Dendritic cell-mediated in vivo bone resorption

Abstract

Osteoclasts are resident cells of the bone that are primarily involved in the physiological and pathological remodeling of this tissue. Mature osteoclasts are multinucleated giant cells that are generated from the fusion of circulating precursors originating from the monocyte/macrophage lineage. During inflammatory bone conditions in vivo, de novo osteoclastogenesis is observed but it is currently unknown whether, besides increased osteoclast differentiation from undifferentiated precursors, other cell types can generate a multinucleated giant cell phenotype with bone resorbing activity. In this study, an animal model of calvaria-induced aseptic osteolysis was used to analyze possible bone resorption capabilities of dendritic cells (DCs). We determined by FACS analysis and confocal microscopy that injected GFP-labeled immature DCs were readily recruited to the site of osteolysis. Upon recruitment, the cathepsin K-positive DCs were observed in bone-resorbing pits. Additionally, chromosomal painting identified nuclei from female DCs, previously injected into a male recipient, among the nuclei of giant cells at sites of osteolysis. Finally, osteolysis was also observed upon recruitment of CD11c-GFP conventional DCs in Csf1r(-/-) mice, which exhibit a severe depletion of resident osteoclasts and tissue macrophages. Altogether, our analysis indicates that DCs may have an important role in bone resorption associated with various inflammatory diseases.

FIGURE 1

Implant of UHMWPE particles in mouse calvaria induces osteolysis. a, H&E staining of mouse calvaria in sham or UHMWPE-implanted mice (upper panels ×20 magnification; lower panels ×40 magnification). Arrows point to areas of osteolysis observed in the UHMWPE-implanted mice but not in the sham controls. b, H&E staining of pericalvarial tissue implanted with UHMWPE. mgcs surround the UHMWPE particles (asterisks) (original magnification ×100). c, Micro-CT scan of mouse calvaria (a representative frontal section is shown out of a total of 162 sections). CT scan was performed in both sham or UHMWPE-implanted mice pre- and postsurgery. Arrow indicates an area of osteolysis in UHMWPE-implanted mice. d, Analysis of pre- and postsurgery bone density parameters in sham and UHMWPE-implanted mice. One of three experiments is shown. In each experiment, three to four mice for each experimental condition (sham or UHMWPE) were used.

FIGURE 2

DCs are recruited to sites of osteolysis. a, Differential interference contrast image of the calvarium from a sham (left panel) and a UHMWPE-implanted (right panel) mouse (original magnification ×5). b, Fluorescence microscopic images of the same calvaria shown in a demonstrating the recruitment of GFP⁺ DCs in a mouse implanted with UHMPE (right panel) but not in the sham-treated mouse (left panel) (original magnification ×5). c, FACS quantification of GFP⁺-DCs recruited to the site of osteolysis in sham (left panel) or UHMWPE (right panel) implanted mice. d, Differential interference contrast image of frozen sections of the calvaria from sham (left panel) and UHMWPE-implanted (right panel) mice (original magnification ×63). e, Fluorescence microscopy of the same frozen sections from sham and UHMWPE-implanted mice shown in d to determine the presence of recruited GFP-labeled DCs (original magnification ×63). f, Overlay of images shown in d and e. One of two experiments is shown. In each experiment, three mice for each experimental condition (sham or UHMWPE) were used.

FIGURE 3

Cathepsin K-positive DCs are recruited to the site of osteolysis. a, Western blot analysis for cathepsin S, L, F, and K and TRAP of bone marrow-derived immature DCs differentiated for 8 d in GM-CSF (DC) or further differentiated with RANK-L and M-CSF for an additional week (osteoclasts). b, FACS analysis of CD11c expression of purified GFP⁺ DCs prior to injection into GFP^– recipient mice (solid line) or isotype control (dotted line). c–f, Confocal light (left panel) and fluorescent (right panel) images of calvarial sections analyzed for expression of cathepsin K⁺ (red) and GFP⁺ (green) DCs recruited to the site of surgery in sham (c) or UHMWPE-implanted mice (d–f). c and d, Original magnification ×63; e and f, original magnification ×100. Arrow indicates cathepsin K⁺ GFP⁺ DCs recruited to areas of osteolysis. g, FACS profile of GFP⁺-cathepsin K⁺ DCs retrieved from the calvaria of sham or UHMWPE-implanted mice. One of three experiments is shown. In each experiment, three mice for each experimental condition (sham or UHMWPE) were used.

FIGURE 4

Female donor DCs are recruited into male recipient mgcs at sites of osteolysis. Confocal light (a, d) and fluorescent (b, c, e, f) images of sections from the calvarium of an UHMWPE-injected mouse. b and e show sections stained for cathepsin K (red) and GFP (green). c and f represent serial sections of b and e, respectively, subjected to chromosomal painting for the Y chromo-some (pink). In c, note the absence of the Y chromosome (white arrow) in a GFP⁺ cell shown in b. d–f show an inflammatory infiltrate adjacent to the bone surface that contains many GFP⁺ cells (e), some of which lack the Y chromosome (f, white arrows). g, H&E staining of mgcs around UHMWPE particles (asterisks) containing a mixture of Y chromosome-positive and -negative nuclei (second and fourth panels, white arrows) within the same cell. One of two experiments is shown. a–g, Original magnification ×100.

FIGURE 5

DCs induce inflammatory osteolysis in absence of endogenous macrophages and osteoclasts. a, FACS profile of T cells (CD3), B cells (CD19), granulocytes (Gr-1), macrophages (CD11b), DCs (CD11c), and NKT cells (NK1.1) from the spleen of mice previously transduced with lentiviral vector expressing GFP under the control of a general promoter (PGK) or a tissue-specific promoter (CD11c). Mice were analyzed 2 mo after receiving the bone marrow transduced with the lentiviral vectors. Ten days prior to the FACS analysis, mice were injected with the B16–GM-CSF cells that produce endogenous GM-CSF, thus increasing the number of splenic DCs and facilitating the analysis of the CD11c promoter tissue specificity. b, Csf1r^+/– recipient mice were lethally irradiated and transplanted with the bone marrow of either Csf1r^–/– or Csf1r^+/– mice. The procedure generates chimeras that only express CD11c⁺ DCs (first and second panels) or DCs, macrophages, and osteoclasts (third and fourth panels), respectively. CT scan of the calvaria of chimeras mice implanted with UHMWPE or sham controls. One of two experiments is shown (in each experiment, three mice for each experimental condition [sham or UHMWPE] were used).