INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

SINCE DEVINE ET AL'S FIRST DESCRIPTION (10) of sarcoplasmic reticulum morphology in vascular smooth muscle, many functions have been attributed to this organelle. The sarcoplasmic reticulum is specialized in Ca²⁺ storage, functioning both to relax smooth muscle by removing Ca²⁺ via sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) from the cytoplasm and activation by release through D-myo-inositol 1,4,5-trisphosphate and ryanodine receptors. Subsequently, more refined aspects of these general functions were discovered. Iino et al. (18) and Saida and van Breemen (37) showed there was calcium-induced calcium release where the Ca²⁺ entry signal is amplified by opening of the Ca²⁺-sensitive ryanodine channels in the superficial sarcoplasmic reticulum. This was later confirmed by Ganitkevich and Isenberg (11) by showing that sarcoplasmic reticulum depletion decreases Ca²⁺ signals accompanying opening of voltage-gated Ca²⁺ channels (VGCC). Another indirect mechanism potentially regulating vascular constriction is activation of Ca²⁺ entry through sarcoplasmic reticulum depletion (5, 8, 36, 42). On the other hand, van Breemen (44) showed that the superficial sarcoplasmic reticulum could also attenuate the VGCC-mediated Ca²⁺ signal by sequestering Ca²⁺ from a restricted subsarcolemmal space. This was made all the more plausible because of Somlyo and Franzini-Armstrong's (39) distinction between the superficial and deep sarcoplasmic reticulum. In fact, Somlyo and colleagues (10, 38, 41) showed specialized junctional areas where the two membranes are separated by only 10-15 nm.

In the 1980s, Bulbring and Tomita (4) suggested that localized release of Ca²⁺ could have a relaxing action on intestinal smooth muscle as seen during -adrenergic stimulation. This mechanism was established on a firm basis by Benham and Bolton (1), who showed localized activation of clusters of K⁺ channels by spontaneous Ca²⁺ release from the sarcoplasmic reticulum, also termed spontaneous transient outward currents (STOCs). More recently, Nelson et al. (30) provided direct evidence for such local Ca²⁺ release by imaging Ca²⁺ sparks. It is of particular interest that sparks more frequently occur peripherally and that there seem to be specific sites that function as hot spots (19) or frequent discharge sites (1, 35). This work led to the suggestion that, at least in small cerebral arteries, the main function of the sarcoplasmic reticulum is to relax the blood vessels as a consequence of sparks and or STOCs. This hypothesis fits well with the earlier evidence that resistance arteries are mainly activated by Ca²⁺ influx rather than by intracellular Ca²⁺ release (7, 30). In addition, this concept may have clinical relevance in that the activity and expression of Ca²⁺-activated K⁺ channels (K_Ca) are enhanced in genetically hypertensive rats. The enhanced Ca²⁺-sensitive K⁺ current through BK_Ca blockers may function as a compensatory mechanism to balance the enhanced activity of VGCC observed in vascular smooth muscle cells of another hypertensive rat model (24).

Considering all of these possible mechanisms of sarcoplasmic reticulum-mediated regulation, we were interested in determining the relative contributions of each of these in a conduit, cerebral blood vessel, the basilar artery. We approached this problem by pharmacologically manipulating Ca²⁺ pumps and channels in the sarcoplasmic reticulum and plasmalemma and by measuring force and intracellular Ca²⁺ concentration ([Ca²⁺]_i) in rabbit basilar arteries. We found that both activation of K channels by sarcoplasmic reticulum Ca²⁺ release and buffering of Ca²⁺ influx are important mechanisms in maintaining the basilar artery in a relaxed state. We found no evidence for store-operated channels (SOCs).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tissue preparation. Adult New Zealand White rabbits weighing 2.0-2.5 kg were killed by CO₂ asphyxiation followed by exsanguination from the carotid arteries. The brain was removed, immediately placed in cold physiological saline solution (PSS), and placed in the refrigerator. The basilar artery was dissected from the brain in cold PSS continually bubbled with 95% O₂-5% CO₂, cut into approximately four 2-mm segments, and mounted on a wire myograph (Multi Myograph 610M, Danish Myo Technology) with two 30-µm stainless steel wires ~2.5 cm in length. The tissues were constantly kept at 37°C and continually bubbled with 95% O₂-5% CO₂ in 5 ml of PSS in their respective chambers. The wire myograph was connected to a personal computer and data were recorded and analyzed via myodaq and myodata (LabView, National Instruments) software. To account for differences in length and contractility, graphs were made in Microsoft Excel and bar graphs show the percentage of high potassium (80 mM) contractions. Although many of the responses had an initial transient component, all increases in force were recorded at plateau levels and these values were used in calculation of the mean data, represented by bar graphs in Figs. 2-4. Error bars represent means ± SE. Results shown are representative of at least three experiments.

[Ca²+]_i measurement. Briefly, the rabbit basilar artery was inverted with forceps to expose the endothelial cell side, and the endothelium was removed by gently rubbing the inverted vessel on filter paper. The tissue was placed in cold HEPES in a specially designed chamber mounted on two stainless steel rods and stretched slightly. Loading solution was added to the chamber with the following composition: HEPES with 5 µM membrane-permeant fura 2-acetoxymethyl ester (fura 2-AM; 1 mM stock in DMSO) and 5 µM pluronic acid to facilitate loading for 2 h in the dark at room temperature. After incubation with fura 2, the tissue was washed two to three times with HEPES and left for ~20 min to remove excessive external fura 2-AM. The chamber, maintained at 37°C, was placed on the stage of an inverted microscope (Nikon Diaphot). The bottom of the basilar artery was exposed to alternating 340- and 380-nm (bandwidth 10 nm) ultraviolet light (1/s) that was passed through a 510-nm (bandwidth 40 nm) cut-off filter before acquisition by an intensified charge-coupled device camera (model 4093G, 4810 series). Fluorescence signals were recorded as digital image data using the software program Northern Eclipse by Empix Imaging Systems housed in a Pentium processor personal computer.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Thapsigargin (TG) concentration-response curve. A: two consecutive applications of 0 Ca2+/0.1 mM EGTA/10 mM caffeine (arrows). The second concentration of caffeine was applied after removing it from the solution by a 10-min washout in physiological saline solution (PSS) indicated by the slashed bars in the time scale. B: same as A but 1 µM TG was applied to all solutions. C: TG concentration-response curve constructed from the ratio of second caffeine responses in B divided by control caffeine responses over time in A. N-nitro-L-arginine methyl ester (L-NAME, 200 µM) and indomethacin (Indo, 10 µM) were added 30 min before the experiment. Note that 1-2 µM TG was found to completely deplete sarcoplasmic reticulum (SR) stores. SERCA, sarco(endo)plasmic reticulum Ca2+ ATPase.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Isometric tension traces in response to Ca2+-activated K+ channel (KCa) antagonist, iberiotoxin (IbTX), followed by application of TG in rabbit basilar arteries. L-NAME (200 µM) and Indo (10 µM) were included in the bathing solution. A, left: representative trace of at least 10 tissues from at least 3 different rabbits. Right, summary of data compared 80 mM K+ contractions with control for tissue variability. No significant additional increases in force after KCa blockade were observed, although the trend was a slight increase in force. B: fura 2-AM spectrofluorometry of inside-out rabbit basilar arteries denuded of endothelium. Left, IbTX is added first, followed by blockade with TG resulting in an increase in [Ca2+]i. Right, summary of data compared with increases from the initial baseline. *P < 0.05 resulting from a paired t-test. Results represent at least 5 different tissues from at least 5 different rabbits. F340/F380, fluorescence ratio.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   A: isometric tension traces in response to the SERCA antagonist TG followed by application of the large conductance KCa (BKCa) blocker IbTX in rabbit basilar arteries. L-NAME (200 µM) and Indo (10 µM) were included in the bathing solution. Left, representative trace of at least 19 tissues from at least 7 different rabbits. Note that even after the SR has been emptied with TG, IbTX additionally increases force. Right, summary of data compared 80 mM K+ contractions with control for tissue variability. **P < 0.005 resulting from an unpaired t-test. B: fura 2-AM spectrofluorometry of inside-out rabbit basilar arteries denuded of endothelium. Similar trends are observed compared with force measurements. Left, TG is added first, followed by blockade with IbTX resulting in an increase in [Ca2+]i. Right, summary of data compared with increases from initial baseline. *P < 0.05 resulting from a paired t-test. Results represent at least 5 different tissues from at least 5 different rabbits.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Ryanodine (Ry, 10 µM) produces trends similar to those observed with 2 µM TG. A, left: isometric tension measurements in rabbit basilar arteries show an increase in force with 10 µM Ry and an additional increase when 3 mM tetraethylammonium ion (TEA) was subsequently added. Results represent at least 5 tissues from at least 5 different rabbits. L-NAME (200 µM) and Indo (10 µM) were included in the bathing solution. Right, summary of results compared with 80 mM K+ contractions. *P < 0.05 resulting from an unpaired t-test. B: fura 2-AM spectrofluorometry in inside-out rabbit basilar arteries denuded of endothelium. Left, 10 µM Ry increases [Ca2+]i and additional blockade of BKCa further increases [Ca2+]i. Right, summary of [Ca2+]i data comparing increases from baseline and analyzed with a paired t-test. The more-specific blocker of BKCa, IbTX, has the same effect as TEA, indicating that BKCa is active without a functional SR.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Contributions of small conductance KCa (SKCa), intermediate conductance KCa (IKCa), and BKCa blockades to additional contraction after SERCA blockade. Blockade with TG increased force and additional blockade with IbTX additionally increased force. Apamin had no additional effects, whereas charybdotoxin (CTx) slightly increased force. Rabbit basilar arteries were pretreated with 200 µM L-NAME and 10 µM Indo. Results represent 8 tissues from 3 different rabbits.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Mn2+ quenching in an endothelium-denuded rabbit basilar artery. The tissue was kept in a 0 Ca2+ HEPES solution before the addition of Mn2+ (see MATERIALS AND METHODS). Fluorescence is measured at a wavelength of 360 nM (F360), the isobestic point of fura 2-AM. A: representative trace of F360 in an endothelium-denuded rabbit basilar artery. Note that the addition of TG and/or Ry had no effect on the slope indicating no additional entry of Ca2+. B: summary of F360 data constructed from the tangent of the slope during application of various agents. K+ (80 mM) increased the steepness of the slope and is included as a positive control. The tangent slope was measured within the first 2 min of high K+ application as it quenches fura 2-AM completely. Each bar represents at least 4 tissues from 3 different rabbits. *P < 0.05 resulting from an unpaired t-test.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Sarcoplasmic reticulum buffering in rabbit basilar arteries. First, 0 Ca2+/0.1 mM EGTA/caffeine was applied to deplete the Ry- sensitive store followed by washout of caffeine with 0 Ca2+/0.1 mM EGTA solution for at least 10 min. A: subsequent application of 80 mM K+ produced results as shown by the first trace. The second trace was treated similarly, except 2 µM TG was added to all solutions. B: 80 mM K+ responses of arteries after depletion with caffeine and with or without TG are superimposed on an expanded time scale. In the absence of TG (control), the increased tension due to depolarization is markedly slower compared with arteries in the presence of TG (+2 µM TG). L-NAME (200 µM) and Indo (10 µM) were also included in all solutions. Refer to MATERIALS AND METHODS for composition and concentrations of solutions. Note that in the presence of TG, tension increases more quickly than under control conditions. Results represent at least 5 tissues from 3 different rabbits. Slashed lines indicate a break in time for recovery of the tissue in PSS and subsequent SR depletion.

Solutions and drugs. PSS contained (in mM) 119 NaCl, 4.7 KCl, 2.5 CaCl₂, 1.17 MgSO₄, 1.18 KH₂PO₄, 24.9 NaHCO₃, and 11.1 glucose (bubbled continuously with 95% O₂-5% CO₂, pH 7.4 at 37°C). For high extracellular K⁺ solution (80 K-PSS), 75.3 mM NaCl was replaced with equimolar KCl to make the final concentration equal to 80 mM. HEPES-PSS contained (in mM) 140 NaCl, 5 KCl, 1.5 CaCl₂, 1 MgCl₂, 10 glucose, and 5 HEPES (pH 7.4 at 37°C). For high extracellular K⁺ solution (HEPES-80 K), 75 mM NaCl was replaced with equimolar KCl to make the final concentration equal to 80 mM. Experimental solutions used for the experiments illustrated in Fig. 7 consisted of Krebs solution with only 0.5 mM CaCl₂ and, in cases where 0 Ca²⁺ was used, no Ca²⁺ was added and 0.1 mM EGTA was included in the solution.

Statistics. Results are shown as means ± SE and are analyzed for significance by paired or unpaired t-tests where noted.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

To examine the contribution of the sarcoplasmic reticulum to Ca²⁺ activation of K_Ca, it was important to determine the maximal concentrations of thapsigargin required for depletion of the store. Therefore, a concentration-response curve for thapsigargin was determined by measuring depletion of the sarcoplasmic reticulum Ca²⁺ content as reflected by inhibition of a caffeine response (Fig. 1). In Fig. 1A, we show reproducible contractile responses to 10 mM caffeine that are initiated by Ca²⁺ release from the sarcoplasmic reticulum and are used in subsequent experiments as a control measure of sarcoplasmic reticulum Ca²⁺ content. The extent of sarcoplasmic reticulum depletion is measured by the ratio of the caffeine response after a 10-min exposure to thapsigargin (Fig. 1B) divided by the initial control response (Fig. 1A). By varying the concentrations of thapsigargin, a range of values for the percentage of emptying of the sarcoplasmic reticulum could be determined, and a concentration-response curve was constructed (Fig. 1C). From this curve, it was concluded that a 10-min exposure to 1 µM thapsigargin completely depleted the caffeine-sensitive store, and, therefore, this was the minimum concentration used in subsequent experiments.

To eliminate the endothelial contributions to the data obtained in subsequent experiments, tissues were exposed to 200 µM L-NAME [inhibits synthesis of nitric oxide (NO)] and 10 µM Indo (inhibits synthesis of prostacylin) for either 30 min before experimentation or incorporated into the initial bathing solution during the experimental time period. These concentrations were used because they eliminated relaxing responses of the basilar artery to 10 µM ACh and 10 µM bradykinin (data not shown). Other studies have shown that all endothelium-dependent relaxations in rabbit basilar artery can be blocked with NO synthase and prostacyclin inhibitors and that there is no indication for the action of an endothelium-derived hyperpolarizing factor other than NO or a prostanoid in this artery (25). Mechanical removal of the endothelium also inhibited vasodilatory responses to ACh and bradykinin, but these tissues were not used in these experiments because of compromised smooth muscle function.

After establishing a procedure for the complete depletion of the sarcoplasmic reticulum, we compared the effects of this procedure with direct blockade of large BK_Ca with IbTX. The rationale for this experiment is that if the only functional consequence of abolishing sarcoplasmic reticulum function is inactivation of BK_Ca, then the two procedures should be indistinguishable and nonadditive (3). Blockade of the BK_Ca channel with 100 nM IbTX added to the bath solution produced a sustained contraction presumably due to depolarization (Fig. 2A, left). Subsequent additive application of 2 µM thapsigargin caused further contraction indicating an action at sites other than BK_Ca (Fig. 2A, right). Figure 2B examines the Ca²⁺ signals measured in parallel with the above contraction experiments. IbTX alone caused an increase in [Ca²⁺]_i (Fig. 2B, left) and additional subsequent blockade of SERCA with thapsigargin increased [Ca²⁺]_i further (Fig. 2B, right). These results suggest that the sarcoplasmic reticulum regulation of vasoconstriction is more complex than exclusively regulating K_Ca activity.

In the reverse situation where SERCA was blocked initially followed by blockade of BK_Ca, we again would expect that the two procedures would be nonadditive if the sarcoplasmic reticulum is solely responsible for releasing Ca²⁺ toward the BK_Ca. Abolishing the sarcoplasmic reticulum function first with 2 µM thapsigargin resulted in an increase in tension (Fig. 3A) paralleled by an increase in [Ca²⁺]_i (Fig. 3B). When BK_Ca was then blocked in the absence of sarcoplasmic reticulum function, there were additional large increases in tension (Fig. 3A) and [Ca²⁺]_i (Fig. 3B). These additional increases induced by IbTX application reached values of significance in the case of force development and [Ca²⁺]_i (Figs. 3, A, right, and B, right). These results suggest that even though the sarcoplasmic reticulum is empty and SERCA is blocked, BK_Ca remain active. Therefore, BK_Ca must also be activated by Ca²⁺ from sources other than the sarcoplasmic reticulum.

When the previous experiments were repeated with 10 µM ryanodine in place of thapsigargin, the same trends were observed. Ryanodine alone increased force and subsequent application of 3 mM TEA additionally increased tension (Fig. 4A). [Ca²⁺]_i measurements confirmed that the increases in tension were due to increases in [Ca²⁺]_i. Subsequent addition of 3 mM TEA, after the sarcoplasmic reticulum was no longer functional, further increased [Ca²⁺]_i (Fig. 4B). These results were repeated with 100 nM IbTX, and the same trends were observed (data not shown). Therefore, regardless of the mode of inactivation of the sarcoplasmic reticulum, i.e., either with thapsigargin (via SERCA blockade) or ryanodine (via depletion due to opening of ryanidine receptors), there still seemed to be activation of BK_Ca independent of the sarcoplasmic reticulum.

Because TEA is a nonspecific blocker of all K_Ca, it was important to determine the relative contribution of each K_Ca channel when blocked with this agent. Therefore, the contribution of small conductance K_Ca (SK_Ca) was tested with 100 nM apamin, a selective inhibitor of SK_Ca (16). When added subsequently to treatment with ryanodine or thapsigargin and IbTX (blocks BK_Ca specifically), apamin had no additional effect (Fig. 5), suggesting that in the absence of a functional sarcoplasmic reticulum, SK_Ca do not play a significant role. Intermediate conductance K_Ca (IK_Ca) were also examined by the addition of 100 nM CTx after IbTX and apamin were present in the bathing solution. Although CTx blocks all K_Ca subtypes (16), its addition after blockade of SK_Ca with apamin and BK_Ca with IbTX would test specifically for the involvement of IK_Ca. Addition of CTx did appear to have a small effect (Fig. 5), and, therefore, IK_Ca may also contribute to the regulation of membrane potential (E_m) under these conditions. In addition, experiments performed with 3 mM TEA tended to show slightly larger effects than in the presence of 100 nM IbTX, which may be explained by a nonspecific action on IK_Ca (data not shown). Thus, although TEA is not entirely specific, we were confident that the major portions of the responses observed after depletion of the sarcoplasmic reticulum using ryanodine were due to activation of BK_Ca. This idea was supported by use of the specific BK_Ca blocker, IbTX, in experiments using thapsigargin.

The increases in [Ca²⁺]_i and force observed with sarcoplasmic reticulum-depleting agents or BK_Ca blockers were completely abolished by application of 1 µM nifedipine or 10 µM diltiazem. The effects of all sequences of SERCA and K_Ca blockers were completely reversed with VGCC blockers (data not shown). In addition, preincubation with 1 µM nifedipine for 20 min almost completely prevented a high K⁺ response (data not shown). These data support the notion that cerebral vessels rely solely on the VGCC as a source of Ca²⁺ as has been shown in smaller, resistance-sized cerebral vessels (7, 14).

To test for a possible contribution by SOC, 5 µM Mn²⁺ was added to a nominally Ca²⁺-free PSS and fura 2 fluorescence recorded at the isobestic wavelength of 360 nm. There was no change in slope when either 10 µM ryanodine and/or 2-5 µM thapsigargin were added to the superfusate, even though subsequent high K⁺ depolarization caused a fourfold increase in the rate of Mn²⁺ entry (Fig. 6). These data are consistent with the presence of VGCC and absence of SOC in the basilar artery of the rabbit.

Finally, it was demonstrated in these cerebral arteries that the sarcoplasmic reticulum is capable of actively buffering Ca²⁺ influx. All solutions for this experiment contained low CaCl₂ (refer to MATERIALS AND METHODS) to slow down the rate of force development. Figure 7 shows rate of force development when Ca²⁺ is restored and the K⁺ concentration is elevated in the superfusate of basilar arteries with a depleted sarcoplasmic reticulum. Before the increase in extracellular [K⁺] the sarcoplasmic reticulum was depleted by caffeine in 0 Ca²⁺ PSS and the caffeine then washed out for 10 min. Under these conditions, addition of thapsigargin to the high-K⁺ solution markedly enhances the rate of force development (Fig. 7B), presumably because it prevents Ca²⁺ uptake into the sarcoplasmic reticulum as it enters the smooth muscle cells through VGCC.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Major findings in this communication are 1) effects of thapsigargin or ryanodine and blockade of K_Ca on force development and [Ca²⁺]_i are additive; 2) depletion of sarcoplasmic reticulum does not enhance Mn²⁺ influx; and 3) inhibition of SERCA enhances force developed during high K⁺-induced depolarization. After SERCA is blocked with thapsigargin, K_Ca blockers additionally increased [Ca²⁺]_i and force, suggesting that K_Ca is at least partially activated in the absence of a functional sarcoplasmic reticulum. In addition, when K_Ca is blocked first, subsequent SERCA blockade still increases [Ca²⁺]_i and force. These data indicate that in the rabbit basilar artery 1) BK_Ca activity is not totally dependent on functional sarcoplasmic reticulum; and 2) SERCA lowers [Ca²⁺]_i at least in part by mechanisms other than regulation of E_m, most likely by buffering Ca²⁺ entry. It has been reported that the Ca²⁺ sensitivity of K_Ca is too low to respond to global [Ca²⁺]_i (12, 43) and, therefore, requires local [Ca²⁺]_i elevation by virtue of release of quanta of Ca²⁺ from the sarcoplasmic reticulum in the form of sparks (2, 30). Yet, after complete blockade of SERCA, which has been shown to eliminate sparks and STOCS (35, 45), K_Ca of the basilar artery is still activated. We postulate that, under control conditions, SERCA removes Ca²⁺ from a restricted subplasmalemmal space and that its blockade leads to accumulation of Ca²⁺ in this space (see Fig. 8). The portion of K_Ca that faces these restricted cytoplasmic regions would then be activated on SERCA blockade.

View larger version (27K): [in this window] [in a new window]   Fig. 8.   Schematic representation of the various SR-related mechanisms involved in regulation of membrane potential and generation of force in vascular smooth muscle. The arrows indicate Ca2+ movements. The dashed arrows follow a path that Ca2+ ions may take through a basilar artery smooth muscle cell. Note that Ca2+ may accumulate in the junctional space between the SR and plasma membrane where interaction with other Ca2+-sensitive channels and pumps may occur. VGCC, voltage-gated Ca2+ channels; NCX, Na+/Ca2+ exchanger; RyR, Ry receptor.

These observations were made in the rabbit basilar artery. Although this artery is not strictly speaking a resistance artery, it is an important conduit cerebral artery, which has many characteristics in common with cerebral resistance arteries. For example, vasoconstriction is completely dependent on Ca²⁺ influx through VGCC and the smooth muscle displays spontaneous activity. Smooth muscle cells isolated from the basilar artery of the rat have been used in the study of sarcoplasmic reticulum function in small, cerebral arteries (35). In addition, Omote and Mizusawa (32) have studied fluctuations in basilar artery myogenic responses generated by variable activity of VGCC. They proposed that this Ca²⁺ entry activated BK_Ca, leading to hyperpolarization of the membrane and, thus, providing negative feedback on the opening of Ca²⁺ channels. Similar mechanisms have been suggested in cerebral arteries from spontaneously hypertensive rats (33) and in rabbit mesenteric arteries (31). In addition, this latter study suggested that the endothelium does not contribute to myogenic activity because its removal did not affect the oscillations. This may further implicate a role for BK_Ca because there is some recent evidence suggesting that BK_Ca is not active in the endothelium (22).

We chose the rabbit basilar artery as an appropriate model for the investigation of the roles the sarcoplasmic reticulum plays in a cerebral artery. The main question to be resolved is: What is the function of the sarcoplasmic reticulum in a blood vessel in which force development is totally dependent on Ca²⁺ entry through VGCC? It appears unlikely that even agonist-induced contractions have a significant component of sarcoplasmic reticulum Ca²⁺ release because histamine contractions are almost completely blocked by nifedipine. In peripheral resistance arteries, norepherine-induced contractions are readily abolished by removal of extracellular Ca²⁺ (6). Ca²⁺ sequestration by the sarcoplasmic reticulum without interaction between the plasma membrane and sarcoplasmic reticulum would also not be able to explain the steady-state results obtained in this study because such a putative function would cease on saturation of the sarcoplasmic reticulum and, therefore, be transient. In 1995, Nelson et al. (30) proposed that in cerebral resistance arteries, the sole function of the sarcoplasmic reticulum is to hyperpolarize the plasma membrane by releasing Ca²⁺ in the vicinity of K_Ca. The authors based their conclusion on the observation that the effects of sarcoplasmic reticulum depletion and K_Ca blockade on diameter, [Ca²⁺]_i, and E_m were identical and nonadditive. Because they subsequently reported on spark activation of K_Ca in cerebral resistance vessels of other species, including humans, their hypothesis appeared to have universal validity. There is, indeed, an abundance of recent evidence suggesting the importance of Ca²⁺ sparks in regulating E_m and vascular tone specifically in resistance-sized vessels (21, 27, 28, 30) (for a review see Ref. 20). However, in the rabbit basilar artery, it appears that activation through Ca²⁺ sparks may not be the only way the E_m is regulated in these larger cerebral vessels.

The idea that localized Ca²⁺ release could cause relaxation though activation of K_Ca was first proposed by Bulbring and Tomita (4) to explain -adrenergic relaxation of intestinal smooth muscle (13). The existence of this mechanism was established by the studies of Benham and Bolton (1) showing that STOCs arose from transient Ca²⁺ releases by the sarcoplasmic reticulum. Nelson et al.'s elegant and extensive studies (30) of sparks and related events in cerebral resistance arteries have established their contribution as a feedback regulatory mechanism. However, evidence for other sarcoplasmic reticulum-mediated mechanisms have been reported for vascular smooth muscle, namely buffering of Ca²⁺ influx by the superficial sarcoplasmic reticulum and the activation of SOC by sarcoplasmic reticulum depletion. Clearly, the significant thapsigargin-induced increase in [Ca²⁺]_i blockage of BK_Ca by IbTX or TEA supports the involvement of either of these two latter mechanisms. In this instance, Ca²⁺ entry through VGCC would be enhanced by the depolarization caused by blockade of K_Ca.

To test for the activity of SOC in the rabbit basilar artery, we measured the effect of thapsigargin on Mn²⁺ quenching of intracellular fura 2-AM. The negative outcome of this experiment makes the SOC possibility unlikely, although it cannot be completely ruled out at this point. In some cells other than smooth muscle, Ca²⁺ release-activated channels have been described, which were highly selective for Ca²⁺ (17, 34). However, in our experiments, the effects of thapsigargin were completely blocked by nifedipine (at 1 or 10 µM), which is not known to inhibit SOC. In contrast, Mn²⁺ entry through VGCC was clearly observed. Additionally, it has been shown previously that either cadmium or diltiazem (both Ca²⁺ channel blockers) inhibited Mn²⁺ influx in aortic smooth muscle cells isolated from rats (24) and in venous smooth muscle of the rabbit (9), respectively.

Having thus exhausted all the other possible mechanisms, we conclude that the most likely candidate for explaining the actions of thapsigargin and ryanodine after blockade of K_Ca is inhibition of the Ca²⁺-buffering action of the superficial sarcoplasmic reticulum. This is supported by our observation that thapsigargin enhanced the rate of force development in response to Ca²⁺ entry through VGCC. In the experiment reported herein, an effect on E_m was ruled out by the large depolarization induced by the 80 mM K⁺ solution. This result would also not be explained by inhibition of calcium-induced calcium release, because this would have the opposite effect.

In addition, our results support the idea of Guia et al. (15) that in rabbit coronary artery smooth muscle cells, there may be coupling of VGCC and BK_Ca, which has been shown previously in hippocampal neurons (26). Ca²⁺ influx through VGCC may directly cause accumulation of Ca²⁺ in the junctional space and directly activate BK_Ca as demonstrated by whole cell patch-clamp and simultaneous [Ca²⁺]_i measurements in vascular smooth muscle cells (15). Guia et al. (15) showed that abolishing the sarcoplasmic reticulum function with CPA or ryanodine had no effect on the transient opening of BK_Ca. In fact, the transient stimulation of K_Ca they observed appeared to be independent of sarcoplasmic reticulum function. Our data are consistent with the notion that local Ca²⁺ entry may directly activate at least a fraction of BK_Ca.

This leads to the concept of the existence of cytoplasmic microdomains within the smooth muscle cell. In some areas of vascular smooth muscle, it has been observed that the relative distance of the sarcoplasmic reticulum from the plasma membrane is on the order of 20 nm (20). Therefore, colocalization and functional interactions of various channels and ion pumps within this small, restricted space seems probable. Although our data are consistent with those of Guia et al. (15), suggesting a direct functional interaction between VGCC and K_Ca, we cannot rule out the possibility that blockade of Ca²⁺ removal from a restricted space may lead to accumulation of Ca²⁺ and activation of K_Ca in the absence of Ca²⁺ sparks. Thus, if SERCA (located in the superficial sarcoplasmic reticulum) removes Ca²⁺ from the restricted cytoplasmic space (Fig. 8), it would not only decrease the flow of Ca²⁺ into the deeper cytoplasmic space, but it would also lower [Ca²⁺]_i in the vicinity of K_Ca. Blocking SERCA would then not only abolish sparks and their activation of K_Ca but, in addition, raise [Ca²⁺] near the plasma membrane, which would partially activate K_Ca. The model presented in Fig. 8, which proposes a closer interaction between sarcoplasmic reticulum Ca²⁺ release channels and K_Ca than between VGCC and K_Ca and, in addition, Ca²⁺ removal from a subplasmalemmal space, would explain all the data reported herein.

In conclusion, our results demonstrate for the first time that in an intact cerebral small artery, BK_Ca remains active in the absence of Ca²⁺ release from the sarcoplasmic reticulum and that the superficial sarcoplasmic reticulum buffers Ca²⁺ entry through VGCC. Additionally, our data suggest that SOC do not significantly contribute to elevation of [Ca²⁺]_i in the basilar artery.