain Reaction (qPCR)-The microarray detected gene expression in ASMCs cultured for 3 weeks at which time elastin deposition was maximal. To confirm whether the differential gene expression was present in V3-expressing ASMCs at earlier time points, qPCR was performed on ASMCs cultured over a time course. DNA-free RNA was obtained from control LXSN-or V3-expressing ASMCs cultured for 1, 4, 7, and 10 days using the total RNA isolation kit from Agilent Technologies, and cDNA was prepared from the isolated RNA by reverse transcription using random primers with a high capacity cDNA archive kit from Applied Biosystems. Real time PCR was carried out using an Applied Biosystems Prism 7500 sequence detection system and a SYBR Green or TaqMan Universal PCR master mix kit from Applied Biosystems as directed by the manufacturer. Primers used with SYBR Green reagents are listed in Table 1. For each group, assays were run on samples isolated from 3 to 4 replicates using the relative standard curve method (Applied Biosystems).
Cytokine ELISA-Conditioned media from control LXSN or V3-expressing ASMCs cultured for 3 days were filtered through 0.22-m pore syringe filters and spun at 3000 ϫ g for 5 min at 4°C. Cytokine levels in the conditioned media were quantified by Rat Cytokine/Chemokine 27 Plex Discovery Array (Eve Technologies).
Western Blotting-Cytoplasmic proteins were isolated from control or V3-expressing ASMCs grown for 7 days using NE-PER nuclear and cytoplasmic extraction reagents (Thermo Scientific) according to the manufacturer's protocols. Total protein concentration was determined by BCA assay (Thermo Scientific). 30 g of total protein from cytoplasmic extracts was run on 8% SDS-PAGE and transferred to a nitrocellulose membrane at 10 V constant voltage overnight at 4°C. Nitrocellulose membranes were blocked in 2% BSA in Tris-buffered saline with 0.1% Tween 20 and incubated with polyclonal antibodies against the C-terminal domain of Fak (EMD Millipore; formerly Upstate), which detects both Fak and Frnk (37), followed by horseradish peroxidase-conjugated anti-rabbit secondary antibody (1:80,000). Membranes were developed with a chemiluminescent substrate, West Dura (Pierce Biotech). Densitometry was performed using ImageJ version 1.44 from the National Institutes of Health.
Statistical Analysis-All data are expressed as the average Ϯ S.E., unless otherwise specified. Differences were identified by two-tailed Student's t tests for the comparison of two groups and by one-way analysis of variance followed by Tukey's post hoc tests for the comparison of three or more groups and were regarded significant if p Ͻ 0.05.

Expression of Versican V3 Significantly Alters Gene Expression Profiles in Rat
ASMCs-DNA microarray analysis was performed on RNA isolated from primary rat ASMCs retrovirally transduced with empty vector control (LXSN) and V3-containing vector (LV3SN) and maintained in culture for 3 weeks. The 3-week period was chosen because it is the period of time required for prominent accumulation of elastin. Analysis of microarray data identified 521 genes that were significantly differentially expressed in ASMCs following V3 transduction (supplemental Table 1; Fig. 1). A subset of genes showing the highest magnitude changes is shown in Table 2. Among the genes most up-regulated were components of actin cytoskeleton such as Acta1, Myh10, and Myl9, and ECM components, including Fbln5, Adamts9, and osteoglycin, a small leucine-rich proteoglycan. Among the most down-regulated genes were complement component 3 and inflammatory cytokine genes such as Cxcl1, Ccl20, and Ccl2.

V3 Expression Affects the Extracellular Microenvironment-
To examine the biological processes influenced by V3 expression, genes significantly affected in V3-transduced cells were subjected to functional enrichment analysis based on their GO annotations. As shown in Table 3, the processes most significantly enriched by V3 were linked with extracellular region, extracellular space, and ECM. Table 4 shows expression data for significantly regulated genes categorized as extracellular region (GO:0005576). Review of the genes, their related functions, and their changes in expression revealed a dynamic and complex alteration of the extracellular microenvironment in response to V3. Some ECM components were up-regulated (fibulin 5, elastin, laminin ␤1, and SPARC-related molecular calcium binding 1 (Smoc1)), whereas others were down-regulated (versican, cartilage oligomeric matrix protein, and proline arginine-rich end leucine-rich repeat protein (Prelp) and decorin). The levels of V0/V1 were significantly down-regulated by expressing V3 when examined in control and V3 overexpressing ASMCs from four independent transductions. In comparison, the levels of V3 expression achieved by transduction were greater than 10 times the V0/V1 levels naturally expressed in the control ASMCs (supplemental Fig. 1). Some proteases and protease inhibitors were up-regulated (Adamts9 and carboxypeptidase E), although others were down-regulated (Timp1, ectonucleotide pyrophosphatase/phosphodiesterase 2, secretory leukocyte protease inhibitor, and extracellular proteinase inhibitor). Some genes involved in ECM-cell interactions were up-regulated (vinculin) and others were downregulated (dystroglycan 1). This response of up-and down-regulation of genes within specific functional groups was also seen for several other categories, including cell-cell adhesion, growth factors, and hormones, and the coagulation and complement cascades. However, the chemokine category was nota- ble in that all of these genes were down-regulated (lipopolysaccharide-binding protein, angiogenin, Wnt4, Cxcl1, Ccl20, Ccl2, Cxcl6, Cxcl12, and macrophage migration inhibitory factor). Based on these findings, V3 expression results in extensive remodeling of the extracellular microenvironment and an alteration in the inflammatory status of ASMCs. V3 Expression Significantly Affects Genes Regulating Cell Survival and Inflammation-To more clearly assess the many processes affected by V3 expression, gene annotations were further analyzed to identify significant groups (clusters) of related processes. This analysis again highlighted that the extracellular environment, cytoskeletal system, and inflammatory status were all influenced by V3 (Table 5). Additionally, this analysis highlighted two clusters of functional groups from the prior analyses as follows: apoptosis/cell death and vascular development. Table 6 shows the changes in expression of genes associated with negative regulation of apoptosis. The gene in this category that was most up-regulated in V3-expressing ASMCs was Ptk2, which can encode for Fak or its dominant negative form Frnk. This was intriguing because Fak is a key regulator of many important cellular processes, including cell proliferation, adhesion, motility, and survival (38 -42). V3 expression up-regulated other signaling molecules and transcription factors involved in the negative regulation of apoptosis, including myocyte enhancer factor 2C (Mef2c), a transcription factor promoting cardiac muscle differentiation and vascular development; cadherin 13, a nonclassical glycosylphosphatidylinositol-anchored transmembrane cadherin negatively regulating EGF signaling and a potential tumor suppressor; Cd59, a regulator of complement-mediated cell lysis; and Rho kinase (Rock1), a key regulator of the actin cytoskeleton promoting contractile force generation. Tcf4 was increased by 2-fold in V3-expressing ASMCs, whereas Wnt4 was decreased by 2.2-fold, and recombination signal-binding protein for immunoglobulin J region (Rbpj), a transcription factor activated by Notch signaling, was decreased 1.74-fold, indicating that the Wnt/␤-catenin as well as Notch signaling pathways are also affected by V3 expression. Notably, Smad3 was decreased by 2.4-fold in V3-expressing ASMCs, indicating that the TGF␤ signaling pathway is affected by V3 expression. Many ECM-degrading enzymes, cytokines, and growth factors discussed above that are also involved in cell survival are included within this GO. A number of key regulators of apoptosis were down-regulated by V3 expression, including caspase 3, cell death-inducing DNA fragmentation factor ␣ subunit-like effector A, clusterin, and Bcl2. Interestingly, the antioxidant enzymes peroxiredoxin 5 and superoxide dismutase 2 were down-regulated by V3 expression, which may reflect reduced metabolic burden of these cells after 3 weeks of culture because V3 expression significantly reduced proliferation. Proinflammatory transcription factors such as Nfkb1 and CCAAT/enhancer-binding protein ␤ (Cebpb) were also decreased by 1.9-and 2.6-fold, respectively, in V3-expressing ASMCs.
The expression patterns of several genes involved in apoptosis were interrogated by qPCR at days 1, 4, 7, and 10 after V3 transduction. Expression levels of Cebpb and Nfkb1 in V3-expressing ASMCs were lower than those of control ASMCs at day 1, but the difference was not statistically significant (Fig. 2, p ϭ 0.22 and 0.21, respectively). However, these levels increased in control ASMCs but were significantly reduced in V3-expressing ASMCs at days 4, 7, and 10. Expression levels of caspase 3 and Rela were also significantly reduced in V3-ex- pressing ASMCs, although Rela was statistically significant only at day 4 (p ϭ 0.004, Fig. 2). These findings indicate that V3 expression prevents the increase of pro-apoptotic and pro-inflammatory transcription factors during cell culture. Table 7 shows the expression changes of genes regulating response to wounding, many of which overlap with the genes listed in Table 5, including components of the complement cascades, growth factors, and cytokines. Notably, retinoid X receptor ␣ was increased by 3.36-fold, which suggests that the elastogenic effect of V3 may be partially dependent on retinoic acid receptor-dependent signaling pathways that have also been shown to promote elastogenesis, especially in the lungs (43,44). Notch3 was increased by 2.48-fold, along with the decrease of Rbpj shown in Table 6, indicating that V3 affects Notch signaling pathways. Moreover, components of the cytoskeletal network, such as myosin heavy chain 10, tropomyosin ␣, and microtubule-associated protein 1B were also increased by 4.92-, 2.69-, and 2.06-fold in V3-expressing ASMCs.  To confirm that V3-expressing ASMCs had a marked reduction in the inflammatory mediators shown in Table 7, we examined transcript levels of C3, Ccl2, Cxcl1, Ccl20, and Cxcl5 from control LXSN or V3-expressing ASMCs cultured for 10 days. As shown in Fig. 3, V3-expressing ASMCs had reduced transcript levels of these inflammatory mediators. Among these, we also examined Ccl2 and Cxcl1 protein levels, verifying that both were significantly lower in conditioned media from V3-expressing ASMCs. These data demonstrate that V3 expression negatively impacts expression of  a number of key complement components and inflammatory chemokines. V3 Expression Promotes Mature, Differentiated ASMC Phenotypes-V3 expression increased the expression of Mef2c, which promotes muscle differentiation and influences the actin cytoskeletal network. We therefore sought to examine whether V3 expression in ASMCs promoted a differentiated SMC phenotype. As shown in Fig. 4, V3 expression significantly up-regulated levels of smooth muscle ␣-actin (SMA) and calponin, which are components of the cytoskeletal network as well as markers of contractile SMCs. V3 expression also up-regulated myocardin, which is a key transcription factor promoting con-tractile phenotypes of SMCs (45). Notably, V3-expressing ASMCs exhibited increases in these SMC markers at day 1, which persisted up to day 10. These findings indicate that V3 expression increases differentiated and mature SMC markers.
TRE for SRF, Stat1, and NFB Are Enriched in Promoter Regions of Genes Affected by V3 Expression-To elucidate transcriptional mechanisms by which V3 influenced differential gene expression in ASMCs, we conducted an analysis of TREs enriched in promoters of the 521 genes significantly affected by V3 expression. Among the genes up-regulated by V3 expression (fold change Ͼ1.5, p Ͻ 0.05), the TRE for binding to the serum-response factor (SRF), a key transcription factor promoting contractile SMC phenotypes, was significantly overrepresented (Table 8, FDR Ͻ0.3). For the genes down-regulated by V3 expression (fold change Ͻ Ϫ1.5, p Ͻ 0.05), TREs significantly over-represented included those for binding to Stat1 and NFB (Table 9, FDR Ͻ0.3), indicating that V3 expression blocks these pro-inflammatory transcription factors.
Frnk, but Not Fak, Is Significantly Up-regulated in V3-expressing ASMCs-One of the genes significantly up-regulated by V3 expression was Ptk2, which can encode either Fak or its endogenous inhibitor Frnk (Table 6). Fak is one of the key signaling molecules in the focal adhesion complex that is activated via integrin or growth factors, and it further activates downstream signaling pathways mediated by PI3K, MAPK, and/or NFB, promoting cell proliferation and migration (38 -42). Therefore, we further explored whether V3 expression affects Fak or its dominant negative isoform Frnk. As shown in Fig. 5, V3-expressing ASMCs had significantly greater transcript levels of Frnk, but not Fak, at day 1 of culture, which persisted up to day 10. V3-expressing ASMCs had greater amounts of Frnk at protein levels, when examined by Western blotting (Fig. 5B). These findings suggest that V3-expressing ASMCs have altered focal adhesion signaling pathways that can further impact other cell phenotypes.

Discussion
In this study, we demonstrated that V3 expression reprograms ASMCs to enhance expression of differentiated, mature SMC markers while repressing components of the complement cascades, inflammatory cytokines, and cell adhesion molecules, promoting an anti-inflammatory phenotype.
Expression of the V3 transcript in vivo has been detected in mouse, rat, and human fetal and adult tissues and cells (46 -50), and most recently, secretion of V3 protein by primary human skeletal muscle cells was detected by proteomic analysis (51). Previous studies have shown that V3 expression induces increases in cell-cell and cell-ECM association (19,(23)(24)(25). This is accompanied by a remodeling of the ECM such that accumulation of aggrecan, hyaluronan, and the larger isoforms of versican, V0/V1, was reduced (22)(23)(24)(25), whereas elastic fiber deposition was greatly enhanced (23)(24)(25), generating a microenvironment resistant to monocyte adhesion both in vitro and in vivo (26,27). Our present findings show that V3 expression induces significant remodeling of the extracellular microenvironment, indicated by the changes observed in the ECM molecules, growth factors, cytokines, components of the complement cascades, proteases, and protease inhibitors, as  well as molecules involved in cell-cell adhesions. Components of elastic fiber, fibulin 5, and elastin were up-regulated by V3 expression, as demonstrated previously in our studies (27,52). Versican expression was down-regulated, confirming our pre-vious findings that V3 negatively regulates the larger isoforms of versican (23). Additionally, basement membrane components such as laminin ␤1 and Smoc1 were up-regulated by V3 expression, whereas collagen-binding proteins such as decorin, Prelp, and cartilage oligomeric matrix proteins were down-regulated. Among the many growth factors and cytokines affected by V3 expression, TGF␤ superfamily members, including TGF␤1, follistatin, inhibin ␤-B, and TGF␤3, were significantly modulated. This is consistent with our previous findings that TGF␤ signaling pathways were affected by V3 expression (27).
We also found that a number of molecules mediating Wnt/ ␤-catenin and Notch signaling pathways were significantly affected by V3 expression, such as Wnt4, Tcf4, Notch3, and Rbpj. Given that these signaling pathways regulate tumorigenesis and that V3 expression in melanoma cells has been shown to significantly reduce tumorigenesis (16 -18), V3 may negatively impact tumorigenesis by affecting Wnt/␤-catenin and Notch signaling pathways. Recently, we have also found that negatively regulating versican in leiomyosarcoma cells significantly blocked tumorigenic potential when subcutaneously implanted into nude mice (53), which suggests V3 as a negative regulator of larger isoforms of versican may provide a novel anti-tumorigenic approach.   Vascular SMCs undergo phenotypic changes during development and pathological processes in vivo and during cell culture in vitro (45, 54 -57). For example, medial SMCs exhibit a mature and differentiated phenotype in which the cells express greater amounts of proteins necessary for the contractile function of the cells, including smooth muscle myosin heavy chain and SMA, compared with intimal SMCs that exhibit a "synthetic" phenotype with greater proliferative and migratory capacity and express more ECM proteins such as type I collagen (45,54,55). Our study clearly demonstrates that V3 expression increases a number of contractile SMC markers such as SMA, calponin 1, and myocardin, a key transcription factor promoting muscle differentiation, which occurred as early as 24 h after culture. Previously, myocardin had been shown to repress versican through induction of miR143 (58,59), consistent with the decrease in total versican transcript induced by V3 expression, suggesting that V3 may promote SMC differentiation by decreasing the amount of larger isoforms of versican. Consistently, genes up-regulated by V3 expression were enriched with TRE for SRF, a transcription factor promoting mature and differentiated SMC phenotypes. Changes in the TGF␤ superfamily induced by V3 expression, as discussed above, may be responsible for the increases in the markers of differentiated mature SMC phenotypes (60,61). Our findings that V3 expression induces increases in the differentiated SMC markers as well as in components of basement membrane and elastic fibers suggest that ASMCs can be reprogrammed to actively synthesize the ECM components present in healthy arteries while maintaining a differentiated and contractile phenotype. Intriguingly, proteomic profiling by mass spectrometry has identified V3 as part of the secretome of primary human skeletal muscle cells (51), suggesting it may play an intrinsic role in promoting muscle cell differentiation. FIGURE 5. A, to validate whether the changes in Ptk2 detected by microarray analysis were specific for Fak or Frnk, we used one set of primers against regions common for both Fak and Frnk (Frnk primers), and another set of primers for kinase domain that is present only in Fak but not in Frnk (Fak primers). V3-expressing cells had significantly elevated transcripts for Frnk but not Fak at day 1, which persisted up to day 10. n ϭ 3, *, p Ͻ 0.05 in unpaired t test. B, when probed by a polyclonal antibody, which detects both Fak and Frnk, we found a significant increase in Frnk but not Fak. n ϭ 4, *, p Ͻ 0.05 in unpaired t test. Our studies have previously demonstrated that V3-expressing ASMCs bind fewer monocytes in vitro and in vivo (26,27). This was due to the generation of an ECM enriched in elastin and depleted in hyaluronan, as well as a decrease in Vcam1, a key monocyte-binding cell surface molecule (27). Our present findings demonstrate that V3 expression also reduces other inflammatory mediators, including chemokines and cytokines, such as lipopolysaccharide-binding protein, angiogenin, Wnt4, Cxcl1, Ccl20, Ccl2, Cxcl6, Cxcl12, and macrophage migration inhibitory factor. This is most likely due to down-regulation of the proinflammatory transcription factors, CCAAT/enhancerbinding protein ␤ and NFB1, shown in this study. Such results indicate that V3 expression reprograms ASMCs to have an anti-inflammatory phenotype.
V3-expressing ASMCs exhibited a marked increase in cell spreading and larger areas of close contact (24). In this study, we found that V3 expression increased Ptk2, which encodes for Fak or Frnk. Fak mediates key signaling pathways elicited by extracellular stimuli activating integrins or growth factors, and it further regulates cell proliferation, survival, and migration via PI3K, MAPK, and/or NFB (38 -42). Down-regulation of Fak activation was correlated with increased expression of differentiated SMC markers when cells were plated on laminin, as opposed to fibronectin, which increased proliferation while decreasing differentiated SMC marker expression (38). Frnk is an endogenous inhibitor of Fak, and it has been shown to be selectively expressed by SMCs and large conduit blood vessels and negatively regulates proliferative and migratory phenotypes (62) while reinforcing expression of differentiated SMC markers (63). Recently, it has been shown that bone marrow mononuclear cells of multiple myeloma patients had elevated transcript levels of versican, ␤-catenin, ␤1 integrin, as well as Fak when compared with healthy control populations (64), suggesting that versican may have functional roles in ␤1 integrin and Fak signaling downstream of ␤-catenin. In our study, V3 expression significantly increased expression of Frnk, but not Fak, at mRNA and protein levels. Frnk has a transcriptional start site within an intron of the Fak gene, which is positively regulated by TGF␤ (63). This is concomitant with the aforementioned TGF␤ signaling pathways affected in V3-expressing ASMCs. Moreover, we have previously shown that activation of PI3K and NFB induced by EGF was significantly down-regulated in V3-expressing ASMCs (27). Based on these findings and our present finding that V3 expression significantly up-regulated Frnk, but not Fak, the effect of V3 on reprogramming ASMC phenotype may be due to altered Fak signaling pathways leading to increases in differentiated smooth muscle marker expression and decreases in activation of PI3K and NFB.
A significant aspect of our observations is that expression of a single ECM gene, V3, elicits a wide range of phenotypic reprogramming exhibited by expression changes in multiple genes. However, the molecular mechanism(s) that causes these changes is not understood. One possibility is that changes in cell shape caused by V3 expression induce such phenotypic reprogramming. Our earlier work showed that forced expression of V3 resulted in a dramatic cell shape change creating a more flattened, highly spread cell with large areas of close contacts potentially altering cell adhesion through Fak signaling (24). Such shape changes most likely came about as a result of the dramatic changes in the composition of the ECM deposited by the V3-expressing cells (23)(24)(25). There is extensive evidence that compositional changes in the ECM can lead to changes in cell shape and micromechanical properties, which in turn will alter gene expression and associated cell phenotypes (65)(66)(67)(68). Alternatively, V3 may directly interact with ␤1 integrins and EGF receptor via the C-terminal G3 domain (15), which will in turn affect Fak signaling downstream of integrins and EGF receptors. We suspect that V3 may modulate Fak signaling pathways by either directly interacting with these cell surface receptors or by altering the micromechanical environment around the cells.
Overall, these findings demonstrate that V3 expression reprograms ASMCs promoting anti-inflammatory and differentiated SMC phenotypes via altering cell-ECM interactions by regulating focal adhesion signaling pathways. This work provides novel insight into the mechanisms by which specific components of the ECM regulate SMC phenotypes, which may offer potential therapeutic targets in the treatment of vascular and other diseases.
Author Contributions-I. K. participated in the conception and design of the study, acquired, analyzed, and interpreted data, and wrote and revised the paper. J. L. B., acquired, analyzed, and interpreted microarray data and participated in drafting and revising the paper. E. P. S. acquired, analyzed, and interpreted microarray data and participated in drafting the paper. D. W. Y. acquired, analyzed, and interpreted qPCR and Western blotting data. G. A. W. acquired, analyzed, and interpreted qPCR data. K. R. B. contributed to the conception and design of the study, and acquired, analyzed, and interpreted microarray and qPCR data. W. S. A. participated in the conception and design of the study, analyzed and interpreted microarray data, and contributed to the writing of the paper. T. N. W. participated in the conception and design of the study, interpreted data, and contributed to the writing and revising of the paper. All authors reviewed the results and approved the final version of the manuscript (with the exception of W. S. A. who passed away before the final version was completed).