FGF2 crosstalk with Wnt signaling in mediating the anabolic action of PTH on bone formation

Abstract

The mechanisms of the anabolic effect of parathyroid hormone (PTH) in bone are not fully defined. The bone anabolic effects of PTH require fibroblast growth factor 2 (FGF2) as well as Wnt signaling and FGF2 modulates Wnt signaling in osteoblasts. In vivo PTH administration differentially modulated Wnt signaling in bones of wild type (WT) and in mice that Fgf2 was knocked out (Fgf2KO). PTH increased Wnt10b mRNA and protein in WT but not in KO mice. Wnt antagonist SOST mRNA and protein was significantly higher in KO group. However, PTH decreased Sost mRNA significantly in WT as well as in Fgf2KO mice, but to a lesser extent in Fgf2KO. Dickhopf 2 (DKK2) is critical for osteoblast mineralization. PTH increased Dkk2 mRNA in WT mice but the response was impaired in Fgf2KO mice. PTH significantly increased Lrp5 mRNA and phosphorylation of Lrp6 in WT but the increase was markedly attenuated in Fgf2KO mice. PTH increased β-catenin expression and Wnt/β-catenin transcriptional activity significantly in WT but not in Fgf2KO mice. These data suggest that the impaired bone anabolic response to PTH in Fgf2KO mice is partially mediated by attenuated Wnt signaling.

Keywords: Bone; FGF2; PTH; Wnt signaling.

Fig. 1

PTH differentially regulated Wnt 10b in bones of WT and Fgf2KO mice. Three-month old female mice were treated with PTH (20 μg/kg body weight) or vehicle for 8 h. Left tibia was harvest for RNA extraction and right tibia was harvest for protein extraction. (A) Wnt10b mRNA expression by qPCR, n = 6–11 mice/group. (B) Wnt10b protein expression by western blot and (C) quantification, n = 3 mice/group. (C) Wnt3a mRNA expression by qPCR, n = 3 mice/group. Data presented are Mean ± SE. *: WT-Veh vs. Fgf2KO-Veh p < 0.05; #: compared with corresponding Veh p < 0.05 by ANOVA.

Fig. 2

SOST mRNA and protein were higher in Fgf2KO mice at basal level and reduced after PTH treatment. Three-month old female mice were treated with PTH (20 μg/kg body weight) or vehicle for 8 h. Left tibia was harvest for RNA extraction and right tibia was harvest for protein extraction. Femurs were used for IF. (A) SOST mRNA by qPCR, n = 6–11 mice/group. (B) SOST protein expression by Western blot and (C) quantification, n = 3 mice/group. (D) SOST protein in osteocyte of cortical bone by IF (arrows). SOST: red; DAPI: blue. C: cortical bone; M: bone marrow. Data presented are Mean ± SE. *: WT-Veh vs. Fgf2KO-Veh p < 0.05; #: compared with corresponding Veh p < 0.05 by ANOVA. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

PTH differentially regulated Dkk2 and Dkk1 mRNA in bones of WT and Fgf2KO mice. Three-month old female mice were treated with PTH (20 μg/kg body weight) or vehicle for 8 h. Left tibia was harvest for RNA extraction. (A) Dkk2 mRNA expression by qPCR, n = 6–11 mice/group. (B) (E) IF staining of DKK2. Osteoblasts on trabecular surface were label with green fluorescence sapphire (arrows). DKK2 was labeled with red fluorescence (dashed arrows). As noted, PTH markedly increased DKK2 in osteoblasts of WT mice but the increase was attenuated in the Fgf2KO mice (yellow, arrow heads). (C) Dkk1 mRNA expression by qPCR, n = 6–11 mice/group. Data presented are Mean ± SE. *: WT-Veh vs. Fgf2KO-Veh p < 0.05; #: compared with corresponding Veh p < 0.05 by ANOVA. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

PTH markedly increased pLRP6 in bones of WT mice but the increase was attenuated in Fgf2KO mice. Three-month old female mice were treated with PTH (20 μg/kg body weight) or vehicle for 8 h. Left tibia was harvest for RNA extraction and right tibia was harvest for protein extraction. Femurs were used for IF. (A) Lrp5 and (B) Lrp6 mRNA expression by qPCR, n = 6–11 mice/group. Western blot analysis of (C) pLRP6 vs. total LRP6 and (D) quantification, n = 3 mice/group. Data presented are Mean ± SE. *: WT-Veh vs. Fgf2KO-Veh p < 0.05; #: compared with corresponding Veh p < 0.05 by ANOVA. (E) IF staining of pLRP6. Osteoblasts on trabecular surface were label with green fluorescence sapphire. pLRP6 was labeled with red fluorescence. As noted, PTH markedly increased pLRP6 in osteoblasts of WT mice but the increase was attenuated in the KO mice (arrows). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

PTH differentially regulated β-catenin expression in bones of WT and Fgf2KO mice. Three-month old female mice were treated with PTH (20 μg/kg body weight) or vehicle for 8 h. Left tibia was harvest for RNA extraction and right tibia was harvest for protein extraction. (A) β-Catenin mRNA expression by qPCR, n = 6–11 mice/group. (B) Active vs. total β-catenin protein expression by Western blot and (C) quantification, n = 3 mice/group. (D) Inactive vs. total GSK3β protein expression by Western blot and (E) quantification, n = 3 mice/group. Data presented are Mean ± SE. *: WT-Veh vs. Fgf2KO-Veh p < 0.05; #: compared with corresponding Veh p < 0.05 by ANOVA.

Fig. 6

Effect of PTH treatment on Wnt/β-catenin transcriptional activity in primary calvarial osteoblasts of WT and Fgf2KO mice. Primary calvarial osteoblasts from 3 days old WT and Fgf2KO mice were plated at 30,000 cells/well/300 μl in 48-well plate in DMEM +10% FBS + 1% P/S. At the confluence of 70–90% confluence, cells were transfected with (A) TOPflash luciferase reporter or (B) FOPflash luciferase reporter and beta-galactosidase (BetaGal) for 36 h, then the cells were treated with Vehicle or 10 nM PTH for 3 h or 8 h. TOPflash luciferase activity or TOPflash luciferase activity was normalized to betaGal and the ratio in WT vehicle group was set to one. FOPflash was used as negative controls. Data are pool results from three independent experiments. Data presented are Mean ± SE. #: compared with corresponding Veh p < 0.05 by ANOVA.