Differential Effects of Parathyroid Hormone and Its Analogues on Cytosolic Calcium Ion and cAMP Levels in Cultured Rat Osteoblast-like Cells*

While the stimulatory effect of parathyroid hormone (PTH) on osteoblast-like cell adenylate cyclase is well known, the effect of PTH on cytosolic calcium ion ([Ca2+]i) mobilization is controversial, one group find- ing no effect but others reporting various increases. We investigated the effects on [Ca2+]i of synthetic rat PTH fragment 1-34 (rPTH(1-34)) and two bovine PTH analogues that inhibit PTH's stimulation of adenylate cyclase (bovine 54Tyr-PTH(3-34) and 34Tyr-PTH(7-34)). [Ca2+Ii was measured before, during, and after exposure to PTH analogues in perifused, attached osteoblast-like rat osteosarcoma cells (ROS that had been scrape-loaded with the lumines- cent photoprotein aequorin. Resting [Ca2+]i was 0.094 pM (mean S.D., and rose in a time- and dose-specific way after exposure to rPTH(1-34). concentrations of PTH initial peaks of followed by plateaus. At the initial peak &fold basal,

While the stimulatory effect of parathyroid hormone (PTH) on osteoblast-like cell adenylate cyclase is well known, the effect of PTH on cytosolic calcium ion ([Ca2+]i) mobilization is controversial, one group finding no effect but others reporting various increases. We investigated the effects on [Ca2+]i of synthetic rat PTH fragment 1-34 (rPTH(1-34)) and two bovine PTH analogues that inhibit PTH's stimulation of adenylate cyclase (bovine 8*18Nle, 54Tyr-PTH(3-34) and 34Tyr-PTH(7-34)). [Ca2+Ii was measured before, during, and after exposure to PTH analogues in perifused, attached osteoblast-like rat osteosarcoma cells (ROS 1712. 8) that had been scrape-loaded with the luminescent photoprotein aequorin. Resting [Ca2+]i was 0.094 f 0.056 pM (mean f S.D., n = 103) and rose in a timeand dose-specific way after exposure to rPTH(1-34). At 10"' M rPTH(1-34), [Ca2+]i rose 100% within 30 s to a plateau; higher concentrations of PTH yielded increasing initial peaks of [Ca2+Ii followed by lower plateaus. At lo-' M, the initial peak was &fold basal, or 0.64 2 0.07 p~. Both analogues of PTH were at least partial agonists for [Ca2+]i mobilization and did not reduce peak [Ca2+Ii when co-perifused with rPTH . However, the analogues did reduce significantly rPTH( 1-34)-induced cAMP accumulation and did not increase cAMP accumulation by themselves. Thus, rPTH(1-34) strongly mobilizes [Ca2+]i in ROS 1712.8 cells, at near-physiologic concentrations. Failure of the PTH analogues to block the effect of PTH on [Ca2+Ii while inhibiting the effect on cAMP accumulation suggests separate pathways for PTH activation of adenylate cyclase and mobilization of calcium.
Parathyroid hormone (PTH)'-mediated effects on target cells have been primarily attributed to stimulation of adenyl-* This work was supported in part by Grants AR-32526, DK-38855, and AG-04875 from the National Institutes of Health. Portions of this work were presented at the National Meeting of the American Society for Clinical Investigation, San Diego, CA, May, 1987. 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 U.S.C. Section (9,10) and mouse osteoblast-like MC3TB-E/cells (11).
The initial peak height for [Ca2+]; decreased with decreasing concentrations of rPTH  in the medium (10-6-10-'o M) (Fig. 2). Moreover, a change in the shape of the response curve occurred over this dose interval. At high doses, the response consisted of an initial sharp peak followed by a lower The arrow is dashed to indicate that the height of L, .
is several orders of magnitude off the scale depicted. The distance between two hatch marks represents 1 min; note expansion and contraction of time scale. plateau. With decreasing doses, the initial peak tended to disappear and a uniphasic, plateau-shaped curve remained. While there was a significant dose-dependent increase of initial peak height ( p < 0.05), the shoulder or plateau height remained constant ( p = 0.48) over the dose interval studied. The dose-response relationship for height of the initial [Ca2+]; peaks is shown in Fig. 3. Preliminary results suggest that the threshold concentration of rPTH( 1-34) for increasing [Ca2+]; lies between 10"' and 10"' M (data not shown). Effect of rPTH(1-34) on [Caz+]i in Fibroblasts-In a control study, rat skin-derived fibroblasts were exposed to rPTH(1-34), M (n = 5). No significant changes in [Ca2+]i were recorded, but [Ca2+]; did increase dramatically (from about 0.06 to 0.50 WM) during perifusion of the aequorin-loaded fibroblasts with trifluoperazine M), a putative calmodulin antagonist and inhibitor of membrane Ca,Mg-ATPase.
[Ca2+]; returned toward basal rapidly upon removal of trifluoperazine. In previous studies, we found only small and inconsistent rises of CAMP content in similarly derived human fibroblast cells after exposure to rPTH(1-34) (25).
Effects of PTH Analogues on presence of an excess of either one of two well characterized PTH analogues: bPTH(3-34) and bPTH  M and M, respectively) (26). The cells were perifused with an analogue for 120 s, followed by a 120-s perifusion of rPTH(1-34) plus analogue.

DISCUSSION
The present study verified that PTH has, among its actions on osteoblast-like bone cells, a powerful capacity to mobilize Ca2+ into the cytosolic compartment. The rapidity of this [Ca2+Ii response is similar to the adenylate cyclase response and is at least as sensitive. The Kd values for displacement of radiolabeled PTH from plasma membrane receptors (27,28), activation of adenylate cyclase (27,28), and release of cAMP from rat bone (29) (30)) was not less than that at higher doses. The [Ca2+]i-raising effect of PTH may therefore be one of the most sensitive indexes of PTH action on bone cells. The [Ca2+Ii response we observed was apparently tissue-specific, as it occurred in rat bone-derived ROS 17/2.8 cells, but not in cultured rat dermal fibroblasts.
Our findings differ considerably from published data on PTH modulation of [Ca2+]i in osteoblast-like cells. For example, Boland et al. (6) used quin2-loaded ROS 17/23 cells to study the effects of various calcium-regulating hormones, and found no effect of PTH on [Ca2+]i. In contrast, Lowik et al. (7), using the same fluorescent Ca" probe, found relatively small 25-50% increases of [Ca2+]i during exposure of UMR-106 rat osteosarcoma cells to PTH. We attribute those authors' difficulties in seeing the large PTH-induced changes of [Ca2+]i that we observed primarily to limitations of the quin2 method, which have been clearly described by Grynkiewicz et al. (31). Chief among the problems is buffering of Ca2+ by quin2, which might limit the ability to detect the earliest action of PTH on [Ca2+Ii (the initial peak). Phenotypic drift of their cultured bone cells, with loss of PTH responsiveness, is another possibility.
More recently, Yamada et al. (11) observed modest monophasic increases in [Ca2+Ii in mouse osteoblast-like MC3T3-E l cells loaded with fura-2 and exposed to PTH as did Reid et al. (9) using UMR-106 cells loaded with indo-1. Yamaguchi

TABLE I Effects of rPTH(1-34) and analogues on [Ca2+]i
Maximum initial effects on [Ca2+], in ROS 17/2.8 cells of 2-min perifusions with medium, bPTH(3-34), or bPTH(7-34) alone, followed by addition of rPTH(1-34) alone or in the presence of the PTH analogues (the protocol design is the same as that shown in Fig. 4). The basal period is designated period A. In all cases, the 2min period immediately after the basal (period B) included perifusion with medium alone (first line) or PTH analogue alone (last six lines). The substances perifused in period B were continued in the third 2-min period   7-8). Differences between groups were tested by the least significant difference procedure in order to correct for multiple comparisons. The test is a t test, using in the denominator the pooled variance for all groups instead of the individual group variances. Brackets and *** indicate comparisons that yielded significant differences. Column F was significantly greater than columns A through E , and none of the latter five differed from one another. Note the absence of partial agonism, but highly significant antagonistic effect of the two PTH analogues in this system. *** = p < 0.001. et al. (8), using fura-2, found a transient initial PTH-mediated increase in [CaZ+Ii in the UMR-106 cell line that may be mediated by protein kinase C (Ca2+-and phospholipid-dependent protein kinase) (32), and, contrary to our findings, a later CAMP-dependent rise in [Ca2+Ii. The [Ca"]i response to rPTH(1-34) that we observed in ROS 17/23 cells was biphasic, with an initial peak resembling but larger than that seen by Yamaguchi et al. (8) and Reid et al. (9). In addition, we observed a sustained plateau elevation of [Ca2+]i after the initial peak and did not see the late increase of [Ca2+]i described by Yamaguchi et al. (8). Whether these discrepancies are due to differences in cell lines or indicator systems remains to be established.
We have not yet examined the sources for the Ca2+ appearing in cytosol, but it is tempting to speculate that the curves seen at high PTH doses represent composites of two different responses. Perhaps the initial peak represents mobilization of Ca2+ from intracellular membranes or organelles, and the plateau represents influx of extracellular calcium, as has been described for ascites tumor cells (33). The concept that PTH mobilizes both intracellular and extracellular Ca2+ into the cytosol is supported by the findings of Yamaguchi et al. (8), showing that the removal of extracellular Ca2+ or addition of calcium channel blocker reduced but did not abolish the [Ca2+]i response to bPTH(1-34) in UMR-106 cells. Reid's data (9) are ambiguous: low-calcium medium (-1 FM) abolished the [Ca2+]i response to bPTH(1-34), whereas removal of extracellular Ca2+ by EGTA had no effect on the response to PTH. Again, the reasons for these inconsistencies are unclear. However, numerous pharmacologic probes (34, 35) now available should permit us to address this issue in future studies.
The two amino-truncated analogues of parathyroid hormone, bPTH(3-34) and bPTH(7-34), both exert inhibitory effects on PTH-mediated adenylate cyclase stimulation (23,24). We verified the ability of the two analogues to inhibit rPTH( 1-34)-induced cAMP accumulation under our experimental conditions. In vivo weak agonism has been described for bPTH  in stimulation of adenylate cyclase, induction of hypercalcemia, and stimulation of urinary phosphate excretion (23,24). Lowik et al. (7) also reported weak agonism of bPTH  in raising [Ca2+]i, but did not describe whether the analogue reduced [Ca2+Ii mobilization by active PTH. On the other hand, Reid et al. (9) found that bPTH  alone had no effect on [Ca2+]i and reduced the effect of bPTH(1-34) on [Ca2+Ii in UMR-106 cells, while Yamaguchi et al. (8) obtained similar results with bPTH . In contrast to the above findings, our experiments showed bPTH  to be a partial agonist and bPTH   Our ability to detect increases in [Ca2+]i in response to bPTH  and bPTH  suggests that the aequorin method may have greater sensitivity to small or local changes in [Ca2+Ii than the fluorometric techniques. We have no explanation for failure of the analogues to inhibit rPTH(1-34)-induced [Ca2+Ii elevations in our cells. It is unlikely that species differences in PTH structure played a role here; bPTH(1-34) and rPTH(1-34) are strongly homologous, differing at only 3 amino acid residues (24).
The [Ca2+]i indicator system used for these studies, aequorin luminescence, differs sharply from other methods used recently with bone cells. In cell systems studied so far, aequorin was distributed evenly throughout the cytosol and exerted only minimal buffering capacity on [Ca2+]i (16). The luminescence is unaffected by most other ions of biological significance, except for Mg2+ (16), which may tend to decrease measured levels of [Ca2+Ii. Unlike the fluorescent indicator quin2, which may undergo photobleaching on prolonged light exposure, the luminescence of aequorin stays essentially constant for many hours (15). The main disadvantage of aequorin is its relative scarcity. Other problems primarily encountered in studies on muscle cells, such as relatively slow response time (milliseconds) and slow diffusion rates of aequorin (16) were of no significance within the time frame of responses in this study.
Until recently, aequorin could only be used in systems wherein it could be loaded by techniques such as microinjection or hypoosmotic shock (16,36,37). This hindered its use in cultured adherent mammalian cells. Recently, however, this problem was solved with the development of scrapeloading of cells by McNeil and Taylor (15). Most techniques employing fluorescent indicators have involved the use of cells in suspension. Detachment of cells may lead to pertubations of membrane structure and receptor populations, which might alter responsivity to a given hormonal stimulus. Studies of cells in suspension necessitates the interruption of the fluorescent signal in order to introduce stimulating agents. This may preclude the detection of the earliest cellular responses to an agent. The use of a perifusion system such as ours, however, allows the recording of very rapid cellular responses and simple manipulation of experimental conditions, with rapid exposure to and disposal of the stimulating agents. Furthermore, repetitive testing of the same cells can be made. The system described in this paper will be broadly applicable to many kinds of adherent cells in culture.
In conclusion, our studies established that rPTH(1-34) elicits complex time-and dose-dependent effects on [Ca2+Ii mobilization in cultured osteoblast-like ROS 17/2.8 cells. The maximal responses were considerably greater than those observed with fluorescent indicators and included a pattern of [Caz+Ii response not previously described for those cells. Our data greatly strengthen the hypothesis that transduction of the PTH signal in bone cells involves not only formation of CAMP, but rapid activation of a [Ca2+]i signal. Furthermore, the methods we describe herein will permit a detailed exploration of PTH effects on [Ca2+]i in these and other cell types, and the techniques should be generalizable to a wide variety of cultured mammalian cells. In contrast to their inhibitory effects on PTH stimulation of adenylate cyclase, the PTH analogues bPTH(3-34) and bPTH  were at least partial agonists for [Ca2+Ii mobilization and ineffective antagonists of rPTH( 1-34)-induced [Ca2+]i mobilization. We speculate that these data are clues to separate mediation of PTH's effects on adenylate cyclase and [Ca2+]i, whether it be by two separate PTH receptors (7) or differential regulation at a post-receptor level.