INTRODUCTION

In human lymphocytes, as well as in other electrically non-excitable cells, receptor stimulation results in a transient increase in the cytosolic free calcium concentration ([Ca²⁺]_i). [Ca²⁺]_i increase primarily occurs due to the calcium release from inositol 1,4,5-trisphosphate-sensitive intracellular pools and is generally accompanied by an entry of extracellular calcium mediated by the activation of receptor-operated or second messenger-activated Ca²⁺ channels (1). In the hypothesis referred to as capacitative model, calcium influx is controlled by the filling state of intracellular calcium pools (2-4). The mechanisms by which capacitative calcium influx is elicited are not clear. Depletion of cellular calcium pools may trigger release of a diffusible cytoplasmic messenger which in turn opens transmembrane Ca²⁺ channels (5, 6).

One approach to study the role of intracellular Ca²⁺ pools for the Ca²⁺ entry utilizes inhibitors of the Ca²⁺-ATPases in the store membrane. Thapsigargin has been shown to selectively inhibit Ca²⁺-ATPases in the endoplasmic reticulum without affecting Ca²⁺-ATPases in the plasma membrane (7, 8). In contrast to receptor agonists that initiate rapid formation of inositol 1,4,5-trisphosphate, thapsigargin depletes Ca²⁺ pools solely by preventing Ca²⁺ reuptake (9). In a variety of cells including human lymphocytes the thapsigargin-mediated depletion of calcium stores leads to a sustained elevation of [Ca²⁺]_i supported by the entry of extracellular Ca²⁺ (10-15).

Although cellular phospholipases are crucial for the regulation of numerous physiological processes, their role in maintaining intracellular calcium homeostasis is little understood. The early phase of Ca²⁺ increase due to Ca²⁺ release from inositol 1,4,5-trisphosphate-sensitive stores is thought to involve activation of the receptor-coupled, phosphoinositide-specific phospholipase C. The role of this enzyme in producing the sustained phase of Ca²⁺ increase which is caused by the transmembrane Ca²⁺ influx is, however, less clear (16-18). There is also evidence pointing to the possible role of phospholipase A₂ in modulating Ca²⁺ entry following depletion of intracellular Ca²⁺ stores (19, 20). To our knowledge, no data exist on the role of phosphatidylcholine-specific phospholipases in the regulation of intracellular Ca²⁺ homeostasis. Therefore, the aim of the present study was to investigate the involvement of phosphatidylcholine-specific phospholipase C (PC-PLC)¹ and phosphatidylcholine-specific phospholipase D (PC-PLD) in the regulation of the calcium entry following depletion of Ca²⁺ stores with thapsigargin. Our results demonstrate that PC-PLC, but not PC-PLD, plays an important role in the down-regulation of Ca²⁺ influx in human lymphocytes.

EXPERIMENTAL PROCEDURES

Tricyclodecan-9-yl xantogenate (D609), Fura-2-AM, propranolol, and phospholipid standards were from Sigma, Deisenhofen, Germany. {3-[1-[3-(Amidinothio)propyl-1H-indoyl-3-yl]-3-(1-methyl-1H-indoyl-3-yl)maleimide methane sulfonate} (Ro-31-8220), 1-(6((17-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl)-1H-pyrrole-2,5-dione (U73122), phorbol myristate acetate (PMA), and thapsigargin were purchased from Calbiochem, Bad Soden, Germany. 4-PMA was obtained from Biomol, Hamburg, Germany. [1-¹⁴C]Arachidonic acid and L-lyso-3-phosphatidylcholine [1-¹⁴C]palmitoyl were from Amersham, Braunschweig, Germany, and [³²P]orthophosphoric acid was from NEN Life Science Products, Dreieich, Germany. Phospholipase C (from Bacillus cereus) and phospholipase D (from Streptomyces chromofuscus) were purchased from Boehringer, Mannheim, Germany. Silica Gel 60 plates and solvents for thin layer chromatography were obtained from Merck, Darmstadt, Germany. Ficoll (Lymphoprep) was obtained from Nycomed, Uppsalla, Sweden. Autoradiography was performed with Kodak X-Omat film (Eastman Kodak). Ultima GOLD scintillation mixture was provided by Packard, Frankfurt, Germany.

Lymphocytes were obtained form heparinized blood of 12 different donors according to previously described methods (21). Briefly, blood was centrifuged at 240 × g for 15 min, and the upper two-thirds of the supernatant were aspirated. The remaining blood was mixed 1:1 with Hanks' balanced salt solution containing 136 mM NaCl, 5.4 mM KCl, 0.44 mM KH₂PO₄, 0.34 mM Na₂HPO₄, 1.0 mM CaCl₂, 5.6 mM D-glucose, and 10 mM HEPES, pH 7.4. Lymphocytes were prepared after centrifugation of blood on a Ficoll gradient (Ficoll 5.6% (w/v), density 1.077 g/ml). The lymphocyte interphase was carefully aspirated, washed three times (400 × g for 5 min), and resuspended in Hanks' balanced salt solution. The lymphocyte viability was greater than 95% as determined by the trypan blue exclusion test.

Intracellular calcium measurements were performed according to the established method (22). Lymphocytes were loaded with 2.5 µM Fura-2 AM for 45 min at 37 °C. The lymphocyte suspension was then washed twice (240 × g, 15 min) to remove unincorporated dye and adjusted to a final concentration of 1 × 10⁶/ml. Fluorescence was recorded at 510 nm (bandwidth 10 nm) with excitation wavelengths of 340 and 380 nm (bandwidth 10 nm) using a fluorescence spectrophotometer (model 2000, Hitachi Ltd, Tokyo, Japan). The intracellular calcium concentration was calculated as described previously (22). Briefly, the maximum fluorescence was obtained after addition of 1.0 mM digitonin. The minimum fluorescence was obtained after addition of 5 mM EGTA. The ratio out of the measured fluorescence values was calculated. [Ca²⁺]]_i was obtained according to Equation 1 by Grynkiewicz et al. (23).

(Eq. 1)

where R_min stands for the ratio in calcium-free solution, R_max for the ratio at calcium saturation, and K* for K_d × F_min2/F_max2, the latter representing the fluorescence maximum and minimum at 380 excitation. K_d of Fura-2 was set to be 224 nmol/liter.

Phospholipid analysis was done as described previously (24, 25). For metabolic lipid radiolabeling the lymphocyte suspension was incubated at 37 °C for 90 min with the following compounds: 0.2 µCi/ml [1-¹⁴C]arachidonic acid, 0.2 µCi/ml [1-¹⁴C]lyso-3-phosphatidylcholine, or 100 µCi/ml [³²P]orthophosphoric acid. In the latter case, phosphate was replaced by carbonate in the incubation buffer. The lymphocytes were then washed, adjusted to a final concentration of 1 × 10⁶ cells/ml, and stimulated with the desired agonist. At different time points 0.5-ml aliquots were withdrawn and added to 1.5 ml of ice-cold chloroform/methanol (2:1, v/v) or chloroform/methanol/hydrochloric acid (2:1:0.01, v/v) in the case of ³²P labeling. The phases were split by adding 0.5 ml of chloroform and 0.5 ml of water. The samples were centrifuged at 3000 rpm for 5 min, and lipid phases were collected. Water-soluble phases were extracted once more with 1.5 ml of chloroform. Lipid phases were then combined, dried under nitrogen, dissolved in 0.3 ml of hexane, and stored at 70 °C until analyzed.

In most experiments, a double one-dimensional TLC as described by Gruchalla et al. (26) was used to separate phospholipids and neutral lipids of interest. In this approach, a series of samples was spotted 12 cm from the bottom of the plate. To resolve labeled neutral lipids (DAG, fatty acids, and triglycerides) from phospholipids that remained at the origin, the plates were twice developed in toluene/ether/ethanol/triethylamine (100:80:4:2, by volume). After the first run to 20 cm, the plates were thoroughly dried and developed a second time with the same solution to 16 cm. Plates were then cut 0.8 cm above the origin (i.e. 12.8 cm above the lower edge of the plate), rotated 180°, and developed to the top with chloroform/methanol/ammonium hydroxide (65:35:5, by volume). After they were dried thoroughly, autoradiography was performed with Kodak X-Omat film for 10-14 days. Radioactive bands were cut from the silica plates, placed in scintillation vials containing 10 ml of Ultima-Gold scintillation fluid, and quantitated by liquid scintillation counting in a 1214 scintillation counter (LKB, Bromma, Sweden). The identities of labeled bands were determined based on R_F values obtained for authentic neutral lipids and phospholipids visualized by iodine staining.

For determination of phosphatidylinositols [³²P]phospholipids were developed with chloroform/methanol/acetone/acetic acid/water (60:20:23:18:12, by volume) using potassium oxalate-impregnated Silica 60 plates. Bands corresponding to PtdIns, PtdInsP, PtdInsP₂, or PtdChol were cut out from the silica plates, placed in scintillation vials, and quantitated as described above. The identities of labeled bands were determined based on R_F values obtained for authentic neutral lipids and phospholipids.

PC-PLD hydrolyzes phospholipids to yield the free polar head groups choline and PtdOH. In the presence of primary alcohols, however, PC-PLD catalyzes a phosphatidyl transfer reaction producing phosphatidyl alcohols. Since the transphosphatidylation is catalyzed solely by PC-PLD, the production of phosphatidyl alcohols is an unequivocal marker for involvement of this enzyme. For the examination of PC-PLD activity lymphocytes were labeled with 0.2 µCi/ml [1-¹⁴C]lyso-3-phosphatidylcholine. Butan-1-ol was added 5 min prior to agonist addition. Radiolabeled phospholipids were extracted as described above and analyzed using Silica 60 plates developed with ethyl acetate/2,2,4-trimethylpentane/acetic acid (9:5:2 by volume) as described by van der Meulen and Haslam (27). In some experiments the amount of PtdBut formed was examined using the double one-dimensional system as described above. The optimal concentration of butanol was determined by performing a butanol concentration curve using 1 µM PMA as an agonist. As shown in Fig. 1 the most effective PtdBut production was seen at butanol concentrations between 0.3 and 0.6% (v/v) and was accompanied by a decrease in PtdOH formation.

[View Larger Version of this Image (18K GIF file)]

Each experiment was performed in duplicate and repeated 3-5 times, as indicated in appropriate legends. Unless otherwise indicated, data represent the mean from duplicate determination in a representative experiment. To avoid possible bias due to the variability of lymphocyte populations obtained from different donors, control and drug responses were determined in the same cell preparations. For calcium measurements data are presented as means ± S.E. The groups were compared with the non-parametric Wilcoxon-Mann-Withney test using the computer software Instat 2.02 (GraphPAD). Two-tailed p values less than 0.05 were considered to be significant.

RESULTS

[Ca²⁺]_i was measured in intact human lymphocytes using the calcium-sensitive fluorescent dye Fura-2. The resting [Ca²⁺]_i level in these cells averaged 94 ± 8 nM (n = 34). The resting [Ca²⁺]_i level did not significantly change at least over 400 s (n = 23). As illustrated in Fig. 2, addition of 5 µM thapsigargin resulted in a time-dependent increase in intracellular calcium. Within 200 s [Ca²⁺]_i increased by 228 ± 27 nM (n = 34) over the resting level. Next, the effects of the PC-PLC-inhibitor D609, the unspecific PLC inhibitor U73122, the indirect PLD inhibitor butanol, and phosphatidic acid phosphohydrolase inhibitor propranolol on the thapsigargin-triggered [Ca²⁺]_i elevation were tested. In the presence of 20 µM D609 or 10 µM U73122 the thapsigargin-induced [Ca²⁺]_i increases were significantly enhanced and averaged 377 ± 54 nM (n = 19; p < 0.02) and 517 ± 99 nM (n = 16; p < 0.01) over the resting level, respectively (Fig. 2). By contrast, preincubation of cells with 0.3% butanol or 50 µM propranolol did not affect thapsigargin-induced [Ca²⁺]_i elevation (Fig. 2). Under these experimental conditions [Ca²⁺]_i rose by 188 ± 24 nM (n = 10; not significant versus thapsigargin alone) and 171 ± 15 nM (n = 6; not significant versus thapsigargin alone). The potentiating effects of D609 and U73122 on the thapsigargin-induced [Ca²⁺]_i elevation were concentration-dependent (Fig. 2). For both compounds the significant enhancement of [Ca²⁺]_i increases was seen at a concentration of about 10 µM.

[View Larger Version of this Image (21K GIF file)]

To test whether potentiating effects of U73122 and D609 on thapsigargin-stimulated intracellular calcium elevation occur as a consequence of increased Ca²⁺ entry, the unidirectional uptake of Mn²⁺, a Ca²⁺ surrogate, was determined. Mn²⁺, which quenches Fura-2 fluorescence, enters cells through physiological pathways, yet is not readily extruded at an appreciable rate. Thus, the rate of fluorescence decrease provides a relative measure of the divalent ion entry. As illustrated in Fig. 3, addition of 5 µM thapsigargin was found to accelerate the rate of Fura-2 quenching, indicating activation of the direct cation permeability pathway at the plasma membrane. When lymphocytes were first treated for 5 min with 20 µM D609 or 10 µM U73122, and subsequently stimulated with thapsigargin, the rate of Fura-2 quenching was considerably more pronounced (Fig. 3). These findings suggest that U73122 and D609 exert their potentiating effects on the thapsigargin-induced [Ca²⁺]_i elevation by increasing trans-plasma membrane calcium influx.

[View Larger Version of this Image (17K GIF file)]

Since diacylglycerol is the principal product released upon activation of PLC, we next investigated whether it is formed following stimulation of lymphocytes with 5 µM thapsigargin. The lymphocytes were labeled for 2 h with 0.2 µCi/ml [1-¹⁴C]arachidonic acid. Upon this treatment phosphatidylinositol (PtdIns) and phosphatidylcholine (PtdChol) incorporated 42 and 34% of the total phospholipid radioactivity, respectively. As demonstrated in Fig. 4, the [¹⁴C]DAG level increased by 98 ± 15% (n = 5) within 120 s following addition of 10 µM thapsigargin. Thereafter, [¹⁴C]DAG remained elevated over the basal level for at least 300 s. Preincubation of lymphocytes with 20 µM D609 for 5 min reduced the thapsigargin-induced [¹⁴C]DAG accumulation to 48 ± 7% (n = 3) at 120 s after stimulation (Fig. 4). Similarly, the reduction of the thapsigargin-induced [¹⁴C]DAG formation to 37 ± 6% (n = 3) at 120 s after stimulation was observed in lymphocytes pretreated with 10 µM U73122 (Fig. 4). In contrast to D609 and U73122, no effect on the thapsigargin-induced [¹⁴C]DAG formation was noted in lymphocytes pretreated with 0.3% butanol (Fig. 4).

[View Larger Version of this Image (18K GIF file)]

Due to the uniform radioactivity incorporation in all major phospholipids, the experiments with [1-¹⁴C]arachidonic acid labeling do not allow us to conclude whether [¹⁴C]DAG is primarily derived from PtdIns or PtdChol. To elucidate the phospholipid substrate for thapsigargin-stimulated phospholipase activation, we labeled lymphocytes with 0.1 mCi/ml [³²P]orthophosphoric acid. As shown in Table I, the radioactivity associated with PtdInsP₂, PtdInsP, and PtdIns was not significantly altered following stimulation of lymphocytes with 5 µM thapsigargin. By contrast, 5 µM thapsigargin induced decrease of the radioactivity associated with PtdChol (Fig. 5). This effect was abolished in the presence of 20 µM D609. These results suggest that phosphatidylcholine rather than phosphatidylinositol is the main source for the thapsigargin-induced DAG production.

Table I. Effect of thapsigargin on phosphatidylinositols in human lymphocytes Human lymphocytes labeled with [32P]orthophosphate were stimulated with 5 µM thapsigargin. After indicated times reactions were terminated, and cellular extracts were examined for [32P]PtdlnsP2 and [32P]PtdlnsP as described under "Experimental Procedures." The data represent mean ± S.E. from three experiments. [32P]PtdlnsP2 [32P]PtdlnsP min dlp/ml/106 cells 0 2084  ± 171 2507  ± 205 1 2290  ± 205 2831  ± 315 2 2203  ± 254 2804  ± 301 5 1935  ± 98 2292  ± 237

[View Larger Version of this Image (14K GIF file)]

To investigate further which phospholipid pool serves as a source for DAG during lymphocyte stimulation with thapsigargin, cells were labeled for 2 h with 0.2 µCi/ml [1-¹⁴C]lyso-phosphatidylcholine. Under this condition, the majority of the radioactivity incorporated into phospholipids was found in PtdChol with less than 2% incorporated into PtdIns. Addition of 5 µM thapsigargin to the lymphocyte suspension resulted in a substantial increase in [¹⁴C]DAG. The maximum response (74 ± 4.1% over the basal level (n = 6)) was attained 60 s after stimulation. The [¹⁴C]DAG level remained elevated above the basal value from 60 s onwards (Fig. 6A). In the presence of 0.3% butanol or 50 µM propranolol, [¹⁴C]DAG increased by 54 ± 7.6% (n = 3) and 71 ± 11.3%, respectively, within 120 s after stimulation. The amount of [¹⁴C]DAG then declined within the next 180 s but remained elevated above the basal level. Preincubation of lymphocytes with 20 µM D609 or 10 µM U73122 virtually abolished the thapsigargin-induced accumulation of [¹⁴C]DAG. As shown in Fig. 6B, the inhibition of the thapsigargin-stimulated [¹⁴C]DAG formation by D609 and U73122 was concentration-dependent. For both compounds the inhibitory effect was noted at concentrations above 10 µM.

[View Larger Version of this Image (22K GIF file)]

In contrast to [¹⁴C]DAG, no increase in [¹⁴C]PtdOH was detected following stimulation of [¹⁴C]lyso-phosphatidylcholine-labeled lymphocytes with thapsigargin (not shown).

In the presence of primary alcohols PC-PLD catalyzes a phosphatidyl transfer reaction yielding poorly metabolized phosphatidyl alcohols. Since trans-phosphatidylation is catalyzed solely by PLD, synthesis of phosphatidyl alcohols is considered to be an unequivocal marker of PLD activation. To confirm the lack of PC-PLD involvement in the thapsigargin-stimulated DAG production, lymphocytes were stimulated with various concentrations of thapsigargin in the presence of 0.3% (v/v) butanol. Phosphatidylbutanol (PtdBut) formation was measured 5 min after stimulation. No significant [¹⁴C]PtdBut formation was observed in cells stimulated with thapsigargin at concentrations up to 10 µM (Fig. 7). By contrast, marked [¹⁴C]PtdBut formation within 5 min after stimulation was observed in the same experiment when PMA was used instead of thapsigargin (Fig. 7). The latter compound stimulates PC-PLD via PKC activation.

[View Larger Version of this Image (18K GIF file)]

To study further the role of PC-PLC and PC-PLD in the regulation of the store-operated Ca²⁺ influx, we examined the effects of exogenous PC-PLC and PC-PLD on the thapsigargin-induced Ca²⁺ elevation. Fig. 8A demonstrates that the addition of 5 units/ml PC-PLC to lymphocytes resulted in a rapid formation of [¹⁴C]DAG, whereas the addition of PC-PLD led to the accumulation of [¹⁴C]PtdOH. As shown in Fig. 8B, depletion of intracellular Ca²⁺ stores with 5 µM thapsigargin for 10 min resulted in an increase of [Ca²⁺]_i by 587 ± 77 nM (n = 24). Subsequent addition of 5 units/ml PC-PLC led to a significant decrease of [Ca²⁺]_i by 309 ± 48 nM (n = 14; p < 0.001 versus control) (Fig. 8, B and C). This effect was completely abolished in the presence of 20 µM D609 (Fig. 8, B and C) or when heat-treated PC-PLC was used instead of the native one (not shown). In contrast to PC-PLC, no significant effect on [Ca²⁺]_i was noted when 10 units/ml PC-PLD were added to lymphocytes prestimulated with thapsigargin (Fig. 8, B and C).

[View Larger Version of this Image (21K GIF file)]

Treatment of cells with thapsigargin causes depletion of intracellular calcium stores and triggers influx of extracellular calcium. In the absence of extracellular calcium the thapsigargin-induced [Ca²⁺]_i increase is due to calcium release from intracellular stores. A typical tracing of thapsigargin-triggered [Ca²⁺]_i elevation in lymphocytes resuspended in Ca²⁺-free medium is shown in Fig. 9A. Under this experimental condition [Ca²⁺]_i rose by 20 ± 4 nM (n = 18; p < 0.01). Similarly, in the presence of 5 mM Ni²⁺, which blocks divalent cation entry pathways, thapsigargin-induced [Ca²⁺]_i increase was significantly reduced and averaged 83 ± 20 nM (n = 14, p < 0.001).

[View Larger Version of this Image (19K GIF file)]

We next investigated whether the thapsigargin-induced DAG formation depends on the store-operated calcium influx. [¹⁴C]Arachidonic acid-labeled lymphocytes were suspended in Ca²⁺-free medium or in Ca²⁺-containing medium in the presence of 5 mM Ni²⁺ and stimulated with 5 µM thapsigargin. Fig. 9B demonstrates that under these experimental conditions the thapsigargin-triggered DAG formation was markedly reduced indicating that PC-PLC activation depends on the extracellular calcium entry.

Since DAG liberated by PC-PLC is the main physiological activator of PKC, we next investigated the effect of PKC modulation on the thapsigargin-induced Ca²⁺ increase. This was accomplished using the direct PKC activator, PMA, or the selective PKC inhibitor Ro31-8220. As shown in Fig. 10A, 10 µM PMA markedly inhibited Ca²⁺ increase induced by 5 µM thapsigargin. Under this experimental condition [Ca²⁺]_i increased by 65 ± 18 nM (n = 6, p < 0.01). By contrast, in the presence of 10 µM Ro31-8220 the thapsigargin-induced [Ca²⁺]_i increase was significantly enhanced and averaged 348 ± 41 nM (n = 12, p < 0.01). The effect of PKC activation on the Ca²⁺ elevation was further examined in cells prestimulated for 10 min with 5 µM thapsigargin. Under this experimental condition, direct stimulation of PKC with 10 µM PMA led to a gradual decrease of [Ca²⁺]_i by 202 ± 22 nM (n = 6, p < 0.01 versus control) (Fig. 10B). By contrast, inactive PMA analogue 4-PMA (10 µM) failed to affect [Ca²⁺]_i in cells prestimulated with thapsigargin ([Ca²⁺]_i 26 ± 29 nM, n = 4, not significant). Furthermore, the PC-PLC-induced decrease of [Ca²⁺]_i was significantly inhibited by the PKC inhibitor Ro31-8220 (Fig. 10B). Under this experimental condition [Ca²⁺]_i decreased by 152 ± 28 nM (n = 9, p < 0.02 versus decrease by PC-PLC alone).

[View Larger Version of this Image (25K GIF file)]

DISCUSSION

In the present study we examined the role of phosphatidylcholine-specific phospholipases in the regulation of the store-operated calcium influx. For this purpose we utilized two structurally unrelated inhibitors, D609 and U73122. Whereas D609 is a PC-PLC inhibitor (28), the specificity of U73122 is less well defined, and additional effects distinct from inhibition of phospholipases have been observed in the presence of this agent (29, 30). The possible role of PC-PLD for [Ca²⁺] regulation was studied with butanol. In the presence of butanol PC-PLD forms metabolically inactive phosphatidylbutanol instead of its physiologically relevant product PtdOH. In several experimental systems primary alcohols were shown to blunt PC-PLD-mediated physiological responses (31-35). Moreover, we noted decreased PMA-induced PtdOH production in the presence of butanol.

Our results demonstrate enhancement of thapsigargin-induced Ca²⁺ mobilization by D609 and U73122. The larger increases in [Ca²⁺]_i in the presence of U73122 may point to additional, phospholipase C-independent effects of this compound. However, both structurally unrelated agents enhanced thapsigargin-induced Mn²⁺ quenching to a similar extent. By contrast, butanol failed to affect the thapsigargin-induced calcium mobilization. We interpret these findings to mean that PC-PLC but not PC-PLD is involved in the regulation of the store-operated calcium entry in lymphocytes. Further support for this notion is derived from experiments in which the effect of exogenously added PC-PLC and PC-PLD on the thapsigargin-induced [Ca²⁺]_i increase was examined. Previously this approach has been used successfully to mimic the physiological action of epidermal growth factor and tumor necrosis factor in various cell systems (28, 36). Addition of PC-PLC or PC-PLD to lymphocytes led to the rapid formation of DAG or PtdOH, respectively. Furthermore, addition of native but not heat-treated PC-PLC to lymphocytes resulted in a gradual decrease of [Ca²⁺]_i elevated following stimulation with thapsigargin, and this effect was abolished by D609. Thus, the effect of exogenous PC-PLC on [Ca²⁺]_i is opposite to that of PC-PLC inhibitors. By contrast, no effect of exogenously added PC-PLD on [Ca²⁺]_i was observed further confirming the lack of PC-PLD involvement in regulation of Ca²⁺ homeostasis in lymphocytes.

It therefore seems clear that PLC affects Ca²⁺ influx following depletion of intracellular calcium stores with thapsigargin. However, thapsigargin may cause phospholipase activation. Here we show for the first time that phosphatidylcholine breakdown and phosphatidylcholine-derived DAG formation occur in thapsigargin-stimulated lymphocytes. These findings are consistent with the activation of a phosphatidylcholine-specific phospholipase. Moreover, the involvement of phosphoinositide-specific phospholipases is unlikely, as no phosphatidylinositol breakdown was observed in lymphocytes treated with thapsigargin. The latter observation agrees with previous findings that depletion of intracellular calcium stores by thapsigargin is not accompanied by the hydrolysis of phosphatidylinositols (9, 37-39).

For several reasons, the present study indicates that the activation of PC-PLD did not contribute to the thapsigargin-induced formation of PC-derived DAG. First, no substantial thapsigargin-induced increase of PtdOH, the primary PC-PLD product, was noted in lymphocytes labeled with lyso-phosphatidylcholine. Second, preincubation of lymphocytes with butanol should blunt PC-PLD-mediated DAG synthesis by shunting PtdOH to PtdBut. However, both in arachidonic acid-labeled and in lysophosphatidylcholine-labeled lymphocytes butanol failed to diminish the thapsigargin-stimulated DAG formation. Third, only a slight increase in PtdBut was observed in thapsigargin-treated cells, whereas stimulation of lymphocytes with the PKC activator PMA resulted in considerable accumulation of PtdBut. Fourth, the phosphatidic acid phosphohydrolase inhibitor propranolol failed to affect thapsigargin-induced DAG formation excluding PtdOH as a DAG precursor. In agreement with these results previous studies have failed to detect an effect of thapsigargin on PC-PLD in lymphocytes, mesangial cells, and airway smooth muscle cells (40-42).

Numerous investigations have shown that constitutive levels of [Ca²⁺]_i may be necessary and sufficient for proper PC-PLD function (43, 44), that large increments in [Ca²⁺]_i due to Ca²⁺ entry may not provide adequate stimuli for PC-PLD activation (42, 45-47), and that PC-PLD does not necessarily substantially contribute to PC-derived DAG formation (48-51). On the other hand, Ca²⁺ plays a crucial role in the activation of PC-PLC. Ca²⁺ was reported to increase the activity of partially purified PC-PLC (52). Furthermore, a transmembrane Ca²⁺ influx was shown to activate PC-PLC in intact cells and hence to stimulate PC-derived DAG formation (53, 54). In the present study, we have demonstrated that Ni²⁺, a calcium channel blocker which does not enter the cells (55), markedly attenuated both thapsigargin-induced calcium elevation and DAG formation. Likewise, the thapsigargin-induced DAG formation was found to be substantially reduced under Ca²⁺-free conditions. These results strongly support the contention that the thapsigargin-induced DAG formation is related to PC-PLC activation and caused by the Ca²⁺ increase resulting from extracellular calcium entry.

How might thapsigargin-induced phosphatidylcholine breakdown influence Ca²⁺ homeostasis in lymphocytes? DAG, the principal product of PC-PLC is a physiological activator of several enzymes including PKC and sphingomyelinase and may serve as a donor of free fatty acids known to mediate various cellular responses. Our findings that PKC stimulation with PMA inhibited the thapsigargin-induced [Ca²⁺]_i increase, whereas PKC inhibition with Ro31-8220 exerted the opposite effect, and that effects of exogenous addition of PC-PLC could be mimicked by PMA and were inhibited by Ro31-8220, are consistent with the interpretation that the inhibition of intracellular Ca²⁺ increase may represent an important effect of PKC activation in human lymphocytes. This contention is further supported by the reported inhibitory effects of phorbol esters on thapsigargin-induced calcium increase in primary human lymphocytes (56). In the latter study, Ca²⁺ influx has been demonstrated to be the primary and specific target of the phorbol ester-mediated inhibition. Moreover, in Jurkat T cells and in HPB-ALL T-lymphocytes phorbol esters and other PKC activators have been reported to impair anti-CD3 monoclonal antibody-induced Ca²⁺ fluxes, whereas PKC inhibitors exerted opposite effects (15, 57, 58), although the latter authors did not observe major potentiating effects of Ro31-8220 on thapsigargin-induced Ca²⁺ influx. Finally, PKC activation was shown to inhibit thapsigargin-induced Ca²⁺ entry in cells other than lymphocytes such as FRTL cells and neutrophils (59, 60). Based on the present results it is not possible to conclude which mechanisms underlie the inhibitory effect of PKC. One plausible mechanism of action is through membrane depolarization. Previous studies demonstrated that calcium influx is strongly dependent on the membrane potential and that PKC induces depolarization in some cells (59, 61, 62).

The increase in [Ca²⁺]_i during lymphocyte activation is thought to have important functional consequences for their proper function (63-65). The mechanism by which lymphocytes are able to maintain Ca²⁺ increase over prolonged periods is not yet clear. Inhibition of PC-PLC by U73122 and D609 and inhibition of PKC by Ro31-8220 potentiated the influx of extracellular Ca²⁺ triggered by depletion of intracellular Ca²⁺ stores. These observations together with the strong dependence of PC-PLC activation on extracellular calcium entry suggest that a physiological feedback mechanism exists, which is activated by Ca²⁺ increase and acts via consecutive activation of PC-PLC and PKC to limit the rise in [Ca²⁺]_i. On the other hand, the Mn²⁺ quenching rates increased following thapsigargin addition and then decreased again implying that the influx rate is inhibited to some extent even in the presence of PLC blockers. Thus, additional mechanisms may exist by which thapsigargin-induced Ca²⁺ entry is modulated. In several cell types including lymphocytes the existence of the feedback inhibition of store-operated calcium increases responsible for maintaining the long term Ca²⁺ homeostasis was postulated (66-68). Furthermore, feedback regulation of the capacitative calcium entry appears to be heterogenous (69). For the first time, the present study provides an explanation of how one of the putative feedback mechanisms might operate.