Both N- and C-terminal Domains of Parathyroid Hormone-related Protein Increase Interleukin-6 by Nuclear Factor-κB Activation in Osteoblastic Cells*

Parathyroid hormone (PTH)-related protein (PTHrP) seems to affect bone resorption by interaction with bone cytokines, among them interleukin-6 (IL-6). Recent studies suggest that nuclear factor (NF)-κB activation has an important role in bone resorption. We assessed whether the N-terminal fragment of PTHrP, and its C-terminal region, unrelated to PTH, can activate NF-κB, and its relationship with IL-6 gene induction in different rat and human osteoblastic cell preparations. Here we present molecular data demonstrating that both PTHrP (1–36) and PTHrP (107–139) activate NF-κB, leading to an increase in IL-6 mRNA, in these cells. Using anti-p65 and anti-p50 antibodies, we detected the presence of both proteins in the activated NF-κB complex. This effect induced by either the N- or C-terminal PTHrP domain in osteoblastic cells appears to occur by different intracellular mechanisms, involving protein kinase A or intracellular Ca2+/protein kinase C activation, respectively. However, the effect of each peptide alone did not increase further when added together. Our findings lend support to the hypothesis that the C-terminal domain of PTHrP, in a manner similar to its N-terminal fragment, might stimulate bone resorption. These studies also provide further insights into the putative role of PTHrP as a modulator of bone remodeling.

Parathyroid hormone (PTH) 1 -related protein (PTHrP), the main factor responsible for humoral hypercalcemia of malig-nancy, is also produced in a broad spectrum of normal tissues, including bone (1). PTHrP is now emerging as an autocrine/ paracrine regulator of cell growth and differentiation in many of these tissues (1,2). Present evidence supports the hypothesis that PTHrP is an important regulator of bone cell function. Targeted disruption of the genes for PTHrP and the common type 1 PTH/PTHrP receptor in mice induces a lethal chondrodysplasia in the perinatal period (3,4). Moreover, overexpression of PTHrP targeted to chondrocytes results in a dramatic delay in the differentiation of these cells and endochondral bone formation (5). In addition, the pattern of PTHrP expression in chondrocytes and osteoblasts at different stages of bone development supports a putative role for this factor in bone formation (6).
A variety of in vitro and in vivo studies indicate that the N-terminal PTH-like region of PTHrP appears to affect osteoblastic function mainly through cAMP activation (7)(8)(9). Interestingly, PTHrP (107-139), a putative C-terminal PTHrP fragment (10), has also been shown to affect osteoblastic growth and differentiation, apparently by a protein kinase (PK) C-dependent mechanism (11)(12)(13)(14)(15)(16). The effects of this PTHrP Cterminal region appear to occur by its interaction with a specific receptor different from the type 1 PTH/PTHrP receptor (11,17).
Interleukin-6 (IL-6) is a pleiotropic cytokine synthesized by osteoblasts which acts as a downstream mediator of various bone resorptive factors (18 -22). In addition, IL-6 promotes osteoblastogenesis and bone formation (23). Thus, this cytokine appears to be an important regulator of bone remodeling. Studies using IL-6 promoter fragments transfected into various types of osteoblastic cells indicate that the intracellular mechanisms regulating IL-6 expression in these cells appear to be complex (24,25). As in other cell types, IL-6 up-regulation by a variety of osteolytic cytokines in osteoblasts has been shown to depend at least in part on the transcription factor nuclear factor B (NF-B) activity (22,24,26). NF-B is a ubiquitous family of transcription factors that regulate the expression of various genes involved in inflammatory and immune responses as well as cell proliferation and/or apoptosis (27). The most common form of NF-B consists of a heterodimer of one p50 subunit and one p65 subunit, which resides in the cytoplasm of unstimulated cells bound to its inhibitor IB. Cell stimulation with a variety of agents, including cytokines, induces IB phosphorylation and degradation, allowing active NF-B to translocate to the nucleus where it binds to DNA and regulates gene expression.
We and other investigators have demonstrated that the Nterminal PTH-like region of PTHrP stimulates IL-6 expression in human osteoblastic (hOB) cells and rat osteoblastic osteosarcoma UMR 106 cells (16,18). PTHrP (107-139) was also found to stimulate IL-6 in both cell types; but in contrast to N-terminal PTHrP, this effect of the C-terminal PTHrP appears to depend on PKC activation (16,17). In the present study, we further assessed the putative intracellular mechanisms involved in IL-6 induction by both PTHrP domains in osteoblasts. We examined whether each PTHrP domain can induce NF-B activity and whether this activation is associated with an increased IL-6 expression in different osteoblastic cell types.
hOB cells were isolated from trabecular bone explants obtained from hip or knee samples discarded at the time of surgery on osteoarthritic patients, as described previously (13). The patients (three women and one man, ages 63-79 years) had no evidence of metabolic bone disorders. Subcultured cells at the first passage from the bone fragments in Dulbecco's modified Eagle's medium with 15% FBS and antibiotics were grown to confluence, and they display features of functional osteoblasts (13). These cells were preincubated for 48 h in phenol red-free Dulbecco's modified Eagle's medium (1 g/liter of glucose) supplemented with 50 g/ml ascorbic acid and antibiotics (differentiation medium), and then the test agents were added for various time periods.
Extraction of Nuclear Proteins and Electrophoretic Mobility Shift Assay (EMSA)-Nuclear extracts were prepared according to a commercially available procedure (NE-PER ® , Pierce Chemical Co., Rockford, IL) following the manufacturer's instructions. This procedure is based on the method of Dignam et al. (31) with some modifications. Briefly, cells were washed with phosphate-buffered saline (PBS), and lysed with 50 l of a hypotonic buffer for 10 min on ice. After centrifugation at 16,000 ϫ g for 5 min, the pellet was resuspended and incubated in a hypertonic buffer for 40 min. The supernatant (nuclear extract) was collected after centrifugation at 16,000 ϫ g for 10 min and kept at Ϫ20°C until assay. All centrifugation steps were performed at 4°C. Protein was determined by the Bradford method (Pierce), using bovine serum albumin as standard.
The oligonucleotide 5Ј-AGTTGAGGGGACTTTCCCAGGC-3Ј was 5Јend-labeled with 10 Ci of [␥-32 P]ATP and T4 polymerase. Nuclear extracts (5 g of protein) were incubated with 200,000 dpm of 32 Plabeled oligonucleotide probe in 20 l of a reaction mixture containing 10 mM Tris-HCl (pH 7.9), 50 mM NaCl, 1 mM MgCl 2 , 0.5 mM EDTA, 0.5 mM dithiothreitol, 4% glycerol, 1 g poly(dI-dC) for 20 min at 4°C. Protein-DNA complexes were resolved on native 5% polyacrylamide and 0.25ϫ TBE gels. Gels were then dried and exposed to radiosensitive film. As controls for specificity of the binding reaction, nuclear extracts were preincubated with a 100-fold excess of either unlabeled NF-B oligonucleotide or another oligonucleotide having an Sp1 binding site, for 20 min at 4°C before addition of the labeled probe. In some experiments, nuclear extracts from hOB cells were preincubated for 2 h at 4°C with 2 l of the anti-p50 antibody.
Western Blot Analysis-Nuclear (10 g of protein) and cytosolic (20 g of protein) extracts were transferred onto nitrocellulose membranes (Amersham Biosciences), blocked with 5% defatted milk in PBS with 0.05% Tween 20, and then incubated overnight with either the anti-p50 or anti-p65 antibodies referred to above, at a 1:2,000 dilution (nuclear extracts) or with the antibody to the IB-␣ isoform at a 1:500 dilution (cytosolic extracts). After extensive washing, the membranes were incubated with peroxidase-conjugated goat anti-rabbit IgG and developed by enhanced chemiluminescence (Amersham Biosciences). The corresponding fluorogram bands were quantitated by densitometric scanning (ImageQuant, Amersham Biosciences). Immunofluorescent Staining-MG-63 cells grown on multiwell chambers (Labtek; Nunc, Naperville, IL) were stimulated with the agonists for 10 min in FBS-depleted medium. Then they were fixed with 64% isopropyl alcohol and 15% polyoxyethylene (Cell-fixx™, Shandon, Pittsburgh, PA) and permeabilized with 0.1% Triton X-100 in PBS for 5 min. After treatment with 10% bovine serum in PBS for 30 min for blocking, the anti-p65 antibody referred to above was added at a 1:500 dilution in the blocking solution for 2 h at room temperature. Then, fluorescein isothiocyanate-conjugated anti-rabbit IgG antibody (Sigma) at a 1:200 dilution in blocking solution was added for 30 min. After extensive washing, cells were mounted in 70% glycerol in PBS, and immunofluorescence analysis was then performed with a Leica DM-IRB confocal microscope.
Total RNA Isolation and mRNA Analysis-Cell total RNA was isolated using guanidinium thiocyanate-phenol-chloroform extraction (Tri-Reagent, MRC, Cincinnati, OH). Semiquantitative reverse transcription followed by PCR (RT-PCR) was carried out with the Access RT-PCR System (Promega), as described (16,17); 200 ng of total RNA was incubated in a 10-l reaction mixture for 45 min at 48°C followed by 32 cycles of 1 min at 95°C, 1 min at 58 -60°C, and 2 min at 68°C, with a final extension of 7 min at 68°C, using specific primers for rat or human IL-6. PCR products were separated on 2% agarose gels, and bands were visualized by ethidium bromide staining. Values obtained after densitometric scanning of the IL-6 PCR product were normalized against those of the corresponding glyceraldehyde-3-phosphate dehydrogenase (GAPDH) PCR product (a constitutive control) (16,17).
The response of IL-6 mRNA to the PTHrP domains was analyzed by Northern blot in MG-63 cells, which express both IL-6 and the type 1 PTH/PTHrP receptor (32). Total RNA (15-20 g) was size fractionated on 1% agarose gel containing 1.2 M formaldehyde and transferred to nylon membranes (Hybond-Nϩ, Amersham Biosciences). The membranes were prehybridized at 42°C for 3 h and hybridized overnight at 42°C with 10 6 dpm/ml of a 32 P-labeled human IL-6 cDNA probe. This probe was synthesized by RT-PCR using human IL-6 primers, as described above, and it was then purified by QIAquick silica gel columns (Qiagen, Hilden, Germany). The probe was labeled with [␣-32 P]dCTP (3000 Ci/mmol, PerkinElmer Life Sciences) using a random-primed DNA labeling kit (Roche Molecular Biochemicals, Germany). Filters were subsequently washed in 2ϫ SSPE, 0.3% SDS at 42°C for 30 min, followed by 1ϫ SSPE, 0.3% SDS at 42°C for 15 min. Filters were then exposed on Kodak X-Omat film at Ϫ20°C, and bands were quantified by densitometric scanning. The filters were stained with ethidium bromide to visualize 18 S and 28 S RNA as RNA loading controls.
Statistical Analysis-Data are expressed as mean Ϯ S.D. Statistical significance was determined by either t test or analysis of variance, when appropriate.

RESULTS
Both N-and C-terminal PTHrP Domains Increase IL-6 mRNA through NF-B Activation in UMR 106 Cells-We first examined the putative role of NF-B activation on the induction of IL-6 gene expression by N-and C-terminal PTHrP in UMR 106 cells. We found that each peptide, at 100 nM, induced an increase of IL-6 mRNA (assessed by RT-PCR) within 1 h, and this effect was abolished by 25 M pyrrolidinedithiocarbamate and 1 M dexamethasone, two NF-B inhibitors (27,33), in these cells (Fig. 1). Nuclear and cytosolic extracts were subsequently isolated from UMR 106 cells to assess NF-B activation after PTHrP stimulation. We found that PTHrP (107-139), in a manner similar to PTHrP (1-36), at 100 nM, stimulated NF-B⅐DNA binding in these nuclear extracts ( Fig. 2A). This was associated with a rapid (observed at 5 min) and transient disappearance of IB-␣ in the cytoplasm of these cells (Fig. 3). Pretreatment with parthenolide or MG-132, at 10 M, which prevent IB degradation by inhibiting IB phosphorylation or proteasome activity, respectively, and thereby NF-B activation (27,34), upregulated the IB-␣ band decreased by each PTHrP domain at 10 min in UMR 106 cells (Fig. 3).
The maximal stimulatory effect of these peptides on NF-B activation occurred at an earlier time period (within 5 min) than that at which each peptide maximally increased IL-6 mRNA in UMR 106 cells (Fig. 2, A and B). In addition, NF-B activation in response to each PTHrP domain persisted for up to 1 h, a time frame corresponding to the maximal induction of IL-6 mRNA triggered by these peptides (Fig. 2, A and B).
Dose-dependent Effects of Each PTHrP Domain on NF-B Activity in UMR 106 Cells-The effect of each PTHrP domain on NF-B activity was dose-dependent, being maximal with a 100 nM concentration of each peptide in UMR 106 cells (Fig.  4A). As EMSA controls, competition experiments were performed, showing that the retarded bands in cell extracts from either nonstimulated (Fig. 4A) or PTHrP-stimulated (not shown) cells disappeared with an excess of unlabeled NF-B consensus oligonucleotide, but not by a noncompetitive oligonucleotide containing binding sites for the transcription factor Sp1, an unrelated nuclear protein. This indicates the NF-B specificity of the binding in UMR 106 cells. This dose-response pattern was similar to that observed for IL-6 mRNA induction by each peptide in these cells (Fig. 4B). The addition of both peptides together at a dose (10 nM) inducing a submaximal effect on either NF-B activity or IL-6 mRNA failed to induce a higher effect in these cells (Fig. 4, C and D).
Both N-and C-terminal PTHrP Domains Induce the Activation of p50 and p65 in UMR 106 Cells-We assessed the possible involvement of p50 and p65 proteins in PTHrP-induced activation of NF-B in UMR 106 cells by Western blot analysis. Both NF-B subunits were increased about 2-fold by 100 nM PTHrP (107-139) within 15 min, and at least up to 1 h (the longest time tested) in these cells (Fig. 5, A and B). A similar maximal increase in both p50 and p65 at 15 min was observed after stimulating UMR 106 cells with N-terminal PTHrP (Fig. 5B). 139)-induced NF-B activation in these cells (Fig. 6A). The effects of these inhibitors were consistent with those observed previously on the PTHrP (107-139)-induced increase in IL-6 mRNA in these cells (17). In addition, the pentapeptide PTHrP (107-111) and PMA, two PKC stimulators in another osteoblastic cell line (15), at 100 nM and 1 M, respectively, increased both NF-B activation (Fig. 6, A and C) and IL-6 mRNA (Fig.  6B) in these cells. The calcium ionophore ionomycin, at 100 nM, also stimulated NF-B activation in UMR 106 cells, an effect that was abrogated by 25 nM BIM (Fig. 6C). This ionophore, in a manner similar to PMA, also increased IL-6 mRNA in UMR 106 cells, and this effect of both stimulators was abolished by 50 nM calphostin C, another PKC inhibitor (37) (Fig. 6B). In contrast, the stimulatory effect of PTHrP (1-36) on either NF-B⅐DNA binding activity or IL-6 mRNA was inhibited by RpcAMPS but not by BIM in these cells (Fig. 7, A and B).

Both N-and C-terminal PTHrP Domains Increase NF-B Activation in Human Normal and Transformed Osteoblastic
Cells-To demonstrate that NF-B activation by both PTHrP regions was a general feature of osteoblastic cells, we also assessed the effect of PTHrP (107-139) and PTHrP (1-36) on NF-B activation in osteoblastic osteosarcoma cells MG-63 and primary cultures of hOB cells.
Each PTHrP peptide, at 10 nM, induced NF-B⅐DNA binding in MG-63 cell nuclear extracts at 15 min (Fig. 8A). Furthermore, treatment with each PTHrP peptide, at 100 nM, for 15 min led to nuclear accumulation of p65 protein in these cells (Fig. 9, B and D). This was associated with a rapid (observed at 5 min) and transient disappearance of IB-␣ in the cytoplasm of these cells (Fig. 10). Moreover, pretreatment with 10 M parthenolide up-regulated the IB-␣ band and abrogated NF-B activation, at 15 min in MG-63 cells (Figs. 8A and 10). The increase in IL-6 mRNA triggered by these PTHrP domains, evaluated by RT-PCR, in these cells was also abolished by this inhibitor (Fig. 8B). Addition of both peptides together, at a submaximal dose (10 nM) inducing both NF-B activation and IL-6 mRNA (assessed by Northern blot analysis), failed to trigger a higher effect than that of each peptide alone in MG-63 cells (Figs. 8A and 11).
The effects of each PTHrP domain on IB-␣ degradation and NF-B activation in osteoblastic osteosarcoma cells does not seem to be related to the transformed phenotype of these cells because they were also observed in PTHrP-stimulated hOB cells (Figs. 12 and 13A). However, although treatment of osteosarcoma cells with these PTHrP peptides induced a single NF-B⅐DNA complex, this treatment increased the binding activity of two NF-B⅐DNA bands at 15 min in hOB cells (Fig.  13A). A dramatic decrease in both bands was evident after preincubating with the anti-p50 antibody some cell nuclear extracts after stimulation with PTHrP (1-36) (Fig. 13A) or PTHrP (107-139) (not shown). Addition of 10 M MG-132 diminished the stimulatory effect of each PTHrP peptide, at 10 nM, on both NF-B activation and IL-6 mRNA in these cells (Fig. 13). DISCUSSION In the present study, we show a rapid and transient induction of NF-B after stimulation with N-and C-terminal PTHrP domains in two osteoblastic osteosarcoma cell lines and also in hOB cells. In these osteoblastic cell preparations, the increased NF-B⅐DNA binding activity induced by the PTHrP domains either correlated with a rapid depletion of IB-␣ or was effectively blocked by specific inhibitors affecting IB degradation (27,33,34). Thus, both PTHrP domains seem to activate NF-B by interfering with the IB degradation pathway in osteoblastic cells.
We found that the PTHrP domains stimulated a single NF-B-binding complex in nuclear extracts from the osteosarcoma cell lines, which appears to be the common p50-p65 heterodimer. A similar finding was reported in rat osteosarcoma cells ROS 17/2.8 treated with tumor necrosis factor-␣ (22). In contrast, each PTHrP domain induced the activation of two specific NF-B-binding complexes in hOB cells. Interestingly, and related to this finding, IL-1␣ was shown to induce two NF-B bands, apparently consisting of a p50-p65 heterodimer and a p50-p50 homodimer, in primary cultures of osteoblastic cells from mouse calvaria (38). In fact, we found herein that preincubation with anti-p50 antibody triggered a dramatic decrease in both NF-B-binding complexes in nuclear extracts from hOB cells. Because the p50-p50 dimer is transcriptionally repressive (27,39), its increase might represent a mechanism, which appears to be absent in osteosarcoma cells, to self-limit NF-B activation upon cytokine stimulation of osteoblastic cells.
The N-terminal PTH-like region of PTHrP stimulates the production by osteoblasts of various osteoclast activators, namely IL-6, receptor activator of NF-B ligand and matrix metalloproteinases, whose genes have an NF-B binding sequence in their 5Ј-flanking region (16, 18, 19, 25, 26, 40 -42). However, the possible involvement of this transcription factor in IL-6 gene induction by either PTH or PTHrP has not been tested so far. In the present report, NF-B activation after stimulation with either PTHrP  or PTHrP (107-139) occurred earlier than the increase in IL-6 mRNA, but with a similar dose-response pattern, in UMR 106 cells. Moreover, IL-6 gene expression induced by each PTHrP domain was abrogated or decreased dramatically by various NF-B inhibitors in different osteoblastic cell preparations. These data indicate that the induction of this cytokine by both the N-and Cterminal regions of PTHrP requires NF-B activation in osteoblastic cells.
Although some previous reports showed an inhibitory effect of PTHrP (107-139) on osteoclastogenesis in rodent osteoclasts (12,43), another study found a stimulatory effect of this peptide in isolated mouse bone cell cultures (44). Moreover, a histomorphometric study in osteopenic rats treated daily for about 2 weeks with PTHrP (107-111), which accounts for the effects of PTHrP (107-139) in various cell types (11,13,15,17,(43)(44)(45), found a decreased trabecular bone formation associated with an increased bone resorption (46). Because, as stated above, NF-B activation appears to be a common mechanism of bone resorption activators in osteoblastic cells, our present findings further support that PTHrP (107-139) might be a stimulator of bone resorption.
Both PKA and PKC stimulators are powerful activators of the transcription factor NF-B (47). Previous studies also indicate that the stimulatory effect of the N-terminal region of both PTH and PTHrP on IL-6 in osteoblastic cells is mediated by a cAMP-dependent mechanism (16,19,25). In addition, other studies have shown that PKC and/or intracellular Ca 2ϩ signaling also appear to be important pathways involved in the action of PTH and several osteolytic cytokines on IL-6 production by bone-derived cells (20,21,48,49). Present results, together with those reported previously (17), indicate that two PKC inhibitors and two calcium channel blockers abrogate the stimulatory effect induced by PTHrP (107-139) on NF-B activation as well as on IL-6 gene expression in UMR 106 cells. In addition, in the present study, PMA, a PKC stimulator (15), and the calcium ionophore ionomycin similarly increased NF-B activity and IL-6 mRNA in these cells. Moreover, these PKC inhibitors suppressed both ionomycin-induced NF-B activation and IL-6 mRNA in UMR-106 cells. Collectively these findings, and our previous results (17), strongly suggest that intracellular Ca 2ϩ signals play a key role in PKC activation leading to an increased NF-B activity in these cells. On the other hand, we found that a PKA inhibitor abolished the effect of PTHrP (1-36) on both NF-B activation and IL-6 mRNA in UMR 106 cells. These results extend previous findings in these cells and hOB cells (16,17), indicating that different mechanisms mediate the NF-B-dependent IL-6 induction by the Nand C-terminal PTHrP domains in osteoblastic cells.
Our previous study has shown that the stimulatory effect of the N-and C-terminal peptides of PTHrP on IL-6 in hOB cells was not increased by different concentrations of both combined peptides (16). Consistent with these earlier findings, we showed herein that these peptides together did not induce a higher activation of either NF-B⅐DNA binding complex or IL-6 mRNA in different osteoblastic cell preparations. These findings, taken together with the aforementioned data, strongly suggest that a cross-talk in signal transduction pathways involving PKA and PKC activation, as has previously been suggested (50), occurs associated with the IL-6 response to each PTHrP domain in osteoblastic cells.

FIG. 13. The N-and C-terminal PTHrP domains induce NF-B activation in hOB cells.
Serum-depleted hOB cells were stimulated with the PTHrP peptides, at 10 nM, for 15 min. EMSA was then performed on isolated nuclear extracts, pretreated or not with anti-p50 antibody, as described under "Experimental Procedures." Specificity controls, using an excess (100ϫ) of either cold NF-B or unrelated Sp1 oligonucleotides with nuclear extracts from PTHrP (1-36)-stimulated cells, are also shown (A). IL-6 mRNA changes were analyzed by RT-PCR with cell total RNA, using IL-6-and GAPDH-specific primers, after cell stimulation for 1 h with each PTHrP peptide, at 10 nM (B). PTHrP stimulation of both NF-B binding bands (indicated by arrows) (A) and IL-6 mRNA (B) was inhibited by 10 M MG-132 (added 1 h before the PTHrP peptides). Results are representative of those obtained in cell cultures from four different patients.
Recent studies have shown that the N-terminal region of PTHrP can internalize into the nucleus, apparently by various cellular pathways, in different cell types, including osteoblastic cells (51)(52)(53). In some of these cells (vascular smooth muscle cells and chondrocytes), this nuclear localization seems to be associated with an altered either cell proliferation or apoptosis, respectively (52,53). Interestingly, the C-terminal domain (108 -139) of PTHrP appears to play a critical role on this intracrine proliferative effect induced in rat vascular smooth muscle cells (53). Clarification of the interaction between the possible intracrine function(s) of PTHrP and the induction of NF-B activation by both its N-and C-terminal domains in osteoblasts as reported herein, awaits further studies.
In summary, our findings support that the C-terminal domain of PTHrP, in a manner similar to that of the PTH-like domain, might promote bone resorption by inducing the transcription factor NF-B in osteoblastic cells. The different intracellular pathways associated with this effect induced by each PTHrP domain in these cells could provide alternative pathways to ensure IL-6 synthesis, and possibly that of other osteolytic factors, in the bone microenvironment. Although the pathophysiological significance of these findings awaits further studies, they provide novel insights into the mechanisms modulating bone remodeling.