Effective contractile response to voltage-gated Na⁺ channels revealed by a channel activator

Abstract

This study investigated the molecular identity and impact of enhancing voltage-gated Na⁺ (Na_V) channels in the control of vascular tone. In rat isolated mesenteric and femoral arteries mounted for isometric tension recording, the vascular actions of the Na_V channel activator veratridine were examined. Na_V channel expression was probed by molecular techniques and immunocytochemistry. In mesenteric arteries, veratridine induced potent contractions (pEC₅₀ = 5.19 ± 0.20, E_max = 12.0 ± 2.7 mN), which were inhibited by 1 μM TTX (a blocker of all Na_V channel isoforms, except Na_V1.5, Na_V1.8, and Na_V1.9), but not by selective blockers of Na_V1.7 (ProTx-II, 10 nM) or Na_V1.8 (A-80347, 1 μM) channels. The responses were insensitive to endothelium removal but were partly (∼60%) reduced by chemical destruction of sympathetic nerves by 6-hydroxydopamine (2 mM) or antagonism at the α₁-adrenoceptor by prazosin (1 μM). KB-R7943, a blocker of the reverse mode of the Na⁺/Ca²⁺ exchanger (3 μM), inhibited veratridine contractions in the absence or presence of prazosin. T16A_inh-A01, a Ca²⁺-activated Cl⁻ channel blocker (10 μM), also inhibited the prazosin-resistant contraction to veratridine. Na_V channel immunoreactivity was detected in freshly isolated mesenteric myocytes, with apparent colocalization with the Na⁺/Ca²⁺ exchanger. Veratridine induced similar contractile effects in the femoral artery, and mRNA transcripts for Na_V1.2 and Na_V1.3 channels were evident in both vessel types. We conclude that, in addition to sympathetic nerves, Na_V channels are expressed in vascular myocytes, where they are functionally coupled to the reverse mode of Na⁺/Ca²⁺ exchanger and subsequent activation of Ca²⁺-activated Cl⁻ channels, causing contraction. The TTX-sensitive Na_V1.2 and Na_V1.3 channels are likely involved in vascular control.

voltage-gated Na⁺ (Na_V) channels are central to the initiation and propagation of action potentials in excitable cells, specifically neurons, as well as cardiac and skeletal myocytes (5). Nine Na_V channel isoforms have been identified (Na_V1.1–Na_V1.9) on the basis of the underlying gene (SCNA1-9), rate of inactivation, and sensitivity to TTX; Na_V1.5, Na_V1.8, and Na_V1.9 channels are much less sensitive to blockade by TTX (6, 14). Each Na_V channel consists of a pore-forming ∼260-kDa α-subunit that is associated with one or more auxiliary ∼35-kDa β-subunit(s) (6).

In vascular smooth muscle cells, the presence and impact of Na_V channels are less well defined. Na_V channels have been characterized in freshly dispersed smooth muscle cells from rat or mouse mesenteric artery and portal vein (3, 27, 34) and are particularly prominent in cultured smooth muscle cells (26, 29, 31). There is also some molecular evidence supporting the expression of Na_V channel isoforms in the smooth muscle of portal vein, which displays spontaneous and rhythmic contractions (34), as well as in rat aorta (12) and mouse cremaster arterioles (11). The physiological role of Na_v channels in the vasculature is less well defined, because channel blockers such as TTX generally have little effect on vasoconstrictor responses (12, 34). However, a new blocker of persistent Na_v currents, F 15845, prevents hypoxia-induced contractions of rat femoral artery (4), and the Na_V channel activator veratridine enhances spontaneous contractions in mouse portal vein (34) or causes contraction of rat aorta (12). In the latter case, high concentrations of veratridine were required, and much of the effect was considered to be due to release of neurotransmitters from sympathetic nerve terminals (12).

In the present study we have determined the expression profile of Na_V genes in the rat mesenteric resistance vessel, as well as the rat femoral artery, and have assessed the effect of the Na_V channel activator veratridine on these vessels. We then utilized a panel of pharmacological agents selective for different Na_V channel isoforms to determine the molecular identity of the channels contributing to changes in contractile state. We also used a Na⁺/Ca²⁺ exchange (NCX) inhibitor and a specific blocker of TMEM16A-encoded Ca²⁺-activated Cl⁻ channels [2-(5-ethyl-4-hydroxy-6-methylpyrimidin-2-ylthio)-N-(4-(4-methoxyphenyl)thiazol-2-yl)acetamide (T16A_inh-A01)] (9) to define the vascular mechanisms underlying veratridine-induced contractions. These studies revealed that Na_v channel activator produces robust contractions of resistance arteries, which were only partially driven by increased release of sympathetic neurotransmitters, and involved entry of Ca²⁺ by reverse-mode NCX and the subsequent activation of Ca²⁺-activated Cl⁻ channels. These findings show that “silent” Na_V channels represent a large contractile reserve with important consequences for vascular tone.

METHODS

Isometric tension recording.

Male Wistar rats (250–350 g; Charles River, Kent, UK) were stunned by a blow to the back of the neck and killed by cervical dislocation. All animal care and use were in accordance with the protocols approved under the UK Animal (Scientific Procedures) Act (1986). The third-order branches of the superior mesenteric artery were removed and cleaned of adherent tissue. Segments (2 mm long) were mounted in a Mulvany-Halpern-type wire myograph (model 610M, Danish Myo Technology, Aarhus, Denmark) and maintained at 37°C in gassed (95% O₂-5% CO₂) Krebs-Henseleit solution (mM: 118 NaCl, 4.7 KCl, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, 2 CaCl₂, and 10 d-glucose), as previously described, unless otherwise stated (16). Arteries were equilibrated and set to a basal tension of 2–2.5 mN. The integrity of the endothelium was assessed by precontraction of the vessel with 10 μM methoxamine (an α₁-adrenoceptor agonist) followed by relaxation with 10 μM carbachol (a muscarinic acetylcholine receptor agonist); vessels showing >90% relaxation were designated endothelium-intact. When endothelium was not required, it was removed by mechanical abrasion of the intima with a human hair; <10% carbachol-induced relaxation indicated successful removal. After the test for endothelial integrity, arteries were left for ≥30 min before they were exposed to veratridine.

Preliminary experiments indicated that veratridine induced contractions in some, but not all, quiescent mesenteric arteries, but consistent contractions were obtained when the vessels were prestimulated with a low concentration (1–3 μM) of methoxamine, which induced a small vascular tone. Thus all experiments, unless otherwise stated, were performed under a small methoxamine tone (average ∼2 mN, ∼15% of contraction attained in the endothelial integrity test at the start of each experiment). To obtain a concentration-contraction curve, veratridine (1–30 μM) was cumulatively added to the myograph bath, and the maximal tension achieved at each concentration was determined. To examine the involvement of Na_V channel subtypes, vessels were first stimulated with methoxamine and then with a Na_V channel inhibitor [TTX, ranolazine, ProTx-II, or 5-(4-chlorophenyl)-N-(3,5-dimethoxyphenyl)-2-furancarboxamide (A803467)] or other putative modulators {2-[2-[4-(4-nitrobenzyloxy)phenyl]ethyl]isothiourea (KB-R7943 mesylate) or suramin} for 5–10 min followed by a single concentration (10 μM) of veratridine. The application of ranolazine and, in some cases, KB-R7943 was found to have relaxant effects. Thus the concentrations of methoxamine were adjusted, where necessary, to achieve a similar vascular tone (∼1–3 mN), in the absence and presence of the modulator, before the addition of veratridine. Some of these modulators were also tested against veratridine in mesenteric arteries precontracted with 60 mM KCl. All experiments were performed in matched vessels; effects of putative modulators or endothelium removal were compared with the control responses obtained in vessels of the same rat. The ethanol vehicle of veratridine [maximum 0.06% (vol/vol)] alone had no contractile effects.

To desensitize perivascular sensory nerves, some vessels were treated for 1 h with 10 μM capsaicin, which was then washed out. Capsaicin is an agonist of transient receptor potential vanilloid type 1 (TRPV1), and the prolonged treatment in this protocol was demonstrated to successfully abolish activation of TRPV1-expressing sensory nerves and the subsequent vasorelaxation (18). To test the involvement of sympathetic nerves, 10 μM veratridine was tested in the presence of the α₁-adrenoceptor antagonist prazosin (1 μM). In these experiments, a complete blockade of contractions to 10 μM methoxamine verified an effective treatment, and a small vascular tone was induced by 5-HT (0.3–1 μM) before addition of veratridine in control and prazosin-treated vessels. Responses to veratridine were also tested after treatment with chemical sympatholytics, guanethidine (5 μM for 30 min) (33) or 6-hydroxydopamine (2 mM for 20 min, with 0.3% ascorbic acid) (36), followed by a small 5-HT (0.3–1 μM)-induced tone. The control vessels for 6-hydroxydopamine treatment were incubated with 0.3% ascorbic acid.

In separate sets of experiments, rat femoral artery and aorta were isolated and mounted in a myograph (Danish Myo Technology), as described above; basal tensions of 5 mN (4) and 10 mN (7) were used for femoral artery and aorta, respectively.

RNA extraction, cDNA synthesis, and PCR.

To explore the expression of Na_V isoforms (Na_V1.1–1.9), we used conventional RT-PCR of total RNA extracted from the isolated vessels. Three separate experiments (using vessels from 3 rats) were performed. Similar to previous studies (8, 22, 28), total RNA was extracted from rat mesenteric and femoral arteries using the RNeasy Micro kit (Qiagen) according to the manufacturer's instruction, which includes an on-column DNase treatment step. RNA quality was measured using a spectrophotometer (Nanodrop, Agilent Technologies), and oligo(dT)_12–18 primers and Maloney's murine leukemia virus were used for RT. Negative controls (RT⁻) were carried out in the absence of RT and used to check for genomic contamination. Primers for all genes were taken from Fort et al. (12) and synthesized by Invitrogen.

Semiquantitative PCR was performed using Platinum Taq DNA polymerase (Invitrogen) with an initial denaturation step at 94°C followed by 35 cycles of 94°C for 30 s, 59°C for 30 s, and 72°C for 1 min; the reaction was completed with a 10-min extension step. β-Actin served as an internal control; no-template controls were performed with every PCR. PCR product amplification was confirmed with 1% agarose gel electrophoresis and subsequent band quantification with National Institutes of Health ImageJ (http://rsb.info.nih.gov/ij/, 1997–2005).

Isolation of mesenteric myocytes.

Mesenteric artery smooth muscle cells were isolated as previously described (7). Briefly, mesenteric arteries were transferred to Ca²⁺-free physiological saline solution (mM: 6.0 KCl, 120 NaCl, 1.2 MgCl₂, 10.0 d-glucose, and 10.0 HEPES, pH 7.3) supplemented with protease type X (0.5 mg/ml) and collagenase type IA (1.5 mg/ml) and incubated at 37°C for 15 min. Dispersed cells were maintained in 0.75 mM Ca²⁺ physiological saline solution and adhered to polylysine slides for 30 min at room temperature.

Immunocytochemistry.

Isolated myocytes were fixed and stained for confocal microscopy as previously described (7). Protein expression and localization were determined by immunofluorescence using antibodies against Na_v (1:200 dilution; catalog no. AB5210, Millipore), NCX-1 (1:500 dilution; catalog no. R3F1, Swant), Na⁺-K⁺-ATPase (1:100 dilution; Ab 7671, Abcam), and smooth muscle α-actin (1:500 dilution; catalog no. A2547, Sigma). For control experiments, primary antibodies were omitted. The labeling was visualized with tetramethylrhodamine isothiocyanate donkey anti-rabbit IgG (1:100 dilution; catalog no. 711-025-152, Jackson ImmunoResearch) and FITC sheep anti-mouse IgG (1:100 dilution; catalog no. 515-095-062, Jackson ImmunoResearch). Images were gain-matched to ensure accuracy between samples.

Data and statistical analysis.

Vascular responses are given as absolute changes in tension (mN) or as percent contraction of the tone induced by methoxamine or KCl. Values are means ± SE; n represents the number of animals. Where concentration-response curves were obtained, the data were fitted using a sigmoidal logistic equation (Prism 4, GraphPad Software, San Diego, CA): y = bottom + (top − bottom)/{1 + 10^{[(logEC₅₀−x)∗Hill slope]}}, where x is a logarithm of drug concentration and y is the response that starts from the bottom and goes to the top in a sigmoid shape. The sigmoidal curves were also used to calculate pEC₅₀ (the negative logarithm of the concentration of agent giving 50% of maximum response) and E_max (the maximal response). Statistical comparisons of vascular responses were made by Student's t-tests or one-way analysis of variance followed by Bonferroni's post hoc tests, where appropriate (Prism 4, GraphPad Software). P < 0.05 was taken as statistically significant.

Drugs.

Methoxamine, carbachol, TTX, suramin, guanethidine, 6-hydroxydopamine, and ascorbic acid (Sigma-Aldrich, Dorset, UK) and 5-HT and ranolazine (Tocris Biosciences, Bristol, UK) were dissolved in deionized water. Veratridine and capsaicin (Sigma) and A-803467 (Ascent Scientific, Bristol, UK) were dissolved in 100% ethanol. ProTx-II (kindly provided by Prof. John Wood, University College London), 9,11-dideoxy-9a,11a-methanoepoxyprostaglandin F_2α (U46619) and KB-R7943 mesylate (Tocris), T16A_inh-A01 (Millipore), and prazosin (Sigma) were dissolved in dimethyl sulfoxide.

RESULTS

Complex effects of veratridine in the rat small mesenteric artery.

Veratridine induced concentration-dependent contractions (pEC₅₀ = 5.19 ± 0.20, E_max = 12.0 ± 2.7 mN, n = 4; Fig. 1) in small mesenteric arteries; these contractions were rapid in onset and rarely sustained (Fig. 1A). Endothelium removal had no effect on contractions to a submaximal concentration (10 μM) of veratridine [9.2 ± 1.0 and 8.7 ± 2.4 mN with (n = 6) and without (n = 4) endothelium, respectively]. These responses were comparable to contractions induced by 10 μM methoxamine in the same vessels [8.1 ± 1.6 and 9.4 ± 3.0 mN with (n = 6) and without (n = 4) endothelium, respectively]. In contrast, ≤50 μM veratridine was a relatively ineffective spasmogen of rat aortic segments (Fig. 1B).

Fig. 1.

A: original trace showing veratridine responses in small mesenteric artery. B: concentration-dependent responses to veratridine in rat small mesenteric artery (n = 4) and aorta (n = 4).

Veratridine also induced relaxations of rat mesenteric artery that were evident at concentrations as low as 1 μM (Fig. 1A) and often preceded the contractile response to 10 μM veratridine (Fig. 2A). However, these relaxations were very variable, particularly at higher concentrations of veratridine. In vessels pretreated with 10 μM capsaicin followed by washout, which has previously been shown to desensitize perivascular sensory nerves (18), relaxation to 1 μM veratridine was abolished [−1.8 ± 0.6 mN (control) and 0.9 ± 0.6 mN (with capsaicin), P < 0.05, n = 4–7], while no significant effects on contractions to veratridine were observed [at 10 μM: 9.8 ± 2.2 mN (control) and 9.4 ± 2.4 mN (with capsaicin), n = 4–7]. Relaxant responses to veratridine were also seen in TTX-treated or endothelium-denuded, precontracted mesenteric arteries [%relaxation of methoxamine tone at 10 μM: 44.9 ± 13.3% (control), 34.2 ± 7.6% (without endothelium), and 54.6 ± 27.5% (with TTX), n = 5–7].

Fig. 2.

A: original trace showing veratridine contraction and its rapid reversal by application of TTX in small mesenteric artery. B: mesenteric contraction to veratridine in the absence (n = 12) and presence of TTX (1 μM, n = 4), ProTx-II (10 nM, n = 4), and A803467 (1 μM, n = 4). C: mesenteric contraction to veratridine in the absence (n = 6) and presence of ranolazine [rano; 1 μM (n = 4) and 10 μM (n = 7)] and ranolazine (10 μM) + TTX (1 μM, n = 4). **P < 0.01 vs. control; *P < 0.05 between treated groups (by 1-way ANOVA, followed by Bonferroni's post hoc tests).

Effects of TTX and other Na_V channel blockers on veratridine-induced mesenteric contractions.

Removal of Na⁺ (replaced by choline) from the Krebs-Henseleit solution abolished contractions to veratridine (n = 3), suggesting that veratridine effects were mediated by Na⁺ influx. We then used an array of pharmacological agents to ascertain the nature of the ion channel proteins involved. TTX at 1 μM is a pore blocker of all Na_V isoforms, except Na_V1.5, Na_V1.8, and Na_V1.9 (25). Pretreatment with 1 μM TTX greatly inhibited (P < 0.01) contractions induced by veratridine (Fig. 2B). Application of TTX after veratridine also caused a rapid reversal of veratridine-induced contraction and sometimes reduced the tone to below the pre-veratridine level (Fig. 2A).

On the other hand, the selective Na_V1.7 channel blocker ProTx-II (10 nM) (35) and the selective Na_V1.8 channel blocker A-803467 (1 μM) (21) had no significant effect on veratridine (10 μM) responses (Fig. 2B). The Na_V channel blockers TTX, ProTx-II, and A-803467 had no effect on the small methoxamine tone prior to veratridine additions [2.5 ± 0.8 mN (control) and 2.6 ± 0.5 mN (with toxin), n = 12].

Effects of ranolazine on veratridine-induced mesenteric contractions.

We found that ranolazine, a Na_V1.4, Na_V1.5, Na_V1.7, and Na_V1.8 channel blocker (13, 32, 37), alone relaxed precontracted tone induced by methoxamine (pEC₅₀ = 5.93 ± 0.10, E_max = 99.4 ± 8.4%, n = 3), but not 60 mM KCl (relaxation at 10 μM = 3.8 ± 8.0%, n = 3). Because of the relaxant effects of ranolazine, a slightly modified protocol was used: vessels were contracted with a higher concentration of methoxamine (3–10 μM), followed by addition of ranolazine (1 or 10 μM) and then veratridine. Using this protocol, ranolazine, at 10 μM (P < 0.01) but not 1 μM, significantly reduced contractions to 10 μM veratridine (Fig. 2C). Treatment with 10 μM ranolazine + 1 μM TTX abolished (P < 0.01) the veratridine-induced contractions (P < 0.01; Fig. 2C).

Contractions to veratridine in mesenteric arteries precontracted with 60 mM KCl.

In vessels precontracted with 60 mM KCl, veratridine (10 μM) induced consistent, monophasic contractions (Fig. 3A). Pretreatment with 1 μM TTX (P < 0.05) or 10 μM ranolazine (P < 0.05) reverted the contractile responses to relaxation (Fig. 3B). Application of TTX after veratridine also reversed the veratridine-induced contraction (Fig. 3A). Treatment with TTX + ranolazine tended to produce an additive effect, but it did not reach statistical significance (Fig. 3B). TTX or ranolazine alone had no effect on KCl-induced tone [3.8 ± 0.8 mN (control) and 3.5 ± 0.6 mN (with toxin), n = 10].

Fig. 3.

A: original trace showing veratridine contraction (in the presence of 60 mM KCl) and its reversal by application of TTX in small mesenteric artery. B: mesenteric contraction to veratridine in the presence of 60 mM KCl. Responses were obtained in the absence (n = 8) and presence of TTX (1 μM, n = 4) or ranolazine (10 μM, n = 5) and TTX + ranolazine (n = 4). **P < 0.01, ***P < 0.001 vs. control (by 1-way ANOVA, followed by Bonferroni's post hoc tests). NS, not significant.

Effects of sympathetic inhibition on veratridine-induced mesenteric contractions.

Activation of Na_V channels on perivascular sympathetic nerves could lead to release of norepinephrine and ATP, which cause mesenteric contractions via activation of α₁-adrenoceptors and P₂ purinergic receptors, respectively (17, 23). The presence of suramin (100 μM), a nonselective P2Y receptor, had no significant effect on contraction to 10 μM veratridine [9.5 ± 2.2 mN (control) and 9.6 ± 0.3 mN (with suramin), n = 4 for both]. In contrast, prazosin (1 μM), an α₁-adrenoceptor antagonist, which completely blocks methoxamine contractions (see methods), partially inhibited (by 61%) veratridine responses (Fig. 4A). Interestingly, the prazosin-resistant contraction to veratridine was abolished by KB-R7943 (Fig. 4A).

Fig. 4.

A: mesenteric contraction to veratridine in the absence (n = 5) and presence of prazosin (1 μM, n = 5) and prazosin (1 μM) + KB-R7943 (3 μM, n = 5). **P < 0.01, ***P < 0.001 vs. control; *P < 0.05 between treated groups (by 1-way ANOVA, followed by Bonferroni's post hoc tests). B: mesenteric contraction to veratridine in the absence (n = 8) and presence of KB-R7943 (3 μM, n = 8). **P < 0.01 vs. control (by Student's t-test). C: mesenteric contraction to veratridine in the presence of prazosin (1 μM, n = 4) and prazosin + T16A_inh-A01 (10 μM, n = 4). *P < 0.05 vs. control (by Student's t-test).

Additional experiments found that destruction of sympathetic nerves with 6-hydroxydopamine (2 mM) caused a 65% reduction in veratridine contractions [8.0 ± 1.3 mN (control) and 2.8 ± 1.3 mN (after 6-hydroxydopamine), n = 6 for both, P < 0.05]. However, it was somewhat surprising that inhibition of vesicular release of norepinephrine/ATP by guanethidine (5 μM) had no effect [10.7 ± 2.3 mN (control) and 10.1 ± 1.0 mN (with guanethidine), n = 4 for both]. We verified that the same guanethidine treatment abolished contractions elicited by electrical stimulation of sympathetic nerves in small mesenteric artery [at 12 Hz, 10 V for 0.5 ms: 5.4 ± 1.4 mN (control) and 0.3 ± 0.1 mN (with guanethidine), n = 6] and vas deferens (data not shown) of the rat. The sympathetic-mediated contractions in mesenteric artery were also blocked by prazosin (by 76 ± 2%, n = 4).

Effects of inhibitors of NCX and Ca²⁺-activated Cl⁻ channels on veratridine-induced mesenteric contractions.

It has been postulated (3, 12, 34) that Na⁺ accumulation following Na_V channel activity results in a rise in Ca²⁺ concentration because the NCX is driven into reverse mode (20, 30). Consequently, we studied whether KB-R7943, a selective inhibitor of the reverse mode of NCX (20), modified veratridine-induced contractions. It is clear from Fig. 4B that, at concentrations selective for NCX, KB-R7943 significantly reduced the veratridine-induced contraction by ∼60% (P < 0.01). We postulated that the rise in intracellular Ca²⁺ concentration ([Ca²⁺]_i) due to reverse-mode NCX would be sufficient to activate Ca²⁺-activated Cl⁻ channels, leading to membrane depolarization and enhanced opening of voltage-dependent Ca²⁺ channels. In rat mesenteric arteries incubated in 1 μM prazosin, preapplication of the TMEM16A-encoded Ca²⁺-activated Cl⁻ channel blocker T16A_inh-A01 (9) resulted in a markedly impaired contractile response to 10 μM veratridine compared with arteries bathed in the vehicle control (Fig. 4C).

Effects of Na_V channel blockers on U46619-induced mesenteric contractions.

Pretreatment with TTX (1 μM) or ranolazine (10 μM) had no significant effect on mesenteric contractions to the thromboxane analog U46619 [at 3 μM: 7.4 ± 1.3 mN (control), 5.9 ± 1.4 mN (with TTX), and 8.9 ± 1.7 mN (with ranolazine), n = 4–5].

Contractions to veratridine in rat femoral artery.

Recently, it has been shown that hypoxia induced contractions of rat femoral artery that were prevented by a Na_V channel blocker (4). We found that, under normoxic conditions, veratridine was an effective constrictor of rat femoral arteries and its response was ablated by application of TTX before or during veratridine challenge (Fig. 5, A and B). In addition, prazosin (1 μM) attenuated veratridine responses by ∼51% (P < 0.05; Fig. 5C). The results with 6-hydroxydopamine (2 mM) treatment were more variable; it reduced contraction to veratridine only in three of six vessels [5.0 ± 2.3 mN (control) and 2.3 ± 0.7 mN (after 6-hydroxydopamine), not significantly different, n = 6 for both]. Veratridine contractions were insensitive to 5 μM guanethidine [3.2 ± 1.0 mN (control) and 4.2 ± 0.9 mN (with guanethidine), n = 3 for both].

Fig. 5.

A: original trace showing veratridine contraction and rapid reversal by application of TTX in femoral artery. B: femoral contraction to veratridine in the absence (n = 5) and presence of TTX (1 μM, n = 4). *P < 0.05 vs. control (by Student's t-test). C: femoral contraction to veratridine in the absence (n = 7) and presence of prazosin (1 μM, n = 7). *P < 0.05 vs. control (Student's t-test).

Expression of Na_V isoforms in rat small mesenteric and femoral arteries.

Among the Na_V isoforms, mRNA transcripts for the α-subunits of Na_V1.2 and Na_V1.3 channels (both are TTX-sensitive) were consistently detected in all samples of mesenteric and femoral arteries (Fig. 6). On the other hand, expression of Na_V1.5 (TTX-resistant) and Na_V1.6 and Na_V1.7 (TTX-sensitive) α-subunits was variable; for example, Na_V1.5 was detected in two of three femoral arteries and one of three mesenteric arteries. The housekeeping gene β-actin was used as internal positive control, as well as no template controls, for all PCRs (see methods; data not shown). The expression pattern of all Na_V isoforms in control tissues, brain, heart (mainly Na_V1.5), skeletal muscle (mainly Na_V1.4), and dorsal root ganglia (Na_V1.8 and Na_V1.9), is also shown in Fig. 6.

Fig. 6.

Analysis of voltage-activated Na⁺ (Na_V) channel transcripts in rat small mesenteric (rMA) and femoral (rFA) arteries. Brain, heart, skeletal muscle, and dorsal root ganglia (DRG) from the rat were positive controls for various Na_V isoforms of α-subunits.

Expression of Na_V and NCX proteins in mesenteric smooth muscle cells.

The working hypothesis is that veratridine-induced contractions are due to accumulation of Na⁺ and NCX being forced into reverse mode. This would necessitate the close association of Na_V and NCX proteins to allow the creation of functional microdomains. We therefore ascertained whether Na_V and NCX colocalized in rat mesenteric artery. Single-cell immunocytochemistry on mesenteric myocytes (Fig. 7), confirmed by positive staining for α-actin, showed significant expression of Na_V protein. This represents the overall level of Na_V expression, since a pan-Na_V antibody was used to visualize all Na_V isoforms in the myocytes. Immunoreactivity for Na_V also costained with the membrane marker Na⁺-K⁺-ATPase, indicating that significant Na_V expression was localized to the plasma membrane (Fig. 7B). Importantly, Na_V displayed considerable colocalization with NCX1 protein (Fig. 7C), the predominant NCX isoform in vascular smooth muscle (40).

Fig. 7.

Identification of Na_V protein in freshly isolated mesenteric smooth muscle cells. Cell type was confirmed by positive staining of smooth muscle α-actin (A; catalog no. A2547, Sigma; 1:500 dilution). NPC, control cells with no primary antibodies. B and C: representative fluorescent images of isolated smooth muscle cells coincubated with anti-Na_V (pan-Na_V; Ab 5210, Millipore; 1:200 dilution) and the membrane marker anti-Na⁺-K⁺-ATPase (Ab 5210, Abcam; 1:200 dilution) or anti-Na_V and anti-Na⁺/Ca²⁺ exchanger type 1 (NCX1; R3F1, Swant; 1:500 dilution). Overlay images indicate coexpression of Na_V channel with Na⁺-K⁺-ATPase and with NCX1. Control cells with no primary antibodies (NPC) showed minimal fluorescence (A), confirming specificity of signals in B and C.

DISCUSSION AND CONCLUSIONS

In rat isolated small mesenteric artery, we demonstrated that veratridine induced endothelium-independent contractions at concentrations (pEC₅₀ ∼5.2) considerably less than those required to contract rat aorta (present study; 12). We did not observe contractile responses to ≤50 μM veratridine in rat aorta. The mesenteric contractions were predominantly inhibited by TTX, suggesting the involvement of TTX-sensitive Na_V channels (Na_V1.1, Na_V1.2, Na_V1.3, Na_V1.4, Na_V1.6, and Na_V1.7). In an attempt to further explore the functional Na_V isoforms in mesenteric arteries, the effects of a range of Na_V channel inhibitors against veratridine contractions were also tested. The selective blockers of Na_V1.7 (ProTx-II) (35) and Na_V1.8 (A-803467) channels (21) had no significant effect. In contrast, the novel antianginal agent ranolazine, which mainly blocks Na_V1.5, but also Na_V1.4, Na_V1.7, and Na_V1.8 (13, 32, 37), channels, partly inhibited veratridine-induced contraction. Ranolazine also abolished the TTX-resistant contraction to veratridine, suggesting the presence of TTX-sensitive Na_V channels (Na_V1.1, Na_V1.2, Na_V1.3, 1.4, and Na_V1.6) and the TTX-resistant Na_V1.5 channel. However, ranolazine, at concentrations commonly used to block Na_V, has been found to antagonize vasocontractions to methoxamine, but not KCl. This could be due to an antagonist action of ranolazine at the α₁-adrenoceptor (1). Thus the ranolazine data must be interpreted cautiously. Additional experiments revealed that, in arteries depolarized with 60 mM KCl [depolarization of ∼20–30 mV (10)], veratridine contractions were also sensitive to TTX and ranolazine, but the two blockers did not exert a significant additive effect. Thus it remains unclear if the Na_V1.5 channel plays a role in the contractile effects of veratridine. Also, although mRNA transcripts for Na_V1.2 and Na_V1.3 α-subunits were consistently found in isolated mesenteric arteries, the Na_V1.5 α-subunit was detected in only some of the arteries. Na_V1.2 and Na_V1.3 channels are sensitive to TTX, and their expression would be consistent with the above-mentioned pharmacological data. We therefore conclude that mesenteric contraction to veratridine is predominantly mediated by Na_V1.2 and Na_V1.3 channels. This represents the first study to characterize the effects of Na_V channel modulators on contractility of the rat small mesenteric artery, which is frequently used to model systemic resistance arteries crucial for blood pressure regulation. Our pharmacological and molecular data also corroborate the detection of Na_V currents, with high sensitivity to TTX, in freshly isolated myocytes from the same vessel (3). Selective blockers of the TTX-sensitive, Na_V1.2 and Na_V1.3 channels are unavailable, but future immunohistochemical studies using isoform-selective antibodies would help pinpoint the isoforms(s) involved.

NCX plays a critical role in regulating [Ca²⁺]_i in vascular myocytes. It operates in a forward mode (for Ca²⁺ extrusion) or a reverse mode (for Ca²⁺ entry), depending on electrochemical gradients of Na⁺ and Ca²⁺ across the plasma membrane. Increasing evidence suggests that elevation of intracellular Na⁺ concentration, which drives NCX into the reverse mode, contributes to agonist-mediated Ca²⁺ entry and contraction (19, 30). Genetic deletion of NCX1 in vascular smooth muscle and pharmacological inhibitors of the reverse mode of NCX lower blood pressure (19, 40). NCX1 also underlies salt-sensitive hypertension (19). Thus there has been much interest in the therapeutic potential of NCX inhibitors. Here, we found that inhibition by KB-R7943 (3 μM) of the reverse mode of NCX (20) compromised mesenteric contraction to veratridine, suggesting a functional coupling between Na_V and NCX. Using single-cell immunocytochemistry in mesenteric myocytes, we have now demonstrated the close proximity of Na_V and NCX1 proteins on the plasma membrane. Although NCX reversal has increasingly been implicated in Ca²⁺ influx and vascular contractility, the ion channel or transporter responsible for the localized Na⁺ accumulation remains unclear. Our data suggest that this could be mediated by activation of Na_V channels, which represents a novel route of Na⁺ entry in myocytes. Our data also suggest that recruitment of Ca²⁺-activated Cl⁻ channels is involved in this process. Reverse-mode NCX driven by raised intracellular Na⁺ concentration activates Ca²⁺-activated Cl⁻ currents in rabbit portal vein myocytes (24), and the electrophysiological interplay of Na_v, NCX, and Ca²⁺-activated Cl⁻ channels in rat mesenteric artery will be the focus of future experiments.

In this study, we also found that veratridine induced TTX-sensitive contraction in the rat femoral artery. This is consistent with a role for the Na_V channel in resistance (e.g., small mesenteric) and conduit (e.g., femoral and aorta) arteries, but our data also suggest a larger impact of the Na_V channel on the diameter of smaller arteries; veratridine efficacy ranked as follows: small mesenteric > femoral >>> aorta. As in the mesenteric artery, messages for the α-subunits of Na_V1.2 and Na_V1.3 are evident in the femoral artery. Na_V1.5, Na_V1.6, and Na_V1.7 channels can also be detected in some, but not all, femoral arteries. This corroborates a recent report of the Na_V1.5 channel in rat femoral artery (4). Interestingly, only the Na_v1.2 channel was detected in the smooth muscle layer of rat aorta (12). Together, these data provide strong evidence for a role of Na_V channels in vascular myocytes throughout the vasculature. It is worth stressing that veratridine contractions in mesenteric and femoral arteries were significantly reduced by sympathetic inhibition, implicating the involvement of Na_V channels in sympathetic nerves surrounding the vessels and the subsequent release of norepinephrine. However, a considerable contraction to veratridine remained in mesenteric and femoral arteries, despite complete blockade of α₁-adrenoceptor-mediated contraction by prazosin or chemical destruction of sympathetic nerves by 6-hydroxydopamine. In contrast, Fort et al. (12) reported that, in rat aorta, TTX-sensitive contractions to veratridine are completely inhibited by prazosin, although they used much higher concentrations of veratridine (100 μM) than those used in the current study. These data point to a differential role of Na_V in vascular myocytes vs. sympathetic nerves depending on the vascular regions, which could have important pathophysiological implications. It is also worth noting that veratridine also evoked relaxation in mesenteric arteries under certain conditions. This could be due to Na_V channels expressed in the endothelium, resulting in relaxation, counteracting the contraction mediated by Na_V channels in vascular smooth muscle (11). This is unlikely, since the relaxation was also observed in mesenteric arteries without endothelium or in the presence of TTX. It was, however, inhibited by capsaicin treatment, indicating the involvement of TRPV1-expressing sensory nerves and the release of vasodilatory neuropeptides (18, 39). Indeed, relaxation to veratridine was absent in femoral arteries, which display little capsaicin-sensitive relaxation (15). Further investigation is warranted to clarify the role of Na_V channels, especially the TTX-resistant isoforms, in vasorelaxation to veratridine.

Our working model is that activation of Na_V channels in vascular myocytes increases intracellular Na⁺, which then induces Ca²⁺ influx through the reverse mode of NCX. The subsequent rise in [Ca²⁺]_i is sufficient to activate Ca²⁺-activated Cl⁻ channels, leading to membrane depolarization, and the subsequent influx of Ca²⁺ through voltage-dependent Ca²⁺ channels results in arterial contractions. This is supported by our pharmacological, molecular, and immunocytochemical evidence. However, contractions induced by methoxamine or KCl were not significantly attenuated by TTX alone. Similarly, TTX and ranolazine do not significantly modify contractions to a thromboxane analog. This suggests that Na_V channels are not recruited by these vasoconstrictor mechanisms and would appear to have little physiological impact. However, veratridine acts by binding to open Na_V channels to increase activation at more negative membrane potential and greatly reduces inactivation of the channels, resulting in sustained Na⁺ influx (2, 38). The generation of large contractions by this agent implies that inactivated Na_V channels are present in these arteries and represent a considerable contractile “reserve.” Consequently, any pathophysiological mechanism that enhances Na_V channel activity will have a large impact on arterial contraction and blood pressure. Our data would also indicate that the Na_V channel contractile reserve in vascular smooth muscle is much larger in resistance arteries, which play a major role in the regulation of regional blood flow and blood pressure, than in conduit arteries, especially the aorta. It has been shown previously that hypoxia can produce robust contractions, which are sensitive to Na_V channel blockers (4), and the focus of future studies will be to ascertain the mechanisms that allow Na_V channel activity to be manifest in blood vessels. Interestingly, application of TTX after veratridine often reduced the tone to below the pre-veratridine level (cf. Fig. 2B), a phenomenon observed on spontaneous activity in mouse portal vein (34). These consistent findings suggest that the role of Na_V channels and NCX in vascular smooth muscle is considerably more complex than envisaged and that once Na_V channels are activated, inhibition of the Na⁺ influx leads to an overdrive of Ca²⁺ extrusion via NCX.

In conclusion, the present study confirms that Na_V channels in vascular myocytes are functionally coupled to contraction. Arterial contractions to agonists of α₁-adrenoceptors and thromboxane receptors or a depolarizing concentration of KCl are not dependent on Na_V channel activity. However, these channels can be “activated” by veratridine, leading to contractions via the reverse mode of NCX. Our results reveal that Na_V channels represent a means to increase arterial tone considerably and highlight the need to elucidate the function of vascular Na_V channels in physiological and pathological conditions.

GRANTS

P. S. Chadha and A. J. Davis were funded by British Heart Foundation Grants PG/09/104 and PG/07/127/24235 to I. A. Greenwood.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

W.-S.V.H. and I.A.G. are responsible for conception and design of the research; W.-S.V.H., A.J.D., and P.S.C. performed the experiments; W.-S.V.H., A.J.D., and P.S.C. analyzed the data; W.-S.V.H., A.J.D., P.S.C., and I.A.G. interpreted the results of the experiments; W.-S.V.H., A.J.D., and P.S.C. prepared the figures; W.-S.V.H. and I.A.G. drafted the manuscript; W.-S.V.H., A.J.D., P.S.C., and I.A.G. edited and revised the manuscript; W.-S.V.H., A.J.D., P.S.C., and I.A.G. approved the final version of the manuscript.