The p85alpha subunit of class IA phosphatidylinositol 3-kinase regulates the expression of multiple genes involved in osteoclast maturation and migration

Abstract

Intracellular signals involved in the maturation and function of osteoclasts are poorly understood. Here, we demonstrate that osteoclasts express multiple regulatory subunits of class I(A) phosphatidylinositol 3-kinase (PI3-K) although the expression of the full-length form of p85alpha is most abundant. In vivo, deficiency of p85alpha results in a significantly greater number of trabeculae and significantly lower spacing between trabeculae as well as increased bone mass in both males and females compared to their sex-matched wild-type controls. Consistently, p85alpha(-/-) osteoclast progenitors show impaired growth and differentiation, which is associated with reduced activation of Akt and mitogen-activated protein kinase extracellular signal-regulated kinase 1 (Erk1)/Erk2 in vitro. Furthermore, a significant reduction in the ability of p85alpha(-/-) osteoclasts to adhere to as well as to migrate via integrin alphavbeta3 was observed, which was associated with reduced bone resorption. Microarray as well as quantitative real-time PCR analysis of p85alpha(-/-) osteoclasts revealed a significant reduction in the expression of several genes associated with the maturation and migration of osteoclasts, including microphathalmia-associated transcription factor, tartrate-resistant acid phosphatase, cathepsin K, and beta3 integrin. Restoring the expression of the full-length form of p85alpha but not the version with a deletion of the Src homology-3 domain restored the maturation of p85alpha(-/-) osteoclasts to wild-type levels. These results highlight the importance of the full-length version of the p85alpha subunit of class I(A) PI3-K in controlling multiple aspects of osteoclast functions.

FIG. 1.

Expression of p85 (p85α, p50α, p55α, and p85β) in wild-type (WT) and p85α^−/− OCs. OCps were cultured in the presence of M-CSF (10 ng/ml) and RANKL (100 ng/ml) for either 3 days or 6 days, after which the cells were harvested and subjected to Western blot analysis using a pan-anti-p85 antibody (this antibody recognizes all regulatory subunits of class I_A PI3-K). Arrows in the top panel indicate the level of expression of p85 regulatory subunits in wild-type and p85α^−/− OCs. The bottom panel demonstrates total levels of β-actin in each lane.

FIG. 2.

Deficiency of p85α in vivo results in increased bone mass. Representative μCT reconstructions from the distal third of the right femur in 18-week-old female (A) and male (B) wild-type (WT) and p85α^−/− mice. The anterior half of the bone has been digitally removed to reveal the trabecular bone compartment within the metaphysis. Note the more numerous trabeculae and the greater proximal encroachment of the trabecular network in the p85α^−/− mice. (C) μCT-derived measurements of the trabecular bone volume fraction (BV/TV), trabecular number (Tb.N), thickness (Tb.Th), and separation (Tb.Sp) revealed significantly more numerous trabeculae, with significantly less spacing between trabeculae, a common feature in high-bone-mass conditions (n = 20 mice including 10 wild-type [7 males and 3 females] and 10 p85α^−/− mice [7 males and 3 females]). Error bars represent ± 1 standard error of the mean. *, P < 0.01. (D) Quantitative analysis of the number of OC per bone volume in vivo in wild-type and p85α^−/− bone sections reacted for TRAP activity (n = 3). *, P < 0.05. TRAP 5b levels in the serum of 16- to 17-week-old (E) or 6- to 7-week-old (F) wild-type and p85α^−/− mice. Solid-phase immunofixed enzyme activity assay for the determination of OC-derived TRAP 5b in mouse serum was assessed as described in Materials and Methods. A significant increase in the serum TRAP 5b levels was noted in both old and young p85α^−/− mice compared to wild-type controls. For the 16- to 17-week age group, five wild-type and seven p85α^−/− mice were used. For the 6- to 7-week-old age group, 11 wild-type and 12 p85α^−/− mice were used. *, P < 0.01. (G to I) Trabecular bone turnover was assessed in the secondary spongiosa by measuring the extent of single label (sL.Pm) and double label (dL.Pm) perimeter and the area of bone (dL.Ar) between the calcein and alizarin labels. Derived histomorphometric parameters include mineralizing surface (MS/BS), a measure of active bone-forming surface; MAR, a measure of the rate of radial expansion of new bone; and the bone formation rate (BFR). Five wild-type and five p85α^−/− mice were used (P > 0.05).

FIG. 3.

Defective proliferation of p85α^−/− OCps. Wild-type (WT) and p85α^−/− OCps were cultured in the presence of increasing amounts of M-CSF (A) or RANKL (B) or both (C). After 2 days, proliferation was evaluated by a [³H]thymidine incorporation assay. Bars represent the mean [³H]thymidine incorporation in OCps (cpm ± standard deviation) from one representative experiment performed in triplicate. Similar results were observed in two independent experiments. *, P < 0.05 for wild-type versus p85α^−/− OCs.

FIG. 4.

Deficiency of p85α in OCps does not alter the survival of these cells. Serum- and cytokine-depleted OCps were stained with annexin V-phycoerythrin and 7-AAD and analyzed by flow cytometry. (A) Representative dot blot showing the percent survival of OCps as determined by the lack of staining of cells by either annexin V and/or 7-AAD (i.e., lower-left coordinate). (B) Bar graph demonstrating percentage of annexin V- and 7-AAD-negative cells at various time points in the absence of serum and cytokines.

FIG. 5.

Defective morphology and differentiation of p85α^−/− OCps. (A) Representative photomicrographs of OCps of the indicated genotypes generated in vitro following culture in M-CSF (10 to 100 ng/ml) and RANKL (100 ng/ml). OCps were identified by staining for TRAP activity. A representative field is shown. (B) Quantitative reduction in the number of multinucleated (>3 nuclei/field) TRAP⁺ OCs is shown. Bars represent the mean numbers of multinucleated cells (mean ± standard deviation) of one representative experiment performed in replicates of three (10 fields were counted per replicate). *, P < 0.05 for wild-type versus p85α^−/−. (C) Relative number of TRAP⁺ cells from five independent experiments is summarized in a line chart. *, P < 0.05 for wild-type versus p85α^−/− OCs. OD, optical density.

FIG. 6.

Impaired migration and adhesion in p85α^−/− OCs. (A) OC migration via αvβ3 integrin or OPN in the presence of M-CSF. (B) OC adhesion via OPN in the presence of M-CSF. Results represent mean ± standard error of the mean of four independent experiments. *, P < 0.01; **, P < 0.001 (comparing p85α^−/− versus wild-type OCs). WT, wild type.

FIG. 7.

Impaired bone resorption by p85α^−/− OCs. Bone resorptive activity was measured by a pit formation assay. (A) Representative photomicrographs of a bone resorption assay from two independent experiments following culture of OCps on dentine slices. The resorbed bone is stained dark blue. The number and area of resorbed regions, referred to as pits, are quantitated in panel B. Data are the mean from four independent experiments. *, P < 0.01, wild-type versus p85α^−/−. WT, wild type.

FIG. 8.

Altered actin organization in p85α^−/− OCs. Representative photomicrograph of OCps following staining with fluorescein isothiocyanate-conjugated phalloidin at magnifications of ×40 (frames 1 and 2) and ×100 (frames 3 and 4). Arrows in frames 1 and 2 indicate clusters. Arrows in frame 3 indicate belt-forming cells. The experiment was conducted on three independent occasions.

FIG. 9.

Reduced activation of PI3-K, Akt, and MAP kinase in p85α^−/− OCps in response to both M-CSF and RANKL. Wild-type (WT) and p85α^−/− OCps were starved and stimulated with a combination of M-CSF and RANKL (A), RANKL (100 ng/ml) (B), or M-CSF (10 ng/ml) (C) for 2 and 5 min. Equal amounts of cell lysates were subjected to a PI3-K lipid assay as described in the Materials and Methods section. Upper panels demonstrate quantitative reduction in the level of PI3-K activity in p85α^−/− OCps. Arrows in the bottom panels indicate the level of activation of PI3P and PIP2 in response to the indicated cytokines. (D and E) Reduced activation of Akt and Erk1/2 MAP kinase in p85α^−/− OCps. Wild-type and p85α^−/− OCps were starved and stimulated with M-CSF (10 ng/ml) or RANKL (100 ng/ml) or a combination of both. Equal amounts of cell lysates were subjected to Western blot analysis using an anti-phospho-Akt or an anti-phospho-Erk1/2 antibody. The amount of total Akt and Erk in each lane is indicated.

FIG. 10.

Quantitative RT-PCR analysis of OC-specific genes in wild-type (WT) and p85α^−/− OCps. Wild-type and p85α^−/− OCps were generated from bone marrow cultured for 6 days with RANKL (100 ng/ml) and M-CSF (10 ng/ml). Total mRNA was extracted as described in Materials and Methods. Expression of mRNA for TRAP, JDP2, cathepsin K (Ctsk), calcitonin receptor (Calcr), MMP-9, and integrin β3 (Itgβ3) was analyzed by real-time RT-PCR using β-tubulin mRNA as an endogenous control. Bars represent the mean ± standard deviation of one independent experiment performed in replicates of three. Similar results were seen in two independent experiments. *, P < 0.01, wild-type versus p85α^−/−.

FIG. 11.

Reexpression of p85α into p85α^−/− OCps restores normal OC formation in vitro. (A) Flow cytometry profiles demonstrating the percentage of transduced cells (as determined by EGFP expression; x axis) expressing either the empty vector (control) or HA-tagged full-length form of p85α (p85α-HA) or an HA-tagged version of p85α with a deletion of the SH3 domain (p85αΔSH3). (B) Wild-type and p85α^−/− cells expressing various versions of p85α indicated in panel A were sorted to homogeneity and subjected to Western blot (WB) analysis. Arrows in the top panel indicate the expression of various p85α constructs as determined by an anti-HA antibody (ab). The middle panel demonstrates a comparison of the level of expression of the HA-tagged full-length version of p85α in p85α^−/− cells with that of endogenous levels of p85α in wild-type cells using an antibody that recognizes only the SH3 domain of p85α. The bottom panel demonstrates the level of protein in each lane as determined by the expression of β-actin. (C) Representative photomicrographs of OC culture following TRAP staining. Wild-type and p85α^−/− OCps expressing the indicated constructs were cultured in the presence of M-CSF and RANKL for 6 days and subjected to TRAP staining. Shown is a representative field. Similar results were observed in two independent experiments. WT, wild type; α, anti.