INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE MECHANISMS by which an increase in neuronal activity generates a rapid increase in cerebral blood flow (CBF) are not completely understood. One simple explanation is that increased oxygen consumption leads to production of tissue hypoxia and consequent production of vasodilator mediators such as adenosine, lactic acid, and extracellular potassium ions. With visual activation in the cat, a rapid increase in cortical deoxyhemoglobin content (33) and a decrease in microvascular PO₂ (48) precede the increase in CBF (32), thereby implying that tissue hypoxia initially occurs as oxygen consumption increases. However, after a delay of a few seconds, the percent increase in CBF exceeds the percent increase in oxygen consumption, resulting in a decrease in oxygen extraction and in deoxyhemoglobin concentration (32). Indeed, the mapping of patterns of neuronal activation by functional magnetic resonance imaging is based on a decrease in paramagnetic deoxyhemoglobin. Also, studies on humans with positron emission tomography generally support a proportionally greater increase in CBF than in oxygen consumption (18, 19). To prevent a decrease in tissue PO₂, which would ordinarily accompany even a small increase in oxygen consumption, a disproportionately large increase in CBF may be required to increase the capillary-tissue PO₂ gradient sufficient to increase oxygen flux across the capillaries (10). If the increase in CBF restores tissue PO₂, then regulation of local CBF during the steady state by neuronal activity probably depends on a feedforward pathway and not solely on a high gain-negative feedback loop, in which tissue PO₂ provides the only error signal.

Nitric oxide (NO) may represent one potential feedforward pathway. Activation of glutamate receptors generates NO, which can diffuse from neurons to vascular smooth muscle and activate guanylyl cyclase. Although inhibitors of NO synthase (NOS) can attenuate the cortical vascular response to sensory stimulation, the attenuation is, at most, ~50% (14, 15). In addition, functional hyperemia is not impaired in knockout mice with neuronal or endothelial NOS gene deletion (8, 31), and NO may play more of a modulatory role than a mediatory role (30). Products of cyclooxygenase-2 metabolism have been implicated, but cyclooxgenase-2 inhibition or gene deletion does not completely eliminate the flow response (36). Thus other pathways also must be involved in coupling CBF to neuronal activation.

One potential pathway is an astrocyte-based epoxygenase pathway that converts arachidonic acid into epoxyeicosatrienoic acids (EETs) (6, 24). Cytochrome P-450 2C11 possesses epoxygenase activity and has been identified in glial cell cultures (4). Exposure of glial cells to glutamate mobilizes arachidonic acid and increases EET formation (3). Application of EETs to cerebral vascular smooth muscle opens calcium-sensitive potassium channels and produces hyperpolarization and dilation (17, 20, 28). Increases in CBF during glutamate and N-methyl-D-aspartate (NMDA) administration are reduced by epoxygenase inhibitors (3, 9). Because astrocyte processes extend to both synapses and small blood vessels, astrocytes connected by gap junctions could serve as a link between neuronal activity and local vasodilation with EETs serving as a paracrine agent released by astrocytes.

To test the role of the epoxygenase pathway in functional hyperemia during physiological sensory stimulation, the cortical blood flow response to whisker stimulation was measured in the rat before and after superfusion of the cortical surface with epoxygenase inhibitors. Two mechanistically distinct inhibitors were tested: 1) N-methylsulfonyl-6-(2-propargyloxyphenyl)hexanamide (MS-PPOH), a selective substrate inhibitor; and 2) miconazole, which acts on the heme moiety of cytochrome P-450 (49). We have previously shown that these agents inhibit the CBF response to NMDA without decreasing NOS catalytic activity (2, 9). In addition, the effect of indomethacin administration was tested because this cyclooxgenase inhibitor has been shown previously to selectively inhibit pial arteriolar dilation to 5,6-EET but not to the other three regioisomers of EETs (17, 28).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experiments were approved by the institutional animal care and use committee. Anesthesia was induced with halothane in male Wistar rats (300-400 g). A tracheostomy was performed, and a polyethylene (PE)-240 tube was inserted into the trachea for mechanical ventilation with 30-40% O₂ and ~1.5% halothane. A femoral artery and femoral vein were catheterized. Rectal temperature was maintained near 37°C. The rats were placed prone in a sterotaxic apparatus. The scalp and temporalis muscle were dissected from the underlying left parietal and temporal bones. The bone was thinned over a 5-mm region located 2-3 mm posterior and 7 mm lateral to the bregma for placement of the laser-Doppler flow probe. Drilling was performed until epidural and pial blood vessels became visible under a stereomicroscope without penetrating the skull. Inhalation of halothane was discontinued, and anesthesia was maintained by -chloralose (50 mg/kg ip plus 40 mg · kg¹ · h¹ iv) for the remainder of the experiment. Epoxygenase inhibitors were administered by subdural superfusion over the somatosensory cortex. A small drill hole was made superior to the laser-Doppler probe site to expose the dura. The dura was pierced, with care taken not to damage the epidural or pial vessels. The tip of a PE-10 catheter was pulled to a diameter of ~200 µm and advanced 1 mm subdurally in a lateral direction. Another drill hole was made inferior to the probe site, and the dura was incised for passive drainage of the superfused fluid. If subdural bleeding was evident, the experiment was discontinued. This subdural superfusion technique was previously used to demonstrate miconazole inhibition of the blood flow response to glutamate (3).

Cortical red blood cell flux was monitored with a laser-Doppler flowmeter (Moor Instruments; Devon, UK) using a flow probe with a tip diameter of 1 mm. The probe was placed in a location over the thinned skull that was devoid of large visible blood vessels and that generated the maximal flow response to whisker stimulation. A drop of mineral oil was applied at the tip of the probe to provide optical coupling with the tissue. The probe position was kept fixed for the entire experiment, and background lighting was not changed.

The whiskers on the right side of the face were mechanically stimulated by placing the whiskers through a plastic screen mesh connected to a solenoid-driven piston with 7 mm of displacement (21). The whiskers were cut to a length of ~3-4 cm. The plane of the mesh was positioned to be approximately orthogonal to the whiskers, and this orientation was held fixed for the duration of the experiment. The whiskers were stimulated at 3 Hz for 60 s. The change in flow was averaged over the entire 60-s period and expressed as a percentage of the baseline flow signal averaged over the previous 60 s. The responses to three trials spaced 5 min apart were averaged at the end of every hour of cortical superfusion for the 3-h duration of superfusion.

Subdural superfusion with artificial cerebrospinal fluid (CSF) started 1 h after completion of the surgery and continued for 3 h at a constant rate of 5 µl/min. The artificial CSF constituents were (in mmol/l) 151 Na⁺, 3 K⁺, 1.25 Ca²⁺, 0.6 Mg²⁺, 134 Cl, 25 HCO, 6 urea, and 3.7 dextrose. In the time-control group (n = 8), the cortex was superfused for 3 h with CSF containing the vehicle (0.5% ethanol) used for MS-PPOH and miconazole. In the MS-PPOH-treated groups, the cortex was superfused for the first hour with vehicle and for the second and third hours with either 5 (n = 8) or 20 µmol/l (n = 8) MS-PPOH. These doses were based on an IC₅₀ of 13 µmol/l in renal microsomes (49). In the miconazole-treated group (n = 8), the cortex was superfused for the first hour with vehicle, for the second hour with 20 µmol/l miconazole, and for the third hour with vehicle. This dose inhibits glutamate-evoked hyperemia (3) without inhibiting NOS activity (2). The 1-h interval for testing blood flow reactivity was based on the data of Irikura et al. (26), who showed that inhibition of NOS and the flow response to whisker stimulation was more effective at 60 than at 30 min after topical application of N-nitro-L-arginine. The additional hour of superfusion with MS-PPOH was included to demonstrate whether there was any additional inhibition of the flow response by this substrate inhibitor in the region subtended by the laser-Doppler signal with longer exposure. In the miconazole experiment, we elected to wash out this reversible inhibitor during the last hour of reperfusion to determine whether vascular reactivity to whisker stimulation recovered.

To test whether indomethacin would prevent MS-PPOH from inhibiting the flow response to whisker stimulation, the response to whisker stimulation was measured before and after MS-PPOH superfusion in the presence of indomethacin in another group of eight rats. After 1 h of superfusion of vehicle, the flow response to whisker stimulation was measured. Indomethacin was then infused intravenously at a dose (10 mg/kg) that has been shown to inhibit pial arteriolor dilation to 5,6-EET (17). Indomethacin was also added to the subdural superfusate (20 µmol/l) to reduce the possibility of washout in the CSF. After the whisker stimulation was repeated, MS-PPOH (20 µmol/l) was added to the superfusate together with indomethacin. The whisker stimulation responses were recorded 1 h later.

Specificity of inhibition of vasodilation by MS-PPOH and miconazole was tested by measuring the blood flow response to the adenosine A₂ agonist 5'-N-ethylcarboxyamide adenosine (NECA) (25, 46). In separate groups of rats, either vehicle (n = 7), 20 µmol/l MS-PPOH (n = 5), or 20 µmol/l miconazole (n = 6) was superfused subdurally. At 1 h of superfusion, 10 µmol/l NECA was added to the superfusate, and the percent increase in laser-Doppler flux signal was measured.

To determine whether MS-PPOH and miconazole inhibit synaptic transmission, somatosensory evoked potentials were measured during electrical stimulation of the foreleg. Evoked potentials were measured with electrical foreleg stimulation rather than whisker displacement because of the accuracy of gating the signal averaging to the electrical stimulus. The foreleg was stimulated with 150-µs pulses of 2-mA current at a rate of 2.9 pulses/s with subcutaneous needle electrodes. A closed cranial window was constructed above contralateral somatosensory cortex by removing skull (~4 mm diameter) and cutting the dura. A silver wire electrode was secured in the window just above the cortical surface for recording the evoked potentials with a reference electrode placed subcutaneously in the anterior midline. The gated average of 128 repetitions was obtained, and the amplitude of the primary cortical wave complex occurring at 12- to 15-ms latency was measured. An average of three trials was obtained for each time point. After baseline measurements were obtained, the window was superfused with either vehicle (0.5% ethanol, n = 5), 20 µmol/l MS-PPOH (n = 5), or 20 µmol/l miconazole (n = 5) for 2 h. The rate was 100 µl/min for the first 20 min, followed by 50 µl/min for the remaining 100 min of superfusion.

Data were analyzed by one-way ANOVA with repeated measures. Comparisons of mean values were made by the Newman-Keuls multiple-range test at the P < 0.05 significance level. Data are presented as means ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There were no significant changes in mean arterial blood pressure, arterial pH, or arterial blood gases over the 3-h superfusion period in any group (Table 1).

Baseline perfusion without whisker stimulation was not significantly changed over the 3-h vehicle superfusion period. Furthermore, there was no significant change in the baseline flow signal during superfusion with MS-PPOH, miconazole, or return of vehicle after miconazole (Fig. 1).

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Laser-Doppler flux signal (means ± SD) during the 60-s baseline period before each whisker stimulation period over the 3-h superfusion period. Compared with baseline flux at 1 h of vehicle superfusion, there was no significant difference in baseline flux during subsequent superfusion with vehicle (n = 8), N-methylsulfonyl-6-(2-propargyloxyphenyl)hexanamide (MS-PPOH, n = 8), or miconazole and return of vehicle (n = 8).

In the time-control group with continuous superfusion of vehicle, whisker stimulation produced a 26% increase in the red blood cell flux signal averaged over the 60-s stimulation period. Three stimulation trials were generated at the end of each hour of superfusion. The coefficient of variation of the evoked increase in the signal (100 SD/mean of the percent response) for the three stimulation trials at each hour averaged 13.4% among the eight rats in the group. There was no significant change in the evoked response at the second or third hour of vehicle superfusion compared with the response at 1 h (Fig. 2).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Percent change in laser-Doppler flux signal (means ± SD) during whisker stimulation in the time-control group (n = 8) superfused with vehicle for 3 h.

An example of the laser-Doppler trace during whisker stimulation after 1 h of vehicle superfusion is shown in Fig. 3. The response was markedly blunted 1 h after subsequent superfusion with 20 µmol/l MS-PPOH. Group data are shown in Fig. 4. With 5 µmol/l MS-PPOH superfusion, the response was attenuated by 28% (from 24.5 ± 9.1% to 17.7 ± 7.6%). With 20 µmol/l MS-PPOH superfusion, the response was attenuated by 69% (from 27.8 ± 8.9% to 8.5 ± 4.1%). There was no further dimunition of the response after an additional hour of MS-PPOH superfusion with either dose.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Trace of laser-Doppler flux signal in rat during 60-s whisker stimulation after 1-h subdural superfusion with vehicle (A) and after 1-h superfusion with 20 µmol/l MS-PPOH (B). Frequency of fluctuations correspond to frequency of mechanical ventilation.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Percent change in laser-Doppler flux signal (means ± SD, n = 8) during whisker stimulation after 1-h superfusion with vehicle and after superfusion with MS-PPOH for 1 h (hour 2) and 2 h (hour 3). A: superfusion with 5 µmol/l MS-PPOH; B: superfusion with 20 µmol/l MS-PPOH. *P < 0.05 from response during vehicle superfusion.

Superfusion of 20 µmol/l miconazole for 1 h also markedly attenuated the increase in the laser-Doppler signal during whisker stimulation (Fig. 5). In the group of eight rats, the response was significantly reduced from 31.3 ± 5.5% to 10.2 ± 1.1%, which represented a 67% attenuation of the response. Subsequent superfusion with vehicle for 1 h largely restored the response to whisker stimulation (Fig. 6).

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Trace of laser-Doppler flux signal in rat during 60-s whisker stimulation after 1-h superfusion with vehicle (A), 1-h superfusion with 20 µmol/l miconazole (B), and subsequent 1-h superfusion with vehicle (C).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Percent change in laser-Doppler flux signal (means ± SD, n = 8) during whisker stimulation after 1-h superfusion with vehicle, 1-h superfusion with 20 µmol/l miconazole (hour 2), and 1-h superfusion with vehicle (hour 3). *P < 0.05 from response after first hour of vehicle superfusion.

To test whether inhibition of the flow response by MS-PPOH required cycloxygenase activity, indomethacin was administered before MS-PPOH in another group of rats. Administration of indomethacin decreased baseline perfusion by 13% (from 220 ± 57 to 192 ± 50 perfusion units), but there was no further change after MS-PPOH superfusion (197 ± 48 perfusion units). Indomethacin administration did not significantly alter the perfusion response to whisker stimulation (Fig. 7). However, the response was attenuated by MS-PPOH after indomethacin administration.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Percent change in laser-Doppler flux signal (means ± SD, n = 8) during whisker stimulation after 1-h superfusion with vehicle, after intravenous infusion of 10 mg/kg indomethacin combined with superfusion of 20 µmol/l indomethacin, and after 1-h superfusion with 20 µmol/l MS-PPOH plus 20 µmol/l indomethacin. *P < 0.05 from response after first hour of vehicle superfusion.

In separate groups of rats superfused with either vehicle, 20 µmol/l MS-PPOH, or 20 µmol/l miconazole, there was no difference in the percent increase in laser-Doppler perfusion in response to the addition of 10 µmol/l NECA to the superfusate (Fig. 8).

View larger version (13K): [in this window] [in a new window]   Fig. 8.   Percent change in laser-Doppler flux signal (means ± SD) during superfusion with 10 µmol/l 5'-N-ethylcarboxyamide (NECA) after 1-h superfusion with either vehicle (n = 7), 20 µmol/l MS-PPOH (n = 5), or 20 µmol/l miconazole (n = 6).

Somatosensory-evoked potential amplitude was measured as a percentage of the baseline response after 1 and 2 h of superfusion of either vehicle, 20 µmol/l MS-PPOH, or 20 µmol/l miconazole. There was no significant change in amplitude over time in any of the three groups (Fig. 9).

View larger version (33K): [in this window] [in a new window]   Fig. 9.   Somatosensory-evoked potential amplitude, expressed as a percentage of baseline, during electrical foreleg stimulation after 1- and 2-h superfusion of vehicle (n = 5), 20 µmol/l MS-PPOH (n = 5), or 20 µmol/l miconazole (n = 5). There was no significant change from 100% in any group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The primary finding of this study is that cortical superfusion of cytochrome P-450 epoxygenase inhibitors markedly reduce functional hyperemia in the whisker barrel cortex in response to vibrissal stimulation in the rat. The magnitude of the reduction in the evoked flow response was 67-69%, which is at least as great as that reported in the literature with other pharmacological probes. For example, others have reported reductions in the response of 40-50% with the NOS inhibitor N-nitro-L-arginine (8, 14, 15, 31), 30% with a guanylyl cyclase inhibitor (30), 40% with theophylline or adenosine deaminase (15), and 40-50% with a cyclooxygenase-2 inhibitor (36). Thus the epoxygenase pathway appears to be at least as important as other potential mediators of functional hyperemia in the cerebral cortex.

We chose a 1-h period of cortical superfusion of inhibitors based on the results of Irikura et al. (26), who showed that a 1-h application of 1 mmol/l N-nitro-L-arginine was more effective than a 30-min application in inhibiting both tissue NOS catalytic activity and the whisker barrel blood flow response. For MS-PPOH, we superfused an additional hour to see whether there would be additional inhibition of the blood flow response. As a substrate inhibitor, inhibition of enzyme activity at submaximal inhibitory concentrations may depend on the duration of exposure as well as the concentration. However, we did not find any additional inhibition of the blood flow response between 1 and 2 h of superfusion with either the 5 or 20 µmol/l dose. Hence, 1 h of superfusion appears adequate for obtaining steady-state inhibition of epoxygenase activity in tissue subtended by the laser-Doppler signal (~1-2 mm depth with an exponential weighting from the surface).

Our conclusions are based on assumptions of the selectivity of MS-PPOH and miconazole as epoxygenase inhibitors. MS-PPOH is a substrate inhibitor of epoxygenase activity with an IC₅₀ of 13 µmol/l in renal microsomes (49). This agent does not inhibit the formation of 20-hydroxyeicosatetraenoic acid (20-HETE) by cytochrome P-450 -hydroxylase when examined at concentrations as high as 50 µmol/l (49). Moderate inhibition of cyclooxygenase was reported at 50 µmol/l MS-PPOH (49). Assuming that the concentration in the underlying tissue is considerably less than the concentration in the infusate, we chose a maximum dose of 20 µmol/l MS-PPOH in the cortical surface superfusate to provide a tissue concentration that would not be too far below the IC₅₀ for epoxygenase activity but well below the IC₅₀ for -hydroxylation and cyclooxygenase pathways. The 28% attenuation of the flow response at 5 µmol/l and the 69% attenuation at 20 µmol/l are consistent with an IC₅₀ of 13 µmol/l MS-PPOH on epoxygenase activity and suggest that the drug is not acting via other arachidonic acid pathways. In addition, we did not observe any inhibition of the NMDA-induced increase in the conversion of labeled arginine to labeled citrulline in microdialysis effluent perfused with 20 µmol/l MS-PPOPH (9). Thus MS-PPOH does not inhibit NOS activity.

Miconazole is a reversible cytochrome P-450 inhibitor with an IC₅₀ of ~0.5 µmol/l for epoxygenase activity and ~5 µmol/l for -hydroxylase activity in renal microsomes (53). However, inhibition of the formation of the vasoconstrictor 20-HETE by -hydroxylase would not readily explain the observed inhibition of vasodilation. Our observation that 1-h superfusion of vehicle largely restored the blood flow response to whisker stimulation after miconazole superfusion is consistent with miconazole acting as a reversible inhibitor that is readily cleared from the underlying tissue. Several nonspecific effects of miconazole have been described, including inhibition of NOS at higher concentrations (16, 27, 34, 47, 51). However, at the 20 µmol/l concentration used in the present study, we previously found no inhibition of NOS activity in cortical homogenates (2) or in vivo using NMDA-evoked increases in the conversion of arginine to citrulline in microdialysates (9). Moreover, the fact that MS-PPOH and miconazole, which inhibit epoxygenase activity by different mechanisms, produce equivalent inhibition of the evoked blood flow response strongly supports a common effect on epoxygenase activity. Furthermore, the observations that MS-PPOH and miconazole did not inhibit the flow response to the adenosine A₂ agonist NECA indicate that that inhibition of vasodilation by these agents was selective. Selectivity is also supported by previous experiments showing that MS-PPOH and miconazole do not inhibit the flow response to nitroprusside (2, 9).

Among the four EET regioisomers, indomethacin selectively inhibits pial arteriolar dilation to 5,6-EET, which can serve as a substrate for cyclooxygenase (17, 28). In the rat, the dilation to 5,6-EET was reduced by superoxide dismutase plus catalase administration, suggesting a role for superoxide generated by cyclooxygenase (17). In the piglet, dilation to 5,6-EET was restored after indomethacin administration by applying as little as 1 pM iloprost (28). Thus, in the piglet, restoration of prostacyclin-receptor activation may act to permit dilation to 5,6-EET rather than a 5,6-EET metabolic product of cyclooxygenase activity mediating the vasodilation. In the present study, the laser-Doppler flow response to whisker stimulation was not consistently reduced by indomethacin, despite using a high intravenous dose of 10 mg/kg plus a topical dose of 20 µmol/l. The systemic dose was similar to that used in studies of pial arteriolar diameter (17, 28). When 20 µmol/l MS-PPOH was superfused after indomethacin administration, the response to whisker stimulation was reduced to 12.5 ± 5.0%, which was similar to the 8.5 ± 4.1% and 10.5 ± 5.3% responses seen after superfusion of MS-PPOH alone for 1 and 2 h, respectively. These results indicate that either cyclooxygenase metabolism of 5,6-EET or a permissive effect of prostacyclin on dilation to 5,6-EET is not a major contributing factor to vasodilation during whisker stimulation. However, segmental differences in the actions of EETs should be considered in interpreting these results. In rat pial arterioles, 5,6-EET is a more potent dilator than other EET regioisomers (17), whereas patch-clamp studies of intraparenchymal vascular smooth muscle indicate that the other EET regioisomers are potent openers of K⁺ channels (4, 20, 24).

The lack of effect of indomethacin on the flow response to whisker stimulation in rats may appear at odds with the decreased response obtained in mice with a cyclooxygenase-2 inhibitor (36), particularly because cyclooxygenase-1 inhibition has no effect on the whisker response (37). However, it is possible that combined inhibition of cyclooxygenase-1 and -2 by indomethacin, or some nonspecific effect of indomethacin, permits redundant pathways to be active in the response that would otherwise be obtunded by selective cyclooxygenase-2 inhibition.

Cytochrome P-450 proteins can be expressed in neurons, but we are not aware of any cytochrome P-450 that possesses epoxygenase activity in cortical neurons. The lack of effect of MS-PPOH and miconazole on somatosensory-evoked potentials suggest that any cytochrome P-450 epoxygenase expressed in neurons is not essential for neurotransmission and that the diminished flow response is unlikely to be the result of decreased synaptic transmission. This conclusion assumes that the lack of effect on evoked potentials with electrical stimulation of the foreleg also holds for the vibrissal sensory pathway. This assumption seems reasonable because there is no a priori reason that EETs would be important in synaptic activity selectively in different somatotopic regions. However, we cannot exclude that EETs may modulate neurotransmission in a population of neurons whose activity is not reflected in the field potentials measured by evoked potentials.

We postulated that EETs may act as a glial-derived hyperpolarizing factor (24) based on the observations that 1) astrocytes generate EETs (6, 42); 2) cytochrome P-450 2C11, which possesses epoxygenase activity, is localized in astrocytes (4); 3) application of glutamate to glial cell culture causes EET formation and release, which are inhibited by miconazole (3); 4) EETs open potassium channels on smooth muscle isolated from cerebral microvessels (4, 20); and 5) EETs relax cerebral arteries and produce cerebral vasodilation (17, 20, 28). In addition, EETs have been identified as a calcium influx factor responsible for replenishing intracellular calcium stores in glia (41) and thus could help assure efficient signaling in the astrocyte syncytium (11-13, 40) for coordinating release of vasoactive mediators.

The rapid flow response to sensory activation requires a rapid release of vasodilator mediators. Because EETs can be stored in membrane phospholipid pools (42, 50), it is conceivable that EETs could rapidly be released and that de novo synthesis is used primarily to replenish these stores. A recent preliminary report (52) using subdural superfusion of radiolabeled EETs showed that localization of label in astrocytes in the whisker barrel cortex was markedly decreased by whisker stimulation, consistent with the concept that neuronal activation leads to release of membrane-bound EETs from astrocytes. The rapid recovery of blood flow after whisker stimulation in the present study (Fig. 3) is also consistent with regulation of a release process rather than solely relying on metabolic clearance of the mediator. In this scenario, EETs may play a more important role in dynamic hyperemic responses than in regulating basal vascular tone. The lack of effect of MS-PPOH and miconazole on basal perfusion in the present study is consistent with this hypothesis. Moreover, others (29, 39) have implicated a role for epoxygenase activity in the hyperemic responses to hypoxia and hypercapnia under specific conditions without a major influence on basal pial arteriolar diameter. Analogous effects have been seen with calcium-sensitive potassium channel inhibitors, which usually do not affect baseline pial arteriolar diameter in vivo but can inhibit specific vasodilator responses (7, 38, 45). On the other hand, decreases in cerebral perfusion have been observed with miconazole superfusion in pentobarbital-anesthetized rats (2). We have no simple explanation for the different effect of miconazole in this study and the present study on chloralose-anesthetized rats other than the different anesthetic.

Other substances implicated in functional hyperemia in the whisker barrel cortex include NO, adenosine, and cyclooxygenase products. Although some laboratories have reported a 40-50% inhibition of the hyperemic response with N-nitro-L-arginine (15, 31), others observed either no significant attenuation of the percent increase in blood flow or only a modest effect dependent on anesthesia (1, 22, 23, 35). Moreover, the hyperemic response is not diminished in neuronal NOS null mice (31) or in endothelial NOS null mice (8). Furthermore, Lindauer et al. (30) found that restoring baseline blood flow levels with NO donors after N-nitro-L-arginine administration or with a cGMP analog after 7-nitroindazole administration restored the hyperemic response to whisker stimulation. Thus a minimal amount of NO and cGMP may be necessary for full expression of the vasodilatory response. In addition to stimulating guanylyl cyclase, NO can inhibit cytochrome P-450 -hydroxylase activity in vascular smooth muscle, leading to decreased 20-HETE formation (5, 43, 44). Because 20-HETE decreases opening of potassium channels, a decrease in 20-HETE formation by NO may permit increased opening of potassium channels by EETs. Thus the role of NO in functional hyperemia is complex.

In summary, our data implicate an important role for EETs in mediating the physiological coupling of cortical blood flow to neural activation. These findings add to the existing array of potential mediators, which can potentially interact in a complex fashion in regulating the integrated vascular response to neural activity.