Connective tissue growth factor is a target of notch signaling in cells of the osteoblastic lineage

Abstract

Connective tissue growth factor (Ctgf) or CCN2 is a protein synthesized by osteoblasts necessary for skeletal homeostasis, although its overexpression inhibits osteogenic signals and bone formation. Ctgf is induced by bone morphogenetic proteins, transforming growth factor β and Wnt; and in the present studies, we explored whether Notch regulated Ctgf expression in osteoblasts. We employed Rosa(Notch) mice, where the Notch intracellular domain (NICD) is expressed following the excision of a STOP cassette, placed between the Rosa26 promoter and NICD. Notch was activated by transduction of adenoviral vectors expressing Cre recombinase (Ad-CMV-Cre). Notch induced Ctgf mRNA levels in a time dependent manner and increased Ctgf heterogeneous nuclear RNA. Notch also destabilized Ctgf mRNA shortening its half-life from 13h to 3h. The effect of Notch on Ctgf expression was lost following Rbpjκ downregulation, demonstrating that it was mediated by Notch canonical signaling. However, downregulation of the classic Notch target genes Hes1, Hey1 and Hey2 did not modify the effect of Notch on Ctgf expression. Wild type osteoblasts exposed to immobilized Delta-like 1 displayed enhanced Notch signaling and increased Ctgf expression. In addition to the effects of Notch in vitro, Notch induced Ctgf in vivo, and calvariae and femurs from Rosa(Notch) mice mated with transgenics expressing the Cre recombinase in cells of the osteoblastic lineage exhibited increased expression of Ctgf. In conclusion, Ctgf is a target of Notch canonical signaling in osteoblasts, and may act in concert with Notch to regulate skeletal homeostasis.

Keywords: CCN proteins; Notch; connective tissue growth factor; osteoblasts; transcription.

Figure 1

Effect of Notch on Hey1, Hey2 (lower panel) and Ctgf (upper left panel) mRNA, Ctgf hnRNA (upper middle panel) and Ctgf protein (upper right panel) expression in osteoblasts. Calvarial osteoblasts isolated from Rosa^Notch mice were transduced with Ad-CMV-Cre, to activate Notch (black bars), or control Ad-CMV-GFP (white bars) and cultured to confluence (day 0) or up to 2 weeks following confluence. Samples for hnRNA and protein determination were obtained 3 days post-confluence. Total RNA was extracted, reversed transcribed and amplified by qRT-PCR. Data for mRNA and hnRNA are expressed as Ctgf, Hey1 and Hey2 copy number corrected for Rpl38 expression relative to the mRNA or to the Ctgf hnRNA expression in Ad-CMV-GFP control cells, arbitrarily set at a value of 1. Data for Ctgf protein, measured by ELISA, are expressed as ng/ml of culture medium. Data for Ctgf, Hey1 and Hey2 mRNA were pooled from 2 experiments. Values are means ± SEM; n = 4. *Significantly different between Ad-CMV-Cre Notch activated cells and control, p < 0.05.

Figure 2

Effect of Notch on Ctgf transcript stability in osteoblasts. Calvarial osteoblasts isolated from Rosa^Notch mice were transduced with Ad-CMV-Cre, to activate Notch (filled circles), or control Ad-CMV-GFP (open circles) and cultured. Seventy-two h after confluence, cells were transcriptionally arrested by the addition of DRB (time 0), and harvested at the indicated times after DRB. Total RNA was extracted, reversed transcribed and amplified by qRT-PCR. Values are means ± SEM; n = 11 to 12. Data are expressed as percent of Ctgf mRNA corrected for Rpl38 expression, relative to the time of DRB addition and plotted versus time, and were pooled from 3 independent experiments. Slopes from Notch activated and control cells are significantly different, p < 0.05.

Figure 3

Effect of Notch on Ctgf expression in the context of Rbpjκ downregulation. Calvarial osteoblasts isolated from Rosa^Notch mice were transduced with Ad-CMV-Cre, to activate Notch (black bars), or with control Ad-CMV-GFP (white bars), transfected with Rbpjκ small interfering RNA (siRbpjκ) or scrambled siRNA (siScram) and cultured for 72 h. Total RNA was extracted, reverse transcribed and amplified by qRT-PCR. Data are expressed as copy number of Rbpjκ mRNA, Ctgf mRNA and Ctgf hnRNA, corrected for Rpl38. Values are means ± SEM; n = 4. * Significantly different between siRbpjκ and siScram, p < 0.05.

Figure 4

Effect of Notch on Ctgf expression in the context of Hes1, Hey1 or Hey2 downregulation. Calvarial osteoblasts isolated from Rosa^Notch mice were transduced with Ad-CMV-Cre, to activate Notch (black bars), or with control Ad-CMV-GFP (white bars), transfected with Hes1, Hey1 or Hey2 small interfering RNA (si) or scrambled siRNA (siScram) and cultured to confluence or for 3 days after confluence. Total RNA was extracted, reverse transcribed and amplified by qRT-PCR. Data are expressed as copy number of Ctgf mRNA corrected for Rpl38. Downregulation of Hes1, Hey1 and Hey2 mRNA in control (Ad-CMV-GFP) and Notch activated (Ad-CMV-Cre) cells, expressed as the mean % of suppression relative to the mRNA expression in siScram cells is indicated in the right upper corners of both panels. Values are means ± SEM; n = 4. * Significantly different between siHes1, Hey1 or Hey2 and siScram, p < 0.05.

Figure 5

Effect of Notch on Hey2 mRNA, Ctgf mRNA and Ctgf hnRNA expression in osteoblasts. Wild type calvarial osteoblasts were cultured on plates coated with the Notch ligand Delta like 1 (Dll1) at the indicated doses for 72 h following confluence. Total RNA was extracted, reverse transcribed and amplified by qRT-PCR. Data are expressed as copy number of Hey2 mRNA, Ctgf mRNA and Ctgf hnRNA corrected for Rpl38 expression. Values are means ± SEM; n = 4. *Significantly different from control, p < 0.05.

Figure 6

Effect of Notch on Ctgf expression in calvarial and femoral extracts. Total RNA extracted from calvariae of male and female mice and femurs from male Osx-Cre^+/−;Rosa^Notch, Col2.3-Cre^+/−;Rosa^Notch; Bglap/Oc-Cre^+/−;Rosa^Notch and Dmp1-Cre^+/−;Rosa^Notch (black bars) and respective control Rosa^Notch littermates of the same sex (white bars) was reverse transcribed and amplified by qRT-PCR. Data are expressed as copy number of Ctgf corrected for Gapdh and expressed as the relative ratio with respect to controls (ratio 1.0). Values are means ± SEM; n = 3 to 10. *Significantly different between Rosa^Notch and control littermates, p < 0.05; # p < 0.07.