Inhibition of store-operated Ca²⁺ entry by extracellular ATP in rat brown adipocytes

Abstract

1. Modulation of intracellular free Ca²⁺ concentration ([Ca²⁺]_i) by extracellular ATP was investigated in cultured adult rat brown adipocytes using the fluorescent Ca²⁺ indicator fura-2.

2. Bath application of ATP in micromolar concentrations caused a large increase in [Ca²⁺]_i in cells previously stimulated with noradrenaline. This ATP-induced [Ca²⁺]_i increase exhibited a monotonic decline to near the resting levels within approximately 2 min, even in the continued presence of the agonist.

3. The magnitude and time course of the [Ca²⁺]_i increase in response to ATP were not significantly affected by removal of extracellular Ca²⁺, suggesting that a mobilization of intracellular Ca²⁺ primarily contributes to the increase.

4. The [Ca²⁺]_i increase in response to ATP was sensitive to inhibition by suramin, suggesting the involvement of P2 purinoceptors in the response.

5. Thapsigargin (100 nm) evoked a gradual and irreversible increase in [Ca²⁺]_i which was entirely dependent upon extracellular Ca²⁺, providing functional evidence for the expression of store-operated Ca²⁺ entry in these brown adipocytes.

6. Extracellular ATP at a concentration of 10 μm depressed this thapsigargin (100 nm)-induced [Ca²⁺]_i increase by 92 ± 3 % (n= 8 cells), strongly suggesting that ATP inhibits an influx of Ca²⁺ across the plasma membrane through the store-operated pathway. Bath application of phorbol 12-myristate 13-acetate (PMA, 5 μm) did not affect the thapsigargin-induced [Ca²⁺]_i increase, indicating that the inhibitory action of ATP is not mediated by activation of protein kinase C (PKC).

7. These results indicate that extracellular ATP not only mobilizes Ca²⁺ from the intracellular stores but also exerts a potent inhibitory effect on the store-operated Ca²⁺ entry process in adult rat brown adipocytes.

There is increasing evidence indicating that ATP is released as a cotransmitter with noradrenaline from sympathetic nerve terminals and elicits potent, diverse and relevant physiological effects by interacting with ATP-specific receptors on the cell surface (for reviews see Harden et al. 1995; Burnstock, 1996). Previous workers have investigated the effects of ATP as well as noradrenaline on the activities of brown adipocytes in order to elucidate the mechanisms underlying the sympathetic control of cell functions. In neonatal rats, micromolar concentrations of extracellular ATP stimulate multiple activities of brown adipocytes, namely, an elevation in cytosolic free Ca²⁺ (Lee & Pappone, 1997), an increase in cell membrane capacitance (Pappone & Lee, 1996), a modulation of voltage-gated K⁺ currents (Wilson & Pappone, 1999) and a modest facilitation of heat production (Lee & Pappone, 1997). These actions of ATP were suggested to be mediated mainly through P2Y purinoceptors. On the other hand, in brown adipocytes α-adrenoceptor activation with noradrenaline has been shown to elevate cytosolic levels of inositol 1,4,5-trisphosphate (InsP₃; Nånberg & Putney, 1986; Schimmel et al. 1986) and free Ca²⁺ (Wilcke & Nedergaard, 1989; Lee et al. 1993), while the stimulation of β-adrenoceptors by noradrenaline is known to primarily produce an acute thermogenic effect by activating cAMP-protein kinase A signalling cascade (Nicholls & Locke, 1984). Thus, intracellular free Ca²⁺ levels, which have been demonstrated to modify diverse cellular processes such as gene expression and cell proliferation in brown adipocytes (Lee et al. 1993), appear to be regulated not only by noradrenaline through an α-adrenoceptor but also by ATP via a P2Y purinoceptor. However, interactions between noradrenaline and ATP in the control of intracellular free Ca²⁺ levels have yet to be fully clarified.

In a variety of electrically non-excitable cells, stimulation of diverse plasma membrane receptors leading to an InsP₃-dependent Ca²⁺ release from endoplasmic reticulum Ca²⁺ stores has been shown to be followed by activation of Ca²⁺ influx across the plasma membrane through a store-operated pathway (Putney, 1986; for reviews see Putney, 1990; Berridge, 1995; Parekh & Penner, 1997). The depletion of intracellular Ca²⁺ stores has been proposed to activate this Ca²⁺ entry process, although the precise signalling cascade responsible for Ca²⁺ entry remains to be fully identified. A store-operated pathway has been suggested to be a major Ca²⁺ entry mechanism in non-excitable cells, and most Ca²⁺ entering the cell through this pathway is likely to refill the intracellular Ca²⁺ stores and thus contribute to subsequent release from the stores. The stimulation of a P2 purinoceptor has been reported to be coupled to activation of the store-operated Ca²⁺ entry mechanism in various cell types (C6-2B glioma cells: Munshi et al. 1993, Chiono et al. 1995; rat megakaryocytes: Somasundaram & Mahaut-Smith, 1994; HT29 colonic epithelial cells, Kerst et al. 1995; human thyrocytes, Schöfl et al. 1995a; human gastric mucous cells: Schöfl et al. 1995b; calf pulmonary artery endothelial cells: Madge et al. 1997; human glioblastoma cells: Hartmann & Verkhratsky, 1998; for review see Parekh & Penner, 1997). It still remains unknown, however, whether this Ca²⁺ entry mechanism also operates in brown adipocytes.

The aim of the present study was to examine (i) the presence of the store-operated Ca²⁺ entry mechanism and (ii) the interactions of ATP and noradrenaline with this Ca²⁺ entry process, using fura-2-loaded rat brown adipocytes. Our results indicate that store-operated Ca²⁺ entry indeed exists and also exhibits a high sensitivity to inhibition by extracellular ATP in these brown adipocytes.

METHODS

Materials

Male 3-week-old Sprague-Dawley rats were purchased from Charles River Japan Inc. (Yokohama, Japan) and fed ad libitum for at least 1 week before use. Fraction V bovine serum albumin (BSA) was purchased from Intergen (Purchase, NY, USA), class II crude collagenase from Worthington Biochemical (Freehold, NJ, USA), and DNase I from Boehringer Mannheim Co. (Tokyo, Japan). Fura-2 acetoxymethyl ester (fura-2 AM) was obtained from Dojin Chemicals (Kumamoto, Japan), and suramin sodium salt, thapsigargin and phorbol 12-myristate 13-acetate (PMA) were from Wako Pure Chemicals Industries Ltd (Osaka, Japan). Adenosine 5′-triphosphate (ATP, disodium salt), adenosine 5′-diphosphate (ADP, sodium salt), uridine 5′-triphosphate (UTP, sodium salt), (−)noradrenaline hydrochloride, adenosine 5′-O-(3-thiotriphosphate) (ATPγS, tetralithium salt), α,β-methylene adenosine 5′-triphosphate (α,β-methylene ATP, lithium salt), ionomycin, valinomycin and poly-L-lysine were purchased from Sigma Chemical Co. (St Louis, MO, USA). Rabbit anti-uncoupling protein (UCP) serum was kindly provided by Dr T. Kawada (Kyoto University, Kyoto, Japan). All other reagents were obtained from Nacalai Tesque (Kyoto, Japan).

Cell isolation and culture

The method used for isolating brown adipocytes was a modification of that described previously (Omatsu-Kanbe et al. 1996). Rats aged 4–7 weeks were kept at 5°C for 5–8 h with free access to food and water in order to deplete stored lipid in brown adipose tissues. Rats were then deeply anaesthetized by an overdose of sodium pentobarbital (50 mg kg⁻¹, intraperitoneal injection) and were killed by decapitation. Brown adipose tissue, carefully dissected out from interscapular regions, was placed in Krebs-Ringer bicarbonate Hepes (KRBH) buffer supplemented with BSA at a concentration of 1 % (w/v) under sterile conditions. KRBH buffer contained (mm): NaCl, 120; KH₂PO₄, 4; CaCl₂, 2; MgSO₄, 1; NaHCO₃, 10; Hepes, 30 (pH adjusted to 7.4 with NaOH). Brown adipose tissue obtained from these cold-exposed rats was found to contain less lipid and therefore sank easily, as has been previously reported for cold-stressed neonatal rat brown adipose tissue (Lucero & Pappone, 1989). Both muscle and other tissues were then carefully trimmed away from the brown adipose tissue mass with round-tipped forceps to avoid any damage of the surface. The tissue was then minced and digested with a combination of collagenase and DNase I (7.5 mg ml⁻¹ and 0.5 mg ml⁻¹, respectively) at 37°C for 20 min. During the incubation, the mixture of cells and tissue fragments was dispersed by pipetting several times with a 3 ml plastic pipette. After digestion, the reaction mixture was filtered through a 100 μm nylon mesh. The filtrate was transferred to a 15 ml centrifuge tube and centrifuged at 100 g for 5 min, and then the floating tissue debris, white adipocytes and the buffer were discarded. The pelleted cells were resuspended in KRBH-BSA buffer and washed twice by centrifugation in this buffer. The cells were then suspended in the culture medium and washed three more times by centrifugation. The culture medium consisted of Dulbecco's modified Eagle's medium (DMEM) with 10 % fetal bovine serum, 100 μg ml⁻¹ penicillin and 0.1 mg ml⁻¹ streptomycin. After a final wash, the cells were resuspended in the culture medium and seeded on glass cover slips pre-coated with 0.1 % poly-L-lysine in 35 mm plastic culture dishes containing 3 ml culture medium. The brown adipocytes isolated by this procedure were found to contain very little fat, and most of the cells sank to the bottom of dishes and adhered to the cover slips within 30 min. All procedures were performed in plastic containers.

The cells were then maintained at 37°C in a humidified atmosphere of 95 % O₂ and 5 % CO₂. The majority of the cells isolated and cultured in this way are considered to be brown adipocytes, as judged by their morphological characteristics, such as the small size of the multilocular lipid droplets and the presence of a large number of mitochondria. These brown adipocytes consisted of mature-type cells with diameters of 20–35 μm and immature-type cells, with diameters of approximately 15 μm, classified as brown pre-adipocytes (Goglia et al. 1992). The mature brown adipocytes accumulated numerous lipid droplets during culture and retained the ability to respond to both noradrenaline and ATP for at least 6 days, while the immature-type cells neither responded to these stimulants nor accumulated any lipid, even after 6 days of culture. The mature-type brown adipocytes cultured for 1–4 days were used for experiments for the measurement of [Ca²⁺]_i. All experiments were carried out according to the guidelines laid down by the Shiga University of Medical Science Animal Care Committee.

Immunofluorescent staining

The cells were fixed and immunostained based on standard immunohistochemical methods for tissue sections. Cells cultured on cover slips were rinsed with ice-cold 0.1 M phosphate-buffered saline (PBS, pH adjusted to 7.4), and were fixed with 4 % paraformaldehyde, 0.2 % picric acid and 0.35 % glutaraldehyde in 0.1 M phosphate buffer (pH adjusted to 7.4) at 5°C for 20 min. The fixed cells were washed three times with 15 % sucrose in 0.1 M phosphate buffer (pH adjusted to 7.4) and washed three more times with 0.1 M PBS containing 0.3 % Triton X-100. The cells were then incubated with anti-UCP antiserum (1:500) at 5°C for 3 days and stained with fluorescein isothiocyanate (FITC)-labelled anti-rabbit immunoglobulin G (1:100). After washing with PBS-Triton X-100, the cells were examined using a confocal laser microscope (MRC 600, BioRad, UK; Diaphot 300, Nikon, Tokyo, Japan).

Measurement of [Ca²⁺]_i

Changes in [Ca²⁺]_i were measured using the fluorescent Ca²⁺ indicator fura-2 from cells in a recording chamber perfused with the standard bath solution consisting of KRBH buffer with 5.6 mm glucose and 0.2 % BSA. The cells cultured on cover slips were loaded with membrane-permeant fura-2 AM (5 μm) in the dark for 30 min at 37°C and were washed twice in the standard bath solution. Fura-2-loaded cells were then incubated in the standard bath solution for 30 min at 37°C and were transferred to the recording chamber (approximately 0.5 ml in volume) mounted on the stage of an inverted microscope (Diaphot, Nikon, Tokyo, Japan). The recording chamber was maintained at 36°C and was perfused continuously at a rate of 1.5 ml min⁻¹ with the standard bath solution or test solutions. The fluorescence measurement of [Ca²⁺]_i was performed using a microspectrofluorometer (CAM-220, Japan Spectroscopic Co., Tokyo, Japan). The fluorescence intensity increased with an increase in incubation time and became saturated after 30–60 min of incubation, as has been described for neonatal rat brown adipocytes (Lee et al. 1993).

The [Ca²⁺]_i values were calculated from the fluorescence ratios (R) acquired at 340 and 380 nm excitation wavelengths using the following equation (Grynkiewicz et al. 1985):

where K_d is the dissociation constant of fura-2 for Ca²⁺ (K_d= 224 nm, Grynkiewicz et al. 1985), S_f2/S_b2 is the fluorescence ratio of the 380 nm signal in the absence of Ca²⁺ to that in the presence of saturating Ca²⁺ (S_f2/S_b2= 2.35 ± 0.11, mean ±s.e.m., n= 8 cells), and R_min and R_max are minimal and maximal fluorescence ratios, respectively. R_min, measured by incubating cells with a nominally Ca²⁺-free bath solution supplemented with 5 mm EGTA and 5 μm ionomycin, was 0.47 ± 0.03 (mean ±s.e.m., n= 8 cells) and an R_max value of 4.85 ± 0.22 was obtained by incubating cells with the standard bath solution supplemented with 3 mm CaCl₂ (total [Ca²⁺]_o= 5 mm) and 5 μm ionomycin. All data were stored either on videotape after being digitized by a PCM processor (PCM-501 modified, SONY, Tokyo, Japan) or on a digital audiotape using a PCM data recorder (RD-120TE, TEAC, Tokyo, Japan) and were then transferred to a magnetic optical disk through a DigiData 1200 interface (Axon Instruments Co., Foster City, CA, USA) for a later analysis with pCLAMP software (Axon Instruments Co.).

The nominally Ca²⁺-free bath solution used for some experiments shown in Figs 4 and 7 was prepared by simply omitting CaCl₂ from the standard bath solution.

Figure 4. Effects of the removal of extracellular Ca²⁺ on the noradrenaline- and ATP-evoked [Ca²⁺]_i responses.

Periods of exposure to 1 μm noradrenaline (NA), 10 μm ATP and the nominally Ca²⁺-free bath solution (0 Ca²⁺) were indicated by the horizontal bars below the records. A, a single cell was exposed to noradrenaline in the standard bath solution ([Ca²⁺]_o= 2.0 mm) and in nominally Ca²⁺-free bath solution, as indicated. Note that noradrenaline evoked only a transient and small increase in [Ca²⁺]_i in the nominally Ca²⁺-free solution. B, the bath solution was changed to the nominally Ca²⁺-free solution with noradrenaline present. The noradrenaline-induced increase in [Ca²⁺]_i was nearly completely abolished in the absence of extracellular Ca²⁺. C, a single cell was stimulated with noradrenaline in standard bath solution and then with ATP in nominally Ca²⁺-free bath solution. Each trace is obtained from different cells and is representative of 6–8 cells.

Figure 7. Presence of store-operated Ca²⁺ entry in brown adipocytes.

Periods of exposure to 1 μm noradrenaline (NA), 100 nm thapsigargin (TG), and the nominally Ca²⁺-free bath solution (0 Ca²⁺) were indicated by the horizontal bars below the records. A, a single cell was successively exposed to noradrenaline and thapsigargin. Extracellular Ca²⁺ was transiently removed during the presence of and after a washout of thapsigargin. B, a cell, previously stimulated with noradrenaline, was exposed to thapsigargin in the absence of extracellular Ca²⁺. Note that even after the withdrawal of thapsigargin a large and irreversible increase in [Ca²⁺]_i was evoked by restoring extracellular Ca²⁺. C, a cell, previously stimulated with noradrenaline, was exposed to thapsigargin, which was followed by the addition of noradrenaline in the presence of thapsigargin. Each trace is obtained from different cells and is representative of 6–8 cells.

Data analysis and statistics

The results are expressed as means ±s.e.m. and n is the number of cells studied. Statistical comparisons were made using Student's t test for paired or unpaired data where appropriate, and differences were considered to be significant at P < 0.05.

RESULTS

Expression of uncoupling protein in cultured brown adipocytes

Brown adipocytes have been demonstrated to express the uncoupling protein (UCP) in their mitochondria, which uncouples substrate oxidation from ATP production, thus resulting in non-shivering thermogenesis. Using immunofluorescence techniques we first examined whether the mitochondrial UCP is expressed in cultured adult rat brown adipocytes prepared using the procedures described above. Figure 1 demonstrates representative results obtained using the well-characterized immune serum (Kawada et al. 1991; Hikichi et al. 1993). The protein expression of UCP can be clearly detected as bright signals in cells which had been cultured for 1 (left panel) and 4 days (right panel). Single brown adipocytes were thus found to maintain the characteristic expression of mitochondrial UCP for a period of 1–4 days of culture.

Figure 1. Immunofluorescent labelling of the mitochondrial UCP in cultured brown adipocytes.

Cells cultured for 1 day (left) and 4 days (right) were fixed and incubated with anti-UCP antiserum. Immunoreactive proteins were stained with FITC-labelled anti-rabbit immunoglobulin G and were then examined using a confocal laser microscope. Immunolabelled proteins are detected as bright signals. Numerous lipid droplets accumulating in the cells are seen as dark circles. Calibration bar in the right panel represents 10 μm and also refers to the left panel.

Properties of noradrenaline- and ATP-evoked [Ca²⁺]_i responses

We then examined [Ca²⁺]_i responses of brown adipocytes to extracellular noradrenaline and ATP using the Ca²⁺-sensitive fluorescent dye fura-2. In a standard bath solution the resting [Ca²⁺]_i averaged 57 ± 4 nm (n= 70 cells) and bath application of 1 μm noradrenaline evoked a sustained increase in [Ca²⁺]_i with a mean amplitude of 1533 ± 154 nm (n= 10; Fig. 2A). This [Ca²⁺]_i response to noradrenaline was concentration dependent and was saturated at 0.5 μm noradrenaline (data not shown), which is in good agreement with the concentration dependence of noradrenaline-induced [Ca²⁺]_i responses in neonatal rat brown adipocytes (Lee et al. 1993). Successive applications of 1 μm noradrenaline at 5–6 min intervals caused increases in [Ca²⁺]_i of a similar magnitude. However, the [Ca²⁺]_i response during the first exposure to noradrenaline was found to develop much more slowly than that during subsequent exposure to noradrenaline. The times required to reach a peak level of [Ca²⁺]_i after the first and second applications of 1 μm noradrenaline were 155.1 ± 12.8 and 30.3 ± 4.0 s (P < 0.0001, n= 10), respectively. The [Ca²⁺]_i response to noradrenaline was completely abolished by 1 μm phentolamine (Fig. 2B), thus indicating the involvement of an α-adrenoceptor in [Ca²⁺]_i responses.

Figure 2. Comparison of [Ca²⁺]_i increases in response to noradrenaline and ATP.

Chart record of [Ca²⁺]_i responses during exposure to 1 μm noradrenaline (NA) and 10 μm ATP, indicated by the horizontal bars below the records. Here and in subsequent figures except Fig. 3, left scales give the fluorescence ratio and right scales the corresponding [Ca²⁺]_i. A, [Ca²⁺]_i responses during four successive applications of noradrenaline. B, inhibition of noradrenaline-induced [Ca²⁺]_i increase by the α-adrenoceptor blocker phentolamine (1 μm, PHT). C, [Ca²⁺]_i response in a cell initially exposed to ATP and then to noradrenaline, followed by a second application of ATP. D, [Ca²⁺]_i responses during two successive applications of ATP at 5 min intervals in a cell previously exposed to noradrenaline. Each trace is obtained from different cells and is representative of 6–8 cells. Note the different time scale in panel B.

Bath application of 10 μm ATP induced only a small increase in [Ca²⁺]_i with a mean amplitude of 53 ± 2 nm (n= 8), while application of ATP at the same concentration to cells previously stimulated with noradrenaline evoked a rapid increase in [Ca²⁺]_i of 1057 ± 114 nm (n= 10; Fig. 2C). The time-to-peak [Ca²⁺]_i in response to 10 μm ATP averaged 9.7 ± 1.9 s (n= 10). It should be noted that the ATP-evoked increase in [Ca²⁺]_i subsided monotonically towards the resting levels within approximately 2 min, even in the continued presence of the agonist. Extracellular ATP at concentrations of 10 μm caused such a transient [Ca²⁺]_i response in almost all (209/212) cells previously stimulated with 1 μm noradrenaline. Repeated applications of equimolar ATP with short wash-out periods were consistently accompanied by the diminished [Ca²⁺]_i responses. As demonstrated in Fig. 2D, a second application of 10 μm ATP after a 5 min interval evoked only about 60 % of the [Ca²⁺]_i response observed during the first application of ATP. When the interval between ATP applications was less than 2 min, the [Ca²⁺]_i response during the second application of 10 μm ATP became negligibly small (data not shown). The ATP-evoked [Ca²⁺]_i responses thus appear to be subject to some desensitization process, as previously reported for P2 purinoceptor-mediated [Ca²⁺]_i increases in various cell types (Nobles et al. 1995; Bouyer et al. 1998).

The concentration-response relationship for the ATP-induced [Ca²⁺]_i increase is illustrated in Fig. 3A. The [Ca²⁺]_i response was evoked by extracellular ATP at concentrations of more than 0.1 μm and the maximal response was obtained with 10 μm ATP. The [Ca²⁺]_i response was somewhat reduced when the concentration of ATP was further increased to 100 μm.

Figure 3. Properties of the ATP-induced [Ca²⁺]_i response.

The [Ca²⁺]_i response to extracellular ATP was measured in cells previously stimulated with noradrenaline (1 μm). A, concentration-response relationship for the increase in [Ca²⁺]_i induced by extracellular ATP. The amplitude of the [Ca²⁺]_i increase was calculated by subtracting the resting [Ca²⁺]_i from the peak [Ca²⁺]_i during exposure to ATP in each cell. Only one concentration of ATP was tested in a given cell to exclude the diminished response to multiple applications of ATP (cf. Fig. 2D). B, concentration-response relationship for the inhibition of the ATP-induced increase in [Ca²⁺]_i by suramin. Cells were pretreated with suramin for 2 min and were then exposed to 10 μm ATP in the presence of suramin. Peak amplitudes of the ATP (10 μm)-induced [Ca²⁺]_i increase are plotted against concentrations of suramin. Data in panels A and B represent the means ±s.e.m. of 8 cells.

A P2 purinoceptor antagonist suramin (for reviews see Burnstock, 1996; North & Barnard, 1997) was found to suppress the ATP-induced [Ca²⁺]_i increase in a concentration-dependent manner (Fig. 3B). The pretreatment of cells with suramin at concentrations of 500 μm or more nearly completely inhibited the peak amplitudes of the [Ca²⁺]_i response evoked by 10 μm ATP. This result suggests that extracellular ATP evokes a [Ca²⁺]_i response in these brown adipocytes by binding to a P2 purinoceptor, consistent with a previous report on neonatal brown adipocytes (Lee & Pappone, 1997).

In order to elucidate whether the flux of extracellular Ca²⁺ into the cells is involved in agonist-induced [Ca²⁺]_i increases, the effect of extracellular Ca²⁺ removal was examined in the experiments shown in Fig. 4. In the absence of extracellular Ca²⁺, bath application of 1 μm noradrenaline evoked a small transient increase in [Ca²⁺]_i (224 ± 25 nm, n= 8; Fig. 4A), thus suggesting that the release of Ca²⁺ from the intracellular stores contributes at least partly to the initial rising phase of the [Ca²⁺]_i response. The stimulation of α-adrenoceptor in brown adipocytes was demonstrated to induce an increase in InsP₃ formation through the activation of phospholipase C (Nånberg & Putney, 1986; Schimmel et al. 1986). The release of Ca²⁺ from InsP₃-sensitive stores such as the endoplasmic reticulum is therefore most likely to underlie noradrenaline-induced increases in [Ca²⁺]_i recorded in the absence of extracellular Ca²⁺ (Fig. 4A), as has been previously suggested by Lee et al. (1993).

When extracellular Ca²⁺ was removed during a sustained elevation in [Ca²⁺]_i in response to 1 μm noradrenaline, the [Ca²⁺]_i declined to near the resting levels and this effect was reversible upon returning to the standard bath solution (Fig. 4B), thus indicating that the sustained elevation in [Ca²⁺]_i during exposure to noradrenaline is produced primarily by an influx of Ca²⁺ from the extracellular solution. However, this noradrenaline-evoked sustained increase in [Ca²⁺]_i was not appreciably affected by the L-type Ca²⁺ channel blocker nifedipine (1 μm), the T-type Ca²⁺ channel blocker flunarizine (1 μm), the P/Q-type Ca²⁺ channel blocker ω-conotoxin MVIIC (1 μm) and the N-type Ca²⁺ channel blocker ω-conotoxin MVIIA (1 μm) (data not shown), thus suggesting that the noradrenaline-induced Ca²⁺ entry which underlies the sustained increase in [Ca²⁺]_i is not mediated through the activation of these voltage-dependent Ca²⁺ channels.

In contrast, in the absence of extracellular Ca²⁺, bath application of 10 μm ATP consistently evoked a rapid (time to peak, 9.2 ± 3.3 s; n= 8) increase in [Ca²⁺]_i with a mean amplitude of 1004 ± 92 nm (n= 8) in cells previously stimulated with noradrenaline in the standard bath solution (Fig. 4C). The time course and amplitude of the ATP-induced [Ca²⁺]_i increase in the absence of extracellular Ca²⁺ are similar to those (time to peak, 9.7 ± 1.9 s; amplitude, 1057 ± 114 nm; n= 10) of the ATP-induced [Ca²⁺]_i increase recorded in the presence of standard bath solution ([Ca²⁺]_o= 2 mm; cf. Fig. 1C and D), indicating that the [Ca²⁺]_i increase evoked by ATP is primarily due to a mobilization of Ca²⁺ from the intracellular stores.

Interactions of P2 purinergic agonists and noradrenaline in the regulation of [Ca²⁺]_i

To further characterize the effects of ATP on changes in [Ca²⁺]_i, the interactions of ATP and noradrenaline in the regulation of [Ca²⁺]_i were examined. In the experiments shown in Fig. 5A, the cell was previously stimulated with 1 μm noradrenaline, and 10 μm ATP was subsequently added to the bath solution about 2 min after starting a second application of noradrenaline. Extracellular ATP was found to greatly diminish the sustained levels of noradrenaline-induced [Ca²⁺]_i increase without producing any additional increase in [Ca²⁺]_i. In a total of eight cells, 10 μm ATP reduced the sustained levels of noradrenaline (1 μm)-induced [Ca²⁺]_i increases by 92 ± 2 %. This ATP-induced depression consistently persisted after withdrawal of the ATP, and a subsequent application of ATP after a washout period of about 5 min did not have any appreciable effect on [Ca²⁺]_i.

Figure 5. Inhibitory effect of extracellular ATP on the noradrenaline-evoked [Ca²⁺]_i increase.

Periods of exposure to 1 μm noradrenaline (NA) and 10 μm ATP are indicated by horizontal bars below chart records of [Ca²⁺]_i. A, ATP was added to the bath twice at an interval of about 5 min in the presence of noradrenaline, as indicated. The initial application of ATP greatly depressed the noradrenaline-induced increase in [Ca²⁺]_i, the depression persisting after the withdrawal of ATP, and the second application was associated with a negligibly small change in [Ca²⁺]_i. B, effects of pretreatment by ATP on the noradrenaline-induced [Ca²⁺]_i increase. Application of noradrenaline in the presence of ATP was associated with a transient increase in [Ca²⁺]_i. Each trace is obtained from different cells and is representative of 6–8 cells.

In the presence of 10 μm ATP, bath application of 1 μm noradrenaline only evoked a transient increase in [Ca²⁺]_i of 254 ± 33 nm, (n= 8; Fig. 5B), which resembles in both magnitude and time course the [Ca²⁺]_i response to noradrenaline recorded in the absence of extracellular Ca²⁺ (cf. Fig. 4A). Assuming that the sustained increase in [Ca²⁺]_i during exposure to noradrenaline is largely due to noradrenaline-stimulated Ca²⁺ influx across the plasma membrane, it is reasonable to propose that extracellular ATP blocks noradrenaline-stimulated Ca²⁺ influx. Extracellular ATP was thus found not only to mobilize Ca²⁺ from intracellular stores (Fig. 4C) but also to exert an inhibitory effect on the process of Ca²⁺ influx mediating the sustained increase in [Ca²⁺]_i in response to noradrenaline (Fig. 5).

We next examined the effect of other P2 purinergic agonists on both resting [Ca²⁺]_i and the noradrenaline-evoked [Ca²⁺]_i elevation. Bath application of 10 μm ADP induced a sustained increase in [Ca²⁺]_i of 734 ± 74 nm (n= 6) in cells previously stimulated with 1 μm noradrenaline (Fig. 6A). In contrast to the effects of ATP, extracellular ADP did not exert any inhibitory effect on the noradrenaline-evoked sustained increase in [Ca²⁺]_i (Fig. 6B, cf. Fig. 5A). As demonstrated in Fig. 6C, the external application of 10 μm UTP produced a sustained [Ca²⁺]_i increase of 485 ± 44 nm (n= 6). Figure 6D illustrates the effects of 10 μm UTP on the noradrenaline-evoked [Ca²⁺]_i increase. We could not detect any depression by UTP at concentrations of up to 100 μm of the noradrenaline-evoked [Ca²⁺]_i increase. Bath application of AMP at concentrations of up to 100 μm did not elicit any effect on resting [Ca²⁺]_i or the noradrenaline-stimulated [Ca²⁺]_i elevation (data not shown).

Figure 6. Effects of ADP and UTP on resting [Ca²⁺]_i and the noradrenaline-evoked [Ca²⁺]_i increase.

Periods of exposure to 1 μm noradrenaline (NA), 10 μm ADP and 10 μm UTP are indicated by horizontal bars below chart records of [Ca²⁺]_i. A, a single cell was successively exposed to noradrenaline and ADP, as indicated. B, effect of addition of ADP on noradrenaline-induced increase in [Ca²⁺]_i. C, a cell was initially exposed to noradrenaline and then to UTP. D, effects of application of UTP on the noradrenaline-induced increase in [Ca²⁺]_i. Increases in [Ca²⁺]_i evoked by ADP and UTP were not accompanied by a monotonic decline during the continued presence of the agonists. Neither ADP nor UTP exerted an inhibitory effect on the noradrenaline-induced increases in [Ca²⁺]_i. Each trace is obtained from different cells and is representative of 6–8 cells.

Presence of store-operated Ca²⁺ entry in rat brown adipocytes

It has been reported in a number of electrically non-excitable cells that the depletion of intracellular Ca²⁺ stores triggers Ca²⁺ entry across the plasma membrane (Putney, 1986; for reviews see Putney, 1990; Berridge, 1995; Parekh & Penner, 1997). In order to determine whether the store-operated Ca²⁺ entry mechanism operates in rat brown adipocytes, we examined the effect on [Ca²⁺]_i of thapsigargin, an agent which gradually depletes endoplasmic reticulum Ca²⁺ stores by specifically and irreversibly inhibiting Ca²⁺ uptake through endoplasmic reticulum Ca²⁺ pumps (Thastrup et al. 1990; Premack et al. 1994). In the presence of extracellular Ca²⁺, bath application of 100 nm thapsigargin consistently caused a gradual increase in [Ca²⁺]_i of 1552 ± 141 nm (n= 8) in cells previously stimulated with 1 μm noradrenaline (Fig. 7A). This [Ca²⁺]_i response to thapsigargin was reversibly abolished by removing extracellular Ca²⁺ and persisted even after withdrawal of the drug in the presence of extracellular Ca²⁺. When 100 nm thapsigargin was applied to the cell in the absence of extracellular Ca²⁺, the cell responded with a small elevation in [Ca²⁺]_i of 84 ± 2 nm (n= 4), which presumably reflects the small amount of releasable Ca²⁺ stored in the endoplasmic reticulum in these brown adipocytes. However, once extracellular Ca²⁺ was restored, a large and irreversible elevation in [Ca²⁺]_i was evoked, even after withdrawal of the thapsigargin (Fig. 7B). A similar [Ca²⁺]_i response to thapsigargin was observed in cells which were not previously exposed to noradrenaline (data not shown). These rat brown adipocytes were thus found to possess store-operated Ca²⁺ entry mechanisms.

Noradrenaline has been suggested to activate InsP₃-dependent mobilization of intracellular Ca²⁺ through stimulation of α-adrenoceptors in brown adipocytes (Nånberg & Putney, 1986; Schimmel et al. 1986; Lee et al. 1993). The question then arises as to whether the resultant emptying of the endoplasmic reticulum Ca²⁺ stores is followed by the activation of store-operated Ca²⁺ entry in association with a noradrenaline-induced sustained increase in [Ca²⁺]_i which is entirely dependent upon extracellular Ca²⁺ (cf. Fig. 4B). In order to address this issue, we tested whether the stimulatory effect of noradrenaline on [Ca²⁺]_i was additive to that of thapsigargin. In the experiments shown in Fig. 7C, the cell was initially exposed to 100 nm thapsigargin, which resulted in a sustained elevation in [Ca²⁺]_i through the activation of store-operated Ca²⁺ entry. After the [Ca²⁺]_i response to thapsigargin reached a steady state, noradrenaline at a maximally effective concentration (1 μm) was then added. In all eight cells examined subsequent addition of 1 μm noradrenaline in the presence of thapsigargin failed to cause any additional increases in [Ca²⁺]_i, thus suggesting that the Ca²⁺ influx pathway activated in response to noradrenaline is the same as that activated by thapsigargin, i.e. store-operated Ca²⁺ entry.

Inhibition of the thapsigargin-induced [Ca²⁺]_i increase by ATP and ATPγS

We then investigated the effects of extracellular ATP on [Ca²⁺]_i elevation evoked by activation of store-operated Ca²⁺ entry (Fig. 8). When the gradual [Ca²⁺]_i elevation in response to 100 nm thapsigargin reached a steady state, 10 μm ATP was added to the bath solution. In a total of eight cells, 10 μm ATP reduced the thapsigargin (100 nm)-induced [Ca²⁺]_i elevation by 92 ± 3 % (upper panel), thus strongly suggesting that an influx of Ca²⁺ through the store-operated pathway is potently inhibited by extracellular ATP. It should be noted that the inhibition persisted for a relatively long period (≥ 10 min) after withdrawal of the ATP. Addition of the Ca²⁺ ionophore ionomycin (10 μm) to the bath solution after ATP application consistently induced an abrupt increase in [Ca²⁺]_i (n= 4 cells), confirming that the dye was functioning normally and that the cells were viable. In addition, fluorescence signals following excitation at 340 and 380 nm (F₃₄₀ and F₃₈₀, respectively) were found to shift qualitatively as if they were mirror images during the course of the experiment (middle and lower panels), confirming again that [Ca²⁺]_i was successfully recorded from intact cells. Therefore, the irreversible nature of the ATP action can be regarded as a genuine cellular response.

Figure 8. Inhibition of the thapsigargin-induced [Ca²⁺]_i elevation by extracellular ATP.

Periods of exposure to 1 μm noradrenaline (NA), 100 nm thapsigargin (TG), 10 μm ATP and 10 μm ionomycin (IO) are indicated by horizontal bars below the chart record of [Ca²⁺]_i (upper panel). Fluorescence intensities following excitation at 340 and 380 nm are shown in the middle and lower panels, respectively. The vertical bar in the lower panel represents 50 arbitrary fluorescence intensity units and refers also to the middle panel. The trace is representative of 6 cells.

We tested whether the synthetic ATP analogues ATPγS and α,β-methylene ATP can mimic the inhibitory action of ATP on thapsigargin-induced [Ca²⁺]_i elevation. Bath application of 10 μm ATPγS not only induced a transient increase in [Ca²⁺]_i of 632 ± 79 nm (n= 6 cells) in cells previously stimulated with noradrenaline, but also reduced the thapsigargin-induced [Ca²⁺]_i increase by 85 ± 2 % (Fig. 9A). This inhibitory effect of ATPγS also persisted for a period of at least 10 min after withdrawal of the ATPγS. In contrast, 10 μmα,β-methylene ATP neither elevated [Ca²⁺]_i nor decreased thapsigargin-induced [Ca²⁺]_i elevation (Fig. 9B).

Figure 9. Effects of ATPγS and α,β-methylene ATP on resting [Ca²⁺]_i and the thapsigargin-evoked [Ca²⁺]_i increase.

Periods of exposure to 1 μm noradrenaline (NA), 100 nm thapsigargin (TG), 10 μm ATPγS, and 10 μmα,β-methylene ATP (αβ-meATP) are indicated by horizontal bars below chart records of [Ca²⁺]_i. The effects of ATPγS (A) and α,β-methylene ATP (B) on resting [Ca²⁺]_i and the thapsigargin-induced [Ca²⁺]_i increase were examined, as indicated. Each trace was obtained from different cells and is representative of 6 cells.

In neonatal rat brown adipocytes extracellular ATP was shown to activate a non-selective cation conductance which would depolarize the cell membrane to some extent (Lee & Pappone, 1997). Membrane depolarization attenuates the magnitude of Ca²⁺ entry through the store-operated pathway due to a reduced driving force for Ca²⁺ (Hoth & Penner, 1992). It is therefore probable that the inhibitory effect of ATP occurred secondarily to membrane depolarization. To test this possibility, we investigated the effect of ATP on thapsigargin-induced [Ca²⁺]_i elevation in the presence of the K⁺ ionophore valinomycin (1 μm) which is expected to clamp the cell membrane near the reversal potential of K⁺. In the presence of valinomycin, 10 μm ATP still decreased thapsigargin-induced [Ca²⁺]_i elevation by 93 ± 2 % (n= 6 cells; Fig. 10), which is similar to the decrease observed in the absence of valinomycin (92 ± 3 %, n= 8 cells; cf. Fig. 8). Thus, the inhibitory effect of ATP was not attenuated by stabilization of cell membrane potential with valinomycin, suggesting that the possible depolarization of cell membrane is not involved.

Figure 10. Effect of valinomycin on the inhibitory action of ATP on the thapsigargin-induced [Ca²⁺]_i increase.

Periods of exposure to 1 μm noradrenaline (NA), 100 nm thapsigargin (TG), 10 μm ATP, 1 μm valinomycin are indicated by horizontal bars below the chart records of [Ca²⁺]_i. The effect of ATP on the thapsigargin-induced [Ca²⁺]_i increase was examined in the presence of valinomycin, as indicated. The trace is representative of 6 cells.

In a number of cell types the activation of protein kinase C (PKC) has been shown to inhibit Ca²⁺ influx through the store-operated pathway (e.g. HL-60 promyelocytes: Montero et al. 1993; Song et al. 1998; rat basophilic leukaemia cells: Parekh & Penner, 1995; rat mesangial cells: Mene et al. 1997; guinea-pig enteric glial cells: Sarosi et al. 1998). In order to address the question of whether PKC was involved in the inhibitory effect of ATP on the thapsigargin-induced [Ca²⁺]_i increase, we examined whether PMA, a direct activator of PKC, could mimic the action of ATP. Bath application of 5 μm PMA did not produce any appreciable effects on the thapsigargin-induced [Ca²⁺]_i increase in all six cells examined (Fig. 11), thus suggesting that an inhibitory action of ATP is not mediated by activation of PKC in rat brown adipocytes.

Figure 11. Effect of PMA on thapsigargin-evoked [Ca²⁺]_i increase.

Periods of exposure to 1 μm noradrenaline (NA), 100 nm thapsigargin (TG) and 5 μm PMA are indicated by horizontal bars below chart records of [Ca²⁺]_i. The trace is representative of 6 cells.

DISCUSSION

In the present study we characterized the [Ca²⁺]_i responses to extracellular ATP in adult rat brown adipocytes. External application of ATP in micromolar concentrations evokes a large and transient increase in [Ca²⁺]_i in cells previously stimulated with noradrenaline (Fig. 2). In addition, extracellular ATP exerts a potent inhibitory effect on both noradrenaline- and thapsigargin-induced [Ca²⁺]_i increases (Figs 5 and 8), thus clearly showing that extracellular ATP exerts a dual effect on [Ca²⁺]_i in these brown adipocytes.

Properties of the stimulatory effects of ATP on [Ca²⁺]_i response

The [Ca²⁺]_i response of brown adipocytes to extracellular ATP was first analysed in neonatal rats by Lee & Pappone (1997). These investigators demonstrated that bath application of ATP in micromolar concentrations evokes an increase in [Ca²⁺]_i by activating both Ca²⁺ influx and Ca²⁺ release from intracellular stores. Our results in adult rat brown adipocytes, however, show that the removal of extracellular Ca²⁺ did not produce any appreciable effect on the ATP-induced increase in [Ca²⁺]_i (Fig. 4C), which indicates that ATP exerts a stimulatory effect on [Ca²⁺]_i mainly through mobilization of Ca²⁺ from the intracellular Ca²⁺ stores. The observation that the initial application of ATP is accompanied by a small increase in [Ca²⁺]_i (Fig. 2C) may reflect the fact that the intracellular stores contain a low amount of releasable Ca²⁺ in these adult rat brown adipocytes under unstimulated basal conditions.

The suramin sensitivity of the ATP-induced [Ca²⁺]_i increase (Fig. 3B) suggests the involvement of a P2 purinoceptor. In recent years P2 purinoceptors have been classified into two main subfamilies, namely, P2X purinoceptors comprising ligand-gated cation channels (an ionotropic receptor; Valera et al. 1994; Brake et al. 1994) and P2Y purinoceptors comprising G protein-coupled receptors (a metabotropic receptor; Lustig et al. 1993; Webb et al. 1993; for reviews see Burnstock, 1996; Burnstock & King, 1996; North & Barnard, 1997). The stimulation of both P2X and P2Y purinoceptors in various cell types has been shown to lead to an increase in [Ca²⁺]_i through distinct intracellular mechanisms. An elevation in [Ca²⁺]_i associated with P2X purinoceptor stimulation is elicited by the influx of Ca²⁺ through the ATP-gated non-selective cation channel (Michel et al. 1996; Troadec et al. 1998) and/or via the voltage-gated Ca²⁺ channel which is secondarily activated by membrane depolarization due to the influx of Na⁺ through the ATP-gated channel (Hirano et al. 1991). Extracellular Ca²⁺ is therefore essential for the [Ca²⁺]_i increases via P2X purinoceptors (for review see Harden et al. 1995).

On the other hand, the P2Y purinoceptor-mediated [Ca²⁺]_i increase in various cell types has been shown to be primarily evoked by the release of Ca²⁺ from InsP₃-sensitive internal stores, i.e. the endoplasmic reticulum; thus this response is not obligatorily dependent upon the presence of extracellular Ca²⁺ (Nobles et al. 1995; Bouyer et al. 1998; Hartmann & Verkhratsky, 1998). Judging from the observation that the removal of extracellular Ca²⁺ was without effect (Fig. 4C), extracellular ATP is most likely to evoke a transient [Ca²⁺]_i increase through a P2Y purinoceptor rather than via a P2X purinoceptor in this preparation. The finding that micromolar levels of UTP increased [Ca²⁺]_i with a high potency (Fig. 6C) further supports the involvement of a P2Y purinoceptor in this [Ca²⁺]_i response, since none of P2X purinoceptor subtypes (P2X₁-P2X₇) was effectively activated by UTP (for review see Burnstock & King, 1996). The post-receptor process mediating the [Ca²⁺]_i increase in response to extracellular ATP is presently unclear; however, most of metabotropic P2Y purinoceptors have been reported to be coupled to phospholipase Cβ through a GTP binding protein, resulting in activation of InsP₃-mediated mobilization of intracellular Ca²⁺ (for review see Burnstock & King, 1996). Further experiments are called for to explore the possible involvement of the InsP₃-dependent process in ATP-induced [Ca²⁺]_i elevation in brown adipocytes.

Presence of store-operated Ca²⁺ entry and its inhibition by extracellular ATP

It has been demonstrated in a variety of electrically non-excitable cells that InsP₃-mediated Ca²⁺ release and the resultant depletion of endoplasmic reticulum Ca²⁺ stores trigger Ca²⁺ entry across the plasma membrane through a store-operated pathway (Putney, 1986; for reviews see Putney, 1990; Berridge, 1995; Parekh & Penner, 1997). This store-operated Ca²⁺ entry mechanism can also be pharmacologically activated by thapsigargin, a highly specific and irreversible inhibitor of endoplasmic reticulum Ca²⁺-ATPase (Thastrup et al. 1990; Premack et al. 1994). The present experiments clearly demonstrated that the application of thapsigargin consistently evoked an irreversible increase in [Ca²⁺]_i that was entirely dependent upon extracellular Ca²⁺ (Fig. 7A and B), which strongly indicates the presence of store-operated Ca²⁺ entry in these brown adipocytes.

The stimulation of α₁-adrenoceptors with noradrenaline has been shown to cause a sustained increase in [Ca²⁺]_i in brown adipocytes of hamster (Wilcke & Nedergaard, 1989) and neonatal rat (Lee et al. 1993). This α₁-adrenoceptor-mediated [Ca²⁺]_i increase was clearly demonstrated to arise from two distinct pathways, i.e. an initial mobilization of intracellular Ca²⁺ followed by a sustained Ca²⁺ influx process (Lee et al. 1993). We confirmed in adult rat brown adipocytes that a similar consecutive activation of Ca²⁺ release and Ca²⁺ influx underlies an increase in [Ca²⁺]_i in the presence of noradrenaline (Fig. 4A and B). The present experiments also show that the stimulatory effect of noradrenaline on [Ca²⁺]_i is not additive to that of thapsigargin (Fig. 7C), which suggests that the Ca²⁺ influx process associated with noradrenaline stimulation is evoked through the store-operated pathway. Taken together, in rat brown adipocytes, noradrenaline appears to increase [Ca²⁺]_i by initially mobilizing internal Ca²⁺ through an InsP₃-dependent pathway, followed by the activation of Ca²⁺ influx through the store-operated pathway triggered by the resultant depletion of endoplasmic reticulum Ca²⁺ stores. The slow onset of the [Ca²⁺]_i increase during the first exposure to noradrenaline (Fig. 2A) is, again, thought to be related to the nature of intracellular Ca²⁺ stores containing a small amount of releasable Ca²⁺ under unstimulated conditions.

Micromolar levels of extracellular ATP nearly completely reduced the sustained increase in [Ca²⁺]_i evoked by thapsigargin (Fig. 8), thus strongly suggesting that in rat brown adipocytes store-operated Ca²⁺ entry is potently inhibited by extracellular ATP. This inhibitory action of ATP was not affected by the K⁺ ionophore valinomycin (Fig. 10), ruling out the possible involvement of membrane potential changes. Stimulation of PKC with PMA did not mimic the inhibitory effect of ATP (Fig. 11), indicating that the inhibitory effect is not mediated by activation of PKC. Extracellular ATP at 10 μm reduced the sustained increases in [Ca²⁺]_i evoked by noradrenaline and thapsigargin to a similar extent (92 ± 2 % reduction with noradrenaline, n= 8 cells; 92 ± 3 % reduction with thapsigargin, n= 8 cells). In addition, ATP irreversibly depressed both the noradrenaline- (Fig. 5A) and thapsigargin-induced [Ca²⁺]_i increases (Fig. 8). These similarities in the sensitivity of depression of [Ca²⁺]_i increases to extracellular ATP are also consistent with the activation of store-operated Ca²⁺ entry in the presence of noradrenaline. The inhibitory effect of ATP on the noradrenaline- or thapsigargin-induced increase in [Ca²⁺]_i was mimicked by ATPγS (Fig. 9A) but not by ADP, UTP or α,β-methylene ATP (Figs 6B and D, and 9B). It still remains unclear whether extracellular ATP inhibits store-operated Ca²⁺ entry by binding to a P2 purinoceptor in brown adipocytes; however, none of the P2X and P2Y purinoceptors identified to date exhibit such a high agonist selectivity for ATP and ATPγS (for reviews see Burnstock & King, 1996; North & Barnard, 1997).

The ATP-induced [Ca²⁺]_i elevation is consistently accompanied by a monotonic decline to near resting levels, even in the continued presence of ATP, which forms a striking contrast to the noradrenaline-induced [Ca²⁺]_i increase, which remains stable near peak levels during the entire period of agonist application (Fig. 2). In several cell types, extracellular ATP has been shown to induce InsP₃-dependent Ca²⁺ release followed by the activation of store-operated Ca²⁺ entry, by stimulating metabotropic purinoceptors (Munshi et al. 1993; Somasundaram & Mahaut-Smith, 1994; Chiono et al. 1995; Kerst et al. 1995; Schöfl et al. 1995a,b; Madge et al. 1997; Hartmann & Verkhratsky, 1998). Assuming that these two cellular processes are consecutively activated during exposure to ATP in brown adipocytes, the subsequently activated store-operated Ca²⁺ entry should be almost completely inhibited by the presence of extracellular ATP; as a result, only cytosolic Ca²⁺ initially released from the endoplasmic reticulum can contribute to the changes in [Ca²⁺]_i, and this phenomenon may account for the characteristic transient form of the ATP-induced [Ca²⁺]_i elevation.