Functional expression and signaling properties of cloned human parathyroid hormone receptor in Xenopus oocytes. Evidence for a novel signaling pathway.

Expression of human parathyroid hormone receptor (hPTHR) was obtained in Xenopus oocytes. Receptor function was detected by hormone stimulation of endogenous Ca2+-activated Cl- current. This current was blocked by injected, but not by extracellular, EGTA, confirming that the hPTHR activates cytosolic Ca2+ signaling pathways. PTH responses were acutely desensitized but were regained in 6 12 h. Injection of cAMP or analogues had no effect on either responsiveness or desensitization to hPTH. The hPTH response was more sluggish than seen with serotonin 5-hydroxytryptamine (5-HT2C) receptor. In oocytes co-expressing both hPTHR and 5-HT2C receptors, homologous desensitization was seen, but cross-desensitization was not observed. Injection of inositol 1,4,5-trisphosphate (InsP3) elicited a fast inward current similar to that induced by serotonin, and complete cross-desensitization occurred between the InsP3 and 5-HT2C responses. Desensitization by hPTH did not affect responses to either InsP3 or serotonin, but cells desensitized to injected InsP3 still responded strongly to PTH. Oocytes did not respond to either cADPR or NAADP+, but NADP+ and analogues were found to be potent inhibitors of PTH signaling. We suggest that PTH cytosolic Ca2+ signaling in oocytes either involves a novel signaling system or proceeds through a Ca2+ compartment whose responsiveness is regulated in a novel way.

Expression of human parathyroid hormone receptor (hPTHR) was obtained in Xenopus oocytes. Receptor function was detected by hormone stimulation of endogenous Ca 2؉ -activated Cl ؊ current. This current was blocked by injected, but not by extracellular, EGTA, confirming that the hPTHR activates cytosolic Ca 2؉ signaling pathways. PTH responses were acutely desensitized but were regained in 6 -12 h. Injection of cAMP or analogues had no effect on either responsiveness or desensitization to hPTH. The hPTH response was more sluggish than seen with serotonin 5-hydroxytryptamine (5-HT 2C ) receptor. In oocytes co-expressing both hPTHR and 5-HT 2C receptors, homologous desensitization was seen, but cross-desensitization was not observed. Injection of inositol 1,4,5-trisphosphate (InsP 3 ) elicited a fast inward current similar to that induced by serotonin, and complete cross-desensitization occurred between the InsP 3 and 5-HT 2C responses. Desensitization by hPTH did not affect responses to either InsP 3 or serotonin, but cells desensitized to injected InsP 3 still responded strongly to PTH. Oocytes did not respond to either cADPR or NAADP ؉ , but NADP ؉ and analogues were found to be potent inhibitors of PTH signaling. We suggest that PTH cytosolic Ca 2؉ signaling in oocytes either involves a novel signaling system or proceeds through a Ca 2؉ compartment whose responsiveness is regulated in a novel way.
Parathyroid hormone (PTH) 1 is the primary regulator of calcium and phosphate homeostasis in higher animals (1). This 84-amino acid peptide and truncated forms such as the 1-34 or 1-36 fragments initiate the biological actions of PTH through a specific receptor (PTHR) on the plasma membrane in target tissues, primarily kidney and bone (2)(3)(4)(5). Recently, the cDNAs of PTHR from rat, mouse, and human cells have been cloned (6 -10), and both transient and stable transfection of PTHR has been described (7,11,12).
The transfection experiments demonstrated that the PTHR can activate multiple signaling events in the same cell, including production of cAMP, activation of phospholipase C, and elevation of cytosolic Ca 2ϩ (7,12). However, this multiple signaling capability is not always present in cells that constituitively express the PTHR. In some systems, a cAMP response is observed in the absence of the cytosolic Ca 2ϩ response (13,14), while in others the reverse appears to be true (15). In addition, N-terminally truncated forms of PTH that do not activate adenylyl cyclase still appear to generate the cytosolic Ca 2ϩ signal (16), and likewise cells in a cultured bone cell line desensitized to PTH with regard to the cAMP response still show a cytosolic Ca 2ϩ signal (17). Thus, the factors that determine signaling in different systems are not understood.
The signaling systems responsible for elevation of cytosolic Ca 2ϩ in response to PTH are also complex. The hormone has been reported to activate Ca 2ϩ influx through a cAMP sensitive plasma membrane channel (18), but elevated cytosolic Ca 2ϩ can also be observed in the absence of influx of extracellular Ca 2ϩ (19). The latter effects have been attributed to generation of InsP 3 through the action of phospholipase C (20,21). However, some studies suggest that production of InsP 3 is not obligatory for elevation of cytosolic Ca 2ϩ by PTH. It has been reported that parathyroid hormone and thrombin release intracellular calcium from different calcium stores in osteoblastlike UMR 106-H5 rat osteosarcoma cells (22), and no effect of PTH on InsP 3 was found in this study. It has also been reported that full-length hPTHR stimulated by hPTH does not increase InsP 3 levels in stably transfected HK-293 cells, while some truncated forms of the receptor do express this activity (23).
These studies indicate a need for further examination of signaling by the PTHR under conditions where different signals can be studied independently in the same cell system. The functional expression of the PTHR in Xenopus oocytes appears to offer that possibility. Specifically, the frog oocyte has a Ca 2ϩ -activated Cl Ϫ channel that is directly utilized to detect cytosolic Ca 2ϩ changes (24 -27) but no endogenous cAMP-activated channels. Also, the ability for direct manipulation of cytosolic contents provide a flexible system that may be used to clarify the PTH signaling system. In this paper, we demonstrate its use for examination of cytosolic Ca 2ϩ signaling.

EXPERIMENTAL PROCEDURES
Materials-Human parathyroid hormone receptor cDNA in pcDNA I vector was generously provided by Dr. G. V. Segre (Massachusetts * This work was supported in part by National Institutes of Health Grants NS28190, NS01557, DA09116, DA09444, and AA07611. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  General Hospital/Harvard Medical School, Boston). D-myo-Inositol 1,4,5-trisphosphate, 8-Br-cAMP, 8-CPT-cAMP, c-2,3-AMP, and serotonin (5-hydroxytryptamine (5-HT)) were purchased from Sigma; hPTH (1-34) was from Bachem, and L-15 medium was from Life Technologies, Inc.
Cloning of Human Parathyroid Hormone Receptor into Expression Vector and in Vitro RNA Synthesis-cDNA of human parathyroid hormone receptor was subcloned into pcDNA3 expression vector (Invitrogen) at BamHI-XbaI sites for DNA injection to the oocytes. For in vitro synthesis of RNA, the subclone was linearized by XbaI, and cRNA was synthesized with T7 mMessage mMachine kit (Ambion). The RNA yield was measured by A 260 and 1-10 ng of RNA was injected into each oocyte.
Preparation of Xenopus Oocytes-Adult female Xenopus laevis frogs were purchased from commercial sources (Xenopus I, Ann Arbor, MI, and African Fish Farm, South Africa). Frog was anesthetized in 0.17% MS-222 (tricaine methanesulfonate) for about 1 h. Ovarian tissue was surgically removed, and the follicular cells were removed by treatment with 2 mg/ml of collagenase in Ca 2ϩ -free ND96: 96 mM NaCl, 2 mM KCl, 1.5 mM MgCl 2 , and 5 mM HEPES, pH 7.5. Stage V and VI oocytes were selected and kept in 50% L-15 medium supplemented with 0.8 mM glutamine and 10 g/ml gentamicin at 18 -20°C.
Microinjection of DNA and RNA-Injection of oocytes was performed in a Ca 2ϩ -free ND96 solution by an automatic Drummond microinjector. 5-10 nl of DNA or 30 -40 nl of RNA solution per oocyte were injected. After injection, oocytes were transferred to 50% of L-15 medium for incubation.
Whole Cell Voltage Clamp-Electrophysiological recording was essentially as described previously (29,30). Briefly, oocytes placed in a recording chamber were voltage-clamped at Ϫ80 mV with two electrodes (filled with 3 M potassium chloride and having a resistance of 0.5-3 megaohms) using an Axoclamp-2A (Axon Instruments). Oocytes were superfused with ND96 containing 1.8 mM of CaCl 2 . hPTH  was directly applied to the bath, and 100 nM of 5-HT was applied to the oocytes by superfusion. Voltage command protocols were generated by pCLAMP software (Axon Instruments), and the resulting membrane current changes were recorded using the same program and on a chart recorder.
Intracellular injection of InsP 3 , EGTA, cAMP, 8-Br-cAMP, 8-CPT-cAMP, and 2Ј,3Ј-cAMP were performed using an automatic Drummond microinjector, similar to that described by Lin et al. (31). InsP 3 (4.8 mM), EGTA (50 mM), cAMP (10 mM), 8-Br-cAMP (10 mM), 8-CPT-cAMP(10 mM), and 2Ј,3Ј-cAMP(10 mM) were dissolved in distilled water and back filled into the pipette (tip broken to about 5 m diameter) of the microinjector. A few droplets of the solution were expelled out of the pipette just before the oocyte impalement to ensure a free flow of the injection pipette. A volume of 10 nl for InsP 3 , EGTA, cAMP, 8-Br-cAMP, 8-CPT-cAMP, or 2Ј,3Ј-cAMP was injected into individual oocytes. Oocytes were perfused with calcium-free NP96 solution during all injections. After recording, injection pipettes were examined under the microscope to ensure that the pipettes were not clogged. For InsP 3 injection, shortly after injection into experimental oocytes, InsP 3 was also injected into control oocytes to ensure that InsP 3 was still capable of eliciting a response.
Receptor Binding Studies-Radioligand binding to expressed receptors on oocytes was performed essentially as described by Yoshii et al. (28). Oocytes were prewashed once in ND96 solution and transferred to 1 ml of ND96 containing 5 nM of either [ 3 H]mesulergine or 125 I-hPTH (1-34 tyrosine amide) for 90-min incubation at room temperature. Mianserin (5 M) or hPTH (1-34) (10 M) were used as competing ligands to define the amount of specific binding for 5-HT 2C receptor or hPTH receptor, respectively. Oocytes were pooled in groups of 5 or 10 for the binding studies, which were done in duplicate. Binding was terminated by vacuum filtration through Whatman GF/B filters pretreated with 1%   was directly added to the recording chamber with a capacity of about 500 l so that the final concentration of hPTH in the recording chamber was 10 Ϫ6 -10 Ϫ7 M. Applying hPTH to uninjected oocytes, no response was observed (top trace). When hPTHR-injected oocytes were stimulated with hPTH, inward currents were induced (middle and bottom trace). After the first hPTH stimulation, the oocyte was washed with ND96 with 1.8 mM CaCl 2 . When a second hPTH was applied about 1 min after the washout, no response was elicited (middle trace). When the second hPTH stimulation was applied several hours after the washout, the response was partially recovered (bottom trace). B, bar graph showing hPTHR-mediated membrane currents and homologous desensitization. hPTH did not induce noticeable membrane currents in uninjected oocytes (control). Following the first hPTH stimulation, a large inward current was observed (581.3 Ϯ 170 nA, n ϭ 4). * indicates a significant difference from the control value (p Ͻ 0.01). Upon washout of hPTH, very little response could be induced by a second application of hPTH within 5 min, indicating complete homologous desensitization. The response slowly recovered after several hours (257.5 Ϯ 85.7 nA, n ϭ 4). ** indicates a significant difference from the 5-min washout value (p Ͻ 0.01).
polyethylenimine. The filters were washed 3 times with 3.5 ml of ice-cold 50 mM Tris-HCl, pH 7.4, and then counted.

Functional Expression of Human Parathyroid Hormone Receptor in X. laevis Oocytes-To express the hPTHR in
Xenopus oocytes, a DNA clone for the hPTHR (10) was used. Aliquots of cDNA of the hPTHR were microinjected into the nucleus of defolliculated X. laevis oocytes. Alternatively, synthetic cRNA was in vitro transcribed from the cDNA clone and microinjected into the cytoplasm of Xenopus oocytes. 3-4 days after cRNA or cDNA injection, stimulation with 1-34 hPTH elicited an inward current in hPTHR-injected oocytes but not in uninjected control cells (Fig. 1A). The hPTH response displays a strong homologous desensitization, as shown by a lack of response from the second hPTH stimulation following washing out the bath hPTH used in the initial stimulation (Fig. 1A, middle trace, and Fig. 1B). The responsiveness of hPTHR-expressing oocytes slowly recovers, with the size of the inward current showing a partial recovery after 6 h (Fig. 1A, bottom trace, and Fig. 1B) and a complete recovery after 12 h (data not shown). The hPTHR was consistently expressed in oocytes, whether DNA or RNA was injected.
Impact of Calcium for the Human Parathyroid Hormone Receptor-mediated Current-To determine whether the hPTHinduced current in oocytes requires cytosolic Ca 2ϩ , 500 pmol of Ca 2ϩ chelator EGTA was microinjected into hPTHR expressing oocytes. This gives a final concentration of EGTA at ϳ0.5 mM for oocytes with an average volume of 1 l. Stimulation with hPTH (1-34) after EGTA injection produced no responses ( Fig.  2A, middle trace, and Fig. 2B), indicating that the hPTHinduced currents were completely eliminated by intracellular EGTA injection. However, the addition of 1 mM EGTA to the calcium-free ND96 superfusion solution had no effect on hPTHinduced currents ( Fig. 2A, bottom trace, and Fig. 2B). This showed that the hPTHR-induced response in Xenopus oocytes is dependent on intracellular Ca 2ϩ but not on extracellular Ca 2ϩ .
The Inward Current Is Mainly Carried by the Chloride Ion-The ionic selectivity of hPTH-induced currents were determined using different channel blockers. There was no effect on hPTH-induced membrane currents when in the bath solution Ca 2ϩ was chelated with EGTA and, with the inclusion of Cd 2ϩ as the Ca 2ϩ channel blocker and 4-AP and TEA as the K ϩ channel blocker (Fig. 3A, upper trace, and Fig. 3B). However, the membrane current could be completely suppressed by using an external ND96 solution containing 1 mM niflumic acid, a Cl Ϫ channel blocker (Fig. 3A, lower trace, and Fig. 3B). It has been reported that reversal potential for Cl Ϫ channel is approximately Ϫ25 mV (35). The hPTH-induced membrane current was also eliminated by clamping the oocytes at the holding potential of Ϫ25 mV (data not shown). Together, these results indicate that the hPTHR-induced current is mainly carried by the Cl Ϫ ion.
Human Parathyroid Hormone-induced Response Is Independent of cAMP-It has previously been reported that PTH can activate production of cAMP in Xenopus oocytes injected with mRNA extracts from target cells that contain the PTHR message (32). Also, in some systems, cAMP has been reported to mediate desenstitization to PTH by a protein kinase A-dependent process (33,34). Therefore it was of interest to examine whether cAMP has any effect on the cytosolic Ca 2ϩ re-sponse to PTH in our system. Intracellular injection of cAMP and cAMP analogues, 8-Br-cAMP, 8-CPT-cAMP, and 2Ј,3Ј-cAMP, did not induce any membrane current in either control oocytes not expressing the hPTHR (data not shown) or hPTHRexpressing cells (Fig. 4). These same hPTHR-expressing cells can produce an inward current by subsequent hPTH application, demonstrating that they are not desensitized to PTH by cAMP, and suggesting that the desensitization process in oocytes is not mediated by protein kinase A.
The Response to PTH Is More Sluggish Than That of a Receptor Coupled to the Phospholipase C Pathway-Xenopus oocytes have been used widely to study membrane receptors that couple to the phospholipase C signaling pathway (24), since stimulation of such a receptor expressed in oocytes gives a robust inward Cl Ϫ current. To compare the hPTHR-induced response to that of a system that utilizes the InsP 3 pathway, we also expressed a cloned serotonin receptor, the 5-HT 2C receptor, that couples to the phospholipase C pathway (36). Two parameters were analyzed for time course studies of receptormediated membrane currents, the latency of onset and the start-peak time (Fig. 5A). The response profiles from the two receptors are very distinct from each other, with the 5-HT response being a fast and robust one, and the hPTH response showing a longer latency and a slower kinetics for start-peak time (Fig. 5, A and B). Specifically, application of 5-HT produces an almost immediate current, whereas hPTH-induced current does not commence until about 13 s after the hormone is added (Fig. 5B, top panel). The 5-HT current also develops quickly, with an average start-peak time at 4 s, whereas hPTHinduced current takes about 7 s to develop (Fig. 5B, bottom  panel). These differences are statistically significant, and they suggest that the two receptors may use different signaling pathways.
On average, the chloride current responses to serotonin are also larger than those to PTH. However, this quantitative difference cannot be interpreted since it could be due to one or a combination of several variables including the efficiency of expression of the receptor message, the stability of the injected nucleic acid messages, the stability of the expressed receptor in the cell membrane, the efficiency of coupling to second messenger systems, and the sensitivity and size of responsive calcium compartments.
The relative levels of receptor expression were examined in ligand binding studies. We found that the levels of expression of hPTH receptor and 5-HT 2C receptors were roughly comparable. The radioligand binding studies showed 2.2 ϫ 10 9 and 4.2 ϫ 10 9 receptors/oocyte for hPTH and 5-HT 2C receptors, respectively. Thus, it is unlikely that the qualitative and quantitative differences between the PTH and the 5-HT 2C systems are related to different levels of receptor expression.

Examination of Desensitization Properties for the Human Parathyroid Hormone Receptor and the 5-HT 2C Receptor-G
protein-coupled pathways often display marked desensitization, operationally defined as a reduced response to continued or repeated stimulation of agonists. The response induced by both the hPTHR and 5-HT 2C receptor show strong homologous desensitization, as the second stimulation by the same agonist produced little or no membrane current (Fig. 6A). It has been well characterized that 5-HT 2C receptor uses inositol trisphosphate as the second messenger to induce calcium release from 5. Human parathyroid hormone receptor-mediated current has a long latency of onset and a slow kinetics. A, diagram showing the measurement of the latency time and the start-peak time.
Top trace shows an hPTH-induced response. As a comparison, the bottom trace shows a serotonin (5-HT)-induced membrane current in an oocyte injected with the RNA for 5-HT 2C receptor. Oocytes were clamped at Ϫ80 mV. B, relative time courses of the latency of onset and the start-peak time between hPTH-and 5-HT-induced responses. Data are presented by mean Ϯ S.E.; n is for numbers of oocytes used. The latency of onset for hPTHR, 12.8 Ϯ 0.8, n ϭ 52; for 5-HT 2C , 0.6 Ϯ 0.06, n ϭ 14. The start-peak time for hPTHR, 6.9 Ϯ 0.5, n ϭ 52; for 5-HT 2C , 3.9 Ϯ 0.3, n ϭ 14. * (top graph) indicates a significant difference for the latency time between the 5-HT response and hPTH response (p Ͻ 0.01). * (bottom graph) indicates a significant difference for the start-peak time between the two groups (p Ͻ 0.01). intracellular stores. To examine the possibility that hPTH induces Ca 2ϩ release by activating the same Ca 2ϩ release mechanism as the 5-HT 2C receptor, the hPTH receptor and 5-HT 2C receptor were coexpressed in oocytes. When hPTH application has caused a complete homologous desensitization, 5-HT stimulation was still effective in releasing Ca 2ϩ (Fig. 6B). The results suggest that the 5-HT 2C receptor and the hPTH receptor cause intracellular calcium release through distinct pathways.
Role of Inositol 1,4,5-Triphosphate in Signaling by Human Parathyroid Hormone Receptor-It has previously been proposed that the cytosolic Ca 2ϩ signal for PTH is activated by InsP 3 (20). However, some studies indicate that this may not be the case in every system (23), and the existence of InsP 3insensitive calcium compartments that are activated by PTH has been proposed (22). Our kinetic and desensitization data also suggest that possibility, and further experiments were designed to test it more directly by studies with microinjection of InsP 3 . Strong activation of the Cl Ϫ channel was observed upon initial injection of InsP 3 , but no response was observed with a second injection (Fig. 7, top trace). Thus, either the InsP 3 receptor is desensitized or the Ca 2ϩ pools, which respond to injected InsP 3 , are emptied. 5-HT and InsP 3 also showed crossdesensitization as expected (Fig. 7, middle two traces), since they share the same calcium release pathway. However, in hPTH-desensitized cells, InsP 3 injection still elicited an inward current, and in cells that were unresponsive to injected InsP 3 , a strong signal was still observed upon addition of PTH (Fig. 7, bottom two traces).
Since the PTHR can activate Ca 2ϩ compartments that are insensitive to injected InsP 3 , we examined whether PTH signaling might proceed through the cADPR pathway (44 -46) or the recently described NAADP ϩ pathway (51). However, we observed no effect of injection of these compounds on the Cl Ϫ current in Xenopus oocytes. In addition, as expected from the above results, inhibitors that block both InsP 3 and cADPR signaling systems were found to have no affect on the PTH responses in oocytes (data not shown). However, unexpectedly we found that NADP ϩ , thio-NADP ϩ , and the NADP ϩ analogue, 3-acetylpyridine-ADP ϩ , are all strong inhibitors of the cytosolic Ca 2ϩ response to PTH in oocytes (Fig. 8). This inhibition persists for a few hours, but the hormone response is eventually regained. In control experiments, injection of these compounds alone had no impact on cytosolic Ca 2ϩ , or on oocyte responsiveness to serotonin. DISCUSSION As with other hormone receptor systems that have been expressed in oocytes (24 -27), this work shows that increases in cytosolic Ca 2ϩ can be detected by activation of an endogenous Cl Ϫ channel. Identification of the PTHR-stimulated signal as a Cl Ϫ current was accomplished by its inhibition with niflumic acid and by elimination of the signal through clamping the voltage at Ϫ25 mV, the known resting potential for Cl Ϫ in this system. The Ca 2ϩ dependence for this current is demonstrated by its inhibition with injected EGTA, and a lack of inhibition by external EGTA establishes that the signal is not mediated by extracellular Ca 2ϩ . The results are consistent with earlier demonstrations that the PTHR activates increases in cytosolic Ca 2ϩ in wild-type and in transfected cells (7,12,15,16).
Earlier studies indicated that PTH can activate increases in cytosolic Ca 2ϩ by a cAMP-mediated stimulation of a plasma membrane channel in some cells (18). This effect is dependent on extracellular Ca 2ϩ and thus appears unlikely to be responsible for the signals we observe. The PTHR expressed in oocytes does activate production of cAMP (32), and thus a lack of any detectable affect of injected cAMP or its nonhydrolyzable analogues represents additional evidence that our signals are not mediated through this plasma membrane channel. The apparent lack of any effect of cAMP on PTH responses or desensiti- zation are also of interest since protein kinase A has been implicated in the down-regulation of PTHR in some cultured cell systems (33,34). Further work should also examine the possible role of protein kinase C in this system, since it has also been implicated in down-regulation of PTHR (37).
This work is the first demonstration of the cytosolic Ca 2ϩ response to PTH in Xenopus oocytes expressing the PTHR. Earlier work with mRNA extracts from target cells for PTH demonstrated that the receptor can be expressed in oocytes, utilizing the cAMP response to measure receptor function (32,38). Thus, the two primary signaling responses to PTH have now been demonstrated in this system. This is significant since cell systems that express the PTHR do not always show both responses. For example, some cultured bone cells show elevated cAMP in response to PTH, but no detectable increases in cytosolic Ca 2ϩ (14), while cultured keratinocytes show a cytosolic Ca 2ϩ response but no elevation in cAMP (15). Also, some forms of PTH can generate one response but not the other (16), and in one study, cells that were desensitized with regard to the cAMP response still elicited a cytosolic Ca 2ϩ response (17).
The most interesting results from this work are the apparent lack of involvement of InsP 3 in PTHR signaling and the inhibition of signaling by NADP ϩ . The most direct interpretation of our data with regard to the role of InsP 3 is that cytosolic Ca 2ϩ signaling by PTHR in oocytes is not mediated by this second messenger. This interpretation is supported by several observations. First, oocytes that cannot respond to injected InsP 3 still show a strong response to added PTH. Second, cells whose InsP 3 -sensitive Ca 2ϩ stores appear to be depleted by activation of coexpressed 5-HT receptors still respond to added PTH. Third, the response to PTH is consistently slower than that to 5-HT, suggesting that signaling is through a different route, and fourth, the PTH response is strongly inhibited by NADP ϩ and related compounds, none of which have any apparent chemical or biological relationship to InsP 3 .
An alternative interpretation is that frog oocytes possess Ca 2ϩ compartments that are not accessed either by injected materials or by InsP 3 generated by other receptors. Such compartments could be activated by cAMP or InsP 3 released in localized sites near the PTHR, while remaining unresponsive to the injected materials or to increased concentrations of these second messengers generated by activation of other receptor systems. Although we know of no evidence for such compartments, this possibility cannot be excluded at this time. Thus, we cannot unequivocally conclude that InsP 3 is not the signal for cytosolic Ca 2ϩ increases triggered by the PTHR in the frog oocyte. However, at the minimum, our studies demonstrate the existence of unique Ca 2ϩ compartments in oocytes that are responsive to the PTH receptor. The unique nature of these compartments is either their unresponsiveness or their inaccessibility to presently known signals and their sensitivity to NADP ϩ .
Although InsP 3 is probably the most widely studied signal for cytosolic Ca 2ϩ , three alternative second messenger candidates are known: cADPR (41)(42)(43)(44)(45)(46), sphingosine-phosphate (47)(48)(49)(50), and NAADP ϩ (51). None of these have yet been shown to be FIG. 7. Xenopus oocytes expressing the PTHR can be desensitized to injected InsP 3 and still respond to PTH. Membrane cur-rent traces were recorded at a holding potential at Ϫ80 mV. InsP 3 microinjected into oocytes induces an inward current and desensitizes the same oocyte to subsequent InsP 3 injection (top trace). 5-HT 2C receptor-mediated response desensitizes the oocyte to subsequent InsP 3 injection (second trace from top), and InsP 3 injection similarly desensitizes the cell to 5-HT stimulation (third trace from top). However, in a cell that is desensitized to hPTH stimulation, InsP 3 injection still elicited a strong inward current (fourth trace from top), and in a cell that is desensitized to InsP 3 treatment with PTH still elicited a strong inward current (fifth trace).
active in Xenopus oocytes, and there is evidence that the cADPR pathway does not function in this system (39). However, it has also been reported that acetylcholine and thyrotropin-releasing hormone mobilize calcium from functionally separate stores in Xenopus oocytes (40), although the signaling systems for these two responses have not been identified. In the present work, we were unable to demonstrate responses of frog oocytes to either injected cADPR or NAADP ϩ , but we have not yet examined the sphingosine-phosphate pathway.
The inhibition of PTHR signaling by NADP ϩ has not been described previously. This effect is not observed with the serotonin receptor, which is known to utilize the InsP 3 pathway, and it thus provides further evidence that InsP 3 may not be the signal for the PTH system. It is of interest that the most sensitive assay for PTH is based on a procedure that requires activation of glucose-6-phosphate dehydrogenase and production of NADPH (52). The biochemical rationale for this assay is not understood, but our results with NADP ϩ now provide further impetus for examination of the possible role of this coenzyme in biochemical actions of PTH. As indicated in Fig. 8, the inhibition of cytosolic Ca 2ϩ signaling is acute and persists for hours following injection of NADP ϩ . This effect could be the result of direct interaction of NADP ϩ with some regulatory element in the PTH signaling pathway or a more general effect of a significant alteration of the NADP ϩ /NADPH ratio in the cytoplasm and its subsequent impact on the oxidation/reduction or catabolic/anabolic state of the cell. It may be that the signaling by PTHR is modulated by metabolic state since this hormone is known to produce both anabolic and catabolic affects on different cell types under different conditions. Further study of the regulation of signaling by the PTHR in the frog oocyte system may shed light on these possibilities.