Factors Limiting Regeneration of ATP following Temporary Ischemia in Cat Brain

Transcription

1 234 STROKE VOL 13, No 2, MARCH-APRIL Yen TT, Powell CE, Pearson DV: An indirect method of measuring the blood pressure of rats without heating. Proc 2nd Int Sym on the SHR. US Dept of HEW, publ No (NIH) , 1976, pp Mueller SM, Ertel PH: The chronic influence of sympathetic nerves offer a protective effect to cerebral vessels of the spontaneously hypertensive rat. (Abstract) Annals of Neurology 10: 104, Mueller SM, Heistad DD, Marcus ML: Effect of sympathetic nerves on cerebral vessels during seizures. Am J Physiol 237: H178-H184, Koletsky S, Shook P, Rivera-Valez J: Absence of hyperactive renal humoral pressor system in spontaneously hypertensive rats. In Okamoto K (ed) Spontaneous Hypertension, Tokyo, Iguku Shoin Ltd., 1972, pp Sen S, Smeby RR, Bumpus FM: Renin in rats with spontaneous hypertension. Circ Res 31: , Carretero OA, Oza NB, Scicli AG, Schork A: Renal tissue kallikrein, plasma renin and plasma aldosterone in renal hyper- tension. Acta Physiol Lat Am 24: , Giese J: Renin, angiotensin and hypertensive vascular damage: a review. Am J Med 55: , Robertson AL, Khairallah PA: Effects of angiotensin II and some analogues on vascular permeability in the rabbit. Circ Res 31: , Fujishima M, Onoyama K, Oniki H, Ogata J, Omae T: Effects of acute hypertension on brain metabolism in normotensive, renovascular hypertensive and spontaneously hypertensive rats. Stroke 9: , Johansson BB, Linder LE: The blood brain barrier in renal hypertensive rats. Clin Exp Hyper 2: , Romero JC, Lazar JD, Hoobler SW: Effects of renal artery constriction and subsequent contralateral nephrectomy on the blood pressure, plasma renin activity, and plasma renin substrate concentration in rabbits. Lab Invest 22: , Gifford RW Jr, Westbrook E: Hypertensive encephalopathy: mechanisms, clinical features, and treatment. Prog Cardiovasc Dis 17: , 1974 Factors Limiting Regeneration of ATP following Temporary Ischemia in Cat Brain FRANK A. WELSH, PH.D., MICHAEL J. O'CONNOR, M.D., VAL ROBERT MARCY, B.S., ANTHONY J. SPATACCO, B.A., AND REBECCA L. JOHNS, B.A. SUMMARY Cerebral ischemia was induced in cats using bilateral carotid artery occlusion coupled with hemorrhagic hypotension. Thirty minutes of ischemia, which depleted levels of ATP and phosphocreatine throughout the cerebral cortex, was followed by 2-4 hours of recirculation. During the recovery period, cortical perfusion and NADH fluorescence were monitored through a cranial window. Postischemic perfusion, as indicated by transit time, was initially higher than control, but declined to subnormal levels by 60 minutes. NADH fluorescence transients, induced by brief anoxia, also decreased steadily during recirculation, indicating a failure of oxidation-reduction capability. The disappearance of anoxic-nadh transients usually preceded the decline of flow, suggesting that 0 2 delivery was not the factor limiting redox reactions. Furthermore, tissue levels of NADH, which were nearly normal after 2-4 hours of recirculation, did not indicate tissue hypoxia. In spite of normalization of NADH, resynthesis of high energy phosphates was severely impaired. The degree of ATP recovery varied widely in different cortical regions; however, there were two general groups of ATP values one at 5% and the other at 70% of control levels. In the energy-depleted areas, NADH levels were normal, but the total pool of NAD (NADH + NAD + ) and the tissue content of K + were 43% lower than control. In contrast, the NAD pool and K + content were only slightly diminished in the regions with greater ATP restitution. The results suggest that postischemic resynthesis of ATP may be limited not by inadequate delivery of 0 2, but rather by defective production of NADH. StrokeVoll3,No2,1982 PERMANENT DISRUPTION of energy metabolism occurs in several models of temporary cerebral ischemia. 1 " 4 Two distinct mechanisms may account for this irreversible energy failure. First, postischemic blood flow may not be adequate for energy restitution even though the capability for ATP resynthesis remains. Alternatively, there may be a primary lesion Division of Neurosurgery, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania Reprint requests should be addressed to Dr. Frank A. Welsh, Division of Neurosurgery, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania This work was supported by U.S. Public Health Service grant NS within the cell which prevents resynthesis of ATP regardless of blood flow. Although postischemic hypoperfusion has been well demonstrated, 6 " 8 it is not known whether flow plays a limiting role in ATP regeneration. If blood flow were delivering insufficient amounts of 0 2, then the resulting tissue hypoxia should be accompanied by increased levels of NADH. However, tissue NADH levels were not higher than normal in regions with permanent energy depletion. 1, 3 In the present study we have measured the time-course of postischemic blood flow and of redox transients of surface NADH fluorescence. In addition, we have determined regional levels of NADH, NAD +, and K + in recirculated brain and correlated these levels with postischemic restitution of ATP.

2 ATP REGENERATION FOLLOWING ISCHEMIA/Welsh et al. 235 Methods Cats, weighing kg, were anesthetized with ketamine (20 mg/kg), immobilized with gallamine, and ventilated mechanically with a mixture of 75% N 2 0 and 25% 0 2. Respiratory rate and volume were adjusted to give an end-tidal C0 2 of 4.5%. Core temperature was maintained at 38 C using a heating lamp and rectal thermistor, and arterial pressure was recorded via a femoral catheter. The common carotid arteries were isolated and encircled with loose-fitting ligatures. The head of the animal was then placed into stereotaxic frame, and a craniotomy (1.5 cm diameter) was made over the parietal cortex of the right hemisphere. The dura was removed, and a glass plate was fastened into the craniotomy at the level of the inner table of the skull. Electrodes were screwed into the skull in a quadrangular array for EEG recording. Cerebral ischemia was initiated by occluding the carotid arteries and simultaneously withdrawing arterial blood into a heparinized syringe. Arterial pressure was lowered until the EEG became isoelectric, and this pressure (25-35 torr) was maintained by withdrawing or reinfusing small quantities of blood. After 30 minutes, ischemia was reversed by releasing the carotid snares and reinfusing the collected blood. During the recirculation period, arterial blood gases and pressure were measured, and systemic acidosis was corrected with intravenous bicarbonate in 22 of the 40 experiments. Cortical perfusion was monitored through the cranial window by recording the cerebrovascular transit of Thioflavin S, a fluorescent, nondiffusible tracer. 9 A fluorometric method was chosen in order to compare perfusion directly with NADH fluorescence in the same volume of cortex. Since both NADH and Thioflavin S absorb light at 366 nm and fluoresce at 450 nm, the cortex was illuminated with ultraviolet light (366 nm) and the 450 nm fluorescence was recorded continuously with a television camera fitted with an image-intensifier. 10 Cortical perfusion was measured during the control period and at intervals of minutes during postischemic recirculation. For each measurement, 0.2 ml of 10 mg/ml thioflavin was injected into the arterial circulation either at the root of the aortic arch via a femoral catheter or into the left atrial appendage through a transthoracic catheter. The cortical transit time was estimated from the fluorescent transient by extrapolating to baseline the steepest portion of the ascending and descending phases of the transient, and measuring the intervening duration. 11 In order to assess the reproducibility of thioflavin injection, a shunt between a femoral artery and vein was inserted in a few animals, and the transit of thioflavin through this shunt was recorded fluorometricahy. The capacity of the cerebral cortex to generate NADH was determined using transient hypoxia. Thus, upon completion of the flow measurement, the transient elevation of cortical NADH fluorescence was recorded during a brief interval (48 sec) of systemic anoxia. Systemic anoxia was produced by replacing 0 2 with N 2 in the ventilation mixture; typically the anoxic increase of cerebral NADH occurred after a delay of seconds. The intensity of 366 nm reflected light was also measured to detect changes in fluorescenceqquenching by hemoglobin. However, during postischemic recirculation, alterations of 366 nm reflectance during transient anoxia were negligible. After 2-4 hours of recirculation, the brain was frozen in situ 12 with liquid N 2 for regional measurements of metabolites. In 3 additional animals, the brain was frozen after 30 minutes of ischemia without recirculation. NADH fluorescence in frozen brain sections was recorded photographically as described previously. 12 The regional distribution of ATP was determined in adjacent thin sections (30 microns) using a modification of the luciferin-luciferase technique. 13 Unfixed, frozen sections were layered onto the luciferin reagent at 15 C, melted at room temperature, and immeditely placed in contact with Polaroid type 52 film for a 10-sec exposure. Although the use of unfixed tissue permits endogenous ATPases to compete with the luciferin reaction, the initial luminescent image was in general agreement with quantitative assay for ATP in regional samples (unpublished data and figures 3 and 4). Regional brain ph was measured qualitatively using the fluorescent ph indictor, umbelliferone. 14 Frozen thin sections were layered at 15 C onto strips of cellulose acetate which had been impregnated with umbelliferone. The layers were melted and iluminated with ultraviolet light (366 nm). The fluorescent image (450 nm), the intensity of which is directly proportional to ph, was recorded photographically on high contrast Polaroid type 51 film. Regional samples, for quantitative measurement of tissue metabolites, were removed from sections of frozen brain used to measure NADH fluorescence with a small cork borer (2 mm diameter) in a 30 C glove box. After weighing (2-5 mg), the samples were extracted by one of two procedures. The first was designed for combined measurements of ATP, phosphocreatine, lactate, NAD +, and NADH. 12 The sample was added to 50 /xl of 0.1 N NaOH in methanol at 30 C, and a suspension was made with a glass stirring rod. In an ice-water bath, 1 ml of 0.01 N NaOH was added, and a 95 [i\ aliquot was made 1 mm in cysteine and heated for 10 min at 60 C to destroy NAD +. The heated aliquot was assayed for NADH. The remaining portion of the tissue suspension was acidified with perchloric acid, centrifuged, neutralized, and analyzed for ATP, phosphocreatine, lactate, and NAD +. The second extraction method was designed for measurement of total tissue K + along with ATP, phosphocreatine, and lactate. The sample was added to 50 ^1 of 0.1 N LiOH in methanol at -30 C, and a suspension was made before transfer to an ice-water bath. After adding 230 fi\ of 0.05 N LiOH, the mixture was heated at 95 C for 5 min to destroy tissue enzymes. LiOH was used rather than NaOH to reduce

3 236 STROKE VOL 13, No 2, MARCH-APRIL 1982 the blank rate of pyruvate kinase activity in the subsequent assay for K +. This extraction method yielded values for ATP, phosphocreatine, and lactate from control brain identical to those obtained by acid extraction. Enzymatic, fluorometric techniques were employed to assay ATP, phosphocreatine, and lactate, as outlined by Lowry and Passonneau. 16 NAD + and NADH were measured with an enzymatic cycling method. 16 K + was determined using the stimulation of pyruvate kinase as described by Outlaw and Lowry. 17 Statistical comparisons were made using Dunnett's modification of the t test. 18 Results Arterial blood gases remained in the normal range during the recirculation period except for a slight increase of pc0 2 at 30 minutes (table 1). In 18 animals (out of 40) in which postischemic bicarbonate infusion was omitted, arterial ph fell to 7.06 ± 0.02 (mean ± SEM) at 30 minutes and recovered to 7.18 ± 0.02 at two hours. Despite this systemic acidosis, there was no difference of metabolite levels in brain between bicarbonate-treated and untreated animals. The arterial pressure declined significantly during recovery, but did not fall below a mean pressure of 100 torr, or the animal was discarded from the series. Thirty minutes of ischemia depleted ATP and phosphocreatine throughout the cerebral cortex (table 2). Lactate was uniformly increased to extremely high levels in all cortical regions analyzed. However, the total content of K + decreased only slightly during the 30-minute insult. Following the period of ischemia, recirculation was measured at 15-minute intervals using the fluorescent tracer, Thioflavin S. Figure 1 illustrates the sequence of thioflavin transients recorded simultaneously from the cortical surface and from a femoral arteriovenous shunt. Thefluorescencetransients in the shunt did not vary during recovery, demonstrating the uniformity of the thioflavin arterial curve following each injection. At 15 minutes of recirculation, the amplitude of the cortical transient was larger than that of the preischemic trace, and the washout was more rapid. By 60 minutes, the cortical transient was similar to the control, but at 150 minutes, the amplitude was markedly diminished. Although the time-course varied in different animals, the gradual decrease of TABLE 1. Arterial Variables Control period Recirculation 30 minutes Recirculation 2 hours Mean pressure (torr) 155 ±3 153 ±4 132* ±5 Po 2 (torr) 110 ±2 118 ±4 122 ±4 Pco 2 (torr) 29.7 ± ± ±0.9 Values are means ± SEM. N = 40 animals, except for ph where N = 22. Different from control period, p < ph 7.35 ± ± ±0.03 TABLE 2. Cortical Alterations during Ischemia Control Ischemia: 30 minutes ATP PCr Lac K ± * ± ± * ± ± * ± ±2 109 ±6 Values are means ± SEM. N = 3 animals per group, 2-6 samples per animal. Values for ATP, PCr (phosphocreatine), Lac (lactate), and K+ expressed as mmol/kg. Different from control, p < perfusion following an initial period of hyperemia was a consistent finding (table 3). Thus, at 15 minutes of recirculation, the transit time was slightly reduced, while the amplitude of the thioflavin transient was moderately increased compared to control values. However, by 60 minutes, transit time had increased by 50%, and the amplitude was no longer elevated. These trends of decreasing amplitude and increasing transit time continued at 120 and 180 minutes of recirculation. However, preceding the decline of perfusion, there was a significant decrease in the amplitude of the anoxic-nadh transient, which is a measure of the capacity of the cortex to generate NADH. Thus, at 15 minutes, the NADH transient was only 59% of control and had diminished to 23% by 60 minutes. At all durations of recirculation, the percentage decrease of the NADH transient was greater than that for blood flow. Indeed, in individual animals, the anoxic-nadh transient disappeared at a time of relative hyperemia (fig. 2). After 45 minutes of recirculation in this animal, the amplitude of the NADH transient was less than 10% of control while cortical perfusion remained markedly increased. Postischemic restoration of brain ATP levels varied greatly in different regions and in different animals. Topographic depiction of ATP revealed focal reductions, which correlated with regional alterations of ph and NADH tissuefluorescence (fig. 3). Regions with reduced ATP (A, C, and D) snowed a remarkable depression of ph and NADHfluorescence.Quantitative measurements in these areas confirmed the decrease of ATP and NADH and demonstrated an extreme elevation of tissue lactate. However, the level of NADH was the same (11 /xmol/kg) in regions A and B despite differences in NADH fluorescence and ATP luminescence. There was a profound decrease of NAD + in regions with impaired metabolic recovery. Regions which failed to reconstitute ATP levels during postischemic recirculation were invariably acidotic, but did not always exhibit diminished NADHfluorescence (fig. 4). In this brain, patches of intense NADHfluorescencewere superimposed upon a darker background in areas depleted of ATP. Despite the increasedfluorescence,the measured value of NADH (sample D) was lower than normal. Thus, the intense tissuefluorescencein this brain was not due to high levels of NADH. As in the previous example, tissue levels of NAD + were markedly decreased in areas with energy failure.

4 ATP REGENERATION FOLLOWING ISCHEMIA/Welsh et al. 237 CEREBRAL CORTEX REFERENCE ARTERY CONTROL PERIOD FLUORESCENCE INCREASE RECIRCULATION^ 15 Minutes -H 5 sec [*- RECIRCULATION : 60 Minutes RECIRCULATION^ 150 Minutes THIOFLAVIN INJECTION I THIOFLAVIN INJECTION FIGURE 1. Thioftavin Fluorescence Transients from Cerebral Cortex and from a Reference A rtery during Postischemic Recirculation. After bolus injection into the arterial circulation, the transit of Thioftavin S was recordedfluorometricallythrough a cranial window (CEREBRAL CORTEX) and simultaneously from a femoral arteriovenous shunt (REFERENCE ARTER Y)prior to cerebral ischemia (CONTROL PERIOD) and at three different times during recirculation. Although the regional restoration of ATP ranged widely in recirculated brain, a frequency histogram of the number of samples at a given ATP value revealed two major groups of samples (fig. 5). In one group, ATP levels were less than 0.5 nmol/kg, while in the other, ATP was resynthesized to levels between 1.3 and 2.0 mmol/kg. In the group of samples with low ATP, lactate remained at end-insult levels, but NADH was slightly lower than control (table 4). NAD + levels were diminished by 43% resulting in an increase of the NADH/NAD + ratio. The total tissue content of K + was also 43% lower than control in regions with ATP less than 0.5 mmol/kg. By contrast, in areas which restored ATP above 1.3 mmol/kg, lactate levels were substantially reduced from end-insult levels, but remained elevated compared to control. Interestingly, NADH levels were 30% greater than control values, and there were minor decrements in TABLE 3. Cortical Thioftavin Fluorescence Transients and Anoxic NADH Fluorescence Transients during Postischemic Recirculation Control Recirculation: 15 min 30 min 60 min 120 min 180 min Amplitude (% of control) 105 ± ±13 72 ±12 55 ±9* 49 ±8* Thioflavin Values are means ± SEM. N = 25 animals for thioflavin, except at 180 min. where N ; 16. N = 9 animals for anoxic NADH. Different from control, p < Transit time (sec) 6.1 ± ± ± ±1.0* 10.4 ±1.1* 12.8 ± 1.4* Amplitude of anoxic increase of NADH (% of control) 59 ± 10* 38 ± 10* 23 ±9* 21 ±8* 17±11*

5 238 STROKE VOL 13, No 2, MARCH-APRIL 1982 THIOFLAVIN FLUORESCENCE NADH FLUORESCENCE CONTROL PERIOD -J \^%^- Fluorescence Increase RECIRCULATION^ 15 Minutes 20 sec RECIRCULATION : 30 Minutes RECIRCULATION^ 45 Minutes THIOFLAVIN INJECTION (- -ANOXIA, 48 see-*] FIGURE 2. Thioflavin Fluorescence Transients and Anoxic NADH Fluorescence Transients in Cerebral Cortex during Postischemic Recirculation. Sequential fluorometric recordings through a cranial window of the transit of Thioflavin and of the anoxic elevation of NADH prior to cerebral ischemia (CONTROL PERIOD) and at three different times during recirculation. ^Hk 'ST V9 0BA» By ATP.5 ^an a! %' ph ) * %.. ^B^'^SS' : ' : >i * ^ NADH v 11 *JPi ''A..Mi : "»2 * fbt A AT P 1.2 PC r 2.7 La c 15.3 REGION R r n F F D V/ U w r IB IP ^ / HI \D NADH II

6 ATP REGENERATION FOLLOWING ISCHEMIA/ Welsh et al. 239 FIGURE 4. Regional Cerebral Metabolites during Postischemic Recirculation. Adjacent coronal sections of frozen brain were processed for topographic depiction of ATP, ph, and NADH and then sampled regionally (A-D) for quantitative determination of metabolites levels. Values for ATP, PCr (phosphocreatine), and Lac (lactate) are given in mmol/kg. Values for NAD + and NADH are given in /xmol/kg. the tissue content of NAD+ and K +. The NADH/ NAD + ratio in regions which resynthesized ATP was comparable to that in energy-depleted areas. Discussion Models of temporary cerebral ischemia permit identification of irreversible alterations which may limit recovery from the ischemic insult. The central aim of the present report was to investigate the factors which interfere with postischemic restitution of brain energy metabolism. Since the initial demonstration of "no-reflow," 5 postischemic blood flow has been implicated as a major factor preventing recovery of brain tissue. More recent reports have confirmed the occurrence of focal hypoperfusion following severe ischemia, 6 " 8 but others have demonstrated significant neuropathology in the absence of impaired flow and have questioned the causative role of postischemic hypoperfusion in the production of ischemic damage. 19 ' 20 In the present study, similar to previous reports, 2123 initial postischemic hyperemia was followed by a steady decline of perfusion to subnormal levels. The critical question is whether this perfusion decrement causes a second and more permanent episode of tissue hypoxia. Evaluation of the adequacy of a given level of flow is difficult if the metabolic demand is not known. Thus, a subnormal flow may be sufficient if the metabolic rate is comparably reduced. Alternatively, if the delivery of 0 2 were insufficient, then the tissue would become hypoxic and levels of NADH would increase. However, the present study demonstrates that NADH was not elevated in areas permanently depleted of ATP. Furthermore, the NADH/NAD+ ratio, while elevated, was the same as that in regions with higher ATP levels. Therefore, postischemic 0 2 delivery does FIGURE 3. Regional Cerebral Metabolites during Postischemic Recirculation. Adjacent coronal sections of frozen brain were processed for topographic depiction of ATP, ph, and NADH. In each picture, increased brightness represents an increased value of ATP, ph, or NADH. A row of NADH standards of decreasing concentration, left to right, is included in the NADH picture. Regional samples (A-F) were assayed quantitatively for metabolite levels. Values for ATP, PCr (phosphocreatine), and Lac (lactate) are given in mmol/kg. Values for NAD + and NADH are given in /xmol/kg.

7 240 STROKE VOL 13, No 2, MARCH-APRIL Number of Samples 20 FIGURE 5. Frequency Histogram of the Number of Samples with a Given Level of A TP in Recirculated Cortex. A total of 175 samples were analyzed for A TP in 35 recirculated brains. 1.0 ATP (mmol/kg) not appear to be the factor limiting restitution of high energy phosphates. Resynthesis of ATP depends not only on the delivery of 0 2, but also on the generation of reducing equivalents (NADH) for reoxidation by the electron transport chain (rig. 6). If the rate of formation of NADH in the citric acid cycle is compromised, then the production of high energy phosphates will be decreased regardless of the supply of 0 2. In the present investigation, the generation of NADH during transient anoxia never returned to normal during recirculation; indeed, the steady decline of the anoxic- NADH transient indicates that the ability to generate NADH becomes progressively compromised during the postischemic period. The marked decrement of tissue NAD + levels is further evidence of a serious disturbance of NAD + - linked redox reactions. A decrease in cerebral NAD + levels following ischemia has not been reported previously; however, a nutritional deficiency of nicotinic acid (Pellagra) has been shown to cause a loss of 30-40% of NAD + in rat brain and was accompanied 2.0 by profound neurologic dysfunction. 24 Thus, the 43% decline of tissue NAD + observed in the present study may severely inhibit the rate of NADH generation in the citric acid cycle and, thus, limit postischemic resynthesis of ATP. Evidence indicating impaired production of reducing equivalents has been detected in synaptosomes derived from hypoxic rat brain. 25 Furthermore, decreased respiratory activity of isolated mitochondria has been demonstrated following 30 minutes of incomplete ischemia in rat brain. 28 Loss of NAD + from the tissue may occur by three routes. First, NAD + may be catabolized by the enzyme, NADase, which is present in high activity in brain, and which rapidly degrades NAD + in cerebral homogenates. 27,28 Secondly, a decrease in the NAD pool may result from the acid-catalyzed destruction of NADH, 29 the rate of which may be accelerated by ischemic acidosis. At ph 6, NADH is destroyed at the rate of approximately 20% per hour. However, assuming tissue NADH levels were maintained at 20 ^tmol/kg, the loss at ph 6 would be only 4 ^mol/kg per hour and could not account for the entire decre- TABLE 4. Cortical Alterations during Postischemic Recirculation Control animals: Recirculated animals: Regions with ATP > 1.3 mmol/kg Regions with ATP < 0.5 mmol/kg ATP 2.22 ±0.07 (26) 1.62* ±0.02 (75) 0.12* ±0.02 (67) Lac 2.2 ±0.1 (26) 10.7* ±0.9 (75) 38.4* ±1.6 (67) NADH 16 ±1 (14) 21* ±1 (45) 14 ±1 (46) Measurements from brain recirculated for 2-4 hours. Values are means ± SEM. Number of samples in parentheses. Number of animals: 3 control, 35 recirculated. Values for ATP, Lac (lactate), and K+ expressed as mmol/kg. Values for NADH and NAD+ expressed as /imol/ kg. Different from control, p < NAD+ 302 ±8 (14) 244* ±5 (48) 171* ±7 (45) NADH/ NAD ± (14) 0.086* ± (45) 0.082* ± (45) K+ 120 ±2 (12) 106* ±2 (27) 69* ±7 (21)

8 ATP REGENERATION FOLLOWING ISCHEMIA/ Welsh et al. 241 Citric Acid Cycle NADH + H + NAD + J Electron Transport jchajj js&gifl 3(ADP+ P;) 3(ATP) ^ \O z HoO FIGURE 6. Central Role of NADH Metabolism in Resynthesis of ATP. NADH formation occurs primarily in the citric acid cycle for reoxidation by the electron transport chain. Generation of high energy phosphates is coupled to the reoxidation of NADH. ment in the NAD pool. Finally, NAD + may simply be washed out of the tissue during postischemic recirculation. Since plasma and mitochondrial membranes are normally impermeable to the pyridine nucleotides, a drastic increase of membrane permeability would have to occur to account for a major efflux of NAD +. Tissue K +, which is primarily intracellular, was also decreased in regions with energy failure. Previously, K + content has not been measured in small regions of brain, nor has K + been correlated directly with metabolite levels determined in the same regional samples. A decrease of the K + /Na + ratio was observed in macrosamples of cortical gray matter following 60 minutes of complete cerebral ischemia in the monkey. 30 Extracellular K + activity, measured with K + -sensitive microelectrodes, increases sharply during ischemia 31 or hypoxia, 32 but returns to control during postischemic recirculation. 31 Normalization of extracellular K + can occur either by active cellular uptake or by clearance from the tissue. The present results demonstrate that regions which resynthesized ATP were able to maintain high levels of K + content, presumably by active transport into the intracellular space. By contrast, the greatest efflux of tissue K + occurred in areas which failed to reconstitute ATP. The regional variations of metabolic recovery, and the factors responsible for the heterogeneity, remain perplexing. Brain regions with permanent energy failure were acidotic as evidenced by decreased umbelliferone fluorescence and by increased levels of tissue lactate. The deleterious effect of excessive lactic acidosis has been noted previously, 3, " 33 and it is likely that NADH generation and energy resynthesis are directly inhibited by decreased ph. However, as in earlier reports, 1,3 end-insult levels of lactate were comparable in all areas of cortex analyzed; thus, there was no apparent regional predisposition for energy failure at the onset of recirculation. The complexity of regional metabolic restitution, as demonstrated by the topographic techniques, emphasizes the necessity for regional measurements. Although the topographic methods used in the present study were sensitive indicators of regional abnormalities, quantitative inferences must be made with caution. Nonspecific changes, such as altered binding of NADH to tissue enzymes, might significantly influence fluorescence intensity. In particular, postischemic elevation of NADH tissue fluorescence in frozen brain did not correspond to increased levels of NADH, measured quantitatively with specific, enzymatic techniques. Thus, while the origin of the intense tissue fluorescence is not known, degradation products of NAD +, some of which fluoresce strongly, might account for the marked increase of tissue fluorescence. Whatever the explanation, it is imperative that qualitative mapping methods be validated by quantitative assay. Although the present study has focused on "irreversible depletion" of ATP, a state representing the initial stages of cerebral infarction, regions which resynthesized ATP to greater than 1.3 mmol/kg cannot be presumed to have escaped permanent injury. Previous reports have demonstrated a severe degree of selective neuronal pathology in tissue with major restitution of energy metabolites. 38 " 38 In the present model of ischemia, there were only a few regions of cortex in which complete normalization of metabolite levels occurred. In areas with substantial, but incomplete, restoration of ATP, the presistent elevation of lactate and NADH may correspond to underlying defects which do not involve a gross depletion of energy metabolites. Indeed, complete restitution of ATP in these areas may be restricted by the decreased size of the adenylate pool 39 and not be a defect in oxidative phosphorylation. Acknowledgments We thank Thomas Langfitt and Walter Obrist for their helpful criticisms and Joan McKee, Christine Gold and Janet Reed for their excellent typing. References 1. Welsh FA, Ginsberg MD, Rieder W, Budd WW: Diffuse cerebral ischemia in the cat: II. Regional metabolites during severe ischemia and recirculation. Ann Neurol 3: , Nordstrom C-H, Rehncrona S, Siesjo BK: Effect of phenobarbital in cerebral ischemia. Part II. Restitution of cerebral energy state, as well as of glycolytic metabolites, citric acid cycle intermediates and associated amino acids after pronounced incomplete ischemia. Stroke 9: , Welsh FA, Ginsberg MD, Rieder W, Budd WW: Deleterious effect of glucose pretreatment on recovery from diffuse cerebral ischemia in the cat. Stroke 11: , O'Connor MJ, Welsh FA, Marcy VR: The effects of bilateral carotid occlusion and systemic hypotension in the cat. III. Regional metabolism and flow during recovery. Submitted to Stroke, March, Ames A III, Wright RL, Kowada M, Thurston JM, Majno G: Cerebral ischemia. II. The no-reflow phenomenon. Amer J Pathol 52: , Hallenbeck JM: Prevention of postischemic impairment of microvascular perfusion. Neurology (Minneap) 27: 3-10, Ginsberg MD, Budd WW, Welsh FA: Diffuse cerebral ischemia in the cat: I. Local blood flow during severe ischemia and recirculation. Ann Neurol 3: , Ginsberg MD, Welsh FA, Budd WW: Deleterious effect of glucose pretreatment on recovery from diffuse cerebral ischemia in the cat. I. Local cerebral blood flow and glucose utilization. Stroke 11: , Schlegel JU, Moses JB: A method for visualization of kidney blood vessels applied to studies of the crush syndrome. Proc Soc

9 242 STROKE VOL 13, No 2, MARCH-APRIL 1982 Exp Biol Med 74: , Schuette WH, Whitehouse WC, Lewis DV, O'Connor M, Van Buren JM: A television fluorometer for monitoring oxidative metabolism in intact tissue. Med Instrum 8: , OldendorfWH: Measurement of the mean transit time of cerebral circulation by external detection of an intravenously injected radioisotope. J Nucl Med 3: , Welsh FA, Rieder W: Evaluation of in situ freezing of cat brain by NADH fluorescence. J Neurochem 31: , Kogure K, Alonso OF: A pictorial representation of endogenous brain ATP by a luminescent method. Brain Res 154: , Sundt TM, Anderson RE, Van Dyke RA: Brain ph measurements using a diffusible, lipid soluble ph sensitive fluorescent indicator. J Neurochem 31: , Lowry OH, Passonneau JV: A Flexible System of Enzymatic Analysis. New York and London, Academic Press, Kato T, Berger SJ, Carter JA, Lowry OH: An enzymatic cycling method for nicotinamide-adenine dinucleotide with malic and alcohol dehydrogenases. Analyt Biochem 53: 86-97, Outlaw WH Jr, Lowry OH: Measurement of 10" 7 to mol of potassium by stimulation of pyruvate kinase. Analyt Biochem 92: , Dunnett CW: New tables for multiple comparisons with a control. Biometrics 20: , Levy DE, Brierley JB, Plum F: Ischemic brain damage in the gerbil in the absence of "no-reflow." J Neurol Neurosurg Psychiatry 38: , Levy DE, Brierley JB, Silverman DG, Plum F: Brief hypoxiaischemia initially damages cerebral neurons. Arch Neurol 32: 450^56, Hossmann K-A, Kleihues P: Reversibility of ischemic brain damage. Arch Neurol 29: , Marshall LF, Durity F, Lounsbury R, Graham DI, Welsh FA, Langfitt TW: Experimental cerebral oligemia and ischemia produced by intracranial hypertension. Part I. Pathophysiology, electroencephalography, cerebral blood flow, blood-brain barrier, and neurological function. J Neurosurg 43: , Levy DE, Van Uitert RL, Pike CL: Delayed postischemic hypoperfusion: A potentially damaging consequence of stroke. Neurology (Minneap) 29: , Garcia-Bunel L, McDougal DM, Burch HB, Jones EM, Touhill E: Oxidized and reduced pyridine nucleotide levels and enzyme activities in brain and liver of niacin deficient rats. J Neurochem 9: , Rafalowska U, Erecinska M, Wilson DF: The effect of acute hypoxia on synaptosomes from rat brain. J Neurochem 34: , Rehncrona S, Mela L, Siesjo BK: Recovery of brain mitochondrial function in the rat after complete and incomplete cerebral ischemia. Stroke 10: , Mann PJG, Quastel JH: Nicotinamide, cozymase and tissue metabolism. Biochem J 35: , Mcllwain H, Grinyer I: Concentration or exclusion of some heterocyclic compounds by brain cortex, in relation to their effects on carbohydrate metabolism. Biochem J 46: , Lowry OH, Passonneau JV, Rock MK: The stability of the pyridine nucleotides. J Biol Chem 236: , Zimmerman V, Hossmann K-A: Resuscitation of the monkey brain after one hour's complete ichemia. Brain Res 85: 1-11, Silver I A: Changes in PO s and ion fluxes in cerebral hypoxiaischemia. Adv Exp Med Biol 78: , Kirshner HS, Blank WF Jr, Myers RE: Brain extracellular potassium activity during hypoxia in the cat. Neurology (Minneap) 25: , Myers RE: Anoxic brain pathology and blood glucose. Neurology (Minneap) 26: 345, Kaplan NO, Collowick SP, Barnes CC: Effect of alkali on diphosphopyridine nucleotide. J Biol Chem 191: , Lowry OH, Roberts NR, Kapphahn JI: The fluorometric measurement of pyridine nucleotides. J Biol Chem 224: , Salford LG, Plum F, Brierley JB: Graded hypoxia-oligemia in rat brain. II. Neuropathologic alterations and their implications. Arch Neurol 29: , Marshall LF, Graham DI, Durity F, Lounsbury R, Welsh FA, Langfitt TW: Experimental cerebral oligemia and ischemia produced by intracranial hypertension. Part II. Brain Morphology. J Neurosurg 43: , Ginsberg MD, Graham DI, Welsh FA, Budd WW: Diffuse cerebral ischemia in the cat. III. Neuropathological sequelae of severe ischemia. Ann Neurol 5: , Ljunggren B, Schutz H, Siesjo BK: Changes in energy state and acid-base parameters of the rat brain during complete compression ischemia. Brain Res 73: , 1974

10 Factors limiting regeneration of ATP following temporary ischemia in cat brain. F A Welsh, M J O'Connor, V R Marcy, A J Spatacco and R L Johns Stroke. 1982;13: doi: /01.STR Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 1982 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Stroke is online at: