INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THROMBOXANE A₂ (TxA₂), a labile product of prostaglandin synthase metabolism, is produced in many tissues, including platelets, vascular smooth muscle, lung, leukocytes, heart, and brain (21, 22, 34). Synthesis of TxA₂ increases during conditions that include endotoxemia, anaphylaxis, circulatory shock, myocardial ischemia, atherosclerosis, coronary vasospasm, and cerebral ischemia (5, 18, 20, 23, 24, 38). TxB₂, the relatively stable metabolite of TxA₂ and a useful index of TxA₂ formation, is normally present in plasma in low concentrations (<50 pg/ml) but may rise into the nanogram-per-milliliter range during these conditions (24). The homeostatic responses to these conditions include the activation of hemodynamic and adrenocortical responses, leading us to hypothesize that these responses may in part be triggered by TxA₂. Low rates of intravenous endotoxin infusion into adult sheep result in the elevation of heart rate, blood pressure, cardiac index, plasma ACTH, and TxB₂ concentrations (8, 18, 38). The blood pressure, heart rate, and hypothalamic-pituitary-adrenal responses to small endotoxin infusions cannot be explained by the modest changes in blood gases secondary to TxA₂-induced pulmonary vasoconstriction (29). In addition, the cardiopulmonary responses to endotoxin are prevented by infusions of the TxA₂ receptor antagonist SQ-29548 (18). And, finally, increased formation of TxA₂ in response to peripheral infusions of mineral acid results in elevations of heart rate and hypothalamic-pituitary-adrenal axis activation and these responses are prevented by prostaglandin synthase inhibition (11) or by PGH₂/TxA₂ receptor antagonism (12). Together, these findings suggest that responses to endotoxin are at least in part mediated by TxA₂. Although excessive and generalized formation of TxA₂ sufficient to result in widespread platelet aggregation and significant vasoconstriction of pulmonary and peripheral vascular beds likely has a negative impact on survival and recovery, a discrete production of TxA₂, perhaps in the brain, may perform a protective function by stimulating hemodynamic and adrenocortical responses that would act to preserve tissue perfusion. In this study, we have begun to test this hypothesis by infusing the TxA₂ analog U-46619 directly into the central nervous system in conscious, chronically instrumented sheep. A second series of experiments was performed using lateral ventricle infusions of U-46619 during prostaglandin synthase blockade to determine whether responses to the infusion of U-46619 are altered by de novo formation of TxA₂ or a second prostaglandin synthase metabolite other than TxA₂.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Adult nonpregnant Rambouillet ewes (n = 6) were anesthetized with halothane, and aseptic surgery was performed. Catheters (polyvinyl chloride, 0.050 in. ID, 0.090 in. OD) were advanced from the femoral arteries and veins into the abdominal aorta and vena cava, respectively. The vascular catheters were then tunneled subcutaneously to the flank region and brought through the skin. A cloth pouch was attached to the skin to contain the catheters. A curvilinear skin incision was made to expose the skull, and a 3-mm hole was bored through the skull 1 cm rostral to bregma and 1 cm lateral to midline. A predrilled plastic plate (2 in. × 1 in. ×  in., Vivak; US Plastics, Lima, OH) was fixed to the skull over the hole using stainless steel screws ( in.). An 18-gauge, 2-in. Teflon catheter (Angiocath; Becton Dickinson, Sandy, UT) was passed into the lateral ventricle. The hub of the catheter was removed, and a polyvinyl chloride catheter (0.050 in. ID, 0.090 in. OD) was fixed to the catheter, and the catheter assembly was fixed to the plastic plate using cyanoacrylate glue. Catheters were directed subcutaneously and brought through the skin in the cervical region cranial to the shoulder and a pouch was secured to the skin to protect the catheters. Correct positioning of the lateral cerebral ventricle catheters was determined postmortem by injecting india ink and verifying the location of the ink in the lateral cerebral ventricle after removal of the brain. Animals were treated postoperatively twice daily for 5 days with 25 mg/kg ampicillin trihydrate subcutaneously (Polyflex; Aveco, Fort Dodge, IA) and 2 mg/kg gentamicin sulfate intramuscularly (Gentavet 100; Velco, St. Louis, MO). Animals recovered for 5 days after surgery before experiments were begun. During the surgery recovery period, the animals were habituated to the laboratory on at least 3 different days.

On the day of an experiment, sheep were placed in individual pens. Sheep were studied in pairs to permit them to maintain visual contact with conspecifics so as to avoid anxiety related to separation from herdmates. Studies were designed to allow at least 48 h between experiments. Experiments consisted of a 30-min preinfusion period (30-0 min), a 30-min infusion period, and a 30-min postinfusion period. During the infusion period, sheep received 9,11-dideoxy-9,11 -epoxy-methanoprostaglandin F₂ (compound U-46619; Cayman Chemical, Ann Arbor, MI) at rates of 0, 100 or 1,000 ng · kg¹ · min¹ in artificial cerebrospinal fluid (CSF). Artificial CSF consisted of 152 meq/l Na⁺, 3 meq/l K⁺, 1.6 meq/l Mg²⁺, 25 meq/l HCO₃, 0.5 meq/l PO³₄, and 135 meq/l Cl. Final pH was adjusted to 7.4 using HCl. The volume infusion rate was 0.5 ml/min. All solutions were infused through a 0.22-µm bacteriostatic filter. Each subject received all three treatments to result in a completely repeated-measures design. The treatment order was randomized.

Phasic arterial blood pressure was measured continuously during the 90-min experiments by connecting an aortic catheter to a strain gauge pressure transducer (Isotec; Quest Medical, Allen, TX). The analog voltage output from the transducer was sampled at a rate of 20 Hz using an analog-to-digital converter (DAS 1402; Keithley Metrabyte, Tauton, MA) and a microcomputer. One-minute mean arterial pressure and heart rate values were calculated offline from the digital recordings (Viewdac, Keithley Metrabyte). Blood (6 ml) was collected into chilled polystyrene tubes containing 300 µl of 0.5 M EDTA from the second aortic catheter at the beginning of the control period, at the beginning of the infusion period, and then every 10 min until the end of the postinfusion period. Tubes were kept on ice until the end of the experiment, then centrifuged for 20 min at 2,800 g at 4°C. Plasma was separated and stored in separate aliquots at 20°C. Blood (0.5 ml) for blood gas and pH measurements was collected anaerobically in heparinized 3-ml syringes. Blood gases and pH were measured using a blood gas analyzer (model 330; Radiometer, Westlake, OH). Blood for hematocrit measurements was collected into heparinized microhematocrit tubes.

A second series of experiments was performed using the same subjects. The prostaglandin synthase inhibitor indomethacin (10 mg/kg; Sigma, St. Louis, MO) was infused intravenously over 10 min. At 30 min after the beginning of indomethacin infusion, artificial CSF or artificial CSF plus 1,000 ng · kg¹ · min¹ was infused into the lateral ventricle over 30 min. Phasic arterial pressure was collected continuously over the 30-min infusion and postinfusion periods. Blood samples were collected at the beginning of the infusion period and every 10 min for 1 h.

Plasma ACTH was measured by commercial radioimmunoassay (Incstar, Stillwater, MN). Validation of this assay for use on sheep plasma has previously been described (9). Cortisol was measured using [1,2,6,7-³H]cortisol (Amersham, Arlington Heights, IL) and rabbit anti-cortisol antiserum kindly provided by Dr. Charles E. Wood, University of Florida. This assay has been completely described elsewhere (40). Before assay, cortisol was extracted from plasma using 20 vol of ethanol. Arginine vasopressin (AVP) concentrations were measured using anti-AVP antiserum, kindly provided by Dr. Wood, and ¹²⁵I-labeled AVP (Amersham). Before assay, AVP was extracted from plasma on bentonite. This assay has been completely described elsewhere (28).

Data are presented as means ± SE. For the first study, treatment (artificial CSF, 100 ng · kg¹ · min¹ and 1,000 ng · kg¹ · min¹ U-46619) and time were subjected to repeated-measures two-way ANOVA. For the second study, an a priori decision was made to compare treatment groups from the first study (artificial CSF and 1,000 ng · kg¹ · min¹ U-46619) with treatment groups from the second study (indomethacin + artificial CSF and indomethacin + 1,000 ng · kg¹ · min¹ U-46619) over time using repeated-measures two-way ANOVA. A posteriori analysis was performed using the Student-Newman-Keuls test. Analyses were performed using SigmaStat software (Jandel Scientific, San Rafael, CA). In all cases, the null hypothesis was rejected when P < 0.05. This study was approved by the Institutional Animal Care and Use Committee of Texas A&M University.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Lateral cerebral ventricle infusions of artificial CSF did not result in significant changes in any of the response variables measured. The high U-46619 infusion rate, 1,000 ng · kg¹ · min¹, but not the 100 ng · kg¹ · min¹ infusion rate, resulted in significant elevations in mean arterial pressure, increasing from 93 ± 3 mmHg at the beginning of the infusion period to peak at 133 ± 5 mmHg 9 min after the end of the infusion period (Fig. 1). Heart rate was significantly elevated in response to the low U-46619 infusion rate, 100 ng · kg¹ · min¹, compared with the control infusion group and compared with the control period, increasing from 70 ± 3 beats/min at the beginning of the infusion period to peak at 86 ± 10 beats/min 8 min after the end of the infusion period. Heart rate did not increase in response to the high rate of U-46619 infusion but declined significantly following the end of the infusion period compared with the other treatment groups and compared with the control period in response to the high U-46619 infusion rate; the lowest value was 55 ± 6 beats/min at 27 min postinfusion compared with 69 ± 6 beats/min at the beginning of the infusion period and the peak value of 76 ± 12 beats/min at 17 min of infusion. The largest decrease in heart rate coincided with the peak elevation of mean arterial pressure occurring early in the postinfusion period.

View larger version (36K): [in this window] [in a new window]   Fig. 1.   One-minute mean ± SE values for mean arterial pressure (left) and heart rate (right) during a 30-min preinfusion period (30-0 min), a 30-min infusion period, (0-30 min), and a 30-min postinfusion period (30-60 min). Treatment groups received 0, 100, or 1,000 ng · kg1 · min1 U-46619 (from top to bottom) in artificial cerebrospinal fluid (CSF) administered into the lateral cerebral ventricle. Mean arterial pressure increased significantly only in response to the highest U-46619 infusion rate (bottom left), whereas heart rate increased in response to the 100 ng · kg1 · min1 infusion rate (middle right). Heart rate did not increase in response to the 1,000 ng · kg1 · min1 U-46619 infusion but decreased significantly following the end of the infusion period (bottom right).

The high U-46619 infusion rate, 1,000 ng · kg¹ · min¹, resulted in changes in arterial blood gases (Fig. 2). Arterial partial CO₂ pressure (Pa_CO2) was significantly decreased by 20 min of infusion compared with the control period in response to the highest U-46619 infusion rate and remained significantly lower than control levels through the end of the experimental period. Arterial pH (pH_a) increased significantly in response to the high U-46619 infusion rate. Values were significantly greater than the control period from 20 min of infusion through the end of the experimental period. The increase in pH_a corresponded temporally with the decrease in Pa_CO2. Arterial partial O₂ pressure (Pa_O2) (Table 1) did not change significantly.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Mean ± SE values for arterial pH (pHa; left) and PaCO2 (right) during the preinfusion (30-0 min), infusion (0-30 min), and postinfusion (30-60 min) periods. pHa increased (bottom left) whereas PaCO2 decreased (bottom right) significantly from 20 min until the end of the experimental period in response to 1,000 ng · kg1 · min1 U-46619 infusions into the lateral cerebral ventricle. * Significantly different from control period.

Infusions of U-46619, 1,000 ng · kg¹ · min¹, resulted in hypothalamic-pituitary-adrenal axis activation (Fig. 3). Plasma ACTH concentrations increased significantly compared with the other treatment groups and compared with the control period beginning 10 min after the end of the infusion period and remained elevated through the end of the experimental period. Plasma cortisol concentrations increased in concert with the ACTH responses, evidencing a significant elevation by 20 min postinfusion compared with the control period and compared with other treatment groups. Plasma AVP concentrations (Table 1) did not change significantly.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Mean ± SE plasma ACTH (top) and cortisol (bottom) responses to infusions (0-30 min) of U-46619 (0, 100, and 1,000 ng · kg1 · min1) into the lateral cerebral ventricle. Plasma ACTH and cortisol concentrations increased significantly at 40-60 min and at 50 min, respectively, in response to the highest U-46619 infusion rate. * Significantly different from the control period and from the artificial CSF and 100 ng · kg1 · min1 treatment groups.

The high infusion rate of U-46619, 1,000 ng · kg¹ · min¹, resulted in significant increases in hematocrit (Fig. 4). Hematocrit was significantly increased compared with the control period and the other treatment groups at 40 and 50 min after the beginning of the infusion period, 10 and 20 min, respectively, after the end of the infusion period.

View larger version (53K): [in this window] [in a new window]   Fig. 4.   Hematocrit (mean ± SE) increased significantly in response to 1,000 ng · kg1 · min1 infusions (0-30 min) of U-46619. * Significantly different compared with control period and compared with artificial CSF and 100 ng · kg1 · min1 infusion groups.

In a second series of experiments, subjects were infused with indomethacin before infusions of artificial CSF or U-46619 (1,000 ng · kg¹ · min¹) into the lateral cerebral ventricle were begun. Mean arterial pressure increased significantly over time and compared with indomethacin + artificial CSF in response to indomethacin + U-46619 (Fig. 5). However, the response to U-46619 was significantly reduced by prostaglandin synthase blockade, with a peak value of 133.2 ± 4.7 mmHg without blockade (Fig. 1) compared with 113.4 ± 3.8 mmHg with blockade. Heart rate did not change during the U-46619 infusions following indomethacin but decreased significantly following the end of U-46619 infusion, and responses were not different compared with the group receiving 1,000 ng · kg¹ · min¹ U-46619 without indomethacin.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   One-minute mean ± SE values for mean arterial pressure (left) and heart rate (right) during infusions of indomethacin + artificial CSF (top) and indomethacin + U-46619 (bottom). Mean arterial pressure increased significantly over time and compared with indomethacin + artificial CSF infusions in response to indomethacin + U-46619 infusions (0-30 min). Heart rate did not increase during U-46619 infusions but decreased significantly following the infusion period.

pH_a increased significantly compared with the control period in response to U-46619 + indomethacin (Fig. 6). This response was not different compared with the response to the same dose of U-46619 without indomethacin. Pa_CO2 was significantly decreased compared with the control period in response to U-46619 following indomethacin, and responses were not different from the group receiving the same dose of U-46619 without indomethacin.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Mean ± SE pHa (left)and PaCO2 (right) responses to infusions of indomethacin + artificial CSF (top) and indomethacin + U-46619 (bottom). pHa increased and PaCO2 decreased significantly over time in response to indomethacin + U-46619. * Significantly different between treatments.

Plasma ACTH concentrations did not change in response to indomethacin + artificial CSF but increased significantly over time and compared with indomethacin + artificial CSF in response to indomethacin + U-46619 (Fig. 7). ACTH responses to indomethacin + U-46619 were not different compared with the 1,000 ng · kg¹ · min¹ U-46619 treatment group (Fig. 3). Plasma cortisol concentrations increased significantly over time and compared with indomethacin + artificial CSF in response to indomethacin + U-46619. Cortisol responses were not different when comparing between groups receiving 1,000 ng · kg¹ · min¹ U-46619 with or without indomethacin.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Mean ± SE plasma ACTH and cortisol concentration responses to infusions of indomethacin + artificial CSF and indomethacin + U-46619. ACTH and cortisol increased significantly over time and compared with indomethacin + artificial CSF in response to indomethacin + U-46619. * Significantly different between treatments.

Hematocrit increased significantly compared with the control period in response to indomethacin + U-46619 but did not change significantly in response to indomethacin + CSF (Fig. 8). Hematocrit responses were not different when comparing between groups receiving U-46619, 1,000 ng · kg¹ · min¹, with or without indomethacin (Fig. 4). AVP and Pa_O2 (Table 1) did not change significantly in response to indomethacin or indomethacin + U-46619. Mean preinfusion values were 1.0 ± 0.2 pg/ml and 101.2 ± 4.8 mmHg, respectively.

View larger version (40K): [in this window] [in a new window]   Fig. 8.   Mean ± SE values for hematocrit in response to indomethacin + artificial CSF and indomethacin + U-46619. Hematocrit increased significantly over time in response to indomethacin + U-46619 infusions. * Significantly different between treatments.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The infusion of the TxA₂ analog U-46619 into the lateral cerebral ventricle resulted in a significant elevation in blood pressure. Eicosanoids formed in the brain move from the extracellular space into the CSF, where they are avidly taken up from the CSF by facilitated transport processes located in the choroid plexus (3, 17). It is likely that U-46619 was subject to removal from the ventricular system to the vascular space by the same mechanisms that perform this function on endogenously produced eicosanoids. Therefore, at least a component of the increase in blood pressure was probably in response to direct peripheral vasoconstriction mediated by U-46619. However, a component of the increase in blood pressure was mediated by a direct action of TxA₂ on the brain. Evidence supporting this conclusion is that the magnitude of the increase in mean arterial pressure was more than threefold higher in response to intracerebroventricular infusions compared with previous experiments in which an equivalent dose of U-46619 was administered intravenously to chronically instrumented sheep; mean arterial pressure increased from 93 ± 3 to 133 ± 5 mmHg in response to intracerebroventricular infusions compared with an increase from 83 ± 5 to 95 ± 5 mmHg in response to an equivalent infusion rate of U-46619 administered intravenously (9). The likely mechanism by which centrally acting U-46619 influences mean arterial pressure is an increase in neurally mediated vasomotor tone and not an increase in cardiac output because heart rate was decreased during the period of highest mean arterial pressure.

Infusion of U-46619 into the lateral cerebral ventricle resulted in prompt changes in heart rate. Heart rate was the most sensitive of the dependent variables measured, increasing in response to U-46619 infusions of 100 ng · kg¹ · min¹. At this infusion rate, there was no detectable increase in blood pressure. The increase in heart rate was in response to U-46619 acting on the brain. Evidence supporting this conclusion is that equivalent or greater rates of peripheral infusion of U-46619 result in far smaller increases in heart rate (9). In addition, the heart rate response began promptly after the beginning of the infusions, likely too soon after the beginning of the infusion period for significant amounts of the U-46619 to have left the central nervous system. At the 1,000 ng · kg¹ · min¹ infusion rate, a dose that resulted in large increases in blood pressure, there was no significant change in heart rate during the infusion period. This was followed by a statistically significant fall in heart rate during the postinfusion period. Although the trend downward in heart rate began near the end of the infusion period, the fall did not achieve statistical significance until after the end of the infusion period. The maximum fall in heart rate occurred together with the maximum increase in mean arterial pressure. The absence of a fall in heart rate during the period of infusion in the group receiving the high infusion rate, when blood pressure was greatly elevated, suggests that U-46619 presented a stimulatory action that offset the expected blood pressure-mediated baroreceptor inhibition of heart rate. On the basis of these observations, we conclude that TxA₂ acts on the brain to stimulate heart rate.

Others have investigated the hemodynamic responses to TxA₂ infusions into the lateral cerebral ventricle. Siren et al. (35) found that lateral cerebral ventricle infusions of U-46619 into anesthetized, spontaneously hypertensive rats but not anesthetized, normotensive rats resulted in prompt increases in blood pressure. Heart rate changes were not detected. Our data differ from theirs likely for several reasons. The dose used in their experiments was far higher than ours on a body-weight basis and was delivered as a bolus. As discussed above, the stimulatory actions of TxA₂ on heart rate may be suppressed by baroreceptor-mediated reflex heart rate inhibition. It is possible that they may have observed an increase in heart rate had they utilized lower doses of U-46619. Additionally, they used anesthetized rats whereas we used conscious sheep; differences in species and in the state of consciousness also may account for the differences in findings.

Lateral cerebral ventricle infusion of U-46619 resulted in prompt decreases in Pa_CO2 and concomitant increases in pH_a. Previously, we reported that peripheral infusions of U-46619 into the carotid artery but not into the vena cava resulted in changes in blood gases like those presented in this study (9). On the basis of these findings, we conclude that the actions of U-46619 on pH_a and Pa_CO2 are mediated at the brain. These responses likely were due to changes in respiration because others have demonstrated that increasing de novo production of TxA₂ in the peripheral circulation results in an increase in minute ventilation (27, 31). Others have reported that interruption of vagal nervous transmission in anesthetized cats prevents the TxA₂-stimulated increases in minute ventilation and concluded that TxA₂ acts on pulmonary vagal afferent nerves to mediate the changes in ventilation (32). However, based on our findings, it is also possible that U-46619 infusions act on central respiratory centers and that preventing vagal signal alters a centrally mediated response. Species differences or differences in the state of consciousness also may have been responsible for the differences in responses. Nevertheless, we conclude that TxA₂ acts on the brain to alter pH_a and Pa_CO2 although it is possible that TxA₂ also may act on pulmonary afferent fibers to mediate respiratory responses.

Infusions of U-46619 into the lateral cerebral ventricle resulted in hypothalamic-pituitary-adrenal axis activation. The magnitude of the hypothalamic-pituitary-adrenal axis response was large, resulting in a fourfold increase for ACTH and an 11-fold increase in cortisol concentrations. The finding that both ACTH and cortisol were elevated in response to intracerebroventricular infusions of U-46619 supports the conclusion that TxA₂ acts on the brain to activate the hypothalamic-pituitary axis and does not act directly at the adrenal. This interpretation is supported by a report that U-46619 stimulates the release of corticotropin-releasing hormone from explanted rat hypothalami (2). Further support that the site of action is the brain is provided by previous studies demonstrating that peripheral infusions of U-46619 equivalent to those used in the present study produced far smaller increases in ACTH and cortisol: peak values of 78 ± 38 pg/ml and 32 ± 8 ng/ml in response to intravenous infusions compared with 227 ± 47 pg/ml and 58 ± 36 ng/ml in response to intracerebroventricular infusions for ACTH and cortisol, respectively (9).

The lateral cerebral ventricle infusions of U-46619 in the present study did not result in a change in plasma AVP concentrations. We previously found that peripheral infusions of U-46619 into the carotid artery also failed to increase circulating concentrations of AVP (9). On the basis of these findings, we conclude that TxA₂ does not stimulate magnocellular release of AVP. However, these findings do not eliminate the possibility that AVP may be released from parvocellular neurons in response to U-46619 infusion to potentially play a role in mediating the hypothalamic-pituitary-adrenal axis responses.

Blockade of prostaglandin synthase did not alter the heart rate, pH_a, Pa_CO2, hematocrit, ACTH, or cortisol responses to U-46619 infusions. These findings provide further evidence that responses to U-46619 infusions are mediated by PGH₂/TxA₂ receptor activation and that these responses do not require the formation of a second species of prostaglandin metabolite or de novo formation of TxA₂. However, mean arterial pressure responses to U-46619 were smaller after blockade. This finding suggests that U-46619 infusions resulted in the formation of additional TxA₂. It is possible that U-46619 gaining access to the peripheral circulation activated platelets to increase de novo TxA₂ formation. A second possibility is that U-46619 infusion might result in the formation of a second prostaglandin synthase metabolite in the central nervous system or in the periphery, one other than TxA₂, that might influence blood pressure.

TxA₂ is one of several eicosanoids that have been demonstrated to act on the brain. Chimoskey and co-workers (4, 14) have reported that PGE₂ acts on the brain to mediate increases in heart rate and blood pressure. Although the specific brain site of action was not determined, they found that carotid infusions of PGE₂ act more potently then intracerebroventricular infusions to mediate heart rate and blood pressure responses. In the sheep, the carotid arteries perfuse all of the brain cranial to the obex (1). Therefore, compounds infused into the carotid arteries would be able to access receptors in all regions of the brain cranial to the obex. On the other hand, intracerebroventricular infusions would only bring agonist in contact with receptors within diffusion distance of the brain ventricular system. Structures a short distance from the ventricular system would be exposed to a relatively higher concentrations via the intracerebroventricular infusion route compared with the carotid artery route given equivalent doses, whereas structures at greater distance from the ventricular system would be subjected to lower concentrations. In contrast to PGE₂, U-46619 acts more potently to alter heart rate and blood pressure when infused into the lateral cerebral ventricle compared with the carotid infusion site. These findings support the conclusion that PGE₂ and TxA₂ act at different sites in the brain to stimulate hemodynamic responses. PGE₂ also acts on the brain to mediate hypothalamic-pituitary-adrenal axis responses, as others have reported that microinjections into the preoptic area result in release of ACTH (15) and we have reported that carotid infusions of PGE₂ stimulate the hypothalamic-pituitary-adrenal axis (10). In contrast to TxA₂, PGE₂ acts at the brain to increase magnocellular vasopressin release as intracerebroventricular PGE₂ infusions increase plasma vasopressin concentrations (4). These findings demonstrate that, although both TxA₂ and PGE₂ act on the brain to stimulate hemodynamic and hormonal responses, their respective collection of actions and sites of actions are different, evidence that both TxA₂ and PGE₂ are neurostimulatory or neuromodulatory substances, each with distinct actions on the brain.

Perspectives

Eicosanoids are generated within the brain in response to nonspecific stimuli, ischemia, trauma, and seizures and in response to specific chemical mediators and act on discrete regions of the brain to perform a variety of functions (19, 33, 37, 39). TxA₂ is a major product of arachidonic acid metabolism in normal brain under basal conditions (6, 13, 25, 26, 30, 36, 39). The putative roles of eicosanoids in the brain include neuromodulation, temperature regulation, control of hormone release, control of blood pressure and regulation of cerebral blood flow (7, 16, 36). Our findings that PGH₂/TxA₂ receptor activation at the brain results in blood pressure, heart rate, pH_a, Pa_CO2, hematocrit ACTH, and cortisol responses suggest that TxA₂ may play a physiological role in activating or modulating these responses. We hypothesize that the physiological role of TxA₂ in mediating the responses reported in this study is as a paracrine substance. We estimate that the high infusion rate in the present experiments created U-46619 concentrations in brain tissues in the micromolar range, a concentration that is physiologically achievable in response to local TxA₂ formation. The finding that all but one of the response variables failed to respond to the lower infusion rate (that created less than micromolar concentrations within the brain) suggests that TxA₂ does not act as an endocrine substance because only local formation of TxA₂ would likely produce micromolar concentrations of TxA₂ within brain tissue. Increases in local TxA₂ formation within the brain would require local stimulation from a circulating signal, a neural signal, or changes in brain perfusion to result in changes in local Pa_O2 and pH conditions. We speculate that TxA₂ formation in the brain increases in response to decreases in brain perfusion to result in physiological responses that serve to promote cardiovascular function and thus cerebral perfusion.