Intracellular ph in the Brain Following Transient Ischemia

Transcription

1 Journal of Cerebral Blood Flow and Metabolism 3: Raven Press, New York Intracellular ph in the Brain Following Transient Ischemia Hideo Mabe, Photjanee Blomqvist, and Bo K. Siesjo Laboratory for Experimental Brain Research, University of Lund, Lund, Sweden Summary: The objective of the present study was to discover whether or not intracellular alkalosis develops in the brain in the recovery period following transient ischemia, Forebrain ischemia of IS-min duration was induced by four-vessel occlusion in rats, with recovery periods of IS, 60, and 180 min. Intracellular ph was derived both by the HC03- - H2C03 method and from the creatine kinase equilibrium. The ischemia was associated with energy failure and marked accumulation of lactic acid in the cerebral cortex. brought about rapid rephosphorylation of adenine nucleotides and gra- dual normalization of lactic acid levels. After 15 min of recovery, the HC H2C03 method indicated persisting acidosis, but the creatine kinase reaction did not. After 60 min, a shift of ph in the alkaline direction was demonstrated in both methods. This alkalosis had disappeared after 3 h of recovery. It is concluded that resumption of ATP production after ischemia is followed by a rapid rise in intracellular ph, which transiently increases above normal. Key Words: Energy metabolism Intracellular ph-ischemia-. It has been repeatedly observed that metabolic recovery following a period of transient brain ischemia often includes an increase above control levels of the phosphocreatine/creatine ratio (Ljunggren et a!., 1974a; Marshall et a!., 1975; Levy and Duffy, 1977; Nordstrom et a!., 1978a,b; Rehncrona et a!., 1981). Theoretically, such an increase may reflect either a change in the equilibrium constant of the creatine kinase (CK) reaction, e.g., due to an altered intracellular Mg 2 + concentration, or it may be due to intracellular alkalosis (see Ljunggren et al., 197 4a, b; Nordstrom et al., 1978a). The latter possibility derives from the fact that the equilibrium of the CK reaction is ph-dependent and that the reaction may, therefore, be used to estimate changes in intraceiiular ph (Kuby and Noltmann, 1962; Rose, 1968). This assumption was substantiated by experiments in which changes in intracellular ph (phi) in the brain Dr. Mabe is on leave of absence from Department of Neurosurgery, Nagoya City University, Nagoya, Japan. Address correspondence and reprint requests to Professor Siesjo at Laboratory for Brain Research, Floor EA-5, Lund Hospital, S Lund, Sweden. Abbreviations used: CK, Creatine kinase; Cr, creatine; ECF, extracellular fluid; PCr, phosphocreatine; ph;, intracellular ph; PtCOz, tissue CO2 tension. during hypercapnia were derived both from the HC H2C03 buffer system and from the CK reaction (Siesjo et a!., 1972). It has become increasingly evident that the degree of final cell damage that occurs after transient brain ischemia is influenced by postischemic metabolic events (for literature review, see Siesjo, 1981). This belief justifies continued efforts to define metabolic conditions prevailing in the recirculation -reoxidation period. In the present study, we induced ischemia of IS-min duration and derived phi during recirculation both from the HC H2C03 system and from the CK equilibrium. METHODS The study was performed on male Sprague-Dawley rats, weighing g. The animals were allowed food pellets (Astra-Ewos, SoderUilje, Sweden) and tap water until operation. Induction of ischemia The objective of the experiments precluded the induction of compression ischemia by means of the intracisternal infusion of an artificial cerebrospinal (CSF) solution used by Ljunggren et al. (1974a, b; see also Nordstrom et ai., 1978b). Furthermore, since we wished to avoid the systemic acidosis associated with decreased arterial blood pressure (see Nordstrom and Siesjo, 1978) 109

2 110 H. MARE ET AL. forebrain ischemia was induced using the method described by Pulsinelli and Brierley (1979). In summary, both vertebral arteries were electrocauterized through the alar foramina during halothane anesthesia delivered through a face mask. The dorsal neck incision was then closed and the animals were left to recover until the following day. They were then anaesthetized with about 3% halothane in oxygen, tracheotomized, and maintained during operation on 1% halothane and 70% N20 in oxygen. Catheters were placed in a femoral artery and a femoral vein. The tracheotomy incision allowed access to the common carotid arteries, which were dissected free from the vagus nerves and the sympathetic trunks. A midline skin incision was placed over the skull bone. Two electroencephalograph (EEG) electrode screws were placed bilaterally in the bone in the frontoparietal region, and a burr hole was placed over the superior sagittal sinus for sampling of cerebral venous blood. The neck muscles were reflected so as to expose the atlanto-occipital membrane for later sampling of CSF. Preparations were then made to insert a plastic funnel in the skin incision over the exposed skull bone for freezing the brain in situ. When the operative procedures had been completed, the halothane supply was discontinued and the animals were ventilated with 70% N20 and 30% O2 for at least 30 min before ischemia was induced. During that time, the arterial Pco2 was adjusted to mm Hg, P02 was maintained at 100 mm Hg or higher, and body temperature was regulated to 37 C. Ischemia of 15-min duration was then induced by occluding the carotid arteries with an atraumatic arterial forceps. Five groups of animals were studied, each consisting of seven animals. Control animals were maintained under anaesthesia for min before arterial and venous blood were sampled. The atlanto-occipital membrane was punctured to allow sampling of cisternal CSF. Brain tissue was frozen in situ with liquid nitrogen (for details of techniques, see Siesjo et ai., 1972; Ponten et ai., 1973). In one experimental group, sampling was done after 15 min of ischemia; in the remaining three groups, sampling took place after recovery periods of 15, 60, and 180 min, respectively, following removal of the carotid clamps. Analytical techniques Arterial Po2, Pco2, and ph were measured immediately after sampling with microelectrodes (for Po2, Eschweiler and Co., Kiel; for ph, Radiometer, Copenhagen), with due correction for any deviation of body temperature from 37 C. Cerebral venous Pco2 was similarly measured. Total CO2 levels of cisternal CSF and parietal cortex were measured by microdiffusion techniques (see Siesjo et ai., 1972). In four animals from each group, a portion of the parietal cortex was extracted with HCI-methanol at -22 C, and the samples were subsequently processed for enzymatic, fluorometric measurements of phosphocreatine (PCr), creatine (Cr), ATP, ADP, AMP, glucose, lactate, and pyruvate (for analytical techniques, see Folbergrova et ai., 1972). Calculations Tissue (and CSF) CO2 tensions were calculated by adding 1 mm Hg to the arithmetic mean of the arterial and cerebrovenous CO2 tensions (Ponten and Siesjo, 1966). The cerebrospinal fluid and intracellular HC03 - concentrations and ph values were calculated according to the equations: [Hcod ('SF = [ C02] CSF - PtC (1) [ ] [HC03-] CSF phcsf = log PtC (2) ( [ C02] t - PtC02' VECF [HC03-]CSF - Vbl [HC03-]bl HCO'l., I = (3) Vi [HC03] i PH = log (4) I PtC02' Here, [C02] is the total CO2 content (/Lmol g-l); PtC02 the tissue (and CSF) CO2 tension in mm Hg; 6.12 the pk' of carbonic acid; and the appropriate CO2 solubility coefficients (/Lmol g-i mm Hg-l); and V ECF, V bl' and Vi the volumes (as volume fractions) occupied by extracellular fluid (ECF), blood, and intracellular water, respectively (see Siesjo et ai., 1972). Since it was assumed that V ECF and V hi were 0.15 and 0.03, respectively, the value of Vi used was In calculating changes in phi from the CK equilibrium, the equation [ PCr ] [ A TP ] K' [ Cr] = [ADP] [H+] was first solved for K', using the concentrations of PCr, Cr, ATP, ADP, and H+ obtained in the control material (see below). This K' value was then inserted in the equation, together with the appropriate values for PCr, Cr, ATP, and ADP, to obtain intracellular H+ (phi) in the postischemic recovery groups. Statistical differences between the groups were calculated with Student's t test. RESULTS In all ischemic and recovery animals included in the series, carotid artery clamping was associated with cessation of spontaneous EEG activity, and the record remained "isoelectric" throughout the period of ischemia. The results (Table 1) show that body temperature was similar among the groups, that mean arterial blood pressure was 130 mm Hg or higher, and that P02 exceeded 100 mm Hg. Arterial Pco2 was in the range of mm Hg, except in the ISO-min recovery group, which had a slightly elevated CO2 tension. Plasma ph remained above 7.40 in all groups. The arteriovenous Pco2 difference was clearly reduced after 15 min of recirculation and increased after 60 min (see below). ) (5) J Cereb Blood Flow Me/abol, Vol. 3, No. I, 1983

3 ISCHEMIA AND BRAIN ph III TABLE 1. Physiological parameters in control animals after 15 min o.f ischemia, and after 15, 60, and 180 min of recirculation (following 15 min of ischemia) in other test animals Ischemia (15 min) Control Ischemia (15 min) (15 min) (60 min) (180 min) Temperature ec) 37.4 ± ± 0.2 MABP(mm Hg) 140 ± ± 7 Pao2 (mm Hg) 114 ± ± 4" Paco2 (mm Hg) 36.9 ± ± 1.4 PVco2 (mm Hg)a 43.9 ± 1.5 PVco2 - Paco, (mm Hg) 7.1 ± 1.0 ph 7.45 ± ± ± ±3 111 ± ± ± ± 0.ge 7.42 ± ± ±3 118 ± ± ± ± 0.8e 7.44 ± ± ±4 112 ± ± ± ± ± 0.02 a The venous samples were taken from the superior sagittal sinus. b p < e p < Values are means ± SEM; n = 7 in all groups. Table 2 lists the calculated PtC02 values, the measured CSF and tissue CO2 content, and the calculated CSF and intracellular HC03 - concentrations and ph values in the control and the recovery groups. Cerebrospinal fluid ph values remained constant in all recovery groups. Calculated intracellular ph was significantly reduced after 15 min, and significantly increased after 60 min, whereas the l80-min value was close to control. It should be noted, though, that PtC02 in that group was increased above control (see also CSF HC03- concentration). Tissue concentrations of labile metabolites are given in Table 3. The values obtained in the ischemic group showed the expected derangement of cerebral energy state and an increase in tissue lactate content to about 20 /Lmol g-'. All recovery animals had ADP and AMP concentrations similar to (or lower) than control values demonstrating adequate reoxygenation. The ATP concentration (and thereby the sum of adenine nucleotides) gradually increased toward normal during the 180-min recirculation period. The lactate content was still elevated after 15 min of recirculation, but had normalized after 60 and 180 min. The PCr concentration was not reduced below control after 15 min, but it was significantly increased after 60 min of recirculation, a normal value again being observed in the 180-min group. Changes in Cr concentrations were reciprocal, but the values were more variable. Figure 1 compares the phj values derived from the HC03 --H2C03 system and those calculated from the CK equilibrium. In drawing the figure, we assumed that phj during ischemia fell to 6. 5 (see Ljunggren et ai., 1974a; Siesjo, 1978, chap. 10). A disparity between the ph values derived using each TABLE 2. Acid-base variables in cisternal CSF and parietal cortex, and calculated intracellular HCo.,- concentrations and ph values in control animals and in those allowed recovery periods of /5,60, or 180 min following 15 min o.f forebrain ischemia CSF Cortical tissue Intracellular fluid [CO2] (J.'mol g-l) [HC03-] (J.'mol g-l) [CO,] [HC03-] Peo, [HC03-] ph (J.'mol g-l) (J.'mol g-l) (mm Hg) (J.'mol g-l) ph Control 15 min 60 min 180 min ± ± ± ± ± ± ± 0.41' ± 0.37' 7.44 ± ± ± ± ± ± b b 39.8 ± ± 1.26b 6.70 ± 0.09b a b 41.9 ± ± 0.68b 7.15 ± 0.02b a a 46.4 ± ± ± 0.02 a p < b P < c P < Values are means ± SEM; n = 7 in all groups. J C{'I'eh Blood FIOII' Me/ahol. Vol. 3. No. I. 1983

4 112 H. MABE ET AL. TABLE 3. Concentrations of PCr, Cr, ATP, ADP, AMP, lactate, and pyruvate, as well as calculated values for adenine nucleotide pool ('Z.Ad), adenylate energy charge (EC) and lactate/pyruvate ratio in control animals, during ischemia, and at 15, 60, and 180 min of recirculation following 15 min of ischemia Ischemia (15 min) Control Ischemia (15 min) (15 min) (60 min) (180 min) PCr Cr ATP ADP AMP 'Z.Ad EC Lactate Pyruvate Lactate/Pyru vate 4.48 ± ± ± ± ± ± ± ± ± ± ± 0.17' 0.35 ± 0.11" ± 0.052' 1.25 ± 0.12" 2.35 ± 0.06' ± e ± 3.71b ± ± 52a 4.10 ± ± 0.09' 4.86 ± 0.12a 6.95 ± ± 0.18a 6.61 ± ± 0.02' 2.02 ± 0.03' 2.32 ± O.03a ± 0.007b ± ± ± ± ± ± 0.02e 2.32 ± 0.04' 2.64 ± 0.02a ± ± ± ± l.13b 1.75 ± ± ± 0.023b ± ± ± ± ± 3.5 a p < bp < O.01., P < Values are means ± SEM (mmoles/kg wet weight). method was observed after 15 min of recirculation, but good agreement was obtained after 60 and 180 min. DISCUSSION Before the results on phi changes are discussed, it seems necessary to consider two other findings. One concerns the changes in the arteriovenous Pco2 difference. After 15 min of recovery, this difference was below 5 mm Hg in six of seven animals, and, after 60 min, the difference exceeded 11 mm Hg in six of seven animals. The results suggest that the ratio of CBF to metabolic rate was increased after 15 min and reduced after 60 min. This suggestion is clearly in agreement with results that show that ischemia is followed by an initial hyperemia, followed by a delayed hypoperfusion (see Pulsinelli et ai., 1982, for data on this model of ischemia). It has been clearly demonstrated that ischemia leads to a reduction in the size of the adenine nucleotide pool due to deamination and dephosphorylation of AMP and to loss of nucleosides and bases (see Siesjo, 1978 for literature and further discussion). In the present study, the pool size was reduced by about 25% during ischemia, yet prolongation of the recirculation period to 180 min led to virtually complete resynthesis of adenine nucleotides. Such recovery has not previously been observed after 15 min of complete ischemia (Ljunggren et ai., 1974b; Levy and Duffy, 1977). It remains to be shown, therefore, if the type and degree of ischemia induced in the present study are less detrimental to cellular viability than is complete ischemia. It has been shown previously that brain extracellular ph normalizes within min following 10 min of complete ischemia in the rat (Siemkowicz and Hansen, 1981), but information on intracellular o 15/0 15 I.. 60 ReCirculation 180min. FIG. 1. Comparison between intracellular ph (ph;) in the brain following 15 min of forebrain ischemia as calculated by the HC H2C03 method (filled circles) and from the CK equilibrium (open circles). It was assumed that ph; fell to 6.5 during the ischemia (dashed line). Values are means ± SEM; "p < 0.01 (significant difference between postischemic and control values). J Cereb Blood Flow Metahol. Vol. 3, No. J. 1983

5 ISCHEMIA AND BRAIN ph II3 ph is not available. When the present acid-base results are considered, it should be recalled that both methods used to derive phj are based on certain assumptions, the validity of which must be scrutinized whenever the methods are applied to pathological conditions. For the HC H2C03 method, which requires a steady state, the critical assumptions concern the size of the ECF pool and the extent to which the CSF HC03 - concentration reflects the HC03 - concentration of the ECF proper. In all probability, the assumption of an invariant ECF volume is not critical if the method is applied to recovery periods of the duration used in this study. Thus, although ischemia is accompanied by a marked decrease in ECF volume, recirculation promptly leads to normalization of the volume (Hossmann et ai., 1977). The second assumption is more critical. Thus, since forebrain ischemia of this type leads to relatively moderate ischemia in brainstem and cerebellum (Pulsinelli et ai., 1982) the acidosis of cisternal CSF must be less pronounced than that of parietal cortex ECF. This was also evident from the fact that CSF sampled at the end of 15 min of ischemia showed a decrease in the HC03- concentration by only 5 ILmol ml-1 (data not shown). It seems likely, therefore, that the correction for the ECF HC03 - content gave rise to an underestimation of the intracellular HCO:l- concentration after 15 min of recirculation, and that the true phj at that time was higher than what was estimated. There also exists direct evidence that the ph of cortical ECF has not normalized after 15 min following the start of recirculation (Siemkowicz and Hansen, 1981). In other words, the phj derived from the CK equilibrium may be closer to the true value. We tentatively conclude, though, that equilibrium in the HC03 --H2C03 system was at hand after 60 and 180 min of recirculation. Derivation of phj changes from the CK equilibrium requires that the compartmentation of the reacting species does not interfere with the results, and that the free Mg 2 + activity does not change. The rapid normalization of the phosphorylation state of the tissue in the recovery period inspires some confidence in the phj values derived. Perhaps the most important result is the agreement obtained between the two methods in the 60- and 180-min recovery groups. This agreement, and the similarity in results previously obtained with the two methods in hypercapnia (Siesjo et ai., 1972), provide strong evidence that recirculation following transient ischemia leads to rapid normalization and a subsequent alkaline shift of intracellular ph in the brain. At first glance, the present results appear to be supported by those reported by Kogure et al. (1980), who induced unilateral brain ischemia in rats by a microembolization procedure and estimated cellular ph by a histochemical technique (based on color transition of neutral red) after postembolization periods of 3 to 30 min. The authors reported normal ph values of , a drop in ph to below 6.5 after 3 min, and areas with ph values of after 30 min. However, these results differ from the present ones, which pertain to the recovery period after transient ischemia characterized by extensive restoration of cerebral energy state and tissue lactate levels. Thus, in the study of Kogure et al. (1980) the markedly alkalotic values were observed in tissue areas with persisting ischemia, low A TP values, and grossly elevated lactate concentrations. Assuming a normal Pco2, the reported values imply that intracellular bicarbonate must have increased from control values of Lmol g-i to above 25 ILmol g-l. It remains to be shown how such an accumulation can occur in the absence of an adequate energy source. The present results, and previous ones demonstrating an increase in the tissue PCrfCr ratio in the postischemic period (see above) suggest that recirculation is accompanied by rapid disappearance of the acidosis arising during ischemia, and that transient intracellular alkalosis precedes the final return of phj to normal. Provided we are allowed to use the CK equilibrium to calculate phj, we can use the present results and those previously published (Ljunggren et ai., 1974a,b; Marshall et ai., 1975; Nordstrom and Siesjo, 1978; Nordstrom et ai., 1978a; Rehncrona et ai., 1981) to conclude the following. First, provided that ischemia is followed by rapid restoration of cerebral energy state, phj may return to normal within about 15 min, i.e., at a time when tissue lactic acid concentrations are still raised above control levels. Second, intracellular alkalosis may occur following ischemic periods as short as 5 min, and the increase in phj may exceed 0.1 units if the period of ischemia is longer (15-30 min). Third, alkalosis still exists at 90 min following the start of recirculation, at least when the preceding ischemia lasts for min. It should be recalled that intracellular alkalosis of the present magnitude occurs in but few conditions. Thus, although an increase in phj is observed in hypocapnia, its severity is considerably curtailed by.i Ccreb Blood FlOlI' Metabol, Vol. 3, No.1, 1983

6 114 H. MABE ET AL. enhanced lactic acid production (MacMillan and Siesj6, 1973a; Pelligrino et al., 1981) unless deep barbiturate anaesthesia is employed (MacMillan and Siesj6, 1973b). Barbiturate anaesthesia of itself induces moderate intracellular alkalosis, possibly because of an associated retardation of glycolytic flux with an ensuing "consumption" of metabolic acids (Nilsson and Siesj6, 1970; MacMillan and Siesj6, 1973a; Anderson et al., 1980). Finally, an increase in phi has been noted in hypoglycemia (Pelligrino et al., 1981), an increase that is grossly exaggerated by combined hypocapnia and moderate hypotension (Pelligrino and Siesjo, 1981). Although postischemic alkalosis seems to be an established phenomenon, its cause is unknown and its pathogenetic importance, if any, cannot be predicted. Possibly, resumption of tissue circulation and ATP production initiates active efflux of H+ from cells at a time when lactate levels are still high. Since oxidation of the remaining lactate will remove a stoichiometric amount of H+ ions, overcompensation for the acidosis would result. In other words, an acidosis due to lactic acid production will be fully compensated for when the accumulated lactic acid is again oxidized; accordingly, if other ph-regulating mechanisms are brought into play, transient overcompensation may result. Predictably, this overcompensation may be more pronounced, and perhaps more protracted, if excessive amounts of lactic acid accumulate during the ischemia. Acknowledgment: This study was supported by grants from the Swedish Medical Research Council, from US PHS, and from the Swedish Work Environment Fund. REFERENCES Anderson RE, Michenfelder JD, Sundt TM Jr (1980) Brain intracellular ph, blood flow, and blood-brain barrier differences with barbiturate and halothane anesthesia in the cat. Anesthesiology 52: Folbergrova J, MacMillan V, Siesjil BK (1972) The effect of moderate and marked hypercapnia upon the energy state and upon the cytoplasmatic NADH/NAD+ ratio of the rat brain. J Neurochem 19: Hossmann K-A, Sakaki S, Zimmerman V (1977) Cation activities in reversible ischaemia of the cat brain. Stroke 8:77-81 Kogure K, Busto R, Schwartzman RJ, Scheinberg P (1980) The dissociation of cerebral blood flow, metabolism and function in the early stages of developing cerebral infarction. Ann NeuroI8: Kuby SA, Noltmann EA (I 962) A TP-creatine transphosphorylase. In: The Enzymes (Boyer PD, Lardy H, Myrback K, eds), New York, Academic Press, Vol 6, pp Levy DE, Duffy TE (1977) Cerebral energy metabolism during transient ischemia and recovery in the gerbil. J Neurochem 28:63-70 Ljunggren B, Ratcheson RA, Siesjil BK (1974a) Cerebral metabolic state following complete compression ischemia. Brain Res 73: Ljunggren B, Norberg K, Siesjil BK (1974b) Influence of tissue acidosis upon restitution of brain energy metabolism following total ischemia. Brain Res 77: MacMillan V, Siesjil BK (1973a) The influence of hypocapnia upon intracellular ph and upon some carbohydrate substrates, amino acids and organic phosphates in the brain. J Neurochem 21: MacMillan V, Siesjil BK (1973b) The effect of phenobarbitone anaesthesia upon some organic phosphates, glycolytic metabolites and citric acid cycle-associated intermediates of the rat brain. J Neurochem 20: Marshall LF, Welsh F, Durity F, Lounsbury R, Graham DI, Langfitt TW (1975) Experimental cerebral oligemia and ischemia produced by intracranial hypertension. Part 3: Brain energy metabolism. J Neurosurg 43: Nilsson L, Siesjil BK (1970) The effect of anesthetics on the tissue lactate, pyruvate, phosphocreatine, ATP and AMP concentrations, and on intracellular ph in the rat brain. Acta Physiol Scand 80: Nordstrilm C-H, Siesjil BK (1978) Effects of phenobarbital in cerebral ischemia. Part I: Cerebral energy metabolism during pronounced incomplete ischemia. Stroke 9: Nordstrilm C-H, Rehncrona S, Siesjil BK (1978a) Restitution of cerebral energy state, as well as of glycolytic metabolites, citric acid cycle intermediates and associated amino acids after 30 min of complete ischemia in rats anaesthetized with nitrous oxide or phenobarbital. J Neurochem 30: Nordstrilm C-H, Rehncrona S, Siesjil BK (l978b) Effects 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: Pelligrino D, Siesjil BK (1981) Regulation of extra- and intracellular ph in the brain in severe hypoglycemia. J Cereb Blood Flow Metabol 1 :85-96 Pelligrino D, Almquist L-O, Siesjil BK (1981) Effects of insulin-induced hypoglycemia on intracellular ph and impedance in the cerebral cortex of the rat. Brain Res 221: Ponten U, Siesjil BK (1966) Gradients of CO2 tension in the brain. Acta Physiol Scand 67: Ponten U, Ratcheson RA, Salford LG, Siesjil BK (1973) Optimal freezing conditions for cerebral metabolites in rats. J Neurochem 21: Pulsinelli W, Brierley JB (1979) A new model of bilateral hemispheric ischemia in the unanesthetized rat. Stroke 10: Pulsinelli WA, Levy DE, Duffy TE (1982) Regional cerebral blood flow and glucose metabolism following transient forebrain ischemia. Ann Neurol 11: Rehncrona S, Rosen J, Siesjil BK (1981) Brain lactic acidosis and ischemic cell damage: I. Biochemistry and neurophysiology. J Cereb Blood Flow Metabol 1 : Rose IA (1968) The state of magnesium in cells as estimated from the adenylate kinase equilibrium. Proc Natl Acad Sci USA 61: Siemkowicz E, Hansen AJ (198 I) Brain extracellular ion composition and EEG activity following 10 minutes ischemia in normo- and hyperglycemic rats. Stroke 12: Siesjil BK (1978) Brain Energy Metabolism. New York, John Wiley & Sons Siesjil BK (1981) Cell damage in the brain: A speculative synthesis. J Cereb Blood Flow Metabo[ I: Siesjil BK, Folbergrova J, MacMillan V (1972) The effect of hypercapnia upon intracellular ph in the brain, evaluated by the bicarbonate-carbonic acid method and from the creatine phosphokinase equilibrium. J Neurochem 19: J Cereb Blood Flow Me/abol. Vol. 3. No. I. /983