Hck contributes to bone homeostasis by controlling the recruitment of osteoclast precursors

Abstract

In osteoclasts, Src controls podosome organization and bone degradation, which leads to an osteopetrotic phenotype in src(-/-) mice. Since this phenotype was even more severe in src(-/-)hck(-/-) mice, we examined the individual contribution of Hck in bone homeostasis. Compared to wt mice, hck(-/-) mice exhibited an osteopetrotic phenotype characterized by an increased density of trabecular bone and decreased bone degradation, although osteoclastogenesis was not impaired. Podosome organization and matrix degradation were found to be defective in hck(-/-) osteoclast precursors (preosteoclast) but were normal in mature hck(-/-) osteoclasts, probably through compensation by Src, which was specifically overexpressed in mature osteoclasts. As a consequence of podosome defects, the 3-dimensional migration of hck(-/-) preosteoclasts was strongly affected in vitro. In vivo, this translated by altered bone homing of preosteoclasts in hck(-/-) mice: in metatarsals of 1-wk-old mice, when bone formation strongly depends on the recruitment of these cells, reduced numbers of osteoclasts and abnormal developing trabecular bone were observed. This phenotype was still detectable in adults. In summmary, Hck is one of the very few effectors of preosteoclast recruitment described to date and thereby plays a critical role in bone remodeling.

Keywords: Src tyrosine kinases; cell migration; osteopetrosis; podosomes.

Figure 1.

Hck^−/− mice have a high trabecular bone mass, due to bone degradation defects. A) High-resolution micro-computed tomography (micro-CT) images of femurs of 3-mo-old wt and Hck-deficient (hck^−/−) mice. White arrowheads show cortical bone; black arrowheads show trabecular bone. Trabecular bone is more dense in hck^−/− mice. Scale bars = 500 μm. B) Bone microarchitecture parameters are modified in hck^−/− mice: bone volume/tissue volume (BV/TV), trabecular number (Tb. N), trabecular separation (Tb. Sep), and connectivity. Data were obtained from 10 mice/phenotype. Error bars = sem. C) Biochemical markers of bone turnover. PINP and DPD levels were measured in serum and urine samples, respectively, from 3-mo-old control and hck^−/− mice. DPD level is decreased in hck^−/− mice. Data were obtained from 10 mice/phenotype. Error bars = sem.

Figure 2.

Formation of hck^−/− mature OCs in vitro is normal. A) Bone marrow mononuclear cells from wt and hck^−/− mice were seeded on glass coverslips in the presence of M-CSF and RANKL to promote OC differentiation. Cells were fixed at the indicated times and stained with Texas red-coupled phalloidin (F-actin, red) and DAPI (nuclei in green). Merged images representative of 7 experiments are shown. White arrowheads show large and mature OCs with their typical F-actin belts. Scale bar = 100 μm. B, C) Automatic quantification (Image J) of OC surface, number (B), and fusion index (C) at d 4 and 5 of differentiation. Data were obtained from 5 independent experiments. Error bars = sd.

Figure 3.

Podosome organization is defective in hck^−/− pre-OCs, but normal in mature OCs. A) Immunofluorescence microscopy of wt and hck^−/− pre-OCs. Cells were stained with antibodies against vinculin (green), Texas red-coupled phalloidin (F-actin, red) and DAPI (nuclei, blue). Merged images are shown. White arrowheads point to podosome rosettes that are mostly absent in hck^−/− pre-OCs. Scale bar = 10 μm. B) Percentage of pre-OCs exhibiting podosome rosettes (means±sd of 3 independent experiments, ≥100 cells/experiment). C) Immunofluorescence microscopy of mature OCs from wt and hck^−/− precursors. Cells were stained as in A. Merged images show individual podosomes (left panels) and podosomes organized as clusters or rings (middle panels) or belts (right panels), which are organized normally in hck^−/− mature OCs. Insets: 2.7-fold magnification of boxed areas. Scale bars = 10 μm.

Figure 4.

Bone resorption is increased and Src is overexpressed in hck^−/− mature OCs. A) Bone marrow mononuclear cells from wt and hck^−/− mice were seeded on bovine bone slides and differentiated into OCs with M-CSF and RANKL for 10 d. Then, OCs were lysed and bone slices were stained with toluidine blue to visualize resorption pits. Representative images of bone-resorption pits (violet, indicated by black arrowheads) generated by OCs. B) Quantitative data of results from A show an increase in bone degradation by hck^−/− OCs. Data were obtained from 6 independent experiments. C) Western blot analysis of OC supernatants (cell medium, left panels) and of OC total extracts (cell lysate, right panel) was performed using antibodies directed against cathepsin K and actin (as a loading control). D) Gelatin zymograph of OC supernatants (cell medium, left panels) and of OC total extracts (cell lysate, right panel) show that MMP9 activity is increased in hck^−/− mature OCs. C and D show a representative experiment out of 3 independent experiments. E) Western blot analysis of total cell extracts was performed using antibodies directed against Hck, Src, Lyn, and actin (as a loading control). The two isoforms of Hck migrated as a single band. F, G) Quantification of Hck (F) and Src (G) expression levels show that Src is overexpressed in hck^−/− mature OC. Arbitrary units (AU) represent the signal intensity measured with Adobe Photoshop, in 3 experiments.

Figure 5.

Migration of hck^−/− pre-OCs is defective in vitro and in vivo. A) Pre-OCs (d 3 of differentiation) were seeded on thick matrices of Matrigel and allowed to migrate. The percentage of hck^−/− pre-OCs infiltrating the matrices quantified after 48 h of migration is decreased compared to wt (means±sd of 3 independent experiments). B) Representative histological sections of metatarsals of 1-wk-old wt and hck^−/− mice stained with TRAP to visualize OCs (black arrows) and counterstained with hematoxylin. Bottom panels show 6-fold magnification of boxed areas in top panels. Scale bars = 100 μm. C) Surface occupied by TRAP-positive signal was quantified per bone surface in 6 separate histological sections per mouse (n=5 mice/phenotype). Number of TRAP-positive cells in hck^−/− is significantly diminished. Error bars = sem. D) Staining of metatarsals with Safranine O/Fast green to visualize cartilage (red) and bone formation (blue) show a defect in trabeculae organization in 1-wk-old hck^−/− mice (black arrows). Scale bar = 100 μm. E) OC surface was quantified per bone surface in 3 separate histological sections of femurs and tibia of 8-wk-old wt and hck^−/− mice (n=6 mice/phenotype). Number of OCs in hck^−/− mice is significantly diminished. F) Representative histological sections of trabecular bone of femurs stained with TRAP to visualize OCs, indicated by black arrows, and counterstained with hematoxylin to visualize bone trabeculae. Scale bar = 100 μm. Error bars = sem.