METABOLITE LEVELS IN BRAIN FOLLOWING EXPERIMENTAL SEIZURES: THE EFFECTS OF MAXIMAL ELECTROSHOCK AND PHENYTOIN IN CEREBELLAR LAYERS

Abstract— The effects of maximal electroshock (MES) and phenytoin on metabolites and cyclic nucleotides in layers of frozen-dried cerebellum have been investigated. The four layers (molecular, Purkinje-cell rich, granular and white matter) had remarkably homogeneous distributions of P-creatine, ATP, glucose, glycogen, lactate, GABA and the cyclic nucleotides. MES caused dramatic decreases in P-creatine, ATP, and glucose at 10 s after treatment, followed by a decrease in glycogen at 30 s. Lactate levels were elevated, and GABA was unchanged. Cyclic AMP concentrations were increased at 10s and cyclic GMP at 30 s. Phenytoin modified most of the MES induced changes in all the layers, although white matter was less affected by MES and/or phenytoin. Lactate concentrations were increased by MES and these effects were not altered when phenytoin was administered. The most dramatic effects of phenytoin were on the changes in cyclic nucleotides. Cyclic AMP concentrations were elevated after MES but the values returned to normal more rapidly when phenytoin was present. The drug almost obliterated the MES induced changes in cyclic GMP. The possible relationship of cyclic nucleotide concentrations and the modulation of seizure activity is discussed.     

Glycogen, ammonia and related metabolities in the brain during seizures evoked by methionine sulphoximine

Abstract— The levels of ATP, P-creatine, glucose, glycogen, lactate, glutamate and ammonia were measured in mouse brain after administration of the convulsive agent methionine sulphoximine (MSO). No changes were observed in ATP and P-creatine levels either before or during the seizures. Lactate levels were unchanged until the onset of seizures (4–5 hr) at which time the levels increased an average of 65 per cent. Glucose and glycogen levels increased progressively. Just before the onset of seizures the levels had increased 95 and 62 per cent, respectively. During the seizures both substances had increased a total of 130 per cent. Comparable changes were found in cerebral cortex, cerebellum and subcortical forebrain. Through the use of quantitative histochemical methods it was found that the greatest increases in glycogen occurred in layers I and III (layers II and IV were not analysed). Progressively smaller changes were found in layers V and VI and no increase at all was found in the subjacent white matter. Glucose, in contrast to glycogen, increased to about the same degree in all cerebral layers and in subjacent white matter. The increase in glycogen after MSO administration may be related to the fact that MSO also causes an increase in the ratio of brain to serum glucose levels. This would indicate that an increase in intracellular glucose had occurred. Ammonia levels were increased 300–400 per cent in both cerebrum and cerebellum. A time study in cerebellum showed that the increase begins early and reaches maximal levels long before the onset of seizures. Glutamate levels were reduced by small but statistically significant amounts in both cerebrum and cerebellum. Administration of methionine sulphoximine completely prevented seizures and the increase in lactate, but did not prevent the increases in glycogen and glucose. The rise in ammonia was reduced but not prevented. During 20 sec of complete ischaemia (decapitation) ATP, P-creatine and glucose fell somewhat more rapidly than normal in brain of animals undergoing MSO seizures. From the changes it was calculated that the metabolic rate had been increased about 20 per cent by the seizure. A new sensitive and specific enzymic method for determination of tissue ammonia is presented together with evised enzymic procedures for lactate and glutamate.     

Changes in metabolites of the energy reserves in individual layers of mouse cerebral cortex and subjacent white matter during ischaemia and anaesthesia

The distribution of glucose, glycogen, ATP, P-creatine and inorganic phosphate was measured in layers I, III, IV, V and VI of cerebral cortex and subjacent white matter of mouse brain. ATP, P-creatine and inorganic phosphate were evenly distributed in all regions examined, whereas levels of glucose and glycogen were higher in white matter than the average for the other layers. Anaesthesia increased levels of glucose and P-creatine in layers I and V and subjacent white matter (other layers were not examined). Anaesthesia doubled the level of glycogen in molecular layer I with lesser increases in layers III, IV, V and VI, but with no change in white matter from the unanaesthetized control value. The metabolic rates in the individual layers were estimated from the rates of expenditure of energy reserves during total ischaemia. In non-anaesthetized mice, white matter had a higher metabolic rate than either layer I or V. Anaesthesia reduced the metabolic rates in all layers; however, the largest reduction occurred in subjacent white matter (86 per cent), with reductions of 54 per cent and 76 per cent respectively in layers I and V.     

LIGHT-INDUCED ALTERATIONS IN RETINAL PYRUVATE AND HIGH-ENERGY PHOSPHATES, IN VIVO

—The effect of illumination upon some metabolic substrates of frog retina was investigated in vivo, using conditions of illumination for which electrophysiological correlates in the retina are well defined. Frogs were frozen immediately after illumination, the tissue was processed for quantitative histochemistry, and the compounds were measured fluorometrically. Levels of P-creatine were lower in flash-illuminated retinas than in either dark- or light-adapted retinas. The high-energy phosphates and pyruvate changed rapidly upon exposure to flashing light, then returned towards the original steady-state level, with ATP preceding pyruvate and P-creatine. ATP and P-creatine were primarily concentrated in the bipolar and ganglion cell layers. The energy reserve of the retina was depleted by an enhanced rate of neural activity in vivo. Levels of P-creatine and ATP decreased in only those cellular layers which initiate neural action potentials. These data suggest that the mechanisms of neural excitation are closely coupled to energy and glucose metabolism in the retina.     

GLYCOGEN AND GLYCOGEN PHOSPHORYLASE IN THE CEREBRAL CORTEX OF MICE UNDER THE INFLUENCE OF METHIONINE SULPHOXIMINE

Abstract— Glucose and glycogen levels in the mouse cerebral cortex in vivo were studied after recovery from methionine sulphoximine seizures. The animals appeared normal 24 h after methionine sulphoximine administration but both glucose and glycogen still persisted at higher levels 72 h after injection (by 64 and 275 per cent, respectively). When seizures were prevented by methionine, the increase in glucose and glycogen at the longer time intervals was significantly smaller than in animals treated with methionine sulphoximine only; glucose reached normal values at 48 or 72 h; the accumulation of glycogen was reduced by about three to five times, but after 72 h the levels were still significantly higher than in control animals (67 or 32 per cent increase, depending on the administered dose of methionine). In contrast to the considerable accumulation of glycogen after administration of methionine sulphoximine in vivo, it had no effect on the level of glycogen in brain cortex slices in vitro. After 3 h incubation in the absence of methionine sulphoximine, glycogen was resynthesized to a level of about 4 μmol/g wet tissue and this value was not significantly affected by the presence of various concentrations of methionine sulphoximine in the incubation medium (10^-5 to 10^-2 M). The total (a+b forms) phosphorylase activity of mouse cerebral cortex in vivo after methionine sulphoximine administration was not affected. The fraction of active phosphorylase was reduced by about 50 per cent at the time of seizures. When seizures were prevented by methionine, the decrease in active phosphorylase was also completely prevented. In the preconvulsive period (1-2 h) and after recovery from the seizures (48 h after methionine sulphoximine administration) active phosphorylase was normal. The possible mechanisms involved in the increased accumulation of glycogen after methionine sulphoximine administration are discussed.     

GLYCOLYTIC SUBSTRATE UTILIZATION AND ENERGY CONSUMPTION IN THE CEREBRAL HEMISPHERES OF THE CHICK EMBRYO DURING THE PERIOD OF EEG DEVELOPMENT

Abstract— The glycolytic substrate utilization and energy consumption of chick embryo hemispheres aged 15 and 19 days, respectively, were studied by converting the heads into a closed system. After transforming the brain into a closed system by decapitation, spontaneous bio-electrical potential changes, as indicated by EEG measurements, stayed normal in the hemispheres until at least 30 s after the procedure. Although the two developmental stages studied represented the start and the final phase of functional development, no changes in carbohydrate utilization and energy consumption were observed when comparing hemispheres of the two stages mentioned. The utilization of substrates other than glucose during embryological development is discussed. The disappearance of the EEG is not related to the overall depletion of high energy phosphates in the hemispheres since ATP is still present at nearly its original level after 2 min of ischaemia, despite the low level of P-creatine occurring at 30 s after decapitation. A compartmentation of the major pools of P-creatine and ATP is suggested.     

CEREBRAL ENERGY METABOLISM DURING TRANSIENT ISCHEMIA AND RECOVERY IN THE GERBIL1

Energy metabolism was studied in the cerebral cortex of gerbils during and following ischemia induced by 1 h of unilateral carotid artery occlusion. An aneurysm clip was applied to the right common carotid artery of 50-70 g gerbils under brief halothane anesthesia, and the clip was removed 1 h later. Clinical state (gait, responsiveness, seizures) was evaluated during carotid occlusion, and 40% of the animals showed clinical evidence of stroke. Cortical energy stores (2 ATP + ADP + P-creatine) were more than half depleted in the ipsilateral cortex of clinically-affected gerbils, and glucose fell by 75%; lactate rose over 7-fold in the same specimens. After release of the carotid clip, clinical state improved, and biochemical abnormalities partially resolved. However, even after 24 h, the concentration of ATP and the total pool of adenine nucleotides remained subnormal. Metabolic activity in the ischemic cortex, assessed as the utilization of high-energy phosphates following decapitation, was normal after 1 h of recovery and decreased (-50%) after 24 h but was increased by more than 50% after 4 h. Cerebral glucose utilization, evaluated from autoradiographs prepared after intravenous administration of 2-[1-¹⁴C]deoxyglucose, was also increased in the cortex, hippocampus, and thalamus after 4 h of recovery. This post-ischemic hypermetabolism in tissue damaged by ischemia may identify a critical period for cell repair, when therapy could be decisive.     

Changes in glucose and energy metabolism in Vivo in developing retinae from visually-competent (DBA-1J) and mutant (C3H-HeJ) mice

Abstract— The distribution in vivo of glucose and lactate between the complete or sub- divided retina and the blood has been evaluated in DBA and C3H mice during postnatal development. Levels in vivo of several intermediates of glucose and energy metabolism were measured by enzyme-linked fluorometric assays of freeze-dried retinae; glucose and lactate were determined in freshly-drawn plasma. DBA retinae. During the first 20 days of postnatal life, the level of glucose in the plasma rose slightly while that in the retina declined: during this period the level of lactate in the plasma rose and became nearly equal to that in the retina. Changes during development in levels of glucose and glycogen were consistent with the interpretation that the rate of utilization of glucose in vivo is enhanced during early postnatal life. C3H retinae. The levels of glucose and glycogen in vivo were abnormally high throughout the developmental period, whereas levels of lactate were normal. The rise in levels of glucose after the 15th postnatal day was not related to an increase in blood levels of glucose but rather to a decreased utilization of glucose during this period. For the first 10 postnatal days the content of glucose, lactate, ATP and P-creatine within the photoreceptor layer of C3H retinae were within normal limits. Then, biochemical changes occurred which were secondary to ultrastructural pathology in the photoreceptors. This observation suggested that glucose metabolism and energy production are not involved in the primary aetiology of the inherited disease.     

An examination of the efficiency of glucose and glutamine as energy sources for cultured chick pigment epithelial cells

Embryonic chick pigment epithelial cells in culture require glucose as their major energy source for long-term growth, pigment formation, and colony organization. Cell number increases with glucose concentration at least up to 5.0 mM. Cells can be grown with glutamine as the major energy source but produce comparable cell numbers for only the first 3 days in culture, after which they cease growing. However, they are able to metabolize glutamine at a two to sixfoid higher rate than cells grown in the presence of glucose as measured by CO₂ release and by incorporation into protein. In cells grown in the presence of both glucose and glutamine, basal ATP levels were 31.1 nmoles/mg protein; P-creatine averaged 15.2 nmoles/mg protein and showed marked variability between experimental groups. During starvation, P-creatine levels fell while ATP levels remained relatively constant. Glucose was required for the recovery of P-creatine to prestarvation levels when measured 5 min after refeeding. Because of these marked changes in P-creatine concentration as a function of nutritional status, the ATP/P-creatine ratio becomes a useful measure of the energy state of the cell.     

POST-ISCHEMIC CHANGES IN CERTAIN METABOLITES FOLLOWING PROLONGED ISCHEMIA IN THE GERBIL CEREBRAL CORTEX

-Eight metabolites were measured in the post-ischemic period following either 1 or 3 h of unilateral ischemia in the gerbil cerebral cortex. The levels of ATP, P-creatine, glucose, glycogen and GABA were essentially restored by 1 h after ischemia. In the 3 h ischemic animals. glycogen continued to increase to greater than control values aftcr 5 and 20 h of recirculation. The Icvels of glutamate were unchanged during the ischemic episode, but decreased to 60% of control at Smin and 1 h after either period of ischemia. The concentrations of cyclic AMP, which were 4-to 5-fold elevated during ischemia. increased an additional 6-fold 5 min after recirculation in both groups. Arter 1 h of recovery. the levels were not different from control values. After the 1 h ischemic period, lactate levels recovered between 5 and 20 h of recirculation. In the 3 h ischemic animals. lactate concentrations were still elevated even after 20 h of recirculation. These data suggest that with the exception of lactate. recovery of metabolites is not sevcrely compromiscd by either 1 or 3 h of ischemia. Furthermore, the changes in glycogen. glutamate and cyclic AMP after recirculation suggest that the recovery process is not just a rcversal of the changes observed during ischemia.     

An Improved Method for In Situ Freezing of Cat Brain for Metabolic Studies

This study introduces a new method for rapid freezing of the cat brain. The method employed a Styrofoam box which was fitted around the head of the animal. Liquid nitrogen was poured into the box until the head was submerged. Temperature changes in three brain sites (ventral hypothalamus, the fourth ventricle, and the corpus callosum) and levels of labile carbohydrate metabolites (glycogen, glucose, ATP, P-creatine, and lactate) in five brain regions (cortex, thalamus, midbrain, cerebellum, and pons) frozen by the box method were compared with those frozen by a conventional cup method in which liquid nitrogen was poured into a hollow Styrofoam cup placed on top of the skull. The box method shortened the time of arrival of the freezing front and improved the freezing rate. The time required to bring the tissue to -20 degrees C was shortened, from 20 min at the ventral hypothalamus and 10-12 min at the fourth ventricle with the cup method, to less than 5 min at both sites with the box technique. Continued perfusion of brainstem prior to freezing was demonstrated. Levels of metabolites frozen by either method were similar. Lactate levels in any of the five brain regions studied by either method were not elevated, indicating no ischemic change.(ABSTRACT TRUNCATED AT 250 WORDS)     

Effects of duration of convulsions on energy reserves of the brain

Abstract— Rapid administration (0·4 ml in 1 sec) of the convulsant inhalant, flurothyl (hexafluorodiethyl ether, indoklon), to mice induced within 10 sec clonic-tonic seizures that were accompanied by marked decrease of P-creatine, decrease of ATP and glucose, and increase of lactate in the cerebral cortex. In contrast, mice to which flurothyl had been administered slowly (0·05 ml every 30 sec for 10 min) exhibited myoclonic jerks after about 3 min, merging into irregular clonus at about 5 min, and intermittent clonus thereafter until onset of tonic hind limb extension at about 10 min. In these mice, P-creatine in cerebral cortex decreased gradually for 6 min and then remained through 10 min at levels nearly as low as those reached 10 sec after rapidly administered flurothyl. Lactate in cerebral cortex increased much more during the 10 min of slow administration of flurothyl than in 10 sec of rapid administration, the greatest increase occurring at 4–6 min, as myoclonic jerks merged into irregular clonus. Glucose and ATP in cerebral cortex fluctuated somewhat but did not decrease greatly when flurothyl was administered slowly.     

The effect of topical application of diethyl- -fluoroglutarate on the metabolism and electrical activity in vivo of cat cerebral cortex

Abstract— The effect of topical application of diethyl-α-fluoroglutarate on the primary response of cat cerebral cortex to thalamic stimulation was investigated, and samples of the involved tissue and of homologous contralateral control tissue were removed at appropriate times for biochemical analyses. Changes in electrocortical response were first noted 30–40 min after topical application of 10 μ1 (52.7 μmol) of the drug to the pericruciate gyrus. A rapid reduction in amplitude of the surface negative component of the primary response was observed initially, followed by amplitude reduction and rapid disappearance of the primary response throughout the cortical field within a few minutes after the change first observed at the surface. The effects were interpreted either as a direct action of the drug on the somadendritic membrane or an inhibition of the excitatory synaptic impingement. Analyses of the tissues removed at the time of maximum electrocortical response indicated profound metabolic changes, including depletion of energy reserves and several Krebs cycle intermediates. Large increases in tissue levels of glucose, glucose 6-phosphate, and pyruvate were found. Changes in amino acids comprised depletion of glutamate with increased tissue levels of aspartate, GABA, NH₃, threonine, serine and alanine. Tissues were also removed at 10 min after topical application of the drug but before the advent of electrocortical changes. Decreased levels of glutamate were associated with a rise in tissue aspartate. Tissue levels of the other amino acids were unchanged. Glucose, glucose 6-phosphate and pyruvate levels were increased, lactate and ATP levels were unchanged, and P-creatine, α-KG and malate levels were reduced. We believe that the observed pattern of changes represents responses of the cerebral cortex to either a block in the synthesis of glutamate or in oxidation of pyruvate, with consequent interference with oxidative and energy metabolism and eventual depression of cortical electrical activity.     

PITFALLS IN THE USE OF RAPID FREEZING FOR STOPPING BRAIN AND SPINAL CORD METABOLISM IN RAT AND MOUSE

The rate at which the metabolism is stopped by means of freezing in Freon-12 (–150°C) was studied in various areas of the rat and mouse CNS, using changes in temperature and levels of glucose and lactate as parameters for this rate. The rat cerebral cortex was frozen after 0.5 min while the hypothalamus reached 0°C after more than 1.5 min. The skin on the skull was found to be the most important temperature isolator for the cortex. Substrate levels can be studied in this area only if this piece of skin is removed previously. In the mouse, the cerebral cortex was frozen after 6 s, the hypothalamus after 0.5 min. The lumbar level of the mouse spinal cord was frozen after 15 s, the cervical level only after 47 s. Liquid nitrogen alone cooled the mouse cerebral cortex at least as fast as did Freon at its melting point. A gradual decrease from dorsal to ventral was observed in the glucose level of the molecular layer of the mouse cerebellum. The existence of a freezing front, moving slowly from dorsal to ventral, and its consequences for the measured levels of biologically labile substrates, are discussed.     

The effect of electroshock on regional CNS energy reserves in mice

ATP, phosphocreatine, glycogen, glucose and lactate levels were measured in the cerebral cortex, thalamus, cerebellum, brain stem and spinal cord of mice following supramaximal electroshock. During the initial 17 s after the onset of a 2 s electrical stimulus high energy phosphate expenditure exceeded formation in all regions but was slower in spinal cord than in the other regions. In cerebral cortex high energy phosphate utilization continued to exceed formation for 32 s which was twice as long as in any other region studied. Altered levels of metabolites recovered most rapidly in spinal cord and least rapidly in cerebral cortex. Pretreatment with a non-anaesthetic dose of phenobarbitone influenced the effect of electroshock. Most of the clinical seizure was prevented, and increased high energy phosphate utilization was sustained for a much shorter period. Only in cerebral cortex did high energy phosphate expenditure exceed formation for as long as 15 s after the electrical stimulus; but even in this region the excess of expenditure over formation was much less than in untreated animals.     

Diversity of metabolic patterns in human brain tumors--I. High energy phosphate compounds and basic composition

Abstract— A total of 25 human brain tumors and 4 specimens of human brain were rapidly frozen at the time of operation and analyzed for ATP, ADP, AMP, UTP, total nucleoside triphosphates, P-creatine, creatine, inorganic P, creatine kinase, lipid and glycogen. Analyses were made on submicrogram samples dissected from frozen dried sections in order to obtain material as free as possible from admixture with brain, necrotic tissue, blood, etc. A method was developed to estimate the original water content of the frozen dried samples. The brain specimens contained five times as much glycogen as small mammal brains, otherwise the values were similar. The tumors were in fair to excellent energy status. Within the areas chosen for assay, most of ATP and total adenylate were substantially higher than in brain in the case of 5 out of 15 gliomas, 3 of 5 meningiomas, and 1 of 4 schwannomas. UTP was almost invariably higher and other nucleotide triphosphates (besides ATP and UTP) lower than in brain. Glycogen was extremely variable, ranging among the gliomas from 0.05% to 6% of dry wt (4 times the level in the human brains). Creatine plus P-creatine, compared to cerebral cortex levels, ranged from 15 to 85% in gliomas, was about 25% in meningiomas and the only medulloblastoma, and varied between 6 and 8% in the schwannomas. P-Creatine varied more or less in keeping with the energy status. Creatine kinase was exceedingly variable. It was almost zero in the schwannomas, the medulloblastomas, 3 of 5 meningiomas, and 2 of 15 gliomas, whereas in some of the gliomas the activity approached that found in brain.     

Quick freezing of the murine CNS: comparison of regional cooling rates and metabolite levels when using liquid nitrogen of Freon-12

Abstract— Murine brains were frozen in situ, either in liquid N₂ or in Freon-12 cooled to its freezing point. The effect of these coolants on cooling rates and times in various CNS regions was determined. In addition, levels of ATP, P-creatine and lactate were measured in selected regions of brains from both intact animals and severed heads frozen in either coolant. For both the intact animals and severed heads, superficial regions of brain cooled to 0°C and deeper regions to 25°C, at the same rate in either liquid N2 or Freon. Subsequent cooling was more rapid in liquid N₂ in both regions. Levels of ATP, P-creatine and lactate were similar in brains frozen in either coolant, probably because CNS utilization of highenergy phosphates decreased markedly as body temperature fell. In brains of animals frozen intact, levels of ATP and P-creatine were higher and levels of lactate were lower than those in brains from heads which were severed prior to freezing. This difference may be a result of the marked stimulation which accompanies decapitation and may also reflect continued cerebral circulation in the intact animal for a brief time after immersing the animal in the coolant.     

Cerebral carbohydrate metabolism during acute hypoxia and recovery

Abstract— The levels of ATP, ADP, AMP and phosphocreatine, of four amino acids, and of 11 intermediates of carbohydrate metabolism in mouse brain were determined after: (1) various degrees of hypoxia; (2) hypoxia combined with anaesthesia; and (3) recovery from severe hypoxia. Glycogen decreased and lactate rose markedly in hypoxia, but levels of ATP and phosphocreatine were normal or near normal even when convulsions and respiratory collapse appeared imminent. During 30 s of complete ischaemia (decapitation) the decline in cerebral ATP and phosphocreatine and the increase in AMP was less in mice previously rendered hypoxic than in control mice. From the changes we calculated that the metabolic rate had decreased by 15 per cent or more during 30 min of hypoxia. Hypoxia was also associated with decreases of cerebral 6-phosphogluconate and aspartate, and increases in alanine, γ-aminobutyrate, α-ketoglutarate, malate, pyruvate, and the lactate :pyruvate ratio. Following recovery in air (10 min), increases were observed in glucose (200 per cent), glucose-6-phosphate, phosphocreatine and citrate, and there was a fall in fructose-1, 6-diphosphale. Similar measurements were made in samples from cerebral cortex, cerebellum, midbrain and medulla. Severe hypoxia produced significant increases in lactate and decreases in glycogen in all areas; γ-aminobutyrate levels increased in cerebral cortex and brain stem, but not in cerebellum. No significant changes occurred in ATP and only in cerebral cortex was there a significant fall in phosphocreatine. Phosphocreatine, ATP and glycogen were determined by quantitative histochemical methods in four areas of medulla oblongata, including the physiological respiratory centre of the ventromedial portion. After hypoxia, ATP was unchanged throughout and the changes (decreases) in phosphocreatine and glycogen were principally confined to dorsal medulla, notably the lateral zone. Thus there is no evidence that respiratory failure is caused by a ‘power’ failure in the respiratory centre. It is suggested that in extremis a protective mechanism may cause neurons to cease firing before high-energy phosphate stores have been exhausted.     

Sleep deprivation decreases glycogen in the cerebellum but not in the cortex of young rats

We tested whether brain glycogen reserves were depleted by sleep deprivation (SD) in Long-Evans rats 20-59 days old. Animals were sleep deprived beginning at lights on and then immediately killed by microwave irradiation. Glycogen and glucose levels were measured by a fluorescence enzymatic assay. In all age groups, SD reduced cerebellar glycogen levels by an average of 26% after 6 h of SD. No changes were observed in the cortex after 6 h of SD, but in the oldest animals, 12 h of SD increased cortical glycogen levels. There was a developmental increase in basal glycogen levels in both the cortex and cerebellum that peaked at 34 days and declined thereafter. Robust differences in cortical and cerebellar glycogen levels in response to enforced waking may reflect regional differences in energy utilization and regulation during wakefulness. These results show that brain glycogen reserves are sensitive to SD.