ACTH-induced hypertension is dependent on the ouabain-binding site of the α₂-Na⁺-K⁺-ATPase subunit

Abstract

ACTH-induced-hypertension is commonly employed as a model of stress-related hypertension, and despite extensive investigation, the mechanisms underlying elevated blood pressure (BP) are not well understood. We have reported that ACTH treatment increases tail-cuff systolic pressure in wild-type mice but not in mutant mice expressing ouabain-resistant α₂-Na⁺-K⁺-ATPase subunits (α2^R/R mice). Since tail-cuff measurements involve restraint stress, the present study used telemetry to distinguish between an effect of ACTH on resting BP vs. an ACTH-enhanced stress response. We also sought to explore the mechanisms underlying ACTH-induced BP changes in mutant α2^R/R mice vs. wild-type mice (ouabain-sensitive α₂-Na⁺-K⁺-ATPase, α2^S/S mice). Baseline BP was not different between the two genotypes, but after 5 days of ACTH treatment, BP increased in α2^S/S (104.0 ± 2.6 to 117.7 ± 3.0 mmHg) but not in α2^R/R mice (108.2 ± 3.2 to 111.5 ± 4.0 mmHg). To test the hypothesis that ACTH hypertension is related to inhibition of α₂-Na⁺-K⁺-ATPase on vascular smooth muscle by endogenous cardiotonic steroids, we measured BP and regional blood flow. Results suggest a differential sensitivity of renal, mesenteric, and cerebral circulations to ACTH and that the response depends on the ouabain sensitivity of the α₂-Na⁺-K⁺-ATPase. Baseline cardiac performance was elevated in α2^S/S but not α2^R/R mice. Overall, the data establish that the α₂-Na⁺-K⁺-ATPase ouabain-binding site is of central importance in the development of ACTH-induced hypertension. The mechanism appears to be related to alterations in cardiac performance, and perhaps vascular tone in specific circulations, presumably caused by elevated levels of circulating cardiotonic steroids.

it is well-established that cardiac glycosides such as ouabain and digoxin exert their effects by directly inhibiting the activity of the Na⁺-K⁺-ATPase, and it is currently understood that these effects are largely mediated through a specific binding site on the first extracellular loop of the catalytic Na⁺-K⁺-ATPase α-subunit (15–17). The highly conserved nature of this so-called ouabain-binding site, as well as the discovery of endogenous circulating cardiotonic steroids, suggests that regulation of the Na⁺-K⁺-ATPase via this site may play an important physiological role, particularly in the cardiovascular system (1, 9, 12, 20). For example, it has been shown that endogenous digoxin-like factors are elevated in a variety of human and experimental forms of hypertension, including ACTH-induced hypertension (7, 14, 31). In support of this hypothesis, we have demonstrated that the ouabain-binding site on both the α₁- and α₂-Na⁺-K⁺-ATPase isoforms participates in the regulation of blood pressure and cardiac function (2–4).

There are four major α-isoforms of the Na⁺-K⁺-ATPase (αNKA), and in most species all four are extremely sensitive to ouabain (5). In mice and rats, however, the α₁-NKA has low affinity for, and is resistant to, the inhibitory effects of ouabain and other cardiac glycosides. To explore the physiological role of the ouabain-binding site, mice have been generated with targeted mutation of the α₂-NKA such that it, like the α₁-isoform, is also resistant to ouabain. Using these mice, we recently reported that ACTH treatment for 5 days significantly increased systolic blood pressure, measured by tail cuff, in the α₂ ouabain-sensitive wild-type mice (α2^S/S) but not in ouabain-resistant α₂-NKA mutant mice (α2^R/R). These data suggested that ACTH-induced cardiovascular effects are mediated through the actions of an endogenous substance acting on the α₂-NKA ouabain-binding site (6). Since the tail-cuff method and attendant restraint were employed in this study, there are two potential interpretations of these earlier findings: first, that the primary effect of ACTH treatment was to elevate resting blood pressure, and second, that ACTH only serves to augment the cardiovascular response to restraint stress. Indeed, this latter possibility that ACTH-induced hypertension is not a true hypertension has been previously argued in a study using rats (8). If this were the case, then perhaps ACTH-induced increases in the stress response would be related to the release of an endogenous cardiotonic steroid on the imposition of acute stress. Thus an important goal of this study was to distinguish between these two possible interpretations, and we hypothesized that ACTH-administration does not elevate resting blood pressure in α2^R/R mice.

ACTH-induced hypertension has been associated with alterations in a variety of cardiovascular variables including cardiac output, total peripheral vascular resistance, endothelial NO production, and sympathetic activity (26, 27, 29). For example, it has been shown in rats that angiotensin-converting enzyme inhibition prevented the ACTH-induced rise in renal vascular resistance and the development of hypertension (26). Moreover, in this same study, β-adrenergic blockade prevented ACTH-induced increases in cardiac output but only partially prevented the increase in blood pressure. It is therefore reasonable to hypothesize that ACTH-induced changes in blood pressure are related to an increase in vascular and cardiac muscle reactivity, which in turn may be dependent on the actions of an endogenous glycoside interacting with the α₂-NKA on vascular smooth muscle and, perhaps, on cardiac muscle. A second goal of the present study, then, was to explore the potential role of altered vascular and cardiac contractility in ACTH-induced hypertension in mice with a ouabain-sensitive vs. ouabain-resistant α₂-NKA.

METHODS

Animals.

Animals were obtained from an established colony at the University of Cincinnati. The development of mice expressing the ouabain-resistant α₂-isoform of the Na⁺-K⁺-ATPase by gene targeting was described previously (2). Homozygous mutant (α2^R/R) and wild-type (α2^S/S) offspring were generated from heterozygous matings, and all mice were maintained on a mixed 129SvJ and Black Swiss background. For all experiments, animals were paired for age, weight, and sex. Genotypes were determined by PCR analysis of DNA from tail biopsies, as described previously (2). All experimental protocols were approved by the Institutional Animal Care and Use Committee at the University of Cincinnati in accordance with established guidelines.

Telemetric measurement of blood pressure.

Continuous ambulatory blood pressure recordings were made in α2^S/S and α2^R/R mice as described previously using the TA11PA-C10 pressure transmitter (Data Sciences International, St. Paul, MN) (19). Transmitters were implanted under isoflurane anesthesia using carotid artery cannulation and subcutaneous transmitter placement. After implantation, mice were allowed to recover for at least 7 days before data collection. The mice were synchronized to a 12:12-h light-dark schedule with lights on at 7:00 AM. After 3–5 days of baseline recording, mice were administered saline vehicle or ACTH every 8 h by subcutaneous injection for 5 days (0.5 μg/g body wt of fragment 1–24; Sigma). Blood pressure was monitored continuously in 1-min episodes at 5-min intervals using a PowerLab system and Chart software (ADInstruments, Colorado Springs, CO). To evaluate the participation of the renin-angiotensin system (RAS) in maintaining blood pressure, we injected mice with losartan (10 μg/g body wt ip) and monitored blood pressure continuously for 60 min. This was performed once before administration of ACTH in each mouse and repeated after 5 days of ACTH treatment. On a different day before ACTH treatment, we also evaluated the contribution of sympathetic tone to blood pressure maintenance by injecting hexamethonium (40 μg/g body wt ip) to induce ganglionic blockade. To evaluate the specificity of hexamethonium for ganglionic blockade vs. neuromuscular blockade of the diaphragm, we administered escalating doses of hexamethonium with and without artificial ventilation and while monitoring blood pressure and ventilation with a spirometer (ADInstruments).

Regional blood flow and vascular resistance.

α2^S/S and α2^R/R mice, treated with vehicle or ACTH as described, were anesthetized with ketamine (50 μg/g body wt ip) and thiobutabarbital (Inactin; 100 μg/g boy wt ip) and instrumented as previously described (18). After tracheostomy and cannulation of the femoral artery and vein, mice were provided a maintenance infusion of 2.5% BSA in PBS at 0.15 μl·min⁻¹·g body wt⁻¹ to maintain euvolemia. For carotid flow measurements, the left common carotid artery, just proximal to the internal/external bifurcation, was isolated and fitted with a 0.5-PSB perivascular flow probe connected to a dual-channel TS420 flowmeter (Transonic Systems, Ithaca, NY). The probe window was filled with acoustical gel, and the probe body was embedded in 4% agar to secure its position. For renal and mesenteric blood flow, a flank incision was made and the renal and/or superior mesenteric artery was carefully isolated and fitted with a 0.5-PSB flow probe, which was held in position with a micromanipulator. Since flow could only be measured in two arteries at a time, the probe configuration was varied systematically from experiment to experiment (i.e., carotid and renal, carotid and mesenteric, renal and mesenteric). Continuous traces of regional vascular resistance were derived by dividing the mean arterial pressure signal by the mean blood flow signal. After completion of the surgical procedure and a 30-min recovery period, dose-dependent responses to phenylephrine (PE; 10, 40, and 80 ng·min·g body wt ⁻¹ iv) were tested. Individual doses were infused until a plateau was achieved (1–2 min), with a 3- to 5-min recovery period between doses.

Vascular smooth muscle reactivity.

Analyses of contractile properties of vascular smooth muscle were performed in intact thoracic aortas as previously described (11). Briefly, aortic rings from vehicle- and ACTH-treated α2^S/S and α2^R/R mice were gently dissected and mounted in a myograph chamber. After the experiment, the aortas were gently blotted and weighed, and dimensions were measured. Aortic wall thickness (t_w) was estimated from the equation t_w = blot weight/(1.05 × length × circumference), and the cross-sectional area (CSA) for force normalization was calculated as CSA = 2(t × length). The bath solution contained (in mmol/l) 118 NaCl, 4.73 KCl, 1.2 MgCl₂, 0.026 EDTA, 1.2 KH₂PO₄, 2.5 CaCl₂, and 5.5 glucose, buffered with 25 NaH₂CO₃; pH when bubbled with 95%O₂-5% CO₂ was 7.4 at 37°C. For force measurement, rings were mounted on 100-μm stainless steel wires and attached to a force transducer (South Natick, MA). Resting tension on each aorta was set to 25 mN, the estimated in vivo tension calculated for a pressure of 100 mmHg and a CSA of 0.42 mm². Before the start of the experiment, each aortic segment was challenged once with 50 mM KCl and three times with 1 μM PE to ensure reproducible forces. Cumulative PE concentration-isometric force relationships were generated for each aorta. EC₅₀ was determined using a logistic nonlinear curve-fitting routine (OriginLab, Northampton, MA).

Left ventricular function.

Left ventricular (LV) function was evaluated in closed-chest anesthetized mice as described previously (18). Since we have previously reported that LV function is nearly identical between vehicle-treated α2^S/S and α2^R/R mice (2), only ACTH-treated mice were evaluated in the current study. Mice were anesthetized with ketamine and Inactin as before and were instrumented with femoral artery and vein catheters and with a micromanometer (model SPR-671; Millar Instruments, Houston, TX) in the LV via the right carotid artery. Arterial and LV pressure and the change in LV pressure over time (dP/dt) signals were recorded at 2,000 Hz and analyzed using a PowerLab system. Hemodynamic measurements were taken at the basal state and after the administration of graded doses of dobutamine (1, 2, 4, 8, 16, and 32 ng·min⁻¹·g body wt⁻¹) to evaluate contractile reserve and β-adrenergic responsiveness. Maximum dP/dt (dP/dt_max) and dP/dt at 40 mmHg of developed pressure (dP/dt₄₀) were calculated from the first derivative of the pressure waveforms.

Adrenal cortical steroid levels.

α2^S/S and α2^R/R mice were treated for 5 days with either saline vehicle or ACTH as described above. After anesthesia was administered, blood was withdrawn by aortic puncture and centrifuged, and the serum was stored at −20°C for later analysis. Aldosterone was measured by enzyme immunoassay (EIA) using a kit from Diagnostic Systems Laboratories (Webster, TX), and corticosterone was measured by EIA using a kit from Immunodiagnostic Systems (Fountain Hills, AZ).

Data analysis.

Statistical analysis was performed by analysis of variance (ANOVA) using either a single-factor within-subjects design or a two- or three-factor mixed design with repeated measures on the last factor. Individual contrasts were used to compare group effects and interactions when needed, and Tukey's post hoc test was used to compare individual means where appropriate (SigmaStat version 3.5; Point Richmond, CA). Data are means ± SE, and differences were regarded as significant at P < 0.05.

RESULTS

Telemetric blood pressure.

Baseline blood pressure was found to be not different between α2^S/S and α^R/R mutant mice, as shown in Fig. 1. Both groups showed normal diurnal variations in blood pressure and heart rate (not shown) over the 3 days of control recording. In response to treatment with ACTH over a 5-day period, α2^S/S mice demonstrated a substantial and significant increase in blood pressure from a control level of 104.0 ± 2.6 mmHg to a final sustained level of 117.7 ± 3.0 mmHg (n = 8), as shown in Fig. 2, left. In contrast, blood pressure did not significantly change in response to ACTH in α2^R/R mutant mice (108.2 ± 3.2 to 111.5 ± 4.0, n = 6). Examination of daytime vs. nighttime blood pressure (Fig. 2, right) suggests that the ACTH -induced hypertension persisted during both sleep and active periods in wild-type mice. Although there appeared to be a transient mild elevation in blood pressure in α2^R/R mice during the light cycle, perhaps related to the stress of the three-times-daily injections, this effect did not persist and there was no significant elevation of pressure through days 4 and 5 of ACTH treatment. Before ACTH administration, administration of losartan resulted in equivalent decreases in blood pressure in α2^S/S (111 ± 2 to 96 ± 3 mmHg, n = 8) and α2^R/R mice (104 ± 4 to 90 ± 4 mmHg, n = 7; ANOVA: genotype effect, P = 0.11; treatment effect, P < 0.001; interaction, P = 0.82). After 5 days of ACTH treatment, the hypotensive response to losartan was not altered in either group (120 ± 5 to 107 ± 4 mmHg in α2^S/S, n = 6, vs. 101 ± 4 to 90 ± 4 mmHg in α2^R/R, n = 6; ANOVA: genotype effect, P < 0.001; treatment effect, P = 0.032; interaction, P = 0.79), suggesting that the systemic RAS was not an important contributor to the development of hypertension in the ACTH-treated wild-type mice. Hexamethonium injection (40 μg/g body wt) did not significantly alter blood pressure in the α2^S/S mice (114 ± 2 to 108 ± 4 mmHg) but, surprisingly, caused rapid death in α2^R/R mice. Since the specificity of hexamethonium for cholinergic receptors of autonomic ganglia vs. motor end plate is, in our experience, not particularly high in mice, we evaluated blood pressure and ventilatory responses in anesthetized mice. We found that administration of hexamethonium at doses >30 μg/g body wt in α2^R/R mice caused immediate but modest decreases in blood pressure (12 ± 2 mmHg, n = 3), followed by complete respiratory arrest, cardiovascular collapse, and death. Moreover, we found that artificial ventilation prevented the cardiovascular collapse and that mice tolerated doses of hexamethonium up to 80 μg/g body wt with only small decreases in blood pressure. When subsequently removed from the ventilator, mice were unable to resume spontaneous breathing.

Fig. 1.

Twenty-four-hour averages of mean arterial pressure (MAP) measured by telemetry for mice expressing ouabain-sensitive (α2^S/S wild type) and ouabain-resistant α₂-Na⁺-K⁺-ATPase subunits (α2^R/R mutant) during the control period 3 days before treatment with ACTH. Data represent the mean blood pressure from each 1-min episode, recorded at 5-min intervals.

Fig. 2.

Telemetric MAP measurements before and during administration of ACTH in α2^S/S and α2^R/R mice. Left: data are shown as 12-h means and are calculated as the mean pressure between 9:00 AM and 5:00 PM (light period) and 9:00 PM and 5:00 AM (dark period). Right: changes in (delta) MAP in response to ACTH during the light (top) and dark period (bottom). *P < 0.05 compared with control period within the same genotype. BW, body weight.

Regional vascular resistance.

We predicted that ACTH-induced hypertension would be associated with an increase in vascular reactivity due to the effects of elevated cardiotonic steroids interacting with the α₂-NKA in vascular smooth muscle. To test this, we evaluated blood pressure and flow responses to PE in α2^S/S and α2^R/R mice. Blood pressure responses are shown in Fig. 3. Consistent with telemetric data, blood pressure under basal conditions and at each PE dose was elevated in ACTH-treated α2^S/S mice (genotype effect: P = 0.05 compared with vehicle) but not in α2^R/R mice (genotype effect: P = 0.24 compared with vehicle). In addition, the data indicate that ACTH treatment did not significantly alter the overall sensitivity of blood pressure to PE challenge in α2^S/S mice (genotype × dose interaction: P = 0.19). By contrast, PE sensitivity was modestly elevated in ACTH-treated α2^R/R vs. α2^S/S mice (genotype × dose interaction: P = 0.04).

Fig. 3.

MAP values from anesthetized vehicle- and ACTH-treated α2^S/S and α2^R/R mice instrumented with perivascular flow probes on the renal, mesenteric, or carotid arteries. Values represent the average from a 20- to 30-s recording taken under resting conditions and at the peak of the response during graded infusions of phenylephrine. Inset shows the percent change in MAP relative to baseline to highlight interactions. ANOVA results: *P < 0.05, significant genotype effect for vehicle- vs. ACTH-treated α2^S/S mice. †P < 0.05, significant genotype × dose interaction for vehicle- vs. ACTH-treated α2^R/R mice. §P < 0.05, significant genotype × dose interaction for ACTH-treated α2^R/R vs. α2^S/S mice.

We further examined regional blood flow to the renal, mesenteric, and cerebral vasculatures to determine whether alterations in vascular tone and reactivity in specific resistance circuits might be altered. As shown in Fig. 4, resting blood flow was not different in any of the vascular beds in vehicle-treated mice. Interestingly, the response of resting blood flow to ACTH treatment was not consistent between the vasculatures. In α2^S/S mice, ACTH treatment increased renal blood flow (P = 0.05), decreased mesenteric blood flow (P = 0.005), and (marginally) had no effect on cerebral blood flow (P = 0.09). In α2^R/R mice, ACTH treatment decreased renal blood flow (P = 0.001) and mesenteric blood flow (P = 0.001) but had no effect on cerebral blood flow (P = 0.58). Thus it appears that the failure of ACTH to elevate blood pressure in α2^R/R mice is not specifically related to higher regional blood flows. Calculations of regional vascular resistance in response to PE challenge support this view and are shown in Fig. 5. Renal vascular reactivity was not different between the two genotypes under baseline conditions and was substantially increased by ACTH in the α2^R/R but not in the α2^S/S mice (Fig. 5, top). A similar pattern of change was observed in the mesenteric vasculature (Fig. 5, middle). In the cerebral vasculature, whereas responsiveness to PE challenge was not different between the genotypes (as evidenced by a nonsignificant genotype × dose interaction, P = 0.48), vascular resistance was elevated in the ACTH-treated α2^S/S mice throughout the dose-response range (Fig. 5, bottom, P = 0.009), consistent with an elevated blood pressure and unchanged carotid blood flow. No such difference was observed in the α2^R/R mice (P = 0.12).

Fig. 4.

Renal, mesenteric, and carotid blood flow under resting conditions in ACTH- and vehicle-treated α2^S/S and α2^R/R mice. In each experiment, flow probes were placed on 2 of the 3 arteries, and values reflect subsets of the data shown in Fig. 3. *P < 0.05 compared with vehicle-treated mice in the same group. †P < 0.05 compared with α2^S/S mice with the same treatment.

Fig. 5.

Regional vascular resistance (VR) from anesthetized mice instrumented with perivascular flow probes on the renal (top), mesenteric (middle), or carotid arteries (bottom). In each experiment, flow probes were placed on 2 of the 3 arteries, and values reflect subsets of the data shown in Fig. 3. ANOVA results: *P < 0.05, significant genotype × dose interaction for α2^S/S vs. α2^R/R mice with the same treatment. †P < 0.05, significant genotype × dose interaction for ACTH vs. vehicle within the same genotype. §P < 0.05, significant genotype effect for ACTH α2^S/S vs. vehicle α2^S/S mice.

Vascular smooth muscle reactivity.

To provide some independent measure of support for this in vivo vascular reactivity data, we evaluated contractile responses to phenylephrine in isolated aortas taken from vehicle- and ACTH-treated wild-type and mutant mice. Maximal force generation was not increased by ACTH treatment in either α2^S/S or α2^R/R aortas (Fig. 6). In fact, force generation at the highest doses of PE was significantly lower in aortas from ACTH-treated α2^S/S mice that were hypertensive. ACTH treatment had no significant effect on maximal force generation in aortas from α2^R/R mice. ACTH treatment also slightly decreased the sensitivity of the aortas to PE in α2^S/S but not α2^R/R mice: in α2^S/S mice, the EC₅₀ for PE was 0.26 ± 0.04 and 0.61 ± 0.14 μM in aortas from vehicle- and ACTH-treated mice, respectively (P < 0.05); in α2^R/R mice, the EC₅₀ was 0.28 ± 0.06 and 0.48 ± 0.13 μM in vehicle- and ACTH-treated mice, respectively (P = 0.23). Together, these in vivo and ex vivo data do not support the hypothesis that ACTH-induced hypertension is due to an overall increase in systemic vascular tone and reactivity.

Fig. 6.

Force development in response to graded doses of phenylephrine in isolated intact aortas from vehicle- and ACTH-treated α2^S/S and α2^R/R mice. *P < 0.05 for vehicle- vs. ACTH-treated α2^S/S at corresponding doses.

LV function.

As an extension of our overall hypothesis that ACTH-induced increases in endogenous cardiotonic steroids contribute to the development of hypertension, we predicted that ACTH treatment would enhance cardiac contractility in the α2^S/S mice but not in the α2^R/R mutants. We have previously reported that untreated α2^R/R mice show no differences in cardiac performance compared with α2^S/S mice and that these mice are insensitive to exogenously applied cardiac glycosides (2). Measurements of LV performance in ACTH-treated α2^S/S and α2^R/R mice are shown in Fig. 7. Under resting conditions, dP/dt_max was significantly higher in ACTH-treated α2^S/S mice compared with α2^R/R mice (11,425 ± 843 vs. 9,405 ± 585 mmHg/s; P = 0.007). Likewise, dP/dt₄₀, a contractile index that attempts to normalize for differences in afterload, was higher in ACTH-treated α2^S/S mice (9,005 ± 565 vs. 8,019 ± 404; P = 0.004). There were no differences in LV relaxation indexes between the two groups (data not shown), and the ability of the heart to maximally increase contractile performance in response to β-adrenergic stimulation was not influenced by ACTH treatment. Thus the development of ACTH-induced hypertension in α2^S/S mice appears to be associated with a significant elevation of LV contractile performance under resting conditions that is absent or blunted in the α2^R/R mice.

Fig. 7.

MAP, left ventricular systolic pressure (LVP_sys), maximum LV dP/dt (dP/dt_max), and LV dP/dt at 40 mmHg of developed pressure (dP/dt₄₀) in anesthetized ACTH-treated α2^S/S and α2^R/R mice instrumented with a high-fidelity transducer in the left ventricle. Measurements were made at baseline and during graded doses of dobutamine. *P < 0.05 compared with corresponding value in α2^R/R.

Serum aldosterone and corticosterone.

ACTH stimulates corticosterone production and may also influence aldosterone production. To determine whether the observed differences in blood pressure were related to alterations in ACTH-induced production of these two adrenal cortical steroids between the two genotypes, we measured serum levels in ACTH- and vehicle-treated α2^S/S or α2^R/R mice (Fig. 8). Although there was a slight tendency for aldosterone to be higher in the α2^R/R compared with α2^S/S mice, this difference was not significant, and there was no stimulation of aldosterone by ACTH treatment in either genotype. Basal corticosterone levels were similar in the two genotypes, and the stimulation caused by ACTH, approximately threefold, was the same in α2^S/S and α2^R/R mice.

Fig. 8.

Serum aldosterone and corticosterone concentrations in vehicle- and ACTH-treated mice. There were no differences in serum aldosterone levels, and corticosterone levels increased to the same degree in both genotypes. *P < 0.01 compared with vehicle-treated mice.

DISCUSSION

We have previously shown using the tail-cuff method that mice with a ouabain-resistant isoform of the α₂-NKA were resistant to ACTH-induced hypertension (6). In those studies, however, it was possible that the degree of stress associated with tail-cuff restraint might have influenced the results such that the observed differences between genotypes might actually reflect differences in the acute response to stress. In this regard, Gruber et al. (8) previously reported that ACTH-treated rats exhibit elevated blood pressure when measured by tail cuff (with restraint) but not when measured by indwelling catheter (unrestrained). To address this potential confounding variable, we used telemetry in the present study to evaluate resting blood pressure before and during ACTH treatment in α₂ ouabain-sensitive and ouabain-resistant mice. We found that baseline blood pressure was nearly identical between the two genotypes, supporting the hypothesis that the α₂ ouabain-binding site is not involved in the maintenance and control of resting blood pressure (see Fig. 1). Furthermore, we also confirmed that 5 days of ACTH treatment increased mean arterial blood pressure by ∼10–15 mmHg in wild-type mice, whereas ACTH treatment had virtually no effect in the α2^R/R mutants. Since we have already shown that circulating levels of endogenous cardiotonic steroids are increased similarly by ACTH treatment in α2^S/S and α2^R/R mice (6), our data suggest that the decreased sensitivity of the α₂-subunit in the mutant mice renders them resistant to the hypertensive influence of ACTH. It is important to note, however, that in our previous tail-cuff studies, the elevation in blood pressure associated with ACTH treatment was considerably more than that achieved in the present study: ∼25 mmHg. It appears likely, therefore, that a substantial component of the hypertension observed in these mice was indeed related to the acute stress associated with the tail-cuff method, in accordance with the previous report from Gruber et al. (8). On the other hand, current understanding of the role of acute stress in the pathogenesis of hypertension suggests that the distinctions between “true hypertension” and stress-induced hypertension are increasingly blurred (21).

The underlying nature of ACTH-induced hypertension has been investigated thoroughly over the last several decades, particularly through the work of Whitworth et al. (29), and although there appear to be clear hallmarks regarding mechanisms of hypertension, a consolidated hypothesis has yet to emerge. Importantly, it is likely that there are substantial species differences in the mechanisms underlying the elevation in blood pressure associated with increased levels of serum ACTH. One important observation is that in all species tested, the rise in blood pressure can be prevented by prior adrenalectomy and is therefore dependent to a substantial degree on adrenally produced steroids. In humans and rats, for example, administration of the primary glucocorticoid (cortisol and corticosterone, respectively) can in large part reproduce the hypertensive state (30). We found that corticosterone increased dramatically in response to ACTH treatment in both ouabain-sensitive and -resistant mice and that the magnitude of this response was not different between the genotypes. Likewise, aldosterone concentration was not different between groups either before or after ACTH and, in fact, did not respond at all to ACTH treatment. Therefore, the difference in the blood pressure response cannot be attributed to altered secretion of these adrenal cortical steroids. On the other hand, endogenous cardiotonic steroids, presumably derived from the adrenal gland as well, also have been implicated in the development of ACTH hypertension (7, 13, 14). In our previous study (6), we demonstrated that ACTH treatment increased plasma levels of endogenous cardiac glycosides in both α2^S/S and α2^R/R mice and that the ACTH-induced hypertension in the α2^S/S mice could be blocked by concomitant treatment with Digibind. In light of these previous findings, and since we have repeatedly found that ACTH-induced hypertension in mice is dependent on the sensitivity of the α₂-NKA to ouabain, it seems reasonable to postulate that endogenous cardiotonic steroids are a necessary mediator of the hypertensive response. Such a role of endogenous cardiotonic steroids does not negate an important role for other adrenal steroids, such as corticosterone, but rather suggests that a common and necessary component is the binding of a ligand to the α₂-NKA subunit. Since the α₂-NKA isoform is widely expressed throughout the mouse cardiovascular system, potential candidates for the downstream targets and mechanism(s) of action are numerous.

As an initial attempt to identify potential targets mediating the elevated blood pressure, we used the telemetered mice to evaluate whether the RAS was activated differentially by ACTH in the two genotypes. Losartan was administered to block AT₁ receptors, and the degree of blood pressure decline was evaluated as an index of the prevailing level of RAS activation. Before ACTH treatment, losartan treatment decreased pressure in the wild-type and mutant mice by 15 and 14 mmHg, respectively. After ACTH treatment, losartan decreased pressure by 13 and 11 mmHg, respectively. Thus the RAS appears to contribute to blood pressure maintenance equally in both genotypes, both before and after administration of ACTH, and does not appear to play a role in the development of hypertension in the wild-type mice, a finding that is consistent with earlier studies (23, 24, 28).

To likewise explore the potential role of the sympathetic nervous system in the development of hypertension, we also attempted to perform a similar experiment using the ganglionic blocker hexamethonium. To our surprise, hexamethonium at a dose of 40 μg/g body wt resulted in only a modest decrease in blood pressure in vehicle-treated α2^S/S mice (∼5–10 mmHg), whereas this same dose resulted in apparent cardiovascular collapse and abrupt death in all of the α2^R/R mice. Although this result at first suggested that blood pressure was maintained in these mice through elevated sympathetic drive, subsequent experiments showed that this dose of hexamethonium was causing skeletal muscle paralysis and cessation of breathing in the mutant mice. When anesthetized mice were artificially ventilated, the applied dose of hexamethonium did not cause cardiovascular collapse and, in fact, the decline in blood pressure was minimal and not different between wild-type and mutant mice. The reason for the altered sensitivity of cholinergic receptors at the motor end plate in α2^R/R mice is unknown, but functional interactions between nicotinic receptors and the α₂-NKA have been described (10).

Conventional paradigms propose that the stimulatory effects of cardiac glycosides on cardiac and vascular smooth muscle are related to decreases in Ca²⁺ extrusion by the Na⁺/Ca²⁺ exchanger secondary to increases in intracellular Na⁺ concentration (5, 9). Accordingly, we hypothesized that ACTH-stimulated increases in endogenous cardiotonic steroids mediate the hypertension that develops in these animals by blocking the α₂-NKA and that mice with a resistant α₂-isoform are therefore resistant to the development of hypertension. We evaluated vascular function, both in situ and ex vivo, and found that blood pressure and vascular resistance responses to PE were not significantly augmented in the ACTH-treated wild-type mice that developed hypertension. In fact, vascular responsiveness in both mesenteric and renal beds was paradoxically elevated in the ACTH-treated mutant mice that were resistant to hypertension, and blood flows were correspondingly low throughout the dose-response range. In the cerebral vasculature, there was a modest elevation in resistance that reached statistical significance in the hypertensive α2^S/S mice but not in the α2^R/R mice. In isolated aortas, responsiveness to PE was actually decreased by ACTH treatment in α2^S/S mice but not in the α2^R/R mutants. Vascular responses to ACTH treatment are therefore complex, and the possibility remains that subtle differences in vascular tone contribute to elevated blood pressure in ACTH-treated wild-type mice. It is further possible that differences in skeletal muscle blood flow, which were not evaluated in this study, may underlie the differences in blood pressure between the two groups. Since skeletal muscle blood flow is dramatically lowered by anesthesia (25), it would likely be necessary to evaluate this possibility in conscious mice.

There have been several prior studies evaluating hemodynamics in ACTH-induced hypertension. Notably, a study in rats reported that ACTH-induced hypertension was associated with increases in cardiac output and renal vascular resistance (RVR) but not in other vascular beds or in overall peripheral resistance, suggesting that the increase in RVR may specifically contribute to the increased blood pressure (27). Although we saw no such increase in renal blood flow in either wild-type or mutant mice, a renal basis for the differences in blood pressure remains a distinct possibility. It is possible, for example, that ACTH-induced elevations in endogenous cardiotonic steroids cause a shift of the pressure natriuresis relationship through a direct effect on the α₂-NKA in the kidney. However, since renal tubular cells express only the α₁-subunit, this effect would likely involve alterations in renal microvascular behavior.

Since previous investigations had consistently reported elevations in cardiac output in the early phases of ACTH-dependent hypertension (22, 26, 27), we sought to determine whether there were any differences in cardiac performance that might directly underlie the differences in blood pressure following ACTH treatment. Our hypothesis predicts that elevations in endogenous cardiotonic steroids following ACTH would enhance contractility in the α2^S/S wild-type mice but not in the α2^R/R mutants. We have previously shown that basal and β-adrenergic-stimulated cardiac contractility is comparable between wild-type and mutant mice and that ouabain administration increases contractility in the α2^S/S but not α2^R/R mice (2). In the present study we found that baseline cardiac contractility was elevated in ACTH-treated α2^S/S vs. α2^R/R mice. The combined data are consistent with the notion that cardiac output may be elevated to some degree in the hypertensive α2^S/S mice and that this increase in output is accommodated by elevated cardiac contractile function.

In summary, the data presented provide important confirmation that mice with a ouabain-resistant α₂-isoform of the Na⁺-K⁺-ATPase are resistant to the development of ACTH-induced hypertension. Use of telemetry confirms that differences in blood pressure between ACTH-treated α2^S/S and α2^R/R mice persist in the absence of the restraint stress associated with tail-cuff measurements. Vascular tone and reactivity appear to be relatively similar between α2^S/S and α2^R/R mice following ACTH treatment, with only small differences in responses between the genotypes. Cardiac function was incrementally elevated in the α2^S/S but not α2^R/R mice. Given that the α₂-NKA is expressed in a wide variety of tissues and cell types relevant to cardiovascular function, it is possible that the difference in phenotype between the wild-type and mutant mice is the result of a combination of subtle differences in cardiac, vascular, endothelial, and neuronal function.

GRANTS

This work was supported by National Institutes of Health Grants DK-57552 (to J. N. Lorenz), HL-66062, and HL-28573 (to J. B. Lingrel).