The specific role of different isoforms of the Na,K-pump in the vascular wall is still under debate. We have previously suggested that the α2 isoform of the Na,K-pump (α2), Na+, Ca2+-exchange (NCX), and connexin43 form a regulatory microdomain in smooth muscle cells (SMCs), which controls intercellular communication and contractile properties of the vascular wall. We have tested this hypothesis by downregulating α2 in cultured SMCs and in small arteries with siRNA in vivo. Intercellular communication was assessed by using membrane capacitance measurements. Arteries transfected in vivo were tested for isometric and isobaric force development in vitro; [Ca2+]i was measured simultaneously. Cultured rat SMCs were well-coupled electrically, but 10 μM ouabain uncoupled them. Downregulation of α2 reduced electrical coupling between SMCs and made them insensitive to ouabain. Downregulation of α2 in small arteries was accompanied with significant reduction in NCX expression. Acetylcholine-induced relaxation was not different between the groups, but the endothelium-dependent hyperpolarizing factor-like component of the response was significantly diminished in α2-downregulated arteries. Micromolar ouabain reduced in a concentration-dependent manner the amplitude of norepinephrine (NE)-induced vasomotion. Sixty percent of the α2-downregulated arteries did not have vasomotion, and vasomotion in the remaining 40% was ouabain insensitive. Although ouabain increased the sensitivity to NE in the control arteries, it had no effect on α2-downregulated arteries. In the presence of a low NE concentration the α2-downregulated arteries had higher [Ca2+]i and tone. However, the NE EC50 was reduced under isometric conditions, and maximal contraction was reduced under isometric and isobaric conditions. The latter was caused by a reduced Ca2+-sensitivity. The α2-downregulated arteries also had reduced contraction to vasopressin, whereas the contractile response to high K+ was not affected. Our results demonstrate the importance of α2 for intercellular coupling in the vascular wall and its involvement in the regulation of vascular tone.
- Na+, Ca2+ exchanger
- smooth muscle cell synchronization
- short interfering RNA
- norepinephrine contractility
- endothelium-dependent hyperpolarizing factor
the electrogenic na,k-pump modifies numerous cellular pathways by modulating Na+-coupled transport. The functional significance of the Na,K-pump is dependent upon the isoform. The catalytic α-subunit exists in three isoforms (5). These isoforms have different affinities for cardiac glycosides, kinetics, and regulation but show high structural homology and are difficult to distinguish at a functional level. The widespread distribution of the relatively ouabain-resistant (in rodents) α1 isoform (51) suggests a ′house-keeping′ function in controlling Na+/K+ homeostasis (24). Many cell types have in addition an ouabain-sensitive isoform (either α2 or α3), which accomplishes more specific regulatory functions. We (30, 31) as well as others (13, 15, 41, 47) have previously identified the expression of α1 and α2 isoforms (α2) of the Na,K-pump in mesenteric small arteries. It has been documented that inhibition of the ouabain-sensitive Na,K-pump has significant effects on vascular function without changing smooth muscle cell (SMC) Na+ concentration (2, 4, 31).
Evidence has been provided that the Na,K-pump is important for nitric oxide- and prostanoid-independent arterial relaxation, i.e., the endothelium-dependent hyperpolarizing factor (EDHF) response (12, 14). EDHF is suggested to be a direct transfer of hyperpolarization through myoendothelial gap junctions or an increase of the local extracellular K+ concentration, although other mechanisms are also suggested. It is possible that K+ efflux through the endothelial Ca2+-activated K+ channels increases near myoendothelial junctions to cause a local increase of the K+ concentration, which, in turn, activates the SMC Na,K-pump providing SMC hyperpolarization (12, 14). Which catalytic subunit of Na,K-pump is important for this signaling is unclear, but it has been shown to be a ouabain-sensitive subunit and α2 has been shown to be present at the myoendothelial junction (12).
Using membrane capacitance measurements, we suggested that a Na,K-pump with high ouabain-sensitivity affects intercellular communication via localized changes in intracellular calcium ([Ca2+]i) through modulation of Na+, Ca2+-exchange (NCX) activity (31). We have further demonstrated that this modulation of intercellular communication could be important for controlling synchronization between SMCs in the vascular wall and hence for producing vasomotion in small arteries (1, 30, 31). Nevertheless, the precise role of α2 in modulation of intercellular communication is lacking.
It is also unclear whether α2 is important for the sensitivity and maximal contraction of blood vessels to vasoconstrictors. Evidence has been provided from work with α2 isoform genetically modified mice that the myogenic tone is increased after α2 knockout consistent with the effect of ouabain (1, 4, 13, 21, 34, 53). The role of α2 for agonist-induced contraction has only been addressed in one study where aorta from α2 knockout mice showed increased contractile response to thromboxane (49). It seems, however, that the role of α2 can vary depending on the type of vasculature and agonist (47). The results from mouse aorta (49) are interesting in light of recent findings indicating a close interaction between expression of the α2 isoform and NCX together with the demonstration that reduction in NCX activity leads to reduced arterial contractility and lowering of blood pressure (27, 41, 43, 45, 54). In the current study, we address the role of α2 for SMCs synchronization and for the control of tone in rat mesenteric small arteries. The α2 isoform was downregulated in arterial segments using a novel short interfering RNA (siRNA) transfection approach in vivo (8), and these arteries were compared with control arteries transfected with nonrelated siRNA.
All experiments were approved by and conducted with permission from the Animal Experiments Inspectorate of the Danish Ministry of Justice.
Transfection of segments of rat mesenteric small arteries in vivo was performed using a modification of our previously described method (8, 32). TKO-based transfection has been used (Mirus Bio). TKO was mixed (1:10 volume ratio) with a sterile isotonic salt solution (9 mg/ml NaCl) and left at room temperature. After 10 min this solution was mixed with siRNA (final concentration, ∼800 nM). Fifteen minutes later the solution was used for transfection.
Wistar male rats were anesthetized with a subcutaneous injection of a combination of hypnorm (1 mg/100 g; VetaPharma) and midazolam (0.5 mg/100 g; Hameln Pharm). Supplementary anesthesia was administered half-hourly during the transfection procedure. At the end of the transfection, and twice during the day of the surgery, a painkiller (Temgesic, 2 ml/100 g; Schering-Plough) was injected.
A medial laparotomy was performed, and a section of the mesentery was gently pulled out. The tissue was kept moist with a sterile isotonic salt solution (9 mg/ml NaCl) and covered with moistened gauze. A short segment (∼1 mm) of a first-order branch from the superior mesenteric artery was gently cleaned of fat under a dissection microscope, and a piece of suture was placed upstream around the cleaned segment temporarily restricting blood flow. A puncture hole was made, and the tip of an elastic micropipette (MicroFill 34G-5; World Precision Instruments) connected to the syringe filled with the transfection solution was inserted ∼0.5 mm into the artery. Downstream second- and third-order mesenteric small arteries (∼250–350 μm outer diameter) were flushed with the transfection solution and clamped downstream to avoid backflow of blood from the mesenteric arcade. Arteries were carefully observed for any contamination of the transfected segment with blood. Transfection was performed within 20 min; each 2.5 min the downstream clamp was released and a new portion of the transfection solution was flushed through. An arterial branch adjacent to the transfected artery was labeled with either one or two (depending on the transfection type) ligature loops without removing fat and without compressing the artery. These ligature loops helped to identify the transfected arteries later in the study. After transfection the mesentery was returned to the peritoneal cavity, and the wound was closed.
Two arterial branches were transfected: one with siRNA directed against the α2 isoform (Ambion; the sense sequence: 5′-GAGAACATCTCCGTGTCAtt-3′) and another with the control, nonrelated siRNA directed against enhanced green fluorescent protein (Ambion; the sense sequence: 5′-CCACUACCUGAGCACCCAGtt-3′).
Isometric force measurement.
Rats were euthanized with CO2 3 days after transfection. The transfected and nontreated second- and third-order branches of the mesenteric artery were dissected out and cleaned of connective tissue in ice-cold salt solution (PSS) containing (in mM) 119 NaCl, 4.7 KCl, 1.18 KH2PO4, 1.17 MgSO4, 25 NaHCO3, 1.6 CaCl2, 0.026 EDTA, and 5.5 glucose, gassed with 5% CO2 in air and adjusted to pH 7.4. The cleaned arterial segments were mounted in an isometric wire myograph (Danish Myo Technology) as described previously (40). The myograph chamber was heated to 37°C while the PSS was constantly aerated with 5% CO2 in air. The artery diameter was set to a value where maximal active force is obtained (37
Concentration-response relationships were constructed by cumulative addition of norepinephrine (NE) or vasopressin. In these experiments, the unstimulated (baseline) wall tension was defined as zero (37) while the contractile responses were expressed as absolute increases in active wall tension (i.e., increase from baseline). The role of intracellular stores was tested by applying 10 mM caffeine or 10 μM NE to arteries immersed in Ca2+-free solution (PSS without CaCl2 and with 5 mM EGTA). Endothelium function was tested by applying 10 μM acetylcholine to submaximally preconstricted arteries (by 5 μM NE) under control conditions and after inhibition of NO production [by 100 μM Nω-nitro-l-arginine methyl ester (l-NAME)] and prostacyclin production (by 3 μM indomethacin) and then in the presence of l-NAME, indomethacin, apamin (50 nM), and TRAM34 (1 μM). Relaxation was expressed in percentage from the preconstricted level (0%) to passive wall tension (100%). Vasomotion was analyzed at the level of tone where the amplitude was highest [∼50% of maximal tone (7)]. Vasomotion amplitude was normalized to the maximal active wall tension to agonist stimulation (in amplitude/tensionmax).
Second- and third-order branches of the mesenteric artery were dissected and cannulated at both ends using glass microcannulas and mounted in a pressure myograph (111P; DMT). Experiments were performed under no-flow conditions. Vessel diameter was measured from video images of the preparation using contrast analysis (DMT Vessel Acquisition Suite; DMT). Passive outer diameters (see below) at 60 mmHg were 298 ± 15 μm (n = 6) and 319 ± 18 μm (n = 9) of arteries transfected with nonrelated siRNA and siRNA against α2 isoform of the Na,K-pump.
Arteries were equilibrated at 37°C and pressure was cycled between 10 and 120 mmHg to decrease mechanical hysteresis. Cumulative concentration-response curves to NE from 0.1 to 30 μM were obtained at a transmural pressure of 60 mmHg. The contractile responses were expressed in percent outer passive diameter measured at the same transmural pressure (60 mmHg).
The transmural pressure was then set to 4 mmHg, and 20 min later the vessel was subjected to a series of pressure steps (4, 20, 40, 60, 80, and 100 mmHg). The pressure-step protocols were repeated for arteries under control conditions (no tone development was seen in any experimental group; data not shown) and in the presence of 0.1 μM NE. At the end of the experiment, the vessel was maintained in Ca2+-free solution with 30 μM papaverine for 15 min and subjected to another series of pressure steps, during which the pressure-diameter relation was obtained. The external diameters obtained under these conditions were considered passive outer diameters. The degree of tone at each level of pressure was quantified as 100%·[(passive outer diameter − the outer diameter under the given experimental conditions)/passive outer diameter].
Measurement of [Ca2+]i in the arterial wall.
To obtain ratiometric measurements of wall [Ca2+]i, arteries mounted in a pressure myograph, as described above, were loaded with 2.5 μM fura 2-acetoxymethyl ester (fura 2-AM) for 2 h. Fura 2-AM was dissolved in DMSO with 0.1% (wt/vol) cremophor and 0.02% (wt/vol) pluronic F127. Arteries were excited by a 75W xenon light source alternately at 340 and 380 nm (frequency 110 Hz), and emitted light was measured at 515 nm (frequency 1 Hz). Background fluorescence (after quenching with 20 mM MnCl2) was determined and subtracted from obtained measurements. Fluorescence was collected and stored digitally using Felix32 software (version 1.2; Photon Technology). [Ca2+]i was expressed as the ratio of fluorescence during 340 and 380 nm excitation.
Membrane capacitance measurement for evaluation of intercellular coupling.
Intercellular coupling between cultured cells was evaluated as described previously (8, 31, 33). Because the membrane capacitance per area of cell membranes is nearly constant, a decrease in capacitance would reflect a reduction in membrane area (11). Rat aortic SMCs (A7r5) were transfected as described previously (29). A7r5 cells were cultured in DMEM medium (In Vitro, Denmark) supplemented with 10% fetal calf serum, 1% L-glutamine, and 0.1% KPS [kanamycin 2 g, penicillin 1 × 106 IU, streptomycin 1 g in 20 ml phosphate-buffered salt solution (PBS)]. Confluent cells were detached from the culture dishes by nonenzymatic cell dissociation solution (Sigma-Aldrich, Denmark) and pipetted into new culture dishes (Falcon, Becton Dickson, Denmark). After 4–6 h the medium was replaced with extracellular solution and paired cells used for patch-clamp.
Capacitance measurements were performed using the Membrane Test tool of Clampex 7 (Axon Instruments) and were based on the time constant of current decay following a voltage step (11). All recordings were made at room temperature (22–24°C). Recordings were made with an Axopatch 200B amplifier (Axon Instruments) in conventional whole-cell configuration with a low access resistance (6–12 MΩ). Data acquisition and analysis were done with the software package Clampex 7 for Windows (Axon Instruments). Drugs and solutions were applied to the bath locally over the patched cell.
A7r5 cells were resuspended as described above. After 4–6 h the medium was replaced with 4% formaldehyde for 15 min, washed three times with the PBS containing (in mmol/L) 137 NaCl, 2.7 KCl, 8.2 Na2HPO4, and 1.8 KH2PO4, at pH 7.4. Unreacted fixative was then quenched with 25 mmol/L glycine in PBS for 15 min. Cells were washed three times with PBS, permeabilized with 0.1% Triton-X in PBS for 15 min, and incubated with primary connexin43 (Cx43) antibody (1:100; Invitrogen) for 2 h at room temperature. After three washes, cells were incubated in the dark with Alexa-488 fluorescent conjugated secondary antibody for 45 min at room temperature (1:4,000; Invitrogen). After a wash, the preparation was transferred to the confocal microscope (LSM-5 Pascal Exciter, Zeiss, Germany). The emission signal at 530 nm (after excitation at 488 nm) was stored on the computer for later analyses of fluorescence intensity using ImageJ (National Institutes of Health). Fluorescence intensities of whole cell, a central area, and a near-membrane area were measured and normalized for total area. The intensity of Cx43 specific fluorescence was used as an estimate of the Cx43 protein expression level.
The RNA isolation was carried out with Qiagen micro kit (Qiagen, VWR, Denmark). Isolated arterial segments were disrupted in a Tissue Lyser (Qiagen). The samples were treated with proteinase K (1 mg/ml, 20 min, 37°C; Invitrogen) after homogenization. PCR was performed to assess the expressions of specific RNAs. The reaction was executed with reverse transcriptase III (Invitrogen) and superase (Ambion) for deactivation of RNAse and DNAse.
Primer sets for quantitative PCR analyses for α1 and α2 isoforms of the Na,K-pump, NCX, Cx43, GAPDH, and transferrin receptor were obtained from Applied Biosystems (Denmark). Quantitative PCR was carried out on MX3000P (Stratagene) using Taqman probe (FAM) technology. Gene expression was normalized to GAPDH and transferrin receptor (average Ct value) presented by a ΔCt value. Comparison of gene expression between nontransfected control and transfected arteries was derived from subtraction of control ΔCt from transfected ΔCt value, producing ΔΔCt. Relative gene expression was calculated as 1/(2ΔΔCt), thereby standardized to nontransfected control arteries from the same rat (8).
Two similarly treated branches (∼4-mm-long segments) of mesenteric small artery from the same rat were lysed in 25 μl lysis buffer containing (in mM) 10 Tris·HCl, 250 sucrose, 1 EDTA, 1 EGTA, 2% Triton X-100 (pH 7.4), and 1 tablet protease inhibitor per 10 ml as described previously (8). The homogenate was centrifuged at 10,000 g, and the supernatant was collected. Due to very small volumes, it was only possible in some cases to measure the protein concentration. The protein-containing supernatant was adjusted with 1 mol/L DTT and 2× Tris·glycine SDS sample buffer (Invitrogen, Denmark) with an approximate ratio of 10:3:3, respectively. In cases where protein concentration was obtained, the same amount of protein was loaded; otherwise up to 14 μl of loading mixture was loaded to gels. Proteins were separated on 10% Tris·glycine gels and electrotransferred onto nitrocellulose membranes, which were then blocked by incubating in 5% nonfat dry milk in PBS with 0.5% vol/vol Tween 20 (PBS-T). The membranes were incubated with primary antibodies α1 isoform of the Na,K-pump antibody (1:2,000) and NCX antibody (1:1,000) (Santa Cruz Biotechnology), α2 isoform of the Na,K-pump antibody (1:2,000; Millipore), Cx43 antibody (1:500; Invitrogen) overnight at 5°C in PBS-T. After a wash, the membranes were incubated with horseradish-peroxidase (HRP)-conjugated secondary antibody (1:4,000; Dako, Denmark) for 1 h in PBS-T. Excess antibody was removed by extensive washing, and bound antibody was detected by an enhanced chemiluminiscence kit (ECL; Amersham). Membranes were then stripped for antibodies and stained for pan-actin (Cell Signaling Technology) with a horseradish-peroxidase-conjugated secondary antibody. Detected protein was quantified using the ImageJ program [National Institutes of Health (NIH)] as a ratio to pan-actin measured for the same probe. Quantitative comparison of the amount of protein between the probes is difficult since it is impossible to set absolute ′0′ and maximum (or 100%) values of expression from the blots. We consider therefore our quantification of protein expression by Western blot to be semi-quantitative.
Microsoft Excel and GraphPad Prism software (v.5.02 for Windows) were used for graphing and statistical analysis. Data are summarized as mean values ± SE of the sample group. Significant differences between means were determined by unpaired and paired one-way ANOVA followed by Bonferroni post test. Where applicable, two-way ANOVA and t-test were used. A probability (P) level of <0.05 was considered significant, and n refers to the number of arteries per rats as equally treated arteries from the same animal were pooled.
Downregulation of the α2 isoform Na,K-pump uncouples electrically coupled SMCs.
We have previously shown that ouabain inhibits intercellular communication between vascular SMCs (31) and suggested α2 to be involved in this function (30). To test this we downregulated α2 in cultured A7r5 SMCs using siRNA (Fig. 1). Transfection with specific siRNA significantly reduced the α2 isoform protein expression compared with the level in nontransfected cells (Fig. 1, A and B) and SMCs transfected with nonrelated siRNA.
The electrical communication between pairs of SMCs, evaluated as membrane capacitance (11), was significantly reduced by downregulation of the α2 isoform compared with nontransfected cells and cells transfected with nonrelated siRNA (Fig. 1, C and D). In accordance with our previous findings (31), 10 μM ouabain reduced membrane capacitance of paired SMCs by half, suggesting electrical uncoupling. The α2 isoform downregulated SMCs did not significantly change their membrane capacitance after ouabain application (Fig. 1, C and D). These results suggest that reduction of α2 expression reduces electrical coupling between SMCs, making them insensitive to micromolar concentrations of ouabain.
Downregulation of the α2 isoform Na,K-pump reduces expression of NCX and Cx43 proteins.
Using a recently described approach (8) for in vivo transfection of mesenteric small arteries with siRNA, we downregulated mRNA for the α2 isoform to 12.1 ± 2.6% (P < 0.05, n = 9) of the nontransfected level in mesenteric small arteries, whereas transfection with nonrelated siRNA was without significant effect on α2 mRNA levels. Importantly, expression of the housekeeping α1 isoform of the Na,K-pump was not significantly changed by transfection (121.9 ± 21.0%, n = 7).
Reduction in α2 mRNA correlates with a significant decrease in α2 protein expression (Fig. 2A). In the arteries transfected with the α2-targeting siRNA, protein expression for α2 was reduced to 32.3 ± 9.9% (n = 4) of the control level and was significantly lower than in arteries transfected with nonrelated siRNA. The α1 isoform protein was unaffected by siRNA transfection (Fig. 2B). Importantly, when α2 was downregulated a significant reduction in NCX expression was observed (Fig. 2C). NCX expression was unchanged in the arteries transfected with nonrelated siRNA. Ouabain has been suggested to modify trafficking and reduce expression of Cx43 (28). In the arteries where α2 was downregulated Cx43 expression was not significantly reduced (72.9 ± 10.4% n = 4; Fig. 2D); however, in α2-downregulated A7r5 cells, Cx43 expression was significantly reduced (to ∼57%; Fig. 2, E and F). Furthermore, the subcellular distribution of Cx43 was affected in the α2-downregulated A7r5 cells, i.e., the proportion of Cx43 in membrane near regions relative to Cx43 in central areas was significantly reduced (Fig. 2G).
Downregulation of the α2 isoform of the Na,K-pump has complex effects on contractile responses of mesenteric small arteries.
Mesenteric small arteries downregulated for the α2 were tested in vitro. Under isometric conditions the sensitivity to NE was reduced in arteries downregulated for α2 (Fig. 3, A and B). Ouabain increased the sensitivity of control arteries to NE but was without effect on arteries downregulated for the α2 isoform (Fig. 3, A and B). In the cumulative concentration-response experiments the contraction to 10 μM NE was not significantly different between the groups. However, a discrete activation with 10 μM NE did reveal a reduced contraction in arteries downregulated for α2 (Fig. 3C) as was the contractile response to 10 μM NE in Ca2+-free conditions in these arteries (Fig. 3D). However, in Ca2+-free conditions the contractile response to 10 mM caffeine was similar between the groups (Fig. 3D). Arteries downregulated for α2 developed less tension in response to high concentrations of vasopressin (AVP; Fig. 3E), although the sensitivity to this agonist was not affected. Finally, there was no difference observed in the contractile responses to K+-induced depolarization between control and α2-downregulated arteries (Fig. 3F).
Under isobaric conditions arteries downregulated for α2 also had reduced maximal contractile response to NE (Fig. 4A) similar to the finding under isometric conditions, although no difference in the sensitivity to NE was seen between the two groups (−logEC50 were 5.79 ± 0.17 and 5.81 ± 0.16, control and downregulated, respectively; n = 5–9). Under isobaric conditions (at 60 mmHg) unstimulated arteries downregulated for α2 had a significantly elevated fura-2 fluorescence ratio ([Ca2+]i) (Fig. 4B). NE increased [Ca2+]i in a concentration-dependent manner (Fig. 4B). Changes in vessel diameter as a function of [Ca2+]i (Fig. 4C) suggest a reduced sensitivity to Ca2+ in the arteries downregulated for α2.
The mesenteric arteries from all experimental groups had no active tone under control conditions, i.e., the diameters in PSS and in Ca2+-free solution with papavarine were similar (n = 5–8, data not shown). In the presence of 0.1 μM NE tone was produced at all pressures. Under these conditions arteries downregulated for α2 had significantly higher tone than control arteries over all pressures (Fig. 5A). The increased tone was associated with an increased [Ca2+]i (Fig. 5B).
Downregulation of the α2 isoform Na,K-pump impairs endothelium-derived hyperpolarizing factor (EDHF) component of endothelium dependent relaxation.
Downregulation of α2 was without effect on the endothelium-dependent relaxation to acetylcholine under control conditions (Fig. 6). Inhibition of the NO-dependent pathway with l-NAME and the prostacyclin-dependent pathway with indomethacin suppressed maximal (10 μM) acetylcholine-induced relaxation of the arteries downregulated for α2 but was without effect on maximal relaxation of the arteries transfected with nonrelated siRNA (Fig. 6). The combination of l-NAME, indomethacin, TRAM34, and apamin completely abolished relaxation in both experimental groups (Fig. 6). Thus the EDHF-like (endothelium-dependent, NO- and COX-independent) vasorelaxation was suppressed in α2-downregulated arteries.
Downregulation of the α2 isoform Na,K-pump suppresses the amplitude of vasomotion.
When stimulated with NE the control arteries produced rhythmic contractions, i.e., vasomotion (Fig. 7A). Downregulation of α2 abolished vasomotion in six out of 10 arteries while the remaining four arteries continued to oscillate with an amplitude similar to that of arteries transfected with nonrelated siRNA (0.12 ± 0.035 vs. 0.12 ± 0.026, n = 4 and 10, respectively). Ouabain significantly reduced vasomotion amplitude in all control arteries. In the four downregulated arteries that oscillated, ouabain had no effect on the vasomotion (Fig. 7B). Ouabain was without effect on the frequencies of oscillations (Fig. 7C).
Based on genetic and pharmacological approaches, the α2 isoform of the Na,K-pump has been suggested to be important for the control of vascular tone (4, 13, 21, 26, 42, 43, 49, 53) and, thus, blood pressure regulation (6). In blood vessels ouabain increases myogenic tone (53), inhibits intercellular communication and thus vasomotion (19, 28, 31), and reduces the EDHF response (12). It has been suggested that inhibition of the α2 isoform is responsible for these effects. This conclusion is based on 1) the expression profile of the Na,K-pump isoforms in the arterial wall (13, 24, 27, 31, 34, 42, 49), 2) the relatively high ouabain-sensitivity of α2 over α1 isoforms of the Na,K-pump in rodents (51), and 3) colocalization of α2 with other proteins, e.g., NCX, Cx43, and intermediate conductance Ca2+-activated K+ channels (12, 30, 35, 41, 49). This suggests that α2 could play an important modulatory role in the control of vascular tone in small arteries. To test the hypothesis we assessed the effect of in vivo knockdown of α2 in a single arterial segment.
Downregulation of α2 isoform inhibits intercellular communication similarly to pharmacological inhibition with ouabain.
In accordance with the effect of ouabain (31, 33) downregulation of the α2 isoform electrically uncoupled cultured SMCs as evident from membrane capacitance measurements. Our laboratory (31) and others (4, 16, 17, 25, 34) have previously shown that ouabain affects spatially restricted submembrane [Ca2+]i (in a NCX-dependent way) and suggested that this reduces gap junction conductance. It has been shown that chronic loss of the α2 isoform in SMCs is accompanied with the loss of NCX. This could possibly inhibit clearance of store-released Ca2+ (27), which would increase spatially restricted [Ca2+]i near the membrane and thereby inhibit gap junction opening. Consistent with this we found downregulation of NCX consequent to downregulation of α2. How these local [Ca2+]i signals affect gap junctions, though, is unclear. Because Ca2+ has previously been shown to reduce permeability of gap junctions formed by Cx43 (48, 50), a direct effect of Ca2+ is possible but an involvement of other Ca2+-sensitive messengers cannot be excluded (36).
Long-term exposure to 100 μM ouabain has previously been shown to suppress the expression of Cx43 (28). Consistent with this, in cultured SMCs where α2 was downregulated we found a reduction in overall Cx43 expression and reduced near-membrane expression relative to central expression suggesting internalization. These changes did not achieve significance in the experiments on transfected arteries. Thus the mechanism responsible for inhibition of electrical communication in the vascular wall of α2 isoform downregulated arteries remains unclear. This could either be due to a moderate reduction in Cx43 (which could not be demonstrated in the intact vascular wall, possibly due to a limited number of experiments) or a consequence of closure of gap junctions by localized Ca2+ signaling (31), or a combination.
Loss of intercellular communication leads to inhibition of vasomotion (1, 3, 18, 23, 38–40). We suggest that the reduced oscillatory activity of the arteries downregulated for α2 was a consequence of reduced synchronization between SMCs. Vasomotion frequency was unaffected by the downregulation, consistent with this suggestion (8, 22, 31). Furthermore, ouabain treatment was without significant effect on vasomotion amplitude in the downregulated arteries that oscillated, whereas in the control group of arteries ouabain reduced the amplitude of vasomotion. These findings strongly indicate the involvement of the α2 in modulation of intercellular communication between SMCs in the arterial wall (30). The reason why some α2-downregulated arteries remained oscillating is unclear, but it is unlikely to be due to remaining expression of α2 since oscillations in these arteries were insensitive to ouabain. An “escape” from inhibition of vasomotion has been reported previously (7, 44) so the oscillations observed in this study may further reflect the complexity of the mechanisms important for vasomotion, where α2 is only one of many modulatory factors (1, 30).
Downregulation of the α2 isoform Na,K-pump increases [Ca2+]i.
The arteries downregulated for α2 had elevated resting [Ca2+]i as evaluated by fura-2 fluorescence under isobaric conditions (Figs. 4B and 5B). The elevated [Ca2+]i could be a consequence of reduced NCX-mediated Ca2+ extrusion since it has been shown that NCX accounts for 90% of Ca2+ efflux in vascular SMCs and knockout of α2 eliminates clearance of intracellular Ca2+ by NCX and increases capacitive Ca2+ entry (27). Our finding of reduced NCX expression is consistent with this. Interestingly, activation of arteries with NE produced a smaller increase in [Ca2+]i in arteries downregulated for α2 compared with control arteries. It will be of interest in future experiments to find the reason for this. The activity of the NCX may partly explain the blunted NE-mediated increase in [Ca2+]i after α2 downregulation; since NCX can mediate both Ca2+ entry and exit (6) it might play a different role during NE stimulation such that when [Ca2+]i is high and the membrane depolarized, the electrochemical gradient for the NCX will differ from that under resting conditions.
Different effects of ouabain and α2 isoform downregulation on arterial contractility.
Micromolar concentrations of ouabain were without effect on the sensitivity of α2-downregulated arteries consistent with disappearance of functional α2, whereas ouabain, as expected, increased the sensitivity of the control arteries (4, 13, 21, 49, 53).
Previous studies have shown that in small arteries from α2 knockout mice myogenic tone is increased (13, 21, 49, 53). The current findings are consistent with this since rat small arteries downregulated for the α2 isoform had higher tone under isobaric conditions. In these experiments it was necessary to apply a low NE concentration to induce a small tone since no myogenic tone was seen otherwise, although a lack of or minimal myogenic tone in rat mesenteric small arteries is a frequent observation (20). The reason for the increased tone in the downregulation of the arteries for the α2 isoform is likely the increased [Ca2+]i as discussed above.
In contrast with the enhanced response at low NE concentrations, we found that the downregulated arteries had reduced maximal contractility to NE compared with the controls (both under isobaric and isometric conditions) and reduced sensitivity to NE under isometric conditions. The reason for the different observations under isobaric and isometric conditions with respect to NE sensitivity is unclear but vascular responses under isometric and isobaric conditions are known to differ (9), possibly due to changes in the Ca2+ entry/release pathways and alternate mechanisms for Ca2+ sensitization under the two conditions (9). The relationship between [Ca2+]i and tone shown in Fig. 4C strongly suggests that the reason for the reduced contractility to NE is a reduced Ca2+-sensitivity. Furthermore, the reduced contractility observed with another agonist (AVP) and normal responses to high potassium additionally support the idea that Ca2+-sensitivity during agonist stimulation is compromised after α2 downregulation. Furthermore, caffeine-stimulated Ca2+-release from the SR (in Ca2+-free bath solution) did not produce a significantly different contraction between the two groups, the contractile response to NE in the same experimental conditions was reduced in the α2-downregulated arteries. These findings suggest that the function and capacity of the SR is unaffected by α2-downregulation and that reduced Ca2+-sensitivity may be responsible for the decreased response to agonists. An alternative explanation could be that Ca2+ is released from different pools during the two modes of stimulation (SR-direct and -indirect). We do not know the reason for the compromised NE-induced Ca2+-sensitivity. It is not seen when ouabain acutely inhibits the Na,K-pump, so it seems this effect is a consequence of a more long-term downregulation of the Na,K-pump. Only one other study has assessed agonist-induced vascular contractility in α2 knockout mice: aorta from these mice was more sensitive to a thromboxane A2 analog than the wild-type control tissue (49). It is difficult to pinpoint why genetic ablation of α2 in mice or downregulation of the isoform in rat mesenteric arteries in this study results in opposing contractile effects, but it is known that the significance of the Na,K-pump for the contractile responses to NE differs between rat aorta, mesenteric, and tail arteries (47).
EDHF-like response is reduced in α2 isoform downregulated arteries.
Previous studies strongly suggested the importance of a ouabain-sensitive Na,K-pump in endothelium-dependent relaxation (EDR) (12, 14, 15). Submicromolar and micromolar ouabain was shown to significantly attenuate the EDR when NO- and prostacyclin-dependent pathways were inhibited, suggesting an important role of a ouabain-sensitive Na,K-pump in the EDHF response (10). Although the specific isoforms involved in this ouabain-sensitive Na,K-pump are not clear due to similar pharmacological properties of the different isoforms, several expression studies have indicated the importance of the α2 for EDHF (12, 15, 47, 52). Expression of α2 protein in SMCs and endothelial cells is controversial, but a recent study indicated localization of this isoform to myoendothelial contacts in the vascular wall (12).
Although the mechanism responsible for EDHF response may vary between different vascular preparations, it has been suggested that Ca2+-activated K+ channels opened during endothelial activation release K+ that accumulates in the myoendothelial restricted space and stimulates α2, which then hyperpolarizes SMCs (12, 14). Attenuation of the EDHF-like component of EDR in α2 isoform downregulated arteries may be attributable to a reduced sensitivity of SMCs to such an endothelial K+-efflux. It has, however, been argued that agonist-induced preconstriction produces a SMC-derived “K+-cloud” that effectively blocks responses to additional K+ ions (46). These authors also suggested that agonist stimulation maximizes the importance of gap junction-mediated communication and minimizes the contribution of endothelium-derived K+ ions. Thus the observed reduction in EDHF-like response in α2-downregulated arteries could be consequent to closure of the myoendothelial gap junctions (31), as discussed above. The current study cannot distinguish between these possibilities.
In the current study we have provided strong experimental support for involvement of the α2 isoform of the Na,K-pump in several vascular functions. Downregulation of the α2 isoform of the Na,K-pump reduced intercellular communication in the vascular wall and the EDHF response and had complex effects on vascular SMC [Ca2+]i and tone. These complex effects are likely associated with regulation of NCX and Cx43 expression as well as changes in vascular SMC Ca2+-sensitivity. Some of the findings are consistent with the effects of low concentrations of ouabain while other observations suggest additional effects of long-term inhibition of the α2 isoform of the Na,K-pump.
The study was supported by the Danish Research Council and the Novo Nordisk Foundation.
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: V.V.M., N.M.-N., D.M.B.B., and C.A. conception and design of research; V.V.M., N.M.-N., V.S.D., and Z.N. performed experiments; V.V.M., N.M.-N., and V.S.D. analyzed data; V.V.M., N.M.-N., D.M.B.B., and C.A. interpreted results of experiments; V.V.M. prepared figures; V.V.M., D.M.B.B., and C.A. drafted manuscript; V.V.M., D.M.B.B., and C.A. edited and revised manuscript; V.V.M., D.M.B.B., and C.A. approved final version of manuscript.
We thank Jane Holbæk Rønn and Jørgen Andresen for excellent technical assistance.
- Copyright © 2012 the American Physiological Society