Bone remodeling comprises balanced activities between osteoclasts and osteoblasts, which is regulated by various factors, including hormones and cytokines. We previously reported that IL-3 inhibits osteoclast differentiation and pathological bone loss. IL-3 also enhances osteoblast differentiation and bone formation from mesenchymal stem cells. However, the role of IL-3 in regulation of osteoblast–osteoclast interactions and underlying mechanisms is not yet delineated. In this study, we investigated the role of IL-3 on the regulation of osteoblast-specific molecules, receptor activator of NF-κB ligand (RANKL), and osteoprotegerin (OPG) that modulate bone homeostasis. We found that IL-3 increases RANKL expression at both the transcriptional and translational levels, and it showed no effect on OPG expression in calvarial osteoblasts. The increased RANKL expression by IL-3 induces mononuclear osteoclasts; however, it does not induce multinuclear osteoclasts. Interestingly, IL-3 decreases soluble RANKL by reducing ectodomain shedding of membrane RANKL through downregulation of metalloproteases mainly a disintegrin and metalloproteinase (ADAM)10, ADAM17, ADAM19, and MMP3. Moreover, IL-3 increases membrane RANKL by activating the JAK2/STAT5 pathway. Furthermore, IL-3 enhances RANKL expression in mesenchymal stem cells of wild-type mice but not in STAT5a knockout mice. Interestingly, IL-3 restores RANKL expression in adult mice by enhancing bone-specific RANKL and decreasing serum RANKL. Furthermore, IL-3 increases the serum OPG level in adult mice. Thus, our results reveal, to our knowledge for the first time, that IL-3 differentially regulates two functional forms of RANKL through metalloproteases and the JAK2/STAT5 pathway, and it helps in restoring the decreased RANKL/OPG ratio in adult mice. Notably, our studies indicate the novel role of IL-3 in regulating bone homeostasis in important skeletal disorders.
Bone maintains its structural and functional integrity by the physiological process of bone remodeling, which is regulated by interactions between bone-resorbing osteoclasts and bone-forming osteoblasts. Osteoclasts are derived from hematopoietic stem cells of monocyte/macrophage lineage, and osteoblasts differentiate from mesenchymal stem cells (MSCs) (1). Osteoblasts regulate bone remodeling by producing both stimulatory and inhibitory factors that tightly regulate osteoclast formation and activity. Stimulatory factors include glycoproteins such as M-CSF and receptor activator of NF-κB (RANK) ligand (RANKL) (2–4). RANKL is also expressed by various cell types such as activated T cells (5, 6), mesenchymal cells (7), keratinocytes (8), B lymphocytes (9, 10), mammary epithelial cells (11), vascular endothelial cells (12), cancer cells (13), and microglia (14). RANKL is a 35-kDa protein found both as a transmembrane glycoprotein on the surface of cells and secreted soluble protein. Binding of RANKL to its receptor RANK on the surface of osteoclast precursors initiates signals (15), which lead to fusion, maturation, survival, and activation of osteoclasts (16, 17). Osteoblasts also secrete inhibitory protein known as osteoprotegerin (OPG), which is a decoy receptor for RANKL that prevents the binding of RANKL to RANK, thereby inhibiting osteoclast differentiation and activation (18).
Any alterations in the RANKL/OPG ratio modulate the bone remodeling in skeletal diseases such as osteoporosis, Paget disease, and osteoarthritis (19). The age-related bone loss due to systemic changes, including growth factors, sex steroids, and parathyroid hormone, may modulate RANKL and OPG expression in vivo (20). It has been reported that an increase in serum parathyroid hormone level with postmenopausal aging causes stimulation of RANKL and inhibition of OPG expression (21, 22). These findings indicate that regulation of RANKL and OPG is very important for the prevention and treatment of bone loss in skeletal diseases and also in aging.
Cytokines secreted by immune and other cell types play an important role in regulation of bone remodeling (23). Previously, we reported that IL-3, a cytokine secreted by activated T cells (24), is a potent inhibitor of osteoclastogenesis and inhibits both RANKL and TNF-α–induced osteoclast formation and bone resorption (25–28). IL-3 also increases in vitro osteoblast differentiation and matrix mineralization from human MSCs, and it enhances ectopic bone formation in immunocompromised mice (29). Precise equilibrium between osteoblast and osteoclast activity is crucial to maintain the structural and functional integrity of bone, which is regulated by RANKL and OPG (30). In this study, we investigated the role of IL-3 in modulation of RANKL and OPG expression in mouse calvarial osteoblasts and also in adult mice.
We found that IL-3 increases the expression of membrane-bound RANKL and decreases secretion of soluble RANKL in vitro in calvarial osteoblasts. The expression of OPG in osteoblasts was not affected by IL-3. Furthermore, IL-3 decreases soluble RANKL in osteoblasts by downregulation of metalloproteases, a disintegrin and metalloproteinases (ADAMs) such as ADAM10, ADAM17, ADAM19, and MMP3, and it increases membrane-bound RANKL by activating the JAK2/STAT5 pathway. Interestingly, IL-3 improves the disturbed RANKL/OPG ratio in serum of adult mice. Thus, IL-3 differentially regulates two functional forms of RANKL in both in vitro and in vivo conditions.
Materials and Methods
BALB/c mice of 2–5 d, 6–8 wk, 3 mo, and 1 y old were obtained from the Experimental Animal Facility of the National Centre for Cell Science (Pune, India). STAT5a [C.129S (B6)-Stat5atm1Mam/J] and STAT5b (C.129-Stat5btm1Hwd/J) knockout mice were obtained from The Jackson Laboratory. Water and food were provided ad libitum. All protocols involving animal use were approved by an Institutional Animal Ethics Committee.
Reagents and culture conditions
RANKL Ab and fluorochrome-conjugated secondary Abs were obtained from Abcam. Polyclonal Abs for IL-3Rα, OPG, pERK1/2, ERK1/2, pJAK2, JAK2, pSTAT5a/b, STAT5, and β-actin were obtained from Santa Cruz Biotechnology. The HRP-conjugated secondary Abs were from Bangalore Genei. Fluorochrome-conjugated anti-mouse RANKL (clone IK22/5) and anti-mouse IL-3Rα (clone 5B11) Abs were obtained from BioLegend. Recombinant mouse IL-3 was obtained from BD Biosciences. 1α,25-Dihydroxyvitamin D3 was obtained from Sigma-Aldrich. FCS, l-glutamine, TRIzol reagent, cDNA synthesis kit, and SYBR Green were obtained from Invitrogen. Collagenase and dispase were purchased from MP Biomedicals. RANKL and OPG ELISA kits were obtained from R&D Systems.
All cultures were incubated in growth medium containing α-MEM, heat-inactivated FCS (10%), l-glutamine (2 mM), penicillin (100 IU/ml), streptomycin (100 μg/ml), and osteogenic factors β-glycerophosphate (10 mM) and ascorbic acid (50 μg/ml, all from Sigma-Aldrich). All incubations were performed at 37°C in a humidified atmosphere of 5% CO2 in air.
Isolation of calvarial osteoblasts
Mouse calvarial osteoblasts were isolated from 2- to 5-d-old BALB/c mice using a modified sequential digestion method as described previously (31). Briefly, surgically resected calvariae were cleaned off adherent soft tissues and subjected to five sequential (5, 15, 10, 10, and 5 min) digestions in enzyme solution containing 0.1% collagenase and 0.2% dispase at 37°C. Cells released from second to fourth digestions were pooled, centrifuged, and resuspended in growth medium as described above. Calvarial osteoblasts of passage 2 were used in all experiments.
Isolation of bone marrow–derived osteoclast precursors
The stromal cell–free and M-CSF–dependent osteoclast precursors were isolated from 6- to 8-wk-old BALB/c mice as previously described (25). Briefly, femoral and tibial bones were aseptically removed and cleaned by removing adherent soft tissues. The bone ends were cut, and the bone marrow cavity was flushed out with α-MEM from one end of the bone using a sterile 21-gauge needle. Bone marrow cells were washed twice and incubated for 24 h in the presence of M-CSF (10 ng/ml) at a density of 3 × 105 cells/ml in a 75-cm2 flask. After 24 h, nonadherent cells were collected and layered on a Ficoll-Hypaque (Sigma-Aldrich) gradient. Cells at the gradient interface were collected, washed twice, and used for further experiments.
Coculture model for osteoclastogenesis
To evaluate the effect of IL-3 on osteoclast differentiation in a coculture model, mouse calvarial osteoblasts (2 × 104 cells per well) and bone marrow–derived osteoclast precursors (2 × 105 cells per well) were cocultured in 48-well plates containing α-MEM and 10% FCS with or without IL-3 for 7 d (32). Vitamin D3 was used as positive regulator of osteoclastogenesis in a coculture system that is known to increase RANKL expression on osteoblasts. Osteoclast formation was evaluated by tartrate-resistant acid phosphatase (TRAP) staining.
Isolation of adipose tissue–derived MSCs
Mouse adipose tissue–derived MSCs (AT-MSCs) were isolated from 10- to 12-wk-old wild-type, STAT5a, and STAT5b knockout mice as described previously (33). Briefly, s.c. adipose tissue was digested with 2 mg/ml collagenase (type 1A; Sigma-Aldrich) in PBS at 37°C for 15–20 min. The cell suspension obtained was centrifuged, resuspended in α-MEM containing 10% FCS, and seeded in a culture flask. After 72 h nonadherent cells were discarded and adherent cells were cultured until they attained 80–90% confluency. Homogeneous populations of AT-MSCs from passage 2 or 3 were used in all further experiments.
The proliferation of mouse calvarial osteoblasts in the presence of IL-3 was examined using an MTT assay. After the incubation period, cell culture media were replaced with 100 μl of MTT (0.5 mg/ml) solution and cells were further incubated at 37°C for 3–4 h. The MTT solution was removed and formazan crystals formed were dissolved in 100 μl of DMSO. The absorbance was measured at 570 nm.
Quantitative real-time PCR
Expression of RANKL, OPG, M-CSF, ADAM10, ADAM17, ADAM19, MMP3, MMP14, TRAP, cathepsin K, calcitonin receptor, integrin β3 and GAPDH was assessed by real-time PCR analysis. RNA was isolated from cells by the TRIzol reagent method (Invitrogen). Two micrograms of total RNA was used for synthesis of cDNAs by reverse transcription (cDNA synthesis kit). For real-time PCR, a 10-μl reaction mixture containing SYBR Green and 10 pmol of each primer were used and PCR was set using the StepOnePlus system (Applied Biosystems). The amplification was performed using one cycle of 95°C for 10 min and 40 cycles of denaturation at 95°C for 15 s, primer annealing and extension at 60°C for 60 s, followed by melt curve analysis. The primer sequences (IDT) used are summarized in Supplemental Table I. Each reaction was run in duplicates and data were analyzed for fold change using the comparative 2−∆∆Ct method.
Mouse calvarial osteoblasts were cultured on glass coverslips in 24-well plates in the absence or presence of IL-3 for the indicated time points. The cells were washed twice with PBS, fixed with 4% paraformaldehyde for 10 min, and permeabilized with 0.1% Triton X-100 in PBS for 5 min. Staining for RANKL, OPG and IL-3Rα was carried out using primary Abs followed by fluorochrome-conjugated secondary Abs. After washing cells were mounted using Dabco (Sigma-Aldrich) and viewed with a Zeiss LSM 510 confocal microscope equipped with argon and helium lasers.
Cells were cultured with or without IL-3 for the indicated time points. At the end of the culture period adherent cells were harvested from culture dishes using cell dissociation buffer (Life Technologies), and the few remaining cells were dislodged by gentle scraping on ice. Cell surface staining was performed by incubating 105 cells in 100 μl of PBS with fluorochrome-conjugated Abs. For intracellular staining, cells were first permeabilized with permeabilization buffer (0.1% Triton X-100 in 1× PBS) and then stained with RANKL and OPG Abs. Cells were washed, acquired, and analyzed with a BD FACSCalibur. Data were analyzed using CellQuest Pro software (Becton-Dickinson). The results are expressed as the percentage of cells and the change in mean fluorescence intensity (∆MFI).
Cell culture supernatants were harvested after the indicated time of incubation, centrifuged, and immediately frozen at −80°C until further analysis. All samples were thawed immediately prior to evaluation by ELISA. Soluble RANKL and OPG proteins were measured in supernatant and serum using the respective Quantikine mouse ELISA kits. The procedure was carried out according to the manufacturer’s instructions and the absorbance was measured at 450 nm with a correction wavelength of 540 nm.
Cells were seeded at a density of 5 × 104 cells/cm2 in α-MEM containing 10% FCS and cultured for the indicated time periods in the absence or presence of IL-3. The cells were lysed in RIPA buffer containing protease inhibitors, and proteins were estimated using the BCA method. Protein samples were then subjected to 12% SDS-PAGE. The proteins were transferred from gels onto a nitrocellulose membrane for immunoblot analysis. Blocking was performed with 5% nonfat dry milk in TBS buffer. The membrane was then incubated with primary Ab (1:1000) for 3 h. After washing, the membranes were incubated with HRP-conjugated secondary Abs, and labeled proteins were detected using ECL reagents (Amersham Biosciences). Relative intensities of protein bands were analyzed by densitometry using ImageJ software.
Cells were washed gently in PBS and fixed in 10% formalin for 10 min at room temperature. TRAP staining was performed at 37°C for 10–15 min by incubating cells in TRAP buffer (70 mM sodium acetate, 30 mM acetic acid, 0.1 mg/ml napthol AS-MX phosphate disodium salt, 0.1% Triton X-100 at pH 5). The number of TRAP-positive mononuclear and multinuclear (three or more nuclei) osteoclasts was counted.
In vivo analysis of RANKL and OPG
To investigate the in vivo role of IL-3 on RANKL and OPG expression, young mice of 3 mo old and adult mice of 1 y old were used. Adult mice were injected i.p. with PBS or IL-3 (3 μg per mouse per day) for 5 d. Young mice injected with PBS were used as a control to compare the changes in RANKL and OPG with adult mice. On day 6, bone and serum samples were collected. Bone-specific RANKL expression was analyzed by immunoblotting from femur and tibia after removal of bone marrow. Serum RANKL and OPG levels were analyzed by ELISA.
Results are represented as mean ± SEM. Statistical significance was calculated using one-way ANOVA with a subsequent post hoc Bonferroni test for multiple comparisons. A p value <0.05 was considered significant.
IL-3 enhances RANKL expression in calvarial osteoblasts
To investigate the role of IL-3 on RANKL and OPG expression we first confirmed the expression of IL-3Rα on mouse calvarial osteoblasts. Calvarial osteoblasts showed expression of IL-3Rα as analyzed by immunofluorescence and Western blotting (Supplemental Fig. 1A, 1B). To further quantify IL-3Rα expression on osteoblasts, we performed flow cytometry analysis and observed that ∼48.97% cells were positive for IL-3Rα expression (Supplemental Fig. 1C). To evaluate the effect of IL-3 on RANKL and OPG in a time-dependent manner, we cultured calvarial osteoblasts for 12, 24, 48, and 72 h in osteogenic media containing β-glycerophosphate (10 mM) and ascorbic acid (50 μg/ml) in the absence or presence of IL-3 (100 ng/ml) or vitamin D3 (10−8 M), and expression of RANKL and OPG was examined by quantitative real-time PCR. Vitamin D3 was used as a positive stimulator of RANKL expression. It was observed that vitamin D3 significantly enhanced RANKL expression at 24 and 48 h and IL-3 significantly increased RANKL expression at 24 h (Fig. 1A). Vitamin D3 significantly decreased OPG expression at 24 h. However, IL-3 did not show any effect on OPG expression at all time points (Fig. 1B). We further observed that vitamin D3 significantly increased the RANKL/OPG ratio at 48 h. However, IL-3 did not show any significant effect on the RANKL/OPG ratio (Fig. 1C). We further evaluated whether IL-3 has any role in regulation of another glycoprotein, M-CSF, which is also secreted by osteoblasts and required for survival and early stages of osteoclast differentiation. Similar to OPG, IL-3 showed no effect on expression of M-CSF at all time points (Fig. 1D).
To rule out the possibility that osteogenic factors could induce RANKL expression in osteoblasts, we evaluated the dose-dependent effect of IL-3 on RANKL and OPG expression in the absence of osteogenic factors. Cells were cultured for 24 h with various concentrations of IL-3 in the absence of osteogenic factors, and the expression of RANKL and OPG was analyzed by real-time PCR. We found that IL-3 increased the RANKL expression and showed no effect on OPG (Supplemental Fig. 2A, 2B). These results suggested that the effect of IL-3 on RANKL expression is direct and not through osteogenic factors. However, to mimic an in vivo–like situation we used osteogenic factors in all further experiments.
Next, to examine the effect of IL-3 on RANKL at the protein level, osteoblasts were cultured for 24 and 48 h with IL-3 or vitamin D3, and expression of RANKL was evaluated by immunoblotting and the relative intensity was analyzed by ImageJ software. It was observed that vitamin D3 significantly increased RANKL expression at 48 h. Similar to vitamin D3, IL-3 also increased RANKL expression significantly at 48 h (Fig. 1E, 1F). To further confirm the effect of IL-3 on RANKL protein, we evaluated total protein expression at 48 h by immunofluorescence microscopy. IL-3 significantly enhanced RANKL expression at the protein level (Fig. 1G). Fig. 1H shows the fluorescence intensity of RANKL analyzed by ImageJ software. All of these results suggest that IL-3 significantly increases RANKL at both the gene and protein levels. We also observed that IL-3 has no effect on proliferation of osteoblasts in a dose- and time-dependent manner, and also IL-3 was not toxic to the cells at 100 ng/ml concentration (Supplemental Fig. 3).
Effect of IL-3 on functional forms of OPG and RANKL
OPG is a decoy receptor for RANKL and the only member of TNF receptor superfamily, which is a secretary protein (18). Because IL-3 showed no effect on OPG at the gene level, we further confirmed the effect of IL-3 on OPG expression at the protein level. OPG protein was evaluated by immunofluorescence by culturing osteoblasts for 48 h with IL-3, as well as with flow cytometry by culturing cells for 12, 24, and 48 h with IL-3. Fig. 2A and 2C show that IL-3 has no effect on OPG expression at the protein level. Fig. 2B represents fluorescence intensity of OPG protein expression. Fig. 2D and 2E show the average percentage of OPG-expressing cells and ∆MFI, respectively. Furthermore, the effect of IL-3 on the functional secretary form of OPG was evaluated by incubating osteoblasts for 12, 24, 48, and 72 h with IL-3, and OPG secretion in the culture supernatant was analyzed by ELISA. IL-3 showed no effect on OPG secretion at all time points (Fig. 2F). These results confirm that IL-3 has no effect on both intracellular and functional OPG expression.
Because IL-3 increased RANKL expression at both the mRNA and protein levels, we further investigated whether this effect of IL-3 on RANKL is sustained at the functional level. Osteoblasts were incubated for 48 h with IL-3 or vitamin D3, and expression of total and membrane-bound RANKL was analyzed by flow cytometry. We found that both IL-3 and vitamin D3 significantly increased expression of total and membrane RANKL (Fig. 3A, 3B). Fig. 3C and 3D show the average percentage of cells and ∆MFI of total RANKL. Fig. 3E and 3F shows the average percentage of cells and ∆MFI of membrane-bound RANKL.
IL-3 differentially regulates RANKL expression
RANKL is an osteoclast differentiation factor that stimulates the differentiation, maturation, and activation of osteoclasts from its precursors of monocyte/macrophage lineage. Because IL-3 increases RANKL at both the gene and protein levels, and also increases its membrane expression in osteoblasts, we further evaluated the role of IL-3 in osteoclastogenesis using a coculture model of osteoclast differentiation. Calvarial osteoblasts (2 × 104 cells per well) were cocultured with bone marrow–derived osteoclast precursors (2 × 105 cells per well) in 48-well plates in a contact-dependent manner in the presence of IL-3 (100 ng/ml). Vitamin D3 (10−8 M) was used as positive control for induction of osteoclastogenesis. After 7 d, TRAP-positive mononuclear and multinuclear cells were counted. We observed that vitamin D3 induced formation of both mononuclear and multinuclear TRAP-positive cells (Fig. 4A). IL-3 significantly increased the number of TRAP-positive mononuclear cells (Fig. 4B); however, it was unable to induce formation of TRAP-positive multinuclear osteoclasts (Fig. 4C). We also analyzed the effect of IL-3 on osteoclast-specific genes such as TRAP, cathepsin K, calcitonin receptor, and integrin β3 in coculture conditions. Vitamin D3 significantly increased the expression of osteoclast-specific genes; however, IL-3 did not increase the expression of these genes (Fig. 4D–G). These results suggest that despite increasing RANKL expression, IL-3 was unable to induce formation of mature osteoclasts. To investigate the possible reasons for this action of IL-3, we further evaluated the effect of IL-3 on another functional form of RANKL by analyzing the secretion of its soluble form in culture supernatants. Osteoblasts were incubated for 12, 24, 48, and 72 h with IL-3 or vitamin D3 and soluble RANKL in the culture supernatant was analyzed by ELISA. Vitamin D3 significantly increased soluble RANKL at 48 and 72 h. Interestingly, we found that IL-3 decreases secretion of soluble RANKL at 24 and 48 h, and significant decrease was seen at 72 h (Fig. 4H). These results suggest that vitamin D3 increased both functional forms of RANKL. However, IL-3 increases membrane RANKL and decreases its soluble form, and hence it was unable to induce mature osteoclast formation in the coculture model.
Differential regulation of RANKL expression by IL-3 is through metalloproteases
Soluble RANKL is formed by the proteolytic cleavage of membrane-bound RANKL (34, 35). This process is called ectodomain shedding, which is regulated by various ADAMs such as ADAM10, ADAM17, and ADAM19 (34, 36, 37). To further investigate the mechanism of downregulation of soluble RANKL by IL-3, osteoblasts were incubated for 12, 24, 48, and 72 h with IL-3 and mRNA expression of ADAM10, ADAM17, and ADAM19 was examined by real-time PCR. We found that IL-3 significantly downregulated the expression of ADAM10 at 24 and 72 h (Fig. 5A). IL-3 significantly decreased ADAM17 expression at 48 and 72 h, which is the main metalloprotease that regulates production of soluble RANKL from its membrane form (Fig. 5B), and also significantly decreases ADAM19 expression at 72 h (Fig. 5C). Besides ADAMs, MMPs such as MMP3 and MMP14 also play a crucial role in ectodomain shedding of membrane RANKL (38, 39). Therefore, we evaluated the effect of IL-3 on MMP3 and MMP14 expression under similar culture conditions. IL-3 significantly decreases MMP3 expression at 24 h (Fig. 5D) and showed no effect on MMP14 gene expression (Fig. 5E). These results suggest that although IL-3 is capable of increasing RANKL expression at transcript and membrane-bound protein levels, it decreases soluble RANKL by downregulation of metalloproteases that eventually hinder the cleavage of soluble RANKL from its membrane form.
IL-3 activates the JAK2/STAT5 pathway and regulates RANKL expression through STAT5 isoforms
It is well established that among various signaling pathways, JAK2/STAT5 is an important pathway activated by IL-3 (24). To further investigate the molecular mechanism for the increase in membrane RANKL by IL-3, we cultured osteoblasts for 15, 30, 60, and 120 min with IL-3, and activation of JAK2/STAT5 was evaluated by Western blotting. We observed that IL-3 phosphorylates JAK2 at 30, 60, and 120 min and STAT5a/b at 30 and 60 min (Fig. 6A). Besides JAK2, few studies show that ERK1/2 also plays a crucial role in phosphorylation of STAT5 (40). We observed that IL-3 also increases phosphorylation of ERK1/2 in osteoblasts (Fig. 6A). These results indicate that IL-3 stimulates phosphorylation of STAT5 by both JAK2- and ERK1/2-dependent pathways. The STAT5 transcription factor regulates expression of various genes (41). At the functional level not only phosphorylation of STAT5 but also translocation of its phosphorylated form into the nucleus is crucial for its activity. To confirm this we cultured osteoblasts with IL-3 for 30 min, and analysis of STAT5 translocation was evaluated by immunofluorescence. IL-3 increases the translocation of pSTAT5 into the nucleus (Fig. 6B). Thus, IL-3 activates STAT5 via JAK2 and also induces its translocation into nucleus.
RANKL expression is regulated by JAK2/STAT5a in mammary epithelial cells (11), and IL-3 increases phosphorylation of both isoforms of STAT5 (42, 43). To further investigate which isoform of STAT5 play a prominent role in IL-3 regulation of RANKL expression, we used AT-MSCs isolated from wild-type, STAT5a, and STAT5b knockout mice. AT-MSCs were selected over osteoblasts or bone marrow MSCs because of ease of isolation and expansion as a homogeneous population that differentiates into osteoblasts (33). AT-MSCs showed all characteristics of MSCs and the differentiation potential into functional osteoblasts (data not shown). AT-MSCs were cultured for 24 or 48 h with IL-3 and the expression of RANKL was analyzed at both the gene and protein levels, and OPG was analyzed at the mRNA level. IL-3 significantly increased RANKL expression at the mRNA (Fig. 6C) and protein (Fig. 6F, 6G) levels in wild-type AT-MSCs, but it showed no change in RANKL expression in STAT5a knockout AT-MSCs. IL-3–treated cells from STAT5b knockout mice showed no effect on RANKL expression at the mRNA level (Fig. 6C) but significantly decreased RANKL expression at the protein level (Fig. 6F, 6G). No comparative effect of IL-3 was observed on OPG and M-CSF expression between wild-type, STAT5a, and STAT5b knockout mice (Fig. 6D, 6E). Our results indicate that STAT5a plays a crucial role in IL-3 regulation of RANKL expression.
In vivo role of IL-3 in regulation of the RANKL/OPG ratio in adult mice
Because IL-3 regulates RANKL expression differentially in vitro in osteoblasts, we further analyzed the in vivo effect of IL-3 on RANKL and OPG modulation. We first compared RANKL and OPG expression in young and adult female and male mice. After removal of bone marrow, femur and tibia bones were used to analyze the expression of bone-specific RANKL. Serum was used to analyze the soluble RANKL. It was observed that bone-specific RANKL expression was significantly decreased in both adult female and male mice as compared with young mice (Fig. 7A). Fig. 7B represents the relative intensity of immunoblots measured by ImageJ software. We observed that serum RANKL expression was significantly increased in both adult female and male mice (Fig. 7C). These results suggested the differential regulation of RANKL expression in adult mice.
Although both female and male mice show similar changes in RANKL expression associated with bone and serum, to evaluate the role of IL-3 on RANKL, male mice were selected over female mice to avoid age-related hormonal changes. Adult male mice were injected i.p. with PBS or IL-3 (3 μg per mouse per day) for 5 d. Young mice injected with PBS were used as a control to compare the age-related changes in RANKL. Immunoblotting analysis showed a significant decrease in bone-specific RANKL in adult mice and it was markedly increased by IL-3 (Fig. 8A). Fig. 8B represents the relative quantitative analysis of immunoblots. Interestingly, IL-3 significantly decreased RANKL and increased OPG in the serum of adult mice (Fig. 8C, 8D). Furthermore, the increased RANKL/OPG ratio in adult mice was decreased by IL-3 (Fig. 8E). All of these results indicate that IL-3 helps in restoring the disturbed RANKL and OPG expression in adult mice.
Osteoclast differentiation involves interactions between osteoblasts/stromal cells and osteoclast progenitors of hematopoietic cell lineage (1). RANKL and M-CSF, expressed by osteoblasts, and the cognate receptor RANK, present on the surface of osteoclast progenitors, play an important role in osteoblast and osteoclast interactions (7). Osteoblasts also secrete OPG that functions as a decoy receptor for RANKL and prevent the binding of RANKL to RANK that eventually inhibits osteoclast formation (18). Thus, the RANKL–RANK–OPG axis is very important for regulation of osteoclastogenesis and maintenance of bone homeostasis (44). In important skeletal diseases such as osteoporosis, osteoarthritis, and bone cancers the RANKL/OPG ratio increases (19). Many osteotropic factors including hormones and cytokines regulate osteoclastogenesis and bone resorption indirectly by modulating the expression and/or activity of RANKL, OPG, and RANK (19). We have previously shown that IL-3 is a potent inhibitor of osteoclast differentiation and pathological bone loss (25–28), and it also enhances osteoblast differentiation and bone formation from human MSCs (29). IL-3 inhibits osteoclast differentiation by direct action on osteoclast precursors and also inhibits RANK expression at the posttranslational level (45). Thus, IL-3 has a novel role in inhibition of pathological bone loss and enhancement of bone formation. However, it is not yet known whether IL-3 regulates crosstalk between osteoblasts and osteoclasts and has any indirect role in regulating bone loss. In the present study, we evaluated both the in vitro and in vivo roles of IL-3 on the regulation of RANKL and OPG expression.
IL-3 significantly increased RANKL expression in calvarial osteoblasts at both the transcriptional and translational levels. Our results are in agreement with a previous report that indicated that IL-3 increases RANKL expression in human basophils (46). However, this increase in RANKL expression by IL-3 was not sufficient to induce the formation of multinuclear mature osteoclasts. IL-3 showed no effect on both gene and protein expression of OPG in osteoblasts. We further analyzed the role of IL-3 on functional forms of RANKL. RANKL has two different functional forms, one membrane anchored and another a secreted soluble form (35). Surprisingly, IL-3 differentially regulates two functional forms of RANKL. IL-3 increases the expression of membrane-bound RANKL and simultaneously downregulates the secretion of soluble RANKL. Our comparative studies showed that both vitamin D3 and IL-3 significantly stimulated RANKL expression at transcriptional and translational levels. However, only vitamin D3 significantly stimulated secretion of soluble RANKL. These results indicate that vitamin D3, a well-known stimulator of osteoclast formation in coculture, stimulates osteoclastogenesis by inducing both membrane and soluble RANKL, whereas IL-3 stimulates membrane-bound RANKL and inhibits soluble RANKL.
Similar to many other transmembrane proteins, RANKL undergoes proteolysis and it is released from its plasma membrane by a process called ectodomain shedding (36, 37). The ectodomain shedding is a highly regulated process that affects biological and pathological significance of many transmembrane proteins. For example, cytokine TNF-α and epidermal growth factor exert their paracrine and endocrine effects only after ectodomain shedding (47, 48). Similar to other members of the TNF family, such as TNF-α (47) and Fas ligand (49), the membrane-bound RANKL is converted to a soluble form through ectodomain shedding (34, 35). Members of the ADAM family such as ADAM10, ADAM17, and ADAM19 exhibit RANKL shedding activity (36). MMP3 and MMP14 also cleave RANKL, although the cleavage site for MMP14 differs from others (38, 39). In our study, IL-3 significantly decreases the expression of ADAM10, ADAM17, ADAM19, and MMP3 at the mRNA level. This reduction in ADAMs in osteoblasts may be involved in downregulation of soluble RANKL by IL-3. Therefore, our results indicate that despite increasing membrane-bound RANKL on osteoblasts, IL-3 does not induce mature osteoclast formation and this effect of IL-3 may be due to a decrease in soluble RANKL.
Among various signaling pathways, JAK2/STAT5 is the predominant pathway activated by IL-3 (42). Activated JAK2 stimulates the phosphorylation of STAT5, which is translocated to the nucleus and binds to specific DNA elements to activate transcription of target genes (50). In our system, IL-3 activates the JAK2/STAT5 pathway by increasing the phosphorylation of STAT5a and STAT5b. It has been reported that STAT5a increases RANKL expression by interaction of STAT5a to the RANKL promoter; however, STAT5b showed no role in the regulation of RANKL (11). Our knockout mice studies also support this finding where we observed that IL-3–activated RANKL expression was regulated by STAT5a; however, the role of STAT5b in the regulation of RANKL expression is not yet clear.
Aging is one of the risk factor for bone loss, although the underlying mechanisms are not yet clear. Similar to other pathological conditions such as osteoporosis and multiple myeloma, RANKL and OPG are also important in the regulation of age-associated bone homeostasis (51–53). Interestingly, we observed that as age increases the RANKL associated with bone decreases dramatically in both male and female mice. This observation raises an interesting question that if RANKL is associated with the development of age-related osteoporosis, then how does the bone become more prone to fracture even after expressing less bone-specific RANKL? We further observed that RANKL expression in serum increased significantly in adult mice. Cumulatively, these observations suggest that although RANKL is important for bone metabolism, it plays a different role at different locations. Our studies show that in bone, a basal level of RANKL is required to maintain its structural integrity, and increased serum RANKL may cause development of bone pathology. Our in vitro results showed that IL-3 regulates the two functional forms of RANKL differently. Considering membrane-bound RANKL as bone specific and serum RANKL as soluble, our in vivo analysis demonstrate that IL-3 helps in maintaining RANKL expression in bone and serum by enhancing bone-specific RANKL and decreasing serum RANKL.
In contrast to RANKL, OPG expression does not change in serum of adult mice. In adult age, increased RANKL expression with no change in OPG level indicates that the increased RANKL/OPG ratio in serum might be responsible for the development of an osteopenia-like bone phenotype at this stage. Our results indicate that the significant increase in OPG expression by IL-3 in the serum of adult mice helps to maintain the RANKL/OPG ratio.
In conclusion, our study shows that IL-3 differentially regulates two functional forms of RANKL under both in vitro and in vivo conditions. In our previous studies, we showed that IL-3 inhibits osteoclast formation in a direct manner (25–28). In the present study we confirmed that IL-3 does not induce multinuclear osteoclast formation in an indirect manner due to differential regulation of RANKL. As reported earlier, IL-3 has a dual role in regulation of bone metabolism, that is, it inhibits osteoclastogenesis (25–28) and stimulates osteoblast differentiation (29). IL-3 also prevents pathological bone loss in murine models of inflammatory arthritis (27, 54, 55). Present in vivo studies show that IL-3 not only reduces the elevated circulatory RANKL expression in adult mice, but it also significantly increases OPG. Thus, IL-3 helps in maintaining RANKL and OPG expression in adult mice. Thus, the direct inhibition of osteoclast formation, the inability to induce osteoclastogenesis in a indirect manner, and downregulation of soluble RANKL by IL-3 indicate its potential in treatment of bone loss associated with aging and important skeletal disorders that develop due to increased serum RANKL (51–53).
We thank Satish T. Pote for technical assistance.
This work was supported by the Department of Biotechnology under Government of India Grant BT/HRD/34/01/2009 (to M.R.W.) and K.S. received a Senior Research Fellowship from the Council of Scientific and Industrial Research (New Delhi, India).
The online version of this article contains supplemental material.
The authors have no financial conflicts of interest.