Mechanisms of enhanced basal tone of brain parenchymal arterioles during early postischemic reperfusion: role of ET-1-induced peroxynitrite generation

Abstract

The contributions of vasoconstrictors (endothelin-1 (ET-1), peroxynitrite) and endothelium-dependent vasodilatory mechanisms to basal tone were investigated in parenchymal arterioles (PAs) after early postischemic reperfusion. Transient middle cerebral artery occlusion (tMCAO) was induced for 2 hours with 30 minutes reperfusion in male Wistar rats and compared with ischemia alone (permanent MCAO (pMCAO); 2.5 hours) or sham controls. Changes in lumen diameter of isolated and pressurized PAs were compared. Quantitative PCR was used to measure endothelin type B (ET_B) receptors. Constriction to intravascular pressure (‘basal tone') was not affected by tMCAO or pMCAO. However, constriction to inhibitors of endothelial cell, small- (SK) and intermediate- (IK) conductance, Ca²⁺-sensitive K⁺ channels (apamin and TRAM-34, respectively) were significantly enhanced in PAs from tMCAO compared with pMCAO or sham. Addition of the ET_B agonist sarafotoxin caused constriction in PAs from tMCAO but not from sham animals (21±4% versus 3±3% at 1 nmol/L; P<0.01) that was inhibited by the peroxynitrite scavenger FeTMPyP (5,10,15,20-tetrakis (N-methyl-4′-pyridyl) porphinato iron (III) chloride) (100 μmol/L). Expression of ET_B receptors was not found on PA smooth muscle, suggesting that constriction to sarafotoxin after tMCAO was due to peroxynitrite and not ET_B receptor expression. The maintenance of basal tone in PAs after tMCAO may restrict flow to the ischemic region and contribute to infarct expansion.

Introduction

Ischemic stroke is a vascular disorder that has adverse effects on many cell types in the brain. Infarction of neuronal tissue is the primary insult in response to vascular occlusion, however, both microvascular and macrovascular dysregulation occurs during ischemia and reperfusion (I/R) that can exacerbate the primary insult to cause secondary brain injury such as edema and hemorrhage.^{1, 2} In addition, the cerebral vasculature directly establishes the depth and duration of ischemia, and thus has a central role in defining the final infarct.³ In fact, the only effective treatment for ischemic stroke is a vascular one—rapid restoration of cerebral blood flow (CBF) by dissolving or mechanically removing the clot. Furthermore, any neuroprotective therapy will likely depend on a patent and functional vasculature to reach the site of neuronal damage to provide protection. Thus, how the cerebral vasculature responds to I/R injury is an important consideration for stroke outcome and therapy.

The cerebral circulation is a major target of injury during I/R. The primary effect of I/R on pial arteries and arterioles appears to be vasodilatory in nature.^{3, 4} Occlusion of a large artery decreases perfusion pressure in downstream arteries and arterioles, causing myogenic vasodilation in the acute phase of ischemia.^{2, 3} Ischemia also decreases clearance of vasoactive metabolites including carbon dioxide and lactic acid, further promoting vasodilation.³ In addition, postischemic reperfusion causes prolonged vasodilation of pial arteries due to generation of large amounts of reactive oxygen and nitrogen species that damages smooth muscle, causing progressive loss of basal tone and diminished myogenic reactivity.^{5, 6} The vasodilatory effect of I/R on pial vessels likely serves to increase CBF to the ischemic region, however, prolonged vascular paralysis can cause hyperemia during reperfusion that can exacerbate neuronal injury and promote edema formation.⁷ Thus, it appears that similar to neuronal injury, early reperfusion may be beneficial for vascular function but causes more permanent damage after prolonged periods of I/R.

While vasodilation and loss of tone appears to be the primary response of pial arteries and arterioles to I/R, this does not appear to be the case with brain parenchymal arterioles (PAs) that remain constricted after I/R.^{8, 9, 10} Parenchymal arterioles branch off pial arteries at right angles and perfuse the brain parenchyma.¹¹ These long and relatively unbranched vessels have no collateral supply and directly connect the pial vessels to the capillaries. The PAs are also distinct from pial vessels in that they have increased basal tone at lower pressures due to smooth muscle that is more depolarized.^{12, 13} This unique architecture and function makes them high resistance vessels in the brain that have been shown to be the bottleneck to flow within the cortex.¹⁴ Within the middle cerebral artery (MCA) territory, the lenticulostriate vessels are PAs that perfuse the corpus callosum and striatum.¹⁵ Thus, proximal MCA occlusion (MCAO) can affect deep cortical structures that these vessels perfuse. While the high resistance of PAs is likely protective of the microcirculation from high hydrostatic pressure, their lack of vasodilatory response during postischemic reperfusion may limit reperfusion and the delivery of neuroprotective agents to these brain regions.

In the present study, we investigated underlying mechanisms by which I/R affect basal tone of PAs. We focused on early postischemic reperfusion because this time point is clinically relevant for salvaging brain tissue as well as delivery of potential neuroprotective agents. Similar to previous studies, basal tone of PAs was not different after 2 hours of ischemia and 30 minutes of reperfusion compared with sham controls, despite apparent increases in the endothelial vasodilatory influence on tone. We therefore also investigated vasoconstrictive mechanisms in PAs during I/R that may counteract the vasodilatory effects such that basal tone remains the same. Thus, we investigated the influence of endothelin-1 (ET-1) receptor activation and vasoconstriction induced by peroxynitrite in PAs after I/R.

Materials and methods

Animal Model of Postischemic Reperfusion

Experiments were performed using male Wistar rats that were ∼350 to 380 g. All procedures involving animals were approved by the Institutional Animal Care and Use Committee at the University of Vermont and complied with the National Institutes of Health guidelines for the care and use of laboratory animals. Rats were housed in the Animal Care Facility at the University of Vermont, an Association for Assessment and Accreditation of Laboratory Animal Care accredited facility and were allowed food and water ad libitum. Proximal MCAO was performed using the filament technique, as previously described.^{5, 8, 9} Briefly, animals were anesthetized with isoflurane (1.5% in oxygen) and a midline neck incision made to expose the right common carotid artery. A 5-O monofilament coated with silicon was inserted into the internal carotid artery, through the external carotid artery and advanced until it occluded the MCA, verified by laser Doppler. Animals were intubated and mechanically ventilated to maintain blood gases within a normal physiologic range (Table 1). Three groups of animals underwent surgery: (1) animals that were ischemic for 2 hours followed by filament removal to allow reperfusion for 30 minutes and thus had transient MCAO (tMCAO, N=50); (2) animals that underwent 2.5 hours of ischemia alone MCAO (permanent MCAO (pMCAO), N=32) as a control for reperfusion; and (3) animals that underwent a midline neck incision for 2.5 hours but had no filament inserted as a control for anesthesia and surgery (Sham, N=45).

Table 1. Physiologic parameters of animals undergoing MCAO surgery.

	pMCAO (n=6)	tMCAO (n=6)
Blood glucose (mg/dL)	91±3	88±2
Weight (g)	362±11	365±10
Arterial blood gases
pH	7.40±0.01	7.42±0.01
PCO2 (mm Hg)	43±2	42±2
PO2 (mm Hg)	289±5	283±8
rCBF start (LDU)	445±56	431±68
rCBF>MCAO (LDU)	165±44	191±51
% Decrease in CBF	62	56

CBF, cerebral blood flow; LDU, laser Doppler unit; MCAO, middle cerebral artery occlusion; pMCAO, permanent MCAO; rCBF, relative CBF; tMCAO, transient MCAO.

Preparation of Isolated Parenchymal Arterioles and Reactivity Studies

After the appropriate time period of I/R or sham procedure, animals were quickly decapitated under anesthesia and the brain was removed and placed in cold, oxygenated physiologic salt solution (PSS). At the time of euthanasia, trunk blood was collected in plasma heparin tubes for measurement of ET-1. Parenchymal arterioles that branched off the MCA between the M1 and M2 regions were dissected from within the brain parenchyma on the ischemic side of the brain and mounted on glass cannulas within an arteriograph chamber, as previously described.^{8, 9} All vessels were equilibrated for 1 hour at 40 mm Hg. To determine the level of basal tone and the reactivity to pressure (myogenic reactivity), intravascular pressure was increased from 40 to 80 mm Hg in steps of 20 mm Hg and lumen diameter recorded once stable, ∼10 minutes. The level of basal tone was determined by adding papaverine (0.1 mmol/L) and diltiazem (10 μmol/L) to the bath to fully inactivate smooth muscle and calculating the percent tone as the difference between active and passive diameters. All subsequent experiments were conducted at an intravascular pressure of 40 mm Hg.

To determine the influence of the endothelium to basal tone, the change in lumen diameter in separate sets of PAs was compared from tMCAO, pMCAO, or Sham animals that were (1) denuded of endothelium by passing an air bubble through the lumen, confirmed by lack of dilation to the endothelium-dependent vasodilator NS309 (n=10); (2) given a single concentration of the SK channel inhibitor apamin (300 nmol/L; n=18); (3) given a single concentration of the IK channel inhibitor TRAM-34 (1.0 μmol/L; n=18); (4) given a single concentration of the nitric oxide (NO) synthase inhibitor NG-nitro-L-arginine (L-NNA, 0.1 mmol/L; n=18).

To determine the effect of I/R on constriction to endothelin type B (ET_B) receptor activation, PAs from tMCAO or Sham animals were given increasing concentrations of the selective ET_B receptor agonist sarafotoxin (0.01 to 10 nmol/L; n=13). Sarafotoxin was cumulatively added to the bath and diameter in response to each concentration recorded. A separate set of PAs from tMCAO animals was perfused with the peroxynitrite decomposition catalyst FeTMPyP (10 μmol/L; n=12) before performing the sarafotoxin concentration–response curve to determine if constriction to sarafotoxin was due to peroxynitrite generation. The FeTMPyP has been shown to be an efficacious peroxynitrite decomposition catalyst with an effective concentration producing 50% decomposition of 3.5 μmol/L.¹⁶

Finally, to determine if peroxynitrite could constrict PAs similar to MCA,^{5, 17, 18} a separate set of PAs from naïve animals (n=4) was given increasing concentrations of peroxynitrite (0.01 to 3 μmol/L) to the bath by cumulative addition, and the change in diameter at each concentration was recorded.

Isolation of Parenchymal Arteriole Myocytes and Quantitative PCR Analysis of Endothelin Type B Receptors

Separate sets of PAs from tMCAO (n=3) or Sham (n=3) were dissected for myocyte isolation using laser capture microdissection for quantitative real-time PCR analysis of ET_B receptor expression. Briefly, after the MCAO or sham procedure, six to eight PAs from the ischemic side of the brain were dissected and incubated 17 minutes at 37°C in Papain Enzyme Solution (see below for composition of all solutions). The vessels were then transferred and incubated for 10 minutes at 37°C in Collagenase Enzyme Solution. To stop the enzyme digestion, vessels were washed for 15 minutes in ice-cold Enzyme Solution. The vessels were then transferred to 1 U/μL Ribolock in 0.1 mmol/L phosphate-buffered saline and triturated with a fire-polished glass pipette to dissociate the myocytes. The digested PAs were plated on a PALM Duplex dish (Zeiss, Thornwood, NY, USA) and easily identified morphometrically as long spindle-shaped cells. The dish membrane surrounding each myocyte was cut and 10 to 15 myocytes were catapulted onto a PALM Adhesivecap collection tube (Zeiss) and stored at −80°C for mRNA extraction. A whole vessel (MCA) that contained endothelium was used as a positive control.

The microlaser-dissected samples and MCA were directly lysed, made into cDNA and amplified by in vitro transcription following the manufacturer's protocols (MessageBOOSTER cDNA Synthesis from Cell Lysates Kit, Epicentre, Madison, WI, USA). After the in vitro transcription amplification step, the resulting RNA was column purified (RNA Clean & Concentrator; Zymo Research, Irvine, CA, USA) and used for another round of cDNA synthesis. This cDNA was then diluted 1:3 with RNase-/DNase-free water, and 2 μL of this dilution was used per quantitative PCR (qPCR) in a final reaction volume of 20 μL, containing primers, buffer, and water. The qPCR was performed using gene-specific primers against β-actin and EDNRB (Applied Biosystems, Assays on Demand, Grand Island, NY, USA) and Taqman Fast Advanced Master Mix (2 × stock, part #444456). The qPCR was conducted on an Applied Biosystems 7500 Fast thermal cycler using the following run profile: 95°C for 20 seconds (1 × ), 95°C for 3 seconds, 60°C for 30 seconds (45 × ). The upper baseline limit was set at two cycles below the amplification starting point for each target gene; the cycle threshold was determined by the operator for each target gene. Data analysis was performed using SDS software v2.0.6 (Life Technologies, Grand Island, NY, USA).

Measurement of Endothelin-1 Concentrations in Plasma

The ET-1 levels in plasma were measured using a commercially available enzyme-linked immunosorbent assay, according to manufacturer's directions (R&D Systems, Minneapolis, MN, USA). Samples were not diluted.

Drugs and Solutions

All isolated PA experiments were performed using bicarbonate-based Ringer's PSS, the ionic composition of which was (mmol/L): NaCl 119.0, NaHCO₃ 24.0, KCl 4.7, KH₂PO₄ 1.18, MgSO₄ × 7H₂O 1.17, CaCl₂ 1.6, EDTA 0.026, and glucose 5.5. The PSS was made each week and stored without glucose at 4°C. Glucose was added to the PSS before each experiment. The PSS was aerated with 5% CO₂, 10% O₂, and 85% N₂ to maintain pH.

Papain Enzyme Solution contained (mmol/L): 55 NaCl, 5.6 KCl, 80 L-Glutamic Acid, 2 MgCl₂, 10 HEPES, 10 Glucose, 6.5 1,4-dithioerythritol, and 0.5 mg/mL Papain.

Collagenase Enzyme Solution contained (mmol/L): 55 NaCl, 5.6 KCl, 80 L-Glutamic Acid, 2 MgCl₂, 10 HEPES, 10 Glucose, 0.1 CaCl₂, 0.7 mg/mL Collagenase F, and 0.3 mg/mL Collagenase H.

Enzyme Solution used to stop the digestion reaction contained (mmol/L): 55 NaCl, 5.6 KCl, 80 L-Glutamic Acid, 2 MgCl₂, 10 HEPES, 10 Glucose, and 0.1 CaCl₂.

Apamin and TRAM-34 were purchased from Tocris Biosciences (Ellisville, MO, USA) and kept frozen until used. NG-nitro-L-arginine, sarafotoxin, diltiazem, and papaverine were purchased from Sigma (St. Louis, MO, USA) and made as stock solutions. The FeTMPyP was purchased from Calbiochem (Darmstadt, Germany) and made fresh before each experiment. Peroxynitrite was purchased from Calbiochem and stored at −80°C until use. During use, peroxynitrite was kept on ice and in the dark at all times.

Statistical Analysis

All data are presented as mean±s.e.m. Student's t-test or one-way analysis of variance was used to determine differences in tone and percent constriction, with a post hoc Student–Newman–Keuls test for multiple comparisons, where appropriate.

Data Calculations

Vascular tone was calculated as a percent decrease in diameter from the fully relaxed diameter in papaverine and diltiazem by the equation: (1−(ϕ_tone/ϕ_papav)) × 100% where ϕ_tone=diameter of vessel with tone and ϕ_papav=diameter in papaverine and diltiazem. Percent constriction was calculated as a percent change in diameter from baseline by the equation: (1−(ϕ_drug/ϕ_baseline)) × 100% where ϕ_drug=diameter of vessel in apamin, TRAM-34, L-NNA, or sarafotoxin and ϕ_baseline=diameter before giving the drug.

Results

Basal Tone Is Maintained in Parenchymal Arterioles after Ischemia and Reperfusion

It is well established that MCA has decreased basal tone after I/R,^{2, 4, 8} but this does not appear to occur in PAs.^{8, 9, 10} Table 1 shows that there was no difference in physiologic parameters between groups of animals that underwent the MCAO surgery, blood glucose, or body weight. Arterial blood gases were kept within physiologic ranges for these animals as well. In addition, there was no difference in the decrease in CBF between animals that underwent tMCA versus pMCAO. Figure 1A shows the diameters of PAs at 40, 60, and 80 mm Hg from animals that were ischemic for 2 hours with 30 minutes of reperfusion compared with Sham controls and animals that did not undergo reperfusion. There was no difference in the diameter of PAs from any of groups, or in the reactivity to pressure. Figure 1B shows that the percent tone of PAs at 40 mm Hg was also not different between the groups, suggesting that similar to previous studies, neither ischemia nor reperfusion has a significant effect on basal tone of PAs.

Figure 1.

Effect of ischemia and reperfusion on myogenic reactivity and basal tone of parenchymal arterioles (PAs). (A) Graph showing inner diameters in response to increased intravascular pressure of PAs from permanent middle cerebral artery occlusion (pMCAO), transient MCAO (tMCAO), and Sham control animals. There was no effect of ischemia or reperfusion on inner diameter or the response to pressure. (B) Basal tone of PAs from the same groups of animals. Ischemia and reperfusion had no effect on basal tone of PA.

Enhanced Constriction to Small and Intermediate Conductance Calcium-Activated Potassium Channel Inhibition after Ischemia and Reperfusion

The endothelium influences basal tone by production of vasodilators that inhibit tone, most notably NO and endothelium-derived hyperpolarization (EDH).¹¹ Endothelium-derived hyperpolarization is a prominent vasodilator in PAs that appears to be basally active in these vessels and shown to affect CBF.^{11, 19} Previous studies have shown that I/R causes enhanced EDH-induced vasodilation in PAs.⁸ Because SK and IK channels are critical mediators of EDH-induced vasodilation,²⁰ we investigated the effect of I/R on constriction to inhibition of SK and IK channels as well as on constriction to NO synthase inhibition. Constriction in response to all inhibitors was performed in separate experiments, i.e., not cumulative addition. Figure 2 shows that PAs from tMCAO animals had significantly increased constriction to SK and IK channel inhibition with apamin and TRAM-34, respectively, compared with PAs from Sham controls and pMCAO animals. The enhanced constriction to apamin and TRAM-34 in PAs from tMCAO animals was due to reperfusion and not ischemia alone, because PAs from pMCAO animals—that did not have reperfusion—had similar constriction to these compounds as Sham controls. Parenchymal arterioles from all groups of animals had considerable constriction to NO synthase inhibition with L-NNA that was not different between groups. The percent constriction to 0.1 mmol/L L-NNA for PAs from Sham, pMCAO, and tMCAO was 35±4%, 27±4%, and 43±7% n=6/group (P>0.05). These results suggest that the effect of I/R on endothelial vasodilatory responses was selective for the EDH pathway.

Figure 2.

Effect of ischemia and reperfusion on constriction to small- and intermediate-conductance calcium-activated potassium (SK, IK) channel inhibition. (A) Graph showing percent constriction to SK channel inhibition with apamin of parenchymal arterioles (PAs) from permanent middle cerebral artery occlusion (pMCAO) and transient MCAO (tMCAO) animals compared with Sham controls. Parenchymal arterioles from all groups of animals constricted to apamin; however, the constriction was considerably greater after tMCAO, suggesting that reperfusion and ischemia alone affects PA vasoconstriction. (B) Representative tracing of inner diameter of PA after tMCAO in response to apamin. (C) Graph showing percent constriction IK channel inhibition with TRAM-34 of PAs from pMCAO and tMCAO animals compared with Sham controls. Similar to apamin, PAs from all groups of animals constricted to TRAM-34, but was considerably greater after tMCAO. (D) Representative tracing of inner diameter of PA after tMCAO in response to TRAM-34.

While enhanced constriction to SK and IK channel inhibition in PAs from tMCAO animals suggests enhanced EDH, basal tone was not different in those vessels. We next investigated whether there was greater vasoconstriction after tMCAO in the absence of endothelium that may be counteracting the enhanced vasodilatory influence. Thus, we compared the level of tone in PAs from tMCAO and Sham animals that were denuded of endothelium. The level of tone in PAs from tMCAO animals denuded of endothelium was significantly greater than that of PAs from Sham animals that were denuded of endothelium: 57±6% versus 36±4% (n=5/group; P<0.05). Thus, it appears that the vascular smooth muscle after I/R is more contractile that may be counteracting the vasodilatory influence of the endothelium.

Ischemia and Reperfusion-Induced Constriction to Endothelin Type B Receptor Activation due to Generation of Peroxynitrite

From the results above, it appears that there is a vasoconstrictor influence on tone in addition to vasodilator influence in PAs after I/R. We therefore investigated mechanisms by which I/R could be enhancing basal tone in PAs. Endothelin-1 is a potent vasoconstrictor peptide synthesized and released by endothelial cells in the cerebral vasculature.²¹ Endothelin-1 regulates vascular tone through the activation of two specific receptor subtypes, ET_A and ET_B receptors.²² Under physiologic conditions, there is a balance between the vasoconstrictor effect induced by ET_A receptors on vascular smooth muscle and the vasodilator mechanism mediated by activation of ET_B receptors on endothelium to produce NO.²³ A previous study showed that MCA has de novo expression of ET_B receptors on vascular smooth muscle after I/R that elicits vasoconstriction to ET_B receptor activation.²⁴ We therefore tested the hypothesis that selective ET_B receptor activation with sarafotoxin would cause vasoconstriction of PAs from tMCAO but not PAs from Sham controls. Figure 3A shows the concentration–response curve of PAs from tMCAO and Sham animals to sarafotoxin. There was no effect of ET_B receptor activation with sarafotoxin in PAs from Sham animals. However, sarafotoxin constricted PAs from tMCAO animals at concentrations below 3 nmol/L. Increasing the sarafotoxin concentration from 3 to 10 nmol/L caused vasodilation. Thus, similar to what has been shown in MCA after I/R, PAs exhibit de novo constriction to ET_B receptor activation.

Figure 3.

Role of peroxynitrite in endothelin type B (ET_B) receptor-induced vasoconstriction of parenchymal arterioles (PAs) after transient ischemia. (A) Graph showing change in diameter of PAs after transient MCAO (tMCAO) and Sham controls in response to cumulative addition of sarafotoxin. Parenchymal arterioles from Sham animals had no response to sarafotoxin. After tMCAO, PAs constricted at concentrations below 3 nmol/L, after which they dilated. (B) Graph showing change in diameter of PA from tMCAO animals in response to sarafotoxin after inhibition with the peroxynitrite scavenger (5,10,15,20-tetrakis (N-methyl-4′-pyridyl) porphinato iron (III) chloride) FeTMPyP. The vasoactive response to sarafotoxin in PAs after tMCAO was prevented by FeTMPyP, suggesting that ET-1_B activation generates peroxynitrite to cause constriction. (C) Graph showing the change in diameter of naïve PAs in response to increasing concentrations of peroxynitrite. Peroxynitrite caused constriction of PAs. (D) ET-1 levels in plasma from Sham control and after tMCAO measured by enzyme-linked immunosorbent assay (ELISA). There was a significant increase in circulating ET-1 after tMCAO.

Because de novo ET_B receptor expression was only found in vascular smooth muscle after 48 hours of reperfusion, a time period significantly longer than the 30 minutes of reperfusion in this study, we reasoned that this may not be the mechanism by which I/R was causing constriction to sarafotoxin. We therefore tested the hypothesis that peroxynitrite generation in response to ET_B receptor activation in PAs during I/R is responsible for the vasoconstriction. Peroxynitrite generation is likely in PAs during I/R because of increased superoxide production during reperfusion that is readily available to combine with ET_B-induced NO production.^{6, 25} In addition, we and others, have previously shown that peroxynitrite causes vasoconstriction of cerebral arteries.^{5, 17, 18} Figure 3B shows that perfusion of PAs from tMCAO animals with the peroxynitrite decomposition catalyst FeTMPyP prevented the constriction to sarafotoxin. Thus, de novo constriction to ET_B receptor activation with sarafotoxin appears to be due to peroxynitrite generation.

To confirm that peroxynitrite could cause constriction of PAs similar to what we have shown previously in cerebral pial arteries, a peroxynitrite concentration–response curve was generated in isolated and pressurized PAs. Figure 3C shows that increasing concentrations of peroxynitrite caused constriction of PAs and thus may contribute to the constriction induced by sarafotoxin after I/R.

To determine that the duration of I/R used in the current study caused increased ET-1 in plasma that may influence basal tone in vivo, we measured ET-1 levels in tMCAO and Sham controls. Figure 3D shows that ET-1 levels were significantly increased in plasma after I/R.

Lack of Endothelin Type B Receptors on Vascular Smooth Muscle after Ischemia and Reperfusion

To confirm that the constriction to sarafotoxin was not due to de novo expression of ET_B receptors on vascular smooth muscle after tMCAO, we used laser capture microdissection of isolated vascular smooth muscle from PAs taken from Sham or tMCAO animals. Figure 4 shows representative micrographs of isolated myocytes from PAs (Figure 4A), during cutting of myocytes from the membrane (Figure 4B), and after catapulting the myocytes into tubes containing RNase inhibitor (Figure 4C). The qPCR analysis of isolated myocytes found no expression of ET_B receptors (C_T>45). This was not due to low RNA input into the reaction since the housekeeping gene β-actin was found to be expressed in these same myocytes and we could detect ET_B receptor RNA in whole MCA that contained endothelium, albeit at low levels (Figure 4D).

Figure 4.

Endothelin type B (ET_B) receptor (ET_BR) expression in isolated myocytes from parenchymal arterioles (PAs) and middle cerebral artery (MCA). (A–C) Photomicrographs showing isolated myocytes from PAs used for quantitative PCR (qPCR). After myocytes were isolated and identified (A), they were cut using laser capture microdissection (B) and catapulted (C) into an RNase-free tube. (D) Graph showing C_T values for ET_BR and the housekeeping gene β-actin in myocytes from Sham control (CTL) animals and after transient MCAO occlusion (tMCAO) and in whole MCA that contained endothelium. There was no detectable mRNA for ET_BR in myocytes from either group of animals (C_T values of >40 are considered as undetectable). This was not the case for β-actin or whole MCA that showed detectable levels of mRNA.

Discussion

The results of the present study found that PAs exposed to 2 hours of ischemia and 30 minutes of reperfusion had no effect on basal tone or myogenic reactivity in intact vessels. This result is distinctly different from upstream MCA that have progressive loss of tone even with such a short duration of postischemic reperfusion.^{2, 4, 8} Interestingly, the maintenance of basal tone in PAs was associated with enhanced vasodilator and vasoconstrictor mechanisms that appear to counteract each other such that active diameters and tone remain unchanged. For example, PAs from tMCAO animals had increased constriction to SK and IK channel inhibition, suggesting enhanced basal influence of EDH in these vessels. Vasoconstrictor responses also appear to be increased in PAs from tMCAO animals, as shown by increased tone in vessels denuded of endothelium and constriction to ET_B receptor activation. The maintenance of basal tone in PAs during I/R has been reported in other studies with longer durations of I/R and may be a protective mechanism in the brain to limit hydrostatic pressure to the microcirculation, but the increased vascular resistance may also contribute to infarct expansion.^{8, 9, 10}

The SK and IK channels are expressed in cerebral endothelium and necessary for endothelial cell hyperpolarization involved in EDH-mediated dilation.^{20, 26} We have previously shown that under non-ischemic conditions, PAs constrict in response to SK/IK channel inhibition, suggesting that basal EDH in these vessels inhibits tone and controls CBF in the brain cortex.^{8, 19} Under conditions of I/R, the EDH pathway has been shown to be enhanced in both PAs and MCA.^{8, 27, 28} In the present study, increased constriction to SK and IK inhibition with apamin and TRAM-34 of PAs from tMCAO animals further support enhanced EDH-induced vasodilation in PAs after I/R. Increased EDH-mediated vasodilation in PAs after I/R is likely due to increased endothelial cell calcium and greater activation of SK/IK channels because expression of these channels in cerebral vessels is not different after I/R,⁸ but endothelial cell calcium is increased.²⁷ The underlying mechanism by which I/R causes increased endothelial cell calcium is not known; however, we speculate that the transient receptor potential vanilloid 4 (TRPV4) channel is involved in this response. The TRPV4 is highly calcium permeable and shown to increase calcium sparklets that activate SK and IK channels.^{28, 29} In addition, there are several means by which I/R could activate TRPV4 in cerebral endothelium, including cell swelling, increased shear stress, and activation of protein kinase C.^{30, 31, 32} Further studies are needed to determine if I/R-induced TRPV4 activation underlies the increase in SK/IK channel activation and EDH vasodilation in PAs.

An alternative explanation for the increased constriction to apamin and TRAM-34 in PAs after I/R is that it reflects greater smooth muscle depolarization in those vessels, an effect that is dependent on other membrane conductances in both smooth muscle and endothelial membranes. For example, it is possible that PA smooth muscle depolarization is limited by activation of voltage-dependent potassium (K_v) channels. If the properties and densities of these channels are affected by I/R, then the ensuing depolarization caused by SK/IK channel inhibition may not be ‘braked' as effectively. In fact, PAs that were denuded of endothelium had increased basal tone after I/R, suggesting greater smooth muscle depolarization. It is therefore possible that I/R have an effect on other membrane conductances that amplifies the constriction to apamin and TRAM-34 without a change in SK/IK channel density. Understanding if the increased constriction to apamin and TRAM-34 after I/R is a direct effect on SK/IK currents or if the downstream response is amplified without a change in channel properties is beyond the scope of the present study, but potentially important to understand how basal tone is maintained in these vessels.

Parenchymal arterioles after I/R also appear to have enhanced vasoconstriction in response to ET_B receptor activation. Endothelin-1 is a potent vasoactive agent that exerts opposing effects on the vasculature depending on the receptor and cell type involved. The ET-1 activation of ET_A receptors on vascular smooth muscle elicits potent vasoconstriction, whereas ET_B receptor activation on endothelium causes vasodilation through enhanced NO production.^{21, 22, 23} Under normal conditions, ET-1 receptor activation is balanced by ET_A receptors on vascular smooth muscle and ET_B receptors on endothelium. Under certain conditions, ET_B receptors are expressed on vascular smooth muscle and contribute to vasoconstriction.^{24, 33} During cerebral ischemia, ET_B receptors have been shown to become expressed in vascular smooth muscle in MCA, causing vasoconstriction.²⁴ In addition, increased ET-1 levels have been found in plasma and cerebral spinal fluid after cerebral ischemia, suggesting altered ET-1 signaling in the pathogenesis of I/R.^{34, 35} However, the role for ET-1 and its receptors in I/R injury is complex. Both ET_A and ET_B receptor blockade have been shown to be beneficial, as well as produce no effect in preventing damage during ischemic stroke despite improved vascular function.^{36, 37, 38} In the present study, we show that ET_B receptor activation causes constriction of PAs only after exposure to I/R that was not due to de novo expression on vascular smooth muscle, but was prevented by the peroxynitrite decomposition catalyst FeTMPyP. Thus, the constriction to sarafotoxin in PAs after I/R appears to be due to peroxynitrite generation. Other studies have shown that ET-1 administration during I/R in mesentery generates peroxynitrite due to enhanced production of superoxide under these conditions that is available to combine with ET_B-induced NO production.²⁵ In addition, administration of ET-1 to culture neurons causes cell death via peroxynitrite generation, suggesting that this may be another mechanism by which increased ET-1 during I/R is damaging to brain.³⁹

Other studies have found that peroxynitrite is vasoactive and constricts cerebral arteries, leading to the suggestion that its generation may contribute to ischemic injury.^{5, 17, 18} However, this is the first report we are aware of to show that ET_B receptor activation induces peroxynitrite that is constricting of cerebral arterioles, providing a novel aspect to elevated ET-1 during I/R. The mechanism by which peroxynitrite causes constriction of PAs is not clear and may be due to an effect on either endothelium or smooth muscle. Using whole cell recordings of isolated smooth muscle from cerebral arteries, Brzezinska et al¹⁸ found that peroxynitrite inhibits large-conductance calcium-activated potassium (BK_Ca) channels. BK_Ca channels are expressed in vascular smooth muscle and their activation leads to membrane hyperpolarization, inhibition of voltage-gated Ca_2.1 channels, and relaxation.^{12, 40} Inhibition of BK_Ca causes vascular smooth muscle depolarization and constriction of cerebral arteries.⁴⁰ However, PAs are different from pial arteries in that under normal conditions, BK_Ca are uncoupled from calcium sparks and inhibition of BK_Ca does not produce constriction of PAs as in cerebral arteries.¹³ Thus, it is unlikely that peroxynitrite is causing constriction of PAs via inhibition of BK_Ca channels because of this unique excitation–contraction coupling in PA vascular smooth muscle.

The authors declare no conflict of interest.