Stummer, Walter; Baethmann, Alexander; Murr, Reinhard; Schurer, Ludwig; Kempski, Oliver S.
Background and Purpose: A previous communication of this laboratory demonstrated reduced mortality and neuronal damage by spontaneous locomotor activity preceding forebrain ischemia in Mongolian gerbils. The present experiments seek to elucidate potential mechanisms of protection by measurement of cerebral blood flow, cerebral tissue conductance as an indicator of ischemic cell swelling, and the cerebral release of eicosanoids.
Methods: Gerbils were maintained either in conventional cages (nonrunners) or with free access to running wheels (runners) for 2 weeks preceding 15 minutes of forebrain ischemia. During ischemia and 2.5 hours of reperfusion, cerebral tissue conductance was determined with a two-electrode system. Simultaneously, prostaglandin D2, prostaglandin F2 alpha, and thromboxane B2 were measured in ventriculocisternal perfusate. In additional animals cerebral blood flow was assessed by hydrogen clearance.
Results: Decreases in tissue conductance during ischemia were similar in nonrunners (56 plus minus 3%) and runners (62 plus minus 3%) but normalized more rapidly in runners during reperfusion. In both groups reperfusion was accompanied by marked increases of perfusate prostaglandin D2, prostaglandin F2 alpha, and thromboxane B2. In nonrunners, however, thromboxane B2 was already elevated during ischemia (147 plus minus 9%, P less than .01) and remained elevated longer during recirculation (P less than .05). Postischemic perfusion maxima were higher in runners (70.8 plus minus 7.4 versus 47.0 plus minus 5.0 mL/100 g per minute, P less than .05) and were observed sooner (27.4 plus minus 6.9 versus 62.2 plus minus 12.3 minutes, P less than .05). Both groups displayed delayed hypoperfusion of a similar magnitude (runners, 29.0 plus minus 2.4 mL/100 g per minute; nonrunners, 30.1 plus minus 2.4 mL/100 g per minute).
Conclusions: Protection by preischemic locomotor activity may involve enhanced postischemic reperfusion, leading to more rapid normalization of conductance and thus of cell volume. Enhanced reperfusion may be the consequence of attenuated thromboxane liberation during and after ischemia.
(Stroke. 1995;26:1423-1430.)
Key Words: cerebral blood flow, cerebral ischemia, eicosanoids, locomotion, gerbils
Previous investigations of this laboratory [1] revealed impressive protection from the sequelae of 15 minutes of forebrain ischemia by spontaneous preischemic locomotor activity in Mongolian gerbils maintained with free access to treadmills for 2 weeks preceding the insult. Postischemic survival was 91% at 2 weeks compared with only 37% in controls kept in conventional cages. Furthermore, neuronal death was reduced in selectively vulnerable areas of the brain, such as hippocampus, striatum, and certain laminae of the cortex. The incidence and extent of accompanying thalamic infarction was also reduced. However, these observations did not offer any obvious explanation as to the underlying mechanisms or whether protection was related to the ischemic episode itself or to early or late reperfusion, for example, inhibiting delayed neuronal death evolving days after ischemia. [2]
In the present experiments we investigated the period of ischemia and subsequent 2 hours of reperfusion. During this period we studied electrical tissue conductance of the brain to assess the extent and reversibility of the extracellular to intracellular fluid shifts indicative of ischemic cell swelling. It is generally accepted that changes in the size of extracellular fluid volume are reflected by changes in tissue conductance or its reciprocal, tissue impedance. [3-6] Accordingly, a complete interruption of cerebral blood flow is associated with a decline in tissue conductance of approximately 50%, corresponding to a decrease of extracellular volume from approximately 20% to 10%. [7] On the other hand, the normalization of conductance or impedance indicates a restoration of energy metabolism and ionic pump function, with recovery of cell and extracellular volumes. [8] Measurements of cerebral blood flow (CBF) were included for evaluation of a major determinant of brain function recovery and thus survival. [9-14] CBF was studied by the H2 clearance method, [15-17] allowing repetitive measurements in the same animal. Furthermore, the release of important degradation products of arachidonic acid were investigated, ie, prostaglandin D2 (PGD2), PGF2 alpha, and thromboxane B2 (TXB sub 2), the stable metabolite of TXA2. [18] Various laboratories have demonstrated a rise in eicosanoids after ischemia [19-21] caused by the liberation of unesterified fatty acids, including arachidonic acid, during ischemia [22,23] and their metabolism by cyclooxygenase and lipoxygenase. Eicosanoids have been investigated extensively in connection with cerebral ischemia because some of their physiological effects, such as alterations of platelet function or vascular tone and permeability, [24] may modulate ischemic injury. We examined the release of these substances using ventriculocisternal perfusion of the brain. The technique permits studies of the temporal dynamics of mediator liberation during and after circulatory arrest of the brain, as opposed to the one-time assay of brain tissue homogenates at the end of the experiment. [19,25-28]
Male Mongolian gerbils (60 to 90 g body weight) raised from a stock of animals obtained from Hoechst (Frankfurt, Germany) were divided into two groups. Before ischemia one group was placed in cages with free access to running wheels with a diameter of 34 cm for 14 days (runners). The control group was kept in conventional cages (nonrunners). All gerbils were anesthetized with halothane (1.5% in room air) and maintained at 37 degrees C by a feedback-controlled heating pad. The femoral artery of all gerbils was cannulated for monitoring of systemic blood pressure. Both common carotid arteries were carefully isolated from surrounding tissue and loosely encircled by a 5-0 monofilament ligature with the use of an operating microscope. Care was taken to avoid damage of adjacent vagal nerve fibers. The gerbils were then placed in a stereotactic holder in the prone position for combined ventriculocisternal perfusion and measurement of cerebral impedance (runners, n equals 7; nonrunners, n equals 9) or assessment of CBF (runners, n equals 7; nonrunners, n equals 8). Ischemia was induced by straining the ligature encircling both common carotid arteries with a weight of 15 g at both ends, leading to simultaneous occlusion of both vessels. After 15 minutes ischemia was terminated by cutting the ligature.
For ventriculocisternal perfusion the right lateral ventricle was punctured 2 mm lateral and 1 mm dorsal to the bregma at a depth of 3 mm below the skull surface. For this purpose a stainless steel cannula (0.4-mm diameter) connected to a Statham transducer and perfused by mock cerebrospinal fluid at a rate of 33 mu L/min was lowered stereotactically through a burr hole into the brain. The intrusion of the cannula tip into the ventricle was confirmed by a drop in perfusion pressure. Subsequently, a polytetrafluoroethylene catheter (0.7-mm diameter) was stereotactically introduced into the cisterna magna. The correct position of the catheter was verified by an ascending, pulsating column of cerebrospinal fluid. All gerbils in which perfusate was macroscopically contaminated by blood were excluded from the study. If patent ventricular perfusion was not achieved at the first attempt, only impedance curves were registered (n equals 4 in both groups).
Perfusate was collected continuously through the cisternal cannula in preweighed Eppendorf vials cooled to 1 degree C by a copper coil perfused with ice water. The vials were pretreated with 20 mu L indomethacin (1%) per 200 mu L perfusate. Two samples of 200 mu L each were collected before ischemia, two during ischemia (at 6 and 13 minutes), and nine during 120 minutes of reperfusion. Perfusate samples were immediately frozen in liquid nitrogen. Concentrations of PGE2, PGF2 alpha, and TXB2 were determined by radioimmunoassay with antisera obtained from Pasteur Diagnostics.
Cerebral conductance, which is the reciprocal of impedance, Z, was calculated from tissue resistance, R; capacity, C; and frequency, f as Equation 1
Resistance and capacity were determined with a two-electrode system with the use of a resistance-capacitance bridge and an alternating current of 1000 Hz at an effective voltage of 10 mVeff. [29] Measurements were obtained at 30-second intervals during ischemia and every minute during reperfusion. The perfusion cannula in the lateral ventricle of the brain was simultaneously used as an impedance electrode. A second electrode was implanted stereotactically in the temporal cortex 4 mm above the right external acoustic meatus to a depth of 4 mm beneath the skull surface. Both electrodes were insulated with epoxy except for 1 mm at the tip.
CBF was measured by hydrogen clearance. [15,16] For each measurement hydrogen gas was added to the inspiration gas mixture for a final concentration of 12% and maintained for 2 minutes. Tissue hydrogen was detected by four platinum electrodes implanted in frontoparietal and parietal cortices. The 2-mm-long electrodes were made of 75-micro meter-diameter platinum wire insulated by polytetrafluoroethylene with the insulation removed at the tip (0.75 mm). With these dimensions the exposed platinum surface of the electrodes penetrated the cortex of the gerbils but not the underlying white matter, as verified histologically in foregoing experiments. An electrode bias of plus 300 mV was produced by a modified circuitry for current-to-voltage conversion constructed by one of the authors (R.M.) according to Prazma et al. [30] The circuitry generated and applied a constant bias voltage to the platinum electrodes. A built-in differential amplifier subtracted bias voltage from the electrode signal, yielding an output voltage proportional to the current generated by oxidation of H2 at the electrode. A low-pass filter with a cutoff frequency of 0.1 Hz was used to diminish current noise. Calculation of blood flow based on the Fick principle and simplifying assumptions [15] was performed with the aid of a personal computer and software developed by one of the authors (R.M.). The H2 clearance curves had a large monoexponential contour allowing unequivocal calculation of a first-order time constant. Measurements were performed before and after ischemia at 15-minute intervals but not during ischemia because of distortions from a fluctuating baseline, probably secondary to spreading depression. [31]
Differences between nonrunners and runners were analyzed by the Mann-Whitney U test. For intraindividual comparisons the Wilcoxon-Wilcox test was used. All data are given as mean plus minus SEM. Differences were considered significant with an alpha error probability of less than 5% (two-tailed).
No differences were registered in the courses of systemic blood pressure of gerbils kept in conventional cages (nonrunners) and those with free access to running wheels (runners). Figure 1 displays the mean arterial blood pressure recorded in both groups. On occlusion of the common carotid arteries, mean arterial blood pressure rose sharply in both groups from a control value of approximately 85 mm Hg to more than 110 mm Hg. After release of occlusion, blood pressure returned to preischemic values and remained close to 80 mm Hg in both groups until the end of the experiment 160 minutes after ischemia.
On induction of ischemia, conductance declined slowly at first. With a delay of approximately 150 seconds, however, a large conductance drop was observed, so that conductance was only approximately 60% at the end of ischemia. No statistical differences in conductance were registered between runners and nonrunners during ischemia. After reestablishment of perfusion, conductance in nonrunners remained low and slowly began to normalize after 100 minutes of reperfusion. After 160 minutes of reperfusion, values were still reduced to approximately 80% of control. In runners, however, decreased conductance resolved rapidly after release of the ligature, reaching 90% of normal after only 80 minutes of reperfusion Figure 2.
The individual conductance courses were evaluated with respect to the following values: latency, the time period between vessel occlusion and a conductance drop to less than 90% of control; min, the minimal conductance measured during ischemia and reperfusion; and tmin, the time span between vessel occlusion and the time point at which minimal conductance was observed. These data are summarized in Table 1. As demonstrated, latency did not differ significantly between runners and nonrunners. However, minimal conductance was significantly higher in runners and was measured approximately 30 seconds after discontinuation of ischemia in this group, suggesting prompt reversal of ischemic cell volume perturbations. This observation stood in sharp contrast to gerbils without prior locomotor activity. In these animals, the minimum in conductance did not coincide with the end of ischemia but occurred approximately 30 minutes after the vascular occlusion was reopened.
(Table 2) presents the mean baseline levels of PGD2, PGF2 alpha, and TXB2 in runners and nonrunners obtained immediately before ischemia. No differences were noted between nonrunners and runners. Figure 3 demonstrates the mean levels relative to their respective baseline values in the 13 perfusate samples collected from each gerbil. During ischemia no changes were seen for PGD2 or PGF2 alpha. However, nonrunners displayed a significant increase in TXB2 in the sample collected between minutes 6 and 12 of ischemia to 147 plus minus 9% of baseline. This increase was not observed in runners. After ischemia PGD2 and PGF2 alpha concentrations increased approximately threefold compared with baseline in both groups. Maximal concentrations were encountered in the samples collected between 13 and 27 minutes after ischemia, returning to baseline toward the end of the experiment (120 minutes of reperfusion). No statistical differences were noted between runners and nonrunners for these two prostanoids. The same holds for the course of TXB2 during the first 30 minutes of reperfusion. Thereafter, however, TXB2 remained elevated in nonrunners, whereas TXB2 concentrations in the cerebrospinal fluid of runners normalized within 45 minutes of reperfusion, eventually descending below baseline level (P less than .01 versus baseline).
We performed a regression analysis to determine whether there was a relationship between TXB2 release during ischemia and the minimum of conductance (min). As demonstrated in Figure 4 an inverse relationship was observed between both parameters taken from nonrunning gerbils (r equals minus .81, P less than .01). A significant relationship was also found when the data of runners and nonrunners were pooled (r equals minus .70, P less than .01), whereas no correlation was obtained when regression analysis was restricted only to the data of the runners.
The findings on regional CBF are given in Figure 5. Since no differences were observed between the parietal and frontoparietal recordings of each gerbil, the flow data obtained at these sites were averaged for final evaluation. Resting blood flow did not differ between nonrunners (45.61 plus minus 1.7 mL/100 g per minute) and runners (44.5 plus minus 1.4 mL/100 g per minute). During ischemia blood flow measurements were severely disturbed by baseline instability so that no data could be obtained during this period. After ischemia was terminated, accustomed changes in CBF were observed. These changes encompassed a period of impaired reperfusion followed by a transient phase in which CBF returned to and generally exceeded the resting value (postischemic hyperperfusion). Subsequently, blood flow dropped below the control value in a phase of delayed hypoperfusion, returning to normal toward the end of the experiment. However, as shown in the bottom panel of Figure 5, there was a broad variation with respect to the time at which postischemic hyperperfusion was observed. Because hyperperfusion was generally a short-lived phenomenon, calculation of an average of blood flow between different gerbils at a fixed point of time would have mitigated the amplitudes encountered in the single animals. Therefore, the individual curves were analyzed for the value of postischemic perfusion maxima and the point of time at which these occurred. The same analysis was performed with respect to the value of minimal perfusion after the initial perfusion maximum. Figure 5 presents the means of these data, including also values for control blood flow, CBF 5 minutes after ischemia (early reflow), and CBF at the end of the experiment (160 minutes of recirculation). As demonstrated, CBF during early reflow was reduced in both nonrunners (15.9 plus minus 1.9 mL/100 g per minute) and runners (31.2 plus minus 6.7 mL/100 g per minute, P equals NS). Maximal postischemic perfusion did not always surpass resting values in nonrunners and occurred rather late (62.2 plus minus 12.3 minutes of recirculation). Mean values for CBF in this phase exceeded baseline slightly (47 plus minus 5 mL/100 g per minute). However, in runners hyperperfusion evolved much earlier (27.4 plus minus 6.9 minutes) and was far more pronounced (70.8 plus minus 7.4 mL/100 g per minute). Conversely, minimal blood flow during delayed hypoperfusion was identical for both groups, just as the point of time at which minimal blood flow was measured (nonrunners, 30.1 plus minus 2.4 mL/100 g per minute after 82.3 plus minus 9.8 minutes; runners, 29.0 plus minus 2.4 mL/100 g per minute at 80.2 plus minus 11 minutes).
In a previous communication [1] we reported unanticipated and impressive protection against ischemic brain damage by spontaneous locomotor activity in the Mongolian gerbil. Subsequent experiments reported in the present article demonstrate that ameliorated survival may be related to pathophysiological mechanisms operative during ischemia and early reperfusion. This conclusion is based on differences between runners and nonrunners with regard to measurements of cerebral tissue conductance, CBF, and thromboxane liberation.
Electrical tissue impedance, which is the reciprocal of tissue conductance, has frequently been measured for assessment of the shift of extracellular electrolytes and fluid into the intracellular space accompanying cell swelling during experimental ischemia. [3,5,6,32-36] This approach is based on the close relationship between the extracellular to intracellular fluid distribution and electrical tissue impedance under normal and pathophysiological conditions. As a consequence of the high electrical resistance of cell membranes compared with extracellular fluid, a low-frequency alternating current passing through biological tissues is mainly conducted by the extracellular compartment. [37] Nevertheless, quantitative estimates of extracellular volume by measurements of tissue impedance [37] have been criticized because of the structural complexity of brain tissue and inherent methodological difficulties. [4] Therefore, the current assessment of cerebral impedance was limited to a comparison of relative changes.
The short-term course of cerebral tissue impedance (or conductance) on flow interruption is characterized by two different phases. [4-6] Initially, tissue conductance remains relatively stable for 1 to 2 minutes, followed by a rapid fall until a plateau is reached that corresponds to approximately 50% of the preischemic level. The onset of the rapid decline of conductance has been attributed to a negative shift of the extracellular steady potential and to the disappearance of evoked potentials as a consequence of terminal cell depolarization. [3,4] An interesting relationship has been observed for plasma glucose levels and the latency between flow interruption and the rapid decline of conductance. [38] This may indicate a dependency of the initial delay of the impedance decrease on tissue energy reserves before occlusion. The ischemic threshold for the beginning impedance rise has been reported at 9.6 mL/100 g per minute in monkeys after middle cerebral artery occlusion. [32] Accordingly, failure of normalization has been related to an absent recovery of neuronal function, [4] to blood-brain barrier disruption, and to tissue damage assessed by elevated Naplus-Kplus ratios. [6] Consequently, normalization of conductance and thus of cell volume implies that water and ions are transported across the cellular membrane, indicating functional recovery. [8]
In the present investigation we encountered no differences in the initial delay or subsequent rapid conductance decrease during ischemia in gerbils with or without prior spontaneous locomotor activity Table 1. Moreover, the decreases in conductance registered were of a magnitude observed by other researchers during cerebral ischemia or cardiac arrest. [3,5,6,34,36,39] Therefore, we conclude that ischemia was severe and similar in gerbils with or without preischemic running. However, recovery of conductance after reopening of the carotid arteries was enhanced in runners, indicating early recuperation of membrane function and a normalization of extracellular space and cell volume.
A factor conceivably contributing to enhanced recovery with regard to cell volume was better and earlier reperfusion in gerbils with preischemic wheel running. As demonstrated in Figure 4 the dynamic profile of postischemic blood flow in runners displayed characteristic features, namely, a brief impairment of reperfusion followed by hyperperfusion and subsequent delayed hypoperfusion. [9-14] Five minutes after restoration of perfusion, blood flow was still distinctly impaired in both nonrunners and runners. Shortly thereafter, CBF in runners clearly exceeded that of nonrunners in a period of hyperperfusion. These differences could not be explained by changes in systemic arterial blood pressure, which was identical for both groups.
Impaired reperfusion has been attributed to a variety of factors, such as increased blood viscosity, endothelial and perivascular glial swelling, [40] endothelial blebs protruding into the capillary lumen, [41] platelet aggregation, [42,43] and vessel obstruction by leukocyte attachment to the endothelial surface. [44,45] Perfusion deficits after ischemia may result in a prolongation of cerebral hypoxia, thus contributing to cerebral damage. Accordingly, adequate postischemic reperfusion has been associated with restoration of brain electrical activity. [46,47] Other researchers have shown that even long periods of ischemia can be survived provided that impaired reflow is prevented. [48,49] The present investigation corroborates these observations, because marked and early hyperperfusion was witnessed in the group of gerbils with superior survival and reduced neuronal damage in selectively vulnerable brain regions. [1] Conceivably, the earlier recovery of brain tissue conductance in runners was similarly the consequence of enhanced reflow.
On the other hand, the magnitude and time course of a delayed decrease in perfusion appeared not to have been affected by preischemic locomotor activity. Generally considered to increase ischemic brain damage, the mechanisms underlying the phenomenon of delayed hypoperfusion are not yet fully understood. [41,45,50,51] Our own data suggest that under the present experimental circumstances prompt and sufficient reperfusion may have been more crucial than delayed hypoperfusion for ameliorating the outcome.
Enhancement of postischemic reflow in runners may be attributed to attenuated release of thromboxane during and after ischemia, as determined by ventriculocisternal perfusion. By this method, time-course determinations are possible in individual animals in contrast to studies in which tissue levels of eicosanoids have been determined in brain homogenates during and after ischemia. [19,25-28] In confirmation of these studies and of others using microdialysis, [52] cerebroventricular perfusion, [21] or an ex vivo analysis of eicosanoid levels in the supernatant of brain slices, [20] we found increased cerebral release of TXB2, PGD2, and PGF2 alpha in the postischemic brain. PGD2 and PGF2 alpha peaked at 27 minutes of reperfusion, and TXB2 was already elevated at 20 minutes. No differences were observed between runners and nonrunners for PGD2 and PGF2 alpha, and the production of these two eicosanoids during ischemia was not stimulated. On the other hand, TXB2 release was higher in nonrunners compared with runners in perfusate collected during ischemia. Furthermore, TXB2 levels remained elevated longer in nonrunners during reperfusion. In view of the potent vasoconstrictor and platelet aggregatory properties of thromboxane, [24,53] the poor and delayed reperfusion in nonrunning gerbils might have been related to an increased production of thromboxane in the brain during and after ischemia. This conclusion is consistent with former evidence suggesting that a suppression of thromboxane formation by cyclooxygenase or thromboxane synthase blockers enhances postischemic CBF [12,26,54] and reduces neuronal death in the CA1. [55] This hypothesis is underscored by the close relationship between TXB2 release during ischemia and the severity of the conductance decrease observed in the present experiments (see Figure 3).
We see no evident explanation for the differences in thromboxane liberation during and after ischemia. Possibly, residual blood flow in the brains of nonrunners during ischemia may have allowed for metabolism of arachidonic acid because of the availability of molecular oxygen, [56] increasing thromboxane in the perfusate in this group. Unfortunately, CBF measurement by hydrogen clearance during ischemia was not possible because of an instability of the electrode current during this period. This instability correlated well with changes in cortical steady potentials (data not shown), which displayed a rapid negative shift approximately 1.5 seconds after onset of ischemia, followed by a slower negative shift throughout the period of occlusion.
However, forebrain ischemia in the gerbil is recognized as being exceptionally dense, [13,14] and the changes in electrical conductance in both groups that were registered during ischemia were similar and of a magnitude comparable to those in other studies of global ischemia. [3,5,6,34,36,38] Also, residual perfusion during ischemia in one or both groups of gerbils would also have resulted in increased metabolism of arachidonic acid by cyclooxygenase to PGD2 and PGF2 alpha. [57] This was not the case, so ischemia of equal density in both groups of gerbils can be assumed.
An alternate explanation may be derived from observations by Eichner [58] that physical exercise modifies the ratio of prostacyclin to thromboxane in an "antithrombotic'' direction by increased endothelial liberation of tissue plasminogen activator and prostacyclin. Consequently, activation of platelets and hence aggregation during ischemia [42,43] may be suppressed, attenuating thromboxane synthesis by platelets that are capable of thromboxane production. [18,24] By this reasoning, however, higher perfusate levels of thromboxane in nonrunners may only be secondary to platelet activation, and platelet aggregation may be the explanation for reflow deficits observed in this group. Interestingly, thromboxane levels in runners dropped significantly below baseline levels in the course of reperfusion. This observation may indicate downregulation of synthesis or enhanced clearance of thromboxane by an intrinsic mechanism activated by ischemia. However, it is an open question by what basic mechanism preischemic, spontaneous locomotor activity affects the release of thromboxane and enhances reperfusion. Wheel running may influence a number of other pathophysiological processes that may potentially be involved in conferring the observed protection, for instance, the release of endogenous opioids and hormones. [59,60] In addition, environmental changes, which were an intricate part of the present experiments, may themselves alter brain weight and metabolism [61] and thus possibly influence outcome after ischemia. Nevertheless, the present experiments suggest that physical exercise is the factor offering protection against the sequelae of cerebral ischemia in the gerbil.
In conclusion, our data are consistent with the hypothesis that markedly improved survival and protection of the brain against ischemic damage by preischemic physical activity may be attributable to the better quality of early postischemic blood flow. Tentatively, we contend that thromboxane mediates reflow deficits because we observed significantly lower levels of thromboxane liberation in a group of gerbils with good reperfusion; rapid functional recovery, as indicated by conductance measurements; and better outcome, as demonstrated in a previous report. [1] However, the basic mechanism by which preischemic, spontaneous locomotor activity affects the release of thromboxane and enhances reperfusion remains unknown. Because physical exercise offers unprecedented protection against the consequences of ischemia in the gerbil, we believe that further research into this intriguing phenomenon is warranted in the hope of establishing new treatment strategies for cerebral ischemia.
This work was supported by grant DFG Ba 452/6-7. It contains substantial elements of the doctoral thesis of one of the authors (W. Stummer). We gratefully acknowledge the excellent technical assistance of Hilde Lainer and Ulrike Gorke.
1. Stummer W, Weber K, Tranmer B, Baethmann A, Kempski O. Reduced mortality and brain damage by locomotor activity in gerbil forebrain ischemia. Stroke. 1994;25:1862-1869. Library Holdings [Context Link]
2. Kirino T. Delayed neuronal death in the gerbil hippocampus following ischemia. Brain Res. 1982;239:57-69. [Context Link]
3. Van Harreveld A, Ochs S. Cerebral impedance changes after circulatory arrest. Am J Physiol. 1956;187:180-192. [Context Link]
4. Hossmann K-A. Cortical steady potential, impedance and excitability changes during and after total ischemia of cat brain. Exp Neurol. 1971;32:163-175. Full Text Library Holdings [Context Link]
5. Matsuoka Y, Hossmann K-A. Cortical impedance and extracellular volume changes following middle cerebral artery occlusion in cats. J Cereb Blood Flow Metab. 1982;2:466-474. Library Holdings [Context Link]
6. De Boer J, Klein HC, Postema F, Go KG, Korf J. Rat striatal cation shifts reflecting hypoxic-ischemic damage can be predicted by on-line impedance movements. Stroke. 1989;20:1377-1382. Library Holdings [Context Link]
7. Hossmann K-A. Development and resolution of ischemic brain swelling. In: Pappius HM, Feindel W, eds. Dynamics of Brain Edema. Berlin, Germany: Springer-Verlag; 1976:219-227. [Context Link]
8. Hansen AJ. Effects of anoxia on ion-distribution in the brain. Physiol Rev. 1985;65:101-148. Library Holdings [Context Link]
9. Rehncrona S, Abdul-Rahman A, Siesjo BK. Local cerebral blood flow in the postischemic period. Acta Neurol Scand. 1979;60(suppl 72):294-295. [Context Link]
10. Hossmann K-A. Treatment of experimental cerebral ischemia. J Cereb Blood Flow Metab. 1982;2:275-297. Library Holdings [Context Link]
11. Siesjo BK. Cerebral circulation and metabolism. J Neurosurg. 1984;60:883-908. Library Holdings [Context Link]
12. Stevens MK, Yaksh TL, Hansen RB, Anderson RE. Effect of preischemic cyclooxygenase inhibition by zomepirac sodium on reflow, cerebral autoregulation, and EEG recovery in the cat after global ischemia. J Cereb Blood Flow Metab. 1986;6:691-702. Library Holdings [Context Link]
13. Nadasy GL, Greenberg JH, Reivich M, Kovach AGB. Local cerebral blood flow during and after bilateral carotid occlusion in unanesthetized gerbils. Stroke. 1990;21:901-907. Library Holdings [Context Link]
14. Kato H, Araki T, Kogure K, Murakami M, Uemura K. Sequential cerebral blood flow changes in short-term cerebral ischemia in gerbils. Stroke. 1990;21:1346-1349. Library Holdings [Context Link]
15. Auckland K, Bower BF, Berliner RW. Measurement of local blood flow with hydrogen gas. Circ Res. 1964;14:164-187. [Context Link]
16. Fieschi C, Bozzao L, Agnoli A. Regional clearance of hydrogen as measure of blood flow. Acta Neurol Scand. 1965;41(suppl 14):46-52. [Context Link]
17. Pasztor E, Symon L, Dorsch NWC, Branston NM. The hydrogen clearance method in assessment of blood flow in cortex, white matter, and deep nuclei of baboons. Stroke. 1973;4:556-567. Library Holdings [Context Link]
18. Hamberg M, Svensson J, Samuelsson B. Novel transformations of prostaglandin endoperoxides: formulation of thromboxanes. Adv Prostaglandin Thromboxane Res. 1976;1:19-27. Library Holdings [Context Link]
19. Gaudet RJ, Alan I, Levine L. Accumulation of cyclooxygenase products of arachidonic acid metabolism in gerbil brain during reperfusion after bilateral common carotid artery occlusion. J Neurochem. 1980;35:653-658. Library Holdings [Context Link]
20. Kempski O, Shohami E, von Lubitz D, Hallenbeck JM, Feuerstein G. Postischemic production of eicosanoids in gerbil brain. Stroke. 1987;18:111-119. Library Holdings [Context Link]
21. Stevens MK, Yaksh TL. Time course of release in vivo of PGE sub 2, PGF sub 2 alpha, 6-keto-PGF sub 1a and TXB sub 2 into the brain extracellular space after 15 min of complete global ischemia in the presence and absence of cyclooxygenase inhibition. J Cereb Blood Flow Metab. 1988;8:790-798. Library Holdings [Context Link]
22. Bazan NG. Free arachidonic acid and other lipids in the nervous system during early ischemia and after electroshock. Adv Exp Med Biol. 1976;72:317-335. Library Holdings [Context Link]
23. Yoshida S, Masuhiro I, Busto R, Santiso M, Martinez I, Ginsberg M. Cerebral phosphoinositide, triacylglycerol, and energy metabolism in reversible ischemia: origin and fate of free fatty acids. J Neurochem. 1986;47:744-757. Library Holdings [Context Link]
24. Moncada S, Vane JR. Arachidonic acid metabolites and interactions between platelets and blood-vessel walls. N Engl J Med. 1979;300:1142-1147. Library Holdings [Context Link]
25. Crockard HA, Bhakoo KK, Lascelles PT. Regional prostaglandin levels in cerebral ischemia. J Neurochem. 1982;38:1311-1314. Library Holdings [Context Link]
26. Black KL, Hoff JT, Deshmukh GD. Eicosapentaenoic acid: effect on brain prostaglandins, cerebral blood flow and edema in ischemic gerbils. Stroke. 1984;15:65-69. Library Holdings [Context Link]
27. Dempsey RJ, Roy MW, Meyer K, Cowen DE, Tai HH. Development of cyclooxygenase and lipooxygenase metabolites of arachidonic acid after transient ischemia. J Neurosurg. 1986;64:118-124. Library Holdings [Context Link]
28. Dorman RV. Effects of cerebral ischemia and reperfusion on prostanoid accumulation in unanesthetized and pentobarbital-treated gerbils. J Cereb Blood Flow Metab. 1988;8:609-612. Library Holdings [Context Link]
29. Van Harreveld A, Murphy T, Nobel KW. Specific impedance of rabbit's cortical tissue. Am J Physiol. 1963;205:203-207. [Context Link]
30. Prazma J, Smith D, Jochem WJ. Current-to-voltage converter for measurement of oxygen. J Appl Physiol. 1978;44:977-980. Library Holdings [Context Link]
31. Lauritzen M. Long-lasting reduction of cortical blood flow of the rat brain after spreading depression with preserved autoregulation and impaired CO sub 2 response. J Cereb Blood Flow Metab. 1984;4:546-554. Library Holdings [Context Link]
32. Branston NM, Strong AJ, Symon L. Impedance related to local cerebral blood flow in cerebral cortex. J Physiol (Lond). 1978;275:81P-82P. Full Text Library Holdings [Context Link]
33. Schuier FJ, Hossmann K-A. Experimental brain infarcts in cats, II: ischemic brain edema. Stroke. 1980;11:593-601. Library Holdings [Context Link]
34. Holder DS. Detection of cerebral ischemia in the anaesthetised rat by impedance measurement with scalp electrodes: implications for noninvasive imaging of stroke by electrical impedance tomography. Clin Phys Physiol Meas. 1992;13:63-75. Full Text Library Holdings [Context Link]
35. Williams CE, Gunn A, Gluckman PD. Time course of intracellular edema and epileptiform activity following prenatal ischemia in sheep. Stroke. 1991;22:516-521. Library Holdings [Context Link]
36. Korf J, Klein HC, Venema K, Postema F. Increases in striatal and hippocampal impedance and extracellular levels of amino acids by cardiac arrest in freely moving rats. J Neurochem. 1988;50:1087-1096. Library Holdings [Context Link]
37. Cole KS, Li C-L, Bak AF. Electrical analogues for tissues. Exp Neurol. 1969;24:459-473. Full Text Library Holdings [Context Link]
38. Siemkowicz E, Hansen AJ. Brain extracellular ion composition and EEG activity following 10 minutes ischemia in normo- and hyperglycemic rats. Stroke. 1981;12:236-240. Library Holdings [Context Link]
39. Sobotka P. Impedance changes of the brain cortex during and after complete ischemia. Act Nerv Sup. 1971;13:250-254. [Context Link]
40. Ames A, Wright RL, Kowada M, Thurston JM, Majno G. Cerebral ischemia, II: the no-reflow phenomenon. Am J Pathol. 1968;52:437-453. Library Holdings [Context Link]
41. Chiang J, Kowada M, Ames A II, Wright RL, Majno G. Cerebral ischemia, III: vascular changes. Am J Pathol. 1968;52:455-476. Library Holdings [Context Link]
42. Dougherty JH, Levy DE, Weksler BB. Experimental cerebral ischemia produces platelet aggregation. Neurology. 1979;29:1460-1465. Library Holdings [Context Link]
43. Hossmann V, Hossmann K-A, Tagaki S. Effect of intravascular platelet aggregation on blood recirculation following prolonged ischemia of the cat brain. J Neurol. 1980;222:159-170. Library Holdings [Context Link]
44. Hallenbeck JM, Dutka AJ, Tanishima T, Kochanek PM, Kumaroo KK, Thompson C, Obrenovitch TP, Contreas TJ. Polymorphonuclear leukocyte accumulation in brain regions with low flow during the early post-ischemic period. Stroke. 1986;17:246-253. Library Holdings [Context Link]
45. Grogaard B, Schurer L, Gerdin B, Arfors KE. Delayed hypoperfusion after incomplete forebrain ischemia in the rat: the role of polymorphonuclear leukocytes. J Cereb Blood Flow Metab. 1989;9:500-505. Library Holdings [Context Link]
46. Miller CL, Lampard DG, Alexander K, Brown WA. Local cerebral blood flow following transient cerebral ischemia, I: onset of impaired reperfusion within the first hour following global ischemia. Stroke. 1980;11:534-541. Library Holdings [Context Link]
47. LaManna JC, Crumrine RC, Jackson DL. No correlation between cerebral blood flow and neurologic recovery after reversible total cerebral ischemia in the dog. Exp Neurol. 1988;101:234-247. Full Text Library Holdings [Context Link]
48. Hossmann K-A, Zimmermann V. Resuscitation of monkey brain after 1 h complete ischemia, I: physiological and morphological observations. Brain Res. 1974;81:59-74. Full Text Library Holdings [Context Link]
49. Hossmann K-A, Grosse Ophoff B. Recovery of monkey brain after prolonged ischemia, I: electrophysiology and brain electrolytes. J Cereb Blood Flow Metab. 1986;6:15-21. Library Holdings [Context Link]
50. Takagi S, Cocito L, Hossmann K-A. Blood recirculation and pharmacological responsiveness of the cerebral vasculature following prolonged ischemia in cat brain. Stroke. 1977;8:707-712. Library Holdings [Context Link]
51. Dietrich WD, Busto R, Ginsberg MD. Cerebral endothelial microvilli: formation following global forebrain ischemia. J Neuropathol Exp Neurol. 1984;43:72-83. Library Holdings [Context Link]
52. Patel MP, Drummond JC, Mitchell MD, Yaksh TL, Cole DJ. Eicosanoid production in the caudate nucleus and dorsal hippocampus after forebrain ischemia: a microdialysis study. J Cereb Blood Flow Metab. 1992;12:88-95. Library Holdings [Context Link]
53. Moskowitz MA, Coughlin SR. Basic properties of the prostaglandins: current concepts of cerebrovascular disease-stroke. Stroke. 1981;12:696-701. Library Holdings [Context Link]
54. Pettigrew LC, Grotta JC, Rhoades HM, Wu KK. Effect of thromboxane inhibition on eicosanoid levels and blood flow in ischemic rat brain. Stroke. 1989;20:627-632. Library Holdings [Context Link]
55. Nakagomi T, Sasaki T, Kirino T, Tamura A, Noguchi M, Saito I, Takakura K. Effect of cyclooxygenase and lipooxygenase inhibitors on delayed neuronal death in the gerbil hippocampus. Stroke. 1989;20:925-929. Library Holdings [Context Link]
56. Shohami E, Rosenthal J, Lavy S. The effect of incomplete ischemia on prostaglandin levels in rat brain. Stroke. 1982;13:494-499. Library Holdings [Context Link]
57. Gaudet RJ, Levine L. Effect of unilateral common carotid artery occlusion on levels of prostaglandins D sub 2, F sub 2 alpha and 6-keto-prostaglandin-F sub 1a in gerbil brain. Stroke. 1980;11:648-652. Library Holdings [Context Link]
58. Eichner E. Platelets, carotids and coronaries: critique on anti-thrombotic role of antiplatelet agents, exercise and certain diets. Am J Med. 1984;77:513-523. Full Text Library Holdings [Context Link]
59. Fraioli F, Moretti C, Paolucci D, Alicicio E, Crescenci F, Fortinio G. Physical exercise stimulates marked concomitant release of beta-endorphin and adrenocorticotropic hormone (ACTH) in peripheral blood in man. Experientia. 1980;36:987-989. Library Holdings [Context Link]
60. Shyu BC, Andersson SA, Thoren P. Endorphin-mediated increase in pain threshold induced by long-lasting exercise in rats. Life Sci. 1982;30:833-840. Full Text Library Holdings [Context Link]
61. Gonzalezlima F, Ferchmin PA, Eterovic VA, Gonzalezlima EM. Metabolic activation of the brain of young rats after exposure to environmental complexity. Dev Psychobiol. 1994;27:343-351. Library Holdings [Context Link]