Runx2 mediates epigenetic silencing of the bone morphogenetic protein-3B (BMP-3B/GDF10) in lung cancer cells

To identify novel Runx2 target genes, we performed cDNA expression analysis on total RNA isolated from calvarial mesenchymal cells of wild type and functional deficient Runx2 mice [5]. In addition to the downregulation of known Runx2 target genes (e.g., matrix metalloproteinases) in a osteogenesis-related cDNA array [24], we found that the expression levels of BMP-3B gene was induced in Runx2 deficient cells compared to wild type cells (Figure 1a). The induction of BMP-3B expression in Runx2 deficient calvarial mesenchymal cells was validated by qRT-PCR analysis (Figure 1b). To further confirm Runx2-mediated downregulation of BMP-3B levels, we re-expressed Runx2 via adenoviral delivery in Runx2 deficient primary calvarial cells and measured BMP-3B levels by qRT-PCR analysis (Figure 1c). Our results show a dose-dependent repression of BMP-3B mRNA levels by Runx2 in primary osteoblastic cells. These results suggested that BMP-3B is a novel Runx2 responsive gene.

A tumor growth inhibitory function was proposed for BMP-3B in lung cancers and BMP-3B is downregulated in most of the lung cancers [22, 23, 25]. In context of Runx2-mediated BMP-3B suppression in mesenchymal cells and to understand the upstream regulatory mechanisms of BMP-3B silencing in lung cancers, we hypothesized that Runx2 downregulates BMP-3B expression in lung cancer. To understand the role of Runx2 in BMP-3B transcriptional regulation in lung cancer cells, we first examined Runx2 and BMP-3B mRNA levels in normal lung fibroblasts of mesenchymal origin (WI-38 and IMR-90), atypical carcinoid (H720) and metastatic non-small cell lung carcinoma (H1299) cells by qRT-PCR analysis. Our results showed that Runx2 expression is increased in metastatic lung cancer cells (H1299) compared to normal lung fibroblast cells. In contrast to the Runx2 expression levels, BMP-3B mRNA was detectable but lower in lung cancer cells compared to normal lung fibroblast cells (Figure 1d and e). The Western blot analysis for Runx2 protein levels further validated increased Runx2 levels in lung cancer cells compared to normal lung fibroblast cells (Figure 1f). A punctate nuclear staining of Runx2 was observed in WI-38 and H1299 cells as examined by immunofluorescence (Figure 1g). Taken together, these studies revealed that the inverse relationship between Runx2 and BMP-3B levels observed in calvarial mesenchymal cells also holds true for normal lung fibroblasts and lung cancer cells.

To investigate whether Runx2 suppresses BMP-3B levels in lung cancer cells similar to observed in primary calvarial cells, we stably overexpressed wild type Runx2 (WT) and Runx2 DNA binding domain mutant (DBD) in normal lung fibroblast cells (WI-38 and IMR-90) by lentiviral-mediated gene delivery. Expression levels of wild type and mutant Runx2 protein in these cell types were confirmed by qRT-PCR and western blot analysis (data not shown). Our results showed that stable expression of wild type Runx2 in normal lung cells resulted in more than 2-fold decrease in BMP-3B levels compared to empty vector control cells (Figure 2a and b). Ectopic expression of DBD mutant of Runx2 failed to downregulate BMP-3B levels in normal lung or lung cancer cells. These results suggested that the Runx2 DNA binding activity is required for BMP-3B regulation. In complementary studies, Runx2 knockdown resulted in increased BMP-3B levels in normal bronchial NL-20 cells (10- fold increase, Figure 2c right panel) and H1299 cells (2-fold increase, Figure 2d right panel) compared to empty vector controls as shown by qRT-PCR analysis. The decrease in Runx2 levels in Runx2 knockdown cells was confirmed by qRT-PCR and western blot analysis (Figure 2c and d). Collectively, these results indicate that Runx2 downregulates BMP-3B levels in normal lung fibroblast and lung cancer cells.

To further investigate the mechanism of Runx2-mediated downregulation of the BMP-3B expression in lung cancer cells, we performed chromatin immunoprecipitation analysis in H1299 cells expressing either wild type Runx2 or shRunx2 (Figure 3). Our results showed 3-fold increased Runx2 binding on the BMP-3B proximal promoter (-1.0kb) in H1299-WT-Runx2 cells, that was abrogated in H1299-shRunx2 cells. We next examined the methylation status of the BMP-3B proximal promoter as methylation of lysine 9 of histone H3 (H3K9) allows the binding of heterochromatin protein- 1 (HP1) to silence gene expression [26, 27]. Our results show increased (3-fold) H3K9 levels of proximal promoter region of BMP-3B in H1299-Runx2 cells compared to H1299-shRunx2 cells or antibody controls (Figure 3a). We next examined the recruitment of Suv39h1 protein, a histone H3 lysine 9 specific methyltransferase, on BMP-3B proximal promoter. A twofold increase in recruitment of Suv39h1 was observed in H1299-Runx2 cells compared to H1299-shRunx2 lung cancer cells (Figure 3a). These findings indicated the possibility of physical interaction of Runx2 and Suv39h1 proteins in lung cancer cells. We performed co-immunoprecipitaion assays with Runx2 and Suv39h1 antibodies and a direct interaction of Runx2 with Suv39h1 proteins was detected in H1229 cells (Figure 3b). Taken together, these results show that the recruitment of Runx2 and Suv39h1 on the BMP-3B proximal promoter sequences resulted in increased H3K9 methylation status and consequently downregulation of BMP-3B expression in lung cancer cells.

To examine the phenotypic effects of Runx2 overexpression in lung cancer cells, we assessed proliferation and migration potential of H1299-Runx2 cells or H1299- empty vector cells. Increased Runx2 levels in H1299-Runx2 cells and a corresponding decrease in BMP-3B mRNA expression were confirmed by western blot and qRT-PCR analysis respectively (Figure 4a). A 40% decline in cell proliferation was observed in Runx2 overexpressing H1299 cells compared to empty vector control cells in absence or presence of TGFβ treatment as examined by cell growth assay (data not shown) and MTT assays (Figure 4b). However, in response to TGF-β treatment the Runx2 overexpression in H1299 cells resulted in a significant (p <0.05) increase in wound healing response compared to the empty vector control for 6-48h as shown by wound healing assay (Figure 4c). The H1299 EV or WT-Runx2 cells did not show any differences in KI-67 (cell proliferation marker) immunoreactivity around wound area (data not shown). These results suggest that Runx2 promotes migratory potential of lung cancer cells by modulating TGF-β/BMP-3B signaling axis.

Normal bronchial and lung fibroblast (NL-20; WI-38, and IMR-90) and lung cancer cells (small cell lung cancer and non-small cell lung cancer cells: H720, and H1299) were cultured in growth medium as specified by American Type Culture Collection. The construction and procedure for wild type Runx2 or DNA binding mutant expressing adenovirus and lentivral transduction in normal and cancer cells are reported previously [24, 45].

Animals were maintained at the University of Massachusetts Medical School following procedures approved by the Institutional Animal Care and Use Committee (IACUC). Primary calvarial cells from Runx2 ^-/- mice were isolated as previously described [24].

Normal bronchial NL-20 or lung cancer H-1299 cells were transduced with lentivirus expressing shRNA-Runx2 target sequence 5’-AAGGTTCAACGATCTGAGATTTG-3’ sequence in pLVTHM vector under H1 promoter [34]. Runx2 knockdown efficiency was confirmed by western blot and real time RT-PCR analysis.

Runx2 protein levels in normal bronchial, fibroblast and lung cancer whole cell lysates or nuclear lysates were detected by western blot analysis as described previously [24]. Runx2 antibody (MBL Inc., Woburn, MA) or Suv39h1 (C-14, Santa Cruz Biotechnology Inc. Santa Cruz, CA) and HRP-conjugated secondary antibodies (Santa Cruz) were used to detect immunoreactive proteins.

Chromatin immunoprecipitation (ChIP) was performed as previously described [34]. Protein-DNA complexes were immunoprecipitated using Runx2 antibody (M-70, Santa Cruz Biotechnology Inc.), Suv39h1 (Santa Cruz Biotechnology Inc.) and histone H3K9 (Abcam, Cambridge, MA) or IgG as a control. Purified DNA was subjected to real time PCR amplification with SYBR Green chemistry on an ABI real time thermocycler. BMP-3B promoter fragment containing Runx elements were amplified using forward primer: 5’ ACT TTG ATG AAT CCG CAA CC-3’ and reverse primer: 5’ TTG TCT TGC CTC TAGCAG GAT-3’.

The mRNA levels of Runx2, BMP-3B, GAPDH and 28S in primary osteoblasts, normal lung fibroblast, bronchial and lung cancer cells were analyzed after adenovirus- or lentiviral-mediated Runx2 transduction. Total RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s specification. Purified RNA was oligo dT primed and cDNA synthesized at 42°C with SuperScript II RNA polymerase (Invitrogen). For PCR amplification, the following primers were used: Runx2, forward primer: 5’- CGG CCC TCC CTG AAC TCT -3’, reverse primer: 5’- TGC CTG CCT GGG GTC TGT A -3’, GAPDH, forward primer: 5’- ATG TTC GTC ATG GGT GTG AA -3’, reverse primer: 5’- TGT GGT CAT GAG TCC TTC CA -3’. BMP-3B, forward primer: 5’-AGC TGC TGG ACT TTG ACG AG-3’, reverse primer: 5’-TGA CAA TGC TCT GGA TGG TG-3’. 28S, forward primer: 5’- GAA CTT TGA AGG CCG AAG TG-3’, reverse primer: 5’-ATC TGA ACC CGA CTC CCT TT-3’. The gene expression levels were quantified by ΔΔCt method of relative quantification by normalizing the data with internal control and expressed relative to appropriate control cell line as indicated in the figure legends.

H1299 cells stably expressing Runx2 or empty vector treated control cells were cultured in triplicates in a 6 well dish with reduced serum conditions (0.2% serum) for overnight. The next day, a scratch was made approximately in the center of the monolayer by a sterile 200μl pipette tip. The detached cells and debris were washed with serum-free RPMI medium. The cells were then supplemented with or without TGF-β (5 ng/ml) containing RPMI medium. Five random images per well were photographed at 0h, 6h, 24h and 48h. The distance of the scratch was measured in ImageJ software at every time point. The wound distance at 0h was assigned as 100% and used to calculate percent wound closure at other time points. The P-value for statistical significance was calculated by unpaired T-test.

H1299 cells stably expressing Runx2 or empty vector treated control were counted in a hemacytometer and 1000 cells per well were seeded in a 96-well plate. To determine the changes in proliferation, the cells were indirectly assayed for cell number via a tetrazolium compound-based colorimetric assay (CellTiter 96 kit from Promega Inc. Madison, WI) according to manufacturer’s instructions. At indicated time points over a period of four days, the cell titer reagent (20μl/well) was added to the plate and incubated at 37°C for 1 hour. The quantity of color developed (formazan product from the tetrazolium compound) was measured by reading absorbance at 490 nm in a spectrophotometer (Fluostar Optima BMG Labtech Inc. Cary, NC).

Lung cancer H1299-WT-Runx2 or -shRunx2 cells were washed with ice-cold PBS and harvested in lysis buffer [50 mM NaCl, 50 mM Tris (pH 8.0), 1% NP- 40, 25 mM MG132, and 1× protease inhibitor mixture (Roche, Indianapolis, IN)]. Lysates were incubated overnight at 4°C with 3 μg of rabbit antibodies against Runx2 antibody (M-70, Santa Cruz Biotechnology Inc.), and Suv39h1 (Santa Cruz Biotechnologies). Lysates were then incubated with protein A/G beads for 2 h, followed by four washes with wash buffer [50 mM NaCl, 20 mM Tris (pH 8.3), 0.5% Na-deoxycholate, 0.5% Nonidet P-40, 2 mM EDTA, 25 mM MG132, and 1× protease inhibitor mixture]. The total cell lysates and immunoprecipitated protein complexes were resolved by 8% SDS/PAGE and transferred to polyvinylidene difluoride membranes (Immobilon-P, Millipore, Billerica, MA). Blots were incubated with Runx2 (M-70) or Suv39h1 (C-14) antibodies. Membranes were then incubated with HRP-conjugated secondary antibodies against rabbit or mouse (1:2,000). Proteins bands were visualized with a chemiluminescence detection kit (Perkin–Elmer Life Sciences,Waltham, MA).

WI-38 and H1299 cells grown on gelatin coated cover slips were processed for immunofluorescence microscopy as previously described [31] using rabbit polyclonal Runx2 antibody (Santa Cruz Biotechnology, Inc.), followed by incubation with Alexa 488 conjugated secondary antibody (Molecular Probes, Eugene, OR). All images were taken using a Zeiss Axioplan digital microscope and analyzed using Metamorph software (Universal Imaging, Downingtown, PA).