Annexin II increases osteoclast formation by stimulating the proliferation of osteoclast precursors in human marrow cultures

Abstract

Annexin II (AXII), a calcium-dependent phospholipid-binding protein, has been recently found to be an osteoclast (OCL) stimulatory factor that is also secreted by OCLs. In vitro studies showed that AXII induced OCL formation and bone resorption. However, the mechanism of action by which AXII acts as a soluble extracellular protein to induce OCL formation is unknown. In this paper, we demonstrate that AXII gene expression is upregulated by 1,25-dihydroxyvitamin D3 [1, 25-(OH)2D3] and that addition of AXII significantly increased OCL-like multinucleated cell formation. Time-course studies suggested that AXII acted on the proliferative stage of OCL precursors and that AXII increased thymidine incorporation in OCL precursors. Moreover, AXII enhanced the growth of CFU-GM, the earliest identifiable OCL precursor, when bone marrow cultures were treated with low concentrations of GM-CSF. This capacity of AXII to induce OCL precursor proliferation was due to induction of GM-CSF expression, because the addition of neutralizing antibodies to GM-CSF blocked the stimulatory effect of AXII on OCL formation. RT-PCR analysis using RNA from highly purified subpopulations of marrow cells demonstrated that T cells, especially CD4(+) T cells, produced GM-CSF in response to AXII. Furthermore, FACS(R) analysis of T-cell subpopulations treated with fluorescein-labeled AXII suggested that the CD4(+), but not CD8(+), subpopulation of T cells express an AXII receptor. Taken together, these data suggest that AXII stimulates OCL formation by activating T cells through a putative receptor to secrete GM-CSF. GM-CSF then expands the OCL precursor pool to enhance OCL formation.

Figure 1

(a) Effects of 1,25-(OH)₂D₃ on AXII mRNA expression during OCL differentiation. Total RNA has been isolated from mature OCL-like multinucleated cells and OCL precursors treated with 1,25-(OH)₂D₃ (10^–8 M) at the times indicated. Semiquantitative RT-PCR was performed using specific primers for AXII as described in Methods. β-actin was used as internal control for amount of RNA isolated and the integrity of the RNA. (b) In situ hybridization of AXII RNA in multinucleated OCLs from human giant cell tumors of bone. Giant cells were prepared, fixed, and incubated with sense and antisense probes for AXII in 4-well LabTek chamber slides, and autoradiography was performed as described in Methods. The dark granules on the cells are magnetic beads used in the immunopurification of the OCL-like cells.

Figure 2

Effect of varying treatment time with AXII on formation of 23c6Ab+ multinucleated cells in culture of normal human bone marrow. Human bone marrow mononuclear cells (10⁵ cells per well) were cultured in 96-well plates in α-MEM and 20% horse serum, with AXII (100 ng/mL) for 1, 2, or 3 weeks, followed by treatment with media alone. In selected experiments, AXII was added to cultures for only the second and third weeks, or the third week only, of culture. Values are mean ± SEM of triplicate determinations in 5 independent experiments and are expressed as a percentage of the number of multinucleated cells formed by cells cultured in media alone for 3 weeks. *P < 0.02, significantly different from control cultures (3 weeks media alone).

Figure 3

[³H]thymidine incorporation of OCL-like multinucleated cells formed in response to AXII. Normal human bone marrow nonadherent mononuclear cells (5 × 10⁵ cells per culture) were cultured in 4-well LabTek chamber slides in the presence or absence of AXII. After 24 hours of culture, [³H]thymidine (0.5 mCi/well) was added for the next 48 hours. Cells were fixed and stained for TRAP and then processed for autoradiography. The number of TRAP⁺ multinucleated cells (a) and percentage of multinucleated cells containing labeled nuclei (b) were scored. Results are expressed as the mean ± SEM for 3 or 4 cultures from a typical experiment. Similar results were seen in 3 independent experiments. *P < 0.05, significantly different from control cultures (media alone).

Figure 4

Effect of AXII on CFU-GM colony formation. Human bone marrow mononuclear cells (10⁵ cells/mL) were cultured in 35-mm plates in α-MEM with 30% FCS in methylcellulose. Varying concentrations of AXII (0–100 ng/mL) were added to the cultures in the presence of 100 pg/mL hGM-CSF. CFU-GM–derived colonies that are composed of 40 or more cells were counted. Results are expressed as the mean ± SEM for 4 cultures from a typical experiment. Similar results were seen in 3 independent experiments. *P < 0.05, significantly different from cultures treated with 100 pg/mL GM-CSF alone.

Figure 5

Effects of anti–GM-CSF on AXII-stimulated OCL formation. Human bone marrow (10⁵ cells/mL) was cultured with AXII (0–100 ng/mL) (a) or 1,25-(OH)₂D₃ (10^–8 M) (b) in the presence or absence of increasing concentrations of neutralizing antibodies to GM-CSF (0.001–5.0 μg/mL) in 96-well plates. Cells were fed weekly by changing half the media. After 3 weeks, multinucleated OCL-like cells were scored as described. Results represent the mean ± SEM of quadruplicate determinations from a typical experiment. The results were similar in 2 experiments.

Figure 6

Effects of AXII on GM-CSF mRNA expression in human bone marrow cultures. Total RNA was extracted from human bone marrow cultures treated with or without 100 ng/mL of AXII for 48 hours. GM-CSF mRNA expression was measured by RT-PCR using specific primers as described. β-actin was used as control for the amount and integrity of the RNA isolated. In parallel experiments, conditioned media were collected and used to quantify GM-CSF protein levels by ELISA, following the manufacturer’s protocol.

Figure 7

Northern blot analyses for expression of AXII mRNA. Northern blot analysis of AXII mRNA was performed using multiple cell lines, and the blot was probed with a ³²P-labeled AXII cDNA probe or a β-actin probe, as described in Methods.

Figure 8

Effects of AXII on GM-CSF gene expression in T cells. GM-CSF expression was assessed by RT-PCR analysis using total RNA isolated from T cells incubated with or without 0–1 μg/mL of AXII. Different cycles of PCR were used as described. β-actin was used as control for the integrity and amount of RNA in each sample. Densitometry analysis was performed to compare the expression levels of GM-CSF.

Figure 9

Binding of AXII to CD3⁺ cells. Sorting was performed as described in Methods. The purity of each sorted fraction, CD3⁺ (a) and CD3^– (b), was determined by analysis of the histogram. Binding of AXII and analyses were performed as described in Methods. Two independent experiments were performed and very similar results were obtained. Binding values were based on 100% conjugation of fluorescein to AXII. Ten thousand cells were analyzed. (c) Binding of AXII to CD3⁺ cells and CD3^– cells.

Figure 10

Analysis of AXII binding to flow-sorted CD4⁺ and CD8⁺ T-cell subsets. Sorting of bone marrow–derived T-cell subsets was performed as described in Methods. The purity of each sorted fraction (CD4⁺, CD8⁺) was determined by analysis of histograms, as was done for the CD3⁺ fraction. AXII-fluorescein binding and AXII competition assays were performed as described above for CD3⁺ (a) or CD3^– (b) bone marrow cells. Ten thousand cells were analyzed. Results are reported as the percent change in fluorescence when the cells were exposed to AXII fluorescein. Basal fluorescence was assigned as 100% to permit comparisons of the 2 T-cell subpopulations. Three independent experiments were performed and very similar results were obtained.

Figure 11

AXII binding to PSV10 cells. Cells (10⁵/0.1 mL of PBS) were used for each analysis, and 10,000 cells were collected and analyzed as described in Figure 9. Four experiments were performed and 1 typical experiment is shown. Binding values are based on the assumption of 100% conjugation of fluorescein to AXII.