Matrix-embedded cells control osteoclast formation

Abstract

Osteoclasts resorb the mineralized matrices formed by chondrocytes or osteoblasts. The cytokine receptor activator of nuclear factor-κB ligand (RANKL) is essential for osteoclast formation and thought to be supplied by osteoblasts or their precursors, thereby linking bone formation to resorption. However, RANKL is expressed by a variety of cell types, and it is unclear which of them are essential sources for osteoclast formation. Here we have used a mouse strain in which RANKL can be conditionally deleted and a series of Cre-deleter strains to demonstrate that hypertrophic chondrocytes and osteocytes, both of which are embedded in matrix, are essential sources of the RANKL that controls mineralized cartilage resorption and bone remodeling, respectively. Moreover, osteocyte RANKL is responsible for the bone loss associated with unloading. Contrary to the current paradigm, RANKL produced by osteoblasts or their progenitors does not contribute to adult bone remodeling. These results suggest that the rate-limiting step of matrix resorption is controlled by cells embedded within the matrix itself.

Figure 1. Deletion of RANKL in Prx1-Cre expressing cells causes osteopetrosis

(a) RANKL mRNA levels in bone and spleen of RANKLf/f (n = 9) and Prx1-Cre;RANKLf/f (n = 10) littermates (here and throughout, values are the mean ± s.d.). *P < 0.05 versus RANKLf/f, using Student’s t-test. (b) Femoral BMD of WT (n = 7), Prx1-Cre (n = 7), RANKLf/f (n = 9), and Prx1-Cre;RANKLf/f (n = 10) littermates. *P < 0.05 versus WT, Prx1-Cre, and RANKLf/f, using 2-way ANOVA. (c) Cancellous bone volume in the distal femur of WT (n = 8), Prx1-Cre (n = 8), RANKLf/f (n = 9), and Prx1-Cre;RANKLf/f (n = 10) littermates. *P < 0.05 versus WT, Prx1-Cre, and RANKLf/f, using 2-way ANOVA. (d) Representative μCT images of the distal femur of 5-week-old RANKLf/f and Prx1-Cre;RANKLf/f mice. Scale bar, 1 mm. (e) Histological sections of distal femurs of 5-week-old RANKLf/f and Prx1-Cre;RANKLf/f mice stained with safranin-O (cartilage stains red). Scale bar, 0.5 mm. (f) Histological sections of distal femurs of 5-week-old RANKLf/f and Prx1-Cre;RANKLf/f mice stained for tartrate-resistant acid phosphatase (TRAP) activity (osteoclasts stain red). Scale bar, 200 μm. (g) Cathepsin K (Ctsk), TRAP (Acp5), and calcitonin receptor (Calcr) mRNA levels in tibial RNA of RANKLf/f (n = 9) and Prx1-Cre;RANKLf/f (n = 10) mice. *P < 0.05 versus RANKLf/f, using Student’s t-test. All values were determined in 5-week-old mice, including both sexes.

Figure 2. Deletion of RANKL in Osx1-Cre and Ocn-Cre expressing cells causes osteopetrosis

(a) RANKL mRNA levels in whole tibia of Osx1-Cre;RANKLf/f (n = 8), Ocn-Cre;RANKLf/f (n = 6), Dmp1-Cre;RANKLf/f (n = 9), and their respective RANKLf/f littermates (n = 4 to 11). *P < 0.05 versus RANKLf/f littermates, using Student’s t-test. (b) Quantitative PCR of loxP-flanked genomic DNA isolated from collagenase-digested femoral and tibial cortical bone of Dmp1-Cre;RANKLf/f (n = 9) mice and their RANKLf/f (n = 11) littermates. *P < 0.05 using Student’s t-test. (c) Cancellous bone volume of the distal femurs of Osx1-Cre;RANKLf/f (n = 8), Ocn-Cre;RANKLf/f (n = 6), Dmp1-Cre;RANKLf/f (n = 9), and their RANKLf/f littermates (n = 4 to 11). *P < 0.05 versus RANKLf/f littermates, using Student’s t-test. (d) X-ray images, representative μCT images of the distal femur (scale bar, 1 mm), safranin-O-stained histological sections of the distal femur (scale bar, 0.5 mm), anti-RANKL immunohistochemistry (IHC) (scale bar, 100 μm), and TRAP-stained histological sections of the distal femur (scale bar, 200 μm) from Osx1-Cre;RANKLf/f, Ocn-Cre;RANKLf/f, Dmp1-Cre;RANKLf/f, and a representative RANKLf/f littermate. Arrowheads in the X-rays indicate position of erupted incisors. μCT images for each of the RANKLf/f control littermates are presented in Supplementary Fig. 3. The region of the growth plate containing hypertrophic chondrocytes in the IHC images is outlined by green dashed lines and non-immune IgG controls are presented in Supplementary Fig. 3. All values and images are from 5-week-old mice and include both sexes.

Figure 3. Deletion of RANKL from Dmp1-Cre expressing cells reduces bone remodeling

(a) Serial BMD of Dmp1-Cre;RANKLf/f (n = 14) and RANKLf/f (n = 19) littermates. *P < 0.05 using Student’s t-test comparing the two genotypes at a given age. (b) Cancellous bone volume in the distal femur or in L4 vertebra of 6-month-old Dmp1-Cre;RANKLf/f (n = 11) and RANKLf/f (n = 7) littermates. *P < 0.05 using Student’s t-test. (c) Representative μCT images of the distal femur and L4 vertebra of 6-month-old Dmp1-Cre;RANKLf/f and RANKLf/f littermates. Scale bar, 1 mm. (d) Left, quantitative PCR of loxP-flanked RANKL genomic DNA using genomic DNA isolated from collagenase-digested femoral cortical bone of 6-month-old Dmp1-Cre;RANKLf/f (n = 11) and RANKLf/f (n = 7) littermates. Right, quantitative RT-PCR for RANKL mRNA in tibia and L5 vertebra of the same mice as in the left panel. *P < 0.05 using Student’s t-test. (e) Osteoclast number per mm bone surface in cancellous bone of the distal femur of 6-month-old Dmp1-Cre;RANKLf/f (n = 4) and RANKLf/f (n = 4) littermates. *P < 0.05 using Student’s t-test. (f) Carboxy-terminal crosslinked telopeptide of type I collagen (CTX), osteocalcin (Ocn), or soluble RANKL in the blood plasma of 6-month-old Dmp1-Cre;RANKLf/f (n = 9) and RANKLf/f (n = 8) littermates. *P < 0.05 using Student’s t-test. (g) RANKL mRNA levels in tibial cortical bone of 6-month-old Dmp1-Cre;RANKLf/f and RANKLf/f littermates, pretreated with vehicle or OPG and then injected with vehicle or PTH(1-34) (n = 6 to 8 per group). *P < 0.05 versus RANKLf/f mice pretreated with vehicle or OPG and then injected with vehicle using 2-way ANOVA. #P < 0.05 versus RANKLf/f mice pretreated with vehicle or OPG and then injected with PTH using 2-way ANOVA. All values include data from both sexes.

Figure 4. Osx1-Cre mediated RANKL deletion in adult mice does not alter osteoclast number in cancellous bone

(a) X-ray images (top) and histological sections of femurs stained with safranin-O (bottom) of 6-month-old Osx1-Cre;RANKLf/f and RANKLf/f littermates that were exposed to doxycycline in utero and maintained on a doxycycline-containing diet until 4 months of age. Scale bar, 500 μm. (b) Quantitative PCR of loxP-flanked genomic DNA isolated from collagenase-digested femoral cortical bone (left) and quantitative RT-PCR of RANKL mRNA in tibia and L5 vertebra (right panel) of Osx1-Cre;RANKLf/f (n = 12) mice and their RANKLf/f littermates (n = 10) exposed to doxycycline as described in (a). *P < 0.05 using Student’s t-test. (c) RANKL (left) and interleukin-6 (IL6) (right) mRNA levels in bone marrow cultures from Osx1-Cre;RANKLf/f and RANKLf/f littermates exposed to doxycycline as described in (a) or from 6-month-old Dmp1-Cre;RANKLf/f and RANKLf/f littermates. Each value represents the mean of 3 wells. *P < 0.05 with the comparisons indicated by the brackets using 2-way ANOVA. (d) Osteoclast number per mm bone perimeter in cancellous bone of the distal femur of Osx1-Cre;RANKLf/f (n = 4) and RANKLf/f (n = 4) littermates exposed to doxycycline as described in (a). (e) Cathepsin K (Ctsk) and TRAP (Acp5) mRNA levels in tibia and L5 vertebra of Osx1-Cre;RANKLf/f (n = 12) and RANKLf/f (n = 10) littermates exposed to doxycycline as described in (a). (f) Percent change in BMD between 4 and 6 months of age in Osx1-Cre;RANKLf/f (n = 14) and RANKLf/f (n = 8) littermates exposed to doxycycline as described in (a). All values include data from both sexes.

Figure 5. Tail-suspension of mice lacking RANKL in osteocytes

(a) Percent change of femoral BMD after 3 weeks of tail-suspension or normal loading (grounded control) in RANKLf/f and Dmp1-Cre;RANKLf/f littermates. (b) Cancellous bone volume (left), trabecular spacing (Tb.Sp.) (center), and cortical thickness (Ct.Th.) (right), in the femur of tail-suspended or grounded control RANKLf/f and Dmp1-Cre;RANKLf/f littermates. (c) RANKL mRNA levels in RNA prepared from collagenase-digested tibial cortical bone of tail-suspended or grounded control RANKLf/f and Dmp1-Cre;RANKLf/f littermates. (d) Osteoclast number per mm of cancellous bone surface in the distal femur of tail-suspended or grounded control RANKLf/f and Dmp1-Cre;RANKLf/f littermates. All values are from 6-month-old mice, include data from both sexes, and represent the following numbers of animals per group: grounded RANKLf/f (n = 8), suspended RANKLf/f (n = 7), grounded Dmp1-Cre;RANKLf/f (n = 8), and suspended Dmp1-Cre;RANKLf/f (n = 7), with the exception of the panel d, in which n = 5 for each group. *P < 0.05 versus grounded control of the same genotype by 2-way ANOVA.