Prolonged ouabain administration (25 μg·kg−1·day−1 for 5 wk) induces “ouabain hypertension” (OH) in rats, but the molecular mechanisms by which ouabain elevates blood pressure are unknown. Here, we compared Ca2+ signaling in mesenteric artery smooth muscle cells (ASMCs) from normotensive (NT) and OH rats. Resting cytosolic free Ca2+ concentration ([Ca2+]cyt; measured with fura-2) and phenylephrine-induced Ca2+ transients were augmented in freshly dissociated OH ASMCs. Immunoblots revealed that the expression of the ouabain-sensitive α2-subunit of Na+ pumps, but not the predominant, ouabain-resistant α1-subunit, was increased (2.5-fold vs. NT ASMCs) as was Na+/Ca2+ exchanger-1 (NCX1; 6-fold vs. NT) in OH arteries. Ca2+ entry, activated by sarcoplasmic reticulum (SR) Ca2+ store depletion with cyclopiazonic acid (SR Ca2+-ATPase inhibitor) or caffeine, was augmented in OH ASMCs. This reflected an augmented expression of 2.5-fold in OH ASMCs of C-type transient receptor potential TRPC1, an essential component of store-operated channels (SOCs); two other components of some SOCs were not expressed (TRPC4) or were not upregulated (TRPC5). Ba2+ entry activated by the diacylglycerol analog 1-oleoyl-2-acetyl-sn-glycerol [a measure of receptor-operated channel (ROC) activity] was much greater in OH than NT ASMCs. This correlated with a sixfold upregulation of TRPC6 protein, a ROC family member. Importantly, in primary cultured mesenteric ASMCs from normal rats, 72-h treatment with 100 nM ouabain significantly augmented NCX1 and TRPC6 protein expression and increased resting [Ca2+]cyt and ROC activity. SOC activity was also increased. Silencer RNA knockdown of NCX1 markedly downregulated TRPC6 and eliminated the ouabain-induced augmentation; silencer RNA knockdown of TRPC6 did not affect NCX1 expression but greatly attenuated its upregulation by ouabain. Clearly, NCX1 and TRPC6 expression are interrelated. Thus, prolonged ouabain treatment upregulates the Na+ pump α2-subunit-NCX1-TRPC6 (ROC) Ca2+ signaling pathway in arterial myocytes in vitro as well as in vivo. This may explain the augmented myogenic responses and enhanced phenylephrine-induced vasoconstriction in OH arteries (83) as well as the high blood pressure in OH rats.
- Na+ pumps
- Na+/Ca2+ exchange
- store-operated Ca2+ entry
- receptor-operated Ca2+ entry
- arterial myocytes
essential hypertension is an endemic, multifactorial disorder with an enormous impact on morbidity and mortality from cardiovascular and renal complications (40). The pathogenesis of essential hypertension is still poorly understood (14, 58), but the roles of excess dietary salt and NaCl retention are widely recognized (38, 54). A common feature of many animal models of hypertension is an elevated plasma level of endogenous ouabain (20, 29, 69), an adrenocortical hormone (29, 66). Plasma endogenous ouabain also is significantly elevated in ∼50% of patients with essential hypertension and is related to blood pressure (BP) even in the normal population (52, 65, 75). Also, short-term dietary salt loading increases plasma endogenous ouabain levels in humans (50). Moreover, in rodents, the prolonged administration of low doses of ouabain induces sustained BP elevation, termed “ouabain hypertension” (OH) (42, 51, 82).
Endogenous ouabain may play an important role in regulating arterial tone and peripheral vascular resistance (8, 11). Arterial smooth muscle cells (ASMCs) express Na+ pumps with an α1-subunit as well as Na+ pumps with a catalytic α2-subunit (18, 84). In rodents, α1-subunits of Na+ pumps have a very low affinity for ouabain (57). Thus, nanomolar ouabain exerts its vascular effects by preferentially inhibiting the high ouabain-affinity α2-subunit Na+ pumps, which are located in plasma membrane (PM) microdomains at PM-sarcoplasmic reticulum (SR) junctions (35). The consequent rise in the local, sub-PM Na+ concentration should, via adjacent Na+/Ca2+ exchangers (NCXs) (37), raise the local cytosolic free Ca2+ concentration ([Ca2+]cyt) in the junctional space, increase SR Ca2+ stores, and augment ASMC Ca2+ signaling (3, 11).
Ca2+ homeostasis plays a crucial role in the genesis of vascular myogenic tone, and increases in [Ca2+]cyt appear to underlie at least part of the increased peripheral vascular resistance in hypertension (34, 70, 84). Accumulating evidence has indicated that Ca2+ influx through PM store-operated channels (SOCs) and receptor-operated channels (ROCs), along with voltage-gated Ca2+ channels, may play a role in regulating myogenic tone and vasoconstriction (23, 56, 71, 77). C-type transient receptor potential channel proteins TRPC1, TRPC4, and/or TRPC5, mammalian homologs of the Drosophila TRP channel, form the endogenous SOCs that are activated by SR Ca2+ depletion (5, 60, 79, 80). In contrast, TRPC3 and TRPC6, which are components of ROCs, can be activated by diacylglycerols in a store depletion-independent manner (31).
Two recently discovered families of transmembrane proteins, Orai1 [also known as Ca2+ release-activated Ca2+ (CRAC) channel modulator] and stromal interacting molecule-1, may also contribute to SOC-mediated Ca2+ entry, mainly in nonexcitable cells (21, 64, 74). The proteins are, however, poorly expressed in native arterial myocytes, but are abundant in cultured ASMCs (6). Recent findings have suggested that Orai1 and stromal interacting molecule-1 are upregulated in aortas from stroke-prone spontaneously hypertensive rats (22).
Interestingly, several groups have reported that Ca2+ homeostasis in ASMCs is influenced not only by direct Ca2+ entry through TRPC channels but also by Na+ entry through these nonselective cation channels. The entering Na+ apparently then also promotes Ca2+ entry through nearby NCX (2, 19, 62, 85).
In view of this profound influence of TRPC channels on ASMC function, it is hardly surprising that these channels have also been implicated in the pathogenesis of various forms of hypertension. For example, TRPC1 and TRPC6 are reportedly involved in hypoxic pulmonary hypertension (46, 76), and TRPC6 upregulation has been implicated in idiopathic pulmonary arterial hypertension (46, 81). Also, in spontaneously hypertensive rats, TRPC3 is upregulated in the vasculature (47).
An important role for circulating ouabain in the pathogenesis of hypertension has now been substantiated (11, 52, 75). Nevertheless, the downstream consequences of ouabain's interaction with Na+ pumps, leading to the elevation of BP, are still not completely understood. Here, we explore the mechanisms that underlie altered Ca2+ homeostasis in mesenteric arteries from OH and control [normotensive (NT)] rats. Our data demonstrate that several key proteins in the pathway from ouabain to augmented Ca2+ signaling, namely, the α2-subunit of Na+ pumps, NCX type 1 (NCX1), TRPC1, and TRPC6, are all upregulated in OH rats. These results also provide further, direct verification of several critical elements in the specific pathway linking salt to hypertension, which was postulated more than three decades ago (8).
All experiments were carried out according to the guidelines of the Institutional Animal Care and Use Committee of the University of Maryland School of Medicine.
Ouabain-induced hypertension in the rat.
Male Sprague-Dawley rats (130–150 g, Charles River Laboratories, Frederick, MD) were maintained in a temperature- and humidity-controlled room with a 12:12-h light-dark cycle. Rats had free access to tap water and were fed standard rat chow ad libitum. Body weight was measured weekly. After rats had acclimated to the facility and been conditioned to BP measurements, baseline data were obtained over several weeks. Under halothane anesthesia, NT male rats received a subcutaneous controlled time-release pellet (Innovative Research of America, Sarasota, FL) containing either ouabain (1.5 mg/pellet) or vehicle (control). The ouabain pellets were designed to release ∼25 μg ouabain/24 h for 60 days. Systolic and mean BPs were recorded weekly by tail-cuff plethysmography using a commercial photoelectric system (model 29 BP Meter/Amplifier, IITC, Woodland Hill, CA) and a device providing constant rates of cuff inflation and deflation. The average values for BP in each rat were obtained typically from five sequential cuff inflation-deflation cycles.
Dissection of arteries for immunoblot analysis.
The aorta and superior mesenteric artery from a euthanized rat were rapidly removed and placed in ice-cold physiological salt solution 1 (PSS1) of the following composition (in mM): 140 NaCl, 5.36 KCl, 0.34 Na2HPO4, 0.44 K2HPO4, 10 HEPES, 1.2 MgCl2, 1.8 mM CaCl2, and 10 d-glucose (pH 7.2). The arteries were cleaned of fat and connective tissue, frozen in liquid nitrogen, and stored at −80°C before protein extraction. Mesenteric arteries or aortas from two to four rats were pooled and homogenized in lysis buffer.
Freshly dissociated ASMCs for Ca2+ imaging.
Myocytes were isolated from rat mesenteric arteries as previously described (6). The superior mesenteric artery was cleaned of fat and connective tissue and digested in low-Ca2+ (0.05 mM) PSS1 containing 2 mg/ml collagenase type XI, 0.16 mg/ml elastase type IV, and 2 mg/ml BSA (fat free) for 35 min at 37°C. After digestion, the tissue was washed three times with low-Ca2+ PSS1 at 4°C. A suspension of single cells was obtained by gently triturating the tissue with a fire-polished Pasteur pipette in low-Ca2+ PSS1. Smooth muscle cells were differentiated by their characteristic elongated morphology. Dispersed cells were directly deposited on glass coverslips for fluorescence microscopy. ASMCs on coverslips were stored at 4°C and used within 4 h. Cells were allowed to settle on the coverslips for 20–30 min before being loaded with fura-2. Freshly dissociated cells that were markedly contracted under resting conditions (<5%) were excluded. At the conclusion of the Ca2+ imaging experiments, the same cells were labeled for smooth muscle α-actin to identify ASMCs (26). In these experiments, nuclei also were identified by labeling for 5 min with a 50 μM solution of 4′,6′-diamidino-2-phenylindole (24).
Primary cultured ASMCs for Ca2+ imaging and Western blot analysis.
The methods used for the isolation and culture of rat ASMCs have been previously published (26). Briefly, the superior mesenteric artery was isolated under sterile conditions from euthanized male 9- to 10-wk-old Sprague-Dawley rats as described above. The artery was incubated for ∼45 min at 37°C in Ca2+- and Mg2+-free HBSS containing 1 mg/ml collagenase type 2. After the incubation, the adventitia was carefully stripped, and the endothelium was removed (26). ASMCs in the remaining smooth muscle were dissociated by digestion for 35–40 min at 37°C in HBSS containing 1 mg/ml collagenase type 2 and 0.5 mg/ml elastase type IV. Dissociated cells were resuspended and plated on either 25-mm coverslips for use in fluorescent microscopy experiments or on 10-cm culture dishes for Western blot analysis. Plated cells were maintained in DMEM supplemented with 10% FBS under a humidified atmosphere of 5% CO2-95% air at 37°C. The medium was changed on days 4 and 7. Experiments were performed on subconfluent cultures on days 7 and 8 in vitro if not indicated otherwise. The purity of ASMC cultures was verified by positive staining with smooth muscle-specific α-actin (24, 26). Most of the cells (>99.5%) were α-actin positive. The cells also did not cross react with fibroblast (CD90/Thy-1)-specific and endothelium (factor VIII, von Willebrand)-specific antigens (26).
[Ca2+]cyt was measured with fura-2 using digital imaging (6). Primary cultured ASMCs were loaded with fura-2 by an incubation for 35 min in culture medium containing 3.3 μM fura-2 AM (20–22°C, 5% CO2-95% O2). After the dye loading, coverslips were transferred to a tissue chamber mounted on a microscope stage, where the cells were superfused for 15–20 min (35–36°C) with physiological salt solution 2 (PSS2) to wash away the extracellular dye. PSS2 contained (in mM) 140 NaCl, 5.9 KCl, 1.2 NaH2PO4, 5 NaHCO3, 1.4 MgCl2, 1.8 CaCl2, 11.5 glucose, and 10 HEPES (pH 7.4). Cells were studied for 40–60 min during continuous superfusion with PSS2 (35°C).
The imaging system included a Zeiss Axiovert 100 microscope (Carl Zeiss, Thornwood, NY). Dye-loaded cells were illuminated with a diffraction grating-based system (Polychrome V, TILL Photonics). Fluorescent images were recorded with a CoolSnap HQ2 charge-coupled device camera (Photometrics, Tucson, AZ). Image acquisition and analysis were performed with a MetaFluor/MetaMorph Imaging System (Molecular Devices, Downingtown, PA). [Ca2+]cyt was calculated by determining the ratio of fura-2 fluorescent emission (510 nm) excited at 380 and 360 nm, as previously described (6, 24). Intracellular fura-2 was calibrated in situ in freshly dissociated and primary cultured ASMCs (6). Intracellular Ba2+ measurements are shown as fura-2 340-to-380-nm excitation ratios with fluorescent emission at 510 nm (6).
Membrane proteins were solubilized in SDS buffer containing 5% 2-mercaptoethanol and were separated by SDS-PAGE as previously described (6). The following antibodies were used: rabbit polyclonal anti-TRPC1, anti-TRPC3, anti-TRPC4, anti-TRPC5, and anti-TRPC6 (Allomone Laboratories, Jerusalem, Israel); mouse monoclonal anti-NCX1 (R3F1, Swant, Bellinzona, Switzerland); rabbit polyclonal anti-Na+ pump α1-subunit isoform (gift of Dr. Thomas Pressley); and rabbit polyclonal anti-Na+ pump α2-subunit isoform (Millipore, Billerica, MA). Gel loading was controlled with polyclonal or monoclonal anti-β-actin antibodies (Sigma-Aldrich, St. Louis, MO). After being washed, membranes were incubated with anti-rabbit horseradish peroxidase-conjugated IgG for 1 h at room temperature. The immune complexes on the membranes were detected by enhanced chemiluminescence plus (Amersham BioSciences, Piscattaway, NJ) and exposure to X-ray film (Eastman Kodak, Rochester, NY). Quantitative analysis of immunoblots was performed using a Kodak DC120 digital camera and 1D Image Analysis Software (Eastman Kodak).
Protein knockdown by short interfering RNA.
Primary cultured ASMCs were transfected with silencer (si)RNA ON-Target plus Smart pool (20 μM) designed against NCX1, TRPC6, or siCONTRL (Dharmacon, Lafayette, CO). The sequences of the TRPC6/siRNA duplexes were as follows: 5′-UUGUGUGCCAGCUGAUUUCUU-3′, 5′-UAAACAUGUAUGCUGGUCCUU-3′, 5′-AAUCCGUACAUAACCUUUAUU-3′, and 5′-AAUGGCGACAGCGAGGACCUU-3′; the sequences of the NCX1/siRNA duplexes were as follows: 5′-CCGAUUCCCUCUACCGUAA-3′, 5′-CAAUAUCAGUCAAGGUAAU-3′, 5′-GGAGAGAGCAGUUCAUUGAU-3′, and 5′-GAAUGUACUGGCUCAUAUU-3′. Twenty-four hours before treatment, ASMCs were placed in the culture medium without antibiotics and further transfected with siRNA using Lipofectamine 2000 reagent in Opti-MEM (Invitrogen). After 24 h of incubation, the medium was aspirated and replaced with DMEM (10% FBS) without siRNA for 48 h before Western blot analysis was performed.
FBS was obtained from Atlanta Biologicals (Lawrenceville, GA). All other tissue culture reagents were obtained from GIBCO-BRL (Grand Island, NY). Fura-2 AM and 4′,6′-diamidino-2-phenylindole were obtained from Molecular Probes (Invitrogen Detection Technologies, Eugene, OR). 1-Oleoyl-2-acetyl-sn-glycerol (OAG) was purchased from Calbiochem (San Diego, CA). Collagenase type 2 was obtained from Worthington Biochemical (Freehold, NJ). Cyclopiazonic acid (CPA), caffeine (CAF), DMSO, β-actin, smooth muscle α-actin, collagenase type XI, elastase type IV, BSA, nifedipine, ionomycin, penicillin G, and streptomycin were purchased from Sigma-Aldrich. All other reagents were analytic grade or the highest purity available.
The numerical data presented in the results are means ± SE from n single cells (one value per cell). Immunoblots were repeated at least four to six times for each protein. The number of animals is presented where appropriate. Data from 6 to 18 rats were obtained for most protocols. Data from five to six transfections were obtained for siRNA protocols. Statistical significance was determined using Student's paired or unpaired t-tests or two-way ANOVA as appropriate. P values of <0.05 were considered significant.
Ca2+ homeostasis in freshly dissociated ASMCs from OH rats.
Sustained hypertension in rats was induced by prolonged in vivo ouabain treatment (Fig. 1A). Baseline BPs were stable in rats over 4 wk of observation. At week 4, all rats received a slow-release implant containing either ouabain or vehicle (control). By week 6 (i.e., 2 wk after the implant), systolic BP increased in the ouabain-implanted (OH) group and remained elevated until death at week 9 (147 ± 3 mmHg in the OH group vs. 118 ± 3 mmHg in NT group, P < 0.05).
Freshly dissociated OH rat mesenteric artery myocytes had significantly higher resting [Ca2+]cyt than did NT myocytes (112 ± 2 vs. 95 ± 3 nM, P < 0.001; Fig. 1, B–D). Furthermore, activation by 1 μM phenylephrine, in physiological media, induced augmented Ca2+ signals in OH rat ASMCs. Both the peak initial response, believed to be the result of inositol (1,4,5)-trisphosphate-mediated SR Ca2+ release, and the later, more sustained Ca2+ signal, perhaps mediated by Ca2+ entry through ROCs and/or SOCs (59), were greater in OH than NT rat ASMCs (Fig. 1, B–D). In subsequent experiments, we investigated the possible mechanisms that contributed to this altered Ca2+ homeostasis and augmented signaling.
Na+ pumps and Na+/Ca2+ exchangers in OH rat artery smooth muscle.
Previously published evidence has suggested that Na+ pumps α2-subunits and NCX1 mediate the effects of low-dose ouabain on smooth muscle Ca2+ signaling and vasoconstriction (18, 84). But what happens to these transporters under the influence of chronic in vivo ouabain administration? As shown in Fig. 2, A and B, the response to tonic low-dose ouabain, which inhibits arterial Na+ pump α2-subunits with an apparent half-maximal effect (EC50) at 660 pM (83), is a large increase in the expression of the target Na+ pump α2-subunits: to ∼250% of the level in NT mesenteric arteries. In contrast, the expression of Na+ pump α1-subunits, which constitute ∼80% of the total Na+ pump population in ASMCs (18, 84) and which have very low affinity for ouabain (57), was not altered in OH arteries (Fig. 2, C and D).
NCX1, presumably the next protein in the ouabain-Na+ pump α2-subunit-Ca2+ signaling pathway (11), is, like the α2-subunit, markedly upregulated (by ∼5- to 6-fold) in the OH rat mesenteric artery and aorta (Fig. 2, E–H). Upregulation of the α2-subunit would be expected to counteract the inhibitory effect of ouabain on Na+ pump α2-subunits and reduce the buildup of the Na+ concentration in the tiny spaces between the plasma membrance and junctional SR (3). Nevertheless, because NCX1 normally mediates Ca2+ entry, rather than exit, in ASMCs of arteries with tone (34), upregulation of this transporter should tend to accelerate Ca2+ entry and promote net Ca2+ gain.
SR Ca2+ stores and SOC-mediated Ca2+ entry in OH rat artery smooth muscle.
If Ca2+ entry is increased and [Ca2+]cyt is elevated in OH rat ASMCs, it is logical to ask whether ASMC SR Ca2+ stores are also increased. To address this question, Ca2+ transients were evoked by unloading the SR Ca2+ stores with either 10 mM CAF (which opens SR ryanodine-sensitive Ca2+ channels) or 10 μM CPA [which inhibits sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA)] (27). Ca2+-free media were used to avoid complications from Ca2+ entry. While the CPA-evoked Ca2+ transients were only marginally (and not significantly) increased in OH arterial myocytes (Fig. 3, A and B), CAF-evoked transients were double the amplitude of those in NT artery myocytes (Fig. 3, C and D). This difference between CPA- and CAF-evoked responses is not surprising in view of our evidence that 1) rat mesenteric artery myocytes possess two functionally distinct types of SR Ca2+ stores (6) and 2) the CPA-sensitive store is quite small in freshly dissociated ASMCs (6).
The Ca2+ transients evoked by restoration of external Ca2+ are often used to detect the activity of SOCs, which apparently play an important role in SR store refilling (60). In both CAF- and CPA-treated cells, these Ca2+ transients, a manifestation of store-operated Ca2+ entry, were significantly greater in OH than NT myocytes (Fig. 3, A–D). The implication is that SOC activity is augmented in OH myocytes. Whether this was simply due to increased entry through an unchanged number of channels or to an increase in the number of channels available was tested by immunoblot analysis. The results revealed that TRPC1 expression in OH mesenteric arteries was 2.8-fold that in NT arteries (Fig. 3, E and F). In contrast, the expression of TRPC5, also a component of SOCs, was unaltered (Fig. 3, G and H). TRPC4, a protein component of SOCs in other cell types, is not expressed in rat mesenteric arteries (6). TRPC4 is, however, expressed in the rat aorta, but its expression level was unaltered in OH rats (Fig. 3, I and J).
ROCs and ROC-mediated Ca2+ entry in OH rat artery smooth muscle.
To determine whether ROC-mediated Ca2+ entry also is increased in mesenteric ASMCs from OH rats, freshly dissociated myocytes were stimulated with the cell-permeable diacylglycerol analog OAG. OAG opens TRPC3 and TRPC6 channels in a PKC-independent manner (31). SOCs have high Ca2+ selectivity and, unlike ROCs, are virtually impermeable to other alkaline-earth cations, such as Ba2+ (73). Therefore, to distinguish ROCs from SOCs, we measured Ba2+ entry. Ba2+ is not transported by SERCA or PM Ca2+ pumps (44). In the presence of extracellular Ba2+, 80 μM OAG induced significantly larger elevations of cytosolic Ba2+ (fura-2 340-to-380-nm ratio; see methods) in myocytes from OH rats than in those from NT rats (Fig. 4, A and B). This evidence of greatly increased ROC-mediated Ca2+ entry correlated with more than sixfold augmentation of the expression of TRPC6 (Fig. 4, C and D), an obligatory component of endogenous ROCs in a variety of cell types, including vascular myocytes (31, 33). The expression of TRPC3, which also belongs to the TRPC3/6/7 subfamily of diacylglycerol-activated ROCs, however, was not significantly affected in ASMCs from OH rats (Fig. 4, E and F); mRNA for TRPC7 was not detected in rat mesenteric artery smooth muscle (30) and, therefore, was not studied.
In vitro replication of the actions of in vivo ouabain in primary cultured ASMCs.
The aforementioned effects of in vivo ouabain administration on NCX1, SOCs, and ROCs in arterial myocytes might be the direct result of ouabain's interaction with Na+ pump α2-subunits or, alternatively perhaps, the consequence of the elevated BP or some other in vivo factor(s). To explore this issue, we tested the effects of prolonged treatment with nanomolar ouabain on primary cultured ASMCs from NT rats. Indeed, after a 72-h exposure to 100 nM ouabain, NCX1 expression in ASMCs was threefold higher than in untreated cells (Fig. 5, A and B).
In vitro exposure of cultured NT rat mesenteric ASMCs to 100 nM ouabain for 72 h modestly, but significantly, elevated resting [Ca2+]cyt and augmented the SOC-mediated Ca2+ entry evoked by depleting SR Ca2+ stores with CPA (Fig. 5, C and D). This was associated with a small, but not significant, increased in TRPC1 expression (Fig. 5, E and F). Neither TRPC4 nor TRPC5 expression was affected by the ouabain treatment (not shown).
Under these same conditions, however, ouabain greatly augmented OAG-activated ROC-mediated Ca2+ entry in the cultured myocytes. In fact, 10 nM ouabain had an effect equal to that of 100 nM ouabain (Fig. 6, A–C), as expected, because the EC50 for ouabain is ∼0.66 nM (83). This increase in ROC-mediated Ca2+ entry correlated with a large, time-dependent increase in TRPC6 expression, so that the TRPC6 level nearly doubled by 72 h (Fig. 6, D and E).
Are NCX1 expression and TRPC6 expression interrelated?
The preceding section demonstrates that many of the effects of in vivo ouabain administration on arterial smooth muscle can be mimicked in vitro, in cultured cells. This may enable us to elucidate the possible underlying mechanisms. As a first step, siRNA transfection methods were used to determine whether suppression of either NCX1 or TRPC6 affected the expression of the other protein. As anticipated, NCX1 siRNA markedly downregulated NCX1 expression and abolished the effect of 100 nM ouabain on NCX1 expression (Fig. 7, A and B). But NCX1 siRNA also significantly reduced TRPC6 expression (by ∼25%) and abolished its augmentation by ouabain (Fig. 7, C and D) but did not affect α2-subunit expression (not shown). ROC-mediated Ba2+ entry was significantly decreased under these conditions (Fig. 7, E and F).
When TRPC6 expression was knocked down by TRPC6 siRNA in cultured ASMCs (Fig. 8, A and B), ROCE (i.e., ROC-mediated Ba2+ entry) was, of course, also greatly reduced (Fig. 8, C and D). Indeed, the OAG-induced increase in the fura-2 fluorescence ratio, indicative of Ba2+ entry through ROCs, declined by ∼90% (Fig. 8C). Furthermore, although NCX1 expression in the absence of ouabain was unaffected, the ouabain-induced increase in NCX1 expression was markedly suppressed by TRPC6 siRNA (Fig. 8, E and F). This raises the possibility that either TRPC6 expression, per se, or the resulting increase in resting [Ca2+]cyt and/or ROC-mediated Ca2+ entry is required to enable ouabain to augment NCX1 expression.
Selective inhibition of TRPC6 also did not affect Na+ pump α2-subunit expression (not shown).
The results described in this report show the effects of ouabain in vivo and in vitro on the expression of membrane proteins that affect arterial myocyte Na+ and Ca2+ metabolism. The results, taken together, provide strong experimental support for the proposed sequence of steps leading from prolonged ouabain administration to augmented Ca2+ signaling, as shown in Fig. 9, and, therefore, to augmented vascular contractile responses. This complements evidence for the early steps obtained from molecular engineering studies in mice (summarized in Ref. 11).
Na+ pump α2-subunits and NCX1 are upregulated in OH rat arterial myocytes.
The first step is the direct interaction between ouabain and the ASMC high-ouabain affinity Na+ pump α2-subunits. Nanomolar ouabain inhibits Na+ pump α2-subunits with an EC50 of 0.5–1 nM (83, 84), but such low concentrations have no effect on the other Na+ pumps (α1-subunits) present in rat arterial myocytes [EC50 > 10 μM (57)]. Thus, it is not surprising that the myocytes respond by selectively upregulating Na+ pump α2-subunits (Fig. 2, A and B) to try to compensate for the ouabain-induced reduction in pump activity. This parallels a recent report (78) showing that the chronic inhibition of NCX in cardiac myocytes induces a compensatory upregulation of NCX1 in those cells.
The data from the OH rat myocytes demonstrate that the Na+ pump α2-subunits are the specific target of the circulating ouabain [in humans, the range of plasma levels is between ∼0.1 and 1 nM (50)]. These results complement studies on mice which show that the high-ouabain affinity binding site on the Na+ pump α2-subunit is essential for ouabain-induced and ACTH-induced forms of hypertension (17, 18, 48).
In arterial myocytes, Na+ pump α2-subunits colocalize with NCX1 in PM microdomains at PM-SR junctions (35, 67, 68). Inhibition of Na+ pump α2-subunits would therefore be expected to increase the cytosolic free Na+ concentration in the local sub-PM spaces and reduce the Na+ electrochemical gradient, thereby driving Ca2+ into the cells via NCX1. Indeed, the expression of these exchangers is also markedly elevated in OH rat ASMCs (Fig. 2, E–H), although the mechanism(s) by which this occurs is(are) not yet understood. Nevertheless, it is noteworthy that this simultaneous upregulation of Na+ pump α2-subunits and NCX1 in ASMCs differs strikingly from observations in hypertensive hearts. In both renovascular hypertension (49) and mineralocorticoid-salt hypertension (63), Na+ pump α2-subunits are upregulated but NCX1 is downregulated in cardiac myocytes. Whether or how this difference between cardiac and vascular smooth muscles in the hypertensive models might relate to the apparently different roles of NCX1 in the heart [primarily mediates Ca2+ extrusion (7, 15)] and in ASMCs [primarily mediates Ca2+ entry (34, 41)] is not clear.
It is noteworthy that mineralocorticoid hypertension in rats and humans is associated with elevated plasma endogenous ouabain (29, 43, 65). This begs the following question: “Is the pathway shown in Fig. 9 upregulated in the arteries of humans and animals with mineralocorticoid-salt hypertension?” Available evidence (cited in the next section) has indicated that the answer may be affirmative in rodent models.
Ca2+ homeostasis is altered in OH rat arterial myocytes.
Arterial contractility is augmented in many forms of hypertension, including OH. Some of the functional changes have been attributed to vascular structural remodeling and artery narrowing (53, 55) and increased arterial stiffness (13). Nevertheless, substantial evidence has also pointed to dynamic increases in contraction that may be associated directly with Ca2+ dysregulation, such as enhanced responses to stretch and to vasoconstrictors (12, 83).
Based on this evidence for arterial functional change, we studied Ca2+ homeostasis in OH rat artery myocytes. Indeed, freshly dissociated myocytes exhibit Ca2+ dysregulation: elevated resting [Ca2+]cyt and augmented CAF-releasable Ca2+ stores and phenylephrine-evoked Ca2+ signals (Figs. 1, B–D, and 3, C and D). The myocytes also exhibited increased SOC- and ROC-mediated Ca2+ entry (Figs. 3, A–D, and 4, A and B). Immunoblots indicated that the latter effects are a consequence of upregulated TRPC1 and TRPC6 expression, respectively (Figs. 3, E and F, and 4, C and D). The upregulated TRPC6 and increased ROC-mediated Ca2+ entry are, in fact, reflected by the augmented phenylephrine-evoked vasoconstriction of intact OH rat arteries, especially at high phenylephrine concentrations (83), where the role of ROCs is more prominent (59). The increased basal [Ca2+]cyt, as well as the increased TRPC6 (16, 32, 77), may contribute to the augmented myogenic reactivity of OH rat arteries (83). Indeed, mineralocorticoid-salt hypertension (4), in which endogenous ouabain has been implicated (29, 43), and human idiopathic pulmonary hypertension (81) are associated with upregulation of TRPC6. Thus, the biochemistry and physiology appear to be directly linked. In other words, upregulation of the pathway shown in Fig. 9 may provide a molecular explanation for the observed augmentation of vascular contractile responses in hypertension (12, 83). This concept is underscored by the evidence that, in human primary pulmonary hypertension as well, NCX1 (85) and TRPC6 (81) are both upregulated.
A noteworthy aspect of this comparison between arterial function and these molecular underpinnings is the relatively modest alterations in the function of small arteries (83) from some of the same OH rats that were used for the cellular and biochemical studies described above. Here, we report a relatively robust enhancement of NCX1, TRPC6, and even TRPC1 expression. This contrasts with the seemingly modest augmentation of myogenic reactivity and phenylephrine-evoked responses and the apparently insignificant augmentation (perhaps due to the low number of arteries studied) of contractions evoked by the release of SR Ca2+ stores in intact small arteries (83). How can these ostensibly disparate observations be reconciled? The answer likely lies in the fact that, in intact arteries, numerous other factors also influence the [Ca2+]cyt level and contractility. Nevertheless, consideration of Poiseuille's equation (45) reminds us that even a modest 5% augmentation of constriction in a 100-μm-diameter artery will increase the resistance to flow by >20%–a large effect when translated into total peripheral resistance and blood pressure.
Treatment of cultured ASMCs with nanomolar ouabain mimics observations on ASMCs from OH rats.
Another key observation made in this study is that, when primary cultured normal rat mesenteric ASMCs are exposed to low-dose ouabain, in vitro, they, too, upregulate components of the pathway shown in Fig. 9. Specifically, they increase NCX1 and TRPC6 expression and elevate resting [Ca2+]cyt. The TRPC6 overexpression is reflected, functionally, by the significantly augmented ROC-mediated Ba2+ entry, similar to observations in freshly dissociated OH rat arterial myocytes. Although TRPC1 was marginally, but not significantly, upregulated in the cultured cells, SOC-mediated Ca2+ entry did modestly increase (Fig. 5, C–F). These effects on cultured myocytes show that the downstream effects are triggered by ouabain itself and not by the increase in BP or by other circulating factors. It is noteworthy that, during cell culture, the expression of TRPC channels and SOC-mediated Ca2+ entry correlated with [Ca2+]cyt (24, 28). Therefore, the increased basal [Ca2+]cyt in ouabain-treated ASMCs may contribute to the upregulation of TRPCs. The effect of ouabain treatment was, however, specific only for TRPC6; other TRPC proteins were not changed under these conditions.
Importantly, this in vitro replication of the in vivo effects of ouabain presents the opportunity to explore the mechanisms that underlie these effects of ouabain on myocytes. Accordingly, to determine if the upregulated TRPC6 is a consequence of the augmented expression of NCX1, we tested the effect of NCX1/siRNA on TRPC6 expression. The results (Fig. 7) demonstrated that marked knockdown of NCX1 not only abolished the TRPC6 response to ouabain but even reduced basal TRPC6 expression by about one-third. Conversely, when TRPC6 expression was downregulated by TRPC6/siRNA, NCX1 expression was no longer influenced by ouabain (Fig. 8). These results imply a mutual interaction between NCX1 and TRPC6 and, thus, between the regulation of Ca2+ homeostasis by NCX1 and regulation of Ca2+ (and Na+) entry through ROCs, which are nonselective cation channels (1, 16, 33).
These considerations bear on the specific sequence of steps shown in Fig. 9. It seems clear that prolonged ouabain administration acting via Na+ pump α2-subunits initiates the upregulation of NCX1 and TRPC6 (and TRPC1). The siRNA data support the view that NCX1 lies upstream of TRPC6. The situation is complex because TRPC6 channels are relatively nonselective Ca2+-permeable channels and mediate substantial Na+ entry (1, 16, 62). The influx of Na+ might depolarize the cells and activate L-type Ca2+ channels in addition to promoting Ca2+ entry via NCX1 (2, 19, 62). This downstream role of NCX1 might explain why knockout of TRPC6 suppresses the effect of ouabain on NCX1 (Fig. 8).
The aforementioned effects should not be confused with the rapid (within seconds) increase in vascular tone induced by the short-term application of low-dose ouabain to arteries with myogenic tone (83, 84). The latter vasoconstrictor effect is apparently mediated by acute Na+ pump α2-subunit inhibition and the consequent rapid Na+ gradient-driven entry of Ca2+ via NCX. These acute effects clearly do not depend on the amplification of the Ca2+ signaling pathway through protein synthesis, which appears to be the response to prolonged Na+ pump α2-subunit inhibition.
In summary, this report describes several novel observations on freshly dissociated myocytes from OH rat arteries and on the effects of low-dose ouabain on normal rat cultured arterial myocytes. These findings bear on the functional changes observed in isolated, intact OH rat arteries and, perhaps more generally, on the function of arteries in various types of hypertension. First, we showed that several molecular entities involved in Na+ and Ca2+ homeostasis and signaling (Na+ pump α2-subunits, NCX1, TRPC1, and TRPC6) are upregulated in OH rat mesenteric arteries. This is consistent with accumulating evidence linking high-ouabain affinity Na+ pumps, NCX1, SOCs, and ROCs to local control, at PM-SR junctions, of Na+ and Ca2+ homeostasis and Ca2+ signaling (3, 9, 10, 25, 36, 72).
Second, we found that the molecular sequence of events triggered by prolonged infusion of ouabain in vivo (Fig. 9) is mimicked by a 72-h exposure of ASMCs to ouabain in vitro. This makes it convenient to explore, with molecular tools, the specific mechanisms that underlie ouabain's augmentation of Ca2+ signaling. In the first such study, we demonstrated a mutual interaction between NCX1 and TRPC6 (and ROC-mediated Ca2+ entry). Considerable evidence has indicated that circulating endogenous ouabain is increased in several forms of salt-dependent hypertension (29, 39, 43) and in as many as half of all patients with essential hypertension (61, 65). Thus, our new findings may have broad implications for elucidating the pathogenesis of the elevated BP in these conditions.
This work was supported by National Institutes of Health Grants NS-048263 (to V. A. Golovina) and P01-HL-078870 Project 1 (to M. P. Blaustein) and Project 2 (to V. A. Golovina) and Core B (to J. M. Hamlyn) and by funds from the University of Maryland School of Medicine.
No conflicts of interest are declared by the author(s).
The authors thank Dr. W. Gil Wier for comments on the manuscript and Katherine Frankel for assistance with the preparation of the manuscript.
Present address of M. V. Pulina: Dept. of Medicine, Div. of Cardiology, Weill Cornell Medical College, New York, NY 10021.
Present address of R. Berra-Romani: School of Medicine, Benemèrita Universidad Autònoma de Puebla, 72000 Puebla, México.
- Copyright © 2010 the American Physiological Society