Parathyroid hormone activates mitogen-activated protein kinase via a cAMP-mediated pathway independent of Ras.

In a previous study, we demonstrated that parathyroid hormone (PTH) inhibits mitogen-activated protein (MAP) kinase activation in osteosarcoma cells via a protein kinase A-dependent pathway. Here, we show that PTH can induce a transient activation of MAP kinase as well. This was observed in both Chinese hamster ovary R15 cells stably expressing high levels of rat PTH/PTH-related peptide receptor and parietal yolk sac carcinoma cells expressing the receptor endogenously. PTH was a strong activator of adenylate cyclase and phospholipase C in Chinese hamster ovary R15 cells. PTH-induced MAP kinase activation did not depend on activation of Gi, phorbol ester-sensitive protein kinase C, elevated intracellular calcium levels, or release of Gβγ subunits. It could, however, be mimicked by addition of forskolin or 8-bromo-cAMP to these cells. Prolonged treatment with forskolin caused sustained protein kinase A activity, whereas MAP kinase activity returned to basal levels. Subsequent treatment with PTH or 8-bromo-cAMP did not result in MAP kinase activation, whereas phorbol ester- or insulin-induced MAP kinase activation was unaffected. Finally, expression of a dominant negative form of Ras (RasAsn-17), which completely blocked insulin-induced MAP kinase activation, did not affect activation by PTH or cAMP. In conclusion, PTH regulates MAP kinase activity in a cell type-specific fashion. The activation of MAP kinase by PTH is mediated by cAMP and independent of Ras.

G protein-coupled receptors (GPCRs) regulate MAP kinase activity, depending on the identity of the G protein, the receptor, and the cell type involved. G i -coupled receptors, such as the M2 muscarine acetylcholine receptor, the ␣ 2 -adrenergic receptor, or the receptors for lysophosphatidic acid (LPA) or thrombin, stimulate MAP kinase in a Ras-dependent manner (5)(6)(7). Recent reports have shown that MAP kinase activation through G i involves the release of G␤␥ subunits (8 -11), which via an as yet unidentified tyrosine kinase induce the phosphorylation of Shc, leading to the formation of a Shc-Grb2-son of sevenless complex (12,13) and activation of Ras.
For receptors coupled to G q , such as the M1 acetylcholine receptor, the ␣ 1 -adrenergic receptor, or the bombesin receptor, both ␤␥ subunit-dependent (10) and ␤␥ subunit-independent (8,11) activation of MAP kinase has been reported. The ␤␥ subunit-induced activation is mediated by Ras (10), whereas the G q ␣-induced activation is mediated by phorbol ester-sensitive protein kinase C (PKC) in a Ras-independent manner (11). G s -coupled receptors, such as the ␤-adrenergic receptor or the pituitary adenylate cyclase-activating polypeptide receptor, can either trigger or inhibit MAP kinase activation (14 -16). This is cell type-dependent and can in most cases be mimicked by addition of cell-permeable cAMP analogues. With respect to the inhibition of MAP kinase activation, it was demonstrated that activation of the cAMP-dependent protein kinase A (PKA) interfered with Ras-mediated activation of MAP kinase at the level of Raf-1 (17). However, the mechanism behind the cAMPmediated activation of MAP kinase, as in PC12 cells, is largely unclear (16,18).
PTH and PTHrP bind to a common receptor, which has been shown to couple to at least two signal transduction systems: (i) a G s -mediated increase in cAMP, leading to activation of PKA (19 -21); and (ii) a G q -mediated activation of PLC-␤, leading to increases in intracellular inositol triphosphate and calcium levels and activation of PKC (19,(21)(22)(23). The identity of downstream effectors and their role in cellular responses to PTH and PTHrP are unclear. We have recently demonstrated that PTH inhibits growth factor-induced MAP kinase activation in osteosarcoma cells via a pathway that is dependent on PKA (24).
Here we show for two cell types that triggering of the PTH/ PTHrP receptor can also lead to activation of MAP kinase. This activation was not dependent on three well established G qmediated events, i.e.: (i) elevation of intracellular calcium levels, (ii) activation of PKC, and (iii) the release of G␤␥ subunits. We provide evidence that the effect is mediated by elevation of intracellular cAMP levels and occurs in a Ras-independent manner. Ala-Ser-Leu-Gly (Kemptide), myelin basic protein, and protein kinase inhibitor were from Sigma. [␥ Ϫ32 P]ATP and ECL were purchased from Amersham Corp. Indo-1 acetoxymethyl ester was obtained from Molecular Probes (Eugene, OR). Polyclonal antibodies against p42 MAP kinase were kindly provided by Drs. J. L. Bos and B. M. T. Burgering (Utrecht University, Utrecht, The Netherlands).

Materials-Rat
Cell Culture and Transfections-CHO-R15, CHO-␤2, and CHO-K1 cells were grown in Dulbecco's modified Eagle's medium supplemented with 7.5% fetal calf serum (FCS). PYS-2 cells were grown in medium consisting of a 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F12 containing 7.5% FCS. Transient transfections were performed using the calcium phosphate precipitation method. One day prior to transfection, the CHO-R15 cells were plated at a density of 8 ϫ 10 3 cells/cm 2 in six-well tissue culture clusters. The following day they were cotransfected with plasmid DNA encoding p44 HA-MAP kinase (2 g/well), G␣ subunit of retinal transducin (G␣ t ) (1 g/well), or Ras Asn-17 (3 g/well). Puc-Rous sarcoma virus plasmid was added to bring the total amount of plasmid DNA to 10 g/well.
Calcium Measurements-Nearly confluent CHO-R15 monolayers, attached to rectangular glass coverslips, were incubated in Dulbecco's modified Eagle's medium containing 0.5% fetal calf serum for 18 h. The cells were loaded with indo-1 by exposing them to 10 M indo-1 ester for 40 min at 37°C. [Ca 2ϩ ]-dependent fluorescence was recorded at an excitation wavelength of 355 nm and an emission of 405 nm (25). Since we were not able to perform proper calibration of the calcium responses in CHO-R15 cells, we related the calcium response to PTH  to that to thrombin (1 arbitrary unit).
Activation and Phosphorylation of MAP Kinase-The CHO and PYS-2 cells were plated in concentrations of, respectively, 2.5 ϫ 10 4 and 1 ϫ 10 4 cells/cm 2 in six-well tissue culture clusters and grown for 24 h. The cells were incubated in medium containing 0.5% FCS for 18 h and subsequently treated with agents as indicated. MAP kinase phosphorylation was measured by Western blotting with anti-p42 MAP kinase antibodies as described previously (24). Phosphorylated p42 MAP kinase is detected as a band with reduced mobility compared with unphosphorylated p42 MAP kinase (26). Experiments were repeated at least three times, and representative results are shown.
For determination of MAP kinase activity, endogenous p42 MAP kinase and epitope-tagged p44 HA-MAP kinase were immunoprecipitated with protein A-Sepharose beads coupled to, respectively, anti-p42 MAP kinase antibodies and monoclonal antibody 12CA5, as described previously (24,27). After the kinase reaction with myelin basic protein as a substrate, the reaction mix was subjected to SDS-polyacrylamide gel electrophoresis. Phosphorylation of myelin basic protein was measured using a PhosphorImager and ImageQuant software (Molecular Dynamics).
PKA Activation-CHO-R15 cells were plated at a concentration of 2.5 ϫ 10 4 cells/cm 2 in 96-well tissue culture clusters and grown for 24 h. The cells were incubated in medium containing 0.5% FCS for 18 h and subsequently treated with agents as indicated.
Activation of PKA was measured as described previously (24,28). Briefly, digitonin-permeabilized cells were incubated with a salt solution containing [␥-Ϫ32 P]ATP and Kemptide as a substrate, with or without protein kinase inhibitor as an inhibitory peptide. After a 10min incubation, the reaction was stopped with 25% trichloroacetic acid, and the trichloroacetic acid-soluble material was spotted on phosphocellulose filters. The difference in radioactivity incorporated in the filters between samples treated with protein kinase inhibitor-versus non-protein kinase inhibitor-treated samples was defined as PKA activity.

RESULTS
PTH Activates MAP Kinase in CHO-R15 Cells-Whereas in most cell types the PTH/PTHrP receptor is a potent activator of G s , the activation of G q is suggested to be dependent on receptor density (29). To investigate the effect of PTH on MAP kinase activity in a situation in which the receptor couples strongly to both G s and G q (21,29), we used Chinese hamster ovary cells stably transfected with the rat PTH/PTHrP receptor (CHO-R15). Binding studies with a radioiodinated PTH analogue, PTH , revealed that these cells express approximately 300,000 PTH/PTHrP receptors per cell. 2 Activation of MAP kinase was detected using a gel mobility shift assay (26).
Treatment of the cells with 10 Ϫ7 M PTH(1-34) induced a transient phosphorylation of MAP kinase, which was maximal after 5-10 min and returned to a basal level within 60 min (Fig. 1A). PTH(1-34) induced MAP kinase phosphorylation in a dose-dependent manner, starting at 10 Ϫ11 M (Fig. 1B).

PTH-induced MAP Kinase Activation Is Not Dependent on G i , G␤␥ Subunit Release, Phorbol Ester-sensitive PKC, or Increase in Intracellular
Calcium-Well established pathways used by GPCRs in the activation of MAP kinase involve G i or PKC (5-7, 11, 30, 31). We tested their involvement in the PTH-induced MAP kinase activation by a prolonged treatment of the cells with PTX and TPA to respectively inhibit G i and down-modulate PKC. As expected, PTX inhibited the thrombin-induced MAP kinase activation by approximately 80%, whereas prolonged TPA treatment completely inhibited the TPA-induced MAP kinase activation. Neither of the treatments affected the activation of MAP kinase by PTH (data not shown). Combined treatment with PTX and TPA completely inhibited the activation of MAP kinase by both thrombin and TPA but had no effect on the activation of MAP kinase by PTH(1-34) (Fig. 2). Thus it appears that the PTH-induced MAP kinase activation is not dependent on G i or TPA-sensitive PKC.
Stimulation of CHO-R15 cells with PTH(1-34) induced a transient increase in intracellular calcium, as shown in Fig. 3A. Several reports demonstrate that an increase in intracellular calcium can be sufficient to activate MAP kinase (32)(33)(34)(35). We first examined whether PTH mobilizes calcium from an intracellular or extracellular source by incubating the cells with thapsigargin. This depletes calcium from intracellular stores (36). As is demonstrated in Fig. 3A for FCS, the rapid component, due to release of calcium from intracellular stores, was completely inhibited by a 40-min thapsigargin pretreatment, whereas the slower component, due to calcium influx, was not. The calcium response to PTH(1-34) was completely inhibited by thapsigargin, suggesting that PTH induces release of calcium from intracellular stores. Thapsigargin treatment by itself induced transient MAP kinase phosphorylation (Fig. 3B), showing that, also in CHO-R15 cells, a strong increase in intracellular calcium can be sufficient to activate MAP kinase.
To determine whether an increase in calcium levels is necessary for the activation of MAP kinase by PTH, we prevented both calcium release and calcium influx by preincubating the cells with EGTA followed by thapsigargin. 30 min after the addition of these reagents, the phosphorylation of MAP kinase had returned to basal levels. Subsequent addition of PTH(1-34) stimulated MAP kinase phosphorylation to a similar level as in control cells, indicating that an increase in intracellular calcium levels is not required for MAP kinase activation by PTH.
Another GPCR-mediated event, described as being involved in the activation of MAP kinase, is the release of G␤␥ subunits (8). To examine a possible role of G␤␥ subunits in the PTHinduced MAP kinase activation in CHO-R15 cells, we expressed G␣ t to sequester G␤␥ subunits after they are released from G proteins on receptor stimulation (8). Fig. 4 shows that G␣ t expression efficiently inhibits the activation of a cotransfected, hemagglutinin-tagged MAP kinase (HA-MAP kinase) (27) by LPA, showing that G␤␥ subunits are indeed efficiently sequestered (9,13). However, PTH(1-34)-induced MAP kinase activation is not affected, suggesting that this is not dependent on the release of G␤␥ subunits.
Elevation of Intracellular cAMP Levels Induces MAP Kinase Activation in CHO Cells and Is Involved in the PTH-induced MAP Kinase Activation-PTH is a strong activator of adenylate cyclase in CHO-R15 cells (data not shown). Elevation of intracellular cAMP levels inhibits MAP kinase activation in many cell types (17), and we have previously shown that it is involved in the inhibition of MAP kinase by PTH in osteosarcoma cells (24). We were therefore surprised to find that elevation of intracellular cAMP levels by addition of forskolin or the cellpermeable cAMP analogue 8-bromo-cAMP induced a transient phosphorylation of MAP kinase in CHO-R15 cells (Fig. 5A). The kinetics and the quantity of this phosphorylation were comparable with the phosphorylation induced by PTH(1-34) (Fig.  1A). Also, in wild-type CHO cells (CHO-K1), elevation of cAMP levels resulted in MAP kinase phosphorylation (Fig. 5B). To examine whether G s -induced cAMP formation is sufficient for MAP kinase activation, we measured the activation of MAP kinase by a typically G s -coupled receptor, the ␤-adrenergic receptor. Addition of isoproterenol to CHO cells stably transfected with the ␤ 2 -adrenergic receptor (CHO-␤2) (37) resulted in a phosphorylation of MAP kinase with the same quantity and kinetics as the phosphorylation induced by PTH in CHO-R15 cells (Fig. 5C). 8-Bromo-cAMP also induced phosphorylation of MAP kinase in these cells (not shown). Thus, elevation of cAMP levels through activation of G s is sufficient to explain MAP kinase activation by PTH in CHO-R15 cells. To determine whether elevation of cAMP could account for MAP kinase activation by PTH, we measured activation of MAP kinase after prolonged incubation (2 h) with forskolin. This resulted in a sustained PKA activation (Fig. 6B), whereas MAP kinase activity had returned to basal levels (Fig. 6A). Subsequent stimulation with insulin or TPA resulted in phosphorylation of MAP kinase to a comparable level as in nonpretreated cells, whereas addition of 8-bromo-cAMP or PTH(1-34) no longer induced phosphorylation of MAP kinase (Fig. 6A). 8-Bromo-cAMP or PTH had no or little effect on PKA activity in cells preincubated with forskolin (Fig. 6B). Similar results were obtained when cells were incubated overnight with cholera toxin (data not shown). Importantly, preincubation with forskolin had no effect on the PTH(1-34)-induced calcium release (Fig. 6C), showing that PTH/PTHrP receptor functioning was not impaired. Taken together, these results suggest that the activation of MAP kinase by PTH is mediated solely by elevation of intracellular cAMP levels.
cAMP and PTH Activate MAP Kinase via a Ras-independent Pathway-Common pathways for the activation of MAP kinase by both protein tyrosine kinase receptors and GPCRs involve the activation of Ras (17). To examine the involvement of Ras in the activation of MAP kinase by PTH, we interfered with Rasmediated signaling by overexpression of a dominant-negative form of Ras, Ras Asn-17 (38). This was performed either by infection of the cells with recombinant vaccinia virus expressing Ras Asn-17 , to interfere with endogenous Ras molecules, or by cotransfection of Ras Asn-17 and HA-MAP kinase. Stimulations with insulin and TPA were used as controls for, respectively, Ras-dependent and -independent activation of MAP kinase (38,39). Infection with wild-type vaccinia virus had no effect. As expected, expression of Ras Asn-17 completely inhibited the activation of MAP kinase by insulin, whereas TPA-induced MAP kinase activation was not affected (Fig. 7A). Interestingly, forskolin and PTH(1-34) also were still able to induce MAP kinase activation. Similar results were seen with transfected RAS Asn-17 on the activation of cotransfected HA-MAP kinase (Fig. 7B). These results suggest the existence of a Ras-independent MAP kinase-activating pathway in CHO cells, which involves elevated cAMP levels and can be triggered by the PTH/PTHrP receptor.
PTH Activates MAP Kinase in PYS-2 Cells-To test the relevance of our findings for PTH/PTHrP receptor signaling, we tested several cell lines that express the receptor endogenously. Fig. 8 shows that PTH  can also induce MAP kinase activation in one such cell line, parietal yolk sac carcinoma (PYS-2) cells (40). This activation was not affected by prolonged TPA treatment, suggesting that TPA-sensitive PKC is not in-volved. PTH is a strong inducer of cAMP formation in PYS-2 cells (40), and forskolin induced an activation of MAP kinase in these cells as well, suggesting a comparable mechanism for PTH-induced MAP kinase activation in PYS-2 and CHO-R15 cells. DISCUSSION In the present study, we show that stimulation of the PTH/ PTHrP receptor induces activation of MAP kinase in CHO-R15 and PYS-2 cells. PTH-induced MAP kinase activation was not dependent on PLC-mediated events but appeared to be mediated by elevation of cAMP levels and to occur in a Ras-independent fashion.
Signaling via the PTH/PTHrP receptor involves the activation of at least two G proteins, G s and G q . We have recently shown that PTH inhibits the activation of MAP kinase in UMR 106 and ROS 17/2.8 cells through activation of PKA (24). Although PLC-␤-mediated events can induce MAP kinase activation (11,30,31,34), the activation of PLC-␤ by PTH was apparently not sufficient to affect MAP kinase activity in these cells. It has been reported that the efficiency of coupling of the PTH/PTHrP receptor to G q and PLC-␤ is related to receptor density (29). Because the PTH/PTHrP receptor couples strongly to PLC-␤ in CHO-R15 cells, most likely because of the high levels of receptor expression, we examined whether this might explain the differences between the action of PTH on MAP kinase in CHO-R15 and osteoblast-like cells.
Our data suggest that typical PLC-␤-mediated events, such as release of calcium from intracellular stores and activation of PKC, are not involved in the activation of MAP kinase by PTH. Thapsigargin treatment prevented the PTH-induced calcium response but had no effect on the activation of MAP kinase by PTH, suggesting that the calcium increase is not necessary for the PTH-induced MAP kinase activation. Nevertheless, the small increase in calcium observed with PTH in untreated cells could still be sufficient for MAP kinase activation. However, since a combined treatment with PTX and TPA completely blocked the thrombin-induced MAP kinase activation (Fig. 2), whereas the thrombin-induced calcium increase was not affected (not shown), and since the calcium response to PTH is weaker then the one observed with thrombin, this suggests that the calcium response to PTH is not sufficient to activate MAP kinase and that a strong response, like that with thapsi-gargin, is needed to activate MAP kinase. This is supported by the observation that prolonged cAMP elevation completely abolished MAP kinase activation by PTH without affecting the calcium response. An essential role of PKC was excluded by the observation that down-modulation of phorbol ester-sensitive PKC did not affect the activation of MAP kinase by PTH. These data suggest that the PTH-induced MAP kinase activation is not depending on PLC-␤ activity.
Other described intermediates between MAP kinase activation and GPCRs are G␤␥ subunits. The action of G i on MAP kinase is established to be fully dependent on G␤␥ subunits (8 -11, 13). Studies concerning the role of G␤␥ subunits in MAP kinase activation via G s have produced contradictory results. It was demonstrated that activation of MAP kinase by the G scoupled ␤-adrenergic receptor in COS-7 cells is mediated by G␤␥ subunits and is fully dependent on Ras (14). However, others reported that expression of the G s -coupled D 1A dopamine receptor in COS-7 cells did not lead to activation of Ras (9). Studies on the activation of MAP kinase by G q -coupled receptors have produced contradictory results as well. It was reported that activation of MAP kinase by the M1 acetylcholine receptor occurred in a G␤␥and Ras-dependent manner (10), while others showed that triggering of the M1 acetylcholine receptor, the ␣ 1B -adrenergic receptor, or the bombesin receptor resulted in MAP kinase activation that was independent of G␤␥ subunits (8,9,11) and mediated by PKC in a Raf-dependent but Ras-independent manner (11). In this study, we show that sequestering of G␤␥ by overexpressed G␣ t blocked LPAinduced MAP kinase activation but did not inhibit the action of PTH, suggesting that the activation of MAP kinase by PTH is not dependent on the release of G␤␥ subunits. As for most pathways involved in receptor-mediated MAP kinase activation, the G␤␥ subunit-mediated activation of MAP kinase also depends on Ras (9 -12, 13, 14). Here we show that inhibition of Ras-mediated signaling by overexpression of RAS Asn-17 completely blocks insulin-induced MAP kinase activation, whereas it does not interfere with PTH-induced MAP kinase activation. This suggests that Ras is not involved in the activation of MAP kinase by PTH. Since expression of Ras Asn-17 inhibits the activation of Ras and not the basal levels of Ras activity (43), it is still possible that PTH acts in cooperation with basal Ras activity, as has been suggested for the activation of MAP kinase by PKC (17).
The time course and extent of MAP kinase activation by cAMP and PTH were identical, and, importantly, both were shown to be independent of Ras. Sustained elevation of cAMP levels by forskolin or cholera toxin prevented activation of MAP kinase by cAMP or PTH, whereas insulin-or TPA-induced MAP kinase activation was not affected. This suggests that elevation of cAMP levels is the sole mediator of PTH on MAP kinase. The PTH-induced calcium transient was, under these conditions, similar to that in control cells. This shows that PTH/PTHrP receptor activation was not impaired, at least with respect to PLC-␤ activation. Taken together, these results strongly suggest that PTH-induced MAP kinase activation is mediated by elevation of intracellular cAMP levels.
Numerous reports have documented effects of cAMP on MAP kinase activity. The mechanism involved heavily depends on the cell type studied. It is well established that elevation of intracellular cAMP levels inhibits MAP kinase activation in fibroblasts, arterial smooth muscle cells, adipocytes, and osteoblasts (17,24). It was demonstrated that PKA interfered with Ras-mediated activation of MAP kinase at the level of Raf-1 (44). Activation of MAP kinase by cAMP was first reported in COS and PC12 cells (8,16). A number of recent reports demonstrated the same phenomenon in other cell types, e.g. Swiss-3T3, melanoma, pituitary, and ovarian granulosa cells (45)(46)(47)(48). The mechanism of cAMP-mediated MAP kinase activation, however, is not clear. We show here that elevation of cAMP levels leads to activation of MAP kinase in CHO and PYS-2 cells as well. We draw the following conclusions about the mechanism of cAMP-induced MAP kinase activation from the data presented in this study. (i) Activation of MAP kinase by cAMP is not inhibited by expression of Ras Asn-17 , indicating that Ras is not involved. This argues either for a site of action downstream of Ras in the Ras-Raf-MEK-MAP kinase cascade or a parallel pathway. (ii) A sustained elevation of cAMP levels leads to down-modulation of cAMP-induced MAP kinase activity. However, insulin and TPA are still fully capable of activating MAP kinase, showing that an effector in the MAP kinaseactivating pathway downstream of cAMP is inhibited or down-modulated, which is not essential for the activation of MAP kinase by insulin or TPA. The observation that PKA is still maximally active suggests that this effector is not PKA. (iii) In fibroblasts, a short or prolonged elevation of cAMP levels inhibits MAP kinase activation at the level of Raf-1 (17,49). In cell types in which cAMP stimulates MAP kinase activity, such as PC12 cells, it has been shown that Raf-1 is inhibited by cAMP elevation (18,45,50). Nevertheless, other factors under these circumstances are still capable of activating MAP kinase, suggesting that there is a Raf-1-independent pathway leading to MAP kinase activation. Here we show that, also in CHO cells, even after prolonged elevation of cAMP, the activation of MAP kinase by TPA and insulin was not inhibited, suggesting the existence of a Raf-1-independent pathway as well. Whether MEK and MEK kinases, such as Raf-1 and B-Raf, are involved in the activation of MAP kinase by cAMP in CHO cells remains to be determined.
Although the inhibitory action of cAMP on MAP kinase activation has been shown to correlate with the cAMP-induced inhibition of cell growth, the implication of cAMP-induced MAP kinase activation is less clear. The activation of MAP kinase by cAMP has been suggested to be involved in the neuronal differentiation of PC12 cells (3,16,51,52) and the melanogenesis in melanoma cells (46). In light of these observations it is interesting that PYS-2 cells resemble parietal endoderm cells (40). Studies with embryonic stem cells and F9 embryonal carcinoma cells have established that addition of retinoic acid together with PTH or dibutyryl cAMP induces differentiation to a parietal endoderm-like phenotype (40,53,54). It will be interesting to determine whether activation of MAP kinase by cAMP is involved in the PTH-induced differentiation to parietal endoderm.
In conclusion, our data illustrate that triggering of the PTH/ PTHrP receptor activates MAP kinase in CHO-R15 and PYS-2 cells. The PTH-induced MAP kinase activation did not depend on PLC-mediated events but appeared to be mediated by elevation of cAMP levels. Although the majority of receptor-mediated activations of MAP kinase occurs in a Ras-dependent fashion, we showed that Ras is not involved in the activation of MAP kinase by cAMP and PTH. This suggests the existence of a Ras-independent pathway involving elevated cAMP levels that is triggered by the PTH/PTHrP receptor. Since we have previously shown that the PTH-induced inhibition of MAP kinase activation in osteosarcoma cells involves elevation of cAMP levels as well, PTH/PTHrP receptor signaling to MAP kinase is correlated with the effect of cAMP on MAP kinase in a cell type-specific fashion. This might determine the effect of PTH on MAP kinase-dependent cellular responses, such as proliferation and differentiation.