Barium Reduces Resting Blood Flow and Inhibits Potassium-Induced Vasodilation in the Human Forearm
Background— Increasing extracellular K+ concentration within and just above the physiological range hyperpolarizes and relaxes vascular smooth muscle in vitro. These actions involve inwardly rectifying potassium channels (KIR) and Na+/K+ ATPase, which are inhibited, respectively, by Ba2+ and ouabain. The role (if any) of KIR in controlling human resistance vessel tone is unknown, and we investigated this in the forearm.
Methods and Results— Blood flow was measured by plethysmography in healthy men. Drugs and electrolytes were infused through the brachial artery. BaCl2 (4 μmol/min, also used in subsequent experiments) increased Ba2+ plasma concentration in the infused forearm to 50±0.8 μmol/L (mean±SEM) and reduced blood flow by 24±4% (n=8, P<0.001) without causing systemic effects. Ouabain (2.7 nmol/min), alone and with BaCl2, reduced flow by 10±2% and 28±3%, respectively (n=10). Incremental infusions of KCl (0.05, 0.1, and 0.2 mmol/min) increased flow from baseline by 1.0±0.2, 2.0±0.4, and 4.2±0.5 mL/min per deciliter forearm, respectively. Responses to KCl (0.2 mmol/min) were inhibited by BaCl2, alone and plus ouabain, by 60±9% and 88±6%, respectively (both P≤0.01). In control experiments, norepinephrine (240 pmol/min) reduced blood flow by 24±2% but had no significant effect on K+-induced vasodilation. BaCl2, alone or with ouabain, did not significantly influence responses to verapamil or nitroprusside.
Conclusions— Ba2+ increases forearm vascular resistance. K+-induced vasodilation is selectively inhibited by Ba2+ and almost abolished by Ba2+ plus ouabain, suggesting a role for KIR and Na+/K+ ATPase in controlling basal tone and in K+-induced vasorelaxation in human forearm resistance vessels.
Received January 9, 2002; accepted January 11, 2002.
Strong inwardly rectifying potassium channels (a family of channels termed KIR) influence the membrane potential and contractile state of several kinds of vascular smooth muscle and contribute to vasodilator responses to modestly increased extracellular potassium ion concentrations ([K+]o <15 mmol/L) in vitro.1–4⇓⇓⇓ Such responses could be physiologically or pathologically important because elevated [K+]o occurs in exercising muscle and during myocardial or cerebral ischemia. K+ has also been implicated as an endothelium-derived hyperpolarizing factor (EDHF) in some5 but not all6 blood vessels. Electrophysiological, molecular biological, and gene knockout experiments have shown that one member of the KIR channel family, KIR2.1, is needed for KIR currents and for K+-induced vasodilation in rat cerebral arteries in vitro, effects that are blocked by Ba2+ concentrations <100 μmol/L.7,8⇓ It is not known, however, whether KIR channels control human resistance vasculature under physiological conditions in vivo, and the present study addresses this.
Ba2+ has been proposed as a probe of the functional role of KIR channels,2 and concentrations of Ba2+ <100 μmol/L are relatively specific for these channels.9 Higher concentrations of Ba2+ (100 to >1000 μmol/L) are needed to block non-KIR potassium channels, including HERG (an inwardly rectifying potassium channel present in human hippocampus and distinct from KIR2.1),10 ATP-sensitive K-channels,11 Ca2+-activated K-channels,12 and KV channels.13 Soluble barium salts are rapidly absorbed from the intestinal tract, and Ba2+ is toxic in systemically active doses. Acute toxic effects of Ba2+, which include hypertension, are consistent with block of KIR.14 Human studies identified a no-observed adverse-effect dose of 0.21 mg barium/kg body wt per day.15,16⇓ These data were used by the US Environmental Protection Agency to calculate a chronic oral reference dose of 0.07 mg/kg per day.17,18⇓ In one study, human volunteers were given up to 10 mg/L in drinking water daily for up to 10 weeks. There were no clinically significant adverse events.16 We calculated that brachial artery infusion of BaCl2 could achieve a local Ba2+ concentration of 30 to 50 μmol/L in plasma of the infused forearm without causing systemic toxicity.
We measured effects of Ba2+ on K+-induced vasodilation as well as on basal blood flow. Increasing [K+]o displaces KIR channel–bound polyamines19; in the relevant concentration range, this increases outward K+ current, hyperpolarizing the membrane.4 Increasing [K+]o also hyperpolarizes vascular smooth muscle by activating the electrogenic Na+/K+ pump.3 Brachial artery infusion of ouabain was therefore used in some experiments to produce unilateral regional block of Na+/K+ ATPase.20–22⇓⇓ Experiments with vasodilators (verapamil and nitroprusside) and vasoconstrictors (norepinephrine) that act by mechanisms distinct from KIR and the Na+/K+ pump were used to control for possible nonspecific effects.
The St Thomas’ Hospital Research Ethics Committee approved all the studies, and subjects gave written informed consent. Studies were performed in healthy normotensive (arterial pressure, 116±4/70±2 mm Hg; mean±SEM), normocholesterolemic (mean serum cholesterol, 3.6±0.2 mmol/L) men (mean age, 33±2 years) who were taking no medication. Studies were performed in the morning in a temperature-controlled room (24±1°C). The left brachial artery was cannulated with a 27-gauge stainless steel needle (Cooper’s Needleworks) after local anesthesia (<1 mL 1% lidocaine). Isotonic saline (140 mmol/L sodium chloride ) with or without dissolved drugs was infused at 1 mL/min. BaCl2 was administered through a syringe filter (Acrodisc, 0.2 μm, Pall Corporation). Subjects rested supine for 30 minutes, and saline was infused for 12 minutes before measuring baseline flow. Blood flow (milliliter per minute per deciliter of forearm volume) was measured simultaneously in both arms by venous occlusion plethysmography23 with the use of electrically calibrated strain gauges.24 During measurements, the hands were excluded from the circulation by inflation of wrist cuffs to 180 mm Hg. Upper arm cuffs were intermittently inflated to 40 mm Hg, and the mean of the last 5 readings of each infusion period was used for analysis.
In preliminary dose-ranging studies, 19-gauge plastic cannulas were inserted into the medial antecubital veins draining each forearm. Venous blood (10 mL) was sampled at baseline, during the final 30 seconds of the infusion of BaCl2, and 1 hour after infusion for determination of plasma Ba2+ concentration. Samples were immediately centrifuged (1600g for 5 minutes), and plasma was stored at −18°C in Ba2+-free tubes. Samples were analyzed in duplicate in the Trace Element Unit, Southampton Hospital, with the use of inductively coupled plasma (ICP) mass spectrometry (Perkin Elmer–Sciex Elan 5000 ICP mass spectrometer), detection limit of 10 nmol/L, with a 0.5-mL plasma sample. Routine serum chemistries including creatinine, electrolytes, and liver function tests, in addition to Ba2+ concentration, were measured at baseline and 1 week after BaCl2 infusion.
In preliminary dose-finding experiments, cumulative infusions of intra-arterial BaCl2 (0.25 to 2.0 μmol/min, each for 4 minutes) were administered. From these pilot studies, a dose of BaCl2 of 4 μmol/min for up to 6 minutes was chosen for the following protocols.
Effects of BaCl2 and Ouabain on Basal Forearm Blood Flow
After measuring basal forearm blood flow as above, BaCl2 (4 μmol/min) was infused into the brachial artery for 6 minutes and blood flow was measured in both arms (n=8). In 6 of these subjects, plasma Ba2+ determinations were made as above. In 10 separate experiments, ouabain (2.7 nmol/min) was infused for 15 minutes and blood flows were measured as described elsewhere.21 Ouabain was continued for a further 3 minutes, coinfused with BaCl2 (4 μmol/min).
Effects of Ba2+ (With or Without Ouabain) on Forearm Responses to K+ or Verapamil
After measuring basal forearm blood flow, cumulative doses (0.05, 0.1, 0.2 mmol/min, n=8) of KCl were infused into the brachial artery, each dose for 3 minutes. Venous blood samples (10 mL) were taken from both arms during the final 30 seconds for determination of plasma K+ concentration.
In separate experiments (n=8), KCl (0.2 mmol/min) was infused for 3 minutes followed by saline for 15 minutes, during which blood flow returned to baseline. BaCl2 (4 μmol/min) was then infused for 6 minutes: alone for 3 minutes and during a second 3-minute infusion of KCl (0.2 mmol/min). In control experiments, after blood flow was measured at baseline, verapamil (80 nmol/min) was infused with saline followed sequentially by saline alone, BaCl2 alone, and verapamil with BaCl2 (n=7).
In further separate experiments (n=6), blood flow was measured as before at baseline and during KCl (0.2 mmol/min), followed by a 15-minute saline recovery period. Ouabain and BaCl2 were infused as described above and continued during a final 3-minute KCl infusion (0.2 mmol/min).
Lack of Effect of Norepinephrine on Forearm Blood Flow Response to KCl
Baseline blood flow was measured, followed by KCl infusion for 3 minutes (0.2 mmol/min, n=5). Saline was then infused alone for a 15-minute recovery period, followed by norepinephrine (240 pmol/min),initially alone for 3 minutes, and then continued throughout a second 3-minute KCl infusion.
Lack of Effect of BaC12 Plus Ouabain on Forearm Blood Flow Response to Nitroprusside
Baseline blood flow was measured during saline, followed by nitroprusside (3.3 nmol/min or 33 nmol/min for 3 minutes, n=4). Saline was then infused for a 15-minute recovery period, followed by sequential cumulative additions of ouabain (2.7 nmol/min), BaCl2 (4 μmol/min), and nitroprusside (3.3 or 33 nmol/min), with the same timings used as above.
Materials and Drugs
Drugs were obtained from Baxter Healthcare (saline); David Bull Laboratories (sodium nitroprusside); Antigen Pharmaceuticals (KCl, lidocaine); Abbott Laboratories (norepinephrine); BDH Laboratories (BaCl2 10% wt/vol); Jenapharm (ouabain); and Baker Norton (verapamil).
Data are presented as mean±SEM. Vasodilator responses were calculated as increases of blood flow (mL/min per deciliter of forearm tissue) above the immediately preceding baseline.25,26⇓ Vasoconstrictor responses were calculated as percentage decrease of blood flow ratio of the infused to noninfused arm.27,28⇓ Differences between means were evaluated for statistical significance by means of Student’s paired t test (2-sided) or repeated-measures ANOVA, as appropriate. Differences were considered significant at a level of P<0.05.
Neither arterial blood pressure nor forearm blood flow in the noninfused arm changed significantly during any study, consistent with lack of systemic hemodynamic effects of the infused agents at the doses used. BaCl2 caused no acute ECG changes and had no significant effect on serum electrolytes and liver function tests measured 1 week after the infusion.
Effects of Ba2+ and Ouabain on Basal Flow
In pilot experiments, cumulative, rising doses of BaCl2 (0.25, 0.5, 1.0, and 2.0 μmol/min, n=6), each infused through the brachial artery for 4 minutes, had no significant effect on baseline flow (percent change in baseline flow, −10±10%, −9±7%, −2±14%, and 1±15%, respectively, P=NS). Plasma Ba2+ concentration at baseline (ie, before BaCl2 infusion) was 0.28±0.04 μmol/L and during the final 30 seconds of infusion (at 2.0 μmol/min) was 22±3.4 μmol/L and 1.7±0.9 μmol/L, in the infused and noninfused arms respectively.
BaCl2 (4 μmol/min for 6 minutes) reduced baseline blood flow by 24±4% (n=8, P<0.001, Figure 1). Plasma Ba2+ concentration in venous blood draining the infused arm was 50±8 μmol/L during the final 30 seconds of the infusion (Table 1).
Ouabain (2.7 nmol/min) reduced baseline blood flow by 10±2% (n=10, P<0.05) and coinfusion of BaCl2 with ouabain reduced baseline flow by 28±3% (n=10, P<0.0005 compared with the effect of ouabain alone, Figure 1).
Effect of K+ on Forearm Blood Flow
All subjects reported warmth and tingling in the infused arm during KCl infusion, and in preliminary experiments discomfort limited the maximum dose that was consistently tolerated to 0.2 mmol/min. KCl (0.05, 0.1, and 0.2 mmol/min) increased blood flow by 1.03±0.24, 1.99±0.4, and 4.21±0.46 mL/min per deciliter forearm over baseline (n=8, P<0.0001, ANOVA; Figure 2). The concentration of K+ in plasma from venous blood from the infused arm at the end of the infusion of KCl was 6.3±0.2 mmol/L, whereas the concentration in plasma from venous blood draining the noninfused arm was 4.0±0.2 mmol/L (n=6, P<0.0001).
Effects of Ba2+(With or Without Ouabain) on Forearm Response to K+ or Verapamil
KCl (0.2 mmol/min) alone increased forearm blood flow by 3.53±0.56 mL/min per deciliter forearm over baseline (n=8, Table 2) and by 1.43±0.28 mL/min per deciliter forearm when coinfused with BaCl2 (60±9% reduction; P=0.01, Figure 3A). Verapamil (80 nmol/min) increased forearm blood flow by 4.50±0.84 mL/min per deciliter forearm when infused with saline and by 4.46±0.86 mL/min per deciliter forearm when infused with BaCl2 (n=7, Table 3 P=NS). In separate experiments, KCl (0.2 mmol/min) increased forearm blood flow by 4.49±0.68 mL/min per deciliter forearm over baseline when infused alone and by 0.57±0.23 mL/min per deciliter forearm when coinfused with BaCl2 and ouabain (88±6% reduction, n=6, P<0.005; Figure 3B and Table 4).
Responses to KCl were measured before and during vasoconstriction with norepinephrine. Norepinephrine (240 pmol/min) reduced baseline blood flow by 24±2% (n=5, P<0.05). KCl (0.2 mmol/min) increased forearm blood flow by 3.01±0.45 mL/min per deciliter forearm over baseline when infused alone and by 3.79±0.64 mL/min per deciliter of forearm when coinfused with norepinephrine (P=NS, Figure 3C).
Vasodilator responses to nitroprusside were measured in the presence and absence of Ba2+ and ouabain (Table 4). Nitroprusside (3.3 and 33 nmol/min) increased forearm blood flow, respectively, by 4.13±0.95 and 8.55±1.1.4 mL/min per deciliter of forearm when infused with saline alone and by 4.19±0.39 and 9.30±2.06 mL/min per deciliter of forearm when coinfused simultaneously with ouabain and BaCl2 (P=NS; Figure 3D).
Ba2+ as a Probe of KIR Function in Human Forearm Resistance Vasculature
Electrophysiological and molecular studies have established the presence of KIR2.1 channels in isolated vascular smooth muscle.7,8,23⇓⇓ However, KIR channels are blocked by Ca2+ and Mg2+,9 each present in millimolar concentrations in extracellular fluid, and other signal transduction mechanisms could overwhelm any contribution to basal tone from KIR in vivo. Ba2+ can be used to probe the functional role of KIR channels.2 We investigated the possible role of KIR channels in human resistance vasculature in vivo by using brachial artery administration of BaCl2 with bilateral measurement of forearm blood flow. Doses were selected to produce concentrations that are pharmacologically active locally in the infused forearm but not systemically, and no significant effects on blood pressure or on blood flow in the contralateral (noninfused) forearm were observed. The dose of Ba2+ administered during each study was less than the daily reference dose, and no adverse effects were observed. The mean concentration of Ba2+ measured in plasma from the infused arm was 50±0.8 μmol/L, sufficient to block KIR channels.2–4,7–9⇓⇓⇓⇓⇓ Ouabain and other drugs were similarly administered through the brachial artery in doses that are not active systemically20–22⇓⇓ and produced no change in blood flow in the noninfused arm.
A major concern with this approach is the possibility that Ba2+ and ouabain, in the doses studied, could have effects in addition to those on KIR and Na+/K+ ATPase. Bilateral measurement of forearm blood flow coupled with brachial artery administration of subsystemically active doses circumvents confounding from drug actions on the central nervous system or heart, which cause bilateral changes in forearm blood flow.26 Thus, central effects of ouabain resulting in increased vascular resistance cannot explain its unilateral effects on the infused forearm. Similarly, actions of Ba2+ on the heart or central nervous system cannot explain its unilateral effects in the infused arm. Concentrations of Ba2+ in the range of 100 to >1000 μmol/L are needed to block other potassium channels,10–13⇓⇓⇓ which have not been implicated in K+-induced vasodilation and are unlikely to underlie the effect of Ba2+ described here. However, it is not possible to exclude completely some other property of Ba2+ in the physiological milieu of the human forearm. In particular, physicochemical properties shared with Ca2+ raise the theoretical possibility that Ba2+ could exert some of the effects we observed by influencing [Ca2+]i signaling. We sought effects of Ba2+ on responses to verapamil, an antagonist of L-type voltage-dependent Ca2+-channels, which are active under physiological conditions, to address this possibility. The lack of effect of Ba2+, in a dose that inhibited K+-responses, on responses to verapamil argues against an action of Ba2+ on [Ca2+]i signaling under the experimental conditions in our studies. Furthermore, doses of Ba2+ and ouabain that almost abolished K+-induced vasodilation had no significant effect on vasodilator responses to nitroprusside, ruling out effects of these antagonists on cGMP-mediated vasodilation. Finally, we considered the possibility that the effects of these antagonists on K+-induced vasodilatation were simply the result of vasoconstriction. However, a dose of norepinephrine that caused a similar reduction in baseline blood flow as did Ba2+ plus ouabain had no significant effect on K+-induced vasodilation. The main new positive findings of the present study, namely that Ba2+ constricts forearm resistance vasculature under basal physiological conditions, selectively inhibits K+-induced vasodilation, and virtually abolishes K+ responses when combined with ouabain, can thus best be explained by selective actions of Ba2+ and ouabain, in the doses used, on, respectively, KIR and Na+/K+ ATPase in the infused forearm.
Implications of Tonic KIR and Na+/K+ ATPase Activity in Resistance Vasculature In Vivo
Physiological conditions that influence KIR-channel activity include membrane potential and K+, Ca2+, and Mg2+ concentrations.2,4⇓ Consequently, unequivocal evidence of the presence and function of KIR2.1 channels in arterial smooth muscle in vitro4,7,8⇓⇓ does not, of itself, prove the functional importance of these channels under in vivo conditions. The modest vasoconstriction caused in the forearm by ouabain (10±2% reduction in basal blood flow, consistent with previous work20–22⇓⇓) and greater effect of Ba2+, demonstrated here for the first time, are consistent with tonic activity of K+ on Na+/K+ ATPase and, more effectively, on KIR-channels in vascular smooth muscle in forearm resistance vessels in vivo. Tonic K+-related vasodilation could have been underestimated by the inhibition observed (24±4% by Ba2+ alone, 28±3% by Ba2+ plus ouabain) because, due to concerns related to toxicity, we did not perform prolonged infusions. Vasoconstriction caused by Ba2+ may therefore not have reached a plateau. Even so, Ba2+-induced vasoconstriction was substantial and compares with a maximum vasoconstrictor effect of l-NG-monomethyl-l-arginine (which blocks basal endothelium-derived nitric oxide synthesis in this vascular bed) of ≈50% reduction of basal blood flow.29 Another limitation of the study is that it was not practicable to investigate the effect of Ba2+ on K+-induced responses throughout a dose range of KCl infusions because responses to lower doses of KCl were small and variable and larger doses could not be used as these caused forearm discomfort.
Tonic vasodilator activity of K+ within the physiological range could influence arterial blood pressure. Dietary supplementation with potassium salts has been reported to lower blood pressure in patients with essential hypertension.30,31⇓ Conversely, elevated blood pressure has been described in association with low dietary K+.32 The ability of a small increase in [K+]o to hyperpolarize vascular smooth muscle could contribute to the hypotensive effect of potassium salts. Local plasma K+ concentrations (mean, 6.3±0.2 mmol/L) achieved in venous blood draining the infused arm are in the range observed in voluntary muscle during exercise,4 supporting a role for KIR and for Na+/K+ ATPase activation in regional blood flow responses during exercise.4 It is likely that mechanisms similar to those we have observed in forearm vasculature also operate in vivo in cerebral and coronary vessels (which are perfused by plasma of the same composition as are forearm vessels) and in which KIR-channels have been detected in vitro.2,3⇓ If so, K+-induced vasodilation could be important in pathological conditions such as myocardial or cerebral infarction, in which local elevations of [K+]o could reduce resistance vessel tone in and around the infarct.4 Furthermore, hyperpolarization-induced vasorelaxation by activation of Na+/K+ ATPase is reported to be upregulated when nitric oxide bioavailability is impaired,33,34⇓ so K+-induced vasodilation could provide a compensatory mechanism in diseases in which the l-arginine/nitric oxide mechanism is impaired, especially if the KIR mechanism also proves to be upregulated in such disorders.
In conclusion, Ba2+ constricts forearm resistance vessels, and K+-induced vasodilation is selectively inhibited by Ba2+ and almost abolished by Ba2+ plus ouabain. These findings support a role for KIR and Na+/K+ ATPase in controlling basal tone and in K+-induced vasorelaxation in human forearm resistance vessels in vivo. Such a role has important implications for the (patho-) physiological regulation of resistance vessel tone in humans.
This work was supported by the British Heart Foundation. We thank Dr Alison Jones (Guy’s Poisons Unit) for advice on the toxicity of barium salts.
- ↵Robertson BE, Bonev AD, Nelson MT. Inward rectifier K+ currents in smooth muscle cells from rat coronary arteries: block by Mg2+, Ca2+ and Ba2+. Am J Physiol. 1996; 40: H696–H705.
- ↵McCarron JG, Halpern W. Potassium dilates rat cerebral arteries by two independent mechanisms. Am J Physiol. 1990; 259: H902–H908.
- ↵Quayle JM, Nelson MT, Standen NB. ATP-sensitive, and inwardly rectifying potassium channels in smooth muscle. Physiol Rev. 1997; 77: 1165–1232.
- ↵Zaritsky JJ, Eckman DM, Wellman GC, et al. Targeted disruption of KIR 2:1 and KIR 2.2 genes reveals the essential role of the inwardly rectifying K+ current in K+-mediated vasodilation. Circ Res. 2000; 87: 160–166.
- ↵Wellman GC, Bevan JA. Barium inhibits the endothelium-dependent component of flow but not acetylcholine-induced relaxation in isolated rabbit cerebral arteries. J Pharmacol Exp Ther. 1995; 274: 47–53.
- ↵Trudeau MC, Warmke JW, Ganetzky B, et al. HERG, a human inward rectifier in the voltage-gated potassium channel family. Science. 1995; 269: 92–95.
- ↵Vergara C, Latorre R. Kinetics of Ca2+-activated K+ channels from rabbit muscle incorporated into planar bilayers: evidence for a Ca2+ and Ba2+ blockade. J Gen Physiol. 1983; 82: 243–368.
- ↵Nelson MT, Quayle JM. Physiological roles and properties of potassium channels in arterial smooth muscle. Am J Physiol. 1995; 268: C799–C822.
- ↵Brenniman GR, Levy PS. Epidemiological study of barium in Illinois drinking water supplies. In: Calabrese EJ, ed. Advances in Modern Toxicology. Princeton, NJ: Princeton Scientific Publications; 1984: 231–249.
- ↵Environmental Protection Agency. Health Effects Assessment Summary Tables. Annual FY-95. Washington, DC: Environmental Protection Agency; 1995.
- ↵Environmental Protection Agency. Integrated Risk Information System (IRIS). Health Risk Assessment for Barium. (Verification date 6/21/90). Washington, DC: Environmental Protection Agency; 1995.
- ↵Mason DT, Braunwald E. Studies on digitalis: effects of ouabain on forearm vascular resistance and venous tone in normal subjects and in patients in heart failure. J Clin Invest. 1964; 43: 532–543.
- ↵Whitney RJ. The measurement of volume changes in human limbs. J Physiol (Lond). 1953; 121: 1–27.
- ↵Benjamin N, Calver A, Collier J, et al. Measuring forearm blood flow and interpreting the responses to drugs and mediators. Hypertension. 1995; 25: 918–923.
- ↵Tobian L. The protective effects of high-potassium diets in hypertension. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis, and Management. 2nd ed. New York, NY: Raven Press; 1995: 299–307.
- ↵Smith SR, Klotman PE, Svetkey LP. Potassium chloride lowers blood pressure and causes natriuresis in older patients with hypertension. J Am Soc Nephrol. 1992; 2: 1302–1309.
- ↵Lawton WJ, Fitz AE, Anderson EA, et al. Effect of dietary potassium on blood pressure, renal function, muscle sympathetic nerve activity, and forearm vascular resistance and flow in normotensive and borderline hypertensive humans. Circulation. 1990; 81: 173–184.
- ↵Taddei S, Ghiadoni L, Virdis A, et al. Vasodilation to bradykinin is mediated by an ouabain-sensitive pathway as a compensatory mechanism for impaired nitric oxide availability in essential hypertensive patients. Circulation. 1999; 100: 1400–1405.
- ↵Taddei S, Virdis A, Mattei P, et al. Effect of insulin on acetylcholine-induced vasodilation in normotensive subjects and patients with essential hypertension. Circulation. 1995; 92: 2911–2918.