The Orexin OX1 Receptor Activates a Novel Ca2+ Influx Pathway Necessary for Coupling to Phospholipase C*

Ca2+ elevations in Chinese hamster ovary cells stably expressing OX1 receptors were measured using fluorescent Ca2+ indicators fura-2 and fluo-3. Stimulation with orexin-A led to pronounced Ca2+elevations with an EC50 around 1 nm. When the extracellular [Ca2+] was reduced to a submicromolar concentration, the EC50 was increased 100-fold. Similarly, the inositol 1,4,5-trisphosphate production in the presence of 1 mm external Ca2+ was about 2 orders of magnitude more sensitive to orexin-A stimulation than in low extracellular Ca2+. The shift in the potency was not caused by depletion of intracellular Ca2+ but by a requirement of extracellular Ca2+ for production of inositol 1,4,5-trisphosphate. Fura-2 experiments with the “Mn2+-quench technique” indicated a direct activation of a cation influx pathway by OX1 receptor independent of Ca2+ release or pool depletion. Furthermore, depolarization of the cells to +60 mV, which almost nullifies the driving force for Ca2+ entry, abolished the Ca2+ response to low concentrations of orexin-A. The results thus suggest that OX1 receptor activation leads to two responses, (i) a Ca2+ influx and (ii) a direct stimulation of phospholipase C, and that these two responses converge at the level of phospholipase C where the former markedly enhances the potency of the latter.

The recently described hypothalamic peptides called orexins (1) or hypocretins (2) mediate their effects through G proteincoupled receptors called OX 1 and OX 2 receptors (1). The peptides and their receptors are widespread in the hypothalamus, cortex, and brainstem (2)(3)(4)(5). The orexin/hypocretin peptides are encoded by a single mRNA giving rise to a 33-residue orexin-A peptide containing disulfide bridges and a linear 28residue orexin-B (1). Orexin-A has a 10 -100-fold higher affinity and potency for OX 1 receptor as compared with orexin-B, whereas no preference is displayed by the OX 2 receptor (1). The orexins cause robust increases in intracellular Ca 2ϩ both in neurons cultured from rat medial and lateral hypothalamus (6) and spinal cord (7), and when studied using recombinant receptors (1). This has led to the suggestion that the receptors are coupled to the G q family G proteins. Interestingly, the response in neurons is partially dependent on extracellular Ca 2ϩ , which may suggest that the receptors are connected to a Ca 2ϩ influx pathway in neurons (6). Several different pathways for receptor-stimulated Ca 2ϩ entry have been suggested based on functional studies with other G protein-coupled receptors. Suggested pathways include store-operated Ca 2ϩ channels, second messenger-operated channels, as well as Ca 2ϩ -activated Ca 2ϩ channels (reviewed in Refs. 8 and 9).
The aim of this study was to examine in detail the Ca 2ϩ mobilizing actions of orexins on recombinant OX 1 receptors expressed in CHO 1 -K1 cells. The results reveal the presence of a novel amplification mechanism at the level of phospholipase C that is dependent on activation of Ca 2ϩ influx pathway upstream of phospholipase C.

EXPERIMENTAL PROCEDURES
Cell Cultures-To prepare the CHO-hOX 1 -C1 cells used in this study CHO-K1 cells were transfected with a bicistronic vector containing the coding sequence of human OX 1 receptor as described previously for chemokine receptors (10). Neomycin resistant clones were then isolated by limited dilution. They were grown in nutrient mixture (Ham's F-12) medium (Life Technologies, Inc., Paisley, United Kingdom) supplemented with 100 units/ml penicillin G (Sigma), 80 units/ml streptomycin (Sigma), 400 g/ml Geneticin (G418; Life Technologies, Inc.) and 10% (v/v) fetal calf serum (Life Technologies, Inc.) at 37°C in 5% CO 2 in an air ventilated humidified incubator in 260-ml plastic culture flasks (75-cm 2 bottom area; Nunc A/S, Roskilde, Denmark). For Ca 2ϩ , inositol phosphate and IP 3 measurements in suspension, the cells were grown on circular plastic culture dishes (inner diameter, 94 mm; Nunc). For microfluorometry, the cells were grown on circular glass coverslips (inner diameter, 25 mm).
Materials-EGTA and probenecid (p-[dipropylsulfamoyl]benzoic acid) were purchased from Sigma and thapsigargin from RBI (Natick, MA). Fura-2, fura-2 acetoxymethyl ester, and fluo-3 acetoxymethyl ester were purchased from Molecular Probes Inc. ( Ca 2ϩ Measurements in Suspension-The fluorescent Ca 2ϩ indicator fluo-3 was used to monitor changes in intracellular Ca 2ϩ instead of fura-2 since it has a higher K d (11), which allows more accurate estimation of the very high Ca 2ϩ elevations induced by OX 1 receptor stimulation. The cells were harvested using phosphate-buffered saline containing 0.2 g/liter EDTA, loaded with fluo-3 acetoxymethyl ester (4 M, 20 min, 37°C) in TBM, and stored on ice as pellets (medium removed). For the measurement of intracellular free calcium, one pellet was resuspended in TBM at 37°C. The fluorescence was monitored in a stirred quartz microcuvette in the thermostated cell holder of either a Hitachi F-2000, F-4000, or PTI QuantaMaster fluorescence spectrophotometer at the wavelengths 480 nm (excitation) and 540 nm (emission). Experiments were calibrated by adding 60 g/ml digitonin, which gives the maximum value of fluorescence, and 10 mM EGTA, which gives the minimum value of fluorescence. The leaked fluo-3 was measured in separate experiments by adding 10 mM EGTA, which chelates Ca 2ϩ bound to extracellular fluo-3. The corrected fluorescence values and the K d ϭ 400 nM (11) were used to calculate [Ca 2ϩ ] i . In each batch of cells, all the experiments were performed at least in duplicate.
Photomultiplier-based Microfluorometry and Patch-Clamp-For microfluorometric Ca 2ϩ measurements, fura-2 was used instead of fluo-3 since it enables ratiometric Ca 2ϩ measurement. The coverslips with cells were loaded with 1 M fura-2 acetoxymethyl ester (30 min, 37°C) in a buffer containing 140 mM NaCl, 2.8 mM KCl, 2 mM MgCl 2 , 2 mM CaCl 2 , 3 mM glucose, 1 mM probenecid, and 10 mM Hepes (pH 7.4), rinsed with fura-2-free medium and used as the bottom of an open 0.4-ml chamber. The chamber was placed in a thermostat-controlled holder (32°C) on the stage of an inverted microscope (Nikon) and perfused at a rate of 0.7 ml/min with the loading medium lacking fura-2. The cells were excited at 340/380 nm using a monochromator (TILL Photonics, Munich, Germany) that was controlled by EPC 9 fura extension software (HEKA Elektronik GmbH, Lambrecht, Germany). Fluorescence signals were detected by a photomultiplier mounted to a viewfinder (TILL Photonics) and saved by the X-chart version of the EPC 9 software (HEKA). A new reading (340/380) was achieved every 0.12 s by interpolation. Calibration constants for fura-2 were determined according to (12). Orexin-A was applied with a puff-pipette placed 30 -50 m away from the recording cell.
Patch-clamp experiments in the whole-cell configuration were performed using the same setup as above. Recordings were performed with 2-4-megohm pipettes (Kimax-51; Kimble, Vineland, NJ) and an EPC 9 patch-clamp amplifier together with Pulse software (HEKA). Cells were held at a holding potential of Ϫ50 mV. Capacitative transient and series resistance were continuously monitored and compensated for. The extracellular solution was as above. The internal pipette solution contained 140 mM KCl, 2 mM MgCl 2 , 3 mM Mg-ATP, 0.1 mM fura-2, and 10 mM Hepes adjusted to pH 7.2 with KOH. The osmolarity of all solutions utilized in the patch-clamp experiments was 295 Ϯ 5 mosmol/liter. Ca 2ϩ Imaging-For Ca 2ϩ imaging, fura-2 was used instead of fluo-3 since it enables simultaneous measurement of Ca 2ϩ -dependent fluorescence increase (at 340 nm) and Mn 2ϩ -dependent fluorescence quenching (at 360 nm). The coverslips with CHO-hOX 1 -C1 cells were loaded with fura-2 acetoxymethyl ester (4 M, 20 min, 37°C) in TBM, rinsed with nominally Ca 2ϩ -free TBM, and used as the bottom of an open 1-ml chamber. The chamber was placed in a thermostat-controlled holder (37°C) on the stage of an inverted microscope (Nikon). The additions in the chamber were made by perfusion or by removing 0.35 ml of the total volume of 1 ml (of nominally Ca 2ϩ -free TBM), mixing with the reagents to be added and re-adding into the chamber. The cells were excited by alternating 340 and 360 nm light with two monochromators, a chopper under the control of a PTI Image Master (version 1.4 for Windows), a dichroic mirror (DM400, Nikon), and the emission measured through the dichroic mirror and a 470-nm barrier filter with a PTI IC-200 CCD camera. A new ratioed (340/380) image was achieved every second by interpolation. Saved data was later analyzed with FeliX (version 1.4; PTI).
Total Inositol Phosphate Measurements-The experiments were performed essentially as described in Ref. 13. CHO-hOX 1 -C1 cells on culture dishes were loaded with 3 Ci/ml myo-[ 3 H]inositol for 20 h in the culture medium. For the experiments, they were detached as for Ca 2ϩ measurement in suspension, washed and suspended in TBM supplemented with 10 mM LiCl. After a 10-min preincubation in this solution in 37°C, the cells were stimulated with orexin-A, ionomycin, or thapsigargin. When relevant, EGTA was added 30 s before the stimulation.
After 5 or 20 min of stimulation, the cells were rapidly spun down and the reactions were stopped by adding 200 l of 0.1 M NaOH and vortexing. After a 10-min incubation at 37°C, the solutions were neutralized with 80 l of 0.2 M formic acid. The debris were removed, and the total inositol phosphate fraction was isolated with anion exchange chromatography (14). The radioactivity was determined by scintillation counting.
IP 3 Measurements-CHO-hOX 1 -C1 cells were cultured and detached as for Ca 2ϩ measurements in suspension. They were spun down, washed once with TBM, and resuspended in TBM, and the cell number was adjusted to 10 7 cells/ml by counting in a Bü rker chamber. The extraction of IP 3 was performed using the method of Palmer et al. (15) essentially as described (16). Briefly, the reactions (either basal level, orexin-A Ϯ Ca 2ϩ for 7 s, 1.5 mM EGTA for 1 min, or 1.5 mM EGTA for 1 min ϩ orexin-A for 7 s) were terminated by adding ice-cold perchloric acid to a final concentration of 4% (v/v). The precipitate was sedimented by centrifugation at 2000 ϫ g for 15 min at 4°C. The supernatants were neutralized with 1.5 M KOH containing 60 mM Hepes, and 0.5 M Tris-HCl (pH 8.6) was added to a final concentration of 0.1 M (final pH Ϸ 8.6). The resulting KClO 4 sediment was removed by centrifugation at 2000 ϫ g for 15 min at 4°C. The IP 3 concentration of the samples was determined using [ 3 H]IP 3 radioreceptor assay kit (NEN Life Science Products). The assay was performed as described in Ref. 16. Briefly, the standards, blanks (0.4% (w/v) inositol hexakisphosphate, to determine nonspecific binding), and the samples were incubated with the receptor preparation and [ 3 H]IP 3 for 1 h on ice. The tubes were spun at 2000 ϫ g for 10 min at 4°C and the resulting pellets dissolved in 50 l of 0.15 M NaOH. The NaOH/protein solutions were transferred to scintillation vials together with 2 ϫ 100 l of water used to wash each tube, and the radioactivity was determined by scintillation counting.
Calculations of Free Ca 2ϩ Concentrations-The extracellular free [Ca 2ϩ ] ([Ca 2ϩ ] o ) was calculated using the freeware Bound and Determined 3.0 (17) and controlled by measurement with fura-2 (free acid form). Addition of 1.5 mM EGTA in TBM gave an approximate free [Ca 2ϩ ] of 140 nM. Nominally Ca 2ϩ -free buffer (used in microfluorometric Ca 2ϩ measurements) had a Ca 2ϩ concentration of approximately 1 M. Addition of 1 mM EGTA resulted in subnanomolar free [Ca 2ϩ ], which is approximated with 0 Ca 2ϩ in the text and in the figure legend. 1.5 mM EGTA immediately prior to addition of orexin-A, the [Ca 2ϩ ] i response was abolished but a robust [Ca 2ϩ ] i elevation was seen when 1.5 mM Ca 2ϩ was added to saturate the EGTA and restore the [Ca 2ϩ ] o of 1 mM (Fig. 1B). A higher concentration of orexin-A (100 nM) caused an increase in [Ca 2ϩ ] i also when [Ca 2ϩ ] o was reduced to 140 nM (Fig. 1C). To test whether the inhibitory effect of EGTA on the orexin-A response was caused by Ca 2ϩ pool depletion by EGTA, 2 M thapsigargin was added to the cells in the presence of 1 mM and 140 nM extracellular Ca 2ϩ (Fig. 1, D and E). No depletion was observed since a response of similar magnitude was obtained under both conditions.

Orexin-induced Increase in Intracellular Free [Ca 2ϩ ] in Sus
A concentration-response relation for the effect of orexin-A at normal and low [Ca 2ϩ ] o is shown in Fig. 2 50 value was around 80 nM. The concentration dependence of the spike response seen upon Ca 2ϩ readdition showed a bell-shaped curve with a halfmaximal rising phase at 3 nM and a half-maximal falling phase at about 50 nM orexin-A. Fig. 2A also shows the response to 2 M thapsigargin in the presence of 140 nM external Ca 2ϩ . When added after 1 nM orexin-A challenge in 1 mM Ca 2ϩ , the response to thapsigargin in 140 nM Ca 2ϩ was the same as in control conditions, indicating that no change in the thapsigargin-releasable pool had occurred. When added after 100 nM orexin-A, the release by thapsigargin was almost completely abolished, indicating that the pool had been almost totally discharged. The concentration-response relation for orexin-A-induced accumulation (20 min) of total inositol phosphates in the presence of Li ϩ is shown in Fig. 2B. In 1 mM Ca 2ϩ a concentration-dependent increase in inositol phosphates was seen with an EC 50 around 1.5 nM. However, when the accumulation was measured in the presence of 140 nM Ca 2ϩ , no detectable accumulation could be seen. Ionomycin (1 M) increased the total inositol phosphate accumulation in the presence of 1 mM Ca 2ϩ by 72.4 Ϯ 4.7%, whereas thapsigargin (2 M) did not (5.3 Ϯ 0.6%; mean Ϯ S.E., n ϭ 3). In the presence of 140 nM Ca 2ϩ , neither of these compounds affected the accumulation of inositol phosphates (3.5 Ϯ 0.3% and Ϫ9.5 Ϯ 0.9%, respectively).
The concentration-response relation for the rapid (7 s) orexin-A-induced IP 3 accumulation in similar conditions is shown in Fig. 2C. Reduction of [Ca 2ϩ ] o to 140 nM caused a considerable shift in the concentration-response relation. Whenever 1 mM [Ca 2ϩ ] o was restored after orexin-A in conditions similar to those for the fluo-3 experiments in Figs. 1B and 2A, the concentration-response relation was almost identical to that in the presence of extracellular Ca 2ϩ except that the maximum response might have been lower. Thus, the IP 3 response has a similar extracellular Ca 2ϩ dependence as the Ca 2ϩ response except that the orexin-A has a 10-fold lower potency for the IP 3 response as compared with the Ca 2ϩ response. To investigate whether phospholipase C is involved in the external Ca 2ϩ -dependent response to 1 nM orexin-A, the effect of the phospholipase C inhibitor U-73122 was tested. This compound, however, caused a considerable rise in [Ca 2ϩ ] i on its own at a concentration as low as 1 M (data not shown). This nonspecific effect is in line with other studies (18), and thus this inhibitor could not be used to probe a role of phospholipase C activation.
A bell-shaped relation was observed when the stable phase of Ca 2ϩ elevation seen in Fig. 1A was plotted as a function of the orexin-A concentration (Fig. 3A). Thus, a more stable phase of the signal dominates at low concentrations of orexin-A as compared with high concentrations where this phase is virtually absent. A similar result was obtained with orexin-B although, as reported earlier for the OX 1 receptor (1), the potency of this ligand was 10 times lower than that of orexin-A (data not shown).
Extracellular Ca 2ϩ Dependence of the Responses to Orexin-A-In order to further characterize the Ca 2ϩ dependence of the response to orexin-A, the stimulation was performed with a puff-pipette in a close contact with the cell being monitored. Cells stimulated with 10 nM orexin-A in Ca 2ϩ -free conditions did, as expected (see Fig. 2A), not respond with an elevation of [Ca 2ϩ ] i (Fig. 4A). If the bath perfusion immediately after the challenge was changed to a solution containing 2 mM Ca 2ϩ , there was an immediate robust increase in [Ca 2ϩ ] i even though orexin-A was no longer present in the extracellular solution (Fig. 4A). If 2 mM Ca 2ϩ was present together with orexin-A in the puff-pipette, the cells perfused either in 0 or 2 mM extracellular Ca 2ϩ responded to the orexin-A-Ca 2ϩ -challenge with an immediate robust increase in [Ca 2ϩ ] i (Fig. 4B). Stimulation with the vehicle instead of orexin-A in any of these cases did not cause any response (data not shown). This indicates that the absence of orexin-A response under Ca 2ϩ -free conditions is not caused by the depletion of intracellular Ca 2ϩ stores, but by the absence of the Ca 2ϩ influx.
Orexin-A response was also investigated in single cells under voltage-clamp conditions. When the normal holding potential of Ϫ50 mV was changed to ϩ60 mV, which nearly nullifies the driving force for Ca 2ϩ influx, the orexin-A response was abolished. A rapid change back to Ϫ50 mV in the presence of continued orexin-A application resulted in an immediate Ca 2ϩ elevation. This indicates that, at this orexin-A concentration, the trigger for the Ca 2ϩ response is a Ca 2ϩ influx, most likely a channel.
IP 3 Receptor Dependence of the Responses to Orexin-A-2-APB has been reported to block IP 3 -mediated Ca 2ϩ mobilization without affecting receptor-dependent IP 3 production (19). Fig. 5A shows that 2-APB did not affect the orexin-A-induced Ca 2ϩ elevation in the presence of 2 mM Ca 2ϩ . However, in the presence of 140 nM Ca 2ϩ , the Ca 2ϩ response was almost completely abolished (Fig. 5B). Xestospongin C has been shown to block some subtypes of the IP 3 receptor. For blocking in intact cells, this compound requires preincubation for about 20 min (20,21). Therefore, these experiments were performed in suspension. Cell suspensions were preincubated for 20 min with 20 M xestospongin C before challenge with 3 or 100 nM orexin-A. In the presence of 1 mM extracellular Ca 2ϩ , xestospongin C did not affect Ca 2ϩ at either concentration of orexin-A. However, with 140 nM extracellular Ca 2ϩ , the response to 100 nM orexin-A was almost totally blocked. Thus, both 2-APB and xestospongin C give the same result, e.g. have no effect in the presence of 1 mM extracellular Ca 2ϩ but block the orexin-A response in the presence of 140 nM extracellular Ca 2ϩ .

Mn 2ϩ Influx in Responses to Orexin-A as Determined with
Fura-2-Many of the Ca 2ϩ influx pathways activated by G protein-coupled receptors are also permeable to Mn 2ϩ , and quenching of the fluorescence of fura-2 by this cation can be used for monitoring the activation of such pathways (22). Cells were challenged with different concentrations of orexin-A in nominally Ca 2ϩ -free solutions to see whether orexin-A activates an influx pathway at concentrations that do not cause intracellular Ca 2ϩ mobilization. The fura-2 response in individual cells was viewed by monitoring the fluorescence at 340 nm (to detect the fura-2-Ca 2ϩ complex) and 360 nm (insensitive to changes in Ca 2ϩ but sensitive to Mn 2ϩ binding). 100 nM orexin-A caused a Ca 2ϩ release as indicated by the fluorescence increase at 340 nm, whereas no response was seen at 360 nm (Fig. 6A). A further addition of 100 M Mn 2ϩ quenched the fura-2 fluorescence at both wavelengths, indicating an influx through, e.g., store-operated Ca 2ϩ channels. The addition of 100 M Mn 2ϩ before stimulation with 100 nM orexin-A did not result in any fluorescence response (Fig. 6B). After the addition of orexin-A, however, a similar Ca 2ϩ release and Mn 2ϩ influx, as in Fig. 6A, was seen. Simultaneous addition of 100 nM orexin-A and 100 M Mn 2ϩ resulted in rapid fluorescence increase at 340 nm and a decrease at 360 nm (Fig. 6C). Simultaneous addition of 100 M Mn 2ϩ and 1 nM orexin-A resulted in a quenching of a the fluorescence at both wavelengths (Fig.  6D). A further addition of 10 nM orexin-A led to a further quenching at both wavelengths. This indicates that 1 and 10 nM orexin-A, which do not activate Ca 2ϩ release in these nominally Ca 2ϩ -free conditions, still activate a cation influx pathway seen as Mn 2ϩ influx.  Fig. 2). ***, p Ͻ 0.001 (Student's non-paired two-tailed t test).

Effect of Thapsigargin-induced
latter phase was absent in the presence of 140 nM extracellular Ca 2ϩ (Fig. 1E), indicating that the sustained phase is dependent on extracellular Ca 2ϩ . When added during the sustained phase of thapsigargin-induced Ca 2ϩ elevation, orexin-A (3 nM) had no effect on [Ca 2ϩ ] i in suspensions of CHO-hOX 1 -C1 cells (Fig. 7A). If the Mn 2ϩ influx was measured under similar conditions of store depletion (i.e. after thapsigargin), but in the absence of added extracellular Ca 2ϩ , a considerable and increasing quenching of fura-2 fluorescence was seen (Fig. 7B). This indicates that thapsigargin itself causes a strong cation influx, which is unveiled by Mn 2ϩ . Orexin-A caused very little additional increase in the Mn 2ϩ influx (Fig. 7B, thin versus thick line; see also the legend to the figure).

DISCUSSION
The results of the present study strongly support the idea that the primary response upon activation of the OX 1 receptor is the activation of a Ca 2ϩ influx pathway, which triggers the phospholipase C activation. Thus, no changes in [Ca 2ϩ ] i , total inositol phosphates, or IP 3 could be detected upon activation of the receptor with low concentration of orexin-A when the [Ca 2ϩ ] o was reduced immediately before the challenge. Additionally, reduction of the driving force for the Ca 2ϩ influx had similar effect. Furthermore, low concentration of orexin-A in nominally Ca 2ϩ -free conditions caused a quenching of fura-2 fluorescence by Mn 2ϩ in conditions where no change in [Ca 2ϩ ] i could be detected. The fact that purified phospholipase C-␤ is activated both by Ca 2ϩ and by G protein subunits ␣ q and ␤␥, would allow such an amplification mechanism (23). Purified phospholipase C-␤1 has been shown to be synergistically activated by both Ca 2ϩ and G q (24). As shown here, neither thapsigargin nor ionomycin alone is able to activate accumulation of inositol phosphates to the same extent as orexin-A. This suggests that Ca 2ϩ alone is not sufficient for phospholipase C-␤ activation but an additional receptor-associated mechanism is also required. An enhancement of the maximum IP 3 production by elevated [Ca 2ϩ ] i has been previously shown for muscarinic receptors (25,26). However, there has not been any indication of any receptor-operated pathway that would enhance the primary IP 3 response; on the contrary, the potentiating effect of Ca 2ϩ on IP 3 production has been seen after activation of voltage-sensitive Ca 2ϩ channels (25) or artificial regulation of [Ca 2ϩ ] i in permeabilized cells (26). The results of the present study suggest that the primary response of OX 1 receptors almost completely relies on a receptor-operated, extracellular Ca 2ϩ -dependent amplification mechanism, since the concentration-response curve is shifted 100-fold to the right when this mechanism is disabled. Such a mechanism has not been described for any other receptor.
The lack of Ca 2ϩ elevation due to depletion of intracellular stores in 140 nM extracellular Ca 2ϩ can be excluded as high concentrations of orexin-A and thapsigargin were able to cause Ca 2ϩ mobilization in these conditions. The same was seen in the microfluorometric experiments, where even longer perfusion with even lower Ca 2ϩ concentrations did not result in reduction of orexin-A response, which was immediately fully restored when orexin-A was added together with Ca 2ϩ . Therefore, it is evident that the response at low concentration of orexin-A is dependent on extracellular Ca 2ϩ . Although xestospongin C and 2-APB block the response to high orexin-A concentrations in 140 nM extracellular Ca 2ϩ , the response to low and high orexin-A concentrations in 1 mM extracellular Ca 2ϩ is unaffected, suggesting that influx of Ca 2ϩ into the cells is the primary pathway. Ca 2ϩ elevation via the OX 1 receptor in these cells may thus reflect the function of a receptor-operated influx pathway.
Previously, a variety of phosphatidylinositol hydrolysis-associated Ca 2ϩ influx pathways have been described using electrophysiological techniques (27)(28)(29)(30) and measurements of [Ca 2ϩ ] i (30 -33). Most of the described pathways for Ca 2ϩ entry appear to be activated secondary to IP 3 production. Inositol phosphates have been shown to activate Ca 2ϩ entry pathways in some cells (34 -36). There are, however, also reports on Ca 2ϩ entry pathways (involving G protein-coupled receptors) which are not secondary to phospholipase C activation (37)(38)(39)(40). Potential activators include, e.g., lipid metabolites (41,42). Most receptors coupled to mobilization of Ca 2ϩ activate so-called store-operated Ca 2ϩ channels. An often found Ca 2ϩ current in this category is I CRAC , which is activated in many cells by Ca 2ϩ store depletion (43). A family of non-voltage-activated Ca 2ϩ channels called Trp has been identified. Although some members of this family show some properties similar to store-operated channels (44 -48), others appear to be fairly distinct (49 -51). At least some members have been reported to be activated by lipid metabolites (52,53). Evidence has been presented that one channel belonging to this family, called TRPL, is directly activated by the G␣ 11 G protein subunits (54).
The results shown here suggest that the orexin-A mediated Ca 2ϩ elevation is not primarily due to an inositol phosphateactivated Ca 2ϩ influx pathway. At low orexin-A concentrations, both the Ca 2ϩ and IP 3 responses required extracellular Ca 2ϩ . The same was reflected in the assay of inositol phosphates. In addition, the IP 3 production in the presence of Ca 2ϩ required 10 times higher concentrations of orexin-A than the Ca 2ϩ response. Accumulation of inositol phosphates was seen at lower concentrations of orexin-A, so it cannot be completely excluded the Ca 2ϩ influx would be amplified by an inositol phosphateactivated mechanism.
The store-operated mechanism for Ca 2ϩ entry also appears unlikely. Store-operated Ca 2ϩ channels are activated by pool depletion (43), whereas the extracellular Ca 2ϩ -dependent response here is independent of IP 3 production and Ca 2ϩ pool depletion as well as insensitive to blockers of IP 3 -receptor mediated Ca 2ϩ release. Also, influx of Mn 2ϩ was initiated immediately by the addition of low concentrations of orexin-A, whereas activation of I CRAC by pool depletion is usually delayed (7 s latency; Ref. 43). The stable level of Ca 2ϩ elevation as well as part of the Ca 2ϩ response seen upon Ca 2ϩ readdition could well be due to a store-operated Ca 2ϩ channel. Secondary Ca 2ϩ influx pathways like I CRAC show very fast inactivation ( around 100 ms) by intracellular Ca 2ϩ (55,56). This could explain the bell-shaped relation of the response to Ca 2ϩ readdition and the steady state Ca 2ϩ level upon the orexin-A concentration. Thus, the elevated Ca 2ϩ could reduce the influx. However, the bell shape could also result from, e.g., (i) a faster inactivation of the receptor response or (ii) a stronger activation of Ca 2ϩ extrusion or both (see also below).
Whether the Ca 2ϩ response at low orexin-A concentrations is exclusively due to influx or whether it is part of a Ca 2ϩ -dependent amplification mechanism is worth some consideration. Mn 2ϩ influx seen at low orexin-A concentrations supports the concept of inositol phosphate-and Ca 2ϩ -independent activation of an influx pathway, and a possible mechanism involving IP 3 -sensitive release is unlikely in light of the discussion above. The lack of an effect of orexin-A at steady state after the thapsigargin elevation could be explained by a markedly en-hanced Ca 2ϩ extrusion. Evidence for this was the considerable Mn 2ϩ influx, which was apparent in conditions where [Ca 2ϩ ] i was close to the basal level. Thus, low Ca 2ϩ levels are maintained despite an active Ca 2ϩ influx pathway. This would mean that strong activation of Ca 2ϩ extrusion by Ca 2ϩ (and calmodulin) limits the Ca 2ϩ elevation.
The results of this study thus suggest that activation of a Ca 2ϩ influx pathway is the primary response upon activation of OX 1 receptors. Moreover, activation of the inositol phosphate pathway appears to be secondary to the activation of this pathway; only very high concentrations (1 M) of orexin-A were able to activate IP 3 production in the absence of Ca 2ϩ influx. The reason for the requirement for the amplification of the Ca 2ϩ release is not clear at the moment. Even a small overall Ca 2ϩ influx can result in high submembrane Ca 2ϩ levels, which are enough to result in full Ca 2ϩ stimulation of phospholipase C. Since the inositol phosphate and IP 3 production at moderate orexin-A concentrations (1-100 nM) required external Ca 2ϩ , the trigger of the influx has to be upstream of phospholipase C. The mechanism would thus be a direct or G protein-mediated activation of an influx pathways in close apposition to phospholipase C. It may be, that the signaling mechanisms, e.g. G proteins, for the Ca 2ϩ influx and direct phospholipase C activation are different, i.e. that the latter is activated only at higher concentrations of orexin-A. This is different from the usually described sequence of activation of G q -coupled receptors. This raises the possibility that activation of a parallel Ca 2ϩ influx pathway may be involved in the signal transduction of other receptors as well. The parallel activation of phospholipase C and Ca 2ϩ influx by the OX 1 receptor may represent compartmentalization of the signal transductions, i.e. signal steering to the interior of the cells or to the submembrane areas. It is thus also possible that the Ca 2ϩ influx is involved in activation of other cellular effector pathways like ion channels etc.