Abstract
Parathyroid hormone (PTH) dose dependently inhibits growth factor– and stress-induced osteoblast proliferation via inactivating mitogen-activated protein kinase (MAPK) signaling pathways. Osteoblasts have recently been shown to express MAPK phosphatase (MKP)-1, a dual-specific phosphatase inactivator of MAPK. Investigated was the role of MKPs in the PTH-induced attenuation of MAPK and Jun N-terminal kinase (JNK) signaling in osteoblast-like UMR106-01 cells. PTH induced a persistent inhibition of p42/44 MAPK and JNK phosphorylation starting at 10 min of incubation and lasting for at least 2 h. Actinomycin D affected both p42/44 MAPK and JNK dephosphorylation by PTH, suggesting a transcription-dependent mechanism of action. PTH rapidly and transiently induced expression of MKP-1. MKP-1 mRNA was already elevated after 10 min of 10−7 M PTH incubation, reached maximal expression after 30 to 60 min, and remained elevated after 4 h. MKP-1 protein was also upregulated within 30 to 60 min of PTH administration. The protein kinase A inhibitor H89 partly reduced PTH-induced MKP-1 expression, but the protein kinase C inhibitor bisindolylmaleimide had no effect, suggesting that PTH induces MKP-1 mainly via the protein kinase A pathway. MKP-2 mRNA was downregulated after 2 h after an early period of induction, and MKP-3 mRNA was immediately reduced. Ro 318-220 did not affect PTH-induced MAPK inactivation but effectively blocked JNK dephosphorylation. The time course of PTH-induced MKP-1 protein expression closely correlated with JNK dephosphorylation. PTH attenuates the stress-induced JNK signaling pathway in osteoblasts via induction of MKP-1 synthesis but inhibits the p42/44 MAPK pathway mainly via transcription-independent mechanisms.
Parathyroid hormone (PTH) is a major regulator of calcium homeostasis and has catabolic and anabolic effects on bone and osteoblasts in vivo and in vitro, depending on the temporal pattern of administration (1,2⇓). Upon binding to its G-protein–coupled receptor (PTH/PTHrP-R), PTH activates both Gs-coupled adenylate cyclase and Gq-coupled phospholipase C, resulting in activation of both cAMP/protein kinase A (PKA) and protein kinase C (PKC)/Ca2+ signal transduction pathways (3,4⇓). Besides these well defined signaling pathways, the mitogen-activated protein kinase (MAPK) signaling cascades have recently received attention as essential downstream targets of G protein coupled receptor signaling (5).
Three MAPK subfamilies have been identified: the p42/44 or extracellular signal-regulated kinases, the stress-activated protein kinase/c-Jun NH2 terminal kinase (JNK), and the p38 MAPK. The p42/44 MAPK cascade is activated by various growth factors binding to receptors with intrinsic tyrosine kinase activity (6). This well-established pathway includes the sequential activation via tyrosine phosphorylation of Raf and MEK, which in turn phosphorylates p42/44 MAPK at both threonine and tyrosine residues. Upon translocation to the nucleus, p42/44 MAPK phosphorylate multiple substrates such as the transcription factors p90RSK, Elk-1, and c-myc (reviewed in (7)). The JNK and p38 MAPK pathways are activated by certain mitogens, inflammatory cytokines, and environmental stresses such as heat, UV exposure, and osmotic and oxidative stress (7), all of which activate GTP-binding proteins of the Rho family (Rac, Rho, cdc42). These proteins phosphorylate MEKKs, ASK, or the mixed-lineage kinases, which in turn activate either the JNK kinases MKK 4 and 7, or the p38 MAPK kinases MKK 3 and 6. Activation of the JNK signaling cascade results in phosphorylation of transcription factors such as c-Jun, p53, ATF-2, and Elk-1, which in turn lead to AP-1 induction (8), a key regulator of cellular apoptosis and cell survival (7), as well as certain stress responses (9). The p38 subfamily is involved in the regulation of cell motility, transcription, and chromatin remodeling (7).
PTH markedly inhibits p42/44 MAPK (10) and JNK activation in osteoblasts (11), an effect mediated via the PKA pathway. PKA signaling can directly interfere with MAPK cascades by nontranscriptional mechanisms, e.g., by inactivation of Ras-dependent signals (12). Another possible regulatory mechanism is the induction of inactivating protein phosphatases. The MAPK are dephosphorylated by a family of dual-specific phosphatases, which target both phosphotyrosine and phospho-serine/-threonine residues (13). PTH-related protein (PTHrP) has recently been shown to induce MAPK phosphatase (MKP)-1 in pancreatic β cells (14). MKP-1 has recently been demonstrated to be upregulated by glucocorticoids in osteoblastic cells, with concurrent inactivation of p42/44 MAPK and reduction of osteoblast proliferation (15).
We investigated a possible regulation of MKPs by PTH in osteoblasts and the role of these phosphatases in the PTH-induced inactivation of mitogenic protein kinase signaling. We observed a PKA-mediated, rapid and transient induction of MKP-1 by PTH. MKP-1 expression was necessary for PTH to inhibit JNK, but not p42/44 MAPK phosphorylation.
Materials and Methods
Materials
Recombinant human PTH (1-34) and PTH (3-34) were purchased from Bachem (Heidelberg, Germany). Actinomycin D, cycloheximide, H-89, bisindolylmaleimide, Ro 318-220, and forskolin were obtained from Sigma-Aldrich Chemicals (Munich, Germany). β-Actin monoclonal antibody was obtained from Abcam (Cambridge, UK), and p42/44 MAPK, phospho p42/44 MAPK, SAPK/JNK, phospho-SAPK/JNK, and HRP-conjugated (anti-rabbit, anti-mouse) antibodies were purchased from Cell Signaling Technology (Frankfurt, Germany). MKP-1, -2, and -3 antibodies were purchased from Santa Cruz (Heidelberg, Germany).
Cell Culture.
UMR 106-01 rat osteoblast-like osteosarcoma cells (provided by David Feldman, Stanford University, CA) were grown in 75-cm2 cell culture flasks at 37°C in humidified 5% CO2 atmosphere in MEM (with Earle salts) supplemented with 10% FBS, 100 U/ml penicillin/streptomycin, and 10 mM HEPES. Cells were passaged every 3 to 4 d, and experiments were performed with cells from passages 15 to 22. Subconfluent cultures were kept in serum-free medium for 24 h before stimulation. To induce JNK phosphorylation, cells were serum starved for 48 h.
Western Immunoblotting.
After incubation with the substances indicated, cells were washed once with cold PBS and lysed with ice-cold lysis buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Triton X-100) containing a cocktail of proteinase and phosphatase inhibitors (20 mM NaF, 2 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 1 mM PMSF, 10 μg/ml leupeptin, 10 μg/ml aprotinine, 3 mM benzamidine), vortexed, and centrifuged for 15 min at 15,000 × g. For separation of nuclear and cytoplasmic proteins, we used the commercially available NE-Per Kit (Perbio Sciences, Bonn) and followed the manufacturer’s instructions. The protein content of the supernatants was measured by the Bradford method (Protein Assay Kit; BioRad, Munich, Germany), and Western blot test was performed as described previously (16).
Real-Time RT-PCR.
RNA was isolated with Tri Reagent (Sigma, Munich), checked for integrity on an agarose gel, and quantified photometrically. One microgram of total RNA was reverse transcribed with oligo(dT)/random hexamer primers (10:1). Real time RT-PCR was performed with the ABI Prism 7000 Real Time PCR system (Applied Biosystems, Darmstadt, Germany) with specific primers for 18S (forward: AGTTGGTGGAGCGATTTGTC; reverse: GCTGAGCCAGTTCAGTGTAGC, amplicon length 205 bp), MKP-1 (forward: TCCAAGGAGGATATGAAGCGTT reverse: GCTACAGGAGCTGCATCCG, amplicon length: 134 bp), MKP-2 (forward: CCATCGAATACATAGACGCAGTGA, reverse: CGAAAGCCTCCTCCAGCC, amplicon length: 138 bp), or MKP-3 (forward: GAGCCAAAACCTGTCCCAGTT, reverse: CAAGCAATGCACCAGGACAC, amplicon length: 91 bp) and Universal Mastermix (also Applied Biosystems) with SYBR green to detect PCR products at the end of each amplification step. Serial dilutions of an arbitrary cDNA pool were used to establish a standard curve. Relative quantities of RNA levels were determined accounting for amplification efficacy by the software provided with the PCR system. mRNA levels were normalized to corresponding 18S quantities determined within the same run. (RT-)RNA controls in which the cDNA synthesis step was omitted and controls without template did not show a detectable amplification product.
Results
Time Course of p42/44 MAPK and JNK Inhibition by PTH
The effects of PTH on p42/44 MAPK and JNK phosphorylation were assessed by Western blot test that used phosphorylation state-specific antibodies. UMR 106-01 cells were coincubated with EGF (1 ng/ml) and maximally effective PTH concentrations (10−7 M (10)) for 10 min to 4 h. Incubation with PTH resulted in a rapid and persistent inhibition of basal and EGF-induced p42/44 MAPK phosphorylation (Figure 1). The maximal inhibitory effect occurred after 60 min of exposure to PTH. The inhibition was seen both in the cytoplasmic and the nuclear compartment (results not shown). To induce JNK activation, the cells were deprived of serum for 48 h. PTH administration caused a rapid, dynamic inhibition of JNK phosphorylation (Figure 2), which was maximal after 45 min.
Figure 1. Inhibition of basal and EGF-induced p42/44 mitogen-activated protein kinase (MAPK) phosphorylation by parathyroid hormone (PTH). Twenty-four-hour serum-deprived cells were left untreated or treated with 10−7 M PTH (1-34) 5 min before the addition of 1 ng/ml EGF for indicated incubation times, and harvested for Western immunoblotting.
Figure 2. Inhibition of Jun N-terminal kinase (JNK) phosphorylation and induction of mitogen-activated protein kinase phosphatase (MKP)-1 protein expression by parathyroid hormone (PTH). Forty-eight-hour serum-deprived cells were left untreated or treated with 10−7 M PTH (1-34) for indicated incubation times, and harvested for Western immunoblotting. (A) MKP-1 expression; loading control, β-actin. (B) JNK phosphorylation.
Rapid Induction of MKP-1 Expression by PTH
The effect of PTH on MKP-1 transcription was investigated by real-time RT-PCR (Figures 3A and 4A⇓). MKP-1 mRNA abundance was already increased after 10 min of incubation with 10−7 M PTH, reached maximal expression after 30 to 60 min, and gradually declined thereafter, but remained elevated after 4 h (Figure 3A). MKP-2 gene expression was induced after 1 h, but decreased after 2 to 4 h of PTH-incubation. MKP-3 mRNA abundance decreased after 1 to 2 h of PTH exposure (Figure 3B, C). MKP-2 and -3 protein abundance did not change during the first 4 h of PTH incubation (data not shown). MKP-1 mRNA and protein were dose-dependently induced by PTH doses between 10−0 M and 10−7 M, with a maximal effect at 10−7 M (Figure 4, A and C). MKP-1 protein was upregulated within 30 to 60 min after PTH administration (Figure 2).
Figure 3. Time-dependent induction of mitogen-activated protein kinase phosphatase (MKP)-1 expression by parathyroid hormone (PTH). RNA isolation, cDNA synthesis, and real-time RT-PCR was performed as described in text. Twenty-four-hour serum-deprived cells were left untreated or incubated with 10−7 M PTH (1-34) for the indicated incubation times, then harvested. (A) MKP-1 expression. (B) MKP-2 expression. (C) MKP-3 expression.
Figure 4. Parathyroid hormone (PTH)-induced mitogen-activated protein kinase phosphatase (MKP)-1 expression is dose and PKA dependent. (A, C) Serum-deprived cells were left untreated or incubated with different concentrations of PTH (1-34). (B, D) Cells were incubated with 10−7 M PTH (1-34) with or without 20 μM H89 or 10−7 M bisindolylmaleimide (BIS), 10−7 M PTH (3-34), or 10−7 M forskolin (FSK) for 1 h. (A, B) MKP-1 mRNA levels. (C, D) MKP-1 protein expression with β-actin as loading control.
Effect of PKA and PKC Inhibition on PTH-Induced MKP-1 Expression
To investigate which of the signal transduction pathways coupled to the PTH/PTHrP receptor is responsible for MKP-1 induction, the PKA pathway was inhibited by H-89 and the PKC pathway by bisindolymaleimide. H-89 partly reduced PTH-induced MKP-1 mRNA and protein expression, whereas bisindolymaleimide had no effect. Furthermore, the PKA activator forskolin induced MKP-1 expression, whereas PTH (3-34), a fragment selectively inducing PKC signaling, showed no induction of MKP-1 (Figure 4, B and D).
Effect of Actinomycin D and Ro 31-8220 on PTH-Induced Attenuation of p42/44 MAPK and JNK Activation
To further characterize the possible mechanism responsible for the inhibition of p42/44 MAPK and JNK phosphorylation by PTH, cells were preincubated with actinomycin D, a transcription inhibitor or cycloheximide (CHX), a translation inhibitor, before exposure to PTH. The PTH-induced attenuation of p42/44 MAPK phosphorylation was affected after 30 to 60 min of actinomycin D incubation, whereas preincubation of UMR 106-01 cells with CHX increased PTH-induced attenuation of p42/44 MAPK phosphorylation. (Figure 5). In contrast, the PTH-induced inhibition of JNK phosphorylation was nearly fully reversed by actinomycin D as well as by CHX preincubation (Figure 6), suggesting a mechanism involving gene transcription and de novo protein synthesis.
Figure 5. Effect of actinomycin D (Act. D) and cycloheximide (CHX) on parathyroid hormone (PTH)-induced attenuation of p42/44 mitogen-activated protein kinase (MAPK) phosphorylation. Serum-deprived cells were left untreated or preincubated for 1 h with actinomycin D or CHX before they were treated with or without 10−7 M PTH (1-34) and 1 ng/ml EGF for the indicated times.
Figure 6. Effect of actinomycin D (Act. D) and CHX on parathyroid hormone (PTH)–induced attenuation of Jun N-terminal kinase (JNK) phosphorylation. Serum-deprived cells were left untreated or preincubated for 1 h with actinomycin D (5 μg/ml) or CHX (35 μM) before they were treated with or without 10−7 M PTH (1-34) for 1 h.
To confirm the involvement of MKP-1 in the attenuating effect of PTH on MAPK phosphorylation, Ro 31-8220, an inhibitor of MKP-1 expression, was administered (14,17,18⇓⇓). We confirmed that Ro 31-8220 inhibited PTH-induced MKP-1 mRNA and protein expression in our model (Figure 7). Ro 31-8220 incubation alone induced p42/44 MAPK (Figure 8) and JNK phosphorylation (Figure 9). PTH still attenuated p42/44 MAPK phosphorylation in Ro 31-8220–treated cells (Figure 8), whereas preincubation of Ro 31-8220 blocked the PTH-induced attenuation of JNK phosphorylation (Figure 9).
Figure 7. Ro 31-8220 and actinomycin D inhibit mitogen-activated protein kinase phosphatase (MKP)-1 expression. Serum-deprived cells were left untreated or preincubated for 1 h with Ro 31-8220 (10−6M) or actinomycin D (Act. D; 5 μg/ml) before they were treated with or without 10−7 M parathyroid hormone (PTH) (1-34) for 1 h. (A) MKP-1 mRNA abundance. (B) MKP-1 protein abundance with β-actin as loading control.
Figure 8. Effect of Ro 31-8220 on parathyroid hormone (PTH)–induced attenuation of p42/44 mitogen-activated protein kinase (MAPK) phosphorylation. Serum-deprived cells were left untreated or preincubated for 1 h with 10−6M Ro 31-8220 before they were treated with or without 10−7 M PTH (1-34) and 1 ng/ml EGF for 1 h.
Figure 9. Effect of Ro 31-8220 on parathyroid hormone (PTH)–induced attenuation of Jun N-terminal kinase (JNK) phosphorylation. Serum-deprived cells were left untreated or preincubated for 1 h with 10−6M Ro 31-8220 before they were treated with or without 10−7 M PTH (1-34) for 1 h.
Discussion
This study provides evidence that PTH uses different mechanisms to attenuate the activation status of individual MAPK. We demonstrate for the first time that MKP-1 induction, a novel regulatory mechanism of MAPK signaling, is involved in the regulation of osteoblasts by PTH. We set out on the observation of a distinct attenuating effect of PTH on the phosphorylation status of both p42/44 MAPK and JNK in osteoblasts, a finding confirming previous results in the same cell line as well as in primary osteoblasts (10,11⇓). Because osteoblast cell turnover is critical to the regulation of bone mass, the suppression of MAPK could be a crucial mechanism mediating the profound catabolic action exerted by PTH at high doses, complementing its indirect activation of osteoclast proliferation via suppression of osteoblastic osteoprotegerin release. The attenuated phosphorylation of both MAPK species by PTH occurs rapidly within 15 to 30 min, persists for more than 1 h, and is mediated via the cAMP/PKA pathway (10,11⇓) (our own results; data not shown).
By pharmacologic blocking experiments, we found that PTH-induced JNK dephosphorylation is mediated by a mechanism requiring active gene transcription and de novo protein synthesis. Inhibition of gene transcription also partly abrogated PTH-induced p42/44 MAPK dephosphorylation, whereas inhibition of translation increased the PTH-induced attenuating effect, suggesting that mechanism(s) requiring translation are responsible for preserving p42/44 MAPK phosphorylation.
The transcription dependence of PTH-induced MAPK inhibition was somewhat surprising given the very rapid time course of PTH-induced MAPK dephosphorylation, and this finding suggested the induction of one or several target genes with immediate-early response gene properties. The recent characterization of dual-specific phosphatases as immediate-early gene products (14,15⇓) with rapid transcriptional induction and short half-lives due to rapid degradation by the 26S proteasome (19) made the members of the MKP subclass of this phosphatase family interesting candidate mediators of PTH-induced MAPK dephosphorylation. Indeed, we were able to demonstrate that PTH rapidly and transiently induces gene transcription and protein synthesis of MKP-1 in UMR106-01 osteoblasts. This finding corresponds well to the recent demonstration of MKP-1 induction by PTHrP in pancreatic β cells (14) and the inducible expression of MKP-1 in osteoblasts exposed to glucocorticoids (15).
In contrast to the rapid and distinct upregulation of MKP-1, PTH suppressed, at a slower time course, MKP-2 and -3 gene expression, and did not alter MKP-2 and -3 protein expression within the first 4 h of PTH incubation. A rapid induction of MKP-1 and slow downregulation of MKP-3 was also seen with dexamethasone treatment of mouse osteoblastic cells (15). The availability of a specific inhibitor of MKP-1 expression permitted us to analyze the relative efficacy of MKP-1 in intercepting the individual MAPK pathways. Interestingly, the administration of Ro 31-8220 increased both the basal phosphorylation of p42/44 MAPK in 24 h serum-starved cells and its immediate stimulation by the growth factor EGF, suggesting the presence of some constitutive MKP-1 activity. However, when the MKP-1 inhibitor was coadministered with EGF and PTH, the attenuating effect of PTH on p42/44 MAPK phosphorylation was largely retained. In contrast, MKP-1 inhibition fully abolished the attenuating effect of PTH on starvation stress-induced JNK phosphorylation. Hence, our results provide evidence that PTH uses differential cellular mechanisms to inactivate the growth factor–dependent and the stress-dependent MAPK signaling systems: MKP-1 is essential for PTH to attenuate JNK phosphorylation, whereas p42/44 MAPK phosphorylation is efficiently suppressed by PTH even in the absence of MKP-1 activity. This result is in keeping with MKP-1 binding affinity studies suggesting preferential binding of the phosphatase to JNK (20–22⇓⇓), although MKP-1 has been found to bind to and dephosphorylate also p42/44 MAPK under certain conditions (21).
The lacking effect of Ro 31-8220 and the in part transcription-independent suppression of p42/44 MAPK phosphorylation suggest that this PKA-mediated effect of PTH is mainly mediated by nongenomic mechanisms. PTH administration and PKA activation in general are known to inhibit upstream mediators of p42/44 activation. In osteoblasts and fibroblasts, it has been reported that PKA interferes with MAPK activation at the level of Raf-1 (10,23⇓). PKA can inhibit Raf-1 either directly or via phosphorylation of the Rap1-GTPase, an antagonist of Ras-induced cell transformation (for a review, see (12)).
An important biologic role of MKP-1 in the regulation of inflammatory signals has recently been postulated (24). Induced by cytokines, cellular stress, and glucocorticoids, MKP-1 might primarily serve to limit inflammatory and stress responses on a cellular level by inhibiting JNK signaling, interfering less markedly with mitogenic signals. The distinct induction of MKP-1 by glucocorticoids (15) and, as demonstrated in this study, PTH in osteoblasts adds the bone to the rapidly expanding range of tissues that use this mode of feedback regulation. According to this line of reasoning, the action of PTH on osteoblast MAPK signaling could be considered both anti-inflammatory (via MKP-1) and antiproliferative (via upstream inactivation of p42/44 MAPK pathway).
Acknowledgments
This study was supported by a grant from the Else Kröner-Fresenius Foundation.
- © 2004 American Society of Nephrology