Fig. 3.
Adipocytes are the principal source for RANKL after loss of IFT20 in MSCs, and IFT20 deficiency in adipocytes reverses bone phenotype by reducing RANKL expression. (A) Representative TRAP-stained image of femurs from 1-month-old Prx1-Cre; IFT20f/f mice and controls. Scale bar, 200 μm. N = 5. High magnification image of red box area was at right. Scale bars, 100 μm. The red arrow indicates adipocytes that are found in close vicinity to TRAP positive osteoclasts. (B, C) qRT-PCR analysis of RANKL in whole bone marrow (B) and bone marrow adipose tissue (C). (D) The serum level of RANKL from 1-month-old Prx1-Cre; IFT20f/f mice were significantly increased compared to age-mated controls. (E) The serum RANKL/OPG ratio was identified in 1-month-old Prx1-Cre; IFT20f/f mice and controls as indicated. (F) Representative micro-CT image of femurs from 1-month-old Adipoq-Cre; IFT20f/f mice and controls. Scale bars, 1 mm. N = 5. (G) Quantitative BMD measurements of femurs from 1-month-old Adipoq-Cre; IFT20f/f mice and controls. (H) Representative H&E-stained image of femurs from 1-month-old Adipoq-Cre; IFT20f/f mice and controls. Scale bars, 200 μm. N = 5. (I) Representative TRAP-stained image of femurs from 1-month-old Adipoq-Cre; IFT20f/f mice and controls. Scale bar, 100 μm. The corresponding quantitative analysis of TRAP staining were at right. (J, K) qRT-PCR analysis of RANKL in whole bone marrow (J) and bone marrow adipose tissue (K). (L) The serum levels of RANKL from 1-month-old Adipoq-Cre; IFT20f/f mice were significantly decreased compared to controls. Error bars were the means ± SEM from three independent experiments. *P < 0.05, **P < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)