The possible roles of endothelial intracellular Ca2+ concentration ([Ca2+]i), nitric oxide (NO), arachidonic acid (AA) metabolites, and Ca2+-activated K+ (KCa) channels in adrenergically induced vasomotion were examined in pressurized rat mesenteric arteries. Removal of the endothelium or buffering [Ca2+]i selectively in endothelial cells with BAPTA eliminated vasomotion in response to phenylephrine (PE; 10.0 μM). In arteries with intact endothelium, inhibition of NO synthase with Nω-nitro-l-arginine methyl ester (l-NAME; 300.0 μM) or Nω-nitro-l-arginine (l-NNA; 300.0 μM) did not eliminate vasomotion. Neither inhibition of cGMP formation with 10.0 μM 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) nor inhibition of prostanoid formation (10.0 μM indomethacin) eliminated vasomotion. Similarly, inhibition of AA cytochrome P-450 metabolism with an intraluminal application of 17-octadecynoic acid (17-ODYA) or 6-(2-propargyloxyphenyl)hexanoic acid (PPOH) failed to eliminate vasomotion. In contrast, intraluminal application of the KCa channel blockers apamin (250.0 nM) and charybdotoxin (100.0 nM), together, abolished vasomotion and changed synchronous Ca2+ oscillations in smooth muscle cells to asynchronous propagating Ca2+ waves. Apamin, charybdotoxin, or iberiotoxin (100.0 nM) alone did not eliminate vasomotion, nor did the combination of apamin and iberiotoxin. The results show that adrenergic vasomotion in rat mesenteric arteries is critically dependent on Ca2+-activated K+ channels in endothelial cells. Because these channels (small- and intermediate-conductance KCa channels) are a recognized component of EDHF, we conclude therefore that EDHF is essential for the development of adrenergically induced vasomotion.
- smooth muscle
- endothelial cell
vasomotion is the rhythmic change in vessel diameter that results from cyclical contraction and relaxation of smooth muscle cells in arteries and veins. Whether spontaneous or evoked by agonists, vasomotion is present in a wide array of vascular tissues (28, 44). Vasomotion is generated by cyclical or oscillatory changes in cytosolic [Ca2+] (35, 24) that are spatially uniform within individual smooth muscle cells (SMCs) and synchronous between SMCs. In contrast, tonic vasocontriction is associated with asynchronous propagating Ca2+ waves in SMCs. The mechanisms within SMCs and endothelial cells (ECs) that produce the synchronous Ca2+ oscillations that underlie vasomotion remain obscure. An important role is indicated for changes in membrane potential of SMCs and endothelial influences. Indeed, it has been shown that vasomotion in rat mesenteric arteries is accompanied by cyclical changes in membrane potential (12, 15, 27, 29) and is dependent on functional endothelium (16, 35). Furthermore, it has been suggested that nitric oxide (NO) from ECs and subsequent elevation of cGMP in SMCs represent the essential contribution of the endothelium to the development of voltage-dependent vasomotion (35). In this theory, oscillatory Ca2+ release from the sarcoplasmic reticulum (SR) of SMCs triggers a cGMP-dependent, Ca2+-activated Cl− current that can synchonize Ca2+ release from the SR in adjacent cells, thus initiating vasomotion (28, 35). Nevertheless, the role of NO and cGMP in the generation of vasomotion is equivocal. Vasomotion was reported to return in endothelium-denuded arteries after incubation with high levels of 8-bromo-cGMP (16); however, others have reported vasomotion or the synchronous Ca2+ signals of vasomotion to be resistant to inhibition of NO synthase (43). Furthermore, NO has been reported to either allow (35) or impede (43) the development of synchronous Ca2+ signals during vasomotion in rat mesenteric arteries. Other vasoactive factors that can have effects on the contractile state of the artery also originate from ECs. Part of the relaxation, in response to endothelial stimulation with acetylcholine (ACh) or bradykinin (BK), is mediated via NO in mesenteric arteries (34), but a significant component is attributed to apamin (Apa)- and charybdotoxin (ChTX)-sensitive EDHF (5, 33). The possibility that an endothelial factor other than NO is responsible for the generation of vasomotion has been examined recently (30), and the results indicated that EDHF is also involved in vasomotion. Nevertheless, the study did not indicate whether inhibition of Ca2+-activated K+ (KCa) channels alone, with Apa and ChTX, without the simultaneous inhibition of NO synthase and cyclooxygenase, was sufficient to abolish vasomotion. The goals of the present study were to 1) reexamine the role of NO and cGMP in the generation of vasomotion, 2) determine whether ECs and cytosolic [Ca2+] changes of ECs are essential for vasomotion, 3) determine whether KCa channels in ECs are essential to vasomotion, and 4) define the changes in adrenergically stimulated Ca2+ signals in SMCs during inhibition of KCa channels in ECs. The results show that inhibition of small-conductance (SKCa) and intermediate-conductance KCa (IKCa) channels of endothelium, by itself, abolishes adrenergically induced vasomotion and changes Ca2+ signals in SMCs from synchronous Ca2+ transients to asynchronous propagating Ca2+ waves. EDHF therefore has an obligatory role in the development of vasomotion in rat mesenteric arteries.
Preparation of Arteries
All experiments were carried out according to the guidelines of the Institutional Animal Care and Use Committee of the University of Maryland School of Medicine. Male Sprague-Dawley rats weighing 150–260 g were anesthetized by an intramuscular injection of Nembutal (50–100 mg/kg) and killed by cervical dislocation. The mesenteric arcade was dissected and placed in a cold dissection chamber (7°C) containing a solution (the “dissection” solution) of the following composition (in mM): 2.0 MOPS, 145.0 NaCl, 4.7 KCl, 2.0 CaCl2, 1.2 MgSO4, 1.2 NaH2PO4, 0.02 EDTA, 2.0 pyruvate, and 5.0 glucose (pH 7.4). Dissected arteries were loaded with fluo-4 at room temperature for 3–4 h in the dissection solution, to which had been added 10.0 μM fluo-4 AM, 1.5% (vol/vol) DMSO, and 0.03% (vol/vol) cremophor EL. All arteries were equilibrated over ∼1 h to initial experimental conditions (22 or 37 ± 0.5°C, 70 mmHg) and studied in a modified Krebs solution of the following composition (in mM): 112.0 NaCl, 25.7 NaHCO3, 4.9 KCl, 2.5 CaCl2, 1.2 MgSO4, 1.2 KHPO4, 11.5 glucose, and 10.0 HEPES (pH 7.4 at 22°C) equilibrated with a gas mixture of 5% O2-5% CO2-90% N2. Ca2+-free Krebs solution contained 2 mM EGTA. Most arteries were studied at 22°C rather than 37°C because fluo-4 was retained better at the lower temperature. Initial studies in arteries free of dye, at the two temperatures (24), established some of the behavioral differences at these temperatures.
Measurements of Fluorescence and Arterial Diameter
We used a custom confocal laser scanning microscope, described in detail previously (32, 48). The scanning methods used were those described previously (24). To obtain a larger field of view during large contractions, we used a “dry” low-power objective lens (×20, numerical aperture 0.75). Resolution with this objective lens was still adequate to resolve individual SMCs. “Radial” confocal sections were used to image SMCs in cross section during active contractions. Individual SMCs can be successfully “tracked” during vasomotion using radial sections. “Tangential” sections were used to investigate propagating Ca2+ waves. Although we may refer to the images as “Ca2+ images,” all the images are simply of Ca2+-dependent fluo-4 fluorescence. Measurements of the arterial wall position were done either by using the edges of the fluorescence image or in arteries not containing fluo-4, from transmitted light images recorded at 2 Hz. In some cases, digital video files (MPEGs) were constructed from the series of confocal images (see on-line supplemental data, http://ajpheart.physiology.org/cgi/content/full/01084.2003/DC1).
Drugs were purchased from Sigma (St. Louis, MO) and were prepared as stock solutions in the appropriate solvent. Miconazole, 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ), and BAPTA-AM were prepared using DMSO. BK, ACh, Nω-nitro-l-arginine methyl ester (l-NAME), iberiotoxin (IbTx), and ChTX were prepared using distilled water. Nω-nitro-l-arginine (l-NNA) was prepared using 1 M HCl. Apa was prepared using 50.0 mM acetic acid. Indomethacin, 17-octadecynoic acid (17-ODYA), and 6-(2-propargyloxyphenyl)hexanoic acid (PPOH) were prepared using ethanol.
Pressurization of Arteries, Deendothelialization, and Intraluminal Delivery of Drugs
Arteries were cannulated at both ends and mounted in a chamber permitting internal pressurization through the use of pressure transducers and a servo-controlled micropump (Living Systems; Burlington, VT). The intraluminal solution contained the dissection solution under control conditions.
To remove ECs, an air bubble was introduced into the pressurizing lines by opening the distal stopcock and allowing air to enter the pressurizing lines. Once the air bubble reached the lumen of the artery, closing the distal stopcock stopped the solution flow. A 20-min wait period was observed, after which the air bubble was then pushed through the lumen by opening the distal stopcock; a 15-min wait period was allowed before measurements were taken. Arteries normally dilated to passive diameter (i.e., that in the absence of Ca2+) in response to ACh at 1.0 μM. Adequate removal of the endothelium was indicated by relaxation of not more than 10% of passive diameter. Perfusion with air was repeated as necessary to achieve this.
Drugs to be introduced intraluminally were dissolved in dissection solution to their final concentrations. Corrections in pH were made, as necessary, after the addition of drugs. A predetermined volume of drug-containing solution was perfused through the pressurizing lines using the pressure servo control system to ensure complete replacement of solution inside the lumen of the artery. To examine the role of changes in cytosolic Ca2+ in endothelial cells, BAPTA-AM (40.0 μM) was introduced into the lumen of the artery for 10 min (22–24°C) and then flushed out. A 15-min equilibration time was allowed before measurements were taken. To inhibit NO synthase, arteries were incubated with 300.0 μM l-NAME or 300.0 μM l-NNA inside and outside the lumen (1 h, 37°C). ODQ (10.0 μM) was delivered in the superfusate for 60 min (37°C) to inhibit soluble guanylyl cyclase. Some arteries were treated with both l-NNA and ODQ simultaneously to inhibit the activity of both NO synthase and soluble G cyclase. All drugs delivered outside the lumen were dissolved in the modified Krebs solution. Arteries were incubated with indomethacin to investigate the role of cyclooxygenase metabolism in vasomotion (10.0 μM, 30 min, 37°C). Arteries were incubated with 10.0 μM miconazole inside, outside, or both inside and outside the lumen for 20 min (37°C) to determine the effects and site of action of miconazole. PPOH (50.0 μM) and 17-ODYA (10.0–50.0 μM, 37°C) were delivered inside the lumen for 30–60 min to determine the effects of cytochrome P-450 metabolism on vasomotion. To determine the influence of various K+ channels on vasomotion, 250.0 nM Apa, 100.0 nM ChTX, and 100.0 nM IbTX were delivered to the lumen of the artery alone or in combinations.
Analysis of Results
The frequency of vasomotion in arteries pressurized at 70 mmHg settled at two distinct values depending on the temperature at which the experiments were conducted (24). The amplitude of vasomotion was determined as the difference between the maximums and minimums of the vasomotion waveform. Contraction and vasomotion amplitudes were normalized to the maximum passive diameter of the artery at 70 mmHg and 0 mM Ca2+ solution. Results are given as mean ± SE of n arteries, each obtained from a different animal. Statistical significance was tested using paired Student's t-tests with a P value of <0.05 taken as significant.
Role of ECs in Vasomotion: Removal of Endothelium and Buffering of Ca2+
We first examined the effects of removal of the endothelium on adrenergic vasomotion (Fig. 1). Removal of the endothelium (see methods) was judged to be successful when the relaxation in response to ACh (10.0 μM; Fig. 1, solid trace) was reduced to ∼10% of control values after the removal of ECs with an air bubble (n = 4; Fig. 1, dashed trace). The removal of ECs also significantly increased the contractile response to high levels of phenylephrine (PE; 10.0 μM) stimulation (control: 0.55 ± 0.01, no endothelium: 0.46 ± 0.02, n = 4; amplitude of contraction is reported as the fractional diameter achieved after stimulation). Adrenergically stimulated vasomotion was entirely abolished in all arteries when removal of the endothelium, as judged by the lack of response to ACh, was achieved (Table 1). In preparations in which removal of the endothelium was not yet complete after the first air bubble, as indicated by the relaxation in response to ACh, oscillatory vasomotion was still observed (3 of 4 arteries). In these arteries, vasomotion promptly disappeared once removal of the endothelium was successful in a further attempt. These results are in contrast to those of a recent report (20) in which vasomotion was observed in mesenteric arteries even after removal of ECs.
Because the production of certain vasoactive substances (e.g., NO) by ECs is dependent on elevation of intracellular Ca2+ concentration ([Ca2+]i), we next sought to determine whether changes in [Ca2+]i of ECs might be necessary for adrenergic vasomotion. Selective buffering of changes in [Ca2+]i of ECs, and not SMCs, was achieved by loading ECs with the Ca2+ chelator BAPTA. BAPTA-AM (40.0 μM) was applied to the inside only (lumen) of pressurized arteries. The efficacy of this procedure in loading ECs selectively was verified by confocal imaging of arteries in which fluo-4 AM (a very similar substance) was applied in similar fashion. In these arteries, fluo-4 fluorescence was detected only in ECs and not in SMCs (not shown). Figure 2 shows the typical response of arteries (n = 5) treated with the Ca2+ chelator BAPTA-AM. All five arteries exhibited vasomotion in response to stimulation with maximally effective doses of PE (10.0 μM). Vasomotion was abolished in all five arteries (Table 1), and contraction was increased significantly after treatment with BAPTA (control: 0.53 ± 0.02; BAPTA: 0.46 ± 0.0, n = 5). The vasodilatory response to application of 1.0 μM ACh was also eliminated in all arteries treated with BAPTA. Treatment of arteries with BAPTA increased the elapsed time to reach maximum contraction in response to PE, but the greater amplitude of contraction suggested that Ca2+ in SMCs was not buffered. The frequency of vasomotion seen in Fig. 2 is lower than that in Fig. 1 because the experiments using BAPTA were done at room temperature (22–24°C).
In summary, either the removal of ECs or the buffering of changes in [Ca2+] in them abolishes adrenergic vasomotion. Furthermore, these procedures significantly increase the maximum contraction in response to high levels of adrenergic stimulation (10.0 μM PE). Thus these results suggest that Ca2+-dependent production of relaxing factors in ECs (such as NO) may be necessary for the generation of vasomotion.
NO and cGMP
Inhibition of NO synthase: l-NAME and l-NNA.
To investigate the possible role of NO in adrenergic vasomotion, we used two well-known l-arginine analogs to inhibit the function of NO synthase (l-NAME and l-NNA). l-NAME, at a concentration of 100.0 μM, is effective at inhibiting the endothelium-dependent relaxation and hypotensive responses induced by ACh (37). We used higher concentrations of l-NAME to ensure complete inhibition of NO synthesis. Incubation of arteries with l-NAME (300.0 μM, 37°C, 1 h, both inside and outside the lumen) resulted in a significant increase in the contractile response to stimulation with 1.0 μM PE (Fig. 3A). The mean contraction was 0.58 ± 0.01 and 0.54 ± .0.02 of maximum diameter for control and l-NAME-treated arteries, respectively (n = 4). ACh (1.0 μM) or BK (1.0 μM) was applied to the artery to test for endothelium-dependent relaxation of arteries preconstricted with 1.0 μM PE. Under control conditions, arteries showed a sustained complete relaxation in response to ACh or BK (a total of 10 arteries; Fig. 3, A and B). In the presence of l-NAME or l-NNA, however, the artery contracted again (after an initial transient relaxation) and then underwent erratic vasomotion. The initial phase of relaxation has been attributed to EDHF, whereas the maintained phase of the relaxation has been attributed to NO (33). Thus the absence of sustained relaxation in the experiment is taken as an indirect indicator of the inhibition of NO production. The amplitude of the adrenergic vasomotion (10.0 μM PE) was significantly reduced by l-NAME treatment (n = 4, control: 0.054 ± 0.005; l-NAME: 0.032 ± 0.005 of maximum diameter). l-NNA (300.0 μM) was also used as an alternative antagonist of NO synthase. The amplitude of adrenergic vasomotion in response to 1.0 μM PE was also significantly attenuated by l-NNA (n = 8, control: 0.058 ± 0.008; l-NNA: 0.034 ± 0.005; Table 1).
To probe further the possible involvement of NO in adrenergic vasomotion, we inhibited the production of cGMP, one of the known effectors of NO. ODQ, at a concentration of 10.0 μM, is reported effective at inhibiting the formation of cGMP mediated by soluble G cyclase (13, 41) in cell-free systems and brain slices. We used the same high concentration of antagonist and a longer incubation period to inhibit soluble G cyclase in the intact artery. ODQ (10.0 μM, 37°C, 1-h bath application) had no significant effect on the maximum contraction in response to stimulation with 3.0 μM PE (n = 6, control contraction: 0.60 ± 0.02 of maximum diameter; contraction in ODQ: 0.68 ± 0.05 of maximum diameter; Fig. 3C) but did affect the adrenergically induced vasomotion. In two of the six arteries exposed to ODQ and stimulated with PE (3.0 μM), vasomotion was abolished. However, exposure to maximal concentrations of PE (10.0 μM) reestablished vasomotion. To inhibit both production of NO and cGMP, three other arteries were treated with both l-NNA (300.0 μM) and ODQ (10.0 μM). Vasomotion still occurred in all arteries in response to 1.0 μM PE (n = 3, control: 0.059 ± 0.008; LNNA + ODQ: 0.031 ± 0.008; Table 1). The results with l-NAME, l-NNA, and ODQ agree well with the results of Sell and colleagues (43) and Okazaki and colleagues (30). NO and cGMP play modulatory roles in the generation of adrenergic vasomotion in rat mesenteric arteries.
Arachidonic Acid Metabolism
Inhibition of cyclooxygenase: indomethacin.
We next investigated the possible involvement of endothelially produced arachidonic acid (AA) metabolites that could be involved in adrenergic vasomotion. Indomethacin is a well-established inhibitor of prostaglandin endoperoxide synthase (cyclooxygenase) and is used for inhibition of prostacyclin synthesis. In microsomes, effective inhibition of cyclooxygenase activity occurs by incubation with indomethacin at a concentration of 1.6 μM (21). In three of four arteries, vasomotion in response to lower concentrations of PE (1.0 μM) was abolished by incubation of arteries with indomethacin (10.0 μM, 37°C, 30 min; Fig. 4). The maximum contraction in response to PE (1.0 μM) was not affected (n = 4, control: 0.61 ± 0.01 of maximal diameter; indomethacin: 0.72 ± 0.05 relative to maximal diameter). In the three arteries where vasomotion was abolished, however, a higher concentration of agonist (10.0 μM) reestablished vasomotion (Fig. 4B). To inhibit both cyclooxygenase and production of NO, two other arteries were incubated with a combination of l-NAME (300.0 μM, 1 h) and indomethacin (10.0 μM, 30 min). In agreement with the results of Okazaki and colleagues (30), this combination also failed to abolish vasomotion in response to PE (10.0 μM). Thus cyclooxygenase-dependent products of AA can modulate vasomotion. The possibility that cyclooxygenase products may vary depending on the level of adrenergic stimulation cannot be ruled out. Therefore, the possibility that cyclooxygenase products are essential for vasomotion cannot be excluded given the present results.
Inhibition of cytochrome P-450 epoxygenation and ω-hydroxylation.
Miconazole is an imidazole derivative compound that has been demonstrated to inhibit the epoxygenation reactions of AA. Miconazole did not eliminate vasomotion and showed a trend of decrease in the magnitude of contraction when delivered to the superfusate, external to the artery (n = 4, control: 0.580 ± 0.009; miconazole: 0.603 ± 0.010; Fig. 5B). However, when miconazole was placed both inside and outside the lumen, vasomotion was eliminated and contraction was significantly reduced (n = 9, control: 0.572 ± 0.011; miconazole: 0.709 ± 0.035; Fig. 5C). Vasomotion was reestablished in two of three arteries after a 30-min wash period.
In arteries loaded with fluo-4 for Ca2+ imaging (Fig. 6A), synchronous Ca2+ signals occurred during vasomotion (n = 4), similar to that described previously (35, 24). Ca2+ fluorescence signals increased and declined synchronously in all cells visible in the scan area of the artery. Each trace in Fig. 6A, bottom, represents a single SMC and clearly demonstrates the synchronous Ca2+ fluorescence between adjacent SMCs. A video clip of the data shown in Fig. 6A shows the synchronous calcium signals in SMCs during vasomotion (http://ajpheart.physiology.org/cgi/content/full/01084.2003/DC1). Miconazole (10.0 μM, inside the lumen and out) eliminated vasomotion, and only asynchronous propagating Ca2+ waves could be elicited (Fig. 6B). The Ca2+ waves propagate within a single SMC. The waves are asynchronous between adjacent SMCs and do not propagate between adjacent cells. Two supplemental video clips demonstrate asynchronous propagating Ca2+ waves visualized in both radial and tangential optical sections obtained in the artery of Fig. 6 (see supplemental data, http://ajpheart.physiology.org/cgi/content/full/01084.2003/DC1).
Ppoh and 17-odya.
The results above suggest that miconazole is likely acting on the endothelium because it eliminated vasomotion only when delivered to the lumen of arteries and that it eliminated vasomotion by affecting Ca2+ transients in SMCs. Although miconazole inhibits cytochrome P-450-mediated epoxygenation reactions of AA, it also affects Ca2+ current, ChTX-sensitive K+ currents, and ATP-sensitive K+ currents. Thus we needed to identify the mechanism by which miconazole abolished vasomotion. PPOH is a highly specific inhibitor of cytochrome P-450-mediated epoxygenation reactions of AA with a reported IC50 of 9.0 μM (47). PPOH at 50.0 μM was delivered inside the lumen of the isobaric preparation, and a 30- to 60-min incubation period at 37°C was allowed for the drug to take effect. PPOH did not eliminate vasomotion in the arteries studied (n = 4 arteries; Fig. 7A). In one artery (of 4 arteries total), the amplitude of vasomotion elicited by PE at 3.0 μM was attenuated by PPOH but was restored when the concentration of PE was increased to 10.0 μM.
Arteries were also incubated with the suicide substrate 17-ODYA inside the lumen (1 at 10.0 μM, 2 at 50.0 μM, 30–60 min) to inhibit epoxygenation and ω-hydroxylation reactions of AA metabolism, which may produce vasoactive substances. The reported IC50 values for the inhibition of epoxygenation and ω-hydroxylation reactions by 17-ODYA are 5.0 and 7.0 μM, respectively (47, 51). Exposure to 17-ODYA failed to eliminate vasomotion in response to PE (3.0 μM, n = 3; Fig. 7B). Because neither PPOH nor 17-ODYA eliminated vasomotion, it is unlikely that inhibition of vasomotion by miconazole was mediated via blockade of cytochrome P-450-dependent metabolites of AA.
EDHF and SKCa and IKCa Channels: Apa + ChTx
Another well-known action of miconazole is to block ChTX-sensitive KCa channel currents (1). To test this possibility, endothelial SKCa and IKCa channels were blocked with Apa (250.0 nM) and ChTX (100.0 nM) delivered to the lumen of pressurized arteries, using a method similar to that described by others (10). A predetermined volume of solution containing each of the channel blockers was infused through the pressurizing lines to ensure complete replacement of the saline in the lumen of the artery. At 37°C, a 15-min incubation period was observed before the artery was contracted with 10.0 μM PE. Incubation of arteries with Apa (250.0 nM, n = 1) or ChTX (100.0 nM, n = 1) alone did not abolish vasomotion (Fig. 8). However, vasomotion was abolished in the arteries first treated with either Apa or ChTX when both antagonists were introduced into the lumen of the same arteries (Apa + ChTX). A total of six arteries was incubated with Apa + ChTX, and vasomotion was not observed in all six arteries (Table 1). Supramaximal concentrations of PE (50.0 μM) did not reestablish vasomotion in two additional arteries treated with the combination of Apa and ChTX. Treatment with Apa + ChTX also significantly increased the magnitude of contraction to 10.0 μM PE (n = 6, control: 0.53 ± 0.02; Apa + ChTX: 0.45 ± 0.02 relative to maximal diameter). The effects of Apa + ChTX were poorly reversible. Forty minutes after removal of the drugs, only small-amplitude, transient vasomotion was observed in two of six arteries. ChTX blocks both large-conductance KCa (BKCa) channels and Ca2+-independent, voltage-gated K+ channels (18, 40, 45). To rule out the involvement of BKCa channels in the development of vasomotion, we used IbTX to specifically block BKCa channels. IbTx by itself (100.0 nM; n = 2) did not eliminate vasomotion. The addition of 250.0 nM Apa to the IbTX-treated arteries (Apa + IbTX) also did not abolish vasomotion (Fig. 8). The inability of IbTX to substitute for ChTX thus ruled out the involvement of BKCa channels. Vasomotion persisted in the presence of TEA (10 mM) or 4-aminopyridine (4-AP; 1–5 mM) (15, 17). In our system, 5 mM 4-AP did not abolish vasomotion in arteries stimulated with 10.0 μM PE (n = 2). It is therefore also unlikely that the action of ChTX was mediated by blockade of Ca2+-independent, voltage-gated K+ channels.
The effects of K+ channel antagonists on Ca2+ signals in SMCs were also studied. Confocal imaging showed that, at resting conditions, Ca2+ sparks and Ca2+ waves can be observed (wave frequency: 0.020 ± 0.005 waves·cell−1·s−1, n = 12 cells), in accord with previous findings (24, 26, 50). Figure 9A shows the Ca2+-dependent fluorescence signal during exposure to high levels of PE (10.0 μM). Vasomotion was again accompanied by synchronous Ca2+ elevations along the entire length of the artery visible in the scan window. The arrows in Fig. 9A correspond to two different time points during vasomotion, and the Ca2+ transients are synchronous between different cells marked by the arrows. In contrast, 10.0 μM PE elicited only asynchronous Ca2+ waves in the same artery after treatment with Apa and ChTX together. Two different cells marked in Fig. 9B illustrate the asynchronous Ca2+ signals. The propagating Ca2+ wave frequency was 0.098 ± 0.005 waves·cell−1·s−1 (n = 34 cells from 2 arteries). The combination of Apa and ChTX, which identifies SKCa and IKCa channels, much like miconazole, completely abolished vasomotion and changed Ca2+ transients in SMCs from synchronous Ca2+ oscillations to asynchronous propagating Ca2+ waves. EDHF is therefore essential to the development of the synchronous smooth muscle Ca2+ transients that generate endothelium-dependent vasomotion in rat mesenteric arteries.
The results indicate that adrenergic vasomotion in rat mesenteric arteries is influenced by multiple factors that originate from the endothelium, including NO and AA metabolites. In the present study, vasomotion was absolutely dependent on changes in [Ca2+]i in ECs and on endothelial SKCa and IKCa channels, because the presence of BAPTA in ECs or block of these channels prevented the development of vasomotion in response to α1-adrenoceptor agonists.
Heretofore, most attention has focused on endothelial NO and cGMP in SMCs as the factors essential for adrenergic vasomotion. Given a system that undergoes oscillation, a factor would be considered “essential” if the oscillatory behavior of the system does not occur in the absence of the factor. Factors that modify the characteristics of the oscillating system (frequency, amplitude, etc.) would be classified as “modulatory.” Our present results with removal of the endothelium as well as buffering of Ca2+ in ECs could be taken to suggest that NO and cGMP are essential for vasomotion. Elevation of cytosolic [Ca2+] in ECs facilitates the binding of calmodulin to an endothelial NO synthase signaling complex, thus increasing production of NO (6, 11, 23, 25), and NO can exert vasodilatory influence either directly or via downstream activation of soluble G cyclase and cGMP accumulation in SMCs (2, 3, 14, 19, 22, 29, 31, 38, 39). Indeed, it has been suggested (35) that a rise in smooth muscle cGMP concentration, as a result of the action of NO derived from ECs, is essential to the development of vasomotion. In that model, cGMP allows for a membrane-depolarizing current, presumably a Cl− current, that can in turn synchronize membrane potential and SR Ca2+ release from neighboring SMCs, thus initiating vasomotion (35). The model is supported by experimental data showing that some form of vasomotion can be reestablished in endothelium-denuded rat mesenteric arteries after incubation in high levels of 8-bromo-cGMP. However, in our experiments, inhibition of soluble G cyclase alone, or in combination with extremely high concentrations of l-NNA (300.0 μM), failed to eliminate vasomotion. Thus our results suggest that neither NO nor cGMP is essential for adrenergic vasomotion in rat mesenteric arteries. Instead, the NO cGMP system can be classified as modulatory to vasomotion because modulating effects on vasomotion were observed.
The present results point rather to EDHF as essential to adrenergic vasomotion because the activity of SKCa and IKCa channels in the endothelium (Refs. 7 and 52; for a review, see Ref. 5) is essential for adrenergic vasomotion. Rat mesenteric arteries have Apa- and ChTX-blockable EDHF (9, 36). This EDHF is separate from the activity of anandamide, cytochrome P-450 metabolites, or NO (12, 36, 46). We did not measure changes in EC [Ca2+]i during adrenergic vasomotion in the present study, but it has already been shown in rat mesenteric, pig coronary, hamster cheek pouch, and hamster cremaster muscle arteries that direct stimulation of SMCs with PE (or high K+) causes a secondary, suprabasal elevation of [Ca2+] in underlying ECs (4, 8, 42, 49). Indeed, Schuster and colleagues (42) have shown, using ratiometric measurements of [Ca2+] in endothelium, that an oscillatory pattern of [Ca2+] develops in ECs during adrenergic vasomotion. Such changes in EC [Ca2+] may activate KCa channels, although locally high [Ca2+] may be required. Because adrenergic vasomotion depended critically on endothelial KCa channels, however, we propose that adrenergically induced vasomotion is somehow produced by cyclical activation of these channels, and this constitutes cyclical EDHF. In fact, a recent study (30) elsewhere has shown similar results and conclusions regarding the role of EDHF in adrenergically triggered vasomotion of rat mesenteric arteries.
The source of the elevated [Ca2+] in ECs is presently unknown. We speculate that it diffuses into the EC from SMCs through myoendothelial gap junctions. A further speculation is that mechanisms may exist to elevate [Ca2+] locally in the region of the KCa channels, to levels sufficient to activate them. If operative, however, such mechanisms are presently unknown. Further studies will be required to investigate the mechanisms involved in the activation of KCa channels during vasomotion, as proposed in our model.
Finally, the mechanism by which EDHF synchronizes [Ca2+] oscillations in all SMCs is also unknown. Adrenergic vasomotion is dependent on intact gap junctional communication and never occurs in the presence of blockers of voltage-dependent Ca2+ channels. This result indicates that voltage-dependent Ca2+ influx in SMCs becomes synchronized and then either triggers SR Ca2+ release synchronously or contributes directly to the elevation of SMC [Ca2+]. By hyperpolarizing SMCs, and reducing voltage-dependent Ca2+ influx, EDHF thus provides negative feedback on SMC Ca2+ entry. This feedback mechanism may represent a strategy to prevent vasospasm under high levels of smooth muscle activation and vasoconstriction, thus defining an important role for EDHF in vascular control by the sympathetic nervous system.
This study was supported by National Institutes of Health Research Grant HL-60748 (to W. G. Wier) and Training Grant T32GM-08181 to the University of Maryland Training Program in Integrative Membrane Biology.
We gratefully acknowledge the assistance of M. R. Saunders with video editing.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 by the American Physiological Society