- Email: [email protected]
Anaerobic glycolysis is crucial for the maintenance of neural activity in guinea pig hippocampal slices
Journal of Neuroscience Methods 103 (2000) 163 – 171 www.elsevier.com/locate/jneumeth
Anaerobic glycolysis is crucial for the maintenance of neural activity in guinea pig hippocampal slices Kyoko Yamane a,b, Koichi Yokono b, Yasuhiro Okada a,* b
a Department of Physiology, Kobe Uni6ersity School of Medicine, 7 -5 -1, Kusunoki-cho, Chuo-ko, Kobe, 650 -0017, Japan Department of Geriatric Medicine, Kobe Uni6ersity School of Medicine, 7 -5 -1, Kusunoki-cho, Chuo-ko, Kobe, 650 -0017, Japan
Received 12 April 2000; received in revised form 7 August 2000; accepted 14 August 2000
Abstract To investigate the functional significance of anaerobic and aerobic glycolysis on neural activity and levels of high energy phosphates, we tested the effects of glucose, mannose, fructose lactate, and pyruvate on the maintenance of neural activity and on the levels of ATP and creatine-P (CrP) in hippocampal slices. For an index of neural activity, population spikes (PS) evoked in the granule cell layer were monitored. Immediately after deprivation of glucose, the PS amplitude was gradually reduced and extinguished within 30 mm. Replacement of glucose with either lactate or pyruvate resulted in a decay and loss of PS with a similar time-course as observed during glucose deprivation. However, after the complete loss of neural activity for 10 – 20 min the PS reappeared and recovered to normal levels. The replacement of glucose with either mannose or fructose resulted in a transient decrease of the PS to 80–70% of the original amplitude in 20 mm, followed by recovery. The time-course of the decrease of PS in the mannose-containing medium was slower than that in the medium containing fructose and the time-course of recovery was faster. ATP and CrP were reduced to 90 and 70% of original level in each slice after glucose deprivation for 30 and 100 mm, respectively. In media containing either lactate, pyruvate, mannose, or fructose, the level of ATP and CrP was maintained at the original level. The anaerobic metabolic rate of glucose, mannose and fructose, determined by the rate of lactate production during complete anoxia, was consistent with the order of the decay and recovery of the PS in mannose and fructose-containing medium. The mode of the transient decay or loss of PS with no apparent reduction in the levels of ATP and CrP in each slice, in the presence of either mannose, fructose or lactate, indicates that anaerobic glycolysis is crucial for the maintenance of PS. The results obtained in this experiment are not in accordance with the reports by Schurr and Fowler in which they showed that lactate can support neural activity, although they did not measure the levels of ATP and CrP in slices. The present experiment disclosed that this discrepancy was due to the difference of slice preparation; namely rapidly prepared slice with shorter period anoxia (1 mm after removal of hippocampal tissue block) gives the results mentioned in the present study whereas slowly prepared slices by vibratome or chopper method with longer period of ischemia (5 – 20 mm), did not show transient loss of PS after application of lactate. Thus present experiment indicates that glycolytic process is essential for maintaining neural activity for physiological state of slices, if it is admitted to say that rapidly prepared slices is more physiological because they are exposed on shorter period of ischemia, and that more careful attention should be paid for the interpretation of the results of slice experiment according to the method of slice preparation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Hippocampal slices; Neural activity; High energy phosphates; Anaerobic glycolysis; Glucose; Lactate
1. Introduction Glucose is generally considered to be the main substrate in the production of the energy required to maintain neural activity. Extended hypoglycemia leads * Corresponding author. Tel.: +81-78-382-5362; fax: + 81-78-3825379. E-mail address: [email protected] (Y. Okada).
to coma and may result in the functional derangement of the brain (Cox and Bachelard, 1982). Pyruvate, lactate and other sugars such as mannose and fructose can also be used as substrates for energy metabolism in cortical slices, although there have been few reports studying the ability of these substrates to preserve neural activity (Mcllwain and Bachelard, 1985). Cox and Bachelard (1988) reported that replacing glucose in the perfusion medium with either pyruvate or lactate
0165-0270/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0165-0270(00)00312-5
164
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
failed to maintain the evoked synaptic response and neural activity. On the other hand, Schurr et al. (1988, 1997c) and Fowler (1993), recording the population spikes in the CAl pyramidal cell layer of hippocampal slices, demonstrated that the replacement of glucose with lactate could preserve neural activity and that lactate is an obligatory aerobic energy substrate for functional recovery after prolonged oxygen deprivation. In contrast, we have reported previously that mannose, fructose and lactate can be metabolized and are available for maintaining the levels of high-energy phosphates in brain slices, but are not sufficient to maintain neural activity (Wada et al., 1998). In this study, using guinea pig hippocampal slices, we investigated the effects of replacing glucose with other sugars and intermediate metabolites on neural activity and on the levels of ATP and CrP. Further, we determined the level of lactate produced during complete oxygen deprivation using either glucose, mannose or fructose as a substrate. We demonstrate that anaerobic glycolysis is crucial for neural activity and that lactate cannot be the obligatory substrate to maintain neural activity in the physiological state.
2. Materials and methods
2.1. Preparation of hippocampal slices Adult guinea pigs, weighing 250 – 350 g, were used in these experiments. Animals were sacrificed according to the guidelines for animal experimentation at the Kobe University, School of Medicine. The hippocampal slices (300 –400 mm thick) were prepared by cutting transversely along the long axis of the hippocampus with a razor blade within 1 min after the removal of the brain from the skull. Details of the preparation of the hippocampal slices have been described elsewhere (Okada, 1982). Each slice was preincubated for at least 20 mm in the standard medium (in mM: NaCl 125, KCl 5, KH2PO4 1.24, MgSO4 1.3, CaCl2 2, NaHCO3 26, glucose l0) bubbled with 95% O2 and 5% CO2. The temperature of the medium was kept at 35°C throughout the experiments.
2.2. Recording of neural acti6ity After preincubation, each slice was transferred to the observation chamber under a stereomicroscope. The chamber in which slices were submerged was perfused continuously with the standard or experimental medium at a flow rate of 3 ml/mm and had a 1.5 ml total volume. As an index of neural activity, orthodromic synaptic field potentials (population spikes PS) were recorded in the granule cell layer of the dentate gyrus of the hippocampal slices after electrical stimulation to the perforant path
through a glass microelectrode filled with 2 M NaCl. To differentiate the synaptic events and the neuronal excitability, PS were also evoked antidromically in the granule cell layer and recorded with the same electrode after mossy fiber stimulation. The strength of stimulation was adjusted to maintain maximum amplitude of PS. After assuring the stability of the PS for at least 20 mm, slices were exposed to a perfusing medium containing either mannose (10 mM), fructose (10 mM), lactate (5 mM) or pyruvate (5 mM) instead of glucose. For glucose deprivation, glucose was completely removed from the standard perfusion medium. Each experimental medium was bubbled with 95% O2 and 5% CO2 and the pH was maintained at 7.4.
2.3. Determination of ATP, CrP and lactate For the determination of ATP and CrP concentrations, each hippocampal slice was homogenized immediately in 0.5 M perchloric acid containing 1 mM ethylenediaminetetraacetic acid (EDTA), followed by centrifugation for 10 mm at 3000 rpm. The supernatant was neutralized with 2 M KHCO3 and re-centrifuged for 5 mm at 2000 rpm. The resulting supernatant was used for the determination of ATP, CrP and lactate. To determine the production of lactate during complete anoxic condition, 2 mM Na2S2O3 was added to the medium containing mannose, fructose and glucose and bubbled with 95% N2 and 5% CO2. The oxygen tension was monitored by the oxygenometer, which showed 0 mmHg of oxygen pressure at 1 mm after the introduction of medium containing Na2S2O3 bubbled with N2 gas instead of Na2S2O3 free medium bubbled with O2. The production of lactate was measured after the complete deprivation of oxygen. Lactate in the tissue slice and in the medium surrounding the slice was measured, due to the ease of lactate release into the incubation medium. ATP, CrP and lactate were determined enzymatically and fluorometrically by measuring the production of NADPH and NADH (Okada and McDougal, 1971). The remaining precipitate of the tissue homogenate was used for the determination of protein by the method of Lowry et al. (1964). Chemicals, used for the assay of ATP, CrP and lactate, were purchased from Nacalai Co. Enzymes, such as hexokinase, glucose-6-phosphate-dehydrogenase (G6PDH) and creatine kinase, were obtained from Boehringer Mannheim Yamanouchi.
3. Results
3.1. Effects of glucose depri6ation and replacement of either mannose, fructose, pyru6ate or lactate on the PS amplitude In order to establish an index of neural activity, PS
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
were elicited in the granule cell layer of the dentate gyrus of the hippocampus. After observing the steady potentials of the PS for at least 20 mm, glucose was removed from the perfusate. Immediately after deprivation of glucose, the PS amplitude gradually decreased and was extinguished within 30 mm (Fig. 1-A2). When glucose was replaced with either lactate or pyruvate, there was a decay or loss of PS with a similar timecourse observed during glucose deprivation. However after a complete loss of the PS for 10 – 20 mm, they reappeared and recovered although the addition of glucose resulted in a faster recovery (Fig. 1-A3, Fig. 2-A3). The replacement of glucose with either mannose or fructose also resulted in a transient decrease of the PS to 70–80% of the original amplitude in 30 mm followed by a recovery of PS without further decay (Fig. 2-Al, A2). The time-course of the decrease of the PS amplitude in mannose containing medium was slower than that observed in fructose medium and the time-course of recovery was faster. Fig. 3 shows the effects of the application of either
165
glucose, mannose, fructose or lactate after a 30 mm deprivation of glucose when the neural activity was completely ceased. The addition of either glucose, mannose or fructose resulted in a faster recovery of the PS than that observed in lactate containing medium.
3.2. ATP and CrP le6els on glucose depri6ation The levels of high-energy phosphates, such as ATP and CrP, in each hippocampal slice after exposure to different perfusion media are shown in Fig. 1 and Fig. 2. The exposure of slices to glucose-free medium for either 30 mm or 100 mm resulted in a decrease in the levels of ATP to 90% and 70% of the original level, respectively. In glucose-free medium, the levels of CrP also decreased to 80% and 50% of control level in 30 and 100 mm exposures, respectively. When the hippocampal slices were exposed to experimental medium containing either mannose, fructose, lactate or pyruvate instead of glucose, the levels of ATP and CrP were maintained at original levels.
Fig. 1. Effect of deprivation of glucose and replacement of pyruvate on (Row A) the neural activity and on (Row B) the levels of ATP (closed circles) and creatine P (CrP; open circles) of hippocampal slices. For an index of neural activity, population spikes (PS) were recorded in the granule cell layer of dentate gyrus. (1) Indicates the PS amplitude and the levels of ATP and CrP in the standard medium containing glucose (10 mM). (2) Indicates the change of those in the medium deprived of glucose and (3) those in the medium containing pyruvate (5 mM) instead of glucose. Vertical bars of each plot indicate SEM (n = 6–8).
166
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
Fig. 2. Effect of mannose (1), fructose (2) and lactate (3) in (Row A) the preservation of PS and on (Row B) the levels of ATP (closed circles) and CrP (open circles) of hippocampal slices. In Row A, closed circles indicate the change in the amplitude of PS evoked by orthodromic stimulation and open squares denote that evoked by antidromic stimulation. Vertical bars of each plot indicate SEM (n = 6 – 8).
3.3. Lactate accumulation during complete anoxia The anaerobic metabolic rates of glucose, mannose and fructose were determined by lactate production during complete anoxia. To obtain a complete anoxic condition, 2 mM Na2S2O3 was added to the medium while bubbling with 95% N2 and 5% CO2 The oxygen tension was monitored by an oxygenometer, which indicated 0 mmHg after 1 min of oxygen deprivation. Production of lactate was measured after complete deprivation of oxygen for 2, 5, 10, 15 and 30 mm. As indicated by the formation of lactate, the anaerobic metabolic rate was fastest in glucose-containing medium, followed by mannose and slowest in the fructose medium (Fig. 4). These results are consistent with the order of the decay and recovery of PS in mannose and fructose-containing medium (Fig. 2-A1, A2).
3.4. Effect of lactate on the preser6ation of PS in slices exposed to longer periods of ischemia We have reported previously, using hippocampal slices, that the presence of glucose is required to main-
tain neural activity and that, as confirmed here, the deprivation of glucose blocks neural activity (Okada, 1990, 1992; Saitoh et al. 1994; Kanatani et al., 1995). However, Schurr et al. (1988, 1997a,b,c) reported that lactate, an obligatory aerobic energy substrate, could support neural function. In their study, hippocampal slices were prepared using a Mcllwain tissue chopper. Although the tissue is immersed in the ice cold standard medium, this method requires at least a 10–20 mm anoxic or ischemic period before preparing the hippocampal slices (slowly prepared slices, see Fig. 7B) because the hippocampal tissue block is removed from the brain and fixed on the cutting table before slicing the tissue. In our studies, hippocampal slices (rapidly prepared slices, see Fig. 7A,C) are prepared within 1 mm after removal of the brain from the skull. We prepare slices from the hippocampus tissue block with a razor blade by free hand and immediately incubate them in the glucose containing standard medium for 20 mm. In our technique, slices of similar thickness (300– 400 mm) could be prepared easily and the slice tissue is exposed to ischemic condition for a maximum of l mm after removal of brain from the skull.
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
167
Fig. 3. Effect of glucose (A), mannose (B), fructose (C) and lactate (D) application on the PS amplitude after cessation of PS during glucose deprivation. Vertical bars of each plot indicate SEM (n=4–5).
To investigate the biochemical and functional difference between these two preparations, we prepared hippocampal slices under the same ischemic conditions as the studies using vibratome technique or tissue chopper method. We left the hippocampus tissue block on an ice-cold table for 20 mm immersed in ice cold standard medium and cut the tissue into slices using a razor blade by free hand. Prior to incubating the slices in the medium containing 10 mM glucose and saturated with oxygen, the ATP levels of slices prepared by either our method (rapid preparation) or by the vibratome technique (slow preparation) were 4.44 9 0.85 mmol/kg protein (Fig. 5A, n =5) and 1.759 0.23 mmol/kg protein (Fig. 5B, n =4), respectively. Incubating the slices in medium containing oxygen and glucose for 20 mm, increased the ATP levels in the slices rapidly prepared by our method to 12.339 0.49 mmol/kg protein whereas that slowly prepared by tissue chopper or vibratome method increased to 6.549 1.26 mmol/kg protein (Fig. 5). This inadequate recovery of the highenergy phosphate levels after a longer period of ischemia confirmed the results of our previous study in which a longer exposure of brain tissue to ischemia greatly reduced the recovery of high-energy phosphates (Mori et al., 1992). The synaptic activity of slices prepared slowly, as in the vibratome or slice chopper method, was tested using medium in which glucose was replaced with lactate. Interestingly the amplitude of PS did not disappear even for 100 mm in lactate-containing medium, although the amplitude was transiently reduced to 85% of
original amplitude for the first 20 mm (Fig. 6-B1). Slices prepared by free hand immediately after removal of hippocampal tissue block from the brain showed the transient complete loss of the PS as indicated in Fig. 6-A1. After complete recovery of the PS amplitude in the lactate medium, the lactate was replaced with glu-
Fig. 4. Effect of glucose (n =6), mannose (n = 4) and fructose (n= 4) on the formation of lactate during complete anoxia. Vertical bars of each plot indicate SEM.
168
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
ATP and CrP were lower in slices prepared slowly as with the chopper method.
4. Discussion
Fig. 5. Recovery of ATP and CrP in rapidly prepared slices (A) (n=5) and slowly prepared slices (B) (n= 4) during preincubation in the standard medium. In A, slices were prepared by razor blade immediately after removal of tissue block of hippocampus from brain. In B, they were prepared after putting hippocampal tissue block once on the ice cold cutting table for 20 mm (see text). After preparation of slices, they were incubated in the standard medium for 5, 10 and 20 mm. Vertical bars of each plot indicate SEM.
cose for 30 mm, and then the glucose was replaced with lactate again. In this case, the PS amplitude was not reduced and did not transiently disappear as observed in the first episode of lactate addition. In addition, the neural activity was stable, even after 100 mm. In slices prepared by both methods, the levels of ATP and CrP were maintained at the original levels, after preincubation of slices, even after replacement of the glucose with lactate (Fig. 6A2, B2). However, the initial levels of
In this study, we examined the correlation between neural activity and the levels of high energy phosphates such as ATP and CrP in hippocampal slices exposed to medium containing either glucose, mannose, fructose, pyruvate or lactate as an energy source. There have been many reports demonstrating the relationship between energy metabolism and neural activity in brain tissue (Yamamoto and Kurokawa, 1970; Okada and McDougal, 1971; Cox and Bachelard, 1982, 1988; Lipton and Whittingham, 1982; Bachelard et al., 1984; Cox et al., 1985; Okada, 1992). Yamamoto and Kurokawa (1970) reported that synaptic potentials were completely blocked when the ATP level was reduced by 20% of its original concentration in olfactory cortical slices. Lipton and Whittingham (1982) showed that, in hippocampal slices, ATP was decreased by 15% when the evoked synaptic response began to diminish 2 mm after exposure to hypoxia. Additionally, Okada (1982) found a discrepancy between the time-course in the reduction of ATP levels and the decrease in the PS amplitude during either oxygen or glucose deprivation. The PS amplitude decayed much faster during glucose deprivation than during hypoxia, even though the level of ATP was reduced in the similar time-course. Further, the addition of either lactate or pyruvate instead of
Fig. 6. Effect of lactate on (1) the PS and on (2) the levels of ATP (closed circles) and CrP (open circles) of slowly prepared slices (Column B) and the effect of repetitive application of lactate on the PS and levels of ATP and CrP of rapidly prepared slices (Column A). Note that in B, replacement of glucose with lactate did not cause the reduction of the PS amplitude and in A, the second addition of lactate caused no reduction of the PS amplitude. Vertical bars of each plot indicate SEM (n = 4 – 6).
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
169
Fig. 7. Schematic drawing showing the effect of lactate on the neural activity (PS) of rapidly prepared (A) and slowly prepared slices (B) and the effect of repetitive application of lactate on the PS of rapidly prepared slices (C). Each experimental procedure for slice preparation is shown in the left and the change of the PS amplitude during application of lactate instead of glucose (G) is shown in the right.
glucose maintained the original levels of ATP and CrP, although neural activity was diminished with similar time-course as in glucose deprivation. In a similar observation, Cox and Bachelard (1988) reported an attenuation of the PS amplitude during the incubation of hippocampal slices in lactate-containing medium. Wada et al. (1998) confirmed the importance of glucose in the preservation of PS and found that neural activity, such as synaptic transmission, differed in the requirement for glucose in the granule cells and in the pyramidal cells of CA3 area. This was observed despite the fact that the levels of high-energy phosphates in these areas were not significantly different during glucose deprivation. These results strongly suggested that the presence of glucose is essential in the maintenance of neural activity, regardless of the ATP level in the brain slices. In contrast to the reports described above, Schurr et al. (1997a) and Fowler (1993) reported that lactate, instead of glucose, can support synaptic transmission in CA1 neurons of rat hippocampal slices. However, they did not determine the ATP concentration in the slices. Using an intracellular recording technique for single CA3 pyramidal neurons in hippocampal slices, Takata and Okada (1995) found that lactate could not support synaptic transmission and induced membrane depolarization, although the ATP levels were not reduced in these slices. As the time-course of decay for the synaptic potential during lactate replacement is very similar to that during glucose deprivation, ATP derived from
the oxidative pathway may not support synaptic transmission and may not be able to maintain the resting potential in CA3 neurons. Thus, our results do not agree with the observations of Schurr et al. (1997a) and Fowler (1993). In connection with the functional significance of the glycolytic pathway, Rose and O’Connell (1964) and Mercer and Dunham (1981) reported that, even in erythrocytes, glycolytic enzymes bind to the cell membrane and membrane-bound ATP fuels the Na/K pump that maintains the resting membrane potential. If this is true for neurons, the ATP produced through anaerobic and aerobic pathways may have different functional significance. It is interesting to note the phenomenon observed by Saitoh et al. (1994) that slices maintained for a longer period in lactate-containing medium recovered their neural activity which was initially extinguished. The presence of lactate, together with glucose prior to glucose deprivation, may result in the maintenance of the PS amplitude. To test this possibility, Saitoh et al. (1994) added lactate to the medium for 30 min before glucose deprivation. The PS amplitude decayed, was transiently lost, and reappeared although the levels of ATP and CrP were maintained at the initial concentrations (Saitoh et al., 1994). In the present experiment, we confirm these observations and find that the addition of either mannose or fructose instead of glucose resulted in the transient reduction of the PS amplitude followed by complete recovery. In addition, the PS was not completely extinguished as observed in the lactate
170
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
medium. Further, we examined neural activity and determined the levels of high energy phosphates during the application of either mannose, fructose, or lactate instead of glucose. Fructose and mannose have been reported to act as a substrate for ATP production in brain tissue (Sapolsky, 1986). Mcllwain and Bachelard (1985), measuring oxygen consumption of cortical slices, showed that either sugars, such as galactose, fructose, and mannose, or metabolites, such as pyruvate and lactate, can also be used as substrates for energy metabolism in isolated cortical slice tissues, although the rates of utilization of lactate, fructose and mannose were much lower than that of glucose. Our results suggest that glucose, mannose, and fructose are able to maintain a steady anaerobic metabolic rate although, by measuring the rate of lactate formation, these substrates differ in the metabolic rate they support. These results, together with the effects of those substrates on the neural activity, indicate that the rate of the non-oxidative pathway may be crucial to the maintenance and preservation of neural activity. Why do long periods of slice incubation in lactatecontaining medium result in the recovery of neural activity that was once completely extinguished? Our results reveal that, in slices perfused with lactate medium repetitively, the second episode of perfusion with lactate is accompanied by well preserved PS. This observation indicates that once brain tissue is exposed to either a hypoglycemic or ischemic episode, lactate begins to be utilized to preserve neural activity. During the preparation of slices from a brain tissue block, the hippocampal slices prepared by our method are exposed to ischemic period for only 1 mm, whereas methods that are more frequently used, such as either the vibratome or chopper method, require a 10 – 20 mm preparatory period before the slices are made. Therefore, it is inevitable that the slices are exposed to longer period of ischemic conditions. The differences in the experimental procedures between our method and others in regards to the effect of lactate addition are schematized in Fig. 7. These procedural differences may explain the discrepancy between our results and the results of Schurr et al. (1997a). Schurr and Fowler performed the experiment using rat preparation and recorded the electrical activity from CA1 region of hippocampus. For this reason, we tried the same experimental procedure of slow and rapid preparation of slices for rat (n=6) and also recording from CA1 region of guinea pig hippocampus (n = 5). We obtained the similar results as shown in the present experiment. Thus our rapid slice preparation, with a shorter ischemic period, may reflect a more physiological (in situ) state of brain tissue, with a faster and better recovery of high energy phosphates. Concerning the importance of glycolysis during physiological activity there have been several reports show-
ing that physiological stimulation increases specifically nonoxidative glucose metabolism in the brain of the freely moving rat (Fellows et al., 1993) and that in the human PET study, transient increases in neural activity cause a tissue uptake of glucose in excess of that consumed by oxidative metabolism (Fox et al., 1988). Thus our results together with others suggest that anaerobic glycolysis is crucial for neural activity in the physiological state. Under well supplied conditions, glucose may be the primary substrate for the production of ATP by cytoplasmic enzymes that is used in the maintenance of neural activity. Mitochondrial ATP may become important for neural function only after hypoglycemic episodes, although 20–30 mm may be required to switch from anaerobic to aerobic mechanisms. The switching mechanisms to utilize ATP produced from either an anaerobic or an aerobic pathway remains to be studied further. Finally, the present study suggests that the interpretation of experimental results from brain slices must consider the method of slice preparation because the tissue may be exposed to varying ischemic conditions dependent upon the methods used.
Acknowledgements This paper was supported by the grant from the Minister of Welfare, Japan.
References Bachelard HS, Cox DWG, Drower J. Sensitivity of guinea-pig hippocampal granule cell field potentials to hexoses in vitro: an effect on cell excitability? J Physiol 1984;352:91 – 102. Cox DWG, Bachelard HS. Attenuation of evoked field potentials from dentate granule cell by low glucose, pyruvate + malate, and sodium fluoride. Brain Res 1982;239:527 – 34. Cox DWG, Bachelard HS. Partial attenuation of dentate granule cell evoked activity by the alternative substrates, lactate and pyruvate: evidence for a postsynaptic action. Exp Brain Res 1988;69:368– 72. Cox DWG, Drower J, Bachelard HS. Effect of metabolic inhibitors on evoked activity and the energy state of hippocampal slices superfused in vitro. Exp Brain Res 1985;57:464 – 70. Fellows LK, Boutelle MG, Fillenz M. Physiological stimulation increases nonoxidative glucose metabolism in the brain of the freely moving rat. J Neurochem 1993;60:1258 – 63. Fowler JC. Glucose deprivation results in a lactate preventable increase in adenosine and depression of synaptic transmission in rat hippocampal slices. J Neurochem 1993;60:572 – 6. Fox PT, Raictile ME, Mintun MA, Dence C. Nonoxidative glucose consumption during focal physiologic neural activity. Science 1988;241:462 – 4. Kanatani T, Mizuno K, Okada Y. Effects of glycolytic metabolites on preservation of high energy phosphate level and synaptic transmission in the granule cells of guinea pig hippocampal slices. Experimentia 1995;51:213 – 6.
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171 Lipton P, Whittingham TS. Reduced ATP concentration as a basis for synaptic transmission failure during hypoxia in the in vitro guinea-pig hippocampus. J Physiol 1982;325:51–65. Lowry OH, Passonneau JY, Hasselberger FX, Schulz DW. Effect of ischemia on known substrates and cofactors of the glycolytic pathway in brain. J Biol Chem 1964;239:18–30. Mcllwain H, Bachelard HS. Metabolic, ionic and electrical phenomena in separated cerebral tissue. Biochemistry and the Central Nervous System, 5th edn. Edinburgh: Livingstone, 1985:54 – 83. Mercer RW, Dunham PB. Membrane-bound ATP fuels the Na/K pump: studies on membrane-bound glycolytic enzymes on insideout vesicle from human red cells membranes. J Gen Physiol 1981;78:547 – 68. Mori M, Nishizaki T, Okada Y. Protective effect of adenosine on the anoxic damage of hippocampal slice Neuroscience 1992;46:301 – 7. Okada Y. Reversibility of neuronal function of hippocampal slices during deprivation of oxygen and/or glucose. In: Somjen GG, editor. Mechanism of Cerebral Hypoxia and Stroke. New York: Plenum, 1982:191 – 203. Okada Y. Energy metabolism and neural activity in brain slice-effect of deprivation of oxygen and glucose. Protein, Nucleic Acid and Enzyme (Jpn) 1990;35:661–74. Okada Y. Energy metabolism and neural activity of brain slices importance of glucose for the maintenance of neural activity. J Brain Med Japanese 1992;4:39–49. Okada Y, McDougal DB. Physiological arid biochemical change in frog sciatic nerve during anoxia and recovery. J Neurochem 1971;18:2335 – 53. Rose IA, O’Connell EL. The role of glucose 6-phosphate in the regulation of glucose metabolism in human erythrocytes. J Biol Chem 1964;230:12 – 7.
.
171
Saitoh M, Okada Y, Nabetani M. Effect of mannose, fructose and lactate on the preservation of synaptic potentials in hippocampal slices. Neurosci Lett 1994;171:125 – 8. Sapolsky RM. Glucocorticoid toxicity in the hippocampus: reversal by supplementation with brain fuels. J Neurosci 1986;6:2240– 4. Schurr A, Payne RS, Miller JJ, Rigor BM. Brain lactate, not glucose, fuels the recovery of synaptic function from hypoxia upon reoxygenation: an in vitro study. Brain Res 1997a;744:105 – 11. Schurr A, Payne RS, Miller JJ, Rigor BM. Glia are the main source of lactate utilized by neurons for recovery of function posthypoxia. Brain Res 1997b;774:221 – 4. Schurr A, Payne RS, Miller JJ, Rigor BM. Brain lactate is an obligatory aerobic energy substrate for functional recovery after hypoxia: further in vitro validation. J Neurochem 1997c;69:423– 6. Schurr A, West CA, Rigor BM. Lactate-supported synaptic function in the rat hippocampal slices preparation. Science 1988;240:1326– 8. Takata T, Okada Y. Effects of deprivation of oxygen or glucose on the neural activity in the guinea pig hippocampal slice-intracellular recording study of pyramidal neurons. Brain Res 1995;683:109 – 16. Wada H, Okada Y, Uzuo H, Nakamura H. The effects of glucose, mannose, fructose and lactate on the preservation of neural activity in the hippocampal slices from the guinea pig. Brain Res 1998;788:144 – 50. Yamamoto C, Kurokawa M. Synaptic potentials recorded in brain slices and their modification by changes in the level of tissue ATP. Exp Brain Res 1970;10:159 – 70.
Anaerobic glycolysis is crucial for the maintenance of neural activity in guinea pig hippocampal slices Kyoko Yamane a,b, Koichi Yokono b, Yasuhiro Okada a,* b
a Department of Physiology, Kobe Uni6ersity School of Medicine, 7 -5 -1, Kusunoki-cho, Chuo-ko, Kobe, 650 -0017, Japan Department of Geriatric Medicine, Kobe Uni6ersity School of Medicine, 7 -5 -1, Kusunoki-cho, Chuo-ko, Kobe, 650 -0017, Japan
Received 12 April 2000; received in revised form 7 August 2000; accepted 14 August 2000
Abstract To investigate the functional significance of anaerobic and aerobic glycolysis on neural activity and levels of high energy phosphates, we tested the effects of glucose, mannose, fructose lactate, and pyruvate on the maintenance of neural activity and on the levels of ATP and creatine-P (CrP) in hippocampal slices. For an index of neural activity, population spikes (PS) evoked in the granule cell layer were monitored. Immediately after deprivation of glucose, the PS amplitude was gradually reduced and extinguished within 30 mm. Replacement of glucose with either lactate or pyruvate resulted in a decay and loss of PS with a similar time-course as observed during glucose deprivation. However, after the complete loss of neural activity for 10 – 20 min the PS reappeared and recovered to normal levels. The replacement of glucose with either mannose or fructose resulted in a transient decrease of the PS to 80–70% of the original amplitude in 20 mm, followed by recovery. The time-course of the decrease of PS in the mannose-containing medium was slower than that in the medium containing fructose and the time-course of recovery was faster. ATP and CrP were reduced to 90 and 70% of original level in each slice after glucose deprivation for 30 and 100 mm, respectively. In media containing either lactate, pyruvate, mannose, or fructose, the level of ATP and CrP was maintained at the original level. The anaerobic metabolic rate of glucose, mannose and fructose, determined by the rate of lactate production during complete anoxia, was consistent with the order of the decay and recovery of the PS in mannose and fructose-containing medium. The mode of the transient decay or loss of PS with no apparent reduction in the levels of ATP and CrP in each slice, in the presence of either mannose, fructose or lactate, indicates that anaerobic glycolysis is crucial for the maintenance of PS. The results obtained in this experiment are not in accordance with the reports by Schurr and Fowler in which they showed that lactate can support neural activity, although they did not measure the levels of ATP and CrP in slices. The present experiment disclosed that this discrepancy was due to the difference of slice preparation; namely rapidly prepared slice with shorter period anoxia (1 mm after removal of hippocampal tissue block) gives the results mentioned in the present study whereas slowly prepared slices by vibratome or chopper method with longer period of ischemia (5 – 20 mm), did not show transient loss of PS after application of lactate. Thus present experiment indicates that glycolytic process is essential for maintaining neural activity for physiological state of slices, if it is admitted to say that rapidly prepared slices is more physiological because they are exposed on shorter period of ischemia, and that more careful attention should be paid for the interpretation of the results of slice experiment according to the method of slice preparation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Hippocampal slices; Neural activity; High energy phosphates; Anaerobic glycolysis; Glucose; Lactate
1. Introduction Glucose is generally considered to be the main substrate in the production of the energy required to maintain neural activity. Extended hypoglycemia leads * Corresponding author. Tel.: +81-78-382-5362; fax: + 81-78-3825379. E-mail address: [email protected] (Y. Okada).
to coma and may result in the functional derangement of the brain (Cox and Bachelard, 1982). Pyruvate, lactate and other sugars such as mannose and fructose can also be used as substrates for energy metabolism in cortical slices, although there have been few reports studying the ability of these substrates to preserve neural activity (Mcllwain and Bachelard, 1985). Cox and Bachelard (1988) reported that replacing glucose in the perfusion medium with either pyruvate or lactate
0165-0270/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0165-0270(00)00312-5
164
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
failed to maintain the evoked synaptic response and neural activity. On the other hand, Schurr et al. (1988, 1997c) and Fowler (1993), recording the population spikes in the CAl pyramidal cell layer of hippocampal slices, demonstrated that the replacement of glucose with lactate could preserve neural activity and that lactate is an obligatory aerobic energy substrate for functional recovery after prolonged oxygen deprivation. In contrast, we have reported previously that mannose, fructose and lactate can be metabolized and are available for maintaining the levels of high-energy phosphates in brain slices, but are not sufficient to maintain neural activity (Wada et al., 1998). In this study, using guinea pig hippocampal slices, we investigated the effects of replacing glucose with other sugars and intermediate metabolites on neural activity and on the levels of ATP and CrP. Further, we determined the level of lactate produced during complete oxygen deprivation using either glucose, mannose or fructose as a substrate. We demonstrate that anaerobic glycolysis is crucial for neural activity and that lactate cannot be the obligatory substrate to maintain neural activity in the physiological state.
2. Materials and methods
2.1. Preparation of hippocampal slices Adult guinea pigs, weighing 250 – 350 g, were used in these experiments. Animals were sacrificed according to the guidelines for animal experimentation at the Kobe University, School of Medicine. The hippocampal slices (300 –400 mm thick) were prepared by cutting transversely along the long axis of the hippocampus with a razor blade within 1 min after the removal of the brain from the skull. Details of the preparation of the hippocampal slices have been described elsewhere (Okada, 1982). Each slice was preincubated for at least 20 mm in the standard medium (in mM: NaCl 125, KCl 5, KH2PO4 1.24, MgSO4 1.3, CaCl2 2, NaHCO3 26, glucose l0) bubbled with 95% O2 and 5% CO2. The temperature of the medium was kept at 35°C throughout the experiments.
2.2. Recording of neural acti6ity After preincubation, each slice was transferred to the observation chamber under a stereomicroscope. The chamber in which slices were submerged was perfused continuously with the standard or experimental medium at a flow rate of 3 ml/mm and had a 1.5 ml total volume. As an index of neural activity, orthodromic synaptic field potentials (population spikes PS) were recorded in the granule cell layer of the dentate gyrus of the hippocampal slices after electrical stimulation to the perforant path
through a glass microelectrode filled with 2 M NaCl. To differentiate the synaptic events and the neuronal excitability, PS were also evoked antidromically in the granule cell layer and recorded with the same electrode after mossy fiber stimulation. The strength of stimulation was adjusted to maintain maximum amplitude of PS. After assuring the stability of the PS for at least 20 mm, slices were exposed to a perfusing medium containing either mannose (10 mM), fructose (10 mM), lactate (5 mM) or pyruvate (5 mM) instead of glucose. For glucose deprivation, glucose was completely removed from the standard perfusion medium. Each experimental medium was bubbled with 95% O2 and 5% CO2 and the pH was maintained at 7.4.
2.3. Determination of ATP, CrP and lactate For the determination of ATP and CrP concentrations, each hippocampal slice was homogenized immediately in 0.5 M perchloric acid containing 1 mM ethylenediaminetetraacetic acid (EDTA), followed by centrifugation for 10 mm at 3000 rpm. The supernatant was neutralized with 2 M KHCO3 and re-centrifuged for 5 mm at 2000 rpm. The resulting supernatant was used for the determination of ATP, CrP and lactate. To determine the production of lactate during complete anoxic condition, 2 mM Na2S2O3 was added to the medium containing mannose, fructose and glucose and bubbled with 95% N2 and 5% CO2. The oxygen tension was monitored by the oxygenometer, which showed 0 mmHg of oxygen pressure at 1 mm after the introduction of medium containing Na2S2O3 bubbled with N2 gas instead of Na2S2O3 free medium bubbled with O2. The production of lactate was measured after the complete deprivation of oxygen. Lactate in the tissue slice and in the medium surrounding the slice was measured, due to the ease of lactate release into the incubation medium. ATP, CrP and lactate were determined enzymatically and fluorometrically by measuring the production of NADPH and NADH (Okada and McDougal, 1971). The remaining precipitate of the tissue homogenate was used for the determination of protein by the method of Lowry et al. (1964). Chemicals, used for the assay of ATP, CrP and lactate, were purchased from Nacalai Co. Enzymes, such as hexokinase, glucose-6-phosphate-dehydrogenase (G6PDH) and creatine kinase, were obtained from Boehringer Mannheim Yamanouchi.
3. Results
3.1. Effects of glucose depri6ation and replacement of either mannose, fructose, pyru6ate or lactate on the PS amplitude In order to establish an index of neural activity, PS
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
were elicited in the granule cell layer of the dentate gyrus of the hippocampus. After observing the steady potentials of the PS for at least 20 mm, glucose was removed from the perfusate. Immediately after deprivation of glucose, the PS amplitude gradually decreased and was extinguished within 30 mm (Fig. 1-A2). When glucose was replaced with either lactate or pyruvate, there was a decay or loss of PS with a similar timecourse observed during glucose deprivation. However after a complete loss of the PS for 10 – 20 mm, they reappeared and recovered although the addition of glucose resulted in a faster recovery (Fig. 1-A3, Fig. 2-A3). The replacement of glucose with either mannose or fructose also resulted in a transient decrease of the PS to 70–80% of the original amplitude in 30 mm followed by a recovery of PS without further decay (Fig. 2-Al, A2). The time-course of the decrease of the PS amplitude in mannose containing medium was slower than that observed in fructose medium and the time-course of recovery was faster. Fig. 3 shows the effects of the application of either
165
glucose, mannose, fructose or lactate after a 30 mm deprivation of glucose when the neural activity was completely ceased. The addition of either glucose, mannose or fructose resulted in a faster recovery of the PS than that observed in lactate containing medium.
3.2. ATP and CrP le6els on glucose depri6ation The levels of high-energy phosphates, such as ATP and CrP, in each hippocampal slice after exposure to different perfusion media are shown in Fig. 1 and Fig. 2. The exposure of slices to glucose-free medium for either 30 mm or 100 mm resulted in a decrease in the levels of ATP to 90% and 70% of the original level, respectively. In glucose-free medium, the levels of CrP also decreased to 80% and 50% of control level in 30 and 100 mm exposures, respectively. When the hippocampal slices were exposed to experimental medium containing either mannose, fructose, lactate or pyruvate instead of glucose, the levels of ATP and CrP were maintained at original levels.
Fig. 1. Effect of deprivation of glucose and replacement of pyruvate on (Row A) the neural activity and on (Row B) the levels of ATP (closed circles) and creatine P (CrP; open circles) of hippocampal slices. For an index of neural activity, population spikes (PS) were recorded in the granule cell layer of dentate gyrus. (1) Indicates the PS amplitude and the levels of ATP and CrP in the standard medium containing glucose (10 mM). (2) Indicates the change of those in the medium deprived of glucose and (3) those in the medium containing pyruvate (5 mM) instead of glucose. Vertical bars of each plot indicate SEM (n = 6–8).
166
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
Fig. 2. Effect of mannose (1), fructose (2) and lactate (3) in (Row A) the preservation of PS and on (Row B) the levels of ATP (closed circles) and CrP (open circles) of hippocampal slices. In Row A, closed circles indicate the change in the amplitude of PS evoked by orthodromic stimulation and open squares denote that evoked by antidromic stimulation. Vertical bars of each plot indicate SEM (n = 6 – 8).
3.3. Lactate accumulation during complete anoxia The anaerobic metabolic rates of glucose, mannose and fructose were determined by lactate production during complete anoxia. To obtain a complete anoxic condition, 2 mM Na2S2O3 was added to the medium while bubbling with 95% N2 and 5% CO2 The oxygen tension was monitored by an oxygenometer, which indicated 0 mmHg after 1 min of oxygen deprivation. Production of lactate was measured after complete deprivation of oxygen for 2, 5, 10, 15 and 30 mm. As indicated by the formation of lactate, the anaerobic metabolic rate was fastest in glucose-containing medium, followed by mannose and slowest in the fructose medium (Fig. 4). These results are consistent with the order of the decay and recovery of PS in mannose and fructose-containing medium (Fig. 2-A1, A2).
3.4. Effect of lactate on the preser6ation of PS in slices exposed to longer periods of ischemia We have reported previously, using hippocampal slices, that the presence of glucose is required to main-
tain neural activity and that, as confirmed here, the deprivation of glucose blocks neural activity (Okada, 1990, 1992; Saitoh et al. 1994; Kanatani et al., 1995). However, Schurr et al. (1988, 1997a,b,c) reported that lactate, an obligatory aerobic energy substrate, could support neural function. In their study, hippocampal slices were prepared using a Mcllwain tissue chopper. Although the tissue is immersed in the ice cold standard medium, this method requires at least a 10–20 mm anoxic or ischemic period before preparing the hippocampal slices (slowly prepared slices, see Fig. 7B) because the hippocampal tissue block is removed from the brain and fixed on the cutting table before slicing the tissue. In our studies, hippocampal slices (rapidly prepared slices, see Fig. 7A,C) are prepared within 1 mm after removal of the brain from the skull. We prepare slices from the hippocampus tissue block with a razor blade by free hand and immediately incubate them in the glucose containing standard medium for 20 mm. In our technique, slices of similar thickness (300– 400 mm) could be prepared easily and the slice tissue is exposed to ischemic condition for a maximum of l mm after removal of brain from the skull.
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
167
Fig. 3. Effect of glucose (A), mannose (B), fructose (C) and lactate (D) application on the PS amplitude after cessation of PS during glucose deprivation. Vertical bars of each plot indicate SEM (n=4–5).
To investigate the biochemical and functional difference between these two preparations, we prepared hippocampal slices under the same ischemic conditions as the studies using vibratome technique or tissue chopper method. We left the hippocampus tissue block on an ice-cold table for 20 mm immersed in ice cold standard medium and cut the tissue into slices using a razor blade by free hand. Prior to incubating the slices in the medium containing 10 mM glucose and saturated with oxygen, the ATP levels of slices prepared by either our method (rapid preparation) or by the vibratome technique (slow preparation) were 4.44 9 0.85 mmol/kg protein (Fig. 5A, n =5) and 1.759 0.23 mmol/kg protein (Fig. 5B, n =4), respectively. Incubating the slices in medium containing oxygen and glucose for 20 mm, increased the ATP levels in the slices rapidly prepared by our method to 12.339 0.49 mmol/kg protein whereas that slowly prepared by tissue chopper or vibratome method increased to 6.549 1.26 mmol/kg protein (Fig. 5). This inadequate recovery of the highenergy phosphate levels after a longer period of ischemia confirmed the results of our previous study in which a longer exposure of brain tissue to ischemia greatly reduced the recovery of high-energy phosphates (Mori et al., 1992). The synaptic activity of slices prepared slowly, as in the vibratome or slice chopper method, was tested using medium in which glucose was replaced with lactate. Interestingly the amplitude of PS did not disappear even for 100 mm in lactate-containing medium, although the amplitude was transiently reduced to 85% of
original amplitude for the first 20 mm (Fig. 6-B1). Slices prepared by free hand immediately after removal of hippocampal tissue block from the brain showed the transient complete loss of the PS as indicated in Fig. 6-A1. After complete recovery of the PS amplitude in the lactate medium, the lactate was replaced with glu-
Fig. 4. Effect of glucose (n =6), mannose (n = 4) and fructose (n= 4) on the formation of lactate during complete anoxia. Vertical bars of each plot indicate SEM.
168
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
ATP and CrP were lower in slices prepared slowly as with the chopper method.
4. Discussion
Fig. 5. Recovery of ATP and CrP in rapidly prepared slices (A) (n=5) and slowly prepared slices (B) (n= 4) during preincubation in the standard medium. In A, slices were prepared by razor blade immediately after removal of tissue block of hippocampus from brain. In B, they were prepared after putting hippocampal tissue block once on the ice cold cutting table for 20 mm (see text). After preparation of slices, they were incubated in the standard medium for 5, 10 and 20 mm. Vertical bars of each plot indicate SEM.
cose for 30 mm, and then the glucose was replaced with lactate again. In this case, the PS amplitude was not reduced and did not transiently disappear as observed in the first episode of lactate addition. In addition, the neural activity was stable, even after 100 mm. In slices prepared by both methods, the levels of ATP and CrP were maintained at the original levels, after preincubation of slices, even after replacement of the glucose with lactate (Fig. 6A2, B2). However, the initial levels of
In this study, we examined the correlation between neural activity and the levels of high energy phosphates such as ATP and CrP in hippocampal slices exposed to medium containing either glucose, mannose, fructose, pyruvate or lactate as an energy source. There have been many reports demonstrating the relationship between energy metabolism and neural activity in brain tissue (Yamamoto and Kurokawa, 1970; Okada and McDougal, 1971; Cox and Bachelard, 1982, 1988; Lipton and Whittingham, 1982; Bachelard et al., 1984; Cox et al., 1985; Okada, 1992). Yamamoto and Kurokawa (1970) reported that synaptic potentials were completely blocked when the ATP level was reduced by 20% of its original concentration in olfactory cortical slices. Lipton and Whittingham (1982) showed that, in hippocampal slices, ATP was decreased by 15% when the evoked synaptic response began to diminish 2 mm after exposure to hypoxia. Additionally, Okada (1982) found a discrepancy between the time-course in the reduction of ATP levels and the decrease in the PS amplitude during either oxygen or glucose deprivation. The PS amplitude decayed much faster during glucose deprivation than during hypoxia, even though the level of ATP was reduced in the similar time-course. Further, the addition of either lactate or pyruvate instead of
Fig. 6. Effect of lactate on (1) the PS and on (2) the levels of ATP (closed circles) and CrP (open circles) of slowly prepared slices (Column B) and the effect of repetitive application of lactate on the PS and levels of ATP and CrP of rapidly prepared slices (Column A). Note that in B, replacement of glucose with lactate did not cause the reduction of the PS amplitude and in A, the second addition of lactate caused no reduction of the PS amplitude. Vertical bars of each plot indicate SEM (n = 4 – 6).
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
169
Fig. 7. Schematic drawing showing the effect of lactate on the neural activity (PS) of rapidly prepared (A) and slowly prepared slices (B) and the effect of repetitive application of lactate on the PS of rapidly prepared slices (C). Each experimental procedure for slice preparation is shown in the left and the change of the PS amplitude during application of lactate instead of glucose (G) is shown in the right.
glucose maintained the original levels of ATP and CrP, although neural activity was diminished with similar time-course as in glucose deprivation. In a similar observation, Cox and Bachelard (1988) reported an attenuation of the PS amplitude during the incubation of hippocampal slices in lactate-containing medium. Wada et al. (1998) confirmed the importance of glucose in the preservation of PS and found that neural activity, such as synaptic transmission, differed in the requirement for glucose in the granule cells and in the pyramidal cells of CA3 area. This was observed despite the fact that the levels of high-energy phosphates in these areas were not significantly different during glucose deprivation. These results strongly suggested that the presence of glucose is essential in the maintenance of neural activity, regardless of the ATP level in the brain slices. In contrast to the reports described above, Schurr et al. (1997a) and Fowler (1993) reported that lactate, instead of glucose, can support synaptic transmission in CA1 neurons of rat hippocampal slices. However, they did not determine the ATP concentration in the slices. Using an intracellular recording technique for single CA3 pyramidal neurons in hippocampal slices, Takata and Okada (1995) found that lactate could not support synaptic transmission and induced membrane depolarization, although the ATP levels were not reduced in these slices. As the time-course of decay for the synaptic potential during lactate replacement is very similar to that during glucose deprivation, ATP derived from
the oxidative pathway may not support synaptic transmission and may not be able to maintain the resting potential in CA3 neurons. Thus, our results do not agree with the observations of Schurr et al. (1997a) and Fowler (1993). In connection with the functional significance of the glycolytic pathway, Rose and O’Connell (1964) and Mercer and Dunham (1981) reported that, even in erythrocytes, glycolytic enzymes bind to the cell membrane and membrane-bound ATP fuels the Na/K pump that maintains the resting membrane potential. If this is true for neurons, the ATP produced through anaerobic and aerobic pathways may have different functional significance. It is interesting to note the phenomenon observed by Saitoh et al. (1994) that slices maintained for a longer period in lactate-containing medium recovered their neural activity which was initially extinguished. The presence of lactate, together with glucose prior to glucose deprivation, may result in the maintenance of the PS amplitude. To test this possibility, Saitoh et al. (1994) added lactate to the medium for 30 min before glucose deprivation. The PS amplitude decayed, was transiently lost, and reappeared although the levels of ATP and CrP were maintained at the initial concentrations (Saitoh et al., 1994). In the present experiment, we confirm these observations and find that the addition of either mannose or fructose instead of glucose resulted in the transient reduction of the PS amplitude followed by complete recovery. In addition, the PS was not completely extinguished as observed in the lactate
170
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171
medium. Further, we examined neural activity and determined the levels of high energy phosphates during the application of either mannose, fructose, or lactate instead of glucose. Fructose and mannose have been reported to act as a substrate for ATP production in brain tissue (Sapolsky, 1986). Mcllwain and Bachelard (1985), measuring oxygen consumption of cortical slices, showed that either sugars, such as galactose, fructose, and mannose, or metabolites, such as pyruvate and lactate, can also be used as substrates for energy metabolism in isolated cortical slice tissues, although the rates of utilization of lactate, fructose and mannose were much lower than that of glucose. Our results suggest that glucose, mannose, and fructose are able to maintain a steady anaerobic metabolic rate although, by measuring the rate of lactate formation, these substrates differ in the metabolic rate they support. These results, together with the effects of those substrates on the neural activity, indicate that the rate of the non-oxidative pathway may be crucial to the maintenance and preservation of neural activity. Why do long periods of slice incubation in lactatecontaining medium result in the recovery of neural activity that was once completely extinguished? Our results reveal that, in slices perfused with lactate medium repetitively, the second episode of perfusion with lactate is accompanied by well preserved PS. This observation indicates that once brain tissue is exposed to either a hypoglycemic or ischemic episode, lactate begins to be utilized to preserve neural activity. During the preparation of slices from a brain tissue block, the hippocampal slices prepared by our method are exposed to ischemic period for only 1 mm, whereas methods that are more frequently used, such as either the vibratome or chopper method, require a 10 – 20 mm preparatory period before the slices are made. Therefore, it is inevitable that the slices are exposed to longer period of ischemic conditions. The differences in the experimental procedures between our method and others in regards to the effect of lactate addition are schematized in Fig. 7. These procedural differences may explain the discrepancy between our results and the results of Schurr et al. (1997a). Schurr and Fowler performed the experiment using rat preparation and recorded the electrical activity from CA1 region of hippocampus. For this reason, we tried the same experimental procedure of slow and rapid preparation of slices for rat (n=6) and also recording from CA1 region of guinea pig hippocampus (n = 5). We obtained the similar results as shown in the present experiment. Thus our rapid slice preparation, with a shorter ischemic period, may reflect a more physiological (in situ) state of brain tissue, with a faster and better recovery of high energy phosphates. Concerning the importance of glycolysis during physiological activity there have been several reports show-
ing that physiological stimulation increases specifically nonoxidative glucose metabolism in the brain of the freely moving rat (Fellows et al., 1993) and that in the human PET study, transient increases in neural activity cause a tissue uptake of glucose in excess of that consumed by oxidative metabolism (Fox et al., 1988). Thus our results together with others suggest that anaerobic glycolysis is crucial for neural activity in the physiological state. Under well supplied conditions, glucose may be the primary substrate for the production of ATP by cytoplasmic enzymes that is used in the maintenance of neural activity. Mitochondrial ATP may become important for neural function only after hypoglycemic episodes, although 20–30 mm may be required to switch from anaerobic to aerobic mechanisms. The switching mechanisms to utilize ATP produced from either an anaerobic or an aerobic pathway remains to be studied further. Finally, the present study suggests that the interpretation of experimental results from brain slices must consider the method of slice preparation because the tissue may be exposed to varying ischemic conditions dependent upon the methods used.
Acknowledgements This paper was supported by the grant from the Minister of Welfare, Japan.
References Bachelard HS, Cox DWG, Drower J. Sensitivity of guinea-pig hippocampal granule cell field potentials to hexoses in vitro: an effect on cell excitability? J Physiol 1984;352:91 – 102. Cox DWG, Bachelard HS. Attenuation of evoked field potentials from dentate granule cell by low glucose, pyruvate + malate, and sodium fluoride. Brain Res 1982;239:527 – 34. Cox DWG, Bachelard HS. Partial attenuation of dentate granule cell evoked activity by the alternative substrates, lactate and pyruvate: evidence for a postsynaptic action. Exp Brain Res 1988;69:368– 72. Cox DWG, Drower J, Bachelard HS. Effect of metabolic inhibitors on evoked activity and the energy state of hippocampal slices superfused in vitro. Exp Brain Res 1985;57:464 – 70. Fellows LK, Boutelle MG, Fillenz M. Physiological stimulation increases nonoxidative glucose metabolism in the brain of the freely moving rat. J Neurochem 1993;60:1258 – 63. Fowler JC. Glucose deprivation results in a lactate preventable increase in adenosine and depression of synaptic transmission in rat hippocampal slices. J Neurochem 1993;60:572 – 6. Fox PT, Raictile ME, Mintun MA, Dence C. Nonoxidative glucose consumption during focal physiologic neural activity. Science 1988;241:462 – 4. Kanatani T, Mizuno K, Okada Y. Effects of glycolytic metabolites on preservation of high energy phosphate level and synaptic transmission in the granule cells of guinea pig hippocampal slices. Experimentia 1995;51:213 – 6.
K. Yamane et al. / Journal of Neuroscience Methods 103 (2000) 163–171 Lipton P, Whittingham TS. Reduced ATP concentration as a basis for synaptic transmission failure during hypoxia in the in vitro guinea-pig hippocampus. J Physiol 1982;325:51–65. Lowry OH, Passonneau JY, Hasselberger FX, Schulz DW. Effect of ischemia on known substrates and cofactors of the glycolytic pathway in brain. J Biol Chem 1964;239:18–30. Mcllwain H, Bachelard HS. Metabolic, ionic and electrical phenomena in separated cerebral tissue. Biochemistry and the Central Nervous System, 5th edn. Edinburgh: Livingstone, 1985:54 – 83. Mercer RW, Dunham PB. Membrane-bound ATP fuels the Na/K pump: studies on membrane-bound glycolytic enzymes on insideout vesicle from human red cells membranes. J Gen Physiol 1981;78:547 – 68. Mori M, Nishizaki T, Okada Y. Protective effect of adenosine on the anoxic damage of hippocampal slice Neuroscience 1992;46:301 – 7. Okada Y. Reversibility of neuronal function of hippocampal slices during deprivation of oxygen and/or glucose. In: Somjen GG, editor. Mechanism of Cerebral Hypoxia and Stroke. New York: Plenum, 1982:191 – 203. Okada Y. Energy metabolism and neural activity in brain slice-effect of deprivation of oxygen and glucose. Protein, Nucleic Acid and Enzyme (Jpn) 1990;35:661–74. Okada Y. Energy metabolism and neural activity of brain slices importance of glucose for the maintenance of neural activity. J Brain Med Japanese 1992;4:39–49. Okada Y, McDougal DB. Physiological arid biochemical change in frog sciatic nerve during anoxia and recovery. J Neurochem 1971;18:2335 – 53. Rose IA, O’Connell EL. The role of glucose 6-phosphate in the regulation of glucose metabolism in human erythrocytes. J Biol Chem 1964;230:12 – 7.
.
171
Saitoh M, Okada Y, Nabetani M. Effect of mannose, fructose and lactate on the preservation of synaptic potentials in hippocampal slices. Neurosci Lett 1994;171:125 – 8. Sapolsky RM. Glucocorticoid toxicity in the hippocampus: reversal by supplementation with brain fuels. J Neurosci 1986;6:2240– 4. Schurr A, Payne RS, Miller JJ, Rigor BM. Brain lactate, not glucose, fuels the recovery of synaptic function from hypoxia upon reoxygenation: an in vitro study. Brain Res 1997a;744:105 – 11. Schurr A, Payne RS, Miller JJ, Rigor BM. Glia are the main source of lactate utilized by neurons for recovery of function posthypoxia. Brain Res 1997b;774:221 – 4. Schurr A, Payne RS, Miller JJ, Rigor BM. Brain lactate is an obligatory aerobic energy substrate for functional recovery after hypoxia: further in vitro validation. J Neurochem 1997c;69:423– 6. Schurr A, West CA, Rigor BM. Lactate-supported synaptic function in the rat hippocampal slices preparation. Science 1988;240:1326– 8. Takata T, Okada Y. Effects of deprivation of oxygen or glucose on the neural activity in the guinea pig hippocampal slice-intracellular recording study of pyramidal neurons. Brain Res 1995;683:109 – 16. Wada H, Okada Y, Uzuo H, Nakamura H. The effects of glucose, mannose, fructose and lactate on the preservation of neural activity in the hippocampal slices from the guinea pig. Brain Res 1998;788:144 – 50. Yamamoto C, Kurokawa M. Synaptic potentials recorded in brain slices and their modification by changes in the level of tissue ATP. Exp Brain Res 1970;10:159 – 70.