Table 1.
Primer sequences used in this study.
Fig 1.
Ex vivo deletion of sirt1 inhibits osteoblast differentiation.
A) Primary osteoblasts obtained from the calvaria of Sirt1flox/flox neonates were infected two days post-confluency with adenoviral-CRE to excise sirt1. B) Cells infected with adenoviral-CRE show excision of SIRT1 catalytic exon 4 (T1Δ4) as indicated by a smaller PCR product obtained with primers flanking exon 4. C) CRE-infected cells show reduced alkaline phosphatase enzymatic activity, an early marker of osteoblast differentiation. (n = 3, * p<.05) D) CRE-infected cells showed reduced staining for two different markers of osteoblast differentiation, alkaline phosphatase and alizarin red, a marker of mineralization.
Fig 2.
Ex vivo deletion of sirt1 decreases expression of RUNX2 downstream targets.
A) A schematic representing the osteoblast transcriptional regulators and markers examined in this study. The homeodomain transcriptional regulators, Msx2, Dlx3, and Dlx5, help establish the early osteoblast transcriptional program, including upregulation of Runx2 expression and activity [17, 21–24]. Runx2 then directly binds to and stimulates the transcription of osteoblast specific genes, including Osterix (Osx), an essential osteoblast transcription factor. B) There are no differences in the expression of Msx2, Dlx3, and Dlx5 in SIRT1 knockout (Cre-infected) versus wildtype (vector-infected) osteoblasts as ascertained by quantitative reverse-transcription PCR (qRT-PCR). C) While SIRT1 knockout osteoblasts (Cre) express comparable amounts of Runx2, they show a near two-fold reduction in the expression of the Runx2 downstream target, Osterix (Osx). D) Three other RUNX2 targets, Osteocalcin, Osteopontin, and Bone Sialoprotein (BSP), also show reduced expression in SIRT1 knockout cells (Cre), suggesting decreased transcriptional activity of RUNX2 in the absence of SIRT1. (n≥3, * p<.05; ** p<.01; *** p<.005).
Fig 3.
A) Tagged versions of SIRT1 and RUNX2 interact in 293T cells: FLAG-tagged SIRT1 is able to co-immunoprecipitate HA-tagged RUNX2, and vice versa. (WB: western blot; IP: immunoprecipitation) B) This interaction also exists at the endogenous level. Two different RUNX2 antibodies (Sigma and Abcam) co-immunoprecipitate SIRT1, but not closely related SIRT6 or abundantly expressed HSP90, in U2OS osteosarcoma cell lysates. The band below the RUNX2 band (in the Runx2 WB panel) represents heavy chain IgG. (AB: antibody).
Fig 4.
SIRT1 increases the transcriptional activity of RUNX2.
A) A RUNX2 luciferase reporter assay (p6OSE2) shows that overexpression (OE) of SIRT1 increases luciferase activity, while RNA-interference (RNAi) of SIRT1 decreases it (n = 4). B) Two specific SIRT1 activators, SRT1720 and SRT2183, also increase luciferase activity (n = 4). C) SIRT1 activator, SRT2183, leads to induction of endogenous RUNX2 target Bone Sialoprotien (BSP) in wildtype (vector) but not SIRT1 excised (Cre) cells, indicating that the stimulatory effects of SRT2183 on RUNX2 is SIRT1-dependent. The expression of Runx2 itself is unchanged. (n≥3, * p<.05; ** p<.01; *** p<.005).
Fig 5.
Pharmacological activation of SIRT1 promotes osteoblast differentiation, and expression of RUNX2 targets in vivo.
A) Treatment of primary osteoblasts with SIRT1 activator, SRT2183, increases markers of differentiation (alkaline phosphatase and alizarin red) in a dose-dependent manner. B) Mice (n = 8) were fed 400mg/kg/day of resveratrol (another SIRT1 activator) or vehicle control and had the expression of RUNX2 downstream targets examined in their calvaria (skullcap) by qRT-PCR. C) Resveratrol (Resv) fed mice show similar bone volume/total volume (BV/TV) and bone mineral density (BMD) as control fed mice (due to the short treatment regimen). D) The calvaria of resveratrol fed mice show increased expression of RUNX2 downstream targets, but not RUNX2 itself. (n = 8, * p<.05; ** p<.01; *** p<.005).