INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

DIETARY DEFICIENCY OF MAGNESIUM as well as abnormalities in Mg²⁺ metabolism have been suggested to play important roles in hypertension, stroke, atherosclerosis, and diabetic vascular disease (for recent reviews, see Refs. 1, 3, 5, 10, 22, 34). During the past five or six years, using specific Mg²⁺-selective electrodes that were pioneered in our laboratory, we have demonstrated that patients with severe untreated essential hypertension and stroke exhibit lowered levels of serum-ionized Mg²⁺ (1, 3, 8, 9). Such low defined serum-ionized extracellular Mg²⁺ concentration ([Mg²⁺]_o) levels result in a rapid concentration-dependent rise in intracellular Ca²⁺ concentration ([Ca²⁺]_i) in cultured cerebral arterial smooth muscle concomitant with contraction of these isolated primary vascular smooth muscle cells (8).

Ca²⁺ is a major determinant of contractile force in all types of muscles. The initiation of contraction in vascular smooth muscle is believed to derive from a rise of free cytosolic [Ca²⁺]. There are two ways to raise [Ca²⁺]_i: 1) Ca²⁺ can enter smooth muscle cells through Ca²⁺ channels in the plasma membrane, such as voltage-gated Ca²⁺ channels; and 2) Ca²⁺ can be released from intracellular stores either by D-myo-inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] or by Ca²⁺-induced Ca²⁺ release through the ryanodine receptor (23). Protein kinase C (PKC) and the thin filament-associated proteins have been considered to contribute in several ways to contractile regulation in vascular smooth muscle (16, 21), including a possible regulation of [Ca²⁺]_i and Ca²⁺ channels. Recently phosphatidylinositol-3 kinases (PI3Ks), which phosphorylate phosphatidylinositol 4,5-bisphosphate (PIP₂) to phosphatidylinositol 3,4,5-trisphosphate (PIP₃), have received more attention because PIP₃ and PI3Ks have also been suggested to act as second messengers (31).

The myogenic tone of cerebral vessels is strongly dependent on the plasma concentration of [Mg²⁺]_o (4, 6, 7) and is believed to play an important role in the autoregulation of cerebral blood flow (7, 11). Recently, several clinical trials have demonstrated the therapeutic usefulness of administration of Mg²⁺ in the treatment of stroke (26). However, the mechanism whereby low [Mg²⁺]_o alters cerebral arterial vascular tone may be complex, and whether the vascular action of Mg²⁺ deficiency is associated with specific intracellular signal transduction pathways is less well defined. This prompted the present study to gain insight into the relationship between [Mg²⁺]_o deficiency-induced contractions, Ca²⁺ influx, intracellular Ca²⁺ release, and potential intracellular signaling pathways, such as PKC and PI3Ks.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

General procedures. Rings of canine basilar arteries were obtained from male mongrel dogs (18-22 kg) after administration of pentobarbital sodium anesthesia (40 mg/kg iv) and were placed in normal Krebs-Ringer bicarbonate 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 (6). The rings were 3-4 mm in length. The segments were mounted on stainless steel pins under 2 g resting tension in isolated organ baths, attached to force transducers (Grass model FT 03) and connected to polygraphs (Grass model 7). The organ baths containing normal Krebs-Ringer bicarbonate 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 collection. At the beginning of an experiment, rings were exposed for 30-45 min to 80 mM KCl, and this was repeated every 30-45 min until responses were stable (2-3 times). Successful removal of the endothelium was assessed by showing that ACh (10⁸-10⁶ M) failed to relax segments precontracted by 0.2 µM phenylephrine, while it did relax the endothelium-intact segments (44, 45). When tissues were pretreated by various drugs, the drug was applied for at least 15 min before the concentration-response curves were obtained.

6

6

6

6

Intracellular Ca²+ measurement. Primary smooth muscle cells from canine basilar arteries for image analysis experiments were seeded on glass coverslips (12-mm diameter; ~1 × 10⁴ cells per coverslip) and used 2-3 days postseeding (24, 43-45). Monolayers of the smooth muscle cells grown on the coverslips were loaded with 2.0 µM fura 2-acetoxymethyl ester (AM) and 0.12% pluronic acid F-127 (60 min, 37°C), and the experimental procedures for [Ca²⁺]_i measurements were carried out as described previously using fura 2-AM (24, 43, 44). The resulting images were then used to calculate [Ca²⁺]_i in smooth muscle cells. [Ca²⁺]_i was calculated according to the following equation (17)
where R is the fluorescence ratios of fura 2 obtained by dividing the 340-nm image by the 380-nm is the image, R_min is the minimum fluorescence ratios of fura 2 obtained in medium containing 0 mM Ca²⁺ plus 10 mM EGTA by dividing the 340-nm image by the 380-nm image, and R_max is the maximum fluorescence ratios of fura 2 obtained in medium containing 2.54 mM Ca²⁺ by dividing the 340-nm image by the 380-nm image. A dissociation constant (K_d) of 224 nM was used for the fura 2-Ca²⁺ complex (24, 43, 44). B is the ratio of fluorescence intensity of fura 2 to the Ca²⁺-fura-2 complex excited at 380 nm. Particular care was taken to minimize photobleaching of the dye. Experiments were carried out in total darkness, and exposure to excitation light was less than 2 s in all experiments.

Experiments with removal of extracellular Ca²+. For the extracellular Ca²⁺-free experiments, the canine basilar arterial ring segments were equilibrated in Ca²⁺-free normal Krebs-Ringer bicarbonate solution containing 0.2 mM EGTA for at least 90 min before initiation of the experiments.

Drugs. The following pharmacological agents were purchased from Sigma Chemical (St. Louis, MO): Bis HCl, ACh HCl, EGTA, heparin ammonium salt, propranolol HCl, ryanodine, and verapamil. Atropine sulfate was bought from MANN Research Laboratory (New York, NY). Cimetidine HCl and diphenhydramine HCl were received from Smith Kline and French Laboratories (Welwyn Garden City, Herts, UK). DMSO, thapsigargin (TSG), Gö-6976, and wortmannin were purchased from Calbiochem (La Jolla, CA). Phentolamine methanesulfonate was purchased from CIBA Pharmaceutical (Summit, NJ). Methysergide maleate was purchased from Sandoz Pharmaceuticals (Hanover, NJ). LY-294002 was purchased from Biomol Research Laboratories, (Plymouth Meeting, PA). All other organic and inorganic chemicals were obtained from Fisher Scientific (Fair Lawn, NJ) and were of the highest purity.

Calculations and statistical analysis. The contractile response (g), percentage of maximal KCl-induced contraction, and [Ca²⁺]_i were expressed as means ± SE of the mean. Statistical evaluation of the results was carried out via analysis using the Newman-Keuls test and ANOVA using Scheffé's contrast test. The results were considered significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

[Mg²+]_o deficiency-induced contractions and extracellular Ca²⁺ influx. In medium in which the external Ca²⁺ concentration ([Ca²⁺]_o) is zero, removal of [Mg²⁺]_o from the medium induces a smaller transient [Ca²⁺]_i peak in the cells, which returns to the baseline level quickly without any plateau phase compared with the rapid rise and sustained plateau in [Ca²⁺]_i in the presence of normal [Ca²⁺]_o (Fig. 1A). In the presence of [Ca²⁺]_o, preincubation with 5 × 10⁶ M verapamil (an antagonist of voltage-gated Ca²⁺ channels) for about 10 min almost diminishes completely the low [Mg²⁺]_o-induced [Ca²⁺]_i plateau and significantly reduces the transient [Ca²⁺]_i peak; calculated mean values for low [Mg²⁺]_o-induced [Ca²⁺]_i peaks in the absence of [Ca²⁺]_o and in the presence of verapamil are shown in Fig. 1B.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Intracellular Ca2+ concentration ([Ca2+]i) changes in single smooth muscle cells obtained from canine basilar arteries (A and B) and contractile responses of endothelium-denuded canine basilar arterial rings (C and D), induced by medium with low extracellular Mg2+ concentration ([Mg2+]o) and no [Mg2+]o, are modified by withdrawal of extracellular Ca2+ concentration ([Ca2+]o) and addition of verapamil (5 × 106 M). A and C: horizontal axis shows time (min). B and D: each point indicates the peak value that was graphed (not the plateau value) and represents the mean ± SE. n = 10-15 (B) and 6 experiments (D). *P < 0.01 and **P < 0.001.

6

[Mg²+]_o deficiency-induced contractions and intracellular Ca²⁺ release. Because ryanodine is a "use-dependent" Ca²⁺ release-blocking agent (36), caffeine (10² M) was used first in the smooth muscle cells from canine basilar arteries for an ~3-min period to open the ryanodine channels. Ryanodine (10⁵ M) was then incubated in the tissue baths alone for ~5 min before the subsequent application of Mg²⁺-free medium (in the presence of ryanodine). Caffeine produces a transient [Ca²⁺]_i peak (Fig. 2A). Addition of ryanodine slightly elevates the resting level of [Ca²⁺]_i and markedly suppresses both the transient peak and sustained plateau of [Ca²⁺]_i normally induced by medium containing no [Mg²⁺]_o (Fig. 2A, compare with Fig. 1A). An Ins(1,4,5)P₃-mediated Ca²⁺-release antagonist, heparin, was next tested in our study. As shown in Fig. 2A, preincubation of the smooth muscle cells with 2 mg/ml heparin can be seen to dramatically reduce the increment in [Ca²⁺]_i (both transient and stable phases). We then examined the effect of a selective antagonist of the sarcoplasmic reticulum Ca²⁺ pump, viz. TSG, on low-[Mg²⁺]_o medium-induced [Ca²⁺]_i increments. In the absence of external Ca²⁺, TSG (10⁷ M) produces a transient [Ca²⁺]_i rise (Fig. 2A). A subsequent challenge with a medium containing 0 mM [Mg²⁺]_o in the presence of TSG (under Ca²⁺-free conditions) fails to induce the [Ca²⁺]_i peak and produces a very small, time-shortened [Ca²⁺]_i plateau (Fig. 2A); calculated mean values for different [Mg²⁺]_o values are shown in Fig. 2B.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   [Ca2+]i changes in single smooth muscle cells obtained from canine basilar arteries (A and B) and contractile responses of endothelium-denuded canine basilar arterial rings (C and D), induced by medium containing low [Mg2+]o and no [Mg2+]o, are modified by heparin (2.0 mg/ml), caffeine (102 M) plus ryanodine (105 M), or thapasigagin (TSG, 107 M, in the absence of [Ca2+]o). A and C: horizontal axis shows time (min). B and D: each point represents the peak value (as in Fig. 1, B and D) and the mean ± SE. n = 12-18 (B) and 6 experiments (D). *P < 0.01 and **P < 0.001.

2

5

7

PKC antagonists attenuate [Mg²+]_o deficiency-induced contractions. As shown in Fig. 3, A and B, pretreatment of endothelium-denuded canine basilar arteries with Gö-6976 [a PKC-- and PKC-I-selective antagonist (20)] or Bis [a specific antagonist of PKC (12)] significantly attenuates low-[Mg²⁺]_o medium-induced contractions (both phasic and tonic components) in a concentration-dependent manner. The concentrations producing 50% of the maximal inhibitory effects (IC₅₀ values) for Gö-6976 and Bis are 9.48 ± 0.41 × 10⁷ M and 5.03 ± 0.19 × 10⁷ M, respectively. Mean values for low [Mg²⁺]_o-induced contractions in the presence of Bis and Gö-6976 are shown in Fig. 3C. After obtaining stable contractions of endothelium-denuded canine basilar arterial rings, which we induced using medium with no [Mg²⁺]_o, the cumulative addition of the IC₅₀ amount of either Gö-6976 or Bis to the organ bath results in a reversal of the endothelium-independent contractile responses, in a concentration-dependent manner (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Concentration-dependent inhibitory effects of protein kinase C (PKC) antagonists on contractile responses of endothelium-denuded canine basilar arterial smooth muscle induced by low-[Mg2+]o medium. A and C: antagonists Gö-6976 (106 M) and bisindolyalmaleimide I (Bis, 5 × 107 M) were preincubated for 15 min. A: horizontal axis shows time (min). B and C: each point represents the peak contractile response that was graphed (not the plateau value) and the mean ± SE. n = 8; #P < 0.05, *P < 0.01, and **P < 0.001.

Effects of PKC antagonists on low [Mg²+]_o-induced elevations in [Ca²⁺]_i. Figure 4, A and B, illustrate that preincubation of primary cultured smooth muscle cells from canine basilar arteries with either Gö-6976 or Bis effectively prevents both the low-[Mg²⁺]_o medium-induced rapid increment in [Ca²⁺]_i and the additional rise of [Ca²⁺]_i. Lower steady states and a loss of the rapid peak increment in [Ca²⁺]_i are then seen. Such inhibitory effects of these two antagonists display concentration-dependent effects. The IC₅₀ values for Gö-6976 and Bis for such attenuation of the increases in [Ca²⁺]_i are 7.68 ± 0.23 × 10⁷ M and 3.54 ± 0.12 × 10⁷ M, respectively, which is consistent with the reduced contractile responses induced by low-[Mg²⁺]_o medium under the same conditions. Mean peak [Ca²⁺]_i values obtained under different low-[Mg²⁺]_o conditions in the absence and presence of the antagonists are shown in Fig. 4C.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Concentration-dependent inhibitory effects of PKC antagonists on [Ca2+]i changes in single smooth muscle cells from canine basilar arteries induced by low-[Mg2+]o medium. A and C: antagonists Bis (5 × 107 M) and Gö-6976 (106 M) were preincubated for 15 min. B and C: each point represents the peak [Ca2+]i that was graphed (not the plateau value) and the mean ± SE. n = 12-15; #P < 0.05, *P < 0.01, and **P < 0.001.

PI3K antagonists attenuate [Mg²+]_o deficiency-induced contractions. Figure 5, A and B, illustrates that the presence of wortmannin or LY-294002 [both are selective antagonists of PI3K (30, 39)] attenuates contractile responses (both phasic and tonic components) of endothelium-denuded canine basilar arteries to low-[Mg²⁺]_o medium in a concentration-dependent manner. The calculated IC_5o values for wortmannin and LY-294002 are 8.85 ± 0.36 × 10⁷ M and 5.65 ± 0.23 × 10⁶ M, respectively. Mean values for varying concentrations of low [Mg²⁺]_o-induced contractions, in the presence of wortmannin or LY-294002, are shown in Fig. 5C. After achieving full contractile responses of the arterial rings to Mg²⁺-free medium, cumulative administration of 5 × 10⁷ M wortmannin or 10⁵ M LY-294002 brings about significant relaxation of the endothelium-independent contractions (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Contractile responses of endothelium-denuded canine basilar arteries to low-[Mg2+]o medium are modified by phosphatidylinositol-3 kinase (PI3K) antagonists. A and C: wortmannin (106 M) and LY-294002 (5 × 106 M) were preincubated for 15 min. B and C: each point represents the peak contractile response that was graphed (not the plateau value) and the mean ± SE. n = 8; #P < 0.05, *P < 0.01, and **P < 0.001.

PI3K antagonists attenuate [Mg²+]_o deficiency-induced rises in [Ca²⁺]_i. Figure 6A demonstrates that preincubation of the cells with 5 × 10⁷ M wortmannin or 10⁵ M LY-294002 effectively inhibits both the low-[Mg²⁺]_o medium-induced transient [Ca²⁺]_i peak and the secondary plateau of [Ca²⁺]_i (to lower steady states) of basilar arterial smooth muscle cells. The inhibitory effects of these two antagonists show concentration-dependent effects (Fig. 6B). The calculated IC₅₀ values for wortmannin and LY-294002 are 7.26 ± 0.18 × 10⁷ M and 4.14 ± 0.11 × 10⁶ M, respectively. Mean values for low [Mg²⁺]_o-induced contractile tension development in the presence of wortmannin or 10⁵ M LY-294002 are shown in Fig. 6C.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Concentration-dependent inhibitory effects of PI3K antagonists on [Ca2+]i changes in single smooth muscle cells from canine basilar arteries induced by low-[Mg2+]o medium. A and C: wortmannin (106 M) and LY-294002 (5 × 106 M) were preincubated for 15 min. B and C: each point represents the peak [Ca2+]i that was graphed (not the plateau value) and the mean ± SE. n = 12-15, #P < 0.05, *P < 0.01, and **P < 0.001.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The main objectives of the present study were to explore whether the contractile effects of low-[Mg²⁺]_o physiological salt solution on cerebral arteries may be in large measure mediated by Ca²⁺ influx, Ca²⁺ release, and activation of PKC and PI3Ks. We have recently found that low defined serum [Mg²⁺]_o levels result in a rapid concentration-dependent rise in [Ca²⁺]_i in cultured cerebral arterial smooth muscle concomitant with contraction of these isolated primary vascular smooth muscle cells (8). The low [Mg²⁺]_o values used herein (i.e., 0.3-0.6 mM) have been found recently in the serum of patients with stroke, hypertension, and ischemic heart diseases using new specific Mg²⁺-selective electrodes (1, 3, 8).

The principal trigger for contraction of vascular smooth muscle is an elevation in [Ca²⁺]_i. In the present study, medium containing no [Mg²⁺]_o produced two phases of Ca²⁺ movement in smooth muscle cells from canine basilar arteries: a transient Ca²⁺ peak and a prolonged Ca²⁺ plateau (Fig. 1A). The transient [Ca²⁺]_i peak, induced by medium free of external Mg²⁺, appears to be derived as a consequence of Ca²⁺ release from intracellular stores because it persists in the absence of extracellular Ca²⁺ and fails to occur after the store is emptied by TSG, a selective antagonist of the sarcoplasmic reticulum Ca²⁺ pump (Fig. 2A) (19). In addition, we have noticed that both Ins(1,4,5)P₃- and ryanodine-sensitive Ca²⁺ stores are mobilized by low-[Mg²⁺]_o medium, because the Ca²⁺ release can be significantly suppressed by both heparin, an Ins(1,4,5)P₃-sensitive Ca²⁺-release antagonist (Fig. 2A) (19), and ryanodine, a specific antagonist of ryanodine-sensitive Ca²⁺ release (36). This finding, as far as we are aware, is the first direct and detailed evidence for low [Mg²⁺]_o-induced intracellular Ca²⁺ release in cerebral vascular smooth muscle cells. The prolonged [Ca²⁺]_i plateau induced by low-[Mg²⁺]_o medium results clearly from Ca²⁺ influx, mainly through voltage-gated Ca²⁺ channels from the extracellular medium, because it disappears in the absence of extracellular Ca²⁺ and can almost be abolished by verapamil, an L-type voltage-gated Ca²⁺-channel antagonist.

The marked attenuation of the contractile actions of low-[Mg²⁺]_o medium on canine basilar arterial smooth muscle in the absence of extracellular Ca²⁺ or that found with employment of verapamil, heparin, caffeine plus ryanodine, or TSG (in the absence of extracellular Ca²⁺) (Fig. 1, C and D, and Fig. 2, C and D) is in agreement with and well supports the above-mentioned Ca²⁺ movements produced by low-[Mg²⁺]_o medium. Collectively, these findings indicate that influx of extracellular Ca²⁺ occurs mainly through voltage-gated Ca²⁺ channels, and that intracellular Ca²⁺ release from Ca²⁺ stores [both Ins(1,4,5)P₃ and ryanodine sensitive] is necessary for these low-[Mg²⁺]_o and [Mg²⁺]_o-free media-induced contractile responses. Our present findings are also well supported by previous investigations. It has been shown that essential hypertension, stroke, and ischemic heart disease patients exhibit deficits in serum ionized Mg²⁺ and demonstrate significant elevation in the serum ionized Ca²⁺-to-serum-ionized Mg²⁺ ratio, a sign of probable increased vascular tone and vasospasm (1, 5, 8). Exposure of primary cultured single canine cerebral vascular smooth muscle cells, rat aortic smooth muscle cells, and piglet single coronary arterial muscle cells to the low concentrations of serum-ionized Mg²⁺ found in the hypertensive, stroke, and ischemic heart disease patients, e.g., 0.3-0.48 mM, resulted in rapid elevation in cytosolic free Ca²⁺ (1, 5, 8, 9). Coincident with the rise in [Ca²⁺]_i, many of the primary single cerebral, aortic, and coronary vascular cells went into spasm (5, 8, 9). Extracellular Mg²⁺ has been shown previously by our group to inhibit Ca²⁺ influx at the vascular smooth muscle membrane (6, 38, 41, 44). Previously, extracellular Mg²⁺ has also been suggested to interfere with Ca²⁺ release from intracellular bound sites in vascular smooth muscle, but this early speculation lacked experimental evidence to support this hypothesis (2, 33, 44). The results of the present study add considerable support to this previously advanced tenet.

Mg²⁺ is known to be a noncompetitive antagonist of [³H]Ins(1,4,5)P₃ binding and Ins(1,4,5)P₃-induced Ca²⁺ release; saturating concentrations of Ins(1,4,5)P₃ release less Ca²⁺ from intracellular Ca²⁺ stores at higher concentrations of free Mg²⁺, and Mg²⁺ controls Ins(1,4,5)P₃-induced Ca²⁺ release by affecting both the binding of Ins(1,4,5)P₃ to its receptor sites and the release of Ca²⁺ via Ins(1,4,5)P₃-gated Ca²⁺ channels (40). It is more than likely that decrements in [Mg²⁺]_o would decrease the [Mg²⁺]_i in cerebral arterial smooth muscle cells (24, 43). The inhibitory effects of Mg²⁺ on Ins(1,4,5)P₃ binding and Ins(1,4,5)P₃-induced Ca²⁺ release would be thus attenuated or eliminated. In concert with this, Ins(1,4,5)P₃-induced Ca²⁺ release from intracellular storage sites in the cells would be elevated, and the smooth muscle cells would then be expected to undergo contraction.

The concept that PKC plays a pivotal role in tonic contraction of smooth muscle is relatively recent and comes from observations that phorbol esters, which are established activators of PKC, can produce slowly developing sustained contraction in a number of vascular tissues (16). Conventional PKCs, cPKCs (PKC-, PKC-I, PKC-II, and PKC-), require Ca²⁺, phospholipids, and diacylglycerol (or phorbol esters). PKC is thought to phosphorylate smooth muscle myosin and myosin light-chain kinase in vitro (21), increase the Ca²⁺ sensitivity of the contractile apparatus (14), and influence the sustained phase of agonist-induced contraction (16). An important observation presented herein is that Gö-6976 (a selective antagonist of PKC- and PKC-I isozymes) and Bis (a specific PKC antagonist) markedly attenuated the low [Mg²⁺]_o-induced contractile responses of canine basilar arterial segments, suggesting the probable involvement of PKC activation in such arterial contractions. This contention is supported by the IC₅₀ values found herein experimentally. The calculated IC₅₀ values for Gö-6976 and Bis were 9.48 ± 0.41 × 10⁷ M and 5.03 ± 0.19 × 10⁷ M, respectively. These values are only slightly higher than the reported inhibitory constant (K_i) values for Gö-6976 [2 × 10⁸ M (18)] and Bis [1.6-2.0 × 10⁸ M (37)] for 50% inhibition of PKC. Because the K_i values of these two antagonists were obtained at cellular levels and the IC₅₀ values herein of these two antagonists were obtained at bioassay and tissue levels, the latter should be consistent with the former.

The involvement of PKC in the low [Mg²⁺]_o-induced contraction pathway is reinforced by the present findings that in single smooth muscle cells from canine basilar arteries preincubated with PKC antagonists (Gö-6976 and Bis), medium that is [Mg²⁺]_o free produces slower and smaller increments in [Ca²⁺]_i, which suggests that both Ca²⁺ influx from the extracellular medium and Ca²⁺ release from intracellular stores were inhibited. Similarly, other investigators have demonstrated that: 1) inhibition of PKC activity with staurosporine or chelerythrine can inhibit availability and long opening of L-type Ca²⁺ channels in A7r5 cells (29); 2) inhibition of PKC activity blunts the relative increase in cytosolic free Ca²⁺ in rabbit afferent arterioles in response to ANG II (32); and 3) inhibition of PKC activity can inhibit, completely, the rise in [Ca²⁺]_i induced by alcohol in cerebral vascular smooth muscle cells (45). It is thus tempting to speculate that inhibition of PKC phosphorylation of Ca²⁺ channels in canine basilar arterial smooth muscle cell membranes prevents the necessary rise in [Ca²⁺]_i produced by [Ca²⁺]_o influx and intracellular Ca²⁺ release, thus promoting relaxation of low [Mg²⁺]_o-induced vascular contractions.

The growing importance of PI3Ks in signal transduction has been pointed out over the past three years (13). A number of proteins have been shown to be associated with PI3Ks following stimulation of cells generally as a result of tyrosine-phosphorylated proteins associating via the PI3K p85 SH₂ domains (15). We demonstrate herein for the first time that two potent antagonists of PI3Ks, wortmannin and LY-294002, significantly suppress low [Mg²⁺]_o-induced contractions in canine basilar arteries and the concomitant elevation of [Ca²⁺]_i in canine basilar single arterial smooth muscle cells, indicating the likely involvement of products of PI3Ks [e.g., PI(3,4)P₂ and PI(3,4,5)P₃] in such vessel contractions. This contention gains further support from the experimentally derived IC₅₀ values. The calculated IC₅₀ values obtained herein for wortmannin and LY-294002 were 8.85 ± 0.36 × 10⁷ M and 5.65 ± 0.23 × 10⁶ M, respectively, which are consistent with the reported K_i values of these two antagonists for PI3Ks (10⁸ M and 1.4 × 10⁶ M) (30, 39). Our present findings are consistent with and well supported by the previous studies for magnesium deficiency and some other vasoactive substances: 1) wortmannin attenuates the contraction of guinea pig gastric longitudinal smooth muscle induced by thrombin and epidermal growth factor (47); 2) the amplitude of contractions of the rat aorta induced by KCl, phenylephrine, and prostaglandin F₂ is decreased by wortmannin (35); and 3) addition of wortmannin or LY-294002 to HepG2 cells inhibits the release of intracellular Ca²⁺ induced by coactivation of phospholipase C and PI3K (31). We would, however, like to exercise a word of caution in that wortmannin, unlike LY-294002, has also been shown to be a potential inhibitor of myosin light-chain kinase. But it has recently been demonstrated that, in rat aortic smooth muscle, these PI3K inhibitors also attenuate low [Mg²⁺]_o-induced vasoconstrictions (42). It is noteworthy that activation of PI3Ks leads to synthesis of its products, at least one of which (e.g., PIP₃) has been pointed out to be a selective activator of PKC- (27). This may be another pathway by which low-[Mg²⁺]_o could activate PKC in cerebral vascular smooth muscle cells.

From all of the above, our data suggest that although deficiency of extracellular Mg²⁺ acts directly on vascular smooth muscle cell membranes, it can initiate contraction in canine basilar arterial smooth muscle cells via Ca²⁺ influx through voltage-gated Ca²⁺ channels, intracellular Ca²⁺ release [both Ins(1,4,5)P₃ and ryanodine sensitive], and activation of PKC and PI3Ks. The subsequent elevation in [Ca²⁺]_i in the vascular smooth muscle cells appears to play a vital role in such contractile responses of the vessels when exposed to Mg²⁺-deficient environments.

[Mg²⁺]_o deficiency-induced cerebral vasoconstriction and vasospasm in situ would be expected to hemodynamically lead to reduced brain blood flows, decreased tissue nutrition, decreased tissue oxygenation, and increased cerebral vascular resistance and potentially a stroke. In this context, our laboratory has recently shown that a rapid reduction in cerebral spinal fluid and brain [Mg²⁺] will result in vasospasms and rupture of cerebral microvessels in the intact living rat (4). There is a clear, documented, and growing shortfall in magnesium dietary intake, particularly among populations in Western world countries (see reviews, Ref. 5). Moreover, almost 100 stroke (both ischemic and hemorrhagic) patients have been reported recently to exhibit low levels of serum-ionized Mg²⁺ on admission to the emergency room (8). Since very early studies on the cerebral circulation, it has been suggested that cerebrovasospasm is a trigger in cerebral ischemia and stroke. However, the etiology of cerebrovasospasm remains elusive. The present studies could thus help shed some new light on the etiologies of cerebrovasospasm and diverse cerebral vascular disease states associated with low serum levels of Mg²⁺ and could be of considerable help in pinpointing potential avenues for pharmacological and therapeutic intervention, particularly in magnesium-deficient states.