Negative regulation of osteoclast precursor differentiation by CD11b and β2 integrin-B-cell lymphoma 6 signaling

Abstract

Negative regulation of osteoclastogenesis is important for bone homeostasis and prevention of excessive bone resorption in inflammatory and other diseases. Mechanisms that directly suppress osteoclastogenesis are not well understood. In this study we investigated regulation of osteoclast differentiation by the β2 integrin CD11b/CD18 that is expressed on myeloid lineage osteoclast precursors. CD11b-deficient mice exhibited decreased bone mass that was associated with increased osteoclast numbers and decreased bone formation. Accordingly, CD11b and β2 integrin signaling suppressed osteoclast differentiation by preventing receptor activator of NF-κB ligand (RANKL)-induced induction of the master regulator of osteoclastogenesis nuclear factor of activated T cells, cytoplasmic 1 (NFATc1) and of downstream osteoclast-related NFATc1 target genes. CD11b suppressed induction of NFATc1 by the complementary mechanisms of downregulation of RANK expression and induction of recruitment of the transcriptional repressor B-cell lymphoma 6 (BCL6) to the NFATC1 gene. These findings identify CD11b as a negative regulator of the earliest stages of osteoclast differentiation, and provide an inducible mechanism by which environmental cues suppress osteoclastogenesis by activating a transcriptional repressor that makes genes refractory to osteoclastogenic signaling.

Figure 1. Bone phenotype of CD11b-deficienct mice

(A) Microcomputed tomography (micro-CT) of proximal femurs of 12 week-old male wild type (WT) and CD11b-deficient mice (n=6). Bone volume (BV/TV), trabecular space (Tb.Sp.), trabecular number (Tb.N.) and trabecular thickness (Tb.Th.) were determined by micro-CT analysis. (B) The parameters of osteoclastic bone resorption of wild type and CD11b-deficient mice were determined by bone histomorphometric analysis (n=8). Representative sections of femur, stained with TRAP and hematoxylin, are shown (left panel). (C) Bone formation parameters including mineral apposition rate (MAR) and bone formation rate (BFR) were measured in wild type and CD11b-deficeint mice. Scale bar, 50 µm. (n=4) *, p <0.05; **, p< 0.01; ns, not significant. Error bars show SEM.

Figure 2. Ligation of CD11b with fibrinogen on osteoclast precursor cells inhibits mouse osteoclastogenesis

(A to C) Mouse bone marrow cells from wild type (A, B) or CD11b-deficient mice (C) were cultured with M-CSF (20 ng/ml) for 4 days and then plated onto control FBS-coated wells or fibrinogen (Fb)-coated wells. At 1 hr after plating, RANKL (100 ng/ml) was added and TRAP-positive multinucleated (more than three nuclei) cells were counted 5 days after RANKL addition. The number of osteoclasts in control conditions is set as 100%. Data are shown as means ± SEM from >15 independent experiments. (B) Cells were cultured as in A and mRNA was measured using real-time PCR. mRNA levels were normalized relative to the expression of GAPDH. (B and C) Representative results from at least three independent experiments are shown. (D) Granuloma was induced by PMMA implantation under mouse skin and seven days later, CD11b-positive cells were isolated from granulomatous tissue. CD11b-positive cells from granuloma and bone marrow cells as a control were cultured as in A. Representative results from two independent experiments are shown.

Figure 3. Ligation of CD11b with fibrinogen on osteoclast precursor cells inhibits human osteoclastogenesis

(A to C) Human monocytes were cultured with M-CSF (20 ng/ml) for 1 day, and then plated onto control FBS-coated wells or fibrinogen (Fb)-coated wells. At 1 hr after plating, RANKL (40 ng/ml) was added and TRAP-positive, multinucleated (more than three nuclei) cells were counted 5 days after RANKL addition. Number of control osteoclasts is set as 100%. Data are shown as means ± SEM from > 20 independent donors. (B) Cells were fixed and stained using FITC-phalloidin to detect actin ring formation. (C) Cells were cultured as in A and mRNA was measured using real-time PCR. mRNA levels were normalized relative to the expression of GAPDH. Representative results from at least three independent experiments are shown. (D) Human OCPs were plated on control FBS-coated plates or polyRGD peptide (20 µg)-coated plates, and after 1 hr RANKL was added for an additional 5 days. (E) Human OCPs were incubated with gamma (γ) fibrinogen (377–395) peptide or scrambled peptide (40 µg/ml) for 30 mins, and then cells were plated on control FBS-coated plates or Fb-coated plates. At 1 hr after plating, RANKL (40 ng/ml) was added and replenished every three days. (F) Human OCPs were plated on FBS-coated, osteopontin (100 µg)-coated, or vitronectin (20 µg)-coated plates for 1 hr and then RANKL was added for an additional 5 days. (D to F) TRAP-positive multinucleated (more than three nuclei) cells were counted 5 days after RANKL addition. Number of control osteoclasts is set as 100%. Data are shown as means ± SEM from at least four independent experiments.

Figure 4. RANKL regulates expression of integrins during osteoclast differentiation

(A and B) Cell surface integrin expressions were measured by flow cytometry. (A) Mouse OCPs were cultured with RANKL (100 ng/ml) for three days and stained with anti mouse CD11b antibody. The filled histogram corresponds to a RANKL-induced sub-population of cells (RANKL-responsive) with increased side scatter, and the open histogram shows all other cells (RANKL-unresponsive). Quantitation of mean fluorescence intensity (MFI) is shown in the bottom panel. *, p <0.05. (B) Human OCPs were stained with antibodies against CD11b, β1, β2, β3, and β5 integrins. (C and D) Human OCPs were cultured in the presence of M-CSF (20 ng/ml) with or without RANKL (40 ng/ml) for 4 days. (C) mRNA was measured using real-time PCR. mRNA levels were normalized relative to the expression of GAPDH. Representative results from at least three independent experiments are shown. (D) Cell surface integrins were measured by flow cytometry Surface phenotype (upper panels) of each subset (β3 integrin-positive versus β3 integrin-negative) and quantitation of mean fluorescence intensity (MFI, lower panels) are shown. *, p <0.05. Plots are representative of at least four independent experiments.

Figure 5. Fibrinogen inhibits RANKL-induced NFATc1 expression

(A) Human OCPs were cultured with 40 ng/ml of RANKL with or without fibrinogen (Fb) for the indicated times. Whole-cell lysates were immunoblotted with NFATc1 and β-tubulin antibodies. (B and C) mRNA was measured using real-time PCR. Suppression of NFATc1 expression by Fb was observed in at least 10 independent experiments.

Figure 6. RANK expression is transiently suppressed by Fb

(A and B) Human OCPs were plated on control FBS-coated or fibrinogen (Fb)-coated wells and stimulated with RANKL for the indicated times. mRNA was measured using real-time PCR. mRNA levels were normalized relative to the expression of GAPDH. (B) Cells were cultured with BAPTA (20 µM) or picetannol (40 µM) for 30 mins prior to stimulation with Fb. (C) Mouse OCPs from wild type BALB/c and genetically matched FcRγ-deficient mice were plated on control FBS-coated wells or Fb-coated wells for 1 hr and then cells were stimulated with RANKL for 2 days. Whole-cell lysates were immunoblotted with NFATc1 and p38 antibodies. (D) Osteoclastogenesis assays of parallel wells corresponding to the experiments are shown in (C). Results are representative of at least four independent experiments. *, p <0.05.

Figure 7. BCL6 mediates inhibition of human osteoclastogenesis by Fb

(A and B) Human OCPs were plated on control FBS-coated plates or fibrinogen (Fb)-coated plates for the indicated times. BCL6 mRNA was measured using real-time PCR. (B) Human OCPs were incubated with Piceatannol (40 µM), BAPTA (20 µM) or Bay11 (10 µM) for 30 mins and then plated on control FBS-coated plates or fibrinogen-coated plates for 6 hrs. (C) Chromatin immunoprecipitation (ChIP) analysis of the recruitment of BCL6 to the NFATc1 promoter in human OCPs. Human OCPs were plated on control FBS-coated plates or Fb-coated plates for 24 hrs in the presence or absence of RANKL (40 ng/ml) and then harvested for ChIP analysis. Results (A–C) are representative of at least 3 independent experiments. (D) Human monocytes from 4 different blood donors were nucleofected with control or BCL6-specific small interfering RNAs, and osteoclasts were generated by culturing cells with M-CSF (20 ng/ml) and RANKL (40 ng/ml). TRAP-positive, multinucleated osteoclast formation was visualized by TRAP staining. Left panel, Data from four independent experiments are shown. Right panel, Representative results obtained from one donor are shown, where strong reversal of Fb-mediated suppression of osteoclastogenesis when BCL6 expression was diminished was observed. Lower panels, Efficacy of BCL6-specific small interfering RNA. Comparable results were obtained with an additional three BCL6-specific siRNAs.