Neuronal NO Mediates Cerebral Vasodilator Responses to K+ in Hypertensive Rats
Potassium ion (K+) normally causes cerebral vasodilatation by activating inwardly rectifying K+ (KIR) channels. We tested whether chronic hypertension affects the magnitude and/or mechanism of K+-induced cerebral vasodilatation in vivo. Basilar artery responses were examined in anesthetized Wistar-Kyoto (WKY; mean arterial pressure, 114±4 mm Hg) and spontaneously hypertensive (SHR; 176±3 mm Hg) rats. In WKY, elevating cerebrospinal fluid K+ concentration from 3 mmol/L to 5 and 10 mmol/L caused vasodilatation (percent maximum, 12±1 and 48±7, respectively). The response to 5 mmol/L K+ was greater in SHR (percent maximum, 17±2 [P<0.05 versus WKY] and 49±4). The KIR channel inhibitor, barium ion (Ba2+, 100 μmol/L) selectively inhibited dilator responses to 5 and 10 mmol/L K+ by ≈75% in WKY. In SHR, Ba2+ had no effect on the response to 5 mmol/L K+, and only partially inhibited (by ≈40%) the response to 10 mmol/L K+. The nonselective NO synthase (NOS) inhibitor Nω-nitro-l-arginine methyl ester, the neuronal NOS (nNOS) inhibitor 1-(2-trifluromethyl-phenyl)imidazole, and the N-type calcium channel inhibitor ω-conotoxin GVIA, were all without effect in WKY, but markedly inhibited the response to 5 mmol/L K+ in SHR. When applied together with Ba2+, each of these inhibitors also profoundly reduced responses to 10 mmol/L K+ in SHR. Immunostaining of basilar arteries revealed that the perivascular nNOS-containing nerve plexus was denser in SHR. Thus, K+ dilates the normotensive basilar artery predominantly via KIR channel activation. During chronic hypertension, small physiological elevations in K+ dilate the basilar artery by an nNOS-dependent mechanism that appears to be upregulated in a compensatory manner.
Potassium ion (K+), released by active neurons, may be an important mediator linking increases in cerebral metabolism and blood flow by producing vasodilatation in the cerebral circulation.1 This powerful vasodilator effect of K+ occurs in response to remarkably small changes in brain extracellular K+ concentration, which is normally ≈3.0 mmol/L. Under most physiological conditions, extracellular K+ concentration rarely rises above 4.0 to 5.0 mmol/L.2
Experimental application of up to ≈20 mmol/L K+ dilates cerebral arteries in vivo3,4⇓ and relaxes isolated cerebral arteries.5–8⇓⇓⇓ Functional and electrophysiological findings made in vitro5–9⇓⇓⇓⇓ suggest the main mechanism underlying the cerebral vasorelaxant effect of K+ is activation of inwardly rectifying K+ (KIR) channels on vascular muscle cells, leading to K+ efflux, membrane hyperpolarization, and closure of voltage-operated calcium (Ca2+) channels. However, other mechanisms—such as activation of the sodium-potassium adenosine triphosphatase (Na+,K+-ATPase),6 non–KIR-related K+ channels,10 or NO synthase (NOS)11—have also been suggested to play a role in the cerebral vasodilator response to K+. In normotensive rats, K+-induced dilatation of the basilar artery in vivo is inhibited by barium ion (Ba2+), suggesting an important role for activation of KIR channels.4
Chronic hypertension is a major risk factor for stroke and is characterized by abnormalities of several constrictor and dilator mechanisms in cerebral arteries. There is some evidence that K+-induced cerebral vasodilatation may be impaired during hypertension,12 but no study has examined this possibility in vivo. We therefore tested the hypothesis that cerebral vasodilator responses to physiologically relevant increases in K+ concentration are impaired in vivo during chronic hypertension, and that expression of KIR channels is lower in cerebral arteries from spontaneously hypertensive rats (SHR) versus Wistar Kyoto rats (WKY). Importantly, we used a basal concentration of ≈3 mmol/L K+ (as opposed to 5 to 6 mmol/L K+, which has been commonly used by others7,8⇓ and is equivalent to normal levels of K+ in cerebrospinal fluid).2 Indeed, we found an augmented vasodilator effect and a fundamental difference in the mechanism underlying the vasodilator response to 5 mmol/L K+ in hypertensive animals. These responses to K+ in SHR are KIR channel–independent and appear to be caused by release of NO from neuronal NOS (nNOS)-containing perivascular nerves.
Experiments were performed in male WKY (7.8±2.5 months-old [mean±SD]; n=41) and SHR (6.7±2.1 month-old, n=57) (Animal Resources Centre, Canning Vale, Western Australia). Weight of WKY was 411±68 g, and SHR was 430±62 g.
In Vivo Protocol
The surgical procedure for preparing and measuring basilar artery diameter in anesthetized rats has been described previously.4 Blood gases were maintained within normal levels (pH, 7.36±0.01; pCO2, 33±1 mm Hg; and pO2, 167±5 mm Hg). Cerebrospinal fluid sampled from the cranial window had the following values: pH, 7.38±0.01; pCO2, 35±1 mm Hg; and pO2, 122±1 mm Hg. In 7 groups of rats, we tested the effect of the following inhibitors on vasodilator responses to K+ and appropriate control dilators: (1) 100 μmol/L Ba2+ in WKY (n=10) and SHR (n=9); (2) 100 μmol/L ouabain in SHR (n=4); (3) a combination of 300 μmol/L 4-aminopyridine (4-AP), 1 μmol/L glibenclamide, and 1 mmol/L tetraethylammonium ion (TEA) in SHR (n=4); (4) 1 μmol/L tetrodotoxin in SHR (n=5); (5) 30 μmol/L Nω-nitro-l-arginine methyl ester (L-NAME) in WKY (n=5) and SHR (n=6); (6) 100 μmol/L 1-(2-trifluromethylphenyl)imidazole (TRIM) in WKY (n=3) and SHR (n=11), and (7) 100 nmol/L ω-conotoxin GVIA in WKY (n=4) and SHR (n=5). In groups 5 through 7, the effect of a second inhibitor treatment was tested: control responses were established followed by treatment of the vessel with L-NAME, TRIM, or ω-conotoxin GVIA, respectively, and responses were re-tested. Ba2+ was then applied to the vessel in combination with the first inhibitor, and responses to vasodilators were again tested. At the completion of experiments, a high-dose mixture of sodium nitroprusside (SNP, 100 μmol/L) and nimodipine (10 μmol/L) was applied to the basilar artery to determine maximum diameter.
Four age-matched pairs of WKY and SHR were anesthetized and perfusion-fixed. The entire length of each basilar artery was dissected out and processed for immunohistochemistry. Arteries from the 4 pairs of animals were shown to 5 observers, who were blinded as to the strains of rat. Observers decided for each pair of arteries whether there was a difference in the density of the nNOS-immunoreactive terminals, and if so, which nerve plexus was denser.
Reverse-Transcription coupled with Polymerase Chain Reaction
Total RNA was extracted from cerebral arteries of eight rats (4 WKY, 4 SHR). RNA was reverse transcribed (RT) to produce cDNA using random hexamers as primers. Semi-quantitative RT–polymerase chain reaction (PCR) was performed by varying the PCR cycle number (20, 25, and 30 cycles). Expression of β-actin was used as an internal control.
Data Presentation and Statistics
Vasodilator-induced increases in artery diameter from baseline are expressed as percentage of the maximum response produced by SNP and nimodipine. Single comparisons were made using Student’s paired or unpaired t test, as appropriate. Multiple comparisons were made using ANOVA for repeated measures, followed by a Tukey-Kramer test. For evaluation of density of nNOS-immunoreactive nerve terminals around basilar arteries from WKY and SHR, the results from 5 observers were analyzed using a χ2 analysis for goodness of fit. The results were tested against expected frequencies generated by a random assignment to each category. For all statistical analysis, P<0.05 was considered significant.
An expanded Methods section can be found in an online data supplement available at http://www.hypertensionaha.org.
Mean arterial blood pressure averaged 114±4 mm Hg in WKY and 176±3 mm Hg in SHR (P<0.05). Baseline diameter of the basilar artery differed between the 2 strains (WKY, 245±4 μm; SHR, 224±4 μm; P<0.05). Maximum diameter, as defined by the diameter achieved in response to combined application of SNP and nimodipine, was 369±5 μm in WKY and 361±6 μm in SHR.
Vasodilator Responses in WKY Versus SHR
Elevating K+ concentration in the cerebrospinal fluid from 3 mmol/L to 5 and 10 mmol/L elicited concentration-dependent vasodilatation in normotensive WKY (Figure 1a). The response to 5 mmol/L K+ was significantly greater by 40% in SHR, whereas responses to 10 mmol/L K+ were similar in the 2 strains (Figure 1a). Acetylcholine (ACh) and SNP caused increases of basilar artery diameter in WKY and SHR (Figure 1b and 1c, respectively). Responses to ACh were smaller in SHR versus WKY (Figure 1b). Responses to SNP were equivalent in WKY and SHR (Figure 1c).
Effect of Ba2+ in WKY
Ba2+ (100 μmol/L) caused constriction of the basilar artery in WKY (baseline, 236±7 μm; Ba2+-treated, 217±5 μm; Δ=−8±1%, n=17). Treatment with Ba2+ markedly inhibited vasodilator responses to K+ in WKY by 50% to 75% (Figure 2a). Aprikalim (10 μmol/L), an activator of ATP-sensitive K+ channels, induced dilatation of the basilar artery, and this response was unaffected by Ba2+ (Figure 2b), indicating that the inhibitory effect of Ba2+ was selective for responses to K+. The response to aprikalim could be inhibited by further treatment with 10 μmol/L glibenclamide (Figure 2b).
Effect of Ba2+ in SHR
The constrictor effect of Ba2+ was slightly greater in SHR (baseline, 210±4 μm; Ba2+-treated, 183±5 μm; Δ=−13±2%, n=9; P<0.05 versus WKY). Surprisingly, treatment with Ba2+ had no effect on the response to 5 mmol/L K+ and only inhibited the response to 10 mmol/L K+ by ≈40% (Figure 2c). Vasodilator responses to SNP were unaffected by Ba2+ (Figure 2d). Thus, it appears there is a much less important role for a KIR channel–mediated contribution to K+-induced vasodilator responses in SHR versus WKY. We therefore conducted the following groups of studies to identify any additional mechanism(s) that may contribute to the dilator response to K+ in SHR.
Effect of Na+,K+-ATPase and K+ Channel Inhibitors on Vasodilator Responses to K+ in SHR
Ouabain (100 μmol/L) decreased diameter of the SHR basilar artery (baseline, 211±6 μm; ouabain-treated, 194±3 μm; Δ=−8±1%, n=4) but had no effect on K+-induced vasodilatation (control responses for 5 mmol/L and 10 mmol/L, 9±4% and 39±13% of maximum, respectively; responses during 100 μmol/L ouabain treatment, 17±15% and 36±19% of maximum, respectively; n=4). Combined application of glibenclamide (1 μmol/L, an inhibitor of ATP-sensitive K+ channels), 4-AP (300 μmol/L, an inhibitor of voltage-dependent K+ channels), and TEA (1 mmol/L, an inhibitor of large conductance calcium-activated K+ channels) caused profound rhythmic activity of the basilar artery, with oscillations in diameter typically occurring with a frequency of ≈2 cycles/min (n=4) and amplitude of 30 to 70 μm (n=4). Mean (mid-point) diameter was decreased by ≈8% during this treatment (baseline, 210±10 μm; glibenclamide/TEA/4-AP treated, 195±4 μm). However, the vasodilator response to K+ was not inhibited by this combination of K+ channel inhibitors (control responses for 5 mmol/L and 10 mmol/L were 12±3% and 25±10% of maximum, respectively; responses during glibenclamide/TEA/4-AP treatment, 13±2% and 50±13% of maximum, respectively; n=4).
Effects of a NOS Inhibitor Alone and in Combination With Ba2+
L-NAME (30 μmol/L) caused similar constriction of the basilar artery in WKY and SHR (WKY: baseline, 229±11 μm; L-NAME, 200±12 μm; Δ=−13±2%, n=5; SHR: baseline, 230±11 μm, L-NAME, 188±8 μm, Δ=−18±2%, n=6). In WKY, vasodilator responses to K+ were unaffected by L-NAME (Figure 3a). Responses to ACh were inhibited by ≈90% in the presence of L-NAME (data not shown), indicating that L-NAME effectively inhibited NO production. Subsequent treatment with L-NAME plus Ba2+ constricted the basilar artery further (−155±15 μm, Δ=−33±3% from control baseline, n=5) and markedly inhibited responses to K+ (Figure 3a), similar to the effect of Ba2+ alone in WKY (Figure 2a). In contrast to WKY studies, in SHR L-NAME inhibited the majority of the response to 5 mmol/L K+ (Figure 3b). However, L-NAME had no effect on the response to 10 mmol/L K+ in SHR (Figure 3b). Combined treatment with L-NAME and Ba2+ in SHR constricted the basilar artery further (to 144±8 μm, Δ=−37±4% from control baseline, n=6) but had no further effect on the response to 5 mmol/L K+ (Figure 3b). The response to 10 mmol/L K+ was substantially inhibited by the combined treatment with L-NAME and Ba2+ (by 65%, Figure 3b), an inhibitory effect that was much greater than that achieved in SHR by either L-NAME (Figure 3b) or Ba2+ treatment (Figure 2c) alone. By contrast, combined treatment with L-NAME and Ba2+ had no effect on responses to the control vasodilator, SNP (baseline, 237±18 μm; L-NAME plus Ba2+-treated, 124±5 μm; P<0.05 versus control; control responses for 0.01 μmol/L and 0.1 μmol/L, 34±12% and 74±8% of maximum, respectively; responses during 30 μmol/L L-NAME plus 100 μmol/L Ba2+-treatment, 40±9% and 69±8% of maximum, respectively; n=4).
Effects of a nNOS Inhibitor Alone and in Combination With Ba2+
TRIM (100 μmol/L), a selective inhibitor of nNOS,13 had no effect on basilar artery diameter in WKY or on vasodilator responses to K+ (Figure 3c). In contrast to WKY, in SHR TRIM treatment slightly increased diameter of the basilar artery (Figure 3d). Importantly, TRIM inhibited the response to 5 mmol/L K+ in SHR by >50% (Figure 3d), similar to the effect of L-NAME (Figure 3b). Also similar to L-NAME, TRIM had no effect on the response to the higher concentration of K+ in SHR (Figure 3d). Combined treatment with TRIM plus Ba2+ had no additional inhibitory effect on the response to 5 mmol/L K+ (Figure 3d). Interestingly (as with L-NAME plus Ba2+), during combined treatment with TRIM and Ba2+ in SHR, vasodilator responses to 10 mmol/L K+ were inhibited to a greater extent (by ≈60%; Figure 3d) than after Ba2+ alone (by ≤40%; Figure 2c). In the same animals, TRIM had no effect on responses to ACh (control responses for 1 and 10 μmol/L, 7±2% and 27±3% of maximum, respectively; responses during TRIM treatment, 13±3% and 33±5% of maximum, respectively; P>0.05 for both).
Neuronal Contributions to the Response to K+ in SHR
Given that our data from the use of TRIM implicated a role for nNOS in K+-induced responses in SHR, we tested whether neuronal activation of nNOS-containing nerves might contribute to K+-induced dilator responses of the basilar artery. Tetrodotoxin (1 μmol/L), a sodium channel blocker that prevents action potential generation in nerves, had no effect on SHR basilar artery diameter (control, 225±8 μm; tetrodotoxin-treated, 225±9 μm, n=5; P>0.05) or on K+-induced vasodilatation (control responses for 5 and 10 mmol/L, 15±5% and 64±17% of maximum, respectively; responses during tetrodotoxin treatment, 19±9% and 64±18% of maximum, respectively; P>0.05 for both).
We next considered that 5 mmol/L K+ may cause small depolarizations of perivascular nerves sufficient to cause activation of N-type Ca2+ channels, thus increasing activity of nNOS and exocytosis of NO. Thus, we tested the effect of the N-type Ca2+ channel blocker, ω-conotoxin GVIA (100 nmol/L). ω-Conotoxin GVIA had no effect on basilar artery diameter in WKY and did not affect vasodilator responses to K+ (Figure 3e). Responses to ACh were unaffected by ω-conotoxin GVIA (control responses for 1 and 10 μmol/L, 11±5% and 33±8% of maximum, respectively; responses during ω-conotoxin-treatment, 10±4% and 27±4% of maximum, respectively; n=5; P>0.05 for both).
Consistent with the effects of L-NAME and TRIM in SHR, ω-conotoxin GVIA inhibited the response to 5 mmol/L K+ in SHR by ≈60% (Figure 3f). Additional treatment with Ba2+ in SHR caused no further inhibition of the response to 5 mmol/L K+, but markedly inhibited responses to 10 mmol/L K+ by ≈70% (Figure 3f).
Expression of nNOS in the Basilar Artery
Immunostaining of whole-mounted basilar arteries with an anti-nNOS antibody showed intense staining of perivascular nerves in both WKY (n=4) and SHR (n=4) (Figure 4). When presented with 4 pairs of nNOS-immunostained arteries (ie, each pair composed from 1 WKY and 1 SHR) in a blinded manner, 4 of 5 observers indicated that in all 4 pairs of vessels the perivascular plexus of NOS-immunoreactive fibers was denser in the SHR than in the WKY basilar artery. The fifth observer reported denser innervation in the SHR vessel of 2 pairs and was unable to discern any difference between WKY and SHR arteries of 2 other pairs. This pattern of results was significantly different (χ2 test for goodness of fit, P<0.05) from a random assignment of vessels to the 3 possible categories (see Methods).
Expression of KIR2.1 in the Basilar Artery
PCR products corresponding to mRNA for KIR2.1 were present in cerebral arteries from WKY and SHR (Figure 5). Increasing the PCR cycle number increased band intensities for KIR2.1 (upper gel) and β-actin (lower gel). Importantly, the intensity of the bands for both genes was similar in the 2 rat strains. These data suggest that expression of KIR2.1 and β-actin in cerebral arteries is similar in WKY and SHR.
Our data indicate that a small but physiological increase in extracellular K+ concentration, from 3 to 5 mmol/L, produces a larger dilator response of the basilar artery in SHR versus WKY. These effects of a seemingly modest alteration in K+ concentration indicate that this is a very sensitive dilator mechanism in the intact cerebral circulation. Most interestingly, unlike in WKY, in SHR these responses are Ba2+-insensitive and may be caused by release of NO from nNOS-containing perivascular nerves, which more densely innervated the cerebral vessels of the hypertensive animals.
Responses to K+ in WKY
Increases in extracellular K+ concentration induced profound concentration-dependent increases in basilar artery diameter of normotensive rats in vivo, consistent with previous findings.3,4⇓ Our findings that Ba2+ constricted the basilar artery and markedly inhibited the dilator responses to K+ in WKY, support the concept that KIR channel activity in vivo normally modulates basal cerebrovascular tone and mediates K+-induced vasodilatation, respectively.4,7–9⇓⇓⇓
Responses to K+ in SHR
We found that within the physiological range, K+ elicited vasodilator responses that were ≈40% larger in SHR than in WKY, whereas responses to a higher concentration of K+ were similar between strains. By contrast, vasodilator responses to the endothelium-dependent agonist, ACh, were impaired in hypertensive rats, and responses to the NO donor, SNP, were preserved, indicating that endothelial cell dysfunction occurs in the basilar arteries of these animals. Augmented responses to physiological levels of K+ may represent a compensatory mechanism to support cerebral perfusion during chronic hypertension.
Despite its lack of inhibition of responses to 5 mmol/L K+ and the relatively weak effect against a higher K+ concentration in SHR, Ba2+ caused even greater constriction of the basilar artery in SHR versus WKY. Furthermore, RT-PCR experiments indicated that expression levels of mRNA for KIR2.1—the KIR channel subtype thought to mediate K+-induced cerebral hyperpolarization and vasorelaxant responses9—were similar in basilar artery from SHR and WKY. Thus, despite expression and function of KIR channels being similar (or greater) in basilar arteries of hypertensive rats, physiological levels of K+ induce slightly greater cerebral vasodilatation, which is surprisingly independent of KIR channel activation.
Evidence for a Role of nNOS-Containing Perivascular Nerves in SHR
Treatment with L-NAME, a NOS inhibitor not selective for any NOS isoform, substantially inhibited the Ba2+-insensitive vasodilator response to 5 mmol/L K+ in SHR. Furthermore, TRIM, a selective inhibitor of nNOS,13 also inhibited these responses in SHR but not in WKY, suggesting an important role for NO generated by nNOS activity. L-NAME, but not TRIM, inhibited vasodilator responses to ACh, consistent with TRIM being a selective nNOS inhibitor.13 The rat basilar artery, like other cerebral arteries, is known to be innervated by nNOS-containing nerves, which can release NO onto vascular muscle.14–16⇓⇓ We confirmed immunohistochemically that basilar arteries from both WKY and SHR are innervated by nNOS-containing nerves. Moreover, density of innervation by nNOS-containing nerves appears to be greater (ie, there is a greater number of these nerves) in basilar arteries from SHR versus WKY. Thus, we considered that these nerves might be the source of the NO that mediates the response to 5 mmol/L K+ in SHR.
A remarkably similar inhibitory effect to that of L-NAME or TRIM was achieved using ω-conotoxin GVIA, a highly selective inhibitor of Ca2+ entry through N-type Ca2+ channels.17 This toxin, a peptide from cone snail venom,18 does not inhibit postjunctional L-type Ca2+ channels, even at concentrations 4 orders of magnitude higher than that required to inhibit transmitter release from sympathetic nerves.19 Thus, it is very likely that the action of ω-conotoxin GVIA was on perivascular nerves and not on vascular smooth muscle or endothelium. We suggest that it is unlikely that this neuronally mediated response to K+ in SHR involved action potential generation, however, because tetrodotoxin did not inhibit the response. The lack of effect of ω-conotoxin GVIA (like TRIM) on ACh-induced vasodilator responses is compatible with this response involving activation of nNOS and not the endothelial isoform of NOS (eNOS).
Proposed Mechanism of Vasodilator Responses to 5 mmol/L K+ in SHR
Taken together, our data suggest that in SHR, increases in extracellular K+ concentration in the physiological range cause depolarization of nNOS-containing nerve endings innervating the basilar artery. This leads to the opening of N-type Ca2+ channels on these nerves, activation of nNOS (a Ca2+-calmodulin dependent enzyme),20 and generation of NO, which then diffuses to the cerebral vascular muscle to cause vasodilatation. Such a mechanism would be expected to be independent of KIR channel activation and is therefore consistent with the K+-induced vasodilator response being Ba2+-insensitive. Because in SHR, but not WKY, responses to a higher concentration of K+ (10 mmol/L) were much more effectively inhibited by a combination of Ba2+ plus either L-NAME, TRIM, or ω-conotoxin GVIA than by treatment with Ba2+ alone, an nNOS-mediated component probably also contributed to the Ba2+-resistant parts of those responses.
The different mechanism of K+-induced vasodilatation present in SHR could be related to the chronic effects of elevated blood pressure, rather than to non–pressure-related strain differences, given that our data suggest no role for NO in K+-induced dilatation of the basilar artery in 2 strains of normotensive rat—the Sprague Dawley4 and WKY (present study). Nevertheless, further studies will be necessary to determine whether the mechanism in SHR is similar in other models of chronic hypertension. This potent nNOS-mediated mechanism may play a compensatory role in regulation of the cerebral circulation during chronic endothelial dysfunction, analogous to reported nNOS-mediated responses occurring in eNOS-deficient mice.21
This study was supported by grants from the National Health and Medical Research Council of Australia (980672), the National Heart Foundation of Australia (GOOM 0659), and National Institutes of Health grants NS-24621 and HL-38901.
- Received November 13, 2001.
- Revision received December 12, 2001.
- Accepted January 17, 2002.
Paulson OB, Newman EA. Does the release of potassium from astrocyte endfeet regulate cerebral blood flow? Science. 1987; 237: 896–898.
Chrissobolis S, Ziogas J, Chu Y, Faraci FM, Sobey CG. Role of inwardly rectifying K+ channels in K+-induced cerebral vasodilatation in vivo. Am J Physiol Heart Circ Physiol. 2000; 279: H2704–H2712.
McCarron JG, Halpern W. Potassium dilates rat cerebral arteries by two independent mechanisms. Am J Physiol Heart Circ Physiol. 1990; 259: H902–H908.
Johnson TD, Marrelli SP, Steenberg ML, Childres WF, Bryan RM Jr. Inward rectifier potassium channels in the rat middle cerebral artery. Am J Physiol Regulatory Integrative Comp Physiol. 1998; 274: R541–R547.
Zaritsky JJ, Eckman DM, Wellman GC, Nelson MT, Schwarz TL. Targeted disruption of Kir2.1 and Kir2.2 genes reveals the essential role of the inwardly rectifying K+ current in K+-mediated vasodilation. Circ Res. 2000; 87: 160–166.
Nguyen TS, Winn HR, Janigro D. ATP-sensitive potassium channels may participate in the coupling of neuronal activity and cerebrovascular tone. Am J Physiol Heart Circ Physiol. 2000; 278: H878–H885.
McCarron JG, Halpern W. Impaired potassium-induced dilatation in hypertensive rat cerebral arteries does not reflect altered Na+,K+-ATPase dilation. Circ Res. 1990; 67: 1035–1039.
McCleskey EW, Fox AP, Feldman DH, Cruz LJ, Olivera BM, Tsien RW, Yoshikami D. ω-Conotoxin: direct and persistent blockade of specific types of calcium channels in neurons but not muscle. Proc Natl Acad Sci U S A. 987; 84: 4327–4331.
Bredt DS, Snyder SH. Isolation of nitric oxide synthetase: a calmodulin-requiring enzyme. Proc Natl Acad Sci U S A. 1990; 87: 682–685.
Meng W, Ayata C, Waeber C, Huang PL, Moskowitz MA. Neuronal NOS-cGMP-dependent ACh-induced relaxation in pial arterioles of endothelial NOS knockout mice. Am J Physiol Heart Circ Physiol. 1998; 274: H411–H415.