TGFβ1-induced Baf60c regulates both smooth muscle cell commitment and quiescence

Abstract

Smooth muscle cells (SMCs) play critical roles in a number of diseases; however, the molecular mechanism underlying their development is unclear. Although the role of TGFβ1 signaling in SMC development is well established, the downstream molecular signals are not fully understood. We used several rat multipotent adult progenitor cell ((r)MAPC) lines that express levels of Oct4 mRNA similar to hypoblast stem cells (HypoSC), and can differentiate robustly to mesodermal and endodermal cell types. TGFβ1 alone, or with PDGF-BB, induces differentiation of rMAPCs to SMCs, which expressed structural SMC proteins, including α-smooth muscle actin (αSMA), and contribute to the SMC coat of blood vessels in vivo. A genome-wide time-course transcriptome analysis revealed that transcripts of Baf60c, part of the SWI/SNF actin binding chromatin remodeling complex D-3 (SMARCD3/BAF60c), were significantly induced during MAPC-SMC differentiation. We demonstrated that BAF60c is a necessary co-regulator of TGFβ1 mediated induction of SMC genes. Knock-down of Baf60c decreased SMC gene expression in rMAPCs whereas ectopic expression of Baf60c was sufficient to commit rMAPCs to SMCs in the absence of exogenous cytokines. TGFβ1 activates Baf60c via the direct binding of SMAD2/3 complexes to the Baf60c promoter region. Chromatin- and co-immunoprecipitation studies demonstrated that regulation of SMC genes by BAF60c is mediated via interaction with SRF binding CArG box-containing promoter elements in SMC genes. We noted that compared with TGFβ1, Baf60c overexpression in rMAPC yielded SMC with a more immature phenotype. Similarly, Baf60c induced an immature phenotype in rat aortic SMCs marked by increased cell proliferation and decreased contractile marker expression. Thus, Baf60c is important for TGFβ-mediated commitment of primitive stem cells (rMAPCs) to SMCs and is associated with induction of a proliferative state of quiescent SMCs. The MAPC-SMC differentiation system may be useful for identification of additional critical (co-)regulators of SMC development.

Figure 1. Differentiation of rMAPCs to smooth muscle cells.

rMAPCs were cultured for 6 days with TGFβ1 and PDGF-BB in serum free medium. On day 6, cells were harvested and re-plated in serum containing medium (Fig. S6), without exogenous growth factors. A. Expression of smooth muscle genes in differentiation Cl19 rMAPC represented as expression relative to GAPDH (Mean±s.e.m of n = 3–5, p<0.05; data for other lines can be found in Fig. S5). B. Cells were fixed and stained on days 0, 2, 6 and after passage/expansion in serum containing media with antibodies against SM22α, CNN1, α-SMA and SM-MHC. Nuclei were identified using Hoechst or Dapi. Images at 40× magnification, scale bar 50 µM; representative example one of 3 clones. C. Undifferentiated MAPCs and MAPC-derived SMCs were implanted in Matrigel plugs supplemented with VEGF and FGF2 under the skin of nude mice, and the plugs were harvested after 21 days. A representative image of a vessel formed by smooth muscle (αSMA) derived from rMAPCs indicated by αSMA colocalized to the GFP cell (white arrow) in the inset i1. The endothelium contribution of the vessel may have been from (GFP) rMAPCs injected (black arrow). Scale bar 10 µM. D. Quantification of the number of vessels wherein GFP and αSMA colocalised, per sq µm of the section (Mean±s.e.m of n = 4; Student's t-test p<0.01).

Figure 2. Genome wide transcriptome analysis of rMAPC-SMC differentiation.

rMAPCs (Cl19 and 3c3) were cultured for 6 days with TGFβ1 and PDGF-BB in serum free medium, and RNA harvested from triplicate samples on days 0, 2, 4 and 6. In addition, RNA from cultured RAOSMCs was obtained in duplicate. A. Ingenuity pathway analysis (IPA) of the differentially expressed genes in both clones of rMAPCs identified predominantly developmental and cardiaovascular system associated genes (table represents data obtained as output from IPA). B. Pathways relating to cardiogenesis were among the top canonical pathways (table represents data obtained as output from IPA). C. Validation by PCR of the target TF genes identified by microarray analysis.

Figure 3. Knock down of Baf60c down regulates αSMA protein levels.

rMAPCs were transduced with a doxycyclin (dox) inducible anti-Baf60c shRNA containing lentiviral vector or a vector containing a scrambled shRNA and subjected to SMC differentiation using TGFβ1. A. Quantitative RT-PCR for Baf60c mRNA levels in cells transduced with anti-Baf60c shRNAs in presence (+Dox) or absence (−Dox) of doxycycline up on TGFβ1 stimulation (Mean±s.e.m of n = 5; Student's t-test p<0.05). B. Western blots for Baf60c, αSMA, SM22α expression in cells transduced with anti-Baf60c shRNAs in the presence (+) or absence (−) of doxycyclin in rMAPC treated with TGF-β. C. Western blots for BAF60c and SM22α in RAOSMC treated with either siRNA against Baf60c or control.

Figure 4. Baf60c is sufficient to induce smooth muscle like fate in rMAPCs.

rMAPCs were transduced with an inducible Baf60c and eGFP cDNA encoding lentiviral vector or a vector encoding only eGFP, and subjected to SMCs without exogenous TGFβ1, and differentiation to SMCs examined by RT-qPCR and immunostaining. A. Expression of smooth muscle specific mRNAs compared to undifferentiated rMAPC (expression normalized to GAPDH) (Mean±s.e.m of n = 4; p<0.05). B. The expression of αSma in Baf60c transduced cells on day 9 cells is significantly higher compared to GFP transduced control cells (Mean±s.e.m of n = 4; Student's t-test p<0.05). C. Baf60c overexpression is sufficient to induce smooth muscle like fate with cells expressing mature smooth muscle proteins as compared to GFP transduced cells (smooth muscle proteins are colored pseudo green).

Figure 5. Baf60c is important for the synthetic/proliferative phase of primary smooth muscle cells.

RAOSMCs were either transfected with an on-target anti-Baf60c siRNA pool vs non-targeting pool (NTP) of siRNAs; or transduced with a lentiviral vector encoding for BAF60c or GFP A. qRT-PCR for the expression of Baf60c, Sm-mhc, and αSma in cells transfected with on-target anti-Baf60c siRNA pool vs non-targeting pool (NTP) of siRNAs (Mean ±s.e.m of n = 4; Student's t-test p<0.05). B. qRT-PCR for the expression of Baf60c, Sm-mhc, and αSma in cells transduced with either GFP or BAF60c (Mean±s.e.m of n = 3–5). C. Cell cycle analysis using fluorescence activated cell sorting in cells transfected with on-target anti-Baf60c siRNA pool vs non-targeting pool (NTP) of siRNAs (Mean±s.e.m of n = 3; Student's t-test p<0.05).

Figure 6. Baf60c is a direct target of SMAD mediated TGF-β signaling.

A. SMAD binding elements between -2096 and -2150 bp in the Baf60c promoter (underlined). B. rMAPCs were cultured with TGFβ1 alone or combined with an inhibitor against SMAD3 phosphorylation (SIS3) or TGFβR inhibitor (TGFβRi) (SB431542) On day 2 after treatment, cells were harvested and RT-qPCR used to detect expression of Baf60c transcripts (Mean±s.e.m of n = 3; Student's t-test p<0.05). C. The intact Baf60c promoter (wild type) or promoter wherein the SBEs between -2096 and -2150 bp had been deleted (mutant) was cloned before a luciferase expression cassette. These constructs together with a CAGA rich promoter sequence (CAGA-control) were transfected in rMAPCs. Following addition of TGFβ1 or in absence of TGFβ1, luciferase activity was measured (Mean±s.e.m of n = 3; Student's t-test p<0.05, NS- not significant).

Figure 7. BAF60c associates with SRF in binding to DNA elements (CArG boxes) in regulating smooth muscle genes.

V5 tagged BAF60c expressing clones of rMAPC were treated with TGFβ1 for 2 days. Cells were lysed, sonicated, and chromatin bound by BAF60c was immunoprecipitated with an anti-V5 antibody. The immunoprecipitated DNA was amplified using primers surrounding CArG boxes in the promoters of Sm22α and αSma. A. Representative example of 3 runs on an agarose gel. B. Real time PCR quantification for CArG box containing sequences in the αSma (Mean±s.e.m of n = 3; Student's p<0.05) and Sm22α (Mean±s.e.m of n = 3; Student's t-test p<0.01) promoters in chromatin IP using V5 antibody compared to isotype (mock) control antibody and negative control αSma intron region (Mean±s.e.m of n = 3; Student's t-test p>0.05). C. Co-immunoprecipitation using an antibody against SRF was probed with an antibody against Baf60c. We detected a Baf60c specific band that was not detected in the IgG control. Protein extracts from 293t cells transiently transfected with Baf60c vector was used as positive control.