The actin depolymerizing factor destrin serves as a negative feedback inhibitor of smooth muscle cell differentiation

Abstract

We have previously shown that several components of the RhoA signaling pathway control smooth muscle cell (SMC) phenotype by altering serum response factor (SRF)-dependent gene expression. Because our genome-wide analyses of chromatin structure and transcription factor binding suggested that the actin depolymerizing factor, destrin (DSTN), was regulated in a SMC-selective fashion, the goals of the current study were to identify the transcription mechanisms that control DSTN expression in SMC and to test whether it regulates SMC function. Immunohistochemical analyses revealed strong and at least partially SMC-selective expression of DSTN in many mouse tissues, a result consistent with human data from the genotype-tissue expression (GTEx) consortium. We identified several regulatory regions that control DSTN expression including a SMC-selective enhancer that was activated by myocardin-related transcription factor-A (MRTF-A), recombination signal binding protein for immunoglobulin κ-J region (RBPJ), and the SMAD transcription factors. Indeed, enhancer activity and endogenous DSTN expression were upregulated by RhoA and transforming growth factor-β (TGF-β) signaling and downregulated by inhibition of Notch cleavage. We also showed that DSTN expression was decreased in vivo by carotid artery injury and in cultured SMC cells by platelet-derived growth factor-BB (PDGF-BB) treatment. siRNA-mediated depletion of DSTN significantly enhanced MRTF-A nuclear localization and SMC differentiation marker gene expression, decreased SMC migration in scratch wound assays, and decreased SMC proliferation, as measured by cell number and cyclin-E expression. Taken together our data indicate that DSTN is a negative feedback inhibitor of RhoA/SRF-dependent gene expression in SMC that coordinately promotes SMC phenotypic modulation. Interventions that target DSTN expression or activity could serve as potential therapies for atherosclerosis and restenosis.

NEW & NOTEWORTHY First, DSTN is selectively expressed in SMC in RhoA/SRF-dependent manner. Second, a SMC-selective enhancer just upstream of DSTN TSS harbors functional SRF, SMAD, and Notch/RBPJ binding elements. Third, DSTN depletion increased SRF-dependent SMC marker gene expression while inhibiting SMC migration and proliferation. Taken together, our data suggest that DSTN is a critical negative feedback inhibitor of SMC differentiation.

INTRODUCTION

Smooth muscle cells (SMCs) provide structural support to blood vessels, the gastrointestinal tract, and airways and regulate flow through these organs through specialized contractile mechanisms. Unlike cardiac and skeletal muscle cells, SMCs retain a remarkable plasticity that allows them to switch between differentiated and proliferative phenotypes in response to environmental cues. This plasticity is important for proper blood vessel development and physiological adaptation but also contributes to the progression of a variety of vascular disorders including atherosclerosis, hypertension, and restenosis.

Extensive evidence indicates that SMC differentiation is controlled by serum response factor (SRF) that binds to the CArG (CC(A/T)₆GG) cis elements that are present within nearly all of the SMC differentiation marker gene promoters (1). The myocardin family of SRF coactivators (myocardin and the myocardin-related transcription factors MRTF-A and MRTF-B) are required for SMC differentiation in many SMC subtypes and strongly activate SMC-specific gene expression when recruited by SRF (2). Importantly, Miralles et al. (3, 4) demonstrated that the MRTFs are sequestered in the cytoplasm by G-actin but translocate to the nucleus upon RhoA-dependent stimulation of actin polymerization. We were the first to demonstrate that RhoA signaling contributed to SMC differentiation and have extensively characterized the RhoA signaling pathways that drive SMC-specific gene expression (4–6). GTP-bound RhoA regulates cytoskeleton dynamics by interacting with various effectors including the Rho kinases ROCKI and ROCKII and the diaphanous family of formins (mDia1-3) (5, 6). mDia1 and mDia2 nucleate linear actin polymerization in cooperate with profilin, while ROCKs inhibit the disassembly of actin filaments through LIM kinase-dependent inhibition of the actin depolymerizing factors (ADFs) (7). LIMK-dependent phosphorylation at Ser-3 inhibits the ability of the ADFs to sever actin filaments thus increasing actin polymerization (8), whereas dephosphorylation of the ADFs by the slingshot phosphatases (SSH1-3) promotes ADF severing activity (9, 10).

It is clear that chromatin modifications that alter transcription factor access to DNA are also crucial for the overall pattern of SMC-specific gene expression observed in vivo. To address these mechanisms, we have characterized chromatin structure and transcription factor binding in human aortic SMCs on a genome-wide scale using DNase hypersensitivity and ChIP seq approaches, respectively (11). Our data suggest that SRF frequently binds in cooperation with Notch/RBPJ, a cell-cell interaction-regulated transcription complex that has been implicated in vascular development and SMC differentiation from multiple origins (12) and that these factors interact to control chromatin structure and gene expression in SMCs. This approach has also led to the identification of previously unrecognized SMC-selective proteins and regulatory elements. For example, we have showed that GRAF3, a Rho-specific GTPase-Activating Protein (GAP), also known as ARHGAP42 is selectively expressed in SMCs and controls blood pressure in mice and humans (13). We went on to identify rs604723 as the causal genetic variant for the BP-associated locus within the GRAF3 gene and that this single-nucleotide polymorphism (SNP) altered GRAF3 expression by promoting SRF recruitment to an intronic enhancer element (14).

The focus of the current study is on the downstream RhoA signaling protein, destrin (DSTN), the founding member of the ADF family of actin severing proteins that also includes the closely related cofilins, CFL1 and CFL2. Interestingly, the chromatin and transcription factor binding patterns observed near the DSTN transcription start site (TSS) suggested that this gene may be regulated by mechanisms similar to those that drive SMC-selective gene expression. Given previous studies demonstrating that DSTN negatively regulated SRF-dependent gene expression in corneal epithelial cells (15), we wanted to characterize DSTN’s role in SMC. We identified transcription mechanisms that drive DSTN expression in vascular SMCs and demonstrate that high DSTN expression levels may help explain the SMC’s unique ability to phenotypically modulate.

MATERIALS AND METHODS

Cell Culture

Human aortic and bronchial SMCs (AoSMCs and BrSMCs, respectively) were purchased from Lonza and maintained in Clonetics smooth muscle growth medium-2 containing 5% fetal bovine serum, insulin, epidermal growth factor, fibroblast growth factor, and gentamicin. Primary aortic SMCs were isolated from Wistar rats or mice, as previously described (16). In brief, thoracic aortas were stripped of the endothelial and adventitial layers by microdissection and then the SMCs were isolated by treatment with trypsin and collagenase. SMCs were cultured in Dulbecco’s modified Eagle’s medium with F12 supplemented with 10% fetal bovine serum and 0.5% penicillin-streptomycin. SMC preparations are routinely tested for smooth muscle-specific marker expression, and only the cell lines that are deemed at least 85% pure by these measurements are used for further experiments. Primary mouse ECs transformed with large T antigen were previously described (17). Mouse multipotential 10T1/2 myofibroblasts were obtained from American Type Culture Collection (ATCC). Both mouse ECs and 10T1/2 cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and 0.5% penicillin/streptomycin.

Plasmids

Potential DSTN regulatory regions were PCR-amplified from human BrSMC genomic DNA and cloned into pGL3 basic vector (Promega). SBE, CArG, and RBPJ mutants were generated by site-direct mutagenesis, and all mutations were verified by Sanger sequencing.

Luciferase Assays

Human BrSMCs, rat AoSMCs, 10T1/2 cells, and mouse ECs were seeded in 24-well culture plates at a density of ∼2.4 × 10⁴ cells/well. Cells were transfected 24 h after plating with 50 ng of plasmid per well. After 48 h of incubation at 37°C, luciferase assays were performed using Steady-Glo Luciferase Kit (Promega), according to manufacturer’s instructions. For agonist addition experiments, transfected cells were serum starved 24 h after transfection. After 24 h of starvation, agonists were added for 24 h followed by luciferase assay. For Notch signaling inhibition experiments, transfected cells were treated with 5 μM γ-secretase inhibitor, N-[(3,5-difluorophenyl)acetyl]-L-alanyl-2-phenyl]glycine-1,1-dimethylethyl ester (DAPT) (Cayman Chemical) 24 h after transfection, and luciferase activity was measured after 24 h.

siRNA Knockdown

Cells were transfected with 20–30 nM siRNA targeting DSTN, RBPJ, or green fluorescent protein (GFP) using the RNAiMAX (Invitrogen) or Dharmafect (Dharmacon) transfection reagent. Cells were assayed after 72 h transfection. Silencer siRNA for DSTN (siDSTN1, ID: 18756; siDSTN2, ID: 185571) was purchased from Invitrogen. SMARTpool siRNA for RBPJ (M-007772-00) was obtained from Dharmacon. The sequence of siRNA targeting GFP shows as follows: 5′-GGUGCGCUCCUGGACGUAGCC-3′.

ChIP Assay

Chromatin immunoprecipitation (ChIP) assay was performed, as described (18) according to the X-ChIP protocol (Abcam). In brief, human BrSMCs were fixed in 0.7% formaldehyde for 10 min. The cross-linking reaction was stopped by incubating cells with 0.125 M glycine for 5 min. Cells were scraped in lysis buffer (5 mM PIPES pH 8.0, 85 mM KCl, 0.5% Nonidet P-40) and then nuclei were isolated by centrifugation for 5 min at 2,300 g. Nuclei were lysed in nuclear lysis buffer containing 50 mM Tris-Cl (pH 8.1), 10 mM EDTA, and 0.13% SDS. Chromatin was sheared into 500-bp fragments by sonication followed by immunoprecipitation with 1–5 μg of antibodies overnight at 4°C. Antibodies used were anti-SRF (Santa Cruz), anti-RBPJ (Cell Signaling), anti-Notch3 (Santa Cruz), nonimmune rabbit IgG (Cell Signaling), and nonimmune mouse IgG (Millipore). In DAPT treatment experiments, cells were treated with DAPT (10 μM) or vehicle for 18 h before formaldehyde fixation.

Cell Proliferation Assay

Cell proliferation was determined by growth curves of human AoSMCs derived from cell counting. Cells were transfected with siRNA targeting DSTN or GFP in the absence of serum to synchronize the cells. After 24 h, 10% FBS was added and cell numbers were determined for the indicated period. Each count was an average of four repeats and each data point was an average of three separate experiments.

In Vitro Scratch Assay

The rate of cell migration was determined by scratch assay, as previously described (19). In brief, rat AoSMCs were transfected with siRNA targeting DSTN or GFP until cells reached high confluence to form a monolayer. A p200 pipette tip was used to create a scratch of the cell monolayer. Fresh growth medium was replaced after one wash with the same medium. Wound closure was then determined after 24 h. Cell movement was measured by ImageJ. Migration area of siDSTN and siGFP control was analyzed by 12 independent images from two sets of siRNA treatment. Experiments were performed two times and representative images were shown.

Carotid Artery Injury Model and Analysis

Carotid artery ligation was performed, as previously described (6). In brief, a suture was tied around the left common carotid artery just below the bifurcation with the right carotid artery serving as a sham control. After 2 wk of surgery, neck tissues were harvested, fixed, and embedded in paraffin using standard protocols. Sections (10 µm) were blocked with 5% goat serum in PBS and incubated with the following primary antibodies overnight at 4°C; anti-DSTN (1:200, GV-13, Sigma-Aldrich), anti-SM α-actin (1:500, Sigma-Aldrich), and anti-PECAM (1:200, BD Bioscience). The specificity of the GV-13 DSTN antibody was confirmed by Western blot in wild-type and DSTN-deficient SMCs. Only a single band was observed even after relative long exposures. This band was of the correct size (17 kDa) and was almost completely abolished by multiple siRNAs targeting DSTN. The appropriate secondary antibody only control experiments were also performed and we did not observe significant staining under any circumstances. Images were acquired with an Olympus BX61 fluorescence microscope using the ×20/0.50 UPlanFLN objective. The settings of exposure time and light source gain were identical between uninjured and injured carotid vessels. To determine DSTN and SM α-actin expression in the SMC media, ImageJ software (NIH) was used to determine medial area (between the inner and outer elastic laminae) and intensity of DSTN and SM α-actin fluorescence. Medial thickness was determined by averaging six independent measurements per artery. Image analyses included neck sections from four separate mice.

MRTF-A Localization

To visualize endogenous MRTF-A in 10T1/2 cells, cells were fixed with 4% PFA and permeabilized with 0.5% Triton X-100 in PBS. Slides were incubated with primary antibodies against MRTF-A (1:250, Santa Cruz) overnight at 4°C. Alexa Fluor488 (1:500, Thermo Fisher Scientific) and DAPI were added for 1 h at room temperature. Cells were imaged on an Olympus BX61 wide field microscope using the ×20/0.50 UPlanFLN objective. MRTF-A localization was scored as nuclear or diffuse in control and DSTN-deficient cells from two separate experiments. For each condition, at least 300 cells were scored from six random fields.

Quantitative PCR

RNA was isolated from cells using the Quick-RNA MiniPrep Kit (Zymo Research) and treated with DNase (Zymo Research) to eliminate contaminating genomic DNA. RNA was converted to cDNA using iScript cDNA synthesis Kit (Bio-Rad). cDNA (30 ng) was used for quantitative PCR. Primers used were mouse SMA, 5′- CGAAGCCGGCCTTACAGAG, 3′- CGCTGTCAGGAACC- CTGAGA; mouse DSTN, 5′- CTTTGTATGACGCCAGCTTTG, 3′- CTAGATACGT- TCGAGCTTCCTA; mouse GAPDH, 5′- ATGGGTGTGAACCACGAGAA, 3′-GG- CATGGACTGTGGTCATGA. GAPDH serves as the reference gene.

Western Analysis

Cells were lysed in RIPA buffer (25 mM Tris-Cl pH 7.2, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS) with protease and phosphatase inhibitors (Thermo Fisher Scientific). Protein concentration was determined by BCA assay (Thermo Fisher Scientific). Protein samples were resolved by 10% or 12% SDS-PAGE and then detected by immunoblotting. Nitrocellulose membranes were incubated with primary antibodies against DSTN (1:2,000, AE-14, Sigma-Aldrich), SM MHC (1:2,000, Abcam), SM α-actin (1:5,000 or 1:1,000, Sigma-Aldrich), CNN1 (1:3,000, Abcam), SM22 (1:500, Santa Cruz), anti-PECAM (1:500, Thermo Fisher Scientific), cyclin E (1:1,000, Santa Cruz), RBPJ (1:2,000, Cell Signaling), and α-Tubulin (1:3,000, Sigma-Aldrich). After incubation with horseradish peroxidase-conjugated secondary antibodies (GE Healthcare), ECL chemiluminescence (Thermo Fisher Scientific) was used to visualize protein bands. Protein expression from at least three separate Western blots was quantified by densitometry using ImageJ software (NIH) and is expressed relative to α-Tubulin as a loading control. Representative blots are shown in all figures.

Statistics

All data represent at least three independent experiments presented as means ± SE. Means were compared by Student’s t test, and statistical significance was considered as a P value less than 0.05.

Study Approval

All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of North Carolina at Chapel Hill. All animals were housed in facilities accredited by Association for Assessment and Accreditation of Laboratory Care International.

RESULTS

DSTN Is Strongly Expressed in Vascular SMCs and Is Downregulated by Vascular Injury

To further characterize the mechanisms that drive SMC-selective gene expression, we have been employing genome-wide analyses of chromatin accessibility and transcription factor binding in human AoSMCs. These approaches have increased our understanding of the control of SMC-specific gene expression and have led to the identification of previously undescribed DNA elements that have SMC-selective regulatory activity (14). Given our previous demonstration that RhoA signaling is critical for the regulation of SMC phenotype, we were particularly intrigued by the identification of several DNase hypersensitive regions near the DSTN gene that were not present in 9 other non-SMC cell types from the ENCODE database (Fig. 1A). When coupled with our ChIP-seq data showing strong SRF and RBPJ binding at the DSTN TSS (11), we hypothesized that DSTN may show SMC-selective expression and be regulated in a manner similar to other SMC-differentiation marker genes.

Figure 1.

Destrin (DSTN) was strongly and selectively expressed in vascular smooth muscle cell (SMC). A: schematic presentation of the DSTN promoter region that summarizes our DNase hypersensitivity (DHS) data, serum response factor (SRF) and RBPJ binding, sequence conservation, and H3K27 acetylation. B: aorta, heart, and kidney from 1-mo-old C57/Black6 mice were sectioned and stained with specific antibodies against DSTN (green), SM α-actin (SMA) (red), and DAPI (blue). Scale bars: 50 μm for aorta and 100 μm for kidney and heart. C: high-power magnification of a kidney blood vessel stained with antibodies against DSTN (green) and platelet endothelial cell adhesion molecule (PECAM) (red). Scale bar: 50 μm. D: Western blot analysis of DSTN and SMC differentiation marker gene expression in mouse aortic SMCs (AoSMCs) and endothelial cells (ECs). CNN1, calponin1; SM MHC, smooth muscle myosin heavy chain; SM22, transgelin.

To examine the localization of DSTN in vivo, we performed immunohistochemistry on SMC-containing tissues and vessels in mice. Importantly, the antibody used in these experiments does not cross-react with the closely related DSTN family members cofilin 1 (CFL1), which is ubiquitously expressed or cofilin 2 (CFL2), which is highly expressed in cardiac and skeletal muscle (20, 21). As shown in Fig. 1B, DSTN staining was strongly and selectively expressed in vascular smooth muscle of the aorta, heart, and kidney, as evidenced by coexpression with SM α-actin (SMA). We did not observe DSTN expression in endothelial cells (ECs) costained for platelet endothelial cell adhesion molecule (PECAM) (Fig. 1C), although we did see trace amounts in cultured mouse ECs (Fig. 1D). Our characterization of DSTN expression in mouse was in good agreement with DSTN expression patterns observed in humans, as reported by the genotype-tissue expression (GTEx) consortium (Supplemental Fig. S1; all Supplemental material is available at https://doi.org/10.6084/m9.figshare.14226266).

It is well known that vascular injury promotes the phenotypic modulation of SMC, as characterized by a reduction in SMC marker gene expression and increased cell proliferation, migration, and matrix synthesis (22). To test whether DSTN, like other SMC markers, is downregulated by vessel injury, we subjected adult mice to carotid artery ligation and measured DSTN expression by immunohistochemistry in control and injured arteries 2 wk after ligation surgery. As in the aorta, DSTN expression in the uninjured carotid artery was strong and completely overlapped with SMA expression in the vessel media (Fig. 2A). As quantified in Fig. 2B, both SMA and DSTN expression were significantly lower in ligated vessels, which exhibited a 2.5-fold increase medial thickening (Supplemental Fig. S2). As PDGF-BB plays a crucial role in suppressing the expression of SMC contractile markers during injury (22), we next tested whether DSTN was downregulated by this SMC mitogen. As shown in Fig. 2C, DSTN protein level was significantly downregulated in AoSMCs treated with PDGF-BB along with other SMC differentiation markers. Taken together, these data indicate that DSTN exhibits strong and at least partially selective expression in differentiated vascular SMCs.

Figure 2.

Destrin (DSTN) expression was downregulated by vessel injury and PDGF-BB. A: 2 wk after subjecting C57Black6 mice to carotid artery ligation surgery, injured and control vessels were sectioned and stained for SM α-actin (SMA) (red) or DSTN (green), as indicated. Elastic fibers (white) bracket the medial SMC layer. Scale bar: 25 μm. B: DSTN fluorescence intensity within the SMA positive medial layer was quantified from representative images and is expressed as fold over uninjured. Data represent means ± SE of n = 4. **P < 0.01. C: rat aortic SMCs (AoSMCs) were serum starved for 24 h and then treated with PDGF-BB (10 ng/mL) for 24 h. Cell lysates were examined for the expression of DSTN and SMC differentiation marker genes. Quantified data of three independent experiments shown at right. Data represent means ± SE. **P < 0.01.

DSTN Promoter Elements Have Strong SMC-Selective Activity

To begin to test whether DSTN expression is transcriptionally regulated in a manner similar to that of other SMC differentiation markers, we identified potential regulatory elements in the DSTN gene using several criteria including sequence conservation, DNase hypersensitivity measurements of open chromatin, binding of SRF and RBPJ, and proximity to the DSTN TSS. As shown in Fig. 3 and Supplemental Fig. S3, several regulatory regions were PCR amplified from human genomic DNA, cloned into the luciferase reporter vectors, and then tested for regulatory activity in SMCs and ECs. As shown in Fig. 3A, a fragment of the DSTN proximal promoter from −1 kb through +1 kb had strong activity in human BrSMCs (∼26-fold over the promoterless) but significantly less activity in ECs. Moreover, the DNase hypersensitivity (DHS) region just 5′ to the TSS that contains a stretch of highly conserved sequence exhibited remarkably strong SMC-selective activity on its own (∼163-fold over promoterless) suggesting that it likely functions as a SMC enhancer. Results from additional deletion analyses (Supplemental Fig. S3A) indicated that the immediate DSTN TSS from −500 to +250 had very high activity in both cell types (∼758-fold and ∼110-fold over promoterless for SMC and EC, respectively) suggesting that it functions as more of a basal promoter, and that the DSTN first intron contains strong repressor activity that likely contributes to the SMC specificity of the proximal DSTN promoter region. The transcriptional activities measured in our human BrSMC model were very similar to those measured in rat AoSMCs (Supplemental Fig. S3B). Lastly, we tested the activity of two upstream elements that were shown to be DNase hypersensitive in human AoSMCs but not in other cell types. When cloned upstream of the DSTN proximal promoter, the DHS element at −10 kb, which was shown to bind SRF by ChIP seq, significantly enhanced DSTN promoter activity, whereas the DHS element at −25 Kb, which was shown to bind RBPJ, inhibited DSTN promoter activity (Fig. 3B).

Figure 3.

A destrin (DSTN) promoter region displayed strong and smooth muscle cell (SMC)-selective activity. A: the DSTN promoter fragments shown on the left were PCR amplified from human genomic DNA, cloned into luciferase vector, and then transfected into bronchial SMCs (BrSMCs) and endothelial cells (ECs). Luciferase assays were performed after 48 h, and transcriptional activity was normalized to the promotorless pGL3 vector. The sequence shown represents the conserved region marked by an asterisk and potential transcription factor binding sites are boxed and labeled. B: the SMC-selective DNase hypersensitivity (DHS) regions depicted were cloned upstream of the DSTN proximal promoter and then transfected into BrSMCs. Luciferase activity was measured 48 h later and is expressed as fold over promotorless pGL3 vector. Data represent means ± SE of n = 3. *P < 0.05, **P < 0.01.

DSTN Functions as an Autoregulatory Negative Feedback Inhibitor of SRF-Dependent SMC Differentiation Marker Gene Expression

We and others have shown that most SMC differentiation marker genes are regulated by SRF binding to multiple CArG elements within their promoters and that RhoA-dependent activation of the SRF cofactors, MRTF-A and MRTF-B, plays an important role in maintaining high levels of SMC differentiation marker gene expression (2, 16, 23). Importantly, a highly conserved relatively low-affinity CArG element was identified near the center of the DSTN promoter region that exhibited very high SMC-selective transcriptional activity (Fig. 3A). This CArG was also within the SRF binding peak that we detected just 5′ to the DSTN TSS by our ChIP seq analyses in human AoSMCs. As shown in Fig. 4A, mutation of this CArG element (but not a second nonconserved CArG further upstream) significantly reduced the transcriptional activity of the SMC-selective enhancer fragment, whereas targeted ChIP assays using PCR primers that flank the conserved CArG (+/−50 bp) provided further evidence that SRF bound to this region (Fig. 4B). As discussed more fully below, we also detected binding of the Notch transcription mediator, RBPJ, to this region. Further supporting the involvement of SRF, endogenous DSTN expression was significantly upregulated by the strong RhoA agonist, S1P, in human BrSMCs (Fig. 4C) and in 10T1/2 SMC precursors (Supplemental Fig. S4, A and B) and the activity of the SMC-selective enhancer was similarly affected (Figs. 4D and Supplemental S4B). In addition, overexpression of myocardin strongly increased the activity of the SMC-selective enhancer and this effect was substantially diminished by mutation of the conserved CArG element (Fig. 4E).

Figure 4.

Destrin (DSTN) negatively regulated smooth muscle cell (SMC) differentiation marker gene expression. A: the indicated CArG box mutations were made in the context of the SMC enhancer-luciferase construct, which were then transfected into bronchial SMCs (BrSMCs). Luciferase activity was measured 48 h later and is expressed as fold over promoterless pGL3 vector. Data represent means ± SE of n = 3. **P < 0.01. B: targeted ChIP assays were performed in BrSMCs using antibodies against serum response factor (SRF) and RBPJ and PCR primers spanning the conserved region of the DSTN SMC-selective enhancer. Representative image from three independent experiments. C: BrSMCs were serum starved for 24 h and then treated with S1P (1 μM). DSTN expression was analyzed after 48 h by Western blot. Quantified data on the right represent means ± SE of n = 3. **P < 0.01. D: BrSMCs were transfected with the SMC-selective enhancer-luciferase construct, serum starved for 24 h, and treated with S1P (1 μM). Luciferase activity was measured 24 h later and is expressed as fold over promotorless pGL3 vector. Data represent means ± SE of n = 3. **P < 0.01. E: the indicated promoter-luciferase construct was cotransfected into BrSMCs along with FLAG-myocardin. Luciferase activity was measured 48 h later and is expressed as fold over plus empty FLAG expression vector. Data represent means ± SE of n = 3. **P < 0.01, *P < 0.05. TK, minimal thymidine kinase promoter. F: 10T1/2 cells were treated with siRNA targeting DSTN or green fluorescent protein (GFP) for 72 h. Protein lysates were analyzed by Western blot using the indicate antibodies. In quantification below, data represent means ± SE. n = 3. **P < 0.01. G: 10T1/2 cells were treated with siRNA against DSTN or GFP for 48 h. Localization of endogenous myocardin-related transcription factor-A (MRTF-A) was visualized by immunofluorescence staining and representative images shown. MRTF-A localization was scored as diffuse or nuclear (colocalized with DAPI) in 349 GFP control and 318 DSTN-depleted cells from multiple random fields. Scale bar: 50 μm. **P < 0.01.

Although DSTN expression is upregulated by RhoA signaling, DSTN’s ability to sever actin filaments and increase G-actin pools would be expected to decrease MRTF/SRF-dependent gene activation. Indeed, examination of a spontaneous DSTN null mouse line by Ikeda et al. (15) revealed increased expression of many SRF-dependent genes including SM α-actin in the corneal epithelium (15). Although the transcriptional effects of DSTN depletion observed by this group were likely mediated by enhanced MRTF nuclear localization, this mechanism was not directly examined. When coupled with our current demonstration that high DSTN expression in vascular SMCs was mediated by RhoA/SRF signaling, we hypothesized that DSTN functions as an autoregulatory negative feedback inhibitor of SMC differentiation. In support of this idea, we observed a significant increase in SM MHC, SMA, and SM22 protein expression in 10T1/2 SMC precursor cells and human BrSMCs in which DSTN was depleted by siRNA (Fig. 4F and Supplemental Fig. S5). DSTN depletion also significantly increased MRTF-A nuclear localization strongly implicating this mechanism (Fig. 4G).

DSTN Is Required for Proper SMC Proliferation and Migration

As actin severing and actin remodeling in general are required for proper cell division and cell migration, we next tested whether DSTN plays a role in regulating these important SMC processes. As shown in Fig. 5, A and B, total cell number and levels of the proliferative marker, cyclin E, were significantly decreased in AoSMCs treated with DSTN siRNA. In addition, DSTN depletion also significantly inhibited SMC migration, as measured in scratch would assays (Fig. 5C).

Figure 5.

Destrin (DSTN) depleted smooth muscle cells (SMCs) display defects in cell proliferation and migration. A: human aortic SMCs (AoSMCs) were treated with DSTN or GFP siRNAs, serum starved for 24 h, and then treated with 10% serum. Cell proliferation was measured by cell counting at each time point. Data represent means ± SE of n = 3. **P < 0.01. B: rat AoSMCs were treated with siRNA targeting DSTN or GFP for 72 h. Lysates were harvested and cyclin E protein level were measured by Western blot. Quantified data represent means ± SE of n = 3. **P < 0.01. C: confluent monolayers of control (siGFP) or DSTN-depleted rat AoSMCs were used for scratch wound assays. After 24 h, migration into the scratched area was measured by microscopy. Migration area of siGFP and siDSTN was analyzed by 12 independent images from two sets of siRNA treatment. **P < 0.01.

DSTN Expression in SMC Is Regulated by TGF-β and Notch Signaling

The data presented so far suggest that DSTN promotes several important aspects of SMC phenotypic modulation making it critical to identify the mechanisms that control its expression in addition to the RhoA/MRTF/SRF pathway. Sequence analysis of the SMC-selective enhancer just 5′ to the DSTN TSS identified potential binding elements for SMAD and RBPJ in close proximity to the functional CArG element within this region (Fig. 3A) and mutation of either of these elements inhibited the regulatory activity of the SMC-selective enhancer fragment (Fig. 6A). We and others have shown that many SMC-specific promoters have a similar arrangement of SRF-, SMAD-, and RBPJ-binding sites suggesting that cooperation between these transcription factor pathways is important for SMC-selective gene expression (11, 24–26). Indeed, TGF-β treatment significantly increased endogenous DSTN expression in human BrSMCs (Fig. 6B) and in 10T1/2 cells (Supplemental Fig. S4A), as well as the activity of the SMC-selective enhancer fragment in both cell lines (Figs. 6C and Supplemental Fig. S4D), confirming an important role for SMAD binding.

Figure 6.

Destrin (DSTN) expression was upregulated by both transforming growth factor β (TGF-β) and Notch signaling. A: the indicated RBPJ and SMAD binding site mutations were made in the context of the smooth muscle cell (SMC) enhancer-luciferase construct. Constructs were transfected into bronchial SMCs (BrSMCs) and relative luciferase activity is expressed as fold over promoterless vector. Data represent means ± SE of n = 3. **P < 0.01, *P < 0.05. B: BrSMCs were serum starved for 24 h and then treated with TGF-β for 48 h. DSTN expression was measured by Western blot and quantified data represent means ± SE of n = 3. *P < 0.05. C: human BrSMCs were transfected with the SMC enhancer-luciferase construct, serum starved for 24 h, treated with TGF-β (2 ng/mL), and then lysed for luciferase assays after 24 h. Luciferase activity is expressed as fold over promotorless pGL3 vector and quantified data represent means ± SE, n = 3. **P < 0.01. D: Western blot analysis of DSTN and SMC differentiation marker gene expression in BrSMCs treated with vehicle (DMSO) or DAPT (10 μM) for 24 h. Quantified data represent means ± SE of n = 4. **P < 0.01. E: rat AoSMCs were transfected with the indicated promoter-luciferase construct, treated with DAPT (5 μM) or DMSO for 24 h, and then assayed for luciferase activity. Data represent means ± SE of n = 4. **P < 0.01. SMA, smooth muscle α-actin; TK, minimal thymidine kinase. F: human BrSMCs were treated with DAPT for 18 h and then subjected to ChIP assays using indicated antibodies and PCR primers spanning the DSTN SMC-selective enhancer. Representative image from three independent experiments. G: human BrSMCs were treated with siRNA targeting RBPJ and GFP for 72 h. Lysates were analyzed by Western blot using the indicated antibodies. Quantified data represent means ± SE of n = 4. **P < 0.01, *P < 0.05.

To further investigate the role of Notch signaling, we treated human BrSMCs with the γ-secretase inhibitor DAPT that prevents ligand-dependent Notch receptor cleavage. As shown in Fig. 6D, DSTN protein levels were significantly reduced by DAPT treatment. Interestingly, DSTN was less sensitive to DAPT-induced downregulation than SM α-actin and SM22 but more sensitive than SM MHC and CNN1, which were not significantly affected in this model (Fig. 6D). DAPT treatment also inhibited the transcriptional activity of the SMC-selective enhancer fragment suggesting that the effects of DAPT were at least partially transcriptionally mediated (Fig. 6E). Further supporting this idea, targeted ChIP assays showed that DAPT treatment inhibited NICD3 recruitment to the SMC-selective DSTN enhancer without significantly affecting RBPJ or SRF binding to the enhancer (Fig. 6F).

Although Notch signaling clearly plays a role in the activation of DSTN and other SMC marker genes (26, 27), we have recently shown that RBPJ acts as a repressor of SMC-selective gene expression in the absence of active Notch signaling. Under these conditions, RBPJ inhibits SRF binding to nearby CArG elements and depletion of RBPJ “derepresses” SMC marker gene expression. To test the importance of this mechanism, we depleted RBPJ in human BrSMCs by siRNA and examined DSTN expression by Western blot. As shown in Fig. 6G, DSTN expression was upregulated by twofold in RBPJ depleted cells. As with DAPT, DSTN was less sensitive to RBPJ depletion than SM α-actin or SM22 but more sensitive than SM MHC (Fig. 6G), indicating that the effects of Notch signaling on SMC-selective gene expression are both context and gene dependent.

DISCUSSION

SMC phenotype is regulated by a complex array of environmental cues and signaling pathways that ultimately control the transcription factor networks that drive SMC-specific gene expression (1). Results from the current study suggest that the actin severing protein, DSTN, is highly expressed in SMC and is regulated by mechanisms that are similar to those that drive SMC differentiation marker gene expression. Using our chromatin structure and transcription factor binding data sets, we identified and characterized several regulatory regions that drive DSTN expression in SMC including a SMC-selective enhancer element near the DSTN TSS. Our demonstration that DSTN depletion upregulated SMC marker gene expression, whereas inhibiting SMC migration and proliferation suggests, for the first time, that DSTN plays an important and perhaps coordinating the role in SMC phenotypic modulation.

Our observation that the SMC-selective enhancer was SRF dependent is in excellent agreement with a previous study by Sun et al. (28) who showed that the mouse sequence containing this CArG element could bind SRF and was transactivated by an SRF-VP16 fusion protein. The SMC-selective enhancer and the endogenous DSTN gene were upregulated by TGF-β, and the SMAD-binding element within the SMC enhancer just upstream of the CArG box likely plays a role in this response. These results are similar to those observed with other SMC-specific promoters (25, 29) and may reflect previously described interactions between SRF and SMAD transcription factors (24). The SMC-selective enhancer and endogenous DSTN expression were also Notch dependent, a result consistent with our previous demonstration that RBPJ and SRF bind to similar genomic regions in human AoSMCs (11) and data from our laboratory and others that SRF interacts with Notch to regulate SMC-specific gene expression (18, 30). The current studies also support our previous finding that RBPJ can act as a repressor of SRF-dependent gene expression under conditions of low Notch activity (11, 12).

Given that DSTN depletion promoted MRTF-A nuclear localization and endogenous SMC differentiation marker gene expression, our results are consistent with the idea that DSTN’s effects on SMC-specific promoter activity are mediated by actin polymerization-dependent control of MRTF nuclear localization. We have recently shown that the RhoA-specific GAP, GRAF3, is also selectively expressed in SMC by RhoA/SRF-dependent mechanisms (13, 14). Thus, it is interesting to postulate that the expression of unique upstream and downstream inhibitors of RhoA signaling is critical for maintaining SMC plasticity. We postulate that the downregulation of DSTN following vessel injury is mediated, at least in part, by PDGF-BB, which is known to be increased under these conditions (31) and that significantly downregulated DSTN expression in our studies. As originally described by Wang et al. (32), DSTN downregulation likely involves PDGF-BB-dependent activation of the ternary complex factor (TCF) family of SRF coactivators that competitively inhibit myocardin factor binding to SRF. In regard to our measurements of DSTN expression following vessel injury, it is difficult to completely rule out the contributions of a potentially DSTN negative progenitor cell population such as the Sca1 positive medial cells detected by Dobnikar et al. (33). However, we did not observe Sca1 positive cells in our model, as measured by immunohistochemistry (data not shown), and DSTN downregulation was fairly uniform.

Our results support a model in which the activation of highly expressed DSTN allows for potential disruption of the feed forward RhoA/actin/MRTF/SRF signaling transcription loop that helps maintain SMC differentiation. We would expect that DSTN deficiency would lead to decreased intima formation and vessel thickness following injury. However, given that DSTN is also an SRF-dependent gene, we further postulate that this effect would be more prominent during the initial stages of injury when DSTN expression is still strong. It is also possible that DSTN’s initial role in SMC phenotypic modulation would eventually be overcome by a major injury stimulus that strongly downregulated SMC differentiation marker gene expression.

Although our data clearly indicate that DSTN has a nonredundant function in SMC, vascular phenotypes have not been reported in the DSTN null mutation mouse line (15). To our knowledge, however, this line has not been specifically examined for SMC defects, and CFL1 compensation has been shown to limit the effects of DSTN depletion (34, 35). Indeed, we observed significant upregulation of CFL1 activation (as measured by decreased Ser3 phosphorylation) in DSTN-deficient SMCs (data not shown), which likely limited the effects of DSTN depletion in our studies.

Dynamic regulation of the actin cytoskeleton is critical for SMC migration, and the effects of DSTN depletion on this process are likely mediated by defective remodeling of existing actin structures. Although DSTN mRNA levels in human arteries are much higher (Supplemental Fig. S1), most studies that have examined ADF function in SMC have focused on CFL1 (36, 37). Separate groups have shown that CFL1 is activated by PDGF-BB, that SSH1 and CFL1 were upregulated by vessel injury (38), and that siRNA-mediated depletion of CFL1 or one of several slingshot phosphatases inhibited SMC migration (38, 39). CFL1 was also shown to be activated in a SMC-specific Fibulin4 knockout model of aortic aneurysm and inhibition of SSH1 upregulation attenuated this effect (40). The fact that CFL1 is upregulated by PDGF-BB and vessel injury (38, 39, 41), whereas DSTN is downregulated by these same cues strongly suggests that these ADFs have specific functions and that their expression levels are maintained by different mechanisms. It is also possible that a switch from DSTN to CFL1 signaling may be important for maintaining SMC phenotypic modulation. It is also interesting that CFL1 is ubiquitously expressed while DSTN and CFL2 are enriched in smooth and striated muscle, respectively (20, 21, 42), perhaps suggesting that CFL2 and DSTN are important for the maintenance and/or remodeling of contractile actin fibers. Finally, a recent study demonstrated that the ADFs bind to G- and F-actin with equal affinities, but that DSTN and CFL2 have more efficient actin severing ability (43).

In conclusion, our data indicate that high expression of the actin severing protein, DSTN, in SMC is mediated by SMC-selective transcription mechanisms and that DSTN protein promotes SMC phenotypic modulation. A closer examination of DSTN and CFL1 expression and activity in SMC during development is warranted as are additional studies to determine whether DSTN contributes to neointima formation and/or SMC contractility.

SUPPLEMENTAL DATA

Supplemental Figs S1–S5: https://doi.org/10.6084/m9.figshare.14226266.

GRANTS

This work was supported by National Institutes of Health grants HL109607 (to C. P. Mack) and HL130367 (to J. M. Taylor and C. P. Mack). The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund (http://commonfund.nih.gov/GTEx) of the Office of the Director of the National Institutes of Health and by National Cancer Institute, National Human Genome Research Institute, National Heart, Lung, and Blood Institute, National Institute on Drug Abuse, National Institute of Mental Health, and National Institute of Neurological Disorders and Stroke.

DISCLOSURES

The data used for the analyses described in this manuscript were obtained from the GTEx Portal (http://www.gtexportal.org) on 11/08/2020. No conflicts of interests, financial or otherwise, are declared by authors.

AUTHOR CONTRIBUTIONS

K.A.L., X.B., and C.P.M. conceived and designed research; K.A.L., K.V.R., X.B., J.M.T., and C.P.M. performed experiments; K.A.L., X.B., J.M.T., and C.P.M. analyzed data; K.A.L., K.V.R., X.B., J.M.T., and C.P.M. interpreted results of experiments; K.A.L., X.B., and C.P.M. prepared figures; K.A.L., X.B., and C.P.M. drafted manuscript; K.A.L., X.B., J.M.T., and C.P.M. edited and revised manuscript; K.A.L., X.B., and C.P.M. approved final version of manuscript.