Pertussis toxin inhibits phospholipase C activation and Ca2+ mobilization by sphingosylphosphorylcholine and galactosylsphingosine in HL60 leukemia cells. Implications of GTP-binding protein-coupled receptors for lysosphingolipids.

Extracellular sphingosylphosphorylcholine (SPC) and galactosylsphingosine (psychosine) induced Ca2+ mobilization in a dose-dependent manner in HL60 leukemia cells. The rapid and transient increase in intracellular Ca2+ concentration ([Ca2+]i elicited by SPC and psychosine at concentrations lower than 30 μM was inhibited by treatment of the cells with pertussis toxin (PTX) and U73122, a phospholipase C inhibitor, as was the case for UTP, a P2-purinergic agonist. The increase in [Ca2+]i induced by these lysosphingolipids was associated with inositol phosphate production, which was also sensitive to PTX and U73122. The inositol phosphate response is not secondary to the increase in [Ca2+]i as evidenced by the observation that thapsigargin and ionomycin, Ca2+ mobilizing agents, never induced inositol phosphate production and, unlike lysosphingolipids, the [Ca2+]i rise by these agents was totally insensitive to PTX and U73122. When HL60 cells were differentiated into neutrophil-like cells by dibutyryl cyclic AMP, inositol phosphate and Ca2+ responses to AlF−4 were enhanced, probably reflecting an increase in the amount of Gi2 and Gi3 compared with undifferentiated cells. In the neutrophil-like cells, however, the responses to SPC and psychosine were markedly attenuated. This may exclude the possibility that the lysosphingolipids activate rather directly PTX-sensitive GTP-binding proteins or the phospholipase C itself. Other lysosphingolipids including glucosylsphingosine (glucopsychosine) and sphingosylgalactosyl sulfate (lysosulfatides) at 30 μM or lower concentrations also showed PTX- and U73122-sensitive Ca2+ mobilization and inositol phosphate response in a way similar to SPC and psychosine. However, platelet-activating factor and lysoglycerophospholipids such as lysophosphatidylcholine and lysophosphatidic acid were less effective than these lysosphingolipids in the induction of Ca2+ mobilization. Taken together, the results indicate that a group of lysosphingolipids at appropriate doses induces Ca2+ mobilization through inositol phosphate production by phospholipase C activation. The lysosphingolipids-induced enzyme activation may be mediated by PTX-sensitive GTP-binding protein-coupled receptors, which may be different from previously identified platelet-activating factor receptor or lysophosphatidic acid receptor.

Sphingolipids have recently been shown to be important participants in the regulation of a variety of cellular processes (1)(2)(3). Sphingosine, one of the metabolites of sphingolipids, was in its early studies demonstrated as a potent endogenous inhibitor of protein kinase C (1,4) and has been implicated to be a negative regulator for a few signaling processes (1,4). Further studies, however, revealed that the exogenous sphingosine also induces various types of positive biological actions, e.g. activation of phospholipase D (5), stimulation of cell proliferation (6), regulation of Ca 2ϩ mobilization from the internal pool (7)(8)(9)(10)(11), and inhibition of Ca 2ϩ influx through the plasma membranes (12). These actions seem to be exerted through phosphatidate (5) or a phosphorylated product of sphingosine, sphingosine 1-phosphate (S1P) 1 (7,8,(13)(14)(15); many of them were suggested to be independent of protein kinase C. S1P was reported to act directly on the internal Ca 2ϩ pool resulting in Ca 2ϩ mobilization in a way similar to inositol 1,4,5-trisphosphate (8,15). This lysosphingolipid has also been proposed as a second messenger of platelet-derived growth factor and serum on cell proliferation in fibroblasts (16). In the brain and other peripheral tissues of inherited sphingolipid disorders, it has been shown that any one of lysosphingolipids, e.g. sphingosylphosphorylcholine (SPC), galactosylsphingosine (psychosine), or glucosylsphingosine (glucopsychosine), is accumulated (4,(17)(18)(19). These lysosphingolipids might be responsible for the respective pathogenesis (4,(17)(18)(19). SPC has recently been shown, similarly to S1P, to be a potent Ca 2ϩ releaser from the internal pool and suggested to cause the Ca 2ϩ release from the 1,4,5-trisphosphate-sensitive pool in various cell types (7)(8)(9)(10).
These observations suggest that, in addition to protein kinase C inhibition, intracellular Ca 2ϩ mobilization is an important action of lysosphingolipids, which may have pathological and physiological significance. This raises the question of whether the Ca 2ϩ mobilization is caused by the activation of the phospholipase C-Ca 2ϩ signal transduction pathway. In fact, recent studies demonstrated that extracellular S1P in Swiss 3T3 fibroblasts (15) and sphingosine in Swiss 3T3 fibroblasts (5), astrocytes (20), and foreskin fibroblasts (21) can induce inositol phosphate production, probably reflecting activation of phospholipase C. Although the S1P-induced [Ca 2ϩ ] i increase in the cells has been suggested to occur independently of the enzyme activation (15), at least a part of the sphingosineinduced Ca 2ϩ mobilization as well as phospholipase C activation in foreskin fibroblasts was sensitive to PTX, showing some similarity to a typical feature of PTX-sensitive G-protein-mediated activation of the phospholipase C-Ca 2ϩ pathway (21). If this is the case, we might be allowed to imagine the presence of a receptor(s) for the lysosphingolipids which lead to the activation of phospholipase C, although the previous findings have not excluded the possibility that the lipids penetrate into the cells and act on the pathway inside the cells.
In the present paper, our study was focussed on the Ca 2ϩ mobilizing actions of SPC and other lysosphingolipids which are accumulated in the respective sphingolipidosis, especially on the mechanisms of their actions. We found that, in HL60 leukemia cells, extracellularly added lysosphingolipids at 30 M or less induced a rapid and transient increase in [Ca 2ϩ ] i , the features of which are indistinguishable from those of the Ca 2ϩ response induced by UTP, a P 2 -purinergic agonist, in the same cells. The transient [Ca 2ϩ ] i rises were associated with inositol phosphate production, and both Ca 2ϩ and inositol phosphate responses were inhibited by the treatments of cells with PTX and U73122, a potent phospholipase C inhibitor. Our results suggest that extracellular lysosphingolipids at appropriate doses induce a [Ca 2ϩ ] i rise due to the activation of the phospholipase C being mediated by a putative receptor(s) coupled to a PTX-sensitive G-protein(s).
Purity Check and Purification of SPC and Psychosine-According to the "certificate of analysis" of the manufacture, the purity of lysosphingolipids is more than 85% for SPC and more than 95% for psychosine, glucopsychosine, and lysosulfatides. The purity of the lipids was checked in the present study by Silica Gel 60 (Merck) TLC using two solvent systems (solvent I, butanol/water/acetic acid, 3:1:1 (v/v); solvent II, CHCl 3 /MeOH/water/acetic acid, 30:30:2:5 (v/v)) (14). In solvent I and II, R F values were 0.12 and 0.04 for SPC, 0.39 and 0.62 for psychosine, 0.42 and 0.65 for glucopsychosine, and 0.37 and 0.70 for lysosulfatides, respectively. These lipids were detected with ninhydrin (all samples), molybdenum blue (SPC), and anthrone/H 2 SO 4 (psychosine, glucopsychosine, and lysosulfatides) sprays (14,30). In the case of the psychosine, glucopsychosine, and lysosulfatides sample, only a single spot was detected that was positive with ninhydrin and anthrone/H 2 SO 4 on TLC with either solvent. In the case of the SPC sample, however, there was a trace of unknown spot that was positive with ninhydrin, but not with molybdenum blue, at R F ϭ 0.26 in solvent I and R F ϭ 0.20 in solvent II. Since the unknown compound does not seem to affect Ca 2ϩ response (see Fig. 2), these lysosphingolipids were used in the present study without further purification unless otherwise stated. In some experiments in Fig. 2, SPC and psychosine were purified by Silica Gel 60 TLC using solvent I. The region corresponding to SPC or psychosine was scraped off and extracted with CHCl 3 /MeOH/water (10:10:1) for SPC and with MeOH for psychosine. SPC and psychosine were quantified by the malachite green method (31) and the anthrone/H 2 SO 4 method (30), respectively. The TLC-purified lipids were checked for the ability to induce Ca 2ϩ mobilization.
Cell Cultures-HL60 cells were routinely cultured in a RPMI 1640 medium (Sigma) supplemented with 10% fetal calf serum (Life Technologies, Inc.) and maintained in a humidified atmosphere of 95% air and 5% CO 2 . In some experiments in Fig. 7, the cells were cultured for 5 days in a medium containing 500 M dibutyryl cyclic AMP to differentiate into neutrophil-like cells. Two days before the experiments, the cells were sedimented (250 ϫ g for 5 min) and transferred to fresh medium for [Ca 2ϩ ] i measurement and membrane preparation. For inositol phosphate response, the cells were transferred to an inositol-free RPMI 1640 medium containing 10% fetal calf serum and myo- [2-3 H] inositol (4 Ci/ml). PTX treatment of the cells was performed by adding the toxin (50 ng/ml) to the medium 4 h before the experiments.

Measurement of [ 3 H]Inositol
Phosphates Production-The [ 3 H]inositol-labeled cells were washed by sedimentation (250 ϫ g for 5 min) and resuspended with Hepes-buffered medium which consisted of 20 mM Hepes (pH 7.5), 134 mM NaCl, 4.7 mM KCl, 1.2 mM KH 2 PO 4 , 1.2 mM MgSO 4 , 2 mM CaCl 2 , 2.5 mM NaHCO 3 , 5 mM glucose, and 0.1% (w/v) bovine serum albumin (fraction V). The washing procedure was repeated and the cells were finally resuspended in the same medium. The cells (about 2 ϫ 10 6 cells) were preincubated for 10 min with 10 mM LiCl and 0.5 units/ml adenosine deaminase in polypropylene vials (20 ml) in a final volume of 1.5 ml. The test agents (ϫ100) were then added to the medium and incubated for 1 min unless otherwise specified. The cell suspension (0.5 ml) in duplicate was transferred to tubes containing 1 ml of CHCl 3 /MeOH/HCl (200:100:1). [ 3 H]Inositol mono-, di, and trisphosphates were separated as described previously (27). The radioactivity of the trichloroacetic acid (5%)-insoluble fraction was measured as the total radioactivity incorporated into the cellular inositol lipids. Where indicated, the results were normalized to 10 5 cpm of the total radioactivity.
Measurement of [Ca 2ϩ ] i -The cells were sedimented, resuspended in Ham's F-10 medium containing 0.1% bovine serum albumin, and then incubated for 20 min with 1 M Fura-2/AM. [Ca 2ϩ ] i was estimated from the change in the fluorescence of the Fura-2-loaded cells as described previously (27,29).
Immunoblot Analysis-Crude plasma membranes and their cholate extracts were prepared as described previously (25,28). The cholate extract (25 g of protein) was resolved on SDS-polyacrylamide (12.5%) slab gel electrophoresis and then electrophoretically transferred to a Millipore Immobilon sheet (23). G i2 ␣ and G i3 ␣ were visualized by incubating the sheet with a specific rabbit-antiserum to the respective ␣ subunit of G i (22), with an alkaline phosphate-conjugated goat antibody against rabbit IgG and finally with 5-bromo-4-chloro-3-indoylphosphate and nitro blue tetrazolium as described previously (23).
Data Presentation-All experiments were performed in duplicate or triplicate. The results of multiple observations were presented as the representative or means Ϯ S.E. of at least three separate experiments unless otherwise stated. Fig. 1, A and B, shows representative traces of [Ca 2ϩ ] i changes in undifferentiated HL60 cells. Both SPC and psychosine, in a dose-dependent manner, increased [Ca 2ϩ ] i very rapidly. The shape of the rapid and transient increase in [Ca 2ϩ ] i by these lipids is very similar to that obtained with UTP, a P 2 -purinergic agonist (Fig. 1A), which activates phospholipase C through a G-protein-coupled receptor in the same cells (32). The lipid-induced [Ca 2ϩ ] i rise was markedly suppressed by prior treatment of the cells with PTX ( Fig.  1, B-D). The response to UTP was also partially inhibited by the toxin treatment ( Fig. 1, B and F). These results suggest that the increase in [Ca 2ϩ ] i by the lysosphingolipids in HL60 cells involves PTX-sensitive G-proteins.

Extracellular SPC and Psychosine Increase [Ca 2ϩ ]i in a Manner Sensitive to PTX-
As mentioned in the Introduction, another well documented action of lysosphingolipids is protein kinase C inhibition, especially in the earlier period of the studies (4). SPC and psychosine therefore might induce Ca 2ϩ mobilization as a result of the enzyme inhibition. To examine this possibility, we also used sphingosine which is a similar or more potent inhibitor of protein kinase C than SPC or psychosine (4). Sphingosine also increased [Ca 2ϩ ] i ; however, the time course was very slow and the net increase was much less than that induced by similar doses of SPC and psychosine (Fig. 1A). Moreover, the sphingosine-induced [Ca 2ϩ ] i increase was totally insensitive to PTX (Fig. 1, B and E). We also examined the effect of another protein kinase C inhibitor, 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H7), but this drug never increased [Ca 2ϩ ] i (data not shown). In fibroblasts (11,21), the sphingosine-induced [Ca 2ϩ ] i rise has been suggested to be protein kinase C-inde-pendent. These results suggest that at least the PTX-sensitive increase in [Ca 2ϩ ] i by SPC and psychosine is independent of the protein kinase C inhibition.
TLC analysis of the SPC sample showed the presence of a small, but detectable, amount of unknown compound that is positive with ninhydrin at R F ϭ 0.26 ( Fig. 2A, lane 1). However, it was confirmed that SPC itself elicited the Ca 2ϩ response and the contaminated unknown compound is inactive to induce the response (Fig. 2). Sphingomyelinase from B. cereus almost completely converted SPC to sphingosine, but did not influence the unknown compound ( Fig. 2A, lanes 2-4). The enzyme-treated SPC never elicited a rapid and transient [Ca 2ϩ ] i increase which is a characteristic to the untreated SPC, instead induced a rather slow increase probably due to sphingosine (Fig. 2C). The enzyme was rather specific to SPC; psychosine was tolerable to the enzyme ( Fig. 2A, lanes 6 and 7) and the lipid-induced [Ca 2ϩ ] i increase was unaffected by its treatment (Fig. 2C). Furthermore, the TLC-purified SPC sample, which is free from the unknown compound ( Fig. 2A, lane 5), induced the Ca 2ϩ response to an extent similar to that of the unpurified SPC (Fig. 2C).
Although psychosine obtained from the drug company showed a single spot that is positive with ninhydrin and anthrone/H 2 SO 4 on TLC using two solvent systems, we further purified the psychosine by TLC (Fig. 2B, lanes 1 and 2). The purified psychosine also induced a rapid and transient [Ca 2ϩ ] i increase as effectively as the unpurified psychosine did (Fig.  2D). Since the active compound to induce Ca 2ϩ mobilization was demonstrated to be SPC or psychosine itself and furthermore, there was no appreciable difference in the ability to induce Ca 2ϩ response between purified and unpurified products, we performed the following experiments without further purification of the lipids.
SPC and Psychosine Mobilize Ca 2ϩ from the Internal Pool in a Manner Sensitive to U73122, a Phospholipase C Inhibitor-As shown in Fig. 3, an addition of excess EGTA to the incubation medium hardly affected the [Ca 2ϩ ] i increases due to SPC, psychosine, and UTP at 30 M, 30 M, and 1 M, respectively. These results suggest that the increased [Ca 2ϩ ] i induced by these lysosphingolipids is derived predominantly from intracellular pools. Although this is not inconsistent with the recent observations where SPC mobilizes Ca 2ϩ by rather direct interactions with intracellular pools in DDT 1 MF-2 smooth muscle cells (7,8), pancreatic acinar cells (9), and basophilic leukemia cells (10), the present finding of the similarity of the Ca 2ϩ response pattern to the UTP actions also suggests that the Ca 2ϩ mobilization by these lysosphingolipids is caused by the activation of the phospholipase C-Ca 2ϩ signaling pathway. In favor on the latter suggestion, U73122, a potent phospholipase C inhibitor (33) completely inhibited the SPC effect at 30 M (Fig. 3A). The situation was similar for psychosine, although the 30 M psychosine effect was not completely abolished by the phospholipase C inhibitor (Fig. 3B). Under these conditions, the UTP effect was totally sensitive to U73122 (Fig.  3C). These results suggest that SPC and psychosine at 30 M induce Ca 2ϩ mobilization predominantly through phospholipase C activation.
SPC and Psychosine Produce Inositol Phosphate- Fig. 4, A, C, and E, show that 30 M SPC and psychosine induced inositol phosphate production, which may reflect activation of phospholipase C. The time courses of the production of three species of inositol phosphate induced by both SPC and psychosine were very similar to those by UTP which activates the enzyme through a P 2 -receptor (32). The actions of the lysosphingolipids as well as UTP were markedly inhibited by a PTX treatment (Fig. 4, B, D, and F), suggesting the involvement of a PTXsensitive G-protein(s) in the lysosphingolipids-induced phospholipase C activation. The PTX treatment suppressed more than 70% of the lipid-induced activation at any dose (Fig. 5, A  and B). As shown in this figure, U73122 markedly inhibited the inositol phosphate production, confirming that these lysosphingolipid actions are due to the activation of phospholipase C.
PTX and U73122-sensitive Activation of Phospholipase C Is not Secondary to the [Ca 2ϩ ]i Rise-The actions of lysosphingolipids, as shown in the previous section, bear characteristics of the activation of the phospholipase C-Ca 2ϩ pathway through receptors coupling to PTX-sensitive G-proteins. On the other hand, another possibility remains that might explain the events in a reverse way, that is, a Ca 2ϩ -induced phospholipase C activation, because previous studies have demonstrated phospholipase C activation by increased [Ca 2ϩ ] i (34) in addition to lysosphingolipid-induced [Ca 2ϩ ] i increase by their direct action on intracellular Ca 2ϩ pools (7)(8)(9)(10). This possibility, however, can be ruled out based on the following observations. In Fig. 6, we examined the effect of ionomycin, a Ca 2ϩ ionophore, and thapsigargin on the cells. Thapsigargin inhibits Ca 2ϩ uptake into its intracellular pool by inhibiting Ca 2ϩ -ATPase, resulting in an increase in [Ca 2ϩ ] i . Both agents increased [Ca 2ϩ ] i to an extent similar to 30 M SPC and psychosine. The Ca 2ϩ increase by these agents, however, was hardly modified by the treatments of the cells with U73122 and PTX (Fig. 6, A and B). Moreover, inositol phosphate was not significantly produced by the incubation of the cells with these Ca 2ϩ mobilizers for at least 5 min, while in the same experiment, an appreciable production of inositol phosphate was found in the presence of SPC at 30 M (Fig. 6C).

Differentiation into Neutrophil-like Cells Was Associated with Attenuation of Responses to Lysosphingolipids-HL60 cells can be differentiated into neutrophil-like cells by treatment of the cells with dibutyryl cyclic AMP or other inducers.
Increase in PTX-sensitive G-proteins, G i2 and G i3 , is accompanied by differentiation (32,35). Because the foregoing results suggest an involvement of the toxin-sensitive G-proteins in the lysosphingolipid signaling, the cell differentiation would potentiate the actions of lysosphingolipids.
As shown in Fig. 7A, the contents of G i2 and G i3 were actually increased by a dibutyryl cyclic AMP treatment of the cells as evidenced from increases in immunodetectable G i2 ␣ and G i3 ␣. The dibutyryl cyclic AMP-treated cells also showed a PTXsensitive formyl-Met-Leu-Phe-induced [Ca 2ϩ ] increase (Fig.  7C) and inositol phosphate production (Fig. 7E), currently recognized to be differentiation markers. AlF 4 Ϫ , a non-selective G-protein activator, induces phospholipase C activation and the subsequent Ca 2ϩ mobilization in many types of cells (34,36). These AlF 4 Ϫ actions shown in Fig. 7, D and E, are very slow, but significant, and are slightly stronger in the differentiated cells than in the undifferentiated ones, probably reflecting higher contents of G i proteins in the neutrophil-like differentiated cells than in the undifferentiated cells (Fig. 7, D and E). Unexpectedly, however, the SPC-induced Ca 2ϩ mobilization was markedly attenuated in the neutrophil-like cells (Fig. 7, B and C). The Ca 2ϩ response to psychosine was also decreased (Fig. 7D). In parallel with the Ca 2ϩ response, the inositol phosphate response to SPC and psychosine was clearly attenuated by differentiation, suggesting that the lipids signaling of the PTX-sensitive G-protein-coupled phospholipase C-Ca 2ϩ pathway is blocked before a G-protein step in the neutrophil-like cells (Fig. 7E).
Glucopsychosine and Lysosulfatides Also Induce Ca 2ϩ Mobilization and Inositol Phosphate Production-We next examined the effects of glucopsychosine and lysosulfatides, which have been suggested to be accumulated in other sphingolipidoses, i.e. Gaucher's disease and metachromatic leukodystrophy, respectively (4,(17)(18)(19), on [Ca 2ϩ ] i and inositol phosphate production. increase in [Ca 2ϩ ] i was followed by the sustained increase. The early transient rises at 10 and 30 M lipids were markedly suppressed by U73122 and PTX treatment, while the later sustained increase was rather resistant to these agents. The U73122-insensitive [Ca 2ϩ ] i increase was also detected in the presence of 2.5 mM EGTA (data not shown), suggesting that the source of Ca 2ϩ is the internal pool. However, because it is also possible that the lipids induced the leakage of the fluorescence indicator, Fura-2, we cannot conclude that the U73122-insensitive fluorescence change reflects [Ca 2ϩ ] i change under the present experimental conditions. 2 In any event, the results shown in Fig. 8, A and B, suggest that at least the transient increase in [Ca 2ϩ ] i at early phase is due to the PTX-sensitive phospholipase C activation. In fact, as shown in Fig. 8C, a significant inositol phosphate production was observed immediately after the addition of either lysosphingolipid, although the lysosulfatides effect was diminished after 1 min. The inositol phosphate production was also abolished by PTX treatment (Fig. 8D). Thus these two lyso compounds caused essentially the same responses in the cells as those induced by SPC and psychosine.
Lysoglycerophospholipids and Platelet-activating factor Also Induce Ca 2ϩ Mobilization, but They Were Less Effective Than Lysosphingolipids-We also examined the effect of sphingomyelin and galactosylceramide on Ca 2ϩ mobilization. These lipids are derivatives of SPC and psychosine, respectively, and each having a fatty acyl moiety linked to their amino group. However, these sphigolipids hardly influenced the [Ca 2ϩ ] i level, confirming again that the SPC and psychosine effects on the [Ca 2ϩ ] i level is not due to the possible contamination of the precursor molecules (Table I). Some of glycerophospholipids and lysoglycerophospholipids, such as platelet-activating factor and lysophosphatidic acid, have already been shown to induce a variety of biological responses including Ca 2ϩ mobilization in many types of cells (37,38). As shown in Table I, plateletactivating factor and some of lysoglycerophospholipids also induced significant Ca 2ϩ mobilization, but none of them was as potent as SPC and other lysosphingolipids. PTX treatment was also inhibitory for their action except for the lysophosphatidic acid-induced one; the Ca 2ϩ mobilization induced by lysophosphatidic acid was hardly affected by toxin treatment (Table I). In contrast to SPC and psychosine action (Fig. 7), plateletactivating factor-induced Ca 2ϩ response was markedly enhanced, but not attenuated, by dibutyryl cAMP-induced differentiation; net [Ca 2ϩ ] i increase by platelet-activating factor at FIG. 7. Differentiation into neutrophil-like cells attenuates phospholipase C and the subsequent Ca 2؉ mobilization in response to SPC and psychosine. In A, cell membranes were prepared from undifferentiated cells (a) and neutrophil-like cells differentiated by dibutyryl cyclic AMP (b). Their cholate extracts were subjected to a SDS-polyacrylamide gel electrophoresis, transferred to an Immobilon sheet, and then probed with G i2 ␣or G i3 ␣-specific antiserum as de- 10 M in the differentiated cells was 708 Ϯ 60% of that in the undifferentiated control cells (number of observations was 3). DISCUSSION In the present paper we have shown that lysosphingolipids (SPC, psychosine, glucopsychosine, and lysosulfatides) at doses lower than 30 M induce phospholipase C activation and the subsequent Ca 2ϩ mobilization in a manner sensitive to PTX and U73122, a phospholipase C inhibitor. This is, to our knowledge, the first indication that these lysosphingolipids activate the phospholipase C-Ca 2ϩ system possibly through receptors coupling to a PTX-sensitive G-protein(s). The putative receptors may be different from the previously identified platelet-activating factor receptor (37) and lysophosphatidic acid receptor (38).
As far as extracellular SPC-induced intracellular Ca 2ϩ mobilization is concerned, a few studies on fibroblasts (39) and FRTL-5 thyroid cells (40) have been reported. However, no significant production of inositol phosphate was observed in these experiments (39,40), despite the fact that Ca 2ϩ mobilizing receptor agonists, such as bradykinin, induced not only Ca 2ϩ mobilization to an extent similar to that with SPC but also phospholipase C activation under the same conditions (39). In addition, in other studies, SPC mobilized Ca 2ϩ from permeabilized cells (7-10) and purified endoplasmic reticulum membrane vesicles (8). On the basis of these previous results, the SPC actions have been currently considered to occur inside the cells by the incorporated SPC molecules, without activating phospholipase C. The present results, however, suggest that at least the early phase of the Ca 2ϩ mobilization induced by lower than 30 M SPC or other lysosphingolipids in intact HL60 cells  is mediated by the activation of the enzyme. This suggestion is based on the following findings. First, SPC and other lysosphingolipids at doses lower than 30 M induced immediate activation of phospholipase C. Second, U73122, a potent phospholipase C inhibitor, suppressed at least the early phase of the lysosphingolipids-induced increase in [Ca 2ϩ ] i . Third, treatment of the cells with either PTX or dibutyryl cyclic AMP attenuated both the lysosphingolipid-induced phospholipase C activation and Ca 2ϩ mobilization. Finally, phospholipase C activation is not a secondary response to the increase in [Ca 2ϩ ] i ; agents such as thapsigargin and ionomycin, which primarily increase [Ca 2ϩ ] i , never activated the enzyme in HL60 cells under the present conditions (Fig. 6).
Several findings in the present study suggest that the lysosphingolipids signaling is performed through G-protein-coupled receptors. The pattern and kinetics of [Ca 2ϩ ] i increase and inositol phosphate production by the lysosphingolipids were very similar to those of the responses to a G-protein-coupled receptor agonist, UTP (a P 2U -purinergic agonist) (Figs. 1, 3,  and 4). Furthermore, as stronger evidence for the involvement of G-protein coupled receptors, the lysosphingolipid actions are suppressed by prior treatment of the cells with PTX which, as is well known, ADP-ribosylates G i -proteins and thereby blocks communication between receptors and effector enzymes. Similar PTX sensitivity has already been shown in the phospholipase C activation induced by several receptor agonists such as formyl-Met-Leu-Phe and UTP in leukocytes such as HL60 cells and neutrophils. This finding has been concluded to reflect the fact that receptors coupling to PTX-sensitive G-proteins mediate the phospholipase C activation (24 -26, 32). In this analogy, it is reasonable to assume that the lysosphingolipid actions are mediated via G i -protein-coupled receptors. It is still possible, however, that amphipathic lysosphingolipids penetrate into the cells and then directly activate G i -proteins. If this was the case, PTX would block the lipid-induced actions. This possibility is excluded from the experiments shown in Fig. 7. Dibutyryl cyclic AMP-induced differentiation into neutrophil-like cells enhanced AlF 4 Ϫ (a nonspecific G-protein activator)-induced phospholipase C activation, probably reflecting the increase in the amount of G i -proteins. This suggests that in the differentiated cells, the downstream region of the G-protein-mediated signaling cascade leading to phospholipase C activation and Ca 2ϩ mobilization is rather fortified by the increase in PTXsensitive G-proteins. On the contrary, the SPC and psychosineinduced enzyme activation was seriously suppressed by differentiation of the cells. This suggests that differentiation impairs the process between the action sites of lipids (or receptors) and G-proteins and hence may rule out the possibility that these lysosphingolipids directly activate G-proteins. Thus, the present pharmacological study suggests the existence of G-proteincoupled receptors for lysosphingolipids, although conclusive evidence for the existence of the receptors will have to await their molecular cloning.
In addition to lysosphingolipids, platelet-activating factor and some lysoglycerophospholipids, such as lysophosphatidylcholine and lysophosphatidic acid, also induced Ca 2ϩ mobilization in HL60 cells, but they were not as effective as SPC and other lysosphingolipids (Table I). Furthermore, in contrast to SPC and psychosine effects which were attenuated in dibutyryl cAMP-induced differentiated cells (Fig. 7), platelet-activating factor-induced response was conversely enhanced by the induction of differentiation, suggesting that the putative receptors for lysosphingolipids are different from platelet-activating factor receptor. Among lysoglycerophospholipids examined, lysophosphatidylcholine was the most effective in the induction of Ca 2ϩ mobilization (Table I). Similarly to the actions of ly-sosphingolipids, the lysophosphatidylcholine effect was PTXsensitive, whereas the lysophosphatidic acid-induced response was not (Table I). Thus, the receptor for lysophosphatidic acid (38) appears to be different from putative receptors for lysosphingolipids. On the other hand, it remains unclear whether lysoglycerophospholipids (including lysophosphatidylcholine, lysophosphatidylethanolamine, and lysophosphatidylinositol) other than lysophosphatidic acid share with lysosphingolipids the same receptor and signaling pathways.
Among lysosphingolipids, sphingosine and S1P have been previously shown to induce phospholipase C activation and the Ca 2ϩ mobilization in a few types of cells (5,15,20,21). In HL60 cells, sphingosine induced the Ca 2ϩ mobilization; however, this action was PTX-insensitive (Fig. 1). Furthermore, the [Ca 2ϩ ] i increase due to the lipid was so slow that it took 1-3 min to reach a peak value (Fig. 1). Thus, the sphingosine signaling pathway seems to be different from that of SPC and other lysosphingolipids. This also suggests that the PTX-sensitive Ca 2ϩ mobilization by lysosphingolipids cannot be explained by the inhibition of protein kinase C, because sphingosine is a protein kinase C inhibitor similar to or more potent than the lysosphingolipids examined in the present study (4). We also preliminarily examined S1P actions on phospholipase C and the Ca 2ϩ mobilization in HL60 cells. This lipid also activated the enzyme and increased [Ca 2ϩ ] i in the cells. In this case, we could not detect any difference between S1P and SPC actions in their sensitivity to PTX and U73122. Thus, S1P seems to share a signaling pathway similar to that of SPC in HL60 cells. In Xenopus oocytes, however, S1P activated a Cl Ϫ channel probably through phospholipase C activation, but SPC could not mimic the S1P action (41). The receptor cloning again would make it clear whether all the lysosphingolipids and some lysoglycerophospholipids share the same receptor or each lipid interacts with its own receptor.
At the present stage of investigation, the physiological roles of the lysosphingolipid-induced activation of the phospholipase C-Ca 2ϩ pathway in leukocytes have not been clarified yet. This type of lysosphingolipid signaling was attenuated by dibutyryl cyclic AMP-induced differentiation of HL60 cells into neutrophil-like cells (Fig. 7). In the preliminary experiments, we found that other differentiation inducers such as dimethyl sulfoxide, retinoic acid, and vitamin D 3 also diminished such lysosphingolipid signaling. This may suggest that only under undifferentiated conditions lysosphingolipids act as physiological and extracellular signals which are oriented to the phospholipase C-Ca 2ϩ pathway. In the previous study in differentiated cells such as fibroblasts (39,42) and thyroid cells (40), SPC has been shown to be a potent mitogen. The Ca 2ϩ mobilizing action of SPC may be involved in the cell proliferation (39,40). A preliminary finding in the undifferentiated HL60 cells, however, showed that SPC rather attenuated the cell growth and instead facilitated cell attachment to culture dishes. This phenomenon might reflect a physiological role of SPC as an inducer of cell differentiation. Further study is now in progress to clarify this point.
The possible existence of cell surface receptors for lysosphingolipids may allow consideration of a novel autocrine or paracrine regulatory mechanism operated by the lysosphingolipids in a way similar to other lipids mediators such as prostaglandins and leukotriens. At present, there are no data on the extracellular occurrence of lysosphingolipids in vivo. To establish the autocrine or paracrine role of the lipids, further studies on the problems are needed, which include characterization of intracellular and extracellular metabolic pathways and physiological functions of the lysosphingolipids as well as identification of their putative receptors.