Insulin-like growth factor-binding protein-2 is required for osteoclast differentiation

Abstract

Global deletion of the Igfbp2 gene results in the suppression of bone turnover. To investigate the role of insulin-like growth factor-binding protein-2 (IGFBP-2) in regulating osteoclast differentiation, we cultured Igfbp2(-/-) bone marrow cells and found a reduction in the number of osteoclasts and impaired resorption. Addition of full-length IGFBP-2 restored osteoclast differentiation, fusion, and resorption. To determine the molecular domains of IGFBP-2 that were required for this effect to be manifest, Igfbp2(-/-) bone marrow cells were transfected with constructs in which the heparin-binding (HBD) or the IGF-binding domains of IGFBP-2 were mutated. We found that both domains were necessary for osteoclastogenesis because expression of the mutated forms of either domain failed to support the formation of functionally mature osteoclasts. To discern the mechanism by which IGFBP-2 regulates osteoclast formation, PTEN abundance and phosphorylation status as well as AKT responsiveness to IGF-I were analyzed. Igfbp2(-/-) cells had elevated levels of PTEN and phospho-PTEN compared with controls. Expression of wild-type IGFBP-2 reduced the level of PTEN to that of wild-type cells. Cells expressing the IGF-binding mutant showed suppression of PTEN and phospho-PTEN equivalent to the wild-type protein, whereas those expressing the IGFBP-2 HBD mutant showed no PTEN suppression. When the ability of IGF-I to stimulate AKT activation, measured by Thr³⁰⁸ and Ser⁴⁷³ phosphorylation, was analyzed, stimulation of Ser⁴⁷³ in response to IGF-I in preosteoclasts required the presence of intact IGFBP-2. This effect was duplicated by the addition of a CK2 inhibitor that prevents the phosphorylation of PTEN. In contrast, in fully differentiated osteoclasts, stimulation of Thr³⁰⁸ phosphorylation required the presence of intact IGFBP-2. We conclude that IGFBP-2 is an important regulator of osteoclastogenesis and that both the heparin- and the IGF-binding domains of IGFBP-2 are essential for the formation of fully differentiated and functional osteoclasts.

Fig. 1. In vitro osteoclastogenesis

Representative images of TRAP5b+ osteoclast cells in the male and female Igfbp2^−/− and Igfbp2^+/+ from 7 day bone marrow cultures (left and middle panels). Igfbp2^−/− cultures were also treated with 200 ng/ml IGFBP-2 (right) rescuing the null phenotype; all images captured at 20X magnification.

Fig. 2. Bone resorption in vitro

Igfbp2^−/− and Igfbp2^+/+ osteoclast cultures plated on the Corning Osteo Assay Surface 96 well plate. Top panels are representative images, at 20X magnification, of osteoclasts in resorbed areas of the plate. A) Igfbp2^+/+ B) Igfbp2^−/− or C) Igfbp2^−/− treated with IGFBP-2 (200 ng/ml). Bottom panels are examples of von Kossa stained wells (taken at 10X magnification) after 14 days in culture. The unresorbed matrix is black while the resorbed surface is white.

Fig. 3. In vitro osteoclastogenesis following expression of wild-type and mutant forms of IGFBP-2

A) Diagram showing regions of IGFBP-2 including the two regions (IGF binding and unique HBD) in which amino acid substitutions were performed. B) Induction of osteoclasts in bone marrow cultures from Igfbp2^−/− mice transduced with either a control construct (#1) or constructs expressing wild-type (#2) or the non-IGF binding mutant of IGFBP-2 (#3) or the heparin binding domain substitution mutant (#4). At day 10, the cells were fixed and stained for TRAP5b. C) Equal amounts of cell lysates from Igfbp2^−/− bone marrow cultures expressing the different forms of IGFBP-2 obtained at day 10, were separated by SDS-PAGE and immunoblotted with the indicated antibody. D) Induction of osteoclast cells in bone marrow cultures of Igfbp2^−/− and treated with vehicle (#1), IGFBP-2 (#2) or heparin binding domain peptide (#3). At day 10, the cells were fixed and stained for TRAP5b.

Fig. 4. Regulation of PTEN by IGFBP-2

A) Equal amounts of cell lysates from Igfbp2 ^−/− (−/−) and Igfbp2^+/+ (WT) derived bone marrow cultures obtained 10 days after stimulation with M-CSF and RANKL (+) were separated by SDS-PAGE and immunoblotted with the indicated antibody. AKT was immunoblotted as a loading control. B) Equal amounts of cell lysates from Igfbp2 ^−/− (−/−) and Igfbp2^+/+(WT) derived bone marrow cultures treated with IGFBP-2 obtained 10 days after stimulation with M-CSF and RANKL (+) were separated by SDS-PAGE and immunoblotted with the indicated antibody. Some cultures received IGFBP-2 (200 ng/ml for 10 days). C) Equal amounts of cell lysates from Igfbp2 ^−/− derived bone marrow cultures treated with IGFBP-2 or HBD peptide for 10 days with M-CSF and RANKL (+) were separated by SDS-PAGE and immunoblotted for PTEN. Immunoblotting for total AKT was used to demonstrate that there was no significant difference in the total amount of protein in each lysate sample. D) Equal amounts of cell lysates from the M-CSF primed bone marrow cultures of Igfbp2^−/− cells that had been transduced with either a control construct (−) or constructs expressing wild-type (WT), heparin binding domain substitution (HBD) or the non-IGF binding mutant of IGFBP-2 (IGF) and treated for 10 days with RANKL and M-CSF were separated by SDS-PAGE and immunoblotted for PTEN. E) Lysates from Igfbp2^−/− bone marrow cultures that were exposed to IGFBP-2 (BP-2) or the caseine kinase inhibitor (CK2) and RANKL+ M-CSF for 10 days were separated by SDS-PAGE and immunoblotted with the indicated antibody. The bar graphs show the results of scanning densitometry units obtained for 3 individual experiments.

Fig. 5. Regulation of AKT phosphorylation by IGFBP-2

Equal amounts of cell lysate from Igfbp2 ^−/− derived bone marrow cultures 5 (A) and 10 (B&C) days after stimulation with RANKL and M-CSF followed by short-term treatment with IGF-I (50ng/ml) (10 min) were separated by SDS-PAGE and immunoblotted with the indicated antibody. IGFBP-2 or the CK2 inhibitor were added for 10 days where indicated. The graph shows the results of the phospho-AKT immunoblots. The blots shown are derived from discontinuous lanes on the same blot. Immunoblotting for total AKT was used to demonstrate that there was no significant difference in the total amount of protein in each lysate sample.