Focal and Perifocal Changes in Tissue Energy State During Middle Cerebral Artery Occlusion in Normo- and Hyperglycemic Rats

Transcription

1 Journal o/cerebral Blood Flow and Metabolism 12: The nternational Society of Cerebral Blood Flow and Metabolism Published by Raven Press, Ltd., New York Focal and Perifocal Changes in Tissue Energy State During Middle Cerebral Artery Occlusion in Normo- and Hyperglycemic Rats laroslava Folbergrova, *Hajime Memezawa, tmaj-lis Smith, and tbo K. Siesj6 nstitute of Physiology, Czechoslovak Academy of Sciences, Prague, Czechoslovakia; *Second Department of nternal Medicine, Nippon Medical School, Tokyo, Japan; and tlaboratory for Experimental Brain Research, University of Lund, Lund, Sweden Summary: The objective of the present study was to assess changes in cellular energy metabolism in focal and perifocal areas of a stroke lesion and to explore how these changes are modulated by preischemic hyperglycemia. A model for reversible occlusion of the middle cerebral artery (MCA) in rats was used to study changes in energy metabolism. Following MCA occlusion for 5, 15, or 3 min in normoglycemic rats, the tissue was frozen in situ, and samples from the lateral caudoputamen and from two neocortical areas were collected for metabolite analyses, together with a control sample from the contralateral, nonischemic hemisphere. Two other groups, subjected to 3 min of MCA occlusion, were made hyperglycemic by acute glucose infusion or by prior injection of streptozotocin. Enzymatic techniques were used for measurements of phosphocreatine, creatine, ATP, ADP, AMP, glycogen, glucose, pyruvate, and lactate. The neocortex of the contralateral, nonischemic hemisphere had labile metabolites that were similar to those measured in control animals. psilateral neocortex bordering the focus, and thus constituting the "penumbra," showed mild to moderate ischemic changes. n the "focus" (lateral caudoputamen plus the overlying neocortex), deterioration of energy state was rapid and relatively extensive (ATP content 2-4% of control). After 5 min of occlusion, no further deterioration of metabolic parameters was observed. Substrate levels were markedly reduced, and lactate content rose to 1 mm kg - 1. n the animals with the most severe energy depletion, no additional accumulation of lactate occurred, suggesting substrate depletion. This was confirmed by the results obtained in the hyperglycemic subjects whose tissue lactate contents rose to 2 rnm kg - 1. However, the energy state of the focus was better preserved in both hyperglycemic groups as compared with the normoglycemic group. t has been shown, in this model, that relatively brief occlusion periods are required to induce infarction. The present results demonstrate that this can occur in spite of the absence of pronounced depletion of energy reserves. After 3 min of MCA occlusion, infarction developed in the lateral caudoputamen, but not in the neocortex. Since a similar perturbation in metabolic state was demonstrated here, other factors must contribute to the degree of tissue damage. The present results suggest that damage is exaggerated by hyperglycemia because it allows additional lactate to accumulate in the partially substrate-depleted tissue. Key Words: Cerebral ischemia-energy metabolites Hyperglycemia-Middle cerebral artery occlusion. There is little doubt that ischemia causes brain damage by disrupting or straining cellular energy metabolism (e.g., Siesjo, 19a). Following shortlasting, transient ischemia, the brain damage in- Received January 25, 1991; revised April 25, 1991; accepted July 3, Address correspondence and reprint requests to Prof. B. K. Siesj6 at Laboratory for Experimental Brain Research, Department of Neurobiology, Experimental Research Center, University Hospital, S Lund, Sweden. Abbreviations used: Cr, creatine; MCA, middle cerebral artery; PCr phosphocreatine. curred can be traced back to the ischemic disruption of cellular energy state, and, at least to a first approximation, the density of damage is proportional to the duration of energy depletion (see, however, Hossmann, 195). Thus, although adverse mechanisms operating during recirculation may contribute to the final damage (see Siesjo, 191, 19a; Siesjo and Wieloch, 195), it has never been shown that damage affects brain areas whose metabolic state remains unperturbed during the primary insult. t has been shown, though, that ischemic damage is aggravated by hyperglycemia. The effect 25

2 26 J. FOLBERGROV A ET AL. is observed when hyperglycemia is induced before, and not after, ischemia (Rehncrona et a., 191; Pulsinelli et a., 192; however, see also Lundy et a., 197). This is presumably because excessive acidosis during the period of energy depletion triggers additional adverse reactions whose effects become manifest after hours of recirculation (Myers, 1979; Siemkowicz and Hansen, 191; Siesjo, 19b; Smith et a., 19). The relationship between loss of energy homeostasis and cell damage is less obvious in focal ischemia due to middle cerebral artery (MCA) occlusion, particularly if the occlusion is permanent. Early studies showed that, in squirrel monkeys, tissue ATP content fell gradually and lactate content increased over hours (Michenfelder and Sundt, 1971). t was not known, though, whether intermediate degrees of energy depletion existed or if the values obtained represented variable proportions of normal and extensively deranged tissue. Such heterogeneity was documented later (Ratcheson and Ferrendelli, 19; Welsh, 194) and led to successful attempts to separate severely underperfused and better perfused tissues, not by anatomical landmarks, but from the distribution of histochemical flow indicators (Selman et a., 197). On the basis of regional sampling, new information was acquired on focal and perifocal changes in energy metabolism during stroke (Hossmann et a., 195; Nowicki et a., 19; Obrenovitch et a., 19; Selman et a., 199). The information obtained is, however, contradictory. Thus, whereas Selman et ai. (199) demonstrated extensive deterioration of cellular energy state in the focus (neocortex) and moderate deterioration in a bordering penumbra zone, Nowicki et al. (19) had previously reported moderate changes in the focus and virtually no changes in the penumbra after 2 h of occlusion. Since the penumbra becomes part of the infarct after 4 h of occlusion (Duverger and Mackenzie 19), the data give no clue as to the metabolic mechanisms involved. Another unresolved issue, equally important, is the cause of the aggravated damage that is observed when the MCA is transiently occluded in hyperglycemic animals (see Nedergaard, 197). Thus, although this result is predictable, considering the data previously obtained following global or forebrain ischemia (see above), the effect of hyperglycemia on infarct size in permanently occluded subjects is not consistent (see Ginsberg et a., 197; Nedergaard and Diemer, 197; de Courten-Myers et a., 19; Duverger and Mackenzie, 19; Zasslow et a., 199). The problems of interpretation were highlighted by Nedergaard and Diemer (197), who suggest that although exaggerated lactic acid pro- duction may contribute to the total destruction of cells in the focus, it may curtail events leading to selective neuronal necrosis outside the infarcts. This important issue has not previously been directly addressed by measurements of cellular energy state. We have recently adapted the model of Koizumi et ai. (196) to studies of transient and permanent MCA occlusion in the normotensive Wistar rat and described changes in blood flow and histopathological outcome (Memezawa et a., 1992a). n this model, MCA occlusion of 3-min duration consistently gives infarction of the lateral caudoputamen but not of the adjacent neocortex, which either is normal or shows selective neuronal necrosis (Memezawa et a., 1992b). With longer occlusion periods, infarction of the neocortex increases, and after 9-12 min, the whole area supplied by the MCA becomes infarcted. The objectives of this study were twofold. First, we wished to assess changes in energy metabolism in focal, perifocal, and normal tissues following MCA occlusion for 5, 15, and 3 min induced in normoglycemic rats to allow correlation of such changes to the reduction in flow during ischemia as well as to the localization of tissue damage observed under these conditions (Memezawa et a., 1992a,b). Second, we wished to study the extent to which preischemic hyperglycemia modulates the metabolic changes and, in particular, to explore whether it enhances accumulation of lactate. MATERALS AND METHODS Operative and MCA occlusion techniques Male Wi star rats (M legaard Breeding Center, Copenhagen, Denmark), weighing g, were used for this study. Anesthesia was induced with 3.% isoflurane in NzO/Oz (7:3). The animals were ventilated with a small animal respirator, isoflufane concentration being reduced to % during operation. Polyethylene catheters were inserted into a tail artery and vein for blood pressure recording, blood sampling, and drug infusions. Thermistor probes were placed on the bregma under the scalp and in the rectum. With the help of these probes, external heating was employed so as to keep both the scalp and the rectal temperature at 37 C. A midline skin incision was made to accommodate a plastic funnel, used for freezing the brain in situ with liquid nitrogen (Ponten et ai., 1973). MCA occlusion (Koizumi et ai., 196) was induced as described in two previous publications (Memezawa et al., 1992a,b). n summary, MCA occlusion was induced by an intraluminal filament. The occluding device consisted of two pieces: occluder filament and guide sheath. A surgical midline incision was made to expose the right common, internal, and external carotid arteries. The external carotid and the occipital arteries were ligated. The pterygopalatine artery, a branch of the internal carotid artery, was encircled with a suture and retracted to prevent incorrect insertion of the guide sheath. After that had been J Cereb Blood Flow Metab, Vol. 12, No.1, 1992

3 ENERGY METABOLTES N STROKE 27 done,.1 ml of heparin (15 V ml-) was given. The internal carotid artery was temporarily closed by a microvascular clip, and the common carotid artery was closed by a suture, 3 mm proximal from the carotid bifurcation. A small incision was then made in the common carotid artery, 1 mm proximal from the carotid bifurcation, and the MCA-occluding device was inserted from the right common carotid artery into the internal carotid artery. After removal of the microvascular clip, the occluder filament was advanced so as to close the origin of the MCA. Sampling technique and analytical methods The animals were subjected to 5, 15, or 3 min of MCA occlusion. At the end of these ischemic periods, the tissue was frozen in situ. After -1 min of freezing, the brain was chiseled out with the help of a chisel and hammer, during intermittent irrigation of the head and the brain with liquid nitrogen. Further preparation of the tissue occurred in a glove box at - 2 C. The tissue was cut into 4- to 5-mm-thick slices in the coronal plane. Tissue samples, representing the lateral caudoputamen and upper and lower neocortical areas in the ipsilateral hemisphere and the lower neocortical area in the contralateral hemisphere, were dissected as shown in Fig. 1. The samples, weighing 25-3 mg each, were extracted at - 2 C with HCUmethanol. Further treatment and preparation of tissue extracts for enzymatic fluorometric analyses were performed according to Lowry and Passonneau (1972), as described previously (Folbergrova et a., 1972). Analytical conditions were the same as published earlier (Folbergrova et a., 1972). The following metabolites were measured: phosphocreatine (PCr), creatine (Cr), ATP, ADP, AMP, glycogen, glucose, lactate, and pyruvate. Experimental groups Two series of animals were studied: normoglycemic and hyperglycemic. n the normoglycemic groups, the plasma glucose concentration was not manipulated. n each of the experimental groups (5, 15, or 3 min of occlusion), four tissue regions were sampled for analysis (see Fig. 1). n addition, a group of sham-operated controls (n = 3) was studied to provide control values for the four regions analyzed. FG. 1. Schematic drawing showing the four brain regions dissected for metabolite analysis. Regions 1-3 were taken from the middle cerebral artery-occluded hemisphere and region 4 from the contralateral hemisphere. The second series consisted of two groups of hyperglycemic animals. n one (n = 4) preischemic hyperglycemia was induced by an intravenous infusion of a 25% glucose solution (2-3 mi) over 3 min to produce a stable blood glucose level of 2 25 mm. The glucose infusion was continued at a slower rate during ischemia to keep the glucose levels within mm. n the second group (n = 4), acute diabetes was induced by an intraperitoneal injection of streptozotocin (6 mg kg-) 2 days before operation (Nedergaard, 197). n both hyperglycemic groups, MCA occlusion was induced for 3 min before the tissue was frozen in situ. Experimental animals, with the exception of the streptozotocin-injected group, were fasted the night before the experiment. Statistics Statistical differences were evaluated with one-way analysis of variance, followed by Dunnett's test for differences between control and experimental groups. The statistical analyses were performed separately for shamoperated versus normoglycemic ischemia of 5-, 15-, and 3-min duration and for normoglycemic ischemia of 3- min duration versus glucose-infused and streptozotocininjected hyperglycemic animals with 3-min ischemia. A p value of <.5 was regarded as statistically significant. The number of analyses in each group was small (n 3-4). RESULTS As shown in Table 1, the animals were normothermic, normotensive, and normocapnic and they had normal plasma ph. Normoglycemic animals had blood glucose values typical for starved animals subjected to anesthesia and artificial ventilation, while animals in the two hyperglycemic groups had glucose concentrations around 2 mm kg - 1. Two streptozotocin-injected animals showed only moderately elevated blood glucose levels and were excluded from the study. Table 2 shows tissue concentrations of labile metabolites in sham-operated and ischemic animals. n all regions analyzed in sham operated animals, the concentrations were in the normal range (e.g., Folbergrova et a., 1972; Ponten et al., 1973; see also Nilsson et a., 1975; Folbergrova et a., 199). Thus, even in regions relatively far from the surface, constituting the initial freezing front, concentrations of PCr and ATP were high and those of ADP and AMP were low. The levels of metabolites measured in MCAoccluded normoglycemic animals are also shown in Table 2. Values for region 4, i.e., neocortical tissue from the contralateral side, represent intraanimal controls since this tissue is nonischemic. However, there was a tendency for ATP, AMP, PCr, adenylate energy charge (Table 2), and lactate levels (Table 3) in the ischemic groups to be somewhat less optimal than those measured in sham-operated controls (statistically significant at p <.1 for AMP J Cereb Blood Flow Metab, Vol. 12, No.1, 1992

4 2 J. FOLBERGROV A ET AL. TABLE 1. Physiological parameters Normoglycemia Hyperglycemia schemic period Glucose Streptozotocin Sham 5 min (n = 3) (n = 3) Before ischemia Rectal temperature ee) 37. ± ±. 1 Scalp temperature ee) ± ±. Blood pressure (mm Hg) 13 ± 6 17 ± 6 Peo2 (mm Hg) ± ±. P2 (mm Hg) 115. ± ± 5.9 ph 7. 4 ± ±. 2 Blood glucose (mm) 5. 5 ± ±. 4 During ischemia Rectal temperature ee) ± Scalp temperature ee) Blood pressure (mm Hg) 114 ± 4 Peo2 (mm Hg) P2 (mm Hg) ph Blood glucose (mm) 15 min 3 min 3 min 3 min (n = 3) (n = 4) (n = 4) (n = 4) 37. ± ± ± ± ± ± ± ±.5 12 ± ± 9 91 ± 3 94 ± ± ± ± ± ± ± ± \ ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±.5 13 ± ± ± 9 95 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±. 1.4 ± ± 1. 1 Values are means ± SO., not determined. and at p <.5 for energy charge and lactate, both in the 5-min ischemia group). Region 1, representing tissue of an area bordering the focus (penumbral tissue), showed mild to moderate metabolic perturbation at all three occlusion times, with decreases in PCr and ATP and rises in Cr, ADP, and AMP concentrations. Nevertheless, with the statistical methods used, these changes were not different from control. However, as Fig. 2 shows, all ATP values were outside the normal ranges. Changes in energy metabolites were pronounced in regions 2 and 3, representing focal neocortical and focal striatal tissue (Table 2; Fig. 2). t is of interest that changes in energy metabolites were at least as pronounced after 5 as after 3 min. There was thus no indication of progressive energy failure during the first 3-min period. The results given in Table 2 show that induced hyperglycemia improved energy state in ischemic neocortex, in both the penumbra (region 1) and the focus (region 2). This was clearly shown by the increased tissue ATP values. No significant difference was evident in the caudoputamen. Changes in glucose metabolites are shown in Table 3 (see also Fig. 2). Although glucose and glycogen concentrations were reduced in region, substrate depletion was not observed. This was also evidenced by the fact that the pyruvate concentration tended to be increased, rather than reduced, during ischemia. The pattern was different in regions 2 and 3. Thus, the reduction in glucose and glycogen concentrations was pronounced, lactate content rose to -1 mm kg - 1, and pyruvate con- centrations tended to be relatively low. This raises the question of whether the relatively low lactate concentrations were, at least in part, due to substrate depletion. Two findings support this notion. First, animals having very low ATP values did not show higher lactate values, but had low values for glycogen, glucose, and pyruvate concentration (data not shown). Furthermore, although hyperglycemia was associated with improved (region 2) or unchanged (region 3) energy state, the lactate contents rose to high levels. This concerns especially region 3, in which the lactate levels increased >2.5 times compared with the normoglycemic group with the same period of ischemia (significant at p <.1; see Table 3). DSCUSSON One objective of this study was to assess focal and perifocal changes in energy metabolism following MCA occlusion induced by the intraluminal filament technique (Memezawa et a., 1992a), with particular emphasis on the phosphorylation state of the adenine nucleotides and on lactate. The former must determine the balance between exergonic and endergonic reactions and thereby cell damage, while the latter is a major determinant of extra- and intracellular ph. n discussing these important modulators of cell damage, we make the tacit assumption that changes in the chemically measured ADP and AMP concentrations reflect corresponding changes in the free ADP and AMP concentrations (see Erecinska and Silver, 199); also that J Cereb Blood Flow Metab, Vol. 12, No. 1, 1992

5 ENERGY METABOLTES N STROKE 29 TABLE 2. Changes of energy metabolites during MCA occlusion Normoglycemia Hyperglycemia schemic period Glucose Streptozotocin Sham 5 min 15 min 3 min 3 min 3 min (n = 3) (n = 3) (n = 3) (n = 4) (n = 4) (n = 4) Region 1 (upper ex) per 4.45 ± ± ± ± ± ±.92 Cr 5.56 ±.9.33 ± ± ± ± ±. 9 ATP 2.6 ± ± ± ± ±. 9a ±.24a ADP.29 ±.3.47 ± ±.2.39 ±.9.25 ±.2b.26 ±.2b AMP. 5 ±.1.6 ± ±.44.2 ± ±. 1.7 ±. EC.94 ±..6 ± ±.23. ± ±..93 ±.1 Region 2 (lower ex) PCr 4.46 ±.1.41 ±.5'.4 ± 1.96c 1.24 ± 1.5c 2.99 ± ± 1.33 Cr 5.31 ± ±.54c.5 ± 1.94d.31 ± 1.3d 6.9 ± ±.7 ATP 2. ±.3.4 ±.53c.69 ±.9c. 75 ±.c 2.26 ±.41a 2. 3 ±.61a ADP. 31 ±.2.73 ± O.13c.63 ±.2d.53 ± ±. 9.4 ±. 1 AMP.6 ± ±.44'.96 ±.44d.61 ± ±.1.35 ±.34 EC.94 ±..29 ±.1c.44 ±.23d.5 ±.29d.5 ± ±. 15 Region 3 (caudoputamen) PCr 4.52 ± ± 1.1c.9 ±.31c 1. 4 ±. 6c 1.6 ± ±.65 Cr 5.49 ± ± 1.1c.9 ±.c.53 ±. 75c.13 ± ±. 59 ATP ± ±.2c.2 ± O.27c.94 ±.49c 1.6 ± ±.4 ADP.3 ±.6.54 ± ± O.13d.53 ± ± ±.12 AMP.5 ±.3 1. ±.62c.93 ±.25d.66 ± ± ±.19 EC.94 ±.1.55 ±.25d.4 ±.11'.56 ±.16c.67 ± ±.13 Region 4 (contralateral CX) PCr 4.54 ± ± ± ± ± ±. 19 Cr 5.34 ± ± ± ± ± ±.46 ATP 2.9 ± ± ± ± ± ±. ADP.29 ±.1.31 ±.1.29 ±. 1.2 ±.2.27 ±.1.2 ±.2 AMP.5 ±.2.14 ±.4c.1 ±. 3. ±.2.1 ±.. ±.2 EC. 94 ±.. 91 ±.2d. 92 ±.1.93 ± ± ±. 1 Values are means ± SD. MCA, middle cerebral artery; CX, neocortex; PCr, phosphocreatine; Cr, creatine; EC, adenylate energy charge. Statistical significances were evaluated by analysis of variance followed by Dunnett's test. The hyperglycemic groups with 3-min MCA occlusion were compared with the normoglycemic 3-min MeA occlusion group: ap <.5, bp < om. The normoglycemic ischemia groups were compared with the sham-operated group: cp <. 1, dp <. 5. changes in lactate concentration give rise to proportional changes in extra- and intracellular ph. The other main objective was to explore the influence of preischemic hyperglycemia on the metabolic changes. n discussing changes in lactate, we will address the question of whether its rate of accumulation in the ischemic tissue is limited by substrate supply. We wish to emphasize that we consider the determinants of tissue damage following transient MCA occlusion to be at least three: the type of cells/cell aggregates (e.g., neocortex versus caudoputamen), the density of energy failure, and the severity of intra- and extracellular acidosis. Labile organic phosphates n the present model, the "focus" consists of the lateral part of the caudoputamen and the overlying neocortex since both these areas have flow rates of 5-1% of control (Memezawa et a., 1992a). Both regions (2 and 3) also showed extensive deterioration of cellular energy state, which was manifested already 5 min following MCA occlusion. t is of interest that neither of these regions showed complete deterioration of the phosphorylation state of the adenine nucleotide pool, since only 6 of 2 ATP values were <.25 mm kg-. n complete ischemia, ATP concentrations are lower (Folbergrova et a., 1974; Ljunggren et a., 1974), and also after severe, incomplete forebrain ischemia, ATP levels are severely depleted (Pulsinelli and Duffy, 193; Hillered et a., 195). t should be remarked, however, that in all 1 animals studied, the deterioration of energy state in the focus was extensive. This attests to the fact that the MCA occlusion model gives a reproducible reduction of CBF. The fall in ATP (and PCr) concentration observed in this study is clearly more rapid and extensive than that originally reported for the squirrel monkey brain by Michenfelder and Sundt (1971). Probably, the reduction in CBF is less marked in the monkey J Cereb Blood Flow Metab, Vol. 12, No.1, 1992

6 3 J. FOLBERGROV A ET AL. TABLE 3. Changes of glucose metabolites during MCA occlusion N ormoglycemia Hyperglycemia schemic period Glucose Streptozotocin Sham 5 min 15 min 3 min 3 min 3 min (n = 3) (n = 3) (n = 3) (n = 4) (n = 4) (n = 4) Region (upper ex) Glucose ± ±.21a ± ± ±.94c ± 1.25c Glycogen 3. 5 ± ±.52b ±. a 1.6 ± 1.22b 2.96 ±.59d 2. 6 ± 1.25 Lactate. 93 ± ±.27b ± ± ± ± 4.52 Pyruvate. 7 ± ± ± ± ± ±. Region 2 (lower ex) Glucose 1. 4 ± ±.6b. 5 ±.2b. 4 ±. 62b 5.25 ± 1.59c ± 1. 3c Glycogen ± ±. 19b. 23 ±.1b. 16 ±.22b 2. ±.74d 1.63 ± 1.36 Lactate 1.13 ± ± 1.41b 1.53 ±.56b.55 ± 1.b 14.6 ± ±.67 Pyruvate. 9 ± ±.5. 6 ± ±.1.17 ±..1 ±. 11 Region 3 (caudoputamen) Glucose 1.36 ± ±.32b. 32 ± O.lOb.34 ±. 16b 2.47 ±.5c 1.43 ±. 79 Glycogen 2.25 ± ±. 2b. 29 ±.2b.2 ±. 16b ±.46c.46 ±. 31 Lactate ± ±. 74b. 2 ±.44b ± 1. 17b ±.2c ± 6.14c Pyruvate. ± ± ± om. 4 ± ±. 4d. ±.3 Region 4 (contralateral ex) Glucose 1. 1 ± ± ± ± ± O.64c ±. c Glycogen 3.25 ± ± ± ± ± ±.32 Lactate 1. ± ±. 1a 1. 3 ± ± ±.42C 1. ±.47 Pyruvate. ±. 2.9 ± ± ±. 1. ±. 1.7 ±.2 Values are means ± SD. MeA, middle cerebral artery; ex, neocortex. Statistical significances were evaluated by analysis of variance followed by Dunnett's test. The normoglycemic ischemia groups were compared with the sham-operated group: ap <.5, bp <.1. The hyperglycemic groups with 3-min MeA occlusion were compared with the normoglycemic 3-min MeA occlusion group: cp <.1, dp <.5. than in the rat. Our data are more in line with those reported for cortical areas 6- by Selman et ai. (199), who used the Tamura model in rats (Tamura et a., 191a). The CBF values reported by Tamura et ai. (191b) were, for these cortical areas 15% of control. Our values were lower (see Memezawa et a., 1992a), probably explaining the more extensive deterioration of energy state. Different values were, however, reported by Nowicki et ai. (19), who found that ATP and PCr in the striatum after 2 h of MCA occlusion (Tamura model) were reduced by 46 and 41 % only, the parietal cortex showing normal values. Since the whole caudoputamen was sampled, the values probably represent a mixture of poorly and better perfused tissue. The absence of changes in neocortex is most easily explained by a higher "penumbral" blood flow in the species used. n our material, region 1, representing the penumbral area, had a clearly better upheld energy state. However, adenine nucleotides and PCr were not normal, and in 3 of 1 animals, relatively extensive loss of energy homeostasis was encountered. At least in part, the variability obtained could reflect interanimal differences in the density of ischemia in this region. t must be emphasized, though, that if flow is reduced to values that seriously curtail cellular energy production, a large variability is expected simply because energy failure and loss of ion homeostasis are mutually reinforcing events, tending to give a threshold effect on cerebral energy state (see Naritomi et a., 19; Obrenovitch et a., 19; Siesjo, 19a). n contrast, region 4, taken from the contralateral neocortex, had values for labile metabolites that were close to control. There was a very moderate lowering of ATP concentrations in the 15- and 3- min groups and an increase in AMP concentration at 5 and 15 min of ischemia, but these changes were indeed moderate. Substrates and lactate n the control animals, glucose, glycogen, pyruvate, and lactate concentrations were as are usually considered normal (e.g., Folbergrova et a., 1972; Ljunggren et a., 1974). Glycogen and glucose concentrations were lower in the caudoputamen. This has been noted before (e.g., Pulsinelli and Duffy, 193). n the densely ischemic areas, tissue glycogen and glucose were virtually depleted already after 5 min, and very low substrate levels were found after 15 and 3 min. n the densely ischemic neocortex (region 2) in the normoglycemic groups, lactate content rose to, and remained at, -1 mm kg-. n the J Cereb Blood Flow Me/ab, Vol. 12, No.1, 1992

7 ENERGY METABOLTES N STROKE 31 ATP 3 o 2 o FG. 2. ndividual values for ATP (top) and lactate (bottom) in regions 1-3, at the different times of middle cerebral artery (MeA) occlusion studied in the normoglycemic animals. Values are given as mm kg -1, and control values are shown as horizontal lines, with SD as a shaded area. o o o o -r---- orl r Lactate S region 1 o e region 2 9 region 3 o S duration of MeA occlusion, min. hyperglycemic animals, the lactate values in neocortex region 2 were higher by 7% as compared with those in normoglycemic animals with the same period of occlusion. Mean lactate contents in the caudoputamen were somewhat lower in all normoglycemic groups, probably reflecting the lower preischemic substrate levels. However, as the present results demonstrate (Table 3), >2.5-fold higher values in this region were observed in hyperglycemic animals, reaching levels of 2 mm kg-. The question arises of why lactate values exceeding 1 mm kg- were not encountered in normoglycemic animals. n principle, two factors limit the production of lactate: substrate availability and ATP concentration (or phosphorylation potential.) The dependence on ATP concentration is clear-cut in hypoxia, where flow does not limit substrate delivery (Gardiner et a., 192). n ischemia, glucose delivery is flow restricted (EkOf and Siesjo, 1972). Thus, when reduction of flow curtails ATP synthesis, the production of lactate may be inhibited by lack of substrate. This was clearly the case in the present experiments. First, in samples with very low A TP values and lactate concentrations around 1 mm kg-, both glucose and glycogen concentrations were low, reflecting the lack of substrate. Sec- ond, when substrate delivery was enhanced by induced hyperglycemia, lactate content rose, even though A TP concentration did not fall further or actually rose. Correlation to histopathological alterations We have shown recently that 3 min of MCA occlusion in normoglycemic animals induces infarction localized to the lateral caudoputamen (Memezawa et a., 1992b), while the adjacent neocortex shows either no change or selective neuronal necrosis. The present results define the corresponding metabolic profiles following similar periods of occlusion. They indicate that there is extensive deterioration of energy state in focal striatal and focal neocortical tissue. However, this deterioration is not as extensive as in forebrain ischemia in the rat, which gives dense damage in the caudoputamen even if ischemia duration is <3 min (see Pulsinelli et a., 192; Pulsinelli and Duffy, 193; Smith et a., 194, 19). Clearly, factors other than energy failure enter as determinants of the final damage in curred after transient ischemia. Since neither the phosphorylation state nor the lactate content differed between the regions, it seems likely that inherent differences in vulnerability play a role, as J Cereb Blood Flow Metab, Vol. 12, No.1, 1992

8 32 J. FOLBERGROV A ET AL. they do in complete or near-complete global or forebrain ischemia. n our recovery studies, we have failed to observe damage to cells in area 1 if reperfusion occurs after 3 min (Memezawa et a., 1992b). The present results showed that 1 of 1 animals showed mild to moderate deterioration of cellular energy state of that area. Obviously, this degree of metabolic perturbation is compatible with preservation of cell structure, unless the ischemia is of even longer duration. A corresponding correlation between histopathology and changes in energy state under hyperglycemic conditions must remain tentative until we have histopathological results. However, since Nedergaard (197) clearly demonstrated an adverse effect of hyperglycemia in animals with 1 and 15 min of transient MeA occlusion, and since our preliminary data confirm these results, the following comments seem justified. First, because the infarct is increased in hyperglycemic subjects, it seems likely that it is the rise in lactate content (and reduction in ph) that is responsible (the phosphorylation potential was either unchanged or improved). Second, if we can confirm that hyperglycemia reduces perifocal neuronal necrosis similarly as was shown by Nedergaard and Diemer (197), the present results obtained in area 1 would suggest that improvement of substrate supply and increase in ATP concentration contribute. n summary, the present article supplements previous ones in allowing definition of focal and perifocal changes in energy metabolism during MeA occlusion (see Nowicki et a., 19; Selman et a., 199). Since part of this information was collected in models for reversible MeA occlusion, it now becomes possible to relate ischemic and postischemic metabolic events to the final damage incurred. The present results on energy metabolism supplement those of Abe et a. (19) who studied edema and lipid metabolism in a similar model. Our data present novel information on the relationship between substrate delivery, energy failure, and lactate accumulation, information that is crucial for an understanding of how hyperglycemia modulates damage in focal and perifocal areas. Acknowledgment: This study was supported by the Swedish Medical Research Council (grant no. 14X-263) and NH, V.S. Public Health Service (grant no. 5. RO NS-73). Maj-Lis Smith was supported by a postdoctoral fellowship from Pfizer, V.K. The authors gratefully acknowledge the skilled technical assistance of Lena Sjoberg on metabolite measurements and the practical advice of Dr. Kenichiro Katsura on data processing. REFERENCES Abe K, Araki T, Kogure K (19) Recovery from edema and of protein synthesis differs between the cortex and caudate following transient focal cerebral ischemia in rats. J Neurochern 51: de Courten-Myers G, Myers RE, Schoolfield L (19) Hyperglycemia enlarges infarct size in cerebrovascular occlusion in cats. Stroke 19: Duverger D, Mackenzie ET (19) The quantification of cerebral infarction following focal ischemia in the rat: influence of strain, arterial pressure, blood glucose concentration, and age. J Cereb Blood Flow Metab : Ekl6f B, Siesj6 BK (1972) The effect of bilateral carotid artery ligation upon acid-base parameters and substrate levels in the rat brain. Acta Physiol Scand 6:52-53 Erecinska M, Silver A (199) ATP and brain function. J Cereb Blood Flow Metab 9:2-19 Folbergrova J, MacMillan V, Siesj6 BK (1972) The effect of hypercapnic acidosis upon some glycolytic and Krebs cycleassociated intermediates in the rat brain. J Neurochern 19: Folbergrova J, Ljunggren B, Norberg K, Siesj6 BK (1974) nfluence of complete ischemia on glycolytic metabolites, citric acid cycle intermediates, and associated amino acids in the rat cerebral cortex. Brain Res : Folbergrova J, Minamisawa H, Ekholm A, Siesj6 BK (199) Phosphorylase a and labile metabolites during anoxia: evidence for a rise in cytosolic free calcium before calcium enters from extracellular fluid. J Neurochem 55: Gardiner M, Smith M-L, Kagstr6m E, Shohami E, Siesjo BK (192) nfluence of blood glucose concentration on brain lactate accumulation during severe hypoxia and subsequent recovery of brain energy metabolism. J Cereb Blood Flow Metab 2: Ginsberg MD, Prado R, Dietrich WD, Busto R, Watson BD (197) Hyperglycemia reduces the extent of cerebral infarction in rats. Stroke 1: Hillered L, Smith M-L, Siesj6 BK (195) Lactic acidosis and recovery of mitochondrial function following forebrain ischemia in the rat. J Cereb Blood Flow Metab 5: Hossmann K-A (195) Post-ischemic resuscitation of the brain: selective vulnerability versus global resistance. Prog Brain Res. 63:3-17 Hossmann K-A, Mies G, Paschen W, Csiba L, Bodsch W, Rapin JR, Poncin-Lafitte ML, Takahashi K (195) Multiparametric imaging of blood flow and metabolism after middle cerebral artery occlusion in cats. J Cereb Blood Flow Metab 5: Koizumi J, Yoshida Y, Nakazawa T, Ooneda G (196) Experimental studies of ischemic brain edema. 1. A new experimental model of cerebral embolism in rats in which recirculation can be introduced in the ischemic area. Jpn J Stroke : 1- Ljunggren B, Norberg K, Siesj6 BK (1974) nfluence of tissue acidosis upon restitution of brain energy metabolism following total ischemia. Brain Res 77: Lowry OH, Passonneau JV (1972) A Flexible System of Enzymatic Analysis. New York, Academic Press Lundy EF, Kuhn JE, Kwon JM, Zelenock GB, D'Alecy LG (197) nfusion of five per cent dextrose increases mortality and morbidity following six minutes of cardiac arrest in resuscitated dogs. J Crit Care 2:4-14 Memezawa H, Minamisawa H, Smith M-L, Siesj6 BK (1992a) schemic penumbra in a model of reversible middle cerebral artery occlusion in the rat. Exp Brain Res (in press) Memezawa H, Smith M-L, Siesj6 B-K (1992b) Penumbral tissues salvaged by reperfusion following middle cerebral artery occlusion in rats. Stroke (in press) Michenfelder JD, Sundt TM (1971) Cerebral ATP and lactate levels in the squirrel monkey following occlusion of the middle cerebral artery. Stroke 5: Myers RE (1979) Lactic acid accumulation as a cause of brain edema and cerebral necrosis resulting from oxygen depriva- J Cereb Blood Flow Metab, Vol. 12, No. 1, 1992

9 ENERGY METABOLTES N STROKE 33 tion. n: Advances in Perinatal Neurology (Korobkin R, Guilleminault G, eds), New York, Spectrum, pp Naritomi H, Sasaki M, Kanashiro M, Kitani M, Sawada T (19) Flow thresholds for cerebral energy disturbance and Na + pump failure as studied by in vivo 31p and 23Na nuclear magnetic resonance spectroscopy. J Cereb Blood Flow Metab : Nedergaard M (197) Transient focal ischemia in hyperglycemic rats is associated with increased cerebral infarction. Brain Res 4:79-5 Nedergaard M, Diemer NH (197) Focal ischemia of the rat brain, with special reference to the influence of plasma glucose concentration. Acta Neuropathol (Bed) 73: Nilsson B, Norberg K, Nordstrom C-H, Siesjo BK (1975) Rate of energy utilization in the cerebral cortex in rats. Acta Physiol Scand 93: Nowicki J, Assumel-Lurdin C, Duverger D, Mackenzie ET (19) Temporal evolution of regional energy metabolism following focal cerebral ischemia in the rat. J Cereb Blood Flow Metab : Obrenovitch TP, Garofalo, Harris RJ, Bordi L, Ono M, Momma F, Bachelard HS, Symon L (19) Brain tissue concentration of ATP, phosphocreatine, lactate, and tissue ph in relation to reduced cerebral blood flow following experimental acute middle cerebral artery occlusion. J Cereb Blood Flow Metab :66-74 Ponten U, Ratcheson RA, Salford LG, Siesjo BK (1973) Optimal freezing conditions for cerebral metabolites in rats. J Neurochem 21: Pulsinelli WA, Duffy TE (193) Regional energy balance in rat brain after transient forebrain ischemia. J Neurochem 4: Pulsinelli WA, Waldman S, Rawlinson D, Plum F (192) Moderate hyperglycemia augments ischemic brain damage: a neuropathological study in the rat. Neurology 32: Ratcheson RA, Ferrendelli JA (19) Regional cortical metabolism in focal ischemia. J Neurosurg 52: Rehncrona S, Rosen, Siesjo BK (191) Brain lactic acidosis and ischemic cell damage:. Biochemistry and neurophysiology. J Cereb Blood Flow Metab 1: Selman WR, VanDerVeer C, Whittingham TS, LaManna JC, Lust WD, Ratcheson RA (197) Visually defined zones of focal ischemia in the rat brain. Neurosurgery 21:25-3 Selman WR, Ricci AJ, Crumrine RC, LaManna JC, Ratcheson RA, Lust WD (199) The evolution of focal ischemic damage: a metabolic analysis. Metab Brain Dis 5:33-44 Siemkowicz E, Hansen AJ (191) Brain extracellular ion composition and EEG activity following 1 minutes ischemia in normo- and hyperglycemic rats. Stroke 12: Siesjo BK (191) Cell damage in the brain: a speculative synthesis. J Cereb Blood Flow Metab 1: Siesjo BK (19a) Mechanisms of ischemic brain damage. Crit Care Med 16: Siesjo BK (l9b) Acidosis and ischemic brain damage. Neurochem Patho9:31- Siesjo BK, Wie1ch T (195) Molecular mechanisms of ischemic brain damage: Ca2+ -related events. n: Cerebrovascular Diseases (Plum F, Pulsinelli W, eds), New York, Raven Press, pp 17-2 Smith M-L, Bendek G, Dahlgren N, Rosen, Wieloch T, Siesjo BK (194) Models for studying long-term recovery following forebrain ischemia in the rat. 2. A 2-vessel occlusion model. Acta Neural Scand 69:35-41 Smith M-L, Kalimo H, Warner DS, Siesjo BK (19) Morphological lesions in the brain preceding the development of postischemic seizures. Acta Neuropathol (Berl) 76: Tamura A, Graham or, McCulloch J, Teasdale GM (l9 a) Focal cerebral ischemia in the rat: 1. Description of technique and early neuropathological consequences following middle cerebral artery occlusion. J Cereb Blood Flow Metab 1:53-6 Tamura A, Graham or, McCulloch J, Teasdale GM (191b) Focal cerebral ischemia in the rat: 2. Regional cerebral blood flow determined by [14Cliodoantypyrine autoradiography following middle cerebral artery occlusion. J Cereb Blood Flow Metab 1:61-69 Welsh FA (194) Regional evaluation of ischemic metabolic alterations. J Cereb Blood Flow Metab 4: Zasslow MA, Pearl RG, Shuer LM, Steinberg GK, Lieberson RE, Larson CP (199) Hyperglycemia decreases acute neuronal ischemic changes after middle cerebral artery occlusion in cats. Stroke 2: J Cereb Blood Flow Metab. Vol. 12. No