Role of Cav1.2 L-type Ca2+ channels in vascular tone: effects of nifedipine and Mg2+

Jin Zhang, Roberto Berra-Romani, Martina J. Sinnegger-Brauns, Jõrg Striessnig, Mordecai P. Blaustein, Donald R. Matteson


Ca2+ entry via L-type voltage-gated Ca2+ channels (LVGCs) is a key factor in generating myogenic tone (MT), as dihydropyridines (DHPs) and other LVGC blockers, including Mg2+, markedly reduce MT. Recent reports suggest, however, that elevated external Mg2+ concentration and DHPs may also inhibit other Ca2+-entry pathways. Here, we explore the contribution of LVGCs to MT in intact, pressurized mesenteric small arteries using mutant mice (DHPR/R) expressing functional but DHP-insensitive Cav1.2 channels. In wild-type (WT), but not DHPR/R, mouse arteries, nifedipine (0.3–1.0 μM) markedly reduced MT and vasoconstriction induced by high external K+ concentrations ([K+]o), a measure of LVGC-mediated Ca2+ entry. Blocking MT and high [K+]o-induced vasoconstriction by <1 μM nifedipine in WT but not in DHPR/R arteries implies that Ca2+ entry via Cav1.2 LVGCs is obligatory for MT and that nifedipine inhibits MT exclusively by blocking LVGCs. We also examined the effects of Mg2+ on MT and LVGCs. High external Mg2+ concentration (10 mM) blocked MT, slowed the high [K+]o-induced vasoconstrictions, and decreased their amplitude in WT and DHPR/R arteries. To verify that these effects of Mg2+ are due to block of LVGCs, we characterized the effects of extracellular and intracellular Mg2+ on LVGC currents in isolated mesenteric artery myocytes. DHP-sensitive LVGC currents are inhibited by both external and internal Mg2+. The results indicate that Mg2+ relaxes MT by inhibiting Ca2+ influx through LVGCs. These data provide new information about the central role of Cav1.2 LVGCs in generating and maintaining MT in mouse mesenteric small arteries.

  • myogenic tone
  • patch clamp
  • calcium current
  • dihydropyridine receptor
  • arterial smooth muscle

myogenic tone (MT) is observed in small resistance arteries, where it plays an important role, independent of humoral, neural, or endothelial influences, in regulating blood flow and blood pressure (BP) (16). Understanding the generation of MT is important for elucidating the pathogenesis of hypertension (63). Here, we define MT, in vitro, as the intrinsic ability of isolated small artery segments to constrict in response to increases in intraluminal pressure when the arteries are pressurized and warmed to physiological temperature in the absence of intraluminal flow.

In small arteries, MT is apparently triggered by an intraluminal pressure-induced membrane depolarization (28, 31). MT requires extracellular Ca2+: it is abolished in Ca2+-free media and is markedly reduced by L-type voltage-gated Ca2+ channel (LVGC) blockers (16), including dihydropyridines (DHPs), such as nifedipine. The implication is that LVGCs mediate the depolarization-induced Ca2+ entry and consequent vasoconstriction (16). Four subtypes of LVGCs (Cav1.1–1.4) have been identified (14); however, Cav1.2 is the dominant isoform in arterial smooth muscle (ASM) (52). Furthermore, in addition to Cav1 (L type) channels, Cav2 (P/Q type) and Cav3 (T type) voltage-gated Ca2+ channels (VGCs) are also expressed in some ASM cells and appear to play roles in regulating vascular tone and arterial contraction (7, 21, 48, 50). Thus it is worthwhile to determine whether Cav1.2 plays a dominant role in regulating MT and BP.

Previously, we found that both nifedipine (0.3–1 μM) and elevated external Mg2+ concentration ([Mg2+]o; 10 mM) blocked most of the MT in rat mesenteric small arteries. In contrast, nifedipine was a far more effective blocker of the 75 mM external K+ solution ([K+]o)-induced vasoconstriction than was high [Mg2+]o in these arteries (64). Both nifedipine and high [Mg2+]o also inhibited the vasoconstriction induced by readmission of extracellular Ca2+ following the depletion of sarcoplasmic reticulum Ca2+ stores (i.e., Ca2+ entry through store-operated channels; SOCs) (23, 64). Thus part or all of the block of MT by high [Mg2+]o and perhaps even nifedipine might not be mediated by LVGCs. Indeed, there are several reports that nifedipine also blocks other Ca2+-entry mechanisms, such as nonselective cation channels (15) and SOCs (23, 64). Moreover, although high [Mg2+]o blocks LVGCs in vascular smooth muscle (3), it is a nonselective blocker: Mg2+ also blocks SOCs (9, 61), inwardly rectifying K+ channels (44), and receptor-mediated nonselective cation channels (38). The nonspecific blocking effects of nifedipine and Mg2+ raise the possibility that other Ca2+-entry pathways may contribute to the generation and/or maintenance of MT. Furthermore, some of these channels may influence the membrane potential (Vm) and thus, indirectly, alter LVGC open probability and Ca2+ entry. The present study was therefore designed to explore the effects of Mg2+ and nifedipine specifically on LVGCs and MT.

To investigate the mechanism of action of Mg2+, we tested its effect on LVGC current in arterial myocytes using patch-clamp methods. Extracellular Mg2+ blocks LVGCs in cardiac (24), neuronal (13), and endothelial cells (17). The effects of intracellular Mg2+ are complex. Intracellular Mg2+ modulates LVGCs in cardiac myocytes (1, 2, 54, 55) and Cav1.2 expressed in tsA-201 cells (12). Furthermore, the function of large-conductance Ca2+-activated K+ (BKCa) channels is altered by intracellular Mg2+: the single-channel conductance is reduced and the open probability is increased (10, 36, 65). Surprisingly, the effects of either extracellular or intracellular Mg2+ on the LVGC current of isolated ASM cells have not been extensively examined by electrophysiological methods. It is possible that the intracellular free Mg2+ concentration ([Mg2+]i), as well as [Mg2+]o, regulates the physiology of the ASM cells, at least in part by modulating Ca2+ influx through LVGCs. Therefore, we characterized the effects of [Mg2+]i and [Mg2+]o on LVGC current in isolated mesenteric artery myocytes.



Wild-type (WT) male mice (C57/BL6 background, 12–18 wk) and littermate males of the same age with a mutated Cav1.2 LVGC α1-subunit (T1066Y = DHPR/R) (45), which dramatically reduces DHP sensitivity of Cav1.2 LVGCs, were used for this study. The mice were killed by CO2 followed by cervical dislocation. The animal protocol was approved by the University of Maryland School of Medicine Institutional Animal Care and Use Committee.

Tissue preparation and diameter measurements.

Fourth-order mesenteric small arteries [∼120 μm outside passive diameter (PD)] were isolated, cannulated at both ends, and pressurized (70 mmHg) as described (64). One cannula was connected to a servo-controlled pressure-regulating device (Living Systems, Burlington, VT); the other cannula was attached to a closed stopcock. Arterial segments were incubated with gassed Krebs solution at 33–35°C and were studied in the absence of intraluminal flow. The arteries were viewed with a ×10 objective with bright-field illumination on a Nikon TMS microscope (Nikon, Melville, NY) equipped with a monochrome video charge-coupled device camera. Outside diameter was continuously monitored with an on-line edge detection system that utilizes a video frame grabber and custom LabView software (National Instruments, Austin, TX). PD was determined by incubating the arteries with Ca2+-free solution for 10 min at the end of each experiment.

Arterial myocyte isolation.

Details of the methods are published (11). In brief, the superior mesenteric artery from WT mice was rapidly removed and transferred to a chamber containing ice-cold low-Ca2+ physiological salt solution (PSS) with the following composition (in mM): 140 NaCl, 5.36 KCl, 0.34 Na2HPO4, 0.44 K2HPO4, 10 HEPES, 1.2 MgCl2, 0.05 CaCl2, and 10 glucose, pH 7.2 (adjusted with Tris). The artery was cleaned of fat and connective tissue and placed in fresh low-Ca2+ PSS at 37°C for 30 min. The tissue was digested in low-Ca2+ PSS containing (in mg/ml) 2 collagenase type XI, 0.16 elastase type IV, and 2 BSA for 35 min at 37°C. After digestion, the tissue was washed three times with low-Ca2+ PSS at 4°C. A suspension of single cells was obtained by gently triturating the tissue in low-Ca2+ solution with a fire-polished Pasteur pipette. Dispersed ASM cells were used immediately or were stored at 4°C and used within 4 h. For an experiment, four or five drops of cell suspension were placed on a coverslip and allowed to sit for 10 min at 4°C to permit the cells to adhere to the coverslip. The coverslip was placed in a perfusion chamber mounted on a Nikon Diaphot inverted phase-contrast microscope. The cells were superfused at 2 ml/min with external solution. Only cells with elongated morphology (i.e., relaxed cells) were studied.

Patch-clamp recording.

Whole cell, patch-clamp configuration was applied to record membrane currents with the use of an Axopatch 200 patch-clamp amplifier (Axon Instruments, Union City, CA). Fire-polished micropipettes (1–3 MΩ resistance) were manufactured from borosilicate capillary tubing (Garner Glass, Claremont, CA) using a micropipette puller (model P-97; Sutter Instruments, Novato, CA). An Ag-AgCl reference electrode was connected to the bath by an agar salt bridge containing 1 M KCl. The offset voltage was zeroed immediately before seal formation. Before the pipette approached the cell, the chamber was perfused with Ca2+-free PSS containing 2 mM [Mg2+]o to prevent cell contraction caused by ATP diffusing from the pipette. The pipette was placed in contact with the cell surface, and gentle suction was applied to form a high-resistance (1–1.2 GΩ) seal. After seal formation, the cells were superfused with normal PSS solution. Additional suction was applied to break the membrane under the pipette tip, allowing low-resistance access to the intracellular space. The holding potential was −70 mV in all experiments. Leak and capacity transients were subtracted with the use of a P/4 protocol (8). Series resistance was compensated to give the fastest possible capacity transient without producing oscillations. Membrane currents were recorded with 12-bit analog-to-digital converters (Digidata 1322A; Axon Instruments). Data were sampled at 500 kHz (unless otherwise stated), filtered at 5 kHz with a 902LPF low-pass Bessel filter (Frequency Devices, Haverhill, MA), and stored for subsequent analysis. All records were obtained at room temperature (25–26°C).


Artery dissection solution contained (in mM) 145 NaCl, 4.7 KCl, 1.2 MgSO4·7H2O, 2.0 MOPS, 0.02 EDTA, 1.2 NaH2PO4, 2.0 CaCl2·2H2O, 5.0 glucose, and 2.0 pyruvate (pH was adjusted to 7.4 with NaOH at 5°C). Krebs perfusion solution (Krebs) for diameter measurements contained (in mM) 112 NaCl, 25.7 NaHCO3, 4.9 KCl, 2.5 CaCl2, 1.2 MgSO4·7H2O, 1.2 KH2PO4, 11.5 glucose, and 10 HEPES (pH was adjusted to 7.4 with NaOH). High [K+]o-Krebs was made by replacing 5–70 mM NaCl with equimolar KCl. Ca2+-free solution was made from Krebs solution by omitting Ca2+ and adding 0.5 mM EGTA. For solutions with an elevated [Mg2+]o, MgCl2 was added to the normal Krebs. Solutions were gassed with 5% O2-5% CO2-90% N2; i.e., the solutions were buffered with both the zwitterion, HEPES, and bicarbonate.

To isolate Ca2+ currents, the pipette was filled with high-Cs+ solution of the following composition (in mM): 110 CsCl, 15 NaCl, 2.5 MgCl2, 10 HEPES, 10 EGTA, and 2 Na2ATP (pH adjusted to 7.2 with CsOH). Intracellular solutions containing free Mg2+ concentrations ([Mg2+]i) of 0.06, 0.5, and 2 mM were designed with the use of a computer program (WEBMAXC version 2.40 obtained at∼cpatton/webmaxcS.htm) and by adding the appropriate amounts of Na2ATP and MgCl2. Bath solutions containing different concentrations of Ba2+, Ca2+, and Mg2+ were prepared by substituting BaCl2, CaCl2, or MgCl2 for an isosmotic amount of NaCl in the Ca2+-free reference external solution (see above) of the following composition (in mM): 150 NaCl and 10 HEPES, pH adjusted to 7.4 with NaOH. All experiments were performed in the presence of 200 nM tetrodotoxin (TTX) to block the Na+ currents previously described in these cells (11). The osmolarity of all internal and external solutions was maintained at 271 ± 2 and 296 ± 1 mosmol/l, respectively. Osmolarity was measured with a model 5500 vapor pressure osmometer (Wescor, Logan, UT).


Nifedipine, phentolamine, collagenase, elastase, BSA, and Cremaphor EL were obtained from Sigma-Aldrich (St. Louis, MO). TTX was obtained from Calbiochem (San Diego, CA). TTX was dissolved in deionized water; nifedipine was dissolved in DMSO. Other reagents were reagent grade or the highest grade available.

Data analysis.

Data are expressed as means ± se; n denotes the number of arteries studied with diameter measurements (1 per animal) or the number of cells studied in electrophysiological experiments. Comparisons of data were made by Student's paired or unpaired t-test, as appropriate; differences were considered significant at P < 0.05.

To estimate the time constant (τ) of the K+-induced vasoconstrictions, their time course was fit to a first-order exponential function of the following form: Math(1) where Y is the diameter at time t and Ymax and Ymin are the maximum and minimum diameters, respectively, for a K+-induced vasoconstriction.

Electrophysiological data analysis was performed by pCLAMP software (version 9.0; Axon Instruments). Ca2+ and Ba2+ current (ICa and IBa) amplitudes were measured relative to the current level before the pulse.

The Mg2+ dose-response effect on LVGC current was fit to an equation of the form Math(2) where X is the response (i.e., %block of LVGC current), [Mg2+]o is the extracellular Mg2+ concentration, IC50 is the [Mg2+]o that reduced the response to 50% of control, and Xmax is the apparent maximum response.


DHPR/R mice have normal MT and [K+]o-induced vasoconstriction.

Mesenteric small arteries develop MT after 30- to 45-min perfusion with Krebs solution under 70 mmHg at 33–35°C (63). MT was similar in WT (23 ± 2% of PD; n = 14) and DHPR/R arteries (22 ± 2% of PD; n = 13). To evaluate LVGC activation in these intact, pressurized small-resistance arteries, vasoconstriction was induced by 10–75 mM [K+]o solutions. Elevation of [K+]o from 5 mM (in normal Krebs) to 10–20 mM caused the arteries with MT to dilate; i.e., vasoconstriction was reduced relative to control MT (Fig. 1A). This vasodilation probably is the result of inward rectifier K+ channel-activated membrane hyperpolarization (43). Elevation of [K+]o from 5 to >20 mM caused a transient vasodilation followed by vasoconstriction relative to MT (Fig. 1B) (64). This vasoconstriction is due to depolarization and the activation of LVGCs (30).

Fig. 1.

Vasoconstriction-induced by high external K+ concentration ([K+]o) in wild-type (WT) mouse mesenteric small arteries. A: effect of raising [K+]o in Krebs solution from 5 mM (normal) to 15 mM on diameter from a representative artery (n = 6). Horizontal solid bar (bottom) indicates time of exposure to 15 mM [K+]o. Horizontal dashed line (top) indicates passive diameter (PD), which was determined by incubating arteries with Ca2+-free solution for 10 min at the end of the experiment. MT, myogenic tone; Kpeak, peak vasoconstriction from maximum relaxation to maximum constriction in the presence of [K+]o. B: representative experiment showing the effect of raising [K+]o in Krebs solution from 5 to 45 mM on artery diameter (n = 9). C: Kpeak, normalized to Kmax, as a function of [K+]o (from ∼5 to ∼75 mM). Kmax, Kpeak induced by 75 mM [K+]o. EC50 for high [K+]o-induced vasoconstriction is 39.5 mM (n = 6 arteries).

For quantitative purposes (e.g., Fig. 1C), the amplitude of the high [K+]o-induced vasoconstriction was measured from the maximum vasodilation to the maximum constriction (peak [K+]o-induced vasoconstriction; Kpeak; Fig. 1, A and B). Phentolamine (1 μM), an α-adrenoreceptor antagonist, was present in all high [K+]o solutions to minimize constriction due to neurogenic catecholamine release. The [K+]o dependence of the Kpeak was determined in arteries from WT mice (Fig. 1C). The EC50 ([K+]o for half-maximal vasoconstriction) was 39.3 mM, and 75 mM [K+]o-induced vasoconstriction was defined as the maximum Kpeak. In DHPR/R mouse arteries, the Kpeak induced by 60 mM [K+]o was 46 ± 4% (n = 4) of PD vs. 53 ± 4% (n = 8) of PD in WT arteries (P = 0.135). The similar MT and similar 60 mM [K+]o-induced vasoconstriction imply that [K+]o-dependent vasoconstriction in DHPR/R is not different from that in WT arteries.

Nifedipine block of [K+]o-induced constriction is greatly reduced in DHPR/R mice.

The inhibitory effect of nifedipine on high [K+]o-induced vasoconstriction was compared in WT and DHPR/R mice. Nifedipine, in a concentration-dependent manner, blocked the 60 mM [K+]o-induced vasoconstriction (a near maximum vasoconstriction) in arteries from both WT and DHPR/R mice (Fig. 2, A and B). The nifedipine concentration range needed for block of the vasoconstriction was, however, much higher in DHPR/R (1–100 μM) than in WT arteries (0.001–1 μM). This block by low-dose nifedipine in WT but not in DHPR/R arteries confirmed that high [K+]o-induced vasoconstriction is mediated by DHP-sensitive Cav1.2 LVGCs. The summarized data in Fig. 2C indicate that the apparent affinity of nifedipine to LVGCs was reduced by ∼1,000-fold; i.e., the IC50 shifted from ∼0.01 μM in WT to ∼23 μM in DHPR/R mouse arteries. For example, 1 μM nifedipine inhibited the 60 mM [K+]o-induced vasoconstriction completely in WT arteries but inhibited this vasoconstriction by only 2% in DHPR/R arteries. Thus the inhibitory effect of low-dose nifedipine (<1 μM) on high [K+]o-induced vasoconstriction in the WT arteries is due solely to its inhibition of Cav1.2 LVGCs.

Fig. 2.

Effects of nifedipine (Nif) on 60 mM [K+]o-induced vasoconstriction in mesenteric small arteries from WT and dihydropyridine-resistant Cav1.2 (DHPR/R) mice. A: representative experiment showing the dose-dependent inhibition of nifedipine (0.001–1 μM) on 60 mM [K+]o-induced vasoconstriction in a WT mouse artery (n = 6). B: representative experiment showing the dose-dependent inhibition of nifedipine (1–100 μM) on 60 mM [K+]o-induced vasoconstriction in a DHPR/R mouse artery (n = 4). C: inhibition of [K+]o-induced vasoconstriction by nifedipine is shifted dramatically to the right in DHPR/R arteries (gray line) compared with that in WT arteries (black line).

Nifedipine markedly reduces MT in WT but not in DHPR/R arteries.

The effects of nifedipine on MT in representative WT and DHPR/R arteries are illustrated in Fig. 3, A and B. Nifedipine (0.3 μM) rapidly reduced MT in WT arteries, but the recovery from nifedipine exposure was slow. Summarized data in Fig. 3B indicate that nifedipine inhibited MT by 78% in WT arteries. In contrast, in arteries from DHPR/R mice, nifedipine did not block MT (Fig. 3, C and D). Moreover, treatment of WT arteries, but not DHPR/R arteries, with 0.3 μM nifedipine for 2 min not only reduced MT but also prevented most of the vasoconstriction normally induced by subsequent application of 75 mM [K+]o (Fig. 3, E and F). Taken together, these data demonstrate that Ca2+ entry via Cav1.2 LVGCs is obligatory for MT. The complete absence of nifedipine effects in DHPR/R arteries demonstrates that it did not inhibit MT by interfering with other mechanisms, such as SOCs, but exclusively by inhibiting LVGCs. This emphasizes the predominant role of the Cav1.2 isoform in MT but does not rule out a minor contribution of other Ca2+-entry pathways.

Fig. 3.

Effects of nifedipine and high extracellular Mg2+ concentration ([Mg2+]o) on MT and 75 mM [K+]o-induced vasoconstriction in mesenteric small arteries from WT and DHPR/R mice. A: effects of 0.3 μM (black line) and 10 mM [Mg2+]o (gray line) on MT from a representative WT mouse artery. B: summarized data showing effects of 0.3 μM nifedipine and 10 mM [Mg2+]o on MT in WT mouse arteries. Numbers in parentheses indicate arteries studied. Ctrl, control. C: effects of 0.3 μM nifedipine (black line) and 10 mM [Mg2+]o (gray line) on MT from a representative DHPR/R mouse artery. D: summarized data showing the effects of 0.3 μM nifedipine and 10 mM [Mg2+]o on MT in DHPR/R mouse arteries. Numbers in parentheses indicate arteries studied. E: both 0.3 μM nifedipine (black line; n = 6) and 10 mM [Mg2+]o (gray line; n = 8) reduced MT, but only nifedipine blocked the 75 mM [K+]o-induced vasoconstriction in WT mouse mesenteric small arteries. K+Mg, 75 mM [K+]o-induced vasoconstriction in the presence of 10 mM [Mg2+]o; K+Nif, 75 mM [K+]o-induced vasoconstriction in the presence of 0.3 μM nifedipine. F: 10 mM [Mg2+]o (gray line; n = 2), but not 0.3 μM nifedipine (black line; n = 4), blocked MT in DHPR/R mouse arteries. Neither nifedipine nor [Mg2+]o affected the 75 mM [K+]o-induced vasoconstriction. ***, P < 0.001.

High [Mg2+]o reduces MT and alters high [K+]o-induced vasoconstriction.

Elevation of [Mg2+]o to 10 mM reduced MT by about the same extent (80%) as did 0.3 μM nifedipine in WT arteries (Fig. 3, B and C); however, Mg2+ was equally effective in DHPR/R arteries (Fig. 3D). Based on the aforementioned evidence that LVGCs play a critical role in MT, this is consistent with block of LVGCs by high [Mg2+]o. In contrast to nifedipine, however, 10 mM [Mg2+]o had little effect on the 75 mM [K+]o-induced vasoconstriction (Fig. 3, E and F).

The reduction of MT, but apparent lack of inhibition of the 75 mM [K+]o-induced vasoconstriction by 10 mM [Mg2+]o (Fig. 3, E and F; see Ref. 64), raises questions about the mechanism by which high [Mg2+]o blocks MT. Is it by blocking some other Ca2+-entry pathways or by blocking LVGCs and/or by shifting LVGC gating voltage? To explore this issue, we examined the effects of 10 mM [Mg2+]o on vasoconstrictions induced by 10–75 mM [K+]o in WT arteries. Representative superimposed vasoconstrictions in normal [Mg2+]o (1.2 mM) or high [Mg2+]o (10 mM) at 20, 35, 45, and 75 mM [K+]o are shown in Fig. 4, AD, respectively. Raising [Mg2+]o from 1.2 to 10 mM, in addition to reducing MT, slowed the development of the vasoconstriction induced by elevated [K+]o, especially at [K+]o between 25 and 45 mM (Fig. 4, B, C, and E). At 75 mM [K+]o, however, [Mg2+]o had relatively little effect on [K+]o-induced vasoconstriction (Fig. 4, D, E, and F). Summarized data reveal that high [Mg2+]o markedly slowed the rate of development of the vasoconstriction, i.e., increased the time constant (τ; Fig. 4E), and decreased the amplitude (Fig. 4F) of the high [K+]o-induced vasoconstrictions.

Fig. 4.

Effects of 10 mM [Mg2+]o on MT and 10–75 mM [K+]o-induced vasoconstrictions in WT mouse mesenteric small arteries. AD: representative experiments showing, respectively, 20, 35, 45, and 75 mM [K+]o-induced vasoconstrictions in 1.2 mM (black lines) or 10 mM [Mg2+]o (gray lines); n = 4–8 arteries in each concentration. KCtrl, maximum vasoconstriction induced by elevated [K+]o, measured from PD to minimum diameter in the presence of high [K+]o. K+Mg = [K+]o-induced vasoconstriction in the presence of 10 mM [Mg2+]o, measured from PD to minimum diameter in the presence of both elevated [K+]o and [Mg2+]o. Horizontal bars indicate time of exposure to [Mg2+]o and [K+]o. E: time constant of [K+]o-induced vasoconstriction in 1.2 mM (Ctrl; n = 8) or 10 mM [Mg2+]o (n = 4). F: dose response of [K+]o-induced vasoconstriction in 1.2 mM (Ctrl) or 10 mM [Mg2+]o; n = 5 arteries.

Replacement of HCO3 and PO43− by zwitterions such as HEPES has been reported to attenuate [K+]o-induced arterial contraction (5). In contrast, our solutions contained HEPES as well as HCO3 and PO43− (materials and methods). HEPES was used to maintain external pH because the tissue chamber (1 ml volume, 3.4-cm2 surface area) is open and the partial pressure of CO2 tends to decline as the fluid flows through (2.8 ml/min). HCO3 was required for C1/HCO3 exchange to maintain intracellular pH (29). The magnitude of the high [K+]o-induced vasoconstriction and inhibition by [Mg2+]o were maintained when HEPES was removed (data not shown). Thus the previously reported results (5) may have been because of substitution of HCO3, per se, rather than because of the presence of zwitterions such as HEPES.

LVGC current characteristics in arterial myocytes.

The preceding data are consistent with the view that LVGCs are partially blocked by high [Mg2+]o. To obtain direct evidence for Mg2+ inhibition of LVGCs, we investigated the effects of extracellular and intracellular Mg2+ on LVGC current using the whole cell patch-clamp technique. In freshly isolated mesenteric artery myocytes from WT mice, the inward current consists of two distinct components: a fast transient Na+ current through voltage-gated Na+ channels and a more slowly activating, DHP-sensitive ICa (11). To isolate ICa, all of the following experiments were carried out in the presence of 200 nM TTX, which is sufficient to block the Na+ current (11). Inward currents recorded in the presence of 10 mM external Ca2+ concentration ([Ca2+]o), i.e., ICa, were activated at voltages positive to −30 mV, reached a maximum amplitude at about +30 mV, and exhibited a reversal potential of approximately +70 mV, as illustrated in the current-voltage curve in Fig. 5A. The mean maximum current amplitude was −58 ± 15 pA (n = 7).

Fig. 5.

L-type voltage-gated Ca2+ channel (LVGC) currents in isolated myocytes from WT mouse mesenteric arteries. A: inward currents were recorded during 5-ms steps to potentials from −50 to +110 mV with either 10 mM external Ca2+ concentration ([Ca2+]o; ○, ICa) or 10 mM [Ba2+]o (•, IBa) present as the current carrier. Current amplitude is plotted as a function of voltage. B: inward current (IBa) generated by a 5-ms step to +10 mV in the absence (Ctrl) or presence of 10 μM nifedipine.

When 10 mM Ba2+ was substituted for extracellular Ca2+, the amplitude of IBa increased about threefold (mean maximum amplitude of −148 ± 33 pA; n = 8). The current-voltage relationship (Fig. 5A) shows that IBa was activated at voltages positive to −50 mV. The maximum IBa was observed at about +10 mV with a reversal potential at about +70 mV. Figure 5B shows IBa recorded in 10 mM [Ba2+]o during a step to 0 mV from a holding potential of −70 mV, before and during application of 10 μM nifedipine. Nifedipine blocked almost all of the TTX-resistant inward current (ICa and IBa) (n = 5 and 4, respectively), suggesting that LVGCs are responsible for these currents.

Elevated [Mg2+]o alters the activation and magnitude of LVGC current.

Next, we examined the effect of [Mg2+]o on ICa and IBa. Experiments in which the current was essentially the same before and after Mg2+, i.e., when the effect of Mg2+ was reversible, were considered “successful”; this was the case in 41 of 49 cells from 14 different mice. The representative superimposed current traces in Fig. 6, AC, show the effect of increasing [Mg2+]o from 0 to 20 mM on the LVGC current evoked by depolarization to +10 mV when the bath solution contained 10 mM [Ca2+]o or 5 or 10 mM [Ba2+]o. The currents were recorded before, during, and after increasing [Mg2+]o. The addition of 20 mM [Mg2+]o decreased the amplitude of the current by ∼50% in 10 mM [Ca2+]o and by ∼80% and ∼70% in 5 and 10 mM [Ba2+]o, respectively. Removal of Mg2+ from the bath solution restored the current within a few minutes. The relationship between [Mg2+]o and the percent block of ICa or IBa is plotted in Fig. 6D. The smooth curves are the fit of the data to Eq. 2, with IC50 at 14.9 mM when the external solution contained 10 mM Ca2+. The IC50 was 1.2 and 0.25 mM when cells were superfused with 10 and 5 mM [Ba2+]o, respectively. Thus [Mg2+]o inhibited both IBa and ICa in a concentration-dependent manner, but Mg2+ reduced IBa more effectively than it did ICa.

Fig. 6.

High [Mg2+]o blocks LVGC current. Current traces were generated by 10-ms steps to +10 mV in 0 mM [Mg2+]o (Ctrl), 20 mM [Mg2+]o, and after returning to 0 mM [Mg2+]o (Wash) in the presence of 10 mM [Ca2+]o (A), 5 mM [Ba2+]o (B), or 10 mM [Ba2+]o (C). D: Percent block of LVGC current (ICa or IBa) by high [Mg2+]o, averaged for 3–5 cells from different arteries, is plotted as a function of [Mg2+]o. Percentage block = [(control current − LVGC current + Mg2+)/control current] × 100.

Elevating [Mg2+]o to 20 mM also caused a positive shift in the voltage dependence of LVGC activation when the currents were measured with 10 mM [Ba2+]o; this probably reflects a decrease in the negative surface potential. Figure 7 shows a plot of the normalized IBa tail current as a function of Vm for a representative experiment in 10 mM [Ba2+]o before and during 20 mM [Mg2+]o perfusion. Mg2+ shifted the Vm at which the current was activated to more positive potentials (from less than −30 mV to greater than −10 mV). The midpoint of the IBa tail current-Vm relationship shifted from 16 ± 2.7 to 23 ± 2.2 mV in the presence of 20 mM [Mg2+]o (P < 0.01, n = 4). High [Mg2+]o similarly shifted the activation for ICa, but the current amplitudes were much smaller and difficult to quantitate.

Fig. 7.

High [Mg2+]o shifts the voltage dependence of IBa in the positive direction. Shown in graph is normalized Ca2+ tail current (Itail; carried by 10 mM [Ba2+]o) as a function of membrane potential (Vm) in control conditions (•) and during the application of 20 mM [Mg2+]o (○) from a representative myocyte. Itail data were fitted with a Boltzmann equation (continuous lines).

LVGC current is inversely related to [Mg2+]i.

The effects of intracellular Mg2+ were also examined. Figure 8 shows the effects of different [Mg2+]i on the maximum ICa and IBa amplitudes. When cells were superfused with solution containing 5 mM [Ba2+]o and the patch electrode contained 0.5 mM [Mg2+]i, a concentration close to the normal basal [Mg2+]i in smooth muscle cells (62), the maximum current was −122 ± 18 pA (n = 21). The IBa was increased by ∼2.5-fold in cells dialyzed with low [Mg2+]i (0.06 mM). In contrast, increasing [Mg2+]i to 2 mM reduced the mean IBa amplitude by 2.8-fold. Low [Mg2+]i similarly increased the current recorded in the presence of 10 mM [Ca2+]o (Fig. 8), whereas ICa at high [Mg2+]i was too small to measure accurately.

Fig. 8.

Effects of changing intracellular Mg2+ concentration ([Mg2+]i) on ICa and IBa amplitudes. The maximum ICa and IBa current amplitudes (Imax) are given in the presence of different [Mg2+]i. Cells were superfused with a bath solution containing 5 mM [Ba2+]o or 10 mM [Ca2+]o, and the patch electrode contained 0.06, 0.5, or 2 mM [Mg2+]i. Currents were generated by a test pulse to +10 mV from a holding potential of −70 mV and measured at 1.5–2.0 min after patch rupture. Numbers in parentheses indicate arteries studied. The mean current amplitudes are significantly different (***P < 0.001; **P < 0.01).

The time course of the IBa amplitude during dialysis following patch rupture also depended on [Mg2+]i. When cells were superfused with a solution containing 5 mM [Ba2+]o and the patch pipette contained 0.5 mM [Mg2+]i (control conditions), a slow rundown of IBa occurred during intracellular dialysis. This rundown was usually observed during whole cell recordings of ICa; it likely was caused by a combination of factors resulting from a change in the intracellular environment during cell dialysis. In Fig. 9, the IBa amplitude is plotted as a function of time starting ∼60 s after patch rupture in representative cells dialyzed with either low (0.06 mM, n = 20 cells) or high (2 mM, n = 16 cells) [Mg2+]i from four different mice. When ASM cells were dialyzed with a pipette solution containing low [Mg2+]i, IBa initially increased in amplitude, reached a peak within 2–2.5 min after patch rupture, and then slowly ran down (Fig. 9A; recordings of IBa taken at the times indicated in Fig. 9A are shown in Fig. 9B). In myocytes dialyzed with high [Mg2+]i, the IBa amplitude following patch rupture was much smaller than with low [Mg2+]i, and little rundown was observed.

Fig. 9.

Time course of IBa amplitude after cell dialysis with high or low [Mg2+]i. A: peak amplitude of IBa was measured during a step to +10 mV from a holding potential of −70 mV during dialysis with pipette solutions containing 0.06 or 2 mM [Mg2+]i, and IBa is plotted as a function of time beginning 1 min after patch rupture. Cells were superfused with bath solution containing 5 mM [Ba2+]o. Time 0 corresponds to the current recorded at ∼60 s after patch rupture. Numbers in parentheses (1–3) correspond to the sample currents displayed in B.


Key role of Cav1.2 channels in MT; evidence from Cav1.2 DHPR/R mice.

The origin and maintenance of vascular MT have important implications for understanding the control of BP and blood flow. Elucidation of the underlying cellular and molecular mechanisms that generate and maintain MT has obvious application, especially in the therapy of hypertension. The fundamental principles governing MT relate to the fact that vasoconstriction depends on the cytosolic Ca2+ concentration and on the Ca2+ sensitivity of the contractile machinery. The view that MT is mediated mainly by membrane depolarization and LVGC activation, a major Ca2+ influx pathway, is widely accepted (16, 39). A confusing aspect, however, is that several VGCs are expressed in ASM cells and appear to play roles in regulating vascular tone and arterial contraction (7, 21, 48, 50). One way to determine which VGC subtype(s) mediates the development and/or maintenance of MT is to use specific channel gene knockout mice. In this context, Moosmang et al. (35) generated mice with a conditional, tamoxifen-induced, smooth muscle-specific Cav1.2 gene. Resistance arteries from these mice exhibit very little myogenic constriction in response to increased intravascular pressure. Interpretation of these data may be ambiguous, however, because the mice have very low BP, show signs of severe illness, and die within 4–5 wk of tamoxifen injection, probably due to shock and other complications of bowel paralysis (35).

To circumvent these problems, we studied small mesenteric arteries from WT mice and mice genetically engineered to express functionally normal but DHP-resistant Cav1.2 (DHPR/R) (45). Thus, the role of Cav1.2 could be verified by comparing the different effects of nifedipine on MT and high [K+]o-induced vasoconstrictions in arteries from WT and DHPR/R mice. Despite the greatly reduced DHP sensitivity of Cav1.2, the expression and function of these channels are unaltered in DHPR/R mice (45). Moreover, these mutant mice are physiologically normal with normal BP (data not shown), MT, and LVGC-mediated vasoconstriction (Figs. 2 and 3; see also Ref. 45). Therefore, the altered DHP sensitivity can be used to separate the effects of DHP on Cav1.2 from its effects on other channels.

Selective blocking effect of nifedipine on Cav1.2 channels and MT.

Nifedipine (0.3 μM) reduced MT by ∼80% and blocked virtually all of the high [K+]o-induced vasoconstriction in WT arteries but not in Cav1.2 DHPR/R mice (Figs. 2 and 3). These findings demonstrate directly the dominant role of Cav1.2 in depolarization-induced Ca2+ entry and in the development and/or maintenance of MT in mouse mesenteric small arteries. These results apparently rule out significant participation of other VGCs such as Cav1.3, Cav2, and Cav3. The high [K+]o data imply that the DHP receptors on Cav1.2 channels are necessary and sufficient for nifedipine block of VGCs in these arteries. The lack of blocking effect of nifedipine on MT in Cav1.2 DHPR/R mice also indicates that nifedipine blocks MT by inhibiting, exclusively, Cav1.2 and not other Ca2+-entry channels (64). Nevertheless, much evidence indicates that other mechanisms that control, directly or indirectly, the availability of Ca2+ also contribute to MT; examples include SOCs (20, 42, 53, 64), the α2-isoform of the Na+ pump (63), and the Na+/Ca2+ exchanger (27, 63). On the other hand, mechanisms that affect Vm, including cation entry through stretch-activated channels, SOCs, and BKCa channels, will indirectly affect the large, Cav1.2-mediated component of MT (25, 28, 34).

High [Mg2+]o attenuates MT.

On the basis of these findings, the action of Mg2+ appears anomalous. Mg2+ blocks LVGCs in arterial (33), as well as urethral (26) and tracheal (47), smooth muscle. This cation is nonselective, however, and also blocks other cation-entry pathways in arteries (9, 38, 61).

In WT arteries, 10 mM Mg2+ reduced MT to about the same extent as did 0.3 μM nifedipine. As anticipated, Mg2+ also reduced MT in DHPR/R arteries (Fig. 3). This is consistent with block of LVGCs (Cav1.2 channels) by Mg2+ binding in the pore and not at the DHP-binding site.

In conduit and muscular arteries, high [Mg2+]o inhibits high [K+]o-induced contractions (3, 37, 49). Our observations are comparable, even though we studied small arteries under isobaric conditions (this report and Ref. 64), whereas the larger vessels were examined under isometric conditions (3, 37, 49).

Effects of high [Mg2+]o are attenuated at very high [K+]o.

In contrast to the uniformly inhibitory effect of nifedipine on high [K+]o-induced vasoconstriction, we observed that 10 mM Mg2+ had little effect at the highest [K+]o concentrations tested (≥60 mM) in mouse and rat mesenteric small arteries (Fig. 4) (64). In addition, in large arteries, high [Mg2+]o (4.8 mM) decreased K+-induced contractile responses and increased the EC50 for Ca2+ but had little effect on the maximal contractile tension at the highest [K+]o (49). Clearly, the effects of [Mg2+]o on [K+]o-induced contraction are complex. Electrophysiological experiments were therefore undertaken to test, directly, the effects of high [Mg2+]o on LVGC current.

One possibility is that Mg2+ attenuates the K+-induced depolarization that opens LVGCs. Another possibility that seems more compatible with block of LVGCs, however, is that the block of LVGCs may be voltage dependent, with reduced block at more depolarized potentials (Ref. 13, and see below). Furthermore, because of a surface charge effect, high [Mg2+]o can be expected to shift LVGC activation in the depolarizing direction (Fig. 7, and see below) (18, 47). Thus, at small depolarizing voltages, LVGC activation kinetics will be slowed when [Mg2+]o is elevated. This may explain the slow K+-induced vasoconstriction in 10 mM Mg2+, at [K+]o <45 mM (Fig. 4). The reduced amplitude of these vasoconstrictions (measured from PD; Fig. 4) at [K+]o ≤50 mM in 10 mM Mg2+ can also be explained by a high [Mg2+]o-induced shift in the activation of LVGCs in the depolarizing direction. Thus, in high [Mg2+]o, fewer LVGCs should be opened by small depolarizations. At higher [K+]o (i.e., larger depolarizations), however, the effect of high [Mg2+]o on channel activation should be minimal, as we observed (Fig. 4).

The inhibitory effects of extracellular Mg2+ on vascular tone also could be due to a secondary rise in [Mg2+]i and to activation of BKCa channels and membrane hyperpolarization (10). Nevertheless, our observations are all consistent with the view that Mg2+ blocks LVGCs and can explain the inhibitory effects of Mg2+ on vascular tone and on BP (6, 19, 46). Electrophysiological experiments are needed, however, to confirm directly the effects of Mg2+ on these channels. Here, we show that concentrations of Mg2+ in the physiological range have significant effects on LVGC current in ASM cells.

Block of Ca2+ channels by extracellular Mg2+.

An increase in [Mg2+]o blocks LVGCs in a variety of cell types, including some smooth muscles (e.g., Refs. 26, 47). Although high [Mg2+]o is known to block LVGCs in vascular smooth muscle (3), this has not been well documented with electrophysiological methods. Langton and Standen (33) reported that 1 mM [Mg2+]o decreased ICa in rat basilar artery cells by 17%. Our results show that [Mg2+]o decreased both IBa and ICa in ASM cells in a dose-dependent manner (Fig. 6). The observed IC50 for Mg2+ block of ICa in ASM cells superfused with 10 mM Ca2+ (14.9 mM) is close to the value of 12 mM in uterine smooth muscle cells superfused with 2 mM Ca2+ (40). Extracellular Mg2+ was a more effective blocker of IBa than ICa in ASM cells (Fig. 6), as in taenia caeci smooth muscle (22). This is probably because Mg2+, Ba2+, and other divalent cations all bind to the same site as Ca2+ in the channel pore (4, 32), but Ba2+ binds less tightly than Ca2+ so that it is easier for Mg2+ to displace Ba2+ and block the channel.

External Mg2+ is an open-channel blocker in rat phaeochromocytoma cells: it produces discrete block of inward single-channel currents through LVGCs by binding to the same high-affinity site as Ca2+ (32). Block of ICa by Mg2+ in sensory neurons is maximal at negative potentials and decreases at more positive potentials, suggesting that Mg2+ does not traverse the channels (13).

Mg2+ shifts the voltage dependence of ICa.

High [Mg2+]o (20 mM) caused a positive shift of 7 mV in the potential at which the conductance (measured as the normalized tail current) is half maximal. A similar shift has been observed in capillary endothelial cells (17), tracheal smooth muscle (47), and ventricular myocytes (18). This shift is likely due to neutralization of surface charge by the divalent cations, thereby reducing the local electric field affecting the channel voltage sensors. Hence, the effect of high [Mg2+]o on ICa is consistent with shielding of surface charge together with simple competitive blockade of the channel pore.

[Mg2+]i modulates LVGC current.

The effect of intracellular Mg2+ on whole cell LVGC current in ASM cells has not previously been investigated. In cardiac myocytes, changes in [Mg2+]i have significant effects on ICa. Increasing [Mg2+]i between ∼0.3 and 7 mM decreases LVGC current in frog and guinea pig cardiac myocytes and Cav1.2 expressed in tsA-201 cells (12, 41, 54, 56, 57). These results are consistent with the effect of [Mg2+]i on ICa in ASM cells reported here. A higher concentration of Mg2+, at least 5–10 mM [Mg2+]i, is needed to produce discrete single-channel blocking events (32). Thus the inhibitory effect of low [Mg2+]i is probably not the result of block of the LVGC channel pore. Another possibility is that the inhibitory effect of [Mg2+]i in ASM is related to its requirement in phosphorylation reactions. In cardiac myocytes, LVGC activity is modulated by phosphorylation, and the extent of phosphorylation appears to be an important factor in the effect of [Mg2+]i on ICa (41, 51, 54). Inhibition of phosphorylation abolishes the effect of [Mg2+]i on ICa (41); conversely, enhancement of phosphorylation augments the effect of [Mg2+]i on ICa (51). In addition, in vascular smooth muscle, Mg2+-regulated Ca2+ transport may be influenced by various cellular signaling pathways, including protein kinase C (58), tyrosine and mitogen-activated protein kinases (60), and nitric oxide (59).

We observed that reduction of [Mg2+]i from 0.5 to 0.06 mM increased the maximum IBa and ICa by ∼2.5-fold. This result indicates that not all LVGCs are active in the physiological range of [Mg2+]i and that small changes in [Mg2+]i can modify ICa significantly. When patch rupture with a pipette solution containing 0.06 mM Mg2+ occurred, the current amplitude first increased, presumably due to the reduction in [Mg2+]i; it reached a maximum in ∼2 min and then slowly ran down. The maximum ICa (in 10 mM [Ca2+]o) was reduced when [Mg2+]i was increased from 0.06 to 0.5 mM.

In conclusion, the results reported here provide new evidence that Cav1.2 channels play a dominant role in the generation and maintenance of MT. Our data demonstrate that the entire inhibitory effect of ≤1.0 μM nifedipine on MT is attributable to block of Cav1.2 channels. These channels are also the target of both extracellular and intracellular Mg2+. Although Mg2+ is known to affect several types of channels, our data indicate that the effects of [Mg2+]o on LVGCs appear to account for most, if not all, of the inhibition of MT and depolarization-induced vasoconstriction caused by K+.


This work is supported by National Heart, Lung, and Blood Institute Grants HL-045215 and HL-078870 to M. P. Blaustein, Austrian Science Fund Grants P-17109 and P-17159 to J. Striessnig, and an American Heart Association Postdoctoral Fellowship to J. Zhang.


We thank Holeatheia Rene for genotyping transgenic mice.

Present address of R. Berra-Romani: Department of Biomedicine, School of Medicine, Benemérita Universidad Autónoma de Puebla, Puebla 72000, México.


  • * J. Zhang and R. Berra-Romani contributed equally to this work.

  • 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.


View Abstract