Na⁺/Ca²⁺ exchanger overexpression in smooth muscle augments cytosolic Ca²⁺ in femoral arteries of living mice

Abstract

Plasma membrane Na⁺/Ca²⁺ exchanger-1 (NCX1) helps regulate the cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT) in arterial myocytes. NCX1 mediates both Ca²⁺ entry and exit and tends to promote net Ca²⁺ entry in partially constricted arteries. Mean blood pressure (telemetry) is elevated by ≈10 mmHg in transgenic (TG) mice that overexpress NCX1 specifically in smooth muscle. We tested the hypothesis that NCX1 overexpression mediates Ca²⁺ gain and elevated [Ca²⁺]_CYT in exposed femoral arteries that also express the Ca²⁺ biosensor exogenous myosin light chain kinase. [Ca²⁺]_CYT and the NCX1-dependent (SEA0400-sensitive) component, ≈15% of total basal constriction in controls, were increased in TG arteries, but constrictions to phenylephrine and ANG II were comparable in TG and control arteries. Normalized phenylephrine dose-response curves and constriction to 30 and 300 ng/kg iv ANG II were virtually identical in control and TG arteries. ANG II-evoked constrictions, superimposed on elevated basal tone, accounted for the larger blood pressure responses to ANG II in TG arteries. TG and control mouse arteries fit the same pCa-constriction relationship over a wide range of pCa (≈125–500 nM). Vasodilation to acetylcholine, normalized to passive diameter, was also comparable in TG and control arteries, implying normal endothelial function. TG artery Na⁺ nitroprusside (nitric oxide donor)-induced dilations were, however, shifted to lower Na⁺ nitroprusside concentrations, indicating that TG myocyte vasodilator mechanisms were augmented. Maximum arterial dilation was comparable in TG and control mice, although passive diameter was ≈6–7% smaller in TG mice. The changes in TG arteries were apparently largely functional rather than structural, despite the congenital hypertension.

NEW & NOTEWORTHY Smooth muscle Na⁺/Ca²⁺ exchanger-1 transgene overexpression (TG mice) increases femoral artery basal cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT) and tone in vivo and raises blood pressure. Arterial constriction to phenylephrine and angiotensin II are normal but superimposed on the augmented basal [Ca²⁺]_CYT and tone (constriction) in TG mouse arteries. Similar effects in resistance arteries would explain the elevated blood pressure. Acetylcholine-induced vasodilation is unimpaired, implying a normal endothelium, but TG arteries are hypersensitive to sodium nitroprusside.

INTRODUCTION

Ca²⁺ homeostasis plays a fundamental role in vasoconstriction and hemodynamics because Ca²⁺ triggers contraction in vascular smooth muscle (47). The plasma membrane (PM) Na⁺/Ca²⁺ exchanger (NCX) is unique among the many ion channels and transporters that normally regulate the cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT) and signaling in arterial smooth muscle cells (ASMCs). NCX can transport Ca²⁺ either into or out of ASMCs, depending on the Na⁺ and Ca²⁺ concentration gradients and membrane potential (V_m). The net NCX-mediated Ca²⁺ flux direction, inward or outward, is determined by the net driving force on the exchanger, V_m – E_Na/Ca, where E_Na/Ca = 3E_Na – 2E_Ca and E_Na and E_Ca are, respectively, the Na⁺ and Ca²⁺ equilibrium potentials (9). In cardiomyocytes during diastole, and in quiescent neurons, V_m is on the order of −70 mV and is more negative than E_Na/Ca; thus, NCX mediates Ca²⁺ efflux, primarily, in these cells. Most ASMCs are tonically contracted in vivo to maintain arterial tone, and V_m is generally about −30 to −45 mV (29, 32) and is more positive than calculated E_Na/Ca.¹ NCX should, therefore, mediate Ca²⁺ influx, primarily in ASMCs (23, 44, 61, 62). In this case, overexpression of NCX would be expected to enhance Ca²⁺ influx, elevate [Ca²⁺]_CYT, augment Ca²⁺ signaling and arterial tone, and increase blood pressure (BP). In fact, transgenic (TG) mice that overexpress NCX 1 specifically in smooth muscle (SM-NCX1-TG mice) have elevated BP and enhanced salt sensitivity (23).

The discovery of NCX in arterial smooth muscle fostered the view that the cooperative action of PM Na⁺ pumps and NCX contributes to the regulation of ASMC [Ca²⁺]_CYT and BP (6, 46). Indeed, NCX1 and ouabain-sensitive α₂-Na⁺ pumps as well as certain receptor- and store-operated (transient receptor potential) channels that are permeable to Na⁺ and Ca²⁺ colocalize in PM microdomains at ASMC PM-sarcoplasmic reticulum (SR) junctions [“PLasmERosomes” (3, 8, 27, 65)]. Together, these transporters all help regulate local cytosolic Na⁺ concentration ([Na⁺]_CYT) and [Ca²⁺]_CYT and SR Ca²⁺ stores in human and rodent ASMCs (34, 39, 54).

Expression of NCX1 and of the SR Ca²⁺ pump (SERCA) as well as several Ca²⁺-selective channels is increased in ASMCs in many animal models of hypertension (10, 42, 65). This Ca²⁺ transporter “protein remodeling” appears to be due to a direct action of ouabain on α₂-Na⁺ pumps, because incubation of primary cultured rodent or human ASMCs with nanomolar ouabain also upregulates these proteins (34, 43). Moreover, mice with mutant ouabain-resistant α₂-Na⁺ pumps do not develop ouabain-induced or adrenocorticotropic hormone-induced hypertension, and the NCX blocker KBR7943 prevents the development of adrenocorticotropic hormone-induced hypertension (14, 15, 36). These studies have indicated that the α₂-Na⁺ pump ouabain-binding site and NCX1 are intimately involved in the control of Ca²⁺ homeostasis and BP and contribute to the generation of many forms of hypertension.

Given this key role of NCX1 in arteries, the present report describes experiments designed to test the hypothesis that arterial NCX1-mediated Ca²⁺ entry is enhanced in SM-NCX1-TG mice, thereby increasing basal [Ca²⁺]_CYT and augmenting Ca²⁺ signaling and constriction in vivo. The experiments were performed on exposed femoral arteries of live, anesthetized mice because of the following: 1) simultaneous [Ca²⁺]_CYT and diameter measurements have been validated in this preparation (60); 2) the effects of local and systemic application of various pharmacological agents can be assessed in vivo, e.g., femoral artery [Ca²⁺]_CYT is increased and diameter is decreased in mice with angiotensin (ANG) II-induced hypertension (58); and 3) the effects of genetic engineering (e.g., arterial NCX1 overexpression) on arterial [Ca²⁺]_CYT and diameter also can be measured directly. Since NCX1-TG mice are modestly hypertensive (23) and endothelium-dependent vasodilation is often attenuated in hypertension (55), acetylcholine (ACh)-induced vasodilation was also examined.

BP is determined largely by the “tone” [constriction as a percentage of passive diameter (PD)] in small “resistance arteries,” with only a small contribution from constriction of large muscular arteries such as the femoral artery (4, 17). Nevertheless, most of the same vasoconstrictor and vasodilator receptors as well as Ca²⁺ transporter proteins, including NCX1, are expressed in resistance and muscular arteries (39, 41, 57, 61, 62), even though quantitative differences in Ca²⁺ handling surely contribute to functional dissimilarities. We reasoned that the effects of NCX1 upregulation and increased Ca²⁺ delivery on femoral artery Ca²⁺ signaling and diameter in vivo might also provide clues to the effects of NCX1 upregulation in resistance arteries. Indeed, the enhanced myogenic tone in isolated small resistance arteries of TG mice (unpublished observations) seems consistent with the enhanced sympathetic drive-dependent tone in the femoral arteries of TG mice (present study). Thus, despite the likely compensatory action of other transporters, which may be differently regulated in different vascular beds, enhanced Ca²⁺ delivery mediated by upregulated NCX1 and its effects on arterial constriction appeared to dominate.

Notwithstanding the elevated basal arterial [Ca²⁺]_CYT and BP in TG mice, vasoconstrictor-evoked Ca²⁺ signals and ACh-induced vasodilation were comparable in control and TG mouse femoral arteries. These results, which are contrary to expectation, may provide novel insights into some of the peripheral mechanisms that contribute to the pathogenesis of hypertension.

MATERIALS AND METHODS

Experimental Mice and Ethical Approval

All husbandry and experimental procedures involving mice complied with the standards stated in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of the University of Maryland-Baltimore. Mice were housed in the University of Maryland-Baltimore Veterinary Resources facility and were fed standard laboratory chow.

SM-NCX1-TG mice (designated N1.3^Tg/Tg-10) were originally generated by Prof. Takahiro Iwamoto (Fukuoka University, Fukuoka, Japan) (23). The transgene was randomly inserted into the genome so that mice would have several copies of the gene and greatly overexpress the protein (Fig. 1 and see Supplementary Fig. 3 in Ref. 23), but the copy number in each mouse was not determined. Mice were generously provided to us by Prof. Iwamoto, and the mouse line was rederived on a C57Bl/6 background. It was then crossed with a line that expresses a modified exogenous myosin light chain kinase transgene (exMLCK) in smooth muscles (60); exMLCK is a green fluorescent protein-based cyan/yellow fluorescent protein (CFP/YFP; eGFP) Förster resonance energy transfer (FRET) Ca²⁺ biosensor (18). The offspring of exMLCK/SM-NCX1-TG mice (on a C57Bl/6 background) were then inbred to generate mice homozygous for both transgenes. For convenience, inbred mice are designated as “NCX1-TG” or “TG” mice. Mice that expressed exMLCK, but not the NCX1 transgene, were used as control (Ctrl) mice. The presence of the exMLCK gene has no detectable effect on the contractile and hemodynamic properties of the arteries (60).

Fig. 1.

Na⁺/Ca²⁺ exchanger-1 (NCX1) is overexpressed in smooth muscle-specific NCX1-transgenic (TG) mice. Immunoblots show deendothelialized aortic membrane proteins from control (Ctrl) and heterozygous and homozygous TG mice that overexpress NCX1 specifically in smooth muscle (SM-NCX1-TG mice) (TG/+ and TG/TG, respectively). TG/TG mice were used for all physiological experiments in this study. The NCX1 transgene was randomly inserted into the genome, and the protein was, therefore, greatly overexpressed, as illustrated. Protein samples had to be markedly diluted so that both Ctrl and TG NCX1 bands could be visualized on the same gel (see also Supplemental Fig. S3 in Ref. 23). β-Actin was used as the loading control.

Fig. 3.

Effects of pretreatment (15 min) with 0.3 μM SEA0400 on basal cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT; A) and basal tone (C) in control (Ctrl) and transgenic (TG) femoral arteries in vivo. The SEA0400-induced decrease in [Ca²⁺]_CYT (Δ[Ca²⁺]_CYT; B) and vasodilation [diameter increase as %passive diameter (PD; D) are also shown. Effects of SEA0400 were augmented in TG mouse arteries. Values are means ± SE. Statistics were by two-way ANOVA. *P < 0.05 vs. the respective Ctrl group; #P < 0.05 and ##P < 0.01 vs. the basal TG group; ΔP < 0.05 vs. the basal TG group. Numbers of mice are in parentheses.

In a few experiments (see results), mice in which NCX1 was conditionally knocked out in smooth muscle (SM-NCX1-KO mice, abbreviated as “KO”) (57) were tested for comparison with Ctrl and TG mice.

Primer set sequences used for genotyping the mice with the two transgenes were as follows: 1) exMLCK, forward 5′-GCACGACTTCTTCAAGTCCGCCATGCC-3′ and reverse 5′-GCGGATCTTGAAGTTCACCTTGATGCC-3′; and NCX1, forward 5′-AGGTGAAGGTATTGCGAACATCTGGA-3′ and reverse 5′-GGCTCCAATGAAAGCTGTCAGTAT-3′.

BP Measurements

Telelemtry.

Telemetric BP transmitters (DSI TA11PA-C10, Data Sciences, Minneapolis, MN) were surgically implanted in 15 to 18 wk-old mice in the University of Maryland-Baltimore Institutional Animal Care and Use Committee-approved Physiology Phenotyping Core. The transmitter catheter was inserted into the right common carotid artery, and the tip was threaded into the ascending aorta. Details have been previously published (62). After 7–10 days of recovery from surgery, 24-h BPs were recorded with Data Sciences receivers and software. After stable baseline BPs were recorded, under isoflurane anesthesia, Alzet osmotic minipumps (Durect, Cupertino, CA) were implanted in the right abdominal wall for subcutaneous infusion of ANG II or vehicle. The 24-h BP data collection was performed “double blind”: the individual measuring BP did not know the genotype. Animal code numbers were matched with genotype and BP after the data had been collected.

Acute aortic BP measurements.

Mice (~15–18 wk old) were anesthetized with 2–3% isoflurane supplemented with 100% O₂; core temperature was maintained at 37.5–38°C. An anterior midline incision was made in the neck, and the right carotid artery and jugular vein were isolated and the trunk of the carotid was ligated with a suture. A 1.1-Fr Mikro-tip pressure catheter (Millar Instruments, Houston, TX) was inserted into the artery proximal to the ligature and threaded forward to the ascending aorta. The jugular vein was then ligated, and a PE-10 catheter (Braintree Scientific, Braintree, MA) for intravenous infusions was inserted caudally and threaded down to the superior vena cava. The incision was then closed with 5-mm EZ clips (Braintree Scientific). Arterial BP was recorded under 1.5% isoflurane anesthesia (25) at a sampling rate of 1 kHz using a PowerLab data acquisition system (AD Instruments, Colorado Springs, CO). Data were later calculated offline (BioPac System, Santa Barbara, CA). After the experiment, the animal was deeply anesthetized with isoflurane and euthanized by cervical dislocation. Further details have been previously described in Ref. 60.

Femoral Artery In Vivo Imaging

The in vivo imaging methods and validation have been previously described in detail (58, 60). In brief, under microscope observation, a femoral artery of an isoflurane-anesthetized mouse was exposed through a cutaneous incision in the midthigh. The connective tissue was lightly dissected over a ~2-mm-long region of the distal part of the femoral artery, taking care to avoid severing nerves. The mouse was then moved to the stage of the fluorescence microscope, and superfusion of the exposed artery and surrounding muscle was initiated. The arrangement of the inlet and outlet for the superfusion fluid was made to achieve a laminar flow over the artery of ~9 ml/min.

The use of a volatile anesthetic (isoflurane) was desirable to maintain the animal in an optimal physiological state. Such agents have, however, been reported to influence sympathetic nerve activity (SNA) (49), alter the activity of the renin-angiotensin system (53), and activate K⁺ channels (30). Nevertheless, femoral arteries under the conditions of these experiments maintained a high degree of vasoconstriction, as evidenced by the large increase in diameter after total autonomic ganglion blockade (see results).

Fluorescence and Diameter Recordings in Femoral Arteries

A xenon arc lamp (Lambda LS, Sutter Instrument, Novato, CA) provided intensity-adjusted, gated fluorescent illumination at 436 ± 10 nm (excites CFP) that was controlled by a programmable shutter (Smart Shutter, Sutter Instrument). A long working distance, high numerical aperture Olympus MVX10 MacroView fluorescence microscope (Olympus America, Center Valley, PA) and a sensitive charge-coupled device ORCA ER camera (Hammamatsu, Bridgewater, NJ) fitted with a DualView image splitter (Photometrics, Tucson, AZ) were used. This provided “CFP” and “YFP” images (470 and 535 nm, respectively). The camera was controlled and images were acquired (1 frame/s) using HCImage (Hamamatsu). Image processing was performed with custom software written using IDL 6.4 (IIT Visual Information Solutions, Boulder, CO). Elevation of [Ca²⁺]_CYT promotes the binding of Ca²⁺ to calmodulin and exMLCK [a calmodulin-based biosensor (22)]. In the latter case, CFP and YFP moieties then move apart, thereby reducing FRET from CFP to YFP and thus increasing the CFP-to-YFP (FRET) ratio. The FRET ratio was calibrated to determine [Ca²⁺]_CYT (Ref. 60 provides details).

Arterial “tone” is the difference between the PD in Ca²⁺-free solution and the observed diameter under basal or other conditions tested (D_Obs) as a percentage of PD as follows:

$$\text{Tone}=\frac{\text{PD}–D_{\text{Obs}}}{\text{PD}}\times 100$$

Changes in D_Obs (ΔD_Obs) due to agent applications are presented as a percentage of PD [(ΔD_Obs/PD) × 100]. Vasodilation was graphed as upward bars or curves; vasoconstriction was graphed as downward bars or curves. Actual diameters were also graphed.

Immunoblot analysis

Previously published methods were used for immunoblot analysis of mouse aorta membrane proteins (34, 61, 62). NCX1 expression was identified by chemiluminescence with R3F1 anti-NCX1 monoclonal antibodies (R3F1; generously provided by Prof. Kenneth D. Philipson, University of California, Los Angeles, CA). β-Actin was used as a loading control and was identified with clone AC-74 anti-β-actin monoclonal antibodies (Sigma-Aldrich, St. Louis, MO).

Reagents and Solutions

Artery superfusion solution composition was as follows (in mM): 112 NaCl, 25.7 NaHCO₃, 4.9 KCl, 2.5 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 11.5 glucose, and 10 HEPES (pH 7.4, equilibrated with gas of 5% O₂-5% CO₂-90% N₂ at 37°C). Ca²⁺-free solution was made by omitting Ca²⁺ and adding 0.5 mM EGTA.

Reagents and sources were as follows. ACh, ANG II, hexamethonium chloride, sodium nitroprusside (SNP), ouabain, phenylephrine (PE), and prazosin HCl were from Sigma-Aldrich. Losartan was from Merck (Kenilworth, NJ). SEA0400 (2-{4-[(2,5-difluorophenyl) methoxy]phenoxy}-5-ethoxyaniline) was a gift from Prof. Iwamoto. All reagents were reagent grade or the highest grade available. SEA0400 was dissolved in DMSO; other agents were dissolved in double-distilled deionized water as stock solutions. The final bath concentration of DMSO did not exceed 0.01% (vol/vol) and did not affect vascular function.

Statistics

Data in Figs. 2–12 are shown as means ± SE. Shapiro-Wilk and Levene tests were performed for data normality and homogeneity, respectively; all data used for ANOVA passed both tests. Paired or unpaired t-tests, two-way ANOVA with post hoc correction using the Bonferroni approach, or linear regression (SigmaPlot 11.0, Systat Software, San Jose, CA) were used to test for significant differences (P < 0.05), as stated in the figures. SigmaPlot was used for post hoc calculation of statistical power.

Fig. 2.

Effects of smooth muscle-specific Na⁺/Ca²⁺ exchanger-1 (NCX1) overexpression on femoral artery cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT), diameter, and tone. Baseline diameter and passive diameter (PD; A), basal tone (PD − baseline diameter as %PD; B), and basal [Ca²⁺]_CYT (C) in the exposed femoral arteries of control (Ctrl) and transgenic (TG) mice in vivo under light anesthesia are shown. Baseline diameter was reduced, and basal tone and [Ca²⁺]_CYT were increased, in TG mouse arteries. *P < 0.05 vs. Ctrl mice (Student’s unpaired t-test); ***P < 0.001 vs. Ctrl mice; ###P < 0.001 vs. the respective basal diameter (two-way ANOVA with pairwise comparisons using the Holm-Sidak method). Numbers of Ctrl and TG mice are shown in parentheses at the top. The difference between mean Ctrl PD and TG PD was not significant (P = 0.085 by two-way ANOVA). When mean PD values from six independent experiments (with 51 Ctrl arteries and 51 TG arteries) were analyzed, however, the respective means of the means from each experiment (Ctrl PD: 397.2 ± 3.9 μm and TG PD: 379.5 ± 5.3 μm) were significantly different (P = 0.023 by a Student’s unpaired t-test). The 95% confidence interval for the difference of the means was 3.043–32.291; the power of the t-test was 0.617.

Fig. 12.

Sodium nitroprusside (SNP) cumulative dose-response curves for exposed femoral arteries in control (Ctrl) and transgenic (TG) mice in vivo; SNP was applied by superfusion. A: SNP dose versus artery diameter. Ctrl passive diameter (PD) = 407 ± 23 μm (n = 6); TG PD = 387 ± 13 μm (n = 6). B: SNP-induced decrease (Δ) in cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT). Basal [Ca²⁺]_CYT was 219 ± 31 nM in Ctrl mouse arteries and 309 ± 27 nM in TG mouse arteries. C: SNP-induced vasodilation (as %PD). TG curves were all significantly shifted to the left (indicating higher affinity and lower SNP EC₅₀ values) relative to the respective Ctrl curves (*P < 0.05). Values are means ± SE. Statistics were by two-way ANOVA. Numbers of mice are in parentheses.

RESULTS

NCX1 Expression Detected by Immunoblot Analysis and Function

Relative NCX1 expression was determined by immunoblot analysis. NCX1 expression was markedly upregulated in aortas from NCX1-TG heterozygous mice and even more so in homozygous TG mice (Fig. 1) because each mouse may have several copies of the transgene. These results resemble those obtained with the original TG line (see Supplemental Fig. 3a in Ref. 23). Homozygous overexpressors were used for all experiments described below.

Functional evidence of NCX1 overexpression was obtained in femoral arteries of anesthetized mice in vivo. First, arterial diameter and [Ca²⁺]_CYT were measured in Ctrl and TG mouse arteries. Under basal conditions, the average diameter of TG arteries was ~20% narrower than that of Ctrl arteries, whereas the PDs of TG arteries (i.e., in Ca²⁺-free solution) were only ~5% narrower than the PDs of Ctrl arteries (Fig. 2A). Basal “tone” was, therefore, significantly greater in TG femoral arteries (Fig. 2B). This increased tone was associated with, and is likely due to, the significantly elevated basal [Ca²⁺]_CYT in TG arteries (Fig. 2C). The modestly smaller PD of the TG arteries in age-matched mice raises the possibility that the congenital increase in arterial tone and BP (Ref. 23 and present study) may have induced some structural changes in TG arteries.

Second, to test for NCX1 function directly, the effects of the NCX1 inhibitor SEA0400 on [Ca²⁺]_CYT and femoral artery diameter were determined. Superfusion with 0.3 μM SEA0400 reduced [Ca²⁺]_CYT by ~50 nM in Ctrl arteries and by ~85 nM in TG arteries (Fig. 3, A and B). This was associated with a significantly greater decrease in tone in TG arteries than in Ctrl arteries (Fig. 3, C and D), as expected if the arterial constriction and tone depend, in part, on NCX1-mediated, SEA0400-sensitive Ca²⁺ entry. The SEA0400-induced 30% decline in basal tone in TG mouse arteries was statistically significant, but the 10% decline in Ctrl arteries was not, even though all six arteries dilated by a small amount.

Furthermore, when arteries were constricted by prolonged (15 min) exposure to 10 μM ouabain, which should elevate [Na⁺]_CYT and promote Ca²⁺ entry via NCX1 (Fig. 4A), TG arteries constricted more than did Ctrl arteries (Fig. 4B). SEA0400 then reduced the constriction in TG arteries to a significantly greater extent than in Ctrl arteries (Fig. 4B). The elevated basal [Ca²⁺]_CYT in arteries from TG mice and the difference in SEA0400-triggered decrease in tone between Ctrl and TG mice fit the view that ASMC NCX1 mediates Ca²⁺ entry, primarily in those arteries with tone. The data shown in Fig. 2, Fig. 3, and Fig. 4 are all consistent with the overexpression of NCX1 protein in TG mouse arteries (Fig. 1).

Fig. 4.

Effects of ouabain and SEA0400 on femoral artery cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT) and diameter. [Ca²⁺]_CYT increase (Δ[Ca²⁺]_CYT; A) and vasoconstriction [diameter decrease, as %passive diameter (PD); B] induced by a 15-min superfusion with 10 μM ouabain as well as the effects of 0.3 μM SEA0400 (15-min pretreatment before ouabain) on these changes. Ouabain-induced, SEA0400-sensitive effects were amplified in transgenic (TG) mouse arteries. Values are means ± SE. Statistics were by a paired t-test. #P < 0.05 vs. ouabain alone. Numbers of mice are in parentheses.

Basal BP and BP Responses to Chronic and Acute ANG II Infusion Are Augmented in TG Mice

The BP data shown in Fig. 5A support previous reports (23, 57, 62) showing that SM-NCX1-KO reduces basal BP and SM-NCX1 overexpression elevates BP relative to BP in Ctrl mice. Furthermore, the BP increase induced by chronic low-dose subcutaneous ANG II infusion (400 ng·kg⁻¹·min⁻¹ for 3 wk) was much greater in TG mice, and less in KO mice, than in Ctrl mice. Also, acute intravenous infusion of 30 and 300 ng/kg ANG II induced a significantly larger increment in BP in TG anesthetized mice than in Ctrl anesthetized mice (Fig. 5B). This raises the possibility that arteries from TG mice may be more sensitive to ANG II than arteries from Ctrl mice and that the ANG II dose-response curve may, therefore, be shifted to the left.

Fig. 5.

The acute and chronic angiotensin II (ANG II)-induced blood pressure elevation is augmented in transgenic (TG) mouse arteries. A: mean blood pressure (24-h MBP) in awake, mobile smooth muscle-specific NCX1 knockout mice (SM-NCX1-KO; KO), control (Ctrl) mice, and TG mice that overexpress NCX1 specifically in smooth muscle (SM-NCX1-TG; TG) mice infused with saline or ANG II (400 ng·kg⁻¹·min⁻¹ sc) for 3 wk. B: increases in aortic MBP (ΔMBP) induced by acute intravenous infusion of 30 or 300 ng/kg ANG II in lightly anesthetized KO, Ctrl, and TG mice. Values are means ± SE. Statistics were by two-way ANOVA. *P < 0.05 and ***P < 0.001 for the pairs indicated. NS, nonsignificant. Numbers of mice are in parentheses.

Femoral Artery Tone Is Largely Dependent on Sympathetic Drive

Femoral arteries in vivo constricted to superfused PE (an α₁-adrenoceptor agonist) in a dose-dependent manner (Fig. 6). Basal [Ca²⁺]_CYT was elevated in TG arteries (Figs. 2C and 3A), and they were modestly, but significantly, more constricted than Ctrl arteries at baseline (Figs. 2A and 6A; see Figs. 7–11). Furthermore, the TG artery diameter-PE dose-response curve was shifted downward relative to the curve for Ctrl arteries (Fig. 6A). When the data were normalized to PD and expressed as PE-induced vasoconstriction and Δ[Ca²⁺]_CYT versus PE concentration curves, however, the Ctrl and TG artery curves were almost superimposable (Figs. 6, B and C). Thus, the sensitivity of TG arteries to PE appeared comparable to that of Ctrl arteries.

Fig. 6.

Phenylephrine (PE) cumulative dose-response curves for exposed femoral arteries in control (Ctrl) and transgenic (TG) mice in vivo; PE was applied by superfusion. A: PE dose versus artery diameter. Ctrl passive diameter (PD) = 404 ± 21 μm (n = 9); TG PD = 374 ± 5 μm (n = 11). B: PE-induced vasoconstriction (as %PD). C: PE-induced increase (Δ) in cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT). Basal [Ca²⁺]_CYT was 221 ± 29 nM in Ctrl mouse arteries and 319 ± 36 nM in TG mouse arteries, but the normalized vasoconstriction curves (B) and Δ[Ca²⁺]_CYT curves (C) from Ctrl and TG mice were virtually superimposable (two-way ANOVA). Values are means ± SE.

Fig. 7.

Effects of α₁-adrenergic blockade [prazosin (Pra)] on femoral artery cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT), diameter, and tone in femoral arteries of control (Ctrl) and smooth muscle-specific Na⁺/Ca²⁺ exchanger-1 (NCX1) overexpression [transgenic (TG)] mice. A: diameters of exposed femoral arteries of Ctrl and TG mice in vivo as well as the effects of superfusion with 300 nM Pra. Passive diameter (PD) (arteries superfused with Ca²⁺-free medium) is also shown. Ctrl PD = 403 ± 13 μm (n = 6); TG PD = 397 ± 12 μm (n = 6). B and C: [Ca²⁺]_CYT and arterial tone (as %PD), respectively, in Ctrl and TG mouse arteries under basal conditions and during superfusion with 300 nM Pra. Pra reduced [Ca²⁺]_CYT and tone to a greater extent in TG mice than in Ctrl mice. [Ca²⁺]_i, intracellular Ca²⁺ concentration. Values are means ± SE. Statistics were by two-way ANOVA. *P < 0.05, **P < 0.01, and ***P < 0.001 vs. basal Ctrl mice; ##P < 0.01 and ###P < 0.001 vs. basal TG mice. Numbers of mice are in parentheses.

Fig. 11.

Acetylcholine (ACh) cumulative dose-response curves for exposed femoral arteries in control (Ctrl) and transgenic (TG) mice in vivo; ACh was applied by superfusion. A: ACh dose versus artery diameter. Ctrl passive diameter (PD) = 390 ± 19 μm (n = 6); TG PD = 361 ± 16 μm (n = 10). B: phenylephrine-induced vasodilation (as %PD). C: ACh-induced decrease (Δ) in cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT). Basal [Ca²⁺]_CYT was 230 ± 23 nM in Ctrl mouse arteries and 335 ± 10 nM in TG mouse arteries, but the normalized vasodilation curves (B) and [Ca²⁺]_CYT curves (C) from Ctrl and TG mice were virtually superimposable. Values are means ± SE.

Superfusion of femoral arteries with 300 nM prazosin (an α₁-adrenoceptor antagonist) reduced arterial [Ca²⁺]_CYT to ≈80–85 nM in both TG and Ctrl anesthetized mice and dilated the arteries and decreased basal tone by ~55% in Ctrl arteries and 70% in TG arteries (Fig. 7). The decrease in tone in wild-type (WT) arteries was only modestly greater than the 46% decline observed with 100 nM prazosin (59). Of note, prazosin reduced tone to about the same level (≈12–14% of PD) in TG and Ctrl arteries. This implies that the TG-induced increase in tone apparently depended on both the NCX1-mediated increase in [Ca²⁺]_CYT (Fig. 3) and SNA.

Blockade of the autonomic ganglia by intravenous bolus infusion of 3 mg/kg hexamethonium reduced [Ca²⁺]_CYT to barely measurable levels and dilated the arteries almost to PD, thereby nearly abolishing tone in both TG and Ctrl mice (Fig. 8). These effects of prazosin and hexamethonium indicate that the higher [Ca²⁺]_CYT and greater tone in TG mouse femoral arteries both depend largely on SNA. The hexamethonium data signify that essentially all of the basal tone in TG arteries, as in Ctrl arteries (see Ref. 59), was due to sympathetic drive. Also, mean PDs of Ctrl and TG arteries were nearly identical (see Fig. 2), even though TG arteries had significantly more basal tone (Fig. 2B, Fig. 3C, and Fig. 7A).

Fig. 8.

Effects of intravenous injection of 3 mg/kg hexamethonium on arterial diameter (representative original recording; A) and mean arterial tone (B) and on cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT) representative original data (C) and mean data (D) in exposed femoral arteries of control (Ctrl) and smooth muscle-specific Na⁺/Ca²⁺ exchanger-1 transgenic (TG) mice in vivo. [Ca²⁺]_i, intracellular Ca²⁺ concentration. Hexamethonium redued [Ca²⁺]_CYT to ≪50 nM and virtually abolished tone in both Ctrl and TG arteries. Ctrl passive diameter (PD) = 382 ± 14 μm (n = 8); TG PD = 386 ± 12 μm (n = 8). Values are means ± SE. Statistics were by two-way ANOVA. *P < 0.05 and ***P < 0.001 vs. basal Ctrl mice; ###P < 0.001 vs. basal TG mice. Numbers of mice are in parentheses.

In both Ctrl and TG femoral arteries, 100 nM ANG II superfusion elevated [Ca²⁺]_CYT, induced vasoconstriction, and increased tone, and the magnitudes of these changes were comparable in TG and Ctrl arteries (Fig. 9). Furthermore, 100 nM losartan [an ANG II type 1 receptor (AT₁R) antagonist] abolished all of these ANG II-induced changes (Fig. 9). Both the PE and ANG II data (Figs. 6 and 9) indicate that arterial contractility was comparable in Ctrl and TG mice. In comparison with the sympatholytic agents (Figs. 7 and 8), however, losartan had a negligible effect on [Ca²⁺]_CYT, basal diameter, and tone in either Ctrl or TG arteries (Fig. 10). Thus, ANG II activation apparently contributes very little to basal tone, in contrast to the dominant role of the sympathetic nervous system and norepinephrine (Figs. 6 and 7) (see Ref. 59).

Fig. 9.

Effects of superfusion of 100 nM angiotensin II (ANG II) with or without 100 nM losartan on the diameter (A), cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT; B), and tone [as %passive diameter (PD); C] of exposed femoral arteries in control (Ctrl) and transgenic (TG) mice in vivo. Losartan abolished nearly all of the ANG II-induced increases in [Ca²⁺]_CYT and tone in both Ctrl and TG arteries. Values are means ± SE. Statistics were by two-way ANOVA. ***P < 0.001 vs. the respective basal Ctrl or TG mice; ##P < 0.01 vs. ANG II alone; ¶P < 0.05 vs. basal Ctrl mice. Numbers of mice are in parentheses.

Fig. 10.

Effects of superfusion with 100 nM losartan on the diameter (A), cytosolic Ca²⁺ concentration ([Ca²⁺]_CYT; B), and tone [as %passive diameter (PD); C] on exposed femoral arteries in control (Ctrl) and transgenic (TG) mice in vivo. Losartan had a negligible effect on [Ca²⁺]_CYT and tone in both Ctrl and TG arteries. Values are means ± SE. Statistics were by two-way ANOVA. *P < 0.05 vs. basal Ctrl mice. Numbers of mice are in parentheses.

Vasodilation Is Not Attenuated in TG Mouse Femoral Arteries

ACh-induced vasodilation is impaired in human hypertension and in many animal hypertension models (for a review, see Ref. 55). Despite the modestly elevated basal BP in TG mice (Fig. 5A and Ref. 23), however, ACh-induced vasodilation of femoral arteries was comparable in TG and Ctrl mouse femoral arteries (Fig. 11): the ACh-induced decline in arterial [Ca²⁺]_CYT and vasodilation, normalized to PD, was virtually identical in Ctrl and TG mice (Fig. 11, B and C). This implies that endothelium-mediated vasodilator mechanisms, including the release of nitric oxide (NO), were not compromised in TG mice, despite the chronic BP elevation. Although the PDs of TG mouse arteries were only slightly smaller than those of Ctrl mouse arteries (Fig. 11), a similar difference was seen when all arteries were averaged (Ctrl PD: 397 ± 5 μM, n = 51; TG PD: 380 ± 5, n = 51; P < 0.03; data from experiments shown in Figs. 2, 6–8, 11, and 12).

Surprisingly, comparison of the vasodilation activated by the NO donor SNP in the two phenotypes revealed that TG mouse femoral arteries were significantly more sensitive to SNP than Ctrl mouse femoral arteries (Fig. 12). TG artery diameter, normalized vasodilation, and Δ[Ca²⁺]_CYT SNP dose-response curves were all shifted leftward (toward lower SNP doses) compared with those from Ctrl arteries (Fig. 11, A–C, respectively). Thus, although the endothelium-mediated vasodilator mechanisms apparently were normal (Fig. 11), either the myocyte PKG response to NO was enhanced or some downstream vasodilator mechanism in ASMCs was upregulated in TG mice.

DISCUSSION

ASMC NCX1 Preferentially Mediates Ca²⁺ Entry Under Basal Conditions In Vivo

NCX1, along with the PM Ca²⁺ pump (50) and various Ca²⁺-permeable channels (16, 24), helps to maintain Ca²⁺ homeostasis in ASMCs. NCX1 can mediate Ca²⁺ influx as well as efflux and is important not only for clearing Ca²⁺ after Ca²⁺ transients (39, 41) but also for mediating net Ca²⁺ entry and maintaining elevated [Ca²⁺]_CYT during the development and maintenance of arterial tone (23, 44, 45, 61). The latter is exemplified by the observations that local inhibition of NCX1 modestly lowers basal arterial [Ca²⁺]_CYT and dilates tonically constricted Ctrl arteries (Fig. 3, B and C). Therefore, it is not surprising that increasing NCX1 capacity by TG overexpression of NCX1 in smooth muscle (Fig. 1) increases basal [Ca²⁺]_CYT and arterial tone and elevates BP (Fig. 2) (23). This demonstrates that NCX1 mediates Ca²⁺ entry in vivo under basal conditions.

To better understand these effects, the influence of vasoconstrictors and vasodilators on BP, and on femoral arteries in vivo (under anesthesia), was studied in Ctrl and smooth muscle-specific NCX1-TG mice. Homozygous TG/TG mice markedly overexpress NCX1 in arterial myocytes (Fig. 1) because of random insertion of the transgene into the genome. The overexpression was functionally confirmed by the augmented SEA0400-induced vasodilation and decline in Δ[Ca²⁺]_CYT (Fig. 3) and the increase in ouabain-induced, SEA0400-sensitive vasoconstriction in TG mouse arteries (Fig. 4).

Role of Arterial NCX1 Activity in the Control of Basal [Ca²⁺]_CYT and Arterial Tone

If NCX1 helps regulate [Ca²⁺]_CYT in ASMCs primarily by promoting Ca²⁺ entry (23, 62), and if arterial constriction and tone depend on Ca²⁺ signals (13, 29, 60), it follows that knockout of myocyte NCX1 would be expected to lower [Ca²⁺]_CYT and reduce arterial tone and BP, as observed (44, 57, 62). Conversely, overexpression of NCX1 in ASMCs may be expected to raise [Ca²⁺]_CYT and augment arterial tone (Figs. 2 and 3). Furthermore, the elevated BP in TG mice (Fig. 5A) (23) fits with the inference that the role of NCX1 observed in muscular femoral arteries also applies to small resistance arteries with myogenic tone (57, 62).

The contribution of NCX1 to basal [Ca²⁺]_CYT and arterial tone is indicated by the effects of genetically altered NCX1 expression and NCX1 blockade by SEA0400. The amplification of the effects of blockade in TG arteries (Fig. 3) and, especially, the absence of effects on arterial tone and BP in NCX1 KO mice (57, 62), indicates the specificity of low-dose (0.3 μM) SEA0400 action. While the SEA0400-induced fall in [Ca²⁺]_CYT was significant, the 17% decline in Ctrl muscular artery basal tone did not reach statistical significance (Fig. 3). (Power analysis indicated that a far larger number of arteries would be required, given the small magnitude of the decrease and the variance of the measurements.) Nevertheless, the trend conforms with the similar, but statistically significant, ≈15% decrease in resistance artery myogenic tone previously observed in 0.3 μM SEA0400-treated isolated Ctrl arteries (23, 44). The SEA0400-induced changes indicate that NCX1 accounted for ~27% of basal [Ca²⁺]_CYT and ~19% of basal tone in TG mouse femoral arteries.

In a post hoc analysis after completion of all experiments, the relationship between femoral artery [Ca²⁺]_CYT and tone was examined for the myriad conditions tested. The data [shown as an in vivo (under light anesthesia) pCa-vs.-arterial constriction plot in Fig. 13] revealed that constriction was nearly linearly related to log[Ca²⁺]_CYT between 100 and 550 nM [Ca²⁺]_CYT. The exMLCK FRET calibration data (58) are included in Fig. 13. For comparison, a plot of pCa versus myogenic constriction in isolated, pressurized rat small cerebral arteries (29) is also shown (Fig. 13), because myogenic tone in small arteries is directly related to intracellular Ca²⁺ concentration measured with fura-2 (13, 29). The inflection point (~120 nM [Ca²⁺]_CYT, pCa = 6.92) is similar for femoral and small cerebral arteries, presumably because this is close to the contraction threshold. However, the normalized small-artery constriction curve is steeper than that for femoral arteries and reaches maximal constriction (diameter ≈40 μm) at [Ca²⁺]_CYT ~ 325 nM (pCa = 6.5). The larger arteries have a larger range of constriction (from ≈400 to ≈150 μm vs. ≈200 to ≈40 μm for small arteries) and are not yet maximally constricted at [Ca²⁺]_CYT = 550 nM (pCa = 6.26). This relationship implies that small NCX1-dependent changes in [Ca²⁺]_CYT may have a large effect on arterial constriction and BP. The evidence that Ctrl and TG artery data fit the same curve implies that the sensitivity of the contractile apparatus to Ca²⁺ did not vary under the conditions examined here, consistent with the data shown in Figs. 6 and 8.

Fig. 13.

Femoral artery constriction [%passive diameter (PD)] in vivo graphed as a function of pCa (−log[Ca²⁺]_CYT; cytosolic Ca²⁺ concentration). Data were taken from Figs. 2–12 [○, control (Ctrl); ▼, transgenic (TG)] and from Wang et al. (58) (Δ). The regression lines for the three sets of data for [Ca²⁺]_CYT = 100–500 nM (R = 0.91, 0.69 and, 0.97, respectively) were indistinguishable and correspond to a constriction of ~10% of PD per 100 nM increase in [Ca²⁺]_CYT. For comparison, the myogenic constriction versus pCa data in isolated, pressurized (60 mmHg) rat small posterior cerebral arteries taken from Fig. 10 of Knot and Nelson (29) are also shown (■); the constriction, as %PD, was calculated based on a PD of 210 μm. The graph also contains the normalized exogenous myosin light chain kinase transgene Förster resonance energy transfer (FRET) ratio curve (●; where 0 = “0” [Ca²⁺]_CYT; 100 = saturating [Ca²⁺]_CYT) from Fig. 1B of Wang et al. (58).

Dominant Role of Sympathetic Drive in Maintaining Basal Femoral Artery Tone

In contrast to the modest effects of SEA0400 in Ctrl mice, blockade of α₁-adrenoceptors with prazosin reduced femoral artery [Ca²⁺]_CYT by ≈65% and basal tone by ≈55% (Fig. 7). Blockade of sympathetic ganglia with hexamethonium reduced [Ca²⁺]_CYT and basal tone by ≥90% in Ctrl mice and by even more in TG mice (Fig. 8). Nevertheless, the data shown in Fig. 3 suggest that a modest fraction of the almost entirely SNA-driven tone (59), perhaps ≈15% in normal femoral arteries, also depends on NCX1-mediated Ca²⁺ entry, but the majority apparently depends on other Ca²⁺ transporters and channels. Basal [Ca²⁺]_CYT and tone in TG as well as Ctrl mouse arteries were, however, negligibly affected by the AT₁R inhibitor losartan (Fig. 10), even though it effectively blocked ANG II-evoked [Ca²⁺]_CYT elevation and vasoconstriction (Fig. 9). These findings corroborate earlier reports showing that in vivo basal arterial tone in mouse femoral arteries is largely determined by sympathetic drive (58, 59).

Why Is ANG II-Induced BP Elevation Greater in SM-NCX1-TG Mice Than in Ctrl Mice?

ANG II, whether infused acutely (intravenously) or chronically (subcutaneously), is widely known to increase BP (37, 48, 58). Interestingly, not only is basal BP modestly elevated in SM-NCX1-TG mice but also both acute and chronic ANG II induced a larger BP increment in TG mice than in Ctrl mice (Fig. 5). Conversely, SM-NCX1-KO mice are modestly hypotensive (57, 62), and the chronic ANG II-induced BP elevation tends to be smaller than in either Ctrl or TG mice (Fig. 5A). Thus, the ANG II-induced BP elevation appears to be arterial NCX1 “dose dependent”.

The results summarized above might suggest that TG mouse arteries are more sensitive to vasoconstrictors than are Ctrl mouse arteries. For example, if basal [Ca²⁺]_CYT is elevated, SR Ca²⁺ stores and evoked Ca²⁺ signals might be expected to increase, but this apparently is not the case. Femoral arteries in anesthetized TG and Ctrl mice exhibited virtually identical vasoconstrictor responses (measured as “Δdiameter”) to locally superfused PE (Fig. 6) and ANG II (Fig. 9A). The downward shift of the PE dose versus TG artery diameter curve (i.e., the dose-response curve), relative to the Ctrl artery curve (Fig. 6A), implies that the PE constriction of TG arteries is superimposed on the augmented basal constriction. The normalized curves (relative to PD; Fig. 6B) are nearly identical, however, indicating that there is no difference in PE sensitivity (i.e., half-maximal activation or K_0.5) between Ctrl and TG mice. It is noteworthy that a very similar result was obtained when PE-induced femoral artery vasoconstriction was compared in Ctrl mice and in mice infused chronically with ANG II to induce hypertension (58).

Complementary effects were observed in SM-NCX1-KO mice, which have low BP. Basal myogenic tone was lower in KO resistance arteries than in Ctrl resistance arteries, and the PE dose-response curve was simply displaced downward in the KOs (62). In vivo infused ANG II dose-BP response data also indicate that the curve for KO mouse arteries paralleled that of Ctrl mouse arteries but was displaced downward (63). Thus basal tone and BP are directly related to the level of NCX1 expression in Ctrl, TG, and KO arteries.

TG and Ctrl arteries also constricted to the same extent when tested with 100 nM ANG II (Fig. 9A). Furthermore, the amplitude of the ANG II-induced rise in [Ca²⁺]_CYT was also virtually identical in Ctrl and TG arteries (i.e., the signal, itself, was not amplified), but it was superimposed on a higher basal [Ca²⁺]_CYT in TG arteries (Fig. 9B). A possible explanation for the lack of signal amplification is that compensatory mechanisms (e.g., reduced SERCA expression) might be triggered to minimize the effects of the enhanced Ca²⁺ entry in TG mouse ASMCs; this needs verification. Importantly, our measurements of [Ca²⁺]_CYT are spatial averages of cytoplasmic Ca²⁺. We cannot eliminate the possibility that overexpression of NCX1 changed Ca²⁺ concentration in other cellular compartments in which Ca²⁺ signaling occurs, such as the endoplasmic reticulum or nuclear envelope. Thus, differences between Ctrl and SM-NCX1-TG mice could, in ways yet unknown, be due also to changes in Ca²⁺ concentration in these compartments.

In contrast, femoral arteries of mice with ANG II-induced hypertension give attenuated [Ca²⁺]_CYT and vasoconstrictor responses to superfused ANG II relative to Ctrl mice (58). In that case, the sustained elevation of plasma ANG II likely caused downregulation of AT₁Rs (56) to compensate (partially) for the hyperstimulation. There is no reason to expect hyperstimulation of either α₁-adrenoceptors or AT₁Rs in NCX1-TG hypertension; consequently, agonist-induced downregulation of the cognate receptors seems unlikely.

If TG arteries do not have enhanced sensitivity to vasoconstrictors, what accounts for the augmented BP elevation in TG mice, versus Ctrl mice, when ANG II is infused (Fig. 5)? A likely explanation is based on the evidence that basal arterial diameter is smaller and arterial tone is greater not only in TG mouse muscular arteries than in Ctrl mouse muscular arteries (Fig. 2), but, importantly, also in resistance arteries (unpublished observations), which account for most of the total peripheral resistance (TPR). Furthermore, artery resistance (R) is proportional to 1/r⁴, where r is artery radius (4). Under these circumstances, the same vasoconstrictor-induced artery diameter decrease in Ctrl and TG mice should cause a greater increase in TPR and BP in TG mice [assuming constant cardiac output, since BP = cardiac output × TPR (4)]. In other words, the vasoconstrictor-induced increase in TPR, and thus in BP, is amplified in TG mice simply because the increase is superimposed on a greater basal tone due to the increased basal [Ca²⁺]_CYT.

SNP- But Not ACh-Induced Vasodilation Is Enhanced in TG Mouse Arteries

Endothelium-dependent vasodilation, as measured with ACh stimulation (38), is often impaired in arteries of hypertensive humans and animal models (55). One might anticipate, therefore, that ACh-induced vasodilation would be attenuated in TG mice with chronically elevated BP. In contrast, the data shown in Fig. 11 reveal that the response of TG mouse arteries to ACh is normal. This implies that the endothelial impairment in hypertension may not simply be the result of the elevated intravascular pressure (19). When vasodilation was induced with SNP, however, the dose-response curve was shifted to the left in TG arteries; i.e., sensitivity to the NO donor was increased: the apparent EC₅₀ was ~100 nM in Ctrl arteries but only ~20 nM in TG arteries (Fig. 11). This is not totally unexpected because there are reports showing that the sensitivity to SNP, largely an endothelium-independent vasodilator (for evidence of endothelium-dependent effects, see Ref. 28), is increased in some forms of hypertension (12, 52). In arterial myocytes, NO stimulates PKG to increase the cGMP level; this activates myosin light chain phosphatase, which promotes vasodilation by reducing myocyte sensitivity to Ca²⁺ (11), although other mechanisms may also contribute (33, 64). It seems possible that one or more of these vasodilator pathways were augmented to help compensate for the tendency of NCX1 overexpression to drive BP up. Indeed, others have suggested that Ca²⁺ sensitivity may be reduced in hypertension (1).

Conclusions

The aim of the present in vivo study was to test the seemingly intuitive hypothesis that elevated arterial myocyte [Ca²⁺]_CYT and enhanced myocyte Ca²⁺ signaling can explain the mild hypertension in NCX1-TG mice. While TG mice have high basal [Ca²⁺]_CYT and enhanced arterial tone, the data indicate that arterial myocytes are not hypersensitive to vasoconstrictors, as also has been reported for human small arteries (1, 51). Aalkjaer and colleagues (1) attributed the enhanced agonist-induced pressor response without a change in agonist sensitivity to “altered vascular structure” but “unchanged or possibly depressed” myocyte function. In contrast, however, the present results indicate that the increased pressor response in TG hypertensive mice is largely due to the superimposition of normal agonist-induced contraction on an artery with already enhanced basal tone. Despite the (presumably) congenitally high BP, the arterial functional responses appeared normal: arteries constricted normally to PE and ANG II and dilated to a PD that was only ~6–7% narrower than that of Ctrl arteries in response to ACh and SNP. This suggests that there is relatively little structural remodeling (21, 31) of TG arteries, despite longstanding hypertension. Thus, at least in NCX1-TG hypertension, the primary defect appears to be the NCX1-mediated elevation of [Ca²⁺]_CYT and its functional consequences, most notably, enhanced basal tone, that lead directly to increased TPR. Because arterial NCX1 is upregulated in many forms of hypertension (10, 42, 43, 65), it seems reasonable to anticipate that similar functional (active contraction) changes likely also apply at least during the early stages of hypertension. We also made the unexpected observation that endothelial function is not attenuated in femoral arteries of modestly hypertensive SM-NCX1-TG mice (Fig. 11) and that vasodilator mechanisms are even augmented (Fig. 12). These observations seem inconsistent with the popular view that BP elevation, per se, impairs endothelium-dependent vasodilation (35).

The key observation is that, while sympathetic drive is the dominant mechanism responsible for sustaining arterial tone, Ca²⁺ entry via NCX1 plays a small but crucial role in helping to maintain [Ca²⁺]_CYT above the contraction threshold. The critical relationship between [Ca²⁺]_CYT and arterial tone (Fig. 13) illustrates how enhanced net NCX1-mediated Ca²⁺ entry can increase tone. Net Ca²⁺ transport by NCX1 is governed by the PM Na⁺ electrochemical gradient (see the introduction and Ref. 9) that is controlled by adjacent α₂-Na⁺ pumps in myocyte PLasmERosomes (see the introduction and Refs. 8, 27, and 39). Apparently, α₂-Na⁺ pump activity is ultimately responsible for modulating arterial tone, TPR, and BP through NCX1 (7). The steep relationship between [Ca²⁺]_CYT and arterial tone (Fig. 13) suggests that very modest increases in SNA (40) and plasma endogenous ouabain (5, 20), which increases arterial NCX1 expression (43, 66), may act in concert to elevate BP in many forms of hypertension.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants R01-HL-107654 (to W. G. Wier) and R01-HL-45215 and R01-HL-107555 (to M. P. Blaustein), American Heart Association Grant 15GRNT24940022 (to J. Zhang and then to M. P. Blaustein), National Natural Science Foundation of China Grants 81100174 and 81570449 (to Y. Wang), Fundamental Research Funds for the Central Universities Grant GK201803097 (to Y. Wang), College Students’ Innovative and Training Project of Shaanxi Normal University Grant cx2018197 (to Y. Wang), and the University of Maryland Baltimore Foundation (Center for Heart Hypertension and Kidney Diseases Fund). This work was also supported by the University of Maryland School of Medicine Physiological Phenotyping Core (L. Chen, Core Manager).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

J.Z., Y.W., and L.C. performed experiments; J.Z., Y.W., L.C., and M.P.B. analyzed data; J.Z., Y.W., W.G.W., and M.P.B. interpreted results of experiments; J.Z. and M.P.B. drafted manuscript; J.Z., Y.W., L.C., W.G.W., and M.P.B. approved final version of manuscript; Y.W. and M.P.B. prepared figures; Y.W., L.C., W.G.W., and M.P.B. edited and revised manuscript.