The effects of Mg2+ and nifedipine (Nif) on vasoconstriction and Ca2+ transients were studied in intact, pressurized rat mesenteric arteries with myogenic tone. Changes in cytosolic Ca2+ concentration ([Ca2+]cyt) were measured with confocal microscopy in fluo 4-AM loaded, individual myocytes. Myogenic tone was abolished by 10 mM Mg2+ or 0.3 μM Nif. Contractions induced by 75 mM K+ depolarization were blocked by 0.3 μM Nif, but not by 10 mM Mg2+. Phenylephrine (PE; 5 μM) evoked sustained [Ca2+]cyt elevation and vasoconstriction with superimposed Ca2+ oscillations and vasomotion. The subsequent addition of 10 mM Mg2+ or 0.3 μM Nif reduced [Ca2+]cyt and abolished plateau vasoconstriction. When added before PE, both Mg2+ and Nif abolished the PE-evoked Ca2+ oscillations and vasomotion. Mg2+ dilated the PE-constricted arteries after a brief (≤180–240 s) vasoconstriction, but Nif did not. Both agents also abolished the vasoconstriction attributed to Ca2+ entry through store-operated channels (SOCs) during internal Ca2+store refilling that followed store depletion. The data suggest that Ca2+ entry through SOCs helps maintain both myogenic tone and α1-adrenoceptor-induced tonic vasoconstriction.
- confocal laser scanning microscopy
- cytosolic Ca2+ concentration
- mesenteric artery
a variety of Ca2+-permeable ion channels, including voltage-gated Ca2+ channels, receptor-operated channels (ROCs), and store-operated channels (SOCs), participate in the “resting” and/or agonist- or depolarization-stimulated influx of Ca2+ in vascular smooth muscle (VSM) (33). Recent evidence indicates that SOCs play an important role in sarcoplasmic reticulum (SR) refilling after VSM activation (2,22, 25, 29, 43), and may even be involved in the generation and/or maintenance of myogenic tone (MT) (31, 39). Mg2+ is a known blocker of SOCs (2, 17, 26,43). Thus if SOCs contribute to MT, we might predict that an elevated extracellular Mg2+ concentration ([Mg2+]o) should interfere with the development or maintenance of tone. Indeed, blood vessel tone can be modulated by [Mg2+]o (1, 23,36). Reduction of [Mg2+]o increases tone. Conversely, elevation of [Mg2+]orelaxes VSM: it dilates small spontaneously constricted arteries (i.e., those with MT) and attenuates agonist-evoked vasoconstriction (1,23). Ca2+ exchange in VSM is enhanced by low [Mg2+]o and reduced by high [Mg2+]o (36), which lowers intracellular Ca2+ without changing intracellular Mg2+ (8). Nevertheless, the mechanism(s) of action of Mg2+ is (are) poorly understood.
Interpretation of the effects of Mg2+ is problematic because various channel types might be targets of Mg2+ action. For example, several investigators have reported that Mg2+ is a blocker of voltage-gated Ca2+ channels (1). External Mg2+has a rapid on-off rate and is a “weak blocker” of Ca2+ currents through L-type voltage-gated Ca2+channels (LVGCCs) (21). Little is known about the effects of external Mg2+ on Ca2+ currents and Ca2+ entry through LVGCCs in VSM or other cell types. Dihydropyridines (DHPs) such as nifedipine (Nif) are widely viewed as selective blockers of LVGCCs (10, 34, 38) and thus are potentially useful in distinguishing the role of LVGCCs from other channels, such as SOCs, in phenomena such as MT. Recently, however, some investigators (5, 7, 35) have reported that DHPs may block other Ca2+ entry channels, including SOCs and/or ROCs. There also is new evidence that SOCs, which should be sensitive to Mg2+ (2, 17, 30, 43), may contribute to MT. These findings raise questions about the widely held view (9, 14, 18) that LVGCCs are the main source of Ca2+ for MT.
There is a paucity of information about the direct effects of [Mg2+]o on the cytosolic Ca2+concentration [Ca2+]cyt in VSM. This, plus the uncertainty about the roles of LVGCCs and SOCs in the generation of MT, led us to reexamine these issues in intact, pressurized small mesenteric arteries. Small arteries were studied because MT keeps these arteries in a state of maintained (tonic) contraction that is critical for the control of blood flow and blood pressure. It is important to understand how this tone is generated by increased intralumenal pressure, and how it is maintained, because altered MT may play a role in the pathogenesis of hypertension.
MATERIALS AND METHODS
Preparation of Arteries
Male Sprague-Dawley rats (150–240 g) were euthanized by cervical dislocation. Part of the distal mesenteric artery arcade was rapidly removed and transferred to a chamber containing ice-cold dissection solution. Small arteries [passive diameter (PD) = 170–260 μm] were isolated and dissected into 3–5 mm segments without branches. While under a microscope, both ends of the segments were tied onto glass cannulas (100–150 μm diameter) as described in previous studies (24, 26). One cannula was connected to a servo-controlled pressure-regulating device (Living Systems; Burlington, VT), whereas the other was attached to a closed stopcock. This enabled the study of diameter changes in pressurized segments in the absence of intraluminal flow.
The arteries were continuously superfused with gassed Krebs solution and were equilibrated (45–60 min) to initial experimental conditions (37°C, 70 mmHg). Arteries with significant leaks or branches were discarded. Only vessels that exhibited stable MT were studied further under isobaric conditions.
Arteries were viewed with a ×10 objective with brightfield illumination on a Nikon TMS microscope equipped with a monochrome video charge-coupled device camera. The outer diameter was monitored by a real-time edge-detection system that utilizes a video frame grabber and custom-designed LabView software (National Instruments; Austin, TX). We determined PD at the end of each experiment by incubating the tissue in Ca2+-free solution for 10 min.
Imaging of Ca2+.
For Ca2+ imaging, dissected arterial segments were loaded (∼3 h at room temperature) with Ca2+ indicator (fluo 4) in albumin-free dissection solution containing 15 μM fluo 4-AM, 1% DMSO (vol/vol), and 0.03% cremaphor EL (vol/vol). Dye-loaded arteries were mounted on the glass cannulas, pressurized to 70 mmHg, and superfused with Krebs solution for 30 min at 20–22°C to remove extracellular Ca2+ dye. The temperature was then raised to 37°C to permit the development of MT. Experiments were carried out at 37°C to mimic in vivo conditions and to maintain MT, although the dye was lost at a faster rate than at 20–22°C.
The arteries were imaged with a confocal imaging system (model MRC 600, Bio-Rad Life Science Group; Hercules, CA) connected to a Nikon Diaphot microscope equipped with a Nikon CFN plan Apochromat (×60, numerical aperture 1.2) water immersion objective. Fluorescence was excited by the 488-nm emission line of an argon ion laser. The video output of the Bio-Rad system was captured with a video capture board (PCI 1407, National Instruments) and stored in a computer. A transverse optical section (211 × 114 μm or 768 × 512 pixels) through the center of the artery was imaged (Fig. 1; see also Fig. 1 in Ref. 24). The cross sections of ∼20–35 individual dye-loaded myocytes could usually be clearly identified within the artery wall in these images. Because the myocytes are spindle-shaped, the cross-sectional areas of cells near their ends were small and were not analyzed. Images were captured at the rate of either 1 or 2 frames/s for periods of up to 10 min, and could be captured at 4 frames/s for short periods. The use of the transverse optical section at the center of the artery had a special advantage. It enabled us to visualize individual cells in cross section and to follow the same portion of each cell during artery constriction and dilation because this portion moved only horizontally along the optical (focal) plane (Fig. 1). Because the cells were visualized in cross section, Ca2+ waves, which are often evoked by vasoconstrictors (24, 26), could not be observed. Consequently, only seemingly uniform rises or falls of [Ca2+]cytthroughout the entire cell cross section were observed.
A relatively high magnification is needed to visualize individual cells. The arteries were positioned longitudinally along thex-axis (211 μm) so that a maximum number of cells could be visualized. Wall movement due to vasoconstriction and vasodilation then occurred in the y-direction (114 μm). Therefore, only a single wall of each artery could be visualized in the 211 × 114 μm2 field. Vasoconstriction and vasodilation were inferred from the movement of the wall (on the assumption that the diameter change, “Δ diameter” = 2× wall displacement; seeresults). Note that the diameters determined in this way were not in the statistical analyses reported in results. Most arteries studied in this way had only a single layer of myocytes (see results). The relatively large visual field area was needed for two reasons: 1) to visualize a sufficient number of myocytes for analysis of variations in [Ca2+]cyt changes from cell to cell, and2) to detect large displacements of the cell wall as the artery constricted and dilated. The time required to scan the large image area plus the need to minimize dye bleaching limited the rate of image capture. Thus both spatial (i.e., Fig. 1 in Ref. 24) and temporal resolutions were compromised to some extent to obtain sufficient data for analysis. Fluo 4 is a nonratiometric dye, and absolute values of [Ca2+]cyt were not determined. Nevertheless, the fluo 4 fluorescence changes are a reliable indicator of changes in [Ca2+]cyt(26).
To quantitate changes in [Ca2+]cyt, fluo 4 fluorescence changes were measured in a small square that remained within the boundaries of the cell even when the artery was maximally dilated and the cell cross-sectional area was at a minimum (24). Under these circumstances, the volume of cytoplasm in which fluorescence was measured remained constant whether the cell cross-sectional area increased or decreased. Thus changes in fluorescence can be attributed solely to changes in [Ca2+]cyt, and not to changes in the amount of fluo 4 within the measured volume.
Drugs and solutions.
The dissection solution was composed of (in mM) 145 NaCl, 4.7 KCl, 1.2 MgSO4 · 7 H2O, 2.0 MOPS, 0.02 EDTA, 1.2 NaH2PO4, 2.0 CaCl2 · 2 H2O, 5.0 glucose, 2.0 pyruvate, and 1% albumin (pH 7.4 at 5°C). The Krebs solution was composed of (in mM) 112 NaCl, 26 NaHCO3, 4.9 KCl, 2.5 CaCl2, 1.2 MgSO4 · 7H2O, 1.2 KH2PO4, 11.5 glucose, and 10 HEPES (adjusted pH to 7.3–7.4 with NaOH). K+-Krebs contained 75 mM KCl and only 42 mM NaCl. In other cases, varying amounts of NaCl were replaced with LiCl, N-methyl-d-glucamine (NMDG)-Cl, or sufficient sucrose to maintain osmolality (seeresults for details). Ca2+-free solution (0 Ca2+-Krebs) was made by the omission of Ca2+ and addition of 0.5 mM EGTA. For solutions with a reduced [Mg2+]o, the [Mg2+] in the Krebs solution was altered by deleting MgSO4 and adding MgCl2 to give the desired final concentration. For solutions with an elevated [Mg2+]o, MgCl2 was usually added to the normal Krebs. In some cases, however, 27 mM NaCl was replaced by either 8.8 mM MgCl2plus 6 mM sucrose (=10 mM Mg2+-Krebs) to maintain osmolality, or 27 mM LiCl to maintain both ionic strength and osmolality. Solution osmolalities were measured with a vapor pressure osmometer (model 5500, Wescor; Logan, UT). All osmolalities reported inresults are means of multiple measurements with SE <0.5%. The normal or modified Krebs solutions were all gassed with a mixture of 90% N2-5% CO2-5% O2. The tissue chamber had a volume of ∼1.0 ml, and the arteries were superfused at a rate of 3–4 ml/min. See results for details of the experimental protocols and solution changes.
Phenylephrine (PE), Nif, verapamil, phentolamine,N,N,N′,N′- tetrakis (2-pyridylmethyl) ethylenediamine (TPEN), NMDG, and cremaphor EL were purchased from Sigma-Aldrich (St. Louis, MO). PE, Nif, verapamil, and phentolamine were prepared as stock solutions, and aliquots were added to the relevant solutions just before each experiment. Fluo 4-AM was purchased from Molecular Probes (Eugene, OR).
Data Analysis and Statistics
Artery diameter data are expressed as means ± SE andn indicates the number of arteries studied (1 per animal). Comparisons of data were made using Student's pairedt-test. Data were considered significantly different atP < 0.05. For confocal laser scanning experiments, data were analyzed using custom software written in Interactive Data Language (IDL, Research Systems; Boulder, CO). IDL was also used to create compressed video files in the MPEG format. Video titles were added with Adobe Premiere software (Adobe Systems; San Jose, CA). Seeresults for web site information regarding the video files. The videos can be viewed with the use of Windows Media Player (Microsoft; Redmond, WA).
Effects of [Mg2+]o on MT
When first cannulated and pressurized to 70 mmHg at 20–22°C, arteries were nearly maximally dilated (i.e., to PD). The arteries then developed MT (i.e., spontaneous, maintained vasoconstriction) during 30–60 min incubation in Krebs solution at 37°C (9). With fully developed MT at 70 mmHg, the average diameter of the arteries was 75 ± 1% (n= 69) of PD; this pressure was used for all experiments described here. Figure 2 A shows an example of control MT (MTc) generated under these conditions. The PD (indicated by the dashed line) was determined in 0 Ca2+-Krebs at the end of the experiment.
Figure 2 A also illustrates the protocol used to test the effect of [Mg2+]o on MT. When [Mg2+]o was increased from normal (1.2 mM) to 6 mM for 2 min, the artery was partially dilated. The effect was reversible; the artery constricted back to the previous diameter when normal Krebs was restored. In Fig. 2 B, the MT is presented as a percentage of MTc, where MT = MT*/MTc × 100%, and MT* is the residual MT at low or high [Mg2+]o (Fig. 2 A).
Data for seven such arteries, in which [Mg2+]o between 0.2 and 10 mM were tested, are summarized in Fig. 2 B. Reduction of [Mg2+]o <1.2 mM increased MT. Conversely, elevation of [Mg2+]o >1.2 mM reduced MT in a concentration-dependent manner. The [Mg2+]oneeded for a 50% reduction in MT was 5 mM.
In all of the aforementioned high-[Mg2+]oexperiments, MgCl2 was simply added to the normal Krebs. Thus both the ionic strength and the osmolality were increased. To control for the high osmolality of the 10 mM Mg2+-Krebs (299 vs. 286 mosmol/kgH2O for normal Krebs), we compared the effects of this solution to Krebs solution, in which 8 mM NaCl or 16 mM sucrose was added. These latter solutions, which had normal [Mg2+]o and osmolalities of 300 and 301 mosmol/kgH2O, respectively, had negligible effects on the normal MT (not shown).
The solutions with high [Mg2+]o also had a high ionic strength, and therefore might be expected to increase screening of membrane surface charges (15). By this mechanism, elevated [Mg2+]o might reduce the voltage-dependent entry of Ca2+ through LVGCCs and reduce MT. To control for this, we examined the effects of elevated [Mg2+]o in solutions in which the ionic strength was kept the same as in the normal Krebs. In these solutions, [Na+]o was maintained constant (85 mM) by reducing the NaCl content of the Krebs; ionic strength was kept the same as in normal Krebs (ionic strength, 159; osmolality, 286 mosmol/kgH2O) with either 27 mM LiCl, or 8.8 mM MgCl2 + 27 mM sucrose. As shown in Fig. 2 C, when the normal Krebs was replaced by modified Krebs, which contained 27 mM LiCl and only 85 mM Na+; there was only a small, transient decline in MT. When the solution containing 10 mM Mg2+ + 27 mM sucrose was introduced, however, MT was nearly completely abolished. These data indicate that high [Mg2+]o inhibits MT by a mechanism that does not involve surface charge screening.
Effects of Nif on MT
A representative experiment showing the effect of Nif on MT, comparable to the experiments with high [Mg2+]o (Fig. 2 B) is illustrated in Fig. 2 D. Nif (0.3 μM), like high [Mg2+]o, rapidly reduced MT (20). The effect of Nif, too, was reversible, but the recovery (not shown) was much slower than that observed with restoration of normal [Mg2+]o. The dose-response curve for the effect of 0.01–0.3 μM Nif on MT is presented in Fig. 2 E. The IC50 for inhibition of MT by Nif was 0.03 μM. A maximal effect was observed with 0.3 μM Nif; increasing the Nif concentration to 1 μM or more had negligible additional effect (not shown).
Effects of Elevated [Mg2+]o and Nif on [Ca2+]cyt in Individual VSM Cells of Pressurized Arteries: Relation to MT
MT is believed to be maintained by elevated [Ca2+]cyt because the tone depends upon the entry of extracellular Ca2+ into the myocytes. Indeed, recent studies on intact (18) and endothelium-denuded (31) small arteries indicate that MT is directly related to the global [Ca2+]cyt. Therefore, to determine the effects of Nif and high [Mg2+]oon the relationship between [Ca2+]cyt and MT, fluo 4 fluorescence was measured in individual identified myocytes, simultaneously with diameter change, with the use of confocal microscopy.
Data from three representative cells in each of two representative experiments are presented in Fig. 3. Figure 3 A, top, shows a fluorescent image; three of the individually identified myocytes are outlined. Under control conditions, the artery generated MT, and many of the cells exhibited increases in [Ca2+]cyt (detected as maintained increases in fluo 4 fluorescence; not shown). When [Mg2+]o was elevated to 10 mM, the steady level of [Ca2+]cyt (i.e., fluo 4 fluorescence) in each of the myocytes declined and the artery dilated, as indicated by the change in diameter. On restoration of normal Krebs (with 1.2 mM Mg2+), the steady [Ca2+]cyt in the myocytes rose and the artery constricted. The effects of 3 and 6 mM Mg2+ were also tested; the effects on [Ca2+]cyt were graded with [Mg2+]o (not shown). A video clip of the original fluo 4 fluorescence data shown in Fig.3 A from the experiment is athttp://ajpheart.physiology.org/cgi/content/full/283/6/H2691/DC1.
Data from a comparable experiment with Nif are shown in Fig.3 B. Nif (0.3 μM) greatly reduced the steady [Ca2+]cyt in all cells and dilated the artery. When Nif was washed out, there was a delay before [Ca2+]cyt started to rise again. The increase in steady [Ca2+]cyt was accompanied by complete recovery of MT.
Effects of Elevated [Mg2+]o and Nif on K+-Evoked Vasoconstriction
The preceding results indicate that Nif and elevated [Mg2+]o inhibit MT in these arteries in comparable fashion. The inhibitory effect of DHPs, such as Nif on MT, has been attributed to block of LVGCCs (9, 20). Furthermore, Mg2+ is a “weak inhibitor” of LVGCCs (21). Therefore, experiments were designed to determine whether the effects of high [Mg2+]o and Nif can be explained by a block of LVGCCs. To this end, the effects of Nif and high [Mg2+]o were tested on 75 mM K+-evoked and depolarization-activated (presumably primarily LVGCC mediated) vasoconstriction in rat small mesenteric arteries.
The protocol and representative experimental results are illustrated in Fig. 4, A–C. Figure 4 A shows a “control” 75 mM K+-evoked vasoconstriction superimposed on the MT. This artery initially constricted to 67% of PD (= MTc). Elevating the external K+ concentration ([K+]o) to 75 mM for 2 min, which should depolarize arterial myocytes to about −20 mV, sufficient to open most LVGCCs, induced a further, reversible vasoconstriction to 46% of PD. This constriction corresponds to the sum of MTc and the control 75 mM K+(depolarization)-evoked vasoconstriction (Kc). In this and all subsequent experiments in this series, all solutions contained 1 μM phentolamine to block the α-adrenoceptors and eliminate any contribution of catecholamine release from the intravascular sympathetic nerves to the observed responses. In control experiments, replacement of 70 mM NaCl by equimolar LiCl or NMDG-Cl, or sufficient sucrose to maintain osmolality (see materials and methods), did not induce vasoconstriction, but relaxed arteries with MT. Thus it seems reasonable to attribute the vasoconstrictor effect of the 75 mM K+-Krebs to the specific depolarizing action of the high [K+]o.
After recovery for 15 min in normal Krebs, this artery was exposed first to 10 mM Mg2+-Krebs (for 2 min), and then to 75 mM K+-Krebs containing 10 mM Mg2+ for an additional 2 min. As expected, (Figs. 2, A and B, and 3 A), elevation of [Mg2+]o to 10 mM rapidly reduced MT. High K+ depolarization, after a very brief dilation, rapidly and markedly constricted the artery despite the continued presence of 10 mM Mg2+. When normal Krebs was restored, the artery rapidly dilated to its PD, and then slowly recovered MT (see beginning of the record in Fig. 4 C, which shows a continuation of the experiment from Fig. 4 B). With the normal (1.2 mM) [Mg2+]o, raising [K+]o to 75 mM evoked a smaller constriction (Kc in Fig. 3 A) than in the medium with 10 mM Mg2+ (K* in Fig. 4 B). A possible explanation is that the ability of the artery to constrict further was limited when there was already a large vasoconstriction due to the MT. The dose-response curve (Fig. 4 D) shows that elevating [Mg2+]o from 1.2 to 10 mM not only did not inhibit the 75 mM K+-induced vasoconstriction but actually increased the K+-induced constriction significantly (by 50 ± 18%; P < 0.05; n = 3). The final diameters reached in 75 mM K+ were, however, similar in both 1.2 and 10 mM Mg2+ (compare Fig. 4, Aand B).
Treatment of the same artery with 0.3 μM Nif also markedly reduced the MT, but in the presence of Nif, subsequent elevation of [K+]o to 75 mM induced very little vasoconstriction (K* in Fig. 4 C). This is expected if Nif blocks the LVGCCs that are opened by the 75 mM K+-induced depolarization. Nif blocked the high [K+]o-induced contractions with high affinity (Fig. 4 E). The IC50 was 0.04 ± 0.01 μM when Nif was added during the plateau of the K+-induced constriction (not shown). The effect of 10 μM verapamil, a member of a different class of LVGCC blocker, was identical to that of Nif on both MT and the response to 75 mM [K+]o (not shown). This seems consistent with the hypothesis that MT and 75 mM K+-stimulated vasoconstriction are both mediated by LVGCCs. However, the fact that high [Mg2+]o blocks MT, but not the K+-stimulated vasoconstriction, implies that the effect of Mg2+ is not due to block of LVGCCs.
The dose-response curves for Mg2+ and Nif are shown in Fig.4, D and E, respectively. These data demonstrate clearly that, whereas Nif and high [Mg2+]oboth reduce MT, Nif, but not 10 mM Mg2+, blocks the 75 mM K+-induced vasoconstriction. The latter is most likely due to Ca2+ entry as a result of depolarization and consequent opening of LVGCCs. The implication is that most of the pressure-induced, Mg2+-sensitive, myogenic vasoconstriction cannot be the result of Ca2+ entry through LVGCCs alone.
Effects of Elevated [Mg2+]o and Nif on [Ca2+]cyt Changes Evoked by 75 mM [K+]o
A protocol similar to that illustrated in Fig. 4, B andC, was used to investigate the effects of [Mg2+]o and Nif on [Ca2+]cyt during exposure to 75 mM K+-Krebs. The data in Fig. 5shows the effect of 10 mM [Mg2+]o. Under control conditions, the artery had modest MT and a number of myocytes exhibited moderate fluo 4 fluorescence (Fig. 5,A,a and B). The fluorescence in these cells decreased (i.e., [Ca2+]cyt declined) and the artery dilated when [Mg2+]o was raised to 10 mM (Fig. 5, A,b and B). When [K+]o was raised to 75 mM in the presence of high [Mg2+]o, [Ca2+]cyt rose (Fig. 5,A,c and B) and the artery constricted markedly (Fig. 5, A,d,e andB). These effects of high [K+]oand high [Mg2+]o were reversible (Fig. 5,A,f–i and B).
Figure 6 shows a comparable experiment in which the effect of Nif was tested. As already shown (Fig.3 B), 0.3 μM Nif reduced [Ca2+]cyt (i.e., fluo 4 fluorescence) in the myocytes and dilated the artery (Fig. 6, A,b andB). Elevation of [K+]o to 75 mM in the continued presence of Nif had little effect on [Ca2+]cyt in most cells (1 and 2 in Fig6 A,a), but induced modest elevation of [Ca2+]cyt in a few cells (3 in Fig6 A,a). There was no arterial wall displacement (i.e., vasoconstriction) under these circumstances. Following washout of the Nif and high K+, the myocyte fluorescence slowly increased and the artery recovered MT (Fig. 6,A,i and B). When this artery was later exposed to 75 mM [K+]o (in the absence of Nif), the myocytes exhibited a large increase in fluorescence and the artery constricted markedly (not shown). The transient 75 mM [K+]o-induced elevation of [Ca2+]cyt seen in some cells in Fig. 6 may have been due to brief opening of LVGCCs because block by Nif is use dependent (19).
Taken together, the data in Figs. 5 and 6 indicate that the 75 mM [K+]o-induced rise in steady [Ca2+]cyt and vasoconstriction observed in the presence of 10 mM Mg2+ were apparently due to opening of Nif-sensitive LVGCCs and consequent Ca2+ entry. Thus 10 mM Mg2+ does not block these channels.
Effects of Elevated [Mg2+]o on Peak and Plateau Phases of PE-Induced Vasoconstriction
Because Nif and high [Mg2+]o have different effects on the 75 mM K+-evoked responses in small mesenteric arteries, we were prompted to test the effects of Nif and Mg2+ on the responses evoked by α1-adrenoceptor stimulation. Two protocols were used, as illustrated in Fig. 7.
The specific α1-adrenoceptor agonist PE, at a concentration of 5 μM, evoked a rapid vasoconstriction. The constriction reached a maximum (Pkc in Fig.7 A = the control peak PE response) and then relaxed slightly to a new plateau level (Plc). During the plateau phase there was substantial vasomotion (Fig. 7 A). These rapid transient changes in diameter are superimposed on the maintained contraction (indicated by the solid line) because individual constrictions and dilations merged on this time scale. After recovery in normal Krebs, [Mg2+]o was raised to 10 mM and MT was abolished (Fig. 7 B). PE (5 μM) then induced a large vasoconstriction (Pk* in Fig. 7 B), comparable in amplitude to that observed with normal (1.2 mM) [Mg2+]o [i.e., (Pk*/Pkc) × 100 ≥ 100%]. Vasomotion was not observed during the plateau vasoconstriction in the presence of 10 mM Mg2+. After the peak vasoconstriction and a relatively brief plateau phase (180–240 s or less), the artery suddenly and rapidly dilated (see Fig. 7 B). Alternatively, PE was applied first and [Mg2+]o was then raised to 10 mM during the plateau vasoconstriction (Fig. 7 C). The artery then promptly dilated to a new diameter (Pl*) that was only slightly more constricted than PD so that (Pl*/Plc) × 100 ≤ 20%. This artery recovered its normal MT when normal Krebs was restored.
Concentration-response data (Fig. 7 D) reveal that [Mg2+]o between 1.2 and 10 mM did not reduce the initial peak response to 5 μM PE (or lower concentrations, not shown). However, the plateaus of both PE-evoked vasoconstriction and MT were similarly reduced by external Mg2+ in a concentration-dependent manner.
The effects of elevated [Mg2+]o on [Ca2+]cyt were investigated as shown in Fig.7 E. When 10 mM Mg2+ was introduced during the plateau phase of the response, [Ca2+]cyt fell rapidly to low levels. The Ca2+ oscillations that normally accompany PE-induced vasomotion (24) also were abolished, as was the vasomotion.
Effects of Nif on Peak and Plateau Phases of PE-induced Vasoconstriction
Comparable data of the effects of Nif on the PE-evoked response are illustrated in Fig. 8. In seven of ten experiments, 0.3 μM Nif attenuated Pk (e.g., compare Pk* in Fig.8 B with Pkc in Fig. 8 A). On average, 0.3 μM Nif reduced the peak response to 47% of control Pk in these seven arteries. In three other arteries, however, the same concentration of Nif actually increased the amplitude of the initial peak vasoconstriction in response to 5 μM PE (e.g., compare Pk* in Fig. 8 D to Pkc in Figs. 7 A from the same experiment); the mean peak amplitude was 163 ± 4% of control. Despite this variation in the peak of the PE-evoked response, Nif, like high [Mg2+]o, invariably abolished the vasomotion induced by PE.
In all 10 of these experiments, the plateau vasoconstriction in response to PE was maintained during a 5- to 7-min exposure to Nif (Fig. 8, B and D) when Nif was added before PE. This contrasts with the sudden, rapid vasodilation (i.e., decline in the plateau) observed when similar experiments were performed with 10 mM Mg2+ (Fig. 7 B). Nevertheless, when 0.3 μM Nif (like 10 mM Mg2+; Fig. 7 C) was applied during the plateau of the PE-evoked vasoconstriction (PlCin Fig. 8 C), most of this vasoconstriction was rapidly blocked (Pl* = residual constriction). Moreover, 0.3 μM Nif, like 10 mM Mg2+, prevented the vasomotion normally observed during the plateau (recorded as the thickened lines during the plateau; compare Fig. 8, A and C, with Fig. 8,B and D). Nif reduced both MT and the plateau PE-induced vasoconstriction when Nif was applied after PE (as in Fig.8 C) in a comparable, dose-dependent manner (Fig.8 E). These results are similar to those for high [Mg2+]o (Fig. 7 D).
The effects of Nif on the underlying PE-evoked changes in [Ca2+]cyt were also determined, along with simultaneous diameter changes. Treatment with 0.3 μM Nif in the presence of 5 μM PE reduced the steady [Ca2+]cyt and the [Ca2+]cyt oscillations evoked by PE, and relaxed the artery. In the experiment illustrated in Fig.8 F, Nif was added only 50 s after PE, just as the vasoconstriction was reaching a maximum. Thus vasomotion was not apparent before the Nif was added. Nevertheless, abortive [Ca2+]cyt oscillations and vasomotion are seen during the Nif-induced vasodilation.
Effects of Nif and Elevated [Mg2+]o on Vasoconstriction After SR Ca2+ Store Depletion
The observations described above indicate that the inhibition of MT by high [Mg2+]o cannot be explained by block of LVGCCs because 10 mM Mg2+ has little, if any, effect on these channels. Mg2+ does, however, block SOCs very effectively in VSM (2, 43). Recently, some investigators (31, 39) have suggested that SOCs may play a role in the generation of MT. Therefore, we also tested the effects of Nif and 10 mM Mg2+ on the external Ca2+-dependent vasoconstriction following SR Ca2+ store depletion in 0 Ca2+-Krebs.
As already noted, removal of external Ca2+ dilates the arteries to PD. Restoration of normal (2.5 mM) Ca2+ is then associated with prompt constriction of the arteries and recovery of MT (Fig. 9 A; the beginning of the blue record was obtained 15 min after the end of the red record). The 0 Ca2+-Krebs (containing 0.5 mM EGTA) should slowly lower the [Ca2+] within the SR ([Ca2+]SR) as well as lowering [Ca2+]cyt. Reduction of [Ca2+]SR can be accelerated and prolonged by adding the membrane-permeable, low-affinity Ca2+ chelator, TPEN, to the 0 Ca2+-Krebs (16). After TPEN treatment, restoration of normal Krebs induced an even greater vasoconstriction than normal (i.e., an “overshoot”; Fig.9 A, end of the blue record, and Fig. 9 B, end of the red record) and slowed recovery to normal MT (start of the green record, ∼15 min after the end of the blue record in A). This overshoot vasoconstriction after TPEN washout, and at least part of the rapid vasoconstriction induced by external Ca2+restoration in the absence of TPEN, are likely due to Ca2+entry through channels opened by SR Ca2+ store depletion (SOCs), and the consequent rise in [Ca2+]cyt.
Figure 9 B illustrates the effects of Nif and 10 mM [Mg2+]o on the external Ca2+-dependent vasoconstriction after treatment with TPEN. In these experiments, 10 mM [Mg2+]o or 0.3 μM Nif were added with 0 Ca2+-Krebs containing TPEN (i.e., during SR store depletion), and were included in the normal Krebs during the first 2 min of TPEN washout and Ca2+restoration. The data demonstrate that 10 mM Mg2+ and 0.3 μM Nif both markedly reduce or abolish the vasoconstriction that normally occurs when TPEN is removed and external Ca2+ is restored. Because SOCs are blocked by high [Mg2+]o (2, 43), and because DHPs may block SOCs in VSM (31, 35) the possibility is raised that the vasoconstriction observed when Ca2+ was restored after a period in 0 Ca2+-Krebs without or with TPEN was due to Ca2+ entry through SOCs.
Another protocol was also used to deplete the SR Ca2+stores (Figs. 9, C and D). In these experiments, after removal of external Ca2+, 5 μM PE was added to the medium to release residual SR Ca2+ by activation of the inositol 1,4,5-trisphosphate-sensitive channels. This caused a brief, modest vasoconstriction, followed by relaxation to the PD. Washout of the PE and, 1.5 min later, restoration of the extracellular Ca2+ caused the arteries to constrict to a greater extent (by 79% on average) than did restoration of Ca2+ when the arteries were not treated with PE (Fig. 9 C). In these experiments, too, the vasoconstriction induced by restoration of external Ca2+ was abolished by both 10 mM Mg2+and 0.3 μM Nif (Fig. 9 D).
Mechanisms Involved in Generating and Maintaining MT
Recent evidence indicates that a maintained, global increase in [Ca2+]cyt within individual myocytes of small arteries is responsible for MT (18, 31). The myogenic constriction is a result of the Ca2+-dependent activation of myosin light chain kinase and the phosphorylation of myosin light chain (44). Regulation of the Ca2+ sensitivity of contraction by protein kinase C (40) and Rho-associated protein kinase (37) also is important in MT. Nevertheless, the change in [Ca2+]cyt that occurs when wall pressure is increased is essential for the development and maintenance of tone. Several types of channels influence this change in [Ca2+]cyt: 1) LVGCCs, which increase [Ca2+]cyt when they are activated by depolarization caused by increased intramural pressure (12,18); 2) Ca2+-activated K+channels, which are activated by Ca2+ sparks and tend to decrease [Ca2+]cyt by hyperpolarizing the membrane (20); 3) Cl− channel closure, which also hyperpolarizes the membrane (28);4) stretch-activated channels, which also mediate Ca2+ entry (for a review, see Ref. 9).
In addition to the mechanisms listed above, several studies (2,22, 25, 30, 43) suggest that SOCs play an important role in Ca2+ entry and the control of vascular contractility. In fact, antisense oligodeoxynucleotide knockdown of transient receptor potential isoform 6 protein, which may be a component of SOCs (3,6), markedly attenuates MT in small cerebral arteries (37). Also, the inositol 1,4,5-trisphosphate receptor antagonist 2-aminoethoxydiphenyl borate inhibits MT without altering [Ca2+]cyt or affecting the pressure-induced increase in [Ca2+]cyt (31). These findings suggest that inositol 1,4,5-trisphosphate receptor-mediated release of SR Ca2+ and SOCs may both also be involved in generating MT. In sum, multiple mechanisms that affect Ca2+ entry and Ca2+ sensitivity apparently all contribute to the generation and maintenance of MT.
Elevated [Mg2+]o inhibits MT (1) but the mechanism of action of Mg2+ is unclear. Extracellular Mg2+ is a “weak blocker” of LVGCCs (21), and the inhibition of myogenic and hormone-induced vasoconstriction by Mg2+ has been attributed to this Ca2+ antagonist action (1). Mg2+also blocks SOCs in VSM myocytes (2, 43) as well as in other types of cells (11, 30, 42).
How does Mg2+ Inhibit MT? Comparison With Action of DHPs
The data in this study demonstrate that Mg2+, like DHPs, lowers [Ca2+]cyt in the myocytes of small arteries with MT and dilates these vessels. Indeed, 10 mM Mg2+ or 0.3 μM Nif dilates small arteries to (or close to) the PD. In contrast to Nif, however, 10 mM Mg2+ has negligible effect on the sustained rise in [Ca2+]cyt or the resultant vasoconstriction induced by elevating [K+]o to 75 mM. Low concentrations of Nif block these effects of high [K+]o (IC50 for inhibition of the vasoconstriction ≈0.04 μM Nif). This implies that these effects are mediated by Ca2+ entry through LVGCCs opened by the K+-induced depolarization. But these data and the apparent lack of a role of surface charge screening by the higher ionic strength also indicate that the inhibition of MT by Mg2+ cannot be due to blockage of LVGCCs.
The final studies in this report (Fig. 9) fit the view that the SR Ca2+ stores in small arteries are refilled through SOCs that can be blocked by high [Mg2+]o. Moreover, the SOCs in these small arteries are also blocked by low concentrations of Nif. The latter finding adds to the growing evidence that DHPs can block certain SOCs (7, 35), although this is not the case in some other vascular preparations (e.g., 2, 25, 29). Indeed, the LVGCC blockers diltiazem and d-600, as well as DHPs, were all significantly more potent in inhibiting adrenergic agonist-stimulated Ca2+ influx and contraction than 80 mM K+-stimulated Ca2+ influx and contraction in rabbit mesenteric arteries (5). This, too, indicates that the LVGCC blockers are also acting on another Ca2+ entry pathway.
The most parsimonious explanation for these findings is that the inhibitory effects of both Nif and high [Mg2+]o on MT are mediated primarily by blockage of SOCs and not LVGCCs. This may also have important implications for the mechanism underlying the antihypertensive effect of DHPs. It has generally been assumed that this effect is mediated by DHP blockage of LVGCCs (34, 38). Our results, however, suggest that the antihypertensive action of DHPs may actually be the result, primarily, of blockade of SOCs in small artery myocytes.
Mg2+, DHPs, and α1-Adrenoceptor-Mediated Vasoconstriction
Comparison of the effects of high [Mg2+]o and Nif on PE-stimulated changes in [Ca2+]cyt and vasoconstriction provides new information about the α1-adrenoceptor-mediated responses. We observed that 0.3 μM Nif (Fig. 8; also see Ref. 5), but not 10 mM Mg2+ (Fig. 7), usually (7 of 10 experiments) attenuated the initial, peak 5 μM PE-evoked vasoconstriction. This raises the possibility that a portion of this initial response to relatively high PE concentrations (EC50≈ 0.4 μM PE; data not shown) is mediated by depolarization and Ca2+entry through LVGCCs. We would expect these channels to be blocked by DHPs but not by 10 mM Mg2+.
In contrast, Mg2+ and Nif have similar inhibitory effects on the following: 1) PE-evoked, maintained [Ca2+]cyt elevation; 2) superimposed [Ca2+]cyt oscillations and associated vasomotion; and 3) PE-evoked tonic (plateau) vasoconstriction; these findings require a different explanation. Indeed, there is a striking similarity between the dose-response curves for inhibition of the plateau vasoconstriction and MT, respectively, by both Mg2+ and Nif (Figs. 7 D and 8 E). This may indicate that the tonic vasoconstrictor action of PE is mediated largely by Ca2+ entry through SOCs (2,25) because, as noted above, both Nif and Mg2+ block SOCs in these small arteries.
Whereas 0.3 μM Nif and 10 mM Mg2+ both block PE-evoked Ca2+ oscillations and vasomotion, and PE-evoked plateau vasoconstriction, details of these findings may provide clues to possible underlying mechanisms. When the Nif or 10 mM Mg2+was introduced before PE, the Ca2+ oscillations and vasomotion were blocked before any effect on the plateau constriction was observed. Thus the mechanisms that underlie the Ca2+oscillations and vasomotion must differ from those responsible for the plateau vasoconstriction. It is not yet clear, however, why 10 mM Mg2+ or 0.3 μM Nif immediately abolished the plateau when added after PE (Figs. 7, C and E, and 8,C and F) whereas, when applied before PE, Mg2+ dilated the artery after a brief (≤180–240 s) vasoconstriction (≈200 s in Fig. 7 C), and whereas Nif did not reduce the plateau even after 400 s (Fig. 8, B andD).
Study of “Integrated” Vascular Responses
An important caveat to our observations about the differential effects of Mg2+ and Nif relates to the fact that the present study was carried out on intact arteries. Thus the responses of the endothelial cells and myocytes were integrated in these experiments. Moreover, all solutions were applied at the perivascular (adventitial) surface. We cannot rule out the possibility that the differential effects of Nif and Mg2+ on Ca2+ entry might be due to differences in penetration to the endothelial cells. Attempts to destroy all of the endothelial cells and none of the myocytes yielded ambiguous results. Also, it is unfortunate that no specific and selective blocker of SOCs is available to directly test our conclusion that both Nif and Mg2+block SOCs in these arteries (32).
A New Look at Generation of MT
The data described in this study raise a critical question about the widely held view that LVGCCs are a major contributor to the generation of MT in small blood vessels (9, 14). This hypothesis is based largely on the assumption that DHPs are selective blockers of LVGCCs. Our results and those of others (5, 7,35) indicate that this assumption may not be valid, and that DHPs may also block SOCs in some vascular preparations. These data, plus other recent observations (31, 39), indicate that Ca2+ entry through SOCs may play a more important role in the generation of MT than previously suspected. Moreover, at least in some small arteries, the contribution from SOCs may predominate over the contribution from LVGCCs.
Control of [Ca2+]cytand MT: Importance of Ca2+Extrusion
If the level of MT is, indeed, directly correlated with, and dependent on, tonic [Ca2+]cyt, then another factor needs attention as well. Maintained [Ca2+]cyt is controlled not only by the rate of Ca2+ entry, but also by the rate of Ca2+exit from the cytosol. There will, of course, be a steady-state traffic of Ca2+ between the cytosol and SR, with equal movements of ions in the two directions over time. In addition, two mechanisms are known to play a role in net Ca2+ extrusion from arterial myocytes, the ATP-driven plasma membrane Ca2+ pump (28, 41) and the Na+ electrochemical gradient-driven Na+/Ca2+ exchanger (4). It should be apparent that inhibition of either of these mechanisms will, like activation of any of the Ca2+entry mechanisms, promote a rise in [Ca2+]cytand the consequently increased vasoconstriction (13). Therefore, it seems important to also consider the role of Ca2+ extrusion mechanisms that have been relatively ignored until now regarding their possible contribution to the generation and maintenance of MT (9). Indeed, it is apparent that multiple Ca2+ transport/regulatory mechanisms must contribute to MT; the SOCs are just one such mechanism.
We thank Joseph R. H. Mauban and Christine Lamont for assistance with techniques.
This study was supported by a postdoctoral fellowship from the American Heart Association, Mid-Atlantic Affiliate (to J. Zhang), and by National Heart, Lung, and Blood Institute Grants HL-60748 (to W. G. Wier) and HL-45215 (to M. P. Blaustein).
Address for reprint requests and other correspondence: M. P. Blaustein, Dept. of Physiology, Univ. of Maryland School of Medicine, 655 W. Baltimore St., Baltimore, MD 21201 (E-mail:).
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.
August 15, 2002;10.1152/ajpheart.00260.2002
- Copyright © 2002 the American Physiological Society