Retinoblastoma binding protein-1 (RBP1) is a Runx2 coactivator and promotes osteoblastic differentiation

Abstract

Background: Numerous transcription factors are involved in the establishment and maintenance of the osteoblastic phenotype, such as Runx2, osterix and Dlx5. The transcription factor retinoblastoma binding protein-1 (RBP1) was recently identified as an estrogen regulated gene in an osteosarcoma cell model. Since the function of RBP1 in osteoblastic differentiation and mineralization is unknown, we investigated the role of RBP1 in these processes.

Methods: To create a cell model with suppressed RBP1 expression, primary calvarial osteoblasts were infected with a shRNA lentiviral vector specific for mouse RBP1 (CalOB-DeltaRBP1) or a scrambled control shRNA lentivirus (CalOB-Control). Stable cell lines were generated and their mineralization potential was determined using osteoblastic differentiation medium, Alizarin Red staining, and quantitative PCR (QPCR) analyses. Runx2 coactivation by RBP1 was determined through the use of transient transfection assays.

Results: Stable expression of the RBP1 shRNA lentivirus in CalOB-DeltaRBP1 cells resulted in a 65-70% suppression of RBP1 expression. Osteoblastic mineralization assays demonstrated that suppression of RBP1 results in a potent delay in osteoblastic nodule formation in the CalOB-DeltaRBP1 cells with a concomitant decrease in the expression of the osteogenic transcription factors Runx2 and osterix, along with decreases in BMP2, alkaline phosphatase, osteocalcin and bone sialoprotein. Regulation of Runx2 expression by RBP1 was shown to be mediated through the proximal P2 Runx2 promoter. Furthermore, RBP1 was demonstrated to be a potent coactivator of Runx2 transcriptional activity on two known Runx2 reporter constructs. These data suggest that the expression and activity of Runx2 is critically dependent on the presence of RBP1.

Conclusions: This study is the first to demonstrate that RBP1 is an important mediator of the osteoblastic phenotype and clearly defines RBP1 as a novel transcription factor involved in the well known Runx2-osterix transcriptional cascade. As such, the effects of RBP1 on these processes are mediated through both regulation of Runx2 expression and transcriptional activity.

Figure 1

shRNA design and confirmation of RBP1 knockdown in mouse calvarial osteoblasts. (A) The sequence at +3299 to +3320 relative to the start of transcription of mouse RBP1 gene was selected and packaged into a shRNA lentiviral expression vector along with a scrambled shRNA control. Both lentiviral constructs were infected into wild-type mouse calvarial osteoblasts and stable cell lines were selected using blasticidin S for 2 weeks. The resulting cell lines are termed CalOB-Control and CalOB-ΔRBP1. (B) Total RNA from the CalOB-Control and CalOB-ΔRBP1 cell lines (n = 3) were isolated and processed for real-time QPCR analysis. The data are expressed as RBP1 expression relative to TBP controls and normalized to the CalOB-Control cell line. (C) The CalOB-Control and CalOB-ΔRBP1 cell lines were plated (n = 6), allowed to grow for 72 hours and a proliferation assay was performed. The data are expressed as relative cellular proliferation normalized to the CalOB-Control cell line. For all panels, the black bars represent the mean ± standard deviation. The asterisk represents statistical significance at the p < 0.01 level (ANOVA).

Figure 2

Alizarin Red staining reveals a significant delay in osteoblastic mineralization of the CalOB-ΔRBP1 cell line. The CalOB-Control (A) and CalOB-ΔRBP1 (B) cell lines were plated (n = 3) and allowed to differentiate in osteoblastic differentiation media for 0, 10, 17 and 31 days. The cells were subsequently stained using Alizarin Red. (C) Total RNA was extracted from identically treated CalOB-Control (solid line) and CalOB-ΔRBP1 (dotted line) cells and processed for real-time QPCR analysis (n = 3). The data are expressed as RBP1 expression relative to a TBP control and normalized to the day 0 CalOB-Control cell line. The data points represent the mean ± standard deviation. The asterisks represents statistical significance at the p < 0.01 level (ANOVA).

Figure 3

Real-time QPCR analysis demonstrates suppressed bone marker gene expression in the CalOB-ΔRBP1 cell line. Total RNA was extracted from CalOB-Control (solid line) and CalOB-ΔRBP1 (dotted line) cells treated as in Figure 2 and processed for real-time QPCR analysis (n = 3) for the indicated genes (A-F). The data are expressed as target gene expression relative to TBP controls and normalized to the day 0 CalOB-Control cell line. The data points represent the mean ± standard deviation. The asterisks represents statistical significance at the p < 0.01 level (ANOVA).

Figure 4

RBP1 regulates Runx2 through the Runx2 P2 promoter. Runx2-specific primer pairs designed to transcripts arising from either the Runx2 P1 promoter (open bars) or the Runx2 P2 promoter (closed bars) were used in a real-time QPCR analysis from RNA isolated from the CalOB-Control or CalOB-ΔRBP1 cell lines. The data are expressed as Runx2 expression relative to a TBP controls and normalized to the day 0 CalOB-Control cell line. The data represent the mean ± standard deviation. The asterisks represents statistical significance at the p < 0.01 level (ANOVA).

Figure 5

RBP1 is a Runx2 coactivator in both U2OS osteosarcoma and mouse calvarial osteoblast cell models. Mouse calvarial osteoblast (A) or human U2OS osteosarcoma (B) cells were transiently transfected (n = 3) with either the p6OSE2-Luc reporter construct or the mouse osteocalcin luciferase reporter construct as described under "Methods." The black bars represent the mean ± standard deviation. A single asterisk (*) represents significance of p < 0.01 (ANOVA) relative to the reporter construct alone and a double asterisk (**) represents significance of p < 0.01 (ANOVA) relative to the reporter and Runx2 expression construct alone. U.D. stands for undetectable.