Distinct β-Arrestin- and G Protein-dependent Pathways for Parathyroid Hormone Receptor-stimulated ERK1/2 Activation*

Parathyroid hormone (PTH) regulates calcium homeostasis via the type I PTH/PTH-related peptide (PTH/PTHrP) receptor (PTH1R). The purpose of the present study was to identify the contributions of distinct signaling mechanisms to PTH-stimulated activation of the mitogen-activated protein kinases (MAPK) ERK1/2. In Human embryonic kidney 293 (HEK293) cells transiently transfected with hPTH1R, PTH stimulated a robust increase in ERK activity. The time course of ERK1/2 activation was biphasic with an early peak at 10 min and a later sustained ERK1/2 activation persisting for greater than 60 min. Pretreatment of HEK293 cells with the PKA inhibitor H89 or the PKC inhibitor GF109203X, individually or in combination reduced the early component of PTH-stimulated ERK activity. However, these inhibitors of second messenger dependent kinases had little effect on the later phase of PTH-stimulated ERK1/2 phosphorylation. This later phase of ERK1/2 activation at 30–60 min was blocked by depletion of cellular β-arrestin 2 and β-arrestin 1 by small interfering RNA. Furthermore, stimulation of hPTH1R with PTH analogues, [Trp1]PTHrp-(1–36) and [d-Trp12,Tyr34]PTH-(7–34), selectively activated Gs/PKA-mediated ERK1/2 activation or G protein-independent/β-arrestin-dependent ERK1/2 activation, respectively. It is concluded that PTH stimulates ERK1/2 through several distinct signal transduction pathways: an early G protein-dependent pathway meditated by PKA and PKC and a late pathway independent of G proteins mediated through β-arrestins. These findings imply the existence of distinct active conformations of the hPTH1R responsible for the two pathways, which can be stimulated by unique ligands. Such ligands may have distinct and valuable therapeutic properties.

The type I PTH/PTH-related peptide receptor (PTH1R), 2 a seventransmembrane receptor (7TMR) highly expressed in the kidney and bone, plays a fundamental role in the regulation of calcium homeostasis, as well as in bone formation and resorption. Ligands for PTH1R including PTHrp and PTH are involved in the etiology and treatment of disease processes such as hypercalcemia of malignancy and osteoporosis. The actions of PTH, however, are complex. PTH is known for both anabolic and catabolic effects on bone, which are dependent upon intermittent or persistent exposure, respectively (1)(2)(3). The mechanistic basis of these effects on bone remodeling are not well understood.
The intracellular signaling pathways activated by PTH and PTHrP via the PTH1R receptor include G s -mediated activation of adenylate cyclase, resulting in cAMP production and PKA activation, and G q/11mediated PLC␤ stimulation, leading to inositol 1,4,5-trisphosphate (IP 3 ) production, calcium mobilization, and PKC activation (4 -7). It has also been demonstrated that PTH activates the Raf-MEK-ERK MAP kinase (MAPK) cascade through both PKA and PKC in a cell-specific and G protein-dependent manner (8 -10). MAPKs activated in response to stimulation by many different classes of cell surface receptors, including growth factor receptor tyrosine kinases and 7TMRs, regulate cell growth, division, differentiation, and apoptosis (11). PTHstimulated activation of MAPK is known to have proliferative effects in kidney and bone (12,13).
There is growing evidence for novel 7TMR signaling mechanisms that are distinct from the classical G protein second messenger-dependent pathways. One such mechanism involves ␤-arrestins, a small family of cytosolic proteins initially identified for their central role in 7TMR desensitization. ␤-Arrestins are recruited to agonist-occupied 7TMRs that have been phosphorylated by specialized G protein-coupled receptor kinases (GRKs) and sterically inhibit receptor-G protein coupling resulting in homologous receptor desensitization. Additionally, ␤-arrestins act as adaptors in clathrin-mediated receptor endocytosis (14,15). The role of ␤-arrestins acting as signal transducers through the formation of scaffolding complexes with accessory effector molecules such as Src, Ras, ERK1/2, JNK3, and MAPK kinase 4 (MKK4) is becoming increasingly recognized (16 -20).
The potential signaling diversity of 7TMRs suggests the possible existence of multiple discrete "active" receptor conformations. This implies that specific ligands might direct distinct signaling responses by preferentially stabilizing one or more of these active conformations. In the simple two-state model of receptor activation, agonists are defined as drugs that stabilize the active receptor conformation, which in turn promotes G protein activation. Conversely, an inverse agonist preferentially binds to the inactive receptor conformational state thereby reduc-ing G protein signaling (21). Recent observations suggest that some inverse agonists for cAMP generation may none the less be capable of recruiting ␤-arrestin to the receptor and inducing biological effects (22).
Agonist stimulation of PTH1R promotes the translocation of both ␤-arrestin 1 and ␤-arrestin 2 to the plasma membrane, the association of the receptor with ␤-arrestins, and the internalization of the receptor/ ␤-arrestin complexes (23)(24)(25). ␤-Arrestin 2 has been shown to influence bone remodeling and the anabolic effects of intermittent PTH-(1-34) administration in murine models (26,27). The mechanistic basis of these physiologic effects has not been established. Accordingly, we set out to determine the roles of G proteins and ␤-arrestins in PTH1Rstimulated ERK1/2 activation. cDNA Constructs-Previously described expression plasmids encoding the wild type human PTH1R and H223R mutant PTH1R were generous gifts from Ernestina Shippani. To facilitate the PTH1R immunopreciptitation, a FLAG epitope tag was introduced using Exsite PCRbased site-directed mutagenesis (Stratagene). The amino acid sequence EKRLK located in the downstream of the predicted PTH1R signal peptide was replaced by the FLAG epitope sequence DYKDDDDK. The expression vectors for His-tagged-␤-arrestin 1 and His-tagged-␤-arrestin 2 vectors have been described previously (28) Cell Culture and Transfection-HEK293 cells were grown in Eagle's minimum essential medium with Earle's salts supplemented with 10% (v/v) FBS and a 1:100 dilution of a penicillin/streptomycin mixture (Sigma). Cells were transiently transfected with 1 g of pcDNA1-PTH1R using FuGENE  siRNA Transfection-The double-stranded siRNA sequences individually targeting ␤-arrestin 1 and ␤-arrestin 2 are 5Ј-AAAGCCUUC-UGCGCGGAGAAU-3Ј and 5Ј-AAGGACCGCAAAGUGUUUGUG-3Ј, corresponding to the positions 439 -459 and 148 -168 relative to the start codon, respectively. The double-stranded siRNA sequence 5Ј-AC-CUGCGCCUUCCGCUAUG-3Ј was used to simultaneously target ␤-arrestin 1 and ␤-arrestin 2. This sequence corresponds to the positions 172-190 and 175-193 relative to the start codons of ␤-arrestin 1 and ␤-arrestin 2, respectively. The chemically synthesized doublestranded siRNAs and a non-silencing control RNA duplex were purchased from Xeragon (Germantown, MD) in deprotected and desalted form. HEK293 cells were transfected simultaneously with 2 g of pcDNA1-PTH1R and 1.44 nmol of ␤-arrestin 2 siRNA, ␤-arrestin 1 siRNA or control siRNA using the GeneSilencer Transfection reagent as previously described (29,30).
Cyclic AMP Determination-PTH-stimulated cAMP levels were determined in HEK293 cells transiently transfected with either the wild type human PTH1R or the constitutively active H223R mutant. Cells were preincubated for 15 min in minimum essential medium supplemented with 10 mM HEPES (pH 7.4) and 1 mM isobutylmethylxanthine and stimulated for 15 min with PTH-(1-34), PTH-NBR, or PTH-IA. Forskolin (10 M) was used as a positive control. The stimulation was terminated with the addition of 0.125 M EDTA, and the samples were boiled for 10 min and clarified by centrifugation for 1 min at 15,000 ϫ g. cAMP levels were determined using a 3 H-labeled cAMP assay as described previously (31). The results are expressed as a percent maximum stimulation normalized to forskolin.

Inhibition of PKA and PKC Reveals Distinct Temporal Components of PTH-induced MAP Kinase
Activation-HEK293 cells transiently transfected to express human PTH1R stimulated with PTH-(1-34) (100 nM) resulted in a time-dependent increase in MAP kinase activation (Fig. 1). There was an early rapid activation of ERK1/2 at 2-5 min. PTH-(1-34)stimulated ERK1/2 activation reached a peak at 5 min and remained greater than 8-fold over basal for at least 60 min (Fig. 1A). To determine the contributions of G s /PKA-and G q /PKC-dependent mechanisms to MAP kinase activation, cells were pretreated with the PKA inhibitor H89 (20 M) or the PKC inhibitor GFX (2.5 M), respectively, for 15 min prior to stimulation with PTH-(1-34) (100 nM). Each of these treatments inhibited MAP kinase activation. The greatest effect on the time course of ERK phosphorylation was apparent at 2-5 min of stimulation. In contrast, the sustained ERK1/2 activation at 30 -60 min after PTH stimulation was significantly less sensitive to H89 or GFX (Fig. 1, B and C). To determine the contribution of a putative MAP kinase signaling pathway independent of both PKA and PKC, cells were pretreated with a combination of H89 (20 M) and GFX (2.5 M) for 15 min prior to stimulation with PTH-(1-34) (100 nM) (Fig. 1D). These results indicate that PTH stimulation of PTH1R results in MAP kinase activation independent of PKA and PKC. As illustrated in Fig. 1E, MAP kinase activa-  tion is comprised of an early G protein-dependent component and a late G protein-independent component. Inhibition with H89 and GFX, individually and combination, demonstrates that the G s /PKA-and G q /PKCdependent mechanisms contribute primarily to the early time component of MAP kinase activation. Subtraction of the time course obtained from the stimulation of PTH1R with PTH-  in the presence of combined inhibitors for PKA and PKC (Fig. 1D) from the curve obtained in the absence of inhibitors predicts the time course for the G protein-dependent activation of ERK1/2. The G protein independent component of MAP kinase activation is the time course obtained from the stimulation of PTH1R with PTH-  in the presence of combined inhibitors for PKA and PKC (Fig. 1E).
Taken together with the PKA and PKC inhibitor data above, these results suggest that the ␤-arrestin-mediated ERK1/2 activation follows a time course that is very different from the G protein-mediated PKA/ PKC-dependent ERK1/2 activation. Specifically, G protein (G s , G q )-dependent activity is rapid in onset and transient, whereas the majority of ␤-arrestin-dependent activity contributes to the late component of ERK1/2 activation and is more prolonged. Consistent with the idea that the time course of PTH-stimulated ERK1/2 activity is composed of independent G protein-mediated and ␤-arrestin-mediated components, Fig. 2F shows that the G protein-dependent component of ERK1/2 activation observed in the absence ␤-arrestin 1 and ␤-arrestin 2 overlays the predicted time course of G protein-dependent ERK1/2 activation illustrated in Fig. 1E.
MAP Kinase Activation by a G s /PKA-selective PTH1R Ligand-Pharmacologic interest in modifying PTH1R-agonist interactions that contribute to ␤-arrestin-mediated desensitization has lead to the design of novel PTH1R agonists that uniquely destabilize receptor-␤-arrestin association (32). These PTH analogues have been shown to induce selective stabilization of the active G protein-coupled state of the PTH1R receptor, without causing ␤-arrestin recruitment. We examined a new PTH1R agonist, PTH-NBR, and its ability to generate G protein second messengers and to stimulate MAP kinase activation. Fig.  3, A and B, illustrates the effect of PTH-(1-34) and PTH-NBR on cAMP generation and phosphoinositide (PI) hydrolysis in HEK293 cells    Fig. 4, A and B. The time course of PTH-NBRstimulated ERK1/2 activation was significantly different from the time course of PTH-(1-34)-stimulated ERK1/2 activation. PTH-NBR stimulated an early and rapid activation of ERK1/2 that reached a peak at 5 min. The peak response of PTH-NBR was 65% of that achieved by PTH- . PTH-NBR-stimulated ERK1/2, while robust, was not sustained and rapidly returned to basal levels by 30 -60 min.
To examine the contributions of G s /PKA-and G q /PKC-dependent mechanisms to PTH-NBR-stimulated MAP kinase activation, cells were pretreated with the PKA inhibitor H89 (20 M) or the PKC inhibitor GFX (2.5 M), respectively, for 15 min prior to stimulation with PTH-NBR (100 nM). Fig. 4C shows that treatment with H89 completely inhibited PTH-NBR MAP kinase activation and the time course of MAP kinase activation was not significantly affected by treatment with GFX. These results indicate that PTH-NBR stimulation of PTH1R results in a time course of MAP kinase activation that is primarily dependent upon a G s /PKA-mediated pathway. To determine whether ␤-arrestins contributed to MAP kinase activation, we again used RNA interference to reduce the expression of endogenous ␤-arrestin 1 or ␤-arrestin 2. PTH-NBR-stimulated ERK1/2 activation was measured in HEK293 cells transfected with PTH1R and siRNA for ␤-arrestin 1 or ␤-arrestin 2. In control siRNA-transfected cells, ERK1/2 activation reached maximal levels within 5 min of treatment with PTH-(1-34) and then rapidly declined to basal levels by 30 -60 min, and reduced amounts of either ␤-arrestin 1 or ␤-arrestin 2 by siRNA did not significantly alter the time course of PTH-NBR ERK1/2 activation (Fig. 5, A and B). There was no difference in the time course of ERK activation when both ␤-arrestin 1 and ␤-arrestin 2 are silenced concomitantly (n ϭ 2; data not shown). The results obtained from the use of G s /PKA and Gq/PKC inhibitors taken together with results demonstrating the absence of a ␤-arrestinmediated component contributing to PTH-NBR-stimulated ERK1/2 activation suggest that PTH-NBR exclusively activates ERK1/2 via a G protein-dependent mechanism, namely G s /PKA.
With the lack of ␤-arrestin recruitment by PTH-NBR, one might expect that the desensitization of PTHR would be impaired and that the ERK activation may be extended to some of the later time points. The results presented here suggest that PTH1R must undergo some form(s) of ␤-arrestin-independent desensitization when stimulated by PTH-NBR. Desensitization of PTH1R by GRK2 in the absence of receptor phosphorylation has been described previously (33). PKA has also been implicated in the desensitization of PTH1R (34).
MAP Kinase Activation by a ␤-Arrestin-selective PTH1R Ligand-It has been previously reported that PTH-IA exhibited inverse agonist activity for cAMP in cells expressing the constitutively active PTH1R H223R mutant (35). The PTH1R point mutation, H223R, constitutively activates adenylate cyclase, but not phospholipase C, and results in Jansens metaphyseal chondrodysplasia (36). The effects of PTH-(1-34) and PTH-IA on cAMP generation in HEK293 cells expressing the WT   PTH1R or the H223R mutant receptor were confirmed (Fig. 6). Reflecting its constitutive activity, the basal cAMP levels in untreated cells expressing H223R were 11-fold greater than basal levels in cells expressing the WT receptor. Stimulation with PTH-(1-34) (100 nM) increased cAMP levels in cells expressing WT PTH1R or H223R to 15-fold WT receptor basal cAMP levels. Treatment of cells expressing WT PTH1R with PTH-IA (1 M) did not activate cAMP. In cells expressing H223R, treatment with PTH-IA (1 M) significantly reduced cAMP levels by 50% compared with the cAMP levels in the untreated cells, consistent with inverse agonist activity. PI hydrolysis provoked by PTH- (1-34), and PTH-IA was measured in HEK293 cells transiently expressing WT PTH1R or H223R. There was no measurable generation of IP 3 in response to PTH-IA in cells expressing the wild type PTH1R or the H223R mutant receptor (data not shown). These experiments demonstrate that PTH-IA acts as an inverse agonist for G s /PKA activation while having no apparent activity for G q /PKC stimulation.
Despite its inability to activate G s /PKA or G q /PKC signaling pathways, the inverse agonist PTH-IA is able to stimulate MAP kinase activation in HEK293 cells expressing PTH1R. The effect of PTH-IA on the time dependence of MAP kinase activation in HEK293 cells expressing WT PTHR1 is shown in Fig. 7, A and B. The maximal stimulated ERK1/2 activity by PTH-IA reached 35% of that achieved by PTH- . Consistent with the inability of PTH-IA to activate G s /PKA or G q /PKC signaling pathways ERK1/2 activation by PTH-IA is unaffected by PKA or PKC inhibition with H89 or GFX, respectively (Fig. 7, A and  B). Moreover, the time course of ERK1/2 activation was prolonged resembling the response to PTH-(1-34) observed in the presence of PKA/PKC inhibitors.
Since PTH-IA is unable to activate G s /PKA or G q /PKC signaling pathways, we hypothesized that the ERK1/2 activation provoked by the "inverse agonist" is mediated by ␤-arrestin. To determine whether ␤-arrestin 1 or ␤-arrestin 2 are contributing to ERK1/2 activation we used siRNA to silence their expression. Fig. 8, A-C, show that in HEK293 cells cotransfected with control siRNA and PTH1R, PTH-IA stimulates ERK1/2 in a time-dependent manner. The reduced expression of ␤-arrestin 1 or ␤-arrestin 2 by siRNA abolished this response at all time points. Taken together with the evidence that PTH1R has the ability to activate ERK1/2 through a ␤-arrestin-mediated pathway, which is independent from the G protein-mediated PKA/PKC-dependent ERK1/2 activation, these results suggest that PTH-IA exclusively activates ERK1/2 via this mechanism. The time course of ␤-arrestin-mediated FIGURE 8. ERK1/2 activation by PTH-IA is ␤-arrestin-dependent. HEK293 cells transiently transfected to express human PTH1R and the indicated siRNAs were stimulated with 1 M PTH-IA for the indicated periods. A, phosphorylation of endogenous ERK1/2 was determined by immunoblotting (IB). B and C, the effect of ␤-arrestin 1 siRNA or ␤-arrestin 2 siRNA transfection on PTH-IA-stimulated ERK/12 activation. The values in the graphs are expressed as percent of the phosphorylation of ERK1/2 obtained at 5 min of stimulation in control (CTL) siRNA-transfected cells. Data represent the mean Ϯ S.E. from at least five independent experiments. Statistical analysis of the curves was performed by using a two-way ANOVA comparison between ␤-arrestin siRNA-transfected cells and control siRNA-transfected cells ( †, p Ͻ 0.0001), and a Bonferroni post-test was used to compare values at each of the time points (***, p Ͻ 0.001). D, the ␤-arrestin-dependent ERK1/2 activation induced by PTH-IA overlies the predicted G protein-independent component of ERK1/2 activation represented by the data obtained from the PTH-  in the presence of combined inhibitors for PKA and PKC (shown in Fig. 1E).

␤-Arrestinand G Protein-mediated ERK1/2 Activation by PTH1R
ERK1/2 activation through the PTH1R demonstrated by stimulating PTH1R with PTH-  in the presence of combined PKA and PKC inhibitors (from Fig. 1E) or by PTH-IA without inhibitors are virtually identical (Fig. 8D).
To examine the interaction of ␤-arrestin with PTH1R, HEK239 cells transiently overexpressing FLAG-epitope-tagged PTH1R and ␤-arrestin 1 and ␤-arrestin 2 were treated with PTH-(1-34) (100 nM), PTH-IA (1 M), or PTH-NBR (100 nM), and the formation of receptor ␤-arrestin complexes was measured by coimmunoprecipitation after cross-linking with dithiobis(succinimidyl propionate). As shown in Fig. 9, A and B, ␤-arrestin coimmunoprecipitates with the receptor after stimulation with PTH-(1-34) as well as with PTH-IA. The association of ␤-arrestin with FLAG-PTH1R stimulated by PTH-IA was 30% of that stimulated by PTH- . Coimmunoprecipitaion stimulated by PTH-NBR was not significantly different from the non-stimulated control. These data support the ability of the PTH-IA to induce a receptor conformation that has the ability to recruit ␤-arrestin independent of G protein activation, while PTH-NBR, which effectively activates G s , does not recruit ␤-arrestin.

DISCUSSION
PTH1R has been recognized as a key regulator of calcium homeostasis and bone metabolism. Despite the importance in PTH-mediated signals in bone remodeling, our understanding of the mechanistic basis for these effects is incomplete. Our findings indicate that PTH-(1-34) stimulation of the PTH1R activates ERK1/2 MAP kinase by two temporally distinct mechanisms: one, a G protein-dependent pathway, which is rapid in onset and quite transient, and the other ␤-arrestin-dependent, which is slower in onset but much more persistent. Furthermore, these distinct mechanisms of MAP kinase activation can be selectively stimulated through the use of novel PTH analogues.
Our data suggest that both ␤-arrestin 1 and ␤-arrestin 2 are required for PKA/PKC-independent stimulation of the ERK MAPK cascade by PTH1R. This differs from the reciprocal effects of ␤-arrestin 1 and ␤-arrestin 2 observed for angiotensin II type 1A receptor-mediated ERK1/2 activation (29). For the AT1A receptor, ERK1/2 activation is increased when the cellular level of ␤-arrestin 1 is down-regulated by siRNA and ERK1/2 activation is decreased or eliminated when the cellular level of ␤-arrestin 2 is diminished. The requirement for both ␤-arrestin isoforms in PTH1R signaling could reflect a unique requirement for ␤-arrestin heterodimers in the transmission of ␤-arrestin-depenedent signals by the PTH1R. Dimerization of ␤-arrestin is supported by structural studies of visual arrestin, a protein analogous to ␤-arrestin, that is responsible for rapid desensitization of the GPCR, rhodopsin. Sedimentation equilibrium analysis of visusal arrestin has provided evidence for arrestin dimerization at physiologic concentrations and crystallographic studies of visual arrestin reveal the association of two asymmetric dimers (38,39). It has also recently been demonstrated, using coimmunoprecipitaion and resonance energy transfer (bioluminescence resonance energy transfer (BRET) and fluorescence resonance energy transfer (FRET)), that ␤-arrestins constitutively form both homo-and hetero-oligimers at physiologic concentrations. Furthermore, hetero-dimerization may influence the subcellular distribution of ␤-arrestin 1 and ␤-arrestin 2 and their functions (40).
We also show that the two mechanisms of PTH1R ERK1/2 activation (G protein versus ␤-arrestin) can be distinguished by the use of ligands that preferentially activate one or the other pathway. Underlying the complexity of 7TMR signaling diversity are discrete receptor conformations that may be preferentially stabilized as a result of binding to specific receptor ligands. It has been previously demonstrated for PTH1R that various N-terminal PTH and PTHrp fragments and truncated hPTH-  analogues are capable of selectively stimulating distinct G protein-mediated signals including adenylate cyclase (G s ) or PKC (G q ) activity (41,42). Biased agonism-simulating G protein-independent/␤arrestin-dependent 7TMR signaling to ERK1/2 has been shown in the AT1A angiotensin receptor system using a synthetic angiotensin agonist peptide, [Sar 1 ,Ile 4 ,Ile 8 ]SII (43).
Recently, some ligands originally classified as inverse agonists based on their effects on G protein-mediated signaling have been shown to promote scaffold assembly and ␤-arrestin-mediated MAPK activation (22). These observations suggest that ␤-arrestin recruitment is not exclusive to "agonist"-stimulated 7TMRs and G protein activation. Furthermore, ligands classically classified as inverse agonists with respect to G protein activation may rely on ␤-arrestin for their positive signaling activity. This phenomenon is not unique to PTH1R. ICI118551, an inverse agonist for the ␤ 2 -adrenergic receptor, and SR121463B, an inverse agonist for the V 2 vasopressin receptor-stimulated adenylate cyclase, have also been shown to recruit ␤-arrestin and stimulate ERK1/2 (22). The data presented here are a demonstration of biased agonism for PTH1R, where unique PTH analogues can selectively stimulate G protein-dependent signaling (specifically G s /PKA) or ␤-arrestindependent signaling mechanisms. Moreover, these findings add to the growing evidence that ␤-arrestin scaffolds can mediate 7TMR signaling to ERK1/2 independent of G-protein activation and that ␤-arrestin recruitment is not exclusive to classical 7TMR agonists of G-proteinstimulated second messengers.
It is evident that 7TMRs employ multiple mechanisms to activate the ERK1/2 cascade. The signaling mechanisms underlying ERK1/2 activation are complex and may result from activation of classical G proteinregulated effectors such as PKA and PKC, from cross-talk between 7TMRs and receptor tyrosine kinases or focal adhesion complexes or from ␤-arrestin scaffolding directly on the 7TMR (11). Depending on receptor and cell type, one mechanism may predominate or multiple mechanisms may be activated simultaneously. The contribution of ␤-arrestin-mediated ERK1/2 activation is widely variable depending upon receptor type. It has been shown in murine embryonic fibroblasts stably expressing the PAR2 receptor that ERK1/2 phosphorylation is predominately mediated through a ␤-arrestin-dependent mechanism, whereas in HEK293 cells expressing AT1A angiotensin, ␤-arrestin-dependent and G protein-dependent mechanisms almost equally contribute to the temporal activation of ERK1/2 (44,45). Here we show that the temporal contribution ␤-arrestin-mediated ERK1/2 activation for the PTH1R is about 35%.
Despite the apparent redundancy of these multiple mechanisms of MAPK activation, emerging data suggest that the different pathways leading to ERK1/2 activation downstream of 7TMRs are not only mechanistically distinct but also perform different signaling functions. ERK1/2 activated by G proteins generally accumulates in the nucleus where it phosphorylates and activates various transcription factors (46). In contrast, ERK1/2 activated by ␤-arrestin is largely excluded from the nucleus and is confined to the cytoplasmic compartment where it presumably phosphorylates a distinct set of effectors (47). While ␤-arrestin 2 has been shown to influence bone remodeling and the skeletal response to intermittent PTH-(1-34) (26), the contributions of G protein-mediated ERK1/2 activation or ␤-arrestin-mediated ERK1/2 activation to cellular responses such as proliferation, differentiation, migration, and apoptosis remain to be determined.
While PTH-(1-34) activates both G protein and ␤-arrestin-mediated signaling, PTH-NBR activates only G s /PKA signaling, and PTH-IA activates only ␤-arrestin signaling (while inhibiting G signaling). Such ␤-Arrestinand G Protein-mediated ERK1/2 Activation by PTH1R APRIL 21, 2006 • VOLUME 281 • NUMBER 16 results directly imply that the receptor conformations initiating the two forms of signaling must be distinct. Over the past several years this idea, variously referred to as "biased agonism" or "ligand-directed signaling", has been gaining currency (21,37). This phenomenon is likely to be of significant physiological and therapeutic impact, since the cellular consequences of the distinct mechanisms are likely quite different. An important implication of this work is that the receptor can exist in more than one active conformation. Thus, the receptor conformation induced by PTH-NBR selectively activates Gs/PKA but is unable to activate G q /PKC or recruit ␤-arrestin. The receptor conformation induced by PTH-IA, which is not able to activate G proteins, is able to activate ERK1/2 via ␤-arrestin. This demonstration of fundamentally distinct agonist-induced conformations of the PTH1R may have significant implications both for understanding the molecular signaling mechanism of PTH as well as for the development of novel therapeutics for the treatment of various bone pathologies. For example PTH stimulation can lead to anabolic or catabolic effects dependent upon intermittent or persistent exposure, respectively. The regulatory mechanisms invoked in these contrary responses are incompletely understood. The availability of agents such as PTH-NRB and PTH-IA, which allow dissection of previously unappreciated signaling pathways, may ultimately herald the development of entirely new agents, which very specifically target a desired subset of PTH actions.