Transforming Growth Factor-β Promotes Recruitment of Bone Marrow Cells and Bone Marrow-derived Mesenchymal Stem Cells through Stimulation of MCP-1 Production in Vascular Smooth Muscle Cells^*

General Materials

Recombinant TGF-β1 and rat MCP-1 were obtained from R&D Systems. Dulbecco's modified Eagle's medium (DMEM) and cell culture reagents were from Invitrogen. Other reagents, if not specified, were purchased from Sigma.

Smooth Muscle Cell Culture

Rat aortic SMC were isolated from the thoracic aorta of male Sprague-Dawley rats based on a protocol described by Clowes et al. (23) and maintained in DMEM containing 10% fetal FBS at 37 °C with 5% CO₂. Cell viability was assayed by the trypan blue exclusion method, which indicated that <5% of the cells took up the dye both before and after the infection of adenoviral vectors or treatment with recombinant TGF-β1 or inhibitors. For experiments using conditioned media, RSMC (2 × 10⁵ cells/well in 6-well plates) were infected with adenovirus (30,000 particles/cell in DMEM containing 2% FBS for 4 h at 37 °C) followed by starvation in DMEM containing 0.5% FBS for 48 h. The cells were then treated with recombinant TGF-β1 (5 ng/ml) or solvent (4 mm HCl with 2% bovine serum albumin) for 6 h. Media were collected and used for ELISA or chemotaxis assays as described below.

Rat Mesenchymal Stem Cell Culture

Rat MSC isolated from rat bone marrow were obtained from Cell Applications Inc. (San Diego, CA). The cells were maintained through the use of mesenchymal stem cell growth medium kit (Cell Applications Inc.). MSC populations were passaged 1–4 times. Cells were harvested from plates using a solution containing 0.25% trypsin and 1 mm EDTA (Invitrogen). Confirmation of MSC was made through the presence of cell surface markers such as CD90, CD44, and CD54 and the absence of CD31 and CD45. For cell differentiation experiments, MSC up to passage 4 were plated in MSC growth medium (non-differentiating).

Isolation of Rat Bone Marrow

Bone marrow from Sprague-Dawley rats was isolated by flushing the long bones with Dulbecco's phosphate-buffered saline containing 2% bovine serum albumin, and heparin (1000 units/ml). Cells from individual animals were pooled. Bone marrow was dissociated by aspiration with a Pasteur pipette. Red blood cells were excluded using Ficoll-Hypaque (GE Healthcare) density centrifugation. The remaining cells were then rinsed twice in phosphate-buffered saline followed by centrifugation.

Construction of Adenoviral Vectors and Infection

Adenoviral vectors expressing Smad3 (AdSmad3), PKCδ (AdPKCδ), and GFP (AdGFP) were constructed as previously described (2).

Cell Migration Assay

2 × 10⁵ bone marrow cells or 2 × 10⁵ MSC were placed in the upper chamber of Costar 24-well transwell plates with 5-μm pore filters (Corning, Inc., Corning, NY). Cultured conditioned medium was placed into the lower chambers or wells. After incubating plates for 6 h at 37 °C, migrated cells were collected from the lower chambers and counted.

siRNA Knockdown of MCP-1 and Smad3

RSMC were plated at 50–60% confluence in DMEM culture medium in 6-well plates and incubated for 24 h. Cells were then transfected in Opti-MEM I medium with 20 nmol of siRNA for rat MCP-1 (Santa Cruz Biotechnology), 100 pmol of siRNA for Smad3 (Invitrogen), or control siRNA using RNAiMax transfection reagent (Invitrogen), as described by the manufacturer's protocol. After 6 h, the Opti-MEM I medium was replaced by DMEM containing 0.5% FBS, and cells were starved for 48 h. Cells that were infected with adenovirus were infected the day after transfection, then starved for the remaining 48 h. The cells were then stimulated with recombinant TGF-β1 (5 ng/ml) or solvent for 6 h, after which conditioned media was collected for chemotaxis assays and MCP-1 ELISA, and cells were lysed for protein used in Western blotting.

Western Blot Analysis

Cells were lysed in RIPA buffer (50 mm Tris, 150 mm NaCl, 1% Nonidet P-40, 0.1% sodium dodecyl sulfate, and 10 μg/ml aprotinin). Fifty micrograms of protein from each sample were separated on 10% SDS-PAGE gels. The protein samples were then transferred to nitrocellulose membranes. Protein expression was confirmed by immunoblotting with the following antibodies: rabbit anti-Smad3 (AbCam), rabbit anti-pSmad3 (AbCam), and mouse anti-β-actin (Sigma). After incubation with the appropriate primary and horseradish peroxidase-conjugated secondary antibodies, the membranes were developed with enhanced chemiluminescence reagent (Amersham Biosciences).

Microarray Analysis

RSMC were infected with AdSmad3 or AdGFP followed by starvation for 48 h in DMEM containing 0.5% FBS, then treated with recombinant TGF-β1 (5 ng/ml) or solvent for 6 h. Cells were harvested, and total RNA was extracted by RNeasy Mini Kit (Qiagen, Valencia, CA). This RNA was then processed according to the protocol recommended by Affymetrix using the Superscript Choice kit (Invitrogen) for double-stranded cDNA synthesis and the Enzo Bioarray kit for in vitro transcription and labeling of cRNA. 15-μg samples of fragmented cRNA were hybridized for 16 h at 45 °C to rat 230.2.0 arrays (Affymetrix). Data analysis was performed using GeneSpring^TM software 5.1 (Silicon Genetics) for normalization, -fold change calculations, and clustering.

RT-PCR

Expression of the mRNA of MCP-1 was detected by RT-PCR. RNA was isolated from rat SMC using a kit from Qiagen. A segment of the cDNA was detected using RT-PCR for MCP-1 using MCP-1 primers, which were obtained from Santa Cruz (Santa Cruz Biotechnology).

ELISA for MCP-1-ELISA to detect MCP-1 secreted by RSMC was performed using rat MCP-1 ELISA kit (BD Biosciences). RSMC were cultured at a density of 1 × 10⁵/ml in 1 ml of complete medium in the presence or absence of different stimuli in 6-well plates (Costar). After incubation for specified periods of time at 37 °C, cell-free culture supernatants were obtained. The concentrations of MCP-1 were then measured according to the manufacturer's instructions.

Rat Balloon Injury Model and in Vivo Gene Delivery

Male Sprague-Dawley rats (450–500 g) underwent balloon injury of the left common carotid artery as described elsewhere in accordance with institutional guidelines and approval (24, 25). Briefly, after induction of anesthesia with isoflurane, a 2-French balloon catheter was inserted through the left external carotid artery into the common carotid and insufflated with 1 ml of air three times. Recombinant adenoviral vectors were constructed to express Smad3 as previously described; an empty viral vector was used as a control (26). After injury, animals received intraluminal administration of adenoviral vectors as indicated (2.5 × 10⁹ plaque-forming unit in 200 μl of serum-free media over 20 min). The external carotid artery was then ligated, and flow was re-established through the common carotid and internal carotid arteries.

Rats were sacrificed 5 days after injury, and perfusion was fixed with 4% paraformaldehyde diluted in phosphate-buffered saline at a pressure of 100 mm Hg for 15 min. The left and right carotid arteries were then harvested and fixed in 4% paraformaldehyde for an additional 30 min. The arteries were then paraffin-embedded and cut into 7-μm sections for histological analysis.

Bone Marrow Labeling in Vitro and Implantation

Bone marrow cells were extracted from rat long bones and labeled with fluorescent dyes using a PKH26 red fluorescent cell linker kit (Sigma-Aldrich). After left carotid balloon angioplasty was performed, a 24-gauge angiocath was inserted into the external carotid arteriotomy and secured. 5 × 10⁷ BM suspended in 1 ml phosphate-buffered saline were injected systemically through the arteriotomy. The external carotid artery was then ligated, and flow was re-established from the common carotid artery to the internal carotid artery. Rats were sacrificed 48 h after BM cell injection, and the tissue was frozen into blocks with optimal cutting temperature medium (Sakura Finetek Inc.).

Immunohistochemistry

Immunostaining for MCP-1 using monoclonal anti-MCP-1 antibody (Santa Cruz) was performed as described previously (24). Slides were visualized with a Nikon Eclipse E800 upright microscope. Digital images were acquired using a RetigaEXi CCD digital camera and processed and analyzed using IPLab software. Semiquantitative immunohistochemical analysis was performed using Adobe Photoshop 7.0 and NIH imaging software (ImageJ) (24).

Statistical Analysis

Data are expressed as the mean ± S.E. Unpaired Student's t test was used to evaluate the statistical differences between control and treated groups. In cases of multiple groups, differences were first analyzed with one-way analysis of variance. Values of p < 0.05 were considered significant. All experiments were repeated at least in triplicate.

TGF-β Induces VSMC to Secrete Bone Marrow and Progenitor Cell Chemoattractants

To begin testing our hypothesis that TGF-β stimulates intimal hyperplasia through the recruitment of BMPC, we first established a bone marrow cell chemotaxis assay using SDF-1α as a positive control (Fig. 1A). We then evaluated whether recombinant TGF-β could directly stimulate either total bone marrow cell or MSC migration. TGF-β alone had no direct effect on progenitor cell migration (Fig. 1, A and B). We then asked whether TGF-β might stimulate VSMC to produce progenitor cell chemoattractants as it is well established that TGF-β levels are elevated after vascular injury. As shown in Fig. 1, C and D, conditioned media derived from rat aortic VSMC treated with TGF-β significantly increased both total bone marrow cell as well as isolated rat MSC chemotaxis. These findings suggest that TGF-β activates VSMC, which in turn produce a secreted factor that is responsible for recruitment or chemotaxis of bone marrow progenitor cells.

The Effect of TGF-β on Progenitor Cell Chemotaxis Is Mediated by Smad3

Smad3 is a major signaling molecule for TGF-β that we and others have shown to be important in mediating many of the TGF-β effects on VSMC. We, therefore, hypothesized that Smad3 might mediate the effect of TGF-β on progenitor cell recruitment. To examine whether Smad3 plays a role in TGF-β-induced secretion of BMPC chemoattractants by VSMC, we infected rat aortic VSMC with adenoviral vectors expressing Smad3 (AdSmad3) and then stimulated cells with recombinant TGF-β1, thus enhancing TGF-β/Smad3 signaling. As shown in Fig. 2, conditioned media from VSMC treated with AdSmad3 and TGF-β significantly increased both total bone marrow cell as well as isolated rat MSC migration, compared with conditioned media from VSMC treated with TGF-β alone.

To further study the role of Smad3 in this process, we examined the effect of inhibition of Smad3 signaling on BMPC recruitment. VMSC were pretreated with SB431542 (10 μg/ml), a TGF-β receptor kinase inhibitor that, at this concentration, is specific for Smad2/3 phosphorylation. As shown in Fig. 3A, treatment with SB431542 effectively inhibited TFG-β-induced phosphorylation of both endogenous and adenovirus-derived Smad3 but did not affect total expression of Smad3. VSMC treated with TGF-β or TGF-β/AdSmad3 were pretreated with SB431542, and the effect of this conditioned media on bone marrow cell and MSC migration was evaluated. As shown in Fig. 3, B and C, pretreatment of VSMC with SB431542 followed by stimulation with TGF-β or TGF-β/AdSmad3 resulted in a significant decrease in the chemotaxis of both bone marrow cells and MSC in response to conditioned media derived from VSMC. These data confirm that phosphorylation of Smad3 is an essential step in TGF-β mediated recruitment of progenitor cells by vascular SMC.

To more specifically inhibit Smad3 function we developed a siRNA to rat Smad3. As shown in Fig. 3D, transient transfection of VSMC with Smad3 siRNA effectively down-regulated Smad3 expression. Furthermore, conditioned media derived from VSMC transiently transfected with Smad3 siRNA followed by stimulation with TGF-β produced significantly decreased total BM and MSC migration compared with conditioned media from VSMC transfected with scrambled siRNA (Fig. 3, E and F). Taken together, these results demonstrate that Smad3 is the signaling protein through which TGF-β mediates its effect on progenitor cell recruitment.

Identification of the Progenitor Cell Chemoattractant Released by VSMC in Response to TGF-β and Smad3

Our next goal was to determine the identity of the progenitor cell chemoattractant released by VSMC in response to TGF-β. To explore the possibilities, we performed a microarray analysis of TGF-β/AdSmad3-stimulated VSMC. Treatment with TGF-β/AdSmad3 increased by more than 4-fold the expression of 566 genes and decreased expression by at least 4-fold of 284 genes. The selected genes were clustered into groups that influence cell differentiation or cell growth using GeneSpring^TM software. We were able to identify a number of known progenitor cell chemoattractants stimulated by TGF-β in Smad3-overexpressing cells, as summarized in Table 1. One of the most highly induced genes was that of MCP-1, with a greater than 80-fold induction. Because MCP-1 has been shown to attract neural progenitors cells to areas of tissue injury (27), we hypothesized that MCP-1 may be the critical factor that allows TGF-β-stimulated vascular SMC to function as a chemoattractant of bone marrow progenitor cells to sites of arterial injury.

TGF-β Induces VSMC Production of MCP-1 through Smad3

As shown in Fig. 4A, TGF-β stimulated the production of MCP-1 in RSMC in a dose-dependent manner. Smad3 overexpression via adenovirus-mediated gene transfer also increased the production of MCP-1, with further enhancement when Smad3-overexpressing cells were stimulated with TGF-β (Fig. 4B). To demonstrate that the increase in protein level was secondary to an increase in gene transcription, RT-PCR confirmed that TGF-β and Smad3 overexpression both increased expression of MCP-1 mRNA in VSMC (Fig. 4C).

We next investigated the impact of inhibiting Smad3 signaling on TGF-β-induced MCP-1. As shown in Fig. 4D, inhibition of Smad2/3 phosphorylation using SB431542 (10 μg/ml) significantly blocked TGF-β-induced MCP-1 in both control RSMC and RSMC overexpressing Smad3. Furthermore, Smad3 siRNA also significantly decreased TGF-β-induced MCP-1 expression in RSMC (Fig. 4E).

MCP-1 Is an Important Chemokine for the Recruitment of Bone Marrow-derived Cells

To determine the role of MCP-1 in the recruitment of bone marrow cells and MSC, we first tested the effect of recombinant MCP-1 on total bone marrow cell and MSC migration. Fig. 5, A and B, show that recombinant MCP-1 markedly induced both total bone marrow cell and MSC chemotaxis. To further examine whether MCP-1 is necessary for the recruitment of bone marrow cells by VSMC treated with TGF-β and AdSmad3, we used a siRNA to silence MCP-1 expression in these cells. As shown in Fig. 5C, MCP-1-specific siRNA significantly reduced the amount of MCP-1 released by SMC treated with TGF-β, whereas scrambled siRNA had no significant effect. Inhibition of MCP-1 gene expression by the siRNA was also confirmed by RT-PCR (Fig. 5D). Transfection of MCP-1 siRNA into RSMC followed by stimulation of these cells with TGF-β/AdSmad3 significantly decreased the chemoattractive property of the conditioned media to both bone marrow cells and MSC, in comparison to conditioned media from cells treated with scramble siRNA (Fig. 5, E and F). These findings were confirmed by repeating identical studies with conditioned media from VSMC pretreated with an MCP-1 blocking antibody (Fig. 5, G and H). Specific inhibition of MCP-1 by MCP-1 antibody resulted in a 26 and 21% reduction in the chemoattractive properties of VSMC-conditioned media for total bone marrow cells and MSC, respectively, thus suggesting that MCP-1 is not the only progenitor cell chemoattractant induced by TGF-β/Smad3. This, however, is consistent with our gene array data, which shows that TGF-β/Smad3 induces several other potential progenitor cell chemoattractants, including vascular endothelial growth factor, fibroblast growth factor-2, and c-kit ligand (stem cell factor). Together, these data provide convincing evidence that MCP-1 is one of the critical factors released by SMC that mediates TGF-β induced chemoattraction of bone marrow progenitor cells.

VSMC-conditioned Media Induces MSC to Express Smooth Muscle Cell Markers

In addition to a small percentage of progenitor cells, total bone marrow contains many cells that differentiate to the leukocyte and monocyte linage and are expected to be recruited by MCP-1. To show that MCP-1 not only induces progenitor cell chemotaxis but can also induce progenitor cell differentiation into smooth muscle-like cells, MSC were cultured with conditioned media derived from VSMC treated with or without TGF-β or AdSmad3. After 3 days of exposure to conditioned media, the morphology of treated MSC changed significantly to resemble SMC (Fig. 6, A–C). Cellular protein levels of α-smooth muscle actin (α-SMA), calponin, and desmin were determined by Western blotting. As shown in Fig. 6D, the expression levels of α-SMA, calponin, and desmin in MSC were significantly increased in response to treatment with conditioned media derived from VSMC treated with TGF-β and AdSmad3. TGF-β itself did not have a significant effect on VSMC morphology or on expression of α-SMA or calponin (Fig. 6E).

Smad3 Overexpression in Vivo Increases MCP-1 Expression by VSMC after Injury

We next turned to the rat carotid balloon injury model to test our hypothesis in vivo that TGF-β through Smad3 signaling stimulates SMC production of MCP-1. Smad3 was overexpressed via adenovirus-mediated gene transfer in rat carotid arteries after balloon injury, and MCP-1 expression was evaluated immunohistochemically. We have previously shown that overexpression of Smad3 after arterial injury produces an intense neointimal response. We found that MCP-1 expression was significantly increased at day 5 after injury in arteries infected with AdSmad3 as compared with those infected with AdNull (Fig. 7, A–C). This in vivo finding confirms our in vitro results showing that Smad3 overexpression stimulates production of MCP-1 by VSMC.

Bone Marrow Cells Migrate to Areas of Arterial Wall Injury

We then examined whether arterial injury induces migration of BM cells into carotid arteries in vivo after injury. Bone marrow cells were labeled and injected systemically at the time of injury as described under “Experimental Procedures.” As shown in Fig. 7, D and E, few labeled bone marrow cells were observed in injured arteries but not in uninjured arteries.

PKCδ Signaling Pathway Is Necessary for TGF-β Stimulation of MCP-1

Although Smad3 is a critical signaling protein for TGF-β, several studies have previously shown that TGF-β can also signal through the ubiquitous kinase, PKCδ (28–30). Moreover, our previous data suggest that there may be cross-talk between the PKCδ and Smad3 pathways (2). To test whether PKCδ plays a role in TGF-β-induced MCP-1 expression, we first examined the effect of a general inhibitor of PKC, Ro 32-0432, on TGF-β-induced MCP-1 production. At a concentration of 1–10 μm, Ro 32-0432 has been shown to block PKC-dependent activities in whole cells (31). As shown in Fig. 8, A and B, Ro 32-0432 prevented both TGF-β- and TGF-β/AdSmad3-induced MCP-1 production in a dose-dependent manner, with complete inhibition at 5–10 μm. VSMC were then treated with TGF-β and AdSmad3 in the absence and presence of the selective PKCδ inhibitor, rotterlin, which has an IC₅₀ of 3–6 μm for PKCδ (32). As shown in Fig. 8C, pretreatment with rotterlin prevented TGF-β/AdSmad3- induction of MCP-1. To further confirm the involvement of PKCδ in the generation of MCP-1, we infected RSMC with an adenovirus expressing PKCδ (AdPKCδ) or AdNull and found that overexpression of PKCδ significantly induced MCP-1 production (Fig. 8D).

To determine whether PKCδ is necessary for TGF-β/AdSmad3-induced recruitment of bone marrow progenitor cells, we pretreated AdSmad3-infected VSMC with rotterlin followed by stimulation with TGF-β. Conditioned media from SMC pretreated with rotterlin induced significantly less BM cell or MSC migration compared with control-conditioned media (Fig. 8, E and F). Taken together, these data suggest that PKCδ signal transduction is necessary for TGF-β/Smad3-induced MCP-1 expression as well as the recruitment of bone marrow progenitor cells by TGF-β.