INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE SPHINGOMYELIN PATHWAY is emerging as an important regulator of membrane signal transduction and, thus, a variety of cellular functions (8, 12-14, 18, 21). Sphingomyelin cell signaling is initiated by activation of sphingomyelinase (SMase), which catalyzes the breakdown of sphingomyelin to form ceramide and phosphorylcholine (13, 14, 18, 21). Ceramide, released as a consequence of sphingomyelinase, is now thought to play roles in fundamental processes such as cell proliferation, membrane-receptor functions, oncogenesis, and immune inflammatory responses (8, 12-14, 18, 21). Ceramide, derived from sphingomyelin, can be phosphorylated by ceramide kinase to ceramide-1-phosphate (11, 21), and <10% of the ceramide generated is converted to free sphingoid base (20), presumably via the action of a neutral ceramidase (18). It is now evident that the sphingomyelin pathway is involved in atherogenesis (5, 7, 10, 29) and vascular biology (10). However, little is known about the vasomotor effects of sphingomyelin metabolites. It has been shown that a cell-permeable synthetic ceramide, C₂ ceramide, induced a sustained contraction in isolated smooth muscle cells of the rabbit rectosigmoid (28). Our recent in vivo studies on the rat mesenteric microcirculation reveal that local administration (perivascular or intra-arterial) of C₂ ceramide produces potent contractile actions on the microscopic resistance and capacitance vessels (arterioles and venules) (unpublished data). In vitro, we have recently found that cerebral arterial smooth muscle can rapidly generate ceramide and other sphingolipids under certain conditions (5, 23), leading us to hypothesize that ceramides and sphingolipids may exert effects on cerebral vasomotor tone.

Bacterial SMase has been shown to activate the sphingomyelin pathway directly (17). Exogenously applied bacterial SMase from Bacillus cereus functions at neutral pH (9) and thus has been considered a useful tool in tissue culture experiments to induce elevation of cellular ceramide levels in an attempt to mimic the biological effect of activation of cellular SMase (19, 24). Because intracellular free Ca²⁺ concentration ([Ca²⁺]_i) plays a key role in regulating vascular tone, the present studies were designed to investigate effects of neutral SMase (N-SMase), C₈ and C₁₆ ceramides, other ceramide analogs, and phosphorylcholine on vascular tone and Ca²⁺ mobilization in isolated canine cerebral arterial smooth muscle.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals, vessels, cell preparations, and solutions. Canine basilar arteries were obtained from male and female mongrel dogs (15-20 kg) after anesthetization with pentobarbital sodium (40 mg/kg iv) and were placed in normal Krebs-Ringer bicarbonate (NKRB) solution at pH 7.4 containing (in mM) 118 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 2.5 CaCl₂, 10 dextrose, and 25 NaHCO₃ (4, 15). In some preparations the endothelium was removed by gently rubbing tissue against the teeth of a pair of fine forceps (33). The vessels were cut into small segments ~3-4 mm in length. These segments were mounted on stainless steel pins under resting tensions of 2.0-3.0 g in an organ bath isometrically and were attached to force transducers (Grass model FT 03, Grass Instruments, Quincy, MA) connected to Grass model 7 polygraphs. The organ baths containing NKRB solution were gassed continuously with 95% O₂-5% CO₂ and warmed to 37°C (pH 7.4). Tissues were allowed to equilibrate for at least 90 min before data were collected. Incubation medium was routinely changed every 15 min as a precaution against interfering metabolites (3). Stimulation of rings with 60 mM KCl was repeated every 30-45 min two to three times until responses were stable. The successful removal of endothelium was assessed by performing successive concentration-response curves to substance P (10⁹-10⁷ M) and by showing that substance P (5 × 10⁸ M) failed to relax segments denuded of endothelium precontracted by 2 × 10⁷ M PGF₂, whereas substance P did relax the endothelium-intact segments (34, 35, 38).

Experimental procedures for canine cerebral arterial rings. Rings of canine basilar arteries with and without endothelium were exposed to N-SMase (0.001-0.1 U/ml), C₈ and C₁₆ ceramide, other ceramide analogs, and phosphorylcholine at various concentrations for periods of at least 15 min for each dose. C₈ and C₁₆ ceramide were dissolved in DMSO, and aliquots were added to an aqueous medium with sonication to a stock concentration of 1 mM. The stock solution was diluted in the NKRB solution, and aliquots were added to the chambers and allowed to rapidly mix at concentrations over the range from 10⁹ to 10⁵ M. N-SMase (0.001-0.1 U/ml) , C₈ ceramide, C₁₆ ceramide, other ceramide analogs, and phosphorylcholine (10⁹-10⁵ M) were added, cumulatively, to the tissue baths, respectively. The results of these experiments are expressed in developed tension (in g) and percentage contraction of the maximal contractions induced by the lipid molecules (reference contraction). In some experiments the extracellular Ca²⁺ concentration ([Ca²⁺]_o) in the NKRB solution was removed to determine the role of, and need for, [Ca²⁺]_o in the vascular actions of N-SMase, C₈ ceramide, and C₁₆ ceramide. For further clarification of the mechanism of sphingomyelinase action and to test for a possible role of [Ca²⁺]_o, the rings were exposed to verapamil and nimodipine (L-type Ca²⁺ channel blockers) and to various specific pharmacological antagonists for 20 min before stimulation with N-SMase. To determine whether myosin light chain kinase (MLCK) and protein kinase C (PKC) activations play any important role in the initiation of vascular smooth muscle contraction and the sustained phase of vascular smooth muscle contraction in N-SMase-induced contraction, two inhibitors of PKC (staurosporine and bisindolylmaleimide I) and an inhibitor of MLCK (ML-9) were tested over a wide concentration range; these drugs were also added 15-20 min before N-SMase was added.

Extracellular and intracellular Ca²⁺ determinations. For intracellular Ca²⁺-buffered experiments, arterial segments were preincubated with 2 × 10⁵ M 1,2-bis(2-aminophenoxy)- ethane-N,N,N',N'-tetraacetic acid (BAPTA)-AM (an intracellular Ca²⁺ chelator) in the bath medium for 20 min before being precontracted by EC₅₀ concentrations of PGF₂. When stable contraction of canine basilar arterial rings was obtained, we investigated the effects of pretreatment of the ring segments with 2 × 10⁵ M BAPTA-AM on both 60 mM KCl- and 1 × 10⁶ M PGF₂-induced contractions. After canine basilar arterial segments were incubated with 2 × 10⁵ M BAPTA-AM for 20 min, 60 mM KCl or 1 × 10⁶ M PGF₂ retained an ability to contract the intact vascular tissues, but the maximal responses of the vessels were suppressed by ~25%. For the extracellular Ca²⁺-free experiments, the canine basilar arterial rings were equilibrated in Ca²⁺-free Krebs-Ringer bicarbonate solution containing 2.0 mM EGTA and 2 × 10⁵ M BAPTA-AM for at least 60 min before the experiments were initiated. Depletion of extracellular Ca²⁺ was confirmed by repeating the high extracellular KCl- and PGF₂-induced contractions. To obtain similar degrees of tone, the concentrations of PGF₂ required to induce precontractions for untreated vessels were lowered, because buffering of intracellular Ca²⁺ and removal of extracellular Ca²⁺ attenuated the vessel tone.

Procedures for cell culture. The procedure employed to isolate and culture single canine cerebral vascular smooth muscle cells and the use of digital imaging microscopy with the fluorescent indicator fura 2 have been reported previously (15, 36). Briefly, primary vascular smooth muscle cells from canine basilar arteries were isolated and cultured in DMEM mixed with Ham's nutrient mixture F-12 (1:1 vol/vol), penicillin (100 U/ml), streptomycin (100 µg/ml), and 0.1% lipid mixture and supplemented with 20% fetal bovine serum (FBS) at 37°C in a humidified atmosphere composed of 95% air and 5% CO₂. Morphological examination of confluent cultures revealed vascular smooth muscle cells exhibiting a crisscross pattern, hills and valleys, and nodular structures when examined by phase-contrast microscopy (15, 36). Immunohistochemical staining with a monoclonal antibody recognizing exclusively -smooth muscle actin indicated that >97% of the cultures were pure vascular smooth muscle cells.

[Ca²⁺]_i studies in single cultured canine cerebral vascular smooth muscle cells. [Ca²⁺]_i in single vascular smooth muscle cells was measured according to previously established methods (15, 36). Cells for image analysis experiments were seeded on glass coverslips (~1 × 10⁴ cells/coverslip) and used 2-3 days postseeding. Monolayers of the cerebral arterial smooth muscle cells, grown on the coverslips, were loaded with 2.0 µM fura 2-AM and 0.12% Pluronic acid F-127 (40 min, 37°C). The monolayers were washed two to three times with PBS and 20 mM HEPES (pH 7.4) and were incubated with this buffer at room temperature until ready to use. The HEPES buffer solution contained (in mM) 118 NaCl, 4.8 KCl, 2.5 CaCl₂, 1.2 KH₂PO₄, 1.2 MgSO₄, 5 HEPES, and 10 glucose. The pH was brought to 7.4 with NaOH. The monolayers were inserted in a leak-proof coverslip holder. Buffer was added to the monolayer on the coverslip. The coverslip holder was mounted onto the stage of a temperature-controlled Nikon TMS inverted microscope with a long working distance Nikon Fluor objective (N.A. 0.5) attached to a 300-W xenon light source and a charge-coupled device camera for image acquisition. Buffer (control), N-SMase, ceramide molecules, or pharmacological antagonists were added to the monolayers in the above-mentioned setup.

Chemicals and reagents. The following pharmacological agents were purchased from Sigma Chemical (St. Louis, MO): neutral sphingomyelinase (EC 3.1.4.12) from Bacillus cereus, phosphorylcholine chloride, ACh HCl, EGTA, staurosporine, bisindolylmaleimide I, and verapamil hydrochloride. BAPTA-AM and fura 2-AM were purchased from Molecular Probes (Eugene, OR). N-octanoyl-D-erythrosphingosine (C₈ ceramide), ML-9, and N-palmitoyl-D-erythrosphingosine (C₁₆ ceramide) were obtained from Biomol (Plymouth Meeting, PA). Nervonic ceramide (C_24:1), lignoceric ceramide (C_24:0), C₆ ceramide, and C₈ ceramide-1-phosphate were obtained from Cayman Chemical (Ann Arbor, MI). The following specific pharmacological antagonists were also used: atropine sulfate (Mann Research, New York, NY), propranolol hydrochloride, nimodipine, and diphenhydramine hydrochloride (Calbiochem, La Jolla, CA); phentolamine methanesulfonate (CIBA Pharmaceutical, Summit, NJ); cimetidine hydrochloride (SmithKline Beecham, Philadelphia, PA); methysergide maleate (Sandoz Pharmaceuticals, Hanover, NJ); naloxone hydrochloride (Sigma), and indomethacin (Merck, Rahway, NJ). All other organic and inorganic chemicals were obtained from Fisher Scientific (Fair Lawn, NJ). These were commercial products of the highest grade available.

Statistical analyses. Where appropriate, results are expressed as means ± SE. Differences between means were analyzed using unpaired t-tests or ANOVA followed by a Newman-Keuls test. Statistical significance was assumed when P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

N-SMase induces endothelium-independent contraction in canine isolated cerebral arterial rings. Addition of N-SMase (0.001-0.1 U/ml) to organ chambers containing resting endothelium-intact and -denuded canine cerebral arterial rings provoked a gradual but sustained constriction in a dose-related manner and reached a maximum at 0.1 U/ml (Fig. 1, A1 and A2). There were no significant differences between rings with endothelium and those without (P > 0.05), suggesting that the contractile effects exerted by N-SMase on the vessels appear to be endothelium independent. The average maximal tension developed in endothelium-intact rings was 1 ± 0.2 g (n = 13), whereas that in endothelium-denuded rings was 1 ± 0.3 g (n = 13).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Typical recordings of contractile responses to neutral sphingomyelinase (N-SMase; 0.001-0.1 U/ml) in canine cerebral isolated arterial rings were obtained in absence (control) (A1) or presence of endothelium (control) (A2). B: response to N-SMase (0.1 U/ml) in denuded, isolated canine cerebral arterial rings exposed in Ca2+-free solution. C: response to N-SMase (0.1 U/ml) in denuded, isolated rings exposed to 5 × 105 M verapamil. Responses to N-SMase (0.1 U/ml) were also determined in denuded isolated canine cerebral arterial rings exposed to protein kinase C (PKC) inhibitors 5 × 106 M staurosporine (D1) or 5 × 106 M bisindolylmaleimide I for a 20-min period before N-SMase administration (D2), respectively. E: response to N-SMase (0.1 U/ml) in denuded isolated canine cerebral arterial rings exposed to Ca2+-free solution for a 20-min period before N-SMase administration. Bisindo, bisindolylmaleimide I; SMase, sphingomyelinase.

Involvement of Ca²⁺ in N-SMase-induced contractions. To assess the role of external Ca²⁺ in N-SMase stimulation of canine cerebral arterial rings devoid of endothelium, we examined the effect of Ca²⁺-deficient solution on N-SMase-evoked vasoconstriction. Incubation of arterial rings for 45-min periods in Ca²⁺-free solution containing 2.0 mM EGTA produced significant inhibition of N-SMase-induced constriction. The N-SMase-evoked constriction was reduced to 21.3 ± 9.4% of control (Figs. 1B and 2). When denuded canine cerebral arterial rings were pretreated for 20 min with verapamil (5 × 10⁵ M), subsequent N-SMase-induced constriction was attenuated markedly (Fig. 1C). Pretreatment with verapamil caused maximal N-SMase-induced constriction to be reduced to 46.2 ± 4.2% of control (Fig. 2). Although not shown, similar results were obtained with nimodipine. However, exposure to 5 × 10⁵ M BAPTA-AM, used for buffering the intracellular Ca²⁺ in canine cerebral arterial smooth muscle, for 20 min did not reduce significantly (data not shown, n = 6) the subsequent N-SMase-induced contractions. When verapamil (5 × 10⁵ M) and BAPTA-AM (5 × 10⁵ M ) were added together to the bath medium, the contractile effects of N-SMase on the denuded canine cerebral arterial rings were inhibited to 50.1 ± 8.5% of control, approximately that of verapamil alone (data not shown, n = 6). These results confirm that the N-SMase-induced contractions in canine cerebral arterial smooth muscle are in large measure [Ca²⁺]_o dependent.

PKC antagonists attenuate N-SMase-induced canine cerebral arterial smooth muscle contractions. The purpose of these experiments was to determine whether the N-SMase-induced contraction is dependent on activation of PKC. The PKC inhibitors staurosporine (5 × 10⁶ M) or bisindolymaleimide I (5 × 10⁶ M) were added 20 min before N-SMase was added to the denuded canine cerebrovascular rings. We found that both staurosporine and bisindolymaleimide I reduced maximal N-SMase-induced contractions to 66.2 ± 7.2% and 72.1 ± 5.5% of control, respectively (Fig. 1, D1 and D2, and Fig. 2). These concentrations of staurosporine and bisindolymaleimide I, in these tissues, exert no effect on resting [Ca²⁺]_o influx (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Cumulative contractile concentration responses to N-SMase (0.001-0.1 U/ml) in presence and absence (control) of Ca2+-free solution and in presence of the Ca2+ antagonist verapamil (5 × 105 M), the PKC inhibitor staurosporine (5 × 106 M), and Ca2+-free medium plus staurosporine on canine isolated cerebral arterial segments. Each point represents mean ± SE expressed as a percentage of maximal tension developed to N-SMase (0.001-0.1 U/ml) as control.

Effects of staurosporine with Ca²⁺-free solution on N-SMase-induced vasoconstriction. Incubation of denuded canine cerebral arterial rings in Ca²⁺-free solution (2.0 mM EGTA) together with staurosporine (5 × 10⁶ M) for 20 min almost completely inhibited the vasoconstriction induced by N-SMase (to ~10% of control, Figs. 1E and 2).

Effect of ML-9 on N-SMase-induced canine cerebral arterial smooth muscle vasoconstriction. The hypothesis of this investigation is that increases of [Ca²⁺]_i and activation of MLCK are important for the initiation of vascular smooth muscle contraction. The specific protein kinase inhibitor ML-9 (an inhibitor of MLCK) was used. However, preincubation of denuded cerebral arterial rings with ML-9 (1 × 10⁵-5 × 10⁵ M) for 20 min did not result in any inhibition of N-SMase (0.1 U/ml)-induced contraction but significantly inhibited KCl (80 mM)- and 5-hydroxytryptamine (10 µM)-induced contractions (data not shown, n = 4).

Effects of C₆ ceramide, C₈ ceramide, C₁₆ ceramide, lignoceric ceramide, C₈ ceramide-1-phosphate, and phosphorylcholine on canine cerebral vascular smooth muscle tone. Because SMase catalyzes the breakdown of sphingomyelin to form ceramides and phosphorylcholine, we tested the effects of C₈ ceramide, C₁₆ ceramide, phosphorylcholine, and other analogs on basal tone in canine isolated and denuded cerebral arterial smooth muscle, respectively. Both C₈ ceramide (10⁹-10⁵ M) and C₁₆ ceramide (10⁹-10⁵ M) constricted denuded canine cerebral arterial rings in a concentration-related manner (Fig. 3). Maximal constriction was produced by 10⁶ M C₈ ceramide and 10⁶ M C₁₆ ceramide; the mean maximal tensions were 0.6 ± 0.1 and 0.5 ± 0.1 g, respectively (n = 6). These maximal tensions are, thus, each ~50-60% of N-SMase maximally induced tensions. However, exposure to phosphorylcholine (10⁹-10⁵ M), C₆ ceramide, nervonic (C_24:1) ceramide, lignoceric (C_24:0) ceramide, and C₈ ceramide-1-phosphate elicited only a slight relaxation in canine denuded cerebral arterial rings, not contraction (data not shown, n = 4).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Typical cumulative contractile concentration responses to C8 ceramide (109-106 M) and C16 ceramide (109-106 M) in canine cerebral isolated arterial rings in absence of endothelium.

Effects of N-SMase, C₈ ceramide, and C₁₆ ceramide on [Ca²⁺]_i mobilization in cultured canine cerebral arterial smooth muscle cells. Because Ca²⁺ plays an important role in the regulation of smooth muscle contraction, we examined the effect of N-SMase on [Ca²⁺]_i mobilization. N-SMase (0.1 U/ml) caused a gradual and sustained increase in [Ca²⁺]_i after a lag time of 5 min in fura 2-loaded, cultured canine cerebral arterial smooth muscle cells (Fig. 4A); although not shown, smaller concentrations of N-SMase (i.e., 0.001-0.01 U/ml) produced smaller concentration-related increases in [Ca²⁺]_i. Addition of N-SMase (0.1 U/ml), C₈ ceramide (10⁶ M), and C₁₆ ceramide (10⁶ M) to cultured canine single cerebral vascular smooth muscle cells caused a significant rise in [Ca²⁺]_i from a mean resting level of 79.5 ± 7.2 nM to 129 ± 5.8, 116 ± 9.5, and 113.3 ± 4.7 nM, respectively (Table 1). To evaluate the sources of activator Ca²⁺ in canine cerebrovascular smooth muscle cells in response to N-SMase, we examined the effectiveness of [Ca²⁺]_i in the presence of nimodipine (5 × 10⁵ M) and Ca²⁺-free solution containing 1.0 mM EGTA, respectively. Exposure of cultured canine cerebral vascular smooth muscle cells to Ca²⁺-free solution showed that the elevation of [Ca²⁺]_i triggered by N-SMase observed in the Ca²⁺-depleted cells reduced the [Ca²⁺]_i rise to almost approximately the level at the resting state (Fig. 4B). We found that pretreatment with nimodipine (5 × 10⁵ M) for 5 min caused a steady decline in [Ca²⁺]_i induced by N-SMase (Fig. 4C). Moreover, exposure of cerebral vascular smooth muscle cells containing N-SMase to Ca²⁺-free solution together with nimodipine (5 × 10⁵ M) failed to elicit a response. Although not shown, similar results were obtained with verapamil. However, after 2.5 mM Ca²⁺ was added, the response was very rapid, reaching nearly maximal levels within 5 s, close to a threefold increase in the intracellular Ca²⁺ basal value observed; it then fell to slightly above the basal level (Fig. 5). In contrast, without N-SMase present, reintroduction of 2.5 mM Ca²⁺ increased the [Ca²⁺]_i more than three times the basal level (Fig. 5). Interestingly, after 2.5 mM Ca²⁺ was added to canine cerebral vascular smooth muscle cells with N-SMase in Ca²⁺-free solution, the [Ca²⁺]_i exhibited a rapid and large increase (>3-fold) and then fell to a level similar to that mentioned above, followed by a secondary increase within 5 min. The latter is most likely the phenomenon of Ca²⁺-induced Ca²⁺ release (22). These results suggest that Ca²⁺ probably enters the cells through L-type (and possibly additional types) of Ca²⁺ channels in the plasma membrane, which causes the rise of [Ca²⁺]_i in canine cerebral arterial smooth muscle cells in response to N-SMase stimulation.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Effects of N-SMase (0.1 U/ml) on intracellular Ca2+ concentration ([Ca2+]i)-induced rises in single primary canine cerebral vascular smooth muscle cells. A: rise in [Ca2+]i in single primary canine cerebral vascular smooth muscle cells exposed to N-SMase (0.1 U/ml) and obtained 5-30 min after N-SMase administration was added. B: effects of N-SMase (0.1 U/ml) in Ca2+-free normal Krebs-Ringer bicarbonate solution, with 1.0 mM EGTA added, on rises in [Ca2+]i in single primary canine cerebral vascular smooth muscle cells. C: effects of N-SMase (0.1 U/ml) after 5-min exposure to nimodipine (5 ×105 M). Traces are representative examples of typical responses of single cells from at least 10 similar, separate experiments under various assay conditions.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Changes in cytoplasmic free Ca2+ concentrations under Ca2+-free conditions with N-SMase (0.1 U/ml) and nimodipine (5 × 105 M). Quiescent, cultured canine single cerebral vascular smooth muscle cells were exposed in Ca2+-free HEPES buffer containing 1 mM EGTA and N-SMase () to either nimodipine and SMase () or placebo control (). At indicated times, 2.5 mM Ca2+ was added. In each case, Ca2+ changes, determined using fura 2 imaging, of 4 responsive cells are shown.

Failure of specific pharmacological antagonists to interfere with the N-SMase- and ceramide-induced contractions. Incubation of canine cerebral arterial rings with various amine antagonists (i.e., phentolamine, propranolol, diphenhydramine, cimetidine, atropine, or methysergide, n = 6 each), opiate antagonists (naloxone, n = 6), and cyclooxygenase antagonists (i.e., indomethacin, n = 6) for 20 min each, before addition of the lipid agonists, all failed to either attenuate or interfere with contractions induced by N-SMase and C₈ or C₁₆ ceramide (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

It has been established that numerous vasoconstricting agonists elevate [Ca²⁺]_i and induce tension development in cerebral vascular smooth muscle cells by inducing both Ca²⁺ influx from the extracellular space and Ca²⁺ release from intracellular stores (1). These results are the first demonstration that the key initiating enzyme in the sphingomyelin pathway, N-SMase, can directly elicit a gradual and sustained vasoconstriction in isolated canine cerebral arterial rings as well a slow, sustained increase in [Ca²⁺]_i in cultured canine cerebral vascular smooth muscle cells.

In the present study, Ca²⁺ depletion, in the presence of 2.0 mM EGTA, significantly reduced the N-SMase contractions to ~20% of control (Fig. 2). A similar suppression of the N-SMase response was observed after pretreatment with the organic Ca²⁺ channel blockers verapamil and nimodipine. Moreover, exposure of the cerebral vascular smooth muscle cells to Ca²⁺-free solution reduced the N-SMase elevation of [Ca²⁺]_i to approximately baseline (Fig. 4B). It has been shown that only L-type Ca²⁺ channel currents have been identified in isolated myocytes from canine middle cerebral arteries (2). These observations inevitably led us to propose that N-SMase stimulation (independent of release of amines, opiates, or prostanoids) of canine cerebral vascular muscle promotes mainly an extracellular Ca²⁺ influx, probably through L-type Ca²⁺ channels, to raise the cytosolic [Ca²⁺]_i level, thus resulting in vasoconstriction. How, indeed, at the molecular level, does N-SMase modulate this enhancement of L-type Ca²⁺ channels? Our studies were not designed to answer this important question. Interestingly, N-SMase-induced constriction of canine cerebral arterial smooth muscle was partially resistant to inhibition by high concentrations of either verapamil or nimodipine (Fig. 2). Moreover, it is clear that attenuation of the rise of [Ca²⁺]_i, in the presence of nimodipine pretreatment, was less than that of Ca²⁺-deficient medium alone (Fig. 4C). This suggests that the nimodipine-resistant contraction could be due to Ca²⁺ uptake through nimodipine-insensitive Ca²⁺ channels, Ca²⁺ release from intracellular stores, or other unknown mechanisms. In the present study, N-SMase produced some contraction in Ca²⁺-deficient solution, and the increase in [Ca²⁺]_i induced by N-SMase in the Ca²⁺-free medium was slightly above the basal level. It is possible that the internal Ca²⁺ release caused by N-SMase may be a result of Ca²⁺ entry through plasma-membrane Ca²⁺ channels to produce the phenomenon of Ca²⁺-induced Ca²⁺ release (22). However, utilization of a high concentration of BAPTA-AM, which presumably would bind [Ca²⁺]_i released by internal stores, failed to inhibit the N-SMase contractions. The residual proportion of the N-SMase contractions (21% of control), elicited in the presence of 2.0 mM EGTA, is more probably via a mechanism independent of [Ca²⁺]_o entry. The possibility must also be entertained that N-SMase may in some way alter myofilament Ca²⁺ sensitivity.

To clarify the possible involvement of PKC activation in the N-SMase-induced contraction, we used two PKC inhibitors, staurosporine (5 × 10⁶ M) and bisindolylmaleimide I (5 × 10⁶ M). It has been suggested that staurosporine, which interacts with the ATP-binding site of PKC that shares substantial homology with other protein kinases, may be a nonselective PKC antagonist (27). Nevertheless, bisindolylmaleimide I, which is known to inhibit most isozymes of PKC without affecting other protein kinases, is a more specific PKC inhibitor (31). In the present study, both of these antagonists inhibited, significantly, N-SMase-induced contractions (Figs. 1 and 2). These results suggest that PKC activation may, indeed, play a role in the N-SMase-induced contractions. Furthermore, pretreatment with staurosporine (5 × 10⁶ M) for 20 min, in Ca²⁺-free solution, caused a greater degree of inhibition of N-SMase-induced contraction (89.5 ± 9.4%) than in the case of Ca²⁺ depletion alone (P < 0.05). These results could be used to suggest a probable involvement of a Ca²⁺-independent PKC isozyme in activation of the Ca²⁺-independent part of the N-SMase-induced contraction in isolated canine cerebral vascular smooth muscle.

It has been shown that treatment of vascular smooth muscle cells by B. cereus SMase can induce an intense incorporation of [³H]thymidine, which is associated with an extensive hydrolysis of radiolabeled sphingomyelin and a concomitant production of ceramide (6). N-SMase treatment of mouse epidermal (HEL-37) and human skin fibroblast (SF3155) cells, preincubated with [³H]serine to label the sphingomyelin pool, causes an accumulation of labeled ceramide but not sphingosine or ceramide-1-phosphate (16). Other recent studies indicate that ceramide levels in vascular smooth muscle are clearly Mg²⁺ dependent (5, 23). In contrast to ceramide exposure, exposure of canine denuded cerebral arterial rings to phosphorylcholine (10⁹-10⁵ M), the other major sphingomyelin metabolite, resulted in a slight relaxation (data not shown, n = 4), not contraction, of these cells. Furthermore, it has been shown that addition of sphingosine (3 × 10⁶ M) to canine cerebral arterial rings alone provokes a gradual relaxation, not constriction (unpublished data), suggesting that the effects of N-SMase are independent of sphingosine or phosphorylcholine generation in this study. These new data thus reasonably led us to suggest that the action of N-SMase, in this study, was caused by the accumulation and activation of a ceramide molecule(s) via the hydrolysis of sphingomyelin in response to SMase.

Recent studies from our laboratory have demonstrated that another ceramide analog alone, i.e., C₂ ceramide, did not elicit any significant changes in either basal tension or the resting level of [Ca²⁺]_i in canine cerebral vascular muscle. C₂ ceramide, however, attenuates PGF₂-induced contractions in canine cerebrovascular muscle rings and the secondary rises of [Ca²⁺]_i evoked by PGF₂ in cultured canine cerebrovascular smooth muscle cells (39); similar results were obtained with C₂ ceramide in rat aortic smooth muscle (40). Interestingly, neither basal tension nor resting levels of [Ca²⁺]_i were changed by N-SMase in rat aortic smooth muscle cells, thus suggesting differences between peripheral and cerebral vascular muscle cells with respect to actions of N-SMase. N-SMase can attenuate phenylephrine-induced contraction as well as inhibit the elevation in [Ca²⁺]_i in cultured rat aortic smooth muscle cells (unpublished findings; 41). To investigate what seems to be at first glance paradoxical, we examined the effects of C₈ ceramide and C₁₆ ceramide on basal tone as well as Ca²⁺ mobilization in isolated canine cerebral arterial smooth muscle. We found that both C₈ ceramide and C₁₆ ceramide caused cerebral arterial ring contractions and increases of [Ca²⁺]_i, respectively.

C₈ ceramide is a biologically active, cell-permeable, but nonphysiological ceramide analog, and C₁₆ ceramide is an abundant molecular species of endogenous ceramide. Both are known to induce phosphorylation on Thr-669 in A-431 cells by stimulation of ceramide-activated protein kinase (20). Some evidence suggests that there may be structural differences between the ceramide molecules produced on sphingoid-signaling activation, depending on the subcellular site of production and/or the composition of the sphingoid precursor (25). In addition, in the same cell, distinct pools of sphingoid signal molecules can be generated after stimulation; an increase of the same molecules in specific compartments (so far documented only for ceramide) has been reported to be responsible for the mediation of different signaling cascades, leading to different cellular effects (25, 32). Moreover, because smooth muscle cells are locally derived from individual organ parenchyma during embryogenesis, smooth muscle cells in different arteries may in fact be different and may respond differently to the same agonists (26, 30), explaining in part why aortic smooth muscle cells respond differently to N-SMase. Clearly, further studies are needed to clarify this puzzling and exciting issue.

In summary, N-SMase can constrict isolated canine cerebral arterial rings and elevate [Ca²⁺]_i in cultured canine cerebrovascular smooth muscle cells in a concentration-related manner. The constriction by N-SMase was blocked only partially by L-type Ca²⁺ channel blockers or PKC inhibitors (i.e., bisindolylmaleimide I, staurosporine) and was markedly inhibited by Ca²⁺-free medium. N-SMase-induced contractions were found to be nearly completely eliminated by the addition of staurosporine to a Ca²⁺-free solution. Both C₈ ceramide and C₁₆ ceramide, but not phosphorylcholine, C₆ ceramide, C₈ ceramide-1-phosphate, nervonic ceramide, or lignoceric ceramide, were found to constrict canine isolated, denuded cerebral vascular smooth muscle in a concentration-dependent manner. The action of N-SMase is probably caused by the accumulation and activation of ceramide molecules via the hydrolysis of sphingomyelin in response to sphingomyelinase.